Corrosion and Electrochemistry of Zinc
Corrosion and Electrochemistry
of Zinc Xiaoge Gregory Zhang Cominco Ltd. Product Technology Centre Mississauga, Ontario, Canada
Springer Science+ Business Media, LLC
Library of Congress Cataloging-in-Pub1ication
Data
Zhang, Xiaoge Gregory. C o r r o s i o n a n d e l e c t r o c h e m i s t r y of z i n c / X i a o g e G r e g o r y p. cm. Includes bibliographical references a n d index. I S B N 978-1-4757-9879-1 1. Z i n c — C o r r o s i o n , 2. E l e c t r o c h e m i s t r y . I. T i t l e . TA480.Z6Z45 1996 620. 1'84223—dc20
ISBN 978-1-4757-9879-1 DOI 10.1007/978-1-4757-9877-7
Zhang,
96-32551 CIP
ISBN 978-1-4757-9877-7 (eBook)
© 1996 Springer Science+Business Media New York Originally published by Plenum Press, New York in 1996 Softcover reprint of the hardcovrer 1st edition 1996
All rights reserved 10 9 8 7 6 5 4 3 2 1 No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording, or otherwise, without written permission from the Publisher
To my mother Youliu Zhang and father Hongtao Zhang
Foreword Humankind's use of zinc stretches back to antiquity, and it was a component in some of the earliest known alloy systems. Even though metallic zinc was not "discovered" in Europe until 1746 (by Marggral), zinc ores were used for making brass in biblical times, and an 87% zinc alloy was found in prehistoric ruins in Transylvania. Also, zinc (the metal) was produced in quantity in India as far back as the thirteenth century, well before it was recognized as being a separate element. The uses of zinc are manifold, ranging from galvanizing to die castings to electronics. It is a preferred anode material in high-energy-density batteries (e.g., Ni/Zn, Ag/Zn, ZnJair), so that its electrochemistry, particularly in alkaline media, has been extensively explored. In the passive state, zinc is photoelectrochemically active, with the passive film displaying n-type characteristics. For the same reason that zinc is considered to be an excellent battery anode, it has found extensive use as a sacrificial anode for the protection of ships and pipelines from corrosion. Indeed, aside from zinc's well-known attributes as an alloying element, its widespread use is principally due to its electrochemical properties, which include a well-placed position in the galvanic series for protecting iron and steel in natural aqueous environments and its reversible dissolution behavior in alkaline solutions. Dr. Zhang has undertaken the monumental task of describing the corrosion properties and electrochemistry of zinc in a single book. The reason why this task is "monumental" is that the literature on this important metal is highly fragmented, no doubt reflecting zinc's diversity of use. Furthermore, the literature stretches from the very fundamental to the very applied, with some of the reports on the properties of the metal and its compounds being anecdotal in nature. The task of assembling all of the relevant information into a single monograph, in a manner that is logical and easy to read. is the task that Dr. Zhang undertook. He has succeeded admirably, and this book will surely become an authoritative source of information on the electrochemistry of this technologically important metal. Digby D. Macdonald The Pennsylvania State University University Park, Pennsylvania
vii
Preface Zinc is one of the most widely used metals. Its most important commercial application is corrosion protection of steel. In the past decades, a tremendous amount of research work has been done on the various aspects of zinc corrosion and electrochemistry. This book provides a systematic review of the enormous volume of technical results generated from these investigations. It is hoped that it will not only be useful to those interested in specific information on this subject but also will stimulate those currently working in the field to carry out further research. This book attempts to combine fundamental information on the electrochemistry of zinc with practical corrosion data for zinc and its alloys and to connect the academic and industrial realms of interest, which are often detached from each other. In general, books on corrosion written from an academic perspective usually treat individual metals or alloys either as examples or in a rather general fashion, while those issuing from the metals industries cover little fundamental information. However, from the viewpoint of a metals user or researcher, it is most beneficial that all the relevant corrosion and electrochemistry information, theoretical and practical, on one metal be systematically organized in one single source. Additionally, only a dozen or so metals, including zinc, are used in massive quantities in today's society, and a compilation of all the corrosion, electrochemistry, and related information in a single book for each of them would be very useful for more effective application of these metals in the future. This book focuses on corrosion and does not cover other applied aspects of zinc electrochemistry. However, as it contains a large collection of electrochemical information on zinc, it can also serve as a source of reference for electrochemical processes such as electroplating, electrowinning, and batteries. Much of the electrochemical information presented in this book is related to the elemental reactions such as dissolution, hydrogen evolution, oxygen reduction, and passivation, which are also important in electrochemical processes other than corrosion. Two general approaches have been taken in the selection and treatment of the information presented in this book. The first is to emphasize the properties pertaining to zinc as a material, rather than those pertaining to specific products (such as coatings, wires, plates, cast alloys, etc.). The behavior of the various zinc products is taken into account in the consideration of the effect of specific physical or chemical factors, such as alloying elements, physical dimensions, temperature, solution composition, and pH. The ix
x
PREFACE
second is to emphasize the specificity of corrosion data in relation to corrosion environments. The environmental conditions pertaining to each corrosion situation are specified, together with detailed data. Generalizations are provided when a consensus exists in the data. Also, because the environment is as important as the material in a corrosion process, information is provided at the beginning of each chapter to describe the corrosion environments and to define the various factors involved. The book consists of 15 chapters. The first chapter of the book presents the basic physical and chemical properties of zinc. The remainder of the first half of the book is concerned with the electrochemistry of zinc. More specifically, Chapter 2 on Electrochemical Thermodynamics and Kinetics, Chapter 3 on Passivation and Surface Film Formation, and Chapter 4 on Electrochemistry of Zinc Oxide deal with the fundamental electrochemistry of zinc, and Chapter 5 on Corrosion Potential and Corrosion Current, Chapter 6 on Corrosion Products, and Chapter 7 on Corrosion Forms. These last three chapters, in different aspects, connect the fundamental electrochemistry with the practical corrosion behavior of zinc. The remaining chapters deal with corrosion performance in various environments. ACKNOWLEDGMENTS I am profoundly grateful to the management of Cominco Ltd., particularly to Dr. E. M. Valeriote and Mr. S. R. Wilkinson, for their support of my undertaking of such a time-consuming task. In addition, I would like to personally thank Dr. Valeriote, who, as the manager of the Co minco Product Technology Centre, was not only instrumental in initiating this project but was also always keen to assist by providing advice and resources. A very special thanks is due to Professor D. D. Macdonald of The Pennsylvania State University, who suggested that I write this book and kindly helped at various stages during the process. Also, I am deeply indebted to Professor M. Pourbaix of I'Universite Libre de Bruxelles, who has had a great influence on my career and whose work and spirit have inspired me in my writing of this book. It would not have been possible for the book to arrive at its present form without the constructive suggestions and criticisms of many people. I sincerely thank the following people, who have helped in reviewing the various chapters of the manuscript: Dr. T. D. Burleigh, University of Pittsburgh, United States Dr. T. G. Chang, Cominco Ltd, Canada Dr. B. R. Conard, INCO Ltd., Canada Dr. F. E. Goodwin, International Lead and Zinc Research Organization, United States Dr. J. A. Gonzalez, Cominco Ltd., Canada Dr. T. E. Graedel, AT&T Bell Laboratories, United States Professor T. M. Harris, University of Tulsa, United States Professor M. B. Ives, McMaster University, Canada Professor D. W. Kirk, University of Toronto, Canada Mr. G. P. Lewis, Lewis Consulting, Canada Professor C. Leygraf, Royal Institute of Technology, Sweden Dr. J. H. Lindsay, General Motors Research Laboratories, United States Professor D. D. Macdonald, The Pennsylvania State University, United States
xi
PREFACE
Dr. 1. Odnevall, Royal Institute of Technology, Sweden Professor P. Searson, Johns Hopkins University, United States Dr. H. E. Townsend, Bethlehem Steel Co., United States Dr. K. Tomantschger, Cominco Ltd., Canada Dr. E. M. Valeriote, Cominco Ltd., Canada Professor R. Wiart, Universite Pierre et Marie Curie, France In particular, the critical reading of the manuscript and the very helpful suggestions made by Mr. G. P. Lewis of Lewis Consulting and Dr. H. E. Townsend of Bethlehem Steel Co. are greatly appreciated. Many people at the Product Technology Centre of Cominco have helped at different stages in the development of the manuscript. To them I am most grateful. In particular, I would like to thank Mr. J. E. Valeriote, who helped in the preparation of the figures, Mr. J. Hwang, who provided feedback by reading the first draft of the manuscript, Mrs. V. Rodic and Mrs. P. L. Doyle, who helped in obtaining and organizing the references, Mrs. H. Laur for handling correspondence and mailing, and Ms. M. F. Haughton, who assisted in preparing the tables and making manuscript corrections. Also, I would like to acknowledge The Second Cup at 292 Dundas West in Toronto, where I read the literature and revised the manuscript during the morning hours of numerous weekends. The work of Amelia McNamara, Arun Das, Kenneth Howell, and Jacqueline Sedman at Plenum Press is greatly appreciated. Finally, I wish to express my deeply felt thanks to my wife, Li, for her understanding and moral support during the long period required for the writing of this manuscript. Xiaoge Gregory Zhang
Mississauga, Ontario, Canada
Contents LIST OF SYMBOLS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
XIX
CHAPTER 1. Properties, Products, and Processes 1.1. Introduction . . . . . . . 1.2. Basic Properties . . . . . . . 1.2.1. Physical Properties .. 1.2.2. Mechanical Properties 1.2.3. Alloying Properties .. 1.3. Main Products and Applications 1.3.1. Zinc Coatings . 1.3.2. Cast Products . . . . . 1.3.3. Rolled Zinc . . . . . . 1.3.4. Zinc Dust and Powder 1.4. Coating Processes . . . . . 1.4.1. Hot-Dip Galvanizing 1.4.2. Electroplating. 1.3. Phosphating 1.4. Chromating . . . . .
2
3 3 3 7 7 7 7 7 13 15 16
CHAPTER 2. Electrochemical Thermodynamics and Kinetics 2.1. Introduction . . . . . . . 2.2. Thermodynamic Stability 2.3. Ionic Properties . . . . . 2.4. Double-Layer Properties 2.5. Kinetics of Elemental Reactions 2.5.1. Dissolution . . . . . 2.5.2. Deposition . . . . . 2.5.3. Hydrogen Evolution 2.5.4. Oxygen Reduction .
19 19 25 27
29 29 36 39
48 xiii
xiv
CONTENTS
2.6. Corrosion Processes . . . . . . . . . . . . 2.6.1. General Considerations . . . . . . . 2.6.2. Impedance of Corroding Electrodes
54 54 54
CHAPTER 3. Passivation and Surface Film Formation 3.1. Introduction . . . . . . . . . . 3.2. Characteristics and Conditions. 3.3. Alkaline Solutions .. .
3.3.1. i-VCurves .. . 3.3.2. Passivation Time 3.3.3. Characteristics .. 3.3.4. Mechanisms of Formation of Passive Films 3.4. Other Solutions. . . . . . . . . . . . . . . . . . 3.4.1. Slightly Alkaline and Carbonate Solutions 3.4.2. Phosphate Solutions. . . 3.4.3. Miscellaneous Solutions 3.5. Anodization . . . . . . . 3.6. Stability of Passivation . . . . 3.6.1. Type of Passivation .. 3.6.2. Passivation Breakdown
65 65
68 68 70 73 75 77 77 80
84
85 87 87
89
CHAPTER 4. Electrochemistry of Zinc Oxide 4.1. Introduction . . . . . . . 4.2. Basic Properties . . . . . . 4.2.1. Physical Properties . 4.2.2. Electronic Properties 4.3. Semiconductor Electrochemical Behavior. 4.3.1. Basic Theories .. . 4.3.2. Flatband Potential . . . . . . . 4.3.3. Band Structure . . . . . . . . 4.3.4. Electrode Kinetics in the Dark 4.3.5. Photoelectrochemical Kinetics 4.3.6. Electroluminescence 4.4. Thin ZnO Films . . . . . . . . . . . 4.5. Stability . . . . . . . . . . . . . . . 4.5.1. Conditions of Stability and Decomposition Reactions 4.5.2. Rate of Decomposition . . . . . . . . . . . . .
93 93 93 95 97 97 100 103 105 109 114 115 119 119 121
CHAPTER 5. Corrosion Potential and Corrosion Current 5.1. Introduction . . . . . . . . . . . . . . . . . . . . . . 5.2. Relation between Corrosion Potential and Corrosion Current 5.2.1. Polarization Resistance and Corrosion Current. 5.2.2. Conversion Factors . . . . . . . . . . . . . . . . . . .
125 125 127 129
CONTENTS
5.3. Corrosion Potential and Reaction Kinetics 5.4. Ecorr and icorr under Various Conditions 5.4.l. Effect of Zinc Ions . . . . . . 5.4.2. Effect of Anions and Cations. 5.4.3. Effect of pH . . . . . . . . . . 5.4.4. Effect of Temperature . . . . 5.4.5. Effect of Aeration and Convection. 5.4.6. Effect of Surface Condition 5.5. Zinc Alloys . . . . . . . . . . . . . . . . 5.6. Effect of Time . . . . . . . . . . . . . . . 5.7. Correlation between Corrosion Current and Weight Loss Rate
xv
130 133 133 135 137 140 141 143 144 149 153
CHAPTER 6. Corrosion Products 6.l. Introduction . . . . . . . . . 6.2. In Atmospheric Environments .. 6.2.1. Composition and Structure . 6.2.2. Quantity and Morphology 6.2.3. Formation Processes 6.3. In Waters . . . . . . 6.3.l. Fresh Waters 6.3.2. Seawater .. 6.4. In Solutions . . . . 6.4.l. Effect of pH . 6.4.2. Formation Processes 6.4.3. Zinc Alloys . . . . . 6.5. In Other Environments .. 6.6. Effect of Corrosion Products on Zinc Corrosion
157 158 158 163 165 168 168 170 171 171 173 176 176 178
CHAPTER 7. Corrosion Forms 7.1. Introduction . . . . 7.2. Galvanic Corrosion 7.2.1. Introduction. 7.2.2. Theoretical Aspects . 7.2.3. Practical Factors .. 7.2.4. Polarity Reversal .. 7.2.5. Galvanic Corrosion in Natural Environments 7.2.6. Galvanic Protection of Steel by Zinc. 7.3. Pitting Corrosion . . . . . . 7.3.l. Introduction. . . . . . 7.3.2. Occurrence of Pitting. 7.3.3. Pitting Potential 7.3.4. Morphology 7.3.5. Mechanisms ..
· 183 183 183 185 196
.203 .208 · 213 .217 · 217 · 217 · 221 .224
.225
xvi
CONTENTS
7 A. Intergranular Corrosion 704.1. Introduction . . 704.2. Occurrence. . . 704.3. Metallurgical Effects. 70404. Effect of Environmental Factors . 704.5. Effect on Mechanical Properties. 704.6. Mechanisms . . . . . . . . . . . 7.5. Wet Storage Stain . . . . . . . . . . . . 7.6. Hydrogen Embrittlement and Corrosion Cracking
227 227 227 229 232 234 235 236 238
CHAPTER 8. Atmospheric Corrosion 8.1. Introduction . . . . . . 8.2. Atmospheric Factors . . 8.2.1. Type of Wetting 8.2.2. Air Pollutants . 8.3. Corrosion in Outdoor Environments. 8.3.1. Typical Corrosion Rates . 8.3.2. Effect of Time of Wetness . . 8.3.3. Effect of Pollutants . . . . . 8.3.4. Effect of Elevation and Distance from Seawater 8.3.5. Effect of Initial Weather Conditions. 8.3.6. Effect of Climate . . . . . . . . 8.3.7. Effect of Sample Configuration 8.3.8. Effect of Sheltering 8.3.9. Galvanized Steel . . . . . . 8.3.10. Effect of Alloying . . . . . 8.3.11. Effect of Surface Treatment 8.3.12. Effect of Corrosion Products 8.3.13. Forms of Corrosion . . . . 8.3.14. Highway Environment . . . . 804. Corrosion in Indoor Environments . 8.5. Corrosion in Simulated Environments. 8.5.1. Humidity Chamber Exposure 8.5.2. Water and Salt Spray .. 8.5.3. Cyclic Test . . . . . . . 8.504. Thin-Layer Electrolytes 8.6. Corrosion Mechanisms . . . .
241 241 241 243 245 245 248 249 252 252 254 254 255 256 258 260 261 261 262 264 266 267 270 272
274 278
CHAPTER 9. Corrosion in Waters and Aqueous Solutions 9.1. Introduction . . . . . . . 9.2. Characteristics of Waters . 9.2.1. Fresh Waters . . . 9.2.2. Seawater . . . . . . 9.3. Corrosion in Pure Water . 9.4. Corrosion in Natural Waters
283 283 283 284 286 288
CONTENTS
9.4.1. Cold Fresh Water. 9.4.2. Hot Fresh Water . 9.4.3. Seawater . . . . . 9.5. Corrosion in Aqueous Solutions 9.5.1. Effect of Dissolved Species 9.5.2. Effect of pH . . . . . . . . . 9.5.3. Effect of Immersion Conditions 9.5.4. Effect of Surface Treatments .. 9.5.5. Effect of Metallurgical Factors.
xvii
.288 .289 · 291 .296 .296 .298 · 301 .302 .302
CHAPTER 10. Corrosion in Soil 10.1. Introduction . . . . . 10.2. Characteristics of Soil 10.3. Corrosion Rates . . . 10.3.1. Effect of Soil Factors 10.3.2. Galvanic Corrosion . 10.4. Electrochemical Measurements
.305 .305 .308 .308 · 312 · 312
CHAPTER 11. Under-Paint Corrosion 11.1. Introduction . . . . . . . . . 11.2. Basic Characteristics of Paint 11.2.1. Components in Paint . 11.2.2. Barrier Properties of Paint . 11.3. Corrosion Tests. . . . . . . . . . . 11.4. Corrosion Behavior . . . . . . . . 11.4.1. Characterization of Corrosion . 11.4.2. Effect of Coating Type. . 11.4.3. Effect of Test Conditions 11.4.4. Effect of Paint System . 11.4.5. Galvanic Action 11.5. Corrosion Mechanisms . . . .
· 315 · 315 · 316 · 316 · 317 · 319 .319 .321 .325 .328 .332 · 333
CHAPTER 12. Zinc-Rich Coatings 12.1. Introduction . . . . . . 12.2. Coating Characteristics 12.3. Protection Mechanism. 12.4. Performance . . . . . . 12.4.1. Effect of Zinc Content . 12.4.2. Effect of Zinc Particle Size 12.4.3. Effect of Binders . . . . . . 12.4.4. Effect of Coating Thickness . 12.4.5. Effect of Additives . . . . . 12.4.6. Effect of Surface Condition 12.4.7. Other Factors . . . . . . . .
.337 .337 .339 · 341 · 341 .343 .344
.345 .345 .347 .348
xviii
CONTENTS
CHAPTER 13. Corrosion in Concrete 13.1. Introduction . . . . . . . . . . 13.2. Concrete Environment . . . . 13.2.1. Formation of Concrete 13.2.2. Characteristics . . . . l3.3. Corrosion of Steel Reinforcement in Concrete. 13.3.1. Effect of Corrosion . l3.3.2. Protection Methods . . . . . . . . . 13.3.3. Galvanized Coatings . . . . . . . . 13.4. Corrosion of Galvanized Steel in Concrete 13.4.1. Testing Methods . . . . . . . 13.4.2. Field Test Results . . . . . . . l3.4.3. Results from Simulated Tests
351 352 352 353 358 358 359 359 360 360 360 365
CHAPTER 14. Corrosion in Batteries 14.1. Introduction . . . . . . . 14.2. Zinc Cells and Batteries 14.2.1. Leclanche Cell . 14.2.2. Zinc Chloride Cell 14.2.3. Zinc Alkaline Cell 14.2.4. Zinc-Air Battery . 14.2.5. Zinc-Nickel Battery 14.3. Corrosion . . . . . . . . . . 14.3.1. Effect of Testing Time 14.3.2. Effect of Electrolyte . 14.3.3. Effect of Chemical Agents 14.3.4. Zinc Electrode . . . . 14.3.5. Operating Conditions .. .
373 373 373 375 375 376 377 377 379 379 382 385 390
CHAPTER 15. Corrosion in Other Environments 15.1. Introduction . . . . . 15.2. Organic Solvents .. 15.2.1. Classification 15.2.2. Corrosion .. 15.3. Gaseous Environments 15.4. Zinc Anodes . . . . . . 15.4.1. Sacrificial Anodes 15.4.2. Anodes for Impressed Current Cathodic Protection
393 393 393 395 399 403 403 407
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 409 INDEX
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463
List of Symbols Symbol A aH+' aOH-
B b
C Cd C,c CH Co C, Ccrit
D d
E Eo EII2 E" E" Ec Ecorr EF ED Etb
E~ Egc
E" E" E,
e-, p+ e
F hv fa
Definition Surface area Activity of hydrogen and hydroxyl ions Stern-Geary constant Tafel slope Capacitance Capacitance of double layer Capacitance of space charge layer Capacitance of Helmholtz double layer Bulk concentration Surface concentration Critical concentration for passivation Diffusion coefficient Distance between anode and cathode Electrode potential Standard potential Half-wave potential Potential of anode Breakdown potential Lower edge of conduction band Potential of cathode Corrosion potential Fermi level Decomposition potential Flatband potential Width of band gap Potential of a galvanic couple Passivation potential Top edge of valence band Solution potential Electron and hole Electron charge Faraday constant Photon Anodic current xix
Section
5.2 2.5 5.2 2.5 2.4 2.6 4.3 4.3 3.3 3.3 3.3 3.3 7.2 2.5 2.2 2.5 7.2 3.2 4.2 7.2 5.2 4.2 4.5 4.3 4.2 7.2 3.2 4.2 4.3 4.3 4.3 2.5 4.5 7.2
xx
LIST OF SYMBOLS
Symbol
Ie Ig
)~ i, io
--7 I
iOa' iOe
i a , ie icorr
ig
i gt im
i" iphoto
Kf L L,
L"
M
n ox ,
nred
n" Ps PH"PO,
Q; -
R
t+, L VH
Vrd
V,
VSCE V SHE
Definition Cathodic current Galvanic current Current density Oxidation and reduction current density of a redox reaction Exchange c'urrent density Exchange current densities for anodic and cathodic reactions Anodic and cathodic current density Corrosion current density Galvanic current density Gravimetric corrosion rate Limiting current density Passivation current Photocurrent density Rate constant for forward reaction Polarization parameter Width of space charge layer Hole diffusion length Molar concentration per liter Molar mass weight Effective density of states in conduction band Dopant concentration Bulk electron and hole density Charge per atom Number of oxidizing and reducing species Surface electron and hole density Partial pressure of hydrogen and oxygen Immobile charge of space charge layer Resistance Gas constant Corrosion rate Charge-transfer resistance Electrolyte resistance Capacitance of corrosion product film Metallic resistance Polarization resistance Capacitance of charge transfer Potential scanning rate Passivation time Temperature Thickness loss Time Transport number Potential drop across Helmholtz double layer Rest dark potential Potential drop across space charge layer Potential versus saturated calomel electrode Potential versus standard hydrogen electrode
Section
7.2 7.2 2.5 5.2 2.5 7.2 3.4
5.2 7.2 5.7
2.5 3.2
4.3
2.5 7.2
4.3 4.3
2.5 5.2
4.3 4.3 4.3
5.2
4.3 4.3
2.5
4.3
7.2 2.5 8.3
2.6 2.6 2.6 7.2 5.2 2.6 3.4 3.3
2.5 5.2 3.3 2.3
4.3 4.3 4.3
xxi
LIST OF SYMBOLS S~mbol
V CSE AVmAVe AVd W
X Z
Zw
a
fJ e
eo 'I 'la' 'Ie 'II' ()
K
A+"L v p (J
¢ w AC CCT ohp mpy PZC RH r.d.s. rpm SCE SHE SST
Definition Potential versus copper sulfate electrode Potential drops across anode and cathode Potential drop between anode and cathode Weight loss Width of steel cathode Galvanic protection width Electrochemical impedance Warburgimpedance Charge-transfer coefficient Photo absorption coefficient Tafel slope Relative permittivity Permittivity of vacuum Overpotential Anodic and cathodic overpotential Passivation overpotential Area factor Rate constant Equivalent conductance Stoichiometric coefficient Density Specific resistivity Cross section Potential difference between electrode and solution Electrode rotation speed Frequency of AC signal Alternating current Cyclic corrosion test Outer Helmholtz double layer Mils per year Potential of zero charge Relative humidity Rate-determining step Number of rotations per minute Potential of saturated calomel electrode Standard potential of hydrogen electrode Salt spray test
Section
7.2 7.2 5.2 7.2 7.2 2.6 2.6 2.5 4.3 5.2 4.3 4.3 2.5 7.2 3.2 7.2 4.3 2.3 2.2 5.2 7.2 4.3 2.5 2.5 2.6 2.6
11.3
4.3 2.4 8.3 2.5 2.5 2.5 2.5
11.3
1 Properties, Products, and Processes 1.1. INTRODUCTION Zinc is 23rd among the elements in relative abundance in the earth's crust, amounting to 0.013%, compared with aluminum's 8.13% and iron's 5.0%. However, it ranks fourth among the metals in worldwide production and consumption, behind only iron, aluminum, and copper [527]. The uses of zinc can be divided into six major categories: (a) coatings, (b) casting alloys, Cc) alloying element in brass and other alloys, Cd) wrought zinc alloys, (e) zinc oxide, and (f) zinc chemicals. The use of zinc coatings for corrosion protection of steel structures is the most important application owing to the high corrosion resistance of zinc in atmospheric and other environments. Nearly half of the zinc produced is used for this purpose. The position of zinc in the electromotive series of metals means that zinc coating provides not only a barrier layer to prevent contact between the coated steel and the environment but also a sacrificial protection if discontinuities in the coating occur. This chapter provides background information on the physical and metallurgical properties of zinc, its main products and applications, and the processes that are important to its corrosion behavior. The treatment of these subjects in this chapter is rather general: more detailed and specific information will be presented in subsequent chapters. 1.2. BASIC PROPERTIES
1.2.1. Physical Properties Zinc is a silvery blue-gray metal with a relatively low melting point (4 19'soC) and boiling point (907°C). The physical properties of zinc are shown in Table 1.1 [218,1295]. Zinc crystals have a close-packed hexagonal structure. The lattice constants a and c are 0.2664 and 0.4947 nm, respectively. The axial ratio cia is 1.856, which is considerably greater than the theoretical value of 1.633 for the system. Although each zinc atom has 12 near neighbors, 6 are at a distance of 0.2664 nm and the other 6 are at 0.2907 nm. Thus, the bonds between the atoms in the hexagonal basal layers are appreciably stronger
CHAPTER I
2
TABLE 1.1. Physical Properties of Zinc" Atomic number Atomic weight Density Solid,20oe Solid,419se Liquid, 419.5°e Velocity of sound, 20°C Melting point Boiling point, I atm Ionization potentials First Second Third Heat of fusion, 419.5°e Heat of vaporization, 907°C Heat capacity Solid,25°e Liquid Resistivity Solid,20oe Liquid,419.7°e Thermal conductivity Solid, 18°C Solid,419.SOe Liquid, 419.5°e Linear coefficient of thermal expansion Polycrystalline a axis c axis Volume coefficient of thermal expansion Surface tension, liquid, 419.5°e Viscosity, liquid, 419Se
30 65.38 7.14 g/cm 3 6.83 g/cm 3 6.62 g/cm 3 3.67 kmls
419.5°e 907°C 9.39 eV 17.87 eV 40.0eV 7.28 kllmol 114.7 kllmol 25.4 llmol 31.4 llmol 5.96IlQ·cm 37.4/lQ·cm 113 W/(m·K) 96 W/(m-K) 61 W/(m-K)
39.7 x 10-6 K- 1 14.3 X 10-6 K- 1 60.8 X 10-6 K- 1 0.9 X 10-6 K- 1 782 mN/m 3.85 mNlm
"Refs. 218 and 50 I.
than those between the layers. This accounts for much of the deformation behavior and anisotropy of the zinc crystal [218]. The grain structure in a polycrystalline zinc product has preferred orientations depending on the casting and mechanical working conditions: for cast products, the (0001) direction is perpendicular to the axis of the cast columnar crystals; for wire, the (0001) plane is parallel to the axis of drawn wire; and, for sheet, the (0001) plane is parallel to the rolling plane and the (1120) direction is parallel to the rolling direction for sheet rolled at 20 0 e [527]. 1.2.2. Mechanical Properties The strength and hardness of unalloyed zinc are greater than those of tin or lead but appreciably less than those of aluminum or copper. The pure metal cannot be used in stressed applications because of its low creep resistance. Except when very pure, zinc is brittle at ordinary temperatures, but it is ductile at about lOOoe [501]. Since the distance between atoms in the basal plane is shorter than that between atoms in adjoining layers, bonding between basal planes is relatively weak, and, under
PROPERTIES, PRODUCTS, AND PROCESSES
3
stress, the lattice tends to first slip along this plane. At higher temperature, slip may also occur along the (1010) plane. Another major deformation mode of zinc crystal is twinnmg, which tends to occur along one of the (1012) pyramidal planes [2181. Pure zinc recrystallizes rapidly after deformation at room temperature because of the high mobility ofthe atoms within the lattice. Thus, zinc cannot be work-hardened at room temperature. Zinc has low resistance to creep due to grain boundary migration. The temperature for recrystallization and the creep resistance can be increased through alloying [218]. Superplasticity, with an extension of up to 1000%, can be obtained for Zn-AI eutectoid alloys of very fine grain size, on the order of I ,um, at temperatures of 200-2700C [218]. The deformation under this condition appears to take place by slip of the small grains over each other, with little distortion of the grains.
1.2.3. Alloying Properties The binary zinc alloy systems of most interest for commercial applications are (1) Zn-AI, which at 4% AI forms the basis of the zinc die-casting alloys, (2) Zn-Cu, which with up to 45% zinc are brass alloys; (3) Zn-Fe, which includes the phases making up the galvanized coatings, and (4) Zn-Pb, which plays an important role in some pyrometallurgical extraction processes [218]. Ternary and quaternary systems, involving the above alloys with addition of such elements as Ni, Mg, Ti, and Cd, are also of commercial importance. Figure 1.1 is the phase diagram for the Zn-AI alloy system. While the solubility of zinc in aluminum is reasonably high, that of aluminum in zinc is rather limited. This system has a eutectic composition at about 5% AI and a eutectoid composition at 22% AI. Many commercially important alloys, such as the standard die-casting alloys and the coating alloy Galfan, have been developed at or near the eutectic composition. The phase diagram of the Zn-Fe system is illustrated in Fig. 1.2. Iron has very little solid solubility in zinc. When the amount of iron in zinc is above 0.001%, its presence can already be detected micrographically by the appearance of an intermetallic phase, possibly FeZn 7 • In a typical hot-dip galvanizing process, a number ofZn-Fe intermetallic compounds can be formed. The relative amounts and metallographic morphology of these compounds depend on the substrate steel and the hot-dipping conditions, particularly on the Al content in the bath. The zinc corner of the phase diagram of the Zn-Fe-AI ternary system is important for the hot-dip galvanizing process and has been investigated in a number of studies [1281, 1282]. 1.3. MAIN PRODUCTS AND APPLICATIONS
1.3.1. Zinc Coatings The many types of zinc and zinc alloy coatings can be classified according to the coating composition and the production methods employed [501, 1250, 1297]. When classified according to chemical composition, zinc-based coatings fall into several major categories: pure zinc, zinc-iron, zinc-aluminum, zinc-nickel, and zinc composites. In terms of methods, zinc coatings can be produced by hot-dipping, electroplating, mechanical bonding, sherardizing, and thermal spraying (metallizing). The hot-dip method can
100
200
300
400
:;00
Al
10 20
10
30
50
I
30 f
60
!
40
Weight Percent Zinc
40
20
Alomi c Per ce nl Zinc
70
.'
50
60
60
.JoT"""'"
I
70
90
I
Zn
100
'11I.M·e
90 100
(Zn)
"
60
FIGURE 1.1. Zinc-aluminum equilibrium diagram. From Baker [1228]. Reprinted with permission from ASM International.
CU b
E
0..
CU
...
....
...cu ::I
0
U
600
~2"(
700
600
0
n
tTl ;Al
-l
'"0
;l>
:c
"'"
f-
QJ
.~.~l~C
,, ,, ,
,,
-I
- " _',
, ............ 4
782"C
60
-2.-'-'---_.- _._._._._'."UII:!:. ._._. _._._._.-
,,
50
70 80
L
I
,90
\-' e
We ig h l Percen l Zinc:
:l00 \· ( · .. ·····, ·"··· .. ·T·· .. .. · .. ,· .. ·· .. ··,··,. ~· ··· , .... n-.'~~ . ,,~1'-- " 1-"~ '~ I o 10 20 :.to "0 :;0 60 70 60 90
400 ·
:;00
600
700
(aFe)p
,, ,, ,,
40
I
100
100
Zn
FIGURE 1.2. Zinc- iron phase diagram. From Baker [1228J . Reprinted with permission from ASM International.
E
a.
600
912"C 900 -
....
cO
1000
1100 ·,
::I
....
u o
, 200
1300
,,
,,
30
;;0
til
'" '" tTl '"
tTl
n
o
-0 ;;0
o
-l )/' ;J>
c
o o
Z
20
1400 1394"C
10
n
I
o
~
tTl
.'"-0
1638"C 1500
1600
Alo m lc P erce n l Zinc
tTl
o-0
-0 ;;0
6
CHAPTER I
TABLE 1.2. Typical Applications of Zinc-Coated Steel Products" Coating
Applications
Coatings by continuous process Zn and Zn-5AI Roofing, siding, doors, culvert. ductwork. housing. appliances. autobody Sheet panels and structural components Nails. staples. guy wires. stand. tension members. rope. utility wire. Wire fencing Autobody panels and structural components Zn-Fe Autobody panels and structural components. housings. appliances. Zn-Ni fasteners Automotive small parts and fasteners Zn-Co Roofing, siding, ductwork, culvert. mufflers. tailpipes. heat shields. ovens. Zn-55Al toasters, chimneys. silo roofs Batch gal vanized Structural steel for power-generating plants. petrochemical facilities. heat exchangers, cooling coils. water treatment facilities. and electrical transmission towers and poles Bridge structural members. culverts, corrugated steel pipe. arches Reinforcing steel for concrete structures Highway guard rails. lighting stands, sign structures Marine pilings. rails Grates. ladders. safety cages Architectural applications of structural steel. lintels. beams. columns. and related building materials Painted galvanized structural steel for aesthetic. color-coded. or extendedlife applicatIOns "Refs. 1238. 1250. and 1296.
be further divided into two processes: continuous hot -dip and batch hot -dip. In continuous hot-dip, long strands of sheet, wire, or tubing are fed through a bath of molten zinc alloy in a continuous process. In batch hot-dip, fabricated parts, such as fasteners, poles, or beams, are dipped into a molten bath either individually or in discrete batches. Similarly, zinc electroplating can be made in a continuous or a batch process. Typical applications for zinc- and zinc-alloy-coated steel sheet products cover a wide range in the construction, automobile, utility, and appliance industries as shown in Table 1.2. As the cost of lumber increases, additional large-scale applications of zinc-coated steel products are also expected to develop in the residential construction markets for roofing, siding, and framing [1250]. Among all coated steel products, continuous-hot-dip zinc-coated steel sheet has the widest range of applications and is predominant in terms of tonnage produced and consumed. The electroplated zinc-based coatings are applied primarily on automotive bodies and have advantages of uniform coating thickness and excellent surface characteristics for subsequent painting, but certain disadvantages in terms of cost [1296]. Zinc coatings can also be produced by sherardizing and thermal spraying [527]. Sherardized coatings are produced by a cementation or diffusion process in which the steel parts are heated with zinc dust in a slowly rotating drum. They are suitable for nuts, bolts, hinges, nails, and similar hardware and fittings. Thermal-sprayed zinc coatings are obtained by melting zinc powder or wire in a flame or electric arc and projecting the molten metal by air or gas onto the surface to be coated. This process is mostly used with structural steel
PROPERTIES, PRODUCTS, AND PROCESSES
7
works that are too large for a galvanizing bath, for example, steel bridges. It is also used for repairing galvanized steel surfaces on which the coating is lost due to mechanical damage or welding. More recently. thermal-sprayed zinc coatings have been used on concrete surfaces to serve as the anode for cathodic protection. Sprayed zinc coatings are relatively rough and porous compared to the coatings produced by other methods.
1.3.2. Cast Products Cast zinc products are mainly produced by the die-casting process. in which liquid metal is forced under pressure into a cooled die and solidifies almost instantaneously to produce a fine-grained product. Die casting is a single high-speed operation that can produce complex but very accurate components requiring little or no tinal shaping. Die-cast products are used for automotive parts, household appliances and fixtures, office and computer equipment, and building hardware. A typical composition for the most commonly used die-casting alloy (Alloy 3) is Al 4.0%, Mg 0.03%. Cu < 0.25%, Fe < 0.1 %. The aluminum content contributes to the alloy's good castability and strength [218, 501].
1.3.3. Rolled Zinc Rolled zinc products are in the form of sheet, strip, foil, plate, rod. and wire, with a variety of compositions. Rolled zinc sheet is widely used in building. in the form of roofing, cladding, gutters, rainwater pipes, and flashings. The rolled zinc used for roofing is typically a Zn-Cu-Ti alloy (0.7-0.9% Cu and 0.08-0.14% Ti), which gives a good combination of tensile strength. creep strength, and formability. Rolled zinc foil has been made into adhesive tape for coating the surfaces of large structures that are difficult to galvanize. Zinc wires are primarily used for metal spraying [501].
1.3.4. Zinc Dust and Powder Zinc dust and zinc powder are particulate materials. The word "dust" is used for fine particles, usually 2-20 .um in diameter, and "powder" for coarser particles. A distinction is often made whereby "zinc dust" refers to the product made by condensation of zinc vapor, and "zinc powder" to the product of atomization of molten zinc by a jet of air or an inert gas [SOl]. Zinc dust may also be made in the form of flakes by milling in a nonreactive fluid such as a hydrocarbon, a process referred to as "flaking." Flake thickness tends to be 1 .um or smaller, and the diameter-to-thickness ratio may be about 10. Zinc dust and powder are used mainly as reagents for producing chemicals, in metal refining processes. as a component for making zinc-rich paints. and as an active material for zinc batteries. They are also used in smaller quantities as a material for sherardizing and thermal spraying and as an additive in plastics [501]. 1.4. COATING PROCESSES
1.4.1. Hot-Dip Galvanizing 1.4.1.1. General Considerations. Hot-dip galvanizing is a process by which an adherent coating of zinc and zinc-iron alloys is produced on the surface of iron or steel
8
CHAPTER 1
products by immersing them in a bath of molten zinc. It is the oldest and the most used process for producing zinc coatings. Hot-dip galvanizing can be further divided into two main processes: batch galvanizing and continuous galvanizing. In general, an article to be galvanized is cleaned, pickled, and fluxed in a batch process or heat-treated in a reducing atmosphere to remove surface oxide in a continuous galvanizing process. It is then immersed in a bath of molten zinc for a time sufficient for it to wet and alloy with zinc, after which it is withdrawn and cooled. Any of these stages can be critical to coating quality. The coating so produced is bonded to the steel by a series ofZn-Fe alloy layers with a layer of almost pure zinc on the surface. The engineering quality of the coating depends on the physical and chemical nature of the Zn-Fe intermetallic layers formed. The thickness and composition of the alloys depend on whether they are produced in a batch or a continuous process, mainly because of the differences in the immersion time in the molten zinc bath and the bath composition employed in the two types of processes. The coating produced by a batch process is thicker and has clearly distinguishable alloy layers as shown in Fig. 1.3, while that produced by a continuous process is thinner and has only a very thin and sometimes not visible (with an optical microscope) alloy layer at the coating/steel interface as shown in Fig. 1.4. Table 1.3 gives the characteristics of the Zn-Fe alloys in a batch hot-dip galvanized coating. The thickness of a given phase in a Zn coating on steel is determined by the rate of diffusion through the phases during their growth [262]. The main diffusion process is diffusion of Zn through the galvanized layer toward the iron interface. The diffusion of the iron moving outward occurs at a much slower rate. During a galvanizing process, the
Fe
FIGURE 1.3. Cross section of a typical batch-galvanized coating, showing the various Zn-Fe alloy layers.
PROPERTIES, PRODUCTS, AND PROCESSES
9
FIGURE 1.4. Cross section of a typical coating produced in a continuous process.
( layer is fonned first, followed by the i5 layer and, finally, the rlayer. The growth rate of these different layers in the coating is shown in Fig. 1.5 [262]. The ( layer grows rapidly at first but then much more slowly, while the growth of the i5 layer is at first slower than that of the ( layer but then becomes faster. The growth rate of the r layer is very slow, and therefore this phase may not be seen under the microscope at short reaction times. The formation of zinc-iron alloy layers greatly depends on the silicon content of the
steel, as shown in Fig. 1.6 [262]. At a nonnal galvanizing temperature, the coatings on steels with low Si concentrations, less than 0.03%, have nonnal thicknesses (Fig. 1.3). On the other hand, at high concentrations of Si, e.g., 3%, the reactivity of the steel is low and thin coatings are produced. At intermediate concentrations of Si, the reactivity of the steel is high and thick zinc-iron alloy layers are produced. The peak reactivities occur at Si contents of 0.06-0.1 % and about 0.5% as shown in Fig. 1.6. Figure 1.7 shows the cross-sectional structure of a typical zinc coating on an Si-containing steel. The addition
TABLE 1.3. Characteristics of Zn-Fe Intermetallic Alloys" Phase
Formula
Fe content (wt. %)
'7 ( ,)1
Zn FeZn13 FeZnw FeSZn21
Max. 0.003
r
"Refs. 262, 312, and 501.
5.7-6.3 7.0-11.5 21.0-28.0
Crystal structure Hexagonal close-packed Monoclinic Hexagonal close-packed Face-centered cubic
Densi ty (glcm 3)
7.14 7.18 7.24 7.36
10
CHAPTER 1
200 r----------------------------------------;
Time, hours
FIGURE 1.5. Rate of growth of r, "I' and (layers at 45TC. After Mackowiak and Short [262].
of small amounts of aluminum to the zinc bath may effectively inhibit the growth of the alloy layers on silicon-containing steels. The as-galvanized coating is typically characterized by the appearance of spangles, which often show a strong (000 1) basal texturing [261]. Also, the surface of fresh galvanized coating is readily oxidized in air to form a very thin oxide film. The surface oxide film on the galvanized steel produced in an aluminum-containing bath is usually enriched with aluminum, owing to the high affinity between aluminum and oxygen, and has a thickness varying from 20 to 100 A depending on the content of aluminum in the coating [253,491]. 1.4.1.2. Batch Galvanizing. In batch hot-dip galvanizing, the articles to be galvanized are first degreased and then pickled to remove mill scale and rust from steel parts. 3~---r----.-----r----r----.----.
o (), • •
3mln.455·C 4 min. 460·C 8mln. 455·C 8mln. 460·C
FIGURE 1.6. Comparison of the results of various studies on the effect of silicon content on OL---"l:---...."J.;;:----:h::----,.'-:;--~--O;:;:·"i6 coating weight. After Mackowiak and Short [262].
PROPERTIES, PRODUCTS, AND PROCESSES
11
FIGURE 1.7. Cross section of a zinc-iron alloy coating structure obtained on a high-Si (0.4%) steel.
Each of the degreasing and pickling steps is followed by a water rinse. The most common degreasing process uses heated (65-82°C) alkaline solution. Aqueous solutions of 3-14 wt. % sulfuric acid or 5-15 wt. % hydrochloric acid are generally used in pickling. To avoid overpickling, inhibitors are often used [312]. Batch galvanizing can be a wet or a dry process. Wet galvanizing involves a kettle-top flux blanket; dry galvanizing uses a preflux but does not use a flux blanket on the kettle. In the dry process, after the steel article is degreased and pickled, it is immersed in an aqueous zinc ammonium chloride solution, dried, and then immersed in the molten zinc bath. In the wet process, the article is not usually prefluxed after cleaning but is placed directly in the molten zinc bath through the top flux blanket. Zinc ammonium chloride is generally used as the flux blanket. The fluxing promotes the alloying process at the steel/molten zinc interface by removal of FeO on the steel substrate and ZnO on the surface of molten zinc through chemical reactions with ZnCI 2 and NH 4Cl. The molten zinc bath generally operates in a temperature range of 445-454°C. The bath temperature affects the fluidity of the molten zinc, the rate of formation of oxides on the bath surface, the rate of coating solidification, the coating thickness, and the amount and structure of the Zn-Fe alloy layers. The immersion time is usually in the range of 3-6 min. The speed of immersion and withdrawal influences the coating uniformity, particularly with large articles. 1.4. J.3. Continuous Galvanizing. In the continuous hot-dip coating process, coils of steel are welded end to end and are coated at speeds of up to 200 mlmin [1254]. In general, there are "hot" and "cold" continuous hot-dipping processes. The major differ-
12
CHAPTER I
ence between the "hot" and "cold" processes is in the preparation of the steel surface after the first cleaning stage and before immersion in the molten zinc bath. In the "hot" process the strip first enters an alkaline bath that removes oils, dirt, and residual iron fines from the rolling process. This is followed by a further cleaning stage with mechanical brushing and electrolytic alkaline cleaning. The sheet then passes into a radiant tube furnace containing a mixture of hydrogen and nitrogen that reduces surface iron oxides. Heating of the steel also takes place to a temperature just above that for subcritical recrystallization. The steel is then cooled to near bath temperature before entering the zinc bath. In the cold process, steel strip is cleaned, pickled, and fluxed in-line with no heating beyond that required to dry an aqueous flux solution of ammonium chloride and zinc chloride on the steel surface before entering the zinc bath. Typically, 0.1-0.2% Al is added to the bath to prevent the formation of a thick, continuous layer of Zn-Fe intennetallic that could lead to poor coating adhesion during forming [1250]. As the steel strip exits the bath, a layer of molten zinc is coated on the surface. The thickness of the layer is controlled by passing the strip between wiping dies to remove excess metal with a stream of gas. Forced-air cooling is used to reduce the sheet temperature, which prevents coating damage from contact with turnaround rolls. Before the sheet is finally wound into the coil fonn, it may be subjected to one or more post-treatments such as oiling, chromating, and phosphating. 1.4.1.4. Galvannealing. In galvannealing, the hot-dipped steel sheet is processed further. After exiting the molten zinc bath and passing through the wiping dies, it is heated to temperatures of 500-550°C for about 10 s to generate interdiffusion of iron from the substrate and zinc from the coating to fonn an Fe-Znalloy coating. The actual alloying time is influenced by the coating thickness and the compositions of both the zinc bath and the steel substrate. Compared to the galvanized coating, the galvannealed coating is generally easier to paint without a special pretreatment, probably because of its rougher surface. The outer surface of a galvannealed coating is a , phase containing about 6% Fe. The intermediate J phase contains iron in the range of 8-12%. Next to the steel substrate is a Tlayer. These intermetallic phases differ significantly in mechanical properties and detennine the forming properties of the coating. 1.4.1.5. Zn-AI Alloy Coatings. Zn-55% Al (Galvalume) and Zn-5% Al (Galfan) coatings are the two major hot-dip commercial zinc-aluminum alloy coatings. Ga1valume, containing 55% AI, 1.5% Si, and 43.5% Zn, was developed by Bethlehem Steel [353, 1250]. This alloy has properties intermediate between those of hot-dipped zinc and aluminum coatings. Galvalume has a higher corrosion resistance but less galvanic action than a zinc coating. The microstructure consists of an outer layer and a thin intennetallic layer that bonds the outer layer to the steel. This thin layer consists of two intennetallic compounds: the inner sublayer is a quaternary AI-Fe-Si-Zn compound, and the outer sublayer is a ternary A1-Si-Fe compound. The silicon moderates the reaction during hot dipping and serves to minimize the thickness of this intermetallic layer [238]. About 80 vol % of the outer layer is composed of cored, aluminum-rich dendrites, representing the first solid fonned during cooling. The last liquid to freeze in the interdendritic volume between the aluminum-rich regions is enriched in zinc. Galfan, containing 95% Zn, 5% AI, and a small amount of misch metal, exhibits improved fonnability and increased corrosion life. The multiphase microstructure of
PROPERTIES, PRODUCTS, AND PROCESSES
13
Galfan is characteristic of its composition, exhibiting a lamellar structure of alternating zinc-rich and aluminum-rich phases. The fineness of the structure increases with increasing cooling rates, and the structure is completely eutectic when fast-cooled. It is also oriented in the direction of cooling. One characteristic of the Galfan microstructure is the virtual absence of a brittle intermetallic phase between the steel and the coating. This has been attributed to the addition of misch metal, which allows complete wetting of the steel surface. The lack of the alloy layer is directly responsible for the high formability of Galfan-coated steel [353].
1.4.2. Electroplating Electroplating is another common method for producing zinc coatings on steel surfaces. The plating process generally comprises three stages: (1) degreasing and cleaning, (2) electroplating, and (3) post-treatment. Various types of plating baths are used in the plating industry, and these can be roughly classified as acid or alkaline. Most commercial zinc plating before 1980 was done in conventional alkaline cyanide baths. The environmental concerns related to cyanide use have led to continuing development and application of other processes. At present, acid zinc plating baths constitute about half of all zinc baths in developed nations, and their use is rapidly increasing throughout the world [1297]. Typical compositions for acid and alkaline plating baths are given in Table 1.4. 1.4.2.1. Continuous Plating Process. Continuous plating is a process for plating a metal coating onto an endless steel sheet or wire. It consists of five main sections: payoff, pretreatment, plating, post-treatment, and delivery [1296]. In the payoff section, the coils of cold-rolled steel are loaded onto the entry reels, with the end of a new coil welded to the tail end of the previous coil. The strip then passes through a precleaning and rinse station, where the bulk of the oils is removed. In the pretreatment section, the residual oil, surface carbon, and light surface oxide are removed as the strip is passed through one or more alkaline cleaning, brushing, pickling, and rinsing stations. The cleaned strip then enters the plating section, which consists typically of multiple plating cells. At the beginning of the plating process, a conditioning cell may be used to prepare the steel surface for plating. As the strip moves through from cell to cell, the coating
TABLE lA. Typical Plating Bath Compositions and Plating Conditionsa Chloride bath Composition
ZnCI 2, 15-56 gil NH4 C\, 100-200 gil Brighteners, 3-5%
Temperature pH Current density
21-27°C 5.2-6.2
aRefs. 312 and 1297.
0.3-5A1dm 2
Cyanide bath Zn(CN)z, 54-86 gil NaCN, 30-41 gIl NaOH, 68-105 gil Na2C03' 15-60 gil Sodium polysulfide, 2-3 gil Brightener, 1-4 gil 21-40°C
-13
1-5 Aldm 2
14
CHAPTER I
thicknessis gradually built up to the required value. Upon exiting the last cell, the coated sheet is immediately rinsed and dried to prevent streaking or staining of the coated surface. In the post-treatment section, the steel strip is treated with processes such as phosphating, chromating, or oiling to prepare it for painting or to provide extra surface protection. Continuous plating processes are generally classified in terms of three characteristics: anode type, electrolyte chemistry, and plating cell geometry [1158]. Either soluble or insoluble anodes are used. Soluble anodes dissolve anodically into the electrolyte, whereas the reaction on insoluble anodes is generally the oxidation of water. Soluble anodes are usually used in chloride baths, and insoluble anodes are usually used in sulfate baths. There are three types of cells based on cell geometry: vertical, horizontal, and radial. The choice of anode type, bath chemistry, and cell geometry in an electroplating line depends on the capacity, productivity, automation. specialized components, etc. 1.4.2.2. Alloy Plating. The development of zinc alloy coatings has addressed the need for more corrosion-resistant automotive bodies. The steel sheets used for automotive body panels must have not only high corrosion resistance but also good paintability, formability, and weldability. Currently, the most prevalent zinc alloy coatings are Zn-Fe and Zn-Ni [1158]. Other alloy and composite coatings such as Zn-Co [229,246], Zn-Mn [330, 349], Zn-Co-Mo [312], Zn-Co-Cr [425], Zn-Ni-Si0 2 [263], Zn-Ni-Ti [487], Zn-Si0 2 [284], and Zn-Co-Cr-Alp3 [351] have also been developed. Table 1.5 shows the typical compositions of Zn-Fe and Zn-Ni alloy plating baths. The content of Fe and Ni in the alloy coating can be greatly varied by changing the bath chemistry and operating conditions [227, 312]. The typical Zn-Fe coatings contain approximately 18% Fe, and Zn-Ni coatings contain 9-13% Ni. The deposition of Zn-Fe and Zn-Ni alloy coatings represents cases of anomalous codeposition [37]. In an anomalous codeposition the less noble metal is predominantly deposited. This is opposite to the normal process, where the more noble metal is predominantly deposited. For Zn-Fe or Zn-Ni coating at low current densities, zinc and iron or zinc and nickel codeposit by a normal process. However, at high current densities, zinc is predominantly deposited. This arises from an increase in pH near the cathode surface under high-current conditions. The higher pH favors the formation of Zn(OH)2 and thus reduces the deposition sites for iron or nickel.
TABLE 1.5. Typical Bath Compositions and Conditions for Plating Zn-Fe and Zn-Ni Alloy Coatings" Zn-Fe bath Composition
Temperature pH Current density "Ref. 312,
ZnS04,7H 20, 50 gIl FeS04,7H 20, 250 gIl (NH4 hS04, 120 gil CSH S07' 0,5 gil 50°C 1 30 Ndm 2
Zn-Ni bath ZnS04,7H20 + NiS0 4,6H20, 500 gil Na2S04, 60 gil 50°C 2 10-15 Ncm 2
15
PROPERTIES, PRODUCTS, AND PROCESSES
1.3. PHOSPHATING Phosphating is a surface treatment process in which metals such as iron, zinc, and aluminum and their alloys are treated with a solution of phosphoric acid and other chemicals. The reaction between the surface of the metals and the solution results in the formation of an integral layer of insoluble crystalline phosphate. Phosphate coatings typically range in thickness from 3 to 50 ,urn and vary in color from iridescent blue to dark gray [163, 578]. Phosphating on zinc-coated steels is used mainly to prepare the metal surface for painting. Phosphate coatings provide uniform surface texture and increased surface area, and, when used as a base for paint, they promote good adhesion, increase the resistance of the paint to humidity and water soaking, and eventually increase the corrosion resistance of the painted system. Phosphate coatings can be produced by spray, immersion, or a combination of the two [578]. There are three principal types of phosphate coatings used in the industry: zinc, iron, and manganese phosphate. In the strip galvanizing lines, zinc phosphate baths are usually used. The zinc phosphating bath is typically operated in the pH range of 1.4-3.4 and at a temperature in the range from 30 to 100°C. The process time can vary from several seconds to several minutes. In general, the spray method produces a coating at a faster rate than the immersion method. A phosphating bath usually contains an accelerating agent to speed up the rate and to reduce crystal size. The composition of phosphating solutions varies greatly and depends on the specific application. Many phosphating solutions used in the industry are proprietary. A simple phosphating bath can, for example, contain 6.4 g ZnO/I, 10 ml H 1POil, 4 ml HNO/I, and I g Ni(N0 3h/1 [94]. A dark gray coating can be obtained on galvanized steel after 5 min in this solution at about 60°C. Most phosphated articles that are used as a paint base are also given a post-treatment with a rinse of chromic acid or other solutions [5781. Phosphate coatings, when used for corrosion protection, are usually topped with an oil or wax to seal the pores in the coating. Phosphate coatings are formed through a dissolution and a precipitation process. The zinc dissolves in the phosphating solution with accompanying hydrogen evolution. As a result of the hydrogen reduction, a thin layer of electrolyte near the surface is depleted of hydrogen ions and becomes neutralized. Since the solubility of zinc phosphate in neutral solution is low, zinc phosphate precipitates on the zinc surface to form the crystalline phosphate coating. This process involves essentially three reactions [93-95, 11751: I. Dissolution of zinc: Zn
~
Zn 2+ + 2e-
2. Reduction of oxidizing agents to the reduced form: Ox (4W, O2 , 4NO:;, etc.) + 4e- ~ Re (2H2' 20 2-, 4NO, etc.) 3. Precipitation of a zinc phosphate layer, hopeite: 3Zn2+ + 2H 2 PO:;- + 4HP ~ Zn/P0 4 )2 . 4H 2 0 (s) + 4W and also a zinc-iron phosphate layer, phosphophyllite, in the presence of Fe 2+:
16
CHAPTER I
The kinetics of nucleation and growth is systematically discussed in the book by Rausch [1175]. The most important reaction in controlling the thickness of the phosphate coating is the nucleation of phosphate crystals. To obtain a large number of nuclei, and thereby a fine-grained coating, the increase in pH at the metal surface should be as fast as possible during the first seconds of fonnation of the phosphate coating. Dipping in a colloidal solution of titanium phosphate is usually used as an activation process to increase the number of nuclei. The phosphate coating process is completed when the surface of the metal is so fully covered by the crystalline phosphate that no further significant neutralization of the near-surface liquid layer can take place. 1.4. CHROMATING Chromating is a process in which an aqueous solution of chromic acid, chromium salts, and mineral acids is used to produce a thin conversion coating on a metal surface. The chemical reactions between the metal and solution cause the dissolution of the metal and fonnation of a protective film containing complex chromium and metal compounds. Chromate conversion coatings can be applied to a number of metals. They are most commonly used to protect coatings based on zinc and its alloys during storage and transportation. They also generally enhance the corrosion resistance at zinc and zincalloy coatings on tubings, fasteners, etc. Since its introduction in the mid-1930s [57], the chromating process has become the most widely used process for surface post-treatment of zinc products. However, because of environmental concerns, its applications have become increasingly limited, which has led to new research activities in search of alternatives [1258]. Most of the chromating formulations used today are proprietary [1258]. A conventional chromating process, for example, the "Cronak process" developed by The New Jersey Zinc Company in 1936, consists of immersion of the zinc article for 5-15 s in a chromate solution (200 g of Na2Cr207·H20 and 6-9 ml of H 2S04 in II of water at 20°C), followed by rinsing and drying [57]. The detailed fonnation mechanism for chromate coatings is not fully understood. In general, the fonnation follows a dissolution and precipitation process similar to that in phosphating [67]. During chromating, a set of electrochemical and chemical reactions occur on the zinc surface in contact with the chromate solution. There are primarily three reactions: zinc dissolution coupled with a cathodic hydrogen reduction, reduction of hexavalent chromium ions to trivalent ions, and precipitation of trivalent chromium hydroxide which incorporates hexavalent chromium compounds and zinc compounds. The precipitation of the hydroxide of chromium is promoted by the rise of surface pH through the local consumption of acid [65]. The thickness, composition, and color of chromate coatings depend mainly on chromate concentration, pH, and dipping time [57,65]. For example, a yellow coating typically ranges in thickness from 0.1 to 0.6 pm [57]. The freshly fonned films are gel-like and do not reveal any crystalline components by X-ray diffraction examination [59,65]. They harden from their original soft state to a reasonable abrasion-resistant coating within 24 h in air. The hardening also results in the crystallization of the coating, which can be
17
PROPERTIES, PRODUCTS, AND PROCESSES
TABLE 1.6. Composition of the Chromate Coating Formed by the Cronak Process" Constituent Chromium(VI) Chromium(III) Sulfate Sodium Zinc Water
Relative amount (%)
7-12 25-30 2-3.5 0.2-0.5 2-2.5 15-20
"Ref. 57.
observed from X-ray diffraction patterns. Cr20 3 , ZnCr0 4 , and ZnO have been detected in chromate coatings. Table 1.6 shows the composition of a chromate coating [57]. Chromate coatings protect the zinc metal through barrier and passivation effects. The complex chromium oxide film serves as a barrier to the environment, while the hexavalent chromium contained in the film serves as a passivating agent. The hexavalent chromium leaches out when in contact with water and produces a local chromate solution that forms a chromate film at an exposed zinc surface [65]. Through leaching, immersion of a chromated surface in distilled water for 24 h results in a marked loss of chromate. A similar effect is seen in outdoor exposure [71].
2 Electrochemical Thermodynamics and Kinetics 2.1. INTRODUCTION Electrochemical processes playa very important role in the production and application of zinc. Electrowinning in zinc refining, electroplating in the production of zinc coatings, zinc batteries for energy storage, and zinc coatings and anodes for corrosion protection are all essentially based on electrochemical processes. In this chapter, the thermodynamic and kinetic properties of the zinc electrode are reviewed. The material presented is organized according to each of the elemental reactions that can occur on a corroding electrode: zinc dissolution, zinc deposition, hydrogen evolution, and oxygen reduction. While zinc dissolution is discussed in detail, zinc deposition is only treated superficially since it is not important from a corrosion perspective. The topics of oxide film formation and passivation are dealt with in a separate chapter because of the large amount of literature on these phenomena and their particular importance in .corrosion processes. The corrosion potential and corrosion current, which are the two key parameters connecting the fundamental electrochemistry and practical corrosion behaviors in various applications, are also considered in another chapter. The information presented in this chapter is limited to aqueous solutions. Some electrochemical information on the zinc electrode in nonaqueous electrolytes is presented in Chapter 15. A discussion of the zinc electrode kinetics in nonaqueous electrolytes and fused salts can be found elsewhere [532]. 2.2. THERMODYNAMIC STABILITY Zinc is divalent in all its compounds. Compounds of Zn(l) do not exist naturally [1253J. The stability of zinc and its compounds in aqueous solutions in the absence of complex formation is determined by the equilibrium conditions listed in Table 2.1 [1,906]. The value for the standard potential of the zinc electrode can be calculated from thermodynamic data [I]:
Eo = LVjJ.123060n = -35,18412 x 23,060 = -0.763 V SHE (2.1) 19
20
CHAPTER 2
TABLE 2.1. Reactions of Zinc in Aqueous Solutions and Equilibrium Conditions" Reaction
Equilibrium
Standard potential or equilibrium condition
Two dissolved substances I Zn 2+ + H20 =ZnOH+ + H+ 2 ZnOH+ + H20 =HZnO;- + 2H+ 3 Zn 2+ + 2H 20 =HZnO;- + 3H+ 4 HZn02 =ZnO~- + H+ Two solid substances
ZnO + H20 =HZnO;- + H+ ZnO + H 20 =ZnO~- + 2H+ Zn Zn 2+ + 2eZn + 2H 20 =HZnO;- + 3H+ + 2eZn + 2H 20 =ZnO~- + 4H+ + 2e-
log(Zn 2+) = 10.96 - 2pH log(HZnOz) =-16.68 + pH log(ZnOh =-29.78 + 2pH Eo =-0.763 + 0.0295Iog(Zn 2+) Eo =0.054 - 0.0886pH + 0.0295 log(HZnO;) Eo =0.441 - 0.1182 pH + 0.0295Iog(ZnO~-)
H2 =2H+ + 2e2H20 =O2 + 4H+ + 4e-
Eo =0.000 - 0.0591 pH Eo = 1.228 - 0.0591 pH
=
Stability of water (a)
(b)
=
Eo =-0.439 - 0.0591 pH
5 One solid and one dissolved substance 6 Zn 2+ + H20 =ZnO + 2H+
7 8 9 10 II
log (ZnOH+)/(Zn 2+) =-9.67 + pH log (HZnO;- )/(ZnOH+) =-17.97 + 2pH log (HZn02 )/(Zn2+) -27.63 + 3pH log (ZnO~-)/(HZnO;- ) =-13.17 + pH
"Ref. I.
where v is the stoichiometric coefficient. n is the number of electrons involved in the reaction, and f..l is the chemical potential of the species involved in the reaction. Temperature has slight effect on the zinc potential as shown in Fig. 2.1 [532]. At equilibrium the potential difference between various crystal surfaces of a zinc single crystal and polycrystalline zinc is less than 10mV [532]. The potential values measured in different zinc salt solutions generally show good agreement with the calculated values.
-0.76
+ Ref. 2 ... Ref. 1
+ Ref. 7 -0.762
• Ref. 6
-W
"
~
0
w
-0.764
-0.766 ' - - -_ _...l.-_ _ _- " -_ _ _ _' - -_ _ _. . . l . - _ - l o 10 20 30 40 Temperature,
°c
FIGURE 2.1. Effect of temperature on the zinc electrode potential determined experimentally. The reference numbers in the figure are from Brodd and Leger [532]. Reprinted by courtesy of Marcel Dekker, Inc.
21
ELECTROCHEMICAL THERMODYNAMICS AND KINETICS
The equilibrium conditions listed in Table 2.1 can be represented by the Pourbaix diagram in Fig. 2.2 [1]. The lines labeled with the letters a and b represent, respectively, the equilibrium conditions of the reduction of water to gaseous hydrogen and of the oxidation of water to gaseous oxygen, when the partial pressure of hydrogen or oxygen is I atm at 25°C. According to Fig. 2.2, the stable region of zinc is below line a, and thus zinc is thermodynamically unstable in water and aqueous solutions and tends to dissolve with the evolution of hydrogen over the whole pH range. In solutions of pH between approximately 8.5 and 12, zinc can be covered with a hydroxide film, which has the effect of inhibiting zinc dissolution. The pH-potential diagram in Fig. 2.2 is valid only in the absence of the chemical species with which zinc can form soluble complexes or insoluble compounds. The stability of zinc oxides and hydroxides in aqueous solutions depends on pH. As shown in Fig. 2.3, zinc hydroxides are amphoteric. They dissolve in acid solutions to give
a
-2 -I
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
2,2,-,----i'---.--T-T--....;-~-T-......;---.:_;___=;...__.;~~~'---~~~~.:,'2
EM
,
2
$,
1,B
1.4
2
$
,,
1,6
1,2
,
1,6
,
1,4
I
1,2
I
o
I I
-2 -6
6
I I I
I I I I
-6
-....
0,8
-2
I
0,6
o
I-_ I
8
I
0,4
0,4 0,2
I
Z~(OH)2 I I
-0,4
I
I I I I
i-_ I
-~81~~9~~~~~~~~~~~=d~~i F 0
-z
°
I
-0,2
-0,6
-4 -6
0,6
,
I
° "0-_
0,8
I
I I
0,2
2
I,B
Zn02
--®- __
'
-0,2
,,
-0,4
I I
-0,6
...,-
-0,8
I I I I I
,,
-I
-1,2
-1,2
-1,4
Zn
-1,6
-I, 8~7--;.;---:--;';---::--:---::---=--=_-::-~---:''::--"-:---:'::--:'!:--,L--,.L.-.-l-1 8 -2
-I
0
2
3
4
5
6
7
8
14
15pH16'
FIGURE 2.2. Potential-pH equilibrium diagram for the zinc-water system at 25"C [established by considering Zn(OHh]. From Pourbaix [I]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
22
CHAPTER 2 1
5
4
6
7
8
9
12
11
10
13
0
,
-1
'/
?" ~ -2
1
a
9/
-\
,~""'/ -$> /
//,
-2
~ /
-)
-3
~ /
~/
~
'1,/
~ -4
-.
15 16
(-
.§
+ ..-...
14
-4
~ + -5
~
-
L:::::J
g>
-5
-6
-6
-7
-7
-8 4
5
6
7
8
10
12
11
13
14
15
pH
-8
FIGURE 2.3. Influence of pH on the solubility of the zinc hydroxides in water at 25°C. From Pourbaix [IJ.
© Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
zincic ions Zn 2+ and in alkaline solutions to give bizincate or zincate ions HZnO;- and Znq- [1]. The solubility varies slightly with the type of hydroxides and oxides. At room temperature, e-Zn(OHh is the most stable compound whereas amorphous Zn(OHh is the most unstable [404]. Figure 2.4 shows the experimentally measured solubility of ZnO as a function of KOH concentration [1131]. Zinc can form insoluble compounds with many chemical agents. The compounds commonly found in corrosion products are zinc sulfate, chloride, and carbonate. The stability of these compounds has been found to affect the corrosion resistance of zinc in many environments [331]. Zinc carbonate is of particular importance because it has been
2 :E
•
c-
o
';a ~ Q)
•
<.l C
. 0
U
'C::: N
0
0
5
10
15
KOH Concentration , M
FIGURE 2.4. Solubility of ZnO in KOH at 23°C. After Hampson et al. r1131].
23
ELECTROCHEMICAL THERMODYNAMICS AND KINETICS
2
3
ZnO,
b
- - - - -- - - - -
--;;:. ~
CUl
'
,
ZnCO, ,
-
I I
-:- ~ I
I
·1
Zn
I
, I
H,CO,
-2
5
I
I
~
Zn"
a
- - -- - - ---
O
, "t -
,
~
4
I
-
-: - Ln(OH),(CO,),
- - ,, - - - - - -, ZnO, - - ----- - L I
IICO,-
I
o
4 pH
Z,,(OH),-
I
Zn(OI1).'-
~
I I I
COll-
12
8
FIGURE 2.5. Potential-pH diagram for the zinc-water-carbonate system at 298 K. [Zn 2-j = 10-4 molldm 3 (dissolved zinc species); [H 2CO]j + [HCO]l + [CO~-l = 10-2 molldm J (dissolved carbonate species). After Kannangara and Conway [3[.
found to be responsible for the high corrosion resistance of zinc in atmospheric environments. Theoretically, the formation of zinc carbonate can occur in solutions containing carbonates and bicarbonates according to the following reactions 13. 906J: ZnO + 2H+ ~ Zn 2+ + H 2O
(2.2)
Zn 2+ + H 2CO] ~ ZnCOJ(s) + 2W
(2.3)
1
0
-1
I
..... -2
u
'"
00
0
.....
-3
Solution
-4
-s
5
6
7
B
9
10
11
pH FIGURE 2.6. Equilibrium diagram for corrosion products of zinc in a chloride environment. After Feitknecht [404].
CHAPTER 2
24
(2.4)
2W + 2HCO:J + HzO + 5ZnO (s)
~
(2.5)
Zns(OHMC03h (s)
The pH-potential diagram for the zinc-water-carbonate system is shown in Fig. 2.5. It can be seen that the presence of carbonates and bicarbonates extends the possible passivation region to near neutral pH values. It has been calculated that when the total concentration of HZC0 3 + HCO:J + CO~- is larger than lO-zAM, the domain of zinc passivation by formation of a zinc carbonate film becomes larger toward neutrality than that of zinc passivation by formation of a zinc hydroxide film [906]. At pH values greater than 9, zinc carbonate is less stable than zinc hydroxide. The stability of zinc sulfate and zinc chloride are determined by the following reactions [331]: (y = 3 or 6) (2.6)
ZnCI
~
Zn (OH )..-_=::l:=:....L_..:::::::L=-4--r-.r,,;:;:::I:::>:I...
I;
zn (OH I 2
ZnCI~
10
N'" v' U r "'_Nr o v C
N
i
V
-!V r
0
N
9
C
N
0
-
--
~ ~
v
-~ I
c
0 c
0 c
'"
N
N
FIGURE 2.7. Relative concentrations of basic zinc chloride species in aqueous solution as a function of pH and Cl- concentration. After Despic [1118].
ELECTROCHEMICAL THERMODYNAMICS AND KINETICS
25
(2.7) Figure 2.6 shows that in a chloride solution, the zinc chloride compound that can be formed varies with the pH of the solution [404]. The relative amounts of the various chloro-hydroxo complexes as a function of pH and concentration of Cl- are shown in Fig. 2.7 [1118]. In concentrated solutions, zinc seldom exists as a simple ion owing to the formation of complexes. 2.3. IONIC PROPERTIES The zinc ion has a radius of 0.74-0.83A and a hydration number of 10-12 [8, 327]. Owing to the electronic configuration of the zinc atom, zinc ions tend to form spJ -hybridized tetrahedrally coordinated complexes in solution [532]. The primary coordination number of the zinc ions is four, although variations have been shown to exist. The equilibrium constants for several ligands are listed in Table 2.2. Generally, complex formation follows the pattern: (2.8a) (2.8b) (2.8c) (L8d) The transport properties of ions in solution are characterized by the ionic diffusion coefficient and mobility. In Table 2.3 the ionic mobilities and diffusion coefficients of
TABLE 2.2. Complexing agent F CIBr-
1-
cw NH3 OW SCW Tartrate" SO~-
Overall Formation Constants for Various Zinc Complexes" Log formation constant"
PI 0.92 0.72 0.22 -0.47 2.32 6.31 1.57 2.30 2.08
Ih
P3
IJ4
-0.49 -0.10 -2.00 11.07 4.61 11.19 1.56 4.10
-0.19 -0.74 -0.74 26.05 6.97 14.31 1.51 5.55
-0.18 -1.00 -1.25 35.67 9.36 17.70 3.02 6.79
"Reprinted from Brodd and Leger [532], by courtesy of Marcel Dekker, Inc.
h/I,-fl. are the formation constants corresponding to the reactions given by Eqs. (2.Sa)-(2.Sd), respectively.
26
CHAPTER 2 TABLE 2.3.
Values of Equivalent Conductances (}.~) and Diffusion Coefficients (D) of Selected Ions at Infinite Dilution at 25°Ca .0
1..;
Cation H+ Li+ Na+ K+ NH; Ag+ TI+ Mg2+ Ca 2+ Sr2+ Ba 2+ Cu 2+ Zn 2+ La3+
Co(NH3)~+
I"i
(cm 2/s)
(Q-I'cm2/equiv) 349.8 38.69 50.11 73.52 73.4 61.92 74.7 53.06 59.50 59.46 63.64 54 53 69.5 102.3
.0
Djx 105
9.312 1.030 1.334 1.957 1.954 1.648 1.989 0.7063 0.7920 0.7914 0.8471 0.72 0.71 0.617 0.908
Anion OW
CI-
Br-
r NO] HC03" HCO;: CH 3CO Z SO~-
Fe(CN)~Fe(CN):10 CI0 BrO; HS0
4 4
4
(Q-I'cm2/equiv) 197.6 76.34 78.3 78.8 71.44 41.5 54.6 40.9 80 101 III 54.38 67.32 55.78 50
Dj x 105 (cm 2/s)
5.260 2.032 2.084 2.044 1.902 1.105 1.454 1.089 1.065 0.896 0.739 1.448 1.792 1.485 1.33
"From Newman [328J. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
zinc at infinite dilution are listed along with those of other ions [328]. In KOH solutions the diffusion coefficient of zincate ions varies with KOH concentration and temperature as shown in Fig. 2.8 [221]. The diffusion coefficient of Zn 2+ in 0.05M ZnCl 2 + 1M KCl solution has been found to be 0.89 x 10-5 cm 2/s [1120]. The mobility of zinc ions is generally lower than that of most uncomplexed anions, as a result of the much larger hydration shell of the zinc ions. For example, the transport number of zinc in Zn(Cl0 4)2 solution is 0.44 in a solution at infinite dilution and 0.335 12.----------------------------------, x
~
E 8
~
c Q)
'(3
~o o 4 c:
o
'00
:l
'I:
is
o o
L -_ _ _ _ _---'--_ _ _ _ _---"_ _ _ _ _ _- '
4
8
Molarity of KOH, M
12
FIGURE 2.8. Variation of diffusion coefficient of Zn(OH)~- (3 x 10-3M) in KOH solutions with temperature and KOH concentration. After McBreen and Cairns [22IJ.
27
ELECTROCHEMICAL THERMODYNAMICS AND KINETICS
100
E E .c 0
;3-
80 60
:~
ti
:l "0
40
c: 0
()
20 0
0
2
6
4
8
10
12
14
ZnCI 2 Concentration , M
FIGURE 2.9. Conductivity of aqueous zinc chloride solutions at 298 K. After Thomas and Fray [8971.
in a molar solution [327]. In general, for a salt with a cation transport number t+ [t+ = A)(},+ + },J J less than the anion transport number t_, t+ will decrease with increasing concentration. The conductivity of a zinc salt solution increases sharply with concentration at low concentrations but decreases at high concentrations due to increased viscosity. Figure 2.9 shows the changes in conductivity as a function of zinc chloride concentration [897]. 2.4. DOUBLE-LAYER PROPERTIES The values of the double-layer capacitance of a zinc electrode near its reversible potential in aqueous solutions range from 16 to 20/lF/cm2 [3, 180, 532, 782]. The double-layer capacitance is a function of potential and solution concentration. Figure 2.1 0 shows typical capacitance-potential curves of a zinc electrode in two different sodium sulfate solutions [1291]. Determination of capacitance at potentials anodic to the Zn rest potential is difficult because of the rapid anodic dissolution. The capacitance curves measured on single-crystal zinc electrodes are, in general, similar to those meas ured on polycrystalline electrodes [782]. Values of capacitance much larger than about 20 /lF/cm 2 are generally associated with faradaic and specific adsorption processes. The pH of a solution has a significant effect on the capacitance of a zinc electrode because a strong interaction between OH- and the electrode occurs at pH> 3.4 [810]. Capacitance values of 40 to 600 /lF/cm 2 have been observed in KOH solutions due to OW adsorption [12, 763, 786]. Also, OW adsorption appears to be associated with surface inhomogenei ty. It has been reported that in strong KOH solutions, OH- adsorption occurs on the zinc electrode but not on amalgamated zinc since the latter may have less surface inhomogeneity [478]. Sustained immersion in neutral or alkaline solutions may cause an increase of the electrode capacitance with time because of the gradual formation of a solid film on the surface [70 I] .
28
CHAPTER 2 30
25
~ 15
<5
10
1 0.5 N Na 2SO., pH=2.87 2 0.01 N Na2SO., pH=2.96
5
OL---~------~~-------L--
-1.5
-2
-1
______ ______ -0.5 ~
~
o
E, Vshe
FIGURE 2.10. Relationship between capacitance and potential for a zinc electrode in Na2S04 solutions. From Keifets and Krasikov [1291].
Measurements in O.lM NaCl04, NH4Cl0 4, and NH4Cl solutions of different pH values indicate that the specific adsorption of Cl- has a negligible effect on the capacitance of the zinc electrode in the potential range between -1.11 and -1.65 VSCE [810]. In acidic solutions, where OH- adsorption is absent, cation interactions may affect the capacitance values. The double-layer capacitance of a zinc electrode in an aerated 0.5M NaCl solution has been found to decrease with increasing thickness of the electrolyte layer as shown in Fig. 2.11 [608]. It is largely independent of the layer thickness in deaerated solution. The potential of zero charge (PZC) can be obtained from the minimum of the capacitance-voltage curve. Experimental investigations typically locate the PZC of zinc around-0.85 to -1 VSCE [180, 782, 532]. It has been found that the PZC varies with crystal
50,--------------------------------------,
Eo
40
~30 ai
o
ffi
.~ 20 Q.
ro
<.)
10
O~------~------~-------L------~
-2
-1
o
2
______~ 3
Layer thickness, mm (Log)
FIGURE 2.11. Interfacial capacitance C as a function of electrolyte layer thickness. After Keddam et al. [608].
29
ELECTROCHEMICAL THERMODYNAMICS AND KINETICS TABLE 2.4.
Electrochemical Techniques Used to Study Zinc Electrode Kinetics References
Technique Dynamic polarization Steady-state polarization Measurement of i-t transients Measurement of V-t transients Impedance measurements Rotating electrode
Corrosion
10, 33, 110, 106, III. 112 106,703,790 114,790
Dissolution! deposition
H+ and 02 reduction
22, 176,796,800, 10,110,797 1251 61,181, 700, 763, 113,683 786, 789 12,324
118,790
12,785,789,887, 7,9 903 113,700,701 61,683,763,786, 110,683, 1251 894 106,113,116,128 405,786,789, 106, 116, 128, 796,888 445,797, 1251
Passivation
526,702,794,890 526,904 18,681,904 18,127,886,889 526, 702 702
orientation [180]. Since the potential of a zinc electrode in a solution is nonnally negative to E pze , the electrical double layer is populated primarily with cations in the inner Helmholtz plane. 2.5. KINETICS OF ELEMENTAL REACTIONS The kinetics of the zinc electrochemical reaction processes such as dissolution, deposition, hydrogen evolution, oxygen reduction, passivation, and surface film formation, has been the subject of many electrochemical studies. Table 2.4 lists the different electrochemical techniques used in these studies.
2.5.1. Dissolution The dissolution of zinc has been extensively studied [12. 785, 786, 790]. Zinc dissolves readily near its equilibrium potential, with the formation of divalent zinc ions. In acidic solutions the dissolution product is simply Zn 2+. In alkaline solutions the predominant zinc species has been identified to be tetrahedral Zn(OH)~-. The charge density for dissolution of one monolayer of solid zinc can be calculated using the lattice constants of the zinc crystal to be 522 jJ.C/cm2 [3]. Table 2.5 lists the exchange current densities and Tafel slopes for zinc dissolution in various solutions. The exchange current density and Tafel slope are a function of many factors. Figures 2.12 and 2.13 show the exchange current density of zinc in KOH solutions as a function of KOH concentration and the concentration of zincate ions [887]. The exchange current density increases with KOH concentration and reaches a maximum at a KOH concentration of about 8M. It changes only slightly with zincate concentration. According to Dirkse [1132], the decrease in the exchange current density at high KOH concentrations shown in Fig. 2.12 is related to the ionic strength, which influences the mobility of the reacting species and modifies the exchange current density. Johnson et al. [181] studied the anodic dissolution of zinc in solutions containing CI-, Br-, 1-,
30
CHAPTER 2
TABLE 2.5 . Tafel Slopes and Exchange Current Densities for Zn Dissolution in Various Solutions b(mV)
Solution IMHCI 1M H ZS0 4
25 42
O.IM NaZS04' pH = 1-5 0.06MHCI
30 90 94.4
1M ZnCl z, pH
=2
1M NaCI , pH = 3.8
25
1M NaZS04' pH = 3.8 O.IM NaCI , pH = 5.3
38 30 20
O.OIM ZnCl z + 0.32M K ZS0 4
Reference(s) 14 14 0.0008 a
1.75
110 1I0 14,118 181
120
IMZnS04 O.IMKOH O.4MKOH
200
13
24
887 794
40 28
IMKOH IMKOH IMKOH I. 8M KOH + ZnO 7MKOH
445 15 1120
93
311 887
786
33-95 50 27
8-370 209
12 785
"Zn2+ concentration is between 10-5 and I0-4M.
SO~-, NO;, and acetate. They found that the Tafel slopes for zinc dissolution tend to be 60-85 m V/decade for solutions containing NO; and 15-40 mV/decade for those without NO;. Dissolution may occur on a bare surface or on a surface covered by a solid film. Solid films formed during Zn dissolution may have different compositions and various degrees
0 .5
• SS· C .., 40· C + 2S' C
0 .4
E ~ .-
· O· C
...
0 .3 0 .2 0 .1
KOH Concentration, M
FIGURE 2.12. Variation of the exchange current density for the zinc electrode in KOH solutions with temperature and KOH concentration. After Dirkse [1132] .
31
ELECTROCHEMICAL THERMODYNAMICS AND KINETICS
KOH Cone ,
0 ,3
03 M
E
+ 7M
$"
)!(
c
10 M
• 12 M
e
:;
u
Q)
FIGURE 2.13, Exchange current density in KOH solutions as a function of zinc concentration, Reprinted from Dirkse and Hampson 18871. with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom,
C>
C
'"
,J;;
" w
0,1 -
)(
°°
0.4
0,6
1.2
1.6
Molarity of Zincate. M
of compactness and thus may significantly affect the dissolution process. In solutions containing no species with which zinc can form insoluble salts (e.g., NaCI and Na 2S04 solutions), the zinc electrode maintains a bare surface during dissolution at a pH below 3.8 [110). At a pH value of 5.8 in 3 M NaCI or Na2S04 , according to Baugh [110], the zinc electrode is coated with an oxide film: this film may affect both the anodic and the cathodic process but is not passivating, In O.lM NaCI solution, the dissolution of zinc at low pH values (up to 3) is kinetically controlled whereas at high pH values (> II) it is controlled by diffusion of ZnO~- or HZn02 away from the surface [ 1161. The anodic dissolution of zinc may be greatly inhibited by the formation of solid surface films in solutions of carbonates [3, 127, 196], nitrates [591, phosphates [481, 800, 797), and chromate, molybdate, and tungstate [57,59,98,199). In phosphate solutions, according to De Pauli et al. [800), the dissolution of zinc occurs through a solid surface film which changes composition and structure with pH [481,800]. At pH > 12 a dissolution-precipitation mechanism operates whereas at pH < 12 a solid-phase process dominates. The presence of PO!- ions promotes zinc dissolution whereas the presence of HPO;- ions inhibits it. Kuznetsov and Podgornova [101) suggested that zinc dissolves to form [ZnH 2P0 4 at pH 4.5 and [Zn(HP0 4 hf- at pH 9.5. In saturated solutions of zinc sulfate, a thin pore-free salt layer forms, and zinc dissolution occurs through a direct reaction of zinc with sulfate ions to form solid zinc sulfate [888). The presence of Pb and Sn in alkaline solutions has been found to inhibit zinc dissolution as a result of the formation of a smooth inert surface film [27). Silicate was found to prevent the precipitation of a solid film on the zinc electrode in alkaline solutions and therefore maintains a high dissolution rate [7871. Davydov et al. [405] measured the limiting anodic current density of the zinc electrode in KOH solutions of different concentrations. They found that the anodic limiting current density, im , increases with KOH concentration up to about 6M, as shown in Fig. 2.14. The larger values of i for higher KOH concentrations are explained in terms of a higher solubility of the anodic reaction products. Because of the dissolution, the concentration of zincate ions at the electrode surface at the i can be much higher than the solubility in the bulk, and thus the solution near the surface is oversaturated. Owing
r
ll1
ll1
32
CHAPTER 2
.------------------------------------, 10
:i:
o c
2
1§ 'E
'"u C
o u
'"u .!'!
:;
(j)
OL---------------~~--------------~
o
5
Concentration of KOH
FIGURE 2.14. Limiting current density, i lll • of dissolution of a zinc disk electrode (l11 = 750 rpm) in KOH and calculated concentrations of hydroxyl ions near the surface of the electrode, c?W, as a function of bulk solution concentration, c~H-. After Davydov et al. [405].
to the formation of Zn(OH)~- ions, which consumes OW ions, the concentration of OW at the surface is considerably lower than that in the bulk, as shown in Fig. 2.14. The surface area, morphology, and other properties of the electrode surface may change as a result of dissolution. For examples, Yamashita [112] found that the surface of Zn becomes very smooth after 10 days' immersion in 1M ZnS04 ; Powers reported that anodic dissolution starts preferentially at grain boundaries in 7M KOH solution [27] and that hexagonal pits are produced on single crystals during low-potential dissolution in 7M KOH solution [29]; and Menzies et al. [488] observed that the anodic dissolution of Zn in sulfamic acid-formamide solution leads to polishing of the surface. 2.5.1.1. Dissolution Efficiency. In most solutions, the dissolution of zinc occurs with a valency of2 and with almost 100% faradaic efficiency [181,532]. However, the apparent valence may be lower than 2 in certain solutions. Figure 2.15 shows that in the absence of NO), the anodic dissolution of zinc occurs with a valency of 2 [181]. In the presence of NO), the dissolution is accompanied by the reduction of NOj', which leads to surface disintegration. Because the disintegration is outside the faradaic circuit, it results in an apparent valence lower than 2. There are two possible explanations for the lower apparent valence values: (1) stepwise oxidation: Zn ~ Zn+ + e- (at anode) followed by Zn+ + oxidant = Zn 2+ + reductant (in solution); and (2) anodic disintegration, generating fine metallic zinc particles. These particles are very active and are oxidized rapidly by the oxidant in the solution [182]. Nonfaradaic dissolution is also found in NaBr03 [1119], in NaCIO z [34], and in ZnBrz solutions containing Brz [706]. 2.5.1.2. Mechanism. The mechanism of dissolution is different for acidic and alkaline solutions and for complexing and noncomplexing solutions. In acidic and noncomplexing neutral solutions, the ZnJZn z+ electrode reaction appears to occur in two consecutive charge-transfer steps [110, 786, 789]: (2.9)
33
ELECTROCHEMICAL THERMODYNAMICS AND KINETICS
.. Q)
g Q)
1.9
+ 0 .10 /0.30 M
c:
0. 0.
'" ~
• 0 .0/0.333 M ... 0.05/0.317 M
Iii >
~
KN0 3 /K 2 S0 4
Concentrations
• 0.30 / 0.2 33 M X O.SO /0.167 M
1.8
• 1.7
• 0.70 / 0.10 M " 1.0
I 0.0
M
X 2.0/0.0 M
1.6 L---------------------~------~
o
0 .04
0.08
i, A I em"
FIGURE 2. I 5. Apparent valency of zinc dissolving anodically in different KNO r K2S04 solutions at 25°C. Reprinted from Johnson el al. [18 I j, with kind permission from Elsevier Science Ltd, The Boulevard. Langford Lane, Kidlington OXS 1GB, United Kingdom.
(r.d.s. )
(2.10)
with the reaction in Eq. (2.1 0) as the rate-determining step (r.d.s.) and Zn+ as an adsorbed and/or a solution-soluble intermediate. This dissolution mechanism gives a Tafel slope of 2.3 x 2RTl3F (40 mV).1t can be seen in Table 2.5 that the dissolution in many solutions may follow this simple mechanism. At low overpotentials, the concentration of adsorbed intermediate is small, and the reaction can be treated as a pseudo-one-step reaction. However, at higher potentials the contribution from the adsorbed species becomes significant [786]. This contribution has been measured by many investigators on solid zinc electrodes as well as on amalgamated zinc electrodes [532]. This simple reaction scheme is also reported to occur in other electrolytes where zinc complexes form. Hurlen and Fischer [789J found that the Zn+/Zn 2+ charge-transfer step in concentrated acidic chloride solutions occurs between the couple ZnCI 2 (Hpr/ZnCI 2(H 20)" but species with one or no chloride ligand take over as the main electroactive species at chloride concentrations below 1M. The reaction scheme represented by Eqs. (2.9) and (2.10) was also proposed by Armstrong and Bell [786] to describe the dissolution of zinc in alkaline solutions, in which hydroxo-zinc complexes generally form. However, the dissolutIOn mechanism becomes more complicated when complexes are formed because more reaction steps are required to account for the formation of these complexes. Figure 2.16 shows that the rate-determining step for zinc dissolution depends on the type of anions in the electrolyte. The dissolution of zinc in NaCI solution, although following Eqs. (2.9) and (2.10), is diffusion-limited, probably because of the diffusion of chi oro-zinc species (e.g., ZnCI~-) away from the electrode surface [110]. Armstrong and Bell [786] found that the concentration of adsorbed intermediate species and the diffusion of zincate ions away from the surface are important parts of the dissolution process in 1M KOH since the dissolution current of zinc depends on the rotation rate of the electrode.
34
CHAPTER 2 0 .13 r - - - - -- - - - - - - -- - - -- - - , -NaCI0 4
0.12
.
10<.>
Na 2S0 4 ..J..NaCI
0 .11
..: 0.1
~========:::;;::;:¢::!:::::::::::::::!:==::::::1:==r=::::I===;
0 .09
0 . 08L---------------~----------------~----~
o
0.05
0.1
W ·" i SH2
FIGURE 2.16. Rotation speed dependence of the anodic dissolution current for zinc in different molar solutions at pH 3.0. Reprinted from Baugh rIlOl. with kind permission from Elsevier Science Ltd. The Boulevard. Langford Lane. Kidlington OX5 1GB, United Kingdom.
Johnson and co-workers [181,182] proposed the reactions in Eqs. (2.11)-(2.13) to explain the dissolution mechanism in neutral solutions containing various anion species that are not reducible by zinc. In this scheme, the desorption of ZnO ads is the ratedetermining step and the Tafel slope is 2.3RT!2F (30 m V). When the solutions contain also NO) ions, the Tafel slope appears to be 2.3RTIF (60mV), which is associated with a cathodic reduction from NO) to NO;. (2.11 ) (2.12) (2.13) Cachet and Wiart [849] proposed a reaction scheme [Eqs. (2.14)-(2.16)] for the dissolution of zinc in de aerated ZnCl 2 and NH 4 CI solutions in which zinc complexes form. The dissolution involves two parallel paths. The major path, Eq. (2.15), is catalyzed by Zn;ds' The minor path, Eq. (2.14), is much more dependent on the diffusion of the chi oro-zinc species than the major one. Both reaction paths are stimulated by chloride anions. The formation of ZnOHadS' is a side reaction and is caused by the chemical oxidation of zinc by the electrolyte. Deslouis et al. [700] confirmed the validity of such a reaction scheme for zinc dissolution in aerated sulfate solutions.
I~ Zn
Zn~ds
Znads 2+ ----+ Zn 2 + + sol e
(2.14 )
Zn~;l + Zn~ds
(2.15)
+ e-
I+Zn
L--=:...
+ 2e-
(2.16)
35
ELECTROCHEMICAL THERMODYNAMICS AND KINETICS
In alkaline zincate solutions, Cachet et ai. [763, 1195] measured four loops on the complex-plane impedance plots with decreasing frequency: (i) a capacitance loop generally highly depressed in connection with the current penetration within pores of a surface film; (ii) an inductive loop corresponding to the presence of a monovalent intermediate Zn+ in the reactive interface; (iii) a capacitive loop resulting from the precipitation and escape of Zn 2+ ions by diffusion from the pore bases; and (iv) an inductive loop consequent on the slow decrease of the pore length with increasing anodic polarization. They postulated that there are at least four adsorbed species at the electrode surface and that the dissolution of zinc does not occur by a series reaction. Some adsorbed species are formed and consumed by slow reactions taking place in parallel with the main reaction path, similar to the reactions in Eqs. (2.14) and (2.16). The active dissolution takes place essentially at the base of pores in a layer of oxidation products whose degradation by the anodic current can be depicted as a slowly decreasing layer thickness with increasing anodic polarization. The zincate ions formed at the pore bases partially precipitate inside the pores but mostly escape from the pores by a diffusion process whose rate increases with decreasing pore length. In addition, these authors suggested that the much faster charge transfer observed for dissolution than for deposition is due to a drastic change III the kinetics occurring within a small potential domain passing through the equilibrium potential. Bockris et ai. [12] proposed a multistep reaction mechanism [Eqs. (2.17)-(2.20)] for the dissolution of zinc in alkaline solutions, with the rate-determining step being reaction in Eq. (2.19). A similar scheme was proposed by Muralidharan and Rajagopalan [790], who, however, pointed out that the mechanism described by Eqs. (2.17)-(2.20) is only valid under transient conditions. Under steady-state conditions, the rate-determining step becomes the diffusion of zincate away from the surface rather than the charge-transfer reaction (Eq. 2.19). Furthermore, under steady-conditions, zinc redeposition occurs owing to the slow diffusion of zincate away from the surface. Zn + ow ;:::::::= Zn(OH) + e-
(2.17) (2.18)
Zn(OH) + OI-t;:::::::= Zn(OH);: Zn(OH);: + OW
~
Zn(OH):; + e-
Zn(OH)3 + OH-;:::::::= Zn(OH)~-
(r.d.s.)
(2.19) (2.20)
According to the reaction scheme proposed by Hampson et ai. [785] [Eqs. (2.21)(2.24)], the adatom surface diffusion step, Eq. (2.21), is the rate-determining step in alkaline solutions. It is considered that the adsorbed atoms are stabilized by OH-, which is extensively adsorbed at the zinc electrode. This mechanism implies that the reaction is insensitive to the zinc ate concentration in the solution. The charge-transfer coefficients appear to be about 0.9 for the anodic reaction and about 0.1 for the cathodic reaction, indicating that the anodic and cathodic reactions are not the same. This mechanism can be used to explain the different dissolution rates observed on zinc electrodes with different crystallographic orientations, which have different densities of kink sites [893].
36
CHAPTER 2
Zn kink + OW ~
Zn(OH)~ds
(r.d.s~
(2.21 ) (2.22) (2.23) (2.24)
The exchange current density in concentrated KOH solutions (Fig. 2.13) depends only slightly on the zincate ion concentration. This seems to be in agreement with either the reaction mechanism given by Eqs. (2.17)-(2.20) or that given by Eqs. (2.21 )-(2.24). The decrease of the exchange current density, io, with increasing KOH concentration from 7M to 12M can be attributed to (a) the formation of different anodic dissolution products in the electrochemical reaction at KOH concentrations below and above 8M and (b) a shortage of water molecules for the hydration process when the KOH concentration is high. It appears that zinc dissolution can follow different mechanisms depending on the electrolyte and experimental conditions. The differences between the various proposed mechanisms arise essentially from the differences in the final dissolution products and their properties, which include the type and number of intermediates, their mobility, and their state of adsorption and solvation. Each mechanism can be characterized by a set of distinctive kinetic parameters such as the Tafel slope, reaction orders, etc. Table 2.6 lists the theoretical values for several different reaction schemes, and Table 2.7 compares the values obtained from different studies [790]. 2.5.2. Deposition In most cases, zinc deposition plays a negligible role in zinc corrosion for two reasons: (a) corrosion generally occurs at a potential anodic to the reversible potential of zinc, where the deposition is insignificant compared to the dissolution, and (b) corrosion is usually encountered in solutions containing very little ionic zinc. Thus, zinc deposition is only discussed here in a rather general fashion. However, much of the information presented above (Section 2.5.1) regarding dissolution can be applied also to deposition. More information on the subject of zinc deposition can be found in the literature [61, 683, 1158]. Zinc deposition occurs at potentials negative to the Zn reversible potential. In aqueous solution, both zinc deposition and hydrogen evolution may occur at potentials negative to the zinc reversible potential. Thermodynamically, hydrogen evolution is a more favorable reaction at a cathodic potential because of its more positive reversible potential. However, the cathodic reactions on zinc near the zinc reversible potential are dominated by Zn deposition when the zinc concentration is higher than 10-4M. This is attributed to the small io and a large Tafel slope for the hydrogen reaction on zinc. For example, in alkaline solutions at potentials at which substantial zinc deposition occurs, the hydrogen evolution current is very small, less than 10 j1A/cm 2 [12].
0.5
0
0
Znad + 20H- ~ Zn(OHh + 2eZn(OHh + 20W ~ Zn(OH)~-
Znad + OH- ~ ZnOH ad + eZnOHad + OH- ~ Zn(OHh + e-
0.75
o 40
120
120
60
Anodic
120
40
40
60
Cathodic
Tafel slope (mV/decade)
3
2
Anodic
-3
-3
-2
Cathodic
Reaction order with respect to OW
o
o
0
0
Anodic
Cathodic
Reaction order with respect to zincate
"Reprinted from Muralidharan and Rajagopalan [7901. with kind permission from Elsevier Science-NL. Sara Burgerhartstraat 25. 1055 KV Amsterdam. The Netherlands.
mr mr
Zn(OH) + OW ~ Zn(OH); Zn(OH)2" + ~ Zn(OH)3" + eZn(OH)3" + ~ Zn(OH)~-
Zn(OH) + e-
Zn + OW
~
0.25
o
Zn + OW ~ ZnOlf,;"d ZnOlf,;"d ~ ZnOHad + e ZnOHad + OH- ~ Zn(OHh + eZn(OHh + 20W ~ Zn(OH)~-
Zn(OHh + 20W ~ Zn(OH)~-
d log CZn(OH)~-
dlogaoH-
Mechanism
0.25
dlog io
dlogi o
TABLE 2.6. Possible Mechanisms for Dissolution/Deposition of Zinca
~
~ Z
" ~
» z
CI:l
n
~
z »
~
~
i
~
~
::r:
\.l
~
38
CHAPTER 2
TABLE 2.7.
Comparison of the Results from Different Studies on the Mechanism of DissolutionlDeposition of Zinc in Alkaline Solutions" Muralidharan and Rajagopaland
Parameter [
d log
io]
dpH
Hampson et al. b Bockris et al. e
Steady state
Current step
Potential step
0.2
0.14
0.10
0.66
0.2
o
0.67
0.65
0.3
0.27
49± 12
50± 10
90±20
90±20
113 ± 30
175 ± 20
200±20
200± 20
0.72 to 1.05
0.53 to 0.7
-0.04 to -0.21
0.0
-0.75 to -1.96
0.7 to 1.3
0.76 to 0.8
0.31 to 0.53
0.06
0.1
0.23 to 0.39
0.04 to 0.07
2.56
0.5
0.91
0.34 to 0.47
2-
cZn(OH)4
dlogio d log CZn(OH)~-
Anodic Tafel slope (mY/decade) Cathodic Tafel slope (mY/decade) Cathodic reaction order w.r.t. zincate Cathodic reaction order w.r.t. OHAnodic reaction order w.r. t. zinc ate Anodic reaction order w.r.t. OH-
62 ± 10e 320f 55 ± 8e 280±40f
to
1.06
"Reprinted from Muralidharan and Rajagopalan [790J, with kind permission from Elsevier Science-NL. Sara Burgerhartstraat 25, 1055 KY Amsterdam, The Netherlands. h
Ref. 785.
c
Ref. 12.
d
Ref. 790.
c
Low overpotential.
! High overpotential.
In solutions in which Zn 2+ exists as a complex, the electrode reaction must begin with the formation of a tetrahedral solution complex, which then undergoes consecutive dissociations until zinc metal forms, which requires two charge-transfer steps [532], e.g.,
The reaction mechanisms for deposition of zinc near the reversible potential are generally considered to be the opposite of those proposed for Zn dissolution [12, 786, 789, 790, 894]. In certain cases the rate-determining step for anodic dissolution is considered to be different from that for cathodic deposition [763, 785, 887]. The Tafel slopes for deposition in alkaline solutions are in general larger than those for dissolution, i.e., 120-300 m V versus 20-40 mV [12, 785, 790]. Kim and lome [1120] found that the deposition of zinc in ZnCl 2 solution is the reverse of the two-step reaction given by Eqs. (2.9) and (2.10), with a reaction order of about one with respect to the Zn 2+ concentration. Epelboin et al. [61] postulated that deposition of zinc in acidic sulfate, LecIanche cell, and alkaline zincate solutions depends on the presence of Hads , Zn;ds, and other adsorbed anions and may have an autocatalytic
ELECTROCHEMICAL THERMODYNAMICS AND KINETICS
39
step, Zn 2+ + Zn;ds + e- = 2Zn;ds, in which a monovalent intermediate is involved (the adsorption of Hads acting primarily as an inhibitor for zinc deposition). On the other hand, according to Cachet and Wiart [683], zinc deposition in highly acidic sulfate electrolytes is associated with an inhibition of hydrogen evolution, possibly due to the formation of the intermediate Zn;ds. The presence of Ne+ ions has been found to destabilize the zinc deposition process in sulfate solutions by stimulating hydrogen evolution [1030, 1251]. The morphology of the surface deposit varies with overpotential, current density, and Zn2+concentration [62,221,324]. Smooth, dark gray porous, or dendritic Zn deposits are formed as the overpotential changes from low to high values [62]. In alkaline solutions, smooth deposits occur at low overpotentials with vigorous stirring; dark gray porous deposits occur at low overpotentials «70 mV); and dendritic deposits occur at high overpotentials (>75 m V) [221]. The transition from moss to dendrites corresponds to the onset of mass-transport control since dendritic growth is a diffusion-controlled process and is influenced by flow of solution, especially at lower concentrations. Dendrites initiate at places where the local current density is high [1120]. They originate from the tips of pyramids arising as a result of rotation of a screw dislocation. As a pyramid grows, its radius of curvature decreases, and eventually the tip becomes a point for a spherical diffusion [221]. Cathodic potential oscillations during zinc deposition in alkaline solutions containing zinc ions have been found to occur in the current range of 0.5 mA/cm 2 to 0.17 A/cm 2 [132]. The phenomenon has been explained as a result of the balancing effect between deposition and diffusion of zinc ions and the competing effect of the zinc deposition and hydrogen evolution reactions at the electrode surface. Impurities (Ni, Co, Cu, Cd, Sb, Ge, As, Bi) in electrowinning solutions induce instability in zinc deposition and alter the deposit morphology [1253].
2.5.3. Hydrogen Evolution 2.5.3.1. Potential of Hydrogen Electrode. The standard potential of the hydrogen electrode, defined by the reaction in Eq. (2.25), is conventionally taken as E~ = 0 [11]. The reversible hydrogen potential in aqueous solutions depends on the hydrogen gas pressure, PH,' and the activity of hydrogen ions, a H\ as expressed by Eq. (2.26).
W+e-~~H2(gs) EH =
E~
- RT I2FlogpH, + RT IF-log aH>
(2.25) (2.26)
In concentrated alkaline solutions the reversible potential at 25°C can be calculated from the equation
EH = 0.0296 log aH,O where aH,o is the activity of water in the solution [7]. The reversible hydrogen potential cannot be measured on a zinc electrode in aqueous solutions owing to the active nature of zinc, which has a reversible potential much lower than that of the hydrogen electrode. Hydrogen gas has a very low solubility in water; under a hydrogen pressure of 1 atm, aqueous solutions contain approximately 0.8 x 1O-3MH 2 [II]. The solubility of hydrogen is greatly decreased in concentrated electrolytes because of salting-out effects, as shown
40
CHAPTER 2 20 • H2 SO4 " KOH
::: 15
.9-
E'"u
;10 :0 ::> (5
(J)
5
I
0
0
8
6
4
2
12
10
16
14
Concentration (N)
FIGURE 2.17. Hydrogen solubility as a function of electrolyte normality at 30°C. After Riietschi [1146].
in Fig. 2.17 [1146]. Ions with large hydration shells are particularly effective in saltingout. Hydrogen solubility is, therefore, lower in KOH than in H2S04 or NH4 Cl. The diffusion coefficient of hydrogen, detennined from limiting currents to a rotating platinum disk electrode, greatly decreases with increasing concentration of the electrolytes as shown in Fig. 2.18 [1146]. 2.5.3.2. Exchange Current Density and Tafel Slope. The hydrogen overpotential, Yf, is related to the exchange current density, io, and Tafel slope, b, through the Tafel equation: Yf
=b log ilio,
(2.27)
b=RTlaF
where a is the charge-transfer coefficient. Figure 2.19 shows the Tafel plots, measured by Lee [7], on a zinc electrode in 6NKOH solution. A clear linear relation between current
5.------------------------------------, 4
~3 ()
'" CIl
o
oL-__ ____ o 2 4 ~
L __ _
~
6
____
~
8
__
~
____
10
~
12
__
~L-
14
__
~
16
Concentrat ion (N)
FIGURE 2.18. Dependence of the hydrogen diffusion coefficient on electrolyte normality at 30°C. After Riietschi [1146].
41
ELECTROCHEMICAL THERMODYNAMICS AND KINETICS
L10
1.00 ~ 0
>
0.90
.J
< ;::: 0.80 z w
~a::
070
w
> 0
0.60 0.50 10
0.1
0 .01
CURRENT DENSITY (ma fern l
1000
100 )
FIGURE 2.19. Hydrogen overpotential on Zn in 6N KOH. After Lee [71.
and overpotential is observed. The values of the exchange current density and Tafel slope determined in various solutions are listed in Table 2.8. As can be seen in Table 2.8, in most cases the Tafel slope for hydrogen evolution on the zinc electrode has a value of about 120 m V/decade, which is also the value found for
TABLE 2.8.
Tafel Slopes and Exchange Current Densities for Hydrogen Reduction on Zinc in Aqueous Solutions
Solution INHCI IN H 2 S04 INH 2 S04 H 2S04, 0.05-2N
O.IM NazS04' pH = 1-8 1M NaCI , pH = 5.8
= 5.8 =6 1M NH4c!' pH = 6 1M NH 4 CI, pH
1M (NH 4lzS0 4, pH IN LiOH IN NaOH INKOH 5NKOH 6NKOH 9NKOH
9NKOH
b(mVj
232 124 120 120 120 200 120 125 174 150 120 140 160 124 145 124
log io
-10.8 -10 to -10.8 -8.9
Reference 14 14 6 II 445 110 III 33 33 311 311 311 311
-9.1 -8.2
7 311 7
o
01
. _0
~
E
N
-10
-5
Cr
Fe
Ni
II ,
Zn Ge
30 I
V Mn Co Cu Go As
I , , , ,
25 <~
Zr
I
I
45 , , I
,
I
50 I
Nb Tc Rh AQ In Sb Me Ru Pd Cd Sn Te
I,
40
I 2 5p
I 2 34
To
75
t
/~.
80
N 6p
W
Re Ir Au TI 8i Os PI HQ PI> Po
1 23
222222112222
23456791010101010
HI
FIGURE 2.20. Values of hydrogen exchange current density, log iQ, for various metals in acid solutions. After Kita [6).
4p
I 2 45 2 2 I 2 2 2 2 I 2 2 2
,
,,
,
.:
,~
H-."'T""'1"""T'"+'-,.,-r-T""T-+-......--,-,,-
55211211012222265
Ti
•
Pi v. I
(I)
(b
° l .,.i ~ ,l
1
4d 2 4 5 5 7810101010101010 5d
Si
i
Alomic number
I
3p
AI
51 '
~
fA f •
..
.;.'
.,.l-'\' . ..
3s 2 2 2 3d 2 3 5 5 6 7 8 1010 10 10
MQ
13
~
!
~I
./ .
\ ..
~
t
~\:
·~r I .
I.
-10
-5
E
.2
01
.!?
~
N
(')
N
;>:l
ITI
-i
» "'0
:c
'"
...
ELECTROCHEMICAL THERMODYNAMICS AND KINETICS
43
hydrogen reduction on many metal electrodes. The presence of chloride ions seems to result in larger b values. The high overpotential for hydrogen reduction on zinc, compared to that on other metals, is mainly due to the low exchange current density as shown in Fig. 2.20. According to Brodd and Leger [532], the low values of the hydrogen exchange current density are a result of the weak interaction between zinc and hydrogen. Because of this weak interaction, zinc is essentially free of a chemisorbed layer of atomic hydrogen. The periodic variation of io with the atomic number for each long period (Fig. 2.20) indicates that the reaction kinetics for hydrogen evolution are essentially determined by the intrinsic properties of the electrode materials. The variations of io and b with changes in the experimental conditions reflect mainly the influence of surface condition, solution chemistry, etc. According to Brodd and Leger [532], who collected data from several studies, the exchange current density of the hydrogen reaction on zinc is largely independent of pH except in concentrated acid or alkaline solutions. as shown in Figs. 2.21 and 2.22 r116,445,532]. 2.5.3.3. Reaction Mechanism. The hydrogen evolution reaction in acidic solution can be expressed by the following equation: (2.28) whereas the following equation describes the reaction in alkaline solution: (2.29)
..
-11
e
e
e
Ref. 135 Ref. 136 • Ref. 137 'Y
-10 0
O"l
0
-'
_O.s'
e--e
'"
-0.4
.
e--I
.
e-
I
0.01
0.1
10
Acid concentration (N)
FIGURE 2.21. Values of exchange current density, io, and charge-transfer coefficient, a, for hydrogen overpotential in sulfuric acid as a function of concentration. The reference numbers in the figure are from Bradd and Leger [532]. Reprinted by courtesy of Marcel Dekker, Inc.
44
CHAPTER 2
---11,----------------------,
-10 • Ref. 138 ~ Ref. 139 _ Ref. 140
Ol
o
-'
-9
--0.5
•
.. ... ~
-.-.--.-------=-.:..-~--
- - 0.4':--:-:---------::'-:---------'::-----=-=--:--:-':-_-_ _ _, 0.01 0.1 10 Alkaline concentration (N)
FIGURE 2.22. Values of io and a for hydrogen overpotential in alkaline solutions as a function of concentration. The reference numbers in the figure are from Brodd and Leger [532]. Reprinted by courtesy of Marcel Dekker, Inc.
The consistency of the a value in acidic and alkaline solutions (Figs. 2.21 and 2.22) suggests that the electric field has the same effect on the electron-transfer reactions involving H30+ and H20. A Tafel slope of 120 mV/decade and a charge-transfer coefficient of 0.5 may indicate that the discharge reaction is the rate-determining step in accordance with the elementary steps generally proposed for hydrogen evolution [11, 532]: Acid
(2.30)
Alkaline
(2.31 ) (2.32)
The rate equations for hydrogen evolution on a zinc electrode can be expressed by the Butler-Volmer equation by neglecting the back reactions because on zinc the reaction occurs at a potential far from its equilibrium value [7]: Acid
(2.33)
Alkaline
(2.34)
where l/J is the potential between the metal electrode and the bulk of the solution, and Kj is the rate constant for the forward reaction. The Nernst equation for the reactions in Eqs. (2.28) and (2.29) can be expressed by Eqs. (2.35) and (2.36), respectively.
ELECTROCHEMICAL THERMODYNAMICS AND KINETICS
45
Acid
(2.35)
Alkaline
(2.36)
The overpotentiall] is I]
= ¢J - ¢Jeq
Thus,
2RT I Fin i
(2.37)
- 2RT IFln i
(2.38)
Acid
I]
= const. + RT IF In( aH,o+aH,o) -
Alkaline
I]
= const. + RT IFln(aH,oao H)
assuming a = 0.5. At a constant temperature and for a given electrolyte, Eqs. (2.37) and (2.38) yield the Tafel equation, i.e., Eq. (2.27). A plot of I] versus log i at 25°C for Eqs. (2.37) and (2.38) displays a slope of 120 mV/decade. Catino [311] postulated that in alkaline solutions of concentrations lower than 5N, water reduction is controlled by the recombination of adsorbed hydrogen atoms lEq. (2.32) rather than Eq. (2.31 )j. In more concentrated alkaline solutions (7-9N), hydrogen reduction is promoted by the alkali metal cation, which acts as an electron bridge at low overpotentials: (2.39) (r.d.s.)
(2.40)
Catino suggested that hydrogen reduction at high overpotentials follows the alkali metal penetration mechanism described by Eqs. (2.41) and (2.42). In very concentrated solutions, the alkali metal cations may become partially dehydrated in the inner portion of the double layer and thus force the reactant water molecule to move away from the zinc surface [532]. (r.d.s.)
(2.41 ) (2.42)
Different processes may be involved in different electrolytes at different overpotentials. In solutions with pH values between 3 and 12, hydrogen evolution may be controlled by nonactivation steps. For example, diffusion of protons to the surface has been found to be the rate-determining step at low overpotentials in slightly acid solutions in the pH range 3.5-6 [116, 110,445]. In near-neutral and slightly alkaline solutions, hydrogen reduction is found to be affected by the formation of a surface oxide film [110, 116,445]. Powers [27, 29] found that formation of an anodic film catalyzes hydrogen evolution in alkaline solution. Baugh [110] reported that in acidic solution of pH 3.8, hydrogen evolution occurs via proton reduction at low overpotentials and water reduction at high
46
CHAPTER 2 1.000 r---------------------------------~--_.
100
• NaCI0 4
10
... Na 25°4 + NaCI
1~----~----~------~----~--~~----~
-1.6
- 1.5
-1.4
-1. 2
-1. 1
-1
FIGURE 2.23. Effect of anions on the polarization curves for zinc in molar solutions at pH 3.8. Reprinted from Baugh [III], with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom.
overpotentials. Figure 2.23 shows the polarization curves measured in three different electrolytes, indicating the effect of electrolyte composition on hydrogen evolution. 2.5.3.4. Effect of Solution and Electrode Composition. The overpotential for hydrogen reduction is strongly affected by the ions present in the solution owing to their specific adsorption on the zinc electrodes and their interaction with water [II]. As shown in Fig. 2.23, hydrogen evolution is affected by the presence of anions, especially at low overpotentials. NH; in weak acid solutions has been found to affect hydrogen reduction by direct reduction (Eq. 2.43) and by changing the concentration of H30+ near the surface (Eq. 2.44) [IIIJ. (2.43) (2.44) The presence of Fe 2+, Cu 2+, N?+, As 3+, Sn 2+, and Sb3+ ions promotes hydrogen evolution on zinc [10, 115, 683, 1251, 1252]. These elements have more positive reversible potentials and lower hydrogen overpotentials than zinc, and the precipitation of these elements on a zinc surface causes an increase of rate in hydrogen evolution. On the other hand, Pb 2+ ions inhibit hydrogen evolution as shown in Fig. 2.24 [10, 115]. H4 PO; has a catalytic effect on hydrogen reduction [943]. The presence of oxyanions through formation of cathodic films has been found to affect hydrogen reduction [199, 597J. The presence of Zn 2+ in acid solutions and Zn(OH)~- ions in alkaline solutions generally results in a reduction of the hydrogen reduction rate [10, 683, 1252]. Figure 2.25 shows the effect of Zn 2+ concentration on the hydrogen evolution rate [1252]. The rise in current at potentials more negative than -1.04 V seE is due to zinc deposition. The decrease in the rate of hydrogen reduction before the current rise at more negative potentials is attributed to the adsorption of zinc ions at potentials less cathodic than
47
ELECTROCHEMICAL THERMODYNAMICS AND KINETICS -0.4 -0.5 -0.6
"1
"i. ~
-0.7
• Cu" ... Sn"
C Q)
-0.8
a.
-0.9
+ base curve • PO"
(5
-1 -1 .1 0 .01
0.1
10
100
(rnA I em')
FIGURE 2.24. Polarization curves for zinc in 6N KOH at 25°C in the presence of 1Q-3 M Zn"+. Cu 2+. Sn 2+. or Pb 2+. After Mansfeld and Gilman [IOj.
required for zinc deposition [683]. In alkaline solutions, the inhibition of hydrogen evolution with the addition of ZnO is due to the formation of zincate ions. Zincate ions lower the activity of water according to the following reaction [10]: (2.45)
Polycrystalline and single-crystal zinc surfaces exhibit nearly the same hydrogen overpotential characteristics [532]. However, impurities present in the zinc electrode may change the kinetics of hydrogen reduction [9, 11,438,891]. In acid solutions the presence of a trace amount of lead in zinc results in a significant decrease of the hydrogen evolution [1250], Lee [9, 1121] reported (Table 2.9) that Hg in Zn decreases the exchange current
[Zn"]. gil
o
100
I
E
5 20 60
()
~
.§. ~
100
10
'0;
c
Q)
"0
C
~
:l
U
FIGURE 2.25. Effect of Zn 2+ concentration on voltammograms in 200-gll H 2 S04 solutions. After Wang et al. [1252].
0.1 L-_ _--'--_ _ _- - l L -_ _ _....J...._ _ _----J -0.8 -1 -0.9 -1 .1
Potential (V seE)
48
CHAPTER 2
TABLE 2.9. Tafel Slopes, Charge-Transfer Coefficients. and Exchange Current Densities for the Hydrogen Evolution Reaction on Surfaces of Various Zn Alloys in 9N KOH" Surface
Tafel slope
a
io (Alcm 2 )
Zn Zn-2%Hg Zn-4%Hg Zn-8% Hg Zn-4%Cd Zn-8%Cd Zn-0.2%Pb Zn-0.8% Pb Zn-2%Pb Zn-0.05%Mn Zn-0.5%Mn Zn-2% Hg-0.2% Pb Zn-0.8% Pb-0.05% Fe
0.124 0.116 0.098 0.086 0.158 0.154 0.137 0.134 0.172
0.48 0.51 0.60 0.69 0.37 0.37 0.43 0.44 0.34 0.43 0.42 0.48 0.48
1.5 x 10-9 2.7 x 10- 10 8x 10- 11 6 x 10- 12 7 x 10-8 1.5 x 10-8 2 x 10-9 1.3 x 10-9
0.138 0.140 0.125 0.125
6.2 x 10-8 5.1 x 10-8 7.5 x 10-8 6 x 10- 10 9 x 10- 10
"Ref. 9.
density io for hydrogen reduction in alkaline solutions and reduces the Tafel slope. Pb causes a significant increase in the Tafel slope while having a varying effect on i o. Mn and Cd increase the Tafel slope slightly and increase io significantly. According to Lee [9, II], the small amounts of impurities in zinc lower the hydrogen overpotential but do not change the mechanism ofthe hydrogen evolution processes in alkaline solutions. Alloying with noble metals generally facilitates hydrogen evolution.
2.5.4. Oxygen Reduction 2.5.4.1. Solubility and Diffusivity. The solubility of oxygen decreases significantly with increasing temperature as shown in Table 2.10. It is also affected by dissolved salts in water. In water with dissolved salts up to 1000 ppm, the oxygen solubility is basically constant [558], but it decreases significantly in concentrated solutions as shown in Fig. 2.26. The process of dissolving oxygen gas in water is, however, not efficient. The water surface acts as a barrier to the incoming oxygen molecules. A water surface at 25°C admits only one in 6,000,000 impinging molecules of oxygen [403]. The diffusivity of O 2 in water at 25°C is about 1.9 x 10-5 cm2/s [496]. It decreases with the amount of salts dissolved in the water. For example, it is 1.24 x 10-5 cm2/s in 0.5MNazS04 solution [113]. The diffusivity of Oz in KOH solutions as a function ofKOH concentration is shown in Fig. 2.27 [496]. 2.5.4.2. Reaction Kinetics. Oxygen reduction is, apart from hydrogen evolution, the most important cathodic reaction in the metal corrosion process. The electrode reaction for oxygen reduction in acid solutions is [1, 1122, 1123]
Eo =1.229 V SHE
(2.46)
ELECTROCHEMICAL THERMODYNAMICS AND KINETICS
TABLE 2.10.
49
Solubilities of Air and Oxygen in Water" Air Percent oxygen in dis sol ved air
a (10 3)"
Temperature (0C)
34.29 34.69 34.47 34.25 34.03 33.82 33.6
29.18 25.58 22.84 20.55 18.68 17.08 15.64 14.18 12.97 12.16 11.26 11.05
0 5 10 15 20 25 30 40 50 60 80 100
Oxygen
a
"
0.0489 0.0429 0.0380 0.0342 0.0310 0.0283 0.0261 0.0231 0.0209 0.0195 0.0176 0.0170
q
c
0.00695 0.00607 0.00537 0.00480 0.00434 0.00393 O.O035,! 0.()()30l 0.00266 0.()()227 0.00138 0.00000
" Data from Ref. 495. /, Volume of gas. in milliliters, measured at DoC and 760 mm, dissolved in I ml of water when the pressure of the gas (without the contribution of the water vapor) is 760 mm. " Weight of gas. in grams, dissolved in 100 g of water when the pressure of the gas plus that of the water vapor is 760 mm.
and in alkaline solutions is
Eo = 0.401 V SHE
(2.47)
The reversible oxygen potential cannot be determined on a zinc electrode owing to the active nature of zinc. The O2 reversible electrode potential and Tafel parameters can be measured on a Pt electrode [1122]. Figure 2.28 shows that the Tafel slopes for Pt in 02-saturated 1M H2S04 solution are 93 and 126 mV/decade for reduction and oxidation,
'0 E
•
.£!
0 .6
]fOA :0 ~ '0 Ul
0.2
KOH
ooL-------2~0--------4~O--~----6~O--------8LO-------l~OO We ight % electro lyte
FIGURE 2.26. Solubility of 02 in KOH, H 2S04 , and H 3 P04 solutions at 25°C. After Drane [558].
so
CHAPTER 2
2 .-------------------------------------~
u'"
1.5
Q)
x
o
0 .5
20
10
30
40
50
Weight % KOH
FIGURE 2.27. Oxygen diffusivity in KOH solutions at 25°C. After Gubbins and Walker [4961.
1).. ------------
log apparent current density (A/cm 2)
.
FIGURE 2.28. Plot of the anodic and cathodic overvoltage obtained galvanostatically on a bright Pt electrode in an Oz-saturated. 2N H 2S0 4 solution. From Hoare [1122]. Reprinted by permission of John Wiley & Sons, Inc.
ELECTROCHEMICAL THERMODYNAMICS AND KINETICS
51
respectively, and the exchange current density is l.3 x 10-9 A/cm". The detailed reaction mechanism for oxygen reduction is complex as reduction of one oxygen molecule involves a four-electron charge transfer. In-depth reviews on the subject can be found in the literature [1122,1123]. As on most other metals, the reduction of oxygen on zinc occurs in two well-defined steps. In the first step, H20 2 is generated according to Eq. (2.48), and it is then reduced in the second step (Eq. 2.49) [113,128,1123]. (2.48) (2.49) These two reactions depend on the type and concentration of anions in the solution. The half-wave potentials for the oxygen reduction in a number of solutions have been represented by an equation of the form [1139] EI/2
=a -
b log C
(2.50)
in which C is the concentration, and a and b are constants. The ease of reduction of oxygen decreases in the order sot> cr > Br- > ClOt > NO] > 1- in the concentration range 10- 2-I M. The effect of anions on oxygen reduction is attributed to surface adsorption to act as an electron bridge for available ions [1139]. As pointed out by Tarasevich et al. [1123], the large negative free energy changes associated with the decomposition of H"02 and HO; suggest that these species should be very unstable in both acid and alkaline solutions. However, in the absence of impurities, the decomposition of hydrogen peroxide is very slow in aqueous solutions. Therefore, reduction of the peroxide (Eq. 2.49), is not complete, and various amounts of peroxide may be produced as a result of the whole reaction process. Boto and Williams [128] studied the electrode behavior of zinc in oxygen-saturated sulfate solutions in the pH range between 4 and 11. They found that the reduction of oxygen produces a mixture of H20 2 and hydroxide formed via a two-electron reduction of the peroxide. Depending on solution composition and pH, the average number of electrons for the oxygen reduction varies between 2.4 and 3.9. Also, according to Boto and Williams, the reduction is controlled by different processes in different pH ranges: • In a low pH range (between 4 and 6), both reactions (2.48) and (2.49) proceed on the zinc surface. In a higher pH range (up to 11), only reaction (2.48) proceeds on the zinc surface because the buildup of corrosion products at high pH prevents reaction (2.49) from occurring on the surface. • Within the pH range 4-6 at the corrosion potentials, E,o,," oxygen reduction on zinc is diffusion-controlled. In the pH range 6-11, on the other hand, it is controlled by the processes inside the passive film. Wroblowa and Qaderi [797] investigated the mechanism of oxygen reduction on zinc in O.IM K3P0 4 solutions in the pH range between 10.5 and 12.25 using a ring-disk electrode. Figure 2.29a shows that above --0.6 VNHE the surface is covered with passive surface films, and below -0.65 VNHE the zinc surface is bare. On the anodic potential
52
CHAPTER 2
- 0.4
0
0.4
0 .8
Disc potential , VNHE FIGURE 2.29. Zinc disk and gold ring currents as a function of disk potential. Electrolyte: 1M borate buffer + O.IM K3P04 ; sweep rate: 0.01 Vis. (a) Disk background currents in deaerated electrolyte; (b) pre reduced (anodic positive sweep) disk currents in oxygen-saturated electrolyte; rotation rates (rpm): (I) 700, (2) 1000, (3) 1500, (4) 2500, (5) 3600; (c) ring currents corresponding to curves in (b); (d) disk currents; cathodic sweep; rotation rates as in curves b; (e) ring currents corresponding to curves in (d); ring potential set at 1.08 V NHE • Reprinted from Wroblowa and Qaderi [797J, with kind permission from Elsevier Science Inc., 655 Avenue of the Americas, New York.
ELECTROCHEMICAL THERMODYNAMICS AND KINETICS
53
sweep (Fig. 2.29b,c) the ring current is only a very small fraction of the disk current, indicating that very few peroxyl ions are produced on the bare zinc surfaces and the reaction proceeds primarily by a direct four-electron reduction to hydroxyl ions, i.e., via the reaction in Eq. (2.51). On the other hand, peroxyl ions are produced on the passive surface, both in the passive and prepassive regions (Fig. 2.29d,e). (2.51a) H02+2e+W~OW
(2.51b)
In oxygenated solutions of NaHS0 3, Rosales and Granese [943] observed that the reduction rate of oxygen on a zinc electrode increases with increasing NaHS0 3 concentration. In neutral 0.5M Na2S04 solutions, according to Deslouis et ai. [113], the oxygen reduction is diffusion-controlled through a layer of corrosion products in the vicinity of the corrosion potential but occurs on an active surface at higher overpotentials. The film of zinc corrosion products acts as a barrier to the diffusion of oxygen but does not directly alter the reaction steps in the oxygen reduction. When the thickness of the electrolyte on the electrode surface is close to or smaller than that of the diffusion layer, the oxygen reduction rate increases significantly. Figure 2.30 shows that the reduction current greatly increases with decreasing electrolyte thickness at thicknesses less than 100 J1m [156]. According to Rosenfeld [336] the increased reduction current density in thinner electrolytes is due not only to the reduction of the diffusion-layer thickness but also to the self-mixing effect in thin electrolytes induced by evaporation and variation in surface tension and temperature. The dependence of oxygen reduction on the thickness of the electrolyte is very important with respect to atmospheric corrosion processes.
FIGURE 2.30. Dependence of the rate of oxygen reduction on Pt on the thickness of the electrolyte layer. 1M Na2S04' E = -0.65 VSHE' Reprinted from Stratmann et al. [156], with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom.
54
CHAPTER 2
2.6. CORROSION PROCESSES
2.6.1. General Considerations Corrosion is an electrochemical process in which the surface of a metal in contact with an electrolyte is oxidized with the simultaneous reduction of some species in the electrolyte on the metal surface and which, over time, results in the deterioration of the metal. Generally, a corrosion process can proceed in one of three modes, depending on the compactness and stability of the corrosion products as shown in Fig. 2.31: (a) direct dissolution without hindrance from corrosion products; (b) direct dissolution with hindrance from corrosion products; and (c) indirect dissolution through the formation of passive films. A corrosion process in a given environment can involve one, two, or all three modes and can change with time from one mode to another. A corrosion process can be studied with various electrochemical techniques as shown in Table 2.4. Compared to such corrosion testing methods as weight loss measurements, electrochemical techniques are fast and can be used to obtain instantaneous information on a corrosion process, which cannot be provided by weight loss measurements. Among the electrochemical techniques, AC impedance technique is a particularly useful method for studying electrode kinetics at the corrosion potential. Also, impedance techniques, along with the linear polarization technique, are the most commonly used methods for determining corrosion rates. In this section, the corrosion information obtained with impedance technique is presented. The linear polarization technique will be discussed in Chapter 5.
2.6.2. Impedance o/Corroding Electrodes 2.6.2.1. Impedance Techniques. The impedance of an electrode is one of the most important quantities that can be measured in electrochemistry. When the impedance is sampled over an infinite bandwidth, the impedance data contain all the information that can be obtained from the system by purely electrical means [137]. Some of the advantages of impedance techniques are (1) the use of very small signals which do not disturb the
Zn(OH),
+
2H+ - Zn"
+
2H,O
Zn - Zn 2 + + 2e
porous film (a)
(b)
(e)
FIGURE 2.3\. Schematic illustration of different modes of corrosion: (a) direct dissolution without hindrance from corrosion products; (b) direct dissolution with hindrance from corrosion products; (c) indirect dissolution through the formation of passive films.
ELECTROCHEMICAL THERMODYNAMICS AND KINETICS
55
electrode processes being studied, (2) the possibility of studying corrosion reactions and measuring corrosion rates in low-conductivity media, where traditional DC methods fail. such as corrosion inside concrete or under paint, and (3) the fact that polarization resistance as well as double-layer capacitance data can be obtained from the same measurement. A metal/electrolyte interface undergoing simple reduction or oxidation reactions can be simplistically described by an electric circuit as shown in Fig. 2.32, in which R Q is the resistance of the electrolyte, Cd is the double-layer capacitance, Ret is the charge-transfer resistance, and Zw is the Warburg impedance, which is related to diffusion processes [133, 137]. By passing a sine-wave potential signal of small amplitude across the electrode and measuring the AC current, one can obtain an AC impedance Z, expressed as
where j = ~ -1, and ZRe and Z,m are frequency-dependent real numbers. When Zim is plotted against ZRe for different frequencies, one obtains the complex plane of impedance (called a Nyquist plot). Alternatively, log IZ I and (/J can be plotted versus log w (Bode plot), where IZ I = (Z~e + Z;m)! 12, tan (/J = -ZR/Z,nl' and w is the frequency of the AC signal. Each plot has its advantages. The complex plane frequently is more useful for mechanistic analysis. On the other hand, the Bode plot directly employs frequency as the independent variable, so that a more precise comparison between experimental and calculated impedance can be made. Figure 2.33 schematically shows the complex plane of impedance for a simple system such as that in Fig. 2.32. From this plot. the values of the elements in the circuit of Fig. 2.32 can be obtained, such as the solution resistance between the surface and the reference electrode, the polarization resistance, R,), or the charge-transfer resistance, Ret. In the case of a simple system such as that shown in Fig. 2.32, R" = Ret. The polanzation resistance can be used to calculate corrosion current, as discussed in Chapter 5. In cases in which pseudoinductance is measured in the low-frequency range, R" may not be equal to Rei' The selection of R,! or RCI for use in corrosion rate calculations depends on the circumstances since there are a number of possible sources for the pseudoinductance [719]. The polarization resistances measured by an AC impedance technique have been found to generally agree with those obtained by a DC linear polarization technique for many systems [718].
FIGURE 2.32. Equivalent electric circuit for a simple-charge transfer reaction at a planar electrode surface.
56
CHAPTER 2
z""
Mass
Kinetic control
transfer control
/
/ 1
/
I
\
\ R + R«
- - - Z-,..
FIGURE 2.33. Impedance plot for an electrochemical system. Regions of mass transfer and kinetic control are found at low and high frequencies, respectively.
In real corrosion systems, the reaction processes are often more complicated than that described with the circuit in Fig. 2.32, as the electrode surface may be porous or covered with a surface coating. Different equivalent circuits are used to describe the impedance data obtained from these systems. For example, for filmed or coated electrodes the impedance can be expressed by a diagram of the type shown in Fig. 2.34, where Cd is the capacitance of the intact coating layers and R, is the resistance inside pores of the coating [134, 135]. The theoretical analysis for the impedance of different electrode/electrolyte systems can be found in the literature [133-137]. 2.6.2.2. Impedance of Zinc Electrodes. The impedance techniques have been used in a number of studies on the corrosion of zinc. Table 2.11 presents some impedance
c.
~t R,
FIGURE 2.34. Equivalent circuit for an electrode surface covered with a solid surface film.
Dissolution and precipitation of hydroxide
Diffusion in solid surface film
Simple dissolution
1M Na2S04' r = 2000 rpm, pH = 10, deaerated
1M Na2S04 + 0.2M Na2P04' r = 2000 rpm, pH = 10, deaerated
O.SM Na2S04 + O.OIM acetate, pH =4.7, deaerated
Revealed Process
6t
1m
10
20
(ncm')
2000
4000
(nem')
Zim
2
(n em')
1m
60
100
10
2000
..... 0.6
20
4000
600
Spectra
0.'"
10
O.OU
(n
em')
Rc (n em')
ZR,
o.m
Rc (n em')
Impedance Spectra of the Zinc Electrode Measured at or near the Corrosion Potential in Various Solutions
Solution and conditions
TABLE 2.11.
93
702
702
(continued)
Reference
-..J
Ul
t/.l
Pi
~
Z
;.::
:.. 3: Pi t/.l :.. Z o
Slz
o
3:
~
@
~ l'
~
~g
=4.7, Zn2P20rtreated, de aerated
1M NaCl, pH =3.8, de aerated
O.SM Na2S04, r =600 rpm
pH
O.SM Na2S04 + O.OIM acetate,
Solution and conditions
Dissolution with diffusioncontrolled proton reduction
Dissolution and diffusion mixed control
Dissolution under passivation
Revealed Process
TABLE 2.11.
. 10
100
I
•
200
300
.
0.002
~
100
Spectra
It
100
200
0.1
300
l00~
(n em')
~
Ii
In',:' Z.
100
ZI", (nem')
(Continued)
ZR, (n em')
ZR< (n em')
0.002
110
700
93
Reference
()
tv
~
~
;J>
:I:
~
=8 Multistep formation of oxide and carbonate surface film
Surface adsorption and surface diffusion controlled reactions
= 2000 rpm
7M KOH + O.IMZnO, r
350 ppm NaHC0 3, pH
Charge-transfer-controlled dissolution
= 2000 rpm
3M KOH + O.IM ZnO, r
Zim
1m
~
...l
(n em')
2
(ncm'1 ~
Z""
(ncm')
20
• s
0.1
11
4
2
em')
(n em ')
Re (n
'7Rc
. 0.1
ZRc (n em'l
(continued)
704
147
147
n
~
'"
""Zttl -l n
Z '0
'";J>
n
~
;J>
-< Z
=::: 0 '0
:;.:l
:c: ttl
t"' -l
;J>
~
ttl
:c:
n
0
:;.:l
-l
n
ttl
ttl
t"'
Charge-transfer-conlrolled reaclions
NH; adsorplion and reduclion + dissoluli on
O.IM (NH4}zS04' pH = 5.9
Revealed Process
3.5% NaCi. pH '" 6.4
Solution and condilions
TABLE 2.11.
I
2
)
(Oem')
Zim
1000
(Oem')
1m
•s
1000
2000
•
ZRc (Oeml)
ZIte (Oem')
~
SpeClra
l /"
(Continued)
427
70 1
70 1
Reference
tv
~
~
>
(")
:I:
~
Charge-transfer-controlled reactions
Charge-transfer-controlled reactions
\0% NH4C\, deaerated
1M Na2S04 + 1.5M ZnS04 + 0.0 1M NBu4Br
r
1000
Zim (Oem')
1000
Zim (Oem')
t-
~
1000
2000
21
ZRe (Oem')
ZRe (Oem')
843
ttl
0....
::l n Vl
ttl
~
0
> Z
Vl
(5
::
>
0 -< Z
::0
ttl
::a
~
> r
(5
~
:r:
n
0
b; n ;d
62
CHAPTER 2
spectra reported in the literature for the zinc electrode at or near the corrosion potential. It can be noted that these spectra differ greatly not only in shape but also in the numerical values, indicating that different electrochemical processes may occur on the zinc surface depending on the nature of the electrolytes. The nature of a corrosion process can often be revealed by an impedance spectrum. Deslouis et al. [700], based on the impedance spectra of zinc in deaerated sodium sulfate solutions, that the corrosion resistance is determined by a dissolution and a diffusioncontrolled process. In anodic dissolution, the first step yields an intermediate Zn;ds' The further oxidation of Zn;ds follows two parallel paths: a major path is to form Zn;~ in the solution, and a minor one involves the formation of Zn;;s on the surface. The overall corrosion is a dissolution through formation of zinc hydroxide with an accompanying diffusion-controlled cathodic process. Cachet et al. [702] found that the presence of HPO~- in Na2S04 solution increases the impedance of a zinc electrode. A strong inhibition of zinc dissolution occurs owing to the formation of a protective surface layer. A Warburg impedance is measured at low frequencies, indicating that the corrosion process is controlled by the diffusion of ions through the phosphate passivating layer [702]. Similar diffusion-controlled processes through a carbonate passive film have been proposed for the corrosion of a zinc electrode in bicarbonate solution [704]. In 1M NaCl at pH 3.8, Baugh [110] found that the corrosion of a zinc electrode is a simple charge-transfer dissolution limited by a dissolution-controlled proton reduction, since at high frequencies the Nyquist impedance plot reduces to a semicircle and at low frequencies a Warburg impedance develops, having a slope of 45° in the complex plane. He also proposed that formation of an oxide film may be involved in the corrosion processes since the double-layer capacitance is considerably smaller around the corrosion potential than in the cathodic region. Deslouis et al. [113] proposed an equivalent circuit for a zinc electrode in 0.5M Na2 S04 as shown in Fig. 2.35, where Ret is the dissolution charge-transfer resistance, R Q is the electrolyte resistance, W is the Warburg impedance, Cd is the double-layer capaci-
FIGURE 2.35. Equivalent circuit for zinc/O.5M Na2S04 interface. Reprinted from Deslouis et al. [113], with kind permission from Elsevier Science-NL. Sara Burgerhartstraat 25, 1055 KV Amsterdam, The Netherlands.
63
ELECTROCHEMICAL THERMODYNAMICS AND KlNETICS
7oo.------------------------------------, 7o
Eu
600
60
500
50
400
40
«
13 l.L
~300
30 ~
u
200
20
100
10
OL---~~~----~------~------~~.--J O
-1.9
-1.8
-1,7
-1 .6
-1,5
- 1.4
FIGURE 2.36, Variation of current density J, double-layer capacity C,b and surface film capacity Cr as a function of potential in 0.5M Na2S04' Reprinted from Deslouis et ai, [1131, with kind permission from Elsevier Science-NL, Sara Burgerhartstraat 25, 1055 KV Amsterdam, The Netherlands.
tance, and Cris the capacitance of the corrosion product film. Figure 2.36 shows the values of the elements in Fig. 2.35 as a function of potential. The lower Cd value and the definite values of Cf near the corrosion potential indicate the presence of a surface film. The corrosion of zinc in sulfate or chloride solutions seems to involve a charge-transfercontrolled dissolution process, with a formation of a corrosion product film on the surface, and a diffusion process through the film [113, 700, 702, 704]. When the zinc surface is free of corrosion products or the corrosion product film is of a porous nature, the corrosion process is controlled by charge-transfer-controlled dissolution and/or the diffusion of the reactants for the cathodic reaction. When the surface is covered with a passive film, the corrosion process may be controlled by a diffusion process through the film. Compared to the dissolution at an anodic potential, at which diffusion of the dissolution products such as Zn 2+ or Zn(OH)~- may be the rate-determining process at a large dissolution current, the dissolution rate at the corrosion potential is seldom controlled by the diffusion of the dissolution products because the oxidation/dissolution rate is usually very small at the corrosion potential.
3 Passivation and Surface Film Formation 3.1. INTRODUCTION Passivation is a process in which the metal surface transforms from an active state to an inactive state owing to the formation of a barrier layer. The passivation of zinc has been the subject of numerous studies as shown in Table 3.1. It should be noted that there is a clear difference between the studies made in strong alkaline solutions and those made in other solutions. In general, the studies made in strong alkaline solutions are related to battery applications, and the focus is on the maximum current prior to passivation and the time to passivation. This focus arises because passivation is a problem in alkaline batteries under a high discharge rate. On the other hand, studies made in neutral and slightly alkaline solutions are generally related to corrosion, and the focus is on the conditions and processes of passivation as well as on the stability of the passive films. Accordingly, the material presented in this chapter is organized in two main sections; dealing with passivation in alkaline solutions and passivation in other solutions. Prior to these two sections, a description of the conditions and characteristics of passivation is provided. A later section is devoted to anodization, an anodic process used to produce a solid surface film which generally passivates the surface. The last section discusses the stability of passivation and passivation breakdown. 3.2. CHARACTERISTICS AND CONDITIONS Passivation can be simplistically characterized by an anodic polarization curve, as shown in Fig. 3.1 [8, 1126]. In the active state, the metal electrode dissolves according to the reaction Me = Me'+ + ze-, and the dissolution current increases sharply with increasing potential. At a certain potential value, E p' the passivation potential, the current stops increasing and starts decreasing rapidly to much lower values, marking the onset of passivity. The current on the passivated surface, called the passivation current, iI" can be several orders of magnitude smaller than that on an active surface at the same potential. With further increase of the potential beyond a certain value, E b , the current may start to sharply increase, and the electrode is said to be in a transpassive state. This sharp increase 65
66
CHAPTER 3
TABLE 3.1. Passivation Overpotentia!, 1] p' breakdown potentia!, E h , and Passivation Current Density, ip , of a Zinc Electrode in Various Solutions Solution 0.5M NaH2P04 0.5M NaH 2P04 0.5M NaH 2P04 O.IM Na3P04 + O.IM Na zH4 350 ppm NaHC0 3 O.OIM NaHC0 3 0.15M Na2B407 + 0.3M H3B0 3 O.IM Na2HAs04 O.IM Na2Cr04 O.IMNaCI 1M NaN0 3 0.IMNaMo04 0.2M H 3B03 + O.IM NaOH H3B04 + NaOH 0.2M Na2HP04 1M Na2S04 + 0.2M Na2HP04 O.OOIMKOH O.OIM NaHC0 3 1M Na2C03 3M NaCI O.IM Na3P04 + O.IM Na2HP04 O.OIMKOH O.OIMNaOH H3B04 + NaOH O.IMNaOH 0.3MNaOH 0.5MKOH 0.5MKOH IMKOH IMKOH 4MKOH 4MKOH 5MKOH 7MKOH 7M KOH + 0.25M ZnO 7M KOH + 0.25M ZnO
pH 4.5 6.2 6.5 7.1 8 8.1 8.4 8.9 9 9 9 9 9.2 9.2 10 10 II 11.5 11.5 11.7 12 12 12.3 12.9 13.5
14
IJ ea (V)
0 0.2 0 0.15 0.2 0.5 0.15 0.2 0 0 0 0 0.22 0.18 0.25 0 0 0.22 0 0 0.23 0 0.2 0.3 0.26 0.3 0.3 0.32 0.39 0.36 0.43 0.37 0.38 0.36 0.27 0.3
Eh (V SCE)
0.9
ie (j.1Ncm z) 0.05 200 0.2 30 28 5
2.0 1.2 -0.75 -0.76 1.5 2.1 1.2 0.6 -0.6 2.8 0.3 1.6 0.8 1.6 l.l 1.4
IO 100 I 0.2 200 200 0.8 300 I I 50 300 200 2 I 9 2
5 100 500 600 17,000 15,000 18.000 3,000 20,000 5,000 5,000 2,500
Reference IOI 603 IOI 481 704 194 526 21 98 45 45 98 16 355 698 702 46 127 3 3 526 46 37 355 526 19 422 1128 24 1128 794 681 1128 27 29 26
aDifference between passivation and corrosion potentials.
in current is either associated with the breakdown of the passive film, leading to a severe dissolution of the electrode, usually localized, or with the onset of another reaction such as oxygen evolution. When it is associated with the breakdown of the passive film, Eb is called the breakdown potential. It is also termed the pitting potential since localized corrosion, such as pitting, generally occurs above the breakdown potential. Generally, passivation occurs when the dissolution of a metal produces a situation in which the solubility of a salt or hydroxide in the electrolyte near the electrode surface is exceeded and a compact solid film forms [1126, 1127]. As a result of the film formation, ions must move from the metal phase into the surface film in order for further dissolution
PASSIVATION AND SURFACE FILM FORMATION
67
passive
active-passive transition
FIGURE 3.1. Schematic plot of a typical current-potential curve showing the transition from the active to the passive state of a metal.
current
of the metal to take place. At least three processes are involved in the dissolution on the passivated electrode: (i) transfer of metal ions from the metal phase into the surface film; (ii) transfer of ionic species from the solution phase into the surface film; and (iii) transfer and hydration of metal ions across the film/solution interface. This last process is the dissolution of the film and determines, in general, the corrosion rate of the metal in the passive state. When there is no other reaction, such as oxidation of water, the passivation current, ip ' equals the net corrosion rate of the metal in the passive state. The occurrence of passivation on zinc surfaces is determined by the thennodynamic and kinetic conditions for formation of a stable and compact solid surface film. According to the potential-pH diagram shown in Fig. 2.2 in Chapter 2, passivation of a zinc surface does not occur in acidic solutions without the presence of film-forming agents. In slightly alkaline solutions containing no complexing agents with which zinc can form soluble salts, passivation of zinc is thermodynamically possible through the formation of zinc oxides or hydroxides. In the presence of ionic species, the possibility of passivation may either increase as a result of the formation of a solid zinc salt layer or decrease as a result of the formation of more soluble zinc compounds in the solution. For example, the presence of carbonate promotes the formation of zinc carbonate in near-neutral or neutral solutions and thus extends the pH range in which passivation is possible to lower values compared to that for carbonate-free solutions (Fig. 2.5). The appropriate thermodynamic conditions do not necessarily guarantee the occurrence of passivation. The actual occurrence of passivation depends also on kinetic conditions. While thermodynamic conditions determine whether formation of stable zinc salts is possible as a result of zinc dissolution, kinetic conditions determine the chemistry near the electrode surface and the nature of the surface film formed. The stability, continuity, and compactness of the film eventually determine the degree of passivation. Depending on the conditions, passivation may occur instantly in some cases while it may take days or months in others. Table 3.1 presents the solution compositions and potential ranges in which the passivation of zinc is observed. It may be noted that in some solutions passivation occurs
68
CHAPTER 3
it the open-circuit potentials whereas in others an overpotential is needed. The corrosion potential of an electrode can be used as an indication of the state of passivation. A corrosion potential that is much more positive than the reversible potential usually indicates the passivation of the electrode surface. On the other hand, the occurrence of passivation mayor may not result in a corrosion potential that is significantly more positive than the reversible potential. 3.3. ALKALINE SOLUTIONS Studies on passivation of zinc electrodes in alkaline solutions are mostly related to zinc alkaline batteries [18, 24, 681, 794,889,903]. The utilization of zinc electrodes in alkaline batteries depends on the ability of the electrode to remain active during the anodic dissolution process. The occurrence of passivation prevents their maximum utilization. Due to this special interest the parameters obtained from these studies are often the peak current density before passivation and the time to passivation at a given current density. The peak current density is generally determined from a dynamic potential-current curve whereas the time to passivation is most often obtained from a potential-time curve. 3.3.1. i-V Curves Figure 3.2 shows a typical anodic current-potential curve for a zinc electrode in an alkaline solution [24]. The curve can be divided into four regions: an active dissolution region (I), a first linear region (II), a second linear region (III), and a passive region (IV). The current values and the limits of the regions vary with hydroxide concentration, temperature, and hydrodynamic conditions. The characteristics of the i- V curves depend on the potential sweep rate and convective conditions in the electrolyte, except in region I the i- V relation is essentially
100
"""' 80 Na u
:;;:
a 60 .£
III
II
'-'
IV
i::
Q
E
1:: ;::l
u
40 20 0
-1.5
-1.4
-1.3
-1.2
-1.l
-1.0
Potential (Vsce) FIGURE 3.2. Current-potential curve measured on a zinc rotating disk electrode. Conditions: IN KOH, 300 rpm, 4-m VIs sweep rate, 25°C. Regions: I, initial dissolution; II, first linear region; III, second linear region; IV, passive region. After Chang and Prentice [24].
69
PASSIVATION AND SURFACE FILM FORMATION
20 r---------------------------------,-1 • First peak cuuenl
... Second peak curr.nl
-1.1
• Firs. peak poteontJa]
X Second
peak potential
~
<- 10 .s
-1.2
~ :s!
c: OJ
x
-1.3
.
•
~
-1.4
O L-----------------~------------~
o
5
(Sweep rate)'" (mV/s) 'A
FIGURE 3.3. EtTect of sweep rate on a stationary disk. Conditions: IN KOH, 25°C. After Chang and Prentice [24].
unaffected by the potential sweep rate. Figure 3.3 shows that the potentials of the first and second peaks are not affected by sweep rate. However, the peak potentials are found to be dependent on the sweep rate in some situations [19,1128]. Figure 3.3 also shows that the current densities at the peaks increase with increasing sweep rate, the relationship becoming almost linear at higher sweep rates [24]. The lack of a fully linear relationship between the peak currents and the square root of the sweep rate indicates that the reaction is not totally controlled by diffusion in the electrolyte. This is in accordance with the dependence of the current peak on the rate of electrode rotation, shown in Fig. 3.4 [1128]. On a rotating electrode, the peak current densities are less dependent on the potential sweep rate. At rotation rates below 600 rpm, a straight line is 200
150
E<>
.s..: 100
J
50
0
,/
0
2
4
e
6
(Revolution
I s)
10
'12
FIGURE 3.4. Peak current at a rotating zinc disk electrode in IN KOH as a function of the square root of the rotation rate. After Hull et al. [1128].
70
CHAPTER 3
0 .----------------------------------,---,
0- 0 . 5 en
;f.
en
:I: V>
:>
>
-1
-1.5
o
20
10
T i me , min
FIGURE 3.5. Typical potential-time curve for a zinc anode in a 7.S4M KOH solution containing 0.5M dissolved ZnO. Current density is 40 mNcm 2. After Sato [1137].
observed, which is indicative of a mass-transfer control. With increasing rotation rate, diffusion control in the overall reaction becomes increasingly less important.
3.3.2. Passivation Time The time for passivation is most often determined by the use of galvanostatic techniques [18, 25, 889, 903, 904], by which the potential is measured as a function of time under a constant current density. Figure 3.5 shows a typical E-t curve. where the time at which the potential rises rapidly is taken as the time to passivation, ~, [1137]. There
2,---------------------------------------, • 7.24 M
"'6
T 4.98 M
1.5
...::
+ 2.92 M . 0.784 M
;E-
....
...
Vi c:
Q)
"0
C ~
8 0 .5
.. .. ... 0.5
1.5
t
-111
2
2.5
3
, S ·l7
FIGURE 3.6. Current density vs. reciprocal of the square root of the passivation time for upward-facing zinc anodes in KOH solutions of various concentrations. After Liu et al. [IS].
PASSIVATION AND SURFACE FILM FORMATION
NE u
71
• 1.0 M
8
,. 2 .0 M
...:: ~
'(i;
3 .5 M
6
• 4.5 M
c:
'"
X 5 .0 M
u
E
~
4
. 7.0 M
:;
0
"12.8 M 2
0
0
20
10
30
40
50
FIGURE 3.7. Anodic current density vs. reciprocal of the square root of the passivation time in KOH solutions of different concentrations at 23°C. Reprinted from Dirkse and Hampson [889[. with kind permission from Elsevier Science Ltd. The Boulevard. Langford Lane. Kidlington OX5 1GB. United Kingdom.
is generally a linear relation between current density and the reciprocal of the square root of the passivation time, (", for current densities up to 1.5 A/cm2 , as shown in Fig. 3.6 [18]. Figure 3.7 shows that the i_t~l /2 curves are also linear for higher current densities [889]. Extrapolating the curves in Figs. 3.6 and 3.7 to the current axis, one notes that the maximum current density attainable without passivation is rather low. It is generally found that for a wide range of conditions, equations of the type
(i - i o)tp112 = k
(3.1 )
can be used to describe the relationship between i, the applied current density, and tp ' the time required for passivation. In this equation, io and k are constants whose values depend on electrolyte concentration, temperature, and convective conditions. The form of Eq. (3.1) indicates the important role of diffusion in the passivation of the zinc electrode [889]. In semilinear diffusion, if the original concentration in the bulk solution is Co, after a time f the concentration, c, of the dissolution product at the electrode surface is C
=
Co
+ (2i1F)(tlnD) I 12
(3.2)
with D the diffusion coefficient and F the Faraday constant. Passivation occurs when the solution in the vicinity of the electrode reaches its capacity limit for the dissolution product. If C"" is the critical concentration required to cause passivation, Eq. (3.2) becomes (3.3) Equation (3.3) is valid when diffusion in solution is the only mode of mass transport. When a certain amount of convection is taken into account, Eq. (3.3) can be modified to an equation of the form of Eq. (3.1) in which io represents the mass transport by the
72
CHAPTER 3
processes other than diffusion. Different modifications of Eq. (3.1), taking into account oxide film growth or chemical reaction steps in addition to diffusion and convection, have been used to describe the relationship between current density and passivation time [18, 25,889]. Increase in temperature appears to prolong the passivation time. Measurement of dissolution rate as a function of temperature indicates that the passivation process on zinc electrodes in alkaline solutions is under mixed control of diffusion and activation processes [889]. The effective activation energy was reported to be in the range of 27.2-41.9 kllmol, which is higher than the typical activation energy for a diffusioncontrolled process, about 12.6 kllmol, and is lower than that for an activation-controlled process, about 41.9 kllmol [24]. The orientation of the electrode surface is found to affect the passivation time. Different passivation times are measured for upward-facing, downward-facing, and vertical electrodes [18, 26, 1130]. Longer passivation times are required for vertical electrodes than for horizontal ones owing to the onset of natural convection [26]. The thickness of the diffusion layer as a function of electrode orientation has been studied by O'Brien et aZ. [1138]. Dirkse and co-workers [886, 902] studied the effect of ionic strength on passivation time. They found that for binary KOH-HzO mixtures the passivation time at a given current density increases with increasing KOH concentration at a relatively low ionic concentration, while it decreases with increasing KOH concentration at a high ionic concentration. As shown in Fig. 3.8, the slope of the i versus t;1 IZ curve increases with KOH concentration up to 7-8M and then decreases with further increase of KOH concentration [889]. This phenomenon was explained in terms of loss of the unbound water as the ionic strength of the solution increases. The presence of zincate ions in the solution seems to shorten the passivation time [18]. Figure 3.9 shows that the slope of the i versus t;112 curve decreases with increasing zincate concentration [1131]. According to Hampson et aZ. [1131], passivation occurs when the concentration of zincate ions at the anode is equivalent to that of OH-. The presence of other ionic species also affects passivation time. In one study, passivation time was found to decrease linearly with increasing carbonate concentration
.
~
-",
~
~
O.B
:J
u
?
:; "00.4 Q)
a. o iii
oOL-..- -2-'---.......I..4 -
-...l.6--....Ja'--- -,l..o- -'.L.2---.J'4
Concentration of KOH (M)
FIGURE 3.8. Slopes of the plots of current density vs. reciprocal of the square root of the passivation time for KOH solutions of different concentrations. Reprinted from Dirkse and Hampson [889], with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom.
PASSIVATION AND SURFACE FILM FORMATION
73
1.35
0.95L---------------~--------------~------~
o
0.5
Zn 2 + Concentration, M
FIGURE 3.9. Variation of k. the slope of the i vs. t;1 12 curve, with Zn 2+ concentration in 7M KOH for horizontal zinc anodes at 23°C. After Hampson [1131 J.
up to 2M in 7M KOH solution; this decrease was attributed to the increase in the viscosity of the solution [1137]. In another study, it was found that Pb and Sn added to the electrolyte give rise to a smooth and compact metallic film over the Zn surface and thus increase the degree of passivation [27].
3.3.3. Characteristics The amount of surface film needed for passivation to occur can be very small. Hull et al. [1128] measured i- V curves using a rotating ring-disk electrode. They found that, as shown in Fig. 3.10, the i- V curves observed with both ring and disk electrodes are, in
b -to
~
+4.0
~
C
...
C
~
...~
:l U
:l
u
u
Vl
Ci
-o.~
bll
c:
+2.0
C2
-1.0
Disc potential, V (Hg/HgO)
-1.0
Disc potential, V (Hg/HgO)
FIGURE 3.10. (a) Current-voltage curves recorded on the disk of a graphite-zinc ring-disk electrode at different rotation rates. Scan rate = 10 mV/s; IN KOH; electrode area = 0.035 cm 2 . (b) Reduction current observed at the graphite ring held at a fixed potential of -1.45 V (Hg-HgO) for the current-voltage curves shown in (a). After Hull et al. [1128].
74
CHAPTER 3
600.0
Thick silver colored film Film darkens Brigh.ly e.chod
K
200.0
Film beeomes G brigh.
F
A
BI.ckfilm_
/
~
I
0.0 L....,6"--_'---_ _ _...I...-_ _ _-'J'--'-H'--_~ - 1.40
0
-1.28
-1.16
- 1.04
Electrode Potential vs Hg-HgO (Volt)
-0 .92
FIGURE 3.11. Current-voltage curve of a zinc wire electrode and the corresponding changes of the electrode surface in unstirred 5N KOH at a potential scan rate of 1.1 mVIs. After Hull et at. [1128}.
general, of almost identical forms. The collection efficiency on the ring electrode is constant for all the regions of the i- V curves, indicating that the fraction of the current that is utilized in the formation of the surface films on the zinc disk at either the peak potential or during the onset of passivity is so small that there is no detectable change in the amount of dissolved species arriving at the ring electrode. Thus, the majority of the current produced at the disk for all potential regions is utilized in direct electrochemical oxidation of the zinc to soluble products. The physical appearance of the zinc electrode surface during anodic polarization in alkaline solutions varies with the potential. Figure 3.11 summarizes the changes in the appearance of the zinc electrode surface visually observed by Hull et al. [1128] during cyclic anodic polarization in an alkaline solution. At potentials less anodic than that at point B, the surface of the electrode remains bright and the development of etch patterns is clearly visible. At point B, a milky white film forms, which, in the absence of stirring, slowly flakes from the electrode surface in the region C-D. At point D, as the current begins to fall, the color of the surface beneath this film can be observed to darken until a region of intense current oscillation is reached. The passivated surface at F is very black; however, the color rapidly lightens as the potential increases further, leaving the surface covered with a white crystalline passivating layer. The oscillation at the onset of passivation is attributed to the changes in the pH of an electrolyte layer adjacent to the surface as a result of the oxide formation and diffusion [1128]. The oscillation has also been explained as a result of the distribution between IR potential drop in the electrolyte and the potential across the double layer [904]. Aurian-BIajeni and Tomkiewicz [20] studied the impedance spectra of zinc electrodes in alkaline solutions and concluded that the anodic passive films are composite layers of oxides and solution. The growth of the layers and the passivity are dictated by the diffusion of the electrolyte across these layers. The passive layer becomes more and more compact, with grains approaching a spherical shape. The layer thickness starts to increase after a certain porosity and conductivity are reached. It was found that the dielectric properties of the oxide-solution composite layer were not the linear combination of the dielectric characteristics of the oxide and the solution [1134].
PASSIVATION AND SURFACE FILM FORMATION
75
The passive films formed on zinc electrodes in alkaline solutions have often been found to have semiconducting properties [422,484,526,687]. For example, Scholl et al. [422] studied the photocurrent spectra of the films formed in 0.5 and 1M KOH solutions under various conditions and found that the films exhibited properties characteristic of an amorphous semiconductor. It is important to note that the oxide films formed in alkaline solutions are not "truly" passive for there is a significant current in the passive potential region. For example as reported by Dirkse [904], the steady-state dissolution current density in the passive region in 10-40% KOH is in the range of 19-33 mA/cm 2. Also, as can be noted in Table 3.1, the passivation current densities in concentrated KOH solutions, e.g. 1-7M, can range from 2.5 to 20 mA/cm 2.
3.3.4. Mechanisms of Formation of Passive Films The detailed mechanisms of zinc passivation in alkaline solution under specific conditions are complex and are still not fully understood. However, in simplistic terms, as Dirkse [904] has summarized, the passivation generally proceeds by the following steps: 1. Zinc is oxidized to form ZnO or Zn(OH)2' 2. They dissolve in the electrolyte to form Zn(OH):1 or Zn(OH)~-. 3. When the electrolyte can no longer dissolve the ZnO or Zn(OHh that is produced by the charge-transfer reaction, a solid film forms and passivates the surface. Powers and Breiter [26, 27, 29] studied the change in the appearance of the surface of zinc electrodes in 7M KOH under a microscope during anodic polarization. Two different films, both zinc oxides, were found to form under different conditions. Type I is white and porous and forms by precipitation from a supersaturated layer of electrolyte covering the electrode. Type II forms directly on the electrode surface and ranges in color from light gray to black. The dark color of the type II film is due to the excess of zinc in the film. Type I film precipitates near the potential of the first peak of the i- Vcurve. Type II film is more coherent and skinlike than Type I film and forms directly on the surface beneath the type I film at a slightly higher potential than that of the first peak. The characteristics of the two types of films are summarized in Fig. 3.12. Absence of convection seems to be important for the formation of type I film; in the presence of convection, only type II film is observed to form [27]. The formation of type II film is considered to be responsible for the transition from the active to the passive state of zinc in alkaline solutions. Cobweblike structures are formed when type II film dissolves due to the gathering together of the excess zinc in this film [29]. The cobwebs are electrically conductive and can be further oxidized. Type II film also appears to serve as a catalyst for hydrogen evolution at potentials anodic to the zinc/zinc oxide equilibrium value. The formation of hydrogen bubbles can mechanically dislodge the passivating film and cause reactivation of the electrode [27]. The duplex nature of the oxide film formed on a zinc electrode surface was also reported by Szpak and Gabriel [28]. Depending on the conditions, the anodic oxides, on the micron scale, have either a carpetlike, a boulderlike, or a thistlelike structure. Figure
76
CHAPTER 3
With convection nearly absent
1.
While. Oocculem, IYpe 1 film. formed by precipitalion from a . / supersaluraled layer of eleclrolyle, ZnO
~
Coherenl, strongly lighl·absorbing, direclly·formed lype II fiLm, ZnO with excess of Zn Zn eleclrode
2.
In presence of convection ~_ _
Type II film Zn electrode
FIGURE 3.12. Schematic illustration of the films formed on zinc electrodes in 7M KOH under different convection conditions. After Powers [27].
(b)
(a)
Adsorption
(c)
Adsorption
Transport
QznloHIV
Adsorption
Transport
iZn'OH'/
ad.y
, /II ~Zn!oHll l~' znloHr;
OZnIOHI;:
_-Q0H
I
M~o..: I
I :
I
V
I
I
Distance
Polymerization
(d)
(f)
(e)
Adsorption
Transport
/1
I
Adsorption
Transport
I
I I
I I
I
!\ I
Nucleation and growth
I I I I
I I I
Oxide densification
Oxide film folding
FIGURE 3.13. Schematic representation of the stages in the development of an anodic ZnO film: (a) quasi·equilibrium state; (b) development of transport region; (c) formation of polymerization region by trapping of monomers; (d) nucleation and growth region ; (e) oxide densification; (f) oxide film folding. After Szpak and Gabriel [28].
77
PASSIVATION AND SURFACE FILM FORMATION -0.80 ,--------,.---------------r-,....-,
o
Zn + 4 ORZn(OH)/ + 2e
~ -1.00-
00
::c:
Zn + 40R -> Zn(OH)/ + 2e Zn(OH)/ -> ZnO + 20R + H20 Growth of type I ZnO
]
is -1.20 .!:!
I
8<1.l -g l:l
I
Zn + 20H-
->
ZnO + H2 0 + 2e :
Growth of type II ZnO
--r : I
I
Lilil -1.40 L - -_ _ _--'-_ _ _ _ _ _ _ _ _ _ _ _.....L---Jw
r- t.
- - 1 - - - - - tb
-----+-1 tel
Time
FIGURE 3.14. Proposed scheme for the processes associated with the anodic passivation of zinc in alkaline solutions. After Liu et at. 1181.
3.13 schematically illustrates the stages in the formation of the oxide film. The nucleation is considered to be associated with the saturation of Zn(OH)3 monomers and to be completed within milliseconds. The growth of the nuclei to an observable size, e.g., to a radius of 10-5 cm, may take a much longer time depending on the conditions. Liu et al. [18] proposed a multistep reaction process for the passivation of zinc in alkaline solutions as shown in Fig. 3.14. In the first step, the anodic dissolution proceeds for a time, t", producing zincate ions which accumulate near the surface. When a critical concentration, Ceril , which may be several times the solubility of ZnO in KOH solution, is reached, type I ZnO begins to precipitate. The anodic dissolution continues through the porous oxide film up to a time th at which the rate of mass transfer of hydroxide ions through the film falls below that required for the formation of zincate ions and formation of type II zinc oxide is initiated. After an additional time, te , the whole surface is covered with type II oxide and becomes passivated. Cabot et al. [681,794] described the formation of a passive film under a potentiostatic condition as consisting of two stages. In the first stage, the process is controlled by diffusion of zincate ions in the solution near the surface. In the second stage, during the film growth the process is controlled by ion migration through the pores in the film. The growth tends to be two-dimensional at low solution concentrations and three-dimensional at high concentrations. According to Cabot et al., a porous solid precipitate, possibly Zn(OH)2' might already form in the linear region before the current peak on an anodic i- V curve, because a linear region on an anodic i-V curve implies solution resistance control in the pores of the hydroxide film. 3.4. OTHER SOLUTIONS
3.4.1. Slightly Alkaline and Carbonate Solutions The typical behavior of a zinc electrode in carbonate solution can be seen in Fig. 3.15, which shows a potential-time curve obtained by Kaesche [127] for a zinc electrode
78
CHAPTER 3 - charging curve - - -discharging curve
D
c ........... oxygen evolution (?) I I
B
--- growth of oxide layer
I ..
F oxide reduction \,--- --.(, G
paSSivatIOn _ hydrogen evolution
Time
\
hydrogen evolution ..... __ [ __ _
FIGURE 3.15. Potential-time curve for zinc in 1M Na2C03 solution, 60°C, during galvanostatic anodic charging and cathodic discharging. Reprinted from Kaesche [1271. with kind permission from Elsevier Science Ltd. The Boulevard. Langford Lane. Kidlington OX5 1GB. United Kingdom.
in 1M Na2C03 solution at 60°C. At point A, the anodic charging is switched on from a preset cathodic potential where hydrogen evolution has been occurring. In the time between points A and B, the electrode is passivated, followed by oxide growth between points Band C, with superimposed oxygen evolution beyond point C. If the current is reversed during charging, then between points F and G the oxide is reduced, and the potential finally returns to the exclusively hydrogen-evolution potential. The passivation time, tp ' corresponding to point B, is found to be a logarithmic linear function of the charging current density, i a , with a slope of about -1.6. This relation indicates that the passivation in the carbonate solution has a constant value of the product iJ;12 , differing significantly from that in concentrated alkaline solutions, where ii ~ 12 is generally a constant. In Kaesche's experiments in 1M Na2C03, the passivation time increases with increasing temperature from about 0.1 s at 20°C to about 0.5 s at 80°C at a charging current density of 10 mA/cm2 [127]. For a tp value of about 0.1 s at room temperature, the product of ii p is about 1 mC/cm 2, corresponding to approximately an oxide monolayer. Apparently, the thickness of the passive film at passivation increases with increasing temperature. The current efficiency in 1M Na2C03 at room temperature is nearly 100% since the ratio icfJiia, with ic and tr the current density and the time for the reduction of the oxide film, is near unity. At a higher temperature the ratio is lower than 1, as shown, for example, in Fig. 3.16 for a temperature of 60°C [127]. The decrease in current efficiency observed with longer charging times is attributed to oxygen evolution. The passivation characteristics vary with the concentrations of carbonate and bicarbonate in the solution. In 1M Na2C03 solution, the passivation is fast and brought about by a small amount of oxide, about a monolayer [127]. On the other hand, in O.OIM NaHC0 3 solution, no passivation occurs after 30 min of polarization at 25°C, but passivation occurs readily at higher temperatures. In O.IM NaHC0 3, the passivation is very slow and is associated with the formation of a thick oxide film. At higher temperatures, this film becomes thinner and has a larger number of nuclei and thus requires a shorter time to reach passivation. According to Kaesche [127], the passivation process in O.lM NaHC0 3 is quite different from that in 1M Na2C03 in the following respects: (a) longer passivation time by an order of magnitude; (b) less than 20% current efficiency; (c) formation of white oxide flakes on the surface; (d) formation of a larger amount of oxide, about 300 mC/cm2, at passivation; (e) reduction of passivation time with increasing temperature, from 24 min at 40°C to less than 1 min at 90°C; and (f) larger oxide nuclei.
79
PASSIVATION AND SURFACE FILM FORMATION
..... . .
~
::·0.8
::>;
0.6
... •
y
.~
o c
Q)
~
0.4
Q)
'E Q)
5 0.2
o
oL--------------L--------------~--------~
10
0.1
Amount of charge, me I cm 2
FIGURE 3.16. The ratio i,t"!Va as a function of the anodic charge ie/a' measured for zinc in 1M Na2C01 at 60°C. I i" I = I ic I = 10 mAlcm 2 Reprinted from Kaesche r127]. with kind permission from Elsevier Science Ltd. The Boulevard. Langford Lane. Kidlington OX5 1GB. United Kingdom.
According to Muralidharan and Rajagopalan [206], in O.OIM NaOH solution zinc can be passivated with the formation of three monolayers of Zn(OHh at the active centers. The formation of these monolayers is two-dimensional with an instantaneous nucleation. D' Alkaine and da Cunha [23] found that the peak current density of the anodic i- V curve is linearly related to the peak potential in carbonate solutions, indicating an ohmic-controlled process in the oxide film: (3.4 )
where p is the ionic resistivity, and II' is the film thickness at the peak potential. This resistance is found to decrease with increasing carbonate concentration as shown in Fig. 100,-----------------------------------------,
a.
0..
oL---------------~--------------
o
__
_ L _ _~
2
Na 2C0 3 Concentration (M) FIGURE 3.17. Representation of oxide film resistance (pip) versus Na2C03 concentration. After D' Alkaine and da Cunha [231.
80
CHAPTER 3
3.17. D' Alkaine and da Cunha reasoned that since the film thickness changes little with concentration, the decrease in the resistance with increasing concentration shown in Fig. 3.17 may be attributed to changes in the resistivity of the film. Kannangara and Conway [3] conducted a detailed study of the passivation of zinc in alkaline and carbonate solutions. Figure 3.1S shows the cyclic voltammograms for Zn in 1M Na2C03 and in 3M NaCI solutions at pH 11.5. The passive potential range in the carbonate solution is significantly larger than that in the chloride solution. The major features of the i- V curves are identified as follows: • Peak Al consists of the dissolution current and formation current for ZnO or Zn(OH)2' • Peak Al consists of Ala and Alb' with Ala due to the dissolution of surface defects (grain boundaries, etc.), which is diffusion-controlled, and Alb due to the fonnation of a compact layer of ZnO or Zn(OH)2, which is controlled by a surface mechanism. The fonnation of this compact oxide layer, corresponding to only one or two monolayers, passivates the Zn surface. • Peak C I is the reduction current for ZnO or Zn(OHh. The slope of the current density of peak Al versus the square root of the potential sweep rate, about 0.7, was explained by Kannangara and Conway [3] as being due to a mixed process of diffusion-controlled dissolution and film formation. The dissolution process (with a peak current density proportional to the square root of the sweep rate) proceeds simultaneously with the film fonnation process (with a peak current density proportional to the sweep rate) until the surface is completely covered by the passive oxide film. The amount of dissolution during the passivation is described by qJ qc, the ratio of the charge under peak Al to that under peak C I . This ratio decreases with decreasing solubility of the oxide and with increasing potential sweep rate as shown in Fig. 3.19. Similar parallel dissolution and film fonnation processes were proposed by Rangel and Cruz [704] for a 350 ppm bicarbonate solution of pH S.O. The dissolution of zinc produces an adherent film with OH- ions diffusing from the bulk into the porous film against a flux of zinc species. They measured the current peak as a function of temperature and obtained an apparent activation energy of IS.4 kJ/mol, which is in the range for a diffusion-controlled process. The dissolution mechanism was considered to follow the reactions given by Eqs. (2.14)-(2.16) in Chapter 2. Using X-ray diffraction, Huber [S02] identified the anodic films formed in hydroxide and carbonate solutions. The oxide film formed in Na2CO, is thin and light in color and is insulating, whereas that fonned in NaOH is of y-zinc hydroxide with a small amount of oxide and is thick and dark. The dark color is due to the presence of an excess of metallic zinc. ZnO has also been identified in other studies as the passive film in carbonate solutions [3]. The ionic resistance of the passive film has been found to decrease with increasing concentration of CO~- [23].
3.4.2. Phosphate Solutions The passivation of zinc in phosphate solutions is of particular importance in the surface treatment of zinc and its alloys. Phosphating, an immersion or spray process by
-8
-4
A
-1
c,
o 2
0, evolution
A:-/
E (Vscd
3
-4
-2
0
2
4
B
- 0.8
C,
11--
- 0.6
reduction of ZnO
reduction of Zn + +
dissolution and film formation
- 0.4
FIGURE 3.18. Typical cyclic voltammograms for polarization of polycrystalline Zn in 1M Na2COj (a) and 3M NaCI solution (b) at pH 11.5. Potential sweep rate = 50 mY/s oThe electrode is polycrystalline, of apparent area 0.07 cm 2, etched in HC!. After Kannangara and Conway 131.
1
(,)
......
ME
0
4
~
00
Z
(5
~
!::
'"
o
.."
!::
m :::::: r
'";;;; n
c
[/)
>Z o
Z
(5
~'-l
[/) [/)
82
CHAPTER 3
15
• pH 13 T
pH 12.7
+ pH
11.5
200
300
Sweep rate (mV s" )
FIGURE 3.19. Values of q,,/qc from cyclic voltammograms for Zn oxidation and redeposition, as a function of potential sweep rate , S, for 1M Na2S04 solutions. After Kannangara and Conway [3J.
which an insulating zinc phosphate coating is formed on the zinc surface, is commonly used in the steel industry to enhance paint adhesion on zinc-coated steels. Passivation of zinc in phosphate solutions can occur at the open-circuit potential in a wide pH range [101,481, 784]. The open-circuit electrode potentials were found to shift to more positive values, a sign for the occurrence of passivation [784]. De Pauli et al. [481] investigated the i- V characteristics of the zinc electrode in O.lM Na3P04 + O.lM Na2HP04 in the pH range between 6.5 and 13.1. Figure 3.20 shows that the shape of the i- V curve strongly depends on pH. The peak current density before the passivation region is a function of potential sweep rate and, when plotted against the pH. has a minimum around pH 9,5, as shown in Fig. 3.2l. In the higher pH range. the peak current density has a linear relation with the square root of potential sweeprate, S. which is similar to the behavior in alkaline solutions, In the lower pH range the peak current density has no linear dependence on 5 112 or 5, indicating a different passivation mechanism, Also, it was suggested that at pH > 12 a dissolution-precipitation mechanism operates whereas at pH < 12 a solid-phase process prevails, Furthermore, different mechanisms may operate in different temperature ranges . At a temperature near ooe, the peak current is independent of the electrode rotation speed, whereas at higher temperatures it depends linearly on the square root of the rotation speed [603]. According to De Pauli et al. [481, 698], PO~- ions promote zinc dissolution since the peak current density, i lll , increases with increasing PO~- concentration; the capacity ofPO~- ions to provide OH- ions at the interface was suggested as a probable explanation for this effect. The passive film can be a monolayer or a multilayer film, depending on the concentration of phosphate and the temperature of the solution. Awad and Kamel [784] attributed the passivation in phosphate solutions to the formation of a highly polymerized zinc phosphate layer on the electrode surface. According to De Pauli et ai, [603], the passivation operates through a dissolution-precipitation mechanism with the participa-
83
PASSIVATION AND SURFACE FILM FORMATION
pH - - - -13.1
- - 9.1 0.9 -'-7.1
.,
0.6
f .
,
0
C .,
·0.3
\
·0.6
·1.2
\.
\ \
J. ".\ I.
.1
/ ,'i!
./ \
r. /
-0.9
\
\
I,
t
8
f
I
0.3
1
.
/
/
12<
I
I
I
./
\ I
,
\ r
./
J'
-Ul
I
Y .1.5
·1.2
·0,9
Potential (V"J FIGURE 3.20. Current-voltage relationship for a stationary Zn electrode in buffered phosphate solutions of different pHs. Potential sweep rate 0.05 Vis. Reprinted from De Pauli et al. [481], with kind permission from Elsevier Science-NL Sara Burgerhartstraat 25, 1055 KV Amsterdam, The Netherlands .
=
3 .5
• 0.1 VIs y 0,05 VIs
+ :;c ,3.
0.025 VIs
3
~
Ol
.2
2.5
24L-----~6------~8------~ 10-------12------~14~~
pH
FIGURE 3.21. Dependence of log it" on pH. [pol-] = 0.2M. Reprinted from De Pauli et al. [4811. with kind permission from Elsevier Science-NL, Sara Burgerhartstraat 25, 1055 KV Amsterdam, The Netherlands.
84
CHAPTER 3
,.-...
"'E
~
5
.Q .;:;:;
.,
0
~
0
;:
-1
0>
I: =>
u
-2
- 1.8
- 1.5
- 1.2
-0.9
-0.6
-0.3
Potential (V ...) FIGURE 3.22. Effect of potential reversal at different anodic potentials on the reduction peaks. S = 0.05 Vis. rpO~-l =0.2M, pH = lO.9. After De Pauli et al. [4811.
tion of phosphate species in the solution and Zn 2+, which diffuses through the thin nonpassivating film, according to the following scheme: (3 .5) (3.6)
Spherical nodules with high phosphorus content can be formed in the multilayer films . The formation of the nodules was considered to be a result of repassivation at the places where breakdown of the passive film has occurred [603, 702]. Passive films may have different phases at different potentials. Figure 3.22 shows that the positions and the number of cathodic peaks change when the potential range of anodic polarization is increased, indicating that phase transformations of the passive film may be involved during the anodic polarization [481]. This phase transformation in the passive region was postulated by De Pauli et al. [481] to involve a slow chemical reaction of the form (3.7)
The dissolution in the passive region appears to be associated with the diffusion of ions through the passive layer [702].
3.4.3. Miscellaneous Solutions Boron compounds are often added as a pH buffering agent in the solutions used for passivation studies on zinc electrodes. However, it has been found that, in addition to their buffering effect, these compounds also participate directly in the passivation process. Zinc becomes passivated very easily in boric acid-sodium hydroxide solutions in the pH range of 9.2-12.3 [355], and the amount of charge needed for the passivation is many times less
85
PASSIVATION AND SURFACE FILM FORMATION
/
}
i
1.0
/
.'I \
\
,
I
,,
_ . - 1M crO," -IMMoO/ - - -- I \VO.'"
I
--1.0 0.01
-
-
-
- -
-
- - - - -
________
-
- -- - - - -
________
- -
- -- ,
________ -w
0.1
10
Current density, rnA/em' FIGURE 3.23. Effect of oxyanions on passivity in aerated 1M solutions at pH 9 and 40°C. After Bijimi and Gabe [98].
than that needed in alkaline solutions of similar pH values [907]. The passivation may be attributed to the formation at nonsoluble zinc borate salts [16, 17] or to the buffering effect of borate ions on the electrolyte near the electrode surface [45]. Pimat et ai. [482] studied the passivation of zinc in chromate solutions at pH 1.5. Zinc exhibits a passive behavior in the potential range from -0.9 to 0.05 VseE' Passivation cannot occur when sulfate is also present because of the competitive adsorption of sulfate on the electrode surface. Bijimi and Gabe [98] found that the passivation current density of zinc in a chromate solution increases with temperature. Compared to other oxyanions, chromate is the most effective in passivating the zinc surface, as shown in Fig. 3.23. Aeration or deaeration has no effect on the passivation of zinc in solutions of oxyanions. According to Macias et ai. [175,202], the passivation of zinc in Ca(OHh-containing solutions is due to the complete coverage of the surface with a compact layer of Ca(Zn(OHh)2·2H20, the formation of which is determined by the concentration of Ca2+ ions. The passivation results in a decrease in corrosion current to about 0.5-1 /lA/cm" and a shift of the corrosion potential to about -0.5 VSCE' De Pinto et ai. [21] studied the passivation of zinc in arsenate solutions and found that arsenate ions increase the dissolution rate in the active potential region but tend to form an insoluble compound with zinc ions to passivate the zinc surface. Thick anodic films can be formed in arsenate solutions. Cracks were found to develop in the solid surface film, more in the solutions with dissolved 02' Passivation can also occur in chloride, sulfate, and other solutions under anodic conditions within certain pH ranges, generally due to the surface saturation of the zinc salts [532]. 3.5. ANODIZATION Anodization is a process used to produce a solid surface film of a certain thickness and properties. Anodic coatings of various colors, from white to gray to black, can be produced in aqueous solutions of sodium hydroxide and sodium carbonate. Figure 3.24 shows that, depending on the current density, a white or a black film can be formed on a zinc surface through anodization in a solution of NaOH and Na 2C0 3 [494].
86
CHAPTER 3
>
• 200 mA I cm 2
:g
cCl>
" 100 mA I cm 2
+ 50 mA l cm 2
(5
0Cl> "'0
• 30 mA I cm 2
o
~
- - - Wh ite films - - - - - Black films _ . __
a
_
_
_ __ _
2 3 4 Time , m in
FIGURE 3.24. Potential-time curves of zinc electrodes anodized at various current densities in O.146N NaOH and O.054N Na2C03' After Whitaker and Fry [494].
Black or dark-colored coatings can be produced by anodization in NaOH solutions with pH values greater than 13.3 at current densities of 70-140 mA/cm 2 • Gray to white oxide coatings can be produced in alkaline solutions having a pH equivalent to that of O.OOl-O.IN NaOH solutions. White or light-colored coatings can also be produced in Na 2C0 3 solutions. The coatings produced in Na2C03 solutions are less porous and are 10 to 100 times thinner than those produced in NaOH [493]. The conditions fortheformation of a black oxide film in NaOH solution are shown in Fig. 3.25. The black color is attributed to metallic zinc particles dispersed in the film, resulting from the reactions (3.8) (3.9) When zinc is anodized in NaOH solutions, first a thin film composed of very tine crystals of oxide (50-150 A in size) is formed. Upon prolonged anodization, this primary film adjacent to the metal becomes thicker, and a porous layer is formed [404]. The anodic films formed in alkaline and carbonate solutions have been found to consist primarily of zinc oxide. The oxide is converted to zinc hydroxide with difficulty and is practically insoluble in pure water. However, if some carbon dioxide is added to the water, the oxide layer is rapidly converted to basic carbonate [493]. The black oxide films prepared from anodization in NaOH and Na2C03 solutions appeared to have a high absorption for wavelengths shorter than 2.2 pm and a high transmission for wavelengths longer than 2.2 pm [1133]. The anodic coatings obtained in Na3P0 4 solutions primarily consist of zinc phosphates [96]. MuItilayers are formed on zinc electrodes in slightly acidic solutions in the presence of NaH 2P0 4 . The films are passivating and contain ZnO and inclusions of phosphate; the amount of inclusions increases with temperature [603]. The anodically produced films in chromate-containing
87
PASSIVATION AND SURFACE FILM FORMATION
Compact black oxide layer 1-
_ _ _ __
/
o
/
I
Thin layer of porous black oxide UJ
:t
>'" W
-0.5 Black oxide layer
l
Thin passive layer
-1.0
0.1
10
Current density, rnA/ern2 FIGURE 3.25. Anodic behavior of zinc in O.SM NaOH. After Bianchi et (//.13591.
solutions vary in color from clear to slightly iridescent to yellow to black [493]. The clear film consists essentially of CrPl with some water. 3.6. STABILITY OF PASSIVATION 3.6.1. Type of Passivation
Stability of passivation refers to the ability of the electrode surface to maintain its state of passivation. It can be characterized by the potentials at which passivation occurs and ends and by the passivation current density in the passive region. Depending upon the state of polarization, zinc passivation in various solutions can be divided into two types: that which occurs only under a certain anodic polarization (type A) and that which occurs also at the open-circuit potentia!, i.e., the nonpolarized condition (type B). In many cases, passivation is associated with the formation of a solid film that is stable only at certain anodic potentials. Sometimes, the solid film is not stable at all potentials, and passivation is achieved only under certain anodic current densities to maintain a metastable solid film. In alkaline solutions the passivation, due to the formation of Zn(OHh, disappears when the anodic polarization is removed as shown in Fig. 3.26, because the metastable passive hydroxide film quickly dissolves in the solution [904]. In general, the passivation of zinc electrodes in concentrated alkaline solutions is of type A since it is necessary to impose an anodic overpotential on the zinc electrode. On the other hand, the passivation in many slightly alkaline solutions is of type B. In the case of type B passivation, the oxide or salt films are relatively stable, and the electrode surface
88
CHAPTER 3 2 1.6
1.2 ~
g
a 0 .8
-t
0.4
0
0
2
4
6
8
Time, seconds
FIGURE 3.26. Voltage decay at 25°C in 30% KOH saturated with ZnO polarized at 1.9 Vzinc for 3 min. After Dirkse [904J.
can maintain the passive state without an external polarization, for example, in slightly alkaline, carbonate, and phosphate solutions. Stability of passivation can be further characterized by the passivation current density in the passive region. Generally, the current in the passive region consists of two parts. A part of the current is due to film growth to maintain the barrier thickness which chemically dissolves. Another part of the current is due to the direct dissolution of zinc through the pores or defects of the film. If the passive film is stable, then the current is mainly from the dissolution through defects in the film. On the other hand, if the film is compact but not stable and dissolves, the current is mainly used for maintaining the film thickness. Table 3.1 shows that the passivation current, iI" can vary by orders of magnitude from one case to another. There appears to be a clear distinction in the data shown in Table 3.1 between concentrated alkaline solutions and other solutions. Generally, in alkaline solutions the passivation current densities are higher than I mA/cm 2, while in other solutions they are much lower than 0.1 mA/cm 2• The current density in the passive region in alkaline solutions as a function of concentration and potential sweep rate is shown in Fig. 3.27 [794]. In situations in which stable salt films can form, the passivation current is determined primarily by the compactness of the films. Generally, the solid films formed through a direct oxidation process are more compact than those formed through a dissolution-precipitation process, the latter tending to result in porous films [607]. The means of transport for charge, reactants, and dissolution products depend on the structure of the surface film, electrolyte composition, and test condition. Mass transport by diffusion, capillary force, and convection are important to the reactions on the passivated zinc surface in alkaline solutions. The passive films are generally porous [26, 27]. Ionic diffusion through the oxide film tends to be the rate-limiting process for zinc in phosphate solutions [702]. The passive oxide films formed in certain borate, borax, and NaOH solutions are found to be electronically conductive since oxygen evolution can
PASSIVATION AND SURFACE FILM FORMATION
80
89
. 3.0 M KOH '" 2.0 M KOH
60
+ 1.0 M KOH
20
+ o L-----~------~------~--------------~~
0 2 3 4 5 s '/2 (mV·/2 s ··,')
FIGURE 3.27. Passivation current (Ip) VS. SII2 plots for KOH solutions of different concentrations. Reprinted from Cabot et al. [7941. with kind permission from Elsevier Science-NL, Sara Burgerhartstraat 25, 1055 KV Amsterdam, The Netherlands.
proceed on the oxide films [355,907, 1129]. The passive film formed in carbonate solutions seems to be an insulator, as the passive region can extend to potentials more positive than 3.0 V SCE without significant increase in the anodic current ]3, 907]. The passive films formed in various alkaline, phosphate, and borate solutions have been found in many studies to conduct current under certain conditions through a semiconducting mechanism [422, 484, 526, 687].
3.6.2. Passivation Breakdown Passive films on a metal electrode tend to break down at certain anodic potentials. Thus, the stability of passivation can also be characterized by the potential above which breakdown of the passive film occurs. Generally, the higher the breakdown potential, the more stable the passive film is. In practice, breakdown can be determined with an anodic polarization curve as shown in Fig. 3.1. The metal dissolution at passivation breakdown is usually localized, leading usually to the formation of pits, and the breakdown potential may be taken as the pitting potential. The rate of pit growth can be very rapid at the breakdown potential because of the large driving force. The value of the breakdown potential is very sensitive to solution chemistry. As shown in Fig. 3.18, breakdown does not occur for an anodic polarization up to 3.0 V SHE in 1M Na 2C0 3 at pH 11.5, while it occurs at about -0.1 VSCE in 3M NaCI solution at the same pH. Depending on the composition of the base solution, the addition of a very small amount of chemical species can greatly affect the value of breakdown potential. The addition of as little as 150 ppm of chloride ions in a borate solution has been found to reduce the breakdown potential by about 0.5 V [355]. Generally, Ct, Br-, r-, F, CIO:;, SO~-, and CH 3CO:; have been found to reduce the breakdown potential [16, 17,46,355], while OW, NUl' HPO~-, CrO~-, CO~-, WO~-, MoO~-, and BOi- have been found to increase the breakdown potential [37,45,98, 1091. Examples of the variation of the
90
CHAPTER 3
20 ,----------------------------------------, • NO ~
T CH,CO', + CIO; ~
.~
• CIO ;
10
Q)
v
X F'
• Sr-
C
6S0: '
~
:s
X I'
o
v CI .
-1
-0.9
-0.8
-0. 7
Potential, V (Hg/HgO)
FIGURE 3.28. Influence of different anions on anodic behavior of zinc. Potential sweep rate = 25 mV/min; pH = 9.2; [H 3B0 3] = O.2M; [NaOH] = O.IM; [anion] =O.IM. Reprinted from Augustynski era!. [16], with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington. OX5 1GB, United Kingdom.
breakdown potential with the addition of some ions are shown in Fig. 3.28 [16, 17]. Figure 3.29 shows that the breakdown potential in slightly alkaline borate solutions decreases with increasing concentration of CIO; and Cl- [16]. Sergi et al. [174] found that passivation breakdown is more sensitive to variations in pH than to changes in chloride concentration. Breakdown can arise from a variety of effects, which can be generally divided into physical effects and chemical effects. Physical effects include changes in field strength, dielectric properties, and mechanical failure of the barrier film caused by internal stress or volume changes resulting from transitions in crystal structure. The chemical effects relevant to oxide breakdown originate at the film/electrolyte interface. These include -0.4 .----------------------------------------,
0- 0 .6 en J: 0; ~
>
-0 .8
OJ
~ Q) (5
n.
-1
_1 .2L-______
~
-2
__________
~
____________L __ _ _ __ _
-1
~
o
Log anion concentration (M)
FIGURE 3.29. Variation of breakdown potential of zinc with concentration ofCIO; and Cl-. pH = 9.2; [H 3B0 3 ]
=O.2M; [NaOH] = O.IM. Reprinted from Augustynski er al. [16], with kind permission from Elsevier Science Ltd. The Boulevard, Langford Lane. Kidlington. OX5 1GB, United Kingdom.
PASSIVATION AND SURFACE FILM FORMATION
91
nonuniform dissolution of the film and defects introduced by ion adsorption on the surface or ion incorporation in the film. In many cases, these physical and chemical effects cannot be separated. Chemical changes can lead to mechanical failure; on the other hand, mechanical failure may result in localized attack. It is the interdependent effects of the physical and chemical processes that control the breakdown of passivity. The mechanism for breakdown of passivation on zinc surfaces has been the subject of several studies. Galvele and co-workers [45. 652J postulated that passivation hreakdown is caused by OH- ion depletion at the zinc/electrolyte interface. The formatIon of hydroxides at the surface produces protons, which reduce the pH near the surface. Since zinc oxide or hydroxide is only stable in slightly alkaline solutions. the decrease in pH may prevent the formation of a stable film. On the other hand. since the dissolution rate of zinc exponentially increases with potential, the dissolution at a higher potential will result in a lower pH value near the surface. When the pH in localized areas near the surface is outside the range in which the oxide film is stable, the dissolution of zinc in these local areas will not produce nor maintain passivation at these areas by a passive film. The breakdown potential is therefore the minimum potential at which an acidified solution can be produced and maintained in contact with an active dissolving metal. Thus, whether breakdown occurs under a given potential is essentially dictated by the thermodynamic equilihrium at the local areas. In the model put forward by Galvele and co-workers. any anion that does not interfere with the zinc dissolution and acid consumption at the interface should not affect the breakdown potential. Anions that can reduce surface acidification. for example. by a buffering action, will generally enhance the stability of a passive film on a zinc surface. On the other hand, according to Augustynski et al. [16, 17], who studied the effects of many anions on the breakdown potential of zinc in borate solutions, the depassivation effect of the anions is due to the ability of these anions to form more soluble salts with zinc. Through specific adsorption on the electrode surface, these anions locally prevent the formation of a passive film. The difference in the effects of the various anions on the breakdown potentials is largely due to the difference in the solubilities of the corresponding zinc salts and in the extent to which the anions are specifically adsorbed. Although the models of Galvele and Augustynski differ in the proposed role of the anions. both models are essentially based on the solubility of the passive film at localized areas. Galvele's model emphasizes the effect of local acidification, which increases the solubility of passive films, while Augustynski's stresses the effect of anion adsorption and accumulation at localized points, which cause changes in the solubility of zinc in the electrolyte near the surface. The combination of the two views perhaps gives a more complete picture concerning the phenomenon of passivation breakdown of a zinc electrode. It can be generalized that anions which enhance surface acidification or form zinc salts more soluble than zinc oxide and hydroxides tend to reduce the stability of passivation while those which reduce surface acidification or form less soluble salts tend to enhance the stability of passivation. In essence, the occurrence of breakdown is determined by the ability to repassivate the local surface area once activated by any physical and chemical inhomogeneities and fluctuations, which are always present within the system. If the condition is such that repassivation is not possible, the active dissolution at these areas will intensify and expand, leading to the breakdown of passivation.
4 Electrochemistry of Zinc Oxide 4.1. INTRODUCTION Zinc oxide is a semiconducting material and is commonly found in the corrosion products of zinc and its alloys. Many passive films formed on zinc electrodes in various electrolytes have been shown to have some of the semiconducting properties ofZnO, and the semiconducting properties have been found to play an important role in the corrosion behavior of zinc in many situations. However, correlations between the electrochemical behavior of zinc oxide and the corrosion of zinc are still lacking. Also, zinc oxide is notable as one of the most frequently used materials in studies of semiconductor electrochemical and photoelectrochemical phenomena. Much of the early understanding of semiconductor electrochemistry was actually obtained from these studies. It is thus felt that a systematic overview of the electrochemical properties of zinc oxide would be useful not only for a deeper understanding of many zinc corrosion phenomena but also for further research on the semiconducting behavior of zinc oxide itself. 4.2. BASIC PROPERTIES 4.2. J. Physical Properties
Zinc oxide exhibits many useful optical and thermal properties and is widely used in the production of rubbers and paints. As a semiconductor, zinc oxide possesses a set of unique electronic and photoelectronic properties and has been used in a number of applications such as varistors and photocopying products. Selected properties of zinc oxide are shown in Table 4.1 [570]. Zinc oxide has a wurtzite structure in which the oxygen atoms are arranged in a hexagonal close-packed lattice with zinc ions occupying half the tetrahedral sites, as shown in Fig. 4.1 [860]. The two types of ions, Zn 2+ and 0 2-, are tetrahedrally coordinated and are therefore positionally equivalent. Due to their marked difference in size, these ions fill only about 44% of the volume in a zinc oxide crystal, leaving some relatively large (O.095-nm radius) open spaces [570]. The natural color of zinc oxide powder is white, but it displays pronounced changes in color when heated or when certain impurities are incorporated into the crystals [11791. In the visible region ofthe spectrum, zinc oxide powder has good hiding power (the ability 93
94
CHAPTER 4
TABLE 4.1. Properties of Zinc Oxide" Molecular weight Lattice Lattice constants Density Dielectric constant Refractive index Energy band gap Enthalpy of formation Melting point Specific heat Solubility in H 20
Zn: 65.38; 0: 16.00; ZnO: 81.38 Hexagonal. wurtzite a =0.324 nm, C =0.519 nm, cia = 1.60 5.78 glcm 3 or 4.21 x 1022 ZnO molecules/cm 3 8.54 2.008 3.2eV I Zn(s) + Z02(g) ~ ZnO(s) -83.17 kcallmol Vaporizes at -1700 D C at normal atmospheric pressure; melts at 197YC under pressure 9.66 call(mol-K) 1.6 x 10-6 g per gram of H 20 at 2YC
=
"Reprinted from Van [570]. with kind permission from Elsevier Science Ltd. The Boulevard, Langford Lane. Kidlington OX5 lOB, United Kingdom.
to prevent light transmission) and tinting strength, depending on the refractive index and particle size. The white color of zinc oxide powder, composed of transparent and colorless microcrystals, is a result of a series of optical processes: surface reflection, transmission through crystals, refraction, and scattering of light rays. Reflection of light at an oxide/air interface is low, about 11 %, the rest of the light being transmitted through the crystal. Oxide particles with sizes that approach the wavelength of the incident light are highly effective in scattering those rays, thereby reducing the degree of light penetration into the powder layer. The hiding power varies with particle size to a maximum of 0.25 pm as shown in Fig. 4.2 [1277]. Zinc oxide with particle size smaller than 0.06 pm attenuates (i.e., scatters and absorbs) ultraviolet radiation most effectively [451]. Zinc oxide possesses a set of unique thermal and optical properties, as summarized by Brown [1179]. It changes from reflector to absorber abruptly at a wavelength of 0.385 pm, close to the border between the UV and visible regions, as shown in Fig. 4.3 [1179]. With dopant additions and proper heat treatment, zinc oxide can be a versatile phosphor that converts ultraviolet light and X-ray radiation into light of various colors [570]. Zinc ,
"
FIGURE 4.1. Crystal wurtzite structure of zinc oxide. 0, zinc; e, oxygen. After Addison [860].
ELECTROCHEMISTRY OF ZINC OXIDE
c
95
ZnO
o
·in en
·E en
c
~
:c
g> Q)
>
~
(j)
a:
o
0.4
0.2
0.6
0.8
Particle size in microns
FIGURE 4.2. Relative light transmission of zinc oxide as a function of particle size. Zinc oxide is availahle in a wide range of particle sizes and provides a broad spectrum of hiding power since the relative light transmission of the oxide is a U-shaped function of the particle size. Optimum hiding power is ohtaincd with particles of O.25-flm size. After Stutz 11277].
oxide also shows significant photoconductivity in the ultraviolet region and throughout most of the visible region of the spectrum. 4.2.2. Electronic Properties
Semiconductors are substances with electronic conductivity between that of metals (10 6 _10 4 Q-I·cm-I) and dielectrics « 10- 10 Q-I·cm-I). Zinc oxide is intrinsically an n-type semiconductor due to electrons excited from ionized zinc interstitials existing in the zinc oxide crystal lattice. At 25°C, the typical electronic conductivity is 1 Q-I·cm-I [11791.
100
~ Q)
()
c cO t5 Q)
50
'lii a:
OL-______ 300
~
400
______L -______L -_ _ _ _ _ _ 500
600
~
700
____
~
800
Wavelength (nm)
FIGURE 4.3. Percentage of light reflected by zinc oxide as a function of wavelength. Zinc oxide exhibits a pronounced absorption edge in the ultraviolet range (low reflectance) at 385 nm. After Brown [1179].
96
CHAPTER 4
,',"" Conduction Band Donor -"'--'-
/
Valence Band ./
./
;'
;'
/
FIGURE 4.4. The effect of donor and acceptor defects on electron transport across the band gap .
The quantum theory of solids presents a complete and rigorous description of the nature of current carriers in semiconductors. According to quantum theory, the energy spectrum of electrons in an ideal crystal consists of energy bands filled with energy levels (allowed bands) and with no energy levels (band gaps). The width of a band gap and the distribution of electrons in the allowed bands determines the electronic nature of a crystal (i.e., metal, semiconductor, or dielectric). For a semiconductor, the upper, unfilled band is called the conduction band while the lower, almost filled band is called the valence band, as shown in Fig. 4.4. The width of the band gap, E~ = E,. - En which is the most important electronic characteristic of a crystal, depends on the strength of the chemical bonds. For ZnO, Ex = 3.2 eV. The electronic conductivity of semiconductors, as expected from the band structure, can be generated by electrons of atoms of the basic substance in the crystal (intrinsic conductivity) as well as by electrons of impurity atoms or by the presence of defects (extrinsic conductivity). In intrinsic semiconductors at T> 0 K, the generation of current carriers occurs as a result of the thermal excitation of some electrons from the valence band to the conduction band, with the corresponding thermal rupture of some chemical bonds. Simultaneously, an equal number of positively charged holes are created in the valence band. In an electric field, these holes behave like particles possessing a positive charge equal in absolute value to the charge of the electron. For extrinsic semiconductors, impurities and defects (which have energy levels located in the band gap) are classified as either donors or acceptors as shown in Fig. 4.4. Donors, usually located at energy levels slightly below the conduction band, give up excess electrons to the conduction band, thereby creating electron conductivity (n-type semiconductors). Acceptors, located at energy levels slightly above the valence band, capture valence electrons from atoms of the basic substance, producing hole conductivity (p-type semiconductors). An important concept in the description of semiconducting properties is that of the Fermi level, EF, which is defined as the energy level for which the probability of being occupied by an electron is ~. For an intrinsic semiconductor at room temperature, EF lies essentially midway between the conduction band and the valence band within the band gap. For a doped material, the location of EF depends on the type and concentration of the dopant. For moderately or heavily doped n-type solids, EF lies slightly below the conduction band. Similarly, for moderately or heavily doped p-type materials, EF lies just above the valence band. The conductivity of zinc oxide samples has been observed to be in the range of 10- 17 to 103 Q-I·cm-I, depending principally upon the method of sample preparation [1179]. A
ELECTROCHEMISTRY OF ZINC OXIDE Zn2+ O~ Zn'+ O~ eZn+ Zn2+ Zn2+ o~ Zn2+ O~ Zn e O~ Zn2+ O~ Zn'+ O~ Zn2+ Zn2+ Zn2+ O~ Zn2+ O~ eO~ Zn2+ O~ Zn2+ O~
O~
(a)
97 Zn'+
O~
Inl+
O~
Zn2+
O~
Zn'+
O~
Zn2+
O~
O~
In 3 +
Zn2+
O~
O~
e-
Zn2+ 0" e O~ Inl+ Zn 2 + 0" Zn'+
O~
Zn'+ e O~
(b)
FIGURE 4.5. Schematic lattice structure of zinc oxide (al and doped zinc oxide (b).
single crystal of pure zinc oxide has very low conductivity and is an insulator. The semiconducting property of zinc oxide depends on the presence of defects in the zinc oxide lattice. Two types of semiconducting zinc oxides are distinguished according to the types of defects; one contains interstitial and the other substituted zinc atoms, as shown in Fig. 4.5. In the interstitial type, zinc oxide is partially reduced by reaction with agents such as carbon monoxide or hydrogen at elevated temperatures (400-900°C). Each atom of oxygen removed releases an atom of zinc and two electrons. The zinc atom moves to the void space to become an interstitial atom, which may be in the form of Zn, Zn+, or Zn 2+, depending mainly on temperature. The substitutional type of zinc oxide is produced in the presence of metallic vapor or salts at elevated temperature. A portion of the zinc atoms in the zinc oxide crystals are replaced by the foreign metallic atoms. The zinc atoms, upon release from their lattice positions, diffuse to the crystal surface, where they vaporize. Depending on the type of metallic atoms, the conductivity of zinc oxide can be either increased or decreased. The electronic structure and surface characteristics of zinc oxide are found to increase the rate of many chemical reactions [570]. Zinc oxide has great absorptivity for Hb CO, and CO 2 after being cleaned of absorbed HP and CO 2 by vacuum heating. The catalytic activity of ZnO is generally increased with increase in conductivity. Upon exposure of zinc oxide to air or to oxygen, some of the electrons near the surface are spontaneously trapped by physically adsorbed oxygen to form negative ions on the surface [1179]. These ions, formed by transfer of electrons from the interior of ZnO to the surface, create an upward bending of the bands. The adsorbed ions can be desorbed either by heating or by generation of holes with light. The holes are driven to the surface to neutralize the adsorbed ions, and the photoelectrons in the conduction band compensate the positive space charge of the ionized donors, resulting in a flatband condition. 4.3. SEMICONDUCTOR ELECTROCHEMICAL BEHAVIOR
4.3.1. Basic Theories When zinc oxide is immersed in an aqueous solution, protons, hydroxyl ions, and other ions adsorb on the surface. In the simplest case of pure water, OH- ions are attracted
98
CHAPTER 4 liquid
H
o
H
H 0
H
H 0
H
H 0
H
I
I
I
I
I
I
I
I
Zn --- 0 --- Zn --- 0 ___ Zn ___ 0 --- Zn ___ 0
solid
FIGURE 4.6. Schematic lattice structure of zinc oxide surface in water.
to the zinc sites and H+ ions are attracted to the oxygen sites on the ZnO surface as shown in Fig. 4.6. The surface generally adsorbs an excess of one species and becomes charged either positively or negatively depending on the reactions at equilibrium: (4.1 )
(4.2) When a semiconductor is brought into contact with a solution containing a redox couple (e.g., Zn 2+/Zn or H+/H2)' if electrostatic equilibrium is attained, the Fermi levels in the two phases must become equal (the electrochemical potentials must become equal). In the case shown in Fig. 4.7 for an n-type semiconductor, where EF of the semiconductor is higher than that in solution, electrons will flow from the semiconductor to the solution phase. The excess charge in the semiconductor does not reside at the surface, as it would in a metal, but instead is distributed in a region near the surface, called the space charge region. The resulting electric field that forms in the space charge region is shown by a bending of the bands. In the case of Fig. 4.7b, where the semiconductor is positively charged with respect to the solution, the bands are bent upward (with respect to the level in the bulk semiconductor), and the degree of band bending is measured by V,. When the semiconductor has no excess charge, there is no space charge region and no electric field and the bands are not bent. The electrode potential under this condition is called the flatband potential, Etb . The flatband potential is a very important quantity for a semiconductor electrode as it connects the parameters that can be experimentally determined to the parameters derived from semiconductor/electrolyte interface physics. The value of the flatband potential of a metal oxide semiconductor is found to be quantitatively related to its electron affinity [1180]. When the interface is irradiated with light of energy greater than the band gap, E~, photons are absorbed and electron-hole pairs are generated (Fig. 4.7c). Some of these electrons and holes, especially those formed in the bulk semiconductor beyond the space charge region, recombine with the evolution of heat. However, the space charge field causes the separation of electrons and holes. Thus, in the case of Fig. 4.7c, the holes arrive at the surface at an effective potential equivalent to the valence band edge and cause the oxidation of the redox species in the solution from R to 0 while the electrons move into the external circuit through the semiconductor electrode lead. The flow of holes and electrons in opposite directions can be measured as current (photocurrent). The larger the band bending, the more holes are driven to the surface and the larger is the photocurrent. Thus, the onset of the photocurrent is near E!b, at which the band bending is zero.
E,
(a)
solution
O/R
vsT
(b)
interface
O/R
(c)
EEl'
--------....-r I - - .-
e
E
°/R
I hv > E,
O/R: (a) Before contact in the dark; (b) after contact (in the dark) and electrostatic equilibration; (c) junction under irradiation.
FIGURE 4.7. Representation of the formation of the junction between a semiconductor and a solution containing a redox couple
semiconductor
Ev
EF
Ec
E
"" 'C
Z \) o >< 6tT1
N
o"T1
~
-I
Vl
s::
tT1
:r:
o\)
;>:J
-I
tT1 \)
r
tT1
100
CHAPTER 4
Similarly, for a p-type semiconductor the bands are usually bent downward and the electrons generated by irradiation are moved by the field in the space charge region toward the surface, causing reduction of 0 to R. The electrode potential of a semiconductor can be changed by an extemal power source so that the degree of band bending is altered. Depending on the type of semiconductor, electron or hole current can be generated by bending the bands in one direction. In the dark under reverse bias (increasing band bending), there is essentially no current flow because for an n-type semiconductor there are few holes and for a p-type semiconductor there are few electrons available in the semiconductor to participate in the reactions. On the other hand, under a forward bias (decreased band bending), there are more electrons (n-type semiconductor) or more holes (p-type semiconductor) at the semiconductor surface. More thorough descriptions of fundamental semiconductor electrochemistry can be found in the literature [1177, 1183].
4.3.2. Flatband Potential The flatband potential of a semiconductor electrode can be experimentally determined by measuring the capacitance as a function of potential. The capacitance of the space charge layer, Csc. the degree of band bending, V, = E - Efb • and the dopant concentration, N D, are related and can be described by the Mott-Schottky equation [1177]:
lIC;c = (2IeeeoND)(-V, - kTIe)
(4.3)
Thus. a plot of lIC~ versus potential E is linear. In this plot the potential at which the line intersects the potential axis yields the value of Efb, and the slope can be used to calculate the doping level ND . Figure 4.8 shows a typical Mott-Schottky plot reported by Dewald [514] for single-crystalline ZnO in 1M KCl at pH 8.5. The flatband potential can also be estimated by determining the onset potential for photocurrent [423]. 1,200
,,
1,000 N
fu
800
u2-
600
b
400
'-0.59 ohm" em"
theoretical SlOPer",
200 0 -0.5
o
0.5
1.5
Pote ntial (V seE)
FIGURE 4.8. Mott-Schottky plots for two crystals under exhaustion conditions. The dashed lines represent the theoretical slopes. The intercepts of the linear plots give the values of the flatband potentials of the crystals. After Dewald [514).
101
ELECTROCHEMISTRY OF ZINC OXIDE
TABLE 4.2. Dopant Concentrations, N D' Slopes of Logarithmic Current vs. Potential (in Millivolts), and Flatband Potentials, En" of ZnO in Various Solutions Solution
pH
10·4_ IO -z M Fe(CN)~-
3 6 8.8 12 8.5 3.8
Acetonitrile O.IMNaOH
13
IMKCI O.IM NaZS04 1M KCI + borate + 10-3M Fe(CN)~1M KCI + O.OlM KOH + 0.5M K 3 Fe(CN)n 1M KCI + borate
ND (cm- 3 )
1020 3 x 10 18 3.3 x 10 17 5 x 10 17 2.6 x 10 10 2 x IO IR 2 x 10 17 3 x 10 18
log i-V slope
60±5 65 ± 5 65 ±5
Efb (SCE)
-0.46 -0.52 -0.45 -0.65 -0.47 -0.2 -0.75 -0.82
Reference
455 474 950 1033 514 951 919 474
The value of the flatband potential is determined by two factors: variation of the bulk Fermi level and interaction of surface states with the electrolyte. The first is related to variations in dopant concentration because the Fermi level with respect to the conduction band edge, Ec, is equal to kTle·ln(NJND ) for a nondegenerated semiconductor, where Ne is the effective density of states in the conduction band. The flatband potentials of ZnO determined in various solutions are shown in Table 4.2. They vary from 60 to 65 m V per decade in N D, in agreement with the theory. The second factor is subject to variations in surface treatment and the nature of the electrolyte. These variations are manifested by changes in VH , which is the voltage drop across the Helmholtz double layer and depends on the excess surface charge. It is generally found that for non degenerated semiconductors VII is primarily determined by adsorption/desorption processes between the surface and the electrolyte. The contribution from electron transfer between the surface and the bulk of the semiconductor is negligible. This is because the amount of charge stored in the semiconductor associated with this transfer is on the order of 10 12lcm 2 or less, which is very small in comparison with the amount of charge (on the order of 10 15lcm 2 ) associated with the adsorption/desorption processes [1177]. In solutions in the absence of specific adsorption of other ionic species, adsorption/desorption of H+ or OH- is responsible for the excess charge stored on the surface. As reported by Morrison [951], the flatband potential of ZnO in aqueous solutions of various pH values is essentially independent of the presence of different redox couples. The Helmholtz potential of ZnO is primarily determined by H+ and OH- ions and is little affected by the presence of other species. In the reaction
where H30+ is a hydronium ion in solution, the free energy of the reaction varies with the double-layer potential Vfj, as the proton must acquire the potential energy (eVH ) to become adsorbed [1177]. Therefore, (4.4)
102
CHAPTER 4
-0.2 -0.4
~
~
-0.6
w
+ Rest dark potential
-0.8
+
Rest potential
ft Flatband potential
-1
-1.2
o
2
4
6
8
10
12
14
pH
FIGURE 4.9. pH dependencies of the potentials of the ZnO electrode: the flatband potential, the rest photopotential, and the rest dark potential. After Matsumoto el al. [474].
where A is a constant. The double-layer potential is proportional to the charge adsorbed. Assuming that CH , the capacitance of the Helmholtz double layer, is independent of VH , the relation between the charge adsorbed and the double-layer potential can be described as (4.5) Since [W] only varies slowly with [H30+], according to Eqs. (4.4) and (4.5) and as an approximation one obtains VH = B + kT/e . In[H30+] = B - 0.059pH
(4.6)
(where B is a constant) which indicates that the Helmholtz potential decreases about 59 m V per pH unit. The flatband potential then also varies 59 m V per unit pH since it is expressed as (4.7) The approximately 60-m V/decade variation of the flatband potential with pH on a ZnO electrode was first confirmed by Lohmann [1196]. This relationship has also been reported in several studies as shown in Fig. 4.9 [474, 514]. As illustrated in Fig. 4.18, there is also a linear dependence of the flatband potential on pH for passive films formed on a zinc electrode [526]. The flatband potential depends on the crystallographic plane and on the surface condition of ZnO [737]. Dewald [514] found that the flatband potential for a ZnO crystal etched in H 3P04 solution was 130 mV more positive than that for one etched in KOH solution. The effect of etching was explained by Dewald on the basis of acid-base equilibrium. On an ideal {1120} surface, there are equal numbers of zinc and oxygen atoms. Each of the surface oxygen atoms has an unshared pair of electrons, and each zinc atom
ELECTROCHEMISTRY OF ZINC OXIDE
103
- 0 . 4 8 , - - - - - - - - - - - - -- -- - - - - - - - , • First run
-0.5 Ul
'I'
:;:
G. ~
Second run
+ Third
-0.52
run
C
'" -0.54 -0 a.
-0
c:
1i
-a
u:
.f
-0 .56
+ -0.58
---'J..2- - - ,'-6- - - - -2'0----=-'24
-0. 6 0L ----'4---..J. 8
Time after etching, hours
FIGURE 4.10. Variation of the flatband potential with time after etching in H,P0 4 . The different symbols correspond to three successive runs on the same crystal. After Dewald [514].
has an empty pair of orbitals. These oxygen electrons are shared, and the zinc orbitals are filled by bonding with the ionic species in the solution, which at equilibrium determine the flatband potential. Different etchants may result in different ratios of surface zinc atoms to oxygen atoms and thus change the excess charge on the surface. Figure 4.1 0 shows that the flatband potential tends to change with time in the solution [514]. This gradual change is attributed to slow dissolution or corrosion of the surface leading to a new surface condition which is independent of the initial surface treatment. According to Dewald [514], the direction of the change depends on the nature of the surface treatment, while the final limiting value is independent of the nature of the surface treatment. The frequency used for measuring the capacitance for the Mott-Schottky plot has been found to have varied effects on the value of the flatband potential. Dewald [5141 reported that the Mott-Schottky plot measured in 1M KCl at pH 8.5 is essentially independent of frequency from 100 to 10,000 Hz. On the other hand, Vanden Berghe et al. [1033] found that the flatband potential of ZnO in 1M KCl containing 0.5M Fe(CN)~- at pH 12 varies by 70 m V over the frequency range 130-10,000 Hz. The surface states, energetically distributed over the whole range of the band gap, are often considered to be responsible for the frequency-dependent flatband potential. 4.3.3. Band Structure The energetic positions of conduction and valence bands of a semiconductor at the surface depend on the interaction between the semiconductor and electrolyte through ionic and electronic exchange at the interface. They are a function of the semiconductor material and of the composition of the electrolyte. The conduction band edge in the bulk for an n-type semiconductor, the valence band edge for a p-type semiconductor, and the Fermi level differ by a fixed amount of energy determined by the doping concentration (e.g., about 0.1 V for moderately doped semiconductors). The position of the band edges at the
104
CHAPTER 4 E (NHE)
....L..--r------~./.-----
E, ____-,__________.-~---
3.2 V
o
E,
ZnO 2
rI
I
3
Ev ----~-----------------
semiconductor
En.
0
L, 4
ohp
electrolyte
interface FIGURE 4.11. Band structure of moderately doped ZnO in the dark at pH 7. ohp. Outer Helmholtz plane.
surface can be determined through the flatband potential using the Mott-Schottky equation (Eq. 4.3). The band structure can be illustrated for given values of the flatband potential, doping concentration, and rest potential. Figure 4.11 shows the band structure of moderately doped ZnO in the dark at pH 7. At this pH value the potential drop in the Helmholtz double layer, VH , is estimated to be about 0.1 V using the pH value of 8.8 at the point of zero charge, where VH = 0 [1181]. The point of zero charge, according to Blok and De Bruyn [1181], is mainly influenced by the presence of impurities in the oxide and the nature of the electrolyte. Several other factors may also influence the value of the point of zero charge, as reviewed by Parks [1185]. The band bending for the ZnO semiconductor in Fig. 4.11 is 0.53 V. Assuming a dopant concentration of 1017/cm3, the width of the space charge layer is calculated to be about 0.07 ,urn according to Ls= (2V,cocfeND)112 [1177]. Accordingly, the amount of immobile charge in the space charge layer is Qs = eNoL, = 1.1 x 10-7 Cfcm 2, and the field strength at the surface is Cs = eNDL/ceo = 1.5 x 105 V fcm. This band structure is a function of the flatband and rest potentials. Figure 4.9 shows that the rest dark potential ofZnO in 0.IMNa 2S0 4 is essentially independent of pH [474]. This means that the degree of band bending, V" also increases with increasing pH since the rest dark potential Vrd is expressed as (4.8) The variation of V, is caused by the charge transfer between the surface and the bulk semiconductor, which has little effect on the magnitude of VH for nondegenerated semiconductors [l177]. The reason V, varies with pH, according to Matsumoto et at.
105
ELECTROCHEMISTRY OF ZINC OXIDE
[474], is that the protons and hydroxyl ions not only adsorb on the ZnO surface but also react with the metal ions on the surface. This leads to a change of the valence of the metal ions so that the potential drop across the space charge layer changes with pH according to
v, = const. -
(4.9)
0.059pH
Combining Eqs. (4.6)-(4.9), one obtains Vrrl
= const.
(4.10)
4.3.4. Electrode Kinetics in the Dark
The typical semiconductor electrode is characterized by the phenomenon that in the dark current can only flow in one direction, depending on the type of semiconductor. For an n-type semiconductor at anodic potentials (reverse bias), the current in the dark, being limited by the availability of holes, is close to zero and is essentially independent of the potential. At cathodic potentials (forward bias), the current is not limited, and it increases with decreasing potential bias because electrons, the majority carrier, are responsible for the current flow. The magnitude of the cathodic current at a given potential depends on the concentration of oxidizing agents in solution while the anodic current is essentially independent of the presence of reducing agents. The electrode kinetics of ZnO have been extensively investigated by Freund and Morrison [892, 921,950,9511. The anodic current in the dark is found to be very low, typical for an n-type semiconductor. For example, the dark anodic current density on ZnO in 1M KCl is less than 5 nA/cm" up to a potential as high as 10 V seE [9511. Under a forward bias in a solution containing Fe(CN)~-, the cathodic current is proportional to the concentration of Fe(CN)~-. A plot of the logarithmic current versus potential over the current range 0.02-20 j1A/cm" gives a 60 m V change in the surface barrier per decade of current change. This behavior is consistent with the simple theoretical model for a semiconductor electrode under a forward bias. The rate of reaction is first-order in the density of electrons at the surface, n, = ND exp( -e V/kT), and first-order in the density of unfilled states at the solid/electrolyte interface (adsorbed Fe(CN)~- ions), no" and can be expressed as (4.11 ) is the number of where J is the cathodic current density, h' is the rate constant, Fe(CN)~- ions per cubic centimeter in the electrolyte, and 11., is the number of conduction band electrons per cubic centimeter at the semiconductor surface. The linear dependence of current on Fe(CN)~- concentration indicates that the adsorption of ferricyanide ion on a zinc electrode follows a linear isotherm. It is also found that the current is independent of the presence of Fe(CN)t, indicating that the reaction is irreversible; i.e., electrons are not transferred from adsorbed Fe(CN)t to the conduction band. This effect is an indication that the filled energy level is far below the conduction band. The cathodic current-potential relationship described by Eq. (4.11) has been observed in solutions containing various redox couples, as shown in Fig. 4.12 1951]. All the 110X
106
CHAPTER 4
curves give slopes in the range of 65 ± 5 mV/decade. The I/C2 versus potential plots give straight lines, agreeing well with Eq. (4.3) [950]. The 60-m V slope for log i versus Vand the linear Mott-Schottky plots suggest that the rate-limiting step in the reduction process is the transfer of electrons within the space charge layer. Thus, the zinc electrode behaves like a typical semiconductor: the Helmholtz potential, VH , does not change with the applied potential, and the potential change occurs only across the space charge layer, V,. Figure 4.12 also shows that for a given current density the surface barrier required in different solutions is different. The different barrier heights are related to the electrode capacitance and can be determined from capacitance measurements. Table 4.3 shows that at a given cathodic current density the capacitance of a ZnO electrode in 1M KC1 solution 10.3r-----.----.-----.-----.-----.-----.----.-----,
o o
10 ' ·CeIHSO.I •. wH IS . 1000 11 lace IQ· J KMnO •. p H 87 100011
£>
IO ' 2M KMnO • . pH ~ 5 1000' 1
• ..,
1O' 2M INH.1 2 I
o
10·2M F. II I
"
1O. 2 M A9INH.12NOJ pH 12 (00011
g
IO· 3M Fe tIl Fhenaolhfoh~ln pH 1.5 (0001]
Cv.n.de.
pH 12.0 100011
10 ,5
'"E
~
i-
0 0
'Vi C
~
....C
10.6
'1:::1
~
U
10'7
10· ~L_
o
__
~~
__
~
____
~
____
~
____- L____- L____ __ 06 07 08 ~
~
Surface barrier Vs, V FIGURE 4.12. Variation of the cathodic current with the surface barrier, for various one·equivalent oxidizing agents. Reprinted from Morrison [951], with kind permission from Elsevier Science-NL, Sara Burgerhartstraat 25, 1055 KV Amsterdam, The Netherlands.
ELECTROCHEMISTRY OF ZINC OXIDE
107
TABLE 4.3. Effect of Different Oxidizing Agents on the Capacity of a ZnO Electrodea Measured during Cathodic Reactions in 1M KCl Solutions b Substance added
Capacity (nF)
O.IMHCOO-
°2
02 + O.IM HCOO0.IMH 20 2 O.IM H20 2 + 0.2M HCOO-
600 580 113 108 60
62
"Geometric area = 0.2 cm': dark current = -5.0 ± 0.1 nA. hRef. 921.
varies with the addition of different oxidizing agents [921]. With no oxidizing agent present, the capacitance is high, corresponding to a large surface barrier (large VJ, Table 4.3 also shows that both H20 2 and O 2 are very active oxidizing agents on ZnO while the formate ion is not active. According to Morrison [951], since the Helmholtz potential of the system is independent of the kind of redox species in the solution, this difference in surface barrier must be due to the difference in electron reactivity, O'[X], with 0' the cross section of the surface state for electron capture and [X] the concentration of the unfilled surface states. For different oxidizing agents, both 0' and [X] can be different. While the amount of unfilled surface states is determined by adsorption of the oxidizing agent, the cross section depends on the extent of overlap between the energy levels in the conduction band on the surface and the energy levels in the adsorbed oxidizing agents. Because different agents have different redox potentials, overlapping of the energy levels in the solution and in the bands of the semiconductor varies. ZnO electrode behavior in the dark has been studied by different techniques. Kohl and Bard [919] measured the cyclic voltammogram of a ZnO electrode in a nonaqueous (acetonitrile) solution containing various redox couples. They compared the reduction/oxidation potentials of the couples obtained for ZnO with those obtained on a Pt electrode. Gomes and Cardon [939] used electrochemical noise measurements to study the oxidation reactions of 1- and aliphatic alcohols on a ZnO electrode. Bindra et al. [477] studied lead deposition on zinc oxide in 1M KN0 3 solution. Eger et al. [806] found that a very strong accumulation layer with a surface electron density approaching 1014/cm2 is achieved on a ZnO surface in 2M KOH solution using a constant-pulse technique. The charge density and thickness of accumulation layer was found to be a function of the barrier height. Breakdown. Breakdown of a semiconductor electrode made from a nondegenerated semiconductor occurs when the near-zero current at reverse bias sharply increases with increasing reverse potential bias. At breakdown the electrode loses its "insulating" character and becomes "conductive." Breakdown is observed on the zinc oxide electrode when itis anodically polarized to high potentials as shown in Fig. 4.13 [946]. The potential at breakdown generally decreases with increasing dopant concentration. As a result of the breakdown, both dissolution of zinc oxide and formation of oxygen from water oxidation occur. The ratio of zinc oxide dissolution to oxygen formation is found to decrease with
108
CHAPTER 4
12,------------------------------------.---,
Eu
2,
>-
10 8
'iiic
6
C
4
CD -0
~
::; <.)
No = 10 '6 cm·3
2 OL---__ 1
~
__
~
__
~
5
_____ L_ _ _ _ 10
~~
____
~
____
50
~
100
Electrode potential E (VscEl
FIGURE 4.13. Potential dependence of breakdown currents at ZnO electrodes with different doping concentration N D • Electrolyte: 1M KCI. pH 6; sweep rate: 50 mV/s. After Pettinger et al. [946).
increasing current. This ratio varies with the crystallographic orientation of the zinc oxide electrode, indicating the specific catalytic effect of surface structure on different reactions. Breakdown of a ZnO electrode at high anodic potentials in aqueous solution is associated with an electron tunneling process. According to Pettinger et al. [455], by applying high voltages to a moderately or highly doped semiconductor electrode, the thickness of the potential barrier for energy levels in the upper part of the forbidden zone is reduced so that electrons can tunnel through the barrier from the redox couple in the solution into the conduction band, as illustrated in Fig. 4.14. The probability for an electron to tunnel increases with decrease of the thickness of the space charge layer [1197]. Thus, the tunneling current increases with increasing dopant concentration since the thickness of the space charge barrier is a function of dopant concentration according to L, = ( 2Vl4;/eND) I 12. This equation also indicates the dependence of the current on potential at breakdown. Since the majority of the applied voltage is across the space charge layer and its thickness is a function of the voltage, the tunneling current can be changed by several orders of magnitude by varying the potential. According to the model illustrated in Fig. 4.14, the tunneling current also depends on the energy levels of the reducing agents and their concentrations in the solution. Thus, at breakdown, at a given potential the anodic current may vary depending on the kind of reducing species in the solution. In addition to the tunneling process from a redox level in the solution into the conduction band, band-to-band tunneling has also been considered possible [946]. However, at a band bending less than 3.2 V, electron tunneling from the upper edge of the valence band to the conduction band is not possible for ZnO, which has a band gap of 3.2 eV Electron tunneling does not appear to be only an anodic process on zinc oxide. Pettinger et al. [455] also observed a cathodic current on a highly doped ZnO electrode at a potential as high as 0.7 VseE. at which considerable band bending occurs. Since the electron concentration at the surface is very low at such a potential, cathodic reaction can
ELECTROCHEMISTRY OF ZINC OXIDE
109
E
electron donor
3.2 V
ZnO electrolyte
Ev----~--------~
FIGURE 4.14. Model representation of electron tunneling from the donor in the electrolyte into the conduction band. After Pettinger et al. [4551.
semiconductor interface
only occur with the participation of bulk electrons. Furthermore, for a highly doped ZnO electrode, the barrier thickness can be sufficiently thin for tunneling to occur at a relatively small band bending. The cathodic current observed at a rather positive potential can then be attributed to the electron tunneling from the conduction band to the oxidizing agent in the solution. Therefore, it appears that electrons can tunnel through the space charge barrier of highly doped ZnO electrodes in both directions when proper redox couples are present in the electrolyte to act as the electron donor and acceptor. 4.3.5. Photo electrochemical Kinetics
When a ZnO electrode/electrolyte interface is under illumination, a photopotential or a photocurrent can be measured depending on whether or not the electrode potential is controlled. The photopotential and photocurrent are essentially the result of the change in the surface concentration of electrons and holes induced by absorption of light. The concentrations of electrons and holes are determined by
n, = no exp(eV/kT)
(4.12 )
p, =Po exp(-eV/kT)
(4.13)
Under a moderate illumination intensity, the concentration of the majority carrier in the bulk of an n-type semiconductor, no, will change very little because the number of electrons generated by moderate illumination is very small compared to no. However, the
110
CHAPTER 4
5 3
o
2 E (V",)
FIGURE 4.15. Current-voltage curves for ZnO in HzS04 (pH 3) under illumination of intensities I L .) and I LZ ' I dark = 10-6 _10- 7 Ncm 2. After Gerischer [129].
minority carrier, Po, can be drastically changed, by orders of magnitude, under illumination. Thus, according to Eq. (4.13), the surface hole concentration is greatly increased as a result of illumination. Figure 4.15 shows an example of thephotocurrent observed on a ZnO electrode in H 2S04 solution [129]. The photocurrent increases with increasing anodic potential since the concentration of surface holes increases with increasing surface potential barrier V,. Above a certain potential, the photocurrent reaches a constant value, the saturation current, when all the photogenerated holes available at the surface are exhausted by the anodic reactions. The saturation current depends only on bulk properties of the semiconductor for a given light intensity [793]. Theoretically, photocurrent is directly proportional to the light intensity 10 and is a function of the space charge layer thickness L" the hole diffusion length L,,, and the absorption coefficient a, as described by Eq. (4.14) [1183,1187]. i photo oc
eIo{ 1 - [exp(-aL,)/(l + aLp)]}
(4.14 )
The depth of light penetration in a semiconductor depends on the wavelength of the light. For example, with 370-nm-wavelength light the penetration in ZnO is about 0.2 Ilm [1188]. The net current and the quantum yield are a function of a number of competing processes as described by Gerischer [793] in Fig. 4.16. For an n-type semiconductor, the externally measurable current i is the difference between the photocurrent and the forward current of electrons. The majority current is decreased to zero at a certain anodic bias. The flux of holes to the surface, defined as idf' is exclusively controlled by the solid-state properties, while all the other reaction steps depend on the surface properties of the semiconductor. The holes arriving at the surface can either be transferred to the electrolyte by a reaction with a redox couple, by an oxidative or reductive decomposition process, or recombine via surface recombination centers. The measured photocurrent, ipho,o, can therefore largely deviate from idf' Consequently, photocurrent-voltage curves of different
1II
ELECTROCHEMISTRY OF ZINC OXIDE Electrolyte
E
n-type semiconductor
Conduction band
~---''''----~'f-~~=-''::....l..;oL...::-=-~- Ee
-E F Surface recombination
Difussion current
hv
Light absorption
Recombination
Valence band
FIGURE 4.16. Kinetic processes controlling the photocurrent yield. Reprinted from Gerischer [793], with kind permission from Elsevier Science-NL, Sara Burgerhartstraat 25, 1055 KV Amsterdam, The Netherlands.
semiconductor samples can have very different shapes and depend differently on changes in the experimental conditions. Thus, in the absence of surface recombination and with a fast rate of electron transfer, the photocurrent increases steeply when the depletion layer starts to form, and a saturation current is quickly reached. On the other hand, with sufficient surface recombination or in cases of slow electron-transfer reactions, the apparent onset of the photocurrent is shifted to higher bias, and the saturation current is only reached at a larger band bending. The crystal orientation, surface structure, and surface treatment have been demonstrated to affect the surface recombination processes [793]. It is found that mechanically induced defects near the surface of ZnO single crystals act as hole traps. which are efficient recombination centers [1188]. A ZnO electrode under illumination exhibits a number of phenomenon including current doubling, photocatalysis. and electroluminescence. Also. single-crystal ZnO is found to be photosensitive to light with energy less than the band gap in a solution containing agents, such as ferrous ion, which act as a dye sensitizer [949]. The sensitization is attributed to electron injection from the photoexcited adsorbed-dye surface states. Masuda et al. [792] detected an acoustic signal during photoelectrochemical reactions at a ZnO electrode in O.2M KN0 3 + 0.0 1M Pb 2+. The acoustic signal was found to associate with the deposition of an oxide layer of Pb0 2 and to be a function of the light wavelength. 4.3.5.1. Current Doubling. Current doubling is a phenomenon observed in the anodic reaction of a ZnO electrode under illumination in an electrolyte containing two or more equivalent reducing agents. In current doubling an anodic current is produced, which is approximately double the original limiting hole current generated by illumination [921,1032]. A two-step mechanism has been suggested for this process (4.15) (4.16)
112
CHAPTER 4
TABLE 4.4. Anodic Reactions on Illuminated ZnO in O.IM KCl at 2 VSCE with and without the Addition of Various Substances" Added substance(s) None HCOOHCOO- +0 2 HCOO- +H 20 2 H 20 2 O2
Current (nA)
H20 2 produced (molecules/hole)
200 350 200 200 200 200
0.1 0 0.1 0.1
"Reprinted from Morrison and Freund [8921. with kind permission from Elsevier Science Ltd. The Boulevard. Langford Lane. Kidlington OX5 1GB. United Kingdom.
where A represents the two equivalent reducing species at the surface, which can be As 3+, CW, HCOO-, methanol, or ethanol. A is oxidized by a hole to become a radical A+. Because of its closer position to the conduction band edge, A+ is further oxidized by injecting an electron into the conduction band to generate the current doubling effect. The occurrence of current doubling on a ZnO electrode depends also on the presence of other oxidizing chemical species. Table 4.4 shows the anodic currents measured under illumination at 2 V SCE in O.IM KCl solution, with or without an oxidizing agent [892]. As seen in the first row, hydrogen peroxide is produced when no active chemicals are present, and the yield is one molecule per 10 holes. With the addition of O.IM formate ions, a current doubling agent H20 2 is no longer produced and the anodic current is almost doubled. The third and fourth rows show that in the presence of oxidizing agents the current doubling effect offormate disappeared. According to Morrison and Freund [892] with no active chemicals present in the electrolyte, the hole p+ oxidizes the lattice oxygen of the ZnO according to (4.17)
When formate is added, the holes now react with formate ions: (4.18)
The radical HCOO· is a highly reactive form, and it spontaneously becomes oxidized by injecting an electron into the conduction band of the ZnO: (4.19) Therefore, for each hole reacted, an electron is injected, leading to current doubling. If an oxidizing agent is introduced into the system (third row in Table 4.4), the free radical reacts with the oxidizing agent rather than injecting an electron into the conduction band: (4.20)
ELECTROCHEMISTRY OF ZINC OXIDE
113
Fujishima et al. [1171 proposed a different mechanism for the current doubling process, with the dissolution of ZnO as the first step. They observed the formation of zinc ions during current doubling on ZnO in O.4M KN0 3 solution containing HCOONa due to the dissolution of the ZnO electrode. The lifetime of the radicals generated during current doubling is short. As was investigated by Cardon and Gomes [1032], the lifetime of the intermediate radicals formed on ZnO during the oxidation reactions is not longer than 10-5 s. The efficiency of current doubling is affected by the presence of nonreactive anions. Micka and Gerischer [942) studied the effect of the presence of various anions on the current doubling of zinc oxide in O.OlM HCI0 4 + 0.00 1M HCOOH. They found that many anions have a suppressing effect on the current doubling reactions on ZnO. The suppressing effect on the photocurrent was found to decrease in the order 1- ::c: Br- > CI- > H 2P04 > NO) > SO~- ::c: ClO:;. According to Micka and Gerischer, the results indicate the importance of anion adsorption at the surface of a ZnO crystal in the reactions with holes. Those anions which have a strong tendency to form complexes with Zn 2+ cations are adsorbed preferentially and react with holes, preventing the oxidation of formic acid or methanol. In strong alkaline solutions, current doubling disappears, also indicating a competitive process of hole capture by OH- ions. 4.3.5.2. Photocatalysis. When a semiconductor is illuminated and electrons and holes move to the surface, the electrons will reduce chemical species at the surface; the holes will oxidize the species at the surface. The chemical changes will occur in the medium in which the semiconductor is immersed. The process of producing these chemical changes is termed photocatalysis [1177). For an n-type semiconductor the hole current flowing to the surface is proportional to light intensity and is dependent on V, as described by Eq. (4.14). The electrons and holes arriving at the surface may recombine with no net chemical change in the solution. On the other hand, catalysis occurs when a net chemical change results from the reactions of the electrons and holes wi th the chemical species in the solution. The photocatalytic effect ofZnO on chemical reactions in aqueous solution has been extensively investigated by Morrison and Freund [921,1177). They found, as an example, that formate ions only react with O 2 when a ZnO electrode is present. The electrons and holes reaching the ZnO surface cause the reduction of oxygen and oxidation of formate ions, respectively, according to the following sequence of reactions: (4.21 ) (4.22) (4.23) (4.24) The quantum efficiency of ZnO-catalyzed reactions depends on many factors. Cunningham and Zainal [807) found quantum efficiencies ranging from 10- 3 to 0.29 for reactions produced by UV illumination of zinc oxides suspended in solutions containing agents such as NaN0 3 and KMn0 4 • Hada et al. [944] reported quantum yields of
114
CHAPTER 4
0.24-0.39 for the reduction of Ag+ by the irradiation with 365-nm light of a ZnO surface in aqueous solutions containing 5 x 10-5-1 x 1O-3M AgCI04 • The electrons generated by irradiation reduce the Ag+ ions adsorbed on the ZnO surface. Bernas [1188] found that the quantum yield of H2 0 2 production in a zinc oxide suspension under illumination is increased when appreciable quantities of interstitial zinc atoms are present in the oxide.
4.3.6. Electroluminescence Electroluminescence is a process in which light of a certain wavelength is emitted from a semiconductor electrode as a result of the electrochemical reactions. Electroluminescence has been observed on zinc oxide under both forward and reverse bias [798, 1038]. Under reverse bias, the light originates from the recombination of electron-hole pairs either by direct band-to-band recombination, when the corresponding luminescence is centered at 390 nm, or by recombination at self-activated centers intrinsic to ZnO, when the light emission is broad and centered around 550 nm. Under forward bias, in addition to the luminescence from the self-activated centers, a light emission is observed at 728 nm which is attributed to the centers activated by electron impact. Luminescence may occur during reduction reactions on a zinc oxide electrode [1035]. The light emission is attributed to homogeneous electron transfer between radical cations and anions which are generated during reduction on the zinc oxide electrode. According to Pettinger et al. [1036], these radical species have very high electron affinity are normally required in order to inject holes into the deep-lying valence band of ZnO. Although not stable in aqueous solution, short-lived species such as S04- and OR are
I
e
- ·R 4
I I
I I3 I
v, > E,
hv I
2
R
ZnO
E, ----'-- - - semiconductor
e lectrolyte
AGURE 4.17. Band structure of a highly doped semiconductor/electrolyte interface. (I) Band-to-band tunneling under the condition V,2: Eg . (2) Reaction of holes with constituents of the electrolyte. (3) Generation of radicals by the process. (4) Electron injection from the radical. (5) Recombination of electrons and holes. After Pettinger et al. [946].
115
ELECTROCHEMISTRY OF ZINC OXIDE
TABLE 4.5. Flatband Potentials, Efb , of Non-Single-Crystal ZnO Electrodes in Various Solutions Solution INKOH INKOH O.IMNaOH O.IM Na3P04 + O.IM Na2HP04 0.02MNaOH
Fonn of zinc oxide Anodic film Polycrystalline Anodic film Anodic film Particulate films
Efb
(SCE)
-0.86 -0.86 -0.77 -0.69 -0.6
Reference 423 423 484 526 580
found to have a long enough lifetime to inject holes into the valence band of ZnO when they are generated by an electrochemical reaction. Anodic luminescence is associated with breakdown such that an anodic reaction becomes possible in the dark [798, 946]. The emission intensity increases with current, and the maximum emission intensity lies at a wavelength of 390 nm, corresponding to the energy of the band gap. According to Pettinger et al. [946], as shown in Fig. 4.17, the anodic breakdown of zinc oxide at high anodic potential is due to two tunneling processes: one involves a band-to-band tunneling from the valence band to the conduction band, and the other involves an interface tunneling from an electron donor in solution to the conduction band of the semiconductor. The holes generated by the band-to-band tunneling are consumed in the anodic dissolution of the zinc oxide. The photoemission occurs only when the two tunneling processes proceed simultaneously because the emission of photons is only possible if both kinds of charge carriers are available for the recombination to occur. 4.4. THIN ZnO FILMS Thin zinc oxide films can be formed by anodization, by heating of metallic zinc, or by physical and chemical deposition [113,423,788,948]. Also, the passive films formed on a zinc surface under many conditions are essentially thin zinc oxide films. Compared to the physical characteristics of bulk single-crystalline ZnO, those of thin-film ZnO, being polycrystalline in general, vary greatly. For example, it has been reported that the conductivity of a thin oxide film formed by thermo-oxidation depends strongly on the environment because of its large surface-to-bulk ratio [788]. The tlatband potential and band gap of thin ZnO films are generally found to be similar to those of a single crystal of ZnO. Table 4.5 shows that the tlatband potentials determined for zinc oxide films and other non-single-crystal zinc oxides are close to that of bulk ZnO (see Table 4.2). Similar results are shown in Fig. 4.18, reported by Bothe et al. [526], for the tlatband potentials at anodic passive films formed in solutions with pH values ranging from 8.4 to 12.9. An almost 60 mY/decade pH dependance indicates that the potential drop in the Helmholtz double layer is primarily due to the adsorption/desorption of W or OW ions. Figure 4.19, reported by Burleigh [423], illustrates that the variation of quantum yield with photon energy for anodic films formed in 1M KOH solution depends on the formation potential. The film formed at -0.5 Y has a similar band edge to that of polycrystalline zinc oxide. Photocurrent has been observed for a very thin
116
CHAPTER 4
- 0 . 4 , - - - - - - - - - - - -- - - - - - - - - - - , • Zinc oxide film . . ZnO single crystal
'""UJ -0.6
~ re
w -0.7
-0.8 -0.9
L..-_ _--'-_ _ _- ' -_ _ _- ' -_ _ _L...._ _......1._....1
8
11
10
9
12
13
pH
FIGURE 4.18. Dependence of the flatband potential Etb on the solution pH for ZnO film and ZnO single crystal obtained from the extrapolation of the corresponding Mott-Schottky plots. Reprinted from Bothe eT al. [526], with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom.
(10-20 A) passive film fonned on a zinc surface in 1M NaOH + O.IM borate solution of pH lO.5 [1184]. Scholl and Prentice [354,422] measured the photocurrent spectra of the passive films formed on a zinc electrode in KOH solutions. They found that the photocurrent spectra at the rest potential exhibit a strong ultraviolet response with a shoulder tailing into the visible region as illustrated in Fig. 4.20. The shoulder is attributed to the existence of interband energy levels characteristic of an amorphous film. There are clear differences in the photospectra of the passive film formed in different potential regions, indicating differences in composition and structure of the films.
2
0
- ZnO polycryslal (1 x)
0.4
.Zn: ·o.so Vsce (100x)
(5
.s= a. C;;
c
e
+ Zn: ·1.00 Vsce (1oox)
0.3
u
Q)
~ "C
a;
0.2
';;'
E :::I
C
a'"
0.1
:::I
0
1
2
3
4
5
6
Photon energy (eV)
FIGURE 4.19. Photospectra for anodic oxide films on zinc formed at - 1.00 VSCE (4 hand id =5.8 mAlcm 2 ) and -0.50 VSCE (2.5 hand id =0.02 mAlcm 2) and polycrystalline ZnO. From Burleigh [423]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
117
ELECTROCHEMISTRY OF ZINC OXIDE
g.--------------------------------------, •
c~o _ :)
N. t':
... C~t _ O N.
•
.....
t
OIl
-t ."IO Y, P.C .
-1.40e v, P.C . C'I .O H. E • -1.<00 V. s.c. ~
~
0
t::
..... ..... ;:\ u
80
.. 0
0
..c::
0.. "0
• •• • •• • • .
~ 0
oa . t:::l
8..... 0
Z
0
o
100
.00
~oo
•
• •••
.
0 0
.
•
•
•
•
••
0 N
•
eoo
•
• 700
800
Wavelength (run) Fl(JURE 4.20. Photocurrent spectra at the rest potential (Ere") in stagnant O.SN and I.ON KOH. T = 293 K, for polycrystallinc WC) and single-crystal (S.C) zinc. The solid line is a B-spine fit of the circles. After Scholl ~t ul·14221·
According to Scholl and Prentice [3541, the photocurrent measured on zinc passive films in alkaline solution is linearly dependent on the illumination power over two orders of magnitude. The photocurrent spectra show that the band edge corresponds closely to that of the bulk material, which suggests that the passive film is largely composed ofZnO. Hotchandani and Kamat 1580J investigated the photoelectrochemical behavior of ZnO particulate films. They found that the photoelectrochemical properties are very similar to those of hulk ZnO. At wavelengths shorter than 320 nm, most of the incident photons are absorbed by the film. The efficiency of the incident photo-to-current conversion at this wavelength was found to be 15%. The electrochemical properties of thin zinc oxide films vary greatly because of the variations of composition, structure, and morphology in films formed under different conditions. Novak and Szucs 1796] found that the solid layer formed on the surface of the zinc electrode during anodic polishing in an organic acid exhibits the characteristics of an n-type semiconductor. According to Deslouis et at. [113], the photoelectrochemical effects induce the dissolution of ZnO or the dissociation of H20 hut do not directly alter oxygen reduction on the ZnO film. Sengupta and Chatterjee [948] found that a zinc oxide film electrochemically deposited on a zinc surface has a very high surface state density, which causes a pinning of barrier light. Fruhwirth et al. [847] found a linear relation with a slope of unity between the photopotential and the electrode potential for a single-crystal ZnO electrode in O.IM Na 2 B4 0 7 ·lOH1 0. By comparison to the single crystal, the zinc oxide film formed on a corroded zinc surface yields much smaller photopotentials, and the potential dependence is not linear. The smaller photopotentials of a corrosion product layer is attributed to the high doping and thinness of the layer.
118
CHAPTER 4
TABLE 4.6. Donor Concentrations, ND, and Flatband Potentials, E fb' Determined from the Linear Portion of Mott-Schottky Plots a
ND
E fb (V)
(I020/cm 3)
Sample Zn Zn-O.4%Co Zn-0.6%Co Zn-1.0%Co Zn-I.2%Co Zn-0.2%Ni Zn-1.2%Ni Zn-6.0%Ni Zn-12.0%Ni
2.8 3.3 4.1 5.0 5.7 3.9 5.1 8.7
-0.77 -0.76 -0.75 -0.74 -0.73 -0.75 -0.69 -0.70 (-0.70?)
"Ref. 484.
Passive films formed on Zn-Co and Zn-Ni alloys are found to exhibit similar properties to ZnO films [373,483,484]. These films behave like n-type semiconductors. Table 4.6 shows the flatbandpotentials and dopant concentrations determined from Mott-Schottky measurements for various Zn-Co and Zn-Ni alloys. The donor concentration of the passive films is very high and appears to increase with increasing Co and Ni concentrations. The band gap of the passive films was found to be a function of the electrode potential as shown in Fig. 4.21 [484]. The physical structure and the degree of surface inhomogeneity of the passive films on Zn-Ni and Zn-Co alloys were investigated by Juttner and Lorenz [373,483] using electrochemical impedance spectroscopy.
3 .35 ,---------------------------------------,
•
3.3 f-
> ~3.25 w'"
3 .2
"
.
• Zn· l .2% Co ... Zn· l .2% Ni
" l",
+ Zn
."''''
. • • ZnO single crysta l
"''''", '" '" ... tt++ "'", "''''''' •• + + •• + + ,
••
. ............ . ... . ... .......... '+'-'1-" '+'"
3 . 15 L-----------~-------------L------------~
-1
0
2
E
(Vsce)
FIGURE 4.21. Band gap energy, Eg , as a function of the electrode potential of a ZnO single crystal and of passive layers on Zn, Zn-1.2% Co, and Zn-1.2% Ni. After ViJche et al. [484).
ELECTROCHEMISTRY OF ZINC OXIDE
119
4.5. STABILITY 4.5.1. Conditions of Stability and Decomposition Reactions
The stability of ZnO in aqueous solution is a function of pH according to the pH-potential diagram (Fig. 2.2). ZnO is thermodynamically stable in the pH range between 6 and 12. In solutions with low or high pH values, ZnO is not stable and dissolves owing to the relatively high solubilities of zinc in these pH ranges. In acidic solution the dissolution of zinc oxide in the dark is attributed to both chemical and electrochemical reactions [1187]. The chemical dissolution is due to direct reaction with hydrogen ions: (4.25) and the electrochemical reaction is (4.26) This reaction is electrochemical because it depends on the electrode potential to regulate the adsorption of protons on the surface of the zinc oxide. In solutions other than acidic or strongly alkaline ones, zinc oxide is usually stable under anodic polarization but dissolves when breakdown occurs under high anodic potentials. Zinc oxide may not be stable under illumination and may decompose according to (4.27)
leading to dissolution of the semiconductor. At the open-circuit potential, the photoinduced decomposition is accompanied by a reduction reaction such as 2H+ + 2e- ~ H2 through a conduction band process. The rate of decomposition can be increased with the addition of oxidizing agents to increase the rate of electron consumption. The decomposition rate increases with anodic potential as the concentration of photogene rated holes increases with the potential (Fig. 4.15). According to Gerischer [129], the overall reaction involved in the photodecomposition of ZnO consists of the following steps: ZnO + 4hv
~
ZnO + 4e- + 4p+
Hole + electron generation First hole trapped on surface (slow) Second hole trapped (slow) Formation of oxygen molecule (fast) Zn 2+ leaving surface (fast)
The whole reaction: 2ZnO + 4hv ~ 2Zn;~ + O 2 + 4e-
(4.28)
120
CHAPTER 4 E(NHE)
-I
ZnO + 2H- + 2e- -Zn + H, O
-
o
2H' + 20'" H, ZoO + OH- + 2h- - Zn(OH)" + 1120, H,O + 2h ' '" 2 H" + 1120 ,
3_2 V
ZnO 2
3
E.----~----------~--
rsem iconductor
I
L, electrolyte
4
interface FIGURE 4.22. The relative positions of various redox couples with respect to the edges of bands at the ZnO surface_
The instability of a ZnO electrode under illumination in aqueous solutions is essentially due to the fact that the decomposition potential of ZnO by holes is more negative than the edge of the valence band E,,, as shown in Fig. 4.22 [941, 1176]. Depending on the decomposition product, the decomposition potential may vary according to the reactions: (4.29) 1
ZnO + OW + 2p+ ~ Zn(OHt + 202
E~ = 0.52 VseE
(4.30) (4.31 )
Figure 4.22 shows that the anodic decomposition potential is lower than the redox potential for water oxidation and thus the decomposition is a thermodynamically more favorable reaction. ZnO decomposition is normally the predominant reaction for hole consumption at low current densities. However, water oxidation occurs as a parallel reaction at high current densities [946]. According to Fig. 4.22, ZnO is stable against cathodic decomposition by electron reduction because the position of the decomposition potential is above the conduction band edge. Cathodic decomposition can only occur at high cathodic polarizations when the conduction band edge is shifted to above the position for the cathodic decomposition reaction.
121
ELECTROCHEMISTRY OF llNC OXIDE 50 ~---------------------------------,0.1
Eu ..:
::L
;!-
-50
'c;; c: (J)
-0
C
!----.......----
o ~----__~------------~====~
• Dissolution fate
o
c 'N
0.05 "0 (J)
"§
-100
c:
.2 "S
~
:s
0-150
"0
6 -200L-----~----~------~----~------~0
-2
-1
0
2
3
Potential (V seE)
FIGURE 4.23. Correlation of zinc dissolution rates with the current-potential relation for an 8.5-Q·cm lnO single crystal in the dark in l.OM KOH. After Justice and Hurd [475J.
4.5.2. Rate of Decomposition The dissolution rate ofZnO depends on many factors. Justice and Hurd [475] found that the dark dissolution rate of ZnO in 1M KOH is relatively low and is essentially independent of anodic potential up to 3.0 VSCE as shown in Fig. 4.23. The change of the cathodic current has no effect on the dissolution rate. On the other hand, under illumination, as shown in Fig. 4.24, the dissolution rate is much higher and increases proportionally to decreasing cathodic current or increasing anodic current. In 0.5M KOH solution the dissolution rate is lower compared to that in 1M KOH but the photocurrent is larger,
100,------------------------------ - - , 0.15
1:
i~
"'E
Cl
u
O. I
E 0
c: 'N
"0 (J)
A Di:ssoh.nion ralt
0.05
Poh~"iz.ation cutvl
"§ c:
.2
"S
"0 (f)
6'"
_200L-~--~----~----~------L-----~ 0
-2
-1
0
2
3
Potential (V SCE )
FIGURE 4.24. Correlation of zinc dissolution rates under illumination with the current-potential relation for an 8.5-Q·cm lnO single crystal in l.OM KOH. The open-circuit potential is -650 mY SCE' After Justice and Hurd [4751.
122
CHAPTER 4
indicating that the photocurrent is not fully generated from the ZnO dissolution process. The dissolution in the dark is primarily a chemical process which depends only on the concentration of OH- ions: ZnO + 20H- + HP ~ Zn(OH)~-
(4.32)
The dissolution under illumination according to Justice and Hurd [475], the amount of zinc actually measured in the solutions, is always greater than the amount calculated from the anodic currents. This indicates that under illumination a photo-assisted dissolution occurs in addition to the chemical dissolution process. Thus, the dissolution rate of ZnO is a function of both OH- and hole concentrations at the surface, which are a function of both potential and light intensity. The photo carriers generated under illumination are consumed either for ZnO decomposition or for water oxidation in the absence of other redox couples. According to Fruhwirth et al. [1187], the contribution of each reaction is a function of pH. At a pH of about 9.2, water oxidation is the predominant reaction because the saturation concentration of Zn 2+ ions is about 10-7 molll and will be reached immediately after immersion of the ZnO electrode. With decreasing pH, the solubility of Zn 2+ ions is increased, and the contribution of ZnO decomposition to the photocurrent increases. Fruhwirth et al. [1187] found that the photocurrent displays a linear dependence on pH in a nonsaturated solution whereas it is independent of pH in a Zn 2+-saturated solution. Erbse et al. [1278] found a linear relation between the dark dissolution rate and pH in the pH range between 4 and 6. Gerischer and Sorg [1186] reported that the dissolution rate of zinc oxide in 0.5M KCl solution decreases with increasing pH in the acidic range. In the alkaline pH range, the presence of NH3 leads to a pronounced increase of the rate, which is followed by a decrease at pH 10-13. The dissolution rate depends on the type of anion in the solution and increases proportionally to the square root of the electrode rotation speed. The latter dependence is attributed to the diffusion of H+ ions. Gerischer and Sorg also found that the dissolution rate is independent of pH (independent of band bending) in the presence of acetic acid as the proton donor. It was thus concluded that the dissolution rate of ZnO is controlled mainly by chemical processes and that the electric forces at the interface play only a limited role in the reaction kinetics. According to Gerischer and Sorg, the rate-determining step in the dissolution of ZnO in the dark is hydrolytic bond splitting between Zn atoms in kink sites and 0 atoms in lattice sites, catalyzed by H+ ion donors, on the one hand, and by molecules or anions that coordinate to the Zn 2+ions, on the other. The dependence of ZnO decomposition on crystal orientation was interpreted by Morrison [1177] in terms of hole capture, mainly by different hydroxyl groups on the different crystalline faces. Because of the structure of the wurtzite crystal, only Zn ions are exposed on a (0001) surface; any residual oxygen over the surface of the zinc plane is weakly bonded and will be dissolved into the solution. Then a preferred adsorption of OH- is expected for this plane because the surface sites are all cationic. Holes coming to the (0001) surface may interact and oxidize these adsorbed OH- ions, but lattice ions will not be oxidized. On a (OOOT) face, however, surface oxygen ions are exposed. For this plane, preferred adsorption of H+ is expected, giving OLH- or O~- groups, where OL is a lattice oxygen ion. Holes coming to the surface will then be able to oxidize the lattice
ELECTROCHEMISTRY OF ZINC OXIDE
123
100
+ Single
crystal
.. Sinter
'0 <:: o
"fii "'C .§ 50 4>
~
0;
a. E o
(.)
o·~----~--------------------~~--~
0.0001
0.001
0 .01
Concentrat ion of I' (mol - dm-')
FIGURE 4.25. Percentage of competitive oxidation of ron ZnO electrodes as a function of 1- concentration. Arter Kobayashi et al. [9181.
oxygen to a -1 valence or to the zero valent (0 2) state, enabling lattice oxygen to evolve, followed immediately by the now weakly bonded underlying zinc ions. Thus, the (0001) plane corrodes much more readily than the (0001) plane. The anodic photodecomposition of zinc oxide can be quenched by competitive hole capture with the addition of a redox couple that has a potential more negative than the ZnO decomposition potential [1186]. Kobayashi et al. [918] reported that the percentage of hole capture by anodic decomposition of ZnO is reduced by addition of 1- in the solution, as shown in Fig. 4.25. The ratio of hole capture through r reduction versus that through ZnO decomposition increases with r- concentration but decreases with photocurrent (Fig. 4.26). A moderate effect is observed with the addition of Br- ions, and a very
100
~ .:....
'0 <::
Electrolyte
0
~
"'C
'x 0
concenlr,
50
CD
.~
(mol o dm· 3)
+10- 2 .. 10 .3
'~
a. E
+ 3.16 X 10-4
(.)
• 10-4
0
0 1 E-06
0
0
o
Disk photocurrent, A
FIGURE 4.26. Percentage of competitive oxidation of 1- on a sintered ZnO electrode as a function of the disk photocurrent and C concentration. After Kobayashi et al. [9181.
124
CHAPTER 4
small effect with the addition of Cl- ions. The effectiveness of 1- in quenching the decomposition of ZnO in aqueous solution is due to its more negative redox potential compared with the anodic decomposition potential for ZnO. On the other hand, the redox couples Br-/Br and Cl-ICl are located at potentials more positive than the decomposition potential and therefore are not effective in competing for hole capture.
5 Corrosion Potential and Corrosion Current 5.1. INTRODUCTION Corrosion potential is a mixed potential (also an open-circuit potential or rest potential) at which the rate of anodic dissolution of the electrode equals the rate of cathodic reactions and there is no net current flowing in or out of the electrode. Corrosion current is the dissolution current at the corrosion potential. Corrosion potential and corrosion current are two important parameters which connect the fundamental electrochemistry and the practical corrosion behavior of metals. The value of the corrosion potential indicates the state of a corroding metal while that of the corrosion current reflects the instantaneous corrosion rate at the time of measurement. In this chapter, the fundamental relationship between corrosion potential and corrosion current is described, and the various methods for measuring the corrosion current and their limitations are briefly discussed. The corrosion potentials and currents are then presented according to the specific effects of the various material, solution, and measurement factors. Lastly, the effect of time and the correlation of corrosion current to real corrosion loss data are discussed, and the errors which may be generated in determining corrosion currents are analyzed. Since in most cases oxidation and/or dissolution is the dominant anodic reaction on a zinc electrode and in most studies the corrosion potentials are determined by measuring the open-circuit potentials, no distinction is made in the discussion regarding the corrosion potential versus the open-circuit potential or the rest potential. 5.2. RELATION BETWEEN CORROSION POTENTIAL AND CORROSION CURRENT The basic theory on the relationship between corrosion rate and electrochemical polarization resistance was formulated in the 1950s [693,694]. According to the theory, for a corroding metal there are two coexisting electrochemical reactions: the dissolutiondeposition of the metal, M r + + re- ~ M, and a reduction-oxidation of a species in the electrolyte, Z"+ + ne- ~ Z. Each of these reactions has its own exchange current and Tafel 125
126
CHAPTER 5
~
G
-0.1
~
# ~
> 0
-0.3 ACTIVE 0.01
LO
0.1
Current density
10
100
{J.IA/cm 2)
FIGURE 5.1. Relationship between overvoltage and current for a corroding electrode system consisting of two coexisting electrochemical reactions. After Stern and Geary [694].
slopes so that the corrosion of the metal occurs at a potential, the corrosion potential, at which the total rate of oxidation equals the total rate of reduction: +-
(5.1)
<--
iz + i", = i: + i", f--
~
---?
f-
where i", is the rate of reduction of Mr+ and i", is the rate of oxidation of M, and iz and iz are the rates of reduction and oxidation of species Z, respectively. The corrosion current Ls th!; difference between the dissolution and deposition currents of M, namely, icorr = i", - im . Figure 5.1 illustrates the potential-current relationships for a mixed electrode system. The rates of each reaction under activation control are:
~ = i~ exp(-2.31J /fJJ
(5.2)
~ = i~ exp(2.31J /fJ)
(5.3)
-->
(5.4)
+-
(5.5)
im = i~, exp(-2.3IJm1fJm) ill! = i ~ exp(2.31J ml Pm)
CORROSION POTENTIAL AND CORROSION CURRENT
127
where i~ and i~, are the exchange currents and", and" '" are the overpotentials for the reactions Z"+ IZ andM r +1M, respectively, and /J, and /J", are Tafel slopes for the two reactions, assuming the same values for the forward and backward directions. When the corrosion Eotential, Ecoff ' is sufficiently removed from the equilibrium potential of the reactions, i, and 7,,, become insignificant in comparison to 7, and 7,,,, and the corrosion current can then be expressed as <-
-->
(orr = im = i:
At any potential E deviating from Eco", the anodic dissolution current can be described by an expression analogous to the rate expression for a single redox couple: <--
=im -
->
iz
= icorr[exp(2.3" I/J",) - exp(-2.3" I/J,)J
(5.6)
with" = E - ECOIT"
5.2.1. Polarization Resistance and Corrosion Current For small values of", exp(2.3" I {Jill) and exp(-2.3" I /J,) may be approximated by I + 2.3" I/J", and I - 2.3" I/J,. Equation (5.6) thus reduces to (5.7) Differentiating" with respect to i in Eq. (5.7), one obtains (5.8)
or icnrr = BIR"
(5.9)
with B =2.3(/J, + /J",)/(/Jfl",), and R" =d" Idi'l~o. RI' is the polarization resistance, and B is called the Stern-Geary constant. It can be seen in Eq. (5.9) that the corrosion current for a corroding metal can be determined by measuring the polarization resistance, RI" at the corrosion potential when B is known. The polarization resistance is most commonly determined by the linear polarization technique. In linear polarization the slope of the polarization curve, the polarization resistance, is determined within an overpotential range close to the corrosion potential. The constant B is determined from the anodic and cathodic polarization curves. Figure 5.2 shows examples of linear polarization curves of zinc in three different solutions [1101. The polarization resistance can also be measured by impedance techniques. The determination of polarization resistance by impedance techniques is described in Chapter 2. Compared to the linear polarization technique, the advantages of the impedance techniques are that mechanistic information is obtained, that the measurement is independent of the solution conductivity, and that electrode capacitance is also obtained along
128
CHAPTER 5
4
Eu «
"NaCI
3
+ NaCI04
2
bNa2S04
a
2-
-3 -4
-5
-4
-3
-1
-2
a
3
2
5
4
E (mV SCE ) FIGURE 5.2. Linear polarization plots in the vicinity of the corrosion potential for zinc in molar Na2S04, NaCI04, and NaCI solutions at pH 5.8. Reprinted from Baugh [110], with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom.
with Rp [718]. The disadvantages are that impedance techniques are more complicated and that the Tafel slopes cannot be obtained from an impedance spectrum. Other electrochemical techniques are sometimes used for measuring polarization resistance. Boto and Williams [114] used a differential pulse method to determine the polarization resistance of zinc in aerated NaCl solutions. In this method a short potential pulse of several millivolts is applied, and the current response is measured. 11£ll1i is then taken as the polarization resistance Rp. It was reasoned that because the polarizing pulse is applied for only a very short time, less than 60 ms, the perturbation to the corrosion system is minimized.
10,000
... 80 2-
+
1,000
4
CI-
NE u
«
"-
100
.-
10
1~----~----~----~------~~--~~--~
1.04
1. 06
1.08
1.1
1. 12
Potential (Vsce )
~.1
1 .16
E corr
FIGURE 5.3. Anodic polarization of zinc in 1M NH4CI and 1M (NH4hS04 solutions. After Dattilo [33].
CORROSION POTENTIAL AND CORROSION CURRENT
129
In addition to the methods for determination of polarization resistance, corrosion current can also be determined by Tafel line extrapolation of the anodic or cathodic i- V curves to the corrosion potential as illustrated in Fig. 5.3 [33]. The underlying methodology for this method is straightforward. The anodic dissolution is described by ia = icorrexp(2.31J / Pm), and thus at the corrosion potential IJ = 0 and i" = iCOJr There is generally good agreement between the corrosion rates determined by the extrapolation method and by the polarization resistance method for zinc in various electrolytes [33, 110,534]. Corrosion current is meaningful only for evaluating certain corrosion situations where the corrosion is essentially uniform over the whole surface area. It may result in large errors if it is used for evaluating localized corrosion, such as pitting corrosion or intergranular corrosion. The corrosion loss caused by localized corrosion must be evaluated with other parameters. Localized corrosion is discussed in Chapter 7.
5.2.2. Conversion Factors In practice, corrosion rates are most often measured with methods other than electrochemical techniques and are expressed in terms of weight loss or thickness loss in units of milligrams per square decimeter per day or microns per year, etc. The conversions between corrosion current density and corrosion rate expressed in terms of weight loss and thickness loss rates are i=-·W
nF m
(5.10)
.
pnF m
(5.11 )
and l=--·T
in which Wand Tare the corrosion rates in weight loss rate (b/cm"·s) and thickness loss rate (cmls), respectively, p is the density of the metal, n is the charge per atom, m is the molar mass weight, and F is the Faraday constant. For zinc, p = 7.3 g/cm 3, n = 2, and m = 65.4 g/mol; thus, 1 j1A/cm"
=4.64 x 10-7 j1m/s = 14.5 j1m1yr = 2.92 mg/(dm2·day) (mdd)
or 1 j1m1yr = 0.07 j1A1cm 2 In the case in which the cathodic reaction is mainly due to hydrogen evolution, the extent of corrosion can also be measured by the amount of hydrogen collected during the corrosion process. For zinc at room temperature and at atmospheric pressure, the conversion between the corrosion current and the volume of hydrogen is [311]: I j1A = 0.4 ttl/hr
130
CHAPTER 5
In North America, weight and thickness for zinc coating are commonly expressed in English units. The conversions between the different weight and thickness units are: I ozlfe =305 g/m2 =30.5 mg/cm 2 = 1.70 mil
=43.2 ,urn
From a practical point of view, 1 ,uNcm2 is a rather large corrosion rate. The typical corrosion rate of zinc in most atmospheric environments is no more than a few microns per year, which is only a small fraction of 1 ,uNcm2 • 5.3. CORROSION POTENTIAL AND REACTION KINETICS As can be noted in Fig. 5.1, corrosion potential is the potential at which the total anodic dissolution current equals the total cathodic current. Any changes in conditions, such as surface preparation, solution composition and concentration, pH, temperature, time, convection, aeration, etc., that affect the anodic and/or the cathodic reactions will affect the value of the corrosion potential. Thus, the corrosion potential as well as the corrosion current of a metal can vary greatly, depending on the specific conditions. Table 5.1 lists the corrosion potentials and corrosion currents of zinc (or galvanized steel) in various environments. In solutions in which passivation does not occur or there is no oxidizing agent, the corrosion potential of zinc generally lies in the vicinity of its equilibrium value. This is primarily due to the high exchange current density for zinc dissolution and the relatively much smaller exchange current density for the cathodic reactions on zinc. The equilibrium electrode potential for zinc in acidic or near-neutral noncomplexing solutions is -1.122 VseE, assuming that the concentration of Zn 2+ ions in the solution is 10-4M, which is a likely concentration level in an aqueous solution containing a zinc electrode without the presence of Zn 2+ initially. From Table 5.1 it can be seen that most corrosion potentials measured in various solutions are close to this value. The small differences in the values among solutions of similar compositions and pH may be attributed to differences in aeration and measurement procedures. Aeration usually affects the cathodic reaction rate, whereas the measurement procedure may affect the surface condition of the electrode. For example, time of immersion in the electrolyte may affect the surface concentration of Zn 2+ ions and therefore the electrode potential of zinc. Large differences can occur between the corrosion potential and the equilibrium value when the surface of zinc is passivated or there are oxidizing agents in the solution. A change in the corrosion potential in either the anodic or the cathodic direction may correspond to a decrease or an increase in the corrosion current. The relative values of the corrosion potential and their relation to the corrosion currents under various conditions can be summarized using the schematic polarization curves in Fig. 5.4. In this figure, the corrosion potential of an active zinc electrode in a solution is E~o", and E ~oIT' E~o", E~o", and E~orr are the corrosion potentials under various conditions. The corrosion potential of an active zinc surface is E~orr with I" and Ie as the anodic and cathodic curve respectively. In the case in which the anodic dissolution is inhibited, for example, by surface adsorption of a chemical species, the anodic curve becomes 2". This will result in a more positive corrosion potential (from E~orr to E ~orr) if the cathodic reaction remains unchanged. In such a situation, the corrosion current is reduced upon a
CORROSION POTENTIAL AND CORROSION CURRENT
131
TABLE 5.1. Corrosion Potentials and Corrosion Currents of Zinc in Various Solutions Material" G90
Galvanized steel
Solution INHCl h INH 2S0 4 0.12MCr03 O.IMHCI
O.IMNaCI" O.IMNaCl h 0.05M H2SO 4" 0.5MNH4C1"
G90
IMKCI 1M NaCI" 1M Na2S0/ IN NaClf 100 ppm NO"] 0.IMNa2S0/ 0.IMNa2S0/ 2.7MNaCI IN NaCl h 0.IMNa2Cr04 O.IM Na2Mo04
0.IMNa2W04 6MNH 4C1 O.IM succinic acid 1M NaCI" 1M Na2S0/ O.IM Na2S04 + O.IM (NH 4)oSO/ I O.IM (NH4)2S04' O.IM (NH4)2S0/ 0.5NNaCl t 1M (NH 4lzS0i 1M NH4C1
0.IMNa3P04 O.IMNaCI 3.5% NaCl d 3.5% NaCl h 1.0NNaCI"
pH -0
I 2 2 3 3 3.8 3.8 4 4 4 4 4.8 5 5 5 5 5.3 5.6 5.8 5.8 5.8 5.9 5.9 6 6 6 6.2 6.3 6.4 6.4
1.0NNaCIf
3% Na2S0l 5% NaClf 0.IMNa2Cr04 99.2% Zn
(V SCE )
-1.00 -0.98 +0.05 -1.02 -1.057 -1.085 -1.07 -1.06 -1.03 -I.I -1.09 -1.04 -0.65 -1.102 -1.125 -1.12 -1.04 -0.93 -0.88 -1.07 -1.17 -1.095 -1.09 -1.13 -1.07 -1.056 -1.143 -1.03 -1.15 -1.133 -0.65 -1.098 -1.187 -1.14 -1.07
120 p~m HCO;- + 10 ppm S04-' NO;-
8.8
-1.07 -0.55 -0.89 -1.06 -1.11 -0.99 -0.995 -1.0 -1.12 -0.79
0.IMNa2Cr04 O.IM Na2Mo04
9 9
-0.57 -0.73
7.3
O.IMKNO/ O.IMKCI
0.IMNa2S0/ IMZnSO/
IMZnSO/ 0.IMNa2S0/ Zn (0001)
Ecorr
8
1M (NH 4 lzS0 4 h
icorr "
(mA/cm")
Reference
400' 3'
14 250 59 1.8' 703 1129 1129 0.91' 64 0.053' 63 75 0.014,·e 110 0.05 cc 110 110 196 1129 1129 0.009"" 534 45 199 199 199 0.028"<" 534 0.8'" 114 0.009"" 110 0.004"" 110 128 701 701 5 0.017"'" 33 0.011c.C 33 710 0.08S'" 114 701 701 1.8 X 10-4 <" 176 0.0014<" 176 0.05'" 41 41 710 38 38 38 0.002'" 112 0.001<" 112 128 0.009'" 446 410 199 199 (continued)
132
CHAPTER 5
TABLE 5.1. (Continued) Material a
Galvanized steel
Zn (0001) Galvanized steel
Solution
pH
0.IMNa2W04 0.IMNa2S0/ 0.IMNa2S0/ 100pprnNO:J NaClO/ O.OIMNaOH O.OOIMKOH
9 9 9 9.5 II 11 II
0.IMNa2S0/ 0.1 gil Ca(OH)2 0.IMNa2S0/ Sat. Ca(OH)z 0.05M NaOH + 0.5M H 20 2 0.IMNa2S04 O.3M KOH + Zn(OH)~-" 0.5MNaOH Sat. Ca(OH)z + 0.5N NaOH
II 11.1 12 12.6 12.7 13
0.5M NaOH + 0.08M H20 2 IN LiOH" IN NaOH b IN NaOH b IN KOH" 3NNaOHb
13.5 13.6 13.7
>14
30% KOH" 5NKOH 6NKOH b
8NNaOH 9NKOHb
>14 >14 >14 >14
Ecorr (V seE) -0.65 -1.077 -1.168 -0.78 -0.74 -0.11 -0.79 -1.141 -0.45 -1.171 -0.43 +0.01 -1.393 -1.43 -1.47
-0.6 -1.47 -1.46 -1.52 -1.47 -1.54 -1.49 -1.55 -0.73 -1.57 -1.61
icorr 2
(rnA/ern)
O.Ole
0.Dl5"
6 X 1O-4e
<0.04 0.06 0.02"
O.Ole
0.018 c 0.14' 0.012 e 0.2' 0.016 0.55 e
0.015 c
Reference 199 1129 1129 196 34 710 104 1129 197 1129 202 359 1129 12 446 202 359 311 311 790 311 790 25 311 10 790 311
" Unless otherwise specified. the material is of pure zinc. b Deaerated. C
d
Extrapolation from i- V curves. Aerated.
, From polarization resistance. Nondeaerated.
f
shift to a more positive potential. On the other hand, if the anodic dissolution kinetics remain unchanged but the cathodic reactions is represented by curve 2c instead of curve Ie (due to, for example, lowering of the pH), the potential also becomes more positive (from E~orr to E~orr)' However, in this case the corrosion current is increased upon a shift to a more positive potential. The anodic curve becomes 30 when the surface is passivated. If the cathodic reaction is unchanged, the corrosion potential of the electrode becomes E~orr' more positive than E~om and the corrosion current is generally much smaller than that at E~orr" Values of the corrosion potential that are much more positive than E~orr are usually associated with both the passivation of the zinc surface and the presence of oxidizing
133
CORROSION POTENTIAL AND CORROSION CURRENT
E
_ .-2, I,
---' FIGURE 5.4. Schematic polarization curves illustrating Eeorr and icorr under various conditions (see text).
rt-r-T1'--._
E' "'"
E'ool'T J30"",
I, i,
o
i.
agents (e.g., chromate ions or hydrogen peroxide) in the solution. The cathodic and anodic polarization curves in such a solution are illustrated by 3c and 3", the coupling of which yields the corrosion potential E~()ff" The corrosion rate in this situation is usually very low. However, if the surface is not passivated by the presence of an oxidizing agent, the corrosion rate can be very high. This can be appreciated by coupling curves la and3cIn a similar manner, a decrease in potential can be caused by either faster anodic dissolution kinetics or a slower cathodic reaction. For example, a decrease in corrosion potential due to deaeration of the solution is usually observed for a zinc electrode in a neutral solution. Deaeration removes the dissolved oxygen and thus reduces the cathodic reaction rate.
5.4. Ecorr AND
icorr
UNDER VARIOUS CONDITIONS
5.4.1. Effect ()f Zinc Ions
The electrode potential of zinc is a function of the zinc concentration in the solution. Thennodynamically, as illustrated by the pH-potential diagram in Chapter 2 (Fig. 2.2), the effect of zinc concentration on the zinc potential depends strongly on pH. It has been found, as shown in Fig. 5.5, that the corrosion potential of zinc in sulfate solutions is almost independent of the concentration of Zn 2+ at low pH values, but at pH > 5 the corrosion potential starts to show a linear dependence on the logarithm of Zn 2+ concentration. The slope of this line is 2.3 x RTl2F, corresponding to the dependence of the equilibrium potential of zinc on the concentration of Zn 2+, according to the Nernst equation [4451. The insensitivity of the corrosion potential to changes in the bulk Zn 2+ concentration at low pH values is probably due to a high surface concentration of Zn 2+ resulting from the dissolution of zinc, which occurs rapidly at low pH.
134
CHAPTER 5
-1.-----------------------------------------,
- 1.05 • pH 6108
+
pH 5.4
• pH 4.45
x
pH 3 .3
• pH 2 .7 'l
- 1.15
pH 2 .2
~------~--------~------~--------~ -0.5 -3 .5 -2.5 -1.5 -4.5
Log [ZnSO.l
FIGURE 5.5. Zinc electrode potential in O.IM sulfate solutions as a function of zinc ion concentration and pH of the solution at a disk velocity of 9 rev/s and a temperature of 2Ye. After Gmytryk and Sedzimir [445].
In alkaline solutions the zinc potential is found to be a logarithmically linear function of Zn(OH)~- concentration, as shown in Fig. 5.6 [131]. The deviation caused by the hydrogen reaction on the zinc electrode is less than 1 mV owing to the fast rate of the zinc reaction and the slow rate of the hydrogen reaction. The slopes of the lines in Fig. 5.6 are between 26.9 and 28.6 mY/decade for KOH concentrations between 3.1 and 1O.8M, which is very close to the slope predicted from the equilibrium equation (5.12)
E= 0.441 - 0.118pH + 0_02951og aZn(OH)~
Bockris et al. [I2] found that the zinc potentials experimentally measured in 0.3-3M KOH solutions containing various amounts of zincate are 20 to 30 mV more positive than those calculated according to the equilibrium described by Eq. (5.12)_ -1 _34
r-----------------------------------------,
-1.36
oOl -1.38
:r:
~
• 3.1 M OH K
-1.4
'" 4 .8 M KOH
~
w -1.42
7.7 M KOH y
10.8 M KOH
-1.44
-1 . 46~--------~---------L----------~------~
0 .001
0.01
0.1
10
K.Zn(OH), Concentration (M)
FIGURE 5.6. Zinc electrode rest potentials as a function ofZn 2+ concentration in KOH solutions at 25°e. After Isaacson et al. [131].
CORROSION POTENTIAL AND CORROSION CURRENT
135
-0.75,--- -- - - - - - - - - -- - -- -- - ,
w
u
~
-0.85
~
c:QI
(5 c.
-0.95
Sr -
QI
... CI-
~
+ NOj
'0
U
QI
W
-1. 05
" 1-
X 50 2 4 • CI0 .j
10"8
10
·5
.,
10
FIGURE 5.7. Potential-log C curves for zinc in solutions of corrosive anions. Reprinted from Gouda et al. l 1139], with kind permission from Elsevier Science Ltd. The Boulevard, Langford Lane. Kidlington OX5 1GB, United Kingdom.
The addition of zinc ions to a solution generally reduces the corrosion current of the zinc electrode. For example, it has been found that the corrosion current of zinc in 1,8M KOH solution is 45 I1A/cm 2 , but the addition of 0.03-0, 1M zincate ions to this solution decreases the corrosion rate to below ll1A!cm 2 [12],
5.4.2. Effect of Anions and Cations According to Gouda et ai, [1139], anions can be classified into two main groups based on the mode of variation of the steady-state corrosion potential of zinc with anion concentration, In the first group, to which SO~-, CI-, 1-, Br-, CIO:!, and NO) belong, an increase in anion concentration is accompanied by a decrease in the potential of the zinc electrode as shown in Fig. 5.7. The anion concentration and the corrosion potential follow the relationship Eeorr = a - b log C
(5.13)
with C the concentration, and a and b constants. For SO~-, Cl-, Be 1-, and CIO:!, b is 33 mY/decade; for NO), it is about 15 mY/decade. The lower slope for NO) is probably related to the slight oxidizing character of the nitrate ion. The reduction in the corrosion potential with anion concentration was suggested by Gouda et al. to be the result of an altered balance between anodic and cathodic areas set up by the competing influence of oxide film formation and destruction by the anions. According to Awad and Kamel [784], the adsorption of the anions on the bare anodic areas of the metal surface accelerates the ionization of the metal, perhaps through reduction of the activation energy of the process. The potential decreases with an increase in the anion concentration in order to increase the cathodic reaction rate. In the second group of anions, to which CrO~-, H2PO;;, WO~-, HPO~-, and NO; belong, the corrosion potential depends on the type of anion but not on concentration at low concentrations; it increases with concentration at high concentrations as shown in
136
CHAPTER 5 -0.45 , - - - --
- - - - - -- - --
- - -- - - - - ,
• H2 PO'; Q; u
A
G -0.65 ~
• HPOt
c:
X N0 2·
'"
2
(5
a.
'"
'0
~
<:5
CrOt
+ wot
If)
-0.85
'" UJ
• .,
_' . 05L----L----~----L----L----~----~--~--~ ~3
10
10
-2
Log C (M)
10
10°
10'
FIGURE 5.8. Potential-log C curves for zinc in solutions of corrosion-inhibiting anions. Reprinted from Gouda
et al. [1139], with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom.
Fig. 5.8 [1139]. The more positive corrosion potentials in these solutions are due to the surface passivation taking place at certain concentrations, depending on the type of anion. The corrosion potential of a passivated zinc electrode is also sensitive to the presence of other ionic species as shown in Fig. 5.9 [59]. The corrosion potential in phosphate solutions varies with the concentrations of the primary, secondary, and tertiary phosphate ions. Figure 5.10 shows the corrosion potential as a function of the concentrations of NaH 2P0 4, Na2HP04, and Na3P04' Generally, a shift of the potential to more positive values takes place at a certain phosphate concentration which depends on the solution pH. According to Awad and Kamel [784], this potential 0.2.---------------------------~
o
--0J,!', ,;:06L.------------l NaN03
______
~~
_ -0.2
0 .3
-
Q)
u
If)
G -0.4 n;
'E
'"
-0.6
(5
c.. -0.8 ~--_----------70i-,0~i-6------------------.:
-1.2
j NaCI JNa"SO,
0,3
-1 ----
0,03 0.3
L...._ _ _ _...L..._ _ _--'-_ _ _ _ _ _ _ _-'--_ _ _ _ _- ' -_ _- '
o
10
20
30
40
Time, hours
FIGURE 5.9. The effects of several anions on the potential-time curves for zinc immersed in 0.12-molll Cr03 solutions at 20°C. The concentrations of the anions, in moles per liter, are labeled on the plot. Reprinted from Williams [59], with kind permission from Elsevier Science-NL, Sara Burgerhartstraat 25, 1055 KV Amsterdam, The Netherlands.
137
CORROSION POTENTIAL AND CORROSION CURRENT
-O.5r-----------------------------------------, • Na,HPO.
-0.7
+ NaH,PO.· Na,HPO. mix • Na,PO.
~
~ -O.9
LoU
- 1. 1
_1.3L-------~-------L------~--------L-------~
-4
-3
-2
-1
o
Log C (M)
FIGURE 5. I O. Effect of concentration of sodium primary phosphate (NaH 2P0 4). primary-secondary phosphate mixture. secondary phosphate (Na2HP04)' and ternary phosphate (Na3P04) on the potential of the zinc electrode. Reprinted from Awad and Kamel [7841. with kind permission from Elsevier Scicnce-NL. Sara Burgerhartstraat 25. 1055 KY Amsterdam. The Netherlands.
shift is attributed to the fonnation of a protective serniglassy interface resulting from the adsorption of phosphate ions in a complex fonn on the surface of a zinc phosphate layer. After reaching a maximum value, the potential decreases upon further increase of concentration owing to adsorption of phosphate ions on the bare cathodic areas of the metal, which decelerates the reduction of oxygen. In a Na 3PO c NaOH mixture, both OHand PO~- ions take part in forming the serniglassy interface. At low concentrations, adsorption of H2PO:; results in a reduction of the anodic dissolution whereas at high concentrations H4 PO; ions exist and catalyze hydrogen reduction through preferential discharge of hydrogen bonded to the ions. In Ca(OHh solutions of concentrations in the range 0.1-3 gil (pH 11.1-12.55), zinc shows a corrosion potential of about -0.5 VseE after 2-4 days of immersion [175, 197, 202J. The ennoblement of the zinc surface is due to the formation of a compact calcium hydroxyzincatc. In HSO:; solutions, when the concentration is above 5 x 1O- 3M, the potential is independent of concentration. This is associated with competitive adsorption between S02 and HSO:; [943]. The presence of NaCI0 2 in aqueous solution leads to more positive Ecorr and larger icorr values due to the effect of ClO:; reduction, similar to the effect of NO:; reduction [34]. The corrosion current in aqueous solutions is significantly reduced by adding rare-earth salts such as CeCl 3 [605]. Many organic species affect the corrosion potential or corrosIOn current of zinc in aqueous solutions, including acrylic type anions [39], dimethyl sulfoxide, N-dimethylformamide, and acetonitrile [463], pyrazole [15], oxalate ions [40], phosphines [64], benzene thiols [164], phenothiazine [63], zinc gluconate [100], and n-decylamine [207]. The corrosion potential and current of zinc and zinc alloys have also been detennined in real environments such as concrete [468], soils [357], and natural waters [709, 565]. 5.4.3. Effect afpH
Figure 5.11 shows the corrosion potentials experimentally detennined in various noncomplexing and nonoxidizing solutions for zinc or zinc coatings as a function of pH
138
CHAPTER 5
0r-----------------------------.
1,
a-0 .5
UJ"
"iii
"E., 15
Q..
c:
·1
i ...
.9
e'"
(;
o
· 1.5
o
2
4
.~ ',...
FIGURE 5.11 . Corrosion potentials in various solutions (data from Table 5.1). The solid line indicates the reversible potential calculated from the Nernst equ ation, assuming IO-4M Zn 2+ in the solution.
~~
6
8
10
12
14
pH
(data from Table 5.1). The solid line indicates the reversible potential calculated from the Nemst equation assuming 1O-4M Zn 2+ in the solution. It can be seen that the corrosion potentials in the pH range 4-8 are close to the calculated values. However, the corrosion potentials in acidic and alkaline solutions are somewhat higher than the reversible potential values. This may indicate that the concentrations of Zn 2+ in these solutions are higher than lO-4M, at least near the surface, where the zinc ions from the dissolution may accumulate. The larger discrepancies between the calculated values and the measured values in the pH range 8-12 are due to the formation of solid oxide or hydroxide films, resulting in various degrees of passivation. Gmytryk and Sedzimir [445] measured the corrosion potential of zinc in deaerated O.IM Na2S04 solutions in the pH range between 0 and 9. Figure 5.12 shows that the corrosion potential changes the most between pH 4 and 6. The more positive values in -1
-1.05
QI
u
~
·1 .1
w
-1.15
-1.2
0
2
6
4
8
pH
FIGURE 5.12. Corrosion potential ofZn as a function of pH in O.IM Na2S04' Reprinted from Gmytryk and Sedzimir [445], with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom.
139
CORROSION POTENTIAL AND CORROSION CURRENT ·0.8 • 02 M PO'"
.
'4
A O.2M PO ~'w ilh roln .
.,.. NaOH
+ 0.2M
PO~ .
·1 " NaOH
..
X NaOH wilh rotation
ell
~
~
-1.2 L---:"'------~-
(; <)
W
· 1.4
_1.6 L-__ __ ____ __ ____L __ __ L_ _ _ _L __ __ L_ _ 13 14 11 12 10 6 7 8 9 ~
~
~
~
~
pH
FIGURE 5.13. Dependence of Zn electrode corrosion potential, Eeorp on pH for different types of solutions: 0.2M [pol-I; 0.2M [pol-] with electrode rotation; NaOH + O.2M [pol-I; NaOH; and NaOH with electrode rotation. Reprinted from De Pauli et al. [481], with kind permission from Elsevier Science-NL, Sara Burgerharstraat 25, 1055 KV Amsterdam. The Netherlands.
the lower pH range are attributed to the anodic polarization required for the relatively larger dissolution rate of zinc. In the pH range between 4 and 6, the cathodic part of the corrosion process is controlled by diffusion of hydrogen to the surface. Figure 5.13 shows the dependence of Eeorr on pH in solutions containing various amounts of phosphate [481]. Within the pH range 7-10 an almost constant potential is found, whereas within the pH range 11.5-14 a slope of 60 mV per unit of pH is obtained. The constant corrosion potential value from pH 7 to 10 is attributed to the reaction Zn + HPO~- = HZnP04 + 2e-. Electrode stirring is found to have an effect on the corrosion potential between pH 10.5 and 11.5, where the corrosion potential values are more positive and scattered, but has no effect at other pH values. -3 r---~----------------------------------------,
NE -4 u
.a
~
0
A
._u
~.5
b C
" d Xe -6
a
8
4
12
pH
FIGURE 5.14. Logarithm of current density for zinc in deaerated O.IM NaCI solutions as a function of pH of the solution. (a) and (d) Reaction control; (b) intermediate control; (c) two simultaneous reactions; (e) diffusion control regime. Reprinted from Zembura and Burzynska [116], with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom.
140
CHAPTER 5 3
+
+
'f2
+ +
u
<:
.s
~
(;
u1
+ 0
4
5
6
7
8
9
10
11
pH
FIGURE 5.15. Corrosion current density of zinc as a function of pH in unbuffered O.IM Na2S04 solutions and oxygen-saturated solutions: acetate buffer, pH 4.7; succinate buffer, pH 5.6; phosphate buffer + 0.1 mol KClIl. pH 7.0; sodium acetate (0.1 molll), pH 8.0; sodium acetate (0.1 molll) + EDTA (0.1 mol/I) + NaOH to pH 9. Reprinted from Boto and Williams [128], with kind permission from Elsevier Science-NL, Sara Burgerharstraat 25, J055 KV Amsterdam, The Netherlands.
In contrast to the corrosion potential, which varies only slightly with pH in acidic solutions, the corrosion current is a strong function of pH. Figure 5.14 shows the corrosion current of zinc in deaerated O.lM NaCI solution as a function of pH [116]. The corrosion current is relatively high in acidic and alkaline solutions and is lowest at pH 9. This U-shaped dependence of the corrosion current on pH agrees well with weight loss data obtained from immersion tests (see Chapter 9, Fig. 9.13). Boto and Williams [128] found that in oxygenated sulfate solutions in the pH range 4-6 the corrosion of zinc is controlled by oxygen diffusion as shown in Fig. 5.l5. At higher pH values, zinc hydroxide builds up on the surface, and the reaction is no longer oxygen-diffusion-controlled. As a result, the corrosion current decreases to much smaller values at pH > 6.
5.4.4. Effect a/Temperature The equilibrium potential of zinc changes by only a few millivolts for a change in temperature of several tens of degrees, as can be seen in Fig. 2.1 in Chapter 2. Figure 5.16 also shows that the zinc potential changes little as a function of temperature in alkaline solutions containing various amounts of zinc ions [131]. Significant changes in the corrosion potential as a result of changing temperature can, however, arise from changing the surface state of the electrode from active to passive. It is a well-known phenomenon that the polarity reversal of a zinc/steel galvanic couple in hot water or aqueous solutions is primarily due to the change of the zinc surface from an active state to a passive state. In distilled water, potential can significantly change with temperature variations in the presence of small amounts of ionic species in the solutions. Figure 5.17 shows that the corrosion potential of zinc increases with increasing temperature in waters containing trace amounts of HCO), SO~-, and NO) [410]. Generally, NO), HCO), and cOj- ions cause an increase in corrosion potential with increasing temperature, while SO~-, Ct, SiO~-, and Ca2+ cause a decrease [196,410]. More infor-
CORROSION POTENTIAL AND CORROSION CURRENT
141
-, .34 r - - - - - - - -- -- - -- - - - - - - - - - - , + 0.0132M • 0.0476M "0.0966M . O.I 54M -1.36
•
•
•
0.35M
•
•
Q)
u
'" > ~ -1.38 w
e
t
-1 .4
'0
40
30
20
Temperature
60
50
(Oel
FIGURE 5.16. Effect of temperature on Zn potential in 4.SM KOH containing Zn 2+ at different concentrations. After Isaacson et al. [13 I].
mation on the variation of corrosion potential in hot water with respect to changing temperature are presented in Chapter 7 (Section 7.2.4). There are very limited data on the effect of temperature on the corrosion current of zinc. Yamashita [112] found that the corrosion current of zinc in 1M ZnS0 4 solution may increase with temperature in the presence of oxygen but decreases with temperature in the absence of oxygen. The difference was explained as probably due to the formation of a surface corrosion product.
5.4.5. Effect of Aeration and Convection Aeration or deaeration changes the concentration of dissolved oxygen in an electrolyte. Figure 5.18 shows that aeration has a significant effect on the corrosion potential of
-, , - - - - - - - - - - - - - - - - - - - - - - , -0.9
30'C
....,-0.8 u
'"
2:,.-0.7 40' C
n3
'g-0.6 Q)
<5 Q.-
0.5
- 0.4
70' C
-0.3~---L---~--~L---~---~--~
o
2
3
4
5
6
Time, hou rs
FIGURE 5.17. Effect of temperature on zinc potential. Solution composition: 115-140 ppm HCOl 10 ppm SO~-, and 10 ppm NO;. From Hoxeng and Prutton [410]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
142
CHAPTER 5
-1,----------------------------------,
c: -1 .2 Ul u
~
-a.rated
C Q)
-0-
o
OIitBe-rated
c.
c:: -1.4
o
';;;
e
o
(j
-'.6
o
8
4
FIGURE 5.18. Effect of aeration on the corrosion potential of zinc in O.IM Na2S04 solutions of different pH values. After Zhang and Hwang r1129J.
12
pH
zinc, particularly in near-neutral or slightly alkaline solutions [1129]. The more negative potential values in the oxygen-free solutions are due to the reduced total cathodic reaction rate, as curve 2e changes to curve Ie in Fig. 5.4. Dissolved oxygen has less of an effect on the corrosion potential in acidic or alkaline solutions, where hydrogen evolution is the predominant cathodic reaction. In addition to the aeration conditions, the form of the electrolyte affects the rate of oxygen reduction. Figure 5.19 shows that the corrosion potential in aerated 0.5M NaCl solution decreases with decreasing solution thickness from 200 mm to 0.1 mm whereas in deaerated solution the corrosion potential increases with decreasing thickness [608]. The decrease in the aerated solution is attributed to the depletion of dissolved oxygen, since the amount of dissolved oxygen decreases with decreaSing volume of electrolyte
-1.06 , - - - - - - - - - - - - - - - - - - - -- - - -- - - - - - -- - ----,
Q) ()
G'"
-1.1
• Aerated
iij
~ Q)
A
De.aerated
(5
Cl.
C
.~ -1.14
e o
o
-,
-1.'8 '--_ _ _-'-_ _ _--'-_ _ _---'-_ __ ----'_ _ _---.J -2 o 2 3 Layer thickness, mm (Log)
FIGURE 5.19. Corrosion potential Ecorr of zinc in aerated and de aerated O.SM NaCI solutions as a function of electrolyte layer thickness. After Keddam et al. [608].
143
CORROSION POTENTIAL AND CORROSION CURRENT
TABLE 5.2. Effects of Grit Size of Grinding Paper on Corrosion and Surface Parameters of Zinc in O.IM Na2S04 at 25°C" Grit size
Eeorr
(V)
-1.053 -1.051 -1.062 -1.077
120 240 320 400
icarr
IJiA/cm 2 )
1187 819.2 483.1 184.2
icath
!JINcm 2 )
623.9 304.4 206.7 178.6
Rp (Q·cm 2 )
fI" (V/decade) fie (V/decade)
0.161 0.158 0.153 0.79
37.51 63.24 97.54 143.2
0.262 0.487 0.373 0.263
._--
"Ref. 383.
on the surface. On the other hand, the increase in the deaerated solution is attributed (0 the slower anodic reaction kinetics due to the buildup of zinc species in the solution. Convection, which enhances the transport of the species in an electrolyte, has little effect on the corrosion current of zinc in acidic solutions because the corrosion in acidic solutions is, in general, controlled by an activation process. Convection usually affects the corrosion current in nonacidic solutions, in which the corrosion process may be associated with different diffusion-controlled reactions depending on aeration and pH [128, 445). Bo(O and Williams [128] found that the corrosion current in aerated sulfate solution at pH 5.8 increases linearly with the square root of the rotation rate of the electrode [128]. Gmytryk and Sedzimir [445] reported that the corrosion current III deaerated O.IM NaCl solution from pH 4 to 6 depends on the convective condition. 5.4.6. Effect of Surface Condition Surface conditions are very important in electrochemical corrosion measurements. The condition of the surface is perhaps one of the most nonreproducible factors and may account for the remarkable differences in the corrosion data reported in different studies. Depending on the surface finishing condition, the degree of cleanness and roughness can vary greatly. The effective surface area can be many times greater than the apparent surface area, depending on the polishing or grinding material. For example, Table 5.2 shows that the grit size of the grinding paper has a significant effect on the corrosion potential and corrosion current density [383]. For a change in grit size from 400 to 120, the apparent current density increases by as much as a factor of 6. For a single-crystalline zinc sample, the corrosion current may be a function of the crystal orientation. Table 5.3 shows the Eeorr and polarization characteristics of single TABLE 5.3. Electrochemical Parameters for Zinc (0001), (1010), and (1120) Surfaces in Deaerated 1M (NH4)2S04 Solutions"
____Eeorr (V seE) ___b" (mV) _____be (mV) ___ Rp(Q·cm-)
Exposed surface
(0001) ( 10TO) ( 1120)
7
=~~~
-1.116 -1.132 -1.132
~
31 31 32
~~
124 132 127
_L~_~_
1052 1447 1499
"Reprinted from Abayarathna et al. [446], with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OXS 1GB. United Kingdom.
144
CHAPTER 5
10,--------------------------------------. • 6-hour immersion
E0 «
A 3.hour immersion
... ...
3~ ~
0 0
0 . 1~----~----~-----L----~
048
12
____~~____~
16
20
24
Drying time (hours)
FIGURE 5.20. Effect of drying time in air after immersion in 5% NaCI solution on corrosion current density of electrodeposited Zn-7% Fe alloy. Reprinted from Sagiyama et ai. [611], with kind permission from The Iron and Steel Institute of Japan, Chiyoda·ku, Tokyo, Japan.
crystals in 1M (NH4)2S04 solution [446]. It appears that the polarization resistance is slightly lower for the (0001) orientation than the others. The opposite relationship is observed in alkaline solutions. The individual grains of a polycrystalline surface may show different potentials. Tsuru et at. [705], using a scanning microelectrode, measured differences in the corrosion potential (up to 15 m V) among individual grains of a polycrystalline zinc electrode. Other surface factors may affect the values of EcoIT and icorr' Chiu and Selman [706] found that the corrosion potential and current varies across a curved zinc surface in a flowing electrolyte owing to the transport and ohmic effects of the electrolyte. The corrosion potential of a chromate-treated zinc surface is usually similar to that without the treatment, but the corrosion current is generally much smaller with the chromatetreated surface because of the passive film formed on the zinc surface [75, 701]. Sagiyama et at. [611] found that drying in air after immersion in the solution drastically reduces the corrosion current of a zinc alloy as shown in Fig. 5.20 and attributed this decrease to the formation of a more compact corrosion product film as a result of drying. The corrosion current of electroplated zinc coatings in neutral chloride and sulfate solutions was found by Dattilo [33] to be similar to that of pure Zn. 5.5. ZINC ALLOYS The corrosion potential of a zinc alloy depends on both the equilibrium potential of zinc and that of the alloying element. Since zinc has a low position in the emf series, alloying with most elements will result in a more positive corrosion potentiaL In general, addition of small amounts of other elements, i.e., a few percent, does not significantly alter the corrosion potential from that of pure zinc. The potential only becomes significantly different from that of pure zinc when a certain level of the alloying element is present. Compared to the corrosion potential, the corrosion current of zinc alloys varies with the alloying element in a more complicated manner, as it is affected not only by the
CORROSION POTENTIAL AND CORROSION CURRENT
145
nature and quantity of the alloying element but also by the fonnation of intermetallic phases and the microstructure of the alloy. Zinc has been alloyed with many elements, such as aluminum, iron, nickel, titanium, and copper. Table SA lists the corrosion potentials and currents of some zinc alloys in various solutions. Figure 5.21, reported by Selvam and Guruviah [330], shows the corrosion potentials and currents of Zn-Mn alloys in 3% NaCI solution. The corrosion potentials of the alloys with Mn contents below 50% are close to that of pure zinc, while the corrosion current is the lowest for alloys with Mn contents between 10 and 30%. For zinc-copper alloys, Fig. 5.22 shows that the corrosion potential is close to that of pure zinc when the Cu concentration is below 30% and is close to that of pure copper when it is above 40% [361.
TABLE 5.4. Corrosion Potentials and Corrosion Currents of Some Alloying Elements and Zinc Alloys in Various Solutions Alloy
Solution
Cu 10%Cu AI
O.IM Na2S04 5% NaCI" INHCI" IN NaCI INHCI" SS%AI INH 2S04 INHCI" 5%AI S5%AI LON NaCl a LON NaCi 94% AI 3% NaCI sr/e AI 5% NaCI 55%AI 5% NaCI INNaCI S%AI INNaCI INHCI" Fe Steel INNaCI 5% NaCI" 10% Fe 10% Mg S% NaCl" 5% NaCl" 10%Ti 5% NaCI" 10%Cr 10% Ni 5% NaCl" 12%Ni 5% NaCl a 5% NaCl d 1O- 3M Cl0.19%Ni Mn 3% NaCI 14.3% Mn 3% NaCI INNaOH 2%Pb O.IMHCI 2%Cd O.IMHCI 1% Ti + 0.5% Cu O.OIN NaOH" ----.--~-
-------~--
/1
Dcaerateo. Extrapolation from i- V curves.
C
From polarization resistance.
d
Aerated.
II
pH
0 4
0
6.5
4 4 ()
4
]
12
Ecorr (V SCE)
-0.05 -0.843 -0.76 -0.64 -1.01 -0.96 -1.05 -1.05 -1.17 -1.00 -1.09 -1.01 -0.99 --1.04 -0.48 -0.69 -0.915 -1.016 -1.003 -0.942 -0.9 -1.11 -0.97 -1.02 -1.267 -1.043 -0.5 -1.01 -1.02 -1.26 ..
-----~.--
icorr
(mAlcm 2 )
Reference ]g
Sh 2h 0.3 h
O.Sh 2. I x 10-4 , 2.3 x 10- 5,
0.4"
0.026' 0.59" 0.36"
1241 14 14 14 250 14 176 176 42 ]47 34R 14 14 14 14 1241 1241 1241 1241 1241 287 287 32 330 330 330 703 703 37 _0. _ _ _ -
146
CHAPTER 5 -1r-----------------------------~
170
wu
~
Corrosion potential
>'" :::- -1 .1
E
o ~
.~
C
120
Cll
<5
..:; Ql
iii c o
a:
Cl.
c o
'0;
.~ -1 .
70
(; ()
e
(;
()
-1.3 ' - - - - ' - -- - - ' - -- -...L..- ---'---...J20
o
W
~
00
~
100
%Mn
FIGURE 5.21. Corrosion potential and current of Zn-Mn alloys in 3% neutral NaCI as a function of manganese content. After Selvam and Guruviah [330].
Figure 5.23 shows the changes in corrosion potential after 35 days' immersion in 5% NaCl solution for Zn-Ni and Zn-Co alloy coatings [44]. The coatings containing less than 10% Ni and 7.5% Co have potential values close to that of zinc and maintain them after 35 days of immersion. The corrosion current as a function of Co and Ni concentration is shown in Fig. 5.24. According to Short et al. [44], the ennoblement of the potential with increasing Ni or Co content is associated with an increase in the anodic Tafel slope. They also found that for the Zn-Ni and Zn-Co alloy coatings more positive corrosion potentials are associated with lower corrosion currents. Similarly, Hosny et al. [606] found that the addition of 8 gil cobalt in the zinc plating bath resulted in 25% reduction of corrosion current compared to pure zinc coating. Darwish [32] reported that, compared to the corrosion potential of pure zinc, the corrosion potentials of Zn-Ni alloys vary more with pH in acidic and near-neutral solutions. -0.2
r---- -- - - - -- - - - - - -- ----,
Q; -0.4 to r/)
G
~ -0 .6 c:
oa. Q)
§
-0,8
'iii
e(;
()
-'~---~~ -, .2 :------::-':-______---'-________-'--_____--'-__-.J o 20 40 60 80 Concentration of Cu ('Yo)
FIGURE 5.22. Potential of Zn-Cu alloys in air-saturated 3.5% sodium chloride solution at 25°C as a function of copper content. From Budinski and Wilde [36]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
147
CORROSION POTENTIAL AND CORROSION CURRENT
-O.6 r----------------------------------------, ..•.. initial
-
35 days immersion
Q)
u
2:.'" -0 .8 ·z'" c
Q)
oa. c
o
'iii ~
-1
o
..... ... ... .. . ..... . .. . .........
u
_ 1.2 L-----~-------L------~------~----~----~
o
4
8
12
16
20
Concentration of Ni or Co (%) FIGURE 5.23. Variation in corrosion potential with time for Zn-Ni and Zn-Co coatings in 5% NaCI solution. Data are taken from Ref. 44.
Figure 5.25 shows the corrosion potential as a function of Fe content in electroplated coatings in 5% NaCI solution at three different pH values [612]. There are roughly three distinct regions of behavior with respect to Fe content. The corrosion potential of Zn - Fe alloys with Fe contents less than a few percent is similar to that of pure zinc. Alloys with 10-50% Fe show a relatively constant corrosion potential 200 m V higher than that of pure zinc. For alloys with Fe content higher than 70%. the corrosion potential is close to that of pure iron.
100
"*'Ni
Ec.>
+Co
..::
3. C ~ 10 :; c.>
c 0
'in ~
<:;
U
1
0
4
8
12
16
20
24
Concentration of N i or Co (%)
FIGURE 5.24. Effect of Ni and Co concentration on the average corrosion rate , between 20 and 40 days. of zinc alloy coatings in 5% NaCI solution. After Short el af. [44].
148
CHAPTER 5 -0.4
pH 9,5 ... pH 11.7
-0 ,6
+ pH
13.6
-;;; -0,8 u
'"
~
~
-1
C Q)
~ -1.2 -1.4
40
20
100
80
60
Fe content in coating (wt 'Yo)
FIGURE 5.25. Relationships between corrosion potential of Zn-Fe alloy coatings measured in alkaline 5% NaCl solutions and Fe content in the coating. Reprinted from Sagiyama et al. [6121, with kind permission from The Iron and Steel Institute of Japan, Chiyoda-ku, Tokyo, Japan,
According to Sagiyamaetal. [611], Zn-Fealloys containing 4-27% Fe exhibit much lower corrosion currents than of pure zinc and alloys containing more than 30% Fe when tested in neutral 5% NaCI solution. The dependence of the corrosion current on iron content is a function of the pH of the testing solution. At pH 9.5 and 11.7, the dependence of corrosion current on Fe content is similar to that observed in the neutral solutions as shown in Fig 5.26 [612]. The same figure shows that at pH 13.6 the corrosion current increases with Fe content up to about 50% Fe and then decreases with further increase in the Fe content. At pH 12, the corrosion current is found to decrease with increasing Fe content over the whole composition range. In another study, Chang and Wei [365] found that electrodeposited zinc-iron alloys with 20-40% Fe have the lowest corrosion current in de aerated O.IM NaCI solution.
100
E'-' «
.3- 10
• pH 9.5
0
... pH 11 .7
+ pH 13.6 1
o
20
40
60
80
100
Fe content in coating (wt 'Yo)
FIGURE 5.26, Relationships between corrosion current density obtained from polarization curves for Zn-Fe alloy coatings measured in alkaline 5% NaCI solution and Fe content in the coating, Reprinted from Sagiyama et ai, [612], with kind permission from The Iron and Steel Institute of Japan, Chiyoda-ku, Tokyo, Japan.
CORROSION POTENTIAL AND CORROSION CURRENT
149
The corrosion of most zinc alloys is associated with a dezincification process, in which the zinc is preferentially dissolved. As a result of dezincification, the surface concentrations of the alloying elements increase. Thus, the corrosion potential and current tend to vary much more with time than in the case of pure zinc. 5.6. EFFECT OF TIME The time at which a measurement is taken is one of the most important factors in determining the corrosion potential and current of a metal in a corrosion system. It is perhaps also a major factor in the differences among the results reported in various studies of similar corrosion systems. Time invariably brings two basic changes to a corrosion system: (I) a change of the physical structure and chemical composition of the corroding metal surface and (2) a change in the composition of the solution, particularly in the vicinity of the surface. Specific changes that may occur include changes in surface area and roughness, adsorption of species, formation of passive films, saturation of dissolution products, precipitation of a solid layer loosely attached to the surface, and exhaustion of reactants. Mechanistically, these changes may lead to alterations in the equilibrium potentials, the type of reactions involved, the rate-controlling process, etc. As a result, the corrosion potential and current may vary drastically depending on the nature and extent of these changes. The potential and current may not reach a constant value if the surface and solution change continuously. For example. in the case of zinc alloy coatings, such as batch hot-dipped zinc coatings, the composition varies from the surface ro the coating/steel interface. Thus, the corrosion potential will change with time as the coating gradually dissolves. Depending on whether the surface is active or passive, the corrosion potential of a zinc electrode may reach different values at a steady state. As shown in Fig. 5.1 I, the corrosion potential of an active zinc surface in neutral nonoxidizing and noncomplexing salt solutions is about -1.1 Vso' near the equilibrium value. In solutions in which passive corrosion products can form, the corrosion potential can be much more positive than the equilibrium value.
TABLE 5.5. Time to Reach a Steady Zinc Corrosion Potential (V seE) under Various Conditions -.---.-~.----
Solution I gil Ca(OH)2' pH 12.3 Distilled water, 40"C 100 ppm cr. 40°C. pH 4 10 ppm NO:;. 80°C, pH 8 30 ppm NaHC0 3 , 4()OC IMZnSO.j,2YC IMZnS04,50°C 0.1 wI. % NaCI, 25°C O.5N Na2S04, pH 4 O.IM Na2S04, pH 6 O.IMNa2S04' pH 13 -------------- - - - -
Time
Initial Ecorr
.------.
3 days 240 min 150min 100 min 50 min 120 min 200 min 20 min 10 min 10 min 2 min
-1.39 -0.69 -0.90 -0.81 -1.0 -1.01 -1.01 -0.99 -1.12 -1.09 -1.41
Steady Eeorr -0.45 -0.93 -0.97 -0.48 -0.84 -0.99 -1.00 -1.05 -1.11 -1.06 -1.42
Reference
------
197 196 196 196 196 112 112 365 14 1129 1129
150
CHAPTER 5 TABLE 5.6. Effect of Time on Corrosion Currents of Zinc Initial Solution
Tap water, pH 8 30 gil NaCI, pH 5.5 30 gil NaCI, pH 7 0.5M Na2S04 a 0.5M Na2S04a IMZnS04 5% NaCI, pH 6.3 O.OIMKOH 8 gil Ca(OH)z 0.2M C 2Hz{COONa)z, pH 5.6 O.IM NaCI, pH 5.3 5% NaCI," pH 5.6 IMNaCI
icoIT
Later iCOIT
(j1Ncm 2)
(j1Ncm 2 )
38.8 200 80 46 66 2.38 25 3 9 640 60 0.9 8
1.1 42 38 18 26 0.1 12 0.1 0.1 830 250 8 13
Duration
Reference
97 days 5 hours 5 hours 23 hours 23 hours 10 days 35 days 30 days 30 days 7 hours 40 hours I day 4 days
697 664 664 700 700 112 44 104 197 114 118 611 607
"Sample rotated at 600 rpm. hZn-7% Fe.
The time required for reaching a steady-state value varies with the test conditions. As shown in Table 5.5, the time for the corrosion potential to reach a steady-value varies from a few minutes to several days. Table 5.6 shows the change in corrosion current with immersion time in various solutions. It is noted that the corrosion current decreases with time in some solutions whereas it increases in others. It generally decreases with time in solutions in which some kind of inhibitive corrosion product film forms. On the other hand, the corrosion current increases with time as the surface becomes more activated or roughens in solutions in which such films do not form. Walter [118] investigated the effect of time on the corrosion potential and current of zinc in a O.IMNaCI solution of pH 5.3 buffered with phthalic acid. As shown in Fig. 5.27. although the corrosion potential remains relatively constant, the corrosion current increases with time by a factor of 3. The value of the equilibrium potential also increases due to the buildup of zinc ions in the solution. Macias and Andrade [104] measured the corrosion potential and current of galvanized steel as a function of time in alkaline solutions of various pH values. Figure 5.28 shows that the corrosion potential of the zinc coating decreases with increasing pH after the first hour of immersion. However, after 33 days of immersion the corrosion potentials shift to more positive values. It was found that the formation of corrosion products during the course of the immersion passivates the surface and is responsible for the change in potential to a more positive value. Figure 5.28 also shows that corrosion currents at pH> 11 after 33 days of immersion decrease significantly as compared to the one-hour values as a result of passivation. A similar effect is observed in Ca(OHh solutions at pH 12-13.8 [175]. Yamashita [112] measured the corrosion current of zinc in 1M zinc sulfate solution and found that it decreased drastically after three days of immersion owing to the formation of corrosion products on the surface as shown in Fig. 5.29. He noted that even
151
CORROSION POTENTIAL AND CORROSION CURRENT
-1.05,------------------------------------------, 50
> .§. 40
ba
.0.
+ + 30
200
,:;-
E
<.J
c:
10
Immersion time, hours
FIGURE 5.27. Effect of immersion time on electrochemical parameters measured for zinc corroding in aerated phthalate-buffered O.IM NaCI solution, pH 5.3. ieof" Corrosion current density; io, metal exchange current density; RIP polarization resistance; Eeof" corrosion potential; Eo, metal equilibrium potential; hI" anodic Tafel slope for the metal; Me electrochemically derived mass loss; M, solution-analysis-derived mass loss. Reprinted from Walter [ 118 [, with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom.
152
CHAPTER 5 o .---------------------------------------~
-0.2 -.:J ~
~
A
-0.4
-0.6
~
C Q)
(5
-0.8
oc:
_1
a.
'iii
e o -1.2
(J
-1.4
...
+
+ +
+ 1-hour immersion
+
• 33-day immersion
-* +
-1f--# ++
-1 .6 L-________---'__________--'-__________-'-____....l 10.5
11.5
12.5
13.5
pH
100
+
B
+ +
10
+
1
4"
+
#"
+
1 1-
0 .1 I-
I-hour immersion 33-day immersion
0 . 01~----------~----------~----------~----~
10.5
11.5
12.5
13 .5
pH
FIGURE 5.28. Corrosion potential (a) and current density (b) of galvanized bars as a function of pH after I hour and 33 day of immersion in test solutions. Data are taken from Ref. 104.
10
E u (;
._0
-
0 2 present
*0 2 absent
0.1L---------~~--------~~--------~~~~
0 .01
0.1
10
Duration (days)
FIGURE 5.29. Influence of immersion time on the apparent exchange current density of the zinc electrode in I .OM ZnS04 solution at 25 °C. After Yamashita [112).
CORROSION POTENTIAL AND CORROSION CURRENT
153
with intense stirring of the solution a corrosion product film could still form. Deslouis et al. [700] reported that the corrosion rate of zinc in neutral sulfate solution decreases within one day to less than half of the initial value because of the accumulation of corrosion products on the surface. 5.7. CORRELATION BETWEEN CORROSION CURRENT AND WEIGHT LOSS RATE
It is always important for the corrosion rates determined with electrochemical techniques to correlate with the corrosion data derived from the more realistic corrosion measurements such as weight loss measurements. The correlation between the corrosion current and the weight loss rate of zinc in aqueous solutions has been investigated in several studies [104, 114, 118, 790]. Macias and Andrade [104] found that by proper! y choosing the Tafel slopes in accordance with the specific anodic and cathodic reactions, good correlation could be obtained between the corrosion current and the corrosion rate determined by weight loss measurement for galvanized steel in alkaline solutions as shown in Fig. 5.30. Similar correlation was found in solutions containing calcium hydroxide [197]. In a study by Chang and Wei [365], a good agreement was found between the corrosion rates determined by weight loss measurement and by an electrochemical technique for zinc-iron alloys in deaerated O.IM NaCI solution, as shown in Fig. 5.31. Muralidharan and Rajagopalan [790] compared the corrosion rates derived from various measurement techniques and found that in concentrated alkaline solutions the corrosion rate derived from steady-state Tafel extrapolation is in close agreement with that determined by weight loss measurement, as shown in Fig. 5.32. Compared to steady-state data, the corrosion currents obtained from transient methods are much higher. The discrepancy between the values obtained by transient and steady-state techniques is attributed to the slow diffusion of zincate ions away from the electrode surface, which Sample
E <.> 0>
E
100
'6
vi
tfl
~
.c0> '0;
15,
10
~
"
,g
'0
"2
"
14
7- - 8 3 •- 6 5
4
Qj
E
_ 3
'5
~
11
0
0. 1 0 .1
2 3 4 5 6 7 8 9 10
10
100
E lectrochem ical wei ght loss , mg / cm'
12 13 14 15 16 17 18
pH 10.95 11.95 12.62 12.79 12.68 12.76 13.00 12.97 13.17 13.43 13.58 13.81 12.59 13.02 13.40 13.52 13.56 13.74
FIGURE 5.30. Comparison of gravimetric and electrochemical weight loss results for galvanized bars. Numbers refer to sample designations in table at right. After Macias and Andrade 1104].
154
CHAPTER 5 10 -
'*
8 >a. E ai
l'!
Weight loss mel hod Electrochemical mel hod
6
c: 0
'0;
e
4
0
u
2
0
20
0
Zn
40
60
Fe (wt%) in Fe-Zn coating
80
100
Steel
FIGURE 5.31. Corrosion rates, in mils per year, at 25°C for steel, Fe-Zn alloys of different Fe contents, and zinc. Reprinted from Chang and Wei [365], with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom.
later causes redeposition of zinc. The transient data extrapolated to time zero give the rates without the effect of redeposition. Walter [118] studied the correlation between the corrosion rates obtained by linear polarization and by solution analysis in a neutral O.lM NaCl solution. The rate calculated from the polarization resistance was found to be typically 50-60% higher than that from solution analysis. Walter attributed the error in the polarization data to neglect of the metal ion deposition when the corrosion potential is near the equilibrium electrode potential. In a study by Boto and Williams [114], the corrosion currents determined by a transient polarization technique were correlated with the weight loss rates determined by
5 ,-------------------------------------~ - Weight loss
+ Steady sta te
NaOH concentration, M
FIGURE 5.32. Comparison of corrosion rates obtained by different methods in NaOH solutions at 40°C. After Muralidharan alld Rajagopalan [790J.
CORROSION POTENTIAL AND CORROSION CURRENT
155
chemical analysis. The differences between the results obtained by the two techniques were generally less than 15%. The errors in the corrosion current values were considered to be related to the number of cathodic reactions involved in the corrosion process, a better correlation between corrosion current and weight loss being obtained when only one cathodic reaction was involved. Deslouis et al. [700] found the corrosion current measured by the impedance technique in an aerated sulfate solution to be 10 to 30% lower than that determined by atomic absorption, which was explained by the fact that the electrochemical measurement is instantaneous while the atomic absorption method is time-averaged. The differences between the electrochemically determined corrosion rate and the gravimetrically determined corrosion rate can arise from many causes, which generally belong to one of two types: (1) time effects and (2) assumptions made in the determination of the corrosion current. The corrosion currents determined by electrochemical methods are intrinsically different from the corrosion rates determined by gravimetric methods. The corrosion current is an instantaneous parameter and is a function of time as discussed in the previous section. Depending on the conditions, the corrosion current may decrease or increase with time. On the other hand, the corrosion rates measured by gravimetric methods are time-averaged. Assuming the equivalent corrosion current for the gravimetric corrosion rate is igt , the relation between the corrosion current, icoIT' and igt can be expressed as (5.14) It is evident from Eq. (5.14) that the value of icorr is close to that of igt when they are measured at a time that is significantly longer than the time required for the system to reach a steady state. On the other hand, the values of i gt and icorr can be very different when they are measured during the time when icorr changes significantly. In such a case, the time-averaged corrosion current will have a better correlation with that of the equivalent corrosion current from gravimetric measurements. It has been reported that the time-averaged corrosion current, compared to a single instantaneous current value, showed a better agreement with the corrosion rate obtained by atomic absorption analysis for zinc in a neutral sulfate solution [700J. For the same reason, corrosion current and gravimetric corrosion rate provide different information about a corrosion system. Corrosion current, when measured properly, reflects the corrosion rate at the time of measurement but provides little information on the total corrosion loss. On the other hand, the time-averaged corrosion rate determined by gravimetric methods, although reflecting the total corrosion loss up to the time of the measurement, provides no information on the corrosion activity at the time of measurement. A unique advantage of the electrochemical techniques is that they can be used to measure the initial corrosion activity of a freshly exposed metal surface, which usually cannot be assessed through the use of a gravimetric method. Another common origin for disagreement between icorr and igt lies within the determination of the corrosion current itself. Depending on the corrosion system and experimental conditions, errors in corrosion current values can be caused by (i) lineari-
156
CHAPTER 5
zation of the current-potential equation; (ii) incorrect estimation of the Tafel slopes; (iii) neglect of reverse partial reactions; (iv) neglect of solution resistance; (v) use of a high potential scanning rate; and (vi) neglect of mass transport. In particular, significant errors are obtained when icorriit (it being the limiting diffusion current) is close to 1, a high potential scanning rate is employed in a low conductive electrolyte, or RQIRp is large (R Q and Rp are the solution resistance and polarization resistance, respectively). The theoretical analysis of these issues can be found in a number of studies [118, 694, 717, 718,790]. Although it is often difficult to obtain a one-to-one correlation between values of corrosion current and gravimetric corrosion rate because of the many possible error sources, experimental results indicate that a certain proportionality exists between the two values [104, 197,365,790]. For example, the curve for the corrosion current as a function of pH shown in Fig. 5.14 agrees very well in shape and position with that of the thickness loss rates shown in Fig. 9.13 in Chapter 9. Under a given set of test conditions, the relation between icorr and igl can perhaps be simplified to igl = aicoIT' with a being a constant. The value of a depends on the specific corrosion system and the measurement techniques and procedures. In situations in which the objective is to compare the relative corrosion rates, it is often not necessary to know the exact value of a.
6 Corrosion Products 6.1. INTRODUCTION The corrosion products discussed in this chapter are the solid materials formed on the surface of a corroding metal. Being a layer between the metal and the environment, corrosion products greatly affect the corrosion behavior of the metal. In general, corrosion products differ in composition, structure, morphology, and properties depending on the specific conditions under which the corrosion process occurs. Based on their effect on corrosion, corrosion products can be roughly divided into two major groups: (i) those having an effect of blocking the anodic and/or cathodic reactions and thus drastically reducing the corrosion rate, and (ii) those having little inhibiting effect or even having an effect of enhancing the corrosive reactions. The corrosion products in group ii are generally thicker, bulkier, and more porous than those in group i. Characterization of corrosion products is essential to the understanding of a corrosion process. Zinc corrosion products have been characterized by various analytical techniques such as X-ray diffraction (XRD), Fourier transform infrared (FTIR) spectroscopy, thermogravimetry (TG), glow discharge optical spectroscopy (GDOS), scanning electron microscopy with X-ray microanalysis (SEM-EDS), X-ray fluorescence spectroscopy (XFS), X-ray photoelectron spectroscopy (XPS), secondary ion mass spectroscopy (SIMS), and ion chromatography (IC) [253,462, 1164]. Each technique has its unique advantage in obtaining information on certain aspects of a corrosion product. In a recent study [1164] the following choice of techniques has been suggested: (a) XRD for determination of crystalline phases; (b) GDOS for qualitative analysis of in-depth distribution of major and minor elements except for Nand CI; (c) SEM-EDS for examination of morphology and quantitative analysis of lateral distribution of elements with atomic numbers of 6 and higher; and (d) IC for quantitative analysis of waterdissolved ionic species. This chapter compiles the information that has been obtained on the composition, structure, morphology, and properties of zinc corrosion products according to the specific environments in which they form. The corrosion environments are divided into four groups: (a) atmospheric environments, (b) waters, (c) aqueous solutions, and (d) environments other than those included in (a)-(c). The fourth group consists mostly of environments used in accelerated tests such as the salt spray test and cyclic tests. A large part of this chapter deals with the characteristics of the corrosion products formed in atmospheric 157
158
CHAPTER 6
environments since they have been the most extensively investigated. Also, the focus in this chapter is on the nonelectrochemical aspects; the electrochemical behavior of zinc surfaces covered with a passive film or corrosion products have been discussed in Chapter 3. 6.2. IN ATMOSPHERIC ENVIRONMENTS
6.2.1. Composition and Structure The composition of zinc corrosion products formed in atmospheric environments has been the subject of many studies [173, 331, 868, 1163]. Table 6.1 lists the zinc compounds that have been identified in the corrosion products formed in various atmospheres. It should be noted that although atmospheres are conventionally defined according to four general types, i.e., rural, industrial, urban, and marine, each atmosphere as a corrosion environment is, in essence, unique. That is, each atmosphere is characterized by its own particular combination of specific factors such as air composition, temperature variation, seasonal climate changes, and type, amount, and frequency of rain. Thus, it is not unusual to find very different kinds of corrosion products formed in atmospheres of the same type. Many zinc compounds can form in each type of atmosphere. However, for a specific atmosphere, only certain compounds dominate. Generally, among the zinc compounds, oxides, hydroxides, and carbonates are most often found in corrosion products [173, 331 J. Zinc sulfate, ZnS04 ·nH 20, and basic zinc sulfate, Zn4S0iOHknH 20. are also frequently found [868, 1259]. In some industrial atmospheres the amount of zinc sulfates in corrosion products can be as high as 50% [555]. In coastal areas, zinc hydroxychloride, Zn5(OH)sCI2·H20, and zinc chlorohydroxysulfate, NaZn4Cl(OH)6S04·6H20, are common compounds [173, 974, 1165]; however, the relative amount of chloride in zinc corrosion products is usually low owing to the high solubility of the chloride-containing compounds [297]. The composition of zinc corrosion products formed in a particular atmosphere generally changes with time. Friel [173] reported that the corrosion products formed on a zinc coating in industrial and marine environments after a 9-year exposure are mainly Zn 5(C0 3MOH)6 with some ZnO, ZnS04, and Zn 5(OH}gCI 2. During the exposure, ZnO becomes increasingly abundant while the amount ofZnS04 decreases. Biestek et al. [868] found that in rural, urban, industrial, and marine atmospheres hydrated zinc oxide, basic zinc carbonate, and zinc sulfate of various compositions are formed in the first two years. However, after 10 years of exposure, the corrosion products contain only basic and neutral zinc sulfates. The distribution of the chemical compounds identified in the corrosion product layer on zinc may not be uniform with depth. Flinn et al. [1171] found O/Zn ratios consistent with ZnO in the outer 15 nm of the surface layer and with ZnC0 3 or Zn(OH)2 deeper into the film. The extreme outer 1 nm is rich in surface contaminants such as chlorine and sulfur. The soluble salt ions may also penetrate into the interior of the surface through cracks and defects [331]. Many zinc corrosion products formed under atmospheric conditions are found to be crystalline [173, 1163]. Data on the crystal structures of the various zinc compounds identified in zinc corrosion products, as recently summarized by Odnevall and Leygraf
CORROSION PRODUCTS
159
TABLE 6,1, Zinc Compounds Detected in Corrosion Products of Zinc Formed in Various Atmospheric Environments Atmosphere Rural
Compounds
------------~--------
ZnO Zn(OHh ZnC0 3 Zn5(C03MOH)6 Zn4S04(OH)6,nH20 (Zn, CU)4S04(OH)6AH20 Zn4CI2(OH)4S04,5H20
Urban
Zn(OHh ZnC01 ZnS04,nH 2O Zns(C03MOH)6 Zn4S04(OH)6,nH20 (Zn, CU)4S04(OH)6AH20 Zn4CI2(OH)4S04,5H20 NaZn 4CI(OH)6S04,6HP
Industrial
iI
II
ZnO Zn(OH)2 ZnS04 ZnCO J Zn4C03(OHkH20 Zns(C03lz(OH)6 Zns(OH)sCI2H2O Zn4S04(OH)6,nH20 NaZn4C](OH)6S04,6H20
Reference( s)
325 297," 1163," 1171" 325, 1171" 297," 511," 868,,,,1> 1164,"./) Cracked, uniform layer" 1259" 297," 868,,,1> 1163," 125917 Spherical particles ( 1-5 pm)' 1164" 974" Uniform, nodular, and finegrainedd Platelet islands" Platelet islands"
ZnO Zn(OH)2 Zns(C03lz(OH)6 ZnS0 4,nH 2O Zn4S04(OH)6,nH20 Zn4CI2(OH)4S04,5H20 Zns(OH)sCI2,H2O
Marine
Structure and morphology
Platelet islands!
Platelet islands g Platelet islandsg
297," 555,,,,1> 1171" 1171" 868,"./' 1163, 1259 b 297," 1164"./' 297," 868,"./' 1163, 1259"./' 1163" 1163"./' 1163"./' 173" 297," 555",1> 173," 297," 511" 173," 555,"./' 868," 1163, 1259" 297," 868,"1> 1163 1163,"./) 1259",h 1163"./' 173," 297," 1163, 1259" 297," 868" 173" 297," 325 325 173," 511," 868,"./) 999"./' 173," 297," 999"./' 297," 868" 999, ",I> 1259",1>
Under an unsheltered condition. Under a sheltered condition.
'Ref. 1164, "Ref. 1171. "Ref. 1165, I Ref. 1163, g
Ref. 999,
[1166], are presented in Table 6,2, According to Odnevall and Leygraf [ 1166], there is a structural resemblance between hydroxycarbonate, hydroxychloride, hydroxysulfate, and sodium zinc chlorohydroxysulfate, as illustrated in Fig, 6,1. These compounds have layered structures with sheets of Zn 2+ in octahedral and tetrahedral coordination, and the main difference is the chemical content and bonding between the sheets, The structural
160
CHAPTER 6
TABLE 6.2. Crystal Structures of Zinc Compounds Found in the Corrosion Products of Zinc in Atmospheric Environments" Compound Oxide,ZnO Hydroxide, p-Zn(OHh
Hydroxycarbonate, Zns(C03h<°H)6
Cell structure
Phase structure
Hexagonal, a = 3.25 A, c = 5.21 A Orthorhombic, a = 8.49 A, b = 5.16 A, Tetrahedrally coordinated zinc atoms c=4.92A forming a three-dimensional network with "hydroxyl ions Monoclinic, a = 13.58 A, b = 6.28 A, Sheets with octahedrally and c = 5.41 A, P= 95.6° tetrahedrally coordinated zinc atoms, held together by carbonate ions
Sulfate ZnS04·H20
Monoclinic, a = 7.51 A, b = 7.59 A, c = 6.94 A, P= 116.25° Monoclinic, a = 5.95 A, b = 13.60 A, c = 7.95 A, P= 90.4° Monoclinic, a = 9.98 A, b = 7.25 A, c = 24.28 A, P= 98.45° Hydroxysulfate, Triclinic, a = 8.36 A, b = 8.37 A, Sheets with octahedrally and c = 20.68 A, a = 90.06°, P= 89.93°, tetrahedrally coordinated zinc Zn4S04(OH)6·4H20 r = 120.11 ° atoms; sulfate groups connect on either side of the sheets, which are held together by hydrogen bonding Hydroxychloride, Sheets with octahedrally and Hexagonal, a = 6.34 A, c = 23.64 A tetrahedrally coordinated zinc Zns(OH)gCI2· H20 atoms, held together by weak O-H···CI bonds Chlorohydroxysulfate, Hexagonal, a = 8.37 A, c = 13.05 A Sheets with octahedrally and tetrahedrally coordinated zinc atoms NaZn4CI(OH)6S04·6H20 are coordinated with sulfate groups; sodium atoms are coordinated between the sheets Chlorosulfate, Monoclinic, a = 10.92 A, b = 4.14 A, c=7.ISA,p= 102.62° Zn4CI2(OH)4S04·5H20 "Data from Ref. 1166.
resemblance between these compounds may facilitate the transformation from one phase into another under the proper environmental conditions. The zinc compounds found in atmospheric corrosion products may be different under rain-sheltered and unsheltered conditions, as shown in Table 6.3, which contains data reported by Johansson and Gullman [1259]. For example, the latter authors found that zinc hydroxycarbonate, Zns(C0 3MOH)6, forms only under an unsheltered condition after exposure for 1 to 5 years, while zinc hydroxysulfate, Zn4S0iOHk4H20, forms only under a sheltered condition in a rural environment. The wet/dry pattern is an important factor affecting the specific composition of a corrosion product. In particular, periodic drying has an important effect on the formation of zinc salts such as hydroxysulfate and hydroxychloride. Precipitation of these salts under atmospheric conditions may occur during every drying period of the wet/dry cycle. However, the situation is very different in a fully immersed condition, where the precipitation of these salts does not necessarily occur because they have a much higher
CORROSION PRODUCTS
161
(b)
(a)
(c)
(d)
~
Tetrahedrally-,
."
Sulfate group,
~
Y
Octahedrally coordinated Zn
Carbonate group,
~
Hydrated Na
FIGURE 6. I. Crystal structures of zinc hydroxycarbonate (a), zinc hydroxychloride (b), zinc hydroxysulfate (c), and sodium zinc chlorohydroxysulfate (d). Structural details are given in Table 6.2. After Odnevall and Lcygraf [11661.
solubility than oxide or hydroxide. As will be seen in later sectIons, these salts are not usually found on zinc surfaces that are fully immersed in solutions. In indoor environments, zinc surfaces may be covered mainly with contaminants from the air. Munier et ai. [406] determined the amount and composition of water-soluble chemical compounds accumulated on a zinc surface that had been exposed to an indoor environment for 40 years. They found that the average contamination rate was 1.2 ,ug/(cm2·yr) for chloride, 1.8 ,ug/(cm2·yr) for sulfate, and 0.33 ,ug/(cm2'yr) for nitrate, The chloride is present mainly as zinc chloride because zinc is a good scavenger for chloridecontaining species. The sulfate and nitrate are present as a mixture of ammonium, sodium, calcium, and zinc salts.
162
CHAPTER 6
TABLE 6.3. Zinc Compounds Detected in the Corrosion Products Formed under Sheltered and Unsheltered Conditions in Four Types of Atmospheric Environments" Duration of exposure Atmosphere Condition Rural
U S
Urban
U
b
S Industrial
Marine
3 months
Zn4S04(OH)6·nH20 ZnO Zn4S04(OH)6,nH20
I year
5 years
Zns(C03h(OH)6 Zn4S04(OH)6·nH20
Zns(C03MOH)6 Zn4S04(OH)6·nH20
Zn4S04(OH)6,nH20
Zn4S04(OH)6,nHzO
Zn4S04(OHknH20 ZnS04-H2O
Zn4S0iOH)6,nH20 ZnS04,H 2O
U
Probably Zn4CI2(OH)4S04,5H20 ZnS04,H 2O Zn4CI2(OH)4S04,5H20
S
ZnS04,H 2O Zn4CI2(OH)4S04,5H20
ZnS04,H2O Zns(OH)gCI2, H2O Zn4C12(OH)4S04,5H20
U
Zns(C03lz(OH)6 NaZn4CI(OH)6S04,6H20 Zns(C03lz(OH)6 NaZn4Cl(OH)6S04,6HzO
ZnO Zns(C03lz(OH)6 NaZn4Cl(OH)6S04,6H20 NaZn4Cl(OH)6S04,6H20 Zns(OH)gCI2, H 2O NaZn4Cl(OH)6S04,6HzO NaZn4Cl(OH)6S04,6H20
S
"Ref, 1259, "S, Sheltered; U. unsheltered,
A very thin surface oxide film fonns on zinc even in dry air. Leroy and Schmitz [253] investigated the surface oxide films of various zinc- and zinc-aluminum-alloy-coated steels. Based on scanning Auger microprobe (SAM) analysis, they concluded that: (a) The surface of electroplated galvanized steel in air is oxidized to a depth of more than 100 A. (b) The surface of hot-dip zinc coating (0.15% AI) is covered with a thin aluminum oxide film less than 50 A in thickness. This film is composed of 18% Zn, 22% 0, and 60% AI. The thickness of the oxide film is related to the cooling rate of the coating during the coating process because the aluminum oxide film is generated by Al diffusing from the bulk to the surface and reacting with oxygen in the air. (c) The aluminum oxide thickness on a 5% Al zinc alloy coating (Galfan) is less than 50 A, and it is less than 20 Aon a 55% Al coating (Galvalume). The thinner oxide on Galvalume could be due to a rapid passivation of the surface, generally observed on pure aluminum. Friel [173] found that amorphous AI-Zn sulfate is the principal corrosion product formed on Zn-55% AI-coated steel in industrial and marine environments. AI(OHh and AlzeOHkHzO were also found in the marine atmosphere. An increase in the exposure time in the marine atmosphere from 3 to 9 years results in a decrease of the zinc concentration in the AI-Zn sulfate and an increase in the AI(OHh concentration. The lower amount of the zinc corrosion product fonned in the longer exposure time was attributed to its more soluble nature, compared to the aluminum compounds on the
CORROSION PRODUCTS
163
surface. In another study, aluminum sulfate and hydroxide were found to be the main corrosion products of Zn-55% AI-coated steel wire in rural and marine environments [456].
6.2.2. Quantity and Morphology The amount of solid corrosion products formed on a zinc surface depends on many factors, among which rain is particularly important. As illustrated in Fig. 6.2 [1264], the amount of corrosion product washed away by rain is comparable to that remaining on the surface. It is also notable in Fig. 6.2 that the relative amount of zinc washed away by rain depends on the time of the year. Schikorr and Schikorr [555] determined the amount of adhering solid corrosion products and the relative amounts of different zinc compounds in the corrosion products formed in several atmospheric environments. As shown in Table 6.4, the amount of corrosion products remaining on the surface increases with exposure time. However, the percentage of zinc compounds in the corrosion products decreases with time, indicating an accumulation of contaminants from the atmosphere. These authors also found that although there is less corrosion under a sheltered condition, the amount of corrosion products is greater than under an unsheltered condition, indicating the washing effect of rain. Because of the lack of rain washing under a sheltered condition, the amount of corrosion products can actually be greater than the weight loss. Flinn et al. [1171] investigated the relation between the corrosion weight loss and the amount of corrosion product retained on the surface of zinc samples exposed for various lengths of time in four different locations in the United States. Figure 6.3 shows
• o 'Q
E
~
Weigh, loss Zinc content in runoff rain \vater Zinc content in corrosion product
0.2
01)
E
'" '"0
..J
.;:: ell .;;;
0. 15
~ -0
c
'"
E
0. 1
;g
0
u <> .s
0.05
ov.
Jan .
Mar.
May
July
Sep.
Month of the Year FIGURE 6.2. Variation of corrosion rate and zinc content of corrosion product and runoff rain water with the time of year. After Slunder and Boyd 1217].
CHAPTER 6
164
TABLE 6.4. Weight Loss (WL), Amount of Corrosion Products (CP), and the Percentage of Zinc Compounds in the Corrosion Products (ZP) on Zinc Sheet Samples Exposed in Several Atmospheric Environments" Test site Berlin-Dahlem, unsheltered
I month 6 months 2 years I month I year I month 1 year 2 years
Berlin-Dahlem, sheltered Beside engine shed
CP (g/m2)
WL(g/m 2)
Duration of exposure
2.5 8 15 3.3 20.2 4.4 13.3 58.8
2.2 20 52 1.4 15 7.6 55 133
ZP(%) 90 87 77 77 62 74 72 50
aRef. 555.
that the amount of corrosion products retained on the surface increases with the weight loss. It also shows that a large part of the corroded zinc is washed away by the rain since the weight of the corrosion products, containing also oxygen, water, and other contaminants, should be several times greater than the zinc weight loss if all the corroded zinc is retained on the surface. The runoff of the corrosion products was found by Flinn et al. [1171] to be mainly due to the effect of hydrogen concentration in the rainwater. The microscopic morphologies of corrosion products formed in many different environments appear to be rather similar [999,1163]. According to Odnevall [1163], small islands with thin platelets form initially, and larger islands or patches with thicker, larger, and more rounded platelets form afterward. The islands and patches grow and merge, gradually covering the whole surface. In a rural environment, small spherical platelet islands less than 1 /lm in diameter were found after 14 days of exposure under a sheltered condition. They grow in size to form larger islands. Under unsheltered conditions, networks of very thin sheets are formed within a few days of exposure. In a marine environment, small islands (20-30 /lm) with layered structures are randomly formed on the zinc surface within 24 hours of exposure [999]. Within the islands, thin hexagonal 100
D
+
• North Carolmill
E
U
0,
~ DIstrict of Columbia
80
.J:. C>
@ 40 c
0
e<;
N9W York
+
Slope = 1.645
60
'0; ~
'iii
ew Jersev Q
E
.~
.
+
t.
•
~.,
".,,,, _
+
+
0
+
D
Slop.= 0. 575
U
40
60
80
Weight lo ss, mgldm'
100
120
FIGURE 6.3. Amount of corrosion products retained on 191 zinc samples after exposure for I, 3, and 14 months at four different sites in the United States. After Zhang and Tran [1117].
CORROSION PRODUCTS
165
platelet (0.5 f1m x 3 ,11m x 2 Jim) crystals of zinc hydroxychloride, Zn s(OH)sCI 2·Hp, form. These corrosion products are not uniform but vary in thickness. They have an average thickness of 0.7 f1m after 14 days, 1.2 f1m after 30 days, and 1.7 f1m after 90 days.
6.2.3. Formation Processes The formation of corrosion products in an atmospheric environment is a complex and continuously changing process. The degree of complexity and the rate of change depend on the type of atmosphere and the various factors involved. Feitknecht [404] investigated the formation of oxide films on a cleavage surface of a zinc single crystal in air. He found that first an amorphous film was formed. After four weeks, the film thickness was -100 A, and the film was amorphous at the surface but crystalline in the interior. The very small crystals were oriented with their a and c axes parallel to the a and c axes of the zinc. The growth rate on anodically polished surface was faster, the film thickness reaching a few hundred angstroms in several days. In a review article on atmospheric corrosion mechanisms, Graedel [331 J proposed the formation reactions for the zinc compounds commonly found in atmospheric corrosion products based on the stabilities of various zinc compounds. In general, the first step in the process is the formation of oxides and hydroxides followed by the formation of zinc carbonates as the system reaches equilibrium with CO 2 in the air. In S02 or CI--containing environments, solid zinc sulfate or zinc chlorides may form. Figure 6.4
I
/
I
/
I
I
I
I
I
/
I I I I I
y(OW.I: H2 0 I
r:---.1,.., 'Zn( HC 02)2' L _____
~
FIGURE 6.4. Schematic representation of the processes involved in the formation of carbonate and organic components during the atmospheric corrosion of zinc. After Graedel133 I I.
166
CHAPTER 6
schematically illustrates the processes postulated by Graedel for the formation of zinc carbonate. The reactions involved in the formation of the various zinc compounds are shown below. For zinc hydroxycarbonate:
For zinc hYdroxychloride: Zn(OHh(s) + 4Zn 2+ + 6(OW) + 2Cr -7 ZI1s(OH)gCI 2 For zinc sulfate and zinc hydroxysulfates: (x = 4, 6, or 7)
Zn(OHh(s) + (y - l)Zn 2+ + (2y - 2)(OW) + SO~- + 4H20 -7 (y =
3 or 6)
Odnevall and co-workers [974, 999, 1163-1166] investigated the sequence of formation of various zinc compounds in the corrosion products formed in various types of atmospheric environments. In a rural environment, zinc hydroxycarbonate, Zns(C03MOH)6, forms initially; this product is also found at later times under an unsheltered condition but gradually transforms to Zn4SOiOHk4H20 under a sheltered condition [999]. Similar results were reported by Johansson and Gullman [1259]. In marine environments, the initial step consists of the formation of basic zinc carbonate, Zns(C03MOH)6, which is subsequently transformed into zinc hydroxychloride, Zns(OH)gCI 2·H20. Later, sodium zinc chorohydroxysulfate, NaZn4Cl(OH)6S04·6H20. is
( Zn(OH)2)
t
FIGURE 6.5. General reaction scheme for the formation of the major corrosion products of zinc under sheltered conditions. After Odnevall [1163].
~
~::--..~':''t~~H)6·nH20
ZnO Zn,(CO,),(OH), Zn,(OH),Cl,.H,O Zn.SO.(OH),.nH,o NaZn4Cl(OH)6S04.6H,o
Major compounds
Zn.S04(OH) •. nH 20 Zn4Cl,(OH)4S04.5H,O
1 ZnS04·nH,O
Zn,(COl),(OH),
Zn(OH), ZnS04·nH,O I Zn,(COl),(OH), Zn.SO.(OH),.nH,O Zn.Cl,(OH)4S04·5H,O
---0>-------+---------11--:1
1 year
FIGURE 6.6. Sequence of formation of the major zinc compounds found in the corrosion products formed in four different types of atmospheric environments under a sheltered condition. The circle below the compounds indicate the earliest detection of the compounds in the corrosion products.
'0
1 month
NaZn.Cl(OH),S04·6H,O
I
1 week
ZnS04·nH,O Zn.SO.(OH) •. nH,O Zn.Cl,(OH)4S04.5H,o
I
Zn,(OH}gCl,.H,O
1 day
--l
'"
(/l
--l
n
c
o
o
;0
'"tl
(5 Z
(/l
o
;0
o;0
n
168
CHAPTER 6
formed [999]. In urban and industrial environments, the initial step is the formation of Zns(C0 3MOH)6, followed by the formation of hydroxysulfate, Zn4 SOiOHknH zO, and eventually by the formation of zinc chlorohydroxysulfate, Zn4CliOH)4S04·5H20 [1165]. The final step can also be the formation of zinc hydroxychloride under some conditions. The sequence of formation of various zinc compounds and the transformations among them depend on the chemical species present in the atmosphere and also on the climatic conditions. Odnevall [1163] proposed a general reaction scheme, shown in Fig. 6.5, for the formation of several major corrosion products formed in different environments. The sequences of formation of the major zinc compounds found in the corrosion products in four typical types of environments are summarized in Fig. 6.6 based on the data in Table 6.1 and the results of Odnevall and Leygraf [1163-1166]. Initially, the zinc surface is covered quickly with zinc hydroxide, which is gradually converted into zinc carbonate. Within one month of exposure, almost all major zinc compounds can be detected in the corrosion products. In the more severe environments, such as marine and industrial environments, the formation of chloride and sulfate compounds can be very fast, occurring within one day. As corrosion continues, the various zinc compounds generally increase in quantity, but some may also disappear as a result of their transformation into other compounds, depending on the specific environmental factors. The sequences of formation of the various corrosion products are generally similar under sheltered and unsheltered conditions, although in most cases the process is faster under an unsheltered condition [974, 999, 1165]. 6.3. IN WATERS
6.3.1. Fresh Waters In fresh waters, as shown in Table 6.5, the most common corrosion products are ZnO and Zn(OH)2' The hydroxide found in corrosion products may have different structures depending on water conditions, but p-Zn(OH)2 is found to occur most frequently [404, 462]. Zinc carbonate may be the major constituent in the corrosion products when the water is saturated with air or carbon dioxide [462, 688]. The corrosion products formed in water are not always crystalline. For example, Gilbert and Hadden [437] reported that the corrosion product formed on a zinc coating immersed in cold supply water is completely amorphous. The relative amounts of zinc oxide and hydroxide in corrosion products formed in water are a function of many factors, such as temperature, aeration condition, and time of immersion. According to Kotnik [462], ZnO is generally the initial corrosion product in water in the temperature range O-lOO°C. The stable corrosion products in the lower temperature region (O-30°C) are mainly zinc hydroxides. However, between 30 and 90°C the products are mainly ZnO whereas at the boiling point they are a mixture of zinc oxide and the hydroxides. Similar results were reported by Gilbert [458]: zinc hydroxide is usually produced in cold water whereas in hot water the corrosion product is usually zinc oxide. Terada [1048] studied the dissolution of zinc oxide in water charged with CO 2, When CO 2 is passed into water in which ZnO is suspended, the oxide dissolves quickly on brisk stirring and forms a metastable solution of Zn(HC0 3h. The zinc concentration in the
CORROSION PRODUCTS
169
TABLE 6.5. Composition of Corrosion Products on Zinc in Different Waters Reference( s)
Water
Composition of corrosion product
Deionized water
ZnS(C03)2(OH)6.a ZnO" ZnO." Zn(OH)/, ZnO ZnO with some Zn(OHh
Cold water
Zn(OHh Amorphous Zn(OHh
Hot water
ZnO
462,458
Seawater
ZnO ZnCI 2 ZnCI 2-4Zn(OH)2 Zn4(C03)(OHkH20 ZnS'
179.209 209 179.209 179.209 209
a
Water saturated with air.
/>
Water saturated with CO 2 -free air.
e
Seawater containing sulfide.
654.688 458,654 462 402 462 437
solution increases to a maximum and then begins to decrease, reaching a constant value within 4 h at 30°e. No solid carbonate is present in the solid phase before the maximum concentration is reached, but afterward the solid phase changes in composition to 5ZnO·2C0 2 AH 20 and slowly approaches the composition of ZnC0 3 • At 40 and 50°C, the zinc concentration in the solution attains a constant value within 40 min, and the solid phase attains also a final composition of 5ZnO·2C02AH 20. Water temperature affects not only the composition but also the compactness of the corrosion products. Table 6.6 shows the characteristics of corrosion products formed in distilled water at various temperatures [412]. At low temperatures the corrosion products exist as a gelatinous and adherent substance. An increase in temperature to 55°C is accompanied by a definite change from the gelatinous form to a granular, nonadherent form. At 65°C the corrosion film is completely granular and somewhat nonadherent and
TABLE 6.6. Characteristics of Corrosion Products Formed in Water at Various Temperatures" Temperature (0C) 20 50
55 65 75 95 100 "Ref. 412.
Tenacity
Appearance Definitely gelatinou~ Slightly less gelatinous Mostly granular Decidedly granular, becoming flaky, and compact Decidedly granular, flaky, and compact Compact, dense, and flaky specimen and fracturing scale Varies from grayish white to black, very dense, resembling enamel
Very adherent Adherent Nonadherent Nonadhercnt Nonadherent Adherent, removed by bending Very adherent and difficult to remove by mechanical means
170
CHAPTER 6
appears to be more compact. Increasing the temperature further is accompanied by an increase in the compactness and tenacity of the film. Temperature also determines the stability of the corrosion products. Zinc oxide was found to be stable in both cold and hot water [458]. Zinc hydroxide transforms to zinc oxide in dry air, in C0z-free air. or in hot water. According to Gilbert and Hadden [437], ZnO·H 20, once dehydrated, cannot take up water to form zinc hydroxide. Kotnik [462] found that Zn(OH)z, whether of the f3 or the e form, and whether present as a layer on zinc or as a powder, is converted to zinc oxide in the presence of water at temperatures near 90°C. Neither low pressure (less than 1 mm Hg) nor storage in a desiccator causes conversion of zinc hydroxide to the oxide at temperatures of 25-30°C. Zinc hydroxide forms when zinc oxide powder is kept in water at low temperature (0-10°C) for extended periods of time. Kotnik also determined the conductivity of the corrosion products formed in water at different temperatures. It was found that the conductivity of the products formed at 88°C is greater than that of those formed at lower temperatures by a factor of 103. The conductivity of the corrosion products, which consist mainly of ZnO, does not remain constant but decreases with time. The structure and composition of zinc corrosion products formed in water varies with time, especially at an early stage. For example. Kotnik [462] reported that a short exposure to water at temperatures below 10°C results in the formation of corrosion products containing a considerable amount of zinc oxide. A longer exposure results in the formation of products composed mainly of hydroxides, with only a small amount of zinc oxide. As another example, according to Feitknecht [404], in distilled water on a cleavage surface of a single crystal, an amorphous film is formed initially. After a few minutes, starting from randomly distributed centers underneath the film, a layer of oriented crystalline oxide begins to grow, extending sideways and finally covering the whole surface. With time. about one day, the crystals gradually lose their orientation, and an uneven thickening of the film occurs, resulting in the initiation of local anodes.
6.3.2. Seawater The zinc compounds that have been identified in the corrosion products formed in seawater are listed in Table 6.5. Calcium and magnesium compounds are also found in the corrosion products, and their relative amount increases with increasing temperature [209]. Also, in seawater containing sulfide at pH > 7.2, the corrosion products were found to be mainly ZnS, which has the lowest solubility in this pH range [209]. The composition of the corrosion products in seawater also depends on the time of exposure. According to Mor and Beccaria [179] when a zinc sample is immersed in seawater, zinc oxide is formed within the first 8 h. At longer times, up to 72 h, ZnCI 2-4Zn(OHh becomes the main corrosion product, and thereafter the main products are ZnCI 2-4Zn(OH)2 and ZniC03)(OH)6·H20. Increasing temperature was found to favor the formation of more compact corrosion products [209]. The corrosion products formed on zinc anodes (i.e., dissolution products as they are formed under an anodic current), which are often used in seawater applications, have been studied by Perkins and co-workers [464, 1169]. They found that the corrosion products formed in seawater on a zinc anode sample, which was galvanically coupled to a steel sample, were mainly zinc oxide. The corrosion film was a porous three-dimensional
CORROSION PRODUCTS
171
network of discrete single-crystal plates of ZnO with a size of 10-100 ,urn. Arrays of plates had a population density on the order of 106 plates/cm2. Individual plates grew to about 30 ,urn in diameter and several microns in thickness after one week of exposure. Side plates nucleated at specific crystallographic orientations to the primary plates. The arrays of plates exhibited a preferred orientation with fast growth directions normal to the basal surface. The effects of anodic current density and the flow rate of seawater over the surface of a zinc anode on the formation of the dissolution products of zinc were investigated by Perkins et al. [1169]. They found that increasing flow velocity favors two-dimensional growth of the corrosion products. At certain velocities, the dissolution product was thin, flat, and compact but cracked. The toughest, most compact films were formed at low current density and with long time exposure under static conditions. 6.4. IN SOLUTIONS Zinc oxide is a common corrosion product in diluted solutions of salts such as NaC! or Na 2S04 with zinc hydroxide present in various amounts as a minor component. In concentrated solutions ofNaCI, Na 2S04, and other salts, the corresponding zinc salts may form in addition to zinc oxide or hydroxide, or both. The presence of solutes, in general, favors the transformation of zinc oxide to zinc hydroxide, which is the main corrosion product in water at room temperature [462]. In solutions of certain salts such as H 3 P0 4 or CrO" with which zinc can form compounds of low solubility, the corrosion products can be concentrated with the corresponding zinc compounds. The formation of zinc salt films such as phosphate films and zinc chromate films has an effect of passivating the zinc surface and has a range of industrial applications. The corrosion products formed in various solutions are shown in Table 6.7. 6.4.1. Effect of pH
The formation of zinc corrosion products in solutions is primarily determined by the pH of the solution in the absence of species with which zinc may form compounds of low solubility. In acidic solutions, zinc has high solubility and dissolves with formation of Zn 2+ ions. Since the solubility of zinc decreases with increasing pH in acidic solutions, precipitation ofZn(OH)2 occurs when a certain pH value is reached. In alkaline solutions, with pH > 9, the solubility of zinc increases with increasing pH, and in the high-pH range, zinc oxide and hydroxides tend to dissolve with the formation of zincate ions. According to Roetheli et al. [497], who investigated the corrosion of zinc in HCI and NaOH solutions of different pH values, no corrosion films are formed on zinc samples held for up to 30 days in solutions with pH lower than 5 or higher than 13.5. Corrosion films may form as a result of corrosion in the pH range 5-13.5, but the films formed at pH < 6 and pH > 12.5 are rather loose. Baugh [II OJ found, in his study of electrochemical corrosion of zinc in slightly acidic solutions of NaCI0 4, NaC!, and Na2S04' that the formation of an oxide film may occur in the pH range 3.8-5.8. The oxide film was found to be porous and not passivating. Macias and Andrade [105] investigated the zinc corrosion products in alkaline solutions. Table 6.8 shows the composition of corrosion products formed on galvanized
172
CHAPTER 6
TABLE 6.7. Zinc Corrosion Products Formed in Various Solutions Solution
a
Corrosion product(s)
Ca(OHh 0.1 gIl 0.8 gIl Saturated O.I-I.SMKOH O.SMNaCI CaC0 3, saturated, 60°C, 10 ppm SO~O.IM NaC!, 70°C, O 2 bubbling O.IM NaI, 70°C, 02 bubbling O.OOSM ZnS04 0.OO2M ZnCI 2 1O-4-2MKCI 0.SMNa2S04 IMZnS04 0.01-IMCoCI2 0.INNa2S Na2Cr20rH20 + H 2SO4 O.OIM NaHC0 3 0.IMH 3P20 7 H3P04 NaOH+ H 20 2
Reference( s)
ZnO Ca[Zn(OHhh· 2H2O Ca[Zn(OHhh·2H20 (pH < 13.3). ZnO (pH> 13.3) ZnO, Zn(OH)2 Zn(OH}z, ZnCI 2·nZn(OHlz ZnO, Zn(OHh, 4ZnO·C0 2AH 20 (s) ZnO ZnO ZnO + Zn(OHh + ZnS04 Zn(OHh + ZnCI·4Zn(OHlz ZnO ZnO+ Zn(OHh Zn(OHh ZnO + CoO ZnS ZnO + Cr203 + ZnCr04 Zn4C03(OH)6·H20 Zn2P207 Zn3(P04hAH20 ZnO + Zn particles
197 197 202 105 404 462 462 462 462 462 402 113 112 97 831 57,66 194 93 1175 359
"Room temperature unless otherwise specified.
steel in KOH solutions of various concentrations. In solutions of various KOH concentrations, the corrosion product formed on the zinc surface is mainly ZnO initially, but it is mainly Zn(OH)2 after about 30 days of immersion. The rate of transformation from ZnO to Zn(OH)2 increases with increasing pH. The morphology of the corrosion products formed in alkaline solutions was found to greatly depend on pH. The zinc oxide formed at pH 11-12 is porous and nonadherent whereas that formed at pH 12-12.8 is thin and
TABLE 6.8. Corrosion Products Identified on Galvanized Steel Bars during Immersion Tests in KOH Solutions" Corrosion product( s) KOH concentration
After 24 hours
Intermediate stage
I.SM
ZnO
ZnO (9 days)
0.6M
ZnO
0.2M O.IM
ZnO Possibly ZnO
Zn(OHh, traces of ZnO (8 days) Zn(OHh (14 days) Zn(OHh (18 days)
"Ref. 105.
End oftest ZnO·ZnFe204, traces of Zn(OH)2 (36 days) Zn(OHh, traces of ZnO (35 days) Zn(OHh (28 days) Zn(OHh (33 days)
CORROSION PRODUCTS
173
2o,-----------------------------------, • EG956, Dip "EG952, Spray + CRS958. Dip • CRS952. Spray ~ 15
(5
"
.c. coo
o
-§. 10
*E
rn 'iii
3:
5
°2~~--~4------~6------~8------~10------~12~~~~ 14 pH
FIGURE 6.7. Dissolution of phosphate coatings obtained by different methods after immersion in stirred HCI or NaOH solutions of different pH values for 30 min at 2SOC. After van Ooij and Sabata [982\.
compact. At pH 12.8-13.4 the product is a layer of well-packed Zn(OHh crystals. At pH
> 13.4 it is a continuous porous ZnO layer. The dissolution products formed in phosphate solutions, mainly zinc phosphates, vary greatly in composition, morphology, structure, and properties depending on the conditions of their formation [93-95, 256, 578, 581]. Zinc phosphate films are stable within the pH range between 3 and 12 as shown in Fig. 6.7 [982,993]. At low pH they dissolve completely, whereas at high pH they are converted to zinc hydroxide, which then decomposes to zinc oxide or dissolves as zincate ions. The possibility of formation of solid corrosion products is determined by the surface pH value, which often differs from the bulk value. Baugh [111] found that the oxide film forms at higher bulk pH values in ammonium salt solutions than in sodium salt solutions, due to the dissociation ofNH; ions within the double layer, which lowers the pH near the electrode surface. Boto and Williams [128] noted that a corrosion product forms on the zinc surface in unbuffered sulfate solution in the pH range 5.6-6.2 whereas it docs not occur in buffered solution in the same pH range. Because of the buffering effect, the surface pH is less likely to be different from that of the bulk, and a higher bulk pH value is therefore required for the formation of hydroxide in the buffered solution compared to the unbuffered solution. 6.4.2. Formation Processes
The formation and transformation of zinc corrosion products in solutions has been systematically investigated by Feitknecht [404]. He found that amorphous hydroxide is precipitated on adding hydroxyl ions to a dilute solution of zinc chloride, and it changes with time to oxide or hydroxides through the following reactions: Zn 2+ + 20H- ---7 am. Zn(OHh
CHAPTER 6
174
am. Zn(OHh
~
ZnO
pH=7-9 pH = 7-9 pH =11-12
If the Cl- concentration is higher than O.OIM and the pH is below 7, two different hydroxide chlorides (I and II) may be formed by the following reactions:
7Zn2+ + 120W + 2Cl- ~ 6Zn(OHh·ZnCI2 (II)
Similarly, basic zinc carbonates can be obtained by precipitating a solution of a zinc salt with a mixture of sodium hydroxide and sodium carbonate. They can also be formed by bubbling air containing CO 2 through a suspension of amorphous zinc hydroxide. The compositon of corrosion products formed on zinc surfaces may not be uniformly distributed. According to Sergi et al. [174], Zn(OH)2 completely covers the surface in 1M KOH for 70 days, but only a fraction of it is in the form of well-defined crystals; the coverage is not complete in the presence ofCl-. Leidheiser and Suzuki [103] found that ZnO formed on single crystals of different zinc alloys by immersion in 3% NaCI solution has different thicknesses. Feitknecht [404] reported that a zinc sample immersed in O.SM NaCI solution for several days is covered by very loose corrosion products consisting of spindlelike ZnO and platelets of p-Zn(OH)2. Craters, up to 1 mm in dimension, are randomly distributed underneath the loose oxide and hydroxide. In the middle of the crater, there is a pit filled with hydroxide chloride. After prolonged exposure, it is covered with a layer of p-Zn(OH)z. As shown in Fig. 6.8, the inner part of the wall around the crater is also composed of hydroxide chloride; then follows p-Zn(OH)2 covered partly by ZnO, and finally a zone with p-Zn(OH)2. Based on the distribution of the corrosion products across a crater, Feitknecht proposed the pH conditions for the formation of corrosion products. The pH rises from about 6.S at the crater to nearly 12 at the zones away from the crater where p-Zn(OH)z is formed. Feitknecht also noted that if a zinc specimen is immersed in a sodium chloride solution through which air is bubbling, it remains bright for several months owing to the formation of a thin layer of zinc carbonate, which is oriented with the c axis perpendicular to the surface of the specimen. The formation process of the corrosion products in alkaline Ca(OH)2 solutions has been studied by Macias, Andrade, and co-workers [197,202, 1214]. According to their results, the initial corrosion product in solutions of all concentrations is invariably ZnO. At a later stage, Ca[Zn(OH)3h·H20 becomes the main corrosion product when the solution contains more than 0.8 g of Ca(OH)2 per liter. The corrosion products are mixtures of ZnO and Zn(OH)2 when the Ca(OH)z content is lower than O.S gil. The
CORROSION PRODUCTS
17S
~
4 Zn{OHh. ZnCI,
[[[[[J
6 Zn(OH),. ZnCI,
~
B Zn(OH), ZnO layer
~
ZnO disperse
$
B Zn(OH),
t
FIGURE 6.S. Schematic illustration of the distribution of the corrosion products on zinc around an active center in O.SM NaCI solution. After Feitknecht [404[.
morphology of the corrosion products varies greatly with changes in the pH of the solution. When the pH is around 12.6, the surface is totally covered by the crystallized corrosion products during the first one or two days. As the pH increases, the size of these crystals increases, and they seal the surface almost completely, leaving only small zones of the metal surface uncovered. At just above the threshold pH value of 13.3, the corrosion products appear as isolated crystals. The corrosion products in chromate solutions form through a dissolution-precipitation mechanism [67]. Zinc dissolves as tetravalent chromium is reduced. The reduced trivalent chromium precipitates as a complex chrome gel on the metal surface. The chromate films can be represented by the general formula Cr203-Cr03·xH20 [66]. The thickness and color of the chromate film depend on the conditions employed [57,651. The color of a film reflects the amount of hexavalent chromium in the film. For example, a yellow film typically has a thickness of 0.1-0.6 flm [57]. The resistance of a yellow film is in the range of 100 to 1000 flQ [65]. Black chromate films can be obtained by adding Cu 2+ or Ag+ to chromate solution [811. Formation of corrosion products that are hygroscopic can cause water condensation at relative humidities much lower than 100% [406,557]. The zinc compounds, particularly zinc chloride, are very hygroscopic and tend to have a higher rate of moisture pickup than zinc metal [406]. Similarly, it was found that the corrosion products formed initially have the effect of enhancing the ability of the surface to retain more corrosion products afterward [1094]. The properties of corrosion products are a function of various material and environmental factors and thus essentially vary from situation to situation. The properties that are important in relation to the corrosion of zinc and its alloys are: (a) electrical conductivity; (b) compactness; (c) adherence to the substrate surface; Cd) hardness; and
176
CHAPTER 6
(e) resistance to dissolution and decomposition. These properties determine the electrochemical activity, permeability, physical stability, and chemical stability of the corrosion products.
6.4.3. Zinc Alloys There is very little systematic information about the corrosion products formed on zinc alloys although the chemical compounds in the corrosion products are mostly reported as supplemental data in many corrosion studies. It has been reported that the IJ phase in a hot-dipped zinc coating was covered by uniaxial blocks of Zn(OHh crystals after immersion in 1M NaOH for 70 days and by a layer of clusters of needle-shaped crystals of ZnO after immersion in chloride solution [174]. The amount of crystal blocks is much smaller for the ( phase compared to the IJ phase. Suzuki and Enjuzi [692] noted that the compactness of corrosion products of Zn-Fe alloy coatings is affected by the presence of metallic ions in the test solutions. Macias and Andrade [105] found the presence of ZnFe204 in the corrosion products of galvanized steel in alkaline solutions. Cheng et al. [393] identified the corrosion products on 5% AI and 55% Al zinc alloys immersed in Ca(OH}z-saturated solution of pH 12.5 to be mainly calcium aluminum hydroxide hydrate Ca2Al(OH)7·3H20. Satoh et al. [339] reported that the corrosion product formed on Zn-Ni alloys is more concentrated with ZnCI-4Zn(OH)2 than that formed on galvanized steel [339]. In another study by Short et al. [44], the corrosion product of a Zn-Ni coating formed during an immersion test in 5% NaCl was identified to be mainly Zn 5(OH)gCl z, and the coating after removal of the corrosion products was found to be rich in Ni. 6.5. IN OTHER ENVIRONMENTS Table 6.9 shows the corrosion products formed in environments other than natural atmospheric environments, water, and aqueous solutions. These environments consist mostly of those associated with laboratory simulated or accelerated test conditions. In general, it appears that either ZnO or Zn(OHh tends to be the corrosion product when the form of moisture is such that a concentrated salt solution does not form on the surface of the zinc. Other zinc compounds may form when a concentrated solution is generated on the surface. In situations in which only a thin layer of moisture is involved, such as in a humidity chamber or under water spray, a small amount of contaminant can produce a concentrated electrolyte, because of the small quantity of moisture. For example, a thin layer of moisture in open air is usually saturated with carbon dioxide, thus providing a favorable condition for the formation of corrosion products enriched with zinc carbonates. It is generally found that in clean damp air the corrosion products formed on zinc surfaces consist mainly of basic carbonate, 2ZnC03 ·3Zn(OHh [437,688]. The relative amount of zinc carbonate in the corrosion product is a function of the concentration of carbon dioxide in the air. In situations in which the access of air is restricted, for example, in the interior of coils of wire, the ratio of zinc oxide to basic zinc carbonate tends to increase with increasing restriction of air. When zinc corrosion occurs in air free of carbon dioxide, the corrosion product, after drying in air for a day or two, consists of zinc oxide only, indicating that dry zinc oxide does not transform into zinc carbonate in dry air [437].
177
CORROSION PRODUCTS
TABLE 6.9. Corrosion Products on Zinc Surfaces Exposed in Environments Other than Atmospheric Environments, Waters, and Aqueous Solutions" Corrosion product( s)
Test
Medium
Reference ---~---
Watcr Water Watcr Damp air Air (90% RH) at 70°C Air (100% RH) containing 1.5% CO 2 Air ( 100% RH) contai ning no CO 2 Air containing HCI Mix of S02 and water vapor Water containing S02 0.1 % solution of 0.88 ammonia 20 wI. % HCI 5% NaCI 1O-3M Na2S04 CI- -laden concrcte 0.05% (NH.jhS04 + 0.35% NaCI Acetic acid, pH 3.8 Alcohols or acetone. etc., + O.OSMHCI
Water sprinkled between two sheets Water trapped between a bundle of wires Fog Exposed Flowing air
ZnO
688
ZnO
437
52
Exposure
4ZnO·C02'4H ZO ZnC0 3 ZnO + Zn(OH)2 + ZnC0 3 2ZnC0 3 ·3Zn(OHh 2ZnC0 3 ·3Zn(OHh
Exposure
ZnO
437
Exposure Exposure Exposure in a sealed tank" Exposure in a sealed tank"
ZnCl 2-4Zn(OHh, ZnCl z S + ZnS + ZnS04 ZnS04·H20 + ZnS04· 6H20 2ZnC0 3 ·3Zn(OHh
811 499 437 437
Exposure in a sealed tank" Spray Cyclic wetting and drying
Zinc oxychloride ZnO Zn(OHh Zn5(OH)8CI2-H20 ZnCI 2-4Zn(OHh + ZnO
437 610 1094 1173 213
2ZnC0 3 ·3Zn(OHh ZnO + amorphous zinc compounds
437 500
Cyclic wetting and drying Spray Immersion
6X~
1170 437
--------~
"Reprinted from Helwig [52]. with kind permission from Elsevier Science Inc., 655 Avenue of the Americas. New York. hHung above the solution.
Helwig [52J identified the corrosion products formed under different types of moist conditions in various accelerated tests; he found that, in short-term tests, abundant air is needed for the formation of zinc carbonate, as shown in Table 6.10. The results of longer TABLE 6.10. Appearance and Identity of Corrosion Products Formed in Various Accelerated Tests" Test 7 cycles (30 min each) of steam pressure 4 days in water film (under glass slide)
Appearance
Identity of corrosion product(s)
White water-spot marks
ZnO ZnO
3 days in condensation cabinet
Moderatel y heavy white powder with dark areas underneath Very heavy, bulky white powder
4 days in water fog
Very heavy. bulky white powder
ZnO and 2ZnC0 3·3Zn(OH)2
1 day in salt spray
Very heavy, bulky white powder
ZnO, ZnC0 3, and ZnCI 2-4Zn(OH)2
4 days on outdoor rack
Spotty gray-whitc powder
4ZnO·C0 2 -4H 20
"Ref. 52.
ZnO and 4ZnO·C0 2-4H 20
178
CHAPTER 6
term tests (e.g., water fog, condensation, and water film for 6 months) reveal that the predominant corrosion product is 4ZnO·C0 2·4H20. Gilbert and Hadden [437] investigated the effect of the presence of acid vapors in air on the formation of corrosion products. In the presence of vapors of organic acids, zinc salts of fatty acids have been detected in the corrosion products. In moist air containing sulfur dioxide vapor, the corrosion products consists mainly of zinc sulfate. In the presence of hydrochloric acid vapor, the corrosion products are largely zinc oxychloride. Lyon et at. [213] found that the corrosion products formed in a cyclic test with a wetting solution of 0.35% NaCI and 0.05% (NH4)zS04 were a mixture of ZnCI 2·4Zn(OH)2, ZndOH)ls(S04)3Cl3(H20)s, and ZnO after 340 hours of test but consisted of only ZnO after 1020 hours. McLeod and Rogers [499] reported that the corrosion product on zinc exposed to a moving mixture of air and sulfurous acid (sulfur dioxide + water vapor) contained sulfur, zinc sulfite, and zinc sulfate. Hronsky [500] reported that the corrosion of zinc in some organic solvents, such as alcohols and acetone, containing 0.05-0.5M HCI resulted in the formation of a poorly adhering, permeable, black layer consisting of amorphous zinc compounds and zinc oxide. The corrosion products formed in organic solvents such as alcohols, acetone, and benzene generally consist of zinc oxide and amorphous zinc compounds. Mabuchi et at. [287] investigated the corrosion products formed under various organic paints. They found that generally under a hydrophilic coating, after a salt spray test, the only corrosion product is ZnCl r 4Zn(OH)z; while under a hydrophobic coating, ZnO is also found. In concrete containing no free chloride, Ca[Zn(OHhh·2H20 is the usual corrosion product formed on a corroding zinc surface [1214]. ZnO and Zn(OHh have also been identified in corrosion products formed on zinc surfaces in concrete. The process for formation of Ca[Zn(OH)3lz·2H20 can be generally expressed as
6.6. EFFECT OF CORROSION PRODUCTS ON ZINC CORROSION Generally, a corrosion product layer that is insulating, compact, adherent, and insoluble serves as a good barrier against a corroding environment. It has been established that the low corrosion rate of zinc in clean atmospheric environments is largely due to the formation of a thin, compact basic zinc carbonate film that has a low solubility in water and dissolves only very slowly. The compact nature of the film prevents the permeation and retention of oxygen and water. The corrosion rate is increased by rain, to an extent that depends on its acidity, and airborne contaminants that promote the dissolution or decomposition of the carbonate film. As a result of such decomposition in the presence of contaminants, other zinc salts may form. Thus, depending on the relative proportions of zinc carbonate and other zinc salts in the corrosion products, further corrosion may also be determined by the stability and compactness of the salt film.
CORROSION PRODUCTS
179
In water and aqueous solutions, the corrosion rate of zinc changes with temperature owing to the different properties of the corrosion products in different temperature ranges. The relatively higher corrosion rate in the 60-90°C range (Fig. 9.3 in Chapter 9) was attributed by Cox [412] to the formation of non adherent corrosion products. However, according to Kotnik [462], the formation of a nonadherent oxide film is not the reason for the relatively higher corrosion rate in this temperature range. He believed that the effect is due to the formation of a zinc oxide film which has a better electrical conductivity than the zinc hydroxide formed at lower temperatures. The zinc/steel galvanic couple exhibits a polarity reversal in hot water that is directly related to the physical and chemical properties of the corrosion product film. The ennoblement of the zinc electrode potential in hot water has been attributed in a number of studies [458,462] to (a) passivation of the zinc surface by zinc oxide, which is the main corrosion product in hot water, and (b) cathodic depolarization due to the semiconducting nature of the oxide, the electrical conductivity of which increases with temperature. Polarity reversal occurs when this ennoblement reaches a certain potential which is more positive than that of the steel. The dependence of the zinc corrosion rate on pH (Fig. 9.13 in Chapter 9) is also determined by the formation and nature ofthe corrosion products. Generally, at pH values below 5, there is no corrosion product formation, and thus corrosion depends solely on the rate of cathodic reactions, which are usually rate-determining on an active zinc surface. In the pH range between 5 and 9, the corrosion rate is hindered to a varying extent by the formation of corrosion products, which are usually bulky and porous. In the pH range between 9 and 12, the corrosion products are thin and compact and are thus passive; the corrosion rate of zinc is the lowest in this pH range. At pH values higher than 13, the corrosion rate becomes high again because the high solubility of zinc prevents the formation of a stable corrosion product film on the surface. Figure 6.9 illustrates the effect of the corrosion products formed on galvanized steel in a Ca(OH)2 solution as a function of pH in the pH range 11-14 [1214]. I
I
I no H, evolution ZnO
HJ
H,
CaHZ
isolated crystals € -Zn(OH),
I
ZnO I € ·Zn(OH), I isol. CaHZ I
localized corrosion 0 11
passivation
corrosion
12 12.2 13 13.3 Alkaline solutions containing Ca'+ ions
14
FIGURE 6.9. Corrosion behavior of galvanized steel immersed in Ca(OH)2 solutions in the pH range 11-14. CaHZ = Ca[Zn(OHh h·2H20. Reprinted from Andrade and Macias [1214 J, with kind permission from Elsevier Science-NL, Sara Burgerhartstraat 25, 1055 KV Amsterdam, The Netherlands.
180
CHAPTER 6
The formation of compact salt films, such as chromate or phosphate films, can greatly reduce the corrosion rate in a wider pH range. For example, it is reported that in distilled water a chromate film on galvanized steel delayed the formation of white rust from I day to 37 days [593]. The protective effect of chromate films on zinc is due to the stability of Cr20 3 and, more importantly, to the presence of hexavalent chromium in the film. The hexavalent chromium leaches out when in contact with water, thus providing a local supply of chromate ions to react with zinc and passivate an exposed surface area. Thus, generally, the more hexavalent chromium the film contains, the longer the protection effect. Compared to chromate films, phosphate films are generally thicker and more insulating but porous, allowing zinc dissolution through pores [993]. However, depending on the pH range of the solution, formation of the dissolution product may seal the pores and prevent further dissolution within the pores. Corrosion product films formed by other surface treatments are found to have various hindering effects on zinc corrosion. In one study it was reported that the corrosion of Zn in 3% NaCI is slightly inhibited by dipping the sample in a solution containing a low concentration of cobalt [73, 296]. In another study, it was found that the corrosion rate of zinc in aqueous solutions can be significantly reduced by adding rare-earth salts, such as CeCI 3, to the solution [605]. The inhibition is attributed to the formation of a passive cerium oxide film. It has been found in several studies [287,339,351,610] that certain zinc compounds are more protective than others. Figure 6.10 shows the corrosion resistances of several zinc alloys as a function of the relative amount of ZnCl z-4Zn(OHh in the corrosion products [610]. It appears that a high percent of ZnCl z-4Zn( OH)2 in the corrosion products indicates a higher corrosion resistance. Among the alloys tested, Zn-Mg has the highest content of ZnCI 2-4Zn(OH)z and thus the highest corrosion resistance. It has also been reported that an Alz03-dispersed Zn-Co-Cr (0.5-2:0.3-1:0.1-2%) coating was more 100
o
~ ~
Q)
50
·zc
a..
O~·~----------------L-----------------~ o 500 1,000 Time to red rust occurrence (h)
FIGURE 6.10. Relation between peak percentage of ZnCl z·4Zn(OHlz in the corrosion products of various vapor-deposited zinc alloy coatings and the time to red rust occurrence in the salt spray test. ZA, Zn-lO.7% AI; ZC, Zn-IO.0% Cr; ZM, Zn-IO.O% Mg; ZN, Zn-9.7% Ni; ZT, Zn-9.8% Ti; EG, electroplating. Reprinted from Kawafuku et al. [610], with kind permission from The Iron and Steel Institute of Japan, Chiyoda-ku, Tokyo, Japan.
181
CORROSION PRODUCTS
1000
• ZnCI 2'4Zn(OH)1
o
....
.... , , /
500
~
,/
Zn-Co-Cr-AllO]
--.,---,.------0-
~o-o
ci
,./'
ZnO
()
2' 1000
C
C
o
Zn-Co
".. ..
500'....... , j ..0-, ......... - - - - ....
,r ,,
-,----------0-
~ ~ 1000
Zn
,~
FIGURE 6.11. Amounts of ZnO and ZnCI 2·4Zn(OHh detected by X-ray diffraction in Zn, Zn-Co, and Zn-Co-Cr-AI 20 J coatings after exposure in a salt spray test (SST) for various amounts of time. Reprinted from Yasuda et al. [351], with permission from ASM International.
500
rl '\. _---------0-
Ijf"-
0---"1:
''-.., . . . . .
'=
._---'_...J 02345 6 SST Time (Oay)
corrosion resistant than a pure zinc coating, which was attributed to a higher content of ZnCI 2 AZn(OHh in the corrosion products of the Al 20 3 coating as shown in Fig. 6.11 [351]. Several explanations for the apparent corrosion-inhibiting nature of ZnCI 2AZn(OHh have been proposed. Kawafuku et al. [610] considered that ZnCI 2AZn(OH)2 is a more compact corrosion product than ZnO because the c axis of the hexagonal close-packed structure of ZnCI 2 AZn(OHh is parallel to the direction of the film growth. Others have suggested that the inhibitive effect of the corrosion products composed mainly of ZnCI 2AZn(OH)2 is probably due to a decrease of the oxygen reduction rate on ZnCI 2 AZn(OH)2 [287, 339].
7 Corrosion Forms 7.1. INTRODUCTION The form of corrosion can be defined according to the nature of the corrosion, e.g., galvanic corrosion and crevice corrosion, or according to the effect of the corrosion on surface morphology, e.g., general corrosion and pitting corrosion, or according to its effect on bulk properties, e.g., intergranular corrosion and stress corrosion cracking. The corrosion forms that are commonly found to occur on zinc are general corrosion, galvanic corrosion, pitting corrosion, and intergranular corrosion. Galvanic corrosion is a particularly important form of corrosion for zinc applications, whether as a coating, an anode, or a zinc-dust paint. For zinc, unlike many other metals, in most situations galvanic corrosion is desirable because it is required for galvanically protecting the coupled metal, usually steel. Pitting corrosion usually occurs under certain conditions when the zinc surface is covered with a passive film, or when the surface is inhomogeneous owing to the presence of impurities, or when the surrounding environment, for example, soil, is inhomogeneous. Intergranular corrosion is found to occur on certain zinc alloys, particularly zinc-aluminum alloys, when used in warm or hot, moist environments. It is an important form of corrosion for the applications of zinc alloys because aluminum is the most used alloying element for die casting and coatings. The galvanic corrosion, pitting corrosion, and intergranular corrosion of zinc are considered below in three separate sections that include detailed data and discussion of the corrosion mechanisms. General corrosion, although it is the most common form of corrosion for zinc, is not specifically considered in this chapter because it is dealt with in almost all the other chapters. Other forms of corrosion, such as stress corrosion cracking and hydrogen embrittlement, are not common in zinc applications. and the limited amount of information available is only briet1y mentioned. On the other hand, a special subsection is devoted to wet storage stain, which, although not a form of corrosion defined in corrosion text books, commonly occurs on galvanized products during periods of storage and transportation. 7.2. GALVANIC CORROSION 7.2.1. lnlroduclion The galvanic corrosion of zinc coupled to other metals, particularly to steels, has been the subject of numerous studies. Table 7.1 lists the coupled metals, the environments, and the specific aspects investigated in a number of studies. 183
184
CHAPTER 7
TABLE 7.1. Studies on the Galvanic Corrosion of Zinc Coupled to Other Metals in Various Electrolytesa Metal
Measurement(s)b
Electrolyte
Effect studied
Reference( s)
Ig Egc,lg Egc,Ig
Area effect Al vs. alloys Ni alloying
356 1102 1192
Humid gas O.lNNaCI O.OIMNaCI 5%NaC! Seawater 0.IMNa2S04,O.lM KCI, O.IM KN03
E distribution E, I distribution E distribution Egc,lg E distribution E distribution, Ig
Kelvin probe Effect of Resistance Modeling Equipment Cathodic protection Corrosion rate
1245 556 1233 1100 516 38
Fe
Seawater 3% NaCI
E distribution I, impedance
Modeling Zn-rich coating
415 5,317
Steel
CO~-,SO~-,N03 in
Egc,lg
Potential reversal
410,194
Egc,lg Egc' C Ec,lg Ea, Ec,Ig Ig Ig Igo pH Egc,Ig
Potential reversal Transient E-t Cathodic protection Galvanic E-I curve Potential reversal Corrosion products Variation of pH Ni alloying Galvanic protection Paint adhesion Surface properties Solution effect Effect of paint Flow rate RH,ctime Galvanic protection SCC of steel d Thin electrolyte Protective power
682,709,925 905 1232 1231 457 286 364,1081 1123 1242 1084 905 1099 827 1237 1261 357 1234 522 1240
Al Al alloys
3.5% NaCI 3.5%NaC! 0.6NNaCI
Cu
Cu-Ni alloy Cu, brass
hot water Hot tap water Seawater Concrete pH 3.8-9.5 0.05MNaC03 3.5% NaC! O.IMNaCI 0.6MNaCI IN NaCI Painted Synthetic seawater NaC!, MgS04, etc. 5%NaC! Seawater Cyclic wet/dry Soils 3% NaCI O.OIM Na2S04 O.OOIM Na2S04
Weight loss, E-I Morphology Transient E-t Weight loss
Ig Ig Morphology Weight loss
Ec E, I distribution E, I distribution
Stainless steel
O.OINNaCI Soils
Ig Ig
Area effect Effect of soi I resistivity
518 517
Passive zinc Pb, Fe Cd, Cu, Ni Sn, stainless steel Ti-6AI-4V AI, Cu, Pd, Fe Cu, Pd. Ni. Mg, anodized AI, Sn, Cr, steel, stainless steel, carbon-filled polyethylene
0.IMKCr04
Ig
Pitting
205
Soils 3.5% NaCI Atmosphere O.lNNaCI
fR drop
Ec-Ea
1239 1103 293, 1093 518
Ege , Ig' weight loss Weight loss
Ig
Corrosion rate Corrosi on rate Area effect
aRef. 1294.
bIg. Galvanic current; Egc. potential of couple; Ec' potential of cathode; Ea. potential of anode; C, capacitance. cRH. relative humidity. dSCC, Stress corrosion cracking.
CORROSION FORMS
185
In this section, the theoretical and practical information on galvanic corrosion of zinc and its alloys coupled to other metals, particularly steel, is organized, and a conceptual and elemental analysis of galvanic coupling between zinc and steel is presented. Various factors that may play roles in galvanic action between zinc and coupled metals are systematically discussed. The principles and practical applications of galvanic protection of steel by zinc coatings, zinc anodes, zinc-rich paints, and other means are also reviewed. A conceptual and elemental analysis is also made for the galvanic action between zinc and steel for geometries of particular importance to applications. 7.2.2. Theoretical Aspects 7.2.2.1. Factors in Galvanic Corrosion. When two dissimilar metals in electrical contact with each other are exposed to an electrolyte, a current, which is called a galvanic current, flows from one to the other. Galvanic corrosion is that part of the corrosion which occurs on the anodic member of such a couple and is directly related to the galvanic current by Faraday's law [5241. Under a galvanic corrosion condition, the simultaneous additional corrosion taking place on the anode of the couple is called the local corrosion or the self-corrosion. The local corrosion mayor may not equal the corrosion taking place when the two metals are not electrically connected, called the normal corrosion. The difference between the local corrosion and the normal corrosion is called the difference effect and may be positive, if the local corrosion decreases when galvanic current flows, or negative. A galvanic current generally causes a reduction in the total rate of corrosion of the cathodic member of the couple. In this case the cathodic member is cathodically protected. The polarity and direction of galvanic current flow between two connccted bare metals is determined by the thennodynamic reversible potentials of the metals. The metal that has a higher reversible potential in the electromotive force (emf) series is the cathode in the gal vanic couple. Table 7.2 shows the standard emf series of common metals [I 124]. In real situations, owing to the formation of a surface oxide or salt film on the surface or owing to the difference in the local electrolytes around the two coupled metals, the polarity may be different from that predicted by the electromotive series. Thus, the relative position of each metal or alloy in a galvanic series depends on the environment. Table 7.3, for example, gives the galvanic series of some commercial metals and alloys in seawater [524, 1246J. As can be noted, some metals that are low in the electromotive series, such as titanium, are actually high in the galvanic series in seawater. Compared to normal corrosion, galvanic corrosion is generally more complex owing to the fact that, in addition to material and environmental factors, it involves also geometric factors. The fundamental relationship in galvanic corrosion is described by Kirchhoff's second law: (7.1 ) where R" is the resistance of the electrolytic portion of the galvanic circuit, R", is the resistance of the metallic portion, Ec is the effective (polarized) potential of the cathodic member of the couple, and E" is the effective (polarized) potential of the anodic member. Generally, R", is very small and can be neglected. Ea and Ec are functions of the galvanic
186
CHAPTER 7
TABLE 7.2. Standard emf Series of Common Metals a
i Noble or cathodic
Metal-metal-ion equilibrium (unit activity)
Electrode potential VS. NHE at 25°C
Au-Au 3+ Pt_Pt 2+ Pd_Pd 2+ Ag-Ag+ Hg-Hg~+ Cu-Cu 2+
+1.498 +1.118 +0.951 +0.799 +0.797 +0.342
HrH+
0.000
Pb_Pb 2+
Active or anodic
J.
Sn-Sn2+ Ni-Ni 2+ Co-Co2+ Cd-Cd2+ Fe-Fe 2+ Cr-Cr3+ Zn-Zn 2+ AI_AI 3+ Mg_Mg 2+ Na-Na+ K-K+
(V)
-0.126 -0.138 -0.257 -0.277 -0.403 -0.447 -0.744 -0.762 -1.662 -2.372 -2.714 -2.931
"Data from Ref. 1224.
current I; hence, the potential difference between the two metals when there is a current flow through the electrolyte does not equal the open-circuit cell potential. In addition to the potential difference between the two coupled metals, many factors play roles in determining galvanic corrosion. Depending on the circumstances some or all of the factors illustrated in Fig. 7.1 may be involved in galvanic corrosion. Generally, for a given couple, the factors in categories a, b, and c vary less from one situation to another than the factors in categories d, e, and f. The effect of the geometric factors on the galvanic actions could, in many cases, be mathematically analyzed. On the other hand, the effect of the factors related to electrode surface condition and its effect on the reaction kinetics in real situations can be very difficult to determine. 7.2.2.2. Analysis. The mathematical description of galvanic corrosion can be very complex because of the many factors involved. It can, however, be simplified for the galvanic corrosion of zinc. In real applications, galvanic corrosion of zinc occurs mainly in two situations: when zinc is used as a coating and when it is used as a sacrificial anode. The specific geometries involved in these applications may be generalized by the scheme illustrated in Fig. 7.2a. The case in which the distance between zinc and steel, d, equals zero, represents galvanized steel on which the zinc coating is partially removed, as shown in Fig. 7.2b. On the other hand, the case in which d» (xae - d) and d» Xce (the lengths of the zinc and steel electrodes) can be considered as representing the situation when the zinc is used as a sacrificial anode, as shown in Fig. 7.2c.
CORROSION FORMS
187
TABLE 7.3. Galvanic Series of Some Commercial Metals and Alloys in Seawater"
i Noble or cathodic
Platinum Gold Graphite Titanium Silver [
Chlorimet 3 (62Ni, ISCr, ISMo) Hastelloy C (62Ni, l7Cr, 15Mo)
IS-S Mo stainless steel (passive) [ IS-S stainless steel (passive) Chromium stainless steel, 11-300/0 Cr(passive) [
Inconel (passive) (SONi, 13Cr, 7Fe) Nickel (passive)
Silver solder Monel (70Ni, 30Cu) Cupronickels (6D-9OCu, 40-1 ONi) Bronzes (Cu-Sn) Copper Brasses (Cu-Zn) [
[
Chlorimet 2 (66Ni, 32Mo. lFe) Hastelloy B (60Ni, 30Mo, 6Fe, IMn) Inconel (active) Nickel (active)
Tin Lead Lead-tin solders [
18-8 Mo stainless steel (active) 18-8 stainless steel(active)
Ni-resist (high-Ni cast iron) Chromium stainless steel. 130/0 Cr (active) [ Cast iron Steel or iron 2024 aluminum (4.5Cu, 1.5Mg, 0.6Mn) Active or anodic
J,
Cadmium Commercially pure aluminum (1100) Zinc Magnesium and magnesium alloys
"Reprinted from Fontana and Greene [524], with permission from The McGrawHill Companies.
188
CHAPTER 7
(f) Geometric factors - area of zinc and steel - distance between zinc and steel - location - shape and orientation
(a) Reversible electrode potentials
(b) Reactions - zinc dissolution - O 2 reduction ---- hydrogen evolution
/ zinc~ (e) Electrolyte properties - ionic species - pH - conductivity - temperature - volume - flow rate
/ (c) Metallurgical factors - alloying - heat treatment - mechanical working
(d) Surface conditions - surface treatment - passive film - corrosion products
FIGURE 7.1. Factors involved in galvanic corrosion of a zinc/steel couple. After Zhang [12941.
The basic relationships for the geometrical arrangement shown in Fig. 7.2a can be expressed as follows: (7.2) and x a S O,xc;::: 0
(7.3)
where Ea and Ec are the corrosion potentials of zinc and steel, respectivel y, under separate open-circuit conditions; I] a andl] care the anodic and cathodic overpotentials under the coupled condition; and ~ VR is the ohmic potential drop across the electrolyte between x" on the zinc surface and XC on the steel surface. (" the total anodic current, and (., the total cathodic current, are given by Ia
(7.4)
=f iaCxa)l dx a d
Ic =
(7.5)
f i,. (xC)l dx' o
in which l is the width of the electrodes, and (,(x") and (eX are the respective current densities on the anode and cathode. Assuming that both the anodic and cathodic reactions are activation-controlled, they can be expressed by the Butler-Volmer equation [1100]: C
)
(a)
x.c d
o X«
XC
insulating material
(c)
~
~~
(b)
zinc coating
d
electrolyte
~~
~
~
zinc anode coupled to a steel cathode. After Zhang [12941.
FIGURE 7.2. (a) General geometry of a zinc/steel galvanic couple. (b) Geometry of zinc-coated steel. (c) Geometry of a
'1..&
electrolyte
A
n
QC \C
[/)
3:::
o ;>:l
.."
Z
(5
[/)
o
;>:l
o;>:l
190
CHAPTER 7
(7.6) (7.7) in which i o" and ioc are the exchange currents for the anodic and cathodic reactions, respectively, Paa, Pac' Pea' and Pee are the kinetic constants, and B" and Bc are the area factors. The area factors vary between 0 and I, being equal to I when the whole surface is active and being close to zero if the surface is fully passivated. In the cases in which the cathodic reaction is limited by oxygen diffusion in the electrolyte, Eq. (7.7) is replaced by
e
(7.8) with F the Faraday constant, Do the diffusion coefficient of oxygen in the electrolyte, Co, the oxygen concentration in the bulk, and J the thickness of the diffusion layer. - The total ohmic potential drop in the electrolyte between any two points on the surface of the anode and the cathode for the situation in Fig. 7.2a consists of three parts: (7.9) where Ll Va' Ll Vc ' and Ll Vd represent the ohmic potential drop in the electrolyte in the x direction across the anode, across the cathode, and across the distance between the anode and cathode, respectively. They can be further expressed by: x"
(7.10)
f
Ll V,,(x") = i,,(x") dR(x") d
Ll
Vc(x c')
c
f
X
(7.11)
= ic(x
C
)
dR(x
C
)
o
(7.12) where Rd = pd/tl, with p the resistivity of the electrolyte, t the electrolyte thickness, d the distance between the anode and cathode, and l the width of the electrodes, andi" andic, given by the following equations, are the sums of the current from x" to Xae on the anode and from XC to Xee on the cathode, respectively.
ia =
f i" (x")l dx"
(7.13)
:/'
ic =
ea
f ic
x
X
c·
(xC)l dx c
(7.14)
CORROSION FORMS
191
It can be seen that the factors listed under categories (a)-( e) in Fig. 7.1, contribute to galvanic action through their effects on the electrochemical reaction kinetics given by Eqs. (7.6) and (7.7). For example, changing the pH of the solution may cause a change in the kinetic parameters, io", ioc ' Pall' Pac' Pca' P"o, or it may cause a change in the effective area, 0" or ()" through passivation. On the other hand, the geometric factors under category (0 affect the galvanic corrosion through the parameters in all the equations from (7.4) to (7.14). Equations (7.4)-(7.14) apply to a rather general geometry. For a specific application, they can be further simplified. In the case of Fig. 7.2b, representing the galvanic action on zinc-coated steel where the bare steel surface is next to the zinc-coated steel surface, that is, d = 0, the term i1Vd in Eq. (7.9) becomes zero. For the geometry in Fig. 7.2c, representing the situation of galvanic protection of steel by a zinc anode when d» (xae - d) and d > > x," !"and!, in Eqs. (7.4) and (7.5) simply become i,,Aaandi,A, with A,,=l(xae-d) and A, = {xce , the areas for the anode and the cathode, respectively. In addition, i1 V" and t, Vc in Eq. 0.9) can be taken as zero because they are very small compared to i1V". In such a case, the geometry of the galvanic cell (i.e. shape and orientation of electrodes, size of the electrode etc.) become insignificant in determining the galvanic action of the couple, and the galvanic corrosion of the anode, as well as the galvanic protection of the cathode surface, becomes uniform. Thus, the galvanic action can be fully described by the polarization characteristics of the anode and the electrolyte resistance. In this case, the relation between the effective potentials, galvanic current, and resistance can be graphically represented by the anodic and cathodic polarization curves as shown in Fig. 7.3. When the solution resistance R is infinite, no current flows, and E( - E" is the open-circuit value of the cell potential. As R is made smaller, ! increases and E( - E" becomes smaller because of polarization. When R is zero, Ec - E" becomes zero, and the galvanic current reaches the maximum, known as the "limiting galvanic current," and is at the intersection of the polarization curves of the anode and cathode. The exact shapes of the anodic and cathodic polarization curves depend on the electrochemical reaction kinetics of each metal in the electrolyte and are thus functions of pH, temperature, solution
IR 1________________ Corro$ion potenlial _
, I
Limidng currenl
: '-l
, I
Galvanic current FIGURE 7.3. Graphic estimation of galvanic current.
192
CHAPTER 7
concentration, diffusion, formation of passive films, etc. Normally, the anodic dissolution of zinc is activation-controlled, with a relatively small Tafel slope (around 40 m V) [43-45]. The cathodic reactions on the steel surface, on the other hand, can either be activation- or diffusion-controlled depending on the conditions, particularly solution pH and aeration conditions. The typical shapes of the anodic polarization curve for zinc (Ea) and the cathodic curve (Ec) for steel are illustrated in Fig. 7.3. A galvanic-corrosion system may operate under different control mechanisms. If the anode does not polarize and the cathode does, then, in solutions of low resistivity, the current flow will be controlled entirely by the cathodic electrode. Such a situation is considered to be under cathodic control. If the anode polarizes and the cathode does not, the status is reversed, and the system is said to be under anodic control. If neither electrode polarizes and the current flow is controlled by the resistivity of the path, mostly in the electrolyte, then the system is said to be under resistance control. 7.2.2.3. Potential and Current Distribution. The galvanic corrosion of the anode and the galvanic protection of the cathode are essentially governed by the potential distribution across the surface of the electrodes. The galvanic current distribution can be determined from the potential distribution when the potential-current relationships for the electrodes are known. The exact description of the potential and current distributions on the surfaces of a galvanic couple can be obtained by solving Laplace's equation: (7.15) There are a number of mathematical models using Laplace's equation for galvanic systems with different cell geometries [321, 1233, 1235, 1236, 1244, 1248]. In these models the polarization parameter, L i , is often used: (7.16) where p is the specific resistivity of the electrolyte; Ii is the current density and '7 j is the overpotential of the anode or the cathode. The polarization parameter, defined originally by Wagner [751], has the dimension of length and provides an electrochemical yardstick for classifying electrochemical systems. Waber [321, 1248] and other authors [415, 1235] used the polarization parameter to describe the behaviors of galvanic corrosion cells. According to Waber [321, 1248], whether the anode and cathode behave "microscopically" or "macroscopically" is determined by the ratio of the dimension of either electrode, C i, divided by the polarization parameter L i • The mathematical modeling indicated that, when the ratio C/L j is small, the variation of current density across an electrode is small; i.e., the electrode behaves microscopically. On the other hand, when the characterizing ratio is large, i.e., when the electrode dimension is much larger than L i , the electrode process can be regarded as macroscopic, and the variation of current density across the electrode surface is large. MaCafferty [1235] modeled the potential distribution of a concentric circular galvanic corrosion cell by assuming a linear polarization for both the anodic and the cathodic reactions. Figures 7.4 and 7.5 show the calculated potential distribution and current distribution as a function of electrolyte thickness for the case in which the polarization parameter of the anode, La> is I cm and that of the cathode, Le, is 10 cm. It can be seen
193
CORROSION FORMS
E~
-1.0
ELECTROLYTE THICKNESS
~
0
~
'3w ...J <{
~
z 0.5
w
f-
a0w
0
ac:: f-
U
FIGURE 7.4. Distrihution of electrode potential for L" = I em and Lc = 10 em for different electrolyte thicknesses. h. Anode radius a = 0.5 em. cathode radius c = 1.0 em; E~; = 0 V. E~) = I V. After McCafferty
112351
W ...J W
E~
-0.0 e==I::1==::.o~--c--.-<.J
o
1.0 RADIUS, r (em)
that, in the bulk electrolyte, the potential variation across the electrodes is small, but both the anode and the cathode are strongly polarized; the actual electrode potentials are far away from E~ and E? . Under a thin electrolyte, the potential variation is large from the anode to the cathode, but both the anode and cathode are only slightly polarized except for the areas near the boundary between the anode and the cathode. The galvanic current increases with increasing electrolyte thickness. Also, the current is distributed on the electrode surface more uniformly in bulk solution than in thin-layer solutions, where the current is more concentrated near the contact line in the thin electrolyte. According to the calculation of Doig and Flewitt [1233], the potential distribution is uniform in the thickness direction under a thin layer of electrolyte, e.g., I mm. It is nonuniform when the cell is under a thick electrolyte. Similar results were reported by Morris and Smyrl [1236] for a galvanic cell with coplanar electrodes. The potential distribution of galvanic corrosion with more general geometrical conditions has been calculated by Munn and Devereux using a finite-element method [415,1244]. One problem in mathematical modeling is the assumption that both the anode and the cathode have a linear or Tafel polarization behavior over the entire potential range. However, the polarization characteristics of a metal electrode are generally different for the anode and for the cathode, and they vary in different potential ranges. Sometimes they also vary with the physical elements in the galvanic cell, such as electrolyte thickness. In addition, the electrode properties of the coupled metals usually change with time due to changes on the surfaces and in the solution. These elements need to be taken into consideration in using a mathematical model for predicting long-term behavior in a real galvanic system.
CHAPTER 7
194
0.3 r - - - - -- -----,,----- -- - - - , CATHODE
ANODE
BULK ELECTROLYTE ~
~ .- 0.2 ~-
iii
zw
0
IZ
w
a: a: ::J u ...J
«
0.1
b=0010/
u 0
...J
o o
~~/
0.5
1.0
RADIUS , r (em)
-0.2
FIGURE 7.5. Current distribution for different electrolyte thicknesses under the same conditions as in Fig. 7.4. After McCafferty [1235] .
r-----------------..."...--~
• 165/lm
-0.4
T 70/lm
+ Bu lk electrolyte
w
ill
>
~ -0.6
a
Q)
a. -0.8
Zn
o
10
20
30
40
50
60
Specimen length, mm
FIGURE 7.6. Distribution of potentials on the electrode surface of a galvanic CulZn couple in a O.IN NaCI solution for different electrolyte thicknesses. Data are taken from Ref. 556.
CORROSION FORMS
FIGURE 7.7. Schematic representation of the electrochemical cell used for obtaining data on potential and current distributions. D is the distance between the zinc and the steel electrode. W is the width of the steel electrode, and X is the position on the steel electrode. Reprinted from Zhang and Valeriote [5221, with kind permission from Elsevier Science Ltd. The Boulevard, Langford Lane. Kidlington OXS 1GB. United Kingdom.
195
Re
/ /
/
/
The potential distribution on the electrode surface of a galvanic couple can be experimentally determined. Rozenfeld [556] showed that the potential variation of the surface of a coplanar zinc/copper couple greatly increases with decreasing electrolyte thickness on top of the surface, as shown in Fig. 7.6. The sharpest potential changes take place on the copper cathode, while the anode does not polarize at all. Zhang and Valeriote [522] measured the potential and current distributions of a coplanar zinc/steel couple under thin-layer electrolytes of various thicknesses and salt concentrations using the cell design shown in Fig. 7.7. The potential distributions on the zinc and steel are similar to that measured on the zinc/copper couple shown in Fig. 7.6. Figure 7.8 shows that the galvanic current is higher for thinner electrolytes, which is opposite to the prediction of the mathematical models [1235, 1236]. In these models, the rate of cathodic reaction on the cathode is assumed to be independent of the electrolyte thickness. However, under thin-layer electrolytes, the oxygen diffusion rate is increased since oxygen reduction is the main reaction on the steel cathode. Figure 7.9 shows that the galvanic current is larger for a thinner electrolyte when the anode and the cathode shown in Fig. 7.7 are close together, but the opposite is observed when the two electrodes are far apart [522]. This change of the relative galvanic current values for small and large distances is due to the change of the rate-limiting process from 500 -0.09
mm + 0.14 mm *0.27 mm
r--0.54mm *1.08mm +2.16mm
('oJ
E 400 FIGURE 7.8. Effect of changes in electrolyte thickness on galvanic current density vs. time curves measured in O.OIM Na2S04 on two electrically connected zinc and steel strips which were 1 mm apart. Reprinted from Zhang and Valeriote [5221, with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom.
u
~ ~ -300 ~
'(ij
c ~200
-
~r'-+~+1---+--~--+---~-+
~ 100 -~~i=~~~==~l====I====~~~I====I u
OL---~--~----~--~----~--~----~~
o
50
100
150
200
Time (sl
250
300
350
196
CHAPTER 7 •
J.36mm
o ~'--~--~--~--~--~--~--~--~----
o
0.5
1.5
2
2.5
3
3.5
4
4.5
Distance (em)
FIGURE 7.9. Galvanic current as a function of the distance between zinc and steel in O.OOIM Na2S04 solutions of different electrolyte thicknesses, t, for a steel width of 1 mm. Reprinted from Zhang and Valeriote r522 J, with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom.
oxygen diffusion at a small distance to ohmic conduction in the electrolyte at a large distance [522]. The galvanic corrosion of zinc under thin-layer electrolytes measured experimentally for the couple illustrated in Fig. 7.7 is summarized in Fig. 7. 10 [522]. The galvanic current (J) increases with the area of the steel (WI) up to a certain size and then decreases slightly with further increases in the area. It decreases sharply as the distance between zinc and steel (D) increases because the system becomes ohmic-resistance-controlled. It is relatively less sensitive to the variation of electrolyte layer thickness (t). The width of the zinc has little effect on the galvanic current because most anodic reactions take place at a very narrow area at the edge closest to the steel.
7.2.3. Practical Factors 7.2.3.1. Effect of Coupled Metals. Different metal alloys have different electrode potentials. However, the extent of galvanic corrosion of a metal does not always parallel the potential difference between the coupled metal alloys. Table 7.4 shows that, although the potential difference between steel and zinc is much less than that between stainless steel and zinc and that between Ti-6AI-4V and zinc, the amount of galvanic corrosion is much larger in the zinc/steel couple than in the other two couples [1103]. This indicates that the difference between the corrosion potentials of the uncoupled metals is not a reliable indicator of the extent of galvanic corrosion. Similar results have been reported on the galvanic corrosion of zinc when coupled to various metal alloys in different atmospheres [293]. As shown in Table 7.4, the amount of corrosion is greater when zinc is coupled to mild steel than to copper, although the potential difference between zinc and steel is smaller than that between zinc and copper. In these situations, other factors, such as reaction kinetics and formation of corrosion products, rather than just the potential difference between the two metals, are the rate-determining factors in the galvanic corrosion. The different galvanic corrosion rates of the anodes coupled with different cathode materials can be explained in some cases in which the cathodic reaction is oxygendiffusion-limited by the different diffusion rates of oxygen through the oxide films. On
197
CORROSION FORMS
200 ~A
D = O.OSmm
.
\
,...... ,, , ,
,
\ I
\
I"
"
'
' \ \
200 ~A
w
50
D =Smm
mm
(a)
200 JJA
w D =40mm
(b)
w (c) FIGURE 7.10. Three-dimensional plots of the galvanic corrosion current of zinc in 0.00 IM Na2S04 solution for different distances (D) between the zinc and the steel. Reprinted from Zhang and Valeriote 1522). with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane , Kidlington OX5 I GB , United Kingdom.
the other hand, when diffusion is not the limiting process, the variation in galvanic corrosion rate can only be due to the cathodic efficiency for oxygen reduction in the oxide scale on the cathode surface 11103]. As a result of the galvanic corrosion of the anodic metal, the corrosion of the coupled metal or alloy is generally reduced, that is, cathodically protected. The extent of protection for different metal alloys galvanically coupled to zinc has been investigated in atmospheric environments [293,545], in seawater [1246], and in soils [357, 517]. The galvanic corrosion of zinc is, however, not always beneficial to the coupled metal. It has been reported that, although zinc is anodic to aluminum, the amount of aluminum corrosion in 3.5% NaCI solution is increased when aluminum is coupled to zinc, compared to that in the uncoupled condition [356]. Similar results were reported by Mansfeld et al. [1102] for a zinc/aluminum alloy couple in 3.5% NaCI solution. The higher dissolution rate of the coupled Al alloy compared to the uncoupled one is attributed to increased alkalinity on the surface of the Al alloy due to the cathodic reaction.
CHAPTER 7
198
TABLE 7.4. Galvanic Corrosion Rates of Zinc (20 cru 2) Coupled to Various Alloys of Equal Size Tested in 3.5% NaCI Solution for 24 Hours" Galvanic corrosion rate r~
Coupled alloy None Stainless steel 304 Ni Cu Ti-6AI-4V Sn 4130 steel Cd
"
t.v d
!.j1mJyr)
!.j1mJyr)
(mV)
0 244 990 1065 315 320 1060 600
101 705 1390 1450 815 810 1550 660
905 817 811 729 435 483 258
o
"Data from Ref. 1103. hMeasllred as galvanic curren(. cMeasured as weight loss. dPotential difference between zinc and the coupled metal before testing.
As can be noted in Table 7.4, the weight loss of zinc when galvanically coupled to other metal alloys can be much larger than the sum of the galvanic corrosion calculated from the faradaic current plus the normal corrosion measured in a noncoupled condition. This implies that self-corrosion (or the local corrosion) of zinc is enhanced by galvanic alloy coupling to another alloy. 7.2.3.2. Effect of Alloying. Addition of alloying elements in zinc changes its electrochemical properties, such as electrode potential, dissolution kinetics, oxygen and hydrogen reduction overpotentials, and formation of solid surface films. Since zinc is widely used in applications in which the galvanic protection of steel is an essential requirement, alloying is usually engineered to improve the normal corrosion resistance but not to reduce by much the electrode potential difference between zinc and steel. In general, additions of small amounts of alloying elements change the corrosion potential of zinc little. With additions of alloying element to about 10%, the potential of the zinc alloy may change by 50-100 illV, usually to a more noble value than the corrosion potential of zinc, as shown in Table 5.4 in Chapter 5. For alloys with more noble elements such as Cu, Ni, and Fe, the potential can be much more positive when the concentration is high. For example, Baldwin et ai. [1222, 1223] found that Zn-Ni alloys galvanically corrode in O.6NNaCI solution when coupled to aluminum alloys or steel up to about 14% Ni concentration, above which the polarity reverses and the corrosion of the coupled aluminum alloys or steel is accelerated. The corrosion potential of an alloy in an electrolyte is a function of time. It tends to change to more positive values as the time of immersion increases because, in most cases, the preferential dissolution of zinc causes an enrichment of the more noble elements on the surface. The polarization behavior of zinc can also be significantly affected by alloying. High polarization resistance is often not desirable for zinc when used for galvanic protection of steel because a large polarization of the zinc anode reduces the potential available for the polarization of steel.
CORROSION FORMS
199
600
5
Eu OJ
- - Weight loss - - - Number of coulombs
500 '"
4
.c E
Zn·Fe
E vi
400 ~ o u
.cC1>
300 '0
'" .2
Q;
'iii it u
.c
200 ~
z
'6 0 c: <:
100 o L-------------~----~------~----~0
o
20
40
60
80
100
Area of steel cathode , cm 2 FIGURE 7.11. Effect of area of steel cathode on the weight loss of Zn anode (100 cm" in area) and on the number of coulombs flowing through the Zn/steel couple over a 96·hr period in 1M NaCI solution at 25 °C. After Pryor and Keir r1242\.
7.2.3.3. Effect of Area. The effect of anode and cathode areas on the galvanic corrosion depends on the type of control over the system. If the galvanic system is under cathodic control, variation in the area of the anode will not change the total amount of corrosion much but variation in the cathodic area will. The converse obtains if the system is under anodic control. The total amount of corrosion will be affected by a change in the area of either electrode if the system is under mixed control. When the system is primarily under resistance control, the corrosion will only change with changes in electrode area if the resistance of the electrolyte also changes with changes in the areas of the electrodes. Pryor and Keir [1242] studied the effect of the areas· of zinc and steel electrodes on the galvanic corrosion of zinc in 1M NaCI aerated solution. Figure 7.11 shows that the 20 r-~-----------------------, 400
~
§
• mg , cm 2 + mg
15
300
OJ
E
vi
.2
.cOJ
10
N
'" .2 200 ~
+
'0;
3: 'iii
+
'0;
it u c:
0>
E
100
5
0
L-____- L_ _ _ _
0
20
~
______
40
~
60
_ _ _ _ _ L_ _ _ _ _ _
80
~
~~o
100
Area of zinc anode, cm'
FIGURE 7.12. Effect of area ofZn anode on the total corrosion and on the intensity of corrosion of Zn coupled to mild steel (100 cm 2 in area) over a 96·hr period in 1M NaCI at 25°C. After Pryor and Keir 11242\.
200
CHAPTER 7 120r---------------------------------~--,
.s 100 Ol
c 0
80
~
(ij
u
.2
.s Q)
40
:l
"C
.2
g
20 0
0
20
60 40 Area of anode, em'
80
100
FIGURE 7.13. Effect of area ofZn anode on the weight loss due to local action in Znlsteel couple in 1M solution at 25°C. The area of the steel cathode was 100 cm2. After Pryor and Keir [1242].
galvanic corrosion of zinc increases with increasing steel cathode area. On the other hand, the galvanic corrosion of zinc changes only very slightly with increasing zinc anode area, as shown in Fig. 7.12. It is noted in Fig. 7.13 that much of the increase in the weight loss of zinc with increase in the area of the zinc anode shown in Fig. 7.12 is due to the increase in the local corrosion. The local corrosion is obtained by the total weight loss minus that calculated from the charge passed through the couple. These results indicate that the galvanic corrosion of zinc is mainly cathodically controlled. This is confirmed by the polarization curves of the zinc/steel couple shown in Fig. 7.14. The shape of the curves suggests that variation of the steel area will significantly change the galvanic current but variation of the zinc area will change the current only slightly.
\'l
>,-0.6
ca E -S -0.7 0..
10 em'
-0.8 -0.9 L -_ _ _--'-_ _ _ _' - -_ _ _--'-_ _ _ _' - - - - - '
o
400
800
1,200
1,600
Current, JlA
FIGURE 7.14. Effect of area of steel cathode on the polarization curves for the Znlsteel couple in 1M NaCl. The area of the Zn anode was 100 cm2. After Pryor and Keir [1242].
CORROSION FORMS
201
TABLE 7.5. Corrosion Rates of Equal-Area Zinc/Steel Couples in Various Solutions" Corrosion rate (jlm/yr)" Uncoupled ------
Solution O.05MMgS04 O.05M Na1S04 O.05MNaCI O.OOSMNaCI Carbonic acid Calcium carbonate Tap water
Coupled
-----_.
Zinc
+' 285 254 112 10.2
+ +
Steel 66 254 254 178 73.7 150 71.1
----~
-----~------
Zinc 86.4 838 762 218 38.1
+ +
~----
---
Steel
----------
+ + + + + + +
"Ref 1099. !'Specimens of equal area partially immersed for 39 day~.
(Pills signs indicate that specimens 1!ained weight.
The effect of area varies from situation to situation. It has been found that the polarity of the zinc/steel couple in hot water reversed faster for a larger steel-lo-zinc area ratio [925]. Schick 1518] found that corrosion of galvanized steel coupled to 301 stainless steel in a solution containing 266 mg cnl and 70 mg SO~-Il is controlled by both anode and cathode areas. Mansfeld and Kenkel 1356] found that, for the Zn/ Al couple in 3.5% NaCI solution, the galvanic current density changed little with the variation of both the zinc anode area and the aluminum cathode area, largely owing to the inactive surface of the aluminum. 7.2.3.4. Effect of Solution Factors. In aqueous solution, zinc is normally anodic to most other common metal alloys and corrodes galvanically. However, in some solutions in which passivation occurs, zinc can be cathodic to other metal alloys owing to the higher corrosion potential of the passive surface. Table 7.5 shows the corrosion loss of zinc and steel in coupled and uncoupled conditions in various solutions [1099]. In all the sol utions, the galvanic action results in a protection of the steel, but the amount of zinc cOiTosion varies with the composition of the solution. The difference in the corrosion rates in magnesium sulfate and sodium sulfate indicates the significant effect of cations on the reaction kinetics. The presence of metal ionic species more noble than zinc, such as Cu 2+ in solution, is known to enhance the corrosion of zinc due to the mini-galvanic cells between zinc and the copper islands deposited on the zinc surface [737 J. In the cases in which there is only a limited amount of electrolyte, the composition of the electrolyte may significantly change as a result of the galvanic action. Massinon and co-workers 1364, 1081] found an increase of pH at a confined electrolyte after a certain time of galvanic action for a zinc/steel couple. Pryor and Keir [12421 pointed out that when the distance between the anode and cathode is small compared to the dimensions of the electrodes, the galvanic corrosion is small because of the limitation in the mass transport of the reactants and reaction products. The position of a galvanic couple in the solution can also affect the galvanic actions between the coupled metals. Shams El Din et al. [205] found that there is a larger potential variation near the solution surface between a zinc anode and a copper cathode, which are
202
CHAPTER 7
half-immersed in the solution, due to the higher oxygen concentration near the surface than in the bulk solution. 7.2.3.5. Effect of Surface Condition. Formation of a surface film, whether a salt film or an oxide film, may significantly change the properties of the surface. It may not only affect the rate of galvanic corrosion but may also affect the polarity ofthe galvanic couple. Usually, in low-pH solution, zinc corrodes without the formation of solid corrosion products on the surface. The corrosion products formed in neutral and slightly basic solutions are oxide and hydroxides, usually only loosely attached to the surface [404, 462]. The corrosion products formed on the zinc surface in the pH range between 9 and 13 have varying degrees of compactness and can result in passivation of the zinc surface [497, 128]. The presence of certain ionic species, such as carbonate, phosphate, and chromate, can enhance the formation of a passive film in a broader pH range [710]. As a result of passivation, the potential of zinc can shift to more positive values, thus changing the galvanic behavior of zinc when coupled to another metal. Typically, for example, if a stable and compact zinc oxide is formed, the zinc electrode may show a potential more noble than -0.5 VSCE ' This potential is considered to be related to the semiconducting properties of zinc oxide. Because zinc oxide is an n-type semiconductor and has a flatband potential of between -0.4 and -0.6 VSCE [30, 514, 526], at equilibrium a positive overpotential is required to balance the charge accumulation at the solid/electrolyte interface. In certain cases, when the formation of a surface film is not complete, a part of the zinc surface is passivated and acts as the cathode to form a local galvanic cell, causing an enhanced corrosion of the rest of the nonpassivated zinc surface [1240]. Shams El Din et al. [205] found that galvanic current was developed between a passivated zinc sample and a partially passivated sample positioned in a two-compartment cell, containing O.IM K2Cr0 4 in one compartment and O.IM K2Cr04 and some NaCl in the other. Pits were found to be generated as a result of such a corrosion situation. The galvanic action can vary also with the surface condition of the metals coupled to zinc. Different kinds of surface films can form on the metals to change the surface condition. For example, aluminum has a low reversible electrode potential but is usually cathodic to zinc in neutral or acidic solutions, owing to the formation of a passive aluminum oxide film. Formation of iron oxide of the form Fe203 may not change the iron corrosion potential much but may change the electrode behavior of iron because Fe 20 3, like ZnO, is an n-type semiconductor, which facilitates the cathodic reaction but hinders the anodic reaction [1124,476]. Jordan [286, 740] studied the effect of corrosion products of zinc and steel on the galvanic corrosion rate of zinc. He found that the galvanic corrosion rate is dependent on the behavior of the corrosion products. In general, fresh, hot-dip galvanized zinc coating is slightly anodic to zinc that is covered with white rust. The rust of steel, as compared to bare steel, may accelerate the galvanic corrosion of zinc. According to Stratmann and Muller [1124], oxygen reduction on an iron electrode is greatly increased by the formation of the rust because oxygen can be reduced in the iron oxide scale, which has a much larger effective surface area.
CORROSION FORMS
203
TABLE 7.6. Polarity of Zinc/Steel Galvanic Couples in Various Waters ~-------
Couple
-----
Water
T(0C)
pH
CO 2(1
02a
8 8
A P
B P
7.7 7.7 7.7
B A B
B A P
7.7 7.7 7.7 7.2
B B B P P
P A B P P
Polarity
---~----
Zn/stcel Zn/steel
Distilled Distilled
85 85
Zn/steel Zn/steel Znlstecl
Hard Hard Hard
Room
Zn-Fc/steel Zn-Fe/steel Zn/Zn-Fe Zn/stcel Zn/stccl
Hard Hard Hard Tap Tap
85
85 8S
85 8S
74 70
Always anodic Sometimes cathodic Always anodic Always anodic Sometimes cathodic Strongly cathodic Always anodic Always anodic Anodic Cathodic
Refcrence
-".-,_.-
709 709 709 709 709 709 709 709 4S7
682
._----_.II
A, ahscnt: P. present; 8, absent or present.
7.2.4. Polarity Reversal The polarity of a zinc/steel galvanic couple may reverse under certain conditions; that is, the steel becomes the anode, and the zinc becomes the cathode. The phenomenon was first reported by Schikorr [682] in 1939. Many studies have since been undertaken to determine the different conditions for polarity reversal, with particular attention to that occurring in hot water and diluted solutions [194, 709, 457, 657]. Table 7.6 lists the polarity of zinc in various waters as reported in the literature. The change in the zinc electrode potential is chiefly responsible for the reversal of polarity since the potential of the steel remains relatively unchanged [709]. It is generally found that polarity reversal does not occur in distilled water up to 6S 0 e, and, without the presence of oxygen, it does not occur in hot distilled water [709, 409, 457]. Many other factors, besides temperature, such as dissolved ions, pH, and time of immersion, are found to affect the polarity of a zinc/steel couple. In normal circumstances, such as in cold water, the zinc is anodic to the steel and thus provides sacrificial protection for the steel. When polarity reversal occurs for a galvanized steel, the steel is not protected cathodically by the zinc coating, and the corrosion of the steel underneath the zinc coating occurs when the coating is completely penetrated by localized corrosion such as pitting. On the other hand, the corrosion of zinc is reduced because the zinc coating is in a passive state. The zinc, instead of being galvanically corroded, is actually cathodically protected by the steel. 7.2.4.1. Temperature. Temperature is a critical factor in the reversal of polarity. The critical temperature, i.e., that at which polarity reversal occurs, is determined by the solution composition [409, 4lO]. At a given composition, the tendency for potential ennoblement of the zinc electrode over time increases with increasing solution temperature as shown in Fig. 7.15. Gilbert [709] found that polarity reversal in hard hot water occurs when the solution temperature is over 60°C and disappears if the previously heated sample is cooled to below 60°C. In some solutions, polarity reversal can also occur at room temperature. For example, polarity reversal was found to occur in a 600 ppm bicarbonate solution at 28°C after 10 days [471].
CHAPTER 7
40°C
50 °C
60°C 70°C'------------2
5
4
3 Time,
6
hours
FIGURE 7.15. Effect of temperature on short-circuit current of zinc1steel couple. Solution composition: 70-80 ppm HCO;, 10 ppm SO~-, and 10 ppm NO;. From Hoxeng and Prutton [4IOJ. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
7.2.4.2. Solution Composition. Solution composition is one of the most important factors determining the polarity of a zinc/steel couple. In distilled water, polarity reversal of the zinc/steel couple does not always occur at high temperatures; this is due to the less protective nature of the corrosion products [709]. The presence of certain ionic species is required for the reversal to occur, as shown in Table 7.6.
-0.2,----------,-------------, potential increasing agents
potential decreasing agents
-0.4
~ ]~ -06 .
potential range of steel
E ~
&:
-0.8
:
1 I
-1
~ I I
130 HCO,-
I 10 so:5 NO,-
130 HCO,-
Solution Composition (in ppm)
40 HCO,5S0 4 2 10 NO,100 HCO,10 SO,'-
FIGURE 7.16. Effect of various chemical agents on the potential of zinc in different solutions.
205
CORROSION FORMS
As previously noted, the change in polarity occurs mainly through change of the zinc electrode potential, because the steel potential is relatively independent of factors such as solution composition. The zinc potentials in solutions containing various amounts of different species are summarized in Fig. 7.16, based on data reported by Hoxeng and Prutton [409,410]. A very small amount of chemical agents, at levels of a few parts per million, can cause the potential of zinc to change significantly. For a given solution composition, increasing the concentration of the species on the left-hand side of the figure, i.e., bicarbonate and nitrate, causes an increase in the potential of zinc, while increasing the concentration of the species on the right-hand side, i.e., chloride, sulfate, calcium, and silicate, decreases the potential. The effect of a particular agent on the potential of zinc depends strongly also on the presence of other agents. This may explain the very different responses of zinc/steel couples in different waters (Table 7.6), because natural waters can contain many different chemical agents at various concentrations. Hoxeng and Prutton [409,410] also investigated the effect of some chemical species in hot water in the presence of oxygen and found that sulfates and chlorides decrease the probability of reversal whereas bicarbonates and nitrates increase it. The addition of even small amounts (up to 20 ppm) of calcium salts or silicates can also decrease the probability of reversal in zinc/steel couples. In the absence of oxygen, the zinc is always anodic to the steel. There appears to be an interaction between ions with opposite tendencies to either promote or inhibit the polarity reversal in hot solutions. Kurr [471] noted that when gypsum was added to a solution at room temperature in which polarity reversal had occurred, the normal zinc potential was restored, as shown in Fig. 7.17. It was also observed in another study that the addition of 5 g of NaCI per liter immediately restores the original polarity for a zinc/steel couple in 0.05M NaHCO, solutions [457]. In a study by von Fraunhofer and Lubinski [457], it was found that polarity reversal occurs at an NaHCO,:NaCI concentration ratio of20: 1. Table 7.7 lists some of the solutions in which the tendency for polarity reversal has been investigated.
_OA
>
'/ I
~
Saturated CaS04 added as gypsum
~ -0.6
'"
:g (j)
J2
-0.8
u
c
N
-t
L __ _ _ _ _ _ _ _
o
~
20
_ _ _ _ _ _ _ _- L_ _ _ _ _ _ _ _ _ _L __ _ _ _ _ _ _ _
40
60
~
80
Time, days
FIGURE 7.17. Restoring effect of added gypsum on zinc potential adversely affected by bicarbonate-rich solution (600 ppm HCO'j. 73 ppm NO.1, 20 ppm CO.1 ) at room temperature. From Kurr [4711. © Copyright by NACE InternationaL All Rights Reserved by NACE; reprinted with permission.
206
CHAPTER 7
TABLE 7.7. Polarity of Zinc in Zinc/Steel Couple in Various Solutions
ot
Polarity
Reference
P
P
Cathodic
P
A
Anodic
P
P
Cathodic
P
P
Cathodic
P
Cathodic
P
Cathodic
457 457 194 410 471 449
Solution
T(°C)
pH
CO 2"
0.05M NaHC03
50-75 50-75 65 50-70
8.5 8.5 8.1 8.5
0.05M NaHC0 3
O.OIM NaHC03
40 ppm HCO; + 5 ppm NO; 600 ppm HCO; + 73 ppm NO; 140 ppm HCO; + 10 ppm NO;- + 10 ppm SO~-
Room
50-75
8.2
P
"P, present: A. absent.
7.2.4.3. pH. The pH of waters or solutions in which reversal occurs is generally slightly basic, between 7 and 9, as can be seen in Tables 7.6 and 7.7. Hoxeng [409] evaluated the effect of pH on the zinc potential in solutions containing bicarbonate ions and found that at about 60°C the most noble potential is reached in the pH range between 7 and 8. This pH range is partially determined by the presence of carbonate and bicarbonate, as they have a buffering effect. Also, as Glass and Ashworth [194] pointed out, raising the solution temperature of a solution that contains HCO;- will result in an increase in pH due to the loss of CO 2 to the atmosphere according to the reaction
7.2.4.4. Time of Immersion. The length of time required for polarity reversal varies greatly with the conditions. The time for polarity reversal generally decreases with increasing temperature. In a solution containing 600 ppm bicarbonate, the zinc potential reaches -0.7 VSCE within 1 h at 80°C, whereas at 28°C it take 280 h to reach that value [471]. In distilled water, it was found that with daily heating to 85°C for 8 h, polarity reversal occurs at about 3-4 days [709]. Von Fraunhofer and Lubinski [457] observed a relationship between the temperature (T) and time (t) for polarity reversal in 0.05M NaHC0 3 solution, as expressed in the following equation with a and b as constants: tT=bt+aT
Once the polarity is reversed, it may require a certain time to return to the normal polarity after the solution is cooled. According to Gilbert [709], it takes a few minutes for the zinc to become the anode again after the previously ennobled zinc sample, which has been immersed in hot water for a few days, is cooled to below 60°C. This process can take a few hours if the sample has been immersed in hot water for a period of weeks. When the zinc specimen is immersed for a long time, e.g., 6 months, the zinc can remain cathodic to the steel for as long as 44 days after the solution is cooled. Shulerner and Lehrman [925] found that for solutions of the same composition, the reversal potential is achieved at a faster rate when the ratio of the area of iron to that of zinc is increased.
CORROSION FORMS
207
7.2.4.5. Mechanism. The mechanisms of polarity reversal of a zinc/steel couple in hot water and very dilute solutions have been the subject of several investigations [194, 559,458, 449J. It is generally concluded that: I. Ennoblement of zinc only occurs in certain waters and solutions; it occurs readily in the presence of bicarbonate and less readily, or not at all, in the presence of chloride or sulfate. 2. The presence of oxygen is necessary for ennoblement. 3. For a given solution, the tendency for ennoblement increases with increasing temperature. The ennoblement of the zinc potential with increasing temperature can be simplistically explained according to Fig. 5.4 in Chapter 5. The anodic polarization curve changes from I" to 3" as a result of the surface passivation. Assuming that the kinetics of the cathodic reaction is little affected, then the corrosion potential will increase from E~orr to E~orr' Most investigators have attributed the ennoblement of zinc in hot waters to surface passivation through the formation of a solid corrosion product film. Potential ennoblement is usually a characteristic of surface passivation. The increasing potential with temperature indicates a transition from an active state to a passive state, due to the formation or transformation of hydroxide, oxide, or salt films on the surface. Very different types of corrosion products can form in supply waters depending on the type and amount of impurities. Gilbert [458] found that different amounts of zinc oxide and hydroxides can be formed depending on whether the zinc is in distilled water or supply water, hot or cold, or whether the water contains CO 2 • In cold water, zinc hydroxide is usually the corrosion product, sometimes accompanied by zinc oxide. The corrosion product at 85°C is usually zinc oxide, sometimes accompanied by some form of hydroxide. The fact that potential reversal is faster when the area ratio of iron to zinc is increased [12381 indicates that the formation of the passive film needed for this change is related to the cathodic reactions on the steel surface. Before potential reversal occurs, the passive film formation on the zinc surface requires an anodic current that equals the cathodic current on the steel. The larger the steel surface is, the larger the cathodic current and the shorter the time for the formation of such a passive film. Cathodic depolarization has been proposed by some authors as a cause of the ennoblement of zinc in hot water [709, 458, 449]. Gilbert [458] argued that the change of the surface state from active to passive is not the main cause of the ennoblement, and, instead, cathodic depolarization must be mainly responsible for the ennoblement. The cathodic depolarization at higher temperatures is attributed to an increase in the electronic conductivity of the semiconducting oxide film. Changes in solution composition lead to a change in the composition and/or the physical form of the oxide and, therefore, determine the extent of ennoblement. Trabanelli et al. [4491 characterized the semiconductor property of the corrosion products by measuring the photo-potential response of the zinc electrode in bicarbonate solutions. They found that the surface film formed at room temperature had p-type characteristics. It was only at temperatures above 50°C, as polarity reversal occurred, that the film on zinc showed n-type characteristics, which facilitate the cathodic reactions. Also, examination of the corrosion products by X-ray
208
CHAPTER 7
diffraction indicated that the film was ZnO at high temperature while it was Zn5(OHMC03)2 at room temperature. On the other hand, polarity reversal is not necessarily associated with a cathodic depolarization because the corrosion product films may have poor conductivity. According to Glass and Ashworth [194], the formation of zinc corrosion products (a very thin, transparent/white film and predominantly basic zinc carbonate) in O.OIM NaHC0 3 actually increases polarization. Also, it is very likely that not all of the passive corrosion products formed in different solutions are semiconducting. It seems that if the formation or transformation of a passive film contributes to ennoblement, the effect of electronic conductivity is only secondary to the effect of passivation, because without passivation any significant departure of the potential of the zinc from its reversible electrode potential must be accompanied by a severe dissolution. On the other hand, only a minimal level of conductivity, whether it is semiconducting, ionic, or via defects, is required for a potential ennoblement to occur once the surface is passivated. Thus, although semiconducting behavior of some corrosion products seems to be associated with the ennoblement of zinc in hot solutions, it is most likely the result, rather than the cause, of the ennoblement. The main role of oxygen in the ennoblement of zinc in hot waters is deemed to be related to cathodic depolarization. In the absence of oxygen, there are not enough oxidizing agents in water for the passivated zinc surface to achieve more noble potentials. A similar role is believed to be played by nitrate and some other oxidizing agents. According to Evans and Davies [401,402], one important effect of the oxygen reduction is that relatively high pH is developed at the local reduction areas, which then regulates the formation of corrosion products, different physical forms of hydroxides and oxides being formed at different pH values [404]. 7.2.4.6. Effect of Reversal on the Steel Corrosion Rate. When polarity reversal occurs, the zinc surface is passivated, and the zinc ceases to provide sacrificial protection to the coupled steel. The effect of this reversal on the corrosion of the steel depends on the type of environment and the arrangement of the zinc/steel couple. If a bulk piece of zinc is used as a sacrificial anode and coupled to a steel article, the fact that the zinc surface became the cathode may not alter the general corrosion rate of the steel very much compared to uncoupled steel. This is because the zinc surface area in such a couple is usually small compared to that of the steel. However, if the steel surface is also in a passive state and the passive film has weak points, the corrosion at these points can be accelerated as a result of the polarity reversal. When zinc is used as a coating, polarity reversal will result in a large zinc cathode: steel anode area ratio at any coating discontinuity. If passivation of the active steel does not occur, much faster penetration rates at the areas of discontinuities can be expected; this is the main cause of premature failure of some hot water tanks.
7.2.5. Galvanic Corrosion in Natural Environments 7.2.5.1. Atmospheric Environments. Field exposure data are valuable for a realistic evaluation of the relative severity of galvanic corrosion. Compared to other types of corrosion, galvanic corrosion in the field has not been well investigated. This is probably due to the more complicated situation; in addition to all the factors that may affect the normal corrosion of a metal, other factors, such as the kind of cathodic materials, the size
CORROSION FORMS
209
of the electrodes, anode and cathode arrangement, etc., are also involved in a galvanic corrosion system. In addition, this complexity makes application of the field corrosion data limited because in a real situation it is seldom, or very rare, that the whole arrangement of material, dimensional, and geometric factors, plus the environmental factors, is closely similar to that of an earlier field test. A test program on galvanic corrosion under atmospheric conditions was started as early as 1931 by the American Society for Testing and Materials (ASTM) [293]. Since then, a number of extensive exposure programs, most of which took zinc as one of the metals for the galvanic corrosion couples, have been carried out all over the world [336. 515,545,620]. In general, galvanic corrosion under atmospheric conditions is evaluated by weight loss measurements. In assessing galvanic corrosion in various other environments, such as in water or soil, the potentials and/or the galvanic current of the two coupled metals can be measured, but it is very difficult to measure the potentials of the metals under atmospheric conditions. For weight loss measurements, two types of assemblies have been mostly used: plate type and wire-on-bolt type [293]. In the plate type of assembly, a strip of one metal IS attached by bolts to a panel of another metal. The bolts are insulated from the strip and panel. The galvanic corrosion is evaluated by visual examination or by weight loss measurement for the strip or panel. In the wire-on-bolt type of assembly, a wire of the metal to be tested is tightly wound in the threads of a bolt of the other metal in the couple. The galvanic corrosion is quantitatively estimated by comparing the weight loss of the coupled wire to that of the same wire wound on the threads of a plastic bolt. Galvanic corrosion of galvanized steel occurs at areas where the coating is damaged and the steel underneath is exposed, such as at cuts or at scratches. At these areas, the exposed steel is cathodically protected while the surrounding zinc coating is galvanically corroded. However, in most cases, for galvanized steel the amount of coating loss due to galvanic corrosion, compared to the loss due to normal corrosion, is small because the exposed areas of bare steel are usually too small to cause significant corrosion of the relatively much larger zinc surface area. As a result, the atmospheric corrosion rate, including the contributions of both galvanic and normal corrosion, of galvanized zinc coatings is usually very similar to that of uncoupled zinc. Galvanic corrosion can, however, be a significant contributor to the total atmospheric corrosion of zinc when it is connected to other metals of similar size. Data in Table 7.8, reported by Kucera and Mattsson [293], show the galvanic corrosion rate of zinc wires when coupled to bolts of various metals in different atmospheric environments. Depending on the connected metal and the type of atmosphere. the galvanic corrosion can be as much as five times the normal corrosion of zinc in a rural atmosphere and three times that in a marine atmosphere. It can be seen from Table 7.8 that the amount of corrosion is not directly related to the difference between the reversible potentials of zinc and the coupled metal. Among the metals, mild steel acts as the most efficient cathodic material, largely owing to the voluminous rust, which can absorb pollutants and retain moisture and thus give rise to an aggressive electrolyte of good conductivity. Table 7.9 shows the galvanic corrosion of zinc and iron in four different atmospheric environments, assessed using metal disks clamped together with insulating washers [1229]. In this galvanic cell, the corrosion rate of zinc disk samples is increased by a factor of l.7-3.7. For zinc/steel couples, the galvanic corrosion of zinc is generally insignificant
210
CHAPTER 7
TABLE 7.8. Galvanic Corrosion Rates of Zinc Coupled to Other Common Commercial Metals in Various Atmospheric EnvironmentsO (/1mJyr) Galvanic corrosion rate (Ilm/yr) Coupled alloy
Rural
Urban
Zinc freely exposed Mild steel Stainless steel Copper Lead Nickel Aluminum Anodized aluminum Tin Chromium Magnesium
0.5 3.0
2.4 3.3 1.8 2.0 2.4 1.9
l.l
2.2 1.6 1.5 0.4 0.9 1.0 0.7 0.02
Marine
1.3 3.9 2.0 3.2 3.4 2.8 1.1 1.0 2.4 1.9
l.l
1.9 2.6 1.4 0.04
l.l
"Data from Ref. 293.
compared to the decrease in the corrosion of steel resulting from the galvanic action. Also, galvanic protection of the steel is more effective in industrial and marine atmospheres than in rural ones. This suggests that the pollutants in the atmosphere are beneficial to the galvanic protection of steel, although they are very harmful to the normal corrosion of the uncoupled steel. Compton and Mendizza [545] showed that the extent of galvanic corrosion of zinc coupled to different metal alloys does not vary much, even though there are wide differences in the reversible potentials among the alloys. They suggested that, under atmospheric conditions, other factors, such as corrosion products on zinc and the other metals, are more important in controlling the galvanic corrosion of zinc than the differences in the metal potentials.
TABLE
7.9. Corrosion of Galvanic Couples in Various Atmospheric Environments after Seven Years' Exposureo.h
Couple ZnlZn ZnlPb ZnlCu ZnlAI Zn/Fe Fe/Fe Fe/Zn
Rd
W' 187 313 292 362 332 1825 43
Industrial, marine
Rural
Industrial
1.7 1.6 1.9 1.8 1 40
W
27 47 48 100 81 470 147
R
1.7
1.8 3.7 3.0
1.
W
195 328 338 440 349 1534 5
Industrial, humid
W
R
1.7 1.7
2.3 1.8 300
43 83 100 141 127 1406 6
R
1.9 2.3 3.3 3.0 230
"Ref. 1229. "Samples consisted of two 1.5-in.-diameter disks, in. in thickness, clamped together with l-in.-diameter Bakelite washers, giving an exposed area of in. all round the edge of the disk and an annular area, ~ in. deep, of 1.275 sq. in. 'Weight loss (in milligrams) of the first metal in a couple, e.g., Zn in Zn/AI. "Ratio of the weight loss of metal M in the couple MIM' (Metal alMetal b) to that measured for M/M.
Tt;
Tt;
CORROSION FORMS
211
Zinc is usually anodic to other metal alloys in atmospheric environments; however, aluminum in urban and marine atmospheres and magnesium in all atmospheres are usually anodic to zinc, and hence their connection to zinc will reduce the corrosion of zinc [293, 544, 551]. It is shown in Table 7.8 that the amount of zinc corrosion is smaller for zinc connected to aluminum than that for a free-standing zinc sample, indicating a galvanic protection of zinc by aluminum. Owing to the formation of a passive film, however, aluminum is cathodic to zinc in many environments. Doyle and Wright [515] have reported that aluminum, when tested with wire on a zinc bolt, is cathodic to the zinc in most industrial atmospheres and some of the marine atmospheres, but the resulting galvanic corrosion of zinc is usually very small. Galvanic action is most significant in marine atmospheres because of the high conductivity of seawater, as shown in Table 7.8. In a marine atmosphere, the galvanic corrosion rate of zinc is found to increase at the beginning of the exposure and then remains at a relatively constant value afterward [620]. Rain, compared to other types of moisture, is particularly effective in causing galvanic corrosion. Table 7.10 shows that the galvanic corrosion rate is several times that of normal corrosion rates in an open exposure while the rates under a rain shelter are similar. This can be explained by the fact that the electrolyte layer formed by rain is thicker and has a smaller lateral electric resistance than the moisture formed by condensation. 7.2.5.2. Soil Environments. Like the atmosphere, soil is a complex medium. There are many sources of variability in soils that can affect the electrochemical behavior of metal alloys and, therefore, the galvanic actions between the alloys. The zinc electrode potential can vary by hundreds of millivolts depending on the type of soil [357, 1239]. Thus, the galvanic series measured in soils often do not follow the emf series [11251. Table 7.11, contained in a study by the National Bureau of Standards (NBS), shows the annually averaged galvanic corrosion rates of zinc coupled to steel with different anode!cathode surface area ratios in different soils [357, 1239]. Figure 7.18 illustrates the potentials of zinc/steel couples in three different soils, recorded over a period of five years. The cathode material was a 10-in. steel ring made of a 0.5-in.-diameter rod. Zinc anodes of different surface areas were located I in. below the steel. It was found that the amount of galvanic corrosion of zinc generally increased with decreasing soil resistivity. However, the degree of galvanic protection for the steel was lower in a soil of higher resistivity. Table 7.1 J shows that, although the total corrosion increases slightly with increasing zinc surface area, the corrosion density decreases fairly significantly, along with a significant
TABLE 7. 10. Galvanic Corrosion Rates of a Zinc Wire on a Steel Bolt after One Year of Testing with and without Rain Sheltera Rural R"
( Jlm/yr) ( Ilffi/yr)
Zinc/iron Zinc freely exposed
0.5 0.4
Urban
OC 3.0 0.5
"Ref. 293. "Galvanic corrosion rate with rain shelter. 'Galvanic corrosion rate without rain shelter.
RIO 0.17 0.8
R
0
(Ilmlyr) ( Jlmlyr)
1.3 1.3
3.3 2.4
Marine RIO 0.39 0.54
R
()
(Ilm/yr) (Jlm/yr)
0.4 0.5
3.9 1.3
RIO 0.1 0.38
CHAPTER 7
212
TABLE 7.11. Galvanic Corrosion of Zinc1Steel Couple in Soilsa Galvanic corrosion (glyr)
Soil characteristics Location
pH
R (,Q·cm)
Zinc:steel area ratiob
Lousville, Miss.
4.3
9390
West Austintown, Calif.
7.1
2582
0/20 1120 2120 3/20 0120
821
2120 3/20 0120
1120
Latex, Tex.
4.5
1120
2120 3/20
Cathode
Anode
10.1 8.3 5.1 4.94 11.2 2.58 1.48 1.45 21.4 0.57 0.19 0.45
0.12 1.55 3.36 6.1 0.1 4.9 7.38 7.79 0.25 13.7 20.7 20.3
"Data from Ref. 357. hSteel area = 2100 cm2 •
reduction in the corrosion of the steel. Escalante [517] also found that a linear relationship existed between the galvanic current and the resistivity of soils for a zinc/stainless steel couple separated 30 cm from each other at a depth of 0.8 m, There is a tendency for the galvanic current to decrease with the time of exposure. This is attributed to the formation of anodic and/or cathodic reaction products that have the effect of hampering the electrochemical reactions. In some soils, a protective cathodic film, which inhibits the cathodic reactions, can be formed on steel that is galvanically coupled to zinc.
-0.2 r - - - - - - - - - - - - - - - - - - - : - - - 1 -Louisville
~
>'
oj -0.4
~
• West Austintown
+ Latex
.~ -0.6 '0 Q)
c.. :> 8 -O.B
-1.2 '---_ _-'-_ _ _.l...-_ _- ' -_ _ _.!..-_ _--'-_ _- - - '
o
2
3
4
5
6
Time, years
FIGURE 7.18. Variation electrode potentials of zinc/steel couples in three soils over five years. Data are taken from Ref. 357. See Table 7.11 for soil characteristics.
CORROSION FORMS
213
Galvanic corrosion of zinc in soil is also found to occur when zinc is connected to a nonmetallic conductive material. Schick [518] reported that in underground telephone cable plants, rust formed on the galvanized steel used to support nonmetallic conductive material hardware that was electrically connected to rebar in concrete. Also, increased corrosion of galvanized steel posts was observed when they were connected to a carbon-black-filled, polyethylene-jacketed power cable.
7.2.6. Galvanic Protection of Steel by Zinc The galvanic corrosion of zinc generally results in galvanic protection of the coupled alloy. This property of zinc has been used in many applications, especially for the protection of steel. Coating steel with zinc is one of the most common ways to prevent the steel from corroding in natural environments. The steel is protected by the zinc coating through a barrier effect and a galvanic effect, in which zinc acts as the sacrificial anode while steel acts as the cathode. Besides galvanizing, zinc is also widely used cathodically as a bulk sacrificial anode material for cathodic protection of steel structures. The principles of protection of steel structure through the use of sacrificial zinc anodes are in essence the same as those for protection through impressed current by a rectifier. When a cathodic current is passed through steel, the potential of the steel will be changed to more negative potentials. When the potential is in the region in which iron is thermodynamically stable, the steel becomes inert. The amount of current required for cathodic protection depends on many conditions including all the factors illustrated in Fig. 7.1. The relations among polarization, electrolyte resistance, and cathodic protection of iron have been systematically studied by Holler [1125]. In most natural environments, zinc corrodes much less than steel, by a factor of 10-100 times [357,539]. The protection of steel by a zinc coating is, thus, mainly through the barrier effect. However, at the places where the zinc coating is removed or defective, leaving the steel exposed, the galvanic action between steel and zinc can protect the exposed steel from corrosion. The galvanic corrosion of galvanized steel is quite unique because in galvanized steel, unlike other galvanic couplings, the combination of materials and the geometry do not change much with different applications. The galvanic corrosion rate of zinc and, at the same time, the extent of galvanic protection for the steel can be determined based on dimensional and environmental factors. In atmospheric environments, galvanic action on galvanized steel depends on factors such as the concentration of the electrolyte, the thickness of the electrolyte, the dimension of the bare steel surface, and the distance between the zinc and steel. These factors have been systematically studied by Zhang and Valeriote [522, 1103] using a coplanarly coupled zinc/steel cell under thin-layer electrolytes (Fig. 7.7). Figure 7.19 shows that the protection area, on which the potential is below -700 mVSCE, proportionally increases with increasing width of the steel up to a certain value and then decreases with further increase in the steel width. It increases with increasing electrolyte thickness, although less sensitively than with changing of steel width. For a smaller steel width, the whole steel surface could be effectively under galvanic protection even when the zinc is at some distance. The protection distance (the distance between the zinc and steel electrodes when the potential of the steel is below -700 m VSCE) increases with increasing conductivity and thickness of the electrolyte and with decreasing area of the steel.
214
CHAPTER 7
X (mm) 20
0.001 M Na,sO,
X (mm) 20
12
4
t (mm)
FIGURE 7.19. Three-dimensional plots of protection width X (galvanic protection area =X x L. see Fig. 7.7) as a function of the electrolyte thickness (t), steel width (W), and the distance between the zinc and steel (D): (a) D = 0; (b) D = 5 mm. After Zhang and Veleriote [1104].
Atmospheric exposure testing of mild steel wire on zinc bolts indicated that galvanic action reduced the corrosion of the steel wire by a factor of 10-40, depending on the type of atmosphere [293]. Similar results can be seen in Table 7.9, which presents the findings of a seven-year exposure test with disks of the metals clamped together with insulating washers, exposing an annular area of each metal-ls in. wide. The galvanic action reduced the corrosion of steel by 40 times in industrial, 230 times in humid-industrial, and 300 times in seacoast industrial atmospheres. The reduction was only about threefold in rural atmospheres. The much lower galvanic effect on the corrosion of steel in rural atmospheres is largely due to the relatively nonconductive nature of the moisture. Table 7.9 also shows that the accelerating effect of galvanic action on the corrosion of zinc, a factor of 1.6-3, is generally insignificant compared with the reduction of steel corrosion. The galvanic protection of steel by zinc anodes in soils can be seen in Table 7.11 [357]. The extent of protection depends on the resistivity of the soils and the zinc:steel area ratio. In the soil with a resistivity of 821 Q·cm, the corrosion of steel is virtually stopped at a zinc/steel area ratio of 1120, but in the soil with a resistance of 9390 Q·cm, at an area ratio of 3/20 the corrosion of the steel is reduced only to half of that of the uncoupled steel. It can be noted in Table 7.8 that the reduction in the amount of steel corroded is accompanied by consumption of a similar amount of zinc owing to the galvanic corrosion. This is quite different from the galvanic effect in atmospheric environments, where the amount of galvanic corrosion of zinc is insignificant compared to the reduction in the corrosion of steel (Table 7.9). Zinc is a common material for making sacrificial anodes. Historically, zinc anodes have been mostly used in seawater-oriented applications [1193]. They are also used for cathodic protection in hot water tanks [447], fuel storage tanks [281], steel-reinforced concrete [392, 1232, 1284], and underground steel structures [471,473]. Anode composition, shape, size, and position can be tailored to specific applications [469, 1192, 1193]. More information on zinc anodes can be found in Chapter 15 (Section 15.4). When zinc is used as an anode material in an electrolyte of low resistivity, it has the advantages, compared to other anode materials such as aluminum and magnesium, of high efficiency and little hydrogen evolution [1247]. Self-corrosion due to hydrogen
CORROSION FORMS
215
evolution is significant for magnesium in solutions with pH below 12 and for aluminum in solutions with pH above 10 [I, 1247, 12431. Owing to the small self-corrosion rate in most natural environments, the zinc anode has a high galvanic efficiency, 95% to almost 100%, because zinc suspended in seawater or buried in the ground does not corrode rapidly by self-corrosion [1105]. Another advantage of zinc as an anode material is its generally low overpotential for anodic dissolution. Because the overpotential on the anode is low, most of the potential difference between zinc and steel is available to polarize the steel. In some situations, the smaller potential difference between zinc and steel compared to that between steel and aluminum or magnesium has the advantage of not causing overprotection, which could result in an excessive hydrogen evolution on the steel and a high anode consumption rate. When a zinc anode is employed in the cathodic protection of a steel ship hull in seawater, an empirical rule is to employ 1 unit of zinc anode area for 100 units of surface area of a painted steel hull or for 5 units of bare surface area [1105]. According to such a rule, the zinc anode consumption rate is about 0.1-0.2 Ib/yr per square foot of painted steel surface. The presence of Fe in a zinc anode, even in a very small amount, e.g., 0.00 I %, is harmful to the performance of the anode [230, 1194]. The presence of Fe causes a reduction of current output and ennoblement of the anode potential owing to the formation of an insulating dissolution product film on the surface of the anode. Addition of Al can reduce the effect of Fe in the zinc anode [470]. Steels are often protected by zinc-rich paints, in which the zinc dust serves as a pigment material and provides some galvanic protection to the painted steel. For zinc-rich coatings, three conditions must be satisfied in order for the galvanic process to occur [5, 1054J: I. The zinc particles in the coating must be in electrical contact with each other. 2. The zinc particles must be in electrical contact with the steel. 3. A continuous electrolyte must exist between the zinc particles and the steel. These conditions imply that the galvanic protection of steel by zinc-rich coatings improves with increasing amounts of zinc. Thus, high zinc contents, higher than 70%, are needed for good galvanic protection of steel [5, 700, 1054]. Galvanic action of the zinc dust is usually effective in the early stage; with time, the oxidized zinc particles causes a gradual loss of the electrical continuity between the dust particles in the interior of the paint and the steel. However, the transformation of the zinc particles from the metallic form to the oxide form exerts a sealing effect on the paint and adds more resistance to the permeation of aggressive agents from the environment [569, 700]. The recent search for ways of producing more corrosion-resistant automobiles has led to the development of many new zinc alloy coatings [351, 985, 1007, 1202]. By comparison with pure zinc coatings, these alloy coatings are in general more resistant to normal corrosion yet are still effective in providing enough galvanic protection to the coated steel. Table 7.12 lists some studies on the galvanic action of different zinc alloy coatings coupled to steel. Hayashi et ai. [869] measured the time-dependent galvanic current and potential of cold-rolled steel, coupled to Zn-Fe at various concentrations in differentially aerated solutions. They found that the initial galvanic current decreases with increasing iron content in the coating. The extent to which Zn-AI coatings provide galvanic protection
CHAPTER 7
216
TABLE 7.12. Studies on Zinc Alloys Coupled to Steel in Various Electrolytes Zinc alloy
Electrolyte
Zn-Co.Zn-Mn Zn-Fe 10% Mg, Ti, Cr, Fe, Ni, Cu Zn-rich paint 13%Ni Zn-Ni,Zn-Fe Zn-Al 5%AI,55%AI Painted alloy 5%AI,55%Al
Zn-Fe Zn-Ni
5% NaCl Supply water 5% NaCl O.IMNaCI IN NaCI 0.03MNaCl IN NaCl 5% NaCl Atmosphere, humid chamber 5% NaCl 0.6MNaCl
Measurement( s)o Egc.lg Egc,lg
19
E gc ' E-I curve Egc Egc,lg Egc
19 Area of rust . Egc,lg Egc,lg
Effect studied
Reference
Alloying Polarity reversal Edge protection Galvanic protection Coating stability
390 709 1241 965 43
Time of protection Galvanic protection Galvanic protection Galvanic protection
167 14 827 1237
Current transient Alloying
869 1223
"Egc ' couple potential; I" galvanic current.
depends on the Al concentration. Coatings with more than 60% Al behave like aluminum and provide little galvanic protection to the steel in atmospheric environments [248]. Zn-Ni coatings with Ni concentration below 14% provide galvanic protection to steel in 0.6N NaCl solution, and the polarity is reversed when the Ni concentration is higher than 18% [1223]. Suzuki et al. [1241] investigated the galvanic effect of different electroplated zinc alloy coatings on the edge protection of painted panels of different zinc-alloy-coated steels. The extent of galvanic corrosion at the cut edge of the painted steel subjected to a wet-dry cyclic test varied with the alloying elements. Figure 7.20 shows the effect of
12
Cu
•
10
'#.
0
~
'In
8 Ti
6
Ni
•
•
2
-0 Q)
a:
Fe
4
•
Zn
•
2
Mg. 0 -1.05
Cr
-1
-0.95
-0.9
Immersion potential,
-0.85
-0.8
VSCE
FIGURE 7.20. Correlation between immersion potential and red rust ratio on the cut edge of various zinc-alloy-coated steels. From Suzuki et al. [124IJ. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
CORROSION FORMS
217
alloying elements on the red rust ratio on the edge. The throwing power of galvanic protection is generally larger for coatings with more negative electrode potentials. In summary many applications of zinc and its alloys, whether as a coating, an anode, or a zinc-rich paint, involve galvanic corrosion, which is desirable for zinc, unlike many other metals, because it is required for protecting another metal, usually steel. Galvanic corrosion is complex because it is a function of many factors, including values of electrode potentials, number of reactions and their kinetics, metallurgical conditions, surface conditions, electrolyte properties, and geometric factors. Depending on the circumstances, some or all of the factors may playa role in the galvanic corrosion. Generally, the effect of the geometric factors on the galvanic action can, in many cases, be mathematically analyzed. On the other hand, the effect of the factors related to electrode surface condition and reaction kinetics in real situations can be very difficult to determine. For galvanized steel, unlike other galvanic couplings, the combination of materials and the geometry do not usually change from one situation to another, simplifying the analysis of the galvanic action. In real situations, each of the various factors needs to be considered in order to maximize the beneficial effect resulting from the galvanic corrosion of zinc. 7.3. PITTING CORROSION
7.3.1. Introduction Pitting corrosion is a form of localized attack that results in holes in a corroding metal. It is a serious type of corrosion: though the extent of the reaction may be small, the damage may be severe, particularly when the metal concerned is used as a container for fluid (e.g., a heat exchanger tube or underground service pipe). Generally, a pit may be described as a cavity or hole whose surface diameter is about the same as or less than the depth [524]. One characteristic feature is that pitting generally occurs on metals and alloys whose surface is in a passive state, with severe dissolution of the metal at localized points and relatively little dissolution of the rest of the exposed surface. Another feature is that it usually occurs in a medium that contains aggressive anions, such as chloride ions, which cause local breakdown of the passive surface. A third characteristic feature of pitting is the existence of a threshold potential below which pitting does not occur but above which pitting occurs [197]. The susceptibility of a metal or alloy to pitting can be evaluated by several methods: (I) determination of the characteristic pitting potential; (2) determination of a critical temperature of pitting; (3) measurement of pit density; (4) evaluation of the size and depth of the pits; and/or (5) determination of the critical concentration of aggressive ions.
7.3.2. Occurrence of Pitting 7.3.2.1. In Distilled Water. Pitting is a common form of corrosion of zinc in distilled water at room temperature, as was reported as early as 1919 by Bengough and Hudson [654]. According to their observations, if zinc panels are placed vertically in distilled water, corrosion pits form, often arranged in straight rows of unconnected pits as shown in Fig. 7.21. According to Evans and Davies [402], the pits are the result of the gravitational sinking of solid corrosion products, which lodge at points where they shield the underlying zinc from oxygen, facilitating the anodic attack at those points. Pitting
218
CHAPTER 7
o , I
r
,,
I
\
,
t
,
"
,.
\
~
•
, "
,
,
, J
~
,
I
\
1
A
I
FIGURE 7.21. Vertical arrangement of pits in zinc exposed to distilled water. After Bengough and Hudson [654].
does not occur without oxygen or under high oxygen pressure. The local depletion of oxygen was found to be necessary for pitting corrosion. Zinc samples moving in distilled water saturated with oxygen show no pitting. The effect of oxygen depletion on pitting formation was demonstrated by Evans and Davies [402] by covering the zinc surface with a polyethylene fiber to produce two parallel lines of pits as shown in Fig. 7.22. At the two crevices on each side of the contact line, the metal, being locally shielded from oxygen, become anodic to the main surface and corroded preferentially [402]. When distilled water contains very small amounts of dissolved salts, general corrosion, rather than pitting corrosion, occurs [401]. The corrosion is no longer confined to
-~~PJl Zinc
or trench
Pits~
o
Trench~ I
I
o
ij
Surface afler removal of fiber
FIGURE 7.22. Two lines of pits produced in zinc in distilled water by contact with polyethylene thread. After Evans and Davies [402]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
CORROSION FORMS
219
the initiation points and tends to spread outward over an arch-shaped area owing to the separation of the anode and cathode when salt is present in distilled water. This is also confirmed by another study [559] in which the pitting corrosion rate for zinc samples immersed in distilled water for 30 days was 22 f.1m/yr, while it was only 6.S f.1m/yr for samples immersed in water containing 0.002% NaCI. Kenworthy and Smith [4001 reported the effect of dissolved carbon dioxide on the formation of pitson zinc and galvanized steel in pure water at room temperature. The depth of pits was found to decrease, but the number was found to increase, with increasing amounts of dissolved carbon dioxide in the water. After 57 days of immersion, the pit depth on zinc samples was about SO f.1m at 0.6 ppm CO 2 and was only I f.1m at 36 ppm CO 2 ; in contrast, the total amount of corrosion was four times more at 36 ppm CO 2 than at 0.6 ppm. Compared to the average pit depth zinc samples, that on galvanized steel samples was much less, being about 10 f.1m at 0.6 ppm CO 2 compared to SO f.1m for zinc; however, the total corrosion rate was similar to that for pure zinc. For galvanized steel, as pits penetrate to the zinc-iron alloy layers, the rate of penetration decreases owing to the more noble nature of the alloy materials; this causes the pits to spread instead of penetrating further. 7.3.2.2. In Hard Water. Kenworthy and Smith [400] found that the pitting penetration in zinc in hard water is only a fraction of that in distilled water. This agrees with the general observation that the corrosion attack is more uniform when the water contains salts [401]. 7.3.2.3. In Hot Water. Pitting is a common form of corrosion for zinc in hot water. This can be a serious problem for galvanized steel hot water tanks [400,655,656, 709]. Gilbert [709] investigated the corrosion of zinc in hot water and found that the corrosion attack is highly localized and very deep. After five months of immersion in hard supply water, saturated with 95% air and l.5% (about 14 ppm) carbon dioxide at S5°C, the maximum pit depth was about 0.3 mm on zinc samples and about 0.15 mm on galvanized coatings. In hot soft water, pitting corrosion is likely to lead to rapid penetration of galvanized coatings. Because of the reversal of polarity for zinc/steel galvanic couples in hot water, the pitting can continue through the substrate steel. In hard water, the corrosion is likely to be stined by the deposition of a protective scale, depending on the heating method. The presence of copper in the water was found to enhance the pitting corrosion of galvanized coatings in hot water [257]. The extent of pitting in hot water also varies with the content of dissolved carbon dioxide. Kenworthy and Smith [4001 reported that the depth of pitting is greatest with 5.6 ppm carbon dioxide and is least with 10 ppm carbon dioxide. The pits formed on solid zinc are, in general, much deeper than those formed on galvanized coatings. As a result of pitting, the corrosion failure of galvanized steel in hot water tanks may occur even when SO% of the coating still remains. 7.3.2.4. In Solutions. The occurrence of pitting corrosion on zinc in solutions depends on the pH and composition. In general, pitting corrosion is not likely to occur in acidic solutions and is most often found in slightly basic or basic solutions. Lorking and Mayne [334] reported that solutions prepared from distilled water with additions of NaOH, at pH values from 5.8 to II, can cause pitting on zinc. Figure 7.23 shows that the pitting rate increases with NaCI concentration from about 20 j.1m/yr at NaCi concentra-
220
CHAPTER 7
200,---------------------------------------, - Average rate
+ Maximum pit depth/ year ...: 150
<:: E
Ol.
ai
1§ 100 c
o
'Vi
e
o
o
50
O~
____ 0.001
0.0001
_L~
____
~
0.01
____
~
_ _ _ _ _ _L __ _ _ __ L_ _
0.1
~
10
Concentration of NaCI, %
FIGURE 7.23. Pitting corrosion rate of zinc in NaCI solutions at room temperature as a function of solution concentration. Data are taken from Ref. 559.
tions of a few parts per million to about 160 f.1mJyr at 3.5% [559]. The pitting rates in the solutions are much faster at higher temperatures. Pit density generally increases with increasing halide ion concentration [45, 355,402]. The number of pits and pit size greatly depend on the initial surface condition [355]. 7.3.2.5. In Atmospheric Environments. Pitting corrosion in atmospheric environments has been seldom reported as the main cause offailure of zinc products. In an ASTM atmospheric exposure program, pits were observed on high-grade zinc and on a 1% Cu-zinc alloy after two years of exposure in four different environments [294,549]. The pits were shallow, their depths generally being smaller than their diameters. Most of them were more like dimples than pits by definition. The depth of the pits increased with time, but the ratio of pit depth to the surface-averaged corrosion penetration decreased with time. Williams et al. [235] reported that pitting corrosion occurs on Zn-22Al alloy at 100% relative humidity and 50°e. Pitting was found to be more severe when sulfur dioxide was also present. Pit depth increased from an average of 30 f.1m without sulfur dioxide to about 100 f.1m with 100 ppm sulfur dioxide after 1000 hours of exposure. External stress did not noticeably change the pit depth. 7.3.2.6. In Soils. Corrosion of zinc in soils occurs unevenly over the surface, as soil is a very inhomogeneous environment. This unevenness may lead to the development of pitting, the extent of which varies significantly depending on the soil chemical composition and texture. In a study by the U.S. National Bureau of Standards, it was found that zinc pits in most soils, and the maximum pit depth is, in general, five or more times the average corrosion penetration [357]. 7.3.2.7. In Zinc Batteries. In zinc cans used in zinc batteries, pitting often occurs along lines of particular stress introduced during can formation [1145]. In inadequately sealed cells, deep pits tends to form at the zinc/electrolyte/air boundary. The occurrence of pitting in Leclanche types of electrolytes is a function of many factors, particularly the presence of impurities such as Pb 2+ and Cd 2+ in the electrolytes [884, 885]. An increase
221
CORROSION FORMS
in ZnCI 2 concentration favors a more isolated pit formation. A trend from general metallographic etching at low concentrations of ZnCI 2 «0.0 1M) to severe pitting at high concentrations (> I.OM) is found. The formation of pits on zinc cans is very detrimental since perforation of the zinc may develop very rapidly, leading to electrolyte extrusion and/or drying out of the battery [1145].
7.3.3. Pitting Potential As described in Chapter 3, the potential at which the passivation begins to break down at localized areas, leading to pitting, is defined as the pitting potential. Pitting potential can be determined from an anodic polarization curve as the potential at which the current begins to sharply rise with increasing potential. Figure 7.24, as an example, shows the potentiodynamic polarization curves for zinc in KOH solution as a function of chloride concentration in the solution [46]. The pitting potential, at which current sharply rises, decreases with increasing KCl concentrations. Table 7.13 shows the pitting potentials measured in various solutions. The value of pitting potential can be used as an indication of pitting tendency of a metal in an electrolyte. Generally, the more positive the pitting potential, the more difficult it is for pitting to occur at the rest potential. The values of the pitting potential can be affected by many factors such as pH, solution composition, and temperature. Solution composition (i.e., the nature and concentration of dissolved chemical species) has been found to be a particularly critical factor in determining the value of the pitting potential. Several investigations have shown that pitting potential becomes more negative with increasing chloride concentration, as shown in Figs. 7.24 and 7.25 [46,355]. According to Keitelman et al. [652], a linear relation exists between the pitting potential and the 5
4
NE <)
3
=<:t
.-
0()
0
...J
FIGURE 7.24. Potentiodynamic anodic polarization curves for Zn electrodes in 10- 2M KOH solutions containing various concentrations of KCI. After Ahd El Haleem [46].
2
o
0.8 E, Vsee
1.6
222
CHAPTER 7
TABLE 7.13. Pitting Potentials Measured from Potentiodynamic Polarization Curves in Various Solutions Solution
pH
0.02M NaCI04 + 0.2M H 2B0 3
-0.85 -0.73 -0.76 -0.9-0.95 +1.1 -0.7 0.6 -0.66 -0.7 -0.8 -0.4 +1 0.1
8 9 9 9 9.2 9.2 10 II 11 II 11.5 12 12
1M NaN0 3a O.IMNaCl a 105 ppm NaCI + borate-NaOH buffer 210 ppm NaCI + borate-NaOH buffer 1M Na2S04 + 0.2M Na2HPO/ 0.02M NaCI04 + 0.2M H 2B0 3 O.OOIMKOH" O.OOIM KOH + 5 x 1O- 4M KCI" 3M NaCI" O.OIMNaOH + 0.00IMNaC1" O.OIM KOH + O.OOIM KCI"
Reference
Epit (V SCE )
16 16 45 45 355 355 701 16 46 46 3 37 46
aDeaerated.
logarithmic concentration of Cl- in borate-buffered sodium chloride solutions. Abd EI Haleem [46] also showed that the pitting potential of zinc in D.DIM KOH decreases linearly with logarithmic concentration of CI-; similar relationships were observed for Br- and r- as shown in Fig. 7.26. Many other ionic species have been identified as pitting-enhancing agents. Augustynski et al. [16] found that, in addition to Cl-, Br- and r-, CIO:;, SO~-, NO), CIO), F", and CH}CO;: added to NaOH-H3BOr buffered solutions reduce passivation breakdown potential. The presence of certain anions in the solution, however, can inhibit the pitting corrosion of zinc. Figure 7.27, reported by Abd EI Haleem [37], shows the effect of three anions that
2 • pH 9.2 .., pH 11 w
:;l
>
cij ~ Q)
0 a. 0
_1L-------L-------L-------L-------L-----~
o
200
100
300
400
500
Chloride concentration, ppm
FIGURE 7.25. Pitting potential of zinc and Lotlikar [355].
VS.
chloride concentration in borate-buffered solutions. After Davies
CORROSION FORMS
223
1.2
• CIT Br-
0.8
- I-
0.4
"' M
>
.
0
w
-0.4 -0.8 -1.2
-4
-3 .5
-3
-2.5
-2
log C
FIGURE 7.26. Variation of the critical pitting potential, £1" of zinc with the concentration of cr. Br-, and lions in 1O- 2M KOH solutions. After Abd EI Haleem [46).
enhance pitting resistance. In this work, the pitting potential was found to increase with the concentration of the anions in the solution in the order PO;- < crO~- < C05-. In another report, the inhibiting effect was reported to be in the order WO~- < PO~- < crO~- [1091. All these anions were found to promote the formation of a solid oxide or salt film on the zinc surface, which results in passivation [3, 57, 65, 93, 98, 127]. The amount of passivating agents required for pitting inhibition depends on the concentration of the aggressive species. It has been reported that in concentrated NaCI solutions. although the total corrosion is reduced with addition of chromate up to 4 gil, the pitting rate is little affected compared to that without chromate addition [559].
2~--------------------------------~
CO,"
1.6 1 .2
>~ 0 .8 ~
t
w
_ _- -. CrO/"
0.4
o -0.4
-0·~2L.5------' _2------1-L.-5------.1..1-------!0.5
FIGURE 7.27. Variation of Epitting for Zn-Ti alloy with the concentration ofC05-, CrO~-, and HPO~- in O.OIM NaOH + O.OOSM NaCl solutions. After Abd El Haleem [37).
CHAPTER 7
224
It is important to note that the pitting occurring at the pitting potential is very different from that occurring at the corrosion potential (the rest potential). The pitting potential can be an indication of the pitting tendency in a particular environment but, by itself, does not provide a basis for predicting whether pitting actually does occur at the corrosion potential. This is because pitting potential is measured with an enforced external anodic current. In real applications, zinc products, except for zinc anodes, are normally used under a condition of no enforced current, i.e., at the corrosion potential. However a comparison between corrosion potential and pitting potential can provide a practical indication about the tendency for pitting corrosion at the corrosion potential. Pitting is less likely to occur when the corrosion potential of the metal is well below the pitting potential than when the corrosion potential is close to the pitting potential.
7.3.4. Morphology The morphology of pits can be characterized by the number, shape, size, and distribution of pits and corrosion products. Evans and Davies [402] investigated the pit arrangement and characteristics of corrosion products on zinc sheets in distilled water. The distribution and arrangement were found to be different for horizontally and vertically placed samples. For the vertically placed sample, the pits were found to arrange in straight rows and disconnect (Fig. 7.21) owing to the oxygen-shielding effect of the corrosion product deposited from an upper pit. The pits were filled and were generally surrounded by rings of the white corrosion product. Outside the white rings, the zinc was unattacked but showed interference colors by specular reflection. The arrangement of the colors showed the film to be thickest close to the pits. When the rings of white matter were brushed off, the film was found to continue below them as shown in Fig. 7.28. The white product around the pits was zinc oxide with some p-zinc hydroxide. The white matter obtained from within the pits gave no lines in X-ray diffraction patterns, suggesting that it was amorphous. The film giving the interference colors was identified to be zinc oxide. Pits on galvanized hot water tanks were found to be usually formed under gas bubbles that adhered to the surface during immersion of the zinc articles so that bubble cups (at the base of the gas bubbles) of corrosion products were formed around the pits [400,709]. Negatively charged colloid particles Ring of white matter
Ring of white matter
Color film
Color film Bare ring
------------~--~/
r~~----------------
Zinc Pit
. Loose white matter
FIGURE 7.28. Schematic illustration of the morphology around a pit. After Evans and Davies [402]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
CORROSION FORMS
225
These bubble cups consisted of white zinc corrosion products, and when failure of the coating occurred, it was always beneath these white products. Very often, the bubbles were removed when the water was changed, but fresh bubbles tended to appear in the same places as before. The center of the cups on the zinc sample had the same color as the surrounding metal, while the center of the bubble cups on galvanized steel had a darker red coloration than the surrounding metal, indicating the dissolution of the Zn-Fe alloy layers in the galvanized coating. In Leclanche battery electrolytes, pitting on zinc has the following characteristics, according to Baugh et al. [884, 885]. (a) Pits appear to be distributed indiscriminately across the surface of the zinc electrode with no preferred tendency to form on or near the grain boundaries. (b) The sites of initiation seem to be at the points of emergent dislocations. (c) Similar pitting characteristics were observed on cleaved and polished single-crystal electrodes, indicating that the effect of surface preparation on the pitting is of secondary importance. (d) The number of pits formed under the open-circuit condition is smaller than that formed under anodically polarized conditions. (e) Almost all the pits observed exhibited radii greater than their depths. The pits formed on zinc often have certain crystallographic features. Both rectangular and hexagonal pits can develop [29,45]. Alvarez and Galvele [45] observed that, after passivity breakdown, zinc dissolution does not propagate in a flat front, but in certain crystallographic planes. Pitting in O.IM NaCl solution, at a pH of9, was found to develop along {OOOI} planes as well as along {IOID} planes. Baugh et al. [885] found that the pits formed on single-crystal zinc electrodes with the basal plane exposed, at slightly anodic potentials in Leclanche types of electrolytes, were hexagonal, and the regularity of the hexagonal shape depended strongly on the composition of the electrolytes. 7.3.5. Mechanisms The detailed processes in pitting corrosion are complicated. Whether pitting corrosion occurs depends on many factors. For a metal in solution, it can generally be said that (1) physical imperfections on a metal surface determine the initial points of attack and (2) chemical factors in the solution determine whether the attack will occur, be healed through repassivation, remain localized and develop into a pit, or spread out and lead to a general attack. The mechanism involved in the pitting on zinc in cold distilled water is in essence different from that in hot water containing trace amounts of salts. In distilled water the zinc surface is initially not passivated. Because of the high resistivity of the water, the electrochemical reactions are highly localized, causing the formation of pits. The areas surrounding these pits then become passivated due to the rise of the local pH as a result of the cathodic reactions [402]. The localized passivation is a result of pitting. In hot water, on the other hand, the whole surface is generally passivated with or without the formation of pits. Pitting occurs at the places where the passive film has undergone breakdown. According to Evans and Davies [333,401,402], when corrosion starts at imperfections, the dissolution of zinc results in an acidification inside the pits due to the formation of zinc hydroxides. When ionic transport is efficient, such as in impure water, the anodes and cathodes can be physically separated, and ionic transport will allow the acid at the anodes to spread out and dissolve the preexisting oxide film, hence leading to a general
226
CHAPTER 7
attack. In distilled water, the ionic conductivity is very poor, and hence the anodes and cathodes stay very close together. In this case, the acid at the anodes is neutralized by the base generated at the cathodes before it can spread out. While the corroding sites keep active, the surrounding areas remain unattacked until pits are developed. The purer the water, the fewer and smaller the pits are. Motion of the sample in water can promote ionic movement along the surface and, therefore, reduce the chance of pitting, as was confirmed by Evans [401]. In solutions, pitting of zinc only occurs on passivated surfaces [45,46,355]. To cause pitting, breakdown of the passive film must occur. The electric field (anodic potential) required for the breakdown depends on the presence of certain ionic species, particularly halide ions, in the solution. The role of halide ions has been postulated to involve specific adsorption on certain sites of the passive film, followed by penetration of the film under the influence of an electric field. As a result. the field required for passivation breakdown at these sites is lowered, and when an anodic potential is applied on the electrode and the corresponding field exceeds the critical value. film breakdown occurs, leading to pitting corrosion. When the concentration of chloride ions is high, the whole film can be weakened and breakdown can occur over the whole surface, leading to a more uniform corrosion. The theory oflocal acidification seems to be widely accepted for the pitting corrosion of zinc. To initiate and maintain an active pit. local acidification is considered to be necessary [45, 128. 402, 652]. This local acidification is a result of metal dissolution: (7.17)
(7.18) The degree of acidification required depends on the bulk solution pH, the amount of chloride ions. the electrode potential, and the presence or absence of inhibiting agents. According to this model. the concentration of chloride ions determines the extent of passivity breakdown; the pH regulates passivation and repassivation; and the electrode potential controls not only the extent of breakdown but also the rate of dissolution at the places where breakdown occurs. If the dissolution in a pit is not fast enough to maintain sufficient acidity to prevent repassivation, a new passive film will form and the pit can become stifled. The fewer aggressive ions and the more passivating agents in solution, the higher is the potential needed to give a sufficient dissolution rate to maintain the acidic condition. The pitting potential is considered to be the minimum potential at which such acidification can be maintained [45]. According to Galvele and co-workers [45, 652], the pitting potential of zinc in aqueous solutions is determined by the local acidification resulting from reactions on the electrode surface and consists of three components: . (7.19)
E;
where is the corrosion potential of the acidified pitlike solution, '7 is the anodic polarization necessary to draw enough current through the pit to maintain the local solution chemistry for an active pit, and t1¢J is the potential drop resulting from the resistance in the electrolyte. The
CORROSION FORMS
227
relative contribution of each component to the pIttmg potential, Ep;1' vanes with pH, concentration of buffering agents, and solution conductivity. The formation of pits on a zinc surface in the Leclanche type of solutions was concluded by Baugh et al. [884, 885] not to be due to the breakdown of passivity by aggressive ions since neither the CI- concentration nor pH does significantly changes the pitting characteristics. Instead, pitting was considered to be due to the presence in the electrolyte of small amounts of impurities, such as Pb 2+ and Cd"+ ions, emanating from ZnCI 2 • The deposition of Pb and Cd on the zinc surface forms an metallic film. This film allows dissolution of the zinc at a few unprotected sites, which leads to pitting. The different effects of various anions on the pitting characteristics are attributed to their specific adsorption on the zinc surface, which affects the number of nucleation sites and the rate of dissolution inside the active pits. 7.4. INTERGRANULAR CORROSION 7.4.1. Introduction
Intergranular corrosion is defined as the localized corrosion at or adjacent to grain boundaries, with relatively little corrosion of the grains [524]. It results, at best, in a loss of ductility and strength and, at worst, in a very rapid, complete destruction of the metal. In most cases, intergranular corrosion of metals is associated with the presence at the grain boundaries of a phase different from the base metal in its electrochemical behavior. The extent of intergranular corrosion is usually evaluated through metallographic examinations of the surface and a cross section of the corroded sample and quantified by the depth of penetration along the grain boundaries. It can also be assessed by measurements of strength and ductility. 7.4.2. Occurrence
Intergranular corrosion of zinc alloys was first reported in the first quarter of this century as a serious problem for die-cast alloys used in hot water and warm humid atmospheres [535]. It was later realized that the problem was associated with the presence of aluminum along with certain impurities such as lead and cadmium in the alloys. Although most reported cases have been related to die-cast zinc-aluminum alloys, intergranular corrosion has been found to occur also in some other zinc products such as zinc coatings under warm and humid conditions [723, 1289]. Intergranular corrosion of zinc-aluminum alloys has been observed to occur in different environments: in the atmosphere, in water, in solution, and in concrete. Defrancq [203] observed intergranular corrosion of lead-containing zinc alloys in warm domestic waters. Ahmed et al. [154] reported that severe intergranular attack occurred in a Zn-0.3AI-0.03Cd anode after two years of service in seawater. The same material was attacked intergranularly after 6 hours of anodic polarization in 70 a C seawater or after one month of immersion in 70 a C distilled water. Roberts [535] found that intergranular attack on a Zn-O.l Al alloy occurred in 150 a C dry water vapor. Mercille [220] observed intergranular corrosion of Zn-4.2AI and Zn-12AI alloys that were embedded inside concrete and exposed in a rural atmosphere. Intergranular corrosion was also found to occur under a cathodic current as well as under an anodic current [154,224].
CHAPTER 7
228
TABLE 7.14. Intergranular Penetration Rates of Some Zinc-Aluminum Alloys in Various Environments Percent Al
Environment
Zn purity
0.1 4 20 21.1
99.999% 99.999% 99.999%
0. 1 0.075 4 0.04 4.2
99.99% 99.999% 99.99% 99.99% 99.99%
95°C water vapor 95°C water vapor 95°C water vapor 100% relative humidity at 50°C 95°C tap water 95°C water vapor 95°C water vapor 90°C, 0.05M KCI Concrete
Duration of exposure
Rate (mmJday)
10 days 10 days 10 days 42 days
10 days 10 days 10 years
Reference
0.18 0.033 0.028 0.002
722 722 722 235
0.1 0.02 0.066 0.07 5 j1mJyr
49 535 535 48 220
Table 7.14 lists the intergranular corrosion rates of some zinc-aluminum alloys in various environments that have been reported in the literature. The penetration rate of intergranular corrosion depends on many factors, among which the composition and structure of the alloy are the most significant. The presence of very small amounts of lead, tin, and some other elements may greatly accelerate the intergranular corrosion rate of AI-containing zinc alloys. Figure 7.29 shows the depth of intergranular corrosion for three zinc-aluminum alloys in water vapor at 95°C as a funtion of time [722]. It is noted that the intergranular penetration rate can be several orders of magnitude larger than the rates of general corrosion or pitting corrosion in similar environments.
10 - 0 .1 % AI
E E C
--r
4 % AI
"* 20 % AI
0 .;:;
~
Q)
c
Q)
0.
'0 .r:
a.
0 .1
Q)
0
0.01
0
2
4
6
8
10
Exposure time, days FIGURE 7.29. Intergranularcorrosion of zinc-aluminum alloys (prepared with 99.999% purity Zn and 99.98% purity AI) in water vapor at 95°C. Data are taken from Ref. 722.
229
CORROSION FORMS
7.4.3. Metallurgical Effects 7.4.3.1. Alloying Elements. Zinc of high purity is not susceptible to intergranular corrosion [48, 722]. The presence of other elements, particularly aluminum, as alloying elements or impurities is necessary to cause the intergranular corrosion. Intergranular corrosion has also been found to occur in zinc alloys containing only lead or magnesium [203,535]. For Zn-AI alloys, intergranular corrosion is observed to occur in a concentration range between 0.03% and 50% Al [48,49, 225, 535]. Below 0.03%AI, intergranular corrosion does not occur. According to Devillers and Niessen [48,49], the presence of impurities is not required for the occurrence of intergranular corrosion on zinc-aluminum alloys. They found that intergranular corrosion occurred on a Zn-O.I % AI alloy prepared with 99.9999% purity Zn and 99.999% purity Al in the hot-rolled and as-cast condition in 95°C water. In addition, the penetration rate for this very high purity alloy was essentially the same as that for a Zn-O.l % Al alloy prepared with 99.99% purity Zn and 99.9% purity AI, indicating that the residual impurities in the lower purity alloy are not the primary cause of the intergranular corrosion of Zn-AI alloys. The intergranular corrosion of zinc-aluminum alloys is attributed to the preferential attack on the aluminum-rich phase at the grain boundaries. Since 0.03 wt. % aluminum is close to the solubility limit of aluminum in zinc at room temperature, in zinc alloys containing an aluminum concentration higher than 0.03%, the aluminum precipitates at the grain boundaries and is thus responsible for the increased corrosion rate at the grain boundaries [49]. Moreover, as reported by Devillers [48], in two-phase Zn-AI alloys, e.g., Zn-5% AI, preferential corrosion occurs not only at grain boundaries of the zinc 200r-----------------------------------------~
>:
{l150
E
.3, c
o
e 1ii
100
c
w
a.
"0
w
~
a:
50
OL----------L~
0.001
0.01
________L __ _ _ _ _ _ _ _
~
_ _ _ _ _ _~
0.1 AI content (wt%)
FIGURE 7.30. Intergranular corrosion rate of zinc-aluminum alloys (prepared with 99.99% purity Zn and 99.9% purity AI) in tap water containing 0.05M KCl at 90°C as a function of aluminum content. Data are taken from Ref. 48.
230
CHAPTER 7
matrix but also in the AI-rich phase. However, for two-phase alloys the penetration rate is much lower than that for single-phase alloys of low Al content [535]. Devillers [48] found that in 0.05M KCI solution at 90°C the penetration rate of intergranular corrosion for Zn-AI alloys increases with increasing concentration of aluminum until it reaches a maximum at 0.2% AI, and then it decreases with further increase in Al content as shown in Fig. 7.30. Similar results obtained in another study [535] are shown in Fig. 7.31, where the corrosion penetration increases, peaking at 0.27% AI, and then decreases to much lower values with further increases in the Al content. According to Devillers [48], aluminum's effect on reducing the grain size is at least partially responsible for the reduction of intergranular corrosion rates for alloys with Al content greater than 0.2%. A decrease in the size of grains results in an increase in the number of grain boundaries; hence, to reach the same depth of corrosion penetration, the total grain boundary path is likely to be longer in a fine-grain alloy than in a coarse one. Melton and Edington [225] found that the intergranular corrosion of alloys containing 40% and 50% Al is much less severe than that of alloys containing 22% AI. The intergranular corrosion rate of zinc-aluminum alloys can be greatly influenced by the presence of small amounts of other elements. More specifically, experimental results have indicated that Mg, Cu, Au, Ni, Pt, and Co are beneficial, whereas Sn, Tl, In, Pb, Bi, Hg, Cd, and Na are harmful [48,49,535,J. Mo, Zr, Ti, Ba, Si, Be, Te, Li, Sb, and Ag are found to have little effect. Tables 7.15 and 7.16 contain data reported by Devillers and Niessen [49] on the effects of some beneficial and harmful elements on the intergranular corrosion of Zn-O.I % Al alloy exposed to 95°C water vapor. According to these results, Mg is the most effective element in suppressing intergranular corrosion, whereas Sn has the most harmful effect. The beneficial elements and the harmful elements are found to counter each other's effects when both exist in the same alloy. It has been noted that all the harmful elements have very low solubilities in zinc, and when their solid solubility limit in zinc is exceeded, they will form precipitates in their elemental form at the grain boundaries of the zinc matrix. Another common characteristic of all these elements is that they have higher hydrogen overvoltages. 1.2.------------------, E 1
E
., g
.50.8
~ 0.6 Q)
...00.4 a.
-5
a. Q) 00.2 OL----~---~---~--~-~
0.001
0.01
0.1
AI concentration, wt%
10
FIGURE 7.31. Average depth of intergranular corrosion penetration of zinc-aluminum alloys (prepared with 99.999% purity Zn and 99.98% purity AI) in water vapor at 95°C for 10 days as a function of aluminum content. Data are taken from Ref. 535.
231
CORROSION FORMS
TABLE 7.15. Effect of Some Beneficial Additions on the Grain Boundary Corrosion Rate of a Zn-O.IAI Alloy at 90°C" Addition (WI. %) 0.05% Mg 0.3% Cu 5%Au 0.5 Ni O.I%Pt
CorrosIOn rate (%l
15 15
5 65 75
"Reprinted from Devillers and Niessen [491. with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom.
hEx pressed as a percentage of the corrosion rate of the pure Zn-AI alloy.
The presence of small amounts of a single harmful element, other than aluminum, in pure zinc does not cause significant intergranular corrosion [48, 535], However, the intergranular corrosion can be severe when there are also other elements in the alloy. As observed by Roberts [535], Mg is the most beneficial element in inhibiting intergranular corrosion of Zn-AI alloys in warm water. If a small amount of magnesium is present alone in pure zinc, it causes only mild intergranular corrosion. However, when a small amount of lead, tin, or cadmium is also present along with the magnesium, severe intergranular attack occurs in the zinc-magnesium alloy, The intergranular corrosion of zinc-magnesium alloys has not been systematically studied. 7.4,3.2. Effect of Mechanical and Heat Treatments, Williams et al. [235] tested the effect of stress on the grain boundary corrosion of Zn-22AI in 100% relative humidity at 49°C and found that the depth of intergranular penetration is greater under tensile stress than under compressive stress. Devillers [48] found that stress has little effect on the grain boundary corrosion of Zn-O.I Al in 80°C water. Heat treatment generally atlects the number of grain boundaries and the extent of grain boundary segregation, Devillers [48] found that the corrosion penetration rate of annealed specimens is similar to that of quenched and aged specimens, whereas it is higher TABLE 7.16. Increase of Boundary Corrosion Rate in a Zn-O.I %AI Alloy with Addition of Harmful Impurity Elements at a Level of 0,01 at.%" Element Sn Pb Bi Cd
Hg TI
Percent increase in corrosion rate
400 300 260 250 50 50
"Reprinted from Devillers and Niessen [49], with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom.
232
CHAPTER 7
than that of quenched samples without aging, indicating that the segregation of aluminum at the grain boundaries can be significantly reduced by annealing and quenching. Annealing and quenching have little effect on the intergranular corrosion rate in water at high temperatures, because the diffusion of Al to the grain boundaries is sufficiently fast to cause a segregation. Melton and Edington [22S] found that the intensity of intergranular corrosion, measured as the dimensional growth of the test specimens, of extruded 22% AI-zinc alloy is lower in the extrusion direction than in the perpendicular direction owing to the directional microstructure of the material. 7.4.4. Effect of Environmental Factors As noted in Table 7.14, intergranular corrosion of zinc-aluminum alloys can occur under different environmental conditions. Among the environmental factors, temperature seems to be the most significant. Roberts [722] studied the effect of water vapor temperature from 80 to IS0°C in an autoclave and found that the corrosion penetration rate increased exponentially with increasing temperature as shown in Fig. 7.32. The relationship may be expressed as log d= a - KIT
(7.20)
where d is the rate of intergranular penetration, T is the absolute temperature, and a and K are constants. Due to this exponential dependence on temperature, the penetration rate can be very high at elevated temperatures. For example, as shown in Fig. 7.32, it can be as much as 4 mrn/day at ISO°C for Zn-O.lAI. A similar exponential relationship is found to exist between the corrosion rate of Zn-AI alloys and temperature in the range 80-120°C in tap water [48]; at higher temperatures, up to 200°C, the relationship becomes linear. Alkaline environments appear to be the most aggressive in intergranular corrosion of zinc-aluminum alloys. Figure 7.33 shows the effect of pH on the grain boundary corrosion rate in 9SoC tap water [49]. Between pH Sand 10, the corrosion penetration 10
.,>
- 0.1 %AI
~
E E C
.,
.0
1
f
+ 4%AI * 20%A1
~
CI ~
Q)
c. 0.1
~
0
~
<0
a:
FIGURE 7.32. Intergranular corrosion rate of zinc-aluminum alloys (prepared with 0.01 70
90
11 0
130
Temperature, C
150
QQ QQQ% nnrltv 7n !lnti QQ QRo;,., nJlr;tv A n!1~
233
CORROSION FORMS 250~----------------------------------------'
:;:;
.g
E
200
2-
~ 150 LD 0)
co Q)
~ 100 c 0
'en
2 0
50
U
oL-__ ____ ____ ____ ~
o
2
~
4
- L_ _ _ _- L____
~
6
8
10
~
12
__
~
14
pH
FIGURE 7.33. Effect of pH on intergranular corrosion rates of a rolled Zn-O. I wI. % Al alloy in 95°C tap water. Reprinted from Devillers and Niessen [491, with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane. Kidlington OX5 1GB, United Kingdom.
rate is almost constant. Below pH 5, it decreases with decreasing pH. On the other hand, for pH values above 10, it increases drastically with increasing pH. Grain boundary corrosion can also occur in dry water vapor where condensation does not occur, although the general corrosion rate is only about one-tenth of that in wet water vapor [722]. It is, therefore, believed that the intergranular attack of zinc alloys in water vapor can be both chemical and electrochemical in nature, depending on condensation conditions. When there is no condensation, the intergranular attack is slow and is chemical in nature. When the surface is wet due to condensation, on the other hand, the grain boundary corrosion is mainly of an electrochemical nature, and the corrosion rate is greatly increased. The presence of air, and thus oxygen, in water vapor was found by Roberts [722] to have no effect on the corrosion penetration rates of zinc alloys of three different Al contents, 0.1 %, 4%, and 20%. Williams et al. [235] studied the effect of sulfur dioxide on the grain boundary corrosion rate ofZn-22AI alloy in air with 100% relative humidity at 49°C. They found that intergranular corrosion occurs in the humid air without S02 but does not occur with 100 ppm S02; in the latter case, the general corrosion is more severe. It was also noted that the corrosion product formed in the humid air containing S02 was very soluble, whereas that formed in air without S02 was not soluble. External current, both anodic and cathodic, can affect the occurrence and the penetration rate of intergranular corrosion of zinc-aluminum alloys [154, 224]. Ahmed et al. [154) observed intergranular corrosion in Zn-0.3AI-0.03Cd in 70°C seawater under an anodic current of 0.5 mA/cm2 . Devillers and Niessen [224] reported that under relatively high applied cathodic current density (>0.5 mA/cm2) hot-rolled Zn-AI alloys in hot water suffer very rapid grain boundary attack and blistering, as shown in Fig. 7.34. This effect does not occur in pure zinc, but the addition of 0.03% Al is sufficient for this attack to occur. The intergranular corrosion rate under cathodic conditions depends less on alloy composition than is the case for corrosion without applied current. However, the
234
CHAPTER 7
~200
-0
E
~
..;
~ c o
.~
(5 u
1a
-0
C :J
o
co 3 Appl ied cathodic current density, rnA I em'
FIGURE 7.34. Intergranular corrosion rate of a hot-rolled Zn-AI alloy (Zn-O.l wt. % AI) in hot water as a function of cathodic current density. From Devillers and Niessen [224]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
rates were found to be strongly dependent on water composition and are particularly sensitive to the presence of carbonates.
7.4.5. Effect on Mechanical Properties As a result of intergranular corrosion, the strength of zinc alloys can be drastically reduced. Table 7.17 contains data reported by Kehrer [1290] showing the reduction in the strength of zinc-aluminum-lead alloys as a function of aluminum content after 10 days of exposure in 95°C steam. There is practically no mechanical strength left after the test for the alloys containing more than 0.05% Ai. Similarly, Melton and Edington [225] observed a 77% reduction of the original strength of a 22% AI-zinc alloy after 7 days in hot steam.
TABLE 7.17. Influence of Aluminum Concentration on the Strength of a Zn-I.l % Pb Alloy after 10 Days' Exposure at 95 °C in Air or Steama Relative strength Al concentration 0 0.01% 0.02% 0.03 % 0.05% 0.08% 0.12% °Ref. 1290.
In air 14 13 12.5 II 12 12 II
In steam II 7.5
6 3 0 0 0
CORROSION FORMS
235
7.4.6. Mechanisms
lntergranular corrosion of metals is, in most cases, associated with the presence at the grain boundaries of a phase whose electrochemical behavior is different from that of the matrix. This can happen in two-phase alloys where the second phase forms a continuous path along the grain boundaries or in unstable solid solutions where the solute segregates toward grain boundaries. The intergranular corrosion of zinc-aluminum alloys in hot steam or water has been attributed to the increased electrochemical reactivity of the grain boundaries caused by the segregation of impurities or precipitation of phases. In earlier days, due to the unavailability of high-purity zinc, the role of different impurities in intergranular corrosion was not clearly identified, and their segregation at the grain boundaries was generally regarded as the formation of the cathode with the adjoining grains as the anodes. It was not until later, as high-purity zinc became available, that the role of aluminum in causing intergranular corrosion was clearly identified. Several theories have been proposed to explain the processes involved in the intergranular corrosion in zinc-aluminum alloys. Roberts [535, 722] proposed that the intergranular corrosion is due to the preferential oxidation of the AI-rich phase at the grain boundaries. The oxidation can be brought about by a direct chemical reaction between water vapor and the interface material, as manifested by the fact that intergranular corrosion can occur in dry steam, where there is no condensation, to form a continuous electrolyte. The much faster corrosion penetration rate in wet steam is attributable to electrochemical reactions in which the aluminum-rich phase is the anode and the zinc matrix acts as the cathode. The effect of the corrosion-accelerating elements such as lead and tin is explained by either a buildup of a cathodic phase adjacent to the grain boundary zone or the increase in width of the distorted boundary zone resulting from the segregation of these elements at the grain boundaries. With regard to the reason for the effect of the impurity or alloying elements on intergranular corrosion, Roberts noted that there is no coherent explanation. The observation has been made that all inhibitive elements tend to form intermetallic compounds with aluminum, while none of the corrosion-accelerating elements do. By the formation of intermetallic compounds, the amount of aluminum available for diffusion to grain boundaries is reduced. According to Roberts, the inhibiting effect of magnesium on the intergranular corrosion of Zn-Al alloys is likely due to its role in reducing grain size because the segregation of impurities is distributed over a greater area with a corresponding reduction in the effective concentration of the impurities in the grain boundary. The severe intergranular corrosion that occurs with aluminum-free zinc containing small amounts of Mg and Pb or Sn is attributed to the precipitation of magnesium intermetallic compounds, probably Mg 2Pb and Mg 2Sn. These intermetallic compounds are found to be unstable in moist air and, hence, are rapidly attacked. In another theory, proposed by Devillers and Niessen [48,49], the preferential anodic dissolution of the aluminum particles precipitated at the alkaline grain boundaries is considered to be the primary cause of the intergranular corrosion of zinc-aluminum alloys. The anodic dissolution is coupled with a cathodic reduction of hydrogen at the tip of the corroding grain boundary. The alkalinity of the grain boundary electrolyte is supposed to be caused by the formation of OH- ions from a cathodic reduction of
236
CHAPTER 7
1.6
§ 1.2 4:
E ,;.
.i;i 0.8 c
C!>
1)
C C!>
~ 0.4
()
AI
0
-1.5
-1
Potential, V
-0.5
o
FIGURE 7.35. Potentiostatic anodic polarization of Al and galvanostatic cathodic polarization of Zn, Pb, Sn, In, and Cu in a pH 10.5 solution. Reprinted from Devillers and Niessen [49], with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom.
hydrogen. This theory is supported by the fact that the intergranular corrosion rate increases with increasing pH (Fig. 7.33). The effect of the different elements on the corrosion rates, in Devillers and Niessen's theory, is explained according to their effects on the corrosion potential of aluminum in the active-passive transition range in alkaline solutions as shown in Fig. 7.35. In the active-passive transition potential range, Al shows an anodic dissolution peak followed by a passivation. The corrosion rate of the AI-rich phase at the grain boundaries is, therefore, strongly dependent on the cathodic reaction rates, which depend on the presence of impurities. According to Fig. 7.35, the presence of impurities with higher hydrogen overpotential than zinc will increase the corrosion rate and thus prevent the passivation of AI. Elements with more noble potentials, such as Cu, if present in large enough quantity at the grain boundaries, will reduce the corrosion rate by shifting the corrosion potential at the grain boundaries to more noble values, thus facilitating the passivation of AI. The special effect of Mg, i.e., its inhibition of intergranular corrosion in Zn-AI alloys, is believed to be the result of its formation of a solid solution with AI, hence lowering the critical potential for passivation, and also its anodic polarization characteristics, allowing easier passivation. Mg has no solid solubility in Zn but has appreciable solid solubility in AI. 7.5. WET STORAGE STAIN "Wet storage stain" is a term used in the galvanizing industry to describe the zinc corrosion products formed on a galvanized steel surface during the period of storage. It is also referred to as "white rust," which is a term generally applicable to all zinc corrosion products. Wet storage stain is voluminous, white, powdery, and bulky and is formed when closely packed galvanized articles are stored under damp and poorly ventilated conditions [51-53]. The crevices formed between the articles can attract and absorb moisture and retain the wetness more readily than the surface area exposed to the open air. This noneven coverage of moisture on the galvanized steel articles results in localized patches of
237
CORROSION FORMS
corrosion. Although wet storage stain can seriously affect the appearance of the galvanized steel articles in some situations, it is generally not harmful in terms of the long-term corrosion performance [53]. The moisture necessary for the formation of wet storage stain may originate in various ways. It may be present on the galvanized parts at the time of stacking or packing, as a result of incomplete drying after quenching. It may result from direct exposure to rain or seawater or from condensation caused by atmospheric temperature changes. Close packing can result in moisture being retained by capillary action between the surfaces in contact, where drying is delayed by the lack of circulating air. In practice, wet storage stain is most often formed by condensed moisture in a confined environment where the relative humidity is high. The corrosion process experienced during storage is generally different from the corrosion process in an open atmosphere, where the zinc surface is periodically wetted and dried. Gilbert and Hadden [437] did an extensive study on white rust formation. They found that the form of moisture is important for extensive white rust formation. Trapped water, either from rain or from condensation, between the surfaces of zinc articles is not readily dried and is usually responsible for considerable amounts of white rusting. The amount of corrosion under the stains may significantly vary with circumstances. Gilbert noted that zinc sheets, exposed in a glass tank in which the relative humidity was 100%, for I week were only tarnished. Considerable corrosion occurs within a few hours, however, when drops of water are deposited on the surface of zinc. The conditions that facilitate condensation promote corrosion. In one example, specimens placed over dishes containing ice corroded at a rate of 0.15 pm/day. In another example, the external surface of a zinc-coated steel pipe, in which cold supply water was passed, corroded at a rate of 0.4 pm/day. White rust formed in uncontaminated air usually consists mainly of a basic zinc carbonate, often mixed with zinc oxide; in the absence of carbon dioxide, it consists only of zinc oxide [52, 437J. Under confined conditions, due to the lack of air circulation and drying, white rust is less compact and less likely to be converted into the more protective zinc carbonates. Table 6.8 in Chapter 6 shows the chemical composition of the white rust formed in various types of tests. The primary electrochemical reaction leading to the formation of white rust in the presence of air and moisture is, according to Gilbert and Hadden [437], Zn 2+ + 20W ~ Zn(OH»),
(7.21 )
The zinc hydroxide is precipitated by interaction of the products from adjacent anodic and cathodic areas. Secondary reactions which then occur are: (7.22)
or (7.23)
or
238
CHAPTER 7 tf>
120
>-
'"
Test
"0
-0 <1l
'"ca
-r
'Y Salt· spray
<:
·16 Vi
-t-
• Condensation·cabinet
80
Q)
Water·lilm • Water·tog
Q)
'-' ~
:; II>
?f-
a
40
'" .9 Q)
E
i=
0 0.5
1.5
2
2.5
Total chromium on sheet surface, 119/cm'
FIGURE 7.36. Effect of chromating on performance of galvanized sheet in various accelerated corrosion tests. Data are taken from Ref. 52.
(7.24) Gilbert reasoned that the formation of zinc carbonate is a secondary reaction because the amount of corrosion in a water spray test is the same whether ordinary air or COr free air is used to saturate the water, but the corrosion product is basic carbonate in the one case and zinc oxide in the other. Zinc hydroxides do not tend to form in damp air, where the conditions favor the formation of zinc oxide or, in the presence of carbon dioxide, basic zinc carbonate. Severe white rust can be produced in moist air containing contaminants such as sulfate and chloride [52, 437]. The presence of flux residues, from galvanizing, also enhances the formation of white rust [437]. Chromating has been used in the galvanizing industry as an effective surface treatment to prevent wet storage stain from forming during storage or transportation periods [51-57,437]. Figure 7.36 illustrates the effect of chromating on the formation of white rust in four different tests [52]. Different types of tests, with various degrees of simulation and acceleration, can be used to study surface staining by corrosion, as also shown in Fig. 7.36. The difference between the zinc losses in the water film and condensation tests indicates that oxygen concentration may play an important role in the formation of wet storage stain because the electrolyte is open to air in the condensation test while in the film test, the electrolyte is isolated from the air. More information on white rust formation in humid air and the characteristics of zinc corrosion products can be found in Chapters 8 and 6. 7.6. HYDROGEN EMBRITTLEMENT AND CORROSION CRACKING Hydrogen embrittlement is a loss of mechanical strength and ductility in a metal caused by the interaction between the metal and hydrogen, which may be a by-product of the corrosion process. Corrosion cracking refers to the cracking caused by the simultaneous presence of tensile stress, either residual or applied, and a corrosive medium.
CORROSION FORMS
239
Hydrogen embrittlement and corrosion cracking are rarely encountered in practical applications of zinc and its alloys, largely because zinc is not usually used as a structural material to bear stresses. In one of the few studies on the subject, Foster et al. [426 J showed that a galvanizing alloy (eta-phase alloy) was susceptible to stress corrosion failure when exposed to 60°C circulating potable water and that the failure always occurred in a direction normal to the applied stress. The time to failure decreased exponentially with increasing applied tensile stress. Since the cracking was observed to be reduced or completely eliminated for the samples tested under a cathodic polarization, it was suggested that the stress corrosion process is controlled by anodic dissolution. Both intergranular and trans granular cracking of zinc have been observed for certain alloys in corrosion tests in laboratory environments [154,203]. The intergranular cracking of the zinc alloys in most investigations was found to be caused by intergranular corrosion [225, 203, 235, 426]. It was also attributed to hydrogen embrittlement in some studies [154,224J.
8 Atmospheric Corrosion 8.1. INTRODUCTION Atmospheric corrosion is the most prevalent type of corrosion for zinc, owing to extensive outdoor applications of galvanized steels. Numerous research programs have been carried out in the past to investigate the corrosion behavior of zinc in various types of atmosphere. The testing methods used in these investigations can be generally divided into two groups: field exposure testing and simulated testing. Data from field exposure represent a real corrosion rate in an atmospheric environment, while those from simulated testing provide specific information on the effects of the atmospheric variables. In view of this intrinsic difference, information from real field exposure testing and that from simulated testing are presented separately in this chapter. A brief description is provided first on the atmospheric factors involved in a corrosion process. The mechanisms for atmospheric corrosion of zinc are discussed at the end of the chapter. 8.2. ATMOSPHERIC FACTORS 8.2.1. Tvpe of Wetting Atmospheres in different geographic locations vary greatly with respect to solar radiation, temperature, moisture, wind, air constituents, and air pollutants. Atmospheric corrosion of a metal alloy is affected by many factors [540-542], particularly those affecting surface wetness and precipitation of pollutants [312, 315, 509]. The duration and form of wetness are also important in determining the corrosion behavior of a metal. Experimental results have indicated, for example, that, for the same time of wetness, rain could cause more corrosion of zinc than dew [614, 1094]. Wetness in an outdoor atmosphere is usually generated by condensation, fog, or rain. In the case of condensation, the level of wetness depends on the water content of the air and the temperature. Figure 8.1 shows that the amount of water in air increases with increasing temperature and relative humidity. When the water content reaches the saturation level, i.e., at a relative humidity of 100%, water will condense in the form of dew. The rate of condensation on a surface is a function of substrate temperature. Figure 8.2 shows that the rate of water condensation decreases with increasing sample temperature [556]. Condensation depends also on the hygroscopicity of the surface contaminants and 241
242
CHAPTER 8
50
100%
40
90%
'OJ
30
65%
.
20
~
~
....
.5 ....
B ~
Ph
45%
~30%
10 -10
0
10
20
30
Temperature (0C)
FIGURE 8.1. Absolute water vapor content in the air at different temperatures and relative humidities. From Rozenfeld [556]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
corrosion products. All water-soluble salts are hygroscopic to some extent. A rough or porous surface structure can result in capillary. condensation, where water condenses inside a pore at a vapor pressure below the saturation level. It has been calculated that the relative humidity for condensation decreases from 98% to 50% as pore radii decrease from 360 to 15 A [542]. Wind is found to increase the relative humidity at which condensation occurs [615]. Rain has an effect of not only wetting the surface of an exposed metal but also washing away pollutants and corrosion products. In some areas, the rain can be polluted (e.g., acid rain), which increases its corrosiveness. Compared to rain, snow has a negligible effect on the corrosion of metals because snow is not an electrolyte and a corrosion process cannot proceed without an electrolyte [13]. Metal surfaces can also receive water through adsorption. It has been reported that a zinc surface at O°C adsorbed 15 molecular layers of water at 55% relative humidity, 17 layers at 93%, and 92 layers at 100% [542]. The amount of adsorption changes drastically with changes in sample temperature, as shown in Fig. 8.3 [1265]. Moisture, in the form of rain or dew, on a metal surface will evaporate under the drying effect of temperature, radiation, or wind, thus causing an exposed metal surface to continuously undergo cycles of wetting and drying. One effect of drying is that solutes (pollutants and corrosion products) are solidified. The duration of wetting and drying as
-.:..c; Ne ~ ~
0.6
~....
0.4
c::
0
'i '"c::
'8"0
u
0.2
5
10
15
Temperature (0 C)
20
FIGURE 8.2. Dependence of condensation on metal temperature. Air temperature is 25°C, and relative humidity is 100%. From Rozenfeld [556]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
243
ATMOSPHERIC CORROSION
20~------------------------------,
(f)
Ql
>-
15
cu
o c
o
E 10
-0 Ql
.D
o(f)
~ FIGURE 8.3. The amount of physically adsorbed moisture on the surface of zinc at 93% relative humidity and various temperatures. After Strekalov el aL. 112651.
5
oL-----~--~--~--~~--~----~
o
10
20
30
40
50
60
70
80
90
Temperature (oe)
well as the frequency of the cycles has a significant effect on the morphology and compactness of corrosion products [614, 1194].
8.2.2. Air Pollutants The normal composition of air is summarized in Table 8.1 [540], In many places, such as in cities, the air is polluted. The major air pollutants are sulfur dioxide, hydrogen sulfide, oxides of nitrogen. and aerosols. Near the seacoast, the air is laden with sea salts, particularly NaCI. In industrial areas, appreciable amounts of S02 and lesser amounts of H 2 S, NH 3 , N02 , and other suspended salts are encountered [542]. The concentration of pollutants varies greatly from location to location. Table 8.2 shows typical levels of gaseous pollution. The type and concentration of pollutants may vary with the form of water precipitation. It is reported that fog contains the highest levels of H+ ions [492]. Rain receives a higher degree of mineralization than mist precipitation. As the raindrops fall, they wash a large volume of air and catch salt particles suspended in the air. The amount of salt
TABLE 8.1. Approximate Constitution of the Atmosphere at 10°C and 760 mm Hg Pressure, Excluding Impurities" ------_._----
Amount present Constituent Air Nitrogen Oxygen Argon Water vapor Carbon dioxide "Ref. 540.
glm
3
1172 879 269
IS 8 0.5
Weight % 100 75 23 1.26 0.70 0.04
Amount present Constituent Neon Krypton Helium Xenon Hydrogen
mglm
3
14 4 0.8 0.5 0.05
ppm 12 3 0.7 0.4 0.04
244
CHAPTER 8
TABLE 8.2. Typical Concentrations (mg/m3) of Impurities in the Atmospherea Hazy winter day
Summer
Impurity
Town
Country
Sulfur dioxide Solid particles Carbon monoxide Sulfur trioxide Ammonia
1.2 1.2 10.0 0.01
0.15 0.24
Town
Country
0.2 0.2 2.0 0.0001
0.03 0.05
0.01
"Ref. 540.
dissolved in raindrops varies between 20 and 200 ppm, depending on the geographic location [556]. Nitrate appears to be predominantly a result of gaseous deposition; deposition of organic substances on a metal surface is higher indoors than outdoors; and sulfate ions may come more from dew than from other forms of precipitation [492]. The most important corrosive constituent of industrial atmospheres is sulfur dioxide, which originates predominantly from the burning of coal, oil, and gasoline [13]. Table 8.3 shows that the concentration of sulfur dioxide in the air is high inside cities and falls off with distance from a city because of the higher fuel consumption inside cities [1266]. Sulfur dioxide slowly oxidizes homogeneously in the presence of oxygen and moisture to form sulfuric acid. The heterogeneous reaction of S02 in the presence of moisture and oxygen is found to be significant after adsorption onto a substrate, such as an airborne particle or an exposed metal surface [542]. The adsorption of S02 on a metal surface increases with increasing relative humidity. Comparative data shown in Fig. 8.4 indicate that surface adsorption on zinc is somewhat lower than that on Fe, but higher than that on Cu and AI. Other air contaminants are hydrogen sulfide and nitrogen compounds. Hydrogen sulfide is produced by putrefaction of organic sulfur compounds or by the action of sulfate-reducing bacteria in anaerobic conditions. Since hydrogen sulfide is easily oxidized, it generally occurs at very low concentrations [540]. Nitrogen compounds are produced during electric storms and by decay of organic matter. Their most significant contribution to corrosion is probably through the formation of ammonium sulfate.
TABLE 8.3. Variation of S02 Content of Air with Distance from Center of Citya.b S02 content (ppm) at a distance (in kilometers) of: City Detroit Philadelphia-Camden Pittsburgh St. Louis Washington, D.C.
0-8
8-16
16-24
24-32
32-40
40-48
0.023 0.030 0.060 0.111 0.003
0.012 0.018 0.030 0.048 0.001
0.006 0.016 0.015 0.029 0.001
0.004 0.021 0.018 0.020 0.001
0.004 0.012 0.009 0.018 0.001
0.005 0.012 0.010 0.014 0.002
"Ref. 1266.
urhe level of S02 has decreased over the years due to environmental regulations.
ATMOSPHERIC CORROSION
245
Fe
1.5
FIGURE 8.4. Rate of adsorption of S02 on Fe. Zn. Cu, and Al surfaces after 200-min exposure to S02 at different relative humidities. Reprinted from Sydberger and Vannerbcrg [1267], with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB. United Kingdom.
0.5 60
70
80
Relative Humidity (%)
Dust is an important solid contaminant of many atmospheres. Industrial atmospheres carry suspended particles of carbon, carbon compounds, metal oxides, H2S04 , (NH4)2S04' NaC!, and other salts [13]. Marine atmospheres contain salt particles that may be carried for miles inland, depending on the speed and direction of the wind. These substances, combined with moisture, initiate corrosion by forming galvanic or differential aeration cells, either by their hygroscopic nature, hence forming an electrolyte on the metal surface, or by the capillary effect of inducing condensation at lower relative humidities. The amount of pollutants in air has dropped over the years due to environmental regulations [623], 8.3. CORROSION IN OUTDOOR ENVIRONMENTS 8.3.1. Typical Corrosion Rates
The corrosion rate of a metal varies significantly from one geographic location to another and from time to time. The corrosion rate of zinc is lowest in dry, clean atmospheres and highest in wet, industrial atmospheres. Seacoast atmospheres not in direct contact with salt spray are mildly corrosive to zinc [546, 549]. Locations near sea level are subject to salt spray, and hence the corrosion rate can be much higher at such locations. The typical atmospheric corrosion rates of zinc and its alloys are [4 J: Rural
0.2-3 JlmJyr
Marine (outside the splash zone)
0.5-8
Urban and industrial
2-16
Table 8.4 lists the rankings of 45 different atmospheres around the world with respect to their corrosivity for zinc and steel in the period between 1960 and 1962 [539].
246
CHAPTER 8
TABLE 8.4. Rankings of 45 Locations with Respect to Corrosivity for Steel and Zinc from Two Years' Exposure a ." Weight loss (g)
Ranking Location Norman Wells, N.W.T., Canada Phoenix, Ariz. Saskatoon, Sask., Canada Esquimalt, Vancouver Island. Canada Detroit, Mich. Fort Amidor Pier, Panama Canal Zone Morenci. Mich. Ottawa, Ont., Canada Potter County. Pa. Waterbury. Conn. State College. Pa. Montreal, P.Q., Canada Melbourne, Australia Halifax (York Redoubt). N.S., Canada Durham. N.H. Middletown. Ohio Pittsburgh. Pa. Columbus. Ohio South Bend. Pa. Trail. B.c.. Canada Bethlehem. Pa. Cleveland. Ohio Miraflores, Panama Canal Zone London (Battersea). England Monroeville, Pa. Newark, N.J. Manila, Philippine Islands Limon Bay, Panama Canal Zone Bayonne, N.J. East Chicago, Ind. Cape Kennedy, Fla., ~ mile from ocean Brazos River, Tex. Pilsey Island, England London (Stratford), England Halifax (Federal Building), N.S., Canada Cape Kennedy, Fla., 60 yards from ocean, 60-ft. elevation Kure Beach, N.C., 800-ft. lot Cape Kennedy, Fla., 60 yards from ocean, 30-ft. elevation Daytona Beach, Fla. Widness, England Cape Kennedy. Fla., 60 yards from ocean, ground level Dungeness, England Point Reyes, Calif. Kure Beach, N.C., 80-ft. lot Galeta Point Beach, Panama Canal Zone "Ref. 539. "Specimen size. 4 x 6 in.
Loss ratio Zinc Steel/Zinc
Steel
Zinc
I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
2 3 4 15 5 II 7 \3 31 10 28 6 20 19 12 30 27 21 18 14 33 8 29 24 35 16 32 39 22 9 23 40 42 43 38
0.73 2.23 2.77 6.50 7.03 7.10 9.53 9.60 \0.00 11.00 I I. 17 11.44 12.70 12.97 13.30 14.00 14.90 16.00 16.20 16.90 18.3 19.0 20.9 23.0 23.8 24.7 26.2 30.3 37.7 41.1 42.0 45.4 50.0 54.3 55.3 64.0
0.07 0.13 0.13 0.21 0.58 0.28 0.53 0.49 0.55 1.12 0.51 1.05 0.34 0.70 0.70 0.54 1.14 0.95 0.7S 0.70 0.57 1.21 0.50 1.07 0.84 1.63 0.66 1.17 2.11 0.79 0.50 0.81 2.50 3.06 3.27 1.94
10.3 17.0 21.0 31.0 12.2 25.2 18.0 19.5 18.3 9.8 22.0 10.9 37.4 18.5 19.0 26.0 13.1 16.8 20.8 24.2 32.4 15.7 41.8 21.6 28.4 15.1 39.8 25.9 17.9 52.1 84.0 56.0 20.0 17.8
37 38
26 36
71.0 80.2
0.89 1.77
80.0 45.5
39 40 41
25 44 37
144.0 174.0 215.0
0.88 4.48 1.83
164.0 39.0 117.0
42 43 44 45
34 17 41 45
238.0 244.0 260.0 336.0
1.60 0.67 2.80 6.80
148.0 364.0 93.0 49.4
Steel
17.0
33.0
247
ATMOSPHERIC CORROSION
Compared to the corrosion rate of steel, that of zinc exhibits less variation with geographic location. Table 8.4 shows that the corrosiveness of the atmosphere from one location to another varies by as much as a factor of 100 for zinc and 500 for steel. It also shows that the atmospheric corrosion rate of zinc in most locations is at least 10 times lower than that of steel, which is why zinc is commonly used, through the galvanizing process, to effectively protect steel from corrosion. The difference in corrosion rates between zinc and steel tends to be larger in marine environments and less in rural environments. Furthermore. it can be seen from Table 8.4 that the relative corrosiveness of different atmospheres toward steel may be very different from that toward zinc. For example. among the locations compared. Waterbury. Connecticut has a ranking of only 10 for steel but 31 for zinc; Cape Kennedy has a ranking of 31 for steel but only 9 for zinc. These variations and differences among atmospheres and materials demonstrate the very complex nature of atmospheric corrosion. Even in atmospheres with little contamination and within one geographic region. corrosion rates can vary significantly. For example. the corrosion rate of zinc measured at 26 sites in one rural area in Spain varied from 0.7 to 2.4 jlm/yr 13061. Also. atmospheric conditions. such as the amoum of rain or the pollution level. change over time. and thiS can cause the corrosion rate to change from year to year. The yearly averaged corrosion rate of zinc does not vary much for a given atmosphere 1190.217.550 I. Figure 8.5 shows that the corrosion loss of zinc measured at a single site in British Columbia was almost linear as a function of time [1901. It shows also the seasonal effects due to variations in time of wetness and the amount of air pollutants. Due to environmental regulations and awareness. the level of pollution in many developed countries has been considerably reduced over the years; similar trends have generally been found for corrosion rates of many metals. The corrosion rate of zinc was found to be higher between 1960 and 1970 than it was in the 1980s 1241. 248. 6161. Table 8.5 presents a comparison of the corrosion rates of zinc and other common commercial metals in various atmospheric environments. Accordingly. thc corrosion resistance of zinc is higher than that of iron and cadmium in all environments, highcr than 1,000 • May t959
Q;
c
+ Mar. 1960
800 ·
(1)
..e0> E
~
Dec, 1958
o
Sep. 1960
X No •. 1960
600 r
III III
.2 C
0
~
0 ~
X
400 L
'iii
g 0
tJ
FIGURE ~.5. Corrosion loss versus time curve for zinc specimens exposed at different times of the year. Dala are taken from Ref. 190.
200 ~
£~
0' 0
l
. 20
" + 40
60
80 100 120 140 160 180 200
Exposure time (weeks)
248
CHAPTER 8
TABLE 8.5. Comparison of Typical Corrosion Rates of Zinc and Other Common Commercial Metals in Various Atmospheric Environments Relative corrosion rate Metal
Industrial
Zn
I 2 0.23 0.13 2.4 0.07
Cd Sn Al Cu Pb Ni Sb Mg Fe
0.06 0.31 30
Marine
Rural
I
1
2 1.6 0.3 0.72 0.3 0.6
2.4 1.9 0.09 0.38 0.28 l.l
1.8 50
1.9 15
Reference(s) 297 292,543 543 543 543 543 543 544 217,539
that of copper in industrial environments, and higher than that of tin and magnesium in marine and rural environments. There is no single set of rules for a reliable estimation of the corrosion rate of a metal in the atmosphere for every geographic location. Field exposure data are the best basis for reliable prediction of the corrosion resistance of zinc and its alloys in the atmosphere. The corrosion rates of zinc and its alloys in many locations can be found in an Intemation Lead Zinc Research Organization (ILZRO) publication, Zinc: Its Corrosion Resistance [217], and review articles [294, 331]. Other corrosion data have been reported from many countries, including Australia and New Zealand [300], Brazil [301], China [302], Finland [303], Norway [304], South Africa [305], Spain [306], Czechoslovakia [308], Poland [297], the Soviet Union [556], and Japan [623]. 8.3.2. Effect of Time of Wetness
The corrosion of zinc is negligible when the relative humidity is low but is significant when the surface is wet at a high relative humidity. A zinc surface can become wet at a relative humidity below 100%. Because of the presence of hygroscopic impurities in the air or in the metal itself, the critical relative humidity for condensation to occur is usually much lower than 100%. The time of wetness can therefore be defined as the total time during which the relative humidity is higher than the critical value. Guttman and Sereda [190, 191] have experimentally determined that the time of wetness for zinc is the time during which the humidity exceeds 86%. They also derived an empirical equation for calculating the corrosion loss of zinc as a function of the time of wetness. Time of wetness can be measured with various types of miniature moisture sensors made of bimetal couples, which often include zinc as the anode material [201,210,537,538,615,312]. It can also be estimated from meteorological data [548]. Time of wetness depends on the position and orientation of the exposed surface. For a particular building, time of wetness was found to be longer on the roof than on a side wall by about a factor of 2 [541]. Time of wetness was about 50% higher near the top of
ATMOSPHERIC CORROSION
249
TABLE 8.6. Average Corrosion Losses of Zinc Coatings on Buildings in Various Locations and Atmospheric Environments after 10 Years" Corrosion loss (mill 55%AI-Zn
Zinc Location
Environment
Rural Urban Industrial Severe industrial Inland, shore of Rural lake or marsh Urban Industrial Severe industrial Coast Rural Urban industrial Inland
Seashore
Severe industrial Severe industrial
Roof
Wall
Roof
Wall
0.42 1.48 1.40 1.59 0.59 1.97 1.40 2.12 0.74 2.47 1.75 2.65 2.06
0.17 0.59 0.56 0.64 0.24 0.79 0.56 0.85 0.29 0.99 0.70 1.06 0.82
0.15
0.06
0.15
0.06
0.20
0.08
0.20
0.08
0.25
0.10
0.25
0.10
0.46
0.18
"Reprinted from Suzuki [312], by courtesy of Marcel Dekker.lnc. hI mil = 25.4 ,urn.
the building than at the bottom [541]. As seen in Table 8.6, the corrosion was at least twice as severe on the roof as on the side wall [312].
8.3.3. Effect of Pollutants Sulfur dioxide, which is one of the most serious atmospheric pollutants, induces an abnormally high corrosion rate in zinc [4, 217]. Other air pollutants, such as NOx , have a relatively insignificant effect on the corrosion of zinc, largely because of the much lower content of these species in the air [331, 616]. Sulfur dioxide has a high solubility in water and, when dissolved, forms sulfurous acid, H 2S0 3• For example, when the concentration of S02 in air is O.l %, the concentration of H2S03 in water is 1.6 x 10-3 molll, and the resulting pH at equilibrium is low (2.8). It has been found that there is a direct correlation between the corrosion weight loss and the amount of S02 adsorbed by zinc, as shown in Fig. 8.6 [217]. Figure 8.7 shows that the corrosion rate followed the winter rise and summer fall of humidity and of the S02 content in air, due to the higher fuel consumption in the winter than in the summer [555]. Haynie and Upham [509] proposed a simple linear function to relate S02 concentration with zinc corrosion rates measured over four years in eight major cities in the United States, as shown in Fig. 8.8:
Y =0.00103(RH - 48.8) x [S02] The variable Y is in microns per year, and [SOz] is in micrograms per cubic meter. Below a relative humidity (RH) of 48.8%, the zinc surface was considered not wet, and no
250
CHAPTER 8
.-... "'E ..... ~
..
"'0 a.>
V
•
S02 ab orption weightlo
2
.c 0
'"
"'0
'" 0 CIl "'0
;a '"
I
'" .9
I
.c!>I) ·v ~
Nov.
Oct.
April
Month of the year FIGURE 8.6. S02 adsorption and corrosion loss of zinc at Stuttgart at different times of the year. Data are taken from Ref. 217.
corrosion should occur [217, 509, 556]. Similar empirical equations have been derived in other investigations [304, 625]. Experimental results indicate that there is a stoichiometric relation between Zn 2+ and SO~- in the corrosion products [426,626]. Spence and co-workers [626-628] analyzed the ionic species in the runoff water collected from zinc samples under field exposure, obtaining the data shown in Table 8.7, and found that zinc loss is directly related to the amount of pollutants precipitated on the zinc surface. The pollutants react with the zinc ~
"t:!'"
ME
.....
•
~
o
c::
o
.E-0 0
Relative humidity S02 absorption Weight loss
.-...
~
0'"
0.4
90
0.3
70
:a
c::
0.2
50
..c::
'" .9
0.1
30
.c '"
CIl "'0
'"'"
.c
.~ a.>
~
Oct.
Oct.
April
1940
April
Oct. 1941
Oct.
April
10
0
·s ;::I
a.>
;> .~
Q) ~
1942
Time (month of the year) FIGURE 8.7. Monthly corrosion rate and S02 adsorption rate of zinc at Berlin (Dahlem). Data are taken from Ref. 217.
ATMOSPHERIC CORROSION
251
8 >;
E
..:!. OJ
E
c 0 'v;
e
::;
u
0
0
6
slope = 1.03
4 0
'!.
00
0
2
FIGURE 8.8. Effect of atmospheric S02 concentration and relative humidity (RH) on the corrosion of zinc. From Haynie and Upham [509]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
2
3
4
5
6
7
Atmospheric factor (RH - 48.8)~S02l (mg/ml )
corrosion products (oxide, hydroxides, and carbonates) and cause their dissolution when the surface is wet. The dissolved species are then washed away in the event of rain. Spence and co-workers derived a linear damage function for galvanized steel structures based on the chemical analysis of the runoffs and on the assumption that a stoichiometric relation exists between the amount of zinc corrosion and the amount of corrosive species (H+, SO;, CO~-) precipitated on the surface. The reactions responsible for the dissolution of zinc corrosion products are: (I) zinc replacing hydrogen ions in rain, [H+lrain: (2) solution of basic zinc carbonate by dissolved carbon dioxide in rain, [HCOllain ; (3) formation of solution of soluble zinc sulfate from rain, [S02Ler; and (4) dry deposition of sulfur dioxide gas, [S02]dry' The overall corrosion rate, R, is
in which ao, at, a2, and a3 are constants. These constants are determined by a number of factors such as wind direction and shape and size of samples r123-125, 522, 556].
TABLE 8.7. Ionic Species Detected in Galvanized Steel Runoff Water at Sites in Ohio (OH) and North Carolina (NCt Rate of detection [nmoll(cm 2 day)] Site
HCOO-
Cl-
NO;-
SO~-
Na+
K+
H+
Ca2+
Zn 2+
NC OH
1.17 0.36
1.04 2.90
0.60 0.62
2.63 13.99
0.21 0.41
0.23 0.47
0.12 0.09
0.38 7.29
5.34 10.05
"Ref. 626.
252
CHAPTERS ~
6 ><
~ ~ ~ ~ ~
0
1'i
~
1
20
... <:
10
~
weight loss salinity
30
..9 ·S
.!:!' Q)
2
40
10
1000
100
Distance from the sea (m)
FIGURE 8.9. Corrosion loss of zinc and air salinity as a function of distance from the seashore. From Brown and Masters [542]. Reprinted by permission of John Wiley & Sons, Inc.
8.3.4. Effect of Elevation and Distance from Seawater Near the seacoast the major pollutants are chloride salts. In relative terms, sea salts are less corrosive than S02' The typical corrosion rate of zinc in a marine atmosphere is about 2.5 flm/yr, which is about 25 times less than that of iron. The corrosion rate increases when zinc is exposed closerto seawater [191,436,542]. Figure 8.9 shows thatthe amount of corrosion decreases with distance from the seashore because the salt content in air drops significantly with distance from the shore [542]. Since the time of wetness also drops with distance, the difference between the corrosion rates inland and at the seashore cannot be entirely attributed to the difference in salt concentration. For example, it has been found that, due to the mist near the sea, the time of wetness at 800 ft from the shore is only two-thirds that at 80 ft [191]. The longer time of wetness near the sea is partially due to the mist and partially to the higher salt concentration, because salt promotes condensation at lower humidities. Table 8.8 shows that the corrosion rate of zinc varies little with changes in elevation, in contrast to the corrosion rate of steel, which is very sensitive to the elevation above sea level [539]. 8.3.5. Effect of Initial Weather Conditions The conditions at the time of initial exposure exert marked effects on the corrosion of zinc, as illustrated in Fig. 8.10 [618, 191]. This figure shows that samples that suffered the more severe attack initially continued to corrode at a higher rate than those which
TABLE 8.8. Weight Losses (in Grams) for Steel and Zinc at Different Elevations and Distances from the Oceana •b 60 yards from ocean Material
~ mile from ocean
Steel Zinc Steellzinc aRef. 539.
bAt Cape Kennedy, Florida. 'Elevation.
42 0.51 82.3
Ground
30 f{
60 ftc
215 1.83 117
80.2 1.77 45.3
64 1.94 32.9
253
ATMOSPHERIC CORROSION
0.4
~
'"
'" £!
August
0.3
.c<>0
·v i3:
0.2
0.1
40
FIGURE 8.10. Effect of conditions at the time of initial exposure on the atmospheric corrosion of zinc. After Elli s 16181.
80
160
120
200
Time (days)
were less severely corroded initially. Long-lasting rainfall or a relative humidity at or near 100% during the first days tends to cause a high corrosion rate. Figure 8.10 also shows that the corrosion rate is, in general, higher at the beginning of the exposure and decreases thereafter. This effect can also be seen in Fig. 8.11 [436]. The decrease of corrosion rate with time is generally due to the gradual formation of protective corrosion products.
4 .---------------------------------------------~
marine atmosphere
~
<0
a:2 c
.2
'"~
0, (;
OL---~---L--~----L---~---L--~----~
o
20
40
60
80
100
1 20
1 40
1 60
__~__~
1 80
200
Time of Exposure (months) FIGURE 8.11. Variation of corrosion rate with time in a marine atmosphere. Data are taken from Ref. 436.
254
CHAPTER 8
8.3.6. Effect of Climate One of the climatic effects on an exposed metal surface is the pattern of periodic wetting and drying. Depending on weather conditions, wetting can take the form of condensation or rain. In the case of condensation, the moisture has little effect on the accumulation of corrosion products. Rainwater, on the other hand, can wash away the corrosion products. The duration of wetting and drying can also have a significant effect on the compactness of the corrosion products. Depending on the type of wetting and drying pattern, the amount and the form of corrosion products can differ greatly from one atmosphere to another. For example, the corrosion rate of sheltered (without effect of rain) zinc samples may be several times less than that of unsheltered samples [307]. Other climatic factors, such as wind and radiation, may also affect condensation and rate of drying as well as the amount of contaminants and corrosion products retained on the surface. Recent studies have shown that wind can be important in the corrosion of zinc coatings, because wind velocity determines the amount of sulfur dioxide deposition on the zinc surface [626]. The corrosion rates also vary depending on the time of day when the samples are exposed [1268]. Table 8.9 shows that the highest corrosion rate was exhibited by specimens exposed in the early morning period when the probability of wetting was high owing to dew formation. In another study, it was reported that nighttime corrosion is about three times more than daytime corrosion [620]. The increased rate was attributed to high humidity during the night.
8.3.7. Effect of Sample Configuration The size, shape, and orientation of test samples may considerably affect the corrosion rate of zinc [617]. The corrosion rate is higher on the skyward surface than on the groundward surface even though the wetting time is longer on the groundward surface [187, 190,315]. This may be attributed to the effect of rain and the larger amount of precipitation of pollutants on the skyward surface. While the corrosion rate of a skyward sample is generally linear with respect to time, the groundward surfaces do not show linearity between the corrosion rate and time [295]. Thin wire is normally found to corrode faster than thick wire. The corrosion rate in terms of weight loss per unit area of a 0.5-mm-diameter zinc wire was found to be about four times that of a 12.5-mm-diameter wire, which had about the same rate as flat surfaces
TABLE 8.9. Weight Losses of Rolled Zinc Alloys Exposed during Different Periods of the Day for Six Years a Weight loss (g)/panel Exposed period
a. 8:15 A.M.-4:15 P.M. b. 4:15 P.M.-12:15 A.M. c. 12: 15 A.M.-8: 15 A.M. Sum of a. b, and c 24 hr, continuous QAt
Zn-0.01Pb-0.OO3Cd
Zn-0.5Pb-0.2Cd
0.7582 0.8248 1.1691 2.7521 2.7224
0.8493 0.9378 1.2578 3.0449 2.8731
Palmerton. Pennsylvania, a mildly industrial environment; data are from Ref. 539.
255
ATMOSPHERIC CORROSION
6 ~-----------------------------'
• Open aIr
5
FIGURE 8.12. Comparison of the corrosion loss of zinc under sheltered and unsheltered conditions in Madrid, Spain. Data are taken from Ref. 306.
Louvered box
~L-----~------~------3 L-----~4--~
Time, years
1217]. The effect of diameter appears to be significant only on wires with a diameter less than a few millimeters [1098]. Flat sheet specimens were reported to corrode more slowly than bent sheet specimens and spherical specimens [217]. Spence and Haynie 1627, 6281 related the effect of sample orientation, shape, and size on the corrosion rates with the changing deposition rates of pollutants. The deposition rate was the highest for fencing wire, followed by small corrosion panels, and was the smallest for large sheets. 8.3.8. Effect of Sheltering
Sheltering prohibits rain from falling directly on the sheltered sample surface and thus can significantly reduce the corrosion rate. In one study, Feliu and Morcillo [306] found that the amount of corrosion of zinc samples exposed in Madrid, Spain, was considerably lower under a sheltered condition than under an unsheltered condition, and the differences seemed to increase with time, as shown in Fig. 8.12. The effect of
3.5 .-----------------------, 3
~
E
:l..2.5 oS
0
Sheltered
•
Unsheltered
2
~ c
.2
~ 1.5
o
u 1 FIGURE 8.13. Corrosion loss for zinc plates exposed for five years in different types of atmosphere under sheltered and unsheltered conditions. Data arc taken from Ref. 1259.
O· : L--...lI........J~r_~~ Rural
Urban
Indus.
Type of Atmosphere
Marine
CHAPTER 8
256
sheltering depends on the type of atmosphere as reported by Johansson and Gullman [1259] and shown in Fig. 8.13. Sheltering (hanging on a rod under a roof) reduced the amount of corrosion by 3.8 times in a rural, 2.4 times in an urban, and 2.2 times in an industrial environment. Interestingly, as shown in Fig. 8.13, in the marine environment the amount of corrosion under a sheltered condition actually increased by 1.4 times. This was explained as possibly due to the fact that zinc chloride is very hygroscopic, and the formation of zinc chloride in the marine environment causes the zinc surface to remain moist for a longer time in a rain-sheltered condition [1259]. However, in another study by Kucera and Mattsson [293] it was found that the corrosion of zinc in a marine environment under a sheltered condition (in a meteorological box with slatted walls) was 2.6 times less than that under an unsheltered condition. On the other hand, the amount of corrosion reduction with sheltering was only 1.25 and 1.85 times in rural and urban environments, respectively. Kucera and Mattsson attributed the significant effect of sheltering in the marine environment to the virtual prevention of transport of NaCl particles and droplets into the meteorological box. The results from these studies indicate that the method of sheltering is also important in affecting the corrosion. It needs to be distinguished whether samples are shielded from rain, wind, orland air for a given sheltering condition.
8.3.9. Galvanized Steel The corrosion of the zinc coating on galvanized steel is generally similar to that of zinc panels. Data from field exposure tests indicate that the time to red rust is largely a linear function of the zinc coating thickness [1098]. Typically, the number of years to reach 50% surface red rust is suggested as the average life of a zinc coating [217]. Figure 8.14 shows that, for a given thickness, the life of the coatings produced by different processes does not vary significantly [1269]. The results of a test program by ASTM [1098] in 11 different geographic locations in the United States over 21 years show that zinc-coated steel wire of different diameters has a coating life directly proportional to the coating thickness, irrespective of the method used to produce the coating. Zinc coating produced by thermal spray has been found to corrode slightly more than that produced by hot-dipping [623, 624]. It was reported that the zinc coating deposited from a cyanide bath had better corrosion resistance than that deposited from a sulfate bath [297].
'"i: ~
15
on
"'"
10
~
5
~
'" ..::.> :J
..
Coating process: sherardizing hot dipping zinc spray a plating
'"
" 50
100
150
Coating thickness (j.tm)
200
FIGURE 8.14. Effect of coating thickness and coating methods on coating life. After Hudson [1269].
ATMOSPHERIC CORROSION
257
TABLE 8.10. Effect of Various Annealing Treatments on the Time for Appearance of Yellow Rust and the Black Layer and the Life of Galvanized Steel Wires" Temperature and time of annealing As received 450°C/30 min 500°CI20 min 550°C/30 min 600°C/20 min 650°Cl30 min
Weeks to appearance
Average iron content(%)
Yellow rust
4.2
113
10.9
44
12.6 15.4 18.0 21.6
46 2
Black layer
147" 75 75 53 75 53
Life (weeks)
159 166 168 221 253 >503
"Ref. 433. "Slight blackening only.
From the time of initial rusting, it takes a longer time for galvanized steel with thicker coatings to become fully covered with red rust [510, 525. 617]. Thicker zinc coating is produced by longer dipping time and has a thicker alloy layer; therefore, it takes a longer time from initial rusting to complete rusting of the galvanized steel. The zinc-iron alloy layers in a galvanized coating are often found to be more corrosion-resistant than the pure zinc layer. Table 8.10 shows that the life of hot-dipped galvanized steel can be substantially increased by a suitable annealing treatment [433], although dicoloration of the coating is accelerated. The life of a coating containing 21.6% iron is more than three times the life of a nonannealed coating. In another study [434], it was reported that coatings containing less than 10% iron are inferior to those containing 20% iron. The corrosion products formed on zinc-iron alloy coatings, e.g., annealed galvanized steel, are yellowish brown in color, reflecting the presence of the corrosion products of iron [433, 394]. Brown stains also appear during atmospheric corrosion of hot-dip coatings when the free zinc layer is consumed or in coatings on steels of high silicon content [219, 394]. Annealed galvanized coating develops yellow rust, which then gradually changes to a uniform dark gray to black color [433]. The black film formed after six years' exposure in an urban environment was found to be about 80 f.1m in average thickness, about three times that on nonannealed samples. The black layer, which is insoluble and adherent, does not completely cover the coating unless the coating itself is rich in iron. The steel substrate can have a significant effect on the galvanizing process. Siliconcontaining steel (or silicon-killed steel), known to galvanizers as "reactive steel," gives rise to the formation of a thick coating containing a high proportion of zinc-iron alloy phases. A I5-year urban atmospheric exposure of six different kinds of silicon-containing steels (0.21-0.45% Si) revealed that corrosion rates are comparable to that of nonreactive steel [219]. However, the iron content in the coatings causes the galvanized silicon-containing steels to show various degrees of brownness (60-100% surface coverage with brown corrosion products), while regular galvanized steel appears mainly gray-white after 15 years of exposure.
258
CHAPTER 8
8.3.10. Effect of Alloying The effect of alloying elements on the atmospheric corrosion performance of zinc is complex. Some elements may have beneficial effects in one situation while having harmful effects in another. Some elements have little effect on atmospheric corrosion but may enhance corrosion when another element is present. For example, trace amounts of lead in zinc, which are otherwise harmless, induces intergranular corrosion when aluminum is present. For galvanized steel, the iron-zinc alloy layers are more resistant to atmospheric corrosion than pure zinc. In general, trace amounts of impurities have very little effect on the corrosion rate of zinc in atmospheric environments [294]. 8.3.10.1. Lead. As a common impurity or alloying element in zinc products, lead has little effect on the corrosion of zinc when synergistic effects with other elements are absent. No significant differences in corrosion rates were found among different grades of zinc with lead concentrations of 0.0055, 0.049, and 0.84% after 20 years of atmospheric exposure [546]. Similar findings were reported in another study for galvanized coatings containing 0.57 and 0.68% lead [1093]. 8.3.10.2. Copper. Copper has been found to have a beneficial effect on the atmospheric corrosion resistance of a galvanized coating. It was reported that addition of up to 0.82% copper increased the corrosion resistance by as much as 20% in a two-year industrial exposure test, as shown in Fig. 8.15 [489]. Adding 1% copper to rolled zinc sheet was found to have little effect on the average rate of corrosion except in a severe marine environment, where the copper-containing sheet showed a higher corrosion rate than unalloyed zinc [549]. This enhancement of the corrosion rate was attributed to the galvanic cell effect between zinc and copper, the latter acting as an effective cathode in the low-resistivity electrolyte formed by the marine environment. Copper-bearing zinc is more likely to develop distinct pits during corrosion than unalloyed zinc [294]. 8.3.10.3. Aluminum. Aluminum is a widely used alloying element for zinc. When present in small quantities, aluminum reduces the atmospheric corrosion resistance with a maximum effect at a concentration of 0.32%, as shown in Fig. 8.15 [489]. Addition of a small amount of copper was found to offset the effect of aluminum at a similar concentration. As shown in Fig. 8.16, with the addition of more than 1% AI, the 200 r-------------------------------~
- 150
N
.€
E:
.c
Ol
~
50
pure
ZinC
coating,
O L--L__~~__~__~_~___ L_ _~~_ _
o
0.1
0.2 0.3 0 .4
0 .5 0.6
0.7
Concentration (wt % )
0.8
0.9
FIGURE 8.15. Coating loss in 2~ years for galvanized steel produced in baths containing different amounts of various alloying elements. After Radeker et al. [489J.
ATMOSPHERIC CORROSION
259 25 ~--------------------------,
-- Sevele marine ...,- Indusui al *
M arine
... Ind u5trial
en en
*
15
Rural
~
FIGURE 8.16. Effect of Al content on corrosion of Zn-AI alloy coatings after 5-yr exposure in various atmospheric environments. After Zoccola et at. [1891.
:::: .3 40 50 60
O ----~~~~L-~--~--~--~--~
o
10
20
30
70
80
Concentration of Al (wI %)
atmospheric corrosion resistance of zinc coatings increases with aluminum content up to 4-7% Al (eutectic composition is 5%), beyond which corrosion resistance decreases with aluminum content to about 21 % (the eutectoid composition being 22% AI). Between 21 % and 70% AI, corrosion resistance increases almost linearly with Al content [189]. Two major commercial zinc-aluminum alloy coatings, Galfan and Galvalume, have been developed for the production of more corrosion-resistant steel sheets. Table 8.11 compares the corrosion resistance of galvanized, Galfan, and Galvalume coatings in various atmospheric environments [242]. In general, Galfan is about two times more corrosion-resistant than a galvanized coating, and Galvalume is two to four times more corrosion-resistant than a galvanized coating. The corrosion of zinc/aluminum alloys proceeds through several stages. First, the zinc preferentially dissolves, leaving an aluminum-rich porous structure. During this stage, the steel is cathodically protected. After the zinc is depleted in the coating, depending on the compactness of the remaining structure and the type of atmosphere, red rust may start to form since the aluminum may be passivated and hence is cathodic to steel. In a chloride environment, the aluminum left in the zinc/aluminum coatings after depletion of zinc can still galvanically protect the
TABLE S.II. Corrosion Losses for Galvanized Steel, Galfan, and Galvalume after One-Year Exposure in Four Different Atmospheric Environments" CorrosIOn loss (um) Galfan Environment
Galvanized
Max.
Min.
Galvalume
Rural Urban Marine Severe marine
1.0 3.0 2.4 5.4
1.0 1.4 2.8 3.8
0.3 0.7 1.4 2.8
0.3 0.6 1.1 2.6
"Ref. 242.
CHAPTER 8
260
140 --
II> II>
pure 2 me COJtlng
80 ~
.2
1: 60 Ol
-Ag
+Sb
'iii
5: 40 ·
*Mg + Bi
20 -
O ~~--~~--L--L--~~--~-L--
o
0.1
0.2 0.3 0.4 0.5
0.6 0.7 0.8 0.9
Concentration (w t %)
1 .1
FIGURE 8.17. Effect of various alloying elements on corrosion of galvanized steel in an industrial environment. After Radeker r489].
steel until all the coating is consumed, since the passive film of aluminum is not stable and aluminum is anodic to steel [189]. 8.3.10.4. Tin. Small additions of tin (from 0.27 to 0.96%) in a galvanizing bath have been reported to have very little effect on the atmospheric corrosion rate of the gal vanized coating [217]. However, Fig. 8.15 indicates that with an addition of 0.3-0.9% Sn, the corrosion rate of zinc in an industrial environment increases by about 20%. Laminar coatings of zinc and tin or zinc alone exhibit equivalent or better resistance to industrial atmospheric corrosion than similar tin-zinc alloy coatings [621]. 8.3.10.5. Other Elements. Figure 8.17 shows the effects of a number of other elements on the corrosion rate of galvanized steel [489, 1093]. A IS-year urban atmospheric exposure showed that addition of 0.08% vanadium slightly increases the corrosion resistance of galvanized coatings [219]. Addition of 0.04 wt. % Mg was found to have no significant effect on the life of galvanized coatings in various atmospheric environments [332].
8.3.11. Effect of Surface Treatment The surface of zinc can be treated to increase the corrosion resistance. The process most commonly used is chromating. However, due to environmental concerns, the use of chromating is becoming increasingly limited. The chromating process involves dipping in a solution containing chromate, resulting in the formation of a conversion coating consisting of chromium oxide and chromate. Chromate treatment is very effective in delaying the onset of corrosion on a zinc surface and in preventing the formation of wet storage stain. However, chromating does not significantly change the long-term atmospheric corrosion performance of zinc, as can be seen in Table 8.12 [593]. Phosphating is another surface treatment extensively used for surface treatment of zinc and its alloys. Because phosphating significantly alters the surface appearance, it is used less as a surface finishing process and more as a pretreatment for painting. Plating a more corrosion-resistant metal or alloy layer on the zinc surface can also be used to increase the service life of zinc products. Copper, nickel, and chromium electroplated coatings have been found to improve the corrosion resistance of zinc
ATMOSPHERIC CORROSION
261
TABLE 8.12. Outdoor Exposure Data on Cronak-Treated Zinc" Months to rusting" Coati ng type
----_.
__.
Zinc thickness (11m)
Untreated
Cronak treated
30 30
30
-----------"----------
Electrodepositcd
Hot-dip
8.6
12.9 17.2
42
2S.R
48
22.R 32.7
54
60
30 .+2 54
60 60 ------
"Ref. 593. "In New York City.
die-casting alloys [188]. Pitting at defects of the coatings is the predominant corrosion feature of these coated products. 8.3.12. Effect of Corrosion Products For most metals, the corrosion resistance is largely determined by the stability, adherence, and compactness of the corrosion products (e.g., oxides, hydroxides, and salt films). In generaL the corrosion rate of zinc after atmospheric exposure decreases with time because of the formation of protective corrosion products on the zinc surface [4, 618]. During the exposure, as the silver-gray surface turns dull gray, a thin adherent surface film (identified most often as zinc carbonate) gradually forms and inhibits further corrosion of zinc. The formation of solid zinc carbonate is a slow multistep process. Zinc is first converted to zinc hydroxide, fonning a gel, which may then be converted by a small amount of carbon dioxide to a tough, thin film of basic zinc carbonate. Under conditions of limited access of air, and hence slow drying and insufficient carbon dioxide, the zinc hydroxide is converted to a fluffy zinc oxide, known as "white rust." Because of the loose nature of the white rust, it has little barrier effect on the access of solution to the zinc metal. Furthermore, the buildup of white rust will prolong the time of wetness by reducing the critical relative humidity for condensation, retaining more moisture, and retarding the drying process. The formation and characteristics of various corrosion products are discussed in Chapter 6. 8.3. J3. Forms of Corrosion The corrosion of zinc in most atmospheric environments is usually general corrosion; that is, the corrosion occurs uniformly across the zinc surface. The corroded surface after years of exposure may be covered with dimples, for which the ratio of depth to diameter is small [2941. The dimple size can be a few millimeters in a marine environment and much smaller (about one-tenth) in a rural environment. Pitting is not a common form of corrosion of zinc in atmospheric environments. Another common corrosion form of zinc is galvanic corrosion. On galvanized steel at places where the coating is damaged, such as at cut edges, the exposed steel is cathodically protected while the surrounding zinc coating is galvanically corroded.
262
CHAPTER 8
Although a common form of corrosion, galvanic corrosion is not a major contributor to the corrosion of zinc coatings because the exposed areas of bare steel are usually too small to cause significant corrosion. Usually, therefore, the atmospheric corrosion rate of galvanized zinc coatings is essentially the same as that of zinc. Galvanic corrosion can be a significant contributor to corrosion when a small piece of zinc is connected to a similar or larger piece of another metal. Table 7.8 in Chapter 7 shows that the corrosion of zinc wire is increased by electrically connecting to bolts of other common metals, with the exception of aluminum and magnesium. The corrosion rate of zinc decreases when it is connected to magnesium in all types of atmospheric environments and to aluminum in urban and marine environments [293, 544, 551]. In other atmospheric environments, zinc is anodic to aluminum, owing to the passive film on aluminum. A zinc rod wired by aluminum is anodic to the aluminum wire in most of the industrial and some of the marine atmospheric exposures [515]. Coatings with more than 60% AI behave like aluminum and provide no galvanic protection to the steel [248]. The galvanic effect is most significant in marine environments because of the high conductivity of surface moisture. The thinner the moisture film on the metal surface, the more localized is the attack at the galvanic couple contact [293]. A detailed discussion of galvanic corrosion is presented in Chapter 7. Premature darkening is a phenomenon sometimes encountered in galvanized roofing sheets, when the darkening, which would otherwise take a few months, occurs very rapidly after a few days of exposure [119, 120]. The characteristics of premature darkening have been reported as the following: 1. The darkening occurs only in rural environments. 2. The darkening occurs within one week of the initial exposure to the atmosphere. (It normally takes a few months to darken the surface.) 3. Only the skyward surface is darkened. 4. Sheets adjacent to the darkened sheets remain bright. The exact cause of premature darkening is not fully understood. The initial surface conditioning during the storage period may be responsible for premature darkening. The darkening may also be due to the particular kind of galvanized coating [119]. Wet storage stain is the voluminous, white, and powdery corrosion product formed when closely packed galvanized articles are stored under damp and poorly ventilated atmospheric conditions [51]. It is most often found on stacked and bundled items such as galvanized sheets, plates, angles, bars, and pipe. Weathered zinc surfaces are seldom attacked. The formation of wet storage stain generally has little effect on the corrosion life of galvanized steels other than changing the surface appearance. It has been observed that when galvanized steel sheets are exposed for six months, the sharp color contrast between the stained and nonstained areas is less noticeable. After two years, there is little difference between ordinary and wet storage stained sheets [53]. More information on wet storage stain is presented in Chapter 7.
8.3.14. Highway Environment The highway environment, experienced by automobiles and highway structures, is particularly aggressive owing to the high pollution level from gas exhaust and, in the
263
ATMOSPHERIC CORROSION
wintertime, from the deicing salts [571]. The salt solution not only increases the corrosivity but also the time of wetness. In the highway environment, zinc is used primarily as a coating material for steel structures such as guardrails, automotive body panels, and rebar. The zinc coating on guardrails is directly exposed in the highway environment. On the other hand, the zinc coating on automotive body panels or rebar is covered with paint or concrete, respectively, and is normally indirectly exposed to the environment. The discussion presented here is limited to the corrosion caused by direct exposure to the highway environment. Information on the corrosion behavior of zinc under paint or inside concrete is presented in Chapters II and 13, respectively. The Society of Automotive Engineers (SAE) conducted an extensive under-vehicle test for various zinc-coated steels [5721. Table 8.13 shows that the corrosIOn performance of zinc-coated steels is superior to that of bare steel or steel coated with other types of coatings. The reduction in weight loss for zinc-coated steel compared to the bare steel is about sixfold. Also, it can be noted in Table 8.13 that a thicker zinc coating is beneficial in reducing the percentage of base metal attack and the pitting depth. As a rough estimate, the corrosion rate of the zinc coatings in an under-vehicle environment, being about 8.5 Jim/yr, is comparable to the corrosion rate in a relatively severe marine atmospheric environment. The corrosion performance of zinc- or alloy-coated steel varies with coating type. In one study, it was reported that, after two-year under-vehicle testing, galvanized and galvannealed steels have less than 5% surface area showing base metal corrosion and pit depth less than 30 Jim. On the other hand, in the case of the electroplated Zn and Zn-Ni coatings, the base metal corrosion is between 50 and 80% and the pit depth is about 80 Jim [3371. Table 8.14 contains data reported by Allegra and Townsend [485] on the corrosion rates of galvanized and 55% AI-Zn-coated steels obtained from under-vehicle tests. The corrosion rate of the zinc coating, varying from 1.1 to 3.3 wn/yr and being higher than that of the zinc-aluminum alloy coating, is comparable to the corrosion rate in a rural environment, which is typically in the range of 0.2-3 Jimlyr [4].
TABLE 8.13. Corrosion of Various Coated and Uncoated Steel Panels after 2.3-Y r Under-vehicle Exposure Test" - - - - - - - - - - - - - - - - - - - - - - - - - - -
Weight loss Sample
h
G90 G60 Anodic electrodeposited primer MOPB Zinc-rich primer Bare steel
Img/cm 2 (jim)]
13.9(19.6) 13.5 (19)
82.7 (ll7)
Average % area of base metal attacked 5.1 16.6 22.8 40.9 58.6 100.0
- - - - - - - - - _... _ -
Average pit depth 8.5 23.1 58.0 93.1 52.0 132.1
-.------------------------------
"Data from Ref. 572. "G90 and GOO. Galvanized zinc coating specified in ASTM Standard AbS3: MOPE. metallo-organic petroleum base coated steel.
264
CHAPTER 8
TABLE 8.14. Corrosion Rates ofZn and 55% Al-Zn-Coated Steel Panels in Under-vehicle Exposure Tests" Corrosion rate (j1m/yrl Sample A B
C D E F
Average
Galvanized
55%AI-Zn
3.33 1.56 1.54 1.4 l.l 1.9 l.81
0.47 0.42 0.21 0.68 0.65 0.24 0.45
"Data from Ref. 485.
The severity of a highway environment varies with location. German [186] reported the results of a 7-yr field exposure test for galvanized steel placed along various highway locations in Ontario and Quebec. It was estimated that the zinc coating life is about 5 yr/mil (corresponding to a corrosion rate of about 5 flm/yr) in an urban highway environment and that about 10-20% longer life is likely in a rural environment. The corrosion of automotive materials is also affected by motion. Talati and Patel [299] found that the corrosion rate of zinc on a moving coach in Bombay and Ahmedabad was several times higher than that under a static condition. The higher corrosion rate was attributed to the falling off of the corrosion product, which has the effect of inhibiting corrosion, from the specimen during movement at high velocity. 8.4. CORROSION IN INDOOR ENVIRONMENTS In normal indoor environments, such as inside residential houses, zinc corrodes very little. Generally, a visible tarnish film forms slowly, starting at spots where dust particles have fallen on the surface [404]. Over a period of time, such films grow gradually until the surface has lost much of its original luster. The appearance of the surface and the degree of corrosive attack are related to the relative humidity. Relative humidity up to about 70% has little influence on the corrosion. Above this point, corrosion activity may occur as it becomes possible for moisture to precipitate on the surface, especially when the surface is covered with zinc corrosion products and contaminants. The corrosion rate of zinc in a clean indoor atmosphere is typically below 0.1 flm/yr. Figure 8.18 presents the testing results obtained by British Steel in 15 residential houses in England over a period of three years, showing that the average corrosion loss of galvanized steel samples over three years is 0.12 flm and the rate decreases with time [1292]. A linear regression analysis of the data in Fig. 8.18 yields an expression for corrosion loss with respect to time of the form: Weight loss = 0.059 t 064
265
ATMOSPHERIC CORROSION
0.4 . - - - - - - - - - - - - - - - - - - - . . . , Carr. loss ~ 0.059 to'"
E 0.3
+
:J..
+
iii II)
E c 0.2 o
'iii
g o
u 0.1
0.5
1.5
2
2.5
3
3.5
Time, years
FIGURE 8.18. Corrosion loss of galvanized steel. exposed in the loft area of 15 residential houses located in three different geographic locations in England, as a function of time. The data are the mean values of 6 samples for each house. The equation in the figure is the best fit from linear regression analysis. Data are taken from Ref. 1292.
Similar low corrosion rates have been found in other indoor situations. In one case It was reported that the average corrosion of galvanized steel on telephone equipment exposed for up to 40 years to the New York City environment was much less than 0.1 j.HnJyr [406J. In another case the weight increase of zinc exposed in a basement room with a window open to the street at an average relative humidity lower than 70% was found to be linear with respect to time, with a rate of about 0.1 f.lrnJyr [746]. As with outdoor corrosion, many factors may affect indoor corrosion. These include the type of climate, thermal insulation, heating, air conditioning, and amount of air contaminants. Corrosion rates higher than the normal values may be found in situations in which moisture precipitates regularly or the air is polluted. Table 8.15 shows that the corrosion rates in industrial indoor environments are in the range of 1 j.HnJyr, much higher than in clean indoor environments [559]. In another case, it was found that in an indoor
TABLE 8.15. Corrosion of Different Alloys in Industrial Indoor Environments for 10 Years" ---------------------------;c------Corrosion rate (/lm/yr)"
Test location Pyrometer shed-ZnO furnace building ZnO furnace building over furnaces Cement crusher building Cement kiln furnace building Coal breaker building
A 0.773 2.06 0.258 0.515 0.515
B
0.773 2.32 0.258 0.515 0.515
C 0.773 2.58 0.258 0.515 0.773
D 0.773 2.58 0.515 0.258 0.773
"Data from Ref. 559. "A, Hot-rolled high-grade zinc; B, Hot-rolled brass special zinc; C, Hot-rolled selected zinc; 0, Hot-rolled zinc plus 1% Cu and 0.0 I % Mg.
266
CHAPTER 8
environment the average formation rate of corrosion products on the exterior surface of a galvanized pipe for cold water was about 0.4 J1.rn/yr over a period of 32 years, while it was only 0.1 J1.rn/yr on a hot water pipe [1293]. It is easier for moisture to condense on a cold surface than a warm surface at a given relative humidity. It has also been reported in another study [613] that no significant corrosion was observed for zinc samples placed above a lead-acid battery electrolyte (concentrated sulfuric acid) reservoir under a nonsealed condition (with a small hole in the cover) in a room for electronic equipment because the normal air flow in the room prevented the generation of concentrated acid vapor. 8.5. CORROSION IN SIMULATED ENVIRONMENTS Atmospheric corrosion is a slow process, usually taking several months or years to show its effects. Thus, laboratory testing methods are used to accelerate the process under simulated conditions. However, because atmospheric corrosion is a complex phenomenon involving numerous variables that are noncontrollable and change irregularly with time, it is very difficult to accelerate and, at the same time, to closely simulate atmospheric corrosion. In many cases, the corrosion rates obtained from a laboratory simulated test bear little resemblance to those resulting from normal exposure. For example, as shown in Fig. 8.19 [547], the corrosion rate of zinc subjected to a spray of natural seawater in the laboratory can be much higher than the rates obtained from real exposure in a marine atmospheric environment. Comparative results shown in Fig. 8.20 indicate that the corrosion rate of zinc greatly varies depending on the kind of test used, with the salt spray (SS) test being the most severe [52]. The accelerating factor of each test is very different for different metals. Thus, caution must be exercised in evaluating and comparing the corrosion rates obtained for different materials from these corrosion tests. For example, the corrosion rate of zinc is only about 2 times lower than that of steel in a salt spray test but is 10 to 100 times slower than that of steel in real atmospheric exposure tests.
13
15
~
00
,s '"'" ..9
10
.E
.!1!l
~
5
~m 500
1000
1500
Exposure time (hours) FIGURE 8.19. Comparison of corrosion of zinc exposed in a marine atmosphere (25 and 250 m from the sea) and in sprays of natural seawater and 3 and 20% NaCl. After Baker and Lee [547].
267
ATMOSPHERIC CORROSION
til til
E
E
0>
' ijj
FIGURE 8.20. Comparison of corrosion of galvanized steel in different accelerated tests: Sp, steam pressure; W film, water film; CC, condensation cabinet; W fog, water fog; SS, salt spray. Data are taken from Ref. 52.
~
0.1
o.ol L0.5
5
50
Time (hours)
However, laboratory experiments can be designed to study the effect of the variables on specific aspects of atmospheric corrosion. The most common tests used to evaluate atmospheric corrosion of a metal are salt spray test, humidity chamber exposure, and wet/dry cyclic test. These test methods have one element in common: they form a thin layer of electrolyte on the metal surface, which is what occurs under real atmospheric exposure. Also, all these tests use enclosed chambers so that the required humidity, spray, and level of pollutants can be generated and controlled. Humidity chamber exposure is a simple test to evaluate the effect of relative humidity and temperature, but it lacks the dynamic effect of raining and drying. On the other hand , salt spray tests simulate the effect of continuous raining but do not represent the chemical composition of rain nor the effect of condensation and drying. The standard salt spray test, ASTM B 117 [259] , although widely used, is a very severe corrosion test and bears little similarity to atmospheric corrosion. The wet/dry cyclic test, incorporating the effect of condensation, spray, and drying, is closer to real atmospheric exposure. Because a large number of variables are involved, the conditions in different cyclic tests can vary greatly. Thin-layer electrolytes can also be used to study the electrochemical reactions and changes in solution chemistry during the corrosion process.
8.5.1. Humidity Chamber Exposure A humidity chamber provides the conditions to create surface wetness by water condensation . Thus, the corrosion inside a humidity chamber is similar to that caused by natural dew. A zinc surface becomes stained when tested in a humidity chamber. The percentage of stained area increases with increasing relative humidity and temperature as shown in Fig. 8.21 [120]. Figure 8.22 shows that the corrosion rate of zinc in a humidity chamber is linear as a function of time [1270]. At room temperature in a closed chamber containing distilled water (relative humidity 100%), zinc is only tarnished after many days of exposure [437]. However, when the zinc article is cooler than the surrounding humid air, large droplets of water may form and cause more corrosion. The presence of pollutants in the air can significantly increase the corrosion rate . Figure 8.23 illustrates that, in a chamber at 21 °C and at a relative humidity between 95
268
CHAPTER 8
60 o
25°C
40
60
FIGURE 8.21. Effect of relative humidity on surface staining of wetted galvanized steel. Reprinted from HelwIg r120], with kind permissIOn from Elsevier Science lnc., 655 Avenue of the Americas, New York.
80
Relative humidity (% )
and 100%, the corrosion rate of zinc increases significantly with increasing sulfur dioxide concentration [1271]. Gilbert and Hadden [437] reported that in a closed tank containing various amounts of saturated S02 solution, the corrosion of a zinc sample, placed horizontally above the solution, increased with increasing S02 concentration in the tank as shown in Table 8.16. The sulfur dioxide appears to facilitate the wetting of the surface and to damage the initial oxide film, which would otherwise maintain its protective properties in pure air [556]. The formation of sulfate compounds as the corrosion products is equally important in promoting corrosion because these compounds are more soluble and more hygroscopic. In moist H2S gas, zinc is very stable in comparison to many other metals, as shown in Fig. 8.24. Thus, a zinc coating can offer effective protection to steel and iron against corrosion in an environment containing hydrogen sulfide [556].
---u E
1.2
Sen en
0.8
'il ~
0.4
M
.3 1: OJ)
97% relative humidity 85% relative humidity
.' 3
6
9
12
Time (months)
FIGURE 8.22. Weight gain versus duration of exposure for zinc in a humidity chamber at two humidities. After Rajagopa1an and Ramaseshan [l270].
269
ATMOSPHERIC CORROSION
SO, volume %
20
. 0
0
00 2;
15
'" :cOIl
10
0.00 0.01 0.05 0.1 0.5
o
'" .S! .;:; ~
b
FIGURE 8.23. Effect of S02 content in air on corrosion of zinc at 100% relative humidity. After Barton and Beranek [12711.
0
b
10
30
20
Time (days)
Askey et ai. [992] studied the effect of fly-ash particulates on the corrosion of zinc in a humidity chamber and found that the fly ash caused a slight increase in the corrosion rate. Coal fly-ash particulates, containing less than 0.3% leachable ionic species. arc generally very much less corrosive than oil fly ashes, with corresponding leachable species contents greater than 1.5%. Liquid-phase catalytic oxidation of SOz by the ionic species leached from fly-ash-bascd particulates is not significant. In addition to relative humidity and pollutants, other factors can also affect the corrosion of zinc in a humidity chamber. It was reported that the corrosion rate of zinc in moving water vapor (40 cm/min) containing 1.8% S02 was high (635 !lm/yr) at the beginning of the test and reached a fairly constant value of 135 {lm/yr after 500 days [499]. Photo-enhanced corrosion was observed when a zinc surface was illuminated at 100% relative humidity, but the enhanced corrosion does not appear to be a dominant factor in the overall corrosion process [331]. The corrosion rate of zinc in moist air can be increased by the presence of a very thin film of noble metals on the surface [835]. TABLE 8. I 6. Corrosion of Galvanized Steel Placed for One Week in Sealed Tanks Containing Various Amounts of SOrSaturated Solution" SO? solution added (mll b
Weioht loss Odor in tank initially Strong Strong Easily detectable
20 8 2
0.25 0.1
o
Just detectable Not detectable Not detectable
Condition of specimens
(m~/cmz)
Large pools of liquid quickly appeared on surface Large pools of liquid quickly appeared on surface Smaller pools of liquid quickly appeared on surface Very fine droplets appeared on surface Only a very little liquid detectable No liquid observed on the surface No liquid observed on the surface
7.72 4.7 1.95
"Ref. 437. b Average
for both sides. Attack was most severe on the upper surface.
0.92 0.15 0.12 O.IS
270
CHAPTER 8
20 .-------------------------,
1 :"5 ~ r o
Zn
AI
C,
Brass
Ni
Cu
FIGURE 8.24. Corrosion of metals in a mixture of air and 5% H 2S saturated with moisture (100% relative humidity) for 35 days. Data are taken from Ref. 556.
Fe
8.5.2. Water and Salt Spray Water spray provides a flow of electrolyte to the metal surface and has the effect of washing away corrosion products and surface contaminants. In a detailed study, Gilbert and Hadden [437] found that in a 100% relative humidity chamber the corrosion rate of zinc sheets, sprayed for 3 s twice a day (5 days/week), was high for the first 5 days and then decreased to a low value, as shown in Fig. 8.25. The water spray, although short in duration, significantly increased the corrosion rate compared to that without spray, as shown in Table 8.17. Table 8.17 also shows that spray in dry air induced much less corrosion than spray in humid air owing to the shorter time of wetness. Table 8.18 shows that CO 2 in ordinary air does not change the corrosion rate in the water spray test. This indicates that the effect of carbon dioxide on the corrosion of zinc is not direct but rather involves a secondary reaction with the zinc hydroxides to form zinc carbonate. When a large amount of CO 2
1.5
Number of limes sprayed
0.5 I
2
I
I
I
8
10
5
14
18
10
22
15
38
28
20
Time of exposure (days) FIGURE 8.25. Weighlloss versus time of exposure to distilled water sprays for galvanized steel. After Gilbert and Hadden [437].
ATMOSPHERIC CORROSION
271
TABLE 8.17. Corrosion Rate of Zinc under Water Spray" Test modeh
Corrosion (pml5 days)
Two WS./day in 100% RH air Two WS./day in lab air (RH about 60%) Two WS./day in 100% RH air + 3 h/day in lab Two 3% NaCI spray in 100% RH air Two pH 4 WS./day in 100% RH air No spray in 100% RH air
1.8 0.09 0.9 3.0 2.0 0.2
"Data from Ref. 437. "Abbreviations:
w.s .. Water spray for three seconds; RH. relative humidity.
is present in air, the corrosion rate in the water spray test is considerably reduced [437]. The corrosion of galvanized steel under continuous water spray was found to be similar to that of galvanized steel dipped in distilled water and then kept in a 100% relative humidity tank. In the same study, it was found that freshly prepared galvanized steel sheet was slightly less resistant to corrosion in the water spray test than that stored for 18 months. Also, electroplated coatings corroded 35% more than hot-dipped coatings [437]. Johansson and Linder [937] used dropping water to simulate rain and found that in the pH range of 4-7 the corrosion rate varied little, but that it was considerably higher at pH 3. The amount of corrosion under the synthetic rain was found to be a linear function of time. Salt spray causes more corrosion than water spray as shown, for example, in Fig. 8.20 and Table 8.17. When zinc is tested in the ASTM salt spray test, using a 5% NaCI solution spray, its surface becomes completely covered with a white corrosion product in less than one hour. Figure 8.26 shows that the zinc coating loss in a salt spray test is almost linear with exposure time [630]. The corrosion rate is about 1000 pm/yr, about 200-1000 times higher than the rates experienced in real atmospheric environments, indicating the severity of the salt spray test. As also indicated in Fig. 8.26, more than half of the corrosion product is washed away by the salt spray, and the amount of corrosion product remaining on the surface become almost constant after a few days of exposure.
TABLE 8.18. Effect of CO 2 in 100% Relative Humidity Air on Corrosion of Galvanized Steel in Air with Two Distilled Water Sprays Daily for a Week" Loss of coating Atmosphere
(llm)
Ordinary air CO 2-free air Air containing 1.5% CO 2
1.85 1.87 0.72
----"--------------_._-----
"Data from Ref. 437.
CHAPTER 8
272
T
Weight loss Corrosion product
;:;
U
"0
c:
rn ~
5 t-
:o~: o
~
~
00 00
+
l001~1~loo1oo200
Exposure Time (Hours)
FIGURE 8.26. Weight loss and amount of corrosion product remaining on the surface as a function of exposure time inside a salt spray chamber. After Zhang [630 J.
8.5.3. Cyclic Test
The conditions in a wet/dry cyclic test can vary greatly according to the wetting and drying methods (condensation, dipping, or spray for wetting, heating or dry air blowing for drying), speed and duration of wetting and drying, and chemical composition of the wetting solution. Therefore, great caution has to be exercised in evaluating and comparing the corrosion rate data obtained from different cyclic tests. Lyon et al. [213] reported that the corrosion rate is about 200 j.lfnJyr when galvanized steel is exposed to a cyclic test with one-hour salt spray (0.35% sulfate and 0.35% chloride solutions) followed by one-hour drying, a factor of20 to 100 times the typical atmospheric corrosion rates [213]. Haynie et al. [185] used a condensation/light -drying environmental chamber to study the effect of different pollutants. Sulfur dioxide and relative humidity were found to be the most important factors among 15 possible direct or synergistic effects on the corrosion of galvanized steel. With a more sophisticated chamber in which solar radiation, dew, rain, and photochemical smog could be simulated, Spence et al. [626] found that the zinc content collected in the condensate is a linear function of S02 content, in agreement with the data from field exposure (Fig. 8.8). It was postulated that for every S02 molecule that is deposited on galvanized steel during periods of surface wetness, one zinc atom reacts to form zinc sulfate (ZnS04 ). Zhang and Tran [1094] studied the effect of cyclic wetting and drying on the corrosion of zinc and steel. They found that on both the zinc and steel surfaces the ratio of weight loss to corrosion product (WIP) decreases with exposure time, indicating that the corrosion products formed initially have the effect of enhancing the ability of the surface to retain corrosion products. Spraying has a strong effect of washing away the corrosion products on the samples, especially for zinc (see Fig. 8.27); the W/P ratio increased with increased spraying time for a given time of wetness. Also the variation of the relative proportion of the time for water condensation and the spraying time for a given time of wetness has a significant effect on the morphology of the corrosion products [1094]. Figure 8.28 shows that hydrogen ions in spray solutions of pH greater than 3 contribute more to the cathodic reaction and less to the dissolution of the corrosion
ATMOSPHERIC CORROSION
273
11
we ight loss
.A - - - - } (
W/P
0.9
~--L-~--~--~~--~--~--~~---J O.4
o
2
6
4
10
8
12 14
16
18
20
Spraying lime (m in. )
FIGURE 8.27. Weight loss and ratio of weight loss to amount of corrosion product remaining on the sample surface for zinc and steel with variation of spraying time. Drying time was 10 min, and total wetting time was 20 min. After Zhang and Tran r10941.
lO r--------------------------------------, 10 N
E u 0, E
__ weight loss
8
8
- - - - corr. product
u
:>
,
"0
a:o
6
t
,
6 "
4~ ~J
steel
4
]
2
.c
.. -
0>
~ o
1
zinc
1.5
2
2.5
3
3.5
4
4.5
5
5 .5
6 6.5
7
0
pH
FIGURE 8.28. Effect of pH of the spraying solution on weight loss and weight of corrosion product after I-day exposure in a wet/dry cyclic test. Cyclic pattern: 5 min of spraying, IS min at 100% relative humidity, and 10 min of drying. After Zhang and Tran r1094].
CHAPTER 8
274
products on the steel surface while the opposite is true on the zinc surface. The results in Figs. 8.27 and 8.28 demonstrate that the cyclic wetting and drying pattern is one of the important factors determining the corrosion rates of zinc alloys and steels. This may explain why the atmospheric corrosion rate of a metal alloy is sometimes quite different at different geographic locations with similar times of wetness and similar levels of air pollutants [539]. The fact that a particular wetting and drying pattern has different effects on zinc and steel might be partially responsible for the different rankings of the corrosiveness of the atmosphere toward the two materials at various locations around the world as shown in Table 8.4. 8.5.4. Thin-Layer Electrolytes A thin-layer electrolyte is an electrolyte having a thickness less than a few millimeters. The corrosion of metals under a thin layer of electrolyte is very different from that in the bulk solution because the diffusion of oxygen is greatly enhanced under a thin electrolyte layer in open air [336, 556]. Also, experimental results have indicated that the solvation capacity of a thin electrolyte layer for the dissolved species is very limited, which can affect the process of formation of corrosion products [183]. In an early study, Gilbert and Hadden [437] reported on the corrosion of zinc under water drops. For a given volume of water, corrosion increased with the number of drops as shown in Fig. 8.29. For the same volume, more drops cause a larger surface area to corrode. In addition, a smaller water drop has a smaller thickness, which facilitates oxygen diffusion and enhances the corrosion. The corrosion generates a band of localized pitting around the perimeter of the drops, which is attributed to the difference in oxygen diffusion at the perimeter compared to the center of a water drop. Valencia et al. [945] found that the formation of various zinc corrosion products can be simulated using different synthetic solutions in an immersion-drying method, originally developed by Pourbaix et al. [258]. Rozenfeld [556] examined corrosion under frequent wetting by periodically immersing the sample in a solution and then exposing it in a chamber of controlled relative humidity and oxygen pressure. The thin electrolyte layer formed by such a method is about 30 J.1m thick. Figure 8.30 shows that the corrosion
20
rr Cl
E
~1+ I 5 L _ - - - ' - _ - - - 1_ _"---_-'-_-'-_--'--_---'_~ o 10 20 30 40 50 60 70 80 Number of Water Drops
FIGURE 8.29. Corrosion caused by equal quantity of distilled water distributed on galvanized steel surface as various numbers of drops for one week. Total volume of water was 70 ml. After Gilbert and Hadden [437].
275
ATMOSPHERIC CORROSION
4,------------------------------,
FIGURE 8.30. Effect of wetting frequency (number of wettings per hour) on zinc corrosion (oxygen consumption) and comparison to zinc corrosion resulting from immersion in bulk solution. Duration of test was 6 hr. Data are taken from Ref. 556.
... 0.5 N NaCI x 0.5 N Na,SO. /'
o
Bulk immersion
1
2
3
4
5
6
7
8
9 10 11 12 13 14
Number of Wettings
rate of zinc under frequent wetting is many times higher than that for the fully immersed condition and that it increases with the frequency of wetting. In this figure, the corrosion rate under frequent wetting is also seen to be higher in sulfate solution than in chloride solution. The relative humidity of the surrounding air determines the retention time of the thin electrolyte on the sample surface and, hence, the corrosion rate. It also affects the corrosion features of zinc. At lower relative humidity, corrosion is more uniform. The effect of temperature on the corrosion rate under a thin-layer electrolyte was attributed by Rozenfeld [556] to, among other things, changes in the diffusion-layer thickness resulting from changes in convection and solubility of oxygen. Mansfeld and Tsai [210] measured the weight loss of zinc plates covered by a O.5-mm-thick electrolyte of various chemical compositions after drying them out at various relative humidity values. The corrosion rate was much higher under a thin electrolyte compared to that for bulk immersion in O.OIM NaCl or O.OIM Na2S04 solution. However, the converse was found for O.OIM Hel or O.OIM H2S04 solutions. In 8r-------------------------------~
6 Cl
E U) U)
24
FIGURE 8.31. Weight loss for zinc under 0.5-mm thin layer of distilled water in air of different humidities. Data are taken from Ref. 210.
o o
__ Bulk immersion I
I
10
20
30
40
50
Relative Humidity (%1
60
70
80
276
CHAPTER 8
16,------------------------------, 14
~ 12
'" .210
6
4L-__ -4
~
-3.5
__
_ L _ _ ~_ _ _ _ ~_ _ _ L _ _ _ i _ _ ~
-3
-2.5
-1.5
-2
-1
Log c
-0.5
FIGURE 8.32. Weight loss for zinc under thin layers of Na2S04 solutions of different concentrations. Reprinted trom Mansfeld and Tsai r210]. with kind permission from Elsevier Science Ltd. The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom.
the salt solutions the main oxidizing agent is oxygen, the concentration of which near the surface is a function of diffusion-layer thickness (electrolyte thickness), whereas in the acidic solutions it is hydrogen ions, the amount of which depends on the volume of the solution. Figure 8.31 shows that the corrosion of zinc increases with increasing relative humidity since the time of wetness is longer at a high relative humidity. At a given relative humidity, the corrosion under a thin-layer electrolyte increases with increasing concentration of the electrolyte, as shown in Fig. 8.32. Stiles and Edney [183] reported that corrosion of Zn under a 0.3-mm thin electrolyte increases with the initial concentration of H+ (Fig. 8.33), independent of the kind of acid.
0.4
-
v
HNO" pH = 3
o
><
0.3
+'
N(:i
~
0.2
(:i
o
.~ o.l
g 8
g N
0.1
HN0 3 , pH = 4
rg
El
8
~~/~o~o--------------~------ j
HN03 • pH = 5
,
500
1000
Residence time (8)
1500
FIGURE 8.33. Zinc concentration as a function of time in thin-layer HN0 3 solutions of different H+ concentrations. From Stiles and Edney [183]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with pennission.
ATMOSPHERIC CORROSION
277
250
g OJ
~
150
E
t
8
100 50 25 5
FIGURE 8.34. Effect of electrolyte film thickness on cathodic polarization of zinc in a 0.1 N NaCl solution. Data are taken from Ref. 556.
~
__-ULL____~____L-~
-0.8
-1
-1.2
-1.4
Electrode potential (V",)
The pH of the thin-film electrolytes increased in less than 100 seconds to about 6.7. The ratio of the final [Zn 2+] to the initial [Wj was about 0.48, which is consistent with the 1:2 stoichiometric ratio ofZn to H required for the corrosion process in an acidic environment. At the end of the test, the thin-layer electrolyte was found to be saturated with zinc ions. Data from electrochemical measurements reveal that the cathodic current of zinc under a thin-layer electrolyte is greatly increased with decreasing thickness of the electrolyte owing to the enhanced diffusion rate of oxygen, as shown in Fig. 8.34 [336, 522,556]. The effective diffusion-layer thickness under a thin electrolyte layer is about 0.3 mm, which is smaller than that in a bulk solution because under a thin electrolyte layer the diffusion is also enhanced by convection caused by evaporation [556]. The anodic polarization behavior of zinc under a thin-layer electrolyte is similar to that in the bulk electrolyte. The reduction of dissolved S02 in solution was also found by Rozenfeld [556] to increase under a thin-layer electrolyte. Instead of being a reducing agent as is commonly assumed, S02 shows oxidizing properties and acts as a powerful cathodic depolarizer. The galvanic action of a metal couple under a thin-layer electrolyte differs also very significantly from that in bulk electrolyte. Because of the difference in polarization resistances between the anodic and the cathodic reactions, the potential and galvanic current distributions over a coplanar zinc/steel couple under thin-layer electrolytes are very different for zinc and steel, as shown in Figs. 7.4 and 7.5 in Chapter 7. Because of this kind of potential distribution, when a piece of galvanized steel is connected with a piece of bare steel under a thin-layer electrolyte, the corrosion will only occur at the edge of contact and will proceed in the direction away from the contact line. Walter II 079] found a correlation between the corrosion current, electrochemical impedance, and weight loss of zinc for a zinc/steel galvanic couple under a thin water layer (0.5 mm thick).
278
CHAPTER 8
8.6. CORROSION MECHANISMS Zinc corrodes very slowly in clean dry air at room temperature with the formation of an oxide film. On a cleaved plane of a single crystal, an amorphous film is slowly produced [404]; after a few weeks, it reaches a thickness of approximately 100 A. It is amorphous on the surface but crystalline in the interior. If the zinc surface is anodically polished, the film thickens faster and reaches a thickness of a few hundred angstroms in several days. The oxide films prevent further oxidation of the metal. Unlike oxidation in dry air, the oxidation under moisture is of an electrochemical nature. When in contact with condensed clean moisture in the form of rain, mist, or dew, zinc corrodes with the formation of zinc hydroxide according to the following reactions: Zn + 20W
~
Zn(OHh + 2e-
(8.1 )
(8.2) and/or (8.3)
Over a certain period of time, ranging from a few days to a few weeks, this hydroxide layer then dehydrates to form oxide or reacts with the carbon dioxide dissolved in the water to form the relatively less soluble zinc carbonate [173, 331, 437]: (8.4)
or (8.5)
or 5Z00 + 2C0 2 + 3HzO ~ 2ZnCO y 3Zo(OHh
(8.6)
The formation of a carbonate film slows down the corrosion rate. In the early months of exposure, when the carbonate film is forming, the corrosion rate is relatively high, but it gradually reaches a lower constant value as the formation and dissolution rates of the carbonate film become equal. Thereafter, the corrosion is determined by the rate of chemical dissolution of the carbonate film, resulting in a linear relationship between the atmospheric corrosion and exposure time. The atmospheric corrosion of hot-dip galvanized sheet occurs in three distinct stages: (1) a short initial period during which a protective surface layer is formed; (2) a long period of corrosion of the zinc coating; and (3) corrosion of steel on the exposed area where the zinc coating has been consumed. As corrosion proceeds, the surface changes in appearance from silver-gray (zinc) to gray-white (corrosion product of zinc) to brown-gray (corrosion product of zinc-iron alloy phases) to red rust (corrosion product
ATMOSPHERIC CORROSION
279
of steel) [550]. Regular galvanized steel is gray-white in appearance for most of the life of the coating because unalloyed zinc constitutes the main pan of the coating. Regular galvanized coatings consist of several layers of different ZnlFe compositions and structures. These layers, from the free surface to the steel surface, are free zinc followed by (, J, and Tintermetallic phases [3941. Initially, the free zinc layer corrodes. This is followed by the corrosion of the ( layer, which is more corrosion-resistant. At some local sites. the ( crystals may break away, leaving the J layer, and. subsequently, the Tlayer, exposed to attack. In the meantime, the surface becomes more and more brown because of the formation of iron oxides. When the T layer is consumed initially at a localized area, the steel base is attacked, leading to the formation of brown voluminous corrosion products, which protrude out of the zinc layer as brown pimples. However. because the zinc surrounding the pimples provides cathodic protection, extensive corrosion of the steel at these pimples does not occur until most of the zinc coating around the pimples is consumed. The whole atmospheric corrosion process follows two irregular cycles: on the one hand, the cycle of wetting (dew or rain) and drying (radiation, temperature, or wind) and, on the other hand, the cycle of zinc dissolution, hydroxide formation, carbonate formation, and dissolution of the carbonate film. The interaction between these two cycles determines the panicular corrosion rate of zinc in a given atmospheric environment. The various parameters involved in atmospheric corrosion affect the corrosion of zinc by changing the patterns of these two cycles. The amount of water per unit area in dew is small, and this water can be quickly saturated with dissolution products to form precipitates. Salt saturation does not easily occur in the surface water formed by rain owing to the flowing nature of rainwater. Therefore, while the length of the wetting time will determine the amount of corrosion within one cycle, the form of wetting determines the amount of corrosion products retained on the surface. The speed and the extent of drying will determine the composition as well as the compactness of the corrosion products. since the formation of zinc oxide and zinc carbonate involves dehydration. Long periods of wetting with infrequent drying result in relatively more corrosion and formation of less compact corrosion products. For the same wetting time, wetting with much rain will result in more dissolution of the carbonate film and less accumulation of corrosion products. while wetting with much dew and little rain will be more likely to result in a pileup of corrosion products. The corrosion process and the formation of corrosion products can cause condensation at a lower relative humidity and increase the absorption of moisture as well as pollutants. Under a thin-layer electrolyte, dissolution and precipitation are two fundamental kinetic processes in the corrosion of a metal. They. in turn, are functions of two chemical variables, acidity and ionic strength, which are determined by the atmospheric conditions. In a nonpolluted atmosphere, hydrogen reduction is not the major cathodic process owing to its high overpotential on pure zinc, and the corrosion rate of zinc is normally controlled by the reduction of oxygen. Under thin-layer electrolytes, such as those formed by rain and dew, the rate of oxygen reduction is greatly increased compared to that in bulk electrolyte because of the thinner diffusion layer. As a result of the cathodic reactions, the pH of the electrolyte increases, reducing the solubility of the corrosion products and thus facilitating solid precipitation. Carbon dioxide is another air constituent that affects the corrosion of zinc. With a concentration within the range of 300-500 ppm in the
280
CHAPTER 8
atmosphere, it equilibrates with the rainwater to form a buffered solution of pH approximately 5.6 [631]. Although carbon dioxide is beneficial in forming the protective zinc carbonate film, the dissolution of much CO 2 in moisture may lower the pH to a range in which the carbonate is not stable and can be dissolved [331, 627]. Abnormal corrosion rates of zinc occur when there are considerable amounts of pollutants present in the air. The major effects of pollutants are increasing the time of wetness, increasing the solubility of zinc oxide and carbonate, decreasing the pH of the wetting solution, and contributing directly to the cathodic process. Low-pH moisture, such as acid rain, is very corrosive to zinc because the protective zinc carbonate film cannot form. Laboratory testing has shown the existence of a stoichiometric relation between the amount of zinc dissolution and the concentration of hydrogen ions in a solution [183, 626]. Salt near the sea and sulfur dioxide in urban and industrial areas are two pollutants that can cause a significant increase in the corrosion rate. The presence of both salt and sulfur dioxide increases the time of wetness and the solubility of zinc in the wetting solution. However, unlike sulfur dioxide, salt does not enhance the cathodic reaction process, which may explain why much more salt, as compared to S02' is needed to cause a comparable amount of corrosion and why the atmospheric corrosion in unpolluted marine environments is much lower than that in heavily polluted industrial areas. There are two different theories on the role of sulfur dioxide in the corrosion process of metals. The commonly accepted one involves the acidification of moisture through hydration and oxidation of sulfur dioxide to sulfuric acid and the reduction of hydrogen ions in the moisture to promote the corrosion of zinc [331, 626, 629]: (8.7)
(8.8) (8.9)
(8.10)
and (8.1l)
The entire process can be considered as corrosion under an acid rain, in which hydrogen acts as the cathodic depolarizer and the sulfate ions promote the solvation of the dissolved zinc ions. The deposition of S02 on zinc surfaces has been found to be enhanced by the presence of 0 3 and N0 2, which also oxidize surface sulfite formed during the corrosion [940]. The theory based on the reactions in Eqs. (8.7)-(8.11) agrees with a number of experimental observations. Data from both laboratory testing and field exposure indicate a one-to-one stoichiometric relation between Zn 2+ and SO~- in the runoff solutions [626]. Furthermore, zinc sulfates are usually found in the corrosion products of zinc in almost all types of atmospheric environments [173, 297, 331, 437].
ATMOSPHERIC CORROSION
281
Another theory, originally proposed by Rozenfeld [556], suggests that S02 acts directly as a cathodic depolarizer: (S.12) (S.13) (S.14 ) This theory is supported by the fact that the whole process of reduction of S02 is enhanced under thin-layer electrolyte conditions. McLeod and Rogers [499] observed that the formation of sulfur is the first reaction to occur on a zinc surface that is exposed to moist air containing S02, while the formation of 2ZnSO,·5H 1 0 and then ZnS0 4 ·H 20 occurs more than one month later, indicating that the reduction of S02 is an important contributor to the corrosion. They also observed the formation of sulfide and thiosulfate in deaerated sulfurous acid as a result of the corrosion of ZInC. However, Fiaud [291] pointed out that the S02 concentration (0.1 %) in Rozenfeld's experiments is much higher than that in real atmospheres (around 10-4 %) and that the reduction reaction at very low levels of S02 may be very different. It was reasoned that under a thin-layer electrolyte the reduction of S02 at atmospheric pollution levels is greatly affected by the presence of oxygen. While laboratory electrochemical and analytical experiments indicate the depolarizing effect of S02, which should lead to the formation of sulfur species of lower valence, only sulfates have heen reported to exist in the corrosion products of zinc from field exposure [173, 331]. This indicates either that the reduction of SOl does not make a significant contribution to corrosion in field exposure or that the products of the reduction reaction are not stable and are oxidized eventually to form sulfates.
9 Corrosion in Waters and Aqueous Solutions 9.1. INTRODUCTION Waters are commonly classified as pure water (e.g., distilled water or deionized water), natural fresh water, and seawater. Waters containing artificially introduced salts are usually called aqueous solutions. Zinc-coated steel articles and structures, such as galvanized tubing and water tanks, are commonly used in waters. They are not usually used in aqueous solutions in practical applications. However, a large number of studies on the corrosion of zinc have been carried out in solutions, primarily (a) to simulate a corrosion phenomenon, (b) to accelerate a corrosion process, or (c) to conduct electrochemical measurements. The corrosion data presented in this chapter are organized into three main sections: (i) pure water, (ii) natural water, and (iii) aqueous solutions. The corrosion rates in aqueous solutions are largely limited to the data obtained with gravimetric methods. The data obtained by electrochemical methods and the general electrochemical behavior of zinc electrodes in solutions are discussed in Chapter 2, 3, and 5. In addition, the information on corrosion forms, which are also mainly studied in solutions, is presented in Chapter 7. 9.2. CHARACTERISTICS OF WATERS 9.2.1. Fresh Waters
Waters contain solids, gases, and sometimes colloidal or suspended solid matter. The concentrations of dissolved substances are relatively low but vary considerably from one source to another. Even distilled water, depending on its aeration condition, can contain varying amounts of oxygen and carbon dioxide. The important constituents in water can be classified as follows: (1) dissolved gases (e.g., oxygen); (2) mineral constituents, including calcium and sodium salts, salts of other metals, and silica; (3) organic matter, including that of animal and vegetable origin; and (4) microbiological life forms [558]. Among the dissolved gases, oxygen is probably the most significant constituent in relation to the corrosion of metals, owing to its cathodic depolarizing effect. In surface 283
284
CHAPTER 9
water, the oxygen concentration approaches saturation. The solubility is slightly less in the presence of dissolved solids than in pure water, but this effect is not very significant in natural waters containing less than 1000 ppm dissolved minerals. Dissolved carbon dioxide is also very important; however, its effect must be considered in relation to other constituents, especially calcium hardness. The amounts of dissolved air and oxygen in water as a function of temperature are listed in Chapter 2 in Table 2.10 [496]. The principal ions found in waters are calcium, magnesium, sodium, bicarbonate, sulfate, chloride, and nitrate. The hardness of water is usually referred to the calcium carbonate content. Water containing less than 50 ppm of CaC0 3 is considered soft, and water containing more than 150 ppm is considered hard [558 J. CaC0 3 tends to precipitate on the surface of the water container to form scale. The amount of scale is less with waters of high carbon dioxide concentration because of the higher solubility of calcium carbonate. Table 9.1 shows the constituents of typical waters. Typical resistivities of different waters are shown in Table 9.2 [659].
9.2.2. Seawater The most characteristic feature of seawater is its high salt content. The salt content of open-sea water, away from inshore influences such as melting ice, freshwater rivers, and areas of high evaporation, is quite constant and is roughly equivalent to that of a 3.5% solution of sodium chloride. As a result of its high salt content, seawater has a very low resistivity compared to other waters, as shown in Table 9.2. Table 9.3 shows the major constituents of seawater. The average temperature of the surface water of the oceans tends to vary directly with the latitude, ranging from about -2°C at the poles to 35°C right on the equator.
TABLE 9.1. Typical Analyses (ppm) for Natural Fresh Waters"
Water Soft lake water Moderately soft surface water Slightly hard river water Moderately hard river water Hard borehole water (chalk formation) Slightly hard borehole water containing sodium bicarbonate Very hard underground water aRef.558. b Also 51 ppm nitrate (NO,).
Alkalinity to methyl Total Calcium orange hardness hardness pH value (CaC0 3 ) (CaC0 3 ) (CaC0 3)
Sulfate (S04)
Chloride (CI)
Silica (Si0 2)
Dissolved solids
6.3 6.8
2 38
53
5 36
6 20
II
0.3
33 88
7.4 7.5
90 180
120 230
85 210
39 50
24 21
3 4
185 332
7.1
250
340
298
17
4
7
400"
8.3
278
70
40
109
94
12
620
7.1
704
559
451
463
149
6
1670
10
5
Trace
CORROSION IN WATERS AND AQUEOUS SOLUTIONS
285
TABLE 9.2. Resistivities of Various Types of Water" Resistivity (Q'cm)
Type of water
20.000,000 500,000 20,000 1-5000 200 30 20-25
Pure water Disti lied water Rainwater Tap water River water (brackish) Seawater (coastal) Seawater (open sea)
"Approximate values; data are trom Ref. 1105.
The amount of dissolved oxygen in seawater is a function of temperature. The amounts dissolved at equilibrium are tabulated below [563]: T(°C) Dissolved O2 (mill)
-2
o
8.52
8.08
10 6.44
5 7.16
IS 5.86
20 5.38
30 5.42
Figure 9.1 shows the temperature, salinity, pH, and oxygen concentration as a function of depth in the sea. The concentration of oxygen decreases with increasing depth up to about 2000 ft but then increases slightiy with further increases in depth. The concentration of dissolved oxygen is also affected by the degree of water movement and by the amount of biological activity. Photosynthesis increases the oxygen concentration. while some bacterial activities can reduce it to zero [659J. The metabolism of some bacteria produces hydrogen sulfide (H 2S), ammonia. and other nitrogenous compounds; H2S concentrations of 30-35 ppm are quite common in seawater. The pH of surface seawater, in equilibrium with carbon dioxide in the atmosphere. normally lies between 8.1 and 8.3, because of the existence of excess amounts of basic radicals, mainly carbonates, but may fall to 7 in stagnant basins [659]. The pH decreases
TABLE 9.3. Major Constituents of Seawater,,·h Species
Concentration (ppt)
Chloride (Cn Sulfate (SOh Bicarbonate (HCUjl Bromide (Br-) Fluoride (F-) Boric acid (H oB0 3 ) Sodium (Na +)
Magnesium (Mg 2+) Calcium (Ca2+) Potassium (K+) Strontium (Sr 2+) "Ref. 563. "Chlorinity, 190/r; density at 20°C. 1.0243.
18.98 2.65 0.14 0.065 0.0013 0.026 10.56 1.27 0.40 0.38 0.013
286
CHAPTER 9
Oxygen, mil I 2
3
4
5
6
1,000
~I
~ :/~~
'I i~
~...,., ~
2,000
~
-..
.c.
I
~
a;
~I
3,000
~
~I
a.
I
0
~
~ r-- PH
4,000
~
5POO
~ 6pooO
~
2
4
6
~
I~ I
~
I
~
~~ ~
~
~~ I
~
12
8
14
16
18
342
34.4
34.6
34.8
7.6
7.8
8.0
82
Temperature, C 33.0
332
33.4
33.6
33.8
34.0
Salinity, ppt 6.4
6.6
6.8
7.0
7.2
7.4
pH FIGURE 9. I. Oceanographic data taken in the Pacific Ocean at a site west of Port Hueneme. California. After Fink and Boyd [660].
with depth as shown in Fig. 9.1. The presence of carbon dioxide also affects the formation of scales. However, precipitation of calcium carbonate in seawater does not occur as readily as in fresh water since the solubility of calcium carbonate in seawater is about 530 times that in fresh water [659]. 9.3. CORROSION IN PURE WATER The corrosion rate of zinc in distilled water varies widely, ranging between 15 and 150 f1rn1yr [217, 559, 654]. The degree of distillation or deionization has negligible effect
on the corrosion rate. In one study, about the same corrosion rate of zinc was found in deionized water with resistivities of 1 MQ'cm and 18 MQ'cm [1263]. The corrosion rate depends strongly on the amount of dissolved oxygen and carbon dioxide as shown in Table 9.4 and Fig. 9.2 [400,559]. According to Kenworthy and Smith
287
CORROSION IN WATERS AND AQUEOUS SOLUTIONS
TABLE 9.4. Effect of Oxygen on the Corrosion of Zinc in Distilled Water".!' - - - - - - - - _.... _ - - - - - - - - - - - - - -
Test condition Boiled distilled water: specimens immersed in sealed flasks Oxygen bubbled slowly through the water
Temperature (OC)
Corrosion rate (pm/yr)
Room 40 65 Room 40 65
25.4 48.3 83.8 218.4 348.0 315.0
"Ref. 559. hHigh-grade zinc specimens, in duplicate, immersed for 7 days. The corrosion rate was calculated after removal of corrosion products.
[400], the form of corrosion in distilled water changes from pitting to unifonn attack with increasing carbon dioxide concentration. Figure 9.3 shows that the corrosion rate of zinc in distilled water increases only slightly with temperature up to about 50°C, then increases quickly with temperature, reaching a maximum at about 65°C, before decreasing [412]. According to Cox [4121, the sharp increase in corrosion rate from 50 to 60°C may be attributed mainly to an abrupt change in the nature of the corrosion products from being protective to being nonprotective, leading to a sharp increase in the corrosion rate. At room temperature the corrosion products precipitate in the form of a gel with an indefinite amount of absorbed water. When the precipitate is heated, it gradually loses water and undergoes changes in its physical characteristics. As shown in Fig. 9.4, Grubitsch and Illi [707] similarly found that a peak in the corrosion rate occurs at around 60°C and reported that the presence of a sufficient amount of oxygen is required for the occurrence of the peak. 0.2 .---------------------------------------, >-
'"
~
5 0.1 6
0, E
+
Hard Supply Wator
-0- Distillad Water
-g 0.12 ".
"0 (/)
III
Ci
u .::
0.08
N
"0
:E 0.04 Ol
~
6
12
18
24
30
36
42
Free Carbon Dioxide Content of Water, ppm
FIGURE 9.2. The effect of free carbon dioxide in distilled water and hard supply water on the dissolution of zinc (immersed for 56 days at 18°C). After Kenworthy and Smith [400].
288
CHAPTER 9
3 ,-------------------------------, >
E 2.5 E
Qi
m
2
a:: c 0
en
g
1.5
0
U
I1J Cl
ro
~ 0.5
<{
oL---~======~~~----~--~~
o
20
40
80
60
100
Temperature, ·C
FIGURE 9.3. Corrosion of zinc in airsaturated water as a function of temperature. The test samples were rotated at a speed of 56 rpm . After Cox [412J.
In distilled water at room temperature and open to air, zinc corrodes with the formation of pits [400-402, 654]. The formation of pits depends on the oxygen content; when the water is depleted of oxygen, there is little corrosion, and when oxygen pressure is high, the corrosion is of a uniform type. The corroding area spreads with time as the surface gradually becomes covered with a thick layer of hydroxide and carbonate. In a few weeks, the specimen is completely covered with the white material. On removal of this layer, pits are found on the surface of the metal. 9.4. CORROSION IN NATURAL WATERS 9.4.1. Cold Fresh Water
In general, the corrosion rate of zinc is lower in hard water than in soft water or distilled water [400, 217]. An example is shown in Fig. 9.2. This lower rate is largely
0.4 . ------------------------------------ ,
+ Ul"1de-r PUt' 0.32 NE
*
1
Under aif, CO 2 ,,••
+ Under Nl.+ O.6vol% 02 . . U nd~r pvre 0 2
~
~0 . 24
.2
.E 0 .1 6 Cl
·w ~
0.08
oL-____ ____ ____ o 20 40 ~
~
~
______
60
Temperature, ·C
L __ _ _ _
80
~~
100
FIGURE 9.4. Effect of several gases on the temperature dependence of zinc corrosion in hot distilled water. After Grubitsch and IIIi [707].
CORROSION IN WATERS AND AQUEOUS SOLUTIONS
289
TABLE 9.5. Corrosion Rates of Zinc and Zinc Coatings Immersed in Various Industrial and Domestic Waters a Type of water Mine water, pH 8.3, 110 ppm hardness, aerated Mine water, 160 ppm hardness, aerated Mine water, 110 ppm hardness, aerated Demineralized water River water, moderate soft River water, moderate soft River water. treated by chlorination and copper sulfate River water, treated by chlorination and copper sulfate Tap water, pH 5.6, 170 ppm hardness. aerated Spray cooling water, chromate treated. aerated Hard water Soft water
Corrosion rate (jlm/yr) 31 30
46 137
97 61" 81
64b 142 IS 16 15
"Refs. 203 and 217. "Galvanized steel.
attributed to the formation of a protective scale in hard waters. As shown in Table 9.5, the corrosion rate can vary significantly, from as low as 8 j1m/yr to as high as 140 j1m/yr, in different waters. Dissolved carbon dioxide in hard water generally increases the corrosion rate but has less effect than in distilled water (Fig. 9.2). Also, the corrosion of zinc in hard supply water is much more uniform than that in distilled water, in which pitting usually occurs. This was attributed by Evans [401] to a more effective ionic exchange in supply waters, which prevents the localization of corrosion activities. According to Kenworthy and Smith [400], in waters oflow carbon dioxide content, the calcium bicarbonate in the water can precipitate as a protective carbonate scale, but at high carbon dioxide contents, this precipitate will not easily form, owing to the lower pH. and the dissolution of zinc proceeds unhindered. Flowing water causes more corrosion than still water. Fujii [561] found a linear relation between the corrosion of a zinc coating and the cubic root of the flow velocity in tap waters, as shown in Fig. 9.5. This figure also shows that, at a given flow rate, the corrosion rates in the tap water of Berlin and Dortmund were many times higher than in that of Tokyo. 9.4.2. Hot Fresh Water Galvanized tanks have been widely used to store hot water and can usually last as long as 30 or 40 years. However, these tanks have also been found to have a very short life in some supply waters [655,656]. Among the many factors affecting the performance of galvanized steel water tanks, temperature and composition of the water appear to be the most important [707, 400, 412, 458, 709]. The area-averaged corrosion rate in hot hard water is usually low. Kenworthy and Smith [400] reported that the corrosion rate of zinc in 75°C hard supply water, free of
290
CHAPTER 9
4
•
Tokyo
~
U
;;;-
--E
3
OJ
ai
"§
c 0 -iii
2
e 0
0
0
0
0 .2
0.4
0_8
0.6
1.2
Flow velocity, (m/s)'/l FIGURE 9.5. Effect of flow velocity on the corrosion rate of zinc-coated steel tubes in tap waters. After Fujii [5611.
carbon dioxide, is lower than I jJ.m/yr. Dissolved carbon dioxide has a significant effect on the corrosion rate, as shown in Fig. 9.6 [400]. Gilbert [709] found that the corrosion rate of zinc over a five-month period in hard supply water saturated with 1.5% (about 14 ppm) CO 2 at 85°C was between 25 and 43 jJ.m/yr. 9.4.2.1. Pitting Corrosion. Pitting is a common form of corrosion for zinc in hot supply water. It is, in general, much worse on pure zinc than on galvanized specimens; the zinc-iron alloy layers in galvanized coatings have greater resistance to the pitting
-+ C0 2 free
5
-- 5 . 6 ppm CO2
E '-' Oi
E -0 Q)
>
-10.2 ppm
4
-
20.2 ppm
.... 35 ppm -<>-62 ppm
3
'0 (/) (/)
'6
2
c'-'
N
' _~L- ~ ------------------------------__~__~__L -_ _L-~~
O~~--~---L---L
o
20
40
60
80
100
120
140
160
180
Duration of test, days FIGURE 9.6. Effect offree carbon dioxide on the dissolution of zinc in hot hard supply water. After Kenworthy and Smith [400].
CORROSION IN WATERS AND AQUEOUS SOLUTIONS
291
attack [400, 709 J. During a corrosion test, as described by Gilbert [709], bubbles were found to adhere to the surfaces at the beginning of the test, and "bubble cups" of corrosion product formed at some of the points of attachment. These bubble cups consist of white zinc corrosion products, and in some cases several bubbles adhere in close proximity such that fairly large areas covered with white zinc compounds are produced. When failure occurs, it is always beneath this white corrosion product. Bonilla [686] reported that although the total amount of corrosion in hot water is much less on galvanized steel than on black steel, pitting may be more severe on galvanized steel. In hot water, the zinc surface may be passivated, leading to a polarity reversal of galvanized steel. When pits are initiated, they may penetrate through the coating and into the steel substrate because of the galvanic action between zinc and steel. The pitting rate in hot water can thus be quite high, especially when carbon dioxide is present. The presence of copper in hot water also enhances pitting. Campbell [257] reported that, in tap water with 0.1 ppm copper and 28 ppm free carbon dioxide at 75°C, a galvanized coating showed numerous pits, extending into the steel to a depth of 0.3-0.4 mm after 33 weeks. The general theories on pitting and polarity reversal are discussed in Chapter 7. Intergranular corrosion is sometimes found to occur on zinc alloys in hot water, especially on zinc-aluminum alloys. It has been established that intergranular corrosion of zinc alloys in hot water occurs in alloys with more than 0.03% Al [48,491. The intensity of intergranular corrosion is increased by the presence of small amounts of impurities such as Ph, Sn, In, Cd, and Hg. The tensile strength of zinc-aluminum alloys can be significantly reduced as a result of intergranular corrosion [225 J. Detailed information on the intergranular corrosion of zinc alloys is presented in Chapter 7. 9:4.2.2. Effect of Dissolved Copper. The presence of trace amounts of copper in water can substantially increase the corrosion of zinc. Kenworthy [737] found that the amount of copper dissolved in hot water increases with increasing carbon dioxide concentration. As little as 0.1 ppm copper causes a definite increase in corrosion rate. With amounts of copper up to about 0.3 ppm, the amount of corrosion is proportional to the amount of copper. The copper appears to deposit as small metallic particles on the surface of the zinc. Enhanced corrosion occurs because of the larger cathodic activity generated by the copper particles. 9.4.2.3. Other Factors. The flow of water over a zinc surface has a significant effect on its corrosion. Nielsen and Y ding [720] found that galvanized pipes in which hot water circulated continuously showed more coating corrosion than pipes without circulation. Also, pitting corrosion was observed to be more severe on the bottom of the pipes. In another work, Weast and Shulman [226] reported that in pressurized waters the corrosion rate of zinc as a function of temperature does not pass through a maximum as it does in nonpressurized water.
9.4.3. Seawater 9.4.3.1. Corrosion Rate. The corrosion rate of zinc in seawater is typically between 20 and 110 f.1rn1yr, varying with location, length of exposure, type of zinc, etc., as shown in Table 9.6. It is generally higher at the beginning of exposure and decreases with time.
292
CHAPTER 9
TABLE 9.6. Corrosion Rates of Zinc in Seawater Material
Duration of expos ure (yr)
Location
99%Zn 99%Zn Galvanized
Pacific Ocean Pacific Ocean Kure Beach. Hawaii
0.5
99.I%Zn
Eastport, Maine
I 3
Cast zinc Cast bar Galvanized Zinc Galvanized Zinc
Panama Bristol Channel Bristol Channel Digha, India Digha, India Tropical
Corrosion rate (p.mJyr)
Reference
110 70 53
0.5
575 575 436 436 436 436 436 436 436 662 662 663 663 217
28 28 25
28 92
4 4 0.1 0.1 4 16
64 34 41 21
18 IS
8 10
6
• X
0
8
•0 •
Monel Zinc Si bronze Lead Low brass Cu-Ni Aluminum
.!:!
'E c·
6
0
~
Gi c CD n.
.,
24 CD
«>
2
Rc:::: 002 mpy
2
8
16
Exposure Time, years FIGURE 9.7. Average penetration of wrought nonferrous metals after 16 years' continuous immersion in seawater. After Fink and Boyd [660].
293
CORROSION IN WATERS AND AQUEOUS SOLUTIONS
140,-----------------------------------------• Lake fresh water
120
•
E
Qi cQ)
0..
Sea water (immersed)
+ Sea water (tidal)
"': 100 c o ~ 80 60
Q)
Ol
'>Q;"
40
20 0
0
2
3
4
5
6
7
8
Years exposed
FIGURE 9.8. Comparison of zinc corrosion rates in lake water and seawater in the Panama Canal Zone. After Alexander et al. [690].
The data in Table 9.6 indicate that in the Pacific Ocean and at Kure Beach, the average corrosion rates for a l-yr exposure were only half of those for a 0.5-yr exposure. Figure 9.7 shows that zinc corrodes relatively fast compared to several other common metals and alloys. It also shows that the corrosion rate of zinc in seawater decreases with immersion time [6601. Alexander et al. [690] investigated the relative corrosion rates of zinc in fresh water and seawater. The corrosion rate in seawater was initially higher than that in fresh water but after about two years of exposure it became similar to that in fresh water, as shown in Fig. 9.8. The data in Table 9.7, reported by Anderson [436], show that a small amount of iron in rolled zinc had little effect on the corrosion rate. Campbell et al. [434], on the other hand, found that heat treatment of a galvanized coating, which introduced 10-20% iron, resulted in rapid failure of the coating in seawater. 9.4.3.2. Corrosion Form. Both uniform and localized forms of corrosion can occur on zinc in seawater, depending on the material [217, 661, 690]. Pitting was found to occur
TABLE 9.7. Corrosion Rates of Iron-Containing Zinc in Seawater" Corrosion rate l.}1m1yd Iron (%) 0.0003 0.0008 0.0014 0.0021 0.006 0.011
Total immersion
Tidal zone
Flowing water (2 ft/s)
35.6 22.4 22.4 22.4 22.4 35.6
20.3 20.3 22.9 22.9 20.3 22.9
76.5 50.8 50.8 50.8 50.8 50.8
"From Anderson [436]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission. b Average
for two I x 4 x
~-inch specimens exposed for
I yr.
CHAPTER 9
294
on 99.7% pure zinc in tropical seawater, the ratio of the deepest pit to the average corrosion rate being 14 [690]. Kweon and Coddet [691] observed blistering of a flame-sprayed 85% Zn-15% Al coating on an aluminum alloy after immersion in seawater for 10 months. 9.4.3.3. Effect of Flow. Tidal Zone, and Depth. In one study, similar corrosion rates were found for iron-containing zinc in a totally immersed condition and a tidal zone condition, as shown in Table 9.7 [436]. Water flow significantly increased the corrosion rate; at a flow velocity of 2 ftls the corrosion rate was more than double that in stagnant water. Khan et al. [662] reported that the corrosion rate at half-tide level is five times more than that in a fully immersed condition. Anderson [436] also reported a higher corrosion rate at mean tide level than in a fully immersed condition. The splash and tidal zones are usually much more corrosive than the submerged zone since they are continuously wetted with well-aerated seawater [660]. The differences between the data of Suzuki and those of others may be due to the fact that Suzuki used sprayed zinc, which is very porous. However, according to Suzuki [562]. the corrosion rate of a sprayed zinc coating, after a two-year exposure, was highest in the submerged zone, being about three times more than that in the splash zone, as shown in Fig. 9.9. The depth of immersion in the sea also affects the corrosion. Reinhart [575] reported that the corrosion rate of zinc in seawater decreased with depth from 113 f1rn1yr near the surface to 58 f1rn1yr at about 2000 ft but increased to 168 f1rn1yr at about 5500 ft. It is noted that the change of the corrosion rate with depth appears to correspond to the change of oxygen concentration with depth as shown in Fig. 9.1. 9.4.3.4. Effect of Temperature. Mor and Beccaria [179] reported that the corrosion rate in synthetic seawater decreased with increasing temperature from 25 to 60°C, as shown in Fig. 9.10. The decrease of corrosion rate with time was found to be associated with the accumulation of corrosion products on the zinc surface. With increasing temperature, the amount of zinc oxide and hydroxide decreased while calcium carbonate 50,----------------------------------------,
40 Zinc spray-coated steel
~ 30 enen
.Q
c 020
'iii ~
(;
u
10
OL---------~--------~------
Submerged zone
Tidal zone
___ L_ _ _ _ _ _ _ _~
Splash zone
FIGURE 9.9. Corrosion rates of zinc coating under various seawater conditions (2 years' exposure at Akashi, Japan). Data are taken from Ref. 562.
295
CORROSION IN WATERS AND AQUEOUS SOLUTIONS
3r----------------------------------------, . 25·C ... 40·C
400 Time, hours
FIGURE 9.10. Weight loss of zinc in synthetic seawater at pH 8.2 at different temperatures. From Mor and Beccaria 1179]. © Copyright by NACE International. All Rights Reserved by NACE: reprinted with permission.
increased in quantity. Other corrosion products were formed whose presence was not detected at room temperature, such as basic sulfates, ZnS04·3Zn(OH)2AH20. By precipitating together with CaC0 3 and MgS0 4 , this basic sulfate possibly makes the barrier layer more compact and thus inhibits the subsequent anodic dissolution of the zinc. 9.4.3.5. Effect of Bacterial Activity. The effect of sulfide, a common by-product of bacterial activities, on the corrosion of zinc was investigated by Mor et al. [209]. Figure 9.11 illustrates that the corrosion of zinc in synthetic seawater is accelerated in the presence of sulfide at pH values greater than 7.2, whereas at lower pH values the corrosion process is partially inhibited. Mor et al. attributed this finding to the different concentra-
• Aerated , no S2·
E
+
0.8
... Aerated , with S 2-
De-aerated, with S 2· XDe-aerated, no S 2·
~
~0 . 6
...
'"o
E 0.4 OJ
'w ~
0 .2
O~------------~------------~--------------~
6
8
7
9
pH FIGURE 9.11. Weight loss of zinc specimens after immersion for 48 hours in aerated and deaerated artificial seawater at different pH values, with and without addition of sulfide. After Mor et al. [2091.
296
CHAPTER 9
tions of S2- ions in the two pH regions. At pH values above 7.2, the S2- ions (at a concentration of 1O- '2M) cause the corrosion products to be predominately ZnS, which reduces the adhesion of corrosion products and, therefore, enhances the corrosion. 9.5. CORROSION IN AQUEOUS SOLUTIONS
9.5.1. Effect of Dissolved Species Solutions differ from natural waters in that a solution contains solutes, usually artificially introduced. A solution is formed when chemical compounds are dissolved in water. As a rough classification, solutions containing more than 0.1 mol per liter of solute can be considered to be concentrated while those containing less than 0.01 mol per liter can be regarded as dilute. As seen in the previous sections, the presence, even in trace amounts, of chemical compounds dissolved in water can greatly change the corrosion behavior of zinc. The corrosion behavior of zinc can vary drastically, depending on the solution composition and test conditions, from virtually noncorroding to rapidly dissolving. The major factors affecting the corrosion form and the rate of zinc dissolution are type of dissolved species, concentration, pH, and temperature. The presence of various chemical species can change the solubility of zinc dissolution products by forming complexes with zinc ions, increasing the electrolyte conductivity, modifying the composition, structure, and compactness of the corrosion products, forming an insoluble salt film on the surface, providing reactants for the anodic and cathodic reactions, and changing the reaction kinetics through catalytic or inhibitive adsorption. The corrosion processes of zinc in a solution are greatly influenced by the nature of the anions present. Depending on the specific effect, anions may be classified into three groups: (a) anions that increase the solubility of zinc, such as chloride or sulfate; (b) anions that reduce the solubility of zinc and thus promote the precipitation of zinc salts which may be protective, such as carbonate or phosphate; and (c) anions that react with the zinc surface and, depending on the reaction products, may form a passive film, such as chromate. Table 9 .8 lists the corrosion rates of zinc in some common solutions. The particularly low values in phosphate and chromate solutions are due to the formation of passive films on the zinc surface. It appears that in neutral solutions, with chemical agents that are neither electrochemically reactive nor capable of forming insoluble salts or complex ions with zinc, the corrosion rate of zinc is not very different from that in distilled water. Figure 9.12 shows the effect of concentration of several salts on the corrosion rate of zinc [559]. The corrosion rate of zinc in chloride and sulfate solutions increases with concentration up to about 5 gil and then decreases with further increase in concentration. At salt concentrations higher than 150 gil, the corrosion rate is actually less than half of that in pure water. In nitrate solutions, the corrosion rate decreases with increasing concentration. Nitrate is known to promote the passivation of zinc in neutral solutions [409,410]. However, as reported by Goodrich and Schmid [223], the presence of nitrate in ammonium-containing solutions enhances the corrosion of zinc because the formation of a zinc-ammonia complex in the solution reduces the pH near the surface and prevents the formation of a passive film. In LiBr solution, the zinc corrosion rate increases with
297
CORROSION IN WATERS AND AQUEOUS SOLUTIONS
TABLE 9.8. Corrosion Rates in Different Solutions in the Neutral pH Range Solution
Duration of exposure
Distilled water Distilled water Distilled water 0.INNa2S04 O. I N benzoate O.INNaCI O.IN Na3P04 O.IN chromate NaCI.5 gil KCI, 5 gil NaN0 3, 5gIl Na2S04' 5 gil K2S04,5 gIl 3.5% NaCl 8% Na2S04,IOH20 5% NaCI 3% NaCI 3% Na2S04 --.------
------
Corrosion rate (um/yr)
4 weeks I month 3 weeks 4 weeks 4 weeks 4 weeks 4 weeks 4 weeks 2 months 2 months 6 months 2 months 2 months I month I month 3 weeks I day I day
Reference
- -_. , - - - -
46 55 48 70 59 62 1.8 0.4 90 92 18 65 52 88 83 89 175 144
.... - ~---.
7\0 559 722 710 710 7\0 710 7\0 559 559 559 559 559 208 208 722 97 97
solution concentration, peaking at about 3M, and then decreases with further increasing concentration [1262]. Lorking [710] found that, although the formation of corrosion products is related to the solubility of zinc hydroxides and zinc salts, there is not necessarily a definite relationship between corrosion rate and solubility. He found that the corrosion rate in sulfate solutions was related to the amount of dissolved zinc in the solution. The corrosion rate of zinc specimens in zinc oxide-saturated sulfate solution was 2 pm/yr. which is less
100 ...... NaCI
>:
60
+KCI
-6: KN0 3
E
,3. Ql
. . Na,SO. · 10H.0
60
.I<,SO.
II:
c 0
'iii
e(;
()
40 20 0
0
50
100
150
200
250
300
Concentration (g/I) FIGURE 9.12. Effect of salt concentration on corrosion rate of zinc (T = 8-13°C, 57-189 days, nonaerated). Data are taken from Ref. 559,
298
CHAPTER 9
than one-tenth of that without zinc oxide in the solution. On the other hand, there was not much difference between the corrosion rates in zinc oxide-saturated and zinc oxide-free benzoate solutions. Leidheiser and Suzuki [97] reported that the presence of small amounts of Co 2+, up to 1O-4M, in 3.5% NaCI solution slightly inhibited the corrosion of galvanized steel, but a significant increase in the corrosion rate was found at higher concentrations of C0 2+. The effect of cobalt was postulated to be due to its incorporation into the zinc oxide, which reduces the flux of electrons for the cathodic reactions. At high cobalt concentrations, in or on the surface of the oxide, elemental cobalt aggregates to form metallic cobalt, which serves as a catalyst for the cathodic reaction. Some chemical agents can inhibit the formation of corrosion products on zinc surfaces. Boto and Williams [128] reported that the addition of a complexing agent (EDTA) prevented the formation of corrosion product in neutral or basic solutions and that the corrosion rate was limited only by the diffusion of oxygen. On the other hand, the presence of some inorganic or organic species may promote the formation of a surface film, which then acts as a barrier to corrosion. Besides chromates and phosphates, certain other inorganic agents (such as carbonates [194,331], molybdates [367, 102], tungstates [199,420], silicates [574,599], and cerium salts [605]) and certain organic species (such as gluconate [100], esters [601], phosphines [64], and benzene thiols [164]) are found to inhibit the corrosion of zinc.
9.5.2. Effect a/pH
In the absence of reducing or passivating agents, the corrosion of zinc in aqueous solutions is primarily determined by the pH of the solutions. The results in Fig. 9.13, reported by Roetheli et at. [497], show that the corrosion rate of zinc in water of pH 6-12 is relatively low. At pH values lower than 6 or higher than 12, the corrosion rate increases substantially. The low corrosion rate at pH values between 6 and 12 is primarily due to the formation of passive corrosion products on the surface of the zinc. According to Roetheli et at. [497], the decrease in the corrosion rate at pH values near 14 shown in the figure is due to a decrease in the solubility of oxygen in strongly alkaline solutions.
(ij Q)
0.5
>-
E
u
c:: 0
0.4
~
a; 0.3 c:
OJ Cl.
c;;
Qj
0 .2
is Q)
0>
'"
0.1
Qj
~
0
0
4
8
pH
12
16
FIGURE 9.13. Corrosion rate in distilled water as a function of pH (addition of N aOH or HCI for pH adjustment). After Roetheli et al. [497].
299
CORROSION IN WATERS AND AQUEOUS SOLUTIONS 1 ,000 -Na 3 P04
+ Na 2SO. 100
'6' Benzoate
"E <.J
c;
g
10
(II (II
.2 1:
OJ
'OJ ~
0 .1
FIGURE9.14. Effect of pH on weight loss of zinc in O. IN solutions. Data are taken from Ref. 710.
0.01
0
2
4
6
8
10
12
14
pH
The form of corrosion is related directly to the range and s-tability of passivity, which is a function of pH. Below pH 6 or above pH 12, the corrosion is normally of a general type, while in the pH range between 6 and 12 the corrosion tends to be more localized. Figure 9.14 shows the corrosion rate measured by Lorking [710] for zinc in various solutions as a function of pH. The pH dependence of the corrosion rates of zinc in sulfate, chloride, and benzoate solutions is similar to that in water, i.e., relatively low in nearneutral or slightly alkaline solutions and high in acidic or strongly alkaline solutions. In phosphate solutions, the corrosion is inhibited between pH 4 and 12, owing to the formation of a less soluble and more protective zinc phosphate film. In chromate solutions, the corrosion of zinc is inhibited almost in the entire pH range of 1-13 owing to the formation of a passive chromate-incorporated chromium oxide film. As shown in Fig. 2.3 in Chapter 2, pH determines the solubility of zinc in water, the solubility being lowest at pH 9. Over pH range 0-4, zinc enters solution as a divalent cation. From pH 4 to 12, zinc corrodes with precipitation of zinc hydroxides, due to their relatively low solubility. At pH values greater than 12, zinc enters solution as the zincate ion [ 1, 710]. It may be noted, by comparing Figs. 9.13 and 2.3 that the pH value at which the corrosion rate of zinc is the lowest does not coincide with the pH value for the lowest solubility. The lowest solubility for zinc hydroxide in water is near pH 9 [1], while the lowest corrosion rate is around pH 12. This indicates that solubility of corrosion products is not the only factor to determine the corrosion rate. If corrosion rates were determined primarily by the solubility of corrosion products, the lowest corrosion rate should be observed at a pH value around 9. pH not only dictates the solubility and stability of the corrosion products but also determines the physical structure. Besides the effect on the formation of corrosion products, the pH of a solution affects the corrosion of zinc by affecting the cathodic reaction. It has been established that in aerated solutions of near-neutral or higher pH values, hydrogen reduction is not the main cathodic reaction in the corrosion process, owing to the relatively high overpotential for hydrogen reduction on a zinc surface. The main reducing agent for the cathodic reaction of the corrosion process in aerated solution is oxygen [116, 128]. As the hydrogen ion
FIGURE 9.15. Distribution of anodic and cathodic areas on an electrode surface at the waterline. Data are taken from Ref. 403.
'0
:>:l
tTl
"1:) ....,
»
n ::c
'"oo
CORROSION IN WATERS AND AQUEOUS SOLUTIONS
301
concentration increases, hydrogen evolution becomes the more important cathodic reaction.
9.5.3. Ejject of Immersion Conditions It has been known that when a zinc sheet is partially immersed vertically in a chloride solution, attack is soon observed at numerous points, distributed sporadically, and spreads out from some of these points, producing corroding areas a short way below the waterline. A corrosion-immune zone is formed along the waterline with the attacked zone below it. As reported by Thornhill and Evans [403], the cause of waterline immunity is generally attributed to the fact that alkali, a cathodic product, is preferentially found in the zone where oxygen, the cathodic stimulator, can be renewed most readily. The zinc salts formed by anodic action in the lower parts must travel a considerable distance before precipitating as hydroxides, whereas any zinc salts formed momentarily in the upper parts are precipitated by the excess alkali in physical contact with the metal, thus stifling any further anodic attack. As a result, interference colors due to the hydroxide film are visible in this regIOn. The electrochemical nature of this reaction process of zinc in 1O-3M NaCl solution can be seen in Fig. 9.15 [403]. The area close to the head of the meniscus is the cathode, while the area around the foot of the meniscus is the anode. This kind of distribution is determined mainly by differential aeration since oxygen can easily reach the head of the meniscus without passing through a thick layer of liquid [403, 708]. From Fig. 9.15, it is seen that at the outset, most anodic attack is at the foot of the meniscus; later, it will be directed mainly on any physically loose areas in the lower part of the specimen; finally, when the supply of unstable matter on these areas has been exhausted, it returns to the foot of the meniscus. The distance between the attack and the cathode is determined by the pH and conductivity of the solution. High conductivities favor attack at a distance from the cathode. Bianchi [708] pointed out that formation of macro cells of differential aeration is connected with passivation of zinc in the more aerated zones. Differential-aeration corrosion cannot exist without passivation of the more aerated zones of the zinc surface. If buffered solutions are used, so that the pH is outside the range for passivation of zinc, the formation of differential-aeration macrocells may not occur. Conditions of aeration and sample rotation have an important influence on the corrosion rate of zinc. Table 9.9 shows the effect of rotation and aeration on the corrosion
TABLE 9.9. Effect of Aeration and Rotation on the Corrosion Rate (j1m/yr) of Zinc in Distilled Water and 3.5% NaCI Solution" Distilled water
---------------------------- - - - - 3.5% NaCI
------
Type of test
15 days
30 days
15 days
30 days
128 249 193 194
124 165 128 122
104 362 127 194
133 242 118 226
--------~-
No aeration; no rotation No aeration; rotation Aeration; no rotation Aeration; rotation
"From Anderson and Reinhard [559J. Reprinted by permission of John Wiley & Sons, Inc.
302
CHAPTER 9
rate. As found by many authors [112. 113. 116, 128,445], the corrosion of a bare zinc surface in neutral nondeaerated solutions is oxygen-diffusion-controlled. while that of a corrosion-product-covered surface may be independent of oxygen concentration in the solution. 9.5.4. Effect of Surface Treatments
Surface treatments can have a significant effect on the corrosion of zinc, especially in the early stages of a corrosion test. Among the many surface treatment processes, chromating is the oldest and the most effective one in increasing the corrosion resistance of zinc in solutions [57, 65]. Chromate conversion coating on galvanized steel that is immersed in distilled water can delay the appearance of white rust for many weeks [591]. Dipping in phosphate, molybdate, tungstate, and carbonate solutions has also been found to be beneficial in increasing the corrosion resistance of zinc [93, 94, 101. 102, 178,404, 420]. Anodization can also produce surface films that inhibit the corrosion of zinc in water [493,596]. 9.5.5. Effect of Metallurgical Factors 9.5.5.1. Crystal Orientation. Using single-crystal zinc, Ashton and Hepworth [222] measured the corrosion currents of different crystal planes in 0.5M NaOH. They found that the corrosion rate decreased in the order (1120) > (1020) > (0001), which is the same as the order of decreasing planar packing density. Abayarathna et al. [446,721] found the order to be the same in terms of the corrosion rate of zinc single crystals in three different orientations. The oxide or hydrated oxide film which formed on the (0001) surface appeared to be the most protective. They also found that particular planes of a zinc single crystal in near-neutral 1M (NH4)S04 solution were corroded preferentially, so that the corrosion on the basal plane resulted in hexagonal pits with facets parallels to the (10TO) planes. On the (10TO) surfaces, elongated and striated structures were observed, with the
3 r--------------------------------------. As roll ed - 3% per pass
en
E
b
~
oi ~
2 .8 . 5%
2.6
.§ 2 .4 en
e o
u 2 .2
2
~------_L
o
20
______
~~
40
______
_ L _ _ _ _ _ _~ _ _ _ _~
60
80
Total reduction , %
FIGURE 9.16. Corrosion rate in 7.5% HCI solution as a function of total reduction of zinc sheet immediately after cold rolling. After Kobayashi et al. [411].
CORROSION IN WATERS AND AQUEOUS SOLUTIONS
303
3 .---------------------------------~
5% per pass
(/) 2.8
.E b
~
oi
+ As rolled ~ ,
Month
2.6
li!
.~ 2.4 (/)
~
o
•
y
•
O 2 .2
25L-------1JO--------1L5--------2~ O -------2~5~----~ 30 Residual stress, MPa
FIGURE 9. 17. Relation between corrosion rate and the original residual stress for cold-rolled zinc sheets. After Kobayashi et al. [411].
elongation aligned along the (1010) direction. Corrosion on the (1120) surfaces resulted in a crested ridge structure. Also, regardless of the initial surface orientation, the morphology of the corroded surface was consistent with the (1120) plane being the most active surface. Leidheiser and Suzuki [97] found that for different galvanized steels with the amount of (1000) grain orientation varying from 50% to 98%, the corrosion rates in 3% NaCI solution were approximately the same. On the other hand, the corrosion resistance of electrolytic zinc coatings, reported by Girin and Panasenko [459], significantly increased with a reduction in the average scattering angle for the axial texture with the (1120) axis and with an increase in the percentage of crystals with a disordered orientation. 9.5.5.2. Deformation. Fedrizzi and Bonora [715] measured the polarization resistances of zinc coatings subjected to different extents of deformation. The polarization resistance of deformed samples in 3.5% NaCI solution was slightly lower than that of the undeformed ones. This was attributed to the increased surface area caused by the coating cracks created by the deformation. Kobayashi et al. [411] studied the effect of cold rolling on the corrosion rate of zinc in 7% HCI solution. The corrosion rate exhibited a maximum at about 30% reduction for the samples immediately after rolling (Fig. 9.16) but was independent of the amount of rolling when the samples were held at room temperature for one month after rolling. The difference III the corrosion rates was found to be related to residual stress, which decreases with time after rolling, as shown in Fig. 9.17.
10 Corrosion in Soil 10.1. INTRODUCTION As a corrosion environment, soil differs from other natural environments, such as atmospheric environments or waters, in two major aspects. Firstly, soil has a much wider range of chemical and physical properties. For example, the pH of soil may vary from as low as 2.6 to as high as 10.2, and the resistivity from several tens of ohms to near 100 k.Q [357]. As a result, the corrosion rate of a metal may drastically vary from soil to soil. Secondly, soil is a highly inhomogeneous environment, both microscopically (e.g., on the scale of a clay particle) and macroscopically (e.g., on the scale of a rock). Thus, the corrosion in soil is seldom uniform across the metal surface. Galvanized steel is commonly used for structures in soil such as lamp posts, culverts, and land enforcements. Zinc anodes are also frequently used in soil for cathodic protection of underground structures such as pipelines and storage tanks. There have been few systematic studies of the corrosion of zinc and its alloys in soils. The most extensive investigations were carried out by the U.S. National Bureau of Standards (NBS) between 1910 and 1955 [357,565]. In these investigations, the corrosion of rolled zinc, cast zinc, and galvanized steels, with different coating thicknesses, was evaluated in more than 50 different soils at various locations across the United States. The data presented in this chapter are abstracted mainly from the reports of these investigations. 10.2. CHARACTERISTICS OF SOIL Soil makes up the first few feet of finely divided, weathered (by climatic and biological processes) rock material covering the level and moderately inclined portions of the earth [357]. It is usually named and classified according to the size range of its particulate matter: the principal categories are sand, silt, and clay, which derive their names from the predominant size range of the inorganic constituents. Particles between 0.07 and about 2 mm in mean diameter are classified as sands. Silt particles range from 0.005 to 0.07 mm, and clay particles range from 0.005 mm down to colloidal matter. The relative proportions of the three types determine many properties of the soil. A common classification scheme is shown in Fig. 10.1 for various proportions of sand, silt, and clay. 305
306
CHAPTER 10
40 PERCENT SAND
20
FIGURE 10.1. Proportions of sand, silt, and clay making up the various groups of soils classified on the basis of particle size. After Romanoff [357].
More detailed information on the classification and the chemical and physical properties of soils related to corrosion can be found in several references [314, 357, 565, 987]. Besides the differences in texture, soils differ in their chemical composition and their interaction with other environmental factors. A large number of chemical compounds exist in soils. The inert compounds are chiefly those of silicon. aluminum, and iron. Iron, in its various oxidation states, is responsible for the color of many soils. Furthermore. this color is an indicator of the degree of aeration of the soil; red, yellow. and brown colors indicate the oxidized forms of iron. Poorly aerated soils are predominantly gray. indicating the presence of reduced forms of iron. The soluble base-forming elements are sodium, potassium, calcium. and magnesium. The acid-forming species are carbonate. bicarbonate, chloride, nitrate, and sulfate. The nature and amount of soluble salts, together with the moisture content of the soil, generally determine the ability of the soil to conduct an electric current, which is very important in relation to corrosion processes. Another important factor influencing the corrosion of a metal is soil acidity. The development of acidity in soils is a result of the natural processes of weathering under humid conditions. In regions of plentiful rainfall, not only are soluble salts removed from soil, but the base-forming constituents normally present in the colloidal materials are partially removed. resulting in increased acidity [565]. Bacteria also affect the chemical properties of soils through oxidation and reduction reactions. Bacterial activities tend to decrease the oxygen content and replace oxygen with carbon dioxide. Most bacterial activities occur in the upper six inches of soil [1280].
307
CORROSION IN SOIL
50,-----------------------------------------~
40 )(
G
30
ai u c
~ 20
'r;;
ill
0:
10
O~----~------~======~====~ -20 -10 o 10 20 Temperature, °C
FIGURE 10.2. Effect of temperature on soil resistance. Data are taken from Ref. 1105.
The physical properties of soils that are important in relation to corrosion are mainly those that determine its permeability for air and water. The particle-size distribution in a soil is an important factor with respect to air and moisture content. Soils of fine texture, due to a high clay content, contain more closely packed particles and have less pore capacity for gaseous diffusion. It is generally assumed that the composition of the gases in the upper layers of a soil are similar to that in the atmosphere above the soil. Soil moisture may derive from free groundwater, which is constantly present at certain depths below the surface, gravitational water, which enters the soil from the surface through rainfall, and capillary water, which is the usual form of water retained in a soil owing to capillary action.
1,000
a )(
E u
100
G
~
S
~
'iIi
10
ill
a:
1L---------~--
5
25
________
~
________
45
~
____
~
65
Moisture content of soil, %
FIGURE 10.3. Effect of moisture content on resistivity of a clay soil. Data are taken from Ref. 1105.
308
CHAPTER 10
Various types of parameters have been used to evaluate the corrosion of metals in soils [987]. It is generally accepted that aeration, moisture content, pH, and resistivity are the main factors affecting corrosion, although there is still a lack of good correlation between the corrosion of a metal and these factors. Among them, resistivity, which is directly related to the amount of soluble salts, and moisture content seem to have the best correlation with the corrosion rate. The resistivity of a soil is typically a function of temperature and moisture, as shown in Figs. 10.2 and 10.3 [1105]. 10.3. CORROSION RATES Tables 1O.l and 10.2 contain the results obtained in the NBS investigation for the corrosion rates of zinc and galvanized steel in various soils. The average annual corrosion rates of rolled zinc and zinc coatings in most soils are below 10 11m/yr. The maximum penetration rates, shown in Table 10.2, are 3-30 times the surface-averaged corrosion rates. Table 10.2 also shows the beneficial effect of zinc coatings on pitting corrosion of the steel. The area-averaged corrosion rates of galvanized steel are 3-6 times lower than those of bare steel whereas the pitting corrosion rates of galvanized steel are 4-20 times lower than those of bare steel. As also shown in Table 10.2, the average corrosion rates of galvanized steel are similar to those of zinc, but the pitting penetration rates of galvanized steel are noticeably lower than those of zinc. This indicates either the effect of galvanic protection by the zinc coating or the higher corrosion resistance of the Zn-Fe alloy layers at the coating/steel interface. It was found in the NBS study [357] that when most of the outer coating corrodes away, the penetration rate slows down, indicating a higher corrosion resistance of the intermetallic layers.
10.3.1. Effect of Soil Factors The factors which may affect the corrosion of zinc and galvanized steel in soils are numerous, but the correlation between the corrosion behavior and the various factors is in general rather poor. The NBS results appear to suggest that the corrosion rates tend to be lower in soils with high resistivities, as shown in Fig. 10.4. Insofar as pH is concerned, there is little correlation between corrosion rate and soil pH, as shown in Fig. 10.5. (It should be noted that the corrosion rate of zinc in aqueous solution has a clear dependence on pH as shown in Fig. 9.13 in Chapter 9.) This lack of correlation between corrosion rate and soil pH and the very weak correlation between corrosion rate and soil resistance reflects the very complex nature of corrosion in soil. The various corrosion rates in different soils are essentially determined by the synergistic effect of soil resistivity, pH, moisture content, aeration, soluble salt concentrations, and other factors [987, 1085, 1086, 1088]. In general, as stated by Romanoff [357], poorly and very poorly aerated soils are more corrosive to zinc, and a high corrosion rate is not always associated with deep pitting. Soils of fair to good aeration, but containing high concentrations of chlorides and sulfates, tend to induce deep pitting. The extreme corrosion rates and deep pitting seem to be associated with soils having very low pH
309
CORROSION IN SOIL
TABLE 10.1. Average Corrosion Rates, R, of Galvanized Steel Pipe Specimens" Embedded in Various Soils for 10 Yearsh No. I 2 3 4 5 6 7 8 9 10 II 12 13
14 15 16 17 19 20 22 23 24 25 26 27 28 29 30 31 32 33 35 36 37 38 40 41 42 43 44 45 46 47
c
Soil type Allis silt loam, Cleveland, Ohio Bell clay, Dallas, Tex. Cecil clay loam, Atlanta, Ga. Chester loam, Jenkintown, Pa. Duhlin clay adobe, Oakland, Calif. Everett gravelly sandy loam, Seattle, Wash. Maddox silt loam, Cincinnati. Ohio Fargo clay loam, Fargo, N. Dak. Genesee silt loam, Sidney, Ohio Gloucester sandy loam, Middleboro, Mass. Hagerstown loam, Loch Raven, Md. Hanford fine sandy loam, Los Angeles, Calif. Hanford very fine sandy loam, Bakersfield, Calif. Hempstead silt loam, St. Paul, Minn. Houston black clay, San Antonio, Tex. Kalmia fine sandy loam, Mobile, Ala. Keyport loam, Alexandria, Va. Lindley silt loam, Des Moines, Iowa Mahoning silt loam, Cleveland, Ohio Memphis silt loam, Memphis, Tenn. Merced silt loam, Buttonwillow, Calif. Merrimac gravelly sandy loam, Norwood, Mass. Miami clay loam, Milwaukee, Wis. Miami silt loam, Springfield, Ohio Miller clay, Bunkie, La. Montezuma clay adobe, San Diego, Calif. Muck, New Orleans, La. Muscatine silt loam, Davenport, Iowa Norfolk fine sand, Jacksonville, Fla. Ontario loam, Rochester, N.Y. Peat, Milwaukee, Wis. Ramona loam, Los Angeles, Calif. Ruston sandy loam, Meridian, Miss. St. John's fine sand, Jacksonville, Fla. Sassafras gravelly sandy loam, Camden, N.Y. Sharkey clay, New Orleans, La. Summit silt loam, Kansas City, Mo. Susquehanna clay, Meridian, Miss. Tidal marsh, Elizabeth, N.J. Wabash silt loam, Omaha, Nebr. Unidentified alkali soil, Casper, Wyo. Unidentified sandy loam, Denver, Colo. Unidentified silt loam, Salt Lake City, Utah
p (Q'cm)
--------
pH
R (jlm/yr)
--------
1,215 684 30,000 6,670 1,345 45,100 2,120 350 2,820 7,460 11,000 3,190 290
7.0 7.3 5.2 5.6 7.0 5.9 4.4 7.6 6.8 6.6 5.3 7.1 9.5
lUI 1.5 1.7 7.9 7.7 0.5 10.R 3.2 5.0 5.2 3.7
3,520 489 8.290 5,980 1,970 2,870 5,150 278 11,400
6.2 7.5 4.4 4.5 4.6 7.5 4.9 9.4 4.5
1.1 1.5 4.2 14.8" 2.9 4.9 5.2 40.6" 1.1
1,780 2,980 570 408 1,270 1,300 20,500 5,700 800 2,060 11,200 11,200 38,600 970 1,320 13,700 60 1,000 263 1,500 1,770
7.2 7.3 6.6 6.8 4.2 7.0 4.7 7.3 6.8 7.3 4.5 3.8 4.5 6.0 5.5 4.7 3.1 5.8 7.4 7.0 7.6
1.45 2.9 3.9 8.8 25.5" 1.9 0.67 2.4 7.4 1.3 1.0 8.7 0.85 4.0 2.2 3.0 5.5 1.9" 7.5 0.7 4.3
2.2 d
3.7
------
"Average coating thickness, 121 JIm. "Ref. 357. 'Original identification. dSheet specimens. "Coating corroded completely: data included the corrosion of steel.
CHAPTER 10
310
TABLE 10.2. Average Annual Surface-Averaged Corrosion Rates (CR) and Maximum Penetration Rates (MPR) of Zinc and Galvanized Steel Embedded in Soils for 9 Years" Rolled zinc No.
p
b
51 53 55 56 58 59 60 61 62 63 64 65 66 67 70
Soil type Acadia clay Cecil clay loam Hagerstown loam Lake Charles clay Muck Carlisle muck Rifle peat Sharkey clay Susquehanna clay Tidal marsh Docas clay Chino silt loam Mohave fine gravelly loam Cinders Merced silt loam
CR
(Q·cm)
MPR
Galvanized stee( CR
MPR
Steel CR
MPR
(pmlyr) {flm/yr) (pmlyr) (pm/yr) (pmlyr) (pm/yr)
pH
190 17,790 5,210 406 712 1,660 218 943 6,920 84 62 148 232
6.2 4.8 5.8 7.1 4.8 5.6 2.6 6.8 4.5 6.9 7.5 8.0 8.0
22.9 5.3 3.3 21.5 35.3 21.9 112d
22.9 3.8 3.9 26.3 43 14.1 77.4" 7.2 d 4.3 9.6 7.6 7.6 11.9
22.5 16.9 16.9 36.7 180.6 22.2 234+ 34 16.9 22.6 28.2 16.9 16.9
83 16.2 19.6 132 82.7 35.8 d
5.2 6.2 9.6 6.7 6.7 28 d
79 36.6 22.5 81.8 163 61.1 416+ 39.5 33.9 69.4 223 158 124
455 278
7.6 9.4
131d 17.2
749+ 237
58 9.9'
286 18.5
151 64
85.5 20.1 25.3 51.1 43" 33.4 58.8 d
361+ 209 260 409+ 277 56.4 164 135 192 226 226 183 409+ 409+ 344
"Ref 357. "Identification in the original report. 'Nominal coating thickness. 130 pm. dData at 4 years. 'Data at 11.2 years.
30,-----------------------------------~
~ ~
c o 'iii ~10 (;
u
,
,
",
O~--------L---------~~------~-----·~·--~
10
100
1,000
10,000
100,000
Resistivity, ohm-em FIGURE 10.4. Average annual corrosion rates of galvanized steel embedded in 53 different soils across the United States for 10 years, as a function of soil resistivity. Data are taken from Ref. 357.
CORROSION IN SOIL
311 30~------------------------------'
2i
E c o
.iii
::10
o
FIGURE 10.5. Average annual corrosion rates of galvanized steel embedded in 53 different soils across the United States for 10 years, as a function of soil pH. Data are taken from Ref. 357.
I·
u
••
"
;' ,: .'
OL---~--~'~'~----~----~----~----~--~.
3
4
5
6
7
8
9
10
pH
values. Muddy clay and peat (as compared to sand) are, in general, more corrosive to zinc [ 1085]. Camitz and Vinka [1085] investigated the corrosion rate of galvanized steel samples in seven different soils and found that the position of galvanized steel samples with respect to the groundwater level had no distinct effect on the corrosion rate, as shown in Fig. 10.6. The level of the water table was found, however, to affect the corrosion rate of the steel samples. Camitz and Vinka also found that the corrosion rate was lower on panels embedded in a homogeneous sandfill than on panels placed directly in the original soil, which was attributed to the formation of a more protective surface film, due to better aeration in the sand. 60
o above G W table o above G W table, in sand fill
'-
~
Q)
~
'">E 40
below GW table below G W table, in sand fill
:J
'§ c
o
g20 o
u
o
-
clay 1
ctay 2
I
S clay
Mclay 1 Mclay 2
B ~
peal
sand
Soils
FIGURE 10.6. Corrosion rates on zinc-coated panels after about three years of exposure above and below the groundwater (GW) table in soil (Data for the Kramfors test site are for only one year of exposure.) AfterCamitz and Vinka [1085] .
CHAPTER 10
312
Different corrosion rates may be found on different parts of a single structure owing to a highly inhomogeneous soil environment. Edgar r1088] investigated the corrosion of culverts that were put into service some 30 to 40 years ago and reported that the corrosion rate for most pipes was below 5 Jim/yr. The corrosion was most severe in the center of the culvert, which was usually the joint between two sections. He reported that the worst corrosion typically occurred at the invert, or the bottom, of the pipe. Edgar also pointed out that the use of poor backfill material was the most obvious defective practice causing poor corrosion performance of galvanized culverts.
10.3.2. Galvanic Corrosion Galvanic corrosion of zinc in soil occurs when the zinc coating on galvanized steel is partially removed, exposing the steel underneath, or when it is used as an anode connected to a steel structure for cathodic protection of the steel. In contrast to selfcorrosion, which seems to lack a clear correlation with the various factors in soil, galvanic corrosion has a definite dependence on the resistivity of the soil. This is because, for a given potential difference between the anode and cathode, the amount of galvanic current flowing between the coupled metals is directly proportional to the resistivity of the soil. The galvanic corrosion of zinc coupled to various metals in different mediums is discussed in detail in Chapter 7 (Section 7.2). 10.4. ELECTROCHEMICAL MEASUREMENTS The electrode potential of zinc in a soil depends on the chemical composition of the soil. In Table 10.3, the electrode potentials reported by Romanoff r357] for zinc, steel, and a zinc-iron alloy layer of galvanized steel in 12 different soils are tabulated. The specimen of the zinc-iron alloy with surface condition A was prepared by electrolytic removal of the outer zinc coating, and the specimen with surface condition B was achieved
TABLE 10.3. Potentials (V seE) of Zinc-Iron Alloy, Zinc, and Steel in "Air-Free" Soils Zinc-iron alloy, surface A. outer zinc coating removed electrolytically Soil
Initial
Steady
51 55 56 58 60 61 62 63 64 65 66 70
-0,96 -0,86 -0,91 -0,87 -0,96 -0.92 -0.90 -0.92 -0.97 -0.95 -0.80 -0.95
-0,62 -0,60 -0,61 -0,63 -0,64 -0.66 -0.54 -0.65 -0.61 -0.64 -0.61 -0.60
Zinc-iron alloy, surface B, outer zinc coating removed by corrosion Initial
Steady
-0.96
-0.103
-0.90 -0.99 -0.85 -0.98
-0.75 -0.96 -0.84 -0.88
Zinc
Steel
-1.02 -1.02 -1.04 -1.04 -1.02 -1.02 -0.92 -0.94 -1.08 -1.01 -0.94 -0.99
-0,71 -0,75 -0,73 -0,74 -0.68 -0.72 -0.72 -0.64 -0.73 -0.71 -0.72 -0.76
313
CORROSION IN SOIL
-0.4.-------------------.-------------------, II
Coltosion in 1he lield
... Electrolytic stri pping
-0.6 w
u
>'" ~ -0.8
c: 0
a. -1
\
_1.2L---~----~----L----L----~----L---------~
-20
-10
- 15
-5
o
5
10
15
20
Current densi ty, JiA I em'
FIGURE 10.7. Anodic and cathodic polarization curves of zinc-iron alloy exposed by corrosion in the field (surface B) and electrolytic stripping in the laboratory in an "air-free" environment (surface A). After Romanoff [357)
by using the section of the galvanized steel that had its outer zinc coating corroded by 13 years of exposure to soiL The measurements were made in water-saturated soils from the test sites. Air was excluded, to prevent the soil from drying in the laboratory, by sealing the test celL The potential of zinc in the 12 soils was found to be between -0.92 and -1.08 VseE' The potentials of the zinc-iron alloy are generally more positive than those of zinc, reflecting the effect of the presence of Fe in the alloy. The difference between the steady-state potentials for the two surface conditions of the alloy was attributed by Romanoff to the presence of a cathodic film formed on surface B during field exposure. Figure 10.7 shows that, because of the presence of this surface cathodic film, the polarization resistance is much larger for surface B than surface A. This figure also explains the higher steady potential values in Table 10.3 for surface A; because of the relatively higher reactivity, the zinc-iron alloy layer is soon corroded away, resulting in the exposure of the substrate iron.
TABLE 10.4. Weight Loss (in Grams) of The Samples with Different Surface Conditions of Galvanized Steel and Bare Steel Samples in Aerated Soil for 60 days" Soil
Surface Ai>
Surface Be
Surface Cd
Galvanized steel
Steel
.-----~---.-----------------------------------
Lake Charles clay. pH 7.1. p = 400 Q·cm Merced silt loam. pH 9.4. p = 280 Q·cm
2.231 1.444
o
0.166
3.234
2.255
o
0.Q45
2.431
2.03
"Ref. 357. /lGalvanized steel with ourer zinc layer removed electrolytically. (,Galvanized steel with outer zinc layer removed by soil in the field. "Galvanized steel with zinc layer removed by soil in the laboratory.
CHAPTER 10
314 0.8r-----------------------------------------~
>- 0.6
a.
E
ai
1ii
~ 0.4
o
'iii
2
o
u 0.2
OL-____
o
~
_______ L_ _ _ _ _ _
10
20
~
_ _ _ _ _ _L __ _ _ __ L_ _
30
40
~
50
Time, months
FIGURE 10.8. Corrosion rate (in mils per year) determined by the linear polarization technique versus time. After Rabeler; data are taken from Ref. 1086.
To further evaluate the properties of the cathodic film. samples with surface condition C, which is similar to that from field exposure, were prepared by exposing a galvanized steel sample to a very corrosive soil under a laboratory condition. Table 10.4 shows the beneficial effect of the surface conditions Band C. The higher corrosion resistance of surface conditions B and C was attributed to the formation of a protective film in the soil by the galvanic action between the zinc coating and the bare steel. It was further found that a similar film was formed on a steel surface galvanically coupled to a piece of zinc immersed for one year in tap water containing 3% NaCl. This film was white and was identified by X-ray analysis to be primarily zinc silicate. The formation of a protective cathodic film on the surface of the zinc-iron alloy may perhaps explain the fact that in many soils the corrosion rate of galvanized steel can be high initially but remain relatively low in value thereafter for a lengthy period of time. Corrosion rates in soils are sometimes determined as corrosion currents by electrochemical methods. Serra and Mannheimer [313] measured the corrosion currents of galvanized steel in four different water-saturated soils. which were held in a sealed cell. They found that, except for the relative higher values initially, the corrosion current was quite constant during the one-month measurement. Rabeler [1086] conducted corrosion current measurements in the field on galvanized steel anchors supporting electrical power lines. The results, shown in Fig. 10.8, indicate that the corrosion rate decreases with time to a relatively low value and were found to be in good agreement with weight loss measurements on anchors embedded for about four years in the field.
11 Under-Paint Corrosion 11.1. INTRODUCTION Knowledge about painted zinc-coated steels has rapidly accumulated during the past two decades owing to the extensive research and development of more corrosion-resistant car bodies. There are numerous published studies on the corrosion perfonnance and corrosion mechanisms of various painted steel products. This chapter aims to compile the results reported in these studies. The basic characteristics of paint and its function as a physical barrier are briefly described in the first section. In subsequent sections, the tests that are used to evaluate under-paint corrosion are described, and infonnation on the effects of various components in a painted system on the corrosion perfonnance IS presented. In the last section, the mechanisms of under-paint corrosion are discussed. 11.2. BASIC CHARACTERISTICS OF PAINT Paint is a coating material consisting essentially of pigments dispersed in a solution of a binding medium. The main components are (a) a binder or vehicle, (b) pigments and extenders, (c) volatile solvents or thinners, and (d) additives such as driers, hardening agents, stabilizing agents, surface-activating compounds, and dispersion agents. The characteristics of various types of paint and the specific functions of each paint component have recently been reviewed by Van Eijnsbergen [1202]. A paint coating is used either to protect the substrate material from deterioration or to give the surface a certain aesthetic appearance. It protects the metal substrate from corrosion primarily by two mechanisms: (1) by serving as a barrier to water, oxygen, and other corrosive chemical species and (2) by serving as a reservoir for corrosion inhibitors [I 109]. The barrier effect of the paint can be improved by increasing its thickness, by utilizing pigments and fillers that make the paint less penneable to water and oxygen, and by increasing its resistance to degradation. A metal surface corrodes very little when it is properly painted. Corrosion occurs only at places where the paint is damaged or has deteriorated. Starting at the damaged places, such as at cuts, scribes, and cracks, the corrosion propagates laterally under the paint, causing paint delamination and blistering. The nature and extent of corrosion are detennined by the type of substrate, type of paint system, and environmental conditions. 315
316
CHAPTER II
11.2.1. Components in Paint The vehicle, which includes solvents and resin or oil, etc., determines the basic physical and chemical properties of the paint, but these can be modified by the nature and proportion of pigments [312, 1107, 1202]. Vehicle materials are usually polymers of relatively low molecular weight. The function of the volatile solvents in a vehicle is to control the viscosity of the paint for ease of manufacture and for subsequent application. Additives are added to the vehicle to improve the brittleness and to inhibit crazing. The primary function of the pigment for a finish paint is to provide color, but in a primer paint the pigment contributes to the corrosion resistance and the durability of the whole system. The corrosion-inhibiting effects of pigments are realized in several ways. The reactions of pigments with vehicles reduce the number of hydrophilic groups, thereby increasing the resistance of the paint to the permeation of aggressive species. In addition, pigments and their reaction products inhibit the corrosion of the substrate metals by their buffering ability, which fixes the pH of the electrolyte in the paint. Furthermore, some pigments, such as chromates, can leach out ionic species to inhibit the reactions at the paint/metal interface or to passivate the metal surface. In general, a single paint does not possess all the required properties, and in real applications it is usually necessary to use multilayer paints comprising a primer, a finish, and possibly one or two intermediate paints. A primer is the first coat of a system. Its principal functions are to provide adequate adhesion and afford good protection to the substrate. It can contain rust-inhibitive pigments such as zinc chromate and zinc dust. The finish or the final paint must make up for the deficiencies of the primer and provide the required color and degree of gloss.
11.2.2. Barrier Properties of Paint For an undamaged paint, corrosive species such as water and oxygen have to diffuse through the paint in order to reach and react with the metal surface. The rates of diffusion have been investigated in a number of studies. It has been estimated [364, 1108] that the amount of water diffusing through various O.I-mm-thick paint films at 85-100% relative humidity is between 0.2 and 1.8 g/( cm2·yr). The diffusion rate for oxygen in various paint films is between 0.002 and 0.05 g/( cm2 ·yr). Tanabe et al. [738] reported that the diffusion coefficient of oxygen in an epoxy paint film in 3 wt. % NaCI solution is between 1.1 x 10-7 and 1.6 x 10-7 cm2/s and varies little with paint thickness and time of immersion. The diffusion rate of ions in paint is, in general, much smaller than that of water and oxygen but varies significantly with the type of resin and pigment used [312]. Water penetration into paint depends on osmotic pressure. Miyoshi et al. [772] evaluated water penetration in distilled water, in atmospheres with 100%,95%, and 85% relative humidity, and in a salt solution. Distilled water and an atmosphere with 100% relative humidity bring about the highest osmotic pressure and were found to give the highest water penetration, as shown in Fig. 11.1. Miyoshi et al. also found that water penetration is independent of the substrate since similar amounts of penetration were found for cold-rolled, galvanized, and zinc/iron-coated steels. The resistance to water penetration may also vary with time. Some paints become more and more penetrated by the solution during immersion while others maintain their diffusion-restraining properties [244].
UNDER-PAINT CORROSION
317
3.0,-------------------,
Water ~
~
2.0
I-
Z
w
I-
Z
FIGURE II. I. Water penetration into a 90-Jin1 coat on galvanized sheets exposed to distilled water. 5% NaCI solution, and atmospheres with relative humidities of 100%, 95%, and 85%. Water content is calculated from the weight increase. Reprinted from Miyoshi el al. [7721, with permission, SAE Paper 850007. © 1985 Society of Automotive Engineers. Inc.
0
U
a:
1.0
w
I-
5%NaCI Sol
«
~
95%RH 85%RH
0
0 TIME
(d)
Many studies have revealed [364, 1108] that the diffusion of water or oxygen is not necessarily the rate-limiting process for under-paint corrosion, since the intake rate of water and oxygen by a paint film can be much larger than that needed for the corrosion of an unpainted metal surface. The durability of a paint is primarily determined by (1) the internal strength of the paint films and (2) adhesion to the substrate [3121. The common degradation mechanisms of a paint include abrasion and impact, cracking or crazing at low or high temperature, bond breakage within the polymer matrix due to hydrolysis reactions, oxidation, or exposure to ultraviolet light, and freeze-thaw cycling. The result of such degradation allows access of reactants to the coating/substrate interface without the necessity of diffusion through the matrix. With increasing age, the elasticity of the film usually decreases. Thus, expansion and contraction of the metal base caused by severe temperature changes can result in the formation of discontinuities in a relatively inelastic paint film. In practice, however, premature failures of paints are usually due to insufficient preparation and treatment of the metal surface. The physical and chemical properties of various paint products and their failure modes have been described in detail elsewhere [312,11091· 11.3. CORROSION TESTS Different types of corrosion tests can be used for evaluating and understanding under-paint corrosion. Depending on the practical purpose, a corrosion test is designed to have one or all of the following properties: (1) to simulate corrosion in a real environment; (2) to accelerate the corrosion process; (3) to be reproducible; and (4) to provide an estimation of corrosion life in service [754]. However, it is usually very difficult to design a corrosion test that possesses all these properties. Practically, corrosion
318
CHAPTER II
tests can be divided into two groups: those conducted in environments in which the conditions, such as temperature, relative humidity, salt spray, spray solution concentration, etc., can be controlled; and those conducted in real environments in which the conditions cannot be controlled. Usually, the first group of tests provides better acceleration and reproducibility whereas the second group of tests provides better simulation and life estimation. Miyoshi [754], in a review article, classified corrosion test methods for painted automotive materials as follows: • Field survey: Used car bodies are evaluated for the degree of cosmetic damage and number of perforated spots. • Monitor car test: Cars are tested under designed driving conditions. • Proving ground test: Cars are driven on salty, dusty, and muddy roads in the daytime and kept in a humidity chamber overnight. • Vehicle exposure test: Cars are exposed at the seacoast and are periodically sprayed with a solution. • Under- and on-vehicle test: Test coupons are mounted under or on a vehicle to be driven under given conditions. • Laboratory tests: Two most frequently used tests are the salt spray test (SST) and the cyclic corrosion test (CCT). A CCT includes various factors such as salt spray, drying, humidifying, freezing, etc. There are many different CCTs. involving various combin.ations of temperature, humidity, solution composition, pressure, duration, chipping with stones, etc. (The test conditions for a typical CCT can be seen in Fig. 11.13.) A CCT is intended to accelerate as well as to simulate the corrosion processes occurring in a car body. Currently, almost all of the world's automakers use some form of laboratory cyclic corrosion test to evaluate coated steel sheet. In order to provide a reliable and universally accepted ranking of automotive sheet steel products for cosmetic corrosion, there has been an effort by the automotive and steel industries to develop a standard CCT. The effects of the various factors in such a standard accelerated corrosion test have been summarized by Townsend [1112]. Test specimens of various configurations may be used. depending on the practical purpose. Shaped structures and lapped panels are usually used for evaluation of perforation damage, and flat panels are mainly used for evaluation of cosmetic damage. Since under-paint corrosion usually starts at paint defects or damaged spots, defects are often artificially introduced prior to the test by scratching, chipping, cutting, etc. The common parameters for evaluating under-paint corrosion are the area of the rusted surface, amount of paint loss, penetration depth in steel, and creep length from the scratch. A cross-hatch test is normally used for evaluating wet adhesion of paint. In this test, the paint is scribed with perpendicular lines spaced a few millimeters apart. After a corrosion test, the number of the paint squares removed by an adhesive tape is used as an indication of the loss of adhesion [772,762]. Besides exposure-type tests, electrochemical measurements are often used to study the mechanistic aspects of under-paint corrosion by measuring paint porosity, diffusion, impedance, anodic and cathodic delamination, etc. [167,244,364,827,985].
UNDER-PAINT CORROSION
319
11.4. CORROSION BEHAVIOR
11.4.1. Characterization of Corrosion The corrosion of painted metals, particularly automotive bodies, is generally characterized as perforation corrosion or cosmetic corrosion. Corrosion of a painted steel sheet that initiates at an interior surface of a car body panel, penetrates the sheet, and eventually shows through as rust at the exterior exposed surface is known as perforation corrosion [lOOT]. It often occurs at locations that are difficult to clean, phosphate, and coat, such as lapped parts and hem flanges, or at crevices that collect dirt, salt, and moisture [1113, 754]. The term cosmetic corrosion is applied to an attack that is initiated at the exterior surface, usually at regions where the paint is damaged. Although this form of corrosion may eventually lead to perforation, the main concerns are with appearance. Cosmetic corrosion is usually related to (l) red rust-rust stain and bleeding at scratches in the paint; (2) paint creep-undercutting of the paint and loss of adhesion at scratches: and (3) chipping-removal of paint due to the combined effects of corrosion and Impact damage by stones and road debris. Simplistically, the directions of corrosion propagation in cosmetic corrosion and perforation corrosion can be described as being parallel to the surface and perpendicular to the surface, respectively, as shown in Fig. 11.2. J1.4.1.1. Peiforation Corrosion. The resistance of a painted material to perforation corrosion is usually evaluated by weight loss or depth of corrosion penetration. Figure 11.3 shows the depth of corrosion penetration for cold-rolled steel and several coated steels after a cyclic test [762]. The depth of penetration at the scribed line is much smaller for zinc- and zinc-alloy-coated steel than for noncoated steels. The corrosion penetration at the places where the paint is damaged, such as at the scribed lines, can be compared to the corrosion of an unpainted surface. Figure 11.4 shows that the weight losses of unpainted zinc-coated steel in a cyclic test are lower than those of cold-rolled steel and decrease with increasing coating weight [1072]. Similarly, Fig. 1l.5 shows that the corrosion penetration of the samples painted with a primer is decreased with increasing zinc coating weight [282]. Unlike cold-rolled steel, coated steel exhibits substantial corrosion penetration only after a period of incubation [339]. The incubation time for coated steel is attributed to the corrosion resistance and the galvanic action provided by zinc coatings. Thus, an increase in coating weight generally increases the incubation time and delays the perforation of the coated steel. After an incubation period, the coating is consumed, and the corrosion rate of coated steel becomes similar to that of cold-rolled steel. Pure zinc coatings are more effective than alloy coatings in reducing the corrosion penetration of the steel, owing to the stronger galvanic action between the steel and the coating.
steel FIGURE 11.2. Corrosion direction in perforation corrosion (P) and in cosmetic corrosion (C).
1I1I11I11I!jjillllJlLiJP'"
~
'"""11111111111"/
---'\~\~\T"""""'\ -\ \~\~\-r-\\r--<,\--'-\~\-"-\\r-.:\:--"-~.....--:\"<""""<"""~'-
coating .
crevice
CHAPTER II
320
perforation I
E E
S
a.
~ 0 .5 c: o
'iii
e o
u
CR
ZM
5Z
GA
FIGURE 11.3. Corrosion depth at scribe line on cathodic electropaint film after 60 cycles of cyclic corrosion test. CR, Cold-rolled steel; ZM, Zincrometal; SZ, Ni-Zn electroplated steel (20 glm2); GA, galvannealed steel (45 g/m2); G90, hot-dip galvanized steel (135 glm 2). Reprinted from Wakano et al. [762], with kind permission from The Iron and Steel Institute of Japan, Chiyoda-ku, Tokyo, Japan.
11.4.1,2. Cosmetic Corrosion. Cosmetic corrosion is most commonly evaluated by measuring the length of under-paint creeping. It is sometimes evaluated also by the extent of rust formation or paint loss [282, 762, 772, 1090]. Many factors, such as coating composition, coating thickness, surface treatment, test condition, type of paint damage, and type of paint, can affect the cosmetic corrosion of painted steels, as will be discussed in the following sections. Data from a field survey (Fig. 11.6), have shown that, in general, steel panels coated with zinc or zinc-alloy coatings have much slower under-paint creeping than cold-rolled steel [1072].
120.------------------------------------------, -CR
100
FIGURE 11.4. Effect of drying time ratio Rdry on corrosion weight loss of unpainted cold-rolled (CR), electrogalvanized (EG, two coating weights), and hot-dip galvanized (GI) steel sheets after three-week wet/dry cyclic test. Reprinted from Ito et al. [1072], with kind permission from The Iron and Steel Institute of Japan, Chiyoda-ku, Tokyo, Japan.
UNDER-PAINT CORROSION
321
I
0.4 , . . , C " " ' o " " ' l d - : - - - - - - - - - - - - - - - - - , . roll ed Elec tropla ted • 1"10 NaCI (;l • In- I 0 0.1"10 aC I 0.3 0 • '" C 0.01% aC I Ga lvannea led '"c: 0.2
.s ""
-><
a
c..>
FIGURE 11.5. Perforation test results for electrodeposition primed samples (dip phosphate, cathodic electrodeposition, 20 pm). After Iric [2821.
:£ -g
~
0.1
0
c..>
:J
~
CI:
0
0
20
•
4D
Galvani zed
60
80
t20
Coating weight (g/m')
11.4.2. Effect of Coating Type Many different zinc alloy and composite coatings have been developed for automotive applications. Table 11, I lists the major zinc- and zinc-alloy-coated steel products used for automotive bodies [1198]. The coated steel products used in automotive bodies must meet many engineering requirements [1007]. Some of the key properties of these products, besides corrosion resistance. are weldability, formability, bondability, and paintability. There is now a large amount of data indicating that almost all coated steels show improved resistance to cosmetic and perforation corrosion in comparison to cold-rolled steel. However, it has been noted that the differences between the various types of coated steels are not always significant, and the corrosion performance rankings for the various coatings reported by different investigators sometimes disagree [985]. In general, as shown in Fig. 11.7 [351], zinc and zinc-alloy coatings substantially reduce the blistering width compared to that for cold-rolled steel. Robbins et al. [3461 tested painted steel samples coated with alloys of various types and thicknesses and found that a pure zinc coating with heavier coating weight exhibited the best cosmetic corrosion resistance, The relative cosmetic corrosion resistance of the different zinc and alloy 6 r---------------------------------------~
.. Cold Rolled Steel .. G.lvanne.led Stee l
+ Electrogalv. St!.
Qj3
.~
:0
02 .s::
'5 1
.~
:J
_______
____
~
~o~------~----r-==~======~~~ o 2 3 Time (years)
FIGURE 11.6. Growth of underfilm corrosion of cold-rolled, galvannealed, and galvanized steel sheets used for monitor-car body outer panel. Reprinted from Ito e/ al. [1072], with kind permission from The Iron and Steel Institute of Japan , Chiyoda-ku, Tokyo, Japan.
322
CHAPTER 11 TABLE 11.1. Sheet Steels Used in Car Bodieso Cold-rolled steel Mild, unalloyed High strength, e.g., micro-alloyed, dual-phase, P-alloy Steel, hot-dipped Zn + 0.1-0.2% Al Zn + about 10% Fe !Galvanneall Steel, electroplated Zn Zn + about 10% Ni Zn + about 16% Fe Zn + about 16% Fe + FeZn topcoat (83% Fe) Steel, hot-dipped and electroplated Zn + about 16% Fe (Galvanneal) + FeZn topcoat (83% Fe) Steel, coated with conductive primer Zinc-rich primer (Zincrometal) Zn or ZnNi (electrolytic) + Zinc-rich paint (Bonazinc 2000) Zn or ZnNi (electrolytic) + organic coating (1 ,urn, e.g .. Durasteel) "Ref. 1198.
coatings, however, varies in different investigations. Johansson and Rendahl [1006] found that Zn and Zn-Fe coatings have better overall corrosion resistance in the hot -dipped state than in the electrolytic state [1006]. In another study, electrogalvanized zinc-iron and zinc coatings were reported to have better paint adhesion properties than hot-dip zinc and Zn-Fe coatings [1073]. Figure 11.7 indicates that the under-paint creeping after an outdoor exposure test is less for the zinc coating than for the Zn-Ni coating. However, in another study it was found that the creeping resistance of Zn-Ni-coated steel in atmospheric environments is
Zn-Ni
Free Zn
e~;~:nnoaled __ Zn·Co·Cr with AI~O~
1.5
2
2.5
3
3.5
Duration (year)
FIGURE 11.7. Growth of blistering on cold-rolled (CR), galvannealed, and zinc- and zinc-alloy-coated steels in the outdoor exposure test. Coating weight was 30 g/m 2. Reprinted from Yasuda et al. [351], with permission from ASM International.
UNDER-PAINT CORROSION
323
comparable to that of painted pure zinc coatings [283]. Van Ooij et ai. [984J compared the blistering of several zinc alloy coatings and found that (1) Zn-Ni coatings have excessive cracking sensitivity and a high rate of paint blistering; (2) no delamination and less blistering occur for a Zn-Fe coating; and (3) for Zn-AI coatings the blistering is extensive owing to poor phosphate adhesion. Figure 11.8 shows that 4% AI-O.I % Mg-misch metal prepainted steel sheet has a better resistance to creeping compared to zinc- and other Zn-AI-alloy-coated steels [1082]. The better corrosion resistance was attributed to magnesium depositing on the a-AI area of the eutectic phase and decreasing the crystallization rate of the Zn-AI eutectic phase. Various methods have been used to further increase the corrosion performance of coated steels. Robbins reported that trace amounts of Co codeposited in Zn-Ni coatings increase the corrosion resistance of the coatings [346]. Yasuda et at. [351 J reported that an AI 20 r dispersed Zn-Co-Cr composite coating exhibited excellent resistance to creeping. They attributed this resistance to the formation of more mechanically and chemically stable corrosion products and the anodic passivation brought about by Cr and AI 20 r • Shastry and Townsend J8l5J found that the presence of an electroplated Cr + CrO, overlay significantly improves the scribe creep resistance of zinc coatings, owing to the improved paint adhesion by mechanical interlocking of the primer to Cr + CrO, on zinc. in addition to the chemical stability ofCrO,. in the paint environment. Memmi et al. [340J reported that Cr-CrO,-topped zinc coatings performed significantly better in terms of the resistance to creeping, under a cyclic corrosion test. than simple Zn, Zn-Ni, and Zn-Fe coatings. For a Cr-CrO,-topped zinc coating, a phosphate treatment is not necessary because paint adhesion is ensured by the Cr-CrO, layer. In addition, the surface of this layer is chemically inert to phosphate solutions. Duplex coatings, such as Fe-ZnJZn-Ni and Zn-Ni/Fe-P, have also been developed [312, IllS]. Duplex coatings consist of a lower layer to provide good corrosion resistance and an upper layer to provide good paint adhesion. Kokubo [IllS] found that Zn-Fe/ZnNi two-layer electroplated steel sheet has better resistance than one-layer coated steel to paint blistering from a scribe, owing to the better corrosion resistance of the Zn-Ni coating and the better wet adhesion of the Zn-Fe coating. Other alloy and composite coatings such as Zn-Co [44.229.246], Zn-Mn [330, 349], Zn-Co-Mo [312], Zn-CoCr [425], Zn-Ni-Si0 2 [263], Zn-Ni-Ti [487], Zn-Si0 2 [284], and Zn-Co-Cr-AI 201 [351] have also been investigated. Coating Weight. The resistance to under-paint creeping for a given coating generally increases with coating weight [767,1064,1072,1073, 1090,1114]. Taking data reported by Ito et al. [I072J as an example, Fig. 1l.9 shows that the resistance of zinc and zinc-alloy-coated steels to creeping increases with coating weight. Franks [11 14] found that the length of under-paint creeping decreased approximately linearly with coating weight from 25 to 125 g/m2 in an on-vehicle exposure test. Davidson et ai. [1073 J reported that the scribe creepage was 37 mm for a 3.5-,um-thick coating and only 2.4 mm for a SS-,um-thick coating after a two-year outdoor exposure test. Light coating weight generally results in poor cosmetic corrosion resistance regardless of the type of paint defects [346,1073J.
o
os
,.-
1.5
+ Zn
·5AJMM
-~AJ
Exposure time (years)
Ichikawa-1
Ichikawa-2
FIGURE 11.8. Developmental changes in the width of blisters at the cut edge of prepainted steel sheets subjected to atmospheric exposure tests. MM, Misch metal. Reprinted from Masuhara and Kumon [1082], with kind permission from The Iron and Steel Institute of Japan, Chiyoda-ku, Tokyo, Japan.
iii
.~
~
.~
-c
-5
E 5
Kiryu
'M
>-
::0
tTl
""-l
n :c
~
N
325
UNDER-PAINT CORROSION
a: w
6
~
5
~~
4
f-
a5
..,
~'-' 3 Qw
•
3:rn ~iI 2 ::l~ ~~
~g
0
••
0
20
..
Zinc-Alloy Coated Steel
,~___ 40
Electrogalvanlzed --_ _--"S:...\.t:.:.,:ee'"-I_ _
8
60
80
100
COATING WEIGHT (g/cm 2 )
FIGURE 11.9. Etfect of coating weight of zinc-alloy-coated steel sheets on paint adhesion of monitor-car body outer panel after three years' running in Okinawa, Japan. Reprinted from Ito et al. r1072], with kind permission from The Iron and Steel Institute of Japan, Chiyoda-ku, Tokyo, Japan.
11.4.3. Effect of Test Conditions Currently, almost all of the world's automakers and steelmakers use some form of laboratory accelerated corrosion test for evaluating coated steel sheets [1112]. The various tests differ in the conditions employed, and thus the results vary accordingly. Under-paint creeping strongly depends on the test conditions such as temperature, rate of drying, impurity level. sequence and cyclic frequency, and type and duration of moisture, as reviewed by Miyoshi [754]. It has been established by many investigators [767,772, 827] that under a field exposure or in a cyclic test the under-paint creeping rate for zinc-coated steels is much lower than for cold-rolled steel, but under a salt spray test condition the converse can be observed, as shown in Fig. 11.10 [827]. Continuous salt spray testing accelerates the corrosion process, but it generally does not renect the corrosion performance of painted steels in real environments. An important factor in a cyclic test is periodic wetting and drying. Figure 11.11 shows that corrosion penetration of unpainted samples varies with the percentage of wetting time in the total time of testing [827]. A similar effect can be seen in Fig. 11.4, which reveals the important effect of drying on the amount of corrosion: a wet/dry ratio of around 0.5 seems to be the most severe condition, leading to the greatest weight loss of the unpainted steels in the cyclic corrosion test. Figure 11.12 illustrates the effect of wetness fraction on the creeping of painted samples [827]. It has been noted that for a given total time of wetness, the corrosion rates of unpainted zinc and steel samples are much higher when the samples are continuously wet than when they are periodically dried [1129]. The type of wetting is also significant in determining the corrosion behavior [980]. It was pointed out by Standish et al. [767] that corrosion product buildup, which occurs in atmospheric service, is prevented in a test in which the solution runs continuously, such as the salt spray test. Zhang and Tran [1117] found that variations in the proportion of spraying time and drying time result in significant changes in the amount of corrosion product formed on the surface. The frequency of cycling was reported by Kurokawa et al. [827] to exert a significant effect on corrosion penetration of cold-rolled and coated
~ 10
90
ok::: o
180
270
450
T ime, days
360
540
630
~
Seashore exposure
~ 720
0
10
20
Time, days
30
40
50
i~~-+--~--
SST
FIGURE 11. 10. Growth of blisters in Okinawa seashore exposure and salt spray test (SST). EG, Electrogalvanized steel; GA, galvannealed steel; CR, cold-rolled steel. From Kurokawa et al. [827]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
CD
]j 5
Q;
-~
i5
.J::.
-CR
·GA
+EG
-------------------------------------r-------------------------------------,
15 r.
tTl ;:0
-l
n >.."
:r:
N 0\
....
327
UNDER-PAINT CORROSION
~ 0.8
S E
E_ 0.6
c .2 ti
::l
'C ~
0.4
Wetness 100 fraction( %)
80
75
66
50
38
Corrosion SST tests
CCTI
CCTI
CCT3
CCT4
CCT5
Modified Volvo
'"
Okinawa
Seashore exposure
FIGURE 11.11. Thickness reduction of the bare steel (specimen coated with IO-flm electropaint film and with I O-mm-wide bare metal in the center) after one-month exposure in various corrosion tests. SST. Salt spray test; CCT, cyclic corrosion test; modified Volvo: seawater spray twice a week; Okinawa. seashore exposure on Japanese islands located in subtropical zone; EG. electrogalvanized steel; GA. galvannealed steel; CR, cold-rolled steel. From Kurokawa et al. [827]. © Copyright by NACE International. All Rights Reserved hy NACE; reprinted with permission.
6.0 .----------------------------------------, " CR
5 .0
Wetness 100 fraclion( %) Corrosion SST tests
80
75
66
50
38
ccn
ccn
CCTJ
CCT4
CCT5
Modified Volvo
Okinawa
Seashore exposure
FIGURE 11.12. Blister widths after one-month exposure in various corrosion tests. See caption to Fig. 11.11 for identification of tests and samples. After Kurokawa et al. [827].
328
CHAPTER II
Cyclic Test
40r-------------------------------,
Salt spray at 35 "C . 10 min.
E E
.p 30
Drying at 60 "C , 155 min.
Cll
cQ)
I
0.
I I
Q)
Wetting at 60 "C and 95 % RH. 75 min.
~ 20
u
...Q)
I I
~ ~
o 0.
Drying at 60 "C. 160 min.
~10 "0'
5 cycles OL-L-~·~·~·
o
___ L_ _ _ _ _ _ _ __ L_ _ _ _ _ _ _ _ 10
~
15
I
I
Wetting at 60 "C and 95 % RH, 80 min.
I
Cyclic test creep length Imm) FIGURE 11.13. Correlation between the results of a cyclic corrosion test (after 28 cycles, one cycle-24 h) and one-year Okinawa exposure test. Reprinted from Sakauchi and Kunimi [1074), with kind permission from The Iron and Steel Institute of Japan, Chiyoda-ku, Tokyo, Japan.
steels. The composition of the spraying solution also has a significant effect on the corrosion rate, as shown in Fig. 11.5. It is important for accelerated tests to produce results that show a definite correlation to those in a real environment. Much research effort has been made to develop such tests [754, 1112]. Figure 11.13, for example, shows that the results from a multistep cyclic corrosion test correlate well with those from a field exposure test [1074]. In simulated tests, wet/dry cycling is critical for creating corrosion products having a structure and morphology similar to those formed under natural environments. To more closely simulate the real corrosion performance of a car body, a chipping effect may also be included in a cyclic test. According to Gray and Shaffer [589], a cyclic corrosion test with a chipping effect is a more realistic test for assessing paint adhesion of coated steels. They found that the results of the chipping corrosion test show no linear relationship to those obtained from the scribed test and that the initial chipping damage is of critical importance in the corrosion performance [169, 338]. 11.4.4. Effect of Paint System The possible variations in paint formulation are infinite, and it is difficult to systematically classify the various paints according to their effect on the corrosion of the substrate materials. In general, corrosion of painted products occurs at places where the paint is damaged. The effect of a paint on corrosion is mainly determined by its internal strength, stability, and adhesion to the substrate, which are not only a function of paint formulation but also a function of the methods used in applying the paint. A paint has to have a certain thickness in order to furnish adequate protection against corrosion penetration, as shown in Fig. 11.14 [827]. A paint less than a few microns in
329
UNDER-PAINT CORROSION
perioration E E
~
.sc a
il
:> "0
0.5
~
C .><
u .s::.
I-
Paint thickness ,pm
FIGURE 11.14. Effect of paint thickness on the thickness reduction of the painted steels, measured after fi ve months' exposure in a c yclic corrosion test (wet: 66%). EG, Electrogalvanized stee l; GA , galvanllealed steel; CR. cold-rolled steel. From Kurokawa el (l1.1827J. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
thickness is not very effective in suppressing corrosion since the barrier effect of the paint against the permeation of oxygen, water, and other aggressive species increases with paint thickness [287, 827,1199]. Paint thickness also affects wet adhesion [772]. Mabuchi et al. [287] found that hydrophilic resins generally promote the formation of ZnCI 2 AZn(OHh, which shows good corrosion resistance, whereas hydrophobic resins promote the formation of ZnO, which is less protective. However, hydrophilic resins promote water permeation, which i sharmful to paint adhesion and corrosion performance [312,754]. It has been reported that the addition of a certain amount of silica strongly affects corrosion resistance by altering the types of corrosion products developed [287, 341]. Zinc-rich paint was found to reduce under-paint creeping of Zn-Ni coatings [768, 770] . van Ooij and Sabata [978] found that paint adhesion changes with changes in the curing conditions for the primer and with a rinse of the phosphated panel with silicates and silanes. Chromating before painting has been found to have various effects on the corrosion performance of coated steels [954, 120 I]. In one study, chromating of the zinc coatings before application of organic paints was found to increase the time to occurrence of red rust [287]. Chromate treatment of Zn-Ni alloy coatings prior to painting has been reported to enhance corrosion resistance [341]. However, it was reported in another study that chromate or nonchromate rinses have little effect on the scribe creep resistance of zinc or zinc-alloy coatings in a cyclic test [815] . 11.4.4.1. Phosphating . Surface treatments of zinc- and zinc-alloy-coated steels are particularly important for paint adhesion and corrosion performance [170, 762, 815]. Currently, zinc- and zinc-alloy-coated steels for car bodies are exclusively treated hy a zinc phosphate process. In an early study, LaQue and Boylan [755] reported that phosphating at least doubled the resistance against corrosion weight loss of various types of painted bare steels after exposure in a marine atmosphere. It becomes more beneficial
330
CHAPTER 11
H
H
100
~ .2
P
+P
100
Q)
I + Wet adhesion
~ c 'iii Q) .c
~
I
0
.Q
I .... P-ratio
50
SO
"0
'"
~
Cl.
a;
~
0
0
Zn
20
40
60
80
Fe content of Zn-Fe layer (%)
0 100 Fe
FIGURE 11.15. Wet adhesion. A 90-,um three-layer coating on Zn-Fe alloy (5-100% Fe) electroplated steels with a coating weight of 20 g/m 2 was evaluated with a cross-hatch test after IO days in distilled water at 40°C. Adhesion failure was measured by a video pattern analyzer. P ratio is defined as P/P+H, where P and H denote phosphophyllite and hopeite, respectively, and was determined from X-ray diffraction peak intensities. Data are taken from Ref. 772.
in cases in which the paint is damaged by scrubbing because phosphating improves the wet adhesion of paints [762, 772]. Phosphating is particularly effective on zinc coatings containing high levels of iron. Figure 11.15 shows that the wet adhesion of paint increases with iron content in Zn-Fe alloy-coated steel [772]. This was attributed to the fact that the amount of the adhesionenhancing phosphate formed increases with increasing iron content in the coating. The type of zinc phosphate tends to be mostly phosphophyllite for coatings with an iron content above 60% and to be mostly hopeite for coatings with an iron content below 20%. The adhesive property of a phosphated surface can be further improved by adding certain amounts of ionic species to control the composition and crystal size of the phosphate layer [1059]. The beneficial effect of phosphating is, however, not always observed because other factors may become predominant in determining adhesion strength. As reported by Hess and Soreide [1111], the phosphate coating weight, crystal structure. and size have little correlation with paint adhesion performance. Also, Shastry and Townsend [815J found that the scribe creep resistance of Zn-Ni and Zn-Fe alloy coatings is less sensitive to the differences in phosphate treatments than is the case with zinc coatings. Loss of paint adhesion can occur due to water penetration in the paint, with or without the effect of corrosion [754]. As reported by Wakano et ai. [762] and shown in Table 11.2, the loss of adhesion due to water penetration is generally more severe for zinc- and zinc-alloy-coated steels than for cold-rolled steel because phosphophyllite, which enhances paint adhesion, easily precipitates on a bare steel surface but not on a zinc alloy surface. Loss of adhesion was also found to be related to the phosphate coating weight and to the uniformity of the galvanized coating surface crystal orientation [1065]. 11.4.4.2. Effect of Paint Defects. Corrosion of painted structures usually starts at damaged areas. Test results have shown that under-paint creepage depends on the kind of
331
UNDER-PAINT CORROSION
TABLE 11.2. Results of Paint Adhesion Tests".!' Peeled area of paint film (%) Substrate
Test I
Test 2
0 0 0 0 0
Cold-rolled steel Electrogalvanized steel Gal vannealed steel Ni-Zn alloy electroplated steel Zincrometal
0 76
100 44
0
"Reprinted from Wakano et al. [762], with kind permission from The Iron and Steel Institute of Japan, Chiyoda-ku, Tokyo, Japan. ''Test I: As painted; test 2: after immersion in warm deionized water (40°C or SO°C) for 240 h.
mechanical damage [754]. There are several methods for generating paint damage: scrubbing, gravel chipping, arrow chipping, and diamond chipping, In one study, it was found that gravel chipping gave the largest creepage, followed by wide scrubbing, as shown in Fig. 11.16 [754]. The extent of paint damage caused by chipping depends on the type of paint, the paint thickness, and the substrate [169, 1075]. Standish et ai. [767] found that there is, in general, more creeping at scribes which cut deeply into the metal panels than at those which cut just through the paint. Vrable [588] found that deformation of zinc- and Zn-Fe-coated steels prior to painting, which causes cracks in the alloy coatings, has little effect on paint creeping performance in a cyclic corrosion test. Deformation of the painted galvanized steel, on
E
O CRS
$10 I-
f2 Zn
:u
~
~ Zn-Fe
E
.~
'0
..... .c
...
'0
'~
5 I-
Q)
01
III Q, Q)
~
u
o
~
a
b
c
~ d
e
Paint damaging method FIGURE 11.16. Influence of paint-damaging method on the creepage width or diameter after one-year weathering test ~f zinc-coated, zinc-alloy-coated, and cold-rolled (CR) steels with a salt solution spray. Coating weight, 40 g/m-. (a) Wide scribing (both sides); (b) narrow scribing (both sides); (c) diamond shot; (d) arrow chipping; (e) gravel chipping. Data are taken from Ref. 754.
332
CHAPTER II
the other hand, reduces the corrosion resistance, with complex strain conditions being worse than linear strain conditions [428]. Cosmetic corrosion of coated steel at welds has been studied by van Ooij et al. [986]. 11.4.5. Galvanic Action
Galvanic action between the coating and the steel substrate is one of the critical factors determining the corrosion behavior of the coated steels. As has been systematically discussed in Chapter 7, the galvanic action of a zinc-coated steel involves a large number of factors, and a specific test condition can be designed to evaluate each of these factors. Generally, the extent of galvanic action can be estimated by measuring the coupled potential and the galvanic current between the zinc coating and the steel, either painted or unpainted. Figure 11.17, as an example, shows the galvanic current between painted zinc-coated specimens and a bare cold-rolled steel specimen in 5% NaCl solution [827]. The painted zinc-coated specimen behaves as an anode while the painted bare steel behaves as a cathode. Thus, the corrosion of bare steel, such as at a paint scratch, in contact with a painted zinc-coated steel would be reduced by the galvanic action whereas in contact with a painted cold steel the corrosion would be enhanced by the galvanic action. Wakano et al. [762] reported that the galvanic current between a painted steel and a bare zinc-alloycoated steel increases with time. Jordan [286] found that the galvanic corrosion rate of a zinc coating is higher when coupled to iron corrosion products than when coupled to bare steel, indicating that the formation of steel corrosion products can accelerate the corrosion of zinc coatings at areas where the paint is damaged. This is, however, a short-lived transient effect that occurs during reduction of FeOOH [1299]. In a study by Sun and Tsujikawa [1261], it was found that the width of the under-paint corrosion of a zinc coating coupled to steel depended on the conditions of the cyclic test, such as the concentration of NaCI, the relative humidity, and the length of drying. The drying 10,--------------------------------------,
EtJ 0.1
~
+GA
~0.01
~ ~
.m
O~--------~~------------~~--------~
E
+~
Q)
§i-o.o (.)
-0.1 -1
3
5
7
9
11
Cycles (2 days for 1 cycle)
FIGURE 11.17. Galvanic currents between electropainted test specimen and cold-rolled sheet. EG, Electrogalvanized steel; GA, galvannealed steel; CR, cold-rolled steel. Data are taken from Ref. 827.
UNDER-PAINT CORROSION
333
condition in a cyclic test has a significant effect on the initiation of the under-paint galvanic corrosion of the zinc coating. Suzuki et at. [1241] determined the rest potential of various alloy coatings in 5% NaCI solutions and found that the galvanic throwing power was generally higher for a coating with a more negative potential. Massinon et al. [167] measured the galvanic protection life of various zinc- and zinc-alloy-coated steels coupled to bare cold-rolled steel in O.03M NaCI solution and found that the protection life generally increased with increasing coating weight. They also found that, as a result of cathodic protection, the steel becomes less efficient as the cathode owing to the precipitation of jJ-Zn(OHh or ZnO on the steel surface. Hayashi et al. [869] studied the relationship between the under-paint chloride penetration rate and the galvanic current for Zn-Fe alloy coatings. They found that the galvanic current was limited by the kinetics of charge transfer within the first 0.1 s and then was controlled by oxygen reduction at the cathode. The efficiency of Zn-Fe alloy coatings as a cathode increases with increasing iron content. The under-paint corrosion at the lap joints of galvanized steel over cold-rolled steel was found by Vincent and Coon [1199] to decrease with paint thickness. The galvanic action between the steel substrate and the zinc coating under paint occurs essentially in a thin layer of high-resistance electrolyte (within the paint film). According to Zhang and Valeriote [522], the area ofthe galvanically protected steel under a thin-layer electrolyte depends on the resistivity and thickness of the electrolyte, the distance between the steel and the zinc coating, and the area of the bare steel surface.
11.5. CORROSION MECHANISMS The mechanisms of under-paint corrosion for zinc- and zinc-alloy-coated steels have been the subject of many studies [167-169, 772, 869, 978, 983, 984]. The multicomponent interactions between paint, coating, steel, and environment make under-paint corrosion a very complicated phenomenon. In general, under-paint corrosion starts at places where the paint is damaged. The corrosion proceeds with corrosion of the coating or with paint delamination followed by corrosion of the substrate and, at longer times, leads to perforation of the steel. Generally, the corrosion products built up at the corrosion front may mechanically delaminate the paint. Delamination can occur at different interfaces in a paint-coating-steel system, depending on the materials and the environmental conditions, as schematically shown in Fig. 11.18. The causes of the delamination at the corrosion front, as reported by different investigators [167-169, 983, 984], can be physical, anodic, cathodic, mechanical, or the combinations of them. The predominant cause in a specific corrosion situation can be due to variations in paints r1081], coatings 1168,983], phosphates [772,978], and test conditions r168, 767, 827]. Shastry and Townsend [168] noted that significant differences exist in the mechanisms of paint creepage from the scribe on steel and coated steel substrates. In the case of a steel substrate, interfacial failure results from the dissolution of phosphate, whereas for zinc-coated steel, anodic dissolution of the coating seems to be the predominant factor controlling scribe creep. According to van Ooij et al. [983, 984], the expansion of the corrosion product causes delamination of paint. The rate of delamination is related to the anodic corrosion rate of the zinc coating. For Zn-Ni alloy coatings, the first step in the corrosion process is also delamination by anodic dissolution, but the dissolution rate is
334
CHAPTER II
FIGURE 11.18. Possible delamination paint phosphate _ coating
steel
~~~~~~~~~~
modes of a painted coated steel: (I) at paint/phosphate interface due to loss of adhesion; (2) within phosphate layer due to mechanical fracture; (3) at phosphate/coating interface due to dissolution of phosphate; (4) dissolution of coating: (5) at coating/steel interface due to mechanical failure.
lower owing to dezincification of the coatings, leading to a better barrier effect of the Ni -enriched layer and the fonnation of a more compact corrosion product. The differences in perfonnance of various types of alloy coatings are, to a large extent, due to the compactness and stability of the corrosion products formed in a corrosion environment [983]. Cathodic delamination has been identified as a common cause for advancement of the corrosion front [168, 772]. In a local corrosion cell at a scribe, the coating serves as the anode, and the surrounding area of the painted surface serves as the cathode. Because the cathodic reaction is primarily the reduction of oxygen diffusing through the paint, the high alkalinity resulting from the cathodic reaction can cause the dissolution of phosphate layers, resulting in paint delamination. Massinon and co-workers [167, 1081] found that cathodic paint delamination can be produced by passing a cathodic current through the phosphated galvanized steel. When the paint is thin or of a porous nature, oxygen and water can easily penetrate through the paint and react with the coating, thus causing a dissolution of the phosphate layer. Oxygen reduction at the delamination head is considered to be the rate-controlling reaction in the delamination process. Figure 11.19 schematically represents the corrosion process caused by a cathodic delamination. The galvanic action between zinc and steel is also found to cause paint/substrate interface deterioration due to a cathodic delamination [1084]. Thus, the ratio of the area of the steel cathode to that of the coating anode at a micro defect may determine the extent of blistering [1083]. The advancement of under-paint creeping can also be directly caused by anodic dissolution of the zinc coating [168, 827, 1072]. Kurokawa et al. [827] found that, in a cyclic corrosion test, the anodic dissolution front under the paint on the coated steel is acidic while the corrosion product behind the dissolution front is basic. According to
Phosphate _ layer
FIGURE 11.19. Schematic representation of the degradation under a paint film. After Massinon el ai. [167) .
UNDER-PAINT CORROSION
335
van Ooij et al. [994], the mechanism of under-paint corrosion propagation is initially anodic, i.e., dissolution of the coating, for all systems. Mechanical paint delamination, due to corrosion product buildup, has been identified as a cause of delamination between phosphate and coating surfaces in many circumstances [983, 984]. Standish et al. [767] found that corrosion buildup, which results in the delamination occurring within zinc phosphate crystals, is the cause of lateral spread of corrosion in painted steels. The better performance of galvanized steel as compared to that of cold-rolled steel is attributed to the lower volume and the barrier effect of the zinc corrosion products under a cyclic or a natural exposure condition. Standish et al. also found that corrosion product does not build up beneath the paint in a salt spray test, while it does in cyclic or natural exposures. van Ooij and co-workers [978, 994] noted that the adhesion of paint does not strongly affect corrosion creeping at damaged areas such as at scribes of zinc-coated steels. According to their corrosion model, as shown in Fig. 11.20, the creeping process (at a scratch in a wet/dry test) starts with anodic dissolution of the exposed zinc layer, with the bare steel as the cathode. The gradual formation of zinc corrosion product causes a wedging effect to delaminate the phosphate crystals at the zinc and phosphate interphase. The cathodic area in this initial stage is behind the corrosion front. As the corrosion front only
~~Iili--~"""'''''U phosphate
'§iimr~
Attacked phosphate above chloride
Chloride only
Phosphate
~
I~~~~~~~~
. pro ducl / ' " CorroSIOn
~~~$¥~~~~~~ Scribe FIGURE 11.20. Schematic lOp view and cross section of cosmetic corrosion propagation mechanism in painted precoated steels. From van Ooij and Sabata [311. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
336
CHAPTER 11
propagates farther from the scratch, small corrosion cells form ahead of the corrosion front, and, when the corrosive species start to diffuse through the paint in quantity, these areas start to serve as the major cathodic areas on which an alkaline environment is generated. At this stage, the cathodic area is ahead of the corrosion front. This results in a partial, and eventually a complete, alkaline dissolution of the phosphate crystals. Once the cathodic area is ahead of the corrosion front, and dissolution of phosphate crystals occurs, the overall rate of under-film corrosion is increased.
12 Zinc-Rich Coatings 12.1. INTRODUCTION Zinc-rich coatings, which contain zinc dust and a binding medium, have been widely used since the 1930s for the protection of steel structures from corrosion [345, 1159). They can be used as a primer or as a final coating and can be applied by spray, dip, or brush. As coatings, they possess a number of advantages over galvanized coatings and regular organic paints: they are easier to apply than galvanized coatings, particularly for large or odd-shaped articles, and, unlike regular organic paints, they provide galvanic protection at edges or the places where the paint coating is defective. In contrast to other types of applications of zinc, where the resistance to corrosion is of primary importance, the role of zinc dust in zinc-rich paints is not to resist corrosion but to react with moisture and oxygen, especially in fresh coatings, and in the process to strengthen the paint and to galvanically protect the coated steel. The corrosion of zinc dust is thus an essential part of the corrosion protection mechanism of a zinc-rich paint. The material presented in this chapter is organized into three sections. First, the basic components and characteristics of zinc-rich paints are briefly described. The corrosion protection mechanism is then discussed, and finally the effects of various factors on the performance of zinc-rich paint are presented. 12.2. COATING CHARACTERISTICS There are basically two broad categories of zinc-rich coatings-organic and inorganic. The classification of a coating as organic or inorganic depends upon the chemical nature of the binder used to bond the zinc particles to each other and to the steel surface (1159). The organic type uses binder materials such as phenoxies, catalyzed epoxies, urethanes, chlorinated rubber, silicones, or vinyls. The inorganic type usually employs sodium silicate or ethyl silicate as the binder. Potassium, lithium, and ammonium silicates are also used in inorganic coatings. Most modem inorganic zinc-rich coatings are of the self-curing, as opposed to the postcuring, type. They are further divided into water-reducible and solvent-reducible. Generally, those using alkali metal silicates are in the water-reducible group, which cure during and after evaporation of water from the coating. The solvent-reducible group consists primarily of the partially hydrolyzed alkyl silicates. A special case involves 337
338
CHAPTER 12
organic silicate vehicles, such as partially hydrolyzed tetraethyl silicate. When such a vehicle cures, the organic portion is hydrolyzed or evaporated, leaving an inorganic silicate film. Because galvanic action is an important protection mechanism of zinc-rich coatings, both organic and inorganic types of coatings must be electrically conductive to ensure galvanic action between the zinc particles and the steel substrate. Binders of either type do not have good electrical conductivity. Thus, the coatings must rely on a heavy load of fine zinc particles in contact with each other to provide the electrical continuity required for galvanic action to occur. Additives and extenders are also added to zinc-rich coatings to obtain specific effects. For example, addition of iron phosphate at levels of 10-20% improves the weldability of paints such as epoxy ester paints [1159]. A small amount of carbon black is sometimes added to enhance the conductivity. Red or yellow iron oxides may be added to give the paint tint. Addition of a small quantity of calcium oxide can reduce gassing from the reaction between zinc dust and water or acid. Zinc oxide, when added to zinc-rich paints at levels of about 20%, has been found to be of special value when the paints are to be applied to galvanized steel, which is more difficult to paint than steel because of adhesion problems. Zinc oxide has also the effect of maintaining coating flexibility. The maintenance of mechanical properties by zinc oxide is attributed to the absorption of ultraviolet light and to its reaction with organic decomposition products. which tend to catalyze further degradation [569]. The chemical effect of zinc oxide in a coating can also be due to pH buffering [839]. Other additives such as iron and lead have been used in epoxy-based coatings to control operation conditions [507]. A sheet steel coated with zinc-rich coating, known as Zincrometal, has been widely used in the automotive industry [5, 1159]. The Zincrometal coating consists of a basecoat called Dacromet and a topcoat called Zincromet. Dacromet is a water-based pretreatment consisting of a combination of zinc dust. chromic acid, and other chemicals. Zincromet is a solvent-based zinc-rich organic coating consisting of approximately 50 vol % (85 wt. %) zinc dust contained in a thermoplastic phenoxy resin. The typical thickness of the coating is about 12 ,urn. Zinc-rich coatings are most often used as primers. They are topcoated with other paints, not containing zinc dust in many cases, to provide longer lasting protection. Generally, organic zinc-rich paints require less rigorous surface preparation than inorganic zinc-rich paints. They are more compatible with a variety of topcoats. more flexible, and less dependent on atmospheric moisture during curing but more difficult to apply under high or low ambient temperatures, as compared to the inorganic ones [1159]. On the other hand, inorganic coatings generally provide better corrosion protection, better toughness and abrasion resistance, faster drying under optimum conditions, and better temperature resistance. Different zinc-rich coatings, depending on the type of binder, offer varying degrees of advantages in terms of economics, amount of protection afforded, and ease of application and topcoating [506]. The formation process of an inorganic zinc-rich coating is very different from that of organic coatings. According to Munger [345], several reactions are involved in the formation process of inorganic zinc-rich coatings. The first reaction is concentration of the silicates by evaporation of the solvent, providing the initial drying and primary deposition of the coating. The second reaction is insolubilization of the silicate matrix by
ZINC-RICH COATINGS
339
TABLE 12.1. Oxygen Permeability of Zinc-Rich Paints (72% Zinc by Volume)" -----------------------------------------------Oxygen reduction CUiTent (flA) Fresh paint Paint type
Dry
Aged paint" Wet
Dry
Wet
---------------------------~------------------------~------------------
Chlorinated rubber, 125-175 flm thick Epoxy polyamide, 100-150 11m thick
104
1.4
28
1.0
16
0.2
12
0.2
"Data from Ref. 726. /, Aged by immersion in O.SM Nae] for six months.
reaction with zinc ions from the surface of zinc particles and iron ions from the steel surface. At this stage, the vehicle hydrolyzes to a form of silicic acid with which zinc ions react to form an insoluble silica-oxygen-zinc polymer. At the zinc silicate/steel interface, an iron-zinc-silicate complex is formed. The zinc silicate and iron-zinc-silicate matrix, surrounding the surfaces of zinc particles and the steel substrate, is very inert to water, weather conditions, solvents, and many other chemicals. The third reaction that takes place within the coating occurs over a long period of time--a number of months or years. During this time, a reaction of zinc compounds with carbon dioxide from the air results in formation of zinc carbonate, which fills the pores within the coating. In addition, a reaction of the alkali in the coating with the moisture from the air releases silicic acid in the silicate polymer. This reaction gradually proceeds through the coating to the steel interface, increasing the adhesion of the coating to the steel. The permeability of zinc-rich coatings is affected by many factors. The data in Table 12.1, reported by Ross and Wolstenholme [726], show that dry paints are much more permeable to oxygen than wet paints and that the chlorinated rubber-based paint is more permeable than the epoxy polyamide-based one. Aged paints have lower permeability than fresh paints owing to the blockage of pores inside the paint by corrosion products. Also, the permeability of the chlorinated rubber coatings to Cl- ions was found to be much lower than that of the epoxy coatings. 12.3. PROTECTION MECHANISM The protection of steel by a zinc-rich coating is mainly via two effects: the barrier effect and galvanic action. While the coating serves as a physical barrier to the environment in which the coated steel is used, the zinc particles in the coating provide the galvanic protection. In addition to the galvanic action and the barrier effect of zinc-rich coatings, two other effects that are also beneficial in terms of corrosion protection have been identified: (a) the absorption of oxygen by the zinc powder and (b) the inhibition of the oxygen reduction by the zinc corrosion products on the steel surface [1191]. The remaining zinc particles, after their insulation from the steel surface, later react with oxygen and thus reduce the amount of oxygen, arriving at the steel surface. In the case of
340
CHAPTER 12
the oxygen that does arrive at the steel surface, the reduction reaction is hindered by the zinc corrosion products. In order for a galvanic process to occur on a steel with a painted zinc-rich coating, three conditions must be satisfied: 1. The zinc particles in the coating must be in electrical contact with each other. 2. The zinc particles must be in electrical contact with the steel. 3. A continuous electrolyte must exist between the zinc particles and the steel. The first two conditions are met by zinc-rich coatings containing a sufficiently high zinc content. The third condition is fulfilled when a steel panel bearing a zinc-rich coating is wetted by a film of electrolyte such as a salt solution. There are two stages in the protective action of zinc-rich coatings [5,726,780]. The first stage is a relatively short period in which galvanic protection of the steel by zinc particles is in effect. After this period, the galvanic action between the steel and zinc gradually disappears. The second stage is a long-term barrier protection that is attributed to (1) greater resistance of the coating to the permeation of aggressive species such as water, oxygen, and salts because the pores in the coating are blocked by the zinc corrosion products or (2) inhibition of the steel surface by the zinc corrosion products. The galvanic action generally decreases with time. The loss of galvanic protection is due to (1) the loss of electrical contact between zinc and steel as a result of corrosion of zinc particles and formation of nonconductive corrosion products at the interface; (2) the loss of electrical contact between zinc particles as a result of the formation of corrosion products on the surface of the zinc particles; or (3) blockage of the coating surface by zinc corrosion products. However, the galvanic protective action of a zinc-rich paint at a later stage is still latently present and may activate at any time after mechanical damage of the paint or chemical removal of the barrier effect of the corrosion products. For example, a commercial zinc-rich paint, after being weathered for 10 years, had a layer of zinc corrosion products removed and afterward showed quite a normal galvanic action [780]. The galvanic effect of a coating can be evaluated by measuring the potential of a coated steel sample having a certain area of bare steel surface or by measuring the current between a coated sample and an uncoated sample [780,965, 1172]. Alternatively, galvanic protection can be evaluated by visually observing signs of red rust on a coated sample at cuts or other defects on the coating. The barrier protection of zinc-rich coatings is often evaluated by the use of electrochemical impedance techniques. Impedance spectra of zinc-rich coatings generally display a behavior corresponding to a high-frequency charge-transfer process in series with a low-frequency diffusion process [432, 805]. Owing to the formation of corrosion products which block the pores inside the paints, the impedance of many zinc-rich coatings gradually increases with exposure time in the testing environment [432, 805, 813, 1160]. Thus, in general, the barrier effect of zinc-rich coatings on the diffusion or permeation of aggressive species increases with time. Armas et al. [805] reported that the impedance of steel coated with zinc-rich paint decreased with immersion time in artificial seawater. Pereira et at. [432] found that the high-frequency impedance of a polyamide cured epoxy and an ethyl silicate zinc-rich paint in a marine environment was associated with charge-transfer reactions and increa~ed with time. Feliu et at. [1160] measured the
ZINC-RICH COATINGS
341
impedance of several zinc-rich paints after three years of atmospheric exposure and found that the barrier effect of the paints increased with exposure time. They also found that the paints still maintained a large part of their capacity for providing cathodic protection, especially in the case of an ethyl silicate vehicle. 12.4. PERFORMANCE
12.4.1. Effect oJZinc Content Zinc dust content and particle size are two important factors determining the extent of galvanic protection provided by a zinc-rich coating, through their effects on the specific surface area and on the electrical continuity between the steel substrate and zinc particles and also among zinc particles. In general, for each type of zinc-rich coating, there is a certain zinc level below which the galvanic protection of the steel substrate by the coating is not effective. Figure 12.1 shows that the resistance to red rust of steel samples coated with two inorganic coatings under salt fog greatly increases with zinc content, particularly in the range of 40-60% by volume [513]. Table 12.2 shows that the effectiveness of inorganic zinc-rich coatings on a steel surface with mill scale increases with zinc content [1162]. Ross and Wolstenholme [726] examined the effect of zinc content on the galvanic protection of a bare steel area on steel panels coated with zinc-rich paints. They found that the coatings with low zinc contents provided no cathodic protection to the bare steel surface. The effect of zinc dust content on the protection of steel strongly depends on the type of binder and pigment materials [506,513,805]. Theiler [780] found that there exists an optimum zinc content level for each binder material that yields the longest cathodic protection for damaged coating spots, as shown in Fig. 12.2. According to Theiler, if too much zinc is incorporated into the paint, then the anodic area is large in comparison with the cathodic area, and, as a result, a small current density is drawn from the active zinc.
10
I'"
n. o
§
8
6
OJ
~
'"
4
Rusting: 10 = no rust completely rusted
o=
~-------------~~~------
~
-Pigment A • Pigment B
IT:
2
~~0~-----------5LO-------------6~0------------~70 % Zn in dry film
FIGURE 12.1. Results of 4000-hour salt fog tests on primers made from silicate vehicle with two types of pigments and varying levels of zinc (vol. 'Yo) in the dry film. Data are taken from Ref. 513.
342
CHAPTER 12
TABLE 12.2. Percent Zinc in the Dry Film of Inorganic Coatings vs. Performance over Mill Scale" Zinc level (wt. %)
Dry film thickness (mil)
o
2.5
20 40 60
2.6
70 85
2.2 2.5
87.5
2.4
Time to failure h 2 months 2 months 2 months 4 ('ears 52 years No failure, less than I % rust after ~ years No failure, less than I % rust ~ years
2.1 2.3
"From Montie and Hasser [1162). © Copyright by NACE International. All Rights Reserved by NACE; reprimed with permission.
The pH of the electrolyte in the vicinity of the zinc particles does not change under this low current density. The zinc surface is thus passivated by the formation of insoluble corrosion products in the neutral or slightly basic electrolyte and loses its protective action after a short time. On the other hand, if there are too few zinc particles present within the paint, the anodic area is very small and the anodic current is very high. The active zinc is consumed very rapidly, and the cathodic protection is quickly lost. The available zinc content can also be evaluated by measuring the apparent polarization resistance. Figure 12.3 shows that the apparent polarization resistance of steel samples coated with zinc-rich paints decreases with increasing zinc content. The total zinc surface area is increased owing to the increase in the total number of zinc particles, which results in more contacts among particles [780]. Thomas et at. [996] determined the available zinc content (the amount of zinc available to afford cathodic protection) in zinc-rich coatings by measuring the galvanic current between painted and bare steel samples. They defined the amount of zinc available 40 r-----------------------------------~
.,"'>-
'C
~
C 30 .9
POlystyrene
.. Chlorinated rubber
+ Vinyl copolymer
(3
o Q)
~20
'6 a .c:
iii
o '010 QI
E
i=
O~----~------~-------L------~------~
50
55
60
65
Zinc content, vol%
70
75
FIGURE 12.2. Time of cathodic protection as a function of zinc content of zinc-rich paints with different binders under immersed conditions. Reprinted from Theiler [780], with kind permission from Elsevier Science Ltd, The Boulevard, Langford Lane, Kidlington OX5 1GB, United Kingdom.
ZINC-RICH COATINGS
343
100,000.-------------------------------------- , ~
Chlorinated rubber
s:}-
Polystyrene
+ Vinyl copolymer
10,000 NE
:-'
c:
1,000
0:0.
100
10L---------~----------~----------~~
50
60
70
80
Zinc content, vol%
FIGURE 12.3. Apparent polarization resistance in relation to the zinc dust content for zinc-rich paints with different binders. Data are taken from Ref. 780.
for galvanic protection as that which is in direct contact with the steel substrate; hence. the available zinc content is independent of coating thickness. Field exposure data show that paints having less than I 0% available zinc fail to prevent rust formation in scribe lines at 6 months whereas those with greater available zinc are all successful in preventing red rust formation at 12 months. The variation of available zinc with vehicle type was explained on the basis that some vehicles are more efficient wetting agents than others. 12.4.2. Effect oJZinc Particle Size
Zinc dust can be classified as regular. fine. or super fine. The typical properties of each class are given in Table 12.3 [822. 1159J. There is a logarithmically linear relation between the mean diameter and the specific area of zinc dust samples. Zinc particle size affects the zinc content available for cathodic protection [996J. Figure 12.4 indicates that the available zinc content is the highest for particle sizes between 2.5 and 3 ,urn. It has been reported that a particle size between 5 and 7 ,urn represents a good combination of particle-to-particle contact and specific area [506J. In one study, the impedance responses of two epoxy zinc-rich paints were found to depend strongly on the particle size distribution [432J. In another study, it was reported that a
TABLE 12.3. Typical Properties of the Three Classes of Zinc Dust" Property Median diameter (pm) Percent smaller than 10 f1.m Percent under 44 f1.m Surface area (m 2/g) Apparent density (glcm3) "Ref. 822.
Regular 8.0 65.0 98.0 0.1 3.04
Fine 5.0 95.0 99.0 0.16 2.4
Super fine 4.0 99.0 100.0 0.2 2.2
344
CHAPTER 12
+
30 ill
:0 C1l
.(ij
~
+
+
+
20
+
u c:
N
'0
*" 10 O~--------------~---------------L------~
2
3
4
Pigment fisher size, J.lm
FIGURE 12.4. Effect of particle size on percent available zinc for cathodic protection in silicate vehicle. Data are taken from Ref. 996.
particle size of 2.5 .urn provides the best protection in a flowing electrolyte [728]. There exists a balance between the magnitude and length of galvanic action. The optimum particle size should be that which provides the right balance for the specific application. 12.4.3. Effect of Binders Binder material is one of the most important factors determining the effectiveness of the corrosion protection provided by a zinc-rich coating. Figure 12.2 shows that the effectiveness of galvanic protection can differ greatly among coatings using different binders. The apparent polarization resistance for a given zinc content also varies with the type of binder material, as shown in Fig. 12.3. The polarization resistance was considered by Theiler [780] to be related to the wettability of the binder to the zinc particles: the better the wetting, the less the active zinc surface area, and thus the larger the polarization resistance. More particles are sealed by the binders with better wetting properties and thus are not available for reactions with the electrolyte. Feliu et al. [1160] measured the electrochemical behavior of zinc-rich paints exposed in the field for up to three years. They found that the coatings retained a considerable amount of their galvanic protection capacity, especially those of the ethyl silicate type, and that the barrier effect increased with time, particularly for coatings of the epoxypolyamide type. Lindqvist et at. [965] reported that commercial ethyl silicate-based paints provided much longer cathodic protective action than epoxy-based ones in a 60-day immersion test in O.IM NaCI solution. Some coatings still provide galvanic protection when a topcoat is applied, while others do not. In one study [726], it was reported that a zinc-rich coating based on chlorinated rubber with a top coating provides cathodic protection to a bare steel surface at defective spots in the coating, while one based on epoxy polyamide does not; on the other hand, when the coatings are not topcoated, both zinc-rich paints provide cathodic
ZINC-RICH COATINGS
345
-0.6r----------------------., a - Initial immersion b - After 1 day c - After 3 days d - After 14 days
-0.7r-_ _ __ w
~
-0.8
~
~
C 0
-0.9
Q)
c..
a
-1 -1.1 0
2
3
4
5
Current, f.lA
FIGURE 12.5. Polarization curves for the galvanic couple comprised of a steel specimen coated with 73% (by volume) zinc-dust paint with chlorinated rubber binder coupled at the centerto a piece of bare steel (0.072-1.27 cm 2 in area) measured in 0.5M NaC!. Reprinted from Ross and Wolstenholme [726], with kind permission from Elsevier Science Ltd, The Boulevard. Langford Lane, Kidlington OX5 1GB, United Kingdom.
protection to the bare steel area. Because of the lesser galvanic action between the bare steel and the painted area for the epoxy polyamide paint, there is also less deposition of the corrosion product on the steel surface than in the case of the chlorinated rubber paint. In addition, it was found that the corrosion product deposited on the steel surface has an effect of cathodic inhibition. Figure 12.5 shows that for a chlorinated rubber paint immersed in 0.5M NaCl solution for 14 days, the open-circuit potential of the steel significantly shifted in the cathodic direction. The fact that the open-circuit potential of the steel after 3 days of immersion was more negative than -900 m V seE indicates that the corrosion products deposited on the steel surface probably contained some metallic zinc.
12.4.4. Effect of Coating Thickness In general, the life of a zinc-rich coating increases with coating thickness. Figure 12.6 shows an example for an organic coating in an accelerated test [996]. As another example, Table 12.4 shows the effect of coating thickness on the corrosion performance in salt fog for an ethyl silicate zinc-rich primer [1162]. According to Thomas et al. [996], the amount of zinc particles available for galvanic protection is the amount that is in direct contact with the substrate. Since the amount of zinc particles in direct contact with the substrate does not vary substantially with coating thickness, the increase in the coating life with increasing thickness must be associated with a stronger barrier effect for the thicker coatings.
12.4.5. Effect of Additives Fawcett et al. [10 10] investigated the effect of Fe2P as a pigment on the corrosion performance of an epoxy ester zinc-rich coating. Figure 12.7 illustrates that a pigment with a composition of 50 wt. % Fe2P and 50 wt. % zinc provides the best performance in a salt spray test. The beneficial effect of Fe 2P is associated with Fe2P serving as cathodic areas for O 2 reduction. In another study [1026], Fawcett et al. found that the presence of
346
CHAPTER 12
300,---------------------------------------, (I)
:;
• With FS 3.7/1 dust
250
'" With FS 2.4/1 dust
o
.s=: ai 200
E ~
:=
150
"t:l Q)
Cil
Q; 100
a;
() ()
50 o~----------~----------~------------~
o
0.05
0.15
0.1
Coating weight, g / em'
FIGURE 12.6. Effect of thickness of zinc-rich coatings on accelerated life for polystyrene vehicle. Data are taken from Ref. 996.
ZnO in zinc dust at levels between 18 and 27% is beneficial to the performance of the coating. The effect of zinc oxide was attributed to the reduction of the total anodic activity of the zinc dust, thus balancing the cathodic and anodic activities. Chua and Ross [729] also reported that the addition of zinc oxide reduces the zinc content in the coating without reducing its functional properties. The effect of the addition of pigments on the performance of zinc-rich coatings is complicated. Fragata and Mussoi [813] found that the presence of extender pigments affects only the physical and chemical characteristics of zinc-rich coatings, but not the galvanic protection. However, in some cases, as shown in Fig. 12.1, the corrosion resistance of zinc-rich coatings may vary significantly with changes in the type of pigment used [513]. Szauer and Miszczyk [1161] modified zinc-rich paint by replacing a small percentage of zinc with carbon black or zinc phosphate. Carbon black was found to decrease the duration of both the cathodic protection and the long-term barrier protection, whereas
TABLE 12.4. Ratings for Inorganic Zinc-Rich Coatings of Different Thicknesses after Salt Fog Exposurea .b Months in test
Dry film thickness (urn) 5.33 7.11 10.67 12.45
9 9 10 10
2
4
9
15
6 7 10 10
2 3 7 7
4 5
1 4 3
"From Montie and Hasser [1162]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission. "Rating scale ranges from 10 (best) to 0 (poor).
ZINC-RICH COATINGS
347
8,------ - - -- - - -- -- -- -- - - -- - - -------1
6
60
20
100
80
Wt % Zn in p igment
FIGURE 12.7. Degree of rust on steel substrate after 200-hour salt fog test at 50 (mixture of zinc and Fe2P)' After Fawcett el al. [10101.
0
e VS. zinc content of pigment
zinc phosphate was found to have no adverse effect on the period of cathodic protection but increased the period of the long-term barrier effect. 12.4.6. Effect qf Surface Condition Mayne [1054] studied the effect of surface condition on the galvanic protection of steel provided by a zinc-rich coating. Figure 12.8 shows that the coating applied on a rusty steel surface (exposed to the atmosphere for one year) provided less cathodic protection to the steel than that applied on a clean surface (pickled). It was also found that the coating on mill scale, which was removed to various extents through weathering, -0.3 .----------------------------~P~ic~k7Ie-d~--------;
- - - Unpa inted specimens
> Qj
E
-
-0.5
o
m
•
Rusty. wire-brushed
+ Rusty
Painted specimens
~,
... Electrolytio
zinc
T _
... T ___ -..- _ _ _ •
(3
__ .... _____ ... ...
..
::.:: -0.7
"lii
!e. ~
~ -0.9
a a.
-1 .1
--- . --- .- -- .---- -~--- .---. L -____L-____
o
10
~
20
____
~
30
____
~
40
_____ L_ _ _ _
50
~
_ __ _
60
~
70
Time, days
FIGURE 12.8. Potential-time curves for painted and unpainted steel specimens partially immersed in natural seawater. Painted specimens are steel samples with various surface conditions (indicated on the figure) on which a zinc-rich coating was applied. Data are taken from Ref. 1054.
348
CHAPTER 12
provided cathodic protection to the steel. Two possible explanations were suggested: either (a) the rust on the surface may not have been continuous so that zinc particles may have been in contact with the steel surface or (b) the rust may have been reduced to a certain extent by the zinc, thereby becoming conductive. Long-term exposure showed that steel samples that had been weathered for six months before being coated remained in good condition after four years of atinospheric exposure. According to Mayne. the best performance was obtained when the exposed steel was coated in the summer. Hendry [513] reported that zinc-rich coatings over mill scale perform well.
12.4.7. Other Factors Figure 12.9 shows the potential of steel samples coated with four commercial ethyl silicate zinc-rich coatings immersed in O.IM NaCl solution. For the samples with 91 % and 87% zinc, the galvanic effect only lasted for about 10 days, while for the samples with 83% and 93% zinc it lasted for more than 60 days. indicating that factors other than the base binder material and zinc content can also be important in determining the length of galvanic protection by a zinc-rich coating. Tanabe et al. [743] reported that zinc-rich paints, when used as primers. greatly increase the durability of epoxy top coatings, as shown in Fig. 12.10. Zinc-rich coatings when used without a topcoat tend to perform better with higher zinc loadings in severe conditions while coatings with lower zinc loadings tend to perform better when topcoated in less severe conditions [506]. Generally, a topcoat reduces the galvanic activity between the zinc particles and the steel substrate [726]. Ross and co-workers [727-729, 731] investigated the effect of solution flow rate on the galvanic protection of a zinc-rich coating on steel, using a rotating coated sample with a bare surface in the center. They noted that the anodic current of the zinc-rich coating increased significantly with increasing flow rate of the 0.5% NaCl solution. At high flow rates, an ennoblement of the zinc-rich coating was observed. The rise in the potential was -0.5r------------------------, 91 wt%
-0.6
67 wt%
"""01-0.7
~
]! -0.8
"E Ql
63 wt"A.
(5
c.. -0.9
93 wt%
solid zinc
-1
-1.1~--~--~--~--~--~--~-~
o
10
20
30
40
50
60
Time, days
FIGURE 12.9. Electrode potential as a function of time during exposure in aerated O.IM sodium chloride solution at 25°C and pH 7 for four commercial ethyl silicate zinc-rich paints on steel and for solid zinc metal. The zinc contents of the paints are labeled on the curves. After Lindqvist et al. [965].
ZINC-RICH COATINGS
349
1,000,----------------------------------------,
100 3 (j)
10
u
C
u;
'0;
~
c
4
o
~ NO.1
.§ o
0... 0.01L-------------------~---------L--------~
o
100
200
300
400
Immersion time, days
FIGURE 12.10. The change of polarization resistance of epoxy-coated steel with immersion time in 3 wI. % sodium chloride solution. (1) and (3) Zinc-rich paint used as a primer; (2) and (4) without zinc-rich paint. After Tanabe et al. [743 J.
attributed to the formation of passive zinc corrosion products caused by cathodically produced hydroxyl ions, which flow from the center to the edge of the plate. An annular band, the size of which depended on the flow rate, of thin gray corrosion product was also formed on the steel adjacent to the surrounding coated surface. while a brown stain was formed in the center of the bare steel area.
13 Corrosion in Concrete 13.1. INTRODUCTION Steel-reinforced concrete is a widely used construction material. In the absence of certain deleterious factors and agents, steel does not rust in the concrete environment. However, in practice, the entrance of moisture, salts, and oxygen by diffusion or through pores or hairline cracks in the concrete may cause the steel to corrode and may eventually result in cracking and spalling of the concrete by the expansive force of the corrosion products [269, 1219]. The problems resulting from corrosion of reinforcing steel in concrete are widespread and very serious. This is especially true for the bridge decks on highways where deicing salts are used. Beginning at about the end of the 1960s, severe deterioration of many reinforced concrete bridge decks in the "snow belt" states of the United States was noted. Later, assessment of the bridges constructed in the 1960s and 1970s with black steel reinforcement showed a drastic increase in the deterioration of the bridge decks and substructures of highway bridges in the United States [1217]. The percentage of bridges that were categorized as suffering moderate to extensive deterioration increased from about 15% in 1986 to around 35% in 1988 and to over 50% in 1990 [1220]. Corrosion is generally also a serious problem for steel-reinforced concrete structures in marine environments. The use of galvanized steel rebars as concrete reinforcement is one of the remedies that have been suggested to alleviate the corrosion problems [268,271]. Galvanized steel was used for concrete reinforcement as early as the 1930s and has been successfully used in many concrete structures. The performance of galvanized steel inside concrete has been investigated in a number of studies. These studies can be grouped into three categories according to the experimental conditions applied in the corrosion tests: (1) studies of rebars embedded in concrete and exposed to natural or field environments; (2) studies of rebars embedded in concrete and exposed to accelerated-test environments; and (3) studies of rebars exposed to simulated concrete solutions. In this chapter, only the information obtained from tests using concrete samples, i.e., falling into categories 1 and 2, is discussed. The information obtained from the solution type of tests, which do not have a real concrete environment and are only marginally relevant to the corrosion inside concrete, is presented in Chapters 5 and 9. 351
352
CHAPTER 13
13.2. CONCRETE ENVIRONMENT
13.2.1. Formation of Concrete Concrete is a two-phase material in which a mineral aggregate is dispersed in a matrix of hardened cement paste [1211]. It is made by mixing together cement, sand, gravel, and water, in proper proportions. Cement is a pulverized material which, when combined with water, forms a paste and, after setting, eventually becomes a hard solid. Hardened cement paste is the fundamental material as it provides the strength that allows concrete to be used structurally. The aggregate, i.e., sand and gravel, is also essential in that it provides rigidity and dimensional stability to the concrete. The raw materials of the commonly used portland cements are limestone and clay. These are fine-ground and sintered in special kilns at a temperature of 1400-1500°C. The product of the sintering process, called "clinker stone," is ground to dust together with a few percent of gypsum (CaS0 4 ·2H20) and other additives to become cement. The typical composition of portland cement is 23% Sial> 6.5% A1 20 3 , 3% Fe z0 3, 64% CaO, 0.6% MgO, and 2.1 % SOz [272]. Upon contact with water, hydration occurs with simultaneous dissipation of heat. Table 13.1 shows the reactions involved during hydration [272]. During hydration, the water is absorbed at a high rate by the cement particles, especially in the initial phase. In the later stages of hydration, the diffusion of water into the interior of the particles takes place at a slower rate and less uniformly. The cement is hardened during the hydration process in several stages, as schematically illustrated in Fig. 13.1 [1211]: (a)
Immediately after mixing, the cement paste is in its most fluid state. The cement grains are dispersed in the mixing water, the spacing being determined by the water/cement ratio. (b) After two hours, the cement paste is much less fluid but can still be worked. Hydration on the surface of the cement grains occurs.
TABLE 13.1. Hydration of Portland Type Cements" Hydrated clinker mineral
Clinker mineral
2Ca·SiO z(n)H 20 + Ca(OH)2 CaO·Si0 2(m)H 20 + Ca(OH)2 3CaO· Al 20 y 12H20 + CaO· Fe20] 3CaO·AI 2°3'6H20
3CaO·Si0 2 2CaO·SiOz 4CaO· Al z0 3·FeZ03 3CaO·AI 20 3
Product of hydration
Accessory substance Free CaO CaS04· H20
Ca(OHh 3CaO· A1 20 3·3CaS04' 31 H 20" 3CaO· Al 20 3·CaSO4.12H 2d Mg(OHh c
MgO "Ref. 272. "With the help of CaO·AI 20
3.
'The hydration of MgO present in crystalline form (periklase) is a very slow process; consequently, hydration accompanied by swelling in the hardened concrete is detrimental.
353
CORROSION IN CONCRETE
~ . . . Unhydrated ~ _ "mom p.n,,'"
131
FIGURE 13.1. Schematic representation of the stages in the hydration of cement particles and the formation of a cement gel. (Sec text for explanation.) After Newman [ 12721.
(c)
(d)
-I bl
_
Cement gel
Capillary pores • • and cavities lei
Id l
After a day, the cement paste is set but has no real strength. The hydration has penetrated further into the grains. The hydrates in the intergranular spaces have grown and interconnected, forming a continuous gel and thus establishing a solid skeletal structure. After seven days, the hardened cement paste has achieved considerable strength. The skeletal structure has been further developed by infilling between the original hydration links to produce a denser gel structure.
The reactions and the hardening process do not terminate within the first few weeks of solidification but continue for a considerable time. The time required to reach each stage can be altered by adjustments to the proportions of the compounds taking place in the chemical reactions.
13.2.2. Characteristics 13.2.2. J. Porosity and Permeability. The hardened cement paste contains small gel pores and large capillary cavities. Gel pores are formed as a result of the shrinkage during solidification, whereas capillary pores result from the evaporation of excess, unbound water [272]. The gel pores are about 2.5 nm in size, and they constitute about 20-30% of the volume of the hardened cement (also called cement stone). The sizes of the capillary pores are in the range of 1-10 ,urn, and the number of pores increases with increasing amount of mixing water but decreases with the progress of hydration. In addition, there are also air pores, whose size ranges from 0.01 to 2 mm and whose volume may range from 1 to 10% of the total volume of the concrete [272]. Concrete is thus permeable to water, air, and other agents owing to the presence of pores. The gel pores have an insignificant effect on the permeability of a cement stone owing to their small size and isolated nature. The permeability of concrete to water is predominantly determined by the capillary pores, the amount of which increases with increasing water/cement ratio. Figure 13.2 shows the relationship between the permeability and the water/cement (W /C) ratio [633]. When the concrete is submerged in water, the capillary pores fill first. Water penetrates into the air pores only after the displacement of air by diffusion from the capillary pores. The air (and thus oxygen) permeability depends on many factors, particularly the water content of the concrete [1221, 1274]. A dry concrete has a large fraction of pores unfilled with water, and these pores are available for air transport. On the other hand, in
354
CHAPTER 13
~
E
1.2
a.
~
..c ~
E
0.8
~0.6
'0 ~ 'u
~o
0.4
0.2
u
0.3
0.5
0.4
0.6
0.7
W/C ratio of cement paste (95% hydrated) FIGURE 13.2. Effect of water/cement (W/C) ratio on permeability of cement mortar specimens. After Power et al. [1273].
water-saturated concrete, air has to diffuse through the liquid phase, which is a much slower process. This can be noted in Fig. 13.3, which shows that the flux of oxygen through concrete increases with decreasing thickness of the concrete cover and relative humidity and with increasing water/cement ratio [1274]. The diffusion coefficient of chloride in concrete has been reported to be of the order of 10- 7-10- 8 cm 2/s [634-638]. There are a number of factors that affect the diffusion of chloride in concrete. The W/C ratio is particularly an important factor affecting the diffusion of chloride in concrete, as shown in Fig. 13.4 [634]. Figure 13.5 shows that the diffusion coefficient decreases with time [634].
ff"' b
.'" E
10,000
)(
()
1,000
0
'0
.S-
100
c
'" ~ 0
10
15
w/c=0.40 100% rh
x :>
u:::
-__
1
0
w/c'=0.60
100%rh
- - - -_ _ _ _ __
25
50
75
Concrete cover, mm
FIGURE 13.3. Effect of relative humidity (rh) and water/cement (W/C) ratio on the flux of oxygen through concrete. Data are taken from Ref. 1274.
355
CORROSION IN CONCRETE
3.----------------------------------,
. (J
+ 1 year
Cl)
E (J
.. 6 months
2
"1
o
E Cl)
'(3
i:
8'" 1 c a
'iii
:J
:t:
(5
oL-________________
- L_ _ _ __ _ _ _
o
0,2
0.4
~
________
0.6
~
0 .8
Water to cement ratio
FIGURE 13.4. Chloride diffusion coefficient as a function of water!cement ratio. From Lin r634]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
13.2.2.2. Moisture Content and Resistivity. As one of the major components, water is held in a number of different states in the hardened cement paste [12Il]: (a) (b) (c) (d)
Water vapor: The voids in hardened cement paste contain water vapor, exerting a vapor pressure corresponding to the relative humidity and temperature. Free water: The free water is mostly located in the capillaries and in larger gel pores. Adsorbed water: Several layers of water molecules are adsorbed on the surface of the gel pores. Interlayer water: Some water can penetrate the lattice layers of gel solids or into the intercrystalline spaces. 12 <..> Q)
V>
Eu ~
"E Q)
'0
i: Q) 0 <..>
c: 0
'iii
::l
:t::
(5
0
0
200
400
600
800
1,000
1,200
Exposure time, days
FIGURE 13.5. Time-dependent chloride diffusion coefficients for two different concretes: (A) Concrete with ordinary portland cement; (B) concrete with ordinary portland cement mixed with fly ash. From Lin 1634]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
356
CHAPTER 13
4r-----------------------------, W /C rat io : 0.50
2cm
3
E
r.
o ..>:
3cm
°OL-------~------~-------3~ 00----~
100
200
Duration of test, days
(e)
FIGURE 13.6. Changes of resistivity at different concrete depths during drying under constant climatic conditions (20°C, 50% relative humidity). Data are taken from Ref. 637.
Chemically combined water: This is the water combining with the unhydrated cement in the hydration reactions and forming an integral part of the solid.
The moisture content of portland cement concrete varies widely depending upon the environment surrounding it. Fully saturated concrete contains 13-15% water by weight, but the moisture content in air-dried concrete can be as low as 3-5%, depending upon the atmospheric humidity level [632]. The resistivity of the concrete is directly related to the amount of water present in the concrete. Figure 13.6 illustrates that the resistivity increases with time during drying; the increase is faster near the surface than in the interior owing to the faster rate of drying near the surface [637]. Resistivity increases with decreasing W/C ratio, and, for a given W/C ratio, it increases with decreasing cement/aggregate ratio [647]. 13.2.2.3. Cement Solution. The water trapped inside concrete pores contains the soluble species leaching out from the solids in the cement. Table 13.2 shows the typical
TABLE 13.2. Typical Composition of Cement Solution (One Part Cement, Th~ee Parts Watert Element Ca 2+ Na+ K+ Fe 3+ A1 3+ SiO~SO~CI-
Ml+
Concentration (mgll)
930 220 660
310 <10
aFrom Maahn and Sorensen [177J. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
357
CORROSION IN CONCRETE
-....
..
+
~ro
o
rn
0.1
O.01L---------------~-----------------L--~
12
14
13
pH
FIGURE 13.7. Concentration ofC}T in saturated Ca(OHh solutions as a function of pH. After Macias and Andrade [175].
composition of the cement solution (made with 1 part cement in 3 parts demineralized water) [177]. The pH of the aqueous solution formed within the concrete pores may vary from 12 to 14 depending on the nature of the cement and its degree of hydration. During the hydration, part of the CaO combines with water to form Ca(OH)2' The solution saturated with Ca(OH)2 has a pH of about 12.6 and a Ca 2+ concentration of about 0.9 gil [175, 177]. The concentration of Ca2+ ions decreases with increasing pH, and it changes also with the ionic strength of the solution. Figure 13.7 shows the concentration of Ca 2+ ions in Ca(OHh-saturated solutions as a function of pH [175]. The pH of the concrete solution can be affected by the addition of other salts. The addition of CaCl 2 decreases the pH of the pore solution whereas addition of NaCl hardly affects it. Also, the intake of chloride from the external environment during service has little effect on the pH of the pore solution [268,624]. On the other hand, it has been found that the presence of a high concentration of metal chlorides, such as FeCI 2 and ZnCI 2, lowers the pH of a Ca(OH)2 solution [1106]. When cement contains a large amount of Na+, K+, or SO~- ions, the pH of the pore solution can be as high as 13-14 [632, 1213}. J3.2.2.4. Chloride Content. Chloride salts affect the corrosion of steel in concrete in several ways. First, cr ions cause breakdown of the passive film on the steel. Secondly, chlorides are hygroscopic, attracting water and maintaining a moist environment inside the concrete. Also, CI- ions increase the concrete conductivity, which typically enhances corrosion activity. Chloride ions can diffuse through the concrete from deicing salts on highway bridges or from seawater in marine structures. They may also be mixed with the concrete as part of the water or aggregate. In some cases, chloride ions are added as calcium chloride to serve as a set accelerator. After hydration, some of the chloride ions are bonded in insoluble compounds such as 3CaO·AI 20,·CaCI 2·10H 20, and only a part of the chloride can be dissolved in the pore water [1039}. For example, the available NaCI is 0.01 % for 0.04% NaCI added to the cement and 4.92% for 5.87% added.
358
CHAPTER 13
13.2.2.5. Carbonation. When in contact with air, calcium hydroxide in concrete reacts with carbon dioxide and is transformed into calcium carbonate. This process is referred as carbonation [272]:
Carbonation reduces the pH of concrete; a pH as low as 6.5 has been measured in completely carbonated mortar samples [454]. The penetration rate of carbonation may vary greatly, depending on concrete mix and environmental conditions [276J. 13.2.2.6. Stability. Deterioration of concrete in a medium can be caused by [272]: (a) leaching of the free lime (b) dissolution of compounds by aggressive species in water (c) the expansion effect due to the formation of gypsum and calcium sulfoaluminate hydrate crystals The deterioration in soft water is essentially caused by dissolution and subsequent leaching offree Ca(OH)2. Leaching is rapid at first but becomes slower later. Because the solubility of calcium oxide is about 100 times that of calcium carbonate, the presence of a carbonate layer formed by carbonation has a barrier effect to leaching of the free lime when the concrete is exposed to soft water. 13.3. CORROSION OF STEEL REINFORCEMENT IN CONCRETE
13.3.1. Effect of Corrosion For a concrete exposed in nonpolluted environments, the steel reinforcement is free of corrosion. The alkaline concrete environment results in passivation of the steel surface. However, in many applications, the protective environment may be offset by the presence of aggressive agents that have entered the concrete from the external environment, and significant corrosion of the steel may occur [632-646]. Chloride salts are particularly aggressive in causing corrosion of steel in concrete. The chloride threshold level for corrosion of steel in concrete has been considered to be in the range of 0.65-0.77 kg/m3 [268, 1219].
Concrete
Corrosion products FIGURE 13.8. Schematic illustration of concrete spalling.
CORROSION IN CONCRETE
359
Severe corrosion of the steel reinforcement may result in weakening and failure of a concrete structure. Analysis of deteriorated concrete structures indicates that the corrosion of steel, which produces voluminous corrosion products, is responsible for the cracking and spalling that can lead to failure of the structure [269,271,642]. A schematic representation of the cracking and spalling phenomenon in reinforced concrete is presented in Fig. 13.8. J3.3.2. Protection Methods
Many methods have been developed for protecting steel reinforcement in concrete. They include concrete modifications [645, 1219], the use of inhibitors [275,644, 1039], cathodic protection [270, 277, 643], and the use of epoxy coatings [269,290], zinc wire [1225], and galvanized zinc coatings [271, 274] on the steel inside concrete. These methods can be generally divided into three groups: 1. Methods based on impeding entrance and transport of deleterious materials (water, oxygen, salts, carbon dioxide, etc.) in concrete. 2. Methods based on modifying the electrochemical processes through the use of inhibitors or by cathodic protection. 3. Methods based on modifying the surface of the steel by an organic or inorganic coating. Detailed information on the performance and features of many protection methods can be found in several review articles [645, 1219]. In the following discussion, only the information related to galvanized steel will be covered. J3.3.3. Galvanized Coatings
Galvanized coatings on reinforcing steel have been reported in a number of studies to be beneficial in delaying the corrosion of the steel and the consequent deterioration of the concrete structure [268, 271, 274, 290, 468]. Bonding between reinforcement and concrete is essential for reliable performance of concrete structures. The bond strengths for reinforcements in concrete are usually determined by either a pullout test or a bending test [274, 654]. Many factors, such as concrete mix and additives, curing conditions, and age, may affect the bonding between the galvanized steel and concrete. The bond strength of galvanized steel is largely similar to that of black steel [380,654]. Particularly, for deformed rebars the bond strength for black steel and galvanized steel is essentially the same because the strength is mainly provided by the mechanical interlocking between the ridges of the deformed bars and the concrete [290]. It has been noted that hydrogen evolution may occur during curing, resulting in a more porous interface between the galvanized steel and the concrete; this may affect the bond strength [963]. The addition of a small amount of chromate, 30-70 ppm, to the concrete has been found to suppress the evolution of hydrogen and thus increase the bond strength [379, 963]. It has also been reported that chromating of the galvanized rebar before embedding in concrete could prevent hydrogen evolution [177]. Annealing a galvanized steel has been found to be harmful to its performance [174, 397]. Annealing may produce a structure in which zinc-iron alloy layers extend to the
360
CHAPTER 13
coating surface, and the rate of hydrogen evolution may be increased as a result of the presence of the alloy layer on the surface. Also, owing to the lack of ductility, an annealed galvanized steel coating that has zinc-iron alloy extending to the outer surface is considered to be unsuitable for reinforcement in concrete [380, 397]. 13.4. CORROSION OF GALVANIZED STEEL IN CONCRETE
13.4.1. Testing Methods Corrosion inside concrete is a very complicated phenomenon involving a large number of factors. Many of these factors are uncontrollable and change irregularly with time. Because corrosion inside concrete is a rather slow process, typically taking years to show its effect, laboratory-simulation types of tests are often used to accelerate the corrosion process. In general, the tests conducted on the performance of galvanized steel in concrete fall into three broad categories: (1) tests conducted under natural exposure using concrete samples embedded with rebars; (2) tests carried out under simulated environments using concrete samples embedded with rebars; and (3) tests conducted under simulated concrete solutions using bare rebar samples. While the amount of data from simulated tests is enormous, the data from natural-exposure tests, which are the only reliable information for predicting the life of galvanized reinforced concrete structures, are limited. Particularly, there is a lack of data from heavy-deicing-salt environments [1219]. The available field data suggest that galvanized steel reinforcement may provide longer life compared to black steel [268, 173]; however, the data from studies using simulated test conditions [1218] are often in disagreement with these field data. It needs to be emphasized that laboratory tests designed to accelerate the corrosion process inside concrete, although providing useful information for the understanding of the various factors affecting the process, may not provide reliable information regarding the performance of the reinforced concrete structure in practice. The corrosion performance in the field involves the synergism of a large number of factors, many of which cannot be controlled, whereas only a limited number of predetermined factors are controlled and enhanced in laboratory tests. In particular, it is known that in atmospheric environments zinc generally corrodes 10 to 100 times slower than steel; it may, however, corrode at a similar rate as steel in a salt spray test. Similarly, it is known that in natural environments the under-paint corrosion rate is much slower for painted galvanized steel than for painted cold-rolled steel, but the opposite may be true in a salt spray test [767, 827]. The accelerating factor of each corrosion test can be very different for different materials.
13.4.2. Field Test Results 13.4.2.1. Corrosion Rate. When galvanized steel is covered with a good-quality concrete free of chloride, the corrosion rate is very low. As reported by Treadaway et al. [271] and shown in Fig. 13.9, the mean corrosion rate over an exposure period of 14 years in an industrial environment is typically in the neighborhood of 0.1 pmJyr for concrete with a WIC ratio of less than 0.6. The corrosion rate increases with increasing WIC ratio, particularly with a thin concrete cover (1.3 cm). The higher corrosion rate with the thin concrete cover was found to relate to carbonation.
CORROSION IN CONCRETE
361
0.8 - - - - - - - - - - - - - - - - - - - - ,
~>
1.3 em
-
E
0.6 -
i0.4~
- 2.5 em ""'3.8 em
" 5 em
.2
'"
~ 0.2~ o--------~--------~--------~--~
0.4
0.6
0.8
Water}:ement ratio FIGURE 13.9. Mean corrosion rate of a zinc coating in concrete over 14 years in an industrial e nvironment as a function of water/cement ratio and concrete cover thickness. Data are taken from Ref. 271.
Mercille [220J reported that a zinc coating suffered no long-term corrosion when embedded in concrete and exposed in an urban atmospheric environment. It was found that most of the corrosion experienced by embedded galvanized steel in concrete occurred during curing of the concrete. As much as 5 flm was corroded in the first day during setting of the concrete, but there was little change thereafter, less than 10 Jim being corroded during exposure for 10 years. High corrosion rates of a zinc coating may occur when the concrete is contaminated by chloride salts. Treadaway et al. [271] reported that, after five years of industrial atmospheric exposure, the corrosion rates for galvanized steel were low when the chloride concentration in the cement was below 1% (equivalent to 1.7 kg/m' for concrete with a 25 ~------------------------------~-,
.,~
20
>
E
:1..15
0;
_ . 20 mm; W/C 0.6
-'- 10 mm; W /C 0.6
r
20mm; W /C 0.75
.. 10mm;W/CO.75
~
g 10
' Vi
e
o
u
5 O~0~:~-~-~&~~L---------2 L---------3~--Amount of chloride added in cement. %
FIGURE 13.10. Effec t of chloride added in cement on the mean corrosion rate of zinc coating exposed in an indu strial atmosphere for three years for different combinations of cover thicknesses and water/cement (W/C) ratios. Data are taken from Ref. 271.
362
CHAPTER 13
60 _ 50
-. shorlblasted steel
t'O
+ pre ·rusted steel
Q)
>
E40
+!
,,
galvanized steel
:::l...
105
,,
.. chromated galv.
oj
~ 30 c
,
0
,
I
.~ 20
I
(3
U
-- .. o
3
2 Amount of salt added in cement, %
FIGURE 13.11. Comparison of the mean corrosion rates for clean and prerusted steel and for galvanized steel and chromated galvanized steel exposed in an industrial environment for three years. Data are taken from Ref. 271.
W/C value of 0.75 and 2.3 kg/m3 for a W/C value of 0.6) but increased significantly with increasing chloride concentration over 1%, as shown in Fig. 13.10. In the same study, Treadaway et ai. found that steel cleaned with shotblasting showed somewhat lower corrosion rates than galvanized steel, but prerusted steel corroded several times faster than galvanized steel at chloride concentrations higher than 2% (Fig. 13.11). In the same test, the effect of chromating on the corrosion of galvanized steel was found to be not significant. High salinity does not necessarily result in an excessive corrosion of galvanized steel inside concrete structures. Stark [273] did a field investigation on galvanized-steel-reinforced concrete structures in the tidal wave zone at four different marine locations for a period of time ranging from 7 to 23 years. He reported that the corrosion rates in concrete structures with high chloride content were generally below 1 l1m/yr, and in most cases below 0.5 l1m/yr, as can be seen in Table 13.3. TABLE 13.3. Average Corrosion Rates, R, of Zinc Coatings inside Various Concrete Structures Exposed in Marine Environments in Bermuda"
Structure Longbird Bridge St. George Dock
Hamilton Dock
Age (years)
Cover (cm)
Sample location
CI - content of concrete ~wt. % (kg/m )]
23 12 10 7 10
10
Above HT Above HT AboveHT Above HT NearMT Above HT BelowHT
0.19(4.4) 0.27 (6.4) 0.22 (4.6) 0.14 (3.0) 0.08 (1.9) 0.14 (3.6) 0.16 (3.7)
10 Bermuda Yacht Club "Ref. 273. "HT. High tide; MT, mean tide.
8
8 6 13 5 6 7
"
R (prnJyr)
0.2
1.1 0.5 0.29 0.5 0.8 0.0
CORROSION IN CONCRETE
363
TABLE 13.4. Potential Values of Galvanized Steel in Concrete Bridge Decks"
Location
Bridge
Age (years)
Cover (cm)
W/C ratioh
CI- content of concrete (kg/m J )
---"-~-"
Longbird' Boca Chica Seven Mile" Flatts Long Dick Spring Street Skokie" a
Refs" 271 and 1274.
II
Water/cement ratio.
Bermuda Key West, Fla. Key West, Fla. Bermuda Ames, Iowa Montpelier, Vt. Skokie, III.
21 3 3 8 7 3 6
6.0 3.8 5.0
0.44
6.4 7.6 1.3
0.40 0.44 0.43
1.0 1.12 0.84 0.53 0.29 0.06 0.9
-"
Potential (mVesE! -~~--
-370 to -450 - \00 to - 3OC) -280 to -3\0 -70 to -140 -100to-200 -200 to -300 -90 to -110
, With 5-em asphalt overlay" d
With 2.'i-cm asphalt overlay"
'Test slab"
Arnold [1260] investigated the effect of mixing galvanized rebars with black rebars in a bridge deck and found that connecting a galvanized rebar directly to a black rebar did not cause excessive corrosion of the zinc coating near the contact points after several years of field testing. Zinc-aluminum alloy corrodes significantly faster than zinc in concrete owing to the low stability of aluminum in an alkaline environment. Corrosion rates of about 10 prnJyr for a 4.5% Al zinc alloy and about 20 prnJyr for a 12% Al alloy were observed when the alloys were embedded in concrete exposed to a rural atmosphere for 10 years [22(n 13.4.2.2. Potential. The electrode potential is an important parameter indicating the surface condition of galvanized steel inside concrete. Although the measurement of the potential of rebar inside concrete is a simple practice, it is to be noted that several factors, particular! y I iq uid j unction potentials at concrete/reference electrode interface, may cause significant errors, as much as 100 m V [778]. Table 13 A lists the electrode potentials of galvanized steel reinforcements in a number of bridge decks installed for 3-21 years in different parts of the United States and in Bermuda r1279]. All the values are in the range of -OA5 to -0.07 VesE , which is considerably more positive than the potential, about -I VCSE' of an active zinc surface in aqueous solutions (see Chapter 5). These potential values suggest, based on the electrochemical behavior of zinc in aqueous solutions, that the zinc surfaces in these structures were probably in a passive state. Metallographic examination of the cross-sectional coating structure indicated that the corrosion in all these structures, except the Longbird bridge, was only superficial, and for the Longbird bridge it was estimated that 60-70% of the coating remained after 20 years of service. Dugan et al. [1226] also reported that the potential values determined for galvanized rebars in six bridge decks located in Iowa, Florida, and Pennsylvania were in the range of -0.385 to -0.07 VesE , and the corrosion of the zinc coatings was found to be rather superficial. Arnold [1260[ also found that most of the potential values of galvanized rebars embedded in concrete decks of several highway bridges in Detroit over a period of five years were in the range of -OA to -0.03 VCSE '
364
CHAPTER 13
13.4.2.3. Effect of Corrosion on Concrete. The corrosion of concrete reinforcement is a main cause of concrete cracking and spalling. Thus, cracking and spalling can be used as criteria for assessing the effect of corrosion on the concrete structures. A field survey by Stark [273] showed that the galvanized-steel-reinforced concrete structures in the tidal wave zone at four different marine locations showed no sign of concrete cracking over periods of time ranging from 7 to 23 years owing to the low corrosion rate of the galvanized steel (Table 13.3). The results of investigations on concrete structures or samples exposed in field environments generally indicate that, compared to black steel, galvanized steel delays the onset of concrete cracking. Baker et al. [1215] investigated the performance of concrete 80
A·615 Steel
tao .£;
'"c
~ 40 .<3
Cracks
g.
~
Spaliing
20
o
,
•
•
2
3
•
4
E
6
I
I
7
8
I
9
123456789-
80
4340 Steel
.£;
Spalling
Cracks
o
~,
2
3
4
I E
6
I 7
I B
No failure Very slight crack Slight crack Moderate crack Severe crack Very slight spalling Slight spalling Moderate spalling Severe spalling
9
80
Galvanized steel
~ao
.£;
Spaliing
Cracks
o
•
2
• 3
I 4
5
6
7
B
9
Failure type
FIGURE 13.12. Comparison of the degree of concrete damage due to corrosion for galvanized steel (48 samples), 4340 steel (28 samples), and A-615 steel (140 samples) after exposure in marine environments for II years. Data are taken from Ref. 1215.
CORROSION IN CONCRETE
365
slabs embedded with several steels exposed for II years in three different marine environments: splash zone, tidal zone, and SO-ft marine atmosphere lot. The concrete samples had a WIC ratio of either 0.4 or 0.5 and 1.3- or 3.S-cm cover thickness. The results presented in Fig. 13.12 show that the galvanized steel clearly had better performance than the black steel in terms of the extent of cracking and spalling. Treadaway et al. [271 J found that for concrete embedded with galvanized steel with 10- or 20-mm cover thickness and a WIC ratio of 0.6 or 0.75 and exposed in an industrial environment, the onset of corrosion-induced cracking of the concrete cover takes twice as long as in the case of black steel.
13.4.3. Results from Simulated Tests The information presented in this subsection is obtained from tests that use concrete samples but are conducted in simulated environments. These tests include immersion of concrete samples in water or in salt solutions. exposure of concrete samples to atmospheres having controlled relative humidities or controlled oxygen or carbon dioxide levels. cyclic immersion and drying tests, salt spray tests, water or salt ponding tests. and atmospheric exposure with periodic salt solution spraying. These tests. simulating the various conditions in field environments, accelerate the corrosion processes to different degrees. While it may take many years to obtain corrosion effects in field environments, it only takes a few months to a few years to complete the tests conducted under simulated environments. The corrosion rate of galvanized steel in concrete in laboratory testing is often measured with electrochemical techniques [745, 1214]. As discussed in Chapter 5, the corrosion rates measured as corrosion currents, due to the many factors and assumptions in the measurement technique and procedure, can have a wide range of deviations from the real corrosion rates. In a concrete electrolyte, this can be a particularly serious problem with polarization techniques due to the high resistivity of concrete [1096-1098] .
• 1 day
E
~ "'c""
10
,. 365 days
~
:; ()
'"
c 0
'in
2 0 0.1
0
'" 0.01
12
12.2
12.4
'" '"
,. '"
12.6
'"
'"
'" 12.8
pH of cement suspensions
FIGURE 13.13. Relation between the corrosion rates of galvanized rebars inside mortar samples on days I and 365 of a partial immersion test and the pH values of the cement suspensions (waterlcement ratio = I). Data are taken from Ref. 1214.
366
CHAPTER 13
TABLE 13.5. Effect of CaCl z Content in Cement, Carbonation, and Test Conditions on the Corrosion Current and Potential of Galvanized Steel in Mortar"'!> After 4 months Test
50% relative humidity
100% relative humidity
Water immersion
After 12 months
CaCl z
Carbonation
E (V SCE)
(pA/cm )
icorr 2
E (V SCE)
No 2% No 2% No 2% No 2% No 2% No 2%
No No Yes Yes No No Yes Yes No No Yes Yes
-0.17 -0.84 -0.01 -0.58 -0.39 -0.45 -0.41 -0.82 -0.42 -0.56 -1.05 -1.00
0.065 0.14 0.004 0.023 0.14 1.9 0.25 1.0 0.15 0.046 0.2 0.2
-0.21 -0.85 0.01 -0.68 -0.43 -0.40 -0.37 -0.53 -0.35 -0.42 -0.99 -1.04
lcorr 2
(pA/cm )
0.03 0.12 0.0013 0.023 0.04 1.0 0.12 0.5 0.06 0.15 0.17 0.23
"Data from Ref. 468. "Cover thickness, 7 mm; water/cement ratio, 0.5.
13.4.3.1. Corrosion Rate. Andrade and Macias [1214] studied the corrosion of galvanized rebars inside mortar samples, made of 11 types of cements, as a function of the pH value of the cement suspension. Figure 13.13 shows that the corrosion currents of all the samples decreased by more than one order of magnitude over one year of a partial immersion test. Gonzalez and Andrade [468] investigated the effect of carbonation, CaCl z content in cement, and test conditions on the corrosion current and potential of galvanized steel in mortar samples. Table 13.5 shows that in most cases the corrosion current decreased with time, The addition of CaCl z to cement significantly increased the corrosion current, particularly in the test with 100% relative humidity. When compared to the effect ofCaCl 2, the effect of carbonation on the corrosion current was found to be minor. However, data obtained by Simm [454], shown in Table 13.6, indicate that carbonation is as important as CaCl z in promoting the corrosion of zinc in concrete. The effect is the most significant when carbonation and CaCl z are both present at the same time. TABLE 13.6. Effect of Carbonation and CaCI 2 on the Corrosion of 0.15-mm-Thick Zinc Foil in Concrete Exposed in 100% Relative Humidity Air at 25°C for Up to 700 Days" Corrosion rate (pm/yr) Concrete condition" High carbonation + 1.5% CaCI 2 Low carbonation + 1.5% CaCI 2 Low carbonation, no CaCl z High carbonation, no CaCI 2 "Ref. 454. "Cover thickness. 17 mm; water/cement ratio, 1.2.
Unchromated
10 0.5 <0.1 0.63
Chromated
1.2 0.25 0.25 <0.1
367
CORROSION IN CONCRETE 1 . 000 ~------------------------------------~
+ '0
• Fe
Zn
100
)(
"'E u
E
10
.r: 0
r:r;G.
o . 1 L-----------~------------~----------~10
0.01
0.1
Concrete conductiv ity. (ohm" em " ) x 10. 3
FIGURE 13.14. Relation between polarization resistance, Rp, for mild steel and zinc rods in concrete and the conductivity of the concrete. After Feliu el al. [278].
Table 13.5 also shows the effect of test conditions. Of the three test conditions investigated, exposure at 100% relative humidity appears to be the most aggressive in terms of the corrosion current measured. In another study, Yeomans [388] found that the corrosion rate of galvanized steel was considerably higher in a test involving a cyclic salt solution immersion followed by oven drying than in a test consisting of continuous exposure in a salt spray chamber. Feliu et al. [278] found that the polarization resistance of zinc in concrete has a logarithmically linear relationship with the conductivity of the concrete as shown in Fig. 13.14. As postulated by Feliu et ai., the logarithmic proportionality is not due to an ohmic control of the corrosion process but may be due to the existence of a relation between the polarization resistance and the fraction of the wetted metal surface, Which, in turn, is related to the measured conductivity. Table 13.6 shows that chromating reduces the corrosion rate of zinc, especially in highly carbonated concrete with a high concentration of chloride [454]. In practice, chromating is used to reduce the corrosion during the curing period, when the zinc surface is active, but it does not seem to have a long-term effect on the corrosion of zinc in concrete [271,275]. The corrosion of galvanized steel is, in general, more uniform than that of steel in concrete [274]. Owing to the lower tendency for pitting, fewer and shallower pits were found to develop on galvanized steel rebars than on black steel rebars. Localized corrosion was found to occur in certain situations. Simm [454] observed pitting corrosion of zinc coatings in carbonated concrete containing chloride. Macias and Andrade [197] found that the corrosion on galvanized steel in Ca(OH)z solutions at pH < 11.5 was localized. Treadaway et ai. [271] noted that in cracked concrete the corrosion tended to be restricted to the vicinity of the cracks in the absence of added chloride in the concrete, whereas it tended to occur at areas remote from the cracks when chloride was added to the concrete. 13.4.3.2. Potential. The potential of galvanized steel in concrete varies greatly with the testing conditions. Table 13.7 lists the potential values reported in a number of studies.
CHAPTER 13
368
The values vary over a wide range, from -1.1 VSCE to -0.2 VSCE' It appears that the cyclic wet/dry type of tests tends to give more negative values than single-mode tests. It is indicated in Table 13.5 that, without the effect of carbonation and chloride salt, the potentials of galvanized steel in all tests are more positive than -0.45 VSCE' The presence of salt in concrete generally results in more negative potential values [l029]. Table 13.5 also shows that carbonation has little effect on the potential in the humiditycontrolled tests but results in more negative values in the water immersion test. The type of cement paste and additions, such as pozzolanic materials, have also been found to affect the potential of galvanized steel [745, 1249]. Figure 13.15 is a plot of corrosion potential versus corrosion current based on the data in Table 13.5. There does not appear to be a clear correlation between the corrosion current and potential except, perhaps, for the lower corrosion currents for the potential values more positive than -0.4 VseE' Considering that the potential values measured in most galvanized-steel-reinforced concrete structures in the field are also generally more positive than -0.4 VCSE (Table 13.4), and that at the same time the corrosion activity in these structures is also very low, it can perhaps be suggested that a large shift, in the positive direction, of the potential from that of an active zinc surface may be taken as an indication of a low corrosion rate of zinc in concrete. 13.4.3.3. Effect of Corrosion on Concrete. The results from different investigations using laboratory accelerating testing methods are rather conflicting with respect to the effect of corrosion of galvanized steel on the performance of concrete.
TABLE 13.7. Potential Values of Galvanized Steel inside Concrete Slabs in Accelerated Tests Test"
WID SID SID SS SP SP RH
PWI SID SP PSI PSI PWI FSI FWI FWI FSI lab air
Time 20 cycles 20 cycles 10-140 days 10-142 days Up to 5 years Up to 5 years 12 months 12 months 10-100 days 1-5 years 300 days 300 days 300 days I year 3 years 3 years 3 years 2 years
Cover (em) 1.0 1.0 1.2 1.2 2.5 2.5 0.8 0.8 1-3 1.2 2.5 2.5 2.5 2.5 2 2 2 3.1
W/C ratio b
0.5 0.5 0.8 0.8 0.4 0.5 0.5 0.5 0.8 0.41 0.63 0.51 0.56 0.5 0.68 0.68 0.55
E(V SCE )
-0.2 -1.0 -0.7 to -1.03 -0.7 to -1.02 -0.14 to -0.7 -0.2 to -0.6 -0.4 -0.35 -0.9 to -1.1 -0.44 to -0.76 -0.82 to -0.99 -0.68 to 0.98 -0.31 to -0.52 -1.02 -0.42 -0.5 -0.68 -0.15
Reference 177 177 388 388 1218 1218 468 468 1225 1260 1095 1095 1095 139 1029 1029 1029 745
"Abbreviations: WID. cyclic immersion in water and drying in oven; SID, cyclic immersion in salt solution and drying in oven; SS, in salt spray chamber; SP, ponding with salt solution; RH, controlled relative humidity; PWI, partial water im· mersion; PSI, partial salt immersion; FSI, full salt immersion; FWI, full water immersion. bWater/cement ratio.
369
CORROSION IN CONCRETE 10 r-----------------------------~
• 4 months X 12 months X
X
..; <: ~
:;
X .
0.1
"<:o
X
oe
X
X
'0;
0.01
U
0.001 w_l---_0..J..8--_0'--.6---_--' O.4----0-:-.-:2 ---0 ~Corrosion potential, V'CE
FIGURE 13.15. Plot corrosion potential vs. corrosion current of galvanized steel in concrete. based on data in Table 13.5.
In an investigation of the fatigue strength of reinforced concrete. Okamura and Hisamatsu [508] reported that concrete with galvanized rebars could endure a larger crack width than concrete with black rebars under a corrosive environment, as shown in Fig. 13.16. Swamy [519] studied the cracking of galvanized-steel-reinforced concrete in an accelerated test (6 h in 60°C seawater and 6 h of drying in air) and reported that the amount of corrosion and the depth of pits on galvanized rebars were considerably less than on plain rebars, as shown in Fig. 13.17. Also, there appeared to be fewer cracks in the concrete embedded with galvanized steel samples. In another study comparing the performance of galvanized steel and black steel reinforcements in concrete, Comet and Bresler [274] found that under corrosive conditions concrete embedded with black steel cracked earlier, and the cracks grew longer, as shown in Fig. 13.18. ~
240 r-------------------------------------------,
~
'V Black lower
a;
g> 220 ~
'" ~ '":>
t>. Black upper
.. Galvanized upper ... Galvani zed lower
200
...
Q)
Ol
~ 180 Q) u ij"
c: 160
~ 'E N
140 0.2
0.3
0.4
Maximum crack width at the side of specimen, mm
FIGURE 13.16. Effects of the maximum crack width in concrete beams on the fatigue strength of reinforcing bars (exposed to a solution of sodium chloride for a duration of 1 year). From Okamura and Hisamatsu [508]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
1.5
2.0
Plain bar
0 c=tc~_~
0.5
.... ~-~:;=;p
J.
2.73 mm
1.5
~ 0.5
0..
of.j
'0 '.0
's..
5
'-'
i
2.0
r
Galvanized bar
2.02 mm !
FIGURE 13.17. Distribution of corrosion pits and of concrete cracks after a cyclic immersion and dry test. Concrete cover was 2 em. Reprinted from Swamy [519]. with kind permission from Elsevier Science-NL. Sara Burgerhartstraat 25. lOSS KV Amsterdam. The Netherlands.
'" Cl
~ 1.0
"'"0
'5s..
I
2.5
.....
'"'"
::0
"" ;l
>-
::c
n
<:>
-.l
371
CORROSION IN CONCRETE
120 "
E (,)
100
Bla ck
a Galvanized
Q) (,)
P lain
.l!1
80
,5
Deformed
II)
""ctI (,)
60
t;
'0 L:
... - . . . . .. - ... - .. ..
c;, 40 c:
2! "iii
~
0
E E
• Black
L:'
1.6
E ~ E
.~
E Q) en ~
Q)
.'
....... - .., .,
..
I
I
0.8
_.. _.. - rr - "
it
-
.'
- - - Deformed
0.4
...
,
__ Plain
(,)
/
"
,
6 Galvanized
""ctI o 1.2
24
20
16
8
2
~
-
,,
20
-...-...'
I
~ I
"
,~
----....-~ Age at o bserv ation , mon ths
FIGURE 13,18, Effect of corrosion of black steel and galvanized steel reinforcements on concrete under stress in a cyclic salt immersion and dry test (sample dimension 10 x 10 x 30 em). After Cornet and Bresler [274],
~100
'"
'0
ci c:
~
'"t;
50
B OJ
E i=
10 L-----------~------------~--~~----~
o
0 ,2
0.4
0.6
Effective current densi ty, rn A /em'
FIGURE 13.19, Time to cracking as a function of effective anodic current density for concrete specimens embedded with zinc samples immersed in natural seawater, Data are taken from Ref. 1106,
372
CHAPTER 13
TABLE 13.8. Time to Cracking of Concrete Cover a for Concrete Slabs Embedded with Galvanized Steel and Black Steel Tested by Partial Immersion in Saturated NaCI Solutionb Steel
0.72 0.72 0.47 0.47
Black Galvanized Black Galvanized
Days to crackinl 149 286 >622 468
"Cover thickness, 2.5 cm. hRef. 1275. C Water/cement ratio. d Average of two samples.
Clear [ 1218] compared the corrosion performance of black steel and gal vanized steel inside concrete slabs with a I-in. cover and water/cement ratios of 0.4 and 0.5. The slabs were exposed outdoors with daily ponding to a 2-mm depth with 3% NaCI solution for 9 years. Based on the visual observation of cracking and spalling, Clear found that the slabs embedded with galvanized rebars performed slightly better than those embedded with black steel for a W /C ratio of 0.5 but slightly worse for a W /C ratio of 0.4. Similar results were reported for W/C ratios of 0.72 and 0.47, as shown in Table 13.8 [645,1095]. Grimes et al. [1106] investigated the effect of anodic current on the time to cracking of concrete slabs embedded with various metals, including iron and zinc. Figure 13.19 shows that the time to cracking decreases with increasing anodic current density. Grimes et at. found that pH gradients, from 2 to 12, in the concrete pore water may be generated by the anodic polarization among the local regions at the metal/concrete interface. They also found that the rate of zinc dissolution and the solubility of zinc in the pore water determine the rate of corrosion product formation and, therefore, the time of cracking. The corrosion products have greater volumes than the metals from which they are formed. This volume expansion is responsible for the cracking of concrete. The extent of volume expansion depends on the degree of hydration as well as on the diffusion and dissolution of the corrosion products inside the concrete. It is not clear how zinc corrosion products form and diffuse in concrete under various conditions. It has been noted that, compared to the corrosion products of iron, which are highly insoluble in concrete and tend to remain at the metal/concrete interface, the zinc corrosion products are more soluble and may, therefore, diffuse farther away from the metal/concrete interface [309, 645].
14 Corrosion in Batteries 14.1. INTRODUCTION Zinc is one of the most commonly used battery electrode materials because of its low equilibrium potential, reversibility, compatibility with aqueous electrolytes, low equivalent weight, high specific energy, high volumetric energy density, abundance, low cost, low toxicity, and ease of handling [685]. The use of zinc has mainly been in pnmary (nonrechargeable) batteries, but developments have been made that allow zinc to be more widely used in secondary (rechargeable) batteries. The discussion in this chapter is limited to the corrosion of a zinc electrode. The correlation between corrosion and battery performance is not within the scope of this discussion. In order to specify the corrosion environments, the first section is used to briefly describe the elemental features of each type of zinc battery. The many factors affecting the corrosion behavior of zinc electrodes, including the roles of the electrolyte, the zinc electrode, and operating conditions, are then discussed. 14.2. ZINC CELLS AND BATTERIES There are a number of batteries that use zinc as the negative electrode material. Zinc-manganese dioxide batteries, the most important zinc batteries, have been the mainstay of primary-battery manufacturing for more than 100 years. Currently, there are three commercial types of zinc-manganese dioxide batteries: Leclanche, zinc chloride, and alkaline cells. The pH-potential diagram in Fig. 14.1 illustrates the stable chemical species at each pH and potential for each type of battery. The cell potential for each type can also be predicted from the diagram. Other zinc cells have been developed over the years but are now of less commercial significance [221]. Besides the primary batteries, various types of zinc secondary batteries have been developed. Currently, zinc-air and zinc-nickel batteries are the two battery types under most active research and development. 14.2.1. LeclancheCell
The Leclanche cell, which is a zinc-manganese dioxide system, was first developed by G. Leclanche in 1860 and has since been improved by many others [1149]. It is typically manufactured in a paper-lined and paste-wall cylindrical configuration as shown 373
CHAPTER 14
374
2 .2 1.8
MnO~
1.4 1.0
(()
~
~
0 .6
Mn++
'"
0 .2
w -0,2
tl::l G c:
N
+
Zn
-0.6 - 1.0 Zn
- 1.4 -1.8 0
2
4
6
8
1O
14
12
16
pH FIGURE 14.1. pH-potential diagram for the zinc-manganese dioxide system. assuming [Zn 2+J [Mn2+] = IO-4M . After Brodd [1143].
= 1M and
in Fig. 14.2 [1143]. The zinc anode is a sheet metal, and the cathode is manganese dioxide powder mixed with carbon powder. Carbon powder is used as the conducting media because Mn0 2 is a poor conductor. The electrolyte which moistens this mixture is an aqueous solution of ammonium chloride and zinc chloride. A typical electrolyte composition is, for example, 1M ZnCl 2 + 2M NH4 Cl [1143]. Although the battery can be used
Zinc can Label rube ;·, ~ ·.~... c.r.;;--Paper
separator
Positive active material Bottom washer Bottom cap FIGURE 14.2. Cross-sectional view of paper-lined construction of the Leclanche cell. After Brodd [1143].
375
CORROSION IN BATTERIES
in a wide pH range (4-9), the pH of the electrolyte is commonly set at slightly acidic values, e.g., pH 5.2. Other additives may be added to increase efficiency, reduce corrosion, and improve mechanical properties. The reaction at the zinc anode during discharge is (14.1 ) This reaction is immediately followed by combination of the zinc ions with the chloride ions of the electrolyte to yield a mixture of species, principally zinc tetrachloride: (14.2) At the manganese dioxide cathode, at least two reactions are observed. The initial reaction is (14.3) The ammonia generated from this reaction usually reacts with the zinc chloride present to form zinc ammonium chloride: (14.4 ) The overall reaction during discharge is
Eo = 1.63 V
(14.5)
14.2.2. Zinc Chloride Cell The zinc chloride cell is designed to provide higher drain capabilities than the Leclanche cell [1143]. A more acidic environment in this cell results in a higher open-circuit voltage, as noted in Fig. 14.l. The construction is similar to that of the paper-lined ammonium Leclanche cell. A typical electrolyte is, for example, 27% ZnCl. The zinc chloride electrolyte often contains a small amount of NH 4 CI to help control pH and prevent zinc passivation. The overall cell reaction is
8MnOOH + ZnCI 2-4ZnO·5H 20
Eo = 1.75 V 04.6)
14.2.3. Zinc Alkaline Cell The alkaline cell possesses superior performance at higher current drains and has longer shelf life and greater freedom from leakage than either the Leclanche or the zinc chloride cell [1140]. While the alkaline cell employs the same active reactants (zinc and Mn0 2) as the Leclanche and the zinc chloride cell, it differs in cell chemistry and construction features. As shown in Fig. 14.3 the cell is contained in a steel can which serves as current collector for the cathode [1139]. The zinc anode is a high-sUlface-area zinc powder suspended in a gelling agent such as carboxymethyl cellulose. A brass sheet or nail serves
376
CHAPTER 14
Can-Stee l
Positive Cover---+-""7"*+---/// Plated Steel ~~S'b=~~~
Meta lized Plastic Film labet
ElectrOlytePotass ium Hvdroxide ~~~r-+-
CathodeManganese Dioxide . ---1~H*f Carbon
~---'~~I--+-
AnodePowdered Zinc
Current Collec t or B r ass
Sea l -Nylon
SeparatorNon - Woven Fabric--+---iI~"f-
Inner Ce ll Cove r Steel
~Negative Cover -
Me t al Washer- - - - - - - - - - '
Rivet-Brass
Pla t ed Steel
FIGURE 14.3. Cross-sectional view of a typical 'D' -size alkaline-manganese battery. After Schumm r1141J.
as the anode collector. The electrolyte is a concentrated KOH solution (25-35%, about 5-1OM). The discharge reaction proceeds more slowly in an alkaline than in an acidic electrolyte. The high surface area of the zinc powder makes up for the more sluggish discharge reaction. The overall reaction in the alkaline cell is Zn + 2Mn0 2 + H20
~
ZnO + 2MnOOH
Eo = l.55 V
(14.7)
14.2.4. Zinc-Air Battery The zinc-air battery has a very high energy density because of its ability to use oxygen from the air as the positive electrode reactant [1143]. It consists of a caustic electrolyte, a zinc anode, and an air electrode on which oxygen is reduced. The electrolyte is usually around 6N NaOH, and the overall reaction is 2Zn + O2 ~ 2ZnO
Eo = l.64 V
(14.8)
When the electrode is cast zinc, the battery is used for low-discharge applications since the electrode can only sustain a maximum current of 30-40 mA/cm2 because passivation occurs at higher current densities. Zinc powder is used as the anode material for higher discharge rates.
377
CORROSION IN BATTERIES
14.2.5. Zinc-Nickel Battery The zinc-nickel battery has been under development as a rechargeable system for some time [221]. The overall reaction for this system may be written as 2NiOOH + Zn + 2H 20
~
2Ni(OHh + Zn(OH)2
Eo = 1.75 V
(14.9)
The electrolyte used is typically a solution of potassium hydroxide at a concentration of 6-7M. The main problems limiting the successful commercialization of this battery system are (a) changes in the shape of the zinc electrode as a result of cycling; (b) disparity in charging efficiency between positive and negative electrodes; (c) dendrite growth during the charging process; and (d) zinc poisoning of the nickel electrode [221, 1207]. 14.3. CORROSION The corrosion of the zinc electrode in zinc cells and batteries is the main cause for self-discharge, relatively short shelf life, and perforation of the zinc can in the case of Leclanche cells, when the corrosion is localized. In sealed batteries, corrosion is also responsible for pressure buildup by hydrogen resulting from the corrosion process. The corrosion of zinc in a battery environment is extremely complicated because it involves a large number of factors. These factors can be classified into three main groups: 1. Properties of the electrolyte (a) Composition: Type of chemicaL concentration. pH. zinc concentration. oxygen concentration, gelling agents. and additives (b) Physical setting: Volume, tlUidity, and texture 2. Properties of the zinc electrode (a) Form: Solid sheet (e.g., methods of production and surface treatments), powder (e.g., shape and size of particles), porous structure (pore dimension and porosity), or amalgamated material (b) Composition: Impurities and alloying elements 3. Operating conditions (a) Temperature (b) Time (c) Current collector (d) Composition of cathode material (e) Rate and depth of discharge (f) Sealing method (g) Cell geometry The corrosion of zinc in batteries has been the subject of a number of investigations. but the understanding of the effect of many of these factors has not progressed beyond an empirical level. Although some agreement is found among the results of different investigations, contlicting results are also found. The discrepancies among the results from different investigations are usually due to the fact that it is not possible to control all the factors affecting the corrosion of zinc in batteries. Moreover, the test conditions in these investigations are practically always different.
CHAPTER 14
378
The main cathodic reaction accompanying the corrosion of zinc in a battery environment is hydrogen evolution. Thus, in Lec1anche cells the overall corrosion reaction is (14.10)
and in alkaline cells it is Zn + 2HP + 20~ ~ Zn(OH)~- + H2
(14.11)
The corrosion rate of zinc battery materials is most commonly measured by a gassing test in which the volume of hydrogen evolved during the corrosion process is collected. Figure 14.4 illustrates a typical setup for the gassing test [1146]. Measurements based on
10
o
FIGURE 14.4. Cell for the study of hydrogen evolution. After Riietschi [1146].
CORROSION IN BATTERIES
379
weight loss are less often used for determination of the corrosion rate. Electrochemical techniques have also been used to determine the polarization characteristics of the corroding electrode under various conditions [9,232,534,790]. The conversion between the corrosion rates from a gassing test and from electrochemical measurement is based on the equivalence of a gassing rate of l,ul!(cm 2·h) to a corrosion current of 2.5 ,uA/cm2 131 I]. It is important to note that the gassing rates reported in different studies should be compared with caution because the test conditions among the various studies differ in temperature, form of the zinc electrode, amount of zinc material and electrolyte, electrolyte composition, and equipment design [1153]. The corrosion rates and basic electrochemical characteristics of zinc electrodes in the corrosion process are presented in the following sections, which are organized according to the major factors listed above. General information on the elemental electrochemical reactions of a zinc electrode is discussed in Chapter 2. 14.3.1. Effect of Testing Time The corrosion rate of zinc powder in a sealed cell or in a typical gassing test generally decreases with time before reaching a steady value. Snyder and Lander [1144] reported that the time required to reach a steady gassing rate varies, by several days, for sealed "dry" cells under different conditions, as shown in Fig. 14.5. According to Gregory et al. 1234], who obtained similar results in a typical gassing experiment, the initial decrease in gassing rate in a gassing experiment may be attributed to the fact that an average time of about one day is required to achieve saturation of the electrolyte with hydrogen and that formation of a surface coating of zinc oxide or hydroxide occurs, which may inhibit the anodic dissolution.
14.3.2. Effect of Electrolyte 14.3.2.1. Concentration. The dependence of corrosion rate on KOH concentration varies with the test conditions. As shown in Fig. 14.6, three types of relationships between 0.4,-----------------------------------------~
C 0.3
E
E ~
5% KOH, 1 % HgO
0.2
OJ
c 'iii
30% KOH, 1% HgO
(f)
'"
<.:J 0.1
_-=:::::::::::~=~~;;;;;5~%~K;O;H,
2% HgO
oL~============~==========~~~40~%~o:KO~H~,~2~%~H~g~O~
o
10
20
30
Time, days
FIGURE 14.5. Gassing rate in sealed "dry" cells vs. time. After Snyder and Lander [11441.
380
CHAPTER 14
8 Reference -(1144)
Qi
+(234)
.~ 6 V
a
.(1146)
E
:> (5 :>
-4
e
a::'"
C>
c
~ 2
'"
CJ
0
0
2
4
6
8
10
12
14
KOH Concentration (M)
FIGURE 14.6. Effect of KOH concentration on gassing rate.
corrosion rate and KOH concentration have been reported in the literature: (a) Corrosion rate increases with increasing KOH concentration [1146]; (b) corrosion rate decreases with increasing KOH concentration [231, 1144]; and (c) corrosion rate decreases with increasing KOH concentration at low concentrations, reaches a minimum, and then increases with concentration at high KOH concentrations [234]. The different dependencies of gassing rate on KOH concentration among various studies indicate that different rate-determining reactions [Eqs. (14.12)-(14.16)] may be involved in each case. Type (a) behavior was attributed by Rtietschi [1146] to the ability of the electrolyte to dissolve interfacial Zn(OH)z formed as a result of discharging. Thus, the reaction in Eq. (14.13) is the rate-determining step in the corrosion process. Type (a) behavior may also be related to the increase in exchange current density of the zinc electrode with increasing KOH concentration [1155]. The decrease in corrosion rate with increasing concentration, type (b) behavior, was explained by Snyder and Lander [1144], as well as by Dirkse and Timmer [231], as a result of reduced water activity, which means that the reduction of water, Eq. (14.12), is the rate-determining reaction. As KOH concentration increases, the water content is decreased. Type (c) behavior indicates a change in the rate-determining reaction on going from low to high KOH concentrations. According to Gregory et al. [234], in such situations water reduction may be the rate-determining reaction at low KOH concentrations, whereas a step in the anodic dissolution that requires the participation of hydroxyl ions, i.e., dissolution of the surface oxide film, may be rate-determining at higher KOH concentrations.
Zn + 20W
~
Zn(OH)z + 2e-
(hydrogen evolution)
(14.12)
(zinc oxidation)
(14.13)
CORROSION IN BATTERIES
381
Zn(OHh + 20W ~ Zn(OH)~-
/
Zn(OH)~blllk
(formation of zincate)
(14.14)
(diffusion to bulk)
(14.15)
(formation of oxide film)
(14.16)
The differences in the rate-determining reaction for the corrosion of zinc electrodes among the different studies must arise from the differences in experimental conditions, such as electrolyte and electrode compositions, amount of electrode material versus electrolyte, form of the electrode, etc. Curve a in Fig. 14.6 was obtained from tests that lasted less than 2 days whereas curves band c were obtained from tests that lasted for more than 20 days [232, 1144, 1146]. Also, curve c was obtained from a "dry" cell test, in which the ratio of electrode material to the volume of electrolyte was small (about 100 gllOO ml) as compared to those (about 1 gllOO ml) for curves a and b. It seems that in the "dry" cell the amount of water available for reactions is more likely to be limiting and that the corrosion is limited by water reduction at all KOH concentrations. The effect of the volume of electrolyte reflects the importance of the ratio of zinc powder to the amount of electrolyte. Figure 14.7 shows that the gassing rate increases with a decrease in the ratio of zinc powder to the amount of electrolyte [1152]. Differences in the relationship between gassing rate and KOH concentration may also be caused by the differences in physical form of the zinc electrodes used in the various studies; some have a porous structure whereas others have a solid structure. A porous electrode is likely to differ from a solid sheet electrode not only in the reaction rate but also in the limiting rates for diffusion, activation, film formation, and dissolution.
30
~ 5
9
·1 0 9 C!>
25
+25 9
'5.
E ::J
~
C!>
15
c
:a
'Vi
10
(!)
5 2
4
6
8
14
Time, days
FIGURE 14.7. Influence of quantity of zinc powder on the gassing rate of zinc powder (3% Hg, 0.02% In) in 100 mI electrolyte (480 g KOH, 60 g ZnO, 460 ml H 20) at 45°C. Reprinted from Meesus el al. 111521. with permission from International Power Sources.
382
CHAPTER 14
14.3.2.2. Zincate Ions. The presence of zincate ions in alkaline electrolytes is generally found to reduce the corrosion rate of the zinc electrode [231, 232, 234, 1146]. The presence of zincate has the effect of reducing the rate of diffusion of corrosion products away from the surface. Hence, for the same period of time, the surface coverage by ZnO in a zincate-saturated electrolyte must be much larger than in electrolytes without zincate, so that the corrosion must be correspondingly smaller. On the other hand, it has been found that when the system is not controlled by an anodic reaction, addition of zincate to the electrolyte can actually cause an increase in the corrosion rate. As reported by Snyder and Lander [1144], the corrosion of zinc powder in a "dry" cell, where the gassing rate increases with increasing amount of ZnO added to the electrolyte, is limited by the activity of water in the electrolyte. According to Snyder and Lander, addition of ZnO to the electrolyte increases the effective water activity. ZnO reacts with KOH in a 1:2 ratio to form a zincate ion, resulting in a lower effective concentration of KOH. It was estimated that a 45% KOH solution saturated by ZnO has an effective concentration of 36%, and a 30% KOH solution has an effective concentration of 26% based on this equivalence. A similar explanation can be given for the increase in gassing rate with the addition of ZnO, as reported by Sato et al. [1147], for a zinc powder in 40% KOH gel. Because of the high alkaline concentration and the gelling effect, the corrosion process may be dependent on water activity. Amalgamation generally reduces the effect of ZnO on the corrosion of the zinc electrode in alkaline electrolytes [1144]. 14.3.3. Effect of Chemical Agents 14.3.3.1. Inorganic Species. Era et al. [115] investigated the effect of several ionic species on the gassing rate of a solid zinc sheet of about 100 cm 2 in 500 ml of Leclanche electrolyte (25% NH 40H + 12% ZnCl). Figure 14.8 shows that the presence ofCu 2+, Ni 2+, As 3+, or Sb3+ at concentrations as low as 1 or 2 mg/l drastically increases the corrosion rate of zinc. As 3+ and Sb3+ ions seem to be the most aggressive among these species in promoting gassing. The presence of Fe 2+ also increases gassing, but Fe 2+ is much less aggressive than the other species. Figure 14.9 shows that the gassing rate generally decreases with addition of Pb 2+ to the electrolyte with or without the presence of the
Impurities
250
-Blank
+2mIJ/lCu" -6'2 mg/l F.~·
:::-200
.s., .,
+ 2 mgJ1 Ni ,.
*2mg/lAs'·
0>150
+ \ mIJ/lSb"
'0 Ql
E
" 100
~
o
2
4
6 Time (days)
8
10
FIGURE 14.8. Effect of ionic impurities in Lec1anche electrolyte on the gassing of a solid zinc sheet of about 100 cm 2 in 500 ml of Leclanche electrolyte (25% NH4 0H + 12% ZnCl). Data are taken from Ref. 115.
383
CORROSION IN BATTERIES Impurities
1,000
-Blank
+2mg/INi
+2 mg/I Cu2 +-fr2 mg/I Fe2 + 2+
3+
3+
*1 mg/IAs +1 mg/ISb
=E (/)
ro
OJ
o Q)
E .2 o FIGURE 14.9. Effect ofPh~+ concentration of Leclanche electrolyte, in the presence and absence of other ionic impurities, on gassing of a solid zinc electrode. Data are taken from Ref. 115.
>
0.1L--------L----__- i_ _ _ _ _ _ _ _L -_ _ _ _ _ _ o 5 20 10 15
-L~
Pb concentration (mg/I)
aggressive ions. Mansfeld and Gilman [10] found that the presence of Sn 2+ has only a slight influence on the gassing rate of a zinc rod in 6N KOH whereas Cu 2+ increases and Pb 2+ decreases the gassing rate significantly. Addition of 1O-3M KCl or KBr has only a marginal effect on the gassing of the zinc rod. The effects of these ions were found to arise mainly from their alteration of the cathodic polarization resistance of the zinc electrode. Yoshizawa and Miura [12051 investigated the effect of various metal oxides and hydroxides on the corrosion of zinc powder in alkaline battery electrolytes. They found that addition of compounds of indium, yttrium, and zirconium resulted in significant inhibition of the corrosion. Sato et al. [1147] reported that addition of a certain amount of In 2 0 3 in zinc powder gel had a beneficial effect on reducing the gassing rate in an alkaline solution (0.2 g of gel in 2 ml of KOH), as shown in Fig. 14.10. The gassinga:; Ol Ol N
150
c::i
OJ
1§ c 0
"-5
....
50
0
>
OJ
(f)
ctS
c.9
0
0
2 In,03 Concentration, wt%
FIGURE 14.10. Gas evolution rate from Zn alloy mixed with an equimolar amount ofZnO containing various concentrations of In203 at 60 e. Reprinted from Sato et al. [11471, with kind permission from Elsevier Science-NL, Sara Burgerhartstraat 25,1055 KV Amsterdam, The Netherlands. 0
384
CHAPTER 14
reducing effect resulting from the presence of In 20 3 in the electrode is attributed to the inhibiting effect of metallic In, formed through the reaction: (14.16) The presence of metallic In was confirmed by X-ray diffraction. The reason for the increase in gassing rate at higher In 20 3 content is not clear [1205]. The effect of electrolytes on the corrosion of a zinc anode has been investigated in a number of studies. Baugh [111] found that the cathodic polarization curves of zinc are similar in Br-, Ct, CIO;; and SO;- ammonium salt solutions with pH range 3.8-5.8. There are, however, significant differences in the corresponding sodium salt solutions. This was attributed to the dissociation of NH! ions, which produces hydrogen ions at the electrode surface. Gregory et al. [234] noted that in NaOH electrolyte the rate of hydrogen evolution is lower than that in KOH. Catino [311] measured the polarization behaviors of zinc electrodes in LiOH, KOH, and NaOH and found that while the anodic polarization for zinc dissolution decreases slightly in the order LiOH > NaOH > KOB, the cathodic polarization for hydrogen reduction significantly increases in the order LiOH < KOH < NaOH. According to Catino, the cathodic reaction at small overpotentials is mainly water reduction; at large overpotentials, it involves alkali discharge. The mechanism of hydrogen evolution with alkali cation participation is discussed in Chapter 2. 14.3.3.2. Organic Additives. Organic additives are usually used in zinc batteries for corrosion inhibition and to prevent passivation of the electrodes. They have acquired great importance in recent years in the production of mercury-free zinc batteries. Also, addition of organic inhibitors together with inorganic inhibitors has been found to produce maximum corrosion inhibition [1205]. Narte.y et al. [1203] investigated 18 organic corrosion inhibitors for suitability in rechargeable alkaline zinc batteries. The inhibitors were evaluated with a hydrogen gas evolution test and polarization curve measurements and in a real battery situation. They found that, except for two, all the organic species tested had various degrees of corrosioninhibiting effect in the out-of-cell hydrogen gas evolution test. Among the 18 additives tested, a-diphenylglyoxime and polyethylene (600) were found to perform the best in the real cell situation. Nartey et al. also found that in the case of the inhibitors performing well in real cell situations, more positive corrosion potentials and lower anodic current densities were obtained than in the standard condition, while the corrosion current densities were similar to those in the standard condition. This means that at the corrosion potential the rate of anodic dissolution is inhibited by to the presence of the inhibitors while the rate of cathodic reactions is enhanced. Brossa et al. [427] studied a number of organic agents as corrosion inhibitors in a Leclanche cell electrolyte. Figure 14.11 shows that all but one have an effect of reducing the corrosion of zinc in the electrolyte. In another study, the addition of perfluoroalkyl polyethylene oxide as a surface activator was found to prevent zinc corrosion without a loss of discharge performance [1205]. Organic inhibitors are considered to inhibit metal corrosion by adsorption on the metal surface. The adsorption can be physical, electrostatic, or chemical [1203]. However, there is still a lack of understanding on the detailed mechanisms of the corrosion inhibition of battery electrodes by various organic inhibitors.
385
CORROSION IN BATTERIES
6 . -------------------------------------, • L-meth ionine
600 Time. minutes FIGURE 14.11. Hydrogen evolution vs. time in pure and inhibitor-containing NH 4 CI solutions at 40°C. NDA, I-Decylaminc; TTH, 1,2,6-HexametrioI-trithiolgIycolate; PoeIe, Polyoxyethylene (23)-Laurylether; Triton, Octylphenoxy-polyethoxyethanol; Forafac, Ethoxylated-polyfluoro-alcoho1. After Brossa et af. [4271.
14.3.4. Zinc Electrode The composition and surface condition of the zinc electrode materials are two very important factors for corrosion. Inhibitors are commonly added to electrode materials to inhibit gassing. Inhibitors can be uniformly mixed within the total mass or applied only on the surface. Gas evolution is a surface phenomenon and must be reduced at the surface of the zinc electrode, whether it consists of powder particles or a solid sheet. Inhibitors distributed uniformly in the bulk material can provide continuous inhibition during discharge, while those deposited on the surface may disappear after the battery is put into operation and surface material is dissolved. However, surface-only inhibition requires much less inhibitor than mass alloying and is therefore more economical. In many applications, the most important gassing inhibition is provided by inhibitors applied to the original electrode surface before the battery is put into operation. For effective surface-only inhibition during battery operation, the inhibitors must be sufficiently mobile to remain on the receding surface or, if they dissolve in the electrolyte, they must have the tendency to coat fresh zinc surface. 14.3.4.1. Amalgamation. Amalgamation is the most effective way of reducing the corrosion rate of zinc in zinc cells and batteries. As an example, Fig. 14.12 shows that the gassing rate for a sealed "dry" cell is significantly reduced with increasing amount of HgO in the electrode material [1144]. However, the use of Hg is now very limited owing to environmental concerns. Through alloying with other elements, the amount of mercury currently used in batteries for corrosion inhibition has been reduced to very low levels. Low gas evolution is achieved for surface-amalgamated special zinc alloy powder containing as little as 0.05% Hg [1204]. With alloying and the use of organic inhibitors, batteries containing no Hg have recently been produced in Western countries 11147].
CHAPTER 14
386 0.12
c
'E
E
0.08
ai
~
OJ
C
'0;
'"
'"
0.04
(9
0
~
1%
:045%
4%
2%
HgO Concentration,
wt%
FIGURE 14.12. Variation of the steady-state gassing rate for a sealed "dry" cell with HgO content in the anode material at 38°C. After Snyder and Lander [1144].
Traditionally, two techniques have been employed for amalgamation of zinc battery material [1152]: Surface amalgamation by mixing zinc powder and mercury metal or salts Mass amalgamation by atomizing a homogeneous Zn-Hg alloy In surface amalgamation, Hg is initially deposited at the outer surface of zinc particles. The mercury then diffuses through the grain boundaries into the grains. A combination of surface amalgamation and low mass amalgamation is also used to obtain optimum corrosion inhibition. Electrochemical measurements indicate that amalgamation affords corrosion inhibition mainly through reducing the cathodic reaction. Baugh et al. [1150] investigated the electrochemical characteristics of amalgamated zinc electrodes. Various amounts of mercury (10-1000 /1g/cm2) were plated onto the zinc surface. At low mercury levels, the mercury deposits were localized. A complete coverage of the surface was obtained at the I-mg/cm2 level. Diffusion into the zinc occurred in grains, particularly at grain boundaries. Figure 14.13 illustrates the anodic and cathodic polarization curves in 30% KOH electrolyte for the zinc electrodes having different levels of amalgamation. It can be seen that while the dissolution of zinc is slightly facilitated by amalgamation, the hydrogen evolution current is greatly diminished. In another study, Baugh et al. [534] found that amalgamation has little effect on the anodic behavior of the zinc electrode in NaCI solution but has a significant effect in NH4CI solutions. The hydrogen-inhibiting effect of amalgamation is less pronounced in NaCI solution than in KOH and NH 4CI solutions; this was attributed to the diffusion-limited hydrogen evolution in the NaCI solution. According to Dirkse et at. [1154], amalgamation also lowers the overvoitage for zinc dissolution in alkaline electrolytes. The overvoltage for the dissolution reaction at amalgamated electrodes is for the charge transfer while at non amalgamated electrodes it is for adatom diffusion.
CORROSION IN BATTERIES
387
~ .;;;
c:
FIGURE 14.13. Polari zation characteristic s for Zn in 6.8 molldm 3 (30% w/w) KOH as a function of amalgama1 tion level: 0, Pure zinc; e, 5 Ilg/cm-; ,-I, 10 pg/cm 2; . , 20 Ilg/cm", <>, 75 jl g/cm 2, x, 100 jlg/cm2, +, 250 jl g/cm 2; 6., 500 Ilglc m2; .... 1000 pg/cm 2 Reprinted from Baugh e! al. 11150], with permissi on from International Power Sources .
.g
100
E
~
:::l
U
10
Potential, V (Hg/ HgO)
J4.3.4.2. Alloying Elements. Variation of zinc electrode composition is perh aps the most effective way to control gassing in zinc batteries. In an extensive investigation, Miura et al. [1151] found that Cd, Pb, In, Bi, and TI inhibit the corrosion of the zinc electrode in 40% KOH while Ag, Cu, Ga, Sn, and Sb promote the corrosion. Al and Ni inhibit corrosion when they are present in very small quantities but promote corrosion when the concentration is increased beyond certain values, as shown in Fig. 14.14. Figure 14.15 shows that when AI, In, Pb, and Hg are present together, AI and In inhibit corrosion until their concentrations reach certain values, beyond which they promote corrosion . The optimum composition obtained in Miura's study, in terms of the effectiveness of corrosion inhibition, was 0.02% In, 0.05 % Pb, 0.05 % AI, and 1 %Hg. This alloy is equivalent in its effect to that containing 9% Hg. Glaeser [1153] also reported that, with multielement alloying, zinc powder containing less than 0.1 % Hg is made equivalent to 3% Hg powder in terms of gassillg rate.
45 Waighl "" . 0.05 (l0.1
35
I 0)
c
0 1
25
., '" CJ
'iii
15
FIGURE 14.14. Amount of gassing for various zinc all oys amalgamated with 1% Hg (tested with 10 of Zn powder in 1.9 ml of 40 WI. % KOHl. Data are taken from Ref. I 151.
5
Ag
AI
rn~ Bi
Cd Cu Ga
Il1J In
Ni
Pb Sb Sn
Alloying elements
J
TI Base Zn (1 % Hg)
CHAPTER 14
388
5
%4
.
"0
~
~
In + X
0,02 0,02
Pb + Al + 1% Hg
0,02
X 0.02
0,05 0,05
X
3
0
In
Ol C
'"'" '" t:)
1 Pb 0 0,001
0.Q1
0.1
Content of elements, X, (wt%) FIGURE 14.15. Effect of alloying elements on gassing. Data are taken from Ref. 1151.
The roles of various alloying elements in the corrosion of amalgamated zinc have been summarized by Miura et al. [1151] as follows: • In, TI, Pb, and Cd increase H2 overpotential, maintain Hg at surface, and inhibit Hg diffusion. • AI, Bi, Ca, Mg, and Sr smooth surface and inhibit Hg diffusion. • Ni and Ag maintain Hg at surface and inhibit Hg diffusion. Lee [9, 1121] measured the hydrogen overpotential for zinc contammg small amounts of alloying elements and found that the presence of 0.05% Cd or Ca increases the overpotential while the presence of 0.05% Fe or Mn reduces the overpotential, as shown in Table 2.9. Bhatt and Udhayan [1206] found that a Zn-0.27Pb-0.09Cd alloy electrode has a slightly higher corrosion rate than a pure zinc electrode in 5.5N ammonium chloride but less internal resistance. One effect of Cd and Pb as inhibitive alloying elements is to reduce the grain size; it is generally agreed that a fine-grain structure is less sensitive to corrosion than a coarse-grain structure [1155]. Also, Cd and Pb were reported to inhibit corrosion perforation of zinc cans. Figure 14.16 shows that pitting tendency is reduced with the addition of Cd and Pb to the zinc electrodes for Leclanche cells [1276] . In a recent study on mercury-free batteries, Yoshizawa and Miura [1205] reported that zinc alloys including Pb, In, and Bi [i.e., Zn(Pb + a), a = In + Bi] as well as Zn(ln + Bi + Ca) alloys as mercury-free electrode materials show the highest corrosion resistance. Zinc alloys containing 0.00 I-I % In, 0.005-1 % Mn, 0.005-1 % Pb, 0.005-1 % AI, and 0.0005-0.1 % rare-earth metals have also been developed as mercury-free electrode materials [1208]. 14.3.4.3. Surface Treatment. Catino [311] found that dipping a zinc wire in 0.05M metal-ion solutions produces various degrees of precipitates on the zinc surface. Immersion in HgCI2 solution produces a dark surface that is removed when immersed in 9N
389
CORROSION IN BATTERIES 100,-----------------------------------------,
75 ~
Pb
cO
0
~ (;
50
0.15 Cd
+ Pb
~
Cd
0-
25
OL-______L -______L -______L -____ o 0.25 0.5 0.75
~~
____
~
1.25
Metal addition, %
FIGURE 14.16. Influence of Pb and Cd additions on pitting of zinc battery cans. After Aufenast [12761·
KOH. Immersion in Pb(C 2H 30 2)2 yields a deposit with a dense, fine, needlelike structure that does not change after immersion in 9N KOH. Immersion in InCl) does not change surface morphology, whereas immersion in TICI yields a uniform distribution of TI-rich particles. With metal-ion treatment, both the anodic and cathodic reactions show a higher polarization than in the case of nontreated zinc. Table 14.1 shows the corrosion rates of zinc samples treated with these metal-ion solutions. 14.3.4.4. Physical Forms. Zinc powder with a larger particle size is found to have smaller gassing rates than that having a smaller particle size [1148, I 152]. The difference between the gassing rates of zinc powders having different particle sizes decreases with amalgamation [1148]. Cachet et al. [843] reported that the corrosion of zinc powder in a Leclanche cell electrolyte causes an increase in the reactive surface area owing to surface roughening but does not significantly change the shape and size of the zinc particles. According to Gregory et al. [234] and Catino [311], the formation of zinc hydroxide and zinc oxide on different areas of the electrode surface may also occur, leading to nonuniform corrosion of the surface because zinc hydroxide can be more easily dissolved than zinc oxide.
TABLE 14.1. Initial and Final Corrosion Rates for Metal-Ian-Treated Zn Wires in 9N KOH Based on R" Measurements and Gas Volumes Produced on Metal-Ian-Treated Zn Powder after 24 Hours at 71°C in 45% KOH" ~------~--
Metal-ion treatment None
HgCI2 Ph(C 2H 30 2 )2 PhCl 2
TICI InCI,] "Ref. 311.
i", initial (j1A/cm 2 ) 13.8 5.1 3.4 10.7 16.6 12.1
ie' final (pAlcm 2 ) 21.7 4.0 1.5 4.3 12.8 22.1
Gas volume (ml) 0.18 0.05 0.07 0.06 0.22 0.6
390
CHAPTER 14
14.3.5. Operating Conditions 14.3.5.1. Temperature. The rate of gassing from corrosion in a zinc cell increases with increasing temperature. Snyder and Lander [1144] found that the steady-state gassing rate for sealed "dry" alkaline cells increased by a factor of about 4-7 as the temperature was increased from 25 to 50°C. Gregory et at. [234] observed a lO-fold increase in the gassing rate in gassing tests when the temperature was increased from 25 to 60°C. The enthalpy of activation for the gassing reaction has been reported to be about 19 kcal. A lO-fold increase in the gassing rate was also observed by Riietschi [1146] for zinc powder immersed in electrolytes containing zincate with a temperature increase from 60 to 95°C. The effect of temperature on gassing rate depends sensitively on the composition of the electrode material [1153]. 14.3.5.2. Discharging. In real situations, a battery may operate under a certain discharging rate for a length of time, or it may stand idle after a certain depth of discharge. The corrosion that occurred on the electrode during or after discharging may be changed by the discharging process, because discharging can cause changes in electrolyte composition and texture, changes in the surface area and composition, formation of surface films, and development of non uniformity across the surface. The effects of most of these changes on the corrosion are little known. Sato et at. [1147] investigated the effect of discharging on the corrosion of a zinc powder gel. Figure 14.17 shows that the gassing rate significantly increases with depth of discharge. It also increases with increasing discharge current, as illustrated in Fig. 14.18. The increase of gassing rate with depth of discharge and current density was explained by Sato et al. as a result of the buildup of zincate ions in the gel, since an increase in the gassing rate with addition of zinc oxide to the gel was observed in a separate experiment. According to Yoshizawa and Miura [1205], in an alkaline manganese battery, more hydrogen is produced in the case of partial discharge than in the case of nondischarge
~ 12 Cl C\I
ci
>-
"'"
8
:i.. cD
~
c:
0
S (5
4
> Q>
'"'"
0
0
20
40
60
Depth of discharge, %
FIGURE 14.17. Gas evolution rate from Zn alloy with 2% Hg at 60°C at various discharge depths after discharge at 30 rnA for 0.2 g of gel. Reprinted from Sato et at. [1147], with kind permission from Elsevier Science-NL, Sara Burgerhartstraat 25, 1055 KV Amsterdam, The Netherlands.
391
CORROSION IN BATTERIES
a:;
OJ
~400
o >-
.g
300
:i.. Q)
"§ 200 c
o ~
~ 100 Q) (/)
ro c.9
O~------~---------L--------~------~~
o
10
20
30
40
Discharge current, mA
FIGURE 14.IS. Gas evolution rate from 50% discharged Zn alloy at various cUlTents for 0.2 g of gel at 60°C. After Sato et al. 111471.
owing to (1) roughening of the zinc surface and (2) decreasing hydrogen overvoltage at contacting faces with ZnO particles. 14.3.5.3. Current Collector. The function of a grid is to conduct electric current and to support the active material. Grid materials with electrode potentials more positive than that of zinc may form a local galvanic cell to cause extra corrosion of the zinc electrode. It has been established that silver and copper grids promote gassing of the zinc electrode l234, 1144]. The grid can be treated to reduce gassing. In one study, it was found that amalgamation of a silver grid reduced the gassing rate in a "dry" alkaline cell. In another study, amalgamation of silver grid material was found to have little or no effect on the gassing rate whereas amalgamation of copper grid material significantly reduced the gassing rate of a porous zinc electrode immersed in alkaline electrolytes [2341. The difference between the silver and copper grids was attributed to the difference in the concentration of mercury on the surface of the two types of grids.
15 Corrosion in Other Environments 15.1. INTRODUCTION This chapter deals with the corrosion situations that are not covered in other chapters. More specifically, it includes information on corrosion in organic solvents and gases and on zinc anodes. 15.2. ORGANIC SOLVENTS
15.2.1. Classification The reactivity of metals in organic solvents depends on the type and structure of the organic compound. Organic solvents can be classified as nonpolar aprotic, dipolar aprotic, and protic, as summarized in Table 15.1 [424]. The classification of a solvent as protic or aprotic is based on whether it has an ability to provide protons. Protic media contain acidic hydrogen atoms, and aprotic media do not. Dipolar aprotic solvents display electrostatic forces due to ion-dipole and dipole-dipole interactions. An important role of a solvent in a corrosion process is solvation, a process whereby solvent molecules form shells around each dissolved ion to compensate the Coulombic forces between the oppositely charged ions in the solution. Changes in the composition of a solvent alter its solvation properties. The solubility of a metal significantly varies from solvent to solvent owing to differences in solvation properties. Figure 15.1, as an example, shows that the solubility of zinc chloride in different organic solvents varies over several orders of magnitude [500]. Usually, an oxidizing group within the molecular structure of a solvent is responsible for the corrosion process [424]. For example, for carboxylic acids such as acetic acid, CH)COOH, the corrosive group is COOH- and/or H;o'v, and for alcohols such as ethanol, C 2HsOH, the corrosive group is H2COH- and/or H;olv. In addition, as in aqueous solutions, the presence of chemical species such as oxygen and halogens may significantly increase the corrosivity of the organic solvents. The corrosion reactions in organic solvents can be grouped into two types: electrochemical and chemical [424]. In the electrochemical type of reaction, the anodic partial reaction for zinc corrosion can be expressed as 393
394
CHAPTER IS
TABLE 15.1. Organic Solvent Systems with Protic and Aprotic Propertiesa Kind of bonding
Species dissolved
Solvent Hydrocarbons Cyclopentadiene Propylene carbonate
o~ HX, halogens Fe +, Co2+, AgCI04 HCI
Acetone Carboxylic acids
FeCI 3 Hydrohalogens
Alcohols
Inorganic and organic acids
Water
Inorganic and organic acids
van der Waals n-Complexes Ion-dipole forces, dipoledipole forces Covalent complexes Dipole-dipole forces, hydrogen bridges Dipole-dipole forces, iondipole forces, hydrogen bridges Dipole-dipole forces, iondipole forces, hydrogen bridges
Type of solvent Nonpolar aprotic Nonpolar aprotic Dipolar aprotic Dipolar aprotic Protic Protic
Protic
"Ref. 424.
Zn + 2X-
~
ZnX 2 + 2e-
(15.1)
where X- is a halogen ion, organic acid anion, etc. The cathodic partial reaction can be represented as
HA + e- ~ -} H2 + A-
(15.2)
where A- = carboxylic acid anion, hydroxyl ion. alcoholate ion, etc. gil Water Methanol Ethanol } Acetone Diethyl ether Ethyl acetate I-Butanol I-Propanol
- - - --
1000 ____ _ -
-
-
-
-
100
lA-Dioxane
10
FIGURE 15.1. Solubility of zinc chloride in solvents containing 0.05MHCl at 20°e. From Hronsky [500]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
395
CORROSION IN OTHER ENVIRONMENTS
The chemical type of reaction can be represented in a general form, for example, for solvents with halogen ligands: (15.3) 15.2.2. Corrosion Hronsky [500] investigated the corrosion of zinc in various organic solvents. Figure 15.2 shows that the corrosion rate varies greatly with the type of solvent. It also shows that the corrosion rate of zinc in some solvents is much higher than in water, while in others it is much lower. Figure 15.3 illustrates that the viscosity of the solvent is a predominant factor determining the corrosion rate of zinc in organic solvents containing a small amount of acid. Hronsky found that the corrosion rate decreases slightly with increasing molecular weight of the sol vent and with decreasing dielectric constant. Also, it may be noted by comparing Figs. 15.1 and 15.2 that the corrosion rate does not appear to be a direct function of the solubility of zinc in the solvents. Furthermore, according to Hronsky [500], neither the dissociation level of the dissolved acid nor the electrolytic conductivity has a major effect on the corrosion rate. He thus concluded that the rate-determining step in the corrosion of zinc in these solvents is probably the transport of the oxidizing agent, which is a function of the viscosity. For a given solvent, the corrosivity is affected significantly by the presence of other chemical agents. Lechner-Knoblauch and Heitz [842] investigated the corrosion of zinc
____ _
-
(4)l'I01olcn P'OO!JC.I Borrltf Film rormtd
,1
90
Dltth,.. cl/"e'
80
10
60
50 1,4
40
-Oio~o t'lt
Wattt
20
FIGURE 15.2. Zinc corrosion in various organic media containing 0.05 mol HCI per liter. From Hronsky [5001. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
I - P,oponol
t -8ul0nl)l _ _ _ _ _ <1
l/iA}'lidtM c.h\l)I't~t
Time (hours)
396
CHAPTER 15
50,--------------------------------------, Acetone.
"Diethyl ether
40 Ethyl acetate
Methanol 1 ,4-Dioxane •
• Benzene
• Ethanol- Water • • 1-Propanol
Vinylidene chloride
°0~__~t-~B=uta=n=ol____L __ _ _ _ _ _ _ _ _ _~2------------~3~
Reciprocal kinematic viscosity (s/mm2j
FIGURE 15.3. Corrosion rate of zinc vs. reciprocal kinematic viscosity of corrosive media containing 0.05 mol HCI per liter at 20°e. From Hronsky [500]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
in alcohols containing various gaseous and solid contaminants. Table 15.2 shows that the addition of other solvents such as formic acid, sodium formate, acetic acid, and sodium acetate results in increases in the corrosion rates. Jaenicke and Schweitzer [1275] determined the apparent exchange current density of the Zn 2+/Zn amalgam electrode in mixtures of organic solvents and water. Figure 15.4 shows the dependence of io on the molar fraction of tetrahydrofuran, acetone, acetonitrile, ethanol, and dioxane in mixtures with water. It can be seen that io decreases with the addition of water to the solvents up to a certain proportion and then rapidly increases with further addition when the water content reaches more than 95%. According to Jaenicke and Schweitzer, in the water-solvent mixture the metal surface is preferentially solvated by organic molecules, while the solvation shells of the dissolved metal ions consist mainly of water molecules, which are the most polar complexing agents present in the medium. At a certain solvent/water ratio, this difference in the solvation of the metal surface and the metal ions in the solution reaches a maximum. Consequently, the energy barrier for the metal atom-ion transition through the double layer at the electrode is maximal, which is reflected by a minimum in io, as shown in Fig. 15.4. In general, according to Miles and Gerischer [1189], the variation in electrode kinetics with different mixtures of solvent and water results from a combination of (a) specific adsorption of the sol vent, (b) changes in the potential across the double layer, and (c) changes in the charge-transfer activated complex as the solvation sheath changes. The results from a number of studies indicate that the corrosion processes in solvents containing small amounts of contaminants are usually diffusion-controlled [424, 488, 500, 749]. This is supported by the fact that viscosity, which determines the diffusion coefficient, is a predominant factor in determining the corrosion rate in a number of solvents. Similarly, for zinc in primary alcohols containing O.01N HCl, the corrosion rate decreases with increasing chain length of the alcohol, as illustrated in Fig. 15.5 [749]. Also, the polarization curves of a zinc electrode in methanol and octanol, shown in Fig. 15.6, indicate that the rate-limiting process in the corrosion is the diffusion of oxidizing agents to the surface. Electrode processes other than diffusion can affect the corrosion rate in some solvents. Biallozor and Bandura [463] postulated that the charge-discharge of Zn 2+ in dimethyl sulfoxide, N,N-dimethylformamide, and acetonitrile proceeds by a two-step
-
N2
1.29
0.02
0.18
O2
.. -
H 2O
~,-------
0.01
..
0.01 0.14
0.09
3.9
0.46
4.2 0.87 0.62
0.38
N2
Air
S.S
0.3
7.0 3.93 0.28 0.16 0.13
0.67
0' 0.01
0,
CH 30H
9.7
11.7
0
Air
0.17 0.24 0.20 0.29 0.07
0.01
0.17 1.06 0.43 0.21 0.03
0.03
0 0
N2
0.08
0
2.8 2.8 0.62 0.29 0.02
0.03
0 0
O2
C 2H s OH
0.13
3.51
0
Air
0.14
0.48
0.43
0.73
0
0.12
0.56
6.28
1.60
0
CH 30H
SO°C
0.14
0.01
0.1
0.37
0
'0 corresponds to weight loss rates of <0.01 g/(m'·h).
"Reprinted from Lechner-Knoblauch and Heitz [842J. with kind permission from Elsevier Science Ltd. The Boulevard. Langford Lane. Kidlington OXS 1GB. United Kingdom.
dCorrosion rates in grams per square meter per hour nearly correspond to millimeters per year.
O.IM O.OSM O.OIM O.OOSM O.OOIM
HCOOH
O.IM
HCOONa
O.IM 0.05M O.OIM O.OOSM O.OOIM
CH 3COOH
O.IM
None 50 ppm CICH 3COONa
Contaminant mol I-I
Room temperature
0.18
0.02
6.19
0.01
0
C 2H sOH
TABLE 15.2. Weight Loss Rates (in Grams per Square Meter per Hour)" of Zn in Alcohols Containing Different Gases and Contaminants at Room Temperature and 50°C"
~
....
C/l
-l
Z
tIl
3:
:;:; o z
~ <
;>:l
tIl
:t
~
Z
oZ
oC/l
:xl :xl
o
n
398
CHAPTER 15 100,---------------________________--, (4 ) H2 0 +CHaCN (5) )CH2)2" (1 ) H20 + CH -CH OH Hz<:> +0" 3 Z (2) H20+CHa-TI-CHa (CH2)2
?
°
(3 )
10
.~
0.1
0.01
L - - L____L __ _~L__ _~_ _ _ _- L____L _ _ I
0.4
0.6
0.8
1.0
Organic solvent
Water
FIGURE 15.4. Exchange current of zinc (2mM ZnCI04 ) in binary water mixtures as a function of mole fraction of organic solvent. (Concentration of supporting electrolyte, NaCI04 : in tetrahydrofuran and acetone, 0.7 M; in acetonitrile and ethanol, 0.6M; in dioxane, 0.2M). After Jaenicke and Schweitzer [1275]. 2.5,---------------------------------------,
2
~ 1.5
g
0.5
C1
C2
C3
C4
C5
C6
C7
C8
C-Atoms FIGURE 15.5. Corrosion rate ofzine in primary alcohols of different ehain lengths with addition of 0.0 IN HCI
+ IN LiCI at 25°C (with hydrogen bubbling). After Heitz et at. [749].
CORROSION IN OTHER ENVIRONMENTS
399
100,-------- - - -- - - -- -- - - -- - - - - -- - - - - - - ---,
Eu
10 Methanot
~
E
:E.u; cQ) u
E
0.1
Octanol
~
:;
U
0.01
0.001
-1.15
-0.95
-1.05
E.
-0 .85
-0.75
Vs ... e
FIGURE 15.6. Polarization curves of zinc in methanol and octanol solutions containing O.OIN HCI and IN LiCI at 25°C. After Heitz et al. [749].
reaction. Singh and co-workers [386,453] found that the corrosion in dimethylformamide is determined by an activation controlled cathodic reaction. The addition of water to the solvent reduces the cathodic polarization and thus increases the corrosion rate. Hronsky [500] noted that in some solvents a poorly adhering, permeable, black layer, identified as consisting of amorphous zinc compounds and zinc oxide, is formed during corrosion. The formation of the surface film appears to have an effect of hindering the further corrosion of zinc in the solvents. Babaqi et al. [1013] determined the corrosion current for zinc in methanol solutions with various water contents. The corrosion current was found to increase linearly with increasing water content at different temperatures. The activation energy values determined experimentally indicate that the rate-determining step in the corrosion is more related to the adsorption of a participating species and less to a diffusion process, which is also supported by the fact that solution stirring has no effect on the corrosion current. Their results seem to suggest that when the solvent is pure and the corrosion rate is low, the corrosion is less likely to be controlled by diffusion and more likely to be controlled by other processes. It should be noted that although some general rules may be applied to estimate the corrosiveness of an organic solvent solution, reliable values for the corrosion rate in a specific solvent solution still need to be experimentally determined, since the corrosion rate of zinc in organic solvents varies greatly with solvent composition. The corrosion rates of zinc and its alloys in many specific organic solutions can be found in the book Corrosion Resistance of Zinc and Zinc Alloys by Porter [1190]. 15.3. GASEOUS ENVIRONMENTS The gaseous environments discussed in this section are not related to the normal atmospheric environments that were considered in Chapter 8. A gaseous environment is defined here as an environment confined in a container or package in which gases other than those in normal air are also present. These gases may be present originally, or they
400
CHAPTER 15
Specimens inHCI
5
E <.l
vapor
4
0)
E
c:
.iij
3
0)
:g,
2
~
in wat er
vapor 20
30
40
50
60
Exposure, days
FIGURE 15.7. Gain in weight of galvanized sheets exposed to the atmosphere in a sealed tank containing 20% hydrochloric acid. After Gilbert and Hadden [437].
may be released from liquids or solids, particularly plastic materials, also present in the confined space. Corrosion in gaseous environments is governed by principles similar to those elucidated for atmospheric corrosion although it has its special features [480]. As in normal atmospheric corrosion, the amount of moisture in the environment plays an important role in gaseous corrosion. Depending on the kind of gases and their concentrations, the critical relative humidity required for corrosion may vary. Also, depending on whether an electrolyte is or is not formed, the corrosion can be electrochemical or chemical in nature. The corrosion of zinc in HCI vapors has been studied by several investigators. Gilbert and Hadden [437] found that the corrosion rate of galvanized sheets hanging in a sealed tank above a 20% HCI solution was many times higher than that of sheets placed above water, as shown in Fig. 15.7. Askey et al. [811] observed that the corrosion rate of zinc in HCl vapor increases approximately linearly with HCl concentration. According to them, at normal atmospheric levels HCl reacts with preexisting Zn(OH}z to form insoluble
TABLE 15.3. Corrosion Rate of Zinc in a Moving Atmosphere Containing S02 and Water Vapor (Average S02 Content 1.8%)" Time (days)
Corrosion rate (J1.mlyr)
7 50 100 200 500
627 238 145 130 127
"From McLeod and Rogers [499]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
CORROSION IN OTHER ENVIRONMENTS
401
NO.8
E -2. OJ E_ 0.6 Q) rJ)
ro ~ ()
.~
..,
0.4
..c
OJ Q)
~ 0.2
Mg
Pb
Cr
AI
Fe Avialite Zn Brass Ag
Ni
Cu
18-8 FIGURE 15.8. Corrosion stability of metals in dry hydrogen sulfide, as indicated by change in weight in 180 days. Data are taken from Ref. 556.
basic zinc chloride but at higher levels transformation of Zn(OH)2 to more soluble ZnCl 2 can proceed. Table 15.3 shows that the corrosion rate of zinc in a moving mixture of S02, water vapor, and air is rather high and that it decreases with exposure time [499]. In hydrogen sulfide gas, the corrosion rate of zinc is low compared to those of other common metals and alloys, as shown in Fig. 15.8 [556]. Variation of the relative humidity of the H 2S-air mix has little effect on the corrosion rate of zinc, although it greatly changes the corrosion rates of some other metals, as shown in Fig. 8.24 in Chapter 8. It is noted that although Fe corrodes little in dry H2S, it corrodes severely in H 2S-water vapor mixtures. This indicates that zinc can be a corrosion-resistant coating material for protection of steel in H 2S environments [556]. Knotkova-Cermakova and Vlckova r480] investigated the corrosion of several metals placed in sealed compartments containing various plastics, rubber, and various types of wood. Table 15.4 shows the corrosive effect of these materials on zinc and iron, which is due to their tendency to produce acetic acid and, in lesser amounts, formic acid by hydrolysis. In general, direct contact causes more corrosion than mere exposure to the vapor released by these materials. Table 15.5 shows the corrosion rates of zinc in a sealed chamber containing different types of wood at 35°C. Oak appears to be distinctively more corrosive to zinc compared to other woods. Table 15.6 contains the corrosion rates reported by Helwig and Bird [69] for galvanized steel over sealed aqueous solutions of several organic acids. The corrosion rate is greatly increased due to the presence of these acids. Table 15.6 also shows that chromate surface treatment reduces the amount of corrosion in these environments. It also shows that chromating the zinc surface can significantly reduce the amount of corrosion under the test condition.
402
CHAPTER 15
TABLE 15.4. Corrosive Effects of Organic High-Molecular-Weight Materials on Steel and Zinc in Sealed Enclosures or by Contact" Degree of corrosive effect on metal" In sealed enclosure Material Phenolic resins Amino plastics Polyformaldehyde Polycarbonates Poly(vinyl chloride} Polyamide alkali Polyamide hydro Glass-reinforced polyesters Epoxides Pol yethy 1ene Poly(methyl methacrylate) Pol ystyrene Poly( vinyl acetate) Cellulose acetate c Rubber Paints oil Epoxide'· Melamine alkyd'" Wood Oak Beech Chestnut Birch Fir Walnut Elm
Steel 2 1
0 0 0-1 0 0 1 0 0 0 1 1 0 0 1
3 3 2 1-2 0-1 0 0
By contact
Zn
Steel
3 2 I 0 0-1 0 0 2 2 0 0 0
Zn
3 3
3 2 I 1 I 0 0 I I 0 0 0 I 1-2
1-2 0 0 2
I I I 0 2 2 0 0 0 I 1-2 I 2
3 3
3 3 2 2 1-2 0 0
3 3 2 2 1-2 0 0
2 1-2 0-1 0 0
"Data from Ref. 480. "Corrosive effect: O. None; I. slightly corrosive; 2. corrosive; 3. highly corrosive 'The corrosive effect is most intense during the curing period.
TABLE 15.5. Corrosion Losses of Iron and Zinc Enclosed with Wood at 35°C for 32 Days" Material (veneer) Oak Beech Ash Maple Spruce Poplar
"Ref. 480.
Relative humidity
Corrosion loss of metal (urn)
(%)
Iron
100 84 100 84 100 84 100 84 100 84 100 84
10.57 1.37 3.02 0.53 3.16 0.53 1.04 0.48 0.58 0.03 1.12 0.07
Zinc 15.42 10.47 3.00 4.33 3.00 4.33 1.53 4.22 0.40 0.46 5.20 6.37
403
CORROSION IN OTHER ENVIRONMENTS TABLE 15.6. Zinc Loss for Specimens Stored for 30 Days at Ambient Temperature over Various Solutions".!' Zinc loss (mg/dm 2 ) for specimens stored over: Cr in surface film (mg/dm 2 ) 0 0.2 1.3 2.2
Water
1.8% Ammonia
2.5% Formic acid
0 0 2 4
185 11 4 4
690 204 131 69
2.1 % Acetic acid 5.1 % Acetic acid 795 400 161 90
303 137 35
"Reprinted from HelwIg and Bird [69J, with kind permission from Elsevier Science Inc., 655 Avenue of the Americas, New York. hOriginai coating weight was 1.12 oz/fr~ of sheet.
15.4. ZINC ANODES
15.4.1. Sacrificial Anodes Zinc is a commonly used material for making sacrificial anodes owing to its low position in the electromotive series. Zinc anodes were first used to provide galvanic protection to the copper-sheathed hulls of warships a century and half ago [1193]. Historically, zinc sacrificial anodes have been used most extensively in seawater applications. They are also widely used for the cathodic protection of hot water tanks [447], fuel storage tanks [281], underground steel structures [471,473], and steel-reinforced concrete structures [329,1232,1284,1285,1288]. Compared to other metals with low electrode potentials, such as aluminum and magnesium, zinc has distinct advantages as an anodic material in many common environments. It does not usually passivate, in contrast to aluminum, and it has a low self-corrosion rate, and thus high efficiency, in contrast to magnesium. The performance of sacrificial zinc anodes in providing galvanic protection depends on a number of factors, many of which are discussed in detail in Chapter 7 . In the following discussion, only the information related to material properties is presented. Anode design considerations, in terms of shape, size, and position, regarding specific applications can be found elsewhere [469,471, 1192, 1193J. 15.4.1.1. Definitions. Anode potential, current output, and anode efficiency are the three most important properties of sacrificial anode materials, The anode potential determines whether there is a sufficient driving force for the cathodic protection in a given environment. Zinc sacrificial anodes are, in most cases, used for protection of steel structures: for complete protection, the potential of steel needs to be cathodically polarized to about -0.75 V SCE' The corrosion potential of an active zinc surface in neutral electrolytes is about -1.05 V SCE' This allows about a 0,3-V potential drop in the electrolyte caused by the galvanic current flow between the anode and steel cathode, assuming that very little polarization is associated with the zinc anode, The actual potential drop in the electrolyte depends also on the shape and geometry of the anode and cathode as well as the resistivity of the electrolyte. Current output is the galvanic current available for cathodic protection under a given set of conditions for an anode-electrolyte system, It can be defined as 1 = (E I - E 2 )IR,
404
CHAPTER 15
where E1 is the anode potential, E2 is the potential of steel under cathodic polarization, and R is the resistance in the electrolyte [1192]. In an active state the dissolution of a zinc anode generally has a low overpotential, and thus E1 is close to the open-circuit value of -1.05 V SCE• Anode efficiency is the percentage of the anodic dissolution due to galvanic action. The total anode consumption also includes self-corrosion (nongalvanic action). The efficiency can be determined by comparing the total electrical charge passing through from anode to cathode to the charge calculated from Faraday's law for the weight loss of the anode. 15.4.1.2. Effect of Anode Composition. Anode composition has long been recognized as critical in determining the performance of the anode. There are many patents on alloy composition, in which claims are made for the beneficial effects of specific alloying additions [1192]. In general, the composition of the alloy is designed to result in an anode possessing the following properties: 1. An electrode potential that is sufficiently negative, but not too negative, for a specific application. 2. A large current output at small polarizations, so that most of the potential difference between the anode and cathode can be used to polarize the cathode. 3. An electrode that remains active and does not become passivated throughout the anode life. 4. A high efficiency: a high galvanic corrosion rate versus a low self-corrosion rate. Specifications of alloy compositions for anodes and their basic perfonnance characteristics can be found in the literature [471,1192]. It has been well established that the presence of Fe in zinc anodes, even in very small amounts, e.g., 0.001 %, is harmful to the performance of the anode [230,470, 1194]. Fe in a zinc anode generally causes a reduction of current output and ennoblement of the anode potential, due to the formation of an insulating dissolution product film on the surface of the anode. Addition of a certain amount of Al can effectively remove the harmful effect of Fe. Carson et al. [470] found that the effects of alloy composition, time, and current density on anode perfonnance in seawater are interrelated. Figure 15.9 shows the potential of zinc and zinc-aluminum alloy anodes as a function of Fe content current density. In the case of the pure zinc anodes, the presence of Fe in the anode results in an ennoblement of the potential. This effect is very small for Fe contents less than 0.0003% but significantly increases for Fe contents higher than 0.001 %. Carson et al. also found that the extent of potential ennoblement increases with time. The effect of Fe disappears when small amounts of Al are present, as shown in Fig. 15.9b, c. Waldron and Peterson [230] investigated the effect of Fe, AI, and Cd on the perfonnance of zinc anodes in seawater. They found that the anodes containing small amounts of Al and Cd give higher total current outputs, as shown in Figs. 15.10 and 15.11. The effects of other common elements on the perfonnance of zinc anodes are relatively less significant compared to those of Fe and Al [469]. Crennel and Wheeler [1194] found that the addition of Sn has little effect whereas surface amalgamation is beneficial.
,.,,,;>
-r""
(a) 0% AI
1;;)':> ~e
(b) 0.3% AI
t
(c) 1.2% AI
Ht j
FIGURE 15 .9. Potential as a function of current density and iron content for zinc-aluminum alloy anodes. From Carson el al. [470]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
Po.
'0
.,c
lJ j //\
til
Q
"'"
(/)
-l
Z
:::: m
z
0
;;0
<
z
:r: m ;;0 m
-l
0
0 z Z
(/)
0
;;0 ;;0
0
n
406
CHAPTER 15
6,-----------------------------------------,
-;;5 til
:; o
s::.
3~--
o
____L __ _ _ _ _ _L __ _ _ _ _ _L __ _ _ _ _ _L __ _ _ _ 0.1
0.2
0.3
0.4
~
0.5
% Aluminum FIGURE 15.10. Effect of small amounts of aluminum in zinc anodes on current output in seawater. From Waldron and Peterson [230]. © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
Perkins and Bornholdt [464] identified the dissolution products formed on zinc anode surfaces in seawater as zinc oxide. The oxide film was porous and consisted of many discrete single-crystal plates on the order of 10-100 Jlm in diameter. The growth of the dissolution products was largely by plate broadening with limited thickening. The morphology of dissolution products of zinc anode has been found to be a function of current density and flow rate of electrolyte [1069]. 15.4.1.3. Efficiency and Effect oJ Temperature. Zinc anodes normally have a high efficiency of around 95% or higher [1192]. Waldron and Peterson [230] tested zinc anodes in seawater and reported that all showed an efficiency of nearly 100%, whether containing small amounts of Fe, AI, and Cd or not. 0.5% AI
6
------ ....
~~
0.1% AI
~ )(
til
:; 0
s::.
'"Q;
5
a. E
«
4L-----------~------------~------------~
o
0.08
0.04
0.12
% Cadmium FIGURE 15.11. Effect of small amounts of cadmium in zinc anodes on current output in seawater. From Waldron and Peterson [230). © Copyright by NACE International. All Rights Reserved by NACE; reprinted with permission.
407
CORROSION IN OTHER ENVIRONMENTS
TABLE 15.7. Effect of Temperature on Efficiency of a Zinc Alloy Tested for 28 Days with Impressed Current in Synthetic Seawater" (DC)
Galvanic efficiency (%)
Anodic potential (V seE)
21 38 43 54 60 69 74
95 91 89 86 84 76 70
-1.04 -1.03 -1.02 -1.01 -1.01 -1.00 -0.97
Ambient temperature
"From Kurr 1471J. © Copyright by NACE International. All Rights Reserved by NACE: reprinted with pennission.
The efficiency of a zinc anode may change with temperature. Kurr [471] reported that the efficiency of an aluminum-containing zinc alloy anode (0.1-0.5% AI) decreased with increasing temperature as shown in Table 15.7. At room temperature the efficiency was 95%, but at 74°C it was only 70%. The drop in efficiency is due to the increased self-corrosion associated with the higher intergranular corrosion of AI-containing zinc alloys at higher temperatures. Ahmed et al. [154] found that a Zn-0.3AI-0.03 Cd anode failed prematurely at 70°C in seabed mud owing to intergranular corrosion of the anode. Another phenomenon that may be associated with the performance of zinc anodes at higher temperatures is polarity reversal (see Chapter 7). It has been found that the predominant factor in polarity reversal in hot water is the interaction between zinc and the water and that the presence of a small amount of Fe in the zinc anode does not alter the extent ofreversal [447]. 15.4.2. Anodes for Impressed Current Cathodic Protection
In addition to its application as a sacrificial anode material, zinc has also been used as an anode material for impressed current cathodic protection of steel-reinforced concrete structures [267, 310]. In this application, zinc is thermally sprayed onto the surface of the concrete to form a continuous coating. During the operation of such a system, an anodic current, sufficient for the protection of the steel inside concrete, is impressed through the zinc coating via a rectifier. Good adhesion of the coating to the concrete surface and a small driving voltage under a given current density are required to ensure an effective performance of such a cathodic protection system [1286]. Brousseau et al. [520, 1286, 1287] investigated the factors affecting the adhesion and driving voltage. They found that the adhesion of sprayed zinc coatings on concrete can be affected by moisture content, temperature, current density, time, coating thickness, and concrete surface texture. It was concluded that, in order to maximize the adhesion, the concrete should be dry, warm, and grit-blasted at low air pressure without exposing too much aggregate. These authors also found that the driving voltage tends to increase with the applied current density and polarization time owing to the formation of zinc dissolution products at the coating/concrete interface. Sprayed Zn-15%AI is not suitable for using as an impressed current anode; severe disbonding and blistering of the coating was observed during the impressed current test with this alloy.
References I. Pourbaix, M.: Alias of Eleclrochemical Equilibria in Aqueous Solutions, 2nd ed., National Association of Corrosion Engineers, Houston, pp. 406-413,1974. 2. Nevison, D. C. H.: Corrosion of zinc, in ASM Mewls Handbook, 9th ed., Vol. 13, 755-769, ASM International, Materials Park, Ohio, 1987. 3. Kannangara, D. C. W., and Conway, B. E.: Zinc oxidation and redeposition processes in aqueous alkali and carbonate solutions, 1. Electrochem. Soc. 134,894-918,1987. 4. Mattsson, E.: The atmospheric corrosion properties of some common structural metals-a comparative study, Mater. Peiform. 21,9-19,1982. 5. Frydrych, D. 1.: Corrosion mechanisms of zinc-rich organic coatings on steel, Ph. D. Thesis, University of Pennsylvania, 1986. 6. Kita, H.: Periodic variation of exchange current density of hydrogen electrode reaction with atomic number and reaction mechanism, 1. Electrochem. Soc. 113, 1095-1111, 1966. 7. Lee, T. S.: Hydrogen overpotentlal on pure metals in alkaline solution, 1. Electrochem. Soc. 118, 1278-1282, 1971. 8. Bockris, J. 0' M., and Reddy, A. K. N.: Modern Electrochemistry, Vol. l, Plenum Press, New York. 1970. 9. Lee, T. S.: Hydrogen overpotential on zinc alloys in alkaline solution, 1. Electrochem. Soc. 122, 171-173,1975. 10. Mansfeld, F., and Gilman, S.: The effect of several electrode and electrolyte additives on the corrosion and polarization behavior of the alkaline zinc electrode, 1. Electrochem. Soc. 117, 1328-1333, 1970. II. Bard, A. J. (ed.): Encyclopedia of Electrochemislry of the Elements, Vol. IX, Part A, Marcel Dekker, New York, 1982. 12. Bockris, J. O. M., Nagy, Z., and Danjanovic, A.: On the deposition and dissolution of zinc in alkaline solutions, 1. Eleclrochem. Soc. 119,285-295, 1972. 13. Uhlig, H. H., and Revie, R. w.: Corrosion and Corrosion Control-An Introduction to Corrosion Science and Engineering, 3rd ed., John Wiley & Sons, New York, 1985. 14. Berke, N. S., and Friel, J. J.: Applications of electrochemical techniques in screening metallic-coated steels for allTIosphenc use, In Laborarory CorrosIOn Tests and Standards, STP 866, pp. 143-158, American Society for Testing and Materials, Philadelphia, 1985. 15. Gad Allah, A. G., Hefny, M. M., Salih, S. A., and EI-Basiouny, M. S.: Corrosion inhibition of zinc in HCI solution by several pyrazole derivatives, Corrosion 45, 574-578, 1989. 16. Augustynski, J., Dalard, F., and Sohm, 1. c.: Oxydation anodique du zinc en milieu faiblement basique, Corros. Sci. 12,713-724, 1972. 17. Augustynski, J.: Etude de la rupture de passivite de certains metaux electrochimiquement actifs, Corros. Sci. 13, 955-965, 1973. 18. Liu, M., Cook, G. M., and Yao, N. P.: Passivation of zinc anodes in KOH electrolytes, 1. Electrochem. Soc. 128, 1663-1668, 1981. 19. Bocchi, N., da Cunha, M. R., and D' Alkaine, C. V: The passivating films of zinc in alkaline solution, Key Eng. Maler. 20-28, 3941-3946, 1988.
409
410
REFERENCES
20. Aurian-Blajeni, B., and Tomkiewicz, M.: The passivation of zinc in alkaline solutions, J. Electrochem. Soc. 132, 1511-1515, 1985. 21. De Pinto, I. S., De Pauli, C. P., Herrera, H .. Mishima, H., and Lopez, B. A: Effect of arsenate anion on the dissolution and passivation of zinc electrode in slightly alkaline solutions, Electrochim. ACTa 31, 527-533, 1986. 22. Bocchi, N., and D' Alkaine, C. Y.: Zinc behavior in slightly alkaline solutions. The reduction processes, Key Eng. Mater. 20-28,417-423, 1988. 23. D' Alkaine, C. Y., and da Cunha, M. R.: Influence of the coj- on the passivation of zinc in slightly alkaline solutions, International Congress on Metallic Corrosion, Toronto, Ontario, pp. 122-125, 1986. 24. Chang, Y, and Prentice, G.: A model for the anodic dissolution of the zinc electrode in the prepassive region, J. Electrochem. Soc. 136,3398-3403,1989. 25. Elder, 1. P.: The electrochemical behavior of zinc in alkaline media, J. Electrochem. Soc. 116,757-762, 1969. 26. Powers, R. W., and Breiter, M. W.: The anodic dissolution and passivation of zinc in concentrated potassium hydroxide solutions, J. Electrochem. Soc. 116,719-729, 1969. 27. Powers, R. W.: Film formation and hydrogen evolution on the alkaline zinc electrode, J. Electrochem. Soc. 118,685-695, 1971. 28. Szpak, S., and Gabriel, C. 1.: The Zn-KOH system: The solution-precipitation path for anodic ZnO formation,J. Electrochem. Soc. 126, 1914-1923, 1979. 29. Powers, R. W.: Anodic films on zinc and the formation of cobwebs, 1. Electrochem. Soc. 116, 1652-1659,1969. 30. Burleigh, T. D.: Anodic photocurrents and corrosion currents on passive and active-passive metals, Corrosion 45, 464-472,1989. 31. van Ooij, W. 1., and Sabata, A: Under-vehicle corrosion testing of primed zinc and zinc alloy-coated steels, Corrosion 46, 162-171, 1990. 32. Darwish, N. A: Contribution to the electrochemical behaviour of some Zn-Ni systems in halide media, J. Electrochem. Soc. India 32, 259-267, 1983. 33. Dattilo, M.: Polarization and corrosion of electrogalvanized steel-evaluation of zinc coatings obtained from waste-derived zinc electrolytes, J. Electrochem. Soc. 132,2557-2561,1985. 34. Wu, 1. K., and Hsu, Y S.: Electrochemical corrosion of zinc in sodium chlorite, J. Mater. Sci. 21, 3475-3478, 1986. 35. Lee, H. H., and Hiam, D.: Corrosion resistance of galvannealed steel, Corrosion 45,852-856, 1989. 36. Budinski, M. K., and Wilde, B. E.: An electrochemical criterion for the development of galvanic coating alloys for steel, Corrosion 43, 60-62, 1987. 37. Abd EI Haleem, S. M.: Environmental factors affecting the pitting corrosion potential of a zinc-titanium alloy in sodium hydroxide solutions, Br. Corros. J. 14, 171-175. 1979. 38. Shams EI Din, AM., Abd EI Kader, 1. M., and Badran, M. M.: Galvanic corrosion in the copperlzinc system. Part i. Potential distribution in relation to the nature of the cathode and to the type of anion in solution, Br. Corros. 1. 16,32-37,1981. 39. Hutchison, P. E, and Turner, 1.: Some aspects of the electrochemical behavior of zinc in the presence of acrylate and methacrylate anions, J. Electrochem. SOC. 123, 183-186, 1976. 40. Hefny, M. M., Gad-Allah, A. G., Salih, S. A, and EI-Basiouny, M. S.: Zinc passivation in oxalate solution, Corrosion 44, 691-695,1988. 41. LeRoy, R. L.: Evaluation of corrosion rates from nonlinear polarization data, J. Electrochem. Soc. 124, 1006-1012,1977. 42. Kapali, V., Srinivasan, K. N., Venkataraman, B., and Balakrishnan, K: Development of a series of AI-Zn-In ternary alloy anodes for the cathodic protection of submerged structures, Key Eng. Mater. 20-28, 955-962, 1988. 43. Lambert, M. R, Hart, R. G., and Townsend, H. E.: Corrosion mechanism ofZn-Ni alloyelectrodeposited coatings, at The 2nd Automotive Corrosion Prevention Conference, Dearborn, Michigan, 1983, Paper No. 831817. 44. Short, N. R., Abibsi, A, and Dennis, 1. K.: Corrosion resistance of electroplated zinc alloy coatings, Trans. Inst. Met. Finish. 67 (Part 3), 73-77, 1989. 45. Alvarez, M. G., and Galvele, J. R.: Pitting of high purity zinc and pitting potential significance, Corrosion 32, 285-294, 1976.
411
REFERENCES
46. Abd El Haleem. S. M.: Dissolution current and pitting potential of zinc in KOH solutions in relation to the concentration of aggressive ions, Br. Corros. 1.11,215-218,1976. 47. Szklarska-Smialowska, Z.: Pitting Corrosion of Metals, National Association of Corrosion Engineers, Houston, 1986. 48. Devillers, L. P.: The mechanism of aqueous intergranular corrosion in zinc-aluminium alloys, Ph. D. Thesis, University of Waterloo, 1974. 49. Devillers, L. P., and Niessen, P.: The mechanism of intergranular corrosion of dilute zinc-aluminium alloys in hot water, Corros. Sci. 16,243-252, 1976. 50. Devillers, L. P., and Niessen, P.: The mechanism of intergranular corrosion of dilute zinc-aluminium alloys in hot water, Corros. Sci. 16,243-252,1976. 51. American Hot Dip Galvanizers Association, Wet Storage Stain, Brochure, 1984. 52. Helwig, L. E.: Comparison of humid-storage tests for chromated galvanized sheet, Met. Finish. 75(5), 33-38, 1977. 53. Bird, C. E., and Strauss, F. J.: Effect of wet storage staining on subsequent atmospheric corrosion of galvanized iron sheets. Mater. Perform. 15(11),27-29, 1976. 54. Cominco PTC Project No. 83-5-11, August 1990. 55. EI-Mallah, A. T.. and Magd, M. R. G. A: Chromate passivation of zinc, Met. Finish. 82(2),49-54, 1984. 56. Helwig, L. E.: Chromate treatment of galvanized sheet, Met. Finish. 68(7), 54-59. 1970. 57. Williams, L. F. G.: Chromate conversion coatings on zinc, Plating 1972, 931-938. 58. Williams, L. F. G.: The formation and performance of chromate conversion coatings on zinc, Surf. Technol. 5,105-117,1977. 59. Williams, L. F. G.: The mechanism of formation of chromate conversion films on zinc, Surf Technol. 4,355-366,1976. 60. Paatsch. w.: Quality testing of coatings produced by chromating zinc, Met. Finish. 76(9), 15-17, 1978. 61. Epelboin, I., Ksouri, M., and Wiart, R.: On a model for the electrocrystallization of zinc involving an autocatalytic step, 1. Electrochem. Soc. 122, 1206-1214, 1975. 62. McBreen, J., and Cairns, E. J.: The zinc electrode, in Advances in Electrochemistry and Electrochemical Engineering, Vol. II, H. Gerischer and C. W. Tobias (eds.), pp. 273-352, John Wiley & Sons, New York. 1978. 63. Abdel Aal, M. S" Radwan, S .. and EI-Saied, A: Phenothiazines as corrosion inhibitors for zinc in NH4 CI solution, Br. Corros. 1.18, 102-106, 1983. 64. Fiaud, c., Bensarsa, S., des Aulnois, I. D., and Tzinmann, M.: Inhibiting properties of phosphines against zinc corrosion in acidic media. Br. Corros. 1. 22,109-112,1987. 65. Ostrander. C. W.: Chromate conversion coatings, in Electroplating Engineering Handbook, 3rd ed., A. K. Graham (ed.), pp. 437-447, Van Nostrand Reinhold, Toronto, 1971. 66. Krishnamurthy, S.: Chromate conversion coatings on zinc plating, 1. Electrochem. Soc. India 30, 47 -51, 1981. 67. Van de Leest, R. E.: Yellow chromate conversion coatings on zinc: chemical composition and kinetics, Trans. Inst. Met. Finish. 56, 51-54,1978. 68. Lazauskas, v., Sarmaitis, R., Butkevicius. J" and Matulis, J.: The role of sulfate ions in chromating zinc and cadmium, Plating Surf. Finish. 66(1),58-60, 1979. 69. Helwig, L. E., and Bird, J. E.: Protection of galvanized coatings from attack by organic acid vapors, Met. Finish. 7l( II), 45-48, 52. 1973. 70. Helwig, L. E., and Swan, J. D.: Use of spot-test reagents for detecting chromates on hot-dip galvanized surfaces, Met. Finish. 80(9),37-40, 1982. 71. Duncan, J. R.: Electron spectroscopy of chromated galvanized steel sheet after heating, immersion in water or outdoor weathering, Surf. Techno!. 17,265-276, 1982. 72. Lengyel, B., Baroti-Labar, Z., and Kahan, R.: Study of chromate coatings by means of electrode impedance measurements, Mater. Chem. 7,183-194,1982. 73. Efimov, E. A, Tok, L. D., and Tverdynina, T. B.: Mechanism of the reduction of chromic acid anions to trivalent chromium ions, SOy. Electrochem. 25,1249-1251,1989 [in English]. Chromate passivation protection of Zn- and Al-Zn-coated steel 74. Townsend, H. E., and Zoccola, J. sheet against wet-storage stain, 1. Electrochem. Soc. 125, 1290-1292, 1978. 75. Van de Leest, R. E.: Yellow chromate conversion coatings on zinc: Electrochemical measurements and corrosion tests, Werkst. Korros. 29.648-653, 1978 [in English].
c.:
412
REFERENCES
76. Pearlstein, E: Selection and application of inorganic finishes. Part II. Chromate coatings, Plating Suif. Finish. 66(1), 30-34,1979. 77. Treverton, J. A., and Amor, M. P.: The structures and surface composition of chromate conversion coatings: An XPS and SEM study, Trans. Inst. Met. Finish. 60, 92-96, 1982. 78. Van de Leest, R E., and Wessels, J.: Stabilization of black chromate conversion coatings on zinc, Plating Surf Finish. 67(5), 86-89, 1980. 79. Spring, S., and Woods, K.: Prepaint treatments-chromate coatings, Met. Finish. 79(6),49-55, 1981. 80. Choosing chromate and phosphate processes for corrosion resistance, Corros. Prevo Control 1981 (February), 16-18,32-33. 81. Rozovsky, V. G., and Sarmaitis, R R: Corrosion resistance of black chromate conversion films on zinc, Plating Suif. Finish. 68(10), 54-56,1981. 82. Sarmaitis, R: Corrosion resistance of chromate films on zinc, Plating Suif. Finish. 68(9), 83-86, 1981. 83. Perfetti, B. M.: A comparison of pretreatments for prepainted galvanized sheets, Met. Finish. 81(5), 57-60, 1983. 84. Timmins, E D.: Adhesion of paint to zinc surfaces, Ind. Carras. 1(3), 15-17, 1983. 85. Murphy, M.: Technical developments in 1983. Inoraganic metallic finishes, processes and equipment, Met. Finish. 82(2), 17-32, 1984. 86. Gallaccio, A., Pearlstein, E, and D' Ambrosio, M. R: Effects of heating chromate conversion coatings, Met. Finish. 64(8),50-54, 1966. 87. Bardin, P. C: Conversion coatings-prepare galvanized steel for painting, Ind. Finish., 1967 (July), 58-61. 88. Pearlstein, E, and D'Ambrosio, M. R: Heat resistant chromate conversion coatings, Plating 55. 345-346,1968. 89. Helwig, L. E.: Heavy chromate conversion treatments on zinc-coated steel, Met. Finish. 70(9),51-55, 1972. 90. Chromate conversion coatings for zinc & cadmium, Met. Finish. Plants Processes 7(1),15-24, 1971. 91. Bridger, R. D. E. L.: Pretreatment of galvanized surfaces prior to powder coating, Trans. Inst. Met. Finish. 61, 13-18, 1983. 92. Centre de Recherches Metallurgiques (Belgium): White Rust Inhibition for Galvanized Steel. International Lead Zinc Research Organization Project ZE-348, Final Report, February 1989. 93. Jardy, A., Rosset, R, and Wi art, R: Diphosphate coatings for protection of galvanized steel: quality control by impedance measurements, 1. Appl. Electrochem. 14,537-545, 1984. 94. Kargol, J. A., and Jordan, D. L.: The influence of phosphorous alloy additions on the zinc phosphate coating formation on cold rolled steels, Corrosion 38(4), 201-206, 1982. 95. Nancollas, G. H.: Phosphate precipitation in corrosion protection: Reaction mechanisms, Corrosion 39(3),77-82, 1983. 96. EI-Mallah, A. T., Morsy, S. M., and Ibrahim, I. G.: Electrophosphate coatings on zinc, Met. Finish. 82(8), 31-72,1984. 97. Leidheiser, H., and Suzuki, I.: Cobalt and nickel cations as corrosion inhibitors for galvanized steel, 1. Electrochem. Soc. 128,242-249, 1981. 98. Bijimi, D., and Gabe, D. R.: Passivation studies using group VIA anions. Part III. Anodic treatment of zinc, Br. Carras. 1. 18, 138-141, 1983. 99. Kargol, J. A., Jordan, D. L., and Palermo, A. R.: The influence of high strength cold rolled steel and zinc coated steel surface characteristics on phosphate pretreatment, Corrosion 39, 213-218, 1983. 100. Wrubl, C, and Mor, E. D.: Zinc gluconate as an inhibitor of the corrosion of copper and zinc in sea water, Br. Carras. 1.18,142-147,1983. 101. Kuznetsov, Y. I., and Podgornova, L. P.: Corrosion and electrochemical behavior of zinc in phosphate solutions, Zashch. Met. 19,98-102,1983 [in English). 102. Wilcox, G. D., and Gabe, D. R: Chemical molybdate conversion treatments for zinc, Met. Finish. 86(9), 71-74,1988. 103. Leidheiser, H., and Suzuki, I.: Towards a more corrosion resistant galvanized steel, Corrosion 36, 701-702, 1980. 104. Macias, A., and Andrade, C.: Corrosion of galvanized steel reinforcements in alkaline solutions. Part I. Electrochemical results, Br. Carras. 1. 22,113-118,1987.
REFERENCES
413
105. Macias. A .. and Andrade. c.: Corrosion of galvanized steel reinforcements in alkaline solutions. Part II. SEM study and identification of corrosion products, Br. Corros . .I. 22. 119~ 129. 1987. 106. Leidheiser. H., Simmons. G. W, Nagy. S., and Music, S.: Mossbauerspectroscopic study of the corrosion inhibition of zinc by cobalt ions,./. Electrochem. Soc. 129. 16S8~1662. 1982. 107. Leidheiser. H., Momose. Y, and Granata, R. D.: Inhibition of the cathodic reaction on zinc in aerated 3% NaCI solution, Corrosion 38, 178~ 179, 1982. 108. LeRoy, R. L.: Organic wet storage stain inhibitor-a substitute for chromates, Mate/: Pofilnn. 19(8). 54~55, 1980. 109. Shams EI Din. A. M., Abd EI Haleem, S. M .. and Abd EI Kader, 1. M.: Studies on the pitting corrosion of zinc in aqueous solutions. Part II. Measurement of pitting corrosion currents operating under natural conditions, .l. Electroanal. Chem. 65, 335~349, 1975. 110. Baugh, L. M.: Corrosion and polarization characteristics of zinc in neutral~acid media. Part I. Pure zinc in solutions of various sodium salts, Elcctrochim. Acta. 24, 657~667, 1979. III. Baugh, L. M.: Corrosion and polarization characteristics of zinc in neutral~acid media. Part II. Effect of NH! ions and the role of battery zinc alloying constituents. Electrochim. Acta 24. 669~677, 1979. I 12. Yamashita, T.: Electro-deposition and electro-dissolution of zinc in the vicinity of equilibrium potential in aqueous zinc sulfate solutions . .I. Electroanal. Chel1l. 106, 95~ 102, 1980. 113. Deslouis. c.. Duprat. M .. and Tuiet-Tournillon, c.: The cathodic mass transport process during zinc cOlTosion in neutral aeratcd sodium sulphate solutions,.I. Electmanal. Chem. 181, 119~136. 1984. 114. Boto. K. G .. and Williams. L. F. G.: The determination of the cOlTosion rate of zinc in solution by the differential pulse method . ./. Electrochem. Soc. 124. 6S6~661, 1977. liS. Era, A .. Takehara. Z .. and Yoshizawa. S.: Influence of impurities especially lead contained in manganese dioxide upon the self-discharge of the leclanche dry cell, Electrochim. Acta 13. 183~396. 196X 116. Zembura. Z .. and Burzynska. L.: The corrosIOn of zinc III de-aerated O. I M NaCIIn the pH range trom I. 6to 13.3. Corms. Sci. 17. 871~878. 1477. 117. Fujishima. A .. Kato. T., Maekawa. E.. and Honda, K.: Mechanism of the currem doubling effect. Part I. The ZnO photaanode in aqueous solution of sodium formate, Bull. Chern. SoC, .lpn. 54. 1671 ~ 1674. 1984. 118. Walter, G. W: Corrosion rates of zinc, zinc coatings and steel in aerated slightly acidic chloride solutions calculated from low polarization data. Corms. Sci. 16, S73~586. 1976. 119. Helwig. L. E.: Premature darkening of galvanized roofing sheets. Part L Mel. Finish. 82(4). 41~46, 1484. 120. Helwig. L. E.: Premature darkening of galvanized roofing sheets. Part II, Met. Finish. 82(5). 61~64. 1984. 121. Mansfeld, F., and Kenkel, J. V: Electrochemical monitoring of atmospheric corrosion phenomena. Corms. Sci. 16. III ~ 122, 1976. 122. Yan, B., Meilink. S. L., Warren. G. W. and Wynblatt. P.: IEEE Trans. Componellls, Hybrids Mallufac· turing Technoi .. CHMT-IO. 247, 1987. 123. Stratmann, M.: The investigation of the corrosion properties of metals, covered with adsorbed electrolyte layers-a new experimental technique, Corros. Sci. 27, 869~872, 1987. 124. Stratmann, M., and Streckel, H.: On the atmospheric corrosion of metals which are covered with thin electrolyte layers. Part I. Verification of the experimemal technique, Corms. Sci. 30, 681 ~696. 1990. 125. Stratmann, M .. and Streckel. H.: On the atmospheric corrosion of metals which are covered with thin electrolyte layers. Part II. Experimental results, Corros. Sci. 30, 697~ 714, 1990. 126. Stratmann, M .. Streckel, H .. Kim, K. T., and Crockett, S.: On the atmospheric corrosion of metals which are covered with thin electrolyte layers. Part III. The measurement of polarisation curves on metal surfaces which are covered by thin electrolyte layers, Carras. Sci. 30, 715~ 734, 1990. 127. Kaesche, H.: The passivity of zinc in aqueous solutions of sodium carbonate and sodium bicarbonate, Electrochim. Acta 9, 383~394, 1964. 128. Bota, K. G., and Williams, L. F. G.: Rotating disc electrode studies of zinc corrosion,./. Electroanal. Cizem. 77. 1~20, 1977. 129. Gerischer. H.: Electrochemical behavior of semiconductors under illumination,.I. Eieelrochem. Soc. 113, 1174~ I 182, 1966. 130. Epelboin, I.. Ksouri, M., and Wi art, R.: On a model for the electrocrystallization of zinc involving an autocatalytic step,.I. Electmehem. Soc. 122, 1206~ 1214, 1975.
414
REFERENCES
131. Isaacson, M. J., McLarnon, E R, and Cairns. E. J.: Zinc electrode rest potentials in concentrated KOH-K2Zn(OH)4 electrolytes. 1. Electrochem. Soc. 137,2361-2364,1990. 132. St-Pierre, J .• and Piron. D. L.: Mechanism of cathodic potential oscillations of the zinc electrode in alkaline solutions. 1. Electrochem. Soc. 137,2491-2498, 1990. 133. Bard, A. J .. and Faulkner, L. R: Electrochemical Methods-Fundamentals and Applications, John Wiley & Sons. New York. 1980. 134. Mansfeld, E, Kendig, M. w., and Tsai. S.: Recording and analysis of AC impedance data for corrosion studies. Part II. Experimental approach and results, Corrosion 38. 570-580. 1982. 135. Mansfeld. E, and Kendig, M.: Evaluation of protective coatings with impedance measurements, Proceedings of the International Congress on Metallic Corrosion, Toronto. Ontario. 1984. Vol. 3. pp. 74-84. 136. Mansfeld, E: Recording and analysis of AC impedance data for corrosion studies. Part I. Background and methods of analysis. Corrosion 36. 301-307,1981. 137. Macdonald, D. D.: Theoretical analysis of electrochemical impedance, National Association of Corrosion Engineers Corrosion '87 Conference, San Francisco, March 9-13,1987, Paper No. 479. 138. Macdonald, D. D.: Some advantages and pitfalls of electrochemical impedance spectroscopy, Corrosion 46.229-242 (1990). 139. Brown, R P., and Kessler. R. J.: One year laboratory testing of plain and galvanized rebars in chloride containing concrete, Dept. of Transportation, Florida Corrosion Report No. 78-2, 1978. 140. Silverman, D. c.: Rapid corrosion screening in poorly defined systems by electrochemical impedance technique. Corrosion 46. 589-598, 1990. 141. Mansfeld, E. Kendig, M. W., and Tsai, S.: Evaluation of corrosion behavior of coated metals with AC impedance measurements. Corrosion 38, 478-485,1982. 142. Sagues, A A: Corrosion measurements of reinforcing steel in concrete exposed to various aqueous environments, National Association of Corrosion Engineers Corrosion' 87 Conference, San Francisco, March 9-13,1987, Paper No. 118. 143. Wheat, H. G., and Eliezer, Z.: Comments on the identification of a chloride threshold in the corrosion of steel in concrete, Corrosion 43,126-128,1987. 144. Kendig, M., and Mansfeld, E: Corrosion rates from impedance measurements: An improved approach for rapid automatic analysis, Corrosion 39, 466-467, 1983. 145. Matsuoka, K.: Monitoring of corrosion of reinforcing bar in concrete, National Association of Corrosion Engineers Corrosion '87 Conference. San Francisco, March 9-13, 1987, Paper No. 121. 146. Berthier, E, Diard. J. P.,Jussiaume, A, and Rameau, J. J.: The instantaneous impedance of non-stationary electrochemical systems: Application to a corroding zinc electrode, Corros. Sci. 30. 239-247, 1990. 147. Hendrikx, J., Visscher, W., and Barendrecht, E.: Impedance measurements of zinc and amalgamated zinc electrodes in alkaline electrolyte, Electrochim. Acta 30, 999-1006, 1985. 148. Rodgers, R .. Rothstein, M. L.. and Palus, A E: Instrumentation for the automated A.C. impedance analysis of electrochemical systems, Proceedings of the National Association of Corrosion Engineers Metallography and Corrosion Conference. July 1986, pp. 203-220. 149. Castle, J. E.: Corrosion test methods, Br. Corros. 1. 22, 77-82,1987. I SO. Siebert, O. W.: Laboratory electrochemical test methods, Laboratory Corrosion Tests and Standards, STP 866, American Society for Testing and Materials, Philadelphia, 65-90, 1985. 151. American Society for Testing and Materials: Standard reference test method for making potentiostatic and potentiodynamic anodic polarization measurements. in Annual Book ofASTM Standards, Vol. 03.02, No. G5-87, pp. 79-85,1989. 152. American Society for Testing and Materials: Standard practice for conventions applicable to electrochemical measurements in corrosion testing, in Annual Book of ASTM Standards, Vol. 03.02, No. G3, pp.62-67, 1989. 153. American Society for Testing and Materials: Standard practice for conducting potentiodynamic polarization resistance measurements, in Annual Book of ASTM Standards, Vol. 03.02. No. G59-78, pp. 221-224, 1989. 154. Ahmed, D. S., Ashworth, v., Scantlebury, J. D., and Wyatt, B. S.: Mechanism of intergranular attack in Zn-AI-Cd sacrificial anodes at elevated temperatures, Br. Corros. 1. 24, 149-152, 1989. ISS. Clarke, S. G., and Andrew, J. E: The chromate passivation of zinc, 1. Electrodepos. Tech. Soc. 20, 119-138,1945.
REFERENCES
415
156. Stratmann, M., Streckel, H., Kim, K. T., and Crockett. S.: On the atmospheric corrosion of metals which are covered with thin electrolyte layers. Part III. The measurement of polarisation curves on metal surfaces which are covered by thin electrolyte layers, Corros. Sci. 30, 715-734, 1990. 157. Stratmann, M., and Streckel, H.: On the atmospheric corrosion of metals which are covered with thin electrolyte layers. Part II. Experimental results, Corros. Sci. 30, 697-714. 1990. 158. Gonzalez, J. A., Otero. E .. and Cabanas. c.: Recording polarisation curves in thin electrolyte layers by electrochemical atmospheric corrosion monitors. Br. Corros. 1. 25.125-130,1990. 159. White, R. T.: A review of corrosion inhibitors. Part I, Corrosion and Coatings South Africa 1986(8), 2. 4,6,8. 160. White, R T.: A review of corrosion inhibitors. Part II, Corros. Coatings S. Afr. 1986( 10), 4, 6, 10. 161. Leidheiser, H.: A review of proposed mechanisms for corrosion inhibition and passivation by metallic cations, Corrosion 36, 339-345, 1980. 162. Leroy, R L.: Polythoiglycolate passivation of zinc, Corrosion 34. 113-119, 1978. 163. Rausch, w., and Mueller, G., (MetallgesellschaftA. G., Germany): Process for phosphating metals, U.S. Patent 4637838. January 20, 1987. 164. Abdel Aal, M. S., Abdel Wahab, A. A., and EI Saied, A.: A study of the inhibiting action of benzene thiols and related compounds on the corrosion of zinc in acidic media, Corrosion 37, 557-563, 1981. 165. Penninger, J. (Henkel Kommanditgesellschaft auf Aktien, Germany): Method for inhibiting corrosion of zinc using bis-triazoles, U.S. Patent 4636359, January 13, 1987. 166. Abd EI Haleem, S. M., and Abdel Fattah, A. A.: The role of some organic anions in promoting or inhibiting the corrosion of zinc, Surf Coatings Techno/. 29, 41-49. 1986. 167. Massinon, D., Dauchelle. D., and Charbonnier, 1. c.: On the contribution of electrochemical methods in the study of corrosion mechanisms in automotive body steel sheets, Mater. Sci. Forum, 44-45, 461-476, 1989. 168. Shastry, C R, and Townsend. H. E.: Mechanisms of cosmetic corrosion in painted zinc and zinc-alloycoated sheet stcels, Corrosion 45, 103-118. 1989. 169. Shastry, C. R., and Townsend, H. E.: Failure mechanisms of painted sheet steel in a chipping corrosion test, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo. September 5-7, 1989, pp. 511-518. 170. Serra, E. T., and de L. Fragata, E: Surface treatments of galvanized steel for painting, International Congress on Metallic Corrosion. Toronto, Ontario, 1984, pp. 384-389. 171. Lengyel, 8., Meszaros, L., Kahan, R., and Jeney, I.: Study of chromate coatings by impedance technique, Proceedings of the Symposium on Electroplating, Budapest, Hungary, April 16-19, 1985. 172. Abu Zahra, R. H .. and Shams EI Din, A. M.: A radiotracer study on the reaction between zinc and chromate, Corros. Sci. 5, 517-524.1965. 173. Friel. J. J.: Atmospheric corrosion products on AI, Zn. and AIZn metallic coatings. Corrosion 42, 422-426, 1986. 174. Sergi, G., Short. N. R., and Page. C L.: Corrosion of galvanized and galvannealed steel in solutions of pH 9. to 14.0, Corrosion 41. 618-624, 1985. 175. Macias, A., and Andrade, c.: Corrosion rate of galvanized steel immersed in saturated solutions of Ca(OH}z in the pH range 12-13. 8, Br. Corros. 1.18,82-87,1983. 176. Jones, D. A .. and Nair, N. R.: Electrochemical corrosion studies on zinc-coated steel, Corrosion 41, 357-362, 1985. 177. Maahn, E .. and Sorensen, B.: The influence of microstructure on the corrosion properties of hot-dip galvanized reinforcement in concrete, Corrosion 42,187-196,1986. 178. Kiss, K., and Coll-Palagos, M.: Cyclic voltammetric and scanning electron microscopic evaluation of the corrosion resistance ofZn-phosphated steel, Corrosion 43,8-14.1987. 179. Mor, E. D., and Beccaria, A. M.: Effect of temperature on the corrodibility of copper and zinc in synthetic sea water, Corrosion 31, 275-279, 1975. 180. Batrakov, Y Y, Bogdanov, N. Y, and Kocherov, A. V.: The application of alternating current method for adsorption and corrosion measurement on iron and zinc, Proceedings of the International Congress on Metallic Corrosion, Toronto, Ontario, June 3-7,1984, Vol. I, pp. 287-290. 181. Johnson, 1. W., Sun, Y. C, and James, W. 1.: Anodic dissolution ofZn in aqueous salt solutions, Corros. Sci. 11, 153-159. 1971.
°
416
REFERENCES
182. James, W. J., Straumanis, M. E., and Johnson, J. W: Anodic disintegration of metals undergoing electrolysis in aqueous salt solutions, Corrosion 23,15-23,1967. 183. Stiles, D. C, and Edney, E. 0.: Dissolution of zinc into thin aqueous films as a function of residence time, acidic species, and pH, Corrosion 45, 896-901. 1989. 184. Yee, S., Stratmann, M., and Oriani, R. A: Application of a Kelvin microprobe to the corrosion of metals in humid atmospheres, 1. Electrochem. Soc. 138, 55-61, 1991. 185. Haynie, F. H., Spence, J. W, and Upham, 1. B.: Effects of air pollutants on weathering steel and galvanized steel: A chamber study, in Atmospheric Factors Affecting the Corrosion of Engineering Metals, STP 646, pp. 30-47, American Society for Testing and Materials, Philadelphia, 1978. 186. German, G.: Behavior of zinc-coated steel in highway environments, in Atmospheric Factors Affecting the Corrosion of Engineering Metals, STP 646, pp. 74-82, American Society for Testing and Materials, Philadelphia, 1978. 187. LegaUlt, R. A, and Pearson, Y. P.: Kinetics of the atmospheric corrosion of galvanized steel, in Atmospheric Factors Affecting the Corrosion of Engineering Metals, STP 646, pp. 83-96, American Society for Testing and Materials, Philadelphia, 1978. 188. Payer, J. H., and Safranek, W H.: Atmospheric corrosion of electroplated zinc alloy die castings, in Atmospheric Factors Affecting the Corrosion of Engineering Metals, STP 646, pp. 115-128, American Society for Testing and Materials, Philadelphia, 1978. 189. Zoccola,1. C, Townsend, H. E., Borzillo, A. R., and Horton, J. B.: Atmospheric corrosion behavIOr of aluminum-zinc alloy-coated steel, in Atmospheric Factors Affecting the Corrosion of Engineering Metals, STP 646, pp. 165-184, American Society for Testing and Materials, Philadelphia, 1978. 190. Guttman, H.: Effects of atmospheric factors on the corrosion of rolled ZInC, in Mewl Corrosion in the Atmosphere, STP 435, pp. 223-239, American Society for Testing and Materials, Philadelphia, 1968. 191. Guttman, H., and Sereda, P. J.: Measurement of atmospheric factors affecting the corrosion of metals, in Metal Corrosion in the Atmosphere, STP 435, pp. 326-359, American Society for Testing and Materials, Philadelphia, 1968. 192. Guttman, H.: Atmospheric and weather factors in corrosion testing, in Atmospheric Corrosion, W. H. Ailor (ed.), pp. 51-68, John Wiley & Sons, New York, 1982. 193. Mansfeld, E, and Kenkel, J. Y.: Electrochemical studies of atmospheric corrosion, Corrosion '75, Symposium on the Protection and Performance of Materials, Toronto, Ontario, April 14-18, 1975, Paper No. 161. 194. Glass, G. K., and Ashworth, Y.: The corrosion behaviour of the zinc-mild steel galvanic cell in hot sodium bicarbonate solution, Corros. Sci. 25, 971-983,1985. 195. Mansfeld, E, Jeanjaquet, S. L., Roe, D. K., and Carey, R.: A new corrosion rate monitor and data logging system for atmospheric corrosion studies, National Association of Corrosion Engineers Corrosion' 87 Conference, San Francisco, March 9- 13, 1987, Paper No.4 I I. 196. Hubbard, D. J., and Shanahan, C E. A.: Corrosion of zinc and steel in dilute aqueous solutions, Br. Corros.l. 8, 270-274,1973. 197. Macias, A., and Andrade, C: Corrosion of galvanized steel in dilute Ca(OHh solutions (pH I I. I - I 2.6), Br. Corros. 1.22,162-171,1987. 198. Calvo, E A., Otero, E., and Pardo, A.: Corrosion protection of galvanized steel pipe in acidic media by surface amalgamation, Br. Corros. 1. 22, 185-189, 1987. 199. Wilcox, G. D., and Gabe, D. R.: Passivation studies using group VIA anions. Part Y. Cathodic treatment of zinc, Br. Corros. 1. 22, 254-258, 1987. 200. Andrade, C, and Castelo, Y.: Practical measurement of the AC impedance of steel bars embedded in concrete by means of a spectrum analyser (fast Fourier transform), Br. Corros. 1. 19, 98-100, 1984. 201. Gonzalez, J. A, Otero, E., Cabanas, C, and Bastidas, 1. M.: Electrochemical sensors for atmospheric corrosion rates. A new design, Br. Corros. 1. 19, 89-94, 1984. 202. Blanco. M. T., Andrade, C, and Macias, A.: SEM study of the corrosion products of galvanized reinforcements immersed in solutions in the pH range 12.6 to 13.6, Br. Corras. 1. 19,41-48,1984. 203. Defrancq, J. N.: Zinc and zinc-lead alloys in domestic water, Br. Corros. 1.17, 125-130, 1982. 204. Saleh, R. M., Ismail, A A., and El Hosary, A. A.: Corrosion inhibition by naturally occurring substances, Part VII. The effect of aqueous extracts of some leaves and fruit-peels on the corrosion of steel, AI, Zn and Cu in acids, Br. Corros. 1.17,131-135,1982.
REFERENCES
417
205. Shams EI Din, A. M., AbdEI Kader,J. M., and Kuhn, A. T.: Pitting corrosion of zinc in aqueous solutions. Part III. Effect of anode to cathode surface area ratio on pitting corrosion currents, Br. Corros. 1. 15, 208-210,1980. 206. Muralidharan, V. S., and Rajagopalan, K. S.: Kinetics and mechanism of passivation of zinc in dilute sodium hydroxide solutions, Br. Corros. 1. 14, 231-234, 1979. 207. Rudresh, H. B., and Mayanna, S. M.: n-Decylamine as a corrosion inhibitor for zinc in acidic solution, Br. Corros. 1.12,54-56,1977. 208. Mor, E. D., and Beccaria, A. M.: Inhibitory action of acrylonitrile on the corrosion of zinc in sea water, Br. Corros. 1.8,25-27,1973. 209. Mor, E. D., Beccaria, A. M., and Poggi, G.: Behaviour of zinc in sea water in the presence of sulphides. Br. Corros. 1. (Quarterly) 1974(1). 53-56. 210. Mansfeld, E. and Tsai. S.: Laboratory studies of atmospheric corrosion. Part I. Weight loss and electrochemical measurements, Corros. Sci. 20, 853-872. 1980. 211. Kesternich, W.: Corrosion testing of metallic and non-metallic coatings in an artificial industrial atmosphere, Stahl Eisen 71( II), 1951. 212. Townsend, H. E.: Status of a cooperative effort by the automotive and steel industries to develop a standard accelerated corrosion test, SAE Technical Paper No. 892569. Society of Automative Engineers 1989 Automotive Corrosion & Prevention Conference & Exposition, Dearborn, Michigan, December 4-6, 1989. 213. Lyon, S. B., Thompson, G. E., Johnson, 1. B., Wood, G. c., and Ferguson, 1. M.: Accelerated atmospheric corrosion testing using a cyclic wet/dry exposure test: Aluminum, galvanized steel, and steel, Corrosion 43,719-726, 1987. 214. Lambert, M. R., Townsend, H. E., Hart, R. G., and Frydrych, D. J.: Accelerated corrosion tests of precoated sheet steels for automobiles, Ind. Eng. Chem. Prod. Res. Dev. 24, 378-384. 1985. 215. Abbas, M. I.. Ahmed, M. S., and Shahin, M. A.: Effect of galvannealing heat treatment on the corrosion characteristics of galvanized steel, CIM Bull. 82, 69-73, 1989. 216. Wilde, B. E.: An electrochemical criterion for galvanic protection of steel exposed to marine environments, C1M Bull. 83, /62-164, 1990. 217. Slunder, C. 1., and Boyd, W. K.: Zinc: Its Corrosion Resistance, 2nd ed., International Lead Zinc Research Organization, Inc .• New York, 1986. 218. Morgan. S. W. K.: Zinc and Its Alloys and Compounds, John Wiley & Sons, Toronto, 1985. 219. Ho, M. S.: Atmospheric Corrosion Results of ZM-1I3 Galvanized Samples after Fifteen Years of Exposure, Cominco PTC Project No. 84-5-511, Report No. I. June 4. 1985. 220. Mercille, G. P.: Corrosion Evaluation of Zinc, No.3 Alloy and ZA-12 Alloy Embedded in Concrete for 10 Years, Cominco PTC Project No. 83-5-511. Report No.3, 1987. 221. McBreen. J .• and Cairns, E. J.: The zinc electrode. in Advances in Electrochemistry and Electrochemical Engineering, Vol. II, H. Gerischer and C. W. Tobias (eds.). pp. 273-352. John Wiley & Sons, New York. 1978. 222. Ashton, R. E. and Hepworth, M. T.: Effect of crystal orientation on the anodic polarization and passivity of zinc. Corrosion 24, 50-53, 1968. 223. Goodrich. 1. D .• and Schmid, G. M.: The dissolution of zinc in the NaNO r NH4CI-NH r H 20 system. Corrosion 24. 252-254. 1968. 224. Devillers, L. P.• and Niessen. P.: Accelerated grain boundary corrosIOn of dilute zinc-aluminum alloys under high applied cathodic currents, Corrosion 32, 152-154, 1976. 225. Melton. K. N .. and Edington, J. W.: Effect of increasing aluminum content on the boiling water corrosion resistance of superplastic Zn-Al alloys, Corrosion 31, 146-147, 1975. 226. Weast. R. c., and Shulman, S. L.: Corrosion of zinc in pressurized dilute aqueous solution systems, Corrosion 18, 417-420,1962. 227. Steinbicker, R. N., and Fountoulakis, S. G.: Production of zinc-nickel electoplated coatings, Iron Steel Eng. 1989(7), 28-31. 228. Dargis, R. G.: Zinc-alloy electroplates as functional finishes, Prod. Finish. 1990(10), 43-54. 229. Sharples. T. E.: Zn-Co: Fighting corrosion in the 90's, Prod. Finish. 1990(3),38-44. 230. Waldron, L. J., and Peterson, M. H.: Effect of iron, aluminum and cadmium additions on the performance of zinc anodes in sea water, Corrosion 16, 81-85, 1960.
418
REFERENCES
231. Dirkse, T. P., and Timmer, R.: The corrosion of zinc in KOH solutions, 1. Electmchem. Soc. 116, 162-165,1969. 232. Vorkapic, L. Z., Drazic, D. M., and Despic, A R.: Corrosion of pure and amalgamated zinc In concentrated alkali hydroxide solutions, J. Electrochem. Soc. 121, 1385-1392,1974. 233. Beard, R. B., Carim, H. M., Dubin, S. E., DeRosa, J, E. and Miller. A S.: Corrosion and histopathological studies on anode materials for implantable power sources. J. Electrochem. Soc. 121. 1129-1133. 1974. 234. Gregory, D. P., Jones, P. c., and Redfearn. D. P.: The corrosion of zinc anodes in aqueous alkaline electrolytes, J. Electrochem. Soc. 119, 1288-1292, 1972. 235. Williams, D. N., Koehl, B. G., Berry. W. E., and Bartlett, E. S.: The corrosion behavior of Zn-22AI alloy sheet, J. Electrochem. Soc. 118, 1684-1688, 1971. 236. Townsend, H. E., and Meitzner, C. E: Corrosion resistance of zinc/4% aluminum and zincl7% aluminum alloy coatings compared to zinc and zinc/54% aluminum alloy coatings, Mater. Perform. 22( 1),54, 1983. 237. Irving, R. R.: Galfan the zinc-aluminum coating, Iron Age 1986(Aug. I)' 47. 49. 238. Witmer, D. A: New markets for corrosion-resistant galvalume steel sheets. ASM Technical Paper No. 8201-035, 1982 American Society for Metals Metals Congress. SI. Louis. Missouri. October 23-28. 1982. 239. Lynch, R. E, and Goodwin, E E.: Galfan coated steel for automotive applications. SAE Technical Paper No. 860658, Society of Automotive Engineers International Congress and Exposition, Detroit, February 24-28. 1986. 240. Baumgartner, M., and Kaesche, H.: Intercrystalline corrosion and stress corrosion cracking of AIZnMg alloys, Corrosion 44, 231-239, 1988. 241. Centre de Recherches Metallurgiques (Belgium): Galfan: Its Corrosion Resistance in an Industrial Atmosphere, International Lead Zinc Research Organization Project ZM 285, Progress Report No. 14A July I-Dec. 31,1985. 242. Centre de Recherches Metallurgiques (Belgium): Atmospheric Exposure Testing of Galfan, International Lead Zinc Research Organization Project ZM 285, Progress Report No. 15B, Jan. I-June 30, 1986. 243. Centre de Recherches Metallurgiques (Belgium): Electrochemical Study of Coil Coated Galfan, International Lead Zinc Research Organization Project ZM 285, Progress Report No. 14C. July I-Dec. 31,1985. 244. Hubrecht, 1., Skenazi, A E, Vereecken, J .. and Coutsouradis, D.: Corrosion resistance of Galfan and coil coated Galfan in chloride containing atmospheres. ATB Metall. 25(3), 243-252, 1985. 245. Herrschaft, D. c., Radtke. S. E, Coutsouradis, D., and Pelerin, J.: Galfan-A New Zinc-Alloy Coated Steel for Automotive Body Use. SAE Technical Paper No. 830517, Society of Automotive Engineers International Congress and Exposition, Detroit, February 28-March 4, 1983. 246. Sharples, T. E.: Zinclzinc alloy plating, Prod. Finish. 1988(4),50-56. 247. Loto, C. A: Corrosion evaluation of cast iron, aluminium alloy, galvanized and stainless steels in cassava juice, Corros. Prevo Control 1990(October). 13] -] 38. 248. Townsend, H. E., and Borzillo, A. R.: Twenty-year atmospheric corrosion tests of hot-dip coated sheet steel, National Association of Corrosion Engineers Corrosion' 87 Conference. San Francisco, March '9-13, 1987, Paper No. 421. 249. Vi art, G.: Galvanised steel wire with high corrosion resistance, Wire Ind. 1984( I), 76-78, 83. 250. Allegra, L., Berke, N. S., and Townsend, H. E.: Resistance of galvanized, aluminum-coated, and 55% AI-Zn-coated steel sheet to atmospheric corrosion involving standing water, in Atmospheric Corrosion, W. H. Ailor (ed.), pp. 595-606, John Wiley & Sons, New York, 1982. 251. Dreulle, N., Frappa, A, Limare, A, Piessen, P., and Sokolowski, R.: Effect of aluminum on the protection of steel by zinc against atmospheric corrosion, Mater. Tech. 1985(Feb.-Mar.), 77-83. 252. Nuenninghoff, R., and Fischer, H.: Corrosion behaviour of zinc-aluminum alloys on steel wire, Stahl Eisen 105(9), 517-522,1985. 253. Leroy, V., and Schmitz, B.: Surface chemistry of Zn and ZnlAI hot-dipped steels: Influence of some processing parameters, Scand. J. Metall. 17, 17-23, 1988. 254. Matthews, P.: Impact of Galvalume on zinc consumption, Cominco Ltd., Metals Market Research Report, October 1981. 255. Townsend, H. E., and Zoccola, J. c.: A comparison of the corrosion resistances of skyward- and groundward-exposed surfaces of 55% aluminum-zinc alloy-coated and zinc-coated steel sheet, Pro-
REFERENCES
256.
257. 258.
259. 260.
261.
262. 263.
264.
265.
266.
267.
268.
269.
270. 271.
272. 273.
419
ceedings of the International Congress on Metallic Corrosion, Toronto, Ontario, Vol. 3, pp. 216-219, 1984. Dreulle, P, Rosset, R., and Jardy, A.: Corrorsion properties of steel sheet coated with Zn-AI alloys, Proceedings of the 12th International Galvanizing Conference, INTERGALVA '79, Paris, France, May 17-23, 1979. Campbell, H. S.: Zinc-aluminium coatings for hot water tanks: Summary report. The British NonFerrous Metals Research Association, Report No, A, 1382, March 1962. Pourbaix, M., Van Muylder, J., Pourbaix, A., and Kissel, J.: An electrochemical wet and dry method for atmospheric corrosion testing, in Almospheric Corrosion, W. H. Ailor (ed.), pp. 167-177, John Wiley & Sons, New York, 1982. American Society for Testing and Materials: Standard test method of salt spray (fog) tcsting, in Annllal Book ofASTM Standards, Vol. 02.05, No. BI17-90, pp. 19-25, 1990. Warnecke, W., Bode, R., Kothe, R., and Meyer, L.: Modern hot dip coated sheet steel-processing, coating characteristics and fabricating properties, in Zinc-Based Sleet Coaling Srslems: Mela/lurgy and Per/(Jrlnance, G. Krauss and D. K. Matlock (cds.), pp. 3-17, The Minerals, Metals and Materials Society, Warrendale, Pennsylvania, 1990. Marder, A. R.: Microstructural characterization of zinc coatings, in Zinc-Based Sleel Coaling :Svslems: Metallurgv and Performance, G. Krauss and D. K. Matlock (eds.), pp. 55-82, The Minerals. Metals and Materials Society. Warrendale, Pennsylvania, 1990. Mackowiak, J., and Short. N. R.: Metallurgy of galvanized coatings, Int. Mel. Rev. No. I, 1979, Review No. 237. Takahashi, A., Miyoshi, Y, and Hada, T.: A study on microstructure ofZn-Ni-Si0 2 composite platings by transmission electron microscope, in Zinc-Based Sleel Coating Systems: Meta/lll~"r and Per/rlr/nance, G. Krauss and D. K. Matlock (eds.), pp. 83-94, The Minerals. Metals and Materials Society. WaITendale. Pennsylvania, 1990. Chen, Y L., and Snyder, D. D.: Microstructural investigation of electroplated Zn-Ni coatings on cold-rolled steel, in Zinc-Based Steel Coating Systems: Melallurgyand Performance, G. Krauss and D. K. Matlock (cds.). pp. 95-107, The Minerals, Metals and Materials Society, Warrendale, Pennsylvania, 1990. Lin, Y, Pak, S., and Meshii, M.: Investigation of pure zinc electrogalvanized coatings by electron microscopy, in Zinc-Based Steel Coating SYstems: Metallurgy and Per/rlrlnance, G. Krauss and D. K. Matlock (cds.), pp. 109-119, The Minerals. Metals and Materials Society, Warrendale, Pennsylvania, 1990. Shaffer, S. 1., Morris, 1. W., and Wenk, H. R.: Textural characterization and its applications on zinc electrogalvanized steels, in Zinc-Based Sleel Coating Systems: Metallurgy and Peljormance, G. Krauss and D. K. Matlock (eds.), pp. 129-140, The Minerals, Metals and Materials Society, Warrendale, Pennsyl vania, 1990. Apostolos. J. A.: Cathodic protection of reinforced concrete using /lame sprayed zinc, National Association of Corrosi on Engineers Corrosion' 83 Conference, Anaheim, California, April I R-22, 19R3. Paper No. 180. Cook, A. R., and Radtke, S. E: Recent research on galvanized steel for reinforcement of concretc, in Chloride Corrosion of'Steel in Concrete, STP 629, pp. 51-60, American Society for Testing and Materials. Philadelphia, 1977. Kilareski, W. P.: Epoxy coatings for corrosion protection of reinforcement steel, in Chloride Corrosion of Sleet in Concl'Cle, STP 629, pp. 82-88, American Society for Testing and Materials, Philadelphia, 1977. Vrable, 1. B.: Cathodic protection for reinforcing steel in concrete, in Chloride Corrosion of Steel in Concrete, STP 629, pp. 124-149, American Society for Testing and Materials, Philadelphia, 1977. Treadaway, K. W. J., Brown. B. L., and Cox, R. N.: Durability of galvanized steel in concrete. in Corrosion olReill/orcillg Sleel in Concrele, STP 713, pp. 102-131, American Society for Testing and Materials, Philadelphia, 1980. Biczok, I.: COllclde Corrosion and Concrete Prolection, Chemical Publishing Co., New York, 1967. Stark, D.: Measurement techniques and evaluation of galvanized reinforcing steel in concrete structures in Bermuda, in Corrosion ofReil!/orcing Steel in Concrete, STP 713, pp. 132-141, American Society for Testing and Materials, Philadelphia, 1980.
420
REFERENCES
274. Cornet, I., and Bresler, B.: Critique of testing procedures related to measuring the performance of galvanized steel reinforcement in concrete, in Corrosion ofReinforcing Steel in Concrete, STP 713, pp. 160-195, American Society for Testing and Materials, Philadelphia, 1980. 275. Corderoy, D. 1. H., and Herzog, H.: Passivation of galvanized reinforcement by inhibitor anions. in Corrosion of Reinforcing Steel in Concrete, STP 713, pp. 142-159, American Society for Testing and Materials, Philadelphia, 1980. 276. Wyatt, B. S., and Irvine. D. 1.: A review of cathodic protection of reinforced concrete, Mater. Perform. 26(12),12-21, 1987. 277. Mudd, C. 1., Mussinelli, G. L., Tettamanti, M., and Pedeferri, P.: Cathodic protection steel in concrete, Mater. Perform. 27(9), 18-24, 1988. 278. Feliu, S., Gonzalez, 1. A., Feliu, S., Jr., and Andrade, Relationship between conductivity of concrete and corrosion of reinforcing bars, Br. Corros. 1. 24, 196-198, 1989. Developments in the continuous galvanizing of steel, 1. Metals 41(8),34-36, 1989. 279. Bush, G. 280. Gustafson, D. P.: Steel reinforcement: Purpose, types, and uses, ASTM Standardization News 1990(12), 38-42. 281. Bogner, B.: Review of internal corrosion of underground fuel storage tanks, Mater. Perform. 29(10), 30-32, 1990. 282. Irie, T.: Development of zinc-based coatings for automotive sheet steel in Japan, in Zinc-Based Steel Coating Systems: Metallurgy and Performance, G. Krauss and D. K. Matlock (eds.), pp. 143-155, The Minerals, Metals and Materials Society, Warrendale, Pennsylvania, 1990. 283. Lee, H. H.: The corrosion resistance of electrodeposited Zn-Ni alloy coatings, in Zinc-Based Steel Coating Systems: Metallurgy and Performance, G. Krauss and D. K. Matlock (eds.), pp. 157-169, The Minerals, Metals and Materials Society, Warrendale, Pennsylvania, 1990. 284. Abe, M., Shiohara, Y., and Okado, A.: Coating characteristics and performance of electrodeposited Zn-Si02 composite coating, in Zinc-Based Steel Coating Systems: Metallurgy and Performance, G. Krauss and D. K. Matlock (eds.), pp. 171-181, The Minerals, Metals and Materials Society, Warrendale, Pennsylvania, 1990. 285. Goodwin, F. E.: Mechanisms of corrosion of zinc and zinc-5% aluminum steel sheet coatings, in Zinc-Based Steel Coating Systems: Metallurgy and Performance, G. Krauss and D. K. Matlock (eds.), pp. 183-193, The Minerals, Metals and Materials Society, Warrendale, Pennsylvania. 1990. 286. Jordan, D. L.: Influence of iron corrosion products on the underfilm corrosion of painted steel and galvanized steel, in Zinc-Based Steel Coating Systems: Metallurgy and Performance, G. Krauss and D. K. Matlock (eds.), pp. 195-205, The Minerals, Metals and Materials Society, Warrendale, Pennsyl vania, 1990. 287. Mabuchi, M., Takao, K., Ogishi, H., Kimura, H., and Ichida, T.: Effect of organic composite coating on corrosion resistance ofZn-Ni alloy coated sheet steel, in Zinc-Based Steel Coating Systems: Metallurgy and Performance, G. Krauss and D. K. Matlock (eds.), pp. 207-218, The Minerals, Metals and Materials Society, Warrendale, Pennsylvania, 1990. 288. Claus, G., Houbaert, Y., Dilewijns, 1., Meers, U., Vanthournout, M., and Verhoeven, G.: Behaviour of prephosphated Zn-coated steel sheet during and after final phosphating, in Zinc-Based Steel Coating Systems: Metallurgy and Performance, G. Krauss and D. K. Matlock (eds.), pp. 219-230, The Minerals, Metals and Materials Society, Warrendale, Pennsylvania, 1990. 289. Yau, Y., and Fountoulakis, S. G.: Characterization of the coating-to-steel interface strength in ZnNi coated sheet steel, in Zinc-Based Steel Coating Systems: Metallurgy and Performance, G. Krauss and D. K. Matlock (eds.), pp. 371-386, The Minerals. Metals and Materials Society, Warrendale, Pennsylvania, 1990. 290. Yeomans, S. R., and Novak, M. P.: Further studies of the comparative properties and behaviour of galvanized and epoxy coated steel reinforcement, International Lead Zinc Research Organization Project ZE-341, Progress Report No.4, Research Triangle Park, North Carolina, July 1990. 291. Fiaud, C.: Electrochemical behavior of atmospheric pollutants in thin liquid layers related to atmospheric corrosion, in Atmospheric Corrosion, W. H. Ailor (ed.), pp. 161-166, John Wiley & Sons, New York,1982. 292. Warwick, M. E., and Hampshire, W. B.: Atmospheric corrosion of tin and tin alloys, in Atmospheric Corrosion, W. H. Ailor (ed.), pp. 509-527, John Wiley & Sons, New York, 1982.
c.:
w.:
REFERENCES
421
293. Kucera. Y. and Mattsson, E.: Atmospheric corrosion of bimetallic structures. in Atmospheric Corrosion, W. H. Ailor (ed.), pp. 561-574. John Wiley & Sons, New York. 1982. 294. Dunbar. S. R., and Showak. W.: Atmospheric corrosion of zinc and its alloys, in Atlllospheric Corrosion, W. H. Ailor (ed.), pp. 529-552, John Wiley & Sons, New York, 1982. 295. Legault, R. A.: Atmospheric corrosion of galvanized steel, in Atmospheric Corrosion, W. H. Ailor (ed.), pp. 607-613, John Wiley & Sons, New York, 1982. 296. Leidheiser, H., and Suzuki, I.: Cobalt and nickel cations as corrosion inhibitors for galvanized steel. in Atmospheric Corrosion, W. H. Ailor (ed.), pp. 615-629, John Wiley & Sons, New York, 1982. 297. Biestek, T.: Atmospheric corrosion testing of electrodeposited zinc and cadmium coatings, in Atmospheric Corrosion, W. H. Ailor (cd.), pp. 63 I -643, John Wiley & Sons, New York, 1982. 298. Altorfer. K. J.: Service evaluation of zinc thermal spraying systems, in Atmospheric Corrosion. W. H. Ailor (cd.), pp. 645-650, John Wiley & Sons, New York, 1982. 299. Talati, J. D., and Patel. B. M.: Atmospheric corrosion of metals under moving conditions, in Atmospheric Corrosion, W. H. Ailor (ed.), pp. 695-703, John Wiley & Sons, New York, 1982. 300. Moresby. J. E, Reeves, E M .. and Spedding, D. J.: Atmospheric corrosion testing in Australasia. in Atmospheric Corrosioll, W. H. Ailor (ed.). pp. 745-754, John Wiley & Sons, New York, 1982. 30 I. Dutra, A. C, and Vianna, R. de 0.: Atmospheric corrosion testing in Brazil, in Atlllospheric Corrosion, W. H. Ailor (ed.). pp. 755-774. John Wiley & Sons, New York, 1982. 302. Biestck, T: Testing clectrodeposited coatings in tropical China, in Atmospheric Corrosion, W. H. Ailor (ed.), pp. 775-785, John Wiley & Sons. New York, 1982. 303. Hakkarainen, T, and Ylasaari. S.: Atmospheric corrosion testing in Finland, in Atlllospheric Corrosion, W. H. Ailor (ed.), pp. 787-795, John Wiley & Sons, New York, 1982. 304. Atteraas. L., and Haagenrud, S.: Atmospheric corrosion in Norway. in Atmwpheric Corrosion. W. H. Ailor (cd.), pp. 873-891. John Wiley & Sons, New York, 1982. 305. Callaghan, B. G.: Atmospheric corrosion testing in southern Africa, At11lospheric Corrosion, W. H. Ailor (ed.). pp. 893-912. John Wiley & Sons, New York, 1982. 306. Feliu, S .. and Morcillo. M.: Atmospheric corrosion testing in Spain, in Atmospheric Corrosion. W. H. Ailor (cd.), pp. 9 I 3-92 I, John Wiley & Sons. New York, 1982. 307. Mikhailovski, Y. N., and Strekalov. P Y: Atmospheric corrosion tests in the USSR, in Atmospheric Corrosion, W. H. Ailor (cd.). pp. 923-942. John Wiley & Sons. New York. 1982. 308. Knotkova, D., Barton. K., and Cerny, M.: Atmospheric corrosion testing in Czechoslovakia, in Atmospheric Corrosion. W. H. Ailor (cd.), pp. 991- 1014, John Wiley & Sons. New York. 1982. 309. Tomlinson, W. 1., and Marsh, P. 1.: Corrosion of iron, steel, zinc, nickel andelectroless nickel in concrete. J. Matn: Sci. Lett. 7. 808-810. 1988. 3 I O. Apostolos. J. A., Parks, D. M .. and CareJlo. R. A.: Cathodic protection of reinforced concrete using metallized zinc. Mater. Perform. 26(2), 22-28,1987. 3 I I. Catino. J. W.: The polarization and corrosion characterization of Zn and modified Zn in concentrated alkali metal hydroxide electrolyte: The effect of a Zn oxide/hydroxide surface, Ph. D. Thesis, Lehigh University, 1988. 3 I 2. Suzuki. I.: Corrosion-Resistant Coatings Technology, Corrosion Technology Series, PA. Schweitzer (ed.), Marcel Dekker. New York, 1989. 3 I 3. Serra, E. T, and Mannheimer, W. A.: On the estimation of the corrosion rates of metals in soils by electrochemical measurements, in Underground Corrosion, STP 741, pp. 1 I 1-122, American Society for Testing and Materials. Philadelphia, 198 I. 314. Miller. F. P., Foss. J. E., and Wolf, D. C: Soil surveys: Their synthesis, confidence limits, and utilization for corrosion assessment of soil, in Underground Corrosion, STP 74 I, pp. 3-23, American Society for Testing and Materials, Philadelphia, 198 I. 3 I 5. Guttman, H.: Effects of atmospheric factors on the corrosion of rolled zinc, in Metal Corrosion in the Atmosphere, STP 435, pp. 223-239, American Society for Testing and Materials, Philadelphia, 1968. 316. Dunbar. S. R.: Effect of one per cent copper addition on the atmospheric corrosion of rolled zinc, in Metal Corrosion in the Atmosphere, STP435. pp. 308-325, American Society for Testing and Materials, Philadelphia, 1968. 317. Baboian, R. (cd.): Automotive Corrosion by Deicing Salts. National Association of Corrosion Engineers, Houston, 1981. 3 I 8. Geduld. H.: Zinc Plming, ASM International. Metals Park, Ohio, 1988.
422
REFERENCES
319. International Lead Zinc Research Organization: A Comparison between Hot Dip Galvanized and Fusion Bonded Epoxy Coatings for Steel Reinforcement in Concrete. ILZRO Project No. ZE-341. Annual Report, Research Triangle Park, North Carolina, 1988. 320. Tomashov, N. D., and Chernova, G. P.: Passivity and Protection of Metals against Corrosion. Plenum Press, New York, 1967. 321. Waber,1. T.: Analysis of size effects in corrosion processes, in Localized Corrosion, pp. 221-237, National Association of Corrosion Engineers, Houston, 1974. 322. Chen, Z. W, Sharp, R M., and Gregory, 1. T.: Fe-AI-Zn ternary phase diagram at 450 C, Mater. Sci. Techno!. 6, 1173-1176, 1990. 323. Albalat, R, Gomez, E., Muller, J., Pregonas, J .• Sarret, M., and Valles, E.: Zinc-nickel coatings: Relationship between additives and deposit properties, 1. Appl. Electrochem. 21, 44-49, 1991. 324. Djoufac Woumfo, E., and Vittori, 0.: Electrochemical behaviour of a zinc electrode in 8 M KOH under pulsed potential loading, 1. Appl. Electrochem. 21,77-83, 1991. 325. Beccaria, A. M.: Zinc layer characterization on galvanized steel by chemical methods, Corrosion 46, 906-912, 1990. Velinov. v.. Riesenkampf. A., and Pawlik, B.: Factors influencing the 326. Tomov, I., Cvetkova, preferential orientations in zinc coatings electrodeposited from chloride baths. 1. Appl. Electrochem. 19, 377-382, 1989. 327. Faita, G., Mazza, E, and Bianchi, G.: Role of water and ionic solvation in localized corrosion phenomena, in Localized Corrosion, pp. 34-44, National Association of Corrosion Engineers, Houston, 1974. 328. Newman, J.: Mass transport and potential distribution in the geometries of localized corrosion, in Localized Corrosion, pp. 45-61, National Association of Corrosion Engineers, Houston, 1974. 329. Fujii, C. T.: A review: Macro-cells for the study of solution chemistry and electrochemistry in localized corrosion, in Localized Corrosion, pp. 144-150, National Association of Corrosion Engineers, Houston, 1974. 330. Selvam, M., and Guruviah, S.: Corrosion of electrodeposited zinc-manganese alloys, Bull. Electrochem. 6(5),485-486. 1990. 331. Graedel, T. E.: Corrosion mechanisms for zinc exposed to the atmosphere. 1. Electrochem. Soc. 136, 193-203,1989. 332. Townsend, H. E.: On the effects of magnesium on the atmospheric corrosion resistance of galvanized sheet steel, Corrosion 44. 229-230, 1988. 333. Evans, U. R: Factors deciding between corrosion and protective film formation. The 1st International Congress on Metallic Corrosion. London, April 1961, pp. 3-9. 334. Lorking, K E. and Mayne, J. E. 0.: Some observations on the corrosion of zinc. The 1st International Congress on Metallic Corrosion, London, April 1961, pp. 144-146. 335. Cismaru, D., and Cismaru, G. D.: Zinc oxidation at high temperatures. Effect of electric field on zinc oxidation, The 1st International Congress on Metallic Corrosion, London, April 1961. pp. 237-239. 336. Rosenfeld, I. L.: Atmospheric corrosion of metals. Some questions of theory, The 1st International Congress on Metallic Corrosion, London, April 1961, pp. 243-253. 337. Neville, R J., and de Souza, K M.: Electrogalvanized or hot dip galvanized-results of five years undervehicle corrosion testing, SAE Technical Paper No. 862010, Society of Automotive Engineers Automotive Corrosion and Prevention Conference, Dearborn, Michigan, December 8- 10, 1986. 338. Stephens, M. L., Davidson, D. D., Soreide, L. E., and Shaffer, R 1.: Developing the chipping corrosion test, SAE Technical Paper No. 862013, Society of Automotive Engineers Automotive Corrosion and Prevention Conference, Dearborn, Michigan, December 8- 10, 1986. 339. Satoh, H., Shimogori, K, Nishimoto, H., Miki, K, Ikeda, K, Iwai, M., Sakai, H., and Nomura, S.: Analysis of perforation corrosion of cold-rolled and galvanized steel sheets by extreme-value statistics, SAE Technical Paper No. Technical Paper No. 862018, Society of Automotive Engineers Automotive Corrosion and Prevention Conference, Dearborn, Michigan, 8-10, 1986. 340. Memmi, M., Ferrari, v., Arrigoni, G., Sarracino, M., Moelzer, H., and Ulivieri, L.: A qualitative and quantitative evaluation of Zn + Cr-CrOx multilayer coating compared to other coated steel sheets, SAE Technical Paper No. 862028, Society of Automotive Engineers Automotive Corrosion and Prevention Conference. Dearborn, Michigan, December 8-10, 1986.
c..
REFERENCES
423
341. Mohri, T, Tsugawa, S., Kobayashi, S .. Ichida, T, and Kurosawa, M.: Newly developed organic composite-coated steel sheet with bake-hardenability, SAE Technical Paper No. 862030. Society of Automotive Engineers Automotive Conosion and Prevention Conference, Dearborn. Michigan December 8-10, 1986. 342. Tator, K. B.: How coatings protect and why they fail, Mater. Perfimn. 17(11),41 -44, 1978. 343. Leidheiser, H.: Conosion of painted metals-a review. Corrosion 38, 374-383. 1982. 344. Barrett, L. D.: Preparation of metals by phosphating, Mater. Prot. 3(7),20-23,1964. 345. Munger, C. G.: Inorganic zinc coatings past, present, and future, Mater. PetjiJrm. 14(5),25-29, 1975. 346. Robbins, D. 1., Smith, D. M., and Roberts, T R.: Corrosion testing and behavior of zinc and zinc alloy coated autobody sheet steels. ASM Corrosion-Resistant Automotive Sheet Steels Conference, Chicago, September 24-30, 1988, pp. 1-14. 347. Lin, K. L., Yang, C. E, and Lee, J. T: Correlation of microstructure with eonosion and electrochemical behavior of the batch-type hot-dip AI-Zn coatings. Part I. Zn and 5% AI-Zn coatings, Corrosion 47, 9-16,1991. 348. Lin, K. L., Yang, C. E, and Lee, J. T.: OllTelation of microstructure with conosion and electrochemical behaviors of the batch-type hot-dip AI-Zn coatings. Part II. 55% AI-Zn coating, Corrosion 47, 17-23, 1991. 349. Watanabe, T, Kubota, T, Yamashita, M .. Urakawa, T, and Sagiyama, M.: Con'osion-resistant precoated steel sheets for automotive body panels, ASM Conosion-Resistant Automotive Sheet Steels Conference, Chicago, September 24-30, 1988, pp. 15-20. 350. Neville, R. L and de Souza, K. M.: Corrosion performance of zinc and zinc alloy coated steels exposed to deicing salt spray under vehicles, ASM Conosion-Resistant Automotive Sheet Steels Conference, Chicago, September 24-30, 1988, pp. 2 I -30. 351. Yasuda, A., Umino, S., Kyono, K., and Yamato, K.: Cosmetic conosion ofZn alloy coated sheet steels, ASM Conosion-Resistant Automotive Sheet Steels Conference, Chicago, September 24-30,1988, pp. 31-37. 352. Yasuda, A., Koumura, H., Yamato, K., Onizawa, K., and Ota, H.: Development of high performance galvannealed steel sheet for outer panels of automotives, ASM Conosion-Resistam Automotive Sheet Steels Conference, Chicago. September 24-30, 1988, pp. 45-53. 353. Lynch, R. E: Hot-dip galvanizing alloys, 1. Met. 39(8), 39-41. 1987. 354. Scholl, p, and Prentice, G.: Photocurrcnt response of the anodic film on zinc electrodes in alkaline elcctrolyte, Electrochemical Society Extended Abstracts, Vol. 91-1, Abstract No. 75, p. I J 3. 1991. 355. Davies, D. E., and Lotlikar, M. M.: Passivation and pilling characteristics of zinc, Br. Corros. 1. L 149-155, 1966. 356. Mansfeld, E. and Kenkel, J. V: Galvanic conosion of Al alloys. Part III. The effect of area ratio, Corm.l. Sci. 15, 239-250, 1975. 357. Romanoff, M.: Underground Conosion, U.S. National Bureau of Standards Circular 579. 1957. 358. Rachlot, B.: Sacrificial anode aJloys-Their specifications and uses, Proceedings of the 2nd International Congress on Metallic Conosion, New York, March 11-15, 1963, pp. 285-289. 359. Bianchi, G., Mazza, E, and Trasatti, S.: Mechanism offormation of black oxide films on zinc, aluminum and titanium in alkaline aqueous solutions, Proceedings of the 2nd International Congress on Metallic Conosion, New York, March 11-15, 1963, pp. 905-915 360. Chou, K. c., and Jagannathan, V: Corrosion and powdering resistance of galvanneal for automotive applications, National Association of Corrosion Engineers Corrosion '91 Conference, Cincinnati, Ohio, March II-IS, 1991,PaperNo.375. 361. Mori, K., Miyawaki, T' and Matsushima, Y.: Effect of zinc composite coating on conosion resistance at high temperature, National Association of Con os ion Engineers Conosion '91 Conference, Cincinnati, Ohio, March I 1-15, 1991, Paper No. 40 I. 362. Wang, Y. M., and Radovic, D.: Porosity measuremcnts of phosphate coatings by electrochemical impedance spectroscopy (EIS) method, National Association of Corrosion Engineers Corrosion '91 Conference, Cincinnati, Ohio, March I J -15, 1991, Paper No. 419. 363. Granata, R. D.: Mechanisms of coatings disbondment, National Association of Conosion Engineers Corrosion '91 Conference, Cincinnati, Ohio, March 11-15, 1991, Paper No. 382.
424
REFERENCES
364. Massinon, D., and Thierry, D.: Rate controlling factors in the cosmetic corrosion of coated steels, National Association of Corrosion Engineers Corrosion '91 Conference, Cincinnati, Ohio, March 11-15,1991, Paper No. 574. 365. Chang, J. c., and Wei, H. H.: Electrochemical and Mossbauer studies of the corrosion behavior of electrodeposited fe-Zn alloys on steel, Corros. Sci. 30, 831-837,1990. 366. Wippermann, K, Schultze, 1. W, Kessel, R., and Penninger, 1.: The inhibition of zinc corrosion by bisaminotriazole and other triazole derivatives, Corros. Sci. 32, 205-230, 1991. 367. Kirihara, S., and Ohshima, K (Kobe Steel, Ltd., Japan): Method for anticorrosive treatment of galvanized steel, U.S. Patent 4385940, May 31, 1983. 368. Sriveeraraghaven, S., Krishnan, R. M., and Natarajan, S. R.: Corrosion behavior of zinc and cadmium deposits obtained from different baths, Met. Finish. 89(8),51-53, 1991. 369. Cremaschi, P., Tantardini, G. E, Muilu, 1., and Pakkanen, T. A.: Adsorption of hydrogen on zinc-calculation of potential energy surface and dynamics, Vacuum 41( 1-3), 260-264, 1990. 370. Girin, O. B., and Panasenko, S. A.: Effect of the texture of electrolytic zinc coatings on their corrosion resistance, Zashch. Met. 25(3),480-482, 1989 [in English]. 371. Novoa, X. R., Izquierdo, M., Merino, P., and Espada, L.: Electrochemical impedance spectroscopy and zero resistance ammeters (ZRA) as tools for studying the behaviour of zinc-rich inorganic coatings, Mater. Sci. Forum 44-45, 223-234, 1989. 372. Fiaud, c., Vastra, I., and Tribollet, B.: Application of the electrohydrodynamical impedance method to the study of corrosion inhibitors, Mater. Sci. Forum 44-45, 205-212, 1989. 373. Juttner, K., and Lorenz, W. J.: Electrochemical impedance spectropscopy (EIS) of corrosion processes on inhomogeneous surfaces, Mater. Sci. Forum 44-45, 191-204, 1989. 374. Guillaume, R., Keddam, M., and Takenouti, H.: In-situ electrochemical investigations of a scratched galvanized steel during a climatic test, Mater. Sci. Forum 44-45, 109-122, 1989. 375. Lay, D. E., and Eckles, W E.: The fundamentals of zinc/cobalt, Plating Suif. Finish. 77(10), 10-14, 1990. 376. Sjoukes, E: Chemical reactions in fluxes for hot dip galvanizing, Anti-Corrosion 1990(April), 12-14. 377. Gu, M., and Marder, A. R.: A new method to fabricate single-phase Zn-Fe coating, Plating Suif. Finish. 78(1),77-79,1991. 378. Arup, H.: Galvanized steel in concrete, Mater. Perform. 18(4),41-44,1979. 379. Hofsoy, A., and Gukild, I.: Bond studies on hot dip galvanized reinforcement in concrete, Amer. Conc. Inst. 1. 66(3), 174-184, 1969. 380. Cornet, I., and Bresler, 8.: Corrosion of steel in prestressed concrete, Mater: Prot. 4(11), 35-37, 1965. 381. O'Donnell, A. 1.: Comparison of the sacrificial protection afforded by continuous galvanised coatings, Corras. Coatings S. Afr. 16(10),2,4-6, 1990. 382. Imagawa, H., Matuno, K, and Konishi, T.: Zinc spraying and corrosion cracking prevention for liquid ammonia tanks, Corros. Eng. 38, 353-361, 1989. 383. Mourad, M. Y., Kandell, A. Y., and Elgaber, A. S.: Effect of surface preparation on the corrosion behaviour of zinc in sodium sulphate solution, Indian 1. Technol. 28, 612-614, 1990. 384. Akhter, R., Watkins, K G., and Steen, W. M.: Electrochemical characterisation of laser welded zinc coated steel, Mater: Lett. 9, 550-556, 1990. 385. Kothari, N. C.: Corrosion ofzinc coated steel in simulated industrial environment, Corrosion-Air, Sea, Soil Conference, Auckland, New Zealand, Nov. 19-23, 1990, Paper No. 25. 386. Singh, W. R., and Singh, R. N.: Corrosion of Zn in propylene carbonate: Role of water, H+ and O2 , 1. Electrochem. Soc. India 39, 104-106, 1990. 387. Helwig, L. E., and Swan, 1. D.: Use of spot-test reagents for detecting chromates on hot-dip galvanized surfaces, Met. Finish. 80(9), 37-40, 1982. 388. Yeomans, S. R.: Comparative studies of galvanized and epoxy coated steel reinforcement in concrete, The 2nd CANMET/ACI International Conference on Durability of Concrete, Montreal, Quebec, August 1991. 389. Ravindran, V., and Muralidharan, V. S.: Corrosion of zinc in NH4CI solution, Bull. Electrochem. 6(2), 237-238, 1990. 390. Wilde, B. E., and Budinski, M. K (Ohio State University): Galvanic protection of steel with zinc alloys, U.S.Patent4917966,AprilI7,1990.
REFERENCES
425
391. Verberne, W. M. 1. C. (OMI International Corporation, Michigan): Passivation. U.S. Patent 4776898, October II, 1988. 392. Kressler, R 1., Powers, R G.:and Lasa, I. R: Cathodic protection using scrap and recycled materials, Mater. Pelform. 30(6),29-31, 1991. 393. Cheng, T P, Lee, 1. T, Lin. K. L., and Tsai, W. T: Electrochemical behavior of galvanized AI-Zn coatings in saturated Ca(OHh solution, Corrosion 47. 436-442, 1991. 394. Sjoukes, F: Brown staining of hot dip galvanised coatings in atmospheric cOlTosion, Br. Corros. l. 26, 103-107,1991. 395. Allen, Corrosion of galvanised steel in synthetic mine waters, Br. Corros. l. 26, 93-101, 1991. 396. Smith, C. 1. E.: Research into the control of corrosion on aircrafts, Aircr. EnK. 1990(August), 15. 397. Bird, C. E.: Bond of galvanized steel reinforcement in concrete, Nature (London) 194, 798. 1962. 398. Sebisty, 1. 1.: Continuous-strip galvanized coatings at elevated temperatures. Electrochem. Technol. 6. 330-336,1968. 399. Cook, T.: Myths and truths in hot dip galvanizing, Met. Finish. 89(6),107-108,1991. 400. Kenworthy, L.. and Smith, M. D.: Corrosion of galvanized coatings and zinc by waters containing frcc carbon dioxide. l. Illst. Mel. 1944,463-489. 401. Evans, U. R.: Pitting and cracking, Chem. Ind. 1956, 1291-1297. 402. Evans, U. R .. and Davies, D. E.: The pitting of zinc by distilled water and dilute solutions. Corrosio/l 8(5),165-170,1952. 403. Thornhill, R S., and Evans, U. R.: The electrochemistry of the corrosion of partly immersed zinc, 1. Chem. Soc. 1938,2109-2114. 404. Feitknecht, W.: Studies on the influence of chemical factors on the corrosion of metals, Chem.lnd. 1959, 1102-1109. 405. Davydov, A. D., Zhukova, T B., and Engel'gardt, G. R.: Concentration changes in the solution layer next to the anode, and the maximum currents of electrochemical zinc dissolution in alkalies, Eleklrokhimiva 26, 1207-1209, 1990 [in Englishl. 406. Munier, G. B .. Psota, L. A., Reagor, B. T.. Russiello, B., and Sinclair, 1. D.: Contamination of electronic equipment after an extended urban exposure, 1. Electrochem. Soc. 127,265-2 fL, 1980. 407. Pushpavanam. M., Natarajan. S. R, Balakrishnan, K., and Sharma, L. R.: Corrosion behaviour of electrodeposited zinc-nickel alloys, l. Appl. Electrochem. 21,642-645,1991. 408. Townsend, H. E., Allegra, L., Dutton. R. J., and Kriner, S. A.: Hot-dip coated sheet steels-a review. Maler. PerjcJI"In. 25(8), 36-46, 1986. 409. Hoxeng, R B.: Electrochemical behavior of zinc and steel in aqueous media. Part II, Cormsion 6(9). 308-312,1950. 410. Hoxeng, R. B., and Prutton, C. F: Electrochemical behavior of zinc and steel in aqueous media, Corrosion 5(10), 330-338,1949. 411. Kobayashi, N., Yamasaki, Y., and Inakazu, N.: Effects of cold rolling on corrosion rate of zinc in acid solution, 1. lpn.lnst. Mel. 53,1255-1262,1989. 412. Cox, G. L.: Effect of temperature on the cotTosion of zinc, Ind. Eng. Chem. 23,902-904. 1931. 4\3. Trabanelli. G.: Inhibitors-An old remedy for a new challenge, Corrosion 47, 410-419. 1991. 414. Sizelove. R R: Developments in alkaline zinc-nickel alloy plating, PlatinK SUli Finish. 78(3),26-30. 1991. 41S. Munn. R S., and Devereux. O. F: Numerical modeling and solution of galvanic corrosion systems. Part II. Finite-element formulation and descriptive examples, Corrosion 47, 61 ~-634. 1991. 416. Hinton, B. R. W.: Corrosion prevention and chromates: the end of an era?, Mel. Finish. 89( 10). IS-20. 1991. 417. Yoshioka, K .. Watanabe, K., and Kato. K. (Kawasaki Steel Corp., Japan): Zn-coated stainless steel welded pipe. U.S. Patent 4885215. December 5, 1989. 418. Rocha, L. A .. and Barbosa, M. A.: Microstructure, growth kinetics, and corrosion resistance of hot-dip galvanize\! Zn-S% AI coatings, Corrosion 47, 536-541, 1991. 419. Dean, S. W.: Corrosion testing of metals under natural atmospheric conditions, in Corrosion Testing and Evaluation: Silver Anniversary Volume. STP 1000, pp. 163-176, American Society for Testing and Materials, Philadelphia, 1990. 420. Hinton, B. R. W.: Corrosion prevention and chromates. The end of an era?, Mel. Finish. 89(9),55-61. 1991.
c.:
426
REFERENCES
421. Richards, C. E.: The atmospheric corrosion of galvanised iron wire and its bearing on specification testing, J.lron Steellnst. 137, l27p-176p, 1938. 422. Scholl, P., Shan, X., Bonham, D., and Prentice, G. A.: Photoelectrochemical characterization of the anodic film on zinc in KOH solution, J. Electrochem. Soc. 138,895-899, 1991. 423. Burleigh, T. D.: Anodic photocurrents and corrosion currents on passive and active-passive metals, Corrosion 45, 464-472, 1989. 424. Heitz, E.: Corrosion of metals in organic solvents, in Advances in Corrosion Science and Technology, Vol. 4, M. G. Fontana and R W. Stachle (eds.), pp. 149-243, Plenum, New York, 1977. 425. Adaniya, T.: The development of corrosion-resistant composite zinc-coated steel sheet, Sheet Met. Ind. Int. 1978(December) 73-79. 426. Foster, G. R, Lee, S. G., and Myers, 1. R: Stress corrosion of eta-phase, hot-dip-zinc alloy in potable-hot water, Corrosion 30, 239-241,1974. 427. Brossa, A., Maja, M., Penazzi, N., Farnia, G., Sandona, G., and Marcuzzi, E: Inhibition of zinc corrosion in chloride media: Green piles, Proceedings of the 7th European Symposium on Corrosion Inhibitors, Ann. Univ. Ferrara. N. S.• Sez. V, 1990(Suppl. No.9), 711-724. 428. Gronostajski. J., Ali, W. 1., and Ghattas, M. S.: The effect of deformation on the corrosion resistance of coated steel sheets, J. Mater. Process. Technol. 23, 21-28, 1990. 429. Brown, W. N., and Mackowiak, 1.: A study of the kinetics of interaction between Fe(s) and Zn(l) in the temperature range 570-740 C. Corros. Sci. 5, 779-785. 1965. 430. Mackowiak, 1., and Short. N. R: The effect of pressure on the reaction between Fe(s)-Zn(l) at 501 and 554 c, Corros. Sci. 16,519-528, 1976. 431. Short, N. R., and Mackowiak, 1.: The effect of pressure on the reactions between Fe(s)-Zn: 1.5% AI(I) at 501 C, Corros. Sci. 17.397-404,1977. 432. Pereira. D., Scantlebury, 1. D., Ferreira, M. G. S., and Almeida, M. E.: The application of electrochemical measurements to the study and behaviour of zinc-rich coatings, Corros. Sci. 30. 1135-1147, 1990. 433. Hadden, S. E.: Effect of annealing on the resistance of galvanized steel to atmospheric corrosion, J.lron Steellnst. 1952(June), 121-127. 434. Campbell, H. S., Stanners, 1. E, and Watkins, K. 0.: Effect of heat treatment on the protective properties of zinc coatings on steel, J. Iron Steellnst. 1965(March), 248-251. 435. Rajagopalan, K. S., and Ramaseshan. G.: Relative corrodibility of zinc and steel in unpolluted atmospheres, J. Appl. Chern. 10, 493-496, 1960. 436. Anderson, E. A.: Zinc in marine environments, Corrosion 15, 409t-4121. 1959. 437. Gilbert, P. T., and Hadden, S. E.: White rust formation on zinc, J. Inst. Met. 78(Part 1),47-70, 1950. 438. Mansfeld, E: The effect of temperature on the corrosion rate of Zn of different purity in 6N KOH. Corros. Sci. 11,557-559, 1971. 439. Blickwede, D.l.: 55% AI-Zn-alloy-coated sheet steel, Iron Steel 1980(7), 821-834. 440. Mathieu, S., and Fenaille, B.: Characterization of zinc-iron and tin-iron coatings by coulometric method associated with x-ray diffraction and glow discharge optical spectrometry, Methodes d'Investigation des Metaux Conference, France, 1980. 441. Katz, W.: The structure of zinc coatings on the basis of electrochemical dissolutions, Arch. Eisenhiittenwes. 25(7/8), 307-314, 1954 [in GermanJ. 442. Chang, S.: Characterization and development of Zn-Fe alloyed coating layer, in Zinc-Based Steel Coating Systems: Metallurgy and Performance. G. Krauss and D. K. Matlock (eds.), pp. 319-330, The Minerals, Metals and Materials Society, Warrendale, Pennsylvania, 1990. 443. Landriault, 1. P., and Harrison, E w.: A comparison of thermally alloyed Zn-Fe coatings produced by mechanical wiping to those produced by post-coating annealing, Proceedings of the 25th Canadian Institute of Mining, Metallurgy, and Petroleum Conference, Toronto, Ontario, August 1986, pp. 71-78. 444. Imagawa, H., Matuno, K., and Konishi, T.: Prevention of corrosion cracking in liquid ammonia tanks by zinc spraying, Boshoku Gijutsu 38(6), 321-326, 1989. 445. Gmytryk, M., and Sedzimir, 1.: Corrosion of Zn in deaerated sulphate solutions at different pH values, Corros. Sci. 7,683-695, 1967. 446. Abayarathna, D., Hale, E. B., O'Keefe, T. J., Wang, Y. M., and Radovic, D.: Effects of sample orientation on the corrosion of zinc in ammonium sulfate and sodium hydroxide solutions, Corros. Sci. 32, 755-768, 1991.
REFERENCES
427
447. Weast, R. C: Zinc anodes for corrosion protection of domestic hot water storage tanks. Proceedings of the 7th International Conference on Hot Dip Galvanizing, Paris, June 1964, pp. 487-507. 448. TOITI~s- Villasenor, G .. Ugalde, A.. Hernandez. L.. and Singer. I. L.: Water vapour corrosion of lamellar. superplastic and cast dendritic Zn-2IAI alloy. Corms. Sci. 24. 159-166. 1984. 449. Trabanelli. G .. Zucchi. E. Brunoro. G .. and Gilli. G.: Characterization of the corrosion or anodic oxidation products on zinc. Electrodeposition SUlj: Treat. 3. 129-138. 1975. 450. Stuart. N .• and Evans. U. R: The effect of zinc on the corrosion-fatigue life of steel. 1. [ron Steel [n.H. 147.131-144.1943. 451. Stamatakis. P. Palmer. 8. R. Salzman. G. C. Bohren. C F., and Allen. T 8.: Optimum particle size of titanium dioxide and zinc oxide for attenuation of ultraviolet radiation. 1. Coatings Teclmol. 62. 95-98. 1990. 452. Solorzano. I. G .. Purdy. G. R .. and Weatherly, G. C: Studies of the initiation. growth and dissolution of the discontinuous precipitation product in aluminum-zinc alloys. Acta Metall. 32. 1709-1717. 1984. 453. Singh. R. N .. Bahadur. L .. and Singh. P.: The corrosion and electrochemical behaviour of zmc in binary mixtures of dimethylformamide and water. Corms. Sci. 27.561-566.1987. 454. Simm. D. W.: The corrosion of zinc in mortar and its relevance to wall tic COITosion. Corms. Prevo ContmI1984(December). 5-10. 455. Pettinger. B .. Schoppel. H.-R .. and Gerischer. H.: Tunnelling processes at highly doped ZnO electrodes in contact with aqueous electrolytes. Part I. Electron exchange with the conduction band. Ber. BUIlsellges. 78. 450-455. 1974. 456. Van Craen. M .. Haemers. G .• and Adams. E: Sims analysis of the atmospheric corrosion of a 55% AI-Zn coated steel wire. [nt. 1. Mass Spectrol1l. [on Phvs. 46,531-534, 1983. 457. von Fraunhofer. 1. A .. and Lubinski. A. T: Polarity reversal in the zinc-mild steel couple. Corms. Sci. 14. 225-232. 1974. 458. Gilbert. P T.: The nature of zinc corrosion products. 1. Electrochem. Soc. 99. 16-21. 1952. 459. Girin. O. B .. and Panasenko. S. A.: Effect of the texture of electrolytic zinc coatings on their corrosion resistance. Zashch. Met. 25(3). 480-482. 1989 Iin English]. 460. Gomes, W. P.. Freund. T. and Morrison. S. R: Chemical reactions involving holes at the zinc oxide single crystal anode. 1. Electrochem. Soc. 115. 818-823. 1968. 461. Shih. Y. R.: Investigation of the zinc-iron potential reversal and its application to cOITosion in domestic hot water systems. Ph.D. Thesis, San Jose State University. 1989. 462. Kotnik. L. J.: An investigation of the properties of zinc COlTosion products. Ph.D. Thesis. Casc Institute of Technology. 1960. 463. Biallozor, S .. and Bandura, E. T: On electrochemical behaviour of zinc in organic solvents. Elcctrochim. Acta 32. 891-894. 1987. 464. Perkins, 1.. and Bornholdt. R. A.: The corrosion product morphology found on sacrificial zinc anodes. Corros. Sci. 17.377-384. 1977. 465. McDonald. R. D.: Steel embrittlement problems associated with hot dip galvanizing-causes. mechanisms. controls. and selected references. Mate!: Perfonn. 14( 1). 31-37. 1975. 466. Keogh. T J.: The influence of pickling on the embrittlement of mild steel. Met. AlislI: 1972(August). 239-244. 467. Johnson. W. H.: On some remarkable changes produced in iron and steel by the action of hydrogen and acids. Pmc. R. Soc. London 23,1875. 468. Gonzalez. J. A .. and Andrade. C: Effect of carbonation. chlorides and relative ambient humidity on the corrosion of galvanized rebars embedded in concrete. Br. Corros. 1. 17,21-28. 1982. 469. Carson. J. A. H.: Zinc as a self-regulating galvanic anode for ship hulls. Corrosion 16(10).99-104. 1960. 470. Carson. 1. A. H .. Phillips. W. L. M .• and Wellington. 1. R.: A laboratory evaluation of zinc anodes in sea water, CorrosioI116(4). 107-113. 1960. 471. Kurr. G. W.: Zinc anodes-underground uses for cathodic protection and grounding. Mater. Perform. 18(4),34-41.1979. 472. Onishi. M., Wakamatsu. Y, and Miura. H.: Formation and growth kinetics of intermediate phases in Fe-Zn diffusion couples, Trans. lpn. Inst. Met. 15,331-337, 1974 473. Zinc Institute. Inc .. and Zinc Development Association: Cathodic protection of pipelines with zinc anodes. Pipei' Pipelines [Ill. 1975(August). 23-27.
428
REFERENCES
474. Matsumoto, Y., Yoshikawa, T., and Sato, E.: Dependence of the band bending of the oxide semiconductors on pH, J. Electrochem. Soc. 136,1389-1391,1989. 475. Justice, D. D., and Hurd, R. M.: Electrochemical dissolution of ZnO single crystals, J. Electrochem. Soc. 118,1417-1420,1971. 476. Craig, B. D.: Fundamental Aspects of Corrosion Films in Corrosion Science, Plenum Press, New York, 1991. 477. Bindra, P., Gerischer, H., and Kolb, D. M.: Electrolytic deposition of thin metal films on semiconductor substrates, J. Electrochem. Soc. 124, 1012-1018, 1977. 478. Dirkse, T. P., and Shoemaker, R: Double layer capacitance of zinc electrodes in KOH solutions, J. Electrochem. Soc. 115, 784-786, 1968 479. Gomes, W. P., Freund, T., and Morrison, S. R.: Chemical reactions involving holes at the zinc oxide single crystal anode, J. Electrochem. Soc. 115, 818-823,1968. 480. Knotkova-Cermakova, D., and V1ckova. J.: Corrosive effect of plastics. rubber and wood on metals in confined spaces, Br. Corros. J. 6. 17-22. 1971 481. De Pauli, C. P.. Derosa. O. A. H., and Giordano, M. c.: Zinc dissolution and passivation in buffered phosphate solutions. Part I. A comparative study with sodium hydroxide solutions. J. EleCfroallal. Chem. 86.335-348, 1978. 482. Pirnat, A., Meszaros, L.. and Lengyel, B.: Study of the formation ofchromate layer on zinc by impedance technique, Electrochim. Acta 35, 515-522. 1990. 483. Juttner. K.: Electochemical impedance spectroscopy (ElS) of corrosion processes on inhomogeneous surfaces. Electrochim. Acta 35.1501-1508,1990. 484. Vi1che. J. R. Juttner. K., Lorenz, W. 1., Kautek, w.. Paatsch, w., Dean, M. H., and Stimming, U.: Semiconductor properties of passive films on Zn, Zn-Co, and Zn-Ni substrates, J. Electrochem. Soc. 136,3773-3778,1989. 485. Allegra, L., and Townsend, H. E.: Undervehicle corrosion resistance of 55% AI-Zn coated steel sheet, Met. Prog. 119(4),33-35, 1981. 486. Uchiyama, Y.: Corrosion behaviour of zinc-aluminium alloys, Corros. Eng. 39, 297-307,1990. 487. Srivastava, S. K., Srivastava, S., Shukla, R, and Srivastava, S. c.: Effect of addition agents on Zn-Ni-Ti alloy deposition from a sulphate bath, Indian J. Technol. 27. 564-568, 1989. 488. Menzies, I. A., Marshall, G. w., Gearey, D., and Griffin, G. B.: Some aspects of the mechanism of the anodic dissolution of zinc in sulfamic acid-formamide solutions, J. Electrochem. Soc. 117, 849-854, 1970. 489. Radeker, w., Peters, F. K., and Friehe, w.: The effect of alloy additions on the properties of hot galvanized coatings, Proceedings of the 6th International Conference on Hot Dip Galvanizing. Zinc Development Association, London, England, June 1961.238-264. 490. Arsenault, B., Champagne, B., Lambert, P., and Dallaire. S.: Zinc-nickel coatings for improved adherence and corrosion resistance, Surf. Coatings Technol. 37, 369-378, 1989. 491. Kim, D. K., and Leidheiser, H.: Chemistry of the surface of hot-dipped galvanized steel and its relation to paint adherence, Surf. Technol. 5, 379-396, 1977. 492. Graedel, T. E.: Corrosion-related aspects of the chemistry and frequency of occurrence of precipitation, J. Electrochem. Soc. 133,2476-2482.1986. 493. MacEwan, J. R: Chemical and electrochemical finishes foq:,i.I)c. Consolidated Mining and Smelting Company of Canada Ltd., Research and Development DiviSIOn, August 1956. 494. Whitaker, M., and Fry, H.: The anodic oxidation of zinc and zinc-aluminium alloys in a sodium hyroxide-sodium carbonate solution, The British Non-Ferrous Metals Research Association, Report No. A. 1172, February 1958. 495. Lange. N. A.: Lange's Handbook of Chemistry, J. A. Dean (ed.), 13th ed., McGraw-Hill, New York, 1985. 496. Gubbins, K. E., and Walker, R D.: The solubility and diffusivity of oxygen in electrolytic solutions, J. Electrochem. Soc. 112,469-471,1965. 497. Roetheli, B. E., Cox, G. L., and Littreal, W. B.: Effect of pH on the corrosion products and corrosion rate of zinc in oxygenated aqueous solutions, Met. Alloys 3(3), 73-76, 1932. 498. Engelbrecht, G. J.: A fundamental study of the influence of aluminium on the white rusting of galvanized steel, Ph.D. Thesis, Delft Technical College, Republic of South Africa, 1963.
REFERENCES
429
499. McLeod, W.. and Rogers. R. R.: The nature of corrosion of zinc by sulfurous aCid at ordmary temperatures, CorrositJl1 25. 74-76, 1969. 500. Hronsky, P.: Corrosion behavior of metallic materials in organic media containing hydrogen chloride, Corrosion 37. 161-170. J 981. 50 I. Porter. F.: Zinc Handbook, Marcel Dekker, New York. 1991. 502. Dean. S. W.: Planning. instrumentation, and evaluation of atmospheric corrosion tests and a review of ASTM testing. in Atmospheric Corrosion. W. H. Ailor (ed.), pp. 195-216. John Wiley & Sons. New York, 1982. 503. American Society for Testing and Materials: Standard practice for conducting atmospheric corrosion test on metals, in Annual Book of'ASTM Standards, Vol. 03.02. No. G50-76. pp. 192-196, 1989. 504. American Society for Testing and Materials: Standard practice for laboratory immersion corrosion testing of metals. in Anllual Book of ASTM Standards. Vol. 03.02, No. G3 J -72, pp. \08-115. 1989. 505. Nakasa. K .. and Suzawa, M.: Liquid-zinc-induced crack propagation in high-strength steels in Proceedings oFlhe 1st Inlernational Conference 011 Environment-Induced Cracking of'Metals. October 2-7, 1988, Kohler, Wisconson, National Association of Corrosion Engineers, Houston, 537-540, 1988. 506. Anon.: Panel discussion of inorganic zinc primers at corrosionl75, Male!: Peifonn. 15(5),35-38, 1976. 507. Hinden, E. R: How zinc rich epoxy coating protect structures, Mater. ProI7(4), 38-40, 1968. 508. Okamura, H., and Hisamatsu. Y: Effect of use of galvanized steel on the durability of reinforced concrete, Mala Peiform. 15(7),43-47, 1976. 509. Haynie. F. H., and Upham. 1. B.: Effects of atmospheric sulfur dioxide on the corrosion of zinc, Male!: Prot. and Perform. 9(8),35-40. 1970. 510. Bird, C. E.: Corrosion behavior of galvanized sheet in relation to variation in coating thickness, Mate!: Perform. 16(4), 14-16, 1977. 511. Townsend, H. E .. and Zoccola, J. c.: Atmospheric corrosion resistance of 55% AI-Zn coated sheet steel: I3 year test results, Mater. Prol. 18(10). 13-20, 1979. 512. Vickers. R.: Commercial hot dip galvanizing of fabricated items, Maler. Pmt. 1( 1).30-39. 1962. 513. Hendry. C. M.: Inorganic zinc rich primers-fact and fancy, Mate!: Perf'orm. 17(5). 19-27. 1978. 514. Dewald. 1. E: The charge and potential distributions at the zinc oxide electrode, Bell S\"-'I. Tech . .1. 39. 615-639. 1960. 515. Doyle. D. P, and Wright. T. E.: Quantitative assessment of atmospheric galvanic corrosion. in Galvanic Corrosion, STP 978, pp. 161-171, American Society for Testing and Materials, Philadelphia. 1988. 516. Scully. J. R .. and Hack, H. P: Prediction of tube-tubesheet galvanic corrosion using finite clement and wagner number analyses, in Galvanic Corrosioll. STP 978. pp. 136-157. American Society for Testing and Materials. Philadelphia. 1988. 517. Escalante. E.: The effect of soil resistivity and soil temperature on the corrosion of galvanically coupled metals in soil. in Galvanic Corrosion. STP 978. pp. 193-202. American Society for Testing and Materials. Philadelphia. 1988. 518. Schick. G.: Avoiding galvanic corrosion problems in the telephone cable plant, in Galvanic Corrosion. STP 9n, pp. 283-290, American Society for Testing and Materials, Philadelphia. 1988. 519. Swamy. R. N.: Resistance to chlorides of galvanized rebars, in Corrosion o/Rein/iJrcCmelll in COllcrete, C. L. Page. K. W. J. Treadaway. and P B. Bamforth (eds.), pp. 586-600. Elsevier Science Publishing Co., New York. 1990. 520. Brousseau, R .. Arnott. M .. Dallaire, S .. and Feldman. R: Factors affecting the adhesion on concrete of arc sprayed zinc, Corrosion 48. 947-952,1992. 521. Geduld. H.: Zinc Plaling, Finishing Publications Ltd., Teddington, Middlesex, England. 1988. 522. Zhang. X. G., and Valeriote. E. M.: Galvanic protection of steel and galvanic corrosion of zinc under thin layer electrolytes, Corros. Sci. 34, 1957-1972. 1993. 523. Chen. Z. W., Kennon, N. E, Sec, J. B., and Barter. M. A.: Technigalva and other developments in batch hot-dip galvanizing, .1. MelLlls 44(1),22-26, 1992. 524. Fontana. M. G .. and Greene, N. D.: Corrosion Engineering, 2nd cd., McGraw-Hill. New York. 1978. 525. Cheng. H. B.: Hot dip galvanizing, Galvaniz:ers f(JlInd. Taiwan 20(4),19-30,1992 [in Chinesel. 526. Bothe, A. E., Vilche, 1. R, Juttner, K., Lorenz, W. J., Kautek, W., and Paatsch, W.: An electrochemical impedance spectroscopy study of passive zinc and low alloyed zinc electrodes in alkaline and neutral aqueous solutions, Corros. Sci. 32, 621-633, 1991.
430
REFERENCES
527. Morrow, H.: Encyclopedia of Materials Science and Engineering, Vol. 7, M. B. Bever (ed.), MIT Press, Cambridge, Massachusetts, 1986. 528. National Research Council: Zinc, University Park Press, Baltimore, 1979. 529. Mongeon, L., and Barnhurst, R. 1.: Zinc and zinc alloys, in ASM Metals Handbook, 9th ed .. Vol. 9, pp. 488-496, American Society for Metals, Materials Park, Ohio, 1985. 530. Barnhurst, R. 1.: Zinc and zinc alloys, in ASM Metals Handbook, 10th ed., Vol. 2, American Society for Metals, Materials Park, Ohio, pp. 527-542,1990. 531. Zhang, X. G.: Electrochemical stripping of galvannealed coatings on steel, Corrosion 50, 308-317, 1994. 532. Brodd, R. J., and Leger, V. E.: Zinc, in Encyclopedia of Electrochemistry of the Elements, Vol. V, A. 1. Bard (ed.), pp. 2-60, Marcel Dekker, New York, 1976. 533. Naybour, R. D.: The effect of electrolyte flow on the morphology of zinc electrodeposited from aqueous alkaline solution containing zincate ions, J. Electrochem. Soc. 116,520-524, 1969. 534. Baugh, L. M., Tye, E L., and White, N. c.: Corrosion and polarization characteristics of zinc in battery electrolyte analogues and the effect of amalgamation, J. Appl. Electrochem. 13, 623-635, 1983. 535. Roberts, C. W: Intercrystalline corrosion in cast zinc-aluminium alloys, J. Inst. Met. 81,301-309, 1952-53. 536. Kolberg, T., Rausch, W, Schubach, P., and Wendel, T. (Metallgesellschaft A. G., Germany): Formation of conversion coatings on surfaces of zinc or zinc alloys, CA Patent 2041892, November 30, 1991. 537. Sereda, P. 1., Croll, S. G., and Slade, H. E: Measurement of the time-of-wetness by moisture sensors and their calibration, in Atmospheric Corrosion of Metals, STP 767, pp. 267-285, American Society for Testing and Materials, Philadelphia, 1982. 538. American Society for Testing and Materials: Standard practice for measurement of time-of-wetness on surfaces exposed to wetting conditions as in atmospheric corrOSIOn testing, in Annual Book of ASTM Standards, Vol. 03.02, No. G84-89, pp. 353-358,1989. 539. Corrosiveness of various atmospheric test sites as measured by specimens of steel and zinc, in Metal Corrosion in the Atmosphere, STP 435, pp. 360-391, American Society for Testing and Materials, Philadelphia, 1968. 540. Evans, E. L. 1.: Chapter 2.1 in Corrosion, Vol. 1, Corrosion of Metals and Alloys, L. L. Shreir (ed.), George Newnes Ltd., London, 1963. 541. Sereda, P. J.: Weather factors affecting corrosion of metals, in Corrosion in Natural Environments, STP 558, pp. 7-22, American Society for Testing and Materials, Philadelphia, 1974. 542. Brown, P. W., and Masters, L. W.: Factors affecting the corrosion of metals in the atmosphere, in Atmospheric Corrosion, W H. Ailor (ed.), pp. 31-49, John Wiley & Sons, New York, 1982. 543. Copson, H. R.: Report of subcommittee VI, of Committee B-3, on Atmospheric Corrosion, ASTM 58th Annual Meeting, Symposium on Atmospheric Corrosion of Non-Ferrous Metals, STP 175, pp. 3-19, American Society for Testing and Materials, Philadelphia, 1955. 544. Teeple, H. 0.: Atmospheric galvanic corrosion of magnesium coupled to other metals, ASTM 58th Annual Meeting, Symposium on Atmospheric Corrosion of Non-Ferrous Metals, STP 175, pp. 89-115, American Society for Testing and Materials, Philadelphia, 1955. 545. Compton, K. G., and Mendizza, A.: Galvanic couple corrosion studies by means of the threaded bolt and wire test, ASTM 58th Annual Meeting, Symposium on Atmospheric Corrosion of Non-Ferrous Metals, STP 175, pp. 116-125, American Society for Testing and Materials, Philadelphia, 1955. 546. Anderson, E. A.: The atmospheric corrosion of rolled zinc, ASTM 58th Annual Meeting, Symposium on Atmospheric Corrosion of Non-Ferrous Metals, STP 175, pp. 126-134, American Society for Testing and Materials, Philadelphia, 1955. 547. Baker, E. A., and Lee, T. S.: Calibration of atmospheric corrosion test sites, in Atmospheric Corrosion of Metals, ST.P 767, pp. 250-266, American Society for Testing and Materials, Philadelphia, 1982. 548. Haynie, E H.: Evaluation of the effects of microclimate differences on corrosion, in Atmospheric Corrosion of Metals, STP 767, pp. 286-308, American Society for Testing and Materials, Philadelphia, 1982. 549. Showak, W, and Dunbar, S. R.: Effect of I percent copper addition on atmospheric corrosion of rolled zinc after 20 years' exposure, in Atmospheric Corrosion of Metals, STP 767, pp. 135-162, American Society for Testing and Materials, Philadelphia, 1982.
REFERENCES
431
550. Tonini, D. E.: Atmospheric corrosion test results for metallic-coated steel panels exposed in 1960, in Atmospheric Corrosion of Metals, STP 767, pp. 163~ 185, American Society for Testing and Materials, Philadelphia, 1982. 551. Baboian, R.: Final report on the ASTM study: atmospheric galvanic corrosion of magnesium coupled to other metals, in Atmospheric Factors Affecting the Corrosion of Engineering Melllls, STP 646, pp. 17~29, American Society for Testing and Materials, Philadelphia, 1978. 552. Longo, F N., and Durrnann, G. 1.: Corrosion prevention with thermal-sprayed ZInc and aluminum coatings, in Atmospheric Factors Affecting the Corrosion of Engineering Metals, STP 646, pp. 97 ~ I 14, American Society for Testing and Matenals. Philadelphia. 1978. 553. Schikorr, G.: Behavior of zinc in the atmosphere, in Corrosion 8ehavior o{Zinc, Vol. I, American Zinc Institute and the Zinc Development Association, 1965; cited on p. 30 in Zinc: Its Cormsion Resistance, 2nd ed .. International Lead Zinc Research Organization. New York, 1986. 554. Barton, K .. and Beranek. E.: Reaction mechanism of the atmospheric corrosion of metals in moist air
555. 556. 557. 558. 559. 560. 561. 562.
563. 564. 565. 566.
contaminated with sulfur dioxide, Werkst. Korros. 10, 377~383, 1959; cited on p. 27 in Zinc: Its Corrosion Resistance. 2nd ed .. International Lead Zinc Research Organization, New York, 1986. Schikorr, G., and Schikorr, I.: The resistance of zinc to atmospheric corrosion. Z. Metallkd. 35(9). 175~181. 1943 [TranslationJ. Rozenfeld, 1. L.: Atlllospheric Corrosion of Metals, National Association of Corrosion Engineers. Houston, 1972. Hudson. J. c.: TrailS. Faradav Soc. 25. 177. 129. Drane, C. W.: Chapter 2.2 in Corrosion, Vol. 1, CorrosionofMetalsandAllo.\'s, L. L. Shreir(ed.), George Newnes Ltd., London, 1963. Anderson, E. A .. and Reinhard. C. E.: Zinc, in The Corrosion Handbook, H. H. Uhlig (ed.), pp. 331 ~346. John Wiley & Sons. New York. 1948. Schikon', G.: Z. Metallkd. 32. 314.1940; cited on pp. 4 and 138 in Corrosion, Vol. I. Corrosion o{Metals and Allo,vs, L. L. Shrcir (ed.). George Newnes Ltd .. London. 1963. Fujii, T.: Rust Prevention (80sei Kanri) 27, 85,1983; cited on p. 97 in Corrosion-Resistant Coatings Technology, P. A. Schweitzer (ed.), Marcel Dekker, New York, 1989. Suzuki, N.: Safety Engineering of Corrosion Protection-Protective Coatings, The High Pressure Gas Safety Institute of Japan, 1986, p. 13; cited on p. 104 in Corrosion-Resistant Coatings Technologv, P. A. Schweitzer (ed.). Marcel Dekker, New York, 1989. Crennell. J. T.: Chapter 2.3 in Corrosion, Vol. I, Corrosion of Metals and Alloys, L. L. Shreir (ed.), George Newnes Ltd., London, 1963. Tsukuda, T., and Kawabe, M.: Therm. Nucl. Power Eng. Soc. Jpn. 25, 985, 1974: cited on p. 103 in Corrosion-ResisIallt Coatings Technologr. P. A. Schweitzer (ed.), Marcel Dekker, New York, 1989. Logan. K. H.: Corrosion by soils, in The Corrosion Handbook. H. H. Uhlig (ed.), pp. 446~466. John Wiley & Sons, New York, 1948. Redfield. A. c.: Characteristics of sea water, in The Corrosion Handhook, H. H. Uhlig (cd.), pp.
III1 ~ 1122. John Wiley & Sons, New York. 1948. 567. Kim, D. K., and Leidheiser, H.: Metall. Trans. 98. 581. 1978; cited on p. 218 in Corrosion-Resistant Coatings Technology. P. A. Schweitzer (ed.). Marcel Dekker. New York. 1989. 568. Standish, J. V: Atmospheric corrosion mechanisms of painted steel. Society of Automotive Engineers International Congress & Exposition, Detroit, February 28~March 4, 1983, SP-538. pp. 1~lJ. 569. Hafford, B. c.: Zinc-containing protective coatings, in Encvc/opedia of'Materials Science and Engineering. Vol. 7, M. B. Bever (ed.), pp. 5521~5528, MIT Press. Cambridge, Massachusetts, 1986. 570. Yan, M. F: Zinc oxide, in Encyclopedia of Materials Science and Engineering, Vol. 7, M. B. Bever (ed.), pp. 5528~5530, MIT Press. Cambridge. Massachusetts, 1986. 571. Baboian. R.: The automotive environment. in Automotive Corrosion b,' Deicing Salts, R. Baboian (ed.), pp. 3~ 12, National Association of Corrosion Engineers, Houston, 1981. 572. Neville, R. J.: A test for under vehicle corrosion resistance, in Automotive Corrosioll by Deicing Salts, R. Baboian (ed,), pp. 182~218, National Association of Corrosion Engineers. Houston, 1981. 573. Rynewicz,1. F: Evaluation of paint coatings tested in the deep Atlantic and Pacific oceans, in Corrosion in Natural Environments. STP 558, pp. 209~235, American Society for Testing and Materials, Philadelphia. 1974.
432
REFERENCES
574. Huot, J. Y: The effects of silicate ion on the corrosion of zinc powder in alkaline solutions, J. Appl. Electrochem. 22,443-447,1992. 575. Reinhart, EM.: Corrosion of metals and alloys in the deep ocean, Naval Facilities Engineering Command, Alexandria, Virginia, Technical Report R834. February 1976. 576. Natorski, T. J.: Zinc and zinc alloy plating in the '90s, Met. Finish. 90(3),15-17.1992. 577. Pearlstein, E: Selection and application of inorganic finishes. Part 1. Black oxide and phosphate coatings, Plating Suif. Finish. 65(] 2), 22-29, 1978. 578. Phosphate coating, in ASM Metals Handbook, 9th ed., Vol. 5, pp. 434-456, American Society for Metals, Materials Park, Ohio. 1982. 579. Cheever, G. D.: Formation and growth of zinc phosphate coatings, J. Paint Techno!. 39. 1-13, 1967. Photoelectrochemistry of semiconductor ZnO particulate films, J. 580. Hotchandani, S., and Kamat, P. Electrochem. Soc. 139. 1630-1634, 1992. 581. Tegehall, P. E., and Vannerberg, N.-G.: Nucleation and formation of zinc phosphate conversion coating on cold-rolled steel, Corros. Sci. 32, 635-652, 1991. 582. International Organization for Standardization: Chromate conversion coatings on zinc and cadmiumtest methods, International Standard, ISO 3613-1980 (E), pp. 243-246. 583. American Society for Testing and Materials: Standard practice for testing chromate coatings on zinc and cadmium surfaces, Annual Book of ASTM Standards, Vol. 02.05, No. B201-80, pp. 58-60, 1980. 584. Centre de Recherches Metallurgiques (Belgium): Continuous gal fan coating of steel sheet and wire, International Lead Zinc Research Organization Project ZM-285, Progress Report No. 20, November 1988-ApriI1989. 585. Johnsson, T., and Kucera, Eight years' field exposure of coatings of zinc, aluminium and their alloys, Proceedings of the 2nd International Conference on Zinc Coated Steel Sheet, Rome, Italy, June 9-10, 1988, pp. SA6/I-SA6/l1. 586. Davin, A., and Goodwin, E E.: Atmospheric corrosion behaviour of gal fan coated steels, Proceedings of the 2nd International Conference on Zinc Coated Steel Sheet, Rome, Italy, June 9-10, 1988, pp. SA2/l-SA2/31. 587. Belisle, S., and Dufresne, R.: Corrosion resistance of ZA alloys, Canadian Institute of Mining, Metallurgy, and Petroleum 25th Annual Conference of Metallurgists, International Symposium on Zinc-Aluminum (ZA) Casting Alloys, Toronto, Ontario, August 17-20, 1986, pp. 109-126. 588. Vrable, J. B.: Effect of forming deformation on corrosion test performance of electroplated zinc and iron-zinc alloy coatings for automotive sheet, SAE Technical Paper No. 862012, Society of Automotive Engineers Automotive Corrosion and Prevention Conference, Dearborn, Michigan, December 8-10, 1986. 589. Gray, B. M., and Shaffer, R. 1.: An evaluation of cosmetic corrosion performance of zinc and zinc alloy coated steels, SAE Technical Paper No. 862014, Society of Automotive Engineers Automotive Corrosion and Prevention Conference, Dearborn, Michigan, December 8-10,1986. 590. Lay, D. E., and Eckles, W. E.: The fundamentals of zinc/cobalt, Plating Suif. Finish. 77(10), 10-14. 1990. 59!. Anderson, E. A.: Chromate coatings on zinc, in The Corrosion Handbook, H. H. Uhlig (ed.), pp. 862-864, John Wiley & Sons, New York, 1948. 592. Taylor, E: Chromate passivation, Met.lnd, 1944(September), 149-150. 593. Anderson, E. A.: The Cronak process, Proc. Am. Electroplaters Soc. 31, 6-13, 1943. 594. Coppins. W. C: Chemical colouring, Met. Ind. 1948(June),482. 595. Stareck, 1. E.: Anozinc and other new coatings on zinc, Proc. Am. Electroplaters Soc. 32, 235-247, 1944. 596. Pearlstein, E: Selection and application of inorganic finishes. Part IV. Anodic coatings for other metals, Plating Suif. Finish. 66, 41-46, 1979. 597. Wilcox, G. D., and Gabe, D. R.: Passivation studies using group via anions. Part iv. Cathodic redox reactions and film formation, Br. Corros. J. 19, 196-200, 1984. 598. Barnes, C, Ward, J. J. B., Sehmbhi, T. S., and Carter, V. E.: Non-chromate passivation treatments for zinc, Trans. Inst. Met. Finish. 60,45-48, 1982. 599. Shalaby, L. A., and Abbas, H.: The behaviour of zinc and zinc-silver alloys in the presence of some inhibitors, Carras. Sci. 13, 545-552, 1973.
v.:
v.:
REFERENCES
433
600. Abdul Azim, A. A., Shalaby, L. A., and Abbas, H.: Mechanism of the corrosion inhibition ofZn anode in NaOH by gelatine and some inorganic anions, Corros. Sci. 14,21-24, 1974. 601. LeRoy. R. L., and Zavorsky, Z.: Cathodic pretreatment of zinc and zinc-coated surfaces, Corms. Sci. 17, 943-944, 1977. 602. Rudresh, H. B., and Mayanna, S. M.: The synergistic effect of halide ions on the corrosion inhibition of zinc by n-decylamine, Corros. Sci. 19,361-370,1979. 603. De Pauli, C P., Giordano, M. C, Lopez, B. A., and Manzur, M. E.: The passivation of zinc in slightly acidic buffered solutions in the presence of phosphate ions, Corros. Sci. 28, 769-785, 1988. 604. Bagger, N. B.: Colouring of metals, Mater. Methods 1949(June), 67-82. 605. Hinton, B. R. W.: New approaches to corrosion inhibition with rare earth metal salts, National Association of Corrosion Engineers Corrosion '89 Conference, New Orleans, Louisiana, April 17-21, Paper No. 170. 606. Hosny, A. Y., El-Rafei, M. E., Ramadam, T A., and El-Gafari, B. A.: Corrosion resistance ofzinc coatings produced from a sulphate bath, Met. Finish. 1995(November), 55-59. 607. Suzuki, l.: The behavior of corrosion products on zinc in sodium chloride solution, Corros. Sci. 25, 1029- 1034, 1985. 608. Keddam, M., Hugot-Le-Goff, A., Takenouti, H., Thierry, D., and Arevalo, M. C: The influence of a thin electrolyte layer on the corrosion process of zinc in chloride-containing solutions, Corms. Sci. 33, 1243-1252,1992. 609. Dante, J. F., and Kelly, R. G.: The evolution of the adsorbed solution layer during atmospheric corrosion, Symposium on Corrosion and Reliability of Electronic Materials and Devices, Electrochemical Society Meeting, Toronto, Ontario, October 1992. 610. Kawafuku, J., Katoh, J., Toyama, M., Nishimoto, H., Ikeda, K., and Satoh, H.: Structure and corrosion resistance of zinc alloy coated steel sheets obtained by continuous vapor deposition apparatus, 1. Iron Steel lilst. Jpn. 77, 995-1002,1991. 611. Sagiyama, M., Hiraya, A., and Watanabe, T.: Electrochemical behavior of electrodeposited zinc-iron alloys in 5'70 NaCI solution, J.lmn Steel Inst. lpn. 77,244-250, 1991. 612. Sagiyama, M., Hiraya, A., and Watanabe, T.: Electrochemical behavior of electrode posited zinc-iron alloys in alkaline solutions, J. Iron Steel blSt. Jpn. 77,251-257, 1991. 613. Schubert. R.: The interaction of sulfuric acid vapors and zinc surfaces, 1. p;fectrochem. Soc. 139, 3450-3453, 1992. 614. Zhang, X. G.: Effect of weather pattern on corrosion of zinc and steel, unpublished results. 615. Grossman, P. R.: Investigation of atmospheric exposure factors that determine time-of-wetness of outdoor structures, in Atmospheric Factors Affecting the Corrosion of Engineering Metals, STP 646, pp. 5-16, American Society for Testing and Materials, Philadelphia, 1978. 616. Baedecker, P. A., Edney, E. 0., Moran, P. J., Simpson, T C, and Williams, R. S.: Effects of acidic deposition on materials, State-of-Science/Technology, National Acid Precipitation Assessment Program, Report No. 19, May 1990. 617. Daescn, J. R.: Corrosion resistance of galvanized coatings, Proceedings of the 2nd International Congress on Metallic Con'osion, New York, March 11-15, 1963, pp, 686-705. 618. Ellis, O. B.: Effect of weather on the initial corrosion rate of sheet zinc, Proc. ASTM COllf 49, 152-170, 1949. 619. Salt, F. W., Stanners, J. F., and Watkins, K. 0.: Atmospheric exposure trials of electrodeposited iron-zinc alloy coatings, Br. Corros. 1. 1,5-8, 1965. 620. Dey, A. K.. Sinha Mahapatra, A. K., Khan, D. K., Mukherjee, A. N., Narain, R., Mukherjee, K. P., and Baneljee. T: Interim report on atmospheric corrosion studies under marine atmosphere, NML (Natl. Mew/!. Lab., Jmnshedpur. India) Tech. J. 8, 11-16, 1966. 621. Cooke, F., Cosslett, J. K., Scott, R. W., and Shanahan, C E. A.: The atmospheric cOlTosion resistance of elcctrodeposited tin-cadmium and tin-zinc coatings on steel, Br. Corros. l. 1. 283-289, 1966. 622. Payer, J. H .. and Safranek, W. H.: Atmospheric corrosion of electroplated zinc alloy die casting, in Atmospheric Factors Affecting the Corrosion of Engineering Metals, STP 646, pp. 115-128, American Society for Testing and Materials, Philadelphia, 1978. 623. Fukushima, T, Sato, N., Hisamatsu, y, Matsushima, T, and Aoyama, Y.: Atmospheric corrosion testing in Japan, in Atmospheric Corrosion, W. H. Ailor (ed.), pp. 841-872, John Wiley & Sons, New York, 1982.
434
REFERENCES
624. Longo, EN., and Durmann, G. 1.: Corrosion prevention with thermal-sprayed zinc and aluminum coatings, in Atmospheric Factors Affecting the Corrosion of Engineering Metals, STP 646, pp. 97-114, American Society for Testing Materials, Philadelphia, 1978. 625. Spedding, D. 1.: Some investigations into hydrogen sulphide adsorption by zinc, Corros. Sci. 17, 173-178, 1977. 626. Spence, 1. W, Edney, E. 0., Haynie, E H., Stiles, D. c., Corse, E. W, Wheeler. M. S., and Cheek, S. E: Advanced laboratory and field exposure systems for testing materials, in Corrosion Testing and Evaluation: Silver Anniversary Volume. STP 1000, pp. 191-207, American Society for Testing and Materials, Philadelphia, 1990. 627. Spence, 1. W, and Haynie, E H.: Derivation of a damage function for galvanized steel structures: Corrosion kinetics and thermodynamic considerations, in Corrosion Testing and Evaluation: Silver Anniversary Volume, STP 1000. pp. 208-224. American Society for Testing and Materials. Philadelphia. 1990. 628. Haynie, E H .• Spence, 1. W, Lipfert, E W, Cramer, S. D., and McDonald. L. G.: Evaluation of an atmospheric damage function for galvanized steel, in Corrosion Testing and Evaluation: Silver Anniversary Volume, STP 1000, pp. 225-240. American Society for Testing and Materials, Philadelphia, 1990. 629. Cramer, S. D., and McDonald, L. G.: Atmospheric factors affecting the corrosion of zinc. galvanized steel and copper, in Corrosion Testing and Evaluation: Silver Anniversary Volume, STP 1000. pp. 241-259, American Society for Testing and Materials, Philadelphia. 1990. 630. Zhang. X. G.: Corrosion of zinc in salt spray test. Cominco Report 651-008, 1992. 631. Spence. J. W, Haynie, E H., Lipfert, E w., Cramer, S. D .• and McDonald, L. G.: Atmospheric corrosion model for galvanized steel structures, Corrosion 48.1009-1019, 1992. 632. Locke, C. E.: Solving rebar corrosion problems in concrete, National Association of Corrosion Engineers. Chicago. September 1982, reprint #2. 633. Locke, C. E.: Solving rebar corrosion problems in concrete. National Association of Corrosion Engineers. September. Chicago. September 1982, preprint #5. 634. Lin, S. H.: Chloride diffusion in a porous concrete slab, Corrosion 46. 964-967. 1990. 635. lang. 1. W.• and Iwasaki, I.: Rebar corrosion under simulated concrete conditions using galvanic current measurements, Corrosion 47. 875-884, 1991. 636. Gau, Y.. and Cornet, I.: Penetration of hardened concrete by seawater chlorides with and without impressed current. Corrosion 41, 93-100. 1985. 637. Tritthart. J.: Pore solution composition and other factors influencing the corrosion risk of reinforcement in concrete. in Corrosion of Reinforcement in Concrete. C. L. Page, K. W. J. Treadaway, and P. 8. Bamforth (eds.), pp. 96-106, Elsevier Science Publishing Co. New York. 1990. 638. Bamforth, P. 8.. and Pocock, D. c.: Minimising the risk of chloride induced corrosion by selection of concreting materials, in Corrosion of Reinforcement in Concrete, C. L. Page, K. W. J. Treadaway, and P. B. Bamforth (eds.), pp. 119-131. Elsevier Science Publishing Co .• New York, 1990. 639. Dehghanian, c., and Locke, C. E.: Electrochemical behavior of steel in salt contaminated concrete: Part I, Corrosion 39. 299-305. 1983. 640. American Society for Testing and Materials: Standard test method for half-cell potentials of uncoated reinforcing steel in concrete, in ASTM Standards for Corrosion Testing of Metals, No. C876-87. pp. 94-98. ASTM, Philadelphia, 1990. 641. Elsener, 8., Muller, S., Suter, M., and Bohnl, H.: Corrosion monitoring of steel in concrete-theory and practice, in Corrosion of Reinforcement in Concrete, C. L. Page, K. W. J. Treadaway, and P. 8. Bamforth (eds.), pp. 348-357, Elsevier Science Publishing Co., New York, 1990. 642. Borgard, 8., Ramirez, c., Somayaji, S., Jones, D., Keeling, D., and Heidersbach, R.: Failure analysis in concrete structures: A comparison of field data with results from laboratory exposure, Corrosion 47, 758-769, 1991. 643. Funahashi, M., and Bushman, J. 8.: Technical review of 100 mv polarization shift criterion for reinforcing steel in concrete, Corrosioll47, 376-386,1991. 644. Rosenberg, A. M., Gaidis, 1. M., Kossivas, T. G., and Previte, R. W: A corrosion inhibitor formulated with calcium nitrite for use in reinforced concrete, in Chloride Corrosion of Steel in Concrete, STP 629, pp. 89-99, American Society for Testing and Materials, Philadelphia, 1977.
REFERENCES
435
645. Cornet, I.. and Bresler. B.: Galvanized steel in concrete: Literature review and assessment of perform· ance, in Galvanized Reinforcement for Concrete~lI. pp. I-56. International Lead Zinc Research Organization. New York. 1981. 646. Gonzalez, 1. A .. Algaba, S., and Andrade. c.: Corrosion of reinforcing bars in carbonated concrete. Br. Corms . .I. 15, 135-139. 1980. 647. Clemena, G. G.:Electrically conductive portland cement concrete, Proceedings of the National Asso· ciation of Corrosion Engineers Corrosion '87 Symposium on Corrosion of Metals in Concrete. pp. 49-64. 648. Krepski. R. P., Zinc silicate thermal spray coating for corrosion control of steel in concrete, Proceedings of Fourth National Thermal Spray Conference. Pittsburgh, Pennsylvania, 1991. p. 433. 649. Yeomans. S. R.: Corrosion testing of black, galvanized and epoxy coated reinforcing steel in concrete, University of New South Wales, internal report Australia, 1993. 650. Piercy, R. c.: Cominco Project No. 67·6·747, Report No.5. 1967. 65 I. Watson. T. w.: Cominco Project No. 67·6·749. Report No.3. 1967. 652. Keitelman. A. D., Gravano. S. M .• and Galvele. J. R.: Localized acidification as the cause of passivity breakdown of high purity zinc. Corms. Sci. 24, 535-545, 1984. 653. Li, S., He. 1.. and Li. S.: The effect of rare earth elements on corrosion resistance of electrodeposited zinc coating. Proceedings of the 7th APCCC Conference on Corrosion Control, Beijing China. 1991. Vol. 2. pp. 1034-1038. 654. Bengough, G. D., and Hudson, O. F.: Fourth report to the corrosion committee of the institute of metals. Presented at the Annual General Meeting. London. March 25, 1919. 655. Featherly. R. L.: Practical experiences in corrosion of hot water tanks. Plumbing Hearing .I. 1952(ApriI/May). 656. Newell, I. L.: The corrosion of domestic galvanized hot water storage tanks, Corrosio/1 9, 46-51. 1953. 657. Duval R.. and Arliguie. M. G.: Passivation du zinc dans I'hydroxyde de calcium. eu, egard au comportement de I'acier galvanise dans Ie beton. Mem. Sci. Rev. Metal/urg. LXXI. 719. 1974. 658. Stein, A.: The prevention of corrosion in galvanized iron water tanks-the results of 12 year field trials. Proceedings of the 5th Annual Conference of the Australasian Corrosion Association. Newcastle. New South Wales. Australia. November 16-20,1964. 659. Chandler. K. A.: Marine and Offshore Corrosion, Butterworth & Co .• London, 1985. 660. Fink. F. W.. and Boyd. W. K.: The corrosion of metals in marine environments, Defense Metals Information Center. DMIC Report No. 245, Columbus, Ohio, May. 1970. 661. Darbin M .. Jailloux, J. M .. and Montuelle, 1.: Durability of reinforced earth structure: The results of a long·term study conducted on galvanized steel. Proc. Inst. Civ. Eng. Part I, 1988, 1029. 662. Khan, D. K .. Mukherjee. K. P., and Banerjee, T.: Correlation of data in laboratory and field studies on sea water corrosion. NML (Natl. Metall. Lab., Jamshedpur, India) Tech. J. 8(4).17-21. 1966. 663. Southwell. C. R .. and Alexander, L. L.: Marine corrosion of cast and wrought non·ferrous metals, National Association of Corrosion Engineers Conference, Houston. 1969, Paper No. 63. 664. Sedzimir. 1. A., and Gmytryk, M. W.: Corrosion of zinc in neutral sodium chloride solutions at constant pH, Proceedings of the 2nd International Congress on Metallic Corrosion. New York, March 11-15, 1963, pp. 972-975. 665. Memmi, M.: Improvement in the corrosion resistance of Zn-Al coatings by addition of magnesium. Proceedings of the 14th International Galvanizing Conference, INTERGALVA '85. Munich. Germany, 1985. pp. 8/1-8/2. 666. Frappe. A. L.. Limare, A .. Piessen. P.. and Vacher, 1. c.: Investigation of the influence of impurities or alloying clements on the atmospheric corrosion of gal vanized steel, Proceedings of the 14th International Galvanizing Conference, INTERGALVA '85. Munich, Germany, 1985. pp. 8/3-8/18. 667. Dreulle. P. Rosset. R .. and Jardy. A.: Hydrophosphate coating, Proceedings of the 12th International Galvanizing Conference, INTERGALVA '79, Paris, May 17-23, 1979, pp. 92-96. 668. Hissel, 1.: Corrosion of galvanized steel related to water composition, Proceedings of the 12th International Galvanizing Conference, INTERGALVA '79. Paris, May 17-23, 1979. pp. 84-91. 669. Rosset, R., Jardy. A., and Parthasarathi, M. N.: Protection of galvanized steel against cold and hot water corrosion by a pyrophosphate coating, Proceedings of the 13th International Galvanizing Conference. INTERGALVA '82, England. 1982.
436
REFERENCES
670. Mihailov, G., and Iovchev, M.: Corrosion studies of galvanised steel and zinc in soft water at various temperatures, Proceedings of the 8th International Congress on Metallic Corrosion. Frankfurt. Germany, 1981,pp.1754-1759. 671. Leclercq, M.: Preweathered zinc for roofing, Proceedings of the 8th International Congress on Metallic Corrosion, Frankfurt, Germany, 1981, pp. 299-303. 672. Abdel Aal, M. S., Abdel Wahab, A A., and EI Saied, A: A study of the inhibiting action of benzene thiols and related compounds on the corrosion of zinc in acidic media, Proceedings of the 8th International Congress on Metallic Corrosion, Frankfurt, Germany, 1981, pp. 1212-1216. 673. Sagiyama, M., Inagaki, 1.-1., and Morita, M.: Fe-Zn alloying behavior and the coating microstructure of galvannealed steel sheets, NKK Technical Report, No. 135, 1991, pp. 49-56. 674. Takimoto, K-I., Suzuki, K.-I., Nishizaka, K, and Ohtsubo, T.: Quantitative analysis of zinc-iron alloy galvanized coatings by glow discharge spectrometry and secondary ion mass spectrometry, Nippon Steel Technical Report, No. 33,1987, pp. 28-35. 675. Hayashi, K, Ito, y', and Miyoshi, Y.: Effect of coating weight and corrosive environment on under-film corrosion of Zn and Zn-Fe alloy coated steel sheets, l. Iron Steellnst. lpn. 78, 127-133. 1992. 676. Odashima, H.: Characteristics of electrolytic chromate films formed on zinc or zinc alloy plated steel sheet in low concentration Cr0 3-Heavy Metal Ions-Halogen Bath, 1. Iron Steellnst. lpn. 78, 121-126, 1992. 677. Saito, T., Wake, R., Oka, J., and Kitayama, M.: Development and properties ofZn-Ni alloy electroplated steel sheet (DURZINKLITE), Nippon Steel Technical Report, No. 25. 1985, pp. 1-10. 678. Hada, T., Kanamaru, T., Nakayama, M., Arai, K, Fujiwara, T., Suemitsu, Y., Sato, M., and Ogawa, Y.: Development and Properties of Zn-Fe Alloy Electroplated Sheet Steel DUREXCELITE, Nippon Steel Technical Report, No. 25,1985, pp. 11-18. 679. Shindo, y', Yoneno, M., Yamada, A., Oka, 1., and Ejima, M.: Development and properties of organic composite-coated steel sheet, Nippon Steel Technical Report, No. 25, 1985, pp. 19-27. 680. Tano, K, and Higuchi, S.: Development and properties of zinc-aluminum alloy coated steel sheet with high corrosion resistance (SUPER ZINC), Nippon Steel Technical Report, No. 25,1985, pp. 29-37. 681. Cabot, P. L., Cortes, M., Centellas, E, and Perez, E.: Potentiostatic passivation of zinc in alkaline solutions, l. Appl. Electrochem. 23, 371-378, 1993. 682. Schikorr, G.: The cathodic behavior of zinc versus iron in hot tap water, Trans. Electrochem. Soc. 76, 247-258, 1939. 683. Cachet, c., and Wiart, R: Zinc deposition and passivated hydrogen evolution in highly acidic sulphate electrolytes, depassivation by nickel impurities, 1. Appl. Electrochem. 20. 1009-1014, 1990. 684. Edney, E. 0., Stiles, D. c., Corse. E. W., Wheeler, M. L., Spence, 1. w., Haynie, E H., and Wilson, W. E.: Controlled field study to determine the impact of dry and wet deposition of air pollutants on the corrosion rate of galvanized steel, National Association of Corrosion Engineers Corrosion '87 Conference, San Francisco, March 9-13, 1987, Paper No. 410. 685. McLarnon, E R, and Cairns, E. J.: The secondary alkaline zinc electrode, l. Electrochem. Soc. 138, 645-664, 1991. 686. Bonilla, C. E: Pipe service tests on Baltimore water-pitting of black and galvanized wrought iron and steel pipe carrying cold water, hot water, and returning condensate, Trans. Electrochem. Soc. 87, 237-254,1945. 687. Aurian-Blajeni, B., and Tomkiewicz, M.: Space charge effects at passive zinc electrodes, 1. Electrochem. Soc. 133, 1766-1769, 1986. 688. Morral, E R: X-ray analysis of corrosion products from galvanized sheets, Trans. Electrochem. Soc.
77n8,279-288.1940.
689. Southwell, C. R, Hummer, C. w., Alexander, A. L, and Forgeson, B. W.: Corrosion of metals in tropical environments. Part VI-galvanic couples, National Association of Corrosion Engineers, 19th Annual Conference, New York, March 11-15, 1963. 690. Alexander, A L., Forgeson, B. Mundt, H. Southwell, C. R, and Thompson, L. J.: Corrosion of Metals in Tropical Environments-Part I-Test Methods Used and Results Obtained for Pure Metals and a Structural Steel. Naval Research Laboratory, Report No. 4929, U.S. Dept. of Commerce, Washington, D.C., June 19,1957. 691. Kweon, Y. G., and Coddet, C.: Behavior in seawater of zinc-base coatings on aluminum alloy 5086, Corrosion 48, 97-102, 1992.
w.,
w.,
REFERENCES
437
692. Suzuki, I., and Enjuzi, M.: The development of the corrosion resistance of an Fe-Zn alloy coating on the basis of the behaviour of the corrosion product, Corms. Sci. 26. 349-355, 1986. 693. Stern, M.: A method for determining corrosion rates from linear polarization data, Corrosion 14, 440t-444t, 1958. 694. Stern, M., and Geary, A. L.: Electrochemical polarization-Part I. A theoretical analysis of the shape of polarization curves, .I. Eleclrochelll. Soc. 104. 56-63, 1957. 695. Kautek, W.: The galvanic corrosion of steel coatings: Aluminum in comparison to cadmium and zinc, Corros. Sci. 28, 173-199. 1988. 696. Campbell, H. S.: Sprayed aluminium and zinc protect against corrosion by aqueous solutions. Anti-Corrosioll 1982(August), 10-12. 697. Rangel, C Moo De Damborenea, 1., De Sa. A. I.. and Simplicio. M. H.: Zinc and polyphosphates as COlTosion inhibitors for zinc in ncar neutral waters. Br. Corros . .I. 27, 207-212, 1992. 698. De Pauli, C Poo Giordano. M. C, and Mishima. H. T.: Zinc dissolution and passivation in buffered phosphate solutions. Part II. About the nature of the passivating film,.!. Eleclmallal. Chelll. 103,95-102, 1979. 699. Belisle. S .. and Dufresne, R.: Performance of galvanized steel in municipal wastewater treatment plants. Maler. Perform. 31(6). 64-67. 1992. 700. Deslouis. C, Duprat. Moo and Tourni II on, C: The kinetics of zinc dissolution in aerated sodium sulphate solutions. A measurement of the corrosion rate by impedance techniques. Corms. Sci. 29. 13-30. 1989. 701. Fedrizzi. L.. Ciaghi, L., Bonora, P. L.. Fratesi, Roo and Roventi, G.: Corrosion behaviour of electrogalvanized steel in sodium chloride and ammonium sulphate solutions, a study by E.I.Soo J. Appl. Eleclrochelll. 22.247-254.1992. 702. Cachet, C. De Pauli, C P, and Wiart, R.: The passivation of zinc in slightly alkaline solutions. Corms. Sci. 25.493-502, 1985. 703. Abdel-Aal, M. S., Ahmed, Z. A., and Hassan. M. S.: Inhibiting and accelerating effects of some quinolines on the corrosion of zinc and some binary zinc alloys in HCI solution, .J. AI)I)I. fJeClmchcm. 22,1104-1109,1992. 704. Rangel, C Moo and Cruz, L. E: Zinc dissolution in Lisbon tap water. Corros. Sci. 33.1479-1493, 1992. 705. Tsuru, Y, Takenouchi, K.. and Hosokawa. K.: Scanning micro-electrode study of zinc corrosion in hydrochloric acid solution. Asahi Glass Foundation for Industrial Technology, Report, Vol. 57. pp. 193-199, 19901 in Japanese I. 706. Chiu, S. L.. and Selman, 1. R.: Determination of electrode kinetics by corrosion potential measurements: Zinc corrosion by bromine . .I. Apl'l. Eleclrochem. 22, 28-37, 1992. 707. Grubitsch, H., and !ilL 0.: The hot water corrosion of zinc. Part II. Korros. Melallsclillr: 16. 197-203, 1940 1in German I. 708. Bianchi, G.: Corrosion of zinc by differential aeration, Corrosion 14, 245t-248t, 1958. 709. Gilbert, P. T.: An investigation into the corrosion of zinc and zinc-coated steel in hot waters, Sheer Mel. Ind. 1948( October-December). 710. Lorking. K. E: The corrosion of zinc, Department of Supply, Australian Defence Scientific Service, Metallurgy Report No. 67, Melbourne, Australia, May 1967. 711. Pryor. M. J., and Keir, D. S.: Galvanic con-osion,.I. Eleclrochem. Soc. 104.269-275, 1957. 712. Shatfer, R. 1., and Belisle, S.: An assessment of the performance of coated sheet steels in vehicle and atmospheric environments, NatIOnal Association of Corrosion Engineers Corrosion '91 Conference, Cincinnati. Ohio, March 11-15. 1991. Paper No. 411. 713. Subramanian, G .. Ananth, v., Chandrasekaran, P, Pairaj, S., and Guruviah, S.: Con-os ion and surface modifications of zinc at mandapam, Bull. Eleclrochem. 6, 494-496, 1990. 714. Arimura, M .. I wai, M., Sakai, H., and Nishimoto. H.: Field test of Zn-Ni alloy electroplated steel sheet formed into automotive door in Okinawa for 5 years, National Association of Corrosion Engineers Con'osion '91 Conference, Cincinnati, Ohio. March 11-15. 1991. Paper No. 370. 715. Fedrizzi, L., and Bonora, P. L.: Effects of mechanical deformation on electrochemical behaviour of galvanised steel, Br. Corms . .I. 28, 37-42. 1993. 716. Zheng. Y, Wang. Z., and Zhang, E: Influence of acid deposition on atmospheric corrosion of linc, Proceedings of the 7th APCCC Conference on Corrosion Control, Beijing China. 1991, Vol. I. pp. 487-490.
438
REFERENCES
717. Nagy, Z., and Wesson. 1. M.: Error analysis of the polarization-resistance technique for corrosion-rate measurements.J. Electrochem. Soc. 139. 1261-1266. 1992. 718. Mansfeld. E: Polarization resistance measurements-today's status. Electrochemical Techniquesfor Corrosion Engineering. pp. 67-71, National Association of Corrosion Engineers. Houston. 1986. 719. Silverman, D. Primer on the AC impedance technique. in Electrochemical Techniquesfor Corrosion Engineering. pp. 73-79. National Association of Corrosion Engineers. Houston. 1986. 720. Nielsen, K.. and Yding. E: Influence of pipe quality on corrosion of galvanized steel pipes for domestic water supply, Werkst. Karras. 34, 547-556,1983. 721. Abayarathna. D.: Corrosion and polarization characteristics of single crystal and polycrystalline zinc surfaces. Ph.D. Thesis. University of Missouri-Rolla, 1989. 722. Roberts, C. W.: An investigation into the causes of intercrystalline corrosion in zinc-aluminium alloys. Metallurgia. 1961(9).57-66. 723. von Riecke. E.: Investigation on the influence of zinc on the corrosion behavior of high strength steels. Werkst. Korros. 30, 619-630.1979. 724. Chen. C. L, Lee. P. Y. Wu. 1. K.. Chiou. D. 1.. Chu. C. y, and Lin, 1. Y: The use of zinc and tin coatings and chemical additives for preventing hydrogen embrittlement in steel. Carras. Prevo Control 40(3), 71-74.1993. 725. Biber, H. E.: Scanning auger microprobe study of hot-dipped regular-spangle galvanized steel. Part I. Surface composition of as-produced sheet. Metall. Trans. 19A. 1603-1608, 1988. 726. Ross, T. K., and Wolstenholme. 1.: Anti-corrosion properties of zinc dust paints. Carras. Sci. 17, 341-351.1977. 727. Johnson. B. V.• and Ross, T. K.: The protection of mild steel by zinc-rich paint in flowing aerated 0.5 M NaCI solution. Part II. The effect of exposed area ratio, Corros. Sci. 18.511-518. 1978. 728. Chua, H. H., Johnson, B. V., and Ross, T. K.: The protection of mild steel by zinc-rich paint in flowing aerated 0.5 M NaCI solutions. Part I. The effect of zinc particle size. Carras. Sci. 18, 505-510, 1978. 729. Chua, H. H., and Ross, T. K.: The protection of mild steel by zinc-rich paint in flowing aerated 0.5 M NaCI solution. Part IV. The effect of zinc oxide addition, Corros. Sci. 18, 523-525, 1978. 730. Yakimenko, G. Ya .. Andryushchenko. E K., Kharchenko. E. P.. and Valenya, E. P.: Colored molybdate films on zinc and their protective properties. Zashch. Met. 13. 123-125, 1977. 731. Pedram, R., and Ross, T. K.: The protection of mild steel by zinc-rich paint in flowing aerated 0.5 M NaCI solution. Part III. The effect of zinc content, Carras. Sci. 18, 519-522, 1978. 732. Vannerberg, N.-G., and Sydberger, T.: Reaction between S02 and wet metal surfaces, Corros. Sci. 10, 43-49, 1970. 733. Strekalov, P. v.. Agafonov, V. V., and Mikhailovskii. Yu. N.: Effect of temperature on the adsorption of moisture and the corrosion rate of zinc under atmospheric conditions. Zashch. Met. 8(5).577-580. 1972. 734. Miki. K.. Shimogori, K., Satoh, H., Ikeda, K., M, and Terada.: Corrosion behaviour of cold rolled and zinc alloy plated steel sheets in various kinds of corrosion tests, Tetsu To Hagane 72, 1090-1097. 1986. 735. Fukuda, Y., Tsuchiya. Y. Terasaka. M., Nakaoka. K., and Hara. T.: Analysis of corrosion layers on electrodeposited Zn-Ni (I3wt%) alloys using AES, SAM, XPS, SEM and XRD, Tetsu Ta Hagane 72, 1782-1789,1986. 736. Mattei, I. V., Martorano, R.. and Johnson. E. A.: Formulating stable latex paints with zinc oxide, Jo. Coatings Techno!. 63. 39-45. 1991. 737. Kenworthy. L.: The problem of copper and galvanized iron in the same water system. J. Inst. Met. 69, 67-90. 1943. 738. Tanabe. H .. Shinohara. T.• Hoshino, M., Itami, K.. and Sato, Y.: Effect of dissolved oxygen on under-film corrosion. Boshoku Gijutsu 30. 622-626. 1981. 739. Westberg, J., and Borjesson, L-G.: Influence of sheet metal surface conditions on the corrosion properties of automotive paint systems, National Association of Corrosion Engineers Corrosion '80, Chicago, March 3-7, 1980, Paper No. 278. 740. Jordan, D. L.: Galvanic interactions between corrosion products and their bare metal precursors: A contribution to the theory of underfilm corrosion, Proceedings of the Symposium on Advances in Corrosion Protection by Organic Coatings, 1989, pp. 30-43. 741. Kurokawa, S., Ban, T., Yamato, K., and Ichida, T.: A study on cosmetic and perforation corrosion test for automotive steel sheets, J. Iron Steellnst. Jpn. 72, 1111-1118, 1986.
c.:
REFERENCES
439
742. Shibuya, A .. Kurimoto, T.. Korekawa, K., and NojI, K.: Corrosion-resistance of electroplated Ni-Zn alloy steel sheet, Tetsu To Hagane 66. 771-778, 1980. 743. Tanabe, H .. Shinohara. T.. Sato. Y. Nii. H .. Hoshino, M., and Itami, K.: Effect of zinc rich paints on the durability of epoxy top coats in sodium chloride solution. Boshoku Gijutsu 30, 390-395, 1981. 744. Ramasamy, N .. Sarangapani, S., Nityanandan, J. P, Venkatesan, S .. and Aravamuthan. V: Gas evolution studies with powder compact Zn anodes, Corros. Sci. 11, 873-884, 1971. 745. Moreno, E .. and Sagues, A. A.: Performance of plain and galvanized reinforcing steel during the initial stage of corrosion in concrete with pozzolanic addition, National Association of Corrosion Engineers. Corrosion '96, Paper No. 326, 1996. 746. Vernon, W. H. J.: Second experimental report to the Atmospheric Corrosion Research Committee (British Non-FelTous Metals Research Association), TrailS. Faraday Soc. 23, 113-204, 1927. 747. Vernon, W. H. J.: A laboratory study of the atmospheric corrosion of metals. Part II. Iron: The primary oxide film. Part III. The secondary product or rust (influence of sulphur dioxide, carbon dioxide and suspended particles on the rusting of iron), Trans. Faradav Soc. 31, 1668-1700, 1935. 748. Vernon, W. H. J.: A laboratory study of the atmospheric corrosion of metals. Part I. The corrosion of copper in certain synthetic atmospheres, with particular reference to the influence of sulphur dioxide in air of various relative humidities, Trans. Faraday Soc. 27,255-277, 1931. 749. Heitz, E., Hukovic, M., and Maier. K. H.: Basic process of corrosion in organic solvents II, Werkst. Korros. 21. 457-462,1970 lin Germani· 750. Fujita, S., Shimizu, Y, and Matsushima, I., Corrosion resistance of painted zinc plated steel in sodium solution, Tetsu To Hagane 72, 1897-1902, 1986. 751. Wagner, C: Contribution to the theory of cathodic protection, l. Electrochem. Soc. 99, 1-12, 1952. 752. Myers, J. R., and Obrecht. M. E: Potable water systems-recognition of cause vital in minimizing cotTosion, Mater. Prot. Perfimn. 11(4),41-46, 1972. 753. Fedot'ev, N. P.. Grilikhes, S. Ya., and Orekhova, E. E: Effect of light on the corrosion of zinc and cadmium, Zh. Prikl. Khilll. 32,1428-1430,1959. 754. Miyoshi, Y: Evaluation technology of corrosion behavior for automotive steel sheet, ISIJ Int. 31(2), 122-133,199L 755. LaQue. E L., and Boylan, 1. A.: Effect of composition of steel on the performance of organic coatings in atmospheric exposure, Corrosion 9(8), 237-241,1953. 756. Ikeda, K., and Satoh, H.: An analysis of corrosion cracking phenomena of electrode posited Zn-Ni alloy layers, J.lron Steel Inst. lpn. 77,1162-1168,1991. 757. Arsalane, M .. and Tosser, A.: Interband photoconduction edge in zinc sulphide thin films, NOliI'. Rev. Opt. 7(5), 323-327, 1976 lin French]. 758. Suzuki. N .. Bando, S .. and Sugisawa. S.: Phosphotability of Zn-Ni alloy electroplated steel sheets, l. lroll Steellnsf. lpn. 77, 1058-1065, 1991. 759. Margulies. P. H.: Peroxygen compounds hold important place in treating metal surfaces, lrollAge 175(4), 71-74,1955. 760. Mukherjee, D., Venkataraman, B., Palaniswamy, N., Natesan, M., and Balakrishnan, K.: COlTosion and oxidation resistance of zinc composite materials, The International Congress on Metallic Corrosion, Madras, India, 1987, Vol. 4, Sessions 14-19, pp. 3689-3694. 761. Fernandez, A., del Carmen Leiro, M., Rosales, B. M., Ayllon, E. S., Varela, E E., Gervasi, C A .. and Vikhe, J. R.: Techniques applied to the analysis of the atmospheric corrosion of low carbon steel, zinc, copper and aluminum, in Proceedings of the 12th International Corrosion Congress on Corrosion Control for Low-Cost Reliability, September 19-24, 1993, Vol. 2, pp. 574-589, 1993. National Association of Corrosion Engineers, Houston. 762. Wakano, S., Shibuya, A., Hobo, Y, and Nishihara, M.: Corrosion performance of Zn alloy precoated steels for automotive body, Trans. Iron Steel In.51. lpn. 23, 967-973, 1983. 763. Cachet, C, Stroder, U., and Wi art, R.: The kinetics of zinc electrode in alkaline zincate electrolytes, Electrochim. Acta 27, 903-908, 1982. 764. Shibuya, A., Kurimoto, T., Hoboh, Y. and Usuki, N.: Development of Ni-Zn alloy plated steel sheet, Trans. lroll Steellnst. ll'". 23,923-929, 1983. 765. Svensson, 1.-E., and Johansson, L.-G.: A laboratory study of the initial stages of the atmospheric corrosion of zinc in the presence of NaCI, influence of S02 and N0 2 , Corros. Sci. 34,721-740, 1993.
440
REFERENCES
766. Miura, N., Saito, T., Kanamaru, T., Shindo, Y., and Kitazawa, Y.: Development of new corrosion-resistant steel sheets for automobiles, Trans. Iron Steel Inst. lpn. 23,913-922, 1983. 767. Standish, J. v., Whelan, G. w., and Roberts, T. R: The corrosion behavior of galvanized and cold rolled steels, SAE Technical Paper No. 831810, Society of Automotive Engineers International Congress and Exposition, Detroit, February 28-March 4, 1983. 768. Yamashita, M., Kubota, T., and Adaniya, T.: Organic-silicate composite coated steel sheet for automobile body panel, SAE Technical Paper No. 8620 17. Society of Automotive Engineers International Congress and Exposition, Detroit, February 24-28, 1986. 769. Lambert, M. R, and Hart, R G.: Corrosion resistance of 0-15% Ni-Zn alloy electroplated coatings, SAE Technical Paper No. 860266, Society of Automotive Engineers International Congress and Exposition, Detroit, February 24-28, 1986. 770. Tsugawa, S.-I., Mohri, T., Kobayashi, S., and Ichida, T.: Newly-developed organic composite-coated steel sheet with high perforation resistance, SAE Technical Paper No. 850004, Society of Automotive Engineers International Congress and Exposition, Detroit, February 25-March I, 1985. 771. Watanabe, T., Ohmura, M., Honma, T., and Adaniya, T.: Iron-zinc alloy electroplated steel for automotive body panels, SAE Technical Paper No. 820424, Society of Automotive Engineers International Congress and Exposition, Detroit, February 22-26, 1982. 772. Miyoshi, Y., Kitayama, M., Nishimura, K, and Naito, S.: Cosmetic corrosion mechanism of zinc and zinc alloy coated steel sheet for automobiles, SAE Technical Paper No. 850007, Society of Automotive Engineers International Congress and Exposition, Detroit, February 25-March I, 1985. 773. Bassi, G., and Hugo, J. P.: Etch pits in zinc, 1. Inst. Met. 87, 376-379, 1958-1959. 774. Obrecht, M. E: Potable water systems in buildings: Deposit and corrosion problems, Heat.lPipinglAir Condo 1973(May), 77-83. 775. Khairy, E. M., and Khater, M. M.: Contribution to the electrochemical behaviour of zinc electrode in pyridine media, Corros. Sci. 12, 121-132, 1972. 776. Shams EI Din, A. M., and Khedr, M. G. A.: Role of anions in the corrosion and corrosion inhibition of Zn in aqueous solutions. Part II. Behaviour in deaerated solutions and the determination of critical currents for passivation, Corros. Sci. 12, 393-404, 1972. 777. Grauer, R, and Kaesche, H.: Electronmicrospic and electrochemical investigations on the passivation of zinc in O.IM NaOH, Corros. Sci. 12,617-624, 1972. 778. Myrdal, R.: Phenomena that disturb the measurement of potentials in concrete, National Association of Corrosion Engineers, Corrosion '96, Paper No. 339, 1996. 779. Williams, L. E G.: The initiation of corrosion of chromated zinc electroplate on steel, Corros. Sci. 13, 865-868, 1973. 780. Theiler, E: The rust preventing mechanism of zinc dust paints, Corros. Sci. 14,405-414,1974. 781. Dirkse, T. P.: The solubility product constant ofZnO, 1. Electrochem. Soc. 133, 1656-1657, 1986. 782. Caswell, P., Hampson, N. A., and Larkin, D.: The differential capacitance of zinc in aqueous solution-a further note, 1. Electroanal. Chem. 20, 335-338,1969. 783. Cescon, P., Marassi, R., Bartocci, v., and Fiorani, M.: Standard electrode potentials of silver, cadmium, cobalt, indium, titanium, zinc in molten alkali thiocyanates, 1. Electroanal. Chem. 23, 255-259, 1969. 784. Awad, S. A., and Kamel, Kh. M.: Mechanism of corrosion-inhibition and corrosion-promotion of zinc by phosphate ions. 1. Electroanal. Chem. 24,217-225,1970. 785. Hampson. N. A., Herdman, G. A., and Taylor, R: Some kinetic and thermodynamic studies of the system Zn, OH, 1. Electroanal. Chem. 25. 9-18, 1970. 786. Armstrong, R D., and Bell, M. E: The active dissolution ofzinc in alkaline solution, Electroanal. Chem. Interfacial Electrochem. 55, 201-211,1974. 787. Drury, J. S., Hampson, N. A., and Marshall, A.: The anodic behaviour of zinc in silicate-containing alkaline solutions, Electroanal. Chem. 1I1Ierfaciai Electrochem. 50, 292-294, 1974. 788. Bonasewicz, P.. Hirschwald, W., and Neumann, G.: Intluence of surface processes on electrical photochemical and thermodynamical properties of zinc oxide films, 1. Electrochem. Soc. 133, 22702278, 1986. 789. Hurien, T., and Fischer, K P.: Kinetics of zinc reactions in acidified solutions of potassium chloride, Electroanal. Chem. Interfacial Electrochem. 61, 165-173, 1975. 790. Muralidharan, V. S., and Rajagopalan, K S.: Kinetics and mechanism of corrosion of zinc in sodium hydroxide solutions by steady-state and transient methods, 1. Electroanal. Chem. 94, 21-36,1978.
REFERENCES
441
791. Lemasson. P., and Gautron. J.: On the anodic decomposition of the zinc selenide electrode. l. Electmanal. Chern. 119,289-299. 1981. 792. Masuda. H., Morishita, S .• Fujishima. A., and Honda, K.: Study of photoelectrochemical reactions at a n-type polycrystalline ZnO electrode using photoacoustic detection, 1. Electmanal. Chern. 121. 363368. 1981. 793. Gerischer, H.: The role of semiconductor structure and surface properties in photoelectrochemical processes. l. Electmanal. Chern. 150,553-569. 1983. 794. Cabot, P. L.. Cortes, M .. Centellas. F. A., Garrido. 1. A., and Perez, E.: Potentiodynamic passivation of zinc in aqueous KOH solutions, l. Electmanal. Chern. 201. 85-100,1986. 795. Hamnett. A .• and Mortimer. R. 1.: The electrochemical oxidation of zinc studied by ellipsometry. 1. Electroanal. Chern. 234.185-192.1987. 796. Novak. M .. and Szucs.: Study of electropolishing of a Zn anode in acidic medium. l. Electroanal. Chern. 266. 157-172. 1989. 797. Wroblowa. H. S .. and Qaderi. S. 8.: The mechanism of oxygen reduction on zinc. l. Electroanal. Chern. 295.153-161.1990. 798. Kouyate, D .. Ronfard-Haret. J.-c.. and Kossanyi. 1.: Electroluminescent properties of polycrystalline zinc oxide electrodes doped with erbium ions, 1. Electroanal. Chern. 319, 145-160, 1991. 799. Cachet. c.. and Wiart, R: The kinetics of zinc dissolution in chloride electrolytes: Impedance measurements and electrode morphology. l. Electroanal. Chern. 111,235-246. 1980. 800. De Pauli. C .. Derosa, O. A. H.. and Giordano, M. C.: Effect of phosphate ions on zinc dissolution in alkaline solutions, l. Electroanal. Chern. 73. 105-108, 1976. 801. Elsdale. R N., Hampson. N. A., Jones, P. c.. and Strachan, A. N.: The anodic behaviour of porous zinc electrodes, l. Appl. Electrochern. 1, 213-217,1971. 802. Huber, K.: Anodic formation of coatings on magnesium, zinc, and cadmium, 1. Electrochern. Soc. 100. 376-382.1953. 803. Bruno. R, and Memmi, M.: The development of new galvanized coatings for use in aggressive environments, Proceedings of the II th International Galvanizing Conference. INTERGALVA '76. Madrid, pp. 213-221. 804. Steinsmo. U .. and Bardal, E.: Use of DC electrochemical methods for evaluation of paint films on steel, aluminum and zinc, Corrosion 48, 910-917, 1992. 80S. Armas, R A., Gervasi, C. A., Di Sarli. A .• Real, S. G .. and Vilche. J. R.: Zinc-rich paints on steels in artificial seawater by electrochemical impedance spectroscopy, Corrosion 48. 379-383, 1992. 806. Eger, D., Many, A.. and Goldstein. Y.: Very strong accumulation layers on ZnO surfaces, Phys. Lett. 55A(3). 197-198, 1975. 807. Cunningham, J., and Zainal, H.: Reactions involving electron transfer at semiconductor surfaces. Part IV. Zinc oxide-promoted photoreductions in aqueous solutions at neutral pH. l. Phys. Chern. 76. 2362-2374, 1972. 808. Edwards. 1. D.: Anodic and surface conversion coatings on metals, Trans. Electrochern. Soc. 81/82, 341-357, 1942. 809. Ohsawa, S., and Takeda. M.: Studies of inhibition effect of crown ether on Zn-4% Al alloy. lbaraki Daigaku Kogakubu Kenkyu Shuh. 37. 37-45, 1989. 810. Baugh, L. M., and Lee. J. A.: Differential capacitance of polycrystalline zinc. Part I. Effect of pH and ammonium ions, Electroana/. Chern. Interfacial Electrochern. 48,55-61, 1973. 81 I. Askey. A .. Lyon, S. 8., Thompson, G. E., Johnson, J. 8.. Wood, G. c., Cooke. M .• and Sage, P.: The corrosion of iron and zinc by atmospheric hydrogen chloride. Corros. Sci. 34, 233-247, 1993. 812. Giridhar. J., and van Ooij, W. J.: Study of Zn-Ni and Zn-Co alloy coatings electrodeposited on steel strips. Part II. Corrosion, dezincification and sulfidation of the alloy coatings, Surf. Coatings Techno/. 53.35-47, 1992. 813. Fragata. F. L., and Mussoi. C. R S.: Influence of extender pigments on the performance of ethyl silicate zinc-rich paints, l. Coatings Technol. 65, 103-109, 1993. 814. Miyoshi. Y.. Oka, 1., and Maeda, S.: Fundamental research on corrosion resistance of precoated steel sheets for automobiles, Trans. Iron Steellnst. lpn. 23, 974-983. 1983. 815. Shastry. C. R., and Townsend, H. E.: Effect of pretreatment on performance of zinc and zinc alloy electroplated sheet in a cyclic corrosion test. SAE Technical Paper No. 892556. Society of Automotive
442
816.
817.
818. 819. 820. 821. 822. 823.
824. 825. 826.
827.
828.
829. 830. 831. 832.
REFERENCES Engineers 1989 Automotive Corrosion & Prevention Conference & Exposition, Dearborn, Michigan, December 4-6, 1989. May, T. P., and Alexander, A. L.: Spray testing with natural and synthetic sea water. Part I. Corrosion characteristics in the testing of metals, American Society for Testi ng and Materials 53rd Annual Meeting, June 26-30,1950, pp. 1131-1141. Cunningham, J., and Corkery, S.: Reactions involving electron transfer at semiconductor sUlfaces. Part VI. Electron spin resonance studies on dark and illuminated aqueous suspensions of zinc oxides, 1. Phys. Chem. 79,933-941,1975. Shaw, B. A., and Moran, P. J.: Characterization of the corrosion behavior of zinc-Aluminum thermal spray coatings, Corrosion '85, Boston, March 25-29,1985, Paper No. 212. Anicai, L., Siteavu, M., and Grunwald, E.: Corrosion behaviour of zinc and zinc alloy depositions, Carras. Prevo Control 1992(August), 89-93. Townsend, H. E.: Effects of zinc coatings on the stress corrosion cracking and hydrogen embrittlement of low-alloy steel, Metall. Trans. 6A, 877-883, 1975. Riedl, G.: Anodes-dissolution and change of shape, Mater. Perform. 20( 12), 34-40. 1981. Montes, E.: Influence of particle size distribution of zinc dust in water-based, inorganic, zinc-rich coatings, 1. Coatings Techno!. 65, 79-82, 1993. Sagiyama, M .. Kawabe, M., and Watanabe, T.: Effects of electrolysis and bath conditions on the surface roughness. morphology and crystal orientation of zinc electrodeposit. Tetsu To Hagane 76. 1301-1308, 1990. Petrovich, I.. and Galvan. J. c.: The application of electrochemical techniques to the study of duplex systems. Galvanotecnica 36(2).29-34, 1985. Sehgal, A.. Li, D .• Kho, Y. T. Osseo-Asare, K, and Pickering, H. w.: Reproducibility of polarization resistance measurements in steel-in-concrete systems. Corrosion 48. 706-714. 1992. Edney, E. 0., and Stiles. D. c.: Laboratory investigations of the impact of dry deposition of S02 and wet deposition of acidic species on the atmospheric corrosion of galvanized steel, Atmos. Environ. 20, 541-548, 1986. Kurokawa, S .. Yamato. K, and Ichida, T.: A study on cosmetic and perforation corrosion test procedures for automotive steel sheets, National Association of Corrosion Engineers Corrosion '91 Conference, Cincinnati, Ohio, March 11-15, 1991, Paper No. 396. Koda, M., Uchida, y', Suzuki, M., Deguchi, T, and Hirose, Y.: Inhibition mechanism of black patina development on a hot dip Zn-AI alloy coated steel sheet by spraying aqueous solution of Co salt, Tetsu To Hagane 76,383-390, 1990. Grunberg, L.: The formation of hydrogen peroxide on fresh metal surfaces, Proc. Phys. Soc. 866, 153-161,1953. Kaneko, F., and Nagasaka, H.: Evaluation of corrosion resistance of metal sprayed and painted steel, Surf. Eng. 6, 125-128, 1990. Chance, D. A., and Nobe, K: Polarization and photopotentials of zinc single crystals in Na2S, Corrosion 28, 170-171, 1972. Satoh, H., Shimogori, K., Nishimoto, H .. Miki, K, Ikeda, K, Iwai, M., Sakai. H., and Nomura, S.: Analysis of perforation corrosion of cold rolled and galvanized steel sheets by extreme-value statistics,
Tetsu To Hagane 72, 1098-1105, 1986. 833. Danilov, F. I., Sukhomlin. D. A., Gerasimov. V. v., and Popovich, V. A.: Electrodeposition of zincmanganese alloys, Sov. Electrochem. 28, 217-220. 1992. 834. Zhdanov, B. B., and Sukharev, N. P.: Corrosion-electrochemical behavior of alloys having a eutectic structure: Zn-Cd, Zn-Pb and Cd-Pb, Zh. Prikl. Khim. 62. 2361-2363, 1989. 835. Hope, G. A .• and Ritchie. I. M.: The effect of gold on the oxidation of zinc in moist oxygen. Corros. Sci. 16. 177-181, 1976. 836. Dorfman, M .• Vargas, J., Clayton, C. R .. and Herman, H.: Cavitation behavior of active thermal-sprayed coatings. Thin Solid Films 64, 351-357, 1979. 837. Hanneman. R. E., and Aust. K T: Solute clustering and intergranular corrosion, Metallurgica 2, 235-238, 1968. 838. Hayashi, K, Ito, Y., Kato, C, and Miyoshi. Y.: Under-film corrosion behavior of zinc and zinc alloy coated steel sheets for automobiles, Tetsu To Hagane 76. 1317-1324, 1990.
REFERENCES
443
839. Johnson. W. C: French process zinc oxide in primers for hand-cleaned structural steel, J. Coatings Technol. 57,67-75,1985. 840. Krishnan. R. M., and Muralidharan. V. S.: Electrochemical and salt spray testing of zinc electrodeposits, Br. Corros. l. 27. 317-318.1992. 841. Takazawa. H .. Saito, H .. and Nakai, H.: Corrosion resistance of hot-dipped Zn-AI alloy coatings in coastal areas. Zaino To Kankyo 40(7), 470-477.1991. 842. Lechner-Knoblauch. U., and Heitz, E.: Corrosion of zinc, copper and iron in contaminated non-aqueous alcohols. Electrochirn. Acla 32. 901-907,1987. 843. Cachet. C, De Pauli, C P. and Wiart. R.: The pore texture of zinc electrodes corroded in acidic electrolytes, Electrochim. Acta 30. 719-723. 1985. 844. Garrels. R. M.: Mineral species as functions of pH and oxidation-reduction potentials. with special reference to the zone of oxidation and secondary enrichment of sulphide ore deposits. Geochim. Cosrnochim. Acta 5, 153-168, 1954. 845. Kuffel, K.: Comparative testing of AI and Zn hot-dip coatings in a fertilizer environment. Ochr. Korol. 29(5).116-117.1986 [in Polishl. 846. Hayashi. K .. Ito. Y. Kato. C. and Miyoshi. Y: Electrochemical study on under-film corrosion mechanism ofZn and Zn-Fe alloy coatings, J.lron Steeilnsl. lpn77. 1122-1129, 1991. 847. Fruhwirth, 0 .• Herzog, G. W., and Poulios, 1.: ZnO photoeffects under galvanostatic and potentiostatic conditions. Suif. Technol. 11.259-267. 1980. 848. Lin. C, Wen. L., and Qian. S.: Corrosion failure analysis of flame sprayed Zn-A115 coating in sea water, Acla Metal!. Sin. 26(2), B 110-BI13. 1990. 849. Cachet. C. and Wiart. R.: Reaction mechanism for zinc dissolution in chloride electrolytes, l.
Electroallal. Chem. 129, 103-114. 1981. 850. Ito. Y. Hayashi. K., and Miyoshi, Y: The influence of drying process in cyclic corrosion test on the underfilm corrosion of galvanized steel sheets. J. Iron Steeilnsl. lpn. 77. 1138-1145, 1991. 851. De Pauli, C P.• Derosa, O. A. H., and Giordano. M. C: Zinc dissolution and passivation in buffered phosphate solutions. Part I. A comparative study with sodium hydroxide solutions. l. Electmwwl. Chem. 86, 335-348, 1978. 852. EI-Sharif. M. R .. Suo Y 1., Watson, A., and Chisholm, CU.: Studies of chemical conversion treatments of electrodeposited zinc-chromium and zinc-nickel-chromium alloys, in Proceedings of the 121h International Corrosion Congress on Corrosion Control for Low-Cost Reliabilitv, Septemher 19-24, 1993, Vol. I, pp. 329-336, National Association of Corrosion Engineers, Houston, 1993. 853. Sarmaitis, R., andJuzeliunas, E.: Protective action of chromate conversion coatings, Plaling Surf: Finish. 78(1). 72-76, 1991. 854. Abdel Haleem, S. M., and Abdel Fattah, A. A.: The role of some organic anions in promoting or inhibiting the corrosion of zinc, Suif. Coatings Technol. 29,41-49, 1986. 855. Fratesi, R., and Roventi, G.: Electrodeposition of zinc alloys in chloride baths containing cobalt ions. Mater. Chern. Phys. 23, 529-540, 1989. 856. Arshadi, M. A., Johnson, J. B., and Wood, G. C: The influence of an isobutane-S02 pollutant system on the earlier stages of the atmospheric corrosion of metals, Corros. Sci. 23, 763-776, 1983. 857. Sakurai, S., and Naruse, T.: Conversion treatment on zinc using titanium chelate, l. Met. Finish. Soc. lpn. 37, 674-676,1986. 858. Ohsawa. S., Takeda, M., and Yamamura, T.: Study on corrosion inhibition of zinc-aluminum alloy by organic compounds. J. Met. Finish. Soc. lpn. 39, 310-314. 1988. 859. Nikolova. M .. Raichevski. G .. Banova, R.. Rashkov, S. T., and Klaus. M.: Corrosion hehaviour of electrodeposited zinc-cobalt alloy coatings, Bull. Electrochern. 5, 314-318, 1989. 860. Addison, W. E.: Slruclural Principles in Inorganic Compounds, John Wiley & Sons, New York. \953, p.57. 861. Almeida. E., Pereira, D .. and Figueiredo, 0.: The degradation of zinc coatings in salty atmospheres, Prog. Org. Coat. 17, 175-189, 1989. 862. Fukuda, Y, and Fukushima, T.: Electrodcposition of nickel and zinc into the pores of anodic oxide film on aluminum, Trans. NaIl. Res.lnsl. Mel. 26,315-321,1984. 863. Hauffe, K., and Range, J.: On the anodic dissolution of zinc-oxide single crystals in the presence of light, Phys. Chern. 71, 690-697, 1967.
444
REFERENCES
864. Sydberger, T, and Vannerberg, N.-G.: The influence of the relative humidity and corrosion products on the adsorption of sulfur dioxide on metal surfaces, Corros. Sci. 12, 775-784, 1972. 865. Zeller, R. L., and Savinell, R. E: Interpretation of a.c. impedance response of chromated electrogalvanized steel, Corros. Sci. 26, 389-399,1986. 866. Wilcox, G. D., Gabe, D. R., and Warwick, M. E.: The development of passivation coatings by cathodic reduction in sodium molybdate solutions, Corros. Sci. 28, 577-587,1988. 867. Felloni, L., Fratesi, R., Roventi, G., and Fedrizzi, L.: Assessment of the corrosion resistance of Zn-Ni electrodeposits, Proceedings of the 11th International Corrosion Congress, Florence, Italy, April 2-6, 1990, Vol. 2, pp. 365-372. 868. Biestek, T, Drys, M., Sokolov, N., Knotkova, D., Ramishvili, R., Kozhukharov, v., and Zeldel, M.: Atmospheric corrosion of metallic systems. Part V. Identification of the chemical compounds in the corrosion products of zinc, Zashch. Met. 19(5), 750-753, 1983. 869. Hayashi, K, Ito, Y, Kato, c., and Miyoshi, Y: Electrochemical study on under-film corrosion mechanism of Zn and Zn-Fe alloy coatings, Tetsu To Hagane 77, 1122-1129, 1991. 870. Calvert, J. G., Lazrus, A., Kok, G. L., Heikes, B. G., Walega, J. Goo Lind, 1., and Cantrell, C. A: Chemical mechanisms of acid generation in the troposphere, Nature (London) 317, 27-35, 1985. 871. Alimov, V. I., Berezin, A. v., Marchuk, S. I., and Lavrent'eva, G. A.: Corrosional resistance of zinc coating on steel after annealing, fzv. Akad. Nauk SSSR Met. 1988(3), pp. 160-163. 872. Ohsawa, S., Takeda, M., and Yoshinuma, K: Studies on corrosion inhibition of zinc single crystals by organic inhibitors, fbaraki Daigaku Kogakubu Kenkyu Shuho, 36(12),141-148, 1988. 873. Aoki, T, Sumiya, 1., Uchida, y, and Hirose, Y: Hydrogen Permeation Behavior in the Coating Layers ofZn-AI Alloy Coated Sheets, Nissin Steel Technical Report, Vol. 52, June 10, 1985. 874. Beranek, E., Barton, K, Sekerka, I., and Smrcek, K: Reactions in aqueous suspensions of zinc oxide. Part II. Reaction mechanism for the formation of hydrogen peroxide and its influence on the photoelectric properties of an iron electrode kept in the zinc oxide suspension, Collect. Czech. Chem. Commun. 25,369-374,1960. 875. Rausch. W.: Chemical surface treatment of steel coated with zinc and zinc alloys, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet. GALVATECH '89. Tokyo. September 5-7, 1989. pp. 197-205. 876. Johnsson. T, and Nordhag. L.: Corrosion resistance of coatings of aluminium. zinc and their alloys: Results offouryears' exposure, United Nations Seminar, Steel Committee, Economic and Technological Aspects of the Protection of Steel Against Corrosion, May 7-11, 1984. Geneva. 877. Akerdida, M. T, and Lyon, S. B.: Comparative corrosion resistance of galvanized and galvannealed steel, in Surface Engineering Practice-Processes, Fundamentals and Applications in Corrosion and Wear. K N. Strafford, P. K Datta, and J. S. Gray (eds.), Ellis Horwood, New York, pp. 543-552. 1990. 878. Condensation in Insulated Domestic Roofs, Department of the Environment, Building Research Establishment. Digest No. 270, Glasgow, Scotland, February 1983. 879. Common Defects in Low-Rise Traditional Housing, Department of the Environment, Building Research Establishment, Digest No. 268, Garston, Watford, England, December 1982. 880. Notowidjojo, B. D., Kennon, N. E, and Wingrove, A L.: Hot dip galvanizing: The process, the economics, the future, Proceedings of the Conference Step into the 90's, August 27-31, 1989, Queensland, Australia, Vol. 2. 881. Sakoda, A, Wakano, S., and Nishihara, M.: Effect of coating weight on cosmetic corrosion performance ofZn andZn-Alloy electroplated steel sheets in automotive body, l.fron Steelfnsf. lpn. 72, 1106-1110, 1986. 882. Bohe, A E., Vilche, J. R., Juttner, K, Lorenz, W. J., and Paatsch, w.: Investigations of the semiconductor properties of anodically formed passive layers on Zn and of ZnO single crystals in different aqueous electrolytes by EIS, Electrochim. Acta 34, pp. 1443-1448, 1989. 883. Stern, E: Iteration methods for calculating self-consistent fields in semiconductor inversion layers, l. Comput. Phys. 6, 56-67, 1970. 884. Baugh, L. M., Gidney, P. M., and Lee, J. A.: Pitting ofzinc in leclanche and related electrolytes. Part II. Mode of growth of pits and rate of propagation, Electrochim. Acta 25,765-777, 1980. 885. Baugh, L. M., Field, R. J., and Lee, 1. A.: Pitting of zinc in leclanche and related electrolytes. Part I. Pitting characteristics on battery alloy and single crystal electrodes, Electrochim. Acta 25, 751-763, 1980.
REFERENCES
445
886. Dirkse. T P. and Hampson. N. A.: The anodic behaviour of zinc in aqueous solution. Part III. Passivation in mixed KF~KOH solutions. Electrochim. Acta 17, 813~818, 1972. 887. Dirkse, T P.. and Hampson. N. A.: The zinc exchange reaction in KOH solution. Part II. Exchange current density measurements using the double-impulse method, Electrochim. Acta 17, 383~386. 1972. 888. Schwabe, K.. and Luck. H.-B.: The anodic behaviour of zinc in saturated solution of zinc sulphate. Electrochim. Acta 17. 99~IOS. 1972. 889. Dirkse, T P, and Hampson. N. A.: The anodic behaviour of zinc in aqueous KOH solution. Part I. Passivation experiments at very high current densities, Electrochim. Acta 16. 2049~20S6. 1971. 890. Breiter. M. W.: Dissolution and passivation of vertical porous zinc electrodes in alkaline solution. Electrochim. Acta IS, 1297-1304. 1970. 891. Krug, H., and Borchers. H.: Results of corrosion experiments with zinc alloys. Electrochim. Acta 13. 2203~2205. 1968. 892. Morrison. S. R., and Freund. T.: Chemical reactions of electrons and holes at the ZnO electrolytesolution interface. Electrochim. Acta 13, 1343-1349. 1968. 893. Bonnemay. M .. Hamelin. A.. and Sanchez de Haro. M.: Etude de la dechargc du zinc et du cadmium sur electrodes monocristallines. Electrochim. Acta 8, S09-S19. 1963. 894. Hurlen. T.: Kinetics of metal/metal-ion electrodes: Iron, copper. zinc, Electrochilll. Acta 7. 6S3~668. 1962. 895. Shoji, T, Hishinuma. M .. and Yamamoto, T.: Zinc~manganese dioxide galvanic cell using zinc sulphate as electrolyte: Rechargeability of the cell, J. Appl. Electrochem. 18, S21-526, 1988. 896. Randin, l-P: Determination of state-of-discharge ofzinc~silver oxide button cells. Part I. Galvanostatic measurements, 1. Apl'l. Electrochem. IS, 293~306, 1985. 897. Thomas, B. K., and Fray. D. 1.: The conductivity of aqueous zinc chloride solutions. 1. Appl. Electmchem. 12. 1-5. 1982. 898. Popov, K. I., Pavlovic, M. G., Spasojevic, M. D., and Nakic, V. M.: The critical overpotential for zinc dendrite formation, J. Apl'l. Electrochem. 9, S33~S36, 1979. 899. Poa, S.-P., and Wu, C. H.: The current distribution and shape change of zinc electrodes in secondary silver-zinc cells, J. Appl. Electmchem. 8. 491~SOI. 1978. 900. Poa, S.-P.. and Wu, C. H.: A quantitative study of shape change in zinc secondary electrodes. 1. Appl. Electrochem. 8, 427 ~436. 1978. 901. Coates, G., Hampson, N. A., Marshall, A., and Porter. D. F: The anodic behaviour of porous zinc electrodes. Part II. The effects orspecific surface area of the zinc compact material. 1. Appl. Electrochem. 4, 7S~80, 1974. 902. Dirkse, T P, and Kroon. D. 1.: Effect of ionic strength on the passivation of zinc electrodes in KOH solutions, 1. Appl. Electrochelll. I, 293~296. 1971. 903. Bushrod, C. J., and Hampson. N. A.: The anodic behaviour of zinc in KOH solution. Part V. Galvanostatic polarization with an interruption. J. Appl. Electrochem. 1, 99~ 101, 1971. 904. Dirkse, T P.: Voltage decay at passivated zinc anodes, J. Appl. Electrochel1l. 1. 27~33. 1971. 90S. Schuldiner, S" and White. R. E.: Studies of time~potential chang~s on an electrode surface during current interruption. Part I. Zinc~steel couple in synthetic sea water. J. Electrochem. Soc. 97. 433~447. 1950. 906. Delahay,P, Pourbaix, M., and Van Rysselberghe, P.: Potential-pH diagram orzinc and its applications to the study of zinc corrosion. 1. Electrochem. Soc. 98, 101 ~ lOS, 195 I. 907. EI Wakkad. S. E. S., Shams EI Din. A. M .. and Kotb, H.: The anodic oxidation of zinc and zinc~tin alloys at very low current density. J. Electrochem. Soc. 105, 47~51. 1958. 908. Gulbransen. E. A .. and McMillan. W. R.: Electron diffraction studies on the oxidation of pure copper and pure zinc between 200 and 500 C, J. Electrochem. Soc. 99, 393~40 1, 1952. 909. Granese. S. L., Fernandez. A., and Rosales. B. M.: Chemical characterization of the corrosion products formed on plain C steel, zinc, copper and aluminum, in Proceedings ()fthe 12th International Corrosion Congress 011 Corrosion COlltrolfor Low-Cost Reliability, September 19~24, 1993, Vo\. 2. pp. 652~661, National Association of Corrosion Engineers, Houston.
910. Fry, H., and Whitaker. M.: The anodic oxidation of zinc and a method of altering the characteristics of the anodic films. 1. Electrochell1. Soc. 106,606-611, 1959. 911. Papazian, H. A., Flinn. P A., and Trivich, D.: Influence of impurities on the photoconductance of zinc oxide, J. Electrochem. Soc. 104, 84~92, 1957.
REFERENCES
446
912. Gulbransen, E. A., and McMillan, W. R: Electron diffraction studies on the oxidation of pure copper and pure zinc between 200 and 500 C, 1. Electrochem. Soc. 100, 292-293, 1953. 913. Jorgensen, P. 1.: Effect of an electric field on the oxidation of zinc, 1. Electrochem. Soc. 110,461-462, 1963. 914. Harrison, D. E., and Hummel, E A: Phase equilibria and fluorescence in the system zinc oxide-boric oxide, 1. Electrochem. Soc. 103,491-498, 1956. 915. Rudresh, H. B., and Mayanna, S. M.: Adsorption of n-decylamine on zinc from acidc chloride solution, 1. Electrochem. Soc. 124,340-342, 1977. 916. Chuang, S.-E, Collins, S. D., and Smith, R. L.: Porous silicon microstructure as studied by transmission electron microscopy, Appl. Phys. Lett. 55, 1540-1542, 1989. 917. Mukherjee, D., Venkataraman, B., Palaniswamy, N., and Balakrishnan, K.: Effect of alloying elements on the corrosion resistance properties of zinc, 1. Electrochem. Soc. India 37, 233-236, 1988. 918. Kobayashi, T., Yoneyama, H., and Tamura, H.: Effect of illumination intensity on stabilization ofZnO photoanodes in halide solutions, Chem. Lett. 1979,457-460. 919. Kohl, P. A., and Bard, A. 1.: Semiconductor electrodes. Part 13. Characterization and behavior of n-type ZnO, CdS, and GaP electrodes in acetonitrile solutions, 1. Am. Chem. Soc. 99, 7531-7539, 1977. 920. Sakoda, A: Corrosion mechanism of zinc-iron alloy mechanical coating, Bosei Kanri 34(4), 131-134, 1990. 921. Morrison, S. R, and Freund, T.: Chemical role of holes and electrons in ZnO photocatalysis, 1. Chem. Phys.47, 1543-1551, 1967. 922. Yamato, K., Umino, S., Yasuda, A .• and Ichida, T.: Corrosion behavior of zinc alloy and finely alumina dispersion coating, Hyomen Gijutsu 40, 162-163, 1989. 923. Diaz, A., Gonzalez, C. L., Gonzalez, S., and Arevalo, A: Activating or inhibiting action of some compounds on zinc corrosion, Bull. Electrochem. 3, 289-294, 1987. 924. Hinton,8. R and Wilson, L.: The corrosion inhibition of zinc with cerous chloride, Corros. Sci. 29, 967-985, 1989. 925. Shuldener, H. L., and Lehrmen, L.: Effect of increasing the area of iron to zinc on the rate of reversal of potential in a zinc-iron system, Corrosion 14(12), 17, 1958. 926. Phibbs, M. K., and Giguere, P. A: Hydrogen peroxide and its analogues. Part III. Absorption spectrum of hydrogen and deuterium peroxides in the near ultraviolet, C. 1. Chem. 29, 490-493, 1951. 927. Jepson, S., Meecham, S., and Salt, E W.: The electrodeposition of iron-zinc alloys, Trans. Inst. Met. Finish. 32, 160-185, 1955. 928. Ruckert, J., Neubauer, E, and Zietelmann, c.: Influence of bitumineous roof coverings on corrosion behaviour of roof draining systems made of zinc and galvanized steel, Werkst. Korros. 34, 355-364, 1983. 929. Obrecht, M. E: The forms of corrosion in potable water piping. Part II., Heat./Piping/Air Condo 1973(July),33-37. 930. Obrecht, M. E: The forms of corrosion in potable water piping, Heat./Piping/Air Condo 1973(June), 57-62. 931. Case, 8.: The diffusivity of oxygen in dilute alkaline solution from 0 to 65 C, Electrochim. Acta 18, 293-299, 1973. 932. Morrison, S. R: Measurement of surface state energy levels of one-equivalent adsorbates on ZnO, Surf. Sci. 27, 586-604, 1971. 933. Dirkse, T. P., Passivation studies on the zinc electrode, Power Source 3, Proceedings of the 7th International Symposium, Brighton, September 1970, p. 485. 934. Bech-Nielsen, G., Larsen, M. K., Jensen, K. A., and Ulrich, D.: The corrosion and dissolution valence of a chromated Zn-Fe alloy studied by CMT and EC measurements, Corros. Sci. 36, 1907-1919, 1994. 935. Leidheiser, H.: Applications of mossbauer spectroscopy to studies of electrodeposits and the chemistry of metal surfaces, 1. Electrochem. Soc. 135, 5-11, 1988. 936. Asami, K., Hashimoto, K, Masumoto, T., and Shimodaira, S.: ESCA study of the passive film on an extremely corrosion-resistant amorphous iron alloy, Carras. Sci. 16,909-914, 1976. 937. Johansson, E., and Linder, M.: The influence of environmental acidification on the atmospheric corrosion of zinc, in Proceedings of the 12th International Corrosion Congress on Corrosion Control for Low-Cost Reliability, September 19-24, 1993, Vol. 2, pp. 549-560, National Association of Corrosion Engineers, Houston, 1993.
w.,
REFERENCES
447
938. Persson, D .. and Leygraf, C: Metal carboxylate formation during indoor atmospheric con'osion of Cu. Zn. and Ni, 1. Electrochern. Soc. 142, 1468-1477, 1995. 939. Gomes. W. P., and Cardon, F.: Study of noise associated with oxidation reactions at the illuminated single crystal zinc oxide anode, 1. Solid State Chern. 3, 125-130, 1971. 940. Svensson, I.-E., and Johansson, L.-G.: Initial stages of the SOrinduced atmospheric corrosion of zinc investigated by in-situ IR spectroscopy and time-resolved trace gas analysis. synergistic effects of N0 2 and 0 3 , in Proceedings of the 12th International Corrosion Congress on Corrosion Controlfor Low-Cost Reliabilitv, September 19-24. 1993. Vol. 2, pp. 662-675, National Association of Corrosion Engineers. Houston, 1993. 941. Gerischer, H.: On the stability of semiconductor electrodes against photodecomposition,./. Electmanal.
Chern. 82. 133-143. 1977. 942. Micka. K .. and Gerischer, H.: Anodic photooxidation of formic acid and methanol at a zinc oxide electrode and the influence of anions, 1. Electroanal. Cheln. 38, 397-402, 1972. 943. Rosales, B. M., and Granese, S. L.: Establishment of the Zn corrOSIOn rate in NaHS01 aqueous solutions,
./. Electrochern. Soc. 132, 1281-1284, 1985. 944. Hada. H., Tanemura. H., and Yonezawa, Y: The photoreduction of the silver ion in a zinc oxide suspension, Bull. Chern. Soc. lpn. 51, 3154-3160.1978. 945. Valencia. A .. Perez. R .. Arroyave, C, and Mesa, S.: Environmental effects in the atmospheric corrosion of zinc: An immersion-drying study, in Proceedinf?s of the 12th International Corrosion Congress on Corrosion Controlfor Low-Cost Reliability, September 19-24, 1993, Vol. 2, p. 748, National Association of Corrosion Engineers, Houston, 1993. 946. Pettinger, B., Schoppel, H.-R. Yokoyama, T, and Gerischer, H.: Tunnelling processes at highly doped ZnO-electrodes in aqueous electrolytes. Part II. Electron exchange with the valence band, Ber. Bunsenges. Phvs. Chern. 78, 1024-1030, 1974. 947. Graedel, T E.: Dissolution and precipitation phenomena in atmospheric corrosion, in Proceedings of the 12th International Corrosion Congress on Corrosion Control for Low-Cost Reliabilitv, September 19-24. 1993, Vol. 2. pp. 711-721, National Association of Corrosion Engineers, Houston, 1993. 948. Sengupta, T K., and Chatterjee, S. D.: Studies on the transport properties and surface states of anodic ZnO film on Zn, Proceedings of the 4th International Symposium on Frontiers of Electrochemistry, November 14-16,1989, Madras, India. 949. Freund, T.: Less-than-bandgap photosensitivity of ZnO and dye sensitization by ferrocyanide ion, SUfi Sci. 33, 295-305. 1972. 950. Freund, T, and Morrison, S. R: Mechanism of cathodic processes on the semiconductor zinc oxide, Surf Sci. 9,119-132,1968. 951. Morrison, S. R.: Electron capture by ions at the ZnO/solution interface, Sutf. Sci. 15,363-379,1969. 952. Schwerdtfeger W. J., and McDorman, O. N.: Potential and current requirements for the cathodic protection of steel in soils, 1. Res. Natl. Bur. Stand. 47, Research Paper RP2233, 1951. 953. Denison, I. A., and Romanoff, M.: Behavior of experimental zinc-steel couples underground, ./. Res. Natl. Bur. Stand. 40, Research Paper RP1876, 1948. 954. Yamashita, M., Enatsu, A., Adaniya, T.. and Hara, T.: Organic-silicate composite coated Ni-Zn plated steel sheet with high corrosion resistance for automobile bodies, 1. Iron Steellnst. lpn. 72, 1038-1043, 1986. 955. Ritter. I. J.: Ellipsometric studies on the cathodic delamination of organic coatings on iron and steel. 1. Coatings Techno!. 54, 51-57, 1982. 956. Leidheiser, H., and Wang, W.: Some substrate and environmental influences on the cathodic delamination of organic coatings, 1. Coatings Techno!. 53, 77-84, 1981. 957. Schmetzer, K., Schnorrer-Kohler, G., and Medenbach, 0.: Wulfingite and simonkolleite, two new minerals from Richelsdorf, Hesse, F. R G., Nelles lahrb. Mineral Monatsh. 1985(4), 145-154. 958. King, R A., Miller, I. D. A., and Stott, J. F. D.: Subsea pipelines: Internal and external biological corrosion, The National Association of Corrosion Engineers International Corrosion Conference, Houston, 1986, pp. 268-274. 959. Chan, L., and Griffin, G. L.: Temperature programmed desorption studies of hydrogen on Zn(OOOI) surfaces, Sutf. Sci. 145, 185-196, 1984.
448
REFERENCES
960. Hudson, J. C.: Atmospheric corrosion of metals. Third (experimental) report to the Atmospheric Corrosion Research Committee (British Non-Frrous Metals Research Association), Trans. Faraday Soc. 25, 177-252, 1929. 961. Sumita, M., Maruyama, N., and Uchiyama, I.: Fatigue crack growth rates and their thresholds of high strength steels in sea water at the zinc potential, Tetsu To Hagane 69( 11), 113-120, 1983. 962. Zinc-Coated Reinforcement for Concrete, Department of the Environment, Building Research Establishment, Digest No. 109, Garston, Watford, England, September 1969. 963. Bird. C. E.: The influence of minor constituents in portland cement on the behaviour of galvanized steel in concrete, Corros. Prevo Control 1964(July), 17-21. 964. Miyoshi, Y., Oie, Y., Amano, M., and Koyahara, H.: Corrosion behavior of electrophoretically coated cold rolled, galvanized and galvannealed steel sheet for automobiles-adaptability of cataphoretic primer to zinc plated steel, SAE Technical Paper No. 820334, Society of Automotive Engineers International Congress and Exposition, Detroit, February 22-26, 1982. 965. Lindqvist, S. A., Meszaros, L., and Svenson, L.: Aspects of galvanic action of zinc-rich paints. Electrochemical investigation of eight commercial primers,J. Oil Colour Chern. Assoc. 1985( I). 10-14. 966. Costa, 1. M., and Vilarrasa, M.: Effect of air pollution on atmospheric corrosion of zinc, Br. Carras. J. 28,117-120,1993. 967. Ahmed, D. S.: The intergranular attack in zinc alloy sacrificial anodes at elevated temperatures, Ph. D. Thesis, University of Manchester, 1984. 968. Justo, M. J., and Ferreira, M. G. S.: The corrosion of zinc in simulated SOrcontaining indoor atmospheres, Corros. Sci. 34, 533-545, 1993. 969. Wenz, R. P., and Claus, J. J.: Influence of abrasive cleaning on the corrosion resistance of coated steel, National Association of Corrosion Engineers Corrosion '81 Conference, Toronto, Ontario, April 6-10, 1981, Paper No. 82. 970. Ito, K., Oohashi, H., and Yoshii, T.: Corrosion behavior of zinc plated stainless steel, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7,1989, pp. 569-576. 971. Leclercq, M.: The electrochemical determination of the proportions of metal and oxide in zinc-base coatings obtained by hot spraying, Anti-Corrosion 1978(September), 5-11. 972. Oohashi, H., and Adachi, T.: Atmospheric corrosion resistance of zinc plated stainless steel, Proceedings of International Conference on Stainless Steels, Chiba, Japan, 1991. 973. Lieberman, E. S., Clayton, C. R., and Herman, H.: Thermally-sprayed active metal coatings for corrosion protection in marine environments, Naval Sea Systems Command, David Taylor Naval Ship R&D Center, Washington, D.C., Final Report, January 1984, Report No. SUSB-84-1. 974. Odnevall, I., and Westdahl, M.: Zinc chlorohydroxosulfates: Newly-discovered corrosion products on zinc, Carras. Sci. 34, 1231-1242, 1993. 975. Swamy, R. N.: In-situ behaviour of galvanized reinforcement, Galvanized Rebar Seminar, Structural Integrity Research Institute, University of Sheffield, April 12, 1991. 976. Gage, N., and Wright, D. A.: Corrosion testing of zinc and zinc alloy electrodeposited coatings, Proceedings of the Asia Pacific Interfinish 90: Growth Opportunities in the 1990's Conference, Singapore, November 19-22, 1990. 977. Yoshihara, R., Wake, R., and Yamamoto, M.: Characteristics of thermal diffused Zn/SnlNi triple coated steel sheets, Proceedings of the Asia Pacific Interfinish 90: Growth Opportunities in the 1990's Conference, Singapore, November 19-22, 1990. 978. van Ooij, W. J., and Sabata, A.: Effect of paint adhesion on the underfilm corrosion of painted precoated steels, National Association of Corrosion Engineers Corrosion '91 Conference, Cincinnati, Ohio, March 11-15,1991, Paper No. 417. 979. Baboian, R.: Chemistry and corrosivity of the automotive environment, National Association of Corrosion Engineers Corrosion '91 Conference, Cincinnati, Ohio, March 11-15, 1991, Paper No. 371. 980. Skerry, B. S., and Simpson, C. H.: Combined corrosion/weathering accelerated testing of coatings for corrosion control, National Association of Corrosion Engineers Corrosion '91 Conference, Cincinnati, Ohio, March 11-15, 1991, Paper No. 412. 981. Giridhar, J., and van Ooij, W. J.: Adhesion and corrosion properties of a new NiZnlZnCo-coated steel tire cord, Suif. Coatings Technol. 53, 243-255, 1992.
REFERENCES
449
982. van Ooij. W. J., and Sabata. A.: Chemical and thermal stability of phosphate layers on c()ld~rollcd and elcctrogalvanized steels . .I. Coatinfis Technol. 61. 51-65. 1989. 983. van Ooij. W. J .• Sabata, A .. Loison, D .. Jossic. 1., and Charbonnier. J.~c.: Paint delamination from electrocoated automotive stcels during atmosphenc corrosion. Part II. Zinc alloy~coatcd steeb. J Adhesion Sci. Technol. 3(2). 79-97.1989. 984. van Ooij. W. J., Sabata, A .. Loison. D., Jossic. T.. and Charbonnier. J.~c.: Paint delamination from electrocoated automotive steels during atmospheric corrosion. Part I. Hot-dip galvanized and electro~ galvanized steel. J. Adhesion Sci. Techno!. 3( I). 1-27. 1989. 985. van Ooij. W. J., and Sabata. A.: Developments towards corrosion~resistant sheet steels in the building and transportation industries. Scand. J. Meta/!. 21. 32-48, 1992. 986. van Ooij. W. 1., Edwards. R. A., and Neihciscl. G. L.: Cosmetic corrosion of welded hot dip galvanized steel panels, SAE Technical Paper No. 912287. Proceedings of the 5th Society of Automotive Engineers Automotive Corrosion & Prevention Conference, Dearborn, Michigan. October 21- 23. 1991. 987. Chaker. V: Corrosion testing in soils-past, present, and future, in Corrosion Testini!, and Evaluation: Silver Anniversary Volume, STP 1000, pp. 95-111, American Society for Testing and Materials, Philadelphia, 1990. 988. Kanda. K.: Properties of black electrogalvanized steel sheet, CAMP-/SIJ 4,614-617, 1991. 989. Molina, A., Blanco, M. 1., and Andrade, c.: Corrosion rate of three different types of galvanized coatings of steel reinforcements in contact with mortar, Proceedings of the 9th International Congress on Metallic Corrosion Conference, Toronto, Ontario, June 3-7,1984. 990. Walter, G. W.: Laboratory simulation of atmospheric corrosion by S02' Part I. Apparatus. e1ectrochemi~ cal techniques, example results. Corros. Sci. 32, 1331-1352, 1991. 991. Radeker, w., and Friehe, w.: Effect of alloying elements on the properties of hot dip galvanized coatings, Proceedings of the 7th International Galvanizing Conference on Hot Dip Galvanizing. Paris. June 1964, pp.167-178. 992. Askey, A., Lyon, S. B .. Thompson, G. E., Johnson, J. B., Wood, G. c.. Sage, P. w.. and Cooke, M. 1.: The effcct of f1y~ash particulates on the atmospheric corrosion of zinc and mild steel. Corms. Sci. 34, 1055-1081. 1993. 993. van Ooij. W. J., and Sabata, A.: Chemical stability of phosphate conversion coatings on cold~rolled and clcctrogalvanized steels. SurF Coatings Technol. 39/40,667-674. 1989. 994. van Ooij. W. 1., Sabata, A .. and Strom, G.: Corrosion of primed and fully painted prccoated automotive steels in atmospheric and scab tests, Prog. (Jig Coat. 18, 147-172,1990. 995. Baboian, R.: Corrosion behavior of painted galvanized auto~body steel in contact with exterior automotive trim, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet. GALVATECH '89. Tokyo. September 5-7. 1989, pp. 297-305. 996. Thomas. D. G., Cox. R. T.. and Richardson. B. A.: The role of metallic zinc in zinc rich paints. Proceedings of the 5th Annual Conference of the Australasian Corrosion Association, Newcastle, New South Wales, Australia, November 16-20.1964. 997. Zhu. Y, Li. S., Wang. J., Wang, J., and Qi, Y: The effect of titanium on improving the corrosion resistance of hot dip galvanizing zinc coating on malleable cast iron, Matet: Mech. Eng. (Chino) 15(3)' 1-4. 1991 I in Chinese]. 998. Ayoub, c.. Davin, A., Goodwin, F. E., and Coutsouradis, D.: Atmospheric corrosion behaviour of galfan coated steels, Proceedings of the 15th International Galvanizing Conference. INTERGAl.VA ' 88, Rome. June 1988. 999. Odnevall. I.. and Leygraf. c.: Formation of NaZn4CI(OH)6S04·6H20 in a marine atmosphere, Corros. Sci. 34,1213-1229,1993. 1000. Guttman. H., and Redden, 1. J.: Handling and storing of galvanized steel, Cominco PTC Project No. 71 ~5~412, Progress Report. 1972. 1001. Jagannathan, Y.: Emcrging technologies in the hot~dip coating of automotive sheet steel. 1. Metals 45(8), 48-51. 1993. 1002. Strom, M., van Ooij, W. J., Sabata, A., Edwards, R. A., Strom, G., and Ramamurthy, A. c.: A statistically designed study of atmospheric corrosion simulating automotive field conditions using a high perform~ ance climate chamber-status report of work in progress, SAE Technical Paper No. 912282, Proceed~ ings of the 5th Society of Automotive Engineers Automotive Corrosion & Prevention Conference, Dearborn, Michigan, October 21-23,1991.
450
REFERENCES
1003. Otero, E., and Royuela, J. J.: The assessment of short term data of pipe corrosion in drinking water. Part I. Galvanized steel, Carras. Sci. 34, 1581-1593, 1993. 1004. Sebisty, J. J., and Palmer, R. H.: Hot dip galvanizing with less common bath additions, Proceedings of the 7th International Conference on Hot Dip Galvanizing, Paris, June 1964, pp. 235-265. 1005. Abu Zahra, R. H.: A radiotracer study on the reaction between zinc and chromate. Part II. The corrosion of zinc in dilute chromate solutions, Corros. Sci. 6,349-355, 1966. 1006. Johansson. E., and Rendahl, B.: A swedish field exposure program of precoated sheet steels intended for the automotive industry. National Association of Corrosion Engineers Corrosion '91 Conference. Cincinnati. Ohio, March 11-15, 1991, Paper No. 390. 1007. Townsend, H. E.: Coated steel sheets for corrosion-resistant automobiles. National Association of Corrosion Engineers Corrosion '91 Conference, Cincinnati. Ohio, March 11-15. 1991. Paper No. 416. 1008. Loar. G. w.. Romer. K R., and Aoe. T. J.: Zinc-alloy electrodeposits for improved corrosion protection, Plating Surf. Finish. 78(3),74-79.1991. 1009. Lindsay. J. H.: Electrogalvanized automotive sheet steel and the manufacturing system. Plating Surf. Finish. 79(5).129-135.1992. 1010. Fawcett. N. C. Stearns. C. E .. and Bufkin. B. G.: Zinc consumption in zinc-rich primers co-pigmented with di-iron phosphide. 1. Coatings Technol. 56. 31-34. 1984. 1011. Lee. H. H.: Galvanized steel with improved resistance to intergranular corrosion, Proceedings of the Galvanizers Committee '77, 1977, pp. 17-23. 1012. Stareck, J. E.. and Cibulskis. W. S.: Recent developments in the use of conversion coatings on zinc. Proc. Am. Electroplaters Soc. 1947, 171-179. 1013. Babaqi. A. S .. Hefny. M. M .. Abdulla, R. M .• and EI-Basiouny. M. S.: Cathodic behaviour of zinc in aqueous methanol. 1. Electrochem. Soc. India 39.198-200.1990. 1014. Ghali, E. L., and Potvin. R. J. A: The mechanism of phosphating of steel, Corros. Sci. 12.583-594. 1972. 10 15. Townsend. H. E.: Twenty-five-year corrosion tests of 55% AI-Zn alloy coated steel sheet. Mater. Petform. 32(4). 68-71. 1993. 1016. Shaldaev, V. S .. Tomashova. N. N .. and Davydov. A D.: A combined protection of zinc in the alkaline current sources. Zashch. Met. 28. 516-519. 1992 [in Russian]. 1017. Notowidjojo. B. D .• Kennon. N. E. and Wingrove. A L.: Zinc-nickel coating: A new galvanizing technology, Proceedings of the Step into the 90's Conference, August 27-31. 1989, Queensland. Australia. pp. 623-632. 1018. Sommer, A. J .. and Leidheiser. H.: Effect of alkali metal hydroxides on the dissolution behavior of a zinc phosphate conversion coating on steel and pertinence to cathodic delamination. Corrosion 43. 661-665.1987. 1019. Leidheiser, H.: Electrochemical methods for appraising corrosion protective coatings, 1. Coatings Techno!. 63, 21-31. 1991. 1020. Jones, D. A. Blitz. R. K, and Hodjati. I.: Alternate immersion testing of coated sheet steel. Corrosion 42,255-262. 1986. 1021. Guttman, H., and Niessen, P.: Reactivity of silicon steels in hot-dip galvanizing, Can. Metall. Q. 11. 609-615. 1972. 1022. Harvey. G. 1.: Structure and cOrrosion resistance of zincalume coatings. BHP Tech. Bull. 25(2). 63-67. 1981. 1023. Kaneko, K, Mochizuki, A., Nakajima, H .. Ikeda, K .• and Toyose, K: Development of galvanized aluminum alloy sheet for body panels with an excellent filiform corrosion resistance, SAE Technical Paper No. 930703. Society of Automotive Engineers International Congress and Exposition, Detroit. March 1-5, 1993. 1024. Cornet, I.. and Bresler, B.: Some recent developments in the use of galvanized steel reinforcement in concrete. Proceedings of the 9th International Conference on Hot Dip Galvanizing, Dusseldorf. Germany, June 1970. pp. 414-426. 1025. Grenon, P.: Steel structures coating with zinc metallizing, National Association of Corrosion Engineers Corrosion '87 Conference, San Francisco, March 9-13. 1987. Paper No. 427. 1026. Fawcett, N. C, Stearns, C E., and Bufkin, B. G.: Effect of zinc oxidation on the conductivity and performance of di-iron phosphide augmented zinc-rich primers. 1. Coatings Technol. 56. 49-52. 1984.
REFERENCES
451
1027. Horstmann, D., and Riecke, E.: Fracture characteristics of hot dip galvanized steels in prestressed concrete in cases of corrosion in neutral and alkaline media, Proceedings of the 12th International Galvanizing Conference, INTERGALVA '79, Paris, May 17-23,1979. 1028. Hampson, N. A., Tarbox, M. J.: The behavior of zinc in potasium hydroxide solution, 1. Electrochem. Soc. 110,95-98,1963. 1029. Thangavel, K., Rengaswamy, N. S .. and Balakrishnan, K.: Corrosion resistance of galvanized steel in concrete. Mater. Perform. 34, 59-63, 1995. 1030. Cachet, C, Wiart, R., Ivanov, I., and Rashkov, S.: Mechanism of the reverse dissolution of zinc in the presence of nickel. Part I. Influence of anodic products and substrate purity, 1. Appl. Electrochem. 23, 1011-1016,1993. 1031. Raichevskii, G., and Vitkova, SI.: The effect of the phase composition and the structure of electrolytic cobalt deposits on their corrosional electrochemical behavior in an acid medium, Zashch. Met. 19, 418-424, 1973 lin Russian]. 1032. Cardon, E, and Gomes, W. P.: Experiments on the zinc oxide/electrolyte interface during anodic oxidation reactions under pulsed illumination, Surf. Sci. 27, 286-296, 1971. 1033. Vanden Berghe, R. A. L., Cardon, E, and Gomes, W. P.: On the electrochemical reactivity of the redox couple Fe(CN)i-lFe(CN)~- at the single crystal zinc oxide electrode, Surf. Sci. 39, 368-384, 1973. 1034. Pyun, S., Bae, J., Park,s., Kim, J., and Lee, Z.: The anodic behavior of hot-galavnized zinc layer in alkaline solution, Corros. Sci. 36, 827-835, 1994. 1035. Grabner, E. W.: Electrochemiluminescence ofrubrene at a zinc oxide electrode, Electrochim. Acta 20, 7-12, 1975. 1036. Pettinger, B., Schoppel, H.-R., and Gerischer, H.: Electroluminescence at semiconductor electrodes caused by hole injection from electrolytes, Ber. Bunsenges. SO, 849-855, 1976. 1037. Hampson, N. A., Shaw, P. E., and Taylor, R.: Anodic behavior of zinc in potassium hydroxide solution, Br. Corros. 1. 4, 207-211. 1969. 1038. Fichou, D., and Kossanyi, J.: Electroluminescence at the n-ZnO/electrolyte interface: A comparative study of the emission processes under cathodic and anodic pulsed polarization, .I. Electrochem. Soc. 133,1607-1617.1986. 1039. Lewis, D. A.: Some aspects of the corrosion of steel in concrete, Proceedings of the First International Congress on Metallic Corrosion, London, April 10-15, 1961, pp. 547-555. 1040. Hair, W. J., and Wall, R.: Metal spraying and painting for the protection of steel structures in polluted railway atmospheres, Proceedings of the First International Congress on Metallic Corrosion, London. April 10-15, 1961, pp. 556-566. 1041. Weast, R. C, Kotnik, L. J., and Geehan, D. M.: Corrosion of zinc in dilute aqueous solutions, Proceedings of the 6th International Conference on Hot Dip Galvanizing, Interlaken, Switzerland, June 1961, pp. 155-176. 1042. Stricker, E: Comparison of various methods for determining the thickness of galvanized coatings, of the iron-zinc alloy layers and of the pure zinc layer, Proceedings of the 7th International Conference on Hot Dip Galvanizing, Paris, June 1964, pp. 277-288. 1043. Johnsson, T., and Kucera, V: Possibilities of improving the corrosion resistance of zinc coatings by alloying, Proceedings of the 13th International Galvanizing Conference, INTERGALVA '82, London, 1982, pp. 4711-47/6. 1044. Iovchev, M., and Mihailov. G.: Restraining the corrosion of galvanized steel by the use of neutralite-a case study, Proceedings of the 15th International Galvanizing Conference, INTERGALVA '88, Rome, June 1988, pp. GH311-GH3/5. 1045. Yang, Z .. Li, J., and Jia, L.: Development and application of Zn-Fe alloy-coated (galvannealed) steel sheet, Proceedings of the 1st Asian-Pacific General Galvanizing Conference, September 15-18,1992, pp.208-216. 1046. Hauffe, K.: Oxidation (if Metals, Plenum Press, New York, 1965, pp. 201-228. 1047. Glayman, J.: Anodization of zinc, Galvallo 1970(February), 147-149. 1048. Terada, K.: Solution of zinc oxide in water charged with CO2 , Bull. It/st. Phys. Chem. Res. Tokyo 10, 864-871, 1931. 1049. Reinhart, F. M.: Twenty-year atmospheric corrosion investigation of zinc-coated and uncoated wire and wire products, STP 290, pp. 1-141, American Society for Testing and Materials, Philadelphia, 1961.
452
REFERENCES
1050. Townsend. H. E .• Borzillo, A R, and Barker, W. D.: Performance of Al-Zn alloy coated steel sheet after twenty-two years of atmospheric corrosion testing, Proceedings of the 2nd International Conference on Zinc Coated Steel Sheet, Rome, June 9-10,1988, pp. SAIII-SAllIl. 1051. Bonaretti, A, Capoccia, A., Giardetti, G., and Mucchino, U.: Process and use properties of lavegal, Proceedings of the 2nd International Conference on Zinc Coated Steel Sheet, Rome, June 9-10, 1988, pp. SC611-SC61l3. 1052. Applegate, L. M.: Cathodic Protection, McGraw-Hill, New York, 1960, pp. 115-117. 1053. Lucas, R W., and Eborall, R: Effects of additions to the metal and atmosphere on the surface oxidation of zinc, The British Non-Ferrous Metals Research Association, Research Report No. A.1463, December 1963. 1054. Mayne, J. E. 0.: Effect of rust and environment on inhibition by zinc-rich paints, 1. Iron Steelln.H. 176, 140-143, 1954; Atmospheric exposure tests with zinc-rich polystyrene paints, 176, 143-146, 1954. lOSS. Mayne, J. E. 0.: Cementiferous paints. Part I & II. Zinc phosphate cements, 1. Iron Steellnsf. 219-228, 1948;289-293,1950 1056. Ryabov, A K., and Gratsansku, N. N.: Influence of small additions of metals on the physicochemical properties of zinc coatings, Ukr. Khim. Zh. (Russ. Ed.) 30, 883-886, 1964. 1057. Pourbaix, A.: Corrosion of galvanized steel pipes used for water supply, CELBELCOR, Brussels, Belgium, Technical Report No. 180, July 1970. 1058. Hisamatsu, Y: Science and technology of zinc and zinc alloy coated steel sheet, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September5-7,1989,pp.3-12. 1059. Tsuzuki, Y, Suganuma, A, Yamamoto, T, Kozuka, N., and Mase, K.: Development of new zinc phosphating for automobile, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7, 1989, pp. 206-213. 1060. Sato, N.: Characterization of zinc phosphate films on zinc alloy coated steels, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7,1989, pp. 214-221. 1061. Suzuki, M., Hayashi, H., Kiyomatsu, 1., Miyawaki, T, and Matsushima, Y: A study on the hopeite crystal deposited on galvanized steel, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7, 1989, pp. 222-229. 1062. Nakazawa, M., and Yoneno, M.: Effect of phosphate and silica additives on the performance and structure of chromate conversion coatings, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7,1989, pp. 238-245. 1063. Koizumi, S., Shima, S., and Matsushima, Y: A development of black chromate-oxide finishes by baking process for galvanized steel, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7, 1989, pp. 246-253. 1064. Fountoulakis, S. G., and Steinbicker, R. N.: Influence of coating thickness on the cosmetic-corrosion performance of zinc, zinc alloy and multilayered electroplated sheet steels, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7, 1989, pp. 306-313. 1065. Petrey, L. D., Shaffer, R 1., Berndt, T, and Macciocco, J.: The effect of automotive plant painting operations on the paint adhesion and cosmetic corrosion resistance of precoated steels, Proceedings of the International" Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7,1989, pp. 314-320. 1066. Soshiki, Y, Ishida, w., Iida, A, Kanamaru, T, and Morita, 1.: Application of Fe-Zn alloy electroplated galvannealed steel sheet to automotive bodies, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7,1989, pp. 329-337. 1067. Tanaka, S., and Miyosawa, Y: Outdoor exposuring of formed specimens of precoated metal sheets, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7,1989, pp. 381-388. 1068. Mizuno, K., Suzuki, K., Ohtsubo, T, and Suzuki, M.: Characterization of phosphate conversion coatings, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7,1989, pp. 428-434.
REFERENCES
453
1069. Jixun, W, Yanping, L., Zhiyuan, Y, Yinhua, L., and Yingting, X.: A study ofthe structure ofZn-Fe alloy plating, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet. GALVATECH '89, Tokyo, September 5-7,1989, pp. 619-624. 1070. Gu. Moo Notis, M. R., and Marder, A. R.: Structure of as-plated electrodeposited zinc-iron coatings, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89. Tokyo. September 5-7,1989, pp. 462-469. 1071. Schedin, E .• Engberg, G .. Karlsson, S .. Kiusalaas, R., and Klang, H.: Plasticity of pure-zinc hot-dip galvanized coatings, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7,1989, pp. 493-499. 1072. Ito, Y, Hayashi, K., Kato, c., and Miyoshi, Y: A study on simulated and accelarated corrosion test methods for automotive precoated steel sheet, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steeel Sheet, GALVATECH '89, Tokyo, September 5-7, 1989, pp. 503-510. 1073. Davidson, D. D., Soreide, L. E., and Stephens, M. L.: Evalution of the cosmetic corrosion resistance of zinc coated sheet steels-using the chrysler chipping corrosion test and long term outdoor exposure testing, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7, 1989. pp. 519-527. 1074. Sakauchi, T., and Kunimi, H.: Evaluation of the corrosion resistance of zinc-alloy-plated steel sheet applied to automobile body, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet. GALVATECH '89, Tokyo. September 5-7,1989, pp. 528-535. 1075. Davidson, D. D., Soreide, L. E., and Stephens, M. L.: Assessment of the chrysler chipping corrosion test using laboratory scab and long term outdoor exposure testing, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet. GALVATECH '89, Tokyo, September 5-7, 1989, pp. 536-542. 1076. Uchiyama, Y, Hasaka, M., and Koga, H.: Effect of structure and mischmetal addition on the corrosion behaviour ofZn-5 mass% AI alloy, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7, 1989, pp. 545-552. 1077. Tajiri. Y.. Shimada, S., Yamaji, T., and Adaniya, T.: Effect of coating structure on corrosion resistance of hot dipped Zn-5%AI + Mg alloy coated steel sheet, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7, 1989, pp. 553-559. 1078. Walter. G. w.: Corrosion rates of zinc based coatings calculated by a galvanostatic non-linear method from low polarization data, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7, 1989. pp. 560-564. 1079. Walter, G. W.· Use of electrochemical sensors to measure corrosion characteristics of Zn and Zn-55% AI alloy coated steel under simulated atmospheric conditions, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7, 1989, pp. 565-568. 1080. Fukumoto, H .. Mizuki. H., and Masuhara, K.: Application of scanning vibrating electrode technique for investigation of chemical surface treatment on zinc coated steel sheets, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo. September 5-7. 1989. pp.577-584. 1081. Massinon, D .. and Dauchelle, D.: Recent progress towards the understanding of underfilm corrosion of coated steels used in the automotive industry, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet. GALVATECH '89, Tokyo, September 5-7, 1989, pp. 585-595. 1082. Masuhara. Koo and Kumon, E: Corrosion resistance of prepainted Zn-4%AI-0.I % Mg-misch metal alloy coated steel sheets, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7, 1989, pp. 596-602. j 083. Matsumoto. M., Shinohara, T., and Tsujikawa. S.: Application of scanning vibrating electrode technique to detect corrosion of zinc and zinc alloy layer under organic coating, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7, 1989, pp.603-61O. 1084. AI-Hashem, A., and Thomas, D.: The effect of corona discharge treatment on the corrosion behaviour of metallic zinc spots contained in polymeric coatings, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, September 5-7, 1989, pp. 611-618.
454
REFERENCES
1085. Camitz, G., and Vinka, T.: Corrosion of steel and metal-coated steel in swedish soils-effects of soil parameters, in Effects of Soil Characteristics on Corrosion, STP 1013, pp. 37-53, American Society for Testing and Materials, Philadelphia, 1989. 1086. RabeJer, R. c.: Soil corrosion evaluation of screw anchors, in Effects of Soil Characteristics on Corrosion, STP 1013, pp. 54-80, American Society for Testing and Materials, Philadelphia, 1989. 1087. Escalante, E.: Concepts of underground corrosion, in Effects of Soil Characteristics on Corrosion, STP 1013, pp. 81-94, American Society for Testing and Materials, Philadelphia, 1989. 1088. Edgar, T. V.: In-service corrosion of galvanized culvert pipe, in Effects of Soil Characteristics on Corrosion, STP 1013, pp. 133-143, American Society for Testing and Materials, Philadelphia, 1989. 1089. International Lead Zinc Research Organization: Corrosion Resistance of Unpainted Galfan Steel, ILZRO Project ZM-285, Progress Report No. 24, Research Triangle Park, North Carolina, 1992. 1090. Linch, R. E, and Goodwin, E E.: Galfan coated steel for automotive applications, SAE Technical Paper No. 860658, Society of Automotive Engineers Automotive Corrosion and Prevention Conference, Dearborn, Michigan, December 8-10, 1986. 1091. Guttman, H.: Corroded Highway Guard Rails, Cominco PTC Internal Report No. TSF 6715, 1967. 1092. Guttman, H.: A Study of Corroded Highway Guard Rail, Cominco Project No. TSF 5-410,1978. 1093. Pryor, M. J.: Bimetallic corrosion, in Corrosion, Vol. I, Shreir, L. L. (ed.), Chapter 1.69, George Newnes Ltd., London, 1963. 1094. Zhang, X. G., and Tran, H. Q.: Effect of cyclic wetting and drying on corrosion of zinc and steel, in Cyclic Cabinet Corrosion Testing, G. S. Haynes (ed.), ASTM STP 1238, pp. 125-135, American Society for Testing and Materials, Philadelphia, 1995. 1095. Hill, G. A., Spellman, D. L., and Stratfull, R. E: Laboratory corrosion tests of galvamzed steel in concrete, Transp. Res. Rec. No. 604, pp. 25-31,1976. \096. Sagues, A. A., and Kranc, S. c.: On the determination of polarization diagrams of reinforcing steel in concrete, Corrosion 48, 624-633, 1992. \097. Sagues, A. A.: Corrosion measurement techniques for steel in concrete, National Association of Corrosion Engineers, Corrosion '93, Paper No. 353, 1993. 1098. Videm, K., and Myrdal, R.: The electrochemical behavior of steel in concrete and how to evaluate the corrosion rate, National Association of Corrosion Engineers, Corrosion '96, Paper No. 348, 1996. 1099. Anderson, E. A.: Zinc and zinc coatings, in: Corrosion Resistance of Metals and Alloys, 2nd ed., E L. LaQue and H. R. Copson (eds.), pp. 223-247, Reinhold Publishing Corp., New York, 1963. 1100. Lauer, G., and Mansfeld, E: Measurement of galvanic corrosion current at zero external impedance, Corrosion 26, 504-506,1970. 1101. Mansfeld, E, and Kenkel, J. Y.: Laboratory studies of galvanic corrosion: Effect of velocity in NaCI and substitute ocean water, Corrosion 33, 236-240, 1977. 1102. Mansfeld, E, Hengstenberg, D. H., and Kenkel, J. Y.: Gal vanic corrosion of AI alloys: Effect of dissimilar metal, Corrosion 30. 343-353, 1974. 1103. Mansfeld, E, and Kenkel, J. Y.: Laboratory studies of galvanic corrosion: I. Two-metal couples. Corrosion 31, 298-302,1975. 1104. Zhang, X. G., and Valeriote, E. M.: Galvanic protection of steel by zinc under thin layer electrolytes, in Atmospheric Corrosion, W. W. Kirk and H. H. Lawson (eds.), ASTM STP 1239, pp. 230-239, American Society for Testing and Materials, Philadelphia, 1995. 1105. Mansfeld, E, and Kenkel, 1. Y.: Laboratory studies of galvanic corrosion: Two-metal couples, Corrosion 31,298-302,1975. 1106. Grimes W. D., Hartt, W. H., and Turner, D. H.: Cracking of concrete in sea water due to embedded metal corrosion, Corrosion 35. 309-316,1979. 1107. O'Reilly, M. W.: Chapter 15:2 in Corrosion, 2nd ed., Vol. 2, L. L. Shreir (ed.), Newnes-Butterworths, London, 1976. 1108. Mayne J. E. 0.: Chapter 15: 10 in Corrosion, 2nd ed., Vol. 2, L. L. Shreir (ed.), Newnes-Butterworths, London, 1976. 1109. Leidheiser, H.: Corrosion of painted metals-a review, Corrosion 38, 374-383, 1982. 1110. Ito, K., Oohashi, H., and Yoshii, T.: Corrosion behaviour of zinc plated stainless steel, Proceedings of the International Conference on Zinc and Zinc Alloy Coated Steel Sheet, GALVATECH '89, Tokyo, 1989, pp. 569-576.
REFERENCES
455
1111. Hess, J. c., and Soreide, L. E.: The effect of phosphate chemistry and rinsing conditions on the corrosion resistance and the paint adhesion of zinc coated and cold rolled sheet steels, Society of Automotive Engineers Automotive Corrosion and Prevention Conference Proceedings, Dearborn, Michigan. 1989, Paper No. 892557, pp. 17-38. 1112. Townsend, H. E.: Status of a cooperative effort by the automotive and steel industries to develop a standard accelerated corrosion test, Society of Automotive Engineers Automotive Corrosion and Prevention Conference Proceedings, Dearborn, Michigan. 1989, Paper No. 892569, pp. 133-145. lin. Ito, Y, and Miyoshi, Y: Corrosion protection of galvanized steel sheet-corrosion investigation offield vehicle and its laboratory evaluation methods, Society of Automotive Engineers Automotive CorrosIon and Prevention Conference Proceedings, Dearborn, Michigan, 1989, Paper No. H92580, pp. 217-227. 1114. Franks, L. L.: An eight-year atmospheric and on-vehicle corrosion test of painted and zinc coated steels, Society of Automotive Engineers Automotive Corrosion and Prevention Conference Proceedings. Dearborn, Michigan, 1989, Paper No. 892581, pp. 229-238. 1115. Kokubo. Z., Nomura, S., Sakai, H., Sakaguchi, M., and Zwai, M.: Newly developed Zn-Fe/Zn-Ni double-layer electroplated steel sheet, Society of Automotive Engineers Automotive Corrosion and Prevention Conference Proceedings, Dearborn, Michigan. 1983, Paper No. 830518, pp. 75-82. 1116. Catanzano, A., Arrigoni, G., Palladino, M., and Sarracino, M.: Multilayer electrogalvanized (Zn-CrCrOx) steel sheet for optimum corrosion protection of car bodies, Society of Automotive Engineers Automotive Corrosion and Prevention Conference Proceedings, Dearborn, Michigan. 1983, Paper No. 830583, pp. 107-124. 1117. Zhang, X. G .. and Tran, H. Q.: Effect of cyclic wetting and drying on corrosion of zinc and steel, in STP 1238, Cyclic Cahinet Corrosion Testing. G. S. Haynes (cd.), pp. 125-135, American Society for Testing and Materials, Philadelphia, 1995. IIIH. Despic. A. R.: Deposition and dissolution of metals and alloys, Part B: Mechanisms, kinetics. texture, and morphology, in Comprehensive Treatise ()( Electrochemistry, Vol. 7, B. E. Conway et al. (cds.), Plenum Press, New York, pp. 451-528. 1119. Straumanis, M. E., and Wang, Y: Surface disintegration of zinc mono- and polycrystals dissolving anodically in sodium bromate solutions and the apparent valency of zinc ions, Corrosion 22, 132-136, 1966. 1120. Kim, J. T, and Jorne, 1.: The kinetics and mass transfer of zinc electrode in acidic zinc-chloride solution, 1. Electrochem. Soc. 127,8-15.1980. 1121. Lee, T S.: Hydrogen overpotential on zinc containing small amounts of impurities in concentrated alkaline solution, 1. Electrochem. Soc. 120,707-709, 1973. 1122. Hoare, J. P.: The Electrochemistrv o(Oxvr;ell, Intersciencc Publishers, John Wiley & Sons. New York, 1968. 1123. Tarasevich, M. R., Sadkowski, A., and Yeager, E.: Oxygen electrochemistry, in Comprehmsil'e 7imtise ofElectrochemistrv, Vol. 7, B. E. Conway et al. (eds.), Plenum Press, New York, pp. 301-398. 1124. Stratmann, M., and MUller, J.: The mechanism of the oxygen reduction on rust-covered metal substrates. Corms. Sci. 36, 327-359, 1994. 1125. Holler, H. D.: Studies on galvanic couples, 1. Electrochem. Soc. 97(12),271, 277, 453,1950. 1126. Sato, N.: The passivity of metals and passivating films, in Passivity of Metals, R. Frankenthal and J. Kruger (eds.), pp. 29-58, The Electrochemical Society, Princeton, New Jersery, 1978. 1127. Vetter, K. J.: EleClrochim. Acta 16,1923,1971. 1128. Hull, M. N., Ellison, J. E., and Toni, J. E.: The anodic behaviour of zinc electrodes in potassium hydroxide electrolytes, J. Electrochem. Soc. 117, 192-198, 1970. 1129. Zhang, X. G., and Hwang, 1.: Cominco Internal Report 651-008, 1994. 1130. Eisenberg, M., Baumann, H. F, and Brettner, D. M.: Gravity field effects on zinc anode discharge in alkaline media, 1. Electrochem. Soc. 108,909-915, 1961. 1131. Hampson, N. A., Shaw, P. E., and Taylor, R.: Anodic behaviour of zinc in potassium hydroxide solution, Br. Corrosion 1. 4, 207-211,1969. 1132. Dirkse, T P.: The behaviour of the zinc electrode in alkaline solutions, 1. FJeclrochem. SoC, 125, 1591-1594, 1978. 1133. van der Lcij, M.: The possibility of black zinc oxide as spectrally selective coating for low temperature solar collectors, 1. Electrochem. Soc. 125, 1361-1364, 1978.
456
REFERENCES
1134. Aurian-Blajeni, B., and Tomkiewicz, M.: Passive zinc electrodes: Application of the effective medium theory, J. Electrochem. Soc. 132,869-870, 1985. 1135. Loftus, D., Roberts, J., Weaver, R, Leach, S., and Nanis, L.: Diffusivity in zinc chloride-potassium chloride electrolyte, J. Electrochem. Soc. 130,332-334, 1983. 1136. Bonasewicz, P., Hirschwald, W., and Neumann, G.: Influence of surface processes on electrical, photochemical, and thermodynamical properties of zinc oxide films, J. Electrochem. Soc. 133, 22702278,1986. 1137. Sato, Y., Niki, H., and Takamura, T.: Effects of carbonate on the anodic dissolution and the passivation of zinc electrode in concentrated solution of potassium hydroxide, 1. Electrochem. Soc. 118, 1269-1272, 1971. 1138. O'Brien, R N., Leja, J., and Beer, E. A: Interferometric study of the ZnlZnS041Zn system, J. Electrochem. Soc. 121,370-375, 1974. 1139. Gouda, V. K., Khedr, M. G. A., and Shams EI Din, A M.: Role of anions in the corrosion and corrosion-inhibition of zinc in aqueous solutions, Corros. Sci. 7,221-230,1967. 1140. Kordesch, K. V.: Alkaline manganese dioxide zinc batteries, in Batteries, K. V. Kordesch (ed.), Vol. I, pp. 241-384, Marcel Dekker, New York, 1974. 1141. Schumm, B.: Zinc-carbon batteries, in Modern Battery Technology, C. D. S. Tuck (ed.), pp. 87-111, Ellis Horwood, Chichester, England, 1991. 1142. Hunter, 1. C.: Alkaline-manganese batteries, in Modern Battery Technology, C. D. S. Tuck (ed.), pp. 111-125, Ellis Horwood, Chichester, England, 1991. 1143. Brodd, R J.: Batteriesfor Cordless Appliances, Research Studies Press, Letchworth, England, 1987. 1144. Snyder, R N., and Lander, 1. 1.: Rate of hydrogen evolution of zinc electrodes in alkaline solutions, Electrochem. Techno!. 3(5-6), 161-166, 1965. 1145. Bell, G. S.: The stability of zinc in zinc-chloride/ammonium-chloride electrolytes, Electrochim. Acta 13,2197-2202,1968. 1146. Riietschi, P.: Solubility and diffusion of hydrogen in strong electrolytes and the generation and consumption of hydrogen in sealed primary batteries, J. Electrochem. Soc. 114,301-305,1967. 1147. Sato, Y., Takahashi, M., Asakura, H., Yoshida, T., Tada, K., Kobayakawa, K., Chiba, N., and Yoshida, K.: Gas evolution behaviour ofZn alloy powder in KOH solution, 1. Power Sources 38,317-325, 1992. 1148. Dube, G., Renaud, R, and Huot, J. Y.: Amalgamation of battery grade indium-zinc powders, Prog. Batteries Battery Mater., 10, 151-155, 1991. 1149. Huber, R: Leclanche batteries, in Batteries, K. V. Kordesch (ed.), Vol. I, pp. 1-239, Marcel Dekker, New York, 1974. 1150. Baugh, L. M., Tye, F. L., and White, N. c.: Corrosion and polarization characteristics of zinc in concentrated alkaline solution and the effect of amalgamation, Power Sources 9, 303-323, 1983. 1151. Miura, A, Takata, K., Okazaki, R, Ogawa, H., Uemura, T., Nakamura, Y., and Kasahara, N.: Development of corrosion resistant zinc alloys for alkaline-manganese batteries, Denki Kagaku 1989(6), 57, 459. 1152. Meesus, M., Strauven, Y., and Groothaert, L.: New development in reduction of mercury content in zinc powder for alkaline dry batteries, Power Sources 11, 281-299, 1987. 1153. Glaeser, W.: Gas evolution data on very-Iow-mercury-content zinc powders for alkaline batteries, Power Sources 12, 265-296, 1989. 1154. Dirkse, T. P., De Wit, D., and Shoemaker, R: The anodic behaviour of zinc in KOH solutions, J. Electrochem. Soc. 115,442-444, 1968. 1155. Dirkse, T. P.: The behaviour of the zinc electrode in alkaline solutions, J. Electrochem. Soc. 126, 541-543,1979. 1156. Meeus, M. L., Crocq, G. G., and Lemaitre, C. A.: Metallurgical, structural and electrochemical factors
influencing the production of optimal zinc cans for leclanche dry cells by impact extrusion, Power Sources 7, 463-479, 1979. 1157. Mayer, S. T.: Anodic film formation on Cu, Zn, Na, Ag in alkaline media, Thesis, University of California, Berkeley, 1989. 1158. Lindsay, 1. H., and O'Keefe, T. J.: Electrogalvanizing, in Modem Aspects of Electrochemistry, No. 26, B. E. Conway, 1. O'M. Bockris, and R. E. White (eds.), Plenum Press, New York, pp. 165-228. 1159. Hafford, B., Pepper, W. E., and Lioyd, T. B.: Zinc Dust and Zinc Powder, International Lead Zinc Research Organization, New York, 1982.
REFERENCES
457
1160. Feliu, S., Jr., Morcillo, M., Bastidas, J. M., and Feliu, S.: Evolution of the protective mechanisms of zinc-rich paints during atmospheric exposure, 1. Coating Techno!. 65, 43-48, 1993. 1161. Szauer, T., and Miszczyk, A.: Zinc pigmented coatings: Ways of improving performance, Polymeric Materials for Corrosion Control. Proceedings of the American Chemical Society Conference, Chicago, Sept. 8-13,1985, pp. 849-854. 1162. Montie, J. E, and Hasser, M. D.: Mater. Peiform. 1976(August), 15-18. 1163. Odnevall, I.: Atmospheric corrosion of field exposed zinc, Ph.D. Thesis, Royal Institute of Technology, Stockholm, 1994. 1164. Odnevall, l., and Leygraf, c.: A comparison between analytical methods for zinc specimens exposed in a rual atmosphere, J. Electrochem. Soc. 138, 1923-1928, 1991. 1165. Odnevall, I., and Leygraf, c.: The formation of Zn4CI2(OH)4S04·5H20 in an urban and an industrial atmosphere, Corros. Sci. 36, 1551-1567, 1994. 1166. Odnevall, 1., and Leygraf, c.: Reaction sequence in atmosphere corrosion of zinc ASTM STP 1239. American Society for Testing and Materials, Philadelphia, 1994. 1167. International Lead Zinc Research Organization: Zinc Chemicals, ILZRO, New York, 1973. 1168. Bornholdt, R. A, and Perkins, 1.: SEM examination of corrosion product morphology for anodically polarized zinc, Metallographv 8, 401-409, 1975. 1169. Perkins, 1., Luebke, W. H., Graham, K. 1., and Todd, J. M.: Anodic corrosion of zinc alloysin seawater, 1. Electrochem. Soc. 124, 819-826, 1977. 1170. Tran, Q. T., and Dabosi, E: International Lead Zinc Research Organization Report ZE416, 1994. 1171. Flinn, D. R., Cramer, S. D., Carter, 1. P, Hurwitz, D. M., and Linstrom, P. J.: Materials degradation caused by acid rain, R. Bahoian (ed.), pp. 119-151, ACS Symposium Series 318, American Chemical Society, Washington, D.C., 1986. 1172. Romagnoli, R., Vetere, V. E, and Armas, R. A.: Influence of the composition of zinc-ethyl silicate paints on electrochemical properties. J. Appl. Electrochem. 24, 1013-10 18, 1994. 1173. Hime. M., and Machin, W. G.: Performance variances of galvanized steel in mortar and concrete, Corrosion 49, 858-860, 1993. 1174. Mattsson, E.: Corrosion: An electrochemical problem, Chemtech 15, 234-243, 1985. 1175. Rausch, The Phosphating of Metals, Finishing Publications Ltd., Middlesex, England, 1990. 1176. Bard, A. J., and Wrighton, M. S.: Thermodynamic potential for the anodic dissolution of IHype semiconductors, J. Electrochem. Soc. 124, 1706-1710, 1977. 1177. Mon'ison, S. R.: Electrochemistrv at Semicolldutor and Oxidized Metal Electmdes, Plenum Press, New York, 1980. 1178. Tench, D. M., and Gerischer, H.: The phototransition in ZnO at 380 nm studied hy anodic photocurrents, J. Elcctrochem. Soc. 124,1612-1618,1977. 1179. Brown, H.: Zinc Oxide-Properties and Applications, International Lead Zinc Research Organization, New York, 1976. 1180. Butler, M. A., and Ginley, D. S.: Prediction of f1atband potentials at semiconductor-electrolyte interfaces from atomic electronegativities, 1. Electrochem. Soc. 125, 228-232, 1978. 1181. Blok, L., and De Bruyn, P. L.: The ionic double layer at the ZnO/solution interface. 1. Colloid Interface Sci. 32, 518-526,1970. 1182. Memming, R.: The role of energy levels in semiconductor-electrolyte solar cells, 1. Electrochem. Soc. 125,117-123,1978. 1183. Pleskov, Yu. V., and Gurevch. Yu. Ya.: Semiconductor Photoelectrochemistry, P N. Bartlett (translation ed.), Consultants Bureau, New York, 1986. 1184. Ismail, K. M., Sikora, E., and Macdonald, D. D.: Electrochemical Society Extended Abstracts, Vol. 94-2, Abstract No. 203, 1994. 1185. Parks, G. A: The isoelectric points of solid oxides, solid hydroxides, and aqueous hydroxo complex systems, Chon. Rev. 65,177-198,1965. 1186. Gerischer, H., and Sorg, N.: Chemical dissolution of oxides: Experiments with si ntered ZnO pellets and ZnO single crystals, Werhl. Korros. 42, 149--157, 1991. 1187. Fruhwirth, 0., Herzog, G. W., and Poulios, 1.: Dark dissolution and photodissolution of ZnO, SlII/ Techno!. 24, 293-300,198:;. 1188. Bernas, A: A few cases of electron emission and transfer in heterogeneous media, J. Ph"s. Chem. 68, 2047-2052,1964.
w.:
458
REFERENCES
1189. Miles, M. H., and Gerischer, H.: The ZnJZn(Hg) electrode reaction in binary mixture of water and n-propand, 1. Electrochem. Soc. 118, 837-841, 1971. 1190. Porter, E C: Corrosion Resistance of Zinc and Zinc Alloys, Corrosion Technology Series, Vol. 6, P. A. Schweitzer (ed.), Marcel Dekker, New York, 1994. 1191. Romagnoli, R, and Vetere, V. E: The mechanism of the anti-corrosive action of zinc ethyl silicate paints, Surf. Coat. Int. 76, 208-210, 212-213,1993. 1192. Mackay, W B.: Chapter 11.2 in Corrosion, Vol. 2, L. L. Sheir (ed.), Newnes-Butterworths, London, 1976. 1193. Morgan, J. H.: Zinc anodes for cathodic protection, Carras. Techno!. 272-277, 1957. 1194. Crennel, J. T., and Wheeler, W C G.: Zinc anodes for use in sea water, 1. App!. Chem. 6,415-421, 1956. 1195. Cachet, C, Saidani, B., and Wi art, R: The behaviour of the zinc electrode in alkaline electrolytes. 1. Eletrochem. Soc. 139,644-654, 1992. 1196. Lohmann, E: Der einfluB des pH auf die elektrischen und chemischen eigenschaften von zinkoxidelektroden, Ber. Bunsenges. Phys. Chem. 70, 428-434. 1966. 1197. Sze, S.: Physics of Semiconductor Devices, John Wiley & Sons, New York, 1969. 1198. Gehmecker, H.: National Association of Corrosion Engineers Corrosion '91 Conference, Cincinnati, Ohio, March II-IS, 1991, Paper No. 380. 1199. Vincent, J. 1., and Coon, C L.: The relative effect of paint film thickness on bimetallic and crevice corrosion, Society of Automotive Engineers, Paper No. 860109,1986. 1200. Cormier, G. 1.: Effective cleaning of zinc and aluminum surface, Society of Automotive Engineers Automotive Corrosion and Prevention Conference and Proceedings, Dearborn. Michigan, 1989, Paper No. 892555, pp. 7-16. 1201. Yoshikawa, y., Itoh, M., Wakano, S., and Shiota, T.: The properties of organic composite coated Zn-Ni plated steel sheets, Society of Automotive Engineers Automotive Corrosion and Prevention Conference Proceedings, Dearborn, Michigan, 1989, Paper No. 892565, pp. 115-120. 1202. Van Eijnsbergen, 1. E H.: Duplex Systems, Elsevier, Amsterdam, 1994. 1203. Nartey, V. K, Binder, L.. and Kordesh. K: Identification of organic corrosion inhIbitors suitable for use in rechargeable alkaline zinc batteries, 1. Power Sources 52, 217-222. 1994. 1204. Glaeser, W.: Gas evolution data on very-Iow-mercury-content zinc powders for alkaline batteries, Power Sources 12, 265-296,1989. 1205. Yoshizawa, H., and Miura, A.: Mercury free alkaline manganese batteries, Prog. Batteries Battery Mater. 12,132-139,1993. 1206. Bhatt, D. P., and Udhayan, R: Electrochemical studies on a zinc-lead-cadmium alloy in aqueous ammonium chloride solution, 1. Power Sources 47, 177-184, 1994. 1207. McBreen, J.: Nickel/zinc batteries, 1. Power Sources 51, 37-44, 1994. 1208. Strauven, Y. A. 1., and Skenazi, A.: European Patent 0441420 AI, 1991. 1209. Gonzalez,1. A., Lopez, W, and Rodriquez, P.: Effects of moisture availability on corrosion kinetics of steel embedded in concrete, Corrosion 49, 1004-1010,1993. 1210. Thompson, N. G., Islam, M., Lankard, D. R, and Virmani, Y. P.: Effect of environmental factors on corrosion induced deterioration of reinforced concrete bridge structures, Corrosion '95, National Association of Corrosion Engineers International, Houston, Paper #4, 1995. 1211. Illston, J. M., Dinwoodie, J. M., and Smith, A. A.: Concrete, Timber and Metals, Van Nostrand Reinhold, New York, 1979. 1212. Suda, K, Misa, S., and Motohashi, K: Corrosion products of reinforcing bars embedded in concrete, Corros. Sci. 25,1543-1549,1993. 1213. Yamashita, T.: Corrosion of iron and zinc in concrete, Corros. Eng. 40, 345-357, 1991. 1214. Andrade, M. C, and Macias, A.: Galvanized reinforcements in concrete, Suiface Coatings-2, Chapter 5, pp. 137-182, Elsevier Applied Science, Amsterdam, 1988. 1215. Baker, E. A., Money, K L., and Sanborn, C 8.: Marine corrosion behavior of bare and metallic-coated steel reinforcing rods in concrete, ASTM STP 629, pp. 30-50, American Society for Testing and Materials, Philadelphia, 1977. 1216. Cornet, I., and Bresler, B.: Galvanized Reinforcement for Concrete-II, International Lead Zinc Research Orgainization, 1981, p. 1.
REFERENCES
459
1217. Tonini, D. E., and Cook, A. R.: The performance of galvanized reinforcement in high chloride environments-Field study results, International Lead Zinc Research Organization, 1981, pp. 57-88. 1218. Clear, K. c.: Federal Highway Administration, Washington, D.C., FHWNRD-82/028, December 1981. 1219. Slater, J. E.: Corrosion of Metals in Association with Concrete, STP 818, American Society for Testing and Materials, Philadelphia, 1983. 1220. Federal Highway Administration, Publication No. SA-91-023, HTA-13/6-91 (2.5M) QE, June 1991. 1221. Hansson, C. M.: Oxygen diffusion through portland cement mortars, Corms. Sci. 35, 1551-1556, 1993. 1222. Baldwin, K. R., Robinson, M. 1., and Smith, C. J. E.: Galvanic corrosion behavior at electrodeposited zinc and Zn-Ni coatings coupled with aluminum alloys, Br. Corros. 1. 29,293-298, 1994. 1223. Baldwin, K. R., Robinson, M. J., and Smith, C. J. E.: Galvanic corrosion behavior of electrodeposited Zn-Ni coatings coupled with steel, Br. Corros. 1.29,299-304, 1994. 1224. Weast, R. C. (ed.): Handbook of Chemist(\' and Physics, 68th ed., CRC Press, Boca Raton, Florida. 1988. 1225. Zhang, X. G .• Corrosion protection of rebar using a zinc wire, Corrosion 51. 679. 1995. 1226. Dugan, B. P., Stejskal. B .• and Stoneman. A. M.: Survey of highway bridges containing galvanized rebar. in Proceedings ofhllernational Conference on Corrosion and Corrosion Protection of Steel in Concrete. Vol. I, R. N., Swamy (ed.). p. 157. Sheffield Academic Press, Sheffield, United Kingdom, 1994. 1227. Lee, H. H.: Galvanized steel with improved resistance to intergranular corrosion, Proceedings of the Galvanizing Committee of the Zinc Institute, Chicago, December 1977. 1228. Baker. H. (ed.), Alloy Phase Diagrams, Vol. 3, ASM Metal Handbook. pp. 2.56. 2.206, ASM International. 1992. 1229. Bailey. R. W, and Ridge, H. G.: Corrosion of metals in buildings, Chon. Ind. 1957. 1222-1227. 1230. Edgar, T. V: In-service corrosion of galvanized culvert pipe, in EfieCls of Soil Characteristics 0/1 Corrosion, ASTM STP 1013, pp. 133-143, American Society for Testing and Materials. Philadelphia. 1989. 1231. Cabrillac, c., and Exertier, A.: The effect of coating composition on the properties of a galvanized coating, in Proceedings of the 7th International Conference on Hot Dip Galvanizing, Paris, June 1964. pp. 289-313, Pergamon Press. 1232. Whiting, D., Stark, D., and Schutt, W.: Galvanic anode cathodic protection system for bridge decksupdated results, Corrosion/Xl. International Corrosion Forum, Sponsored By the National Association of Corrosion Engineers. Toronto. Ontario, April 6-10, 1981, Paper #41. 1233. Doig. P, and Flewitt. P. E. 1.: A finite difference numerical analysis of galvanic corrosion for semi-infinite linear coplanar electrodes. 1. Electrochem. Soc. 126. pp. 2057-2063, 1979. 1234. Asphahani. A .. and Uhlig. H. H.: Stress corrosion cracking of 4140 high strength steel in aqueous solutions. 1. Electrochem. Soc. 122, pp. 174-179, 1975. 1235. McCafferty, E.: Distribution of potential and current in circular corrosion cells having unequal polarization parameters, 1. Electrochem. Soc. 124, 1869-1878, 1977. 1236. Morris, R., and Smyrl, W: Current and potential distribution in thin electrolyte layer galvanic cells, 1. Electrochem. Soc. 136,3229-3236, 1989. 1237. Tierra, P, Bernal, M., Molera. P., and Fernandez, L.: Comparative study on the cathodic protection supplied to steel by hot dip coatings, INTERGALVA '91, 3rd International Zinc Coated Sheet Conference. Barcelona, June 1991. 1238. Wetzel, D.: Batch hot dip galvanized coatings, inASM Metals Handbook, Vol. 5, pp. 360-371, Amcrican Society for Metals, Materials Park, Ohio, 1994. 1239. Compton, K. G.: Sources of underground corrosion potential differences, Corrosion 16, 87-90, 1960. 1240. Wilhelm, S. M.: Galvanic corrosion caused by corrosion products, in Galvanic Corrosion, H. P. Hack (ed.), ASTM STP 978, pp. 23-34, American Society for Testing and Materials, Philadelphia, 1988. 1241. Suzuki, S., Kanamaru. T.. Arai. K., and Morita, J.: Edge corrosion resistance of various coated stecl sheets, Corrosion/91. National Association of Corrosion Engineers Annual Conference and Corrosion Show, Cincinnati. Ohio. March I I - I 5. 1991, Paper No. 415. 1242. Pryor. M. J., and Keir, D. S.: Galvanic cOITosion: current flow and polarization characteristics of the aluminum-steel and zinc-steel couples in sodium chloride solution, 1. Electrochell1. Sue. 104,269-275, 1957. 1243. Pryor, M. J.. and Keir, D. S.: Galvanic corrosion: Effect of pH and dissolved concentration on the aluminum-steel couple, 1. Electrochem. Soc. 105,629-635, 1958.
460
REFERENCES
1244. Munn, R. S., and Devereux, O. E: Numerical modeling and solution of galvanic corrosion systems, Corrosion 47, 612-634, 1991. 1245. Huang, S., and Oriani, R. A.: The corrosion potential of galvanically coupled copper and zinc under humid gases, Fall Meeting of the Electrochemical Society, Minneapolis, Minnesota, 1993, Abstract No. 91. 1246. LaQue, E L.: Corrosion by sea water, in Corrosion Handbook, H. H. Uhlg, (ed.), p. 383, John Wiley & Sons, New York, 1948. 1247. Wellington, J. R.: The low potential zinc anode in theory and application, Corrosion 17, 122-128, 1961. 1248. Waber, J. T.: Mathematical studies on galvanic corrosion, J. Electrochem. Soc. 101, 271-276, 1954; 102,344-353,420-429,1955;103,64-72, 138-147,567-570, 1956. 1249. Gukild, 1., Husevag, R., and Markestad, A.: Reactivity of zinc and aluminum in fresh concrete, 5th Scandinavian Corrosion Congress, 28/1-12, 1968. 1250. Townsend, H. E.: Continuous hot dip coatings, inASM Metals Handbook, Vol. 5, pp. 339-348, American Society for Metals, Materials Park, Ohio, 1994. 1251. Wi art, R., Cachet, C., Bozhkov, c., and Rashkov, S.: On the nature of the "induction period" during the electrowinning of zinc from nickel containing sulphate electrolytes, J. Appl. Electrochem. 20, 381-389, 1990. 1252. Wang, Y., O'Keefe, T. J., and James, W. J.: Voltammetric evaluation of zinc electrowinning solution containing nickel, J. Electrochem. Soc. 127,2589-2593, 1980. 1253. Maja, M., Penazzi, N., Fratesi, R., and Roventi, G.: Zinc electrocrystallization from impurity-containing sulfate baths, J. Electrochem. Soc. 129, 2695-2700, 1982. 1254. Rashkov, S., Petrova, M., and Bozhkov, c.: The effect of nickel on the mechanism of the initial stages of zinc electrowinning from sulphate electrolytes, Part I, J. Appl. Electrochem. 20, 11-16, 1990. 1255. Rashkov, S., Petrova, M., and Bozhkov, c.: The effect of nickel on the mechanism of the initial stages of zinc electrowinning from sulphate electrolytes, Part II, 1. Appl. Electrochem. 20, 17-21, 1990. 1256. MacKinnon, D. J., and Brannen, J. M.: Aluminum cathode effects in zinc electrowinning from industrial acid sulphate electrolyte, J. Appl. Electrochem. 16, 127-133, 1986. 1257. O'Keefe, T. J., Chen, S. E, Cole, E. R., Jr., and Dattilo, M.: Electrochemical monitoring of electrogalvanizing solutions, J. Appl. Electrochem. 16,913-919,1986. 1258. Hagans, P. L., and Haas, C. M.: Chromate conversion coatings, in ASM Metals Handbook, Vol. 5, pp. 405-411, American Society for Metals, Materials Park, Ohio, 1994. 1259. Johansson, E., and Gullman, J.: Corrosion study of carbon steel and zinc, in Atmospheric Corrosion, W W. Kirk and H. H. Lawson (eds.), ASTM STP 1239, pp. 240-256, American Society for Testing and Materials, Philadelphia, 1995. 1260. Arnold, C. J.: Galvanized Steel Reinforced Concrete Bridge Decks, Michigan State Highway Commission, Research Report No. R-1033, Lansing, December 1976. 1261. Sun, X., and Tsujikawa, S.: Initiation of under film corrosion for zinc in galvanic contact with steel under cyclic corrosion tests, Corros. Eng. 41, 741-750,1992 [in Japanese]. 1262. Garcia-Anton J., Perz-Herranz, V., Guinon, J. L., and Lacoste, G.: Use of diiferential pulse polarography to study corrosion of galvanized steel in aqueous lithium bromide solution, Corrosion 50, 91,1994. 1263. Hwang, J., and Zhang, X. G.: Cominco PTC Internal Report, 1995. 1264. Schikorr G.: Relation between the atmospheric corrosion of zinc and nickel and the sulphur dioxide content of the air, Metal 15, 981,1961 [in German]. 1265. Strekalov, P. v., Agafonov, V. v., and Mikailovskii, Y. N.: Prot. Met. 8, 521, 1972. 1266. Greenburg, L., and Jacobs, M.: Ind. Eng. Chern. 48,1517,1956. 1267. Sydberger, T., and Vannerberg: The influence of the relative humidity and corrosion products on the adsorption of sulfur dioxide on metal surface, Corros. Sci. 12,775, 1972. 1268. Anderson, E. A., and Reinhard: Chemical removal of corrosion products in the determination of the corrosion rate of in zinc, ASTM 139, 691, 1939. 1269. Hudson, J. c.: Sixth Report of the Corrosion Committee, Special Report No. 66, Iron and Steel Institute, 1959, p. 25. 1270. Rajagopalan, K. S., and Ramaseshan, G.: Relative corrodibility of zinc and steel in unpolluted atmospheres, J. Sci. Ind. Res. (India) 18B, 87,1959. 1271. Barton, K., and Beranek, E.: Reaction mechanism of the atmospheric corrosion of metals in moist air contaminated with sufur dioxide, Werkst. Korros. 10, 377, 1959.
REFERENCES
461
1272. Newman, K.: Composite Materials, L. Holliday, ed., Elsevier, 1966. 1273. Power, T. C. et al: 1. Am. C(lIlcr. Inst. 1954. 1274. Tuutti, K.: Corrosion of steel in concrete, presented at the 6th European Congress on Metallic Corrosion, London, 1977. 1275. Jaenicke, w., and Scheitzer, H.: Z. Phvs. Chelll. (Frankfurt) 52,104, 1967. 1276. Aufenast, E: 1st International Symposium on Batteries, Christchurch. New Zealand, 1958. 1277. Stutz, G.: Ind. Eng Chem. 19(1),52,1927. 1278. Erbse, H., Hauffe, K., and Range, J.: Z. Phys. Chem. N.F 74, 248, 1971. 1279. Stark, D., and Perenchio, W.: The Performance of Galvanized Steel Reinforcement in Concrete Bridge Decks, Portand Cement Association, Project No. 2e-206, 1975. 1280. Harris, 1. 0.: Soil in the con'osion process, in Corrosion, Vol. 1, L. L. Shreir, cd., p. 2.37, George Newnes Ltd., London, 1963. 1281. Tang, N. Y: Comment on Fe-AI-Zn.1. Phase Equilihria 15, 237, 1994. 1282. Tang, N. Y, Adams, G. A., and Kolisnyk, P. S.: On determining effective aluminum in continuous galvanizing baths, GALVATECH '95, Chicago, Iron and Steel Society, 1995, pp. 777-782. 1283. Schikorr, G.: Relation between the atmospheric corrosion of zinc and nickel and the sulfur dioxide content in air, Metal 15, 981, 1961 lin Germani. 1284. Whiting, D. A., Nagi, M. A., and Broomfield, 1. P.: Laboratory evaluation of sacrificial anode materials for cathodic protection of reinforced concrete bridge decks, Corrosionl95, Paper No. 513, National Association of Corrosion Engineers, Houston, 1995. 1285. Saques, A., and Powers, R.: Sprayed-zinc sacrificial anodes for reinforced concrete in marine service, Corrosionl95, Paper No. 515, National Association of Corrosion Engineers, Houston, 1995. 1286. Brousseau, R., Arnott, M .. and Baldock, B.: Laboratory performance of zinc anodes for impressed current cathodic protection of reinforced concrete, Corrosion 51, 639-644, 1995. 1287. Brousseau, R., Arnott, M .. and Baldock, B.: Improving the adhesion of zinc coatings used to metallize concrete, Mater. Perform. 33,40-42, 1994. 1288. Kessler, R. J., Powers, R. G., and Lasa, 1. R.: Update on sacrificial anode cathodic protection on steel reinforced concrete structures in seawater, Corrosiol1195, Paper No. 516, National Association of Conosion Engineers, HOllston. 1995. 1289. Wills, D. J.. De Lisco, A., and Gleeson, B.: Intergranular corrosion in continuously galvanized steel, GALVATECH '95, Chicago, Iron and Steel Society, 1995, p. 139. 1290. Kehrer, H. P.: The effect of small impurities on grain boundary corrosion of zinc die castings, Metal!. 28,883, 1974 [in Germani. 1291. Keifets, V L., and Krasikov, B. S.: Dokl. Akad. Nauk SSSR 109, 586,1956. 1292. John, V: Durability of Galvanized Steel Building Components in Domestic Housing, Report No. WL/SMPIl106EIl09IfD, British Steel, 1991. 1293. Zhang, X. G.: Unpublishcd results, Cominco Product Technology Centre, 1995. 1294. Zhang, X. G.: Galvanic corrosion of zinc and its alloys, 1. Electrochem. Soc. 143, 1472-1484, 1996. 1295. Zinc Chemicals, International Lead Zinc Research Organization, New York, 1983. 1296. Fountoulakis, S. G.: Continuous electrodeposited coatings for steel strip, in ASM Metals Handhook, Vol. 5, pp. 349-359, Matcrials Park, Ohio, 1994. 1297. Sato, A.: Zinc plating, in ASM Metals Handbook, Vol. 5, pp. 227-235, American Society for Metals, Materials Park, Ohio, 1994. 1298. Zhang, X. G., and Hwang, J.: Zinc wired rebar, National Association of Corrosion Engineers, Corrosion '96, Paper No. :lOO, 1996. 1299. Townsend, H.: Bethlehem Steel Co., private communication.
Index Absorption atomic, 155 gas in ZnO, 97 light, 86, 94, 95. 98. 109-117. 350 water. 279, 236 Accelerate tests, 266, 317, 318, 325, 360, 368 Acceleration factor, 266 Accumulation layer, 107 Acetate, 30, 57 Acetic acid, 122, 40 I Acetone, 178, 395, 396 Acidification, 91, 225, 226, 280 Acids, 249, 266, 276, 395, 401 Acoustic effect, III Activation process, 16,45,126, 143, 188, 192,381 Activation energy, 72, 80, 135, 390, 399 Adhesion coatings, 12,407 cOlTosion products. 163, 296, 169, 175 paint, 15.82.328-330 Adsorption capacitance, 27 corrosion products, 135, 137, 149 current doubling, 113 dissolution and deposition, 34, 38 dissolution of ZnO, 119, 122 double layer, 97,101, 105 f1atband potential, 115 hydrogen evolution, 45 inhibitors, 384 intergranular corrosion, 59 passivation. 85. 91 pitting. 226, 227 in sol vents. 399 of sulfur dioxide, 244 of water layer, 242 Aeration, 141, 305 capacitance, 28 corrosion potential and current, 130, 133
corrosion rates, 30 I dissolution, 34, 35, 36 impedance, 57, 62, surface film, 85 Aerosol, 243 Aging, 231 Agitation: see Convection Air, 243 aeration, 174 in concrete, 353 cooling, 12 corrosion products, 162, 174, 176- 178 pollutants, 243, 244 soil, 307 wet storage stain, 237 Alcohols, 112, 178, 293, 393 Alloying, 3 anode, 214, 215 corrosion potential. 144-149 corrosion rates, 258, 260, 385, 387 galvanic corrosion, 184, 198,216 intergranular corrosion, 227-229 Aluminum adsorption of SO" 244 affinity to oxygen, 10 alloying element, 145,363,387,404 anode, 214,403 atmospheric corrosion, 258 corrosion potential, 145,263 galvanic corrosion, 184, 188, 197,202,210 intergranular corrosion, 181, 227 -231 Zn-AI phase diagram, 4 Aluminum oxide, 323 Amalgamation, 385, 391, 404 Ammonia, 285, 375 Annealing, 232, 257, 359 Antimony, 260 Anodes, 184, 186,212,213,305,403 alloying, 214, 215, 404
463
464 Anodes (cant.) cathodic protection, 407 efficiency, 213, 215,403 passivation, 403, 407 Anodization, 85-87,115 Area adsorption, 135 effective surface, 143,202, 343 gal vanic corrosion, 184, 196, 199-121, 214 polarity reversal, 206, 217 surface roughness, 143 Atmosphere, 241 corrosion forms, 184, 261 corrosion product, 158-168, 261 corrosion rates, 245-248 galvanic corrosion, 208 intergranular corrosion, 227 pitting, 220 Auger, spectroscopy, 157, 162 Automobile, 6, 215, 263, 315 Axis, 174 Bacteria, 244, 285, 295, 305 Band bending, 97-100, 104, 108 Band edge, 98,101-103,115,117,120 Band gap, 96, 98, 103, 108, III, 115, 118 Band structure, 96, 103-105, 114, 120 Barrier height, 106, 117 Batteries, 68, 373 cathode, 374 electrolyte, 374 pitting, 220, 225 Benzene, 78, 395,396 Benzoate, 299 Binding medium, 337, 344 Black film, 86 Blistering, 233, 315, 321 Bond strength, 359 Bonding, 159; see also Adhesion Brass, 3, 184, 375 Breakdown of passivation, 66, 89, 221, 225 pitting. 221 of semiconductor, 107-109, 115, 119 Bridge, 51, 351, 363 Buffering, 84, 91, 173 Bundle, 262 Cadmium, 184, 198,231,388,404 Capac itance adsorption, 35 corrosion product film, 62, 63 double layer, 27, 55, 62, 63 space charge layer, 100, 103, 106 surface coatings, 56
INDEX Capillary effect, 88, 237, 242, 245, 307, 353 Car: see Automobile Carbon, 245,338, 346.374 Carbon dioxide anodiztion, 86 corrosion products, 168, 170, 176 corrosion rates, 270, 287, 289 galvanic corrosion. 203 pitting, 219 polarity reversal, 203 solubility in water, 284 Carbonates corrosion products, 158-181, 237, 278 dissolution. 31 formation condition, 23, 24 278 intergranular corrosion, 234 passivation, 77-80, 85-87 polarity reversal, 203 Carbonation, 358, 366 Casting, 7 Catalytic reaction, 34, 38, 45,108,113,269,338 Catalyst, 75 Cathodic protection, 213, 359, 407 Cells, 195,201,301,334,373 Cement, 352 Charge carriers, 96, 105, 109, 110 Charge transfer. 32, 59, 62, 10 I, 104, 115, 122, 333 Charge transfer coefficient, 35, 40, 44, 48 Chemicals: see Ions Chipping, of paint, 328, 331 Chlorides; see also Ions adsorption, 27, 28 battery electrolytes, 374, 384 complexing with zinc, 24 in concrete, 351, 357 conductivity, 13, 127 corrosion current, 128, 135 corrosion products, 171, 172, 174, 177 corrosive agents, 357 galvanic corrosion, 198-202 intergranular corrosion, 229 marine, 252, 255, 266, 291 pitting, 221-223 in solutions, 296 current doubling, 113 dissolution. 31. 124 flux, II impedance, 60 oxygen reduction, 51 passivation, 78 passivation breakdown, 89 plating bath, 13 polarity reversal, 204 road salt, 263 salt spray test, 270
INDEX Chlorides (cont.) in solvents, 395, 399 vapor, 400 Chromium, 184. 260 Chromate, 85; see also Chemicals Chromating,16 Chromate film characterization, 16, 17, 171. 175 effect on corrosion, 144, 238, 260, 329, 362 properties, 17, 180 Clay, 305, 311 Cleaning, 11, 12 Cleavage, 165, 170 Climate, 254 Coating life, 256 Coating weight, 323 Coating thickness, 9 Coatings, 12, 14 Cobalt, 188 Colloidal solution, 16,305 Composite coatings, 323 Complexes, 25, 33-39 Condensation, 233, 237, 341, 252 Concrete, 184,227,351,403,407 Conductivity concrete, 354 corrosion products. 187 galvanic corrosion, 188,213 metal. 2 pitti ng, 225 soil, 305, 307, 310 solution, 26, 27, 301 surface films, 74, 79, 115 waters, 285 zinc oxide, 95, 96, lIS Contact line, 195,218,277 Contaminants, 161, 163, 176,241,265 Convection, 72, 75. 88, 139, 141,275 Conversion coatings, 16,260 Conversion factors, 129, 130 Cooling, 12, 162 Coordination number, 25 Copper, 145,245,391 alloying elements, 145, 146,307 galvanic corrosion, 184, 194, 196,210 Corrosion, 54 Corrosion current, 125,314,365 calculation, 125-127 correlation to weight loss, 153 definition, 125 measurement techniques, 127, 128 relation to corrosion potential, 130, 368 Corrosion environments atmosphere, 241 batteries, 373
465 concrete, 351 Corrosion environments (cont.) gaseous, 399 lab testing chambers, 267, 270, 272 organic sol vents, 393 paints, 315 soils, 305 solutions, 296 waters, 283 Corrosion forms, 183,293 Corrosion potential, 125-156 in concrete, 363, 367 definition, 125 painted, 332 pitting. 224 relation to corrosion current, 153 respect to passivation. 68. 85, 207 in soil, 211, 212. 312 in solution, 85.133-153 zinc-rich paints, 347, 348 Corrosion products, 157,261,399 characterization techniques, 157 classification, 157 corrosion current, 141, 149. 152 corrosion potential, 149, 152 effect on corrosion, 54, 178 effect of climate, 254 galvanic corrosion, 184,202,210.224,340 impedance, 62, 63 oxygen reduction, 51, 53 relative to weight loss. 272 stability, 31, 165, 170, 176 transformation, 168-176 Corrosion rates, 130 atmospheric, 245 in concrete, 360-363, 366 galvanic corrosion, 208-212 gaseous, 400-402 intergranular corrosion, 228-234 metals other than zinc and steel, 248, 292 in organic sol vents, 395-398 pitting, 219 simulated tests, 266-277 in soil, 308-311 in solutions, 296-304 under-paint, 319 units, 129, 130 in waters, 286-296 Corrosion tests, 211, 237, 317, 325 Cosmetic corrosion, 319, 320, 360. 378 Countries, 248 Cracking, 183,238, 358, 364 Creep, 2 Creeping, 319-323 Crevice, 218, 236,357
466 Crevice corrosion, 183 Cross section, 107 Crystal polycrystal, 115-117, 144 powder, 94 size, 94, 171, 175,320 zinc single crystal, 27, 32, 47 165, 174 Crystal structure corrosion products, 160, 161, 165, 175 zinc, I zinc oxide, 94 Crystal orientation breakdown, 108 capacitance, 28 cast and rolled products, 2 corrosion potential, 143 corrosion rates, 143, 302 decomposition, 122 dissolution and deposition, 35 photocurrent, III pitting, 225 standard potential, 20 Curing, 329, 337, 352 Current dark current, 105 distribution, 184, 192, 194 limiting current, 31 peak current, 68, 69, 82, 83 saturation current, 110 versus potential curves: see I-V curves Current collector, 375, 390 Current distribution, 192 Current doubling, 111-113 Current efficiency, 32, 79, 80 Current output, 404 Curvature, 39 Cut edge, 216, 262 Crystallization, 16, 323 Cyclic test, 216, 272, 318, 325 Cyclic wet/dry pattern, 254, 272 Deaeration: see Aeration Decomposition corrosion products, 173, 176 HP2,51 zinc oxide crystal, 110, 119-124 Deformation, 303, 331 Deicing, 263 Delamination, 315, 333 Dendrite, 39, 377 Depassivation,91 Depletion layer, III Depolarization, 207 Deposition, 36-39, liS, 117
INDEX Depth intergranular corrosion, 227 pitting, 217, 220, 225 sea water, 285, 294 Desorption, 34, 101; see also Adsorption Dew, 242 Dezincification, 149 Die casting alloy, 3, 7, 227 Dielectric properties, 74, 89, 395 Diffusion in concrete, 354 deposition, 39 dissolution, 31, 35, 162 of electron and hole, 110 of hydrogen, 40, 139 of mercury in zinc, 386 in organic solvents, 396 in paint, 316 passivation, 62, 69-73, 77, 80, 84, 88 of salts, 316, 357 of various ions, 26 of zinc in zinc coatings, 8 in zinc oxide, 97 Diffusion coefficients, 25, 40, 71, 190, 316, 354 Diffusion length, 110 Discharging, 377, 390 Disintegration, 32 Distance from city center, 244 gal vanic action, 20 I, 209, 213 from sea coast, 244, 252 from waterline, 301 Dislocation, 39, 225 Dissolution, 29 at corrosion potential, 125-129 corrosion products, 30, 36, 168, 171 efficiency, 32 exchange current, 30, 31 inhibition, 62 paint delamination, 333 at passivation, 66, 80, 87-89 pitting, 226 process, 62 of zinc oxide, 119-124 Distilled water, 140, 169, 170, 286 intergranular corrosion, 227 pitting, 217, 224 polarity reversal, 203, 204 Donor,96, liS, 118, 122 Doping, 117 Double layer, 27-29, 55, 62,115, 173,396 Drying accelerated test, 267, 325 battery, 221
INDEX Drying (coni.) concrete, 356 corrosion current, 144 corrosion products, 160, 177, 242, 254 wet storage stain, 257 Ductility, 227, 238, 359 Duplex coating, 323 Dust, 215, 245, 316 Dye, III Edge, 196,217,277 Electrochemical techniques, 29, 54, 153 Electrogalvanizing, 13, 14 Electrolumineseces, 114 Electron acceptor, 96 Electrolyte battery, 374, 379 conductivity, 26, 27. 191, 192 near surface, 32 pitting, 225, 227 resistance, 55, 62, 144, 211 thickness, 190, 192-197,213,276 Electroplating, 13, 14,260 Electrowinning, 19, 39 Elevation, 252 EMF series, 144, 185, 186,211 Enthalpy, 390 Epoxy, 339 Equilibrium, 19,20-25, 120,127,133 Equivalent electrical circuit: see Impedance Etch, 74, 102,221 Eutectic phase, 3, 13, 323 Exchange current, 30, 36, 42,125,127,130,151, 396 Extracts, 356 Extrusion, 232 Faradaic process, 32 Fatigue, 369 Fermi level, 96, 98, 100-103 Field strength, 90, 104 Film: see Surface films Flake, 7 Flame: see MetaJlizing Flat band, 97, 98 Flatband potential, 100 definition, 98 determination, 100 oxide film, lIS, 116, 118,202 single crystal, 100, 10 I Flow, 144, 178, 184,266,289,294,348 Flux, II, 110,237,354 Fluxing, II Fly-ash,269 Fog, 243
467 Formability, 321 Formic acid, 112, 113 Galfan, 12, 162,259,263 Galvalume, 12, 162,263,324 Galvanneal, 12, 145, 147, 148,263 Galvanic corrosion, 183, 261, 312, 332, 339 corrosion rate, 198,201,210-213 of coupled metals, 184, 196-198, 210 definition, 183 factors, 188, 209 testing, 184,209,211 theory, 185-196 Galvanic current, 185, 191, 194-197,213 Galvanic protection, 213 anodes, 215 coatings, 215 definition, 185 field data, 212 throwing power, 213 zinc-rich paint, 339 Galvanic series, 187,211 Galvanizing: see also Electroplating batch process, 10 classification, 7, 8 continuous process, II Galvanized steel, 256 coating methods, 7, 256 galvanic protection, 213 life, 256 pitting, 203, 219, 225, 308 polarity reversal, 203 premature darkening, 262 surface treatment, 12 wet storage stain, 183, 236 Galvanized rebar, 359 Gases, 184,339 CO,: see Carbon dioxide H,: see Hydrogen H,S,243,268,285,401 N,,242 NH 3,243 0,: see Oxygen SO,: see Sulfur dioxide Gassing, 224, 338, 378 Gel, 353, 382 General corrosion, 183, 225 Geometrical effect, 185, 186, 189-192, 213 Gravitation, 217 Grain, crystal orientation, 2 potential, 144 shape, 74 size, 3, 230, 235, 388 Grain boundary, 3, 32, 225, 352
468 Grid,390 Grinding, 143 Hardness, 284 Hardware, 213 Heat treatment, 231 Highway, 262 Hole, 96, 217 Hot dip: see Galvanizing Hot water corrosion products, 168-170, 179 corrosion rates, 288, 289 pitting, 219, 291 polarity reversal, 203, 208 Humidity atmosphere, 241 intergranular corrosion, 226 test, 267 time of wetness, 248, 252, 280 wet storage stain, 183, 236 Hydration number, 25 Hydration process, 36, 45, 67, 352 Hydration shell, 39 Hydrogen, gas anode,214 corrosion rate in water, 299 diffusion, 40, 139 intergranular corrosion, 235 measurement device, 378 rebar/concrete surface, 359 solubility, 40 surface preparation, 78 units, 129 Hydrogen embrittlement, 183, 238 Hydrogen peroxide, 51, 107, 112 Hydrogen reaction, 39 corrosion potential, 134 exchange current, 40, 42 inhibition, 47 over potential, 41, 45, 230 oxide film, 45 thermodynamics, 21 Ice, 237 Illumination, 109 Impedance, 54 corrosion current, 115 dissolution, 35 equivalent circuit, 56, 62 paints, 318, 340 polarization resistance, 127 semiconducting film, 118 spectrum, 56-61, 74 thin-layer electrolyte, 277
INDEX Impurities, in electrodes, 47 electrolyte, 39, 51, 227 grain boundaries. 235 water, 207 zinc, 47, 183,258,377,404 zinc oxide, 93, 96, 100 Incubation, 319 Indium, 387 Indoor atmosphere, 16 I. 162, 264 Inductance, 55 Inhibitors, 359, 384 Interfaces, 55, 67, 90, 103, 137,235,316,339,359 Intergranular corrosion. 129, 183.225.239,407 Intermetallic compounds, 3. 8-12, 145, 227-238 Ionic properties, 25-27 Ionic strength, 72 Ionization, 135 Ions. in solutions Ag+, 114, 175 As'+, 46, 112, 382 AsO~-, 85, 89 BO;-, 89, 116, 117,221 Br-, 51, 60, 89,113, 124, 135,383 BrO,,32 Ca2+~ 85. 137. 140, 174, 179,205,357 Cd'+, 220 Ce ' +, 137, 180 CH,CO;,89 Ct: see Chlorides CW, 13, 112 C10:;, 32, 137 CIO.;, 28, 51, 89, 113, 128. 171,384 Co2+, 298 col-: see Carbonates complexing agents, 25 concentration, 298 corrosion current, 131, 135, 145, 150 corrosion potential, 131, 135-137, 145, 149 corrosion products, 172, 177 corrosion rates, 297, 298 current doubling, 113 crO;-, 16,31,89,223 Cu'+' 46,175,201,291,382 diffusion coefficients, 26 exchange current, 30 F-,89 Fe'+, 46, 382 Fe(CN)t, 103 f1atband potential, 101 galvanic corrosion, 184,201,204 HCOO-112 HP-, 51, 53 hydrogen evolution, 41
469
INDEX Ions, in solutions (collt.) 1-,89, 107. 113, 123 In'+.383 I-V curve ofZnO. 106 K+, 45 Mg2+, 201 MnO:;, 31, 113 MoOt, 89 Ni 2+, 39, 46. 382 Na+,201 NH;, II. 13,28,34,46,60,128, 173,244,374, 384 NO~, 30, 51, 89,107,112,135,137,140,205 passivation, 66, 89, 90 Pb2+, 36, 47. 73, III. 220. 383 pitting, 222 PO~-, 51. 115, 139. 171,223 S2-,296 Sb'+, 46, 382 SiOl-,205 Sn 2+, 31, 46, 73, 383 SOl-, 137 S01-, 14,40,51,57, 113, 135, 140, 171,201, 244,384 WO~-, 31. 135, 136, 223 Zn 2+, 133 battery, 373 current doubling, 113 corrosion current, 134 corrosion potential, 130, 133. 138 dissolution, 29, 63, 396 hydrogen reaction, 46 photo decomposition, 122 standard potential, 19 Zn(OH)i-. 113 battery, 381 corrosion potential, 134 dissolution, 29, 63 hydrogen reaction, 46, 47 passivation, 75, 77 solubility, 22 Iron, steel adsorption of S02' 244 alloying element, 198,216,293,404 concrete reinforcement, 358 corrosion potential, 145,312 corrosion products, 202, 332 corrosion protection, 208, 210-217, 259 corrosion rates, 208, 210, 212 atmospheric, 246, 252 gaseous, 402 soil,310 under-paint, 263 Fe-Zn phase diagram, 5
galvanic action, 184, 188.203,208. 345 Iron. steel (cant.) intermetallic compounds, 3, 219, 279 phosphatising, 330 solubility in zinc, 3 IR drop, 74, 403 Isotherm, 105 I-V curves breakdown of ZnO, 108 cathodic reaction on ZnO, 106 corrosion current, 128 decomposition ofZnO, 121 galvanic corrosion, 184,200 intergranular corrosion, 233 passivation, 67, 68, 73. 74, 79-84 passivation breakdown, 89 photocurrent, 110 pitting, 221 in soil, 313 in solvents, 399 zinc rich paints, 345 Kelvin probe, 184 Kinetics corrosion processes, 125-127 dissolution, 29-39 electrochemical techniques, 29 galvanic corrosion, 188-191 hydrogen reaction, 39-46 oxygen reduction, 48-53 passivation, 68-85 semiconductor electrochemistry, 105-115 Kink sites, 35, 122 Lattice constant, I, 94 Leaching, 356, 358 Lead, 258 galvanic corrosion, 184,210 intergranular corrosion, 227, 229-231 Leclanche cell, 38, 220, 225, 373 Life, 256, 333, 360, 375 Lifetime, 113. 115, 346 Light absorption, 86, 94,109-115,350 deflection, 94, 95 emission, 114, 115 penetration depth, 110 transmission, 94, 95 Linear polarization, 54 Loam, 305 Local corrosion, 185, 200 Localized corrosion, 66, 91,129,225 Luminescence, 114
470 Magnesium, 145,260,388 anode,214,403 gal vanic corrosion, 184, 210 intergranular corrosion, 229 Manganese, 145, 146 Mass transport, 71, 77, 88 Mechanisms atmospheric corrosion, 278 carrier recombination, III current doubling, III deposition, 38 dissolution, 32 formation of corrosion products, 181 hydrogen reaction, 43 intergranular corrosion, 235 metallic surface film, 86 oxygen reaction, 49 passivation, 75, 80 passivation breakdown, 89, 225 pitting, 225 polarity reversal, 207 premature darkeni ng, 262 protection by zinc rich paint, 339 under-paint corrosion, 333 wet storage stain, 236 Mechanical properties, 2. 90. 231. 234 Melting point. I Meniscus, 30 I Mercury. 385 Metallizing. 6, 407 Metallography. 8, 9. II Metallurgical effect, 229. 302 Metals (other than zinc and steel) adsorption, 244 atmospheric corrosion, 248 gal vanic corrosion, 184. 198. 210 hydrogen exchange current, 42 sea water corrosion. 292 Microstructure. 232 Mill scale, 347 Minerals, 283 Mixing, 53 Mobility. 25-27. 36 Moisture. 175,237,242.355,400 Monolayer, 29, 78. 79, 82 Morphology, 39, 74. 75,163-165,172.175,244 Mortar. 358, 365 Mott-Schottkyequation. 100 Nernst equation. 133, 138 Neutralization, 16 Nickel, 184, 198,210,260.387 Normal corrosion, 185.210,211 Nucleation, 16,77,79.171
INDEX Ohmic effect CUlTent measurement, 156 gal vanic corrosion, 188. 190. 196 sacrificial protection. 403 Open circuit condition, 68. 82. 87. 125, 188,225. 345 Orbital. 103 Organic substances. 384. 393. 402 Oscillation. 39, 74 Osmatic pressure, 316 Overpotential, 127. 188.386 anode. 215.403 dissolution. 33 hydrogen evolution, 41. 43 oxygen reduction. 50 passivation. 66 Oxidation of zinc, liS. 162 of water. 120. 122 Oxide: see Zinc oxide Oxygen in concrete. 354 diffusion, 51. 53. 274,195.334 diffusion coefficient, 48 effect on corrosion, 141-143, 181 evolution, 78, 88 galvanic corrosion, 190 intergranular corrosion, 287 kinetics, 48-54, 202 paints, 334, 339 pitting, 218 polarity reversal, 203, 206-208 stability, 21 standard potential, 48 solubility, 48, 284, 298 Paint, 328. 337 adhesion, 15,82.328-330 galvanic corrosion, 184 paintability, 12, 14 permittivity, 316 properties, 316. 337 thicknesi>, 324 Particles, 339, 341, 343, 352, 389 Particulate film, liS, 117 Passivation, 65-92 in batteries, 375, 376 breakdown, 66. 81, 87. 89-91. 221, 225 condition, 21, 23, 67, 78 definition. 65, 87 effect on corrosion, 130-133, 138, 191,233 overpotential, 66 passivation current, 67, 75, 87-89 passivation potential, 67,81, 84, 85
INDEX Passivation (COlli.) polarity reversal 207-208 process. 67. 74-77 time to passivation. 65. 70-73. 77. 78 Passive films: see a/so Surface films characterization. 75. 78 properties. 75. 93.115-118.149.179.207 stability. 67. 86-92. 299 Passive state. 65. 67. 87. 88, 208, 363 Paste, 352 Permeation, 176. 178. 215, 316, 339, 353 Perforation, 221, 319 pH, 137.206.298 buffer, 82. 84, 173 capacitance. 27 concrete. 357 corrosion current. 131. 137-140, 148, 152 corrosion potential, 131, 137-140, 148, 152 corrosion products. 171-173. 273 corrosion rates, 273. 298, 311 dark potential, 102 decomposition ofZnO, 119, 122 flatband potential, 102-105, 115, 116 intergranular corrosion, 233 passivation, 66, 80, 83, 299 passivation breakdown, 89 pitting. 222 polarity reversal, 206 soil, 309-311 solubility, 22. 23. 299. 357 waters. 285 pH-potential diagram, 21,23,374 Phase diagram, 4, 5 Phosphates, as chemicals corrosion potential. 136 dissolution, 31. 62 oxygen reduction, 5 I passivation, 80-84, 86, 89 zinc-rich paint, 346 Phosphate coating, IS, 171, 180,329,334 Phosphating. 15,80.260.329 Photo: see Light Photo catalysis, 113. 269 PholOcurrent. 95, 98. 100. 109 current doubling, III dissolution of ZnO. 121-123 luminesces, 115 passive film, 116, 117 Photoeleetrochemistry, 109-118 Photoluminensece, 114 Photopotential, 109, 110, 117, 207 Physical properties, I. 93, 94 Pickling. I I Pigment, 316, 338. 341, 345 Pipe, 217. 262, 291, 312
471 Pitting, 217, 290 characterization, 66, 89. 217 corrosion current, 129 corrosion rate 219-220 galvanic, 184 in hot water, 290 occurrence, 202. 217, 261, 263. 293. 308 measurement, 217 mechanism, 89-91 Pitting potentia!, 66, 89, 217, 221-224, 226 Plastics, 184,213,218,401 Plate, 209 Polarity reversal, 179, 184, 202-208. 219, 407 Polarization resistance, 127 conductivity, 367 corrosion rate, IS I, 154 impedance, 51 zinc-rich paint, 342, 349 Polarization parameter, 192 Polishing, 32, 143.225 Pollutants, 242, 249, 269 Polymers. 316, 338 Polymerization, 82 Porosity, 74, 353 Potential, electrode breakdown potential. 66 dark potential, 102, 104 half-wave potential, 51 Helmholtz potential, 101-105 mixed potential, 125 open circuit potential. 68,82,87. 125, 188 photopotential, 109, 117,207 pitting potential, 221 rest potential, 102, 104, 125,221, 222 reversible potential, 39, 65, 138, 185 standard potential, 19. 39 Potential barrier: see Surface barrier Potential bias, 105, 114 Potential distribution, 184. 192, 195 Potential of zero charge, 28, 104 Pourbaix diagram, 21 Powder, 170, 177 Premature darkening, 262 Pressure, 170, 291 Primer, 316, 323 Quenching, 232 Quantum yield, efficiency, 110, 113-117 Radiation, 241 Radicals, 112, 114,285 Rain, 158, 163, 164,242 galvanic corrosion, 211 synthetic, 271 Rebar, 359
472
Recombination, 98,110, III, 113-115 Recrystallization, 2, 12 Red rust, 256, 320 Redox couple, 98,105-108 corrosion potential, 127 dissolution of ZnO, 122, 123 Redox potential, 120 Repassivation,91 Resin, 316 Resistivity: see Conductivity Resource, zinc, I Rock,305 Rolling, 303 Rolled zinc, 7 Roof,249 Rotating electrode, 40, 51, 68-70, 82, 123, 143 Roughness: see Surface Rubber, 337,401 Runoff, 164,250 Sacrificial protection: see Galvanic protection Salinity, 252 Salt, 358 Salt layers, 67, 88, 223, 266, 270, 318, 325 Salt spray test, 266, 270, 318, 325 Salvation, 36 Sample, for testing, 188, 195, 196 orientation, 72, 191, 254, 262 position, 201, 301 preparation, 143 shape, 191, 254 size, 191,201 Sand,305,311,352 Scanning rate, 156 Scribe, 323 Sea water, 169-171,284,291 corrosion rates, 291 galvanic corrosion, 184, 187,211 intergranular corrosion, 227 Sealing, 313, 377 Season, 158,247,249 Self corrosion, 214, 403 Semiconducting behavior, 75, 89, 93-124 polarity reversal, 202, 208 Sensitization, 110 Sheet, 237,255,258,262 Sheltering corrosion products, 160-166 corrosion rates, 255 galvanic corrosion, 211 Sherardising, 7 Ship, 215 Silt, 305 Silicate, 313, 338 Silicon, 9, 257
INDEX Silver, 260, 387, 391 Slip, 3 Snow, 242 Soil,305 galvanic corrosion, 184,211 pitting, 220 Solidification, II. 353 Solubility air, 49 alloying elements, 3, 229, 230 hydrogen, 39 minerals, 284 oxygen,48,49, 298 passive films. 87. 91, 299 salts in organic solvents, 393 zinc compounds, 22, 66, 91, 158, 171. 299 Solutions, 171-176,201,204,219.296; see also Electrolytes and Ions Solvation, 274, 280, 393, 396 Solvent, 170.393 Space charge layer, 98, 100. 104-106, 108 Spalling, 358, 364 Spangle, 10 Spectra, 116, 117 Splash zone, 294, 365 Spray, 6, 7 Stability carbonates, 23 concrete, 358 corrosion products, 31, 165, 170, 176 hydroxides, 21 oxide, 21,119-124 passivation, 87-92, 299 Stain. 257, 268, 349 Stainless steel. 184, 196, 198, 210 Steam, 235 Steel: see Iron Steady state, 35, 135, 153. 155 Stern-Geary constant. 127 Stirring, 139. 168, 173 Strength, 227, 234, 317, 352. 359 Stress, 90, 231, 239, 303 Strain, 332 Stress corrosion, 183. 184, 238 Superplasticity, 2 Surface area, 143, 149,341.343,375,390 Surface barrier, 105-110 Surface condition, 143, 144.202,347,363 Surface films; see also Corrosion products and Passive films catalytic effect, 45, 75 classification, 67 color, 74, 75, 80, 85-87, 204 corrosion potential, 207 formation, 75-78, 80, 84, 399
473
INDEX Surface films (conI.) impedance. 62. 63. 79 inhibitive effect. 3 I. 5 I metallic, 73, 80, 86 nature, 75, 77-80, 84-86, 89, 178-181 phase transformation, 168-17 I. 208 semiconducting, 102, 115-118,202 stability, 78,87-92 Surface roughness, 149, 150 Surface states, 101.103, 105, 107, 117, 140 Surface structure, 242 Sulfur dioxide adsorption, 245 air pollutants, 243 corrosion mechanism, 280 corrosion rates, 249, 280, 400 intergranular corrosion, 233 solubility, 249 Surface treatment. 260, 302, 388 chromating, 16, 180 gal vanizing, 12 phosphating, 15 plming, 13 ZnO crystal, 102, 103, I I 1 Sweep rate, potential, 69, 80, 82, 88 Tafel slope corrosion current, 125-127, 153, 156 determination, 127, 128, 156 dissolution, 30, 33, 34, 36-38 hydrogen reduction, 40, 41, 48 oxygen reduction, 49 Tank,208, 214,219,237,268,403 Temperature, 140,203 anode efficiency, 406 corrosion products, 168-170 corrosion rates, 14 I. 390 passivation, 78, 80, 85 pi tli ng, 217, 220 polarity reversal, 203 potential,20, 140, 141 semiconductor, 97 solubility, 48 Texture, 10, 15, 232, 303, 306, 390 Thermodynamics, 19-25 Thin-layer electrolyte, 274 capacitance, 28 concentration, 276 corrosion potential, 142 gal vanic action, 176, 193, 333 hydrogen reduction, 276 oxygen reduction, 53, 279 Throwing power, 213, 217, 333 Tidal zone, 293, 294, 364 Time, effect, 149, 144,206
concrete curing, 353 Time, effect (conI.) corrosion current. 144, 150, 152 corrosion potential, 149. 151. 152 corrosion products, 158, 163-165, 170 corrosion rates, 253, 293 galvanic corrosion, 193 gassing, 379 paint curing, 339 polarity reversal, 206 Time of the day, 253 Time of the year, 163 Time of wetness, 248, 252, 280 Tin, 184, 198,230,260,387 Titanium, 184, 185, 196, 198 Transient effect, 35, 154, 184 Transpassive state, 65 Transport number. 27 Tunnelling, 108. 109. 115 Twinning, 3 Ultraviolet light, 94, 317 Under-paint corrosion. 315 Under-vehicle corrosion. 263 Underground. 213. 217, 305 Uniform corrosion, 185 Valence, 32 Vanadium, 260 Vapor, 178.228, 269. 387 Viscosity, 73, 393 Wall. 249 Water, 9 in concrete, 355 content in air, 242 corrosion products, 168-171. 177 intergranular corrosion. 227 pitting, 219 polarity reversal, 203 in solvents, 393 Water drop, 274 Water line, 301, 31 I Water spray, 270 Weather, 252, 339 Weathering, 305, 347 Weight loss relation to corrosion current, 129. 151-156 relative to corrosion products, 163 units, 141-143 Wetting, 177,241,271,325,344 Wet storage stain. 183.236.262 White rust. 237 Wire, 176,209. 254. 359 Wind. 242. 245
474 Woods, 401 X-ray, 94, 157,207,224 Yellow rust, 257 Zinc alloys; see also Alloying and Zinc coatings battery anodes, 387 corrosion current and potential, 144-149 corrosion products, 176 die-casting. 3, 7 sacrificial anodes, 403 Zinc coatings, 12, 14,321. 407 applications, 6. 322 classification, 3 composite coatings, 323 corrosion products, 180 duplex coatings, 337 gal vanic corrosion, 216 production methods, 7-14 under-paint corrosion, 321 under-vehicle corrosion, 263 Zn-AI, 12, 145,324
INDEX Zn-Co, 118, 145-147, 181 Zinc coatings (cant.) Zn-Cr, 145,216 Zn-Cu, 145. 146,216 Zn-Fe, 12, 145, 147, 148,263,323 Zn-Mg. 145,216 Zn-Ni, 118, 145-147, 184,216 Zn-Ti, 145,216 Zinc dust, 7, 316, 338, 341 Zinc oxide, 93-124 applications, I, 93, 338 corrosion products. 157.263,399 oxide films: see Surface films photo electrochemistry. 107 physical properties, 93-97 semiconducting property, 97 stability, 107-109 Zinc powder, 7, 375, 381,389 Zinc products, 3, 6 Zinc-rich paint, 183, 184,215,337 Zinc tape, 7 Zincate, 382 Zincrometal, 338
