Composition effects on the whiteness and physical-mechanical properties of traditional sanitary-ware glaze

Available online at www.sciencedirect.com

ScienceDirect Journal of the European Ceramic Society 35 (2015) 3735–3741

Composition effects on the whiteness and physical-mechanical properties of traditional sanitary-ware glaze K. Boudeghdegh a,∗ , V. Diella b , A. Bernasconi c , A. Roula d , Y. Amirouche a a

LEAM, University of Jijel—B.P 98 Ouled Aissa, 18000 Jijel, Algeria National Research Council, IDPA, Section of Milan—Via Botticelli 23, I-20133 Milan, Italy c European Synchrotron Radiation Facility, 71 Avenue des Martyrs, 38000 Grenoble, France d LIME, University of Jijel—B.P 98 Ouled Aissa, 18000 Jijel, Algeria

b

Received 9 January 2015; received in revised form 28 April 2015; accepted 3 May 2015 Available online 19 May 2015

Abstract Whiteness and physical-mechanical characteristics of sanitary glazes are usually controlled and improved by changing chemical composition of the raw materials. In this study, ten glazes of industrial interest, with different composition, were prepared by spraying, using the traditional ceramic substrate, and then thermally treated in industrial tunnel kiln at temperature of 1250 ◦ C. The obtained glazes were characterized by X-ray diffraction in order to reveal their mineralogical composition, also confirmed by FTIR spectra, and were observed by backscattered electron images to study their microstructure and to derive the thermal expansion coefficients. Whiteness characteristics were obtained by a Micro color colorimeter data station. The results showed a difference of 10% between calculated and experimentally derived thermal expansion coefficients and the improvement of whiteness (up to 85.7%) and flexural strength (47.61 MPa) at a maximal content of zircon (14.5 wt%) and a low content of ZnO (2.5 wt%). © 2015 Elsevier Ltd. All rights reserved. Keywords: Physical-mechanical properties; Whiteness; Glaze; Ceramic sanitary-ware

1. Introduction Ceramic glazes are thin glassy layer on ceramic products surface which play both decorative and functional basic roles. Those roles are mutually combined, and, from a commercial point of view, they mainly depend on surface properties of the glazes.1–4 The whiteness of the ceramic bodies (granted by the deposited glaze) is the main esthetical factor that determines the quality of any product. It must meet the requirements for domestic sanitary ware (Wh. > 80%) and be attractive, at lower prices, for users. ZrO2 , ZnO, TiO2 , and SnO2 are used as opacifying agents. ZrSiO4 was proved to enhance opacity, especially in the presence of ZnO with a SiO2 /A12 O3 molar ratio equal to 10.1,2 The coverage capacity and wetting characteristics of a glaze are controlled by its surface tension, contact angle, viscosity and opacity.5 The properties of glaze depend on a variety

∗

Corresponding author. Tel.: +213 662 154111; fax: +213 34501189. E-mail address: [email protected] (K. Boudeghdegh).

http://dx.doi.org/10.1016/j.jeurceramsoc.2015.05.003 0955-2219/© 2015 Elsevier Ltd. All rights reserved.

of aspects: the effects of raw materials and oxide compositions on one hand and the firing conditions on the other hand.6,7 The final formulation of the enamel depends on both physical as well as esthetical properties. Among the first, important properties are the maturation time, the thermal expansion coefficient values and the reactivity with the ceramic support. In particular, the thermal expansion coefficient value of the enamel must be close to the body’s one (in most cases around 7 × 10−6 ◦ C−1 ) to avoid strains on the tile8 and several investigators have proposed methods for calculating these coefficients.9,10 Whiteness, depending on the amount of coloring oxides in the raw materials, is used to compare ceramic pieces and evaluates their quality. For a white ceramic, the amount of (Fe2 O3 , TiO2 ) should not exceed 0.3–0.4 wt%. Experiments showed that the zircon gives a large increase in whiteness and shine, extends the softening range up to 150 ◦ C, and hence correspondingly increases the firing range of the glaze.11 The zircon amounts should be not less than 10%.12 From a phase composition point of view, glaze is generally constituted by a dominant amorphous phase in which other crystalline phases (such as zircon, diopside, wollastonite,

3736

K. Boudeghdegh et al. / Journal of the European Ceramic Society 35 (2015) 3735–3741

and quartz) are dispersed providing opacity and giving a thermal shrinkage fitting with the ceramic bulk.13,14 The assessment of the quality of a glaze mainly relies on its fusibility behavior and rheological properties, on its color, opacity and on its thermal expansion.15 In the present study, we investigated the effect of varying the chemical composition (amount of raw materials and accordingly the oxides contents) on the glaze properties (whiteness, flexural strength, etc.) in order to improve the quality of the obtained sanitary ware glaze. We prepared ten different glazes: (i) three samples varying the zircon and ZnO quantities, (ii) four samples varying the dolomite grain size distribution, (iii) two samples using talc in substitution of dolomite and one sample as reference. The glazes were characterized using X-ray powder diffraction image processing of backscattered electron images, electron microprobe analysis and the results used to calculate the thermal expansion coefficients. We measured whiteness, flexural strength, porosity and chemical durability, and calculated fusibility, surface tension and thermal expansion coefficients, these lasts were compared with the values resulted from experimental data.

Table 2 Starting molar composition in oxides of the glazes on study. Sample

Basic oxides

G1

0.128 Na2 O 0.007 K2 O 0.602 CaO 0.164 MgO 0.099 ZnO 0.130 Na2 O 0.007 K2 O 0.610 CaO 0.166 MgO 0.086 ZnO 0.132 Na2 O 0.007 K2 O 0.619 CaO 0.168 MgO 0.073 ZnO 0.157 Na2 O 0.009 K2 O 0.527 CaO 0.168 MgO 0.139 ZnO 0.127 Na2 O 0.007 K2 O 0.593 CaO 0.161 MgO 0.112 ZnO 0.239 Na2 O 0.013 K2 O 0.379 CaO 0.322 MgO 0.047 ZnO 0.127 Na2 O 0.009 K2 O 0.603 CaO 0.179 MgO 0.082 ZnO

G2

G3

G4

G5, G6, G7, and G8

2. Experimental procedure 2.1. Preparation of glazes

G9

All experiments were prepared in the laboratories of the Société Céramique Sanitaire El-Milia, (Jijel, Algeria). The raw materials used for preparation of the glazes were kaolin, sodium feldspar, quartz, calcite, zircon, dolomite, and zinc oxide, all of industrial grade, and the overall composition in weight % is reported in Table 1, together with a reference sample Gref . The raw materials were milled in humid conditions up to 1% residue using 63 microns sieve.16,17 and the experimental glazes were prepared by grinding the necessary weight of quartz and dolomite or (talc) in the rotary jar and mixing it with 3/4 of the 40 wt% of water for 2–3 h with few drops of sodium silicate (to avoid flocculation) and adding the remaining water and raw materials afterwards. The oxide compositions of the studied glazes (in molar fraction: Table 2) are calculated according to the Seger formula18 where SiO2 is the major oxide ranging from 1.147 to 4.236,

Gref

Neutral oxides

Acid oxides

0.199 Al2 O3 0.003 Fe2 O3

2.203 SiO2 0.002 TiO2 0.164 ZrO2

0.201 Al2 O3 0.003 Fe2 O3

2.242 SiO2 0.002 TiO2 0.172 ZrO2

0.204 Al2 O3 0.003 Fe2 O3

2.281 SiO2 0.002 TiO2 0.181 ZrO2

0.269 Al2 O3 0.006 Fe2 O3

2.862 SiO2 0.002 TiO2 0.193 ZrO2

0.196 Al2 O3 0.003 Fe2 O3

2.166 SiO2 0.001 TiO2 0.156 ZrO2

0.409 Al2 O3 0.010 Fe2 O3

4.236 SiO2 0.003 TiO2 0.263 ZrO2

0.222 Al2 O3 0.003 Fe2 O3

2.148 SiO2 0.002 TiO2 0.105 ZrO2

while Al2 O3 content varies in the range from 0.145 to 0.409. Finally, for each composition, raw mixture was sprayed upon a traditional ceramic slip (6–9 wt% Kaolin Remblend (RMB), 33–34 wt% sodium feldspar, 24–26 wt% quartz), and then thermally treated at temperature of 1250 ◦ C under industrial conditions.

Table 1 Raw materials (wt%) used for glazes preparation. Sample

Kaolin RMBa

Sodium feldspar

Quartz

Calcite

Zircon

Dolomite

Dolomite grain size (␮m)

Talc

ZnO

G1 G2 G3 G4 G5 G6 G7 G8 G9 Gref

6 6 6 6 6 6 6 6 6 6–9

34 34 34 34 34 34 34 34 38 33–34

26 26 26 26 26 26 26 26 27 24–26

10 10 10 10 10 10 10 10 5 8–10

13.5 14 14.5 13 13 13 13 13 13 7–9

7 7 7 0 7 7 7 7 0 7–8

63 63 63 63 63 65 45 <45 63 63

0 0 0 7 0 0 0 0 10 0

3.5 3 2.5 4 4 4 4 4 1 2–3

a

Kaolin Remblend.

K. Boudeghdegh et al. / Journal of the European Ceramic Society 35 (2015) 3735–3741

2.2. Experimental data Glaze phase composition and microstructure were investigated by X-ray powder diffraction and Backscattered Electron Images. For the former, small portion of glaze were removed from each fired tile by using a metallic indentation. After checking with stereo-microscope the mere presence of glaze, the material were grounded in a zirconia mortar and then analyzed with an X-Pert Panalytical Diffractometer. Data were collected in the 5–60◦ 2θ range, with a step size of 0.02◦ and a counting time of 30 s. Afterwards, diffraction patterns were used to qualitatively evaluate the crystalline phases involved in each glaze. For the latter, each sample was inspected using a JEOL JXA8200 Electron Microprobe at the laboratory of the Department of Earth Science, University of Milan, Italy. All the different ceramic tiles were transversely cut and embedded into araldite, for each sample, after polishing; four backscattered electron images were collected and then processed with Image Pro Plus software, Media Cybernetics, Inc., 2001, in order to extract information about the spatial distribution of the crystalline phases. Following the strategy described in,15 Image Fraction Value (IFVi,j ) for each i-phase was determined as the number of pixel attributed to the ith phase (in the jth region) divided by the total number of pixel of the jth region (note that the black pixel, assigned to pores, were not included in this quantity). Moreover, due to the fact that: (i) the residual number of crystalline phases was very small (from X-ray powder diffraction patterns only zircon and quartz were identified) and, (ii) such crystalline phases were distinguishable from the dominant glassy matrix in terms of gray scale pixel, we were able to extract a quantitative information about them, following Eq. (1): Wij (%) =

ΣN Vi (%)ρi Σn ΣN Vi (%)ρi

(1)

where Wi,j (%) is the wt% of the ith phase in the jth region; Vi is the volume of the ith phase; ρi is the density of the ith phase and N, n are the numbers of regions and phases, respectively. Note that this equation is based on the numerical approximation that Vi (%) is equal to the sum of all the IFV values of the ith region, divided by N, while phase densities values were taken from tabulation. Additionally, zircon determinations obtained from imaging were cross-checked by a chemical recalculation approach: 40 chemical micro-analyses were collected in the glassy matrix and then zircon determinations were performed by applying Eqs. (2) and (3): WZrO2 , starting = Xglass WZrO2 + 65(100 − Xqtz − Xglass )

(2)

Xzircon = 100 − Xqtz − Xglass

(3)

where WZrO2 , starting is the starting weighed amount of zirconium oxide; WZrO2 is the measured zirconium oxide amount in glass (obtained by micro-analyses); Xqtz is the quartz amount obtained with imaging and 65 is the stoichiometric zirconium oxide amount in zircon.

3737

Infrared transmission spectra were carried out using the same weight of the glaze powder dispersed in KBr pellets and the data recorded by IRAffinity-1 Shimadzu Japan spectrophotometer in the range from 400 to 1500 cm−1 . Values of residue, density and fluidity were obtained using classical methods and apparatus: residue sieving the samples up to 63 ␮m, density weighing a specific enamel volume (ranging from 1.75 to 1.77 g/cm3 ), and fluidity corresponding to a outflow time from a 2.6 mm Ford cup viscosimeter of equivalent to 23–40 s. Whiteness (Wh) was estimated with a colorimeter Dr Lange micro color DATA STATION (in accordance with DIN 5033). Flexural strength (Sfl ) was determined according to the three points test method ASTM D 790-80, measuring three points flexural strength and using the formula: Sfl = (3 × P × L)/(2 × b × h2 ), where P, L, b and h are the breaking load (N), the distance (mm) between supports, the width (mm) and the height (mm) of the test piece, respectively. Porosity was determined (according to the reference standard NF.D 14512) using the formula Ptot = 100 × (Pi − Pf )/Pi measuring Pi and Pf , the initial and final weight of the sample. Finally, chemical durability was evaluated washing the glazes surfaces of the samples with methyl alcohol, then treated in 3% HCl and 3% NaOH solutions for 7 days at room temperature.19 Afterwards, the glazes surfaces were visually inspected (with a hand lens for color, gloss, crazing, bubbles peeling20 ). 2.3. Theoretical calculations Because of the dependence from the chemical composition, the coefficients of thermal expansion, coefficient of fusibility and surface tension of glazes were computed as following: - thermal expansion coefficients (α) using the additivity law13,21,22 that admits that each oxide has a different contribution to the final α value, according to Appen formula by Eq. (4):  αι pi 10−7 αglaze = (4) 100 where αι is the thermal expansion coefficient of oxide i (◦ C−1 ), taken from22 and pi is the starting molar % of oxide i in the sample. Furthermore, αglaze can be also predicted by: (i) assuming αzircon = 12.3 ◦ C−1 in the 293–1293 K range from,23 αquartz = 24.3 ◦ C−1 in the 298–773 K range from,23 (ii) using Xquartz from imaging, Xglass from chemical microanalysesby Eq. (2), and Xzircon from chemical micro-analyses by Eq. (3), (iii) calculating αglass(20–400 ◦ C) with pi that come from the chemical micro-analysis in the glassy matrix, and (iv) referring to the Fei simple model,23 and we can write: αglaze(20−400 ◦ C) = (αglass(20−400 ◦ C)Xglass + αquartz (20−400 ◦ C)Xquartz + αzircon (20−400 ◦ C)Xzircon )

(5)

3738

K. Boudeghdegh et al. / Journal of the European Ceramic Society 35 (2015) 3735–3741

neglecting the effect due to the contact surfaces between phases and assuming that expansion for crystalline phases is isotrope, which means that αvol = 3αlin . - Coefficient of fusibility (F) calculated by Eq. (6): F=

a1 n1 + a2 n2 + · · · + ai ni b1 m 1 + b 2 m 2 + · · · + b i m i

(6)

where ai is the coefficient of fusibility of easy-fusible oxides; bi is the coefficient of fusibility of difficult-fusible oxides and n, m are their corresponding amounts.22 - Surface tension (K) estimated by Eq. (7): K=

n 

Table 3 Image Fraction Value (IFV) of zircon and quartz in the studied glazes. Sample

IFV (zircon)a

SDb

IFV (qtz)c

SDb

Delta zircond

G1 G2 G3 G4 G5 G6 G7 G8 G9 Ref

6.87 7.6 7.93 8.19 5.03 6.98 7.34 6.05 6.82 3.72

0.53 11 1.46 0.8 1.25 0.46 1.37 1.16 0.32 0.16

0.17 0.43 0.28 1.46 0.22 0.21 0.34 0.32 7.32 0.92

0.22 0.41 0.31 0.29 0.18 0.24 0.47 0.23 2.05 0.63

2.41 2.47 2.51 1.33 2.39 2.53 2.54 2.16 1.02 2.25

a

ci χi

(7)

i=1

where c is the percentage weight of the oxide components, x the related surface tension value and n the number of components.24,25 3. Results and discussions Preliminary X-ray powder diffraction measurements of glazed samples treated at 1250 ◦ C underlined the presence of zircon and quartz as crystalline phases while the large hump in the 15–35◦ 2θ range is addressed to a large glass (amorphous phase) content, as displayed in the example of Fig. 1. Subsequently, image processing was performed as described in Section 2.2 and the results of crystals distributions are summarized in Table 3. Zircon IFV ranges from 3.72 to 8.19 and, in some cases, as highlighted by the standard deviation, zircon dispersion is not optimal, especially in the samples with the highest starting zircon amount. This is probably caused by the fact that an increased number of zircon particles may promote their aggregation, as displayed in Fig. 2A, belonging to sample G8, where some aggregates can reach up 100 ␮m of size. Thus, a longer milling time, coupled with the use of some additives, could minimize this effect.

Fig. 1. X-ray powder diffraction patter of the glaze G9 in the 5–60◦ 2θ range.

b c d

Zircon Image Fraction Value. Standard deviation. Quartz Image Fraction Value. Difference between starting zircon and the zircon.

On the other side, the very low quartz IFV (from 0.17 to 1.46) evidences that its amount is limited. The only exception is given by the sample G9 where an IFV of 7.32 was observed probably due to a low content of calcite and dolomite that will be explained later on. Fig. 2B shows such remarkable dispersion of quartz grains in sample G9 (gray color, intermediate between pores and glass). In all sample, quartz IFV standard deviation is close to IFV itself, suggesting that, at the adopted scale, quartz is a heterogeneously distributed phase in glaze. If IFV values are converted into weight % following Eq. (1), zircon and quartz contents range from 6.04 to 12.50 wt% and 0.15 to 6.82 wt%, respectively. On the opposite, zircon weight % obtained following Eqs. (1) and (3), lies between 6.75 and 11.98; with a linear R2 factor of 0.83 between these two distinct zircon determinations. For zircon, chemical micro-analysis are more accurate than imaging one, due to the fine particle size of zircon that makes the results extremely sensitive to the selected gray scale threshold values. Therefore, only zircon contents obtained with Eq. (1) will be considered. Furthermore, the evaluation of the difference between the starting and the final zircon contents is very important to understand how dissolution acts on the opacifier phase and to calibrate the most appropriate amount to load before firing. In all samples, this difference ranges from 2.16 to 2.54 (Delta zircon in Table 3) with the exclusion of sample G9 and G4 where the extremely low fluxing agents amount (no dolomite is loaded differently from all the others) makes the dissolved zircon halved. A similar effect is evident also if one considers the difference between starting and residual quartz: this difference is lower for sample G9 and G4, especially for sample G9 where a halved starting calcite amount (i.e., 5 wt% instead 10 wt% in all the others) brings to a less marked dissolution of quartz, as remarked by the diffused residual quartz grains in Fig. 2B. Fig. 3 shows the mid-infrared region (1500–400 cm−1 ) FTIR spectra of the investigated glazes. The characteristic FTIR bands of silica are observed in (1032–1035 cm−1 ) range corresponding to the asymmetric vibration of the bridging Si–O–Si bands within [SiO4 ] tetrahedra. The band at 785 cm−1 , is associated with Si–O–(Si,Al) symmetric stretching vibrations between the

K. Boudeghdegh et al. / Journal of the European Ceramic Society 35 (2015) 3735–3741

3739

Fig. 2. Backscattered electron images of two selected samples G8 (A) and G9 (B).

Fig. 3. IR spectra of studied glazes.

tetrahedra. The band at 468 cm−1 is due to the bending vibration of Si–O–Si and Si–O–Al linkages.26,27 The band at 615 cm−1 in the case of the all samples could be attributed to the Si–O–Zr vibrational modes.19,28 These observations suggest that the studied glazes exhibit a disordered structure29 and FTIR spectra are in agreement with X-ray powder diffraction findings. Porosity (ranging from 0% to 0.4%) and chemical durability are comparable with those of the reference sample, such as the calculated surface tension (ranging from 340 to 363 dyne/cm). Coefficients of fusibility resulted 0.2 for all studied glazes, including the reference one.

Whiteness (Wh), flexural strength (Sfl ) and thermal expansion coefficients of studied samples are reported in Table 4 where, in particular, the thermal expansion coefficients were calculated by Eq. (4) using the starting compositions (fourth column), by Eq. (5) (fifth column) and by Eq. (4) using the chemical composition of the glass from the microprobe analyses (sixth column). All reported values are the average calculated, at minimum, on three samples. Considering G1, G2, and G3, the substitution of zinc oxide with zircon improves the flexural strength (up to 47.61 MPa) and the whiteness, even if for this latter the trend is not monotonic. This can mean that zircon has a role in the enhancement of whiteness. Comparing samples G4 and G5, with the same zinc oxide quantity but containing dolomite (G5) instead of talc (G4), it is possible to evidence the role of dolomite which lowers the melting temperature, decreases the final zircon content and therefore the whiteness. Different grain size of dolomite does not appear to affect the final zircon amount and the other considered properties even if a little increase in the flexural strength may be noted. The flexural strength of the samples are greater than the reference one, save G9 where glass phase has the lowest value. The highest value of the flexural strength (47.61 MPa) together with a high value for whiteness (85.7%) was registered in the sample G3, containing the highest amount of zircon (14.5 wt%) and the lowest content of ZnO (2.5 wt%). The calculated and experimental thermal expansion coefficients of the glaze differ by about 10%, both similar to the typical substrate values (4.21 × 10−6 ◦ C−1 ). Sample 4 and 9

Table 4 Whiteness, flexural strength, thermal expansion coefficients of studied samples. Sample

Whiteness (%)

Flexural strength (MPa)

Calculated thermal expansion coefficient αglaze (10−6 ◦ C−1 )

Experimental thermal expansion coefficient αglaze (10−6 ◦ C−1 )

Experimental thermal expansion coefficient αglass (10−6 ◦ C−1 )

G1 G2 G3 G4 G5 G6 G7 G8 G9 Gref

84.5 80.9 85.7 86.7 82.6 85.4 83 82.6 79.9 81.5

41.61 42.14 47.61 37.52 36.60 42.80 47.10 40.26 30.28 38.25

6.03 6.01 5.99 5.57 6.05 6.00 6.00 6.00 5.15 6.22

6.86 6.79 6.84 6.14 6.95 6.82 6.78 6.68 5.92 6.83

7.21 7.14 7.19 6.38 7.28 7.13 7.09 6.99 6.00 6.96

3740

K. Boudeghdegh et al. / Journal of the European Ceramic Society 35 (2015) 3735–3741

Fig. 4. (a–d) Influence of (a) ZnO on whiteness; (b) ZrO2 on flexural strength; (c) SiO2 /Al2 O3 ratio on whiteness; and (d) SiO2 /Al2 O3 on flexural strength of studied glazes.

exhibit the lowest values, due to the presence of talc in place of dolomite, which in turn determines a higher content of SiO2 , that is an oxide with a small Appen coefficient, respect the others. Moreover, systematically, when comparing experimental αglaze and αglass (columns 4 and 5 in Table 4, respectively), one can observe that αglaze is higher than αglass , due to the presence of zircon that expands quite less than glass phase. In Fig. 4a we report the plot of whiteness values versus ZnO content; a positive influence of this oxide on the esthetical property may be recognized. Moreover, as previously reported in literature,1,2 also the amount of zircon contributes to the improvement of this property. Flexural strength depends mainly on the ZrO2 content (Fig. 4b) which has a stronger effect at high SiO2 /Al2 O3 ratios. In Fig. 4 c and d, plots of whiteness and flexural strength against SiO2 /Al2 O3 ratio values show that the simultaneous presence of high ZrO2 and SiO2 /Al2 O3 contents enhances the two properties.

- the replacement of dolomite with talc improves the whiteness, probably because talc preserves zircon in the final product, and reduces the thermal expansion coefficients, but implies a low flexural strength; - the granulometry of dolomite seems to be the parameter less significant in the evolution of glaze; - the experimental thermal expansion coefficients resulted close to the desired values of the body (4.21 × 10−6 ◦ C−1 ); - whiteness and flexural strength depend on starting chemical composition, in particular on the SiO2 /Al2 O3 ratio and ZrO2 content; - G3 sample, containing 14.5 wt% of zircon and 2.5 wt% of ZnO, shows better physical-mechanical properties and esthetic whiteness when compared to the reference sample (Gref ), and those enhancements were achieved with a lower cost production of the final product. Acknowledgements

4. Conclusions In this work the influence of the composition on the properties of different zircon-containing glazes was investigated. It is concluded that: - zircon in substitution of zinc oxide improves the mechanical properties of the glaze but zinc oxide plays an important role in the control of whiteness;

The authors are grateful to the manager and to the production staff of SCS El-Milia Company. Thank are due to the colleagues from laboratories of Earth Sciences Department at University of Milan. References 1 Eppler RA, Eppler DR. Glazes and glass coatings. Westerville, Ohio: American Ceramic Society; 2000.

K. Boudeghdegh et al. / Journal of the European Ceramic Society 35 (2015) 3735–3741 2 Taylor JR, Bull AC. Ceramic glaze technology. Oxford: Pergamon Press; 1986. 3 Esposito L, Tucci A, Rastelli E, Palmonari C, Selli S. Stain resistance of porcelain stoneware tiles. Am Ceram Soc Bull 2002;81(10):38–43. 4 Yekta BE, Alizadeh P, Razazadeh L. Floor tile glass-ceramic glaze for improvement of glaze surface properties. J Eur Ceram Soc 2006;26:3809–12. 5 Sainz IG. Physical-chemical characteristics of ceramic glazes and their influence on quality of floor and wall tiles. Tile Brick Int 1990;6:21. 6 Froberg L, Kronberg T, Hupa L. Effect of soaking time on phase composition and topography and surface microstructure in vitro crystalline whiteware glazes. J Eur Ceram Soc 2009;29:2153–60. 7 Karasu B, Dolekcekic E, Ozdemir B. Compositional modifications to floor tile glazes opacified with zircon. Br Ceram Trans 2001;100:81–5. 8 Amorós JL. Vidrados para pavimentos e revestimentos cerâmicos. Evoluc¸ões e perspectivas: Parte I. Cerâm Ind 2001;6(4):41–50. 9 Cook RL, Brunner CD. Correlation of glaze-body stresses with thermal properties of white ware bodies. J Am Ceram Soc 1949;32(12):401–8. 10 Fluegel A. Thermal expansion calculation of silicate glasses at 210 ◦ C based on the systematic analysis of global databases. Jena, Germany: Glass properties.com; 2007. 11 Swiecki Z. White opaque zircon glaze for sanitary ceramics. Sklar a Keramika (Czechoslovakia) 1964;3:73–5. 12 Avgustinik AI. Ceramic. Leningrad: Stroy Izdatelstvo; 1975. 13 Rasteiro MG, Gassman T, Santos R, Antunes E. Crystalline phase characterization of glass-ceramic glazes. Ceram Int 2007;33:345–54. 14 Schabbach LM, Bondioli F, Ferrari AM, Manfredini T, Setter CO, Fredel MC. Color in ceramic glazes: analysis of pigment and opacifier grain size distribution effect by spectrophotometer. J Eur Ceram Soc 2008;28:1777–81. 15 Bernasconi A, Diella V, Marinoni N, Pavese A, Francescon F. Influence of composition on some industrially relevant properties of traditional sanitaryware glaze. Ceram Int 2012;38(7):5859–70.

3741

16 Atkinson I, Teoreanu I, Zaharescu M. Correlation among composition properties of SnO2 opacified glazes. UPB Sci Bull Ser B 2008;70(3):11–22. 17 Atkinson I, Teoreanu I, Zaharescu M. Characterization of multicomponent oxides system for glazes applications. Rev Him (Bucureti) 2008;59(9):990–3. 18 Singer F, Singer SS. Industrial ceramics. London: Chapman and Hall; 1979. 19 Popa M, Calderón-Moreno JM, Popescu L, Kakihana M, Torecillas RJ. Crystallization of gel-derived and quenched glasses in the ternary oxide Al2 O3 –ZrO2 –SiO2 system. J Non-Cryst Solids 2002;297:290–300. 20 Mete Z, Tanisan H. Ceramic technology and applications, Izmir, vol. I; 1986 [in Turkish]. 21 Amorós JL, Blasco A, Carceller JV, Sainz V. Acordo esmalte-suporte (II): expansão térmica de suporte e esmaltes cerâmicos. Cerâm Ind 1997;2(1/2):8–16. 22 Appen AA. The chemistry of glass. Khimiya: Leningrad; 1974 [in Russian]. 23 Fei Y. Thermal expansion. In: Ahrens TJ, editor. Mineral physics and crystallography: a handbook of physical constants, vol. 2. AGU Reference Shelf; 1995. 24 Yalc¸in N, Sevinc¸ V. Utilization of bauxite waste in ceramic glazes. Ceram Int 2000;26:485–93. 25 Pamelee CW. Ceramic glazes. New York: Cahners Books Division of Cahners Publishing Company; 1987. 26 ElBatal FH, Azooz MA, Hamdy YM. Preparation and characterization of some multicomponent silicate glasses and their glass–ceramics derivatives for dental applications. Ceram Int 2009;35(3):1211–8. 27 Merzbacher CI, McGrath KJ, Higby PL, Si NMR. and infrared reflectance spectroscopy of low-silica calcium aluminosilicate glasses. J Non-Cryst Solids 1991;136:249–59. 28 Popa M, Kakihana M, Yoshimura M, Calderon-Moreno JM. J Non-Cryst Solids 2006;352:5663–9. 29 Roy BN. Infrared spectroscopy of lead and alkaline-earth aluminosilicate glasses. J Am Ceram Soc 1990;73:846–55.