Accred Qual Assur (2003) 8:231–234 DOI 10.1007/s00769-003-0588-x
J. M. F. Nogueira P. Serôdio
Received: 13 June 2002 Accepted: 9 December 2002 Published online: 1 April 2003 © Springer Springer-V -Verlag erlag 2003 J. M. F. No Nogu guei eira ra (✉) · P. P. Ser Serôdi ôdio o Departamento de Química e Bioquímica da Faculdade de Ciências da Universidade de Lisboa, Campo Grande Ed. C8, 1749–016 Lisboa, Portugal e-mail:
[email protected] Tel.: +351-21-7500899 Fax: +351-21-7500088 J. M. F. No Nogu guei eira ra Centro de Ciências Moleculares e Materiais, Campo Grande Ed. C8, 1749–016 Lisboa, Portugal
PRACTITIONER’S REPORT
Determination of the expiration date of chemical solutions
Abstract Thirty-five chemical solutions, acids, bases, complexing, argentometric, reducing, oxidising, salts and eluent mixtures, were studied continuously over a 7.5-month period. Twenty-seven remained reasonably stable and had acceptable expiration dates, which were equal to or more than 2 months longer than the established average deviation of ±2% relative to the initial value. Exceptions were made for potassium permanganat perma nganatee (0.1 M), perchloric perchloric acid (0.1 M) and disodium disodium hydrogen hydrogen phosphate (10%) solutions, as well as for HPLC eluent mixtures, which showed stability with expiration
dates equal to stated values or shorter by 1 month. The eight chemical solutions presenting expiration dating higher than 6 months were hydrochloric drochlo ric acid (0.1 (0.1 and 1 M), sulphuric acid acid (0.5 and 1.5 1.5 M), ethylethylenediamine tetraacetic acid (EDT (EDTA) A) (0.1 M), silver nitrate (0.1 M), ammonium thiocyana thiocyanate te (0.1 M) and iodine (0.1 M). The present present data data is a suitable guideline for the date of several chemical solutions routinely used in the analytical laboratories. Keywords Dat Datee · Chemical Chemical solution solu tionss · Stab Stabilit ility y · She Shelflf-life life · Chemical analysis
and regularly at routine intervals. Each time a bottle is opened, the reagent or solution is exposed to various exThe stability requirements for analytical laboratories de- ternal influences that may alter its integrity for a specific mands careful control of the expiration date of chemi- use. cals, reagents and solutions. The use of deteriorated or Some reagents and solutions, e.g. sodium hydroxide, expired chemical solutions affects the validity of analyti- easily react with carbon dioxide in the atmosphere. Temcal procedures, leading to erroneous data, which can perature, light, air, moisture and storage bottle material, cause consequences for the community [1]. etc. can also affect the physical appearance of the chemiHow can a laboratory establish the expiration dating cal solutions (easily recognizable as a colour change) of chemical solutions and reagents not provided by sup- sedimentation and chemical decomposition, leading pliers, or for chemical solutions prepared in-house? This sometimes to an erroneous concentration. On the other is of increasing concern in analytical laboratories which hand, the number of times a bottle is opened and for how follow good laboratory practice (GLP), especially for long are factors which may also influence the stability of those which implement a quality system [2]. the chemical solutions. The determination of the expiration date of a chemiShelf-life is defined as the period of time a product is cal solution is generally associated with its stability, considered usable for each application taking into acwhich is an intrinsic property affecting the maintenance count physical and/or chemical changes. Products conof the functional capacity of the reagent for a particular sidered to have extensive shelf-life when stored unanalytical application. In the laboratory, the stability of opened and under the proper conditions are not given an reagents and solutions should be checked just before use expiration date. To clarify a common misconception, “no
Introduction
232
expiration date” and “forever” are not synonymous. The expiration date may be understood as the date at/or after which the chemical solution may no longer be stable for a specific analytical application. Taking in consideration a wide range of applications, the expiration date of a chemical solution should be established on the basis of their specific use. Some solutions may be too old for one analytical application, but still suitable for another. For example, a sodium hydroxide solution used to standardize hydrochloric acid solutions may have reached its expiration date for this type of application, but it would still be suitable to adjust pH of acids prior to disposal, a less critical application. In this context, laboratories should establish their own internal standard operating procedures (SOPs) for the replacement of chemical solutions after a specific period of time. An expiration date should be assigned to each type of chemical solutions, i.e. acids, bases, complexing, argentometric, reducing, oxidising, salts, mixture of solvents, etc. For most reagents, with the exception of those with a manufacturer-specified expiration date, a colour code system may be also implemented to better identify expiration dates [3]. Chemical solutions are used for a wide variety of applications. Users should assess a chemical solution for its actual expiration date based on the application for which the chemical or reagent is being used, taking into consideration the time spend on preparation and the associated costs, as a policy to improve laboratory performance. In this context, there is no information in the literature regarding expiration dating of chemical solutions used in analytical laboratories. The aim of the present work was to evaluate the expiration date of several chemical solutions in common use, particularly in pharmaceutical laboratories, in order to establish a suitable guideline for analytical departments.
Table 1 Typical reactions involved in the titrimetric procedures adopted in the present study [4, 5, 6, 7, 8, 9]
Acids
Bases
Complexing Argentometric Reducing Oxidizing a Titrators b Acetic
in italics medium
Experimental In the present study calibrated glass material was used and all the reagents were analytical grade. The solutions were stored in amber glass bottles. Plastic bottles were used only for salt solutions. The bottles were opened exclusively for analytical purposes and kept in a closed cabinet, protected from light, at room temperature (20 °C). The solutions were studied by means of frequency of use and application demand, and rigorous concentration, such as acidic (HCl, H2SO4 and HClO4), basic (NaOH and NH3), complexing (ethylenediamine tetraacetic acid – EDTA; C10H16N2O8), argentometric (AgNO3 and NH4SCN), reducing (Na2S2O3), oxidising (I2 and KMnO4), salts (NH4CH3COO, (NH4)2HPO4, K2HPO4, KH2PO4, Na2HPO4, NaH2PO4, NaCl and Na 2SO4) and eluent mixtures (CH3OH/H2O, CH3CH2OH/H2O and CH3CN/H2O) for high performance liquid chromatography (HPLC). The expiration date was studied by the periodic determination of the concentration of each solution. Titrimetric procedures were employed to analyse acidic, basic, complexing, argentometric, reducing and oxidizing solutions. Table 1 shows the main reactions involved in the titrimetric procedures. Ion-exchange chromatography was used to analyse salt solutions, using a Dionex DX-500, with a quaternary pump (GP40), cationic and anionic suppressor columns (CSRS-I, 4 mm and ASRS-I, 4 mm), a conductivity detector (CD20), a column (IonPac CS12; AS4A, 250×4 mm) and a pre-column (IonPac AG4, 50×4 mm). For cationic (Na+ and K+) and anionic (CH3COO−, SO42− and PO43−) determination, methane sulphonic acid (0.02 M) and carbonate/bicarbonate (0.02 M/0.02 M) solutions were used as mobile phases having flow rates of 1.0 and 1.3 ml/min, respectively. Karl-Fisher (Metrohm
Chemical solutions
Main reactionsa
HCl H2SO4 HClO4b NaOH NH3
Na2CO3(aq)+2HCl(aq)→2NaCl(aq)+H2CO3(aq) Na2CO3(aq)+H2SO4(aq)→Na2SO4(aq)+H2CO3(aq) KHC 8 H 4O4(ac)+HClO 4(ac)→H2C8H4O4(ac)+KClO4(ac) KHC 8 H 4O4(aq)+NaOH(aq)→NaKC8H4O4(aq)+H2O(l) HCl(aq)+NH3(aq)→NH4Cl(aq) NaOH (aq)+HCl(aq)→NaCl(aq)+H2O(l)(back titration) Ca2+(aq)+Y−4(aq)→CaY2−(aq)
EDTA (Y4−) AgNO3 NH4SCN Na2S2O3 I3− KMnO4
NaCl(aq)+AgNO3(aq)→AgCl (s)+NaNO3(aq) Ag+(aq) + SCN-(aq) → AgSCN(s) BrO3− (aq)+9I−(aq)+6H+(aq)→Br−(aq)+3I3−(aq)+3H2O(l) I 3−(aq)+2S2O32−(aq)→3I−(aq)+S4O62−(aq) (back titration) H 3 AsO3(aq)+I3−(aq)+H2O(l)→H3AsO4(aq)+2H+(aq)+3I−(aq) 5C 2O42− (aq)+2MnO4−(aq)+16H+(aq)→10CO2(g)+2Mn2+(aq)+8H2O(l)
233
720 KFS Trinitro) was used to analyse the composition of the HPLC eluent mixtures. All analytical procedures employed were based and validated according to American (USP) and British (BP) Pharmacopoeias or in agreement with GLP procedures [4, 5, 6, 7, 8, 9]. The uncertainty of titrimetric (<0.25%), chromatographic (<0.60%) and Karl-Fisher (<0.36%) procedures were calculated according to the EURACHEM/ CIAC Guide [10]. The solutions were analysed weekly during the first month and then monthly over a 7.5-month period , where three assays and a blank were performed each time a solution was studied. In some cases, to cover a wide range of concentrations, the study of a higher and a lower concentration was performed, assuming that all the solutions of intermediate concentration would have the lower observed expiration date . The expiration date of each chemical solution studied Fig. 1 Diagram showing the expiration dating for the acidic, bacomplexing, argentometric, reducing and oxidizing solutions was establish on the basis of the average deviations system- sic, studied atically higher than ±2% relative to the initial value, obtained by standardization. It was also established, if deterioration signs occurred, e.g. colour changes or sedimentation, the solutions were percolated before standardization. If those signs reminded, the study was concluded and the solution rejected. The period until those occurrences was assumed to be the expiration dating for those solutions.
Results and discussion To assign an expiration date to each of the chemical solution studied, an average deviation systematically higher than ±2% relative to the initial value was established. This criteria was established taking into consideration the uncertainty usually accepted as GLP by analytical departments to reject solutions outside of the expected concentration, due the external influences. Therefore, the average deviation established (±2%) is an overestimate when compared with the initial value (µ) of each chemical solution: a µ±3σ confidence bands would include 99.74% of Gaussian data. From the 35 chemical solutions studied continuously over the 7.5-month period, Fig. 1 shows the expiration dating behaviour for the acidic, basic, complexing, argentometric, reducing and oxidizing solutions, some of them having more than one concentration level. The acidic solutions of chloridric (0.1 and 1 M) and sulphuric acids (0.5 and 1.5 M) reminded stable over 7 months, showing a similar behaviour within the established average deviation of ±2% in relation to the initial concentration, whereas perchloric acid (0.1 M) presented an expiration dating of only 1 month. In general, the more diluted solutions presented a decreasing stability in relation to the initial concentration, where the lower acidic content of the solutions suggests a higher vulnerability to external factors.
Fig. 2 Shewart chart showing the average relative deviation observed for ethylenediamine tetraacetic acid (EDTA) solutions (0.01 and 0.1 M) studied over a 7.5-month period
The basic solutions, such as sodium hydroxide (0.2 and 5 M) and ammonia (5 and 10 M), proved to be sensitive to atmospheric carbon dioxide and presented an expiration date of 5 and 3 months, respectively. The EDTA-complexing solutions seemed to lose some stability during the present study, as observed in Fig. 1. Although the 0.1 M solution remain very stable over the 7.5 months, the 0.01 M solutions become unstable after 2 months. The Shewart chart from Fig. 2, exemplifies the degradation kinetics profile for both EDTA solutions within the established average deviation of ±2%, relative to the initial concentration. A clear systematic deviation could be observed after 2 months for the 0.01 M EDTA solution, which confirms once again that the lower the concentration the higher the susceptibility to external factors. Chemical solutions used for argentometric procedures, such as 0.1 M silver nitrate, showed an expiration
234
phosphate, sodium dihydrogen phosphate (0.05 and 0.1 M), sodium chloride (5 ppm and 0.1 M) and sulphate (10 ppm). Exceptions were observed for disodium hydrogen phosphate (10%, w/v), potassium dihydrogen phosphate (0.01 and 1 M) and dipotassium hydrogen phosphate (0.01 and 0.2 M), showing expiration dates lower than 1 and 4 months, respectively. The 10% disodium hydrogen phosphate solutions showed a very high instability and it is advisable to always use a fresh solution for analytical purposes. The HPLC eluent mixtures (v/v), methanol/water (75/25 and 35/65), ethanol/water (30/70) and acetonitrile/water (20/80 and 90/10), presented an expiration dating of only 1 month, showing some instability behaviour for the established average deviation of ±2% relaFig. 3 Diagram showing the expiration dating for the salt solutive to the initial composition. tions and HPLC eluent mixtures studied In summary, from the 35 chemical solutions studied over the 7.5-month period, 27 remained reasonably stadating of 7 months. Nevertheless, for lower concentra- ble and had suitable expiration dates, equal or longer tions (0.01 M) the stability decreased to 5 months, due than 2 months for the established average deviation of probably to the much higher incidence of photochemical ±2% in relation to the initial concentration. Exceptions degradation. For the 0.1 M ammonium thiocyanate solu- are made for potassium permanganate (0.1 M), perchlotion, usually used for back-titration, an expiration dating ric acid (0.1 M) and disodium hydrogen phosphate of 7 months was achieved. (10%) solutions, as well as for HPLC eluent mixtures, The redox chemicals, like the reducing solution 0.1 M which showed stability with expiration dates equal to or sodium thiosulphate, showed a expiration dating of shorter than 1 month. The top 8 chemical solutions that 5 months within the deviation establish. Unexpectedly, the presented expiration dating longer than 6 months were oxidizing solution, 0.1 M iodine, presented a long expira- hydrochloric acid (0.1 and 1 M), sulphuric acid (0.5 and tion date of 7.5 months; 5 months for a 0.01 M solution. 1.5 M), EDTA (0.1 M), silver nitrate (0.1 M), ammoniThe lack of stability of the iodine solutions, due to volatil- um thiocyanate (0.1 M) and iodine (0.1 M). ity and air-oxidation processes, presented, as expected, a From the present work, it’s obvious that real solutions low shelf-life. Nevertheless, the low frequency of use of used in the laboratory are often opened much more freiodine solutions, observed in analytical laboratories in re- quently and for longer periods of time than in the present lation to other chemicals, accented the importance of a study, where the external factors should have a more long expiration date. The 0.1 M potassium permanganate substantial influence in the overall expiration dating. solutions also showed low stability, showing expiration Nevertheless, the data obtained must be understood as a dating of only 1 month, probably owing to the occurrence guideline to the expiration dating of several chemical soof photo degradation and air-oxidation reactions. lutions used for analytical purposes that could bring sevFigure 3 shows the expiration dating behaviour for eral benefits, enhancing laboratory’s performance. aqueous salt solutions and eluent mixtures currently used in HPLC. In general, the salt solutions studied showed Acknowledgements The authors acknowledge the financial supan expiration dating equal to or longer than 5 months: port of Hovione SA, R. Agudo for technical assistance and Prof. 2 M ammonium acetate, 0.03 M diammonium hydrogen Dinis Pestana for statistics helpful.
References 1. Prichard E (1995) Quality in the analytical chemistry laboratory. Wiley, UK 2. Lawn RE, Thompson M, Walker RF (1997) Proficiency testing in analytical chemistry. The Royal Society of Chemistry, UK 3. Garner WE, Barge MS, Ussuary JP (1992) Good laboratory practice. ACS, Washington D.C., USA
4. Serôdio P (2001) Validation of the stability of solutions, Report AA/RE181, Hovione SA, FCUL, Portugal 5. (1985) United States Pharmacopoeia, Vol. XXI 6. (1990) United States Pharmacopoeia, Vol. XXII 7. (1995) United States Pharmacopoeia, Vol. XXIII
8. (1980) British Pharmacopoeia, A45, Vol. I 9. (1988) British Pharmacopoeia, Vol. I 10. EURACHEM/CITAC (2000) Quantifying uncertainty in analytical measurement, EURACHEM/CITAC Guide 2nd edn. EURACHEM/ CITAC, Teddington, UK
