ANALYST. JANUARY 1992, VOL. 117 27 Determination of Lead, Cadmium, Zinc and Tin in Samples of Poly(viny1 chloride) by Square-wave Voltammetry and Atomic Absorption Spectrometry Rao V. C. Peddy, G. Kalpana and Valsamma J. Koshy Research Centre, Indian Petrochemicals Corp. Ltd., Baroda-397 346, India Many metallic compounds used as stabilizers are incorporated in poly(viny1 chloride) (PVC) to improve its characteristics for various commercial applications. The determination of Pb, Cd, Zn and Sn in PVC samples by square-wave voltammetry (SWV) and atomic absorption spectrometry (AAS) is described. Two different procedures, using H2S04-H202-N H3-ethylenediami netetraacetic acid and H2S04-H202-H NO3 reagents to dissolve lead sulfate, were developed in order to suit the experimental requirements of AAS and SWV methods.The Pb-Cd-Zn mixed metal system was anaiysed by SWV and AAS and good recoveries were found for synthetic samples. The advantage of SWV over AAS is that it is more sensitive and the simultaneous determination of these elements is feasible. A good correlation between the data from SWV and AAS was obtained for all the metals ( r > 0.995). Tin can be readily determined using SWV and AAS methods. The SWV determination of Sn using 1 mot dm-3 HCI-4 mol dm-3 NH4CI buffer was carried out and the recovery was comparable t o that obtained using AAS. The same method was extended to a commercial sample. Statistical evaluation showed no significant bias between the methods used. Keywords: Square-wave voltammetry; atomic absorption spectrometry; lead, cadmium, zinc and tin determination; poly(vin yl chloride) Poly(viny1 chloride) (PVC) is the most important vinyl polymer, and to achieve processability and optimum perfor- mance characteristics, various compounding additives must be added.1 Many of these additives are metallic compounds, which affect the physico-mechanical properties of the end product.2 Lead compounds are used to improve the electrical and heat-stabilizing properties in applications such as cable sheathing and high-temperature extrusions and zinc com- pounds are used as mild stabilizers for non-toxic applications, particularly with epoxy resins.3 Cadmium laurate and stearate are used as light stabilizers and to confer good lubricating properties for obtaining transparent compounds.Mixtures of these metal additives are used for synergistic effects.3 The effectiveness of these metallic stabilizers depends on many factors, such as their concentration and their compatibil- ity with the polymer system. Although it is obviously important to be able to determine these additives in PVC samples, not many methods are available. Usually the most problematic and time-consuming part of the analytical process is the dissolution of the plastic. The classical method of calcination and subsequent dissolution of the resin in acid536 has the disadvantage of the possible loss of volatile com- ponents, and the treatment with concentrated nitric acid proposed by Rombach et af.7 is too long, as the attack by acid takes 8 h. Taubinger and Wilson8 studied the use of 50% hydrogen peroxide together with concentrated sulfuric acid to treat various organic samples, including PVC.Mendiola et af.”lO tested other methods of attack, and concluded that the most satisfactory results were obtained with sulfuric acid and hydrogen peroxide. However, the treatment with sulfuric acid causes difficulties if the PVC contains Pb or alkaline earth metals because of the precipitation of the corresponding sulfates. To overcome this problem, Belarra and co-work- ersl l , I 2 used an NH3-ethylenediaminetetraacetic acid (EDTA) reagent to dissolve the sulfates, but this reagent interfered in analysis by voltammetry. In this paper, an alternative method for the dissolution of PVC samples is suggested. The determination of trace metals, viz., Pb, Cd, Zn and Sn, in different combinations in a PVC matrix using square-wave voltammetry (SWV) and atomic absorption spectrometry (AAS) techniques was carried out and the results are compared.Experimental Instrumentation A GBC Model 902 atomic absorption spectrometer, equipped with a deuterium-arc background corrector, a 10 cm laminar- flow air-acetylene burner, a 5 cm laminar-flow dinitrogen oxide-acetylene burner, an Epson LX-800 printer and hollow cathode lamps, was used under the conditions given in Table 1. A Princeton Applied Research (PAR) Model 384B-4 polarographic analyser system together with a PAR 303-A static mercury drop electrode (SMDE) were used for all SWV measurements and the voltammograms were recorded on a Houston DMP-40 digital plotter.Potentiostatic control of the electrode potential was established by means of a three- electrode system consisting of an SMDE, a platinum-wire counter electrode and an Ag-AgCI-KCI (saturated) electrode as the reference electrode. Triply distilled mercury was used. Reagents and Chemicals All solutions were prepared from analytical-reagent grade chemicals, unless indicated otherwise, and water purified using a Milli-Q system (Millipore). Aldrich AAS standard solutions of Pb, Cd, Zn and Sn of 1000 pg ml-1 were used for preparing calibration standards. AnalaR-grade concentrated sulfuric acid (density 1.84 g cm-3), hydrogen peroxide (30% v/v), concentrated nitric acid (density 1.42 g cm-3) and concentrated ammonia solution (density 0.89 g cm-3) were Table 1 Instrumental parameters used for AAS analysis Distance of Lamp burner below Working Wavelength/ current/ optical range/ Element nm mA Flame* axis/mm pgml-’ Pb 217.0 5 AA 6 0-8 Cd 228.8 3 AA 6 0-2 Zn 213.9 5 AA 6 0-1.4 Sn 235.5 5 NA 6 0-80 * AA = air-acetylene; NA = dinitrogen oxide-acetylene.28 ANALYST, JANUARY 1992, VOL.117 used during different stages of dissolution. The EDTA solution (4% m/v) was prepared by dissolving 4.00 g of EDTA (as the disodium salt) in 100 ml of water together with a few drops of ammonia. Acetate buffer (0.1 rnol dm-3) was prepared by dissolving 8.2 g of sodium acetate in water, adjusting the pH to 4.5 with glacial acetic acid and diluting to 1 dm3 with de-ionized water. The HCI (1 rnol dm-3)-NH4CI (4 rnol dm-3) buffer was prepared by placing 8.25 ml of concentrated HCI and 21.4 g of NH4CI in a 100 ml calibrated flask and diluting to volume with de-ionized water. Sample Preparation Method A For AAS analysis, the dissolution procedure reported by Belarra et al." was followed for preparing sample solutions, the insoluble sulfate being brought into solution by converting it into its EDTA complex. A reagent blank was prepared by a similar procedure.For the determination of Sn in PVC, the sample solutions were clear after the H2S04-H202 digestion step, hence the same solutions were subjected to the analysis. Method B The sample preparation up to the stage of H202 treatment in method A was followed. The solution was then boiled to eliminate the excess of H202, 2 ml of concentrated HN03 were added and the mixture was heated vigorously to dissolve the precipitate. Subsequently the solution was cooled and diluted to 50 ml with water. A reagent blank was prepared by a similar method.Calibration For the AAS determination of all the elements, calibration solutions were prepared from their AAS standard solutions in the range suggested in Table 1. Matrix matching was carried out using the reagent blank, depending on the sample dilution. For obtaining the background current, square-wave voltam- mograms were recorded between an initial potential of -0.2 V and a final potential of - 1.2 V on 10 ml of de-aerated solution containing only the supporting electrolyte. This background current was deducted from the respective peak currents of the analyte. The experimental conditions for the SWV determina- tion of Pb, Cd, Zn and Sn are given in Table 2.Results and Discussion Sample Dissolution The dissolution procedure of Belarra et al.11 for the AAS determination of trace metals in a PVC matrix was followed. For SWV analysis, after the stage of H2S04-H202 digestion, the precipitate of lead sulfate was brought into solution using concentrated HN03 because the NH3-EDTA reagent caused Table 2 Experimental conditions for SWV Calibration Supporting Peak range1 Element electrolyte pH potentiallv yg ml-l Pb 0.1 mol dm--? Cd 0.1 rnol dm--? Zn 0.1 mol dm-3 Sn 1 rnol dm-3 HCI- acetate buffer 2.49 -0.496 1-5 acetate buffer 2.49 -0.670 1-5 acetate buffer 2.49 - 1.080 5-25 4 mol dm-3 NH4CI buffer -* -0.660 1-5 * Highly acidic. serious interference in the voltammetric analysis.It has been found that the presence of EDTA masks the analyte, resulting in a reduction in peak currents in SWV analysis. Pb-Cd-Zn Mixed Metal System As mentioned under Experimental, dissolution of the insol- uble sulfates was effected using the NH3-EDTA reagent for the AAS method and concentrated HN03 for the SWV method. The effect of the reagents was studied for the AAS method. Fig. 1 shows the change in analyte signal with increase in reagent concentration. Although sample dilution was suggested for the interference-free determination of Cd and Pb by Belarra and co-workers, 11.12 interactive matrix matching was found to be more suitable. For SWV analysis, the experimental conditions were optimized (Table 2) with respect to pH and the effect of buffer and its concentration, etc.Fig. 2(a) shows the square-wave voltammograms recorded in 0.1 rnol dm-3 acetate buffer containing Pb2+, Cd2+ and Zn2+ (all at 1 ppm concentration). The voltammetric pattern is characterized by three well defined peaks at -0.438, -0.628 and -1.048 V, which correspond to the two-electron reduction of Pb2+, Cd2+ and 0.130 0.120 0.140 I* It P 0.110 0.30 9 2 4 6 8 Reagent volume/ml Fig. 1 Effect of reagent concentration on atomization of various elements. Reagent composition: 2 ml of concentrated H2S03 + 10 ml of NH3 + 10 ml of 4% EDTA in 100 ml. (a) Pb; (b) Cd; and (c) Zn (cach 1 pprn) t c 2 3 0 -0.4 -0.6 -0.8 -1.0 -1.2 Poten tia IN versus Ag-Ag CI Fig. 2 Effect of dissolution reagent on SWV analysis of Pb-Cd-Zn system. (a) 10 pg ml-1 admixture of Pb, Cd and Zn standards in acetate buffer; ( b ) 10 pg ml-1 admixture of concentrated HN03 digest; and ( c ) 10 yg ml-1 admixture of NH3-EDTA digestANALYST, JANUARY 1992, VOL.117 29 Table 3 Calibration characteristics for the different metal additives in PVC Additive Analyte Method Regression equation* Pb-Cd-Zn Pb SWV y = 2.179~ + 4.800 X AAS AAS AAS AAS y = 3.392 x10-2x + 1.493 X y = 0.353~ + 2.653 X lop2 y = 0.240~ + 3.600 x lop2 y = 4.900 X lO-’x + 6.000 X Cd SWV y = 3.663~ + 0.348 Zn SWV y = 4.681~ + 0.562 Sn Sn SWV y = 3.900~ + 4.348 * x = Concentration (pg ml-1); y = peak current (nA) for SWV and absorbance for AAS. t n = 5 . Correlation coefficient (Y) 0.997 0.999 0.999 0.998 0.999 0.993 0.984 0.999 Relative standard deviation 1 .o 1.2 0.5 1.1 1.0 1 .o 0.8 2.0 (%)t 8.0 $ 6.0 E (3 4.0 2 2.0 Pb Zn n -0.4 -0.6 -0.8 -1.0 -1.2 Potent i a IN versus Ag -Ag C I Fig.3 Square-wave voltammograms of Pb, Cd and Zn with increasing analyte concentrations Zn2+, respectively. It can be seen that the voltammetric peaks of the metal ions are clearly separated under the same experimental conditions using an SMDE. Fig. 2(b) and (c) shows the square-wave voltammograms for a sample solution of the mixed metal salt system using methods A and B for dissolution. It can be seen that the peak potentials of Pb, Cd and Zn were shifted in the cathodic direction with HN03 digestion, probably owing to complexation, and there was no substantial difference in the peak currents. The voltammo- grams for Pb-Cd-Zn solution prepared by the NH3-EDTA route showed a distorted pattern [Fig.2(c)] and the Zn2+ signal was absent. The sensitivity in terms of the peak currents of the metal ions decreased markedly owing to the strong masking nature of EDTA. Hence the solutions prepared by method B were adopted subsequently for all the SWV analyses. Voltammograms for the simultaneous determination of Pb2+, Cd2+ and Zn2+ in a 1 : 1 : 1 molar ratio are shown in Fig. 3. Calibration graphs for Pb, Cd and Zn in the concentration ranges indicated in Tables 1 and 2 were prepared using standard solutions of these metals for analysis by AAS and SWV. The calibration graphs obtained by both methods were linear and the correlation coefficients and relative standard deviations are given in Table 3.The effectiveness of the procedure was tested by analysing admixtures of Pb, Cd and Zn in different proportions by both AAS and SWV. The recoveries were virtually quantitative in all instances (Table 4). Regression analysis was carried out on the data obtained by AAS and SWV and the regression coefficients were greater than 0.995. Interferences from other concomitants in the determination of Pb, Cd and Zn by AAS and SWV were studied and the results are presented in Table 5. Interferences from Sb and Sn were found to be high even at 1 : 1 molar ratios Table 4 Results of recovery assays of Pb, Cd and Zn Total metal Metal found/mg Recovery (%) added/ Element Sample mg SWV AAS SWV AAS Pb s- 1 s-2 s-3 Cd s-1 s-2 s-3 Zn s- 1 s-2 s-3 0.88 0.89 0.88 1.48 1.49 1.49 1.68 1.71 1.68 1.00 0.97 1.00 1.20 1.21 1.21 1.40 1.44 1.39 1.00 1.04 0.98 2.00 1.97 2.00 2.20 2.20 2.22 101.1 100.0 100.7 100.7 101.8 100.0 Mean: 101.2 100.2 97.0 100.0 101.0 101.0 102.8 99.3 Mean: 100.3 100.1 104.0 98.0 98.5 100.0 100.0 100.9 Mean: 100.8 99.6 ~~ Table 5 Results of interference study Tolerated metal to interferent ratio Pb Cd Zn Interferent Sb A1 Pb Cd Zn Ba Ca Sn Mg AAS SWV 1:lOO 1 : l O 1 : l O O 1 : l O O 1:100 1 : l O 1:100 1 : l O 1:100 1 : l O O 1:100 1 : l O 1:loo 1:lO 1:lOO 1:lO - - AAS SWV 1 : l O O 1 : l O 1:lOO 1:100 1:lOO 1:lO 1 : l O 1 : l O 1:100 1:lOO 1:100 1:lO 1:100 1:lOO 1:lO 1 : l - - AAS SWV 1:loo 1:l 1:100 1:lOO 1:100 1:100 1 : l O O 1 : l O O 1:lOO 1:lOO 1:lO 1:100 1:loo 1:lO 1:lO 1:lO - - in the SWV determination of Zn and Cd, respectively.However, the tolerance limits for other metals are good (metal to interferent ratio >1: 100) in both the AAS and SWV methods. The results obtained by the two methods were compared by using standard statistical techniques to assess the systematic errors.13 With the F-test at the 95% confidence level, the calculated Fvalue did not exceed the tabulated value. Hence it was concluded that on a statistical basis there was no significant bias between the methods used. It is clear that the procedure provides an accurate and rapid method for the determination of Pb, Cd and Zn, each with the others present by both AAS and SWV. Sn-PVC System Tin additives can be used with PVC as heat and light stabilizers for obtaining transparent sheeting and flexible compositions.1 In the determination of Sn in this work the NH3-EDTA dissolution step was not necessary as the solution was clear and free from precipitates after the H2S04-H202 digestion30 ANALYST, JANUARY 1992, VOL.117 10.0 8.0 5 6.0 2 2 3 u 4.0 2.0 0 PotentialN versus Ag-AgCI 0.5 0.7 0.9 Fig. 4 centration: A, 2; B, 3; C, 4; and D, 5 ppm Square-wave voltammograms of Sn with increasing con- Table 6 Results of recovery assays of Sn Total metal Metal found/mg Recovery (%) added/ Sample mg SWV AAS SWV AAS s-4 0.500 0.501 0.492 100.2 98.4 s-5 0.750 0.741 0.769 98.8 102.5 S-6 1.OOO 0.989 1.007 98.9 100.7 Mean: 99.3 100.5 s-7* - 0.240 0.246 * S-7 = commercial sample of PVC with Sn stabilizer. step. However, if Sn is present together with additives of alkali or alkaline earth metals or of lead salts, methods A and B described above should be followed for the sample dissolu- tion.In the SWV method, Sn determination was attempted in various buffers, viz., acetate, tartrate, citrate and HCl- NI&CI, but it was found that there is no quantitative response in any of these buffer systems except HCI-NH,CI, probably owing to strong complexation of Sn with the ligand. Calibra- tion graphs for Sn in the concentration ranges indicated in Tables 1 and 2 were prepared using a standard Sn solution for analysis by SWV and AAS. There was a systematic trend observed for the SWV method in 1 mol dm-3 HC1-4 mol dm-3 NH4CI supporting electrolyte and typical square-wave voltam- mograms with increasing concentration of the analyte are shown in Fig.4. The calibration graph equations obtained from a least-squares fit, correlation coefficients and relative standard deviations for Sn using both the AAS and SWV methods are given in Table 3. Recovery assays of Sn in synthetic samples of PVC (S-44-6) were carried out by both the AAS and SWV methods and the results are given in Table 6. An F-test at the 95% confidence level showed no significant error. The analysis was extended to a commercial sample (S-7) and there was reasonable agreement between the SWV and AAS results. The influence of other metals often present as additives in PVC, viz., Cd, Mg, Ba, Zn, Pb, Al, Sb and Ca, was studied and it was found that the tolerance limits were higher in the AAS method (metal to interferent ratio 1 : 100) than in the SWV method (metal to interferent ratio 1 : 10).There was serious interference from Pb, Ba and Ca in both the AAS and SWV determination of Sn owing to the formation of insoluble sulfates even at a 1 : 10 metal to interferent ratio. The SWV method was found to be more sensitive (1-5 pg ml-1 linear calibration range) than the AAS method (10-80 pg ml-1 linear calibration range) for the determination of Sn. Conclusions The results indicate that treatment with H2S04-HN03 solu- tion is suitable for dissolving insoluble sulfates formed when the PVC contains lead compounds. This method serves as an alternative to NH3-EDTA dissolution and its specific advan- tage is that it can be adopted in SWV techniques where the NH3-EDTA reagent gives severe interferences. However, the NH3-EDTA dissolution procedure can be used for all the metals studied when using the AAS method.The reliability of the methods is shown by the results of the recovery studies and by the satisfactory determination of these metals in PVC samples. The procedure is rapid and simple, the effects of the reagents on the AAS analysis are easily compensated for by interactive matrix matching and good accuracy and precision are obtained. The potential of SWV in the determination of trace metals in polymers was explored and its versatility for both qualitative and quantitative analyses was demonstrated. It is possible to determine Pb, Cd and Zn simultaneously by SWV, in contrast to the single-element AAS technique. The tolerance limits were higher for AAS than for the SWV technique.In conclusion, both methods are suitable for the determina- tion of trace metals in PVC but, depending on the combina- tion of metals and their amounts present, appropriate methods have to be chosen. The authors thank N. R. Shah, T. K. Joshi and R. H . Patel for experimental assistance. They are grateful to Dr. I. S. Bhardwaj, Director (Research Centre, IPCL), for his kind permission to publish this work. 1 2 3 4 5 6 7 8 9 10 11 12 13 References Titow, W. V., PVC Technology, Elsevier Applied Science, Barking, 4th edn., 1984. Whelan, A., Developments in PVC Production and Pro- cessing-I, Elsevier Applied Science, Barking, 1977. Nasc, L. I., Encyclopedia of PVC, Marcel Dekker, New York, 1976, vol. 1. Owen, E. D., Degradation and Stabilization of PVC, Elsevier Applied Science, Barking, 1984. Druckmann, D., At. Absorpt. Newsl., 1967, 6, 113. Fassy, H., and Lalet, P., Chim. Anal. (Paris), 1970, 52, 1281. Rombach, N., Ape], R., and Tschochner, G., GZT Fachz. Lab., 1980,24, 1165. Taubinger. R. P., and Wilson, J. R., Analyst, 1965, 90, 429. Mendiola, J. M., Gonzalez, A., and Arribas, S., Afinidad, 1980, 37,39. Mendiola, J. M., Gonzalez, A., and Arribas, S . . Afinidad, 1980, 37,251. Belarra, M. A., Gallarta, F., Anzano, J. M., and Castillo, J. R., J. Anal. At. Spectrom., 1986, 1, 141. Belarra, M. A., Azofra, M. C., Anzano, J. M., and Castillo, J. R., J. Anal. At. Spectrom.. 1989, 4, 101. 1985 Annual Book of ASTM Standards, American Society for Testing and Materials, Philadelphia, 1985, vol. 03.06, Standard E 876 82, p. 346. Paper 1f03505C Received July 11, 1991 Accepted August 22, 1991