ANALYST, JANUARY 1992, VOL. 117 39 Digestion of Soil Samples for the Determination of Trace Amounts of Lead by Differential-pulse Anodic Stripping Voltammetry Angelo Ransirimal Fernando and James Alan Plambeck" Department of Chemistry, University of Alberta, Edmonton, Alberta, Canada T6G 2G2 An HN03-HC104 dissolution procedure followed by evaporation, re-dissolution and dilution with a CH3C02H-KN03 electrolyte is shown to be effective for the treatment of soil samples prior to the determination of trace amounts of lead by differential-pulse anodic stripping voltammetry. Keywords: Digestion procedure; lead determination; differential-pulse anodic stripping voltammetry; soil sample Differential-pulse anodic stripping voltammetry (DPASV) is now a commonly used technique for determining heavy metals such as lead at trace levels.However, recalcitrant real samples such as soils usually require spectroscopic methods for their analysis. In a previous paper,' some of the possible adverse effects of the reagents used for dissolution were examined. The present work applies the results of that study'.' to the determination of lead in actual soil samples. Experimental A local standard (LS) soil sample was collected from an urban backyard in Edmonton, Alberta, in an area known to have grown only grass and to have been unfertilized and undis- turbed for at least 10 years. Topsoil (exclusive of grass, vegetation and roots) was collected to the extent of 5 kg in polyethylene pails pre-cleaned with a 10% HN03 soak; soil that came into contact with the shovel was discarded.The black soil was oven-dried overnight at 30°C on aluminium foil. Initial sieving to remove stones and other foreign matter with a No. 10 stainless-steel sieve removed very little (<20 g) of the sample. The entire sample was then re-sieved, with grinding (CRC Micro Mill, The Chemical Rubber Co., Cleveland, OH, USA), to 25 mesh and stored in sealed glass jars. The results of the sieving (Canadian Standard Sieves, W. S . Tyler Co. of Canada, St. Catherines, Ontario, Canada) of the LS soil sample are given in Table 1. After mixing, sub-samples were withdrawn and re-sieved with grinding to 100 mesh and then to 200 mesh, then mixed by rotation. Sub-(sub-samples) (-20 g) were taken using a weighing bottle. After drying overnight at 100 "C and cooling in a desiccator, samples ( = I g) were removed for digestion and analysis. The closed weighing bottle prevented absorption of moisture (about 0.17% observed increase in mass if open bottles were used).In addition to the LS soil sample, soil certified reference materials (CRMs) SO-1 Regosolic Clay Soil, SO-2 Podzolic B Horizon Soil, SO-3 Calcareous C Horizon Soil and SO-4 Chernozemic A Horizon Soil [Canada Centre for Energy and Mineral Technology (CANMET). Canada] were used after drying according to the instructions.3.4 Open Beaker Digestions As HC104 digestion with a final evaporation step has been applied' to biological samples prior to analysis by DPASV, and as such a digestion procedure will result in only small amounts of organic residue, this procedure was considered to be the 'normal' method.The exact procedure used was that of the Canadian Society of Soil Science,h which has found widespread use.7.8 In this procedure, 1 g of soil ground to 300 mesh or below is digested successively in 20 ml of HN03, 20 ml of HC104 and 20 ml of HF followed by heating to near dryness, re-dissolution in 25 ml of 1 mol dm-3 HN03 and dilution to 50 ml. The procedure is designed for use with atomic absorption analysis, and was used with the 100 and 200 mesh soil samples without any modifications. Teflon Bomb Digestions The HN03 procedure for dissolution in Teflon-lined steel bombs (Parr 4745, Parr Instrument Co., Moline, IL, USA) described by Reddy et u1.9 was modified on the basis of other workI0.l1 to give the following procedure: weigh 0.3 g of sample into the Teflon cup, add 3 ml of concentrated HN03 and heat the mixture in an oven at 150 "C for 1.5-3 h.Cool with air (or in ice), transfer the solution into a 50 ml poly(propy1ene) calibrated flask and make up to volume with distilled water. This procedure did not dissolve the sample completely. Modification of the procedure by heating 0.2 g of sample at 150 "C for 3 h with 4 ml each of concentrated HN03 and HF produced much more complete dissolution. Although this procedure is a modification of that of Hsu and Locke,l' who used a mixture of HN03, HCI04 and HF, the use of HCI04 contrary to the manufacturer's recommendations'? was not considered safe. Perfluoroalkoxy( PFA)-Teflon bombs (Digestion Vessel 561, Savillex Corporation, Minnetonka, MN, USA) were also used for some microwave-assisted dissolutions.The procedure was the same as that used for the HN03-HF dissolution described above except that the 3 h of heating at 150 "C was replaced by 7 min of heating at 50% power (400 W) for eight of the bombs. The oven used was a standard household oven (Kenmore 88760, Sears Canada, Toronto, Canada) using a Table 1 Rcwlts of the sieving of SO g samples of LS soil Retention as a % of sample massY Sievc size (meshNo.) Run 1 Run2 Run3 Mean 40 18 17 17 17 60 19 19 I Y I9 80 11 I 1 1 1 1 1 100 S 5 5 s I40 7 7 8 7 170 5 S 5 5 200 4 4 4 4 325 10 10 9 1 0 (-325) 21 23 22 22 Bottom plate Total: 100 100 101 100.33 *: Rounded off to the nearest whole number. * To whom correspondence should be addressed.40 ANALYST, JANUARY 1992, VOL. 117 Table 2 DPASV data for the 100 mesh open-beaker digcsted LS soil samples Sample No.1 2 4 5 6 7 9 10 11 13 14 15 17 19 Lead found 29.7 28.0 38.0 33.5 38.0 29.8 35.5 37.8 32.9 23.7 23.1 28.7 30.0 30.1 (PPm) Mean: 31.4 Standard deviation: 4.9 Standard Relative deviation standard 1.8 6 1.1 4 3.3 9 2.6 8 4.4 11 0.5 2 3.3 9 3.2 8 3.2 10 2.0 8 1.5 7 1.9 7 0.92 3 0.96 3 (ppm) deviation (%) Table 3 DPASV data for the open-beaker digested 200 mesh LS soil samples Lead Standard Relative found deviation standard Sample No. (ppm) (ppm) deviation (YO) 1 31 .5 1.7 5 2 27.3 0.67 2 4 25.3 0.61 2 5 28.0 2.5 9 6 27.0 2.3 9 Mean: 27.8 Standard deviation: 2.3 Table 4 DPASV data for open-beaker digested samples of soil CRMs SO-1 and SO-2 Lead Standard Relative Soil Sample found deviation standard CRM No.(ppm) (ppm) deviation (%) SO- 1 * 1 19.7 0.73 4 2 18.9 0.91 5 3 22.9 2.06 9 so-2* 1 23.1 1.6 7 2 24.1 1.6 7 3 21.0 0.65 3 * Certified lead content for SO-1 and SO-2 = 21 k 4 ppm. standard household microwave turntable. In order to prevent leakage of acid fumes, the bombs were placed in a plastic container with a snap-fit lid, which prevented loss of fumes without the danger of explosion inherent in the use of a tightly sealed small container. Reagents, Apparatus and Other Procedures The reagents and the procedures used for their purification were as described previously.'.' All of the reagents were of analytical-reagent grade and were used as received, except as follows. Water was re-distilled from alkaline permanganate at very slow rates (0.5 1 h-1) by using an insulated vertical open column.Aqueous stock solutions of 2 mol dm-3 KN03 and of acetate buffer (1 mol dm-3 in both CH3COONa and CH3COOH) were further purified by using bulk electrolysis in a large-volume cell with a mercury cathode and a large-volume calomel anode separated from the reduction compartment by glass frits. The potential of the mercury pool was held at - 1.4 V against the calomel electrode for one month while the solution was slowly stirred for bubbling with de-oxygenated Table 5 Dissolution data for HN03 digestions in Teflon-lined steel bombs Lead Standard Sample Residue Dissolved found deviation Sample mass/g mass/g (YO) (ppm) (ppm) M200-01 0.3041 0.2109 30.7 27.5 2.3 M200-02 0.3448 0.2429 29.6 28.9 2.9 Average values: 30.1 28.19 Table 6 DPASV data for the 200 mesh microwave-digested soil samples Sample No.1 2 3 4 5 6 7 9 10 11 12 13 14 15 16 Lead found 22.2 20.7 22.0 18.2 20.1 15.8 23.3 19.5 14.6 20.3 18.6 15.4 15.1 14.9 18.5 (PPm) Mean: 18.6 Standard deviation: 2.9 Standard deviation 2.1 1 .o 1.1 1.4 0.9 1.8 1.9 1 .0 1.3 1.8 1.5 1 .0 0.5 4.9 1.4 (PPm) Re 1 at i ve standard deviation (YO) 10 5 5 8 4 11 8 5 9 9 10 6 4 33 7 and water-saturated nitrogen. High-purity nitrogen was further de-oxygenated using acidified vanadium(i1) chloride solution and zinc amalgam bubblers followed by aqueous washing towers. Potassium chloride was further purified by recrystallization and crystal adsorption. Triply distilled mercury was purified by anodic oxidation at +0.3 V against a saturated calomel electrode under a continuous stream of filtered laboratory air obtained by applying a mild vacuum to the cell.Positioning of the air inlet below the mercury surface also allowed the air to agitate the mercury. Acetic acid was further purified by isopiestic distillation. The apparatus used for analysis was as described previ- ously.1 The DPASV parameters1 (Model 174A Polarographic Analyzer, Princeton Applied Research, Princeton, NJ, USA) were: deposition potential, -0.9 V; modulation, 25 mV; clock, 1 s; scan rate, 5 mV s-1; deposition time, 90 s; and equilibration time, 30 s. Results Complete dissolution of the samples was achieved in the normal open-beaker digestion method, as expected, if the samples were heated to near dryness and then re-dissolved.This procedure also eliminates the residual HF and HCI04. The results of the analyses of the LS soil sample and of soil CRMs SO-1 and SO-2, obtained using this dissolution method, are given in Tables 2-4. The use of Teflon pressure bombs with HN03 alone dissolved only about 30% of the sample mass for 200 mesh LS soil samples, but apparently all of the lead (Table 5). Quantitative sample dissolution could be achieved by using HF together with HNO,; this resulted in a more yellowish solution . The same HF-HN03 mixture was also used with microwave heating. Dissolution of the sample was again essentially complete (>98.5%). However, a lead blank of 3.5 ppb together with significantly lower observed lead levels ('TableANALYST, JANUARY 1992, VOL.117 i7j 41 (c) 6) casts doubt on the validity of this protocol. Moreover, both of the HF-HN03 protocols produced solutions which, even when diluted, were reactive towards mercury, resulting in film formation on the mercury surface and plugging of the hanging mercury drop electrode capillary. These effects, which were presumably caused by HF, could be eliminated by conven- tional open-beaker heating of the microwave-digested solu- tions, but only with heating times comparable to those for open-beaker digestions. (Attempts to eliminate these effects by the addition of analytical-reagent grade H3B03 gave significantly higher lead blanks.) Discussion All of the acid digestion methods used required re-dissolution in a highly acidic medium after heating to near dryness.Attempts to re-dissolve the residue quantitatively in 0.1 rnol dm-3 HN03 resulted only in incomplete dissolution even after boiling for 1 h, whereas re-dissolution in 1.0 rnol dm-3 HN03 was rapid and apparently complete. Although the high acidity might not be a problem in atomic spectrometric analysis,lI.l~-1s it reduces the potential window available in DPASV to the point where analysis might become impossible (amplification does not help because the noise caused by the onset of hydrogen evolution is also amplified). Storage of the solutions in a concentrated and acidified form is better, a) J I I J -1.0 -0.5 0.0 +0.5 -1.0 -0.5 0.0 +0.5 PotentialN Fig. 1 Effect of dilution. The interfering signal observed at about -0.7 V in solutions prepared by diluting 1 ml of digested solution with 5 ml of water and sufficient 1 rnol dm-3 NaOH to adjust the pH to 2 [ ( u ) and ( h ) ] disappears completely on reduction of the volume of digested solution added to 0.1 ml [(c) and ( 4 1 .Deposition time was I min for ( a ) and ( b ) and 2 min for (c) and (d). For (c) and (d). current scnsitivity was also increased and no additional pH adjustment was made however, because adsorption on the container walls is then less of a problem. Three possible solutions to the problem of the analysis of very acidic solutions by DPASV are: partial neutralization (e.g., with NaOH); dilution with water; and dilution with a constant weak acid electrolyte. Partial neutralization was not successful. Attempts to dilute 1 ml of the digested solution to 5 ml followed by adjustment of the pH to 2 with NaOH required different amounts of NaOH for repeated samples.As a consequence, an aliquot was taken and titrated with NaOH to determine the amount required and this amount was added to a second aliquot. The volume of NaOH solution required was 600-1000 pl of 1 mol dm-3 NaOH. In the course of the titrations it became clear that a yellow coloration developed at a pH value above 2.7 followed by precipitation of Fe(OH)3 as the pH increased further (the LS soil sample contained 1.25 mmol of Fe per gram of soil as measured by atomic absorption spectrometry). Even when the pH was not allowed to rise above 2 and the solution was apparently clear, DPASV of solutions to which NaOH had been added showed, in addition to the clear and well-defined lead peak, an abnormal and ill-defined peak near -0.7 V (Fig.1). This peak prevented accurate measurement of the baseline of the lead peak. It is probable that the interference is caused by iron, as it disappears on precipitation of Fe(OH)3 at higher pH, but the lead is then removed also. Addition of ascorbic acid or hydroxylamine hydrochloride was not effective, Dilution of the solution, either with water or constant weak acid electrolyte, did remove the interfering peak, as would be expected if this peak were caused by interference from another matrix component, which was not reduced to a form that was soluble in mercury. Therefore, the dilution of the stripping solution with a comparable increase in deposition time results in a satisfactory voltammogram.Dilution of the dissolved sample has been suggested'.9 and is indeed effective in reducing matrix effects. Excellent baseline stability was obtained in repetitive runs on a single solution produced by diluting 100 pl of the digested sample to 5 ml with water. However, the pH of the solution (about 2) depended markedly on the dissolution method and on the extent to which the sample was evaporated prior to re- dissolution. The difference in the results of the analysis of the LS soil sample after different dissolution procedures had been applied was so large as to be unacceptable even though the precision of repetitive runs on a single dissolution was excellent. Dilution of the dissolved sample with a constant weak acid electrolyte makes the procedure less dependent on the amount of acid remaining from the dissolution stcp and is therefore preferable. Chloroacetic acid buffers produced unacceptably high background currents, viz., more negative than -0.8 V; however, solutions of CH3C02H at pH 3 were found to give excellent results if sufficient supporting elec- trolyte (KNOT) was present. Accordingly, the procedure chosen was dilution of 100 pl of the highly acidic dissolved sample with 5.00 ml of CH3C02H + K N 0 3 diluent (0.10 + 0.10 rnol dm-3).De-aeration of the diluent overnight in situ (a 500 ml repipette bottle) reduced the de-aeration time in each analysis from 5 to 1 min with no adverse effects. The dispenser was precise to 0.1%. (Acetic acid alone is not a buffer at pH 3. However, 0.1 mol dm-3 CH3C02H provides considerably greater protection against loss of protons during DPASV than does approximately 0.001 rnol dm-3 HN03.Acetic acid can also be easily purified.') The final procedure involved dilution with CH3C02H and standard additions of two 10 ~1 aliquots of a 5 ppm standard lead solution for each analysis. Four voltammograms were recorded for each of the three solutions. The mean values and standard deviations were calculated by computer programs.? The baseline was taken as tangential to the curve before and after the lead peak. Hsu and Lockel? have suggested that the colour of the42 ANALYST, JANUARY 1992, VOL. 117 digested solution might be caused by incomplete oxidation of organic matter. I n the present work it was found that the colour of the final solution was dependent on the extent of evaporation to dryness.Evaporation of the solution until the first signs of a solid were observed give a brown solid that on dissolution yielded a yellow solution. Further evaporation left a light-yellow solid residue which produced a nearly colourless solution on dissolution. The ultraviolet-visible spectrum of the yellow solution described above showed peaks near 300 and 260 nm, similar to those reported1619 for aqueous solutions of iron(il1) salts. The spectrum is very similar to that of 1.0 mmol dm-3 Fe2(S04)3 in 0.025 rnol dm-3 KCl, 0.02 mol dm-3 K2C2O4 and 0.1 rnol dm-3 ethylenediaminetetraacetic acid (EDTA) or KCI recorded in this work. The colourless digested solution turned yellow on addition of KCI, K2C204 or EDTA, whereas the yellow colour of the coloured solution was extracted into isobutyl methyl ketone, suggesting that chloro complexes of iron are the source of the yellow colour.Chloride might be present in the original sample or from HC104, whereas oxalate might be produced by the oxidation of organic matter by HCI04,Z0 as with the oxidation of soils by peroxide .2 * .22 The lead content of the LS soil sample and of the soil CRMs SO-1 and SO-2 is given in Tables 2-4. The over-all results for the LS soil sample are 31.4 +_ 4.9 pprn of lead (100 mesh sample) and 27.8 k 2.3 pprn of lead (200 mesh sample), whereas the values for SO-1 and SO-2 are well within the range given for their certified lead content, 21 5 4 ppm. Conclusion The DPASV method developed previously' for the analysis of synthetic samples can be used successfully for the determina- tion of total lead in soil samples following dissolution of the soil samples in HN03-HC104, evaporation, re-dissolution in 1.0 rnol dm-3 HN03 and dilution with 0.1 rnol dm-3 CH3COZH-KNO3.The authors are grateful to Dr. Byron Kratochvil and Dr. Gary Horlick for the loan of the different types of Teflon bomb. This research was supported by the Department of Chemistry of the University of Alberta. Taken in part from the thesis of A. R. F. submitted to the Faculty of Graduate Studies and Research in partial fulfilmcnt of the requirements for the degree of Ph.D. 1 2 3 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 References Fcrnando, A. R.. and Plambeck. J . A., Anal.Chem.. 1989.61, 2609. Fernando, A. R.. Ph.D. Thesis. University of Alberta, 1988. Bowman. W. S . , Faye, G. H . , Sutarno, R.. McKcaguc. J . A.. and Kodama. H., CANMET Report 79-3, Soil Sumples SO-I, SO-2, SO-3 und SO-4-Certified Reference Materials, Energy Mines and Rcsourccs Canada, Ottawa, 1979. Steger, H. F., CANMET Report 80-6E, Certified Reference Materials, Energy Mines and Resources Canada, Ottawa, 1980. Oehmc. M., and Lund, W., Fresenius Z. Anal. Cliem., 1979, 298, 260. Munuul on Soil Sumpling unti Methods of Anulysis. cd. McKcague, J. A , , Canadian Society of Soil Science, 2nd cdn., Ottawa, 1978. McKeague, J . A.. Desjardins. J . G., and Wolynctz, M. S . , Minor Elements in Cunudiun Soils. Agriculture Canada. Research Branch, Ottawa, 1979. Dudas. M. J . , and Pawluk, S . . Can. J . Soil Sci.. 1980, 60, 763. Reddy, S. J . , Valcnta, P . , and Nurnbcrg, H. W.. FreseniuA Z . Anal. Chem., 1982, 313. 390. Stoeppler, M.. and Backhaus, F., Fresenius Z. Anal. Cliem.. 1978, 291, 116. Stoepplcr. M., Muller. K . P., and Backhaus. F., Fresenius Z. Anal. Cliern., 1979, 297, 107. Hsu, C.-G., and Lockc, D. C., Anal. Chim. A m , 1983, 153, 313. Microwave Acid Digestion Bombs, Bulletin 4780, Parr Instru- ment Co., Moline, IL, 1986. McLarcn, J . W., Berman, S. S . , Boyko. V. J.. and Russcll, D. S . , Anal. Chem., 1981, 53, 1802. McQuaker. N . R.. Brown, D. F.. and Kluckncr. P. D.. Anal. Cliem., 1979, 51, 1082. Bastian. R., Wcbcrling, R., and Palilla, F., Anal. Cliem.. 1953, 25, 284. Ferguson, R. C., and Banks. C. V.. Anal. Chem., 1951.23,448. Medalia, A . I . . and Byrne. B. J.. Anal. Chem., 1951, 23. 453. Buck. R. P.. Singhadeja, S.. and Rogers. L. B., Anal. Ciiem., 1954. 26, 120. Smith, G. F.. The Wet Chemical Oxidution of Orgunic Composi- tions Employing Perchloric Acid, The G. F. Smith Chcmical Co., Columbus, OH, 1965. Martin. R. T., SoilSci., 1954. 77, 143. Farmcr, V. C., and Mitchcll. B. D., Soil Sci., 1963, 96. 221. Paper 1/01 9921 Receiiied April 29, 1991 Accepted August 8, 1991