Selective Extraction and One-drop Flame Atomic Absorption Spectrometric Determination of Trace Amounts of Silver in Highly-pure Copper and Lead Analytical Atomic Spectrometry ISAO KOJIMA AND ASAKO TAKAYANAGI Laboratory of Analytical Chemistry Department of Applied Chemistry Nagoya Institute of Technology Gokiso-cho Showa-ku Nagoya 466 Japan After the dissolution of copper and lead in sulfuric or nitric acid silver was selectively extracted as a ternary silver- 1,2-bis(ethylthio)ethane-picrate complex into a small volume of chloroform at pH 1.0-1.5 leaving copper and lead in an aqueous solution without a masking agent. Direct nebulization of 50 yl of the chloroform extract into a fuel-lean air- acetylene flame gave a sensitive signal height without background correction from chloroform.The present method was successfully applied in the determination of silver in copper and highly-pure copper and lead metals. Keywords Flame atomic absorption spectrometry; silver; ternary siluer-l,2-bis(ethylthio)ethane-picrate complex; highly-pure copper and lead Silver can be determined extremely well by AAS even with use of low temperature flames.' For the determination of trace silver the combination of solvent extraction and direct nebuliz- ation of the extract is very efficient.2 Even by use of 'one-drop' FAAS with direct nebulization of the chloroform extract of a silver complex the sensitivity for the determination of silver was improved about 4.4 fold.3 Various organic reagents have been proposed for the spectro- photometric determination of silver. Of these reagents some were used for the selective extraction of silver in the presence of a masking agent4 Recently it has been reported that thiacrown ethers obtained by replacing the oxygen atoms in crown ethers with sulfur atoms have a great affinity for soft metals such as silver copper(1) and mercury(11)~ and extract these metal ions as ternary complexes in the presence of a bulky anion such as p i ~ r a t e .~ - ~ In addition 1,2- bis(alky1thio)ethane was used as an extracting agent for these metals in the presence of a counter anion.*.'' The present paper describes the selective extraction of silver(1) with a commercially available thioether [ 1,2-bis(ethylthio)ethane] at pH 1.0-1.5 into a small volume of chloroform in the presence of picric acid leaving copper and lead ions in aqueous solution without use of a masking agent followed by 'one-drop' FAAS determination of silver.The method was applied in the determination of silver in highly-pure copper and lead. EXPERIMENTAL Apparatus An atomic absorption spectrometer equipped with a 100 mm burner head and a deuterium background corrector (Seiko Chiba Japan Model SAS-727) was used for metal measure- ments with a discrete nebulization technique using a fuel-lean air-acetylene flame under the same operating conditions as previously r e p ~ r t e d . ~ Sample solutions were injected with a micropipette (Gilson Villiers le Bel France P-200) into a PTFE funnel coupled directly to the nebulizer needle. Signal intensity was recorded using a strip-chart recorder (National Osaka Japan VP-6612a).The pH of the solutions was meas- ured with a Radiometer (Copenhagen Denmark) Type PHM-22 pH meter with a combined electrode. Reagents Stock standard silver solution 2 mg g-' in 0.5 mol dm-3 nitric acid was prepared by dissolving silver (99.99% purity; Nacalai Tesque Kyoto Japan) in nitric acid directly in a PTFE bottle (120cm3) followed by dilution by mass with water. Stock standard copper and lead solutions 2 mg g-' in 0.10 mol dm-3 nitric acid were prepared by dissolving spectrographic stan- dard grade copper and lead (99.999% purity; Mitsuwa Chemicals Osaka Japan) in nitric acid directly in a PTFE bottle (120 cm3 capacity) followed by dilution by mass with water. Working aqueous standard metal solutions were pre- pared by diluting the stock solution in 5-10 cm3 PTFE bottles to the appropriate concentrations by mass with 0.1 mol dmP3 nitric acid.Working standard silver solutions of chloroform extracts were prepared by extracting silver with 1,2- bis(ethy1thio)ethane (8-2s) at pH 1.0-1.5 in the presence of picric acid into an appropriate volume of chloroform the precise volume of which was calculated from the mass of the volume taken using a micropipette and the density. Chloroform of analytical-reagent grade ( Wako Osaka Japan) was used as received. Commercially available 8-2s (Wako) of analytical-reagent grade was dissolved in chloroform (0.05 mol drnp3). Aqueous ammonium pyrrolidin-l-yldithiof- ormate (APDC) solution (1%) was prepared by dissolving APDC (Nacalai Tesque) in water.Picric acid (0.01 mol dm-3) and sodium tetraphenylborate (0.01 mol drnp3) (Wako) of ana- lytical-reagent grade were also dissolved in water. Working standard chloroform solutions containing the copper- and lead-APDC complex were prepared by extracting the mixed solution containing 2.5 cm3 of copper (5 pg) or lead (10 pg) standard solution 0.5 cm3 of 1% APDC solution and 2 cm3 of water with 0.5cm3 of chloroform to test the percentage extraction of copper and lead with 8-2s at different pH values. Sodium acetate of analytical-reagent grade (Wako) was used to control the extraction pH. The other acids used were of analytical-reagent grade. De-ionized Milli-Q water was used throughout. Journal of Analytical Atomic Spectrometry August 1996 Vol. 11 (607-61 0) 607Dissolution of Metals Into a PTFE bottle of 100 cm3 capacity were added 1 g of copper 4 cm3 of concentrated nitric acid 0.5 cm3 of concen- trated sulfuric acid and 3 cm3 of water.The bottle was then warmed on a hot plate to dissolve the metal and then heated to remove nitric acid at a temperature of about 150°C. The residue was diluted to 50 g with water. One g of lead taken in a 100cm3 PTFE bottle was dissolved with 9.7cm3 of warm 2mol dm-3 nitric acid and finally diluted to 50g with 0.1 mol dm-3 nitric acid. Copper solution was prepared as a sulfate solution to avoid the possibility of the extraction of the ternary Cu-8-2s-nitrate complex because divalent sulfate ion does give a ternary complex. For lead nitric acid was used to avoid the precipitation of lead sulfate.Extraction Procedure Into a 10 cm3 stoppered glass test-tube were placed 0.5-1.5 cm3 of sample solution (weighed) 0.5 cm3 of 0.2 mol dm-3 sodium acetate solution 0.5 cm3 of picric acid solution 3.5-2.5 cm3 of water and 0.5 cm3 of 8-2s chloroform solution (weighed). After shaking for 20 s (three times) by hand the test-tube was set aside until complete phase separation had occurred. A 50 p1 portion of the chloroform extract was then directly injected into the flame for AAS measurement of silver. Signal heights were measured for the determination of metal content. Calibration Graph To a 10cm3 stoppered glass test-tube were added aliquots (0-0.5 cm3 weighed) of a standard aqueous silver solution of 1.0 pg g-l 0.5 cm3 of picric acid and 4.5-4.0 cm3 of water.After shaking for 20 s (three times) by hand with 0.5 cm3 of 8-2s chloroform solution (weighed) the test-tube was set aside until complete phase separation had occurred. Portions (50 pl) of the chloroform extracts were directly nebulized into the flame for AAS measurement of the silver calibration standards. For copper and lead the calibration graph was also constructed by nebulizing 50 p1 of the chloroform extracts of the corresponding APDC complexes. 9 0 1 - I V 1 2 3 4 5 PH Fig. 1 Effect of pH on percentage extraction of silver (10 pg; circles) copper (25 pg; squares) and lead (30 pg; triangles) with 8-2s (0.05 mol dm-3) in chloroform in the presence of picric acid (open circles 2 x mol dm-3; filled circles squares and triangles 1 x lo-' mol dm-3 and open triangles 2 x mol dm-3) and tetra- phenylborate 2 x mol dm-' in aqueous phase; open squares).Vaq Vorg= 5 cm' 0.5 cm3 and the amount of organic substances extracted was reduced; hence the background signal was reduced. Effect of Injection Volume on the Signal Height The direct injection of the chloroform extract of a ternary silver-8-2s-picrate complex gave sensitive spike-like and/or very smooth signal profiles as in the case of zinc.3 The background corrected signal obtained by nebulizing a small volume of the chloroform extract increased with an increase in the injection volume up to a volume of about 70 pl at a constant silver concentration in the extract. But the increase in the signal intensity was not large for an injection volume of 50-70 pl. The background signal appeared only at injection volumes greater than 70 pl.This means that background correction is not necessary if an injection volume of less than 601.11 of chloroform extract is used. In order to enhance the extraction concentration of silver and to reduce the background signal an injection volume of 50 pl was used in the present study. Even an injection volume of 50 pl gave a satisfactory reproducibility. RESULTS AND DISCUSSION Effect of pH on Extraction Effect of Some Factors on the Signal Height The effect of pH on the extraction of silver copper and lead was studied using a chloroform solution of 0.05 moldm-3 8-2S in the presence of various amounts of picric acid (2 x 10-4-2 x mol dm-3) or tetraphenylborate anion (2 x lop4 mol dm-3). The results are shown in Fig.1. At a picrate concentration of 1.0 x low3 mol dmp3 98.5% of silver was extracted at pH 1.0-4.0. At a picrate concentration of 2.0 x mol dm-3 silver was quantitatively extracted at pH 1.0. This is expected as the picrate ion is the main species in the picric acid-picrate system at pH 1.0 because the logarith- mic formation constant of picric acid is 0.33." On the other hand only a few ppm of copper were extracted at pH 2.0 under the same conditions. Also extraction of copper increased with increasing extraction pH but lead was not extracted at all at pH 1-3. In addition the extraction of silver decreased with decreasing picrate concentration. When tetraphenylborate anion was used as a counter ion silver was completely extracted at pH 0.5-4.0 (this was omitted in Fig.1 for simplicity) and a relatively high concentration of copper was also extracted (see Fig. 1 ) but lead was not extracted at all. Thus a picric acid concentration of 1.0 x mol dmP3 was used to extract silver selectively from a solution containing large amounts of copper or lead. Copper or lead was left in the sample solution With a constant injection volume of 50 pl of a chloroform extract containing 1.0 p g ~ m - ~ of Ag the effects of the burner height (10.5-15.0 mm above burner head) sample flow rate (3.0-4.0 cm3 min-' of HzO) and acetylene flow rate (3.0-3.8 dm3 min-') on the signal height were studied with and without deuterium background correction. Judging from the baseline stability and the reproducibility of the signal height the same operating conditions as already reported in ref.3 were selected. Calibration Curve The calibration curve obtained by injecting 50 p1 of the chloro- form extracts was a straight line with a zero intercept. A typical example is presented in Fig. 2 with the calibration curve obtained by nebulizing aqueous solutions. The signals were obtained by repeated nebulization of standard extracts and were compared with the signals obtained by injecting sample extracts. The signal height obtained by injecting 50 pl of the chloroform extract containing 0.1 p g ~ r n - ~ of Ag was very constant the RSD obtained in this case was 1.2% (n= 10) (see A in Fig. 2) and the detection limit was 4 ng cm-3 of Ag in chloroform based on the 3 (T criterion. The calibration 608 Journal of Analytical Atomic Spectrometry August 1996 Vol.1 1Table 1 Analytical results for silver in copper metals Sample 5N-CU-A 5N-CU-B Np-Cu Cu taken/mg 10.25 10.32 10.31 10.66 10.52 31.78 31.01 31.01 31.45 31.60 10.50 10.52 10.54 31.61 31.53 31.46 4.23 4.23 4.25 Ag found/pg 0.092 0.096 0.090 0.097 0.095 0.282 0.282 0.286 0.278 0.280 Standard addition method 0.093 0.096 0.096 0.29 1 0.288 0.288 0.385 0.381 0.383 Ag content (ppm) 9.0 9.3 8.8 9.1 9.0 9.2 9.1 9.2 8.9 8.9 8.9 9.1 9.1 9.2 9.1 9.1 91.2 90.8 90.1 Mean s (%RSD) 9.1 k0.2 (1.7) 9.0 9.0 t- 0.1 (1.3) Table 2 Recovery of silver added to sample solution Cu taken/ 10.60 10.57 10.30 30.78 31.01 31.01 Sample mg 5 N - C u - A Ag added/ Ag found/ Recovery Pg I % (%) - 0.096 - 0.101 0.196 99.0 0.098 0.192 98.0 0.097 0.379 102.1 0.248 0.534 102.4 - 0.280 - curve for silver was not affected even in the presence of up to 50 pg cmP3 of Cu in the chloroform extract.Determination of Silver in Metals The present method was applied in the determination of silver in three copper samples C99.999 plus % purity (5N-Cu-A) 99.999% purity (SN-Cu-B) and normal purity (NP-Cu)] and one lead sample C99.999 plus % (5N-Pb)]. The final results were easily calculated from the sample mass used the final sample solution (weighed) an aliquot (weighed) taken from the final sample solution the volume (or mass) of chloroform and the silver content found in the chloroform extract. The results are given in Tables 1-3 and the signal height obtained for the lead sample is given in Fig. 2B. The analytical results obtained by the calibration method and the standard addition method compared well with one another and were reproduc- ible.The recovery of silver added to the sample solution was satisfactory (see Table 2). Table 3 Analytical results for silver in lead [Agl/pg ml-’ in CHCI [Agl/wg ml-’ in aq. soh. Fig. 2 Typical signals obtained by injecting 50 pl of chloroform extracts and 100 p1 of aqueous solutions where V Vorg = 5 cm3 0.5 cm3; 8-2s 0.05 mol dm-3 in chloroform; picric acid 1 x mol dmP3 in aqueous phase; A reproducibility (0.1 pg Ag in cm3 of chloroform); and B sample signal height in pure lead REFERENCES 1 Welz B. in Atomic Absorption Spectrometry 2nd edn. VCH Weinheim 1985 pp. 324-325. 2 West F. K. West P. W. and Ramakrishna T. V. Environ. Sci. Technol.1967 1 717. 3 Kojima I. Inagaki K. and Kondo S. J . Anal. At. Spectrom. 1994 9 1161. Sample Pb taken/mg Ag added/ng 5N-Pb 101.6 101.9 102.1 102.6 102.3 102.6 102.3 - 6.1 12.1 Standard addition method Ag founding 4.1 4.1 4.0 4.1 4.0 10.2 16.1 Ag content (ppb) Recovery (%) 40 40 40 40 39 40 39 41 - 102 100 Journal of Analytical Atomic Spectrometry August 1996 Vol. 11 6094 El-Ghamry M. T. and Frei R. W. Anal. Chem. 1968 40 1986. 5 Oue M. Akarna K. Kimura K. Tanaka M. and Shono T. Anal. Sci. 1989 5 165. 6 Sevdec D. Fekete L. and Meider H. J. Inorg. Nucl. Chem. 1980 42 885 (and papers cited therein) 7 Saito K. Masuda Y. and Sekido E. Anal. Chim. Acta 1983 151 447. 8 Ohki A. Takagi M. and Ueno K. Anal. Chim. Acta 1984 159 245. 9 Oue M. Kimura K. and Shono T. Anal. Chim. Acta 1987 194 293. 10 Dietze F. Gloe K. Jacobi R. Muhl P. Beger J. Petrich M. Beyer L. and Hoyer E. Solvent Extr. Ion Exch. 1989 7 223. Martell A. E. and Smith R. H. in Critical Stability Constants vol. 3 Other Organic Ligands Plenum Press New York and London 1977. 11 Paper 6102372 J Received April 4 1996 Accepted June 11 1996 61 0 Journal of Analytical Atomic Spectrometry August 1996 Vol. 11