Determination of trace metal impurities in high purity silver by two step selective precipitation separation followed by neutron activation analysis M. Y. Shiue,a Y. C. Sun,b J. J. Yeh,a J. Y. Yangb and M. H. Yang*a a Department of Nuclear Science, National Tsing-Hua University, 30043 Hsinchu, Taiwan b Nuclear Science and Technology Development Center, National Tsing-Hua University, 30043 Hsinchu, Taiwan Received 14th September 1998, Accepted 20th November 1998 A neutron activation analysis method for the determination of Au, Co, Cu, Fe, Hg and Zn in high purity silver materials with prior isolation of the analytes by a two-step selective precipitation separation from the silver matrix was developed. In the first step, the silver matrix was separated from the trace impurities of interest through the addition of hydrochloric acid to form a silver chloride precipitate.The principle of this separation is based on the extreme difference in the solubilities of the chlorides of silver and the trace elements of interest.In the second step, the trace elements remaining in the solution were subsequently coprecipitated with the Pb salt of pyrrolidine dithiocarbamate, Pb(PDC)2. The concentrations of six elements (Au, Co, Cu, Fe, Hg and Zn ) collected in the precipitate were determined by neutron activation analysis. Limits of detection of 0.001, 0.1, 0.08, 10, 0.1 and 1 mg g21 for Au, Co, Cu, Fe, Hg and Zn, respectively, were obtained.The proposed method was validated by the analysis of NIST SRM 8171 Fine Silver and applied to the determination of metal impurities in two high-purity silver samples (EM9465 and EM9343). Introduction The determination of trace impurities in high purity materials is essential in order to control and to improve manufacturing technology. An additional reason for the determination of trace element impurities in high purity silver is the effect of impurities on the freezing and melting properties when silver is used for primary and secondary temperature calibration.1,2 For the determination of trace impurities in high purity materials, a combination of chemical pre-treatment processes with instrumental analysis can generally achieve the best analytical performance.Atomic spectrometric methods, e.g., ETAAS and ICP-OES,3–10 and electrochemical methods,11–13 combined with preconcentration and/or separation procedures, have been reported for the determination of trace impurities in various materials.With the advent of ICP-MS, many attempts10,14 at the determination of trace element impurities in high purity materials both with or without preconcentration processes have been made. Even with this powerful analytical technique, the interference effect caused by the parent matrix may restrict its direct applicability to the determination of trace impurities in the samples. Among sample preconcentration methods, solid–liquid extraction, 10,14 ion exchange,12 coprecipitation5,7 and electrodeposition4,15 –19 are most commonly used for the separation of trace impurities from the sample matrix.However, in these multi-stage combined procedures there is a risk of increasing contamination and consequent worsening of the detection limits. To attain high sensitivity and reliability, the analytical blank and systematic error inherent in extreme trace analysis should be critically controlled.20 Instrumental neutron activation analysis (INAA) is a unique analytical technique which can sometimes be used for the direct determination of trace impurities in high purity metals.21 However, owing to the unavailability of suitable standards and difficulties connected with matrix interference, quantitative applications of this physical method may be severely restricted. 22 It is impossible to achieve the direct determination of trace impurities in high purity silver by NAA without pre- or post-irradiation separation, because the very high levels of gamma activity produced by the silver matrix may cause serious spectral interference in the determination of trace analytes.A method has recently been developed in our laboratory for the analysis of high purity silver by isolating analytes from the silver matrix with selective precipitation followed by ICP-MS determination.23 The elements including Al, Au, Cu, Cd, Co, Fe, Mg, Mn, Ni, Pb and Sn can be quantitatively separated from the precipitate of silver chloride.However, the determination of Hg was difficult because of the serious memory effect in the measurement by ICP-MS. In this study, the feasibility of applying a coprecipitation method for preconcenting Au, Cu, Fe, Hg and Zn from high purity silver followed by NAA was investigated. The effect of hydrochloric acid concentration on the recovery of trace elements from the Ag matrix and the effect of pH on the coprecipitation of the analytes by Pb(PDC)2 were investigated.Experimental Reagents, containers and samples All reagents were of analytical-reagent grade, unless stated otherwise. High-purity water was obtained by purification through de-ionization and double distillation. The purification of nitric acid and hydrochloric acid was carried out by subboiling distillation of the analytical-reagent grade acids. PTFE and glass containers were used throughout and were cleaned by immersion in HNO3 (1 + 1) for at least 24 h.Prior to use, they were rinsed with doubly distilled, de-ionized water and air-dried in a class 100 clean bench. Ammonium pyrrolidinedithiocarbamate (APDC) was used as a coprecipitant together with Pb(NO3)2. pH measurements were Analyst, 1999, 124, 15–18 15made with a conventional pH meter and the pH was adjusted with HCl and NH3 solution after adding originally 1% of 1 m acetate buffer (1 + 1). Samples of high purity silver EM9343 (6A9 grade silver shot) and EM9465 (5A9 grade silver shot) were obtained from Johnson Matthey Electronics (Royston, Hertfordshire, UK).Silver SRM 8171 (Fine Silver FS 14 Block) was obtained from NIST (Gaithersburg, MD, USA). Sample cleaning A 2 g sample of high purity silver was weighed into a 30 mL PTFE beaker and 25 mL of cold 0.1 m nitric acid were added with approximately 10 min of agitation, followed by thorough rinsing in doubly distilled, de-ionized water and air drying in a class 100 clean bench.A similar procedure was also applied to silver standard samples. Sample pre-treatment A flow chart of the proposed separation procedure is given in Fig. 1. A 0.1–0.3 g silver sample was weighed into a 20 mL PTFE beaker and 1 mL of water and 1 mL of concentrated nitric acid were added. The sample was heated below the boilingpoint of nitric acid until complete dissolution of silver was achieved, then 4 mL of 3 m hydrochloric acid were added progressively to form a fine precipitate of silver chloride.The solution was filtered with a 0.45 mm membrane filter and the filtrate collected was heated to near dryness. To the residue, 0.5 mL of 3 m hydrochloric acid was added to form a fine precipitate of silver chloride and the solution was filtered again. The filtrate collected was adjusted to pH 4 with 1% of 1 m acetate buffer (1 + 1), then 50 mg of Pb(NO3)2 and 5 mL of 1% APDC were added to the solution. The solution was allowed to stand for 0.5 h and the precipitate was filtered off.The filterpaper was inserted into quartz ampoules and was heat-sealed. Neutron activation analysis studies Several precipitated samples from silver were irradiated either with multi-element standards or using the monostandard method. Neutron irradiation was performed at fluxes of 1 3 1012–5 3 1013 cm22 s21 in THOR for 1 min (for the determination of Cu) and 30 h (for the determination of Au, Co, Fe, Hg and Zn). The counting system consists of a 38 cm3 Ge(Li) detector coupled with a TN-1710 4096-channel pulseheight analyser (Tracor Northern) and high voltage supplier (Canberra).The energy resolution of the system was 2.4 keV for 1332 keV. The irradiated samples were counted for 10 and 30 min for the measurement of short- (64Cu) and long-lived nuclides (203Hg, 198Au, 60Co, 59Fe and 65Zn), respectively. Results and discussion Precipitation separation procedure To achieve high sensitivity and accuracy for the determination of trace impurities in silver, a method based on the separation of the matrix element prior to the determination of isolated trace elements was developed. Coprecipitation of trace analytes should be avoided during the precipitation of silver chloride in order to achieve effective separation of impurities from the sample matrix.Furthermore, as Ag ions will be simultaneously collected with the trace analytes in the APDC coprecipitation step, the Ag matrix remained in the supernatant should be minimized to prevent the occurrence of interferences during the analysis by NAA.The effect of hydrochloric acid concentration on the recovery of trace elements and Ag matrix remaining in the supernatant after precipitation of silver chloride is shown in Table 1. The experiments were carried out in high purity silver spiked with 25 mg of the elements of interest followed by the proposed procedure and determination by ICP-OES. It can be seen from Table 1 that quantitative recoveries of most of the spiked elements, except Hg, from the silver chloride precipitate are obtained for concentrations of hydrochloric acid from 2.4 to 4 m.In contrast, the Ag matrix remaining in the supernatant was only 0.03–0.12% over the same acid range. It is interesting that the recoveries of Hg and Au show a significant dependence on the concentration of HCl. As Table 1 shows, quantitative recovery of Au can be achieved as the acid concentration increases to about 2.0 m, and that of Hg can only be achieved by further increasing the acid concentration to about 4.0 m. This observation can probably be explained on the basis of the relative tendency for the formation of soluble complexes (chloroauric and chloromercuric complexes) between Hg and Au with chloride ion in the HCl medium.Basically, the complexation behavior of Hg and Au towards chloride ion can be reflected in their respective stability constants. As indicated in the literature, the stepwise stability constants of Au3+ and Hg2+ with chloride ion are 8.5/6.7(log k1), 8.1/6.5(log k2), 7.0/6.9(log k3) and 6.1/1.0 (log k4), respectively. 24 Obviously, the higher stability constants of Au3+, compared with those of Hg2+ will result in a stronger interaction Fig. 1 Flow chart of the separation procedure for the determination of trace metal impurities in high purity silver by NAA. Table 1 Recovery of (%) of trace impurities and Ag matrix remaining in the supernatant after precipitation of silver chloride at different concentrations of HCl HCl concentration/m Element 1 2 2.4 3 4 Au 74 94 101 104 101 Co 93 87 97 94 98 Cu 96 93 100 104 104 Fe 94 87 99 96 98 Hg 40 55 75 81 105 Zn 101 96 105 104 104 Ag 2 0.04 0.03 0.04 0.12 16 Analyst, 1999, 124, 15–18between Au3+ and Cl2 and thus result in a quantitative recovery of Au at lower chloride concentrations than that of Hg.The differences in recovery between Hg and Au at different HCl concentrations can be explained on this basis.Coprecipitation of analytes with Pb(PDC)2 Following the separation of the Ag matrix, the trace analytes that remain in the supernatant should be further concentrated to a minimum sample size in order to be used effectively for NAA. For NAA, solid samples are preferred as they can be irradiated for longer times at higher fluxes of neutrons without the difficulties arising from the considerable gas generation in the radiolysis of water.Many processes can be used for the enrichment of trace elements in solid matrices,15,25 such as ion exchange, sorption on activated carbon, sorption on chelating agents immobilized on silica gel or a polymer chain and coprecipitation. Each of the different methods has its own advantages and the choice between them depends on the elements to be determined, the exact matrix and the method used for the detection of these elements. Coprecipitation is one of the most appropriate methods for our purpose because it provides a convenient way to collect trace elements from a large volume of sample solution on a small solid precipitate ( < 100 mg).The trace elements collected in this small sample can be used as a whole for NAA and can therefore result in a large enhancement of the analytical sensitivity. Concerning the coprecipitating agent, Pb(PDC)2 is considered to be one of the best choices for collecting trace elements in water for NAA,26 because APDC is a well recognized chelating agent which can react with over 30 elements,27 and the major constituent elements of the chelating agent (C, H, O, N) and also Pb do not form radioisotopes or form a beta emitter(209Pb) or have a very short half-life (207Pb, t1/2 = 0.8 s).28 The effect of pH on the coprecipitation of trace impurities in aqueous solution by Pb(PDC)2 was studied and the results are given in Table 2.The experiments were carried out by adding 25 mg of the respective elements to the supernant which was obtained by following the sample pre-treatment process shown in Fig. 1. Quantitative recoveries of most of the spiked elements with Pb(PDC)2 coprecipitation can be obtained from pH 3 to 6. A relatively lower recovery of Fe (90%) is observed however. This may be due to the formation of soluble complexes of iron hydroxide at higher pH.15 On the basis of the above experimental results, a pH in the range 3–5 is suitable for the coprecipitation of trace impurities by Pb(PDC)2 from aqueous solution.A pH of 4.0 was chosen for subsequent studies. Neutron activation analysis of high purity silver The determination of trace impurities in a silver sample is performed first by separation of the silver matrix in a medium of 3 m HCl followed by coprecipitation of the analytes of interest with Pb(PDC)2 at pH 4.0, and finally NAA of the coprecipitate sample. Six elements, Au, Co, Cu, Fe, Hg and Zn, were to be determined in this study.Among the radionuclides produced by the (n, g) reaction, only 64Cu has a short half-life of 12.7 h; the others (198Au, 60Co, 59Fe, 203Hg and 65Zn) have much longer half-lives from days to years. To achieve the quantitative determination of all the elements in the sample, two analytical schemes with a prescribed program of irradiation, cooling and counting were designed. For the determination of 64Cu (t1/2 = 12.7 h), a short irradiation time of 1 min followed by 1 h cooling and 10 min counting was applied, whereas for the longer lived nuclides a longer irradiation time of 30 h followed by 1 week cooling (to allow the decay of shorter interfering nuclides) and 30 min counting was applied.Fig. 2 shows the g-ray spectrum of a silver sample (NIST SRM 8171) which was treated by a two step precipitation separation followed by neutron irradiation. The g-spectrum was obtained for the sample subjected to 30 h irradiation, 1 week cooling and 30 min counting.The gamma peaks of the nuclides of interest including 198Au, 60Co, 59Fe, 203Hg and 65Zn are all clearly identified, but the g-spectrum of 110mAg, resulting from silver matrix remaining in the coprecipitate, also appeared. Since the majority of silver matrix was removed in the separation process, as evidenced in Table 1, the presence of minor activity arising form 110mAg would not constitute a perceivable interference effect on the accurate determination of the analyte nuclides.For the determination of 64Cu, the measurement of the 511 keV g-ray which resulted from the annihilation of b+-emission from 64Cu was performed. Basically, the 511 keV g-ray is not an appropriate choice for the characteristic identification of a specific nuclide because it can originate from both b+-emission and any g-decay with energy higher than 1022 keV. In this study, however, the applicability of determining 64Cu via measurement of 511 keV was tested by the selection of a suitable neutron irradiation time.The assumption underlying this study is that for a short irradiation time (1 min), 64Cu can be produced at a certain level owing to its relatively short half-life, whereas the long-lived nuclides would not be produced with such a short irradiation time at a perceivable level to cause interference in the measurement of 64Cu at 511 keV. In order to prove the correctness of this assumption, a neutron irradiated sample obtained by following the procedure in Fig. 1 was analyzed by tracing the half-life of the 511 keV g-peak. The result indicated that the half-life of this peak is very close to the 12.7 h of 64Cu . From the good coincidence of the half-life of the 511 keV g-peak with that of 64Cu, it may be concluded that the radioactivity measured at the 511 keV g-peak should come Table 2 Effect of pH on the recovery (%) of trace elements by coprecipitation with Pb(PDC)2 pH Element 3 4 5 6 Au 95 99 99 98 Co 97 99 99 99 Cu 98 100 99 100 Fe 97 99 99 90 Hg 94 100 100 100 Zn 95 97 99 99 Fig. 2 g-Ray spectra of a silver sample (NIST SRM 8171) treated by precipitation separation followed by neutron irradiation. Irradiation time, 30 h; cooling time, 1 week; counting time, 30 min. Analyst, 1999, 124, 15–18 17solely from the decay of 64Cu and can therefore be used for the quantification of this nuclide in high purity silver samples. The reliability of the proposed method was evaluated by the analysis of a certified silver standard.Table 3 gives the method detection limit and the analytical results for trace impurities in the sample of NIST SRM 8171 (Fine Silver FS14). Since no certified values are given for Co and Hg in the sample, the analytical reliability was tested with the method of standard additions. From Table 3, it can be seen that the analytical data obtained are in reasonably good agreement with the certified values and the spike recoveries are also within the reasonable range obtained in the previous study.The limits of detection were determined experimentally based on the lowest concentration of the analytes which produced the observed analytical signal on the addition of the analyte standard solution, and were 0.001, 0.1, 0.1, 10, 0.1 and 1 mg g21 for Au, Co, Cu, Fe, Hg and Zn, respectively. Further increase in the detection sensitivity would be possible if an increased neutron flux of the irradiation site can be applied. The developed method was applied to the analysis of the high purity silver samples EM9343 and EM9465 (Johnson Matthey) and the results are given in Table 4. The concentration levels of Co, Cu, Fe and Zn are found to be close to or below the limits of detection, and therefore quantification of these elements in these two samples is not possible under the present experimental conditions. However, the determination of Au and Hg was possible and the results were 0.10 ± 0.02 and 2.2 ± 0.4 mg g21 for EM9343 and 1.3 ± 0.3 and 4.3 ± 0.5 mg g21 for EM9465, respectively.An intercomparison study with the use of ICP-MS for the analysis of the EM9343 sample was also performed, and concentrations of 0.09 ± 0.01 and 2.0 ± 0.3 mg g21, for Au and Hg respectively, were found, which are in good agreement with those obtained by the proposed method and the values provided by Johnson Matthey. Conclusion A method for the determination of trace metal impurities in high purity silver which consists of selective separation and coprecipitation followed by NAA determination is established.Silver in the dissolved sample solution can be nearly quantitatively separated from the trace impurities by the formation of a silver chloride precipitate and the trace impurities remained in the aqueous solution can be subsequently coprecipitated by Pb(PDC)2 under the established conditions. The concentrations of Au, Co, Cu, Fe, Hg and Zn in the solid precipitate are finally determined by neutron activation analysis.The data with reasonably good accuracy and precision can be achieved by the proposed method. Acknowledgements The authors gratefully acknowledge the financial support of the National Science Council of Taiwan, (NSC 87-2212-E- 007-057). References 1 E. H. MacLaren, Can. J. Phys., 1957, 35, 1086. 2 B. W. Mangun, E. R. Pfeiffer, G. F. Strouse, J. Valencia-Rodriguez, J. H. Lin, T. I. Yeh, P. Marcarino, R.Dematteis, Y. Liu, Q. Zhao, A. T. Ince, F. Cakiroglu, H. G. Nubbemeyer and H. J. Jung, Metrologia, 1996, 33, 215. 3 J. D. Mullen, Talanta, 1976, 23, 846. 4 T. Tanaka, Y. Maki, Y. Kobayashi and A. Mizuike, Anal. Chim. Acta, 1991, 252, 211. 5 W. Reichel and B. G. Bleakley, Anal. Chem., 1974, 46, 59. 6 W. Lund, B. V. Larsen and N. Gundersen, Anal. Chim. Acta, 1976, 81, 319. 7 M. W. Hinds, J. Anal. At. Spectrom., 1992, 7, 685. 8 M. Hiraide, Y. Mikuni and H. Kawaguchi, Fresenius’ J.Anal. Chem., 1990, 354, 212. 9 I. G. Yudelevich, B. I. Zakda, V. P. Shabarova and A. S. Chereko, At. Spectrosc., 1992, 13, 108. 10 G. Kudermann, Fresenius’ J. Anal. Chem., 1988, 331, 697. 11 R. Naumann, W. Schmidt and G. Höh, Fresenius’ J. Anal. Chem., 1990, 347, 133. 12 S. R. Kayasth, A. K. Basu, N. Chattopadhyay and H. B. Desai, Anal. Chim. Acta, 1990, 231, 133. 13 T. Matsuda and T. Nagai, Anal. Sci., 1990, 7, 75. 14 R. J. Stummeye and G. Wünsch, Fresenius’ J. Anal.Chem., 1991, 340, 269. 15 A. Mizuike, Enrichment Techniques for Inorganic Trace Analysis, Springer, Berlin, 1983. 16 R. C. Chirnside, H. J. Chuly and P. M. C. Proffit, Analyst, 1957, 82, 18. 17 A. Mizuike, N. Mirsuya and K. Yammgai Bull. Chem. Soc. Jpn., 1969, 42, 253. 18 B. H. Vassos, R. F. Hirsch and H. Letlterman, Anal. Chem., 1973, 45, 792. 19 H. Malissa and I. L. Man, Mikrochim. Acta, 1971, 2, 241. 20 P. Tschöpel and G. Tölg, J. Trace Microprobe Tech., 1982, 1, 1. 21 G.Kudermenn, K. H. Blanfuss, C. Lührs, W. Vielhaber and V. Collisi, Fresenius’ J. Anal. Chem., 1992, 343, 734. 22 G. Tölg, Pure Appl. Chem., 1978, 50, 1075. 23 Y. C. Sun, J. Mierzwa, C. F. Lin, T. I. Yeh and M. H. Yang, Analyst, 1997, 122, 437. 24 J. Bjerrum, G. Schwarzenbach, and L. G. Sillen, Stability Constants, Special Publication No. 7, Chemical Society, London, 1958. 25 J. Minczewski, J. Chwastowska and R. Dybczynski, Separation and Preconcentration Methods in Inorganic Trace Analysis, Ellis Horwood, Chichester 1982. 26 C. R. Lan, Y. C. Sun, J. H. Chao, M. H. Yang, N. Lavi and Z. B. Alfassi, Mikrochim. Acta, 1990, 50, 225. 27 D. K. John and V. L. Jon, Anal. Chem., 1974, 46, 1894. 28 G. Erdtmann, Neutron Activation Tables, Verlag Chemie, Weinheim, 1976. Paper 8/07137C Table 3 Analytical results for trace impurities in high purity silver (NIST SRM 8171) determined by precipitation separation followed by NAA MDL Impurity concentration/mg g21 Detection limitb/ Element Certified Measureda Recovery (%) mg g21 Au 26.7 ± 6.4 23.5 ± 2.4 88 0.001 Co 2 (117 ± 6)c 88 0.1 Cu 65.2 ± 3.7 65.7 ± 9.4 100 0.1 Fe 48.9 ± 2.6 50.9 ± 2.0 104 10 Hg 2 (105 ± 5)c 79 0.1 Zn 7.2 ± 0.8 7.7 ± 1.4 107 1 a Mean ± s (n = 3). b Calculated based on the lowest concentration of the analyte which produced the observed analytical signal in the spectra on addition of the analyte standard solution. c 133 ppm of standard solution was added to the sample prior to sample pre-treatment. Table 4 Analytical results for trace elements (mg g21) in high purity silver samples (from Johnson Matthey Company) EM9465 EM9343 Reference Reference Element This worka value This worka ICP-MSb value Au 1.3 ± 0.3 1.0 0.10 ± 0.02 0.09 ± 0.01 2 Co N.D.c 2 N.D.c 2 2 Cu N.D.d 0.1 N.D.d 2 0.1 Fe N.D.e 2.0 N.D.e 2 0.3 Hg 4.3 ± 0.5 2 2.2 ± 0.4 2.0 ± 0.3 2 Zn N.D.f 2 N.D.f 2 2 a n = 5. b n = 3. c < 0.1 mg g21. d < 0.1 mg g21. e < 10 mg g21. f < 1 mg g21. 18 Analyst, 1999, 124, 15–18