ANALYST, MAY 1991, VOL. 116 525 Determination of Trace Amounts of Copper, Nickel and Zinc in Palladium Compounds by Solvent Extraction Flame Atomic Absorption Spectrometry Sijka A. Popova, Stefanka P. Bratinova and Christina R. lvanova Central Research Laboratory, Higher Institute of Chemical Technology, I 156 Sofia, Bulgaria A method is described for the determination of trace amounts of Cu, Ni and Zn in diamminedichloropalladium and diamminedinitropalladium by flame atomic absorption spectrometry after an extraction procedure using ammonium pyrrolidin-I-yldithioformate [ammonium pyrrolidinedithiocarbamate (APDC)] as complexing agent and isobutyl methyl ketone (IBMK) as extractant. Another extraction system, NH4SCN-IBMK, was used for the preliminary removal of Pd which also forms extractable complexes with APDC.Optimum conditions for the selective separation of Pd and for the simultaneous extraction of Cu-, Ni- and Zn-PDC complexes into IBMK were determined. The sensitivity and precision of the proposed method are sufficient for quality control requirements. Keywords: Copper, nickel and zinc determination; solvent extraction flame atomic absorption spectrometry; diamminedichloropalladium; diamminedinitropalladium; ammonium pyrrolidin-7-yldithioformate Diamminedichloropalladium [ Pd(NH3)2C12] and diamminedi- nitropalladium [Pd(NH3)2(N02)2] are used as raw materials in the production of microelectronic systems; hence they have to conform to a high degree of purity necessitating exacting quality control requirements. The content of impurities such as Cu, Ni and Zn in these materials needs to be be controlled and should not exceed 0.001-0.0001Y0.There are no methods described in the literature for the atomic absorption spec- trometric determination of Cu, Ni and Zn in the Pd compounds studied here. Hence there is a need for an accurate method for the determination of Cu, Ni and Zn in Pd(NH3)2C12 and Pd(NH3)2(N02)2 with high sensitivity and selectivity. This problem has been solved by using solvent extraction prior to flame atomic absorption analysis. An extraction system extensively used and preferred in atomic absorption analysis is a combination of ammonium pyrrolidin- 1-yldithioformate [ammonium pyrrolidinedithiocarbamate (APDC)] as complexing agent and isobutyl methyl ketone (IBMK) as extractant. The main problem in applying the APDC-IBMK extraction procedure to the determination of Cu, Ni and Zn in Pd compounds is the co-extraction of Pd.1 This necessitates its prior removal from the system; precipitation is not recom- mended for this purpose, because of the large amount of Pd involved and the possibility of coprecipitation of Cu, Ni and Zn.It has been recommended that the separation of Pd be carried out by extraction of its rhodanide (sulphocyanide) complex into IBMK.2.3 Extraction of 99.9% of the Pd into the organic phase from 3-6 mol dm-3 HC1 can be achieved with a single extraction procedure.2 In order to develop a suitable method it is necessary to examine the conditions under which the extraction of Cu, Ni and Zn together with Pd as their rhodanide complexes can be avoided, and to optimize the conditions for the solvent extraction pre-concentration of Cu, Ni and Zn with the use of APDC-IBMK and their subsequent determination by flame atomic absorption spectrometry.Experimental Sample Preparation A 1.000 g sample of Pd(NH3)2C12 or Pd(NH3)2(N02)2, previously dried at 105 "C, was dissolved in 10 ml of 1 mol dm-3 HCI by heating at about 90 "C for 1 h. After cooling to room temperature, the solution was diluted to 50 ml with de-ionized water. Reagents All reagents were of analytical-reagent grade, and de-ionized, doubly distilled water was used throughout. Ammonium thiocyanate solution, 20% . Previously purified by extraction with APDC-IBMK. Ammonium pyrrolidin-1-yldithioformate solution, 1 YO.A 1.00 g amount of the reagent was dissolved in 50 ml of water containing 1 ml of 25% ammonia solution and the solution was diluted to 100 ml. The residue was filtered. The solution was prepared daily, Sodium hydroxide solution, 25% (metal-free). Sodium hydroxide pellets (250 g) were dissolved in 1 1 of water and the solution was transferred into a separating funnel. A 1 ml volume of APDC solution was added and the mixture extracted with 20 ml of IBMK. The addition of APDC and extraction was repeated until the extracts were colourless. The solution was stored in a polyethene bottle. Isobutyl methyl ketone (Merck). Standard solutions of Cu, Ni and Zn, 1 mg ml-1 each (BDH). Working standard solutions of Cu, Ni and Zn were prepared daily by appropriate dilution of the stock standard solutions.Instrumentation A Perkin-Elmer Model 3030B atomic absorption spec- trometer equipped with hollow cathode lamps for Cu, Ni and Zn as light sources was used. The absorption signals were measured under the conditions shown in Table 1. A pH-meter (Radiometer) was also used. Results and Discussion Conditions for the Separation of Pd As has been reported previously,2 virtually complete extrac- tion of Pd as its rhodanide complex can be achieved over a wide range of acidity (3-6 mol dm-3 HCl). It is known, however, that Cu, Ni and Zn also form stable rhodanide complexes.4 This necessitates optimization of the conditions for the selective extraction of Pd into IBMK. The Ni-rhodanide complex is not extracted into IBMK and remains entirely in the aqueous phase under the optimum conditions for the extraction of Pd as its rhodanide complex (Fig.1, curve 1). In addition, the Cu- and Zn-rhodanide complexes are not co-extracted with the Pd-rhodanide com-526 ANALYST, MAY 1991, VOL. 116 100 80 s - 60 0 1 .- ta 40 20 E ta Table 1 Optimum conditions for the determination of Cu, Ni and Zn in the aqueous and organic phases -: 2 - - - - Acetylene flow-rate/ Aspiration rate/ 1 rnin-1 ml min-1 Linear Air range of Wave- Slit- Lamp flow-rate/ Aqueous IBMK Aqueous IBMK calibration Metal lengthhm width/nm current/mA 1 min- solution solution solution solution graph/mg 1-' c u 328.1 0.7 15 18.4 4.0 2.6 6 5 0.1-4.0 Ni 232.0 0.2 25 18.4 4.0 2.6 6 5 0.5-6.0 Zn 213.8 0.7 6 18.4 4.0 2.6 6 5 0.1-2.0 - 3 - 1 00 80 - s 60 0 .- +I g 40 CI 20 0 w I x'x\x 1.0 2.0 3.0 4.0 5.0 6.0 [HCll/mol dm-3 Fig.1 Effect of hydrochloric acid concentration on the percentage extraction of 1, Ni-; 2, Cu-; and 3, Zn- rhodanide complexes into IMBK: CuII, 3.10 X 10-5 rnol; N P , 6.10 x 10-5 rnol drn-3; Zn", 7.65 x rnol drn-3; aqueous phase, 20 ml; organic phase, 5 rnl; and NH4SCN, 0.4 rnol dm-3 plex from a strongly acidic medium (6 rnol dm-3 HCl) (Fig. 1, curves 2 and 3). Hence, extraction of the Pd-rhodanide complex should be carried out from 6 mol dm-3 HCI in order to remove Pd selectively, thus allowing the subsequent determination of Cu, Ni and Zn by flame atomic absorption spectrometry. It was established that two extractions for 1 min each in the presence of at least a 5-fold molar excess of the reagent (NH4SCN) with respect to Pd was sufficient for the complete removal of Pd into IBMK.Determination of Cu and Ni The medium, after separation of Pd, is strongly acidic (6 rnol dm-3 HCI). Extraction of Cu-, Ni- and Zn-PDC complexes into IBMK under these conditions has not been fully described in the literature.5-8 However, it has been shown9 that these metals form complexes with APDC over a wide pH range (1-14). Reducing the acidity of the solutions to be analysed from 6 rnol dm-3 to a pH >1 causes additional difficulties, because of the need to introduce large amounts of NaOH, which might lead to contamination. Hence, the possibility of carrying out the extraction procedure from strongly acidic media was investigated. It has been reportedlo that 90% extraction can be achieved from 6 mol dm-3 HCI media.Our investigations, however, do not entirely support these data (Fig. 2). Complete extraction of Cu can in fact be obtained in the range 1-6 rnol dm-3 HC1. However, in acidic media >3 mol dm-3, the total extraction of Ni into IBMK as its APDC complex is not possible. The extraction procedure must therefore be carried out at HCI concentrations of up to 3 mol dm-3 in order to obtain the simultaneous extraction of Cu and Ni. Unfortunately, under these conditions Zn remains O t - PH 8.0 6.0 4.0 2.0 2.0 4.0 6.0 [HCl]/mol dm-3 Fig. 2 Effect of ( a ) pH; and (b) hydrochloric acid concentration on the percentage extraction of 1, Cu-; 2, Ni-; and 3, Zn-PDC complexes into IBMK: Cull, 1.55 x 10-5 mol dm-3; Ni", 3.05 X 10-5 rnol dm-3; Zn", 3.88 x lo-" rnol aqueous phase, 40 ml; organic phase, 5 ml; and APDC.3.00 x rnol drn-3 entirely in the aqueous phase and hence its determination together with Cu and Ni in the organic phase is not possible. It was found that the ratio of the volume of the aqueous phase to the volume of the organic phase (Va : V,) in the range from 2 : 1 to 10 : 1 has no effect on the extent of extraction of Ni and Cu under the optimum conditions for the acidity where the extraction is nearly 100%. Hence, the acidity of the aqueous phase can be decreased by a 2-fold dilution with water. Total extraction of Cu and Ni as their APDC complexes into lBMK is achieved by using a 10-fold molar excess of the reagent for Cu and a 60-fold molar excess for Ni (Fig. 3). A large excess of the chelating agent is required to allow for decomposition of the reagent in strongly acidic media.6 A single extraction procedure for a period of 10 s is sufficient because the rate at which the metal chelates are formed and extracted increases as the acidity of the aqueous phase is increased.7 If the absorption signals for Cu and Ni are measured less than 30 min after the extraction procedure, then separation of the organic layer from the aqueous layer only is required.However, if the absorption signals are measured after more than 30 min it is necessary for the organic phase to be washed with water in order to remove the remaining acid and so prevent it from decomposing the Cu- and Ni-PDC complexes ,6 which would reduce the absorption signals. Hence, it was established that it is possible to determine Cu and Ni simultaneously in Pd(NH3)2C12 and Pd(NH&(N02)2 by employing extraction with APDC-IBMK in 3 rnol dm-3 HCl media after preliminary extraction of Pd into IBMK as its rhodanide complex.As mentioned above, under these condi- tions Zn remains in the aqueous phase. For the simultaneous determination of Zn, Cu and Ni, the extraction should be carried out at pH 2. Particular attention has to be paid to the blank; therefore, the base used forANALYST, MAY 1991, VOL. 116 527 Table 2 Determination of Cu, Ni and Zn in Pd(NH3)2C12 and Pd(NH3)2(N02)2 Compound Method Cucontent (YO) RSD* (%) Nicontent (YO) RSD* (YO) Zncontent Determination of Cu and Ni in strongly acidic media (3 mol dm-3 HC1)- Pd(NH3)2C12 Calibration graph (2.63 f 0.04) x 10-4 5.2 (5.25 4 0.04) x 3.6 - Standard additions (2.68 t 0.04) x 4.8 (5.32 t 0.04) x 4.8 - Pd(NH3)2(N02)2 Calibration graph (2.72 k 0.05) x lo--' 6.2 (7.20 4 0.03) X lo--' 2.5 - Standard additions (2.67 t 0.05) x 7.2 (7.25 t 0.04) x 10F4 3.9 - Determination of Cu, Ni and Zn at p H 2.0- Yo) RSD*(%) Pd(NH3)2C12 Calibration graph (2.75 k 0.05) x 10-4 7.1 (5.20 t 0.04) x 10-4 4.2 (3.61 k 0.04) x 10-4 5.1 Standard additions (2.70 t 0.05) x 10-4 6.8 (5.27 4 0.04) X 10-4 4.8 (3.67 * 0.04) X 10-4 4.7 Pd(NH3)2(NO& Calibration graph (2.67 4 0.04) x 10-4 5.3 (7.14 t 0.04) x 10-4 3.7 (4.72 k 0.04) x 3.9 Standard additions (2.63 4 0.04) x 10-4 5.5 (7.19 k 0.04) X 10-4 5.5 (4.80 f 0.04) X 5.2 * RSD = relative standard deviation.100 x-x-x--x t l / ** Y D Table 3 Recovery of Cu, Ni and Zn added to Pd(NH3)2C12 and Pd(NH3)2(N02)2 by flame atomic absorption spectrometry using the APDC-IBMK extraction system 0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 APDC/10-6 rnol Fig.3 Effect of the amount of APDC on the extraction of 1, Cu; and 2, Ni into IBMK: Cu", 1.55 x 10-5 mol dm-3; NiII, 3.05 x 10-5 rnol dm-3; HCl, 3 mol dm-3; aqueous phase, 40 ml; organic phase, 5 ml adjusting the pH must have been purified previously. As is shown in Fig. 2 the most suitable pH range is 2.0-2.5. At lower pH values the extraction of Zn is decreased considerably, whereas at higher values serious problems arise with the blank even when purified NaOH is used. At pH 2.0-2.5, a period of 10 s is not sufficient for the extraction procedure; a shaking time of 60 s is required.Because the extent of APDC decomposition decreases under these conditions, the mini- mum amount of the chelating agent required for quantitative extraction is much smaller; however, it is recommended that the procedure be carried out using the same amount of APDC as required for the procedure in strongly acidic media, viz., 3.00 X 10-3 rnol dm-3. Absorption signals measured for Cu, Ni and Zn in the organic phase are stable for more than 24 h, hence washing of the organic phase is not necessary. The following procedures were used for the determination of Cu, Ni and Zn in Pd(NH3)2C12 and Pd(NH3)2(N02)2. Separation of Pd An aliquot of Pd(NH3)2C12 or Pd(NH3)2(N02)2 solution was placed in a 100 ml separating funnel. A 10.3 ml volume of concentrated HCI plus 3 ml of a 20% solution of NH4SCN were added and the mixture was diluted to 20 ml with water.Then, 5 ml of IBMK were added and the mixture was extracted for 1 min. After the two phases had separated, the organic layer was discarded. The procedure was repeated with another 5 ml aliquot of IBMK and the organic layer was again discarded. Atomic Absorption Spectrometric Determination of Cu and Ni The aqueous phase from above was transferred into another separating funnel, and 10 ml of water plus 1 ml of a 1% solution of APDC were added. The mixture was shaken Recovery Expected/ Element Found/pg Added/pg yg M* Yo * c u 2.7 10 12.7 12.4 97.6 Ni 5.2 10 15.2 15.1 99.3 Zn 3.6 1 0 13.6 13.3 97.8 * Mean of five determinations. vigorously for about 10 s and allowed to stand for 10 min after which 5 ml of IBMK were added.The mixture was shaken for 30 s and, after the two phases had separated, the aqueous layer was discarded. The organic phase was collected in 5 ml calibrated tubes and diluted to the mark with pure IBMK. The organic layer was aspirated directly into the flame and the absorption signals for Cu and Ni were measured with background correction under the conditions given in Table 1. Atomic Absorption Spectrometric Determination of Cu, Ni and Zn The aqueous phase remaining after the separation of Pd was transferred into 50 ml beakers. An 18 ml volume of 25% NaOH was added and the pH adjusted to 2.0-2.5. One millilitre of a 1% solution of APDC was added and the mixture was transferred quantitatively into another 100 ml separating funnel and diluted to 50 ml with water.A 5 ml aliquot of IBMK was added and the mixture was shaken for 30 s. The aqueous phase was discarded and the organic layer collected in 5 ml calibrated tubes and diluted to the mark with pure IBMK. The organic phase was injected directly into the flame and the absorption signals for Cu, Ni and Zn were measured with the background correction under the con- ditions given in Table 1. Standard solutions for calibration graphs were prepared each time by the same extraction procedure, extracting the appropriate volumes of Cu, Ni and Zn solutions each with a concentration of 10 mg 1-1. The determination of Cu, Ni and Zn in Pd(NH3)2C12 and Pd(NH3)2(N02)2 was carried out by using the calibration graph and standard additions pro- cedures.The results are shown in Table 2. As can be seen, the results obtained using the standard additions method corre- spond to those obtained using the calibration graph method. In addition, similar results for Cu and Ni were obtained with the two procedures. Hence, the proposed method is accurate and free from interferences. Recoveries of Cu, Ni and Zn from the two Pd compounds were studied by carrying out standard additions of these metals to Pd(NH3)2C12 and Pd(NH3)2(N02)2 during the dissolution procedure. The results obtained for the recovery study are shown in Table 3.528 ANALYST, MAY 1991, VOL. 116 In conclusion, a reproducible and highly sensitive method for the determination of Cu, Ni and Zn by extraction with APDC-IBMK followed by flame atomic absorption spec- trometry has been developed. The method was applied to the determination of these elements in Pd(NH3)2C12 and Pd(NH3)2(N02)2- References 1 Sychra, V., Slevin, P. J . , Matousek, J., and Bek, F., Anal. Chim. Acta, 1970, 52, 259. 2 Teruo, I . , Bunseki Kagaku, 1966, 15, 109. 3 Popova, S. A., and Bratinova, S. P., J. Anal. At. Spectrom., 1990, 5,35. 4 Marczenko, Z., Photometric Determination of Elements, Mir, Moscow, 1971. 5 Dellien, I . , and Persson, L., Talanta, 1979, 26, 1101. 6 Takada, T., Talanta, 1982, 29, 799. 7 Murakami, M., and Takada, T., Talanta, 1985,32,513. 8 Murakami, M., and Takada, T., Talanta, 1990, 37, 229. 9 Watson, C. A . , Ammonium Pyrrolidinedithiocarbarnate, Mono- graph 74, Hopkin and Williams, 1974. Brook, R. R., Hoashi, M., Wilson, S. M., and Zhang R., Anal. Chim. Acta, 1989, 217, 165. 10 Paper 0102924F Received June 28th, 1990 Accepted January 3rd, 1991