Determination of Trace Amounts of Antimony Germanium and Tin in High-purity Iron by Electrothermal Atomic Absorption Spectrometry After Reductive Coprecipitation With Palladium TETSUYA ASHINO AND KUNIO TAKADA Institute for Materials Research Tohoku University Katahira 2-2-1 Aoba Sendai Miyagi 980-77 Japan Trace amounts of antimony germanium and tin in high-purity iron were quantitatively separated by a reductive coprecipitation technique with palladium and determined by electrothermal atomic absorption spectrometry. When sodium phosphinate (NaPH,O,) was used as a reductant antimony and germanium could be separated simultaneously from large amounts of iron. Similarly when sodium tetrahydroborate (NaBH,) was used germanium and tin could also be separated simultaneously. The atomic absorbances of antimony germanium and tin were increased by about 1.5 3.7 and 4.5 times respectively in the presence of palladium.The limits of detection (corresponding to three times the standard deviation of the blank) of antimony germanium and tin were 0.Ol9 O.Ol0 and O.o& pg g-' respectively. Keywords Atomic absorption spectrometry; reductive coprecipitation; high-purity iron; antimony; germanium; tin; palladium The characteristics of high-purity metals have been widely investigated. For example it has become apparent that 99.999% (5N) iron differs from 99.95% (3N5) iron in physical and chemical properties.' In order to assess the purity trace elements have to be determined. The reported methods for the determination of antimony and tin2-5 cannot really be applied to 5N iron.Methods for the determination of germanium in iron have not been reported. Therefore more sensitive methods for the determination of trace amounts of antimony germanium and tin in high-purity iron are needed. When the elements in metal or alloy samples are to be determined the sample is usually decomposed and dissolved and the sample solution is used for the determination. For the deter- mination of trace amounts of antimony germanium and tin spectr~photometry,~.~ voltammetry,8-'0 HG-ICP-AES," HG-MIP-AES,I2 HG-1CP-MS,l3 FAAS,I4 HGAAS"-I8 and ETAAS'9-23 have been reported. In the HG methods the elements for determination can be separated from the metal matrix and the sensitivity may be higher than in ETAAS. However the conditions for generation of the hydride must be chosen according to the sample composition and it is not possible to determine many elements from a single sample.Therefore it is difficult to determine trace amounts of anti- mony germanium and tin in a large number of iron samples. In NAA,24 decomposition of the sample is not required but special apparatus is needed. ETAAS is suitable for the deter- mination of trace amounts of antimony germanium and tin in high-purity iron because these elements can be determined from a single sample by the same procedure and no special techniques are needed. However it may be necessary to separate and concentrate trace amounts of these elements from iron not only because the sensitivity of the deter- mination of these elements by ETAAS is not sufficient but Journal of Analytical Atomic Spectrometry also because the sensitivity may be decreased by the presence of iron.Therefore preconcentration methods have also been r e p ~ r t e d . ' > ~ * ~ ~ ~ ~ We have reported on the determination of trace amounts of selenium and tellurium in several metals and alloys by ETAAS after reductive coprecipitation with p a l l a d i ~ m . ~ ~ . ~ ~ This method is useful because trace amounts of selenium and tellurium can be separated simultaneously from large amounts of matrix metals and these elements can be sensitively determined by ETAAS when palladium is used as a chemical modifier. In order to develop a procedure for the highly sensitive determination of trace amounts of antimony germanium and tin in high-purity iron we investi- gated and employed ETAAS after reductive coprecipitation with palladium.EXPERIMENTAL Apparatus For the determination of antimony germanium and tin a 2-9000 simultaneous multi-element atomic absorption spec- trometer (Hitachi Tokyo Japan) was used with Zeeman-effect background correction. A hollow cathode lamp was used as the light source. The operating conditions are listed in Table 1. For dissolution and coprecipitation of the sample a PTFE beaker was used. For sealing the PTFE beaker a sealon film (Fuji Film Tokyo Japan) was employed. For filtration a membrane filter (Nuclepore polycarbonate pore size 0.2 pm Coster Cambridge MA USA) was used. Reagents Antimony standard solution (1.00 mg 1-I). A 0.100 g amount of metallic antimony (99.999%) was dissolved in 20ml of 7 moll-' HN03 together with 1 g of tartaric acid on a hot- plate and diluted to 100ml with water.The solution was diluted with tartaric acid and water before use. Germanium standard solution A (1.00 mg 1- '). A 0.100 g amount of metallic germanium (99.999%) was dissolved in 20 ml of 7 moll-' HNO together with 1 g of tartaric acid on a hot-plate and diluted to 100ml with water. The solution diluted with tartaric acid and water before use. Germanium standard solution B (1.00 mg 1-I). A 0.144 g amount of germanium(rv) oxide (GeO 99.999%) was dissolved in 2.5 moll-' NaOH solution on a hot-plate after which 20 ml of 14 moll-' HN03 were added and the solution was trans- ferred into a calibrated flask. The solution was then diluted to 100 ml with water.It was diluted with water before use. Tin standard solution (1.00 mg 1-l). A 0.100 g amount of metallic tin was dissolved in 20 ml of 6 mol I-' HCl on a hot- plate after which the solution was transferred into a calibrated flask and 40 ml of 12 mol I-' HCl and 1 g of tartaric acid were Journal of Analytical Atomic Spectrometry August 1996 Vol. 11 (577-583) 577Table 1 Instrumental and operating conditions for ETAAS Instrument Hitachi 2-9000 :simultaneous multi-element atomic absorption spectrometer Sample injection Light source Background correction Carrier gas flow rate Interrupted gas flow rate Furnace Injection volume Element Wavelength Lamp current Temperature programme- Drying Ashing Atomization Cleaning Autosampler Hollow cathode lamp Polarized Zeeman effect 300 ml min-l 0 ml min-' Pyrolytic graphite coated graphite tube 10 pl Sb Ge 231.2 nm 8.0 mA 80-150°C 30 s 150-1200 "C 30 s 1200-1200 "C 30 s 25OO0C 5 s 3000"C 10 s 265.2 nm 10.0 mA 80-150 "C 30 s 150-1400 "C 30 s 1400-1400 "C 30 s 2600"C 5 s 3000"C 10 s Sn 224.6 nm 8.0 mA 80-150°C 30 s 150-1600 "C 30 s 1600-1600 "C 30 s 3000 "C 10 s 2700"C 5 s added.The solution was then diluted to 100ml with water. It was diluted with tartaric acid solution (5% m/v) before use. Palladium standard solution ( 3 g I-' solution was used for calibration 0.6 g 1-' solution was used for coprecipitation). A 0.300 g amount of metallic palladium was dissolved in 5 ml of 14moll-' HN03 and diluted to 100ml with water. The solution was used for calibration. A 20ml aliquot of the solution was diluted to 100ml with water and used for coprecipitation.Sodium ethylenediaminetetraacetate (EDTA-Nu) solution (0.4 moll-' = 15% m/v). A 75 g amount of EDTA-Na was dissolved in 20 ml of 15 moll-' aqueous NH3 on a hot-plate and diluted to 500ml with water. Sodium phosphinate (NaPH,O,) solution (1.14 mol I-' = 10% m/u). A 12g amount of NaPH2O2.H2O was dissolved in water and diluted to 10 ml with water. Sodium tetrahydroborate (NaBH,) solution (1.59 moll-' = 6% m/v). A 6 g amount of NaBH was dissolved in water and about 0.1 g of NaOH was added. The solution was then diluted to 100ml with water. EDTA-Na NaPH,O and NaBH solutions were kept in a polyethylene bottle and NaPH202 and NaBH solutions were kept at a temperature of less than 10°C. Distilled de-ionized water was used for the preparation of all standard and sample solutions.All the reagents used were of analytical-reagent grade. Samples The high-purity iron used was Grade 1 iron powder (Johnson Matthey Materials Technology Royston Hertfordshire UK). JSS 001-4 and 002-4 Pure Iron (Japanese Iron and Steel Certified Reference Materials The Japan Iron and Steel Federation Tokyo Japan) were used as reference samples. Procedure Preparation of sample A 1 g amount of the sample was weighed into a PTFE beaker then 10 ml of HN0,-HCl (1 + 1) and 10 ml of water were added and the sample was dissolved on a hot-plate. Subsequently 20 ml of H2S04-H,P04 (1 + 1) were added and the solution was heated to fumes. After cooling to room temperature the solution was diluted with 30 ml of water and 50 mi of EDTA-Na solution were added. Then 6.25 moll-' NaOH solution was added until the pH reached 4.0 whereupon the solution was boiled in order to produce the EDTA-Fe complex.After cooling to room temperature 1 g of ascorbic acid 578 Journal of Analytical Atomic Spectrometry August 1996 was added. For the determination of antimony 6.25 moll-' NaOH solution was added until the pH reached 7.0 after which 5 ml of 0.6 g 1-' palladium standard solution and 5 ml of NaPH solution were added. The solution was sealed with sealon film and left at room temperature for at least 3 h. For the determination of tin 6.25 mol I-' NaOH solution was added until the pH reached 10.0 after which 5 ml of 0.6 g 1-' palladium standard solution and 5 ml of NaBH solution were added.The solution was sealed with sealon film and left at room temperature for at least 4 h. For the determination of germanium both methods could be adopted. After standing the precipitate was collected on a membrane filter. Immediately thereafter the membrane filter with the collected precipitate was transferred into a glass beaker and the precipitate was dissolved in 1 ml of 0.66mol1-' (=lo% m/v) tartaric acid solution 1.5 ml of HNO and one drop (about 0.05 ml) of HC1 at room temperature. The solution was transferred into a calibrated flask and diluted to exactly 10 ml with water. The blank was prepared by the same procedure described above but without the sample. Determination by E TAAS A lop1 aliquot of the sample or blank solution was injected into the graphite furnace.The atomic absorbance was measured under the conditions shown in Table 1. Preparation of solutions for the calibration graph Volumes (0-15 ml) of the antimony germanium and/or tin standard solution were transferred into calibrated flasks. Then 5 ml of 3 g 1-' palladium solution 7.5 ml of HNO 5 ml of 0.66 moll-' tartaric acid solution and five drops (about 0.25 ml) of HC1 were added to each solution. Finally the solutions were diluted to exactly 50 ml with water. RESULTS AND DISCUSSION Effects of Iron and Chemical Modifier on the Determination of Antimony Germanium and Tin by ETAAS The effects of iron and chemical modifiers on the determination of antimony germanium and tin by ETAAS under the con- ditions mentioned in Table 1 were examined and the results are shown in Fig.1. The method employed was as follows solutions containing 0.1 mg I-' of antimony germanium and/or tin and lOOmgl-' (1000 times the concentration of these elements) of various chemical modifiers or 10 g 1-' (100000 times the concentration of these elements) of iron VOl. 115 I t n I I I I I 4 f !I3 !I U 0 Fig. 1 Effect of chemical modifier on the atomic absorbance of Sb Ge and Sn; Sb Ge and Sn 0.1 mg 1-l; Pd Mg and Cu 100 mg 1-'; Fe logl-'. Atomic absorbances were normalized to 1 when no modifier was present were prepared. Then the atomic absorbances of antimony germanium and/or tin were measured. Palladium palladium- magnesium and palladium-copper were used as chemical modifiers. For antimony germanium and tin the increase in atomic absorbance was highest in the presence of palladium only.Therefore of the chemical modifiers studied palladium appeared to be the most useful for the ETAAS determination of these elements. Also it was found that the absorbances of antimony germanium and tin were decreased in the presence of a 100 000-fold concentration of iron. In spite of the addition of palladium to the iron-containing solution the same results were obtained. Therefore for the determination of trace amounts of antimony germanium and tin in high-purity iron by ETAAS it is necessary to separate these elements from iron; in the method adopted palladium was used as a carrier for coprecipi t a t ion. Concentration of Palladium as Chemical Modifier The effect of the amount of palladium on the atomic absorbances of antimony germanium and tin was examined.The results obtained are shown in Fig. 2. The method employed was as follows solutions containing 0.1 mg I-' of antimony ger- manium and/or tin and palladium concentrations ranging from 0.06 to 0.66 g 1-' were prepared. Then the atomic absorbances of antimony germanium and/or tin were measured. For all three elements the highest atomic absorbance was obtained when the concentration of palladium was 0.3 g 1-'. Therefore the concentration of palladium adopted as a chemical modifier for the determination by ETAAS was 0.3 g 1-' i.e. the amount of palladium used as a carrier for coprecipitation was 3 mg. Ashing Temperature for ETAAS The effect of ashing temperature on the atomic absorbances of antimony germanium and tin was examined.The results o*20 I 0.15 8 0.10 $ < 0.05 0.00 I 0 400 800 1200 1600 2000 Ashing temperaturePC Fig. 3 Relationship between the ashing temperature and the atomic absorbance of Sb; Sb 0.1 mg I-'; Pd 300 mg 1-'; + Sb and Pd; -+- Sb only 0.16 0.12 1 0.08 51 .n 4 0.04 0.00 1 . I . I . I . I 400 800 1200 1600 2000 Ashing temperaturePC Fig. 4 Relationship between the ashing temperature and the atomic absorbance of Ge; Ge 0.1 mg 1-I; Pd 300 mg 1-'; + Ge and Pd; -+- Ge only [Pd] I g r' 0 400 800 1200 1600 2000 Ashing temperaturePC Fig. 2 Relationship between the palladium concentration and the absorbance of Sb Ge and Sn; Sb Sn and Ge 0.1 mg 1-'; + Sb; + Ge; and + Sn Fig. 5 Relationship between the ashing temperature and the atomic absorbance of Sn; Sn 0.1 mg 1-I; Pd 300 m g 1-I; - Sn and Pd; + Sn only Journal of Analytical Atomic Spectrometry August 1996 Vol.11 579obtained are shown in Figs. 3-5. The method employed was as follows solutions containing 0.1 mg I-' of antimony ger- manium and/or tin with and without 300 mg 1-' of palladium were prepared. The atomic absorbances of antimony ger- manium and/or tin were then measured in the ashing tempera- ture range from 200 to 2000°C. For the determination of antimony the highest absorbance was obtained at 1200 "C; the absorbance decreased above 1300 "C regardless of whether or not the solution contained palladium. Therefore the ashing temperature adopted for antimony was 1200 "C. For the deter- mination of germanium the absorbance was almost constant below 1600 "C; the highest absorbance was obtained at 1400 "C when palladium was not present.Therefore the ashing tem- perature adopted for germanium was 1400 "C. For the determi- nation of tin the highest absorbance was obtained at 1600°C; the absorbance decreased above 1700 "C when palladium was present. However the absorbance decreased above 1300 "C when palladium was not present. Therefore the ashing tem- perature adopted for tin was 1600°C. Atomization Temperature for ETAAS The effect of atomization temperature on the atomic absorbances of antimony germanium and tin was examined. The results obtained are shown in Fig. 6. The method employed was as follows solutions containing 0.1 mg I-' of antimony ger- manium and/or tin and 300 mg 1-' of palladium were prepared.The atomic absorbances of antimony germanium and/or tin were then measured in the atomization temperature range from 1700 to 3000°C. For the determination of antimony the absorbance was almost constant between 2000 and 3000"C and the highest absorbance was obtained at 2500 "C. Therefore the atomization temperature adopted for antimony was 2500 "C. For the determination of germanium the absorbance was almost constant between 2600 and 3000 "C. Therefore the atomization temperature adopted for germanium was 2600 "C. For the determination of tin the highest absorbance was obtained at 2700 "C. Therefore the atomization temperature adopted for tin was 2700 "C. Selection of Reductant for Coprecipitation In order to select a suitable reductant for the coprecipitation and separation of trace amounts of antimony germanium and tin from high-purity iron the effect of reductants was examined.The method employed was as follows high-purity iron (1 g) was weighed into a PTFE beaker and 1 ml of antimony germanium and/or tin standard solution was added. The Oa30 * mixture was dissolved in HNO and HCI then H2S04 and H,PO were added and the solution was heated to fumes. The analyte elements were separated and concentrated by coprecipitation with palladium using various reductants. The precipitate was dissolved and the solution diluted to 10m1 then the elements were determined by ETAAS. Ascorbic acid sodium hydrogensulfite-hydrazine sulfate (NaHS0,-hydrazine) NaPH202 NaBH and metallic zinc were examined as reductants and the results obtained are shown in Table 2.Ascorbic acid and NaHS0,-hydrazine have been used for coprecipitation of selenium and t e l l u r i ~ m . ~ ~ . ~ ~ NaPH 02 NaBH and metallic zinc are stronger reductants than ascorbic acid and NaHS0,-hydrazine. Antimony germanium and tin are not reduced as easily as selenium or tellurium because the redox potentials of Sb'" Ge" and Sn" are lower than those of SeIV and Te". Therefore it was decided to investigate stronger reductants than ascorbic acid or NaHS0,-hydrazine. When ascorbic acid or NaHS0,-hydrazine was used as the reductant antimony germanium and tin were not recovered. It was assumed that these elements could not be reduced to the metal by these reductants. When NaBH was used the same result was obtained.However it was assumed that antimony germanium and tin were vaporized as their hydrides because the reducing power of NaBH was too strong. On the other hand when NaPH,O was used hydrogen was generated on the surface of palladium formed by reduction. In solutions containing large amounts of iron these elements and palladium were not precipitated. It is believed that since the iron(1IF phosphinate complex is formed NaPH,O cannot act as a reductant. In order to overcome these problems a procedure was adopted in which the matrix iron was masked by EDTA-Na28.29 and antimony germanium and tin were coprecipitated with NaPH,O or NaBH at pH > 4.0. When NaPH,O was used both antimony and germanium could be quantitatively recovered at pH 7.0. When NaBH was used both germanium and tin could be quantitatively recovered at pH 10.0.When metallic zinc was used only tin was quantitatively recovered. Therefore for the determination of antimony NaPH,O2 is suitable as a reductant. For the deter- mination of tin NaBH is suitable. For the determination of germanium both NaPH20 and NaBH can be used. Effects of pH on the Coprecipitation The effect of pH on the recoveries of antimony germanium and tin by coprecipitation with palladium using NaPH20 or NaBH was examined. The results obtained are shown in Figs. 7 and 8. The method employed was as follows high- purity iron (1 g) was weighed into a PTFE beaker and 1 ml of antimony germanium and/or tin standard solution was added. The mixture was dissolved and heated to fumes using the same procedure as described above after which the analyte 1W ZOO0 2400 2800 3200 Atomizing temperaturePC Fig.6 Relationship between the atomization temperature and the atomic absorbance of Sb Ge and Sn; Sb Ge and Sn 0.1 mg 1-I; Pd 300mg1-'; + Sb; + Ge; and + Sn 580 Journal of Analvtical Atomic Svectrometrv. August 1986. Table2 Recoveries of Sb Ge and Sn by coprecipitation using several reductants Recovery (YO) Reductant Ascorbic acid* NaHS0,-hydrazine sulfate* NaPH2O2* NaPH202 (pH 4.0) NaPH,02 (pH 7.0) NaBH,* NaBH (pH 7.0) NaBH (pH 10.0) Zn metal* Sb NDt ND ND 3.45 101.0 ND 15.6 69.4 2.35 Ge ND ND ND 19.7 96.9 ND 51.0 97.5 90.2 Sn ND ND ND 19.2 41.1 ND 85.1 99.8 103.4 * No pre-reduction with ascorbic acid and no addition of EDTA-Na. t ND = Not detected. VOl. 11Fig.7 Relationship between the pH for coprecipitation using NaPH,O and the recoveries of Sb Ge and Sn; Fe 1 g; Sb Ge and Sn 1 pg; + Sb; It Ge; and + Sn p 6 4 4 0 20 0 Fig. 8 Relationship between the pH for coprecipitation using NaBH and the recoveries of Sb Ge and Sn; Fe 1 g; Sb Ge and Sn 1 pg; + Sb; + Ge; and -0- Sn elements were separated by coprecipitation with palladium. When NaPH,02 was used as the reductant the pH was adjusted to 4.0-8.0. On the other hand when NaBH was used the pH was adjusted to 7.0-12.0. The precipitate was dis- solved and the solution diluted to 10ml; finally the three elements were determined by ETAAS. When NaPH,O was used as the reductant antimony was quantitatively recovered at pH 6.5-7.0; germanium was quantitatively recovered at pH 7.0.However tin was not completely recovered in the pH range examined. Therefore for the determination of antimony or germanium NaPH,O was used as the reductant and the pH was adjusted to 7.0. On the other hand when NaBH was used germanium and tin were quantitatively recovered at pH 10.0. However antimony was not completely recovered in the pH range examined. Therefore for the determination of germanium or tin NaBH was used as the reductant and the pH was adjusted to 10.0. Amount of Reductant The effect of the amount of NaPH,O or NaBH used as reductant for coprecipitation on the recoveries of antimony germanium and tin was examined. The results obtained are shown in Fig. 9. The method employed was as follows high- purity iron (1 g) was weighed into a PTFE beaker and 1 ml of antimony germanium and/or tin standard solution was added.The mixture was dissolved and heated to fumes using the same procedure as described above after which 0.5-2.5 g of NaPH202 or 0.15-0.75 g of NaBH was added. The analyte 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Reductant mass / g Fig. 9 Relationship between the mass of reductant and the recoveries of Sb Ge and Sn; Fe 1 g; Pd 3 mg; Sb Ge and Sn 1 pg. With NaPH,O, + Sb; It Ge. With NaBH, + Ge; + Sn elements were separated by coprecipitation with palladium. The precipitate was dissolved and the solution diluted to 10 ml; finally the elements were determined by ETAAS. It was found that when 0.5-1.Og of NaPH202 was used both antimony and germanium were quantitatively recovered. On the other hand when 0.15-0.45 g of NaBH was used germanium and antimony were quantitatively recovered.It was assumed that when more than 1.0 g of NaPH202 and more than 0.45 g of NaBH was used all three elements were vaporized as their hydrides. Therefore it was decided to use 0.5 g of NaPH202 and 0.3 g of NaBH,. Standing Time for Coprecipitation The standing time for coprecipitation for the separation of antimony germanium and tin was examined. The results obtained are shown in Fig. 10. The method employed was as follows high-purity iron (1 g) was weighed into a PTFE beaker and 1 ml of antimony germanium and/or tin standard solution was added. The mixture was dissolved and heated to fumes using the same procedure as described above after which the analyte elements were separated by coprecipitation with palladium using standing times of 1-4 h.The precipitate was dissolved and the solution diluted to 10ml; finally the elements were determined by ETAAS. It was found that when the standing time was more than 3 h quantitative recoveries 1 2 3 4 5 Standing time / h Fig. 10 Relationship between the standing time for coprecipitation and recoveries of Sb Ge and Sn; Fe 1 g Sb Ge and Sn 1 pg. With NaPH,O, * Sb; -t Ge. With NaBH, + Ge; -+- Sn Journal of Analvtical Atomic Soectrometrv. AuPust 1996. Vil. 1 1 581of antimony and germanium were obtained whereas tin was quantitatively recovered after more than 4 h. Therefore when NaPH202 was used as the reductant the standing time employed was 3 h; when NaBH was used the standing time employed was 4 h. Acid Dissolution of Sample It has been reported that germanium is vaporized as its chloride during dissolution of samples in hydrochloric acid.,' The effect of hydrochloric acid on the recoveries of antimony germanium and tin was examined.The results obtained are shown in Fig. 11. The method employed was as follows high- purity iron (1 g) was weighed into a PTFE beaker and 1 ml of antimony germanium A (made from the metal) germanium B (made from the oxide) and/or tin standard solution was added. The mixture was dissolved in HCl-HNO solutions containing different proportions of HCl and heated to fumes by the same procedure as described above after which the analyte elements were separated by coprecipitation with p 8s cr! 8 0 - 7s - 0.5 1 .o 1.5 2.0 HCl HN03 Fig. 11 Relationship between the volume ratio of acid (HCI HNO,) used for dissolution and the recoveries of Sb Ge and Sn; Fe 1 g; Sb Ge and Sn 1 pg; .Sb; A Ge A (from metal); V Ge B (from oxide); and 0 Sn Table 3 Analytical results for mixed solution sample palladium. The precipitate was dissolved and the solution diluted to 10ml; finally the elements were determined by ETAAS. It was found that when the HCI:HNO ratio was greater than 1.5 the recovery of tin was also lower. It is assumed that not only germanium but also antimony and tin are vaporized as their chlorides during dissolution and heating to fumes. The recovery of germanium from solution B was less than from solution A when the HC1 HNO ratio was greater than 1. This was because the tartaric acid present in the germanium A standard solution reduced the vaporization of germanium markedly.Therefore a 1 + 1 HC1-HNO3 mixture was used for dissolution of samples. Calibration Graph For the preparation of calibration graphs an aqueous standard solution was used and the relationship between the atomic absorbance and the concentration of antimony germanium and/or tin was examined. A straight line passing through the origin was obtained for a concentration of <0.36 mg 1-1 of antimony c 0.40 mg 1-l of germanium and < 0.30 mg 1 - of tin. The limit of detection (three times the standard deviation of the blank n = 10) of the proposed method was 0.Ol9 pg g-' of antimony O.Ol0 pg g-' of germanium and 0.031 pg 8-l of tin in 1 g of sample. Analysis of Mixture Solution Sample Antimony germanium and tin in a mixture solution sample which consisted of 1 pg of each of these elements and 1 g of high-purity iron were determined according to the proposed method.The results obtained are shown in Table 3. All three elements were quantitatively determined. Analysis of Reference Samples The proposed method was applied to the determination of antimony germanium and tin in reference samples. The results obtained are shown in Table 4. In the determination of antimony and tin the analytical values agreed with the Added/ Reductant Pg Fe ( 1 g) Fe ( 1 g) NaBH Fe (1 g)+Sn 1.02 Fe (1 g) Fe (1 g)+Ge 1.00 Fe (1 €9 NaPH,O Fet (1 g)+Sb 1-04 - Fe (1 g)+Ge 1 .OO - - - Found/ 1.18 1.04 f 0.01 0.98 & 0.00 2.32 & 0.00 < 0.02 < 0.01 1.328 f 0.04 1.01 * 0.00 < 0.01 Recovery ("/.I n 100.3 6 4 98.1 6 4 98.0 5 4 100.9 5 4 - - - - RSD* 1.33 0.660 W O ) - - 0.392 3.29 2.18 - * Relative standard deviation.t High-purity iron; Johnson Matthey Grade 1 powder. Table 4 Analytical results for reference samples Sample Reduct ant Element Pure Iron NaPH,02 Sb (JSS 001-4) Ge NaBH Sn Ge Pure Iron NaPH 0 Sb (JSS 002-4) Ge NaBH Sn Ge Reference value*/ Analytical value/ P8 g-' pg g-' - 0.18 f 0.011 - 0.05~ 0.00 < 0.4 0.29 0.00 < 1 0.55 & 0.03 0.566 & 0.01 < 1 0.64 +_ 0.03 - 0.056 f 0.006 - - 0.59 0.02 RSDt 5.86 2.02 5.94 3.21 5.23 4.09 (%) 11.2 12.4 * Non-certified value. t Relative standard deviation. 582 Journal of Analytical Atomic Spectrometry August 1996; Vol. 1 1reference values. In the determination of germanium good agreement was found between the value obtained using NaPH,O and that obtained using NaBH,.CONCLUSION Trace amounts of antimony germanium and tin in high-purity iron were quantitatively separated and concentrated by reductive coprecipitation with palladium. The advantages of the proposed method are that when palladium is used as a chemical modifier the ETAAS determination of all three elements is highly sensitive. When using NaPH202 both antimony and germanium were separated simultaneously and when using NaBH both germanium and tin were separated simultaneously and quantitatively by a similar procedure from matrix iron. The limit of detection for these elements using the proposed method was about 10-100 times lower than with existing method^.^.^ Therefore the proposed method is useful for the determination of trace amounts of antimony germanium and tin in high-purity iron.The authors thank Professor Kichinosuke Hirokawa for his continuing guidance. REFERENCES Abiko K. Sci. Am. Jpn. Version 1993 23 20. Takada K. Inamoto I. and Okano T. Microchem. J. 1994 49 291. Takada K. Muter. Jpn. 1994 33 84. JIS G1235-1981 Methods for Determination of Antimony in Iron and Steel 1982. Hosoya M. Konno H. and Takeyama S. Bunseki Kagaku 1983 32 444. Danzaki Y. Bunseki Kagaku 1988 37 153. Sun Q. Wang H. T. and Mou S. F. J. Chromatgr. 1995,708,99. 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Adeloju S. B. O. and Pablo F. Anal. Chim. Acta 1992 270 143. Sun C. Q. Gao Q. A. Xi J. B. and Xu H. D. Anal. Chim.Acta 1995 309 89. Waller P. A. and Pickering W. F. Talanta 1995 42 197. Uggerud H. and Lund W. J. Anal. At. Spectrom. 1995,10,405. Lunzer F. Pereiro-Garcia R. Bodelgarcia N. and Sanz- Medel A. J. Anal. At. Spectrom. 1995 10 311. Feldrnann J. Grumping R. and Hirner A. V. Fresenius’ J. Anal. Chem. 1994 350,228. Venkaji K. Naidu P. P. and Rao T. J. P. Talanta 1994,41 1281. Liu X. Z. Xu S. K. and Fang Z. L. At. Spectrosc. 1994 15,229. Schulze G. and Lehmann C. Anal. Chim. Acta 1994 288 215. Burguera M. Burguera J. L. Rivasa C. Carrero P. Brunetto R. and Gallignani M. Anal. Chim. Acta 1995 308 143. Burns D. T. Chimpalee N. and Harriott M. Fresenius’ J. Anal. Chem. 1994 349 530. Rettberg T. M. and Beach L. M. J. Anal. At. Spectrom. 1989 4 427. Xuan W. Spectrochim. Acta Part B 1992 47 545. Sahayam A. C. and Gandadharan S. Can. J. Appl. Spectrosc. 1994 39 61. Ivanova E. Stoimenova M. and Gentcheva G. Fresenius’ J. Anal. Chem. 1994 348 317. Morishige Y. Hirokawa K. and Yasuda K. Fresenius’ J. Anal. Chem. 1994 350 410. Ira P. and Macfarlane A. M. J. Radioanal. Nucl. Chem. 1994 182 427. Sayama Y. Fukata T. and Kuno Y. Bunseki Kagaku 1995 44 569. Ashino T. Takada K. and Hirokawa K. Anal. Chim. Acta 1994 297 443. Ashino T. and Takada K. Anal. Chim. Acta 1995 312 157. Ishikuro M. Hosoya M. and Takada K. Bunseki Kagaku 1991 40 71. Itagaki T. Ishikuro M. and Takada K. Bunseki Kagaku 1994 43 569. Gridchina G. I. Zauod. Lab. 1968 34 1194. Paper 6/01 042C Received February 13 1996 Accepted May 17 1996 Journal of Analytical Atomic Spectrometry August 1996 Vol. 1 1 583