UDK 669.15'777'4:543.428.2 ISSN 1580-2949 Izvirni znanstveni članek MTAEC9,36(6)325(2002) T. DRGLIN: A SIMULTANEOUS DETERMINATION OF THE LEVELS OF LEAD, SELENIUM AND TELLURIUM ... A SIMULTANEOUS DETERMINATION OF THE LEVELS OF LEAD, SELENIUM AND TELLURIUM IN STEEL USING ICP-AES SIMULTANO DOLOČEVANJE Pb, SeIN TeV JEKLU Z ICP-AES Tatjana Drglin Institute of Metals and Technology, Lepi pot 11, 1000 Ljubljana, Slovenia tatjana.drglinŽimt.si Prejem rokopisa - received: 2002-11-12; sprejem za objavo - accepted for publication: 2002-12-05 We have developed a useful procedure for the simultaneous determination of Te and Pb in low-alloy steels in the 5.0-100 µgg-1 range using ICP-AES after the chemical separation of the iron and the pre-concentration of Te and Pb by a batch process. The procedure is based on the liquid-liquid extraction of macro quantities of iron (III) from 7.7 M hydrochloric acid solution. The analytical and instrumental parameters were optimised. The interference caused by some alloying elements is discussed. The bias of the analytical procedure and the accuracy of the proposed method were tested by means of "spiked" samples and standard reference materials of steels. The limit of detection (LOD) and the limit of quantification (LOQ) were calculated for the investigated elements. Key words: low-alloy steel, Se, Te, Pb, separation, ICP-AES V prispevku je opisan postopek za simultano določevanje Te in Pb za koncentracijsko območje (5,0 - 100) µg g-1 v malo legiranem jeklu z ICP-AES po predhodni ločitvi Fe in predkoncentriranju. Ločitev makrokoličin Fe(III) je potekala na osnovi ekstrakcijeiz 7,7 M raztopineklorovodikovekisline. Izbrali smo optimalneanalitskein operacijskerazmereter opisali interference elementov, ki so po ločitvi ostali v raztopini. Uspešnost ločitve in pravilnost metode smo ocenili z analizo internih dodatkov in certificiranih referenčnih materialov jekel. Za preiskovane elemente smo določili mejo detekcije (LOD) in mejo določevanja (LOQ). Ključne besede: malo legirano jeklo, določitev Se,Te, Pb, ločitev, ICP-AES 1 INTRODUCTION The quality of a steel can be influenced both positively and negatively by the presence of metallic or semi-metallic elements like antimony, arsenic, bismuth, selenium, tellurium, tin and lead. Trace levels of certain elements can accumulate in grain boundaries and weaken the microstructure. Traces of surface-active elements such as As, Sb, Se, Sn, Te and others can also affect both the hot- and cold-working properties of many steels and alloys deleteriously because of their tendency to accumulate in grain boundaries, thereby weakening themicrostructureand influencing themagnetic properties of non-oriented steel sheets 1,2. The catastrophic failure of critical components in high-stress, high-temperature, high-corrosion, or high-radiation applications can often be directly related to trace levels of certain impurities 3. All these influences on the quality of steel are already detectable at very low concentrations (0.0002-0.02%). The determination of trace levels of these elements is therefore very important and the development of a suitably sensitive method is necessary. Different techniques have been reported for determining trace amounts of selenium or tellurium, these include: neutron activation analysis (NAA), X-ray fluorescence spectro-metry (XRF), spectrophotometry, voltametry 4, hydridegeneration inductively-coupled-plasma atomic emission spectrometry (hydride-ICP-AES) 5, hydride-generation atomic absorption spectrometry (hydride-AAS) 6-9, and electro-thermal atomic absorption spectrometry (ET-AAS) 10-12. Preconcentration methods have also been reported 13,14. Selenium and tellurium can be separated simultaneously after reductive coprecipitation with palladium using ascorbic acid 4. The determination of hydride-forming elements (As, Bi, Sn, Se, Te, Pb) in low-alloy steel was widely investigated, and HG AAS was suggested as a useful method 15-19. Lead may be extracted into a methyl isobutyl ketone solution of tri-n-octylphosphineoxide(TOPO-MIBK) and th organic layer read out by flame atomic absorption 20,21. High concentrations of iron can cause spectral interferences and increase the detection limit in the case of theICP-AES technique22. Theseparation of iron 23 often has to be carried out in order to lower the detection limit. In this paper a method for the extractive isolation of iron is proposed. Iron can be relatively easily extracted from 7,7 M hydrochloric acid solution prior to the determination of trace elements. Subsequent trace determinations are then free of interferences. The determination of Te and Pb that remained entirely in aqueous solution were free of interference and they were determined simultaneous using the ICP-AES method. Due to a lack of certified reference materials with such a MATERIALI IN TEHNOLOGIJE 36 (2002) 6 325 T. DRGLIN: A SIMULTANEOUS DETERMINATION OF THE LEVELS OF LEAD, SELENIUM AND TELLURIUM low content of Teand Pb wealso confirmed the accuracy by means of recovery. 2 EXPERIMENTAL 2.1 Apparatus A Perkin Elmer, model OPTIMA 3100 RL, simultaneous inductively-coupled-plasma spectrometer was used. The emission intensity measurements were carried out in the conditions shown in Table 1. The calibration curve(0-10.0 µg mL-1) for Se, Te and Pb was carried out with a multi-standard solution prepared from 1.0 mg mL-1 stock solutions of Se, Te and Pb in 2 % nitric acid solution. Table 1: Operating parameters for the inductively-coupled-plasma spectrometer Tabela 1: Operacijski pogoji delovanja atomskega emisijskega spektrometra z induktivno sklopljeno plazmo SPECTROMETER PERKIN ELMER, simultaneous radial, model OPTIMA 3100 RL Frequency 40 MHz, free-running Power output 1200 W Output stability < 0,1 % ICP SOURCE Plasma torch quartz / Al2O3 injector - 2 mm Plasma argon flow 15 L min-1 Auxiliary argon flow 0,5 L min1 Nebulizer argon flow 0,8 L min1 SAMPLE COMPARTMENT Spray chamber Scott-type, Ryton Nebulizer Gem Type Cross-flow, pneumatic Solution uptakerate 1 mL min-1 Flush time 30 s Signal integration time 10 s Integration for determination 3 2.2 Reagents The reagents were as follows: Pb standard solution (1.00mg L-1), Sestandard solution (1.00mg L-1) and Te standard solution (1.00 mg L-1): 100 mg of each metal was dissolved in HNO3 (1+1), and diluted to 100 mL with water. A multi-standard solution of Pb, Se and Te was prepared from stock solutions before use. Distilled water was used for all the preparations of the standard and sample solutions. All the reagents used were of analytical grade. Nitric and hydrochloric acid were of p.a. quality (Merck). The di-isopropyl ether used for extraction was of p.a. quality (Carlo Erba). 2.3 Separation procedure 1, 3 or 5 g of Fe was carefully dissolved in 50 mL of hydrochloric acid (1+1) and oxidised with 5 mL of concentrated nitric acid. The multi-standard solutions of 326 Se, Te and Pb (10.0 µg mL-1) were "spiked" into solutions of iron. The spiked solutions were evaporated to thewet soils. 50 ml of 7.7 M hydrochloric acid was added to each spiked sample. When the soils were dissolved the solution was transferred to the separating funnel. The beaker was rinsed with 7.7 M hydrochloric acid and 35 mL of di-isopropyl ether was added. Solutions were shaken for 30 seconds. When aqueous and ether phases were separated the aqueous phase was transferred to another separating funnel. The extraction procedure was repeated three times for the 1 or 3g samples and four times for the 5g samples. The di iso-propyl ether phase with extracted iron was discarded and the water phase was transferred into a beaker. 1 mL of nitric acid was added and the solution was evaporated to less than 25 mL. The test solution was cooled and diluted to 25 mL in a volumetric flask. Thesameprocedurewas used for thecertified reference materials. The Si and C that remained in solution were filtered immediately after the dissolution of thechips. 2.4 Calibration Thecalibration was carried out with a multi-standard solution in therange0.5-10.0 µg mL-1 Se, Te and Pb. The blank values for the reagents were discarded and the concentration of the investigated elements that were determined in iron without the spike was negligible. 3 RESULT AND DISCUSSIONS The interference effect of the Fe matrix on the simultaneous determination of Se, Te and Pb was substantially eliminated by the extraction of iron into the di iso-propyl ether. The efficiency of the extraction was >99 % for the1 or 3 g samples, and in thecaseof the5 g sampleit was >95 %. Nitric acid was added at thestage of dissolving thesampleand at thestageof evaporating thesolution after extraction to minimisetheSelosses. The spectral interferences caused by the alloying elements (Cr, Mo, Ni) that remained in solution after the extraction were checked up. The interferences free spectral lines for Se, Te and Pb determinations were chosen. Thebias of theanalytical procedurewas investigated by means of a recovery study 24. The different amounts of iron samples with spikes of different, but accurately known, amounts (Cref) of investigated elements were analysed. The recovery (R = Cobs/Cref ) as a ratio of the observed concentration Cobs obtained by the application of an analytical procedure to a material containing analyte at a reference level Cref and mean recovery (Re c) with standard deviation (s) was calculated (Table 2). The standard uncertainty was calculated as thestandard deviation of themean u(Rec) = s/n. In a perfect separation R would be exactly unity. A significance test was used to determine whether the mean MATERIALI IN TEHNOLOGIJE 36 (2002) 6 T. DRGLIN: A SIMULTANEOUS DETERMINATION OF THE LEVELS OF LEAD, SELENIUM AND TELLURIUM Table 2: Bias estimations by means of recovery study Tabela 2: Ocena odmika metode z uporabo izkoristka Value Element SeTe Pb (Rečc) 1)Š%] 74.3 93.9 98.0 s1)Š%] 17.8 3.9 3.4 u(Rec) 0.059 0.013 0.011 t 4.33 4.66 1.73 1) Themean (Rec) and sample standard deviation (s) are given as percentage recoveries. Table 3: Accuracy assessment by means of certified reference materials Tabela 3: Ocena pravilnosti z uporabo certificiranih referenčnih materialov CRM Se(µg g-1) Te(µg g-1) Pb (µg g-1) certified value determined ± std.dev. certified value determined ± std.dev. certified value determined ± std.dev. NBS 362 12a 9.6 ± 0.67 11a 8.7 ± 0.67 4.8 5.1 ±0.61 NBS 363 1.6a 1.1 ±0.25 9a 7.8 ± 0.35 19 17.8 ± 0.60 BS 2 n.d.b < d.l.c 14 13.1 ± 0.34 20 19.1 ±0.38 a value is not certified, but is given as approximate values for information b not determined c under detection limit recovery was significantly different from unity. The test statistic t was calculated using equation 1. 1-Re c u(Rec) (1) This valuewas compared with the2-tailed critical value tcrit, for n-1 degrees of freedom at 95 % confidence (where n is the number of results used to estimate (Re c). If t was greater than the critical value (tcrit = 2,26) then (Re c) is significantly different from 1. In this example a correction factor 1/(Re c) should beused. The accuracy of the method was also tested with some certified reference materials of steels (Table 3). The contents of Se, Te and Pb were sometimes only approximate. The limits of detection (LOD) for Te and Pb were 2 µg g-1 and limits of quantification (LOQ) for Teand Pb were4 µg g-1. 4 CONCLUSIONS Iron can be successfully removed by extraction into di iso-propyl ether. The interference that is caused by Cr that remains in solution after extraction can be overcome when the proper emission lines are chosen. The most sensitivespectral linefor Seat 203,985 nm is freeof interference. A strong interference is noticed for the Se determination at 203.990 nm, due to the presence of a small amount of Cr. TheCr spectral lineat 238.574 nm interferes with the Te determination at 238.578 nm. Te is determined at 214.281 nm, and the spectral line at 220.353 nm is chosen for the Pb determination. The results for the spiked samples are acceptable for the Pb determination, whereas the correction factor should be used for Te determination. Only a 74.3 % mean recovery is detected for the Se determination. The correction factor for the Se determination should not be applied, rather a wide investigation should be carried out. Preliminary work showed that the method is not suitable for the Sn, Sb and As determinations because these elements form chloride compounds that are partly extractable into di-isopropyl ether. 5 REFERENCES 1 F. W. Boulger, J. R. Becker, C. T. Olofson, and H. J. Henning, Metals Engineering Quarterly 12 (1972) 1, 17-22 2 C. A. Fischer, Wire World International 16 (1974) 245 3 T. R. Dulski, R. R. Bixler, Analytica Chimica Acta, 91 (1977) 199-209 4 A. Tetsuya, T. Kunio, Anal Chim Acta 312 (1995) 157-163 5 G. S. Pyen, S. Long, R. Browner, Appl Spectrosc 40 (1986) 2, 246-251 6 H. D. Fleming, R. G. Ide, Anal Chim Acta 83 (1976) 67-82 7 B. Welz, M. Melcher, Spectrochim Acta 36B (1981) 439-462 8 Ni Zhe-ming, He Bin, Han Heng-bin, J Anal At Spectrom 8 (1993) 995-998 9 W. Lund, T. Wickstrom, R. Bye, Anal Chim Acta 208 (1988) 347-350 10 W. Frech, Talanta 21 (1974) 565-571 11 W. Frech, Anal Chim Acta 77 (1975) 43-52 12 B. Welz, G.Schlemmer, R. Mudakavit, J Anal At Spectrom 3 (1988) 695-701 13 O. Kujirai, T. Kobayashi, K. Ide, E. Sudo, Talanta 29 (1982) 27 14 O. Kujirai, T. Kobayashi, K. Ide, E. Sudo, Talanta 30 (1983) 9 15 A. Osojnik, T. Drglin, Vacuum 43 (1992) 717-721 t MATERIALI IN TEHNOLOGIJE 36 (2002) 6 327 T. DRGLIN: A SIMULTANEOUS DETERMINATION OF THE LEVELS OF LEAD, SELENIUM AND TELLURIUM A. Osojnik, T. Drglin, Kovine, zlitine, tehnologije 27 (1993) 301-306 B. Welz, G. Bozsai, M. Sperling, B. Redzuik, J Anal At Spectrom 7 (1992) 505 S. J. Kummer, S.Gangadharen, J anal Spectrom 8 (1993) 127 R. Bye, Analyst 110 (1985) 85-86 K. E. Burke, Analyst 97 (1972) 19-28 K. Thornton, K. E. Burke, Analyst 99 (1974) 469-470 M. G. A. Korn, H. V. Jaeger, A. C. Ferreira, A. C. Spinola Costa, Spectroscopy Letters 33 (2000) 127-145 M. B. O. Giacomelli, J. B. B. Silva, A. J. Curtius, Talanta 47 (1998) 877-881 IUPAC, Pureand Appl Chem 71 (1999) 2, 337-348 328 MATERIALI IN TEHNOLOGIJE 36 (2002) 6