Comparision of Graphite Furnace - and Hydride Generation AAS for Trace Analysis of Tin in Steels and Nickel Alloys Primerjava elektrotermične - in hidridne tehnike AAS za analizo sledov kositra v jeklih in nikljevih zlitinah A. Osojnik, T. Drglin, Inštitut za kovinske materiale in tehnologije, Ljubljana The determination of tiri in various types of steel and nickel superalloys at lovv concentration level using graphite furnace atomic absorption spectrometry (GF AAS) and batch system hydride generation atomic absorption spectrometry (HG AAS) is described. The analytical and instrumentaI parameters for both methods vvere optimized. The interferences of matrix elements and some metalloids vvere investigated. Certified standard reference materials of steels and nickel alloys vvere used to test the methods. Some performances and characteristic data (detection limit, characteristic mass, accurace and relative standard deviation) of the tvvo methods are established and compared. The critical estimate of the both methods is performed. Key words: graphite furnace AAS, hydride generation AAS, interferences, steel, nickel alloys, tin determination. Opisana je metoda ter optimizirani instrumentalni in analizni parametri za določanje sledov kositra v jeklih in nikljevih zlitinah z metodo elektrotermične atomizacije (GF AAS) in hidridne tehnike-AAS (HG AAS). Študirali smo interference elementov osnove in nekaterih metaloidov. Rezultati so bili preverjeni s certificiranimi referenčnimi materiali jekel in nikljevih zlitin. Podani so nekateri karakteristični podatki (meja zaznavnosti, karakteristična masa, točnost, relativni standardni odmik) ter primerjava in kritična ocena obeh uporabljenih metod. Ključne besede: elektrotermična AAS, hidridna tehnika AAS, interference, jeklo, nikljeve zlitine, določanje kositra. Introduction Mechanical, phvsical and technological properties of various types of steel, and especiallv vacuum čast nickel superalloys for high temperature application strongly depend on trace elements contents such as Bi. Sb. Sn. As. Se. Te. and others. Because of their harmful effect alreadv at the p,g g 1 levels and lovver, the permissible concentrations of these elements are strongly limit ed. depends upon the element, the alloy type and application pur-pose. The traces of surface aetive elements such as Sb. Sn, Se, Te, and others influence the magnetic properties of nonoriented steel sheets. The knovvledge of their contents is one of the useful factors for studv of segregation phenomena. Therefore the determination of these elements is extremely important and the development of a suitable, sensitive analytical method is necessary. Graphite furnace - and hydride generation atomic absorption spectrometry scems to be the appropriate tech-niqucs for this purpose, because of their sensitivity and relative simplicitv. The main problems in the determination of tin by GF AAS are the formation of volatile Sn compounds, interactions of tin vvith graphite during the atomization step (1,2, 3, 4, 5) and mata interferences (9, 13). In order to overcome these problems, different chemical modifiers (5, 6. 7. 8). the oxidation of solution vvith nitric acid (1,5, 7. 9) and pretreatment of the graphite tubes vvith refraeto- ry metals (1, 4, 6, 10, 11) and aluminium solution (9) have been suggested. In this way the losscs of tin are diminished and effi-ciency of tin atomization is improved. The use of coated graphite tubes for tin determination has been proposed by many authors (1,4, 6. 9, 10, 11). This treatment results in the enhancement of sensitivity (1,4, 6, 7, 9, 10) and reproducibility of signal (9. 10). a reduetion of interferences (6, 10), and in the increased life time of the graphite tubes (10). The knovvledge and explanation of chemical reactions vvhich occur in graphite furnace during tin determination (1,2, 12) contribute to better understanding of the actions and the rolc of metal coatings, matrix modifiers and in-terfering elements. Determination of tin by HG AAS has been described by a number of authors (14-20), although many problems exist for this element. It is vvell knovvn that sensitivity of Sn signal depends strongly on the pH of the sample solution (14. 15, 19). Therefore for tin determination saturated solution of boric acid vvith addition of lovv concentrations of nitric (15) or hydrochlo-ric acid (19, 20) for standards, sample and carrier solutions is rec-ommended. Different reagents (acid, sodium tetrahydroborate reduetant solution, sodium hydroxide) and their concentrations significantly influence not only sensitivity and peak shapes but also interferences in tin determination by HG AAS. Among the difficulties described in the literature are also high blank values (18. 20. 21). memory effects (20, 21), and interferences from transition metals ions such as Fc. Ni, Co. Cu vvhich cause very serious reduetion of the lin signal (17, 22). The interferences caused hy those elements can hc partly or completely eliminat-ed. The most common way to eliminate the interferences is masking of interfering ions by different masking agents (17. 22. 23). although the changes of acid and reduetant solution concentrations are also useful for this purpose (15. 20). An additional problem in tin determination by HG AAS reported by B. Welz et al. (20) is the appearance of pre-peaks originated from the silica of quartz tube atomizer which can be volatilized and atomized in the presence of hvdrogen. most probably via hydrogen radicals. These prepeaks are difficult to separate from the analvtical signal and mav cause errors in signal evaluation. The present work involved an extensive studv of optimal analvtical and instrumental parameters for low level tin determination in steels and nickel alloys using GF AAS and HG AAS. The determination has been discussed regarding: GF AAS - influence of graphite tube coatings and modifier used on sen-sitivity and reproducibility of signal - interferences of matrix elements - seleetion of optimal pyrolysis and atomization temperature with regard to volatilization of analyte, background, interferences, sensitivity of signal and lile time of graphite tube - evaluation of results HG AAS - influence of acid concentration on analyte signal - interferences of matrix elements and some metalloids - evaluation of results Experimental Aparatus The GBC 902 atomic absorption spectrometer, equipped with deuterium-arc background correction system. automated graphite furnace GF 2000, programmable auto-sampler PAL 2000 and CL 2000 controller vvas used for the measurements of anafvte absorbances using GF AAS. The instrumental parameters and operating conditions are given in Table 1. The furnace program is shovvn in Table 2. A Perkin-Elmer 2380 atomic absorption spectrometer. equipped vvith hvdride generator MHS-10 and printer PRS-10 vvas used for hvdride generation and absorbances measurements using HG AAS. The instrumental parameters and operating conditions are listed in Table 3. Table 1:Instrumental parameters and operating conditions for GF-AAS Table 2: Graphite furnace temperature program for the determination of tin in steels and nickel allovs Spectrometer Wavelength Slit Light source Measurement mode Furnace Graphite tube Char temperature Atomization temp. Sampler Sample volume Standard preparation Stock solution Standard solutions Sample preparation Dissolved in Mass/volume GBC, double beam, 902 286.3 nm 1.0 nm HCL. 10 niA peak height coated vvith Na,W04 800°C 2600 C 20 nI 1000 p,g/ml Sn in 1 M HCI serial dilutions vv ith 0.3 M HNO, 20 ml aqua regia 0.5 g/50 - I to 10/100 ml (diluted vvith 0.3 M HNO,) Step Temp. Ramp time Hold time Ar flovv number (°C) (s) (s) (1 min 1) 1 90 1 9 1.3 2 120 10 10 1.3 3 80 10 10 1.3 4 800 1 1 - 5 2600 1 3 - 6 2650 1 6 1.3 7 20 1 5 1.3 Table 3: Instrumental parameters and operating conditions for HG-AAS Spectrometer Wavelength Slit Light source Hvdride svstern Stock solution Standard solutions Carrier solution Calibration volume Reduetant Flame Sample Dissolved in Mass/volume Measuring volume Elimination of interferences Perkin-Elmer, 2380 286.3 nm 0.7 nm F.DL. 6 W Perkin-Elmer, MHS-10 1000 p.g/ml Sn (in 1 M HCI) serial dilutions vvith 0.1 M HCI 11 BO . sat./O.l M HNO, 25 ml 3 g NaBH4 + 0.5 g NaOH/lOO ml air/acetv lene: blue 20 ml aqua regia 0.5 g/50 mf 0.1-1.0 ml 3 g sodium oxalate/10() ml Reagents Ali reagents vvere of highest available puritv (p.a. or puriss. p.a.) obtained from Merck or Fluka. The solutions prepared vvere: G F AAS - aqua regia - nitric acid. 0.3 M - Pd/Mg nitrate modifier: 300 mg Pd (dissolved in nitric acid) +200 mg Mg(NOi), . 6H O in 100 ml of vvater - sodium tungstate dihvdrate. 5 g in 100 ml of vvater H G AAS - aqua regia - carrier solution: saturated boric acid containing 0.1 M nitric acid - reduetion solution: 3 g of sodium tetrahvdroborate (Fluka) in 100 ml of vvater stabilised vvith 0.5 g of sodium hydroxide - sodium oxalate, 3 g in 100 ml of vvater Standard solutions Stock solution of 1000 p.g ml 1 Sn vvas prepared by dissolv-ing of 1.000 g of tin metal in 100 ml hvdroehlorie acid (1.16) and diluting tii 1 1 vvith deionized vv ater. The other standard solutions vvere prepared from stock solution by serial dilution vvith 0.3 M nitric acid for GF AAS or vvith 0.1 M hvdroehlorie acid for HG AAS. Standard Sn solutions containing the interfering ions vvere prepared by adding the appropriate amounts of interfering ions to the standard solutions. Sample preparation 0.5 g of sample vvas carefullv dissolved in 20 ml of aqua regia (2 hours at 90"C). After eooling the digest vvas diluted to 50 ml vvith deionized vvater. Further dilution of sample solution (10-10(1 times) vvith 0.3 M nitric acid vv as used for GF AAS measurements. 0.800 x= 286,3 nm , TQtorTI = 2600 °C □ Na2W04 coated tube • Na2W04 coated tube * Pd/Mg modifier o TPG tube • TPG tube * Pd/Mg modifier 200 400 600 800 1000 1200 Pyrolysis temperature t °C D 1400 Figure 1: Effect of graphite tube coating and matrix modifier on signal for 2 ng Sn at different pvrolvsis temperatures Slika I: Vpliv prevleke grafitnih cevk in modifikatorja na signal za 2 ng Sn pri različnih razkrojnih temperaturah d) O c CJ _a i— o < 0,700 0,600 0,500 0,400 0,300 0,200 0,100 0,000 1 - 2600°C - 2 - 2300°C - 1 1 3 - 2000°C 2 A \ 3 _ te I *