Thermally Stabilized Iridium on an Integrated Carbide-coated Platform as a Permanent Modifier for Hydride-forming Elements in Electrothermal Atomic Absorption Spectrometry Part 3.* Effect of L-Cysteine Journal of Analytical Atomic Spectrometry DIMITER L. TSALEV YALESSANDRO D'ULI'JO AND LEONARD0 LAMPUGNANI CNR Istituto di Chimica Analitica Strumentale Via Risorgimento 35 561 00 Pisa Italy MARCO D I MARCO AND ROBERTO ZAMBO'VI Universitci di Pisa Dipartimento di Chimica e Chimica Industriale Via Risorgimento 35 561 00 Pisa Italy The effect of L-cysteine (0.2-2% m/v) sodium tetrahydroborate (0.05-0.4% m/v) and hydrochloric acid concentrations (0.01-0.15 mol I-') in an automated flow injection hydride generation system with in situ collection of hydrides of Sb As monomethylarsonate (MMA) dimethylarsinate (DMA) Bi and Sn in an electrothermal atomizer with an integrated platform pre-treated with 110 pg of Zr and 2 pg of Ir for permanent modification was studied.The hydride generation step was optimized by means of a full factorial design of the 33 type and the corresponding regression equations response surfaces and contour plots were derived and are discussed with a view to the improved chemical yields and possible unified conditions for the simultaneous hydride generation/trapping of the three As species or all six analytes. The best characteristic masses for integrated absorbance measurements are 83 35 31 32 110 and 104 pg for Sb As MMA DMA Bi and Sn respectively. The importance of the molar supply rates of the reagents (in pmol s-l) and their ratios rather than reagent concentralions as well as the critical role of sample acidity are emphasized.Keywords Hydride generation and trapping; electrothermal atomic absorption spectrometry; iridium and zirconium coliting; in situ enrichment; L-cysteine; antimony; arsenic; monomethylarsonate; d imethy lars inate ; bismuth; tin In previous parts of this series,'.2 we have proposed and evaluated a permanent Ir modifier stabilized on a Zr- or W-treated platform which has proved efficient for the stabiliz- ation of numerous volatile analytes (As Bi Cd Pb Sb Se Sn Te T1) in direct ETAAS' and more promisingly for in situ trapping/enrichment of volatile hydride-forming elements (HFES)~?~ and some organoelement species of As Se arid Sn with known biological and environmental occurrence and importance3 in automated FI-HG-ETAAS.2 The aim of this study was to evaluate the performance characteristics of this system in the presence of ~-cysteine,~ which appears to be one of the most promising reagents in HGAAS,3 being both an efficient pre-reductanl for As"' and Sb"' and a suitable reaction medium for several HFEs.* For Parts 1 and 2 of this series see refs. 1 and 2. t On leave from Faculty of Chemistry University of Sofia. Sofia 1 126 Bulgaria. In a series of papers by Brindle and co-worker~~-'~ on HG-MIP-AES4-8*'0-'5 and HGAAS,9 ~-cysteine~-'~ and a similar reagent ~-cystine,~.'~-'~ were introduced and analyti- cally evaluated. Further studies by several other research groups on the effects and applications of ~ - c y s t e i n e ~ * ' ~ - ~ ~ and ~-cystine~' in HGAAS,'6,'7,22 FI-HGAAS,19-21*24*25 FI-HG- ETAAS2,18 and HG-non-dispersive atomic fluorescence spectrometry ( HG-NDAFS)23 have been documented in recent years.Several analytes have been studied AS,2.5-9,14,18-21,24,25 monomethylarsonate ( MMA)24 and sn9. 15-19.22 and organotins.22 L-Cystine appears to be less convenient and efficient than L- cysteine in practical work because of its lower solubility in water; therefore higher concentrations of reagent and acid are required e.g. 3% m/v L-cystine in 5 moll-' HCl14 or 4 moll-' HCl,25 or alternatively 0.4 g of solid reagent is added to 5 ml of sample solution in a batch system;" excessive foaming is also entailed. L-Cysteine is compatible with modern continuous flow ( ~ ~ ) 7 - 9 1 6 .1 7 . 2 3 and FI ~ y s t e r n s ~ ~ ' ~ ~ ' ' - ~ ~ ~ ~ ~ and at lower con- centrations of reagent (0.5-2%) and acid (generally below 1 moll-'). Several advantages of this reagent over other pre- reductants and HG media3 have been found (i) faster and more efficient pre-reduction of AsV and SbV to their trivalent states even at lower acid concentrati~ns;~-'~~'~~~~ thus success- ful on-line pre-reduction of AsV has been a t t e m ~ t e d ; ~ ~ ' (ii) sta- bilization of analyte solutions on storage (e.g. for at least 1 (iii) better tolerance to interferen~es;~-'~"~-''~~~ (iv) sensitivity impr~vement;~-~~~'~~'~~'~ (v) identical responses from As"' AsV MMA and DMA;24 (vi) lower blanks; (vii) mild conditions for HG resulting in smaller amounts of H2 being produced' and aerosol being formed,23 and thus less disturb- ance with certain techniques such as HG-DCP-AES,4-8i'0-15 HG-NDAFS23 and HG-ETAAS.2 dimethylarsinate ( DMA),24 Sb,239,10719*20323 Ge,4*1 '-13 EXPERIMENTAL Apparatus The automated FI-HG-ETAAS system based on a Perkin- Elmer (Uberlingen Germany) Model 4100 ZL atomic absorp- tion spectrometer and a transverse-heated graphite atomizer (THGA) longitudinal Zeeman-effect background correction and an AS-7 1 autosampler,26 interfaced with a Perkin-Elmer Journal of Analytical Atomic Spectrometry October 1996 Vol.11 (989-995) 989FIAS-400 FI system has been described el~ewhere.'.~.~~ The standard THGA graphite tubes with integrated platforms (Part No. B 300-0643) were pre-treated with 1.2 pmol of Zr (1 10 pg) and then with 10.4 nmol of Ir (2 pg) as detailed in ref.1. The system was operated under the Software Version 7.21 (Part No. B0509524) using the optimized parameters and tempera- ture programmes for the FIAS-400 manifold and the THGA electrothermal atomizer previously given in refs. 1 and 2. Reagents The preparation of standard solutions and other reagents has been detailed elsewhere.'T2 Working solutions of reductant (0.05-0.400/ m/v NaBH,) were always made alkaline with 0.05% m/v NaOH (i.e. supply rates of 0.833 pmol s-' of OH-). A small amount of HC1 (50mmol 1-l) was added to the stock solution of L-cysteine (10% m/v) in order to facilitate dissolution of the reagent; this contributed to small supply rates of H+ due to this reagent solution varying from 0.1 to 1 pmol s-'.Experimental Design for Studying Hydride Generation and Trapping The effect of the concentrations of reagents viz HCl L-cysteine and NaBH was studied over the following ranges 0.01-0.15 moll-' HC1 0.2-2% m/v L-cysteine and 0.05-0.4% m/v NaBH by means of a full factorial design of the 33 type (see Table 1) according to the plan suggested by the software programme of the Statistical Graphics System Version 6 (Manugistics Rockville MD USA). Randomized experiments ( n = 29) with three replicate measurements were performed the highest mean value of the integrated absorbance for each analyte species being assigned a value of 100% response (Y%). The analyte level was fixed at 5 ng for all species. RESULTS AND DISCUSSION The effect of the concentrations of the three reagents in this chemical system viz. HCl (mol 1 - ') L-cysteine pre-reductant (YO m/v) and NaBH reductant (YO m/v) on the response for each analyte species (Y%) can be described by a regression Table 1 Factors and levels for a 33 full factorial design Level Factor -1 0 + 1 HCl/mol 1- * 0.01 0.08 0.15 NaBH (YO m/v) 0.05 0.225 0.40 L-Cysteine ( O h m/v) 0.2 1.05 2.0 equation of the following type Y% = k o + k [HCl] + k [NaBH,] + k [~-cysteine] + k [ HCl] [NaBH,] + k [ HCl] [~-cysteine] + k [ NaBH,] [~-cysteine] + k7[ HC1I2 + k8 [NaBH,I2 + k9 [~-cysteine]~ In Table 2 are given the coefficients (ko-k,) for each analyte in the above equation as derived from the 33 full factorial design.All experimental values are fitted satisfactorily by these four- dimensional (4D) response surfaces.The correspondence between the concentrations and supply rates (pmol s-') of the reagents as well their molar ratios can be evaluated from the diagram in Fig. 1. It seems that the most convenient and practically useful mode of graphic representation of these 4D surfaces would be as a series of contour plots i.e. the cross-sections of response surfaces at three selected levels of L-cysteine (a) 'low' 0.2% m/v or 1.76 pmol s-' (b) 'the centre of the experimental plan' 1.05% m/v or 9.24 pmol s - l and (c) 'high' 2% m/v or 17.6 pmol s-'. These diagrams are plotted for inorganic As (i-As"') MMA DMA Sb"' Bi"' and Sn" in Figs. 2-7 respectively while the contours at Y>90% are overlayed for all species in Fig. 8 thus showing the optimum conditions for simultaneous HG and determination.Replotting these diagrams in coordi- nates representing the actual molar concentrations of reagents HCI/ mol I-' - 0 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0 2 4 6 8 10 12 14 16 18 HCI/ pmoi s-' 0.5 0.4 3 E s 0.3 . I" m 0.2 2 O.? 0 1 Fig. 1 Calculated molar ratios of reagents within the field of exper- imental parameters. Solid line [H+]/[BH,-]; dashed line [L- cysteine]/[BH,-1; dotted line [~-cysteine]/[H+ 1. L-Cysteine concen- tration fixed at 1.05% m/v (9.24 pmol s-l) Table 2 m/v)] derived from the 33 full factorial design Model fitting results for the regression equations Y% uersus reagent concentrations [HCl/mol l - l NaBH (Yo m/v) and L-cysteine (Yo Analyte species As"' 89.8 - 9.43 - 6.40 0.109 0.113 0.0530 0.241 1.46 2.66 - 0.704 MMA 27.5 - 7.03 - 0.606 62.4 0.0806 1.39 0.361 0.385 - 2.64 - 19.7 DMA 71.5 -4.05 - 2.95 -22.6 0.195 3.23 0.145 0.628 - 2.17 3.71 BiI" 68.7 1.64 8.55 2.30 0.0187 - 1.13 - 0.806 - 0.047 - 0.696 -0.815 Sb"' 81.9 0.763 0.372 1.48 0.0123 0.0862 -0.124 - 0.038 0.103 -0.500 Sn" 91.0 0.760 3.64 0.141 0.700 - 19.9 -0.332 - 0.07 1 - 0.565 3.33 990 Journal of Analytical Atomic Spectrometry October 1996 Vol.11a HCI/ pmol s-l Contour plots for As"' normalized integrated absorba ace signal) cersus supply rates of acid and reductant at three levels of L-cysteine (a)j0.2; ( h ) 1.05; and (c) 2% m p - 0 2 4 6 8 10 12 14 16 (b) MMA I . 1 ' 1 . I ' 1 ' I ' 1 2 4 6 8 10 12 14 16 HCI/ pmol s-l Fig. 3 (a) 0.2; (b) 1.05; and (c) 2% m/v Contour plots for MMA (normalized integrated absorklance signal) versus supply rates of acid and reductant at three levels of 1.-cysteine 0 0 2 4 6 8 10 12 14 16 HCI/ pmol s-' Fig.4 Contour plots for DMA (normalized integrated absorbmce signal) uersus supply rates of acid and reductant at three levels of L-cysteine (a) 0.2; (h) 1.05; and (c) 2% m/v 2 0 2 4 6 8 10 12 14 16 HCI/ pmol s-' 0 93 9 1 d 86 I I ' 0 2 4 6 8 1 0 1 2 1 4 ' S Fig. 5 (a) 0.2; (b) 1.05; and (c) 2% mjv Contour plots for Sb"' (normalized integrated absorbance signal) versus supply rates of acid and reductant at threc levels of L-cysteine .I'ournul of Analytical Atomic Spectrometry October 1996 Vol. 11 991'na 0 . 1 . i ' - I ' 1 . I ' I ' - I 0 2 4 6 8 10 12 14 16 4w9* HCI/ pmol s-l Fig. 6 Contour plots for Bi"' (normalized integrated absorbance signal) cersus supply rates of acid and reductant at three levels of L-cysteine (a) 0.2; (b) 1.05; and (c) 2% m/v 6 8 10 12 14 16 HCI/ pmol s-' Fig.7 Contour plots for Sn" (normalized integrated absorbance signal) cersus supply rates of acid and reductant at three levels of L-cysteine (a) 0.2; (b) 1.05; and (c) 2% m/v 0 2 4 6 0 10 12 14 16 0 0 2 4 5 8 10 12 14 16 HCI/ pmol s-' I 0 2 4 6 8 10 12 14 16 4 Fig. 8 Overlay of the contour plots at 90% response (normalized integrated absorbance signal) for all analytes versus supply rates of acid and reductant at three levels of L-cysteine (a) 0.2; (b) 1.05; and (c) 2% m/v. The species symbol is placed on the side of >go% response before mixing uiz. H + (mol 1-') uersus NaBH (mol 1-') gives similar pictures to those in Figs.2-8 (not shown). Several important effects are observed in the presence of L- cysteine compared with the same system in the absence of this reagent:2 (i) improved sensitivity in FI-HG-ETAAS especially for i-As"' MMA DMA and Sn"/'" by 1.2- 2 . 5 1.8- and 1.5-fold respectively; thus the best characteristic masses (mo) for inte- grated absorbance measurements are decreased to 35 31 32 83 110 and 104 pg for i-As"' MMA DMA Sb'" Bi"' and Sn"/" respectively which agree well with the corresponding figures for the 'injected' analytes in direct ETAAS uiz. 31 30 26 92 176 and 156 pg;' (ii) substantial change in the HG pattern due to L-cysteine especially for i-As"' (shift towards much lower H + concen- trations) and Sn''/'" (signal decreases below certain H+ levels < 0.005-0.025 moll-' depending on L-cysteine concentration); (iii) strong effect of acid concentration on HG being most marked for i-As"' MMA and DMA (Figs.2-4); (iv) apparent importance of the molar supply rates and ratios of reagents uiz. HCl NaBH and L-cysteine; (v) similar behaviour of all three As species which makes possible their simultaneous determination under unified con- ditions [Fig. 8(b) and (c)]; (vi) high relative responses for Bi"'( 2 68% Fig. 6 ) Sn'1(2710/~ Fig. 7) and Sb"' (>84% Fig. 5 ) within the whole field of chemical parameters thus offering several sets of experimental conditions for the simultaneous determination of all four elements with only a minor sacrifice of sensitivity (Fig.8 ) . 992 Journal of Analytical Atomic Spectrometry October 1996 Vul. 11While all these effects are analytically important and practi- cally useful they are much easier to observe than to explain; in particular that the FI system is apparently under kinetic rather than thermodynamic control as observed also with other FI-HGAAS systems2 but not with batch HGAAS wherein a longer time is available for the generation and stripping of hydrides from the reaction medium to thr gase- ous phase.3 Expected Behaviour of the Chemical System The chemistry of the system L-cysteine-BH,-H +-analyte is complex because L-cysteine may act simultaneously in a diverse manner being (i) a pre-reductant (ii) an acid-base buffer (iii) a complexing agent for tetrahydroborate and (iv) 21; com- plexing agent for analytes. L-Cysteine as a pre-reductant L-Cysteine [ 2-amino-3-mercaptopropionic acid (HS- CH2- CHNH2-COOH) denoted as HSR] behaves as a reductant in aqueous solutions being oxidized to the disulfide form L-cysteine [ 3,3'-dithio-bis( 2-aminopropionic acid) HOOC- CHNH2-CH2-S-S-CH,-CHNH2-COOH denoted as RSSR] 2HSR -+ RSSR + 2H + + 2e L-Cysteine by itself may also further act as a reductant. The standard oxidation potential of the cystine-cysteine system is 0.074V (on the basis of polarographic data within pH 3.3-6.6).28*29 The formal potential is slightly increased at lower pH values.Thus L-cysteine is able to reduce As" and SbV to their trivalent states MMA and DMA to their trivalent thi~lates,~ Sn" to Sn" Se" to Se(0) and Te" to Te(0).The resulting solutions exhibit improved stability on storage for As"' Sb"' and Sn1',5,19,20 which could be due to complex formation between these ions and the HSR reagent. In contrast Se and Te are unstable and cannot be determined in solutions containing ~-cysteine;~' moreover the interference by Se on Ge and Sb may become stronger in the presence of this pre- r e d ~ c t a n t ' ~ ~ ~ ~ presumably due to coprecipitation of these analytes with the elemental form of the Se or Te interferent. The reduction of Bi"' to its elemental state is b rde~line;~' hence it would be worth studying the stability of L-cysteine solutions containing this element. L-Cysteine as an acid-base bufSer The acid-base equilibria for L-cysteine are also complex,32 as can be seen in the distribution diagram for this reagent (Fig. 9).Without acidification the neutral form of the reagent HSR 0 2 4 6 8 10 12 '14 PH Fig. 9 Distribution diagram for L-cysteine mole fraction (a) for different forms of reagent uersus pH behaves as a very weak acid with ~ K ~ ~ z t 3 . 2 - 8 . 4 and pKa3 NN 10.0-10.8,32-35 being in effect a dipolar ion (zwitterion HS-CH2-CH-NH3+-COO-).32 At lower pH values (<4-5) the protonated form of the reagent HSRH+ with a much higher acid constant uiz. pK z 2 ( 1.88 +0.02;35 1.90;34 1.9632) is gradually formed on acidification. The ratio of HSRH' HSR must be rapidly increasing with decreasing pH below 3 being approximately lo-' 1 10 and lo2 at pH 3 2 1 and 0 respectively (Fig. 9). Thus L-cysteine also acts as an acid-base buffer on acidification(!) just within the analytically useful range of pH 2 f 1 and its buffering capacity must depend on the concentration of reagent and the molar ratios of L- cysteine HC1 as well as on the initial acidification (or not) of the stock solution of reagent.Indeed pH measurements show that the actual pH of a 1% m/v L-cysteine solution in water is 5.0 which then decreases on acidification to pH 1.69 in equimolar HCl medium (0.0825 moll-' of the added H f ) ; this is in agreement with the distribution diagram in Fig. 9. On the other hand 0.01 mol 1-1 solutions of HCl exhibit actual pH values of 2.0 2.4 2.9 and 3.1 in the presence of 0 0.2 1.05 and 2% m/v L-cysteine respectively i.e. pH values are higher than expected on the basis of only formal concentrations of the added HC1.Further complication of the acid-base equilibria also takes place after merging the reagents in that BH,- and OH- also act as bases and are able to neutralize part of the H + in the resulting solution instantaneously. Concurrent reac- tions of analyte reduction and tetrahydroborate hydrolysis/ decomposition are obviously taking place over very short periods of time. Hence even the mode by which the stock solutions of NaBH are made alkaline36 (or not37) will also affect the HG processes. L-Cysteine as a complexing agent for tetrahydroborate Brindle and Le12 explained the higher efficiency of tetrahydro- borate reductant in the presence of L-cysteine by the possible formation of more efficient intermediate species of the reductant BH4- +HSR-+H3B-SR- +H2 This scheme has been supported by 13B NMR spectra for L- cysteine and analogous reagents uiz.thioglycerol and penicilla- mine but not for non-thiol-containing compounds such as S- methyl-L-cysteine histidine and glycine.12 The scheme is not in conflict with that for the action of BH4- in acidic media given by Mal'tseva and K h a i ~ ~ ~ * wherein the formation of more reactive intermediate species of the reductant BH has been supposed. Hydride generation by the H3B-SR- form of the reductant can be expected to proceed faster and with higher chemical yields than HG by the protonated hydroborate BH but obviously takes place at much lower acid concentrations. Some possible explanations for this effect of H + are (i) either the H,B-SR- species cannot be formed at low pH when L-cysteine is already protonated and/or they are rapidly decomposed at low pH or (ii) higher H + levels are more detrimental in the presence of L-cysteine because of their more pronounced effect on that form of the reductant which is a weaker acid i.e.H3B-SR-Hf rather than H+BH4-. L-Cysteine as a complexing agent for the analyte Yet another possible effect of L-cysteine uiz. complex formation with the analyte ions cannot be excluded from c~nsideration;~~ moreover it may play a decisive role in the HG processes. If such complex formation is taking place it may either impair HG by 'masking' the analyte ions or improve reduction processes owing to higher chemical yields or better kinetics or both. A 'masking' effect on the analytes does not seem very probable because no signal depression is observed in the Journal of Analytical Atomic Spectrometry October 1996 Vol.1 1 993presence of L-cysteine; if still taking place it must be more pronounced for those elements that are complexed more strongly by the HS-reagents (Bi"' Sn") and at excessive amounts of L-cysteine and/or at lower acidity. In fact such a slight impairment of relative responses is observed at 2% m/v levels of L-cysteine and low H + concentrations for Bi [Fig. 6(c)] and Sn [Fig. 7(c)]. Although there are published data on the stability constants of L-cysteine with Cu" Hg" Mn'l Nil' Pb" and Zn" (within 109-1020),34,35 no data could be found on complex formation with AS'" Sb"' Sn" and Bi"'. Nevertheless these ions could be expected to be complexed to some extent by the thiol group of the reagent presumably in an increasing order in the series As"' < Sb"' < Sn" <Bi'" by analogy with the corresponding solubility products of their sulfides.As"' (a 'hard acid') should be the least inclined ion to form complexes with the HSR reagent. Hence As should be the most tolerant of these four analytes to low acid and high L-cysteine levels (see the upper left corner of the diagrams in Fig. 8). On the other hand the complexation of As species with L-cysteine should be more vulnerable to higher H + levels. This is in fact observed experimentally. Several ~ ~ r k e r ~ ~ ~ ~ ~ - ~ ~ have assumed that the formation of complexes (thiolates) between As species and reagents containing the thiol (-SH) group is taking place; eg.in the presence of mercaptoacetic a ~ i d ~ " ' ~ or its methyl ester,41 ~-cysteine~ and thi~glycerol,~~ while no positive effects have been observed with other sulfur-containing additives such as histidine12 or m e t h i ~ n i n e ~ ~ which do not contain an HS group. Haraguchi and Takatsu4' derivatized MMA and DMA to their corresponding volatile products CH,As- (SCH,COOCH,) and (CH,),AsSCH,COOCH respectively and proved by GC-MIP-AES that both As and S were present in the eluates with identical retention times i.e. As and S are present in the same molecule. Le et aL2 have proposed that the reduction of AsV MMA and DMA to their trivalent states and complex formation are taking place in the presence of L- cysteine As( SR) CH,As(SR) and (CH,),AsSR respectively and that the rate of HG from these species is faster than from hydrolysed As in aqueous solutions without added L-cysteine.If these logical assumptions by Le et aL2 are adopted then one might expect that HG will be favoured under conditions where complex formation between analyte ions and L-cysteine is promoted i.e. at lower acid levels. Experimentally HG in the presence of L-cysteine always takes place at lower acidities than in the absence of this reagent. The effect of the concen- trations of reagents and their molar ratios is particularly critical for As (Figs. 2-4). At low concentrations (deficiency) of L-cysteine [Figs. 2(a) 3(a) and 4(u)] marked differences between As"' MMA and DMA are observed. The response Y% generally decreases in the order As"' > DMA > MMA i.e.the stronger the acid the lower the efficiency of HG cf pKal =9.2 6.2 and 4.633,42 for H3As03 (CH,),AsOOH and CH3AsO(OH) respectively. This is the opposite order to that observed in the absence of L-cysteine wherein hydrides from more acidic As species viz. H3As04 and MMA are generated in more acidic media.,,,' In the absence of L-cysteine the distribution of the neutral and charged As species within this particular pH range3I may also account for the above effects log [HAs02]/[As0+] =0.34+pH According to this equation the neutral species of the analyte should prevail at higher pH values e.g. the [HAsO,]/[AsO+] ratio should be about 2.2 22 221 and 2188 at pH 0 1 2 and 3 respectively. Indeed the stepwise attack of the neutral molecule of hydrated arsenic (111) oxide by the reductant has been assumed t o be part of the reaction mechanism of HG in the absence of ~-cysteine.~* DMA is also protonated at low pH values to form a cation (CH3)2As(OH)2+ with pK,,= 1.28 k0.05 as shown by ion-exchange studies by Hansen et al.43 A detailed study of these complex effects is obviously required in order to utilize the potential of L-cysteine completely but is outside the scope of this series of papers.CONCLUSIONS Analytically useful effects on HG are observed in the presence of L-cysteine dilute acid and tetrahydroborate viz. sensitivity improvement due to better yields and kinetics of HG; unified conditions for species-independent determination of the sum of i-As"' i-AsV MMA and DMA; appropriate conditions for the simultaneous HG of arsine stibine bismuthine and stan- nane etc.The generation of arsine and methylarsines takes place within a fairly narrow range of acid concentration and at molar excess of L-cysteine and tetrahydroborate. Although working at lower acid levels is advantageous for lower blanks and reduced reagent consumption the pH adjustment of real sample digests becomes more critical in the presence of L- cysteine. The molar ratios of reagents and their supply rates are important because these chemical parameters are closely inter- related and the molar concentrations after mixing the reagents may critically affect the distribution of neutral charged and complexed species of the analyte. Among the factors contribu- ting to the total balance of reagents could also be 'details' such as the mode of stabilization of stock solutions acidification (:or not) of the L-cysteine and basification of the NaBH solutions.The actual pH obtained after mixing the reagents could be very different from the expected value because of the buffering properties of L-cysteine in the pH range 1-3 and the instantaneous neutralization of HC1 and NaOH. Financial support to D. L. T. by the Consiglio Nazionale delle Ricerche Istituto di Chimica Analitica Strumentale Pisa is gratefully acknowledged. This work was supported in part by the Ministry of University Scientific Research and Technology (Progetto Speciale Sistema Lagunare Veneziano) Italy. One of the authors (R. Z.) thanks the CNR National Committee for the Environment and Habitat Technologies Italy for financial support.REFERENCES 1 2 3 4 5 6 7 8 9 10 11 12 13 84 15 16 17 994 Journal of Analytical Atomic Spectrometry October 1996 Vol. Tsalev D. L. D'Ulivo A. Lampugnani L. Di Marco M. and Zamboni R. J. Anal. At. Spectrom. 1995 10 1003. Tsalev D. L. DUlivo A. Lampugnani L. Di Marco M. and Zamboni R. J. Anal. At. Spectrom. 1996 11 979. Dgdina J. and Tsalev D. L. Hydride Generation Atomic Absorption Spectrometry Wiley Chichester 1995. Brindle I. D. and Ceccarelli Ponzoni C. M. Analyst 1987 112 1547. Brindle I. D. and Le X.-c. Anal. Chem. 1989 61 1175. Brindle I. D. and Chen H.-w. Talanta 1991 38 1137. Brindle I. D. Alarabi H. Karshman S. Le X.-c. Zheng S.-g. and Chen H.-w. Analyst 1992 117 407. Chen H.-w. Brindle I. D. and Le X.-c.Anal. Chem. 1992,64,667. Le X.-c. Cullen R. Reimer K. J. and Brindle I. D. Anal. Chim. Acta 1992 258 307. Chen H.-w. Brindle I. D. and Zheng S.-g. Analyst 1992 117 1603. Brindle I. D. Brindle M. E. Le X.-c. and Chen H. J. Anal. At. Spectrom. 1991 6 129. Brindle I. D. and Le X.-c. Anal. Chim. Acta 1990 229 239. Brindle I. D. Le X.-c. and Li X.-f. J. Anal. At. Spectrom. 1989 4 227. Boampong C. Brindle I. D. Le X.-c. Pidwerbesky L. and Ceccarelli Ponzoni C. M. Anal. Chem. 1988 60 1185. Brindle I. D. and Le X.-c. Analyst 1988 113 1377. Beach C. and Shrader D. in 7th Annual Spring Optical Spectroscopy Workshop Varian Sugar Land TX 199 1 Shrader D. and Beach C. Presented at the X X V I I Colloquium Spectroscopicum Internationale June 9-1 4 1991 Bergen Norway Abstract D-4.3.pp. 11 1-1 19. 1118 19 20 21 22 23 24 25 26 27 28 29 30 31 32 An Y. Willie S. N. and Sturgeon R. E. Spectrochim. Actil Part B 1992 47 1403. Welz B. and Sucmanova M. Analyst 1993 118 1417. Welz B. and Sucmanova M. Analyst 1993 118 1425. Welz B. He Y. and Sperling M. Talanta 1993 40 1917 Martin F. M. and Donard 0. F. X. J. Anal. At. Spectrom.. 1994 9 1143. D'Ulivo A. Lampugnani L. Pellegrini G. and Zambclni R. J . Anal. At. Spectrom. 1995 10 969. Le X.-C. Cullen W. R. and Reimer K. J. Anal. Chim. Acta 1994 285 277. Tesfalidet S. and Irgum K. Anal. Chem. 1989 61 2079. The THGA Graphite Furnace Techniques and Recomrvended Conditions Publication B3210.20 Part No. B05(!-5538 Bodenseewerk Perkin-Elmer Uberlingen 1992. The FIAS-Furnace Technique Recommended Analytical Conllitions and General Information Publication B3212.20 Part No. 0993-5204 Bodenseewerk Perkin-Elmer CJberlingen 1993.Kolthoff I. M. Stricks W. and Tanaka N. J. Am. Chem Soc. 1955 77 4739. Encyclopedia of Electrochemistry of the Elements. Organic Section Vol. X I I Carbonyl Compounds Carboxylic Acids Esters and Anhydrides Organic Sulfur Compounds eds. Bard A. J and Lund H. Marcel Dekker New York 1978 p. 409. Romboli L. Thesis University of Pisa 1995. Pourbaix M. Atlas d 'Equilibres Electrochimique Grau thier- Villars Paris 1963. Butler J. N. Ionic Equilibrium. A Mathematical Approach Addison-Wesley Publishing Co. Reading MA 1964 p. 23 :1. 33 34 35 36 37 38 39 40 41 42 43 CRC Handbook of Chemistry and Physics ed. Weast R. C. CRC Press Boca Raton FL 67th edn. 1986-1987. Perrin D. D. Stability Constants of Metal-Ion Complexes Part B-Organic Ligands Pergamon Oxford 1979 pp. 152-153. Martell A. E. and Smith R. M. Critical Stability Constants Plenum New York 1974 vol. 1 p. 47. Qiu D. R. Vandecasteele C. Vermeiren K. and Dams R. Spectrochim. Acta Part B 1990 45 439. Mandjukov P. B. Djarkova V. and Tsalev D. L. in 6. Colloquium Atomspektrometrische Spurenanalytik ed. Welz B. Bodenseewerk Perkin-Elmer Uberlingen 1991 pp. 597-608. Mal'tseva N. N. and Khain B. S. Borogidrid Natriya [Sodium Tetrahydroborate in Russian] Nauka Moscow 1985. Anderson R. K. Thompson M. and Culbard E. Analyst 1986 111 1143. Anderson R. K. Thompson M. and Culbard E. Analyst 1986 111 1153. Haraguchi H. and Takatsu A. Spectrochim. Acta Part B 1987 42 235. Sarjeant E. P. and Dempsey B. Ionization Constants of Organic Acids in Aqueous Solution Pergamon Oxford 1979. Hansen S. H. Larsen E. H. Pritzl G. and Cornett C. J . Anal. At. Spectrom. 1992 7 629. Paper 6/04887K Received July 1 1996 Journal of Analytical Atomic Spectrometry October 1996 Vol. 11 995