Determination of Chromium by Electrothermal Atomic Absorption Spectrometry with Various Chemical Modifiers Journal of Analytical Atomic Spectrometry NIKOLAOS S. THOMAIDIS EFROSINI A. PIPERAKI,* CHRISTOFOROS K. POLYDOROU AND CONSTANTIN 0s E. E F STAT H I 0 U Laboratory of Analytical Chemistry Chemistry Department University of Athens University Campus 157 71 Athens Greece The determination of Cr in the presence of various isomorphous metals has been studied. The atomic absorption signal for Cr was increased and stabilized by the presence of 20 pg of Mg(NO& 1 pg Rh and 1 pg Pt. Magnesium Rh and Pt gave comparable characteristic masses of 3.2 3.0 and 2.8 pg respectively when integrated absorbance was measured. The limits of detection were 0.18 0.14 and 0.091 pg l-' respectively. The efficiency of these modifiers was tested with the direct determination of Cr in rain-water and serum samples.Quantification was performed with aqueous standards in the case of the rain-water samples and with matrix- matched standards in the case of the serum samples. Recovery tests and a serum reference material were used to check the accuracy of the proposed methods. Accurate results and good agreement with the certified serum values were found in the presence of platinum as a modifier. Chemical modifiers were not necessary for the determination of Cr in rain water. Keywords Chromium determination; electrothermal atomic absorption spectrometry; chemical modification; rain water; serum Electrothermal atomic absorption spectrometry (ETAAS) is the technique of choice when low amounts of Cr (about 1 pg 1-I) have to be determined.However the determination of Cr is considered to be difficult owing to the following problems (a) background correction with continuum light-sources is difficult owing to the low emission intensity of the deuterium- arc background corrector at the resonance line of Cr (357.9 nm); and (b) pre-atomization losses and decreased sensi- tivity can occur in certain complex matrices. Attempts to solve the first problem led to the use of a modified spectrometer with a quartz-halogen light source for background Extensive s t ~ d i e s ~ . ~ on the back- ground signal showed that the problem arises from the emission caused by Cr together with the main matrix elements (such as K and Na) and probably graphite. The magnitude of this interference depends on the lamp current slit-width and the atomization temperature.Halls and Fell4 and Berndt and Sopczak5 found that using a ramp atomization mode at 2400 "C reduced the interference. Pre-atomization losses and reduced sensitivity are severe in the presence of chlorides of various metal^,^,^ and in the presence of carbide forming elements such as Mo and organic solvents that promote the formation of chromium carbide.' Tungsten has been proposed as a chemical modifier for the determination of Cr in ~ a t e r ~ ' ~ ~ because it enables pyrolysis temperatures of 1500"C9 or 1600°C10 in acidic media to be obtained without losses. Vanadium and a V-Mo modifier have been used for the determination of Cr in serum and lake-water * To whom correspondence should be addressed.samples using peak height measurements." Magnesium nitrate has often been used as a chemical modifier to stabilize Cr during the pre-atomization step8-10y12 and for its determination in water sample^.^^^^^ It has also been used as an ashing aid/ chemical modifier for the determination of Cr in s e r ~ m . ~ ~ ' ~ ~ The Mg-Ca mixed modifier has been used to determine Cr in serum because it has facilitated pyrolysis at higher tempera- tures (1400 versus 1200°C without a chemical modifier) and has given a better characteristic mass (2.56 versus 5.13 pg for integrated absorbance measurements).16 In addition to the above problems very low Cr concen- trations are present in various types of sample even though Cr is a ubiquitous element on the Earth's crust.In order to avoid contamination extreme care is needed and direct methods of determination become necessary. The purpose of the present study was to investigate the direct determination of Cr in various matrices using alternative chemical modifiers (Rh and Pt) to optimize their use and compare them with modifiers that have been proposed pre- viously (Mg and W). The direct determination of Cr in samples with extremely low concentrations such as rain water and serum was investigated and the alternative modifiers proved to be essential especially in the latter type of sample. The choice of potential chemical modifiers was based on the studies of Tsalev and c o - ~ o r k e r s l ~ * ~ ~ on the possible substitution of the modifier atoms by an isomorphous analyte.EXPERIMENTAL Instrumentation A Perkin-Elmer Model 5000 atomic absorption spectrometer equipped with an HGA 400 graphite furnace was used for the atomic absorption measurements at the 357.9 nm resonance line of Cr with a 0.7nm spectral bandwidth. A Cr hollow cathode lamp was employed as the radiation source and was operated at 18 mA. Pyrolytic graphite coated graphite tubes (Perkin-Elmer Part No. B0135653) were used throughout the study. Solution volumes of 20 pl were dispensed into the graphite tubes with an AS-1 autosampler and an Eppendorf micropipette with disposable poly (propylene) tips. A tungsten- halogen light source was used for background correction. The graphite furnace operating conditions are summarized in Table 1.The time-resolved atomic absorption peaks were recorded with an IBM compatible PC Quest 286/16 computer. This system has been described elsewhere." Reagents All chemicals used in this study were of analytical-reagent grade. The glass and poly(propy1ene) apparatus were kept in 10% v/v HN03 for at least one night and then rinsed with 1% v/v HN03 three times and subsequently ten times with Journal of Analytical Atomic Spectrometry January 1996 Vol. 11 (31-36) 31Table 1 Temperature programme for the comparison of chemical modifiers in the determination of Cr Temperature Step 1°C Drying 120 Cooling 20 Pyrolysis Various* Atomization 2300 Cleaning 2650 Ramp time/s 10 10 1 0 1 Hold time/s 20 20 9 4 2 Ar flow ratelm1 min- Read 300 - 300 300 - 0 On 300 - - * Various Tpyr values are presented in Table 3. doubly distilled de-ionized water before use.The acids were of Suprapur grade (Merck). Chromium standards were pre- pared by diluting a 1 g I-' Cr (as CrC1,) stock solution (Titrisol Merck) with doubly distilled de-ionized water and acidified to a final HN03 concentration of 1% v/v. Modifier stock solutions were prepared by dissolving appropriate amounts of their salts in acidic media and diluting to a final volume with doubly distilled de-ionized water. The modifiers studied were Mg Ca Sr Sc Y and La (as the nitrates) Zr (as ZrOCl,) W (as Na,WO,) Re (as HReO prepared by dissolv- ing Re powder in H202) Ru Rh Pd and Pt (as the chlorides). All of these metals are isomorphous with Cr.,' A 1 g I-' NH,SCN solution was prepared from solid NH,SCN (Merck).Procedure Comparison of chemical modiJers For the choice of the maximum pyrolysis temperature (TPyr) and the optimum mass of the modifier 5 or 10 pl of the modifier solution were dispensed into the graphite tube (depending on the mass of the modifier) followed by 20 p1 of the Cr solution (200 pg). The temperature programme in Table 1 was followed. The peak height and the integrated absorbance were recorded simultaneously. Calibration curves were constructed by injecting 20 p1 of standard solutions containing 0.5 1.0 2.0 3.0 5.0 7.5 10.0 15.0 and 20.0 pg 1-' of Cr on to the graphite wall. The characteristic mass m (pg) was calculated from the slope (b) of the calibration curve using the equation m =0.0044 x 20/b for a sample volume of 20 pl.The limit of detection (LOD pg I-') was calculated from the equation LOD = 3 x SB,/b where SBL was the standard deviation of ten blank firings. Determination of Cr in rain water A full description of the type of sampler and methods used for collection has been given in a previous study.21 The samples were acidified with HNO to a final concentration of 1% v/v and stored at 4°C until required for analysis. A 20 p1 volume of the sample was injected into the graphite tube with or without 5 pl of the Pt solution. The temperature programme in Table 1 was followed. The value of Tpyr in the absence of Pt was 1300"C whereas in the presence of 1 pg of Pt it was 1500 "C. Quantification was performed with aqueous solutions containing 0.5 1.0 1.5 2.0 3.0 and 4.0 pg 1-l of Cr and using integrated absorbance measurements.Recovery experiments were carried out by spiking a rain water sample with the same concentrations of Cr. Determination of Cr in serum Serum samples were diluted 1+1 with a 0.2% solution of Triton X-100. A 5 pl volume of the modifier solution (Mg or Pt) was injected into the graphite tube followed by 20p1 of diluted serum. The temperature programme given in Table 2 was followed. The quantification was performed with a matrix- matched calibration curve prepared by spiking a diluted serum 32 Journal of Analytical Atomic Spectrometry January 1996 sample with 0 0.5 1.0 2.0 and 5.0 pg 1-l of Cr. Recoveries were calculated from the matrix-matched curve. The accuracy was tested by analysing a Biological Reference Material (Freeze-Dried Human Serum) with a certified Cr value of 0.76 (0.67-0.87) ng g-' supplied by Dr.J. Viersieck University Hospital Ghent Belgium. The freezed-dried material was taken into solution by simply adding doubly distilled de-ionized water mixing vigorously for 5 min on a Vortex and diluting 1 + 1 with a 0.2% solution of Triton X-100. Various portions of this material were used of different masses giving different Cr concentrations in the final liquid sample. These concentrations are quoted as 'certified values' in Table 7. Serum samples from healthy people (n = 22) and samples from patients with various thoracic diseases (lung cancer tuberculosis and pleurisy n = 15) were analysed using the above procedure with Pt as the chemical modifier taking all necessary precautions to avoid contamination.RESULTS AND DISCUSSION Comparison of Chemical Modifiers Amount of chemical modijier and temperature studies The primary role of a chemical modifier is to stabilize the analyte during the pre-atomization step. Although Cr is a moderately volatile element in oxyacidic media such as the 1% v/v HNO used in the present work with a maximum pyrolysis temperature of 1200"C this temperature has to be decreased when samples with a difficult matrix have to be analysed (e.g. in the presence of high concentrations of chloride salts). Therefore the Tpyr was determined in the presence of various potential chemical modifiers. The results are summar- ized in Table 3 where the optimum amounts of the modifiers are also given. The optimum atomization temperature was found to be 2300"C since the integrated absorbance did not change for higher temperatures and an over-correction of the specific absorbance was seen at temperatures higher than 2500 "C confirming previously described y h e n ~ m e n a .~ . ~ ' ~ The modifiers were tested in sequence according to their position in the Periodical Table and the results are discussed below. Group IIA Mg Ca and Sr. Magnesium was the only element in this group that stabilized Cr. The appearance temperature was shifted to higher values in the presence of 20 pg of Mg(NO,),. As the atomic radius of these metals increased Cr atoms were produced earlier and the maximum pyrolysis temperature decreased. Despite the small increase in sensitivity in the presence of 20 pg of Ca(N03)2 it could not be used as a chemical modifier for the determination of Cr since no stabilization was observed.The influence of increasing amounts of Mg(NO,) on the Cr signal at a Tpyr of 1500°C is shown in Fig. 1. The optimum amount of Mg(N03) was 20 pg because larger amounts produced peaks with tailing and decreased the sensitivity. It is apparent that when peak height absorbance was measured the atomic absorption (AA) signal recovered only in the presence of 20 pg of Mg(NO,) whereas the integrated absorbance signal recovered with only 5 pg of VOl. 11Table 2 Temperature programme for the determination of Cr in serum Temperature Step 1°C Drying Pyrolysis 1 Pyrolysis 2 Cooling Atomization Cleaning 120 600 Various* 20 2200 2650 Ramp time/s 10 5 10 1 0 1 Hold time/s 20 15 20 14 4 2 Ar flow rate/ml min- 300 300 300 300 0 300 1 Read * Various qyr are no modifier 1100 "C; Pt 1300 "C; and Mg 1400 "C.Table 3 Maximum pyrolysis temperature (T,,,) for the determination of 0.1 ng of Cr in the presence of various modifiers of different masses Modifier mass/pg - 20 20 20 20 20 20 20 20 20 10 1 1 5 1 +20 1 +20 Tpyr/OC 1200 1500 1200 1150 1300 1300 1200 1200 1200 1200 1200 1400 1300 1250 1500 1400 * Plus 20 pg of NH,SCN. 0.8 I I I 0.6 0 \o. -- L O I 0 20 40 60 80 100 Mass of Mg (NO&/pg Fig. 1 Influence of increasing amounts of Mg(N03)2 on A the peak height and B the integrated absorbance of 0.2 ng of Cr at a pyrolysis temperature of 1500 "C. The units of integrated absorbance (B) are seconds Mg(NO,),. Thus 20pg of Mg(NO,) were used in the sub- sequent studies.Group IIIA Sc Yand La. This group did not really stabilize Cr. In fact Sc produced erratic and noisy peaks and the background was increased. Yttrium produced flat peaks with increased tailing probably by promoting the formation of chromium carbides and 20 pg of La(NO,) had no effect on the AA signal of Cr. Carbide forming elements (Groups IVA VIA and VIIA) Zr Wand Re. None of these elements stabilized Cr. Zirconium led to slow atomization therefore flat peaks with tailing were recorded. Memory effects were observed and the sensitivity decreased. Zirconium and Cr apparently formed stable mixed carbides. With 20 pg of Re the sensitivity of the determination and Tpyr were not affected. When 20 pg of Na,WO were introduced with the Cr solution in the graphite tube neither significant enhancement of the signal nor thermal stabilization were observed and the appearance temperature did not increase substantially. When the pyrolytic-graphite coated graphite tube was impregnated by a solution containing l o g 1-' of Na,WO following the procedure proposed by Ortner and Kantuscher, an increase of about 20% in the peak height was noticed but the peak area did not alter.Noble metals Ru Rh Pd and Pt. All of these metals enhanced the AA of Cr although to different extents. When masses of their chloride salts of greater than 10 pg were injected the excess of chlorine depressed the AA signal of Cr probably owing to the formation of volatile Cr-chloro compounds' in the atomization step.The same was apparent with lower masses of the Pt modifier. It was found that 20 pg of NH,SCN completely eliminated this interference. It is likely that NH,SCN promoted the early reduction of these compounds to the metals and chlorine was driven off from the graphite tube during the early stage of the pyrolysis step (probably as NH,Cl or HCl). Therefore when masses greater than 10 pg were employed 20 pg of NH,SCN were injected with the modifier solution. Careful optimization of the masses of modi- fier used was carried out because it was shown that the mass of modifier influenced the sensitivity and the Tpyr greatly,19924725 especially in the case of the noble metals. The optimum amounts of modifiers were 1 pg of Rh; 5 pg of Ru; 10 pg of Pd; and 1 pg of Pt.The influence of increasing amounts of Rh on the Cr signal at a Tpyr of 1400°C is shown in Fig. 2. It is evident from this figure that 0.25 pg of Rh were sufficient for restoring the integrated absorbance signal of Cr. The same pattern was also observed with the other modifiers except for Pd. However Ru could not be used as chemical modifier because the background absorption at the resonance line of O L Ll 0 1 2 3 4 Rh mass/pg Fig. 2 Influence of increasing amounts of Rh on A the peak height and B the integrated absorbance of 0.2ng of Cr at a pyrolysis temperature of 1400 "C. The units of integrated absorbance (B) are seconds Journal of Analytical Atomic Spectrometry January 1996 Vol. 11 33Cr was elevated; it also showed a considerable blank value and a decrease in the sensitivity was observed after 30 success- ive injections of 10 pg of Ru.Palladium was useful as a modifier only when 10 pg of Pd were mixed with 20 pg of Mg(N03),. This mixed modifier neither changed the appearance tempera- ture of Cr nor increased the sensitivity but it caused a shift in Tpyr to 1300"C compared with 1500°C achieved with only 20 pg of Mg(N03)2. Therefore the mixed Pd-Mg modifier was not tested further. Only Pt and Rh stabilized Cr during the pyrolysis stage and increased the peak area signal even when using remarkably low masses of modifier. These modifiers were used further in this study. The transient signals of 10 pg 1-l of Cr in the presence of different modifiers are shown in Fig. 3. This figure indicates that both modifiers (Mg and Pt) decreased the peak width and increased the peak height which implies an increase in the rate of atomization in their presence.Moreover the Cr peak showed less tailing which means that the interaction with graphite was less favourable in the presence of Pt. The same phenomena were observed with Rh as a chemical modifier. Various mechanisms have been proposed for the atomization of Cr including the thermal dissociation of its oxides,26-28 and the thermal decomposition of chromium carbide (Cr,C2) and desorption of adsorbed C~(S).~' Magnesium nitrate has been p r o p ~ s e d * ~ ' ~ * ~ ~ ~ ~ ~ as a chemical modifier to stabilize chromium oxide avoiding losses of the volatile suboxide. Since masses of modifiers much larger than the stoichiometric amounts had to be used in order to recover the Cr signal at high Tpyr (Figs.1 and 2) trapping of Cr species in the bulk of the modifier mass and/or the formation of a solid solution of Cr with the modifier is a more probable explanation. It is likely that losses attributed to the formation of volatile CrO at the active sites of graphite are prevented owing to blockage of these sites by the modifiers. AnalyticalJigures of merit Calibration curves were constructed following the procedure described under Experimental and applying the temperature programme given in Table 1. The limits of detection and characteristic masses are given in Table 4. These parameters are in agreement with those previously rep~rted.~."-~~ Both noble metals (Pt and Rh) gave slightly improved characteristic masses.However Pt did not affect the baseline noise giving LOD values similar to those obtained in the absence of a modifier. The relative standard deviations for nine replicate injections of a solution containing 5 pg 1-l of Cr in 1% v/v HN03 were 1.2 and O.8% when peak height and integrated absorbance were measured respectively. The corresponding 1.0 I I I I I 0.8 A 0) c 2 0.6 3 0.4 0.2 0 8 1 2 3 4 Atomization tirnels Fig. 3 Atomization signals for 0.2 ng of Cr in the presence of various chemical modifiers A no modifier Tpyr= 1200 "C peak height 0.587 integrated absorbance 0.269 s; B 20 pg of Mg(N03)2 qyr= 1500 "C peak height 0.746 integrated absorbance 0.277 s; C 1 pg of Pt +20 pg of NH4SCN qyr= 1300 "C peak height 0.866 integrated absorbance 0.312 s.In all instances wall atomization was used at 2300 "C Table 4 Sensitivity and LODs for the determination of Cr LOD/'pg 1-' molpg Modifier PH* IAt PH IA None 0.066 0.072 1.9 3.3 Mg(N03)2 0.17 0.18 1.5 3.2 Na2W04 0.21 0.23 1.9 3.2 Rh 0.11 0.14 1.4 3 .O PtS 0.058 0.09 1 1.4 2.8 * PH peak height absorbance. t IA integrated absorbance. $With NH4SCN. values in the presence of modifiers were 1.4 and 4.0% [Mg(NO,),]; 1.9 and 4.5% (Na,WO,); 1.7 and 1.0% (Rh); and 1.4 and 0.9% (Pt). Determination of Cr in Rain Water The Tpyr in the absence of any chemical modifiers was 1300 "C whereas in the presence of 20 pg of Mg(NO,) it was 1550 "C and in the presence of 1 pg Pt 1500°C. It is likely that the matrix components in combination with Pt allowed Tpyr. to be increased substantially.The increased thermal stabilization of Cr in this type of sample led to the determination being carried out without any interferences. No difference in the slopes of the calibration curves with aqueous and matrix-matched stan- dards was observed therefore quantification was possible with aqueous standards and integrated absorbance measurements. It is shown in Table 5 that excellent recoveries were obtained with or without chemical modification by Pt. The within-batch precisions were 2.8 and 3.2% without and with Pt respectively for 0.55 pg 1-l of Cr (n=5). The between-batch precisions were 3.9 and 4.0% without and with Pt respectively for 0.55 pg 1-l of Cr (n = 5). Rain water samples from two stations one urban (St. A) and one rural (St. B) were analysed and the concentrations ranged from 0.21 to 3.15 pg 1-I in St.A (n= 14) to 0.55 to 1.48 pg 1-l in St. B (n=8). The wet deposition (surface concentration pg m-2) of Cr ranged from 2.7 to 194 pg m-2 in St. A and 6.4 to 148 pg rn- in St. B. Determination of Cr in Serum The pyrolysis and the atomization curves for Cr in a serum sample are shown in Fig.4. The atomic absorption signal of Cr was thermally stable up to 1100°C in the absence of chemical modifiers. The 7byr was 1300°C in the presence of 2 pg of Pt and 1400°C in the presence of 20 pg of Mg(N03),. The optimum mass of Pt was found to be 2 pg (higher than in the case of aqueous solutions) and addition of NH4SCN was not necessary. The organic serum components and/or Triton X-100 probably promoted the early formation of Pt during the drying and /or the pyrolysis step.Recovery experi- ments were carried out with and without modifiers and the results are summarized in Table 6. Consistently better recovery Table5 Recovery of Cr added to rain water with and without Pt using integrated absorbance measurements Recovery (n=3)(%) Cr added/pg I-' No modifier 1 pg of Pt 0.5 97.7 f 3.6 99.0 k 1 .o 1 .o 98.0 i- 2.0 98.3 & 1.2 1.5 99.0 f 3.4 100.1 & 2.4 2.0 101.4 f 1.7 100.9 rfI 2.9 3.0 100.7f 1.0 101.0f 1.1 4.0 99.4k 1.8 99.9 f 1.4 34 Journal of Analytical Atomic Spectrometry January 1996 Vol. 110.20 - u) a V \ 4 -8 c. ; 0.10 2 0.15 - (d E CI) - - 0.05 b I 1000 1200 1400 1600 1800 2000 2200 2400 2600 Temperature/OC Fig. 4 Pyrolysis (hollow symbols) and atomization (full symbols) curves for a 1 + 1 diluted serum sample spiked with 5 pg I-' of Cr in the presence of various chemical modifiers (0) no modifier (0) 20 pg of Mg(NO,) and (0) 1 pg of Pt.Table6 Recovery of Cr added to serum with and without chemical modifiers using integrated absorbance measurements Recovery (n=4) (%) ~~ ~~ Cr added/pg 1-' No modifier Mg(NO,) Pt 0.5 114.4 & 6.6 120.1 k8.9 98.0 & 1.0 1 .o 117.9 5 3.2 120.0 2 3.2 98.3 f 1.2 2.0 109.0 f 1.5 1 14.6 k 2.1 99.9 & 0.9 5.0 102.7f 1.8 104.6+ 1.5 99.7f1.1 was obtained in the presence of Pt. Since Mg(N0,)2 has been used successfully in the past for the determination of Cr in serum the increased recoveries we observed with this modifier could be attributed to the significantly lower concentration range of Cr used in the present study (0.5-5.0 pg l-') compared with those used in previously reported studies (generally 2.0-20.0 pg 1-I).The results for the determination of Cr in the Freezed-Dried Human Serum Reference Material are summar- ized in Table 7. Better agreement with the certified value was obtained in the presence than in the absence of Pt. It was found that build-up of a carbonaceous residue after successive injections affected the sensitivity and frequent recalibration was therefore required. The rn in the absence of a chemical modifier as well as in the presence of 1 pg of Pt or 20 pg of Mg(NO& were 1.7 and 2.9 pg for peak height and integrated absorbance measurements respectively with a new graphite tube. After 50 cycles the m was found to be 3.2 pg and after 150 cycles 3.7 pg for integrated absorbance measurements.The LOD in the absence of a chemical modifier was 0.051 pg Table 7 Determination of Cr in freezed-dried human serum refer- ence material Cr found/pg 1 - ' Certified Sample value/pg 1-' No modifier 1 PLg Pt 72 0.185 0.126 0.178 0.118 0.187 198-1 0.175 0.125 0.147 0.126 0.151 198-2 0.151 0.107 0.156 0.100 0.149 198-3 0.192 0.150 0.199 0.142 0.198 825 0.219 0.180 0.223 0.177 0.25 1 0.170 0.223 1-' whereas in the presence of 1 pg Pt or 20 pg Mg(N0,)2 the LODs were 0.041 and 0.091 pg 1-' respectively for integrated absorbance measurements. The within-batch pre- cisions were 3.0 and 3.8% without and with Pt respectively for 0.90 pg I-' of Cr (n=5). The between-batch precisions were 8.9 and 8.0% without and with Pt respectively for 0.62 pg 1-' of Cr (n=5).The proposed method with Pt as a chemical modifier was used to establish the Cr range in the serum of 22 healthy adults and 15 patients suffering from various thoracic disease^.^' The concentrations in samples ranged for the former group from < 0.05 pg I-' (below the LOD) to 0.69 pg 1-' whereas for the latter group they ranged from 0.18 to 1.13 pg 1-'. CONCLUSIONS Based on the results obtained in the present study it is concluded that several elements can stabilize Cr. The modifiers giving the best results in terms of thermal stabilization and sensitivity were Mg Rh and Pt. Palladium did not stabilize Cr. Tungsten (as 20 pg of Na,WO,) neither increased the Gyr nor the sensitivity. No interferences were observed in the determination of Cr in rain water and quantification could be performed with aqueous standards and integrated absorbance measurments without using chemical modifiers.The successful determination of Cr in serum was accomplished in the presence of Pt. Matrix-matched calibration was essential in order to achieve accurate and precise results. Periodic recalibration was required owing to a small decrease in sensitivity from the build-up of carbon. One of the authors (N. T.) thanks the Greek State Scholarships Foundation for financial support of this work. The Agricultural Bank of Greece is also thanked for providing part of the instrumentation. REFERENCES 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Kayne F. J. Komar G. Laboda H. and Vanderlinde R.E. Clin. Chem. 1979 24 2151. Veillon C. Patterson K. Y. and Bryden N. A. Anal. Chim. Acta 1982 136 233. McAughey J. J. and Smith N. J. Anal. Chim. Acta 1987 193 137. Halls D. J. and Fell G. S. J. Anal. At. Spectrom. 1986 1 135. Berndt H. and Sopczak D. Fresenius' Z . Anal. Chem. 1987 329 18. Matsusaki K. Yoshino T. and Yamamoto Y. Anal. Chim. Acta 1981 124 163. Matousek J. P. and Powell H. K. J. Spectrochim. Acta Part B 1988 43 1167. Castillo. J. R. Mir J. M. and Bendicho C. Fresenius' Z . Anal. Chem. 1988 332 783. Arpadjian S. and Krivan V. Fresenius' 2. Anal. Chem. 1988 329 745. Beceiro-Gonzalez E. Bermejo-Barrera P. Bermejo-Barrera A. Barciela-Garcia J. and Barciela- Alonso C. J. Anal. At. Spectrom. 1993 8 649. Manzoori J. L. and Saleemi A. J. Anal. At.Spectrom. 1994 9 337. Slavin W. Carnrick G. R. and Manning D. C. Anal. Chem. 1982 54 621. Manning D. C. and Slavin W. Appl. Spectrosc. 1983 37 1. Veillon C. Anal. Chim. Acta 1984 164 67. Lewis S. A. O'Haver T. C. and Harnly J. M. Anal. Chem. 1985 57,2 Cimadevilla E. A.-C. Wrobel K. Gayon J. M. M. and Sanz- Medel A. J. Anal. At. Spectrom. 1994 9 117. Tsalev D. L. Slaveykova V. I. and Mandjukov P. B. Spectrochim. Acta Rev. 1990 13 225. Tsalev D. L. and Slaveykova V. I. J. Anal. At. Spectrom. 1992 7 147. Thomaidis N. S. Piperaki E. A. and Efstathiou C. E. J. Anal. At. Spectrom. 1995 10 221. Tsalev D. L. Slaveykova V. I. and Mandjukov P. B. Fifth Journal of Analytical Atomic Spectrometry January 1996 Vol. 11 35Colloquium Atomspektrometrische Spurenanalytik ed. Welz B. Bodenseewerk Perkin-Elmer Uberlingen 1989 pp. 177-205. Smirnioudi V. N. and Siskos P. A. Atmos. Environ. 1992,26B 845. Versieck J. Vanballenberghe L. De Kesel A. Hoste J. Wallaeys B. Vanderhaute J. Baeck N. Steyaert H. Byrne A. R. and Sunderman F. W. Jr. Anal. Chim. Acta 1988 204 63. 23 Ortner H. M. and Kantuscher E. Talanta 1975 22 581. 24 Frech W. Li K. Berglund M. and Baxter D. C. J. Anal. At. Spectrom. 1992 7 141. 25 Mandjukov P. B. Vassileva E. T. and Simeonov V. D. Anal. Chem. 1992,64 2596. 26 L'vov B. V. and Fernandez H. A. Zh. Anal. Khim. 1984,39,221. 27 Frech W. Lindberg A. O. Lundberg E. and Cedergen A. Fresenius' Z . Anal. Chem. 1986 323 716. 21 22 28 Castillo J. R. Mir J. M. and Bendicho C. Spectrochim. Acta Part B 1988 43 263. 29 Fonseca R. W. Wolfe K. I. and Holcombe J. A. Spectrochim. Acta Part B 1994 49 399. 30 Frech W. Lundberg E. and Cedergren A. Prog. Anal. Atom. Spectrosc. 1985 8 257. 3 1 Daunderer M. Metallvergiftungen Diagnostik und Therapie in Kompendium der Klinischen Toxikologie Part III-Volume 9 ecomed Verlagsgesellschaft Landsberg Miinchen 1988. Paper 51035720 Received June 5 1995 Accepted September 20 1995 36 Journal of Analytical Atomic Spectrometry January 1996 Vol. 11