JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 199 1 VOL. 6 Determination of Lead in Soil by Slurry-Electrothermal Atomic Absorption Spectrometry With a Fast Temperature Programme 473 Michael W. Hinds Royal Canadian Mint 320 Sussex Drive Ottawa Ontario KIA OG8 Canada Kathryn E. Latimer Department of Chemistry University of Saskatchewan Saskatoon Saskatchewan S7N OWO Canada Kenneth W. Jackson Wadsworth Center for Laboratories and Research New York State Department of Health and School of Public Health State University of New York Albany NY 12201-0509 USA A fast temperature programme for the determination of Pb in soil by slurry-electrothermal atomic absorption spectrometry was developed by omitting the charring step and drying at a higher temperature for a shorter time.The effects of drying temperature and time were systematically studied. Work with phosphate and Mg-Pd as chemical modifiers indicates that better results can be obtained without the use of modifiers. Keywords Electrothermal atomic absorption spectrometry; slurry; fast temperature programme; soil analysis A long analysis time is a major disadvantage of trace element determinations by electrothermal atomic absorp- tion spectrometry (ETAAS) owing to the temperature programme steps needed for sample drying charring furnace cooling atomization and furnace cleaning. Halls' showed that analysis time could be reduced significantly by drying at a fast rate and omitting the charring step. Subsequent ~ o r k ~ - ~ applied this fast temperature pro- gramme technique to other elements in a variety of biological matrices.Slurry-ETAAS reduces the sample preparation time required for the determination of a number of different elements in various It also decreases the risks of sample contamination and high blank values that can occur during conventional sample digestion procedures. Accurate results can be obtained with aqueous calibration standards provided that slurry particle diameters are less than 50 This method has been further enhanced by the introduction of an autosampler for ~lurries.~ Recently there have been reports of the application of a fast temperature programme to slurry samples.10-12 Brad- shaw and Slavinll determined a variety of elements in coal and coal fly ash using a rapid temperature programme and automated slurry sample introduction. They reported ana- lysis times of less than l min per sample which allowed for high sample throughput.Bendicho and de Loos-Volle- bregt12 found that omission of the charring step made no difference to the determination of trace metals in glass by the slurry technique. The work presented here is a systematic study of the effect of drying time and temperature on Pb atomization from aqueous solutions and soil slurry samples without a charring step. The effccts of two common chemical modifiers Mg-Pd13 and phosphate,14 are described. Experimental Apparatus Two spectrometers were used in this work. One was a modified Perkin-Elmer Model 2280 with fast-response signal recording and measurement of the graphite furnace temperature with an optical pyrometer (Series 1 100 Ircon Niles IL USA).A programme written in Turbo Pascal (Borland Scotts Valley CA USA) correlated absorbance temperature and time data.15 The other spectrometer was a Perkin-Elmer Model 5000 with signals collected via a Perkin-Elmer Model 3600 data station. In both instances the atomizer was a Perkin-Elmer HGA- 500 with pyrolytic graphite coated graphite tubes and solid pyrolytic graphite L'vov platforms. The purge gas was Ar. A Pb hollow cathode lamp (Perkin-Elmer Part No. 303-5039) was operated at 8 mA. The wavelength used was 283.3 nm and the spectral bandwidth was 0.7 nm. Deuterium arc background correction was used throughout this study. Samples and Reagents Three Soil Certified Reference Materials from the Canada Centre for Mineral and Energy Technology (CANMET) were used.16 They were oven-dried at 105 "C for 24 h prior to use and did not require further grinding.The following stock solutions were prepared from analyt- ical-reagent grade chemicals 1000 mg 1-l of Pb prepared from Pb(N@& (Analchemia Mississauga Ontario Canada) 10% m/v Mg(N03)2-6H20 (BDH Toronto Ontario Canada) 10% m/v (NH,),HP04 (BDH) and 3000 mg 1-1 of Pd prepared from Pd(N03)2 (Johnson Matthey/ Aesar Toronto Ontario Canada). Solution and Slurry Preparation Aqueous solutions The stock Pb solution was serially diluted with l0/o HN03. Appropriate concentrations were introduced into the fur- nace with a 10 or 20 pl micropipette (Socorex ISBA Renens Switzerland). Chemical modifiers were pipetted on top of the aqueous standards so that the total volume of sample and modifier did not exceed 20 pl.Slurries Slurries were prepared by adding exactly 20 ml of distilled water to weighed amounts of soil (typically 45 mg) in a 50 ml beaker. The slurry was stirred magnetically for 5 min before sampling. Appropriate amounts of the modifier were added to the slurry. The volumes added were small so dilution effects were negligible. Results and Discussion Optimized Fast Temperature Programme The conventional drying and charring stages were replaced by a modified fast drying stage. It was considered that solutions and soil slurries might have different optimum fast drying temperatures and times hence this was investi-474 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1991 VOL. 6 a c 0 0.5 $ $ gated.Table 1 lists the fast temperature programme used in this study. The drying temperature (A') and drying time (r) were studied systematically by setting the drying tempera- ture (at 200 "C initially) and varying the drying time (10 s increments to a maximum of 60 s). Once a series was completed the temperature was increased by 100 "C (600 "C maximum) and the experiment was repeated. A record was kept of the absorbance signal peak area and peak profile. The drying process was also monitored visually through the use of a dental mirror. Occurrences of incom- plete drying and liquid sputtering off the platform were noted. The whole experiment was carried out for 10 and 20 pl aliquots of aqueous solution and soil slurry. When incomplete sample drying was observed the temperature programme was interrupted in order to prevent damage to the tube and platform.A programme that would completely dry the sample was then used in order to clear the atomizer ready for the next injection. The data for the entire optimization experiment are summarized in Fig. 1 which shows the minimum time with respect to a given drying temperature for a specified aliquot size to dry and which gives the same peak area disturbance for Pb runs with longer drying times. Peak area absorbances for both solution and slurry samples gave a relative standard deviation (RSD) of 4% with manual pipetting. The area above the curves indicates conditions where the solutions and slurries are completely dried. The area below each curve shows conditions where drying was incomplete.As expected the larger aliquots required longer . Table 1 gate drying temperatures and times Outline of the temperature programme used to investi- Step Temperature/"C Ramp time/s Hold time/s Drying 'P 1 r* Cooling 50 1 10 Atomization 1700 O f 5 Clean 2 700 1 5 * X and Y varied. t Gas flow stopped. read - 1 s. 50 40 30 20 10 v) . .- E o .w (a) .'\ 30 - 20 10 - - I * I I 0 100 200 300 400 500 600 TemperaturePC Fig. 1 Minimum time and temperature required for an aliquot of (a) aqueous solution and (b) slurry to dry without sputtering. A 10 pl aliquot; and B 20 p1 aliquot A A 0 1 2 3 4 Timels Fig. 2 Comparison of absorbance peak profiles for 0.6 ng of Pb using a fast temperature programme and 20 pl aliquots A aqueous solution (0.274 A s); and B soil slurry (0.252 A s) drying times.Temperatures above 400 "C resulted in changes in peak shapes (shoulders and double peaks). When drying at 600 "C instances of sputtering were observed and the peak areas were irregular (RSD >4%) probably owing to unobserved sputtering. However work carried out with a Perkin-Elmer 5 100 AA spectrometer at a different location (Ottawa) at 600 "C showed none of these effects. This is probably owing to differences in cooling-water temperature which varies with location and season. l 7 The experiments were carried out in winter (Perkin-Elmer 5 100 instrument) and in summer (Perkin-Elmer 2850 instrument). It has been observed that in Canada temperature differences in municipal water can range from 20 to 25 "C between summer and winter.This implies that analysts wishing to adopt this technique may have to adjust drying parameters to account for differences in cooling-water temperature. The peak shapes of Pb from aqueous solutions and soil slurries are shown in Fig. 2. The peaks are nearly coinciden- tal. Little variation was observed for various soil types contrary to previous observations1* when a charring step was used. The peak position did not vary with changes in the drying programme. The optimum fast temperature programme is shown in Table 2. This programme is suitable for aliquots of up to 20 p1 and results in single-peak profiles for aqueous standards. The temperature programme cycle time was reduced from 161 (typical soil slurry programme) to 43 s. Effect of Chemical Modifiers Chemical modifiers typically change the atomization char- acteristics of Pb.The optimum atomization temperatures were established to be 1800 "C for phosphate and 2000 "C for Mg-Pd. Phosphate Phosphate either as HP042- or H2P04- was shown previously to be an effective chemical modifier for Pb in Table 2 Optimized fast temperature programme for solutions and slurries (20 pl aliquots) Step TemperaturePC Ramp time/s Hold time/s Drying 400 1 20 Cooling 50 1 10 Atomization 1700 O* 5 Clean 2700 1 5 * Gas flow stopped read - 1 s.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 199 1 VOL. 6 475 H I I 0 1 2 3 4 Timels Fig. 3 Comparison of absorbance peak profiles for 0.6 ng of Pb plus 14 pg Of POA3- using a fast temperature programme and 10 p1 aliquots A aqueous solution (0.279 A s); and B slurry (0.262 A s) 8 0.3 0 x 0.1 - a 0 1 2 3 4 5 6 7 Timels Fig.4 Comparison of absorbance peak profiles for 0.6 ng of Pb plus 15 pg of Pd+lO pg of Mg(N03)*.6H20 using a fast temperature programme and 20 pl aliquots A aqueous solution (0.156 A s); B slurry (0.154 A s); and C background absorbance soil s1urTies.l8 It was considered that this modifier would delay the Pb signal without a charring step because the modifier is reported to act either in the gas phase19 and/or in the condensed phase,20 which would occur after drying. The experimental results support this contention as shown in Fig. 3 for both Pb in aqueous solution and Pb from a soil slurry. However with the addition of phosphate the Pb signal from the soil slurry is delayed more than the aqueous Pb signal.Without the phosphate the solution and slurry Pb signals coincide which is advantageous (Fig. 2). Work by Falk et a1.21 showed that temporal non-isothermal- ity exists within the graphite furnace during atomization. This suggests that the aqueous standard and soil slurry peaks ought to coincide in order to minimize the effect of furnace temperature differences on peak area measure- ments. Magnesium and palladium Peak profiles for Pb using Mg-Pd as a modifier are shown in Fig. 4. The modifier combination delays Pb atomization from the aqueous solution and soil slurry which was expected.22 However peak profiles of Pb absorbance from aqueous solution and slurry do not coincide. The slurry peak appears later than the solution peak.As discussed previously this detracts from the usefulness of the modifier combination. Another disadvantage in using this modifier is that the peak area absorbances for Pb in both aqueous solution and slurries are lower with the modifier than for Pb alone. The sensitivity is reduced using this modifier combination as shown in the differences in the characteristic masses (pg per 0.0044 A s) of 12.5 for Pb alone and 18.2 with the Mg-Pd modifier. This was not so when a 900 "C charring step was used previously.22 Table 3 Determination of Pb in soil with a fast temperature programme Concentration of Pb in soillpg g- I Soil Experimental* Certified? so- 1 $ 18.1 kO.01 21 k 4 so- 1 + Pod3- 16.1 k2.9 SO- 1 + Mg-Pd 19.3k 1.1 SO-34 SO-3 + Po43- SO-3 + Mg-Pd SO-47 SO-4 P043- SO-4 + Mg-Pd 12.3 + 0.1 1 4 k 3 l l s k 1.4 14.0 k 0.4 14.6k 0.7 1 6 f 3 12.9k 1.7 17.1 k 0 .7 SO-4 + 10% humic acid 13.9 f 0.8 21 + 4 SO-4 + 10% humic acid (with charring step) 9.6k 1.0 * +One standard deviation. t 95% confidence interval. $ Regosolic Clay Soil. 4 Calcereous C Horizon Soil. 7 Chernozemic A Horizon Soil. Quantitative determination Three CANMET Soil Certified Reference Materials were analysed using the fast temperature programme. Chemical modifiers (phosphate and Mg-Pd) were also used in some determinations (Table 3). Values for the determination of Pb without chemical modifiers are within the 95% confi- dence limits of the certified concentrations for each soil. From previous using a 900 "C charring step the addition of 10% humic acid reduced the Pb recovery substantially. However without the charring step much better results were obtained.It appears that large amounts of organic matter enhance volatilization losses during the charring step but do not interfere with the atomization of Pb from soil slurries. Experimental values obtained using an Mg-Pd modifier overlap with the 95% confidence limits of the certified concentrations although the peak profiles of aqueous solutions do not coincide with soil slurry profiles. The determinations with phosphate as a modifier tended to be low compared with the certified values. Precision was somewhat poorer for these determinations. Overall the addition of a modifier does not appear to assist in the determination of Pb in soil by slurry-ETAAS with a fast temperature programme.The risks of higher blanks temporal differences between standard and sample peaks and reduced sensitivity (for Mg-Pd) are not suffici- ently offset by any substantial benefits. Conclusion This study makes it evident that ETAAS temperature programmes for the direct determination of Pb in soil slurries may be shortened by drying at higher temperatures and eliminating the charring step. Ultimately this means that the analysis time can be reduced by a factor of four. Volatilization losses from organic-rich soil can be decreased by removing the charring step. For the determination of Pb in soil by slurry-ETAAS chemical modifiers do not appear to be of benefit in combination with the fast temperature programme used in this study. M. W. H. thanks the University of Saskatchewan and the Royal Canadian Mint for supporting this work.476 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 199 1 VOL.6 1 2 3 4 5 6 7 8 9 10 11 12 References Halls D. J. Analyst 1984 109 1081. Halls D. J. and Fell G. S. Analyst 1985 110 243. Halls D. J. Mohl C. and Stoeppler M. Analysf 1987 112 185. Halls D. J. Black M. M. Fell G. S. and Ottaway J. M. J. Anal. At. Spectrom. 1987 2 305. Keating A. D. Keating J. L. Halls D. J. and Fell G. S. Analyst 1987 112 1381. Hinds M. W. Jackson IS. W. and Newman A. P. Analyst 1985 110 947. Miller-Ihli N. J. J. Anal. At. Spectrom. 1988 3 73. Hoenig M. and Van Hoeyweghen P. Anal. Chem. 1986 58 2614. Miller-Ihli N. J. J. Anal. At. Spectrom. 1989 4 295. Hinds M. W. Allen K. and Jackson K. W. paper presented at XVI Federation of Analytical Chemistry and Spectroscopy Societies Meeting Chicago IL 1st-6th October 1989 paper No. 465. Bradshaw D. and Slavin W. Spectrochim. Acta Part B 1989 44 1245. Bendicho C. and de Loos-Vollebregt M. T. C. Spectrochim. Acta Part B 1990 45 679. 13 Schlemmer G. and Welz B. Spectrochim. Acta Part B 1986 41 1157. 14 May T. W. and Brumbaugh W. G. Anal. Chem. 1982 54 1032. 15 Allen E. and Jackson IS. W. Anal. Chim. Acta 1987 192 355. 16 Steger H. F. Bowman W. S. and McKeague J. A. Geostand. Newsl. 1985 9 213. 17 Slavin W. personal communication. 18 Hinds M. W. and Jackson K. W. J. Anal. At. Spectrom. 1987 2 441. 19 Czobik E. J. and Matousek J. P. Talanta 1977 24 837. 20 Bass D. A. and Holcombe J. A. Anal. Chem. 1987 59 974. 21 Falk H. Glismann A. Bergann L. Minkwitz G. Schubert M. and Skole J. Spectrochim. Acta Part B 1985 40 533. 22 Hinds M. W. and Jackson IS. W. J. Anal. At. Spectrom. 1990 5 199. 23 Hinds M. W. and Jackson K. W. J. Anal. At. Spectrom. 1988 3 997. Paper 1/00717C Received February 15th 1991 Accepted April 30th I991