JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 199 1 VOL. 6 439 Verification of a Correction Procedure for Measurement of Lead Isotope Ratios by Inductively Coupled Plasma Mass Spectrometry* Michael E. Ketterer United States Environmental Protection Agency National Enforcement Investigations Center Box 25227 Building 53 Denver Federal Center Denver CO 80225 USA Michael J. Peterst and Preston J. Tisdaleg ICF/Kaiser Engineers 165 South Union Boulevard Suite 802 Lake wood CO 80228 USA Inductively coupled plasma mass spectrometry (ICP-MS) is a suitable method for determining Pb isotope ratios. This study investigates the effectiveness of a mass discrimination correction technique which is based upon the addition of TI to the sample and measurement of the 205TI:203TI ratio. The TI correction method has been found to be highly effective for minimizing bias and drift in Pb isotope ratios under a variety of operating and sample matrix conditions and for various samples independently analysed by thermal ionization mass spectrometry (TIMS).Biases for ratios with respect to *04Pb are generally controlled at <0.5% and relative standard deviations of 0.2% are attainable. Results using ICP-MS for environmental samples compare favourably with those obtained by TIMS. This correction technique is sutiable for routine analysis of environmental samples. Keywords Lead isotope ratio determination; inductively coupled plasma mass spectrometry; thallium based mass discrimination correction; reference material analysis; environmental samples analysis The determination of Pb isotope ratios and their use in geochemical and environmental studies is a subject that has evoked much interest since the first measurements were reported by Nier in 1938.' The details of theories and applications relating to this subject have been extensively Variations in Pb isotope ratios have been used in environmental studies1*-18 to determine the source of Pb contamination in specimens of interest.Typically Pb in environmental samples reflects contributions from multiple sources and thus isotope ratios are observed that are weighted averages of individual ore sources. Large varia- tions in 2oaPb:204Pb and 208Pb:204Pb are expected to be based upon consideration of the major worldwide sources of lead ores. Lead isotope ratios have been determined using induc- tively coupled plasma mass spectrometry (ICP-MS).19-27 Systematic biases have been observed for reference ma- terials such as the National Institute for Standards and Technology (NIST) Standard Reference Material (SRM) Common Lead and for environmental samples compared with those obtained by thermal ionization mass spectrome- try (TIMS). A method of correcting biases due to mass discrimination has been proposed by Longerich et The Longerich study demonstrated that T1 could be added as an internal standard for mass discrimination correction and that this correction technique could possibly lead to the removal of mass discrimination effects. For applications related to enforcement of environmental regulations the precisions attainable by ICP-MS for Pb isotope ratio measurements are generally satisfactory.How- ever the need to produce ratios that are minimally biased so that samples can be compared with published ore results and to ensure between-batch and interlaboratory compara- bility is of great concern. The purpose of the present study was to investigate the effectiveness of the use of T1 as an internal standard correction. *Presented in part at the 32nd Rocky Mountain Conference ?Present address J. F. Sat0 Associates 1667 Cole Boulevard $ Present address US Department of Agriculture Agricultural Denver CO USA July 28th-August 2nd 1990. Suite 175 Golden CO 8040 1 USA. Research Center NAL Building Beltsville MD 20705 USA. Experimental Materials and Reagents Trace-metal grade HNO (16 mol drn-,) and HCl (12 mol dm-,) were used without further purification.Distilled de-ionized water was the solvent for all solutions. Thallium stock solution (1000 mg 1-l) was obtained from Spex Industries (Edison NJ USA). The following chemicals also obtained from Spex were used to prepare matrix element solutions Na2C03 RbNO CsNO and U03.H20. A stock NIST SRM 981 Common Lead solution (1500 mg 1-l of Pb) was prepared by dissolving a portion of the metal in 1 mol dm- HNO,. Solutions analysed in this study were prepared using the above reagents as needed. Preparation of Environmental Samples Lead ore smelter fly ash and environmental samples were prepared for ICP-MS analysis by open- or closed-vessel digestion with a mixture of HN0,-HCl. Prior to ICP-MS measurement digestates were diluted with 0.16 mol dm- HNO to give solutions of 0.5- 1 .O mg 1-l of Pb and were spiked with 0.5 mg 1-l of T1.Instrumentation A Sciex Elan Model 250 ICP mass spectrometer equipped with mass flow meters for all gas streams and a peristaltic sample delivery pump was used in these studies. A refrigerated circulating bath was used to maintain the nebulizer spray chamber at a temperature of 10 "C. Meinhard TR-C concentric glass nebulizers (J. E. Meinhard Associates Santa Ana CA USA) were used. The ion optics of the spectrometer are the updated version; voltage adjustments consist of a B lens (barrel) P lens (plate) El (Einzel) and S2 (photon stop). The instrument was operated in the multichannel (peak-hopping) mode with single measurements being taken at the nominal mass value of each peak.For all scans an equal measurement time was used for each isotope. The low resolution mode was used producing peak widths of 1.0-1.1 mlz at 10% height. The program 'Spectrum Display' was used to collect data which were directed to a personal computer for storage and manipulation.440 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 199 1 VOL. 6 Table 1 Partial factorial design for investigation of the effects of multiple parameters upon Pb isotope ratio bias and precision Experiment Nebulizer type B Lens setting Measurement Cycle Sample flow Nebulizer flow time/s time/s rate/ml min-' rate/l min-l TR-C-0.5 TR-C-0.5 TR-C-0.5 TR-C-0.5 TR-C-3.0 TR-C-3.0 TR-C-3.0 TR-C-3.0 25 25 30 30 25 25 30 30 20 20 100 100 100 100 20 20 0.5 2.0 0.5 2.0 0.5 2.0 0.5 2.0 1 .o 2.0 1 .o 2.0 2.0 1 .o 2.0 1 .o 1 .oo 1.12 1.12 I .oo 1 .oo 1.12 1.12 1 .oo Isotope Ratio Measurements by ICP-MS The influence of the parameters of nebulizer flow rate nebulizer type sample pump speed total measurement time cycle time and €3 lens setting upon the precisions and biases of 206Pb:204Pb 207Pb:204Pb and 208Pb:204Pb measure- ments for NIST SRM 981 were investigated (Table 1).For each set of conditions three scans of a blank solution (0.5 mg 1-l of T1 in 0.16 mol dm-3 HN03) and ten scans of a 0.75 mg 1-l NIST SRM 981 Pb solution containing 0.5 mg 1-l of T1 in 0.16 mol dm3 HN03 were obtained. The precision and bias of 206Pb:204Pb 207Pb:204Pb and 208Pb:204Pb for three blank-corrected scans of NIST SRM 981 were measured at ion optics settings which produced varying degrees of mass discrimination.Settings for lens B of 10 20 25 30 35 40 and 50 were used with the lens values E1=42 P=07 and s2=37 being held constant. Also investigated were measurements of 206Pb:204Pb 207Pb:204Pb and 208Pb:204Pb for five blank-corrected scans of NIST SRM 981 in a reference matrix of 0.16 mol dm-3 HN03 and in matrices containing 1000 mg 1-l of Na Rb Cs and U. Isotope ratios were measured in environmental samples using a mass integration time of 100 s and with correction for reagent blank contributions and 204Hg isobaric interference. Blank and Mass Discrimination Correction Equations It was necessary to measure and correct for the influence of reagent blanks. Furthermore for the ICP-MS instrument used in this laboratory the addition of T1 to the solution to be analysed generally produced a signal at m/z 204 that is 1 x 10-3-1 x that of the 205 m/z signal (for a low resolution of 1 .O- 1.1 m/z).For the sample signals 2204-i208 and the corresponding blank signals i204b-i208b the following equations apply i204c= 2204 - (i204b)(i205)/(2205b) (1) i206c= 2206 - (i206b)(i205)/(~205b) (2) i207c= 2207- (i207b)(i205)~(~205b) (3) Z208c= 2208 - (Z208b)(~205)~(~205b) (4) The terms i204c i206c i207c and z20gc are blank corrected signals. The term (i205)/(i205b) corrects for time-dependent and sample-derived differences in analyte sensitivity be- tween the blank and sample solutions in a similar way to internal standardization in quantitative analysis. The term i205b is the intensity of a reagent blank spiked with TI.No additional correction is performed for intensities obtained at m/z 203 and 205 in the absence of TI. Inclusion of the (i205)/(i205b) term has been found to be significant under some circumstances and can alter the resulting ratios by up to 1 %. The isotope ratios corrected for mass discrimination effects are obtained from (6) (7) The terms (i206c/i204c) etc. are the isotope ratios uncorrected for mass discrimination effects. The naturally occurring 205Tl:203Tl ratio is 2.3871 and (i205/i203) is the same ratio measured in the sample. Eqns. (1)-(7) were applicable to all experiments investigating the isotope ratios of NIST SRM 98 1. The 204Hg isobar was readily corrected for in environ- mental samples using the following equations (additional isobaric corrections for Os Ir and Pt polyatomic ions were unnecessary) 207Pb:204Pb = ( i207c/i204c)[2.387 1 /( i20s/i203)]1.5 208Pb:204Pb = ( i208c/i204c)[2. 387 1 /( i2o5/i2O3)l2 Z204hc= i2O4c-0.5 [i201 -(i201b i20S/i205b)l (8) The term i204hc is the intensity at m/z=204 which is corrected for both reagent blank Pb contribution and the 204Hg isobar. The term 0.515 is the 204Hg:201Hg natural abundance ratio iZol is the reagent blank signal at m/z 201 and i204c is from eqn. (1). The isotopic ratios are then computed by using i204hc in place of i204c in eqns. (5)-(7). Results and Discussion Effect of Operating Parameters on Isotope Ratios Table 2 lists the means and statistics of both the uncor- rected Pb and T1 isotope ratios and the ratios corrected for mass discrimination effects obtained using the instrumental parameters shown in Table 1.It is evident that the uncorrected isotope ratios are significantly positively bi- ased; this bias is nearly eliminated by applying eqns. (5)-(7). In the corrected ratios a small amount of residual negative bias is seen with the 207Pb:204Pb measurements in experiment 1 having the largest bias of -0.96%. The residual bias appears to diminish with time; experiments 1-8 were run in order. The precision values are generally comparable between the uncorrected and corrected ratios. Table 2 shows the isotope ratios of NIST SRM 981 corrected for bias using eqns. (1)-(7) and that this correc- tion is applicable under a variety of instrumental operating conditions. Statistical analysis of the results of Table 2 demonstrates that measurement time is the most influential variable among those investigated and the higher flow rates of the nebulizer argon supply and sample solution are also associated with improved precision.These two variables produce effects which are related to the counting statistics as increasing them also raised the ion-count rates for all isotopes. The B lens setting nebulizer type and cycle time all produced insignificant effects upon the precision of the ratio measurements. The isotope ratio bias is influenced by the B lens setting and by the nebulizer type but neither of these factors is of great significance. The effect of the nebulizer type is convoluted with the experimental run order in this study as seen in Table 1 since it was most practical to run experiments 1-8 in numerical order.The run order parameter is probably the more influential givenJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 199 1 VOL. 6 44 1 Experiment 1 2 3 4 5 6 7 8 Experiment 1 2 3 4 5 6 7 8 Uncorrected Corrected Uncorrected Corrected Table 2 Uncorrected and mass discrimination corrected results for the determination of NIST SRM 981 lead isotope ratios with varying nebulizer type B lens setting measurement time cycle time sample flow and nebulizer flow values * Certified values for 206Pb:204Pb 207Pb:204Pb and 20sPb:204Pb are 16.937 f 0.0 19 15,497 k 0.0 10 and 36.722 f 0.037 respectively. 17.111 20.103 17.332 f 0.039 17.341 f 0.034 17.225 f 0.040 17.316 k0.035 17.413f0.048 17.263 f 0.048 17.384 -t 0.098 16.879 f 0.089 16.862 k 0.059 16.9 10 k 0.044 16.922 f 0.042 16.905 -t 0.043 16.939 +- 0.037 16.9 15 +- 0.056 16.935-tO.108 15.661 kO.101 15.928 f0.042 16.02 1 f 0.033 16.036 k 0.032 15.883 k 0.034 15.999 k 0.033 16.138 k 0.057 16.094 k 0.089 15.344 f 0.086 15.457 f 0.048 15.442 f 0.036 15.450 f 0.056 15.376 f 0.065 14.440 f 0.045 15.480 k 0.044 15.475 +O.11 7 Uncorrected 37.36 1 f 0.2 10 38.304 k 0.147 38.464 f 0.093 38.449 k 0.083 37.992 k 0.078 38.355k0.083 38.761 f0.103 38.540f0.213 Corrected 36.355 -t 0.2 11 36.546 k 0.241 36.61 3 f 0.135 36.611 k0.147 36.593 k 0.110 36.705 k 0.120 36.575 f 0.21 1 36.578 f 0.318 Uncorrected Corrected 2.4 199 f 0.0054 2.4439 f 0.0044 2.4467 f 0.0027 2.4323 k 0.0024 2.4402 k 0.0027 2.4574 k0.0067 2.4503 f 0.0058 2.4463 k 0.0036 (2.3871) (2.387 1) (2.3871) (2.3871) (2.3871) (2.3871) (2.3871) (2.3871) Table 3 Uncorrected and mass discrimination corrected results of the determination of NIST SRM 98 1 lead isotope ratios for several B lens settings B Lens setting 10 20 25 30 35 40 50 Uncorrected Corrected 16.592f0.102 16.923 k 0.099 16.846 k 0.070 16.923 k 0.06 1 16.83 1 f 0.026 16.855 f0.035 17.269 f 0.133 16.938k0.148 17.097 f 0.034 17.2 12 f 0.042 17.56 1 f 0.060 16.928 k 0.070 16.939 f 0.075 16.920 f 0.024 Uncorrected Corrected 15.005 k0.108 15.455 k0.106 15.322 f 0.050 15.427 f 0.033 15.350 k 0.024 15.382 -t 0.037 16.656 k 0.01 9 15.425 f 0.077 15.791 k0.042 15.4 16 k 0.087 15.936 k0.130 15.480k 0.151 16.273 f 0.022 15.390 f0.025 B Lens setting 10 20 25 30 35 40 50 Uncorrected 35.171 k0.167 36.13 1 f 0.089 36.373 f 0.049 37.175 f 0.037 37.599f 0.054 38.085 f 0.320 39.1 17 f 0.086 Corrected 36.584k 0.203 36.459 f 0.046 36.474 f 0.065 36.447 f 0.225 36.41 5 f 0.184 36.637 k 0.376 36.3 13 k 0.063 Uncorrected 2.3405 2 0.0054 2.3763 f 0.0023 2.3838 k 0.0033 2.4 109 f 0.0069 2.4256 f 0.0048 2.4776 -t 0.0049 2.4338 f 0.0040 Corrected (2.3871) (2.3871) (2.3871) (2.3871) (2.8371) (2.3871) (2.387 1) the tendency of ICP-MS instrumentation to show drift over time.29 In Table 3 precisions and biases are shown for uncor- rected and TI corrected Pb isotope ratios at various B lens settings. As previously shown by Longerich et al.,28 the B lens has the effect of shifting the mass response curve; shifts in the setting from 10 to 50 produce uncorrected results ranging from strongly negatively biased (setting 10) to strongly positively biased (setting 50).The mag- nitude of the effect clearly increases in the order 2MPb:204Pb<207Pb:204Pb<208Pb:204Pb. It was also noted that the zero-bias B lens setting is affected by the other ion optics settings; however the type of behaviour seen in Table 3 usually exists at other sets of ion lens conditions. The correction is thus capable of alleviating instrumentally induced bias over a small range of ion optic conditions. On a practical note use of this correction procedure alleviates the requirements for careful tuning prior to analysis; drift during analysis in the degree of mass discrimination is tolerable also. Effect of Sample Matrix Table 4 shows the influence of 1000 mg 1-l matrices upon the raw and corrected isotope ratios.It is evident that the addition of matrix elements produces large negative shifts in the uncorrected Pb ratios with respect to 204Pb. However this trend is the opposite of the expected sample matrix effect which is a preferential attenuation of the lighter isotopes. An explanation for this result is not apparent at present. It is further noted that the experiments were run in the matrix addition order (none Nay Rb Cs and U) followed by a repeat of the no-matrix NIST SRM 981 solution which produced biased results (prior to TI based442 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 199 1 VOL. 6 Table 4 Uncorrected and mass discrimination corrected results for the determination of NIST SRM 98 1 lead isotope ratios under varying matrix conditions 206pb:204pb 207pb:204pb Matrix Uncorrected Corrected Uncorrected Corrected None 16.91540.057 16.974 k 0.064 15.480 k 0.042 15.56 1 k 0.053 1000 mg I-' of Na 16.830 k0.064 16.9 15 f 0.03 1 15.374 f 0.049 15.490f 0.044 1000 mg I-' of Rb 16.770 k 0.056 16.887k0.054 15.28 1 k0.064 15.443 f 0.06 1 1000 mg I-' of Cs 16.605 k 0.087 16.853 k0.074 15.120 f 0.088 15.460 k 0.071 1000 mg 1 - I of U 16.608 k 0.01 9 16.856 k 0.053 15.163 k 0.027 15.505 -t- 0.076 208pb 204pb 205T1:203Tl Matrix Uncorrected Corrected Uncorrected Corrected None 36.485 f 0.122 36.761 rf~ 0.167 2.3788 k0.0033 (2.3871) 1000 mg 1-I of Na 36.186k0.119 36.550k0.167 2.3752k0.0077 (2.3871) 1000 mg I-' of Rb 35.942 4 0.172 36.449 f 0.126 2.3704 k 0.0053 (2.387 1) 1000 mg I-' of Cs 35.446 f 0.233 36.5 12 k 0.174 2.3520 k 0.0027 (2.3871) 1000 mg 1-1 of U 35.538 k 0.032 36.610k0.124 2.3519k0.0049 (2.3871) Table 5 Comparison of ICP-MS and TIMS lead isotope ratio results for ore tailings soil and smelter fly ash samples Sample Technique 206pb:204Pb 207Pb:*04Pb 20sPb:204Pb [Pb]/mg kg-I Galena ICP-MS (n=3)* TIMS (n=2)* Tailings No.1 ICP-MS (n=3) TIMS (n=2) Tailings No. 2 ICP-MS (n=3) TIMS (n=3) Soil no. 1 ICP-MS (n= 3) TIMS (n=2) Soil no. 2 ICP-MS (n= 3) TIMS (n=3) Soil no. 3 ICP-MS (n= 5 ) TIMS (n=2) TIMS (n=2) Fly Ash ICP-MS (n=3) 16.26 k 0.07 15.33 k 0.07 16.25 k 0.01 15.40 k 0.02 18.13 k 0.07 15.53 4 0.05 18.11 50.02 15.58k0.03 16.32f0.06 15.37k0.06 16.27k0.01 15.39k0.02 18.18 4 0.05 15.50 4 0.09 18.17 k 0.02 15.58 k 0.02 17.64 k 0.05 1 5.52 4 0.05 17.5740.01 15.53 k0.02 16.65 k0.13 15.36 k0.12 16.60 k 0.01 15.40 -t 0.01 18.1 1 k0.02 15.4720.02 18.13 k 0.03 1 5.58 t- 0.04 35.79 k 0.13 388 000 36.01 f 0.05 37.58 k 0.18 26 000 37.67 -t- 0.10 35.95 k0.19 36 300 35.99 rt 0.06 37.57 k0.24 5 470 37.72 k 0.08 37.2620.24 1 1 100 37.32 f 0.06 36.24 +- 0.32 600 36.30 f 0.01 37.47 f 0.06 63 1 000 37.71 f0.19 * Replicates are of the digestion and measurement stages for ICP-MS and the measurement stage for TIMS. correction) similar to the U reference matrix. Following a 2 d idle period the uncorrected reference matrix ratios with respect to 204Pb returned to the initial values. Thus some time dependency exists. Regardless of origin and mecha- nism the sample-induced bias effect is amenable to remedy using T1 to correct for mass discrimination effects.Environmental Samples In Table 5 comparative results are shown for the determi- nation of 206Pb:204Pb 207Pb:204Pb and 20sPb:204Pb for seven environmental samples using ICP-MS and TIMS. Gener- ally excellent agreement is seen for the two analytical methods for this group of samples; in some instances biases of 0.5-1 .O% exist. The relative standard deviations (RSDs) of the results of the analyses of environmental samples using ICP-MS are similar in most instances to those obtained with NIST SRM 981. These samples have been developed as in-house batch quality control materials and isotope ratios for these samples measured by ICP-MS with T1 based correction exhibit long-term reproducibility.Thus data obtainable by the methods presented here are of a suitable quality for the enforcement of environmental regulations. Conclusions Measurements of Pb isotope ratios by ICP-MS can be significantly affected by both instrumental and sample- induced sources of bias. It was observed that substantial changes in mass discrimination could be brought about by the presence of 1000 mg 1-' of concomitant elements. Both sources of bias are readily corrected for by using the value of 205Tl:203Tl measured in the sample. The T1 correction method first proposed by Longerich et is apparently successful because T1 closely emulates the mass discrimina- tion level of Pb under a variety of instrumental and sample conditions. For a measurement time of 100 s the RSDs for 206Pb:204Pb 207Pb:204Pb 20sPb:204Pb and 205Tl:203Tl measure- ments are apparently limited to about 0.20 0.25 0.30 and 0.10% respectively.These values are larger by a factor of 2-3 than are predicted by counting statistics considerations. Inductively coupled plasma MS is a robust and practical method of measuring Pb isotope ratios in environmental samples where data quality requirements are of the order of 0.5% acceptable bias and 0.2-0.5% precision. The authors thank R. E. Zartman and L. M. Kwak of the US Geological Survey for providing comparative TIMS results for the environmental samples. This work was supported by the US Environmental Protection Agency. Specific vendors are mentioned for information purposes only. References 1 Nier A.O. J . Am. Chem. Soc. 1938 60 1571. 2 Bate G. L. and Kulp J. L. Science 1955 122 970. 3 Cahen L. Eberhardt P. Geiss J. Houtermans F. G. Jedwab J. and Signer P. Geochim. Cosmomchim. Acta 1958,14 134.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 199 1 VOL. 6 443 4 Russell R. D. and Farquhar R. M. Lead Isotopes in Geologjj Interscience New York 1960. 5 Cannon R. S. Pierce A. P. Antweiler J. C. and Buck K. L.. Econ. Geol. 1961 56 1. 6 Brown J. S. Econ. Geol. 1962 57 673. 7 Stacey J. S. Zartman R. E. and NKomo I. T. Econ. Geol. 1968 63 796. 8 Doe B. R. and Delevaux M. H. Econ. Geol. 1972 67 409. 9 Doe B. R. and Stacey J. S. Econ. Geol. 1974 69 757. 10 Saager R. and Koppel V. Econ. Geol. 1976 71 44. 11 Koppel V. and Gruenfelder M. in Lectures in Isotope Geologv eds.Jager E. and Hunziker J. C. Springer-Verlag Berlin 1979 p. 134. 12 Chow T. J.. and Johnstone M. S. Science 1965 147 502. 13 Ault W. U. Senechal R. G. and Erlebach W. E. Environ. Sci. Technol. 1970 4 305. 14 Rabinowitz M. B. and Wetherill G. W. Environ. Sci. Technol. 1972 6 705. 15 Gulson B. L. Tiller K. G. Mizon K. J. and Merry R. H. Environ. Sci. Tech. 1981 15 691. 16 Facchetti S. Mass Spectrom. Rev. 1988 7 503. 17 Gulson B. L. Mizon K. J. Korsch M. J. and Noller B. N. Environ. Sci. Tech. 1989 23 290. 18 Facchetti S. Garibaldi P. in Proceedings of the International Symposium on Environmental Health Aspects of Lead Com- mission of the European Communities Luxembourg 1973 p. 995. 19 Gray A. L. Analyst 1975 100 289. 20 Houk R. S. Fassell V. A. Flesch G. D. Svec H. J. Gray A. L. and Taylor C. E. Anal. Chem. 1980 52 2283. 21 Data A. R. and Gray A. L. Analyst 1981 106 1255. 22 Date A. R. and Gray A. L. Analyst 1983 108 159. 23 Date A. R. and Gray A. L. Int. J. Mass Spectrum. Zon Phys. 1983 48 357. 24 Smith R. G. Brooker E. J. Douglas D. J. Quan E. S . K. and Rosenblatt G. J. Geochem. Explor. 1984 21 385. 25 Date A. R. and Cheung Y. Y. Analyst 1987 112 1531. 26 Hinners T. A. Heithmar E. M. Spittler T. M. and Henshaw J. M. Anal. Chem. 1987 59 2658. 27 Russ G. P. and Bazan J. M. Spectrochim. Acta Part B 1987 42 49. 28 Longerich H. P. Fryer B. J. and Strong D. F. Spectrochim. Acta Part B 1987 42 39. 29 Houk R. S. and Thompson J. J. Mass Spectrom. Rev. 1988 7 425. Paper 0/04859C Received October 29th I990 Accepted May 2Ist 1991