Ž . Chemical Geology 148 1998 95–104 Uncertainties on lead isotope analyses: deconvolution in the double-spike method Roger Powell ) , Jon Woodhead, Janet Hergt School of Earth Sciences, UniÕersity of Melbourne, Victoria 3052, Australia Received 9 June 1997; accepted 29 January 1998 Abstract A weighted least-squares method is used to generate the ‘true’ lead isotopic composition from measurements on unspiked and spiked aliquots of a sample using the double-spike method. A form for the weighting based on theoretical mass spectrometer errors is used.The least-squares approach is more reliable for this deconvolution of lead isotopic compositions than current methods, and allows uncertainties and correlations on the true isotopic composition to be estimated, a feature of importance in many applications. The approach allows the prediction of run parameters for ‘optimum’ spike analysis. q 1998 Elsevier Science B.V. All rights reserved. Keywords: Pb isotopes; Techniques; Statistics 1. Introduction In the determination of radiogenic isotope ratios, particularly within the earth sciences, a correction is usually made for the mass fractionation effects en- countered during thermal ionization mass spectrome- try by normalisation to an accepted ratio of two nonradiogenic isotopes. Thus, for example, in the case of Sr, the measured 86 Srr 88 Sr ratio is compared 86 88 Ž . to the known constant Srr Sr s 0.1194 . This measure of fractionation is then employed to correct the other ratios of interest. However, of the naturally occurring Pb isotopes, only 204 Pb can be considered as nonradiogenic and therefore this method cannot be used. Thus, while ) Corresponding author. the current generation of multicollector mass spec- trometers can measure Sr and Nd isotope ratios with Ž unparalleled levels of precision and accuracy e.g., . Thirlwall, 1991 , the inability to make adequate cor- rections for instrumental mass fractionation has, in many ways, hindered the further development of Pb isotope applications. The usual approach adopted is to determine a mass fractionation factor by analysis of a Pb standard with well-known isotopic composi- Ž . tion e.g., SRM 981 and apply this same correction factor to all unknowns. Unfortunately, this is fraught with problems since standards and real rock samples generally will not run under identical conditions of mass fractionation, and indeed suites of geologically similar samples may exhibit considerable differences in their fractionation behaviour, depending upon fac- tors which are often beyond the control of the ana- lyst, for example the presence of trace impurities in the sample which suppress ionisation. 0009-2541r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. Ž . PII: S0009-2541 98 00023-0