High-precision analysis of Pb isotope ratios by multi-collector ICP-MS

We investigated high-precision Pb isotope ratio analysis by multi-collector -inductively coupled plasma-mass spectrometry (MC-ICP-MS) using added thall ium as an internal isotopic standard to correct for mass dependent isotopic fractionation. We compared MC-ICP-MS analysis of both an inter-laboratory standard, NBS 981, and geological samples to conventional thermal ionizatio n mass spectrometry (TIMS). As expected, we found that analytical error in the latter was dominated by mass fractionation, In MC-ICP-MS, we found that fractionation appears to follow the exponential law, but that the fraction ation coefficients, f, of Tl and Ph were nor identical. This difference in fractionation coefficients cannot be compensated for by renormalizing to a different T1 isotopic composition as done in other studies. We found, howev er, that the f(Pb)/f(Tl) ratio was constant over the course of an analytica l session, allowing f(Pb) to be calculated from f(Tl). An exponential law c orrection was then applied to the Pb isotope measurements which effectively eliminates enters associated with mass fractionation, Precision for the MC -ICP-MS analyses ranged from a factor of 2 to a factor of 6 better than for TIMS analyses for the Pb-206/Pb-204 and Pb-208/Pb-206 ratios respectively. Residual error in the MC-ICP-MS analyses was dominated by error in the ana lysis of Pb-204, perhaps in part clue to random errors introduced by correc ting for a Hg-204 isobaric interference. We also found systematic errors in the MC-ICP-MS analyses compared to TIMS determinations that may be due to uneven background and collector biases in the instrument used. We found tha t these systematic errors were the same for both NBS 981 and the geological samples, so accurate correction factors could be generated from the standa rd analyses to correct the sample analyses. MC-ICP-MS has the additional ad vantages of requiring less preparative chemistry, less instrument time, and considerably less labor overall. (C) 2000 Elsevier Science B.V. All rights reserved.