HR: 1330h
AN: V42D-0386    [PDF]
TI: High Precision Pb and Tl Isotope Measurements by MC-ICP-MS: Problems and Solutions
AU: * Kamenov, G
EM: kamenov@ufl.edu
AF: Univ of Florida, PO Box 112120, Gainesville, FL 32611
AU: Mueller, P
EM: mueller@geology.ufl.edu
AF: Univ of Florida, PO Box 112120, Gainesville, FL 32611
AU: Perfit, M
EM: perfit@geology.ufl.edu
AF: Univ of Florida, PO Box 112120, Gainesville, FL 32611
AU: Heatherington, A
EM: aheath@geology.ufl.edu
AF: Univ of Florida, PO Box 112120, Gainesville, FL 32611
AB: Measurements of NBS 981 Pb mixed with Tl for mass bias corrections were conducted on a MC-ICP-MS (Nu Plasma) during the period of April-August, 2003. During the initial stages of the experiments, Pb and Tl from single concentrated stock solutions were mixed in 2%HNO3 and then measured from 48 hours to 60 days after preparation (mixing). Initial results (n=28) with these solutions revealed relatively poor precision and accuracy with 206Pb/204Pb=16.921 (+/-0.039 2s), 207Pb/204Pb=15.469 (+/-0.052 2s), and 208Pb/204Pb=36.630 (+/-0.160 2s). Similarly, using NBS 981 for mass bias correction of the 205Tl/203Tl yielded large epsilon Tl variations (-4 to +31). Comparison of measured Pb and Tl beam intensities for these solutions over this period showed higher Pb:Tl intensity ratios compared to the known elemental ratios in the original mixtures. In addition, the longer the mixed solutions aged, the greater the Pb:Tl decoupling was, and poorer precision and accuracy in the isotope measurements were observed. To address this problem, we prepared new mixed solutions and analyzed them immediately ($<1$ hour) after mixing. Analyses (n=26) showed excellent precision and accuracy with 206Pb/204Pb=16.937 (+/-0.001 2s), 207Pb/204Pb=15.491 (+/-0.001 2s), and 208Pb/204Pb=36.694 (+/-0.004 2s). Measured epsilon Tl in these fresh solutions was 1.5 (+/-0.6 2s). In addition, measured Pb and Tl ion beam intensities always reflected those calculated for the mixtures. Poor precision and accuracy in the isotope measurements and decoupling of Pb and Tl elemental abundances vs. ion beam intensities were observed only when the aged solutions were aspirated through the DSN. The same solutions aspirated in wet plasma mode did not show variations in the measured Pb:Tl ratios, suggesting that precipitation of Tl in the PFA bottles cannot explain the decrease in the Tl signal. Thallium commonly exists as Tl+ in solution, but at low pH-high Eh it can be oxidized to the less soluble form Tl+3. If this oxidation leads to the formation of complex Tl molecules (colloids), then during wet plasma mode the colloids will be introduced into the plasma together with the rest of the analyte, so no decoupling between measured Pb and Tl abundances and intensities is observed. If, however, the colloids behave differently than the rest of the analyte ions in the DSN during the liquid-vapor transition, that may account for the observed elemental fractionations and subsequent inapplicability of Tl isotope ratios for mass bias corrections for Pb. Measured intensities in single Pb and Tl control solutions does not show changes over time suggesting that the reactions occur only in mixed solutions. In addition, experiments currently under way with solutions with different HNO3 concentrations suggest that the interaction between Pb and Tl occurs preferentially at certain pH-Eh range. This study demonstrates that highly precise Pb isotopic measurements can be produced using Tl normalization with MC-ICP-MS. However, significant analytical errors, resulting in poor precision and accuracy in the Pb isotope measurements and up to 30-40 epsilon Tl units variation, can be generated during aging of mixed solutions as a result of complex interactions between the analyte ions. Experiments utilizing thallium for mass bias corrections for Pb, and lead for mass bias corrections for Tl isotope measurements and desolvation are most reproducible when conducted on freshly prepared analyte mixtures.
DE: 1040 Isotopic composition/chemistry
DE: 1094 Instruments and techniques
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2003 Fall Meeting