Enhanced sensitivity for Os isotope ratios by magnetic sector ICP-MS with a capacitive decoupling Pt guard electrode
You are here
Townsend, AT, Enhanced sensitivity for Os isotope ratios by magnetic sector ICP-MS with a capacitive decoupling Pt guard electrode, Fresenius J Anal Chem, 367, (7) pp. 614-620. ISSN 0937-0633 (2000) [Refereed Article]
A magnetic sector ICP-MS with enhanced sensitivity was used to measure Os isotope ratios in solutions of low Os concentration (-1 ng g-1 or less). Ratios with 192Os as the basis were determined, while the geologically useful 187Os/188Os ratio was also measured. Sample introduction was via the traditional nebuliser-spray chamber method. A capacitive decoupling Pt shield torch was developed "in-house" and was found to increase Os signals by -5 × under "moderate" plasma conditions (1050 W) over that found during normal operation (1250 W). Sensitivity using the guard electrode for 192Os was -250-350,000 counts s-1 per ng g-1 Os. For a 1 ng g-1 Os solution with no guard electrode, precisions of the order of 0.2-0.3% (189Os/192Os and 190Os/192Os) to -1% or greater (186Os/192Os, 187Os/192Os and 187Os/188Os) were found (values as 1 σ for n = 10). With the guard electrode in use, ratio precisions were found to improve to 0.2 to 0.8%. The total amount of Os used in the acquisition of this data was -2.5 ng per measurement per replicate. At the higher concentration of 10 ng g-1, precisions of the order of 0.15-0.3% were measured (for all ratios), irrespective of whether the shield torch was used. Ratio accuracy was confirmed by comparison with independently obtained NTIMS data. For both Os concentrations considered, the improvement in precision offered by the guard electrode (if any) was small in comparison to calculated theoretical values based on Poisson counting statistics, suggesting noise contributions from other sources (such as the sample introduction system, plasma flicker etc). At lower Os concentrations (to 100 pg g-1) no appreciable loss of ratio accuracy was observed, although as expected based on counting statistics, poorer precisions of the order of 0.45-3% (l σ, n = 5) were noted. Re was found to have a detrimental effect on the precision of Os ratios involving 187Os, indicating that separation of Re and Os samples is a necessary pre-requisite for highly accurate and precise Os isotope ratio measurements. © Springer-Verlag 2000.
Repository Staff Only:
item control page