Mass discrimination correction in multiple-collector plasma source mass spectrometry: an example using Cu and Zn isotopes

Abstract

Multiple-collector magnetic sector ICP-MS affords the possibility of applying instrumental mass discrimination corrections using a reference isotope ratio of an element (dopant) other than the analyte. Much attention has focused on the use of this approach for lead isotope analysis using a thallium dopant and the technique has also been applied to copper–zinc isotope analysis. The most successful applications of the doping approach have established empirical mass bias relationships between dopant and analyte isotope ratios but this often has to be done for single analytical sessions. Insufficient intra-session variation in mass discrimination often leads to poor constraints on these relationships. Moreover, with the Tl–Pb system there is some doubt over whether samples and standards exhibit the same relationship. Here we show that for the Cu–Zn system, two improvements on previous approaches lead to precise and accurate isotope ratios for unknowns. Firstly, addition of Sr to mixed Cu–Zn standard solutions generates extreme variation in mass bias so that empirical mass bias relationships between analyte and dopant are much better constrained. Secondly, we show that inadequate chemistry, specifically the inefficient removal of matrix Fe and Ti, seriously compromises the Cu–Zn isotope analysis of samples but that with clean chemistry, samples with complex matrices demonstrably yield similar mass discrimination relationships between Cu and Zn isotopes to standards. We also document previously unreported aspects of Cu–Zn isotope analysis: (1) that Cu–Zn mass bias relationships depend critically on the Cu/Zn ratio of the solution; (2) that for an introduction system with a desolvating membrane, the behaviour of standards is highly variable, perhaps due to variations in the oxidation state of Cu in the solution, and that this can be overcome by the passage of standards through the ion exchange procedure used to purify samples. Finally, we document chemical separation and mass spectrometric techniques that permit the isotopic analysis of order of magnitude smaller samples than previously achieved and report values for BCR-1 basalt standard of δ⁶⁶Zn = 0.20 ± 0.09‰ (n = 12) and δ⁶⁵Cu = 0.07 ± 0.08‰ (n = 6) at the 95% confidence level.