Matrix Effects on the MC-ICPMS Analysis of Zn and Cd Isotopes

In order to understand fractionation in heavy stable isotope systems, matrix and auto-matrix effects need to be carefully monitored as they have a major impact on the precision and accuracy of isotope measurement by MC-ICP-MS. Understanding their impact for any given isotopic system is an important aspect of method development. In this study we look at auto-matrix and matrix effects for Zn (Cu mass bias correction) and Cd (Ag mass bias correction). Zn and Cd have similarly high ionization potentials, 9.4 and 9.0 respectively, and ionization potentials for Cu and Ag are similar, 7.7 and 7.6 respectively, making for an interesting comparison. All experiments were performed on a Nu Plasma MC-ICP-MS with 5x molar concentration of matrix elements. Experiments were also performed at concentrations more representative of those likely to be found in purified samples. In addition, we tested auto-matrix effects for Cd and Zn by varying the concentration of the analyte and mass bias correcting element. Cd sensitivity is enhanced in the presence of Rb and Pb, and is suppressed in the presence of Al, Sr and Cs. Spiked Cd runs are isotopically lighter than un-spiked runs, as light as 0.4‰/amu. Ag sensitivity was enhanced by all matrix elements. Cu sensitivity is enhanced by Al, Sr and Ba, and Zn sensitivity is enhanced by the latter two and Pb. Zn spiked with Al, Sr and Pb, or with poorly matched Cu intensities is isotopically lighter than un-spiked standards, as light as 1.0‰/amu. Auto-matrix effects for Cd and Ag, if present, are within the analytical error. We demonstrate that the presence of matrix elements or differences in concentration of analyte and mass bias correcting element between sample and standard can produce fractionation that is not true. The matrix effect is also found to be variable, as replicate analyses do not always agree. Cd and Zn react differently to different matrix elements, demonstrating the need for independent evaluation of each isotopic system. These experiments highlight the need to isolate the analyte before analysis and match the analyte and mass bias correcting element intensities in samples and bracketing standards. This is especially valid for Zn and Cu, which are more susceptible to dramatic matrix and auto-matrix effects.