Determination of Cu and Mo Isotope Fractionations in Natural Materials Using MC-ICP-MS

Abstract

Our aim is to determine the deviations (8) in the isotopic compositions of Cu and Mo between samples and standards, using the VG Plasma54 multiple-collector inductively-coupled plasma mass spectrometer (MC-ICP-MS) at the University of Rochester. MC-ICP-MS systems are well-suited to this research because they feature high ionization efficiency for all transition metals, simple sample introduction, simultaneous collection of multiple ion beams for precise determination of isotope ratios, and no fractionation of the sample reservoir during analysis (Halliday et al., 1995). As a result, the largest source of uncertainty in isotopic analyses by MC-ICP-MS is the variability of instrumental mass bias with time. We can correct for this effect if we take advantage of another unique characteristic of MC-ICP-MS: The minimal variation in instrumental mass bias among different elements over a limited mass range (Hirata, 1996). As a result, if samples and standards are 'spiked' with an element of similar mass to the element of interest, the magnitude of mass bias, ~, can be determined by comparing the measured isotopic composition of the spike element to the 'true' value. The measured composition of the element of interest is then corrected using cz and the appropriate mass fractionation law. This 'elemental spike' method, analogous to use of an isotopic 'double spike', has proven useful in precise and accurate determination of Pb isotopic compositions (Longerich et al., 1987; Hirata et al., 1996). For fractionation studies, the use of elemental spikes avoids some of the complexities of a double spike. It also facilitates isotopic analysis of elements with fewer than four stable isotopes. Significantly, 8 values do not depend on cc being identical for the element of interest (cci) and the spike element (%). Although differences in ~z produce inaccuracies in isotope ratio determinations, such errors cancel in the 8 notation as long as ~i/CCs is constant. Similarly, inaccuracy in the 'true' value of the spike element will not affect 8, as long as the isotopic composition of the spike is identical in samples and standards.