Simple robust estimation of uranium isotope ratios in individual particles from LA-ICP-MS measurements

Abstract

Integration of transient signals is a commonly applied supervised approach to estimating isotope ratios from single-particle LA-ICP-MS measurements. Spectra collected on individual particles suffer from relatively weak signal strength and short duration, both of which present challenges to the integration method. Moreover, since numerous particles are measured in an analytical run, analyst supervision is required for every individual particle to estimate isotope ratios by the integration method (e.g. specification of time window containing background and specification of time window for integration of the signal for every particle spectrum). In this paper we present a simple, unsupervised, robust alternative for estimating isotope ratios from LA-ICP-MS spectra, which can be rapidly applied to a large number of particle measurements. The approach relies on point-by-point data reduction followed by application robust statistical estimation methods.