Mass discrimination during MC-ICPMS isotopic ratio measurements: investigation by means of synthetic isotopic mixtures (IRMM-007 series) and application to the calibration of natural-like zinc materials (including IRMM-3702 and IRMM-651)

Abstract

Calibration of mass spectrometry (MS) isotope ratio measurement results by means of gravimetrically prepared isotopic mixtures is a long known way of anchoring isotopic values to the SI system. Thermal ionisation (TI) MS is the technique traditionally associated with this strategy, while multi-collector double focusing inductively coupled plasma (MC-ICP) ? MS, with a flexible ion source operated at atmospheric pressure and the possibility to achieve 0.01% or less repeatability on isotope ratio measurements, now appears to be an attractive alternative. However, this absolute calibration strategy necessitates that mass discrimination effects remain invariant in time (at least during the duration of all the experiments) and across the range of isotope ratios measured, which is not the case with MC-ICPMS measurements. This issue is illustrated here with Zn isotopic measurements obtained using locally produced synthetic Zn isotope mixtures. The necessary adjustments to specific characteristics of MC-ICPMS are described. Firstly, variations in the mass discrimination effects across the measurement sequence are propagated as an uncertainty component. Secondly, linear proportionality during each individual measurement between normalised mass discrimination and the average mass of the isotope ratios is used to evaluate mass discrimination for the ratios involving 66Zn and 70Zn. Thirdly, linear proportionality after iterative calculations between mass discrimination and the isotope ratio values for n(67Zn)/n(64Zn) and n(68Zn)/n(64Zn) in the mixtures is used to evaluate mass discrimination for the same ratios but specific to the isotopically enriched materials, and additional iterations are introduced until the last significant digit in the isotope ratios in the mixtures is stabilised. Fourthly, ratios in final unknown materials (various natural Zn solutions, including IRMM-3702 and IRMM-651) are calibrated by external bracketing using the isotopic mixtures.The relative expanded uncertainty (k = 2) estimated for n(68Zn)/n(64Zn) and n(67Zn)/n(64Zn) ratio values in the synthetic isotopic mixtures and the natural zinc samples was in the range of 0.034 - 0.048%. The uncertainty on the weighing (r.s.u. = 0.01%, k = 1) was by far the major contributor to the total budget, followed by the uncertainty on the correction for the impurity content in the enriched materials. The effect of the repeatability on isotope ratio measurements was minimal (=2% of the total uncertainty budget). This work was further validated from the agreement achieved for the same set of samples between these results and those obtained with a single detector TIMS (much larger uncertainty estimates) and with another MC-ICP-MS.The absolute isotope ratio values found for the candidate isotopic certified reference material IRMM-3702 ? also proposed as ?Delta 0? for delta-scale isotopic measurements ? were n(66Zn)/n(64Zn) = 0.56397(30), n(67Zn)/n(64Zn) = 0.082166(36), n(68Zn)/n(64Zn) = 0.37519(13) and n(70Zn)/n(64Zn) = 0.012422(24). The derived Zn atomic weight value Ar(Zn) = 65.37778 (22) differs significantly from the current IUPAC value by Chang et al.[ ]. Re-measurement with the IRMM-007 series of synthetic isotopic mixtures of the Zn isotope ratios in the natural material measured by Chang et al. [1] have revealed also large systematic differences (1.73% to 5.6%) that suggest unrecognized measurement biases in their results.