Collision course; high-precision mass-independent and mass-dependent calcium isotope measurements using the prototype collision cell MC-ICPMS/MS, Proteus

This paper reports measurements of mass-independent and mass-dependent Ca isotope variations using the prototype collision cell MC-ICPMS/MS, Proteus. Helium and hydrogen gases are introduced into the collision cell and the well characterised resonant charge exchange and proton transfer reactions between Ar+ ions and H2 molecules results in almost complete removal of the 40Ar+ ions allowing the measurement of the 40Ca isotope by MC-ICPMS without resorting to 'cold' plasma. A range of reference materials have been characterised for both their mass-independent and mass-dependent Ca isotope compositions relative to the international reference material NIST SRM 915a. Notably, the radiogenic 40Ca/44Ca isotope ratio (internally normalised to 42Ca/44Ca) can be accurately measured with typical reproducibility better than 40 ppm (pooled 2 SD) for a single measurement using 12 μg of Ca. Reference materials USGS GSP‐2 and CRPG FKN have notably radiogenic 40Ca/44Ca, with values of 434 ± 7 and 5513 ± 17 ppm respectively relative to NIST SRM 915a (pooled 2 SE, n = 35 and 5 respectively). In contrast, multiple repeat measurements of mantle derived reference materials BIR, BHVO-2, BCR-2 and JP-1 yield values that are slightly unradiogenic (mean − 23 ± 5 ppm, pooled 2 SE, n = 71) relative to NIST SRM 915a, bone ash and seawater reference materials, likely reflecting crustal components in the latter. Coupled measurements of the minor 48Ca and 43Ca isotopes give 48Ca/44Ca and 43Ca/44Ca (internally normalised to 42Ca/44Ca) with typical reproducibilities of ± 40 and ± 16 ppm respectively (pooled 2 SD) for single (24 μg Ca) analyses. Multiple repeats yield mean values for terrestrial reference materials that are within their pooled 2 SE uncertainties of SRM 915a, to precisions of ± 15 and ± 6 ppm respectively. Measurements of mass-dependent Ca isotope variations (δ44/40Ca) have been made using a 42Ca-43Ca double spike isotope tracer together with measurements of 40Ca and 44Ca and have typical reproducibilities of ± 0.035 ‰ (pooled 2 SD) for single (12 μg) analyses. Measured δ44/40Ca values for reference materials are within uncertainty of previously reported values. Moreover, given that the double spike inversion uses the same isotopes as the mass independent determinations of 40Ca/44Ca and 43Ca/44Ca, our documentation of the absence of anomalies at a precision of better than 20 ppm in a range of low K/Ca terrestrial reference materials implies that the accuracy of our δ44/40Ca analyses is at this level of uncertainty.