Isotopic abundance measurements of K and Ca using a wide-dispersion multi-collector mass spectrometer and low-fractionation ionisation techniques

Procedures have been established for routine high-precision isotopic analyses of all aspects of the K-Ca isotopic system, for virtually any natural sample, using a set of K-40 + Ca-43 + Ca-46 spikes and a VG354 (Multi-II) mass spectrometer. Loading samples as KI and CaI2 on triple (or double) filament arrays results in reduced mass fractionation, approximately as expected from Rayleigh distillation theory. For K, this improves the fractionation (or K-41 anomaly) detection limit to similar to 0.8 parts per thousand/amu for single analyses. For Ca, the reduced fractionation renders data insensitive to the choice of fractionation law used for internal corrections. The 2 sigma reproducibility of normalised Ca-40/Ca-44 data (expressed as epsilon(Ca)) is similar to 4, approximately twice the theoretical limits, but for measurements of natural fractionation (delta(44/42)Ca) it is similar to 0.4 parts per thousand, about five times theoretical. These measurement procedures yield an atomic weight for the Ca standard which is consistent with the IUPAC value, and provide strong support for a reduced uncertainty (by a factor of two). A value of 40.078 +/- 0.002 would include all known isotopic fractionations in natural and commercial materials. Analyses of four Nerita marine mollusc samples from a wide geographic (and water temperature) range failed to detect any thermally controlled Ca isotope fractionation (delta(44/42)Ca) corresponding to the observed delta(18)O variations. (C) 1997 Elsevier Science B.V.