Precise analysis of Ca isotope ratios (δ44/40Ca) using an optimized 43Ca–42Ca double-spike MC-TIMS method

Abstract

We used a Monte Carlo error model to optimize a 43Ca–42Ca double-spike technique for measuring Ca isotope ratios (δ44/40Ca) by Multi-Collector Thermal Ionization Mass Spectrometry (MC-TIMS). Optimization is non-unique and method-specific because errors for counting statistics and Johnson noise depend on ion beam intensities, integration times, and the total number of duty cycles, while the cumulative effects of measurement-induced Faraday collector damage restrict the double-spike 43Ca/42Ca and spike/sample ratios that can be practically employed. Factors that improve counting statistics and limit Johnson noise also accelerate collector damage and associated drift in measured δ44/40Ca values. Accordingly, better internal precision occurs at the expense of accuracy and external reproducibility. For a 20V 40Ca ion-beam implemented in a three-hop, dynamic multi-collection routine, the model predicts that a wide range of 43Ca/42Ca ratios should yield internal precisions of ±0.020–0.025‰ (2σSEM) for a runtime of 2.5h/sample, but theoretical constraints for minimizing drift exclude options having relatively high spike/sample ratios. Using a Thermo Fisher MC-TIMS (Triton), we tested 43Ca/42Ca=1 [42Ca/(42Ca+43Ca)=0.50mol/mol] and spike/sample=0.33 [Cadsp/(Cadsp+Casmp)=0.25mol/mol] by repeatedly analyzing OSIL Atlantic seawater (sw), NIST SRM 915a, NIST SRM 915b, and CaF2 over 11 sessions spanning 3 months. The average internal precision for 171 measurements is ±0.024‰, in excellent agreement with the model prediction. We adopted an experimental protocol that eliminates drift for a single measurement session (≤30 runs), thereby increasing accuracy and external reproducibility across consecutive sessions. The average δ44/40Ca values agree well with accepted values: δ44/40Casw-sw=0.000±0.005‰ (n=62), δ44/40Ca915a-sw=−1.865±0.006‰ (n=42), δ44/40Ca915b-sw=−1.134±0.006‰ (n=37), and δ44/40CaCaF2-sw=−1.392±0.008‰ (n=30). The global, long-term external reproducibility for the method is ±0.041‰ (2σSD), which represents a two- to ten-fold improvement over Ca isotope measurements made with existing double-spike MC-TIMS methods.