Continuous-flow analyses of calcium isotopes using a Neoma MC-ICPMS/MS with H2 or SF6

Abstract

This paper reports Ca isotope ratio measurements (n = 1000) from May 2024 to October 2025 using a Neoma MC-ICPMS/MS equipped with a continuous flow sample introduction system (microFAST Isotope 2) and dry plasma. Using H₂ in the collision/reaction cell, the sensitivity ranges from ∼500 to 1000 V ppm⁻¹ in the low-resolution mode for ⁴⁰Ca with typical noise-to-signal ratios of ca. 0.2%. The associated reproducibility is 120, 130, 80, and 70 ppm (±2 SD, n = 716) on the δ^44/40Ca, δ^43/40Ca, δ^42/40Ca, and δ^44/42Ca values, respectively. Effects of concentration or matrix mismatches between the sample and standard were scrutinised over four sessions. We find that mismatch offsets, despite being expectedly linear, can have a positive or negative influence on the Ca isotope ratios, confirming that these are session-dependent rather than instrument-dependent. In all cases, the concentration mismatch is corrected using the adjustable injection flow rate of the sample introduction system. The overall accuracy of the method is acknowledged using a series of twelve certified reference materials with various biological matrices. While the Sr²⁺ double charge interferences are generally corrected up to a Sr/Ca ratio of 1–2% by monitoring ⁸⁷Sr²⁺, we, however, explore the potential of using SF₆ as a reaction gas to measure Ca isotope ratios free of these interferences. Using SF₆ in the collision/reaction cell, the sensitivity drastically drops to ca. 100 V ppm⁻¹ in the low-resolution mode with representative blank/signal ratios of ca. 0.2%. Under these conditions, the typical reproducibility is 140 and 90 ppm (±2 SD, n = 265) on the δ^44/40CaF and δ^42/40CaF values, respectively, with the ⁴³Ca¹⁹F⁺ signal being too low to be measured. We find that the ⁴²CaF/⁴⁰CaF ratio is remarkably exempt from concentration mismatch up to 30% and from Sr²⁺ double charge interferences up to a Sr/Ca ratio of 20%. However, the ⁴⁴Ca¹⁹F⁺ mass does not react in a mass-dependent manner, such that the δ^44/40CaF and δ^44/42CaF values are correlated with a slope of ca. 1 instead of ca. 2. The immunity of the ⁴²CaF/⁴⁰CaF ratio from analytical mass-biases suggests strong potential for measuring Ca isotopes with high throughput in solution or with laser ablation.