A comprehensive evaluation of sulfur isotopic analysis (δ34S and δ33S) using multi-collector ICP-MS with characterization of reference materials of geological and biological origin

Sulfur isotope ratios are often used as biogeochemical tracers to gain understanding of abiotic and biological processes involved in the sulfur cycle in both modern and ancient environments. There is however a lack of matrix-matched well-characterized isotopic reference materials that are essential for controlling the accuracy and precision. This study therefore focused on expanding and complementing the currently available sulfur isotope ratio data by providing the bulk sulfur isotopic composition, as determined using multi-collector inductively coupled plasma-mass spectrometry (MC-ICP-MS), for a comprehensive set of commercially and/or readily available biological and geological reference materials. A total 7 isotopic reference materials and 41 elemental reference materials were studied. These reference materials include standards of terrestrial and marine animal origin, terrestrial plant origin, human origin, and geological origin. Different sample preparation protocols, including digestion and subsequent chromatographic isolation of S, were evaluated and the optimum approach selected for each matrix type. For achieving enhanced robustness, the sample preparation and sulfur isotope ratio measurements were done at two different laboratories for selected reference materials, while at one of the laboratories the measurements were additionally performed using two different MC-ICP-MS instruments. Determined δ34SVCDT and δ33SVCDT values compared well between the different laboratories, as well as between the different generation MC-ICP-MS instruments, and for standards that were previously characterized, our data are similar to literature values. The δ34SVCDT ranges determined for the different categories of the reference materials – terrestrial animal origin: +2 to +9‰, marine animal origin: +15 to +20‰, human origin: +6 to +10‰, terrestrial plant origin: −20 to +7‰, and geological origin: −12 to +21‰ – fit the expected values based on previous studies of similar types of matrices well. No significant mass-independent fractionation is observed when considering the expanded uncertainties for Δ33SV-CDT.