Mass-dependent and mass-independent sulfur isotope fractionation (δ34S and δ33S) from Brazilian Archean and Proterozoic sulfide deposits by laser ablation multi-collector ICP-MS

Mass-independent sulfur isotope fractionation (MIF) has been observed in rocks of the geological record older than about 2.45 Ga, a characteristic which is thought to be related to processes in the Neoarchean atmosphere. Samples recording a MIF effect therefore have to contain sulfur of the exogenic sulfur cycle, while endogenic sulfur should not show this effect. The sulfides analysed from six Brazilian deposits represent either exogenic or endogenic sulfur sources, with supposed ages ranging from about 1.9 to 2.7 Ga. Sulfur isotopes were analysed by in‐situ laser ablation MC-ICP-MS. A range of experiments were conducted using international and in-house isotope standards, which were run under various conditions and set-ups. These include the reference materials IAEA-S1, IAEA-S3, NBS123 (sphalerite), NBS127 (barite), and in-house standards BSB-py (pyrite) and BSB-cpy (chalcopyrite). During six days of analysis, an internal precision of sulfide analyses of 0.10–0.15‰ (1 s) for δ34S and 0.40–0.60‰ (1 s) for δ33S, and an accuracy of ~ 0.30‰ for δ34S and δ33S was achieved. The standard measurements define a δ33S/δ34S relationship of δ33S = 0.513 * δ34S + 0.149, with R² = 0.9997, which is close to the theoretical relationship for mass-dependent fractionation. The by far best analytical errors were obtained for natural pyrite, reaching a within-run precision of about 0.05–0.15‰ for δ34S (1 s), and 0.10–0.15‰ (1 s) for δ33S determinations. Including all precision and accuracy data, we arrive at the following 1 s error limits to which sulfur isotope analyses for MIF studies obtained with this method are reliable: 0.32, 0.34 and 0.46‰ for pyrite (δ34S, δ33S and Δ33S determinations, respectively), 0.34, 0.42 and 0.54‰ for pyrrhotite, and 0.34, 0.50 and 0.58‰ for chalcopyrite.Pyrites and pyrrhotites from 2200 Ma old sediments and volcanics associated with the Pilar greenstone belt, Goiás, have δ34S from − 15.0 to + 2.0‰, and δ33S from − 8.0 to + 1.0‰, with Δ33S being slightly but not significantly positive. Pyrites from supposed ca. 2060 Ma old clastic sediments of the Águas Claras Formation (Carajás, Pará), have δ34S = 9.5–12.5‰ and δ33S = 5.0–6.5‰ with Δ33S close to zero. Chalcopyrites from the Salobo and Sequerinha sulfide deposits, also Carajás, are genetically related to 2570 Ma old A-type alkaline granites with endogenic sources. They have δ34S = 1.6–2.9‰ and δ33S = 1.0–1.5‰, and δ34S = 0.2–0.9‰ and δ33S = 0.65‰, respectively, with Δ33S ranging from − 0.15 to 0.56. The only Δ33S values of this study which clearly suggest sulfur isotope MIF come from the mineralized banded iron-formations of the Quadrilátero Ferrífero, Minas Gerais. Pyrites from the Mina de Cuiabá deposit (2700 Ma) have δ34S = 3.8–5.1‰ and δ33S = 2.9–4.8‰ with Δ33S = 0.9–2.2‰, while pyrites and chalcopyrites from São Bento (2720 Ma) have δ34S = 1.6–3.2‰ and δ33S = 1.5–2.8‰ with Δ33S = 0.2–1.5‰. The sulfides represent an epigenetic mineralization, derived from supracrustal rocks making part of an exogenic sulfur cycle. The study shows, that in-situ laser ablation ICP-MS is sufficiently sensitive for detecting MIF effects in Neoarchean rocks. Whether or not such samples show MIF-related Δ33S values depends on the age and provenance of the sulfur.

► Laser ablation ICP-MS may provide sufficiently precise and accurate δ34S, δ33S and Δ33S data.
► The analytical reliability for pyrite is: 0.32‰ for δ34S, 0.34‰ for δ33S and 0.46‰ for Δ33S (1 s SD).
► The detection of MIF depends on the source and the age (> 2.45 Ga) of the sulfur.
► Sulfur MIF was detected in Brazilian Neo-Archean iron formations from Minas Gerais.
► No MIF was detected in other Brazilian samples, either due to their age or the source of their sulfur.