Early Archean sedimentary sulfates (barites) from Africa, Australia, and India are only slightly enriched in the heavy isotope of sulfur (³⁴S) in comparison with contemporaneous sulfides. The small separation of oxidized and reduced reservoirs has long been used as evidence against the activity of sulfate-reducing bacteria prior to about 2.5-2.7 Ga ago. This view has been challenged recently with the discovery of sulfides fractionated by as much as 22‰ from associated sulfates (Shen et al., Nature, March 1, 2001).

We measured sulfur isotope (³²S, ³³S, ³⁴S) compositions of (1) sedimentary and vein barites; (2) pyrite intimately associated with barite; and (3) pyrite disseminated in black cherts, all from the 3.5 Ga-old Dresser Formation, North Pole area, Western Australia, using multi-collector ion microprobe as well as more conventional methods. Microscopic pyrites from growth laminae within vein barite have the same mass-independent isotopic depletion as the barite (D³³S=–1‰) but are about 15‰ lighter in d³⁴S compared to sulfates. This suggests in situ secondary mass-fractionation by hydrothermal processes. Disseminated pyrites in black cherts, on the other hand, commonly exhibit large positive mass-independent effects (D³³S ~ +4‰) and very little depletion in ³⁴S. These results support the suggestion that the oxidized and reduced sulfur cycles were decoupled during the Archean (Farquhar et al., Science, August 4, 2000). They also argue against the notion that disseminated pyrite in Archean black shales resulted from the bacterial reduction of seawater sulfate. Instead, the pyrite seems to have been derived from the reduction of elemental sulfur formed, at least in part, by photochemical processes in the anoxic Archean atmosphere.