Extreme fractionation and micro-scale variation of sulphur isotopes during bacterial sulphate reduction in deep groundwater systems

This study conducted at the Äspö Hard Rock Laboratory, SE Sweden, determines the extent and mechanisms of sulphur-isotope fractionation in permanently reducing groundwater in fractured crystalline rock. Two boreholes >400 m below the ground surface were investigated. In the 17-year-old boreholes, the Al instrumentation pipes had corroded locally (i.e., Al-[oxy]hydroxides had formed) and minerals (i.e., pyrite, iron monosulphide, and calcite) had precipitated on various parts on the equipment. By chemically and isotopically comparing the precipitates on the withdrawn instrumentation and the borehole waters, we gained new insight into the dynamics of sulphate reduction, sulphide precipitation, and sulphur-isotope fractionation in deep-seated crystalline-rock settings. An astonishing feature of the pyrite is its huge variability in δ34S, which can exceed 100‰ in total (i.e., −47.2 to +53.3‰) and 60‰ over 50 μm of growth in a single crystal. The values at the low end of the range are up to 71‰ lower than measured in the dissolved sulphate in the water (20-30‰), which is larger than the maximum difference reported between sulphate and sulphide in pure-culture experiments (66‰) but within the range reported from natural sedimentary settings. Although single-step reduction seems likely, further studies are needed to rule out the effects of possible S disproportionation. The values at the high end of the range (i.e., high δ34Spy) are much higher than could be produced from the measured sulphate under any biogeochemical conditions. This strongly suggests the development of closed-system conditions near the growing pyrite, i.e., the rate of sulphate reduction exceeds the rate of sulphate diffusion in the local fluid near the pyrite, causing the local aqueous phase (and thus the forming pyrite) to become successively enriched in heavy S (34S). Consequently, the δ34S values of the forming pyrite become exceptionally high and strongly decoupled from the δ34S values of the sulphate in the bulk fluid. The Al-(oxy)hydroxide and calcite precipitates are explained by a combination of deposit and galvanic corrosion initiated by Al corrosion by H2S produced by sulphate-reducing microorganisms.