Separate zones of sulfate and sulfide release from subducted mafic oceanic crust

Abstract

Liberation of fluids during subduction of oceanic crust is thought to transfer sulfur into the overlying sub-arc mantle. However, despite the importance of sulfur cycling through magmatic arcs to climate change, magma oxidation and ore formation, there has been little investigation of the metamorphic reactions responsible for sulfur release from subducting slabs. Here, we investigate the relative stability of anhydrite (CaSO4) and pyrite (FeS2) in subducted basaltic oceanic crust, the largest contributor to the subducted sulfur budget, to place constraints on the processes controlling sulfur release. Our analysis of anhydrite stability at high pressures suggests that this mineral should dominantly dissolve into metamorphic fluids released across the transition from blueschist to eclogite facies (~450–650 °C), disappearing at lower temperatures on colder geothermal trajectories. In contrast, we suggest that sulfur release via conversion of pyrite to pyrrhotite occurs at temperatures above 750 °C. This higher temperature stability is indicated by the preservation of pyrite–bornite inclusions in coesite-bearing eclogites from the Sulu Belt in China, which reached temperatures of at least 750 °C.Thus, sulfur may be released from subducting slabs in two separate pulses; (1) varying proportions of SO2, HSO4− and H2S are released via anhydrite breakdown at the blueschist–eclogite transition, promoting oxidation of remaining silicates in some domains, and (2) H2S is released via pyrite breakdown well into the eclogite facies, which may in some circumstances coincide with slab melting or supercritical liquid generation driven by influx of serpentinite-derived fluids. These results imply that the metallogenic potential in the sub-arc mantle above the subducting slab varies as a function of subduction depth, having the greatest potential above the blueschist–eclogite transition given the association between oxidised magmas and porphyry Cu(–Au–Mo) deposits. We speculate that this zoned sulfur liberation might be one of the factors that lead to the apparently redox-influenced zoned distribution of ore deposit types in the Andean arc. Furthermore, given the lack of sulfate-associated sea floor oxidation prior to the second great oxidation event, the pattern of sulfur transfer from the slab to the sub-arc mantle likely changed over time, becoming shallower and more oxidised from the Neoproterozoic onwards.