A non-primitive origin of near-chondritic S–Se–Te ratios in mantle peridotites; implications for the Earthʼs late accretionary history

Abstract

The chalcophile and highly siderophile elements Se and Te, like the other Highly Siderophile Elements (HSE) in the terrestrial mantle, may constitute powerful key tracers for meteoritic materials that hit the Earth in its latest accretionary stages (“Late Veneer”). Here the Se and Te systematics of mantle-derived peridotites (orogenic peridotites, ophiolites, cratonic peridotite xenoliths) are assessed. Combined with published in-situ analyses of HSE host minerals, whole-rock data are modelled with respect to current petrogenetic models that affect mantle composition, for example partial melting and magmatic refertilisation. We demonstrate that the near-chondritic Se/Te signature (SeN/TeN≈9±4; N = CI-chondrite normalised) of “fertile” ophiolitic and orogenic lherzolites cannot be a primitive signature of the Earthʼs mantle. This signature can however be explained by simple refertilisation models. The HSE–Se–Te budget of these fertile rocks can be modelled by mixing various proportions of a residual assemblage of Fe–Ni monosulphide solid solutions (Mss) and/or refractory platinum group minerals (PGMs – Ru–Os–Ir sulphides + Pt–Ir–Os alloys) with a metasomatic assemblage comprising low-temperature Pt–Pd–Te phases and Cu–Ni-rich sulphides. On the other hand, the reported Se and Te ratios in fertile peridotites are not consistent with melt depletion alone. Additions of late-stage metasomatic S–Se–Te–HSE-rich phases render Primitive Upper Mantle (PUM) estimates for Se and Te highly debatable, especially without appropriate consideration of refertilisation and metasomatism. Our results indicate that there is currently no firm evidence for chondritic S–Se–Te signatures in the Primitive Upper Mantle. This conclusion challenges the simplistic perception that near-chondritic Se/Te ratios may readily trace the Late Veneer composition.