Investigating metasomatic effects on the 187Os isotopic signature: A case study on micrometric base metal sulphides in metasomatised peridotite from the Letlhakane kimberlite (Botswana)

Abstract

The peridotite xenoliths of the Letlhakane kimberlite (Botswana), which intrude the Proterozoic Magondi Belt on the western margin of the Zimbabwe craton, represent highly depleted melting residues. These residues suffered subsequent variable metasomatic overprinting, evidenced by cryptic trace element enrichments in the spinel peridotites to modal addition of phlogopite, clinopyroxene and spinel within the garnet peridotites. In order to assess the robustness of the Re–Os chronometer in such highly metasomatised peridotites, detailed investigations of base metal sulphide (BMS) petrography and single-BMS grain 187Os/188Os analyses have been undertaken in three representative peridotites.The BMS occur as <10μm–50μm inclusions and interstitial grains that are associated with metasomatic phases or metasomatised rims of primary silicates or display melt-like morphology, all attesting of their metasomatic origin. Their 187Os isotopic compositions vary from 0.1016 to 0.6109 yielding TRD ages from 3.75±0.54 (2se) to future ones. They vary independently of the cryptic or modal silicate metasomatic overprinting on the peridotites and independently of the BMS-silicate textural habits (e.g. isolated inclusions, pseudo-inclusions, intergranular melt-like pools), contrary to what is commonly assumed. In such highly depleted peridotites that must have been sulphide-free after the partial melting event, the Eoarchean age is likely inherited from residual PGM (platinum group minerals; i.e. laurite and Os-alloys) that formed in response to the exhaustion of the primary BMS and were later redissolved within the metasomatic BMS. In contrast, the younger single grain TRD ages represents an increasing dilution of the residual PGM signals within the metasomatic BMS, with the single grain 187Os/188Os signatures becoming increasingly dominated by Os derived from metasomatic BMS.Taken as a whole, the single BMS grain Eoarchean age suggests a lithospheric stabilisation age in Letlhakane similar to the oldest crustal samples of the Zimbabwe craton, which thus supports the Letlhakane mantle root to be a westerly extension of the Zimbabwe cratonic mantle.The TRD ages at the single BMS grain scale show a much larger range than their respective whole-rock TRD age, most notably from BMS in the two metasomatic end members: the cryptically-metasomatised peridotite and the most modally overprinted peridotite containing BMS 1 to 2.5Ga older than their whole-rock. Base metal sulphides are thus better time capsules than the whole-rock, regardless of the silicate metasomatism the whole-rocks have experienced. Furthermore, the ancient ages are obtained from interstitial Ni-rich BMS indicating that all possible hosts of Os, not only BMS which appear to be ‘primary’ (i.e. Fe-rich BMS isolated inclusions in olivine) should be investigated. Single BMS grain Re–Os dating should then be considered as the only reliable dating approach, across all tectonic contexts (orogenic, ophiolitic, cratonic), which have experienced extensive metasomatic overprinting or refertilisation.