Silicon isotopic evidence for post-Archean silica enrichment of the cratonic mantle lithosphere

The Archean cratonic mantle formed as residues of extensive melt extraction, which is widely (but not universally) thought to have occurred in oceanic settings before being subducted and involved in the growth of the early continents. Despite its depleted nature, cratonic mantle peridotites, particularly from the Kaapvaal craton of southern Africa, often show geochemical and mineralogical evidence for intense secondary silica addition. Silica addition to the peridotitic cratonic mantle has been suggested to result from metasomatism by eclogite-derived silicic melts in subduction channels, serpentinization of oceanic protoliths, or, conversely, unrelated to oceanic protoliths and subduction tectonics, and results from interaction of the cratonic mantle with rising silica-enriched mantle-derived melts. 

Here we use the Si-O isotopic compositions of cratonic mantle peridotites to constrain the source of silica enrichment, and thus improve understanding of processes that operated during the formation of Earth's first continental lithosphere. Mineral separates from coarse, low-T (<1000°C), orthopyroxene-enriched peridotite xenoliths from the Kaapvaal craton (South Africa) have δ30Si values from -0.56 to +0.40 ‰ and δ¹⁸O from +3.7 to +5.6 ‰. Silica-enriched peridotites with mantle-like δ¹⁸O-δ³⁰Si indicate silica addition did not manifest in any isotopic change, in contrast to peridotites with high δ³⁰Si at low δ¹⁸O. Rising mantle-derived silica-enriched melts (formed by hydrous fluxing of harzburgite or wall rock assimilation) could be the culprits of silica enrichment, based on recent oxygen isotope and geochemical studies of lithospheric mantle peridotites, which reconciles with the mantle-like Si-O isotopic signatures observed here. Post-3.8 Ga granitoids are characterized by elevated δ³⁰Si, interpreted to be sourced from subduction-recycled Archean cherts with universally high δ³⁰Si. Since Archean siliceous sediments are generally characterized by δ¹⁸O>>5‰, this is likely not a feasible method to produce the elevated δ³⁰Si observed here. However, Precambrian ocean waters, with higher Si contents and δ³⁰Si > 0‰ could instead have facilitated the high δ³⁰Si with low δ¹⁸O observed for some Kaapvaal peridotites. However, both Si and O isotope disequilibrium observed in some of our samples raises questions regarding the timing of SiO₂-addition, suggesting that the Si-addition, and isotopic signatures, may be a post-Arcehan feature related to Proterozoic subduction-driven metasomatism.