An oxygen fugacity profile through the Siberian Craton - Fe K-edge XANES determinations of Fe 3+ ∑Fe in garnets in peridotite xenoliths from the Udachnaya East kimberlite

Abstract

The Udachnaya East kimberlite sampled garnet peridotite xenoliths from a pressure range of 1.2 to 7.1GPa in the underlying Siberian cratonic lithosphere. Samples derived from <5.2GPa lie close to a typical cratonic geotherm of 40mWm-2, whereas more deeply derived samples have temperatures ≥100°C above this geotherm. Minor and trace element compositions of garnet and clinopyroxene indicate the presence of both depleted and metasomatically enriched material in the suite. Depleted material derives from the entire sampled depth interval, but enriched material is confined to pressures of 4.5 to 6.6GPa. Thus, the Siberian cratonic lithosphere under the Undachnaya pipe consisted of a relatively cool and depleted upper layer about 150km deep, underlain by a hotter layer which extended to at least 210km depth and contained both depleted and enriched material. Fe K-edge XANES was applied to garnets from this suite to measure their Fe3+/∑Fe values, enabling determination of a redox profile through the lithospheric section represented by the xenolith suite. δlogO2[FMQ] varied from -2.5 to nearly -6.0 over the sampled pressure interval. An overall trend to lower δlogO2[FMQ] values with increasing pressure was defined mostly by the depleted samples. A superimposed oxidation trend to δlogO2[FMQ] values 1-2 units higher than the main trend mostly affected the deeper, enriched samples, indicating a clear link between metasomatism and oxidation. The amount of oxidation was insufficient to de-stabilize diamond in the deep lithosphere. A possible mechanism for metasomatic enrichment relates to localized, low degree "redox melting", whereby upwardly percolating CH4±H2O fluids would encounter progressively more oxidizing peridotite wall-rock resulting in diamond crystallization and increased water activity in the fluid. This could lead to local partial melting and enriched melts could migrate into cooler parts of the lithosphere and crystallize, thus enriching parts of the lithosphere. Melts thus formed are expected to be relatively enriched in Fe3+ as it is moderately incompatible during partial melting. Lithospheric domains metasomatised by solidification of these melts would be relatively enriched in Fe3+ and garnets may therefore have higher Fe3+/∑Fe values, thus recording relatively higher δlogO2[FMQ] values.