The capturing and decompression of melt inclusions in olivine and ilmenite megacrysts from the Monastery kimberlite, South Africa

Abstract

Ilmenite and olivine megacrysts from the 89 Ma Monastery kimberlite (Kaapvaal Craton, South Africa) captured abundant and large melt inclusions containing quenched Si-Mg-rich melt, calcite, spinel, perovskite, phlogopite, and serpentine. Textural observations and 3D X-ray tomography of ilmenite and olivine megacrysts show melt inclusion shapes, sizes and distribution patterns indicative of melt capture during primary crystal growth near the base of the subcontinental lithospheric mantle (SCLM). Patterns supporting secondary melt injection along fractures or veins, such as planar arrays of melt inclusions, are absent. Melt inclusions in olivine, in some examples reaching the dimension of centimetres, likely were captured in skeletal voids forming in fast growing, up to decimetre-sized olivine megacrysts. These large melt inclusions commonly decrepitated, forming apophyses, radial fractures, and veins, along which residual volatile- and Si-Mg-rich melt was extracted. We attribute the decrepitation of melt inclusions in olivine to the rapidly increasing difference between the melt pressure in the inclusions, captured at mantle depth, and the decreasing confining stress to which the host olivines were exposed during magma ascent and after emplacement. In ilmenite, melt inclusions up to ∼ 6 mm in diameter remained commonly intact during the kimberlite ascent from its mantle source to the shallow crust. The quenched silicate melt in olivine- and ilmenite-hosted melt inclusions, in some places preserved as unaltered hydrous and CO3-bearing glass, shows systematic major element compositional variations that suggest that this melt formed by similar fractionation and depletion processes, irrespective of the hosting megacryst phase. Apparent modal variations in quenched silicate melt, calcite, and oxide contents suggest that the melt batches captured as inclusions in ilmenite and olivine either record different evolution stages in the megacryst magma, or document compositional heterogeneities in this magma at the time of megacryst growth.