Experimental constraints on the nature of multiphase solid inclusions and their bearing on mantle wedge metasomatism, Bohemian Massif

Abstract

<p>Fluids are the primary agents for mass transfer in subduction zones. These fluids can be captured as primary inclusions within minerals crystallizing during subduction processes. Some of these inclusions, referred to as multiphase solid inclusions (MSI), are characterized by the high proportion and variety of minerals, hence by a high concentration of solute in the trapped fluid. Kotková et al. (2021) have described primary MSI in garnets of subduction-related ultra-high pressure (UHP) peridotites (P-T of 1030-1150 ºC/3.6-4.8 GPa) of the Bohemian Massif. MSI range in size between ≈5-40 µm and are mostly composed of hornblende, the barian mica kinoshitalite, dolomite and magnesite. MSI have been interpreted as trapped residual liquids produced after partial UHP crystallization of carbonate-silicate melts that now form garnet pyroxenite veins in the peridotites. Experimental re-melting of MSI is the best procedure to investigate the precise nature of trapped fluids. We have conducted re-melting experiments of MSI present in garnets of a lherzolite, taking the inclusions to P-T around their entrapment conditions at or close to host rock peak P-T, 4-4.5 GPa and 1000-1225 ºC. The inclusions (re-)crystallized into a garnet fringe at the boundary between inclusions and host garnet, barian mica and carbonatite melt towards the center of the inclusion, and a large irregular and empty space in between the garnet fringe and the central silicate-carbonate component. Microstructures and mass balance indicate that the empty space was occupied by a Na-K-Cl-F-rich saline aqueous fluid (brine). Hence experiments did not produce a single melt at any experimental conditions, but systematically show the stability and coexistence of barian mica + carbonatite melt + brine at the entrapment conditions, and a garnet fringe indicating reaction between trapped fluids and host garnet. This suggest that growing garnet trapped a carbonatite melt and a saline aqueous fluid coexisting in the matrix, together with solid crystals of barian mica likely produced by metasomatism of the percolating fluids through the host peridotite. It is intriguing, however, that neither single mica crystals nor separate former carbonate melt and brine have been found included in garnets. Mass balance shows that carbonate melt is the main host for incompatible elements such as Ba. This presentation will discuss the bearings of the experimental results on the nature and origin of these MSI, potential links to diamond formation and their implication on mass transfer processes in subduction zones.</p><p>Kotkova et al. (2021) Lithos 398-399, 106309</p>