Microstructures of olivine-plagioclase corona in meta-ultramafic rocks from Sefuri Mountains, NW Kyushu, Japan

Abstract

Two types of reaction rims occur between olivine and plagioclase in ultramafic rocks from the Sefuri Mountains, NW Kyushu, Japan, which were metamorphosed under granulite-facies conditions. One occurs as a thin film of orthopyroxene along the boundary between olivine and plagioclase (orthopyroxene zone). The other is composed of two zones: symplectite of calcic amphibole and spinel on the plagioclase-side (symplectite zone) and calcic amphibole with sporadic orthopyroxene on the olivine-side (tremolite zone). In the tremolite zone, calcic amphibole shows a systematic decrease in Al content and increase in Mg/(Fe +Mg) with decreasing distance from olivine. Local equilibria maintained during the diffusion-controlled corona-forming reaction enable us to apply equilibrium thermodynamics to calcic amphibole and adjacent orthopyroxene. An integrated formulation of the Gibbs method for an Fe–Mg exchange reaction constrains the equilibrium temperature recorded in the tremolite zone to be 600–710 °C. It is significantly lower than the temperature of the granulite-facies metamorphism (800–900 °C) estimated using conventional geothermobarometry. Except for H 2O, the association of calcic amphibole and spinel in the symplectite zone is chemically equivalent to the association of olivine, plagioclase and orthopyroxene that was stable before the corona formation. This suggests that the following orthopyroxene-consuming reaction describes the paragenetic change taking place between 800–900 °C and 600–710 °C, olivine + plagioclase + orthopyroxene + aqueous fluid = calcic amphibole + spinel. In contrast, the overall reaction inferred from microstructures produces orthopyroxene as well as calcic amphibole and spinel at the expense of olivine and plagioclase. This reaction requires removal mainly of MgO that is also responsible for destabilizing the local association of olivine and plagioclase. These features suggest that the presence of orthopyroxene as a product in the corona is not always indicative of an orthopyroxene-producing reaction being responsible for the change of paragenetic relation. Microstructural features should be carefully applied to infer the reaction describing paragenetic change by which we argue the P– T path of the rocks.