Validation of clinopyroxene-garnet magnesium isotope geothermometer to constrain the peak metamorphic temperature in ultrahigh-temperature ultramafic-mafic granulites

Conventional Fe-Mg exchange geothermometers generally are not effective means to measure the peak metamorphic temperatures of granulites because of Fe-Mg diffusion during the re-equilibration during the exhumation and cooling. Therefore, it is essential to find alternative geothermometers. In this study, we investigated the magnitude of Mg isotope fractionation between the coexisting clinopyroxene and garnet in garnet pyroxenites and high-pressure mafic granulites from southern India. The clinopyroxene and garnet from the garnet pyroxenites have delta Mg-26 values of -0.04 to -0.07 parts per thousand, and -0.65 to -0.64 parts per thousand, respectively, with Delta Mg-26(Cpx-Grt) = delta Mg-26(Cpx) - delta Mg-26(Grt) = 0.62 and 0.57 parts per thousand. The delta Mg-26 values of the coexisting clinopyroxene and garnet in the high-pressure mafic granulites are 0.03 to 0.07 parts per thousand and -0.54 to -0.55 parts per thousand, respectively, with Delta Mg-26(Cpx-Grt) = 0.57 and 0.62 parts per thousand. The inter-mineral Mg isotope fractionations between the clinopyroxene and garnet of the garnet pyroxenites are similar to those of the high-pressure mafic granulites and more or less fall within the equilibrium fractionation lines at peak metamorphic temperatures. The measured peak temperatures for all four samples indicate equilibrium Mg isotope fractionations between the clinopyroxene and garnet. The clinopyroxene-garnet Mg isotope geothermometer of Li et al. (2016) yields temperatures of 994 +/- 60 degrees C and 1048 +/- 89 degrees C for the garnet pyroxenites, and 1048 +/- 89 degrees C and 994 +/- 65 degrees C for the high-pressure mafic granulites, which are slightly lower than those at both 10 and 12 kbar from the clinopyroxene-garnet Mg isotope thermometer of Huang et al. (2013). Compared with the peak metamorphic conditions from phase equilibrium modeling, the clinopyroxene-garnet Mg isotope thermometry yielded temperatures corresponding to the peak metamorphic conditions of the garnet pyroxenites and the high-pressure mafic granulites, whereas the conventional clinopyroxene-garnet Fe-Mg exchange thermometry yielded lower temperatures corresponding to the retrograde metamorphism. These results underscore the inter-mineral Fe-Mg exchange between clinopyroxene and garnet during the retrograde cooling of the ultrahigh temperature (>900 degrees C), which, however, did not disturb their Mg isotope fractionation equilibrium attained during the ultrahigh-temperature metamorphic condition. Therefore, we conclude that clinopyroxene-garnet Mg isotope thermometry, together with phase equilibrium modeling, is a valid tool to constrain the peak metamorphic temperature conditions even for ultramafic-mafic granulites that have undergone ultrahigh-temperature metamorphism.