PARTIAL MELTING AND PHASE-RELATIONS IN HIGH-GRADE METAPELITES - AN EXPERIMENTAL PETROGENETIC GRID IN THE KFMASH SYSTEM

Petrogenetic grids are a powerful tool for understanding metamorphic t errains and many theoretical grids have been suggested for the process of granulite formation in metapelitic rocks, via fluid-absent biotite melting reactions. However, application of these grids has been diffi cult due to the lack of suitable experimental constraints. We present here an experimentally determined and tightly constrained petrogenetic grid for KFMASH system metapelites which extends from 840-1000 degree s C and 5.0-12.5 kbar. Sixty Symbols four experiments on three KFMASH, mineral-mix, bulk compositions (X(Mg) = 0.62,0.74,0.86) provide bt ph ase composition and assemblage data from which a grid can be derived a nd constrained. Reversal experiments and consideration of the phase co mposition data show the experiments to be close to equilibrium. The KF MASH univariant fluid-absent biotite melting reactions occur between 8 50 and 870 degrees C at 5 kbar and between 900 and 915 degrees C at 10 kbar. These reactions are connected to equilibria beyond the stabilit y of biotite to develop a fixed framework within which the phase assem blage evolution of metapelitic rocks can be interpreted. The effect of minor components on phase equilibria is evaluated using the experimen tally determined grid as a simple-system reference. The temperature at which melting occurs in metapelites is strongly controlled by the con centrations of titanium and fluorine in biotite. Pressure-temperature pseudosections presented for each of the experimental compositions sho w both the univariant and divariant reactions available to a particula r bulk composition, clearly illustrating the possible evolution of the phase assemblage. The pseudosections also constrain the stability lim its of many distinctive assemblages, which can provide tight pressure- temperature estimates for natural rocks. These pseudosections, in conj unction with previous work, show that the bulk X(Mg) of a rock will co ntrol the amount of melt produced for a given rise in temperature.