EXPERIMENTAL INTERACTION OF GRANITIC AND BASALTIC MAGMAS AND IMPLICATIONS FOR MAFIC ENCLAVES

Abstract

We have performed time series experiments for periods ranging from 3 min to 44 h on the interaction of granite melt and partially molten basalt at 920°C and 10 kbar, in the presence of 5 wt.% water. With time, the assemblage of the basalt domain changes from predominantly amphibole+plagioclase to clinopyroxene+garnet; the melt fraction increases from {small tilde}2?5 to 40%; and between the two domains, the melt compositions progressively equilibrate. Initially in each run, melts of the basalt domain have uniform plateau concentrations for SiO2, Al2O3, CaO, MgO, and FeO because the activities of these components are regulated by the mineral assemblage, but at advanced stages of reaction, no such control is evident. We have derived analytical expressions to describe and simulate the diffusion profiles. The concentration profiles for SiO2, Al2O3, CaO, and Na2O in the granite, emanating from the basalt–granite interface, have been used to estimate effective diffusivities. The values from the shorter runs are compared with those of the experiment of longest duration for which we assumed finite couples in our calculations. In the diffusion calculations for K2O the difference in melt fraction between the two domains is accounted for. The resulting values (in cm2/s) are: DNa2O=6 ? 10–7, DK2O=3 ? 10–7, DMgO=9 ? 10–8, DCaO=(4–6) ? 10–8, and DSiO2 and DAl2O3=(3–0?6) ? 10–8. They are in reasonable agreement with values from other studies. On the basis of our experiments we calculate that mafic enclaves of magmatic origin should equilibrate to a large degree with their host magma in slowly cooling non-convecting granitic plutons. Enclaves approaching complete re-equilibration retain distinctly higher modal amounts of mafic minerals. They do not compositionally resemble binary magma mixtures, but are more like host magma with accumulated crystals. We show that the modal differences between enclave and host are indicative of the temperature of homogenization and that, in principle, this temperature can be deduced from equilibrium phase diagrams.