EXPERIMENTAL-EVIDENCE AGAINST CO2-PROMOTED DEEP CRUSTAL MELTING

THERE has been considerable controversy regarding the role of carbon dioxide in the metamorphism and melting of the Earth's crust, fuelled principally by reports of experimental evidence for CO2-enhanced partial melting of the assemblage phlogopite mica + quartz in the model system KAlO2-MgO-SiO2-H2O-CO2(1,2). The analogous process in the crust would lead to the formation of granulite-facies rocks at temperatures characteristic of amphibolite-facies metamorphism; the granulite facies3-5 would thus be indicative not of unusually high temperatures, but of the influx of carbonic fluids''. Moreover, the similarity of their low-temperature melts to the compositions of some calc-alkaline lamprophyres led Peterson and Newton2 to suggest that such magmas might be generated by the introduction of CO2-rich fluids, providing a means of large-scale mass transfer in the deep crust. These ideas have been highly influential6, but there are aspects of Peterson and Newton's results that are hard to reconcile with other available data. I have accordingly tried to replicate the results of refs 1 and 2 but, as I report here, have been unable to do so. The present results thus cast doubt on the general idea of melt fluxing by CO2, and the formation of magnesium- and CO2-rich magmas at crustal pressures.