A major challenge to modeling the partial melting process at deep mantle depths is the lack of information on the density of magmatic liquids at high pressure. The largest uncertainty is with the pressure dependence on melt compressibility (K0' = dK0/dP), and how K0' varies with melt composition. The goal of this proposal is to tightly constrain values of K0' for four silicate (NaAlSi3O8, CaTiSiO5, Na2Si2O5, K2Si2O5) and three carbonate (Li2CO3, Na2CO3, K2CO3) liquids that are highly compressible at one bar (KT,0 between 4 and 17 GPa at 1600 degrees C). These results, when combined with K0 and K0' data in the literature will provide a robust test of the hypothesis that the K0' of a silicate liquid is inversely correlated with its one-bar bulk modulus (K0). If this new hypothesis is verified, it will allow the K0' of magmatic liquids to be estimated from their one-bar bulk modulus, for which there are well-calibrated models. Conversely, if the postulated relation between K0 and K0' is shown not to hold, the new K0' data for silicate and carbonate liquids will substantially augment the existing data set for liquids for which K0' is known independently of K0. The method that we will use to determine liquid K0' is fusion curve analysis, combined with high-quality phase-equilibrium experiments. This project will provide training to two female graduate students in X-ray diffraction, FTIR spectroscopy, and electron microprobe analysis, as well as with high-pressure and high-temperature equipment including a cold-seal pressure vessel and piston-cylinder apparatus. It will also provide them with an education in thermodynamic calculations of crystal-liquid equilibria, and on the importance of their results to studies of magmatism and volcanism. The skills that they will develop will prepare them equally well for a career in either the earth sciences or the high-tech materials industry.
