CONSTRAINTS ON MIXED-FLUID (H2O AND CO2) SOLUBILITY IN ARC BASALTS: NEW EXPERIMENTS REQUIRE NEW MODELS

Sensible interpretations of magmatic volatiles depend on our knowledge of solubility relationships. One example is the observation of high water and moderate CO2 abundances in arc melt inclusions. Estimations of volatile saturation pressures and identification of degassing trends require reliable mixed-fluid (i.e. H2O-CO2) solubility estimates. A series of more than 30 piston-cylinder experiments have been conducted at conditions of H2O-CO2 saturation and pressures of 300 to 700 MPa to determine how volatile solubility varies in arc basalts in response to calcium content. Two natural basalts from Central America and a third Ca-spiked composition provided low-, medium- and high-CaO basalt compositions (9.1, 12.7 and 18.5 wt.%, respectively). After quench the fluid from each capsule was extracted in a vacuum line and analyzed by manometry. Experimental glasses were analyzed for H2O and CO2 by spectroscopy (FTIR), high-temperature manometry and ion probe. Results show that even modest CaO variation (from low- to moderate-CaO) results in a significant increase (40%) in CO2 solubility (400 MPa, fluid XH2O = 0.42). At higher CaO contents the increase in CO2 solubility is even greater. These new data address poorly constrained parameters relevant to solubility behavior in arc magmatism: 1) characterization of associated H2O-CO2 fluid at volatile saturation, 2) moderate pressures corresponding to middle to upper crust, 3) calcic and calc-alkaline compositions. A model based on a step-wise least-squares regression has been developed to address the influence of compositional variation on the solubility of arc basalts. It is expected that this model will allow more precise estimates of volatile saturation pressures and H2O-CO2 fluids in equilibrium with arc magmas. Initial application of the model to published melt inclusion data from Cerro Negro, Nicaragua indicates a sharply lower estimate of volatile saturation pressure for CaO-rich (12 wt.%) compositions with ~5 wt.% H2O and 1,100 ppm CO2 (300 MPa versus 500 MPa).