Improved Understanding of H2O-CO2 Solubility in Alkali Basalts at Mid-Crustal Pressures

High concentrations of alkali elements (Na and K) in basaltic magmas are known to increase H2O-CO2 solubility. Basaltic magmas rich in alkalis can consequently erupt explosively from exsolution of the abundant gases driving rapid ascent. Existing alkali basalt solubility models are well calibrated below 200 MPa, corresponding to a range of volatile contents that have been measured in melt inclusions (MIs) from natural systems. However, MIs from some highly explosive basaltic eruptions record CO2 contents above the calibrated range. Additionally, recent emphasis on volatiles, primarily CO2, sequestered in MI bubbles has revealed a need for understanding volatile solubility in alkali basalts at higher pressures. Some MI bubble studies have found as much as 90% of the total MI CO2 content in the MI bubble, which can easily raise MI volatile contents beyond the calibrated range and therefore inhibit MI interpretation. To allow for detailed interpretation of MI data at higher CO2 contents and better understand magmatic plumbing systems, we conducted a set of mixed H2O-CO2 volatile solubility experiments in several basaltic compositions with variable alkali contents. We targeted CO2-rich fluid compositions at pressures between 400 and 600 MPa. The compositions cover a range of total alkali contents ( 4 to 9 wt.%), but also varying ratios of Na vs. K to study the relative effect of each alkali component. Results from our experiments suggest that existing volatile solubility models for alkali basalts, if extrapolated beyond their calibrated range, are over-predicting CO2 solubility at mid-crustal pressures. We have adjusted existing models to accommodate our higher pressure experimental data and re-evaluated some available alkali basalt MI data sets. Our revised model indicates deeper magma storage in this mid-crustal region than previously interpreted from MIs with high CO2 contents.