C. M. Tappen

Volatiles, magmatic degassing and eruptions of Mt. Somma-Vesuvius: Constraints from silicate melt inclusions, Cl and H2O solubility experiments and modeling

The eruptive styles of Mt. Somma-Vesuvius vary dramatically and appear to correlate with pre-eruptive magma compositions. Explosive eruptions occurring in the past 3550 years involved magmas enriched in H2O, Cl, S, F, and P and characterized by elevated (S/Cl). To interpret volatile exsolution and magmatic degassing, we have determined the solubilities of H2O and Cl for melts of natural Mt. Somma-Vesuvius phonolite and phonotephrite in hydrothermal experiments conducted at 500 and 2000 bars and temperatures of 946–1166°C. The solubility of Cl in phonolite melts is a complex function of pressure and the assemblage of volatile phases. For melts with low H2O contents and saturated in hypersaline liquid, Cl solubility decreases marginally with decreasing pressure; whereas, Cl solubility increases with decreasing pressure with vapor plus hypersaline liquid stable. Chlorine solubility in hypersaline liquid-saturated melts also varies strongly with melt composition; the Cl content of phonolite melt is approximately one-half that of phonotephrite melt at both pressures. Differentiation of the Cl-enriched Mt. Somma-Vesuvius magmas occurred via crystallization and fractionation of clinopyroxene, olivine, plagioclase, and other phenocrysts. Crystallization increased the volatile abundances of residual fractions of melt while simultaneously reducing Cl solubility, and this increased the potential for pre-eruptive saturation of compositionally evolved magmas in one or more volatile phases. Comparison of the experimental results with the pre-eruptive abundances of volatiles in Mt. Somma-Vesuvius magmas (determined from melt inclusions) and with modeled Cl solubilities in H2O-poor, hypersaline liquid-saturated melts indicates that many compositionally evolved magmas did exsolve or would have exsolved hypersaline liquid (with or without vapor) initially. The passively-erupted, lava- and scoria-forming magmas were more likely to exsolve hypersaline liquid (with or without vapor); whereas, the H2O-enriched, explosively erupting, pumice-forming magmas were more likely to exsolve vapor (with or without hypersaline liquid). Prior experimental investigations have argued that the H2O and Cl concentrations of silicate melt, aqueous vapor, and hypersaline liquid should be invariant as crystallization occurs at constant pressure and temperature. However, the ascent, crystallization, and differentiation of magmas in nature involves simultaneous changes in bulk melt composition, temperature, and pressure, and the solvus for vapor plus brine (i.e., hypersaline liquid) varies as a function of each of these variables. Consequently, the solubilities of H2O and Cl in vapor- plus hypersaline liquid-saturated magmas, undergoing ascent and compositional evolution, will not be fixed by the phase assemblage. Instead, the volatile concentrations of vapor and hypersaline liquid will and do change during eruptive and pre-eruptive magmatic degassing.