New Constraints on Temperature, Oxygen Fugacity and H2O of Subduction Zone Basalts Based on Olivine-Melt Equilibrium

Abstract

Erupted basalts are windows into the deep Earth. This dissertation provides evidence that mantle-derived basalts grow their phenocrysts during rapid ascent to the surface, and that the most Mg-rich olivine in erupted samples approximates the first mineral to crystallize from a liquid with a composition of the whole rock. This presents opportunities to constrain the temperature, oxidation state and water content of these basalts on the basis of olivine-melt equilibrium. Chapter II develops a new olivine-melt thermometer based on the partitioning of Ni (DNiol/liq), and provides evidence that it is far less sensitive to the effects of pressure and dissolved H2O in the melt than DMgol/liq, which is the basis of most olivine-melt thermometers in the literature. The application of both thermometers to a set of subduction-zone basalts allows the depression of the olivine liquidus due to the effect of dissolved water to be determined based on the different temperatures calculated from the two thermometers; a minimum H2O content in the melt at the onset of olivine crystallization can be determined. Chapter III investigates the sensitivity of DNiol/liq to dissolved H2O in basaltic melts through a series of olivine-melt equilibrium experiments. Four 1-bar experiments, one anhydrous experiment at 0.5 GPa, and five hydrous experiments at 0.5 GPa are presented. The Ni-thermometer developed in Chapter II recovers the experimental temperature for all ten experiments within 14˚C on average, including those where the melt contained at least 4.4 wt% H2O. In contrast, the Mg-thermometer (Chapter II) recovers the Texpt of the anhydrous experiments within error (±26˚C), but overestimates Texpt by 88-141˚C for the hydrous experiments. The results shows a negligible dependence of DNiol/liq on pressure and dissolved water under crustal conditions, which confirms that the olivine-melt Ni-thermometer can be applied to hydrous basalts at <1 GPa without corrections for H2O content in the melt and pressure. Chapter IV applies the new Ni-based olivine-melt thermometer developed and tested in Chapter II and III to a set of mantle-derived, high-K melts that erupted within the Colima rift in western Mexico, where a mid-ocean spreading ridge is interacting with a subduction zone. Application of the Ni-thermometer, together with phase-equilibrium experiments on phlogopite lherzolite from the literature, shows that the K-rich Colima melts segregated near the base of the relatively thick lithosphere of the Jalisco block at relatively high pressures (~2.5 GPa) compared to most subduction-zone melts (~1.5 GPa). These temperatures and pressures of melt segregation provide key constraints on geodynamic models of this complex tectonic setting. Chapter V investigates the cause of the relatively high oxygen fugacity (fO2) of the pristine K-rich Colima lavas (Chapter IV). The pre-eruptive oxidation state (Fe3+/Fe2+ ratio), derived from olivine-melt Fe2+-Mg exchange equilibrium, indicate fO2 values that are ~2 log units lower than the post-eruptive value; yet they are still significantly higher than those found in most terrestrial basalts. This elevated pre-eruptive fO2 leads to higher solubility of sulfate in the melt, which degasses as SO2 and H2S, and drives melt oxidation. This is the first documentation of sulfur degassing-induced oxidation observed in natural samples.