No effect of H2O degassing on the oxidation state of magmatic liquids

Abstract

Abstract The underlying cause for why subduction-zone magmas are systematically more oxidized than those formed at mid-ocean spreading ridges is a topic of vigorous debate. It is either a primary feature inherited from the subduction of oxidized oceanic crust into the mantle or a secondary feature that develops because of H 2 O degassing and/or magma differentiation. Low total iron contents and high melt H 2 O contents render rhyolites sensitive to any effect of H 2 O degassing on ferric–ferrous ratios. Here, pre-eruptive magmatic Fe 2+ concentrations, measured using Fe–Ti oxides that co-crystallized with silicate phenocrysts under hydrous conditions, are compared with Fe 2+ post-eruptive concentrations in ten crystal-poor, fully-degassed obsidian samples; five are microlite free. No effect of H 2 O degassing on the ferric–ferrous ratio is found. In addition, Fe–Ti oxide data from this study and the literature show that arc magmas are systematically more oxidized than both basalts and hydrous silicic melts from Iceland and Yellowstone prior to extensive degassing. Nor is there any evidence that differentiation (i.e., crystal fractionation, crustal assimilation) is the cause of the higher redox state of arc magmas relative to those of Iceland/Yellowstone rhyolites. Instead, the evidence points to subduction of oxidized crust and the release of an H 2 O-rich fluid and/or melt with a high oxygen fugacity ( f O 2 ), which plays a role during H 2 O-flux melting of the mantle in creating basalts that are relatively oxidized.