Quantitative evaluation of the effect of H2O degassing on the oxidation state of magmas

Abstract:

The extent to which degassing of the H₂O component affects the oxidation state of hydrous magmas is widely debated. Several researchers have examined how degassing of mixed H-C-O-S-Cl fluids may change the Fe³⁺/Fe^T ratio of various magmas, whereas our focus is on the H₂O component. There are two ways that degassing of H₂O by itself may cause oxidation: (1) the reaction: H₂O (melt) + 2FeO (melt) = H₂ (fluid) + Fe₂O₃ (melt), and/or (2) if dissolved water preferentially enhances the activity of ferrous vs. ferric iron in magmatic liquids. In this study, a comparison is made between the pre-eruptive oxidation states of 14 crystal-poor, jet-black obsidian samples (obtained from two Fe-Ti oxides) and their post-eruptive values (analyzed with the Wilson 1960 titration method tested against USGS standards). The obsidians are from Medicine Lake (CA), Long Valley (CA), and the western Mexican arc; all have low FeO^T (1.1-2.1 wt%), rendering their Fe²⁺/Fe³⁺ ratios highly sensitive to the possible effects of substantial H₂O degassing. The Fe-Ti oxide thermometer/oxybarometer of Ghiorso and Evans, (2008) gave temperatures for the 14 samples that range for 720 to 940°C and ∆NNO values of –0.9 to +1.4. With temperature known, the plagioclase-liquid hygrometer was applied and show that ≤ 6.5 wt% H₂O was dissolved in the melts prior to eruption. In addition, pre-eruptive Cl and S concentrations were constrained on the basis of apatite analyses (Webster et al., 2009) and sulfur concentrations needed for saturation with pyrrhotite (Clemente et al., 2004), respectively. Maximum pre-eruptive chlorine and sulfur contents are 6000 and 200 ppm, respectively. After eruption, the rhyolites lost nearly all of their volatiles. Our results indicate no detectable change between pre- and post-eruptive Fe²⁺ concentrations, with an average deviation of ± 0.1 wt % FeO. Although degassing of large concentrations of S and/or Cl may affect the oxidation state of magmas, at the pre-eruptive levels in these 14 rhyolitic magmas, no effect is detected. Therefore, it can be robustly concluded that degassing of substantial amounts of the H₂O component (≤ 6.5 wt%), by itself, does not induce oxidation in erupted magmas, particularly those more iron-rich than rhyolites (e.g., arc basalts).