Graphite oxidation in the Apollo 17 orange glass magma: Implications for the generation of a lunar volcanic gas phase

Abstract

An Apollo 17 picritic orange glass composition has been used to experimentally investigate the conditions at which graphite would oxidize to form a CO-rich gas, and ultimately produce lunar fire-fountain eruptions. Isothermal decompression experiments run above the A17 orange glass liquidus temperature (>1350 °C) suggest that the initial CO-rich gas phase produced by graphite oxidation would be generated during magma ascent at a pressure of 40 MPa, ∼8.5 km beneath the lunar surface. Additional experiments with 2000 ppm S and 1000 ppm Cl showed that the presence of these dissolved gas species would not affect the depth of graphite oxidation, verifying that the first volcanic gas phase would be generated by the oxidation of graphite. A simple ideal chemical mixing model for calculating melt FeO activity in a Fe-metal/silicate melt system was tested with a series of 0.1 MPa controlled oxygen fugacity experiments. Agreement between the model and experiments allows the model to be used to calculate oxygen fugacity in picritic lunar glass compositions such as the A17 orange glass. Using this model in a reanalysis of chemical equilibria between the natural A17 orange glass melt and the metal spherules (Fe 85Ni 14Co 1) trapped within the glass beads indicates a log oxygen fugacity of −11.2, 0.7 log units, more oxidized than previous estimates. At the A17 orange glass liquidus temperature (1350 ± 5 °C), this f O2 corresponds to a minimum pressure of 41 MPa on the graphite–C–O surface. The fact that the same critical graphite oxidation pressure was determined in decompression experiments and from the Fe–FeO activity model for the natural A17 orange glass–metal assemblage strongly supports this pressure (8.5 km depth) for volcanic gas formation in lunar basalts. Generation of a gas by oxidation of C in ascending magma is likely to have been important in getting dense lunar magmas to the surface as well as in generating fire-fountain eruptions. The vesicles common in many lunar basalts and the ubiquitous Fe-metal in these rocks are also likely generated by the oxidation of carbon. The presence of carbon in the lunar basalts and the recent discovery of ppm levels of water in lunar basalts indicate that at least parts of the lunar interior still contained volatiles at 3.9 bybp.