Water (hydrogen) in the lunar mantle: Results from petrology and magma ocean modeling

Measurements of hydroxyl or water in lunar materials are being used to suggest that melting source regions within the Moon contained tens to hundreds of ppm water. The compositions of magmatic source regions in the Moon can be calculated using magma ocean fractional solidification models, given an initial bulk composition. Our magma ocean models demonstrate that to make a source region with 10 to 200 ppm water, the bulk magma ocean must contain 100 to 1000 ppm water or more (and in a 1000-km deep lunar magma ocean, 1000 ppm water is equivalent to ~ 5% of an Earth ocean by weight).As solidification of the magma ocean continues water and other incompatible elements are progressively enriched in the evolving liquids. The final KREEP-like liquids would contain the greatest fraction of water. We suggest that a bulk magma ocean with over ~ 100 ppm water is prohibited by constraints on these KREEP-like final magma ocean liquids, and by the thermodynamic equilibrium requirement with metallic iron in mare basalts.We conclude that bulk lunar interior water contents must be at the low end of the range posited by recent observations (<~ 10 ppm), though some water from later impacts may have been assimilated into lunar magmas. Most importantly, given the oxygen fugacity of lunar rocks, hydrogen would be expected to be an important volatile species equal to or within an order of magnitude of the abundance of water in a lunar magma.

► Water in lunar rocks is suggested to imply 10–200 ppm water in the source regions.
► The magma ocean would have had ~ 1000 ppm water to produce such a mantle.
► Constraints on KREEP liquids prohibit a bulk magma ocean with over ~ 100 ppm water.
► Oxygen fugacity requires that hydrogen be close to water abundance in lunar magmas.
► Hydrogen, along with hydroxyl and water, must have been important in lunar melts.