Colloquium: Zachary D. Sharp
Abstract: A number of provocative and apparently contradictory studies have been published in the last several years regarding the water content of the Moon. Several authors claim initial water contents equal to that of the Earth, others argue for a dry Moon and still others fall midway between the extremes. Whether or not the Moon was always an anhydrous planet is critical for understand volatile retention of planets. Glass beads and melt inclusions from lunar fire fountain samples have high H abundance, suggesting water contents approaching those of the Earths mantle. The anion compositions of lunar apatites indicate a moderate hydroxyl melt component during late crystallization. However, the extremely elevated Cl isotope compositions of lunar samples indicate degassing from an anhydrous melt, and thermodynamic calculations of evolved metallic iron-bearing KREEP basalts constrain the water content of the lunar interior to <10ppm. Finally, the elevated D/H ratio of apatite (~600?) is interpreted as indicating a large contribution of cometary, rather than indigenous, water. In other words, convincing data have been presented spanning the entire range from a wet to dry Moon.
All of these apparent discrepancies are reconciled in terms of the oxygen fugacity ( f(O2)) of the Moon. The (f(O2) of the lunar mantle may be as much as ~5 orders of magnitude lower than that of Earth. On Earth, H2O is the dominant phase in the H-O system for the upper mantle, whereas on the Moon, H2 will predominate. The different speciation has important implications. First, the diffusion rate of H2 in basalt is >100 times higher than for water. A shallow-level magma can degas H2 far more rapidly than H2O, explaining the immeasurably low H content in most lunar basalts. In an unbuffered system, loss of H2 would increase the H2O/H2 ratio, but as long as Fe metal is present, the f(O2) and H2O/H2 ratio will be buffered, allowing for continued loss of H2. In contrast to the Moon, the Earths upper mantle does not presently contain Fe metal. Any early loss of H2 would have raised the f(O2) of the mantle and increased the H2O/H2 ratio, effectively arresting further H2 loss.
H2 loss has the following consequences. 1) Only samples that were rapidly quenched (glass beads) or preserved as melt inclusions could retain any measurable H. 2) H2 degassing and loss to space would significantly increase the D/H ratio of the melt. Assuming a Earth-like initial D/H ratio, a D value in excess of 600? is achieved by kinetic considerations alone if only 95% of H2 is lost from a magma. Such levels of deuterium enrichment are nearly impossible to achieve if H2O were the degassing species. 3) Fe metal is stable at high f(H2) values, eliminating the thermodynamic/mineralogical constraint on lunar water contents. 4) H2 will degas far more rapidly than HCl or metal chlorides. The high Cl isotope ratios of the Moon require an anhydrous magma during Cl degassing, but do not preclude an H-rich melt prior to Cl loss. 5) The moderately high OH- contents of apatites reflect the OH- content of the melt. If saturated in both H2 and OH-, then this ratio will be fixed by solubility and f(O2) buffers.
The limited post-accretionary differentiation of the Moon may therefore explain its anhydrous character. On Earth, the more efficient partitioning of Fe to the core (stronger gravity, longer duration at elevated temperature?) removed the Fe buffer, allowing for a rapid rise to higher f(O2) during H2 loss and stabilization of water rather than H2. Thus, both size and oxidation state will dictate whether or not a newly forming planetary body will stabilize and retain water.
Zachary D. Sharp, Professor
Department of Earth & Planetary Sciences, University of New Mexico