Volatiles in lunar felsite clasts: Impact-related delivery of hydrous material to an ancient dry lunar crust

Abstract

In this detailed geochemical, petrological, and microstructural study of felsite clast materials contained in Apollo breccia samples 12013, 14321, and 15405, little evidence was found for relatively enriched reservoirs of endogenic lunar volatiles. NanoSIMS measurements have revealed very low volatile abundances (≤2–18 ppm hydrogen) in nominally anhydrous minerals (NAMS) plagioclase, potassic alkali feldspar, and SiO2 that make up a majority of these felsic lithologies. Yet these mineral assemblages and clast geochemistries on Earth would normally yield relatively high volatiles contents in their NAMS (∼20 to ≥80 ppm hydrogen). This difference is particularly notable in felsite 14321,1062 that exhibits extremely low volatile abundances (≤2 ppm hydrogen) and a relatively low amount of microstructural evidence for shock metamorphism given that it is a clast of the most evolved (∼74 wt.% SiO2) rock-type returned from the Moon. If taken at face value, ‘wet’ felsic magmas (∼1.2–1.7 wt.% water) are implied by the relatively high hydrogen contents of feldspar in felsite clasts in Apollo samples 12013 and 15405, but these results are likely misleading. These felsic clasts have microstructural features indicative of significantly higher shock stress than 14321,1062. These crustal lithologies likely obtained no more water from the lunar interior than the magma body producing 14321,1062. Rather, we suggest hydrogen was enriched in samples 12013 and 15405 by impact induced exchange, and/or partial assimilation of volatiles added to the surface of the Moon by a hydrated impactor (asteroid or comet) or the solar wind. Thus, the best estimate for magmatic water contents of felsic lunar magmas comes from 14321,1062 that leads to a calculated magmatic water content of ≤0.2 wt.%. This dry felsic magma has a slightly greater, but comparable water content to the ancient mafic magmas implied by the other lithologies that we have studied. Based on this and expanding evidence for a significantly dry ancient or early degassed Moon it is likely that some recent estimates (100's ppm) of the water abundances in the lunar parental magma ocean have been overestimated.