Abstract The origins of lunar silicic magmas are not well understood. While clast to grain sizes of silicic materials have been returned, none of the Apollo missions visited a silicic target. Despite the availability of a wealth of remote data sets, their low spectral resolution has limited our ability to constrain the bulk composition of lunar silicic volcanic landforms and their pressure–temperature–composition ( P – T – X ) evolution. Here we test two proposed formation mechanisms for the formation of silicic melts using rhyolite-MELTS models. Models of crustal melting show that only 10%–20% partial melting of two likely starting compositions produces melts with >60 wt% SiO 2 and 65 wt% SiO 2 and >5 wt% for returned samples. Alternatively, models of fractional crystallization of KREEP basalt residual liquids can produce silicic melt compositions with >68 wt% SiO 2 and >7 wt% total alkali, consistent with returned silicic samples. Furthermore, thorium partitioning calculations show that the fractionated silicic melts contain >60 ppm thorium, consistent with returned samples. We argue that fractional crystallization of KREEP basaltic magmas at shallow depths along a relatively high geotherm at low f O 2 is the most likely formation mechanism for lunar silicic magmas.
Ravi et al. (Mon,) studied this question.