Fluorine and chlorine abundances in lunar apatite: Implications for heterogeneous distributions of magmatic volatiles in the lunar interior

Apatite has been analyzed from mare basalts, the magnesian-suite, the alkali-suite, and KREEP-rich impact-melt rocks using an electron probe microanalysis routine developed specifically for apatite. We determined that all the lunar apatite grains analyzed are predominantly fluorine rich; however, they also contain varying concentrations of chlorine and a missing structural component that, after ruling out other possibilities, we attribute to OH. Apatite grains from mare basalts are compositionally distinct from the apatite grains in the magnesian-suite, the alkali-suite, and KREEP-rich impact-melt rocks, which all had similar apatite compositions. Apatite grains in mare basalts are depleted in chlorine, and many of the analyzed grains have stoichiometry that suggests a significant OH component (i.e., >0.08 structural formula units), whereas apatite grains in the magnesian suite, alkali suite, and KREEP-rich impact melts are enriched in chlorine and do not typically have a missing structural component that could be attributed to OH− (within the detection limit of 0.08sfu). From these data, we infer that residual liquids in the mare basalts were enriched in H2O and fluorine relative to chlorine at the time of apatite crystallization, whereas residual liquids in magnesian-suite, alkali-suite, and KREEP-rich impact melts were enriched in chlorine relative to H2O and fluorine at the time of apatite crystallization. The relative volatile abundance that we determined for the mare basalts is identical to the previously determined relative volatile abundance for the lunar picritic glasses. This result indicates that the observed relative volatile abundance signature of the picritic glass source is the same as that in the mare basalt source regions. The magnesian-suite, alkali-suite, and KREEP-rich impact-melt rocks likely reflect a volatile source with different volatile abundances than the sources of mare volcanics. Moreover, the magnesian-suite, alkali-suite, and KREEP-rich impact-melt rocks may reveal the relative volatile abundance of urKREEP, the residual melt of the magma ocean. This difference in relative magmatic volatile abundance among the lithologic groups investigated cannot be explained by degassing of a single source composition (relative to magmatic volatiles). The most reasonable explanation for the compositional disparity is a difference in the relative volatile abundances in the magmatic source regions of the Moon. Therefore, we conclude that the Moon has a heterogeneous distribution of magmatic volatiles within its interior, with a chemical divide (with respect to magmatic volatiles) existing between magmas that arise by partial melting of the lunar mantle and magmas that have seen significant contamination by a KREEP component.