The systematics of light lithophile elements (Li, Be and B) in lunar picritic glasses: Implications for basaltic magmatism on the Moon and the origin of the Moon

Abstract

Lunar picrites, represented by high-Mg volcanic glasses, are thought to be products of either partial melting of the deep lunar mantle followed by rapid ascent or polybaric partial melting initiated in the deep lunar mantle. The near primary compositions of these volcanic glasses provide us with a unique perspective for evaluating basaltic magmatism, the characteristics and evolution of the lunar mantle, and the origin of the Moon. The light lithophile elements (LLE = Li, Be, B) in planetary materials have been used to estimate planetary compositions and evaluate magmatic processes. Ion microprobe analyses of these glasses for LLEs were conducted using a Cameca 4f ion microprobe. This suite of glass beads ranged in TiO 2 from 0.3 to 17 wt%. Seventy-one individual glass beads were analyzed for the LLEs. In addition, core-rim analyses of individual glass beads were made. The LLEs show a wide range of variability with Li ranging from 1.2 to 23.8 ppm, Be ranging from 0.06 to 3.09 ppm and B ranging from 0.11 to 3.87 ppm. B/Be ranges from 0.40 to 4.6. Li/Be ranges from 2.7 to 41.7, although 90% of the Li/Be values range from 14 to 30. Both B/Be and Li/Be values for the picritic glasses are less than chondrite. Be/Nd for the glasses ranges from .04 to .06 and are similar to chondrite (.058). Traverses across individual beads indicate that they are generally homogeneous with regards to LLEs regardless of TiO 2 content. The individual glass groups show limited variations in LLE characteristics. The exceptions to this observation are the A17 VLT and the A15 yellow glasses. At individual Apollo sampling sites, the LLE content is generally correlated to TiO 2. The high-Ti glasses are displaced toward higher Li at similar B and Be relative to the very low-Ti glasses. LLE concentrations also parallel the enrichments of other lithophile elements such as Ba, Zr, Sr and REEs. As noted for other trace element characteristics, glasses from each sampling site have similar LLE signatures. For example, the Apollo 14 glasses generally have higher LLE concentrations relative to glasses of similar TiO 2 content from other sites. The LLE data support mantle inhomogeneity and Lunar Magma Ocean (LMO) cumulate overturn models suggested by previous studies. A KREEP component had been incorporated into some of these picritic glasses. This is consistent with other trace elements and probably reflects the recycling of KREEP and/or other late stage LMO cumulates into the deep lunar mantle. The picritic glasses are compositionally distinct from the crystalline mare basalts in LLEs. They are not related by either fractional crystallization or partial melting processes. This suggests that they were derived from distinctively different mantle sources. Estimates of the bulk compositions of the Earth and the Moon have previously been made based on the assumption that the ratio of Li to Be is a direct measure of the ratio of the high temperature condensates (HTC, refractory components) to the Mg-silicates (less refractory components) in a planet. We assert that if Li/Be is to be used to estimate bulk Moon composition, the picritic glasses provide fewer pitfalls and a better estimate than the crystalline mare basalts. Differences in partition coefficients ( D) for Li and Be indicate that fractional crystallization and partial melting will modify the Li Be ratio. Estimates based on the picritic glasses infer a higher Li/Be for the bulk Moon than estimated from the mare basalts. This would indicate that the bulk Moon is less refractory than previously calculated by Li/Be and approaches the bulk composition of the Earth.