Major-element vapor fractionation on the lunar surface: An unusual lithic fragment from the Luna 20 fines

A 250-μm fragment in the Luna 20 fines has a very fine-grained “igneous” texture and has the composition (wt.%): SiO₂, 41.1; TiO₂, 0.35; Al₂O₃, 27.2; Cr₂O₃, 0.14; FeO, 4.2; MnO, 0.06; MgO, 8.5; CaO, 17.8; Na₂O, 0.05; and K₂O < 0.02. It contains ～ 65% plagioclase An_99–100, ～ 15% olivine Fo₉₀, ～ 2% Mg-Al spinel and the remainder an unusual interstitial phase with composition SiO₂, 34.8; TiO₂, 1.78; Al₂O₃, 18.3; Cr₂O₃, 0.04; FeO, 14.1; MnO, 0.22; MgO, 5.0; CaO, 24.1; Na₂O, 0.34; K₂O < 0.02. This fragment probably represents a portion of a normal highland rock (anorthositic norite) which was heated to a very high temperature by impact, lost volatiles including SiO₂, and then partially crystallized. The observed phases and their inferred crystallization sequence are consistent with experimental results in the system CaOMgOAl₂O₃SiO₂ (Schairer and Yoder, 1969), assuming the unusual phase to be a residual glass. This type of internal fractionation, leading to silica depletion in the residuum, is different from that normally observed in lunar rocks and is attributed to slightly lower bulk SiO₂ resulting from vapor fractionation due to impact (which also results in lower Na₂O and other volatiles). Because differentiation of the type shown by this fragment is rare in lunar materials, we infer that such major-element vapor fractionation is uncommon on the surface of the moon. The experimental CaOMgOAl₂O₃SiO₂ phase relations also have a bearing on the lunar model proposed by D.L. Anderson in 1973: his “refractory” original lunar composition would differentiate to produce silica deficient liquids, like the unusual phase in our fragment, rather than the normal lunar crustal rocks.