Search

Delineation of low pressure phase equilibria in the composition space relevant to titaniferous lunar basalts demonstrates a significant degree of control by those equilibria on the compositions of the basalts. The existence of two distinct chemical groups of basalts (high and low K) which cannot be related one to the other by fractional crystallization at any pressure, suggests that melting is responsible for the two groups. Consideration of the pressure shift required to produce the differences between groups constrains magma segregation to have occurred in the outer 150 km of the moon. It is difficult to relate low-Ti and high-Ti basalts to the same source region. The preferred source region of high-Ti basalts, based on phase equilibrium considerations, is a late ilmenite-rich cumulate produced from the residual liquid of the primordial differentiation of the outer portions of the moon. This ilmenite-rich layer is sandwiched between the lunar feldspathic crust and a complementary mafic cumulate.

Mass-balance calculations indicate that the molten layer originally covering the moon may have been several hundred kilometers deep. The solidification of this magma ocean involved some unusual effects resulting from the large-pressure differential in the thick magma layer. The sunken cumulates from this differentiation would be more iron-rich, less refractory, and would contain more incompatible elements than would be expected in simple isobaric fractional crystallization and crystal sinking models of magma differentiation. These lower cumulates can have the mineralogical and chemical properties of the source regions of the low-Ti mare basalts.

Melting experiments over the pressure range 0 to 20 kilobars were conducted on Apollo 14 igneous rocks 14310 and 14072 and on comprehensive fines 14259. The mineralogy and textures of rocks 14310 and 14072 are presumed to be the result of near-surface crystallization. The chemical compositions of the samples show special relationships to multiply-saturated liquids in the system: anorthite-forsterite-fayalite-silica at low pressure. Partial melting of a lunar crust consisting largely of plagioclase, low calcium pyroxene, and olivine, followed by crystal fractionation at the lunar surface is proposed as a mechanism for the production of the igneous rocks and soil glasses sampled by Apollo 14.

Petrographic studies of Apollo 17 mare basalts indicate that 70215 and 71569 arrived at the lunar surface as liquids. Low-pressure melting experiments show that compositional variations within the Apollo 17 and Apollo 11 ophitic basalt suites may be generated by near-surface fractional crystallization of liquids with compositions similar to 70215 and 70017. High-pressure melting experiments show that liquids similar in composition to 70017 and 70215 can be generated by partial melting of an olivine+clinopyroxene+Fe-Ti-oxide source at depths of 100-150 km within the moon.

Melting experiments and petrographic studies of lunar feldspathic rocks reveal possible genetic relationships among several compositionally and mineralogically distinct groups of lunar rocks and soil fragments. Dry, low PO2 partial melting of crustal anorthositic norites of the anorthositic-noritic-troctolitic (ANT) suite produces liquids of the KREEP-Fra Mauro basalt type; dry, low PO2 partial melting of pink spinel troctolite (PST) produces liquids of the 'very high alumina basalt' or microtroctolite type. Both ANT and PST are probable components of the primitive terra crust. If crystal fractionation in a cooling basaltic liquid could have produced such a crust, it would also produce a mafic interior capable of yielding mare basalts by later remelting at depth.

Petrographic studies of Apollo 16 samples indicate that rocks 62295 and 68415 are crystallization products of highly aluminous melts. 60025 is a shocked, crushed and partially annealed plagioclase cumulate. 60315 is a recrystallized noritic breccia of disputed origin. 60335 is a feldspathic basalt filled with xenoliths and xenocrysts of anorthosite, breccia, and anorthite. The Fe/(Fe+Mg) of plagioclase appears to be a relative crystallization index. Low pressure melting experiments with controlled Po2 indicate that the igneous samples crystallized at oxygen fugacities well below the Fe/FeO buffer. Crystallization experiments at various pressures suggest that the 62295 and 68415 compositions were produced by partial or complete melting of lunar crustal materials, and not by partial melting of the deep lunar interior.