Experimental petrology and petrogenesis of mare volcanics

Abstract

Mare volcanics consist of basalts and picritic pyroclastic glasses spanning a wide range of TiO 2 concentration. The more primitive low-Ti basalts and picritic glasses have olivine alone on their lowpressure liquidi. Most of the chemical variation among the low-Ti basalts is the result of olivine fractionation in a series of parental MgO-rich liquids differing in TiO 2 concentration. With one possible exception (Apollo 17 VLT) none of the picritic compositions is a suitable parent for any of the observed low-Ti basalts. Most of the chemical variation among the high-Ti basalts is the result of a series of magmas fractionated along the low-pressure olivine + armalcolite/ilmenite cotectic. All of the picritic high-Ti glasses have olivine alone on the liquidus, but none is a suitable parent for any of the basalts. Volcanics with intermediate TiO 2 concentrations (5 to 10 wt%) are widespread in the maria, even though they are not well represented in the sample collections; however, there is no evidence either among the samples or from remote sensing studies of basalts with > 13 wt% TiO 2 that would be expected as differentiates of the picritic glasses with the highest TiO 2 concentrations. Controlled-cooling-rate crystallization studies on a variety of mare compositions have provided the basis for reconstructing the size and, in some cases, stratigraphy of mare flows. Groundmass textures, crystal size, crystal morphology, nucleation density, and zoning patterns have all been employed to quantify cooling histories of mare basalts. A single-stage linear cooling rate may produce a porphrytic texture. Rapid cooling may also cause plagioclase to crystallize after a mineral that it precedes during slow cooling. Mare basalts are highly reduced. Mineral assemblages and intrinsic oxygen fugacity measurements indicate ƒ O 2 below the wüstite-iron buffer and at or near iron metal saturation. Accordingly, experiments run in high-purity iron capsules gained or lost little iron. Most basalts are undersaturated with respect to sulphur, so reduction through sulphur volatilization cannot be invoked to explain the presence of iron in olivine phenocrysts. The low oxidation state is most likely the result of melting a reduced interior under fluid-absent conditions. The progressive reduction of Cr 3+ → Cr 2+ and Ti 4+ → Ti 3+ in lunar melts permits the elimination of the Fe 3+ that is present at iron-saturation in simple systems. Crystal-liquid partition coefficients determined from melting experiments have been used in a wide range of calculations of major and trace element evolution. The coefficient for Fe-Mg exchange between olivine and liquid apparently varies with TiO 2 concentration of the liquid and is particularly useful in assessing whether fine-grained rocks have excess olivine. Nb and Ti have excess concentrations in mare basalts relative to adjacent REE in incompatibility diagrams. These excesses or positive anomalies are consistent with ilmenite accumulation in light of measured partition coefficients and imply continuous variation of accumulated ilmenite even in the low-Ti mare source regions. Pressures of multiple saturation (olivine + pyroxene ± Cr-rich spinel ± ilmenite) are in the range of 5 to 12.5 kbar for primitive mare basalts and in the range of 18 to 25 kbar for the picritic glasses. Low-Ca pyroxene is the only pyroxene along the liquidus of the low-Ti basalts and glasses; however, augite is the pyroxene most commonly observed along the high-pressure liquidi of the high-Ti basalts: High-Ti picrites have augite in the subliquidus region at intermediate pressures where olivine is the liquidus phase, but orthopyroxene is the liquidus phase at multiple saturation. Because of the steep depth/pressure gradient in the outer portion of the Moon (20 km/kbar), these pressures imply: 1. (a) great depths of melting within the Moon. 2. (b) some means of transporting magmas hundreds of kilometers to the surface without significant chemical modification, if both olivine and pyroxene were left in the residuum. Modeling of major elements during polybaric partial melting suggests that it is possible to reproduce the composition and high-pressure signature of the low-Ti (green) picritic glasses by accumulating small degrees of melt extracted from an upwelling source region. In such a case, melting of the differentiated source must begin at > 1000 km depth (40 kbar) and cease at ~ 100 km (5 kbar). If this model is correct, then experimental determination of the pressure of multiple saturation gives an average pressure of melting: the onset of melting is at higher pressure and actual segregation of the melt from the mantle is at lower pressure.