Formation of Pure Anorthosite During Lunar Magma Ocean Solidification: Implications for the Melt-Solid Segregation Process

E. M. Parmentier,Y. Liang
2010-01-01
Abstract:SELENE remote sensing data indicate that anorthosite consisting of nearly pure plagioclase is exposed in the central peaks of lunar craters. A mafic mineral content of less that 2% inferred from this data places an important constraint on the mechanisms of solid-solid and solid-melt segregation in the lunar magma ocean (MO). Segregation of plagioclase from denser mafic mineral grains can occur by shearing in the granular flow at the top of the low viscosity, convecting, liquidrich region of the MO. Segregation of residual mafic liquid from a compacting solid plagioclase-rich matrix should occur by the buoyant downward migration of the melt. We will explore the role of both mechanisms in the formation of lunar anorthosite. Introduction: Anorthositic crustal rocks on the Moon have long been attributed to plagioclase segregation in a magma ocean (MO); in fact, anorthite-rich Apollo samples provided a primary motivation for the MO hypothesis [cf. 1]. While other mechanisms to generate lunar anorthosites have been explored [2,3], plagioclase flotation in a MO remains a frequently cited hypothesis. Recent results from the Multiband Imager on the KAGUYA (SELENE) spacecraft indicate the presence of anorthosite with plagioclase abundance exceeding 98% [4] in the central peaks of lunar craters. These occurrences of nearly pure anorthosite motivate the need to explore whether plagioclase segregation in a MO can produce lithologies that are almost free of mafic minerals; and, if so, the conditions under which this might occur. The magma ocean setting for anorthosite formation: Figure 1 illustrates schematically temperature distributions expected in a cooling, solidifying lunar MO. The separation of plagioclase has frequently been thought of as crystal flotation in a mafic liquid. However, as recognized previously [5,6], the lunar MO during crystallization of the anorthositic crust is not expected to be fully liquid at any depth, but rather mixture of liquid and solid in varying proportions. The column inset in Figure 1, reproduced from [7], shows the composition of solids that would crystallize at each depth during the progressive fractional solidification of a particular lunar bulk composition. Liquid fractions are likely to exceed 50% during a significant fraction the plagioclase crystallization interval. Figure 2 shows that the MO should consist of two regions: a deeper mostly liquid region in which the viscosity is close to that of the liquid. Cooling at the top of this region combined with its low viscosity should allow vigorous convection [12]. In the mostly solid region at shallower depths mafic liquids can be segregated from a compacting plagioclase-rich solid matrix by buoyantly driven melt migration.
What problem does this paper attempt to address?