Crystallization of a Lunar Magma Ocean: Preliminary Experimental Results

Introduction: Although the Lunar Magma Ocean (LMO) hypothesis has achieved wide acceptance in the lunar science community, it has never been fully tested experimentally. Presented here are the first results of equilibrium crystallization experiments using the Taylor Whole Moon (TWM) composition [1]. The first mafic cumulates in this LMO are dominated by forsteritic olivine, low-Ca pyroxene and, at near Moon center pressures, slightly majoritic garnet. These cumulates and the compositions of the residual liquids provide important insights into the constitution of the deep lunar mantle and constrain the maximum depth of crystallization in the LMO. Methods: The TWM starting material was synthesized as a mechanical mixture of anhydrous powdered reagents and conditioned at the IW fO2 buffer at 1000°C for ~24 hr in a Deltech gas mixing furnace. Experiments were conducted using graphite capsules in a Walker-style multi-anvil press. Cell assemblies, procedures and calibrations are identical to those described by Agee et al. [2]. Run products were analyzed using a JEOL 8200 electron microprobe. Early Crystallization in the LMO: Our experiments simulate equilibrium crystallization in a LMO that represents whole Moon melting. Experimental phase assemblages are shown in figure 1. At 4.5 GPa, just below the Moon center pressure (4.7 GPa), TWM appears to be multiply saturated with forsteritic olivine and slightly majoritic garnet. Olivine is the sole crystallizing phase at 3.7 GPa, 1800°C and the remaining experiments reported here contain olivine + low-Ca pyroxene. Several physical models suggest that early LMO crystallization did not occur at the base of the magma ocean, but rather in an inertial zone created by vigorous convection [3,4,5]. In such models, equilibrium crystallization is thought to prevail until some critical crystallinity is reached (~60-80%), after which fractional crystallization takes place. These early formed crystals make up the deeper lunar mantle and become the source regions for mare basalts and picrite glasses. Presence of Garnet in the Lunar Mantle: Arguments regarding whether garnet is present in the lunar mantle have largely focused on garnet’s effect on trace elements [6, 7]. In our experiment at 1800°C and 4.5 GPa (~1600 km depth) garnet makes up ~13% of the crystalline assemblage but is absent at pressures of 4.1 GPa and below. Thus garnet appears to be stable under a comparatively narrow range of conditions. If the Moon experienced slightly less than complete melting, the base of the magma ocean may not have reached the garnet stability field. However, even if the Moon did experience complete melting, the inertial zone of crystallization may not have extended to depths as great as this narrow window of garnet stability and thus garnet would not have crystallized.