EXPERIMENTAL CONSTRAINTS ON THE ORIGIN OF LUNAR HIGH-TI ULTRAMAFIC GLASSES

Phase equilibria and dissolution rate experiments are used to develop a petrogenetic model for the high-Ti lunar ultramafic glasses. Near-li quidus phase relations of the Apollo 14 black glass, the most Ti-rich lunar ultramafic glass, are determined to 2.2-GPa. The liquidus is sat urated with Cr-spinel at 1-atm, olivine between similar to 0.5- and 1. 5-GPa, and low-Ca pyroxene + Cr-spinel above 1.5-GPa. Ilmenite does no t crystallize near the liquidus and implies that high-Ti ultramafic gl asses are not produced by melting of an ilmenite-saturated source. We infer that high-Ti ultramafic magmas are derived from low-Ti ultramafi c parent magmas by assimilation of ilmenite +/- clinopyroxene +/- urKR EEP +/- pigeonite in the shallow lunar interior. Heat is provided by a diabatic ascent of the low-Ti ultramafic primary magmas from the deepe r lunar interior and crystallization of olivine during assimilation. T he assimilation reaction is modeled by mass balance and requires that ilmenite and high-Ca pyroxene are assimilated in a similar to 3:1 rati o, a much higher ratio than the proportion in which these minerals are thought to exist in the lunar interior. In an effort to understand th e kinetic controls on this reaction, the dissolution of ilmenite is ex amined experimentally in both low- and high-Ti lunar magmas. We find t hat ilmenite dissolves incongruently to Cr-spinel and a high-Ti melt. The dissolution reaction proceeds by a diffusion-controlled mechanism. An assimilation model for the origin of high-Ti melts is developed th at leaves the magma ocean cumulates in their initial stratigraphic pos itions and obviates source hybridization models that require lunar ove rturn. Copyright (C) 1997 Elsevier Science Ltd.