Mare basalt petrogenesis—A review of experimental studies

Abstract

Melting experiments on mare basalts are complicated by technical problems. The quality of the experimental results can be evaluated from the following criteria: (1) attainment of equilibrium (except in rate studies), (2) acceptable monitoring and control of pressure and temperature, (3) acceptable control of oxygen fugacity, and (4) preservation of constant bulk composition of the sample; i.e., reaction of the sample with its container must be controlled within narrow limits. Melting experiments in vacuum and in controlled gas atmospheres confirm petrologic and geochemical observations that most of the chemical variety among the (olivine normative) low‐Ti basalts is due to fractionation of olivine plus or minus minor spinel. The chemical variety among the high‐Ti basalts is likewise controlled by near‐surface fractionation processes, but the low‐K and high‐K groups cannot be interrelated in such a manner. Crystallization experiments have also shown that porphyritic textures can sometimes be produced by a single‐stage cooling history. High‐pressure studies indicate that the low‐Ti suite originated by partial melting of an olivine‐pyroxenite mantle at depths ranging from 200 to 350 km (to 500 km including green glass). It is generally agreed that a spectrum of low‐Ti parental magmas could be produced over this depth interval, the most olivine normative liquids originating at greatest depths. Experimental results suggest that the high‐Ti suite (and orange glass) formed by partial melting of an intrinsically Ti rich olivine‐pyroxenite mantle at depths similar to those at which the low‐Ti suite is produced. Alternatively, the high‐Ti basalts can be interpreted as hybrid liquids whose petrogenesis involves the assimilation of subcrustal Ti‐rich material by parental magmas of low‐Ti type and origin. The Ti‐rich zone is believed to have formed during an early differentiation event. Isotopic data indicate a prior differentiation history for mare basalt source regions.