Silicate Liquid Immiscibility in Lunar Magmas, Evidenced by Melt Inclusions in Lunar Rocks

Abstract

Examination of multiphase melt inclusions in 91 sections of 26 lunar rocks revealed abundant evidence of late-stage immiscibility in all crystalline rock sections and in soil fragments and most breccias. The two individual immiscible silicate melts (now glasses) vary in composition, but are essentially potassic granite and pyroxenite. This immiscibility may be important in the formation of the lunar highlands and tektites. Other inclusions yield the following temperatures at which the several minerals first appear on cooling the original magma: ilmenite (?) liquidus, 1210 degrees C; pyroxene, 1140 degrees C; plagioclase, 1105 degrees C; solidus, 1075 degrees C. The glasses also place some limitations on maximum and minimum cooling rates.

Petrology and Geochemistry of DSDP Leg 16 Basalts, Eastern Equatorial Pacific

“…The tendency of the globules to coat grains of Plagioclase preferentially to olivine suggests further that the interfacial tension between globule and Plagioclase was lower than that between globule and olivine at the time of quenching. Experimental evidence supporting the theory of liquid immiscibility in silicate magmas has been presented by Ferguson and Currie (1972), Roedder and Weiblen (1970), and Philpotts and Hodgson (1968). The immiscible liquids produced by Ferguson and Currie closely approach the compositions of the globules analyzed in this study.…”

Magma chamber–scale liquid immiscibility in the Siberian Traps represented by melt pools in native iron

“…The presence of the metallic (reduced) iron and related effects of oxygen fugacity (f O 2 ) on melt unmixing, coupled with the studies of immiscibility in lunar rocks (Roedder and Weiblen, 1970;Rutherford et al, 1974), seemingly limits applications of this particular case to global processes in crystallizing terrestrial tholeiitic magmas. On the other hand, the two-liquid fi eld has been experimentally shown to expand under oxidizing conditions (Naslund, 1983;Philpotts and Doyle, 1983), although the latter also promote earlier crystallization of magnetite, which decreases the degree of iron enrichment that is necessary to reach the immiscibility fi eld (Philpotts and Doyle, 1983).…”

“…in common silicate magmas proved to be diffi cult, largely because of the transient nature of the phases involved. The existing record of the silicate-silicate liquid immiscibility is supported by observations in matrix glasses and melt inclusions (Charlier et al, 2011;De, 1974;Jakobsen et al, 2005;Kamenetsky and Kamenetsky, 2010;McBirney and Nakamura, 1974;Philpotts, 1978Philpotts, , 1981Philpotts, , 1982Roedder and Weiblen, 1970). However, the compositional evolution of immiscible melts with cooling is unknown and the small size (<10 µm) and disequilibrium state of coexisting immiscible liquid and crystal phases render much of the evidence controversial.…”

A Rb–Sr and Sm–Nd isotope geochronology and trace element study of lunar meteorite LaPaz Icefield 02205

“…The mesostasis areas have been proposed to be the result of silicate liquid immiscibility that occurs after >95% melt crystallization (Anand et al, 2006). Glassy inclusions observed in mineral grains are frequently interpreted as trapped portions of the melt from which the mineral grains crystallized (Roedder and Weiblen, 1970). Assuming that the Si-K-rich inclusions in ilmenite sample the evolved host melt from which ilmenite crystallized, and perfectly incompatible behavior of potassium, the remaining liquid fraction F (represented by the Si-K-rich inclusions) is calculated from the whole rock potassium concentration C 0 in LAP 02205 (0.11 wt% K 2 O) and the concentration C L in the Si-K-rich inclusions (7 to 8 wt% K 2 O) as F = C 0 /C L = 1.4 to 1.6 wt%.…”