The ubiquitous presence of silica‐rich glass inclusions in mafic minerals: Examples from Earth, Mars, Moon and the aubrite parent body

Abstract

Abstract— Highly silicic glass inclusions are commonly present in mafic minerals of xenolithic terrestrial upper mantle rocks (Schiano and Clocchiatti, 1994). They are believed to be the products of volatile‐rich silicic melts for which several sources have been proposed (Francis, 1976; Frey and Green, 1974; Schiano et al, 1995), but their origin(s) and, consequently, that of the glasses, remains unknown. However, in situ formation by very low‐degree partial melting seems to be possible as has been shown by experiments (e.g., Baker et al, 1995; Draper and Green, 1997).Glass inclusions of silicic chemical composition are also present in some mafic minerals of achondritic meteorites (e.g., Fuchs, 1974; Okada et al, 1988; Johnson et al, 1991). The enstatite achondrites (aubrites) Aubres and Norton County, which record early planetesimal and planet formation in the solar nebula, and the olivine achondrite (chassignite) Chassigny, a rock believed to originate from Mars, contain abundant glass inclusions in their main minerals enstatite and olivine, respectively. Glasses of glass‐bearing inclusions have a highly silicic and volatile‐rich chemical composition similar, but not identical, to that of glass inclusions in terrestrial upper mantle peridotite minerals. Furthermore, glass inclusions in olivines from the Moon (e.g., Roedder and Weiblen, 1977) are also silica‐rich.Because different physicochemical conditions prevail in the source regions of these rocks, the process of melting is, perhaps, not generally applicable for the generation of silica‐rich glasses. Alternatively, the glasses could have been formed via precipitation from silicate‐loaded fluids (Schneider and Eggler, 1986) or vapors. Another possible mechanism, not previously identified, could be dehydrogenation of nominally nonhydrous mafic minerals by heating or depressurization that should be accompanied by expulsion of excess silica and incompatible elements. This process will mimic low‐temperature, very low‐degree partial melting. It could account also for the highly variable glass/bubble ratios observed in glass inclusions in aubrite enstatites. We suggest that such a process could have been operating in the solar nebula, the Moon and Mars, and could be operating still on Earth.