PHASE EQUILIBRIUM CONSTRAINTS ON THE ORIGIN OF BASALTS, PICRITES, AND KOMATIITES

Abstract

Experimental phase equilibrium studies at pressures ranging from 1 atm to 10 GPa are sufficient to constrain the origin of igneous rocks formed along oceanic ridges and in hotspots. The major element geochemistry of MORB is dominated by partial crystallization at low pressures in the oceanic crust and uppermost mantle, forcing compliance with liquid compositions in low-pressure cotectic equilibrium with olivine, plagioclase and often augite too; parental magmas to MORB formed by partial melting, mixing, and pooling have not survived these effects. Similarly, picrites and komatiites can transform to basalts by partial crystallization in the crust and lithosphere. However, parental picrites and komatiites that were successful in erupting to the surface typically have compositions that can be matched to experimentally-observed anhydrous primary magmas in equilibrium with harzburgite [L+Ol+Opx] at 3.0 to 4.5 GPa. This pressure is likely to represent an average for pooled magmas that collected at the top of a plume head as it flattened below the lithosphere. There is substantial uniformity in the normative olivine content of primary magmas at all depths in a plume melt column, and this results in pooled komatiitic magmas that are equally uniform in normative olivine. However, the imposition of pressure above 3 GPa produces picrites and komatiites with variations in normative enstatite and Al2O3 that reveal plume potential temperature and depths of initial melting. Hotter plumes begin to melt deeper than cooler plumes, yielding picrites and komatiites that are enriched in normative enstatite and depleted in Al2O3 because of a deeper column within which orthopyroxene can dissolve during decompression. Pressures of initial melting span the 4 to 10 GPa range, increasing in the following order: Iceland, Hawaii, Gorgona, Belingwe, Barberton. Parental komatiites and picrites from a single plume also exhibit internal variability in normative enstatite and Al2O3, indicating either a poorly mixed partial melt aggregation process in the plume or the imposition of partial crystallization of olivine–orthopyroxenite on a well-mixed parental magma. Plume shape and thermal structure can also influence the petrology and geochemistry of picrites and komatiites. Liquids extracted from harzburgite residues [L+Ol+Opx] will dominate magmatism in a plume head, and can erupt to form komatiites in oceanic plateaus. Liquids extracted from garnet peridotite residues in a plume axis will gain in importance when the plume head partially solidifies and is removed from the hotspot by a moving lithosphere, as is the case for Hawaii. The paradoxical involvement of garnet indicated by the heavy rare earth elements in picrites that otherwise have a harzburgite signature in Hawaii can be explained by the mixing and collection of magmas from the plume axis. Volcanic rocks from Hawaii and Gorgona and xenoliths from cratonic mantle provide evidence for the importance of partial crystallization of plume magmas when they encounter a cold lithosphere. Harzburgite residua and olivine–orthopyroxene cumulates formed in plumes can yield compositionally distinct lithospheric mantle which is buoyant, and this could have provided an important foundation for the stabilization of the first continents.