Experimental Petrology of a Highly Potassic Magma

Abstract
The melting behaviour of a highly potassic biotite mafurite of the Central African olivine leucitite kindred has been studied experimentally as a function of pressure (to 30kb) temperature, and water content (0%, 5%, 15%, 25%, and 40% H2O). Olivine is the liquidus phase up to 30 kb for all water contents studied except for anhydrous (clinopyroxene on the liquidus) and 15% H2O (phlogopite on the liquidus) conditions. Analyses of phases crystallizing from the biotite mafurite show that pressure has very little effect on the composition of clinopyroxene which is extremely calcium-rich, and low in Al2O3 and TiO2 for all conditions investigated. Phlogopite has low TiO2 content and titanphlogopite cannot be a refractory phase in the upper mantle causing Ti-depletion in partial melts in equilibrium with titanphlogopite. There are apparently no conditions where the extremely potassic biotite mafurite could be a partial melt from pyrolite but derivation from an olivine+clinopyroxene+phlogopite+ilmenite assemblage occurring as ‘enriched’ patches in the upper mantle, is possible. Liquids in equilibrium with phlogopite as a residual phase at ∼30 kb would be olivine nephelinites with approximately 5% K2O, Na2O/K2O ≽ 1 and TiO2 > 5+. Crystal elutriation with transported residual phlogopite reacting (phlogopite+liquid 1 ⇌ olivine+liquid 2) at lower pressures provides a mechanism for K-enrichment and generating Na2O/K2O < 1.