Effect of water on tholeiitic basalt phase equilibria: an experimental study under oxidizing conditions

Abstract

To investigate the effect of water on phase relations and compositions in a basaltic system, we performed crystallization experiments at pressures of 100, 200 and 500 MPa in a temperature range of 940 to 1,220°C using four different water contents. Depending on the water activity, the oxygen fugacity varied between 1 and 4 log units above the quartz-magnetite-fayalite buffer. Addition of water to the dry system shifts the solidus > 250°C to lower temperatures and increases the amount of melt drastically. For instance, at 1,100°C and 200 MPa, the melt fraction increases from 12.5 wt% at a water content of 1.6 wt% to 96.3% at a water content of 5 wt% in the melt. The compositions of the experimental phases also show a strong effect of water. Plagioclase is shifted to higher anorthite contents by the addition of water. Olivine and clinopyroxene show generally higher MgO/FeO ratios with added water, which could also be related to the increasing oxygen fugacity with water. Moreover, water affects the partitioning of certain elements between minerals and melts, e.g., the Ca partitioning between olivine and melt. Plagioclase shows a characteristic change in the order of crystallization with water that may help to explain the formation of wehrlites intruding the lower oceanic crust (e.g., in Oman, Macquarie Island). At 100 MPa, plagioclase crystallizes before clinopyroxene at all water contents. At pressures > 100 MPa, plagioclase crystallizes before clinopyroxene at low water contents (e.g. 3 wt%. This change in crystallization order indicates that a paragenesis typical for wehrlites (olivine–clinopyroxene–without plagioclase) is stabilized at low pressures typical of the oceanic crust only at high water contents. This opens the possibility that typical wehrlites in the oceanic crust can be formed by the fractionation and accumulation of olivine and clinopyroxene at 1,060°C and > 100 MPa in a primitive tholeiitic basaltic system containing more than 3 wt% water. The comparison of the experimental results with evolution trends calculated by the thermodynamic models “MELTS” and “Comagmat” shows that neither model predicts the experimental phase relations with sufficient accuracy.