Experimental observations of TiO2 activity in rutile-undersaturated melts

Abstract

Abstract An estimate of TiO2 activity (aTiO2melt-sat) is necessary for the application of trace-element thermobarometry of magmatic systems where melts are typically undersaturated with respect to rutile/anatase. Experiments were performed in the system SiO2-Na2O-TiO2 to develop two independent methods of estimating aTiO2melt-sat—one based on the commonly applied rutile-saturation technique and another utilizing a novel Ti-in-tridymite thermometer. It is demonstrated that the rutile-saturation model can lead to an overestimate of aTiO2melt-sat relative to TiO2 activity calculated using the solubility of Ti in tridymite (SiO2) coexisting with rutile. Overestimation via the rutile-saturation technique is due to variations in the solubility mechanisms of Ti in the melt phase as a function of Ti content. In natural systems, overestimates of aTiO2melt-sat will lead to an underestimation of crystallization temperatures by Ti-based trace-element thermobarometers. Although this study is not directly applicable to natural systems, it lays the groundwork for future research on natural composition magmas to constrain TiO2 activity in melts.