TitaniQ revisited: expanded and improved Ti-in-quartz solubility model for thermobarometry

New experiments to study titanium solubility in quartz were conducted at conditions not previously explored to extend and improve existing Ti-in-quartz solubility models for thermobarometric applications. Starting materials for experiments included silica glass, anatase, synthetic and natural rutile, Ti-enriched silica gel, Ti-enriched melts, zirconia, and HF and H 2 O fluids. Additional experimental data enabled us to characterize Ti-in-quartz solubility across much of the α- and β-quartz stability fields from 2 to 30 kbar and 550 to 1050 °C. Mutual occurrences of mineral inclusions in one another and Raman spectroscopy of mineral phases confirmed co-crystallization of quartz, rutile, and zircon. Electron microprobe measurements and cathodoluminescence images show that Ti concentrations in quartz crystals from all experiments are relatively uniform, and Ti concentrations of quartz crystals grown at the same experimental conditions using several Ti–rich starting materials and several different growth media are the same within experimental and analytical uncertainties. There are no significant differences in Ti concentrations of quartz across the α–β quartz transition. The Ti concentration in quartz crystals, X TiO 2 quartz , systematically increases with temperature, but the quantity R T ln X TiO 2 quartz is a constant at fixed pressure. The Ti concentration in quartz decreases non-linearly with pressure. To account for the observed P–T dependent changes to Ti in quartz, we developed the Ti-in-quartz solubility model: R T ln X TiO 2 quartz = - 55.287 - [ P kbar ∙ ( - 2.625 + 0.0403 P kbar ) ] + R T ln a TiO 2 rutile where R is the gas constant 0.0083145 kJ/K, T is temperature in Kelvin, P is the pressure in kbar, X TiO 2 quartz is the mole fraction of TiO 2 in quartz, and a TiO 2 rutile is the activity of TiO 2 in the growth media (e.g., fluid, melt) referenced to rutile at standard state conditions of 1 bar and 25 °C. Experiments that co-crystallized quartz, rutile, and zircon permitted us to cross-check thermobarometric results from our Ti-in-quartz solubility models against the widely accepted Zr-in-rutile solubility models. We further tested our Ti-in-quartz solubility models using experiments that co-crystallized quartz, wollastonite, and titanite to fix a TiO 2 rutile < 1. Concentrations of Ti in quartz crystallized from the sub-unity a TiO 2 rutile experiments in the α- and β-quartz fields predict activities that match those calculated using the min