Rutile solubility in NaF–NaCl–KCl-bearing aqueous fluids at 0.5–2.79 GPa and 250–650 °C

Abstract

The complex nature of trace element mobility in subduction zone environments is thought to be primarily controlled by fluid–rock interactions, episodic behavior of fluids released, mineral assemblages, and element partitioning during phase transformations and mineral breakdown throughout the transition from hydrated basalt to blueschist to eclogite. Quantitative data that constrain the partitioning of trace elements between fluid(s) and mineral(s) are required in order to model trace element mobility during prograde and retrograde metamorphic fluid evolution in subduction environments. The stability of rutile has been proposed to control the mobility of HFSE during subduction, accounting for the observed depletion of Nb and Ta in arc magmas. Recent experimental studies demonstrate that the solubility of rutile in aqueous fluids at temperatures >700°C and pressures <2GPa increases by several orders of magnitude relative to pure H2O as the concentrations of ligands (e.g., F and Cl) in the fluid increase. Considering that prograde devolatilization in arcs begins at ∼300°C, there is a need for quantitative constraints on rutile solubility and the partitioning of HFSE between rutile and aqueous fluid over a wider range of temperature and pressure than is currently available. In this study, new experimental data are presented that quantify the solubility of rutile in aqueous fluids from 0.5 to 2.79GPa and 250 to 650°C. Rutile solubility was determined by using synchrotron X-ray fluorescence to measure the concentration of Zr in an aqueous fluid saturated with a Zr-bearing rutile crystal within a hydrothermal diamond anvil cell. At the PT conditions of the experiments, published diffusion data indicate that Zr is effectively immobile (logDZr ∼10−25m2/s at 650°C and ∼10−30m2/s at 250°C) with diffusion length-scales of <0.2μm in rutile for our run durations (<10h). Hence, the Zr/Ti ratio of the starting rutile, which was quantified, does not change during the experiment, and the measured concentration of Zr in the fluid was used to calculate the concentration of Ti (i.e., the solubility of rutile) in the fluid. The salts NaF, NaCl, and KCl were systematically added to the aqueous fluid, and the relative effects of fluid composition, pressure, and temperature on rutile solubility were quantified. The results indicate that fluid composition exerts the greatest control on rutile solubility in aqueous fluid, consistent with previous studies, and that increasing temperature has a positive, albeit less pronounced, effect. The solubility of Zr-rutile in aqueous fluid increases with the addition of halides in the following order: 2wt% NaF<30wt% KCl<30wt% NaCl<3wt% NaF<(10wt% NaCl+2wt% NaF)<4wt% NaF. The solubility of rutile in the fluid increases with the 2nd to 3rd power of the Cl− concentration, and the 3rd to 4th power of the F− concentration. These new data are consistent with observations from field studies of exhumed terranes that indicate that rutile is soluble in complex aqueous fluids, and that fluid composition is the primary control on rutile solubility and HFSE mobility.