Chlorine partitioning between a basaltic melt and H2O-CO2 fluids at Mount Etna

Partitioning experiments between a basaltic melt from Mt. Etna and a low-density hydrous fluid or vapor containing H2O or H2O-CO2 were performed at 1200-1260 °C, at pressures between 1 and 200 MPa, either near the nickel-nickel oxide (NNO) buffer or at two log units above it (NNO+2), and with different chloride concentrations. Most of the experiments were done at chloride-brine-undersaturated conditions, although at the highest Cl concentrations explored brine saturation might have been reached. The average partition coefficients (DClfluid/melt) over the range of Cl concentrations were derived on a weight basis by plotting the calculated concentrations of Cl in the fluid phase versus the measured ones in the melt. For H2O-Cl experiments in which the Cl concentration in the melt was ≤ 0.4 wt.%, a negative dependence between DClf/m and pressure is observed. DClfluid/melt in H2O+Cl-bearing experiments ranges between 11-14 at 1 and 25 MPa to 6 at 200 MPa at NNO; and between 4 at 50 MPa and 13 at 100 MPa at ΔNNO ≥ 2. Addition of CO2 at NNO yields lower partition coefficients than in CO2-free conditions over the pressure range investigated. The negative pressure dependence observed for H2O-Cl experiments disappears when CO2 is present in the system. Overall, once CO2 is introduced in the system, Cl fugacity in the silicate melt tends to increase, thus resulting in a decrease of DClf/m. Application to Mt. Etna shows that the composition of the volcanic plume in terms of Cl records very shallow pressures of magma equilibration with its exsolved fluid.