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 H(2)O or H(2)O-CO(2) were performed at 1200-1260 degrees 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 (D(Cl)(fluid/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 H(2)O-Cl experiments in which the Cl concentration in the melt was <= 0.4 wt.%, a negative dependence between D(Cl)(f/m) and pressure is observed. D(Cl)(fluid/melt) in H(2)O + 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 Delta NNO >= 2. Addition of CO(2) at NNO yields lower partition coefficients than in CO(2)-free conditions over the pressure range investigated. The negative pressure dependence observed for H(2)O-Cl experiments disappears when CO(2) is present in the system. Overall, once CO(2) is introduced in the system, Cl fugacity in the silicate melt tends to increase, thus resulting in a decrease of D(Cl)(f/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.