The effects of pressure, fO2, fS2 and melt composition on the fluid–melt partitioning of Mo: Implications for the Mo-mineralization potential of upper crustal granitic magmas

Abstract

In order to better constrain the partitioning behavior of Mo during the crystallization of upper crustal granitic magmas, partition coefficients of Mo between aqueous fluids and granitic melts were experimentally determined. The experiments were conducted at 750–850 °C, 80–300 MPa, and fO2 + fS2 conditions being controlled either by the magnetite–pyrrhotite assemblage and an external NiNiO buffer, or fully internally by the magnetite–pyrrhotite–pyrite assemblage. During the experiments, fluids that were equilibrated with silicate melt, molybdenite and the internal buffer minerals were trapped in the form of synthetic fluid inclusions in quartz. Through analyzing the recovered synthetic fluid inclusions and run product glasses by LA-ICP-MS, the abundances of Mo in the fluids and the melts were determined, and corresponding fluid–melt partition coefficients (DMofluid/melt) were calculated. The DMofluid/melt values obtained in our experiments with supercritical fluids (DMosf/melt; ±1σ) range from 9 ± 2 to 54 ± 9 and vary as function of pressure, melt composition, and fO2 + fS2. In slightly peraluminous, magnetite–pyrrhotite-buffered melts at 750 °C, DMosf/melt increases from 15 ± 5 at 160 MPa to 34 ± 6 at 300 MPa. In contrast, over the range of 750 °C to 850 °C, no significant temperature effect was observed. Vapor–melt (DMovapor/melt; ±1σ) and brine–melt (DMobrine/melt; ±1σ) partition coefficients determined in two experiments with subcritical fluids are 4 ± 1 vs 58 ± 8, and 9 ± 2 vs 117 ± 15, respectively, suggesting that Mo favors both fluid phases relative to the silicate melt and partitions more strongly into brine relative to vapor. Most of our experimentally determined DMofluid/melt values agree well with those derived from coexisting fluid and melt inclusions in natural granitic samples and those found in a previous experimental study conducted on sulfur-bearing systems, but they are generally higher than those reported in previous experimental studies conducted on sulfur-free systems. Based on the observed DMofluid/melt values a quantitative Rayleigh fractionation model for the behavior of Mo during the crystallization of upper crustal granitic magma chambers was developed. The results suggest that the confining pressure exerts a strong effect on the evolution of Mo concentration in the residual melt during fractional crystallization. Specifically, magmas that crystallize at deeper levels (i.e., higher confining pressure) become more strongly depleted in Mo during fluid-saturated crystallization, which means that the Mo extraction efficiency of deeply exsolving fluids is higher. We conclude that Mo-mineralization potential of upper crustal granitic magmas may be critically affected by the magma emplacement depth.