The partitioning of Sn between a mildly peraluminous granitic melt and saline aqueous fluids was studied at 800 °C and 150 MPa for oxygen fugacities ranging from the Fe-FeO to the Fe3O4-Fe2O3 buffer. Experiments were carried out in rapid-quench cold-seal pressure vessels using water as pressure medium. Oxygen fugacity was buffered by double capsules, except for oxygen fugacities about 0.5 to 1 log units above the Ni-NiO buffer, which correspond to the intrinsic redox conditions imposed by the Ni alloy vessel in contact with water. Run durations were 3–7 days. Fluid/melt partition coefficients were determined by analyzing fluid inclusions and neighboring quenched glass compositions by laser-ablation ICP-MS, assuming that they represent local equilibrium. At all redox conditions studied, the fluid/melt partition coefficient increases roughly linearly with salinity. The highest partition coefficients are observed 0.5 to 1 log units above the Ni-NiO buffer, where DSnfluid/melt = 22.8 is reached for a salinity of 25 eq. NaCl in the fluid. Towards more oxidizing conditions (Re-ReO2 and Fe3O4-Fe2O3 buffer), the partition coefficient decreases, by about a factor of five for the Fe3O4-Fe2O3 buffer. This probably reflects the transition from the fluid-soluble Sn2+ to the less soluble Sn4+. However, at redox conditions that are more reducing than Ni-NiO, DSnfluid/melt also strongly decreases. This may be due to the increasing abundance of non-polar H2 in the fluid, which reduces solvation of polar species. Moderately reducing conditions near the Ni-NiO buffer are therefore optimal for extracting tin out of a residual melt, as they correspond to a maximum in the fluid/melt partition coefficient. However, the association of hydrothermal tin deposits with reduced granites is likely also due to the effect of oxygen fugacity on the partitioning of tin between minerals and silicate melt. Indeed, observations made in the course of this study suggest that the solubility of cassiterite SnO2 in silicate melts under oxidizing conditions (Fe3O4-Fe2O3 buffer) is even much lower than previously thought. Overall, the results show that tin extraction by magmatic fluids is rather inefficient and tin enrichment requires extensive fractional crystallization. A relatively efficient way to extract tin out of a granitic magma may be fluid release by decompression due to the formation of fractures followed by rapid crystallization induced by water-loss from the melt.
Keppler et al. (Wed,) studied this question.