DOI: 10.1093/petrology/egag097 ISSN: 0022-3530

Experimental Differentiation of a H2O-CO2-S-Cl-Bearing Arc Basalt at 750-1250 °C and 0.4-1.0 GPa, and Implications for the Formation of Porphyry Copper Deposits

Yi-Bo Qiao, Andreas Audétat

Abstract

Porphyry copper deposits tend to be associated with differentiated arc magmas characterized by high Sr/Y ratios, which feature is thought to result from magma fractionation at high pressure. However, significant debate persists regarding the question which specific magma property/properties rendered them so fertile. Potential candidates are: (1) higher Cu content, (2) higher S content, (3) higher Cl content, (4) higher H2O content. To understand how pressure affects these factors during magma fractionation we conducted crystallization experiments on a natural arc basalt doped with 3.5 wt% H2O, 3200 μg/g CO2, 2200 μg/g S, 1000 μg/g Cl, and 200 μg/g Cu, which concentrations are geologically realistic. The experiments were performed at 750-1250 °C and 0.4, 0.7 and 1.0 GPa using piston cylinder presses. In order to obtain minerals and melt pools that are large enough to be analyzed by LA-ICP-MS and FTIR the samples were first heated to 1100-1250 °C and then slowly cooled to the target temperature. Loss of sulfur, iron and carbon to the PtRh capsule was minimized by using capsule liners made of gem-quality zircon crystals. Oxygen fugacity was buffered internally at ~2.3±0.7 log units above the fayalite-quartz-magnetite fO2 buffer through the coexistence of anhydrite and monosulfide solid solution. To verify the results, a set of isothermal experiments was performed without zircon liners in Au capsules. Modal mineral abundances were quantified via SEM-EDS mapping, and minerals and quenched silicate melt were analyzed by EPMA, LA-ICP-MS and FTIR. The slow-cooling experiments produced crystals that are up to several hundred micrometers in size and show textures resembling those of natural samples. The crystallization sequence at 0.7 and 1.0 GPa was spinel + clinopyroxene → amphibole + magnetite + anhydrite + sulfide → plagioclase + apatite, whereas at 0.4 GPa plagioclase started to crystallize before amphibole. Correspondingly, evolved residual melts reached high Sr/Y ratios at 1.0 GPa, but retained low Sr/Y ratios at 0.4 GPa, confirming previous predictions. The CO2 and S concentrations in the residual melts increased until the melt reached ~55 wt% SiO2, after which point they decreased rapidly. In the case of CO2 the decline is due to exsolution of CO2-rich fluid, whereas in the case of S it is due to the precipitation of anhydrite + sulfide. Copper concentrations declined already earlier due to Cu-loss to the noble capsule despite the use of zircon liners. In contrast, H2O and Cl concentrations in the melt increased until ~65-70 wt% SiO2, after which point they remained more or less stable (H2O) or declined (Cl). The highest Cl concentrations were reached at low pressure, whereas the highest H2O concentrations were reached at high pressure. We conclude that the high mineralization potential of high-Sr/Y magmas is primarily a consequence of unusually high H2O contents, which causes the ascending magmas to stop and accumulate at the optimal depth level of ~10-15 km. This depth range seems to be required for the generation of fluid-focusing structures such as porphyry fingers and breccia pipes without resulting in catastrophic volcanic eruptions.

More from our Archive