The magmatic to hydrothermal transition in the Late Cretaceous Elatsite porphyry Cu-Au-(Mo-platinum group element) deposit has been studied in a suite of samples with clear timing relations between porphyry dikes, magmatic-hydrothermal veins, silicate melt inclusions in quartz veins, fluid inclusion generations, and ore minerals. Ore mineralization occurs late, at temperatures ~200° to 300°C below those of the early, multistage interplay between magmatic and hydrothermal processes at near-magmatic temperature-pressure conditions.
Crosscutting relations and petrography indicate that shortly after the intrusion of the earliest, monzodioritic dikes, fluids precipitated a first generation of granular quartz veins accompanied by potassic alteration. The second vein generation with crystalline quartz textures and K-feldspar alteration halos formed at the same time as the second, granodioritic pulse of porphyry intrusions. Cathodoluminescence imaging of quartz growth textures reveals that the earliest fluid inclusions in the crystalline quartz veins are of intermediate density and ~8 wt % NaCl equiv salinity, probably trapped at near-lithostatic pressures of ~1,200 to 1,300 bars at near-magmatic temperatures (≤730°C), and implies a depth of ~4 to 5 km. Depressurization led to phase separation, indicated by a first generation of coexisting brine and vapor inclusions trapped at temperatures of ≥640°C and suprahydrostatic pressures of ≥920 bars. A second quartz generation in crystalline quartz veins precipitated during progressive depressurization and hosts assemblages of coexisting brine, vapor, and silicate melt inclusions, trapped at temperatures in excess of 600°C and suprahydrostatic pressures of 630 to 880 bars. Some open-spaced quartz veins were filled with aplite during this stage. Field relations and geochemical evidence suggest that the aplites as well as the silicate melt inclusions in hydrothermal quartz veins represent volumetrically minor residual melts that evolved directly from granodiorite porphyries at the level of deposit formation, and do not represent aliquots of metal-supplying magma at depth. Fluid inclusions coexisting with the silicate melt inclusions are metal rich, but these fluids predate sulfide precipitation and are, therefore, not the dominant fluids responsible for the Cu-Au mineralization at Elatsite. Melt-fluid-metal separation processes recorded in these co-trapped silicate melt and fluid inclusions in vein quartz are small-scale, local phenomena and do not appear to be suited for understanding and quantifying metal segregation in the deeper source.
Microthermometry data of fluid inclusions, trapped during the later stages of the formation of the second quartz generation, show further depressurization to ~260 to 325 bars and cooling to ~460°C, representing a hot hydrostatic regime. Bornite, chalcopyrite, and magnetite seem to be slightly later precipitated at temperatures ≤460°C in separate veins or opened spaces of preexisting veins. Dissolution textures of the second quartz generation indicate a subsequent local redissolution of vein quartz, most likely as a result of passing through a window of retrograde quartz solubility upon further cooling below 460°C. The next, economically most important chalcopyrite-pyrite stage is largely devoid of quartz precipitation and is mostly expressed as “paint veins.” Due to the lack of fluid inclusions, the temperature-pressure conditions of formation of these veins could not be constrained. The waning phase of hydrothermal activity is represented by a quartz-carbonate-zeolite stage, formed at low temperature (~145°C), as indicated by microthermometry of fluid inclusions trapped in quartz from this stage.
Laser ablation-inductively coupled plasma-mass spectrometry analyses of fluid inclusions show very high Cu contents in early intermediate-density fluid inclusions and in the first vapor inclusion generation, followed by a drastic decrease in the second generation of vapor inclusions. The decrease correlates with the appearance of anhydrite inclusions in the second quartz generation and may indicate that a lack of sulfur in these later vapor inclusions led to Cu partitioning into the brine. Alternatively, it is also possible that, unlike early vapor inclusions, the later vapor inclusions were not susceptible to postentrapment copper enrichment, in accordance with recent experiments. A progressive Cu enrichment in the brine phase correlates well with depressurization prior to mineralization. Mass balance considerations based on analyzed fluid components and fluid phase relations indicate that brine was volumetrically minor and therefore likely stagnant, but this may have prepared ore precipitation by accumulating Cu stripped from ascending vapor.