Magma evolution and the formation of porphyry Cu–Au ore fluids: evidence from silicate and sulfide melt inclusions

Silicate and sulfide melt inclusions from the andesitic Farallón Negro Volcanic Complex in NW Argentina were analyzed by laser ablation ICPMS to track the behavior of Cu and Au during magma evolution, and to identify the processes in the source of fluids responsible for porphyry-Cu-Au mineralization at the 600 Mt Bajo de la Alumbrera deposit. The combination of silicate and sulfide melt inclusion data with previously published geological and geochemical information indicates that the source of ore metals and water was a mantle-derived mafic magma that contained approximately 6 wt.% H2O and 200 ppm Cu. This magma and a rhyodacitic magma mixed in an upper-crustal magma chamber, feeding the volcanic systems and associated subvolcanic intrusions over 2.6 million years. Generation of the ore fluid from this magma occurred towards the end of this protracted evolution and probably involved six important steps: (1) Generation of a sulfide melt upon magma mixing in some parts of the magma chamber. (2) Partitioning of Cu and Au into the sulfide melt (enrichment factor of 10,000 for Cu) leading to Cu and Au concentrations of several wt.% or ppm, respectively. (3) A change in the tectonic regime from local extension to compression at the end of protracted volcanism. (4) Intrusion of a dacitic magma stock from the upper part of the layered magma chamber. (5) Volatile exsolution and resorption of the sulfide melt from the lower and more mafic parts of the magma chamber, generating a fluid with a Cu/Au ratio equal to that of the precursor sulfide. (6) Focused fluid transport and precipitation of the two metals in the porphyry, yielding an ore body containing Au and Cu in the proportions dictated by the magmatic fluid source. The Cu/S ratio in the sulfide melt inclusions requires that approximately 4,000 ppm sulfur is extracted from the andesitic magma upon mixing. This exceeds the solubility of sulfide or sulfate in either of the silicate melts and implies an additional source for S. The extra sulfur could be added in the form of anhydrite phenocrysts present in the rhyodacitic magma. It appears, thus, that unusually sulfur-rich, not Cu-rich magmas are the key to the formation of porphyry-type ore deposits. Our observations imply that dacitic intrusions hosting the porphyry–Cu–Au mineralization are not representative of the magma from which the ore-fluid exsolved. The source of the ore fluid is the underlying more mafic magma, and unaltered andesitic dikes emplaced immediately after ore formation are more likely to represent the magma from which the fluids were generated. At Alumbrera, these andesitic dikes carry relicts of the sulfide melt as inclusions in amphibole. Sulfide inclusions in similar dykes of other, less explored magmatic complexes may be used to predict the Au/Cu ratio of potential ore-forming fluids and the expected metal ratio in any undiscovered porphyry deposit.