Epigenetic gold deposits in metamorphic terranes include those of the Precambrian shields (approx 23,000–25,000 t Au), particularly the Late Archean greenstone belts and Paleoproterozoic fold belts, and of the late Neoproterozoic and younger Cordilleran-style orogens (approx 22,000 t lode and 15,500 t placer Au), mainly along the margins of Gondwana, Laurentia, and the more recent circum-Pacific. Ore formation was concentrated during the time intervals of 2.8 to 2.55 Ga, 2.1 to 1.8 Ga, and 600 to 50 Ma. Prior to the last 25 years, ores were defined by grades of 5 to 10 g/t Au in underground mines; present-day economics, open-pit
mining, and improved mineral processing procedures allow recovery of ores of ≤1 g/t Au, which has commonly
led to the recent reworking of lower grade zones in many historic orebodies. Most of these deposits formed
synchronously with late stages of orogeny and are best classified as orogenic gold deposits, which may be subdivided into epizonal, mesozonal, and hypozonal subtypes based on pressure-temperature conditions of ore formation.
A second type of deposit, termed intrusion-related gold deposits, developed landward of Phanerozoic accreted terranes in the Paleozoic of eastern Australia and the Mesozoic of the northern North American Cordillera. These have an overall global distribution that is still equivocal and are characterized by an intimate genetic association with relatively reduced granitoids. The majority of gold deposits in metamorphic terranes are located adjacent to first-order, deep-crustal fault zones, which show complex structural histories and may extend along strike for hundreds of kilometers with widths of as much as a few thousand meters. Fluid migration along such zones was driven by episodes of major pressure fluctuations during seismic events. Ores formed as vein fill of second- and third-order shears and faults, particularly at jogs or changes in strike along the crustal fault zones. Mineralization styles vary from stockworks and breccias in shallow, brittle regimes, through laminated crack-seal veins and sigmoidal vein arrays in brittle-ductile crustal regions, to replacement- and disseminated-type orebodies in deeper, ductile environments (i.e., a continuum model). Most orogenic gold deposits occur in greenschist facies rocks, but significant orebodies can be present in lower and higher grade rocks. Deposits typically formed on retrograde portions of pressure-temperature-time paths and thus are discordant to metamorphic features within host
rocks. Spatial association between gold ores and granitoids of all compositions reflects a locally favorable structural
trap, except in the case of the intrusion-related gold deposits where there is a clearer genetic association.
World-class orebodies are generally 2 to 10 km long, about 1 km wide, and are mined downdip to depths of 2 to 3 km. Most orogenic gold deposits contain 2 to 5 percent sulfide minerals and have gold/silver ratios from 5 to 10 and gold fineness >900. Arsenopyrite and pyrite are the dominant sulfide minerals, whereas pyrrhotite is more important in higher temperature ores and base metals are not highly anomalous. Tungsten-, Bi-, and Te-bearing mineral phases can be common and are dominant in the relatively sulfide poor intrusion-related gold deposits. Alteration intensity, width, and assemblage vary with the host rock, but carbonates, sulfides,muscovite, chlorite, K-feldspar, biotite, tourmaline, and albite are generally present, except in high-temperature systems where alteration halos are dominated by skarnlike assemblages. The vein-forming fluids for gold deposits in metamorphic environments are uniquely CO2 and 18O rich, with low to moderate salinities. Phanerozoic and Paleoproterozic ores show a mode of formation temperatures at 250° to 350°C, whereas Late Archean deposits cluster at about 325° to 400°C. However, there are also many important lower and higher temperature deposits deposited throughout the continuum of depths that range between 2 and 20 km. Ore fluids were, in most cases, near-neutral pH, slightly reduced, and dominated by sulfide
complexes. Globally consistent ore-fluid δ18O values of 6 to 13 per mil and δD values of –80 to –20 per mil
generally rule out a significant meteoric water component in the gold-bearing hydrothermal systems. Sulfur
isotope measurements on ore-related sulfide minerals are concentrated between 0 and 10 per mil, but with many higher and much lower exceptions, indicating variable sulfur sources and an unlikely dominant role for mantle sulfur. Drastic pressure fluctuations with associated fluid unmixing and/o rdesulfidation during water/rock interaction are the two most commonly called-upon ore precipitation mechanisms. The specific model(s) for gold ore genesis remains controversial. Although the direct syngenetic models of the 1970s are no longer applicable, the gold itself may be initially added into the volcanic and sedimentary crustal rock sequences, probably within marine pyrite, during sea-floor hydrothermal events. Gold transport
and concentration are most commonly suggested to be associated with metamorphic processes, as indicated by
the volatile composition of the hydrothermal fluids, the progressive decrease in concentration of elements enriched
in the gold deposits with increasing metamorphic grade of the country rocks, and the common association of ores with medium-grade metamorphic environments. Gold deposits of typically relatively low grade, which formed directly from fluid exsolution during granitoid emplacement within metamorphic rocks, are now also clearly recognized (i.e., intrusion-related gold deposits), but there are limited definitive data to implicate such an exsolved fluid source for most gold deposits within orogenic provinces. The fact that orogenic gold deposits are associated with all types of igneous rocks is a problem to a pure magmatic model. Hybrid models, where slab-derived fluids may generate rising melts that drive devolatilization reactions in the lower crust, are also feasible. Although involvement of a direct mantle fluid presents geochemical difficulties, the presence of lamprophyres and deep-crustal faults in many districts suggests potential mantle influence in the overall, largescale tectonic event controlling the hydrothermal flow system.