Search

Textural, chemical and multiple sulfur isotope analyses of sulfides from the Archean world-class Cuiaba orogenic Au deposit, Quadrilatero Ferrifero district, Brazil, were conducted to track distinct trace element signatures from syngenetic to epigenetic-hydrothermal pyrite types in different host units, and to deduce the nature and source of the mineralizing fluids. Gold mineralization in Cuiaba is hosted in greenschist facies volcanic and sedimentary host rocks belonging to the Archean Rio das Velhas Supergroup, especially in association with quartz veins and sulfide minerals. The BIF-hosted Fonte Grande Sul orebody is controlled by quartz veins and disseminated gold-bearing pyrite, pyrrhotite and arsenopyrite in the wall rocks. Five pyrite types are classified based on textural relations as spongy, syngenetic (Py1), porous early-(Py2), smooth main-(Py3) and smooth isolated and overgrown late-stage (Py4 and Py5) types variably present in metamorphosed carbonaceous pelites, BIF and andesite . Besides pyrite, pyrrhotite and arsenopyrite are also defined according to textural and paragenetic differences, and they display significant trace element contents. Trace element abundance maps and LA-ICP-MS analyses display trace element incorporations in Py1 (present only in carbonaceous pelite), Py2 and Py3 (in all host rocks), and Py4 and Py5 (only in BIF). Py1 yields high As, Co, Ni, Pb and Ag concentrations, whereas Py2 (formed by agglomeration of Py1) maintains high trace element concentrations with slightly less of those elements. Trace element incorporation in Py3 is similar to Py1 in carbonaceous pelite. Early-stage Py2 and main-stage Py3 in BIF and andesite have increased Co and Ni, but have less Au and As. Late-stage Py4 is characterized by lower trace element concentrations, whereas Py5 is further enriched in As, Bi, Co, Ni and Pb. The hydrothermal alteration is divided into early-, main- and late-stage, which correlate to the sulfide hydrothermal evolution. The study shows that carbonaceous pelite is pre-enriched in Co, Ni and Pb, whereas certain elements like Ag, Au, Bi and As are only hydrothermally concentrated either during the early, main or late stage in BIF and andesite. This geochemical pattern supports a syngenetic versus hydrothermal origin of distinct elements, including some trace elements now in the ore zones. Multiple sulfur isotope data suggest that mineralizing fluids at the Cuiaba gold deposit reflect a complex mixing of sulfur evolved from three possible sources: seawater, mantle, and reduced elemental sulfur. Spongy, syngenetic Py1 yields Δ 33S values ranging from −2.28 to −0.25‰, separated into two ranges, the first from −2.28 to −1.97‰, and the second from −0.96 to −0.25‰. The first range suggests Py1 in the carbonaceous pelite was deposited in a seawater environment. These pyrites probably mixed with later mantle sulfur carrying a near-zero Δ33S signature, which is indicated by the second Δ33S range, or could have mixed with fluids sourced from sedimentary rocks at depth carrying a positive Δ33S signature. The second scenario is more common in sedimentary and diagenetic pyrites. An unequivocal contrast is shown by early-, main- and late-stage pyrite types (Py2, Py3, Py4 and Py5) in carbonaceous pelite, BIF and andesite, which present a continuous process of crustal assimilation values towards increasingly positive δ34S and Δ33S values. This confirms evidence for a sedimentary-derived sulfur source as also indicated by fluid inclusion data of the andesite-hosted orebody at the Cuiaba deposit.

Archean alkali-rich (Syenitic Group) intrusions in the Eastern Goldfields Superterrane, Yilgarn Craton, are chemically similar to intrusions associated with Phanerozoic alkali porphyry systems. High-temperature potassic and ferruginous alteration associated with the Yilgarn intrusions are also similar to those in Phanerozoic counterparts but these are generally not well mineralized. Disseminated and fracture-controlled gold mineralization associated with Syenitic Group intrusions is commonly hosted by mesothermal albite-carbonate-pyrite assemblages, the origin of which is ambiguous. This communication aims to stimulate discussion on the origin of gold in and around Syenitic Group intrusions and the apparent absence of significant gold or copper mineralization in associated high-temperature alteration. Gold associated with Syenitic Group intrusions in the Yilgarn Craton has many similar characteristics to disseminated gold deposits associated with syenitic intrusions in the Abitibi greenstone belt, Canada, and may be present in other Archean greenstone belts.

The Picarrao and Liberdade deposits contain high-grade iron orebodies (>65% Fe) hosted in the Guanhaes Group itabirite, that are associated with pegmatite veins and bodies. Fluid inclusion studies in quartz veins associated with the high-grade orebodies show that medium to high salinities (25–28 wt% NaCl eq.) and temperatures (275–375 °C) fluids are associated with the silica leaching that led to the iron enrichment. Mineral chemistry studies by LA-ICP-MS in the iron oxides demonstrate that metasomatic processes were responsible for the mineralogical transformations of magnetite to hematite and for subsequent hematite recrystallization. These processes are related to the iron upgrade in the itabirite and the formation of high-grade orebodies. The oxidation of the magnetite to martite is associated with an enrichment in P and As, and depletion in Mg, Ti and Co; as observed in martite crystals compared to their matching kenomagnetite rims. On the other hand Ti and Mo are enriched in hematite crystals that recrystallized from martite. In this case Ti behaved as an immobile element, and its enrichment is accompanied by the depletion of most of the trace elements. A second stage of magnetite formation precipitated with quartz in discordant veins and is oxidized to martite-II. These quartz-martite-II veins contain low salinity and temperature fluid inclusions that record an episode of meteoric fluid influx. The results of the LA-ICP-MS analyses on the fluid inclusions from pegmatite and quartz veins associated with the high-grade iron bodies indicate the contribution of anatectic fluids in the evolution of the metasomatic events.

Iron ore deposits hosted by Archean banded iron-formation (BIF) in the Weld Range greenstone belt are representative of most of the documented iron ore deposits in the Yilgarn Craton. They include near-surface, supergene goethite-hematite orebodies that overlie and partly modify deeper occurrences of hypogene magnetite and specular hematite ores. The Cenozoic goethite-hematite-rich orebodies in these deposits are unequivocally the product of meteoric fluid alteration affecting BIF in the near-surface supergene environment; however, the deeper and likely older magnetite- and specular hematite-rich orebodies have a more contentious origin. This study is the first to present a fluid-alteration model for hypogene iron mineralization in the Yilgarn Craton that uses fluid inclusion and mineral chemistry data to constrain the physical-chemical characteristics and source of hypogene fluids responsible for mineralization. High-grade (>57 wt% Fe), magnetite-rich iron ore at the Beebyn deposit defines a discontinuous series of 440 °C) and CO2-rich. Paired O and C stable isotope data for Stage 1 ferroan dolomite suggest that these fluids had a magmatic source, while Stage 1 magnetite chemistry (e.g. enrichments in Mg, Mn, Ca, and Na) indicates chemical exchange took place between the fluids and mafic igneous rocks prior to crystallization of magnetite. The presence of monophase carbonic fluid inclusions in Stage 1 ferroan dolomite suggest that phase separation of a bicarbonate-rich aqueous fluid took place in deeper parts of the hydrothermal system, which led to the separation of the resultant volatile-rich and brine phases during transport of the Stage 1 fluid to shallower crustal levels. Cooling of the hydrothermal system during the Stage 2 fluid event involved (i) an early brine (>275–327 °C; 36–40 equiv. wt% NaCl) with Cl/Br and C and O isotopes values that overlap the ranges for magmatic fluids, with minor involvement of Archean seawater, and extensive chemical exchange with country rocks; followed by (ii) pulses of moderate- and lower-temperature Stage 2 brines (>125–260 °C; 2–24 equiv. wt% CaCl2) with Cl/Br, O and C isotope, Na/Br, and Ca/Ca + Na signatures that suggest cooling of magmatic-derived fluids that mixed with Archean seawater and reacted with mafic igneous country rocks in areas more distal to fluid pathways. The last stage of formation of magnetite-rich ore at the Beebyn deposit involved the flow of Stage 3 fluids through the existing fault network that controlled earlier fluids. Stage 3 fluids are lower-temperature (>98–175 °C), low to high-salinity brines with Cl/Br values that overlap reported ranges for 3.2 Ga vent fluids and seawater. Thus, they are likely the product of heated Archean seawater that was chemically-modified through interaction with mafic country rocks. Magnetite-talc veins at the Madoonga deposit have fluid halogen ratios for fluid inclusions hosted by magnetite that are compatible with a range of possible sources, including low-grade metamorphic fluids, geothermal brines, or oil field formation waters. Although, their common spatial association with semi-massive sulfides suggests their likely precipitation from brines derived from heated Archean seawater. More locally developed specular hematite-quartz veins that cut folded magnetite-rich ores at the Beebyn and Madoonga deposits are the product of iron redistribution in BIF by heated meteoric fluids or seawater, with precipitation of specular hematite as a consequence of oxidation and cooling of the fluid.

“Magmato-Hydrothermal Space” is a new mathematical construct where 24 ore and pathfinder elements are used to quantify ore-element signatures. Quantification allows relationships between ore deposit samples to be mapped. The broad-scale view of Magmato-Hydrothermal Space reveals three important trends: (1) Zn-Pb sediment-hosted mineralisation to igneous-associated Cu-Au mineralisation, (2) Cu-Au mineralisation to Au-only mineralisation, and (3) ultramafic associated magmatic Ni-Cu-PGE mineralisation through Cu-Au mineralisation to granitoid-associated Mo, W and Sn mineralisation. The view provided by Magmato-Hydrothermal Space reveals that there is a spectrum of ore element signatures that mirrors the spectrum of ore deposit classes described in the literature. Geochemical variations within individual ore deposit classes are examined for orogenic-Au, VHMS, epithermal and sediment-hosted Cu samples. Sub-groups within each of these classes are compared on element enrichment diagrams and described in the context of Magmato-Hydrothermal Space. Orogenic Au samples are divided into two sub-groups of As-Sb rich mineralisation and four sub-groups of relatively As-Sb poor mineralisation. The As-Sb poor sub-groups include a Te-Cu-Ag rich sub-group that overlaps with the porphyry Cu class and a Te-W-(Bi-Mo) rich class dominated by granitoid hosted deposits. The VHMS class ranges from a Cu-rich sub-group that overlaps with porphyry Cu and IOCG classes through a Cu-Zn sub-group to two Zn-Cu sub-groups, one of which overlaps with the SHMS class. The epithermal class is divided into Zn-rich, Cu-rich and base metal – poor sub-groups.

The concept of a mineral system was first formally introduced in 1994, two decades after the analogous petroleum system concept had been developed. Mineral systems involve: (1) tectonic events that trigger and define temporal windows for mineralising events, (2) geochemical and tectonic processes that produce ‘source regions’ often enriched in metals and hydrothermal fluids and magmas capable of mobilising these metals, (3) tectonic, sedimentary, diagenetic and structural processes that produce conduits that can act as pathways for fluid/magma flow at lithospheric, crustal, province and district scales, (4) processes that drive fluid flow, (5) focussing mechanism that concentrate (or throttle) the flow of fluids or magmas into depositional ‘trap’ sites at the camp to deposit scales, (6) chemical and physical processes that cause metal deposition at the trap site, (7) post-depositional geochemical processes that produce geochemical and geophysical features that allow later detection, and (8) processes that enhance exhumation, preservation and upgrading of mineralised material. Although mineralising events are generally short-lived, important features of mineral systems can develop well before, or well after the mineralising event, and the evolution of a mineral system can extend over hundreds of millions to billions of years. For example, the architecture of mineral systems, which controls the location of source regions and conduits, is commonly established (long) before the mineralising event, and exhumation and upgrading through metamorphism and/or supergene enrichment occur (long) after the mineralising event. The advantage of the mineral system approach is that it focusses on the critical geological processes necessary to form a major mineral deposit and is not restricted to descriptive elements of a specific mineralisation style. Mineral systems can explain the spatial and temporal co-occurrence of mineral deposits within a specific mineral province. The mineral system concept can also explain families of coeval mineral systems that potentially formed in the same or adjacent terranes or provinces. Presently, mineral systems are broadly subdivided into orthomagmatic and hydrothermal classes, with several transitional systems between these end-members not well defined (e.g., sediment-hosted deposits with hydrothermal overprints or enrichment). Coeval, or nearly coeval, mineral systems, such as IOCG and porphyry Cu–Au systems in Chile or orogenic Au and Cu–Au porphyry systems in the Lachlan Orogen in eastern Australia, can develop during the tectonic evolution of a given terrane. Overprinting mineral systems, potentially separated by tens of millions of years, can also develop where products of early-formed mineralising events (e.g. syn-volcanic or syn-sedimentary events) can be overprinted at a later stage by mineralisation formed during orogenesis or basin inversion (e.g., orogenic Au system). The mineral system concepts and models can be translated into effective exploration targeting criteria and serve as the basis of robust, testable exploration models. Arguably, this is the main practical purpose for a systematic mineral system analysis and its main advantage over traditional descriptive deposit models. This advantage is enhanced in under-explored areas, particularly covered areas with poor or no outcrop, where known existing deposits, from which empirical models are developed, are lacking. Defining and mapping expressions of the critical elements of mineral systems can effectively focus exploration targeting. Major deep-crustal domain boundaries, often overlain by younger geological provinces and not clearly expressed in many traditional geological datasets, provide steep crustal zones of focused regional fluid flow for many orthomagmatic and hydrothermal mineral systems. Significant challenges in the development of mineral system models and their effective practical applications remain. For example, there is not enough information on the tectonic setting of productive mineral systems and the existing information is commonly conflicting. This is often caused by a short-lived transient nature of specific tectonic conditions favourable for triggering and driving a productive mineral system, exacerbated by a lack of outcrop and/or lack of high resolution geophysics, geochemistry and geochronology in poorly explored geological terranes. An important challenge that mineral system geoscience can help to address is exploration area selection at the district to camp scale. At larger scales, conceptual understanding of the relationships between tectonism and metallogeny can identify potentially mineralised provinces, and at smaller scales, well established empirical relationships between deposits and geochemical and geophysical anomalies provide vectors to ore at the camp to deposit scale. It is the intermediate, district to camp scale where both conceptual- and empirical-based targeting commonly fail, providing a challenge to mineral exploration, particularly in greenfields terranes. Finally, mostly end-member hydrothermal and orthomagmatic mineral systems have received much attention in recent years; sedimentary deposits (e.g., placers, phosphorites) have yet to be defined within the mineral system framework. However, recent developments in the mineral system geoscience provide an increasingly robust framework for more effective predictive regional exploration targeting.

Tectono-metallogenic systems are geological systems that link geodynamic and tectonic processes with ore-forming processes. Fundamental geodynamic processes, including buoyancy-related processes, crustal/lithospheric thinning and crustal/lithospheric thickening, have occurred throughout Earth's history, but tectonic systems, which are driven by these processes, have evolved as Earth's interior has cooled. Tectonic systems are thought to have evolved from magma oceans in the Hadean through an unstable “stagnant-lid” regime in the earlier Archean into a proto-plate tectonic regime from the late Archean onwards. Modern-style plate tectonics is thought to have become dominant by the start of the Paleozoic. Mineral systems with general similarities to modern or geologically recent systems have been present episodically (or semi-continuously) through much of Earth's history, but most of Earth's present endowment of mineral wealth was formed during and after the Neoarchean, when proto- or modern-style plate tectonic systems became increasingly dominant and following major changes in the chemistry of the atmosphere and hydrosphere. Changes in the characteristics of some mineral systems, such as the volcanic-hosted massive sulphide (VHMS) system, reflect changes in tectonic style during the evolution towards the modern plate tectonic regime, but may also involve secular changes in the hydrosphere and atmosphere. Whereas tectono-metallogenic systems have evolved in general over Earth's history, specific tectono-metallogenic systems evolve over much shorter time frames. Most mineral deposits form in three general tectono-metallogenic systems: divergent systems, convergent systems, and intraplate systems. Although fundamental geodynamic processes have driven the evolution of these systems, their relative importance may change as the systems evolved. For example, buoyancy-driven (mantle convection/plumes) and crustal thinning are the dominant processes driving the early rift stage of divergent tectono-metallogenic systems, whereas buoyancy-driven processes (slab sinking) and crustal thickening are the most important processes during the subduction stage of convergent systems. Crustal thinning can also be an important process in the hinterland of subduction zones, producing back-arc basins that can host a number of mineral systems. As fundamental geodynamic processes act as drivers at some stage in virtually all tectonic systems, on their own these cannot be used to identify tectonic systems. Moreover, as mineral systems are ultimately the products of these same geodynamic drivers, individual mineral deposit types cannot be used to determine tectonic systems, although mineral deposit associations can, in some cases, be indicative of the tectono-metallogenic system. Ore deposits are the products of geological (mineral) systems that operate over a long time frame (hundreds of millions of years) and at scales up to the craton-scale. In essence, mineral systems increase the concentrations of commodities through geochemical and geophysical processes from bulk Earth levels to levels amenable to economic mining. Mineral system components include the geological (tectonic and architectural) setting, the driver(s) of mineralising processes, metal and fluid sources, fluid pathways, depositional trap, and post-depositional modifications. All of these components link back to geodynamic processes and the tectonic system. For example, crustal architecture, which controls the spatial distribution of, and fluid flow, within mineral systems, is largely determined by geodynamic processes and tectonic systems, and the timing of mineralisation, which generally is relatively short (commonly The geochemical characteristics of many mineral systems are a consequence of their geodynamic and tectonic settings. Settings that are characterised by low heat flow and lack active magmatism produce low temperature fluids that are oxidised, with ore formation caused largely by redox gradients or the provision of external H 2 S. The characteristics of these fluids are largely governed by the rocks with which they interact, rocks that have extensively interacted with the hydrosphere and atmosphere, both environments that have been strongly oxidised since the great oxidation event in the Paleoproterozoic. In settings characterised by high heat flow and active magmatism, ore fluids tend to be higher temperature and reduced, with deposition caused by cooling, pH neutralisation, depressurisation and fluid mixing. Again, the characteristics of these fluids are governed by rocks with which they interact, in this case more reduced magmatic rocks derived from the mantle or lower crust.

The BIF-hosted iron ore system represents the world's largest and highest grade iron ore districts and deposits. BIF, the precursor to low- and high-grade BIF hosted iron ore, consists of Archean and Paleoproterozoic Algoma-type BIF (e.g., Serra Norte iron ore district in the Carajas Mineral Province), Proterozoic Lake Superior-type BIF (e.g., deposits in the Hamersley Province and craton), and Neoproterozoic Rapitan-type BIF (e.g., the Urucum iron ore district). The BIF-hosted iron ore system is structurally controlled, mostly via km-scale normal and strike-slips fault systems, which allow large volumes of ascending and descending hydrothermal fluids to circulate during Archean or Proterozoic deformation or early extensional events. Structures are also (passively) accessed via downward flowing supergene fluids during Cenozoic times. At the depositional site the transformation of BIF to low- and high-grade iron ore is controlled by: (1) structural permeability, (2) hypogene alteration caused by ascending deep fluids (largely magmatic or basinal brines), and descending ancient meteoric water, and (3) supergene enrichment via weathering processes. Hematite- and magnetite-based iron ores include a combination of microplaty hematite–martite, microplaty hematite with little or no goethite, martite–goethite, granoblastic hematite, specular hematite and magnetite, magnetite–martite, magnetite-specular hematite and magnetite–amphibole, respectively. Goethite ores with variable amounts of hematite and magnetite are mainly encountered in the weathering zone. In most large deposits, three major hypogene and one supergene ore stages are observed: (1) silica leaching and formation of magnetite and locally carbonate, (2) oxidation of magnetite to hematite (martitisation), further dissolution of quartz and formation of carbonate, (3) further martitisation, replacement of Fe silicates by hematite, new microplaty hematite and specular hematite formation and dissolution of carbonates, and (4) replacement of magnetite and any remaining carbonate by goethite and magnetite and formation of fibrous quartz and clay minerals. Hypogene alteration of BIF and surrounding country rocks is characterised by: (1) changes in the oxide mineralogy and textures, (2) development of distinct vertical and lateral distal, intermediate and proximal alteration zones defined by distinct oxide–silicate–carbonate assemblages, and (3) mass negative reactions such as de-silicification and de-carbonatisation, which significantly increase the porosity of high-grade iron ore, or lead to volume reduction by textural collapse or layer-compaction. Supergene alteration, up to depths of 200 m, is characterised by leaching of hypogene silica and carbonates, and dissolution precipitation of the iron oxyhydroxides. Carbonates in ore stages 2 and 3 are sourced from external fluids with respect to BIF. In the case of basin-related deposits, carbon is interpreted to be derived from deposits underlying carbonate sequences, whereas in the case of greenstone belt deposits carbonate is interpreted to be of magmatic origin. There is only limited mass balance analyses conducted, but those provide evidence for variable mobilization of Fe and depletion of SiO2. In the high-grade ore zone a volume reduction of up to 25% is observed. Mass balance calculations for proximal alteration zones in mafic wall rocks relative to least altered examples at Beebyn display enrichment in LOI, F, MgO, Ni, Fe2O3total, C, Zn, Cr and P2O5 and depletions of CaO, S, K2O, Rb, Ba, Sr and Na2O. The Y/Ho and Sm/Yb ratios of mineralised BIF at Windarling and Koolyanobbing reflect distinct carbonate generations derived from substantial fluid–rock reactions between hydrothermal fluids and igneous country rocks, and a chemical carbonate-inheritance preserved in supergene goethite. Hypogene and supergene fluids are paramount for the formation of high-grade BIF-hosted iron ore because of the enormous amount of: (1) warm (100–200 °C) silica-undersaturated alkaline fluids necessary to dissolve quartz in BIF, (2) oxidized fluids that cause the oxidation of magnetite to hematite, (3) weakly acid (with moderate CO2 content) to alkaline fluids that are necessary to form widespread metasomatic carbonate, (4) carbonate-undersaturated fluids that dissolve the diagenetic and metasomatic carbonates, and (5) oxidized fluids to form hematite species in the hypogene- and supergene-enriched zone and hydroxides in the supergene zone. Four discrete end-member models for Archean and Proterozoic hypogene and supergene-only BIF hosted iron ore are proposed: (1) granite–greenstone belt hosted, strike-slip fault zone controlled Carajas-type model, sourced by early magmatic (± metamorphic) fluids and ancient “warm” meteoric water; (2) sedimentary basin, normal fault zone controlled Hamersley-type model, sourced by early basinal (± evaporitic) brines and ancient “warm” meteoric water. A variation of the latter is the metamorphosed basin model, where BIF (ore) is significantly metamorphosed and deformed during distinct orogenic events (e.g., deposits in the Quadrilatero Ferrifero and Simandou Range). It is during the orogenic event that the upgrade of BIF to medium- and high-grade hypogene iron took place; (3) sedimentary basin hosted, early graben structure controlled Urucum-type model, where glaciomarine BIF and subsequent diagenesis to very low-grade metamorphism is responsible for variable gangue leaching and hematite mineralisation. All of these hypogene iron ore models do not preclude a stage of supergene modification, including iron hydroxide mineralisation, phosphorous, and additional gangue leaching during substantial weathering in ancient or Recent times; and (4) supergene enriched BIF Capanema-type model, which comprises goethitic iron ore deposits with no evidence for deep hypogene roots. A variation of this model is ancient supergene iron ores of the Sishen-type, where blocks of BIF slumped into underlying karstic carbonate units and subsequently experienced Fe upgrade during deep lateritic weathering.

Archean metabasalts enclose the Serra Norte iron deposits in the Carajas Mineral Province, Brazil. They experienced seafloor hydrothermal alteration, regional metamorphism under low greenschist facies conditions, and hematite-associated hydrothermal alteration, characterizing (i) least-altered, (ii) chlorite, and (iii) hematite-chlorite basalts in the N4 and N5 deposits. Their petrographic and geochemical changes were caused by both chlorite and hematite hydrothermal alteration. Nevertheless, part of the original, igneous textures and some of their primary minerals (plagioclase and clinopyroxene) are commonly preserved. Geochemical investigation of the least-altered basalt indicates a calc-alkaline affinity. Trace element behavior (e.g., Nb and Th) is consistent with an arc-related setting and continental crustal contamination. Where hydrothermal alteration affects basalts the most, distal, intermediate, and proximal alteration zones are distinguished based on the amount of chlorite ± hematite (less magnetite). Chlorite composition in (i), (ii) and (iii) marks these three mafic lithotypes, with variations in the Fet/(Fe + Mg) ratios that clearly reflect buffered through unbuffered conditions. Mineralogical and textural features, such as ruptured amygdales, indicate fluid interaction with hydrothermal minerals forming veins and breccias caused by hydraulic fracturing. Both chlorite and hematite-chlorite basalts experienced renewed isotopic equilibrium due to hydrothermal alteration, showing a decrease in δ18O (7.4 to 1.9 ‰) in comparison to the least-altered basalt reflecting an increase in fluid-rock ratio. On most discrimination diagrams and MORB-normalized spidergram, the investigated rocks are comparable to a volcanic arc environment associated with subduction zone. These and other evidence such as (a) a seafloor-associated volcanogenic sequence; (b) arc-type sedimentation of the Grao Para Group indicated by hyaloclastic tufts, breccias and volcaniclastic rocks; (c) jaspilites in the Serra Norte; and (d) a large subduction signature expressed by enrichment in Th, all suggest that the Grao Para basalts overflowed on the continental crust in a back-arc environment, with the primary magma having experienced significant upper crustal contamination. The arc-association evolution and the variability of lithological, sedimentological, igneous volcanic, structural, metamorphic, and other evolutionary aspects of the Itacaiunas Supergoup is suggestive of a greenstone-belt-type setting.An arc-associated evolution and the variability of lithological, sedimentological, igneous volcanic, structural, metamorphic, and other aspects of the Itacaiunas Supergoup supports a greenstone-belt-type setting.

The Faina greenstone belt is located in central Brazil and contains the Cascavel and Sertao gold deposits (500,000 oz Au total resource and past production). The bulk of the gold is hosted in quartzite and biotite schist at Cascavel, and carbonaceous schist and banded iron formation at Sertao. The metamorphic grade of the host rocks is greenschist facies. Cascavel is a vein-hosted gold deposit with orebodies mainly controlled by F2 fold hinges, with minor mineralization in D3 shear zones. Hydrothermal alteration in quartzite is characterized by distal quartz-K-feldspar ± muscovite, intermediate quartz-fuchsitic mica-biotite-pyrite, and proximal quartz-white mica-pyrite ± chalcopyrite. The biotite schist is characterized by distal quartz-chlorite-calcite and proximal white mica-biotite-tourmaline-siderite-pyrite. Native gold is located within intensely deformed quartz veins, and locally as free gold disseminated in proximal alteration zones. Sertao is a shear zone-hosted gold deposit with orebodies controlled by D3 shear zones including gold-bearing shear veins and surrounding hydrothermally altered wallrocks. Hydrothermal alteration is characterized by distal quartz-ankerite-chlorite-pyrite ± chalcopyrite and proximal quartz-white mica-siderite-pyrite-arsenopyrite ± chalcopyrite. Gold is in equilibrium with arsenopyrite and within arsenopyrite disseminated in the proximal alteration zone. Chlorite geothermometry suggests temperatures of 335–407 °C and 328–397 °C for the gold-bearing hydrothermal fluids in the Sertao and Cascavel deposits, respectively. Arsenopyrite geothermometry on arsenopyrite within gold-bearing veins in the Sertao deposit yields temperatures between 310 and 480°C. Structures controlling hydrothermal alteration and gold mineralization in both deposits are developed during a regional D2 event and subsequently deformed during a D3 event. Hydrothermal alteration minerals are consistently replacing metamorphic minerals. The alteration assemblages in both deposits are similar, albeit their modal mineralogy is controlled by the host rock composition. Therefore, both the Cascavel and Sertao deposits are interpreted as orogenic, mesozonal gold systems comparable with other orogenic gold systems in Brazil and worldwide.

The Lamego orogenic gold deposit (1.3 Mt measured resources at 7.26 g/t Au) is located at the south-western end of the 5 km-long Cuiaba–Lamego trend, Quadrilatero Ferrifero region, Brazil. Both Archean orogenic gold deposits are situated in the Rio das Velhas greenstone belt with the lithological succession at Lamego consisting of metamorphosed (greenschist facies) mafic volcanic rock, chert and banded iron formation (BIF), and carbonaceous and micaceous pelites. The Lamego fold, which controls the Lamego deposit, is the most visible structure related to the D1–D2 event, with a perimeter of about 4.8 km and a maximum width of 450 m. The fold is defined by the layering of the mafic unit with minor BIF and large exposures of carbonaceous and micaceous pelites. It is a rootless, reclined, isoclinal, cylindrical fold with an axial trace striking northwest–southeast dipping 20° to 30°. The hinge zone is thickened and the limbs are thinned, with the limbs dipping 20° to 30° to the SE. Orebodies consist of the Lamego BIF, where gold-mineralized zones are related to iron-rich bands, and associated silicification zones. Replacement-style mineralization is associated with sulfide bands; mainly pyrite, As-rich pyrite, and arsenopyrite. Two structural generations, G1 and G2, are recognized and encompass a set of structural elements. The G1 structural generation developed in a progressive deformation event and resulted in structures oriented from NE to SW and dipping to the SE. Structures pertaining to the G2 structural generation are oriented N–S dipping to the W. The four major high-grade gold orebodies are Carruagem, Queimada, Arco da Velha, and Cabeca de Pedra. Their gold grade shows a spheroidal pattern and a distribution that varies along the S1–2 foliation. These lenses represent the hinge zone of F2 reclined folds with the plunge of the orebodies controlled by the F2 fold axes. The lower-grade gold lenses are controlled by pinch and swell, and locally expressed quartz boudins developed during D1–D2. They have two orthogonal directions, one to the NW–SE and the other to the NE–SW, thereby defining chocolate-tablet style boudinage. Hydrothermal monazite grains in a mineralized mafic volcanic rock indicate that mineralization formed at 2730 ± 42 Ma (U–Pb SHRIMP). Younger monazite dated at 2387 ± 46 Ma, and xenotime dated at 518.5 ± 9 Ma suggest Siderian and Cambrian imprint in the Lamego deposit area. The Cambrian age reflects the late stages of the Brasiliano orogenic cycle, which is expressed in the Lamego deposit by the S3 crenulation cleavage (trend NS and dip steeply to the E). Importantly, the Cambrian structural modification is responsible for the present geometry of the Lamego orebodies, but is not associated with hydrothermal alteration; it has not introduced any new gold and also has not caused remobilization of the orebodies or ore minerals.

In-situ laser ablation ICP-MS analyses on iron oxides in itabirite and iron ore from the Quadrilatero Ferrifero (Brazil) reveal a wide range in trace element abundances (e.g., average concentrations in hematite: Al = 40–2200 ppm, Mg = 1–930 ppm, Mn = 5–540 ppm, Ti = 3–500 ppm, V = 2–390 ppm, Cr = 1–98 ppm, As = 0.5–60 ppm). The chemistry of early hematite stages is mostly inherited from host rock and precursor magnetite, e.g., Mn concentrations correlate with bulk Mn content in itabirite. With progressive iron ore formation and modification, external fluids play a more prominent role. This is reflected by REE-Y switching from seawater-like Y/Ho ratios (> 44) in early-, to more chondrite-like Y/Ho ratios ( The following ore formation stages and key processes, supported by mineral scale mass balance calculations, are constrained: (1) martitisation, cogenetic with gangue leaching, is driven by large volumes of oxidising, Si-undersaturated fluids resulting in an absolute depletion of Mg, Mn, Al, Ti, Ni and Zn, and enrichment of Pb, As, LREE and Y; (2) the formation of granoblastic hematite and locally microplaty hematite represents a largely isochemical recrystallisation of magnetite and/or martite accompanied by a depletion of Mg and Y and an elevated Ti mobility at the mineral scale; and (3) precipitation of schistose and vein-hosted specular hematite along shear and fracture zones is driven by an external Fe–Si-rich hydrothermal fluid likely under high fluid/rock ratios.

The Corrego do Sitio lineament is defined as a 16-km long, NE-SW-trending ductile shear zones, which controls fourteen gold deposits, including the Carvoaria Velha deposit. The dominant lithotypes of this deposit are metagrewackes with subordinate carbonaceous phyllites lenses of the Archean Rio das Velhas greenstone belt, which host gold mineralization. Metamafic dikes and sills are parallel and crosscut the host metasedimentary sequence. All these rocks exhibit alteration to quartz, carbonate and sericite, besides sulfides and sulfosalts. The main gold mineralization styles at the Carvoaria Velha deposit, and at Corrego do Sitio as a whole, are quartz-carbonate-sulfide ± sulfosalt veins of varied distribution. The veins are classified as: V1 veins – quartz-ankerite-pyrite-berthierite-gold – parallel to the main regional foliation S n ,; V2 veins – quartz-ankerite-pyrite – developed at extensional crenulation cleavage S n + 1 , and rarely gold mineralized; V3 veins – quartz-ankerite – filling S n + 3 fractures, usually free of sulfides and sulfosalts; V4 veins – quartz-calcite – of restricted occurrence in metamafic dikes and sills. The latter crosscut the metasedimentary sequence, are extensional and display no preferential orientation. The most common ore minerals in all vein types are arsenopyrite, pyrite, berthierite, and pyrrhotite. Microprobe analyses reveal the presence of metallic phases rich in Sb, Pb and Co, such as stibnite, ullmanite, tetrahedrite, galena, cobaltite, which commonly overgrow the sulfides. Fluid inclusion studies trapped in quartz from V1, V2 and V4 veins have identified a H 2 O-CO 2 ± CH 4 -NaCl mineralizing fluid that may contain KCl, MgCl 2 e FeCl 2 . The presence of CH 4 in fluid inclusions of the V1 and V2 veins reflect interaction of the fluid with the Corrego do Sitio and Santa Quiteria carbon-bearing metapelitic host units. Based on the microthermometric data, the hydrothermal fluid is interpreted to have evolved in at least two stages: i) an early fluid stage, trapped in smoky quartz, of moderate salinity (~ 8.5 eq. wt% NaCl), and minimal trapping temperature of 330 ± 17 °C; and ii) a late-stage fluid trapped in recrystallized quartz with lower salinity (~ 4.6 eq. wt% NaCl), and a minimal trapping temperature of 365 ± 33 °C. Isotopic values of δ 18 O fluid (+ 7.9 and + 13.0‰), Δ 33 S (+ 0.3 to + 3.5‰) and δ 34 S (− 2.9 to + 6.1‰) suggest that indeed the hydrothermal fluid responsible for the gold mineralization at the Carvoaria Velha deposit must have had a metamorphic origin, and interacted with metasedimentary sequences during its ascending path. The addition of CH 4 during fluid-rock interaction may have caused some decrease in ƒO 2 of the fluid which, as a consequence, destabilized gold-bearing sulfur complexes, liberating S − 2 for the formation of Fe sulfides and sulfosalts (arsenopyrite, pyrite, pyrrothite and berthierite, etc.), and outcome gold deposition.

Mesoarchean Algoma-type BIF in the Yilgarn Craton are locally enriched by secondary processes to iron concentrations approaching 72 wt% Fe. While their near-surface goethite–hematite ores are unequivocally the product of supergene fluid alteration, their deeper magnetite and crystalline hematite–martite ores have a contentious origin. This study shows that magnetite ores formed after diagenesis and peak greenschist facies metamorphism of BIF, coinciding with regional deformation events. The ores formed by a multi-stage process involving initial replacement of primary quartz bands in BIF by hypogene carbonate minerals, followed by the dissolution of the carbonate and concentration of magnetite. At the Beebyn deposit, the earliest fluids were hot (∼440 °C), CO2–vapor-rich, and were likely derived from felsic magmatic fluids. In contrast, the fluids responsible for carbonate alteration at the Matthew Ridge, Koolyanobbing, and Windarling deposits were cooler (366 to 125 °C) and aqueous-rich with salinities from 0.3 to 43.9 wt% NaCl equiv. Their Cl–Br and C–O isotopic ratios suggest they are the product of mixing between magmatic fluids and heated seawater. In all deposits, later cooling (∼178 to 98 °C) of these fluids led to dissolution of carbonate minerals and concentration of magnetite. Dated monazite and xenotime from magnetite ores demonstrates that fluid alteration of BIF began in the Mesoarchean and continued episodically through to the Paleoproterozoic. Later specular hematite-rich veins that cut and modify magnetite ores crystallized from aqueous-rich fluids that are more oxidized, with salinities from 0 to 28.7 wt% NaCl equiv. and crystallization temperatures of 243 to 85 °C. Their Cl–Br and C–O isotopic fluid ratios indicate that they formed from mixing between heated seawater and meteoric waters. The BIF-hosted iron deposits in the Yilgarn Craton show remarkable similarities with the well-studied iron deposits in the Carajas Mineral Province and Hamersley Basin. They all show the same transition through time from magnetite- to hematite-rich ores. Subsequent episodic, tectonically-driven, fluid alteration events led to further iron enrichment in these deposits. The transition in ore types most likely corresponds to a common guiding process, namely the uplift of crustal rocks containing BIF – resulting in the progressive shallowing of fluid alteration pathways and the greater influence of cooler, more oxidized, hydrothermal fluids. Unlike iron deposits hosted by Algoma-type BIF in the Yilgarn and Carajas, iron deposits hosted by Superior-type BIF in the Hamersley Basin display a greater influence of oxidized marine and meteoric waters, without involvement of magmatic fluids.

The Mt Richardson deposit contains high-grade (>57 wt% Fetotal) iron ore zones hosted by several banded iron-formations (BIFs) in the c. 3504 Ma Illaara greenstone belt. The BIFs and intervening mafic igneous rocks are exposed along a 700 m-wide, 140 m-high, NNW-trending ridge that demarks the eastern limb of a greenstone belt-scale tight, upright, moderately SSE-plunging syncline. Local expressions of this fold generation include parasitic metre-scale fold hinges, an axial planar fabric, and intersection lineation in BIFs. Margins of tightly folded BIFs are commonly sheared and altered by hypogene magnetite and quartz. Although primary bands are strongly overprinted by disseminated magnetite, the concomitant growth of quartz translates to a negligible increase in the iron content of hypogene-altered BIFs. Reactivation of magnetite–quartz-altered shear zones and interaction with more oxidized fluids led to the replacement of magnetite by martite, precipitation of crystalline hematite, and recrystallization of quartz. Mafic igneous rocks are altered to fine-grained chlorite via replacement of igneous amphibole, biotite, and plagioclase within 30 m of sheared and hypogene-altered BIFs. Hypogene-altered mafic igneous rocks are enriched in Fe2O3total and W, and depleted in MnO, K2O, Ca, Na2O, and Sr compared to least altered rocks. Subsequent reactivation of shear zones and the circulation of supergene fluids led to dissolution and replacement of primary and hypogene quartz by goethite, and further oxidation of magnetite to martite. Resultant goethite–hematite ore zones are silica-poor, iron-rich, porous, and friable. Alteration indices for supergene-modified, hypogene ore zones in BIFs include hypogene magnetite and hematite in BIFs, chlorite-replacement of primary minerals in mafic igneous rocks, intense supergene goethite–hematite–kaolinite alteration, thinning of primary quartz bands in BIFs, and the presence of broad high-strain zones. Prime exploration targets in the Illaara greenstone belt are areas where these alteration indices coincide with fold-thickened BIFs. Given the friable nature of high-grade iron ore zones, mineralized occurrences are often at least partly hidden and require remote detection methods, followed by drilling, for testing. Mt Richardson differs from most iron deposits in the Yilgarn Craton by the absence of early hypogene carbonate mineral replacement of primary quartz bands in BIFs. Instead, Mt Richardson demonstrates the greater role of supergene fluids that exploit existing structures and permeability via reactivation of hypogene magnetite- and hematite-rich shear zones. In this regard, the Mt Richardson deposit has greater affinity with iron deposits in the Pilbara Craton.

The Paleozoic Pataz–Parcoy gold mining area is located in a right-stepping jog on the regional Cordillera Blanca fault, in northern Peru. Most of the 8 million ounces of gold production from this area has come from quartz–carbonate–sulfide veins hosted by the Pataz batholith. Despite a subduction zone setting since at least the Cambrian, the area records several periods of extension and its present structure is that of a rift and graben terrain. The Pataz district (the northern part of the Pataz–Parcoy area) is dominated structurally by northwest to north northwest-striking (NW–NNW) faults and northeast to east northeast-striking (NE–ENE) lineaments, both of which have been active periodically since at least the Mississippian (Early Carboniferous). NW–NNW faults control the margins of a central horst that exposes basement schist and the Pataz batholith, and step across NE–ENE lineaments. The Lavasen graben, to the east of the central horst, contains the Lavasen Volcanics, and the Chagual graben, to the west, contains an allochthonous sedimentary sequence derived from the Western Andean Cordillera. New SHRIMP zircon geochronological data indicate emplacement of the Pataz batholith during the Middle Mississippian, at around 338–336 Ma, approximately 10 Ma earlier than previous estimates based on 40 Ar/ 39 Ar geochronology. The calc-alkaline, I-type batholith comprises diorite and granodiorite, the latter being the major component of the batholith, and was emplaced as a sill complex within the moderately NE-dipping sequence of the Eastern Andean Cordillera. Moderate- to high-temperature ductile deformation took place on the batholith contacts during or shortly after emplacement. Following emplacement of the batholith, differential uplift occurred along NW–NNW faults forming the Lavasen graben, into which the Lavasen Volcanics were deposited. SHRIMP U–Pb in zircon ages for the Lavasen Volcanics and the Esperanza subvolcanic complex, which was intruded into the western margin of the graben, are within error of one another at ca 334 Ma. The ductile batholith contacts were cut by renewed movement on NW–NNW faults such that the margins of the batholith are now controlled by these steep brittle-ductile faults. The NW–NNW faults were oriented normal to the principal axis of regional shortening (ENE–WSW) during formation of the batholith-hosted, gold-bearing quartz–carbonate–sulfide veins. The misoriented faults were unable to accommodate significant displacement, leading to high fluid pressures, vertical extension in the competent batholith and formation of gold-bearing veins. Brittle failure of the batholith was most extensive in the northern Pataz district where the fault-controlled western contact of the batholith is offset by a swarm of NE–ENE lineaments. The timing of vein formation is not established, despite published 40 Ar/ 39 Ar ages of 312 to 314 Ma for metasomatic white mica, which are interpreted as minimum ages of formation. Gold-bearing veins formed during or shortly after uplift of the Pataz batholith and formation of the Lavasen graben; they were therefore broadly coeval with deposition of the Lavasen Volcanics and emplacement of the Esperanza subvolcanic complex. These K-rich, weakly alkalic, ferroan (A-type) magmas may provide a viable source for the ore fluid that deposited gold in the Pataz batholith.

Detailed hydrothermal alteration investigations, including petrography, infrared reflectance spectroscopy (IRS) and XRD of the low sulfidation epithermal Co–O mine, located in Eastern Mindanao (Philippines) revealed that both distal and intermediate hydrothermal alteration zones contain dominantly illite and chlorite, whereas the proximal alteration zone comprises mainly illite, chalcopyrite and pyrite. The gold-bearing veins and the proximal hydrothermal alteration zone display a distinct absence of K-rich hydrothermal alteration minerals such as K-feldspar (adularia). Gold mineralization in the Co–O mine is controlled by an extensive quartz-breccia vein system, which is characterized by three distinct stages of vein (incl. breccias) formation. Gold is mainly observed in stages 2 and 3 veins. Stage 1 veins appear as fragments in stage 2 veins and display boiling textures such as quartz pseudomophs after bladed calcite. These veins further display colloform to crustiform banding and contain pyrite, chalcopyrite and minor gold located in the colloform bands and between bladed quartz pseudomorphs. Stage 2 veins comprise mostly banded to massive quartz and contains sulfides parallel to bands or disseminated. These veins are fine-grained with mosaic/jigsaw quartz and contain calcite blebs and/or fragments of stage 1 veins. Gold is in textural equilibrium with chalcopyrite, sphalerite, and locally pyrite. Stage 3 veins consist of quartz and carbonate (locally Mn-rich), and display irregular banded and comb textures. In auriferous veins of this stage gold is in textural equilibrium with chalcopyrite and pyrite (with local abundance of sphalerite). Other sulfide minerals observed with gold in stages 2 and 3 are galena, acanthite and locally jalpaite. The XRD and IRS provide inconsistent results regarding the abundance of K-rich clays (e.g., illite) associated with auriferous veins. Illite, with possibly interlayered swelling clays, such as Al-smectite, was identified in auriferous vein stages 2 and 3 using IRS, but could not be confirmed by XRD. Comparative analysis of the results of these techniques with respect to the ordering of micaceous minerals, suggest less ordered white mica proximal to the veins. Vein textures such as banded quartz, the absence of K-feldspar and the abundance of illite (interlayered Al-smectite) suggest relatively low temperatures of formation of the hydrothermal alteration system and point to a potential boiling horizon located deeper or marginal to the currently exploited levels of the Co–O mine. The absence of K-feldspar may also be related to relatively low temperatures of the hydrothermal fluid, the medium potassium-rich magma series of the host rocks, and/or a relatively low oxidation state of the hydrothermal fluid.

The Yilgarn Craton hosts three types of Archean granitoid-associated metallogenic systems: orogenic gold, intrusion-related, and porphyry systems. These systems may occur in the same terrane and share many broad characteristics, including spatial coincidence with granitoids, gangue silicate mineralogy, metal associations, and local structural controls. Features of Archean orogenic gold systems spatially associated with granitoids include (i) a broad range of granitoid host compositions; (ii) gold mineralization is mostly late with respect to Yilgarn-wide, granitoid emplacement, peak metamorphism, and deformation; (iii) orebodies are structurally controlled, and (iv) gold deposition occurs over a large range of temperatures, pressures, and crustal depths (2 to 15 km) from a CO 2 -rich, low to moderate salinity, reduced ore fluid that was derived from a metamorphic and/or distal magmatic fluid source. These systems do not display deposit-scale vertical or lateral zonation of metals away from granitoid contacts. Host granitoid age, composition, or oxidation state are not important in controlling gold mineralization processes, whereas structural setting and fluid flux are paramount. Archean intrusion-related Au–Mo–W and Mo ± Au systems in the Yilgarn Craton are commonly small ( 2 -bearing, moderate to high salinity fluid during ore deposition. Archean porphyry Cu–Mo–Au, Cu–Mo, and Cu–Au systems are spatially and temporally associated with volumetrically small, pervasively altered, felsic plutons and dikes surrounded by altered and mineralized supracrustal countryrock. Lateral zonation of metals away from porphyry centers commonly includes a decrease in Cu:Au and Ag:Au ratios with distance from the source pluton. Microthermometry studies on fluid inclusions in gold-bearing veins suggest that early ore fluids were high-temperature and high-salinity, and that mineralization occurred at 2 O 3 /FeO, are demonstrated in Phanerozoic terranes but are presently ill-defined in the Yilgarn Craton. Crustal depth estimation is important in defining areas likely to have preserve shallow-forming, Archean porphyry systems.

The Chalice gold deposit (∼20 t Au produced), in the southwestern portion of the Late-Archean Norseman–Wiluna greenstone belt of the Yilgarn Craton, Western Australia, is hosted both in a sequence of intercalated mafic and ultramafic amphibolites and a post-peak metamorphism monzogranite dike. The deposit is flanked on the western side by calc-alkaline plutonic rocks, and to the east by a predominantly monzogranitic pluton, with at least four generations of monzogranitic dikes that intrude the local mine stratigraphy. Locally, four deformation events (D 1 –D 4 ) have affected this sequence of amphibolites, and two stages of gold mineralisation are recognised, controlled by the progressive D 2 event. Main Stage gold mineralisation (95% of the resource), which is demonstrably later than all granitic bodies except second generation and subsequent monzogranite dikes, is controlled by localised D 2 asymmetric folds developed in the mafic amphibolite. It is expressed as veins and wall-rock replacement, characterised by a prograde assemblage of quartz–diopside–plagioclase–K-feldspar–titanite–pyrrhotite–pyrite–magnetite±garnet, hornblende, scheelite and biotite. Second Stage gold mineralisation (5% of the resource) is associated with the intrusion of a monzogranite dike and is expressed as disseminated gold in the dike, as well as foliation-discordant veins of quartz–gold, quartz–diopside–gold, actinolite–gold and molybdenite–tellurobismuthite–gold. Detailed field mapping and petrography suggest that Main Stage gold mineralisation at Chalice is equivalent to hypozonal orogenic lode–gold mineralisation in other Archean greenstone belts of the Yilgarn Craton, whereas monzogranite-associated Second Stage gold mineralisation is atypical of such deposits. The Main Stage mineralisation cannot be correlated to any particular granitoid in the local environment, although temporally, it is bracketed by granitic magmatism. A deep magmatic or metamorphic source cannot be resolved from the field data. It is also uncertain as to whether Second Stage mineralisation represents a true magmatic-gold event, or is a product of the assimilation and remobilisation of Main Stage gold, during the intrusion of the monzogranite dike. However, its distinctive metal association of Au–Bi–Mo–Te–W suggests the former. Irrespective of this uncertainty, Chalice remains as a deposit which is more intimately related both spatially and temporally to granitic magmatism than other Yilgarn deposits.

The so-called `mesothermal' gold deposits are associated with regionally metamorphosed terranes of all ages. Ores were formed during compressional to transpressional deformation processes at convergent plate margins in accretionary and collisional orogens. In both types of orogen, hydrated marine sedimentary and volcanic rocks have been added to continental margins during tens to some 100 million years of collision. Subduction-related thermal events, episodically raising geothermal gradients within the hydrated accretionary sequences, initiate and drive long-distance hydrothermal fluid migration. The resulting gold-bearing quartz veins are emplaced over a unique depth range for hydrothermal ore deposits, with gold deposition from 15–20 km to the near surface environment. On the basis of this broad depth range of formation, the term `mesothermal' is not applicable to this deposit type as a whole. Instead, the unique temporal and spatial association of this deposit type with orogeny means that the vein systems are best termed orogenic gold deposits . Most ores are post-orogenic with respect to tectonism of their immediate host rocks, but are simultaneously syn-orogenic with respect to ongoing deep-crustal, subduction-related thermal processes and the prefix orogenic satisfies both these conditions. On the basis of their depth of formation, the orogenic deposits are best subdivided into epizonal ( mesozonal (6–12 km) and hypozonal (>12 km) classes.