Fluid inclusion and isotope (O, H, C, Sr) constraints on the orogenic gold mineralization at the Enche Concha and Tunel prospects, Gurupi Belt, Brazil

The gold mineralization at the Enche Concha and Tunel prospects, northwestern portion of the Gurupi Belt, is hosted in dacites and phyllites that belong to the ca. 2160 Ma-old metavolcano-sedimentary Chega Tudo Formation. These host rocks are variably deformed, from a spaced cleavage to schistose and mylonitic varieties. The hosting structures have previously been defined as produced in the Paleoproterozoic. The hydrothermal alteration processes include carbonatization, sericitization, sulfidation, and quartz veining. These processes are both pervasive and fissure-filling in style. Pyrite is largely predominant over chalcopyrite and sphalerite and gold occurs as a refractory phase in pyrite. Fluid inclusion, stable (O, H, C) and Sr isotopes studies indicate that the hydrothermal alteration and gold mineralization occurred between 260 and 370°C (mostly ~328°C), and 0.85 to 2.8 kbar, from a low-salinity (~5% NaCl), aqueous-carbonic (XCO2 = 0.03-0.13), relatively reduced (log fO2 = -32.1), and near neutral (pH = 5.0-6.2) metamorphic fluid (d18O = +9.4 to +11.3‰, dD = -25 to -28‰). Contributions from deep-seated (lower crust and mantle) sources are suggested by a d13CCO2 value of -7.6‰, and 87Sr/86Sr ratios between 0.702699 and 0.705141, which are lower than those found in the country rocks. Gold was transported by the Au(HS)2- complex and precipitation occurred in response to phase separation, identified by fluid inclusions, and fluid-rock interactions, given by the lowering of the d13CCO2 values. The geological and genetic characteristics are equivalent to those of the orogenic gold system previously defined for the Gurupi Belt.


Introduction
The Gurupi Belt, in northern Brazil, contains shear zonehosted gold deposits that totalize over 158 t of gold (~5 Moz). These deposits, including Cachoeira, Chega Tudo, and Cipoeiro, have been defined as belonging to an orogenic gold system (Klein 2014), and formed at the late stages of an accretionary-collisional Rhyacian orogeny (Klein et al. 2020;Tavares et al. 2017). Enche Concha and Tunel are exploration targets (prospects, separated by ca. 2.4 km) located in the central portion of the Gurupi Belt ( Fig. 1) and they have received intermittent exploration work. Unpublished company reports inform the presence of several mineralized intersections, most notably 3.5 m @ 0.82 g/t at Enche Concha, and 7.39 m @ 0.83 g/t (including 1 m @ 2.50 g/t) in the Tunel prospect.
Although the orogenic model is well-defined for the Gurupi Belt, no information on host rocks, hydrothermal alteration, and genetic aspects for several prospects, including Enche Concha and Tunel, is available. In this work, we discuss these aspects for both prospects, supported by fluid inclusions, stable isotopes (C, O, H), and provide the first Sr isotope compositions of hydrothermal calcite associated with gold mineralization in the Gurupi Belt. The results allow us to discuss the physico-chemical characteristics and sources of the mineralizing fluid, and make comparisons with known deposits in the Gurupi Belt and elsewhere.

Regional geology
The Gurupi Belt is composed of magmatic, sedimentary and metamorphic rocks that range in age from 2695 Ma to 549 Ma (Fig.  1, Klein et al. 2020, and references therein). The oldest rock is the Archean Igarapé Grande metatonalite (2695 ± 4 Ma), which forms part of the basement of the belt. Most of the rock units were formed during the Rhyacian period and include arc-related sequences, such as the orthogneisses of the Itapeva Complex (2167-2148 Ma), voluminous juvenile calc-alkaline granitoids of the Tromaí Intrusive Suite (2174( -2148 and Canoa quartz diorite (2162 ± 11 Ma), amphibolites (2166( -2136, the juvenile metavolcano-FIGURE 1. Location map (inset) and simplified geological map of the Gurupi Belt (modified from Klein et al. 2020), with location of the Enche Concha and Tunel prospects and main gold deposits: 1 -Cachoeira, 2 -Chega Tudo, 3 -Serrinha, 4 -Montes Áureos, 5 -Cipoeiro.
sedimentary Chega Tudo Formation (2160 ± 3 to 2146 ± 1 Ma), and the potassic and weakly peraluminous granitoids of 2163 ± 4 Ma (Cantão), 2142 ± 9 Ma (Jonasa), which show crustal signature. Variably deformed peraluminous, muscovite-and biotite-bearing granites and leucogranites with ages between 2116 Ma and 2072 Ma and with crustal signature (Japiim Intrusive Suite) are related to the collisional phase of the Rhyacian orogeny; whereas highly evolved to shoshonitic granite and quartz-syenite of 2100 ± 21 Ma (Anelis Suite), and potassic granite of 2084 ± 5 Ma (Timbozal Granite) are post-collisional units (Klein et al. 2020). Siliciclastic sequences are included in orogenic sedimentary basins, such as the Santa Luzia do Pará Formation (younger than 2140 Ma) and the Igarapé de Areia Formation (younger than 2078 Ma) (Klein et al. 2020). All these Rhyacian units are part of an accretionary to collisional orogen, which continues to the north in the São Luís cratonic fragment (Klein et al. 2020, and references therein).
In terms of structural evolution, Tavares et al. (2017) defined three phases for the Gurupi Belt. (1) D 1 is ductile, oriented predominantly to N40°-50°W, progressively folded by D1b, and coeval to the peraluminous granites of ca. 2100-2070 Ma. The development of the Tentugal shear zone is associated with this phase. (2) D 2 is also ductile, affecting only Paleoproterozoic rocks, and strikes WNW-ESE to E-W. (3) D 3 affects both Paleoproterozoic and Neoproterozoic rocks, with direction averaging N-S. The Tentugal shear zone was reactivated in this phase.

Summary of gold metallogeny
According to Klein (2014), the orogenic gold deposits of the Gurupi Belt are hosted in Paleoproterozoic (2160-2147 Ma) metavolcano-sedimentary and calc-alkaline granitic rocks formed in magmatic arc and/or back-arc settings during a Rhyacian orogeny (2240-2080 Ma) (Fig. 1). The distribution of the gold deposits is controlled by the Tentugal shear zone and its subsidiary structures, which were developed during the D1 deformation phase (Tavares et al. 2017) and broadly coeval to the intrusion of collision-type peraluminous granites at ca. 2100-2070 Ma. The hydrothermal alteration comprises varied amounts of sericite, chlorite, carbonate minerals and sulfides. Gold occurs in quartz and quartz-carbonate veins (with or without sulfides) and in association with disseminated sulfides in strongly altered and variably deformed host rocks. Gold precipitation occurred mostly between 300 and 370°C and up to 3 kbars in response to fluid immiscibility and fluid-rock reactions and minor fluid mixing and oxidation. Geological characteristics, petrographic, fluid inclusion, and isotopic evidence indicate near-neutral, reduced aqueouscarbonic metamorphic fluids, with local contributions from host rocks at the deposit site (Klein 2014, and references therein).

Analytical procedures
All studies were performed in samples obtained from three drill cores. The fluid inclusion studies were carried out in the Laboratório de Microtermometria of the Federal University of Pará (UFPA), in Belém, Brazil, using a Linkam MDSG600 heating-freezing stage with a temperature range of −196° to 600°C. The calibration was done with synthetic fluid inclusions containing pure CO 2 (melting at -56.6 ºC) and H 2 O (melting at 0.0 ºC). The precision was estimated at ± 0.5º C and ± 5ºC for temperatures lower and higher than 50ºC, respectively. The analytical procedures followed criteria from Roedder (1984), Shepherd et al. (1985), and Goldstein and Reynolds (1994). Physical and chemical properties were estimated with the software FLINCOR (Brown 1989), following the equations of Brown and Lamb (1986) for pure H 2 O and CO 2 system, and Bowers and Helgeson (1983) for the H 2 O-CO 2 -salt system.
Regarding the stable isotopes, the carbon and oxygen isotope compositions of calcite were carried out in the Laboratório de Isótopos Estáveis of the University of Brasília, Brazil, using the continuous flow approach. After cleaning of the samples in a He flux at about 72°C, the CO 2 was obtained in a vacuum line by reaction with H 3 PO 4 at 25°C, and subsequently purified cryogenically. The isotope compositions were measured with triple collectors of a SIRA II mass spectrometer. Oxygen isotopes in quartz and hydrogen isotopes in fluid inclusions were analyzed in the Stable Isotope Laboratory of the Department of Geological Sciences, University of Cape Town, South Africa. The isotope ratios were measured offline using a Finnigan Delta XP mass spectrometer in dual inlet mode. Oxygen was analyzed by the laser fluorination method described by Harris and Vogeli (2010). Oxygen from quartz was liberated after reaction with BrF5 and conversion to CO 2 by reaction with graphite at 750°C. An internal standard (Monastery garnet -MONGT, δ 18 O= +5.55‰) was analyzed to calibrate the data to the V-SMOW (Vienna Standard Mean Ocean Water) scale. Hydrogen isotopes were also measured on fluid inclusion water extracted from quartz vein samples by thermal decrepitation of the fluids inclusions. The quartz vein fragments were degassed under vacuum at 200°C using a hot air gun to eliminate any moisture before heating it to about 800 °C. The obtained water was trapped cryogenically and transferred to pyrex tubes containing Zn. The hydrogen isotope composition of the H 2 O was determined using the closed tube Zn reduction method of Vennemann and O'Neil (1993), and the internal water standard CTMP2010 and RMW were used to calibrate the data to the V-SMOW scale. The stable isotope results are reported in ‰ in the delta notation (δ 18 O, δD, δ 13 C) relative to the V-SMOW (oxygen, hydrogen), and V-PDB (Vienna Pee Dee Belemnite, carbon) scales.
Strontium isotopic analyses were carried out at the Laboratório de Geologia Isotópica (Pará-Iso) of the UFPA. About 500 mg of powdered sample was first washed with ultrapure water and dissolved with 150 µl of HCl 2N distilled twice, until reaction was over. Extraction and purification of Sr were carried out by ion-exchange chromatography using a Eichrom® Sr 50-100 μm resin, conditioned in HNO 3 medium followed by ultrapure water as eluant. The Sr concentrates were loaded on degassed tungsten single filament. The Sr isotopic ratios were measured on a Finnigan MAT 262 thermal ionization multicollector mass spectrometer (TIMS) and they were corrected from mass discrimination using 84 Sr/ 88 Sr = 0.1194. During the period of Sr isotopic determinations, the accuracy, precision, and reproducibility of the analyses were monitored by repeated measurements of the NBS987 reference material.

Host rocks
Petrographic investigation of drill cores down to 265 m in depth allowed the recognition of different rock types hosting hydrothermal alteration and gold mineralization. The rocks comprise an intercalation of dacite and phyllite, along with brecciated zones (Fig. 2).  The phyllite is a dark grey to greenish rock (Figs. 2 and 3A) composed of quartz, white mica, chlorite, and carbonate. This rock shows a penetrative and continuous cleavage, which gradually pass to a spaced foliation defined by the alternation of thin layers of mica and quartz. Locally, the rocks resemble schists, where the minerals are recrystallized and coarsergrained (Fig. 4). Porphyroclastic texture is also observed, with ovoid fragments of carbonate that are contoured by finegrained micas and quartz (Fig. 4C). The variation in the planar structures may indicate progressive deformation (S1 followed by S1 b , according to Tavares et al. 2017).
The dacites are light grey rocks with variable grain size, from aphanitic, to fine-grained and locally porphyritic (Figs. 2 and 3B). The granular varieties are composed of plagioclase (75%) and quartz (25%). The plagioclase is tabular and euhedral to subhedral (Figs. 3B and 4H), and partially replaced by carbonate (Fig. 4J). The quartz grains are anhedral and show subgrains. In the porphyritic varieties, which are also porphyroclastic due to deformation, the phenocrysts/ porphyroclasts made up 25% of the rocks, and comprise plagioclase (20%) and quartz (5%) set in a quartzo-feldspathic matrix (Figs. 4K, 4L). The plagioclase is anhedral to subhedral, rather ovoid, and strongly replaced by sericite and carbonate ( Fig. 4), whereas quartz is anhedral and rounded.
The fault breccia forms centimeter-thick layers that cut across the phyllite and dacite layers. The rocks are fragmentary, non-cohesive, and without evident planar tectonic fabric (Fig. 3C). The fragments consist of vein quartz, vein carbonate and phyllite (Figs. 4M, 4N), which range in shape from angular to rounded, immersed in a quartzofeldspathic matrix (20% of the breccia volume), which locally contains carbonate cement.

Hydrothermal alteration and gold mineralization
The hydrothermal alteration comprises four processes: carbonatization, sericitization, sulfidation, and quartz veining. These processes are both pervasive and fissure-filling, are distributed in all host rocks and have slightly different timing regarding deformation (Fig. 5).
Carbonatization is the main and more widespread alteration process observed in all lithotypes and along the entire investigated depths. The carbonate alteration comprises the formation of anhedral calcite, with grain-size ranging from 0.5 to 1.0 mm, and two types/stages are observed: (1) disseminated, by replacement of the magmatic plagioclase (Figs. 6A, 6B), and (2) crystallization in mm-thick veinlets (Fig. 6C). The veining is more intensive in the more deformed portions of the host rocks.
The sericitic alteration occurs predominantly in the dacites and is present in different forms. The alteration is mainly placed along the foliation planes (Fig. 6D), being syn-tectonic to S1 or S1b. Sericite in pressure shadows of pyrite (Fig. 6E) indicates crystallization during the late stages (S1b?) of foliation development (Witt 1993). Locally, the sericite flakes cut across plagioclase crystals, which are partially or completely replaced by carbonate (Fig. 6F). Disseminated patches and replacement of plagioclase are also common (Fig. 6G).
The sulfide alteration affects all host rocks, but is more abundant in the phyllites and quartz veins, being probably associated with the late stages of the ductile deformation. Pyrite is largely the predominant sulfide phase, with subordinate amounts of sphalerite and chalcopyrite occurring in spatial association with pyrite (Fig. 6H). Thin layers of pyrite are placed along the foliation planes or cut across this foliation (Figs. 6I, 6J), and isolated euhedral crystals associated or not with carbonate are also common (Figs. 6K, 6L).
Veins and veinlets of quartz and quartz-carbonate, with or without sulfides (Figs. 6I, 6K, 6L), crosscut all lithotypes. These veins are mm-to a few cm-thick and are in general discordant to the foliation, but show deformation features, such as micro-folding and subgrain generation (Figs. 6L, 6M). The occurrence of thicker (up to 30 cm-thick), massive to laminated gold-bearing quartz veins, which have been mined by artisanal miners (indicating the presence of free-milling gold), has been informed by unpublished company report (Kinross).
Free gold particles have not been observed in our investigation. Gold was detected by scanning electron microscopy, ranging in grade mostly from 0.83 to 1.57%, but reaching higher grades (5-6%) in deformed pyrite crystals present in quartz veins.

Fluid Inclusions
The fluid inclusion study was carried out on quartz crystals from sulfide-bearing (mineralized) veins from the Tunel prospect, sampled from drill core JB08 at the depths of 73.2 and 74.5 m (Fig. 7A). Over 180 fluid inclusions have been analyzed microthermometrically.

Types and distribution
The quartz crystals that host fluid inclusions are anhedral, with irregular rims, and often forming subgrains (Fig. 7B).  The fluid inclusions occur in clusters, in isolation, or in trails (Fig. 8). Necking down features are common in larger inclusions (>5 µm) from all types. These necked inclusions were not included in microthermometric studies. Microscopic investigation at room (22-24°C) and sub-zero temperatures allowed the identification of three types of fluid inclusions. One type is comprised of one-phase inclusions that did not show phase transitions during the microthermometric work. These monophasic inclusions are probably metastable or empty cavities and will not be further addressed in this paper.
The Type-1 consist of aqueous-carbonic fluid inclusions having two phases. These phases comprise liquid water plus liquid CO 2 , or liquid water and gaseous CO 2 (Fig. B). The type-2 consists of aqueous fluid inclusions and is formed predominantly by two-phase (liquid + vapor) and subordinately one-phase (liquid) inclusions (Figs. 8C, 8D).
Type-1 fluid inclusions are less abundant than Type-2 and occur predominantly as isolated individuals, and subordinately in clusters (primary inclusions) and trails (secondary inclusions). The morphology is rounded to ellipsoidal, and they vary in size between 3 and 5 µm. The carbonic phase occupies 15 to 30% of the inclusion cavities.
The more abundant Type-2 occurs as clusters and intragranular trails, and rarely in isolation. These inclusions are primary and pseudo-secondary in origin, and are rounded or polygonal, with sizes ranging from 2 to 5 µm. The vapor phase occupies 80% of the inclusion cavities, but some inclusions have 70 to 60% of the vapor phase.
When in clusters, Type-1 and 2 inclusions occur as an association, and no temporal (crosscutting) relationships could be established, suggesting that the two types were trapped at the same time and belong to a single Fluid Inclusion Assemblage (FIA -Goldstein and Reynolds 1994).

Microthermometric results
The following abbreviations are used in text and figures. TmCO 2 : temperature of melting of the carbonic phase, ThCO 2 : temperature of homogenization of the carbonic phase, Tmclat: clathrate melting temperature, Tmice: final ice melting temperature, Teu: eutectic melting temperature, Tht: temperature of final/total homogenization. Considering the state of homogenization, L stands for liquid, and V for vapor.
In the two-phase Type-2 fluid inclusions, only a few Teu have been observed, ranging from -24.5 to -20.0°C, and Tmice varies between -0.1 and -4.1°C (Fig. 9E). Tht is in the range of 167 and 342°C (Fig. 9D).

Density and composition
Density and compositional data were calculated based on the microthermometric results. The global and CO 2 densities of the aqueous-carbonic fluid inclusions vary from 0.7 to 1.0 g/cm 3 , and 0.2 to 1.0 g/cm 3 , respectively (Bowers and Helgeson 1983). The salinity of these fluid inclusions are in the range of 4.4 to 5.3 wt.% NaCl equivalent (based on Collins 1979), and the small variation in the TmCO 2 values in relation to the pure CO 2 values, indicates only minor amounts (up to 5%) of additional components, such as N 2 and/or CH 4 in the carbonic phase (e.g., Van den Kerkhof and Thiéry 2001). The estimated XCO 2 values range from 0.03 to 0.13. For the aqueous inclusions, densities of 0.7 to 0.9 g/cm 3 , and salinities of 0.2 to 6.3 wt.% NaCl equivalent are estimated according to Brown and Lamb (1986).

Origin of the fluid inclusions
Petrographic and microthermometric results have shown that the two types of fluid inclusions occur in association in the same microscopic domains of the hosting quartz, and that the two types homogenized broadly in the same range of temperatures (Figs. 9D, 10, and 11). These features suggest that the two types are coeval and belong to a single system. A few aqueous fluid inclusions with Tht of 300-350°C have lower salinities (1-2%) when compared with the majority of the inclusions (3-6%) (Fig. 9C). We don't have a clear answer to this fact, which might be attributed to local conditions of trapping of more diluted fluid.
The contemporaneous trapping together with the variation in Tht in a single domain (Fig. 10) is indicative of phase separation (e.g., Wilkinson 2001), which may be triggered by different processes, such as pressure fluctuation (Robert et al. 1995), as observed in the range of estimated pressures, and CO 2 removal due to carbonate crystallization (Rimstidt 1997). These are common features identified in the other orogenic gold deposits of the Gurupi Belt (e.g., Klein 2014).

Stable isotopes
The stable isotope compositions were measured in hydrothermal minerals and fluids: oxygen in vein quartz; oxygen, and carbon in vein and disseminated calcite from the carbonate alteration; hydrogen in inclusion fluids hosted in quartz (Table 1).
The δ 18 O values of quartz from Enche Concha vary between +16.6 and +16.9‰ in mineralized veins and is +17.7‰ in a barren quartz vein, whereas mineralized samples from Tunel range from +15.3 to +16.3‰. The δD values of inclusion fluids from two mineralized samples are -28 and -25‰ in the Enche Concha and Tunel prospects, respectively.
The δ 13 C values of calcite vary between -10.6 and -11.4‰ in the disseminated type, and is -9.8‰ in the quartz-calcitesulfide vein from Enche Concha. In the same prospect, the δ 18 O values are in the range of +14.3 to +16.1‰ for the disseminated variety, and is +14.5‰ for the vein calcite. In the Tunel prospect, the δ 13 C and δ 18 O values range from -14.1 to -15.1‰, and +14.8 to +16.2‰, respectively. As a whole, the oxygen compositions are similar, but the δ 13 C values are quite lower in the Tunel samples.

Sr isotopes
The isotope composition of Sr was measured in hydrothermal calcite from the disseminated carbonate alteration ( Table 2). Irrespective of the host rock type (phyllite or dacite), the obtained 87Sr/86Sr ratios are weakly radiogenic and range from 0.702699 to 0.705141 for the Enche Concha samples, and from 0.703055 to 0.705080 at Tunel. Therefore, no difference exists in the Sr isotopic composition in the two prospects.

Physico-chemical conditions of mineralization (T-P-fO 2 -pH)
The estimated temperature and pressure conditions of gold mineralization is based on the microthermometric results from fluid inclusions and isotope thermometry. The oxygen isotope compositions of the coexisting quartz-calcite pair from sample JB04_238.6 yielded a temperature of 328°C using the equation of Sharp and Kirchner (1994). This value is in line with the interval of total homogenization temperatures measured in the aqueous-carbonic fluid inclusions (~260-370°C).
For pressure estimations, we used the above temperature data, along with salinity, CO 2 density, and isochores calculated according to Bowers and Helgeson (1983) for the system CO 2 -H 2 O-NaCl. These data yielded a pressure range of 0.85 to 2.8 kbar (Fig. 12A), implying 3 to 9 km in depth, assuming lithostatic conditions, which is compatible with the structural setting of mineralization.
For the calculated T-P-XCO 2 range, a log fO 2 value of -32.1 is estimated for the oxygen fugacity of the fluid using the method of Huizenga (2005). This value lies above the CO 2 -CH 4 buffer (Fig. 12B), which is in line with the fluid inclusion TmCO 2 data, and indicates relatively reduced conditions during gold mineralization. Additionally, the presence of calcite and sericite in the hydrothermal alteration, together with the lack of K-feldspar, indicates a solution with near neutral to slightly alkaline pH (ca. 5.0 to 6.2 ;Romberger 1990;Hayashi and Ohmoto 1991).

Fluid composition and potential sources
The composition of the hydrothermal fluid was estimated from the stable isotope results, using relevant fractionation factors and temperature (328°C), and from direct measuring of the isotopic composition of fluid inclusions (Table 1). The  Final homogenization temperature versus salinity plot for Type-1 and Type-2 fluid inclusions. Each symbol represents the average values for a single Fluid Inclusion Assemblage. calculated and measured fluid compositions, along with the Sr isotope and fluid inclusion results, were then compared to those of potential sources.
The δ 18 O values of the water in equilibrium with quartz is +10.7 to +11.0‰ at Enche Concha, and +9.4 to +10.4‰ at Tunel (quartz-H 2 O fractionation factor of Matsuhisa et al. 1979). Calculated for calcite the fluid δ 18 O values are in the range of +9.5 to 11.3‰ for the disseminated calcite, and is + 9.7‰ for the vein calcite in the Enche Concha prospect. At Tunel, the values are +10.0 and +11.4‰ (calcite-H 2 O fractionation factor of Friedman and O'Neil 1977). As a whole, the fluid δ 18 O values have similar ranges in quartz and calcite from the two prospects (+9.4 to +11.4‰). These values, in combination with the δD values from inclusion fluids suggest a metamorphic source for the fluid (Fig. 13A).
The δ 13 C values of the CO 2 in equilibrium with calcite (calcite-CO 2 fractionation factor of Ohmoto and Rye 1979) range from -8.3 to -9.1‰ for the disseminated stage, and is -7.6‰ for the vein calcite at Enche Concha. The values for Tunel are -11.8 and -12.8%. Theses CO 2 δ 13 C values are not diagnostic of a specific carbon source. Instead, the variation in the isotopic composition might indicate variable contribution of organic carbon from the carbonaceous schists and phyllites of the hosting Chega Tudo Formation, which is a common feature in the Gurupi Belt (e.g., Klein et al., 2005). In addition, the fluid δ 13 C value of the quartz-carbonate-sulfide is the highest among the studies samples (Fig. 13B), and this composition may be closer to that of the mineralizing fluid, indicating a deep seated source (Ohmoto and Rye 1979).
In closed systems, minerals with insignificant contents of Rb, such as calcite, which does not incorporate this element in its structure, preserve the initial 87 Sr/ 86 Sr ratio (Mueller et al. 1991). In consequence, the 87 Sr/ 86 Sr ratios of hydrothermal calcite may be used as an approximation to the Sr isotope composition of hydrothermal fluids, and in the investigation of changes that the fluids may undergo by interaction with regional rocks or with mineralized host rocks (Kontak and Kerrich 1995;Scanlan et al. 2018). Furthermore, even for carbonates that have recrystallized or re-equilibrated at low temperatures, the Sr and C isotopic systems are sufficiently resistant to preserve the original compositions (Kerrich 1987). The relatively low 87Sr/86Sr ratios  (~0.7027-0.7051) measured in the hydrothermal calcite from Enche Concha and Tunel indicate a little evolved source with low Rb/Sr ratio. Comparing with the composition of orogenic gold deposits from different cratons and ages, we can observe that our results are lower than those of the orogenic deposits (Fig. 14). Our results are also lower than those presented by the regional Rhyacian rocks of the Gurupi Belt and adjoining São Luís cratonic fragment (Fig. 14). This indicates that the Sr present in the mineralizing fluid is externally-derived and that the regional rocks have not contributed with Sr to the fluid. Furthermore, the data indicate a probable deep source for Sr, such as the lower crust or mantle (Fig. 14).

Genetic constraints
Some interpretations about the genesis of the gold mineralization can be drawn for the studied prospects, based on the fluid inclusion and isotopic data. The physico-chemical characteristics of the fluid, the presence of pyrite as the main sulfide mineral, and the lack of significant amounts of base metals suggest the H 2 S (alternatively HS-) as the main sulfur species in the fluid. This, together with the oxygen fugacity estimation indicate Au(HS) 2 -as the gold transporting complex (Fig. 12B). The destabilization of this complex and deposition of gold from solution might be ascribed to phase separation, FIGURE 13. (A) Oxygen and hydrogen isotopic composition of the hydrothermal fluids at Enche Concha and Tunel compared with available data for orogenic gold deposits of the Gurupi Belt (Klein 2014). The fields for magmatic and metamorphic waters are from Sheppard (1986) and the meteoric water line is from Craig (1961). SMOW = Standard Mean Ocean Water. (B) Carbon and oxygen isotopic composition of hydrothermal calcite.

FIGURE 14.
Strontium isotope composition of hydrothermal calcite from Enche Concha and Tunel compared to the isotopic compositions of orogenic gold deposits worldwide, and with potential sources. References: (1) this study, (2) Kontak and Kerrich (1995), (3)  as documented by the fluid inclusion study, along with fluidrock interactions, which is inferred from the carbon isotope data (reaction with organic carbon) and from the refractory character of gold, which is present in the structure of pyrite. Changes in the pH of the fluid by sericitization and removal of CO 2 of the fluid to produce calcite might have also contributed to the breakdown of the sulfur complex.
Our results and interpretations are similar to those described for other gold deposits in the Gurupi Belt (Klein 2014) and elsewhere, being consistent with the orogenic gold system (e.g., Goldfarb and Groves 2015, and references therein).

Conclusions
The Enche Concha and Tunel exploration targets (prospects) have gold mineralization hosted in dacites and phyllites that belong to the ca. 2160 Ma-old metavolcanosedimentary Chega Tudo Formation. Petrographic, fluid inclusion, and isotopic (O, H, C, Sr) data allowed the following observations and conclusions to be drawn.
(1) The hydrothermal alteration comprises the formation of pervasive and fissure-filling carbonate, sericite, and sulfides (pyrite and subordinate chalcopyrite and sphalerite), in addition to quartz (±carbonate, ±sulfides). Refractory gold occurs in pyrite.
(2) The alteration and mineralization are associated with previously defined Paleoproterozoic deformation phase.
(3) The hydrothermal alteration and gold mineralization occurred between 260 and 370°C, from 0.85 to 2.8 kbar, from an aqueous-carbonic, relatively reduced and near neutral metamorphic fluid, with contributions from deep-seated (lower crust and mantle) sources.
(4) Gold was transported by a sulfur-bearing complex and precipitation occurred in response to phase separation and fluid-rock interactions.
(5) The geological and genetic characteristics are equivalent to those of the orogenic gold system previously defined for the Gurupi Belt.