The gold mineralization of the Shwetagun deposit is located about 3 km south of the Modi Taung deposit (Figures 3, 4). The Kogwe mudstone and Poklokkale Pebbly wackestone are mainly exposed at the Shwetagun area (Mitchell et al., 2004; Myo Kyaw Hlaing, 2013; Mitchell, 2018) and the gold mineralized veins are hosted by the Kogwe Mudstone Unit which consists of pebbly mudstone, siltstone and sandstone (Mitchell et al., 2004). Gold-quartz veins are oriented in NE-SW and N-S directions and can be grouped into one of the three vein systems within exploration Adit I, II and III (Figures 4, 5A–C). The Adit I and II are located along the NE-SW strike, whereas Adit III trends north-south. Each vein system consists of either a single vein or multiple parallel veins separated by host rocks. Quartz veins in Adits I and II show stylolitic lamination and sheeted structure (Figures 5A–C) which indicate the movement on the shear planes during the formation of gold veins (e.g., Mitchell et al., 2004).

In Adit III, the strong post-vein faulting more or less along the plane of the shear resulted in pinched and swelled structure, and brecciation. F1 and F2 faults are mainly related to the mineralized veins where bedding is usually low angle with southeast-northwest trending fold axes; most veins intersect bedding at a high angle, cut crenulation cleavage where visible, cut the bedding plane, and accompanied by pervasive silicification of the mudstone (Figures 5A–C). The general trend of bedding is NE-SW direction paralleling the mineralized veins (Figure 4). En-echelon sigmoidal tension gashes filled with barren to weakly auriferous fibrous quartz are widespread in the mudstone sometimes kilometers away from the mineralized veins (Figures 6A–D). The tension gash veins in the shear zone imply ductile deformation which is evident for later faulting (Figure 6C). The Shwetagun area is structurally complex; there have been at least two or three deformation phases in both small-scale and major structures (e.g., Michell et al., 2004). No igneous rocks have been found at the Shwetagun area, and the nearest igneous rocks to the Modi Taung deposit are the microgranite porphyry at Theingi vein and the late dacite and andesite porphyry dykes which cut across the ore zone at Shwesin vein (Traynor et al., 2015, 2017) and Momi Taung, approximately 3 km northeast of Shwetagun (Figure 3).

The Kyaikhto gold district in the southern part of MB consists of several gold deposits and occurrences, including Kunzeik, Zibyaung and Meyon (Figure 7). The Kyaikhto gold occurrences are formed as sheeted veins, massive veins, network veinlets, stockworks, and disseminations. The structural setting at the Kyaikhto gold occurrences is dominated by NNE-SSW, and NW-SE striking faults (Figure 7). Mineralization in the Kyaikhto gold district is hosted by metasedimentary rocks of sandstone, slate, phyllite, and schist, and associated with the biotite granite and biotite granodiorite (Figure 7). The granitic rocks in this area can be correlated to the Mokpalin quartz diorite and the Kyaikhtiyo granite that have yielded the LA-ICP-MS zircon U-Pb ages of 90.8 ± 0.8 Ma and 63.3 ± 0.6 Ma, respectively (Figure 7; Mitchell et al., 2012; Myo Kyaw Hlaing et al., 2019). The metamorphic grade of the metasedimentary rocks is greenschist facies, characterized by a mineral assemblage of epidote, biotite, muscovite, and chlorite. Locally, fine-grained pyrite crystals are disseminated in the slate, phyllite, and schist (Myo Kyaw Hlaing et al., 2019).

The quartz vein systems almost invariably appear to be related to regional NNW -SSE trending structures which are commonly recorded along the MB (Figure 2). Brecciation and fracturing of the quartz vein contributed favorable sites for supergene gold enrichment processes (Zaw Naing Oo and Khin Zaw, 2009; Zaw Naing Oo et al, 2010; Zaw Naing Oo and Khin Zaw, 2015; 2017). The thickness of the tension gash veins and stockwork veins varies from a few 20 cm to 2 m wide. All of these gold deposits feature extensive, steeply dipping, vein systems, usually with multiple stages of vein development. In some deposits, the veins are slightly narrow (1–2 m) and separated, and there may be several parallel veins set within broad structural zones (e.g., Meyon). In other areas, such as at Kunzeik, the multiple vein systems form stockwork structures. At Zibyaung, the mineralized quartz veins are found as white to milky quartz. Mineralization in the Kyaikhto gold district is dominantly related to the deformation of the host rocks and occurred within structurally-controlled dilatant fractures.

Microscopically, the ore samples show various mineral assemblages and indications of fracture fillings as well as replacement textures. Based on microscopic observation, there are two types of gold-bearing sulfide ore textures: the disseminated and fracture-filling sulfide types (Figures 9A,B,D,E). The electrum occurs as submicron-sized inclusions in the fracture of sulfides at Shwetagun (Figures 9A,B). Most of the electrum grains are identified as free grains, but some are intergrown with pyrite, chalcopyrite, galena and sphalerite. Locally, it is concentrated as clumps of irregular electrum grains dispersed in alteration minerals. Some electrum grains are lighter and gradually change to whitish color with very high reflectivity, due to presence of high silver content. The electrum grains at Shwetagun yielded an Au content of 72.9–74.3 atomic %. Sphalerite is one of the major ore minerals and appears in euhedral to subhedral aggregates up to 5 mm in size (Figure 9C). The FeS mol% of sphalerite is between 4.02 and 8.41. Chalcopyrite is a primary important hypogene mineral in the ore textures (Figure 9E).

Pyrite is the most abundant sulfide mineral at Shwetagun. It occurs as fine-grained, massive aggregate and euhedral to subhedral crystals along the outer margin of the quartz veinlets. The earliest sulfide mineral is invariably pyrite, with deeply fractured sub-euhedral crystals that are locally replaced by other base-metal sulfides (Figures 9A–E). A few grains of pyrite are replaced by chalcopyrite along their grain boundaries and microfractures (Figure 9E).

At Kunzeik, the quartz-carbonate-sulfide vein consists of chloritized host rock fragments surrounded by irregular to massive aggregates of pyrite, chalcopyrite, and sphalerite together with quartz and chlorite (Figure 9F; Myo Kyaw Hlaing et al., 2019). Electrum, sphalerite, and galena are mostly found in pyrite at Zibyaung (Figures 9G,H; Myo Kyaw Hlaing et al., 2019). Sphalerite forms as a fissure filling mineral within pyrite. Galena occurs as fine-to medium-sized grains filling cracks in pyrite and sphalerite. Electrum grains with an atomic % of 74.5–78.1 Au can be detected as minute grains in pyrite with galena, and sphalerite (Figure 9F–H). The FeS mol% of sphalerite ranges from 4.98 to 7.26. At Meyon, chalcopyrite is associated with pyrite and covellite. Covellite occurs as a secondary replacement mineral of chalcopyrite (Myo Kyaw Hlaing et al., 2019). Native gold also develops as free grains in the oxidized zone and fracture fillings within pyrite. Pyrite occurs as two generations, idiomorphic grains intergrown with chlorite aggregates or in contact with massive chalcopyrite, pyrrhotite, arsenopyrite, marcasite, magnetite and hematite (Zaw Naing Oo and Khin Zaw, 2009; Zaw Naing Oo et al, 2010; Zaw Naing Oo and Khin Zaw, 2015; 2017; Myo Kyaw Hlaing et al., 2019).

Fluid inclusions studies have been performed in quartz, quartz-sulfide and quartz-carbonate-sulfide veins from the Shwetagun, Kunzeik and Zibyaung deposits (Figures 10A–D). At Shwetagun, fluid inclusions were found in quartz in the mineralized veins quartz. There are two types of fluid inclusions: Type A (vapor-rich) and Type B (liquid-rich). Type A (vapor-rich) and Type B (liquid-rich) aqueous fluid inclusions were observed in all quartz samples and occur as isolated groups of primary inclusions as clusters and pseudo-secondary trails. Secondary fluid inclusions (trail-bound) are also present but are extremely small (<2 μm) and have not been studied. Most fluid inclusions were aqueous (H₂O) fluid inclusions. At Kunzeik and Zibyaung, two types of fluid inclusions were found in quartz and calcite of quartz-carbonate-sulfide and quartz-sulfide veins at room temperature (Figures 11A–H). Based on the criteria described by Roedder (1984), the fluid inclusions were classified as primary, pseudo-secondary, and secondary, and given below: Type A (aqueous-carbonic) inclusions, Type B (aqueous) inclusions.

Type A (aqueous-carbonic inclusions): This type belongs to the H₂O–CO₂–NaCl system and occur in quartz and calcite of the mineralized vein quartz. They are 5–10 µm in size and have bi-phase (liquid and vapor) at room temperature. These inclusions also show presence of three phase (liquid H₂O, liquid CO₂, vapor CO₂) and clathrate during freezing runs. The CO₂: H₂O ratio in these inclusions range from 80: 20 to 50: 20. Occasionally vapor bubbles are moving but no liquid CO₂ is present (Figures 11C,E–G).

Type B (aqueous inclusions): have been identified in quartz and calcite of the mineralized vein quartz. These inclusions are bi-phase with a vapor bubble and liquid at room temperature (Figures 11C–F,H).

Several geological and geochemical characteristics of the MB gold deposits are comparable with those of the typical orogenic gold deposits worldwide (Table 3; e.g., Groves et al., 1998; Hagemann and Cassidy, 2000; Kerrich et al., 2000; Poulsen et al., 2000; Goldfarb et al., 2001; Robert and Poulsen, 2001; Groves et al., 2003; Goldfarb et al., 2005; Robert et al., 2005; Large et al., 2011; Groves et al., 2019). The gold deposits of the MB have a number of characteristics that are consistent with an orogenic gold system, including a close association between mineralization and deformation, and mineral assemblages of veins and alteration that are similar to Archean and Paleozoic gold systems (e.g., Groves et al., 1998; Bierlein and Crowe 2000; Hagemann and Cassidy 2000; Large et al., 2007). The orogenic gold systems are formed from the crustal fluids produced during the prograde metamorphism at the greenschist-amphibolite facies transformation and developed at intermediate depths (2–10 km) (Groves et al., 1998, 2003, 2020; Bierlein and Crowe, 2000; Goldfarb et al., 2001, 2005; Phillips and Evans, 2004; Dubé and Gosselin, 2007; Phillips and Powell, 2009, 2010; Tomkins, 2013; Goldfarb and Groves, 2015). Metamorphic fluids are generally regarded to be significant in the formation of orogenic gold deposits (e.g., Groves et al., 1998, 2003). There is still some debate over the appropriate source of orogenic gold ore fluids and the metals they transport (Tomkins, 2013; Goldfarb and Groves, 2015; Wyman et al., 2016; Groves et al., 2020). In comparison, Wyman et al. (2016) explored the presence of magmatic fluids in various orogenic gold deposits worldwide, as well as providing a comprehensive review of mineralizing fluids in orogenic gold systems. The formation of the widespread alteration of albite-silica-chlorite-sericite-carbonate-pyrite associated with the gold-bearing quartz veins in the MB was a major enhancing mineralizing mechanism for many other orogenic gold deposits worldwide (e.g., Phillips, 1993; McCuaig and Kerrich 1998; Goldfarb et al., 2005). The worldwide examples of these structural setting of the orogenic gold deposits are recorded at Golden Mile in Kalgoorlie, Western Australia, Western Lachlan Orogen in Victoria, SE Australia, Buller Terrane in western South Island, New Zealand, Meguma Terrane in Nova Scotia, Canada and Main Divide and Pingfengshan gold mine in Taiwan (Table 3; Shackleton et al., 2003; Bierlein et al., 2004; Groves et al., 1998, 2003; Goldfarb et al., 2001, 2005; Craw et al., 2010). Similar structural settings of the gold deposits are found in Myanmar at Phayaung Taung, Modi Taung and Shwegyin (Table 3; Mitchell et al., 2004; Win Phyo et al., 2016; May Thwe Aye et al., 2017; Ye Myint Swe et al., 2017).

The microthermometric characteristics of fluid inclusions, alteration types, host rocks and tectonic setting related to various stages of vein formation of the mineralized quartz vein systems of Shwetagun (Modi Taung-Nankwe) and the Kyaikhto gold districts are remarkably similar. A majority of orogenic gold deposit studies have reported ore-mineralizing fluids to be of low salinity and aqueous-carbonic in composition (e.g., McCuaig and Kerrich 1998; Ridley and Diamond, 2000; Groves et al., 2003; Goldfarb et al., 2005). The abundant aqueous fluid inclusions are similar to gold-related fluids throughout the belt, and typical of orogenic gold deposits elsewhere (e.g., Ridley and Diamond, 2000; Groves et al., 2003).

At Shwetagun, Type A (vapor-rich) inclusions are much less abundant than the Type B (liquid-rich) fluid inclusions, but they were observed as primary and pseudo-secondary inclusions in all samples. The secondary fluid inclusions have similar microthermometric properties to the primary inclusions, indicating that they represent events of vein reactivation in an evolving hydrothermal environment (Roedder, 1984). The ore fluid of Shwetagun ore-forming fluids evolved from a low to medium temperature (250–335°C), low salinity (0.2–4.6 wt% NaCl equivalent).

According to microthermometric results, a drop in salinities is found simultaneously with a decrease in temperature, suggesting that the system is diluted with cooling during the fluid evolution at Shwetagun. The major mechanism of ore and gangue mineral precipitation in veins is temperature drop (Figure 14). The absence of CO₂ in the gold-bearing veins at the Shwetagun deposit is in contrast to many ancient orogenic gold deposits, where separate phases of CO₂ are commonly found in fluid inclusions (Groves et al., 1998; McCuaig and Kerrich 1998; Bierlein and Crowe 2000; Goldfarb et al., 2004, 2005; Craw et al., 2010). However, the abundant CO₂ in the fluid is not a prerequisite for orogenic gold mineralization. (e.g., Main Divide and Pingfengshan gold mine in Taiwan; Craw et al., 2010 and references therein). The bulk densities of ore fluids from the Shwetagun deposit range from 0.62 to 0.83 g/cm³ (Table 2). Although no depth estimation of the Shwetagun deposit is possible by fluid inclusion evidence, Mitchell et al. (2004) estimated a depth of 4–7 km based on stratigraphic and structural evidence.

In comparison, fluid inclusions in these deposits at Kunzeik and Zibyaung are dominated by earlier Type A (aqueous-carbonic) and Type B (aqueous) fluids. Histograms of homogenization temperatures and salinities, as well as a diagram of homogenization temperatures versus salinities, show a wide range of homogenization temperatures between 242 and 376°C, and low to medium salinities between 0.9 and 11.8 wt% NaCl equivalent, respectively (Table 2; Figures 12–14).

According to microthermometric results, the range of total homogenization temperatures (242–376°C) indicate that there may have been more than one episode of CO₂ phase separation during vein formation. The wide range of salinity also indicates a source of fluid, possibly mixed with low-salinity exotic fluid (Figure 14). Based on the relative timing of fluid inclusions, Type B (aqueous) fluids formed first, followed by Type A (aqueous-carbonic) fluids (Figure 14). The Kunzeik and Zibyaung fluids show two totally identical processes of total homogenization of Type A (aqueous-carbonic) and Type B (aqueous) inclusions within a similar temperature range, indicating that they were trapped together and may represent the same H₂O-CO₂-NaCl system (Table 2; Figures 12–14).

The coexistence of Type A (aqueous-carbonic) and Type B (aqueous) fluid inclusions allows to estimate the depth of ore formation using the method of intersecting of the isochores based on the CO₂–H₂O–NaCl system. Using the CO₂ density formulae of the state of the isochores of Span and Wagner (1996) and the H₂O density isochores of Steele-MacInnis et al. (2012), and the pressure conditions of Sterner and Pitzer (1994), the bulk composition and density of fluid inclusions of these gold deposits were determined as below. The isochores of the Type A (aqueous-carbonic) and Type B (aqueous) fluid inclusions were plotted using the Roedder and Bodnar (1980) and Bowers and Helgeson (1983) state of equations, respectively (Figure 15).

The trapping pressures of fluid inclusions have been determined using CO₂-rich inclusions in quartz veins of the Kunzeik and Zibyaung deposits. The bulk densities vary from 0.53 to 0.91 g/cm³ and the densities of CO₂ range between 0.53 and 0.77 g/cm³ (Table 2; Figure 15). The Types A and B fluid inclusions from the Kunzeik deposit indicate that the deposit was formed at 376–246°C and 54–164 MPa (Figure 15), and for Zibyaung, it is estimated to be 358–242°C and 53–156 MPa (Figure 15). These data suggest a temperature and pressure variation during the ore-forming process. Depending on the pressure, the depth of mineralization was estimated. Given that studied areas are in a compressive orogenic environment, the formula can be used to estimate the metallogenic depth by lithostatic pressure: H = P/(ρ × g) (ρ represents average density of rocks, such as 2.70 g/cm³). These pressures correspond to the calculated depths of ore formation between 2.0 and 6.1 km under lithostatic conditions. These temperatures, depth and recorded alteration mineralogy in the Kunzeik and Zibyaung deposits are consistent with mesozonal orogenic gold deposits elsewhere (e.g., Groves et al., 1998, 2003; McCuaig and Kerrich 1998; Goldfarb et al., 2001, 2005; Elmer et al., 2006).

The structural setting of the MB is ideal for generation of orogenic gold deposits as described by Groves et al. (2003), which includes a first-order structure that focuses and channelizes metal-bearing fluids and neighboring second-order and possibly third-order structures that contain fluids, and the major processes comprising the formation of quartz veins and metal deposition around them (e.g., Goldfarb et al., 2001; Groves and Bierlein, 2007). A large number of orogenic or lode gold deposits are structurally-controlled by brittle-ductile shear zones and faults and are distinguished by fault-valve nature (Sibson et al., 1988; Cox et al., 1991; Cox 1995; Robert et al., 1995; Dubé and Gosselin, 2007). In the case of the MB, these structures can be defined by the transcrustal, first-order, the second-order shear zone, and reactivated Papun fault system located in the middle to south of the belt and the third-order structures semi-parallel to the fault zone and hosting the gold mineralization (Figures 2, 16).

The Shwetagun gold-bearing quartz veins and, in particular, the Zibyaung and Khunzeik veins in MB show a structural relationship between the first-order structures, mostly N-S to NNE-NNW oriented, and second order one trending NW–SE. The fault-valve model for orogenic gold deposits represents the cyclic fluctuation in fluid pressures, from lithostatic to hydrostatic, due to episodic shear stress (Sibson et al., 1988; Ridley 1993; Robert et al., 1995; Cox et al., 1995, 2001; Sibson and Scott 1998; Kisters et al., 2000; Oliver and Bons 2001; Groves et al., 2003). The available sulfur isotopic data along MB indicate that the δ³⁴S values of pyrite range from ˗2.80–4.43‰ in Meyon, (Zaw Naing Oo and Khin Zaw, 2009) and a range of +1.33–4.75‰ in Modi Taung (Traynor et al., 2015). These data suggest that the sulfur responsible for gold mineralization is likely to have been derived from a dominant magmatic source with a possible other minor fluid such as metamorphic fluid (e.g., Wyman et al., 2016; Groves et al., 2020).

The age of the gold district of Modi Taung-Nankwe was stratigraphically considered between the Upper Permian and the Middle Jurassic (Mitchell et al., 2004). They considered that it was an orogenic style gold, linking its genesis to the crustal fluids generated during the metamorphism of the MMB (Mitchell et al., 2004; Gardiner et al., 2016). This metamorphism was defined as Jurassic in age, contemplating this age for gold mineralization (Mitchell et al., 2004). However, Searle et al. (2007) suggested a much younger age of Paleogene for peak metamorphism in MMB. Here, at Modi Taung, the Theingi vein is cut by a 20 m wide dyke, and the Shwesin and Momi Taung veins are cut by small dykes and sills (Mitchell et al., 2004; Traynor et al., 2015, 2017). LA ICP-MS U-Pb zircon dating of granitoid unit intruding the host sequence of the Modi Taung gold deposit yielded 95 ± 30 Ma and the age of the andesitic dyke cross-cutting the mineralization by LA ICP-MS U-Pb dating of apatite was determined to be 49 ± 1 Ma (Traynor et al., 2015, 2017). Further work on U-Pb dating of zircon from the cross-cutting dyke is required to ascertain the robust age for the dyke intrusion, the age of the orogenic gold mineralization at the Shwetagun deposit in the Modi Taung-Nankwe gold district appeared that it was formed relatively late. The age of mineralization of the Meyon is proposed to be Lower Cretaceous to Paleogene and may have been consistent with deformation and metamorphism as being structurally-controlled and linked to movement along the Papun Fault Zone (Figure 16; Zaw Naing Oo and Khin Zaw, 2009, 2017).

During the Lower Cretaceous to Oligocene, Myanmar was influenced by at least two main orogenic events (Cretaceous and Eocene), which may have contributed to widespread regional metamorphism which is possibly linked to gold mineralization in the MB (e.g., Mitchell et al., 2004; Gardiner et al.,2014; Khin Zaw et al., 2014a; Zaw Naing Oo and Khin Zaw, 2015; Khin Zaw, 2017; Gardiner et al., 2018). Gold veins were formed during the ascent of metamorphic fluids following the uplifting and prograde metamorphism of the entire MMMB. The generation of this high-grade metamorphic process was associated with the subduction and collision of the India-Eurasia Plate (Khin Zaw et al., 2014b). Thus, we suggest that the age of mineralization for the orogenic gold deposits in the MB may have been Lower Cretaceous to Oligocene, and the emplacement of the mineralizing fluids was associated with retrograde metamorphism of the MMMB and dehydration at a depth of 2–12 km (e.g., Mitchell et al., 2004).

All the evidence presented for the orogenic gold deposits of the MB also suggest that the mineralizing fluids were derived from multi-sources and the initial ore-forming fluids were thought to be either due to fluids produced by metamorphic devolatization and/or mixed with magmatic fluid (e.g., Wyman et al., 2016; Groves et al., 2020). These multi-sourced fluids may have been transported from the deep crust to the shallower level via first-order faults during the evolution of the ore fluid, especially in the late stages of ore precipitation (e.g., Kerrich and Fyfe, 1981; Higgins and Kerrich, 1982; Cameron and Hattori, 1987; Phillips and Powell, 1993; Mikucki, 1998; Diamond, 2001; Wilkinson, 2001). The derivation of the ore fluids from the supracrustal rocks below the continental crust is widely accepted for the source (s) of orogenic gold deposits (Phillips and Powell, 2010; Goldfarb and Santosh, 2014), either from the subducted slab, with overlying oceanic sediments, or from the lithosphere below (Figure 16; Groves and Santosh, 2015). Dehydration and metamorphism within the oceanic crust and sediments have been well-observed mechanisms for producing abundant aqueous-carbonic and low-salinity fluids, as documented in orogenic gold deposits from fluid inclusions that induce massive energy and mass transport, culminating in the fractionation of mobile and even immobile elements (Pearce and Peate, 1995; Ridley and Diamond, 2000; Spandler et al., 2003).

In this context, the MB is a tectonic component of the Sibumasu terrane during the Jurassic to Cretaceous when the main tectonic transition from N-S compression to NNE-SSW shearing was progressively occurred. The large-scale circulation of fluids, granitic magmatism and metallogenesis took place during the Lower Cretaceous to Oligocene in the MB, as a result of which the initial ore fluids of the orogenic gold deposits were likely formed by dehydration, decarbonization and desulfidation of the subducting Neo-Tethys slab (Figure 16; e.g., Ridley and Diamond 2000; Phillips and Powell, 2010; Goldfarb and Santosh, 2014; Yardley and Bodnar, 2014; Khin Zaw et al., 2014b; Yardley and Cleverley 2015; Groves and Santosh, 2016; Groves et al., 2020).