**6. Discussion**

#### *6.1. Vein-Related Mineralization Processes*

The vein formation temperatures and SEM–CL observations provide insights into the conditions that formed the quartz and ore/sulfide minerals of the Erdenet deposit. The qtz–mol veins were precipitated at high pressures and temperatures (500–700 ◦C) [37], and the relatively low Al content in the qtz-mol veins indicates that the veins formed at high temperatures [23]. Platy molybdenite occurs close to the CL-dark fractures. The sulfide mineralization was formed from fluids derived from intermediate–silicic magmas [38]; the CL-dark quartz associated with molybdenite formed at ~600 ◦C and crosscuts the oscillatory zoning. The molybdenite mineralization forms part of the first stage of mineralization at the Erdenet deposit. In contrast, the qtz–py veins show identical growing patterns, with the qtz-mol vein where the pyrite or chalcopyrite occur on the margins of the vein and grain boundary. The mineralization corresponds to CL-dark quartz that crosscuts primary quartz with oscillatory zoning and is oriented parallel or perpendicular to the vein wall. The formation temperature recorded by the pyrite-bearing quartz is ~600 ◦C. The qtz veins record the latest stage of quartz precipitation within this system and are relatively unmineralized, but some pyrite occurs within these veins in association with the CL-dark quartz. We infer that the barren quartz with oscillatory zoning formed at the beginning of the vein formation and that this quartz was precipitated on the vein wall. The molybdenite and pyrite mineralization formed after the initial quartz in association with the CL-dark fractures that cut the oscillatory zoning.

#### *6.2. Estimated Fluid Pressures*

The fluid pressures were estimated by combining isochores calculated from the homogenization and ice-melting of fluid inclusion [35,36] with temperatures from the Ti-in-qtz thermometry. The maximum value of Th was used to construct the isochores to represent the minimum pressure condition. The temperature of ice melting could not be determined, but halite is not present within the inclusions, so the salinity was assumed to be 0–25 wt%. This assumption is consistent with previous reports of salinities of 5–25 wt% for fluid inclusions from surface samples [17]. The calculated pressure for the qtz–mol veins is 1.6–3.0 kbar, assuming a temperature of ~600 ◦C (Figure 10). A pressure of 1.1–2.5 kbar was estimated for the qtz–py veins, based on a Ti-in-quartz temperature of ~600 ◦C (Figure 10). A pressure of 2.1–2.8 kbar was calculated for the qtz veins, based on an assumed temperature of ~500 ◦C. However, the crosscutting relationships and the relatively lower Ti content as well as homogenization temperature (Th) of the fluid inclusion compared to other vein types indicate that qtz veins record the later fluid activity at the Erdenet deposit and were formed at considerably lower pressures. Therefore, we sugges<sup>t</sup> that the qtz veins must have formed at temperatures of <500 ◦C, based on their low Ti concentrations. In this case, the estimated pressure is 1–2 kbar assuming temperatures of 400–450 ◦C, although it could be lower (green b, Figure 10).

**Figure 10.** Estimated pressures of vein formation at the Erdenet deposit. Temperatures are derived from the Ti-in-quartz geothermometer, and the isochores correspond to salinities of 0–25 wt% [35].

#### *6.3. Fluid Evolution at the Erdenet Porphyry Deposit*

Porphyry copper deposits have distinctive characteristic veining and alteration patterns [31–33]. Systematic observations of the crosscutting relationships among veins the and textures within veins provide insights into fluid behavior, and quartz solubility models [39] have revealed fluid processes and evolution at individual porphyry deposits. At the Erdenet deposit, the magma chamber started to solidify and generate a crystal mush as the temperature decreased to ~700–800 ◦C after magma emplacement. Fluids released from crystallization processes (dehydration) moved upwards so that the cupola was saturated with fluid (Figure 11a). The initial pulses of magma injection caused fracturing in the host rock, and silica-saturated fluids percolated through these fractures to form the pre-mineralization quartz that grew from the vein walls (Figure 11b,e). This fluid might represent the magmatic fluid, based on the temperatures of 650–700 ◦C calculated for a high Ti activity (solid arrow in Figure 12).

**Figure 11.** Schematic model of the evolution of magmatic–hydrothermal fluids at the Erdenet deposit and illustration of the crystallization of various quartz veins. Initial magmatic fluid accumulated at the cupola (**a**). Magmatic pulse (**b**) induced the precipitation of the initial euhedral zoning quartz, which has a temperature close to the host rock (**e**). Precipitation of the initial quartz sealed the fracture and increased the fluid pressure. Episodic transience occurred and the fluid was reinjected into the existing fracture, followed by the precipitation of mineralization-bearing quartz (**f**). In the final stages, the magmatic front descends (**<sup>c</sup>**,**d**) and the quartz veins precipitate and cut previous generations of veins with cooler temperatures (hydrothermal system) (**g**).

The molybdenite and pyrite mineralization were formed from later fluids by one of two possible mechanisms as follows: (1) The mineralization of molybdenite and pyrite formed by later fluid processes possibly occurred in two scenarios. The first is that both molybdenite and pyrite were precipitated together. This mechanism is suggested by the occurrences of molybdenite and pyrite relating to fractures and grain boundaries (represent by CL-dark). Additionally, the fractures cut the primary oscillatory zoning of qtz, which we assume as the initial fluid activity in Erdenet. Moreover, the temperature inferred for the mineralization related to the CL-dark quartz is close to 600 ◦C (Figure 11f). Mineral precipitation might have been induced by fluid decompression after the initial pulse of magma. Meanwhile, the fluid temperature decreased and fluids were released episodically into the existing fracture network, where they precipitated quartz and sulfides (dashed arrow Figure 12). (2) Molybdenite was precipitated first from the cooling fluid, followed by pyrite. However, regardless of the precise mechanism, we infer that these minerals were precipitated within a lithostatic pressure regime from supercritical fluids.

**Figure 12.** Evolution of supercritical fluid based on inferred pressures and temperatures (solid and dashed arrow are the inferred P-T path) and the model of quartz solubility in H2O [40]. Two possible scenarios are presented here. The solid arrow represents the magmatic fluid upon gradual cooling that precipitated various quartz vein stages. The dashed arrow depicts a second scenario, where the decompression of the initial fluid takes place and induces the precipitation of mineralization-bearing quartz. Both mechanisms will tend to decrease at lower temperatures and precipitate the last stage of the qtz vein.

The final stages of fluid activity in this system involved the precipitation of qtz veins. The continued cooling and retreat of the magmatic front away from the position of initial emplacement to deeper parts of the system [3,40] are associated with a shift towards a hydrostatic pressure regime. The prograde quartz solubility caused the precipitation of qtz veins and related CL-dark quartz ± calcite (Figure 11g). The CL images show that the fluid generated intense fractures within the veins and intact rock at all stages. Fracturing might record infiltration by over-pressured fluid, whereby the build-up of pressure within the cooling pluton generated additional fluids that were transported into the cupola (Figure 11c). This process caused the injection of fluids into the surrounding rocks, which was controlled by the existing fractures [41]. This mechanism was associated with transient fluid pressure changes during the shift from a lithostatic to a hydrostatic fluid pressure regime (Figure 11d) [5].
