*4.2. Veins*

4.2.1. Vein Types

Veins occur throughout the Erdenet deposit as a stockwork within the quartz and tonalite/granodiorite porphyries. The veins are millimeters to centimeters wide and contain sulfide minerals such as molybdenite, pyrite, chalcopyrite, and bornite, with minor calcite and rutile. At least three types of vein were recognized on the basis of crosscutting relationships in the samples Er-22 (567 m) and Er-24 (646 m). The different vein generations recorded progressive stages of fluid movement and hydrothermal activity. From early to late, they were as follows (Figure 4d):


wide. A veinlet of sericitized material was observed parallel to the vein wall. Most directions of elongation of the quartz grains were oriented perpendicular to the vein wall, but some quartz was oriented parallel to the vein wall (Figure 4b).


**Figure 4.** Photographs of the three vein types of the Erdenet deposit. (**a**) Quartz–molybdenite vein cut by later veins, (**b**) Quartz–pyrite vein cut by the later qtz vein. (**c**) Quartz vein within the quartz porphyry. (**d**) Sketch of samples showing analytical spots and crosscutting relationships among vein generations on tonalite/granodiorite porphyry.

## 4.2.2. Vein Textures

SEM–CL images reveal multiple generations of complex qtz veins that show textures that are not visible under an optical microscope and which record diverse processes (e.g., the dissolution of quartz grain cores, recrystallization, and fracturing). All three stages of veins are characterized by primary oscillatory zoning formed during initial quartz precipitation and the physicochemical changes of hydrothermal fluid cause precipitation of a secondary texture—for instance, recrystallization, fracturing, and the dissolution of quartz. These features crosscut or overgrow on the primary textures and can be distinguished from the primary quartz in the SEM–CL images [22].

The qtz–mol veins are the earliest generation of veins. The constituent grains have oscillatory zoning and are CL-gray in the SEM–CL images. The crystals grew from the vein wall towards the center of the vein, and calcite is common in the vein centers (Figure 5a,b). These features indicate that the quartz growth was syntaxial and that fractures provided voids for quartz precipitation from a silica-saturated fluid within the fluid conduit [25]. The zoned quartz grains are cut by interconnected CL-dark fractures knows as cobweb texture [22]; this texture records dissolution after the formation of the qtz–mol veins.

**Figure 5.** Images of veins. (**a**) Crystallographic orientation in cross-polarized light of the qtz–mol vein. (**b**) SEM–cathodoluminescence (CL) image of the qtz–mol vein shown in (**a**). (**<sup>c</sup>**,**d**) SEM–CL images of the qtz–py veins showing Ti analyses and temperatures. (**<sup>e</sup>**,**f**) SEM–CL images of the qtz veins showing Ti analyses and temperatures.

The qtz–py veins are associated with a dense fracture network and developed during the second stage of vein formation. Small euhedral crystals with vaguely defined oscillatory zoning occur on the walls of the veins. CL-dark quartz records a fracturing event after the formation of the initial qtz–py veins. Sulfides and calcite within the fractures and on the grain boundaries are associated with the CL-dark bands (Figure 5c,d). These veins are inferred to be syntaxial, based on the oscillatory zoning, crystallographic orientation, and presence of calcite.

Quartz veins, which represent the last stages of fluid activity on this system, are characterized by less CL-dark compared to the prior quartz veins. These veins contain a granular CL-bright quartz that contains a cobweb-like pattern of fractures filled with CL-dark quartz (Figure 5e,f). Primary textures, such as oscillatory zoning, are absent within this type of vein.

#### **5. Estimation of Formation Temperatures**

#### *5.1. Application of Ti-in-Quartz and Ti-in-Biotite Geothermometers to the Host Rock*

The crystallization temperature of the host rock was estimated using the Ti-in-quartz [26,27] and Ti-in-biotite [28] geothermometers (Figure 6). The biotite-bearing tonalite/granodiorite is part of the PA, which is associated with mineralization [15].

**Figure 6.** (**<sup>a</sup>**,**b**) Temperatures estimated using the Ti-in-biotite geothermometer of [26] (apfu = atom per formula unit). The host rock records temperatures of 700–750 ◦C. (**<sup>c</sup>**,**d**) Temperatures estimated using the Ti-in-quartz geothermometer. The quartz phenocrysts record temperatures of 400–750 ◦C.

Application of the Ti-in-quartz geothermometer to two quartz phenocrysts from the granodiorite porphyry yielded temperatures of 520–760 ◦C (Table 1). The Ti content was measured for a profile across the quartz grains (rim to rim) to determine if there was a temperature gradient (e.g., high and low temperatures recorded by the core and rim, respectively). Quartz grain (QtzHr1) (Figure 6c) does not show a systematic pattern, and the measured temperatures range from 550 to 750 ◦C. The core and rim of grain QtzHr2 (Figure 6d) record higher and lower temperatures, respectively, and a wide range of temperatures. The temperature recorded by two biotite grains was estimated using the Ti-in-biotite thermometer, and this provided more consistent results than those of the Ti-in-quartz thermometer (Figure 6b). Measurements were taken at points located on a profile from the core to the rim of the biotite grains (Figure 6a). The calculated temperatures ranged from 700 to 750 ◦C, based upon the biotite compositions provided in Table 2. The maximum temperature of quartz phenocryst calculated using the Ti-in-quartz geothermometer is similar to the temperatures calculated using the Ti-in-biotite

thermometer which represent initial quartz formation or nucleation. Therefore, we infer that the host rock was emplaced at a temperature of 700–750 ◦C.


**Table 1.** Compositions of quartz phenocrysts in the host granodiorite porphyry (number in parentheses next to each analysis represents 1σ and given in the term of the least unit cited; n.d. = not detected).

**Table 2.** Representative biotite compositions (wt%) used for Ti-in-biotite geothermometry [28]. Numbers in parentheses next to each analysis represent 1σ and given in terms of least unit cited; n.d. = not detected.


#### *5.2. Application of the Ti-in-Quartz Thermometer to Veins*

Quartz is ubiquitous within the veins, so the Ti-in-quartz geothermometer from [26] was applied to constrain the vein formation temperatures and the pressure-dependent calibration of [27] was applied as a comparison. The pressure of ~1.6 kbar, at which the Diorite porphyry was (fifth stage) emplaced, was chosen to estimate the temperature [16]. The Ti activity (α) was assumed to be unity (α = 1), which provides a minimum estimate of the temperature of vein formation (from hereafter referred as quartz formation temperature); if α were assumed to be 0.5, then the calculated temperatures would be ~65 ◦C higher. The detection limit for Titanium (Ti) by EPMA is 7 ppm, a concentration that corresponds to 500 ◦C, and the uncertainties are exponentially larger for Ti concentrations of <7 ppm [29]. Ti measurement points were selected within areas of different CL intensities to investigate the different physicochemical and fluid dynamic properties of fluids that percolated through fractures in the rocks.

#### 5.2.1. Quartz ± Molybdenite Veins

The qtz–mol veins are associated with the early and middle stages of mineralization within the Erdenet deposit [15,18]; the Ti content of quartz and estimated temperatures for these veins are shown in Figure 7a. In general, quartz within the qtz–mol veins has Ti contents of 6–82 ppm, which correspond to temperatures of 500–750 ◦C (Figure 8). The calculated temperature for barren CL-gray quartz with oscillatory zoning is consistent at ~500–600 ◦C. The CL-bright areas yield temperatures as high as 700 ◦C, whereas the CL-dark quartz associated with calcite has no Ti contents, so the temperature is inferred to be <500 ◦C. Molybdenite has a platy shape and is associated with CL-dark fractures. The CL-gray which is associated with molybdenite truncates the oscillatory zoning and yields temperatures of ~600 ◦C. Molybdenite is also associated with an interconnected fracture which cuts the CL-bright quartz.

**Figure 7.** Representative qtz–mol vein. (**a**) SEM–CLimage. (**b**) BSEimage showing Ti analyses. The estimated temperature is shown in parentheses. (**c**) Ti and Al concentrations of quartz. (**d**,**<sup>e</sup>**) Trace element maps showing a positive relationship between CL-brightness, Ti, and Al.

**Figure 8.** (**a**) Ti concentrations of quartz phenocrysts and vein quartz from the Erdenet deposit. (**b**) Minimum temperature of quartz vein formation calculated using the model of [28]. D.L. = detection limit.

Element maps show that the Ti and Al concentrations are proportional to the CL-brightness (Figure 7d,e), so it is assumed that these elements record the fluid conditions and diffusional states of Ti during quartz precipitation [30]. Therefore, the Ti concentration and temperature of CL-bright quartz are higher than those of CL-gray. Al content is positively correlated with Ti, and the Al concentrations are low (<250 ppm) relative to the average Al contents of other porphyry deposits (>2000 ppm), which show the negative correlations of Al and Ti contents with CL-brightness [23].

#### 5.2.2. Quartz ± Pyrite Veins

The quartz–pyrite veins were formed during the early and middle stages of mineralization at the Erdenet deposit [15]. The middle stages of mineralization are also commonly associated with the formation of metasomatic sericite and low homogenization temperatures of fluid inclusions (<250 ◦C). Multiple measurements of the Ti content of primary quartz with oscillatory zoning and quartz in the CL-dark fractures associated with pyrite yielded Ti concentrations of 3–74 ppm, which correspond to temperatures of 450–700 ◦C (Figure 8). The relationship between the CL-brightness and Ti concentration is similar to that of the qtz–mol veins. The CL-gray quartz, which shows oscillatory zoning between the vein wall and vein center, yields temperatures of ~500–600 ◦C. The Ti concentration of CL-dark quartz have lies below the detection limit (Figure 5c,d), which suggests it occurs as low-temperature quartz. Pyrite occurs on margins of grains that record temperatures of ~600 ◦C and on fractures that connect to the calcite at the center of the vein.

The low formation temperatures inferred for CL-dark quartz, which is associated with sericite and calcite, indicate that the phyllic and propylitic alteration zones, which formed after the sericite alteration, also formed at low temperatures [31–33].

## 5.2.3. Quartz Veins

Only two points within the qtz veins yielded detectable Ti concentrations (Table 3, Figure 5e), which indicate temperatures of ~500 ◦C. However, most of the measured Ti concentrations were less than the detection limit. Thus, we sugges<sup>t</sup> that th qtz veins formed at lower temperatures than those of the qtz–mol veins and qtz-py veins (Figure 8).

The moderate brightness of the CL signal indicates that the qtz veins formed under stable conditions and that the magmatic fluids mixed with lower-temperature fluids derived from meteoric water to produce a relatively low-temperature hydrothermal solution. This inference is supported by the fluid inclusion data (see below).

**Table 3.** Representative trace element contents of quartz veins and minimum temperatures calculated using Ti-in-quartz geothermometry (number in parentheses next to each analysis represents 1σ and given in the term of the least unit cited; n.d. = not detected).


#### *5.3. Fluid Inclusion Microthermometry and Compositions*
