Numerical Simulation of the Mineralization Process of the Axi Low-Sulfidation Epithermal Gold Deposit, Western Tianshan, China: Implications for Mineral Exploration

Abstract

The Axi gold deposit, a low-sulfidation epithermal deposit in the Western Tianshan, China, hosts over 50 t of gold resources and is widely regarded as the result of coupled processes of rock deformation, heat transfer, pore fluid flow, and chemical reactions. However, research on the ore-forming processes of this gold deposit from a coupled perspective remains limited, resulting in its ore-forming mechanisms being incompletely understood. In this paper, we use the concept of mineralization rate based on computational modeling to indicate the 3D spatial distribution of mineralization. The simulation results reveal the following: (1) temperature gradients play a key role in influencing mineral precipitation, whereas the effect of pore fluid pressure gradients is relatively negligible; (2) gold precipitation, characterized by a negative mineralization rate, predominantly took place along fault zones that exhibit vertical transitions from steep to gentle slopes or lateral bends, which are further distinguished by the accumulation of fluids and the presence of significant temperature gradients. Notably, this particular distribution pattern of gold precipitation closely mirrors the spatial arrangement of known gold orebodies. These findings suggest that the coupling of multiple physical and chemical processes at specific fault sites plays a critical role in ore formation, providing new insights into the mechanisms governing the development of the Axi gold deposit. Furthermore, based on these observations, it can be inferred that the deeper regions of the Axi gold deposit hold considerable mineralization potential.

1. Introduction

The Tulusu Basin in Western Tianshan, NW China (Figure 1), hosts significant epithermal gold deposits, with the Axi gold deposit being the largest (Figure 2a), containing more than 50 t gold resources. The mineralization at the Axi deposit is linked to volcanic structural activities that enhance permeability and create ore-forming fluid conduits, which effectively concentrate ore-forming fluids in favorable locations, such as fault bends, dip variations, and structural intersections, thus facilitating critical fluid–rock interactions for gold deposition [1,2,3,4]. The ore-forming fluids are largely typified by magmatic-hydrothermal attributes, intricately connected to the post-volcanic stage [5,6,7,8], and present a composite assemblage predominantly composed of meteoric water [8,9], with magmatic fluids contributing in a subsidiary capacity [10,11,12]. However, the ore-forming mechanisms of the Axi gold deposit have not been thoroughly examined in terms of the coupled processes of rock deformation, heat transfer, pore fluid flow, and chemical reactions.

Hydrothermal mineralization is the result of complex physical and chemical reaction processes coupled nonlinearly in a certain spatial and temporal domain [14,15,16,17,18]. From a physical standpoint, thermodynamic instability can trigger convective pore fluid flow within the hydrothermal system [19,20], facilitating the upward transport of minerals from the lower crust, contributing to orebody formation. From a chemical perspective, chemical dissolution-front instability [21,22] can alter the mechanical properties of wall rock through porosity evolution during fluid–rock interactions [23,24,25,26,27], thereby affecting the permeability and distribution of ore-forming fluids. Thus, the coupling of physical and chemical processes in the mineralization process can provide a more accurate reflection of the ore-forming mechanisms, which has triggered the emergence of the mineralization rate concept. The mineralization rate concept refers to the weight of minerals dissolved or precipitated per unit time per unit volume within a mineralization system [19,28]. This value can be either positive, indicating mineral dissolution, or negative, indicating mineral precipitation. Previous studies, based on the FLAC2D simulation results, have calculated the mineralization rate and further revealed the associated ore-forming mechanisms [29,30,31]. Meanwhile, FLAC3D 5.0 also fails to account for chemical reaction processes. Moreover, owing to the more intricate nodal topology inherent in FLAC3D 5.0 computational models as compared to those of FLAC2D [32], there has been a relatively limited number of studies focusing on the calculation of the mineralization rate using this numerical platform.

Figure 2. A geological map showing (a) the distribution of porphyry and epithermal deposits in the Tulasu Basin [6,10] and (b) the Axi district [5,33,34].

Figure 2. A geological map showing (a) the distribution of porphyry and epithermal deposits in the Tulasu Basin [6,10] and (b) the Axi district [5,33,34].

Given the aforementioned considerations, this study focuses on the Axi gold deposit as a case study of the gold deposits in the Tulasu basin, utilizing computational simulation methods to explore the ore-forming processes. The computational modeling in this study is conducted in two primary stages: First, a coupled thermo-hydro-mechanical simulation is performed using FLAC3D, from which fluid temperature, velocity, and spatial coordinates are extracted to calculate the temperature gradient. Subsequently, this gradient is used to estimate the mineralization rate and to indicate the 3D spatial distribution of mineralization. This study may provide valuable insights into the mineralization mechanisms and offer new perspectives for deep exploration in the Axi gold deposit and Tulasu basin.

2. Geological Background

2.1. Regional Geology

The Tulasu Basin is bounded by the South Keguqin Mountain and the north Yili Basin thrust fault zones in the Borohoro region (Figure 2a). These two fault systems exhibit NNE strikes with dips ranging from 50° to 70°, and each persists for >10 km. Their orientation aligns with the tectonic boundary separating the Yili and Kazakhstan terranes [13]. The secondary structures in the Tulasu basin are dissected by steeply dipping, NW- to N-striking strike–slip faults that influence the distribution of Paleozoic rock units (Figure 2). Around volcanic edifices in the Jingxi, Axi, and Abiyindi areas, concentric and radial fracture systems are well developed [13,35].

The deposits in the Tulasu Basin are hosted within a Late Paleozoic continental volcanic–sedimentary sequence dominated by andesite, dacite, and volcaniclastic rocks. They exhibit a distinct spatio-temporal concentration, with their genesis being fundamentally controlled by NWW-trending fault systems and coeval magmatic-hydrothermal activity [36]. This is exemplified by representative low-sulfidation epithermal gold deposits, including the Axi, Jingxi-Yermand, and Abiyindi deposits, which collectively define a major epithermal-porphyry metallogenic belt (Figure 2).

2.2. Ore Deposit Geology

The fault F₂, which governs the Axi ore deposit, extends in directions of NW 340°, NE 25°, and NW 325°, while its inclination varies from 60° to 70° in the northern segment (No. 44), 50–80° in the middle segment (No. 44~No. 27), and 55–85° in the southern segment (No. 27), respectively (Figure 2b and Figure 3). F₂ not only serves as a primary structure for ore hosting structure but also acts as a major conduit for the large-scale pore fluid flow and accumulation of mineralizing fluids (Figure 3). The hanging wall of F₂ comprises dacitic C1d⁵⁻²(breccia lava) and C1d⁵⁻¹ (andesitic tuff, tuffaceous volcanic breccia, and andesite), whereas the footwall rock consists of andesitic C1d⁵⁻² and C1d⁵⁻⁴ (volcanic agglomerate, andesitic dacite, and dacite) (Figure 2b).

In the Axi gold deposit, thick orebodies generally are located in the gently dipping and curvilinear parts of the F₂ fault, with a progressive thinning observed as distance from these parts increases. Orebodies stretch approximately 1 km along the strike direction (Figure 3) and exhibit variable thickness (11–15 m), grade (2–16 g/t and mean 5.6 g/t) and dip (46–78°) [10]. The ore primarily consists of silicified vein-type mineralization, which constitutes the main economically valuable ore. Gold-bearing minerals within this type of ore are predominantly native gold, electrum, and sulfide that are commonly hosted within quartz-sulfide aggregates and intensely altered breccias (Figure 4). Notably, the strong positive correlation between gold grade and sulfide abundance underscores the essential role of sulfidation and related chemical processes, particularly through the reduction of gold bisulfide complexes ${Au\left(HS\right)}_2^{-}$ and sulfide-driven precipitation, in controlling gold concentration and deposition [10,37,38].

The mineralization of the deposit is subdivided into two principal stages [10,39]. The initial stage was driven by a dominantly magmatic-hydrothermal fluid system, manifesting as sericite–quartz–pyrite alteration. During this phase, ascending deep magmatic fluids, channeled along fault structures, underwent water-rock interaction with the basement series. Concurrently, a downward percolation of meteoric fluids along fractures initiated propylitic alteration in the wall rocks. Within this mixed but magmatic-dominated fluid system, gold transport occurred predominantly as chloride–sulfur complexes [40], with precipitation localized around fluid conduits within the phyllic alteration zone [6,33,34,41]. The late stage records a fundamental shift in fluid composition, marked by extensive silicification and a major influx of meteoric water (Figure 5). A sustained deep thermal engine drove vigorous hydrothermal convection [10], enabling the circulating fluids to act as an efficient leaching medium. These fluids progressively extracted metals, sulfur, and carbon from the basement and wall rocks. This process culminated in the concentration of multi-sourced gold into the evolving hydrothermal fluid, ultimately forming the deposit [42].

3. Methodology

3.1. Mathematical Model of Mineralization Rate

The mineralization rate of a given mineral refers to the variation in its mass per unit volume of rock in unit time during the mineralization process [19,28]. The mineralization rate can be positive, indicating the decomposition of mineralizing materials from the rock, or negative, signifying their precipitation from pore fluid. The mineralization rate can be derived from the perspective of mass conservation. To achieve this, a specific metal species (e.g., species q) is assumed under equilibrium conditions within a pore fluid flow system. The transport equation of this species mass is expressed as follows:

$$u{\displaystyle \frac{\partial C_q^e}{\partial x}}+v{\displaystyle \frac{\partial C_q^e}{\partial y}}+w{\displaystyle \frac{\partial C_q^e}{\partial z}}={\mathsf{\varnothing }D}_0\left({\displaystyle \frac{{\partial }^2{C}_q^e}{\partial x^2}}+{\displaystyle \frac{{{\partial }^{2}C}_q^e}{\partial y^2}}+{\displaystyle \frac{{{\partial }^{2}C}_q^e}{\partial z^2}}\right)+\mathsf{\varnothing }{R}_q$$

(1)

where $u$, $v$ and $w$ are the velocities of pore flow in the x-direction, y-direction and z-direction, respectively; $C_q^e$ is the equilibrium concentration of species $q$ in the specific chemical reactions; $\mathsf{\varnothing }$ is the porosity of the porous rock; $D_0$ is the diffusivity of the chemical species; $R_q$ is the source/sink term.

During the process of orebodies formation, the diffusion term on the right-hand side of Equation (1) is generally much smaller than the advection term and can be considered negligible. Consequently, Equation (1) can be simplified as follows:

$${MR}_q=\mathsf{\varnothing }{R}_q=u{\displaystyle \frac{\partial C_q^e}{\partial x}}+v{\displaystyle \frac{\partial C_q^e}{\partial y}}+w{\displaystyle \frac{\partial C_q^e}{\partial z}}$$

(2)

where ${MR}_q$ is the mineralization rate linked to species $q$.

In general, the equilibrium concentration of a specific metal species depends on temperature, pressure, and the presence of other relevant chemical species, as expressed by the following:

$$C_q^e=f\left(T,P,C_1,C_2\dots ,C_n\right)$$

(3)

where $T$ is temperature; $P$ is pore fluid pressure; $C$ is the concentration of a certain chemical substance; $n$ is total count of relevant chemical species. Then substituting Equation (3) into Equation (2), we obtain the following expression:

$$\begin{array}{ll}M{R}_q& ={\displaystyle \frac{\partial C_q^e}{\partial T}}\left(u{\displaystyle \frac{\partial T}{\partial x}}+v{\displaystyle \frac{\partial T}{\partial y}}+w{\displaystyle \frac{\partial T}{\partial z}}\right)+{\displaystyle \frac{\partial C_q^e}{\partial P}}\left(u{\displaystyle \frac{\partial P}{\partial x}}+v{\displaystyle \frac{\partial P}{\partial y}}+w{\displaystyle \frac{\partial P}{\partial z}}\right)\\ & +{\displaystyle \sum _{r=1}^n}{\displaystyle \frac{\partial C_q^e}{\partial C_r}}\left(u{\displaystyle \frac{\partial C_r}{\partial x}}+v{\displaystyle \frac{\partial C_r}{\partial y}}+w{\displaystyle \frac{\partial C_r}{\partial z}}\right)\end{array}$$

(4)

3.2. Numerical Simulation Workflow

The numerical simulation model is a structural translation of the 3D geological model (Figure 6), which was initially built on the premise of a conceptual model and subsequently refined with data from geological maps, drillings, and geophysical prospecting [2,43,44], and its accuracy was further guaranteed by supplementary geological investigations aimed at resolving ambiguities in original geological maps and cross-sections. In this study, a 3D geological modeling approach [45,46,47] was employed, which integrated diverse data for the Axi gold deposit, including one topographic map, 22 cross-sections and level plans, and 132 drillholes. Based on these data, the 3D geological model (a closed-surface model) was constructed using the GOCAD 17, followed by the discretization of its spatial domain into regularly shaped tetrahedral elements to form a solid model. This solid model was then converted from the GOCAD to the FLAC3D 5.0 format via a custom program bridging the structural differences in their data schemas. Following its import into FLAC3D 5.0 and the configuration of parameters (e.g., boundary conditions and rock properties), thermo-hydro-mechanical coupled simulations were conducted. Subsequently, from these simulation results, the nodal pore fluid pressure and temperature were extracted. Their respective spatial gradients were subsequently calculated using the central difference method. Simultaneously, the chemical equilibrium concentrations of key reactions in the pore fluid were determined, leading to the development of a mathematical model that describes how the first-order partial derivatives of mineralizing substance concentrations vary with temperature and pressure under evolving conditions. Mineralization rates were then computed using the data obtained from the above steps. Finally, the results, including temperature gradients, fluid migration dynamics, and mineralization rates, were cross-referenced with verified geological data (e.g., known ore body distributions) to assess their coherence. The analysis revealed key controls on the localization of gold mineralization and helped delineate prospective targets for further exploration.

3.3. Model Setup and Related Parameters

The aforementioned 3D geological modeling approach [45,46,47] integrates established data from well-constrained regions with inference in poorly constrained domains. The latter include deep unexplored areas, modeled using multi-source geophysical data and extrapolated structural trends, and denuded strata, reconstructed by projecting trends from adjacent known regions. Moreover, following previous studies [5,6], the model is constructed with a metallogenic depth of 1 km and a 1 km thick erosion layer, within a domain that extends 1.7 km north–south and 1.2 km east–west (Figure 7). The geological units in the domain comprise stratigraphic units (C1d⁵⁻¹ and C1d⁵⁻⁴), volcanic pipe facies (C1d⁵⁻²), faults, and gold orebodies, which were delineated primarily by integrating similar lithological characteristics obtained from drillholes. These units are further discretized into approximately regular tetrahedra.

Mechanical, thermal, and hydrological properties were assigned to all units in the computational models based on published studies [10,41,48,49,50]. Some properties in the geological environment varied significantly with burial depth, so these property assignments used in our model were derived from extensive comparative experiments within the general range of physical parameters for rock lithological [48], rather than from specific, fixed values. Subsequently, a parameter sensitivity analysis was conducted to systematically evaluate the influence of parameter variations on simulation outcomes, thereby minimizing the discrepancy between modeled results and field observations. On this basis, the most reasonable parameter combination was ultimately determined (

Table 1. Initial parameters for numerical modeling of the Axi gold deposit.

Property	C1d5−1	C1d5−2	Faulted Zone	C1d5−4
Density (kg·m−3)	2580	2600	2500	2590
Bulk modulus (1010 Pa)	1.6	3.0	0.6	1.7
Shear modulus (1010 Pa)	2.6	5.0	0.4	3.1
Cohesion (106 Pa)	4.9	3.5	5.0	4.0
Tensile strength (106 Pa)	8.3	4.0	2.0	3.0
Dilation angle (°)	3.0	5.0	3.2	5.3
Friction angle (°)	28	30	15	31
Porosity (%)	20	24	31	22
Permeability (10−15)	2.09	4	10	2.81
Viscosity (10−3 N s m−2)	1.0	1.0	1.0	1.0
Thermal conductivity (Wm−1K−1)	3.6	3	2.75	2.0

). In addition, the initial and boundary conditions are specified as follows:

The initial temperature at the top of the model, excluding the fault zone, was fixed at 20 °C, with a geothermal gradient of 20 °C/km increasing with depth. Additionally, the temperature at the base of the fault was set to 400 °C, decreasing upwards at a gradient of 20 °C/km. All porous spaces were initially saturated with fluids. The four-sided boundaries of the model were configured to be impermeable and thermally insulating. Conversely, the top boundary was specifically designed to allow for unrestricted fluid flow, while the bottom boundary maintained a constant pore pressure gradient to facilitate steady upward fluid influx from deeper regions.

Based on previous studies of the tectonic stress field during the mineralization period of the Axi gold deposit [51,52], the stress conditions applied to the computational model in this study are as follows:

①

The model is subjected to east–west (X-axis) extensional stress. At the top of the model, the extensional stress is set to 8.0 × 10⁷ N/m², and it decreases along the negative Z-axis at a gradient of 1.0 × 10⁴ N/m² per meter (Figure 7).

②

Simultaneously, the model experiences north–south (Y-axis) compressive stress. At the top of the model, the compressive stress is set to 8.0 × 10⁶ N/m², and it decreases along the negative Z-axis at a gradient of 1 × 10³ N/m² per meter (Figure 7).

3.4. Chemical Reaction Equilibrium Concentrations of ${Au\left(HS\right)}_2^{-}$

The ore-forming process at the Axi low-sulfidation epithermal deposit comprises two distinct stages: an early stage with slightly acidic and reducing conditions, and a subsequent later stage characterized by the input of low-salinity meteoric water [10,53,54]. This specific fluid regime, characterized by its distinct pH and redox state, established a geologically favorable setting for the stabilization of gold-transporting complexes. Under these physicochemical conditions, gold was predominantly transported as bisulfide complexes, specifically ${Au\left(HS\right)}_2^{-}$, rather than chloride complexes [37,38]. The formation of the predominant ${Au\left(HS\right)}_2^{-}$ complex can be represented by the following reaction:

$${Au\left(HS\right)}_2^{-}+0.5H_2\stackrel{R}{\iff }Au+H_{2}S+{HS}^{-}$$

(5)

where R represents the reaction rate of the chemical process.

The ore-forming fluids may contain carbonate ions, which may play a significant role in the transport and concentration of gold. Due to the presence of CO₂ in the fluids, the pH is buffered through the reactions ${CO}_2+H_{2}O\leftrightarrow H_2{CO}_3$, $H_2{CO}_3\leftrightarrow {HCO}_3^{-}+H^{+}$, and ${HCO}_3^{-}\leftrightarrow {CO}_3^{2-}+H^{+}$. This buffering effect enhances the stability of ${Au\left(HS\right)}_2^{-}$ and increases the solubility of gold in the fluid [55,56,57].

The homogenization temperatures of fluid inclusions in the Axi gold deposit range from 198.7 °C to 384.9 °C, with salinities between 2.2% and 2.6% ${NaCl}_{eq}$, and mineralization pressures ranging from 15 to 25 MPa [6,51]. In this study, the CHNOSZ 2.2.0 was used to calculate the equilibrium concentrations of the chemical reactions (Equation (5)) by setting the mineralizing chemical environment based on the preceding analysis.

The simulation results (Figure 8a) reveal that the equilibrium concentration of the chemical reaction involving ${Au\left(HS\right)}_2^{-}$ follows an approximately parabolic relationship with temperature, from 100 °C to 400 °C and from 400 °C to 500 °C. The equation for the fitting curve can be determined using the equations below:

$$\hspace{0.33em}\begin{array}{cc}\mathit{log}\left(T_i\right)=a{T}_i^2+b{T}_i+c& \left(i=1,2,...,n\right)\end{array}$$

(6)

$$f\left(T\right)=min{\displaystyle \sum _{i=0}^n}{w}_i{\left[\mathit{log}\left(T_i\right)-e_i\right]}^2\hspace{0.33em}\left(i=1,2,...,n\right)$$

(7)

where $\mathit{log}\left(T_i\right)$ is the logarithm of the equilibrium concentration of the chemical reaction; $T_i$ and $e_i$ are the horizontal and vertical coordinates of the point $i$ chosen from the curve displayed in Figure 8a. $a$, $b$ and $c$ are the three coefficients that need to be determined; $w_i$ is the weight of the chosen point $i$.

Ultimately, the mathematical model that describes how the equilibrium concentration varies with temperature is expressed as follows:

$$\begin{array}{c}log{C}_{Au(HS{)}_2^{-}}^e=-9.5\times 10^{-5}{T}^2+0.0755T-22.8\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\left(100 °\mathrm{C}\le T\le 400 °\mathrm{C}\right)\end{array}$$

(8)

$$\begin{array}{c}log{C}_{Au(HS{)}_2^{-}}^e=-2.4\times 10^{-4}{T}^2+0.216T-55.8\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\left(400 °\mathrm{C}\le T\le 500 °\mathrm{C}\right)\end{array}$$

(9)

The equilibrium concentration of ${Au\left(HS\right)}_2^{-}$ exhibits a clear correlation with temperature variation. Within the temperature range of 100–450 °C, the equilibrium concentration of ${Au\left(HS\right)}_2^{-}$ increases with temperature, showing a strong positive correlation. However, as the temperature reaches 450–500 °C, this relationship reverses, becoming negatively correlated. Furthermore, by observing the trend of the first-order partial derivative of ${Au\left(HS\right)}_2^{-}$ equilibrium concentration with respect to temperature (Figure 8c), it is evident that around 350 °C and 400 °C, the rate of change in ${Au\left(HS\right)}_2^{-}$ equilibrium concentration reaches local maxima. This indicates that the equilibrium concentration is highly sensitive to temperature variations near these points.

Figure 8b illustrates that at the fixed temperature points of 100 °C, 300 °C, and 500 °C, the equilibrium concentration of $Au(HS{)}_2^{-}$ does not exhibit significant changes with increasing pore fluid pressure, remaining relatively stable. This indicates that during the mineralization process of the Axi gold deposit, pore fluid pressure variations are not the primary factor controlling changes in the chemical reaction equilibrium concentration of $Au(HS{)}_2^{-}$.

Therefore, the equilibrium concentration of $Au(HS{)}_2^{-}$ during the mineralization of the Axi gold deposit is predominantly controlled by temperature, whereas the influence of pore fluid pressure variations is negligible. As deep ore-bearing hydrothermal fluids (~400 °C) migrate toward shallower areas, the pronounced decrease in temperature leads to a substantial reduction in the equilibrium concentration of $Au(HS{)}_2^{-}$, thereby driving efficient gold precipitation. However, once the fluid temperature falls below 200 °C, the impact of the temperature change on the equilibrium concentration of $Au(HS{)}_2^{-}$ chemical reaction gradually diminishes.

4. Simulation Results and Discussion

4.1. Temperature Gradient

In ore-forming systems, the temperature gradient not only influences the flow pathways of hydrothermal fluids but also significantly enhances the mineralization potential of specific geological regions [19,57,58]. Furthermore, the temperature gradient significantly affects ore grade and scale, serving as a critical parameter for calculating mineralization rates. This study estimates the temperature gradient in the X, Y, and Z directions using the central difference method, based on pore fluid temperature data obtained from numerical simulation results.

On the horizontal planes, the distribution characteristics of the X and Y temperature gradients exhibit significant differences (Figure 9). Specifically, the X temperature gradient shows a symmetrical distribution of positive and negative bands on both sides of the fault, with local maxima, shown in red, and corresponding minima, shown in blue, forming at the fault bends. In contrast, at locations where the fault is not obviously bent, the variations in the positive and negative bands of the X temperature gradient are modest, with no pronounced local extrema formed. In contrast, the Y temperature gradient, although generally symmetrical like the X temperature gradient, shows two intersections of positive and negative bands along the fault, occurring precisely at the fault bends. At these intersections, the colors gradually shift from red/blue to blue/red, reflecting a reversal of values from positive/negative to negative/positive, whereas in regions without significant curvature, the Y temperature gradient retains its symmetrical pattern. This distinctive distribution suggests that heat transfer along the Y temperature direction at the fault bends is particularly sensitive to multiple factors, including local stress variations, fault dip steepness, and changes in fault inclination.

On the cross-section, the local maxima and minima of the Z-temperature gradient exhibit a symmetrical distribution on both sides of the fault, transitioning from steep to gentle inclinations (Figure 10). However, as the elevation decreases below 1.2 km, the curvature of the fault gradually diminishes, and correspondingly, the bands of the Z temperature gradient also decrease. Overall, the temperature gradient field closely aligns with the transitions from steep to gentle fault segments, potentially indicating a potential connection between localized temperature gradient anomalies and the migration, accumulation, and intensification of thermal fluids.

The ore-forming fluids of the Axi gold deposit, a mixture of low-temperature meteoric water and high-temperature magmatic fluids, accumulated, filled, and ultimately precipitated to form orebodies [6]. Numerical simulations reveal that fluids from depths greater than 1.4 km flow downward along the fault, whereas those from depths of less than 1.2 km flow upward, converging and mixing at approximately 1.3 km depth. Subsequently, these mixing fluids continue to migrate laterally along the fault, and reaccumulate in regions characterized by complex geometries (e.g., bends and irregularities) (Figure 9). These phenomena may be related to the influence of the stress field and temperature gradient. The shallow fluids flow downward along the fault under the influence of the stress field, while the deep ore-forming fluid, with higher temperatures and a continuous influx of deep heat, are likely driven upward by these thermodynamic conditions. The mixing of these fluids can reduce the overall flow velocity, thus prolonging the residence time of the ore-forming fluids and facilitating the decomposition of gold complexes (e.g., $Au(HS{)}_2^{-}$) and subsequent gold precipitation.

4.2. Metallogenic and Exploration Implications

On the horizontal planes, the negative mineralization rate of the Axi gold deposit is primarily located in the hanging wall, consistent with the distribution characteristics of the known ore bodies, and its spatial distribution clearly demonstrates significant control by the orientation of the fault (Figure 11). The negative mineralization rate increases with the degree of curvature of the fault, and its intensity shows a significant high–weak intermittent distribution. For instance, in the areas along exploration lines 19 and 44, which feature pronounced bends, locally elevated values of negative mineralization rates are observed (Figure 11a). Notably, the negative mineralization rate in the area corresponding to line 44 surpasses that of line 19, suggesting a greater precipitation of gold. This observation is consistent with the actual distribution pattern of gold ore bodies along these two exploration lines. Meanwhile, the negative mineralization rate exhibits relatively weaker intensity in the areas where fault bends are not significant and, with the higher negative mineralization rate in the fault bending areas, results in a high-weak interval distribution (Figure 11b).

On the cross-section, negative mineralization rates are primarily concentrated in the fault breccia zone above an elevation of 1.2 km, with their orientation roughly parallel to the main fault plane (Figure 12). The intensity of negative metallogenic rates exhibits a significant positive correlation with the angle of transition from steep to gentle faulting and their lengths. For instance, in exploration lines 43, 0, and 16, as the length of gently dipping fault segments progressively increases, the scope of negative mineralization rate gradually intensifies. Notably, the negative mineralization rates show a progressively increasing trend in the deeper regions of the profiles from exploration line 16 to 56, with a particularly pronounced enhancement near exploration line 56.

Overall, the mineralization process of the Axi gold deposit is the result of coupled processes such as geological structure, heat transfer, fluid flow, and hydrothermal reactions. As the temperature decreases and sulfide reactions occur, causing the gold-bearing fluid to become unstable, previous studies have shown that temperature gradient variations in the Axi gold deposit significantly influence the equilibrium concentration of ${Au\left(HS\right)}_2^{-}$ [10,37,38,58], playing a critical role in the ore precipitation process. Numerical simulation results further confirm that a substantial concentration of known orebodies is distributed in areas where mineralization occurs (mineralization rate < 0). These areas are distinctly characterized by the overlap of significant temperature gradients and fluid accumulation, strongly suggesting that such overlapping patterns are critical indicators for identifying potential exploration targets.

Insights from the spatial correlation model, integrated with previous studies [10,49], were used to delineate a prospective target with significant mineralization potential at depth north of the known orebodies (Figure 13). This target is aligned with the NE-trending downward extension of the mineralization, suggesting it may represent a conduit for paleo-fluid flow or a structurally controlled branch from the main orebody. In conclusion, the combination of favorable geological conditions, predictive basis, and presence of mineralized indicators underscores the considerable exploration potential of this area, which should be considered a high-priority target for anomaly verification.

5. Conclusions

This paper establishes a thermal–fluid–mechanical–chemical (THMC) coupling model for the Axi gold deposit and reveals through numerical simulation the patterns of favorable structural features for gold mineralization and precipitation. However, due to the limitations inherent in the existing FLAC3D 5.0, certain physical processes such as mass transport, magma solidification, and hydrothermal boiling were not incorporated into the model. Furthermore, only equilibrium chemical reaction processes were considered, while non-equilibrium chemical reaction processes were overlooked. The main conclusions are as follows:

(1)

Chemical reaction equilibrium concentrations of ${Au\left(HS\right)}_2^{-}$ indicate that temperature gradients are crucial factors influencing the precipitation of ore-forming fluids, whereas pressure gradients play a negligible role.

(2)

The precipitation of mixing ore-forming fluids is preferentially localized in zones with sharp temperature gradients. Following the mixing of meteoric and magmatic fluids, pronounced temperature gradients are generated at locations with relatively large fault dips and gentle undulations. These gradients destabilize the ${Au\left(HS\right)}_2^{-}$ complexes, thereby initiating efficient gold deposition.

(3)

The coupling of multiple physical–chemical processes at the same location of the fault is likely to be the key factor controlling the formation of the Axi gold deposit, as quantitatively expressed by the negative mineralization rate. Based on this metallogenic regularity, it is inferred that a potential mineralization zone has been delineated in the deep northern extension of the known ore bodies.