Composition of Chlorite as a Proxy for Fluid Evolution and Gold Precipitation Mechanisms in the Jinshan Gold Deposit, Dexing District, South China

Abstract

The physicochemical controls on gold precipitation in orogenic gold deposits remain poorly constrained, with traditional fluid inclusion and isotopic studies often yielding ambiguous results due to overprinting or incomplete records. This study addresses this challenge using chlorite—a sensitive mineral proxy for fluid conditions—as a quantitative sensor in the Jinshan orogenic gold deposit (>200 t Au) of the Jiangnan orogenic belt, South China. Hosted in Neoproterozoic phyllite within NE–NNE-trending ductile–brittle shear zones, Jinshan features auriferous quartz–polymetallic sulfide veins with prominent chlorite alteration. Integrating high-resolution SEM-EPMA analyses of multi-generational chlorite with thermodynamic modeling, we reconstruct the temporal evolution of temperature, oxygen fugacity (fO₂), pH and sulfur fugacity (fS₂) during ore formation. Four paragenetic stages are identified: Stage 1 (ankerite–quartz), Stage 2 (pyrite–arsenopyrite–quartz), Stage 3 (quartz–gold–polymetallic sulfide), and Stage 4 (chlorite–carbonate–quartz). Electron microprobe analysis reveals that the chlorite composition changes from Fe-rich chamosite (Stage 2) to Mg-rich clinochlore (Stage 3) and then to Fe-rich chamosite (Stage 4). Chlorite from Stage 2 (Chl-1) formed metasomatically at low fluid/rock ratios, while Stage 3 and 4 chlorites (Chl-2 and Chl-3) precipitated directly from higher fluid/rock ratio fluids. Chlorite compositions record a critical Stage 2–3 transition involving cooling from ~320 °C to ~260 °C, reduction (log fO₂ from −33.6 to −39.7), and alkalinization, and sulfur fugacity remained stable within a narrow range (log fS₂ = −13.6 to −8.0), followed in Stage 4 by minor reheating to ~280 °C, re-acidification, and a slight rebound in oxygen fugacity. Thermodynamic simulations reveal that the destabilization of Au(HS)₂⁻ complexes, primarily driven by the synergistic effects of cooling, pH increase, and decreasing oxygen fugacity, triggered gold precipitation during the main ore stage. Results demonstrate that abrupt cooling coupled with fluid alkalinization and reduction exerted the dominant control on gold precipitation in Jinshan, resolving long-standing debates on ore-forming mechanisms and highlighting chlorite as a robust quantitative sensor for fluid evolution.

1. Introduction

Chlorite, a group of 2:1-type (T-O-T) layered hydrous phyllosilicates, exhibits a general formula of (Mg, Fe²⁺, Fe³⁺Al)₆[(Si, Al)₄O₁₀](OH)₈, characterized by alternating talc-like tetrahedral-octahedral-tetrahedral (T-O-T) sheets and brucite-like octahedral interlayers enriched in Mg²⁺, Fe²⁺, Al³⁺, and structural hydroxyl groups [1,2,3]. This structural framework accommodates extensive solid solution substitutions, including tetrahedral Al^IV-for-Si (Al^IV), octahedral Mg-for-Fe²⁺/Fe³⁺, and variable Fe³⁺/Fe²⁺ ratios, governed by coupled Tschermak (Al^IVAl^VI ↔ Si-Mg/Fe) and di-trioctahedral (3Mg/Fe ↔ 2Al^VI + □) exchanges [4,5,6,7]. These substitutions are thermodynamically sensitive to prevailing P-T-X conditions, yielding end-members such as clinochlore (Mg-rich), chamosite (Fe-rich), and sudoite (Al-rich), which serve as proxies for fluid–rock equilibria in hydrothermal systems [1,8,9].

Recent advances in chlorite thermodynamics have transformed it from a qualitative alteration mineral into a quantitative sensor of ore-forming conditions (e.g., [8,9]). Calibrated thermodynamic models incorporating non-ideal mixing and accurate Fe³⁺ determination now allow simultaneous retrieval of crystallization temperature (150–400 °C), fO₂, and aH₂S with uncertainties commonly <25 °C and <1 log unit fO₂ [1,9,10,11]. Beyond thermodynamics, coupled EPMA and LA-ICP-MS mapping reveals cryptic zoning and trace-element (such as V, Cr, Li, Ba, Sn, Zn) partitioning that track fluid mixing, redox oscillations, and metal scavenging during sulfide precipitation [11,12,13,14].

The Jiangnan orogenic belt (JOB) in South China is a world-class orogenic gold province with >1000 t combined gold endowment formed during Late Neoproterozoic continental collision [15,16]. The Jinshan gold deposit (>200 t Au resource; grade: ~3–20 g/t) in the northeastern Jiangxi Province represents a classic shear-zone-hosted orogenic gold system within Neoproterozoic phyllite along NE–NNE-trending ductile–brittle faults. Previous geochronological studies have proposed a protracted mineralization history from the Neoproterozoic to the Mesozoic (e.g., [17,18]). Furthermore, the consensus from H-O-C-He-Ar isotopic systems favors a dominant metamorphic fluid source for the ore formation (e.g., [19,20,21,22]). This robust isotopic evidence corroborates the orogenic gold deposit model, which has consequently gained broad scholarly acceptance. Mineralization is characterized by auriferous quartz–polymetallic sulfide veins accompanied by intense silicification, sericitization, and chloritization, with chlorite associated with gold mineralization [17].

Despite extensive prior work on fluid inclusions, stable isotopes, and geochronology [23,24,25], the physicochemical triggers for gold deposition remain contentious. Proposed mechanisms include: (1) phase separation induced by fault-valve cycling, (2) mixing between metamorphic and meteoric fluids, (3) sulfidation of Fe-rich wall rocks, and (4) redox shifts driven by interaction with graphitic schists [26]. Critically, none of these models have been tested against quantitative mineral-scale records of evolving T-fO₂-pH- fS₂ conditions. To address this gap, we integrated high-resolution SEM and EPMA analyses of multi-generational chlorite with state-of-the-art thermodynamic and redox modeling. We reconstructed continuous T-fO₂-pH-fS₂ trajectories through the main ore stage, established spatial zonation patterns, and directly linked chlorite Fe³⁺/ΣFe and Mg# evolution to gold precipitation efficiency. These results provide a mineralogically constrained fluid evolution model for Jinshan gold deposit, help to reconcile long-standing debates on ore-forming mechanisms, and establish chlorite-based vectoring criteria applicable to orogenic gold exploration throughout the Jiangnan Belt.

2. Geological Setting

2.1. Regional Geology

The South China Block formed through Neoproterozoic (ca. 970–820 Ma) amalgamation of the Yangtze and Cathaysia blocks along the Jiangnan orogenic belt (Figure 1A; [26,27,28,29,30]). This belt comprises a folded and thrust-faulted assemblage of Meso- to Neoproterozoic greenschist facies, metavolcaniclastic and metasedimentary rocks (Shuangqiaoshan, Lengjiaxi, and equivalent groups) unconformably overlain by a thin, discontinuous Sinian–Phanerozoic cover dominated by Cretaceous continental red beds [31,32]. Post-amalgamation evolution involved three major tectono-magmatic episodes that profoundly influenced metallogeny: (1) early Paleozoic (ca. 460–400 Ma) intracontinental orogeny (Wuyi–Yunkai orogeny), (2) Triassic (ca. 250–205 Ma) Indosinian continent–continent collision, and (3) Jurassic–Cretaceous (ca. 180–90 Ma) subduction and extension along the eastern Eurasian margin [15,27,28,33]. These events generated widespread NE- to NNE-trending fault systems that served as conduits for multiple generations of hydrothermal fluids, resulting in one of China’s premier gold (±Cu, Pb-Zn-Ag) metallogenic provinces with combined endowments exceeding 1200 t Au [15,16].

Figure 1. Schematic geological maps of South China and the Jiangnan Orogen. (A) Simplified geological map of South China showing regional tectonics (modified from [34]). (B) Simplified geological map of the Jiangnan Orogen showing structures, Proterozoic successions, magmatic rocks, and ore deposits (modified from [35]).

The Dexing ore district in northeastern Jiangxi Province lies in the eastern segment of the Jiangnan orogen and exemplifies this polyphase mineralization history (Figure 1B). The district is dominated (>70% exposure) by the Neoproterozoic Shuangqiaoshan Group, a >8 km-thick sequence of weakly metamorphosed (greenschist facies) turbiditic phyllite, metasiltstone, and minor metavolcanic rocks that host most gold and polymetallic deposits in the region [36,37]. In the southeast, the unconformably overlying Dengshan Group comprises shallow-marine to continental volcaniclastic rocks [17]. A markedly incomplete Phanerozoic cover is preserved only in fault-bounded basins and includes isolated Cambrian–Ordovician carbonate and siliciclastic units, Devonian–Triassic shallow-marine strata, and voluminous Middle–Late Jurassic continental volcanic rocks of the Ehuling Formation (rhyolitic to dacitic ignimbrites and lavas) together with Cretaceous red-bed basins [38,39,40].

The Dexing district hosts three world-class deposits developed in close spatial association: (1) the giant Dexing porphyry Cu-Mo-Au system (~8 Mt Cu, ~300 t Au), (2) the Yinshan epithermal-polymetallic Cu-Pb-Zn-Ag-Au deposit, and (3) the Jinshan orogenic gold deposit [18,19,41,42]. Gold mineralization at Jinshan is structurally controlled by second- and third-order splays of the regional NE-trending Jiangnan deep fault zone, occurring as auriferous quartz–sulfide veins and disseminated ores within ductile–brittle shear zones cutting Shuangqiaoshan Group phyllites [17,43]. The intimate association of orogenic gold with earlier porphyry and later epithermal mineralization in the same structural corridor highlights the exceptional fertility and protracted hydrothermal history of the eastern Jiangnan orogen.

2.2. Ore Deposit Geology

The Jinshan gold deposit is situated in the southeastern part of the Dexing ore district within the eastern Jiangnan orogenic belt, northeastern Jiangxi Province, South China (Figure 1B). The NEE-trending orefield spans ~60 km² and contains proven gold resources of >200 t Au [18,44]. The Dexing district is defined by the regional Northeast Jiangxi deep fault zone along its southeastern margin, the Le’an River deep fault zone to the northwest, and the intervening Sizhoumiao synclinorium [18,45]. Between these bounding structures, a network of subsidiary NE- to NNE-trending, NW-dipping shear zones has developed, including the ductile Jiangguang–Fujiawu and Bashiyuan–Tongchang zones, as well as subordinate, nearly EW-trending shears such as the Jinshan–Xijiang and Yangshan zones [18,23,46]. Regional NNW-trending faults, in concert with folds like the Sizhoumiao synclinorium and Yinshan anticline, control the distribution of Au-Ag-Cu polymetallic mineralization [47,48]. The Jinshan gold deposit is specifically hosted in the subordinate, nearly EW-trending Jinshan-Xijiang ductile–brittle shear zone, with kinematic indicators suggesting sinistral strike–slip motion synchronous with Middle Jurassic (ca. 167–161 Ma) regional tectono-magmatic events [15,23,48].

Ore-hosting strata consist of the Mesoproterozoic Shuangqiaoshan Group, a greenschist facies metavolcaniclastic sequence striking 110–150° and dipping 10–35° NW [18,45,49]. This group is subdivided into three lithostratigraphic [23,25,50]: (1) the lower segment (Pt₃sh^III−1), ~300 m thick, comprises carbonaceous, sericite-, quartz-, and chlorite-bearing slates dominant in the southern orefield, constituting ~20% of the area and bounded by shear contacts; (2) the middle segment (Pt₃sh^III−2), the principal ore host (~500 m thick in the mine area), consists of tuffaceous, sandy, and silty phyllites intercalated with lenticular metabasalts, distributed in the northwestern and central sectors and in fault contact with the lower unit [19,51,52,53,54]; and (3) the upper segment (Pt₃sh^III−3), >300 m thick, is characterized by medium- to thick-bedded gray silty slates with interbeds of sandy slate, tuffaceous slate, and metasandstone, occurring in the northeastern area and delimited by shear planes from the middle segment. Magmatic activity is subdued, with cryptic intrusions exposed locally as dikes of pyroxene diorite, diabase, and granodiorite [19,46,47]. The dikes exposed in the mining area, especially the diabase, show significant spatial, temporal, and genetic links with the Jinshan gold deposit [54]. The ages of the granodiorite porphyry and diabase are reported by previous studies as approximately 905 Ma (n = 10) and about 854 Ma (n = 1), respectively, but the latter, based on only a single sample, remains limited in representativeness. Recent geochronological studies indicate that the diabase in this area likely formed during the Late Jurassic (ca. 170 Ma). This age aligns with the previously proposed timing of Yanshanian mineralization, suggesting a potential link to mineralization associated with the Yanshanian period.

The NE-trending F6 fault bisects the deposit into the western Wanjiawu segment and eastern Yangshan segment (Figure 2B; [55]). Mineralization manifests as altered-rock (disseminated) style dominant in Wanjiawu (3–10 g/t Au) and auriferous quartz-vein style prevalent in Yangshan (5–20 g/t Au), both controlled by progressive ductile–brittle deformation within the shear zones [19,30]. Six principal orebodies are delineated: I and II in the Wanjiawu shear zone, and III–VI in the Yangshan shear zone, collectively accounting for the ~200 t Au endowment [18,44]. Hydrothermal alteration assemblages include carbonatization, chloritization, silicification, sericitization, albitization, and pyritization, with carbonatization, silicification, and chloritization exhibiting the strongest spatial and genetic links to gold precipitation [15,30,46,49]. Two generations of native gold have been identified [19]; EDS analyses reveal exceptionally high fineness (940–971), with Ag present as the sole minor alloying element [45].

Figure 2. Geological map and section of the Jinshan gold deposit. (A) Simplified geological map of the Jinshan gold deposit (modified from [23]). (B) Cross-section of the ore body at the Jinshan gold deposit (modified from [23]).

3. Sampling and Analytical Methods

The samples for this study were collected from the gold orebody in the Yangshan segment of the Jinshan gold deposit, at the −112 m, −115 m, and −180 m levels. These samples represent the two main mineralization types: quartz vein-hosted gold mineralization at the −112 m level (Figure 3A,B) and altered rock-hosted gold mineralization at the −115 m and −180 m levels (Figure 3C,D). To focus on chlorite paragenesis and geochemistry, representative specimens of mineralized quartz veins, altered ore, and adjacent wall rock were selected. Following initial hand-specimen description, samples were cleaned, labeled, and prepared as 16 polished thin sections and mounts for petrographic and in situ analysis. Based on ore microscopy, chlorites from distinct paragenetic stages were targeted for microstructural characterization and major element analysis by electron probe microanalysis (EPMA). Analytical protocols are detailed below.

Figure 3. Photographs of rock and ore samples from the Jinshan Gold Deposit. (A) Phyllite rich in coarse-grained pyrite; (B) banded light grayish-green phyllite ore; (C,D) banded quartz vein ore, with pyrite veinlets crosscutting ankerite–quartz veins; (E) grayish-black mylonite, containing sulfide veinlets cut by late-stage chlorite–carbonate veins; (F) calcite vein within a late-stage fault containing breccia fragments in mylonite. Mineral abbreviations: Qz—quartz, Py—pyrite, Ccp—chalcopyrite, Apy—arsenopyrite, Ank—ankerite, Chl—chlorite.

3.1. Petrography and SEM-EDS

Petrographic observations were conducted at the National Key Laboratory of Uranium Resources Exploration-Mining and Nuclear Remote Sensing, East China University of Technology, using a Leica DM2700 optical microscope (Leica Microsystems, Wetzlar, Germany) for polished thin sections. To characterize chlorite microstructure, morphology, and preliminary composition, samples were carbon coated and examined using a Zeiss Sigma 300 field-emission scanning electron microscope (SEM) (Carl Zeiss AG, Oberkochen, Germany) equipped with an Aztec X-Max 100 energy-dispersive spectrometer (EDS) (Oxford Instruments, Abingdon, UK). Backscattered electron (BSE) imaging was utilized to delineate compositional zoning, heterogeneity, and textural relationships in chlorite. Operating conditions included an accelerating voltage of 10–25 kV, beam current of 10 nA, and high-vacuum mode, with BSE image acquisition times of 120 s.

3.2. Electron Probe Microanalysis

Major element compositions of chlorite were determined using a JEOL JXA-8100 electron probe microanalyzer (JEOL Ltd., Tokyo, Japan) at the National Key Laboratory of Uranium Resources Exploration-Mining and Nuclear Remote Sensing, East China University of Technology. Samples were carbon coated to ensure conductivity, and analytical points were selected based on BSE imaging to maximize representativeness and avoid inclusions. Analyses were performed at 15 kV accelerating voltage, 20 nA beam current, and 2 μm beam diameter, with data corrected in real time using the ZAF (atomic number, absorption, fluorescence) procedure. Standards included rutile (Ti), apatite (Ca), orthoclase (K), jadeite (Na), rhodonite (Mn), pyrope (Fe, Al, Mg), sodalite (Si), and olivine (Si). The suite of standards covers all major elements in chlorite, enabling accurate quantification. Under these operating conditions and with ZAF matrix corrections applied, the relative error for major elements (e.g., Si, Al, Fe, Mg; typically >1 wt%) is better than ±2%, with an absolute error generally within ±0.5 wt%. For minor elements (e.g., Na, K, Mn, Ti), good precision is also maintained at concentrations above their detection limits.

Under the above analytical conditions, to rigorously distinguish fine-grained Chl-1 from Chl-2 and avoid mixed analyses, we implemented the following sampling protocol: all analysis points were selected under the guidance of high-resolution BSE images, and rapid BSE X-ray mapping was performed on the target area prior to point analysis to identify and avoid interference from micro-inclusions, compositional zoning, and mineral boundaries. The analysis points were focused on the central part of the target mineral phase and analyzed using the minimum beam diameter (1–2 μm).

Chlorite structural formulae were calculated on the basis of 14 oxygen atoms (anhydrous basis). The octahedral sheet has a fixed capacity of 6 cation sites per formula unit, which are occupied by both cations and structural vacancies (□). All anion sites (O + OH) are fully occupied, corresponding to a theoretical negative charge of 28. The Fe³⁺/Fe²⁺ ratio was estimated using the charge balance method [56]. Tetrahedral Al (Al^IV) was calculated as Al^IV = 4 − Si, with the remaining Al assigned to octahedral coordination (Al^VI). The total number of octahedral cations was determined by iteratively adjusting Fe³⁺and vacancies until charge balance was achieved. To mitigate potential contamination from micro-inclusions (e.g., feldspar or mica) or secondary alteration [57], strict quality filters were applied—only analyses with total alkali oxide contents (Na₂O + K₂O + CaO) < 0.5 wt% were retained—following established criteria [58,59,60].

3.3. Thermodynamic Modeling

Thermodynamic calculations were performed using the Geochemist’s Workbench software package (v.12.0; [61]). The database was based on thermo.com.V8. R6+. dat, supplemented with thermodynamic data for Au-, As-, and Fe-bearing species from SUPCRT92 [62,63]. Equilibrium constants (log K) for all relevant reactions were recalculated according to the pressure–temperature conditions of this study. Simulations were conducted in the REACT module using a reaction path model, with a fixed pressure of 2 kbar and salinity of 5 wt% NaCl equivalent. The system was buffered by pyrite, and log aH₂S was fixed at −3.5. Gold solubility is defined by the following key equilibrium reactions: (1) Au(s) ⇌ Au⁺ + e⁻; (2) Au⁺ + 2HS⁻ ⇌ Au(HS)₂⁻; (3) H₂S(aq) ⇌ H⁺ + HS⁻; and (4) FeS₂(pyrite) + 2H⁺ + 2e⁻ ⇌ Fe²⁺ + 2H₂S(aq). By systematically varying temperature (260–320 °C), pH (ΔpH ≈ +2.0), and oxygen fugacity (Δlog fO₂ ≈ −6.0), the individual and combined effects of cooling, alkalization, and oxidation on gold solubility were simulated. The magnitude of gold precipitation was quantitatively evaluated by the decrease in total dissolved gold concentration.

4. Results

4.1. Mineralogy and Paragenesis

Systematic petrographic observations, field mapping, hand-specimen examination, and optical and electron microscopy, reveal that the Jinshan gold deposit is characterized by a suite of ore minerals dominated by pyrite (Py), arsenopyrite (Apy), chalcopyrite (Ccp), galena (Gn), sphalerite (Sp), and native gold (Au), with subordinate pyrrhotite (Po), and tetrahedrite (Ttr). Gangue minerals include quartz (Qz), ankerite (Ank), albite (Ab), calcite (Cal), chlorite (Chl), siderite (Sd), dolomite (Dol), rutile (Rt), and sericite (Ser). Based on mineral assemblages, paragenetic sequences, cross-cutting relationships, and textural features, the continuous evolution of a single hydrothermal system can be divided into four stages (Figure 4): (1) ankerite–quartz, (2) pyrite–arsenopyrite–quartz, (3) quartz–gold–polymetallic sulfide, and (4) chlorite–carbonate–quartz. Chlorite is extensively developed from Stage 2 to Stage 4 and exhibits a close spatial, temporal, and genetic association with the gold mineralization process (Figure 4, Figure 5 and Figure 6). Specifically, chlorite co-precipitates with pyrite and arsenopyrite in Stage 2; Fe-Mg-rich chlorite is intimately associated with native gold in Stage 3; while in the post-ore Stage 4, Fe-rich chlorite again becomes dominant. The compositional evolution of these multi-generational chlorites systematically records changes in the physicochemical conditions of the ore-forming fluids and indicates key shifts in the efficiency of gold precipitation.

Figure 4. Mineral paragenesis and stage division in Jinshan gold deposit. The line thickness represents the relative abundance of the corresponding mineral, and the dashed line indicates that the mineral rarely appears.

Figure 5. Photomicrographs of sulfide and gangue minerals from the Jinshan Gold Deposit. (A) ankerite–quartz vein; (B) albite and quartz (Qz1), with Qz1 crosscut by Qz2; (C) pyrite (Py2) overgrowing around Py1, with fractures filled by galena, gold, etc.; (D) pyrite (Py2) containing inclusions of arsenopyrite and pyrrhotite, coexisting with arsenopyrite and pyrite; (E) pyrite–arsenopyrite occurring as veins, surrounded by chlorite (Chl) and sericite (Ser); (F) sericite (Ser) and minor chlorite (Chl) replacing albite, with albite showing relict texture; (G) chlorite (Chl) replacing albite, coexisting with rutile; (H) native gold filling within or along fractures of Py2, with minor chalcopyrite distributed around pyrite; (I) gold–polysulfide–quartz vein; (J) pyrite (Py3) in gold–polysulfide–quartz vein, often exhibiting core–mantle–rim texture; (K) Paragenesis of native gold, galena, sphalerite, chalcopyrite, and ankerite, with chalcopyrite and sphalerite forming a solid solution texture; (L) Galena, native gold, etc., filling fractures in pyrite. Mineral abbreviations: Qz—quartz, Au—native gold, Ser—sericite, Py—pyrite, Apy—arsenopyrite, Gn—galena, Po—pyrrhotite, Ccp—chalcopyrite, Sp—sphalerite, Chl—chlorite, Cal—calcite, Ab—albite. Dashed lines indicate the boundaries between different pyrite stages.

Figure 6. Microtexture of chlorite from the Jinshan gold deposit. (A) Pyrite–arsenopyrite–quartz vein with chlorite (Chl-1) distributed along fractures in the ore vein; (B) chlorite (Chl-1) occurring in intergranular spaces or fractures of pyrite, coexisting with sericite; (C) chlorite (Chl-1) replacing feldspar, with the feldspar mineral showing a relict replacement texture; (D) chlorite (Chl-2) present within the gold–polysulfide–quartz vein; (E) chlorite (Chl-2) exhibiting a vermicular form and distributed adjacent to chalcopyrite; (F) chlorite (Chl-2) paragenetically associated with chalcopyrite, sphalerite, and tetrahedrite; (G) calcite–chlorite vein cross-cutting the pyrite–arsenopyrite–quartz vein; (H) chlorite (Chl-3) and calcite forming a vein, with quartz displaying a comb texture; (I) chlorite (Chl-3) forming scaly aggregates, developing fan-shaped, fibrous-radiating, and needle-like spherulitic textures. Mineral abbreviations: Qz—quartz, Ser—sericite, Py—pyrite, Apy—arsenopyrite, Gn—galena, Ccp—chalcopyrite, Sp—sphalerite, Chl—chlorite, Cal—calcite, Ab—albite.

Stage 1: Ankerite–Quartz

This initial stage is marked by barren to weakly mineralized carbonate–quartz veins (assemblage: Ank + Qz + Py ± Ab ± Ser), signifying the onset of hydrothermal activity and commonly overprinted by later veins. Ankerite forms euhedral to subhedral grains in disseminations or veinlets (~500 μm wide), filling fractures to establish the vein framework (Figure 5A). Quartz occurs as subhedral to anhedral grains (~200 μm) with undulatory extinction, intergrown with ankerite as the primary vein fill (Figure 5B). Subhedral to anhedral pyrite (10–200 μm) displays porous interiors, scattered distribution, and replacement by later pyrite generations (Figure 5C). Albite appears as subhedral to anhedral crystals with serrated margins, locally associated with sericite, indicative of early sodic alteration (Figure 5B).

Stage 2: Pyrite–Arsenopyrite–Quartz

Stage 2 veinlets (assemblage: Py + Apy + Qz ± Ser ± Chl-1 ± Au ± Rt ± Po ± Ccp) represent sulfide precipitation, cross-cutting Stage 1 assemblages and hosting minor gold. Pyrite forms euhedral to subhedral cubes (0.03–2 mm) enclosing inclusions of arsenopyrite, pyrrhotite, rutile, and Stage 1 pyrite remnants (Figure 5C,D); fractures are infilled by gold and galena, and oriented veinlets are typically <1 mm wide (Figure 5C). Needle-like to columnar, subhedral to anhedral arsenopyrite (10–40 μm) coexists with pyrite in composite veins, where banded textures reflect shear-zone control (Figure 5E). Subhedral to anhedral quartz (10–50 μm) transects Stage 1 quartz and lines vein selvages with pyrite–arsenopyrite (Figure 5B). Lath-shaped sericite intergrows with chlorite, replacing albite to produce alteration envelopes (Figure 5E,F). First-generation chlorite (Chl-1) manifests as vermicular or fine-scaly aggregates (<50 μm), irregularly rimming or veining albite, and associated with pyrite and rutile (Figure 5F,G); this reflects Fe-Mg metasomatism of feldspar, with local radial-fibrous habits. Sparse native gold (<10 μm) fills pyrite fractures or pores (Figure 5H). Pyrrhotite occurs as ~10 μm inclusions (Figure 5D), and chalcopyrite is sporadic in sericite–pyrite veins (Figure 5H). These veinlets transect ankerite–quartz veins, with pyrite encapsulating earlier relicts before succumbing to polymetallic overprint.

Stage 3: Quartz–Gold–Polymetallic Sulfide

The principal ore-forming stage (assemblage: Qz + Au + Ccp + Sp + Gn ± Py ± Cal ± Sd ± Ttr ± Chl-2 ± Rt) yields high-grade quartz–gold–polymetallic sulfide veins, accounting for the bulk of gold endowment. Fine-grained anhedral quartz (~40 μm) intergrows with sulfides in open-space fillings (Figure 5I). Pyrite (0.2–2 mm, euhedral to subhedral) exhibits core–mantle–rim zoning with porous mantles, overprinted by chalcopyrite–sphalerite–galena (Figure 5J). Irregular chalcopyrite (30–100 μm) locally associates with tetrahedrite and forms chalcopyrite disease (emulsion exsolutions) in sphalerite (Figure 5J,K). Anhedral sphalerite (~100 μm) hosts chalcopyrite exsolutions and tightly intergrows with galena (Figure 5K). Fine-grained anhedral galena (5–100 μm) occurs as massive aggregates or veinlets infilling pyrite voids (Figure 5L). Native gold (5–30 μm, embayed or lamellar) predominantly occupies fractures in quartz interstices or pyrite pores, associated with galena–sphalerite–chalcopyrite, comprising >70% of total gold (Figure 5K,L). Fine-grained anhedral calcite (~100 μm) coexists with gold and sulfides (Figure 5K). Minor second-generation chlorite (Chl-2) accompanies siderite and rutile, with siderite locally replacing ankerite. These veins transect Stage 2 pyrite–arsenopyrite assemblages, with gold–polymetallic sulfides occupying early pyrite porosities prior to late carbonate veining.

Stage 4: Chlorite–Carbonate–Quartz

Late barren carbonate–chlorite veins (assemblage: Cal + Chl-3 + Qz ± Dol) herald the waning of hydrothermal activity. Subhedral quartz (~100 μm) forms clean, sulfide-free vein margins (Figure 6H). Calcite veinlets (0.5–2 mm wide) intergrow with chlorite (Figure 6H,I). Third-generation chlorite (Chl-3) appears as scaly aggregates (50–200 μm) in veins or halos, replacing early pyrite and quartz without associated gold (Figure 6I). These veins crosscut all preceding stages (Figure 6G), with chlorite preferentially corroding pyrite rims.

4.2. Petrography and Composition Characteristics of Chlorite

Chlorite is a ubiquitous and multi-generational alteration mineral at Jinshan gold deposit, appearing continuously from Stage 2 through Stage 4 and exhibiting systematic changes and textural changes throughout the paragenetic sequence, and a composition that tracks the evolving hydrothermal system (Figure 6 and Figure 7; Table 1 and Appendix A Table A1).

Figure 7. Variations in the major element composition of chlorite in samples from the Jinshan gold deposit; data derived from electron probe microanalysis during this study. (A) Al^VI versus Si (apfu). (B) Al^VI versus Mg + Fe (apfu). (C) Al^VI versus Al^IV (apfu). (D) Mg versus Fe (apfu). apfu—atoms per formula unit. (E) Al^IV versus Fe²⁺/(Fe²⁺ + Mg)(apfu)—atoms per formula unit. (F) Al^IV versus Mg/(Fe²⁺ + Mg)(apfu)—atoms per formula unit.

Table 1. Chemical composition of chlorite from the Jinshan gold deposit.

	Chl-1			Chl-2			Chl-3
	Max	Min	Average	Max	Min	Average	Max	Min	Average
SiO2	28.38	24.50	26.29	29.77	25.96	28.28	29.11	26.03	27.50
TiO2	0.24	0.00	0.02	0.06	0.00	0.01	0.06	0.00	0.02
Al2O3	23.10	17.11	20.66	19.45	14.32	17.24	19.37	16.44	18.04
FeO	35.39	27.07	30.14	31.99	24.41	26.85	33.30	26.97	29.76
MnO	0.66	0.02	0.24	0.31	0.14	0.20	0.55	0.21	0.34
MgO	14.43	6.57	10.87	16.46	10.16	14.32	15.38	10.28	12.47
CaO	0.26	0.00	0.03	0.19	0.00	0.08	0.12	0.00	0.05
Na2O	0.06	0.00	0.01	0.08	0.00	0.02	0.06	0.00	0.02
K2O	0.24	0.00	0.03	0.14	0.00	0.03	0.32	0.00	0.03
Total	89.72	87.02	88.32	89.04	85.03	87.02	89.78	85.61	88.23

In Stage 2, microscopic observation and backscattered electron (BSE) images reveal that Chl-1 appears as fine-grained (<50 μm) worm-like, feathery, or scaly aggregates. Key textural indicators include oriented infilling or replacement of earlier structures, such as occurrence within microfractures (Figure 6A), formation of thin rims around pyrite and rutile (Figure 5E and Figure 6A,B), or replacement of albite margins (Figure 6C). Chl-1 is stably associated with lath-shaped sericite, pyrite, and rutile, and shows no direct contact with main-stage sulfides such as galena, chalcopyrite, or sphalerite (Figure 5E and Figure 6A,B). In sharp contrast, the main ore Stage 3 Chl-2 consists of medium-grained (50–120 μm) flaky to tabular crystals, typically occurring as interstitial fillings in open cavities and quartz–sulfide interstices. It is associated with visible gold, sphalerite containing chalcopyrite disease, and galena (Figure 6D–F). Notably, Chl-2 is in direct contact with main-stage chalcopyrite (Figure 6D,F), providing direct spatial and temporal evidence for its formation during the peak period of gold precipitation. Stage 4 chlorite (Chl-3) is the coarsest (80–250 μm) and morphologically most diverse generation, occurring as fan-shaped rosettes, radial-fibrous bundles, and rare spherulites in clean, sulfide-free calcite–chlorite ± quartz veinlets or as pervasive replacement halos that corrode earlier sulfides (Figure 6G,I).

Seventy-seven electron probe microanalyses that passed rigorous quality filtering (Na₂O + K₂O + CaO < 0.5 wt%; totals 86–90 wt%) were recalculated on the basis of 14 oxygen atoms per formula unit. All compositions lie in the tri-trioctahedral field of Wiewióra A, Weiss Z (1990) [64] and define three tightly clustered, stage-specific populations (Figure 8). Stage 2 Chl-1 (n = 41) is Fe-rich chamosite with the highest Al^IV (1.02–1.30, avg 1.20 apfu) and Fe²⁺/ (Fe²⁺ + Mg) (0.45–0.74, avg 0.60). Stage 3 Chl-2 (n = 20) shifts sharply to a composition dominated by Mg-rich clinochlore with a minor component of Fe-rich chamosite, characterized by higher Si (2.92–3.12, avg 3.01 apfu) and markedly lower Al^IV (0.88–1.08, avg 0.99 apfu) and Fe²⁺/(Fe²⁺ + Mg) (0.45–0.62, avg 0.50). Stage 4 Chl-3 (n = 16) reverts toward Fe-rich chamosite with elevated Al^IV (0.93–1.17, avg 1.07 apfu) and Fe²⁺/(Fe²⁺ + Mg) (0.50–0.64, avg 0.57) relative to Chl-2 (Figure 7).

Figure 8. (A) Chlorite classification diagrams, modified after Wiewióra A, Weiss Z (1990) [64]. (B) Major element ternary diagrams showing variations in the Al(apfu)-Mg(apfu)-Fe(apfu) concentrations of chlorite from the Jinshan gold deposit, China, modified after Zane and Weiss (1998) [65]. apfu—atoms per formula unit.

Substitution trends are governed by coupled Tschermak [Al^IVAl^VI ↔ Si(Mg, Fe²⁺)] and Fe-Mg exchange (Figure 7). Strong negative correlations are observed between Al^IV and Si (R² = 1) and between Al^IV and (Fe²⁺ + Mg) (R² = 0.90), with moderate Fe²⁺–Mg anti-correlation (R² = 0.38). Al^IV correlates positively with Fe/(Fe²⁺ + Mg) (R² = 0.64) and negatively with Mg/(Fe²⁺ + Mg) (R² = 0.64). These systematic correlations define a clear evolutionary trend in chlorite composition from Stage 2 to Stage 4. The resulting compositional path—strongly Fe- and Al-rich (Stage 2) → Mg-rich and Al-poor (Stage 3) → intermediate Fe-rich again (Stage 4)—constitutes direct mineralogical evidence for the continuous hydrothermal evolution described above. The complete compositional cycle is not a series of isolated events but represents a direct response to continuous fluctuations in physicochemical conditions, such as temperature, oxygen fugacity (fO₂), and sulfidation state, within a single hydrothermal system. The compositional signature of Stage 4 marks the system’s entry into a late-stage phase characterized by renewed fluid–rock interaction following the main ore-forming period (Stage 3). Consequently, the evolutionary trajectory recorded by these multi-generational chlorite compositions serves as a highly sensitive indicator for deciphering dynamic changes in fluid properties during the continuous mineralization process of the Jinshan gold deposit (discussed in Section 5).

5. Discussion

5.1. Chlorite Formation Mechanisms and Compositional Trends

The textural and paragenetic habits of chlorite provide critical insights into its formation mechanisms and broader hydrothermal evolution [66,67,68,69]. In hydrothermal systems, chlorite typically forms via replacement (dissolution–recrystallization) or direct precipitation (dissolution–transport–precipitation), with the former dominant in low fluid/rock (W/R) ratios and the latter reflecting open-system transport [68]. At Jinshan gold deposit, Stage 2 chlorite (Chl-1) is unequivocally replacive, manifesting as vermicular aggregates that preserve relict textures from early albite and filling pyrite pores (Figure 6C), consistent with near-isochemical, low W/R metasomatism where Al, Si, and alkalis were locally sourced from dissolving carbonates and feldspars. This aligns with early-stage alteration in many orogenic gold systems, where initial fluid–rock buffering limits element mobility (e.g., [70]).

In contrast, Stages 3 and 4 chlorites (Chl-2 and Chl-3) exhibit euhedral to subhedral habits, infilling quartz vugs or forming discrete veinlets without inherited substrates (Figure 6D,H), indicative of precipitation from transported Fe-Mg-Al-Si complexes in a high W/R, fluid-dominated regime. This shift from replacement to precipitation between Stages 2 and 3 signals a transition from a relatively closed, wall-rock-buffered system to an open, advective one, driven by enhanced permeability along shear zones—a common hallmark of main-stage mineralization in shear-hosted orogenic deposits (e.g., [71]).

Compositional trends mirror this mechanistic pivot. In Figure 8A,B, chlorites evolve from Fe-rich chamosite (Chl-1) → Mg-rich clinochlore (Chl-2) → Fe-chamosite (Chl-3), a trajectory governed by Tschermak [Al^IVAl^VI ↔ Si(Mg,Fe²⁺)] and Fe-Mg exchange (Figure 7). The Stage 2–3 shift features sharply increased Si (2.80→3.01 apfu) and Mg# (0.40→0.50), with decreased Al^IV(1.20→0.99 apfu) and Fe²⁺ (2.58→2.29 apfu), reflecting dilution of wall-rock-derived Fe-Al signatures by more Mg-Si-enriched fluids (Figure 7A,F). Stage 4 reversion (Al^IV to 1.07 apfu; Mg# to 0.43) suggests renewed Fe-Al input, possibly from late-stage fluid–rock interaction. These patterns are not anomalous; similar Fe → Mg-Fe reversals occur in multistage orogenic systems like the Guocheng (China; [72]) and Haopinggou deposit [73] where they track fluid sourcing and redox fluctuations. At Jinshan gold deposit, the Stage 3 inflection—coincident with peak gold grades—underscores chlorite’s fidelity as a proxy for the physicochemical pivot enabling efficient metal transport and deposition.

5.2. Fluid Physicochemical Conditions

Chlorite geochemistry, particularly Al^IV and Fe/Mg ratios, serves as a robust archive for quantifying ore-fluid P-T-X evolution [3,74,75]. The observed trends at Jinshan delineate a coherent pathway of declining temperature, subtle fO₂ oxidation, and pH alkalinization from Stages 2–3, followed by partial reversal in Stage 4—conditions primed for Au(HS)₂⁻ destabilization and precipitation [76,77].

5.2.1. Formation Temperature

Tetrahedral Al^IV substitution in chlorite correlates positively with crystallization temperature via Tschermak exchange, modulating octahedral occupancy and basal spacing [74,78,79]. This study applies the empirical thermometers proposed by Cathelineau (1988) [74] and Kranidiotis and MacLean (1987) [79] for calculation, with specific formulas and detailed results provided in Appendix A Table A2. Applying these empirical geothermometers (uncertainty ± 25 °C) yields mutually consistent results: Stage 2 (Chl-1) at 269–351 °C (mean 321 °C); Stage 3 (Chl-2) at 239–289 °C (mean 262 °C); and Stage 4 (Chl-3) at 249–312 °C (mean 284 °C) (Figure 9A; Appendix A Table A2). This ~59 °C drop from Stage 2 to Stage 3 aligns precisely with the Al^IV decline (Figure 7A). This cooling trend is consistent with and corroborated by the reported decrease in fluid-inclusion temperatures [21,24,80], e.g., from ~275–310 °C to ~250–285 °C [21], as also seen in regional analogs (e.g., Shanggong; [81]). Furthermore, the mineralization temperature range (~200–350 °C) estimated by chlorite geothermometry strongly corresponds to the typical interval for orogenic gold deposits (e.g., the Wangfeng deposit; [82]). Recent thermodynamic refinements (e.g., incorporating non-ideal mixing; [83]) affirm the reliability of these calibrations in low-grade orogenic settings, where chlorite thermometry outperforms quartz veins by resolving < 50 °C pulses [84]. The post-peak rebound in Stage 4 may reflect late conductive reheating or fluid mixing, a pattern echoed in the Raja prospect [85].

Figure 9. Temperatures and oxygen fugacities of Stages 2, 3, and 4 in the Jinshan gold deposit. (A) Calculated based on the results of Cathelineau (1988) [74] and Kranidiotis and MacLean (1987) [79], using average values; (B) calculated based on the solid solution model proposed by Walshe (1986) [75].

5.2.2. Oxygen Fugacity

The mineral assemblages of Stage 2 (pyrite + arsenopyrite + pyrrhotite + minor native gold) and Stage 3 (pyrite + chalcopyrite + sphalerite + galena + native gold) qualitatively indicate a reducing fluid environment during mineralization (Figure 4 and Figure 5). This interpretation is quantitatively supported by geochemical data from alteration minerals. Specifically, the Fe³⁺/Fe²⁺ ratio in chlorite—which is directly controlled by fluid oxygen fugacity (fO₂)—governs sulfide stability and gold occurrence [3,75,86]. In this study, the oxygen fugacity of chlorite formation was calculated using the thermodynamic solid solution model of Walshe (1986) [75], combined with the equilibrium constant-temperature function fitted by Zhang et al. (2014) [87]. Detailed calculation procedures are provided in Li et al. (2017) [88] and Zhang et al. (2014) [87]. Calculated log fO₂ values show a distinct evolutionary trend (Figure 9B; Appendix A Table A2): Stage 2 ranges from −39.9 to −30.4 (mean −33.6), dropping significantly in Stage 3 (−45.0 to −36.2; mean −39.7), before increasing modestly in Stage 4 (−41.7 to −34.7; mean −37.7). These estimates align with prior data from the Jinshan gold deposit [24,80] and analogous arsenide-bearing systems like Bou Azzer [89].

The sharp decrease in fO₂ (~5.8 log units) from Stage 2 to Stage 3, accompanied by a marked temperature drop and Mg# increase, suggests the influx of a reduced, alkaline fluid derived from the devolatilization of graphitic metasediments [77]. This reduction event is independently corroborated by carbon isotope data (δ¹³C = −18‰ to −16‰) from the same stage [19]. Subsequently, the transition to Stage 4 is marked by a ~2 log unit increase in fO₂ and a decrease in Mg#. This oxidative shift cannot be attributed to simple cooling or wall–rock interaction (which typically lowers fO₂) but instead reflects the waning of the reduced fluid pulse and enhanced mixing with oxidizing meteoric water. This interpretation is supported by H–O isotopic compositions (δD = −62.9‰ to −43.9‰; δ¹⁸O = +2.2‰ to +11.4‰), which plot between metamorphic and meteoric water fields [45]. Similar fO₂ oscillations recorded by chlorite chemistry have been documented in the Guocheng [72] and Haopinggou [73] deposits.

5.2.3. pH Evolution

Octahedral Fe²⁺-Mg²⁺ exchange in chlorite is acutely sensitive to fluid a(Mg²⁺)/a(Fe²⁺) and pH, with experimental data indicating preferential Mg uptake under alkaline conditions [73,90]. The pronounced Mg# increase from Stage 2 (mean 0.40) to Stage 3 (mean 0.50; Figure 7F)—despite concurrent cooling—cannot be explained by temperature alone [79] and instead implicates alkalinization as the dominant driver, consistent with [73] calibration for orogenic fluids. This shift likely arose from fluid mixing with carbonate-bearing wall rocks or CO₂ degassing, elevating pH and destabilizing Au(HS)₂⁻ via reduced H⁺ activity [91]. The Stage 3–4 Mg# decline (to 0.43) signals renewed acidification, possibly from pyrite oxidation or influx of immature fluids [92]. Such pH oscillations are increasingly recognized as key triggers in shear-hosted gold systems (e.g., Wangu; [93]), where chlorite Mg# serves as a superior proxy to fluid inclusions for resolving sub-pH unit changes.

5.2.4. Sulfur Fugacity Evolution

Sulfur fugacity (ƒS₂) is a critical physicochemical parameter governing the migration and precipitation of metallic elements in hydrothermal deposits [94]). The sulfur fugacity of Stage 2 and Stage 3 was constrained through an integrated analysis of mineral paragenetic sequences, mineral-pair thermodynamics, and phase diagrams. Stage 2 is characterized by the paragenesis of pyrite, arsenopyrite, and minor pyrrhotite (Po). Utilizing the equilibrated pyrite–pyrrhotite (Py-Po) buffer pair, a logƒS₂ range of −12.6 to −9.0 was obtained for this stage, with a best estimate of −10.3. Stage 3 is marked by the appearance of pyrite, chalcopyrite, and sphalerite. By incorporating the pyrite–chalcopyrite–sphalerite (Py-Ccp-Sp) three-phase equilibrium system with measured temperatures (~239–289 °C), the logƒS₂ for this stage was constrained to −13.9 to −8.0 via phase diagram projection (Figure 10).

Figure 10. Temperature versus sulfur fugacity (log fS₂) diagram, adapted from Einaudi et al. (1982) [95]. The lines and shaded areas represent the ranges of sulfur fugacity for Chl-1 and Chl-2, respectively.

Key geochemical constraints indicate that, despite mineral assemblage evolution and a slight temperature decline from Stage 2 to Stage 3, sulfur fugacity remained stable within a narrow and overlapping interval (logƒS₂ ≈ −13.9 to −8.0), without undergoing an order of magnitude variation. Collectively, these constraints demonstrate that the mineralization at the Jinshan gold deposit occurred within a moderate- to low-temperature, reduced, and hydrodynamically stable fluid system.

5.3. Coupled Substitutions: Mineral-Scale Fluid Fingerprinting

The systematic compositional evolution of Jinshan chlorite results from the interplay of several well-established substitution mechanisms operating under changing T-fO₂-pH- fS₂ conditions, providing a detailed mineral-scale archive of fluid history that is more nuanced than can be resolved by bulk geochemical or inclusion studies [8,66,72,92].

The dominant controls are Tschermak (TK) and Fe-Mg (Al-Mg or “AM”) exchange. TK exchange [Al^IVAl^VI ↔ Si(R²⁺)^VI; R²⁺ = Mg + Fe²⁺] is primarily temperature-sensitive [74,92,96]. The ~59 °C cooling from Stage 2 to Stage 3 displaces the TK equilibrium to the right, markedly increasing Si (2.80 → 3.01 apfu) and decreasing tetrahedral Al^IV (1.20 → 0.99 apfu) while creating octahedral vacancies that are filled predominantly by Mg (Figure 7A,B). This TK-driven response to cooling is now widely documented in orogenic systems (e.g., Kofi deposit, Canada; [97]; Haopinggou, China; [73]).

Concurrently, the pronounced rise in Mg# from 0.40 to 0.50 (Figure 7F) is too large to be explained by temperature-controlled TK alone and requires a pH-driven Fe-Mg exchange [Fe²⁺ ↔ Mg²⁺] that strongly favors Mg under more alkaline conditions [72,73]. The resulting tight negative correlation between Al^IV and Mg# (R² = 0.64) is diagnostic of coupled T-pH control: cooling drives TK substitution, while simultaneous alkalinization drives AM substitution in the same direction, producing the characteristic Mg-enriched, Al-depleted chlorite that coincides with peak gold precipitation (cf. Guocheng and Wangu deposits; [72,93]).

Deviations from ideal TK behavior are evident in the Al^VI vs. Al^IV plot (Figure 7C), where most analyses plot above the 1:1 line expected for pure TK exchange [98]. This excess octahedral Al is best explained by di-trioctahedral (DT) substitution [3(R²⁺)^VI ↔ 2Al^VI + □], which introduces additional Al^VI without requiring a corresponding increase in Al^IV [10,96]. Minor contributions from Fe³⁺-bearing vectors (e.g., Si + 2Fe³⁺ ↔ Al^IV + 2R²⁺), permitted by the slight fO₂ increase from Stage 2 to Stage 3, further complicate Al partitioning [72]. The combined operation of TK, AM, DT, and minor Fe³⁺ substitutions produces the observed polyhedral distortion and compositional dispersion, mirroring the complex crystallochemical response documented in other greenschist facies orogenic systems (e.g., Pic de Port Vieux, Pyrenees; [99]; Central Alps; [11]).

Thus, the chlorite lattice at Jinshan functions as an integrated sensor: TK records the sharp cooling pulse, AM captures the accompanying pH increase, and DT + Fe³⁺ substitutions register subtle vacancy and oxidation effects. This multi-mechanism interplay explains why simple Al^IV-based thermometry slightly overestimates the temperature drop when applied in isolation and underscores why coupled TK-AM evolution is the most reliable mineral-scale fingerprint of the precise physicochemical conditions that triggered massive gold deposition in Stage 3.

5.4. Ore-Forming Fluid Evolution and Gold Precipitation Mechanism

The multi-generational chlorite record at Jinshan delineates a punctuated, rather than gradual, fluid evolution path. This mineralogical constraint directly addresses the long-standing debate regarding the dominant physicochemical triggers for gold deposition in the Jinshan gold deposit and, by extension, in similar shear-zone-hosted systems within the Jiangnan orogenic belt [24,25,100,101].

The Stage 2 fluids were characterized by relatively high temperatures (~320 °C), moderate reduced conditions (log fO₂ = −33.60), and mild acidity. These fluids operated under low fluid/rock (W/R) ratios, resulting in the formation of replacive, Fe-Al-rich Chl-1 and sparse early sulfides with only trace gold. A fundamental shift occurred at the transition to Stage 3, where permeability enhancement along the shear zone significantly increased W/R ratios. This structural reactivation shifted chlorite genesis from wall-rock replacement to open-space precipitation, accompanied by distinct changes in intensive parameters: a cooling of ~60 °C, an increase in pH, and a sharp decrease in oxygen fugacity (~6.0 log units).

Thermodynamic studies indicate that under the moderate temperatures (200–350 °C), relatively reducing conditions (log fO₂ = −45 to −31), near-neutral pH (5–8), and low salinities [24] observed at Jinshan, gold was primarily transported as Au(HS)₂⁻ complexes [76,102,103]. To quantify the precipitation drivers, thermodynamic modeling was conducted at 2 kbar and 5 wt% NaCl equiv. (Figure 11; see Methods). The simulation results demonstrate that fluid cooling and chemical desulfidation were the synergistic mechanisms destabilizing the gold-bisulfide complex. Modeling indicates that isolated cooling from ~320 °C to 260 °C reduces Au(HS)₂⁻ solubility by more than 1.5 orders of magnitude, accounting for a significant portion of the total gold precipitation (Figure 11A). Crucially, this cooling effect was compounded by a shift in fluid chemistry. Superimposed alkalinization drives the dissociation of aqueous hydrogen sulfide (H₂S(aq) + OH⁻ (aq) ⇌ HS⁻ (aq) + H₂O(l)), a process that our modeling indicates further destabilizes the gold-bisulfide complex. This chemical shift contributes an additional ~50% to the total gold precipitation event, bringing the cumulative efficiency to >90% (Figure 11B).

Figure 11. Thermodynamic modeling of gold precipitation mechanisms at the Jinshan gold deposit (calculated using Geochemist’s Workbench at 2 kbar, 5 wt% NaCl equiv., pyrite-buffered with log aH₂S = −3.5). (A) Effect of isolated cooling from ~320 °C to ~260 °C on Au(HS)₂⁻ solubility, accounting for the majority (~40%) of total gold precipitation. (B) Additional effect of superimposed modest alkalinization (ΔpH ≈ +2.0), further reducing solubility and contributing an additional ~50% to precipitation (cumulative > 90%) via enhanced dissociation of H₂S(aq) to HS⁻. (C) Effect of redox state on gold transport. The diagram shows gold solubility (mg/kg) versus oxygen fugacity (log fO₂). The trend indicates that Au(HS)₂⁻ complexes are destabilized by decreasing oxygen fugacity.

Furthermore, the dramatic decrease in oxygen fugacity (Δlog fO₂ = −6) played a pivotal role in the mineralization process. Contrary to models where redox shifts are negligible, our simulation reveals a strong positive correlation between fO₂ and gold solubility within the reduced field of the Jinshan system (Figure 11C). As the fluid evolved from the moderately reduced Stage 2 (log fO₂ = −33.6) to the strongly reduced Stage 3 (log fO₂ =39.7), the solubility of gold dropped precipitously by over one order of magnitude (from 8 × 10⁻⁴ to 2 × 10⁻⁵ mg/kg). This indicates that the intense reduction event—likely driven by fluid interaction with graphitic wall rocks—was not merely a side effect but a direct and potent driver of gold precipitation. Thus, the superimposition of rapid cooling, alkalinization, and intense reduction created a highly efficient trap, forcing the quantitative removal of gold from the fluid into the high-grade ore shoots.

Consequently, this hierarchy of controls decisively favors rapid cooling coupled with alkalinization and reduction as the primary triggers for high-grade mineralization. Fluid desulfidation (sulfidation of Fe-bearing phyllites) plays a supporting role, whereas phase separation (fault-valve boiling) appears to be subordinate or absent. The lack of characteristic boiling textures (e.g., adularia, lattice-bladed calcite) and abrupt shifts in stable isotopes (δ¹⁸O: +6.9‰ to +11.2‰; δD: −71‰ to −46‰; δ³⁴S: +3.3‰ to +4.6‰) is consistent with the mesozonal depths (~6–10 km) inferred from fluid inclusions [23,24]. The subsequent Stage 4 represents a partial reversal—warming to ~280 °C, renewed acidification, and slight reduction—reflecting a waning fluid flux re-equilibrating with wall rocks, which explains the sharp termination of high-grade mineralization and the barren nature of late calcite–chlorite veins.

The “cooling + alkalinization + reduction” mechanism identified at Jinshan aligns with an emerging consensus for greenschist facies orogenic gold systems worldwide, particularly those hosted in graphite- or carbonate-bearing sequences (e.g., Bou Azzer, Morocco [89]; Guocheng [72]; and Wangu [93]). At the Jinshan gold deposit, the pH increase during Stage 3 was likely driven by the reaction of ascending metamorphic fluids with carbonate-rich lenses or CO₂ effervescence during fracture dilation, while the rapid cooling and reduction reflect adiabatic ascent and interaction with graphitic schists along the reactivated structure.

In summary, the chlorite-based reconstruction provides the first mineralogically calibrated, quantitative resolution of the ore-fluid pathway at the Jinshan gold deposit. It demonstrates that the deposit’s significant endowment (>200 t Au) was precipitated within a narrow physicochemical window defined by a ~50–60 °C temperature drop, rapid alkalinization, and intense reduction. This model reconciles previously conflicting interpretations (e.g., boiling vs. mixing vs. sulfidation) into a coherent framework, establishing a predictive template for exploration in the Jiangnan orogenic belt and analogous Neoproterozoic–Paleozoic collisional orogens.

6. Conclusions

(1)

The Jinshan gold deposit is a fault-controlled, multistage orogenic system with four consecutive paragenetic stages: Stage 1 (ankerite–quartz), Stage 2 (pyrite–arsenopyrite–quartz), Stage 3 (quartz–gold–polymetallic sulfide), and Stage 4 (chlorite–carbonate–quartz).

(2)

Chlorite occurs from Stage 2 onward and is formed by two mechanisms: metasomatic replacement in Stage 2 (Chl-1) under low fluid/rock ratios, and direct precipitation in Stages 3 (Chl-2) and 4 (Chl-3) under higher fluid/rock ratios.

(3)

Chlorite compositions record a key Stage 2–3 transition: cooling from ~320 °C to ~260 °C, pH increase (alkalinization), and reduction (mean log fO₂ decreases from ~−33.6 to~−39.7). Sulfur fugacity remained stable within a narrow range (log fS₂ = −13.6 to −8.0) throughout Stages 2–3. Stage 4 shows minor reheating (~280 °C), re-acidification, and a slight rebound in oxygen fugacity (log fO₂ ≈ −37.7), yet still maintains a reducing environment.

(4)

Chlorite chemistry reflects coupled Tschermak and Fe–Mg substitutions driven mainly by cooling and rising pH during the Stage 2–3 shift.

(5)

Thermodynamic simulations show that gold precipitation in Stage 3 was the result of continuous fluid evolution, with its triggering mechanisms attributed to rapid cooling (a temperature drop of ~50–60 °C), an increase in pH, and an intensely reducing environment.