Fluid Evolution and Ore Genesis of the Juyuan Tungsten Deposit, Beishan, NW China

Abstract

The newly discovered Juyuan tungsten deposit is hosted in Triassic granite in the Beishan Orogen, NW China. The tungsten mineralization occurred as quartz veins, and the main ore minerals included wolframite and scheelite. The age, origin, and tectonic setting of the Juyuan tungsten deposit, however, remain poorly understood. According to the mineralogical assemblages and crosscutting relationships, three hydrothermal stages can be identified, i.e., the early stage of quartz veins with scheelite and wolframite, the intermediate stage of quartz veinlets with sulfides, and the late stage of carbonate-quartz veinlets with tungsten being mainly introduced in the early stage. Quartz formed in the two earlier stages contained four compositional types of fluid inclusions, i.e., pure CO₂, CO₂-H₂O, daughter mineral-bearing, and NaCl-H₂O, but the late-stage quartz only contained the NaCl-H₂O inclusions. The inclusions in quartz formed in the early, intermediate, and late stages had total homogenization temperatures of 230–344 °C, 241−295 °C, and 184−234 °C, respectively, with salinities no higher than 7.2 wt.% NaCl equiv (equivalent). Trapping pressures estimated from the CO₂-H₂O inclusions were 33−256 MPa and 36−214 MPa in the early and intermediate stages, corresponding to mineralization depths of 3–8 km. Fluid boiling and mixing caused rapid precipitation of wolframite, scheelite, and sulfides. Through boiling and inflow of meteoric water, the ore-forming fluid system evolved from CO₂-rich to CO₂-poor in composition and from magmatic to meteoric, as indicated by decreasing δ¹⁸O_water values from early to late stages. The sulfur and lead isotope compositions in the intermediate-stage suggest that the Triassic granite was a significant source of ore metals. The biotite ⁴⁰Ar/³⁹Ar age from the W-bearing quartz shows that the Juyuan tungsten system was formed at 240.0 ± 1.0 Ma, coeval with the emplacement of granitic rocks at the deposit. Integrating the data obtained from the studies including regional geology, ore geology, biotite Ar-Ar geochronology, fluid inclusion, and C-H-O-S-Pb isotope geochemistry, we conclude that the Juyuan tungsten deposit was a quartz-vein type system that originated from the emplacement of the granites, which was induced by collision between the Tarim and Kazakhstan–Ili plates. A comparison of the characteristics of tungsten mineralization in East Tianshan and Beishan suggests that the Triassic tungsten metallogenic belt in East Tianshan extends to the Beishan orogenic belt and that the west of the orogenic belt also has potential for the discovery of further quartz-vein-type tungsten deposits.

1. Introduction

Tungsten deposits in China are widely distributed in the South China block, the Himalaya, Tibetan, Sanjiang, Kunlun, Qilian, Qinling, Dabie, and Sulu orogens, as well as the Central Asian orogenic belt. The tungsten resource occurs mainly as skarn (63%), quartz-vein (17%), porphyry (17%), and greisen (3%) deposits in China [1]. The quartz-vein tungsten deposits are concentrated mostly in the South China block and are associated with Jurassic granitic magmas [2,3,4,5,6,7,8,9]. However, in recent years, a cluster of tungsten deposits associated with Triassic intrusions was discovered in the East Tianshan-Beishan belt of the Central Asian orogenic belt (CAOB) [10,11]. These tungsten deposits, in which the orebodies are mainly quartz-vein and skarn type, are currently under exploration. Most previous studies of these tungsten deposits have focused on geological characteristics, metallogenic granite chronology and geochemistry, metallogenic chronology, and metal source [12,13,14,15,16,17,18,19,20,21]. These studies have confirmed the existence of a Triassic tungsten metallogenic belt in East Tianshan-Beishan, China.

The Juyuan W deposit, which is situated in the western part of the Shuangyingshan–Huaniushan arc, is a quartz-vein tungsten deposit and was discovered in 2018 by the Xinjiang Institute of Geological Survey. The ore bodies occur mainly in the form of quartz veins and are hosted in Triassic monzonitic granite and two-mica granite. Previous studies of the deposit have focused mainly on the ages of the host granite, the tectonic setting of the granite inferred from geochemical data, and the trends of quartz veins to determine the principal stress direction applied to the deposit [22,23]. However, few studies have focused on the origin and evolution of ore-forming fluids of the deposit, and there is a lack of systematic research on fluid inclusions (FIs).

In this paper, we report a systematic study of the Juyuan tungsten deposit based on FIs in quartz C-H-O isotopes from hydrothermal minerals and S-Pb isotopes of sulfide minerals to investigate the nature of ore-forming fluids and deposited metals, the evolution of the fluids, and the mechanism of wolframite and scheelite deposition. In addition, we used Ar-Ar dating of muscovite to establish the formation age of the deposit.

2. Regional Geology

The CAOB is one of the largest accretionary orogens in the world and was formed by multiple subduction-accretion and collisional events associated with the closure of the Paleo-Asian Ocean [24,25,26]. Tectonically, the CAOB is bounded by the Siberian and Eastern European cratons to the north and the Tarim and North China cratons to the south [27,28,29].

As a significant part of the southern CAOB, the Beishan orogenic belt is located between the Southern Tian Shan suture to the west and the Solonker suture to the east and is tectonically regarded as the eastern extension of the Chinese Tian Shan [27,30]. The Southern Tian Shan suture is on the west of the Beishan orogenic belt, and the Solonker suture is on the east of Beishan orogenic belt. They are bordered but the boundary is not clear. Geographically, Beishan is a mountainous area in western Gansu Province and to the east of Xinjiang Uygur Autonomous Region.

The Beishan orogenic belt comprises Precambrian crystalline basement and overlying sedimentary rocks (Figure 1). The ages of the constituent rocks range mainly from late Precambrian to Mesozoic, including Precambrian metasedimentary rocks and Paleozoic arcs, ophiolitic mélanges, accretionary complexes, and arc-related basin sediments [31,32,33].

The E-W-trending tectonic units of the Beishan orogenic belt are separated by E-W-trending strike-slip faults and ophiolitic mélanges and are cut by later NE-SW-trending strike-slip faults with complicated shear senses [27,28,33].

The orogenic evolution of the Beishan belt can be divided roughly into pre-orogenic (Precambrian–early Cambrian), ocean–continental (Ordovician–Permian), and post-orogenic (Triassic) evolutionary periods. Based on the geological, geochemical, tectonic, and geophysical data, the Beishan orogenic collage can be subdivided into several tectonic units, which from north to south are the Queershan arc, Hongshishan ophiolitic mélange, Heiyingshan-Hanshan arc, Shibanjing-Xiaohuangshan ophiolitic mélange, Mazongshan arc, Hongliuhe-Niujuanzi-Xichangjing ophiolitic mélange, Shuangyingshan-Huaniushan arc, Liuyuan ophiolitic mélanges, and Shibanshan arc. The Heiyingshan-Hanshan arc and the Mazongshan and Shuangyingshan-Huaniushan arcs constitute the Gongpoquan arc-accretionary system [18,27,28,33]. These terranes consolidated and evolved at various times from the Cambrian to the Permian and were affected by multiple periods of magmatism, resulting in large numbers of granitic intrusions and mafic volcanic rocks, accompanied by lesser numbers of mafic–ultramafic intrusions.

Two episodes of magmatism related to coeval tungsten mineralization have been recognized in the Beishan orogenic belt: (1) 424–314 Ma (e.g., Hongjianbingshan, Guoqing, Yingzuihongshan, and Baixianishan deposits); the related intrusions were derived from the partial melting of Precambrian metasedimentary rock or juvenile lower crust [12,15,16,21,34,35]; (2) 286–220 Ma (e.g., Xiaobaishitou, Liushashan and Yushan deposit); the granites were derived from the ancient crustal sediments [17,21,36,37,38,39,40]. These two magmatic events are mainly distributed in the terranes with Precambrian metasedimentary strata, Shuangyingshan-Huaniushan arc, and Hanshan block.

3. Deposit Geology

The representative strata in the Juyuan tungsten ore district comprise mainly the Pingtoushan Formation of the Jixian System and the Kuqa Formation of the Tertiary System (Figure 2). The Pingtoushan Formation is composed of fragmented grayish marble with well-developed schistosity and strong weathering. The Tertiary Kuqa Formation is mainly reddish argillaceous clay that weathers to reddish-brown loose soil.

The structure of the ore district is relatively simple (Figure 3). There are no obvious faults, and quartz-vein dislocation can be seen locally, with displacements of no more than 1 m. Intrusive rocks are common in the ore district and include monzogranite, granite, and numerous dikes. The shape of monzogranite and granite is plutonic rock, and the dikes are mainly diabase dikes, with a small amount of quartz porphyry. The main intrusive phase is represented by Triassic monzogranite (Figure 4a), which is located in the central and northeastern parts of the ore district. The zircon U-Pb age of the monzogranite is 240.7 ± 0.5 Ma (unpublished data). The monzogranite is red in color, medium grained, and composed of quartz, with equal proportions of K-feldspar and plagioclase, and biotite, with accessory zircon, magnetite, apatite, and other minerals. The monzogranite shows argillitization and sericitization. Granite is distributed mainly in the north of the ore district and has a medium-grained granitic texture and massive structure. The surface of the granite is mostly weathered into granules, and ferritization and kaolinization are locally developed. Quartz porphyry, located in the central part of the ore deposit and running through the monzogranite (Figure 4a), is porphyritic and composed of quartz and K-feldspar as phenocrysts in a groundmass of quartz, K-feldspar, and plagioclase. Phenocrysts of K-feldspar have undergone intense argillitization, and the fine groundmass displays sericitization. Diabase and diorite dikes are also developed in the ore district. Diabase dikes are mostly 100–500 m in length and show weak chloritization. Diorite dikes are widely distributed, with lengths of 80–300 m, cutting through diabase dikes locally. The main mineral composition of diabase are pyroxene and basic plagioclase, as well as a small amount of biotite, apatite, magnetite, etc. The mineral composition of diorite is mainly amphibole and plagioclase, and the accessory minerals are mainly apatite, magnetite, and sphene.

At present, 22 ore bodies can be identified in the ore district (Figure 2). The tungsten ore body is 80–420 m long and 0.4–2.8 m wide, with a strike of 124°–152° and an overall south dip of 65°–80° (Figure 3). The tungsten grade 0.065–0.14%. The orebodies occur mainly in quartz veins in monzogranite. Wolframite and scheelite are the major ore minerals of the Juyuan tungsten deposit (Figure 4e,f,h,i), with a small amount of galena, pyrite, and other sulfides. Wolframite is mostly euhedral, with grain sizes ranging from 1 mm to 15 mm. Scheelite occurs in two forms: (1) euhedral, with a grain size smaller than that of wolframite, up to 5 mm (Figure 4i), and (2) distributed in fissures, with a shape structure, which commonly cuts quartz and wolframite. The ore structure is star shaped, vein shaped, or sparsely disseminated.

According to the occurrence and cross-cutting relationships of quartz veins, combined with the characteristics of microstructure and mineral assemblages, the quartz veins in the ore district can be divided into three stages. Early-stage quartz veins are ore veins and are generally gray white in color. Veins are 0.5–2 m thick, and the metallic minerals in the veins appear mainly as wolframates (e.g., wolframite and scheelite). Intermediate-stage quartz veins are smokey gray or gray white in color, and sulfides (e.g., galena and pyrite) (Figure 4l) are more abundant in the veins compared with the early stage. The quartz veins of the intermediate stage cut through the early stage of mineralized quartz veins. Late-stage quartz veins (post-metallogenic quartz veins) are white and 0.1–0.2 m thin. Calcite is present in the quartz veins. These late veins cross-cut the smokey gray quartz veins of the intermediate ore-bearing hydrothermal stage and no obvious metallic minerals are observed in them.

4. Samples and Analytical Methods

4.1. Fluid Inclusion

We undertook a microthermometric FI study of quartz crystals associated with the three different stages of mineralization in quartz veins. More than 150 double-polished thin-sections (200 μm thick) were prepared for FI analysis. Fluid inclusions were carefully observed to identify their genetic types, vapor-liquid ratios, and spatial clustering. Nineteen typical samples were selected for microthermometric measurements and laser Raman spectroscopy analysis.

Microthermometric measurements were conducted at the Institute of Geomechanics, Chinese Academy of Geological Sciences (CAGS), Beijing, China, using a Linkam THMSG 600 heating-freezing stage with a temperature range of −195 °C to +600 °C, attached to a Leitz Ortholux transmitted-light microscope connected to a television camera and screen. The stage was calibrated using synthetic FIs made by Fluid Inc., Oakland, CA, USA. The precision of the measured temperature was ±0.1 °C in the range of −195 °C to 30 °C and ±1 °C in the range of 30 °C to 600 °C. The heating rate was kept at 0.2 °C to 5.0 °C per minute and when approaching the phase-change conditions, the heating rate was changed into 0.2 °C per minute. To avoid inclusion decrepitation, the freezing processes were operated first. The salinities for aqueous fluids were calculated using the equations of [41] and for aqueous-carbonic fluids using the equations of [42,43]. According to the microthermometic data [44,45], the CO₂-H₂O-NaCl system was defined, and the densities and pressures were determined using the Flincor software package.

Laser Raman spectroscopic analysis of single fluids including vapor and liquid phases was done at the Raman Laboratory of the Beijing Institute of Uranium Geology, Beijing, using a LABHR-VIS LabRAM HR800 Laser Raman microspectrometer, (Paris, France). The analysis used an argon atom laser with a wavelength of 532 nm. The spectrum range was set between 100 cm⁻¹ and 4200 cm⁻¹. The accumulation time for each scan was 5 s, and the spectrum resolution was 2 cm⁻¹.

4.2. Isotope Analysis

The samples for isotope analyses were obtained from different mineralization stages, confirmed by microscopy observation. The minerals were separated by handpicking under a binocular microscope to achieve >99% purity. Analyses of isotopes of O, H, C, and S were performed on a Finnigan MAT253 mass spectrometer, (Waltham, MA, USA) at the Analytical Testing Center of the Beijing Research Institute of Uranium Geology, using the methods outlined by [46].

For O isotope analyses of quartz, the quartz was ground to 200 mesh and then dried. Oxygen gas was liberated by reaction with BrF₅ [47] in externally heated nickel vessels and converted to CO₂ on a carbon rod. Isotope ratios (δ¹⁸O) were determined, with all values being reported relative to the V-SMOW standard, with an uncertainty of ±0.2‰.

Hydrogen isotope analyses were conducted on the same samples as investigated for fluid inclusions hosted in the quartz. The selected quartz samples were dried at a low temperature to remove the adsorbed water on the surface of the mineral, and secondary inclusions were removed by gradual heating under vacuum. The fluid inclusions decrepitated when the temperature was increased to 600 °C. Hydrogen in the liberated water was replaced by zinc and released for mass spectrometry analysis. The values are reported relative to the V-SMOW standard, and the uncertainty was better than ±1‰.

The method of analysis of C isotopes was similar to that of H isotopes. Quartz samples were dried and placed in a vacuum system to remove the gas generated by the burst of secondary inclusions during the process of heating. After heating to 600 ℃, the fluid inclusions were burst, and CO₂ was liberated. CO₂ was separated by combined cold traps and then sent for mass spectrometry analysis. We reported the C isotope data per mil relative to the Peedee Belemnite (PDB) standard, with an uncertainty of ±0.2‰.

Four galena and four pyrite samples were analyzed for S and Pb isotopic compositions at the Analytical Testing Center of the Beijing Research Institute of Uranium Geology.

For S isotope analysis, each sample was weighed to 15 mg, mixed with about 150 mg of CuO powder, ground to 200 mesh, and placed in a vacuum quartz tube. The powder was heated to 980 °C when the vacuum reached 2.0 × 10⁻² Pa, and then SO₂ was produced. After purification, SO₂ was transferred to the sample tube, and the ³⁴S/³²S ratio was measured with a Finnigan MAT 251 mass spectrometer. The isotope data were reported relative to the Canyon Diablo Troilite (CDT) standard, and the uncertainty of δ³⁴S values is ±0.2‰.

The O isotope ratios of water (δ¹⁸O_W) were calculated using oxygen-isotope fractionation equilibrium formulas reported by [47].

4.3. Muscovite ⁴⁰Ar/³⁹Ar Dating

The muscovite grains were hand-picked carefully under a binocular microscope, followed by soaking in 5% dilute nitric acid. The separates were then washed repeatedly in the ultrasonic bath using acetone and deionized water for 15 min and placed in an oven to dry at 60 °C. Aliquots of ~10.8 mg were wrapped in aluminum foil and sealed in a quartz tube along with several neutron fluence monitors (ZBH-25 biotite) for calculation of the fluence gradient. The quartz tube was then vacuum-sealed and irradiated for 24 h at the HFETR facility of the Northwest Institute of Nuclear Technology, Beijing, China. Step-heating ⁴⁰Ar/³⁹Ar analyses were conducted using an Argus VI noble gas mass spectrometer in the Argon Laboratory of the Beijing Research Institute of Uranium Geology. The irradiated separates were incrementally heated by a double-vacuum tantulum furnace from 800 °C to total fusion at 1300 °C at a rate of 40 °C to 200 °C every 10 min in eight temperature steps to 1300 °C. Then, the gas released from the sample was purified by a stainless U-type trap with liquid nitrogen, and by two Zr-Al getters, with one at 450 ℃ and the other at room temperature, leaving only the noble gases, which entered the noble gas mass spectrometer for Ar isotope analysis. The interference correction values of the reactor, which were derived from irradiated CaF₂ and K₂SO₄, are (³⁶Ar/³⁷Ar)_Ca = 0.0003011, (³⁹Ar/³⁷Ar)_Ca = 0.0007694, and (⁴⁰Ar/³⁹Ar)_K = 0.0020826. J values for each irradiation position were calculated from the ⁴⁰Ar/³⁹Ar ratios by total fusion of the ZBH-25 biotite, using an assumed age of 132.9 ± 1.3 Ma [48]. The mass discrimination factor was calculated according to the atmospheric Ar ratio of 295.5 ± 0.5 [49], and the decay constant for ⁴⁰K used in the calculation was 5.543 × 10⁻¹⁰ year⁻¹ [50]. The data-processing software used was ArArCALC 2.4 [51].

5. Fluid Inclusion Geochemistry

5.1. Fluid Inclusion Petrography

Through phase relationships, relative phases present at room temperature (21 °C), phase changes during freezing processes, and laser Raman spectroscopic measurements [52], four types of FI were identified (Figure 5), namely, CO₂-H₂O or carbonic-aqueous (C-type), pure carbonic or CO₂ (PC-type), aqueous (W-type), and fluid inclusion with daughter minerals (S-type).

Petrographic criteria suggested by [52] were used to distinguish primary and secondary FIs. Fluid inclusions occurring in growth zones or as isolated clusters in crystal cores are regarded as primary.

C-type were present in quartz from early- and intermediate stages. They usually occurred as two or three phases at room temperature (liquid CO₂ + liquid H₂O or vapor CO₂ + liquid CO₂ + liquid H₂O); Figure 5a–e), with the CO₂ phase volumetrically occupying typically 10–95 vol.%. When the temperature dropped to ≤10 °C, the vapor CO₂ was observed in two-phase. The C-type FIs were commonly isolated individuals or appeared as clusters of randomly arrangement inclusions. They were variable in shape, including subrounded shapes and polygonal forms, with sizes ranging from 4 μm to 27 μm. According to the vapor/liquid ratio by volume, C-type FIs can be further divided into liquid-rich fluids (vapor/liquid < 50 vol.%; Figure 5a) and the vapor-rich fluids(vapor/liquid > 50 vol.%; Figure 5a). The liquid-rich fluids were more common and accounted for more than 90% of the total fluid (Figure 5).

PC-type FIs were rare and showed generally oval to subrounded shapes with a uniform grain size about 6–7 um. They appeared as monophase (liquid CO₂) at room temperature (Figure 5d). During cooling, the monophase inclusions developed a CO₂-rich vapor bubble.

W-type FIs occurred in all three stages, and at room temperature they usually appeared as two-phase (liquid H₂O + vapor H₂O) fluid inclusions (Figure 5d–g), with a vapor/liquid ratio of about 10–45 vol.%. Morphologically, the fluid inclusions (5–13 μm) showed subrounded, oval, or irregular forms. A small amount of CO₂ was identified in the vapor bubbles by laser Raman spectroscopy, although the CO₂ phase was not observed during heating or cooling runs. The primary W-type FIs occurred as isolated individuals or as clusters, but the secondary W-type FIs often occurred along healed fractures, showing irregular forms.

S-type FIs were scarce and occurred only in a few cases in early-stage quartz samples, appearing as three-phase inclusions (liquid CO₂ + liquid H₂O + rectangular-shaped daughter minerals; Figure 5b). The content of daughter minerals in the S-type FIs was low and unknown. The low-temperature phase transformation behavior of the inclusions was similar to that of W-type FIs. The daughter minerals did not melt at the bursting temperature, indicating that daughter minerals may have been captured accidentally during the formation of S-type FIs.

5.2. Microthermometry

5.2.1. Early-Stage Fluid Inclusions

The early-stage quartz included all four types of FIs identified in the Juyuan tungsten deposit (

Table 1. Microthermometric data for fluid inclusions in quartz from the Juyuan tungsten deposit.

Stage	Sample No.	Type	Number	Size (μm)	Vapor (vol.%)	Tm-CO2 (°C)	Tm-cl (°C)	Th-CO2 (°C)	Tm-ice (°C)	Th-tot (°C)	Salinity (wt.% NaCl equiv.)	CO2 Density (g/cm3)	Bulk Density (g/cm3)
Early	D1936-1	C	13	5–9	25–85	−60.9 to −68.8	5.4–8.3	26.9–31.1 (L/V)		261–343 (L/V)	3.3–8.4	0.54–0.69	0.69–0.92
		PC	2	5–7	20–30	−62.5 to −63.2	5.4–8.3	28.3–29.1 (L)				0.63–0.65
	D1938-3a	C	12	6–27	15–50	−58.2 to −69.6	6.2–8.7	27.0–31.1 (L)		266–327 (L)	2.6–7.1	0.59–0.69	0.83–0.95
		W	2	6–9	20–30				−2.8 to −5.8	275–282 (L)	4.7–9.0
		S	1	15	15				−6.8	246 (L)	10.2
	D1958-2-2	C	8	4–12	20–75	−59.0 to −67.1	4.4–8.6	17.2–29.6 (L/V)		253–318 (L/V)	2.8–9.9	0.65–0.81	0.79–0.97
		PC	1	6	40	−60.2		26.5 (L)				0.69
	D1958-4a	C	2	5–7	15–20	−61.0 to −61.9	6.0–6.7	28.0–30.6 (L)		249–321 (L)	6.2–7.4	0.62–0.68	0.95–0.99
	D1965-1	C	17	6–11	25–95	−58.2 to −66.7	6.0–8.7	28.8–31.1 (L/V)		249–333 (L/V)	2.6–7.4	0.52–0.64	0.69–0.92
	D1966-1	C	11	5–10	15–45	−60.5 to −67.0	7.2–10.0	27.2–30.7 (L)		265–336 (L)	0.02–5.3	0.62–0.70	0.83–0.97
		W	1	7	15				−5.5	280 (L)	8.6
	D1967-1	C	19	4–9	15–90	−59.8 to −63.9	5.5–8.5	26.1–30.6 (L/V)		240–340 (L/V)	3.0–8.2	0.62–0.71	0.71–0.96
		PC	2	5–7	25–30	−61.9 to −62.6		15.9–25.7 (L)				0.70–0.81
	D2016-1	C	8	5–9	30–40	−56.7 to −62.7	6.1–9.7	24.5–30.4 (L)		248–290 (L)	0.6–7.2	0.62–0.74	0.86–0.95
	D2017-1	C	5	5–7	30–35	−57.4 to −64.3	6.9–8.8	29.7–30.8 (L)		260–324 (L)	2.4–5.9	0.61–0.67	0.88–0.92
		W	1	5	30				−6.5	232 (L)	9.9
	D2022-1	C	6	5–6	30–40	−56.7 to −59.7	6.9–9.2	27.7–29.8 (L)		241–340 (L)	1.6–5.9	0.64–0.69	0.87–0.91
		W	2	5–6	35–40				−3.1 to −7.4	230–278 (L)	5.1–11
	D2032-1	C	7	5–7	25–40	−58.0 to −64.7	6.8–9.7	29.5–31.0 (L)		257–338 (L)	0.6–6.0	0.60–0.67	0.86–0.92
	D2033-1	C	32	5–11	20–40	−56.9 to −67.3	4.0–9.3	25.8–31.0 (L)		230–309 (L)	1.4–10.5	0.60–0.72	0.84–0.96
		W	1	10	15				−5.8	264 (L)	9.0
	D2045-1	C	10	5–11	30–45	−57.1 to −63.1	8.6–10.0	28.9–30.8 (L)		275–334 (L)	0.02–2.77	0.61–0.66	0.83–0.90
	D2048-1	C	8	5–8	30–45	−57.2 to −65.4	6.7–9.3	25.7–30.9 (L)		257–344 (L)	1.4–6.2	0.61–0.65	0.85–0.90
Intermediate	D1935-1	C	8	5–13	15–40	−58.8 to −68.7	7.4–8.5	30.0–31.1 (L)		270–290 (L)	3.0–5.0	0.61–0.64	0.86–0.96
	D1938-1b	C	11	5–17	20–70	−60.2 to −75.5	6.8–8.7	29.3–31.1 (L/V)		241–286 (L/V)	1.8–6.0	0.60–0.65	0.76–0.92
		PC	2	8–12	25–45	−62.0 to −76.3		24.7–30.0 (L)				0.60–0.72
	D2070-1	C	6	6–7	20–40	−57.2 to −62.1	6.3–8.4	28.7–30.9 (L)		275–295 (L)	3.2–6.9	0.61–0.67	0.89–0.95
Late	D1936-4	W	13	5–13	10–25				−3.2 to −8.5	184–234 (L)	4.6–12.3
	D1959-2a	W	1	6	10				−8.1	196 (L)	11.8

1 Fluid inclusion type: C = CO₂-rich inclusions; W = liquid-rich two-phase inclusions; PC = pure CO₂ inclusions; S = liquid-rich three-phase inclusions with one daughter mineral; 2 T_m-CO2—Final solid CO₂ melting temperature. 3 T_m-cl—Final CO₂ clathrate dissociation temperature. 4 T_h-CO2—CO₂ phase homogenization temperature. 5 T_h-tot—Total homogenization temperature. 6 T_m-ice—Final ice melting temperature.

), with C-type FIs being the most common. The melting temperatures of solid CO₂ (T_m-CO2) were between −69.6 °C and −56.7 °C, below the triple point of pure CO₂ (−56.6 °C), implying the presence of minor amounts of dissolved components in the carbonic phase [52,53]. The melting temperatures of clathrates (T_m-cl) ranged from 4.0 °C to 10.0 °C. The carbonic phase (T_h-CO2) in liquid-rich FIs was partially homogenized to liquid at 15.9 °C to 31.1 °C, and the vapor-rich FIs to vapor at 17.2 °C to 31.1 °C. The C-type FIs were totally homogenized to liquid or vapor at temperatures ranging from 230 °C to 344 °C (T_h), and mostly in the range of 280–300 °C (Figure 6), although many vapor-rich inclusions decrepitated prior to final homogenization at about 300 °C. The CO₂ densities calculated using the formula of [44] ranged from 0.52 g/cm³ to 0.81 g/cm³ and bulk densities from 0.61 g/cm³ to 0.99 g/cm³. Raman analysis indicated that CO₂ and H₂O were the principal species, with small amounts of CH₄ (Figure 7a).

The PC-type FIs displayed two phases and were frozen to solid below −100 °C. The solid part melted at −60.2 °C to −63.2 °C. The T_h-CO2 values of CO₂ (L + V to L) were between 15.9 °C and 29.1 °C, with corresponding densities of 0.63 g/cm³ to 0.81 g/cm³.

The W-type FIs in early-stage quartz yielded ice-melting temperatures, measured at −2.8 °C to −7.4 °C, with salinities ranging from 4.7 to 11.0 wt.% NaCl equiv. They were homogenized to the liquid phase at 230 °C to 282 °C (Figure 6).

An S-type FI with an unidentified solid was observed, with an ice-melting temperature of −6.8 °C, a salinity of 10.2 wt.% NaCl equiv., and homogenization to the liquid phase at a temperature of 246 °C. The daughter mineral did not dissolve during heating to 500 °C, so its homogenization temperature was not observed.

As described above, coexisting C-type inclusions with different vapor/liquid ratios in the early-stage quartz had similar homogenization temperatures but commonly had different salinities, implying that fluid immiscibility occurred during the early stage of mineralization of the Juyuan deposit or that the FIs were trapped at the same time and thus represent an FIA.

5.2.2. Intermediate-Stage Fluid Inclusions

The intermediate-stage quartz, which was closely associated with polymetallic sulfide, contained many FIs. The FIs in this stage quartz were dominantly C-type with lesser PC-type. The T_m-CO2 values of C-type FIs were measured at −75.5 °C to −57.2 °C, below the triple point of pure CO₂ (−56.6 °C). The CO₂ clathrate (T_m-cl) melted at a temperature of 6.3 °C to 8.7 °C, and the corresponding salinities reached 6.9 wt.% NaCl equiv. The CO₂ phases were totally homogenized to liquid at 30.0–31.1 °C (T_h-CO2). The CO₂ densities were calculated at 0.60 g/cm³ to 0.67 g/cm³ and bulk densities at 0.76 g/cm³ to 0.96 g/cm³. They were generally homogenized to liquid or vapor at 241–295 °C, peaking at 260–280 °C.

PC-type FIs were rarely observed in quartz of the intermediate stage. They yielded T_m-CO2 values of −76.3 to −62.0 °C and homogenized to liquid at 24.7 °C to 30.0 °C (T_h-CO2). Raman analysis shows that the component of these FIs was dominated by CO₂ and H₂O (Figure 7c).

5.2.3. Late-Stage Fluid Inclusions

Only W-type FIs were observed in late-stage quartz. Their T_m-ice values were observed at −3.2 °C to −8.5 °C (Table 1), and corresponding salinities at 4.6 to 12.3 wt.% NaCl equiv. They were totally homogenized to liquid at 184–234 °C (Figure 6). The marked decrease in the abundance of C-type FIs from the intermediate to late stage indicates that the fluid system transformed from a relatively closed state to a relatively open state, which enabled a large amount of CO₂ gas to escape and allowed a large volume of meteoric water to enter and mix with the ore-forming fluid.

5.3. Trapping Pressure and Mineralization Depth

Measurements of the partial and total homogenization temperatures and the ratios of vapor CO₂ and liquid CO₂ of C-type FIs allowed the minimum trapping pressures of FIs to be estimated using the Flincor program and the formula of [44,45] for the H₂O-CO₂-NaCl system. The calculated minimum trapping pressures of FIs of the early stage were estimated as 33–256 MPa, and those of the intermediate stage as 36–214 MPa, with the CO₂ densities and bulk densities varying from 0.61 g/cm³ to 0.63 g/cm³ and 0.87 g/cm³ to 0.89 g/cm³, respectively.

Therefore, the minimum pressures estimated for the early-stage corresponded to a hydrostatic depth of about 3 km. The maximum pressure corresponded to a lithostatic depth of about 8 km, given that the density of the overburden rocks of the Triassic granite was ca. 2.90 g/cm³. Hence, the W mineralization depth of Juyuan deposit was about 3–8 km.

5.4. Laser Raman Spectroscopy

Phases in representative FIs were measured using laser Raman microspectroscopy to constrain their compositions. Results of the analyses show that vapor bubbles in FIs in early-stage quartz contained CO₂, H₂O, and CH₄ (Figure 7a) and the liquid phases contained CO₂ and H₂O (Figure 7b). This is consistent with the microthermometric analyses results, which showed that the melting temperatures of solid CO₂ were lower than −56.6 °C.

Vapor bubbles of C-type FIs in intermediate-stage quartz were dominated by H₂O and CO₂ with no CH₄ detected by laser Roman spectroscopy (Figure 7c). Raman spectrum for the liquid and vapor of W-type inclusions in late-stage quartz was almost entirely H₂O. However, the microthermometric measurement and the Laser Raman spectroscopic detection showed the absence of CO₂ in any of the late-stage fluid inclusions (Figure 7), indicating that the fluid system evolved from CO₂-rich to CO₂-absent aqueous solution.

6. Isotope Geochemistry

6.1. Hydrogen and Oxygen

The H and O isotope compositions of selected minerals from different ore-forming stages were analyzed by determining the source of the mineralizing fluids. The O isotope data of fluids in equilibrium with quartz were calculated using the quartz-water equation of [47]. The analytical data and the calculated δ¹⁸O_water values of ore fluids are listed in

Table 2. The δ¹⁸O, δD, and δ¹³C ratios (‰) of the Juyuan tungsten deposit.

No.	Sample	Mineral	δ18Om	δ18Ow	δD	δ13CCO2	T(°C)	Stage	Reference
1	D1934-1a	Quartz	12.2	4.8	−86	−7.8	287	Early	This study
2	D1935-5	Quartz	13.6	6.0	−82	−7.7	280	Early	This study
3	D1937-5	Quartz	12.3	4.9	−98	−14.1	287	Early	This study
4	D1937-6	Quartz	12.9	5.5	−94	−7.4	287	Early	This study
5	D1938-3a	Quartz	13.2	5.7	−91	−3.3	285	Early	This study
6	D1958-2-1	Quartz	12.9	5.3	−95	−4.0	282	Early	This study
7	D1958-4a	Quartz	12.9	5.3	−88	−3.4	282	Early	This study
8	D1960-1	Quartz	12.9	5.5	−91	−6.2	287	Early	This study
9	D1963-1	Quartz	12.9	5.5	−88	−2.3	287	Early	This study
10	D1965-1	Quartz	12.4	5.3	−85	−1.5	294	Early	This study
11	JY	Quartz	12.7	5.3	−72	−9.0	287	Early	This study
12	D1935-2	Quartz	13.1	4.6	−92	−7.4	260	Intermediate	This study
13	D1938-1a	Quartz	12.7	4.2	−93	−6.6	260	Intermediate	This study
14	D1936-4	Quartz	13.0	1.3	−91	−6.5	200	Late	This study
15	D1959-2a	Quartz	12.2	0.5	−89	−7.5	200	Late	This study

. Measured δ¹⁸O values from the early stage ranged from 12.2‰ to 13.6‰. Using the equation, and corresponding fluid inclusion homogenization temperature data, the δ¹⁸O values for fluids from early-stage quartz were calculated to range from 4.8‰ to 6.0‰. The same calculation method applied to intermediate- and late-stage quartz yielded δ¹⁸O values of 4.2‰ to 6.0‰ and 1.3‰ to 0.5‰, respectively.

As the homogenization temperature of FIs provide only a lower limit of the trapping temperature, unless the FIs were entrapped from an immiscible or boiling fluid [47], the δ¹⁸O values of late-stage fluids calculated using FI homogenization temperatures may have been underestimated.

In a δD–δ¹⁸O diagram (Figure 8), the data from early-stage quartz and intermediate-stage approached those of the magmatic water box, with the δ¹⁸O values of the late-stage deviating towards the meteoric water line, suggesting that fluid in the Juyuan tungsten deposit was likely of magmatic origin, with possible later mixing of meteoric water.

6.2. Carbon

Fluid inclusions in different-stages quartz from the Juyuan deposit yielded δ¹³CCO₂ values of −14.1‰ to −1.5‰, with a mean of −6.3‰ (Table 2). Comparing the values of various available carbon reservoirs, such as organic matter (−27‰; [48]), atmospheric CO₂ (−8‰; [55]), dissolved CO₂ in fresh water (−9‰ to −20‰; [56]), continental crust (−7‰; [57]), mantle (−5‰ to −7‰; [56]), and igneous rocks (−3‰ to −30‰; [56]), we concluded that the carbon of the fluid could not have been supplied by any of the above carbon reservoirs alone, instead showing a value range characteristic of mixed sources. Two carbon values (−2.3‰ to −1.5‰) of samples from the Juyuan deposit early-stage are higher than those of other carbon reservoirs but fall in the range of marine carbonate (−3‰ to 2‰ [56]; Figure 9). The higher δ¹³CO₂ values may be due to diagenesis and dehydration of marine carbonates. Therefore, at least a portion of CO₂ in ore fluids of the Juyuan deposit probably originated from Pingtoushan formation marine carbonates.

6.3. Sulfur

Sulfides formed in different geochemical systems have different isotopic compositions. For determining the sources of hydrothermal deposits, the sulfur isotopes of sulfides are known to be a key tracer [63,64,65,66,67]. Sulfur isotope compositions were obtained for intermediate-stage pyrite and galenite. Values of the eight δ³⁴S_V-CDT determinations made in this study are listed in

Table 3. The δ³⁴S values of ores and rocks at the Juyuan tungsten deposit.

No.	Sample No.	Sample Description	Mineral	δ34S	Stage
1	D1935-6	Sulfides -bearing Quartz Vein	Pyrite	6.3	Intermediate
2	D1936-3	Sulfides -bearing Quartz Vein	Pyrite	6.2	Intermediate
3	D1936-4	Sulfides -bearing Quartz Vein	Pyrite	6.2	Intermediate
4	D1962-1	Sulfides -bearing Quartz Vein	Pyrite	6.3	Intermediate
			Average	6.25
5	D1935-6	Sulfides -bearing Quartz Vein	Galena	4.4	Intermediate
6	D1936-3	Sulfides -bearing Quartz Vein	Galena	4.0	Intermediate
7	D1936-4	Sulfides -bearing Quartz Vein	Galena	3.7	Intermediate
8	D1962-1	Sulfides -bearing Quartz Vein	Galena	3.7	Intermediate
			Average	3.95

and presented in Figure 10. The results of the isotope analyses of pyrite were remarkably homogeneous, ranging from +6.2‰ to +6.3‰ (n = 4). Galenite separates were also fairly homogeneous, with δ³⁴S_V-CDT values ranging from +3.7‰ to +4.4‰ and a mean of 4.0‰ (n = 4).

Sulfur isotope was much more enriched in pyrite than in galena, suggesting sulfur isotope equilibrium between the coexisting pyrite and galena in intermediate-stage quartz [63]. The mineral assemblage in this stage suggest a reducing environment, and it is agreement with the fact that no sulfate mineral was observed in our study. This means that the sulfur isotopes of these sulfides should reflect the sulfur isotopic signatures of the hydrothemal fluids [63].

Sulfides from the intermediate stage had reasonably homogeneous δ³⁴S values (+3.7‰ to +6.3‰), suggesting that the ore-forming fluids were derived mainly from crustal rocks. These values are higher than the average values of magmatic sources (mainly −3‰ to 3‰ δ³⁴S) and the deep mantle (~0‰; [63]), and fall within the range of crust-derived granite (−4.0‰ to 9.0‰; [62]), indicating a crustal source (Figure 10). During the process of sulfide crystallization, sulfides underwent a degree of exchange reaction with the surrounding rock.

6.4. Lead

Lead isotope data for selected intermediate-stage pyrite and galena from the Juyuan deposit had ²⁰⁶Pb/²⁰⁴Pb, ²⁰⁷Pb/²⁰⁴Pb, and ²⁰⁸Pb/²⁰⁴Pb values of 18.476–18.538, 15.640–15.706, and 38.437–38.648, with mean values of 18.509, 15.673, and 38.541 (

Table 4. Lead isotope ratios of ores and rocks at the Juyuan tungsten deposit.

No.	Sample No.	Testing Object	206Pb/204Pb	207Pb/204Pb	208Pb/204Pb
1	D1935-6	Pyrite	18.538	15.706	38.648
2	D1936-3	Galena	18.476	15.641	38.437
3	D1936-4	Pyrite	18.534	15.706	38.643
4	D1936-4	Galena	18.510	15.672	38.534
5	D1962-1	Galena	18.487	15.640	38.441
Average (n = 5)			18.509	15.673	38.541

), respectively. In Figure 11, the Pb isotope data plot between the orogen and lower crust fields or orogen and upper crust fields showed the complexity of lead sources. The sulfide Pb isotopes yielded two-stage model ages of 209–161 Ma, suggesting that the metal was likely sourced from Late Triassic–Middle Jurassic strata (Figure 11). The two-stage model ages of sulfides are younger than the timing of tungsten mineralization (240.0 ± 1.0 Ma), suggesting that the formation of sulfides postdated the tungsten deposition at Juyuan.

6.5. Muscovite ⁴⁰Ar/³⁹Ar Dating

Step-heating ⁴⁰Ar-³⁹Ar analytical results for muscovite (Sample D18105-5) are reported in

Table 5. ⁴⁰Ar-³⁹Ar data for muscovite from the Juyuan tungsten deposit.

T (°C)	(40Ar/39Ar) m	(36Ar/39Ar) m	(37Ar/39Ar) m	40Ar* (%)	F (40Ar*/39Ar)	39Ar (×10−14mol)	39Ar (%)	Age ± 1σ (Ma)
D18105-5, Sample weight = 10.8 mg, J = 0.006506
800	22.7382	0.0024	0.0183	96.88	22.0291	21.59	5.53	243.9 ± 1.2
900	21.8982	0.0008	0.0094	98.93	21.6644	41.32	10.59	240.1 ± 1.1
950	21.8163	0.0007	0.0053	99.09	21.6184	52.28	13.4	239.6 ± 1.1
1000	21.7323	0.0005	0.0019	99.26	21.5722	74.14	19	239.1 ± 1.1
1050	21.6891	0.0005	0.0007	99.26	21.5288	79.66	20.42	238.7 ± 1.1
1100	21.6752	0.0007	0.0033	99.1	21.4797	60.88	15.6	238.2 ± 1.1
1150	21.7859	0.0007	0.0038	99.04	21.5765	39.48	10.12	239.2 ± 1.1
1200	22.0963	0.0011	0.0281	98.49	21.7627	16	4.1	241.1 ± 1.2
1300	22.4804	0.0027	0.0749	96.41	21.6744	4.78	1.22	240.2 ± 1.2

Notes: ⁴⁰Ar* refer to radiogenic ⁴⁰Ar.

and displayed in Figure 12. Uncertainties of individual step-heating analyses and on the plateau age are reported at the 1σ level. The age spectrum shows a flat plateau that accounted for 100% of the ³⁹Ar released, indicating that K and radiogenic ⁴⁰Ar were distributed homogeneously and that the K-Ar isotopic system remained closed and away from heating disturbance during the geological history of this sample. Nine continuous steps (800–1300 °C) for the muscovite sample yielded a well-defined weighted plateau age of 240.0 ± 1.0 Ma (1σ; MSWD = 2.2; Figure 12a), an isochron age of 238.3 ± 1.1 Ma (MSWD = 1.3) with an initial ⁴⁰Ar/³⁶Ar ratio of 440 ± 53 (Figure 12b), and an inverse isochron age of 238.3 ± 1.1 Ma (MSWD = 1.2) with an initial ⁴⁰Ar/³⁶Ar ratio of 441 ± 51 (Figure 12c). The 240.0 ± 1.0 Ma age is considered to be a reliable estimate for the crystallization age of muscovite in this sample from the Juyuan deposit.

7. Discussion

7.1. Fluid Nature, Evlution, and Tungsten Deposition

Fluid boiling is an important mechanism for ore-metal deposition in many hydrothermal deposits [71,72,73,74]. Fluid boiling occurred at the Juyuan deposit. Four different types of FIs coexisted in the early-stage quartz. The coexisting vapor-rich and liquid-rich FIs homogenized divergently to respective vapor and liquid phases at around the same temperature range (Table 1 and Figure 5a). It can be concluded that the boiling of fluids resulted in the early-stage precipitation of wolframite and scheelite in the Juyuan deposit, given that fluid boiling is inferred to have occurred in early-stage quartz. Fluid inclusions of the late-stage showed low homogenization temperatures, indicating cooling of the fluids, most likely reflecting the progressive influx and mixing of meteoric water. Of note, the late-stage fluids had relatively high salinity. A decrease in temperature from the early to late stage resembles the fluid-mixing trend proposed, but the slightly higher salinity in the late-stage may be due to high-salinity fluid that was carried by meteoric water during its descent (Figure 13). Furthermore, H-O isotope data (Figure 8) for the intermediate- and late-stage fluid systems plotted towards the meteoric water line, indicating the increasing input of meteoric water. This fluid mixing and/or inflow of meteoric fluids would have resulted in rapid deposition of polymetallic sulfides and gangue minerals.

In summary, boiling of the ore-forming fluids, mixing of the fluids with meteoric water, further pH increase, and decreasing temperature were likely the most important mechanisms that led to tungsten precipitation in the Juyuan deposit.

Various types of FI are present in quartz of different stages, providing insights into the nature and evolution of the fluid system of the Juyuan deposit. From the early to late stages, some regularities of fluid evolution are summarized below. The total homogenization temperatures of fluid inclusions decreased from 230–344 °C (mostly 280–300 °C) during the early-stage, through 241–295 °C (mostly 260–280 °C) during the intermediate-stage, and to 184–234 °C during the late-stage. Fluid cooling and influx of meteoric water occurred during the late-stage.

The C-, W-, S-, and PC-type inclusions in early-stage quartz indicate that the initial fluid system was carbonic and of low salinity. The existence of W-, C-, and PC-type inclusions in early- and intermediate-stage quartz show that the fluid system underwent boiling during these two stages.

It is common for C- and W-type inclusions to coexist in the same FIA [75], indicating that they were captured simultaneously. Microscope measurement results show that the C- and W-type FIs differed in their homogenization style. The W-type FIs totally homogenized to liquid, but the C-type FIs with greater vapor/liquid ratios (>50%) homogenized to vapor. The C-type FIs had higher homogenization temperatures than those of the W-type FIs, but lower salinity. These features indicate that the fluid became immiscible after being captured. The higher salinity of W-type inclusions compared with C-type inclusions can be explained by fluid immiscibility causing the CO₂ phase that was dissolved in the fluid under higher pressure and temperature to become separated as a result of the decrease in pressure and temperature, leading to an increase in the salinity of the remaining fluid owing to the loss of gas.

Laser Raman spectroscopy analyses identified CO₂ and H₂O without CH₄ in fluid inclusions from the intermediate-stage. We propose that an increase in oxygen fugacity, which occurred owing to inflow of oxidized meteoric waters, can explain the absence of CH₄ (transition CH₄ → CO₂) [63]. Only W-type inclusions were observed in late-stage quartz, indicating that the late-stage fluid was a CO₂-poor aqueous solution fluid.

Data points for δD_water–δ¹⁸O_water from the early-to late-stage shifted from the magmatic water field towards the meteoric water line (Figure 8), and a consistent feature of fluid inclusions is some anomalously low-δD fluids. As mentioned above, boiling of the early-stage fluid occurred, and vapor separation of ore-forming fluids by this boiling may explain the anomalously low-δD fluids [66]. Experimental studies have suggested that D fractionates preferentially into the vapor phase relative to H during fluid boiling, reducing the δD_fluid values of the residual fluid by ~28‰ while scarcely changing the δ¹⁸O_fluid values [76,77].

It can be concluded that the fluid system evolved from a CH₄-rich and CO₂-rich system to one that was CO₂ rich but lacked CH₄. The system then evolved further into a CO₂-poor and CH₄-poor aqueous fluid system through CO₂ escape, fluid boiling, and inflow of meteoric water. This conclusion is supported by H-O isotope signatures that indicate that the ore-forming fluids evolved from almost primary magmatic water to meteoric.

7.2. Timing and Tectonic Setting of Formation of the Juyuan Deposit

A series of Triassic tungsten deposits are found in the East Tianshan Mountains, including the Shadong (239 Ma) [18], Jiangshan (230 Ma) [18], and Xiaobaishitou (245 Ma) [40] tungsten deposits. These and other deposits form a Triassic tungsten metallogenic belt in the East Tianshan Mountains, whose metallogenic tectonic setting was an intra-plate extensional regime. Two episodes of magmatism related to coeval tungsten mineralization have been recognized in the Beishan orogenic belt: (1) 424–314 Ma; for example, the Hongjianbingshan, Guoqing, Yingzuihongshan, and Baixianishan deposits. The related intrusions were derived from the partial melting of Precambrian metasedimentary strata or juvenile lower crust [12,15,16,34,46]; and (2) 286–220 Ma; for example, the Liushashan (262–260 Ma; [36,37]), Huaheitan (226 Ma; [78]), and Yushan deposits [79]. The granites were derived from ancient crustal sediment [17,21,36,37,38,39,40,79]. Rocks of these two magmatic events are distributed mainly in terranes with Precambrian metasedimentary strata, the Shuangyingshan-Huaniushan arc, and the Hanshan block.

Structural data from Permian to Triassic rocks in regions adjacent to the Beishan belt indicate the occurrence of extensive thrusting, strike-slip faulting, and possibly also extension associated with the amalgamation of different blocks [80,81,82,83,84,85,86,87,88]. Furthermore, paleomagnetic, geochronological, and stratigraphic data show that the convergence of the archipelago and syn-collisional crustal shortening had ended by the end-Carboniferous to Permian, following which post-collisional extension began [28]. The extensional environment is inferred from the widespread post-collisional mafic-ultramafic rocks and A-type granitic stocks that occur throughout the Beishan orogenic belt [78,79,87,89], as well as from the widespread N-W-trending ultrabasic dikes near the Juyuan deposit (Figure 2). The post-collisional extension stage is interpreted as being responsible for the widespread emplacement of 286–232 Ma granites and associated contemporaneous W-(Mo) deposits.

To discuss the tectonic setting, we used the available granite U-Pb geochronological and geochemical data from the Xinjiang Institute of Geological Survey. The Juyuan granite had low A/CNK ratios (<1.1), high Th and Y contents that increase with increasing Rb content, and a negative relationship between P₂O₅ and SiO₂, indicating an I-type granite affinity [90,91]. The granite also had high K, Rb, and Th values that are consistent with intra-plate magmatism. Isotopes of O and S support the interpretation that the hydrothermal fluids that formed the Juyuan deposit had a magmatic origin.

The orebodies of the Juyuan deposit run through the host granite, with mineralization styles dominated by quartz veins. The hydrothermal alteration consists of greisenization, sulfidization, and carbonation. The hydrothermal fluids were low in salinity and CO₂ rich. These features, as well as the isotope signatures discussed above, suggest that Juyuan is a quartz-vein-type tungsten deposit. The studied muscovite from the Juyuan deposit yielded an Ar-Ar plateau age of 240.0 ± 1.0 Ma (1σ; Figure 12a), which is consistent with zircon ²⁰⁶Pb/²³⁸U ages of 240.9 ± 0.8 Ma and 240.7 ± 0.5 Ma for monzonitic and two-mica granites from the deposit, respectively (unpublished data). As muscovite generally coexists with scheelite at Juyuan, the excellent agreement between the U-Pb and ⁴⁰Ar-³⁹Ar chronometers confirms that the emplacement of the monzonitic granite and the tungsten mineralization in the Juyuan deposit formed contemporaneously during the Middle Triassic (240 Ma).

On the basis of the regional geology, characteristics of magmatic rocks, the formation of the Juyuan tungsten deposit and other similar Triassic deposits is considered to have taken place in an intra-continental plate setting, with the magmatism and metallogenesis being associated with regional extension. The discovery of the Juyuan tungsten deposit extends the known metallogenic belt eastward from East Tianshan to the Beishan structural belt, where quartz-vein-type tungsten deposits that formed in the same intra-plate extensional environment might be found.

8. Conclusions

We conducted a systematic geological and geochemical study of the evolution of mineralizing fluids and ore genesis of the Juyuan tungsten deposit, Beishan, NW China. Our main conclusions are as follows:

(1) The Juyuan tungsten deposit is hosted in Triassic granite. The vein-type mineralization can be divided into three stages, namely, early-stage quartz-oxide, intermediate-stage quartz-sulfide, and late-stage carbonate-quartz.

(2) Quartz in the Juyuan deposit contained CO₂-H₂O, pure CO₂, NaCl-H₂O, and daughter-mineral fluid inclusions. Fluid inclusion analysis reveals that the fluid system evolved from CO₂-rich and CH₄-rich fluid to CO₂-rich and CH₄-absent fluid, then further evolved into CO₂-poor and CH₄-poor aqueous fluid through CO₂ escape, fluid boiling, and inflow/mixing of meteoric water. Fluid boiling and mixing were major processes leading to tungsten mineralization at Juyuan.

(3) Sulfides from the intermediate-stage had reasonably homogeneous δ³⁴S values (+3.7‰ to +6.3‰), indicating the ore-forming fluids were derived mainly from granitic crust. Lead isotope data suggest that the formation of sulfides postdated tungsten deposition at Juyuan.

(4) Zircon U-Pb and muscovite ⁴⁰Ar-³⁹Ar geochronology confirms that the tungsten mineralization of the Juyuan deposit formed contemporaneously with the host granites during the Triassic (240.0 ± 1.0 Ma). The formation of the Juyuan tungsten deposit and other similar Triassic deposits is considered to have taken place in an intra-continental plate setting, with the magmatism and metallogenesis being associated with regional extension.

(5) A comparison of the tungsten deposits in East Tianshan and Beishan suggests that a single Triassic tungsten metallogenic belt extends from East Tianshan to the Beishan structural belt.