Isotope Geochemistry of the Heihaibei Gold Deposit within the Kunlun River Area in the Eastern Kunlun Orogen in Northwest China and Its Metallogenic Implications

Abstract

The Heihaibei gold deposit is located in the Eastern Kunlun Orogen in Northwest China. The gold mineralization here occurs predominantly in quartz veins within faulted granite zones. The sulfide mineral assemblage is dominated by pyrite and arsenopyrite, with minor chalcopyrite, galena, sphalerite, tetrahedrite, and micro-native gold. Weak alterations in Heihaibei granites include silicification and sericitization, with minor chloritization and carbonatization. The measured δD_H2O and δ¹⁸O_quartz values of quartz in auriferous quartz veins range from −104.2‰ to −81.1‰ and +9.2‰ to +13.9‰, respectively. The δ³⁴S values of sulfides in auriferous quartz veins range from +7.60‰ to +8.65‰, and the lead isotope compositions of sulfides in ores range from 18.7219 to 19.0007 for ²⁰⁶Pb/²⁰⁴Pb, 15.6959 to 15.7062 for ²⁰⁷Pb/²⁰⁴Pb, and 37.7359 to 38.8055 for ²⁰⁸Pb/²⁰⁴Pb. The Pb isotope compositions of potassic feldspars from Heihaibei granites vary from 18.3532 to 19.4864 for ²⁰⁶Pb/²⁰⁴Pb, 15.6475 to 15.6812 for ²⁰⁷Pb/²⁰⁴Pb, and 37.1750 to 38.4598 for ²⁰⁸Pb/²⁰⁴Pb. Collectively, the isotope (H, O, S, and Pb) geochemistry suggests that the ore-forming fluid was a special metamorphic water evolved from the deep slab-derived fluids, and the sulfur and lead were predominantly sourced from such metamorphic fluids, and from the deep parts of the Heihaibei granites. Therefore, the Heihaibei gold deposit can be classified as an orogenic gold deposit, which is closely associated with the subduction of the Paleo-Tethys oceanic plate, and even the final closure of this ocean by the Later Triassic.

1. Introduction

The Eastern Kunlun Orogen (EKO) extends from east to west for more than 1200 km, with a width of 50–200 km along the northern margin of the Qinghai–Tibetan Plateau, and records the earliest amalgamation history of the Qinghai–Tibetan Plateau (Figure 1) [1,2]. The EKO, which is bounded by the Qaidam block (QDM) on the north and the Bayan Har–Songpanganzi Terrane (BH–SG) on the south, is usually considered to be composed of three sub-parallel tectonic belts from north to south, the North Kunlun belt (NKL), the Middle Kunlun belt (MKL), and the South Kunlun belt (SKL), separated by the North Kunlun fault (N. KLF), the Middle Kunlun fault (M. KLF), and the South Kunlun–Aryan Maqin fault (S. KLF–AMF), respectively [3]. Since 1999, numerous ore prospecting projects within the EKO have been deployed by the China Geology Survey [4]. So far, the EKO has become an important ore belt in which ca. 140 mineral occurrences regarding Au, Fe, Cu, Pb, Zn, W, Mo, Ni, Co, and Sb were discovered, and ca. 26 of them have become different sizes of mineral deposits in the EKO and the adjacent BH–SG (Figure 1b) [5,6]. Especially, the EKO hosts numerous gold deposits, as best represented by those in the Wulonggou gold field [4,7,8,9,10,11,12,13], Gouli gold field [14,15,16,17,18,19], and Kaihuangbei gold field [14,20].

In recent years, the Kunlun River goldfield (KLRA) in the SKL has gradually become an important gold mineralization belt because more than 10 gold deposits or occurrences were discovered throughout this area (Figure 2), among which the Heihaibei gold deposit is the largest [22,23,24,25]. Thus far, few geologists have carried out geological research on the Heihaibei gold deposit or on other deposits in the KLRA. Most of the published papers or reports emphasize the geological characteristics and gold prospecting potentials for the Heihaibei gold deposit [22,24,25,26,27]. Because the dominant ore types are auriferous quartz veins, it was thought that the Heihaibei gold deposit belongs to a hydrothermal-vein-type gold deposit [22], a magmatic hydrothermal gold deposit [24], or an orogenic gold deposit [25]. Based on the geological characteristics and a few stable isotope compositions, some other gold deposits in the KLRA such as the Heihainan, Heicigou, and Dazaohuo gold deposits were also considered to belong to orogenic gold deposits [28,29] or altered-rock-type gold deposits [30], and the ore-forming fluids and materials were derived predominantly from metamorphic water and wall rocks, respectively [28,29]. However, Yu et al. (2022) argued that the Heihaibei gold deposit was a granitic intrusion-related gold deposit rather than an orogenic gold deposit, and the ore-forming fluids and materials were derived from a magmatic hydrothermal source that was closely related to the Heihaibei granites, according to their three sets of H, O, S, and Pb isotope compositions [23]. In a word, the origins of the ore-forming fluids, sources of metals, and genetic types responsible for the Heihaibei gold deposit or other gold mineralization in the Kunlunhe River area remain controversial. Systematic isotope studies on the Heihaibei gold deposit have not been carried out, and especially, a controversy between the orogenic gold deposit and the granitic intrusion-related gold deposit has persisted for a long time.

A systematic study on the H, O, S, and Pb isotope geochemistry for auriferous quartz vein ores in the Heihaibei gold deposit was carried out. Comparing our new data and a few previously published data on the Heihaibei gold deposit, we can better interpret the sources of H, O, S, and Pb in the Heihaibei gold deposit, determine the genetic type of it, and even further constrain the sources of ore-forming fluids and materials for gold deposits within the whole KLRA.

2. Geological Background and Ore Deposit Geology

2.1. Geological Background

Usually, it is thought that the EKO has stages of orogenies that are closely related to the subduction and closure of the Proto-Tethys and Paleo-Tethys oceans [1,3,31,32,33,34,35,36,37]. Two ophiolite belts, which lay along the M.KLF and S.KLF–AMF, represent the suture zones of the Neo-Proterozoic to Early Devonian Proto-Tethys, and the Carboniferous to Late Triassic Paleo-Tethys oceans, respectively [3,31,37]. The EKO is considered to be divided into three fault-bounded tectonic belts, i.e., NKL, MKL, and SKL (Figure 1b) [3]. The NKL and MKL have dual tectonic layers of a common Precambrian craton, which are characterized by the Paleo-Proterozoic basement, with intermediate to high-grade metamorphism, and the Meso- to Neo-Proterozoic cover strata, with low-grade metamorphism [2]. The NKL is characterized by thick Ordovician clastic and carbonate sedimentary rocks interlayered with mafic volcanic rocks, which were unconformably overlain by the Devonian molasse sediments [38,39]. Widespread granitoids intruded into these Ordovician and Devonian strata, primarily in the late Paleozoic, in this belt [40,41]. The MKL, in which the Precambrian metamorphic rocks vary from gneiss to schist to phyllite [3,39], was predominantly emplaced by the widespread Neo-Proterozoic, early Cambrian–Early Devonian, and Late Permian–Early Triassic granitic intrusions [13,42,43,44,45,46]. The SKL is dominated by Precambrian to Jurassic siliciclastic and carbonate rocks [3,43], and it was locally intruded upon by the Paleozoic and Triassic granites [23,47,48]. The KLRA is tectonically located in the central segment of the SKL within the EKO (Figure 1b).

2.2. Ore Deposit Geology

The KLRA, as the voluminous area to the north bank of the Kunlun River is usually referred to, is a recently discovered gold mineralization belt in the SKL (Figure 2). The northwest-striking Fault F51, which is a segment of the M.KLF, is a regional fault that bounds the MKL and SKL in the KLRA. To the north of Fault F51, the part of the MKL is characterized by widespread Permian to Triassic granitoids [22]. To the south of Fault F51, part of the SKL is composed of sedimentary rocks which were deposited predominantly in the Early Paleozoic and Quaternary, with minor parts in the Proterozoic and Tertiary. A few Silurian and Triassic granitoids were intruded in the Early Paleozoic and Mesozoic sedimentary rocks (Figure 2). Subsidiary faults of Fault F51 that were usually considered to be the products of the closure of the Proto-Tethys Ocean [49] mostly affect the SKL. Moreover, Fault F51 and its subsidiary faults in the KLRA would have been reactivated from the Early Triassic onwards because of the subduction of the Paleo-Tethys oceanic plate [2,49]. Ore-forming hydrothermal fluids can migrate upward through such reactivated faults. More than 10 gold deposits are controlled by the WNW-striking subsidiary faults to the south of Fault F51 in the KLRA (Figure 2) [22,25].

The Heihaibei gold deposit, which is located ca. 20 km NW of the Heihai Lake, is the westernmost of 10 gold deposits in the KLRA (Figure 3). The dominant sedimentary rocks in the Heihaibei area are mainly composed of Silurian–Ordovician Nachitai Group metamorphic feldspathic quartz sandstone and sandy slate in the north, and Triassic Hongshuichuan Formation feldspathic quartz sandstone in the south, with minor Middle-Neo-Proterozoic Wanbaogou Group sandy limestone and basalt, and Triassic Naocangjiangou Formation feldspathic quartz sandstone in the southeast (Figure 3a) [22,23]. An EW- to WNW-striking >15 km long and >100 m wide monzogranite–granodiorite dyke (i.e., the Heihaibei granites) was intruded along the south of the Silurian–Ordovician Nachitai Group. We argued that this monzogranite–granodiorite dyke was classified as A-type granite [50] and was emplaced in the Early Devonian (416 Ma) in a post-collisional setting [50]), rather than in the Ordovician–Silurian (442 Ma [27]), 443 Ma [25], and 453 Ma [23]). The Triassic Hongshuichuan Formation feldspathic quartz sandstone and other minor Triassic sediments were deposited over older sedimentary rocks and the Heihaibei granites [24]. Subsidiary faults in the Heihaibei area predominantly include the NW-striking Fault F52 and Fault F53 (Figure 3a). Fault F52, ca. 18 km long and 10–100 m wide, generally strikes WNW across the Nachitai Group sandstone in the northeast part of the Heihaibei mining area and dips 25°–52° to the north; meanwhile, Fault F53, ca. 15 km long and 10–100 m wide, generally strikes EW across the central part of the study area and dips 30°–63° to the north [23,24]. Cataclasites, fault breccias, and fault gouges are developed in the faulted zones of Faults F52 and F53. Fault F53 with secondary branch faults, i.e., Fault F53-1, is commonly developed along contacts of different geological bodies in the Heihaibei area, i.e., between the granites and the Triassic Hongshuichuan Formation, and the granites and the Nachitai Group, as well as the slate and sandstone in the Nachitai Group. Some auriferous quartz veins, pyritizations, and weak hydrothermal alterations are developed along the faulted monzogranite zone of Fault F53, but little mineralization occurs in Fault F52, which is limited in the Nachitai Group sandstone (Figure 3b).

Weak hydrothermal alterations within monzogranite wall rocks (Figure 4a) are mostly silicification and sericitization, with minor chloritization and carbonatization. In monzogranites, biotite is basically altered into chlorite, plagioclase is partially altered into muscovite, and hydrothermal quartz and/or calcite fill into fine fractures (Figure 4b,c). More than 20 gold ore bodies (with Au abundances larger than the cut-off grade) and 30 auriferous bodies (with Au abundances below the cut-off grade) were delineated in the faulted monzogranite zone within the fault F53, and they yielded a total reserve of 5.7 t Au, with an average grade of 4.8 g/t, in which the 660 m long and average 5 m wide ore body M3 contributes to ca. 90% of the Au reserves [22]. Most of the gold ore bodies strike EW to ENE (east-northeast) and occur as lenses along the faulted monzogranite zone within Fault F53 (Figure 3b and Figure 5). Ore bodies consist of predominant smoky grey quartz veins with minor stockwork quartz veinlets or breccias in the altered monzogranite, and smoky grey quartz veins are usually bordered by fault breccia belts in the faulted monzogranite belt [22,24].

Three mineralized stages were proposed in the Heihaibei gold deposit, i.e., the K-feldspar-quartz-pyrite-arsenopyrite-sericite-epidote (I), the quartz-pyrite-native gold-chlorite (II), and the quartz-carbonate (III) by [23]. Based on the field survey and the optical microscopic study, we argued that the so-called K-feldspar-quartz-pyrite-arsenopyrite-sericite-epidote stage (I) represents silicified granites with few pyrite veinlets, and that the K-feldspar in such ores is a primary rock-forming mineral rather than a hydrothermal mineral in the Heihaibei granites. Additionally, we cannot find any typical products (e.g., quartz calcite veins) of the so-called quartz-carbonate stage (III) in the ore deposit. Therefore, we considered that the paragenetic sequences in the Heihaibei gold deposit might be divided into three major stages, i.e., the hydrothermal stage, the deformation stage, and the supergene stage (Figure 6). The hydrothermal stage can be subdivided into the sulfide–poor–quartz epoch (I) and the polymetallic–sulfide–quartz epoch (II). A few white quartz veins (here called Qv1), which are composed of predominant quartz with minor sericite, pyrite, and arsenopyrite, represent the sulfide–poor–quartz ore epoch (I) (Figure 4e,f), and some white quartz veinlets or silicification in weakly altered monzogranite wall rocks (Figure 4b,c) are the products of this epoch. Predominant smoky grey quartz veins (here called Qv2), which represent the ores from the polymetallic–sulfide–quartz epoch (II), usually contain predominant quartz with a few sulfide veinlets, disseminations, or massive aggregations, which consist of pyrite and arsenopyrite, with minor native gold, chalcopyrite, galena, sphalerite, and tetrahedrite (Figure 4g–j,m–o). Micro-native gold is mostly trapped in pyrite, and arsenopyrite in the Qv2 (Figure 4n). Usually, the contents of sulfides in the Qv2 are less than 5%. Some pyrite–quartz-bearing veinlets and stockworks, which can be found in the weakly altered monzogranite wall rock or Qv1, constitute gold mineralized zones, and some of them may be delineated to ore bodies. Comparatively, the sulfide–poor–quartz epoch (I) and polymetallic–sulfide–quartz epoch (II) in this study are roughly referred to as the stage I and stage II, respectively, as classified by [23]. Partially, both Qv1 and Qv2 can be deformed within the faulted monzogranite zone of Fault F53 and become breccia ores (Figure 4k,l). The oxidized ores from the supergene stage mostly consist of goethite, lepidocrocite, and covellite, which result from the replacement of the hydrothermal stage of primary Qv1 and Qv2 (Figure 4n and Figure 6). In a word, the geological characteristics of the Heihaibei gold deposit show similar features to those of the orogenic gold deposits defined by [51].

3. Sampling and Analytical Methods

3.1. Sampling

The samples for isotope analyses were collected from drill cores, i.e., ZK0602, ZK0603, ZK01, ZK1002, and ZK1003, or surface outcrop sites, i.e., HHB01, HHB07, and HHB08. Detailed sample locations and numbers can be found in Figure 3b and Figure 5, and

Table 1. Sample list in the Heihaibei gold deposit.

No.	Sample Name	Rock Type	Ore Type	Mineral Assembly	Location
1	HHBZK0602-B2	brecciated white quartz vein	Qv1	quartz with minor sericite, pyrite, and arsenopyrite	69.7 m of the drillhole core of the ZK1002
2	HHBZK1002-B4	brecciated white quartz vein	Qv1		62.9 m of the drillhole core of the ZK1004
3	HHBZK1002-B6	brecciated white quartz vein	Qv1		72.3 m of the drillhole core of the ZK1004
4	HHBZK1002-B8	brecciated white quartz vein	Qv1		75.5 m of the drillhole core of the ZK1004
5	HHB08-JK01	white quartz vein	Qv1		93°06′39″ E, 36°08′35″ N
6	HHB08-JK10	white quartz vein	Qv1		93°06′39″ E, 36°08′35″ N
7	HHB08-JK17	white quartz vein	Qv1		93°06′39″ E, 36°08′35″ N
8	HHBZK01-B3	brecciated smoky grey quartz vein	Qv2	quartz with some pyrite and arsenopyrite, and minor native gold, chalcopyrite, galena, and sphalerite, as well as tetrahedrite	26.4 m of the drillhole core of the ZK01
9	HHBZK01-B5	brecciated smoky grey quartz vein	Qv2		29.6 m of the drillhole core of the ZK01
10	HHBZK01-B6	brecciated smoky grey quartz vein	Qv2		30.3 m of the drillhole core of the ZK01
11	HHBZK0602-B3	smoky grey quartz vein	Qv2		70.8 m of the drillhole core of the ZK0602
12	HHBZK0603-B2	smoky grey quartz vein	Qv2		114.7 m of the drillhole core of the ZK0603
13	HHBZK0603-B3	brecciated smoky grey quartz vein	Qv2		115.4 m of the drillhole core of the ZK0603
14	HHBZK0603-B5	brecciated smoky grey quartz vein	Qv2		117.0 m of the drillhole core of the ZK0603
15	HHBZK1003-B11	brecciated smoky grey quartz vein	Qv2		115.9 m of the drillhole core of the ZK1003
16	HHBZK1002-B11	brecciated smoky grey quartz vein	Qv2		82.8 m of the drillhole core of the ZK1002
17	HHB08-JK02	smoky grey quartz vein	Qv2		93°06′39″ E, 36°08′35″ N
18	HHB08-JK03	smoky grey quartz vein	Qv2		93°06′39″ E, 36°08′35″ N
19	HHB08-JK05	smoky grey quartz vein	Qv2		93°06′39″ E, 36°08′35″ N
20	HHB08-JK08	smoky grey quartz vein	Qv2		93°06′39″ E, 36°08′35″ N
21	HHB08-JK13	smoky grey quartz vein	Qv2		93°06′39″ E, 36°08′35″ N
22	HHB08-JK15	smoky grey quartz vein	Qv2		93°06′39″ E, 36°08′35″ N
23	HHB08-JK18	smoky grey quartz vein	Qv2		93°06′39″ E, 36°08′35″ N
24	HHB08-JK19	smoky grey quartz vein	Qv2		93°06′39″ E, 36°08′35″ N
25	HHBZK01-B11	monzogranite	wall rock	plagioclase, 25–40% alkali feldspar, 25–45% quartz, <5% biotite	175.0 m of the drillhole core of the ZK01
26	HHB01-B9	monzogranite	wall rock		93°07′14″ E, 36°08′30″ N
27	HHBZK01-B14	granodiorite	wall rock	35–50% plagioclase, 35–45% quartz, 10–15% alkali feldspar, <5% biotite	209.4 m of the drillhole core of the ZK01
28	HHBZK1003-B7	granodiorite	wall rock		107.2 m of the drillhole core of the ZK1003
29	HHB07-B1	granodiorite	wall rock		93°07′09″ E, 36°08′36″ N

. Fifteen auriferous quartz samples (Qv1 and Qv2) were collected for oxygen and hydrogen isotope analysis (

Table 2. H and O isotope data for quartz in the Heihaibei gold deposit. δD was measured in fluid inclusions; δ¹⁸O was measured in quartz; δ¹⁸O of water in equilibrium with quartz was calculated at the temperature indicated by Th [23] using the equation of 1000lnα = 3.38 × 10⁶/T² − 3.4 [52].

No.	Sample Name	Rock Type	Ore Type	Mineral	δDVSMOW (‰)	δ18OVSMOW (‰)	δ18OH2O (‰)	Average Th (°C)	References
1	HHBZK0602-B2	brecciated white quartz vein	Qv1	Quartz	−97.2	13.4	7.8	340.0	this study
2	HHBZK0602-B3	smoky grey quartz vein	Qv2	Quartz	−91.0	12.1	2.7	241.0	this study
3	HHBZK0603-B2	smoky grey quartz vein	Qv2	Quartz	−101.8	12.6	3.2	241.0	this study
4	HHBZK0603-B3	brecciated smoky grey quartz vein	Qv2	Quartz	−97.7	12.8	3.4	241.0	this study
5	HHBZK0603-B5	brecciated smoky grey quartz vein	Qv2	Quartz	−102.3	13.3	3.9	241.0	this study
6	HHB08-JK01	white quartz vein	Qv1	Quartz	−81.1	9.2	3.6	340.0	this study
7	HHB08-JK02-1	smoky grey quartz vein	Qv2	Quartz	−94.1	13.7	4.3	241.0	this study
8	HHB08-JK02-2	smoky grey quartz vein	Qv2	Quartz	−94.6	13.7	4.3	241.0	this study
9	HHB08-JK03	smoky grey quartz vein	Qv2	Quartz	−89.9	11.3	1.9	241.0	this study
10	HHB08-JK05	smoky grey quartz vein	Qv2	Quartz	−90.4	13.6	4.2	241.0	this study
11	HHB08-JK08	smoky grey quartz vein	Qv2	Quartz	−97.4	11.6	2.2	241.0	this study
12	HHB08-JK10	white quartz vein	Qv1	Quartz	−92.8	13.9	8.3	340.0	this study
13	HHB08-JK15	smoky grey quartz vein	Qv2	Quartz	−97.6	11.6	2.2	241.0	this study
14	HHB08-JK17	white quartz vein	Qv1	Quartz	−96.8	11.0	5.4	340.0	this study
15	HHB08-JK18	smoky grey quartz vein	Qv2	Quartz	−104.2	13.0	3.6	241.0	this study
16	HHB08-JK19	smoky grey quartz vein	Qv2	Quartz	−101.5	11.0	1.6	241.0	this study
17	K1	silicified monzogranite	Qv1+wall rocks	Quartz	−84.1	15.1	9.5	340.0	[23]
18	K5	smoky grey quartz vein	Qv2	Quartz	−81.1	14.1	4.7	241.0	[23]
19	K7	smoky grey quartz vein	Qv2	Quartz	−84.0	14.0	4.6	241.0	[23]

). Seventeen sulfide-bearing ore samples (Qv1 and Qv2) were collected for sulfur isotope analysis (

Table 3. Sulfur isotope data for sulfides in the Heihaibei deposit. δ³⁴S_H2S means δ³⁴S_VCDT values for H₂S in equilibrium with sulfides. The assumed temperatures of equilibrium for different ore types are the same as Th [23] listed in Table 1.

No.	Sample Name	Rock Type	Ore Type	Mineral	δ34SCDT(‰)	δ34SH2S(‰)	References
1	HHBZK01-B3	brecciated smoky grey quartz vein	Qv2	Pyrite	8.5	7.0	this study
2	HHBZK01-B5	brecciated smoky grey quartz vein	Qv2	Pyrite	8.0	6.5	this study
3	HHBZK01-B6	brecciated smoky grey quartz vein	Qv2	Pyrite	7.7	6.2	this study
4	HHBZK0603-B2	smoky grey quartz vein	Qv2	Pyrite	8.0	6.5	this study
5	HHBZK0603-B3	brecciated smoky grey quartz vein	Qv2	Pyrite	7.6	6.1	this study
6	HHBZK0603-B5	brecciated smoky grey quartz vein	Qv2	Pyrite	8.0	6.5	this study
7	HHBZK1003-B11	brecciated smoky grey quartz vein	Qv2	Pyrite	7.6	6.1	this study
8	HHBZK1002-B4-1	brecciated white quartz vein	Qv1	Pyrite	8.1	7.0	this study
9	HHBZK1002-B4-2	brecciated white quartz vein	Qv1	Pyrite	8.0	6.9	this study
10	HHBZK1002-B6	brecciated white quartz vein	Qv1	Pyrite	8.0	6.9	this study
11	HHBZK1002-B8	brecciated white quartz vein	Qv1	Pyrite	8.1	7.0	this study
12	HHBZK1002-B11	brecciated smoky grey quartz vein	Qv2	Pyrite	7.8	6.3	this study
13	HHB08-JK02	smoky grey quartz vein	Qv2	Pyrite	8.5	7.0	this study
14	HHB08-JK08	smoky grey quartz vein	Qv2	Pyrite	8.0	6.5	this study
15	HHB08-JK13	smoky grey quartz vein	Qv2	Pyrite	8.3	6.8	this study
16	HHB08-JK15	smoky grey quartz vein	Qv2	Pyrite	8.2	6.7	this study
17	HHB08-JK17	white quartz vein	Qv1	Pyrite	8.7	7.6	this study
18	K1	silicified monzogranite	Qv1 + wall rocks	Arsenopyrite	8.5		[23]
19	K5	smoky grey quartz vein	Qv2	Pyrite	7.7		[23]
20	K8	smoky grey quartz vein	Qv2	Pyrite	7.9		[23]

). Four sulfide-bearing ore samples (Qv1 and Qv2) and five potassic feldspars in the Heihaibei granites were collected for lead isotope analysis (

Table 4. Pb isotope compositions for sulfides in the Heihaibei gold deposit.

No.	Sample	Rock Type	Ore Type	Mineral	206Pb/204Pb	207Pb/204Pb	208Pb/204Pb	References
1	HHBZK01-B3	brecciated smoky grey quartz vein	Qv2	pyrite	18.8122	15.7002	38.8055	this study
2	HHBZK1002-B4	brecciated white quartz vein	Qv1	pyrite	19.0007	15.7062	38.7655	this study
3	HHB08-JK08	smoky grey quartz vein	Qv2	pyrite	18.7219	15.6959	38.7356	this study
4	HHB08-JK19	smoky grey quartz vein	Qv2	pyrite	18.8262	15.6998	38.7836	this study
5	HHBZK01-B11	monzogranite	wall rock	K-feldspar	18.4283	15.6789	38.4342	this study
6	HHBZK01-B14	granodiorite	wall rock	K-feldspar	18.4363	15.6808	38.4532	this study
7	HHBZK1003-B7	granodiorite	wall rock	K-feldspar	18.4864	15.6475	38.1750	this study
8	HHB01-B9	monzogranite	wall rock	K-feldspar	18.4368	15.6812	38.4598	this study
9	HHB07-B1-1	granodiorite	wall rock	K-feldspar	18.3532	15.6750	38.3713	this study
10	HHB07-B1-2	granodiorite	wall rock	K-feldspar	18.3537	15.6753	38.3724	this study
11	K1	silicified monzogranite	Qv1 + wall rocks	pyrite	19.188	15.713	39.235	[23]
12	K5	smoky grey quartz vein	Qv2	pyrite	18.722	15.695	38.728	[23]
13	K12	smoky grey quartz vein	Qv2	pyrite	18.724	15.699	38.746	[23]

). Previously published data for other gold deposits in the EKO, compared with data in this study, can be found in the Supplementary Materials (Tables S1–S3).

3.2. Oxygen and Hydrogen Isotope Analysis

Oxygen isotope analyses were carried out with the bromine pentafluoride method of [53] at the Key Laboratory of Metallogeny and Mineral Assessment, Institute of Mineral Resources, Chinese Academy of Geological Sciences (CAGS), China. Oxygen was liberated from 6 mg of quartz via reaction with BrF₅ and converted to CO₂ on a platinum-coated carbon rod. The δ¹⁸O determinations were made on a Thermo Scientific 253 plus mass spectrometer, with analytical precisions of ±0.2‰, and the detailed analytical procedures are described by [54]. Hydrogen isotope analyses were carried out at the Beijing Createch Testing Technology Co., Ltd, Beijing, China (BCTT). Hydrogen isotopic ratios were measured on water released via decrepitation (at 1420 °C), of crushed and sieved (to the size range of 60 mesh) 10–20 mg quartz, according to the method of [55]. Samples were first degassed by heating under a vacuum at 90 °C for 12 h; then, the water was released from fluid inclusions by heating the samples to 1420 °C in a Thermo Scientific Flash 2000 HT elemental analyzer. The subsequently released water was trapped, reduced to H₂ by glassy carbon, and then analyzed for δD values using a Thermo Scientific 253 plus mass spectrometer, with a precision of ±1‰. The isotope results are expressed in the delta (δ) notation as the per mil (‰) deviation relative to the Vienna Standard Mean Ocean Water (V-SMOW) standard. The isotopic fractionation of oxygen between quartz and water was calculated using the equation of 1000lnα = 3.38 × 10⁶/T² − 3.4 [52].

3.3. Sulfur Isotope Analysis

Sulfur isotope analyses of sulfide separates were carried out at BCTT using the conventional combustion method [56]. The detailed analytical procedures are described by [57]. Approximately 15 mg of sulfides were combusted at 960 °C for quantitative conversion to SO₂, which were then analyzed for sulfur isotope with a Thermo Scientific 253 plus mass spectrometer. The results are reported in the δ notation as the ‰ deviation of the isotope ratio (³⁴S/³²S) relative to the Vienna Canyon Diablo Troilite (V-CDT) standard, with an analytical precision of ± 0.2‰.

3.4. Lead Isotope Analysis

Lead isotope analyses were conducted on four pyrite separates and five potassic feldspar separates at BCTT, using the analytical procedure as described in [58]. About 150 mg of the pyrite or feldspar sample were completely dissolved in an ultrapure acid mixture of 1 mL HNO₃ + 2 mL HF. After drying, the residue was redissolved in concentrated HNO₃ and dried again. Finally, the sample residue was dissolved in 1mL 3.5 M HNO₃ and then loaded into a column with 100–150 μm of Triskem Sr-Spec resins to separate the Pb fraction by washing with 3.5 M HNO₃, 8 M HCl, and Milli-Q water. Finally, the extracted sample residue was dissolved in 1 mL 2% HNO₃. Isotopic ratios of Pb analyses were undertaken using a Thermo Scientific Neptune Plus Multi Collector-Inductively Coupled Plasma Mass Spectrometer (MC-ICP-MS), and the results were corrected using ²⁰³Tl/²⁰⁵Tl = 0.418922 [58]. The data quality was monitored via repeated analyses of the CAGS Pb reference material, and the error margin (2σ) was <0.002 for all three ratios.

4. Results

4.1. Oxygen and Hydrogen Isotope Results

Quartz samples from quartz veins show a δ¹⁸O_quartz range from +9.2 to +13.9‰, with an average of +12.3‰ (Table 2, Figure 7). It is difficult to calculate the precise δ¹⁸O_H2O values of fluids that were in equilibrium with quartz, due to a large variation of temperatures [23] obtained from fluid inclusions (e.g., [59]). According to the microthermometry of fluid inclusions in auriferous quartz from the Heihaibei gold deposit finished by [23], the homogenization temperatures of Qv1 type of quartz from epoch I range from 268 °C to 412 °C, with an average of 340 °C; meanwhile, the homogenization temperatures of the Qv2 type of quartz from epoch II range from 183 °C to 288 °C, with an average of 241 °C. Usually, the trapping temperature could be defined by the average homogenization temperature to calculate the isotopic fractionation values. Given 340 °C as the trapping temperature for epoch I, the δ¹⁸O_H2O values of the ore-forming fluids in the fluid inclusions from the quartz in the Qv1 samples range from +3.6‰ to +8.3‰ (Table 2, Figure 7b). Meanwhile, given 241 °C as the trapping temperature for epoch II, the δ¹⁸O_H2O values of the ore-forming fluids for fluid inclusions from quartz in the Qv2 samples range from +1.6‰ to +4.3‰ (Table 2, Figure 7b). Generally, the trapping temperatures are most likely higher than the homogenization temperatures for the orogenic fluids.

The hydrogen isotope results for the fluid inclusions from the quartz samples are also shown in Table 2, in which the δD_VSMOW values range from −104.2‰ to −81.1‰, with an average of −95.3‰ (Figure 7). The quartz in the Qv1 samples has δD_VSMOW values ranging from −97.2 ‰ to −81.1‰, and the quartz in the Qv2 samples has δD_VSMOW values ranging from −104.2‰ to −89.9‰ (Figure 7b).

Consequently, the analytical hydrogen isotopes and the calculated δ¹⁸O_H2O values for the fluid inclusions from the quartz in Qv1 and Qv2 samples are different from each other (Figure 7b). Three sets of hydrogen–oxygen isotope compositions for the Heihaibei gold deposits were previously reported by [23], in which two samples from the smoky grey quartz veins (Qv2) have similar hydrogen–oxygen isotope compositions as our data from Qv1. However, the sample K1 is slightly different from our data from Qv1, likely because it might have been collected in the mineralized Heihaibei granitic wall rock, thus representing a mixed sample of Qv1 plus the granitic wall rock.

4.2. Sulfur Isotope Results

Only pyrite separates had been selected from 16 ore samples and analyzed for sulfur isotope compositions. Other sulfide separates were insufficient for analysis because the ore samples were collected from drill cores and had limited volumes. The δ³⁴S_VCDT values of 17 pyrite separates from 16 ore samples in the Heihaibei gold deposit range from +7.6‰ to +8.7‰, with a narrow variation and an average of 8.1‰ (Table 3, Figure 8a). Pyrite separates from the Qv1 type of ores have almost invariable δ³⁴S values ranging from +8.0‰ to +8.7‰, with an average of + 8.2‰. Pyrite separates from the Qv2 type of ores have similar invariable δ³⁴S values ranging from +7.6‰ to +8.5‰, with an average of +8.0‰. The δ³⁴S values of all pyrites have very narrow variations, and the δ³⁴S values of different ores have a comparably small range, usually indicating a relatively unique source.

Furthermore, our sulfur isotope data, ranging from +7.6‰ to +8.7‰, essentially coincide with the three previously published δ³⁴S_VCDT values, i.e., +7.7‰, +7.9‰, and +8.5‰ [23].

4.3. Lead Isotope Results

Only four pyrites separated from four ore samples in the Heihaibei gold deposit were large enough and yielded relatively invariable Pb isotope compositions, with ²⁰⁶Pb/²⁰⁴Pb ratios of 18.7219 to 19.0007, ²⁰⁷Pb/²⁰⁴Pb ratios of 15.6959 to 15.7062, and ²⁰⁸Pb/²⁰⁴Pb ratios of 37.7359 to 38.8055 (Table 4, and Figure 9a,b). Different types of quartz veins have relatively similar Pb isotopic compositions (Figure 9c,d). One pyrite separate from the Qv1 type of ores has ²⁰⁶Pb/²⁰⁴Pb ratios of 19.0007, ²⁰⁷Pb/²⁰⁴Pb ratios of 15.7062, and ²⁰⁸Pb/²⁰⁴Pb ratios of 38.7655. Three pyrite separates from the Qv2 type of ores have invariable Pb isotopic ratios, with ²⁰⁶Pb/²⁰⁴Pb ratios from 18.7219 to 18.8262, ²⁰⁷Pb/²⁰⁴Pb ratios from 15.6959 to 15.7002, and ²⁰⁸Pb/²⁰⁴Pb ratios from 38.7356 to 38.8055. Three sets of Pb isotope compositions for the Heihaibei gold deposits have previously been reported by [23], in which two samples from smoky grey quartz veins (Qv2) have similar Pb compositions as our data from Qv2, but the sample K1 has more radiogenic Pb isotope than our data from Qv1.

Six potassic feldspar separates from five Heihaibei granite wall rocks have invariable Pb isotope compositions, with ²⁰⁶Pb/²⁰⁴Pb ratios of 18.3532 to 19.4864, ²⁰⁷Pb/²⁰⁴Pb ratios of 15.6475 to 15.6812, and ²⁰⁸Pb/²⁰⁴Pb ratios of 37.1750 to 38.4598, indicating the initial lead isotope for the granite wall rocks when they were emplaced at ca. 416 Ma [50].

5. Discussion

5.1. Origins of the Ore-Forming Fluid

The δ¹⁸O_VSMOW values (+9.2–+13.9‰) of quartz separates from quartz veins at the Heihaibei gold deposit are basically similar to those of the Phanerozoic orogenic gold deposits elsewhere (δ¹⁸O = +7–+13‰, [63,64]). Given the average homogenization temperatures of 340 °C and 241 °C as the trapping temperature for the Qv1 and Qv2 samples, the calculated δ¹⁸O_H2O values of the fluids in epoch I and epoch II show relatively large variations, from +3.6 to +8.3‰ and +1.6 to +4.3‰, respectively (Table 2, the solid contour lines in Figure 7b). Generally, the trapping temperatures are most likely higher than the homogenization temperatures for the ore-forming fluids. Therefore, the calculated δ¹⁸O_H2O values of the fluids would be slightly higher than the values of +1.6–+8.3‰ in the Heihaibei gold deposit. In this regard, application of the arsenopyrite geothermometer, calculated using our unpublished electron microprobe analysis (EMPA) data, even higher average temperatures of 417 °C are obtained for epoch I and 385 °C for epoch II, which could also roughly represent the trapping temperature for the pyrite + arsenopyrite assembly in the Qv1 and Qv2, respectively. Given 417 °C and 385 °C as the trapping temperatures for Qv1 and Qv2, the calculated δ¹⁸O_H2O values of the fluids in epoch I and epoch II would be obviously higher than the previously calculated ones, and they would be up to +5.5–+10.2‰ and +6.6–+9.3‰, respectively (the dashed contour lines in Figure 7b). Sheppard (1986) indicates that the δ¹⁸O_H2O values for fluids of magmatic origin are mostly less than +9‰, whereas the metamorphic fluids are usually higher than +3‰ [61]. Therefore, the relatively high and variable δ¹⁸O_H2O values of the ore-forming fluids in the quartz separates in the Heihaibei gold deposit are consistent with a metamorphic origin or alternatively a magmatic fluid that was related to the Heihaibei granites hosting the ore, as proposed by [23]. However, we argue that the ore-bearing Heihaibei granites are not the potential source granites because of their initial Pb isotope ratios, which are totally different from the Pb isotopes of ores (see below). Therefore, the ore-forming water is not possibly derived from the magmatic water related to the Heihaibei granites. A deep magmatic water is still possible, but there is no evidence of another granitic intrusion deeper than the Heihaibei granites, and usually, magmatic water that potentially contributes to hydrothermal alteration and corresponding mineralization should be spatio-temporally related to a nearby-granitic intrusion, i.e., the Heihaibei granites. Moreover, the ore-forming fluids in the Heihaibei gold deposit are characterized by low to intermediate salinity, and intermediate to high homogenization temperatures, and no daughter mineral was trapped in the fluid inclusions [23], also lowering the possibilities of magmatic water. Apparently, higher calculated δ¹⁸O_H2O values of the fluids are closer to those of most of the orogenic gold deposits suggested by [59]. In a word, we argue a likely metamorphic origin rather than a magmatic water one, although we cannot totally preclude a magmatic origin. The calculated δ¹⁸O_H2O values of the Qv1 type of samples (+3.6–+8.3‰, calculated with 340 °C) are commonly higher than those of the Qv2 type of samples (+1.6–+4.3‰, calculated with 240 °C) (Figure 7b), perhaps indicating the mixing of minor meteoric water in the epoch II, because the mixing of metamorphic and meteoric fluids will make the δ¹⁸O_H2O values lower. If equilibrium calculations are made at the temperatures indicated by the arsenopyrite geothermometer, both the calculated δ¹⁸O_H2O values of the Qv1 type of samples (+5.5–+10.2‰, calculated with 417 °C) and the comparably similar ones of the Qv2 type of samples (+6.6–+9.3‰, calculated with 385 °C) (Figure 7b) are all closer to the field for most orogenic gold deposits suggested by [59], further suggesting a metamorphic origin with relatively high but variable δ¹⁸O_H2O values.

Fluid inclusions in 16 quartz separates within the quartz veins at the Heihaibei gold deposit have variable δD values varying from −104.2‰ to −81.1‰ (Table 2), which are apparently lower than those of most orogenic gold deposits suggested by [59], and they are partially plotted in the field of the devolatilization of organic matter in sediments (Figure 7a). However, no report shows organic matter in the sedimentary wall rock in the Heihaibei gold deposit. In some cases, bulk fluids extracted from quartz may be a mixture of primary, pseudosecondary, and secondary inclusions, then low δD values from the bulk extraction of fluid inclusions are often considered to reflect the secondary inclusions formed in the presence of meteoric water during the uplift of the deposits [65,66,67,68]. Therefore, we believe that the secondary inclusions formed in the presence of meteoric water during the uplift of the deposits contribute to the depletion of H isotope compositions from the bulk extraction of fluid inclusions within the quartz in the Heihaibei gold deposit. In this way, the potentially higher δD values may be closer to the δD values of the primary metamorphic ore-forming fluids (Figure 7b). Moreover, low δD values due to the presence of H₂ and/or CH₄ in the aqueous fluid [69] cannot be precluded either; unfortunately, laser Raman spectroscopy has not been carried out. Finally, meteoric water mixing is possible, in order to lower the δD values. Actually, the δD_VSMOW values of quartz in the Qv1 type samples are slightly higher (−97.2‰ to −81.1‰) than those of the Qv2 type samples (−104.2‰ to −89.9‰), suggesting mixing with minor meteoric water in epoch II. Overall, both the δD and δ¹⁸O values of the ore-forming fluids showed a slight decrease from epoch I (Qv1 ores) to epoch II (Qv2) (Figure 7b), further indicating mixing with minor meteoric water, especially in epoch II.

The Dazaohuo and Heicigou gold deposits in KLRA have relatively different H and O isotope compositions compared to each other [29], and both are totally different from those of the Heihaibei gold deposit (Figure 7a), suggesting a different source of ore-forming fluids for these gold deposits in the KLHA. However, other gold deposits, i.e., Shenshuitan [4,8,9,10], Guoluolongwa [19] in the MKL, and Kaihuangbei [14] in the SKL, in which there are both quartz veins and phyllic rock-type ores, have relatively variable H and O isotope compositions relative to gold deposits which have predominant quartz vein ores, such as the Heihaibei gold deposit (this study), as well as the Hongqigou gold deposit [60], perhaps due to higher degrees of water–rock reaction/mixing, and more meteoric water mixing during their mineralization. By the way, the H and O isotope compositions (Figure 7b) for sample K1 are slightly different from other ore samples, perhaps because sample K1 (Qv1 plus Heihaibei granitic wall rock) had experienced a greater extent of water–rock reaction than ore samples in the QV1 or QV2 veins.

5.2. Sulfur Sources

The pyrites from the gold-bearing quartz veins of the Heihaibei gold deposit have almost invariable δ³⁴S_VCDT values ranging from +7.6‰ to +8.7‰, with an average of +8.1‰ (Figure 8a), which span three previously published values of 7.7‰, 7.9‰, and 8.5‰ reported by [23]. Moreover, our sulfur isotopic compositions are within the published δ³⁴S values for most orogenic gold deposits elsewhere (typically δ³⁴S_VCDT = 0–9‰, [51,67,70,71]). Pyrites from the Qv1 type of samples in the study area have almost invariable δ³⁴S values ranging from +8.0‰ to +8.7‰ (Figure 8b), which are totally similar to those of pyrites from the Qv2 type samples ranging from +7.6‰ to +8.5 (Figure 8c). As for the large similarities among the sulfides in the Qv1 type samples and those in the Qv2 samples (Figure 8f), as well as the filling mineralization of quartz veins in the Heihaibei gold deposit, we argue that the sulfur in the ores may be derived directly from an ore-forming fluid that had only experienced a temperature decrease and mixing with minor meteoritic water, rather than the apparent sulfur isotope exchange with the wall rock from epoch I to epoch II. The presence of pyrite + arsenopyrite, and the weak alteration assemblage of quartz + sericite in the Heihaibei gold deposit, together with the lack of oxidized phases (hematite) and sulfate minerals, indicate that sulfur was present in the hydrothermal fluids, mainly as reduced sulfur (H₂S) [72,73]. The δ³⁴S_VCDT values for H₂S (δ³⁴S_H2S) in equilibrium with sulfides were also estimated by evaluating the δ³⁴S_VCDT values of pyrite using the equilibrium isotopic fractionation factor of sulfide for H₂S suggested by [73], and the average temperature of the hydrothermal fluid during the gold–sulfide mineralization event is inferred as 340 °C for Qv1 (epoch I), and 241 °C for Qv2 (epoch II) (based on the homogenization temperatures reported by [23]). The calculated δ³⁴S_H2S values for H₂S in the hydrothermal fluids ranging from +6.9‰ to +7.6‰, with an average value of +7.08‰ in epoch I, are also very similar to those in the hydrothermal fluids, ranging from +6.1‰ to +7.0‰ in epoch II (Table 3, Figure 8g). Although we did not analyze all types of sulfides in the Heihaibei gold deposit, pyrites dominate the sulfur mass balance, and such an evaluation for the δ³⁴S_H2S values of H₂S in the hydrothermal fluids using the sulfur isotopes of pyrites might be roughly reliable.

Previous researchers have proposed that the relatively high δ³⁴S_VCDT values of the ores were probably derived from the Heihaibei granites [23]. However, if the ore-forming fluids collected and exchanged sulfur from the neighboring Heihaibei granites, the ores in the Heihaibei gold deposit would possibly have more various δ³⁴S values rather than a narrow range, and different ores would have different δ³⁴S values because the sulfur isotope ranges of typical granite reservoirs have apparently various δ³⁴S values and could even range from −13.4‰ to +26.7‰ [74]. Unfortunately, we did not analyze δ³⁴S values of primary sulfides in the Heihaibei granites, so we are only making a rough guess here. We argue that the relatively high and invariable δ³⁴S values of sulfides for the Heihaibei gold deposit likely indicate a unique sulfur source directly from deep ore-forming fluids, rather than the neighboring granitic wall rock, because both Qv1 and Qv2 formed directly from primary ore-forming fluids, rather than from the apparent reaction between the ore-forming fluid and the granitic host rock during mineralization. Apparently, prior to upward migration, the deep ore-forming fluids leached sulfur, from the deep source rock in the upper crust, that was homogenized before mineralization.

In the KLHA, compared with the high and invariable δ³⁴S values of sulfides for the Heihaibei gold deposit, the Dazaohuo gold deposit has more variable δ³⁴S values of sulfides (Figure 8d, [29]), which are consistent with its phyllic rock-type ores, rather than the quartz veins. As we know, an apparent water–rock reaction which formed phyllic rock-type ores and related hydrothermal alterations possibly resulted in widespread sulfur mixing and exchange, as well as sulfur isotope fractionations in different ores. Therefore, other gold deposits which have predominant phyllic rock-type-ores, i.e., Shenshuitan [4,7,8,15,16], Guoluolongwa [15,16,17] in the MKL, and Kaihuangbei [20] in the SKL, likely have more variable δ³⁴S values of sulfides than those in the Heihaibei gold deposit (Figure 8d). Additionally, the Hongqigou gold deposit, which has predominant quartz vein ores, has relatively invariable heavy δ³⁴S values of sulfides that are similar to this study, indicating an original metamorphic S-bearing fluid [4]. So, once more, the relatively high and invariable δ³⁴S values of sulfides suggest that the sulfur source is derived directly from a deep metamorphic origin itself, which is consistent with the filling mineralization by the quartz vein, rather than an apparent water–rock reaction.

5.3. Lead Sources

Four pyrite minerals from quartz veins in the Heihaibei gold deposit have relatively invariable Pb isotope compositions (Figure 9a,b). Combined with three previously published sets of Pb isotope ratios ([23] in Figure 9), most of the Pb isotope compositions for sulfides in the Heihaibei gold deposit roughly cluster as a linear group (the blue contour lines in Figure 9a,b), close to the upper crust lead reservoir in the uranogenic plot (Figure 9), which might suggest two-component mixing between a high radiogenic end-member, and another low radiogenic end-member. Apparently, the fields for the Pb isotope compositions of sulfides in the Heihaibei gold deposit are totally different from the fields of the initial Pb isotopes (i.e., the Pb isotopes of K-feldspars) of the Heihaibei granites (the pink dashed lines in Figure 9a,b), and the ores (sulfides) show more radiogenic Pb isotopes and plot roughly along the upper crust growth line (Figure 9), indicating that Au-mineralization was not caused by the Heihaibei granites at ca. 416 Ma. In that way, a granitic intrusion-related gold deposit related to the Heihaibei granites proposed by [23] can be precluded for the Heihaibei gold deposit. Based on the below discussions, we argue that the gold mineralization might have occurred in the Triassic, which was likely around 200 Ma after the formation of the Heihaibei granites that host the veins. The opinion that Pb sources were derived from the neighbor Triassic granites in the north of Fault F51 within the MKL, as shown in Figure 2, might not be a good option, because the Pb isotope compositions of sulfides in the Heihaibei are also apparently different from the Pb isotope compositions of K-feldspars from some Triassic granites in the MKL (e.g., Triassic granites in the Wulonggou area within the MKL reported by [7]). Alternatively, apart from the K-feldspars, the rest of the minerals in the bulk Heihaibei granites will continue to evolve to higher ²⁰⁶Pb/²⁰⁴Pb, ²⁰⁷Pb/²⁰⁴Pb, and ²⁰⁸Pb/²⁰⁴Pb due to the U and Th decay for ca. 200 Ma, since they were intruded in the Early Devonian. Using the average U, Th, and Pb abundances of the Heihaibei granite [50] and the possible mineralization age (Triassic, ca. 200 Ma younger than the Early Devonian emplacement), the age (200 Ma)-corrected Pb isotopic ratios, roughly representing the Triassic Pb isotope compositions for the Early Devonian Heihaibei granitic wall rocks, can be estimated (initial Pb isotopes of K-feldspars, plus radiogenic Pb isotopes after ca. 200 Ma decay), and the results show that the age-corrected Triassic Pb isotope groups for the Devonian Heihaibei granites are partially localized within the ore Pb isotope group (Figure 9), indicating that the high radiogenic Pb isotope end-member might be the Triassic Pb isotope group of the Devonian Heihaibei granites. Furthermore, the sulfides within the Sample K1 from [23] have apparently higher radiogenic Pb isotope compositions than other sulfides (Figure 9c,d), perhaps indicating more Pb sources from granitic wall rocks, because Sample K1 was collected in the mineralized granitic wall rock, which would experience more water–rock reactions than the filling ores in QV1 or QV2. Finally, the low radiogenic Pb isotope end-member should be the deep ore-forming fluid itself.

Meanwhile, the Pb isotope compositions of sulfides in the Heihaibei gold deposit are also different from the Dazaohuo gold deposit, which has predominant phyllic rock ores [29] in the KLRA, or other gold deposits in the EKO (Figure 9a,b), also showing they have different Pb sources among them. The variable Pb isotope compositions in the Heihaibei gold deposit have more uranogenic and thorogenic Pb isotope compositions than other gold deposits, which have predominant phyllic rock ores, i.e., Shenshuitan [4,7,9], Guoluolongwa [16,17] in the MKL, and Kaihuangbei [20] in the SKL, also suggesting that Pb sources are distinctive for the Heihaibei gold deposit. Possibly in the Heihaibei gold deposit, the primary ore-forming fluids would have washed through deep parts of the Heihaibei granites with high levels of U/Pb and Th/Pb in the upper crust before final gold mineralization, because the apparent exchanging of the Pb isotopes between the ore-forming fluid and the Heihaibei granitic wall rock is relatively difficult during the gold filling mineralization accompanying the weak water–rock reaction in the Triassic. Pb isotope compositions of sulfides are more variable than sulfur isotopes, likely because the Heihaibei granites commonly contain more variable radiogenic Pb isotopes, which were continuously increased from U and Th decay than from sulfur stable isotopes. Moreover, the Qv1 type of ore has slightly more uranogenic and thorogenic Pb isotopes than the Qv2 type (Figure 9c,d), likely indicating less of a mixing of the Triassic Pb source of the Heihaibei granites in epoch II than epoch I, because the low-temperature ore-forming fluid of the epoch II might have experienced less Pb isotope exchanging with the deep parts of the Heihaibei granites or the upper crust before gold mineralization.

5.4. Genetic Type and Possible Metallogenic Model

Genetic classification of the Heihaibei gold deposit remains controversial: a hydrothermal-vein-type gold deposit [22], a magmatic hydrothermal gold deposit [24], an orogenic gold deposit [25], or a granitic intrusion-related gold deposit related to the Heihaibei granites [23] have all been proposed. However, as mentioned above, the ore bodies in the Heihaibei gold deposit are usually delineated in quartz veins along subsidiary faults within the Heihaibei granites, and the Heihaibei granites (monzogranites and granodiorites) are wall rocks for gold mineralization. Furthermore, many of the geological and geochemical features of the Heihaibei gold deposit are similar to those of the orogenic gold deposits, as described by [51]. These features include: (1) the geodynamic setting of the Heihaibei gold deposit related with the Triassic evolution of the EKO; (2) the gold mineralization controlled and hosted by the subsidiary faults of Fault F51 (a segment of the M. KLF) within the Heihaibei area, as well as in the KLRA; (3) the major sulfide minerals (contents < 5%) represented by pyrite and arsenopyrite, with minor chalcopyrite, galena, sphalerite, tetrahedrite, and micro-native gold; (4) the weak alteration assemblage, including silicification, and sericitization, with minor chloritization and carbonatization, similar to some typical orogenic gold deposits; (5) the major ore-forming fluids showing H₂O-NaCl fluids with low salinity (3.69~16.63 equiv. wt% NaCl), and intermediate to high homogenization temperatures from 183 to 412 °C [23]; (7) the δD values (−104.2 to −81.1‰) and estimated δ¹⁸O values (+1.6 to +8.3‰, or even from +5.5 to +10.2‰) for ore-forming fluids, indicating a metamorphic water rather than a magmatic water; (8) the δ³⁴S values of sulfides from +7.6‰ to +8.7‰, suggesting a sulfur source that is directly from the deep ore-forming fluids themselves, with sulfur isotope homogenization; and (9) the Pb isotopic compositions of sulfides, with ²⁰⁶Pb/²⁰⁴Pb ratios of 18.7219–19.0007, ²⁰⁷Pb/²⁰⁴Pb ratios of 15.6959–15.7062, and ²⁰⁸Pb/²⁰⁴Pb ratios of 37.7359–38.8055, showing partial similarities to the age-corrected Triassic Pb isotope compositions of the Devonian Heihaibei granitic wall rocks, rather than to the initial Pb isotope compositions of the granite at its time of emplacement 416 Ma ago, indicating a mixed Pb source between the deep ore-forming fluids and the Triassic Pb isotope compositions of the deep Devonian Heihaibei granitic wall rock, and some affinities to upper crust lead reservoirs [62]. Based on the results described above, and the comparison with typical orogenic gold deposits, we propose that the Heihaibei gold deposit belongs to the orogenic gold category, as described by [51], rather than a granitic intrusion-related gold deposit proposed by [23].

To date, mineralization in the Heihaibei gold deposit, whose timing has not been determined directly by radiometric dating but must be later than the Early Devonian (416 Ma, [50]) of the monzogranite wall rocks. As mentioned above, the EKO has undergone at least two orogenies that are closely related to the consumption of the Neo-Proterozoic to Early Devonian Proto-Tethys Ocean, which lays along the M.KLF, and the Carboniferous to Late Triassic Paleo-Tethys Ocean, which lays along the S.KLF–AMF, respectively [3,31,37]. We have previously considered that the EKO during the Late Early Silurian–Middle Devonian (430–389 Ma) was in the post-collisional stage related to the closure of the Proto-Tethys Ocean [75], and that the A-type Heihaibei granites were exact products of post-collisional magmatic emplacement during the post-collisional stage [50]. The northward subduction of the Paleo-Tethys oceanic plate was not initiated until the Carboniferous [3,31,37]. From the Middle Devonian (389 Ma) to the Carboniferous, the tectonism and magmatism in the EKO would have been relatively rare, as well as the gold mineralization here. Actually, numerous magmatic emplacements which are related to the evolution (the subduction and subsequent collision) of the Paleo-Tethys Ocean between the SKL and BH–SG happened predominantly from the Late Permian to the Triassic [42,44,46,75,76]. Additionally, it was in the Early to Middle Triassic that the EKO was strongly affected by the northward subduction of the Paleo-Tethys oceanic plate, and it was in the Late Triassic that the EKO was dominated by the closure of the Paleo-Tethys oceans between the EKO and BH–SG [46]. Accompanying the Early to Middle Triassic subduction and the Late Triassic collision, most typical gold deposits within the EKO have predominant Triassic mineralization ages varying from 237 Ma to <215 Ma, e.g., Shenshuitan (zircon U-Pb dating, <215 Ma, [13]; sericite Ar-Ar dating, 237.0 Ma and 230.8 Ma, [8]), Shihuigou (sericite Ar-Ar dating, 236.5 Ma, [77]; zircon fission track thermochronology, 235–216 Ma [12]), Hongqigou (zircon fission track thermochronology, 223 Ma, [12]), Guoluolongwa (sericite Ar-Ar dating, 202.7 Ma for the late stage of mineralization [78]; and Re-Os dating, 375 Ma and 354 Ma for the early stage of mineralization [18]). With this in mind, we consider that a Triassic age of gold mineralization in the Heihaibei gold deposit may be reasonable.

We previously argued that the ancient metamorphic fluid, perhaps produced during the Precambrian metamorphism in the EKO, cannot explain the predominant Triassic orogenic gold mineralization found in the Wulonggou area in the EKO [4]. Even in the Late Triassic, the soft collision between EKO and BH–SG [33] cannot produce metamorphism of the deep rocks and subsequent metamorphic fluids at depth in the lower crust [4]. If this is the case, the slab devolatilization model proposed for the Jiaodong gold province [79] may be also applicable to the Heihaibei gold deposit, as with the Shenshuitan gold deposit (e.g., [4]). In their model, the late-orogenic metamorphic devolatilization of stalled subduction slabs and oceanic sediments may directly produce metamorphic fluids, and the overpressured slab-derived metamorphic fluids might also be channeled directly into crustal-scale faults with subsequent upward flow [63,79]. Based on the above discussion, the Heihaibei gold deposit may be associated with the subduction of the Paleo-Tethys oceanic plate in the Early to Middle Triassic, or even the collision between BH–SG and EKO in the Later Triassic. Therefore, during the Triassic in the EKO, the slab-derived metamorphic fluids may be initiated following the EKO Triassic geodynamic evolution, and they might have migrated through the first-order crustal-scale faults (i.e., M.KLF) upward into the subsidiary faults at higher levels (i.e., Fault F53) (Figure 10). During the migration of these deep slab-derived metamorphic fluids, sulfur, lead, and gold may be leached from various types of surrounding source rocks in the upper crust. According to the highly radiogenic Pb isotope ratios in this study, we likely infer that such slab-derived fluids may wash up high U/Pb and Th/Pb source rock, i.e., the deep part of the Heihaibei granites, in the upper crust, before upward gold mineralization; then, such slab-derived fluid actually becomes a special ore-forming fluid, since it migrates across such a “strongly leached zone” in the upper crust (Figure 10). Eventually, in the Heihaibei gold deposit, such evolved slab-derived metamorphic water with leached ore-forming materials (such as gold, sulfur, arsenic, iron, lead, copper, zinc, etc.) may continue to migrate upward, become a special ore-forming metamorphic fluid, and periodically deposit quartz veins within shallow subsidiary faults, likely because of a sudden change of physico-chemical conditions rather than strong water–rock reactions between ore-forming fluids and wall rocks. A similar filling mineralization of quartz veins in subsidiary faults could be found in other gold deposits, such as the Heicigou gold deposit [29] in the KLRA, and the Hongqigou gold deposit [18] in the MKL, etc. During the predominant filling of quartz veins in subsidiary faults in the Heihaibei gold deposit, minor water–rock interactions may also occur in the granitic wall rocks and contribute to weak hydrothermal alterations within brecciated granitic wall rocks, minor ores of stockwork quartz veinlets, or breccias in such altered granites. The late stage of tectonic movements due to the soft collision between EKO and BH–SG may deform previously formed quartz veins (Qv1 and Qv2) along fault belts in the Heihaibei gold deposit and form breccia ores, indicating that the deformation stage occurred. Given that there is a strong water–rock interaction between the ore-forming metamorphic water and wall rocks, important phyllic-type ores and strong hydrothermal alterations would have occurred, such as in other gold deposits, e.g., Dazaohuo [29], and the Shenshuitan gold deposit in the Wulonggou gold field [4,7,8,9,10]. Certainly, both the filling of quartz veins and phyllic rocks could be found simultaneously in certain other gold deposits in the EKO, such as the Guoluolongwa gold deposit [14,15,16,17,18,19] and the Kaihuangbei gold deposit [14,20], mostly because of various characteristics of the ore-forming faults.

6. Conclusions

From the H, O, S, and Pb isotope compositions of ores from the Heihaibei gold deposit in the KLRA reported in this study the following conclusions can be derived.

The measured δ¹⁸O_quartz values for the quartz from the auriferous quartz veins in the Heihaibei gold deposit range from +9.2‰ to +13.9‰, and the estimated δ¹⁸O values for ore-forming fluids range from +1.6‰ to +8.3 ‰, or even up to +5.5 to +10.2‰. The δD_H2O values for the fluid inclusions of the quartz range from -104.2‰ to -81.1‰. The results suggest that the ore-forming fluid is a special metamorphic water, perhaps somewhat modified by interaction with meteoric waters.

The δ³⁴S values of sulfide separates from auriferous quartz veins in the Heihaibei gold deposit range from +7.6‰ to +8.7‰, and the calculated δ³⁴S_H2S values for H₂S in the ore-forming fluids range from +6.1‰ to +7.6‰, suggesting that the sulfur sources may be directly derived from the deep ore-forming fluids themselves.

Six potassic feldspar separates from the Heihaibei granites have apparently invariable Pb isotope compositions, with ²⁰⁶Pb/²⁰⁴Pb ratios of 18.3532 to 19.4864, ²⁰⁷Pb/²⁰⁴Pb ratios of 15.6475 to 15.6812, and ²⁰⁸Pb/²⁰⁴Pb ratios of 37.1750 to 38.4598. Additionally, the Pb isotope compositions of sulfide separates from the auriferous quartz veins in the Heihaibei gold deposit range, from 118.7219 to 19.0007 for ²⁰⁶Pb/²⁰⁴Pb, 15.6959 to 15.7062 for ²⁰⁷Pb/²⁰⁴Pb, and 37.7359 to 38.8055 for ²⁰⁸Pb/²⁰⁴Pb, showing partial similarities to the age (200 Ma)-corrected Triassic Pb isotope compositions of the Devonian Heihaibei granitic wall rocks, rather than the initial Pb isotope compositions of the granites, indicating a mixed Pb source between the deep ore-forming fluids and the Triassic Pb isotope compositions of the deep parts of the Devonian Heihaibei, and showing some affinities to the upper crust lead reservoir.

The Heihaibei gold deposit may be classified as an orogenic gold deposit, which is associated with the subduction of the Paleo-Tethys oceanic plate in the Early to Middle Triassic, or even the collision between BH–SG and EKO in the Later Triassic. The slab-derived metamorphic fluids may be generated during the above Triassic geodynamic evolution and subsequently leached ore-forming materials from high U/Pb and Th/Pb source rock, i.e., the deep parts of the Heihaibei granites and their comagmatic granitoids in the upper crust. Then, they continuously migrated upward to the subsidiary faults in the KLRA or even the whole EKO, mixed with minor meteoric water, especially during the later epoch of gold mineralization, and periodically deposited quartz veins within shallower subsidiary faults in the Heihaibei gold deposit. Late stages of tectonic movements may deform previously formed quartz veins along fault belts and form breccia ores.