Introduction

In the eastern North China Craton (NCC), lode gold deposits are observed within uplifted Precambrian basement. These deposits are characterized by quartz vein and altered rock (hydrothermal altered rock accompanied by addition of disseminated pyrite and gold without development of quartz veins) types, and are spatially associated with the Late Mesozoic intrusive bodies (Zhang et al. 2003; Fan et al. 2003; Li et al. 2013; Zhu et al. 2015; Yang et al. 2016; Groves and Santosh 2016). The Liaodong gold province in NE China contains mesothermal lode deposits, similar to the Jiaodong gold province which is the most important gold producer in China. Few studies have been conducted on lode gold deposits within the Liaodong province, and the source of gold, timing of gold deposition, mechanism of gold precipitation, and overall classification of the deposits remain unclear. The Baiyun gold deposit in the central region of the Liaodong gold province provides an opportunity to study the fluid history and timing of gold deposition in the area. Previous studies focused on geological features, the timing and mode of ore formation, and thought that the gold mineralization was related to the Triassic magmatism (Liu and Ai 2000; Liu et al. 2012a; Zhang et al. 2016; Zhou 2017). However, previous experimental data on ore fluids were rather rare (Liu and Ai 1999; Hao et al. 2017). This paper describes the characteristics of the Baiyun gold deposit, focusing on the pressure–temperature evolution of ore fluids and the timing of gold mineralization. We applied multiple microanalytical techniques, such as scanning electron microscopy–based backscattered electron (SEM–BSE) imaging, Raman spectroscopy, microthermometry, and laser ablation–inductively coupled plasma–mass spectrometry (LA–ICP–MS). Our data are used to characterize the physical and chemical evolution of the ore fluids and to constrain the timing of gold deposition.

Geological setting

Regional geology

The Liaodong gold province is located at the northeastern margin of the NCC (Fig. 1a). The region underwent deformation and metamorphism during the Proterozoic, and extensive deformation and magmatism in the Mesozoic (Wu et al. 2005b). Regional lithostratigraphy is dominated by the Precambrian metamorphic rocks (Fig. 1b). The Archean Anshan complex, which is a series of 3800–3000 Ma tonalite–trondhjemite–granodiorite (TTG) gneisses (the oldest in North China) associated with 2500 Ma granitoids and supracrustal rocks, crops out in the north of the study area (Liu et al. 1992, 2008; Song et al. 1996; Wang et al. 2015a). The Paleoproterozoic units in the region comprise the Liaohe Group and the Liaoji Granitoids, which were metamorphosed under greenschist- to lower-amphibolite-facies conditions at 1900 Ma (Luo et al. 2004; Lu et al. 2006; Xie et al. 2011; Meng et al. 2014; Wang et al. 2016, 2017). The NWW-trending Liaohe Group unconformably overlying Archean units is overlain by Neoproterozoic strata, and, from bottom to top, comprises the Langzishan, Dashiqiao, and Gaixian formations. Deposition of the Liaohe Group is constrained to 2200–1950 Ma by zircon U–Pb dating of sedimentary and volcanic rocks (Luo et al. 2004, 2008; Wan et al. 2006; Liu et al. 2012b; Hu et al. 2015). The Liaoji granitoids crop out over an area of 300 × 70 km2 in the central and southern sections of the Liaoji belt, and were emplaced mainly during the period of 2200–2140 Ma (Luo et al. 2004; Lu et al. 2006). The region also contains ~ 20,000 km2 of Phanerozoic intrusions that were emplaced primarily during the Mesozoic, especially the Jurassic and Cretaceous (Fig. 1b). Jurassic granites (180–153 Ma), comprising monzogranite and granodiorite, had undergone regional ductile deformation. In contrast, Cretaceous intrusions (131–117 Ma), which comprise diorite, granodiorite, monzogranite, syenogranite, and syenite, are typically undeformed and more widely distributed than the Jurassic granites (Wu et al. 2005a, b).

Fig. 1
figure 1

Regional geology of the study area. a Map of the China Craton and distribution of major gold deposits in several gold districts (after Zhu et al. 2015). b Geological map of Liaodong region showing the location of major deposits

Full size image

Gold deposits in the Liaodong gold province are concentrated in three orefields (the Qingchengzi, Wulong, and Maoling) (Fig. 1b). The gold deposits occur within auriferous quartz veins hosted by Mesozoic granites (e.g., Wulong), and within altered Paleoproterozoic metamorphic rocks (e.g., Baiyun, Xiaotongjiapuzi, Maoling, and Sidaogou) (Fig. 1b). The Baiyun gold deposit is located within the northern Qingchengzi orefield.

The Qingchengzi Pb–Zn–Au–Ag polymetallic orefield is located in the north of the Liaodong gold province (Fig. 1b). Reserves in the orefield are estimated be ~ 1.5 Mt of Pb and Zn, 2000 t of silver, and 100 t of gold (Yu et al. 2009). The orefield contains 5 gold deposits, 1 silver deposit, and 15 Pb–Zn deposits. The major gold occurrences include the Baiyun, Xiaotongjiapuzi, and Linjiasandaogou deposits (Fig. 2). The detailed geology characteristics have been described by Yu et al. (2009) and Duan et al. (2014, 2017).

Fig. 2
figure 2

Geological map of Qingchengzi district showing the location of major deposits (the square shows the location of Baiyun deposit)

Full size image

Geology of the Baiyun deposit

The strata in the Baiyun gold deposit are composed of the Paleoproterozoic Dashiqiao and Gaixian formations (Fig. 3). The Dashiqiao Formation comprises amphibolite, marble, and tremolite schist, whereas the Gaixian Formation comprises marble, granulite, and schist. Quartz porphyry, diorite porphyry, and lamprophyre dikes intrude the Paleoproterozoic metamorphic units (Fig. 3). The quartz porphyry dikes typically occur along E–W-trending faults and are generally confined to the hanging wall of the orebodies (Figs. S1a, b). The quartz porphyry contains 10 vol% subhedral to euhedral quartz phenocrysts (0.1–1.0 mm) in a fine-grained groundmass (0.02–0.04 mm) of quartz and plagioclase. Adjacent to ore veins, quartz porphyry is pyritized and sericitized (Figs. S1c, d). Unaltered microdiorite dikes striking NW–SE and E–W crosscut orebodies within the Baiyun deposit (Figs. S1e, f). These dikes post-date mineralization and comprise plagioclase, quartz, amphibole, and biotite.

Fig. 3
figure 3

Geological map of Baiyun gold deposit (after Liaoning Zhaojin Baiyun gold Mining Corporation). GF, Gaixian Formation; DF, Dashiqiao Formation

Full size image

The Baiyun deposit contains gold reserves of > 20 t, and more than 20 orebodies have been identified, with gold grades ranging from 1.20 to 42.0 g/t (mean = 5.86 g/t). The orebodies occur primarily within altered sillimanite–biotite schist and granulite of the Gaixian Formation. Most orebodies dip south at 30° to 40°. Two styles of gold mineralization are observed in the Baiyun deposit. The dominant type is altered rock ore (e.g., the No. 1, 2, and 11-4 veins) (Fig. 4a, b). These orebodies typically occur as veins (without quartz veins) and are spatially associated with faults that dip 30°–40° toward south. The minor mineralization occurs in auriferous quartz veins (e.g., the No. 60-2 vein) (Fig. 4c), and these veins show different types of quartz (Fig. 4d). Auriferous quartz veins are mainly controlled by faults dipping 30°–40° toward southeast. Characteristics of the major orebodies are listed in Table 1.

Fig. 4
figure 4

Field view photographs showing crosscutting relationships of the geologic bodies from the Baiyun deposit. a Underground photograph of the Baiyun deposit 11-4 vein on level 280 m. b Underground photograph of the Baiyun deposit 11-4 vein on level 280 m displaying alteration-type orebodies. c Underground photograph of the Baiyun deposit 60-2 vein on level 100 m displaying extensional veins. d Stage 2 quartz (Qz2)-pyrite (Py2) veins crosscutting the primary quartz (Qz1)-pyrite (Py1) vein. Qz, quartz; Py, pyrite

Full size image
Table 1 Features of the major orebodies from the Baiyun deposit
Full size table

Alteration mineralogy

Host rocks associated with gold mineralization were extensively altered by fluids, resulting in vein formation and gold precipitation. In schists far from the ore veins, relict primary metamorphic minerals are preserved (Fig. S2). Adjacent to ore veins, schists are pervasively sericitized (Ser1) (Figs. S2b–d), and alteration produced quartz (during silicification related to vein-stage Qz1) (Figs. S2a–c), muscovite (Ms1) (Fig. S2d), and chlorite (Chl1) (Fig. S2e). Evidence for minor carbonation is also observed (Figs. S2a, f). The altered schists contain 2–5 vol% pyrite and trace amounts of chalcopyrite, mainly as blebs within pyrite grains. Pyrite grains are typically disseminated, or locally form polycrystalline masses or bands.

Mineral paragenesis and mineralization stages

In the Baiyun gold deposit, the dominant hydrothermal minerals are quartz, sericite, and pyrite, with lesser amounts of calcite, barite, chalcopyrite, gold, and hessite (Figs. 4 and 5). Gold is present as electrum and is concentrated within quartz veins and altered rocks. Fine-grained electrum (5–70 μm) has an Au/Ag ratio from 67:33 to 75:25 (at.%). Three mineralization stages have been identified in the Baiyun deposit, based on crosscutting relationships, and mineralogical and textural characteristics (Fig. 6): pyrite–quartz (stage 1), pyrite–quartz–chalcopyrite (stage 2), and quartz–carbonate (stage 3).

Fig. 5
figure 5

Photomicrographs and BSE images showing textural relationships between electrum and other vein minerals from the Baiyun deposit. a Chalcopyrite occurring along primary pyrite (Py1) fractures. b Chalcopyrite and electrum occurring along primary pyrite (Py1) boundaries. c Electrum occurring along fractures of Py1, Qz2-Py2 vein crosscutting Py1. d Electrum occurring along Py1 boundaries and fractures or hosted by secondary quartz (Qz2). e Chalcopyrite-electrum veins crosscutting Py1 (BSE image). f Electrum occurring as inclusions within Py2 (BSE image). g Py2 containing inclusions of hessite (BSE image). h Hessite and galena occurring as wedge shaped in Py1 (BSE image). Ccp, chalcopyrite; Py, pyrite; Au, electrum; Qz, quartz; Hes, hessite; Gn, galena

Full size image
Fig. 6
figure 6

Paragenesis of gangue and ore minerals at the Baiyun deposit. Line thickness represents relative amount of the minerals

Full size image

Stage 1 mineralization is characterized by quartz–pyrite veins comprising coarse-grained quartz (Qz1) (94%) and pyrite (Py1) (6%). The thickness of these veins range from 0.2 to 2.0 m. Coarse (1–3 cm) crystals of quartz are milky. Pyrite 1 (Py1) is (0.2–5.0 cm), subhedral to euhedral, commonly fractured, and is typically disseminated within the vein (Fig. 4d). Some of electrum post-date Py1 as it is dominantly observed as veinlets along fractures in Py1 (Fig. 5c–e) and along Py1 boundaries (Fig. 5b). These gold-bearing veins cut altered schists along brittle fractures.

Stage 2 mineralization is characterized by quartz–pyrite–chalcopyrite veinlets and stockworks within stage 1 ore minerals, and consists of fine-grained quartz (Qz2), fine-grained pyrite (Py2), and chalcopyrite. The thickness of these veins ranges from 0.2 to 8.0 cm. Finer gray quartz grains (< 0.2 mm; Qz2) are observed within veins that crosscut Qz1 veins (Fig. 4d). Finer-grained pyrite (< 1 mm; Py2) occurs coexisting with Qz2 veins that crosscut Qz1 veins (Fig. 4d). Py2 is intergrown with Qz2, which may together represent a second hydrothermal event. Electrum occurs as inclusions in Py2 or is intergrown with Qz2 (Fig. 5f, d). In alteration-related ore bodies, hessite (Ag2Te) is present as inclusions within Py2 (Fig. 5g, h).

Stage 3 mineralization is characterized by quartz–carbonate veinlets that cut veins and alteration zone related to stages 1 and 2. Barite can also be observed in these veinlets. Gold is not associated with this stage of mineralization.

Samples and analytical methods

Microthermometry and Raman spectroscopy

We analyzed 67 samples of mineralized schist and quartz veins from the Baiyun deposit, which were prepared as polished thin sections, as well as 54 fluid inclusion plates. Fluid inclusion microthermometry was conducted using a Linkam THMSG 600 heating–freezing stage (− 198 to 600 °C) and a Leitz microscope at Fluid Inclusion Laboratories, China University of Geosciences, Beijing, China. Synthetic fluid inclusions were used to calibrate the stage to ensure the measurement accuracy. Most measurements were made using a heating rate of 0.2–0.5 °C/min. The temperatures of solid CO2 phase melting (TmCO2) and CO2 clathrate melting (Tmclath) were determined by temperature cycling (Roedder 1984; Diamond 2001); the heating rate at temperatures close to TmCO2 and Tmclath was set at 0.1–0.2 °C/min. The stage uncertainty is ± 0.1 °C for temperatures below 30 °C and ± 1 °C for homogenization temperatures. The salinity of H2O–Na2Cl and H2O–NaCl–CO2 fluid inclusions was determined as follows: WNaCl = 0.00 + 1.78Tmice − 0.0442Tmice2 + 0.000557Tmice3 (Hall et al. 1988) and WNaCl = 15.52022–1.02342Tmclath − 0.05286Tmclath2 (Roedder 1984). Fluid inclusions densities were estimated using the equation from Brown and Lamb (1989) by the Flincor software package (Brown 1989). The trapping pressures were estimated using the equation from Bowers and Helgeson (1983) by the Flincor software package.

For Laser Raman spectroscopy, we prepared double-polished thin sections. Analysis was performed at room temperature using a Jobin-Yvon Horiba LabRam HR confocal Raman microscope equipped with an 800-mm spectrograph at the Institute of Geology and Geophysics, Chinese Academy of Science (IGGCAS), Beijing, China.

Oxygen and hydrogen isotope analyses

Oxygen and hydrogen isotope analyses of quartz were performed at the Laboratory for Stable Isotope Geochemistry, IGGCAS. Samples were carefully handpicked under a binocular microscope after the samples had been crushed, cleaned, and sieved to 40 to 60 mesh, resulting in a separate of 99% pure quartz. Oxygen was liberated from quartz by reaction with BrF5 and converted to CO2 on a platinum-coated carbon rod. The δ18O determinations were made using a MAT-252 mass spectrometer. The δ18Oquartz values were corrected using in-house quartz standard GBW04409, with a value of + 11.11 ± 0.06‰ (Cao et al. 2014). Reproducibility for isotopically homogeneous pure quartz is about ± 0.2‰ (1σ). δ18Owater values of ore fluids were calculated using the equation for equilibrium isotope fractionation of oxygen between quartz and water (Clayton et al. 1972).

Analyses of hydrogen isotopic compositions of the fluid inclusions in quartz were carried out on the splits of the samples for oxygen isotope analyses. Weighted amounts of quartz were loaded into a quartz tubes which have been roasted at 800 °C and stored at 110 °C prior to use. Water was released by heating the samples to approximately 500 °C in an induction furnace. Samples were first degassed by heating under vacuum to 120 °C for 3 h. Water obtained by degassing at 500 °C was converted to hydrogen by reaction with heated zinc powder at 410 °C and the hydrogen was analyzed with a MAT-252 mass spectrometer. The δD water values were corrected using water standard GBW04402, with a value of − 64.8 ± 0.11‰ (Cao et al. 2014). The results are reported relative to V-SMOW with analytical uncertainties ± 2‰.

Zircon U–Pb analysis

Two samples were selected for zircon U-Pb analysis in the Baiyun gold deposit. Sample BY-52 is from a pre-ore quartz porphyry, located at the hanging wall of orebodies. Sample BY17-53 is from a post-ore microdiorite that intruded the orebodies. Zircon crystals were obtained using a combination of heavy liquid and magnetic separation techniques. Individual crystals were handpicked and embedded in epoxy resin, and then polished to expose the grain centers. Cathodoluminescence (CL) images were obtained using a JEOL 6510 electron microprobe at Beijing GeoAnalysis, Beijing, China. LA–ICP–MS zircon analyses were conducted at Sample Solution Analytical Technology Corporation, Wuhan, China, using an Agilent 7700 ICP–MS equipped with a GeolasPro 193-nm laser ablation system. A 24–32-μm spot size was used. Background signals were measured with the laser off for 20 s, followed by a data collection period with the laser firing for 50 s. Time-integrated signals, time-drift correction, and quantitative calibration were made using the ICPMSDataCal software package (Liu et al. 2010). Zircon 91500 and GJ-1 were used as standards and the standard silicate glass NIST 610 was used to optimize the instrument. The detailed analytical method has been described by Liu et al. (2010). U, Th, and Pb concentrations were calibrated using 29Si as an internal calibrant and NIST 610 as a reference material. Final interpretation of the analytical results was performed using ISOPLOT software (Ludwig 2003).

Results

Fluid inclusion petrography

Three types of fluid inclusion were observed in quartz and calcite at room temperature. Figure 7 shows the various fluid inclusion types developed in vein stages 1 through 3.

Fig. 7
figure 7

Photomicrographs showing petrographic characteristics of fluid inclusions. a Low C-type and W-type inclusions in Qz1. b W-type (Laq + Vaq) and C type (Laq + Vco2) inclusions in Qz1. c W-type inclusion in Qz2. d W-type inclusions in calcite. Laq, liquid aqueous phase; Lco2, liquid carbonic phase; Vco2, vapor carbonic phase; Vaq, vapor aqueous phase

Full size image

W-type (aqueous type) inclusions are two-phase (Laq + Vaq) H2O–NaCl ± CO2 inclusions hosted in Qz1, Qz2, and calcite (Fig. 7a–d). W-type inclusions hosted in quartz have variable morphologies, are 5–16 μm in size, and have V/L (volume/liquid) ratios of 0.1–0.4. W-type inclusions in calcite have oval or rectangular shapes, range in size from 4 to 8 μm, and have V/L ratios of 0.1–0.3.

Low C-type (low carbonic type) inclusions are three-phase (Laq + \( {\mathrm{L}}_{{\mathrm{CO}}_2} \) + \( {\mathrm{V}}_{{\mathrm{CO}}_2} \)) H2O–NaCl–CO2 inclusions hosted in Qz1 and Qz2 (Fig. 7a). The inclusions have rectangular, oval, or circular shapes; show no evidence of decrepitation or necking down; and range in size from 5 to 20 μm. Low C-type inclusions typically occur in clusters, but are also observed along linear features interpreted as healed fractures. The volume of the carbonic phase, 25–50 vol%, is relatively consistent in each assemblage. The proportion of \( {\mathrm{L}}_{{\mathrm{CO}}_2} \) to \( {\mathrm{V}}_{{\mathrm{CO}}_2} \) (and the corresponding CO2 homogenization temperature) is similar for all assemblages.

C-type (carbonic type) inclusions are two-phase (Laq +\( {\mathrm{V}}_{{\mathrm{CO}}_2} \)) CO2 ± CH4 ± N2 ± H2O inclusions hosted in Qz1 within stage 1 veins (Fig. 7b). These inclusions generally have oval or rectangular shapes and are 8–12 μm in size. The vapor phase typically occupies ~ 90 vol% and the Laq phase coats the inclusion walls (Fig. 7b).

Fluid inclusion study

Microthermometry results are listed in Table 2. Microthermometric data were obtained from low C-type and W-type inclusions. The C-type inclusions are not common. Microthermometric data were not collected as inclusions commonly decrepitated.

Table 2 Microthermometric data for fluid inclusions of the Baiyun deposit
Full size table

Stage 1

Most inclusions hosted in Qz1 are W-type liquid-rich, two-phase, aqueous inclusions. Inclusions homogenized to liquid between 230 and 289 °C (Fig. 8). Salinities range from 3.4 to 16.5 wt% NaCl eq.; a strong mode occurs between 7 and 15 wt% NaCl eq. (Fig. 8). Neither clathrate formation nor CO2 ice melting were observed, indicating that CO2 constitutes less than 4 mol% of the fluid (Hedenquist and Henley 1985). Low C-type three-phase, H2O–CO2 inclusions yielded TmCO2 from − 57.2 to − 56.6 °C, indicating the lack of other gases. Tmclath of 6.1–8.3 °C indicate low fluid salinities ranging from 3.4 to 7.5 wt% NaCl eq. CO2 phases homogenized to liquid at 28.8–30.3 °C, indicating a range of CO2 densities from 0.58 to 0.64 g/cm3. Total homogenization (ThTOT) to liquid occurred at 261–285 °C (Fig. 8).

Fig. 8
figure 8

Histograms of total homogenization temperature (Th) and salinities of fluid inclusions in different stages

Full size image

Stage 2

Qz2 contains dominantly W-type and rare low C-type fluid inclusions. W-type inclusions yielded ThTOT of 203°–259 °C (Fig. 8). Inclusions have salinities from 5.7 to 16.8 wt% NaCl eq. Low C-type inclusions yield TmCO2 between − 57.4 and − 56.8 °C, indicating CO2 was the only gas. Clathrate melting varied from 5.5 to 6.0 °C, indicating fluid salinities between 7.5 and 8.3 wt% NaCl eq. ThCO2 was between 26 and 30.2 °C, indicating that CO2 densities range from 0.82 to 0.89 g/cm3. Inclusions homogenized to liquid from 241 to 253 °C (Fig. 8).

Stage 3

W-type inclusions in calcite have salinities range from 0.2 to 14.5 wt% NaCl eq and exhibit two modes at 3.0–8.0 wt% NaCl eq and 9.0 to 15.0 wt% NaCl eq. W-type inclusions homogenized by vapor bubble disappearance at temperatures between 138 and 208 °C (Fig. 8).

Laser Raman spectroscopy of inclusion gases

Low C-type inclusions from stages 1 contain a CO2 vapor phase, without CH4 (Fig. 9a). C-type inclusions from stage 1 show well-defined CO2, CH4, and N2 peaks (Fig. 9c). W-type inclusions from stages 1 and 2 contain traces of CO2 and CH4 vapor phases (Fig. 9b–f).

Fig. 9
figure 9

Representative laser Raman spectra of fluid inclusions. a Carbonic phase in low C-type inclusions hosted in Qz1. b The H2O vapor phase in W-type inclusions hosted in Qz1. c Carbonic phase in C-type inclusions hosted in Qz1. d The H2O liquid phase in W-type inclusions hosted in Qz1. e The H2O vapor phase in W-type inclusions hosted in Qz2. f The H2O liquid phase in W-type inclusions hosted in Qz2

Full size image

Oxygen and hydrogen isotopes

Table 3 lists oxygen and hydrogen isotope ratios obtained from Qz1, and calculated δ18Owater values of ore fluids. The δ18Owater from stage 1 are close to the primary magmatic water (5.5 ‰ to 10 ‰; Taylor 1974). The δD from stage 1 also approach to that of the primary magmatic water (− 80 to − 40‰; Taylor 1974) and deviate from the range of the metamorphic water (− 65 to − 20‰; Taylor 1974).

Table 3 Oxygen and hydrogen isotope analyses in the Baiyun deposit
Full size table

Zircon U–Pb geochronology

CL images of zircons are shown in Fig. 10. Results of U–Pb zircon analyses are shown in Fig. 10 and listed in Table 4. Zircons selected from sample BY-52 range in size from 50 to 130 μm and are pale brown in color. Most grains are euhedral, and their aspect ratios vary from 2:1 to 3:1 (Fig. 10a). Nine euhedral zircon grains yield ages from 129 to 127 Ma (Table 4). These analyses form a coherent group and yield a weighted mean 206Pb/238U age of 127.8 ± 0.8 Ma (MSWD = 0.17; Fig. 10c).

Fig. 10
figure 10

CL images of zircons and zircon LA-ICP-MS diagrams for the dikes from the Baiyun deposit

Full size image
Table 4 LA-ICP-MS U-Pb zircon data for dykes from the Baiyun deposit
Full size table

Zircon grains selected from Sample BY17-53 range in size from 70 to 130 μm. Most grains are subhedral, and their aspect ratios vary from 2:1 to 4:1 (Fig. 10b). Twenty-three subhedral zircon grains yield ages ranging from 2454 to 123 Ma (Table 4). The inherited zircons show ages between 2454 and 202 Ma. These ages are consistent with the Liaohe Group and Triassic intrusions in this region (Lu et al. 2006). The remaining 12 analyses form a coherent group and yield a weighted mean 206Pb/238U age of 125.6 ± 1.3 Ma (MSWD = 0.5; Fig. 10d).

Discussion

Ore formation and gold deposition inferred from fluid inclusions

Fluid inclusion data can be used to constrain the pressure–temperature evolution of ore fluids during gold mineralization in lode-gold deposits (Wilde et al. 2001; Wang et al. 2015c; Chai et al. 2016; Neyedley et al. 2017). In the Baiyun deposit, three stages of hydrothermal minerals are identified, and they contain distinct fluid inclusion assemblages that can be used to trace the evolution of the ore fluid.

Stage 1 quartz veins contain W-type, low C-type, and minor C-type inclusions. The variable CO2:H2O ratios in coexisting inclusions (Fig. 7a, b) could have induced by mixing or un-mixing of two immiscible fluids (Anderson, et al. 1992) or through post-entrapment modifications. The lack of a continuum in degree of filling of the mixed CO2–H2O inclusions and absence of homogenization into vapor may rule out the immiscibility as a potential mechanism in the Baiyun deposit (Wang et al. 2015c; Chai et al. 2016; Neyedley et al. 2017). Absence of a vertical trend in THtot versus salinity correlation, which is typical for necking down and leakage during heating, discards the selective water loss via diffusion in measured inclusions. Mixing of fluids is suggested by a salinity change of aqueous fluid inclusions at a near-constant temperature (Lécuyer et al. 1999). Evidence of mixing is also stipulated by the occurrence of discrete clusters of relatively low-salinity low C-type inclusions close to moderate salinity aqueous inclusions. Two distinct fluids can be distinguished based on the fluid inclusion studies. The inferred ore fluid was a moderate salinity fluid that had a temperature of 230–290 °C and contained little CO2 and CH4. The other fluid was carbonic aqueous, moderate temperature, low salinity, and without CH4. We therefore infer that the formation of stage 1 quartz veins was due to fluid mixing of a moderate salinity, reduced ore fluid with a carbonic aqueous, and oxidized fluid.

Fluid mixing has been invoked to explain the deposition of gold in several gold deposits (Hofstra et al. 1991; Cline and Hofstra 2000). The presence of strong silicification and sericitization indicates that the hydrothermal fluids were slightly acidic (Mikucki 1998). In such fluids, gold is transported as gold bisulfide complexes [Au(HS)–2, Au(HS)0] (Seward 1973; Hayashi and Ohmoto 1991; Zezin et al. 2007). Fluid mixing with a second fluid having a relatively lower concentration of H2S would dilute the ore fluid, decreasing the H2S concentration (Arehart 1996). A significant decrease of H2S from ore fluids results in the destabilization of gold bisulfide complexes (Mikucki 1998). Furthermore, mixing with the oxidized fluid can result in a sharp decrease in bisulfide concentrations. The decrease in bisulfide concentrations drove Reactions (1) and (2) (Tombros et al. 2007) to the right side and resulted in the precipitation of hessite and native gold.

$$ \mathrm{HTe}\hbox{--} (aq)+2\mathrm{Ag}\left(\mathrm{HS}\right)\hbox{--} 2(aq)={\mathrm{Ag}}_2{\mathrm{Te}}_{(s)}+4\mathrm{HS}\hbox{--} (aq)+\mathrm{H}+(aq) $$
(1)
$$ {2\mathrm{H}}_2{\mathrm{O}}_{(l)}+4\mathrm{Au}\left(\mathrm{HS}\right)\hbox{--} 2(aq)={4\mathrm{Au}}_{(s)}+{\mathrm{O}}_{2(g)}+4\mathrm{H}+(aq)+8\mathrm{HS}\hbox{--} (aq) $$
(2)

Fluid inclusions from stage 2 have a wide range of salinities, indicating that fluid mixing is still the main process. The homogenization temperatures of fluid inclusions from stage 2 are lower than those from stage 1, illustrating that fluid cooling occurred during stage 2. However, the stability for gold bisulfide complexes change very little with temperature over the range 150–300 °C (Seward 1973; Hayashi and Ohmoto 1991). Therefore, we suggest that fluid mixing may be responsible for the deposition of the gold during stage 2.

During stage 3, the ore fluid was dominated by H2O–NaCl. W-type inclusions that were trapped during stage 3 yield relatively low homogenization temperatures and salinities, with trends toward lower values (Fig. 11). These data are consistent with mixing of the ore fluid with dilute meteoric water, which typically has a low salinity and temperature. In conclusion, fluid mixing might have triggered the formation of quartz–carbonate veins.

Fig. 11
figure 11

Temperature versus salinity plot of fluid inclusions showing fluid evolution at the Baiyun deposit

Full size image

Ore-forming temperature and pressure

For the fluid inclusions, the ThTOT serves only as minimum constrains on the entrapment temperature (Diamond 2001). Therefore, the measured ThTOT values yield the minimum ore-forming temperature. The minimum trapping temperatures estimated from the inclusions were 230–290 °C for stage 1 and 200–260 °C for stage 2 assemblages.

The trapping pressure conditions can be estimated by the trapping temperatures (Wang et al. 2015c; Chai et al. 2016; Lambert-Smith et al. 2016). We performed calculations in the H2O–CO2–NaCl systems and calculated using the FLINCOR software package (Brown 1989). The homogenization temperatures were used to substitute for trapping temperatures calculated for minimum trapping pressures. Stage 1 inclusions yield minimum trapping pressures of 58–139 MPa. The minimum trapping pressures during stage 2 were constrained to 24–68 MPa.

Source of ore fluids

The H–O isotope compositions of quartz from stage 1 are close to the primary magmatic water (Taylor 1974) (Fig. 12). This implies that ore fluid may be dominant by a magmatic fluid at the Baiyun deposit. The quartz samples from stage 3 plot away from the primary magmatic water field (Fig. 12). δD values of quartz samples from stage 3 are lighter and are depleted up to 20–30‰. Such a shift of δD values might reflect that (1) ore fluids reacted with δD-depleted organic matter (Polya et al. 2000) or (2) ore fluids mixed with low δD meteoric water (Jia et al. 2001). Hypothesis 1 can explain such a variation in hydrogen composition, but there is no organic-rich sediments reported in this orefield. Hypothesis 2 may be a more plausible explanation, because fluid inclusion studies show the characteristics that ore fluids mixed with meteoric water during stage 3.

Fig. 12
figure 12

Isotopic composition of oxygen and hydrogen at the Baiyun deposit. The isotopic fields for common geological waters are from Taylor 1974

Full size image

According to Wang et al. 2015b the calculated δ18Owater and δD values of ore fluids for the Xiaotongjiapuzi gold deposit range between 4 and 9‰ and − 89 to − 92 ‰, respectively. These values are consistent with those of ore fluids for the Baiyun deposit and indicate a magmatic source. Therefore, gold mineralization related to magmatic water may be common in the Qingchengzi orefield.

Constraints on the timing and geodynamic mechanisms of gold mineralization in Liaodong

The pre-ore porphyry and post-ore microdiorite yield weighted mean 206Pb/238U ages of 127.8 ± 0.8 Ma and 125.6 ± 1.3 Ma, respectively, indicating that the Baiyun gold deposit formed at ~ 126 Ma. Combining with the fluid inclusion study, we conclude that the Baiyun gold deposit is an intrusion-related vein gold deposit related to the Early Cretaceous magmatism. Wei et al. (2003) dated gold mineralization in the Wulong deposit of the Liaodong gold province at 120 Ma using Rb–Sr dating of pyrite in Au-bearing quartz veins. This result is consistent with a 125–121 Ma zircon SHRIMP U–Pb age of the host granitoids and a monzonitic dike that cuts through the orebodies (Wu et al. 2005a). These ages are similar to the age of gold mineralization in the Baiyun deposit. Therefore, there is a Cretaceous (126–120 Ma) gold mineralization in the Liaodong gold province, similar to that of the Jiaodong gold province.

Wu et al. (2005a) proposed that magmatic activity on the Liaodong Peninsula at 131–117 Ma was associated with extension of the Liaodong gold province and breakup of the NCC in the Early Cretaceous (Yang et al. 2008; Wu et al. 2014). This extension, thinning, and breakup of the lithospheric root likely resulted from subduction of the Pacific plate beneath the NCC (Zhu et al. 2012; Wu et al. 2014). In the Liaodong gold province, gold mineralization (126–120 Ma) was coeval with Early Cretaceous magmatism (131–117 Ma), suggesting that mineralization was related to the breakup of the NCC in the Late Mesozoic.

Dehydration of the subducted Pacific plate led to continuous enrichment in chalcophile elements (Cu, Au, Ag, and Te) of mantle beneath the NCC in the Late Mesozoic (Zhu et al. 2015). When the lithospheric root thinning and breakup, partial melting of enriched mantle would have produced voluminous hydrous, Au- and S-bearing basaltic magma. The mantle-derived magma ascended and served as an important source for ore fluids. Auriferous magmatic fluids would migrate along faults for a long-distance, precipitate ore materials in shallow depth.

Conclusions

  1. (1)

    The initial ore fluid at the Baiyun gold deposit was moderate salinities, reduced fluid with a moderate temperature. The ore fluid evolved to lower temperature and salinity due to mixing processes. Fluid mixing with meteoric water might result in the precipitation of gold and hessite.

  2. (2)

    H–O isotope compositions of quartz in stage 1 suggest that the ore fluids were magmatic hydrothermal fluids.

  3. (3)

    Zircon U–Pb ages of dikes associated with orebodies indicate that gold mineralization occurred at ~ 126 Ma. The Baiyun gold deposit is an intrusion-related vein gold deposit that formed during continental extension triggered by the breakup of the NCC in the Early Cretaceous.