Dating Ore Deposit Using Garnet U – Pb Geochronology : Example from the Xinqiao Cu – S – Fe – Au Deposit , Eastern China

The large Xinqiao Cu–S–Fe–Au deposit in the Tongling ore district, Eastern China, is characterized by a large-scale stratiform orebody, in which garnet is widely distributed as the main gangue mineral associated with mineralization. Xinqiao garnet can be divided into early (Grt1) and late (Grt2) generations based on extensive back-scattered electron (BSE) imaging observations. Laser ablation (LA)-ICP-MS trace element and U–Pb isotope composition analyses indicate that uranium occurs homogeneously within the Xinqiao garnet, and Grt1 and Grt2 have weighted average 207Pb-corrected 206Pb/238U ages of 137.0 ± 7.8 Ma (Mean standard weighted deviation (MSWD) = 4.9) and 129.6 ± 7.1 Ma (MSWD = 1.6), respectively, similar to the zircon U–Pb age (139.6 ± 1.5 Ma) of the Jitou intrusion. These garnet U–Pb ages, combined with the low MnO content and various Y/Ho ratios, suggest that the Xinqiao garnet is likely to have a magmatic hydrothermal replacement origin associated with the Jitou stock. Based on previous studies of the Xinqiao deposit, we infer that the Xinqiao stratiform orebody may have formed from the Early Cretaceous magmatic hydrothermal fluids associated with the Jitou stock, and may have been generated by the Early Cretaceous tectono-thermal event in Eastern China.


Introduction
Skarn deposits are an abundant ore type and represent a globally important source of Cu, Fe, Pb, Zn, W, Ag, and Au [1,2].The direct and precise dating of skarn mineralization is commonly conducted by Re-Os of molybdenite [3,4], U-Pb of titanite [5,6], 40 Ar/ 39 Ar of micas [7] and K-feldspars [8,9], and Sm-Nd of calcite [10].However, due to the lack of suitable datable minerals, direct dating of skarn-type mineralization is still locally unsuccessful.
Garnet is a common mineral in skarn systems [11], and its distinct oscillatory chemical zonation can track the fluid-rock interaction history and provide a continuous physicochemical record of the hydrothermal evolution [12][13][14][15][16][17][18].Moreover, due to high Lu/Hf and Sm/Nd ratios, garnet can be directly dated by Lu-Hf and Sm-Nd dating [19].However, the commonly low Lu and Hf contents Minerals 2018, 8, 31 2 of 18 in garnet constrain the application of Lu-Hf dating [20].Garnet Sm-Nd isochron dating is severely complicated because of variable Nd initial isotope ratios [21].On the other hand, grandite (the most common garnet type in skarn systems) commonly contains variable U and Th coupled with insignificant common Pb [22][23][24], enabling direct U-Pb dating of garnet to be a new method which has been recently applied in some skarn deposits [22,25].
The Tongling ore district is located in the Middle-Lower Yangtze River Valley metallogenic belt along the northern margin of the Yangtze Craton in Eastern China (Figure 1; [26]).It is characterized by widespread Early Cretaceous granitoids and numerous skarn deposits [26].However, the large-scale stratiform orebodies in the Xinqiao Cu-S-Fe-Au deposit and the Dongguashan Cu-Au deposit (Figure 2; [27]) in this region are thought to be obviously different from typical skarn deposits.Due to a lack of convincing mineralization age data, the genesis of these stratiform orebodies has been controversial over the past forty years, and the hypotheses focus on (1) Late Paleozoic submarine exhalative mineralization [28][29][30][31]; (2) Early Cretaceous magmatic hydrothermal mineralization [26,[32][33][34][35]; and (3) Early Cretaceous magmatism overprinting on Late Paleozoic submarine exhalative processes [36][37][38][39].Extensive skarn alteration in these stratiform orebodies powerfully supports the magmatic hydrothermal hypothesis [32].However, previous studies have documented that garnet can form via not only magmatic hydrothermal replacement but also magmatism and submarine sedimentary exhalative processes [2,[40][41][42].The two-stage Xinqiao grandite (And 100 to And 50 Gro 46 ), which is suggested to have formed by magmatic hydrothermal replacement, was discovered based on their major and trace element geochemical characteristics [18]; however, no robust absolute age data support this hypothesis.garnet constrain the application of Lu-Hf dating [20].Garnet Sm-Nd isochron dating is severely complicated because of variable Nd initial isotope ratios [21].On the other hand, grandite (the most common garnet type in skarn systems) commonly contains variable U and Th coupled with insignificant common Pb [22][23][24], enabling direct U-Pb dating of garnet to be a new method which has been recently applied in some skarn deposits [22,25].
The Tongling ore district is located in the Middle-Lower Yangtze River Valley metallogenic belt along the northern margin of the Yangtze Craton in Eastern China (Figure 1; [26]).It is characterized by widespread Early Cretaceous granitoids and numerous skarn deposits [26].However, the largescale stratiform orebodies in the Xinqiao Cu-S-Fe-Au deposit and the Dongguashan Cu-Au deposit (Figure 2; [27]) in this region are thought to be obviously different from typical skarn deposits.Due to a lack of convincing mineralization age data, the genesis of these stratiform orebodies has been controversial over the past forty years, and the hypotheses focus on (1) Late Paleozoic submarine exhalative mineralization [28][29][30][31]; (2) Early Cretaceous magmatic hydrothermal mineralization [26,[32][33][34][35]; and (3) Early Cretaceous magmatism overprinting on Late Paleozoic submarine exhalative processes [36][37][38][39].Extensive skarn alteration in these stratiform orebodies powerfully supports the magmatic hydrothermal hypothesis [32].However, previous studies have documented that garnet can form via not only magmatic hydrothermal replacement but also magmatism and submarine sedimentary exhalative processes [2,[40][41][42].The two-stage Xinqiao grandite (And100 to And50Gro46), which is suggested to have formed by magmatic hydrothermal replacement, was discovered based on their major and trace element geochemical characteristics [18]; however, no robust absolute age data support this hypothesis.This study focuses on the garnet from the Xinqiao Cu-S-Fe-Au deposit (0.5 Mt Cu @ 0.71%, 75.5 Mt of sulfur @ 29.3%, 24.9 Mt Fe @ 46%, and 11.2 t Au @ 4.7 g/t; [43]) with the aim of constraining its mineralization age by using newly developed garnet U-Pb geochronology based on detailed field and mineralogical observations, discusses its implications on the Xinqiao ore genesis, and further introduces a new dating method for the skarn ore deposit type.[27]).
This study focuses on the garnet from the Xinqiao Cu-S-Fe-Au deposit (0.5 Mt Cu @ 0.71%, 75.5 Mt of sulfur @ 29.3%, 24.9 Mt Fe @ 46%, and 11.2 t Au @ 4.7 g/t; [43]) with the aim of constraining its mineralization age by using newly developed garnet U-Pb geochronology based on detailed field and mineralogical observations, discusses its implications on the Xinqiao ore genesis, and further introduces a new dating method for the skarn ore deposit type.

Regional Geology
The Tongling ore district is the largest Cu-Au-Fe-Mo ore district in the Middle-Lower Yangtze River Valley metallogenic belt and hosts numerous skarn deposits associated with the Early Cretaceous tectono-thermal event (namely, widespread tectonics and magmatic activities generated from the subduction of the paleo-Pacific plate and the continental crust delamination of the Yangtze Craton) in Eastern China (Figures 1 and 2; [33,44]).More than 50 ore deposits have been discovered in the Tongling district, and are clustered in the Tongguanshan, Shizishan, Xinqiao, Fenghuangshan, and Shatanjiao orefields from east to west (Figure 2; [45]).Sedimentary rocks exposed in the area have ages spanning from the Silurian to the Cretaceous, excluding the Middle-Late Devonian, with dominant lithologies of sandy conglomerate, pyroclastic rocks, (quartz) sandstone, and limestone [27,46,47].The Lower Permian and the Lower Triassic systems are dominated by limestone and are important ore-hosting rocks of the widespread skarn mineralization in this district.The unconformity between the Upper Devonian System and the Upper Carboniferous System hosts the economically significant stratiform mineralization in this district, such as the Xinqiao and Dongguashan deposits.Structurally, the region contains NE-trending folds and NNE-and NW-trending faults which controlled the emplacement of intrusions.More than 70 igneous intrusions are widely distributed in the region, and dominated by high-K calc-alkaline granite porphyry, (quartz) diorite porphyry, and quartz monzodiorite with ages from 156 ± 2 Ma (muscovite 40 Ar/ 39 Ar dating) to 137 ± 1 Ma (laser ablation (LA)-ICP-MS zircon U-Pb dating), occurring as composite stocks, dikes, and sills (Figure 2; [26,[47][48][49]).

Regional Geology
The Tongling ore district is the largest Cu-Au-Fe-Mo ore district in the Middle-Lower Yangtze River Valley metallogenic belt and hosts numerous skarn deposits associated with the Early Cretaceous tectono-thermal event (namely, widespread tectonics and magmatic activities generated from the subduction of the paleo-Pacific plate and the continental crust delamination of the Yangtze Craton) in Eastern China (Figures 1 and 2; [33,44]).More than 50 ore deposits have been discovered in the Tongling district, and are clustered in the Tongguanshan, Shizishan, Xinqiao, Fenghuangshan, and Shatanjiao orefields from east to west (Figure 2; [45]).Sedimentary rocks exposed in the area have ages spanning from the Silurian to the Cretaceous, excluding the Middle-Late Devonian, with dominant lithologies of sandy conglomerate, pyroclastic rocks, (quartz) sandstone, and limestone [27,46,47].The Lower Permian and the Lower Triassic systems are dominated by limestone and are important ore-hosting rocks of the widespread skarn mineralization in this district.The unconformity between the Upper Devonian System and the Upper Carboniferous System hosts the economically significant stratiform mineralization in this district, such as the Xinqiao and Dongguashan deposits.Structurally, the region contains NE-trending folds and NNE-and NW-trending faults which controlled the emplacement of intrusions.More than 70 igneous intrusions are widely distributed in the region, and dominated by high-K calc-alkaline granite porphyry, (quartz) diorite porphyry, and quartz monzodiorite with ages from 156 ± 2 Ma (muscovite 40 Ar/ 39 Ar dating) to 137 ± 1 Ma (laser ablation (LA)-ICP-MS zircon U-Pb dating), occurring as composite stocks, dikes, and sills (Figure 2; [26,[47][48][49]).

Ore Deposit Geology
The sedimentary rocks that crop out in the mining area range from Middle-Upper Silurian sandstone to Upper Triassic sandy conglomerate, whereas the Lower Carboniferous unit is absent.The NE-trending Dachengshan anticline and the NNE-trending Shenchong syncline are the major structures, and the junction of their hinges is an advantageous location of magmatic rocks and orebody (Figure 3a; [44]).The Jitou stock, as the dominant igneous rock, is a multiphase intrusion with quartz diorite at the center and diorite porphyry along the margin.The quartz diorite was dated to 139.6 ± 1.5 Ma by LA-ICP-MS zircon U-Pb geochronology [50]).

Ore Deposit Geology
The sedimentary rocks that crop out in the mining area range from Middle-Upper Silurian sandstone to Upper Triassic sandy conglomerate, whereas the Lower Carboniferous unit is absent.The NE-trending Dachengshan anticline and the NNE-trending Shenchong syncline are the major structures, and the junction of their hinges is an advantageous location of magmatic rocks and orebody (Figure 3a; [44]).The Jitou stock, as the dominant igneous rock, is a multiphase intrusion with quartz diorite at the center and diorite porphyry along the margin.The quartz diorite was dated to 139.6 ± 1.5 Ma by LA-ICP-MS zircon U-Pb geochronology [50]).The major mineralization type is the economically significant stratiform mineralization (accounting for 90% of the Cu, S, and Fe reserves) which is confined along the unconformity between the Upper Devonian quartz sandstone and the Upper Carboniferous limestone (Figure 3b; [51]).The major stratiform orebody is NW-dipping and has length and width of 2560 m and 1810 m, respectively, with an average thickness of 21 m (Figure 3b; [51]).In the footwall of the stratiform orebody, the quartz-pyrite stockwork occurs in the Upper Devonian quartz sandstone.Field geologic and petrographic observations indicate that ore minerals in the stratiform orebody include magnetite, chalcopyrite, pyrite, pyrrhotite, and hematite, whereas gangue minerals include primarily garnet, diopside, wollastonite, epidote, chlorite, quartz, and calcite.Detailed field investigation indicates that the wallrock alteration products of the hanging wall of the stratiform orebody include garnet, sericite, quartz, chlorite, and kaolinite, with silicic alteration developed in the footwall of the orebody [18,35,50,52].Based on mineral assemblages and textural relationships, the Xinqiao mineralization The major mineralization type is the economically significant stratiform mineralization (accounting for 90% of the Cu, S, and Fe reserves) which is confined along the unconformity between the Upper Devonian quartz sandstone and the Upper Carboniferous limestone (Figure 3b; [51]).The major stratiform orebody is NW-dipping and has length and width of 2560 m and 1810 m, respectively, with an average thickness of 21 m (Figure 3b; [51]).In the footwall of the stratiform orebody, the quartz-pyrite stockwork occurs in the Upper Devonian quartz sandstone.Field geologic and petrographic observations indicate that ore minerals in the stratiform orebody include magnetite, chalcopyrite, pyrite, pyrrhotite, and hematite, whereas gangue minerals include primarily garnet, diopside, wollastonite, epidote, chlorite, quartz, and calcite.Detailed field investigation indicates that the wallrock alteration products of the hanging wall of the stratiform orebody include garnet, sericite, quartz, chlorite, and kaolinite, with silicic alteration developed in the footwall of the orebody [18,35,50,52].Based on mineral assemblages and textural relationships, the Xinqiao mineralization was divided into five stages [18,35]

Sample Descriptions
Two garnet skarn samples were collected from the hanging wall of the stratiform orebody in the southwest part of the 13 m platform at the Xinqiao open pit (XQ39-1; Figure 4a) and W401 stope at −300 m depth (XQ37-1; Figure 4b).The garnet in sample XQ39-1 is beige to dark green in color, coexisting with coarse-grained wollastonite and crosscut by late-stage quartz veins (Figure 4a).Sample XQ37-1 is dominated by beige garnet coexisting with trace pyroxene and wollastonite (Figure 4b).The skarn minerals in Sample XQ37-1 are locally replaced by late irregular quartz and calcite (Figure 4b).Under the microscope, the garnet grains are typically subhedral or anhedral granular, homogenous, and contain no distinct oscillatory zoning (Figure 4c,d).The late quartz and calcite occur among garnet, pyroxene, and wollastonite particles, and pervasively replace these minerals (Figure 4c,d).
In addition to quartz and calcite, late epidote also strongly replaces garnet, pyroxene, and wollastonite (Figure 4e-g), and occurs locally as pseudomorphs of garnet (Figure 4e).

Sample Descriptions
Two garnet skarn samples were collected from the hanging wall of the stratiform orebody in the southwest part of the 13 m platform at the Xinqiao open pit (XQ39-1; Figure 4a) and W401 stope at −300 m depth (XQ37-1; Figure 4b).The garnet in sample XQ39-1 is beige to dark green in color, coexisting with coarse-grained wollastonite and crosscut by late-stage quartz veins (Figure 4a).Sample XQ37-1 is dominated by beige garnet coexisting with trace pyroxene and wollastonite (Figure 4b).The skarn minerals in Sample XQ37-1 are locally replaced by late irregular quartz and calcite (Figure 4b).Under the microscope, the garnet grains are typically subhedral or anhedral granular, homogenous, and contain no distinct oscillatory zoning (Figures 4c,d).The late quartz and calcite occur among garnet, pyroxene, and wollastonite particles, and pervasively replace these minerals (Figures 4c,d).In addition to quartz and calcite, late epidote also strongly replaces garnet, pyroxene, and wollastonite (Figures 4e-g), and occurs locally as pseudomorphs of garnet (Figure 4e).Two types of garnet with different shades of gray (abundant dark gray Grt1 and local gray Grt2) were identified in samples XQ39-1 and XQ37-1 based on back-scattered electron (BSE) imaging observation.Neither have obvious chemical oscillatory zoning in BSE imaging (Figure 4h,i).Grt1 is anhedral granular and always occurs in the core of individual garnet grains (Figure 4h,i).It is locally resorbed by late irregular calcite and quartz (Figure 4h,i).Grt2 is anhedral granular, occurs around Grt1 (Figure 4h,i) and is locally resorbed by late irregular calcite (Figure 4i).Importantly, the contact boundary between Grt1 and Grt2 is irregular (Figure 4h), indicating replacement texture, rather than growth zoning.These two garnets, therefore, correspond to two different generations, i.e., the early generation (Grt1) and the late generation (Grt2), rather than the growth zoning of garnet normally documented in other studies [11,53].

Analytical Methods
Laser mount preparation and petrographic microscopy were conducted at the Key Laboratory of Mineralogy and Metallogeny, Guangzhou Institute of Geochemistry, Chinese Academy of Sciences, Guangzhou, China.Prior to LA-ICP-MS trace element and U-Pb isotope analysis, Electron Probe Microanalyzer (EPMA) analysis, mainly including in situ major element analysis and back-scattered electron (BSE) observation, was carried out in the School of Geosciences and Info-Physics of the Central South University, using a 1720 EPMA (Shimadzu Corporation, Tokyo, Japan).Analytical parameters were 15 Kv (acc.voltage), 2.0 × 10 −8 A (probe current), and 1 µm (spot size), Natural quartz was used as the standard to correct the SiO 2 concentration in garnet.Detection limits for the elements are below 0.01 wt %.Data were corrected using the internal ZAF (Z = Atomic number, A = Absorption and F = Fluorescence) correction program.
Subsequently, garnet U-Pb dating, trace element analysis, and mapping were performed using laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) at the Key Laboratory of Marine Resources and Coastal Engineering, Sun Yat-sen University (Guangzhou, China).The areas selected for U-Pb dating and trace element analysis were free of inclusions and fractures and approximate to EPMA analysis points.Laser ablation for garnet was performed using an ArF excimer laser ablation system (GeoLasPro; Microlas, Gottingen, Germany), and ion signal intensities were acquired using an Agilent 7700 × ICP-MS (Agilent, Santa Clara, CA, USA).A 44 µm spot was used with an energy density of 5 J/cm 2 and a repetition rate of 5 Hz.The trace element compositions of the garnet were calibrated against the National Institute of Standards and Technology Standard Reference Material 610, using the Si (Table 1) determined by electron microprobe as the internal standard.Zircon 91500 was used as the external standard to correct U-Pb isotopic ratios.Time-dependent drifts of U-Th-Pb isotopic ratios were corrected using a linear interpolation (with time) for every 10 analyses, based on the variations of Zircon 91500.The final uncertainties were propagated from uncertainties of the preferred and measured Zircon 91500 values, and from the measured sample values [54].Garnet QC04 was used as the secondary standard for monitoring the precision and accuracy of the U-Pb dating results.The obtained mean 206 Pb/ 238 U age for QC04 is 131 ± 2 Ma (2σ; MSWD = 0.86; n = 10), which is consistent with the recommended values (130 ± 2 Ma, [22]).Each analysis consisted of 20 s of background measurement (laser-off) followed by 45 s of data acquisition.Data reduction was performed using ICPMSDataCal software [54].Meanwhile, ISOPLOT 3.0 software [55] was used to construct the Tera-Wasserburg diagram and weighted mean calculations.

Trace Element Geochemistry
A total of 86 spot analyses for LA-ICP-MS trace element compositions were conducted on XQ37-1 and XQ39-1, including 37 spots in Grt1 grains and 49 spots in Grt2 grains.Trace element concentrations of the garnets are shown in Table 1.Grt1 and Grt2 have U contents ranging from 1.60 to 13.93 ppm (average: 4.95 ppm) and 3.58 to 17.81 ppm (average: 9.58 ppm).LA-ICP-MS trace element mapping data (Figure 5) reveal that the contents of Al, Fe, Y, and rare-earth elements are homogeneous within a specific garnet generation.Grt1 has higher concentrations of Al, Y, and rare-earth elements (except for La) than Grt2.Uranium is relatively homogeneous and enriched within Grt1, but is heterogeneously distributed in Grt2.

Trace Element Geochemistry
A total of 86 spot analyses for LA-ICP-MS trace element compositions were conducted on XQ37-1 and XQ39-1, including 37 spots in Grt1 grains and 49 spots in Grt2 grains.Trace element concentrations of the garnets are shown in Table 1.Grt1 and Grt2 have U contents ranging from 1.60 to 13.93 ppm (average: 4.95 ppm) and 3.58 to 17.81 ppm (average: 9.58 ppm).LA-ICP-MS trace element mapping data (Figure 5) reveal that the contents of Al, Fe, Y, and rare-earth elements are homogeneous within a specific garnet generation.Grt1 has higher concentrations of Al, Y, and rareearth elements (except for La) than Grt2.Uranium is relatively homogeneous and enriched within Grt1, but is heterogeneously distributed in Grt2.

U-Pb Dating
A total of 39 spot analyses for LA-ICP-MS U-Pb isotopic compositions were conducted on Sample XQ37-1, including 25 spots in Grt1 grains and 14 spots in Grt2 grains.U-Pb isotope data of the two-stage garnet are listed in Table 2.
Garnet commonly contains insignificant common Pb [23,24], and, thus, the Pb isotopic composition should be corrected prior to calculating ages.In this study, the 207 Pb-correction method was adopted [56,57].The uncorrected data of Grt1 are plotted in the Tera-Wasserburg diagram (Figure 6a), and a regression through these analyses yields a lower-intercept age of 138 ± 16 Ma (MSWD = 1.5;N = 25) with an upper intercept value of 0.9975, which represents the initial 207 Pb/ 206 Pb [57] and can be used in an algorithm to allow a form of 207 Pb-correction [56].Subsequently, the individual 207 Pb-corrected 206 Pb/ 238 U ages for Grt1 yield a weighted average age of 137.0 ± 7.8 Ma (MSWD = 4.9; Figure 6b).Similarly, the uncorrected data of Grt2 in the Tera-Wasserburg diagram (Figure 6c

U-Pb Dating
A total of 39 spot analyses for LA-ICP-MS U-Pb isotopic compositions were conducted on Sample XQ37-1, including 25 spots in Grt1 grains and 14 spots in Grt2 grains.U-Pb isotope data of the two-stage garnet are listed in Table 2.
Garnet commonly contains insignificant common Pb [23,24], and, thus, the Pb isotopic composition should be corrected prior to calculating ages.In this study, the 207 Pb-correction method was adopted [56,57].The uncorrected data of Grt1 are plotted in the Tera-Wasserburg diagram (Figure 6a), and a regression through these analyses yields a lower-intercept age of 138 ± 16 Ma (MSWD = 1.5;N = 25) with an upper intercept value of 0.9975, which represents the initial 207 Pb/ 206 Pb [57] and can be used in an algorithm to allow a form of 207 Pb-correction [56].Subsequently, the individual 207 Pb-corrected 206 Pb/ 238 U ages for Grt1 yield a weighted average age of 137.0 ± 7.8 Ma (MSWD = 4.9; Figure 6b).Similarly, the uncorrected data of Grt2 in the Tera-Wasserburg diagram (Figure 6c) yields a lower-intercept age of 131 ± 19 Ma (MSWD = 0.37; N = 14) with an upper intercept value of 0.9920, and the individual 207 Pb-corrected 206 Pb/ 238 U ages for Grt2 yield a weighted average age of 129.6 ± 7.1 Ma (MSWD = 1.6; Figure 6d).6. Discussion

Occurrence of Uranium in the Garnet
Previous studies concluded that uranium primarily occurs as U-rich mineral inclusions in garnet, thus limiting the application of garnet U-Pb dating [58,59].However, a relatively homogeneous distribution of U has been discovered in grandite from Nanminghe and Qicun iron skarn deposits in the North China Craton [22], indicating that U with high enough content for U-Pb dating can be hosted within the garnet structure.In this study, we found minor wollastonite, calcite, and quartz inclusions (Figure 4h,i), rather than U-rich mineral inclusions (monazite, zircon, titanite, and so on), in the grandite.LA-ICP-MS trace element mapping reveals not only the homogeneous U in Grt1 and locally enriched U in Grt2, but also the absence of chemical oscillatory zoning within Grt1 and Grt2 and the lack of mineral inclusions within Grt1.More importantly, the time-resolved signals of U, Al, Fe, Ca, Si, Ti, Y, and rare-earth elements (REEs), obtained by depth profile analyses of the observed two-stage garnets, are flat and stable (Figure 7), indicating the homogeneous distributions of these elements and possible absence of the U-rich mineral inclusions in grandite.This further suggests that U possibly occurs within the grandite structure.

Occurrence of Uranium in the Garnet
Previous studies concluded that uranium primarily occurs as U-rich mineral inclusions in garnet, thus limiting the application of garnet U-Pb dating [58,59].However, a relatively homogeneous distribution of U has been discovered in grandite from Nanminghe and Qicun iron skarn deposits in the North China Craton [22], indicating that U with high enough content for U-Pb dating can be hosted within the garnet structure.In this study, we found minor wollastonite, calcite, and quartz inclusions (Figures 4h,i), rather than U-rich mineral inclusions (monazite, zircon, titanite, and so on), in the grandite.LA-ICP-MS trace element mapping reveals not only the homogeneous U in Grt1 and locally enriched U in Grt2, but also the absence of chemical oscillatory zoning within Grt1 and Grt2 and the lack of mineral inclusions within Grt1.More importantly, the time-resolved signals of U, Al, Fe, Ca, Si, Ti, Y, and rare-earth elements (REEs), obtained by depth profile analyses of the observed two-stage garnets, are flat and stable (Figure 7), indicating the homogeneous distributions of these elements and possible absence of the U-rich mineral inclusions in grandite.This further suggests that U possibly occurs within the grandite structure.The radius of U 4+ is similar to those of heavy rare earth element (HREE), commonly found in an eight-fold coordination site within the garnet, and, thus, U and HREE contents within garnet commonly show a positive correlation [60].However, this positive correlation was not observed in the Xinqiao garnet (Figure 8a), indicating that the incorporation of U was not dominated by substitution mechanisms in the eight-fold coordination.On the other hand, the poor correlation between U and HREE also indicates that the primary U incorporation mechanism for the Xinqiao garnet is not surface sorption during rapid crystal growth of the garnet related to fluid infiltration, because surface sorption commonly causes U to correlate positively with both light rare earth element (LREE) and HREE in garnet [16,21].For grandite or andradite, the incorporation of U is primarily achieved by coupled substitution of Ca 2+ in the dodecahedral position on the basis of similar ionic radii and charge balance ( [11,16,60,61], which can be interpreted to be the U incorporation mechanism for the Xinqiao garnet.Uranium in both Grt1 and Grt2 has an obviously positive correlation with LREE (Figure 8b), especially La (Figure 8c), Ce (Figure 8d), Pr (Figure 8e), and Nd (Figure 8f), implying a similar substitution mechanism for LREEs in the Xinqiao garnet.Additionally, previous studies have documented that garnet has the distinct ability to fractionate REEs [11], which is evidenced in the Xinqiao garnet [18].The positive correlation between U and LREEs (Figure 8b-f) further indicates that U is hosted within the garnet structure.The radius of U 4+ is similar to those of heavy rare earth element (HREE), commonly found in an eight-fold coordination site within the garnet, and, thus, U and HREE contents within garnet commonly show a positive correlation [60].However, this positive correlation was not observed in the Xinqiao garnet (Figure 8a), indicating that the incorporation of U was not dominated by substitution mechanisms in the eight-fold coordination.On the other hand, the poor correlation between U and HREE also indicates that the primary U incorporation mechanism for the Xinqiao garnet is not surface sorption during rapid crystal growth of the garnet related to fluid infiltration, because surface sorption commonly causes U to correlate positively with both light rare earth element (LREE) and HREE in garnet [16,21].For grandite or andradite, the incorporation of U is primarily achieved by coupled substitution of Ca 2+ in the dodecahedral position on the basis of similar ionic radii and charge balance ( [11,16,60,61], which can be interpreted to be the U incorporation mechanism for the Xinqiao garnet.Uranium in both Grt1 and Grt2 has an obviously positive correlation with LREE (Figure 8b), especially La (Figure 8c), Ce (Figure 8d), Pr (Figure 8e), and Nd (Figure 8f), implying a similar substitution mechanism for LREEs in the Xinqiao garnet.Additionally, previous studies have documented that garnet has the distinct ability to fractionate REEs [11], which is evidenced in the Xinqiao garnet [18].The positive correlation between U and LREEs (Figure 8b-f) further indicates that U is hosted within the garnet structure.

Timing of the Garnets and Mineralization
In this study, garnet U-Pb dating of Grt1 and Grt2 yielded weighted average 206 Pb/ 238 U ages of 137.0 ± 7.8 Ma and 129.6 ± 7.1 Ma, respectively.These two ages are similar within uncertainty, indicating that the timing between Grt1 and Grt2 was relatively short, and also close to the zircon U-Pb age (139.6 ± 1.5 Ma; [50]) of the Jitou stock, the pyrite Re-Os isochron age (126 ± 11 Ma; [62]) of the stratiform orebody, and the quartz fluid inclusion Rb-Sr isotope isochron age (138.0 ± 2.3 Ma, [50]) of the footwall stockwork mineralization.Although the Re-Os isochron age (319 ± 13 Ma; [39]) of pyrite from the footwall stockwork mineralization may imply Late Paleozoic submarine exhalative origin for the Xinqiao deposit, the initial value of 187 Os/ 188 Os (0.017) is inconsistent with the initial 187 Os/ 188 Os (1-8; [63]) of typical sedimentary exhalative deposit (SEDEX) deposits.Furthermore, the 187 Os and 187 Re contents in pyrites vary in a fairly narrow range, which does not meet the conditions of Re-Os isochron age; thus, this Re-Os isochron age is most likely not the true age of the footwall stockwork mineralization [64].The combined geochronology shows that the stratiform mineralization formed in the Early Cretaceous (ca.138 Ma), consistent with the primary formation age range (144-135 Ma; [65]) of porphyry-skarn Cu-Fe-Au-Mo deposits in the Middle-Lower Yangtze River Valley metallogenic belt and the Jurassic-Cretaceous tectono-thermal event in Eastern China [34].

Timing of the Garnets and Mineralization
In this study, garnet U-Pb dating of Grt1 and Grt2 yielded weighted average 206 Pb/ 238 U ages of 137.0 ± 7.8 Ma and 129.6 ± 7.1 Ma, respectively.These two ages are similar within uncertainty, indicating that the timing between Grt1 and Grt2 was relatively short, and also close to the zircon U-Pb age (139.6 ± 1.5 Ma; [50]) of the Jitou stock, the pyrite Re-Os isochron age (126 ± 11 Ma; [62]) of the stratiform orebody, and the quartz fluid inclusion Rb-Sr isotope isochron age (138.0 ± 2.3 Ma, [50]) of the footwall stockwork mineralization.Although the Re-Os isochron age (319 ± 13 Ma; [39]) of pyrite from the footwall stockwork mineralization may imply Late Paleozoic submarine exhalative origin for the Xinqiao deposit, the initial value of 187 Os/ 188 Os (0.017) is inconsistent with the initial 187 Os/ 188 Os (1-8; [63]) of typical sedimentary exhalative deposit (SEDEX) deposits.Furthermore, the 187 Os and 187 Re contents in pyrites vary in a fairly narrow range, which does not meet the conditions of Re-Os isochron age; thus, this Re-Os isochron age is most likely not the true age of the footwall stockwork mineralization [64].The combined geochronology shows that the stratiform mineralization formed in the Early Cretaceous (ca.138 Ma), consistent with the primary formation age range (144-135 Ma; [65]) of porphyry-skarn Cu-Fe-Au-Mo deposits in the Middle-Lower Yangtze River Valley metallogenic belt and the Jurassic-Cretaceous tectono-thermal event in Eastern China [34].

Origin of the Garnet
As mentioned above, garnet can form via magmatic hydrothermal replacement, magmatism, and submarine sedimentary exhalative processes [2,[40][41][42].The two-stage garnet stages have weighted average 206 Pb/ 238 U ages of 137.0 ± 7.8 Ma and 129.6 ± 7.1 Ma, respectively, indicating that they are not linked to the late Paleozoic submarine exhalative processes.Moreover, garnet originating from submarine sedimentary exhalation is commonly almandine-and/or spessartine-rich due to the abundance of Mn and Fe on the seafloor [66,67].The Xinqiao garnets were formed from grossular-andradite solid solution with low MnO (0.19-0.89%; [18]).Melt and fluid-melt inclusions are suggested to be a direct indicator for magmatic garnets [68][69][70], yet they were not found in the garnets from the stratiform orebody [71].Moreover, magmatic processes are not expected to significantly fractionate Y from Ho [72]; hence, magmatic garnets would have Y/Ho close to the chondrite value of 28 [73].Grt1 and Grt2, however, have Y/Ho ratios ranging from 36.9 to 47.8 (38.6-47.8 for Grt1 from XQ37-1; 36.9-45.8 for Grt1 from XQ39-1) and 31.5 to 60.0 (36.4-60.0 for Grt2 from XQ37-1; 31.5-53.3for Grt2 from XQ39-1), respectively.Previous studies on the Xinqiao stratiform orebody, including the H−O isotope data obtained from the ore-bearing quartz, reveal the magmatic hydrothermal characteristics of the ore-forming fluid [74].Iron isotope compositions of pyrites from Xinqiao (δ 57 Fe ranges from −1.22‰ to 0.15‰; [75]) are similar to the pyrites in the skarn system (δ 57 Fe ranges from −2.58‰ to 1.62‰; [76]), and the initial 87 Sr/ 86 Sr value (0.71138 ± 0.00014; [50]) of the footwall stockwork mineralization is close to that of the Jitou stock.All this evidence indicates that the stratiform orebody may have been associated with the magmatic hydrothermal fluids derived from the Jitou stock, and generated by the Early Cretaceous tectono-thermal event in Eastern China.Therefore, we infer that the garnet hosted in the stratiform orebody is likely to be of magmatic hydrothermal replacement origin, associated with the Jitou stock, as confirmed by the similar ages of garnet and magmatism.

Conclusions
The major findings of this study can be summarized as follows: (1) Garnets from the Xinqiao stratiform orebody can be divided into early garnet (Grt1) and late garnet (Grt2) generations.Grt1 and Grt2 yielded weighted average 207

Figure 5 .
Figure 5. LA-ICP-MS trace element maps of the Xinqiao garnet; all maps show ppm abundances on a log scale.

Figure 5 .
Figure 5. LA-ICP-MS trace element maps of the Xinqiao garnet; all maps show ppm abundances on a log scale.

Figure 7 .
Figure 7. Representative time-resolved signals obtained by depth profile analyses of Grt1 (a) and Grt2 (b) from XQ37-1, showing flat and stable signals during laser ablation.

Figure 7 .
Figure 7. Representative time-resolved signals obtained by depth profile analyses of Grt1 (a) and Grt2 (b) from XQ37-1, showing flat and stable signals during laser ablation.

Table 2 .
LA-ICP-MS U-Pb isotopic data of the Xinqiao garnets.
Pb-correction206Pb/ 238 U ages of 137.0 ± 7.8 Ma (MSWD = 4.9) and 129.6 ± 7.1 Ma (MSWD = 1.6), respectively, close to the zircon U-Pb age of the Jitou stock, indicating that they formed in the Early Cretaceous.(2)The Xinqiao garnet and stratiform mineralization may have formed from Early Cretaceous magmatic-hydrothermal fluids associated with the Jitou stock, and were generated by the Early Cretaceous tectono-thermal event in Eastern China.