Vein Formation and Reopening in a Cooling Yet Intermittently Pressurized Hydrothermal System: The Single-Intrusion Tongchang Porphyry Cu Deposit

: Porphyry deposits are the dominant sources of copper and major sources of several base and precious metals. They are commonly formed via the repeated emplacement of hydrous magmas and associated ﬂuid exsolution. As a result, mineralized hydrothermal veins may undergo multiple deposition and reopening processes that are not fully accounted for by existing ﬂuid models. The Tongchang porphyry Cu deposit is a rare example of being related to a single intrusion. The simplicity in intrusive history provides an ideal starting point for studying ﬂuid processes in more complex multi-intrusion porphyry systems. Detailed scanning electron microscope (SEM) cathodolumines-cence imaging (CL) revealed rich microtextures in quartz and anhydrite that point to a ﬂuid timeline encompassing early quartz deposition followed by ﬂuid-aided dynamic recrystallization, which was succeeded by an intermediate stage of quartz dissolution and subsequent deposition, and ended with a late stage of continuous quartz deposition, brecciation, and fracturing. Vein reopening is more common than expected. Fifteen out of seventeen examined vein samples contained quartz and/or anhydrite that was older or younger than the vein age deﬁned by vein sequences. Thermobarometry and solubility analysis suggests that the ﬂuid events occurred in a general cooling path (from 650 ◦ C to 250 ◦ C), interspersed with two episodes of ﬂuid pressurization. The ﬁrst episode occurred at high-T (>500 ◦ C), under lithostatic conditions alongside dynamic recrystallization, whereas the second one took place at a lower temperature (~400 ◦ C), under lithostatic to hydrostatic transition conditions. The main episode of chalcopyrite veining took place subsequent to the second overpressure episode at temperatures of 380–300 ◦ C. The results of this study reafﬁrm that thermal and hydraulic conditions are the main causative factors for vein reopening and growth in porphyry deposits.


Introduction
Porphyry deposits are the world's dominant sources of copper and molybdenum as well as important reservoirs of other base and precious metals [1].They are one of the areas of focus of critical raw material studies [2].The deposits are normally found along the convergent plate tectonic margins, formed via optimally-integrated magmatic and hydrothermal processes rooted in the subduction lithospheric mantle [3][4][5][6][7].Despite several decades of studies, many details still await to be revealed, particularly regarding the mechanism of vein growth, reopening, and sulfide deposition [8][9][10][11].
The most widely-used porphyry fluid model classifies hydrothermal veins as early A-type quartz vein (700-550 • C), transitional B-type molybdenite quartz vein (500-400 • C), and late D-type pyrite veins (350-250 • C) based on vein morphology, mineralogy, and chronology.However, this model assumes that the veins are formed in a single event that precipitates one generation of vein mineral.This assumption rarely holds for most porphyry deposits due to vein reopening, as revealed by several SEM-CL studies [12][13][14][15][16][17][18].
Erroneous interpretations on fluid source and conditions may result if they are based on bulk analyses of fluid inclusions, mineral chemistry, and isotopes of reopened veins.
Studies have shown that vein reopening in magmatic hydrothermal deposits may be linked to the number of igneous intrusions, which dictates the thermal and hydraulic properties of the system [8,19].Ore deposits with a single intrusion may experience less reopening compared to multi-intrusion ones due to monotonic cooling, and thus less complicated hydraulic fracturing.The Tongchang Cu deposit is genetically associated with a single granodiorite intrusion [20].This study presents a novel fluid model for the Tongchang Cu deposit by using integrated techniques of SEM-CL imaging [13,21,22], TitaniQ geothermobarometry [23][24][25][26][27], fluid inclusion microthermometry [13,[28][29][30], and mineral solubility analysis [31,32].

Regional Geology
Most porphyry deposits form in terrains overriding subduction zones, where crustal magma chambers are fed by arc magmas [4].Repetitive magma replenishment causes magma chambers to evolve chemically and mechanically, and drives the intrusion of multiple intrusive stocks [33].In contrast, porphyry deposits with a single intrusion are rarely reported.Their scarcity is likely linked to the unique tectonic histories of the host terrain.The Tongchang porphyry deposit is a mono-intrusion system located in the inland of the South China Block (SCB), over 500 km from contemporaneous subduction belts [34] (Figure 1a).Interestingly, the deposit is situated within a much older subduction zone that sutured the SCB in the Neoproterozoic [35][36][37].Despite the multiple reworking events in the Paleozoic and Mesozoic eras, geochemical and isotopic studies of the porphyry magmas have established connections with the Proterozoic subduction event [20,25,38,39].For the sake of relevance and clarity, we describe Neoproterozoic and Mesozoic geology below.
event that precipitates one generation of vein mineral.This assumption rarely holds for most porphyry deposits due to vein reopening, as revealed by several SEM-CL studies [12][13][14][15][16][17][18] .Erroneous interpretations on fluid source and conditions may result if they are based on bulk analyses of fluid inclusions, mineral chemistry, and isotopes of reopened veins.

Deposit Geology
The Tongchang deposit is one of three mineralizing centers in the Dexing Cu district (Figure 2).Along with Zhushahong and Fujiawu, the three deposits appear in the intersections of regional and local faults that delineate a NW-SE alignment.The three granodiorite porphyries are identical in mineralogy and rock chemistry [20,38], and emanated from a common batholith at depths confirmed by borehole drillings.Alteration patterns are clearly separated despite the fact that the intrusions are close to one another (Figure 2).The ore types are similar except for different Au and Mo contents.[19,38]).Dash lines denoted with L0-8 represents exploration lines; black pentagons represent the locations of drill hole, and the sampling localities are marked with small rectangles.

Wall Rock and Intruding Porphyries
Wall rocks of the Dexing district are phyllite and tuffaceous slate of the Shuangqiaoshan Group (Figure 3a,b).Biotite and actinolite hornfels are developed in the wall rocks surrounding the granodiorites (Figure 3c).The granodiorites are obliquely emplaced, and account for over 95 vol.% of igneous intrusions in this district.They show porphyritic texture in the peripheral and upper part of the intrusions (Figure 3d), and the center and lower portions grade into an equigranular texture (Figure 3e).Phenocrysts and groundmass have similar mineral composition to andesine, K-feldspar, quartz, hornblende, and biotite.Both the granodiorite and wall rocks are hydrothermally altered and mineralized (Figure 3f).LA-ICP-MS zircon U-Pb dating suggested that the granodiorite formed at 171 ± 3 Ma [20], [38], [63], contemporaneous with Cu-Mo mineralization (molybdenite Re-Os age of ~170 Ma [64]).The porphyry magmas are interpreted as a partial melt of Neoproterozoic subduction-modified lithosphere [20], [38].There is minor (<5 vol.%) post-ore diorite porphyry and aplite (Figure . 3g-i).The diorite truncates mineralized veins (Figure 3h,i), and has been dated at ca. 154 Ma by the zircon U-Pb method [65].[19,38]).Dash lines denoted with L0-8 represents exploration lines; black pentagons represent the locations of drill hole, and the sampling localities are marked with small rectangles.

Wall Rock and Intruding Porphyries
Wall rocks of the Dexing district are phyllite and tuffaceous slate of the Shuangqiaoshan Group (Figure 3a,b).Biotite and actinolite hornfels are developed in the wall rocks surrounding the granodiorites (Figure 3c).The granodiorites are obliquely emplaced, and account for over 95 vol.% of igneous intrusions in this district.They show porphyritic texture in the peripheral and upper part of the intrusions (Figure 3d), and the center and lower portions grade into an equigranular texture (Figure 3e).Phenocrysts and groundmass have similar mineral composition to andesine, K-feldspar, quartz, hornblende, and biotite.Both the granodiorite and wall rocks are hydrothermally altered and mineralized (Figure 3f).LA-ICP-MS zircon U-Pb dating suggested that the granodiorite formed at 171 ± 3 Ma [20,38,63], contemporaneous with Cu-Mo mineralization (molybdenite Re-Os age of ~170 Ma [64]).The porphyry magmas are interpreted as a partial melt of Neoproterozoic subduction-modified lithosphere [20,38].There is minor (<5 vol.%) post-ore diorite porphyry and aplite (Figure 3g-i).The diorite truncates mineralized veins (Figure 3h,i
The potassic alteration is preserved only in the deep parts of the porphyry shallow levels, potassic alteration is largely obscured by chlorite-sericite overpr ure 4a).In the potassic altered granodiorite, porphyritic textures are preserved mary minerals are widely replaced by secondary K-feldspar, biotite, and iro Propylitic alteration is pervasive in the peripheral rocks.Rock textures are larg fected with carbonation and hydration reactions.
The potassic alteration is preserved only in the deep parts of the porphyry stock.At shallow levels, potassic alteration is largely obscured by chlorite-sericite overprints (Figure 4a).In the potassic altered granodiorite, porphyritic textures are preserved, but primary minerals are widely replaced by secondary K-feldspar, biotite, and iron oxides.Propylitic alteration is pervasive in the peripheral rocks.Rock textures are largely unaffected with carbonation and hydration reactions.Moving away laterally from the granodiorite-wall rock contact, chlorite-sericite alterations overprint on potassic and propylitic alterations (Figure 4b,f).While rock textures are largely preserved, most previous minerals were converted to chlorite, sericite, chalcopyrite, rutile, quartz, and anhydrite (Figure 4d).
The spatial distribution of sericitic alterations is structurally controlled.It is mainly developed in the granodiorite-wall rock contacts and surrounding faults and fractures.It obliterated earlier rock textures and converted all mafic minerals and feldspars into muscovite, rutile, pyrite, and quartz (Figure 4c,e).

Hydrothermal Veinlets and Mineralization
Stockworks of veins and veinlets are widely developed alongside the hydrothermal alterations.A previous work recognized three vein types conformable to the typical A-B-D classification, as mentioned previously.[67] further distinguished four subtypes of Atype veins, seven subtypes of B-type veins, and three subtypes of D-type veins.The early A veins are mostly wavy quartz dominant veins with variable amounts of K-feldspar, biotite, and magnetite.The transitional B veins are mainly straight-walled chlorite and quartz veins with pyrite, chalcopyrite, molybdenite, and bornite, along the vein centerline and edges.The late D pyrite-dominant veins have variable amounts of chalcopyrite and quartz.A detailed vein petrography is provided in Section 4.1.
The district has a total resource of 1870 Mt ore at an average grade of 0.46% Cu.Individually, Tongchang has 5.2 Mt Cu at 0.45%, 0.128 Mt Mo at 0.01%, 215 t Au at 0.19 g/t and 1279 t Ag; Fujiawu has 2.57 Mt Cu at 0.5% and 0.168 Mt Mo at 0.03%; and Zhushahong has 0.6 Mt Cu at 0.42% [20].Over 85% of the ore is localized in the chloritesericite and sericitic alteration zones, displaying a hollow cylinder shape (Figure 1d).Copper minerals are predominantly chalcopyrite (>90%), with lesser bornite, tennantite, and chalcocite.The molybdenum is mainly in molybdenite.Gold is recovered as native gold and electrum [68].Sulfide occurrences are different in the chlorite-sericite and sericitic zones.In the former, copper sulfides and molybdenite are primarily contained in the B-type chalcopyrite veins and altered mafic minerals with high chalcopyrite to pyrite ratios (2-0.5).In the sericitic zones, copper sulfides occur primarily as inclusions in pyrite and occasionally as tennantite aggregates [64] with low chalcopyrite to pyrite ratios Moving away laterally from the granodiorite-wall rock contact, chlorite-sericite alterations overprint on potassic and propylitic alterations (Figure 4b,f).While rock textures are largely preserved, most previous minerals were converted to chlorite, sericite, chalcopyrite, rutile, quartz, and anhydrite (Figure 4d).
The spatial distribution of sericitic alterations is structurally controlled.It is mainly developed in the granodiorite-wall rock contacts and surrounding faults and fractures.It obliterated earlier rock textures and converted all mafic minerals and feldspars into muscovite, rutile, pyrite, and quartz (Figure 4c,e).

Hydrothermal Veinlets and Mineralization
Stockworks of veins and veinlets are widely developed alongside the hydrothermal alterations.A previous work recognized three vein types conformable to the typical A-B-D classification, as mentioned previously.Ref. [67] further distinguished four subtypes of A-type veins, seven subtypes of B-type veins, and three subtypes of D-type veins.The early A veins are mostly wavy quartz dominant veins with variable amounts of K-feldspar, biotite, and magnetite.The transitional B veins are mainly straight-walled chlorite and quartz veins with pyrite, chalcopyrite, molybdenite, and bornite, along the vein centerline and edges.The late D pyrite-dominant veins have variable amounts of chalcopyrite and quartz.A detailed vein petrography is provided in Section 4.1.
The district has a total resource of 1870 Mt ore at an average grade of 0.46% Cu.Individually, Tongchang has 5.2 Mt Cu at 0.45%, 0.128 Mt Mo at 0.01%, 215 t Au at 0.19 g/t and 1279 t Ag; Fujiawu has 2.57 Mt Cu at 0.5% and 0.168 Mt Mo at 0.03%; and Zhushahong has 0.6 Mt Cu at 0.42% [20].Over 85% of the ore is localized in the chlorite-sericite and sericitic alteration zones, displaying a hollow cylinder shape (Figure 1d).Copper minerals are predominantly chalcopyrite (>90%), with lesser bornite, tennantite, and chalcocite.The molybdenum is mainly in molybdenite.Gold is recovered as native gold and electrum [68].Sulfide occurrences are different in the chlorite-sericite and sericitic zones.In the former, copper sulfides and molybdenite are primarily contained in the B-type chalcopyrite veins and altered mafic minerals with high chalcopyrite to pyrite ratios (2-0.5).In the sericitic zones, copper sulfides occur primarily as inclusions in pyrite and occasionally as tennantite aggregates [64] with low chalcopyrite to pyrite ratios (<0.1).Molybdenite is contained in D-type pyrite veins and altered wall rocks.Native Au, electrum and Au-rich pyrite are primarily hosted in the sericitic zone.

Samples and Methods
Sixteen representative rock samples were selected to represent different alteration and vein types.The sampling localities are marked in Figure 2 and Table 1, which include mining platforms (20 and 230 m) and drill cores of the Tongchang open pit (SC212), as well as drill cores (ZK0-2, ZK4-4, ZK6-3, ZK8-4, ZK821).Detailed descriptions of sample petrography are given in the Appendix A.

Vein Petrography
Three subtypes of A-type veins have been selected: hematite-quartz (A.V hq ), K-feldspar-quartz (A.V kq ), and anhydrite-quartz (A.V aq ).In the abbreviation, A denotes A-type, and the subscripts denote vein minerals in increasing mounts from left to right.The same abbreviation rule is used for all vein types.

Vein Petrography
Three subtypes of A-type veins have been selected: hematite-quartz (A.Vhq), Kfeldspar-quartz (A.Vkq), and anhydrite-quartz (A.Vaq).In the abbreviation, A denotes Atype, and the subscripts denote vein minerals in increasing mounts from left to right.The same abbreviation rule is used for all vein types.

Analytical Methods
300 µm-thick sections were made by conventional cutting, grinding, and polishing.Mineral textures and parageneses were studied with petrographic microscopes and scanning electron microscopes (SEM) equipped with secondary electron (SE), backscattered electron (BSE), cathodoluminescence (CL), and energy dispersive spectroscopy (EDS) detectors.A total of 774 grey-scale CL, 579 color CL, 782 BSE high-resolution images, and 42 X-ray maps were taken.Trace element concentrations in different generations of quartz were measured in situ by LA-ICP-MS.Overall, 656 analyses for quartz from 16 samples were made.After all analyses, the actual ablation spots were confirmed by new CL imagery.

SEM-CL Microscopy
Greyscale CL imagery for anhydrite and quartz were firstly obtained on a Tescan Vega3 SEM equipped with a Centaurus CL detector and a photomultiplier at the Western Washington University.The accelerating voltage was 15 kV, and the probe current was about 10 nA for both quartz and anhydrite.The photomultiplier had a spectral range from ~300 nm to ~600 nm.A later CL study was carried out at the Institute of Geology and Geophysics, Chinese Academy of Sciences (IGGCAS), with a Nova NanoSEM 450 SEM equipped with a Gatan MonoCL4 panchromatic detector.An acceleration voltage of 15 kV and beam current of 9-12 nA were used.
Color CL images of quartz and anhydrite were collected using a Tescan Vega3 SEM equipped with a Gatan ChromaCL2 multichromatic detector at the GeoRessources Lab, CNRS (UMR7359), Université de Lorraine.An acceleration voltage of 15 kV and beam current of ~10 nA was used.The CL detector records spectra in ranges of red, green, and blue, and produces an RGB image.A dwell time of 300 µs was set for each pixel.During a separate session aiming to test the lifespan of luminescence, images were taken after the samples were exposed to electron beams for a certain amount of time.

In-Situ LA-ICP-MS Trace Elements
Trace elements in quartz were first measured with an Agilent 7500 ce ICP-MS coupled with a NewWave 213 nm Nd:YAG solid state laser ablation system at WWU.A laser fluence of 15 to 20 J/cm 2 , a repetition rate of 10 Hz, and a spot size of 65 µm were used.Helium was used as a carrier gas and mixed with make-up argon gas before entering the ICP torch.The Octupole Reaction System (ORS) was operated in hydrogen reaction mode (H 2 flow rate of 2 mL/min) to increase sensitivity.Oxide production was monitored with 132 Th 16 O/ 132 Th ratios, which were always less than 0.2%.Twenty-two masses were monitored: 7 Li, 9 Be, 11 B, 23 Na, 27 Al, 29 Si, 30 Si, 31 P, 39 K, 43 Ca, 49 Ti, 55 Mn, 56 Fe, 63 Cu, 71 Ga, 74 Ge, 75 As, 88 Sr, 118 Sn, 121 Sb, 137 Ba, and 208 Pb.The dwelling time for 49 Ti was 0.1 s, and it was 0.01-0.03s for the other masses.About 30 s of gas blank was collected as the baseline, which was followed by 60 s of data acquisition with laser ablation.Instrument performance was checked by a rough on-the-fly data reduction with the GLITTER program.Data were reduced with care later on using GLITTER. 30Si was used as the internal standard using stoichiometric values.NIST 610 glass was used as an external standard, with reference values from [69].A natural quartz standard (QZ7 [70]) and a synthetic quartz with a known chemical composition was used as quality controls.Results for QZ7 were within the range of recommended values for all elements except lower Al values.A correction factor was obtained by dividing the recommended value by measured values of Qz7 (ranging from 0.51 to 0.76, Table 4), which was applied to unknowns of the same analytical session.Additional analyses were made with an Agilent 7900 ICP-MS equipped with a MICRO/Las Pro 193 nm ArF excimer laser ablation system at the Institute of Geochemistry of the Chinese Academy of Sciences (IGCAS).A repetition rate of 10 Hz and an energy density of 10 J/cm 2 were used.Spot size of the laser beam was 44 µm.NIST 610 glass was used as an external standard and analyzed twice every 10 analyses.An internal standard-independent calibration strategy, implemented in the ICPMSDataCal program, was used for quantification [71].NIST 612, GSE-1G, and Qz7 were analyzed as quality controls.Analytical uncertainties of most elements for both setups were around 5-10% (1σ).
Although every effort was taken to avoid mineral and fluid inclusions, the influence of sub-microscopic inclusions was inevitable.During data reduction, abnormal spikes and bumps were filtered and excluded from integration.Even so, values of Na, K, Ca, Al, Mn, Fe, Cu, Zn, Sr, As, and Pb were too high to be real.Co-variations in Na-K-Al, Ca-Sr, Fe-Cu, Ca-Pb, K-Ca-Al, Cu-As, and Fe-Mn was interpreted to be the result of feldspar, carbonate, chalcopyrite/pyrite, tennantite, mica, and fluid inclusions, and were used as the basis to discard those values.

Fluid Inclusion Microthermometry
Fluid inclusions were classified as primary, secondary and pseudosecondary, following the principles of [72].Phase transition temperatures were measured on a Linkam THMS 600 cooling/heating stage at the GeoRessources Lab of the Université de Lorraine.The stage was calibrated against the melting point of solid CO 2 (−56.6 • C) and ice (0 • C), and a homogenization temperature (165 • C) of well-characterized natural fluid inclusions.The heating rate varied between 0.2 and 10 • C/min.When fluid inclusions approached the phase transition, the rate was lowered to 0.2-0.5 • C/min.Fluid salinity, density and isochores were calculated using the HokieFlincs_H 2 O-NaCl program [73].

SEM-CL Imaging
Quartz growth sequencing was based on CL microtextures and vein crosscutting relations.The earliest quartz generation (Qz1, Figure 6a), most common in the A.V hq , B.V cpq , and B.V macpq veins, was characterized by variable grain sizes and straight grain boundaries.The grains commonly show undulose extinction under polarized light.They exhibit mottled CL with faint zoning, and abundant microfractures that are filled with CL-dark "splatters" and "cobwebs" (Figure 6a).The second quartz generation of (Qz2), common in A.Vkq, A.Vaq, B.Vcpq, B.Vmacpq, D.Vpq, and D.Vpd comprises anhedral grains with contrast sizes and irregular boundaries filled with CL-dark quartz.Qz2 grains are further divided into larger Qz2a grains that have concentric CL zoning and smaller Qz2b grains that show mottled CL (Figure 6b).The contacts between Qz2a and Qz2b can be both transitional and crosscutting.They form discrete layers in most samples.They also contain "splatters" and "cobwebs".Under color CL, both grains show long-lived blue, with lower intensity for Qz2b (Figure 6c).Qz2 grains truncate Q1 locally, suggesting a later deposition.
The fourth generation of quartz (Qz4), mainly observed in the A.Vaq, B.Vcpq, and B.Vmacpq veins, is low in abundance.The Qz4 grains show dark, red CL and occur as thin rims or aggregates coating earlier quartz (Figure 6e,g,h).Qz4 crosscuts Qz3b grains locally.
The fifth generation of quartz (Qz5) commonly consists of a CL-dark core (Qz5a) and oscillatory overgrowth (Qz5b) (Figure 6i).They are observed in the D.Vmpq, D.Vpq, The second quartz generation of (Qz2), common in A.V kq , A.V aq , B.V cpq , B.V macpq , D.V pq , and D.V pd comprises anhedral grains with contrast sizes and irregular boundaries filled with CL-dark quartz.Qz2 grains are further divided into larger Qz2a grains that have concentric CL zoning and smaller Qz2b grains that show mottled CL (Figure 6b).The contacts between Qz2a and Qz2b can be both transitional and crosscutting.They form discrete layers in most samples.They also contain "splatters" and "cobwebs".Under color CL, both grains show long-lived blue, with lower intensity for Qz2b (Figure 6c).Qz2 grains truncate Q1 locally, suggesting a later deposition.
The fourth generation of quartz (Qz4), mainly observed in the A.V aq , B.V cpq , and B.V macpq veins, is low in abundance.The Qz4 grains show dark, red CL and occur as thin rims or aggregates coating earlier quartz (Figure 6e,g,h).Qz4 crosscuts Qz3b grains locally.
The fifth generation of quartz (Qz5) commonly consists of a CL-dark core (Qz5a) and oscillatory overgrowth (Qz5b) (Figure 6i).They are observed in the D.V mpq , D.V pq , and D.V pd veins, and are intimately associated with pyrite.Locally, they occur as dark bands crosscutting or filling the interstitial space of earlier quartz (Figure 6j).
The latest generation of quartz (Qz6) comprises a CL-dark to grey interior (Qz6a) and a very bright-CL exterior (Qz6b) (Figure 6k).They were observed in the A.V aq , D.V pq , and D.V pd .Qz6a exhibited an oscillation of red and blue CL, whereas Qz6b displayed a bright yellowish blue (Figure 6l).
Four generations of anhydrite have been identified.The first generation of anhydrite (Anh1) exhibited bright, homogeneous to patchy CL (Figure 7a).They are subdivided to Anh1a and Anh1b according to CL intensity.Anh1a grains have higher CL intensity and irregular boundaries (Figure 7a) relative to Anh1b grains that are commonly overgrown, recrystallized, or crosscut by later anhydrites.
Geosciences 2023, 13, x FOR PEER REVIEW 14 of 61 and D.Vpd veins, and are intimately associated with pyrite.Locally, they occur as dark bands crosscutting or filling the interstitial space of earlier quartz (Figure 6j).The latest generation of quartz (Qz6) comprises a CL-dark to grey interior (Qz6a) and a very bright-CL exterior (Qz6b) (Figure 6k).They were observed in the A.Vaq, D.Vpq, and D.Vpd.Qz6a exhibited an oscillation of red and blue CL, whereas Qz6b displayed a bright yellowish blue (Figure 6l).
Four generations of anhydrite have been identified.The first generation of anhydrite (Anh1) exhibited bright, homogeneous to patchy CL (Figure 7a).They are subdivided to Anh1a and Anh1b according to CL intensity.Anh1a grains have higher CL intensity and irregular boundaries (Figure 7a) relative to Anh1b grains that are commonly overgrown, recrystallized, or crosscut by later anhydrites.The second generation (Anh2) shows bright to grey CL with "wavy" oscillatory boundaries (Figure 7a,b), with Anh2a being brighter and Anh2b darker.
The third generation of anhydrite (Anh3) occurred as euhedral crystals overgrowing Anh1 or infilled interstitial spaces of Anh1 and Anh2 (Figure 7a,c).They are subdivided to Anh3a and Anh3b according to CL intensity and zoning patterns (Figure 7c).Anh3a is brighter in CL and shows evident oscillatory zoning in contrast to Anh3b.
The last generation of anhydrite (Anh4) is anhedral and CL-dark (Figure 7d).Anh4 may cement Anh1 and Anh2 fragments (Figure 7e).All anhydrite generations except Anh3 have been found in both A.Vaq and B.Vmacpq.
The third generation of anhydrite (Anh3) occurred as euhedral crystals overgrowing Anh1 or infilled interstitial spaces of Anh1 and Anh2 (Figure 7a,c).They are subdivided to Anh3a and Anh3b according to CL intensity and zoning patterns (Figure 7c).Anh3a is brighter in CL and shows evident oscillatory zoning in contrast to Anh3b.
The last generation of anhydrite (Anh4) is anhedral and CL-dark (Figure 7d).Anh4 may cement Anh1 and Anh2 fragments (Figure 7e).All anhydrite generations except Anh3 have been found in both A.V aq and B.V macpq .

Fluid Inclusions
Fluid inclusions at Tongchang were named using the nomenclature of [74] based on the relative volume fraction of bubbles and solid species (e.g., B15HS denotes 15% vapor phase, a halite daughter, and an opaque solid).Fluid inclusion assemblages (FIA) and their origin were determined with the aid of CL microtextures.Microthermometric temperatures and estimated salinity, density, and pressure are listed in Table 3. Due to intense recrystallization, Qz1 grains are devoid of primary FIAs, but flooded with secondary FIAs in healed fractures.Qz2a and Qz2b quartz host primary FIAs of B15HS and B80 inclusions (Figure 8a Qz3b contains primary B50 inclusions (Figure 8c).They had ice melting tem tures of −2.7-−4.2°C and were homogenized at 300-350 °C via critical mode or v expansion.

Trace Elements
Results are shown and depicted in Figure 9, Table 4, and the Appendix.In all q generations, only Ti, Li, Na, Al, K, and Ge occurred in appreciable concentrations.nic and Sb becomes significant in Qz5 and Qz6.Qz3b contains primary B50 inclusions (Figure 8c).They had ice melting temperatures of −2.7-−4.2• C and were homogenized at 300-350 • C via critical mode or vapor expansion.
Qz4 commonly contains only B30 inclusions, which are considered primary.They exhibited ice melting at −3.2-−5.6 • C and were homogenized at 250-300 • C.They also occurred as secondary inclusions in Qz2 and Qz3 (Figure 8c).B10 occurred mainly as secondary inclusions in Qz2 and Qz3.It had ice melting temperatures at −2.6-−3.4• C and was homogenized at 187-208 • C.

Trace Elements
Results are shown and depicted in Figure 9, Table 4, and the Appendix A. In all quartz generations, only Ti, Li, Na, Al, K, and Ge occurred in appreciable concentrations.Arsenic and Sb becomes significant in Qz5 and Qz6.Lithium increases slightly from 11 ppm in Qz1 to av. 19 ppm in Qz3b, then decreases to 8 ppm in Qz4, and increases to 70 ppm in Qz6a which decreases to 29 ppm in Qz6b.Al increases from 172 ppm in Qz1 to 578 ppm in Qz3b, and then decreases to 300 ppm in Qz4, and then increases to 923 ppm in Qz6a which decreases to 718 ppm in Qz6b.Ge increases from 1 ppm in Qz1 to 14 ppm in Qz6b.As increases from 1 ppm in early quartz to 31 ppm in Qz6b, and Sb increases from 0 ppm (below detection limits) in early quartz to 20 ppm in Qz6a and 207 ppm in Qz6b.

Temperature Estimation
At a chemical equilibrium, the incorporation of Ti in quartz is dependent on pressure and temperature.The dependency has been devised as a geothermometer known as TitaniQ by various experiments [23,26,27,75,76].Different calibration equations have been derived [77], but one developed by [26], as shown below, was selected in this study based on the arguments presented in [76].An overview of the arguments is out of the scope of this study; interested readers are referred to the publications mentioned above.) increases with T but decreases with P. In a P-T space, equal X quartz TiO 2 (isopleth) defines vertical lines that have increasing slopes (Figure 10).For the Tongchang deposit, a TiO 2 = 1 can be assumed for Qz2-, Qz3-and Qz4-and Qz5-depositing fluids since the rutile was coprecipitated.For Qz1 and Qz6, a TiO 2 = 0.5 and 1, are used to calculate two isopleths for each X quartz TiO 2 .Exact pressures and temperatures can be obtained by combining TitaniQ and other geothermobarometers.Fluid inclusion isochores are an option but, unfortunately, they are only available for Qz3 and Qz4.B15HS inclusions in Qz1 and Qz2 are not used since their origin remains unclear [19] and their isochores intersect with TitaniQ isopleths at unreasonably high pressures (Figure 10).For the remaining quartz generations, a range of geologically reasonable pressures are assumed, i.e., 1500-500 bars for Qz1, Qz2, and Qz3a, 500-100 bars for Qz5, and 500-10 bars for Qz6.The assumption of decreasing pressures is supported by the increasing Al and Li contents consistent with a recent study by [78].At a constant pressure, temperature is negatively related to a TiO 2 .The minimum temperatures for Qz1 and Qz6 are calculated with a TiO 2 = 1 and minimum pressure, and maximum temperatures are calculated with a TiO 2 = 0.5 and maximum pressure.Accordingly, the following averaged temperature ranges are obtained for different quartz generations (Figure 10; Table 4)   Note that only Qz3a and Qz4 had reliable fluid inclusion isochore data that allowed for the use of the interception method.For other quartz generations, a pressure range was assumed to produce a reliable temperature estimate.

An Improved Timeline Based on Quartz CL Sequencing
Vein crosscut relations are often used to sort fluid events in the porphyry deposits [8].However, this technique could be imparted by the existence of reopened veins.In those cases, a detailed look into mineral microtextures could help construct a reliable timeline of superimposed fluid events.
CL microtextures in hydrothermal quartz falls in two general categories: primary crystallization textures and secondary modification textures attributable to quartz deposition, dissolution, and deformation processes [79,80].Concentric and euhedral oscillatory zoning in the Qz2a, Qz3, Qz5, and Qz6 grains belong to the primary crystallization textures, indicating a direct deposition from hydrothermal fluids.Mottled CL in the Qz1 and Qz2b grains is indicative of fluid-aided dynamic recrystallization.Similar Ti contents in both quartz generations might indicate a common event of recrystallization.The seemingly contradictory coexistence of un-recrystallized Qz2a and recrystallized Qz2b is likely linked to a difference in crystal sizes and shapes and resultant preferential deformation, as proposed for the Butte porphyry Cu-Mo deposit [22].Qz2b resorbing Qz2a, "splatter" and "cobweb" in the Qz1 and Qz2 grains as well as overgrowth of Qz3 on earlier quartz gains indicate quartz dissolution.Qz3 and Qz4 fragments and fracture-filling Qz5 grains are indicative of brittle fracturing.
Based on these pieces of textural evidence, a refined timeline of fluid evolution is proposed for the Tongchang deposit.The early stage involved the deposition of the Qz1 precursor, Qz2a, and the Qz2b precursor followed by dynamic recrystallization and a first-time dissolution.The subsequent intermediate stage started with dissolution of earlier quartz grains followed by progressive deposition of Qz3a, Qz3b, and Qz4.The late stage commenced with the fracturing and brecciation of earlier quartz grains followed by the progressive deposition of Qz5 and Qz6.
Petrographic observations suggest that the early Qz1-and Qz2-depositing stage also precipitated hematite and anhydrite (Anh1) in the veins, and bornite, chalcopyrite, and potassic silicates in the alteration halos.The early intermediate stage precipitated pyrite, chalcopyrite, anhydrite (Anh3), chlorite, and sericite, along with Qz3 quartz, as well as chlorite-sericite alteration halos.The late intermediate stage precipitated abundant chalcopyrite, pyrite, anhydrite (Anh4), chlorite, sericite, and calcite in the veins.The late stages generated pyrite and sericite with Qz5 and Qz6 grains.

Fluid Cooling and Compression during
Quartz Deposition, Recrystallization, and Dissolution 6.2.1.Quartz Deposition TitaniQ thermobarometry has been widely utilized to estimate the P-T conditions of quartz deposition in magmatic [81,82] and hydrothermal environments [22,74,83].Quartz solubility in various fluid compositions at varied P-T conditions can be used to investigate the fluid processes responsible for quartz deposition and dissolution, and, by inference, vein formation processes [31,32,84].
According to TitaniQ temperatures, early quartz deposition (Qz2a) occurred at around 529-618  11), the quartz deposition temperature ranges are located in the single liquid phase region where quartz solubility decreases dramatically with temperature, and much less with pressure.It is inferred that the quartz deposition is mainly the result of fluid cooling.Anhydrite has a similar prograde solubility behavior at temperatures above 350 • C in H 2 O-10wt.%NaClsolutions, but shows retrograde solubility below that temperature (the grey region in the Figure 11).Increasing salinity tends to decrease the deflection point of the temperature, thus shifting the retrograde region to lower temperatures.H 2 O-NaCl fluids containing silica and calcium sulfate would precipitate quartz and anhydrite upon cooling to 350 • C. Below 350 • C, quartz would continue to precipitate, whereas anhydrite deposition would cease.This inference is consistent with the coprecipitation of quartz and anhydrite in the early to intermediate stage and the disappearance of anhydrite in the late stage in the Tongchang deposit.
grey region in the Figure 11).Increasing salinity tends to decrease the deflection point of the temperature, thus shifting the retrograde region to lower temperatures.H2O-NaCl fluids containing silica and calcium sulfate would precipitate quartz and anhydrite upon cooling to 350 °C.Below 350 °C, quartz would continue to precipitate, whereas anhydrite deposition would cease.This inference is consistent with the coprecipitation of quartz and anhydrite in the early to intermediate stage and the disappearance of anhydrite in the late stage in the Tongchang deposit.[79]; and quartz data are from [31].The two areas in grey color are the retrograde solubility regions of anhydrite (An-hRS) and quartz (QzRS).The color-filled rectangles accompanied with error bars represent the deposition conditions of quartz and anhydrite (colors are the same as the in the Figure 10).The thick arrows (decompression cooling in red; quasi-isothermal compression in black) and connecting dash lines represent the possible P-T paths of the hydrothermal ore fluids.Note that the quasiisothermal fluid overpressure occurred prior to precipitation of Qz3/Anh2, which was followed by pressure fluctuations.Pressure fluctuations also happened during the formation of Qz5.Solubility values (in mm/kgH2O) are marked on the curves.GD solidus: H2O-saturated granodioritic melt solidus; CP10/20: critical point for H2O-10 wt.% NaCl or H2O-20 wt.% NaCl.L: liquid; V: vapor; H: halite.

Quartz Recrystallization
Like many other minerals, quartz undergoes dynamic recrystallization during deformation to minimize free strain energy [86].Three distinctive recrystallization mechanisms have been recognized [87], i.e., bulging recrystallization (BLG), subgrain rotation  [85]; and quartz data are from [31].The two areas in grey color are the retrograde solubility regions of anhydrite (AnhRS) and quartz (QzRS).The color-filled rectangles accompanied with error bars represent the deposition conditions of quartz and anhydrite (colors are the same as the in the Figure 10).The thick arrows (decompression cooling in red; quasi-isothermal compression in black) and connecting dash lines represent the possible P-T paths of the hydrothermal ore fluids.Note that the quasi-isothermal fluid overpressure occurred prior to precipitation of Qz3/Anh2, which was followed by pressure fluctuations.Pressure fluctuations also happened during the formation of Qz5.Solubility values (in mm/kgH 2 O) are marked on the curves.GD solidus: H 2 O-saturated granodioritic melt solidus; CP10/20: critical point for H 2 O-10 wt.% NaCl or H 2 O-20 wt.% NaCl.L: liquid; V: vapor; H: halite.

Quartz Recrystallization
Like many other minerals, quartz undergoes dynamic recrystallization during deformation to minimize free strain energy [86].Three distinctive recrystallization mechanisms have been recognized [87], i.e., bulging recrystallization (BLG), subgrain rotation (SGR), and grain boundary migration (GBM).These mechanisms occur at different temperature regimes and result in characteristic microstructures [22].BLG occurs at temperatures between 250 • C and 400 • C, resulting in small, recrystallized grains surrounding large, un-recrystallized grains.SGR occurs at temperatures between 400 • C and 500 • C, generating straight and commonly polygonal grain boundaries.GBM occurs at temperatures between 500 • C and 600 • C, producing lobate grain boundaries.TitaniQ applied in a metamorphic environment [88] and experiments [76,89,90] suggests that SGR and GBM could re-equilibrate Ti in quartz, and thus that TitaniQ would record the temperature of recrystallization.
The grain boundaries shown in Qz1 and Qz2b grains are similar to those of SGR (Figure 6a,b).TitaniQ temperatures of 400-550 • C suggest SGR/GBM recrystallization consistent with the microstructures.A noteworthy feature is that less-recrystallized Qz2a grains are commonly rimmed by CL-dark edges.A similar feature in B-type veins of the Butte porphyry deposit was proposed to be a result of mismatch in lattice alignment of adjacent crystals during competitive grain growth [22].However, we noted that the CLdark rims in the Qz2a grains are comparable to those quartz experimentally deformed with the presence of fluid [76].We infer that the CL-dark rims in Qz2 (Figure 6b) may have been a product of fluid-mediated dissolution-reprecipitation during dynamic recrystallization.If this inference is correct, the quartz dissolution event requires an increase in silica solubility, which cannot be explained by a drop in fluid pressure in a deflating magma chamber as proposed by [22].Conversely, we consider it more likely to reflect an increase in fluid pressure and/or temperature based on a solubility analysis (Figure 11).Fluid compression is a plausible process in that overpressure in the overlying carapace that could be caused by fluid exsolution and accumulation, which is common for the deposition of early quartz veins [91].In addition, silica solubility can be enhanced by new pulses of mafic recharge, which may introduce hotter fluids.A possible magmatic recharge was suggested by a geochemical study of the mafic enclaves of the Tongchang deposit [20].

Quartz Dissolution and Fracturing
Immediately after the deposition of Qz2b, a quartz-and anhydrite-dissolution took place prior to the deposition of Qz3 and Anh2 at lower temperatures (350-450 • C) and fluctuating pressures.The temperature constraints indicate a transition from lithostatic to hydrostatic regimes (Figure 11).The quartz dissolution was previously attributed to quasi-isobaric cooling through the retrograde quartz solubility region [17,31,92].However, this explanation is not likely for Tongchang because quasi-isobaric cooling would cause lower Ti contents in later deposited Qz3b, which is inconsistent with higher Ti contents in Qz3b (around 15 ppm) than in Qz3a (around 10 ppm).Another observation not in favor of quasi-isobaric cooling is that anhydrite displays prograde solubility in the quartz retrograde solubility field, which cannot account for the coeval dissolution of anhydrite and quartz (Figure 11).Therefore, it is inferred that fluid pressurization is seemingly the only viable way to elevate the solubility for both minerals.
Quartz and anhydrite show a similar solubility behavior at temperatures between 350 • C and 400 • C and pressures below 2000 bars, corresponding to a hydrostatic regime [92].Under hydrostatic conditions, quartz and anhydrite deforms in a brittle fashion, which is consistent with early anhydrite breccia cemented by Anh4 (Figure 7h).The quartz microbreccia enclosed in Qz5 suggests the occurrence of another episode of fluid overpressure.With the space created by brecciation, pressure fluctuated in response to fluid exsolution and mineral precipitation, resulting in the zoning patterns in Qz5b (Figure 11).
Preexisting veins can be reopened during a later fluid event by dissolution, mechanical breaking, brecciation, and recrystallization [17].At Tongchang, these phenomena are abundant and occurred at different stages.Dissolution and recrystallization occurred in the early to transitional stage when host rock behaved in ductile fashion, whereas breaking and brecciation occurred in the late stage when host rocks behaved in a brittle fashion.
The Tongchang porphyry deposit presents a good example of such challenges.According to their deposition temperatures, Qz1-2, Qz3 and Qz5 can be considered as indigenous quartz of A-, B-, and D-type veins [31].Three fourths of the examined veins contain quartz that precipitated earlier or later than the vein formation defined by vein classification (Figure 12).Visual estimates suggested that the early A veins, except the A.V kq vein in 10DX197, contain younger quartz (Qz3 and Qz4) in volumes <50%.The B veins particularly involve the addition of younger Qz4 quartz to older quartz.For instance, the B.V macpq vein in 12DX328 comprises a central lining of pyrite, chalcopyrite, Qz4, Anh4 and carbonates, which is sandwiched by Qz1 and Qz2 quartz.Overprints in late veins are dramatic, and are characterized by the presence of earlier relict quartz (Qz1-3) and minor younger Qz6 quartz.Qz4 quartz are typically absent.D.V mqp is only different from D.V pd veins in that it has a larger amount of relict quartz.
The Tongchang porphyry deposit presents a good example of such challenges.According to their deposition temperatures, Qz1-2, Qz3 and Qz5 can be considered as indigenous quartz of A-, B-, and D-type veins [31].Three fourths of the examined veins contain quartz that precipitated earlier or later than the vein formation defined by vein classification (Figure 12).Visual estimates suggested that the early A veins, except the A.Vkq vein in 10DX197, contain younger quartz (Qz3 and Qz4) in volumes <50%.The B veins particularly involve the addition of younger Qz4 quartz to older quartz.For instance, the B.Vmacpq vein in 12DX328 comprises a central lining of pyrite, chalcopyrite, Qz4, Anh4 and carbonates, which is sandwiched by Qz1 and Qz2 quartz.Overprints in late veins are dramatic, and are characterized by the presence of earlier relict quartz (Qz1-3) and minor younger Qz6 quartz.Qz4 quartz are typically absent.D.Vmqp is only different from D.Vpd veins in that it has a larger amount of relict quartz.A striking feature of the studied veins is that they all contain early quartz (Qz1-2), regardless of vein types.If these samples are representative of veining processes at Tongchang, it is plausible to infer that all the vein types are reopening early quartzdominated veins.The present early veins are the product of slight overprints.The chalcopyrite veins (B.Vcpq and B.Vmacpq) may be produced by the reopening of the earlier A striking feature of the studied veins is that they all contain early quartz (Qz1-2), regardless of vein types.If these samples are representative of veining processes at Tongchang, it is plausible to infer that all the vein types are reopening early quartz-dominated veins.The present early veins are the product of slight overprints.The chalcopyrite veins (B.V cpq and B.V macpq ) may be produced by the reopening of the earlier quartz veins by Qz4-depositing fluids, which is consistent with the conclusion of [31].The late pyritic veins (D.V mqp and D.V pd ) are products of overprints by Qz5-depositing fluids on earlier quartz veins.

Implications for Cu Sulfide Deposition
In the early fluid stage, hematite and anhydrite with little sulfide were deposited in the A.V hq veins (Figure 13), attesting to oxidized sulfur species (SO 2 , SO 2− 4 ) at high temperatures [98,99].In the potassic altered rocks, however, the low-sulfidation assemblage of chalcopyrite, bornite, magnetite and hematite developed [100].These Fe-Cu sulfides were formed coevally with Qz1 and Qz2 at high temperatures (650 • C to 450 • C) (Figure 13), and the required H 2 S was probably produced through the chemical reduction of sulfate in the fluids by mafic minerals such as biotite and hornblende [101].It is worth mentioning that potassic ores of Tongchang are much less significant than in usual porphyry deposits [74], likely due to sulfide remobilization [1].
peratures [98,99](.In the potassic altered rocks, however, the low-sulfidation assemblage of chalcopyrite, bornite, magnetite and hematite developed [100].These Fe-Cu sulfides were formed coevally with Qz1 and Qz2 at high temperatures (650 °C to 450 °C) (Figure 13), and the required H S was probably produced through the chemical reduction of sulfate in the fluids by mafic minerals such as biotite and hornblende [101].It is worth mentioning that potassic ores of Tongchang are much less significant than in usual porphyry deposits [74], likely due to sulfide remobilization [1].The first overprint is characterized by the deposition of pyrite ± chalcopyrite in the Qz3-Anh2 veins and chlorite-sericite alterations (Figure 13).The disappearance of ferric iron oxides and the appearance of pyrite indicate an increase in the sulfidation state with continuous cooling [100].The H S required for sulfide deposition is likely derived from the SO disproportionation favored by fluid cooling (350 °C to 450 °C) [98].The remobilization of early-stage sulfides might not be a significant source of H S because chalcopyrite solubility is significantly impeded during this process [102].In the alteration halos, the precipitation of anhydrite (Anh2) occurred coevally with sulfide deposition and the hydrolysis of feldspars, indicating a potential chemisorption reaction between SObearing fluids and andesine [10].
Cu sulfides of the Tongchang deposit were mainly deposited in the vein assemblage of pyrite-chalcopyrite-Qz4-Anh4 at 300 °C-380 °C (Figure 13).The increased availability of Cu may be a result of the dissociation of CuCl complexes caused by fluid dilu- The first overprint is characterized by the deposition of pyrite ± chalcopyrite in the Qz3-Anh2 veins and chlorite-sericite alterations (Figure 13).The disappearance of ferric iron oxides and the appearance of pyrite indicate an increase in the sulfidation state with continuous cooling [100].The H 2 S required for sulfide deposition is likely derived from the SO 2 disproportionation favored by fluid cooling (350 • C to 450 • C) [98].The remobilization of early-stage sulfides might not be a significant source of H 2 S because chalcopyrite solubility is significantly impeded during this process [102].In the alteration halos, the precipitation of anhydrite (Anh2) occurred coevally with sulfide deposition and the hydrolysis of feldspars, indicating a potential chemisorption reaction between SO 2 -bearing fluids and andesine [10].
Cu sulfides of the Tongchang deposit were mainly deposited in the vein assemblage of pyrite-chalcopyrite-Qz4-Anh4 at 300 • C-380 • C (Figure 13).The increased availability of Cu may be a result of the dissociation of CuCl − 2 complexes caused by fluid dilution in response to the entrainment of groundwaters and metamorphic fluids [103].Another contributing factor might be the remobilization of existing sulfides associated with potassic alteration.Inflow of air saturated groundwater may significantly elevate the oxidation state and acidity of the magmatic fluids, which are favorable for chalcopyrite dissolution [102].
Although a large proportion of Cu sulfides are hosted in sericitic alteration zones, petrographic evidence of chalcopyrite inclusions and high-sulfidation assemblage of pyrite and tennantite indicates that they are remobilized from earlier ores (Figure 13).The mineral assemblage in the phyllic altered rocks reflects the stability of muscovite, quartz and pyrite, and the instability of chlorite, feldspars, mafic minerals, and chalcopyrite.This is similar to the Butte porphyry deposit, where the early ores (chalcopyrite-pyrite-magnetite) with potassic-chloritic alterations were remobilized by circulating oxidized groundwaters and re-deposited as chalcocite and covellite [104].

Conclusions
The Tongchang porphyry deposit represents a rare case of a porphyry copper deposit resulting from a single magmatic intrusion.It was expected that it would have undergone less fluid overprint compared to multi-intrusion counterparts.However, CL imaging of vein quartz and anhydrite revealed vein overprints as complicated (if not more complicated) as those in multi-intrusion deposits.This observation suggests that vein overprinting is likely an intrinsic property of porphyry deposits, regardless of intrusive history and other geological factors.The failure to recognize such vein overprints may result in erroneous fluid chronology and the incorrect interpretation of geochemical data (e.g., fluid inclusion and isotopes).Therefore, one should be careful when conducting geochemical analysis based on macroscopic features and a conventional vein model.Whenever possible, detailed CL imaging is an imperative.
As expected, the P-T history of Tongchang is characterized by an overall cooling and decompression trajectory.However, in contrast to monotonic cooling model, the P-T analysis of Tongchang porphyry deposit uncovered two significant episodes of fluid compression, likely corresponding to two periods of magmatic fluid accumulation under lithostatic and lithostatic to hydrostatic regimes.There might have been an increase in temperature during the second episode of compression.These irregularities in pressure should be considered for fluid models of porphyry deposits.
The Cu deposition process is better elucidated with the aid of CL petrography and integrated geothermometry.It was clearly shown that the high oxidation state of early high temperature fluids precludes Cu sulfide deposition in veins.However, Cu sulfide could be deposited via a chemical reaction with mafic minerals.The Cu sulfide deposition commences at temperatures lower than 450 • C when SO 2 disproportionation starts.Most chalcopyrite was precipitated at lower temperatures of 350 • C to 300 • C.

Figure 2 .
Figure 2. (a) A geological map of the Dexing porphyry Cu district; (b) Two cross sections (a-b, and c-d) of the Cu district (adapted from[19,38]).Dash lines denoted with L0-8 represents exploration lines; black pentagons represent the locations of drill hole, and the sampling localities are marked with small rectangles.

Figure 2 .
Figure 2. (a) A geological map of the Dexing porphyry Cu district; (b) Two cross sections (a-b, and c-d) of the Cu district (adapted from[19,38]).Dash lines denoted with L0-8 represents exploration lines; black pentagons represent the locations of drill hole, and the sampling localities are marked with small rectangles.
), and has been dated at ca. 154 Ma by the zircon U-Pb method [65].

Figure 3 .
Figure 3. Major lithologies (phyllite, slate, hornfel, and porphyries) and their spatial rela Tongchang porphyry Cu deposit.(a) an outcrop of phyllite; (b) a slate with pyrite stock sericitic alteration halo; (c) a biotite hornfel; (d) a fresh granodiorite porphyry; (e) a fr porphyry; (f) a granodiorite porphyry crosscut by an anhydrite quartz vein; (g) a gr porphyry truncating phyllite, both of which are altered by sericitic alteration; (h) porphyry truncating slate, both of which are crosscut by a pyrite vein; (i) an altered truncating D-type pyrite vein (D.Vpd) hosted in the slate.The scale in (c) is a Chinese coin.

Figure 3 .
Figure 3. Major lithologies (phyllite, slate, hornfel, and porphyries) and their spatial relation at the Tongchang porphyry Cu deposit.(a) an outcrop of phyllite; (b) a slate with pyrite stockworks and sericitic alteration halo; (c) a biotite hornfel; (d) a fresh granodiorite porphyry; (e) a fresh diorite porphyry; (f) a granodiorite porphyry crosscut by an anhydrite quartz vein; (g) a granodiorite porphyry truncating phyllite, both of which are altered by sericitic alteration; (h) a dioritic porphyry truncating slate, both of which are crosscut by a pyrite vein; (i) an altered porphyry truncating D-type pyrite vein (D.Vpd) hosted in the slate.The scale in (c) is a Chinese one-dime coin.

Figure 4 .
Figure 4. Major types of hydrothermal alteration at the Tongchang porphyry Cu deposit.(a) Potassic alteration in the granodiorite porphyry; (b) Chlorite-sericite overprinting on potassic alteration in the granodiorite porphyry; (c) Chlorite-sericitic alteration in the granodiorite porphyry; (d) Sericitic overprinting on chlorite-sericitic alteration in the granodiorite porphyry; (e) Sericitic alteration with pyrite-dominated veins in the slate; (f) Propylitic alteration in the phyllite.Scale bars are 1 cm.The scale in (d) is a Chinese one-dime coin.

Figure 4 .
Figure 4. Major types of hydrothermal alteration at the Tongchang porphyry Cu deposit.(a) Potassic alteration in the granodiorite porphyry; (b) Chlorite-sericite overprinting on potassic alteration in the granodiorite porphyry; (c) Chlorite-sericitic alteration in the granodiorite porphyry; (d) Sericitic overprinting on chlorite-sericitic alteration in the granodiorite porphyry; (e) Sericitic alteration with pyrite-dominated veins in the slate; (f) Propylitic alteration in the phyllite.Scale bars are 1 cm.The scale in (d) is a Chinese one-dime coin.

Figure 6 .
Figure 6.CL microtextures of quartz at the Tongchang porphyry Cu deposit.(a) The earliest CLhomogeneous quartz (Qz1) reopened by later veinlet; (b,c) The second generation (Qz2a and Qz2b) that shows dark blue CL with faint zoning.Also shown are three generations of anhydrite (purple); (d) The third generation of quartz (Qz3a and Qz3b) resorbing and overgrowing on early Qz2b.Note that the Qz3a has lower CL intensity than Qz3b does; (e) Qz3b directly overgrown on Qz2b, showing conformable contact with the surrounding pyrite; (f) Qz2 and Qz3 displayed blue CL with different intensities; (g,h) Coprecipitation of CL-dark Qz4 (overgrown blue-CL Qz3b) and Anh4 (cementing purple-CL Anh2 breccia); (i) Qz5a and Qz5b overgrew on Qz3b.(j) Qz5a fill in fractures in Qz3a.Note that Qz5a has lower CL than Qz5b; (k) CL-homogeneous Qz6a and CLoscillatory Qz6b.Note that Qz6b has bright CL. (l) Qz6a shows red CL, whereas Qz6b shows bright blue C. Scale bars are 300 µm.

Figure 6 .
Figure 6.CL microtextures of quartz at the Tongchang porphyry Cu deposit.(a) The earliest CLhomogeneous quartz (Qz1) reopened by later veinlet; (b,c) The second generation (Qz2a and Qz2b) that shows dark blue CL with faint zoning.Also shown are three generations of anhydrite (purple); (d) The third generation of quartz (Qz3a and Qz3b) resorbing and overgrowing on early Qz2b.Note that the Qz3a has lower CL intensity than Qz3b does; (e) Qz3b directly overgrown on Qz2b, showing conformable contact with the surrounding pyrite; (f) Qz2 and Qz3 displayed blue CL with different intensities; (g,h) Coprecipitation of CL-dark Qz4 (overgrown blue-CL Qz3b) and Anh4 (cementing purple-CL Anh2 breccia); (i) Qz5a and Qz5b overgrew on Qz3b.(j) Qz5a fill in fractures in Qz3a.Note that Qz5a has lower CL than Qz5b; (k) CL-homogeneous Qz6a and CL-oscillatory Qz6b.Note that Qz6b has bright CL. (l) Qz6a shows red CL, whereas Qz6b shows bright blue C. Scale bars are 300 µm.

1 Figure 8 .
Figure 8. Petrography of fluid inclusions hosted in different generations of quartz i Tongchang porphyry Cu deposit.The coexistence of brine (B15HS) and vapor-rich (B80) inclu in Qz2a (a) and Qz2b (b), which distributed along CL growth zones indicative of primary o primary B50 inclusions in Qz3b with secondary trails of B80 and B30 inclusions (c); secondar of B10 inclusion in Qz2b (d)."B10" denotes the vapor phase and its volume proportion rela the total inclusion volume; "L" denotes aqueous liquid, "V" denotes vapor, "H" denotes a daughter phase, and "S" denotes unidentified solids.CL growth zones are marked with d curves in the inclusion images.Scale bars are 10 µm.

Figure 8 .
Figure 8. Petrography of fluid inclusions hosted in different generations of quartz in the Tongchang porphyry Cu deposit.The coexistence of brine (B15HS) and vapor-rich (B80) inclusions in Qz2a (a) and Qz2b (b), which distributed along CL growth zones indicative of primary origin; primary B50 inclusions in Qz3b with secondary trails of B80 and B30 inclusions (c); secondary trail of B10 inclusion in Qz2b (d)."B10" denotes the vapor phase and its volume proportion relative to the total inclusion volume; "L" denotes aqueous liquid, "V" denotes vapor, "H" denotes a halite daughter phase, and "S" denotes unidentified solids.CL growth zones are marked with dashed curves in the inclusion images.Scale bars are 10 µm.

Figure 10 .
Figure 10.Pressure and temperature estimates by combining TitaniQ isopleths (dash-d in red) and fluid inclusion isochores (dash lines in black).Ti isopleths of 1 ppm, 10 pp 50 ppm, and 100 ppm are shown; isochores of B15HS inclusions in both Qz2a and Qz2b sions in Qz3b, B30 inclusions in Qz4, and B10 inclusions are shown.The minimum and range (solid open diamonds in colors) as well as an averaged range (solid color-filled

Figure 10 .
Figure 10.Pressure and temperature estimates by combining TitaniQ isopleths (dash-dotted lines in red) and fluid inclusion isochores (dash lines in black).Ti isopleths of 1 ppm, 10 ppm, 30 ppm, 50 ppm, and 100 ppm are shown; isochores of B15HS inclusions in both Qz2a and Qz2b, B50 inclusions in Qz3b, B30 inclusions in Qz4, and B10 inclusions are shown.The minimum and maximum range (solid open diamonds in colors) as well as an averaged range (solid color-filled) of Ti contents for all quartz generations are shown.Orange indicates Qz1, red indicates Qz2a, pink indicates Qz2b, yellow indicates Qz3a, green indicates Qz3b, blue indicates Qz4, dark brown indicates Qz5b, purple indicates Qz6a, and grey indicates Qz6b.Note that only Qz3a and Qz4 had reliable fluid inclusion isochore data that allowed for the use of the interception method.For other quartz generations, a pressure range was assumed to produce a reliable temperature estimate.

Figure 11 .
Figure11.P-T phase diagrams for H2O-10 wt.% NaCl and H2O-20 wt.% NaCl aqueous systems, contoured with quartz and anhydrite solubilities (red: quartz; green: anhydrite; solid curve: H2O-20 wt.% NaCl; dashed curve: H2O-10 wt.% NaCl.Anhydrite data are from[79]; and quartz data are from[31].The two areas in grey color are the retrograde solubility regions of anhydrite (An-hRS) and quartz (QzRS).The color-filled rectangles accompanied with error bars represent the deposition conditions of quartz and anhydrite (colors are the same as the in the Figure10).The thick arrows (decompression cooling in red; quasi-isothermal compression in black) and connecting dash lines represent the possible P-T paths of the hydrothermal ore fluids.Note that the quasiisothermal fluid overpressure occurred prior to precipitation of Qz3/Anh2, which was followed by pressure fluctuations.Pressure fluctuations also happened during the formation of Qz5.Solubility values (in mm/kgH2O) are marked on the curves.GD solidus: H2O-saturated granodioritic melt solidus; CP10/20: critical point for H2O-10 wt.% NaCl or H2O-20 wt.% NaCl.L: liquid; V: vapor; H: halite.

Figure 11 .
Figure 11.P-T phase diagrams for H 2 O-10 wt.% NaCl and H 2 O-20 wt.% NaCl aqueous systems, contoured with quartz and anhydrite solubilities (red: quartz; green: anhydrite; solid curve: H 2 O-20 wt.% NaCl; dashed curve: H 2 O-10 wt.% NaCl.Anhydrite data are from[85]; and quartz data are from[31].The two areas in grey color are the retrograde solubility regions of anhydrite (AnhRS) and quartz (QzRS).The color-filled rectangles accompanied with error bars represent the deposition conditions of quartz and anhydrite (colors are the same as the in the Figure10).The thick arrows (decompression cooling in red; quasi-isothermal compression in black) and connecting dash lines represent the possible P-T paths of the hydrothermal ore fluids.Note that the quasi-isothermal fluid overpressure occurred prior to precipitation of Qz3/Anh2, which was followed by pressure fluctuations.Pressure fluctuations also happened during the formation of Qz5.Solubility values (in mm/kgH 2 O) are marked on the curves.GD solidus: H 2 O-saturated granodioritic melt solidus; CP10/20: critical point for H 2 O-10 wt.% NaCl or H 2 O-20 wt.% NaCl.L: liquid; V: vapor; H: halite.

Figure 12 .
Figure 12.A qualitative estimation diagram of quartz types contained in A-type, B-type, and Dtype veins examined in this study.All examined veins (17 in total) contained quartz younger or older than the vein formation, except two A-type veins.All vein contained early quartz (Qz1, Qz2a, and Qz2b), regardless of vein types.The presence of younger quartz in A-type veins indicates low-degree vein reopening, whereas the presence of older quartz in B-type and D-type veins indicates high-degree vein reopening.Note that there were two veins in the sample 12LDX37.The A.Vhq vein is named as 12LDX37-1, and the B.Vcpq vein is named 12LDX37-2.

Figure 12 .
Figure 12.A qualitative estimation diagram of quartz types contained in A-type, B-type, and D-type veins examined in this study.All examined veins (17 in total) contained quartz younger or older than the vein formation, except two A-type veins.All vein contained early quartz (Qz1, Qz2a, and Qz2b), regardless of vein types.The presence of younger quartz in A-type veins indicates low-degree vein reopening, whereas the presence of older quartz in B-type and D-type veins indicates high-degree vein reopening.Note that there were two veins in the sample 12LDX37.The A.V hq vein is named as 12LDX37-1, and the B.V cpq vein is named 12LDX37-2.

Figure A2 .
Figure A2.Tukey Boxplot of trace elements in vein quartz of the Tongchang porphyry Cu deposit.Figure A2.Tukey Boxplot of trace elements in vein quartz of the Tongchang porphyry Cu deposit.

Figure A2 .
Figure A2.Tukey Boxplot of trace elements in vein quartz of the Tongchang porphyry Cu deposit.Figure A2.Tukey Boxplot of trace elements in vein quartz of the Tongchang porphyry Cu deposit.

Table 1 .
A summary of sampling locality and petrography of the analyzed rock samples of the Tongchang porphyry Cu deposit.The rock is significantly affected by chlorite-sericite alteration.Primary mafic minerals are replaced by chunks of chlorite, sericite, hematite, muscovite, and chalcopyrite.Feldspar phenocrysts are replaced by sericite.Quartz phenocrysts remain largely unaffected.Groundmass is replaced by fine-grained hematite and chalcopyrite.The rock is intensively crosscut by hematite quartz veins (V hq ), which are subsequently cut by chalcopyrite pyrite quartz veins and veinlets (V cpq ).
kqPlagioclase phenocrysts are partly (less than 50% of the surface) altered by sericite along cleavages and cracks.K-feldspar phenocrysts are completely replaced by sericite.Unidentified mineral pseudomorphs of inner sericite and outer chlorite, hematite and magnetite are present.Primary mafic minerals are replaced by chunks of small biotite, chlorite, hematite, and magnetite.Groundmass consists of sericite, quartz, biotite, hematite, and magnetite.10DX197DrillholeZK8-4777mK-feldspar phenocrysts are partly altered by muscovite and chlorite, and plagioclase are pseudomorphed by sericite and chlorite.Quartz grains are overgrown and enlarged.Primary biotites are pseudomorphed by muscovite, chlorite, rutile, and quartz.Numerous chunks of chlorite, sericite, muscovite, rutile, chalcopyrite, epidote, and calcite are present.12DXF05Openpit230m platformFeldspar phenocrysts are overgrown by K-feldspar, which are partly altered by sericite.Primary biotites are replaced by quartz, sericite, chalcopyrite, pyrite, and rutile.Quartz phenocrysts remain largely unaffected.Numerous chunks of chlorite are present.Groundmass consists of fine-grained quartz.K-feldspar phenocrysts are pseudomorphed by sericite, and some anhedral K-feldspar grains are partly altered by sericite, chlorite, and epidote.Clusters of chlorites, sericite, anhydrite, epidote and molybdenite are present.Some quartz phenocrysts are overgrown, and others remain unaffected.

Table 2 .
A summary of generations, CL textures, mineral assemblages and Ti contents of quartz and anhydrite of the Tongchang porphyry Cu deposit.

Table 3 .
Fluid inclusion microthermometry of the Tongchang porphyry Cu deposit.

Table 4 .
LA-ICP-MS trace elements in quartz of the Tongchang porphyry Cu deposit.
• C; early intermediate quartz deposition (Qz3a and Qz3b) occurred at around 350-450 • C; late intermediate quartz deposition (Qz4) occurred at around 310-400 • C; and late quartz deposition (Qz5 and Qz6) occurred at around 240-400 • C. On a P-T phase diagram for H 2 O-NaCl overlapped with quartz solubility curves (Figure