Rutile and Chlorite Geochemistry Constraints on the Formation of the Tuwu Porphyry Cu Deposit, Eastern Tianshan and Its Exploration Significance

The chemical composition of rutile has been used as an indicator in magmatic and metamorphic-related diagenetic systems, but rarely in porphyry-style ore systems. The Tuwu deposit (557 Mt at 0.58% Cu) is a large porphyry-style Cu mineralization in Eastern Tianshan, Xinjiang, with typical disseminated, stockwork mineralized veins hosted in tonalite and diorite porphyry, and to a lesser extent in volcanic rocks of the Qi’eshan Group. We first present determination of rutile minerals coupled with chlorite identified in mineralized porphyries from Tuwu to reveal their geochemical features, thus providing new insights into the ore-forming processes and metal exploration. Petrographic and BSE observations show that the rutile generally occurs as large crystals (30 to 80 µm), in association with hydrothermal quartz, chlorite, pyrite, and chalcopyrite. The rutile grains display V, Fe, and Sn enrichment and flat LREE-MREE patterns, indicating a hydrothermal origin. Titanium in rutile (TiO2) is suggested to be sourced from the breakdown and re-equilibration of primary magmatic biotite and Ti-magnetite, and substituted by Sn4+, high field strength elements (HFSE; e.g., Zr4+ and Hf4+), and minor Mo4+ under hydrothermal conditions. The extremely low Mo values (average 30 ppm) in rutile may be due to rutile formation postdating that of Mo sulfides (MoS2) formation in hydrothermal fluids. Chlorite analyses imply that the ore-forming fluids of the main stage were weakly oxidized (logfO2 = −28.5 to −22.1) and of intermediate temperatures (308 to 372 °C), consistent with previous fluid inclusion studies. In addition, Zr-in-rutile geothermometer yields overestimated temperatures (>430 °C) as excess Zr is incorporated into rutile, which is likely caused by fast crystal growth or post crystallization modification by F-Cl-bearing fluid. Thus, application of this geothermometer to magmatic-hydrothermal ore systems is questionable. Based on the comparison of rutile characteristics of porphyry Cu with other types of ore deposits and barren rocks, we suggest that porphyry Cu-related rutile typically has larger grain size, is enriched in V (average 3408 ppm, compared to <1500 ppm of barren rocks) and to a lesser extent in W and Sn (average 121 and 196 ppm, respectively), and has elevated Cr + V/Nb + Ta ratios. These distinctive signatures can be used as critical indicators of porphyry-style Cu mineralization and may serve as a valuable tool in mineral exploration.


Geology of Tuwu Deposit
The Tuwu porphyry Cu deposit located in the southern part of Dananhu-Tousuquan Arc, Xinjiang ( Figure 1) contains ore reserve of 557 Mt at average grade of 0.58 wt % Cu and 0.2 g/t Au [26,46]. The ore deposit is hosted in diorite porphyry and tonalite porphyry that were emplaced into the Carboniferous Qi'eshan Group, and mainly controlled by EW trending faults and several subordinate NW-trending faults ( Figure 2). The Qi'eshan Group is primarily divided into three units from bottom to top, including the lower andesite and basalt lavas intercalated with tuff, the middle andesite and brecciated andesite lavas, and the upper pebbly lithic sandstone and minor tuffaceous siltstone intercalated with basalt, andesite and dacite lavas [26,47]. Previous U-Pb zircon dating for the diorite porphyry and tonalite porphyry indicates that they were emplaced at 338.6 ± 2.9 Ma and 332.3 ± 5.9 Ma, respectively [47,48]. Minor younger diorite and diabase dykes are present in the district, which intruded the earlier magmatic rocks. Orogen showing major tectonic units, faults and ore deposits (modified from [28]).

Geology of Tuwu Deposit
The Tuwu porphyry Cu deposit located in the southern part of Dananhu-Tousuquan Arc, Xinjiang ( Figure 1) contains ore reserve of 557 Mt at average grade of 0.58 wt % Cu and 0.2 g/t Au [26,46]. The ore deposit is hosted in diorite porphyry and tonalite porphyry that were emplaced into the Carboniferous Qi'eshan Group, and mainly controlled by EW trending faults and several subordinate NW-trending faults ( Figure 2). The Qi'eshan Group is primarily divided into three units from bottom to top, including the lower andesite and basalt lavas intercalated with tuff, the middle andesite and brecciated andesite lavas, and the upper pebbly lithic sandstone and minor tuffaceous siltstone intercalated with basalt, andesite and dacite lavas [26,47]. Previous U-Pb zircon dating for the diorite porphyry and tonalite porphyry indicates that they were emplaced at 338.6 ± 2.9 Ma and  [47,48]. Minor younger diorite and diabase dykes are present in the district, which intruded the earlier magmatic rocks. The three recognized mineralized zones are referred to as I, II, and III ore zones (Figure 2a), where the ore zone II accounts for 90% of the total copper reserves. Hydrothermal alteration at Tuwu affecting porphyries, and wall-rocks can be divided into potassic, phyllic, propylitic, and limited argillic alteration zones, which are shown in Figure 2c [23,26]. Copper mineralization, mostly present as chalcopyrite and bornite, is closely related to the potassic and phyllic alteration zones. The mineralization process at Tuwu includes four stages based on mineral assemblages and crosscutting relationships [23,28,30]. Stage I is represented by two types of veins, including quartz-magnetite-biotite and quartzmagnetite ± K-feldspar ± albite ± pyrite veins associated with local potassic and late pervasive chlorite alterations. Stage II and III are composed of chalcopyrite, bornite, quartz, pyrite, sericite, chlorite, albite, with minor magnetite and enargite, some of which is locally intergrown with rutile minerals (Figure 3). Stage II and III veins are characterized by significant Cu and Mo mineralization and are considered to be the main hydrothermal stages at Tuwu. The copper enrichment is generally linked to phyllic and chlorite-sericite alterations. Three typical Cu-sulfide-bearing veins are recognized in the main stage, including quartz-chalcopyrite-magnetite ± pyrite, quartz-bornite ± chalcopyrite ± chlorite ± epidote ± pyrite, and quartz-chalcopyrite ± chlorite ± pyrite ± rutile veins [23]. Stage IV is defined by mineral associations of calcite, quartz, chlorite, anhydrite, epidote, and pyrite. The molybdenite Re-Os dating for Tuwu suggests that the main mineralization event occurred at 335.6 ± 4.1 Ma [28], in agreement with the emplacement age of the local porphyry intrusions [48][49][50]. The three recognized mineralized zones are referred to as I, II, and III ore zones (Figure 2a), where the ore zone II accounts for 90% of the total copper reserves. Hydrothermal alteration at Tuwu affecting porphyries, and wall-rocks can be divided into potassic, phyllic, propylitic, and limited argillic alteration zones, which are shown in Figure 2c [23,26]. Copper mineralization, mostly present as chalcopyrite and bornite, is closely related to the potassic and phyllic alteration zones. The mineralization process at Tuwu includes four stages based on mineral assemblages and crosscutting relationships [23,28,30]. Stage I is represented by two types of veins, including quartz-magnetitebiotite and quartz-magnetite ± K-feldspar ± albite ± pyrite veins associated with local potassic and late pervasive chlorite alterations. Stage II and III are composed of chalcopyrite, bornite, quartz, pyrite, sericite, chlorite, albite, with minor magnetite and enargite, some of which is locally intergrown with rutile minerals (Figure 3). Stage II and III veins are characterized by significant Cu and Mo mineralization and are considered to be the main hydrothermal stages at Tuwu. The copper enrichment is generally linked to phyllic and chlorite-sericite alterations. Three typical Cu-sulfide-bearing veins are recognized in the main stage, including quartz-chalcopyrite-magnetite ± pyrite, quartz-bornite ± chalcopyrite ± chlorite ± epidote ± pyrite, and quartz-chalcopyrite ± chlorite ± pyrite ± rutile veins [23]. Stage IV is defined by mineral associations of calcite, quartz, chlorite, anhydrite, epidote, and pyrite. The molybdenite Re-Os dating for Tuwu suggests that

Sampling and Analytical Methods
Six representative mineralized samples analyzed in this paper were collected from ore zone II in the Tuwu porphyry Cu deposit (sample locations are shown in Figure 2a). The samples were polished and prepared for micro-observations and in situ chemical composition analyses, including EPMA and LA-ICP-MS. Detailed petrographic studies were carried out using an Olympus BX51 (Olympus Corporation, Tokyo, Japan) petrographic microscope at the China University of Geosciences, Beijing (CUGB), China.
Backscattered electron (BSE) images of rutile, chlorite and EPMA were taken on polished thin sections using a SIMADAZU EPMA-1720 instrument (Shimadzu Corporation, Kyoto, Japan) equipped with five wavelength dispersive spectrometers at the Resources Exploration Laboratory, CUGB. An accelerating voltage of 15 keV, a beam current of 10 nA, and a beam spot diameter of 5 µm were used for analyses. Different elements have different standard minerals, and the detection limits of each element are varied. The suite of analyzed elements mainly included Na, Cr, Si, Ti, Mg, Fe, Mn, Al, Ni, and K; the detection limits of these elements are shown in Table 1. Counting times were 20 s on the main peak and 10 s on the background. The standards used for calibration included albite for Na, chromite for Cr, diopside for Si and Mg, rutile for Ti, hematite for Fe, rhodonite for Mn, garnet for Al, synthetic NiO for Ni, and orthoclase for K. ZAF3 routine was used for

Sampling and Analytical Methods
Six representative mineralized samples analyzed in this paper were collected from ore zone II in the Tuwu porphyry Cu deposit (sample locations are shown in Figure 2a). The samples were polished and prepared for micro-observations and in situ chemical composition analyses, including EPMA and LA-ICP-MS. Detailed petrographic studies were carried out using an Olympus BX51 (Olympus Corporation, Tokyo, Japan) petrographic microscope at the China University of Geosciences, Beijing (CUGB), China.
Backscattered electron (BSE) images of rutile, chlorite and EPMA were taken on polished thin sections using a SIMADAZU EPMA-1720 instrument (Shimadzu Corporation, Kyoto, Japan) equipped with five wavelength dispersive spectrometers at the Resources Exploration Laboratory, CUGB. An accelerating voltage of 15 keV, a beam current of 10 nA, and a beam spot diameter of 5 µm were used for analyses. Different elements have different standard minerals, and the detection limits of each element are varied. The suite of analyzed elements mainly included Na, Cr, Si, Ti, Mg, Fe, Mn, Al, Ni, and K; the detection limits of these elements are shown in Table 1. Counting times were 20 s on the main peak and 10 s on the background. The standards used for calibration included albite for Na, chromite for Cr, diopside for Si and Mg, rutile for Ti, hematite for Fe, rhodonite for Mn, garnet for Al, synthetic NiO for Ni, and orthoclase for K. ZAF3 routine was used for data correction. The mineral formulas were recalculated using MINPET 2.0 software. Element mapping of one sample was performed using a spectrometer setup similar to the spot analyses, which consumed approximately 4 h of instrument time. Trace elements in rutile samples (13TW2-75, 13TW2-83, and TW-33) were measured by LA-ICP-MS analyses, which were carried out on an ESI NWR 193 nm laser ablation system (Elemental Scientific, Montana, USA) coupled with an Agilent 7500a quadrupole ICP-MS (Agilent Technologies Inc., Tokyo, Japan) at the Beijing Kuangyan Geoanalysis Laboratory Co., Ltd., Beijing, China. Helium and argon were used as the carrier gas and make-up gas, respectively. Argon was mixed with helium via a Y-joint before entering the ICP. Each analysis included the acquisition of a background signal for approximately 15-20 s (gas blank) followed by 45 s of data acquisition from the samples. A spot size of 30 µm and a 10 Hz ablation frequency were used for single spot ablation. NIST SRM610 and NIST SRM612 standard silicate glass were used as the external standard minerals. Titanium content was used as the internal standard to correct the time-dependent drift of sensitivity and mass discrimination for the trace elements analyses [51]. The detection limits for each element are shown in Table 2. Quantification of element concentrations was obtained according to the GeoReM database [52]. Off-line selection and integration of background and analytic signals, time-drift correction, and quantitative calibration were performed using the ICPMSDataCal program [51,53].

Mineralogical and Chemical Composition of Rutile
The samples analyzed in this study include mineralized porphyries containing disseminated chalcopyrite or quartz-chalcopyrite ± chlorite veins, with different alteration types (e.g., phyllic and propylitic). Rutile in mineralized porphyries is typically dark brownish-red or yellowish in color, varies from subhedral to anhedral in shape, and is approximately 30 to 60 µm in length (up to 80 µm; Figure 3). Most rutile occurs as aggregates, including 3 to 25 grains within a small area, and is commonly intergrown with quartz, chlorite, chalcopyrite, and/or pyrite, or formed prior to these minerals (Figure 4).
High-contrast BSE images reveal that rutile grains were not zoned or twinned ( Figure  4), also verified in EPMA element mapping images ( Figure 5). Single rutile crystals exhibited homogeneous distribution of Ti, V, Cr, and Sn, with slightly heterogeneous W distribution. Trace element contents of rutile crystals from samples of 13TW2-75, 13TW2-83, and TW-33 are present in Table 2

Zr-in-Rutile Thermometry
The solubility of Zr in rutile coexisting with zircon and quartz turned out to be strongly temperature-dependent, and thus Zr content in rutile was confirmed as a geothermometer [14,15,54]. Rutile crystals from Tuwu generally occur in quartz-bearing veins of ore stage II with presence of minute zircon inside the quartz crystals (Figure 3a,b). The rutile samples had variable Zr concentrations ranging from 50 to 519 ppm ( Table 2). Using the rutile Zr concentrations and equation of [14] without considering the pressure effect, the temperatures calculated for these samples were between 482 and 644 °C. Based on the estimation of fluid pressure (~90 bar) constrained by previous fluid inclusion analyses [23], the pressure-corrected temperatures were calculated to be 488-646 °C (Figure 7) using the experimentally calibrated method of Tomkins et al. [15], and 434-610 °C using the alternative method proposed by Kohn [54].

Zr-in-Rutile Thermometry
The solubility of Zr in rutile coexisting with zircon and quartz turned out to be strongly temperature-dependent, and thus Zr content in rutile was confirmed as a geothermometer [14,15,54]. Rutile crystals from Tuwu generally occur in quartz-bearing veins of ore stage II with presence of minute zircon inside the quartz crystals (Figure 3a,b). The rutile samples had variable Zr concentrations ranging from 50 to 519 ppm ( Table 2). Using the rutile Zr concentrations and equation of [14] without considering the pressure effect, the temperatures calculated for these samples were between 482 and 644 • C. Based on the estimation of fluid pressure (~90 bar) constrained by previous fluid inclusion analyses [23], the pressure-corrected temperatures were calculated to be 488-646 • C (Figure 7) using the experimentally calibrated method of Tomkins et al. [15], and 434-610 • C using the alternative method proposed by Kohn [54].

Zr-in-Rutile Thermometry
The solubility of Zr in rutile coexisting with zircon and quartz turned out to be strongly temperature-dependent, and thus Zr content in rutile was confirmed as a geothermometer [14,15,54]. Rutile crystals from Tuwu generally occur in quartz-bearing veins of ore stage II with presence of minute zircon inside the quartz crystals (Figure 3a,b). The rutile samples had variable Zr concentrations ranging from 50 to 519 ppm ( Table 2). Using the rutile Zr concentrations and equation of [14] without considering the pressure effect, the temperatures calculated for these samples were between 482 and 644 °C. Based on the estimation of fluid pressure (~90 bar) constrained by previous fluid inclusion analyses [23], the pressure-corrected temperatures were calculated to be 488-646 °C (Figure 7) using the experimentally calibrated method of Tomkins et al. [15], and 434-610 °C using the alternative method proposed by Kohn [54].  [14,15,54]. For comparison, the homogenization temperatures estimated from quartz-hosted fluid inclusions [23] and temperatures yielded by chlorite thermometry are shown [55].

Chlorite Geothermometry
Chlorite group minerals from Tuwu were commonly intergrown with or occur later than the hydrothermal rutile of stage II (Figure 4). The major elements in chlorite coexisting with rutile were determined in this study, and the results are presented in Table 3. They yielded consistent SiO 2 (25.40-26.51 wt %), TiO 2 (0.09-0.24 wt %), Al 2 O 3 (20.10-20.63 wt %), and FeO (14.92-22.79 wt %) contents. The (Na 2 O + K 2 O + CaO) contents were lower than 0.5 wt %, indicating negligible contamination [56][57][58]. The chemical formula of the analyzed chlorite was calculated based on 14 oxygen atoms per formula unit (a.p.f.u). The Si (a.p.f.u) and Al (a.p.f.u) values for chlorite range from 2.65 to 2.76 and 2.45 to 2.56, respectively ( Table 3). The precipitation temperatures calculated for chlorite ranged from 308 to 330 • C (T 87 ; Table  3), using semiempirical thermometry [59]. Alternatively, the temperatures were calculated to be in the same range of 338 to 372 • C (Figure 7), using the other two different empirical equations from [60] for T 88 and [61] for T 91 .

Precipitation of Hydrothermal Rutile with Copper Mineralization
Rutile (TiO 2 ) is a common accessory mineral in igneous, metamorphic, and sedimentary rocks, which also occurs as a secondary alteration mineral in magmatic-hydrothermal ore deposits [3,62]. In calc-alkaline porphyry-related systems, high pressure (1.3-1.5 GPa) is required for primary rutile crystallized from magma. Given the shallow formation depth of porphyry deposits, rutile generally formed during hydrothermal alteration instead of magmatic processes [8,[63][64][65][66]. The origin of rutile can be distinguished through its paragenetic mineral assemblages, textures, and chemical compositions [67,68].
Primary rutile is commonly interstitial, intergrown with pyrite and/or phosphate minerals (e.g., monazite and xenotime), or rimmed by K-feldspar, quartz, ilmenite, and magnetite [2,4]. In contrast, rutile from the Tuwu porphyry Cu deposit mostly coexists with hydrothermal quartz, chlorite, pyrite, and chalcopyrite (Figure 3), and occurs as aggregates in mineralized porphyries (Figure 4). These features are similar to those of hydrothermal rutile associated with porphyry-type Cu-Mo and various types of gold mineralization, such as El Teniente porphyry Cu-Mo [8], Zhesang Carlin-type Au [6], orogenic Au from Precambrian terranes [7], and Wulong lode Au [68] deposits. Hydrothermal rutile can be distinguished from primary rutile hosted in igneous and metamorphic rocks with its high abundance of W, V, Fe, Cr, and Sn [6,11,16,69,70]. Enrichment of V (V 2 O 3 = 0.22-0.76 wt %), Fe (FeO = 0.03-0.43 wt %), and Sn (SnO 2 = 0.01-0.04 wt %) contents in rutile from Tuwu (Table 1) suggest that it is hydrothermal in origin. In addition, rutile displays U-shape REE patterns and has relatively high REE content for unaltered rutile (type 1; Figure 6a), which differs from igneous rutile [67,68]. As for type 2a and b rutile from Tuwu (Figure 6b,c), depletion of REE, especially for LREE, may result from post modification by later hydrothermal fluids due to higher mobilities of LREE than HREE. Taken together, the Tuwu rutile is interpreted to be hydrothermal and has spatial and genetic associations with Cu mineralization.

Origin and Substitution Mechanisms of Ti in Rutile
Rutile is generally produced by breakdown and/or re-equilibration of Ti-rich and/or Ti-bearing minerals during water-rock interaction in porphyry Cu systems [8,11,19,20,71]. It occurs predominantly in potassic and phyllic alteration zones associated with chalcopyrite and molybdenite-bearing quartz veins. Given the low solubility of titanium in ore-forming fluids at temperatures below 700 • C [72], fluids cannot be a significant Ti source of the hydrothermal rutile at Tuwu. Four possible processes have been reported for rutile formation driven by the introduction of S and CO 2 in fluids [11,73], including (1) biotite altered to phlogopite, (2) titanomagnetite altered to magnetite, (3) ilmenite altered to rutile in areas of intense alteration, and (4) FeTiO 3 (ilmenite) + S 2 = FeS 2 + TiO 2 (rutile) CaTiSiO 5 (titanite) + CO 2 = TiO 2 (rutile) + CaCO 3 + SiO 2 (4) No primary ilmenite and titanite were detected in porphyries at Tuwu. Hence, it is more likely that biotite and magnetite released Ti for rutile generation as a function of high temperature and influx of S-rich oxidized fluids [28,49]. Pan et al. [49] and Rui et al. [46] determined the compositions of primary biotite, and the results show that biotite has Mg/(Mg + Fe + Mn) values ranging from 0.35 to 0.60. Petrographic observations of the earlystage biotite replaced by later hydrothermal rutile (Figure 4a,b) and chlorite further suggest that biotite can be a source of Ti for rutile formation [49]. This interpretation is consistent with our unpublished EPMA results of hydrothermal biotite from the potassic alteration zone, in which the biotite was Fe-rich and had TiO 2 ranging from 0.10-1.11 wt %. Moreover, the magmatic/early-stage magnetite at Tuwu is Ti-bearing (TiO 2 = 0.04 to 1.09 wt %), as supported by EMPA results reported by Yuan et al. [24], implying that magnetite can be a candidate for releasing some Ti into hydrothermal fluids. During early potassic alteration, we propose that rutile was partially formed by breakdown and re-equilibration of primary biotite, which released residual Ti by exsolution under fluid-dominated conditions [74]. The Ti-magnetite breakdown also contributed some Ti, resulting in the formation of Ti-poor magnetite, pyrite, and rutile (reaction 2) [8,63]. In the phyllic and subsequent propylitic stages, owing to the instability of biotite [8], quartz-chlorite-rutile assemblage replaced the pre-existing biotite, during which abundant sulfides, mainly including chalcopyrite, pyrite, and molybdenite, precipitated from the hydrothermal system. This is consistent with the intimate relationships of rutile with copper mineralization and chlorite alteration in the study area (Figure 4).
The LA-CP-MS and EPMA analyses of rutile indicate that cations such as Fe, Ta, Nb, V, W, Zr, Sn, Ta, and Cr with different valences potentially substitute for the main Ti sites in rutile (Tables 1 and 2; Figure 6d), which dominantly depends on their ionic size, charge neutrality, and cation mobility [75][76][77][78]. In the Tuwu porphyry Cu deposits, significant Sn 4+ , some high field strength elements (HFSE; e.g., Zr 4+ and Hf 4+ ), and minor Mo 4+ may directly replace for Ti 4+ (r = 0.67 Å) in rutile. Furthermore, two coupled-substitution mechanisms were proposed for trivalent, pentavalent, and hexavalent cations incorporating into the rutile structure [8,11] according to the following coupled substitutions (1) (Fe, V, Cr, Sc) 3+ + (Nb, Ta) 5+ = 2 Ti 4+ , and (2) 2(Fe, V, Cr, Sc) 3+ + W 6+ = 3 Ti 4+ . Considering the excess total content of trivalent ions compared to Nb 5+ , Ta 5+ , and W 6+ ions, the incorporation of hydrogen in the rutile structure has been suggested for charge-balancing [79]. Low Cu and Mo concentrations in rutile indicate that these two elements are more incompatible with respect to hydrothermal rutile (Figure 8a,b), although previous experimental studies reported the D Mo rutile / fluid value could be up to 1.5 [74]. Importantly, it is easier for Mo than Cu to exchange Ti in Ti-rich mineral phases, as it can directly substitute Ti 4+ as Mo 4+ under low f O 2 condition, or as Mo 6+ via coupled substitutions such as W 6+ under oxidized systems [74]. Rutile at Tuwu contains some Mo, but only up to 110 ppm (Figure 8a), which is controlled by Mo content in Ti-rich parental minerals and in hydrothermal fluids [8]. Previous studies have demonstrated that the ore-related porphyry melt at Tuwu was hydrous and highly oxidized without significant early sulfide saturation prior to fluid exsolution [28]. In this case, more Mo would have been removed from silicate melts into aqueous fluids, which would have limited the amount of Mo partitioned into Ti-bearing magmatic mineral phases. The oxidized, Mo-bearing hydrothermal fluids can be an alternative Mo source for hydrothermal rutile, which is supported by the low measured

Application of Zr-in-Rutile Thermometer to Magmatic-Hydrothermal Ore System
The zircon-in-rutile temperatures calculated in this study were based on the measurements of Zr contents in rutile and empirical equations from [14,15,54]. They yielded a variable formation temperature of 434-646 °C with an average of 531 °C (Table 1), largely higher than those estimated by chlorite geothermometer (Figure 7) [55], as well as fluid inclusion analyses for Tuwu [23]. Several factors are potentially responsible for the overestimation of temperatures from rutile, including (i) absence of quartz in veins, (ii) incorporation of Zr-rich mineral inclusions (e.g., zircon and zirconolite), and (iii) excess Zr incorporation into rutile caused by a high crystal growth rate [7,68,85]. In the case of Tuwu, petrographic observations support that the rutile is related to quartz-bearing veins, and thus the first factor can be excluded. No zircon inclusions (ZrSiO4) were identified in the Inoue et al. [83] proposed that redox conditions (logf O2 ) can be estimated following the method of Walshe [84], based on a semiempirical thermometer [55]. The logf O2 for the main stage calculated from chlorite compositions varied from −28.5 to −22.1, which is consistent with a weakly oxidized environment. In addition, the precipitation temperature was calculated to be 308 • C to 372 • C (average 345 • C) using the chlorite geothermometer [59][60][61], which overlapped with the temperature (290-400 • C) determined by fluid inclusion measurements (Figure 7) [23]. Therefore, we suggest the main ore stage at Tuwu with the formation of Cu-bearing sulfides and associated chlorite occurred in intermediate temperature and weakly oxidized conditions.

Application of Zr-in-Rutile Thermometer to Magmatic-Hydrothermal Ore System
The zircon-in-rutile temperatures calculated in this study were based on the measurements of Zr contents in rutile and empirical equations from [14,15,54]. They yielded a variable formation temperature of 434-646 • C with an average of 531 • C (Table 1), largely higher than those estimated by chlorite geothermometer (Figure 7) [55], as well as fluid inclusion analyses for Tuwu [23]. Several factors are potentially responsible for the overestimation of temperatures from rutile, including (i) absence of quartz in veins, (ii) incorporation of Zr-rich mineral inclusions (e.g., zircon and zirconolite), and (iii) excess Zr incorporation into rutile caused by a high crystal growth rate [7,68,85]. In the case of Tuwu, petrographic observations support that the rutile is related to quartz-bearing veins, and thus the first factor can be excluded. No zircon inclusions (ZrSiO 4 ) were identified in the analyzed rutile samples, excluding the influence of mineral inclusions on temperature estimations. The Zr and Si concentrations in rutile are up to 519 ppm and 31,072 ppm, respectively ( Table 2), but are negatively correlated, further indicating the absence of zircon inclusions. In addition, the correlations between the Ca and Zr are weak (Table 2), differing from the case for Zr-rich rutile in the study of [7], which suggests absence of zirconolite (CaZrTi 2 O 7 ).
The best reason for overestimated temperatures at Tuwu is excess Zr 4+ substituting for Ti 4+ in rutile. Agangi et al. [7] proposed that the equilibrium between rutile, zircon, and quartz is a prerequisite to use the rutile thermometer, a condition that can be met in slowly heating and cooling geological processes, especially at medium to high temperatures, under which the element diffusion is weak. Existing fluid inclusion studies have revealed the occurrence of a boiling event during the main ore stage [23], evidenced by the coexisting low-salinity vapor-rich and moderate-to low-salinity liquid-rich inclusions within individual fluid inclusion assemblage hosted in quartz. The rutile from the Tuwu porphyry Cu deposit is suggested to be generated in an intensively boiled magmatichydrothermal system that is characterized by pulsating fluid flow at rapidly changing temperatures [23], which would have promoted excess Zr incorporating into rutile crystal. Besides, the occurrence of fluorapatite suggests the ore-forming fluids at Tuwu are F-bearing [27,47,86], which can dissolve, transport, and precipitate remarkable quantities of Zr and Ti [87][88][89][90]. Similar halogen-rich (e.g., F and Cl) aqueous fluids have been reported in other deposits that significantly affect redistribution of Zr in rutile [68]. Large variations in temperatures estimated from the Zr-in-rutile geothermometry have been addressed in other studies [7,18,68,91]. Therefore, the application of a Zr-in-rutile thermometer to magmatic-hydrothermal ore systems should be approached with caution. Detailed microinvestigations coupled with multiple mineral thermobarometers would be helpful for evaluating formation temperatures and other conditions.

Implications for Mineralization Potential and Exploration
In recent years, rutile geochemistry has attracted more and more attention as an indicator or fertility tool for mineral deposits, especially when sulfides from deposits have been leached out, as they can retain chemical signatures through post-weathering and most hydrothermal events [3,11,16,62,63,92]. For example, previous work on the E26N porphyry Cu-Au deposit in Australia indicates that rutile grains from proximal orebody (less than 100 m) are larger in size (length × width > 4000 µm 2 ) and enriched in V (commonly > 2000 ppm), compared to those from distal locations (length × width < 1500 µm 2 ; V concentration < 1500 ppm) [11][12][13]16]. Such rutile is the product of relatively early hydrothermal fluid activity close to the potassic zone, corresponding to some Cu mineralization in porphyry-style systems [11,19].
In this study, petrographic and BSE observations on rutile from the Tuwu porphyry Cu deposit show that the rutile collected from proximal orebody (Figure 2) had large sizes of 30 to 80 µm (Figure 4). Furthermore, we compiled available chemical compositions of rutile from 12 mineral deposits in different genetic types and metal associations, and sediments and metamorphic rocks (Figure 10), including Duobuza porphyry Cu, Duolong porphyry Cu, Zhunuo porphyry Cu-Mo, Dasuji porphyry Mo, Northparkes porphyry Cu-Au, Panasqueira wolframite, Zhesang Carlin-type Au, Precambrian orogenic Au, Wulong lode Au, barren granite-related hydrothermal rutile in implacer from Menderes massif, Dulan eclogite-type rutile deposits, and siliciclastic rocks from Capricorn orogen. The results suggest that rutile from porphyry Cu and hydrothermal W deposits generally contain higher V concentrations (up to approximately 8000 ppm for porphyry deposit; Figure 10). Elemental mapping of rutile from Tuwu indicates homogenous and enriched V distribution ( Figure 5). In contrast, rutile from Au-dominant deposits (e.g., lode, Carlin, and orogenic-type) have variable V concentrations, and rutile formed in sedimentary rocks and metamorphic deposits unrelated to metal mineralization record very low V content ( Figure 10). Plotting samples of rutile from the Tuwu porphyry Cu and Wulong lode Au deposits in the Ti − 100 × (Fe + Cr + V) − 1000 × W ternary diagram (Figure 8d) indicate that the porphyry-related rutile has intermediate W and Sn content (Sn = 73-346 ppm). In addition, the Cr + V/Nb + Ta ratios of rutile from porphyry Cu deposit, such as Tuwu, are higher than those of rutile formed in porphyry Mo or other Au types and rutile deposits (Figure 8c), which can be used as an indicting proxy to discriminate porphyry Cu systems. In summary, we suggest that hydrothermal rutile from porphyry Cu deposits is V-rich, has intermediate W + Sn contents, and is high in Cr + V/Nb + Ta, criteria that can be used as critical indicators for copper exploration. addition, the Cr + V/Nb + Ta ratios of rutile from porphyry Cu deposit, such as Tuwu, are higher than those of rutile formed in porphyry Mo or other Au types and rutile deposits (Figure 8c), which can be used as an indicting proxy to discriminate porphyry Cu systems.
In summary, we suggest that hydrothermal rutile from porphyry Cu deposits is V-rich, has intermediate W + Sn contents, and is high in Cr + V/Nb + Ta, criteria that can be used as critical indicators for copper exploration. Figure 10. Vanadium contents of rutile of the Tuwu porphyry copper deposit compared with several other deposits. The compiled porphyry deposits include Duobuza Cu [12], Duolong Cu [93], Zhunuo Cu-Mo [20], Dasuji Mo [94], and Northparkes Cu-Au [11]. The hydrothermal W deposit includes the Panasqueira deposit in Portugal [17]. The Au deposits include the Zhesang Carlin-type Au [6], the orogenic Au deposit in Precambrian terranes [7], and the Wulong lode Au [68]. Data from sedimentary rocks are from Menderes massif [70] and Capricorn orogen belt [95]. The metamorphic deposit includes Dulan eclogite-type rutile [9].

Conclusions
(1) Rutile in the Tuwu porphyry Cu deposit generally occurs as individual crystals and aggregate grains coexisting with hydrothermal quartz, chlorite, pyrite, and chalcopyrite. It contains an abundance of high V, Fe, and Sn, and displays flat LREE-MREE patterns, suggesting a hydrothermal origin and association with Cu mineralization.
(3) Titanium for rutile formation was made available by the breakdown and re-equilibration of primary biotite and Ti-magnetite in phyllic and subsequent propylitic stages. Sn 4+ , some high field strength elements, and minor Mo 4+ may directly replace for Ti 4+ in rutile.

Conclusions
(1) Rutile in the Tuwu porphyry Cu deposit generally occurs as individual crystals and aggregate grains coexisting with hydrothermal quartz, chlorite, pyrite, and chalcopyrite. It contains an abundance of high V, Fe, and Sn, and displays flat LREE-MREE patterns, suggesting a hydrothermal origin and association with Cu mineralization.
(3) Titanium for rutile formation was made available by the breakdown and reequilibration of primary biotite and Ti-magnetite in phyllic and subsequent propylitic stages. Sn 4+ , some high field strength elements, and minor Mo 4+ may directly replace for Ti 4+ in rutile.
(4) Rutile of larger grain size and high V (up to 8000 ppm), intermediate W + Sn, and high Cr + V/Nb + Ta, is typical of porphyry Cu mineralization. The combined parameters can be considered as useful proxies for ore exploration in Eastern Tianshan and other regions.
Author Contributions: Conceptualization, F.Z.; sample collection, Y.L., H.Z. and Y.C.; methodology and data curation, X.W., M.S. and W.Z.; writing and organizing the paper, X.W.; writing, review, and editing, F.Z. and Y.W. All authors have read and agreed to the published version of the manuscript.