Contrasting apatite geochemistry between ore-bearing and ore-barren intrusions of the giant Kalmakyr gold-rich porphyry Cu deposit, Tien Shan, Uzbekistan

The Kalmakyr deposit in Uzbekistan is one of the world’s largest gold-rich porphyry Cu deposits, and mineralization was proposed to be closely associated with the emplacement of the latest intrusive phases, that is, granodiorite porphyry. However, it is still unclear which features of these ore-bearing intrusions are crucial for the enormous metal enrichment and how they correlate with magmatic fertility. In this study, we address these questions by comparing the petrological and in situ major and trace element geochemistry of apatite for the ore-bearing and ore-barren intrusions. Magmatic apatite from the ore-bearing granodiorite porphyry has higher Eu/Eu* (0.50–0.58), Sr/Y (1.09–1.64), and La/Yb (36.19–53.42) ratios than the ore-barren quartz porphyry (0.01–0.02, 0.02–0.03, and 2.01–3.75, respectively) and monzonite (0.20–0.11, 0.36–1.12, and 19.71–56.32, respectively). Moreover, the ore-bearing granodioritic magma has higher S (average of 111 ppm) and Cl (average of 8602 ppm) contents than the magma of quartz porphyry (S contents average of 9 ppm; Cl contents average of 2652 ppm) and monzonite (S contents average of 9 ppm; Cl contents average of 8483 ppm). These features collectively suggest that the ore-bearing granodiorite porphyry has relatively high oxygen fugacity, H2O, Cl, and S contents compared to those from the ore-barren intrusions. By comparing our data with the published dataset from porphyry Cu–Mo and Mo-only deposits in the world, we suggest that the high Cl content and low F/Cl ratio in magmatic apatite might be an important indicator to prospect for gold-rich porphyry Cu deposits.


Introduction
Gold-rich porphyry Cu deposits, an essential source of gold and copper in the world, are usually associated with calc-alkaline magma, accompanied by multiphase intermediate-felsic intrusions (Sillitoe, 2010;Chiaradia, 2014;Lipson, 2014). It is noteworthy that many of these deposits appear to be related to a particular intrusive phase (Cooke et al., 2005;Chiaradia et al., 2013;Chiaradia et al., 2014;Qin et al., 2022).
Although the tectono-magmatic setting is thought to be a prerequisite for the onset of economic mineralization (Cooke et al., 2005;Sillitoe, 2010), questions as to which features of the evolving magma control mineralization potential and how they correlate with diverse metal associations are still not fully understood.
The Kalmakyr porphyry Cu-Au deposit in eastern Uzbekistan ( Figure 1) has the third-largest gold endowment among all porphyry Cu deposits in the world (Cooke et al., 2005;Seltmann et al., 2014;Zhao et al., 2017). It contains measured and indicated resources of 2000 Mt of ore grading 0.4% Cu and 0.6 g/t Au (Golovanov et al., 2005). Previous studies have indicated that Cu and Au mineralization at Kalmakyr is genetically related to the latest intrusive phase, i.e., the late Carboniferous granodiorite porphyry (Golovanov et al., 2005;Zhao et al., 2017). However, the physiochemical conditions, evolution processes, and magmatic fertility of different intrusive phases at the deposit remain enigmatic and require thorough exploration. Apatite is a common accessory mineral in porphyry-associated intrusions and is resistant to physical or chemical weathering and most types of hydrothermal alteration (Sha and Chappell, 1999;Pan and Fleet, 2002;Bouzari et al., 2016;Zhong et al., 2018a;Zhong et al., 2021b). Its geochemical composition can provide diagnostic information on magma composition, physiochemical condition, evolution processes, and magmatic fertility (Bouzari et al., 2016;Chelle-Michou and Chiaradia, 2017;Xu et al., 2021).
This paper presents the results of in situ major and trace element geochemical analyses for apatite from the ore-barren and orebearing porphyry intrusions of the giant Kalmakyr porphyry Cu-Au deposit. The goal was to explore the critical features of the orebearing magma. These results are further integrated with available apatite geochemical data from porphyry Cu-Mo and Mo deposits to define controlling factors that resulted in diverse metal associations in porphyry deposits.

Regional setting and geology of the Kalmakyr deposit
The Tien Shan orogenic belt in Uzbekistan and Kyrgyzstan comprises three suture-bounded tectonic domains, i.e., the northern, middle, and southern Tien Shan (Figure 1; Yakubchuk et al., 2002;Biske and Seltmann, 2010;Zhao et al., 2015;Zhao et al., 2022). These tectonic domains were displaced by the NW-striking Talas-Fergana strike-slip fault during the Mesozoic (Biske and Seltmann, 2010). The Kalmakyr deposit is situated within the Valerianov-Beltau-Kurama volcano-plutonic arc of the Middle Tien Shan (Figure 1; Golovanov et al., 2005). This arc is accepted to be an Andean-type arc that was formed by the northward subduction of the Turkestan Ocean during the late Paleozoic (Seltmann et al., 2011).
Porphyry Cu-Au mineralization at the Kalmakyr deposit is hosted by a late Paleozoic intrusive complex, which consists of late Devonian quartz porphyry (421 ± 4 Ma; U-Pb zircon age), middle Carboniferous monzonite (327.2 ± 5.6 Ma; U-Pb zircon age), and late Carboniferous granodiorite porphyry (313.6 ± 2.8 Ma; U-Pb zircon age; Zhao et al., 2017). The pre-mineralization quartz porphyry is exposed at the southern end of the deposit area ( Figure 2). It contains phenocrysts of quartz, plagioclase, and biotite in a fine-grained quartz and feldspar groundmass (Figures 3A,D). Monzonite also undergoes pre-mineralization (Golovanov et al., 2005;Zhao et al., 2017), although it hosts most of the highgrade mineralization (Figure 2). It crops out in the central and northern sectors and gradually transitions to diorite in the western sector of the deposit area ( Figure 2). Monzonite contains significant

FIGURE 1
Simplified geological map of the Tien Shan orogenic belt showing the location of major copper and gold deposits (modified from (Zhao et al., 2017). The area shown in Figure 2 is marked.

Frontiers in Earth Science
frontiersin.org 02 amounts of coarse-grained plagioclase and K-feldspar, with minor amounts of biotite, hornblende, and quartz (Figures 3B, E). The granodiorite porphyry represents the latest intrusive phase of the deposit and is thought to be part of a larger, deep-seated intrusion with only limited exposure at the surface (Golovanov et al., 2005). This unit comprises phenocrysts of plagioclase, quartz, K-feldspar, hornblende, and biotite within a felsic groundmass ( Figures 3C, F).
Hydrothermal alteration and Cu-Au mineralization occur widely in the deposit area and are associated with the emplacement of the granodiorite porphyry (Golovanov et al., 2005;Zhao et al., 2017). Upward and outward from the barren core of granodiorite porphyry, hydrothermal alteration shows zonation from early K-silicate to the ore-hosting phyllic zone and later the propylitic zone (Golovanov et al., 2005). The phyllic alteration contains the quartz-sericite and quartz-sericitechlorite-biotite zones, with extensive Cu and Au mineralization and high-grade ores spatially related to the latter zone (Golovanov et al., 2005;Turamuratov et al., 2011). Mineralization is characterized by stockwork and veinlets of chalcopyrite, pyrite, and molybdenite, as is common in many gold-rich porphyry Cu deposits (Sillitoe, 2010;Li et al., 2022;Yang et al., 2022;Tang et al., 2023).

Sampling and analytical methods
Intrusive samples with limited alteration were collected from outcrops around the Kalmakyr open pit for apatite geochemical analyses. The granodiorite porphyry (samples KM-2, KM-3, KM-4, and KM-6) is regarded as ore-bearing, whereas quartz porphyry (sample KM-12) and monzonite (samples KM-24, KM-25, and KM-26) represent ore-barren intrusions. All samples were polished to 100μm-thick and -thin sections for petrological and in situ geochemical analyses, respectively. Apatite grains were selected by optical and backscattered electron (BSE) microscopy. Major and trace element geochemistry of apatite grains were analyzed in situ by electron microprobe (EMP) analyses and laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS), respectively.
A JEOL-JXA8100 electron microprobe performed BSE imaging and the EMP analyses of apatite with a 15-kV accelerating voltage, 10-nA beam current, and 3-μm probe beam at the MNR Key Laboratory of Metallogeny and Mineral Assessment, Institute of Mineral Resources, Chinese Academy of Geological Sciences, Beijing. To detect the characteristic X-ray patterns, 20 s of counting time was chosen for Si, Al, Ti, Fe, Ma, Mn, K, Na, Ba, Cl, and S, 40 s for F, and 10 s for Ca and P. Trace element contents in apatite were analyzed by LA-ICP-MS at the  (Golovanov et al., 2005) and (Zhao et al., 2017)].

Frontiers in Earth Science frontiersin.org
Ore Deposit and Exploration Centre, Hefei University of Technology. The analyses consist of an Agilent 7900 ICP-MS instrument and a Cetac Analyte HE laser ablation system with helium (0.9 L min −1 ) as the carrier gas and argon (0.87 L min −1 ) as the compensation gas to enhance the sensitivity, and they mixed via a T-connector before entering the ICP. After testing the gas blank for 20 s, the laser was run for 40 s at a constant repetition rate of 7 Hz with an energy density of 4 J cm −2 and a beam aperture of 30 μm. NIST 610, NIST 612, and BCR 2G, as the external standards, were run after every 10 analyses. Measured by EMP analyses, the Ca concentrations were used as the internal standards for apatite. The detection limits were calculated for each element in each spot analysis. The analysis accuracy of most trace elements is greater than 10%. The data reduction and processing were carried out using ICPMSDataCal software (Liu et al., 2008).

Apatite petrography
Typical microscopic and BSE images of apatite from quartz porphyry, monzonite, and granodiorite porphyry of the Kalmakyr deposit are presented in Figure 4. Apatite grains from quartz porphyry are euhedral to subhedral columnar crystals of 50-200 μm in length, occurring mostly as inclusions in biotite phenocryst ( Figures 4A, B). The majority of these apatite grains display a homogenous texture with a weak core-rim texture and few gaps or cracks in the BSE images ( Figures

Apatite major and trace element composition
Apatite major and trace element compositions from the orebearing and ore-barren intrusions of the Kalmakyr deposit are summarized in Supplementary Table S1. A total of 27 apatite grains, including 7 grains for granodiorite porphyry, 6 grains for quartz porphyry, and 14 grains for monzonite, were selected for EMP analyses and LA-ICP-MS in this study.
Chondrite-normalized rare earth element (REE) patterns of apatite grains from the three intrusive phases all show enrichment of light REE and depletion of heavy REE, with weak positive Ce anomalies and variably negative Eu anomalies ( Figure 6). The total REE contents of apatite from granodiorite porphyry (3929-7428 ppm, average of 4926 ppm) are lower than those from quartz porphyry (9532-13260 ppm, average of 11073 ppm) and monzonite (3032-10199 ppm, average of 6546 ppm). Eu/Eu* of apatite from granodiorite porphyry (average of 0.53) is higher than that from quartz porphyry (average of 0.01) and monzonite (average of 0.14; Figure 7)

Origin of apatite in multiphase intrusions
Magmatic apatite is generally formed in the early stages of the evolving magma, while hydrothermal apatite appears to be formed during the magmatic-hydrothermal transition or later hydrothermal stages (Bouzari et al., 2016). However, only those magmatic apatite crystals without alteration overprinting can provide original information on the parental magma (Xing et al., 2020;Pan et al., 2021;Xing et al., 2021). Most of the analyzed apatite grains from the three intrusive phases are hosted by biotite, hornblende, and plagioclase phenocrysts (Figure 4). BSE images show that the majority of apatite grains are homogeneous, and only some apatite grains from monzonite contain mineral or fluid inclusions ( Figure 4E). These grains are ignored during our in situ geochemical analyses. In this study, almost all apatite grains are plotted in the magmatic apatite field in the SiO 2 vs. MnO diagram (Chen and Zhang, 2018; Figure 5A). Moreover, the Cl contents of all analyzed apatite grains (<1.64 wt %) are significantly lower than those of hydrothermal apatite (usually exceeding 3.0 wt%; Palma et al., 2019). The apatite grains for a single intrusive phase show parallel REE patterns ( Figure 6). It is, therefore, indicated that the analyzed apatite crystals from all three intrusive phases are magmatic apatite and can document the initial information on their parental magma. Although these apatite grains are magmatic, there are subtle differences in the apatite of the three intrusions. In petrography, apatite crystals from granodiorite porphyry are more fine-grained and homogeneous on the surface compared to those from quartz porphyry and monzonite (Figure 4).

Magmatic oxidation state
Oxidized magma has a higher ability to convey metals and sulfur from the mantle to the shallow crust because sulfur in oxidized magma mainly exists as sulfate (SO 4 2-) that is incompatible with chalcophile elements (Jugo, 2009;Richards, 2011;Zhong et al., 2018b;Zhong et al., 2019;Zhong et al., 2021c). The apatite Eu/ Eu* value is an important indicator of magmatic oxygen fugacity since Eu 3+ has a similar ionic radius with Ca 2+ and is easier to substitute Ca 2+ in the apatite lattice than Eu 2+ , which will result in high apatite Eu/Eu* values in the more oxidized magma (Sha and Chappell, 1999;Piccoli and Candela, 2002). At the Kalmakyr deposit, all the analyzed apatite grains show negative Eu/Eu* anomalies ( Figure 8A). However, it should be noted that the Eu/ Eu* values of apatite from the granodiorite porphyry (0.50-0.58,  Ketcham, 2015); and (D) Apatite Cl vs. F plot.

Frontiers in Earth Science
frontiersin.org average of 0.53) are higher than those from the quartz porphyry (0.01-0.02, average of 0.01) and monzonite (0.20-0.11, average of 0.14) but are identical to ore-bearing intrusions from porphyry deposits elsewhere ( Figure 8A). This indicates that the granodiorite magma may be more oxidized than the quartz porphyry and monzonitic magma. This inference is supported by the relatively high Ce 4+ /Ce 3+ values (average 890) for zircon grains from the granodiorite porphyry (Zhao et al., 2017). Feldspar crystallization may also lead to negative Eu anomalies, accompanied by a significant reduction in Sr contents in the coexisting melt (Pan et al., 2021). There is no apparent positive correlation between apatite Eu/Eu* and Sr from a single intrusive phase, which excludes the previous conjecture ( Figure 7). Therefore, the negative Eu anomalies are more likely to be caused by magma in an oxidized state. However, apatite grains from ore-bearing and ore-barren intrusions display similar Ce/Ce* (weakly positive; Figure 8A), indicating that apatite Ce/Ce* does not have a relationship with magmatic oxygen fugacity (Zhong et al., 2018a;Xing et al., 2021).

Magmatic H 2 O content
A high H 2 O content in oxidized magma is necessary for the exsolution of an aqueous volatile phase to release copper and gold (Richards, 2011;Lu et al., 2015;Zhong et al., 2021a). Magma with high H 2 O contents favors hornblende and garnet crystallization in early deep crust fractionation and would suppress plagioclase crystallization, resulting in elevated Sr/Y and La/Yb ratios (Richards, 2011;Nathwani et al., 2020). Many studies have shown that the characteristics of high Sr/Y and La/Yb in the hydrous magma can be recorded by the early crystallization of apatite (Nathwani et al., 2020;Xing et al., 2021). Petrographic observation indicates that apatite in all three intrusive phases occurs dominantly as an inclusion in phenocrysts such as biotite, plagioclase, and hornblende (Figure 4), suggesting that they may have been formed during the early stage of the evolving magma. The apatite in granodiorite porphyry shows relatively high Sr/Y ratios (1.09-1.64) and La/Yb ratios (36.19-53.42). This characteristic resembles ore-bearing intrusions from many other porphyry Cu-Au±(Mo) deposits (Pan et al., 2016;Qu et al., 2021) but it is obviously higher than those of quartz porphyry (Sr/Y = 0.02-0.03; La/Yb = 2.01-3.75) and other orebarren intrusions  Figure 8B). This indicates that the H 2 O content of the granodiorite porphyry is higher than that of the ore-barren quartz porphyry. Although the apatite from the monzonite intrusion has roughly similar Sr/ Y (0.36-1.12) and La/Dy (19.71-56.32) ratios to those from granodiorite porphyry (Figure 8B), the apatite from the  Sun and McDonough (1989).

FIGURE 7
Plot of Sr vs. Eu/Eu* in the apatite of the Kalmakyr deposit. There is no positive correlation between the Sr and Eu/Eu* of apatite from the same intrusion.

Magmatic F, S, and Cl contents
Magmatic volatile compositions play a critical role in porphyry mineralization, largely due to the fact that they can form chemical ligands with metals (Cu and Au) that are favorable for migration and metal enrichment (Piccoli and Candela, 2002;Yardley, 2005;Jugo, 2009;Zhong et al., 2018a). According to many past studies, apatite wrapped in other igneous minerals as inclusions can preserve the initial volatile content of magma (Stock et al., 2016;Zhu et al., 2018;Li et al., 2021;Pan et al., 2021;Zhu et al., 2022). As we have previously discussed, most of the analyzed apatite grains fit this feature and could theoretically record the original volatile content of the magma. The Cl and F contents of parental magma for the three intrusive phases were calculated according to the apatite/melt partition coefficients proposed by Mathez and Webster (2005), whereas the S abundance in the parental magma was calculated according to the apatite/melt partition coefficients proposed by Parat et al. (2011). The results (Table 1) show that the Cl contents of granodioritic magma (average of 8602 ppm) are significantly higher than the magma of quartz porphyry (average of 2652 ppm) but are consistent with the monzonitic magma (average of 8483 ppm). We, therefore, infer that the Cl-rich hydrothermal fluids exsolved from the ore-bearing granodiorite porphyry may have altered the Cl content of apatite from ore-barren monzonite. The Cl-rich hydrothermal fluids not only affect the apatite Cl content but are also capable of carrying or extracting metals from the monzonite intrusion (Zajacz et al., 2012). For the F content, it is generally recognized that F is partitioned more difficultly into aqueous fluids relative to Cl (Stock et al., 2016;Li et al., 2021); thus, the F content of apatite is basically not altered by the late hydrothermal fluids. The F contents of granodioritic magma (average of 6329 ppm) are lower than those of the magmas for quartz porphyry (average of 12010 ppm) and monzonite (average of 8706 ppm), which indicate that F, unlike Cl, possibly has little influence on the extraction and transport of Cu and Au during the magmatic-hydrothermal evolution. Furthermore, the S  contents of granodioritic magma (average of 111 ppm) are higher than those of magma for quartz porphyry (average of 9 ppm) and monzonite (average of 9 ppm). This result is also supported by the development of numerous anhydrite veins in the mineralized zones since anhydrite is an indicator of sulfur-rich and oxidized melts (Golovanov et al., 2005;Zhao et al., 2017;Hutchinson and Dilles, 2019;Li et al., 2021). To summarize, we conclude that the ore-bearing granodioritic magma is characterized by being Cl-and S-rich, and F-poor compared with the magma for the ore-barren quartz porphyry and monzonite.

Exploration indication for gold-rich porphyry deposits
The porphyry deposit contains significant resources of Cu, Mo, and Au, forming different subgroups with diverse metal associations (e.g., Cu-Au, Cu-Mo, and Mo; Sillitoe, 2010). Magmatic volatile (F, Cl, and S) contents have been proposed to be an important controlling factor for their difference (Zajacz et al., 2012); however, which magmatic features are most critical remains unclear. In this study, we attempted to address this question by comparing the volatile contents of apatite from the ore-bearing granodiorite porphyry at Kalmakyr with available data from porphyry Cu-Au, Cu-Mo, and Mo deposits around the world. As shown in Figures 9A, B, the apatite SO 3 and F contents for ore-bearing intrusions in porphyry Cu-Au, Cu-Mo, and Mo-only deposits are essentially similar, suggesting that these contents in magma may not be key factors controlling metal differences. In contrast, the Cl content and F/Cl ratio in apatite vary significantly ( Figure 9C), i.e., gold-rich porphyry deposits contain the highest Cl content (0.4-2.0 wt%) with the lowest F/Cl ratio (0.8-8.0), Cu-Mo porphyry deposits have a moderate Cl content (0.1-0.6 wt%) and F/ Cl ratio (5.0-30), and Mo-only porphyry deposits show the lowest Cl content (0.01-0.12 wt%) with the highest F/Cl ratio (25-360). This variation might be related to the higher solubility of Au relative to Cu and Mo in Cl-rich melts, resulting in elevated Au enrichment in evolving magma (Zajacz et al., 2012). It is, therefore, suggested that the Cl content in evolving magma might be one of the crucial factors controlling the diverse metal associations in porphyry deposits, and a high apatite Cl content (0.4-2.0 wt%) and low F/Cl ratio (0.8-8.0) in ore-bearing intrusions could be used as an important exploration indicator for gold-rich porphyry Cu deposits.

Conclusion
1) Apatite in multiphase intrusions occurs as inclusions within early phenocrysts, with a homogeneous surface and low Cl and moderate SiO 2 contents, indicating a magmatic origin. 2) The ore-bearing granodiorite porphyry at Kalmakyr is characterized by high oxygen fugacity and high H 2 O, Cl, and S contents compared to the ore-barren intrusions. 3) A high Cl content (0.4-2.0 wt%) and low F/Cl ratio (0.8-8.0) in magmatic apatite can be used as an important indicator to prospect gold-rich porphyry Cu deposits.

Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

Author contributions
YL: investigation, analysis of the data, methodology, and writing-original draft. XZ: investigation, conceptualization, sample collection, designing the experiment, supervision, and writing-review and editing. CX: investigation, supervision, and writing-review and editing. BN: investigation and writing-review and editing. JC: conducting the experiment.

Funding
The authors declare that this study received funding from Zijin Mining Group Co., Ltd. The funder was not involved in the study design, collection, analysis, interpretation of data, the writing of this article, or the decision to submit it for publication.  Richards et al. (2017) and Zhu et al. (2018); data for porphyry Cu-Mo deposits are from Xu et al. (2021); and data for porphyry Mo deposits are from Xing et al. (2020) and Xing et al. (2021).

Frontiers in Earth Science
frontiersin.org