Introduction

Cooling of sulfur-rich oxidized hydrous magma and exsolution of hydrothermal fluids in the middle to upper crust are the primary controls on porphyry Cu-Au systems (Burnham 1980; Blevin and Chappell 1992; Hedenquist and Lowenstern 1994). The solubility of Cu and Au in hydrous silicate melts was proven to be affected by the amount and proportion of dissolved sulfate and sulfide in S-bearing magma, whereas oxygen fugacity controls the speciation of sulfur (Barnes 1979; Hamlyn et al. 1985; Bornhorst and Rose 1986; Jugo 2004; Jugo et al. 2005). The magma oxidation state can be constrained by empirical redox proxies (Soloview 2015). Typical methodologies, such as whole-rock Fe3+/Fe2+ ratio measurements (Brett and Sato 1984), amphibole oxygen thermobarometry (Scaillet and Evans 1999; Ridolfi et al. 2010; Ridolfi and Renzulli 2012), and ilmenite-magnetite oxygen thermobarometry (Buddington and Lindsley 1964; Ghiorso and Sack 1991; Lepage 2003), are all adequate in evaluating the magmatic redox state of unaltered intrusive rocks. However, hydrothermal alteration and magmatic hydrothermal ore formation invariably affect the primary texture and bulk-rock chemistry of Cu-Au porphyry intrusions and may conceal the original igneous rock signature. Accessory minerals such as zircon can better retain the primary chemical information due to slow growth and dissolution (Watson 1996) and low intracrystalline diffusivities (Cherniak et al. 1997). Thus, chemically stable zircon populations are well suited to record the evolution of a magmatic system, and the associated fluctuations in f O2-T conditions.

The Sanjiang orogenic belt in southwest China hosts several large Cu-Au porphyry deposits that formed in distinct tectonic settings (Mao et al. 2014). These include the subduction-related Triassic Pulang porphyry Cu deposit in the Zhongdian arc (the southern part of Yidun arc), and the post-collisional Tertiary Beiya porphyry-skarn Au-Cu deposits within a potassic igneous belt along the Ailaoshan-Red River shear zone. The two deposits are the focus of our study. We attempt to evaluate the f O2-T conditions and magma evolution through the trace element concentration of zircons that crystallized in the two porphyries. We then compare these data with other published global data to propose a discrimination diagram for ore-bearing and ore-barren porphyries.

Geological background

Yidun arc

The NNW-trending Yidun arc is surrounded by the Qiangtang terrane to the west, Songpan-Garze fold belt to the northeast, and Yangtze craton to the southeast, respectively. The Garze-Litang and Jinshajiang suture zones lie to the east and west of the Yidun arc, respectively (Fig. 1). The Zhongza massif in the western part of the Yindun arc, divided by the NS-trending Xiangcheng-Geza fault, comprises carbonate-dominated Paleozoic metasedimentary rocks intercalated with mafic volcanic rocks. It is suggested to have been separated from the western Yangtze craton by opening of the Garze-Litang Ocean in the Late Permian (Chen et al. 1987; Chang 1997; Hou et al. 2003). In contrast, the eastern Yidun arc is dominated by Late Triassic volcano-sedimentary successions consisting of flysch and mafic-felsic arc-related volcanic rocks (BGMRSP 1991). The Late Triassic through Miocene dioritic-granitic plutons extend for more than 500 km from north to south, and become progressively younger towards the east (BGMRSP 1991; Hou et al. 2003).

Fig. 1
figure 1

(a) Geological map showing the distribution of major continental blocks and sutures in southeast Asia (Metcalfe 2006; Metcalfe 2013). (b) Tectonic framework of the Sanjiang domain in southwest China, showing the major terranes, suture zones, volcanic arcs, and location of the Pulang porphyry Cu-Au deposit and the Beiya porphyry-skarn Au-Cu district (Zi et al. 2012; Deng et al. 2014). Original figure from Zhu et al. (2015), and with publication permission from Economic Geology

Full size image

The Yidun arc underwent three main tectonic, magmatic, and metallogenic events during Triassic subduction of oceanic crust, Jurassic-Cretaceous collision, and post-collision, and then affected by Cenozoic intracontinental strike-slip shearing and convergence (Mo et al. 1993; Li et al. 2011; Deng et al. 2014). A continuous glaucophane schist belt (Sha 1998) and an ophiolitic suite along the suture zone (Zeng et al. 2004) are products of the subduction. The NW-oriented Triassic stocks and dikes comprise hypabyssal quartz diorite porphyry, quartz monzonite porphyry, and granite porphyry. Some of these are the causative intrusions for Cu ± Mo ± Au mineralization. These adakite-like calc-alkaline porphyries (e.g., Pulang, Xuejiping, and Langdu) resulted from the Triassic westward-dipping subduction of the Garze-Litang Ocean below the Yidun arc (Pan et al. 2003; Li et al. 2011; Wang et al. 2014a). The collision between the Yidun arc and the Yangtze craton took place in the Jurassic and Cretaceous following final closure of the Garze-Litang Ocean. The Late Cretaceous A-type granites in the northern Yidun arc are associated with W-Sn mineralization, whereas Cu-Mo mineralization is mainly related to the coeval magmatism in the southern Yidun arc. The Late Cretaceous magmatism in the southern Yidun arc formed in a post-subduction setting. However, different viewpoints on the controls of the magmatism include the collision between the Lhasa and Qiangtang terranes (Wang et al. 2014b; Peng et al. 2014) or lithospheric-scale transtensional faulting within the Yidun arc (Yang et al. 2015) during Late Cretaceous.

Jinshajiang-Ailaoshan shear zone

Lateral tectonic escape caused by the collision between India and Asia resulted in the formation of the Ailaoshan-Red River shear zone in Paleocene (Tapponnier et al. 1990; Leloup et al. 1995; Yin and Harrison 2000). Strike-slip motion on the shear associated with lithospheric extension resulted in the emplacement of the Eocene-Oligocene alkaline magmatic belt that is 2000-km long and 50–80-km wide (Chung et al. 2005). Contemporaneously, formation of a series of Tertiary rift basins across the Qiangtang terrane and the western Yangtze craton, extrusion of alkaline basalt, and development of a positive gravity anomaly pattern along the Jinshajiang-Red River fault system took place. These formed in a post-collisional extensional setting (Zhang et al. 1987; Turner et al. 1996; Deng et al. 2010), or transpressional setting during continental subduction (Wang et al. 2001) that possibly corresponded to the opening of the South China Sea (Leloup and Kienast 1993; Liang et al. 2007). The Eocene-Oligocene alkaline-rich intrusions are associated with many porphyry and skarn Au-Cu deposits (Hu et al. 2004; Xu et al. 2007; Lu et al. 2013; Tran et al. 2014). Three typical porphyry suites (Beiya, Yao’an, and Machangqing) are located to the east of Jinshajiang suture, which marks the closure of the Jinshajiang Ocean in the Late Triassic (Mo et al. 1993), whereas the Eocene Yulong porphyry copper belt is located to the west along the Jinshajiang suture.

Deposit geology

Pulang Cu deposit

The Pulang Cu deposit (99° 59′ 23″ E, 28° 02′ 19″ N, elevation 3500–4600 m) is hosted by a series of Late Triassic subvolcanic intrusions emplaced into the Late Triassic clastic rocks and andesite of the Tumugou Formation (Figs. 2a and 3). The deposit contains inferred reserves of 1.3 Gt with an average grade of 0.34% Cu, 0.18 g/t Au, and 1.27 g/t Ag (Li et al. 2011). It is locally controlled by NW-trending faults. The Pulang igneous stocks comprise of quartz diorite, quartz monzonite, granodiorite, and minor monzodiorite covering an area of 8.9 km2.

Fig. 2
figure 2

Simplified geological map of the Pulang deposit area (a) and the Beiya district area (b) showing the distribution of felsic intrusive phases and the orebodies, modified from Li et al. (2011) and Yunnan Gold & Mineral Group Co., Ltd (2011), respectively. Pentagram represents the sampling location. Sample BY04 was collected outside Fig. 2b

Full size image
Fig. 3
figure 3

Southwest-northeast cross section 4 (modified from Li et al. 2011) showing lithological units and the spatial relationship of alteration at the Pulang porphyry copper deposit. Location is shown in Fig. 2a

Full size image

Mineralization in the Pulang porphyry deposit is genetically associated with both quartz diorite porphyry and quartz monzonite porphyry (228–212 Ma) (Wang et al. 2008; Pang et al. 2009; Liu et al. 2013). The quartz diorite porphyry is the most abundant rock and forms nearly 80% of the suite. The rock is gray in color and exhibits a phaneritic matrix of plagioclase (65%), quartz (5%), hornblende (< 1%), and minor K-feldspar, with phenocrysts of plagioclase (25%) and hornblende (< 5%). Phyllic and propylitic alteration are present. The quartz monzonite porphyry forms 20% of the suite. The rock is gray with a phaneritic groundmass of oligoclase (25%), K-feldspar (25%), quartz (15%), and biotite (5%), and phenocrysts of plagioclase (15%), K-feldspar (10%), biotite (5%), and minor quartz. Phyllic, potassic, and silicic alteration are present. A post-mineral granodiorite porphyry dike cuts the quartz monzonite porphyry dike. The granodiorite porphyry contains phenocrysts of plagioclase, K-feldspar, biotite, quartz, and hornblende in a matrix of plagioclase, quartz, hornblende (15%), and biotite (5%). The orebodies are mainly within the phyllic and potassic alteration zones which are superimposed on the quartz monzonite porphyry and the quartz diorite porphyry (Fig. 3). Disseminated, veinlet, and breccia-related mineralization developed. A detailed petrographical description is available in Liu et al. (2015b) and Li et al. (2011).

Beiya Au-Cu district

The Beiya Au-Fe-Cu porphyry and skarn district (100° 08′ 38″ E, 26° 07′ 43″ N, elevation 1700–1900 m) is spatially and temporally associated with several small and buried Eocene-Oligocene potassic porphyritic intrusions covering a total area of 0.34 km2 (Figs. 2b and 4). The estimated ore reserves are 130 Mt at a grade of 2.42 g/t Au, 0.48% Cu, 25.5% Fe, 38.85 g/t Ag, 1.24% Pb, and 0.53% Zn (He et al. 2016). The Beiya ore district is located along the limbs of the N-S-trending Beiya syncline. It mainly consists of two mineralized zones, with the Wandongshan, Hongnitang, and Dashadi deposits on the west limb of the syncline, and the Weiganpo and Bijiashan zones on the east limb. The peripheral Matouwan and Bailiancun deposits occur in the periphery of the district. The Wandongshan deposit is the currently the location of the main workings and is developed by an open pit (Figs. 2b and 4b). It yields supergene Au ores that were weathered and leached from the hypogene porphyry and skarn ores. Hypogene mineralization is primarily manifested as skarn-type, with subordinate porphyry, fracture-controlled, and quartz (ankerite)-sulfide types (Mao et al. 2017; Xu et al. 2007a).

Fig. 4
figure 4

West-east cross section A-A′ and B-B′ (modified after YunNan Gold & Mineral Group Co., 2011) showing lithological units and the spatial relationship of mineralization assemblages of the Beiya district. Location is shown in Fig. 2b

Full size image

The country rocks in the Beiya district include ca. 260 Ma flood basalts (Xu et al. 2001) of the Upper Permian Emeishan Formation, sandstones and hornfels of the Lower Triassic Qingtianbao Formation, and limestones of the Middle Triassic Beiya Formation (He et al. 2015). The contact zone between the Triassic limestone and alkaline porphyries is the main host for the metalliferous skarn (Fig. 4). Hydrothermal breccia associated with Au-Fe-Cu mineralization is also present at the Hongnitang deposit. The period of this igneous activity is coeval with regional potassic mafic magmatism in the western part of Yunnan province (37–34 Ma; Lu et al. 2013; Deng et al. 2015; He et al. 2015; Liu et al. 2015a).

The Tertiary intrusions in the Beiya district comprise volumetrically abundant quartz monzonite porphyry and quartz syenite porphyry stocks (37.0–34.6 Ma, LA-ICP-MS zircon U-Pb), granite porphyry (36.6–34.7 Ma, LA-ICP-MS zircon U-Pb), and post-mineralization lamprophyre dikes (Deng et al. 2015; Liu et al. 2015a). The porphyry and skarn mineralization is associated with the quartz monzonite and quartz syenite porphyries (Table 1, Fig. 4). The quartz monzonite porphyry at the Dashadi deposit is light gray and porphyritic, with phenocrysts of K-feldspar (30–35%), plagioclase (20–25%), and minor quartz in a groundmass of K-feldspar and quartz. The quartz monzonite porphyry at the Weiganpo, Bailiancun, and Matouwan deposits has high contents of biotite and amphibole phenocrysts. The quartz syenite porphyries at the Wandongshan, Hongnitang, and Bijiahan deposits are gray-white in color and microgranular and porphyritic, with phenocrysts of K-feldspar (35–45%), quartz (5–10%), and minor plagioclase, and a groundmass of K-feldspar and quartz. The main rock phases are described in Liu et al. (2015a). Subordinate porphyry-type disseminated and vein-type Au-Cu mineralization is typically hosted within fractures and small shears (Mao et al. 2017). Pervasive potassic alteration overprinted the intrusions, altering plagioclase to K-feldspar. Silification, and phyllic and propylitic alteration also occur. Overall, porphyry mineralization within these stocks is weak and they lack typical alteration zoning.

Table 1 Mineralogy and location of porphyry samples from Pulang Cu deposit and Beiya Au-Cu district in the Sanjiang orogenic belt
Full size table

The granite porphyry intrusion hosts porphyry-type mineralization with abundant Au-rich pyrite and subordinate chalcopyrite, galena, and other sulfides. Potassic and phyllic alteration has also overprinted the intrusion. The mineralization in the granite porphyry is roughly synchronous with the other porphyry and skarn mineralization in this district (Deng et al. 2015; Liu et al. 2015a). The vein-type pyrite within the granite porphyry intrusions has a higher Au content (1–10 ppm) than the pyrite (0.01–1 ppm Au) from the Fe-Au orebodies and skarn in limestone associated with porphyritic monzogranite (Deng et al. 2015). These two magma pulses share similar geochemical and isotopic signatures to the amphibole xenoliths in the potassic felsic intrusions in the western part of Yunnan province, suggesting an origin from K-rich thickened lower crust during the melting of metasomatized subcontinental lithospheric mantle (Deng et al. 2015; Liu et al. 2015a).

Sample preparation and analytical methods

Samples from the quartz diorite porphyry (PL01, PL02) and quartz monzonite porphyry (PL03) at the Pulang deposit, and quartz monzonite porphyry (BY01, BY04) and quartz syenite porphyry (BY02, BY03, BY05) in the Beiya district were prepared for zircon separation. Zircon separation was conducted by conventional heavy liquid and magnetic separation techniques. Representative grains were handpicked and mounted in epoxy resin, and then polished to reveal their internal structures. Reflected and transmitted light photomicrographs were examined for all zircons, as were cathodoluminescence (CL) scanning electron microscope images, to identify the internal structure of zircons and ensure that the laser ablation analytical points were wholly within the rim or the core. The xenocrystic zircons in the Beiya porphyries (Deng et al. 2015) are characterized by partially overgrown clear core-rim structures, or by corroded textures. Such crystals were carefully removed prior to analysis. Most zircons were previously prepared for laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) U-Pb dating (not reported here) to exclude inherited zircons. Where possible, if the zircon crystal was large enough, multiple analyses were conducted on single grains to evaluate core-to-rim chemical variation. Additionally, melt inclusions and fractures were avoided during the laser ablation analyses of zircons according to details from the CL images. The measured P, Ca, Sr, and Al contents of zircons were also used to monitor sub-surface inclusions. In this study, most zircon analyses have very low P, Ca, Sr, and Al concentrations.

Trace elements, including rare earth element (REE) contents of the zircons, were determined by LA-ICP-MS at the National Research Center of Geoanalysis in Beijing, China. The LA-ICP-MS analyses used a Finnigan MAT ELEMENT II high-solution inductively coupled plasma spectrometer (HR-ICP-MS) with a high-performance interface coupled with a New-Wave UP-193nm Nd: YAG UV laser. Ablation was carried out by a He-Ar gas mixture (flow rate of 0.73 L He/min and 0.85 L Ar/min) and the resulting vapor was combined with argon before delivery into the ICP-MS. The analyses were conducted using a 35-μm spot diameter, a 10-Hz frequency, a 0.25-mJ/pulse power, during a 70-s analysis including 20 s measurement of gas blank, followed by ablation of the sample for approximately 40 s by raster. Each group of ten zircon analyses was bracketed by analysis of standard glass NIST612, NIST 610, and Kl2g to correct for mass bias drift during analysis. The calibration procedure using internal standards and matrix normalization followed Hu et al. (2011). The full data set for minor and trace elements obtained by LA-ICP-MS is provided as ESM I. The f O2-T calculation methods are documented in the ESM II, in which f O2 is calculated through the La-Pr interpolation method. The CeIV/CeIII ratio is estimated through the lattice-strain model that was extensively used in previous studies.

Results

Trace elements

Zircons from the Pulang and Beiya porphyries have Hf concentrations ranging from 8500 to 10,700 ppm and 6900 to 16,700 ppm, respectively (Fig. 5a). Similar to other studies of felsic stocks (Claiborne et al. 2010; Dilles et al. 2015), the hafnium content decreases, and Zr/Hf and Th/U ratios of zircon increase with declining Ti-in-zircon model crystallization temperature (Fig. 5a–c). Zircons in the studied fertile porphyries typically exhibit low light REE and elevated heavy REE contents, with distinctive positive Ce but comparable negative Eu anomalies (Fig. 6). The zircon Ce anomalies of the Pulang porphyries are comparable to those of the Beiya porphyries (Fig. 5d, e). The Eu anomalies are modest and similar among the porphyries, yielding average values between 0.5 and 0.7, but a small proportion of zircons from the Beiya porphyries exhibits lower negative Eu anomalies than the Pulang porphyries (Fig. 6).

Fig. 5
figure 5

Trace element variation diagrams for the studied zircons. a TTi-in-zircon vs. Hf. b TTi-in-zircon vs. Th/U. c TTi-in-zircon vs. Zr/Hf. d Ce/Ce* vs. Eu/Eu*. e Eu/Eu* vs. Hf

Full size image
Fig. 6
figure 6

Chondrite-normalized rare earth element (REE) patterns (a) and primitive mantle-normalized trace element diagrams (b) for zircons from the Pulang and Beiya porphyry stocks. Normalization values are from Sun and McDonough (1989)

Full size image

Estimates of magmatic oxidation state by different methods

The estimated ∆FMQ values (La-Pr interpolation) of the Pulang porphyries are 2.9–4.6 (average = 4.0 ± 1.0, n = 3) which are slightly higher than those of the Beiya porphyries (0.6–3.5, average = 1.9 ± 1.3, n = 5), but are comparable with the results obtained in the published amphibole chemistry (Fig. 7). Our results for the Beiya porphyries resemble those calculated from amphibole chemistry (NNO + 1.7, 660–805 °C; W. Y. He, oral commun., February 24, 2017).

Fig. 7
figure 7

Plot of oxidation state and temperature (logf O2 vs. temperature, at 2 kbar pressure). The points represent average values of each sample in this study or results from the literature. The oxygen fugacities of the Pulang Cu, Machangqing Cu, and Yao’an Au porphyries are compiled for comparison; these values are calculated using the data from Liu et al. (2013) and Bi et al. (2009). FMQ fayalite-magnetite-quartz buffer curve (Huebner 1971), HM hematite-magnetite buffer curve (≈FMQ + 5; Chou 1978), MW wüstite-magnetite buffer curve (Eugster and Wones 1962), NNO nickel-nickel oxide buffer curve (≈FMQ + 0.7; Huebner and Sato 1970)

Full size image

Discussion

Zircon/rock partition coefficients

The ratios of rare element concentrations in zircons versus those of the host rocks are compared with the experimental and natural partition coefficients from previous studies (Fig. 8). The obtained average zircon/rock partition coefficients of the studied porphyry samples are consistent with the D zircon/melt value of Sano et al. (2002) but slightly higher than the D zircon/rock from other studies (Rubatto and Hermann 2007; Hanchar and Van Westrenen 2007; Luo and Ayers 2009). As discussed in Claiborne et al. (2006), Hanchar and Van Westrenen (2007), and Nardi et al. (2013), the estimated REE patterns calculated through D zircon/melt values of Sano et al. (2002) and zircon REE compositions are consistent with the whole-rock REE patterns, indicating the behavior of trace elements in the melt is predictable from partition coefficient mineral/melt (Fujimaki 1986; Green 1994; Luo and Ayers 2009). The consistency of our D zircon/rock values with Sano et al. (2002) may suggest that the whole-rock composition reflects the composition of the magma from which it crystallized. To support this suggestion, we use the calculated average D zircon/rock values of the zircon populations from the published literature for comparison and obtain a good agreement (Fig. 8, ESM-III and reference in). One may argue that the zircon grains could crystallize late during the solidification of felsic magmas, thereby reducing the validity of whole-rock REE values as melt REE composition. However, the phosphorus contents of the studied porphyries are low (< 0.5 wt%), indicating that the effects of other REE-bearing minerals (e.g., apatite and monazite, which are early crystallization phases in magma) on bulk composition should be limited. This result also suggests that the REE elements are mostly controlled by zircon structure and are little affected by undetected mineral inclusions. We thus propose that the melt composition mainly determines the rare element contents in the studied zircons.

Fig. 8
figure 8

Diagrams illustrating the zircon/rock partition coefficients (D) obtained in this study. The gray lines represent the zircons from previous literature (Xin and Qu 2008; Lee 2008; Peytcheva et al. 2009; Guo et al. 2011; Luo et al. 2011; Wainwright et al. 2011; Wang et al. 2012; Simmons 2013; Farmer 2013). Comparative experimental values (zircon/melt partition coefficients) are represented by squares (Bea et al. 1994), circles (Sano et al. 2002), and diamonds (Luo and Ayers 2009). See data in ESM III

Full size image

Zircon as recorder of magmatic evolution

The above discussion provides us insights that the rare element composition of zircon is determined by the composition of the melt coexisting with crystallizing zircon. Using Hf as an indicator of fractionation (Claiborne et al. 2006) and Ti as a proxy for temperature (Ferry and Watson 2007), the rimward decrease of Ti in zircons from the Pulang porphyry samples is consistent with crystallization by oversaturation upon cooling. A systematic decrease in the Zr/Hf ratio and increase in Y, Th, and U concentrations support this conclusion. The Ti-in-zircon temperatures of the Pulang felsic stocks yield a tight range of 704–715 °C, averaging 709 ± 6 °C, whereas those of the Beiya porphyry system exhibit a wider range of 767–793 °C, averaging 780 ± 22 °C (Table 2). The highly variable geochemical parameters (e.g., REE content, CeIV/CeIII, Ce/Nd, Hf) of zircons from the Beiya porphyries, and opposing zonation patterns in single crystals, argue against a simple magmatic evolution of the Beiya porphyries (Fig. 9, Table 3).

Table 2 Ti-in-zircon temperature, CeIV/CeIII (lattice-strain model) and f O2 (La-Pr interpolation) values calculated using zircon chemistry
Full size table
Fig. 9
figure 9

Cathodoluminescence images of several representative large zoned zircons from the Pulang (a–c) and Beiya (d–f) porphyry stocks showing LA-ICP-MS spots and corresponding trace element data. T (°C) = temperature (Ti-in-zircon)

Full size image
Table 3 Representative chemical parameters of core-rim pairs of zircons from Pulang and Beiya porphyries
Full size table

The negative correlation between Ce/Nd and Th/U ratios indicates that zircon fractionation controls Ce content. The CeIV/CeIII ratio inversely correlates with the Th/U ratio, suggesting that elevated Ce may be impacted by the effects of variable oxygen fugacity. The narrow range of CeIV/CeIII, Th/U, and Zr/Hf ratios, as well as the Hf concentrations, are expected in an undisturbed oxidized parental magma reservoir. However, the Beiya porphyries seem to crystallize over a wider range of f O2 and over a longer time interval (wide range of Hf), and the redox state remains constant or tends to become more reduced with Hf increases above 11,000 ppm (Fig. 10b). The opposite trend of CeIV/CeIII versus Hf compared to the typical pattern for the Pulang porphyry stock samples indicates that the magma chamber generating the Beiya porphyry intrusions was perturbed.

Fig. 10
figure 10

Correlation between CeIV/CeIII and (a) Th/U, (b) Hf, (c) T, and (d) Eu/Eu* values. The CeIV/CeIII values are calculated by the lattice strain model (Ballard et al. 2002). The large symbols in Fig. 10d represent the average values of each samples. The boundary values of maximum Ti-in-zircon temperature of unaltered igneous zircon and hydrothermal zircons are following Hoskin and Schaltegger (2003) and Fu et al. (2009). Near eutectic temperatures are after Dilles et al., (2015). (1–2) Chuquicamata and El Abra Cu-fertile (1) and barren porphyries (2) in northern Chile (Ballard et al. 2002). (3) El Teniente Cu-Mo-fertile porphyries in central Chile (Muñoz et al. 2012). (4) Porphyries from Oyu Tolgoi Cu-Au district in southern Mongolia (Wainwright et al. 2011). (5) Quellaveco Cu-Mo-fertile porphyries in southern Peru (Simmons 2013). (6–7) Medet Cu-fertile (6) and barren porphyries (7) in Bulgaria (Peytcheva et al. 2009). (8) Battle Mountain Cu-Au-fertile porphyries and (9) Carlin Au-related porphyries in northern Nevada, USA (Farmer 2013). The porphyry rocks in China are from (10) Dabaoshan Mo-Cu (Li et al. 2012a), (11) Yulong Cu-Mo (Li et al. 2012b), (12) Machangqing Cu (Guo et al. 2011), and (13) Langdu Cu (Jin et al. 2013), (14) Hongniu skarn Cu deposit, and (15) Triassic Cu-fertile porphyry in Zhongdian arc. (16) I-type granite and (17) A-type ore-barren granite, and (18) S-type ore-barren granite in Tibet (Wang et al. 2012)

Full size image

In summary, the rimward increase of Ti-in-zircon model temperature and decrease of CeIV/CeIII ratio in the zircon crystals (Fig. 9d–f), and the larger scatter in all geochemical parameters (Fig. 10) for the Beiya porphyries, are best interpreted as the rejuvenation by hotter, younger, relatively reduced magmas of an initial oxidized magma pool. In contrast, the typical compositional zonation of zircons from the Pulang porphyries mimic simple cooling and chemical differentiation of the host magma (Ginibre et al. 2007).

The thermal and physical state of upper crustal magma chambers may be disrupted by the incremental injection of small magma pulses. The injection of external magma pulses revealed by reversal of chemical (e.g., Ti-in-zircon temperature, CeIV/CeIII) zonation in zircon indicates mass and energy exchange between a magma body with its external environment (Streck 2008). In the Beiya district, the presence of Au-bearing granite porphyry dikes intruding the quartz monzonite porphyry stock (Deng et al. 2015) also supports multiple pulses from a deep magma chamber. The fluctuation of temperature and oxygen fugacity estimated from elemental variation of magnetites cystallized from the different stages of hydrothermal fluids in the Beiya deposits also indicates the introduction of younger melt into the pre-existing unconsolidated magma system (Sun et al. 2017). Hence, we believe that the zircon signature of the Beiya porphyries resulted from a mixing process in the upper crust.

Magmatic f O2 conditions for generating Cu-Au-bearing melts

A high magmatic fluid flux ensures the exsolution of an aqueous volatile phase, which is considered as the sine qua non for magmatic hydrothermal ore-forming systems (Richards 2011). The investigated Cu-Au-related porphyries commonly contain amphibole or biotite phenocrysts and are characterized by high Sr/Y and La/Yb ratios, as well as listric-shaped normalized REE patterns, which indicate high water content (Li et al. 2011; Lu et al. 2013). Another key issue in generating a magmatic hydrothermal Cu-Au deposit is the ability of magma to transport metals into the upper crust.

Magma generated at f O2 ≥ FMQ + 2 (Mungall 2002) or +2.3 (Jugo 2009) would preserve Cu and Au contents in silicate melts due to the instability of sulfide phases. The estimated oxygen fugacity using amphibole chemistry for the Pulang Cu porphyry is NNO + 2.8 (Liu et al. 2013). This result is consistent with the values calculated by zircon chemistry, implying a relatively oxidized magma for the porphyry. The determined zircon CeIV/CeIII values are comparable to those of the Yulong and Machangqing Cu porphyries along the Jinshajiang belt, and the Dexing Cu porphyry in southeastern China (Fig. 10). Our samples also mostly fall close to those of the Chuquicamata-El Abra samples, indicating that the elevated redox state of these porphyry suites favors Cu mineralization.

The qualitative oxygen fugacities of other two adjacent post-collisional porphyry systems generated from 37 to 33 Ma in western Yunnan (i.e., the Machangqing porphyry Cu-Mo system and the Yao’an porphyry Au system) have ever been estimated through other methods (Bi et al. 2009).We re-use the data from Bi et al. (2009) to obtain estimated crystallization temperatures and redox states. The results show the Yao’an Au-fertile porphyry (820 °C, ~FMQ + 0.9) and Machangqing Cu-fertile porphyry (730 °C,~FMQ + 2.7) are also the products of oxidized magma, similar to the Pulang (~709 ± 6 °C, ~FMQ + 4.0 ± 1.0, n = 3) and Beiya porphyries (~780 ± 22 °C, ~FMQ + 1.9 ± 1.3, n = 5) within our defined error range. We thus propose that formation of Cu-Au-bearing porphyries requires elevated oxygen fugacities but varible temperatures.

Evaluating the economic potential of porphyry-related Cu-Au deposits

Magmatic oxidation state controls sulfur speciation and metal solubility in magmas (Hamlyn et al. 1985; Bornhorst and Rose 1986; Jugo 2004; Jugo 2009) and high magmatic water contents ensure exsolution of aqueous volatile phases. The CeIV/CeIII ratio calculated through the lattice-strain model is a common indicator of magmatic oxidation state, and the (Eu/Eu*)N ratio is instrumental in indicating either high oxidation state or magmatic water contents (Richards 2011; Ballard et al. 2002; Dilles et al. 2015). Ballard et al. (2002) examined fertile and barren porphyry stocks and showed that the productive porphyries exhibited relatively high ratios (CeIV/CeIII > 300 and EuN/EuN* > 0.4). However, the later published large amounts of data are not well explained in this discrimination diagram. Based on the compiled database, an optimal more constrained interval for Cu-Au porphyry formation is thus shown in Fig. 10c, d. Most zircons from the porphyries show CeIV/CeIII values lower than 3000, and Eu/Eu*N less than 0.8, and crystallization temperatures ranging from 650 to 850 °C.

Conclusions

  1. (1)

    In contrast to the Pulang Cu-fertile porphyry, the Beiya porphyries formed by at least two different melt pulses. Zircon geochemistry obtained by LA-ICP-MS can be used to estimate magma redox state. The calculated Ti-in-zircon temperatures and FMQ values of Pulang are 709 ± 6 °C and 4.0 ± 1.0 (n = 3), whereas those values for Beiya porphyries are 780 ± 22 °C and 1.9 ± 1.3 (n = 5), respectively.

  2. (2)

    Combined with a compilation of published zircon data, an optimal interval on CeIV/CeIII-TTi-in-zircon and (Eu/Eu*)N diagrams for producing magmatic hydrothermal Cu-Au deposits is identified. We foresee that the diagrams can help reduce risk at the first exploration step for porphyry Cu-Au deposits, therefore decreasing the economic and environmental cost.