Introduction

Apatite is a common accessory mineral and rich in elements such as rare earth elements (REE), Sr, U and Th. It can record magmatic information well after its generation and thus is an ideal mineral that can provide valuable information for the origin and petrogenesis and even mineralization potential of its host rocks (Liu and Comodi 1993; Belousova et al. 2002; Xie et al. 2018). Although apatite is known for its stability in various geological conditions and processes, making it resistant to metamorphism and hydrothermal alteration (Ayers and Watson 1991), its interactions with highly intense hydrothermal fluids can still be identified through geochemistry (Bouzari et al. 2016). Apatite contains a variety of elements and the variation of trace element concentration (e.g. REE, Sr, Mn and Th) in apatite is related to the silica and total alkali contents, oxygen fugacity and aluminum saturation index in magma and can reflect the geochemical properties, petrogenesis and degree of differentiation of its parental magmas (e.g. Belousova et al. 2001; Chu et al. 2009; Miles et al. 2014; Mao et al. 2016). In addition, the F and Mn contents can also determine the degree of differentiation and aluminum saturation of the magma (Belousova et al. 2001; Chu et al. 2009; Mao et al. 2016). Furthermore, the halogen composition of apatite can be used to estimate the content of F, Cl, H2O and the saturation of volatile in the melts (Boudreau and Kruger 1990; Schisa et al. 2015; Bao et al. 2016). Therefore, apatite plays an important role in the study of volatile components of ore-forming parent magma, and also in the understanding of migration and precipitation of ore-forming elements (Heinrich et al. 2004; Williams and Heinrich 2005).

The Yidun terrane in eastern Tibet Plateau is characterized by widespread Triassic magmatism and among them, the porphyry plutons in the Zhongdian arc (southern Yidun Terrane) are closely related to the porphyry Cu polymetallic deposits, while no mineralization was found related to the granitic plutons in the northern Yidun terrane (Fig. 1b; e.g. Liu et al. 2017; Wang et al. 2017a, b, 2018; Dong et al. 2020, 2022a). Ore-bearing porphyries are susceptible to alteration by hydrothermal fluids, which can alter the whole-rock compositions, resulting in the difficulty in constraining the source and magmatic processes of ore-forming magmas by whole rock compositions (e.g. Pan et al. 2016; Xing et al. 2020).

Fig. 1
figure 1

Geological map of tectonic magmatic rocks in research area (after Dong et al. 2020) (a)Regional terranes accreted during Late Palaeozoic time in southeast Asia; (b) Simplified geological map of the Yidun Terrane; (c) Simplified geological map of the Zhongdian arc

Full size image

Apatite can well preserve the compositional information of its host magmas and thus is widely used to study the processes during magma evolution and related mineralization (e.g. Chu et al. 2009; Pan et al. 2016; Chelle-Michou and Chiaradia 2017; Xing et al. 2020). This study shows the mineralogical and geochemical characteristics of apatite from Late Triassic plutonic rocks in the Yidun Terrane, further discussed the differences of petrogenesis and related mineralization between the ore-bearing porphyries and the ore-barren granites.

Geological setting

The Yidun Terrane is located in the eastern Tibetan Plateau (Fig. 1a), and is also a microcontinent hosting abundant porphyry Cu polymetallic deposits in the Sanjiang Tethyan Orogenic Belt, which has been recognized as an important metallogenic belt in SW China (Deng et al. 2014; Dong et al. 2020; Sun et al. 2020; Wang et al. 2020; Dong et al. 2022a, b). The Yidun Terrane is connected to the Qiangtang Terrane and Songpan–Garzê Terrane by the Jinshajiang suture and Garzê–Litang suture, respectively (Fig. 1a). The Yidun Terrane is mainly composed of Zhongza massif in the west, and Triassic volcano–sedimentary successions and N-S-trending granitic plutons (Hou et al. 2007). The Zhongza massif is a microcontinent from the western Yangtze Block during the Late Permian (Fig. 1b; Hou 1993; Xiao et al. 2007).

The granitic plutons are mainly emplaced during Late Triassic and Late Cretaceous. The Late Triassic granitic plutons in the northern Yidun Terrane are mainly diorite, quartz diorite, granodiorite and monzogranite, and were emplaced at 235-206 Ma with no metallic mineralization was found (Hou et al. 2003; He et al. 2013; Peng et al. 2014; Wu et al. 2017; Dong et al. 2022a). By contrast, the Late Triassic granitic rocks in the southern Yidun Terrane are mainly intermediate porphyries emplaced at 230-203 Ma and are closely related to the porphyry Cu polymetallic deposits (e.g., Li et al. 2011, 2017; Liu et al. 2017; Wang et al. 2017b). The Triassic magmatism is related to the westward subduction or the slab roll-back of the Garzê–Litang oceanic slab (Hou 1993; Hou et al. 2003, 2007; Wang et al. 2016; Liu et al. 2017; Pan 2017; Chen et al. 2017b; Wang et al. 2018; Li et al. 2020; Dong et al. 2020, 2022a). The Late Cretaceous granites in the northern Yidun Terrane are N-S trending and spatially located in the west of the Triassic granites belt, while those in the Zhongdian arc are outcropped in relatively small areas and overprinted with the Triassic porphyries (Fig. 1b).

Samples and methods

Samples

Samples collected in the northern Yidun Terrane are from Cuojiaoma and Dongcuo batholith. Cuojiaoma batholith is located in the eastern Changtai County, with a length of about 1,200 km from north to south and about 40 km from east to west. It is mainly controlled by NW- and NE-trending faults and paralleling to the Ganzi-Litang suture (Fig. 1). Cuojiaoma batholith intrudes into the Triassic Lamma Formation, Lanasan Formation, Tumugou Formation, and Qugasi Formation, which are composed of sandstone, shale, limestone, andesite, rhyolite, basalt and dacite (Wang et al. 2017b). The Cuojiaoma batholith is dominated by monzonite granite (e.g. Wang et al. 2017b), which is mainly composed plagioclase (20-25 vol%), K-feldspar (20-25 vol%), quartz (35-50 vol%), biotite (10-15 vol%), and hornblende (3-5 vol%), with accessory minerals including zircon, apatite and magnetite (Fig. 2a). A zircon U-Pb isotope study by laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) showed the age of the monzonite granite is 236±1.9 Ma (Wang et al. 2017b). The apatite from the monzonite granite is mainly present as euhedral or subhedral inclusions in plagioclase and biotite, colorless and transparent, with size of 100-120 μm. The cathodoluminescence (CL) images show that most apatite grains have homogeneous texture with few showing zoning textures (Fig. 2e).

Fig. 2
figure 2

Photomicrographs of (a)Cuojiaoma monzonite granite; (b) Daocheng monzonite granite; (c) Zhongdian quartz-diorite-porphyrite; (d)Zhongdian quartz-monzonite-porphyry and Cathode Luminescence (CL) images of (e)Cuojiaoma apatite; (f)Daocheng apatite; (g)Zhongdian apatite. Abbreviations of minerals: Amp-amphibole, Pl-plagioclase, Bt-biotite, Qz-quartz, Zrn-zircon, Ap-apatite

Full size image

The Dongcuo batholith is located in the southern end of the Late Triassic granitic batholiths (Fig. 1b) and emplaced at ca. 217 Ma (e.g. He et al. 2013; Dong et al. 2022a). The batholith is mainly composed of quartz diorites, granodiorites and monzogranites (Fig. 2b). The Dongcuo batholith, situated north of Daocheng County, stretches in a north-south direction with an exposed area of >2000 km2 (Fig. 1). The Dongcuo batholith coexists with contemporaneous arc-related andesites, while simultaneously intruding into the Triassic volcanic-sedimentary sequence. The surrounding rocks primarily comprise Late Triassic basalt, dacite, rhyolite, volcanic tuff, and volcanic sedimentary rocks. The granodiorite and monzogranite occupy 95% of the outcrop of the Daocheng batholith. The monzogranite consists of plagioclase (15-20 vol%), quartz (20-30 vol%), K-feldspar (20-30 vol%), biotite (10-15 vol%) and accessory minerals including zircon, apatite, titanite and magnetite (Fig. 2b). Apatite is colorless and transparent and mainly present as inclusions in the biotite and plagioclase, with size of 100-150μm. The CL images show homogeneous texture of the apatite grains (Fig. 2f).

Granitic porphyry samples in Zhongdian arc were collected from Langdu, Disuga, Telangyong, Xuejiping and Chundu plutons. They emplaced into the Triassic clastic and volcanic rocks of the Tumugou Formation. The porphyries are characterized by porphyritic textures, with phenocrysts consisting of 15-25 vol% plagioclase, 5-15 vol% amphibole, 3-5 vol% biotite and few quartz. The matrix is composed of plagioclase, biotite, K-feldspar, quartz and accessory minerals including apatite, zircon, titanite and magnetite (Fig. 2c and d). Apatite is colorless or light brown and mainly occurred as inclusions in plagioclase, biotite and amphibole, with size of 50-200 μm. Some apatite grains from the Zhongdian porphyries show alteration textures in the CL images (Fig. 2e).

Analytical methods

The rock samples were crushed to natural particle size, and mineral grains were selected by shaking, electromagnetic and gravity sorting. Then, pure apatite grains were selected under the microscope. The sorted apatite grains were cleaned and mounted into epoxy resin, which was used for cathode-luminescence (CL) photography of apatite and electron probe micro-analysis (EPMA). The major compositions of the apatite were analyzed with a JEOL JXA-8230 electron probe micro-analyser. The analyses were under conditions of an accelerating voltage of 15 kV and electron beam current of 10 nA, and the beam diameter was 10 μm. F, Cl, and Ca were measured first with shorter counting times of 10s for peak after 5s for background, while the rest elements were analyzed subsequently with longer counting times on a 30s for peak and 15s for background. The following natural minerals and synthetic oxides were used for calibration: K-feldspar (K), apatite (P, Ca, F), barite (Ba), plagioclase (Si), celestine (Sr), albite (Al), anhydrite (S), sodium chloride (Cl), magnetite (Fe), rhodonite (Mn), and rutile (Ti). All the raw data were corrected with standard ZAF correction procedures. The analytical uncertainties were <5%.

Trace elements in apatite were measured by LA-ICP-MS, using an Agilent 7900 ICP-MS instrument, with a spot size of 43 μm. The duration for the analysis for each spot is 50 s after measuring the gas blank for 20 s. BCR-2G, BHVO-2G and NIST SRM 610 were used as the external standards. NIST SRM 610 was analyzed as an unknown to monitor the data quality. The Ca content obtained by EPMA was used as internal standard to correct the results from the LA-ICP-MS, and the concentrations of trace element were finally obtained.

Results

Major elements

The apatite from the Cuojiaoma monzonite granite (Cuojiaoma apatite), Daocheng monzonite granites (Daocheng apatite) and Zhongdian porphyries (Zhongdian apatite) have similar CaO contents of 53.24-54.13 wt%, 53.04-53.93 wt% and 52.62-54.54 wt%, respectively. While the Cuojiaoma apatite shows low FeO contents of 0.02-0.09 wt% and MnO contents of 0.00-0.04 wt%, the Daocheng apatite and Zhongdian apatite exhibit higher FeO contents of 0.02-0.14 wt% and 0.00-0.35 wt%, and MnO contents of 0.11-0.24 wt% and 0.00-0.18 wt%, respectively. Apatite has a chemical formula of Ca5(PO4)3(F, Cl, OH) and contains a total of 12 oxygen atoms. Accordingly, the quantities of different cations present in the crystal unit of apatite can be determined by using EPMA data (Saxena 1973). The Ca contents for the Zhongdian apatite are wider, while the Mn contents of Daocheng apatite is higher. The Fe-Ca correlation in apatite is better than the Mn-Ca correlation (Fig. 3).

Fig. 3
figure 3

Correlation diagram of Fe-Ca (a) and Mn-Ca (b) for apatite

Full size image

The Cuojiaoma apatite, Daocheng apatite and Zhongdian apatite have similar P2O5 contents of 41.22-42.50 wt%, 40.49-43.01 wt% and 40.48-42.74 wt%, and SiO2 contents of 0.13-0.35 wt%, 0.05-0.63 wt% and 0.05-0.47 wt%, respectively. The Cuojiaoma apatite and Daocheng apatite show low SO3 contents of 0.00-0.02 wt% and 0.00-0.03 wt%, while the Zhongdian apatite displays higher SO3 contents of 0.01-0.31 wt%. The Si-P correlation in apatite is better than the S-P correlation (Fig. 4).

Fig. 4
figure 4

Correlation diagram of Si-P (a) and S-P (b) for apatite. Same symbols as in Fig. 3

Full size image

The Cuojiaoma apatite has low F and Cl contents of 2.40-2.52 wt% and 0.04-0.11 wt%, respectively. The Daocheng apatite shows similar Cl contents of 0.00-0.11 wt% but higher F contents of 2.63-3.10 wt%. By contrast, the Zhongdian apatite exhibits higher F and Cl contents of 2.46-3.67 wt% and 0.00-0.68 wt%, respectively.

Trace elements

The Cuojiaoma apatite, Daocheng apatite and Zhongdian apatite have similar REE concentrations of 3707-6286 ppm, 2990-8308 ppm and 2678-6692 ppm, respectively (Tables 1, 2, and 3). While the LREE/HREE ratios for the Cuojiaoma apatite and Daocheng apatite are 1.61-4.08 and 1.12-3.26, the Zhongdian apatite exhibits significantly high LREE/HREE ratios of 5.92-23.01, showing obvious differentiation between LREE and HREE (Fig. 5). The Cuojiaoma apatite has Mg concentration of 25.36-38.22 ppm, V of 0.46-5.61 ppm, Sr of 87.23-152.3 ppm, Y of 1061-2692 ppm, Hf of 0.01-0.16 ppm. The Daocheng apatite exhibits similar Sr (74.32-173.2 ppm), Y (1113-3990 ppm) and Hf (0.01-0.10 ppm) concentrations with the Cuojiaoma apatite, while the Mg (35.98-85.00 ppm) and V (0.25-12.37 ppm) contents are higher. The Zhongdian apatite has higher Sr (613.5-3397 ppm), V (11.53-29.19 ppm) and Mg (>100 ppm) but lower Y (221.3-1039 ppm) and Hf (<0.03 ppm) concentrations than the Cuojiaoma apatite and Daocheng apatite.

Table 1 Sampling list of intrusive rocks
Full size table
Table 2 Chemical compositions (EPMA results) and calculated apatite formulae
Full size table
Table 3 Apatite trace element content (ppm)
Full size table
Fig. 5
figure 5

Chondrite-normalized REE concentrations in apatite. Same symbols as in Fig. 3

Full size image

Discussion

Implications for petrogenesis

Apatite is a common accessory mineral in magmatic rocks and can record geochemical properties of its parental magma, while alteration can affect the internal structure as well as the chemical compositions of apatite (Bouzari et al. 2016; Pan et al. 2016). CL images show that the Cuojiaoma apatite, Daocheng apatite and most of the Zhongdian apatite are unaltered magmatic apatite and displaying homogeneous textures, while some of the Zhongdian apatite grains are hydrothermally altered and show irregular crystal shapes and inhomogeneous patches (Fig. 2). All the analyzed apatite grains are unaltered and plotted into the magmatic apatite field, indicating a magmatic origin and the compositions thus can be used to infer the compositional information of the parent magmas (Fig. 6).

Fig. 6
figure 6

Discrimination diagrams of hydrothermal apatite and magmatic apatite (from Chen et al. 2017a). (a) FeO vs. SiO2 and (b) MnO vs. SiO2. Same symbols as in Fig. 3

Full size image

The stability field of Grt increases with pressure while the field of Pl decreases at a given temperature, resulting in more garnet but less plagioclase as residue and melts with adakitic compositions during the melting processes of the deep crust (Moyen 2009; Castro 2013). Adakitic magma displays high Sr, low Y and Yb contents, and a lack of negative Eu anomaly (e.g. Defant and Drummond 1990). The Sr/Y ratios and Eu/Eu* ratios in apatite can be related to the competition with plagioclase in the magma. However, the apatite from the Zhongdian porphyries and the Cuojiaoma and Daocheng granites is mainly included in the plagioclase, biotite and amphibole (Fig. 2), indicating that apatite is an early phase and its trace element concentrations can represent the chemical characteristics of its parental magmas. In the Eu/Eu/ versus Sr/Y diagram (Fig. 7), the Zhongdian apatite is plotted into the field of adakite-like rocks, while the Cuojiaoma apatite and Daocheng apatite are plotted into the non adakite-like field, indicating that the parental magma of Zhongdian porphyries has adakite affinity while the Cuojiaoma and Daocheng granites not. This is consistent with the previous whole-rock geochemical studies of these rocks, which revealed that the porphyries in the southern Yidun Terrane exhibit high Sr/Y values (15-103) with adakitic affinity while the Triassic granites in the northern Yidun Terrane are characterized by low Sr/Y ratios that are comparable to those of typical arc magmatic rocks (Liu et al. 2017; Wu et al. 2017; Wang et al. 2017b; Cao et al. 2018; Wang et al. 2018; Dong et al. 2020, 2022a).

Fig. 7
figure 7

Rock-type determination based on Eu/Eu* and Sr/Y ratios (classification from Sun et al. 2019). Same symbols as in Fig. 3

Full size image

Mg and V are generally enriched in the mantle or oceanic crust, while Hf is mainly enriched in the crust (McDonough and Sun 1995; Rudnick and Gao 2014). The Mg and V concentrations and La/Sm ratios in Zhongdian apatite is relatively high with large variations, while the Hf concentration is low with narrow range (Fig. 8). By contrast, the Mg and V concentrations and La/Sm ratios in Cuojiaoma apatite and Daocheng apatite are relatively low with narrow range, while the Hf content is high with large variation (Fig. 8). The differences in Mg, V and Hf concentrations and La/Sm ratios are likely related to different magma sources. The apatite crystalized from I-type, metaluminous magmas with the contribution of mantle-derived melts commonly has lower Y and higher Ce contents than that from S-type, peraluminous and crust-derived granites (Laurent et al. 2017). The Zhongdian apatite is mainly plotted into the field of mantle-derived granitoids, and some of the apatite grains are in the overlapped area of mantle-derived granitoids and mantle-derived granitoids (Fig. 9a), suggesting that the Zhongdian porphyries derived from a mixed source with both mantle and crustal contributions. However, most of the Cuojiaoma and Daocheng apatite grains are plotted into the field of crust-derived granite, with few are in the overlapped area (Fig. 9a), indicating that the sources of the Cuojiaoma and Daocheng granites are dominantly crustal derived with contributions of a small amount of mantle-derived materials. This is consistent with previous geochemical and isotopic studies, which revealed that the Cuojiaoma and Daocheng granites have enriched Sr-Nd-Hf isotopic compositions, with 87Sr/86Sr(t) values ranging from 0.707280 to 0.708250, εNd(t) values ranging from -7 to -6, and zircon εHf(t) values ranging from -8 to -3, indicating that they were sourced from crustal dominated sources. Conversely, the Zhongdian porphyries displayed depleted Sr-Nd isotopes [87Sr/86Sr(t)= 0.704770-0.706308 and εNd(t) ranging from -3.6 to -1.4], suggesting a blend of both mantle and crustal materials (Peng et al. 2014; Wu et al. 2017; Wang et al. 2017b; Cao et al. 2018; Gao et al. 2018; Wang et al. 2018; Dong et al. 2020; Wu et al. 2020; Dong et al. 2022a).

Fig. 8
figure 8

Correlation diagram of V-Mg (a) and La/Sm-Hf (b) for apatite. Same symbols as in Fig. 3

Full size image
Fig. 9
figure 9

Apatite source determination using Y versus Ce (a) (classification from Laurent et al. 2017) and Sr/Th versus La/Sm (b) (after Ding et al. 2015) plots. Same symbols as in Fig. 3

Full size image

The main host of Sr is plagioclase and its fractionation will decrease Sr concentration in the melt, while the apatite from the studied rocks is mainly included in plagioclase (Fig. 2), indicating that plagioclase fractionation had less influence on the Sr concentration in apatite. The partition coefficients of these elements between apatite and melt are relatively constant, thus they can be used to trace the composition of parental magmas (Prowatke and Klemme 2006; Ding et al. 2015). Sr is more incompatible than Th during the partial melting of mantle wedge induced by the metasomatism of slab fluids, resulting in higher Sr/Th ratios in the produced melts. Contrastingly, the influence of partial melting or fractional crystallization on La/Sm ratios is not significant. Instead, these ratios are primarily determined by the degree of melting of subducted sediments. As the amount of sedimentary melting increases, the range of La/Sm ratios also tends to increase (Labanieh et al. 2012). The Zhongdian apatite has a wide range of La/Sm ratios and Sr/Th ratios, indicating the influences from the slab sediments and fluids. The La/Sm and Sr/Th ratios for the Cuojiaoma and Daocheng apatite are relatively low and plotted in a narrow field, suggesting less influences from the slab sediments and fluids.

In summary, the apatite chemistry reveals that the Zhongdian porphyries were derived from a mixed source with both mantle and crustal contributions, which was influenced by slab sediments and fluids. While the Cuojiaoma and Daocheng granites are crustal derived with a small amount of mantle-derived materials and limited contributions from slab sediments and fluids.

Magmatic conditions and S-Cl concentrations in the melt

The abundances of Eu and Ce in apatite can be used to evaluate the oxidation state of magma, since these elements are easier to replace Ca2+ in the apatite at low valence, while be oxidized to high valence under a more oxidized environment (Belousova et al. 2002). The Eu content in magma may also be influenced by the fractionation of feldspar. Previous studies show that the REE concentrations and Eu anomaly of apatite are related to the magma compositions (e.g. A/CNK ratio; Chu et al. 2009; Ding et al. 2015). The Zhongdian porphyries display similar A/CNK ratios (0.74-1.08) and REE patterns and have no or slight Eu anomaly (Eu/Eu*=0.78-1.08; e.g. Wang et al. 2017b; Dong et al. 2020). The Zhongdian apatite has various Eu/Eu*, indicating that the Eu anomaly of Zhongdian apatite has little correlation with the whole rock Eu anomaly and is mainly controlled by the oxygen fugacity of the magma. The Eu/Eu* for Zhongdian apatite is relatively high (0.28-0.69), and mostly higher than the boundary value between oxidation and reduction (0.3; Cao et al. 2012) with Ce/Ce* [CeN/(LaN×PrN)0.5]close to 1.0, indicating an oxidized environment for the magmas. Whereas, the Cuojiaoma and Daocheng apatite show significant negative Eu anomaly with Eu/Eu* lower than 0.1 and Ce/Ce*=0.78-1.05, indicating a reduced environment (Fig. 10).

Fig. 10
figure 10

Apatite Eu/Eu* versus Ce/Ce* correlation diagram. Same symbols as in Fig. 3

Full size image

The S and Cl in the melt play a key role in transporting ore-forming elements that are important for porphyry Cu-Au deposits. The apatite in the rock samples usually occurs as euhedral crystals included in other silicate minerals such as hornblende, plagioclase and biotite (Fig. 2), indicating that the apatite saturation temperatures are more than 850℃ and it occur as an early phase during the magmatic process (Wang et al. 2017c). The calculations of zircon saturation thermometer and Si-in-amphibole thermometer indicate that the Zhongdian porphyries crystalized at c. 800 ℃ (Dong et al. 2020). Both apatite and zircon were crystalized in the early stage of magmatic evolution and apatite is included in amphibole (Fig. 2), suggesting that the crystallization temperature of apatite is ≥800 ℃. Cl and S are essential for the transport and deposition of ore-forming elements in magmatic-hydrothermal systems. Cl, F, and OH occupy the same crystallographic position in apatite, and their abundance is controlled by the stoichiometry of apatite, which means it is incorrect to simply estimate the Cl concentration in the melt by using the partition coefficient of Cl between apatite and the melt (Chelle-Michou and Chiaradia 2017). By utilizing experimental data from Webster et al. (2009) at 200 MPa and 900 °C, Chelle-Michou and Chiaradia (2017) employed the thermodynamic apatite/melt chlorine partitioning model calibrated by Li and Hermann (2017):

$${Cl}_{melt}(wt\%)=\frac{{x}_{Cl}^{ap}}{{x}_{OH}^{ap}}\frac{1}{{Kd}_{Cl-OH}^{ap-melt}}\times 10.79$$
(1)
$${Kd}_{Cl-OH}^{ap-melt}={e}^{(25.81+({X}_{Cl}^{ap}-{X}_{OH}^{ap})\times 17.33)\times \frac{1000}{8.314\times T}}$$
(2)

\({X}_{Cl}^{ap}\) and \({X}_{OH}^{ap}\) represent the mole fractions of Cl and OH in apatite, respectively (Chelle-Michou and Chiaradia 2017).

Parat et al. (2011) developed an empirical non-Henrian partitioning relationship for sulfur between apatite and melt by calibrating a set of natural and experimental apatite/melt partitioning data for andesitic to rhyolitic melts, with the highest melt content limited to 0.1% (ca. 0.3% in apatite), described by Eq. (3):

$${S}_{ap}(wt\%)=0.0629\times \mathrm{Ln}{S}_{melt}(wt\%)+0.4513$$
(3)

Our calculation indicates that the S content in the melts of the Cuojiaoma and Daocheng granites are both 0.001 wt%, while that for the Zhongdian porphyries are significantly higher (0.002-0.065 wt%; Table 4). The Cl concentrations in the melts of the Cuojiaoma and Daocheng granites are 0.01-0.06 wt% (850 ℃) and 0.02-0.05 wt% (850 ℃), respectively. The Zhongdian porphyries exhibit much higher Cl concentrations in the melts (0.03-0.31 wt% (800 ℃); Table 4). Therefore, the S and Cl are significantly rich in the melts of Zhongdian porphyries, which are favorable for the mineralization.

Table 4 Calculation of Cl and S contents in melt by apatite
Full size table

Implications for mineralization

The Late Triassic granitic plutons in the northern Yidun Terrane (e.g. Cuojiaoma and Daocheng batholith) have no related metallic mineralization, while the Late Triassic Zhongdian porphyries are characterized by their close relationship with the porphyry Cu polymetallic deposits (He et al. 2013; Peng et al. 2014; Li et al. 2017; Liu et al. 2017; Wu et al. 2017; Wang et al. 2017b; Dong et al. 2020; Dong et al. 2022a). It is well revealed by the apatite chemical compositions of this study.

Si4+, P5+ and S6+ commonly form a continuous isomorphism in apatite, [SiO4]4- compensates the imbalance produced by the substitution of [PO4]3- by [SO4]2-. When [SO4]2- is insufficient to compensate the imbalance and produce a negative charge surplus, it is inevitable that high-valence elements such as REE3+ will substitute Ca, resulting in Si/S ratio higher than 1 (apfu). Since Si is the major element in the magma, the isomorphism of apatite has limit effect on Si content. Low Si/S ratio suggests relatively high S6+ in magma, reflecting an oxidized environment, which is conducive to the precipitation of ore-forming elements. The Zhongdian apatite display low Si/S ratios (Fig. 11a), while the Si contents in Cuojiaoma and Daocheng apatite are much higher than S, indicating that the Si/S ratio is one favorable indicator for predicting the metallogenic potential of the magmas.

Fig. 11
figure 11

Prediction of mineralization potential by apatite, using S versus Si (a), Y versus Sr (b), Fe versus Mn (c) and Sr versus Mn (d) plots. Same symbols as in Fig. 3

Full size image

Mn, Fe, Sr and Y can substitute Ca in apatite, while Mn and Fe have completely different ordered ubstitution behaviors for Ca, which is an effective "geochemical screen" and can trap Mn in the lattice and strongly restrict Fe from entering the lattice (Liu and Peng 2003). The Zhongdian apatite has higher Fe and Sr, but lower Mn and Y than the Cuojiaoma and Daocheng apatite, it is clear in the plots of Y versus Sr (Fig. 11b), Fe versus Mn (Fig. 11c) and Sr versus Mn (Fig. 11d) to distinguish the differences in between ore-forming (Zhongdian porphyries) and ore-barren rocks (Cuojiaoma and Daocheng granites). Therefore, Mn, Fe, Sr and Y in apatite can be good indicators for predicting the metallogenic potential of the magmas.

Conclusions

In the present study, the apatite chemistry shows that the parental magma of Cuojiaoma and Daocheng granites has non-adakite affinity and was dominantly crust-derived, while that of the Zhongdian porphyries has adakite affinity with high Sr and low Y contents and was derived from a mixed source with both mantle and crustal contributions. The Eu/Eu* and Mn in apatite can reflect the oxidation state of the magma. The apatite from the Zhongdian ore-bearing porphyries has higher Eu/Eu*, S and Cl than that from the Daochang and Cuojiaoma ore-barren granites, indicating that the high O fugacity and S and Cl concentrations in the melts are favorable for Cu-Au mineralization. The S, Si, Sr, Y, Mn and Fe concentrations in apatite can be good indicators for predicting the metallogenic potential of the granitic rocks.