Introduction

Titanite (CaTiSiO5) is a common accessory mineral in igneous and metamorphic rocks, and hydrothermal ore deposits (Frost et al. 2000; Li et al. 2010). Through chemical substitutions, titanite can concentrate significant amounts of REEs, U, Th, Sr, Y, Mn and Pb in the sevenfold Ca site (Higgins and Ribbe 1976; Deer et al. 1982; Tiepolo et al. 2002), and can also accommodate significant amounts of HFSEs, such as Nb, Ta, Zr and W, in the octahedral Ti site (Groat et al. 1985; Bernau and Franz 1987; Lucassen et al. 2011). To a large extent, variations in trace-element compositions of titanite reflect changes of relevant geological conditions, and thus titanite can act as an important indicator of geological processes (Frost et al. 2000; Smith et al. 2009; Ismail et al. 2013). In general, the stability filed of titanite expands in systems with high Ca/Al ratios and relatively oxidizing conditions (~FMQ + 1 < fO2 < ~FMQ + 2) (Frost et al. 2000), and the valence state of Ce in titanite records relatively high-fO2 conditions (King et al. 2013). Titanite with high Fe/Al and Th/U ratios is generally darker in colour (Aleinikoff et al. 2002). Titanite, like zircon, can account for a large proportion of total REE present in the rock (Tiepolo et al. 2002; Vuorinen and Hålenius 2005; Buick et al. 2007; Marks et al. 2008), but titanite of magmatic origin commonly has a higher Th/U ratio than that for metamorphic titanite, which is different from zircon (Gao et al. 2012). Because titanite may incorporate appreciable amounts of U and Th into its structure and has a high closure temperature for Pb diffusion, it can be used as a U-Th-Pb geochronometer (Frost et al. 2000; Buick et al. 2007; Massimo et al. 2013). Zirconium in titanite is sensitive to variations of temperature and pressure and thus can be used as an important thermobarometer (Hayden et al. 2008). Increasing number of studies have indicated that titanite can act as an important host for Sn, W and Mo, and thus prove useful in certifying Sn-, W- and Mo-enriched intrusions, and has a potential application for exploration of Sn, W and Mo deposits (Aleksandrov and Troneva 2007; Xie et al. 2010; Wang et al. 2012; Che et al. 2013).

The Jinshajiang–Red River alkaline igneous belt, extending for over 2000 km adjacent to the NNW–NW-trending Jinshajiang–Red River fault zone, is an important region of Cenozoic Cu–Mo–Au mineralization within the eastern Indo-Asian collision zone (Xu et al. 2012a). This belt contains several hundred alkaline intrusions (Zhang and Xie 1997; Chung et al. 1997, 1998), and dozens of magmatic-hydrothermal Cu–Mo–Au deposits including the Yulong Cu (Mo-Au) and Beiya Au-Pb-Zn deposits, which are associated with alkaline magmatism. While the relationship between individual ore deposits and localized, minor intrusions have been demonstrated (e.g., the Yulong and Machangqing Cu (Mo-Au) deposits, Hou et al. 2006; Xu et al. 2012a), the majority of the alkaline intrusions have not been studied in detail and their mineralization potential is unknown. Hence, a relatively rapid, convenient and economical method for distinguishing between barren and mineralized intrusion would be useful. In recent years, in-situ micro-analytical techniques such as laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS), have been applied widely to major- and trace-element analysis of various rock-forming and accessory minerals (Liu et al. 2008). Geochemical analyses by LA-ICP-MS of accessory minerals, such as zircon, apatite, titanite and rutite, have shown that accessory minerals can be key indicators of mineralization (Ballard et al. 2002; Belousova et al. 2002; Scott 2005; Liang et al. 2006a, b; Xie et al. 2010; Cao et al. 2012; Li et al. 2012; Wang et al. 2012; Che et al. 2013).

Titanite is a common accessory phase in the Jinshajiang–Red River alkaline intrusions. In this paper, we have determined the major- and trace-element compositions of titanites from several representative Cu-mineralized and barren intrusions within the Jinshajiang – Red River alkaline igneous belt. Our study is focused on understanding the mineral chemical characteristics of titanites from mineralized and barren intrusions, and discussing their geological implications for exploration of porphyry Cu deposits in the Jinshajiang – Red River alkaline igneous belt, and elsewhere.

Geological background

The Jinshajiang–Red River alkaline igneous belt extends for over 2000 km and is distributed along the NNW–NW-trending Jinshajiang–Red River deep fault zone along the eastern edge of the Tibetan plateau (Fig. 1; Zhang and Xie 1997; Chung et al. 1997, 1998). The Indo–Asian collision, starting at ~65 Ma, created the Tibetan plateau and resulted in the eastward extrusional tectonics facilitated by strike-slip motion along major faults (e.g., the Jiali, Batang–Lijiang, Gaoligong faults, etc.; Fig. 1), including the Jinshajiang–Red River fault. Strike-slip motion along the Jinshajiang–Red River fault caused lithospheric-scale extension and emplacement of numerous alkaline igneous rocks, including volcanic and intrusive rocks, in the 2000 km long and 50–80 km wide Jinshajiang–Red River alkaline igneous belt (Chung et al. 1997, 1998; Bi et al. 1999, 2009; Yin and Harrison 2000; Ding et al. 2003, 2005; Mo et al. 2003; 2008; Hou et al. 2006, 2007). These alkaline igneous rocks have zircon U–Pb ages between ~43 Ma and ~35 Ma (Xu 2011; Xu et al. 2012a). The rocks range in composition from basaltic to trachytic and rhyolitic (Chung et al. 1998), and are characterized by high alkali contents (K2O + Na2O > 8 wt.%) and potassium enrichment (K2O/Na2O > 1), allowing for their characterization as high-K calc-alkaline or shoshonitic series. Discoveries of abundant alkaline intrusion-associated porphyry Cu(Mo-Au) deposits in the Jinshajiang–Red River alkaline igneous belt make it one of the most important porphyry Cu (Mo-Au) ore belts of Cenozoic age in SW China (Bi et al. 1999, 2002, 2004, 2009; Hu et al. 1998, 2004; Gu et al. 2003; Wang et al. 2004; Hou et al. 2007, 2011; Xu et al. 2012a). Representative porphyry Cu (Mo-Au) deposits including the Yulong, Machangqing, Tongchang and Chang’anchong deposits, are distributed in the northern, central and southern segments of the belt, respectively (Fig. 1). In addition, there are many barren alkaline intrusions in the belt, and representative barren intrusions include the Liuhe, Songgui and Yanshuiqing intrusions in the central part of the belt (Fig. 1).

Fig. 1
figure 1

A simplified geological map of the Jinshajiang–Red River alkaline igneous belt showing Cu-mineralized and barren alkaline intrusions (modified from Wang et al. 2001)

Full size image

Yulong, Machangqing, Tongchang and Chang’anchong porphyry Cu (Mo-Au) deposits

The Yulong deposit is located in the north of the Jinshajiang–Red River alkaline igneous belt (Fig. 1). It is part of the Yulong porphyry copper belt which extends for over 400 km and contains porphyry Cu (Mo-Au) ore deposits with ages of 41–35 Ma, such as Narigongma, Yulong, Malasongduo, Duoxiasongduo, Zhanaga and Mangzong , and more than 20 other Cu (Au)-mineralized porphyries (Hou et al. 2003; Liang et al. 2006a; Xu et al. 2012a). The Yulong deposit is the largest in the Yulong belt, having reserves of ~6.5 Mt Cu and 0.15 Mt Mo with average grades of 0.38 wt.% Cu, 0.04 wt.% Mo, and 0.35 g/t Au (Xu et al. 2012a). Cu (Mo-Au)-mineralization at the Yulong deposit occurred within and around the ~41 Ma monzogranite porphyry stock with an outcrop area of 0.64 km2 (Fig. 2a; Xu et al. 2012a), which intruded into Triassic sandstone and limestone (Li et al. 2012). The Yulong monzogranite porphyry has phenocrysts of K-feldspar, plagioclase, quartz, hornblende and biotite in a phanerocrystalline groundmass (Fig. 3a). Titanite, zircon and apatite are the main accessory minerals.

Fig. 2
figure 2

Simplified geological maps of (a) the Yulong deposit (modified from Hou et al. 2006), (b) Machangqing deposit (modified from Li 2009), (c) Tongchang and Chang’anchong deposits (modified from Xue 2008), and (d) Liuhe, Songgui and Yanshuiqing alkaline intrusions (modified from Zhao et al. 2003)

Full size image
Fig. 3
figure 3

Photomicrographs showing the textural features and mineral compositions of the Yulong monzongranite porphyries (a), Machangqing granite porphyries (b), Tongchang quartz syenite porphyries (c), Chang’anchong quartz syenite porphyries (d), Songgui syenite porphyries (e), Liuhe syenite porphyries (f) and Yanshuiqing syenite porphyries (g)

Full size image

The Machangqing deposit is situated in the central area of the Jinshajiang–Red River alkaline igneous belt (Fig. 1). Cu (Mo-Au)-mineralization at the Machangqing deposit was related to the ~35 Ma granite porphyry stock with an outcrop area of 1.3 km2 (Fig. 2b; Liang et al. 2006b; Xu et al. 2012a), which intruded into lower Ordovician and lower Devonian limestone and sandstone (Xu et al. 2012a). The Machangqing granite porphyry has phenocrysts of K-feldspar, plagioclase, quartz, hornblende and biotite in a phanerocrystalline groundmass (Fig. 3b). Titanite, zircon and apatite are the main accessory minerals. The deposit contains ~0.25 Mt Cu with 0.44 wt.% Cu, 0.03 wt.% Mo and 0.03 g/t Au (Hou et al. 2006; Xu et al. 2012a).

The Tongchang and Chang’anchong deposits are situated in the southern segment of the Jinshajiang–Red River alkaline igneous belt (Fig. 1). Tongchang is located about 5 km east of Chang’anchong. Cu (Mo-Au)-mineralization at both deposits is associated primarily with ~35 Ma quartz syenite porphyry intrusions (Fig. 2c; Huang et al. 2009; Xu et al. 2012a), which were emplaced into middle Silurian limestone and sandstone. The intrusions form stocks and dykes with outcrop areas of ~0.2 km2 and >0.18 km2for the Tongchang and Chang’anchong deposits, respectively (Xu et al. 2012a). Both the Tongchang and Chang’anchong quartz syenite porphyries have phenocrysts of K-feldspar, plagioclase, quartz, hornblende and biotite in a phanerocrystalline groundmass (Fig. 3c and d). Titanite, zircon and apatite are the main accessory minerals. The quartz syenite porphyries forming the two deposits may have originated from a shared magma source, as suggested by the similar mineral assemblages, major- and trace-element, and Sr–Nd isotopic compositions, and zircon U–Pb ages (Xu 2011). Tongchang contains 8,621 t Cu and 17,060 t Mo with 1.24 wt.% Cu, 0.218 wt.% Mo and 0.13 g/t Au, whereas Chang’anchong contains 29,337 t Cu and 13,310 t Mo with 1.48 wt.% Cu, 0.13 wt.% Mo and 0.25 g/t Au (Xue 2008).

Liuhe, Songgui and Yanshuiqing barren alkaline intrusions

Between Dali in the south and Lijiang in the north, the Jinshajiang–Red River alkaline igneous belt contains a group of barren alkaline porphyry intrusions of Cenozoic age. The alkaline porphyry intrusions in this group generally occur as stocks and dykes, which intruded into the Caledonian to Himalayan tectonic layers (Fig. 2d). Tectonically, these porphyry intrusions are located in the central part of the Jinshajiang–Red River alkaline igneous belt (Fig. 1). From south to north, this group comprises the Yanshuiqing, Songgui and Liuhe syenite porphyry intrusions (Fig. 2d) , which have outcrop areas of ~15 km2, ~25 km2 and ~25 km2, respectively (Xu et al. 2014). Both the Yanshuiqing and Songgui syenite porphyries have phenocrysts of major K-feldspar, plagioclase, pyroxene and quartz, and minor hornblende in a phanerocrystalline groundmass (Fig. 3e and f). Zircon U–Pb ages for the Yanshuiqing and Songgui porphyries are ~37 Ma and ~39 Ma, respectively (Xu 2011). The ~37 Ma Liuhe syenite porphyry (Xia et al. 2005; Xu 2011) has phenocrysts of K-feldspar, plagioclase and pyroxene, and minor biotite and hornblende in a phanerocrystalline groundmass (Fig. 3g). Titanite, zircon and apatite are the principal accessory minerals. Various xenoliths have been incorporated into the Liuhe syenite porphyry intrusion, including mainly garnet diopside gneiss and garnet diopside amphibolite, most likely derived from the lower crust (Zhao et al. 2003).

Sampling and analytical methods

Seven titanite-bearing magmatic rock samples, selected from four Cu-mineralized and three barren alkaline porphyry intrusions (Figs. 1 and 2) were investigated in this study. Titanite grains from the samples were prepared as standard petrographic polished thin sections. Optical microscopy and back-scattered electron (BSE) imaging were used to determine the shape and internal structure of the titanite grains prior to LA-ICP-MS analysis. BSE imaging was carried out using EPMA (EPMA-1600, Shimadu, Japan) at the State Key Laboratory of Ore Deposit Geochemistry, Institute of Geochemistry, Chinese Academy of Sciences, Guiyang, China.

Major- and trace-element analyses of titanites were obtained directly from polished thin sections using laser-ablation inductively coupled plasma-mass spectrometer (LA-ICP-MS) housed in the State Key Laboratory of Geological Processes and Mineral Resources, China University of Geosciences (Wuhan), China. Detailed operating conditions for the instrumentation and data reduction are described in Liu et al. (2008). Laser sampling was performed using a GeoLas 2005. The diameter of the laser spot was 44 μm. An Agilent 7500a ICP-MS instrument was used to acquire ionsignal intensities. Helium was used as a carrier gas. Argon was used as the make-up gas and mixed with the carrier gas via a T-connector before entering the ICP. Nitrogen was added into the central gas flow (Ar + He) of the Ar plasma to decrease the detection limit and improve precision (Hu et al. 2008). Each analysis incorporated a background acquisition of approximately 20–30 s (gas blank) followed by 50 s data acquisition from the sample. Agilent Chemstation was utilized for the acquisition of each individual analysis. Off-line selection and integration of background and analysis signals, and time-drift correction and quantitative calibration were performed by ICPMSDataCal (Liu et al. 2008). Element contents (major and trace elements) of titanite were obtained using the method similar to that for anhydrous silicate minerals described by Liu et al. (2008). The USGS reference glasses BCR-2G, BHVO-2G and BIR-1G were used as reference materials for external calibration of element contents; 29Si was used as normalizing element and the sum of all metal oxides was normalized to 100 wt.% (Liu et al. 2008). The technique used here was not applied to titanite previously and is thus untested. Therefore, in order to evaluate the reliability of major-element data by LA-ICP-MS, major element coompositions of several titanite grains analysed by LA-ICP-MS from sample YL907 were measured by EMPA (EPMA-1600, Shimadu, Japan) at the State Key Laboratory of Ore Deposit Geochemistry, Institute of Geochemistry, Chinese Academy of Sciences. The results (see Appendix A and Table 1) indicate that among major elements the largest errors are associated with LA-ICP-MS values for Si (9.3 %) and Al (7.5 %) and Ti (3.5 %). These differences would lead to propagated systematic errors of ~5 % for the reported trace element concentrations. Analyses of BCR-2G, BHVO-2G and BIR-1G reference glasses (see Appendix B) as unknown samples are generally consistent with recommended values (http://georem.mpch-mainz.gwdg.de/) within 10 % for most of trace elements. However, due to significant matrix differences between basaltic glasses and titanite, the errors caused by the matrix differences between calibration standards and the unknowns are difficult to estimate. Given that the differences in elemental concentrations in titanite as determined in our study often exceed an order of magnitude, we are confident that our analytical results reflect real variations in sample compositions.

Table 1 Comparison between EMPA and LA-ICP-MS major element analyses of titanite from sample YL907
Full size table

Results

Petrography of titanite

Titanite grains from the mineralized intrusions are mainly yellow/brown in colour varying in size between 200–2000 μm (Fig. 4). They appear homogeneous in transmitted light, and occur mostly as euhedral to subhedral grains, and partially as crystallographic twins, which are adjacent to, or included in, hornblende, biotite, plagioclase, K-feldspar and quartz (Fig. 4). Some titanite grains from the mineralized intrusions are in contact with chlorite formed by alteration of hornblende or biotite, and the BSE images indicate that some titanite grains from the mineralized intrusions contain very small mineral inclusions, but they demonstrate no evidence of inheritance, zoning or overgrowth textures that would indicate multiple periods of mineral growth or resorption (Figs. 4 and 5; Storey et al. 2007; Li et al. 2010). However, titanite grains from the barren intrusions are primarily pale in colour with relatively smaller sizes of 100–200 μm (Fig. 4). They also appear homogeneous in transmitted light and occur mostly as euhedral to subhedral grains or as crystallographic twins, which are mainly adjacent to, or included in, plagioclase K-feldspar, pyroxene and quartz (Fig. 4). Whereas some titanite grains from the barren intrusions are in contact with chlorite formed by alteration of pyroxene, and the BSE images indicate that some titanite also contain very small mineral inclusions, we found no evidence of inheritance, zoning or overgrowth textures that would indicate multiple periods of mineral growth or resorption (Fig. 5; Storey et al. 2007; Li et al. 2010).

Fig. 4
figure 4

Photomicrographs showing the occurrences of representative titanite grains from the: Yulong monzongranite porphyries (a, b), Machangqing granite porphyries (c, d), Tongchang quartz syenite porphyries (e, f), Chang’anchong quartz syenite porphyries (g, h), Songgui syenite porphyries (i, j), Liuhe syenite porphyries (k, l), and Yanshuiqing syenite porphyries (m, n). Mineral abbreviations: Pl plagioclase; Ttn titanite; Qz quartz; Bt biotite; Ch chlorite

Full size image
Fig. 5
figure 5

Backscattered electron images showing morphological and textural features of representative titanite grains from the: Yulong monzongranite porphyries (a, b), Machangqing granite porphyries (c, d), Tongchang quartz syenite porphyries (e, f), Chang’anchong quartz syenite porphyries (g, h), Songgui syenite porphyries (i, j), Liuhe syenite porphyries (k, l), and Yanshuiqing syenite porphyries (m, n)

Full size image

Chemical compositions of titanite

Mean major- and trace-element compositions of titanite samples are presented in Table 2 and plotted in Figs. 6, 7 and 8. Detailed results of major and trace elements of titanite samples are listed in Appendixes A and B, respectively.

Table 2 Mean major- and trace-element compositions of titanite samples from Cu-mineralized and barren intrusions from the Jinshajing–Red River alkaline igneous belt
Full size table
Fig. 6
figure 6

Plots of CaO vs. Al2O3 (a), Fe2O3 (b), Na2O (c), TiO2 (d), MgO (e) and Fe2O3/Al2O3 (b) for titanite samples from Cu-mineralized and barren intrusions from the Jinshajiang–Red River alkaline igneous belt (the symbols for intrusions in the following figures are same to those in this figure)

Full size image
Fig. 7
figure 7

Chondrite-normalized REY patterns (normalized values for chondrite are from Sun and Mcdoonough 1989) (a), and plots of LREE/HREE vs. ΣREE + Y (b), Eu/Eu* vs. Ce/Ce* (c) and Th/U vs. (Eu/Eu*)/(Ce/Ce*) (d) for titanite samples from Cu-mineralized and barren intrusions from the Jinshajiang–Red River alkaline igneous belt

Full size image
Fig. 8
figure 8

Plots of Th vs. U (a), Nb vs. Ta (b), Zr vs. Hf (c), Zr/Hf vs. Nb/Ta (d), LREE/HREE vs. Sr (e), and Ga (f) for titanite samples from Cu-mineralized and barren intrusions from the Jinshajiang–Red River alkaline igneous belt

Full size image

Major-element compositions of titanite

Titanites from the mineralized intrusions have relatively uniform major element compositions (Table 1), whereas titanites from the barren intrusions exhibit larger variations (Fig. 6a, b and d; Table 2). In addition, titanites from both mineralized and barren intrusions have variable Na2O and MgO contents (Fig. 6c and e; Table 2).

Trace-element compositions of titanite

Titanites from the mineralized intrusions show REE + Y patterns with enrichment in LREEs, depletion in HREEs, notable negative Eu and Y anomalies and weakly positive Ce anomalies on the chondrite-normalized plots (Fig. 7a). They have total REE plus Y contents (ΣREE + Y) ranging from 14,249 to 40,420 ppm (average 28,326 ppm), LREE/HREE ratios from 5.01 to 12.9 (average 8.21), Eu/Eu* values from 0.49 to 0.88 (average 0.67), Ce/Ce* values from 1.14 to 1.38 (average 1.23) and Y/Y* values from 0.71 to 1.02 (average 0.85) (Table 2). They have concentrations of 62.1–248 ppm U (average 122 ppm), 337–985 ppm Th (average 618 ppm), 93.0–500 ppm Ta (average 253 ppm), 989–4097 ppm Nb (average 2082 ppm), 40.5–203 ppm Sr (average 101 ppm) and 29.6–128 ppm Ga (average 77.9 ppm) (Table 2). Th/U ratios are 3.72–7.31 (average 5.18), Nb/Ta ratios are 5.78–11.3 (average 8.52), and Zr/Hf ratios are 10.6–17.5 (average 14.3).

Titanites from the barren intrusions also show REE + Y patterns with enrichment in LREEs, depletion in HREEs, notable negative Eu and Y anomalies and weakly positive Ce anomalies on the chondrite-normalized plots (Fig. 7a), but they have lower ΣREE + Y of 2322–10,954 ppm (average 5586 ppm), lower LREE/HREE ratios of 1.77–7.04 (average 4.49), and higher Eu/Eu* values of 0.73–0.99 (average 0.86), whereas Ce/Ce* values (1.11–1.28; average 1.19) and Y/Y* values (0.62 to 1.00; average 0.79) are similar (Table 2). Concentrations of other trace elements are also lower: 6.48–109 ppm U (average 41.1 ppm), 6.90–270 ppm Th (average 119 ppm), 21.3–139 ppm Ta (average 65.7 ppm), 171–1898 ppm Nb (average 1018 ppm), 66.3–620 ppm Sr (average 255 ppm) and 6.31–32.6 ppm Ga (average 15.3 ppm) (Table 2). Although Th/U ratios are similar between the two intrusion types (3.72–7.31; average 5.18 in the barren intrusions), Nb/Ta ratios and Zr/Hf ratios are different (8.03–29.9; average 15.5 and 10.4–38.4; average 19.7, respectively).

Discussion

Compositional comparisons of titanites between Cu-mineralized and barren intrusions

Major-element compositional results indicate that titanites from the mineralized intrusions have higher Fe2O3/Al2O3 ratios, and lower Al2O3 and CaO contents than those from the barren intrusions (Fig. 6a and f). The trace-element compositions of titanites from the mineralized and barren intrusions display large variations (Table 2 and Appendix B), demonstrating that titanite is capable of accommodating a wide range of trace elements (Higgins and Ribbe 1976; Deer et al. 1982; Cérny’ et al. 1995; Frost et al. 2000). Furthermore, titanites from mineralized and barren intrusions can be distinguished on basis of trace-element concentrations and ratios, (such as ΣREE + Y, LREE/HREE, Eu/Eu*, U, Th, Ta, Nb, Ga, Sr, Nb/Ta and Zr/Hf, etc.; Figs. 7 and 8).

Titanites from both the mineralized and barren intrusions show REE + Y patterns with enrichment in LREEs and depletion in HREEs, notable negative Eu and Y anomalies and weakly positive Ce anomalies on the chondrite-normalized plots, but titanites from the mineralized intrusions have much higher ΣREE + Y contents, higher LREE/HREE ratios, obviously larger negative Eu anomalies and slightly larger positive Ce anomalies than those for titanites from the barren intrusions (Figs. 7a, b and c). In addition, the Eu/Eu*–Ce/Ce* diagram also highlights a negative correlation between Eu/Eu* and Ce/Ce*(Fig. 7c), and titanites from the mineralized intrusions show higher (Ce/Ce*)/(Eu/Eu*) ratios than those from the barren intrusions (Fig. 7d).

Uranium and Th concentrations in titanites from the mineralized intrusions are notably higher than those for titanites from the barren intrusions, and the two types of titanites can be distinguished from each other on the Th–U diagram (Fig. 8a). Tantalum and Nb in titanites from the mineralized intrusions are extremely enriched, demonstrating that titanite is an important host for Ta and Nb, which are much higher than those for titanites from the barren intrusions. Furthermore, titanites from the mineralized and barren intrusions have different enrichment trends for Nb–Ta and Zr–Hf (Fig. 8b and c), consequently producing both lower Zr/Hf and Nb/Ta ratios for titanites from the mineralized intrusions, which can be easily distinguished from titanites from the barren intrusions (Fig. 8d). In addition, titanites from the mineralized and barren intrusions show different enrichment trends for Sr and Ga (Fig. 8e and f). Strontium in titanites from barren intrusions is much higher than in titanites from the mineralized intrusions (Fig. 8e), whereas Ga in titanites from mineralized intrusions is significantly higher than in titanites from the barren intrusions (Fig. 8f).

Possible factors causing compositional differences of titanites between Cu-mineralized and barren intrusions

TiO2 contents of titanites from both the mineralized and barren intrusions have negative correlations with Al2O3 + Fe2O3 (Fig. 9a), indicating the substitution (Fe, Al)3+ + (OH, F)− ⟺ Ti4+ + O2− (Frost et al. 2000, and references therein). In addition, CaO contents of titanites from both the mineralized and barren intrusions show a negative correlation with Sr2+ (Fig. 9e), and the CaO + TiO2 contents negatively correlate with ΣREY3++Al3++Fe3+, Na++Nb5++Ta5+ and U4++Th4++Mg2+ (Fig. 9b, c and d), indicating the likely substitution mechanisms: Ca2+ + Ti4+ = REE3+ + Y3+ + (Al, Fe)3+, Ca2+ + Ti4+ = Na+ + (Nb, Ta)5+, Ca2++Ti4+ = (U, Th)4+ + Mg2+ and Ca2+ = Sr2+. Additional possible substitutions include: (1) (Zr, Hf) 4+ = Ti4+; (2) (Al, Fe)3+ + (Nb, Ta)5+ = 2Ti4+ (Higgins and Ribbe 1976; Groat et al. 1985; Frost et al. 2000; Tiepolo et al. 2002; Liferovich and Mitchell 2005; Prowatke and Klemme 2005; Aleksandrov and Troneva 2007; Xie et al. 2010; Anand and Balakrishnan 2011; Olin and Wolff 2012).

Fig. 9
figure 9

Plots of TiO2 vs. Al2O3 + Fe2O3 (a), and CaO + TiO2 vs. ΣREE3++Y3++Al3++Fe3+ (b), Na++Nb5++Ta5+ (c) and U4++Th4++Mg2+ (d), and CaO vs. Sr2+ (e), and TiO2 vs. Ga (f) for titanite samples from Cu-mineralized and barren intrusions from the Jinshajiang–Red River alkaline igneous belt

Full size image

The darker colour of titanite is suggested to be related to higher Fe/Al ratios (Frost et al. 2000; Aleinikoff et al. 2002). Darker colour titanites from the mineralized intrusions commonly exhibit higher Fe2O3/Al2O3 ratios than the lighter colour titanite from the barren intrusions, consistent with this suggestion.

Trace-element incorporation in titanite depends upon the crystal-chemical behavior of titanite, coexisting melt compositions and mineral phases, pressure, temperature and fO2 (Watson 1976; Adam and Green 1994; Tiepolo et al. 2002; Prowatke and Klemme 2005, 2006; Jung and Hellebrand 2007; Smith et al. 2009; Anand and Balakrishnan 2011; Olin and Wolff 2012). Concerning the crystal-chemical behavior of titanite, the availability of suitable substitution mechanisms to achieve electrovalent balance and the similarity in size of major- or trace-elements are the main factors (Tiepolo et al. 2002). The coexisting melt compositions and mineral phases, which control the availability of trace-element compositions in melts during titanite growth, thus can exert significant influences on the concentrations and patterns of trace elements in titanite (Watson 1976; Tiepolo et al. 2002; Prowatke and Klemme 2005; Smith et al. 2009; Anand and Balakrishnan 2011; Olin and Wolff 2012). Previous works have indicated that temperature, instead of pressure, has a notably inverse effect on the trace-element concentrations in titanite (Tiepolo et al. 2002; Anand and Balakrishnan 2011). The fO2 exerts influence mainly on trace elements with different valances such as Ce and Eu (King et al. 2013). In this study, the crystallization temperature of magmas for the mineralized and barren intrusions is likely to be similar, ~750 °C (Xu 2011). As mentioned above, REE + Y, U, Th, Nb, Ta, Zr, Hf and Sr in titanites from the both mineralized and barren intrusions were incorporated through suitable substitution mechanisms. Therefore, these trace-element concentrations in titanites were mainly controlled by availability of trace-element compositions in melts during titanite growth. Plots of ΣREE + Y, LREE/HREE, U, Th, Ta and Nb of titanites vs. those of whole-rock samples produce general positive correlations (Fig. 10a, b, c, d, e and f). We suggest that the coexisting melt compositions, rather than the coexisting mineral phases, exert a more important influence on these trace-element compositions of titanites from both intrusion types.

Fig. 10
figure 10

Plots of ΣREE + Y, LREE/HREE, U, Th, Ta, Nb, Ga and Sr values of titanites vs. those of whole-rock samples (a-h) for Cu-mineralized and barren intrusions from the Jinshajiang–Red River alkaline igneous belt (trace-element concentrations of whole-rock samples from Xu et al. 2014)

Full size image

Our study has identified very high Ga contents in titanites from the mineralized intrusions. Gallium concentrations in whole-rock samples for both the mineralized and barren intrusions are similar and are about 18–20 ppm (Xu 2011; Xu et al. 2014), but Ga concentrations in titanites from the mineralized and barren intrusions are notably different (Fig. 10g), suggesting that Ga concentrations in titanites are not controlled by the coexisting melts. In igneous rocks, Ga is strongly concentrated in Al- and Fe-enriched minerals such as plagioclase, K-feldspar, nepheline, spinel and magnetite (Paktunc and Cabri 1995; Hieronymus et al. 2001; Tu et al. 2003; Luo et al. 2007; Macdonald et al. 2010). The reason is that under oxidizing conditions Ga3+ has geochemical characteristics similar to Al3+ and Fe3+, and thus commonly substitutes for Al3+ and Fe3+ (Tu et al. 2003; Luo et al. 2007; Macdonald et al. 2010; Breiter et al. 2013). As with Al3+ and Fe3+, Ga would most likely substitute for Ti in titanite, as deduced from a negative correlation between TiO2 and Ga (Fig. 9f). Furthermore, effective ionic radius of the hexahedral Ga3+ is closer to Ti in titanite than for hexahedral Al3+ and Fe3+ (Shannon 1976), implying that Ga3+ enters titanite more readily than Al3+ and Fe3+. Therefore, magmatic fO2 should be a significant control on Ga uptake by titanite, and the pronounced high Ga concentrations in titanites from the mineralized intrusions suggest oxidized conditions, consistent with conclusion made on the basis of other methods (Liang et al. 2006a, b; Bi et al. 2009; Xu 2011; Xu et al. 2012b, 2014).

In contrast to the barren intrusions, mineralized intrusions have higher Sr concentrations, but contain titanites with lower Sr concentrations (Fig. 10h), indicating that Sr concentrations in titanites do not reflect Sr concentrations in the parental melts. They are probably controlled by the nature of other crystalizing mineral phases. As Sr is compatible in plagiolcse, small proportions of plagioclase crystallization can strongly decrease Sr concentrations in melts (Icenhower and London 1996; White et al. 2003; White 2003; Ren 2004), and consequently, less Sr can be incorporated into titanites. Therefore, lower Sr concentrations in titanites from the mineralized intrusions may be related to a higher proportion of plagioclase crystallization.

Both Eu and Ce have two oxidation states (Eu3+–Eu2+ and Ce3+–Ce4+), and variations in fO2 can change the oxidation state from Eu3+ to Eu2+ and Ce3+ to Ce4+. Like other REEs, trivalant Eu and Ce are most favored in the heptahedral Ca site in titanite (Frost et al. 2000; Tiepolo et al. 2002; King et al. 2013). Under reducing conditions, Eu as Eu2+ is not favored by titanite, thus easily producing negative Eu anomalies on the chondrite-nomalized plots (Colombini et al. 2011; Ismail et al. 2013). Therefore larger negative Eu anomalies of titanites from mineralized intrusions should imply more reduced conditions. However, previous studies (Liang et al. 2006a, b; Xu 2011; Xu et al. 2012b, 2014) have indicated that mineralized intrusions have higher magmatic fO2 than barren intrusions. This contradiction can be explained by crystallization of feldspar, because Eu is enriched in plagioclase and minor plagioclase crystallization may produce Eu-depletion in melts (Ballard et al. 2002; Bi et al. 2002; Buick et al. 2010; Anand and Balakrishnan 2011; Ismail et al. 2013). Hence, larger negative Eu anomalies of titanites from the mineralized intrusions may be caused by a larger proportion of plagioclase crystallization, a feature that is consistent with the conclusion made from the lower Sr concentrations in titanites from the mineralized intrusions. Under oxidizing conditions, Ce4+ is favored by titanite’s hexahedral Ti site (King et al. 2013). Therefore, high fO2 facilitates incorporation of Ce4+ into titanite, readily leading to positive Ce anomalies on the chondrite-nomalized plots. Different degrees of positive Ce anomalies of titanites from the mineralized and barren intrusions are consistent with different magmatic fO2 conditions for the mineralized and barren intrusions.

Implications for exploration for porphyry Cu deposits

This study has shown that titanites from the Cu-mineralized intrusions have higher values of Fe2O3/Al2O3, ΣREE + Y, LREE/HREE, Ce/Ce*, (Ce/Ce*)/(Eu/Eu*), U, Th, Ta, Nb and Ga, and lower Al2O3, CaO, Eu/Eu*, Sr, Zr/Hf and Nb/Ta values than those for titanites from the barren intrusions. High Ga concentrations and large positive Ce anomalies of titanites from the Cu-mineralized intrusions suggest oxidized conditions, and confirm that an oxidized magma has potential to produce porphyry Cu-mineralization (Ballard et al. 2002; Mungall 2002). Key compositional differences of titanites between the Cu-mineralized and barren intrusions may provide a useful tool to discriminate between ore-bearing and barren intrusions, at an early stage of exploration, and have potential applications in exploration for porphyry Cu deposits in the Jinshajiang – Red River alkaline igneous belt, and other areas.

Conclusions

Titanites from the Cu-mineralized and barren intrusions have different Fe2O3/Al2O3, Al2O3, CaO, ΣREE + Y, LREE/HREE, Ce/Ce*, Eu/Eu*, (Ce/Ce*)/(Eu/Eu*), U, Th, Ta, Nb, Zr/Hf, Nb/Ta, Ga and Sr values. These differences suggest that melts parental to Cu-mineralized and barren intrusions had different chemical compositions and different magmatic fO2 conditions, and underwent different degrees of crystallization of plagioclase. The Cu-mineralized intrusions are more oxidized. The pronounced compositional differences of titanites between the Cu-mineralized and barren intrusions indicate that titanite has a potential as an indicator for porphyry Cu deposits.