Introduction

Mantle-derived mafic magma, emplaced at different crustal levels, can play an important role for the generation and evolution of granites (Huppert and Sparks 1988; Bergantz 1989; Annen and Sparks 2002; Kemp et al. 2007; Koteas et al. 2010). Mafic magma can act as parental magma, heat source and end-member of mixing or assimilation (Tepper et al. 1993; van de Flierdt et al. 2003). Annen et al. (2006, 2008) proposed a model of intermediate and silicic melt generation in Deep Hot Zone that was formed by repeatedly underplated or interplated mafic sills. The melt can be generated by incomplete crystallization of the mafic magma (residual melt), by partial melting of crustal rocks and by mixing between residual melt (or mafic magma) and crustal melt. The role of mafic magma for the origin of granitoids can be achieved in areas where granitoids are associated with gabbros, diorites, mafic dykes and mafic magmatic enclaves (Altherr et al. 1999; Bonin 2004; Tepper and Kuehner 2004; Barbarin 2005). Mantle-derived magma closely associated with crustal magma can be used to better constrain the lithosphere evolution and the geodynamic processes (Bonin 2004).

The Qinling orogen in central China separates the North China block from the South China block and links the Kunlun and Qilian orogens to the west and the Dabie–Sulu orogen to the east (Fig. 1a), which is an important tectonic domain in eastern Asia (Meng and Zhang 2000; Ratschbacher et al. 2003). The Qinling orogen can be divided into the East and West Qinling (Zhang et al. 2001; Feng et al. 2002). Previous studies focused mainly on the East Qinling and the neighboring northwestern margin of the Yangtze block and eastern part of the West Qinling. Geochronological and geochemical data show that most granitoids in the East Qinling and its neighboring areas have magma crystallization ages ranging from 227 to 205 Ma (Late Indosinian) and that they formed in a post-collisional setting (Sun et al. 2002; Zhang et al. 2007b, 2008; Qin et al. 2008, 2009, 2010a, b and references therein; Cao et al. 2011). However, recent study shows that some granitoids from the middle and western parts of the West Qinling have magma crystallization ages of ~245–235 Ma (Feng et al. 2002; Jin et al. 2005; Zhang et al. 2006; Wang et al. 2010), which are obviously older than those of the East Qinling granitoids. There are two competing tectonic models for the generation of the West Qinling granitoids: (1) active continental margin (Jin et al. 2005; Meng et al. 2005) and (2) early stage of post-collision setting (Zhang et al. 2006, 2008). Therefore, additional studies are required.

Fig. 1
figure 1

a Simplified geological map of China, showing major tectonic units of China (after Zheng et al. 2010); b Simplified map showing distribution of the Early Mesozoic granitoids in the Qinling orogen (after Feng et al. 2002); and c Geological map of the Shuangpengxi area, West Qinling. Abbreviations in a and b: WQ West Qinling, EQ East Qinling, SPGZ Songpan-Ganzi terrane, DB Dabie belt, SL Sulu belt, QD Qaidam, QL Qilian Shan belt, KL (EKL) Kunlun Shan belt, NQL North Qinling, SQL South Qinling. Pluton names and zircon U–Pb age data sources in b: East Qinling (Sun et al. 2002; Qin et al. 2009, 2010b; Zhang et al. 2008 and references therein; Jiang et al. 2010); Northwestern margin of South China block (Li et al. 2007; Zhang et al. 2007b; Qin et al. 2008); Eastern part of the West Qinling (Sun et al. 2002; Qin et al. 2009; Cao et al. 2011): MB Miba, MSL Mishuling, WQ Wenquan, Middle part of the West Qinling (Feng et al. 2002; Jin et al. 2005; Wang et al. 2010): YLG Yeliguan, XH Xiahe, LWX Longwuxia, TR Tongren, Western part of the West Qinling (Zhang et al. 2006): HMH Heimahe, WQ Wenquan

Full size image

In this paper, we carry out an integrated study of U–Pb zircon dating, geochemical and Sr–Nd–Hf isotopic compositions for the Xiekeng (XK) diorite–granodiorite pluton and Shuangpengxi (SPX) granodiorite pluton in the middle part of the West Qinling orogen. The XK and SPX plutons provide an excellent opportunity to study the role of mafic magmas on the origin of granitoids and to examine the model of Deep Hot Zones (Annen et al. 2006, 2008). We also use these data to discuss their tectonic implications.

Geological backgrounds

The Qinling orogenic belt is a multistage orogenic belt and extends for more than 1,500 km in central China (Fig. 1a) (Meng and Zhang 2000; Ratschbacher et al. 2003). The East Qinling orogen is composed of three blocks and two sutures (Meng and Zhang 1999). The Shangdan suture zone separates the North China block (including the North Qinling) from the Qinling microplate (the South Qinling) (Fig. 1b). The North Qinling is predominantly composed of Proterozoic to Paleozoic medium-grade meta-sedimentary and meta-volcanic (Li et al. 1996; Zhang et al. 2001). The South Qinling consists mainly of Late Proterozoic to Triassic sediments overlying Neo-Proterozoic crystalline basement. The Triassic collision of the South China block with the South Qinling along the Mianlue suture led to the widespread fold-thrust deformation and granitic magmatism throughout the East Qinling (Zhang et al. 2001; Sun et al. 2002).

The West Qinling is bounded by the East Kunlun and Qaidam terranes along the Wenquan-Wahongshang fault to the west, separated from the Qilian orogen by the Qinghai Lake-Baoji fault to the north, and separated from the Bayan Har/Songpan-Garze block by the A’nimaque-Mianlue suture zone to the south (Fig. 1b). The A’nimaque-Mianlue suture zone along the southern margin of the Qinling orogenic belt contains abundant ophiolite fragments that record collision between the North and South China blocks, which is considered to be a Late Paleozoic Paleo-Tethys oceanic subduction zone dipping to the north (Li et al. 1996; Xu et al. 2002; Konstantinovskaia et al. 2003; Lai et al. 2008; Yang et al. 2009). In the West Qinling, sedimentary cover is mainly the Devonian to Cretaceous sediments and Precambrian basement is rarely exposed (Feng et al. 2002).

Granitoids are widespread in the West Qinling. Most of them have sharp contact with their wall-rock of the Phanerozoic sedimentary cover. The Mishuling, Miba, Wenquan and Wuduojinghua granitoid plutons in the eastern part of the West Qinling have U–Pb zircon ages of ~225–211 Ma (Fig. 1b) (Sun et al. 2002; Qin et al. 2009; Cao et al. 2011). The granitoids in the middle part have U–Pb zircon ages ranging from 245 to 238 Ma (Fig. 1b) (Feng et al. 2002; Jin et al. 2005; Wang et al. 2010). The granitoids in the western part of the West Qinling are located along the Wenquan-Wahongshan fault and the southern margin of the Qinghai Lake fault (Fig. 1b). They have U–Pb zircon ages of 235–218 Ma (Zhang et al. 2006).

In this paper, samples from the XK and SPX plutons in the middle part of the West Qinling were collected. The XK and SPX plutons intrude Permian and Early Triassic sediments (Fig. 1c). The XK pluton, with an area of ~14 km2, comprises mainly diorite with minor granodioritic porphyry. In the field, no sharp contact between these rocks is observed, indicating that they could be coeval. The diorite is dark gray or gray green, medium grained, massive and consists of 3–5 % quartz, 60–75 % plagioclase, 5–15 % orthopyroxene, 5–8 % clinopyroxene, 3–5 % hornblende and biotite with minor Fe–Ti oxide, apatite and titanite. The hornblende and biotite mostly occur as thin rims on pyroxene (Fig. 2a, b). The granodioritic porphyry has a groundmass of fine-grained quartz and feldspar with minor biotite, and phenocrysts of plagioclase, biotite and hornblende (Fig. 2c). Accessory minerals include zircon, Fe–Ti oxide and apatite. The SPX granodiorite intrudes the XK pluton and occurs as a stock with an area of ~14 km2 (Fig. 1c). This granodiorite is medium-coarse grained and massive and consists of 15–20 % quartz, 40–45 % plagioclase, 10–15 % K-feldspar, 5–8 % hornblende and 5–8 % biotite (Fig. 2d). Accessory minerals include apatite, magnetite and zircon.

Fig. 2
figure 2

Photomicrographs of representative samples of the intrusive rocks from West Qinling. XK pluton: a high-Al diorite (0942); b high-Mg diorite (0946); c granodioritic porphyry (XK-8); SPX pluton: d granodiorite (SPX-6). Amp amphibole, Bt biotite, Opx orthopyroxene, Cpx clinopyroxene, Pl plagioclase, Kfs K-feldspar, Qz quartz

Full size image

Analytical methods

Fresh rock samples were crushed in a steel crusher and then powdered in an agate mill to a grain size <200 mesh. Major elements were analyzed at the Hubei Geological Analytical Center, Wuhan. For analytical methods see Zhang et al. (2002). The analytical uncertainty is generally <5 %. Trace elements, including REE, were measured using Agilent 7500a ICP-MS at the State Key Laboratory of Geological Processes and Mineral Resources (GPMR), China University of Geosciences, Wuhan. Sample-digesting procedure for ICP-MS analyses and analytical precision and accuracy are the same as described by Liu et al. (2008).

Whole-rock Sr and Nd isotopic ratios were measured by a Triton thermal ionization mass spectrometer at GPMR. 87Rb/86Sr and 147Sm/144Nd ratios were calculated from Rb, Sr, Sm and Nd contents measured by ICP-MS. The measured Sr and Nd isotopic ratios were normalized to 86Sr/88Sr = 0.1194 and 146Nd/144Nd = 0.7219, respectively. During the period of analysis, NBS987 standard yielded an average 87Sr/86Sr value of 0.710239 ± 10 (2σ), and BCR-2 standard gave an average 143Nd/144Nd value of 0.512620 ± 2 (2σ). For details of the Sr and Nd isotopic analytical procedures see Gao et al. (2004).

Zircons were separated using conventional techniques. Zircons, selected by examination with a binocular microscope, were mounted in epoxy resin and polished to approximately half thickness. Cathodoluminescence (CL) images, taken at Northwest University, Xi’an, were used to check the internal textures of individual zircon and to guide U–Pb dating and Hf isotope analysis. U–Pb zircon dating was carried out using LA-ICP-MS on an Agilent 7500 equipped with a 193-nm ArF excimer laser at GPMR. Operating conditions are the same as described by Liu et al. (2010). A beam diameter of 32 μm was used for sample SPX-6 and XK-8 and 24 μm for sample 0951. Zircon 91500 and the GSE-1G glass were used as an external standard for Pb/U ratio and concentration, respectively. Common Pb correction is made by using the program of ComPbCorr#3-17 (Andersen 2002). Off-line selection and integration of background and analyte signals, and time-drift correction and quantitative calibration were conducted by ICPMSDataCal (Liu et al. 2010). The data were processed using the ISOPLOT program (Ludwig 2003).

Zircon Hf isotope analysis was carried out in situ using a Neptune MC-ICP-MS at GPMR. Analytical spots were located close to or on the top of LA-ICP-MS spots or in the same growth domain as inferred from CL images. The instrumental conditions and data acquisition were described by Hu et al. (2012). The analyses were conducted with a beam diameter of 44 μm, a hit rate of 8 Hz and laser energy of 60 mJ. A major problem to the accurate determination of the Hf isotopic composition of zircon using LA-MC-ICP-MS is to make the isobaric interference corrections of 176Yb and 176Lu on 176Hf (Woodhead et al. 2004; Yuan et al. 2008; Kemp et al. 2009; Fisher et al. 2011). Recent studies have suggested that the Yb isotope abundances for the mass bias and isobaric interference correction may be instrument and/or technique specific (Kemp et al. 2009; Fisher et al. 2011). During this study, the mass fractionations of Hf and Yb were calculated using an exponential law and values of 0.7325 for 179Hf/177Hf and 1.1265 for 173Yb/171Yb, respectively (IUPAC 1991). The mass bias behavior of Lu is typically assumed to be similar to that of Yb, and βYb was determined using the 173Yb/171Yb measured during each zircon analysis (Woodhead et al. 2004). Isobaric interference of 176Lu and 176Yb on 176Hf was monitoring by measuring the intensity of the interference-free 175Lu and 173Yb isotope, respectively. Ratios used for the corrections were 176Lu/175Lu = 0.02659 and 176Yb/173Yb = 0.79108, respectively (IUPAC 1991). Off-line selection and integration analyte signals, and isobaric interference and mass fractionation correction of Lu–Hf isotopic ratios were also performed by ICPMSDataCal (Liu et al. 2010).

For most natural zircon samples that have relatively high 176Yb/177Hf ratio, the interference correction of 176Yb on 176Hf must be able to accommodate a wide range of 176Yb/177Hf ratios (Valley et al. 2010, Fisher et al. 2011). To monitor the accuracy of this correction, every 5–7 sample analyses were followed by analysis of the reference zircons. Reference zircons 91500 (176Yb/177Hf = 0.0069–0.0077), GJ-1 (176Yb/177Hf = 0.0057–0.0062), Mud Tank (176Yb/177Hf = 0.0009–0.0010) and Temora (176Yb/177Hf = 0.008–0.042) were analyzed as unknown (Table S1). During the analysis, the standard zircons gave 176Hf/177Hf ratios of 0.282309 ± 6 (2σ, MSWD = 0.68, n = 11) for 91500 and 0.282024 ± 8 (2σ, MSWD = 1.7, n = 10) for GJ-1 and 0.282500 ± 5 (2σ, MSWD = 0.72, n = 7) for Mud Tank and 0.282675 ± 7 (2σ, MSWD = 1.2, n = 11) for Temora, respectively. They are all well consistent with the recommended 176Hf/177Hf ratios of 0.282308 ± 6 (Blichert-Toft 2008) for 91500, 0.282015 ± 19 for GJ-1 (Elhlou et al. 2006), 0.282504 ± 26 for Mud Tank (Woodhead and Hergt 2005) and 0.282680 ± 24 for Temora (Woodhead et al. 2004) within analytical errors (Fig. S1), respectively. Most of the samples analyzed in this study have less 176Yb/177Hf ratio than our standards and do not display any obviously correlation between the 176Hf/177Hf and 176Yb/177Hf ratios (Fig. S1), indicating that our interference corrections are reasonable and the zircon Hf isotope data are accurate.

The decay constant for 176Lu of 1.865 × 10−11 year−1 was adopted (Scherer et al. 2001). Initial 176Hf/177Hf ratio, denoted as εHf(t), is calculated relative to the chondritic reservoir with a 176Hf/177Hf ratio of 0.282772 and 176Lu/177Hf of 0.0332 (Blichert-Toft et al. 1997). Single-stage Hf model ages (T DM1) are calculated relative to the depleted mantle, which is assumed to have a linear isotopic growth from 176Hf/177Hf = 0.279718 at 4.55 Ga to 0.283250 at present, with 176Lu/177Hf ratio of 0.0384 (Vervoort and Blichert-Toft 1999), and two-stage Hf model ages (T DM2) are calculated by assuming a mean 176Lu/177Hf value of 0.015 for the average continental crust (Griffin et al. 2002). The single-stage Hf model age (T DM1) is taken for positive εHf(t) values, and two-stage Hf model age (T DM2) is taken for negative εHf(t) values (Zheng et al. 2006).

Results

U–Pb zircon ages

LA-ICP-MS zircon U–Pb data are listed in the Table S2, and representative zircon CL images and their U–Pb concordia plots are shown in Figs. 3 and 4, respectively.

Fig. 3
figure 3

Representative the CL images of zircon samples. a SPX granodiorite (SPX-6); b XK granodioritic porphyry (0951); c XK granodioritic porphyry (XK-8). The smaller solid line circles show LA-ICPMS dating spots and corresponding U–Pb ages (in Ma), and the bigger broken line circles show Lu–Hf isotope analysis and corresponding εHf(t) values

Full size image
Fig. 4
figure 4

Zircon U–Pb Concordia diagrams. a SPX granodiorite (SPX-6); b XK granodioritic porphyry (0951); c XK granodioritic porphyry (XK-8). Shaded and dashed ellipses are not included in the age calculation. Ellipses represent 1-sigma uncertainty for individual analyses

Full size image

Zircons from the SPX granodiorite (sample SPX-6) are colorless, transparent and euhedral. Most of them are short to long prismatic, with length to width ratios ranging from 1:1 to 3.5:1. In CL images, these zircons have planar zoning or oscillatory zoning, consistent with a magmatic origin (Fig. 3a) (Corfu et al. 2003). Some zircons display inherited zircon cores with weak zoning or no zoning (Fig. 3a). Four analyses on the inherited cores give 206Pb/238U ages of 357 ± 5, 264 ± 4, 263 ± 4 and 261 ± 3 Ma, respectively (Fig. 4a). One analysis has a young age of 222 ± 4 Ma, probably due to the Pb loss (Fig. 4a). These analyses have 120–321 ppm U, 91.5–236 ppm Th and Th/U ratios of 0.57–0.78. Eighteen analyses on magmatic domains show 64.5–366 ppm U, 64–258 ppm Th and Th/U ratios of 0.67–1.51 (Table S2). They yield 206Pb/238U ages between 231 ± 4 and 251 ± 5 Ma, with a weighted mean of 242 ± 3 Ma (2σ; MSWD = 2.6) (Fig. 4a), which is interpreted to be the magma crystallization age of the SPX granodiorite.

Two samples from the XK granodioritic porphyry (sample 0951 and XK-8) were chosen for U–Pb zircon dating. Zircons from sample 0951 are colorless, transparent, short prismatic (up to 200 μm long) and euhedral or subhedral. CL images show no or weak zoning (Fig. 3b). Fourteen analyses show 162–786 ppm U, 93.5–641 ppm Th and Th/U ratios of 0.41–0.95 (Table S2). Among them, three analyses give 206Pb/238U ages of 270 ± 4, 264 ± 4 and 258 ± 4 Ma, respectively (Fig. 4b), and the remaining eleven analyses yielded 206Pb/238U ages ranging from 240 to 248 Ma, with a weighted mean of 244 ± 2 Ma (2σ; MSWD = 0.46) (Fig. 4b), representing its magma crystallization age.

Zircons from sample XK-8 display short to long prismatic, transparent and colorless. They have lengths of 100–200 μm with length/width ratios from 1:1 to 2:1, weak or oscillatory zoning, and a few have core-rim structure (Fig. 3c). All analyses show U of 207–1,230 ppm, Th of 128–1,499 ppm and Th/U ratios of 0.19–1.22 (Table S2). Five analyses on the inherited cores give 206Pb/238U ages of 324 ± 4, 264 ± 3, 261 ± 4, 260 ± 4 and 256 ± 3 Ma, respectively (Fig. 4c). The remaining 20 analyses show 206Pb/238U ages ranging from 234 to 249 Ma, with a weighted mean of 242 ± 2 Ma (2σ; MSWD = 2.9) (Fig. 4c), which is consistent with the age obtained from sample 0951 within error, representing the magma crystallization age of the XK pluton.

Major and trace elements

Major and trace element data for the SPX and XK plutons are given in Table 1. Four samples for the XK diorite from Yang (2008) are also included.

Table 1 Major (wt. %) and trace element (ppm) data for the SPX and XK Plutons
Full size table

The SPX granodiorite (Table 1 and Fig. 5) is metaluminous and high-K calc-alkaline (Fig. 6a, b). In the trace element spider diagram, all the samples show negative Nb, Ta, Sr, P and Ti anomalies and are relatively enriched in Rb, Ba, Th, U and K (Fig. 7a). Their REE data display strongly fractionated REE patterns ((La/Yb) N  = 10.1–13.7) and moderately negative Eu anomalies (Eu/Eu* = 0.58–0.61) (Fig. 7b).

Fig. 5
figure 5

Harker plots of selected major and trace elements of the SPX and XK plutons. a SiO2 versus Al2O3; b SiO2 versus MgO; c SiO2 versus Mg#; d SiO2 versus CaO; e SiO2 versus Na2O; f SiO2 versus K2O/Na2O; g SiO2 versus Cr; i SiO2 versus Ni; j SiO2 versus Sr; k SiO2 versus Ba

Full size image
Fig. 6
figure 6

Plots of a A/NK [molar ratio Al2O3/(Na2O + K2O)] versus A/CNK molar ratio [Al2O3/(CaO + Na2O + K2O)] and b SiO2 versus K2O (Peccerillo and Taylor 1976). Symbols as in Fig. 5

Full size image
Fig. 7
figure 7

Primitive-mantle (PM)-normalized trace element spider diagrams and chondrite-normalized REE patterns. Chondrite and PM values are from Sun and McDonough (1989). Symbols as in Fig. 5

Full size image

The XK diorites show SiO2 variation ranging from 52.43 to 56.82 %. Based on the contents of MgO and Al2O3 (Fig. 5a–c), they can be divided into two distinct groups: high-Al diorite and high-Mg diorite. The high-Al diorite has Al2O3 = 17.72–21.11 %, and MgO = 2.50–3.62 % with Mg# of 43–54. The high-Mg diorite is characterized by Al2O3 = 15.32–16.58 %, and MgO = 5.62–7.62 % with Mg# of 59–64. The contents of CaO, Na2O, TiO2 and P2O5 of the high-Al diorite are relatively higher than those of the high-Mg diorite (Table 1 and Fig. 5). The high-Al diorite straddles the transition between medium-K and high-K calc-alkaline, and the high-Mg diorite falls in the medium-K calc-alkaline field (Fig. 6b). The high-Mg diorite shows high Cr (233–358 ppm) and Ni (51–78 ppm), and relatively low Rb (37.3–57.2 ppm), Sr (305–370 ppm) and Ba (203–297 ppm), whereas the high-Al diorite shows low Cr (21–156 ppm) and Ni (12–40 ppm), and relatively high Rb (38.3–86.9 ppm), Sr (305–725 ppm) and Ba (240–573 ppm) (Fig. 5g–k). All samples from the XK diorites show similar primitive-mantle-normalized trace element patterns with Nb, Ta, P and Ti negative anomalies (Fig. 7a). Their REE data show weakly fractionated REE patterns with (La/Yb) N ratio of 4.1–6.5 and variable Eu anomalies (Eu/Eu* = 0.69–1.12) (Fig. 7b).

The XK granodioritic porphyry (Table 1 and Fig. 5) is metaluminous (A/CNK = 0.87–0.97) and lies within the medium-K to high-K calc-alkaline fields (Fig. 6). All samples show enrichment of LILE (Rb, Ba, Th, U and K) and depletion of Nb, Ta, P and Ti (Fig. 7a) and display concave-upward REE patterns with weak negative Eu anomalies (Eu/Eu* = 0.84–0.86) (Fig. 7b). The concave-upward REE patterns and weak Eu anomalies in the XK granodioritic porphyry suggest its source containing significant residual amphibole and minor plagioclase.

Isotope geochemistry

Whole-rock Sr and Nd isotopic data are listed in Table 2 and plotted in Fig. 8. For all samples, their initial Sr isotopic ratio (I Sr) and εNd(t) values are calculated at t = 243 Ma. The SPX granodiorite shows a narrow variation of I Sr values ranging from 0.7081 to 0.7083 and εNd(t) values ranging from −8.0 to −7.6, with mantle-depleted Nd model ages (T DM) ranging from 1.53 to 1.70 Ga. The XK high-Al diorite has I Sr values of ~0.7060 and εNd(t) values of −1.5 to −1.1, with T DM = 1.35–1.40 Ga. The XK high-Mg diorite has I Sr values of 0.7060–0.7063 and εNd(t) values of −2.6 to −2.5, with T DM = 1.55–1.63 Ga. The XK granodioritic porphyry has I Sr = 0.7071–0.7075, εNd(t) = − 4.5 and T DM = 1.30 Ga.

Table 2 Data for whole-rock Sr and Nd isotopes
Full size table
Fig. 8
figure 8

I Sr versus εNd(t) plot for the SPX and XK plutons. Symbols as in Fig. 5. Triassic granites of the West Qinling from Zhang et al. (2007a); Triassic granites of the east Qinling from Zhang et al. (2007a), Qin et al. (2010b), Jiang et al. (2010). εNd(t) and I Sr values are calculated at t = 243 Ma. The parameters used in the modeling are as follows: Mantle-derived end-member: Sr = 305 ppm, Nd = 20 ppm, I Sr = 0.7060, εNd(t) = −1.1; Crustal-derived end-member: Sr = 370 ppm, Nd = 15 ppm, I Sr = 0.7083, εNd(t) = −7.9

Full size image

Zircon Lu–Hf isotopic data are given in Table 3. The initial εHf(t) and T DM2 were calculated using their U–Pb zircon ages in Table 3. For sample SPX-6 from the SPX pluton, twelve analyses show εHf(t) values of −4.7 to −3.6 and T DM2 of 1.49–1.57 Ga (Fig. 9a, b). For the XK granodioritic porphyry, nine analyses from sample 0951 show εHf(t) values of +3.5 to +5.3 with an average of +4.3 ± 0.5 and T DM1 of 0.68–0.76 Ga (Fig. 9a, b). Fourteen analyses from sample XK-8 show εHf(t) values ranging from +0.2 to +5.2, which displays a lager variation than the sample 0951(Fig. 9a). The single-stage model ages range from 0.68 to 0.89 Ga (Fig. 9b).

Table 3 Zircon Lu–Hf isotopic data
Full size table
Fig. 9
figure 9

Zircon Hf isotopic compositions of the SPX granodiorite and XK granodioritic porphyry. a 206Pb/238U age-εHf(t) variations and b Histograms of zircon T DM values for the SPX and XK plutons. The single-stage Hf model age (T DM1) is taken for positive εHf(t), and two-stage Hf model age (T DM2) is taken for negative εHf(t) (Zheng et al. 2006)

Full size image

Discussion

Petrogenesis

Shuangpengxi granodiorite

The SPX granodiorite has I Sr (~0.708), negative εNd(t) (~−8.0) and negative εHf(t) (~−4.2), coupled with Nd and Hf model ages of ~1.5–1.7 Ga, suggesting that its magma was predominantly derived from Meso-Paleoproterozoic crustal materials. As shown in Fig. 8, the granodiorite overlaps with the field of the West Qinling crust-derived granitoids (Zhang et al. 2007a), interpreted to have originated from partial melting of high-K basaltic protoliths in the lower crust.

Experiments of studies have demonstrated that partial melting of mafic rocks can generate melts of metaluminous granitic composition (Rushmer 1991; Wolf and Wyllie 1994). Moyen and Stevens (2006) have compiled nearly all experimental data about the partial melting of amphibolites, showing that melts formed by partial melting of low-K mafic rocks usually have low K2O and Na2O/K2O >1. Medium- to high-K calc-alkaline granitic magmas can be derived from partial melting of hydrous calc-alkaline to high-K calc-alkaline, mafic to intermediate metamorphic rocks (Roberts and Clemens 1993; Sisson et al. 2005). The melting reactions would result from fluid-absent breakdown of biotite and hornblende, leaving a plagioclase-enriched and amphibole-poor residue (Sisson et al. 2005). The SPX granodiorite is metaluminous and has high K2O with Na2O/K2O <1, indicating that a medium- to high-K basaltic protolith was required. Strongly fractionated REE patterns and moderate negative Eu anomalies (Table 1) of the granodiorite suggest magma generation by amphibole dehydration reaction with plagioclase-enriched residue. All these characteristics support a derivation of the SPX granodiorite from partial melting of medium- to high-K basaltic protolith.

Xiekeng diorites

The XK diorites contain two distinct groups: high-Mg diorite and high-Al diorite. The high-Mg diorite has high MgO (5.62–7.62 %), Mg# (59–64), Cr (233–358 ppm) and Ni (51–78 ppm), indicating that its magma was dominantly mantle-derived. The high-Al diorite has relatively low Mg#, Cr and Ni (Table 1), but its Mg# is still higher than the experimental melts produced by melting of metabasalts and eclogites (usually, Mg# <45) (Rapp and Watson 1995; Rapp et al. 1999), implying that its magma was also mantle-derived. The XK diorites (including high-Mg diorite and high-Al diorite) have I Sr = ~0.706 and εNd(t) = ~−3 to −1 (Table 2 and Fig. 8), suggesting that their magma could be derived from enriched lithospheric mantle.

At convergent margins, Sisson and Grove (1993) defined high-alumina basalts (HABs) with SiO2 ≤52 % (some <54 %) and high-alumina basaltic andesites (HABAs) with SiO2 ≤57 %. Both HABAs and HABs have Al2O3 >17 %, with a typical range from 18.5 to 20.5 %. Most HABs and HABAs have low MgO (<6 % and <5 %, respectively) and low Mg# (40–50). The chemical composition of the XK high-Al diorite (Table 1) is very similar to that of the HABAs, suggesting that they share a similar petrogenesis.

Two types of petrogenetic models have been proposed for the origin of low-MgO HABs and HABAs: (1) derivative melts produced by fractional crystallization of a primitive-mantle-derived magma (Sisson and Grove 1993) and (2) derivative melts that had undergone preferential accumulation of plagioclase (Crawford et al. 1987; Wagner et al. 1995). The above two models suggest that the low-MgO HABs and HABAs are not primitive-mantle-derived magma and that H2O plays an important role in their magma evolution. The main effects of H2O on crystallization of mafic melts are to decrease melt temperature and to suppress plagioclase crystallization relative to olivine and clinopyroxene (Danyushevsky 2001). Fractionation of some Al-poor mafic phases such as olivine and pyroxene will reduce the contents of MgO, Cr and Ni, but increase the contents of Al2O3 and other incompatible elements in the residual melts.

The XK high-Al diorite contains some hydrous minerals such as hornblende and biotite (Fig. 2a), indicating that its parental magma carried some extent of water. Both clinopyroxene and orthopyroxene are locally enclosed in plagioclase (Fig. 2a), indicative of early crystallization. The XK high-Al diorite displays low MgO, Cr and Ni, and high Al2O3, Rb, Sr and Ba, probably due to the fractionation of olivine and pyroxene. Sisson and Grove (1993) pointed out that low-MgO HABs and HABAs with Al2O3 >20 % are rare and those with >21 % Al2O3 are very uncommon. Some samples of the XK high-Al2O3 diorite contain Al2O3 up to 21.11 %, implying plagioclase accumulation, consistent with abundant plagioclase (60–75 %) in these samples (Fig. 2a). The XK high-Al diorite shows obvious positive correlations between Eu/Eu* ratios and the contents of CaO and Sr (Fig. 10a, b), together with positive Eu anomalies (Eu/Eu* = 1.1–1.2) further supporting accumulation of plagioclase. Thus, we suggest that the XK high-Al diorite was not directly solidified from a primitive-mantle-derived magma, but it had experienced fractional crystallization of olivine and pyroxene and/or preferential accumulation of plagioclase.

Fig. 10
figure 10

a CaO versus Eu/Eu* diagram; b Sr versus Eu/Eu* diagram; c Mg# versus I Sr diagram; d Mg# versus εNd(t) diagram; e Th/Yb versus Ba/La diagram (after Dokuz 2011); and f Th/Nb versus Ba/Th diagram (after Elliott et al. 1997). Symbols as in Fig. 5. FC fractional crystallization, AFC assimilation fractional crystallization

Full size image

The XK high-Mg diorite is similar to the high-Mg andesites (HMAs) in the Setouchi Volcanic Belt in MgO (Mg#), Cr and Ni (Tatsumi 2006 and references therein). However, high-Mg diorites should not simply interpreted to be the solidification of the HMA magma, because the high-Mg diorites may be the cumulate of some mafic magma (Kamei et al. 2004). The XK high-Mg diorite has relatively less MgO, Mg# and Ni (Table 1) than the primary arc basalts (usually MgO of 8–10 %, Mg# of 63–71 and Ni of 85–245 ppm) (Kelemen et al. 2003), suggesting olivine fractionation. The high Cr (233–358 ppm) for the XK high-Mg diorite suggests accumulation of pyroxene, which is also in good agreement with our petrographic observation that the pyroxene volume of the high-Mg diorite (sample 0946) is more abundant than that of the high-Al diorite (sample 0942) (Fig. 2a, b). For the high-Mg diorite, positive correlations between Eu/Eu* ratios and the contents of CaO and Sr (Fig. 10a, b) indicate plagioclase accumulation. Thus, we suggest that the XK high-Mg diorite was not directly solidified from the HMA magma, but was derived from fractional crystallization of olivine and/or preferential accumulation of pyroxene of a hydrous basaltic magma.

The I Sr of the XK high-Al diorite and the high-Mg diorite is uniform over a wide range of Mg#, and εNd(t) shows a small variation (Fig. 10c, d), suggesting that the two diorites are consanguineous, but have undergone different magmatic differentiation processes. Both of them have subduction-related geochemical signatures, characterized by enrichment of LILE (Rb, Ba and K) and LREE and depletion of HFSE (Nb, Ta and Ti) (Fig. 7a, b), suggesting that their mantle source was modified by slab-derived fluid or melt. We use some trace element ratios to assess the modification of slab-derived components.

Large-ion lithophile elements (e.g., Rb, Ba, Sr, K and U) are effectively transported by the fluid phases, but thorium, LREE and HFSE could be mobilized mainly by melts (Pearce and Peate 1995; Elliott et al. 1997; Class et al. 2000). Mafic magma showing high Ba/La and Ba/Th is commonly interpreted to reflect the addition of an aqueous fluid from the slab (Pearce and Peate 1995; Elliott et al. 1997), whereas high Th/Nb and Th/Yb ratios in mafic magma are ascribed to modifications by slab melts (Johnson and Plank 1999; Class et al. 2000). The XK diorites have high Th/Yb ratios (1.12–2.49) and Th/Nb ratios (0.36–0.94) and relatively low Ba/La ratios (10.0–26.3) and Ba/Th ratios (46.6–98.7) (Fig. 10e, f), suggesting the significant involvement of sediment-derived melt in their mantle source, as consistent with I Sr = ~0.706 and εNd(t) = ~−3 to −1 (Fig. 8) (e.g., Hawkesworth et al. 1997; Dokuz 2011).

Xiekeng granodioritic porphyry

The XK granodioritic porphyry has similar SiO2 as the SPX granodiorite (Table 1), but distinct Sr, Nd and Hf isotopic compositions (Fig. 8 and Fig. 9). The XK granodioritic porphyry is closely associated with the XK diorites, suggesting that there is a genetic link between them. Two possible processes can explain the generation of the XK granodioritic porphyry: (1) crustal assimilation and fractional crystallization (AFC) of the diorite magma and (2) mixing between mantle-derived and crust-derived magmas.

The potential crustal assimilation components could be Devonian to Triassic sediments in the West Qinling. According to Sm–Nd isotopic data from Chen et al. (2008), the calculated εNd(243 Ma) values for those sediments range from −8.3 to −16.3, lower than the values (−4.5) of the XK granodioritic porphyry. This indicates that assimilation of sediments in the magma was rather limited. Alternatively, we propose that the XK granodioritic porphyry was produced by mixing of mantle-derived melt with crustal melt. A simple two end-members mixing model was tested using Sr–Nd isotopic data. In this modeling, the XK diorite represents the mantle-derived magma end-member and the SPX granodiorite represents the crustal end-member. Sample XK-8 from the XK granodioritic porphyry plots on the mixing curve, showing that the XK granodioritic porphyry could contain ~40 % mantle-derived component (Fig. 8).

Zircon Hf isotopic data provide additional evidence for the magma mixing. Zircons in the sample 0951 have a homogeneous Hf isotopic composition, with average εHf(t) = 4.3 ± 0.5. Zircons in the sample XK-8 have εHf(t) = + 0.2 to +5.2, which shows a larger variation (with ~5 εHf(t) units) than the sample 0951 (Fig. 9a). Such variation can only be reconciled by the operation of open-system processes or reflect the hybrid melt at the time of crystallization (Kemp et al. 2007; Yang et al. 2007).

Tectonic implications

The Early Indosinian tectonic setting in the West Qinling is still controversial. Some researchers explain the Early Indosinian granitoids in the West Qinling to be subduction-related (Jin et al. 2005; Meng et al. 2005). However, other studies advocate that they formed in an early stage after collision (Zhang et al. 2006, 2008). The debate focuses mainly on the timing of final closure of the A’nimaque-Mianlue oceans. In the middle part of the West Qinling, two quartz diorite plutons (Yeliguan and Xiahe) have been dated at 245 ± 6 and 238 ± 4 Ma (Fig. 1b), respectively, and interpreted to have formed in the active plate margin (Jin et al. 2005). Meng et al. (2005) have synthesized the Triassic sedimentary stratigraphy in the East and West Qinling, pointing out that Lower Triassic-Anisian sequences in the East Qinling are composed exclusively of shallow-marine facies, with an angular unconformity above the early Anisian (245–237 Ma) sequence. However, in the middle part of the West Qinling, the shallow-marine facies did not disappear until the early Ladinian (237–228 Ma), and terrestrial facies were dominant in the Late Triassic (Yin et al. 1992; Meng et al. 2005). Therefore, Meng et al. (2005) proposed that the Mianlue Ocean in the East Qinling had been closed during the Middle Triassic due to the South China block’s clockwise rotation and diachronous collision with the North China block, while the A’nimaque Ocean in the West Qinling was still in the process of subduction and formation of the magmatic arc.

There are many lines of evidence arguing that the West Qinling was not an active margin in the Early Indosinian. Metamorphic ages of 242–221 Ma of the Mianlue ophiolite suggest the Mianlue Ocean had been closed at least in the Early Triassic (Li et al. 1996). The regional angular unconformity in the East Kunlun between the Late Permian Gequ Formation and underlying Middle Permian Shuweimenke Formation represents the closure of the A’nimaque Ocean and continental collision (Yin and Zhang 1998; Ren 2004; Chen et al. 2010). The Late Gequ Formation, a typical molasse formation, consists of a basal conglomerate with a Changxinggian biostratigraphic age (254–251 Ma) and also unconformably overlies ophiolitic mélange within the A’nimaque suture zone (Ren 2004; Chen et al. 2010). Sedimentologically, an abrupt change of facies occurred in the northwestern part of the West Qinling during the middle Triassic (Meng et al. 2005). However, the cessation of marine deposition cannot necessarily mark the continental collision, because marine deposition may continue in the residual sea after the collision, like for the Cenozoic India-Asia collision (e.g., Wu et al. 2008). As documented by this and previous studies (Feng et al. 2002; Jin et al. 2005; Zhang et al. 2006; Wang et al. 2010), the widespread magmatism in the middle and western parts of the West Qinling occurred in the Early Indosinian with an age-span of ~245–235 Ma (Fig. 1b), which post-dates Late Permian closure of the A’nimaque Ocean. We suggest that the A’nimaque Ocean had disappeared at the end of the Middle Permian and the Early Indosinian magmatism in the West Qinling formed in an early stage after collision.

For post-collisional granitoids, many studies proposed that slab break-off or lithosphere delamination can account for their magma generation (e.g., Davies and von Blanckenburg 1995; Wortel and Spakman 2000; Atherton and Ghani 2002; Bonin 2004; Massonne 2005; Altunkaynak 2007; Mahéo et al. 2009; Dokuz 2011). Lithospheric delamination may result from gravitational instability due to lithospheric thickening, and cause asthenosphere upwelling, that can induce partial melting of the remaining lithospheric mantle. Alternatively, the break-off of a subducted oceanic slab will create a gap, which is rapidly filled with the upwelling mantle asthenosphere, triggering partial melting of the overlying lithosphere. Generally, magmatism caused by the lithospheric delamination shows a regional distribution (Mahéo et al. 2009 and references therein), whereas magmatism resulted from slab break-off shows a linear distribution with rapidly crustal uplift parallel to the suture zone (Rogers et al. 2002; Buiter et al. 2002; Gerya et al. 2004; Mahéo et al. 2009; Duretz et al. 2011). The NW–SE linear distribution of the Early Indosinian intrusive rocks in the West Qinling (Fig. 1b) suggests that slab break-off is a possible mechanism for the Early Indosinian magma generation in the West Qinling.

Based on numerical modeling (Davies and von Blanckenburg 1995; Van de Zedde and Wortel 2001; Li et al. 2002; Duretz et al. 2011), slab break-off occurs at 40–300 km depth starting 4–19 M.yr after continental lithosphere subduction. The Early Indosinian magmatism in the West Qinling took place at ~245 Ma, ~9 M.yr after the onset (~254 Ma) of continental subduction (Wang et al. 1997; Chen et al. 2010). This is consistent with the timescale for slab break-off. The abrupt change from Middle Triassic marine facies to Upper Triassic terrestrial facies and the absence of the upper part of Middle Triassic (Ladinian) sequences in the northwestern part of the West Qinling (Yin et al. 1992; Meng et al. 2005) may reflect rapid crustal uplift due to slab break-off (Buiter et al. 2002; Rogers et al. 2002; Gerya et al. 2004). Furthermore, K-feldspar 40Ar/39Ar multiple diffusion domain (MDD) modeling, combined with the biotite 40Ar/39Ar age from the pluton in the east of Xiahe, reveals a rapid cooling history (17.75 °C/Ma) from 240 to 230 Ma, implying a fast uplift (Zheng et al. 2004), responding to the influx of mantle asthenosphere following slab break-off.

Conclusions

U–Pb zircon dating indicates that the SPX and the XK plutons were formed contemporaneously at ∼244 Ma. The Early Indosinian magmatism in the West Qinling fits well to the model of Deep Crustal Hot Zones (Annen et al. 2006, 2008). The XK diorites were derived from partial melting of enriched lithospheric mantle that had been modified by slab-derived melt. The hydrous basaltic magmas were successively emplaced in the lower crust, providing the necessary heat source and volatiles to induce the partial melting of the lower crust. The XK high-Al diorite was formed by the hydrous basaltic magmas that had experienced fractional crystallization of olivine and pyroxene and/or preferential accumulation of plagioclase, while the XK high-Mg diorite was formed by fractional crystallization of olivine and/or preferential accumulation of pyroxene. The SPX granodiorite originated from dehydration melting of lower crust due to heat input from mantle magma. The XK granodioritic porphyry was generated by mixing of residual mantle-derived melt and crustal melt. Combined with regional studies, we interpret this Early Indosinian magmatism in the West Qinling to result from break-off of the subducted A’nimaque oceanic slab soon after collision. The slab break-off model can explain the linear distribution of the Early Indosinian plutons and rapid crustal uplift during the Middle Triassic in the West Qinling orogen.