Introduction

The Buqingshan–A’nyemaqen tectonic mélange belt (BTMB) is in the western part of the A’nyemaqen Suture Zone, which is at the intersection of the East Kunlun, West Qinling, and Songpan–Bayan Har orogenic belts (Fig. 1a). The mélange belt forms a junction with the Qinghai–Tibetan Plateau in the Qinling orogen and is an important tectonic component of the complex eastern section of the Proto-Tethyan and Paleo-Tethyan ocean systems (Bian and Zheng 1992; Jiang et al. 1992; Xu et al. 1996, 2001, 2006a, 2013; Wang et al. 1997a; Pei 2001; Zhang et al. 2003; Pei et al. 2018). The region experienced two main tectonic phases: the Hercynian and the Indosinian. This multi-stage history of the region is vital to understand the tectonic evolution of continental China (Jiang et al. 1992; Xu et al. 1996; Yin and Zhang 1997; Pan et al. 2012; Dong et al. 2018).

Fig. 1
figure 1

a Geological map of the middle-western sections of the Central Orogenic Belt of China; b geological sketch map of the Buqingshan-A’nyemaqen tectonic mélange belt (BTMB) and its adjacent areas; c geological map of the BTMB, southern margin of the EKOB (modified from China University of Geosciences (Wuhan) Geological Survey 2000)

Full size image

Controversy regarding the evolution process and geodynamic setting of BTMB remains. Xu et al. (2006b, c, d) suggest that the BTMB subduction–accretion complex is the product of subduction of the northern branch of the Paleo-Tethys oceanic crust. Other researchers have proposed that the East Kunlun orogenic belt (EKOB) and BTMB together form a large subduction–accretion–type tectonic mélange belt, and the paleo–ocean basin represented by the Buqinshan ophiolite continued to evolve from the Cambrian to the Early Triassic. The central of the East Kunlun and the Buqinshan ophiolites had accumulated together along the southern margin of the EKOB in the Middle and Late Triassic (Jiang et al. 1992, 2000; Wang et al. 1997b, 1999; Bian et al. 2001a, b, c, 2004, 2007; Li et al. 2007; Li et al. 2015a; Xiong et al. 2015; Liu et al. 2011a, b, c). To clarify the evolution of the BTMB, more research and data are needed and provided in this paper.

Due to their characteristic stability, zircons and their U–Pb ages are extremely useful in constraining the tectonic evolution of a region (Hartmann 2001; Wu et al. 2008; Wang et al. 2011; Zhang et al. 2015; Hoskin and Black 2000; Hoskin and Ireland 2000; Griffin et al. 2000; Belousova et al. 2002; Rubatto 2002; Hoskin and Schaltegger 2003; Hoskin 2005; Hanchar and Westrenen 2007). Previous studies mainly focused on the EKOB along the north side of the BTMB and obtained considerable evidence from zircon U–Pb geochronology (Chai et al. 1984; Harris et al. 1988; Chen et al. 2002a; Liu et al. 2004; Mo and Pan 2006; Sun et al. 2009; Chen et al. 2013a, b, c, 2017a, b, 2018a, b, c; Li et al. 2013a, b, 2017a, 2018a, b, 2019; Li et al. 2013c, 2018c; Chen et al. 2016; Deng et al. 2016; Hu et al. 2017; Zhang et al. 2017). These studies found granitoids with both Early Paleozoic ages in the BTMB (Liu et al. 2011a; Li et al. 2014a, 2015a, b; Li et al. 2014b, 2017b), as well as younger ones in the eastern section of the EKOB, dated at 230–256 Ma (Liu et al. 2004; Chen et al. 2013a, b, c, 2017a, 2018a, b, c; Ding et al. 2014; Chen et al. 2016; Li et al. 2018a). Late Triassic ages of granitoids have so far only been found in the Gerizhuotuo diorite (Li et al. 2013c). To expand the database on igneous rocks in the BTMB, we here analyze the petrology, geochronology, and geochemistry of intermediate–basic dikes in the BTMB. We also discuss the source and petrogenesis of the dikes, and made an attempt to provide new evidence for the Late Triassic tectonic evolution of the BTMB.

Geological background

The E-W striking BTMB discontinuously is more than 700 km long and approximately 10–20 km wide. It starts from Maqin in the east and extends across the Majixueshan and Tuosuohu to Buqingshan and southeast Heicigou to connect with the mafic–ultramafic rocks of Muzitage (Molnar et al. 1987; Burchfiel et al. 1989; Bian et al. 2004; Fig. 1a). To the north, the BTMB is separated from the East Kunlun and West Qinling orogens by the southern East Kunlun Fault, and to the south, it is separated from the Songpan–Bayan Har Orogen by the Changshitou Fault. The BTMB itself forms a suture zone between the Bayan Har and East Kunlun Blocks and is a product of two phases of ocean–continent subduction–collision in the Early and Late Paleozoic (Zhang et al. 1999). It also is part of the East Tethys Ocean tectonic domain (Jiang et al. 1992; Bian et al. 1999a, b, 2001c; Chen et al. 1999, 2000a, 2004; Zhu et al. 1999; Yang et al. 2004; Guo et al. 2007; Liu et al. 2011a, b, c; Pei et al. 2018).

The BTMB comprises the Lower–Middle Permian Maerzheng Formation (P1–2m), including Early Paleozoic and Late Paleozoic ophiolites, Paleozoic rock mass, seamount basalts, and limestone (Liu et al. 2011b; Li et al. 2013c, 2014b, 2017b; Li et al. 2014a, 2015a, b; Pei et al. 2015, 2018; Yang et al. 2016; Pei et al. 2017). In the northern part of the ophiolite belt is the Middle Proterozoic Kuhai Group (Pt2K), which comprises marble, biotite–quartz schist, gneiss, and amphibolite and constitutes the metamorphic basement rocks. A nappe composed of the Upper Carboniferous to Lower Permian Shumenweike Formation (C2P1–2sh), which primarily comprises carbonates with apparent reef affinities (Fig. 1b), covers the entire area.

The BTMB also includes various mafic to felsic dikes that intruded rock masses of various ages and diverse stratigraphy. The dikes intruded parallel to bedding or intersect it at an oblique angle. Their NW orientation appears fault controlled. The rock types primarily include diabase, diorite, granodiorite, granite, and syenogranite. According to the intersection relationships of the dikes, their relative intrusion order may be represented as diabase → diorite → granodiorite → granite → syenogranite. Indosinian dikes are widely distributed in the EKOB, indicating that the Indosinian magmatic event was widespread across the region.

Analytical methods

Petrographic sample preparation and microscopy

Thin sections of rocks were completed at the Laboratory of Mineralization and Dynamics of Chang’an University. During sample preparation, after cutting the rock sample, it is ground with different specifications of sandpaper. Then, the polished rock slab and the glass slide are dried in a preheated oven, and the surface of the rock block is attached to the glass slide using epoxy resin. After cooling, it was cut down to a thickness of 1 mm using a precision cutter and then ground to a light-transmission level (~ 30 μm), and polished with diamond paste and alumina for about 10 min each until the surface of the section was flat and free of mechanical scratches. Petrographic observation and photomicrography were performed using an ORTHOPLAN partial and reflective research microscope from Leitz, Germany.

LA–ICP–MS testing

One sample (MNT-21) of the diorite dike was evaluated for isotopic dating, whose geographic coordinates are N 35° 27′ 30.7″ and E 97° 39′ 42.3″ (Fig. 1).

Rock samples were crushed to 80–100 meshes using conventional methods and separated by flotation and electromagnetism techniques. Well-formed, crystal-shaped, and transparent zircons were handpicked using a binocular microscope. Zircon grains were mounted on a two-sided adhesive tape and fixed with colorless transparent epoxy resin until fully solidified; the surface was polished to expose the interior of the zircons. Cathodoluminescence (CL) microphotography images were taken with a Cameca electron probe X-ray microanalyzer at the Institute of Geology and Geophysics, Chinese Academy of Sciences. The analysis voltage was 15 kV and the current was 19 nA.

The in situ U–Pb isotopic age analysis of zircons was carried out following the standard test procedure using an LA–ICP–MS at the State Key Laboratory of Continental Dynamics, Northwest University. The analysis instruments were an Elan 6100DRC Type Quadrupole Perch Mass Spectrograph and a Geolas200M excimer laser ablation system (193 nm, Geolas200M, Lambda Physic). The facula beam’s diameter of laser ablation was 30 μm, and the depth of laser ablation samples was 20–30 μm. For the calculation of zircon ages, the international standard zircon 91500 was used as an external standard; for the element content analysis, the artificial synthetic silicate glass NIST SRM610 of the American National Standard Substance Bureau was adopted as an external standard. 29Si was used as the internal standard element. The isotopic ratio and element content data were analyzed with GLITTER (ver. 4.0, Macquarie University) software, general plumbum adjustment was conducted using the Andersen software (Andersen 2002), and age calculation and concordia diagram drafting were completed using ISOPLOT (3.0 edition) (Ludwig 2003). The detailed experimental principles, technological process, and instrumentation parameters were the same as those reported by Yuan et al. (2003, 2004).

Geochemical analysis

Nine samples were selected for the analysis of major and trace elements. The samples were ground to 200 meshes and the major and trace elements were determined by the State Key Laboratory of Lithosphere Evolution, Institute of Geology and Geophysics, Chinese Academy of Sciences. The major elements were tested using the method of X-ray fluorescence spectrometry (XRF–1500). To determine the content of oxide, a sheet glass made of 0.5-g samples and 5 g lithium tetraborate was tested using the Shimadzu XRF–1500, with a precision of > 2–3%. The contents of the trace and rare earth elements (REEs) were analyzed by ICP–MS (ElementII). The samples were prepared using the acid-solubility method, which has an analytic precision of > 10% (according to the national standards GSR–1 and GSR–2), but the precision is > 5% when the element content is > 10 ppm. The detailed analysis methods were described by Chen et al. (2000b, 2002b).

Results

Petrography

The sampled dikes can be classified as diabase and diorite. The diabase, gray–black and fine-grained with typical diabasic texture (Fig. 2a, b), is composed primarily of plagioclase (50–60%), pyroxene (± 20%), and minor amounts of hornblende, biotite, and opaque minerals (8–10%). The diorite, dark–gray and fine- to particle-grained with hypidiomorphic granular texture, is composed primarily of plagioclase (65–70%), hornblende (± 20%), quartz (3–5%), and biotite (5–8%) with accessory amounts of zircon, apatite, and magnetite (Fig. 2c). Plagioclase is the most abundant phenocryst and commonly has polysynthetic twins (Fig. 2d). Biotite and hornblende are dark–green or brownish. Some euhedral plagioclase grains are included within quartz crystals (Fig. 2d).

Fig. 2
figure 2

a, b Field photographs and microscopic photos of the diabase. c, d Field photographs and microscopic photos of the diorite in the BTMB, southern margin of the EKOB Am, amphibole; Ap, apatite; Bi, biotite; Cpx, clinopyroxene; Pl, plagioclase; Qz, quartz; b, d, crossed polarizers, 20 times

Full size image

Whole-rock chemistry

The intermediate–basic dikes of BTMB samples have relatively low concentrations of SiO2 (51.96–59.33 wt%), K2O (1.45–2.70 wt%), K2O+Na2O (4.00–5.29 wt%), TiO2 (0.30–0.60 wt%), Al2O3 (10.49–13.95 wt%), and high concentrations of total Fe2O3 (6.29–8.68 wt%), MgO (5.16–11.79 wt.%), and CaO (5.86–10.15 wt%) (Table 1). All samples exhibit subalkaline trends in a total alkali–silica (TAS) diagram (Fig. 3), and the intermediate dikes plot in the andesite and basalt fields.

Table 1 Major element data components (wt%) and trace element abundance (ppm) for the intermediate–basic dikes in the BTMB, southern margin of the EKOB
Full size table
Fig. 3
figure 3

TAS diagrams for the intermediate–basic dikes (after Rickwood 1989) in the BTMB, southern margin of the EKOB

Full size image

From the characteristics of the intermediate–basic dikes of the BTMB in Table 1, we can conclude that the concentration of REEs are 80.23–189.19 ppm, and the ratios of light RREs (LREEs) to heavy RREs (HREEs) (LREE/HREE = 6.53–13.15), with (La/Yb) N = 5.62–19.16, (La/Sm) N = 2.67–3.35. The HREE losses may be due to residual garnet and amphibole at the source (Patino–Douce and Johnston 1991). Chondrite-normalized REE patterns display an incline to the right (Fig. 4a). The diorite samples generally show similar patterns, with differences in abundance and weak negative Eu anomalies (δEu = 0.50–1.10) that may have been induced by the differentiation of plagioclases and K-feldspars.

Fig. 4
figure 4

a Chondrite-normalized REE patterns (chondrite data for normalization taken from Sun and McDonough 1989). b Trace element spider diagram (primitive mantle data for normalization taken from McDonough and Sun 1995) for the intermediate–basic dikes in the BTMB, southern margin of the EKOB

Full size image

The intermediate–basic dikes are characterized by high Rb, Th, and low Rb/Sr ratios (0.14–0.37); Ra/Ba ratios (0.04–0.28); and high K/Rb ratios (75.52–355.14) (Table 1). On the primitive mantle-normalized spidergram (Fig. 4b), they are characterized by enrichment in LILEs relative to HFSEs, showing notable negative Ta, Nb, Sr, P, and Ti anomalies and positive U, Ce, Nd, Sm, and Hf anomalies. The diorite and diabase dikes show different characteristics, with anomalous Zr, Th, Rb, and Ba.

Zircon U–Pb ages

Zircons from the diorite dike (sample MNT-21) are euhedral to subhedral (Fig. 5a) with lengths in the range 100–250 μm and width/length ratios of 1:1 to 1:10.

Fig. 5
figure 5

a CL images and ages of single zircon U–Pb of diorite dike (the yellow circle is the sample test point position). b, c LA-ICP-MS zircon U–Pb concordia diagram of diorite dike in the BTMB, southern margin of the EKOB

Full size image

Thirty zircon analyses yielded Th concentrations in the range 70–1267 ppm and U concentrations in the range 103–1979 ppm (Table 2). A plot of total U versus Th shows a good linear relation (Fig. 6a). The chondrite-normalized REE patterns (Sun and McDonough 1989) (Fig. 6b) indicate that the zircons are depleted in LREEs relative to HREEs with positive Ce anomalies and negative Eu anomalies (δEu = 0.13–0.48) (Table 3).

Table 2 LA-ICP-MS zircon U–Pb analytic data and Ti geological temperature calculation for diorite dikes in the BTMB, southern margin of the EKOB
Full size table
Fig. 6
figure 6

Th-U diagrams and chondrite-normalized REE patterns (chondrite data for normalization taken from Sun and McDonough 1989) of Zircon for diorite dike in the BTMB, southern margin of the EKOB

Full size image
Table 3 Zircon trace element (ppm) analytical results of diorite dikes in the BTMB, southern margin of the EKOB
Full size table

Except for the six apparent discordia data (e.g., analysis spot numbers 002, 005, 015, 022, 023, 029) and captured or inherited zircon ages (e.g., analysis spot numbers 020, 627 Ma) (Table 2), the remnant 23 zircon grains form a single population (Fig. 5b). The 23 testing points are perfectly concordant among 206Pb/238U and 207Pb/235U ages, the concordia age of 204 ± 2 Ma (MSWD = 0.46) (Fig. 5b), and the 206Pb/238U weighted average age of 205 ± 1 Ma (MSWD = 0.88) (Fig. 5c). We interpret this to represent the Late Triassic (Rhaetian) crystallization age of the diorite dike.

Discussion

Formation time of the dikes

The zircons analyzed were taken from the diorite dike (sample MNT-21) in the BTMB. The majority of these zircons show oscillatory zoning structures, indicating that they are magmatic zircons (Belousova et al. 2002; Wu and Zheng 2004; Siebel et al. 2005). This is corroborated by the high Th/U ratios (0.47–2.43) (Table 2), indicating a magmatic origin (Vavra et al. 1999; Claesson et al. 2000; Wu and Zheng 2004). A plot of total U versus Th shows a good linear relation, which also is a feature typical of magmatic zircons (Fig. 6a; Vavra et al. 1999; Claesson et al. 2000; Wu and Zheng 2004). The chondrite-normalized (Sun and McDonough 1989) REE patterns (Fig. 6b) indicate that the zircons are depleted in LREEs relative to HREEs. The zircons display positive Ce anomalies (δCe = 1.02–79.30) and negative Eu anomalies (δEu = 0.13–0.48) (Table 3), which are consistent with characteristics of crustal magmatic zircons (Hoskin and Schaltegger 2003).

Cathodoluminescence (CL) imaging of sample MNT-21 shows the complete zircon morphology, obvious oscillatory zonal structure, and Th/U values of all the analysis points greater than 0.1, demonstrating magmatic zircon characteristics (Tapia–Fernandez et al. 2017). The 23 analysis points obtained a relatively consistent 206Pb/238U apparent age, with the 206Pb/238U weighted average age of 205 ± 1 Ma, considered to be the crystallization age of the diorite dike.

Petrogenesis

The diabase dikes are relatively enriched in LREE and LREE, with Nb and Ta negative anomalies, and have relatively low Nb/U ratios (2.19–4.10) and high Zr/Nb ratios (13.07–16.61) that are significantly different from those of the standard OIB (Nb/U = 47.06 and Zr/Nb = 5.83, according to Sun and McDonough 1989). The diabase dikes in the BTMB have obvious Nb, Ta, and Ti negative anomalies, which reflect the geochemical characteristics of island-arc basalt, suggesting either crustal contamination or metasomatism by subduction fluid (Green and Pearson, 1987; Rollinson 1993; Green 1995; Barth et al. 2000). High primitive mantle-normalized (Th/Nb)N ratios (≫ 1, Saunders et al. 1992) and low Nb/La ratios (< 1, Kieffer et al. 2004) are two reliable trace element indicators for crustal contamination. The BTMB diabase dikes do have high (Th/Nb)N ratios (7.89–12.56) and low Nb/La ratios (0.25–0.33), suggesting crustal contamination. The diorite dike shows characteristics that are opposite to those of the diabase dikes, with lower (Th/Nb)N ratios (1.93–5.30) and higher Nb/La ratios (1.02–2.22). This contrast suggests that the diorite and diabase magmas may have had different source areas.

Magma temperatures

High-temperature and high-pressure experiments indicate that the Ti content in zircon is closely related to temperature, producing a logarithmically linear relationship. That is, the zircon Ti thermometer has an empirical formula temperature estimation whose error is generally no more than 10 °C. This thermometer is simple and practical, so it is utilized by many researchers for great practicality (Watson et al. 2006; Zhao 2010; Gao and Zheng 2011).

We obtained data for 30 measuring points in the BTMB diorite dike (MNT-21), among which 23 are effective points used in our calculations. Plugging the Ti content measured in the sample zircon into the formula, the lowest temperature of the intermediate–basic dikes in BTMB ranges from 702 to 981 °C, averaging ~ 777 °C. TiO2–SiO2 and P2O5–SiO2 geothermometric measurements also verify that the formation temperature of the diorite dike is ~ 800 °C (Fig. 7a, b), confirming that they were formed in a high-temperature environment.

Fig. 7
figure 7

a TiO2-SiO2 (after Harrison and Watson 1984) and b P2O5–SiO2 (after Green and Pearson 1986) diagrams for diorite dike in the BTMB, southern margin of the EKOB

Full size image

Tectonic setting

Heat flow at an extremely high temperature is required to melt large amounts of mafic rocks (England and Thompson 1986; Thompson and Connolly 1995). However, in general, this condition holds only in the tectonic setting represented by continental collision → continental crust thickening → crustal extension and thinning → asthenosphere uplift (Han et al. 2000; Vanderhaeghe 2009; Bea 2012; Hasterok and Webb 2017). The intermediate–basic dikes in the BTMB formed in high-temperature environments, and the depletion of Sr, P, and Ti indicates characteristics of continental arc granite, which may be related to the collision of the Bayan Hara and East Kunlun Blocks along the BTMB in Late Hercynian–Early Triassic.

Generally, the geochemical–element ratios of island arc basalts and partially depleted mid-ocean ridge basalts are Nb/La < l, Hf/Ta > 5, La/Ta > 15, and Ti/Y < 350, while those of intraplate basalts, transitional mid-ocean ridge basalts, and enriched mid-ocean ridge basalts are considerably different (Condie 1989; Fitton 2007; Niu 2016). The diabase dikes in the BTMB have Nb/La ratios of 0.25–0.33, Hf/Ta ratios of 3.25–8.19, La/Ta ratios of 30.36–74.63, and Ti/Y ratios of 174.85–346.53. This suggests that the formation settings of these diabase dikes are unrelated to the lithotectonic settings of intraplate basalt, transitional mid-oceanic ridge basalt, and enriched mid-oceanic ridge basalt, as is the case for island arc basalt and depleted mid-ocean ridge basalt. The element ratios of island arc basalts are Th/Yb > 0.1, Th/Nb > 0.07, Nb/La < 0.8, and Hf/Th < 8, whereas those of depleted mid-ocean ridge basalts are considerably different (Condie 1989; Fitton 2007; Niu 2016). The diabase dikes in the BTMB have Th/Yb ratios of 6.37–8.38, Th/Nb ratios of 0.95–1.52, Nb/La ratios of 0.25–0.33, and Hf/Th ratios of 0.11–0.54, which are clearly of island arc nature. The diorite dike in the BTMB has Nb/La ratios of 1.02–2.22, Hf/Ta ratios of 0.97–2.20, La/Ta ratios of 5.80–13.44, Ti/Y ratios of 96.18–405.17, Th/Yb ratios of 4.95–7.70, Th/Nb ratios of 0.23–0.64, and Hf/Th ratios of 0.11–0.76, mostly similar or close to the ratios of the diabase dikes.

Therefore, it is after the Bayan Hara Block subduction or collision with the East Kunlun Block in Late Hercynian–Early Triassic that mantle melt was enriched during crustal relaxation, was emplaced up along tensile faults, and was contaminated by crustal materials, forming the Late Triassic dikes.

Tectonic significance

Since the Buqingshan–A’nyemaqen ocean subduction, the EKOB developed several calcium metaluminous potassium basic character arc magmatic rocks during the subduction stage of 260–240 Ma (Guo et al. 1998; Yang et al. 2005; Mo et al. 2007; Chen et al. 2013a, b, c, 2017a, 2018a, b, c; Liu et al. 2014; Chen et al. 2016; Hu et al. 2016; Hu et al. 2017; Zhang et al. 2017; Li et al. 2018a; Fig. 8a). Intrusive rocks in the East Kunlun area that are associated with the 240 to 225 Ma collision between the Bayan Hara and East Kunlun Blocks are rare. This stage of intrusive rocks have the syn-collision granite characteristics (Zhang et al. 2012; Xia et al. 2014; Xiong et al. 2015; Chen et al. 2018b; Fig. 8b). After the Late Triassic, post-collision granitoids are represented by the stitching Gerizhuoto diorite pluton in the BTMB, which suggests that by then, the Buqingshan–A’nyemaqen Ocean had already been closed (225.8 ± 1.5 Ma, Li et al. 2013c; Fig. 8c). In the southern part of the EKOB, north of the BTMB, the angular unconformities between the Babaoshan Formation (T3b) and the overlain Middle Triassic Naocangjiangou Formation (T2n) and between continental volcanic rocks and the underlying strata of the Upper Triassic Erashan Formation (T3e) in the northern part of the East Kunlun area mark the end of continental collision between the Bayan Hara and East Kunlun Blocks (Liu et al. 2011a; Li et al. 2012).

Fig. 8
figure 8

Schematic diagrams illustrating the tectonic and magmatic evolution of in the BTMB, southern margin of the EKOB a subduction stage (260–240 Ma). b Collision stage (240–225 Ma). c Post-collision stage (225–205)

Full size image

Comparison of dikes in the BTMB with the 226 Ma Binggou mafic dike swarm of the EKOB (Liu et al. 2017) reveals that at least since the Late Triassic, the EKOB had shifted from compressional to extensional environments (Liu et al. 2017), and intermediate–basic dikes in the BTMB formed later than the various abovementioned intrusive rocks. These results show that at ~ 225 Ma, the BTMB and the adjacent area had entered a post-orogenic extension stage (Fig. 8c), as also indicated by the development of the stitching pluton and mafic dike swarm. This implies that the intermediate–basic dikes in BTMB are the product of the post-orogenic intracontinental extension tectonic setting, suggesting that in 225–205 Ma, the BTMB was at the post-orogenic stress–relaxation stage of Bayan Hara Block subduction–collision with the East Kunlun Block (Fig. 8c). Decompression melting of the post-collision stage lifted the thermal interface, thereby inducing the melting of the mantle wedge and the formation of the abovementioned dikes.

Conclusions

Our comprehensive geochronological and geochemical study of the intermediate–basic dikes in the BTMB (along the southern margin of the EKOB, Tibetan Plateau, China) provides the following conclusions:

  1. (1)

    Geochemical analyses show that the dikes have the characteristics of arc magmatic suites. REEs have low concentrations and incline to the right with weak negative Eu anomalies. The diorite dike magma was formed at high temperature (~ 777 °C) by partial melting of mantle material and possibly was contaminated by crustal material when ascending.

  2. (2)

    The zircon U–Pb age of the diorite dike is 205 ± 1 Ma (MSWD = 0.88), which shows that the intrusion formed in the Late Triassic (Rhaetian) and thus also represents the latest tectonic magmatism in the BTMB.

  3. (3)

    The intermediate–basic dikes in the BTMB are the products of enriched mantle upwelled and emplaced in the crustal relaxation period after the subduction–collision of the Bayan Har and East Kunlun Blocks during the Late Hercynian–Early Triassic.