Introduction

The spatio-temporal evolution of the Paleo-Tethys Ocean has been a topic of debate for decades, for instance, the Shuanghu Paleo-Tethys in northern Tibet (e.g., Li 1987; Deng et al. 1996; Kapp et al. 2000; Li e al., 2008a, b, 2019; Pullen et al. 2008; Yin and Harrison 2000; Kapp et al. 2003; Zhang et al. 2006a, b, 2007, 2011, 2016, 2018; Zhang and Tang 2009). The LongmuTco–Shuanghu Suture Zone in the Tibet Plateau of western China runs through the hinterland of the Qiangtang Basin and separates it into northern and southern (or eastern and western) Qiangtang blocks (Fig. 1a). The suture zone is marked by the central Qiangtang metamorphic belt (CQMB) (Fig. 1b), including an assemblage of eclogites, eclogitic metasediments, metasedimentary schists and gneisses, and marble (e.g., Zhang et al. 2006a, b, 2012), which is key to understand the tectonic evolution of the Paleo-Tethyan realm and the crustal architecture of the Tibetan plateau. Although at least three distinct models have been proposed to interpret the nature and origin of the CQMB: (i) a continental rift model (e.g., Deng et al. 1996); (ii) an allochthonous underthrust mélange model (Kapp et al. 2000, 2003); (iii) an in situ Paleo-Tethyan suture model (e.g., Zhang et al. 2006a, b), more and more evidences from abundant ophiolite slices (Li et al. 2008a, b; Wu 2013; Zhai et al. 2004, 2007; Zhang et al. 2016), low-temperature, high–ultrahigh-pressure metamorphic complexes (Li et al. 2006; Zhai et al. 2011a, b; Zhang et al. 2006a, b, 2018), paleomagnetic and faunal–floral data (Lu et al. 2019), and widespread magmatism related to subduction and collision paralleled to the suture zone (e.g., Hu et al. 2014; Li et al. 2015a, b; Lu et al. 2017; Peng et al. 2015; Zhang et al. 2011) supported the third model, and i.e., the LongmuTco–Shuanghu suture zone most likely represents the remnant of the Paleo-Tethys Ocean and the nature and origin of CQMB are mainly constrained by the evolution of the Paleo-Tethys Ocean (e.g. Li et al. 2006; Zhang et al. 2006a2006b201120162018; Zhai et al. 2011a2011b; Lu et al. 2019). Researchers have proposed the opening of this ocean in the Cambrian (e.g., Zhai et al. 2013) or the Devonian (Zhang et al. 2014a, b) by ophiolite studies and the closure in the Middle Triassic (e.g., Hu et al. 2014; Peng et al. 2015; Tao et al. 2014) or late Triassic (Wang et al. 2019) through granite data or after late Triassic based on metamorphic rock evidence (Zhang et al. 2018). Hence, the opening and closure time of the Shuanghu Paleo-Tethys Ocean remains controversial. In addition, whether bi-directional subduction of the Shuanghu Paleo-Tethys Ocean occurred during its evolution is also a topic of debate. Predominant studies insisted in a unidirectional subduction of the oceanic plate between the Southern and Northern Qiangtang blocks beneath the latter during the evolution of the Shuanghu Paleo-Tethys Ocean (e.g., Zhang et al. 2006a, b; Zhai et al. 2011a, b; Li et al. 2019). However, the other studies asserted that bi-directional subduction of the Shuanghu Paleo-Tethys Ocean occurred during its evolution (e.g., Liu et al. 2011; Zhao et al. 2015). Hence, it is necessary to further explore these issues.

Fig. 1
figure 1

a Simplified tectonic map of Tibet, western China; b simplified geological map of the central Qiangtang metamorphic belt (modified after Zhang et al. 2011)

Full size image

Although ophiolite suites have provided a strong evidence for the existence of an ancient ocean, they are usually sporadically preserved during subduction and surface processes. Thus, the accurate evolution of the ocean cannot be constrained. The study on petrogenesis with integrated metamorphic and igneous rocks will give a proper avenue to determine the secular evolution of an ancient ocean (Zhou et al. 2009; Wu et al. 2011; Yu et al. 2012; Wilde and Zhou 2015; Zhu et al. 2015). Hence, here we carry out petrographic, chronological, and geochemical study on garnet–phengite–quartz schist and mafic rocks in the Rongma area, northern margin of the Southern Qiangtang block, to determine the metamorphic processes of the schist and the petrogenesis of the mafic rocks and to further reveal the early Mesozoic tectonic evolution of the Shuanghu Paleo-Tethys Ocean in northern Tibet.

Regional geology

The Qinghai–Tibet Plateau consists of a series of tectonically splicing terranes, including the Qaidam, Songpan–Ganzi, Qiangtang, and Lhasa terranes, separated by several well-known sutures (Yin and Harrison, 2000; Zhang et al. 2012; Fig. 1a). The Qiangtang Basin is located in the northern margin of the Tibet Plateau and is separated into the southern and northern (or eastern and western) Qiangtang blocks by the LongmuTco–Shuanghu Suture Zone. The southern and northern Qiangtang blocks differ greatly in sedimentary strata and palaeontologic characteristics; the Southern Qiangtang block might have an affinity to the Gondwana, whereas the Northern Qiangtang block might have an affinity to the Yangtze plate in the Eurasia (Li et al. 2004, 2005). The study area is located in Yadan, Rongma Town, northern margin of the southern Qiangtang, and is geotectonically located ~ 80–100 km south of the LongmuTco–Shuanghu Suture Zone (Fig. 1a, b). Regional stratigraphy mainly comprises the late Plaeozoic–Cenozoic strata, such as the Devonian sandstone and siltstone, the Carboniferous–Permian limestone with sandstone, the Upper Triassic limestone with sandstone, the Upper Cretaceous conglomerate and sandstone, the Paleogene terrigenous clastic rocks, and the Quaternary terrigenous talus accumulations. Accreting complex and sedimentary cover zones widely occurred in the region (Wang et al. 2009). The accreting complexes contain argillaceous, quartz schist, marble, pillow basalt, gabbro and diabase, and so on, which have suffered from intense deformation, metamorphism, and feature disordered sequence (Liu et al. 2011); The sedimentary cover zone features intense deformation and weak metamorphism (Wang et al. 2009). Regional strike-slip fault is developed, displaying multi-staged metamorphism–deformation mainly with a NE–SW trending (Fig. 2a, b).

Fig. 2
figure 2

a Simplified geological map of Rongma in the northern margin of the Southern Qiangtang block, northern Tibet (modified after Liu et al. 2011); b geological sketch map of Yadan area in the Rongma Town, in the northern margin of south Qiangtang

Full size image

The late Plaeozoic, Mesozoic, and the Cenozoic igneous rocks widely occurred in the region (Wang et al. 2019). The late Paleozoic magmatic rocks are predominantly outcropped in the northwest, with the Mesozoic and the Cenozoic magmatic rocks mainly in the central area (Fig. 2a, b). Intrusive rocks include mafic dykes [e.g., sillite, gabbro, and diabase–gabbro, as well as granitoids (Fig. 2a, b)]. The mafic dykes were mainly formed in the late Carboniferous–early Triassic (Li et al. 2005; Zhai et al. 2009; Dan et al. 2020, 2021). Volcanic rocks mainly comprise the Paleozoic basalt fragments and the Cenozoic Nadingcuo Formation basalts. Regional metamorphic rocks mainly include garnet–phengite–quartz schist, blue schist, slab, phyllite, ampholite, marble, and metamorphic quartz sandstone (Fig. 2a, b). Field sections show that the garnet–phengite–quartz schist spreads nearly NW–SE, and is in fault contact with mafic rocks and marble (Fig. 3a). In addition, pillow basalt unconformable contact with garnet–phengite–quartz schist is also recognized in the area (Fig. 3b, c). Two stages of tectonic deformation can be observed in the region, with moderate–shallow layered shearing in the early stage, and secondary intense deformation due to N–S trending overthrusting in the late stage (Wang et al. 2019).

Fig. 3
figure 3

Samples and microscopic characteristics of diabase and garnet–phengite–quartz schist from the Rongma area in the northern margin of south Qiangtang. a Field photo of garnet–phengite–quartz schist in fault contact with mafic rocks and marble; b, c pillow basalt and siliceous rocks; d field photo of the Early Triassic diabase dyke; e the Early Triassic diabase hand specimen; f Microscope photo of the diabase under cross-polarized light; g field photo of garnet–phengite–quartz schist; h microscope photo of garnet–phengite–quartz schist under cross-polarized light; i microscope photo of garnet in the garnet–phengite–quartz schist under single-polarized light. Ab albite, Ms phengite, Pl plagioclase, Px pyroxene, Q quartz, Grt garnet

Full size image

Petrography and methodology

Petrography

Diabase and garnet–phengite–quartz schist and gabbro samples were collected in Rongma Town, northern margin of southern Qiangtang, Tibet (Fig. 2b). Diabase (A1638) is colored off-grayish green, massive, with a fine diabasic texture. It mainly includes plagioclase (~ 55 vol.%) and secondary pyroxene (~ 40 vol.%) with minor minerals, such as apatite and magnetite. Plagioclase occurs as automorphic and hypautomorphic plates; plate lengths range from 0.2 to 1.5 mm. Plate crystals show intense sericitization and argillization, making the crystals muddy. Some of the crystals are wrapped by pyroxene, forming an ophitic texture. Pyroxene (Px) occurs as automorphic and hypautomorphic columns and grains, with minor octangular sections; grain sizes range from 0.2 to 2.2 mm. Some crystals wrap plagioclase, forming an ophitic texture. Minor crystals show serpentinization, with pseudomorphs of columnar crystals of primary minerals preserved. Also, minor nontransparent minerals are sporadically distributed and were allotriomorphic granular, with grain sizes of 0.1–1 mm (Fig. 3d–f). The Mayigangri gabbro (PM20) colored off-grayish green with a massive structure and a coarse grained texture. It mainly composed of plagioclase (~ 50 vol.%), clinopyroxene (~ 44 vol.%), amphibole, and biotite pseudoimage (~ 5 vol.%). The minor minerals include magnetite, chlorite, and epidote. Plagioclase is in the shape of hypidiomorphic plate or strip, which is distributed in the form of frame. Some plagioclases have been completely metasomatized and appear as pseudomorphs. Clinopyroxenes are subhedral columnar granular, filled between plagioclase crystals, and amphibole reaction edges can be seen locally (Gao et al. 2019).

Garnet–phengite–quartz schist shows a lepidogranoblastic texture and flaky structure, and it mainly comprises quartz (45 vol.%), phengite (35 vol.%), albite (10 vol.%), chlorite (5 vol.%), and garnet (5 vol.%). Structural fissures widely occurred and were filled with silicified quartz, minor sericite, and fragmented primary rocks as veins (Fig. 3g, h). In addition, overgrowth of garnet is common in garnet–phengite–quartz schist (Fig. 3i).

Methodology

In this study, one diabase sample was collected for U–Pb zircon dating and in-situ Hf isotope analyses, one garnet–phengite–quartz schist sample for U–Pb zircon dating, two groups of phengite samples of the schist for Ar–Ar dating, seven diabases and ten gabbros samples for whole-rock geochemical analysis, and four diabase samples for whole-rock Nd isotope analyses. Separation and selection of individual zircon grains were completed at the Laboratory of Beijing Zircon Chronology Navigator Technology Co., Ltd. Selected zircon grains and standard samples were imaged by transmission light, reflected light, and cathode luminescence microphotography for determination of zircon structure and genesis type. Appropriate testing positions away from cracks and inclusions were selected to avoid affecting data analysis quality. LA-ICP-MS U–Pb dating of zircon was performed using Thermo Element II at the Test Center of the Institute of Geology, Chinese Academy of Geological Science, using the instrument of the Finnigan Neptunetype MC-ICP-MS, and supporting Newwave UP 213 laser ablation system. Analytical spot sizes ~ 25 μm, with the frequency of 10 Hz, and the density of primary ion ~ 2.5 J/cm2. Zircon GJ1 (U, 923 ppm; Th, 439 ppm; and Th/U, 0.475) is used as the external standards in zircon U–Pb dating calibration. The zircon Plešovice and standard samples GJ1 were tested twice, respectively, and once for every 10 spots to ensure the stability of the instrument. The method proposed by Andersen (2002) was adopted for common Pb correction and the international standard program Isoplot (ver3.0) was employed for age calculation. For analytical points with ages obtained as < 1 Ga, 206Pb/238U and the correspondent 1σ were used as the age, and 207Pb/235U and 206Pb/238U ratios were used to calculate the concordant degree; for analytical points with ages obtained as > 10 Ga, 207Pb/206Pb and the correspondent 1σ were used as the age, and the ratios of 207Pb/206Pb and 206Pb/238U were used to calculate the concordant degree. In the concordant degree calculation, 0.93–1.0 were adopted as the concordant points.

40Ar/39Ar dating of phengite was performed at the Laboratory of Isotope Thermochronology, Institute of Geology, Chinese Academy of Geological Sciences. Sample heating by stage was completed using an electro-bombardment furnace. Mass spectrometry (MS) was conducted using a MM-1200B mass spectrometer. Quality discrimination correction, ammonia correction, blank correction, and interference element isotope correction were performed on all of the data. The sorted minerals (purity > 99%) were wrapped in aluminum foil, and placed in quartz vials and irradiated in the H4 position of the nuclear reactor at the China Institute of Atomic Energy, after being cleaned by ultrasonic waves. The standard biotite sample 2BH-25 is irradiated with integrated. The neutron flux was 2.60 × 1013 n cm−2 S−1 with the total irradiation time of 1440 min and an integrated neutron flux of 2.25 × 1018 n cm−2. The irradiated samples were step heated in a graphite furnace. Each sample was heated for 10 min. Before being analyzed in a Helix MC noble gas mass spectrometer, the released gas was purified for 30 min. 20 cycles were measured at each peak. They were corrected by mass discrimination, atmospheric Ar contamination, interfering nuclear reactions, and procedural blanks, after all the data returned to zero-time values. The coefficients of interfered nuclear reactions generated by K and Ca during the neutron irradiation process were obtained from irradiated K2SO4 and CaF2 samples with (36Ar/37Aro)Ca = 0.0002389, and (39Ar/37Aro)Ca = 0.000806, and (40Ar/39Ar)K = 0.004782. The 40 K decay constant (λ) used was 5.543 × 10–10 year−1, and 37Ar was corrected for radioactive decay. The plateau age, isochronous age, and inverse isochronous age were calculated by the ISOPLOT program (Ludwig 2003).

Whole-rock major, trace, and rare-earth element analyses were completed at the Lab Center of Hebei CREC Geophysical Exploration Co., Ltd., and major elements were analyzed with an ARL Advant XP + XRF instrument, with a precision of more than 3%. Trace elements were analyzed using X-series 2/SN01831C ICP-MS, with a precision of more than 3%. Analytical procedures were similar to those described by Liu et al. (2016). In the digestion of samples, 0.5 g sample was mixed with 3 mL HF, 9 mL HNO3, and 2 mL concentrated HCl at 108℃ for 24 h. The solution was evaporated to dryness and then was re-dissolved using 50 mL 2% (v/v) purified HNO3 for analysis. The reference materials, including GSS1, GSS2, GSS3, and GSS4a from the Center of National Standard Reference Material of China, were chosen to check the precision of the analytical methods.

Diabase Nd isotope analyses were performed at the Lab Center of Nanjing University. Isotope testing was conducted by MC-ICPMS in a static mode at the Lab. USGS standard materials BHVO-2, BCR-2, and AGV-2 were employed in the entire analytical process for quality monitoring. 146Nd/144Nd = 0.7219 was adopted for the internal correction of the equipment test according to index laws. Zircon in-situ Hf isotope analysis was performed by Laser Ablation MC-ICPMS at the Lab of Tianjin Geological Survey Center. The instrument system consisted of ESI NEW WAVE 193 nm FX laser and ThermoFisher NEPTUNE MC-ICPMS produced in the U.S.A. Performance parameters included a mass number range of 4–310 amu, resolution > 450 (flat peak, 10% peak valley definition), and abundance sensitivities < 5 ppm (without RPQ) and < 0.5 ppm (with RPQ). Measured peak stability included magnetic field and electric field drift (< 50 ppm/h), a laser pulse frequency of 8 Hz, and an ablation pore size 45 μm. Zircon GJ-1 was used as the standard for isotope monitoring, and zircon 91,500 as the correction standard.

Analytical results

Zircon U–Pb chronology

Zircon U–Pb dating is conducted on diabase and a garnet–phengite–quartz schist samples. A CL photo shows zircons from diabase occurred as plates and short columns, without oscillatory zonation (Fig. 4a, b); zircons from garnet–phengite–quartz schist have a complex shape, mainly ellipses and long and short columns. Some overgrowth zircons occurred in the schist samples and zircon punctuations are concentrated in the margins, but avoiding the accreting rims. Results of the analysis are shown in Table 1, and a zircon U–Pb concordia age diagram is shown in Fig. 4c, d.

Fig. 4
figure 4

Representative cathodoluminescence (CL) images of zircons from diabase sample A1638 in the northern margin of south Qiangtang. a Phengite–quartz schist sample 7.29RZ, and b U–Pb concordia diagrams of zircon data from samples A1638 (c) and 7.29RZ (d) from the Rongma area, Central Qiangtang

Full size image
Table 1 U–Pb isotope compositions of the zircons from diabase and garnet–phengite–quartz schist from Rongma town in the northern margin of south Qiangtang
Full size table

The diabase (A16138) has zircon Th/U ratios between 0.6 and 3.7, indicating magmatic origin. Concordia ages of 36 zircon grains were obtained, and the weighted mean age was 241.2 ± 1.1 Ma (n = 36, MSWD = 0.68), representing the formation age of the diabase and indicating that it was formed in the Early Triassic Time (Table 1, Fig. 4c).

Zircons from the garnet–phengite–quartz schist (7.29RZ) display a complex configuration and roundness, reflecting complex zircon sources. Zircon Th/U ratios varied greatly, mainly from 0.04 to 0.98. The 16 analyses of zircons have a concordant degree of > 92% and seven were < 90%. The concordant ages are concentrated between 393 and 2102 Ma, indicating the protolith of this sample deposited after 393 Ma. The sample also have Precambrian zircon age data, including ~ 700 Ma, ~ 1000 Ma, ~ 1400 Ma, ~ 1800 Ma, and 2100 Ma (Table 1, Fig. 4d).

Phengite 40Ar/39Ar chronology

40Ar/39Ar dating is conducted on two groups of phengite from the garnet–phengite–quartz schist. Stage heating dating analytical results of the phengite are shown in Table 2 and Fig. 5. Isochron diagrams of the two groups are shown in Fig. 5a, c, in which 12 stages of 790–1200 ℃ and 700–1400 ℃ constituted good age plateaus. tp1 = 229.0 ± 1.4 Ma and tp2 = 225.3 ± 1.3 Ma (Fig. 5b, d) correspond to 39Ar release amounts of 98.4% and 99.65%, respectively. The corresponding 36Ar/40Ar-39Ar/40Ar isochron ages are ti1 = 227.7 ± 3.8 Ma and ti1 = 225.2 ± 2.1 Ma. Initialization values of 40Ar/36Ar were 305 ± 16 (MSWD = 36) and 296.0 ± 2.6 (MSWD = 4.6), respectively. Thus, ~ 229 to ~ 225 Ma could represent the formation time of the phengite, and reflect one stage of regional metamorphism event.

Table 2 Result of 39Ar/40Ar stepwise heating dating for phengite of garnet–phengite–quartz schist from Rongma area in the northern margin of south Qiangtang
Full size table
Fig. 5
figure 5

a, c Isochronic age, and b, d 40Ar/39Ar plateau age diagrams of phengite from the garnet–phengite–quartz schist from the Rongma area in the northern margin of south Qiangtang

Full size image

Major/trace/rare-earth elements

Elemental geochemical analyses are conducted on the Early Triassic diabase and gabbro samples in the Rongma area, and the data are shown in Table 3. The gabbro formed in the Early Triassic (~ 237 Ma, Gao et al. 2019). Seven diabase samples have SiO2 of 47.82–49.60 wt.%, Al2O3 of 13.42–14.64 wt.%, TiO2 of 1.55–1.94 wt.%, MgO of 8.10–8.76 wt.%, MgO/FeOt of 1.23–1.42, and LOI of 3.10–6.27 wt.%. Ten gabbro samples show SiO2 of 46.12–48.38 wt.%, Al2O3 of 10.24–15.14 wt.%, TiO2 of 1.31–2.47 wt.%, MgO of 4.52–12.60 wt.%, MgO/FeOt of 0.50–1.70, and LOI of 3.09 ~ 8.35 wt.%. Given the relatively high LOI, some relatively stable elements, e.g., high field-strength elements (HFSEs) were used to determine petrogenesis. On the whole-rock SiO2-Nb/Y and Zr/TiO2-Nb/Y diagrams (Winchester and Floyd 1977; Zhou et al. 2009), all the diabase and most gabbro samples fall in the sub-alkaline basalt field (Fig. 6a, b). On the trace element and rare-earth element (REE)-normalized diagrams (Fig. 7a, b), the seven diabase and ten gabbro samples overall showed geochemical features similar to continental arc basalt despite with high LOI values (Table 3). Diabase samples mainly show weak negative Nb, Ta, and Th anomalies and positive Pb anomaly with total rare-earth elements concentrations (ΣREEs) of 79–89 ppm and chondrite-normalized REE patterns imply they have a typical right-inclined tendency with (La/Yb)N = 5.65–6.14 and with weak positive Eu anomalies (δEu = 1.02–1.20) (Table 3, Fig. 7a, b). However, gabbros are weak depletion in Nb, Ta, and Th, enrichment in Pb, and have variable ΣREEs of 56–119 ppm. They have a typical right-inclined tendency with (La/Yb)N = 4.74–5.65 and with weak negative to weak positive Eu anomalies (δEu = 0.76–1.12) in chondrite-normalized REE patterns (Table 3, Fig. 7a, b). In addition, both diabases and gabbros mainly display signatures of a calc-alkali-potassium basaltic series (Fig. 8a, b).

Table 3 Major elements (wt.%), trace elements (ppm), and REE (ppm) concentrations of the Early Triassic diabase and gabbro from Rongma area in the northern margin of south Qiangtang, and average trace elements (ppm) and REE (ppm) concentrations of OIB, E-MORB, N-MORB, and continental arc basalt (CAB)
Full size table
Fig. 6
figure 6

Geochemical discrimination of the Early Triassic diabases and gabbros from the Rongma area in the northern margin of south Qiangtang. a Diagram of SiO2 vs. Nb/Y (after Le Bas 2000); b diagram of Zr/TiO2 (× 10−4) vs. Nb/Y (after Winchester and Floyd 1977)

Full size image
Fig. 7
figure 7

a N-MORB-normalized trace element diagrams for the Early Triassic diabase and gabbro from the Rongma area in the northern margin of south Qiangtang. b chondrite-normalized REE diagram for the same mafic rocks as in a. The values of primitive mantle, OIB, N-MORB, and E-MORB are from Sun and McDonough (1989). The values of CAB are from Kelemen et al. (2007). The data of the Early Triassic diabase and gabbro in the Rongma area are from this study. OIB = island arc basalt, N-MORB = Normal mid ocean ridge basalt, E-MORB = Enriched mid ocean ridge basalt, CAB = continental arc basalt

Full size image
Fig. 8
figure 8

Geochemical discrimination diagrams of magma series of diabase and gabbro from Rongma area in the northern margin of south Qiangtang. a Diagram of Th/Yb vs. Ta/Yb; b diagram of Ce/Yb vs. Ta/Yb. The symbols are the same as those in Fig. 6a

Full size image

Whole-rock Nd and Hf-in-zircon isotopes

Whole-rock Nd and zircon Hf isotopes analyses are conducted on the Early Triassic diabase. Analytical results are shown in Tables 4 and 5. The rock have a 143Nd/144Nd ratio between 0.512504 and 0.512514, and εNd(t) of − 0.8 ~ 0.9. Zircon 176Hf/177Hf ratio ranges widely from 0.282629 to 0.283184, with εHf (t) of − 0.6 ~ 19.1, and TDM of 498 ~ 991 M (removed two bad data t ˃ TDM of 105 Ma and 131 Ma) (Tables 4, 5; Fig. 9a, b).

Table 4 Whole-Rock Nd isotopic compositions of the Early Triassic diabase from Rongma area in the northern margin of south Qiangtang
Full size table
Table 5 Zircon Hf isotopic compositions of the Early Triassic diabase from Rongma area in the northern margin of south Qiangtang
Full size table
Fig. 9
figure 9

a whole-rock εNd(t) vs. crystallization age of the early Triassic diabase from Rongma area in the northern margin of South Qiangtang; b zircon εHf(t) vs. zircon crystallization age of the early Triassic diabase in (a)

Full size image

Mineral chemistry and thermobarometric evaluation

Based on petrographic observations for the garnet–phengite–quartz schist from Rongma area, three stages of mineral assemblages have been recognized. Peak metamorphic stage mineral assemblages are garnet + phengite + quartz. The early retrograde metamorphic mineral assemblages are composed of garnet, phengite, biotite, chlorite and quartz, and the late retrograde metamorphic mineral assemblages consist of garnet, phengite, biotite, chlorite, albite, and quartz. On the basis of detailed petrographic and mineralogical observation, representative samples were selected for further phase equilibrium model (Fig. 10). It mainly include garnet, phengite, quartz, albite and minor sphene, rutile, albite, and chlorite. The phase equilibrium model was performed using Domino/theriak software (de Capitani and Petrakakis 2010) through the following mineral solid solution models: epidote (Holland and Powell 1998), plagioclase (Baldwin et al. 2005; Holland and Powell 2003), garnet (White et al. 2005), omphacite (Green et al. 2007), chlorite (Holland and Powell 1998), muscovite (Coggon and Holland 2002), biotite (White et al. 2007), talc (Holland and Powell 1998), chlorite (White et al. 2000), magnetite (White et al. 2007), and hornblende (Diener et al. 2007). Other minerals, such as quartz/coesite, kyanite, water, and andalusite, are end member minerals.

Fig. 10
figure 10

Chemical composition of garnet of phengite–quartz schist from Rongma area in the northern margin of south Qiangtang. a cross-polarized light microscopic photo of phengite–quartz schist for microprobe analysis; b plane-polarized light microscopic photo of phengite–quartz schist for microprobe analysis; c BSE microscopic photo of garnet for microprobe analysis; d compositional profile of garnet. Xalm = Fe2+/(Fe2+  + Mn + Mg + Ca), Xpy = Mg/(Fe2+  + Mn + Mg + Ca), Xspss = Mn/(Fe2+  + Mn + Mg + Ca), Xgr = Ca/(Fe2+ + Mn + Mg + Ca). Ab albite, Ms Phengite, Q quartz, Grt garnet, alm almandine, gr grossular, py pyrope, spss spessartine

Full size image

Since garnet in the sample has obviously experienced two stages of growth (Fig. 10a–d), and the core and overgrowth rim record distinct compositions of grossular, pyrope, spessartine, and almandine (Fig. 10d). The pressure and temperature conditions (P–T) of garnet core formation is simulated according to whole-rock geochemical data (Table 6). Electron probe data (Table 7) and the effective whole-rock composition obtained from mineral volume (Carson et al. 1999) were used to simulate phase equilibria for the P–T conditions of garnet overgrowth rim. Considering that TiO2 and P2O5 only exist in accessory minerals and have little content in main silicate minerals, their influence on phase diagram is ignored. MnO is mainly concentrated in garnet core, which has great influence on garnet stability (Wei et al. 2004). Fe2O3 has great influence on the stability domain of amphibole (Du et al. 2014). Therefore, the model system of MnO-Na2O-CaO-K2O-FeOt-MgO-Al2O3-SiO2-H2O (MNCKFMASHO) is used to simulate the metamorphic process of rocks, and the excess of Q and H2O is assumed.

Table 6 Measured and effective whole-rock compositions of the garnet–phengite–quartz schist from Rongma area in the northern margin of south Qiangtang
Full size table
Table 7 Garnet chemical composition of the garnet–phengite–quartz schist from Rongma area in the northern margin of south Qiangtang
Full size table

Measured P–T profile (Fig. 11) is calculated by whole-rock composition of garnet phengite schist in MNCKFMASHO system. The phase diagram shows that garnet is stable in most regions and only disappears at 350–460 ℃/14–21 kbar and 350–440 ℃/5–13 kbar. Biotite is stable in the range of 440–650 ℃/5–14 kbar. Lawsonite is stable in the range of 350–540 ℃/9–30 kbar. Epidote is stable in the range of 350–500 ℃/5–15 kbar. Chlorite is stable in the range of 350–520 ℃/5–17 kbar. The isopleth of grossular, pyrope, and spessartine contents in garnet are also calculated. In most regions, the grossular isopleth has relatively steep slopes, and grossular content decreases gradually with the increase of temperature. The content of pyrope increases gradually with the increase of temperature and their isopleth have variable slopes in most regions. Spessartine isopleth have relatively steep slopes and their contents decrease with the increase of temperature. Because of the low content of pyrope in garnet core, there is a big error to define the P–T conditions of garnet core using the pyrope isopleth. Therefore, here we use the isopleth of spessartine and grossular content to define the P–T conditions of garnet core. The plot of chemical composition of garnet core indicates a P–T range of 25–27 kbar and 430–450 ℃ for the peak metamorphic stage (Fig. 11a).

Fig. 11
figure 11

Measured (a) and effective (b) P–T profile of garnet–phengite–quartz schist under MnNcKFMASHO system in Rongma area, in the northern margin of south Qiangtang. The apparent profile is calculated based on the measured (a) and effective (b) whole-rock composition in Table 6. In addition, the isopleth of spessartine content (e.g., blue dotted line s4), grossular content isopleth (e.g., green dotted line g28), and pyrope content isopleth (e.g., cyan dotted line p4) in garnet are also calculated. The green gradient shadow area represents the projection area of garnet core component, and the blue gradient shadow area represents the projection area of garnet edge component. The black thick solid line represents the metamorphic P–T path from the core to the edge of garnet, and the black thick dotted line represents the inferred partial late retrograde metamorphism path. Peak metamorphic stage mineral assemblages are garnet + phengite + quartz. The early retrograde metamorphic mineral assemblages are garnet + phengite + biotite + chlorite + quartz, and the late retrograde metamorphic mineral assemblages are garnet + phengite + biotite + chlorite + albite + quartz. (c–d) carton model of metamorphic process of garnet. It shows that garnet-phengite-quartz schist would be re-involved into subduction when its buoyant force was lower than the resultant of gravity and slab-pull forces. alm almandine, ep epidote, gr grossular, law andalusite, py pyrope, spss spessartine, jd jadeite, ta talc, coe coesite, q quartz, hem hematite, chl chlorite, g garnet, ab albite, bi biotite, ph phengite, pl plagioclase

Full size image

P–T profile calculated using effective whole-rock composition by EPMA analysis and mineral volume content in MnNcKFMASHO system is shown in Fig. 11b. The phase diagram shows that garnet is stable in most areas and only disappears only at 350–460 ℃/5–22 kbar. Biotite is stable in the range of 370–650 ℃/5–16 kbar. Epidote is stable in the range of 350–540 ℃/5–15 kbar. Andalusite is stable in the range of 350–630 ℃/14–30 kbar. Chlorite is stable in the range of 350–570 ℃/5–22 kbar. Albite is stable at 460–650 ℃/5–14 kbar. In addition, the isopleth of the contents of pyrope and grossular in garnet are also calculated. The phase diagram shows that isopleth of the pyrope contents has steep negative slopes, and its contents increase gradually with the increase of temperature. In the stable region of lawsonite, isopleth of the grossular contents have positive slopes and its content decreases gradually with the increase of pressure. In the stable region of epidote, isopleth of the grossular contents has steep negative slopes and its content gradually decreases with the increase of pressure. We can see that the isopleth of grossular and pyrope contents is only affected by pressure and temperature in the stable region of lawsonite, which can be used to define the formation conditions of garnet edge. The plot of chemical composition of garnet overgrowth rim indicates a P–T range of 17–19 kbar and 470–500 ℃ for the early retrograde metamorphic stage (Fig. 11b). However, the late retrograde metamorphic stage of T–P condition was not obtained.

Discussion

Metamorphic evolution of garnet–phengite–quartz schist

Based on petrographic observations, protoliths of garnet–phengite-quartz schist experienced at least three stages of metamorphic evolution (i.e., Peak metamorphic, early retrograde metamorphic, and late retrograde metamorphic stages). Model results indicate garnet core of garnet–phengite–quartz schist might form at depths of ~ 74–80 km (pressure gradient taking 0.03 GPa/km), whereas its overgrowth rim might experience burial to depths of 50–56 km. It also implies that the metamorphic evolution from core to overgrowth rim of garnet experienced a process of increasing temperature but decompression (Fig. 11b). Although the late retrograde metamorphic stage of T–P condition was not obtained, we infer that garnet–phengite–quartz schist exhumed to the upper crust or the surface with late retrograde metamorphism with cooling and decompression. The evolution path of increasing temperature but decompression from garnet core to overgrowth rim might respond to four complicated processes: (i) subduction—exhumation along subduction channel—involving subduction again—exhumation again; (ii) subduction—exhumation with magma emplacement and overgrowth of garnet core; (iii) crustal thickening—extensional exhumation with magma emplacement and overgrowth of the core, or (iv) subduction1exhumation—late hydrothermal alteration with overgrowth of garnet core. The garnet core domain shows a metamorphic characteristic with enrichment in spessartine and depletion in almandine which is distinct with magmatic garnet (duBray 1988; Dahlquist et al. 2007); thus, we exclude the two hypotheses of “ii” and “iii.” Hydrothermal garnets would display obvious oscillating zonation (Clechenko and Valley 2003; Dziggel et al. 2009), while the garnet in this study does not have such characteristics of hydrothermal alteration. Hence, the hypothesis “iv” also be precluded and we infer that the evolution of garnet experienced a complicated process corresponding to hypothesis “i.”

Obviously, the formation of garnet core should represent an earlier metamorphic event than its overgrowth rim. On the basis of the published data of the metamorphic rocks in the CQMB, different stages of metamorphism, such as eclogitic metamorphism subduction related of 230–244 Ma (Liang et al. 2017), were reported. Thus, we infer that the formation of garnet core might reflect early peak metamorphism related to deep subduction (~ 74–85 km), but growth rim of garnet might respond to early retrograde metamorphs related to re-involvement of subduction (50–56 km). As phengite has a closure temperature of ~ 350 ℃ and is a product of low-temperature, high-pressure metamorphism, it is usually formed in collisional orogenic belts (Jäger 1979). Consequently, we infer that ~ 227–225 Ma phengite age might reflect the late retrograde metamorphism related to exhumation after collision of Southern and Northern Qiangtang blocks. Then, how did the garnet–phengite–quartz schist exhume during the complicated subduction process? Due to garnet–phengite–quartz schist having lower density than eclogite, density difference between garnet–phengite–schist and meta-mafic rocks will increase with the increasing of subduction depth. This would result in the exhumation of schists along high-temperature low-pressure area (low-stress zone) in subduction channel (Wei et al. 2009). However, the meta-mafic rock blocks with higher density might continue to subduct into deeper mantle . In addition, compressive stress maintains in subduction channel during exhumation of garnet–phengite–quartz schist; thus, it will hinder the continuing exhumation of the schist. Hence, we infer that garnet–phengite–quartz schist would be re-involved into subduction when its buoyant force was lower than the resultant of gravity and slab-pull forces (Fig. 11c, d).

Formation and metamorphism time of metamorphic complexes

The U–Pb zircon dating results showed that the concordant ages of the zircon from the schist ranged from 393 to 2102 Ma, implying that the protolith of garnet–phengite–quartz schist deposited after 393 Ma. Metamorphic accreting rim occurred in the zircons of the schist, indicating that the protolith suffered from metamorphism, which surely occurred after 393 Ma. Abundant Precambrian ages, including ~ 700 Ma, ~ 1000 Ma, ~ 1400 Ma, ~ 1800 Ma, and 2100 Ma, can be observed in the zircons of the schist (Table 1, Fig. 3d), suggesting Precambrian basements might exist in the region.

Two groups of phengite from the garnet–phengite–quartz schist yielded 40Ar/39Ar plateau ages of ~ 229 Ma and ~ 225 Ma, respectively (Fig. 5b, d), representing that the garnet–phengite–quartz schist was formed at ~ 229–225 Ma. As stated above, phengite is usually formed in collisional orogenic belts (Jäger 1979). Combined with regional dynamic evolution data (Table 8), the phengite 40Ar/39Ar plateau age of ~ 229–225 Ma obtained in this study most likely represented the time of regional exhumation after the collision of the Southern and Northern Qiangtang blocks.

Table 8 The chronological data for the metamorphic complexes in the CQMB in the Central Qiangtang
Full size table

Petrogenesis of mafic rocks

Given the relatively high LOI of early Triassic diabase and gabbro samples, some relatively stable elements, e.g., high field-strength elements (HFSEs) and SiO2, are used to determine their petrogenesis. As shown in Fig. 12a, Nb/U ratios of mafic rocks do not show significant variation with LOI increase. On the whole-rock Zr/Y–Zr and Zr/TiO2–Nb/Y diagrams (Fig. 6a, b), both seven diabase and eight gabbro samples fall into the sub-alkaline basalt field. Both the diabase and gabbro samples mainly show weak negative Nb, Ta, and Th, positive Pb anomalies, and low Nb/U (~ 13–24) and Ce/Pb (~ 3–11) ratios, with a right-inclined chondrite-normalized REE pattern, similar to average compositions of continental arc basalt (Kelemen et al. 2007) but distinct with OIB and MORB (Table 3, Fig. 7a, b). The Triassic diabase samples have εNd(t) of − 0.8 ~ 0.9, and zircon εHf(t) values of − 0.6 ~ 19.1. Two types of potential petrogenesis can be used to interpret these geochemical signatures: (i) the primary magma of diabase possibly is sourced from depleted mantle but suffered from contamination of minor crustal material, or (ii) the primary magma of diabase might be mainly sourced from enriched lithospheric mantle. Crustal contamination will increase abundance of SiO2, Th/Nb, and lower Nb/U and εNd(t) (Cheng et al. 2018). However, on the Nb/U- SiO2, εNd(t)- SiO2 and εNd(t)- Nb/U diagrams (Fig. 12b-d), they showed opposite correlations, which indicates significant crustal contamination did not occur during formation of the diabase, and the primary magma of diabase might have formed by partial melting of enriched lithospheric mantle. In addition, similar whole-rock geochemical compositions between the diabase and gabbro samples indicate that they might have same petrogenesis related to the subduction of the Shuanghu Paleo-Tethys Ocean.

Fig. 12
figure 12

Geochemical diagrams of the Early Triassic diabase and gabbro from Rongma area in the northern margin of south Qiangtang. a zircon εHf(t) vs. crystallization age (t); b diagram of whole-rock εNd(t) vs. Ba/Th; c whole-rock εNd(t) vs. K2O; d Ba/Th vs. Nb/U. Values of continental crust, BAB, OIB, N-MORB, and E-MORB are from Sun and McDonough (1989) and Gale et al. (2013)

Full size image

On the geochemical diagrams of the tectonic setting (Fig. 13a–f), both the early Triassic diabase and gabbro samples mainly fall into the intraplate extensional environment and display transition from OIB to MORB and arc basalt signatures; both the diabase and gabbro samples show weak depletion in HFSEs (e.g., Nb, Ta, and Th), suggesting that their primary magmas did not suffer from intense fluid metasomatism and crustal contamination. Thus, the diabase may have formed in a continental rift or back-arc extensional environment. Based on previous researches, regional metamorphism related to continent–continent collision occurred in the middle-late Triassic (227–209 Ma) (Zhai et al. 2009; Zhang et al. 2010; Zhu et al. 2010), with post-collision during ~ 225–200 Ma intervals (Wu et al. 2016; Xu et al. 2020; Zhu et al. 2010), suggesting that continental rift extension might occur after 200 Ma. In addition, ~ 244 Ma (Lu–Hf isochron age, Pullen et al. 2008) or ~ 237 Ma (U–Pb zircon age, Zhai et al. 2011b) eclogites related to northward subduction of the Shuanghu Paleo-Tethys Ocean were reported in the CQMB, which indicates a continental arc environment existed in the region during the early Triassic. Thus, a continental rift setting can be excluded in the northern margin of the Southern Qiangtang during the early Triassic. Consequently, combined with the whole-rock geochemical characteristics of the Early Triassic diabase and gabbro, we infer that they might form in a back-arc extensional setting, responding to the model of Zheng (2019).

Fig. 13
figure 13

Discrimination diagrams of the tectonic settings of the Early Triassic diabases and gabbros from the Rongma area, northern margin of south Qiangtang: a diagram of whole-rock TFeO/MgO vs. TiO2; b diagram of whole-rock TiO2 vs. Zr; c diagram of whole-rock Zr/Y vs. Zr (after Pearce and Norry 1979); d diagram of whole-rock Ti vs. Zr (after Pearce and Cann 1973); e diagram of whole-rock Th/Hf vs. Ta/Hf diagram (Zhang et al. 2004a, b, c), and f diagram of whole-rock Ti/100 vs. Zr vs. Y*3 (after Pearce and Cann 1973). The symbols are the same as those in Fig. 5a

Full size image

Tectonic implications

The early Paleozoic ophiolitic mélange (e.g., Middle Ordovician ~ 467 Ma, Li et al. 2008a; Late Cambrian, ~ 517 Ma, Wu, 2013), including cumulate gabbro, plagiogranite, and basalt, was recognized along the CQMB and Geochemical data implies that they are characterized by N-MORB (Zhai et al. 2011a). Furthermore, Early Permian ophiolite (Li et al. 2008a, b) was also found in the LongmuTco–Shuanghu Suture Zone. In addition, the Late Carboniferous-Early Permian adakitic rocks and andesites related to subduction of Paleo-Tethys Oceanic were reported in the CQMB (Fig. 13a; Jiang et al. 2014; Zhang et al. 2014a, b). Therefore, the Paleo-Tethys Ocean might have been open since the Early Paleozoic and remained open in the Early Permian.

Many authors reported the Permo-Triassic arc-related igneous activity mainly in the Northern Qiangtang; thus, they interpreted these igneous rocks to be related to northward subduction of the Shuanghu Paleo-Tethys Ocean underneath the Northern Qiangtang (e.g., Zhang et al. 2006a, b; Zhai et al. 2011a, b; Li et al. 2019). However, some metamorphic rocks related to subduction of ocean plate were also recognized in the Southern Qiangtang (e.g., Liu et al. 2011; Zhu et al. 2015), favoring that view southward subduction of the Ocean underneath the Southern Qiangtang also occurring during the Early Triassic. In addition, recently Dan et al. (2020, 2021) reported abundant the Early Permian (~ 290–285 Ma)—Early Triassic (~ 239 Ma) mafic intrusions in the Southern Qiangtang block, and they proposed that the Early Permian mafic magmas were generated by a mantle plume, whereas the Early Triassic mafic rocks are derived from passive-margin magmatism caused by enhanced oceanic slab-pull forces. However, these mafic rocks show the geochemical features similar to continental arc basalt as the early Triassic mafic rocks in this study, such as low Nb/U (average on ~ 18) and Ce/Pb (average on ~ 9.6), weak–strong depletion in Nb and Ta, and enrichment in Pb (Dan et al. 2020, 2021; Zheng 2019); hence, we infer these mafic rocks might form in a back-arc extensional setting caused by the subduction of the oceanic plate between the Northern and Southern Qiangtang blocks beneath the latter. Combined with the result from the metamorphic evolution of overgrowth zoning garnet from the schist in this study, we favor that view bi-directional subduction of the Shuanghu Paleo-Tethys Ocean in the Early Triassic (Fig. 14a).

Fig. 14
figure 14

Tectonic model of the Paleo-Tethys ocean in central Qiangtang during the Triassic Time. Chronological data of the metamorphic complex are shown in Table 8

Full size image

In addition, our Ar–Ar ages of phengites from the schist indicate the collision time of the two blocks might be earlier (> ~ 229 Ma). Combining with previous geochemical data, we propose that the Paleo-Tethys Ocean was finally closed at ~ 237– ~ 229 Ma, which is also consistent with some previous studies (e.g., Li et al. 2015a, b; Zhang and Tang 2009; Zhang et al. 2018). For instance, on the basis of previous Ar–Ar dating results of phengite from schists (Table 8), the collision of the southern and northern blocks with intense regional metamorphism had occurred before ~ 209 Ma, even tracing back to ~ 227 Ma. Li et al. (2015a, b) discovered simultaneous S-type granite in the Rongma area (222–210 Ma) (Fig. 14b), which also supported that the south and north Qiangtang blocks assembled in the late Triassic. In addition, Zhang et al. (2018) reported the Triassic Baqing eclogites across the CQMB in the north Qiangtang block, favored that the collision between the south and north Qiangtang blocks had occurred at ~ 223 Ma. Also, significant denudation occurred across the CQMB in temporal interval of 222–204 Ma (Zhang and Tang, 2009 and references therein). Thus, in this study we favor that the Paleo-Tethys Ocean had been open since the Early Paleozoic and remained open until the Early Triassic and was finally closed at ~ 237– ~ 229 Ma. This time model will help us better understand the tectonic evolution of the Paleo-Tethys Ocean.

Conclusions

(1) One diabase from the Rongma area yielded a zircon U–Pb age of 241.2 ± 1.1 Ma, one phengite–quartz schist yielded zircon-marginal concordant ages ranging from 1936 to 393 Ma, and two groups of phengite yielded Ar–Ar plateau ages of 229.0±1.4 Ma and 225.3±1.3 Ma, respectively. These ages indicate that the diabase was formed in the early Triassic, that sedimentation of the protolith of the garnet–phengite–quartz schist occurred after ~393 Ma, and that phengite formed in the Middle Triassic (~229 to 225 Ma).

(2) Whole-rock geochemistry of the early Triassic diabase and gabbro and whole-rock Nd and zircon-Hf isotopes of the diabase suggest that both the two early Triassic mafic rocks in the Rongma area are formed in a back-arc extensional setting.

(3) Chronological and geochemical evidence of the garnet–phengite–quartz schist and the early Triassic diabase and gabbro indicate that the Paleo-Tethys Ocean might be finally closed at ~237 to 229 Ma, and that the time of metamorphism of the phengite–quartz schist might respond to the exhumation of post-collision of the Southern and Northern Qiangtang blocks.

(4) Overgrowth zoning garnet in garnet–phengite–quartz schist with three stages of metamorphic evolution is recognized from garnet core to overgrowth rim, indicating two subduction in a short time might be involved its metamorphic evolution.