SHRIMP U–Pb Zircon Ages, Geochemistry and Sr–Nd–Hf Isotope Systematics of the Zalute Intrusive Suite in the Southern Great Xing’an Range, NE China: Petrogenesis and Geodynamical Implications

An integrated zircon geochronological, elemental geochemical, and Sr–Nd–Hf isotopic investigation was carried out on a suite of dioritic–granitic rocks at Zalute in the southern Great Xing’an Range (SGXR), NE China, in order to probe the source and petrogenesis of these granitoid rocks and further constrain the geodynamical setting of early Early Cretaceous magmatism. The results of Sensitive High-Resolution Ion Micro Probe (SHRIMP) zircon U–Pb dating reveal that the Zalute dioritic–granitic rocks have a consistent crystallization age of ca. 137–136 Ma, consisting of quartz diorite (136 ± 1.4 Ma), monzogranite (136 ± 0.8 Ma), and granite porphyry (137 ± 1.3 Ma), which record an early Early Cretaceous magmatic intrusion. Geochemically, the quartz diorites, monzogranites, and granite porphyries are mostly high-K calc-alkaline and show features of typical I-type affinity. They possess uniform and depleted Sr–Nd–Hf isotopic compositions (e.g., initial 87Sr/86Sr ratios of 0.7035 to 0.7049, εNd(t) of −0.02 to +2.61, and εHf(t) of +6.8 to +9.6), reflecting a common source, whose parental magma is best explained as resulting from the partial melting of juvenile source rocks in the lower crust produced by underplating of mantle-derived mafic magma, with minor involvement of ancient crustal components. Evidence from their close spatio–temporal relationship, common source, and the compositional trend is consistent with a magmatic differentiation model of the intermediate-felsic intrusive suite, with continued fractional crystallization from quartz diorites, towards monzogranites, then to granite porphyries. Combined with previously published data in the SGXR, our new results indicate that the Zalute intermediate-felsic intrusive suite was formed during the post-collisional extension related to the closure of the Mongol–Okhotsk Ocean and subsequent slab break-off.


Introduction
Granitoid rocks are widely considered to be one of the most important components of the continental crust, which are mainly generated by a massive transfer of heat and/or mantle materials to the crust in the various geodynamical settings [1][2][3][4][5]. An understanding of petrogenesis and geodynamical setting of granitoids are important for the interpretation of the crustal evolution and crust-mantle interaction [6][7][8][9][10][11][12].
The southern Great Xing'an Range (SGXR), situated in the southeastern Central Asian Orogenic Belt, is well-known for Late Mesozoic granitoid rocks [4,5,13]. There have been important advances in the research of Late Mesozoic granitoid magmatism in recent years. It is widely accepted that these granitoids were mainly formed in the Early Cretaceous, with subordinates in the Late Jurassic [14][15][16], which were mainly derived from partial melting of juvenile materials with near-zero to positive Hf-Nd isotopic characteristics [17][18][19] and mantle-like δ 18 O compositions [16]. The large-scale Early Cretaceous granitoid magmatism is thought to be linked with the intracontinental extension [9,14]. Nevertheless, the magma genesis and geodynamic process of the Early Cretaceous giant granitoid-forming event are still not fully understood.
Several models have been proposed to decipher the origin and geodynamics of the Early Cretaceous granitoids in the SGXR, including (1) the "extreme fractional crystallization model of mantle-derived magma" which emphasizes evolved mantle-derived basaltic magma associated with the continental rifting [4]; (2) the "slab window model" advocates partial melting of subducting oceanic crust near the Mongol-Okhotsk ridge, accompanied with extension in the overlying lithosphere [20]; (3) the "mixing model" stresses on crust-mantle magma mixing, which is related to a post-collisional extensional setting [13,21,22]; (4) the "delamination model" proposes the gravitational collapse of thickened continental crust related to closure of the Mongol-Okhotsk Ocean [5,11,23,24] or subduction of the Paleo-Pacific Plate [25,26]; and (5) the "underplating model" argues for partial melting of underplated lower crust related to closure of the Mongol-Okhotsk Ocean [5,16] or superimposed by a far-field effect of western subduction of the Paleo-Pacific plate [14,15].
When compared to the northern part of the Great Xing'an Range, relatively few comprehensive studies have concentrated on the Early Cretaceous granitoids of the SGXR [13,27], which has hindered our understanding of Early Cretaceous tectonic-magmatism evolution in the SGXR [1].
In this work, we report new Sensitive High-Resolution Ion Micro Probe (SHRIMP) U-Pb zircon ages, major and trace elements, zircon Hf isotopic and whole-rock Sr-Nd isotopic data for quartz diorite, monzogranite, and granite porphyry at Zalute in the SGXR. The purpose of this study is to (1) constrain the magma source; (2) shed light on the petrogenesis of the dioritic-granitic rocks; and (3) provide insights into the tectonic setting that controlled the formation of the Early Cretaceous intermediate-felsic intrusive suite at Zalute.

Regional Geology
The Great Xing'an Range (GXR), located in the eastern part of Central Asian Orogenic Belt (Figure 1a), is sandwiched between the Siberia Craton to the north and North China Craton to the south [4,[28][29][30][31][32][33][34]. It experienced a remarkable geological amalgamation as a result of continental collision events during the Phanerozoic [34], which were related to the subduction of the Paleo-Asian Ocean, Mongol-Okhotsk Ocean, and Paleo-Pacific Ocean during Paleozoic [35,36]. The Mongol-Okhotsk Suture to the northwest of the GXR (Figure 1b) is widely accepted as representing the final closure of the Paleo-Asian Ocean [15,[37][38][39]. The subduction of the Paleo-Pacific oceanic plate is suggested to start in the Early Jurassic [40].
The SGXR lies at the southern segment of the GXR, bound to the north by the Erlian-Hegenshan fault, to the south by the Xar Moron fault, and to the northeast by the Nenjiang fault ( Figure 1c) [36,41,42]. The Xilinhot Complex is mainly composed of Paleozoic mica schist, paragneiss, and biotite-bearing granitic gneiss, representing the oldest strata exposed in the SGXR [43]. Late Paleozoic to Mesozoic magmatic rocks widely intruded into the Permian metamorphic volcano-sedimentary associations,  [28]) showing the location of NE China; (b) Simplified geotectonic division of NE China (after Zhang et al. [37]), showing the location of the SGXR; and (c) Geologic sketch map of the southern Great Xing'an Range (SGXR) (modified after Ouyang et al. [14] and Guo et al. [41]).
The Zalute area is situated in the east part of the SGXR (Figure 1c). The lower Permian strata in the region is a volcanic rocks suite of neritic facies, interlayered with clastic rocks, mainly including meta-andesite, meta-dacitic tuff, and meta-sandstone. The Upper Permian strata are dominated by a suite of clastic rocks of lacustrine facies, consisting of meta-siltstone, mudstone, felsic slate, and marlstone [27]. The Jurassic strata are mainly felsic volcanic rocks and clastic rocks of lacustrine facies, comprising andesite, volcanic breccia, volcanic tuff, conglomerate, sandstone, and shale. The  [28]) showing the location of NE China; (b) Simplified geotectonic division of NE China (after Zhang et al. [37]), showing the location of the SGXR; and (c) Geologic sketch map of the southern Great Xing'an Range (SGXR) (modified after Ouyang et al. [14] and Guo et al. [41]).
The Zalute area is situated in the east part of the SGXR (Figure 1c). The lower Permian strata in the region is a volcanic rocks suite of neritic facies, interlayered with clastic rocks, mainly including meta-andesite, meta-dacitic tuff, and meta-sandstone. The Upper Permian strata are dominated Minerals 2020, 10, 927 4 of 22 by a suite of clastic rocks of lacustrine facies, consisting of meta-siltstone, mudstone, felsic slate, and marlstone [27]. The Jurassic strata are mainly felsic volcanic rocks and clastic rocks of lacustrine facies, comprising andesite, volcanic breccia, volcanic tuff, conglomerate, sandstone, and shale. The Cretaceous strata consist of andesitic breccia, pyroclastic rock, tuff, limestone, and sandstone. The volcanic rocks exposed in this region involve the Late Jurassic Manketou'ebo Formation, and the Early Cretaceous Baiyingaolao, Manitu, and Meiletu Formations ( Figure 2) [24]. The Late Jurassic Manketou'ebo Formation is characterized by rhyolite and tuff, whereas the Early Cretaceous Baiyingaolao, Manitu, and Meiletu Formations are comprised of andesite, basaltic andesite, dacite, rhyolite, tuff, and pyroclastic rock [24,44,47]. The Ealy Permian, Early Triassic, Late Jurassic, and Early Cretaceous intrusive rocks exposed in this region ( Figure 2) are mainly composed of quartz diorite, quartz monzonite, monzogranite, granite porphyry, alkaline granite, and albite granite. These Early Permian to Early Cretaceous granitoid rocks commonly intruded in the form of batholiths, stocks, and dikes [18]. Major faults in the region are dominated by NW-and NE-striking faults, with subordinate E-W-striking faults.

Sample Descriptions
The granitoid samples collected from Zalute area include quartz diorite (Figure 3a Supplementary Table S1. Quartz diorites are characterized by a fine-grained, holocrystalline texture (Figure 3a), which mainly consists of plagioclase, quartz, biotite, and amphibole ( Figure 3b) with accessory zircon, apatite, and titanite. Occasionally, amphibole and biotite grains occur as inclusions in plagioclase.
Plagioclase grains exhibit euhedral to subhedral with sizes spanning from 0.2 to 1.0 mm. Quartz grains mainly show anhedral with sizes varying from 0.1 to 0.4 mm. Biotite grains occur as euhedral crystals up to 0.8 mm long. Amphibole grains are euhedral to subhedral with sizes in the range of 0.1 to 0.3 mm.
Granite porphyries are typically porphyritic with phenocrysts (up to 2 mm) of plagioclase, quartz, K-feldspar, and biotite (Figure 3e,f). Plagioclase and biotite phenocrysts are mainly euhedral or subhedral. K-feldspar phenocrysts occur as subhedral grains. Quartz phenocrysts occur as subhedral to anhedral grains in shape with rounded edges.

SHRIMP U-Pb Zircon Dating
Zircon grains were separated using conventional heavy liquids, hand-picking, and then mounted on epoxy resin and polished to expose the crystal interiors. To characterize internal structures of zircon grains and select better target sites for U-Pb dating, cathodoluminescence (CL) imaging was performed using a JXA-8800R Electron Probe at the Institute of Geology, Chinese Academy of Geological Sciences (CAGS), Beijing, China. Zircon U-Pb isotopic analyses of all the samples were carried out using a Sensitive High-Resolution Ion Micro Probe (SHRIMP) II following the method of Williams [48], at the Beijing SHRIMP Center, CAGS, Beijing, China. International standard SL13 (572 Ma) and TEMORA (417 Ma) were used for the test procedure as discussed by Black et al. [49]. For the sake of monitoring the contents of U, Th, and Pb, one SL13 reference sample was run for every nine analyses of the samples with a beam diameter of 30-40 μm. To maintain the precision of the measured results, one TEMORA reference sample was run for every four analyses of the samples. Detailed analytical procedures followed those described by Song [50]. The weighted mean ages were calculated, and Concordia diagrams were conducted using Isoplot (Ex ver3, Berkeley Geochronology Center, Berkeley, CA, USA) [51]. Monzogranites show fine-to-medium-grained texture (Figure 3c). They mainly consist of plagioclase, K-feldspar, quartz, biotite, and amphibole ( Figure 3d). Accessory minerals include apatite, zircon, monazite, and ilmenite. Plagioclase commonly occurs as euhedral grains up to 2 mm long. Occasionally, K-feldspar (up to 0.8 mm) contains small quartz inclusions. Quartz (0.1-0.8 mm) generally display subhedral grains. Biotite and amphibole grains are mainly euhedral to subhedral, occurring between the intervals or enclosed by quartz.
Granite porphyries are typically porphyritic with phenocrysts (up to 2 mm) of plagioclase, quartz, K-feldspar, and biotite (Figure 3e,f). Plagioclase and biotite phenocrysts are mainly euhedral or subhedral. K-feldspar phenocrysts occur as subhedral grains. Quartz phenocrysts occur as subhedral to anhedral grains in shape with rounded edges.

SHRIMP U-Pb Zircon Dating
Zircon grains were separated using conventional heavy liquids, hand-picking, and then mounted on epoxy resin and polished to expose the crystal interiors. To characterize internal structures of zircon grains and select better target sites for U-Pb dating, cathodoluminescence (CL) imaging was performed using a JXA-8800R Electron Probe at the Institute of Geology, Chinese Academy of Geological Sciences (CAGS), Beijing, China. Zircon U-Pb isotopic analyses of all the samples were carried out using a Sensitive High-Resolution Ion Micro Probe (SHRIMP) II following the method of Williams [48], at the Beijing SHRIMP Center, CAGS, Beijing, China. International standard SL13 (572 Ma) and TEMORA (417 Ma) were used for the test procedure as discussed by Black et al. [49]. For the sake of monitoring the contents of U, Th, and Pb, one SL13 reference sample was run for every nine analyses of the samples with a beam diameter of 30-40 µm. To maintain the precision of the measured results, one TEMORA reference sample was run for every four analyses of the samples.
Detailed analytical procedures followed those described by Song [50]. The weighted mean ages were calculated, and Concordia diagrams were conducted using Isoplot (Ex ver3, Berkeley Geochronology Center, Berkeley, CA, USA) [51].

Zircon Hf Isotopes
In situ Lu-Hf isotopic analyses were conducted by using a Neptune Plus multi-collector inductively coupled plasma mass spectrometry (MC-ICP-MS) (Thermo Fisher Scientific, Erlangen, Germany), attached to a GeoLas HD excimer ArF laser ablation (LA) system (Coherent, Göttingen, Germany) that is hosted at the Wuhan Sample Solution Analytical Technology Co., Ltd., Hubei. To correlate LA-MC-ICP-MS zircon Hf isotope data with U-Pb isotope data, analytical spots were situated close to or overlapping with the SHRIMP measure spots. A spot size of the laser was set to 44 µm, with a laser repetition rate of 8 Hz and an energy density of~7.0 J/cm 2 . To track the accuracy of analysis, zircon standards 91,500 and GJ-1 were used as reference samples, yielding weighted mean 176 Hf/ 177 Hf ratios of 0.282295 ± 0.000005 (2σ) and 0.282000 ± 0.000009 (2σ), respectively, which are good agreement with their recommended ratios. Detailed operating conditions can be found in Hu et al. [52]. Off-line selection, integration of analyte signals, and mass bias calibrations were made using soft ICPMSDataCal10.7 [53].

Major and Trace Elements
The samples were first powdered to~200 mesh. Then, major and trace element analyses of whole-rock were conducted at Createch Testing Technology Co., Ltd., Beijing. China. Major element analyses were performed by X-ray fluorescence spectrometry (XRF-1800). The analytical error and accuracy for major elements are generally better than 1% based on analyses of national rock standards (GSR-1, GSR-2, GSR-3, GSD-9, BHVO-2, and AGV-2). Trace element analyses of whole-rock were analyzed using Agilent 7500ce inductively coupled plasma mass spectrometry (ICP-MS). The detailed analytical procedures followed those described by Wei et al. [54].

Sr and Nd Isotopes
The Sr and Nd isotopic compositions were determined by using a multi-collector (MC) -ICP-MS (Neptune Plus, Thermo-Fisher) at Createch Test Technology Co. Ltd., Beijing, China. Detailed analytical procedures can be found in Yang et al. [55]. The ratios of 87 Rb/ 86 Sr and 147 Sm/ 144 Nd were calculated from Rb, Sr, Sm, and Nd concentrations measured by ICP-MS. During the analytical process, the average 147 Sm/ 144 Nd ratios for international basalt standard BHVO-2 and BCR-2 were 0.1542 (2σ, n = 17) and 0.1405 (2σ, n = 17), respectively. The average 143 Nd/ 144 Nd ratios for standard BHVO-2 and BCR-2 were 0.512964 ± 5 (2σ, n = 17) and 0.512608 ± 5 (2σ, n = 17), respectively, which are within the error of those reported by Li et al. [56]. The Sr isotope international standard NBS 987 was repeatedly tested for monitoring the accuracy of the test results, yielding an average 87 Sr/ 86 Sr ratio of 0.710247 ± 13 (2σ, n = 17), which is close to that reported by Zhai et al. [57].

Zircon U-Pb Age Data
The zircon SHRIMP U-Pb age data are listed in Supplementary Table S1. Zircon grains from the quartz diorite (sample TW28) show euhedral to subhedral, prismatic, and transparent, spanning from 60 to 140 µm in length. Their weakly oscillatory zoning is visible in CL images ( Figure 4a). Fifteen spots from fifteen zircon grains were carried out for U-Pb dating. These zircon grains have 824-891 ppm U and 178-912 ppm Th, with Th/U ratios in the range of 0.6-1.0, indicating a magmatic origin. The results gave 206 Pb/ 238 U ages ranging from 134 to 139 Ma. These analyses fall closely on the Concordia, yielding a weighted mean 206 Pb/ 238 U age of 136 ± 1.4 Ma (MSWD = 0.25; Figure 4a), which represents the crystallization age of quartz diorite.
Zircon grains from the monzogranite (sample TW8192) are euhedral to subhedral, varying from 70 to 160 µm in length. They are transparent and prismatic, with well-developed oscillatory zoning visible in CL images ( Figure 4b). The zircon grains possess 506-1407 ppm U and 198-1065 ppm Th, with Th/U ratios in the range of 0.33 to 0.76, suggesting a magmatic origin. Sixteen spots were performed on sixteen zircon grains for U-Pb dating, yielding 206 Pb/ 238 U ages spanning from 133 to 139 Ma. The concordant to sub-concordant analyses yield a weighted mean 206 Pb/ 238 U age of 136 ± 0.8 Ma (MSWD = 0.91; Figure 4b), which is interpreted as the crystallization age of monzogranite.   Zircon grains from the granite porphyry (sample TW3127) are transparent, prismatic, and euhedral, and are 80-200 µm in length. CL images reveal that they have apparent concentric zoning (Figure 4c). Fourteen spots were performed on fourteen zircon grains for U-Pb dating. These zircon grains have 227-1937 ppm U and 158-1412 ppm Th, with Th/U ratios of 0.4-1.18, indicating a magmatic origin. The results gave 206 Pb/ 238 U ages varying from 129 to 140 Ma. The concordant to sub-concordant analyses yield a weighted mean 206 Pb/ 238 U age of 137 ± 1.3 Ma (MSWD = 1.7; Figure 4c), which represents the crystallization age of granite porphyry.

Zircon Hf Isotopic Data
The zircon Hf isotopic data are given in Supplementary Table S2 and shown in Figure 5a. Eight zircon grains from the quartz diorite (sample TW28) were measured for in situ Hf isotope analyses. Their ε Hf (t) are restricted to high values of +8.1 to +9.6, corresponding to crustal Hf model ages (T DM C ) of 0.58 to 0.68 Ga.
Nine zircon grains from the monzogranite (sample TW8192) were determined for in situ Hf isotope analyses. The variation of ε Hf (t) is limited (+6.8 to +7.8), corresponding to T DM C ages of 0.7 to 0.76 Ga. Eight zircon grains from the granite porphyry (sample TW3127) were measured for in situ Hf isotope analyses. They also possess a restricted range of ε Hf (t) values from +7.4 to +9.1, corresponding to T DM C ages of 0.61 to 0.73 Ga.

Major and Trace Element Data
The major and trace element data for the Zalute granitoids are present in Supplementary Table S3. As shown in the total alkalis-silica (TAS) ((K 2 O + Na 2 O) versus SiO 2 ) diagram, the quartz diorite samples straddle the line between the diorite and monzonite field, and the monzogranite and granite porphyry samples plot mainly in the field of quartz-monzonite and granite, respectively (Figure 6a). Zircon grains from the granite porphyry (sample TW3127) are transparent, prismatic, and euhedral, and are 80-200 μm in length. CL images reveal that they have apparent concentric zoning (Figure 4c). Fourteen spots were performed on fourteen zircon grains for U-Pb dating. These zircon grains have 227-1937 ppm U and 158-1412 ppm Th, with Th/U ratios of 0.4-1.18, indicating a magmatic origin. The results gave 206 Pb/ 238 U ages varying from 129 to 140 Ma. The concordant to subconcordant analyses yield a weighted mean 206 Pb/ 238 U age of 137 ± 1.3 Ma (MSWD = 1.7; Figure 4c), which represents the crystallization age of granite porphyry.

Zircon Hf Isotopic Data
The zircon Hf isotopic data are given in Supplementary Table S2 and shown in Figure 5a. Eight zircon grains from the quartz diorite (sample TW28) were measured for in situ Hf isotope analyses. Their εHf(t) are restricted to high values of +8.1 to +9.6, corresponding to crustal Hf model ages (TDM C ) of 0.58 to 0.68 Ga.
Nine zircon grains from the monzogranite (sample TW8192) were determined for in situ Hf isotope analyses. The variation of εHf(t) is limited (+6.8 to +7.8), corresponding to TDM C ages of 0.7 to 0.76 Ga.
Eight zircon grains from the granite porphyry (sample TW3127) were measured for in situ Hf isotope analyses. They also possess a restricted range of εHf(t) values from +7.4 to +9.1, corresponding to TDM C ages of 0.61 to 0.73 Ga.

Major and Trace Element Data
The major and trace element data for the Zalute granitoids are present in Supplementary Table  S3. As shown in the total alkalis-silica (TAS) ((K2O + Na2O) versus SiO2) diagram, the quartz diorite samples straddle the line between the diorite and monzonite field, and the monzogranite and granite porphyry samples plot mainly in the field of quartz-monzonite and granite, respectively (Figure 6a).   (Figure 6d). Moreover, they show relatively high Sr (530-962 ppm; Figure 7g) and Ni contents (6.7-13.8 ppm; Figure 7h). The quartz diorite samples exhibit fractionated REE patterns (Figure 8a; La N /Yb N = 7-13), with negative to weakly positive Eu anomalies (Eu/Eu* = 0.56-1.09). In the primitive mantle-normalized trace element patterns (Figure 8b), they are enriched in large ion lithophile elements (LILE) (e.g., K, Pb, and Sr) and depleted in high field strength elements (HFSE) (e.g., Nb, Ta, and Ti).

Whole-Rock Sr-Nd Isotopic Data
Whole-rock Sr-Nd isotopic data for the Zalute granitoids are listed in Supplementary Table S4 and shown in Figure. 5b. Initial Sr and Nd isotopic ratios were calculated at their zircon weighted mean 206 Pb/ 238 U ages. The quartz diorite samples exhibit high and uniform initial 87

Whole-Rock Sr-Nd Isotopic Data
Whole-rock Sr-Nd isotopic data for the Zalute granitoids are listed in Supplementary Table S4 and shown in Figure. 5b. Initial Sr and Nd isotopic ratios were calculated at their zircon weighted mean 206 Pb/ 238 U ages. The quartz diorite samples exhibit high and uniform initial 87

Whole-Rock Sr-Nd Isotopic Data
Whole-rock Sr-Nd isotopic data for the Zalute granitoids are listed in Supplementary Table S4 and shown in Figure 5b. Initial Sr and Nd isotopic ratios were calculated at their zircon weighted mean 206 Pb/ 238 U ages. The quartz diorite samples exhibit high and uniform initial 87 Sr/ 86 Sr ratios of 0.7044-0.7049 and near-zero to positive ε Nd (t) values of −0.02 to +2.61, corresponding to two-stage Nd model ages (T DM2 ) of 0.72 to 0.93 Ga. The monzogranite samples also show depleted Sr-Nd isotopic compositions, with initial 87 Sr/ 86 Sr ratios of 0.7048-0.7049 and ε Nd (t) values of +0.98 to +1.5, corresponding to T DM2 ages of 0.81 to 0.82 Ga. The granite porphyry sample has an initial 87 Sr/ 86 Sr ratio of 0.7035, ε Nd (t) value of +1.37, and T DM2 ages of 0.82 Ga, which are close to those of the quartz diorite and monzogranite samples.
Although Shen et al. [72] reported a SHRIMP zircon U-Pb age of 137.4 ± 0.9 Ma for a quartz monzonite sample, few studies were undertaken for widespread quartz diorites, monzogranites, and granite porphyries at Zalute, which are the focus of this study. Here we present new ages of quartz diorite (136 ± 1.4 Ma), monzogranite (136 ± 0. 8 Ma), and granite porphyry (137 ± 1.3 Ma) by SHRIMP zircon U-Pb dating (Figure 4). Based on the above chronological constraints, the intermediate-felsic intrusive rocks at Zalute show a consistent crystallization age of~137-136 Ma, which is earlier than the alkaline intrusive rocks of~127-121 Ma.
The Zalute intermediate-felsic intrusive suite is broadly coeval with or slightly earlier than the Early Cretaceous granitoids in the surrounding regions, such as diorites, monzogranites, and granite porphyries at Linxi (~135-125 Ma) [16,18,58], monzogranites at Ganzhuermiao (~139-125 Ma) [73], and monzogranites and syenogranites at Xilinhot-Huolinguole (~139-121 Ma) [60,74]. The widespread occurrences of these granitoids in the SGXR record an Early Cretaceous giant igneous event of the eastern Central Asian Orogenic Belt [5,16]. Integrating our new data with previous studies in the SGXR (Figure 5a), the early Early Cretaceous intermediate-felsic intrusive suite and middle Early Cretaceous alkaline intrusive suite can be distinguished, representing two magmatic pulses with an age gap of more than 10 Ma. Although many studies have been conducted on the late-stage alkaline granitoids associated with the Baerzhe rare metal deposit [68][69][70][71], little is known about the early-stage intermediate-felsic granitoids at Zalute. In what follows, we focus only on the early Early Cretaceous intermediate-felsic intrusive suite.

Source Nature
The quartz diorite, monzogranite, and granite porphyry at Zalute have uniform and highly positive ε Hf (t) values ( +8.1 to +9.6, +6.8 to +7.8, and +7.4 to +9.1, respectively; Figure 5a), near-zero to positive ε Nd (t) values (0.98 to 2.61) and low initial 87 Sr/ 86 Sr ratios (0.7035 to 0.7049; Figure 5b), suggesting that they share the similar parental magma derived from juvenile source rocks. Furthermore, the occurrence of negative ε Nd (t) value (−0.02) of one sample implies the involvement of minor ancient crustal materials during magma generation.
2. Niu et al. [2] and Xiao et al. [82] argued that the juvenile isotopic signatures of granitoids can be inherited from the underplated oceanic crust. The resultant granitoids generally mixed with a large amount of ancient crustal materials in the sources, leading to large variation in ε Hf (t) and enrichment ε Nd (t) values, as exemplified by the granitoids from the West Kunlun Orogen [83], North Qilian Orogen [84,85], and West Qinling Orogen [86]. However, this is not in agreement with the case of the Zalute granitoids, which are characterized by uniform and high positive ε Hf (t) (Figure 5a) and depleted ε Nd (t) values ( Figure 5b). Furthermore, the granitoids derived from the underplated oceanic crust are expected to be formed in a syn-collision setting [2,77]. Nevertheless, a plot of Y + Nb versus Rb (Figure 6e) reveals that the Zalute intermediate-felsic granitoid samples are mainly plotted in the post-collisional granite field, indicating an extension setting. In addition, it is widely accepted that the SGXR has experienced lithospheric extension since the early Early Cretaceous [21,27,87]. Thus, the Zalute granitoids are difficult to explain with the underplated oceanic crust model.
3. The granitoids formed by magma mixing commonly show resorption texture or reversed zoning in plagioclase [84], which is not consistent with what we observed in the studied granitoids ( Figure 3). Moreover, the lack of mafic microgranular enclaves in the granitoids at Zalute also excluded the magma mixing model [14,22]. Moreover, the granitoids formed by magma mixing are expected to have a wide range of ε Hf (t) values (generally more than 10 units) due to incomplete mixing [88], which conflict with the uniform ε Hf (t) values (+6.8 to +9.6) of the studied samples (Figure 5a). Therefore, the magma mixing model is insufficient to explain the source of the Zalute granitoids.

Granite Type and Fractional Crystallization
In terms of source nature and geochemical characteristics, granitoids have been generally classified as I-, S-, and A-types [78,91,92]. The early Early Cretaceous granitoids at Zalute have relatively low zircon saturation temperatures (665 • C-816 • C; average 753 • C) according to the zircon saturation thermometer of Boehnke et al. [93], similar to those of typical I-type granite (~760 • C) but significantly lower than those of typical A-type granite (~840 • C) [94]. Most samples (exceptionally samples D9723 and D4449) are metaluminous to weakly peraluminous (Figure 6d), which are different from strongly peraluminous S-type granites [95]. Furthermore, the absence of aluminous minerals such as garnet, muscovite, and tourmaline further rules out their affinity with S-type granites. Only two samples display strongly peraluminous features with high A/CNK values of 1.12 and 1.32, which may be caused by magmatic fractionation [78,96].
The metaluminous to weakly peraluminous melts forming I-type granites are characterized by the low solubility of P and its decrease as the melt evolves, resulting in P 2 O 5 decreasing with increasing SiO 2 in the I-type melts [97]. By contrast, the strongly peraluminous melts forming S-type granites are unsaturated in P, hence no definite trend of decreasing P 2 O 5 with the increase in SiO 2 in the S-type melts [57,98]. Therefore, the contrasting behaviors of P between I-and S-type melts can be used to effectively distinguish between them [78]. Plotting P 2 O 5 against SiO 2 for the Zalute granitoids reveals a strongly negative correlation between them (Figure 7e), indicating a chemical affinity to I-type granite. This observation can be further reinforced by the K 2 O versus Na 2 O discrimination diagram that all the samples were plotted in the field of I-type granite (Figure 6c) [63]. Accordingly, the intermediate-felsic intrusive suite at Zalute is I-type granites.
Combination of geochronological, Sr-Nd-Hf isotopic, and whole-rock chemical data support a magmatic differentiation model of an intermediate-felsic granitic system, with a close petrogenetic relationship between quartz diorite, monzogranite, and granite porphyry at Zalute, as indicated by the following lines of evidence.
2. In plots of ε Hf (t) value versus age ( Figure 5a) and initial 87 Sr/ 86 Sr ratio versus ε Nd (t) value (Figure 5b), these granitoid samples are indistinguishable or overlapping within a narrow range in terms of their Sr-Nd-Hf isotopic compositions, indicating a common source.
3. The quartz diorites, monzogranites, and granite porphyries display similar chondrite-normalized REE patterns (Figure 8a) and primitive mantle-normalized trace element patterns (Figure 8b). Moreover, these granitoids show clear trends in chemical compositions defining linear regressions on variation diagrams, where the selected elements correlate negatively with SiO 2 (Figure 7), resulting from either partial melting or fractional crystallization. It is widely assumed that the effects of these two processes can be evaluated and distinguished by a La/Sm versus La diagram [29,66]. As shown in Figure 6f, the fractional crystallization process played a more important role in the evolution of the Zalute intermediate-felsic granitoid magmatism. This observation is further supported by the differentiation index increasing gradually from quartz diorites (60)(61)(62)(63)(64)(65)(66)(67)(68)(69)(70), toward monzogranites (73)(74)(75)(76)(77)(78)(79)(80)(81)(82)(83)(84)(85)(86), then to granite porphyries (83)(84)(85)(86)(87)(88)(89)(90)(91)(92)(93)(94). 4. Fractional crystallization has taken place in the intermediate-felsic magmatic system, as indicated by the pronounced depletion in Nb, Ti, and Eu (Figure 8a,b). The marked negative anomalies of Nb and Ti are thought to be related to the fractionation of biotite [99]. Previous studies have demonstrated that biotite/melt partition coefficients are significantly higher for V and Sc relative to Th, resulting in Th partitioned more strongly into the residual melts than V and Sc during biotite crystallization, and hence lower V/Th and Sc/Th ratios, accompanied with increasing SiO 2 /Al 2 O 3 ratios in the residual melts [100,101] with SiO 2 , which are related to the fractionation of amphibole [102]. A negative correlation between P 2 O 5 and SiO 2 (Figure 7e) resulted from apatite fractionation [103]. Taking all the above considerations into account, the parental melts equivalent to the early Early Cretaceous granitoids at Zalute were generated by partial melting of juvenile source rocks in the lower crust induced by mantle-derived magmatic underplating, followed by continued fractional crystallization from quartz diorites, towards monzogranites, then to granite porphyries. negatively correlated with SiO2, which are related to the fractionation of amphibole [102]. A negative correlation between P2O5 and SiO2 (Figure 7e) resulted from apatite fractionation [103].
Taking all the above considerations into account, the parental melts equivalent to the early Early Cretaceous granitoids at Zalute were generated by partial melting of juvenile source rocks in the lower crust induced by mantle-derived magmatic underplating, followed by continued fractional crystallization from quartz diorites, towards monzogranites, then to granite porphyries.

Geodynamic Implications
The geodynamic setting of the Early Cretaceous magmatic rocks in the SGXR has been hotly debated over the past several decades. Five scenarios have been proposed as follows: (1) mantle plume-rift-extension or other intraplate processes [35,76,104]; (2) lithospheric extension related to ridge subduction of the Mongol-Okhotsk Ocean and a resultant slab window [20]; (3) delamination of thickened crust related to the closure of the Mongol-Okhotsk Ocean [5,11,23,24] or subduction and subsequent retreat of the Paleo-Pacific oceanic slab [25,26,45,105]; (4) combination effects of lithospheric extension associated with break-off of the Mongol-Okhotsk oceanic slab and rollback of the Paleo-Pacific oceanic slab generating back-arc spreading [14,15]; and (5) post-collisional extension related to the scissor-type closure of the Mongol-Okhotsk Ocean [5,11,16,106,107].
Based on Early Cretaceous basaltic rocks characterized by both intraplate and volcanic arc geochemical properties, Ge et al. [76] proposed that the mantle plume hypothesis could interpret the Early Cretaceous magmatism in the SGXR. Larson and Olson [108] suggested that the process of rifting above a mantle plume commonly produces a circular-or ring-shape distribution pattern of magmatic rocks, which conflicts with the Early Cretaceous NE-trending granitoid belt in the SGXR [9,16]. Furthermore, such Early Cretaceous granitoid magmatism lasted for at least~25 Ma, which also argues against the mantle plume model that related magmatism is thought to continue for only a few million years due to the rapid upwelling mantle plume [44].
Widespread adakitic rocks, high-Mg andesites, and tholeiite are typical products of the slab window-related subduction [20,58,109]. However, these magmatic rocks were not found at Zalute, precluding the possibility of a slab window model in the generation of the studied intermediate-felsic intrusive suite.
The Early Cretaceous granitoids occur in the interior of the continent far from (>10 3 km) the active continental margin of the Paleo-Pacific subduction zone during the Early Cretaceous, responding to the intracontinental extension [14,15,110]. Additionally, recent seismological studies suggested that the back-arc extension of the Paleo-Pacific Ocean did not extend to the continental hinterland of the SGXR [111]. There is still a lack of robust evidence to demonstrate the far-field effect of the Paleo-Pacific plate on the intracontinental magmatism [16]. Therefore, the model involving the far-field effect of the Paleo-Pacific oceanic plate may not be responsible for the Zalute granitoid magmatism.
Here an alternative scenario that post-collisional extension related to the closure of Mongol-Okhotsk Ocean is favored based on the following observations. First, numerous studies have shown that post-collisional granitoid magmatism has a wide range in the compositions from high-K calc-alkaline (volume-dominated) to shoshonitic or alkaline granitoids [87,112]. The Zalute granitoids consist of the early Early Cretaceous intermediate-felsic intrusive suite (mostly high-K calc-alkaline in nature; Figure 6b) and middle Early Cretaceous alkaline intrusive suite, supporting the post-collisional magmatism. Secondly, the intermediate-felsic granitoids exhibit LREE enrichment relative to HREE (Figure 8a) and negative Nb-Ta anomalies (Figure 8b), commonly encountered in typical post-collisional granitoids [113]. These observations are further supported by the discrimination diagram for tectonic setting that all the samples were plotted in the post-collisional granite field (Figure 6e).
The closure of the Mongol-Okhotsk Ocean occurred gradually from southwest to northeast by the scissor-type manner, terminating in the northeast in the Late Jurassic-early Early Cretaceous [11,25,107,114]. This interpretation is also supported by recent palaeomagnetic studies (e.g., [115]). The SGXR witnessed a transient Mongol-Okhotsk collisional orogeny in the latest Jurassic-earliest Cretaceous [16,39]. After early Early Cretaceous, the SGXR experienced intensive lithospheric extension, causing the Early Cretaceous giant granitoid-forming event, accompanied by widespread basin-and-range tectonism, metamorphic core complexes, rift basins, and large-scale magmatic-hydrothermal deposits [5,14,27].
The post-collisional granitoid magmatism could be generated by slab break-off or delamination of thickened crust, both of which can provide sufficient heat for crustal melting to produce resultant magmas [116,117]. The lack of S-type granite and crustal eclogite formed in thickened crust-related settings [114,118] implies that the Zalute granitoids seem not to be related to a thickened crust. Furthermore, the lithospheric delamination process generally produces a planar distribution pattern of granitoid rocks, whereas the slab break-off process can result in a narrow, linear distribution pattern of counterparts [117,119,120]. The Early Cretaceous granitoid magmatism in the SGXR is characterized by a NE-trending magmatic belt (Figure 1c) [9,16], implying break-off of the Mongol-Okhotsk oceanic slab.
Considering the above information, the parental magma of intermediate-felsic intrusive suite at Zalute may be interpreted tentatively to be formed dominantly by the partial melting of the underplated juvenile basaltic crust, followed by magmatic fractionation, which is associated with break-off of the Mongol-Okhotsk oceanic slab during the post-collisional extension ( Figure 10). settings [114,118] implies that the Zalute granitoids seem not to be related to a thickened crust. Furthermore, the lithospheric delamination process generally produces a planar distribution pattern of granitoid rocks, whereas the slab break-off process can result in a narrow, linear distribution pattern of counterparts [117,119,120]. The Early Cretaceous granitoid magmatism in the SGXR is characterized by a NE-trending magmatic belt ( Figure 1c) [9,16], implying break-off of the Mongol-Okhotsk oceanic slab.
Considering the above information, the parental magma of intermediate-felsic intrusive suite at Zalute may be interpreted tentatively to be formed dominantly by the partial melting of the underplated juvenile basaltic crust, followed by magmatic fractionation, which is associated with break-off of the Mongol-Okhotsk oceanic slab during the post-collisional extension ( Figure 10).

Conclusion
1. SHRIMP zircon U-Pb dating for a suite of dioritic and granitic samples from Zalute gives a consistent crystallization age of ~137-136 Ma, involving quartz diorite dated at 136 ± 1.4 Ma, monzogranite at 136 ± 0.8 Ma, and granite porphyry at 137 ± 1.3 Ma. The new geochronological data reveals that the Zalute intermediate-felsic granitoids were formed in the early Early Cretaceous.
2. The quartz diorite, monzogranite, and granite porphyry, constituting an I-type intrusive suite, are characterized by uniform and depleted Sr-Nd-Hf isotopic compositions (initial 87 Sr/ 86 Sr ratios of 0.7035 to 0.7049, εNd(t) of −0.02 to 2.6, and εHf(t) of +6.8 to +9.6), indicating that they share a common origin. Their parental magma was dominantly derived from the partial melting of underplated mafic lower crust with minor ancient crustal components. Combined with their close spatial-temporal distribution pattern, common source, and compositional trend, a magmatic fractionation model of the Zalute intermediate-felsic intrusive suite is proposed, with a close petrogenetic relation between quartz diorites, monzogranites, and granite porphyries.
3. The post-collisional extension related to the closure of the Mongol-Okhotsk Ocean and subsequent slab break-off played an important role in the generation of the Zalute intermediate-felsic intrusive suite.

Conclusions
1. SHRIMP zircon U-Pb dating for a suite of dioritic and granitic samples from Zalute gives a consistent crystallization age of~137-136 Ma, involving quartz diorite dated at 136 ± 1.4 Ma, monzogranite at 136 ± 0.8 Ma, and granite porphyry at 137 ± 1.3 Ma. The new geochronological data reveals that the Zalute intermediate-felsic granitoids were formed in the early Early Cretaceous.
2. The quartz diorite, monzogranite, and granite porphyry, constituting an I-type intrusive suite, are characterized by uniform and depleted Sr-Nd-Hf isotopic compositions (initial 87 Sr/ 86 Sr ratios of 0.7035 to 0.7049, ε Nd (t) of −0.02 to 2.6, and ε Hf (t) of +6.8 to +9.6), indicating that they share a common origin. Their parental magma was dominantly derived from the partial melting of underplated mafic lower crust with minor ancient crustal components. Combined with their close spatial-temporal distribution pattern, common source, and compositional trend, a magmatic fractionation model of the Zalute intermediate-felsic intrusive suite is proposed, with a close petrogenetic relation between quartz diorites, monzogranites, and granite porphyries.
3. The post-collisional extension related to the closure of the Mongol-Okhotsk Ocean and subsequent slab break-off played an important role in the generation of the Zalute intermediate-felsic intrusive suite.