December 2006 LIP of the Month

⁴⁰Ar/³⁹Ar geochronology on the Late Mesozoic volcanism in the Great Xing’an Range (NE China): implications for the dynamic setting of NE Asia

Fei Wang,  Xin-hua Zhou,  Lian-Chang Zhang,  Ji-Feng Ying,  Yu-Tao Zhang,  Fu-yuan Wu,  Ri-xiang Zhu

State Key Laboratory of Lithospheric Evolution, Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing 100029, China

[Excerpted and modified from Fei Wang et al., 2006, Late Mesozoic volcanism in the Great Xing’an Range (NE China): Timing and implications for the dynamic setting of NE Asia, Earth and Planetary Science Letter 251: 179-198]

Introduction

Late Mesozoic volcanic rocks occur over a vast area in northeastern China (NEC) and its adjacent areas, including eastern and southern Mongolia, Korean Peninsula, and Japan (Fig.1). Many geological, petrological and chronological studies have previously been carried out on these extensive late Mesozoic volcanic rocks (Zhou et al., 1982; Kinoshita, 1995, 2002; Zhou et al., 2000; Fan et al., 2003; Wu et al., 2005), and have not resulted in a consensus about the mechanism and petrogenesis. The mantle plume hypothesis (Ge et al., 1999), post-orogenic diffuse extension mechanism(Fan et al., 2003), lithospheric mantle delamination model(Wu et al., 2005), and many other hypothesis have been suggested for this magmatism in NEC.  

It was noted that the igneous rocks turn younger in age oceanwards along the Asia continental margin (Kinoshita, 1995, 2002; Zhou et al., 2000), which is largely supported by the K-Ar ages. Ridge subduction of the Kula-pacific Plate was employed to exlpain this migration by Kinoshita (1995, 2002), whereas Zhou and Li (2000) argue that simple subduction of the Kula-pacific Plate with an increased slab dip angle could also be a mechanism for it. 

       The lack of precise geochronological data on these volcanics, especially on the basalts, therefore still causes such disagreements and hampers further advances in the study of the mechanism of magmatism in NEC. Recently, an igneous event ranging from 120 to 130 Ma in northeastern China was found, based on the geochronology on plutons (Wu et al., 2005). However, these are still indirect timing from the intrusive rocks whose exact time relation with the volcanics eruption is unknown. More importantly, precise dating of the volcanic successions in the Great Xing’an Range is highly desirable and crucial for the understanding of the relationship between NEC and Mongol- Okhotsk suture zone. In this paper, new ⁴⁰Ar/³⁹Ar dates from the volcanic rocks of the Great Xing’an Range are presented. This new data, in conjunction with the previous compilations of datings from adjacent areas, including southern and eastern Mongolia (Meng, 2003; Yarmolyyuk and Kovalenko, 2001), northeast China (Wu et al., 2005; Meng, 2003;Wanh et al., 2002), Korean Peninsula and southwest Japan (Kinoshita, 1995), put new constraints on the temporal-spatial distribution of the volcanism in this region. As a conclusion, a new model is suggested attempting to explain this temporal-spatial distribution of volcanism in a frame of plate tectonic mechanisms.

2        Geological setting

Topographically, there is a sharp altitude contrast between the high plateaus to the west and hilly plains to the east along the Great Xing’an Range. The Great Xing’an Range, which represents the steepest altitude gradient from the east to west of NEC, also coincides with the steepest gradient in gravity anomalies and crustal thickness (Niu, 2005). Interestingly, it also marks the steepest gradient in mantle seismic velocity clearly seen at depths of 100 km and 150 km (Niu, 2005). The sudden seismic velocity decreases across the Great Xing’an Range from the west to the east. This is consistent with the interpretation that in such depths, the mantle beneath the plateaus in the west comprises “cold” and “fast” lithosphere whereas beneath the east the asthenosphere is “hot” and “slow”. The latter is also consistent with the recognition that the lithosphere beneath eastern China, including NEC is anomalously thin, considering the geologically perceived cratonic nature in the North China. Petrologically, the occurrance of Paleozoic diamondiferous kimberlites in the NEC (e.g., Fuxian in Liaoning Province, Mengyin in Shandong Province ) (Sun et al., 1993; Chi et a., 1992; Lu et al, 2000; Griffin et al., 1998), indicates that the NEC lithosphere must have been ~200 km thick in Paleozoic times. However, recent studies of mantle xenoliths (Song and Frey, 1989,1990; Zhi et al., 1990; Xu et al., 1998, 2000; Xu, 2001) indicate a much thinner present-day lithosphere, perhaps no more than 80 km thick beneath NEC.  This is confirmed both by seismic studies (Chen et al., 1991) and mantle tomography (Niu et al., 2005). Hence, the lithosphere beneath NEC must have lost a portion of 120 km thickness (Menzies et al., 1993), probably in the Mesozoic (Griffin et al., 1998; Xu, 2001; Menzies et al., 1993; Deng et al., 1998, 2004; Zheng et al., 2001; Gao et al., 2002; Zhang and Zheng, 2003; Yan et al., 2003).

Extensional structures reworked the NEC during late Mesozoic time. Basins began developing, such as the Erlian, Hailar, Songliao basins and the eastern Gobi basin in southeastern Mongolia. In the eastern Gobi basin across the boarder of eastern Mongolia and China, a volcanic interlayer at the lower part of the basin boreholes yielded an ⁴⁰Ar/³⁹Ar age of 155±1 Ma (Graham et al., 2001).  Although there are no dates on the cored volcanic rocks in the Erlian basin, equivalents of the Xingganling Group in West Liaoning are well constrained by ⁴⁰Ar/³⁹Ar and K-Ar of ~156 Ma (Meng, 2003). Volcanism started in Songliao basin at late Jurassic based on the ⁴⁰Ar/³⁹Ar and K-Ar ages of 157.9±2.7Ma from a basalt at the base of a borehole(Wang et al., 2003). These dates suggest that rifting of the NEC basins started ~155Ma synchronously. The wide occurrence of A-type granites and alkaline rocks of late Mesozoic age in NEC and its adjacent area also suggest that the NEC was not in a extensional regime during that time (Wu et al., 2005). Other evidences for the extensional setting in NEC and its adjacent areas in late Mesozoic time come from the intrusion of dyke swarms (Shao and Zhang, 2002), exhumed metamorphic core (Zheng et al., 1001; Webb et al., 1999), and formation of basins ( Meng, 2003).

Late Mesozoic volcanic rocks cover ~100,000 km² in the Great Xing’an Range (BGMRNM, 1991). Some successions have cumulative thicknesses of up to ~ 4-5 km (Song and Dou, 1997; Xie, 2000). These rocks comprise a wide spectrum of rock types, including basalts, basaltic andesites, trachytes, rhyolites, volcanic clastic and tuffs. Based on the lithological associations and lava flow sequences, three group divisions are widely accepted for this region: Tamalan, Sangkuli and Yiliekede Formations.

The volcanic rocks also occur in the other parts of NEC, including Songliao Basin, Jiaodong peninsular, Yanshan and Liaoxi area. In the south and east of Mongolia along the Chinese border, volcanics are also the dominant rock, as well as in the Gobi Basin (Yarmolyyuk and Kovalenko, 2001). The distribution of these volcanics are constrained by the tectonic lines parallel to both, the Mongol-Okhotsk suture and Pacific Plate suduction zone (Fig. 1).

3  The ⁴⁰Ar/³⁹Ar dating techniques

Twenty groundmass samples were obtained from basalts and basaltic andesites collected from lavas as showed in Fig.2 and table 1.

These samples were dated by using the step-heating ⁴⁰Ar/³⁹Ar method. In order to constrain the eruption age and avoiding excess argon, a binocular microscope was used to carefully remove the 60-80 mesh granules of processed groundmass. ⁴⁰Ar/³⁹Ar measurements were performed at Institute of Geology and Geophysics of Chinese Science Academy (IGGCAS), Beijing. A number of neutron fluence monitors (standards) have been intercalibrated at ⁴⁰Ar/³⁹Ar Lab. of IGGCAS (Wang et al., 2006): relative to 18.6 Ma Brione muscovite monitor (Flish, 1982), nine total fusion analyses of Mt Dromedary ( NW Wales, Australia ) biotite (Ga 1550), gave a mean age of 98.5±0.6 Ma, consistent with the 98.5±0.8 Ma and 98.8±0.5 Ma ages determined by Spell et al. (2003) and Renne et al. (1998) respectively.

 Groundmass wafers weighing between 3-16 mg, multiple samples of the 18.6±0.4 Ma neutron fluence monitor mineral Brione muscovite were irradiated in vacuo within a cadmium-coated quartz vial for 45.8 hours in position H8 of the facility of Beijing Atomic Energy Research Institute reactor (49-2). Six to eight replicate analyses of the monitors from each position in the vials were conducted to constrain the vertical neutron fluence gradient to within ±0.7%. This additional uncertainty was propagated into the plateau and inverse isochron ages. However, complete external errors including those arising from the decay constants and primary K-Ar standards were not propagated.

Interfering nucleogenic reactions were checked for every irradiation by using CaF and K₂SO_4,. The correction factors in this study are (³⁶Ar/³⁷Ar)_Ca=0.000261±0.000014; (³⁹Ar/³⁷Ar)_Ca= 0.000724±0.000028; (⁴⁰Ar/³⁹Ar)_K=0.000880±0.000023. Mass discrimination was monitored using an on-line air pipette from which multiple measurements are made before and after each incremental-heating experiments.  The mean over this period is 1.00831±0.00017 per amu and the uncertainty of this value is propagated into all age calculations.

Groundmass wafers were placed into a Ta tube resting in the Ta crucible of an automated double-vacuum resistance furnace.  These were incrementally-heated in 15 steps of 10 minutes each from 700 or 750 ^oC to 1500 ^oC. Following 5 additional minutes of gas purification on Al-Zr getters, isotopic measurements were made on a mass spectrometer MM5400 with a Faraday cup and an electron multiplier of which the latter was used as the collector during this study. Hot system blanks determined several times each day prior to degassing the samples were typically 3x10^-16 mols of ⁴⁰Ar and 9x10^-19 mols of ³⁶Ar in nearly atmospheric ratios and 2-3 orders of magnitude smaller than sample signals. Although the mean blank errors were generally ~2% for ⁴⁰Ar and ~5% for ³⁶Ar, the large size of the samples relative to the blank minimized the impact of propagating these errors into the final age calculations.

Plateau ages were determined from 3 or more contiguous steps, comprising >50% of the ³⁹Ar released, revealing concordant ages at the 95% confidence level. The uncertainties in plateau ages reflect multiplication by the MSWD and were obtained by standard weighting of errors for individual steps according the variance (Taylor, 1982). Thus, more precise determinations were given greater weight than those of lower precision and the overall uncertainty about the mean value may be greatly reduced. Because no assumption is made regarding the trapped component, the preferred ages are inverse isochrones, calculated from the plateau steps using the York (York, 1969) regression algorithm. Errors are reported at the 2σ confidence level.

4  ⁴⁰Ar/³⁹Ar analysis results

The age results from the step-heating experiments are presented in Table 1. The age spectra and isotope correlation (inverse isochron) diagrams are illustrated in Figs.3,4,5. For each sample the argon release age spectra and inverse isochrones are presented. Both the plateau and inverse isochron age uncertainties are given at 2σ level, and do not include systematic errors related to standards or the ⁴⁰K decay constants, which should be considered if these results are compared to ages estimates obtained from other radioisotopic systems (Begemann et al., 2001).

The detailed discussion on the ⁴⁰Ar/³⁹Ar data quality is seen in Wang et al. (2006).

Tamulan Formation  Seven samples (MZL04-6, MZL10, MZL13, MZL16, ERBY 04-1, ERBY04-4 and ERBY1-9) were collected from the Tamulan Formation (Fig.2), which yielded well-defined age spectra with plateau ages in two ranges from 160.0±0.8 to 162.6±0.7 Ma and 139.7±0.7 to 147.0±0.8 Ma (Table 1, Fig.3). Apart from ERBY04-4, all these plateau ages are quite consistent with their respective intercept ages (Table 1, Fig.3) obtained from the isotope correlation diagrams (Fig.3). Regression of the data on the isotope correlation diagram indicate that the trapped initial ⁴⁰Ar/³⁶Ar ratios  no measurable excess argon was caught when they erupted as shown in Table 1 and Fig.3.. Fig.3f shows that the inverse isochron age for ERBY04-4 is 142.4±1.0 Ma, which is a little higher than its plateau age 140.3±0.7 Ma (Fig.3F). The initial ⁴⁰Ar/³⁶Ar ratio of this sample, 270±19, is apparently lower than the atmospheric value, implying that the background contribution in the data and should be considered when interpreting plateau ages. As no assumptions are made about the initial ⁴⁰Ar/³⁶Ar ratios on the inverse isochron age, we regard that the inverse isochron age of ERBY04-4, 142.4±1.0 Ma, is preferred.

The dating results of samples from Tamulan formation clearly show that there are two distinct time periods: from ~160 to ~163 Ma and ~140 to ~147 Ma. This suggests that these lavas with different ages, which were previously defined as “Tamulan Formation”, should be two formations. However in this paper, we follow the tradition to use the term “Tamulan Formation” for these lavas of the two periods.

Shangkuli Formation  Eight samples (GH07, GH10, TH08, TH24, TH22, ELC04-1, ZLT04-8and GH04-1) were collected from Shangkuli formation at different sites (Fig.2). Six samples show fine-defined age spectra over 70% of ³⁹Ar released (Fig.4) giving a plateau age range from 121.2±0.6 to 124.5±0.6 Ma (Table 1, Fig. 4). Apart from TH24 and GH04-1, the samples exhibit consistent inverse isochron ages ( Table 1, Fig.4) with their plateau ages respectively. The initial values of ⁴⁰Ar/³⁶Ar of these samples agree with that of the air (Table 1, Fig.4) suggesting they trapped only atmospheric argon when they formed. TH24 and GH04-1 show different spectrum ages with inverse isochron ages as well as initial ⁴⁰Ar/³⁶Ar values higher than atmospheric value ( 389±70 and 325±16 respectively,  Table 1 and Fig.4). This implies that excess argon was trapped when they formed. Therefore, we regard the inverse isochron age as being more objective, as no assumptions are made about the initial ⁴⁰Ar/³⁶Ar ratios. These dates suggest that the Shangkuli formation was formed during ~120-~125 Ma.

Yiliekede Formation  Five basaltic rocks (YKSNQ 04-4, GH04-4, YKS04-3, JGD04-4 and YKSNQ04-1) were sampled from Yiliekede formation in different localities (Fig.2). Four of these samples (YKSNQ 04-4, JGD04-4, and GH04-4) yielded age spectra with plateau ages from 106.2±0.6 to 115.8±0.6 Ma ( Table 1, Fig.5). In the ³⁶Ar/⁴⁰Ar versus ³⁹Ar/⁴⁰Ar correlation diagrams (Fig.5), the data of the four samples define intercept ages consistent with their respective plateau ages ( Table 1); and the ⁴⁰Ar/³⁶Ar initial ratios are quite consistent with the atmospheric ⁴⁰Ar/³⁶Ar ratio (295.5) as well (Table 1, Fig.5). However, dating on the sample YKSNQ04-1 shows no statistically meaningful plateau age (Fig.5V), but its consistent integrated age (113.2±0.6 Ma) and inverse isochron age (114.3±1.0 Ma), and initial ⁴⁰Ar/³⁶Ar (300±11) indistinguishable from atmospheric value suggest these ages are reasonable assumptions.

The sample YKS04-3 shows a much lower age than other samples from the same formation. As discussed above, this may be caused by the ⁴⁰Ar loss from the alteration. Therefore, if we exclude YKS04-3, the dating results from Yiliekede formation indicate that the eruption of the formation is constrained for a duration between ~113- and 116 Ma.

5  Discussion and conclusion

5.1  The volcanic succession in the Great Xing’an Range

Previously several workers believed in continuous volcanic activity in the Tamulan Formation during the late Jurassic time ( ~160) and throughout the Late Jurassic-Early Cretaceous times (~160-~90) in the Great Xing’an Range. This assumption was based on data from widespread volcanic bodies of ages determined by Rb-Sr and Sm-Nd dating methods (BGMRNM, 1991; Shao et al., 1998). A recent study (Wu et al., 2005) showed that this giant igneous event in the Great Xing’an Range can be constrained in a much shorter duration of 10 Ma from ~120 to ~130 Ma, based on U-Pb dating on intrusive bodies. However, our dating results on the three main volcanic formations in the Great Xing’an Range do not support these views.

Our dates indicate that the three main formations, Tamulan, Shangkuli and Yiliekede, formed in short durations of ~163-160 and ~147-140, ~120-125, ~113-116 Ma respectively. It should be noted that such a conclusion needs to be ascertained by further work due to the representability of the samples. The Tanmulan Formation indicates the start of the igneous activity during late Jurassic times in this region.

5.2  Migration of Late Mesozoic volcanic activity in Northeast Asia

Volcanism during the Late Mesozoic time occurred not only in the Great Xing’an Range, but widespread in NEC and its adjacent areas, such as south and east Mongolia, the Korean Peninsula and southwest Japan. Our dating results are the first to place a precise constraint on the timing of this volcanic succession and the main stage as the Great Xing’an Range. This data, combined with recent age compilations of igneous activity in the adjacent regions (Kinoshita, 1995;Wu et al., 2005; Yarmolyyuk and Kovalenko, 2001; Wang et al., 2002), strongly suggests that igneous activity migrated from west to east in Northeast Asia during the Late Mesozoic (Fig.6). This migration is placed between the Mongol-Okhotsk suture and Pacific Plate subduction zones.

In south and east Mongolia, southeast of the Mongol-Okhotsk suture, the volcanic activity continued for the past 160 Ma peaking around 160-140 Ma (Yarmolyyuk and Kovalenko, 2001).  Based on the K-Ar isotopic ages of the volcanic successions, there are similarities in compositional parameters of most of the volcanic rocks throughout the entire period of formation of this region. The various volcanic associations are dominated by basic rocks, accounting for no less than 95% of the total volume of the entire volcanic rocks (Yarmolyyuk and Kovalenko, 2001). They are presented by subalkaline olivine-basalts and alkali basaltoids, displaying a geochemistry similar to that of their equivalents in northern China. Studies of their isotopic compositions and associated mantle xenoliths like lherzolites suggest that their primary melts were derived from the lithospheric mantle (Yarmolyyuk and Kovalenko, 2001).

Volcanic rocks also spread in Yanshan range, south of the Great Xing’an Range (Fig.1), with a cumulative thickness of the volcanic succession up to ~4 km (Xie, 2000). Zircon U-Pb ages and ⁴⁰Ar/³⁹Ar plateau ages for the volcanic rocks (Davis et al., 2001; Smith et al., 1995; Swisher et al., 1999; Luo et al., 2001; Miao, 2002; Chen et al., 1997; Liu et al., 2003; Li et al., 2004) constraint a duration of 130-150 Ma, with a peak of ~135-145 Ma. They take similarities with those in the Great Xing’an Range geochemically and petrologically.

To the east of the Great Xing’an Range lies the Songliao basin, the largest oil- and gas- producing basin in China. Mesozoic igneous rocks are widespread throughout the basin. A newly obtained core section and high-quality deep seismic reflections lines provide good selection of samples and better understanding of the structure of the volcanic successions buried in the basin. Recent ⁴⁰Ar/³⁹Ar and K-Ar ages (Wang et al., 2002) revealed that the volcanic activities mainly took place between 120-130 Ma.

Late Mesozoic volcanic rocks are also widespread in the Liaoxi area (Western Liaoning Province) to the south of Songliao Basin. Ages for the volcanic rocks from this area have ages between 128.4±0.2 to 120.3±0.7 Ma (He et al., 2004; Smith et al., 1995; Swisher et al., 1999; Wang et al., 2001; Zhu et al., 2004a, 2004b; He et al., 2006). The compilation of ages younger than 150 Ma, including those of intrusive rocks, indicate that 120-130 Ma is the peak time of igneous activity in this area (Wu et al., 2005).

Based on the description above, a general view of the spatial and temporal distribution for Late Mesozoic volcanic activity in the NEC and its adjacent areas can be derived. From west to east, the peak time of igneous activities changed from 160-140 Ma (in east Gobi of Mongolia and west Great Xing’an Range) to 120-130 Ma (Songliao-Liaoxi area and Liaodong peninsula ), supporting the migration of a volcanic front in the Late Mesozoic time in NEC and its adjacent areas.

5.3  Nature and mechanism of the late Mesozoic magmatism in northeastern Asia

It has been recognized that NEC and its adjacent areas are characterized by lithospheric thinning during the Mesozoic (Xie, 2000; Griffin et al., 1998; Menzies et al., 1993; Zheng et al., 2001; Gao et al., 2002; Menzies and Xu, 1998) and Os isotopic constrains indicate that thinning was accomplished by delamination (Gao et al., 2002; Wu et al., 2003), which coincides with the extension setting spatially and temporally. Therefore, it is vital to understand the tectonic regime controlling delamination and its link to widespread Mesozoic magamtism in NEC and its adjacent areas.

Several models have been proposed to explain the extensional features, widespread magmatism and lithospheric thinning. For example, the Pacific backarc extension model (Watson et al., 1987; Traynor and Sladen, 1995; Ratschbacher et al., 2000), the hotpot and plume model (Castillo, 1988; Duncan and Richards, 1991), the subduction model of Pacific Plate beneath eastern China (Zhou et al., 2000; Ratschbacher et al., 2000), the intraplate rifting model (Li, 2000), the Triassic collision model between the Yangze and North China cratons (Griffin et al., 1998; Gao et al., 2002), and the lithospheric mantle delamination model (Wu et al., 2005).

Backarc extension (Watson et al., 1987; Traynor and Sladen, 1995; Ratschbacher et al., 2000) may be the most popular view due to some arc signatures of the widespread calc-alkaline volcanic and I-type granitic rocks (Zhou et al., 2000). But this mechanism fails to account for the fact that late Mesozoic extension occurred over a vast area, as manifested above, more than 2000 km from the Pacific subduction zone. The hotpot and plume model (Castillo, 1988; Duncan and Richards, 1991) argue that a mantle avalanche, induced by the closure of Tethys (Machetel and Humler, 2003) and breakup of Gondwana (Wilde et al., 2003) of 180 Ma ago, caused temperature rising in the upper mantle and ensuing erosion of the overlying lithospheric mantle from the rising asthenosphere, which then resulted in lithospheric thinning. However, basalts and gabbros of mantle-derived mafic rocks, the predicted products of mantle plume activity, are rarely documented in the NEC and its adjacent areas (Wu et al., 2005). A super plume was inferred around ~125 Ma (Larson, 1991), which may have affected the whole earth. It has been proposed that the Early Cretaceous mid-Pacific super-plume increased subduction rates at its outer margins, which assisted in the lithospheric delamination in Eastern China (Wu et al., 2005), but it is hard to explain the migration of magmatism.

Delamination of the lithospheric mantle would bring hot asthenosphere into contact with the Moho (Bird, 1979). This should promote massive crustal melting and predicts a progressive migration of the resultant volcanism in the direction of delamination propagation (Turner et al., 1999). The distribution of the massive igneous rocks in NEC and its adjacent areas temporally and spatially implies a possible relationship of the magamtism with the Pacific Plate subduction and the closure of Mongol-Okhotsk Ocean in some way. Therefore, we propose a shears-shaped delamination model of the lithospheric mantle beneath NE Asia, induced by paleo-Pacific Plate subduction, as illustrated in Fig.7. Continued subduction of paleo-Pacific Plate moved the Northeast China–Mongolia Block northwestwards, and the closure of Mongol-Okhotsk Ocean eventually led to collision between Northeast China-Mongolia and Siberia by ~160 Ma (Zorin, 1999). A recent study suggested that the Mongol-Okhotsk suture in eastern Mongolia formed at least by ~172Ma (Tomurtogoo et al., 2005).  This collision then obstructed the northwestward movement of the region, which shortened and thickened the lithosphere of NEC and its adjacent areas. The strain intensified gradually enough to have the thickened lithosphere delaminated at ~160 Ma starting from the southeast Mongolia-the great Xin’an Range, like a shears opened eastwards (Fig.7). This led to asthenosphere upwelling, extensional tectonic setting, underplating, and ensued extensive magmatism activity propagation eastwards. The high degree of crustal melting accompanied this process to produce the geochemical and isotopic features as discussed above. As the “shears” opened eastwards, alkali granitic plutonism became more pronounced, indicative of existence of voluminous magma ponds in the lower crust and involvement of mantle melts.  The magmatic underplating played a crucial role in generation of the alkali granite plutons (Shao et al., 2000). Climax of plutonism was then followed by widespread normal basins and formation of metamorphic core complexes in the upper crust. During ~130-120Ma, the delamination reached its climax and left the most widespread volcanic formation (such as Shangkuli Formation in the Great Xing’an Range) in NEC.

Geochemical and timing evidences were reported for the magmatic underplating beneath the crust of northeastern China. Granulite xenoliths form the Cenozoic Hannuoba basalts, North China, was regarded as metamorphosed facies of magmatic underplating in northeastern China (Fan et al., 1998; Liu et al., 2004). Zircons from these granulite xenoliths yielded two U-Pb age populations of ~160-140 and ~140-80 Ma (Fan et al., 1998; Liu et al., 2004). Combination of Nd, Sr and Pb isotopic compositions, these two age periods were explained that the granulites were the products of ~160-140 Ma basaltic underplating and ~140-80 Ma granulite-facies metamorphism (Liu et al., 2004). Those Precambrian protoliths underwent granulite-facies metamorphism at 150-80Ma (Liu et al., 2004). Zircons of magmatic origin from an olivine pyroxenite xenolith suggest basaltic underplating at 97-158 Ma (Liu et al., 2004). The overlapping timing for the granulite-facies metamorphism and the basaltic underplating indicates that the Mesozoic granulite-facies metamorphism was induced by heating from the basaltic underplating at the base of the crust.

Recent studies on the tectonic transition from contractional to extensional deformation during late Mesozoic in NEC and its adjacent areas (Meng, 2003; Zhai et al., 2004; Ren et al., 2002) lends further support to this interpretation. Studies on the basins and extensional structure in NEC and its adjacent areas show that contracted high-standing plateau caused by subduction of Pacific Plate and collision of north China and Siberia had transited from crustal compression to extension during 150-140 Ma (Meng, 2003; Zhai et al., 2004; Ren et al., 2002).  This is consistent with the scenario of our proposed shears-shaped delamination model.       

Reference