Apr 7, 2016 - 4. Department of Earth, Environmental and Planetary Sciences, ...... domains can be constrained by the first episode of zircon growth, ..... Hickmott, D.D., Shimizu, N., Spear, F.S., Selverstone, J., 1987. ..... Geology 35, 9-12.
Growth of metamorphic and peritectic garnets in ultrahigh-pressure metagranite during continental subduction and exhumation in the Dabie orogen Qiong-Xia Xia, Haozhen Wang, Li-Gang Zhou, Xiao-Ying Gao, YongFei Zheng, James Ashton Van Orman, Haijun Xu, Zhaochu Hu PII: DOI: Reference:
Please cite this article as: Xia, Qiong-Xia, Wang, Haozhen, Zhou, Li-Gang, Gao, Xiao-Ying, Zheng, Yong-Fei, Van Orman, James Ashton, Xu, Haijun, Hu, Zhaochu, Growth of metamorphic and peritectic garnets in ultrahigh-pressure metagranite during continental subduction and exhumation in the Dabie orogen, LITHOS (2016), doi: 10.1016/j.lithos.2016.08.043
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
SC R
IP
T
ACCEPTED MANUSCRIPT
Growth of metamorphic and peritectic garnets in
NU
ultrahigh-pressure metagranite during continental subduction
MA
and exhumation in the Dabie orogen Qiong-Xia Xia1, 2*, Haozhen Wang3, Li-Gang Zhou1, 3, Xiao-Ying Gao1, Yong-Fei Zheng1,
D
James Ashton Van Orman4, Haijun Xu5, Zhaochu Hu5
TE
1. CAS Key Laboratory of Crust-Mantle Materials and Environments, School of Earth and Space Sciences, University of Science and Technology of China, Hefei 230026, China
CE P
2. State Key Laboratory of Mineral Deposits Research, School of Earth Sciences and Engineering, Nanjing University, Nanjing, China 3. State Key Laboratory of Lithospheric Evolution, Institute of Geology and Geophysics,
AC
Chinese Academy of Sciences, Beijing 100029, China 4. Department of Earth, Environmental and Planetary Sciences, Case Western Reserve University, 10900 Euclid Avenue, Cleveland OH 44106, USA 5. State Key Laboratory of Geological Processes and Mineral Resources, Faculty of Earth Sciences, China University of Geosciences, Wuhan 430074, China
ACCEPTED MANUSCRIPT Abstract Two generations of garnet are recognized in ultrahigh-pressure (UHP) metagranite from the Dabie orogen by a combined study of petrography, major and trace element profiles in garnet,
T
and phase equilibrium modeling for metagranite. The results enable distinction between
IP
metamorphic and peritectic garnet on the basis of BSE images, and major and trace element compositions. Our research provides new insights into the growth of anatectic garnet due to
SC R
dehydration melting of UHP metamorphic rocks during exhumation from mantle depths. The first generation of garnet (Grt-I) occurs as a broad domain in the center, which is related to metamorphic growth during prograde subduction. This garnet is dark in BSE images, rich in
NU
grossular and poor in almandine and pyrope. The chondrite-normalized rare earth element (REE) patterns show LREE depletion and flat MREE-HREE patterns. The second generation
MA
of garnet (Grt-II) occurs as a rim of euhedral garnet, or as patches in Grt-I domains, recrystallized after dissolution of preexisting metamorphic garnet in the presence of anatectic melts during exhumation. It is bright in BSE images, poor in grossular, and rich in almandine
D
and pyrope contents. Trace element analyses on Grt-II domains yield high contents of Sc, Cr,
TE
Y and HREE and low contents of Ti and MREE. The chondrite-normalized REE patterns exhibit LREE depletion, and steep MREE-HREE patterns. Based on REE partitioning
CE P
between garnet and zircon/titanite, the last growth times for metamorphic and anatectic garnets are constrained by zircon and titanite U-Pb ages to be ~240 Ma and ~220 Ma, respectively. Based on anatectic microstructures and a modeled P-T pseudosection, it is
AC
suggested that dehydration melting occurred at 2.0-2.5 GPa during exhumation. Melting occurred through the breakdown of phengite via the peritectic reaction: garnet (I) + phengite + plagioclase + quartz → garnet (II) + biotite + K-feldspar + melt.
ACCEPTED MANUSCRIPT 1. Introduction Dehydration melting of deeply subducted continental crust during exhumation from mantle depths has been demonstrated by several lines of evidence at different scales. These include
T
the occurrence of multiphase solid (MS) inclusions in ultrahigh-pressure (UHP) metamorphic
IP
refractory minerals such as garnet, omphacite and kyanite (e.g., Ferrando et al., 2005; Frezzotti et al., 2007; Gao et al., 2012a; Chen et al., 2013a), felsic veinlets in UHP eclogites
SC R
and gneisses (e.g., Zhao et al., 2007; Xia et al., 2008; Chen et al., 2012, 2013b; Zhang et al., 2015), and migmatitic leucosomes and granitic pegmatites in UHP metamorphic rocks (e.g., Wallis et al., 2005; Liu et al., 2010a; Zong et al., 2010; Zhao et al., 2012; Song et al., 2014;
NU
Wang et al., 2014). Dehydration melting of UHP metamorphic rocks commonly produces felsic melts with very small ratios of melt to residue (Li et al., 2013; Zheng and Hermann,
MA
2014). Such melts are referred to as anatectic melts and are different from differentiated magmatic melts, in that they did not completely separate from their parent metamorphic rocks and experienced little crystal fractionation (Li et al., 2013; Zheng and Hermann, 2014). In
D
contrast, magmatic melts have completely separated from their parental rocks, represent
TE
mixing of different batches of anatectic melts, and have often experienced significant chemical differentiation during their transport upward. The composition of anatectic melts
CE P
varies dramatically depending on source mineralogy and fluid regime (Clemens and Droop, 1998; Stevens et al., 1997, 2007; Zheng and Hermann, 2014), which exerts a strong control on the composition of peritectic and anatectic minerals. For instance, zircon grown from
AC
peritectic reaction during continental exhumation can be distinguished from that grown from magmatic melts and metamorphic fluids in terms of mineral inclusions, U-Pb ages and trace element compositions (e.g., Xia et al., 2010; Liu et al., 2010a; Chen et al., 2013b). Garnet is a very useful major phase in metamorphic rocks. Its stability over a wide range of temperatures and pressures permits its use in remarkably diverse tectonic settings and rock types (e.g., Baxter and Scherer, 2013). It can preserve strong zonation in major and trace elements as well as mineral inclusions, which bears information on re-equilibration with coexisting minerals during high-grade metamorphism (e.g., Skora et al., 2006; Konrad-Schmolke et al., 2008a, b; Dragovic et al., 2012; Baxter and Caddick, 2013). Thus garnet plays an important role in revealing the evolution of subduction-zone processes, including the pressure, temperature and duration of metasomatism, deformation and anatexis (e.g., Angiboust et al., 2014; Caddick et al., 2010; Caddick and Kohn, 2013; Dragovic et al., 2015; Habler et al., 2007; Martin et al., 2014; Pollington and Baxter, 2010; Storey and Prior,
3
ACCEPTED MANUSCRIPT 2005). As a major phase in UHP metamorphic rocks, metamorphic garnet is unavoidably modified when anatectic melting has taken place (e.g., Perchuck et al., 2005, 2008; Liu et al., 2014). Some garnet may be of peritectic origin in the presence of anatectic melts. However,
T
the characteristic features of peritectic garnets that allow them to be distinguished from
IP
metamorphic or magmatic garnets in UHP metamorphic rocks are still uncertain. Identifying such features that constrain the origin of garnets in UHP metamorphic rocks is critical to the
SC R
P-T-t reconstruction of deeply subducted crustal rocks.
In this paper, we combine petrology and geochemistry in garnet and related minerals from UHP metagranite (granitic orthogneiss) in the Dabie orogen to evaluate the origin of different
NU
types of garnet. The metagranite is a direct host of UHP eclogites and biotite paragneiss in this region. Partial melting has been identified in the eclogitic garnet based on MS inclusions
MA
with radial cracks (Gao et al., 2012a, 2013; Liu et al., 2013). It is possible that partial melting also occurred in the country rock metagranites. Garnet porphyroblasts from the UHP metagranite show evident core-rim zoning or patchy structures in back scattered electron
D
(BSE) images, major and trace element profiles. These observations provide a sound basis for
TE
distinction between garnets of different origins in the UHP granitic gneiss, with insights into
CE P
the P-T-t regime of deeply subducted continental crust in the continental collision orogen.
2. Geological setting
The Dabie-Sulu orogenic belt is located in east-central China (Fig. 1), and was built by the
AC
continental collision between the North China Block and the South China Block in the Triassic (e.g., Wang et al., 1995; Cong, 1996; Liou et al., 1996; Li et al., 1999; Zheng et al., 2003, 2009). The orogenic belt is separated into eastern and western segments by the Tanlu Fault, and are referred to as the Sulu and Dabie orogens, respectively (Fig. 1a). The Dabie orogen extends across the Anhui, Henan and Hubei provinces in the west. It consists of a series of fault-bounded metamorphic units that can be subdivided into five main tectonic zones from north to south (Zheng et al., 2005a): (1) the Beihuaiyang low-T/low-P greenschist-facies zone, (2) the North Dabie high-T/UHP granulite-facies zone, (3) the Central Dabie mid-T/UHP eclogite-facies zone, (4) the South Dabie low-T/UHP eclogite-facies zone, and (5) the Susong low-T/high-P blueschist-facies zone.
The eclogite occurs not only as small lenses and enclaves in the granitic orthogneiss in the southern unit, but also as large-scale blocks interlayered with the granitic orthogneiss, biotite
4
ACCEPTED MANUSCRIPT paragneiss or marble in the northern unit (Fig. 1b). Coesite and its pseudomorphs are observed as inclusions in garnet, jadeite and dolomite from the eclogites, marbles and surrounding gneisses (e.g. Gao et al., 2012a; Liou et al., 1997; Liu et al., 2001; Wang et al.,
T
1989; Wang and Liou, 1991; Schertl and Okay, 1994). Petrological studies suggest that the
IP
peak metamorphic pressures for the UHP eclogites are greater than 3.0 GPa at temperatures of 700 to 800 °C, based on garnet-clinopyroxene Fe-Mg partitioning geothermometry (Okay,
SC R
1993; Cong et al., 1995; Carswell et al., 1997). The granitic gneiss was re-equilibrated under P-T conditions of 400±50oC and 400 MPa (Cong et al., 1995; Liu et al., 2003), corresponding to retrograde metamorphism at low amphibolite-facies. However, several observations
NU
indicate that the granitic orthogneiss shares the same UHP metamorphic evolution as the eclogite-paragneiss units that are enclosed by it. Carswell et al. (2000) reported the occurrence
MA
of rutile inclusions in titanite, high-Ca garnet (XCa up to 0.49) and high-Si phengite (up to 3.49) in zoisite from the granitic orthogneiss at this locality. In addition, Liu et al. (2001) discovered coesite inclusions in zircon from all types of para- and orthogneisses in the central
D
Dabie zone. Zircon U-Pb dating for the granitic orthogneiss also yields two groups of Triassic
TE
ages of ~240 Ma and ~220 Ma (e.g., Zheng et al., 2005b). A detailed description of the eclogite, gneiss and jadeite-quartzite at Shuanghe has been
CE P
made by Okay (1993), Cong et al. (1995) and Liou et al. (1997). The UHP eclogite slice at Shuanghe is separated into two units by a dextral strike-slip fault (Fig. 1b). Both units, with outcrop area about 1 km2, are surrounded by the UHP granitic orthogneiss. The present study deals with garnet-bearing granitic orthogneiss at Shuanghe. Three samples (95M16, 12DB37
AC
and 12DB38) were collected from the granitic orthogneiss in the northern unit, whereas another two samples (95M18 and 95M19) were collected from the granitic orthogneiss in the southern unit (Fig. 1b). The five granitic gneisses have the similar mineral assemblages and modal abundances. They are mainly composed of plagioclase, K-feldspar and quartz with minor amounts of garnet, muscovite, biotite, titanite and epidote/allanite (Fig. 2). The detailed modal abundances (%) for minerals in granitic gneisses are listed in Table 1. However, detailed observations under the microscope show that the granitic gneiss in the northern unit contains fewer hydrous minerals such as epidote, muscovite and biotite than those in the southern unit (Table 1).
Mineral abbreviations used in this paper are after Whitney and Evans (2010). Abbreviations for trace elements follow the common conventions: light rare earth elements (LREE), middle 5
ACCEPTED MANUSCRIPT rare earth elements (MREE) and heavy rare earth elements (HREE).
3. Analytical methods
T
3.1 Whole-rock major elements
IP
The whole-rock major elements were analyzed at ALS Mineral/ALS Chemex Co. Ltd. at Guangzhou, using XRF spectrometry. Fused glass disks with lithium borate were used and
SC R
repeated analyses of the standards GSR-2 and GSR-3 indicate that the analytical precisions were better than ±0.01%.
NU
3.2 Major and trace elements in garnet and other minerals
Garnet and mineral inclusions therein from the target samples were carefully examined in
MA
thin sections under a microscope and by Raman spectroscopy before major and trace element analyses were obtained. The laser Raman analysis was made on a Renishaw RM 2000 Raman spectrometer at CAS Key Laboratory of Crust-Mantle Materials and Environments in
D
University of Science and Technology of China, Hefei. The beam size for Raman
TE
spectroscopy was ~2 μm.
Major element mapping of garnet was undertaken using the Fei Quanta 450 FEG scanning
CE P
electron microscope (SEM) at State Key Laboratory of Geological Processes and Mineral Resources in China University of Geosciences, Wuhan, based on energy dispersive spectrometry (EDS) using an Oxford Inca X-Max 50 silicon drift detector (SDD). The
AC
measurements were carried out with an accelerating voltage of 20 kV, a spot size of 6.0 μm and a working distance of 12 mm. Major element analyses were made for some garnet, titanite, allanite, epidote, micas and feldspar using a JEOL JXA-8100 electron microprobe (EMP) at State Key Laboratory of Mineral Deposits Research in Nanjing University, Nanjing. Major elements for other garnets were also analyzed on a JEOL JXA-8230 EMP at Hefei University of Technology, Hefei. The working conditions were set at 15 kV of accelerating voltage and a beam current of 2×10-8 A, with a beam size of 1-5 μm. Each element includes a background acquisition of ~ 10-20s (gas blank) followed by 20s data acquisition from the sample. The precision for most major elements (with atomic numbers greater than 10 and concentrations higher than 10%) are better than ±2% (1σ). Trace elements in garnet and titanite were analyzed by means of the laser ablation-inductively coupled plasma mass spectrometer (LA-ICPMS) at State Key Laboratory
6
ACCEPTED MANUSCRIPT of Geological Processes and Mineral Resources in China University of Geosciences, Wuhan. These analyses were made on the same garnet grains that had been analyzed for major elements by EMP, and BSE images were used to efficiently avoid cracks and mineral
T
inclusions. Detailed operating conditions for the laser ablation system and the ICP-MS
IP
instrument and data reduction are the same as described by Liu et al. (2008). Laser sampling was performed using a GeoLas 2005 with a diameter of ~ 24 or 44 μm depending on the grain
SC R
size of the target mineral, to a depth of ~10-20 μm. The laser frequency was 8 Hz and the energy density was 14 J·cm-2. An Agilent 7500a ICP-MS instrument was used to acquire ion-signal intensities. Helium was applied as a carrier gas. Argon was used as the make-up
NU
gas and mixed with the carrier gas via a T-connector before entering the ICP. Nitrogen was added into the central gas flow (Ar+He) of the Ar plasma to decrease the detection limit and
MA
improve precision (Hu et al., 2008).
Each analysis incorporated a background acquisition of approximately 20s (gas blank) followed by 50s data acquisition from the sample. The Agilent Chemstation was utilized for
D
the acquisition of each individual analysis. Off-line selection and integration of background
TE
and analyzed signals, and time-drift correction and quantitative calibration were performed by ICPMSDataCal (Liu et al., 2008, 2010b). Trace element concentrations were calibrated by
CE P
using 29Si as an internal calibrant and NIST SRM610 as the reference material. The precision and accuracy of the NIST-610 analyses are ±2-5% for most elements at the ppm concentration level. Element contents were calibrated against multiple reference materials (BCR-2G, BIR-1G and BHVO-2G) without applying internal standardization (Liu et al., 2008). The
AC
preferred values of element concentrations for the USGS reference glasses are from the GeoReM database (http://georem.mpch-mainz.gwdg.de/). Analyses of USGS rock standards (BCR-2G and BHVO-2G) indicate that the precision and accuracy (1σ) are better than ±10% for trace elements, and ±2% for major elements. Limits of detection (LOD) for these USGS standards were described in detail by Gao et al. (2002). LOD for each element and analysis were calculated individually as three times the standard deviation of the background signal (taken before ablation) divided by element sensitivity during the respective ablation. For non-significant analytical signals, LOD values are also reported (marked by < LOD). It has to be mentioned that the area and depth of laser ablation are much bigger than the volume sampled by the EMP analysis. Hence, the major element concentrations determined by EMP and LA-ICPMS are not exactly the same. Because the EMP analyses have better spatial resolution, and also may have higher precision than the LA-ICPMS analyses, the EMP results are used in this study for major element compositions. 7
ACCEPTED MANUSCRIPT
3.3 Zircon U-Pb ages and trace elements Zircon U-Pb dating and trace element analyses were simultaneously performed by the
T
LA-ICPMS in-situ method at State Key Laboratory of Geological Processes and Mineral
IP
Resources in China University of Geosciences, Wuhan. Laser ablation sampling was performed using a Geolas 2005 system equipped with a 193 nm ArF-excimer laser. An
SC R
Agilent 7500a ICP-MS was used to acquire ion-signal intensities. Detailed instrumental conditions and data acquisition were described by Liu et al. (2010b, 2010c) and Zong et al. (2010). For zircon trace element and U-Pb isotope analyses, the blank was very low because
NU
high purity argon and helium was used. The ICP-MS measurements were carried out by using time-resolved analysis and peak hopping at one point per mass and the dwell time for each isotope was set at 6 ms for Si, Ti, Nb, Ta, Zr and REE, 15 ms for 204Pb, 206Pb, 207Pb and 208Pb,
MA
and 10 ms for 232Th and 238U. Each spot analysis includes 20s of background acquisition and 40s sample data acquisition. Trace element concentrations were calibrated by using 29Si as an
D
internal calibrant and NIST SRM610 as a reference material. The precision and accuracy of
TE
NIST-610 analyses are ±2-5% for most elements at the ppm concentration level. Isotope ratios, including 207Pb/206Pb, 206Pb/238U, 207Pb/235U and 208Pb/232Th, were calculated
CE P
using Glitter 4.0 software (Macquarie University), which were then corrected using the zircon 91500 as an external calibrant with a
206
Pb/238U age of 1065.4±0.6 Ma (Wiedenbeck et al.,
1995). All the measured isotope ratios of zircon 91500 were regressed over the course of the analytical session and used to calculate correction factors. These correction factors were then
AC
applied to each sample in order to correct both instrumental mass bias and depth-dependent elemental and isotopic fractionation. The common Pb correction was carried out by using the EXCEL program of ComPbCorr#_151 (Andersen, 2002), assuming that the observed 206
Pb/238U, 207Pb/235U and
208
Pb/232Th ratios for a discordant zircon can be accounted for by a
lead loss at a defined time. Apparent and discordant U-Pb ages were calculated by the ISOPLOT program (Ludwig, 2003).
4. Results 4.1 Petrography The occurrence of quartz and feldspar in the target granitic gneiss shows petrographic evidence for partial melting. In most cases the two minerals are found in coarse-grained, xenomorphic and hypidiomorphic blocks. In some cases, plagioclase occurs as veinlets along
8
ACCEPTED MANUSCRIPT the quartz-quartz and quartz-feldspar grain boundaries, showing low feldspar-quartz-quartz dihedral angles (Fig. 2a). Quartz is also fine-grained, or occurs as elongate polycrystalline aggregates filling in plagioclase-plagioclase grain boundaries (Fig. 2c). Some highly cuspate
T
or vein-like quartz is also found to occur along plagioclase-biotite or plagioclase-plagioclase
IP
grain boundaries (Fig. 2b and c). EMP analyses show that plagioclase in the studied orthogneiss is generally Na-rich, with variable but generally low Ca contents (Table 2).
SC R
Garnet grains in samples 12DB37, 12DB38 and 95M16 from the northern unit are idioblastic and euhedral with hexagonal prophyroblasts (Figs. 2d-e and 3a-c), whereas garnets in samples 95M18 and 95M19 from the southern unit occur as islands or peninsulas (Figs. 2f,
NU
2h and 3d-f). Quartz, K-feldspar and plagioclase are the most common mineral inclusions in the garnet porphyroblasts (Fig. 3a and f). Multiple phase inclusions of albite+K-feldspar are
MA
also occasionally found in the garnet (Fig. 3f). Fractures filled with later chlorite and carbonates are common in garnet (Fig. 4d). Most garnet grains are partially replaced by epidote, biotite and plagioclase (Fig. 2e and g).
D
TE
Epidote/zoisite also occurs as large prophyroblasts (>500 µm) or small inclusions (500 µm) (Fig. 2g). Polygenetic titanites have been identified in sample 95M19 by Gao et al. (2012b), in which magmatic and metamorphic origin were distinguished using BSE images, mineral inclusions, major and trace elements and U-Pb ages. The metamorphic
AC
titanite was explained as the product of retrograde metamorphism during exhumation.
4.2 Major and trace element zoning in garnet Major element X-ray mapping was made on two garnet grains in samples 12DB37 and 12DB38 (Fig. 4). Major element profiles were measured on six grains in five samples (Figs. 3 and 5; supplementary Table S1). BSE images show that these garnets have patchy microstructures (Figs. 3 and 4). Major element X-ray maps and profiles indicate that garnet crystals are strongly zoned in Ca and Fe, and weakly zoned or unzoned in Mn and Mg (Figs. 4 and 5). For instance, the contents of almandine vary from 20.3 to 39.3 mol.% and grossular from 20.7 to 48.0 mol.% for garnet grain 12DB38-2, but pyrope contents only vary from 0.20 to 1.29 mol.% (Table S1; Fig. 5c). Major element profiles on the six garnet grains show that the dark and bright BSE domains have distinct compositions, which are denoted as Grt-I and
9
ACCEPTED MANUSCRIPT Grt-II, respectively (Figs. 3-6). The Grt-I domains occur as a broad domain in the center, with low almandine and pyrope concentrations, and high grossular concentrations. In contrast, the Grt-II domains occur as rims of euhedral garnet, or as patches within the Grt-I domains. They
T
have relatively high almandine and pyrope concentrations and low grossular concentrations.
IP
Spessartine in the two garnet grains from samples 12DB37 and 12DB38 is weakly zoned (Fig. 4b and c). However, spessartine in the other garnets from samples 95M16, 95M18 and 95M19
SC R
are strongly zoned, with high contents in the Grt-II domains (Figs. 5a, d, e and f).
Trace elements analyses show that Grt-I and Grt-II domains also have much different
NU
compositions, except in grain 12DB38-2 (Figs. 7 and 8). The Grt-I domains have high Ti (342-939 ppm) and MREE (119-296 ppm) concentrations, and are low in Sc (4.94-32.8 ppm),
MA
Cr (0-2.72 ppm), Y (948-3901 ppm) and HREE (335-718 ppm) (Table 3; Fig. 7). The chondrite-normalized REE patterns show increasingly enriched LREE to MREE, negative Eu anomalies with Eu/Eu* = 0.22-0.52, and flat MREE-HREE distributions with (Yb/Dy)N =
D
1.68-2.86 (Fig. 8). In contrast, the Grt-II domains have low Ti (132-515 ppm) and MREE
TE
(78-415 ppm) concentrations, and high Sc (12.4-141 ppm), Cr (0-10.6 ppm), Y (1079-5004 ppm) and HREE (763-3044 ppm) (Table 3; Fig. 7). The chondrite-normalized REE patterns
CE P
show depletion in LREE, negative Eu anomalies with Eu/Eu* = 0.09-0.60, and steep middle to heavy REE patterns with (Yb/Dy)N = 2.40 to 12.5 (Fig. 8). Both domains exhibit obviously negative Eu anomalies with Eu/Eu* = 0.09-0.60. For garnet grain 12DB38-2, Grt-I and Grt-II domains have comparable trace element compositions. There is no obvious distinction
AC
between the two domains in trace elements such as Sc, Ti, Cr, Y and REE (Table 3; Fig. 7).
4.3 Zircon U-Pb ages and trace elements Detailed U-Pb, Sm-Nd and Rb-Sr geochronological studies have been published for the UHP granitic gneiss at Shuanghe (e.g., Li et al., 2000; Chavagnac et al., 2001; Ayers et al., 2002; Zheng et al., 2005b). In this study, we have only analyzed some zircon grains from three samples: 12DB37, 12DB38 and 95M19. These zircon grains are short to long prismatic, light yellow and translucent, with lengths of 100 to 200 μm. They are euhedral, and display obvious core-rim structure (Fig. 8a, d and g). CL images show that most grains have large cores with obscured oscillatory zoning and very thin rims without zonation. Raman analysis on these zircon grains indicates the presence of many separate mineral inclusions (e.g., apatite, quartz, feldspar and muscovite) and MS inclusions (e.g., Qtz+Kfs, Qtz+Pl, Mus+Kfs) in the 10
ACCEPTED MANUSCRIPT core, but only a few grains of quartz and apatite in the rim. Forty-eight LA-ICPMS analyses were made on zircon core and/or rim from the three metagranites for U-Pb ages and trace elements (Tables 4 and 5). The U-Pb dating yields
T
similar intercept ages, with upper intercept ages of 771±84 to 809±72 Ma and lower intercept
IP
ages of 215±29 Ma to 255±140 Ma (Fig. 9b, e and h). However, there are different age distributions for the three samples. Most zircon from the northern unit orthogneiss (12DB37
SC R
and 12DB38) exhibits discordant U-Pb ages of 316 to 809 Ma. Concordant Triassic ages of 234 to 257 Ma are only obtained from one spot (#20) among eighteen analyses from sample 12DB37 and two spots (#7 and #14) among sixteen analyses from sample 12DB38. On the
NU
other hand, seven spots among fourteen analyses from the southern unit orthogneiss (95M19) yield concordant Triassic ages of 224 to 254 Ma (Fig. 9).