Geoscience Frontiers 6 (2015) 759e770

HOSTED BY Contents lists available at ScienceDirect

China University of Geosciences (Beijing)

Geoscience Frontiers

journal homepage: www.elsevier .com/locate/gsf

Research paper

Retrograde metamorphism of the eclogite in North Qaidam, westernChina: Constraints by joint 40Ar/39Ar in vacuo crushing and steppedheating

Rongguo Hu a,b,c,*, Jan Wijbrans b, Fraukje Brouwer b, Linghao Zhao a, Min Wang a,c,Huaning Qiu a,**

a State Key Laboratory of Isotope Geochemistry, Guangzhou Institute of Geochemistry, Chinese Academy of Sciences, P.O. Box 1131,Wushan, Guangzhou 510640, PR ChinabDepartment of Earth Sciences, VU University Amsterdam, De Boelelaan 1085, 1081 HV Amsterdam, The NetherlandscUniversity of Chinese Academy of Sciences, Beijing 100049, PR China

a r t i c l e i n f o

Article history:Received 18 February 2014Received in revised form28 August 2014Accepted 11 September 2014Available online 11 November 2014

Keywords:40Ar/39Ar techniquesIn vacuo crushingAmphibolite-facies retrogressionFluid flowNorth Qaidam

* Corresponding author. State Key Laboratory of IsotoInstitute of Geochemistry, Chinese Academy of ScienGuangzhou 510640, PR China. Tel.: þ86 20 85290696,fax: þ86 20 85290130.** Corresponding author.E-mail address: rongguo.hu@gmail.com (R. Hu).Peer-review under responsibility of China University

http://dx.doi.org/10.1016/j.gsf.2014.09.0051674-9871/� 2015, China University of Geosciences (BND license (http://creativecommons.org/licenses/by-n

a b s t r a c t

Two amphiboles and a syn-metamorphic quartz vein from the Yuka terrane, North Qaidam, westernChina, have been analyzed by joint 40Ar/39Ar crushing in vacuo and stepwise heating techniques. Thecrushing in vacuo results provide information to directly constrain the timing of fluid activity and the ageof amphibolite-facies retrogression. The stepwise heating results could further be used to decipher thethermal history of the UHP rocks. Amphiboles from amphibolites and quartz vein within garnet-amphibolite lens analyzed by in vacuo crushing yield similarly shaped age spectra and exhibit rela-tively flat age plateaus for the last several steps. The characteristics of gas release patterns andgeochronological data testify to the presence of significant excess 40Ar within the fluid inclusions. Theage plateaux with weighted mean ages (WMA) ranges from 488 to 476 Ma for amphiboles and 403 Mafor quartz (2s). These data points constitute amphibole WMA yielding excellent isochrons with isochronages of 469 and 463 Ma with initial 40Ar/36Ar ratios of 520 and 334, respectively. The isochron ages areinterpreted to represent initial amphibolite-facies retrogression. The data points constituting the quartzage plateaux give an isochron age of 405 Ma with initial 40Ar/36Ar ratio of 295, recording a significantaqueous fluid flow episode during the early Devonian. Age spectra obtained by stepwise heating ofamphibole residues remaining after crushing experiments are characterized by younger and relativelycomplex age spectra, which are probably influenced by the combined effects of resetting argon and/ormineral inclusions. Nevertheless, we note that the spectra shapes have features in common: excludingthe last two steps, minimum apparent ages are found at temperatures of around 500 �C, corresponding to319 and 249 Ma, perhaps representing the time of isotopic resetting or resulting from release gas frommineral inclusions of, e.g., biotite or feldspar. Maximum apparent ages are obtained at temperatures ofaround 800 �C, corresponding to 418 and 413 Ma, which probably reflect mixed ages of radiogenicresetting and original amphibole. These results indicate that the Yuka eclogites and their retrogressedequivalents were overprinted by multiple thermal events in the Silurian and possibly as young as theTriassic.

� 2015, China University of Geosciences (Beijing) and Peking University. Production and hosting byElsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/

licenses/by-nc-nd/3.0/).

pe Geochemistry, Guangzhouces, P.O. Box 1131, Wushan,þ86 189 2404 4095 (mobile);

of Geosciences (Beijing).

eijing) and Peking University. Produc-nd/3.0/).

1. Introduction

Although it is generally accepted that continental subduction ischaracterized by a scarcity of fluidwhen comparedwith subductionof oceanic crust, a considerable amount of aqueous fluid can still bereleased. A significant and abrupt decrease in either temperature orpressure along the retrograde part of the P-T loop may trigger

ction and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-

R. Hu et al. / Geoscience Frontiers 6 (2015) 759e770760

recrystallization of hydrous and hydroxylated minerals, and thedecrepitation of primary fluid inclusions. As a result, amphiboliteand greenschist facies retrogression of eclogites and granulites,formation of quartz veins and even syn-exhumation magmatismare likely to be caused by aqueous fluid flow during exhumation ofdeeply-subducted continental crust (Baker et al., 1997; Franz et al.,2001; Fu et al., 2003; Zheng, 2004). Several groups have studied thefluid inclusions and petrological phase relations of HP/UHP meta-morphic rocks to constrain the origin, the volume and theelemental and hydrogen and oxygen isotopic composition, of thesemetamorphic fluids (e.g., Li et al., 2001; Touret, 2001; Xiao et al.,2002; Fu et al., 2003; Zheng et al., 2003). These studies havegreatly improved our understanding of the nature and effect offluid activity during exhumation of deeply-subducted continentalcrust.

In comparison with research on the chemistry and isotopiccomposition of metamorphic fluids, it is more difficult to constrainthe timing of fluid infiltration during retrogression of UHP meta-morphic rocks. Rather than attempting to indirectly date the quartzsample by using, e.g., the RbeSr isotopic decay system (e.g., Wanget al., 2000, 2003) or to directly date metamorphic minerals suchas garnet and amphibole by extracting gas from fluid inclusions for40Ar/39Ar dating (e.g., Qiu and Wijbrans, 2006, 2008; Qiu et al.,2010), perhaps the greatest insights into the age of fluid activityin UHP rocks have been attained from studies of accessory mineralslike zircon and rutile associated with quartz veins in eclogite-faciesrocks (e.g., Gao et al., 2006; Zheng et al., 2007;Wu et al., 2009; Zonget al., 2010; Wang et al., 2011; Chen et al., 2012).

The 40Ar/39Ar in vacuo crushing method has been utilized inseveral studies of fluid inclusions-rich light-colored rocks andminerals such as chert (Alexander, 1975; Wang et al., 1988), veinquartz (Kelley et al., 1986; Turner, 1988; Qiu, 1996; Kendrick et al.,2001, 2006) and feldspar (Burgess et al., 1992; Turner and Wang,1992; Harrison et al., 1993; Burgess and Parsons, 1994), with theaim of investigating either the paleo-atmospheric 40Ar/36Ar ratio orthe source and halogen chemistry of fluid inclusions. In addition tothat, sulphide minerals (e.g., pyrite and sphalerite) (Phillips andMiller, 2006; Qiu and Jiang, 2007; Jiang et al., 2012), volatile-richmetamorphic minerals from hydrothermal deposits (e.g., scapo-lite) (Kendrick and Phillips, 2009b), and low-potassium mineralsfrom HP/UHP metamorphic rocks (e.g., garnet) (Qiu and Wijbrans,2006, 2008) have been analyzed with some success using thisdating method. In a critique of this method, Kendrick and Phillips(2009b) argued that a significant amount of lattice-hosted noblegas from scapolite can be released during prolonged crushing, but anumber of studies have indicated that crushing has very little effecton the gas component that is trapped within the crystal lattice(Dunlap and Kronenberg, 2001; Qiu et al., 2002; Qiu and Jiang,2007; Qiu and Wijbrans, 2008; Qiu et al., 2010; Jiang et al., 2012;Bai et al., 2013). This suggests that a combination of 40Ar/39Ardating during progressive crushing and stepwise heating of thecrushed powder could provide an opportunity to distinguish thegas components trapped inside fluid inclusions and on the grainboundaries of solid inclusions from the component of the gaswithin the crystal lattice (Villa, 2001).

Here, we report a 40Ar/39Ar study of three samples from theYuka terrane, north Qaidam, western China. By applying the invacuo crushing and stepwise heating 40Ar/39Ar techniques, we havebeen able to extract trapped argon from fluid inclusions and, grainboundaries of solid inclusions, as well as from the crystal lattice ofthe metamorphic amphiboles and quartz. We use Ar isotope sys-tematic studies to: (1) constrain the timing of fluid flow andamphibolite-facies retrogression of Yuka eclogite; (2) reveal thethermal history of these UHP rocks; and (3) identify extraneous40Ar when they are present.

2. Geological setting

The north Qaidam orogenic belt, western China, was created byearly Paleozoic deep subduction of the South Qilian Sea and theQaidam continental crust beneath the Qilian block (Xu et al., 2006;Xiao et al., 2009). It is bounded by the Qilian block in the north andthe Qaidam block in the south, and extends NWW-striking for400 km long from the Altyn-Tagh Fault in the northwest to theWulan Fault in the southeast (Fig. 1a and b). Based on field re-lationships, petrology and rock association, the belt can be dividedinto four HP/UHP metamorphic units. From east to west they are:the Dulan eclogite-gneiss unit, the Xitieshan eclogite-gneiss unit,the Luliangshan garnet peridotite unit, and the Yuka eclogite-gneissunit, which are separated by Paleozoic to Cenozoic sediments.Meanwhile, the occurrence of diamond-bearing garnet peridotite(Yang et al., 1994; Song et al., 2005b), coesite-bearing eclogite andpelitic gneiss (Yang et al., 2001; Song et al., 2003b; Zhang et al.,2009a) imply that the HP/UHP rocks were formed at depthsaround w100 km, or even deeper than 200 km.

During the past ten years, considerable and significant studiesaddressing petrology, major and trace element geochemistry,geochronology and geotectonics have been carried out on the belt.These works have shed light on the protolith nature of the deeplysubducted crust involved (Yang et al., 2006; Song et al., 2007b,2010), the timing, timescale and spatial distribution of the UHPmetamorphism (Song et al., 2003a; Mattinson et al., 2006, 2009;Chen et al., 2007a, b; Zhang et al., 2008; Zhang et al., 2011a, b),the possible depth of continental subduction (Song et al., 2005a,2009a; Zhang et al., 2009a), the mechanisms of subductionand exhumation (Yang et al., 2002, 2009; Xu et al., 2006; Songet al., 2009b; Xu et al., 2010), and the pressure-temperature-time(P-T-t) histories constraining the tectono-metamorphic processesof different units in this region (Chen et al., 2005; Zhang et al.,2005, 2007; Song et al., 2007a; Menold et al., 2009; Zhang et al.,2009b).

The present study area in the Yuka eclogite-gneiss unit is locatedin the westernmost segment of the north Qaidam HP/UHP meta-morphic belt, 40 km north of Da Qaidam town (Fig. 1b). It mainlyconsists of a variety of gneisses, with several eclogite andamphibolite lenses or blocks up to a fewmeters across cropping outalong the north bank of the Yuka River (Fig. 1c). Two main types ofeclogite have been recognized within the granitic and peliticgneisses in this area, coarse-grained phengite eclogite and fine-grained massive eclogite (Chen et al., 2005). The country rockmainly consists of two types of gneisses. Muscovite-bearing gneissconsists of feldspar, quartz and muscovite with minor garnet, cli-nozoisite and zircon, whilst amphibole-bearing gneiss is composedof garnet, phengite, amphibole, plagioclase, and minor rutile andquartz (Zhang et al., 2005; Chen et al., 2009).

Most previous geochronological data obtained by variousmethods have concentrated on constraining the peak eclogite-facies metamorphism and cooling ages of the UHP rocks. Eclogitezircon dated using the TIMS UePb method gave what was inter-preted to be an eclogite-facies metamorphic age of 488 � 6 Ma(Zhang et al., 2005). In contrast, zircon in situ LA-ICP-MS UePbdating of the eclogite yields a mean weighted 206Pb/238U age of436 � 3 Ma for eclogite-facies metamorphism (Chen et al., 2009).40Ar/39Ar measurements on amphibole and phengite concentratesfrom the eclogite gave isochron ages of 477 � 8 and 466 � 5 forcooling ages, respectively, and their isotope correlation diagramsindicated no extraneous 40Ar within these minerals (Zhang et al.,2005). However, by using the 40Ar/39Ar incremental heatingmethod, Menold (2006) noticed that eclogite phengites yieldedanomalously old ages ranging from 611 to 750 Ma, implying thepresence of extraneous 40Ar, i.e. inherited or excess 40Ar.

Figure 1. (a) Geological sketch map showing the locality of north Qaidam orogen in China. (b) Distribution of UHP metamorphic terranes in the north Qaidam orogenic belt.YKUeYuka eclogite-gneiss unit; LLUeLuliangshan garnet peridotite-granulite unit; XTUeXitieshan eclogite-gneiss unit; DLUeDulan eclogite-gneiss unit. (c) Sketch map of Yukaterrane and sample localities (Modified after Zhang et al., 2005).

R. Hu et al. / Geoscience Frontiers 6 (2015) 759e770 761

The mineral assemblages, compositions of minerals, and ther-mometric data indicate that the eclogite-gneiss rocks were meta-morphosed at pressures of 2.3e3.4 GPa and temperatures of600e720 �C (Zhang et al., 2003; Chen et al., 2005). Three distinctmetamorphic stages have been recognized for the eclogites in theYuka unit: (1) a pre-eclogitic metamorphic stage, recorded by agarnet þ amphibole þ zoisite þ plagioclase þ quartz inclusionassemblage preserved in the cores of garnet porphyroblasts,yielding metamorphic temperatures of w450e577 �C at0.6e1.2 GPa; (2) a coesite-bearing eclogite facies stage, recorded bythe assemblage garnet þ omphacite þ phengite þ quartz/coesite þ rutile preserved as inclusions in the rims of garnet por-phyroblasts, yielding metamorphic temperatures of w566e720 �Cand pressure estimates of 2.35e3.4 GPa; and (3) a post-eclogiteretrograde stage recorded by garnet reaction rims, the disappear-ance of omphacite, and the pervasive retrograde mineral assem-blage of garnet þ amphibole þ plagioclase þ quartz, which yieldsmetamorphic temperatures of 550e630 �C and 0.6e1.1 GPa (Chenet al., 2005; Zhang et al., 2005, 2009a).

3. Petrology of samples used for dating

The samples used in this study for 40Ar/39Ar dating are twogarnet-amphibolites (09NQ13Amp and 09NQ29Amp) and onequartz vein within garnet amphibolite (09NQ16Qz) collected fromoutcrops on the northern bank of the Yuka River, Yuka terrane(Fig. 1c). The mineral abbreviations in this paper follow theconvention of Whitney and Evans (2010).

Sample 09NQ13Amp was taken from a 1 m � 0.6 m lens-shapedamphibolite block hosted by muscovite schist (Fig. 2a). It is char-acterized by medium to coarse-grained massive structure, andconsists of amphibole (w60%), clinozoisite (w10%), epidote(w10%), plagioclase (w5%), quartz (w5%), garnet (w5%), sym-plectites (Amp þ Pl w5%) with minor titanite (Fig. 3a). Amphibolecrystals are lepidoblastic with variable grain size from 1 to 5 mmwith biotite, plagioclase and quartz as mineral inclusions.

Amphiboles are classified as edenite andmagnesio-hornblende (c.f.Leake et al., 1997) with K/Ca ratios ranging from 0.08 to 0.11 and amean value of 0.1 (see Table 1).

Sample 09NQ29Amp is obtained from a garnet amphibolite lensand displays a gradual transition from eclogite-facies toamphibolite-facies mineralogy, with fresh eclogite preserved in thecore (Fig. 2c). It is composed of amphibole (w60%), symplectites(Amp þ Pl, w10%), garnet (w10%), epidote (w10%) and plagioclase(w10%), with minor chlorite, titanite and biotite (Fig. 3b). Garnetcrystals display as subhedral granular texture with grain sizes from0.5 to 1 mm. Amphibole, which is characterized by intense alter-ation along the rim, shows lepidoblastic texture with grain sizesfrom 1 to 3 mm and normally contains mineral inclusions such asgarnet, chlorite, biotite and plagioclase. Amphiboles are classifiedas winchite, barroisite, magnesio-hornblende and tschermakite (c.f.Leake et al., 1997) with K/Ca ratios ranging from 0.01 to 0.04 with amean value of 0.03 (see Table 1).

Vein sample 09NQ16Qz occurs as a thin-sheet shaped quartzlens of 6 cm � 8 cm in size, coexisting with phengite and amphi-boles on the surface of a retrograde eclogite block (Fig. 2b, d).

Preliminary petrographic observations have been carried out onpolished quartz and amphibole thin sections (Fig. 3c and d). Most ofthe fluid inclusions in the amphiboles are primary and small in size(ca. 0.5e3 mm), and are randomly distribution and characterized byirregular, tubular and elongated shapes (Fig. 3c). Secondary fluidinclusions in amphibole were identified in healed fractures cross-cutting grain boundaries and growth zones, and generally biggerthan primary types. Fluid inclusions in quartz samples are veryabundant which are 5e20 mm in size and displaying negativecrystal morphologies and observed as isolated or random distri-bution (Fig. 3d).

4. Experimental techniques

Samples were crushed and sieved to obtain a size fraction of250e500 mm. After magnetic separation, heavy liquid techniques

Figure 2. Field views showing occurrence and relationship of garnet amphibolite, quartz vein and associated country rocks in Yuka terrane, western China. (a) Garnet amphibolitelens (09NQ13Amp) in muscovite schist. (b and d) Garnet amphibolite block contains thin-sheet shaped phengite-bearing vein quartz (09NQ16Qz). (c) Fresh eclogite preserved in thecore of garnet amphibolite (09NQ29Amp). The length of the hammer used in (a) is 41 cm; the pen in (b) and (c) is 14 cm; the diameter of the coin used in (d) is 2 cm.

R. Hu et al. / Geoscience Frontiers 6 (2015) 759e770762

were applied to separate amphibole (3.1e3.3 g/cm3) and quartz(2.62e2.65 g/cm3). Finally, all the samples were hand-picked underthe binocular microscope. Samples were wrapped in aluminum foiland loaded into quartz tubes together with standards and irradiated

Figure 3. Photomicrographs in plane-polarized light (PPL) showing the textural relationshipthe Yuka terrane. (a) Euhedral to subhedral amphiboles contain clinozoisite, biotite andamphibole (09NQ29); (c) Primary fluid inclusions are randomly distributed with irregularlhypersaline primary fluid inclusions in quartz vein (09NQ16). These inclusions occur as iso

at the TRIGA reactor at the Oregon State University Reactor Centre.Irradiation timeswere20h for irradiationVU83.Correction factors forinterferences of Ca and K isotopes are: (39Ar/37Ar)Ca ¼ 0.000673,(36Ar/37Ar)Ca ¼ 0.000264, (40Ar/39Ar)K ¼ 0.00086 and

s of amphibolites (a, b) and primary fluid inclusions in amphibole and quartz (c, d) fromplagioclase as inclusions (09NQ13); (b) Garnet relics as inclusions in coarse-grainedy, elongated and tubular in shapes in amphibole (09NQ13); (d) Biphase liquid/vapourlated round inclusions and with extremely small size CO2 bubble.

Table 1Representative microprobe analyses of amphibole (23 O following Holland and Blundy, 1994).

Sample 09NQ13 09NQ29

Spot-position 35 42 46 47 36 38 41 43 6 26 20 24 21 23 17 14

Mineral Ed Ed Ed Ed Mg-hbl Mg-hbl Mg-hbl Mg-hbl Wnc Wnc Brs Brs Mg-hbl Mg-hbl Ts Ts

SiO2 44.02 43.84 43.89 43.78 45.02 43.95 46.20 45.52 53.63 53.89 51.86 53.13 47.39 47.86 42.57 42.33TiO2 0.44 0.47 0.47 0.46 0.50 0.51 0.42 0.40 0.10 0.10 0.13 0.08 0.25 0.17 0.42 0.25Al2O3 13.72 14.29 14.15 12.77 13.68 14.21 11.53 12.09 7.57 8.18 10.05 8.06 10.25 10.50 18.36 16.73FeO 14.42 14.38 15.18 15.02 13.41 14.52 14.17 13.95 4.61 4.54 4.81 5.03 12.93 12.10 9.78 13.77MnO 0.09 0.12 0.18 0.10 0.12 0.15 0.19 0.15 0.04 0.05 0.05 0.06 0.31 0.25 0.25 0.28MgO 10.19 9.97 9.25 9.87 10.48 9.63 10.92 10.58 18.02 17.97 16.94 17.95 12.29 12.67 12.10 9.84CaO 10.28 10.17 10.32 10.79 9.84 10.32 10.26 10.69 9.03 8.59 8.86 8.69 11.48 11.88 10.22 10.71Na2O 2.43 2.44 2.37 2.17 2.56 2.12 1.71 1.98 2.49 2.71 2.99 2.47 1.40 1.32 2.88 2.11K2O 1.04 1.14 1.11 1.17 0.95 1.12 0.95 0.90 0.24 0.28 0.18 0.33 0.25 0.17 0.20 0.37Total 96.62 96.81 96.91 96.12 96.55 96.51 96.36 96.24 95.73 96.32 95.85 95.81 96.55 96.92 96.78 96.39

Si 6.56 6.52 6.55 6.60 6.66 6.55 6.86 6.78 7.53 7.51 7.30 7.47 6.97 6.97 6.18 6.29AlIV 1.39 1.43 1.40 1.35 1.29 1.39 1.10 1.18 0.46 0.48 0.69 0.52 1.01 1.01 1.78 1.68AlVI 1.02 1.08 1.09 0.92 1.10 1.11 0.92 0.94 0.79 0.87 0.98 0.82 0.77 0.79 1.36 1.25Ti 0.05 0.05 0.05 0.05 0.06 0.06 0.05 0.04 0.01 0.01 0.01 0.01 0.03 0.02 0.05 0.03Mn 0.01 0.01 0.02 0.01 0.01 0.02 0.02 0.02 0.00 0.01 0.01 0.01 0.04 0.03 0.03 0.04Mg 2.26 2.21 2.06 2.22 2.31 2.14 2.42 2.35 3.77 3.73 3.55 3.76 2.69 2.75 2.62 2.18Ca 1.64 1.62 1.65 1.74 1.56 1.65 1.63 1.70 1.36 1.28 1.34 1.31 1.81 1.85 1.59 1.71Na 0.70 0.70 0.69 0.63 0.73 0.61 0.49 0.57 0.68 0.73 0.81 0.67 0.40 0.37 0.81 0.61K 0.20 0.22 0.21 0.22 0.18 0.21 0.18 0.17 0.04 0.05 0.03 0.06 0.05 0.03 0.04 0.07Fe3þ 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00Fe2þ 1.80 1.79 1.89 1.89 1.66 1.81 1.76 1.74 0.54 0.53 0.57 0.59 1.59 1.47 1.19 1.71Cations 15.63 15.63 15.61 15.64 15.55 15.55 15.42 15.49 15.19 15.20 15.28 15.22 15.34 15.31 15.63 15.56K/Ca 0.10 0.11 0.11 0.11 0.10 0.11 0.09 0.08 0.03 0.03 0.02 0.04 0.02 0.01 0.02 0.03

R. Hu et al. / Geoscience Frontiers 6 (2015) 759e770 763

(38Ar/39Ar)K¼ 0.01211.We used DRA1 Sanidinewith an assumed ageof 25.26� 0.2Ma asfluxmonitor standards for j-value calculation forthis project (Wijbrans et al., 1995).

Crushing in vacuo and stepwise heating of the powder residuesexperiments were carried out in an in-house designed crushingapparatus that was connected to a three stage extraction line and aquadrupole mass spectrometer (Schneider et al., 2009) in the argonisotope laboratory in VU University Amsterdam. Details of theexperimental procedure can be found in Qiu and Wijbrans (2006).Sampleswere loaded into a 40 cm long, 4 cm diameter Inconel� tubeand crushed by an iron pestle that is lifted and dropped with a fre-quency of one time per second (1 Hz) using external electromagnet-control. The extracted gases were exposed to a cold trap filled withliquid nitrogen (kept at ca. e 90 �C), and then purified at 250 �C by aFe/V/Zr getter pump and 450 �C by a Zr/Al getter pump. The purifi-cation process lasted for around 16 min before the gas was analyzedfor the argon isotopes in the mass spectrometer. Experiments beganand endedwith cold blank analyses to correct for system blankswiththe procedure described above but without moving the pestle. Toanalyze residues in stepwise heating experiments, the crusher tubewas rotated 90� and horizontally put into an externally temperature-controlled oven. The pestle was moved to the end of the tube toavoid degaussing it during stepwise heating. System blanks weremeasured in the initial steps at room temperature.

5. Results

The 40Ar/39Ar dating results are calculated using the ArArCALCsoftware package which was written by Dr. A.A.P Koppers (2002).The 40Ar/39Ar analytical results are listed in Table 2 and Appendix A.Age spectra and isochrons for each of the samples are plotted inFigs. 4e6. All errors are reported at the 2s confidence level in thetext and figures.

5.1. Amphibole

The 40Ar/39Ar dating results of amphiboles 09NQ13Amp and09NQ29Amp by crushing in vacuo yield similar monotonically

decreasing staircase shaped age spectra characterized by anoma-lously high apparent ages in the initial steps (Fig. 4a and c). Thepossible role of 39Ar loss by recoil during sample irradiation withfast neutrons can result in extremely old apparent age has beenrecognized since the development of the 40Ar/39Ar dating method.The mean recoil distance of 39Ar produced from 39K by neutronirradiation was calculated of w0.1 mm with energies of w300 KeV(Onstott et al., 1995; Villa, 1997), while most of the fluid inclusionsin the analyzed amphiboles in size of 0.5e3 mm. Therefore, it seemsthat the extremely high ages at the first steps of in vacuo crushing(w5 Ga) cannot be caused by 39ArK loss by recoil. More likely, theextremely apparent ages can be interpreted as the contamination ofexcess 40Ar which derived from the bigger, easily crushed, sec-ondary fluid inclusions (Qiu and Wijbrans, 2006, 2008; Jiang et al.,2012).

Subsequently, upon continued crushing, the apparent agesgradually stabilize to reveal age plateaus. After applying their initial40Ar/36Ar ratios from the atmospheric air as 295.5 to exclude thenon-radiogenic 40Ar, the two amphibole samples produce ageplateaus in the final steps with concordant early Paleozoicweighted mean ages of 487.9 � 6.5 Ma (39Ar¼ 48%，MSWD¼ 190)and 475.7 � 3.6 Ma, respectively (39Ar ¼ 47%，MSWD ¼ 14.8,Fig. 4a and c). The K/Ca release spectra for these two samples arerelatively flat with mean values of 0.140 � 0.003 and 0.073 � 0.001,respectively (Fig. 4a and c). The data points of the steps defining theabove ages yield isochrons with younger ages of 469 � 2.1 Ma and462.6 � 3.6 Ma, corresponding to initial 40Ar/36Ar ratios of519.9 � 20.1 and 333.8 � 9.5 (Fig. 4b, d). These results indicate thatthere is a significant amount of excess 40Ar presence within thesmall primary fluid inclusions that survive in the powderedmineralseparates after crushing. By applying this initial 40Ar/36Ar ratios toexclude the non-atmospheric/non-radiogenic 40Ar component, therecalculated ages reveal flat age spectra with plateau ages of470.0 � 1.9 Ma (09NQ13Amp) and 462.5 � 2.6 Ma (09NQ29Amp)for approximately 48% of the total 39Ar released.

All crushed residual powders were subsequently investigated bymeans of stepwise heating. In comparison to the in vacuo crushingexperiments, the 40Ar/39Ar dating results by stepwise heating yield

Table 2Complete 40Ar/39Ar crushed powder stepped heating data.

Step T (�C) 36Arair 37ArCa 38ArCl 39ArK 40Ar* Age � 2s (Ma) 40Ar* (%) 39ArK (%) K/Ca (�2s)

(a) Amphibole 09NQ13Amp, by stepped heating, J ¼ 0.0050371 300 1032.2 29433.9 15.8 12668.6 593595.2 382.3 � 2.0 28.33 2.66 0.185 � 0.0192 350 2015.9 14147.5 42.9 5727.6 235519.8 339.6 � 7.8 66.06 5.88 0.174 � 0.0193 400 1699.2 17630.2 70.6 5980.0 242015.2 334.7 � 6.0 32.52 2.78 0.146 � 0.0164 450 1579.1 21254.1 51.8 6754.3 259115.1 318.7 � 4.8 35.70 3.14 0.137 � 0.0175 550 1521.0 31424.0 71.3 9247.7 407232.5 361.4 � 6.8 47.54 4.29 0.127 � 0.0136 600 2065.1 35735.6 80.8 9835.0 451296.9 375.1 � 4.1 42.51 4.57 0.118 � 0.0137 650 3074.3 52962.9 126.9 12147.8 607088.9 405.1 � 4.3 40.66 6.15 0.099 � 0.0108 750 3846.4 100014.9 247.5 19092.3 986969.4 417.5 � 3.4 44.58 9.39 0.082 � 0.0089 850 5769.6 373530.1 452.3 58739.1 2577524.3 360.3 � 2.2 62.50 27.85 0.068 � 0.00710 900 9774.9 423760.1 596.8 71283.0 2156061.1 255.8 � 2.6 42.45 33.31 0.072 � 0.007

(b) Amphibole 09NQ29Amp, by stepped heating, J ¼ 0.0050311 300 674.7 30623.5 6.4 8887.9 369100.3 340.6 � 2.2 64.93 8.35 0.125 � 0.0132 400 1581.6 17060.5 26.5 4275.9 141523.9 276.4 � 8.0 23.24 4.02 0.108 � 0.0123 450 1515.2 22699.5 36.6 4481.1 143131.7 267.5 � 7.0 24.22 4.21 0.085 � 0.0104 500 1508.8 24914.2 69.0 5070.9 149727.0 248.6 � 5.6 25.14 4.77 0.088 � 0.0115 550 2467.5 29992.9 102.8 5495.2 163851.1 250.8 � 7.8 18.35 5.16 0.079 � 0.0086 600 2731.5 33184.9 79.3 5862.2 181332.0 259.6 � 8.5 18.34 5.51 0.076 � 0.0087 650 2801.7 66477.0 88.7 7292.3 297638.5 335.2 � 6.8 26.44 6.85 0.047 � 0.0058 700 2767.6 81856.7 83.1 7038.9 283340.1 331.0 � 7.9 25.73 6.62 0.037 � 0.0049 750 3233.9 133252.4 137.6 8721.9 427645.7 395.8 � 6.1 30.92 8.20 0.028 � 0.00310 800 4010.9 335260.5 189.2 17076.4 877643.6 412.8 � 4.3 42.54 16.05 0.022 � 0.00211 850 8663.3 297400.7 293.7 14142.9 448258.2 265.5 � 10.6 14.90 13.29 0.020 � 0.00212 900 11315.3 412001.2 397.7 18056.8 426914.1 201.7 � 11.2 11.32 16.97 0.019 � 0.002

These 40Ar/39Ar experiments were measured using a quadrupole mass spectrometer in VU University Amsterdam. The argon isotopes are listed in cps.

R. Hu et al. / Geoscience Frontiers 6 (2015) 759e770764

relatively complicated age spectra and younger apparent ages.Saddle shaped age spectrawith minimum apparent ages of 319 and249 Ma at temperatures of 450 �C and 500 �C, and maximumapparent ages of 418 and 413 Ma at temperatures of 750 �C and800 �C are obtained from samples 09NQ13Amp and 09NQ29Amp,respectively (Fig. 4a and c). Argon released in the early steps hashigher K/Ca values than those in the later steps and with total gasK/Ca values of 0.083 � 0.004 and 0.031 � 0.001, respectively. Thereis too much scatter in the data of all amphibole samples to formwell-defined isochrons (Fig. 4b and d).

5.2. Vein quartz

Vein sample 09NQ16Qz was subjected to 32 crushing steps of14,964 impacts, and produced a descending staircase-shaped agespectrum. Apparent ages decrease from as old as 8448 to 506 Mawith steps 1 to 23, and then formed a relative flat age spectrumfrom stages 24e32 with a weighted mean age of 402.5 � 13.3 Ma(39ArK ¼ 21.3%, MSWD ¼ 0.5) (Fig. 5a). The data points defining theage plateau yields a well-defined isochron line with age of405.4 � 41.6 Ma (MSWD ¼ 0.5), corresponding to an initial40Ar/36Ar ratio of 295.2 � 4.1, which is indistinguishable from themodern atmospheric ratio of 295.5 (Fig. 5b).

6. Discussion

6.1. The source of Ar isotopes in fluid inclusions

To search for elemental correlations, we performed multipleregressions on the amount of 40Ar*, Cl (38ArCl) and K (39ArK) withthe aim of identifying the different Ar sources in the fluid inclusions(e.g., Kelley et al., 1986; Kendrick et al., 2001; Qiu and Wijbrans,2006). Here, 40Ar* denotes total 40Ar minus atmospheric 40Ar(40ArA). In other words, 40Ar* includes both in situ derived radio-genic 40Ar (40ArR) from the decay of 4 K and parentless excess 40Ar(40ArE).

The plots of 40Ar*/39ArK vs. 38ArCl/39ArK based on the 40Ar/39Ardata of the amphiboles and quartz by crushing are shown in Fig. 6,

and reveal the following characteristics: (1) the crushing steps(blue points with labels in Fig. 6) yield clear linear correlations with40Ar*/39ArK intercepts of 41e33.3, corresponding to ages of397e295 Ma (Fig. 6). The intercept ages (337e295 Ma) can prob-ably be taken as the best estimate for the age of secondary fluidinclusions (SFIs), as discussed later; (2) in contrast, the late crushingsteps (pink points near the origin of Fig. 6) define the isochronswithout any correlation between 40Ar*/39ArK and 38ArCl/39ArK, butwith constant low 40Ar*/39ArK ratios, implying a unique reservoirdominates the degassing. Two amphibole samples show average40Ar*/39ArK ratios of 63e59 and with average ages of around 485and 475 Ma, respectively (Fig. 6a and b). The quartz sample giveslower 40Ar*/39ArK ratios of 46e41, corresponding to an average ageof 409 Ma (Fig. 6c). These ages are based on the data from the latecrushing steps and are interpreted as the contributions of the pri-mary fluid inclusions (PFIs); (3) in addition to that, we note that thecrushing steps (green points in Fig. 6) between the initial and finalcrushing steps have low 40Ar*/39ArK ratios, low 38ArCl/39ArK ratios,and a scattered distribution. These ratios are interpreted to indicatethat the gases in these crushing steps were a mixture with variableproportions of the PFIs and SFIs end members.

6.2. Dating episodic flow of metamorphic fluid

Fluids in UHP rocks have been extensively studied and arewidely accepted to be derived from dehydration reactions of hy-drous phases (Hermann et al., 2006). For the Sulu-Dabie orogenicUHP metamorphic rocks, based on studies involving stable iso-topes, fluid inclusions and petrological phase relationships, Zheng(2004) proposed that the fluid activity during the exhumation ofdeep-subducted continental crustal is characterized by pervasivefluid flow resulting in amphibolite-facies retrogression, and chan-nelized fluid flow leading to the formation of HP quartz veinswithin eclogites. In many other UHP metamorphic belts, fluid in-clusion studies suggest that minerals in eclogites and their retro-gressed equivalents normally contain a variety of types of fluidinclusions, and there is some correlation between metamorphicgrade and metamorphic fluid composition (Andersen et al., 1989,

Figure 4. Plots of the age and K/Ca spectra (a, c) and inverse isochrons (b, d) of amphiboles based on the 40Ar/39Ar results by in vacuo crushing and crushed powders stepwiseheating. The distributions of the data points in the inverse isochrons have features in common. The results of crushing gradually move away from the excess 40Ar end-member to theradiogenic 40Ar end-member, indicating that ArE and ArR contribute successively to different parts of the degassing. In contrast, the data points from stepwise heating experimentsshow a radiogenic 40Ar end-member to air 40Ar end-member trend with increasing heating temperature, but they are too scatter to form well correlate linear arrays. The crushingdata points contributing to the weighted mean ages define the isochron lines with intercept ages of 469 and 463 Ma. Here WMA stands for weighted mean age and TGA is the totalgas age.

R. Hu et al. / Geoscience Frontiers 6 (2015) 759e770 765

1993; Philippot et al., 1995; Han and Zhang, 1996; Touret, 2001;Zhang et al., 2011c). In other words, fluid flow activities in UHPmetamorphism are generally considered to be episodic. As illus-trated in Fig. 7, the P-T evolution of eclogites in the Yuka terrane canbe divided into five metamorphic stages: pre-eclogite-facies, laterpre-peak, peak pressure, peak temperature and post peak-T retro-gression (Zhang et al., 2009a). Therefore, Yuka HP/UHP mineralscorresponding to the various metamorphic stages could haveencapsulated fluid inclusions with a variety of ages and originsduring the mineral growth.

The purpose of the in vacuo crushing technique is to extract theAr-bearing gas that was trapped in fluid inclusions and in inclusionsalong sealed cracks and cavities in crystals. Previous studies haveshown that crushing has little effect on the gas trapped within thecrystal lattice (Dunlap and Kronenberg, 2001; Qiu et al., 2002; Qiuand Wijbrans, 2006, 2008). An exception to this may be scapolite(Kendrick and Phillips, 2009b), for which argon release by crushingwas described. This may be due to the relatively open crystal

structure of scapolite where volatiles may have been trapped incavities within the crystal lattice. In most cases, the results of the40Ar/39Ar in vacuo crushing should be able to reflect the age infor-mation of the fluid flow as recorded by fluid inclusions in amphi-boles and quartz that crystallized during amphibolite-facies togreenschist-facies retrogression.

6.3. Implications of gas release patterns and spectra of crushingexperiments

6.3.1. Gas release patternThe new results from amphibole and vein quartz 40Ar/39Ar in

vacuo crushing analysis all yield descending staircase-shaped agespectra with anomalously high initial apparent ages in the first fewsteps and relative flat age plateaus over the final several steps.Following the interpretation of Qiu and co-workers (2006, 2008),we suggest that the extremely high initial apparent ages arederived from the largest, most easily crushed secondary fluid

Figure 5. Plot of the age spectrum (a) and inverse isochron (b) of quartz based on the40Ar/39Ar results by crushing in vacuo. The distribution of the data points throughcrushing experience shows the excess 40Ar end-member to air 40Ar end-membertrend, showing that ArE and ArA gradually contribute to different parts during thegas extraction. Here WMA represents the weighted mean age.

Figure 6. Plots of the correlation between 40Ar*/39ArK vs. 38ArCl/39ArK for in vacuocrushing of amphiboles (a, b) and quartz (c). 40Ar* shows strong correlation with 38ArClin the early crushing steps of all the samples.

R. Hu et al. / Geoscience Frontiers 6 (2015) 759e770766

inclusions (SFIs) that are dominated by excess 40Ar and that weremost likely incorporated in the amphibole and quartz after theinitial amphibolite-facies conditions. Meanwhile, in the diagrams ofFig. 6, we discovered that the 40Ar* (representing 40ArRþ40ArE)shows a good correlation with 38ArCl in the initial crushing steps,implying the excess 40Ar in the SFIs and Cl came from the samesource. With continued crushing, the apparent ages decreasegradually probably reflecting a mixture between SFIs with excess40Ar and PFIs. Finally, once the high-excess 40Ar reservoir isexhausted, subsequent gas components are mainly derived fromsmaller PFIs. During this stage of the experiment a relative flat ageplateau is formed over the final several steps. The amphibole datapoints constituting the age plateaus result in isochrons with rela-tive younger intercept ages (469e463 Ma), corresponding to initial40Ar/36Ar values from 334 to 520, implying excess 40Ar is alsopresent in the fine PFIs of amphibole samples. In contrast, thequartz data points constituting the age plateau define an isochronwith a concordant intercept age (405 Ma), corresponding to initial40Ar/36Ar ratio of 295 � 4 indicative of the absence of excess40Ar in PFIs.

In an alternative model to explain the results of Qiu andWijbrans (2006, 2008) on the release patterns of Bixiling by step-wise crushing (Kendrick and Phillips, 2009a) postulated the pres-ence of small sericite crystals dispersed throughout the host garnetthat contributed to the release of radiogenic argon after prolonged

crushing as the source of the observed age plateaus. Although thisis a plausible model, it critically hinges upon the presence of finesericite evenly dispersed throughout the crystals used for crushing.In the case of the Bixiling garnets, Qiu and Wijbrans (2006, 2008)pointed out that in these garnets they have found no evidence forsuch sericite inclusions whereas the presence of numerous smallprimary fluid inclusions was demonstrated. Moreover, for the caseof Bixiling garnet inclusions Qiu and Wijbrans pointed out that theradiogenic component had a positive correlation with the 38ArClconcentration, which would point to a chloride enriched reservoirthat was releasing the radiogenic argon signal. A chloride-enrichedradiogenic argon signal is considered to be more consistent with aninclusion-hosted brine source than with a mica crystal source. Forthe present study sericite inclusions in amphibole are consideredan unlikely source for the radiogenic signal is considered unlikely assericite is not normally part of mafic amphibolite mineralassemblages.

Amphibole and syn-metamorphic vein quartz are mainlyformed during amphibolite facies and greenschist facies retro-gression. During these stages, significant amounts of fluid could beobtained frommolecular H2O in nominally anhydrous minerals likegarnet and omphacite, the breakdown of hydrous minerals likezoisite, phengite and lawsonite, the exsolution of hydroxyl minerals(Zheng et al., 2003; Zheng, 2004; Song et al., 2005a; Frezzotti et al.,2007; Wu et al., 2009), and even from the host gneiss (Chen et al.,

Figure 7. P-T-t path for eclogite from the Yuka terrane, North Qaidam UHP metamorphic belt, western China (after Zhang et al., 2009a), amended with our new ages for fluid flowevents. A: an earlier pre-eclogite-facies stage; B: a later pre-peak stage; C: peak pressure stage; D: peak temperature stage; E: a later stage of retrogression.

R. Hu et al. / Geoscience Frontiers 6 (2015) 759e770 767

2007c; Zong et al., 2010). As a consequence, this kind of meta-morphic fluid probably was trapped by amphibole and vein quartzcrystals during crystallization or at a later stage by a crack-sealmechanism. Such fluid inclusions can therefore be considered asprimary. The ages of these PFIs therefore not only record aqueousfluid flow during the exhumation stages of HP/UHP rocks, but canalso be taken as the best estimate for the ages of amphibolite faciesretrogression and quartz-vein formation.

6.3.2. Implications of crushing spectraCrushing experiments of two amphibole samples yielded

40Ar/39Ar isochron ages of 469� 2.1 and 462.6� 3.6 Ma. These agesare consistent with the 40Ar/39Ar stepwise heating isochron ages(477e466 Ma) from eclogite phengite and amphibole that havebeen interpreted as the timing of early cooling in the eclogite faciesafter peak pressure metamorphism (Zhang et al., 2005). This sug-gests that the 40Ar/39Ar ages of 469e458 Ma constrain the age offluid flow responsible for initial amphibolite-facies retrogressionduring the exhumation of the UHP rocks (see Fig. 7). On the otherhand, the granite zircon UePb dating from the Da Qaidam regionperformed by various analytical different techniques produced agesfrom 397 Ma to as old as 490 Ma with a cluster in the range of465e457 Ma, which have been interpreted to date intrusion ofregionally widespread Ordovician plutons (Gehrels et al., 2003; Wuet al., 2007). Therefore, regional syn-exhumation arc-magmatism

accompanied by the release of aqueous fluid may be an alternativeexplanation for enhanced fluid flow at 469e457 Ma.

In contrast, crushing analysis of the quartz vein within a garnetamphibolite lens yields a younger 40Ar/39Ar isochron age of ca.405.4 Ma. Because the crushing experiment effectively extract thefluid inclusion gases, rather than argon from the crystal lattice orsolid mineral inclusions of the quartz sample, we interpret thesedata as constraining the age of the fluid inclusions locked in duringquartz mineralization. This age is very close to the youngestexhumation ages obtained by muscovite 40Ar/39Ar dating (Xu et al.,2006). Therefore, the 405 Ma age from quartz in vacuo crushingmay record aqueous fluid flow during the late exhumation of YukaUHP rocks (Fig. 7).

6.4. Implications of gas release patterns and spectra of steppedheating

The two amphibole samples have similar complex age spectra,which have commonly been attributed to the presence of excess40Ar and partial resetting or contamination by other phases(Harrison and McDougall, 1981; Harrison and Fitzgerald, 1986).The relatively young apparent ages (418e249 Ma) and total gasages (333e305 Ma) obtained from two amphibole samples aremuch younger than the published amphibole 40Ar/39Ar incre-mental heating result of w477 Ma (Zhang et al., 2005). This

R. Hu et al. / Geoscience Frontiers 6 (2015) 759e770768

indicates that the complex release patterns are not caused byvariable incorporation of excess 40Ar. The obvious presence ofminute mineral inclusions such as biotite, feldspar, and quartz inamphibole porphyroblasts, characterized by various shapes andsizes from 0.01 mm to 0.1 mm, suggest that mineral inclusionsmay have played a role in producing the complicated releasepatterns during stepwise heating of the residual powder aftercrushing. Furthermore, the effect of feldspar or/and biotite in-clusions is also supported by their relatively higher K/Ca valuesduring the early steps (see Fig. 4a and c). The 319e249 Maapparent ages obtained at heating temperatures around 500 �C,probably represent the formation or resetting ages of late retro-grade minerals, whereas the apparent ages of 418e413 Ma, ob-tained at heating temperatures around 800 �C, may representamphibolite cooling ages. It should be pointed out that the step-wise heating experiment on amphibole residues achieves only amaximum temperature of w900 �C while amphiboles are prob-ably not completely degassed at such temperature. Thus, it ispossible that after finishing step heating at temperatures higherthan 900 �C, higher ages due to release from a phase that is moreresistant to 40Ar loss could be obtained (e.g., Siebel et al., 1998). Ifso, the apparent ages of 418e413 Ma can only be interpreted asrepresenting cooling ages of the less 40Ar resistant amphibolephase, which may be taken as the best estimate for the age of lateramphibole-facies metamorphism (see Fig. 7).

On balance, based upon the young apparent ages from thestepwise heating experiments, we suggest that there is no excess40Ar in the amphibole crystals or their inclusions and that almostall excess 40Ar in amphibole powders resides in primary andsecondary fluid inclusions. Considering all in vacuo crushing andresidual powders stepped heating results, we suggest thatamphibole in the Yuka amphibolites crystallized at ca. 469e463Ma, and then cooled below the amphibole KeAr system closuretemperature of about 500 � 50 �C (Harrison, 1982) at or before 418Ma (09NQ13Amp) and 413 Ma (09NQ29Amp). Amphiboles werepartly reset by subsequent thermal overprinting and thencompletely closed at 319 Ma (09NQ13Amp) and 249 Ma(09NQ29Amp).

7. Conclusions

An investigation of 40Ar/39Ar dating by in vacuo crushing andstepwise heating of amphibole separates from amphibolite-faciesoverprinted HP/UHP metamorphic rocks from the Yuka terraneresulted in the following conclusions:

(1) The 40Ar/39Ar in vacuo crushing technique appears to aneffective and promising approach to directly date the mineralformation ages in retrogressed HP/UHP rocks and to decipherthe occurrence of excess 40Ar. In addition, gas trapped in solidinclusions in minerals could be liberated by subsequent step-ped heating analysis of the crushed powder and provide usefulgeological information about the retrograde metamorphicoverprinting of HP/UHP rocks.

(2) In vacuo crushing 40Ar/39Ar analysis was applied to amphiboleand quartz separates from amphibolite samples from the YukaHP/UHP terrane in western China. The results indicate that thesecondary fluid inclusions were formed subsequent to UHPmetamorphism and that the amphibole and quartz incorpo-rated significant amounts of heterogeneous excess 40Ar influid inclusions. The excess 40Ar is generally released early inthe crushing experiments. Primary fluid inclusions trappedduring amphibole growth and released during the later stagesof the crushing experiments contain relatively homogeneous

excess 40Ar. There is virtually no excess 40Ar in the PFIs in veinquartz.

(3) In vacuo crushing of amphiboles provides a geochronologicalrecord of episodes of fluid flow with ages ranging from 469 to463 Ma, reflecting the timing of the Yuka eclogite initialamphibolite-facies retrogression. In contrast, vein quartzwithin garnet amphibolite gave an age of 405 Ma, which mayrecord aqueous fluid flow during the later stages of exhumationof these UHP rocks.

(4) Based on the 40Ar/39Ar stepwise heating ages of amphiboleresidual powders and their equilibration and closure temper-atures, we deduce that the Yuka metamorphic complex and itscountry rocks cooled from peak metamorphic temperature to500 � 50 �C and reached medium to upper crustal levels at orbefore 418e413 Ma. The KeAr isotope system in amphibolewas partially reset by later tectonothermal events at ca. 319 to249 Ma, indicating that the Yuka eclogites and the products oftheir retrogression were overprinted by multiple later thermalevents in the Silurian and possibly as young as the Triassic.

Acknowledgments

This work was funded by the Chinese Academy ScienceseRoyalNetherlands Academy of Arts and Sciences (CAS-KNAW) Joint PhDTraining Programme (ISK/3523/PhD) and the joint research projectbetween the CAS and the De Koninklijke Nederlandse Akademievan Wetenschappen (06CDP002). We sincerely thank Mr. R VanElsas, Mrs. W Koot, Mr. O Postma, Mr. A Bikker, and Mr. W van derPlas from the Faculty of Earth and Life Sciences, VU, for their kindhelp in mineral separation, thin section preparation, and technicalsupport for 40Ar/39Ar analyses. Dr. Huaiyu He and an anonymousreviewer are thanked for their valuable reviews. We are alsograteful to Dr. Inna Safonova for her comments and editorialhandling of the manuscript.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.gsf.2014.09.005.

References

Alexander, C.J., 1975. 40Ar/39Ar studies of precambrian cherts; an unsuccessfulattempt to measure the time evolution of the atmospheric 40Ar/36Ar ratio.Precambrian Research 2, 329e344.

Andersen, T., Austrheim, H., Burke, E.A.J., Elvevold, S., 1993. N2 and CO2 in deepcrustal fluids: evidence from the Caledonides of Norway. Chemical Geology 108,113e132.

Andersen, T., Burke, E.A.J., Austrheim, H., 1989. Nitrogen-bearing, aqueous fluidinclusions in some eclogites from the Western Gneiss region of the NorwegianCaledonides. Contributions to Mineralogy and Petrology 103, 153e165.

Bai, X.J., Wang, M., Jiang, Y.D., Qiu, H.N., 2013. Direct dating of tin-tungstenmineralization of the Piaotang Tungsten deposit, South China, by 40Ar/39Arprogressive crushing. Geochimica et Cosmochimica Acta 114, 1e12.

Baker, J., Matthews, A., Mattey, D., Rowley, D., Xue, F., 1997. Fluid-rock interactionsduring ultra-high pressure metamorphism, Dabie Shan, China. Geochimica etCosmochimica Acta 61, 1685e1696.

Burgess, R., Kelley, S.P., Parsons, I., Walker, F.D.L., Worden, R.H., 1992. 40Are39Aranalysis of perthite microtextures and fluid inclusions in alkali feldspars fromthe Klokken syenite, South Greenland. Earth and Planetary Science Letters 109,147e167.

Burgess, R., Parsons, I., 1994. Argon and halogen geochemistry of hydrothermalfluids in the Loch Ainort granite, Isle of Skye, Scotland. Contributions toMineralogy and Petrology 115, 345e355.

Chen, D.L., Liu, L., Sun, Y., Liou, J.G., 2009. Geochemistry and zircon U-Pb dating andits implications of the Yukahe HP/UHP terrane, the North Qaidam, NW China.Journal of Asian Earth Sciences 35, 259e272.

Chen, D.L., Liu, L., Sun, Y., Sun, W.D., Zhu, X.H., Liu, X.M., Guo, C.L., 2012. Felsic veinswithin UHP eclogite at xitieshan in North Qaidam, NW China: partial meltingduring exhumation. Lithos, 136e139, 187e200.

Chen, D.L., Sun, Y., Liu, L., 2007a. The metamorphic ages of the country rocks of theYukahe eclogites in the northern margin of Qaidam Basin and its geological

R. Hu et al. / Geoscience Frontiers 6 (2015) 759e770 769

significance. Earth Science Frontiers 14, 108e116 (in Chinese with Englishabstract).

Chen, D.L., Sun, Y., Liu, L., Zhang, A.D., Lin, C.L., 2007b. In situ LA-ICP-MS zircon U-Pbage of ultrahigh-pressure eclogites in the Yukahe area, northern Qaidam Basin.Science in China Series D: Earth Sciences 50, 322e330.

Chen, D.L., Sun, Y., Liu, L., Zhang, A.D., Luo, J.H., Wang, Y., 2005. Metamorphicevolution of the Yuka eclogite in the North Qaidam, NW China: evidences fromthe compositional zonation of garnet and reaction texture in the rock. ActaPetrologica Sinica 21, 1039e1048.

Chen, R.X., Zheng, Y.F., Gong, B., Zhao, Z.F., Gao, T.S., Chen, B., Wu, Y.B., 2007c.Origin of retrograde fluid in ultrahigh-pressure metamorphic rocks: con-straints from mineral hydrogen isotope and water content changes in eclo-giteegneiss transitions in the Sulu orogen. Geochimica et Cosmochimica Acta71, 2299e2325.

Dunlap, W.J., Kronenberg, A.K., 2001. Argon loss during deformation of micas:constraints from laboratory deformation experiments. Contributions toMineralogy and Petrology 141, 174e185.

Franz, L., Romer, R.L., Klemd, R., Schmid, R., Oberhänsli, R., Wagner, T., Dong, S.W.,2001. Eclogite-facies quartz veins within metabasites of the Dabie Shan (easternChina): pressure-temperature-time-deformation path, composition of the fluidphase and fluid flow during exhumation of high-pressure rocks. Contributionsto Mineralogy and Petrology 141, 322e346.

Frezzotti, M.L., Ferrando, S., Dallai, L., Compagnoni, R., 2007. Intermediate Alka-lieAlumino-silicate aqueous Solutions released by deeply subducted Conti-nental crust: fluid evolution in UHP OH-rich TopazeKyanite Quartzites fromDonghai (Sulu, China). Journal of Petrology 48, 1219e1241.

Fu, B., Touret, J.L.R., Zheng, Y.F., Jahn, B., 2003. Fluid inclusions in granulites, gran-ulitized eclogites and garnet clinopyroxenites from the Dabie-Sulu terranes,eastern China. Lithos 70, 293e319.

Gao, T.S., Wang, S.S., Gong, B., Wu, Y.B., Zheng, Y.F., 2006. Postcollisional flow ofaqueous fluid within ultrahigh-pressure eclogite in the Dabie orogen. Journal ofGeochemical Exploration 89, 115e118.

Gehrels, G.E., Yin, A., Wang, X.F., 2003. Magmatic history of the northeastern Ti-betan Plateau. Journal of Geophysical Research 108, 2423.

Han, Y.J., Zhang, Z.M., 1996. Fluids in high-pressure and ultrahigh pressure meta-morphism. Earth Science Frontiers 4, 3222e3229 (in Chinese with Englishabstract).

Harrison, T.M., 1982. Diffusion of 40Ar in hornblende. Contributions to Mineralogyand Petrology 78, 324e331.

Harrison, T.M., Fitzgerald, J.D., 1986. Exsolution in hornblende and its consequencesfor age spectra and closure temperature. Geochimica et Cosmochimica Acta 50,247e253.

Harrison, T.M., Matthew, H., Oscar, L., 1993. In vacuo crushing experiments and K-feldspar thermochronometry. Earth and Planetary Science Letters 117, 169e180.

Harrison, T.M., McDougall, I., 1981. Excess 40Ar in metamorphic rocks from BrokenHill, New South Wales: implications for 40Ar/39Ar age spectra and the thermalhistory of the region. Earth and Planetary Science Letters 55, 123e149.

Hermann, J., Spandler, C., Hack, A., Korsakov, A.V., 2006. Aqueous fluids and hydrousmelts in high-pressure and ultra-high pressure rocks: implications for elementtransfer in subduction zones. Lithos 92, 399e417.

Holland, T.J.B., Blundy, J., 1994. Non-ideal interactions in calcic amphiboles and theirbearing on amphibole-plagioclase thermometry. Contributions to Mineralogyand Petrology 116, 433e447.

Jiang, Y.D., Qiu, H.N., Xu, Y.G., 2012. Hydrothermal fluids, argon isotopes andmineralization ages of the Fankou PbeZn deposit in south China: Insights fromsphalerite 40Ar/39Ar progressive crushing. Geochimica et Cosmochimica Acta84, 369e379.

Kelley, S., Turner, G., Butterfield, A.W., Shepherd, T.J., 1986. The source and signifi-cance of argon isotopes in fluid inclusions from areas of mineralization. Earthand Planetary Science Letters 79, 303e318.

Kendrick, M.A., Burgess, R., Pattrick, R.A.D., Turner, G., 2001. Halogen and Ar-Ar agedeterminations of inclusions within quartz veins from porphyry copper de-posits using complementary noble gas extraction techniques. Chemical Geology177, 351e370.

Kendrick, M.A., Phillips, D., 2009a. Discussion of ‘the Paleozoic metamorphic historyof the Central Orogenic Belt of China from 40Ar/39Ar geochronology of eclogitegarnet fluid inclusions by Qiu Hua-Ning and Wijbrans J.R.’. Earth and PlanetaryScience Letters 279, 392e394.

Kendrick, M.A., Phillips, D., 2009b. New constraints on the release of noble gasesduring in vacuo crushing and application to scapolite Br-Cl-I and 40Ar/39Ar agedeterminations. Geochimica et Cosmochimica Acta 73, 5673e5692.

Kendrick, M.A., Phillips, D., Miller, J.M., 2006. Part I. Decrepitation and degassingbehaviour of quartz up to 1560�C: analysis of noble gases and halogens incomplex fluid inclusion assemblages. Geochimica et Cosmochimica Acta 70,2540e2561.

Koppers, A.A.P., 2002. ArArCALC-software for 40Ar/39Ar age calculations. Computers& Geosciences 28, 605e619.

Leake, B.E., Woolley, A.R., Arps, C.E.S., Birch, W.D., Gilbert, M.C., Grice, J.D.,Hawthorne, F.C., Kato, A., Kisch, H.J., Krivovichev, V.G., Linthout, K., Laird, J.,Mandarino, J.A., Maresch, W.V., Nickel, E.H., Rock, N.M.S., Schumacher, J.C.,Smith, D.C., Stephenson, N.C.N., Ungaretti, L., Whittaker, E.J.W., Youzhi, Guo,1997. Nomenclature of amphiboles; report of the subcommittee on amphibolesof the International Mineralogical Association, Commission on new Mineralsand Mineral names. American Mineralogist 82, 1019e1037.

Li, Y.L., Zheng, Y.F., Fu, B., Zhou, J.B., Wei, C.S., 2001. Oxygen isotope composition ofquartz-vein in ultrahigh-pressure eclogite from Dabieshan and implications fortransport of high-pressure metamorphic fluid. Physics and Chemistry of theEarth, Part A: Solid Earth and Geodesy 26, 695e704.

Mattinson, C.G., Wooden, J.L., Liou, J.G., Bird, D.K., Wu, C.L., 2006. Age and durationof eclogite-facies metamorphism, North Qaidam HP/UHP terrane, WesternChina. American Journal of Science 306, 683e711.

Mattinson, C.G., Wooden, J.L., Zhang, J.X., Bird, D.K., 2009. Paragneiss zircongeochronology and trace element geochemistry, North Qaidam HP/UHP terrane,western China. Journal of Asian Earth Sciences 35, 298e309.

Menold, C.A., 2006. Tectonic and Metamorphic Evolution of the North QaidamUltrahigh-pressure Metamorphic Terrane, Western China. PhD Thesis. Univer-sity of California, Los Angeles, 261 pp.

Menold, C.A., Manning, C.E., Yin, A., Tropper, P., Chen, X.H., Wang, X.F., 2009.Metamorphic evolution, mineral chemistry and thermobarometry of orthog-neiss hosting ultrahigh-pressure eclogites in the North Qaidam metamorphicbelt, Western China. Journal of Asian Earth Sciences 35, 273e284.

Onstott, T.C., Miller, M.L., Ewing, R.C., Arnold, G.W., Walsh, D.S., 1995. Recoil re-finements: implications for the 40Ar/39Ar dating technique. Geochimica etCosmochimica Acta 59, 1821e1834.

Philippot, P., Chevallier, P., Chopin, C., Dubessy, J., 1995. Fluid composition andevolution in coesite-bearing rocks (Dora-Maira massif, Western Alps): impli-cations for element recycling during subduction. Contributions to Mineralogyand Petrology 121, 29e44.

Phillips, D., Miller, J.M., 2006. 40Ar/39Ar dating of mica-bearing pyrite from ther-mally overprinted Archean gold deposits. Geology 34, 397e400.

Qiu, H.N., 1996. 40Ar/39Ar dating of the quartz samples from two mineral deposits inwestern Yunnan (SW China) by crushing in vacuum. Chemical Geology 127,211e222.

Qiu, H.N., Jiang, Y.D., 2007. Sphalerite 40Ar/39Ar progressive crushing and stepwiseheating techniques. Earth and Planetary Science Letters 256, 224e232.

Qiu, H.N., Wijbrans, J.R., 2006. Paleozoic ages and excess 40Ar in garnets from theBixiling eclogite in Dabieshan, China: new insights from 40Ar/39Ar dating bystepwise crushing. Geochimica et Cosmochimica Acta 70, 2354e2370.

Qiu, H.N., Wijbrans, J.R., 2008. The Paleozoic metamorphic history of the CentralOrogenic Belt of China from 40Ar/39Ar geochronology of eclogite garnet fluidinclusions. Earth and Planetary Science Letters 268, 501e514.

Qiu, H.N., Wijbrans, J.R., Brouwer, F.M., Yun, J.B., Zhao, L.H., Xu, Y.G., 2010.Amphibolite facies retrograde metamorphism of the Zhujiachong eclogite, SEDabieshan: 40Ar/39Ar age constraints from argon extraction using UV-lasermicroprobe, in vacuo crushing and stepwise heating. Journal of MetamorphicGeology 28, 477e487.

Qiu, H.N., Zhu, B.Q., Sun, D.Z., 2002. Age significance interpreted from dating ofquartz samples from the Dongchuan Copper Deposits, Yunnan, SW China, bycrushing and heating. Geochimical Journal 36, 475e491.

Schneider, B., Kuiper, K., Postma, O., Wijbrans, J., 2009. 40Ar/39Ar geochronologyusing a quadrupole mass spectrometer. Quaternary Geochronology 4, 508e516.

Siebel, W., Henjes-Kunst, F., Rhede, D., 1998. High-temperature memory in calcicamphiboles and constraints on compositional control of their 40Ar/39Ar ages.Geology 26, 31e34.

Song, S.G., Niu, Y.L., Zhang, L.F., 2007a. Two different metamorphic paths of eclo-gites and Differential exhumation of subducted Continental crust: a case studyof the dulan UHP Terrane in the North Qaidam UHP Belt. Geological Journal ofChina Universities 13515e13525 (in Chinese with English abstract).

Song, S.G., Niu, Y.L., Zhang, L.F., Bucher, K., 2009a. The Luliangshan garnet peridotitemassif of the North Qaidam UHPM belt, NW China - a review of its origin andmetamorphic evolution. Journal of Metamorphic Geology 27, 621e638.

Song, S.G., Niu, Y.L., Zhang, L.F., Zhang, G.B., 2009b. Time constraints on orogenesisfrom oceanic subduction to continental subduction, collision, and exhumation:an example from North Qilian and North Qaidam HP-UHP belt. Acta PetrologicaSinica 25, 2067e2077.

Song, S.G., Su, L., Li, X.H., Zhang, G.B., Niu, Y.L., Zhang, L.F., 2010. Tracing the 850-Macontinental flood basalts from a piece of subducted continental crust in theNorth Qaidam UHPM belt, NW China. Precambrian Research 183, 805e816.

Song, S.G., Su, L., Niu, Y.L., Zhang, L.F., Zhang, G.B., 2007b. Petrologicaland geochemical constraints on the origin of garnet peridotite in the NorthQaidam ultrahigh-pressure metamorphic belt, northwestern China. Lithos 96,243e265.

Song, S.G., Yang, J.S., Liou, J.G., Wu, C.L., Shi, R.D., Xu, Z.Q., 2003a. Petrology,geochemistry and isotopic ages of eclogites from the Dulan UHPM Terrane, theNorth Qaidam, NW China. Lithos 70, 195e211.

Song, S.G., Yang, J.S., Xu, Z.Q., Liou, J.G., Shi, R.D., 2003b. Metamorphic evolution ofthe coesite-bearing ultrahigh-pressure terrane in the North Qaidam, NorthernTibet, NW China. Journal of Metamorphic Geology 21, 631e644.

Song, S.G., Zhang, L.F., Chen, J., Liou, J.G., Niu, Y.N., 2005a. Sodic amphibole exso-lutions in garnet from garnet-peridotite, North Qaidam UHPM belt, NW China:implications for ultradeep-origin and hydroxyl defects in mantle garnets.Mineralogical Society of America 90, 814e820.

Song, S.G., Zhang, L.F., Niu, Y.L., Su, L., Jian, P., Liu, D.Y., 2005b. Geochronology ofdiamond-bearing zircons from garnet peridotite in the North Qaidam UHPMbelt, Northern Tibetan Plateau: a record of complex histories from oceaniclithosphere subduction to continental collision. Earth and Planetary ScienceLetters 234, 99e118.

Touret, J.L.R., 2001. Fluids in metamorphic rocks. Lithos 55, 1e25.

R. Hu et al. / Geoscience Frontiers 6 (2015) 759e770770

Turner, G., 1988. Hydrothermal fluids and argon isotopes in quartz veins and cherts.Geochimica et Cosmochimica Acta 52, 1443e1448.

Turner, G., Wang, S.S., 1992. Excess argon, crustal fluids and apparent isochronsfrom crushing K-feldspar. Earth and Planetary Science Letters 110, 193e211.

Villa, I.M., 1997. Direct determination of 39Ar recoil distance. Geochimica et Cos-mochimica Acta 61, 689e691.

Villa, I.M., 2001. Radiogenic isotopes in fluid inclusions. Lithos 55, 115e124.Wang, D.H., Xu, J., Chen, Y.C., Li, H.Q., Yu, J.J., 2003. Dating on the eclogite-hosted

quartz crystal and its significance for tracing the exhumation history of theUHP Belt in North Jiangsu. Acta Geologica Sinica 77, 544e548.

Wang, H., Wu, Y.B., Gao, S., Liu, X.C., Gong, H.J., Li, Q.L., Li, X.H., Yuan, H.L., 2011.Eclogite origin and timings in the North Qinling terrane, and their bearing onthe amalgamation of the South and North China blocks. Journal of MetamorphicGeology 29, 1019e1031.

Wang, S.S., Ge, N.J., Qiu, J., Sang, H.Q., 2000. Isotopic ages of phengite from eclogiteand quartz vein and their geological significance. Progress in Natural Science 10,1024e1028.

Wang, S.S., Turner, G., Burgess, R., 1988. An unusual 40Ar-36Ar age determination ofillite encased in chert. Chemical Geology 70, 18.

Whitney, D.L., Evans, B.W., 2010. Abbreviations for names of rock-forming minerals.American Mineralogist 95, 185e187.

Wijbrans, J.R., Pringle, M.S., Koppers, A.A.P., Scheveers, R., 1995. Argon geochro-nology of small samples using the Vulkaan argon laserprobe. ProceedingsKoninklijke Nederlandse Akademie van Wetenschappen 98, 185e218.

Wu, C.L., Gao, Y.H., Wu, S.P., Chen, Q.L., Joseph, L.W., Frank, K.M., Chris, M., 2007.Zircon SHRIMP U-Pb dating of granites from the Da Qaidam area in the northmargin of Qaidam basin, NW China. Acta Petrologica Sinica 23, 1861e1875 (inChinese with English abstract).

Wu, Y.B., Gao, S., Zhang, H.F., Yang, S.H., Liu, X.C., Jiao, W.F., Liu, Y.S., Yuan, H.L.,Gong, H.J., He, M.C., 2009. U-Pb age, trace-element, and Hf-isotope composi-tions of zircon in a quartz vein from eclogite in the western Dabie Mountains:constraints on fluid flow during early exhumation of ultrahigh-pressure rocks.American Mineralogist 94, 303e312.

Xiao, W.J., Windley, B.F., Yong, Y., Yan, Z., Yuan, C., Liu, C.Z., Li, J.L., 2009. EarlyPaleozoic to Devonian multiple-accretionary model for the Qilian Shan, NWChina. Journal of Asian Earth Sciences 35, 323e333.

Xiao, Y.L., Hoefs, J., Van den Kerkhof, A.M., Simon, K., Fiebig, J., Zheng, Y.F., 2002.Fluid evolution during HP and UHP metamorphism in Dabie Shan, China:constraints fromMineral chemistry, fluid inclusions and stable isotopes. Journalof Petrology 43, 1505e1527.

Xu, Z.Q., Yang, J.S., Ji, S.C., Zhang, Z.M., Li, H.B., Lai, L.F., Zhang, J.X., Wu, C.L., Hai, L.Z.,Liang, F.H., 2010. On the continental tectonics and dynamics of China. ActaPetrologica Sinica 84, 1e29 (in Chinese with English abstract).

Xu, Z.Q., Yang, J.S., Wu, C.L., Li, H.B., Zhang, J.X., Qi, X.X., Song, S.G., Qiu, H.J., 2006.Timing and mechanism of formation and exhumation of the Northern Qaidamultrahigh-pressure metamorphic belt. Journal of Asian Earth Sciences 28,160e173.

Yang, J.J., Zhu, H., Deng, J.F., Lai, S.C., Zhou, T.Z., 1994. The discovery of garnetperidotite in North Chaidam Mountains and its significance. Acta Petrologica EtMineralogica 13, 97e105 (in Chinese with English abstract).

Yang, J.S., Wu, C.L., Zhang, J.X., Shi, R.D., Meng, F.C., Wooden, J., Yang, H.Y., 2006.Protolith of eclogites in the north Qaidam and Altun UHP terrane, NW China:earlier oceanic crust? Journal of Asian Earth Sciences 28, 185e204.

Yang, J.S., Xu, Z., Song, S.G., Zhang, J.X., Wu, C.L., Shi, R.D., Li, H.B., Brunel, M., 2001.Discovery of coesite in the North Qaidam Early Palaeozoic ultrahigh

pressure (UHP) metamorphic belt, NW China. Comptes Rendus de l’Academiedes Sciences e Series IIA - Earth and Planetary Science 333, 719e724.

Yang, J.S., Xu, Z.Q., Zhang, J.X., 2002. Early Palaeozoic North Qaidam UHP meta-morphic belt on the north-eastern Tibetan plateau and a paired subductionmodel. Terra Nova 14, 397e404.

Yang, J.S., Xu, Z.Q., Zhang, J.X., Zhang, Z.M., Liu, F.L., Wu, C., 2009. Tectonic setting ofmain high-and ultrahigh-pressure metamorphic belt in China and adjacentregion and discussion on their subduction and exhumation mechanism. ActaPetrologica Sinica 25, 1529e1560.

Zhang, C., Zhang, L.F., Roermund, H.V., Song, S.G., Zhang, G.B., 2011a. Petrology andSHRIMP UePb dating of Xitieshan eclogite, North Qaidam UHP metamorphicbelt, NW China. Journal of Asian Earth Sciences 42, 752e767.

Zhang, G.B., Ellis, D.J., Christy, A.G., Zhang, L.F., Ling, N., Song, S.G., 2009a. UHPmetamorphic evolution of coesite-bearing eclogite from the Yuka terrane,North Qaidam UHPM belt, NW China. European Journal of Mineralogy 21,1287e1300.

Zhang, J.X., Christopher, G.M., Meng, F.C., Wan, Y.S., Tung, K.A., 2008. Polyphasetectonothermal history recorded in granulitized gneisses from the northQaidam HP/UHP metamorphic terrane, western China: evidence fromzircon U-Pb geochronology. Geological Society of America Bulletin 26,732e749.

Zhang, J.X., Mattinson, C.G., Meng, F.C., Yang, H.J., Wan, Y.S., 2009b. U-Pb geochro-nology of paragneisses and metabasite in the Xitieshan area, north QaidamMountains, western China: constraints on the exhumation of HP/UHP meta-morphic rocks. Journal of Asian Earth Sciences 35, 245e258.

Zhang, J.X., Mattinson, C.G., Yu, S.Y., Li, J.P., Meng, F.C., 2011b. UePb zircongeochronology of coesite-bearing eclogites from the southern Dulan area of theNorth Qaidam UHP terrane, northwestern China: spatially and temporallyextensive UHP metamorphism during continental subduction. Journal ofMetamorphic Geology 28, 955e978.

Zhang, J.X., Meng, F.C., Yang, J.S., 2003. Eclogitic metapelites in the western segmentof the north Qaidam basin and their geological implications. Geological Bulletinof China 22, 655e657.

Zhang, J.X., Meng, F.C., Yu, S.Y., Qi, X.X., 2007. Metamophic history recorded in highpressure mafic granulites in the Luliangshan Mountains to the north QaidamBasin, northwest China. Earth Science Frontiers 14, 85e97 (in Chinese withEnglish abstract).

Zhang, J.X., Yang, J.S., Mattinson, C.G., Xu, Z.Q., Meng, F.C., Shi, R.D., 2005. Twocontrasting eclogite cooling histories, North Qaidam HP/UHP terrane, westernChina: petrological and isotopic constraints. Lithos 84, 51e76.

Zhang, Z.M., Shen, K., Liou, J.G., Dong, X., Wang, W., Yu, F., Liu, F., 2011c. Fluid-rockinteractions during UHP metamorphism: a review of the Dabie-Sulu orogen,east-central China. Journal of Asian Earth Sciences 42, 316e329.

Zheng, Y.F., 2004. Fluid activity during exhumation of deep-subducted continentalplate. Chinese Science Bulletin 49, 985e998.

Zheng, Y.F., Fu, B., Gong, B., Li, L., 2003. Stable isotope geochemistry of ultrahighpressure metamorphic rocks from the Dabie-Sulu orogen in China: implicationsfor geodynamics and fluid regime. Earth-Science Reviews 62, 105e161.

Zheng, Y.F., Gao, T.S., Wu, Y.B., Gong, B., Liu, X.M., 2007. Fluid flow during exhu-mation of deeply subducted continental crust: zircon U-Pb age and O-isotopestudies of a quartz vein within ultrahigh-pressure eclogite. Journal of Meta-morphic Geology 25, 267e283.

Zong, K.Q., Liu, Y.S., Hu, Z.C., Kusky, T., Wang, D.B., Gao, C.G., Gao, S., Wang, J.Q., 2010.Melting-induced fluid flow during exhumation of gneisses of the Suluultrahigh-pressure terrane. Lithos 120, 490e510.