Journal Pre-proof

Permian ultrahigh–temperature reworking in the southern Chinese Altai: Evidence

from petrology, P–T estimates, zircon and monazite U–Th–Pb geochronology

Zhao Liu, Omar Bartoli, Laixi Tong, Yi–Gang Xu, Xiaolong Huang

PII: S1342-937X(19)30248-5
DOI: https://doi.org/10.1016/j.gr.2019.08.007
Reference: GR 2204

To appear in: Gondwana Research

Received Date: 15 January 2019

Revised Date: 26 July 2019
Accepted Date: 6 August 2019

Please cite this article as: Liu, Z., Bartoli, O., Tong, L., Xu, Y.–G., Huang, X., Permian ultrahigh–
temperature reworking in the southern Chinese Altai: Evidence from petrology, P–T estimates, zircon
and monazite U–Th–Pb geochronology, Gondwana Research, https://doi.org/10.1016/j.gr.2019.08.007.

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. 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.

© 2019 International Association for Gondwana Research. Published by Elsevier B.V. All rights
reserved.
 1 Permian ultrahigh–temperature reworking in the southern Chinese Altai:

2 Evidence from petrology, P–T estimates, zircon and monazite U–Th–Pb

3 geochronology

5 Zhao Liua, b, c, Omar Bartolic, Laixi Tongd, *, Yi–Gang Xua, Xiaolong Huanga

a
7 State Key Laboratory of Isotope Geochemistry, Guangzhou Institute of Geochemistry, Chinese

8 Academy of Sciences, Guangzhou 510640, China

b
9 University of Chinese Academy of Sciences, Beijing 100049, China

c
10 Dipartimento di Geoscienze, Università di Padova, Via Gradenigo 6, 35131 Padua, Italy

d
11 State Key Laboratory of Continental Dynamics, Department of Geology, Northwest University,

12 Xi’an 710069, China

13

14

15

16

17

*
18 Corresponding author. Tel: +86 29 88302312; Fax: +86 29 88302202.

19 E–mail address: [email protected] (L. Tong).

20

21

22
23 ABSTRACT

24

25 The Chinese Altai orogen formed in the Paleozoic is an important part of the Central Asian

26 Orogenic Belt (CAOB), and the study on the metamorphism will provide novel and robust

27 constraints on its tectonic evolution. In this study, we investigate our newly recognized garnet–

28 orthopyroxene–cordierite granulites at Wuqiagou area in the southern Chinese Altai. Detailed

29 petrographic study and P–T estimates suggest four distinct metamorphic stages of mineral

30 assemblages: (1) pre–peak (M1) stage containing the spinel–cordierite–bearing association or

31 biotite–plagioclase–quartz–bearing inclusion–phase assemblage, with P–T conditions of 3.0–4.0

32 kbar/700–750 °C; (2) peak ultrahigh–temperature (UHT) (M2) stage represented by relatively

33 coarse–grained garnet–orthopyroxene–cordierite–bearing porphyroblastic assemblage, with high–

34 Al2O3 contents (up to ~8.7 wt.%) in orthopyroxene and P–T conditions of ∼8.0 kbar/~980 °C; (3)

35 post–peak high–temperature granulite facies (M3) stage consisted of orthopyroxene–cordierite and

36 cordierite–quartz corona assemblages, formed during cooling and moderate decompression; and (4)

37 post–peak upper amphibolite facies (M4) stage represented by retrograde biotite–plagioclase–quartz

38 intergrowths. These four discrete metamorphic stages define an anticlockwise P–T path involving a

39 post–peak moderate decompression followed by nearly isobaric cooling process. LA–ICP–MS U–

40 Pb age dating results of metamorphic zircons for UHT samples show two weighted mean ages of

41 ~390 Ma and ~280 Ma. We propose that the M1 stage might occur in the middle Devonian, whereas

42 the near–peak UHT stage probably occurred in the early Permian. The Permian UHT

43 metamorphism was further supported by the monazite U–Th–Pb dating results (287.9 ± 2.1 Ma),

44 reflecting a prominent HT–UHT reworking event in the late Paleozoic. We proposed that the
45 Permian–age UHT reworking event in the southern Chinese Altai probably occurred in a post–

46 orogenic or intraplate extensional tectonic setting associated with the input of external heat, related

47 to the underplating of deep–derived magma as a result of the Tarim mantle plume activity.

48

49 Keywords: UHT metamorphism; Chinese Altai; P–T path; Geochronology; Mantle plume

50

51 1. Introduction

52

53 Ultrahigh temperature (UHT) metamorphism is the most thermally extreme type of crustal

54 metamorphism, with temperatures exceeding 900 °C at moderate pressure (7–13 kbar; 20–40 km)

55 (Kelsey, 2008; Harley, 2008; Santosh et al., 2012). The study of UHT terranes can provide

56 important insights into the formation and evolution of deep continental crust (Kelsey and Hand,

57 2015; Korhonen et al., 2014). Some mineral assemblages in rocks of highly aluminous and

58 magnesian bulk composition (Mg–, Al–rich granulites), including assemblages such as sapphirine +

59 quartz, orthopyroxene (> 8.0 wt.% Al2O3) + sillimanite ± quartz, low Zn spinel + quartz and

60 osumilite + garnet, are the key indicators of such extreme metamorphic conditions (Harley, 1998,

61 2008; Tsunogae et al., 2011; Santosh et al., 2012). But under highly oxidizing or some other certain

62 conditions, even these assemblages may be stabilized below 900 °C (Kelsey et al., 2008).

63 Conventional geothermometers, which are mainly based on Fe–Mg exchange reactions, generally

64 yield erroneous low temperature estimates due to post–peak diffusional cation exchange (e.g.,

65 Harley, 1989). This drawback, however, could be potentially overcome by using alternative

66 approaches such as garnet–orthopyroxene thermobarometer based on Al–solubility in

67 orthopyroxene (Pattison et al., 2003), two–feldspar thermometry (Jiao et al., 2011), Ti–in–zircon

68 thermometry (Ferry and Watson, 2007), and Zr–in–rutile thermometry (Tomkins et al., 2007).

69 A long–term dispute exists on the mechanisms which make crust extremely hot with

70 geothermal gradient of ≥ 20 °C km–1 (Brown, 2007; Kelsey, 2008; Santosh and Kusky, 2010), to

71 produce UHT granulites. Although about 60 UHT granulite terranes of Neoarchean to Miocene

72 have been identified so far (Brown, 2007; Kelsey, 2008; Pownall et al., 2014; Kelsey and Hand,

73 2015), only few UHT metamorphic events are considered to have occurred in the past 500 Ma (e.g.,

74 Nam et al., 2001; Zhao et al., 2010; Galli et al., 2011; Li et al., 2014; Pownall et al., 2014; Tong et

75 al., 2014; Zhang et al., 2015b). Underplating of hot mafic magmas (Guo et al., 2012; Li et al., 2014;

76 Tong et al., 2014) or exhumation of subcontinental lithospheric mantle (Pownall et al., 2014) have

77 been proposed as heat sources for UHT metamorphism of crustal rocks.

78 The Chinese Altai orogenic belt formed in the Paleozoic is an important part of the Central

79 Asian Orogenic Belt (CAOB), accompanying with remarkable metamorphism (e.g., Li et al., 2004,

80 2014; Wei et al., 2007; Wang et al., 2009b; Tong et al., 2014; Broussolle et al., 2018). The early

81 accretionary wedge in the Chinese Altai pervasively experienced a middle Devonian tectono–

82 metamorphic event (e.g., Wei et al., 2007; Jiang et al., 2010). The Devonian orogenic architecture

83 was subsequently reworked by the HT–UHT metamorphism on its southern margin (e.g., Wang et

84 al., 2009b; Tong et al., 2013, 2014; Jiang et al., 2015; Broussolle et al., 2018). Contrasting P–T

85 paths and tectonic models were proposed to explain the late Paleozoic UHT metamorphism in the

86 southern Chinese Altai:

87 (i) Altai UHT granulites followed a clockwise P–T path (Li et al., 2010; Yang et al., 2015b), with a

88 first isothermal decompression (ITD) and a second isobaric cooling (IBC) retrograde P–T paths
 89 (Yang et al., 2015b). Li et al. (2010) correlated the UHT metamorphism with the collisional

90 orogeny between Siberia and Kazakhstan–Junggar plate.

91 (ii) Altai UHT granulites from Wuqiagou area were exhumed along an anticlockwise P–T trajectory

92 with a post–peak near–ITD process at 5–6 kbar/890–940 °C (Li et al., 2014). Slab break–off

93 which caused asthenospheric upwelling and heat flux at 320–290 Ma might contribute to Altai

94 UHT metamorphism (Li et al., 2014; Yang et al., 2015b);

95 (iii) P–T conditions of the Altai UHT granulites from Kalasu area define an anticlockwise P–T path

96 of initial prograde heating and increase in pressure followed by a post–peak near–IBC process

97 (Tong et al., 2014). Underplating and heating of mantle–derived mafic magma as a result of the

98 Tarim mantle plume might provide the heat flux necessary for the Permian HT–UHT

99 metamorphism (Wang et al., 2014; Tong et al., 2013, 2014; Yang et al., 2015a; Liu and Tong,

100 2015).

101 Accordingly, the P–T path of the Altai UHT granulites is not well constrained and the

102 tectonothermal evolution of the southern Chinese Altai is still ambiguous and debated. In this study,

103 we investigate the petrology, mineral chemistry, zircon and monazite geochronology, and P–T

104 trajectory of garnet–orthopyroxene–cordierite–bearing UHT granulites at Wuqiagou area. UHT

105 conditions are retrieved from the high alumina contents of orthopyroxene (~8.7 wt.% Al2O3),

106 conventional geothermobarometry calculations and preliminary phase equilibria modelling. The P–

107 T path recorded in the UHT metapelitic granulites will place important constraints on the

108 continental crustal growth and tectonothermal evolution of the southern Chinese Altai in the

109 Paleozoic.

110
111 2. Geological background

112

113 The Altai orogen is an important part of the CAOB (Jahn et al., 2004). The NW–SE trending

114 Chinese Altai orogenic belt is bounded by the Siberian plate to the north and the Kazakhstan–

115 Junggar plate to the south (Windley et al., 2007). The Chinese Altai orogen comprises various

116 lithological types, mainly including volcanic, pyroclastic and metasedimentary rocks, high–grade

117 metamorphic rocks and large amounts of granitoids (Windley et al., 2002; Jiang et al., 2015).

118 Five fault–bounded terranes have been identified based on the stratigraphy, metamorphism,

119 deformation patterns and chronology (Fig. 1) (Windley et al., 2002; Wang et al., 2006, 2009a).

120 Terrane Ⅰ consists mainly of the late Devonian to the early Carboniferous meta–clastic rocks and

121 limestone intercalated with minor arc–like volcanic rocks. Terrane Ⅰ is composed mainly of the

122 Neoproterozoic to the middle Ordovician sedimentary and volcanic rocks of the Habahe Group

123 (Yuan et al., 2007), which experienced lower greenschist facies metamorphism. Terrane Ⅰ in the

124 central part of the Altai orogenic belt is the largest one and is composed mainly of the early Silurian

125 and the early Devonian flysch sequence of the Habahe Formation (Long et al., 2010), among which,

126 the minor ~502 Ma felsic volcanic rocks which experienced greenschist– to upper amphibolite

127 facies metamorphism have been interpreted to represent components of a continental arc (Windley

128 et al., 2002; Yang et al., 2011). Terrane Ⅰ consists of the late Silurian to the early Devonian arc–like

129 volcanic and pyroclastic rocks in the lower part and the middle Devonian turbidites and pillow–

130 basalts in the upper part, showing a spectrum of metamorphic zones from greenschist to upper

131 amphibolite and locally granulite facies conditions (Wang et al., 2009b; Tong et al., 2014). Terrane

132 Ⅰ is bounded by the Erqis fault fault in the south. It includes a complex sequence including a
133 possible Precambrian basement, the early Paleozoic–Devonian sediments and the late

134 Carboniferous volcanoclastics, metamorphosed under greenschist to amphibolite facies conditions.

135 Mafic granulites and UHT pelitic granulites were reported from this terrane (Li et al., 2004, 2010,

136 2014; Chen et al., 2006; Yang et al., 2015b; Liu and Tong, 2015). Rocks in the Junggar plate (south

137 of the Erqis fault belt) are dominated by the Devonian–Carboniferous volcanoclastics, which have

138 been metamorphosed to greenschist facies.

139 The tectonic evolution of the Chinese Altai orogen mainly involves five stages based on

140 previous studies (e.g., Windley et al., 2002, 2007; Wang et al., 2006; Niu et al., 2006; Jiang et al.,

141 2010; Yang et al., 2011; Zhang et al., 2012; Wang et al., 2009b, 2014; Tong et al., 2014; He et al.,

142 2018): (i) a passive continental margin or peri–Gondwana terrane during the Neoproterozoic–early

143 Paleozoic; (ii) the development of a late Silurian to early Devonian arc environment related to the

144 northward subduction of Junggar plate; (iii) continent–arc collision, subduction of ridge, or the

145 development of a possible back–arc basin in the middle to late Devonian; (iv) Permian (300–260

146 Ma) post–orogenic setting with a possible overprinting by the Tarim mantle plume; and (v)

147 intraplate magmatism beginning in the Jurassic.

148 High–grade gneissic rocks crop out extensively in the terranes Ⅰ and Ⅰ and are currently

149 assigned to the Kemuqi and Fuyun Groups. Their presumed Precambrian age is widely accepted by

150 many workers and led to the proposal of an Altai–Mongolia Precambrian basement or

151 microcontinent (Windley et al., 2002; Li et al., 2006). However, the above interpretation is not

152 supported by SHRIMP zircon U–Pb data for high–grade metamorphic rocks from the southeastern

153 part of the Chinese Altai, which, instead, suggested that these high–grade rocks metamorphosed in

154 the Paleozoic (Chen et al., 2006; Wang et al., 2009b; Jiang et al., 2010; Li et al., 2014; Tong et al.,
155 2014).

156 Two metamorphic events have been documented in the Chinese Altai orogenic belt (Broussolle

157 et al., 2018). The first tectono–metamorphic cycle was dated at 390–365 Ma (e.g., Zhuang, 1994;

158 Hu et al., 2002; Zheng et al., 2007; Jiang et al., 2010), and considered to be linked with two distinct

159 metamorphic stages characterized by M1 Barrovian–type MT–MP and M2 Buchan–type HT–LP

160 field gradients, with metamorphic degrees ranging from greenschist– to amphibolite facies

161 conditions (up to 750 °C; Jiang et al., 2010, 2015). In terms of the tectonic process in the Devonian,

162 researchers have envisaged models of active continental margin (He et al., 2018), arc–continent

163 collision (Windley et al., 2002; Wei et al., 2007), back–arc spreading (Wang et al., 2006), slab

164 break–off (Niu et al., 2006) or ridge–subduction and the development of slab–window (Windley et

165 al., 2007; Jiang et al., 2010).

166 Recently, many HT–UHT metamorphic rocks were reported from the southern Chinese Altai,

167 which mainly consisted of HT–UHT metapelitic granulites (Wang et al., 2009b; Li et al., 2014;

168 Tong et al., 2014; Liu and Tong, 2015; Yang et al., 2015b), mafic granulites (Li et al., 2004; Chen

169 et al., 2006; Liu and Tong et al., 2015), calc–silicate granulites (Yang et al., 2015a). The Permian

170 high–grade rocks are mainly cropping out along NW–SE trending zone in the terranes Ⅰ and Ⅰ

171 (e.g., Tong et al., 2014; Liu and Tong, 2015; Broussolle et al., 2018). Zircon U–Pb and monazite

172 U–Th–Pb geochronology of Altai HT–UHT granulites and gneisses yielded metamorphic ages of

173 293–260 Ma (Chen et al., 2006; Briggs et al., 2007; Zheng et al., 2007; Wang et al., 2009b; Li et al.,

174 2014; Tong et al., 2014), which were interpreted by Broussolle et al. (2018) as the second tectono–

175 metamorphic cycle in the Chinese Altai orogeny.

176 Our samples were collected at Wuqiagou area (Fig. 2). The UHT granulites occur as lenses
177 within garnet–biotite–plagioclase gneiss, with the country rocks generally intruded by granitic dikes.

178 The Altai UHT granulites experienced partial melting as testified by migmatitic appearance (Fig. 3).

179

180 3. General petrography and reaction history

181

182 The Altai UHT granulite samples typically preserve a porphyroblastic texture composed of

183 garnet (12–15%), orthopyroxene (15–20%), cordierite (13–15%), sillimanite (~3%), spinel (~5%),

184 biotite (15–20%), plagioclase (10–12%), quartz (20–25%), with accessory anorthite, zircon,

185 monazite, apatite, rutile and Fe–Ti oxides (< 5%).

186

187 M1 stage: interpretation of the prograde evolution

188 Garnet, orthopyroxene and cordierite porphyroblasts preserve a variety of single–phase or

189 multiphase mineral inclusions, which are ascribed to the pre–peak assemblages. These inclusions

190 show various sizes and shapes, generally with no significant preferred orientation. Tiny composite

191 grains of spinel + cordierite ± sillimanite are present in garnet and cordierite porphyroblasts (Figs.

192 4b and c). Spinel, in addition to occurring as inclusions preserved in garnet, is also found in the

193 matrix, and generally associated with cordierite, fibrous sillimanite, biotite, Fe–Ti oxides and/or

194 anorthite (Figs. 4a and b). The spl1 + crd1 assemblage might reach equilibrium by the following

195 reaction (Bindu, 1997; White et al., 2007):

196 Bi + Sill ± Qtz → Sp1 + Crd1 ± Ksp + Liq (1)

197 The spinel–bearing assemblages were most likely to be formed in the pre–peak stage because

198 they are occasionally preserved as inclusions in garnet. Spinel in direct contact with quartz was
199 never observed, which is different from the nearby UHT granulites reported by Li et al. (2014).

200 Further inclusions in garnet porphyroblasts are biotite, plagioclase, quartz, Fe–Ti oxides and more

201 rarely, intergrowths of cordierite + magnetite ± biotite (Fig. 4b). Porphyroblastic orthopyroxene

202 generally encloses biotite, plagioclase, quartz and Fe–Ti oxides (Figs. 4e and f).

203 In conclusion, pre–peak (M1) assemblage mainly comprised garnet cores, spinel (spl1),

204 cordierite (crd1), biotite (bi1), plagioclase (pl1), sillimanite, quartz and Fe–Ti oxides.

205

206 M2 stage: peak metamorphic assemblage

207 Garnet (grt2), orthopyroxene (opx2) and texturally equilibrated cordierite (crd2) porphyroblasts

208 probably represent the peak metamorphic assemblage. Magnetite, rutile and minor ilmenite are

209 present in the rock matrix (Figs. 4d, 5e and f). As described above, the growth of garnet and

210 orthopyroxene might document progress of the following reaction (Vielzeuf and Montel, 1994):

211 Bi1 + Pl1 + Qtz1 → Grt2 + Opx2 ± Ksp + Liq (2)

212 Cuspate to lobate grains are present in leucocratic portions (Fig. 4g). These microstructures are

213 considered to be melt pseudomorphs (Holness and Sawyer, 2008; Sawyer, 2008) and are indicative

214 of crystallization of localized melts among mineral boundaries. Some polycrystalline inclusions

215 occur isolated or are localized in clusters within garnet porphyroblasts (Fig. 4h). They often display

216 a round to negative crystal shape, range from 1 to 8 µm in diameter and contain multiple daughter

217 crystals of biobite, plagioclase, quartz and ilmenite (Figs. 4i and j), resembling the melt inclusions

218 (MI) described in peritectic garnet of other migmatitic and granulitic terranes (e.g., Acosta-Vigil et

219 al., 2010; Cesare et al., 2015; Bartoli et al., 2013, 2014, 2016). The observed inclusions would

220 represent an additional evidence of the former presence of melts in these rocks.
221 The garnet porphyroblasts are typically pale pink in colour and 0.7–2.0 mm across (Figs. 4b, 5c

222 and g). Orthopyroxenes are generally pale grayish to dark brown, and occur as anhedral grains and

223 have sizes ranging from 0.1 to 2.5 mm (Figs. 4e, f, 5a, e and f).

224 Cordierite grains vary in size from 0.02 to 2 mm and locally contain spinel, tabular biotite and

225 sillimanite needles (Figs. 4a and c). The peak cordierite might represent the only hydrous phase

226 during the peak metamorphic stage, suggesting a notably dry composition.

227

228 M3 stage: corona textures formation

229 Opx2 is extensively replaced by corona textures, which are mainly composed of orthopyroxene

230 (opx3), fine–grained cordierite (crd3) and Fe–Ti oxides (Figs. 5a and b). These opx3–crd3

231 symplectitic rinds are interpreted to be formed subsequently to the peak metamorphic stage and they

232 probably are the result of the release of the Mg–Tschermak’s component from opx2 according the

233 following retrograde reaction (Brandt et al., 2003):

234 High–Al Opx2 + Qtz → low–Al Opx3 + Crd3 (3; Fig. 6a)

235 Garnet may be mantled by cordierite moats (crd3), tiny biotite flakes and vermiform quartz,

236 with a distinct ‘spongy’ appearance (Figs. 5c and d). These microstructures likely reflect the melt–

237 consuming reaction (Cenki et al., 2002):

238 Grt + Liq → Crd + Bt + Qtz (4)

239 Formation of the retrograde opx3–crd3 corona texture was interpreted to be resulted from

240 decompression (Brandt et al., 2003). The melt consuming reaction 4 requires a combination of

241 cooling and decompression, which is consistent with an uplift path (Cenki et al., 2002).

242
243 M4 stage: formations of retrograde garnet, biotite, plagioclase and quartz

244 Retrograde platy and needle–like biotite (bi4), fine–grained plagioclase (pl4) and quartz might

245 form subsequently to the M3 stage around garnet and orthopyroxene porphyroblast rims (Figs. 5f

246 and g). The succession of M3 and M4 reactions (i.e., the M4 stage is likely to have occurred after

247 M3) is clearly visible in Figure 5f, where bi4–pl4–qtz intergrowths are locally grown around opx3 +

248 crd3 symplectites. Bi4–pl4–qtz intergrowths were generally considered to be resulted from

249 interaction of the melt with minerals, suggesting the progress of the following retrograde bi–

250 forming hydration reaction (Brandt et al., 2003; Holness et al., 2011):

251 Opx (Grt) + Liq → Bi4 + Pl4 + Qtz. (5)

252 Symplectitic opx3 is locally overgrown by fine–grained anhedral garnet (< 0.1 mm) (Fig. 5e).

253 These small garnets locally contain inclusions of opx3, indicating the growth of a second generation

254 of garnet at the expense of orthopyroxene. This garnet growth is usually considered to be related to

255 cooling (Brandt et al., 2003) linked to the following reaction (Harley, 1989):

256 Opx + Crd → Grt4 + Qtz. (6; Fig. 6a)

257 In places, some rutile needles were also observed in garnet (Fig. 5h). These rutile needles were

258 probably precipitated during M4 stage or further cooling under subsolidus conditions.

259

260 4. Whole–rock geochemistry and mineral chemistry

261

262 4.1. Whole–rock geochemistry

263

264 Major element oxides (wt.%) were determined on fused glass disks with a 1:8 sample to
265 Li2B4O7 flux ratio, using a Rigaku ZSX100e X–ray fluorescence (XRF) spectrometer in the Key

266 Laboratory of Isotope Geochronology and Geochemistry, Guangzhou Institute of Geochemistry.

267 The accuracy of the analyses is within 1 % for most major elements. Sample preparation techniques

268 and other details of procedures are described in the reference (Long et al., 2011). The geochemical

269 data are presented in Table 1.

270 The Altai UHT metapelitic granulites show variable SiO2 contents (51.42–56.49 wt.%). They

271 typically have moderate Al2O3 and MgO contents, with A/AFM (Al/(Al + Fetotal + Mg)) and Mg#

272 values of 0.32–0.36 and 52.03–52.77, respectively. The K2O + Na2O contents are in the range of

273 2.35–2.56 wt.%, with K2O higher than Na2O. P2O5 and MnO contents are negligible (< 0.39 wt.%).

274 The LOI (= loss on ignition) is normally ranging from 1.10 to 1.32 wt.%. In S(SiO2)–A(Al2O3 +

275 Fe2O3)–FM(FeO + MgO) diagram, they are plotted in the grt–opx–crd–qtz field (Fig. 6), which is

276 consistent with the major minerals observed in these samples.

277

278 4.2. Mineral chemistry

279

280 Mineral compositions were analyzed with a JXA–8100 microprobe at State Key Laboratory of

281 Isotopic Geochemistry, Guangzhou Institute of Geochemistry, Chinese Academy of Sciences, with

282 an accelerating voltage of 15 kv, a beam current of 3x10–8 Å, a beam width of 1 µm, and data

283 correction by using a ZAF method. Structural foumulae are given for fixed oxygen values with Fe3+

284 calculated by stoichiometric charge balance. Representative mineral compositions for the studied

285 samples are listed in Table 2 and Table 3.

286 Garnet is essentially an almandine–pyrope solid solution with minor spessartine and grossular
287 components (Alm61–72Py31–17Sps5–8Grs3–3) (Fig. 6). Slight chemical variations are observed for

288 garnets of distinct generations. The garnet porphyroblasts exhibit an obvious compositional zoning,

289 with a typically “M–like” profile from the core to rim (Figs. 6d and f). In details, the garnets display

290 a relatively low pyrope content in the core (Alm68–64Py22–28Sps7–4Grs3–4) and a slight outward

291 increase in Mg# from core (Mg# = 24.39) to mantle (Mg# = 32.90–33.85). Thereafter, the

292 composition becomes less magnesian rimward to Alm70–71Py18–18Sp7–8Grs5–3 (Mg# = 20.22–20.45).

293 Furthermore, the low Mg# contents of garnet rims are probably associated with the production of the

294 neighboring symplectites of bi4–pl4–qtz. Late grt4 grains have a much stricter compositional

295 variation of Alm65–68Py25–22Sps7–7Grs3–3.

296 Orthopyroxene from different positions has remarkably different Al2O3 contents. The core of

297 the orthopyroxene porphyroblast (Opx2) occasionally preserves markedly higher Al2O3 contents (up

298 to ~8.7 wt.%) than those of its rim (5.48–7.25 wt.%), with the corresponding XAl ( = Al/2) values of

299 0.19 and 0.12–0.16 in the formula, respectively. Symplectitic orthopyroxene (Opx3) shows wider

300 and lower Al2O3 contents (1.94–5.50 wt.%). However, they have similar XMg values ranging from

301 0.57 to 0.64.

302 Spinel inclusions preserved in garnet have moderate ZnO contents (~2.33 wt.%), relatively

303 higher Cr2O3 contents (~0.42 wt.%), and the highest XMg values (~0.34). The matrix spinel grains

304 contain ZnO of 0.82–3.25 wt.%, Cr2O3 of 0.15–0.39 wt.% and XMg of 0.21–0.27. The calculated

305 Fe2O3 contents of spinel by stoichiometric charge balance in this study generally vary from 1.83 to

306 2.86 wt.%, but occasionally as low as 0.36 wt.%, indicative of lower oxygen fugacity.

307 In all analyzed samples, the total sum of oxides in the cordierites is much lower than 100%,

308 suggesting the presence of H2O + CO2 in the structure. They have highly magnesian compositions
309 with XMg values ranging from 0.76 to 0.85. Cordierite included in garnet has the highest XMg value

310 of ~0.85. Peak Crd2 has relatively lower XMg values (0.76–0.78) than Crd1 (~0.80) stable with

311 spinel. Crd3 in the symplectitic intergrowths of opx3–crd3 has a typical XMg value of ~0.81 and crd3

312 moats around garnets have XMg values of 0.80–0.84.

313 Biotite preserved in garnet porphyroblasts has higher XMg values (0.66–0.72) and lower TiO2

314 contents (1.5–3.2 wt.%) than biotite in the spinel–bearing domains (XMg = 0.55–0.58, TiO2 = 4.80–

315 5.00). The matrix biotite shows XMg values of 0.59–0.65 and TiO2 contents of 1.81–3.35. Biotite

316 flakes display an increase in XMg values from the core (~0.61) to the rim (~0.64).

317 Both of matrix–phase and inclusion–phase plagioclase are Na–rich (An32–43Ab68–57), with no

318 obvious chemical compositional zoning. In most cases, M4 plagioclase has lower anorthite contents

319 (An24–28) than those in other positions. Anorthite in this study has XAn of 0.87–0.94 and XAb of

320 0.13–0.06.

321 Sillimanite typically contains Fe2O3 contents of 0.86–0.93 wt.%.

322

323 5. Metamorphic P–T evolution

324

325 5.1. P–T estimates

326

327 Representative texturally–equilibrated mineral pairs and assemblages are used to estimate P–T

328 conditions through geothermobarometers and average P–T approach. The P–T estimates are

329 tabulated in Table 4.

330 Traditional thermobarometers usually underestimate the peak temperatures of UHT

331 metamorphism as they fail to consider the effects of Fe2+–Mg reset that can occur between mineral

332 pairs during the post–peak cooling process. Therefore, in this contribution, we adopt the garnet–

333 orthopyroxene thermobarometer corrected by Pattison et al. (2003) to estimate the peak P–T

334 conditions of the studied UHT metapelitic granulites. Garnet mantles and high–Al orthopyroxene

335 yielded peak temperatures of 915–1024 °C, with an average value of ~973 °C, approximately

336 representing the temperature of peak metamorphism. The average P–T calculation method of

337 Powell and Holland. (1994) can also be utilized to estimate peak P–T conditions, and the peak

338 assemblage of grt + opx + crd + q + mt + ilm + ru + H2O provides P–T estimates of ~8.8

339 kbar/~980 °C.

340 The pre–peak (M1) assemblage garnet (core)–plagioclase–biotite–quartz (GBPQ)

341 thermobarometry results in P–T conditions of ~4.1 kbar/~690 °C (Holdaway, 2000; Wu et al., 2004).

342 The M3 assemblage gives an average P–T condition of ~7.0 kbar/~790 °C (Powell and Holland,

343 1994). For the post–peak (M4) stage metamorphism, GBPQ thermobarometry yields results of ~6.8

344 kbar/~715 °C (Holdaway, 2000; Wu et al., 2004).

345

346 5.2. Phase equilibria modelling

347

348 The phase equilibria modelling for the UHT granulites sample FY15–49 was carried out using

349 Perple_X (Connolly, 2005) with the internal thermodynamic data set of Holland and Powell, 1998

350 (updated November 2003) in the MnO–Na2O–CaO–K2O–FeO–MgO–Al2O3–SiO2–H2O–TiO2–

351 Fe2O3 (MnNCKFMASHTO) chemical system. The phases considered in the calculation include:

352 garnet (g), orthopyroxene (opx), plagioclase (pl), K–feldspar (ksp), cordierite (crd), biotite (bi),
353 muscovite (ms), sillimanite (sill), kyanite (ky), ilmenite (ilm), magnetite (mt), spinel (sp), quartz

354 (qtz), rutile (ru) and silicate melt (liq). Among these phases, sillimanite, kyanite, rutile and quartz

355 are considered as pure end member phases. The activity–composition (a–x) models for garnet and

356 orthopyroxene are from Holland and Powell. (2001), plagioclase from Newton et al. (1980), K–

357 feldspar from Thompson and Hovis. (1979), cordierite and spinel from Holland and Powell. (1998),

358 biotite from Tajčmanová et al. (2009), muscovite from Coggon and Holland. (2002), ilmenite and

359 magnetite from White et al. (2000), and silicate melt from White et al. (2007).

360

361 5.2.1. Peak and post–peak evolution

362

363 The investigated rock is a residual sample as a consequence of melt loss (SiO2 = 51 wt.%; FeO

364 + MgO > 20 wt.%). A substantial melt loss is also consistent with the degree of preservation of the

365 peak metamorphic assemblage (White and Powell, 2002). Therefore, the bulk rock composition

366 obtained from XRF analysis is only appropriate for investigating the peak P–T conditions and the

367 post–peak retrograde evolution (White et al., 2004). The X(O) value (representative of ferric iron

368 content) was fixed at 0.28 mol.%, in agreement with values assumed by Li et al., (2014) for similar

369 rocks in this region. The amount of H2O component involved in the calculation was assumed as the

370 loss of ignition (LOI) of XRF analysis.

371 The P–T pseudosection for sample FY15–49 was constructed for the ranges of 2.0–9.0 kbar and

372 700–1100 °C. As is shown in Figure 8a, the solidus is predicted at 750–770 °C for P <5.5 kbar,

373 reflecting the residual bulk chemical composition of this sample. The peak phase assemblage (grt +

374 opx + crd + qtz + mt + ru + liq ± ilm) is represented by the orange fields in Figure 8a and is
375 predicted to be stable at 7–8 kbar and 950–1050 °C, compatible with P–T estimates from the

376 mineral pairs within errors. Rutile– and orthopyroxene–out curves define the lower and upper

377 pressure limits, respectively, whereas biotite disappears just before reaching peak conditions. The

378 proportion of melt predicted at peak conditions is 35–40 vol.%, significantly lower than the amount

379 expected from UHT anatexis of fertile protoliths. This discrepancy is likely due to the residual bulk

380 composition used for calculations.

381 M3 stage is characterized by the formation of opx3–crd3 symplectitic rinds around

382 orthopyroxene and of cordierite + biotite + quartz replacing garnet (see above). In order to consume

383 garnet and melt and produce orthopyroxene, cordierite, biotite and quartz, the post–peak P–T

384 evolution should be characterized by a combination of cooling and decompression (Figs. 8b–g).

385 However, the subsequent M4 stage needs a near–IBC path to form bi4–pl4–qtz intergrowths and

386 consume orthopyroxene, garnet and melt (Figs. 8b–g). Notably, the proposed near–IBC path can

387 also explain the formation of small garnets via reaction (6) at T < 750 °C.

388

389 5.2.2. Prograde evolution

390

391 The reconstruction of a probable bulk chemical composition for the protolith is needed to

392 recover the prograde history of melt–depleted granulites. This approach involves the reintegration

393 of a certain amount of melt to the residual composition and the calculation of phase equilibria for

394 the new protolith composition (Bartoli, 2017). The single–step approach was adopted for the

395 selected sample: the composition of melt in equilibrium with the inferred peak mineral assemblage

396 was calculated at ~7.6 kbar and 1000 °C (SiO2 = 69.50, Al2O3 = 16.67, FeO = 2.55, MgO = 0.77,
397 CaO = 0.96, Na2O = 1.81, K2O = 4.78, H2O = 2.95 wt.%) and an amount of this melt (30%),

398 sufficient to produce a H2O–saturated solidus at ~9 kbar and < 700 °C, was reintegrated.

399 Figure 9 represents the P–T pseudosection for the melt–reintegrated composition. The most

400 evident changes in phase diagram topology are i) the shift of the solidus to lower temperatures at P

401 < 5.5 kbar, ii) the H2O–saturated character of entire solidus curve from 2 to 9 kbar and iii) a

402 reduction of the stability field of orthopyroxene at T < 900 °C (Fig. 9). However, this modelling

403 also fails to predict the pre–peak (M1) mineral phase assemblage. For instance, sillimanite is

404 present only at P > 7 kbar whereas spinel is not predicted to be present. This discrepancy could be

405 related to the presence of chemical–mineralogical microdomains during the prograde evolution of

406 these rocks (Guevara and Caddick, 2016). In this case, the bulk rock composition cannot be

407 representative of the effective bulk composition (EBC) from which such a spinel–bearing mineral

408 assemblage grew.

409 To investigate the possible effect of local variations of the EBC, a T–XMg section based on the

410 melt–reintegrated composition was constructed at 3 kbar, from 650 to 850 °C, with the XMg value

411 ranging from 0 to 1 (Fig. 10). Orthopyroxene–bearing assemblages appear when XMg > 0.2–0.4,

412 whereas garnet is stable for XMg < 0.6–0.7. Assemblages containing spinel or sillimanite are present

413 in the low–XMg side of the diagram (< 0.1). For these low–XMg effective bulk compositions, the

414 cordierite–spinel and cordierite–spinel–sillimanite pre–peak assemblages are predicted to be stable

415 at ~3 kbar, 730–750 °C. The model also indicates that some amounts of garnet could have been

416 already produced during the pre–peak evolution, in agreement with petrographic inferences.

417

418 6. Zircon and monazite U–Th–Pb geochronology

419

420 6.1. Analytical methods

421

422 Conventional magnetic and heavy liquid techniques followed by hand–picking under a

423 binocular microscope were used for separation of zircons from the UHT metapelitic granulites

424 FY15–49 and FY15–51. The morphology and internal structure of the zircons were documented

425 with transmitted and reflected light microphotographs and cathodoluminescene (CL) images.

426 Zircon and monazite U–Th–Pb dating of the samples FY15–49 and FY15–51 were carried out

427 using LA–ICP–MS at the Wuhan Sample Solution Analytical Technology Co., Ltd., Wuhan, China.

428 Detailed operating conditions for the laser ablation system and the ICP–MS instrument and data

429 reduction are described in Zong et al. (2017). Laser sampling was performed using a GeolasPro

430 laser ablation system that consists of a COMPexPro 102 ArF excimer laser (wavelength of 193 nm

431 and maximum energy of 200 mJ) and a MicroLas optical system. Each analysis incorporated a

432 background acquisition of approximately 20–30 s followed by 50 s of data acquisition from the

433 sample.

434 The spot size and frequency of the laser for zircons were set to 32 µm and 10 Hz, respectively.

435 Zircon 91500 and glass NIST610 were used as external standards for U–Pb dating and trace

436 element calibration, respectively. The spot size and frequency of the laser for monazites were set to

437 16 µm and 2 Hz, respectively. Monazite standard 44069 and glass NIST610 were used as external

438 standards for U–Pb dating and trace element calibration, respectively.

439 An Excel–based software ICPMSDataCal was used to perform off–line selection and

440 integration of background and analyzed signals, time–drift correction and quantitative calibration
441 for trace element analyses and U–Pb dating (Liu et al., 2010). Software SQUID 1.0 and ISOPLOT

442 (V.3.0; Ludwig, 1999) were used for data processing.

443

444 6.2. Zircon morphology and U–Pb geochronology

445

446 Representative zircon CL images of UHT metapelitic granulites from the southern Chinese

447 Altai are shown in Figure 11. LA–ICP–MS zircon U–Pb analysis data and age results for the sample

448 FY15–49 are shown in Figure 12a and listed in Table 5. Zircons in this sample are subhedral to

449 anhedral, 60–120 µm in size, and elongate, prismatic, stubby and occasionally round in shape. They

450 usually have core–rim structures, with dark, oscillatory or sector–zoned cores mantled by anhedral

451 overgrowths. Some weekly–zoned or sector–zoned dark cores with bright overgrowths, are also

452 interpreted to be produced during the metamorphism. Thirty–three spots were analyzed on thirty

453 zircon grains from sample FY15–49, with U concentrations of 133–2877 ppm and Th/U ratios of 0–

454 0.87. Zircon rims have relatively low Th/U ratios (≤0.26, mostly less than ~0.1) and lack oscillatory

455 zoning in CL images, suggesting they are metamorphic overgrowths and/or recrystallization

456 features. Among them, six analyses of metamorphic zircon rims show 206Pb/238U ages ranging from

457 266.1 to 297.5 Ma and yield a weighted mean 206Pb/238U age of 281 ± 12 Ma (Fig. 12a). In addition,

206
458 five analyses of zircon rims have Pb/238U ages ranging from 354.5 to 399.0 Ma and produce a

206
459 weighted mean Pb/238U age of 393 ± 7.8 Ma (Fig. 12a). These ages indicate the records of two–

206
460 phases of metamorphic events in the Chinese Altai orogen. Other grains give Pb/238U ages of

206
461 428.9–1945.4 Ma, and these are interpreted as xenocrysts. Two Pb/238U ages between 360 and

462 320 Ma might reflect mixture ages.

463 CL images of zircons from sample FY15–51 are similar to those from FY15–49 (Fig. 12). LA–

464 ICP–MS zircon U–Pb analysis data and age results are shown in Fig. 12b and listed in Table 6.

465 Forty–three spots were analyzed on thirty–seven zircon grains for sample FY15–51, with U

466 concentrations of 87–1210 ppm and Th/U ratios of 0.01–0.83. Zircon rims have relatively low Th/U

467 ratios (≤ 0.05) and lack oscillatory zoning in CL images, suggesting they are metamorphic

206
468 overgrowths and/or recrystallization features. Among them, twelve analyses show Pb/238U ages

206
469 ranging from 261.4 to 300.6 Ma and yield a weighted mean Pb/238U age of 282 ± 11 Ma (Fig.

206
470 12b). Additionally, nine analyses of zircon rims have Pb/238U ages ranging from 364.1 to 397.6

206
471 Ma and yield a weighted mean Pb/238U age of 387 ± 6.7 Ma (Fig. 12b). These ages also record

206
472 two–phases of metamorphic events in the southern Chinese Altai. Igneous zircons give Pb/238U

473 ages ranging from 441.6 Ma to 1250.8 Ma. Other obtained 206Pb/238U ages between 360 and 320 Ma

474 are interpreted as mixture ages.

475

476 6.3. Monazite morphology and U–Th–Pb geochronology

477

478 Most monazites in the sample FY15–51 are 50–250 µm in size，round, stubby or irregular in

479 shape, and weekly core–rim zoned (Figs. 13c–f). They generally occur as inclusions preserved in

480 garnet or cordierite (Figs. 13a and b). Eighteen available analyses were conducted on nine

206
481 monazites of sample FY15–51. Their Pb/238U ages show a range from 293.3 Ma to 271.9 Ma

208
482 (Table 7). The most concordant 18 data give a weighted mean Pb/232Th age of 287.9 ± 2.1 Ma

483 (Fig. 14a). This cluster includes monazite from different textural positions, and thus it does not

484 reveal statistically distinguishable ages. The weighted mean U–Pb concordia age is 289.45 ± 1.4 Ma
 208
485 (Fig. 14b), which is in agreement with the weighted mean Pb/232Th age within the analytical

208
486 errors. The lower intercept age (277 ± 30 Ma) is slightly younger than the Pb/232Th age. The

487 absence of a microstructural control on the monazite ages may suggest a complete resetting of the

488 U–Th–Pb isotope system that probably affected all the monazite crystals (Langone et al., 2010).

489 Accordingly, the age of 287.9 ± 2.1 Ma is regarded as an appropriately estimated metamorphic age

490 for this sample.

491

492 7. Discussion

493

494 7.1. P–T path and timing constraints

495

496 UHT metamorphism has been documented in two localities in the southern Chinese Altai (Figs.

497 1 and 2): Kalasu area (Tong et al., 2013, 2014), and Wuqiagou area (Li et al., 2010, 2014; Yang et

498 al., 2015b). Both of clockwise and anticlockwise P–T paths were defined for the Altai UHT

499 granulites (Fig. 15), as discussed above. In this study, P–T conditions of four different stages define

500 an anticlockwise P–T path with peak P–T conditions of ∼8.0 kbar/~980 °C, which is roughly

501 consistent with that of Tong et al. (2014). P–T path proposed by Tong et al. (2014) involved a short

502 post–peak P–T fragment from UHT conditions to ~870 °C at 8–9 kbar. Our UHT granulite samples,

503 however, recorded an integrated P–T path with a post–peak decompression and a subsequent near–

504 IBC process from ~900 °C (Fig. 15). It is important to note that the pressure and temperature

505 constraints obtained from classic geothermobarometers and average P–T approach would suggest

506 the presence of near–IBC path characterized by a minimal decompression (Fig. 15). However, this
507 P–T evolution is not consistent with the formation of coronas and thermodynamic calculations. It

508 follows that the P–T path constructed from phase equilibria modelling is more reliable (see above).

509 A combination of zircon and monazite chronological data generally allows for a higher

510 temporal resolution of high–grade metamorphism than when each method is applied separately (e.g.,

511 Wu et al., 2014). The LA–ICP–MS zircon U–Pb age data obtained from the Altai UHT granulites

206
512 indicate the existence of at least two metamorphic events. Weighted mean Pb/238U ages of 393 ±

513 7.8 and 387 ± 6.7 Ma yielded from metamorphic zircons are consistent with the timing proposed for

514 the regionally extensive Devonian metamorphism (390–365 Ma; Zhuang, 1994; Hu et al., 2002;

515 Windley et al., 2002; Wei et al., 2007; Zheng et al., 2007; Jiang et al., 2010; Broussolle et al., 2018).

516 Considering M1 metamorphic temperature conditions (700–750 °C) are roughly in agreement with

517 those (650–700 °C) of amphibolite and paragneiss in the southern Chinese Altai (Jiang et al., 2010),

518 M1 stage may be reasonably inferred to be linked with the Devonian tectono–metamorphic event.

519 Furthermore, weighted mean 206Pb/238U ages of ∼280 Ma produced from metamorphic zircons

520 are in agreement with the Permian tectono–metamorphic event (295–260 Ma; Chen et al., 2006;

521 Briggs et al., 2007; Zheng et al., 2007; Wang et al., 2009b; Li et al., 2014; Tong et al., 2014). Our

522 monazite U–Th–Pb dating results (287.9 ± 2.1 Ma) further support the existence of the Permian–age

523 metamorphic reworking event. We interpret these ages as the timing of the Altai UHT

524 metamorphism. Our monazite age dating results, however, did not document the Devonian

525 metamorphic event, which is probably because monazite recrystallized and reset during UHT

526 metamorphism (Wu et al., 2014; Morrissey et al., 2016).

527 Granitoids are extensively cropping out at Wuqiagou area and also documented two prominent

528 thermal–magmatic events of 393.5 ± 4.5 and 293.5 ± 6.0 Ma (Zhang et al., 2015a). Structurally, the
529 southern Chinese Altai documented two episodes of deformation D1 and D2, among which the

530 early sub–horizontal fabric S1 was modified to different degrees by the late near–vertical

531 deformation D2 in the Devonian (Jiang et al., 2015). Locally, the originally S1 and D2 fabrics were

532 reworked by the Permian upright close and asymmetric F3 folding (Broussolle et al., 2018). Our

533 zircon U–Pb ages also provide a solid support that the Altai UHT granulites experienced two

534 prominent tectono–metamorphic cycles: the first Devonian (~390 Ma) metamorphic and the second

535 Permian (~280 Ma) reworking process.

536

537 7.2. Tectonic Implications

538

539 A long–lived single arc north–dipping subduction and multiple arc subduction of the

540 Kazakhstan–Junggar plate have been proposed to explain the continental growth of Altai orogen in

541 Phanerozoic (e.g., Briggs et al., 2007; Windley et al., 2007). Two tectono–metamorphic cycles with

542 time interval of ~100 Ma are recognized and considered to have affected the whole edifice of the

543 Chinese Altai (Zhuang et al., 1994; Wei et al., 2007; Broussolle et al., 2018), as described above.

544 The first tectono–metamorphic cycle is extensive in the whole Chinese Altai Belt and took place in

545 the middle Devonian (Wei et al., 2007; Jiang et al., 2010; Broussolle et al., 2018). The Devonian

546 orogenic architecture was subsequently reworked by the Permian HT–UHT metamorphism on its

547 southern margin (Wang et al., 2009b, 2014; Li et al., 2014; Tong et al., 2014; Liu and Tong, 2015).

548 Particularly, previous zircon U–Pb and monazite U–Th–Pb dating results of Altai high

549 temperature gneisses and granulites showed metamorphic ages of 293–260 Ma (Chen et al., 2006;

550 Briggs et al., 2007; Zheng et al., 2007; Wang et al., 2009b), indicating that the Chinese Altai
551 orogenic belt experienced a significant HT–UHT reworking event in the late Paleozoic (Xiao et al.,

552 2008). In terms of the Permian HT–UHT metamorphism, researchers have proposed various genetic

553 models, such as Paleo–Asian oceanic crustal subduction (Li et al., 2004; Chen et al., 2006; He et al.,

554 2018), collision of the Junggar arc with the Chinese Altai terrane (Li et al., 2010; Broussolle et al.,

555 2018), slab break–off (Li et al., 2014; Yang et al., 2015b), ridge–subduction and the development of

556 slab–window (Windley and Xiao, 2018), and the thermal pulse of Tarim mantle plume activities

557 (Tong et al., 2013, 2014; Wang et al., 2014; Liu and Tong, 2015; Yang et al., 2015a). However,

558 until recently, the tectonic nature of the southern Chinese Altai is still a controversial topic.

559 The P–T path may be diagnostic of a particular tectonic environment. In particular, the

560 anticlockwise P–T path coupled with UHT peak conditions and a post–peak near–IBC process

561 generally reflects a tectonic evolutionary history involving initial crustal thickening in areas of

562 voluminous magmatic accretion (Fig. 15) (Harley, 1989; Sandiford and Powell, 1991). In addition,

563 the occurrence of a decompression segment is consistent with the extensional thinning of a

564 thickened crust (Harley, 1989). The fact that crustal thickening characterized by a subsequent near–

565 IBC is normally accompanied by underplating/accretion of deep–derived magma has been

566 discussed in many studies (e.g., Brandt et al., 2003; Clark et al., 2014; Jiao et al., 2015). In the

567 southern Chinese Altai, Permian granites have A– or I/A–type characteristics (Wang et al., 2010;

568 Shen et al., 2013), with a few high–temperature S–type granites (Zhou et al., 2007). These

569 granitoids and voluminous contemporary mafic intrusions show a bimodal magmatic association

570 (e.g., Wang et al., 2010; Shen et al., 2013), consistent with a post–orogenic extensional tectonic

571 setting (Shen et al., 2013, Wang et al., 2014). Accordingly, the models involving only a simple

572 subduction process or an arc–continent collision seem to be impossible.

573 The slab break–off was proposed by Li et al. (2014) and Yang et al. (2015b) to explain the

574 UHT metamorphism at Wuqiagou area, which we also considered unlikely. Freeburn et al. (2017)

575 proposed that in most cases slab break–off occurs too deep to trigger melting and cause thermal

576 perturbation within the overriding plate. On the other side, mantle wedge might be unable to

577 provide enough free space for voluminous hot asthenosphere to flow from beneath the slab and fill

578 the wedge above the slab (Niu et al., 2017). Collectively, the model of slab detachment might be

579 unable to explain the regionally extensive magmatic activities, crustal melting and the late orogenic

580 extension in the Chinese Altai orogenic belt.

581 A regional significant event that is broadly coeval with this HT–UHT metamorphism was the

582 formation of the Permian–aged Large Igneous Province (LIP) in northwestern China (Zhang et al.,

583 2012; Liu et al., 2014; Xu et al., 2014). The Permian Tarim LIP covering an area of more than 250

584 000 km2 (Tian et al., 2010; Yang et al., 2006) has recently become a focus of research (Xu et al.,

585 2014 and references therein). Some recent studies have suggested that the coeval Permian mafic

586 magmatic activities were associated with the rapid ascent of plume–related magma beneath the

587 lithosphere (e.g., Chen and Han, 2006; Zhang et al., 2012; Shen et al., 2013; Yang et al., 2015a).

588 Considering the time consistence with the Tarim mantle plume activity (~275 Ma; Zhang et al.,

589 2012), we proposed that the Altai HT–UHT metamorphism was most likely linked to the

590 underplating and heating of mantle–derived mafic magma as a result of the Tarim mantle plume

591 activity. Although we could not exclude the possibility of ridge subduction, the Tarim mantle plume

592 activity seems to be a better genetic model.

593 A possibly textural evidence for the existence of the Tarim mantle plume activity in the

594 southern Chinese Altai is the sinistral strike–slip motion of the Erqis fault belt (290–280 Ma;
595 Laurent–Charvet et al., 2003; Briggs et al., 2007; Zhang et al., 2012). Both of the Altai UHT

596 granulites and low–pressure metapelitic gneisses occurred along the large Erqis fault belt. The

597 movement of Erqis fault belt was ascribed to be at least partially influenced by the Tarim mantle

598 plume activity (Zhang et al., 2012). This probably resulted in the emplacement of voluminous A–

599 type granites and post–orogenic mafic–ultramafic intrusions during the same time. Accordingly, the

600 high heat flow necessary for the HT–UHT metamorphism of the southern Chinese Altai was

601 provided by coeval mafic intrusions which were probably generated by the Tarim mantle plume

602 activity.

603

604 8. Conclusions

605

606 Through detailed petrographic observations and P–T estimates for the UHT metapelitic

607 granulites from the southern Chinese Altai, in combination with both zircon and monazite U–Th–Pb

608 geochronology, some suggestions can be derived as follows:

609 (1) Petrography, mineral compositions and metamorphic P–T estimates for the Altai Permian

610 metapelitic granulites from Wuqiagou area suggest an anticlockwise P–T path characterized by

611 UHT peak conditions (~980 °C), and a post–peak decompression and a subsequent near–IBC

612 processes.

613 (2) The LA–ICP–MS zircon U–Pb dating results indicate the existence of at least two separated

614 high–grade metamorphic events. The M1 stage might occur in the middle Devonian (~390 Ma).

615 And then, the Devonian metamorphic terranes were locally overprinted by the Permian–age

616 UHT metamorphism (∼280 Ma). Monazite U–Th–Pb age dating results for our samples (287.9
617 ± 2.1 Ma) provide a further constraint on the timing of the UHT reworking event.

618 (3) The Altai Permian UHT reworking event was most likely associated with the underplating and

619 heating of deep–derived mafic magma as a result of the Tarim mantle plume activity.

620

621 Acknowledgements

622

623 This study has been supported by the Strategic Priority Research Program (B) of the Chinese

624 Academy of Sciences (XDB18030601), a One Hundred Talents Project of Shaanxi Province granted

625 to L. Tong, and by SIR RBSI14Y7PF grant by Italian Ministry of Education, University, Research

626 to O. Bartoli. We are really grateful to China Scholarship Council for its financial support during a

627 visit of Zhao Liu to Università di Padova, Italy. The Electron Microprobe analysis for mineral

628 composition was finished with help of Ms L.L. Chen at State Key Lab of Isotope Geochemistry,

629 Guangzhou Institute of Geochemistry. LA–ICP–MS zircon and monazite analysis were completed

630 with help of Mr Wei Gao at Wuhan Sample Solution Analytical Technology Co., Ltd., Wuhan,

631 China. We very much appreciate the journal editor Prof. T. Tsunogae and two anonymous reviewers

632 for their helpful and constructive comments on the early version of this paper.

633

634 References

635

636 Acosta–Vigil, A., Buick, I., Hermann, J., Cesare, B., Rubatto, D., London, D., Morgan VI, G.B.,
637 2010. Mechanisms of crustal anatexis: a geochemical study of partially melted metapelitic
638 enclaves and host dacite, SE Spain. Journal of Petrology 51, 785–821.

639 Bartoli, O., Cesare, B., Poli, S., Acosta–Vigil, A., Esposito, R., Turina, A., Bodnar, R.J., Angel, R.J.,
640 Hunter, J., 2013. Nanogranite inclusions in migmatitic garnet: behavior during piston cylinder

641 re–melting experiments. Geofluids 13, 405–420.

642 Bartoli, O., Cesare, B., Remusat, L., Acosta–Vigil, A., Poli, S., 2014. The H2O content of granite

643 embryos. Earth and Planetary Science Letters 395, 281–290.

644 Bartoli, O., Acosta–Vigil, A., Tajčmanová, L., Cesare, B., Bodnar, R.J., 2016. Using nanogranitoids

645 and phase equilibria modeling to unravel anatexis in the crustal footwall of the Ronda peridotites

646 (Betic Cordillera, S Spain). Lithos 256–257, 282–299.

647 Bartoli, O., 2017. Phase equilibria modelling of residual migmatites and granulites: An evaluation

648 of the melt–reintegration approach. Journal of Metamorphic Geology 35, 919–942.

649 Bindu, R.S., 1997. Granulite Facies Spinel–Cordierite Assemblages from the Kerala Khondalite

650 Belt, Southern India [J]. Gondwana Research 1, 121–128.

651 Brandt, S., Klemd, R., Okrusch, M., 2003. Ultrahigh–Temperature Metamorphism and Multistage

652 Evolution of Garnet–Orthopyroxene Granulites from the Proterozoic Epupa Complex, NW

653 Namibia [J]. Journal of Petrology 44, 1813–1856.

654 Briggs, S.M., Yin, A., Manning, C.E., Chen, Z.L., Wang, X.F., Grove, M., 2007. Late Paleozoic

655 tectonic evolution history of the Ertix Fault in the Chinese Altay and its implications for the

656 development of the Central Asian Orogenic System. Geological Society of America Bulletin 119,

657 944–960.

658 Broussolle, A., Aguilar, C., Sun, M., Schulmann, K., Štípská, P., Jiang, Y.D., Yu, Y., Xiao, W.J.,

659 Wang, S., Míková, J., 2018. Polycyclic Palaeozoic evolution of accretionary orogenic wedge in

660 the southern Chinese Altai: Evidence from structural relationships and U, Pb geochronology [J].

661 Lithos 314–315, 400–424.

662 Brown, M., 2007. Metamorphic conditions in orogenic belts: a record of secular change.

663 International Geology Review 49, 193–234.

664 Cenki, B., Kriegsman, L.M., Braun, I., 2002. Melt–producing and melt–consuming reactions in the

665 Achankovil cordierite gneisses, South India. Journal of Metamorphic Geology 20, 543–561.

666 Cesare, B., Acosta–Vigil, A., Bartoli, O., Ferrero, S., 2015. What can we learn from melt inclusions

667 in migmatites and granulites? Lithos 239, 186–216.

668 Chen, H.L., Yang, S.F., Li, Z.L., Yu, X., Xiao, W.J., Yuan, C., Lin, X.B., Li, J.L., 2006. Zircon

669 SHRIMP U–Pb chronology of the Fuyun basic granulite and its tectonic significance in the

670 Altaid orogenic belt. Acta Petrologica Sinica 22, 1351–1358 (in Chinese with English abstract).

671 Chen, L., Han, B., 2006. Geochronology, geochemistry and Sr–Nd–Pb isotopic composition of

672 mafic intrusive rocks in Wuqiagou area, north Xinjiang: constraints for mantle sources and

673 deep processes. Acta Petrologica Sinica 22, 1201–1214 (in Chinese with English abstract).

674 Clark, C., Kirkland, C.L., Spaggiari C.V., Oorschot, C., Wingate, M.T.D., Taylor, R.J., 2014.

675 Proterozoic granulite formation driven by mafic magmatism: An example from the Fraser Range

676 Metamorphics, Western Australia [J]. Precambrian Research 240, 1–21.

677 Coggon, R., Holland, T.J.B., 2002. Mixing properties of phengiticmicas and revised garnet–

678 phengite thermobarometers. Journal of Metamorphic Geology 20, 683–696.

679 Connolly, J.A.D., 2005. Computation of phase equilibria by linear programming: a tool for

680 geodynamic modeling and its application to subduction zone decarbonation. Earth and Planetary

681 Science Letters 236, 524–541.

682 Ferry, J.M., Watson, E.B., 2007. New thermodynamic models and revised calibrations for the Ti–

683 in–zircon and Zr–in–rutile thermometers [J]. Contributions to Mineralogy and Petrology 154,
684 429–437.

685 Freeburn, R., Bouilhol, P., Maunder, B., Magni, V., van Hunen, J., 2017. Numerical models of the

686 magmatic processes induced by slab breakoff. Earth and Planetary Science Letters 478, 203–213.

687 Galli, A., Le Bayon, B., Schmidt, M.W., Burg, J.P., Caddick, M.J., Reusser, E., 2011. Granulites and

688 charnockites of the Gruf Complex: evidence for Permian ultra–high temperature metamorphism

689 in the Central Alps. Lithos 124, 17–45.

690 Guevara, V.E., Caddick, M.J., 2016. Shooting at a moving target: phase equilibria modelling of

691 high–temperature metamorphism [J]. Journal of Metamorphic Geology 34, 209–235.

692 Guo, J.H., Peng, P., Chen, Y., Jiao, S.J., Windley, B.F., 2012. UHT sapphirine granulite

693 metamorphism at 1.93–1.92 Ga caused by gabbronorite intrusions: Implications for tectonic

694 evolution of the northern margin of the North China Craton. Precambrian Research 222–223,

695 124–142.

696 Harley, S.L., 1989. The origins of granulites: a metamorphic perspective. Geological Magazine 126,

697 215–247.

698 Harley, S.L., 2008. Refining the P–T records of UHT crustal metamorphism. Journal of

699 Metamorphic Geology 26, 125–154.

700 He, D., Dong, Y., Xu, X., Chen, J., Liu, X., Li, W., Li, X., 2018. Geochemistry, geochronology and

701 Hf isotope of granitoids in the Chinese Altai: Implications for Paleozoic tectonic evolution of the

702 Central Asian Orogenic Belt. Geoscience Frontiers 9, 1399–1415.

703 Holdaway, M.J., 2000. Application of new experimental and garnet Margules data to the garnet–

704 biotite geothermometer. American Mineralogist 85, 881–892.

705 Holland, T.J.B., Powell, R., 1998. An internally consistent thermodynamic data set for phases of
706 petrological interest. Journal of Metamorphic Geology 16, 309–343.

707 Holland, T.J.B., Powell, R., 2001. Calculation of Phase Relations Involving Haplogranitic Melts

708 Using an Internally Consistent Thermodynamic Dataset [J]. Journal of Petrology 42, 673–683.

709 Holness, M.B., Sawyer, E.W., 2008. On the pseudomorphing of melt–filled pores during the

710 crystallization of migmatites. Journal of Petrology 49, 1343–1363.

711 Holness, M.B., Cesare, B., Sawyer, E.W., 2011. Melted rocks under the microscope: microstructures

712 and their interpretation. Elements 7, 245–250.

713 Hu, A.Q., Zhang, G.X., Zhang, Q.F., Li, T.D., Zhang, J.B., 2002. A review on ages of precambrian

714 metamorphic rocks from Altai orogeny in Xinjiang, NW China. Chinese Journal of Geology 37,

715 129–142 (in Chinese with English abstract).

716 Jahn, B.M., Windley, B., Natal’in, B., Dobretsov, N., 2004. Phanerozoic continental growth in

717 central Asian. Journal of Asian Earth Sciences 23, 599–603.

718 Jiang, Y.D., Sun, M., Zhao, G.C., Yuan, C., Xiao, W.J., Xia, X.P., Long, X.P., Wu, F.Y., 2010. The

719 ~390 Ma high–T metamorphic event in the Chinese Altai: A consequence of ridge–subduction?

720 American Journal of Science 310, 1421–1452.

721 Jiang, Y.D., Štípská, P., Sun, M., Schulmann, K., Zhang, J., Wu, Q.H., Long, X.P., Yuan, C., Racek,

722 M., Zhao, G.C., Xiao, W.J., 2015. Juxtaposition of Barrovian and migmatite domains in the

723 Chinese Altai: a result of crustal thickening followed by doming of partially molten lower crust.

724 Journal of Metamorphic Geology 33, 45–70.

725 Jiao, S.J., Guo, J.H., 2011. Application of the two–feldspar geothermometer toultrahigh–

726 temperature (UHT) rocks in the Khondalite belt, North China cratonand its implications.

727 American Mineralogist 96, 250–260.

728 Jiao, S.J., Guo, J.H., Wang, L.J., Peng, P., 2015. Short–lived high–temperature prograde and

729 retrograde metamorphism in Shaerqin sapphirine–bearing metapelites from the Daqingshan

730 terrane, North China Craton [J]. Precambrian Research 269, 31–57.

731 Kelsey, D.E., 2008. On ultrahigh–temperature crustal metamorphism. Gondwana Research 13, 1–

732 29.

733 Kelsey, D.E., Hand, M., 2015. On ultrahigh temperature crustal metamorphism:phase equilibria,

734 trace element thermometry, bulk composition, heat sources, timescales and tectonic settings.

735 Geoscience Frontiers 6, 311–356.

736 Korhonen, F.J., Clark, C., Brown, M., Taylor, R.J.M., 2014. Taking the temperature of Earth's

737 hottest crust [J]. Earth and Planetary Science Letters 408, 341–354.

738 Langone, A., Godard, G., Prosser, G., Caggianelli, A., Rottura, A., Tiepolo, M., 2010. P–T–t path of

739 the Hercynian low–pressure rocks from the Mandatoriccio complex (Sila Massif, Calabria, Italy):

740 new insights for crustal evolution. Journal of Metamorphic Geology 28, 137–162.

741 Laurent–Charvet, S., Charvet, J., Monié, P., Shu, L., 2003. Late Paleozoic strike–slip shear zones in

742 eastern central Asia (NW China): New structural and geochronological data. Tectonics 22, 237–

743 241.

744 Li, H.J., He, G.Q., Wu, T.R., Wu, B., 2006. Confirmation of Altai–Mongolia microcontinent and its

745 implications. Acta Petrologica Sinica 22, 1369–1379 (in Chinese with English abstract).

746 Li, Z.L., Chen, H.L., Yang, S.F., Dong, C.W., Xiao, W.J., 2004. Petrology, geochemistry and

747 geodynamics of basic granulite from the Altay area, North Xinjiang, China [J]. Journal of

748 Zhejiang University Science 5, 979–584.

749 Li, Z.L., Li, Y.Q., Chen, H.L., Santosh, M., Xiao, W.J., Wang, H.H., 2010. SHRIMP U–Pb zircon
750 chronology of ultrahigh–temperature spinel–orthopyroxene–garnet granulite from South Altay

751 orogenic belt, northwestern China. Island Arc 19, 506–516.

752 Li, Z.L., Yang, X.Q., Li, Y.Q., Santosh, M., Chen, H.L., Xiao, W.J., 2014. Late Paleozoic tectono–

753 metamorphic evolution of the Altai segment of the Central Asian Orogenic Belt: Constraints

754 from metamorphic P–T pseudosection and zircon U–Pb dating of ultra–high–temperature

755 granulite. Lithos 204, 83–96.

756 Liu, H.Q., Xu, Y.G., Tian, W., Zhong, Y.T., Mundil, R., Li, X.H., Yang, Y.H., Luo, Z.Y., Shang–

757 Guan, S.M., 2014. Origin of two types of rhyolites in the Tarim Large Igneous Province:

758 Consequences of incubation and melting of a mantle plume. Lithos 204, 59–72.

759 Liu, Y.S., Gao, S., Hu, Z.C., Gao, C.G., Zong, K.Q., Wang, D.B., 2010. Continental and oceanic

760 crust recycling–induced melt–peridotite interactions in the Trans–North China Orogen: U–Pb

761 dating, Hf isotopes and trace elements in zircons from mantle xenoliths. Journal of Petrology 51,

762 537–571.

763 Liu, Z., Tong, L.X., 2015. The relationship between high–temperature metamorphism in Altay

764 orogen and Tarim mantle plume: some evidence from mafic and metapelitic granulites. Acta

765 Petrologica Sinica 31, 1761–1773 (in Chinese with English abstract).

766 Long, X.P., Yuan, C., Sun, M., Xiao, W.J., Zhao, G.C., Wang, Y.J., Cai, K.D., Xia, X.P., Xie, L.W.,

767 2010. Detrital zircon ages and Hf isotopes of the early Paleozoic flysch sequence in the Chinese

768 Altai, NW China: new constrains on depositional age, provenance and tectonic

769 evolution.Tectonophysics 480, 213–231.

770 Long, X.P., Yuan, C., Sun, M., Kröner, A., Zhao, G.C., Wilde, S., Hu, A.Q., 2011. Reworking of the

771 northern Tarim Craton by underplating of mantle plume–derived magmas: evidence from
772 Neoproterozoic adakitic rocks in the Kuluketage area, NW China. Precambrian Research 187, 1–

773 14.

774 Ludwig, K.R., 1999. Using Isoplot/EX, version 2. A Geological Toolkit for Microsoft Excel.

775 Berkeley Geochronological Center, Special Publication 1a, 47.

776 Morrissey, L.J., Hand, M., Kelsey, D.E., Wade, B.P., 2016. Cambrian High–temperature Reworking

777 of the Rayner–Eastern Ghats Terrane: Constraints from the Northern Prince Charles Mountains

778 Region, East Antarctica [J]. Journal of Petrology 36, 997–1004.

779 Nam, T.N., Sano, Y., Terada, K., Toriumi, M., Van Quynh, P., Dung, L.T., 2001. First SHRIMP U–

780 Pb zircon dating of granulites from the Kontum massif (Vietnam) and tectonothermal

781 implications. Journal of Asian Earth Sciences 19, 77–84.

782 Newton, R.C., Charlu, T.V., Kleppa, O.J., 1980. Thermo–chemistry of high structural state

783 plagioclases. Geochimica et Cosmochimica Acta 44, 933–941.

784 Niu, H.C., Yu, X.Y., Xu, J.F., Shan, Q., Chen, F.R., Zhang, H.X., Zheng, Z.P., 2006. Late Paleozoic

785 Volcanism and Associated Metallogenesis in the Altay area, Xinjiang, China. Geological

786 Publishing House, Beijing, pp. 83–136 (in Chinese).

787 Niu, Y.L., 2017. Slab breakoff: a causal mechanism or pure convenience? Science Bulletin 62, 456–

788 461.

789 Pattison, D.R.M., Chacko, T., Farquhar, J., McFarlane, C.R.M., 2003. Temperatures of granulite–

790 facies metamorphism: Constraints from experimental phase equilibria and thermobarometry

791 corrected for retrograde exchange. Journal of Petrology 44, 867–900.

792 Powell, R., Holland, T.J.B., 1994. Optimal geothermometry and geobarometry. American

793 Mineralogist 79, 120–133.

794 Pownall, J.M., Hall, R., Armstrong, R.A., Forster, M.A., 2014. Earth's youngest known ultrahigh–

795 temperature granulites discovered on Seram, eastern Indonesia [J]. Geology 42, 279–282.

796 Sandiford, M., Powell, R., 1991. Some remarks on high–temperature–low–pressure metamorphism

797 in convergent orogens. Journal of Metamorphic Geology 9, 333–340.

798 Santosh, M., Kusky, T., 2010. Origin of paired high pressure–ultrahigh–temperature orogens: a

799 ridge subduction and slab window model. Terra Nova 22, 35–42.

800 Santosh, M., Liu, S.J., Tsunogae, T., Li, J.H., 2012. Paleoproterozoic ultrahigh–temperature

801 granulites in the North China Craton: implications for tectonic models on extreme crustal

802 metamorphism. Precambrian Research 222–223, 77–106.

803 Sawyer, E.W., 2008. Atlas of migmatites. Mineralogical Association of Canada. The Canadian

804 Mineralogist Special Publication 9, Quebec City.

805 Shen, X., Zhang, H., Ma, L., 2013. LA–ICP–MS Zircon U–Pb Dating for Jieerkuduke Acidic Dykes

806 in the Southern Altay Range [J]. Xinjiang Geology 31, 157–161 (in Chinese with English

807 abstract).

808 Tajčmanová, L., Conolly, J.A.D., Cesare, B., 2009. A thermodynamicmodel for titanium and ferric

809 iron solution in biotite. Journal of Metamorphic Geology 27, 153–165.

810 Thompson, J.B., Hovis, G.L., 1979. Entropy of mixing in sanidine. American Mineralogist 64, 57–

811 65.

812 Tian, W., Campbell, I., Allen, C., Guan, P., Pan, W., Chen, M., Yu, H., Zhu, W., 2010. The Tarim

813 picrite–basalt–rhyolite suite, a Permian flood basalt from northwest China with contrasting

814 rhyolites produced by fractional crystallization and anatexis. Contributions to Mineralogy and

815 Petrology 160, 407–425.

816 Tomkins, H.S., Powell, R., Ellis, D.J., 2007. The pressure dependence of the zirconium–in–rutile

817 thermometer. Journal of Metamorphic Geology 25, 703–713.

818 Tong, L.X., Chen, Y.B., Xu, Y.G., Zhou, X., Liu Z., 2013. Zircon U–Pb ages of the ultrahigh–

819 temperature metapelitic granulite from the Altai orogen, NW China, and geological implications.

820 Acta Petrologica Sinica 29, 3435–3445 (in Chinese with English abstract).

821 Tong, L., Xu, Y.G., Cawood, P.A., Zhou, X., Chen, Y.B., Liu Z., 2014. Anticlockwise P–T

822 evolution at ~280 Ma recorded from Opx–Sil–bearing metapelitic granulite in the Chinese Altai

823 orogen, possible link with the Tarim mantle plume? Journal of Asian Earth Sciences 94, 1–11.

824 Tsunogae, T., Liu, S.J., Santosh, M., Shimizu, H., Li, J.H., 2011. Ultrahigh–temperature

825 metamorphism in Daqingshan, Inner Mongolia Suture Zone, North China. Gondwana Research

826 20, 36–47.

827 Vielzeuf, D., Montel, J M., 1994. Partial melting of metagreywackes. Part I. Fluid–absent

828 experiments and phase relationships [J]. Contributions to Mineralogy and Petrology 117: 375–

829 393.

830 Wang, T., Hong, D.W., Jahn, B.M., Tong, Y., Wang, Y.B., Han, B.F., Wang, X.X., 2006. Timing,

831 petrogenesis, and setting of Paleozoic synorogenic intrusions from the Altai Mountains,

832 Northwest China: Implications for the tectonic evolution of an accretionary orogen. The Journal

833 of Geology 114, 735–751.

834 Wang, T., Jahn, B.M., Kovach, V.P., Tong, Y., Hong, D.W., Han, B.F., 2009a. Nd–Sr isotopic

835 mapping of the Chinese Altai and implications for continental growth in the central Asian

836 orogenic Belt. Lithos 110, 359–372.

837 Wang, W., Wei, C.J., Wang, T., Lou, Y.X., Chu, H., 2009b. Confirmation of pelitic granulite in the
838 Altai orogen and its geological significance. Chinese Science Bulletin 54, 2543–2548.

839 Wang, W., Wei, C.J., Zhang, Y.H., Chu, H., Zhao, Y., Liu, X.C., 2014. Age and origin of sillimanite

840 schist from the Chinese Altai metamorphic belt: Implications for late Palaeozoic tectonic

841 evolution of the Central Asian Orogenic Belt. International Geology Review 56, 224–236.

842 Wei, C.J., Clarke, G., Tian, W., Qiu, L., 2007. Transition of metamorphic series from the kyanite– to

843 andalusite–types in the Altai orogen, Xinjiang, China: Evidence from petrography and calculated

844 KMnFMASH and KFMASH phase relations. Lithos 96, 353–374.

845 White, R.W., Powell, R., Holland, T.J.B., Worley, B.A., 2000. The effect of TiO2 and Fe2O3 on

846 metapelitic assemblages at greenschist and amphibolite facies conditions: mineral equilibria

847 calculations in the system K2O–FeO–MgO–Al2O3–SiO2–H2O–TiO2–Fe2O3. Journal of

848 Metamorphic Geology 18, 497–511.

849 White, R.W., Powell, R., 2002. Melt loss and the preservation of granulite facies mineral

850 assemblages. Journal of Metamorphic Geology 20, 621–632.

851 White, R.W., Powell, R., Halpin, J.A., 2004. Spatially–focussed melt formation in aluminous

852 metapelites from Broken Hill, Australia. Journal of Metamorphic Geology 22, 825–845.

853 White, R.W., Powell, R., Holland, T.J.B., 2007. Progress relating to calculation of partial melting

854 equilibria for metapelites. Journal of Metamorphic Geology 25, 511–527.

855 Windley, B.F., Kröner, A., Guo, J.H., Qu, G.S., Li, Y.Y., Zhang, C., 2002. Neoproterozoic to

856 Paleozoic geology of the Altai orogen, NW China: New zircon age data and tectonic evolution.

857 Journal of Geology 110, 719–737.

858 Windley, B.F., Alexeiev, D., Xiao, W., Kroner, A., Badarch, G., 2007. Tectonic models for accretion

859 of the Central Asian Orogenic belt. Journal of Geological Society 164, 31–47.
860 Windley, B.F., Xiao, W.J., 2018. Ridge subduction and slab windows in the Central Asian Orogenic

861 Belt: Tectonic implications for the evolution of an accretionary orogen. Gondwana Research 61,

862 73–87.

863 Wu, C.M., Zhang, J., Ren, L.D., 2004. Empirical garnet–biotite–plagioclase–quartz (GBPQ)

864 geobarometry in medium–to high–grade metapelites. Journal of Petrology 45, 1907–1921.

865 Wu, Y.B., Wang, H., Gao, S., Hu, Z.C., Liu, X.C., Gong, H.J., 2014. LA–ICP–MS monazite U–Pb

866 age and trace element constraints on the granulite–facies metamorphism in the Tongbai orogen,

867 central China. Journal of Asian Earth Sciences 82, 90–102.

868 Xiao, W.J., Han, C.M., Yuan, C., Sun, M., Lin, S.F., Chen, H.L., Li, Z.L., Li, J.L., Sun S., 2008.

869 Middle Cambrian to Permian subduction–related accretionary orogenesis of Northern Xinjiang,

870 NW China: Implications for the tectonic evolution of central Asia. Journal of Asian Earth

871 Sciences 32, 102–117.

872 Xu, Y.G., Wei, X., Luo, Z.Y., Liu, H.Q., Cao, J., 2014. The Early Permian Tarim Large Igneous

873 Province: Main characteristics and a plume incubation model. Lithos 204, 20–35.

40
874 Yang, S.F., Li, Z.L., Chen, H.L., Chen, W., Yu, X., 2006. Ar–39Ar dating of basalts from Tarim

875 Basin, NW China and its implication to a Permian thermal tectonic event. Journal of Zhejiang

876 University–Science A 7, 320–324.

877 Yang, T.N., Li, J.Y., Zhang, J., Hou, K.J., 2011. The Altai–Mongolia terrane in the Central Asian

878 Orogenic Belt (CAOB): A peri–Gondwana one? Evidence from zircon U–Pb, Hf isotopes and

879 REE abundance [J]. Precambrian Research 187, 79–98.

880 Yang, T.N., Li, J.Y., Liang, M. J., Wang, Y., 2015a. Early Permian mantle–crust interaction in the

881 south–central Altaids: High–temperature metamorphism, crustal partial melting, and mantle–
882 derived magmatism. Gondwana Research 28, 371–390.

883 Yang, X.Q., Li, Z.L., Wang, H.H., Chen, H.L., Li, Y.Q., Xiao, W.J., 2015b. Petrology and

884 geochemistry of ultrahigh–temperature granulites from the South Altay orogenic belt,

885 northwestern China: Implications for metamorphic evolution and protolith composition. Island

886 Arc 24, 169–187.

887 Yuan, C., Sun, M., Xiao, W., Li, X., Chen, H., Lin, S., Xia, X., Long, X., 2007. Accretionary

888 orogenesis of the Chinese Altai: Insights from Palaeozoic granitoids. Chemical Geology 242,

889 22–39.

890 Zhang, C.L., Santosh, M., Zou, H.B., Xu, Y.G., Zhou, G., Dong, Y.G., Ding, R.F., Wang, H.Y., 2012.

891 Revisiting the “Irtish tectonic belt”: Implications for the Palaeozoic tectonic evolution of the

892 Altai orogen. Journal of Asian Earth Sciences 52, 117–133.

893 Zhang, Y., Chen, J.L., Bai, J.K., Tang, Z., 2015a. LA–ICP–MS zircon dating of gneissic granitic

894 intrusive mass in Wuqiagou on the southern margin of Altay Orogenic Belt and its geological

895 significance. Northwestern Geology 48, 127–139 (in Chinese with English abstract).

896 Zhang, Z., Dong, X., Xiang, H., Ding, H., He, Z., Liou, J.G., 2015b. Reworking of the Gangdese

897 magmatic arc, southeastern Tibet: post–collisional metamorphism and anatexis. Journal of

898 Metamorphic Geology 33: 1–21.

899 Zhao, L., Guo, F., Fan, W.M., Li, C.W., Qin, X.F., Li, H.X., 2010. Origin of the granulite enclaves

900 in Indo–Sinian peraluminous granites, South China and its implication for crustal anatexis.

901 Lithos 150: 209–226.

902 Zheng, C.Q., Xu, X.C., Kato, T., Enami, M., 2007. Permian CHIME ages of monazites for the

903 kyanite–sillimanite type metamorphic belt in Chonghuer area, Altai, Xinjiang and their
904 geological implications. Geological Journal of China University 13, 566–573 (in Chinese with

905 English abstract).

906 Zhou, G., Zhang, Z.C., Luo, S.B., He, B., Wang, X., Ying, L.J., Zhao, H., Li, A.H., He, Y.K., 2007.

907 Confirmation of high–temperature strongly peraluminous Mayin’ebo granites in the margin of

908 Altay, Xinjiang: age, geochemistry and tectonic implications. Acta Petrologica Sinica 23, 1909–

909 1920 (in Chinese with English abstract).

910 Zhuang, Y.X., 1994. The PTSt evolution of metamorphism and development mechanism of the

911 thermal–structural–gneiss domes in the Chinese Altaides. Acta Geologica Sinica 68, 35–47 (in

912 Chinese with English abstract).

913 Zong, K.Q., Klemd, R., Yuan, Y., He, Z.Y., Guo, J.L., Shi, X.L., Liu, Y.S., Hu, Z.C., Zhang, Z.M.,

914 2017. The assembly of Rodinia: The correlation of early Neoproterozoic (ca. 900 Ma) high–

915 grade metamorphism and continental arc formation in the southern Beishan Orogen, southern

916 Central Asian Orogenic Belt (CAOB). Precambrian Research 290, 32–48.

917

918 FIGURE CAPTIONS

919

920 Fig. 1. A simplified metamorphic geological map of the Chinese Altai (modified after Wei et al.,

921 2007).

922

923 Fig. 2. A geological sketch map of the Fuyun area and sampling location of the Altai granulites

924 (modified after Li et al., 2004).

925 ★: Sample location: O2–3: Middle to Late Ordovician Habahe Group; Sk1: Silurian Kulumuti Group;
926 D1k: Early Devonian Kangbutibao Group; D2a: Middle Devonian Aletai Group ； C3k: Late

927 Carboniferous Kala–Erqis Group; J3: Late Jurassic Shishugou Group; Cz: Cenozoic Group; ψ4:

928 Variscan mafic and ultramafic rocks; γ4: Variscan granitic rocks.

929

930 Fig. 3. The field photograph of the UHT metapelitic granulites from the Wuqiagou area in the

931 southern Chinese Altai. The marker pen for scale is about 10 cm long.

932

933 Fig. 4. Photomicrographs and back–scattered electron (BSE) images showing the pre–peak and

934 peak mineral assemblages and textures in the UHT metapelitic granulites from Wuqiagou area in

935 southern Chinese Altai.

936 (a), spinel–cordierite–magnetite–anorthite associations in the matrix (FY15–49); (b), a garnet

937 porphyroblast with inclusions of spinel, cordierite, biotite, plagioclase and magnetite (FY15–49);

938 (c), a cordierite porphyroblast with spinel, cordierite, biotite, sillimanite and magnetite inclusions

939 (FY15–51); (d), a peak rutile grain preserved in retrograde quartz (FY15–49); (e), plagioclase and

940 quartz inclusions preserved in the orthopyroxene porphyroblast (FY15–49); (f), inclusion–phase

941 biotite, cordierite and magnetite contained in the orthopyroxene porphyroblast (FY15–51); (g), melt

942 pseudomorphs occurring among plagioclase. The outlines of quartz pools and films are indicative of

943 crystallization from melts (FY15–49); (h), MI clusters preserved in peritectic garnet (FY15–49); (i)

944 and (j), BSE images of selected nanogranites preserved in garnets.

945 Mineral abbreviations: grt, garnet; opx, orthopyroxene; sp, spinel; crd, cordierite; bi, biotite; sill,

946 sillimanite; pl, plagioclase; an, anorthite; qtz, quartz; mt, magnetite; ilm, ilmenite; ru, rutile; apt,

947 apatite; monz, monazite; zr, zircon; MI, melt inclusions.

948

949 Fig. 5. Photomicrographs and back–scattered electron (BSE) images illustrating reaction textures

950 resulting from the breakdown of garnet and orthopyroxene.

951 (a), opx2 rimed by opx–crd symplectitic rinds (FY15–51); (b), a BSE image illustrating the opx–crd

952 symplectite resorbing opx2 (FY15–51); (c), ‘spongy’ appearance around garnet consisted of

953 cordierite moats (crd3) and vermiform quartz (FY15–49); (d), the enlarged BSE image of black box

954 in (c) showing crd–qtz symplectite replacing garnet (FY15–49); (e) a photograph showing the

955 regrowth of the late garnet around orthopyroxene (FY15–51); (f), the intergrowths of bi–pl–qtz

956 destructing the early orthopyroxene (FY15–51); (g), the intergrowths of bi–pl–qtz bypassing the

957 garnet porphyroblast, suggesting their retrograde origin (FY15–49); (h), rutile needles preserved in

958 the garnet (FY15–49).

959 Mineral abbreviations see Figure 4.

960

961 Fig. 6. (SiO2)–A(Al2O3 + Fe2O3)–FM(FeO + MgO) projection from plagioclase, K–feldspar and

962 biotite for bulk rock compositions of the samples.

963

964 Fig. 7. (a) and (b), compositional variations of garnets from the Altai UHT granulites; (c)–(f),

965 compositional zoning profiles across garnet porphyroblasts.

966

967 Fig. 8. (a), P–T pseudosection for the UHT granulite sample FY15–49, calculated in

968 MnNCKFMASHTO system and considering the residual bulk rock composition. Orange fields

969 reflect the predicted mineral assemblages. (b–g), mineral and melt proportions (in vol.%). Green
970 arrow represents the probable post–peak retrograde evolution.

971

972 Fig. 9. P–T pseudosection for the UHT granulite sample FY15–49, calculated in

973 MnNCKFMASHTO system and considering the melt–reintegrated composition.

974

975 Fig. 10. T–XMg pseudosection constructed in MnNCKFMASHT system at 3 kbar and showing

976 different mineral assemblages due to heterogeneous effective bulk compositions.

977

978 Fig. 11. Cathodoluminescence (CL) images and 206Pb/238U ages of the zircon grains separated from

979 the Altai UHT granulite samples FY15–49 and FY15–51, respectively. The circles stand for

980 analytical spots and the neighboring white numbers are the respective 206Pb/238U ages.

981

982 Fig. 12. U–Pb concordant age diagrams showing the LA–ICP–MS zircon age results for FY15–49

983 and FY15–51, respectively.

984

985 Fig. 13. Representative photomicrographs and BSE images of analyzed monazite grains for sample

208
986 FY15–51 and locations of the analytical spots. The individual apparent Pb/232Th age is also

987 labeled.

988 (a) and (b), monazite grains within matrix and cordierite porphyroblasts; (c), an enlarged BSE

989 image of the monazite grain in (a); (d), a monazite inclusion preserved in garnet; (e) and (f),

990 enlarged BSE images of monazite grains in (b).

991
 208
992 Fig. 14. (a), the weighted mean Pb/232Th age for sample FY15–51; (b), concordia diagrams of

993 monazite LA–ICP–MS U–Pb analytical results for sample FY15–51.

994

995 Fig. 15. Suggested P–T path of the UHT granulites from the southern Chinese Altai (after Wei et

996 al., 2007). Biotite–dehydration reactions in the KFMASHTO system from White et al. (2002) are

997 marked in this figure. Also shown are P–T estimates and P–T paths suggested by Wang et al.

998 (2009b, 2014), Li et al. (2014), Tong et al. (2014) and Yang et al. (2015b) for the HT–UHT rocks

999 in this region.

1000

1001 Table captions:

1002

1003 Table 1. Bulk rock compositions of the UHT granulites at Wuqiagou area.

1004

1005 Table 2. Representative garnet and orthopyroxene compositions of the Altai UHT granulites.

1006

1007 Table 3. Representative spinel, cordierite, biotite and plagioclase compositions of the Altai UHT

1008 granulites.

1009

1010 Table 4. P–T estimates for four–stage mineral assemblages in the UHT metapelitic granulites from

1011 the southern Chinese Altai.

1012 Abbreviations: H00 (Holdaway, 2000); PH94 (Powell and Holland, 1994); P03 (Pattison et al.,

1013 2003); W04 (Wu et al., 2004).

1014

1015 Table 5. LA–ICP–MS U–Th–Pb analysis results for zircons from the Altai UHT granulite sample

1016 FY15–49.

1017

1018 Table 6. LA–ICP–MS U–Th–Pb analysis results for zircons from the Altai UHT granulite sample

1019 FY15–51.

1020

1021 Table 7. LA–ICP–MS monazite U–Th–Pb isotopic analyses for the Altai UHT granulite sample

1022 FY15–51.

1023
Table 1
Sample SiO2 Al2O3 Fe2O3T MgO CaO MnO Na2O K2O TiO2 P2O5 Cr2O3 LOI Total
FY15-49 51.42 19.31 15.24 8.52 0.65 0.39 0.64 1.71 1.48 0.04 0.06 1.12 100.58
FY15-51 56.49 18.72 11.56 6.87 0.94 0.20 0.96 1.90 1.24 0.08 0.04 1.32 100.35
Table 2

garnet orthopyroxene
core core mantle mantle rim rim grt4 core core core rim rim symplectite symplectite
SiO2 37.70 38.37 38.86 39.05 38.56 37.82 38.20 51.73 50.73 52.90 49.58 49.54 55.81 56.02
TiO2 0.02 0.00 0.04 0.06 0.03 0.05 0.00 0.07 0.01 0.12 0.00 0.07 0.12 0.03
Al2O3 20.32 21.40 21.68 21.71 21.29 20.70 20.82 8.71 8.14 8.07 5.68 5.85 2.25 2.00
Cr2O3 0.12 0.07 0.06 0.00 0.04 0.05 0.00 0.10 0.12 0.12 0.07 0.04 0.03 0.08
FeO 32.04 30.86 29.41 28.44 31.33 31.50 30.99 21.66 22.06 20.01 24.24 24.15 21.38 21.20
MnO 2.97 2.04 1.93 1.83 2.71 3.14 3.20 0.61 0.71 0.46 0.54 0.50 0.82 0.80
MgO 5.50 6.83 7.68 8.30 5.75 4.58 5.84 16.05 16.27 17.16 19.88 19.89 19.07 20.15
CaO 1.21 1.25 1.12 1.20 1.13 1.66 1.17 0.31 0.33 0.24 0.06 0.06 0.14 0.15
Na2O 0.07 0.01 0.02 0.05 0.00 0.01 0.06 0.80 0.67 0.65 0.02 0.00 0.05 0.06
K2O 0.00 0.01 0.01 0.00 0.00 0.01 0.01 0.01 0.00 0.01 0.00 0.02 0.34 0.00
Total 99.95 100.84 100.81 100.64 100.84 99.52 100.29 100.05 99.04 99.74 100.07 100.12 100.01 100.49
Cation(O) 12 12 12 12 12 12 12 6 6 6 6 6 6 6
Si 2.995 2.988 3.006 3.010 3.022 3.026 3.012 1.907 1.899 1.936 1.857 1.854 2.052 2.045
Ti 0.001 0.000 0.002 0.003 0.002 0.003 0.000 0.002 0.000 0.003 0.000 0.002 0.003 0.001
Al 1.903 1.964 1.977 1.973 1.967 1.952 1.935 0.379 0.359 0.348 0.251 0.258 0.098 0.086
Cr 0.008 0.004 0.004 0.000 0.002 0.003 0.000 0.003 0.004 0.003 0.002 0.001 0.001 0.002
Fe3+ 0.108 0.058 0.007 0.007 0.000 0.000 0.052 0.000 0.000 0.000 0.034 0.029 0.000 0.000
Fe2+ 2.020 1.951 1.896 1.827 2.053 2.107 1.991 0.668 0.690 0.612 0.725 0.727 0.657 0.647
Mn 0.200 0.135 0.126 0.119 0.180 0.213 0.214 0.019 0.023 0.014 0.017 0.016 0.026 0.025
Mg 0.651 0.793 0.885 0.954 0.671 0.546 0.686 0.882 0.908 0.936 1.110 1.109 1.045 1.096
Ca 0.103 0.104 0.093 0.099 0.095 0.142 0.099 0.012 0.013 0.009 0.002 0.002 0.006 0.006
Na 0.011 0.002 0.003 0.007 0.000 0.002 0.010 0.057 0.049 0.046 0.001 0.000 0.004 0.004
K 0.000 0.001 0.001 0.000 0.000 0.001 0.001 0.000 0.000 0.000 0.000 0.001 0.016 0.000
Sum 8.000 8.000 8.000 7.999 7.992 7.995 8.000 3.929 3.945 3.907 3.999 3.999 3.908 3.912
Alm 0.68 0.65 0.63 0.61 0.68 0.70 0.67 En 0.43 0.43 0.40 0.40 0.40 0.39 0.37
Py 0.22 0.27 0.30 0.32 0.22 0.18 0.23 Fs 0.57 0.57 0.60 0.60 0.60 0.61 0.63
Sp 0.07 0.05 0.04 0.04 0.06 0.07 0.07 AlⅣ 0.19 0.18 0.17 0.13 0.13 0.05 0.04
Grs 0.03 0.03 0.03 0.03 0.03 0.05 0.03
Mg# 24 29 32 34 25 21 26 57 57 60 60 60 61 63
Table 3

spinel cordierite biotite plagioclase

matrix matrix sp in grt matrix crd in sp crd in grt crd3 crd3 core rim bi in grt bi in crd matrix pl in grt pl in opx pl4
SiO2 0.45 0.04 0.02 49.56 48.54 50.12 49.65 49.16 37.52 37.40 37.82 35.65 59.80 57.01 59.43 60.75
TiO2 0.08 0.00 0.00 0.00 0.00 0.00 0.01 0.00 3.18 3.35 1.51 4.99 0.00 0.00 0.00 0.00
Al2O3 57.44 57.39 59.30 33.26 33.49 32.75 31.70 31.69 15.69 16.00 17.11 16.34 24.28 26.47 24.56 23.58
Cr2O3 0.26 0.39 0.42 0.73 0.50 0.14 0.06 0.02 0.05 0.08 0.16 0.12 0.00 0.02 0.02 0.02
FeO 34.77 33.80 28.71 5.38 5.58 3.80 4.61 4.70 14.21 14.54 14.45 15.87 0.06 0.25 0.17 0.17
MnO 0.24 0.23 0.11 0.10 0.15 0.06 0.14 0.05 0.05 0.09 0.04 0.12 0.01 0.02 0.02 0.00
MgO 5.80 5.60 7.96 9.75 10.12 11.27 10.33 10.48 14.65 14.54 14.81 11.69 0.00 0.00 0.00 0.01
CaO 0.00 0.00 0.00 0.02 0.05 0.02 0.00 0.02 0.00 0.02 0.06 0.00 7.05 9.15 7.51 6.14
Na2O 0.03 0.10 0.06 0.14 0.24 0.15 0.22 0.15 0.45 0.22 0.12 0.36 8.04 6.55 7.67 8.81
K2O 0.01 0.00 0.01 0.00 0.00 0.00 0.20 0.00 8.95 8.89 9.01 9.25 0.07 0.03 0.03 0.05
Total 99.08 97.55 96.59 98.94 98.67 98.31 96.92 96.27 94.75 95.13 95.09 94.39 99.31 99.50 99.41 99.53
Cation(O) 4 4 4 18 18 18 18 18 11 11 11 11 8 8 8 8
Si 0.013 0.001 0.001 5.010 4.924 5.050 5.098 5.080 2.797 2.779 2.793 2.707 2.687 2.572 2.670 2.721
Ti 0.002 0.000 0.000 0.000 0.000 0.000 0.001 0.000 0.178 0.187 0.084 0.285 0.000 0.000 0.000 0.000
Al 1.909 1.972 1.972 3.964 4.005 3.890 3.837 3.861 1.379 1.402 1.489 1.463 1.286 1.408 1.301 1.245
Cr 0.006 0.009 0.009 0.058 0.040 0.011 0.005 0.002 0.003 0.005 0.009 0.007 0.000 0.001 0.001 0.001
Fe3+ 0.059 0.021 0.021 0.000 0.095 0.029 0.030 0.007 0.000 0.000 0.073 0.000 0.002 0.009 0.006 0.006
Fe2+ 0.761 0.656 0.656 0.455 0.379 0.292 0.366 0.400 0.886 0.904 0.820 1.008 0.000 0.000 0.000 0.000
Mn 0.006 0.003 0.003 0.009 0.013 0.005 0.012 0.004 0.003 0.006 0.003 0.008 0.000 0.001 0.001 0.000
Mg 0.244 0.335 0.335 1.469 1.530 1.692 1.581 1.614 1.628 1.610 1.630 1.323 0.000 0.000 0.000 0.001
Ca 0.000 0.000 0.000 0.002 0.005 0.002 0.000 0.002 0.000 0.002 0.005 0.000 0.339 0.442 0.362 0.295
Na 0.002 0.003 0.003 0.027 0.028 0.029 0.044 0.030 0.065 0.032 0.017 0.053 0.701 0.573 0.668 0.765
K 0.000 0.000 0.000 0.000 0.000 0.000 0.026 0.000 0.851 0.843 0.849 0.896 0.004 0.002 0.002 0.003
Sum 3.002 3.000 3.000 10.994 11.019 11.000 11.000 11.000 7.792 7.768 7.772 7.750 5.019 5.008 5.011 5.037
Mg# 24.000 34.000 34.000 76.000 80.000 85.000 81.000 80 65.000 64.000 67.000 57.000
An 0.32 0.43 0.35 0.28
Ab 0.67 0.56 0.65 0.72
Table 4
M1 T = ~690 °C H00 P = ~4.10 kbar W04

M2 P03 Modle 1 Modle 2 Modle 3 Modle 4 Average

at P = 7 kbar T = 915 °C T = 975 °C T = 979 °C T = 914 °C T = 946 °C
T = 950 °C T = 999 °C T = 1002 °C T = 951 °C T = 976 °C
at P = 8 kbar T = 940 °C T = 1000 °C T = 1004 °C T = 939 °C T = 971 °C
T = 974 °C T = 1024 °C T = 1027 °C T = 975 °C T = 1000 °C
Average — — — — T = 973 °C
PH94 P = ~8.8 kbar, T = ~980 °C cor = 0.54, When a(H2O) = 0.3,
sigfit = 1.73 a(CO2) = 0.7

M3 PH94 P = ~7.0 kbar, T = ~790 °C cor = 0.41 sigfit = 1.23

M4 T = ~715 °C H00 P = ~6.85 kbar W04

Abbreviations: H00 (Holdaway, 2000); PH94 (Powell and Holland, 1994); P03 (Pattison et al., 2003);
W04 (Wu et al., 2004).
Table 5

206
Spot U Th Th/U Pb/238U 1σ
207
Pb*/206Pb* 1σ
207
Pb*/U235 1σ
206
Pb/238U(Ma) 1σ 207
Pb/206Pb(Ma) 1σ
207
Pb/235U(Ma) 1σ
FY15-49-01 757 7 0.01 0.042 0.00045 0.052 0.00134 0.304 0.00762 266.1 2.80 301.9 62.03 269.7 5.93
FY15-49-02 1136 7 0.01 0.043 0.00038 0.055 0.00120 0.330 0.00717 274.5 2.37 405.6 43.52 289.5 5.48
FY15-49-03 697 17 0.02 0.045 0.00049 0.058 0.00243 0.361 0.01713 281.0 3.02 542.6 90.73 313.2 12.78
FY15-49-04 1096 5 0.00 0.046 0.00049 0.053 0.00117 0.336 0.00750 287.9 3.02 342.7 50.00 294.2 5.70
FY15-49-05 925 23 0.03 0.047 0.00071 0.053 0.00153 0.350 0.01214 296.9 4.35 350.1 64.81 305.1 9.13
FY15-49-06 785 7 0.01 0.047 0.00058 0.055 0.00126 0.354 0.00832 297.5 3.58 394.5 56.48 307.8 6.24
FY15-49-07 680 25 0.04 0.057 0.00116 0.053 0.00114 0.418 0.01210 354.5 7.07 350.1 50.00 354.7 8.66
FY15-49-08 471 20 0.04 0.055 0.00102 0.055 0.00135 0.417 0.01347 345.7 6.26 390.8 55.55 354.2 9.65
FY15-49-09 507 21 0.04 0.063 0.00136 0.057 0.00251 0.494 0.02435 390.9 8.27 487.1 102.77 407.7 16.55
FY15-49-10 437 26 0.06 0.063 0.00104 0.050 0.00134 0.436 0.01394 393.1 6.33 183.4 67.58 367.3 9.86
FY15-49-11 739 6 0.01 0.063 0.00106 0.057 0.00264 0.497 0.02190 393.7 6.45 494.5 101.84 409.7 14.85
FY15-49-12 536 138 0.26 0.063 0.00134 0.059 0.00178 0.514 0.01640 396.2 8.09 576.0 66.66 421.3 11.00
FY15-49-13 944 187 0.20 0.064 0.00114 0.052 0.00120 0.466 0.01319 399.0 6.92 301.9 51.85 388.2 9.14
FY15-49-14 2877 366 0.13 0.158 0.00128 0.066 0.00189 1.461 0.03923 943.5 7.10 820.4 59.26 914.5 16.19
FY15-49-15 932 190 0.20 0.082 0.00070 0.057 0.00153 0.650 0.01694 509.7 4.16 487.1 59.25 508.3 10.43
FY15-49-16 558 239 0.43 0.084 0.00120 0.057 0.00138 0.670 0.01792 519.7 7.14 505.6 56.48 520.8 10.90
FY15-49-17 220 83 0.37 0.089 0.00169 0.059 0.00168 0.726 0.02320 549.0 10.02 572.3 61.87 553.9 13.65
FY15-49-18 565 443 0.78 0.088 0.00119 0.060 0.00124 0.731 0.01709 544.7 7.05 598.2 46.29 557.3 10.03
FY15-49-19 520 98 0.19 0.078 0.00101 0.058 0.00136 0.631 0.01677 481.8 6.06 542.6 47.22 496.5 10.44
FY15-49-20 646 139 0.22 0.080 0.00130 0.058 0.00120 0.634 0.01464 496.2 7.77 522.3 50.92 498.6 9.10
FY15-49-21 317 56 0.18 0.070 0.00085 0.055 0.00166 0.533 0.01649 433.9 5.15 433.4 66.66 434.0 10.92
FY15-49-22 431 27 0.06 0.071 0.00138 0.055 0.00147 0.540 0.01616 442.5 8.29 433.4 59.25 438.6 10.65
FY15-49-23 203 151 0.74 0.078 0.00111 0.060 0.00199 0.646 0.02214 487.0 6.66 590.8 72.21 506.3 13.66
FY15-49-24 143 35 0.25 0.230 0.00323 0.082 0.00192 2.616 0.06484 1333.7 16.91 1253.7 45.53 1305.0 18.21
FY15-49-25 401 146 0.36 0.076 0.00126 0.056 0.00145 0.589 0.01851 471.5 7.56 450.0 57.40 470.3 11.83
FY15-49-26 245 55 0.23 0.071 0.00165 0.057 0.00212 0.559 0.02453 442.2 9.91 494.5 83.33 450.9 15.98
FY15-49-27 299 96 0.32 0.082 0.00120 0.057 0.00172 0.652 0.02218 509.0 7.15 501.9 66.66 509.8 13.64
FY15-49-28 515 244 0.47 0.085 0.00109 0.055 0.00113 0.650 0.01529 525.8 6.50 420.4 46.29 508.8 9.41
FY15-49-29 761 260 0.34 0.069 0.00095 0.055 0.00107 0.521 0.01199 428.9 5.71 394.5 44.44 425.7 8.00
FY15-49-30 133 53 0.40 0.352 0.00565 0.124 0.00250 6.060 0.16365 1945.4 26.91 2009.3 35.80 1984.6 23.54
FY15-49-31 701 544 0.78 0.088 0.00118 0.059 0.00106 0.718 0.01494 545.1 7.00 564.9 43.51 549.7 8.83
FY15-49-32 501 273 0.54 0.080 0.00113 0.059 0.00146 0.655 0.01836 494.4 6.75 583.4 53.69 511.8 11.26
FY15-49-33 153 134 0.87 0.138 0.00170 0.067 0.00150 1.274 0.03061 832.2 9.65 838.9 46.30 834.0 13.67
Table 6

206
Spot U Th Th/U Pb/238U 1σ 207
Pb*/206Pb* 1σ 207
Pb*/U235 1σ 206
Pb/238U(Ma) 1σ 207
Pb/206Pb(Ma) 1σ 207
Pb/235U(Ma) 1σ
FY15-51-01 621 5 0.01 0.041 0.00051 0.054 0.00187 0.312 0.01023 261.4 3.15 372.3 77.77 275.8 7.92
FY15-51-02 1162 7 0.01 0.043 0.00071 0.055 0.00155 0.328 0.01021 272.2 4.38 466.7 60.18 287.9 7.80
FY15-51-03 532 6 0.01 0.044 0.00067 0.053 0.00174 0.326 0.01049 279.8 4.12 316.7 75.92 286.5 8.04
FY15-51-04 582 7 0.01 0.046 0.00076 0.055 0.00167 0.353 0.01061 289.9 4.70 416.7 68.51 307.0 7.96
FY15-51-05 1020 8 0.01 0.046 0.00078 0.053 0.00132 0.341 0.00939 290.0 4.80 344.5 55.55 298.0 7.11
FY15-51-06 806 7 0.01 0.046 0.00050 0.054 0.00223 0.345 0.01384 290.0 3.10 388.9 97.21 300.9 10.45
FY15-51-07 1011 15 0.01 0.046 0.00144 0.056 0.00558 0.337 0.01571 292.5 8.86 435.2 224.81 294.9 11.93
FY15-51-08 448 5 0.01 0.046 0.00053 0.059 0.00251 0.381 0.01512 292.5 3.24 588.9 123.13 327.5 11.12
FY15-51-09 796 5 0.01 0.047 0.00047 0.053 0.00206 0.341 0.01324 294.6 2.89 322.3 88.88 297.9 10.03
FY15-51-10 1121 9 0.01 0.047 0.00042 0.056 0.00194 0.366 0.01218 295.0 2.58 464.9 75.92 316.4 9.06
FY15-51-11 1210 12 0.01 0.048 0.00044 0.053 0.00177 0.347 0.01119 300.6 2.68 322.3 75.92 302.8 8.43
FY15-51-12 1077 9 0.01 0.048 0.00045 0.054 0.00190 0.356 0.01231 300.6 2.79 364.9 79.62 309.2 9.22
FY15-51-13 857 5 0.052 0.00075 0.053 0.00138 0.383 0.01066 327.9 4.60 324.1 59.25 329.0 7.83
FY15-51-14 939 8 0.055 0.00120 0.056 0.00143 0.428 0.01397 343.2 7.32 477.8 57.40 362.1 9.93
FY15-51-15 715 5 0.057 0.00075 0.060 0.00224 0.470 0.01432 359.5 4.60 609.3 81.47 391.5 9.89
FY15-51-16 700 5 0.01 0.048 0.00066 0.048 0.00145 0.320 0.01039 302.9 4.06 101.9 67.59 282.0 7.99
FY15-51-17 811 5 0.01 0.058 0.00129 0.062 0.00174 0.502 0.01573 364.1 7.86 733.3 60.33 412.8 10.64
FY15-51-18 510 11 0.02 0.058 0.00099 0.057 0.00166 0.455 0.01399 364.2 6.04 476.0 64.81 381.1 9.76
FY15-51-19 1024 8 0.01 0.059 0.00081 0.053 0.00157 0.437 0.01398 368.2 4.91 331.5 66.66 367.8 9.88
FY15-51-20 676 4 0.01 0.059 0.00135 0.055 0.00139 0.447 0.01432 368.8 8.21 466.7 57.40 375.2 10.05
FY15-51-21 780 24 0.03 0.060 0.00076 0.061 0.00210 0.509 0.01845 373.4 4.60 638.9 78.69 418.0 12.41
FY15-51-22 701 37 0.05 0.061 0.00144 0.061 0.00163 0.513 0.01571 381.9 8.74 655.6 57.40 420.2 10.55
FY15-51-23 524 12 0.02 0.061 0.00085 0.053 0.00157 0.449 0.01400 383.3 5.18 331.5 66.66 376.9 9.81
FY15-51-24 930 8 0.01 0.063 0.00105 0.055 0.00194 0.480 0.01750 391.4 6.36 427.8 77.77 398.1 12.01
FY15-51-25 885 85 0.10 0.064 0.00170 0.055 0.00173 0.500 0.02118 397.6 10.29 431.5 70.36 411.7 14.34
FY15-51-26 776 89 0.11 0.073 0.00151 0.059 0.00192 0.596 0.02065 454.7 9.10 564.9 67.58 474.4 13.14
FY15-51-27 567 82 0.15 0.083 0.00181 0.060 0.00213 0.702 0.03071 514.7 10.77 594.5 105.54 539.9 18.32
FY15-51-28 1908 268 0.14 0.087 0.00122 0.055 0.00145 0.660 0.01871 535.7 7.24 466.7 59.25 514.5 11.45
FY15-51-29 332 26 0.08 0.072 0.00095 0.059 0.00262 0.584 0.02627 446.1 5.71 564.9 96.28 466.7 16.85
FY15-51-30 718 137 0.19 0.130 0.00205 0.066 0.00178 1.184 0.03500 785.6 11.68 1200.0 55.55 793.2 16.28
FY15-51-31 457 225 0.49 0.076 0.00120 0.054 0.00158 0.566 0.01834 469.6 7.21 376.0 66.66 455.2 11.90
FY15-51-32 218 150 0.69 0.081 0.00134 0.058 0.00193 0.649 0.02264 500.4 7.98 538.9 72.21 507.7 13.94
FY15-51-33 489 223 0.46 0.077 0.00123 0.055 0.00165 0.587 0.01943 480.9 7.37 398.2 66.66 468.9 12.43
FY15-51-34 441 59 0.13 0.071 0.00148 0.055 0.00172 0.535 0.01930 441.6 8.94 398.2 74.99 435.2 12.77
FY15-51-35 581 154 0.27 0.074 0.00099 0.055 0.00149 0.559 0.01551 461.7 5.95 390.8 62.96 451.1 10.10
FY15-51-36 225 118 0.52 0.119 0.00255 0.063 0.00167 1.038 0.03485 727.1 14.66 694.5 57.40 723.1 17.36
FY15-51-37 429 385 0.90 0.085 0.00147 0.057 0.00146 0.667 0.01936 525.1 8.76 479.7 89.80 518.9 11.79
FY15-51-38 314 133 0.42 0.079 0.00125 0.057 0.00173 0.617 0.01900 489.0 7.46 483.4 66.66 487.8 11.93
FY15-51-39 397 144 0.36 0.078 0.00154 0.056 0.00175 0.610 0.02172 485.5 9.18 435.2 70.36 483.3 13.70
FY15-51-40 819 676 0.83 0.077 0.00112 0.058 0.00149 0.620 0.01487 478.2 6.69 516.7 57.40 490.0 9.32
FY15-51-41 837 236 0.28 0.152 0.00218 0.068 0.00144 1.428 0.03401 910.3 12.20 857.4 44.45 900.6 14.23
FY15-51-42 815 414 0.51 0.146 0.00292 0.072 0.00170 1.452 0.03822 880.8 16.41 988.9 48.15 910.8 15.83
FY15-51-43 87 23 0.26 0.214 0.00590 0.092 0.00299 2.765 0.12166 1250.8 31.30 1533.3 61.27 1346.2 32.82
Table 7

208
Spot U Th 206
Pb/238U 1σ
207
Pb*/206Pb* 1σ
207
Pb*/U235 1σ Pb/232Th 1σ
206
Pb/238U(Ma) 1σ 207
Pb/206Pb(Ma) 1σ
207
Pb/235U(Ma) 1σ
208
Pb/232Th 1σ
FY15-51-01 8284 40535 0.044 0.00023 0.052 0.00104 0.314 0.00618 0.014 0.00007 276.0 1.46 287.1 46.29 277.4 4.78 271.9 1.43
FY15-51-02 5808 40878 0.045 0.00027 0.053 0.00116 0.325 0.00700 0.014 0.00009 281.5 1.66 316.7 17.59 285.6 5.37 283.7 1.73
FY15-51-03 9121 34978 0.045 0.00027 0.051 0.00107 0.319 0.00649 0.014 0.00008 283.8 1.65 253.8 48.14 280.8 5.00 278.6 1.62
FY15-51-04 9263 35772 0.045 0.00029 0.051 0.00103 0.320 0.00645 0.014 0.00009 285.0 1.81 250.1 41.66 281.7 4.97 283.9 1.75
FY15-51-05 5812 43082 0.046 0.00028 0.053 0.00124 0.332 0.00744 0.014 0.00008 287.1 1.75 324.1 21.30 291.1 5.68 287.0 1.65
FY15-51-06 6493 44967 0.046 0.00032 0.052 0.00123 0.329 0.00764 0.014 0.00008 287.3 2.00 301.9 49.07 289.2 5.84 285.2 1.65
FY15-51-07 6277 44195 0.046 0.00033 0.052 0.00158 0.328 0.01013 0.014 0.00009 288.7 2.05 279.7 68.51 288.3 7.75 290.6 1.79
FY15-51-08 5310 42551 0.046 0.00029 0.053 0.00121 0.336 0.00736 0.014 0.00008 289.2 1.78 344.5 51.85 294.1 5.60 283.0 1.51
FY15-51-09 4936 53027 0.046 0.00037 0.055 0.00122 0.348 0.00754 0.014 0.00009 289.9 2.31 413.0 50.00 303.3 5.68 285.3 1.72
FY15-51-10 6793 40938 0.046 0.00031 0.054 0.00113 0.342 0.00689 0.014 0.00009 290.1 1.92 364.9 48.14 298.5 5.22 290.9 1.73
FY15-51-11 5906 44361 0.046 0.00036 0.052 0.00116 0.329 0.00743 0.015 0.00009 290.2 2.23 279.7 51.85 289.2 5.68 293.3 1.89
FY15-51-12 5521 37842 0.046 0.00033 0.054 0.00121 0.341 0.00768 0.015 0.00009 291.1 2.05 353.8 51.85 298.3 5.82 291.2 1.70
FY15-51-13 6230 43786 0.046 0.00031 0.053 0.00101 0.337 0.00633 0.014 0.00009 291.1 1.94 324.1 44.44 294.9 4.81 288.5 1.75
FY15-51-14 5464 50715 0.046 0.00034 0.050 0.00127 0.321 0.00810 0.014 0.00008 291.9 2.10 211.2 59.25 282.4 6.23 287.7 1.51
FY15-51-15 8602 42066 0.046 0.00030 0.052 0.00103 0.333 0.00656 0.015 0.00007 292.5 1.84 279.7 44.44 291.6 5.00 292.4 1.43
FY15-51-16 7215 45416 0.047 0.00029 0.054 0.00115 0.346 0.00774 0.014 0.00008 293.6 1.78 353.8 80.55 301.4 5.85 289.3 1.59
FY15-51-17 5940 41740 0.047 0.00029 0.054 0.00137 0.346 0.00888 0.015 0.00009 294.3 1.78 366.7 57.40 302.0 6.70 292.2 1.85
FY15-51-18 5626 43620 0.047 0.00037 0.053 0.00114 0.342 0.00710 0.015 0.00010 295.9 2.30 324.1 48.14 299.0 5.37 291.9 1.99
Highlights

Altai UHT granulites documented an anticlockwise P–T path involving a post–peak

ITD to near-IBC process.

U–Th-Pb chronological results for metamorphic zircons and monazites show two
weighted mean ages of ~390 Ma and ~280 Ma.

The UHT metamorphic event was likely associated with Permian reworking and the
Tarim mantle plume activity.