Petrology, phase equilibria modelling, and in situ zircon and monazite geochronology of ultrahigh-temperature granulites from the khondalite belt of southern India

LITHOS 348-349 (2019) 105195

Contents lists available at ScienceDirect

LITHOS journal homepage: www.elsevier.com/locate/lithos

Petrology, phase equilibria modelling, and in situ zircon and monazite geochronology of ultrahigh-temperature granulites from the khondalite belt of southern India Bing Yu a, M. Santosh a, b, c, *, Shan-Shan Li a, d, E. Shaji e a

School of Earth Sciences and Resources, China University of Geosciences Beijing, 29 Xueyuan Road, Beijing 100083, China Deparment of Earth Science, University of Adelaide, Adelaide, SA 5005, Australia Yonsei Frontier Lab, Yonsei University, Seoul, Republic of Korea d Department of Geology and Geological Engineering, Colorado School of Mines, Golden, CO 80401, USA e Department of Geology, University of Kerala, Kariyavattom Campus, Trivandrum 695 581, India b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 23 June 2019 Received in revised form 22 August 2019 Accepted 30 August 2019 Available online 3 September 2019

The granulite facies metapelite (khondalite) belt in the Trivandrum Block of southern India has been central to investigations on extreme crustal metamorphism associated with the final assembly of the Gondwana supercontinent during late Neoproterozoic-Cambrian. Here we investigate garnet-sillimanite-cordieritespinel-bearing metapelites from this khondalite belt using state-of-the-art petrologic, mineral phase equilibria modelling, and coupled zircon and monazite U-Pb geochronology to characterize the nature and timing of metamorphism and their tectonic implications. From textural studies and mineral phase equilibrium modelling, we infer that Sill þ Grt þ Crd representing prograde metamorphism was stable at 6 e9 kbar and 760e790  C. Equilibrium spinel-quartz assemblage suggests peak metamorphic conditions of 6.5e7 kbar and 1010e1030  C consistent with ultra-high temperature metamorphism in the Trivandrum Block. Growth of cordierite and biotite at the expense of garnet correlates with retrograde metamorphism at 4.5e6.5 kbar and 770e950  C. Our results allow an interpretation of the prograde, peak, and retrograde P eT conditions of the khondalites, where from the peak ultrahigh-temperature stage of >1000  C, the rocks underwent isothermal decompression as indicated by garnet breakdown to cordierite, followed by isobaric cooling, corresponding to an overall clockwise PeT evolution. We present results from U-Pb geochronology of zircon as well as LA-ICPMS data on monazite grain separates. The zircon grains show two distinct age peaks, with Paleoproterozoic cores surrounded by Late Neoproterozoic-Cambrian rims or recrystallized domains. The younger ages around 550e560 Ma are similar to the lower intercept age, whereas the wide range of Paleoproterozoic ages fall along a discordia. The LA-ICP-MS monazite U-Pb data also show two distinct age populations at Paleoproterozoic and latest Neoproterozoic-Cambrian, although the Paleoproterozoic population is scarce. We correlate the late Neoproterozoic-Cambrian ages to the ultrahightemperature metamorphic event, associated with the assembly of the Gondwana supercontinent. The wide range of monazite ages indicate that high temperature metamorphism was long-lived. © 2019 Elsevier B.V. All rights reserved.

Keywords: Petrology and mineral phase equilibria Zircon and monazite geochronology Khondalite Trivandrum block Gondwana supercontinent

1. Introduction The breakup and assembly of supercontinents involve some of the fundamental processes of the Earth's dynamic system including rifting, subduction, crust-mantle interaction, crustal growth and recycling, and orogenesis (Nance et al., 2014). The Late Neoproterozoic e Cambrian assembly of Gondwana supercontinent which involved the amalgamation of several large crustal blocks resulted in the construction of the major “Pan-African” orogenic * Corresponding author at: School of Earth Sciences and Resources, China University of Geosciences Beijing, 29 Xueyuan Road, Beijing 100083, China. E-mail address: [email protected] (M. Santosh). https://doi.org/10.1016/j.lithos.2019.105195 0024-4937/© 2019 Elsevier B.V. All rights reserved.

€ner, 1984; Santosh et al., 2009). The belts (e.g., Collins et al., 2007; Kro southern Indian peninsula lies at the junction of the Gondwanaforming orogenic belts, between the East African Orogen and the Kuunga Orogen (Meert and Lieberman Meert and Lieberman, 2008). The crustal blocks in the Southern Granulite Terrane (SGT) of Peninsular India preserve evidence for a prolonged subductionaccretion history during Neoproterozoic leading to Himalayanstyle collision in latest Neoproterozoic-Cambrian associated with the final amalgamation of the Gondwana supercontinent (Collins et al., 2007; Santosh et al., 2009, 2017) During this period, the rocks of central and eastern Madagascar, the Trivandrum, Nagercoil and Madurai Blocks in India and the Highland Complex of Sri Lanka underwent a protracted history of high to ultrahigh temperature

2

B. Yu et al. / LITHOS 348-349 (2019) 105195

metamorphism (Clark et al., 2015; Collins et al., 2007; Sajeev et al., 2010). Granulite facies rocks that were subjected UHT conditions occur widely in the different crustal blocks and tectonic domains of the SGT (Raith et al., 1997; Santosh et al., 2003, 2006, 2009; Collins et al., 2007; Tsunogae and Santosh, 2010; Clark et al., 2015; Johnson et al., 2015; Harley and Nandakumar, 2016). The granulite facies supracrustal belt in the Trivandrum Block at the southern part of the Peninsula, previously termed as the Kerala Khondalite Belt (Chacko et al., 1987) exposes voluminous highgrade metasedimentary gneisses that make up an important component of the record of collision and amalgamation of Gondwana. The aluminous metapelites in this belt intercalated with garnet-and biotite-bearing leucogneisses and charnockites (Fig. 1) with minor mafic granulites and calc-silicate rocks. Previous studies reported high to ultra-high temperature metamorphic conditions from the khondalite belt, although the protolith history, sedimentation and the timing and nature of metamorphism remain debated (e.g., Collins et al., 2007; Harley and Nandakumar, 2016; Johnson et al., 2015; Liu et al., 2016; Santosh et al., 2003). In this work, we present results from an integrated study on the petrology, including mineral phase equilibria modelling, zircon U-Pb

geochronology, and dating of monazite both in thin sections and mineral separates with a view to constrain the timing of metamorphism, pressure-temperature conditions of prograde, peak and retrograde metamorphism and the P-T path. Our results provide further insights into long-lived high to ultrahigh-temperature metamorphism in the collisional orogen associated with the assembly of Gondwana. 2. Geological background and sampling The khondalite belt of Trivandrum Block represents the deeply eroded mid- to lower-crustal section of a Precambrian orogen. The dominant rock type is represented by garnet-sillimanite- and cordierite-bearing metapelitic gneisses with or without graphite which occurs as bands and layers within leucocratic felsic gneisses carrying garnet and biotite. In an earlier study, Morimoto et al. (2004) several mineral reaction textures from cordierite-, sillimanite- and spinel-bearing metapelites in the Trivandrum Block. They also reported spinel þ quartz in equilibrium assemblage from these rocks, suggesting ultrahigh temperature metamorphism in the khondalite belt. The clockwise metamorphic P-T path obtained in their study

Fig. 1. Geological map of the Trivandrum Block (modified after Liu et al., 2016). The sample locations of present study are also shown. Insert shows the major cratonic blocks in Peninsular India. AC- Aravalli craton; BKC- Bundelkhand craton; SC-Singhbhum craton; BC-Bastar craton; DC-Dharwar craton. The area covered in the main figure is marked by box.

B. Yu et al. / LITHOS 348-349 (2019) 105195

3

Fig. 2. Representative field photographs of the rocks analyzed in this study. (a) sample KLD from Koliakode quarry. (b) Migmatitic khondalite with porphyroblasts of garnet, cordierite, and sillimanite (KLD 1-1), (c) sample KLP from Kulappara, (d) khondalite showing aluminous layers rich in garnet, sillimanite, spinel and cordierite (KLP 1-1).

which includes an isothermal decompression stage after the peak metamorphism was correlated to rapid exhumation history associated with the extensional collapse following collision of continental blocks during the assembly of Gondwana in the late Pan-African. Santosh et al. (2006) reported results from a comprehensive study of the petrology, P-T evolution and reported electron-probe (EPMA) ages for zircon and monazite from five key metapelite localities in the Trivandrum Block. Their data suggested that the protolith sediments of the khondalite were derived from Archean to Proterozoic sources with deposition after 569 Ma. A late Neoproterozoic depositional age for the precursor sediments of khondalite suite was also reported in other studies (e.g., Collins et al., 2007, 2014; Santosh et al., 2009), thus correlating the metapelitic assemblage as part of a Neoproterozoic (Pan-African) accretionary belt in a subduction environment (Santosh et al., 2009, 2017). However, recent studies on the khondalites from Trivandrum Block identified possible Paleoproterozoic metamorphic domains, in addition to the Late Neoproterozoic-Cambrian metamorphic overprint (e.g., Harley and Nandakumar, 2016). Liu et al. (2016) performed a detailed zircon Pb study on khondalites from the Trivandrum Block and interpreted that the detrital zircon grains were derived from a ca. 2 Ga or slightly older Paleoproterozoic source, with some of the grains indicating a late Paleoproterozoic highgrade metamorphic event, probably some time prior to 1.8 Ga. Similar Paleoproterozoic monazite and zircon grains were also reported in some other previous studies (Braun et al., 1998; Santosh et al., 2005, 2006). Thus, Liu et al. (2016) interpreted that the khondalites in Trivandrum Block were part of an extensive Paleoproterozoic metasedimentary assemblage that might have

extended from southern Madagascar via the Trivandrum Block to the Highland Complex of Sri Lanka. In this study, we investigate two classic sections in the Trivandrum Block (Fig. 1), at Koliakode and Kulappara, where granulite facies metapelites are exposed in fresh rock quarries (Fig. 2). The details of sample locations are shown in Table 1. These localities were also investigated in some of the previous studies such as Harley and Nandakumar (2014) and Liu et al. (2016). At Koliakode, medium to coarse grained bands of cordierite-garnet-spinel-sillimanite-biotite±graphite ranging in width from few cm up to a meter alternate with quartzo-feldspathic layers of variable width. The felsic layers are composed of K-feldspar, plagioclase-, and quartz, with minor biotite, and the rock shows migmatitic feature. Bluish domains rich in coarse cordierite, sometimes associated with flakes and books of graphite, are common in this locality. The Kulappara quarry exposes similar rocks, although with less cordierite and spinel, and more felsic components including leucocratic pegmatites and veins, some which carry coarse flakes and flaky aggregates of graphite. In both locations, garnet is a common constituent, and occurs both as porphyroblastic grains and as medium to fine grains in the matrix. Large garnet grains show partial breakdown with a corona of bluish Table 1 Sample numbers, rock types, localities and GPS readings. Sample

Location

Block

Rock type

GPS reading

KLD 1-1,2,3,4

Koliakode

Trivandrum

Khondalite

KLP 1-1,2,3

Kulappara

Trivandrum

Khondalite

76 530 2700 E 8 380 2100 N 76 530 3200 E 8 300 05” N

4

B. Yu et al. / LITHOS 348-349 (2019) 105195

cordierite and needles of sillimanite warping around. Spinel occurs in association with garnet, sillimanite and cordierite as medium to fine grained aggregates. We collected khondalite samples from the two localities which carry representative mineral assemblages for petrologic and geochronologic studies. 3. Analytical methods 3.1. Petrography Polished thin sections for petrographic studies and electron probe microanalysis (EPMA) were prepared from the representative samples of the metapelites at the School of Earth and Space Sciences, Peking University, China. Petrographic and thin section studies were carried out at China University of Geosciences Beijing, China. 3.2. Geochemistry Whole-rock geochemical analyses of the metapelite samples were carried out at the using X-ray fluorescence (XRF) at the National Research Centre of Geoanalyses, Beijing, China, after fragmentation in a jaw crusher and manual powering to 200 mesh. The instrument used was XRF model PW 4400, following procedures described in Gao et al. (2008). The Fe2O3 and FeO contents were determined by conventional wet chemical analysis and titration. Loss on ignition was obtained using about 1 g of sample powder heated at 980  C for 30 min. Counterpart chips of the samples used for petrological studies were crushed and powdered to 200 mesh. Loss on ignition was obtained with sample powder (1 g) heated at 980  C for 30 min. The analytical uncertainties for the major element oxides are below 0.5%. The results are presented in Table 2. 3.3. Zircon and monazite geochronology 3.3.1. Zircon U-Pb LA-ICP-MS geochronology Zircon separation was performed at the Yu'neng Geological and Mineral Separation Survey Centre, Langfang city, Hebei Province, China using conventional heavy liquid and magnetic techniques from crushed rock samples, followed by hand selecting under a binocular microscope. Before gold sputter coating, the zircon grains were mounted in epoxy resin discs and polished to reveal midsections at Beijing Geoanalysis Centre, China. For choosing appropriate target sites for U-Pb analyses and to check the internal zoning, the zircon grains were imaged under both transmitted and reflected light, as well as under cathodoluminescence (CL). The CL imaging at the Beijing Geoanalysis Centre used scanning electron microscope (JSM510) equipped with Gantan CL probe, and transmitted and reflected light images were examined by a petrological microscope. Table 2 Major element composition of metapelites from the khondalite belt of Trivandrum Block analyzed in this study. Sample no.

KLD 1-1

Rock type Major element (%) SiO2 71.4 TiO2 0.44 Al2O3 14.33 T Fe2O3 3.53 MnO 0.06 MgO 1.32 CaO 3.68 Na2O 3.47 K2O 0.88 P2O5 0.14 LOI 0.56 Fe2O3 0.25 FeO 2.95

KLP 1-2

KLD 1-3

KLD 1-4

67.66 0.6 14.54 7.76 0.08 1.87 2.31 3.44 1.22 0.09 0.29 1.04 6.05

51.68 0.02 28.97 8.67 0.074 6.48 1.12 1.15 0.38 0.09 1.02 1.34 6.6

Khondalite 57.19 0.77 22.36 9.57 0.15 2.94 0.74 1.48 4.28 0.11 0.38 0.35 8.3

The U-Pb analyses of zircon was conducted by LA-ICP-MS at the Wuhan Sample Solution Analytical Technology Co., Ltd., Wuhan, China. Laser sampling was performed using a GeolasPro laser ablation system that consists of a COMPexPro 102 ArF excimer laser (wavelength of 193 nm and maximum energy of 200 mJ) and a MicroLas optical system. An Agilent 7700e ICP-MS instrument was used to acquire ion-signal intensities. Helium was applied as a carrier gas. Argon was used as the make-up gas and mixed with the carrier gas via a T-connector before entering the ICP. A “wire” signal smoothing device is included in this laser ablation system. The laser spot diameter and frequency were set to 30 mm and 10 Hz in this study. Zircon 91,500 was used as the external standard, zircon GJ-1 was analyzed as an unknown to monitor the data quality and silicate glass NIST 610 was used to optimize the instrument (Morel et al., 2008). Each analysis incorporated a background acquisition of approximately 20e30s followed by 50s of data acquisition from the sample. An Excel-based software ICPMSDataCal was used to perform off-line selection and integration of background and analyzed signals, time-drift correction and quantitative calibration for trace element analysis and U-Pb dating (Liu et al., 2008, 2010). Concordia diagrams and weighted mean calculations were made using ISOPLOT R software (Vermeesch, 2018). 3.3.2. Monazite U-Pb LA-ICP-MS geochronology Monazite separation was also performed at the Yu'neng Geological and Mineral Separation Survey Centre, Langfang city, Hebei Province, China. The U-Pb dating of monazite was conducted by LA-ICP-MS at the Wuhan Sample Solution Analytical Technology Co., Ltd., Wuhan, China. Laser sampling was performed using a GeolasPro laser ablation system that consists of a COMPexPro 102 ArF excimer laser (wavelength of 193 nm and maximum energy of 200 mJ) and a MicroLas optical system. An Agilent 7700e ICP-MS instrument was used to acquire ion-signal intensities. Helium was applied as a carrier gas. Argon was used as the make-up gas and mixed with the carrier gas via a T-connector before entering the ICP. A “wire” signal smoothing device is included in this laser ablation system, by which smooth signals are produced even at very low laser repetition rates down to 1 Hz (Hu et al., 2015). The spot size and frequency of the laser were set to 30 mm and 5 Hz, respectively. Monazite standard 44,069 and glass NIST610 were used as external standards for U-Pb dating and trace element calibration, respectively. Each analysis incorporated a background acquisition of approximately 20e30 s followed by 50 s of data acquisition from the sample. An Excel-based software ICPMSDataCal was used to perform off-line selection and integration of background and analyzed signals, time-drift correction and quantitative calibration for trace element analysis and U-Pb dating (Liu et al., 2008; Liu et al., 2010). Concordia diagrams and weighted mean calculations were made using Isoplot R (Vermeesch, 2018). 3.4. Phase diagram methodology Pseudosection construction and analysis are based on defining the bulk composition of the equilibration volume of a sample (e.g., Palin and White, 2016) (Table 3). The whole rock analytical data on counterpart rock chips used for petrologic studies were used in this study. All phase diagrams were constructed using the petrological modelling software TheriakeDomino (De Capitani and Brown,1987; De Capitani and Petrakakis, 2010) and internally consistent thermodynamic data set ds-62 (Holland and Powell, 2011) in the 11component MnOeNa2OeCaOeK2OeFeOeMgOeAl2O3eSiO2eH2Oe TiO2eO2 (MnNCKFMASHTO) system. The activityecomposition (aex) relations for solid-solution phases used in this study are as follows: garnet, plagioclaseeK-feldspar, haplogranite melt, white mica (muscoviteeparagonite), biotite, orthopyroxene, staurolite, cordierite, chlorite, spinelemagnetite, and ilmeniteehematite (White et al.,

B. Yu et al. / LITHOS 348-349 (2019) 105195

2014a, 2014b). Pure phases comprised rutile, titanite, quartz, kyanite, sillimanite, andalusite, and aqueous fluid (H2O). Uncertainty on the absolute positions of assemblage field boundaries in any individual phase diagram are estimated to be less than ±1 kbar and ± 50  C at the 2s level (Palin et al., 2016; Powell and Holland, 2008), with such variation being largely a function of propagated uncertainty on endmember thermodynamic properties within the data set (Forshaw et al., 2019). However, as all phase diagrams were calculated using the same data set and aex relations, similar absolute errors associated with end-members cancel, and calculated phase equilibria are thus relatively accurate to within ±0.2 kbar and ± 10e15  C (Powell and Holland, 2008). The fluid content of the modelled bulk composition was converted from calculated LOI contents in each sample.

5

coarse grained, and plagioclase shows polysynthetic twinning. The biotite content in this rock is less than that in the other samples (Fig. 3e). 4.1.2. Sample KLP (KLP 1-2) The main minerals in this rock are garnet, sillimanite, cordierite, biotite, and feldspars. In contrast to sample KLD, the KLP samples do not show well defined compositional layer, and also contains less spinel and more biotite. Coarse mesoperthitic feldspar is common these samples. Vermicular intergrowth of Crd þ Qz symplectite surrounding almost completely resorbed garnet grains, similar to the texture in KLP samples also occurs (Fig. 3f). 4.2. Petrological modelling

4. Results 4.1. Petrography The samples labelled KLD and KLP selected for detailed study were collected from fresh exposures in working quarries at Koliakode and Kulappara in the khondalite belt of Trivandrum block. The khondalites analyzed in this study are granulite-facies metapelites composed of Grt þ Sil þ Crd þ Spl þ Kfs þ Bt þ Qz þ Pl þ Gr with accessory zircon and monazite. Samples KLD1-1, KLD1-2, KLD1-3, KLD1-4 and KLP1-2 were used for petrographic analysis. Representative thin section photomicrographs showing representative mineral assemblages are given in Fig. 3. In the petrographic analysis below, peak and retrograde mineral assemblages are interpreted on the basis of grain size and microstructures. 4.1.1. Sample KLD (KLD 1-1, 2, 3, 4) The major minerals of samples from KLD are garnet, feldspar, cordierite, biotite and sillimanite. The secondary minerals are zircon, magnetite and ilmenite. In thin section, all the samples display coarse grained garnet and cordierite with oriented sillimanite. Sample KLD 1-1 is a coarse-grained well-foliated rock with garnetrich and sillimanite-rich layers. Garnet forms porphyroblasts of up to 1e2 mm across and occurs as coarse grain in the matrix and also as smaller subhedral grains surrounded by biotite and sillimanite. Kfeldspar is medium to small grained. Plagioclase occurs as minor small grains. Brown biotite laths occur parallel to foliation (Fig. 3 a,b). Sample KLD 1-2 shows similar assemblage as KLD 1-1 with garnet occurring as coarse grains in the matrix and also as small, resorbed grains within porphyroblastic moats of cordierite. Large garnet porphyroblasts contain inclusions of sillimanite and quartz. The cordierite contains inclusions of Sill þ Grt. K-feldspar is medium to small grained, and the biotite around 10% also similar as KLD 1-1, shown it broken type and most parallel to foliation (Fig. 3c). Cordierite also occurs as coarse grains in sample KLD 1-3. Crd þ Qz symplectites and cordierite corona around garnet are also present (Fig. 3d). These textures represent the KFMASH bivariant reaction: Grt þ Sil þ Qz ¼ Crd. K-feldspar is medium to small grained. Sillimanite occurs as elongate tabular grains or as small inclusions within garnet and K-feldspar (Fig. 3d). Sample KLD 1-4 has abundant green spinel and less sillimanite Garnet shows breakdown to cordierite. K-feldspar is medium to

4.2.1. KLD 1-3: Prograde metamorphism The calculated PeT pseudosection for sample KLD 1-3 is shown in Fig. 4a. The observed matrix assemblage Sill þ Grt þ Crd is considered to be stable at 6e9 kbar and 760e790  C (Fig. 4a), as constrained at low pressure (<6 kbar) by the stabilization of cordierite and at higher temperature by the stabilization of K-feldspar. The occurrence of mesoperthite indicates that peak metamorphism reached ultrahigh temperatures of probably >900  C. Foliated sillimanite and garnet, and sillimanite inclusions within garnet indicate textures developed during prograde metamorphism. Biotite breaks down to react with sillimanite and quartz forming garnet, K-feldspar and cordierite, which also indicate a prograde metamorphic stage. Biotite and sillimanite in sample KLD 1-3 show strong foliation, and may have recrystallized during retrograde metamorphism as evidenced from the late-stage fabric-forming deformational event. 4.2.2. KLD 1-4: Peak metamorphism The computed phase equilibria for sample KLD 1-4 is shown in Fig. 4b, which show peak PeT conditions in the range of 6.5e7 kbar and 1010e1030  C, where Sp, Sill, Grt and Ksp coexist. Spinel is stabilized at pressures below 8 kbar, and at temperatures above 1000  C. Spinel is intercalated with sillimanite and occurs together with elongated K-feldspar, all of them wrapped around garnet porphyroblast, indicating equilibrium assemblage. Inclusions of sillimanite and quartz within spinel indicate reactions of sillimanite and garnet to form the spinel and quartz and suggest peak metamorphism above 1000  C. Spinel in direct contact with quartz in these rocks also suggest UHT conditions of 8 kbar, >1000  C. The foliated texture of sillimanite and spinel, and inclusions of sillimanite in garnet indicate from the prograde to peak grade metamorphic stages, with sillimanite and spinel fabric formed during the peak stage. A T-Xo model was computed at 7 kbar to show the oxidation changes between minerals from all Fe as Fe2þ to 27% Fe as Fe3þ, where sappharine is stable at high temperature > 1000  C (Fig. 5). Spinel and ilmenite coexist g at low oxidation field, and the peak assemblage for this sample is stable at 0.25e0.60 for Mo content. Spinel-quartz remains stable when oxidation state changes at ultra-high temperatures (Boger et al., 2012; Korhonen et al., 2013; Yakymchuk and Brown, 2019). 4.2.3. KLP 1-2: Retrograde metamorphism The calculated PeT pseudosection for sample KLP 1-2 is shown in Fig. 4c, where the observed assemblage Grt þ Sill-Crd-Q-Mt-Bt is

Table 3 Bulk-rock compositions utilised in petrological modelling. Sample no.

H2O

SiO2

Al2O3

CaO

MgO

FeOtot

K2O

Na2O

TiO2

MnO

O

XFe3þ

XMg

KLP-1-2 KLD-1-3 KLD-1-4

1.43 1.04 3.76

64.40 73.08 57.10

14.84 9.26 18.86

0.72 2.54 1.19

4.93 3.01 10.67

8.11 6.31 7.21

3.07 0.84 0.27

1.62 3.60 1.23

0.65 0.49 0.02

0.14 0.07 0.07

0.15 0.42 0.56

0.04 0.13 0.15

0.38 0.32 0.60

FeOtot is total iron expressed as FeO. O is oxygen, which combines with FeO via the equation 2FeO þ O ¼ Fe2O3; hence, bulk O is identically equal to bulk Fe2O3, while true bulk FeO is given by FeOtot e (2  O). Bulk-rock XMg ¼ MgO/(MgO/FeOtot), and XFe3þ ¼ (2  O)/FeOtot.

6

B. Yu et al. / LITHOS 348-349 (2019) 105195

Fig. 3. Representative thin section photomicrographs of the rocks analyzed in this study. (a) garnet-rich and sillimanite-rich layers khondalite showing the (b) khondalite on Orthogonal showing garnet-biotite-sillimanite symplectite (KLD 1-1) (c) khondalite showing assemblage of garnet, sillimanite and cordierite (KLD 1-2)(d) khondalite showing the garnet porphyroblasts break down to Oz þ Crd symplectite (KLD 1-3) (e) khondalite showing the assemblage of spinel, garnet and plagioclase (KLD 1-4), (f) khondalite showing vermicular intergrowth of Crd þ Qz symplectite (KLP 1-2).

stable at 4.5e6.5 kbar and 770e950  C. Cordierite is stabilized at low pressures (<6 kbar), and K-feldspar is stabilized at higher pressure and high temperatures. Pokiloblastic garnet with abundant inclusions of sillimanite, cordierite, and quartz suggest retrograde metamorphism at temperatures below 900  C. Foliated biotite layers suggest that the rock formed during later metamorphic deformation. Matrix sillimanite within cordierite indicates reaction of sillimanite with garnet to form cordierite and quartz. This rock appears to be at a lower-grade than the interpreted peak conditions for sample KLD 1-4, and it is likely that this sample fully re-equilibrated during exhumation, and preserved PeT conditions defining the retrograde path. 4.3. Zircon morphology and geochronology Three representative samples of the khondalites, KLD 1e1, KLD1e3 and KLP 1e3, were chosen for zircon geochronology using the LA-ICP-MS U-Th-Pb method. The representative zircon cathodoluminescence images are shown in Fig. 6.The U-Pb analytical

results are presented in Supplementary Table 1 and the data are plotted in concordia diagrams in Figs. 7 and 8. A summary of the zircon characteristics and age results are given below. 4.3.1. Khondalite (KLD 1-1) The zircon grains from the khondalite sample KLD 1-1 range in length from 50 to 100 mm, with length ratio ranges from 1:1 to 2:1. Most of the grains are colorless, transparent to translucent and occur as euhedral to subhedral grains with long-prismatic morphology. In CL images (Fig. 6), most zircon grains display heterogeneous texture and some of them show core-rim texture. The majority show clear igneous domains and zoning in CL images, and are surrounded by thin rim interpreted as metamorphic recrystallization or overgrowth. A total of 44 U-Th-Pb spots were analyzed from sample KLD 1e1. The Th contents range from 13 to 1273 ppm and U contents range from 41 to 1759 ppm, with Th/U ratios in the range of 0.07 to 3.3 (Supplementary Table 1). Excluding the data from several grains which are discordant (discordance <25%, considered as statistical

B. Yu et al. / LITHOS 348-349 (2019) 105195

7

Fig. 4. Pressureetemperature (PeT) pseudosections for samples (a) KLD 1-3, (b) KLD 1-4, (c) KLP 1-2, calculated using the bulk compositions given in Table 3. Not all fields are labelled for clarity, with numbered fields as follows: 1-Pl Grt Bt Crd Ilm Sill Q; 2-Pl Grt Crd Mt. Ilm Sill Q; 3-Liq Pl Grt Mt. Ilm Sill Q; 4-Liq Pl Grt Mt. Ilm Q; 5-Liq Ksp Grt Crd Mt. Ilm Q; 6-Liq Pl Grt Bt Ilm Ky Q Ru; 7-Ksp Grt Bt Mt. Ilm Sill Q; 8-Liq Ksp Grt Bt Mt. Ilm Sill Q; 9-Ksp Grt Bt Crd Mt. Ilm Sill Q; 10-Liq Pl Grt Bt Crd Mt. Ilm Q; 11-Liq Pl Grt Opx Mt. Ilm Q; 12-Liq Ksp Grt Opx Crd Mt. Ilm Q; 13-Liq Ksp Opx Mt. Ilm Q; 14-Liq Grt Sapp Crd Ilm Sill Q; 15-Liq Ksp Grt Sapp Crd Ilm Sill Q; 16-Liq Ksp Grt Sapp Crd Sp Ilm Sill Q; 17-Liq Ksp Grt Sapp Crd Sp Sill Q; 18-Liq Ksp Grt Sapp Crd Sp Q; 19-Liq Grt Sapp Mt. Ilm Q; 20-Liq Grt Opx Sapp Mt. Ilm Q; 21-Liq Grt Opx Sapp Crd Mt. Q; 22-Liq Ksp Opx Crd Sp Q; 23-Liq Ksp Crd Mt.

outliers), 26 of the 44 analyses form a coherent age group within analytical error. The twenty-six analyses are plotted on the concordia in Fig. 7b. The 207Pb/206 Pb spot ages range from 1798 Ma to 2105 Ma, and yield an upper intercept age of 2100 ± 20 Ma and lower intercept age of 549.2 ± 41 (MSWD ¼ 1.2, N ¼ 26). Although most of the grains are discordant, a few concordant grains show minor distinction with the upper intercept ages. These features suggest variable leadeloss of the zircon grains. The zircon cores mostly define ages in the range of 2.0 and 2.6 Ga with some older grains in the range of 2.9e3.1 Ga (Fig. 7a).

4.3.2. Khondalite (KLD 1-3) The zircon grains from the khondalite sample KLD 1-3 range in length from 80 to 120 mm, and are mostly around 100 mm, with length: width ratio of 1:1 to 2:1. The grains are mostly subhedral to euhedral, with elliptical to rounded morphology. Morphologically these can divided into two populations. One group comprises larger grains with irregular shape, colorless to light brown in color, and transparent to translucent. The core domains are mostly weakly zoned but some of them show oscillatory zoning and fir-tree zoning. The zircon grain with spot 1 shows chaotically zoned

8

B. Yu et al. / LITHOS 348-349 (2019) 105195

Fig. 5. Calculated T-Mo diagram at 7 kbar, yellow bar indicates O content used for P-T pseudosection modelling; method and data are using Theriak Domino and ds-62 (De Capitani and Brown, 1987; De Capitani and Petrakakis, 2010; Holland and Powell, 2011). Top data are composition used for modelling.

cores, with some domains displaying sector zoning. As revealed by CL images (Fig. 6), about half of the zircon population shows multiple phases of recrystallization. Spots 21 display core-rim texture with bright cores and dark overgrowth rims possibly formed during the metamorphic event.

Fig. 6. CL images of zircon grains from samples KLD and KLP.

A total of thirty U-Th-Pb spots were analyzed on zircon grains from sample KLD 1-3. The Th and U show low contents in the range of 10 to 803 ppm and 99 to 1911 ppm respectively. The Th / U ratios range from 0. 01 to 1.25, and most of the values are lower than 0.1, suggesting that these grains are mostly of metamorphic origin (Supplementary Table 1). Excluding the several discordant spots, the data define coherent age groups within analytical error (Fig. 7c). The 207Pb/206 Pb spot ages range from 487 Ma to 2200 Ma, and yield an upper intercept age of 2189 ± 24.8 Ma and lower intercept age of 568 ± 4.7 (MSWD ¼ 2.3, N ¼ 30). The concordant spots of the younger group yield 206Pb/238U weighted mean age of 559 ± 4 Ma (MSWD ¼ 11; N ¼ 5) (Fig. 7c). We interpret this age to represent recrystallization newly grown grains associated with late Neoproterozoic metamorphism (Yakymchuk and Brown, 2019). 4.3.3. Khondalite (KLP 1-3) Zircon grains from sample KLP 1-3 are mostly colorless or light brown. They show elliptical morphology with length around 100 mm, and a length to width ratio of 1:1 to 2:1. In CL images (Fig. 6), they display distinct core-rim structures and dark overgrowth rims, which are considered to be the result of metamorphic recrystallization or overgrowth caused by crystallization from fluids. More than half of the grains show broken euhedral terminations in the core with dark overgrowth rims possibly formed during the metamorphic event. Another three grains (spots 41, 9, 6) possess chaotically zoned and sector zoned cores, with few cores showing clear magmatic oscillatory zoning indicating a metamorphic origin.

B. Yu et al. / LITHOS 348-349 (2019) 105195

9

Fig. 7. LA ICPMS U-Pb data on zircon for KLD 1-1 & 3. Zircon U-Pb concordia plots and weighted average plots for khondalite (a) all data on sample KLD 1-1, with the details on (b) and (c) KLD 1-3. Inset shows data for KLD 1-3. All data point uncertainties 2s, and the weighted mean ages bar width is the same as uncertainties and include systematic error.

Fig. 8. LA ICPMS U-Pb data on zircon for KLP 1-3. Zircon U-Pb concordia plots and weighted average plots for khondalite (a) all data on sample KLP 1-3, with the details on (b). Inset shows data for KLP 1-3, for overgrowth grains. All data point uncertainties 2s, and the weighted mean ages bar width is the same as uncertainties and include systematic error.

10

B. Yu et al. / LITHOS 348-349 (2019) 105195

A total of 40 analyses were carried out on both magmatic core domains and metamorphic rim domains from 59 zircon grains. The Th and U show variable contents ranging from 15 to 1116 ppm and 31 to 1749 ppm respectively. The Th / U ratios show a wide range from 0.008 to 1.4. Most of values are lower than 0.1 (Supplementary Table 1). Excluding the discordant analyses, 40 analyses form three age groups within analytical error. Among these, 30 spots yield 207Pb/206 Pb ages ranging from 543 to 2150 Ma, with an upper intercept age of 2164 ± 7.4 Ma and lower intercept age of 560.3 ± 12.2 Ma (MSWD ¼ 2, N ¼ 30; Fig. 8b). The concordant spots of younger age group yield 206Pb/238U weighted mean age of 528.9 ± 20 Ma (MSWD ¼ 4; N ¼ 3) (Fig. 8a). The remaining grains show concordant spot ages around 2.4 and 2.7 Ga, and one grain shows 2995 Ma. 4.4. Monazite morphology and U-Pb geochronology In situ Laser Ablation-Inductively Coupled-Mass Spectrometry (LA-ICP-MS) was conducted on monazites from the same three samples from which the zircon grains were analyzed. The analytical results are presented in Supplementary Table 2 and representative BSE images of grains are given in Fig. 9. The monazite grains in the khondalite samples range in length from 50 to 100 mm, with length ratio ranges from 1:1 to 2:1, and mostly around 1:1. Some grains have irregular shape. Most are colorless, transparent to translucent and occur as subhedral prisms. In backscatterred electron (BSE) images (Fig. 9), they are relatively homogenous without any prominent core-rim texture. 4.4.1. Khondalite (KLD 1-1) A total of twenty-two U-Th-Pb spots were analyzed on monazite from sample KLD 1-1. The Th contents range from 5980 to 212,114 ppm and U contents range from 5018 to 27,545 ppm. All of the 30 analyses within analytical error are plotted on the concordia curve in Figs. 10a, b. Except for five spots with higher discordance around 1.9 Ga, the remaining 16 analyses plot on the concordia. The 206 Pb/238U ages could be divided into three groups: the younger age

Fig. 9. BSE images of monazite from samples KLD and KLP.

group yield 206Pb/238U weighted mean age of 466.0 ± 11.2 Ma (MSWD ¼ 4; N ¼ 13) (Fig. 10b), the middle age group is represented by one grain at 509 Ma, and the older age group shows 206Pb/238U weighted mean age of 551.8 ± 3.2 Ma (MSWD ¼ 6; N ¼ 3). 4.4.2. Khondalite (KLD 1-3) A total of seventeen U-Th-Pb spots were analyzed on monazite from sample KLD 1-3. The Th contents range from 14,011 to 258,620 ppm and U contents range from 1372 to 14,209 ppm. Seventeen analyses plot on the concordia curve and are divided into two age groups. The younger group yields 206Pb/238U weighted mean age of 461.2 ± 7.5 Ma (MSWD ¼ 12; N ¼ 3), whereas the other group falls along a well-defined discordia line forming two coherent age groups within analytical error (Fig. 10 c,d). The 207 Pb/206 Pb spot ages range from 561 Ma to 1962 Ma and yield an upper intercept age of 1976.0 ± 25.7 Ma and lower intercept age of 560.7 ± 11.2 (MSWD ¼ 1.2, N ¼ 13). The concordant spots around the lower intercept yield 206Pb/238U weighted mean age of 571.0 ± 9 Ma (MSWD ¼ 0.89; N ¼ 6). 4.4.3. Khondalite (KLP 1-3) A total of twenty-two U-Th-Pb spots were analyzed on monazite from sample KLP 1-3. The Th contents range from 36,552 to 271,920 ppm and U contents range from 873 to 25,822 ppm. All of the 22 analyses within analytical error are plotted on the concordia curve (Fig. 11a, b). Two concordant spots show ages around 1. 8 Ga. The 207Pb/206 Pb ages of the remaining 20 grains can be divided into three age groups. The he youngest age group yields 206Pb/238U weighted mean age of 489.6 ± 6.8 Ma (MSWD ¼ 2.5; N ¼ 6) (Fig. 11 b), whereas the middle age group shows 206Pb/238U weighted mean age of 521.3 ± 6.7 Ma (MSWD ¼ 0.9; N ¼ 8). The older age group yields 206Pb/238U weighted mean age of 543.2 ± 5.5 Ma (MSWD ¼ 0.2; N ¼ 6). 5. Discussion 5.1. Pressureetemperature evolution Several previous petrological and phase equilibria studies were carried out on the granulite facies metapelites from the khondalite belt of Trivandrum Block. Harley and Nandakumar (2014) reported Grt þ Crd þ SillþQtz þ Bt þ Ilm þ 2feldspar assemblage from Kulappara in the Trivandrum Block, with peak P-T conditions of 7.5 kbar, 900  C, which shows similar mineral phases with our sample KLD 1-3. In a further study, Harley and Nandakumar (2016) combined isopleth of garnet and cordierite with phase equilibrium, and defined pressures of 6.2e6.6 kbar and temperature at 840 ± 20  C. Based on zircon trace element composition and Tizircon thermometry, these authors constrained temperatures at c.779e792  C for zircon crystallization and modification during cooling or retrograde metamorphism. The spinel-quartz assemblage in our samples indicate peak P-T conditions at 6.5e7 kbar and 1010e1030  C. Morimoto et al. (2004) reported Grt þ Sill þ Bt þ Crd þ Spl þ Kfs þ Pl þ Qz ± graphite assemblage from Chittikara in the khondalite belt and reported temperatures from spinelcordierite thermometry at 900  C, garnet- cordierite thermometer at 818  C, and feldspar ternary composition at c.1000  C. Garnetþ sillimanite þ plagioclase þ quartz barometer shows decompression from 8 kbar to 4 kbar, suggesting a clockwise P-T path (Morimoto et al., 2004). In another study, Tadokoro et al. (2008) compared the P-T estimates from eastern part of Trivandrum Block which are consistent with those from the western part and show peak temperature at 900e1000  C, whereas spinel-quartz barometer shows higher pressure (12 kbar) in the eastern part than that in the western part (10 kbar). In a recent study, Kadowaki et al. (2019 ) reported khondalites from Elavinmoodu quarry, west of

B. Yu et al. / LITHOS 348-349 (2019) 105195

11

Fig. 10. LA ICPMS U-Pb data on monazite for KLD 1-1 & 3. Monazite U-Pb concordia plots and weighted average plots for khondalite (a) all data on sample KLD 1-1, with the details on (b) and (c) KLD 1-3, with details on (d). Inset shows data for KLD 1-3. All data point uncertainties 2s, and the weighted mean ages bar width is the same as uncertainties and include systematic error.

Fig. 11. LA ICPMS U-Pb data on monazite for KLP 1-3. Monazite U-Pb concordia plots and weighted average plots for khondalite (a) all data on sample KLP 1-3, with the details on (b). All data point uncertainties 2s, and the weighted mean ages bar width is the same as uncertainties and include systematic error.

12

B. Yu et al. / LITHOS 348-349 (2019) 105195

Trivandrum Block, where the assemblage of Grt þ Qtz þ Pl þ Kfs þ Sill þIlm þ Mag/Spl indicates peak metamorphism at 6e7.6 kbar and 920e1030  C. Garnet growth is associated with zircon HREE depletion at >810  C during the prograde metamorphism, and retrograde metamorphism is represented by breakdown of garnet at a pressure of 4 kbar and temperature of 750  C. Johnson et al. (2015) reported Grt þ Kfs þ Sill þPl þ Mag þ Ilm þ Spl assemblage from a khondalite layer in the Nagercoil Block with peak P-T conditions of 6e8 kbar and > 900  C, followed by retrograde decompression at 5 kbar and 800  C. They also defined a clockwise P-T path. Phase equilibrium modelling from sample KLD 1e3 shows that the Sill þ Grt þ Crd assemblage was stable at 6e9 kbar and 760e790  C, with the breakdown of biotite and growth of garnet indicating the prograde stage. Spinel-quartz from sample KLD 1-4 suggests peak metamorphic conditions of 6.5e7 kbar and 1010e1030  C consistent with ultra-high temperature metamorphism in the Trivandrum Block and surrounding areas, as also reported in previous studies from this block as well as other blocks of the SGT (Cenki et al., 2002; Li et al., 2019; Shazia et al., 2015; Tsunogae et al., 2008). Growth of cordierite and biotite at the expense of garnet, and the foliation texture suggest retrograde metamorphism at 4.5e6.5 kbar and 770e950  C. Our results from the metapelites from KLD and KLP allow an interpretation of the prograde, peak, and retrograde PeT conditions of the khondalites as shown in Fig. 12. Spinel and quartz assemblage indicates that the rocks equilibrated at temperatures >1000  C. From this peak condition, isothermal decompression resulted in garnet breakdown and cordierite formation, followed by retrograde cooling, defining an overall clockwise PeT evolution. The P-T path defined in our study is broadly similar to those obtained in previous studies, although the detailed modelling of the prograde, peak and retrograde metamorphic conditions in our study provides better constrains on the overall metamorphic evolution of the khondalites. The ultra-high temperatures obtained in our study correlate with the thermal event associated with the late Neoproterozoic-Cambrian collisional assembly of continental fragments within the Gondwana supercontinent (Johnson et al., 2015; Shimizu et al., 2009; Tateishi et al., 2004).

sediments of the metapelites were probably deposited >2.1 Ga ago and were subsequently intruded by granitoid rocks at ca. € ner et al., 2015; Liu et al., 2016). We show a 1765e2100 Ma (Kro compilation of the zircon data from our study and its comparison with other studies in Fig. 13, where there are two distinct major peaks, one at Paleoproterozoic, and the other at latest Neoproterozoic-Cambrian. The younger peak corresponds to the high- to ultrahigh temperature metamorphic event. The zircon grains from both localities in this study show both similar late Neoproterozoic metamorphic ages around 550 Ma. They also carry Paleoproterozoic grains with ages in the range of 1.5e2.2 Ga, together with a few grains showing Archean ages.. The younger ages around 550e560 Ma are similar to the lower intercept age, whereas the wide range of Paleoproterozoic ages fall along the discordia, although with similar 207Pb/206Pb ratios. We infer that the major Pb-loss in the older zircon grains occurred during late Neoproterozoic-Cambrian associated with the ultrahightemperature metamorphism. The zircon grains also show ages around 2.4e2.7 from late Achaean to early Paleoproterozoic, and

5.2. U-Pb ages and implications Recent geochronological studies on khondalites from the Trivandrum Block concluded, based on the concordant ages and discordant minimum 207Pb/206Pb ages that the precursor

Fig. 12. Summary plot showing the interpreted PeT evolution of khondalite.

Fig. 13. Zircon U-Pb ages using the kernel density distribution approach for (a) khondalite (this study); (b) khondalite (Liu et al., 2016); (c) granulite (Taylor et al., 2014).

B. Yu et al. / LITHOS 348-349 (2019) 105195

13

metamorphism and deformational events associated with the amalgamation of the Gondwana supercontinent from the southern part of the Southern Granulite Terrane in India including the Trivandrum Block (e.g., Braun et al., 1998; Braun and Kriegsman, 2003; Cenki and Kriegsman, 2005; Santosh et al., 2003, 2006, 2009; Collins et al., 2007, 2014; Clark et al., 2015; Johnson et al., 2015; Taylor et al., 2014). The metamorphism involved significant partial melting and formation of the garnet-bearing assemblages (Harley and Nandakumar, 2016), with multiphase recrystallization. The growth or recrystallisation of monazite at ca. 550 Ma and its further modification at ca. 470 Ma may indicate that the high-T metamorphism was long-lived. Recent studies have addressed the possibility that the khondalites of Trivandrum Block witnessed two metamorphic events, one during Paleoproterozoic and the other during late Neoproterozoic (Harley and Nandakumar, 2016; Liu et al., 2016), in deviation from the earlier concept that the precursors of the khondalites were deposited during mi- to late Neoproterozoic, followed by single stage metamorphism (e.g., Collins et al., 2007). Rare metamorphic zircon grains with Paleoproterozoic cores were reported by Liu et al. (2016) from the khondalites, and it was proposed that the khondalite sedimentary precursors were derived from Achaean to Paleoproterozoic sources. Harley and Nandakumar (2016) also suggested the possibility of multiple metamorphic events, an earlier Paleoproterozoic event and a subsequent late Neoproterozoic-Cambrian event. Although the zircon data in our study indicate Paleoproterozoic provenance, and the rare Paleoproterozoic monazite grains suggest the case for an earlier event, the data are not conclusive to evaluate multiple metamorphic events, and we correlate the timing of the dominant ultrahigh-temperature metamorphism of the khondalites to the Late NeoproterozoicCambrian thermal event associated with the birth of Gondwana. 6. Conclusions

Fig. 14. Monazite U-Pb ages using the kernel density distribution approach for (a) khondalite (this study); (b) khondalite (Santosh et al., 2006); (c) granulite (Taylor et al., 2014).

the oldest age we found at 3.2 Ga. These ages may suggest the khondalite sedimentary precursors are from different source and then suffer the two metamorphic event. Monazite is considered as a reliable geochronometer for granulite-facies metamorphism as it remains closed to Pb diffusion up to temperatures of around 900  C (Cherniak et al., 2004). Monazite has been demonstrated to be more reactive and responsive than zircon to metamorphism (e.g. Williams, 2001). Thus, our study that combines U-Pb geochronology of monazite grains and those in thin sections together with zircon age data provide robust insights into the metamorphic history of the khondalites. The LAICP-MS monazite U-Pb data show two distinct age populations at Paleoproterozoic and latest Neoproterozoic-Cambrian, although the Paleoproterozoic population is scarce (Fig. 14). The in situ monazite analysis in thin sections using EPMA U-Pb show results similar to those from LA-ICP-MS U-Pb analyses (our unpublished data). Several previous studies have reported widespread evidence for latest Neoproterozoic to Cambrian (ca. 570e515 Ma) high-grade

➢ Textural studies and mineral phase equilibria modelling suggest that the Sill þ Grt þ Crd represent prograde metamorphism at 6e9 kbar and 760e790  C. Equilibrium spinel-quartz assemblage corresponds to peak UHT metamorphism at of 6.5e7 kbar and 1010e1030  C. Garnet breakdown to cordierite and subsequent crystallization of biotite mark retrograde metamorphism at 4.5e6.5 kbar and 770e950  C. The metamorphic path corresponds to an overall clockwise PeT evolution. ➢ Zircons grains in the UHT metapelites display two distinct age peaks at Paleoproterozoic and Late Neoproterozoic. The 550e560 Ma metamorphic zircons are similar to the lower intercept age. ➢ The LA-ICP-MS monazite U-Pb data from grain separates show dominantly latest Neoproterozoic-Cambrian ages with a few Paleoproterozoic grains. ➢ The late Neoproterozoic-Cambrian ages are considered to mark the ultrahigh-temperature metamorphic event, associated with the assembly of the Gondwana supercontinent. The wide range of monazite ages indicate that the high temperature event was long-lived. Supplementary data to this article can be found online at https://doi.org/10.1016/j.lithos.2019.105195. Acknowledgements This manuscript greatly benefitted from the suggestions of Editor-in-Chief Prof. Marco Scambelluri and reviewer Dr. V. Nandakumar as well as the detailed and in-depth review by Dr. Chris Yakymchuk. We are thankful to them. This work forms part of the undergraduate research of Bing Yu. Funding for this study was provided by the Foreign Expert Grant to M. Santosh from the China University of Geosciences, Beijing.

14

B. Yu et al. / LITHOS 348-349 (2019) 105195

References Boger, S.D., White, R.W., Schulte, B., 2012. The importance of iron speciation (Feþ2/ Feþ3) in determining mineral assemblages: an example from the high-grade aluminous metapelites of southeastern Madagascar. J. Metam. Geol. 30 (9), 997e1018. Braun, I., Kriegsman, L.M., 2003. Proterozoic crustal evolution of southernmost India and Sri Lanka. In: Yoshida, M., Windley, B.F., Dasgupta, S. (Eds.), Proterozoic East Gondwana: Supercontinent Assembly and Breakup, vol. 206. Special Publication of the Geological Society, London, pp. 169e202. Braun, I., Montel, J.-M., Nicollet, C., 1998. Electron microprobe dating monazites from high-grade gneisses and pegmatites of the Kerala Khondalite Belt,southern India. Chem. Geol. 146, 65e85. Cenki, B., Kriegsman, L.M., 2005. Tectonics of the Neoproterozoic Southern Granulite Terrain, South India. Precambrian Res. 138, 37e56. Cenki, B., Kriegsman, L.M., Braun, I., 2002. Melt-producing and melt-consuming reaction in the Achankovil cordierite gneiss, South India. J. Metam. Geol. 20, 543e561. Chacko, T., Ravindra Kumar, G.R., Newton, C.R., 1987. Metamorphic PT conditions of the Kerala (South India) Khondalite Belt, a granulite facies supracrustal terrain. J. Geol. 95, 343e358. Cherniak, D.J., Watson, E.B., Grove, M., Harrison, T.M., 2004. Pb diffusion in monazite: a combined RBS/SIMS study. Geochim. Cosmochim. Acta 68, 829840. Clark, C., Healy, D., Johnson, T.E., Collins, A.S., Taylor, R.J.M., Santosh, M., 2015. Hot orogens and supercontinent amalgamation: a Gondwanan example using SHRIMP U-Pb geochronology, REE geochemistry, phase equilibria and numerical modelling. Gondw. Res. 28, 1310e1328. Collins, A.S., Santosh, M., Braun, I., Clark, C., 2007. Age and sedimentary provenance of the Southern Granulites, South India: U-Th-Pb SHRIMP secondary ion mass spectrometry. Precambrian Res. 155, 125e138. Collins, A.S., Clark, C., Plavsa, D., 2014. Peninsula India in Gondwana: the tectonothermal evolution of the Southern Granulite Terrain and its Gondwanan counterparts. Gondw. Res. 25, 190e203. De Capitani, C., Brown, T.H., 1987. The computation of chemical equilibrium in complex systems containing non-ideal solutions. Geochim. Cosmochim. Acta 51, 2639e2652. De Capitani, C., Petrakakis, K., 2010. The computation of equilibrium assemblage diagrams with Theriak/Domino software. Am. Mineral. 95, 1006e1016. Gao, S., Rudnick, R.L., Xu, W.L., Yuan, H.L., Liu, Y.S., Walker, R.J., Puchtel, I.S., Liu, X., Huang, H., Wang, X.R., Yang, J., 2008. Recycling deep cratonic lithosphere and generation of intraplate magmatism in the North China Craton. Earth Planet. Sci. Lett. 270 (1-2), 41e53. Forshaw, J.B., Waters, D.J., Pattison, D.R., Palin, R.M., Gopon, P., 2019. A comparison of observed and thermodynamically predicted phase equilibria and mineral compositions in mafic granulites. J. Metam. Geol. 37, 153e179. Harley, S.L., Nandakumar, V., 2014. Accessory mineral behaviour in granulite migmatites: a case study from the Kerala Khondalite Belt, India. J. Petrol. 55, 1965e2002. Harley, S.L., Nandakumar, V., 2016. New evidence for Palaeoproterozoic high grade metamorphism in the Trivandrum Block, Southern India. Precambrian Res. 280, 120e138. Holland, T.J.B., Powell, R., 2011. An improved and extended internally consistent thermodynamic dataset for phases of petrological interest, involving a new equation of state for solids. J. Metam. Geol. 29, 333e383. Hu, Z.C., Zhang, W., Liu, Y.S., Gao, S., Li, M., Zong, K.Q., Chen, H.H., Hu, S.H., 2015. “Wave” signal smoothing and mercury removing device for laser ablation quadrupole and multiple collector ICP-MS analysis: application to lead isotope analysis. Anal. Chem. 87, 1152e1157. Johnson, T.E., Clarks, C., Taylor, R.J.M., Santosh, M., Collins, C., 2015. Prograde and retrograde growth of monazite in migmatites: an example from the Nagercoil Block, southern India. Geosci. Front. 6 (3), 373e387. Kadowaki, H., Tsunogae, T., He, X.F., Santosh, M., Takamura, Y., Shaji, E., Tsutsumi, Y., 2019. Pressureetemperatureetime evolution of ultrahigh-temperature granulites from the Trivandrum Block, southern India: implications for long-lived high-grade metamorphism. Geol. J. (in press). Korhonen, F.J., Brown, M., Clark, C., Bhattacharya, S., 2013. Osumilite-melt interactions in ultrahigh temperature granulites: phase equilibria modelling and implications for their, P-T-t evolution of the Eastern Ghats Province, India. J. Metam. Geol. 31 (8), 881e907. € ner, A., 1984. Evolution, growth and stabilization of the Precambrian lithoKro sphere. Phys. Chem. Earth 15, 69e106. € ner, A., Santosh, M., Hegner, E., Shaji, E., Geng, H., Wong, J., Xie, H., Wan, Y., Kro Shang, C.K., Liu, D., Sun, M., Nanda-Kumar, V., 2015. Palaeoproterozoic ancestry of Pan-African high-grade granitoids in southernmost India: implications for Gondwana reconstructions. Gondw. Res. 27 (1), 1e37. Li, S.S., Palin, R.M., Santosh, M., Shaji, E., Tsunogae, T., 2019. Extreme thermal metamorphism associated with Gondwana assembly: evidence from sapphirine-bearing granulites of Rajapalayam, Southern India. Geol. Soc. Am. Bull. https://doi.org/10.1130/B35378.1 (in press). Liu, Y.S., Hu, Z.C., Gao, S., Günther, D., Xu, J., Gao, C.G., Chen, H.H., 2008. In situ analysis of major and trace elements of anhydrous minerals by LA-ICP-MS without applying an internal standard. Chem. Geol. 257, 34e43. Liu, Y.S., Gao, S., Hu, Z.C., Gao, C.G., Zong, K.Q., Wang, D.B., 2010. Continental and

oceanic crust recycling-induced melt-peridotite interactions in the Trans-North China Orogen: U-Pb dating, Hf isotopes and trace elements in zircons of mantle xenoliths. J. Petrol. 51, 537e571. €ner, A., Wan, Y., Santosh, M., Shaji, E., Dhanil Dev, S.G.D., 2016. Late Liu, S., Kro Palaeoproterozoic depositional age for khondalite protoliths in southern India and tectonic implications. Precambrian Res. 283, 50e67. Meert, J.G., Lieberman, B.S., 2008. The Neoproterozoic assembly of Gondwana and its relationship to the EdiacaraneCambrian radiation. Gondw. Res. 14 (1), 5e21. Morel, M.L.A., Nebel, O., Nebel-Jacobsen, Y.J., 2008. Hafnium isotope characterization of the GJ-1 zircon reference material by solution and laser-ablation MCICPMS. Chem. Geol. 255, 231e235. Morimoto, T., Santosh, M., Tsunogae, T., Yoshimura, Y., 2004. Spinel þ Quartz association from the Kerala khondalites, southern India: evidence for ultrahightemperature metamorphism. J. Mineral. Petrol. Sci. 99, 257e278. Nance, R.D., Murphy, J.B., Santosh, M., 2014. The supercontinent cycle: a retrospective essay. Gondw. Res. 25, 4e29. Palin, R.M., White, R.W., 2016. Emergence of blueschists on Earth linked to secular changes in oceanic crust composition. Nat. Geosci. 9, 60e64. Palin, R.M., Weller, O.M., Waters, D.J., Dyck, B., 2016. Quantifying geological uncertainty in metamorphic phase equilibria modelling; a Monte Carlo assessment and implications for tectonic interpretations. Geosci. Front. 7, 591e607. Powell, R., Holland, T.J.B., 2008. On thermobarometry. J. Metam. Geol. 26, 155e179. Raith, M., Karmakar, S., Brown, M., 1997. Ultra-high-temperature metamorphism and multistage decompressional evolution of sapphirine granulites from the Palni hill ranges, Southern India. J. Metam. Geol. 15, 379e399. Sajeev, K., Williams, I.S., Osanai, Y., 2010. Sensitive high-resolution ion microprobe U-Pb dating of prograde and retrograde ultrahigh-temperature metamorphism as exemplified by Sri Lankan granulites. Geology 38 (11), 971e974. Santosh, M., Yokoyama, K., Bijusekhar, S., Rogers, J.J.W., 2003. Multiple tectonothermal events in the granulite blocks of Southern India revealed from EPMA dating: implications on the history of supercontinents. Gondw. Res. 6 (1), 29e63. Santosh, M., Collins, A.S., Morimoto, T., Yokoyama, K., 2005. Depositional constraints and age of metamorphism in southern India: U-Pb chemical (EPMA) and isotopic (SIMS) ages from the Trivandrum Block. Geol. Mag. 142, 255e268. Santosh, M., Morimoto, T., Tsutsumi, Y., 2006. Geochronology of the khondalite belt of the Trivandrum Block, Southern India: electron probe ages and implications for Gondwana tectonics. Gondw. Res. 9, 261e278. Santosh, M., Maruyama, S., Sato, K., 2009. Anatomy of a Cambrian suture in Gondwana: implications for the timing of Gondwana assembly. Gondwana Res. 16, 321e341. Santosh, M., Hu, C.N., He, X.F., Li, S.S., Tsunogae, T., Shaji, E., Indu, G., 2017. Neoproterozoic arc magmatism in the southern Madurai Block, India: subduction, relamination, continental outbuilding, and the growth of Gondwana. Gondw. Res. 45, 1e42. Shazia, J., Harlov, D., Suzuki, K., Kim, S., Girish-Kumar, M., Hayasaka, Y., IshwarKumar, C., Windley, B.F., Sajeev, K., 2015. Linking monazite geochronology with fluid infiltration and metamorphic histories: nature and experiment. Lithos 236e237, 1e15. Shimizu, H., Tsunogae, T., Santosh, M., 2009. Spinel þ quartz assemblage in granulites from the Achankovil Shear Zone, southern India: implications for ultrahigh-temperature metamorphism. J. Asian Earth Sci. 36, 209e222. Tadokoro, H., Tsunogae, T., Santosh, M., 2008. Metamorphic P-T path of the eastern Trivandrum Granulite Block, southern India: implications for regional correlation of lower crustal fragments. J. Mineral. Petrol. Sci. 103, 279e284. Tateishi, K., Tsunogae, T., Santosh, M., Janardhan, A.S., 2004. First report of sapphirine þ quartz assemblage from southern India: Implications for ultrahigh-temperature metamorphism. Gondw. Res. 7, 899e912. Taylor, R.J.M., Clark, C., Fitzsimons, I.C.W., Santosh, M., Hand, M., Evans, N., McDonald, B., 2014. Post-peak, fluid-mediated modification of granulite facies zircon and monazite in the Trivandrum Block, southern India. Contrib. Mineral. Petrol. 168, 1044. https://doi.org/10.1007/s00410-014-1044-0. Tsunogae, T., Santosh, M., 2010. Ultrahigh-temperature metamorphism and decompression history of sapphirine granulites from Rajapalaiyam, southern India: implications for the formation of hot orogens during Gondwana assembly. Geol. Mag. 147 (01), 42. Tsunogae, T., Santosh, M., Ohyama, H., Sato, K., 2008. High-pressure and ultrahightemperature metamorphism at Komateri, northern Madurai Block, southern India. J. Asian Earth Sci. 33, 395e413. Vermeesch, P., 2018. Isoplot R: a free and open toolbox for geochronology. Geosci. Front. 9, 1479e1493. White, R.W., Powell, R., Holland, T.J.B., Johnson, T.E., Green, E.C.R., 2014. New mineral activityecomposition relations for thermodynamic calculations in metapelitic systems. J. Metam. Geol. 32, 261e286. White, R.W., Powell, R., Johnson, T.E., 2014. The effect of Mn on mineral stability in metapelites revisited: new aex relations for manganese-bearing minerals. J. Metam. Geol. 32, 809e828. Williams, I.S., 2001. Response of detrital zircon and monazite, and their UePb isotopic systems,to regional metamorphism and host-rock partial melting, Cooma Complex, southeastern Australia. Austr. J. Earth Sci. 48, 557e580. Yakymchuk, C., Brown, M., 2019. Divergent behaviour of Th and U during anatexis: implications for the thermal evolution of orogenic crust. J. Metam. Geol. 37, 899e916.