Journal of Earth Science  2018, Vol. 29 Issue (5): 1060-1073   PDF    
0
Phase Equilibrium Modeling of Retrograded Eclogite at the Kekesu Valley, Eastern Segment of SW Tianshan Orogen and Tectonic Implications
Bin Xia1,2, Qin Yang2, Neng-Song Chen1,3, Zhiqiang Zhou1,2    
1. School of Earth Sciences, China University of Geosciences, Wuhan 430074, China;
2. State Key Laboratory of Geological Processes and Mineral Resources, China University of Geosciences, Wuhan 430074, China;
3. Center for Global Tectonics, China University of Geosciences, Wuhan 430074, China
ABSTRACT: In this study, we report newly found strongly retrograded eclogite blocks in the Kekesu Valley, eastern segment of southwestern Tianshan, northwestern China. Based on detailed petrographical studies and mineral chemistry analyses, we constrain P-T evolution of two representative samples via phase equilibrium modeling method using THERMOCALC software. The peak stage is evidenced by the primary inclusions of omphacite+glaucophane+lawsonite (inferred)+chlorite+white mica in garnet. Using grossular and pyrope isopleth thermobarometry in garnet, the P peak conditions for this stage were constrained to be~23 kbar, 480-500℃ and the T peak conditions were constrained to be 19-20 kbar, 500-520℃. Subsequent exhumation led to the transformation of lawsonite to epidote. Using Si-in-phengite barometry, the P was constrained to be 16-20 kbar at T of 500-520℃ for this stage. Further retrograde stage was evidenced by the mineral assemblages of glaucophane+garnet+epidote+paragonite+ chlorite+white mica+quartz. Later stage retrogression is evidenced by glaucophane+epidote+chlorite+ quartz+white mica+albite+actinolite with P of 5-9 kbar when T < 490℃, indicating the eclogite may have been exhumed to the middle crust level and overprinted by blueschist-greenschist facies metamorphism. The results show that eclogite at the eastern segment of SW Tianshan shares similar P-T path to the HP rocks from the western segment, implying a similar P-T evolution for these rocks.
KEY WORDS: eclogite    P-T path    phase equilibria    southwestern Tianshan    Kekesu Valley    

0 INTRODUCTION

Eclogite and eclogite facies rocks developed in orogenic belt typically represent for paleo-convergent plate margins (Li et al., 2009; Liou et al., 2009; Ernst and Liou, 2008). According to their lithological associations, two types of tectonic background have been defined for their formation environment. One is oceanic type and the other is continental type (Song et al., 2006). The former has lithologies characterized for low temperature eclogite, blueschist, metasedimentary rocks, marble and serpentinized peridotite, mainly developed in subduction trenches (Zhang et al., 2007). On the other hand, the later has lithologies characterized for middle temperature eclogite and granitic gneiss, mostly developed along continental subduction/collision zones (Song et al., 2006).

Chinese southwestern Tianshan is an essential component of the central Asian orogenic belt (CAOB) and marks the final collision of the western CAOB (Zhang H C et al., 2017; Gao et al., 2011; Zhang L F et al., 2007). In SW Tianshan, rock associations are dominantly pelitic and felsic schists mixed eclogite, blueschist, serpentinite, marble and rodingite, belonging to an oceanic type eclogite-blueschist belt (Gao et al., 2009; Zhang et al., 2007). These rocks were interpreted to represent an accretionary mélange formed during the subduction of South Tianshan paleo-ocean under Yili Block (Zhang et al., 2007; Gao et al., 1999). Since the first confirmation of eclogite in this belt, numerous outstanding achievements have been made over the last twenty years (Lü et al., 2009, 2008; Zhang et al., 2005, 2003, 2002; Gao et al., 1999). However, for many years, studies have focused on the western segment of SW Tianshan where eclogite is well-developed and mineralogical evidences for UHP metamorphism have been identified (Li et al., 2016; Tian and Wei, 2014; Lü et al., 2013, 2012, 2009, 2008; Wei et al., 2009; Zhang et al., 2005, 2003, 2002; Gao and Klemd, 2003, 2001). For the eastern segment of SW Tianshan, rocks are mainly blueschist and pelitic and felsic schists and are considered to have undergone blueschist facies metamorphism (Gao et al., 1999, 1995). The distinct difference between the peak P-T conditions for rocks from the western and eastern segments was interpreted to be due to tectonic juxtaposition (e.g., Gao et al., 1999). In recent years, our group has found two pieces of eclogite blocks in the Kekesu Valley from the eastern segment of SW Tianshan Orogen (Xia et al., 2014). The P-T constraint for the eclogites indicates they may share similar metamorphic evolution to the HP eclogite in the western segment (Xia et al., 2014). However, the scale of eclogite facies metamorphism in the eastern segment is uncertain. In this study, we report new strongly retrograded eclogite blocks in one tributary of the Kekesu Valley. Based on detailed petrographical studies and mineral chemistry analyses, we constrain P-T evolution of two representative samples via phase equilibrium modeling method using THERMOCALC software and discuss their tectonic implications for the evolution of the SW Tianshan orogen. Mineral abbreviations in this study are as follows: g (garnet), o (omphacite), gl (glaucophane), law (lawsonite), c (chlorite), pa (paragonite), ep (epidote), q (quartz), hb (hornblende), act (actinolite), mu (muscovite), ph (phengite), ab (albite), cc (calcite), ilm (ilmenite), alm (almandine), sps (spessartine), pyr (pyrope) and grs (grossular).

1 GEOLOGICAL BACKGROUND

The SW Tianshan eclogite-blueschist belt is situated in western China and stretches for nearly 200 km along SWW- NEE strike (Fig. 1). Based on rock associations, this belt can be divided into the western segment and the eastern segment (Xia et al., 2014). Lithologies in the western segment comprise predominantly felsic and pelitic schists mixed with eclogite, serpentinite, marble, rodingite and rare blueschist which occur as blocks, lenses and/or intercalated layers (Zhang et al., 2007). As for lithologies in the eastern segment, they are mainly felsic and pelitic schists with interlayered blueschist and enclosed blueschist blocks and rare eclogite lenses (Xia et al., 2014). In some places in the western segment, based on phase equilibrium modeling and mineralogical evidences, UHP metamorphism was confirmed to eclogite, pelitic and felsic schists and marble (Lü et al., 2013, 2009, 2008; Wei et al., 2009). The western segment was then further subdivided into a north UHP slice and a south HP slice (Lü et al., 2013). In the UHP slice, coesite in combination with phase equilibrium modeling constrain peak P-T conditions to be 25–32 kbar, 470–630 ℃ (Lü et al., 2012, 2009, 2008; Wei et al., 2009). In northernmost of the UHP slice, recent research on the serpentinized wehrlite wall rock adjacent to eclogite generated P-T conditions of > 37±7 kbar, 520±10 ℃ (Shen et al., 2015). For rocks in the HP slice, peak P-T conditions constrained by phase equilibrium modeling were 21–25 kbar, 500–550 ℃ (Du et al., 2014, 2011; Wei et al., 2009). All those rocks follow an "Alpine type" P-T path, with a clockwise path whose temperature peak occurred after the pressure peak and isothermal decompression during its early exhumation (Du et al., 2011; Lü et al., 2009; Gao and Klemd, 2003). In the eastern segment, early studies on blueschist drew peak P-T condition of 9–10 kbar, ~450 ℃, corresponding to epidote-blueschist facies metamorphism (Gao et al., 1995). However, our recent studies on newly found eclogite using phase equilibrium modeling gave peak P-T conditions of 20–23 kbar, 480–515 ℃ (Xia et al., 2014).

Download:
Figure 1. Simplified geological map of the Chinese southwestern Tianshan. Sample locality has been marked.

In situ zircon U-Pb dating for eclogite and pelitic schist in the western segment gave ages of 321–319 Ma (Du et al., 2014, 2011; Yang et al., 2013; Su et al., 2010), comparable to in situ rutile U-Pb ages (318±7 Ma; Li et al., 2011) and garnet Lu-Hf isochron ages (ca. 315 Ma; Klemd et al., 2011), interpreted to represent for the eclogite facies metamorphism in the SW Tianshan. In the eastern segment, Ar-Ar dating for white micas from blueschist and pelitic schist with mineral assemblages of blueschist facies gave Ar-Ar plateau ages of mainly 320–310 Ma (Xia et al., 2016; Wang et al., 2010), while for epidote-mica schist with mineral assemblages of greenschist facies, Ar-Ar plateau ages for white micas are 293.3±1.5 Ma, reflecting greenschist facies overprinting during exhumation (Xia et al., 2016).

2 PETROLOGY 2.1 Analytical Method

For mineral composition analysis, a JEOL JXA-8100 electron microprobe was applied at College of Earth and Space Sciences, Peking University, Beijing. The operating conditions were 15 kV acceleration voltage, 10 nA beam current and 1 μm beam size. Matrix corrections were obtained using the PRZ correction program supplied by the manufacturer. Representative results are given in Tables 1, 2.

Table 1 Representative compositions (wt.%) of rock-forming minerals from eclogite QB-3 at Kekesu, SW Tianshan
Table 2 Representative compositions (wt.%) of rock-forming minerals from eclogite QB-11 at Kekesu, SW Tianshan

For BSE images, an FEI Quanta 200 scanning electron microscope (SEM) equipped with an EDAX energy dispersive spectrometer (EDS) system was applied at the State Key Laboratory of Geological Process and Mineral Resources, China University of Geosciences, Wuhan. The images were obtained at an accelerating voltage of 20 kV with a spotsize of 200–400 nm, an emission current of about 100 μA and a working distance of 11–12 mm.

2.2 Petrography

Two samples, QB-3 and QB-11, were selected for detailed petrographical studies and phase equilibrium modeling to qualify their P-T evolution process. Both samples are lense-structured and enclosed in (garnet-bearing) mica schist. The sample QB-3 comprises minerals of garnet (3%–8%)+glaucophane (25%–30%)+ epidote (30%–35%)+chlorite (20%–25%)+albite (10%–15%)+ phengite (3%–5%)+quartz (1%–3%)+omphacite (1%–3%) and accessory rutile, ilmenite, sphene, apatite and calcite. It shows a porphyroblastic texture (Fig. 2a) and the porphyroblasts are garnet and epidote. Porphyroblastic garnet is euhedral to subhedral, 0.4–1.0 mm in diameter and rich in inclusions. The inclusions are dominantly omphacite, with subordinate glaucophane, phengite, sphene, ilmenite and chlorite (Figs. 2b, 2c, 2d). At places, glaucophane has been replaced by chlorite and albite along rims (Fig. 2b), indicating the reaction g+gl→ab+chl may occur. Composite inclusions of epidote+paragonite showing a boxed shape may imply pseudomorphs after lawsonite due to the reaction g+law→ ep+pa (Fig. 2d; Lü et al., 2009). Some porphyroblastic garnet is fractured and the fractures are filled with chlorite (Fig. 2d). Garnet also develops as inclusions in large epidote. Those garnets are euhedral, smaller (0.1–0.3 mm in diameter) and have fewer inclusions (Figs. 2g, 2h). Omphacite is mainly present as tiny inclusions in porphyroblastic garnet, with less included in epidote. Relict omphacite in the matrix has mostly been retrograded to glaucophane and chlorite, but still preserves a pyroxene-type pseudomorph (Figs. 2e, 2f). Two types of glaucophane are present in the matrix. One is coarse-grained (> 0.5 mm in length) and constitutes the schistosity with omphacite pseudomorphs, phengite and some chlorite, while the other is fine-grained, randomly distributed and occurs around large glaucophane and omphacite (Fig. 2f). Fibrous actinolite is present around glaucophane (Fig. 2a). Porphyroblastic epidote is 1–2 mm in diameter. It contains inclusions of garnet, omphacite, glaucophane, phengite, chlorite, quartz, calcite and sphene (Figs. 2g, 2h). Albite in the matrix is anhedral. It has numerous inclusions of tiny glaucophane and less chlorite and sphene. Those inclusions are commonly randomly distributed. Phengite is fine-grained (0.1–0.2 mm in diameter). Paragonite only develops as inclusions in garnet or as relics in albite.

Download:
Figure 2. Photomicrographs showing the mineralogy and microstructures of eclogite QB-3 under plane polarized light (a), (e), (g), cross polarized light (h) and in BSE images (b), (c), (d), (f). (a) Prophyroblastic garnet and mineral inclusions; (b) mineral inclusions of omphacite, glaucophane, epidote and paragonite and phengite in garnet; (c)–(d) composite inclusions of epidote, paragonite and phengite in garnet, which may indicate the existence of lawsonite; (e)–(f) relics of omphacite, mostly replaced by glaucophane which again was replaced by actinolite; (g)–(h) euhedral tiny garnet included in porphyroblastic epidote. This garnet has no inclusions.

Sample QB-11 has a similar mineral assemblage of garnet (3%–8%)+glaucophane (20%–25%)+epidote (30%–35%)+chlorite (20%–25%)+albite (5%–10%)+phengite (3%–5%)+quartz (1%– 3%)+omphacite (3%–5%) and accessory rutile, ilmenite, sphene, apatite and calcite. Garnet porphyroblast is subhedral to anhedral and has been partially retrograded to chlorite along fractures or rims (Figs. 3a, 3b). Garnet included in epidote is euhedral, clean and commonly fractured (Fig. 3d). Coarse-grained epidote porphyroblast (> 3 mm in diameter) has numerous inclusions of tiny glaucophane, phengite, chlorite and sphene. These inclusions are parallel to the foliation defined by glaucophane, omphacite and phengite in the matrix (Fig. 3c). Similar to QB-3, omphacite in the matrix has been partially transformed to glaucophane and chlorite (Figs. 3e, 3f). Omphacite included in epidote also shows microstructures indicating replacement by glaucophane and chlorite (Fig. 3c).

Download:
Figure 3. Photomicrographs showing the mineralogy and microstructures of eclogite QB-11 under plane polarized light (a), (c), (e), cross polarized light (d) and in BSE images (b), (f). (a)–(b) Prophyroblastic garnet and inclusions of omphacite, glaucophane, epidote and paragonite. Cracks in garnet have been filled with secondary chlorite; (c) larger epidote porphyroblast (> 3 mm in diameter) has numerous inclusions of tiny glaucophane, phengite, chlorite and sphene. These inclusions are parallel to the schistosity defined by glaucophane and phengite; (d) garnet included in epidote, euhedral, clean and commonly fractured; (e)–(f) omphacite in the matrix has been partially transformed to glaucophane and chlorite.
2.3 Metamorphic Evolution

Based on petrographical observations, three main metamorphic stages are identified in QB-3 and QB-11. The prograde or peak pressure stage is evidenced by mineral inclusions of omphacite, glaucophane, phengite and chlorite in porphyroblastic garnet. Lawsonite was assumed to be present during prograde evolution according to its pseudomorphs composed of epidote and paragonite in garnet. The early retrograde stage is evidenced by the main rock-forming minerals of glaucophane, epidote, muscovite, chlorite and quartz. The relics of paragonite in albite may imply its development at this stage. Garnet rim may have been in equilibrium with these minerals. Late stage retrogression during exhumation is evidenced by fibrous glaucophane, albite, chlorite, quartz, sphene/ilmenite and actinolite. This mineral assemblage indicates the eclogite may have been exhumed to middle crust level and overprinted by blueschist-greenschist facies metamorphism.

2.4 Mineral Chemistry 2.4.1 Garnet

For sample QB-3, a traverse of point analyses on one half of a large porphyroblastic garnet from the apparent core to rim shows systematic variations of the spessartine and pyrope contents (Fig. 4). The sps content decreases distinctly from 5.3 mol% to 2.6 mol% and the pyr content increases from 6.3 mol% to 9.0 mol% (Tables 1, 2). The grs content roughly increases from 24.8 mol% to 27.8 mol% from core to rim and the alm content varying in the range of 60.3 mol% to 65.3 mol% shows no systematic variations from core to rim (Fig. 4). For the clean garnet included in porphyroblastic epidote, point analyses from core to rim show a decrease of sps content from 10.2 mol% to 1.6 mol% and an increase of the pyr content from 6.1 mol% to 7.7 mol%. The grs content varies from 23.2 mol% to 24.8 mol% and the alm content varies from 60.5 mol% to 66.3 mol% (Fig. 4). For sample QB-11, a traverse of point analyses on one large porphyroblastic garnet from core to rim shows a distinct decrease of sps (from 13.8 mol% to 1.3 mol%) content and an increase of pyr (from 5.8 mol% to 8.0 mol%), grs (from 21.9 mol% to 25.2 mol%) and alm (from 55.9 mol% to 65.6 mol%) contents (Fig. 4). For the clean garnet included in porphyroblastic epidote, from core to rim, the sps content decreases from 9.8 mol% to 1.8 mol% and the pyr, grs and alm contents increase from 6.5 mol% to 8.4 mol%, 23.4 mol% to 25.5 mol% and 60.3 mol% to 64.5 mol%, respectively (Fig. 4; Tables 1, 2).

Download:
Figure 4. EPMA for garnet porphyroblasts and garnet inclusions in epidote from samples QB-3 and QB-11.
2.4.2 Omphacite

For both samples, the primary omphacite, whether as inclusions in garnet or as a main rock-forming mineral, has indistinguished compositions (Tables 1, 2). For eclogite QB-3, omphacite inclusions in garnet have Jd of 0.27–0.37 and Ae of 0.17–0.21 and rock-forming omphacite has Jd of 0.36 and Ae of 0.14–0.20. For eclogite QB-11, omphacite as inclusions in garnet has Jd of 0.33–0.37 and Ae of 0.17–0.23 and rock-forming omphacite has Jd of 0.32–0.34 and Ae of 0.11–0.21.

2.4.3 Phengite

For QB-11, phengite included in garnet has Si contents of 3.34 p.f.u. (11 O basis) and phengite flakes out of garnet have similar Si contents of 3.33–3.37 p.f.u (Tables 1, 2). For QB-11, phengite included in garnet and outside garnet has Si contents of ~3.38 and 3.35–3.38 p.f.u., respectively (Tables 1, 2).

2.4.4 Amphibole

Sodic amphibole in QB-3 is glaucophane (Leake et al., 1997), with NaB of 1.84–1.91 p.f.u., Mg# (Mg/(Mg+Fe2+)) of 0.59–0.68 and (Na+K)A of 0.04–0.13, whereas calcic amphibole in QB-3 is actinolite, with CaB of 1.57 p.f.u., Mg# of 0.70 and (Na+K)A of 0.10 (Tables 1, 2). Sodic amphibole in QB-11 is glaucophane and has NaB of 1.89–1.90 p.f.u., Mg# of 0.65 and (Na+K)A of 0.19–0.20. Calcic amphibole in QB-11 is actinolite, with CaB of 1.73 p.f.u., Mg# of 0.77 and (Na+K)A of 0.14 (Tables 1, 2).

2.4.5 Other minerals

The epidote-group minerals in QB-3 and QB-11 have Ps=Fe3+/(Fe3++Al) values of 0–10.9 mol% and 12.5 mol%–17.6 mol%, respectively, belonging to clinozoisite (Tables 1, 2). Sodic feldspar in both samples is albite with Ab0.99–1.00. Paragonite in QB-3 as inclusions in garnet is close to end-member composition with Na content of 0.93 p.f.u. and K content of 0.04 p.f.u. (Tables 1, 2).

3 PHASE EQUILIBRIUM MODELING

To determine P-T conditions for the prograde/peak to retrograde metamorphism, phase diagrams was calculated for both retrograded eclogite QB-3 and QB-4. The phase equilibria were modeled using the THERMOCALC software (version 3.33, updated 19-10-2009) and the associated internally consistent thermodynamic dataset ds55 (Holland and Powell, 1998; updated in November 2003) for compositions in the MnNCKFMASHO (MnO-Na2O-CaO-K2O-Fe2O-MgO-Al2O3-SiO2-H2O-O) system. TiO2 is neglected due to its trivial amount, which is incorporated in rutile and ilmenite. Fe2O3 (O) has been added in the chemical system for it can affect the phase relations for metapelite (e.g., White et al., 2000). Activity-composition relationships used in the modeling are as follows: garnet (White et al., 2007); amphibole and clinopyroxene (Diener and Powell, 2012); paragonite and phengitic muscovite (Coggon and Holland, 2002); plagioclase (Holland and Powell, 2003); biotite (combined Mahar et al., 1997 with Powell and Holland, 1999); and, chlorite and epidote (Holland and Powell, 1998). Quartz, lawsonite and aqueous fluid (H2O) are treated as pure end-member phases. H2O was assumed to be present in excess. All primary sample compositions (in wt.%) together with the modified compositions for phase equilibrium modeling (in mol%) are given in Table 3.

Table 3 Primary sample compositions (in wt.%) together with modified bulk rock compositions used for phase diagram calculation (in mol%) for eclogites at the Kekesu Valley. FeOT represents total iron. O (in mol%) is equal to Fe2O3 (in mol%)

For modeling the late prograde, peak and early retrograde P-T evolution, the bulk rock compositions obtained by experimental method (wet chemical method) have been applied for both samples QB-3 and QB-11. The CaO, FeO or MgO contents were corrected for the P2O3 contained in apatite, TiO2 contained in ilmenite and sphene and CO2 contained in carbonate minerals (Fig. 3). For the late retrograde evolution, effective bulk rock composition was calculated due to the element fractionation caused by the residual of garnet and omphacite for sample QB-3 (Fig. 3). In QB-3, considering the garnet porphyroblast and garnet inclusions in epidote are euhedral to subhedral, we assume all garnet was not involved in the domainal equilibration volume during retrogression. On the other hand, the development of omphacite relics implies its major proportion (about 90% in volume proportion) has been retrograded to sodic amphibole and/or chlorite. The procedure for recalculating the effective bulk rock composition follows the tutorial by Richard White in THERMOCALC website via the mode and composition of phases revealed in the rbi (read bulk information) code in THERMOCALC (29/03/2010).

Pseudosections for the P-T range of 10–25 kbar, 450–600 ℃ were calculated for both samples QB-3 and QB-11 with excess H2O (Figs. 5, 6). In the pseudosections, phengite and H2O are present across the P-T range modeled. Because of similar bulk rock compositions, both pseudosections have similar phase relations (Figs. 5, 6). The modeling shows that the omphacite- and lawsonite-present phase assemblages develop in the top left field, while the epidote-present phase assemblages survive in most of the bottom right field. The narrow fields where both lawsonite and epidote are present represent the transformation from omphacite and lawsonite to glaucophane, epidote and paragonite via the reaction of law+o (+c)→g+gl+ep+q+H2O when T increases or P decreases (Figs. 5a, 6a). Chlorite develops in most of the modeled fields, probably due to high Mg/Si ratio of the bulk rock composition, while omphacite survives in limited field may be related to low Na, high Ca values of the bulk rock composition (Xia et al., 2014; Tian and Wei, 2014). At higher T, low P, glaucophane and chlorite transform to hornblende via the reaction of gl+c+ep+ mu→g+hb+pa+q+H2O.

Download:
Figure 5. P-T pseudosection for eclogite QB-3. The bulk rock composition is given in Table 3. "cg" and "mg" represents calculated isopleths for grossular and pyrope contents in garnet, respectively, "Si" represents calculated isopleths for Si in phengite and "h" represents calculated isopleths for H2O mode contained in the phase assemblages.
Download:
Figure 6. P-T pseudosection for eclogite QB-11. The bulk rock composition is given in Table 3. "cg" and "mg" represents calculated isopleths for grossular and pyrope contents in garnet, respectively, "Si" represents calculated isopleths for Si in phengite and "h" represents calculated isopleths for H2O mode contained in the phase assemblages.

Compositional isopleths of grossular (cg) and pyrope (mg) in garnet (Figs. 5b, 6b) and Si in phengite (Figs. 5c, 6c) were modeled. In lawsonite- and omphacite-present phase assemblage fields, grossular isopleth in garnet is almost parallel to the T axis and decreases with P (Figs. 5b, 6b). Thus it can be applied as compositional barometry for low T eclogite with mineral assemblages of g+o+law+gl+ph (+c, quartz, ep). In the omphacite-absent phase assemblage field (g+gl+law+ep+c+ ph+H2O±q), although grossular isopleth is still subparallel to the T axis, it increases at higher P. In the phase assemblage field of g+gl+ep+c+pa+q+mu+H2O, grossular isopleth has medium negative slope and decreases when P or T increases. On the other hand, in the phase assemblage field of g+ep+ pa+q+mu+H2O, grossular isopleth has medium positive slope and increases when P decreases or T increases. For pyrope isopleth in garnet, in the modeled fields with phase assemblages of g+o+gl+law+ph+H2O and g+ep+pa+q+hb+mu+H2O at high T, it is parallel to the T axis and increases with P (Figs. 5b, 6b). In other fields across the P-T range modeled, pyrope isopleth shows similar variations with P and T. It is subparallel to the P axis and increases with T. In general, Si isopleth in phengite increases with P (Figs. 5c, 6c). However, in the lawsonite present fields, Si isopleth in phengite is sensitive to P while in the lawsonite absent field, it is insensitive to P. In conclusion, for low temperature eclogite with mineral assemblages of g+o+law+gl±c±q, compositional isopleths of grossular in garnet and Si in phengite can be applied as barometry and isopleth of pyrope in garnet can be applied as thermometry.

To evaluate fluid behavior during P-T evolution of the retrograded eclogite, mode isopleths of H2O bounded in phase assemblages with P and T were modeled and presented in Figs. 5d and 6d. The value of H2O in the bulk rock composition was chosen to be just sufficient to saturate all the modeled phase assemblages across the P-T range. The highest H2O content corresponds to the H2O contained in the phase assemblage of g+o+gl+law+c+ph at P=25 kbar and T=450 ℃. Let the H2O content in this phase assemblage to be 100 mol% at this P-T conditions. The modeling shows that isopleths of H2O content in the phase assemblages of g+o+gl+law+c+ph and g+gl+ep+ c+pa+q+ph have steep to vertical slope and decrease when T decreases, consistent with dehydration reactions (Figs. 5d, 6d). Therefore, fluid would be lost from the rock when a mineral assemblage starts to dehydrate along the prograde evolution. However, in the narrow fields where both lawsonite and epidote are present, the H2O isopleths have gentle positive slope and decrease dramatically when P decreases, consistent with the quick transformation of more hydrous mineral lawsonite (contain ~12 wt.% H2O as hydroxyl) to less hydrous mineral epidote (contain ~2 wt.% H2O as hydroxyl).

To determine the P-T regime for the late retrograde evolution, phase equilibrium modeling was calculated for QB-3 using the recalculated bulk rock composition in Table 3. The pseudosection was calculated for the P-T range of 5–15 kbar, 400–500 ℃, with H2O in excess (Fig. 6). The modeling shows that garnet is absent across the calculated P-T range, consistent with our petrographic observations. In the narrow fields where both lawsonite and epidote are present, omphacite and lawsonite react to form epidote, paragonite and glaucophane when P decreases. Albite and actinolite develop at P of 7–9 kbar at T=400–500 ℃ via the reaction gl+pa+ep+mu+H2O→act+ab+c+q. At higher T, hornblende develops at the consumption of actinolite and chlorite.

4 DISCUSSIONS 4.1 P-T Path for the Retrograded Eclogite at Kekesu Valley

An advantage of pseudosection method is that it uses all the available information from the samples studied, e.g. the bulk rock composition, the mineral assemblages, mineral compositions and proportions, to determine P-T conditions for different stages of the metamorphic evolution (Powell and Holland, 2008). Therefore, by combining petrographic observations and the calculated P-T pseudosections and compositional isopleths from Figs. 5 and 6, we may reconstruct a portion of P-T path for the retrograded eclogite at the Kekesu Valley.

The primary inclusions of o+gl+law (inferred)+c+mu in garnet from QB-3 and QB-11 indicate that the P-T path passed through the phase assemblage of g+o+gl+law+c+mu+H2O (Figs. 5, 6). Although lawsonite has not been observed in our samples, pseudomorphs of ep+pa after lawsonite have developed as inclusions in garnet. Such pseudomorphs in garnet have been found in many rocks from the low temperature eclogite terranes and have been interpreted to represent the existence of lawsonite. In SW Tianshan, lawsonite and pseudomorphs of ep+pa after lawsonite have been reported in the western segment, indicating the HP/UHP rocks in SW Tianshan may have been passed through the lawsonite eclogite facies field (Du et al., 2014, 2011). In addition, compositional isopleths of grossular and pyrope contents in garnet and Si content in phengite from our samples plot in the P-T pseudosection (Figs. 5, 6). The results show that both garnet and phengite were stable in lawsonite-present fields. Therefore, it is reasonable to assume lawsonite was contained in the prograde or peak mineral assemblages.

As mentioned above, the compositions of garnet and phengite can be used as thermobarometry to constrain P-T conditions. The distinct decrease of sps content in garnet from core to rim implies garnet was progressively formed and its composition may have not been equilibrated during exhumation. For sample QB-3, both porphyroblastic garnet and garnet as inclusions in epidote have a compositional zoning grs roughly decreasing and pyr roughly increasing from core to rim, indicating a P decrease and T increase process during evolution. The maximum P constrained by the garnet core composition with minimum grs content that is included in epidote is about 23 kbar and the maximum T constrained by the porphyroblastic garnet rim composition with maximum pyr content is about 520 ℃ (Fig. 5b), similar to that constrained by grs and pyr contents in garnet from sample QB-11. Therefore, the portion of P-T path in the g+o+gl+law+c+mu+H2O assemblage field could be from ~23 kbar, 480–500 ℃ to 19–20 kbar, 500–520 ℃ in both samples, indicating a thermal relaxation during initial exhumation, similar to that of the lawsonite-bearing eclogite from western segment of SW Tianshan (Du et al., 2014).

Subsequent exhumation may lead to the P-T path going tangential to the H2O contours, as at the "T" peak, free fluid in the mineral assemblage becomes absent, so dehydration reaction ceases and retrogression begins (Guiraud et al., 2001). Under this assumption, P constrained by the Si content of 3.33–3.37 p.f.u. in phengite from QB-3 is 16–19 kbar and from QB-11 is 17–20 kbar at T of 500–520 ℃ (Figs. 5b, 6b), which represents the P-T field passed by the P-T path during decompression.

During further decompression, when the P-T path crossed the epidote- and lawsonite-present fields, the transition from lawsonite to epidote released a large amount fluid. According to our modeling, nearly half of the H2O content in the phase assemblage of g+o+gl+law+c+mu would be lost during this transition (Figs. 5d, 6d). The fluid along grain boundaries would promote rehydration of the phase assemblage of g+gl+ep+pa+c+mu+q and the formation of glaucophane around relict omphacite and large porphyroblastic epidote as shown in Fig. 2.

Further decompression and local fluid migration along grain boundaries led to the development of actinolite and albite at the consumption of paragonite and glaucophane. The final stage of the retrograde evolution is represented by the development of anhedral albite and tiny prismatic actinolite in the matrix, corresponding to the phase assemblage of gl+ep+c+ q+mu+ab+act in the P-T pseudosection with P-T conditions of 5–9 kbar when T < 490 ℃ (Fig. 6). This retrogression was occurred at amphibolite facies conditions in middle crust. In summary, this P-T evolution is similar to may HP/UHP terranes in the world (e.g., Liou et al., 2009).

Along P-T evolution, variations of mode proportions (one-oxide based mole proportions, similar to volume proportions) for different phases have been calculated (Fig. 7). The result shows that two major reactions have been developed along the P-T evolution. The first reaction is evidenced by the variations of the mode proportions of the main rock forming minerals lawsonite and omphacite. Both decrease distinctly at early decompression stage along isothermal decompression process (from 20.0 to 16.3 kbar at 510–520 ℃). Correspondingly, epidote develops and together with glaucophane, their mode proportions increase distinctly via the reaction law+o+c→g+ gl+ep+q+pa+H2O. A large volume of H2O is released during this process. This process demonstrates the transition of the rock from lawsonite eclogite facies metamorphism to epidote blueschist/eclogite facies metamorphism. The second reaction is evidenced by the decrease of the mode proportions of glaucophane, paragonite, epidote and H2O and the increase of the mode proportions of albite, chlorite and actinolite. This rehydration reaction could be gl+pa+ep+H2O→act+ab+c. During this process, the rock was further overprinted by greenschist facies metamorphism at later retrogression.

Download:
Figure 7. P-T pseudosection for eclogite QB-3. The bulk rock composition is given in Table 3.
Download:
Figure 8. Calculated mode variations for different phases along the inferred P-T path.

Our studies show that, though strongly retrograde, these eclogite blocks have been undergone lawsonite eclogite facies metamorphism. The country rock, which is dominantly garnet-free mica schist or albite schist, only shows mineral assemblages of blueschist to greenschist facies metamorphism. Up till now, based on mineral assemblages and conventional thermobarometer, it is difficult to determine P-T conditions for these rocks for their mineral assemblage could be stable at greenschist, blueschist, even eclogite facies P-T fields (Du et al., 2011; Wei et al., 2009). Phase equilibrium modeling, combined with detailed petrographic observations may provide information on P-T evolution for these kinds of rocks, as shown in this study and works by other researchers (e.g., Tian and Wei, 2014; Du et al., 2011).

4.2 Tectonic Implications

Since no eclogite has been found in the eastern segment of the SW Tianshan LT-HP/UHP belt, former researchers claimed for a tectonic overlay of the eastern and western segments of SW Tianshan based on different P-T evolutions of rocks from the eastern and western segments. Our group in recent years has identified several eclogite blocks in the Kekesu Valley and its tributaries. Combined with phase equilibrium modeling, P-T path for these eclogites has been reconstructed. The results show that eclogites at the Kekesu Valley share similar P-T path to the HP rocks from the southern part of the western segment, implying a similar P-T evolution for these rocks.

Exhumation mechanisms for eclogite and associated lithologies in SW Tianshan are controversial. The key point lies on different recognitions for the relationships between HP and UHP rocks in this belt. As mentioned above, for the eclogite facies metamorphic rocks in the western segment and the epidote-blueschist to greenschist facies metamorphic rocks in the eastern segment, a tectonic overlay relationship has been previously proposed based on their distinctly different metamorphic evolution (Gao et al., 1999, 1995). In recent years, coesite inclusions in a variety of lithologies in the western segment has been observed (e.g., Zhang et al., 2013; Lü et al., 2012, 2009, 2008). Mapping for UHP rocks based on coesite, phase assemblages, as well as pseudosection modeling, constrains a coherent UHP terrane in the north and a coherent HP terrane in the south (Shen et al., 2015; Lü et al., 2012, 2009; Zhang et al., 2005, 2002). Their spatial relationship has been interpreted to represent for the northward subduction polarity of the paleo-oceanic crust. Thus in western segment, the exhumation of the UHP and HP units has been believed to be coherent in the subduction channel. Alternatively, Klemd et al. (2011) and Li et al. (2016) claimed for an intimate interlayering of HP and UHP eclogites with metasediments in the western segment. They interpreted the intermingling of HP and UHP rocks to be as a consequence of juxtaposition during subduction and/or exhumation within the subduction channel. To the eastern segment of SW Tianshan, rare occurrences of eclogite or eclogite facies rocks have been reported (this study and Xia et al., 2014). However, unlike the western segment where UHP rocks and eclogite are common, in the eastern segment, eclogite and eclogite facies rocks are scarce while blueschist or blueschist facies rocks are common. Therefore, the spatial distribution of HP and UHP rocks in SW Tianshan shows that UHP rock can only be found in the northern part of the western segment of SW Tianshan, while to the south and east, only rocks undergone HP eclogite facies metamorphism have been reported (Xia et al., 2014; Du et al., 2011). Further to the east in the Kekesu Valley, blueschist is commonly developed while eclogite blocks are extremely rare. We suggest a decrease of maximum P attained by rocks from the west to the east in SW Tianshan, which is in consistence with the hypothesis of a southeast to northwest polarity of subduction (Xia et al., 2014; Lü et al., 2012, 2009). In the subduction zone, UHP rocks from deep trench exhumed near the continental side and HP rocks from shallow depth exhumed near the oceanic side. The exhumation of rocks from different depths could be occurred during the collisions between the Tarim Craton to the south and the Yili and Central Tianshan blocks to the north. Whether the UHP rocks developed as imbricated slices to the north of a shallower one (Lü et al., 2009), or the UHP rocks dispersed in HP rocks and were mingled together during exhumation (Li et al., 2016) is unknown. Future studies on detailed field mapping and peak P-T constraints using consistent method like pseudosection modeling might shed light on this issue.

5 CONCLUSION

Four stages have been identified for the newly found strongly retrograded eclogite at the Kekesu Valley in the eastern segment of the SW Tianshan Orogen. The peak stage is evidenced by the primary inclusions of o+gl+law (inferred)+ c+mu in garnet and P-T conditions for this stage was constrained to be from ~23 kbar, 480–500 ℃ to 19–20 kbar, 500–520 ℃. Subsequent exhumation was evidenced by the mineral assemblages of o+gl+law (inferred)+c+mu+g+ep and Si-in-phengite barometry constrains this stage with P of 16–20 kbar at T of 500–520 ℃. Further retrograde stage transferred lawsonite to epidote and omphacite to paragonite. Later stage retrogression is evidenced by the mineral assemblages of gl+ep+c+q+mu+ab+act with P of 5–9 kbar when T < 490 ℃. Our stages show that the retrograded eclogites at eastern segment of SW Tianshan share similar P-T path to rocks from the HP slice in western segment of SW Tianshan, implying a similar P-T evolution for these rocks.

ACKNOWLEDGMENTS

This paper is dedicated to Prof. Zhendong You on the oc casion of his 90th birthday. We appreciate the reviewer Dr. Zuolin Tian and an anonymous reviewer for their careful re views and constructive comments which helps a lot to improve the manuscript. We also thank the editors for their handling of the manuscript. This study was supported by the National Natural Science Foundation of China (No. 41502043) and China Postdoctoral Science Foundation (No. 2016T90742). The final publication is available at Springer via https://doi.org/10.1007/s12583-018-0844-y.


REFERENCES CITED
Coggon, R., Holland, T. J. B., 2002. Mixing Properties of Phengitic Micas and Revised Garnet-Phengite Thermobarometers. Journal of Metamorphic Geology, 20(7): 683-696. DOI:10.1046/j.1525-1314.2002.00395.x
Diener, J. F. A., Powell, R., 2012. Revised Activity-Composition Models for Clinopyroxene and Amphibole. Journal of Metamorphic Geology, 30(2): 131-142. DOI:10.1111/j.1525-1314.2011.00959.x
Du, J. X., Zhang, L. F., Lü, Z., et al., 2011. Lawsonite-Bearing Chloritoid-Glaucophane Schist from SW Tianshan, China:Phase Equilibria and P-T Path. Journal of Asian Earth Sciences, 42(4): 684-693. DOI:10.1016/j.jseaes.2011.04.003
Du, J. X., Zhang, L. F., Shen, X. J., et al., 2014. A New P-T-t Path of Eclogites from Chinese Southwestern Tianshan:Constraints from P-T Pseudosections and Sm-Nd Isochron Dating. Lithos, 200/201: 258-272. DOI:10.1016/j.lithos.2014.04.009
Ernst, W. G., Liou, J. G., 2008. High-and Ultrahigh-Pressure Metamorphism:Past Results and Future Prospects. American Mineralogist, 93(11/12): 1771-1786. DOI:10.2138/am.2008.2940
Gao, J., He, G. Q., Li, M. S., et al., 1995. The Mineralogy, Petrology, Metamorphic PTDt Trajectory and Exhumation Mechanism of Blueschists, South Tianshan, Northwestern China. Tectonophysics, 250(1/2/3): 151-168. DOI:10.1016/0040-1951(95)00026-6
Gao, J., Klemd, R., 2001. Primary Fluids Entrapped at Blueschist to Eclogite Transition:Evidence from the Tianshan Meta-Subduction Complex in Northwestern China. Contributions to Mineralogy and Petrology, 142(1): 1-14. DOI:10.1007/s004100100275
Gao, J., Klemd, R., 2003. Formation of HP-LT Rocks and Their Tectonic Implications in the Western Tianshan Orogen, NW China:Geochemical and Age Constraints. Lithos, 66(1/2): 1-22.
Gao, J., Klemd, R., Qian, Q., et al., 2011. The Collision between the Yili and Tarim Blocks of the Southwestern Altaids:Geochemical and Age Constraints of a Leucogranite Dike Crosscutting the HP-LT Metamorphic Belt in the Chinese Tianshan Orogen. Tectonophysics, 499(1/2/3/4): 118-131. DOI:10.1016/j.tecto.2011.01.001
Gao, J., Klemd, R., Zhang, L. F., et al., 1999. P-T Path of High-Pressure/Low-Temperature Rocks and Tectonic Implications in the Western Tianshan Mountains, NW China. Journal of Metamorphic Geology, 17(6): 621-636. DOI:10.1046/j.1525-1314.1999.00219.x
Gao, J., Long, L. L., Klemd, R., et al., 2009. Tectonic Evolution of the South Tianshan Orogen and Adjacent Regions, NW China:Geochemical and Age Constraints of Granitoid Rocks. International Journal of Earth Sciences, 98(6): 1221-1238. DOI:10.1007/s00531-008-0370-8
Guiraud, M., Powell, R., Rebay, G., 2001. H2O in Metamorphism and Unexpected Behaviour in the Preservation of Metamorphic Mineral Assemblages. Journal of Metamorphic Geology, 19(4): 445-454. DOI:10.1046/j.0263-4929.2001.00320.x
Holland, T. J. B., Powell, R., 1998. An Internally Consistent Thermodynamic Data Set for Phases of Petrological Interest. Journal of Metamorphic Geology, 16(3): 309-343. DOI:10.1111/j.1525-1314.1998.00140.x
Holland, T. J. B., Powell, R., 2003. Activity-Composition Relations for Phases in Petrological Calculations:An Asymmetric Multicomponent Formulation. Contributions to Mineralogy and Petrology, 145(4): 492-501. DOI:10.1007/s00410-003-0464-z
Klemd, R., John, T., Scherer, E. E., et al., 2011. Changes in Dip of Subducted Slabs at Depth:Petrological and Geochronological Evidence from HP-UHP Rocks (Tianshan, NW-China). Earth and Planetary Science Letters, 310(1/2): 9-20. DOI:10.1016/j.epsl.2011.07.022
Leake, B. E., Woolley, A. R., Arps, C. E. S., et al., 1997. Nomenclature of Amphiboles:Report of the Subcommittee on Amphiboles of the International Mineralogical Association, Commission on New Minerals and Mineral Names. The Canadian Mineralogist, 35: 219-246.
Li, C., Zhai, Q. G., Dong, Y. S., et al., 2009. High-Pressure Eclogite-Blueschist Metamorphic Belt and Closure of Paleo-Tethys Ocean in Central Qiangtang, Qinghai-Tibet Plateau. Journal of Earth Science, 20(2): 209-218. DOI:10.1007/s12583-009-0021-4
Li, J. L., Klemd, R., Gao, J., et al., 2016. Poly-Cyclic Metamorphic Evolution of Eclogite:Evidence for Multistage Burial-Exhumation Cycling in a Subduction Channel. Journal of Petrology, 57(1): 119-146. DOI:10.1093/petrology/egw002
Li, Q. L., Lin, W., Su, W., et al., 2011. SIMS U-Pb Rutile Age of Low-Temperature Eclogites from Southwestern Chinese Tianshan, NW China. Lithos, 122(1/2): 76-86. DOI:10.1016/j.lithos.2010.11.007
Liou, J. G., Ernst, W. G., Zhang, R. Y., et al., 2009. Ultrahigh-Pressure Minerals and Metamorphic Terranes-The View from China. Journal of Asian Earth Sciences, 35(3/4): 199-231. DOI:10.1016/j.jseaes.2008.10.012
Lü, Z., Bucher, K., Zhang, L. F., 2013. Omphacite-Bearing Calcite Marble and Associated Coesite-Bearing Pelitic Schist from the Meta-Ophiolitic Belt of Chinese Western Tianshan. Journal of Asian Earth Sciences, 76: 37-47. DOI:10.1016/j.jseaes.2013.07.034
Lü, Z., Bucher, K., Zhang, L. F., et al., 2012. The Habutengsu Metapelites and Metagreywackes in Western Tianshan, China:Metamorphic Evolution and Tectonic Implications. Journal of Metamorphic Geology, 30(9): 907-926. DOI:10.1111/j.1525-1314.2012.01002.x
Lü, Z., Zhang, L. F., Du, J. X., et al., 2008. Coesite Inclusions in Garnet from Eclogitic Rocks in Western Tianshan, Northwest China:Convincing Proof of UHP Metamorphism. American Mineralogist, 93(11/12): 1845-1850. DOI:10.2138/am.2008.2800
Lü, Z., Zhang, L. F., Du, J. X., et al., 2009. Petrology of Coesite-Bearing Eclogite from Habutengsu Valley, Western Tianshan, NW China and Its Tectonometamorphic Implication. Journal of Metamorphic Geology, 27(9): 773-787. DOI:10.1111/j.1525-1314.2009.00845.x
Mahar, E. M., Baker, J. M., Powell, R., et al., 1997. The Effect of Mn on Mineral Stability in Metapelites. Journal of Metamorphic Geology, 15(2): 223-238. DOI:10.1111/j.1525-1314.1997.00011.x
Powell, R., Holland, T. J. B., 1999. Relating Formulations of the Thermodynamics of Mineral Solid Solutions:Activity Modeling of Pyroxenes, Amphiboles, and Micas. American Mineralogist, 84(1/2): 1-14.
Powell, R., Holland, T. J. B., 2008. On Thermobarometry. Journal of Metamorphic Geology, 26(2): 155-179. DOI:10.1111/jmg.2008.26.issue-2
Shen, T. T., Hermann, J., Zhang, L. F., et al., 2015. UHP Metamorphism Documented in Ti-Chondrodite-and Ti-Clinohumite-Bearing Serpentinized Ultramafic Rocks from Chinese Southwestern Tianshan. Journal of Petrology, 56(7): 1425-1458. DOI:10.1093/petrology/egv042
Song, S. G., Zhang, L. F., Niu, Y. L., et al., 2006. Evolution from Oceanic Subduction to Continental Collision:A Case Study from the Northern Tibetan Plateau Based on Geochemical and Geochronological Data. Journal of Petrology, 47(3): 435-455. DOI:10.1093/petrology/egi080
Su, W., Gao, J., Klemd, R., et al., 2010. U-Pb Zircon Geochronology of Tianshan Eclogites in NW China:Implication for the Collision between the Yili and Tarim Blocks of the Southwestern Altaids. European Journal of Mineralogy, 22(4): 473-478. DOI:10.1127/0935-1221/2010/0022-2040
Tian, Z. L., Wei, C. J., 2014. Coexistence of Garnet Blueschist and Eclogite in South Tianshan, NW China:Dependence of P-T Evolution and Bulk-Rock Composition. Journal of Metamorphic Geology, 32(7): 743-764. DOI:10.1111/jmg.12089
Wang, B., Faure, M., Shu, L. S., et al., 2010. Structural and Geochronological Study of High-Pressure Metamorphic Rocks in the Kekesu Section (Northwestern China):Implications for the Late Paleozoic Tectonics of the Southern Tianshan. The Journal of Geology, 118(1): 59-77. DOI:10.1086/648531
Wei, C. J., Wang, W., Clarke, G. L., et al., 2009. Metamorphism of High/Ultrahigh-Pressure Pelitic-Felsic Schist in the South Tianshan Orogen, NW China:Phase Equilibria and P-T Path. Journal of Petrology, 50(10): 1973-1991. DOI:10.1093/petrology/egp064
White, R. W., Powell, R., Holland, T. J. B., et al., 2000. The Effect of TiO2 and Fe2O3 on Metapelitic Assemblages at Greenschist and Amphibolite Facies Conditions:Mineral Equilibria Calculations in the System K2O-FeO-MgO-Al2O3-SiO2-H2O-TiO2-Fe2O3. Journal of Metamorphic Geology, 18(5): 497-511. DOI:10.1046/j.1525-1314.2000.00269.x
White, R. W., Powell, R., Holland, T. J. B., 2007. Progress Relating to Calculation of Partial Melting Equilibria for Metapelites. Journal of Metamorphic Geology, 25(5): 511-527. DOI:10.1111/j.1525-1314.2007.00711.x
Xia, B., Zhang, L. F., Bader, T., et al., 2016. Late Palaeozoic 40Ar/39Ar Ages of the HP-LT Metamorphic Rocks from the Kekesu Valley, Chinese Southwestern Tianshan:New Constraints on Exhumation Tectonics. International Geology Review, 58(4): 389-404. DOI:10.1080/00206814.2015.1084487
Xia, B., Zhang, L. F., Xia, Y., 2014. Petrology and Phase Equilibrium of Newly Found Eclogites from Kekesu Valley in Eastern Part of Southwest Tianshan HP-UHP Metamorphic Belt, China, and Its Tectonic Significance. Science China Earth Sciences, 57(1): 117-131. DOI:10.1007/s11430-013-4663-9
Yang, X., Zhang, L. F., Tian, Z. L., et al., 2013. Petrology and U-Pb Zircon Dating of Coesite-Bearing Metapelite from the Kebuerte Valley, Western Tianshan, China. Journal of Asian Earth Sciences, 70/71: 295-307. DOI:10.1016/j.jseaes.2013.03.020
Zhang, H. C., Zhu, Y. F., Feng, W. Y., et al., 2017. Paleozoic Intrusive Rocks in the Nalati Mountain Range (NMR), Southwest Tianshan:Geodynamic Evolution Based on Petrology and Geochemical Studies. Journal of Earth Science, 28(2): 196-217. DOI:10.1007/s12583-016-0922-1
Zhang, L. F., Ai, Y. L., Li, X. P., et al., 2007. Triassic Collision of Western Tianshan Orogenic Belt, China:Evidence from SHRIMP U-Pb Dating of Zircon from HP/UHP Eclogitic Rocks. Lithos, 96(1/2): 266-280. DOI:10.1016/j.lithos.2006.09.012
Zhang, L. F., Du, J. X., Lü, Z., et al., 2013. A Huge Oceanic-Type UHP Metamorphic Belt in Southwestern Tianshan, China:Peak Metamorphic Age and P-T Path. Chinese Science Bulletin, 58(35): 4378-4383. DOI:10.1007/s11434-013-6074-x
Zhang, L. F., Ellis, D. J., Arculus, R. J., et al., 2003. "Forbidden Zone" Subduction of Sediments to 150 km Depth-The Reaction of Dolomite to Magnesite+Aragonite in the UHPM Metapelites from Western Tianshan, China. Journal of Metamorphic Geology, 21(6): 523-529. DOI:10.1046/j.1525-1314.2003.00460.x
Zhang, L. F., Ellis, D. J., Jiang, W. B., 2002. Ultrahigh-Pressure Metamorphism in Western Tianshan, China:Part Ⅰ. Evidence from Inclusions of Coesite Pseudomorphs in Garnet and from Quartz Exsolution Lamellae in Omphacite in Eclogites. American Mineralogist, 87(7): 853-860. DOI:10.2138/am-2002-0707
Zhang, L. F., Song, S. G., Liou, J. G., et al., 2005. Relict Coesite Exsolution in Omphacite from Western Tianshan Eclogites, China. American Mineralogist, 90(1): 181-186. DOI:10.2138/am.2005.1587