旃噶陔蔭陝嫌怍刓懦儒坒曹傖綻翐坒

楷票衾ㄩ2021-09-13 08:30:04

Transition of metamorphic series from the Kyaniteto andalusite-types in the Altai orogen, Xinjiang, China: Evidence from petrography and calculated KMnFMASH and KFMASH phase relations
Chunjing Weia,
a ,

, Geoffrey Clarkeb,

, Wei Tiana and Lin Qiua

MOE Key Laboratory of OBCE, School of Earth and Space Sciences, Peking

University, Beijing 100871, China
b

School of Geosciences, University of Sydney, NSW 2006, Australia

Received 8 January 2006; accepted 7 November 2006. Available online 8 January 2007.

Abstract
Metamorphic zones in the Chinese Altai orogen have previously been separated into the kyanite- and andalusite-types, the andalusite-type being spatially more extensive. The kyanite-type involves a zonal sequence of biotite, garnet, staurolite, kyanite, sillimanite and, locally, garnet每cordierite zones. The andalusite-type zonal sequence is similar: it includes biotite, garnet and staurolite zones at lower-T conditions and sillimanite and garnet每cordierite zones at higher-T conditions, but additionally contains staurolite每andalusite and andalusite每sillimanite zones at intermediate-T conditions. As relic kyanite-bearing assemblages commonly persist in the staurolite每andalusite, andalusite每sillimanite and sillimanite zones, it is not clear that the distinction is valid. On the basis of a reevaluation of phase relations modelled in KMnFMASH and KFMASH pseudosections, kyanite and andalusite-bearing rocks of

the Chinese Altai orogen record, respectively, the typical burial and exhumation history of the terrane. Mineral assemblages distributed through the various zones reflect a mix of portions of the ambient P每T array and the effects of evolving P每T conditions. The comparatively low-T biotite, garnet and staurolite zones mostly preserve kyanite-type peak assemblages that only experienced minor changes during exhumation. Rocks in the comparatively high-T sillimanite and garnet每cordierite zones are dominated by mineral assemblages of a transitional sillimanite type, having formed by the extensive modification of earlier higher pressure assemblages during exhumation. Only rocks in the intermediate-T kyanite and probably some lower sillimanite zones were clearly recrystallized by late stage andalusite metamorphism, producing the staurolite每andalusite and andalusite每sillimanite zones. This andalusite metamorphism could not reach an equilibrium state because of limited fluid availability. Keywords: Kyanite type; Andalusite type; P每T pseudosection; Chinese Altai orogen

Article Outline
1. Introduction 2. Geological setting 3. Distribution and characteristics of progressive zonal sequences 4. Mineral chemistry 5. Phase equilibria and P每T evolution of each metamorphic zone 5.1. Biotite zone 5.2. Garnet zone 5.3. Staurolite zone 5.4. Kyanite zone 5.5. Staurolite每andalusite zone 5.6. Andalusite每sillimanite zone 5.7. Sillimanite zone 5.8. Garnet每cordierite zone 6. Discussion 7. Conclusions Acknowledgements References

1. Introduction

Low-pressure metamorphic belts have recently received considerable attention, as they commonly preserve apparently anomalous thermal conditions and evidence for &anti-clockwise* P每T每t paths ([De Yoreo et al., 1991], [Azor and Ballevre, 1997], [Grae? and Schenk, 1999] and [Pattison et al., 1999]). Unlike the products of ner high-pressure metamorphism, which are mainly linked to oceanic and continental subduction processes ([Ernst, 1973], [Miyashiro, 1973] and [Ernst and Liou, 2000]), the products of low-pressure metamorphism (LPM) are linked to tectonic environments of elevated heat flux that may include arc magmatism, crustal extension, thickened crust and regions with uncommonly large aqueous fluid fluxes (see summary of De Yoreo et al., 1991). As many LPM belts require a large component of heat input by anomalously high basal heat flow and/or magmas and/or aqueous fluids, there is commonly a strong coincidence between granitic plutonism and LPM. Examples of LPM terranes in orogenic belts include the Hercynian massifs in Europe ([Azor and Ballevre, 1997] and [Grae? and Schenk, 1999]), the Acadian LP ner terrane in northern New England (De Yoreo et al., 1989), the Buchan zone of the Dalradians in Scotland ([Harte and Hudson, 1979] and [Hudson, 1980]), and the Adelaide fold belt in southern Australia (Dymoke and Sandiford, 1992). Irrespective of tectonic environment, the P每T conditions of LPM are generally in the range of 500每750 ∼ and 2每4 kbar, at pressures lower than the aluminosilicate triple point (e.g. C Vernon et al., 1990) and the inferred P每T每t paths nearly always involved isobaric heating ([De Yoreo et al., 1991], [Pattison and Tracy, 1991] and [Pattison et al., 1999]). Regional metamorphic zones in the Chinese Altai orogen include kyanite-type and andalusite-type rocks; Zhuang (1994) reported a strong spatial relationship between the distribution of metamorphic zones and oval-shaped thermal domes cored by granitic plutons. Zhuang (1994) proposed that the two types of zonal sequence were simultaneously developed at different structural locations 〞 the kyanite type developed in the flanks, and the andalusite-type in terminal ends of the thermal domes. In contrast, He et al. (1990) interpreted that the andalusite-type metamorphism was

related to a rifting event, which preceded plate convergence that resulted in kyanite type metamorphism. This paper integrates the results of three approaches to resolve relationships between the kyanite and andalusite-type rocks in the Altai orogen. (i) A systematic study of the distribution, petrography and mineralogy of metapelitic assemblages in various metamorphic zones. (ii) The resolution of peak conditions for the representative samples of each metamorphic zone on the basis of the KMnFMASH and KFMASH P每T pseudosections calculated with the software THERMOCALC 3.1 (Powell et al., 1998), constructed using the petrogenetic grids of Wei et al. (2004) and internally-consistent dataset of Holland and Powell (1998; subsequent upgrades in 2001). (iii) The definition of P每T vectors for each metamorphic zone in the Altai zonal sequences by comparing details from the calculated P每T pseudosections with textural relations preserved in the metapelitic assemblages in the context of spatial relationships of the various metamorphic zones. We show that the metamorphic zones cannot be simply described as kyanite- and andalusite-types; the metapelitic rocks record a combination of different portions of the P每T arrays and the effects of evolving P每T conditions involving early kyanite and late andalusite conditions related to a clockwise P每T path from burial to exhumation. Mineral abbreviations used in this paper are after Holland and Powell (1998). The thermodynamic models and a-X relations for the relevant solid solutions are those given by Wei et al. (2004).

2. Geological setting
The Altai orogen extends for approximately 2500 km from Russia and East Kazakhstan in the west, through Northern Xinjiang of China to southwestern Mongolia in the east (Xiao et al., 1992). It represents the southwest continental margin of the Siberian plate and is separated from the Junggar plate in the south by the NW-trending Irtysh fault. Rocks in the Chinese Altai orogen record the effects of

Neoproterozoic to late Paleozoic tectonism. He et al. (1990) and Windley et al. (2002) divided the orogen into five fault-bound terranes (see Fig. 1a). Altai terrane 1 comprises Late Devonian每Early Carboniferous neritic clastic sediments and limestones intercalated with minor island arc volcanics, metamorphosed mostly at lower greenschist facies conditions. Altai terrane 2 consists predominantly of a thick Neoproterozoic每Middle Ordovician turbidite sequence metamorphosed mostly at lower greenschist facies conditions. These turbidites are unconformably overlain by a Late Ordovician volcanic and sedimentary sequence. Altai terrane 3 forms the central part of the Altai orogen in China and comprises mainly Neoproterozoic每Middle Ordovician turbidites. Minor ca. 505 Ma felsic volcanics intercalated in the turbidite sequence have been interpreted to represent components of a continental arc (Windley et al., 2002). Rocks from this terrane were metamorphosed at greenschist to upper amphibolite facies conditions. Altai terrane 4 consists of Late Silurian每Early Devonian arc-type volcanic and pyroclastic rocks in the lower part, and of the Middle Devonian turbidites together with pillow-basalts in the upper part. Xu et al. (2003) reported a typical ophiolite sequence dominated by mafic rocks with MORB-affinity in the northwest of Fuyun (Fig. 1a), and inferred that they formed in a back-arc basin environment. These rocks show the spectrum of metamorphic zones from greenschist to upper amphibolite, and locally granulite, facies conditions. Altai terrane 5 is bounded by the Irtysh fault in the south. It includes a complex sequence including Precambrian basement, Early Paleozoic每Devonian sediments and Late Carboniferous volcanoclastics, metamorphosed at greenschist to amphibolite facies conditions. Rocks in the Junggar plate (south of the Irtysh fault) are dominated by Devonian每Carboniferous volcanoclastics, weakly metamorphosed at lower greenschist facies conditions.

Full-size image (176K) Fig. 1. (A) Geological map of the Chinese Altai orogen showing the main terranes, referred to by number in the text, and distribution of metamorphic zones (modified from [Zhuang, 1994] and [Windley et al., 2002]); (B) Detailed metamorphic map around Altai City showing the distribution of metamorphic zones and sample localities (modified from Zhuang, 1994).

Granitoids and orthogneiss form approximately 40% of rocks exposed in the Chinese Altai orogen. They are predominantly tonalite, granodiorite and biotite granite, with minor two-mica granite. Rb每Sr whole-rock isochrons and biotite K每Ar methods give late Caledonian (377每408 Ma, Zou et al., 1988) and Hercynian (290每344 Ma, [Zou et al., 1988] and [Zhang et al., 1996]) ages. More recent U每Pb dating of zircon (207Pb/206Pb age and SHRIMP) confirmed that most plutons were emplaced between 360每390 Ma ([Windley et al., 2002] and [Chen and Jahn, 2002]). These plutons are mostly related to arc magmatism (Chen and Jahn, 2002), but many S-type granite bodies originated from the crustal melting related to high-grade metamorphism. There are few data to constrain the age of metamorphism. Zhuang (1994) obtained an Rb每Sr whole rock isochron of 365 Ma (Late Devonian) from high-grade schist and gneiss in Altai terrane 3, interpreted by Windley et al. (2002) to represent the age of metamorphism. Hu et al. (2002) recalculated discordant U每Pb zircon data in Li et al. (1996) from a high-grade gneiss in the east of Altai terrane 4 and obtained a lower intercept age of 367 ㊣28 Ma that they interpreted as the age of metamorphism. These ages are synchronous with ages inferred for granitoid emplacement.

Deformation fabrics in the Altai orogen are divided into three stages of development (Zhuang, 1994): stage one is characterized by tight to isoclinal folds, inferred to have developed in the Late Devonian (375每355 Ma); stage two is dominated by doming structures that are interpreted as having developed between 345每320 Ma; and stage three is characterized by late thrusting that developed retrograde mylonite zones, probably between 310每280 Ma on the basis of limited Ar每Ar data (Zhuang, 1994). A tectonic scenario for the Altai area based on previous studies (e. g. [He et al., 1990], [Xiao et al., 1992], [Chang et al., 1995], [Chen and Jahn, 2002], [Windley et al., 2002] and [Xu et al., 2003]) involves three main stages: (i) a passive continental margin during the Neoproterozoic每Early Paleozoic; (ii) the development of a Late Silurian to Early Devonian arc environment related to the north-directed subduction of Junggar plate, with a back-arc basin having formed during the Middle Devonian; and (iii) continent-arc collision at the beginning of Late Devonian.

3. Distribution and characteristics of progressive zonal sequences
Altai terranes 3, 4 and 5 expose a pattern of progressive zonal sequences differentiated into kyanite- and andalusite-types (Zhuang, 1994) with the following characteristics: (i) both types have biotite, garnet and staurolite zones at low-T conditions, with sillimanite and, locally, garnet每cordierite zones at high-T conditions; (ii) whereas the kyanite-type includes the kyanite zone, the andalusite-type includes staurolite每andalusite and andalusite每sillimanite zones at intermediate-T conditions (see Fig. 1b); (iii) the andalusite-type zonal sequence is more spatially extensive than the kyanite-type, which sometimes occurs as relic patches scattered in the other (Fig. 1b); and (iv) rocks at and above conditions recorded by the kyanite zone include various felsic veins that contain quartz, plagioclase, K-feldspar and muscovite, with or without garnet, sillimanite and biotite, most probably derived from the partial melting of high-grade metapelites. The characteristics of each metamorphic zone are presented below.

The Biotite zone is generally 1每4 km wide with metapelitic assemblages involving biotite, muscovite, quartz and plagioclase with or without chlorite and calcite; K-feldspar also occurs in potassium-rich compositions. Accessory minerals include ilmenite, tourmaline, and graphite. No aluminosilicate or chloritoid has been found in this zone. The rocks are usually fine-grained and show a well-developed schistosity; they sometimes include coarse-grained detrital K-feldspar and plagiclase grains (see Fig. 2a).

Full-size image (321K) Fig. 2. Photographs showing representative textural relationships in metapelites from the Chinese Altai orogen. (a) Mica schist with subhedral K-feldspar porphyroblasts of probable detrital origin from the biotite zone. Crossed polars. Sample A294. (b) Garnet mica schist with subhedral garnet porphyroblast from the garnet zone. The garnet contains a few inclusions of quartz, plagioclase, and biotite. Plane polarized light. Sample A291. (c) Garnet每staurolite mica schist from the staurolite zone where the euhedral每subhedral staurolite porphyroblasts make up approximately 50 vol. % of the rock. (d) Garnet每staurolite mica schist with euhedral garnet and anhedral staurolite porphyroblasts from the staurolite zone. The porphyroblasts have inclusions of quartz and ilmenite that constitute a weakly developed internal foliation almost parallel to the external foliation composed of biotite, muscovite, quartz, plagioclase and chlorite. Plane polarized light. Sample A340. (e) A hand specimen from a quartz每aluminosilicate vein from the kyanite zone where pinkish euhedral andalusite megacrysts commonly contain bluish kyanite in their cores. (f) Garnet kyanite mica schist from the kyanite zone, containing large porphyroblasts of kyanite up to 10 mm. Crossed polars. Sample A350. Mineral abbreviations are and, andalusite, bi, biotite;

chl, chlorite; g, garnet; ksp, K-feldspar; ky, kyanite; q, quartz. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

The Garnet zone is no more than 500 m wide and separates the biotite and staurolite zones; it cannot always be in between due to the presence of insensitive (high variance) lithologies. Metapelitic assemblages commonly include garnet, biotite, muscovite, quartz and plagioclase with or without chlorite and calcite. Common accessory minerals include ilmenite, tourmaline, monazite and graphite. Garnet commonly occurs as euhedral and subhedral porphyroblasts 1每2 mm across with inclusions of quartz, plagioclase, biotite, rutile/ilmenite, and monazite (see Fig. 2b). Garnet may be partially retrogressed along grain margins and fractures to chlorite, biotite and Fe-oxide. The Staurolite zone is several hundred meters wide with metapelitic assemblages involving staurolite, garnet, biotite, muscovite, quartz and plagioclase with or without chlorite. Common accessory minerals include ilmenite, tourmaline, monazite and graphite. Staurolite sometimes forms porphyroblasts 1每3 cm in size that dominates the rock texture (Fig. 2c). Staurolite and garnet commonly have inclusions of quartz, biotite, muscovite, ilmenite and graphite. Occasionally, euhedral garnet occurs as inclusions in staurolite. The inclusions define an internal foliation that is, in most cases, parallel to and continuous with the foliation external to the porphyroblast (Fig. 2d). The Kyanite zone ranges from 1每10 km across and either occurs as an independent metamorphic zone or, in most cases, is mixed with the staurolite每andalusite zone. Detailed field mapping shows that kyanite-bearing rocks occur as relics in extensive andalusite-bearing rocks (Fig. 1b). In the kyanite-zone, decimeter-wide quartz-kyanite/andalusite veins are common, with bluish kyanite surrounded by

pinkish andalusite in hand specimen (Fig. 2e). Metapelitic assemblages involve kyanite, staurolite, garnet, biotite, quartz and plagioclase, with or without muscovite and chlorite. Kyanite grains are commonly up to 5每10 mm long and may have quartz, mica, and rare garnet inclusions (Fig. 2f). The Staurolite每andalusite zone is more extensive than the kyanite zone and includes metapelitic assemblage involving andalusite, staurolite, garnet, biotite, muscovite, quartz and plagioclase with or without garnet. Accessory minerals include ilmenite, zircon, tourmaline, apatite and graphite. Large subhedral andalusite porphyroblasts 5每10 mm across have inclusions of staurolite, biotite, muscovite, quartz and graphite that define an internal foliation continuous and parallel to the external foliation (Fig. 3a). Staurolite occurs both as independent crystals in the matrix and as inclusions in andalusite. Retrograde chlorite rims biotite and occurs as porphyroblasts 2每3 mm across that cut the penetrative foliation.

Full-size image (202K) Fig. 3. Photographs showing the textural relationships in metapelites from the Chinese Altai orogen. (a) Staurolite每andalusite mica schist with megacrysts of andalusite with inclusions of staurolite, biotite, muscovite, quartz and plagioclase from the staurolite每andalusite zone. The inclusions in andalusite generally constitute an internal foliation that is parallel to and continuous with the external foliation. Crossed polars. Sample A2121. (b) Andalusite每sillimanite mica schist from the andalusite每sillimanite zone where the fibrous and prismatic sillimanite crystals are aligned parallel to the principal foliation, with andalusite and muscovite porphyroblasts cutting the foliation. Crossed polars. Sample A2115. (c)Sillimanite biotite gneiss from the sillimanite zone with large prismatic sillimanite crystals up to

10 mm. (d) Garnet每feldspar gneiss from the garnet每cordierite zone with voluminous coarse-grained and subhedral perthitic plagioclase. Crossed polars. Sample A2152. Mineral abbreviations see Fig. 2.

The Andalusite每sillimanite zone occurs as a transition between the staurolite每andalusite and sillimanite zones, and cannot be defined everywhere (Fig. 1b). Metapelitic assemblage involves andalusite, sillimanite, garnet, biotite, muscovite, quartz and plagioclase. Accessory minerals include ilmenite, tourmaline, zircon and graphite. Sillimanite occurs in both fibrous and prismatic forms, mostly aligned with the penetrative foliation (Fig. 3b). Andalusite grains are commonly aligned with the penetrative foliation, but sometimes cut it. Many large muscovite grains also cut the penetrative foliation (Fig. 3b). Granitic veins with muscovite and garnet occur in this zone. The Sillimanite zone is extensive, sometimes being more than 10 km across. Metapelitic assemblages involve sillimanite, garnet, biotite, muscovite, quartz and plagioclase. Accessory minerals include ilmenite, tourmaline, zircon and graphite. Sillimanite occurs in both fibrous and prismatic forms, and prismatic grains may be 5每10 mm long (Fig. 3c). Muscovite occurs both as small grains and large porphyroblasts that overgrow the foliation, as in the andalusite每sillimanite zone. Rocks in the sillimanite zone contain numerous granitic veins and leucosomes. The Garnet每cordierite zone is less than 2 km across and occurs in rocks of Altai terrane 4. Muscovite was not observed in this zone. In places, garnet每cordierite每K-feldspar assemblages replace the ubiquitous biotite每sillimanite assemblage of the sillimanite zone. Metapelitic assemblage involves garnet, sillimanite, K-feldspar, biotite, cordierite, plagioclase and quartz. Accessory minerals include tourmaline, apatite, rutile and ilmenite. The rocks show schlieren and stromatic migmatite structures, dominated by voluminous coarse-grained anti-perthite

(Fig. 3d). Myrmekitic intergrowths of quartz and plagioclase are developed between plagioclase and K-feldspar.

4. Mineral chemistry
Eight samples selected from each of the metamorphic zones were studied with a JXA-8100 microprobe at Peking University in the wave length-dispersive mode with 15 kV acceleration potential, 10 nA beam current and a beam diameter of 1 micron. Matrix corrections were carried out using the PRZ correction program. The main mineral components and modal proportions for the eight samples are summarized in Table 1, and representative microprobe analyses for the main minerals are presented in Table 2, Table 3 and Table 4. Table 1. Main mineral components and modal proportion (vol. %) of metapelites from the Chinese Altai orogen
Samples A29 A29 A340 4 1 Metamo rphic zones Biot Gar ite net Staur olite A35 0 Kya nite A2121 A2115 A367 A215 2

Staurolite每an dalusite

Andalusite每sil limanite

Sillima Garne nite t每 cordie rite 5 5 1 12 12 每

Garnet Biotite

每 30

10 25 5

4 30 5

3 28 每

每 28 4

4 23 8

Muscovi 10 te Chlorite Staurolit e Cordieri te 每 每

每 每

5 10

4 15

s 14

每 每

每 每

每 每















10

Samples A29 A29 A340 4 1 Kyanite Silliman ite Andalus ite Plagiocl ase 每 每 每 每 每 每

A35 0 8 每

A2121 每 4

A2115 每 18

A367 每 25

A215 2 每 3









9

5





10

50

15

10

8

8

10

15

K-feldsp 10 ar Quartz Calcite 30 8













5

8 每

30 每

30 每

32 每

32 每

50 每

40 每

Full-size table

Table 2. Selected microprobe analyses of garnet from metapelites of the Altai orogen
Samp le Positi on SiO2 A291 A340 A350 A2115 A367 A152

Cor e 37.5 0 0.02

Rim

Cor e 37.5 7 0.00

Ri m 36. 75 0.0 4 20. 96 0.0 3 29.

Cor e 37.7 5 0.02

Rim

Cor Ri e m 37. 20 0.0 0 21. 25 0.0 0 29. 36. 93 0.0 0 20. 74 0.0 7 30.

Cor Ri e m 36. 80 0.0 0 20. 82 0.0 0 23. 37. 13 0.0 2 21. 08 0.0 0 25.

Cor e 38.2 0 0.01

Ri m 37. 91 0.0 1 21. 91 0.0 3 27.

37.3 6 0.00

38.0 4 0.03

TiO2

Al2O3

20.9 7 0.03

21.1 2 0.01

21.2 4 0.00

21.0 3 0.00

21.1 3 0.08

22.2 1 0.00

Cr2O3

FeO

15.4

26.7

27.8

25.5

28.8

26.7

Samp le Positi on

A291

A340

A350

A2115

A367

A152

Cor e 0

Rim

Cor e 3 9.10

Ri m 73 7.1 8 3.1 7 1.5 7 99. 40

Cor e 8 10.4 8 2.61

Rim

Cor Ri e m 74 7.4 9 2.9 1 1.2 7 99. 86 48 7.3 0 2.2 2 1.4 1 99. 08

Cor Ri e m 32 14. 48 3.6 1 0.0 7 99. 10 05 11. 99 3.8 7 0.7 1 99. 85

Cor e 3 7.21

Ri m 95 7.0 8 4.1 0 0.9 1 99. 90

9 10.8 1 2.67

8 6.51

MnO

16.0 7 0.75

MgO

2.98

3.61

5.40

CaO

9.60

1.46

1.60

2.60

2.19

0.94

Total

100. 34

100. 22

100. 32

100. 07

100. 47

100. 70

Normalized to 8 cations and 12 oxygens Si 2.99 5 1.97 5 0.00 2 0.00 1 0.03 0 0.99 8 1.08 7 0.08 9 0.82 3.00 4 2.00 2 0.00 1 0.00 0 0.00 0 1.80 2 0.73 7 0.32 0 0.12 3.00 9 2.00 5 0.00 0 0.00 0 0.00 0 1.86 4 0.61 7 0.35 6 0.13 2.9 74 1.9 99 0.0 02 0.0 02 0.0 46 1.9 66 0.4 92 0.3 82 0.1 3.02 7 1.90 0 0.00 0 0.00 1 0.00 0 1.71 5 0.71 2 0.31 2 0.22 3.02 3 1.98 0 0.00 5 0.00 2 0.00 0 1.92 0 0.43 8 0.42 8 0.18 2.9 98 2.0 19 0.0 00 0.0 00 0.0 00 2.0 05 0.5 11 0.3 50 0.1 3.0 12 1.9 94 0.0 05 0.0 00 0.0 00 2.0 79 0.5 04 0.2 70 0.1 2.9 87 1.9 92 0.0 00 0.0 00 0.0 34 1.5 49 0.9 95 0.4 37 0.0 2.9 81 1.9 95 0.0 00 0.0 01 0.0 39 1.6 43 0.8 15 0.4 63 0.0 2.99 2 2.05 1 0.00 0 0.00 1 0.00 0 1.75 1 0.47 8 0.63 0 0.07 3.0 11 2.0 52 0.0 02 0.0 01 0.0 00 1.8 57 0.4 76 0.4 85 0.0

Al

Cr

Ti

Fe3+

Fe2+

Mn

Mg

Ca

Samp le Positi on

A291

A340

A350

A2115

A367

A152

Cor e 2

Rim

Cor e 7 0.63

Ri m 36 0.6 6 0.1 7 0.1 3 0.0 5

Cor e 4 0.58

Rim

Cor Ri e m 10 0.6 7 0.1 7 0.1 2 0.0 4 23 0.7 0 0.1 7 0.0 9 0.0 4

Cor Ri e m 06 0.5 2 0.3 3 0.1 5 0.0 0 61 0.5 5 0.2 7 0.1 6 0.0 2

Cor e 9 0.60

Ri m 78 0.6 4 0.1 6 0.1 7 0.0 3

6 0.60

7 0.65

XFe

0.33

XMn

0.36

0.25

0.21

0.24

0.15

0.16

XMg

0.03

0.11

0.12

0.11

0.14

0.22

XCa

0.27

0.04

0.05

0.08

0.06

0.03

XFe = Fe2+/(Fe2+ + Mn + Mg + Ca), XMn = Mn/(Fe2+ + Mn + Mg + Ca), XMg = Mg/(Fe2+ + Mn + Mg + Ca), XCa = Ca/(Fe2+ + Mn + Mg + Ca).

Table 3. Selected microprobe analyses of mica from metapelites of the Altai orogen
Sam ple A294 A291 A340 A3 50 mu bi A2121 A2115 A367 A21 52 mu bi

Mine bi ral SiO2 38. 97 1.9 7 17. 71 0.0

mu

bi

mu

bi

bi

mu

bi

mu

bi

47. 76 0.6 0 32. 91 0.0

35. 95 1.6 1 18. 52 0.0

47. 50 0.9 2 32. 73 0.0

37. 01 1.4 2 19. 66 0.0

48. 38 0.3 8 36. 28 0.1

37. 19 1.5 0 19. 09 0.1

35. 70 1.3 8 19. 07 0.0

46. 58 0.3 6 37. 15 0.0

35. 06 1.6 8 19. 63 0.0

48. 13 0.5 9 36. 70 0.0

36. 98 1.1 7 19. 46 0.0

47. 38 0.0 7 36. 56 0.0

35.6 2 2.56

TiO2

Al2O
3

19.3 2 0.08

Cr2O

Sam ple

A294

A291

A340

A3 50 mu bi

A2121

A2115

A367

A21 52 mu bi

Mine bi ral
3

mu

bi

mu

bi

bi

mu

bi

mu

bi

4 11. 04 0.2 5 14. 91 0.0 2

3 1.2 9 0.0 0 1.8 3 0.0 3 0.2 5 10. 81 95. 51

8 18. 27 0.1 7 11. 24 0.0 0 0.1 5 9.3 1 95. 30

0 1.7 1 0.0 0 1.3 2 0.0 4 0.3 7 10. 03 94. 63

5 16. 41 0.1 5 12. 03 0.0 2 0.2 3 9.2 0 96. 19

0 0.9 3 0.0 6 0.5 3 0.0 5 1.2 3 8.1 5 96. 09

4 16. 27 0.0 4 12. 56 0.0 1 0.1 4 9.0 6 96. 00

4 17. 50 0.2 2 11. 67 0.0 1 0.3 1 8.4 5 94. 35

8 1.0 5 0.0 7 0.3 6 0.0 1 1.9 3 8.0 8 95. 67

5 19. 11 0.0 5 9.9 9 0.0 0 0.1 3 8.7 3 94. 43

3 0.7 8 0.1 1 0.4 8 0.0 0 1.2 1 8.3 4 96. 37

2 13. 56 0.2 5 13. 36 0.0 3 0.5 2 8.8 1 94. 16

4 0.5 5 0.0 2 0.4 7 0.0 2 2.0 0 8.6 1 95. 72 15.3 9 0.09

FeO

MnO

MgO

11.1 2 0.00

CaO

Na2O 0.0 7 K2O 9.9 6 94. 94

0.13

9.77

Total

94.0 8

For muscovite: normalized on the sum of tetrahedral and octahedral cations = 6.05 for 11 oxygens; and for biotite: normalized on the sum of tetrahedral and octahedral cations = 6.9 for 11 oxygens Si 2.8 51 1.1 49 0.3 78 0.0 02 0.1 08 3.1 72 0.8 28 1.7 46 0.0 02 0.0 30 2.7 19 1.2 81 0.3 70 0.0 05 0.0 92 3.1 77 0.8 23 1.7 58 0.0 00 0.0 46 2.7 34 1.2 66 0.4 46 0.0 03 0.0 79 3.1 36 0.8 64 1.9 09 0.0 05 0.0 19 2.7 48 1.2 52 0.4 11 0.0 08 0.0 83 2.6 92 1.3 08 0.3 87 0.0 02 0.0 78 3.0 50 0.9 50 1.9 18 0.0 04 0.0 18 2.6 75 1.3 25 0.4 41 0.0 03 0.0 96 3.1 14 0.8 86 1.9 13 0.0 02 0.0 29 2.7 53 1.2 47 0.4 61 0.0 01 0.0 65 3.0 99 0.9 01 1.9 18 0.0 02 0.0 03 2.69 8 1.30 2 0.42 3 0.00 5 0.14 6

AlIV

AlVI

Cr

Ti

Sam ple

A294

A291

A340

A3 50 mu bi

A2121

A2115

A367

A21 52 mu bi

Mine bi ral Fe3+ 0.0 00 0.6 75 0.0 15 1.6 25 0.0 02 0.0 10 0.9 30 7.7 47 0.2 93

mu

bi

mu

bi

bi

mu

bi

mu

bi

0.0 00 0.0 72 0.0 00 0.1 81 0.0 02 0.0 32 0.9 16 6.9 81 0.2 85

0.0 03 1.1 53 0.0 11 1.2 67 0.0 00 0.0 22 0.8 99 7.8 21 0.4 76

0.0 00 0.0 96 0.0 00 0.1 32 0.0 03 0.0 48 0.8 57 6.9 39 0.4 21

0.0 00 1.0 14 0.0 09 1.3 25 0.0 02 0.0 33 0.8 68 7.7 79 0.4 33

0.0 00 0.0 50 0.0 03 0.0 51 0.0 03 0.1 55 0.6 75 6.8 71 0.4 95

0.0 00 1.0 06 0.0 03 1.3 83 0.0 01 0.0 20 0.8 55 7.7 70 0.4 21

0.1 01 1.0 05 0.0 14 1.3 12 0.0 01 0.0 45 0.8 14 7.7 60 0.4 34

0.0 00 0.0 57 0.0 04 0.0 35 0.0 01 0.2 45 0.6 76 6.9 57 0.6 20

0.0 18 1.2 02 0.0 03 1.1 36 0.0 00 0.0 19 0.8 51 7.7 70 0.5 14

0.0 00 0.0 42 0.0 06 0.0 46 0.0 00 0.1 52 0.6 89 6.8 78 0.4 77

0.0 00 0.8 44 0.0 16 1.4 82 0.0 02 0.0 75 0.8 37 7.7 84 0.3 63

0.0 00 0.0 30 0.0 01 0.0 46 0.0 01 0.2 54 0.7 18 6.9 74 0.3 95

0.00 0 0.97 5 0.00 6 1.25 5 0.00 0 0.01 9 0.94 5 7.77 3 0.43 7

Fe2+

Mn

Mg

Ca

Na

K

Total

XFe

Full-size table XFe = Fe/(Fe + Mg).

Table 4. Selected microprobe analyses of feldspar, staurolite, chlorite, and cordierite from metapelites of the Altai orogen

Sam A294 ple Min eral pl kf

A2 A340 91 pl pl st

A350

A2121

A2 A3 A152 115 67 pl pl pl kf cd

chl pl

st

pl

st

SiO2 57. 56 TiO2 0.0 3 Al2 O3 Cr2 O3 FeO 26. 98 0.0 0 0.0 6 0.0 4 0.0 0 每

65. 38 0.0 2 18. 42 0.0 0 0.0 4 0.0 5 0.0 0 每

60. 60. 28. 26. 59. 84 80 10 22 50 0.0 0.0 0.5 0.1 0.0 3 0 1 9 0 24. 24. 54. 22. 25. 53 58 24 00 98 0.0 0.0 0.0 0.0 0.0 4 0 5 0 1 0.0 0.1 12. 20. 0.0 8 9 46 33 2 0.0 0.0 0.4 0.2 0.0 0 4 4 0 0 0.0 0.0 2.0 16. 0.0 1 1 0 68 0 每 每 0.1 每 0 每

27. 62. 28. 61. 60 36 24 82 0.2 0.0 0.6 0.0 7 0 4 0 54. 23. 54. 24. 15 72 50 27 0.1 0.0 0.0 0.0 3 4 8 1 13. 0.0 13. 0.0 33 0 04 7 0.4 0.0 0.5 0.0 1 0 6 0 2.1 0.0 1.7 0.0 5 0 1 0 0.0 每 8 0.1 每 9

63. 60. 64. 48. 54 89 19 47 0.0 0.0 0.0 0.0 5 1 6 0 22. 24. 18. 32. 48 28 60 35 0.0 0.0 0.0 0.0 1 1 0 6 0.0 0.0 0.0 5.3 0 0 2 6 0.0 0.0 0.0 0.3 0 0 0 6 0.0 0.0 0.0 9.5 1 0 1 3 每 每 每 每

Mn O Mg O ZnO

CaO

8.8 7 6.4 2 0.1 3

0.0 0 0.5 1 15. 66

6.0 6.3 0.0 0.0 7.5 5 5 0 1 8 8.2 7.8 0.0 0.0 7.1 5 9 0 7 6 0.0 0.0 0.0 0.4 0.0 8 8 0 8 5

0.0 4.8 0.0 5.6 0 0 0 4 0.0 8.9 0.0 8.2 0 2 0 9 0.0 0.0 0.0 0.0 0 5 0 9

3.9 5.6 0.2 0.0 8 8 0 4 9.5 8.3 1.0 0.1 9 3 5 1 0.0 0.1 15. 0.0 9 7 21 2

Na2 O K2O

Tota l

100 100 99. 99. 97. 86. 100 98. 99. 98. 100 99. 99. 99. 96. .09 .10 91 94 90 18 .30 12 89 96 .19 85 37 36 63

Normalized to 8 oxygens and total Fe assumed as ferric for feldspar, to 46 oxygens for staurolite; to 11 cations for 18 oxygens for cordierite; and to 10 cations for 14 oxygens for chlorite Si 2.5 3.0 2.7 2.7 7.7 2.7 2.6 7.6 2.7 7.7 2.7 2.8 2.7 2.9 5.0

Sam A294 ple Min eral pl kf

A2 A340 91 pl pl st

A350

A2121

A2 A3 A152 115 67 pl pl pl kf cd

chl pl

st

pl

st

76 Al 1.4 23 0.0 00 0.0 01 0.0 02 每

07 0.9 99 0.0 00 0.0 01 0.0 02 每

07

05

51

38

43

37

63

34

36

15

21

80

21

1.2 1.2 17. 2.7 1.3 87 89 63 09 61 0.0 0.0 0.0 0.0 0.0 01 00 11 00 00 0.0 0.0 0.1 0.0 0.0 01 00 06 15 00 0.0 0.0 0.0 0.0 0.0 03 06 00 00 01 每 每 2.8 1.7 每 74 76

17. 1.2 17. 1.2 67 39 59 66 0.0 0.0 0.0 0.0 28 01 17 00 0.0 0.0 0.1 0.0 56 00 32 00 0.0 0.0 0.0 0.0 00 00 00 03 3.0 每 85 2.9 每 87

1.1 1.2 1.0 3.9 81 79 18 51 0.0 0.0 0.0 0.0 00 00 00 05 0.0 0.0 0.0 0.0 02 00 02 00 0.0 0.0 0.0 0.0 00 00 01 26 每 每 每 0.4 65

Cr

Ti

Fe3+

Fe2+

Mn

0.0 00 0.0 00 每

0.0 02 0.0 00 每

0.0 0.0 0.1 0.0 0.0 00 02 03 18 00 0.0 0.0 0.8 2.5 0.0 01 01 22 96 00 每 每 0.0 每 20 每

0.0 0.0 0.1 0.0 96 00 30 00 0.8 0.0 0.6 0.0 87 00 98 00 0.0 每 16 0.0 每 38

0.0 0.0 0.0 0.0 00 00 00 32 0.0 0.0 0.0 1.4 01 00 01 71 每 每 每 每

Mg

Zn

Ca

0.4 25 0.5 57 0.0 07 每

0.0 00 0.0 45 0.9 20 每

0.2 0.3 0.0 0.0 0.3 88 03 00 01 61 0.7 0.6 0.0 0.0 0.6 12 81 00 14 17 0.0 0.0 0.0 0.0 0.0 05 05 00 64 03 每 每 0.7 0.4 每 57 05

0.0 0.2 0.0 0.2 00 28 00 67 0.0 0.7 0.0 0.7 00 66 00 11 0.0 0.0 0.0 0.0 00 03 00 05 0.7 58 0.8 每 11

0.1 0.2 0.0 0.0 87 72 10 04 0.8 0.7 0.0 0.0 13 22 95 22 0.0 0.0 0.9 0.0 05 10 02 03 每 每 每 0.2 40

Na

K

XFe

Full-size table XFe (st, chl, cd) = Fe/(Fe + Mg).

Garnet rim compositions lie in the range of XFe = Fe/(Fe + Mn + Mg + Ca) = 0.55每0.70, XMn = Mn/(Fe + Mn + Mg + Ca) = 0.15每0.27, XMg = Mg/(Fe + Mn + Mg + Ca) = 0.09每0.17, and XCa = Ca /(Fe + Mn + Mg + Ca) = 0.03每0.06 (Table 2). The XMg at garnet rims increases from 0.11每0.17, and the XMn decreases from 0.25每0.16, with metamorphic grade (Fig. 4). Garnet grains from different metamorphic zones have distinct chemical zonation. Garnet from the garnet zone has zoning typical of single-stage growth (e. g. Tracy, 1982), with distinctive progressive increases of almandine (32.6每60.1%) and pyrope (3.0每10.7%) and antithetic deceases of grossular (27.6每1.9%) and spessartine (34.6每24.6%) from core to rim (Fig. 5 and Table 2). Garnet from the staurolite and kyanite zones also has this single-stage growth pattern. Garnet in the staurolite zone involves a moderate increase of almandine (61.6每65.7%) and decrease of spessartine (20.8每16.3%), subtle increase of pyrope (11.6每13.4%) and almost no change in grossular content (4.6每4.8%) from core to rim. Similarly, garnet from the kyanite zone is zoned in almandine (57.1每63.9%), spessartine (23.0每15.9%), pyrope (10.9每14.0%), and grossular (7.0每6.2%) content from core to rim. Garnet from the andalusite每sillimanite zone has weak zoning in almandine (66.6每70.0%), spessartine (17.2每14.7%), pyrope (12.0每10.4%), and grossular (3.5每3.6%) content from core to rim. This progression involving diminishing zoning profiles linked to increasing metamorphic grade increase is similar to that documented by Dempster (1985) from another medium-pressure metamorphic series. Garnet from the garnet-cordierite zone is distinct: it has chemical zonation involving increasing almandine (58.5每62.2%), decreasing pyrope (21.1每16.3%) with constant grossular (2.6%) and spessartine content (16.0%), from core to rim (Table 2).

Full-size image (34K) Fig. 4. Ternary plots with apices pyrope, grossular and spessatine showing the composition of garnet grains from various metamorphic zones in the Chinese Altai orogen.

Full-size image (90K) Fig. 5. Representative zoning profiles of almandine (Alm); grossular (Gross), pyrope (Py) and spessartine (Spss) through garnet grains from the garnet, staurolite, kyanite and andalusite每sillimanite zones in the Chinese Altai orogen. A, B, C, D and E denote locations corresponding to the garnet compositions calculated in the P每T pseudosections.

Biotite has a similar composition throughout most metamorphic zones (Table 3; Fig. 6): XFe (Fe/(Fe + Mg)) = 0.36每0.51, AlIV = 1.22每1.33, AlVI = .37每0.46, and Ti = 0.08每0.11 on a stoichiometric formula of 11 oxygens. Biotite in the biotite zone has lower AlIV (1.14每1.18) and XFe = .28每0.29, and biotite in the garnet每cordierite zone has higher Ti content (0.15每0.19) than these nominal values.

Full-size image (31K) Fig. 6. AlIV vs. Fe/(Fe + Mg) diagram showing the composition of biotite from the various metamorphic zones in the Chinese Altai orogen.

Muscovite is also uniform in composition throughout most metamorphic zones with Si = 3.05每3.14 and Na = 0.15每0.25 on a stoichiometric formula of 11 oxygens. However, muscovite from the biotite zone and some grains from the garnet zone have higher Si (3.17每3.21) and lower Na (0.03每0.05) contents than average. Muscovite from the sillimanite zone and the S-type granitic veins is richer in Na (0.20每0.25; Table 3; Fig. 7) than elsewhere. These observations suggest that the compositions of both biotite and muscovite were mostly reset during retrogression, but in cases, record P每T conditions.

Full-size image (29K) Fig. 7. Si vs. Na/(K + Na + Ca) diagram showing the composition of muscovite from the various metamorphic zones in the Chinese Altai orogen.

Staurolite has XFe (=Fe/(Fe + Mg + Mn)) = 0.75每0.76 in the staurolite and kyanite zones, and 0.76每0.81 in the staurolite每andalusite zone. Inclusions of staurolite in andalusite have lower XFe = 0.76 than staurolite in the matrix (0.78每0.81; Table 4).

There seems to be a general correlation between the Fe content in staurolite and inferred pressure conditions. All staurolite has very low ZnO content: 0.08每0.19 wt.% (Table 4). Plagioclase has anorthite contents from 0.16每0.43, being mostly less than 0.30 (Table 4). There was no discernable relation between plagioclase composition and metamorphic grade, indicating that plagioclase composition was mostly governed by bulk rock composition. Plagioclase grains from rocks below the sillimanite zone are anhedral and 0.1每0.5 mm across; grain size increases subtly with metamorphic grade. Plagioclase grains from the garnet每cordierite zone are mostly subhedral and 2每5 mm across; some contain K-feldspar in patches of varying size and shape (Fig. 3d). K-feldspar in the lowest biotite zone is subhedral and coarse-grained (1每3 mm, Fig. 2a), whereas K-feldspar in the highest garnet每cordierite zone forms anhedral grains 0.5每1 mm across that generally contain fine albite lamellae. Two microprobe analyses of K-feldspar from the biotite and garnet每cordierite zones have orthoclase contents of 0.95 and 0.91, respectively. Chlorite is mostly retrograde as it usually cuts the penetrative foliation or occurs as pseudomorphs after biotite and garnet. A small amount of chlorite in the staurolite zone is apparently in textural equilibrium with peak minerals. All chlorite is similar in composition with Fe/(Fe + Mg) = 0.33每0.43 and AlIV = 1.26每1.46 on a stoichiometric formula of 14 oxygens.

5. Phase equilibria and P每T evolution of each metamorphic zone
To correlate observed mineral assemblages with those predicted in calculations, the chosen model system must reflect the composition of the rocks being studied as completely as possible. Pseudosection calculations have been undertaken in large systems, such as NCKMnFMASHTO to model partial melting process in metapelitic rocks (White et al., 2003). However, it is commonly more convenient to consider smaller systems from the point of view of focusing on equilibria of the

ferromagnesian minerals, along with the other main minerals that occur in KFMASH, for example muscovite, K-feldspar and the aluminosilicates in metapelitic rocks. The KFMASH system has proved to be robust in systematic studies of metapelitic mineral equilibria ([Powell and Holland, 1990], [Xu et al., 1994], [Pitra and De Waal, 2001] and [Wei et al., 2004]). Non-KFMASH components, such as Na2O, CaO, MnO, TiO2 and Fe2O3 etc. have various effects on the KFMASH equilibria ([Mahar et al., 1997], [Worley and Powell, 1998], [White et al., 2000], [White et al., 2001] and [White et al., 2003]), but, with the exception of MnO and as discussed in Wei et al. (2004), they are generally not critical for understanding the stability of ferromagnesian assemblages. The role of MnO is commonly critical, primarily through its role in stabilizing garnet-bearing equilibria. Thus, in modelling phase equilibria of the metapelites from the Altai orogen, we have calculated pseudosections on the basis of the KMnFMASH and KFMASH grids (Wei et al., 2004) respectively for the garnet-bearing and garnet-absent assemblages. It is critical to generate an effective bulk composition (Vance and Holland, 1993) in the chosen model systems for the phase equilibrium calculation metamorphic rocks which may have had various equilibrium domains, even on the thin-section scale (Waters and Charnley, 2002). There have been several methods for calculating effective bulk composition modification due to elemental partitioning related to garnet growth ([Marmo et al., 2002], [Evans, 2004] and [Tinkham and Ghent, 2005]). For example, effective bulk compositions may be generated by subtracting the values of components fractionated into garnet from a bulk-rock X-ray-fluorescence analysis of the rock (Tinkham and Ghent, 2005). However, these methods may be inappropriate for rocks with heterogeneities related to composition layering or porphyroblastic clusters in any given rock. A key part of this study is to evaluate the peak conditions for each metamorphic zone. Accordingly we use a simple approach to obtain an effective bulk composition, involving integrating data from the modal abundance of all phases relevant to the modal system with the microprobe analyses of grain rims ([Carson et al., 1999], [Wei et al., 2003] and [White et al., 2003]). Although the

uncertainties in estimation of the modal abundance cannot be precisely defined, our systematic calculations show that the pseudosections calculated with the effective bulk compositions generated by this approach more accurately reflect observations than pseudosections calculated with only XRF analyses. The effective bulk rock compositions used in the calculations, which are estimated from a combination of point-counted modal proportions (Table 1) and electron microprobe data (Table 1, Table 2, Table 3 and Table 4) are presented in Table 5. The calculated pseudosections in the KMnFMASH and KFMASH are shown in Fig. 8, Fig. 9, Fig. 10, Fig. 11 and Fig. 12. Table 5. Effective bulk rock compositions for the representative samples of each metamorphic zone from the Altai orogen
Sample Al2O3 MgO FeO A294 A291a A340 A350 A2121 A2115 A367a A2152
a

K2O

MnO FeO/(FeO + MgO) 0.29 0.56 0.51 0.51 0.51 0.58 0.51 0.57

29.99 25.02 36.61 45.41 51.60 54.17 76.82 33.35

36.86 15.30 17.55 0.30 26.59 33.72 10.84 3.82 26.31 27.55 7.97 23.36 24.42 5.80 20.49 21.36 6.55 15.20 20.94 7.89 8.73 9.13 2.75 1.57 1.01 每 1.80 2.57 5.22

23.91 31.47 6.06

For the large garnet grains with clear zonation in samples A291 and A367, only half

of their volume proportions are involved in generating the bulk rock composition (Carson et al., 1999).

Full-size image (97K) Fig. 8. (a) KFMASH (+ q + H2O for the subsolidus condition) P每T pseudosection for a mica schist (sample A294) from the biotite zone, and (b) KMnFMASH (+ q + H2O for the subsolidus condition) P每T pseudosection for a garnet mica schist (sample A291) from the garnet zone of the Chinese Altai orogen, their effective bulk rock compositions are listed in Table 5. The pseudosections are contoured with isopleths of H2O-content for the corresponding mineral assemblages (grey dashed lines with labels such as h0.225), and of the observed mole proportions on one-oxide basis for the main phases labeled such as bi = 0.69. The H2O-content refers to H2O/(K2O + FeO + MgO + Al2O3 + H2O) in KFMASH (a) and to H2O/(K2O + MnO + FeO + gO + Al2O3 + H2O) in KMnFMASH (b) on mole proportion. (a) is contoured with isopleths of Si content in muscovite (black dashed lines with labels such as Si = 3.20), and (b) contoured with isopleths of garnet composition shown as dotted lines with labels such as x0.62 for x(g) and m0.26 for m(g). A每B每C每D每E (solid arrow) in (b) shows garnet compositions matching those at the corresponding locations of garnet zone in Fig. 5. The heavy dotted line represents the H2O-saturated solidus.

Full-size image (127K) Fig. 9. KMnFMASH (+ q + H2O for the subsolidus condition)P每T pseudosection for a garnet staurolite mica schist (sample A340) from the staurolite zone in the Chinese Altai orogen with the effective bulk rock composition listed in Table 5. Isopleth details and line labeling as listed for Fig. 8.

Full-size image (107K) Fig. 10. KMnFMASH (+ q + H2O for the subsolidus condition)P每T pseudosection for a garnet kyanite mica schist (sample A350) from the kyanite zone in the Chinese Altai orogen with the effective bulk rock composition listed in Table 5. The big Arrow with &HO* in the right diagram directs decrease of H2O-content. Isopleth details and line labeling as listed for Fig. 8.

Full-size image (145K) Fig. 11. (a) KFMASH (+ q + H2O for the subsolidus condition) for a staurolite andalusite mica schist (sample A2121) from the from the staurolite每andalusite zone. (b) KMnFMASH (+ q + H2O for the subsolidus condition) P每T pseudosection for a garnet sillimanite mica schist (sample A2115) from the andalusite每sillimanite zone in the Chinese Altai orogen; their effective bulk rock compositions are listed in Table 5. The solid line with H2000 in (b) denotes the result calculated using the garnet每biotite geothermometer of Holdaway (2000). Isopleth details and line labeling as listed for Fig. 8.

Full-size image (132K) Fig. 12. P每T pseudosections in KMnFMASH (+ q + H2O for the subsolidus condition and + q + melt for the suprasolidus condition) for: (a) a garnet sillimanite mica schist

(sample A367) from the sillimanite zone; and (b) for a garnet feldspar gneiss (sample A2152) from the garnet每cordierite zone. The effective bulk rock compositions are listed in Table 5. Isopleth details and line labeling as listed for Fig. 8.

5.1. Biotite zone The calculated KFMASH P每T pseudosection for sample A294 is presented in Fig. 8a. It is dominated by trivariant fields, with the observed mineral assemblage involving muscovite, biotite, K-feldspar and quartz (+ plagioclase) stable over a wide P每T domain. To refine the P每T conditions for the observed mineral assemblage, isopleths of the measured mineral modes and Si content in muscovite were contoured for the appropriate fields in Fig. 8a. The measured modes (kf = 0.16, mu = 0.15 and bi = 0.69 on one-oxide basis) lie in either the andalusite or the sillimanite field, and the measured muscovite composition (Si = 3.17每3.21) lies in the kyanite field. This could be explained by the observed mineral assemblage having incompletely equilibrated during either prograde or retrograde events, or by the equilibration domains having been smaller than the area used to constrain the mineral proportions. Textural relations involving large grains of K-feldspar in a comparatively fine-grained matrix would be consistent with much of the K-feldspar being detrital in origin, supporting the interpretation of poor equilibration. Moreover, as discussed by Guiraud et al. (2001) and Clarke et al. (in press) the amount of structurally-bound H2O in minerals plays a critical role in the formation and preservation of mineral assemblages, controlling key changes in rocks undergoing progressive metamorphism. Isopleths of H2O-content of the mineral assemblage in the calculated P每T pseudosections can provide much useful information on plausible evolution trajectories for any given mineral assemblage. For instance, an equilibrium mineral assemblage in a closed system can only evolve via dehydration, crossing contours of decreasing H2O-content and a metamorphic peak should represent a point at which the metamorphic P每T path becomes tangential to a H2O-content contour. Contours of H2O-content saturated for

the relevant mineral assemblages in Fig. 8a indicate that the mineral assemblage was stable in the kyanite field, and a decompressional P每T path from the inferred peak conditions would evolve to mineral assemblages with lower water contents, most probably facilitating changes from the peak assemblage. Even if the mineral assemblage has equilibrated in the kyanite field, it would have most probably recrystallized during exhumation. On the basis of the measured Si contents in muscovite and with a reference to the P每T condition of the adjacent garnet zone, P每T conditions of 4每5 kbar and approximately 500 ∼ are inferred for the biotite zone. C 5.2. Garnet zone In the KMnFMASH P每T pseudosection calculated for sample A291 (Fig. 8b), the observed mineral assemblage involving garnet, biotite, muscovite and quartz (+ plagioclase) is quadrivariant and stable across a wide P每T range at P > 4 kbar and T = 550每700 ∼ Modal isopleths appropriate to the observed proportions of garnet C. (0.17), biotite (0.71) and muscovite (0.12) are consistent with a pressure range of 7每9 kbar, with 1 ? 考 uncertainties of 1每1.2 kbar at constant temperature for the location of each isopleth. Contours of garnet compositions x(g)(= Fe/(Fe + Mg + Mn) and m(g)(= Mn /(Fe + Mg + Mn) are consistent with the observed garnet zonation (Fig. 5) having formed as P每T conditions evolved from 6.1 kbar/510 ∼ (core) to C 6.8 kbar/580 ∼ (rim; see the solid arrow A每B每C每D每E in Fig. 8b). Garnet would have C mostly grown in the trivariant field involving garnet, muscovite, chlorite and biotite, in which garnet composition is mostly influenced by temperature conditions. P每T conditions of approximately 6.8 kbar and 580 ∼ inferred on the basis of the garnet C rim composition match well with the observed assemblage and modal proportions, consistent with the mineral assemblage having equilibrated at the inferred peak. Isopleths of H2O-content for mineral assemblages (Fig. 8b) predict that the mineral assemblage would probably evolve with P每T changes from point A每E as the P每T path crosses contours of decreasing H2O-content, but would remain unchanged during the decompression stage as the PT path traverses over H2O-content invariant or

increasing fields (Guiraud et al., 2001). As a consequence, the mineral assemblage at peak stage (point E) is predicted to be preserved. 5.3. Staurolite zone In the KMnFMASH P每T pseudosection calculated for sample A340 from the staurolite zone, the observed mineral assemblage involving garnet, staurolite, biotite, muscovite, chlorite and quartz (+ plagioclase) is divariant and stable across a narrow temperature range that overlaps the kyanite and sillimanite fields (Fig. 9). Isopleths appropriate to the observed mole proportions for garnet (0.09), biotite (0.54) and staurolite (0.18) are consistent with P每T conditions of 7.8 kbar and 598 ∼ with a C 1 ? 考 uncertainty of 1.0 kbar and 9每11 ∼ (point C). Isopleths of garnet compositions C are consistent with garnet growth in this zone (Fig. 5) having occurred over P每T conditions from 560 ∼ C/6每7 kbar (core; point A) to 6.9 kbar/588 ∼ (rim) with 1 ? 考 C uncertainties of 0.8每1.0 kbar and 9每11 ∼ Combining the results from the mole C. proportions and the garnet rim composition, P每T conditions of 7.3 kbar and 593 ∼ C (point B) are inferred for the zone. Isopleths of H2O-content predict that the mineral assemblage would evolve along the prograde P每T path as it moves to mineral assemblages with lower H2O-content (Guiraud et al., 2001). The most plausible P每T vector to preserve this peak assemblage would involve a decompression P每T path crossing the narrow divariant field involving garnet, muscovite, biotite, chlorite and staurolite, almost parallel to the isopleths of H2O-content, or with decompression being accompanied by cooling moving to assemblages with higher H2O-content. 5.4. Kyanite zone In the KMnFMASH P每T pseudosection calculated for sample A350 from the kyanite zone, the observed mineral assemblage involving garnet, biotite, staurolite, kyanite and quartz (+ plagioclase) is trivariant and stable over a P每T range of 615每650 ∼ and C 6每10 kbar, mostly being in the kyanite field. Isopleths of mole proportions that match those observed for garnet (0.07), kyanite (0.13) and staurolite (0.28) are consistent

with this P每T range. Isopleths of garnet compositions x(g) and m(g) in Fig. 10 predict that: (i) garnet growth mostly occurred in the trivariant field involving garnet, muscovite, chlorite and staurolite before the appearance of biotite; (ii) compositions of garnet core and mantle record a P每T vector from 7.7 kbar/570 ∼ (point A) to C 8 kbar/590 ∼ (point C); and (iii) the garnet rim composition does not correlate with a C specific P每T condition (being close to both points D and E in Fig. 10). Using appropriate mode isopleths of the observed assemblage, we choose point E (8.7 kbar and 630 ∼ as representing peak conditions for this sample. Isopleths of H2O-content C) in mineral assemblages in Fig. 10 show that the prograde portion of the P每T path moves to assemblages with decreasing H2O-content, predicting that assemblage evolution would accompany this P每T vector. However, the probable exhumation portion of P每T path moves to assemblages with higher H2O-content, predicting preservation of the peak assemblage. 5.5. Staurolite每andalusite zone As there is no garnet in sample A2121, phase relations can be modelled in the KFMASH system. In the calculated P每T pseudosection (Fig. 11a), the observed mineral assemblage involving staurolite, biotite, muscovite, andalusite, sillimanite and quartz (+ plagiclase) would be invariant in the KFMASH with a P每T condition at the intersection (point D) of the univariant reactions and = sill and mu + chl + st = bi + als (labeled as reaction 1 in Fig. 11a) were the rock in an equilibrated state. However, the mineral compositions and modal proportions predicted at point D are quite different to the measured values (Table 6), and the assemblage lacks the predicted chlorite. It is more probable that the observed assemblage is divariant relative to reaction 1, all chlorite having been consumed to form biotite and aluminosilicate. For this rock composition, staurolite is only predicted to occur with kyanite or sillimanite (not andalusite) at P每T conditions above reaction 1. Isopleths of the measured modal proportions for staurolite, biotite and aluminosilicate lie in the kyanite-stable divariant field involving muscovite, biotite, staurolite and aluminosilicate. There is thus a

probability that the observed assemblage could be metastable and transitional from the kyanite to andalusite-bearing equilibria. In this context, three probable P每T conditions are evaluated for the observed assemblage (points A, B and C in Fig. 11a). A comparison of the calculated mineral compositions and modal proportions at points A, B and C and the measured values (Table 6) shows that the measured modal proportions for biotite, staurolite and aluminosilicate are closer to values for point A, but the measured compositions of biotite and staurolite lie much closer to those for point B (Table 7). The presence of andalusite suggests that the rock did witness conditions appropriate to the andalusite field, such as, at point C, but the calculated mineral compositions at this point are remarkably different from the measured values. A probable P每T path is proposed in Fig. 11a involving decompression from the kyanite to andalusite-fields. Along such a path, a rock with the composition of A2121 would initially move to assemblages with lower H2O-content, leading to the predicted disappearance of muscovite at P > 7.5 kbar, and further decompression would move to assemblages with higher H2O-content. This could have prevented further evolution of the mineral assemblage, with the exception of the polymorphic transformation from kyanite or sillimanite to andalusite. Table 6. Comparisons between the calculated mineral compositions and modal proportions and the measured ones for samples A2121, A2115 and A367
P(kb ar) T(∼ C ) x ( g ) m x( y( ( m m g 米) 米) ) x( c hl ) y( c hl ) x ( b i) y ( b i) x g ( s t) m c b 米 h i l a l s st

Calcula ted for sample A2121 from

P(kb ar)

T(∼ C )

x ( g )

m x( y( ( m m g 米) 米) )

x( c hl )

y( c hl )

x ( b i)

y ( b i)

x g ( s t)

m c b 米 h i l

a l s

st

stauroli te每anda lusite zone A 8.50 626 每 每 0. 3 9 0. 8 9 每 每 0 . 4 2 0 . 0 4 0 . 4 3 0 . 0 1 3.20 545 每 每 每 每 0. 4 4 0. 6 3 0 . 5 3 0 . 0 1 0 . 6 1 0 . 3 8 0 . 0 8 0 . 4 7 0 . 0 6 0 . 5 8 0 . 0 7 0 . 6 3 0 . 8 9 0 . 8 0 0 . 0 2 0 . 8 1 0 . 0 1 每 0 . 0 3 0 . 5 1 每 每 0 . 4 9 8 . 3 4 0 . 5 3 0 . 0 0 0 . 5 0 0 . 0 0 0 . 0 4 0 . 1 8 0 . 1 4 0 . 2 1 0 . 0 3 0 . 3 6 0 . 0 1 0 . 2 0 0.30

Standar deviatio n





0. 0 3

0. 0 1









6.14

B

5.50

595

















0.26

Standar d deviatio n C







0.03





0 . 1 4 0 . 0 1 0 . 5 7



Standar d deviatio n D 3.40 539 每 每 0. 4 9 0. 9 5

0. 0 2

0. 0 2







0. 5 1

0. 6 6



0 . 4 1

0.02

P(kb ar)

T(∼ C )

x ( g ) 每

m x( y( ( m m g 米) 米) ) 每

x( c hl )

y( c hl )

x ( b i)

y ( b i)

x g ( s t) 每

m c b 米 h i l

a l s

st

Standar d deviatio n Measur ed

0. 0 4

0. 0 0

0. 0 6

0. 0 3

0 . 0 5 0 . 4 3

0 . 0 8 0 . 4 9

0 . 0 1 0 . 8 1

0 . 0 1 0 . 0 6

0 . 0 7 每

0 . 0 4 0 . 4 8

0 . 0 1 0 . 1 9

0.01





0. 6 2

0. 9 2







0.28

Calcula ted for sample A2115 from andalus ite每silli manite zone A 6.00 650 0 . 6 6 0 . 0 3 3.00 560 0 . 5 2 0 . 0 5 0 . 2 0 0 . 0 4 0 . 4 0 0 . 0 5 0. 4 7 0. 9 2 每 每 0 . 5 2 0 . 0 2 每 每 0 . 5 5 0 . 0 0 0 . 5 2 0 . 0 8 0 . 6 1 0 . 0 8 每 每 0 . 1 0 0 . 0 3 0 . 0 5 0 . 0 1 0 . 1 2 0 . 0 2 0 . 0 5 0 . 0 1 每 每 0 . 4 4 0 . 0 5 0 . 5 5 0 . 0 3 0 . 3 4 0 . 0 1 0 . 3 5 0 . 0 1 每 每

Standar d deviatio n B

0. 0 1

0. 0 0

0. 4 5

0. 9 5

Standar d deviatio n

0. 0 2

0. 0 0

P(kb ar)

T(∼ C )

x ( g )

m x( y( ( m m g 米) 米) )

x( c hl ) 每

y( c hl ) 每

x ( b i)

y ( b i)

x g ( s t) 每

m c b 米 h i l

a l s

st

Measur ed

0 . 7 2

0 . 1 8

0. 4 5

0. 9 1

0 . 5 1

0 . 4 4

0 . 0 9

0 . 1 3



0 . 4 4

0 . 3 4

Calcula ted for sample A367 from silliman ite zone 6.70 680 0 . 5 3 0 . 0 1 0 . 5 6 0 . 2 8 0 . 0 2 0 . 2 8 0. 3 6 0. 9 1 每 每 0 . 3 8 0 . 0 2 每 每 0 . 3 6 0 . 4 7 0 . 0 7 0 . 4 6 每 每 0 . 1 1 0 . 0 1 0 . 1 1 0 . 0 3 0 . 0 1 0 . 0 3 每 每 0 . 1 7 0 . 0 2 0 . 1 7 0 . 6 9 0 . 0 0 0 . 6 9

Standar d deviatio n Measur ed

0. 0 2

0. 0 0

0. 3 9

0. 9 2

Full-size table

Table 7. A comparison between the mineral compositions and modal proportions calculated at points A, B and C in Fig. 12b and those measured for sample A2152 selected from the garnet每cordierite zone

P(kba r) Calculat ed A 6.00

T(∼ C)

x(g m( ) g)

x(b i)

y(b i)

x(c d)

g

bi

cd

kf

sill

liq

780

0.6 2 0.0 1

0.1 3 0.0 1

0.4 2 0.0 3

0.4 4 0.0 8

0.3 0 0.0 2

0.3 7 0.0 4

0.1 2 0.0 5

0.2 5 0.0 1

0.0 4 0.0 0



0.2 2 0.0 2

Standar d deviatio n B 5.70 763



0.6 3 0.1 6

0.1 6 0.1 3

0.4 5 0.1 1

0.4 7 0.0 5

0.3 2 0.0 9

0.3 2 13. 4

0.2 3 10. 2

0.2 4 10. 8

0.0 2 0.3 3

0.0 2 0.6 7

0.1 7 1.9 2

Standar d deviatio n C 5.60 753

0.6 2 0.0 1

0.1 7 0.0 1

0.4 6 0.0 2

0.4 8 0.0 6

0.3 2 0.0 1

0.3 1 0.0 3

0.3 2 0.0 4

0.1 9 0.0 4



0.0 4 0.0 1

0.1 3 0.0 2

Standar d deviatio n Measur ed



0.6 1

0.1 7

0.4 4

0.4 2

0.2 0

0.3 6

0.3 0

0.2 0

0.0 9

0.0 6



5.6. Andalusite每sillimanite zone In the KMnFMASH P每T pseudosection calculated for sample A2115 from the andalusite每sillimanite zone (Fig. 11b), the observed mineral assemblage involving garnet, muscovite, biotite, sillimanite, andalusite and quartz (+ plagioclase) with two aluminosilicate polymorphs would be stable near point B (Fig. 11b), at the andalusite每sillimanite transition. However, the mineral compositions and modal proportions predicted for point B are different to the measured values (Table 6).

Isopleths reflecting the measured mole proportions for the main minerals garnet, biotite, muscovite and aluminosilicate are consistent with a narrow pressure range of 5.5每6.5 kbar with a 1 ? 考 uncertainty between 0.5每0.7 kbar, and a wide temperature range from 600每700 ∼ for this rock. Application of the garnet每biotite C geothermometer of Holdaway (2000) to this sample returns T > 650 ∼ for P = 6 kbar. C The presence of andalusite is consistent with the rock having experienced decompression from the sillimanite to andalusite fields as shown by the change from point A to B (the thick solid arrow in Fig. 11b). Isopleths of H2O-content do not assist in discriminating between P每T vectors due to there being minimal change in H2O content of the relevant equilibria. The mineral compositions and modal proportions predicted for this rock at 6 kbar and 650 ∼ lie considerably closer to those measured C in sample A2115 (Table 6), suggesting that the change in conditions from the sillimanite to andalusite fields had limited effect on peak mineral compositions. The garnet compositions preserved in zoned grains in sample 2115 (Fig. 5) do not correlate with isopleths of x(g) and m(g) predicted for path AB in Fig. 11b. Observed garnet core to rim zoning involves m(g) decreasing from 0.18每0.15, which most probably lies in the divariant field involving garnet, muscovite, biotite, staurolite and kyanite. These relationships suggest that the mineral assemblage observed in sample A2115 may have evolved from higher pressure conditions and is not well equilibrated. 5.7. Sillimanite zone In the KMnFMASH P每T pseudosection calculated for sample A367 from the sillimanite zone (Fig. 12a), the observed mineral assemblage involving garnet, muscovite, biotite, sillimanite and quartz (+ plagioclase) is trivariant and stable over a wide P每T field. Isopleths of the measured mole proportions for the four main minerals garnet, biotite, muscovite and sillimanite in sample A367 are consistent with a restricted pressure range from 6.3每6.8 kbar with a 1 ? 考 uncertainty of 0.4每0.6 kbar and an approximate temperature range from 650每690 ∼ The predicted mineral C.

compositions and modal proportions at the average condition of 6.7 kbar and 680 ∼ C closely match the values measured from sample A367, except that the garnet rim compositions have higher x(g) values (Table 6). Isopleths of H2O-content for the mineral assemblages in Fig. 12a indicate little change in H2O-content along the inferred decompressional path from peak conditions, conditions favorable for preservation of the peak assemblage. Isopleths of garnet compositions in Fig. 12a show that the analyzed garnet compositions with x(g) = 0.51/m(g) = 0.34 at core and x(g) = 0.56/m(g) = 0.28 at rim equate with P每T conditions at points A and B with temperature uncertainties of 7每9 ∼ on one-考 level, respectively in the divariant g每ctd每mu每chl每ky and trivariant C g每mu每chl每ky fields at temperatures below biotite stability. These garnet compositions record prograde conditions at substantially lower grade than the metamorphic peak. The isopleth of m(g) = 0.28 is consistent with the inferred peak P每T conditions, but the isopleth of x(g) = 0.56 is not. As a consequence, the mineral assemblage in sample A367 is inferred to have not fully equilibrated. 5.8. Garnet每cordierite zone In the KMnFMASH pseudosection calculated for sample A2152 from the garnet每sillimanite zone (Fig. 12b), the observed mineral assemblage garnet, biotite, cordierite, K-feldspar, sillimanite and quartz (+ plagioclase) with the presence of melt inferred on the basis of migmatitic textures, is divariant and stable over a narrow P每T range at 4.5每7.5 kbar and 720每820 ∼ This corresponds to the biotite dehydration C. melting reaction bi + sill = g + cd + ksp + melt in KFMASH (White et al., 2001). Unlike the other samples from the lower metamorphic zones, it is difficult to estimate P每T conditions from measured modal proportions as the rock was probably exposed to open system processes involving melt gain and/or loss. Calculations on the mineral compositions and mole proportions at three inferred P每T conditions (points A, B and C in Fig. 12b; Table 6) indicate that garnet and cordierite modes decrease whereas biotite and sillimanite modes increase as melt crystallizes across the narrow divariant

field related to the biotite dehydration melting reaction. When contrasted with the measured mineral compositions and mole proportions for sample A2152, we proposed that the observed mineral assemblage would have been formed in a limited cooling process such as from A to C (Fig. 12b) rather than representing an equilibrated divariant assemblage. Thus, the observed mineral assemblage was inferred to have formed over a range of P每T conditions at P = 5.5每6.0 kbar and T = 750每780 ∼ The C. observed garnet rim compositions x(g) = 0.61 and m(g) = 0.17 are consistent with this P每T range.

6. Discussion
The metamorphic zonal sequence from the Altai orogen (Fig. 1a/b) mostly involves a progression from biotite, through garnet, staurolite, staurolite每andalusite, andalusite每sillimanite, to sillimanite, and locally, garnet每cordierite zones, if the spatially-restricted kyanite zone is ignored. Such a sequence matches the facies series 2b of Pattison and Tracy (1991), which is characteristic of: (i) the coexistence of staurolite and andalusite in low to medium grade; (ii) sillimanite occurring at lower grade than muscovite + quartz breakdown; (iii) cordierite being generally at high grade; and (iv) the local development of kyanite with uncertain significance. There are numerous regional examples of facies series type 2b around the world (Pattison and Tracy, 1991), many of which are characterised by a P每T array involving isobaric heating. For example, a low-P metamorphic zonal sequence described by Grae? ner and Schenk (1999) in the Hercynian Aspromonte massif, southern Calabria, involves a sequence of chlorite, biotite, garnet, staurolite每andalusite, and sillimanite每muscovite zones linked to a P每T array of nearly isobaric heating at P = 1.5每2.5 kbar. However, the application of the conventional garnet每plagioclase每muscovite每biotite barometer (Hodges and Crowley, 1985) to mineral assemblages in these rocks yielded pressure estimates of 7每8 kbar for the garnet zone, and 4每5 kbar for the staurolite每andalusite zone. Grae? and Schenk (1999) interpreted these higher pressure results as ner reflecting disequilibrium with respect to CaO partitioning between coexisting garnet

and plagioclase. In the Sierra Albarrana area, Variscan belt, Iberian Massif, a similar zonal sequence described by Azor and Ballevre (1997) was also proposed to have an isobaric heating P每T path at P = 3.5每4 kbar, although the relic character of kyanite in the staurolite每andalusite and sillimanite zones in the Sierra Albarrana may represent prograde relics. A summary of the P每T path inferred for each metamorphic zone from the Altai orogen is shown in Fig. 13. Peak conditions experienced by the biotite, garnet, staurolite and kyanite zones construct a P每T array with a geothermal gradient of 21每26 ∼ C/km, typical of the kyanite-type facies series ([Miyashiro, 1973] and [Miyashiro, 1994]). Only trivial changes are interpreted to have affected the peak assemblages of rocks forming the biotite, garnet and staurolite zones during their exhumation, whereas the polymorphic transformation of kyanite to andalusite clearly affected peak assemblages in the kyanite zone. This resulted in the kyanite-bearing assemblages from the Altai orogen, in the most cases, only occurring as relic patches in the more extensive staurolite每andalusite zone (Zhuang, 1994). We prefer an interpretation whereby the staurolite每andalusite zone in the Altai orogen, represents the poorly equilibrated overprint of andalusite-bearing equilibria on kyanite-bearing assemblages in the kyanite zone. The coexistence of staurolite and andalusite widely reported in orogenic belts and also in contact aureoles has been considered as a low-pressure assemblage in the andalusite-stability field ([Pattison and Tracy, 1991], [Dymoke and Sandiford, 1992], [Azor and Ballevre, 1997], [Pattison et al., 1999] and [Grae? and Schenk, 1999]). On the basis of mineral equilibria modeling involving ner KFMASH P每T pseudosections, Dymoke and Sandiford (1992) established that the zonal sequence with the andalusite每staurolite paragenesis in the Mount Lofty Ranges, South Australia, and in the classic Buchan sequences documented by Harte and Hudson (1979) and Hudson (1980) in the eastern Scottish Highlands, can be predicted for bulk rock compositions with XFe = 0.7每0.8 in. As shown in Table 5, the XFe in the metapelites from the Altai orogen ranges from 0.29每0.58, and the staurolite每andalusite paragenesis is not predicted in the P每T pseudosections. This

reminds us that we should be cautious in assuming equilibrium characteristics in LP rocks that may have experienced a decompression of only a few kilobars that may or may not be clear in textural relations.

Full-size image (112K) Fig. 13. Summary of P每T path of each metamorphic zone from the Altai orogen in a KMnFMASH P每T pseudosection for a typical metapelite used by Mahar et al. (1997). The densely dashed lines around the point i1 show the NKASH phase relations calculated by White et al. (2001). The five univariant reactions in the NKASH are: (1) ab + mu + ksp + q + H2O = liq; (2) ab + mu + q + H2O = sill + liq; ab + mu + q = ksp + sill + liq; ab + ksp + sill + q + H2O = liq; and (5) ab + mu + q = ksp + sill + H2O. Isopleth details and line labeling as listed for Fig. 8.

Peak conditions recorded by rocks in the andalusite每sillimanite and sillimanite zones construct a P每T array with a geothermal gradient of 28每32 ∼ C/km, which is different from that in the kyanite type, though the sillimanite zone has been generally considered as a member of the zonal sequence of kyanite type. Field and textural relations indicate that the mineral assemblages in the sillimanite zone probably had higher pressure precursors. As shown in Fig. 13, the metamorphic characteristic of the appearance of sillimanite takes place at P每T conditions above the minimum solidus in the NKASH, which, as discussed by White et al. (2001), controls the solidus relations in more complicated systems. It is uncertain what role the melt played in assisting the formation of sillimanite-bearing assemblages from a kyanite-dominated precursor.

Peak conditions of rocks in the garnet每cordierite zone establish a P每T array with a geothermal gradient about 38 ∼ C/km, quite distinct to that of the former two types. Such high temperature (sillimanite-type) metamorphic conditions may have resulted in considerable partial melting in fertile lithologies (up to 30 vol. %, [Spear et al., 1999] and [White et al., 2001]) and belong to the granulite facies (Winkler, 1979). The typical low pressure metamorphism ([Miyashiro, 1973] and [Miyashiro, 1994]) is represented by the andalusite overprinting on the earlier higher pressure types with a P每T array along a geothermal gradient about 52 ∼ C/km.

7. Conclusions
Metamorphic zones in the Chinese Altai orogen record a series of P每T and mineral assemblage changes along distinct P每T arrays due to the effects of progressive metamorphism in a typical burial and exhumation P每T path; they do not represent prograde evolution along unique P每T arrays and preserve complexity that invalidates their simplistic division into kyanite and andalusite types. It seems probable that all metamorphic zones originally involved prograde assemblages of the kyanite type. Rocks from the lower grade biotite, garnet and staurolite zones mostly preserve peak mineral assemblages of the typical kyanite type, and mostly experienced trivial recrystallization during exhumation. Rocks in the higher grade zones, including the sillimanite and garnet每cordierite zones, are dominated by transitional assemblages formed during the exhumation of higher pressure assemblages. The clear low-pressure metamorphism characteristic of andalusite overprinting on kyanite- and sillimanite-type assemblages only occurred in intermediate zones, and involves poorly-equilibrated assemblages.

Acknowledgements
This work was financially supported by the National Natural Science Foundation of China (Grant numbers: 40525006 and 40372032). We are most grateful to the two anonymous reviewers for helpful reviews of the manuscript.

References
Azor and Ballevre, 1997 A. Azor and M. Ballevre, Low-pressure metamorphism in the Sierra Albarrana Area (Variscan belt, Iberian Massif), Journal of Petrology 38 (1997), pp. 35每64. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (10) Carson et al., 1999 C.J. Carson, R. Powell and G.L. Clarke, Calculated mineral equilibria for eclogites in CaO每Na2O每FeO每MgO每Al2O3每SiO2每H2O: application to the Pou谷 Terrane, Pam Peninsula, New Caledonia, Journal of Metamorphic bo Geology 17 (1999), pp. 9每24. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (61) Chang et al., 1995 E.Z. Chang, R.G. Coleman and D.X. Ying, Tectonic Transect Map Across Russia每Mongolia每China, Stanford University Press, Stanford, CA (1995). Chen and Jahn, 2002 B. Chen and B.-M. Jahn, Geochemical and isotopic studies of the sedimentary and granitic rocks of the Altai orogen of NW China and their tectonic implications, Geological Magazine 139 (2002), pp. 1每13. View Record in Scopus | Cited By in Scopus (54) Clarke et al., in 2006 G.L. Clarke, R. Powell and J.A. Fitsherbert, The lawsonite paradox: a comparison of field evidence and mineral equilibria modeling, Journal of Metamorphic Geology 24 (2006), pp. 716每726. Dempster, 1985 T.J. Dempster, Garnet zoning and metamorphism of the Barrovian type area, Scotland, Contributions to Mineralogy and Petrology 89 (1985), pp. 30每38. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (19) De Yoreo et al., 1989 J.J. De Yoreo, D.R. Lux, C.V. Guidotti, E.R. Decker and P.H. Osberg, The Acadian thermal history of western Maine, Journal of Metamorphic Geology 7 (1989), pp. 169每190. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (32) De Yoreo et al., 1991 J.J. De Yoreo, D.R. Lux and C.V. Guidotti, Thermal modelling in low-pressure/high-temperature metamorphic belts, Tectonophysics 188 (1991), pp. 209每238. Abstract | View Record in Scopus | Cited By in Scopus (83)

Dymoke and Sandiford, 1992 P. Dymoke and M. Sandiford, Phase relationships in Buchan facies series pelitic assemblages: calculations with application to andalusite每staurolite parageneses in the Mount Lofty Ranges, South Australia, Contributions to Mineralogy and Petrology 110 (1992), pp. 121每132. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (29) Ernst, 1973 W.G. Ernst, Blueschist metamorphism and P每T regimes in active subduction zones, Tectonophysics 17 (1973), pp. 255每272. Abstract | View Record in Scopus | Cited By in Scopus (14) Ernst and Liou, 2000 W.G. Ernst and J.G. Liou, Overview of UHP metamorphism and tectonics in well-studied collisional orogens. In: W.G. Ernst and J.G. Liou, Editors, Ultra-High Pressure Metamorphism and Geodynamics in Collision-type Orogenic Belts, Bellwether Publishing, Ltd., Columbia, USA (2000), pp. 3每19. Evans, 2004 T.P. Evans, A method for calculating effective bulk composition modification due to crystal fractionation in garnet-bearing schist: implications for isopleth thermobarometry, Journal of Metamorphic Geology 22 (2004), pp. 547每557. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (29) Grae? and Schenk, 1999 T. Grae? and V. Schenk, Low-pressure metamorphism ner ner of Paleozoic pelites in the Aspromonte, southern Calabria: constraints for the thermal evolution in the Calabrian crustal cross-section during the Hercynian orogeny, Journal of Metamorphic Geology 17 (1999), pp. 157每172. Guiraud et al., 2001 M. Guiraud, R. Powell and G. Rebay, H2O in metamorphism and unexpected behaviour in the preservation of metamorphic mineral assemblages, Journal of Metamorphic Geology 19 (2001), pp. 445每454. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (55) Harte and Hudson, 1979 B. Harte and N.F.C. Hudson, Pelitic facies series and the temperatures and pressure of Dalradian metamorphism in eastern Scotland. In: A.L. Harris, C.H. Holland and B.E. Leake, Editors, The Caledonides of the British Isles Reviewed, Special Publication-Geological Society of London vol. 8 (1979), pp. 323每337.

He et al., 1990 G.Q. He, B.F. Han, Y.J. Yue and J. Wang, Tectonic division and crustal evolution of Altai orogenic belt in China, Geosciences of Xinjiang 2 (1990), pp. 9每20 (in Chinese with English abstract). Hodges and Crowley, 1985 K.V. Hodges and P.D. Crowley, Error estimation and empirical geothermobarometry for pelitic systems, American Mineralogist 70 (1985), pp. 702每709. View Record in Scopus | Cited By in Scopus (145) Holdaway, 2000 M.J. Holdaway, Application of new experimental and garnet Margules data to the garnet每biotite geothermometer, American Mineralogist 85 (2000), pp. 881每892. View Record in Scopus | Cited By in Scopus (86) Holland and Powell, 1998 T.J.B. Holland and R. Powell, An internally consistent thermodynamic data set for phases of petrological interest, Journal of Metamorphic Geology 16 (1998), pp. 309每343. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (1167) Hu et al., 2002 A.Q. Hu, G.X. Zhang, Q.F. Zhang, T.D. Li and J.B. Zhang, A review on ages of Precambrian metamorphic rocks from Altai orogen in Xinjiang, NW China, Chinese Journal of Geology 37 (2002), pp. 129每142 (in Chinese with English abstract). View Record in Scopus | Cited By in Scopus (13) Hudson, 1980 N.F.C. Hudson, Regional metamorphism of some Dalradian pelites in the Buchan Area, N.E. Scotland, Contributions to Mineralogy and Petrology 73 (1980), pp. 39每51. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (15) Li et al., 1996 T.D. Li, Z.M. Qi and B.Q. Wu et al., New improvement of comparative study of geology and mineralization of Altai between China and Kazakhstan. In: Chinese Geological Society, Editor, Thesis Volume of the Symposium of the 8th Five-Year Plan of Geoscience for Contribution to 30th IGC, Metallurgical Industrial Publishing House, Beijing (1996), pp. 256每259. Mahar et al., 1997 E.M. Mahar, J.M. Baker, R. Powell, N. Holland and T.J.B. Howell, The effect of Mn on mineral stability in metapelites, Journal of Metamorphic Geology 15 (1997), pp. 223每238. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (93)

Marmo et al., 2002 B. Marmo, G.L. Clarke and R. Powell, Fractionation of bulk rock composition due to porphyroblast growth: effects on eclogite facies mineral equilibria, Pam Peninsula, New Caledonia, Journal of Metamorphic Geology 20 (2002), pp. 151每165. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (58) Miyashiro, 1973 A. Miyashiro, Metamorphism and Metamorphic Belts, George Allen & Unwin, London (1973) 492 pp.. Miyashiro, 1994 A. Miyashiro, Metamorphic Petrology, UCL Press Limited, London (1994), pp. 198每237. Pitra and De Waal, 2001 P. Pitra and S.A. De Waal, High-temperature, low-pressure metamorphism and development of prograde symplectites, Marble Hall Fragment, Bushveld Complex(south Africa), Journal of Metamorphic Geology 19 (2001), pp. 311每325. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (13) Pattison and Tracy, 1991 D.R.M. Pattison and R.J. Tracy, Phase equilibria and thermobarometry of metapelites. In: D.M. Kerrick, Editor, Contact Metamorphism, Reviews in Mineralogy vol. 26, Mineralogical Society of America, Washington, DC (1991), pp. 105每206. Pattison et al., 1999 D.R.M. Pattison, F.S. Spear and J.T. Cheney, Polymetamorphic origin of muscovite + cordierite + staurolite + biotite assemblages: implications for the metapelitic petrogenetic grid and for P每T paths, Journal of Metamorphic Geology 17 (1999), pp. 685每703. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (26) Powell and Holland, 1990 R. Powell and T.J.B. Holland, Calculated mineral equilibria in the pelite system KFMASH (K2O每FeO每MgO每Al2O3每SiO2每H2O), American Mineralogist 75 (1990), pp. 367每380. View Record in Scopus | Cited By in Scopus (112) Powell et al., 1998 R. Powell, T.J.B. Holland and B. Worley, Calculating phase diagrams involving solid solutions via non-linear equations, with examples using Thermocalc, Journal of Metamorphic Geology 16 (1998), pp. 577每588.

Spear et al., 1999 F.S. Spear, M.J. Kohn and J.T. Cheney, P每T paths from anatectic pelites, Contributions to Mineralogy and Petrology 134 (1999), pp. 17每32. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (197) Tinkham and Ghent, 2005 D.K. Tinkham and E.D. Ghent, Estimating P每T conditions of garnet growth with isochemical phase-diagram sections and the problem of effective bulk-composition, Canadian Mineralogist 43 (2005), pp. 35每50. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (13) Tracy, 1982 R.J. Tracy, Compositional zoning and inclusions in metamorphic minerals, Reviews in Mineralogy 10 (1982), pp. 355每394. Vance and Holland, 1993 D. Vance and T.J.B. Holland, A detailed isotopic and petrological study of a single garnet from the Gassetts Schist, Vermont, Contributions to Mineralogy and Petrology 114 (1993), pp. 101每118. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (70) Vernon et al., 1990 R.H. Vernon, G.L. Clarke and W.J. Collins, Local, mid-crustal granulite facies metamorphism and melting: an example in the Mt Stafford area, central Australia. In: J.R. Ashworth and M. Brown, Editors, High Temperature Metamorphism and Crustal Anatexis, Mineralogical Society Special Publications (1990), pp. 272每319. Waters and Charnley, 2002 D.J. Waters and N.R. Charnley, Local equilibrium in polymetamorphic gneiss and the titanium substitution in biotite, American Mineralogist 87 (2002), pp. 383每396. View Record in Scopus | Cited By in Scopus (18) Wei et al., 2003 C.J. Wei, R. Powell and L.F. Zhang, Eclogites from the south Tianshan, NW China: petrologic characteristic and calculated mineral equilibria in the Na2O每CaO每FeO每MgO每Al2O3每SiO2每H2O system, Journal of Metamorphic Geology 21 (2003), pp. 163每179. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (57) Wei et al., 2004 C.J. Wei, R. Powell and G.L. Clarke, Calculated phase equilibria for low- and medium-pressure metapelites in the KFMASH and KMnFMASH systems,

Journal of Metamorphic Geology 22 (2004), pp. 495每508. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (25) White et al., 2000 R.W. White, R. Powell, T.J.B. Holland and B. Worley, The effect of TiO2 and Fe2O3 on metapelitic assemblages at greenschist and amphibolite facies conditions: mineral equilibria calculations in the system K2O每FeO每MgO每Al2O3每SiO2每H2O每TiO2每Fe2O3, Journal of Metamorphic Geology 18 (2000), pp. 497每511. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (80) White et al., 2001 R.W. White, R. Powell and T.J.B. Holland, Calculation of partial melting equilibria in the system Na2O每CaO每K2O每FeO每MgO每Al2O3每SiO2每H2O (NCKFMASH), Journal of Metamorphic Geology 19 (2001), pp. 139每153. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (173) White et al., 2003 R.W. White, R. Powell and G.L. Clarke, Prograde metamorphic assemblage evolution during partial melting of metasedimentary rocks at low pressures: migmatites from Mt Stafford, Central Australia, Journal of Petrology 44 (2003), pp. 1937每1960. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (37) Windley et al., 2002 B.F. Windley, A. Kr? ner, J.H. Guo, G.S. Qu, Y.Y. Li and C. Zhang, Neoproterozoic to Paleozoic geology of the Altai Orogen, NW China: new zircon age data and tectonic evolution, The Journal of Geology 110 (2002), pp. 719每737. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (99) Winkler, 1979 H.G.F. Winkler, Petrogenesis of Metamorphic Rocks (5th edition), Springer-Verlag, New York (1979), pp. 256每275. Worley and Powell, 1998 B. Worley and R. Powell, Singularities in NCKFMASH (Na2O每CaO每K2O每FeO每MgO每Al2O3每SiO2每H2O), Journal of Metamorphic Geology 16 (1998), pp. 169每188. View Record in Scopus | Cited By in Scopus (23) Xiao et al., 1992 X.C. Xiao, Y.Q. Tang, Y.M. Feng, B.Q. Zhu, J.Y. Li and M. Zhao, Tectonics in Northern Xinjiang and its Neighboring Areas, Geological Publishing House, Beijing (1992) 169 pp., in Chinese with English abstract.

Xu et al., 1994 G.W. Xu, T.M. Will and R. Powell, A calculated petrogenetic grid for the system K2O每FeO每MgO每Al2O3每SiO2每H2O, with particular reference to contact-metamorphosed pelites, Journal of Metamorphic Geology 12 (1994), pp. 99每119. Full Text via CrossRef Xu et al., 2003 J.F. Xu, P.R. Castillob, F.R. Chen, H.C. Niu, X.Y. Yu and Z.P. Zhen, Geochemistry of late Paleozoic mafic igneous rocks from the Kuerti area, Xinjiang, northwest China: implications for backarc mantle evolution, Chemical Geology 193 (2003), pp. 137每154. Zhang et al., 1996 X.B. Zhang, J.X. Sui, Z. Li and W. Liu, Tectonic Evolution of the Irtysh Zone and Metallogenesis, Science Press, Beijing (1996), pp. 89每91 In Chinese with English Abstract. Zhuang, 1994 Y.X. Zhuang, Tectonothermal Evolution in Space and Time and Orogenic Process of Altaide, China, Jilin Scientific and Technical Press, Changchun, China (1994) 402 pp., In Chinese with English abstract. Zou et al., 1988 T.R. Zou, H.Z. Cao and B.Q. Wu, Orogenic and anorogenic granitoids of the Altay mountains, Xinjiang and their discrimination criteria, Acta Geologica Sinica 62 (1988), pp. 228每245 (in Chinese with English abstract).

Corresponding author. Tel.: +86 10 62754157; fax: +86 10 62751159.


眈壽芢熱

郔陔載陔

笨斕炰辣