Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.geol.msu.ru/deps/petro/Perchuk_publ/Perchuk_Petrol02.pdf
Дата изменения: Wed Feb 15 16:08:10 2012
Дата индексирования: Tue Oct 2 01:53:35 2012
Кодировка: koi8-r

Petrology, Vol. 10, No. 2, 2002, pp. 99118. Translated from Petrologiya, Vol. 10, No. 2, 2002, pp. 115137. Original Russian Text Copyright © 2002 by Perchuk. English Translation Copyright © 2002 by ЕдIД "Nauka / Interperiodica" (Russia).

Eclogites of the Bergen Arcs Complex, Norway: Petrology and Mineral Chronometry
A. L. Perchuk
Institute of Geology of Ore Deposits, Petrography, Mineralogy, and Geochemistry (IGEM), Russian Academy of Sciences, Staromonetnyi per. 35, Moscow, 109017 Russia e-mail: alp@igem.ru
Received March 16, 2001

Abstract--A considerable portion of the Bergen Arcs polymetamorphic complex (southwestern Norway) is composed of anorthosite rocks strongly reworked in the course of two regional metamorphic events, granulite (Grenville time) and eclogite (Caledonian time). Eclogitization is manifested only in the areas of hydrous fluid penetration. They are represented mainly by plastic deformation zones and, occasionally, nondeformed granulites. The investigation of mineral equilibria demonstrated moderate water activity in the fluid ( a H2 O < 0.4) at T = 730°C and Pmin = 19 kbar at the peak of eclogitic metamorphism. Under such conditions, anatexis does not occur in basic rocks. Local retrograde transformations were estimated at ~500°C and 6 kbar (epidote amphibolite facies). Garnet crystals in the eclogite show a zonal structure: magnesium-rich cores formed at granulite metamorphism are overgrown by thin (~200 µm) iron-rich rims of the eclogitic stage. Some of the cores contain stringers, microscopic cracks healed by "eclogitic" garnet, showing sharp transitions to the enclosing "granulitic" garnet. Simulation demonstrated that the diffusion process at the stringerenclosing garnet boundary continued for 0.8 m.y. and terminated when eclogite cooled down to a temperature of 600°C and diffusion mass transfer became inefficient. At linear temperature and pressure variations, this corresponds to a cooling rate of 160°C/m.y. and an exhumation velocity of 3 cm/y. The calculated temperature of Ar system closure in amphibole at such a cooling rate is 510590°C, i.e., ArAr dating of amphibole is relevant to the temperature interval of a retrograde stage. It was assumed that the peak of the eclogite stage at a decompression duration of 0.8 m.y. corresponded to an ArAr age of the amphibole of 448455 Ma.

INTRODUCTION The process of continental plate interaction at convergent boundaries (collision) has given rise to many puzzles. Some researchers believe that the subduction of a thick continental crust is impossible because of the lower density (buoyancy) of the continental lithosphere with respect to the asthenosphere (Cloos, 1982 and reference wherein). On the other hand, studies of ultrahigh-pressure complexes (Sobolev and Shatskii, 1989, 1990; Smith, 1984; Chopin, 1987; Zhang et al., 1995; When Continents..., 1998) demonstrated that continental margins might descent to depths of 90 km and more, where coesite, diamond, and other ultrahigh-pressure phases and associations were stable. A problem arises of the mechanism of ascent toward the surface of large continental blocks (Schreyer, 1995). This problem is not yet solved despite the diversity of related geotectonic models (see Dobretsov and Kirdyashkin, 1994 and Maruyama et al., 1996 for a review). Advances in methods of the computer simulation of geodynamic processes (Schmeling et al., 1999; Ellis et al., 1999; Burg and Podladchikov, 2000) allow us to develop and test models of the burial and exhumation of high-pressure complexes by means of numeric experiments. Of prime importance in this work could be the qualitative information on PT paths (Korikovsky,
99

1995; Perchuk et al., 1996; Gerya et al., 2000) and rates of the thermal evolution and vertical movement of rocks in the lithosphere (Perchuk and Philippot, 2000 and reference wherein), which is obtained in the course of petrological studies of natural samples. The goal of this study is to determine the rates of the temperature and pressure evolution of eclogite from the Bergen Arcs polymetamorphic complex in southwestern Norway, whose formation and exhumation were related to the Caledonian collision (Boundy et al., 1996, 1997). These data will be used for the correct interpretation of the ArAr age of amphibole (Boundy et al., 1996) and determination of the absolute age of eclogitic metamorphism. GEOLOGIC BACKGROUND AND MAIN STAGES OF METAMORPHISM (REVIEW) The polymetamorphic complexes of SW Norway crop out over an area of tens of thousands of square kilometers and are a unique testing ground for the investigation of a wide spectrum of issues related to the high-pressure metamorphism and geodynamics of regions of lithospheric plate convergence and divergence. In the Caledonian history of the region, there is a stage of large-scale continental collision (Laurentia

100

PERCHUK

and Baltia), which resulted in orogeny and crust thickening (Krogh, 1977). The subsequent global change of geodynamic setting (from compression to extension) led to the collapse of the Caledonian orogenic belt. These processes can be conveniently studied using two adjacent high-pressure blocks, the Western Gneiss Region and Bergen Arcs complex. The former is marginal part of the Baltic shield in an autochthonous position. Its rising occurred in an extensional environment (Boundy et al., 1996; Andersen, 1997). The second complex is, in contrast, allochthonous and was moved during the Caledonian collision epoch into the middle horizons of the Earth's crust (Boundy et al., 1996). Rocks that experienced eclogitic-facies metamorphism were found both in the autochthonous and allochthonous parts of the section (Krogh and Carswell, 1995). The Bergen Arcs polymetamorphic complex (Fig. 1), which is the major topic of this study, is structurally represented by a series of arclike nappes thrust over the rocks of the Western Gneiss Region (Roberts and Gee, 1985). The contact between the complexes is major-shear zone composed of eclogite-grade metamorphic rocks. The rocks of the Bergen Arcs complex are overlain with a structural discontinuity by the Middle Devonian conglomerates of the Frensfjorden basin (Wenberg and Milnes, 1994). In the western portion, the complex is cut by the Krossnes granite intrusion with a RbSr isotopic age of 430 ± 6 Ma (Fossen and Ingdahl, 1988). The protolith of the Bergen Arcs complex was dominated by the metamorphosed rocks of the anorthosite complex, which ranged in composition from anorthosite to gabbro (Austrheim and Griffin, 1985). Lenslike bodies of spinel lherzolite occur occasionally among the meta-anorthosites. In some places, metaanorthosite associates with charnockite and mangerite. The complex was emplaced at 1000°C and a maximum pressure of ~9 kbar (Austrheim and Griffin, 1985) during the Grenville orogeny (Cohen et al., 1988; Austrheim and Mork, 1988). Isotopic dating yielded age values in the range 1300870 Ma (Scharer, 1980). The scatter is probably related to the partial disturbance of the isotopic system by thermal events, which were numerous in the history of the region. In particular, the Bergen Arcs complex experienced three major metamorphic events (Table 1), whose extent declined with time. (1) At the earliest evolution stage, the anorthosite complex was affected by global recrystallization under granulite-facies conditions (T = 800850°C and P < 10 kbar). The newly formed mineral association included plagioclase + diopside (Al-rich) + garnet ± orthopyroxene ± scapolite ± hornblende ± spinel (Austrheim and Griffin, 1985; Cohen et al., 1988). The isotopic age estimates (UPb, SmNd, and RbSr methods) for this event lie within 896945 Ma (Cohen et al., 1988; Burton et al., 1995; Boundy et al., 1997) and overlap in part with the timing of the emplacement of

the anorthosite complex. The fluid regime of this metamorphic stage is not known in detail. Probably, the fluid was dominated by water and carbon dioxide, which is suggested by the presence of hornblende, which was reported in an earliest petrologic publication (Austrheim and Griffin, 1985), and findings of primary carbon dioxide inclusions in quartz veins of the granulite stage (Andersen et al., 1993). (2) During the Caledonian orogeny, the granulites experienced partial recrystallization under the conditions of the eclogite facies. The mineral formation of this stage was described in detail by Norwegian researchers (Austrheim and Griffin, 1985; Jamtveit et al., 1990; Erambert and Austrheim, 1993) and occurred only in the areas of fluid penetration. The thickness of such zones vary from a few centimeters to one kilometer. The presence of water in the fluid phase is indicated by the mineral composition of the rocks (garnet + omphacite + quartz + micas + amphibole + epidote + rutile) containing various hydrous minerals, which are lacking or scarce in the parental rock (granulite). According to Jamtveit et al. (1990), the eclogite assemblage formed at T ~ 700°C, P = 1821 kbar, water mole fraction up to 0.95, and was accompanied by significant heat release (~150 000 kJ per one cubic meter of initial granulite). According to other authors, the P T parameters of the eclogitic metamorphism peaked at T = 700800°C and P > 16 kbar (Austrheim and Griffin, 1985) or T = 670 ± 50°C and P > 14.6 kbar (Boundy et al., 1992). Recent isotopic work (UPb and SmNd methods, Boundy et al., 1997) yielded ages of 450 ± 10 Ma (zircon), 462 ± 42 Ma (allanite, sphene, and epidote), and 442 ± 12 Ma (garnetrock) for this event. Previously, the eclogite metamorphism was dated by the RbSr and SmNd methods at 507 ± 109 and 421 ± 29 Ma, respectively (Cohen et al., 1988). (3) The subsequent amphibolite-facies metamorphism imprinted both the granulites and eclogites. Retrograde plagioclase, biotite, amphibole, and sphene formed in the granulites (Austrheim and Robins, 1991) and plagioclase + clinopyroxene + amphibole + epidote ± biotite, in the eclogites. No special thermobarometric investigations were carried out on this mineral formation stage. The only value is a pressure of 8 ± 2 kbar estimated from fluid inclusions in minerals of a shear zone (Andersen et al., 1991b). In calculations, we assumed that the temperature of this stage was 600 ± 100°C. The ArAr dating of amphibole and micas yielded 448455 Ma and 429463 Ma, respectively (Boundy et al., 1996). The Bergen Arcs complex directly contacts the autochthonous Western Gneiss Region. The latter is made up of gneisses and supracrustal rocks gently dipping to the west. Eclogite outcrops may be encountered among most of the lithologies including paragneiss, orthogneiss, gabbro, anorthosite, and peridotite (Medaris and Wang, 1986; Bryhny, 1989; Wain et al., 2000; Krabbendam et al., 2000). The PT conditions of
PETROLOGY Vol. 10 No. 2 2002

ECLOGITES OF THE BERGEN ARCS COMPLEX (a) 1
e e e e

101

2 3 WGR 4 5 6 7
e

60°45

8 9

e ee e e

e

H

A'

A

e

e

BA N
Norway

60°15 0 km 5°00 (b) A
Caledonian nappes

10

N

5°30 A'
Caledonian nappes

ee Allochthon

e

e

e Baltic shield (autochthon)

Fig. 1. (a) Tectonic map of southwestern Norway (modified after Boundy et al., 1997) and (b) schematic cross-section along line AA' through the Bergen Arcs complex. BA, polymetamorphic Bergen Arcs complex; H, Holsnoy Island; and WGR, Western Gneiss Region. (1) Devonian sediments; (25) Caledonian nappes: (2) ophiolitic complex, (3) island-arc series, (4) Lindas nappe and (5) Bergsdalen nappes, (6) gneisses; (7) basement; (8) eclogite exposure; and (9) shear zone. The inset shows the position of the Bergen Arcs complex in the Scandinavian Caledonides.

eclogitic metamorphism increased gradually in the northwestern direction (Krogh, 1977) up to formation of ultrahigh-pressure coesite at the western margin of the complex (Salje region) (Smith, 1984).
PETROLOGY Vol. 10 No. 2 2002

The evolution of some eclogites from the Western Gneiss Region was conspicuous in heating at the initial stage of decompression. During this event, the rocks occasionally attained the PT conditions of granulitegrade metamorphism (Smith, 1995; Krogh and Car-

Cal edo nide s

Sweden

102

PERCHUK

Table 1. Major stages of anorthosite rock metamorphism in the Bergen Arcs complex, Norway after published data (see text for reference) Metamorphic stage Granulite Mineral association Pl + Di (Al-rich) + GrtMg ± Opx ± Scp ± Hbl ± Spl GrtFe + Omp + Qtz + Par + Phn + Amph + Ep + Rut PT conditions 800850°C, < 10 kbar 670800°C, 1421 kbar Isotopic age, Ma 896945 (UPb, SmNd and RbSr methods) 450 ± 10 (zircon, UPb method) 462 ± 42 (allanite, sphene, and epidote; UPb method) 507 ± 109 (RbSr method) 442 ± 12 (garnetrock, SmNd method) 421 ± 29 (garnet, omphacite, and epidote; SmNd method) 448455 (amphibole, ArAr method) 429463 (phengite, ArAr method)

Eclogite

Amphibolite

Pl + Amph ± Cpx ± Ky ± Tit ± Ep ± Bt

500700°C, 610 kbar

swell, 1995; Austrheim et al., 1997 and reference wherein). Interestingly, such PT metamorphic paths were never found in the eclogites of the Bergen Arcs complex. The age of eclogitic metamorphism in the Western Gneiss Region was determined by the SmNd and UPb isotopic methods as 400447 Ma (Griffin and Bruekner, 1980, 1985; Mork and Mearns, 1986; Gebauer et al., 1985). Recent dating of monazite by SHRIMP did not conclusively clarified the situation; inclusions in garnet yielded 400440 Ma, whereas monazite from the matrix showed 370440 Ma (Terry et al., 2000, Fig. 3). These authors demonstrated that the UThPb ages of monazite determined by the microprobe formed three broad peaks within the range 410370 Ma. In comparison with the large uncertainty in the timing of the eclogite stage, the ArAr dating of amphibole and phengite (retrograde stage) showed reproducible values of 410395 and 403395 Ma, respectively (Chauvet and Dallmeyer, 1992; Boundy et al., 1996). It is characteristic that these age values are ~30 m.y. lower than those determined for the minerals of the Bergen Arcs complex. PETROGRAPHIC DESCRIPTION The petrography and mineralogy of the Bergen Arcs eclogite were thoroughly described by Norwegian geologists (Austrheim and Griffin, 1985; Jamtveit et al., 1990; Erambert and Austrheim, 1993). In this study, we used a sample of eclogite from Holsnoy Island, which bears a record of PT a H2 O metamorphic evolution and contains heterogeneous garnet allowing the efficient use of mineral chronometry. The major rock-forming minerals of the eclogite are garnet, omphacite, kyanite, phengite, amphibole, epidote, and quartz. The rock has a heteroblastic texture with a fine-grained matrix enclosing relatively large (5 mm across) garnet crystals resorbed at margins. The garnet size was not due to the high crystallization

strength of the mineral in this case, and most of its volume was merely inherited from the crystals of an earlier generation. The detailed petrologic study of the eclogitization of the Bergen Arcs granulite (Austrheim and Griffin, 1985; Erambert and Austrheim, 1993) demonstrated that the early garnet was formed at the granulite stage. The chemical heterogeneity of the garnet is clearly seen in back-scattered electron images (obtained on a JSM-840A scanning electron microscope in Paris 7 University): the cores are composed of an early garnet generation of the granulite stage (Grt1), while eclogite-stage garnet (Grt2) occurs as mantles around the earlier crystals and linear zones (Fig. 2a), which are referred to as stringers following Cliff et al., (1998). Mineral inclusions in garnet are rather rare. The most common are isolated quartz grains and trails of omphacite crystals (sometimes in association with quartz and kyanite), which trace the central parts of some stringers (Fig. 2b). Thus, high-pressure omphacite always associates with eclogitic garnet. Weakly manifested rock schistosity is formed by the predominant orientation of grains of phengite, kyanite, and clinozoisite coexisting with the above-mentioned highpressure minerals. Retrograde changes affected most of the rocks. They are recorded in the formation of Hbl ±Pl ± Ep1 rims around garnet at the contact with omphacite, growth of thin biotite scales after phengite, and replacement of marginal parts of matrix omphacite by symplectites composed of varying amounts of plagioclase, clinopyroxene, and amphibole. Clinopyroxene and amphibole grains in the symplectites may be isolated (enclosed by plagioclase) or form thin intergrowth with each other (Fig. 3).
1

Mineral abbreviations are after Kretz (1983). PETROLOGY Vol. 10 No. 2 2002

ECLOGITES OF THE BERGEN ARCS COMPLEX () Cpx Hbl Qtz

103

symp

Grt

1

Grt

2

Omp

Phn 100 µm

100 µm

(b)

Fig. 3. Symplectites of Cpx + Hbl + Pl (symp) at the boundary between omphacite (Omp) and quartz (Qtz). Clinopyroxene (light gray) and hornblende (dark gray) often form intergrowth in plagioclase (black).

Mg, Fe, Mn, Ca, formula unit Grt
2

1.5

Mg

A' 1.0 Grt
1

Fe2+

A Ca 100 µm
2

Grt

0.5 Mn

Fig. 2. Back-scattered electron image of a portion of eclogite with garnet. (a) Concentrically zoned garnet (rounded crystal in the center) with a Mg-rich core (darker) formed at the granulite stage and Fe-rich rim (lighter) formed at the eclogite stage. The matrix contains abundant hydrous minerals. (b) Stringers of the eclogite stage manifested as light linear zones in garnet sometimes containing trails of omphacite inclusions, one of which is shown by the arrow. AA' is the microprobe profile line.

0

50

100

150 Distance, µm

Fig. 4. Distribution of bivalent cations in the microprobe profile through a stringer and adjacent garnet zones. The profile line (AA') is shown in Fig. 2b. Note the sharp transition from the stringer to host garnet and constant (not affected by diffusion) composition of the central part of the stringer.

COMPOSITIONS OF MAIN ROCK-FORMING MINERALS The chemical compositions of phases were determined on a CAMECA SX-50 electron microprobe at
PETROLOGY Vol. 10 No. 2 2002

Paris 6 University, CAMPARIS. The analyses were obtained at an accelerating voltage of 15 kV, a beam current of 10 nA, and an electron beam diameter of 1 µm. Natural and synthetic compounds were used as standards. The representative compositions of minerals

104

PERCHUK

Table 2. Microprobe analyses and crystal chemical formulas of garnet from the Bergen Arcs eclogite Component SiO2 TiO2 Al2O3 Cr2O3 FeO1 MnO MgO CaO Na2O K2O Total Si Al Cr Ti Fe3+ Fe2+ Mn Mg Ca Na K Total XMg XCa XAlm XSps XPrp XCrs 41.10 0.12 23.02 0.02 13.45 0.19 14.85 7.29 0.00 0.00 100.04 3.00 1.98 0.00 0.01 0.02 0.80 0.01 1.61 0.57 0.00 0.00 8.00 0.67 0.19 0.27 0.00 0.54 0.19 41.16 0.13 23.08 0.00 13.59 0.12 14.93 7.24 0.00 0.00 100.25 2.99 1.98 0.00 0.01 0.02 0.81 0.01 1.62 0.56 0.00 0.00 8.00 0.67 0.19 0.27 0.00 0.54 0.19 Garnet core 41.07 0.10 23.08 0.10 13.88 0.11 14.67 7.02 0.00 0.00 100.04 3.00 1.99 0.01 0.01 0.00 0.84 0.01 1.60 0.55 0.00 0.00 8.00 0.65 0.18 0.28 0.00 0.53 0.18 40.81 0.08 23.05 0.00 14.18 0.31 14.55 7.04 0.00 0.00 100.02 12 O 2.99 1.99 0.00 0.00 0.02 0.85 0.02 1.59 0.55 0.00 0.00 8.00 0.65 0.18 0.28 0.01 0.53 0.18 3.00 2.00 0.00 0.01 0.00 1.35 0.02 1.08 0.55 0.01 0.00 8.00 0.44 0.18 0.45 0.01 0.36 0.18 3.00 1.98 0.00 0.01 0.02 1.47 0.03 0.86 0.64 0.00 0.00 8.00 0.37 0.21 0.49 0.01 0.29 0.21 2.99 1.96 0.00 0.00 0.05 1.44 0.03 0.94 0.58 0.01 0.00 8.00 0.40 0.19 0.48 0.01 0.31 0.19 2.99 2.00 0.00 0.00 0.01 1.35 0.03 1.07 0.56 0.01 0.00 8.00 0.44 0.19 0.45 0.01 0.36 0.19 stringer 39.49 0.10 22.35 0.00 21.25 0.37 9.55 6.72 0.03 0.00 99.86 39.12 0.11 21.96 0.00 23.11 0.54 7.50 7.79 0.03 0.00 99.61 39.05 0.08 21.75 0.00 23.25 0.52 8.26 7.11 0.03 0.00 100.04 rim 39.36 0.04 22.34 0.00 21.31 0.46 9.45 6.83 0.04 0.02 99.83

Note: XMg = Mg/(Mg + Fe2+); XCa = Ca/(Ca + Mn + Mg + Fe2+). 1Hereafter, total iron in the FeO form.

and their crystal chemical formulas are given in Tables 2 4. The recalculation of minerals to crystal chemical formulas was carried out using the procedures described by Perchuk (1993) and Perchuk and Philippot (1997). Garnet grains are concentrically zoned: homogeneous cores (#Mg2 ~ 67 and #Ca3 ~ 19) are changed abruptly by thin (no more than 200 µm) rims of a constant composition (#Mg ~ 45 and #Ca ~ 19) (Table 2). The homogeneous cores are often disturbed by the
2 3

presence of stringers, whose contacts show the same changes as the core to rim transition (Fig. 4; Table 2). Thus, the composition of eclogitic garnet (Grt2) differs from that of granulite garnet (Grt1) in lower magnesium content at similar concentrations of other components. Remarkable is the absence of diffusion zoning at the contact of garnet with omphacite, which is typical of eclogite. The detailed study of mineral reactions demonstrated (Austrheim and Griffin, 1985) that the granulite garnet (Grt1) was never a reactant. This explains its preservation in the cores of garnet xenoblasts and the
PETROLOGY Vol. 10 No. 2 2002

#Mg = 100Mg/(Mg + Fe2+). #Ca = 100Ca/(Ca + Mn + Mg + Fe2+)

ECLOGITES OF THE BERGEN ARCS COMPLEX Table 3. Microprobe analyses and crystal chemical formulas of clinopyroxene from the Bergen Arcs eclogite Component SiO2 TiO2 Al2O3 Cr2O3 FeO MnO MgO CaO Na2O K2O Total Si Al Cr Ti Fe3+ Fe2+ Mn Mg Ca Na K Total XMg XJd Clinopyroxene crystals in stringers 56.44 0.13 13.33 0.06 4.02 0.04 7.41 12.10 7.49 0.02 101.03 1.98 0.55 0.00 0.00 0.00 0.12 0.00 0.39 0.45 0.51 0.00 4.00 0.77 0.51 55.68 0.12 12.99 0.04 4.03 0.08 7.66 13.17 6.84 0.04 100.63 1.96 0.54 0.00 0.00 0.00 0.12 0.00 0.40 0.50 0.47 0.00 4.00 0.77 0.47 55.98 0.11 13.90 0.04 3.70 0.00 6.97 12.20 7.57 0.02 100.47 1.97 0.58 0.00 0.00 0.00 0.11 0.00 0.37 0.46 0.52 0.00 4.00 0.77 0.52 56.44 0.13 13.33 0.06 4.02 0.04 7.41 12.10 7.49 0.02 101.03 1.98 0.55 0.00 0.00 0.00 0.12 0.00 0.39 0.45 0.51 0.00 4.00 0.77 0.51 matrix 55.68 0.12 12.99 0.04 4.03 0.08 7.66 13.17 6.84 0.04 100.63 1.96 0.54 0.00 0.00 0.00 0.12 0.00 0.40 0.50 0.47 0.00 4.00 0.77 0.47 6O 1.97 0.58 0.00 0.00 0.00 0.11 0.00 0.37 0.46 0.52 0.00 4.00 0.77 0.52 2.00 0.50 0.00 0.01 0.02 0.09 0.00 0.40 0.44 0.52 0.00 3.98 0.81 0.50 1.99 0.06 0.00 0.00 0.08 0.16 0.00 0.72 0.88 0.12 0.00 4.00 0.82 0.03 1.98 0.19 0.00 0.00 0.05 0.13 0.00 0.65 0.78 0.22 0.00 4.01 0.83 0.15 55.98 0.11 13.90 0.04 3.70 0.00 6.97 12.20 7.57 0.02 100.47 56.79 0.39 12.09 0.00 3.84 0.00 7.61 11.68 7.66 0.01 100.07 53.96 0.09 1.29 0.11 7.65 0.09 13.05 22.23 1.67 0.00 100.15 symplectite 54.25 0.13 4.41 0.00 5.97 0.16 11.86 19.90 3.13 0.02 99.83

105

54.62 0.05 4.33 0.00 5.22 0.10 12.65 20.23 3.04 0.00 100.24 1.98 0.18 0.00 0.00 0.05 0.11 0.00 0.68 0.79 0.21 0.00 4.01 0.86 0.14

sharp change in composition despite partial diffusion smoothing at the transition Grt1 Grt2. In fact, we deal with the discontinuous growth of garnet related to a transition from the granulite to eclogite facies. A similar example of discontinuous growth was reported for eclogite from the Sulu-Tyube segment of the Kokchetav massif (Perchuk et al., 1998), where amphibolite garnet even retained its ideomorphic shapes. Clinopyroxene occurs in the eclogite in three structural positions: in matrix, as mineral inclusions (in stringers), and in symplectites. The former two pyroxenes are high-pressure phases. According to the generally accepted nomenclature, they are omphacites with similar high contents of the jadeite end-member (4853 mol %) and low acmite content (05 mol %). The retrograde pyroxene from the Pl + Cpx ± Amph assemblage of symplectites after omphacite is signifiPETROLOGY Vol. 10 No. 2 2002

cantly depleted in jadeite, up to 3 mol % (Table 3, Fig. 5). Among hydrous minerals of the matrix, phengite and zoisite can be distinguished. The content of Si in phengite is about 3.23 f.u. It is also rich in paragonite molecule, Na/(Na + K) = 0.25 (Table 4). The zoisite composition is rather stable at ~4 mol % of pistacite. The symplectites contain abundant Ca amphibole of edenite-dominated composition (Table 4). PT EVOLUTION OF METAMORPHISM Metamorphic Peak (Eclogite Stage) In a well-known study on the petrology of the Bergen Arcs complex, Jamtveit et al. (1990) pointed out that GrtCpx geothermometry yielded too high temperature values (750°C and higher) for the peak of eclogite metamorphism. This conclusion relied on the lack of

106 A 1 2 3

PERCHUK P, kbar 26 a 22 0.5 0.4 0.3 0.2 Jd50 +K y+ HO Par 2
H2O
25

km

70
H2 O

J

18

B a sa lt so lidus a

1
AG

b hA

70

= 0.5

50

14

Jd

50

tz +Q
70

B

lAb

(H

2

O)

FM

C

10
Fluid

30
inclusions

Fig. 5. ACFM diagram for clinopyroxenes occurring in various structural positions in the rock. (1) Clinopyroxene inclusions in garnet stringers, (2) matrix omphacite, and (3) clinopyroxene from symplectites.

6

2
Ky d An

Jd3 + Qtz Sil 600 hAb
76

10

a
O H2

partial melting caused by the high temperature and H2O-rich metamorphic fluid ( X H2 O ~ 0.9, Jamtveit et al., 1990). Since the solidus temperature depends strongly on water activity in the fluid, we revised the determination of water activity using the following two mineral equilibria (Jamtveit et al., 1990): Dol + Qtz = Di (in omphacite) + CO2 (in fluid) (1) and Par (in mica) = Jd (in omphacite + Ky + H2O (in fluid). (2) The correctness of the former equilibrium application was questioned by Krogh and Carswell (1995) on the basis of dolomite isotopic composition, which suggested its superimposed character. However, primary dolomite associating with eclogite assemblage minerals was reported by Erambert and Austrheim (1993, Fig. 5), which suggests that two dolomite generations can be present in the eclogite. In our opinion, the second equilibrium is subject to more serious criticism. It was used with the composition of pure paragonite although this mineral was never documented as a separate phase in the eclogite of the Bergen Arcs complex. In contrary, there are several lines of evidence that phengite with high (2030%) paragonite mole fraction is the only white mica occurring in the eclogite (Table 4; Austrheim and Griffin, 1985; Boundy et al., 1992). This observation changes drastically the position of univariant lines of equilibrium (2) in the PT diagram. This undoubtedly affects the estimate of water activity in the fluid, which will be demonstrated below. The peak metamorphic temperature of the Bergen Arcs complex was determined by the GrtCpx thermometer (Krogh, 2000). The compositions of coexist-

=1

400

500

700 T, °C

800

900

Fig. 6. Evolution of the PT parameters of metamorphism for the Bergen Arcs eclogite, southwestern Norway. Circles show temperature and the minimum pressure estimates for the peak of metamorphism determined by the GrtCpx thermometer (Krogh, 2000) and CpxQtzPl barometer (Perchuk, 1992). Arrow 12 is the exhumation path. Solid lines: Jd + Qtz = Ab (Perchuk, 1992); Par25 = Jd50 + Ky + H2O (see text for further details); and stability fields of aluminum silicates (Ky, And, and Sil) (Salje, 1986). Dashed lines are isochores of H2O-dominated fluid inclusions from a plastic deformation zone of the amphibolite stage (Andersen et al., 1991a). Dotted lines are basalt solidus at a H O = 1 (Lambert and Wyllie, 1972) and with
2

watercarbon dioxide fluid at a

H2 O

= 0.5 (Perchuk, 1973).

Rectangles show the estimates of PT parameters for the Bergen Arcs eclogites after B, Boundy et al. (1992); AG, Austrheim and Griffin (1985); and J, Jamtveit et al. (1990).

ing garnet and omphacite (both in the matrix and stringers) show a temperature of 730 ± 50°C. The corresponding pressure was determined by the CpxPlQtz barometer (Perchuk, 1992) based on the equilibrium Ab (in plagioclase) = Jd (in clinopyroxene) + Qtz. (3) Since there is no plagioclase in the eclogite, the value obtained from this barometer corresponds to the minimum pressure estimate (Pmin). Omphacites from the matrix and stringers contain up to 53 mol % of jadeite and associate with quartz. They yield a pressure of about 19 kbar at 730°C (Fig. 6). Note that these calcuPETROLOGY Vol. 10 No. 2 2002

ECLOGITES OF THE BERGEN ARCS COMPLEX

107

Table 4. Microprobe analyses and crystal chemical formulas of hydrous minerals and plagioclase from the Bergen Arcs eclogite Clinozoisite Component matrix SiO2 TiO2 Al2O3 Cr2O3 FeO MnO MgO CaO Na2O K2O Total Si Al Cr Ti Fe3+ Fe2+ Mn Mg Ca Na K Total XMg, XAn 39.32 0.09 32.53 0.05 1.69 0.00 0.01 24.69 0.01 0.00 98.38 12.5 2.98 2.91 0.00 0.00 0.11 0.00 0.00 2.01 0.00 0.00 8.01 39.47 0.01 32.59 0.01 1.51 0.00 0.01 24.19 0.03 0.01 98.38 O 3.00 2.92 0.00 0.00 0.10 0.00 0.00 1.97 0.01 0.00 7.99 3.13 2.68 0.01 0.01 0.10 0.00 0.21 0.00 0.28 0.68 7.10 0.69 46.11 0.18 33.54 0.11 1.71 0.00 2.12 0.03 2.14 7.82 93.757 matrix 46.71 0.23 33.09 0.00 1.50 0.07 1.75 0.08 1.68 8.03 93.131 47.18 1.02 32.36 0.00 2.08 0.00 1.63 0.06 2.01 8.15 94.471 Phengite Biotite retrograde 34.39 1.75 20.32 0.00 14.70 0.10 12.18 0.05 0.35 9.20 93.04 2.87 2.00 0.00 0.11 1.02 0.01 1.52 0.00 0.06 0.98 8.56 0.60 Plagioclase symplectite 65.20 0.06 21.92 0.00 0.18 0.00 0.00 2.95 10.03 0.13 100.47 2.86 1.13 0.00 0.00 0.01 0.00 0.00 0.14 0.85 0.01 5.00 0.14 8O 2.86 1.13 0.00 0.00 0.00 0.00 0.01 0.14 0.86 0.01 5.01 0.14 7.44 1.05 0.00 0.03 0.17 1.34 0.01 2.95 1.67 0.49 0.00 15.16 0.69 65.40 0.00 21.92 0.03 0.13 0.07 0.10 3.07 10.18 0.10 101.00 51.65 0.30 6.20 0.00 12.51 0.09 13.74 10.80 1.74 0.00 97.03 Amphibole symplectite 50.52 0.59 5.62 0.00 11.08 0.05 15.58 11.62 1.77 0.07 96.89 23 O 7.29 0.96 0.00 0.06 0.24 1.10 0.01 3.35 1.80 0.50 0.01 15.30 0.75 50.21 0.24 6.03 0.08 13.15 0.08 15.02 10.93 1.64 0.06 97.44 7.17 1.02 0.00 0.03 0.76 0.82 0.01 3.20 1.67 0.45 0.01 15.14 0.80

11 O 3.18 3.19 2.65 2.58 0.00 0.00 0.01 0.05 0.09 0.00 0.18 0.01 0.22 0.70 7.03 0.67 0.12 0.00 0.16 0.00 0.26 0.70 7.07 0.58

lations were performed using plagioclase composition (Ab70) from the initial granulite rather than pure albite. The nonideality of the AbAn solid solution was accounted for using the data of Perchuk et al. (1991). A decrease in the mole fraction of albite in plagioclase shifts isopleths into a higher pressure region. Water activity in the fluid at the peak eclogite metamorphism of the Bergen Arcs complex can be estimated from equilibrium (2). The PT diagram (Fig. 6) shows the respective a H2 O isolines calculated for omphacite (Jd53) and phengite (Prg25) on the basis of experimental data (Holland, 1979) accounting for nonideality in the clinopyroxene (Perchuk, 1992) and paragonitemuscovite solid solutions (Eugster et al., 1972). Water fugacity was calculated using the model of Holland and Powell (1990). The relative position of the univariant lines of equilibrium (2) and thermobarometric estimates for the peak of metamorphism (~730°C and ~19 kbar) sugPETROLOGY Vol. 10 No. 2 2002

gests that water activity in the fluid was not high ( a H2 O < 0.4). This estimate is in conflict with the widely accepted opinion on water-rich fluid composition in the Bergen Arcs eclogite. However, the use of real compositions of coexisting minerals allowed us to resolve the above-mentioned contradiction between the thermometric estimates and the position of hydrous basalt solidus. Indeed, at moderate water activity in fluid ( a H2 O < 0.4), the solidus curve of basalt shifts to higher temperatures removing thus the eclogite from the melting field (Fig. 6). Retrograde Metamorphism (Amphibolite Stage) For this stage, pressure was determined by the Cpx PlQtz barometer (Perchuk, 1992) and compositions of augite (Jd3)plagioclase (An14) from symplectites formed after omphacite and quartz during decompres-

108

PERCHUK

sion. The isopleth corresponding to this stage is shown in Fig. 6. Temperature estimation is less straightforward, because garnet and clinopyroxene do not exchange Fe and Mg in the course of retrograde metamorphism (Perchuk et al., 1996; Aranovich and Pattison, 1995) and no other mineral equilibria in the rock could be used as thermometers. Because of this, we used the results of a fluid inclusion study (Andersen et al., 1991a) in quartz from the same shear zone at the amphibolite stage in Honsloy Island, where the eclogite was sampled. The PT diagram (Fig. 6) shows isochores for the unaltered aqueous fluid inclusions. Assuming that the development of shear at the amphibolite stage occurred simultaneously with symplectite formation in the eclogite, the PT parameters of retrograde metamorphism are determined from the intersection of these isochors with the isopleth (Jd3) corresponding to the symplectite formation stage. The resulting temperature and pressure estimates (about 500°C and 6 kbar) correspond to the amphibolite-facies PT conditions and are within the stability field of kyanite (Fig. 6), which is widespread in the rock. Thus, in the course of decompression by 13 kbar, the Bergen Arcs eclogite cooled by about 230°C. If total pressure was equal to lithostatic load, the total pressure decrease corresponded to a vertical displacement of the rock of approximately 45 km. DURATION OF THE PT EVOLUTION Some evidence on the rate of the metamorphism of the Bergen Arcs eclogite was reported by Erambert and Austrheim (1993) and Boundy et al. (1996, 1997). The latter authors calculated the minimum cooling rate (10 15°C/m.y.) from the isotopic ages and temperatures of the main metamorphic stages. Taking into account the uncertainty in the closure temperatures of isotopic systems and great scatter in isotopic age estimates, the real cooling rates could be much different from the values estimated by these authors. Erambert and Austrheim (1993) estimated the duration of metamorphism from characteristic diffusion distances in the transition zone between a stringer and garnet. Such estimates can hardly be regarded as accurate, because the authors solved an isothermal problem (1 4 m.y. at 700°C and 622 m.y. at 650°C), whereas the process was evidently not isothermal (cooling from 730 to 500°C, Fig. 6). We attempted to overcome this drawback at solving the following diffusion problem. Problem Formulation and Solution In order to determine the temperature and pressure evolution of the eclogite, we modeled smoothing the FeMg concentration gradient at the boundary of a stringer with the enclosing garnet. The appropriate nonisothermal diffusion problem was considered in our

recent publications (Perchuk and Philippot, 2000a, 2000b). Thin stringers (2h ~ 2040 µm) are formed through healing of microscopic cracks in garnet at an early stage of rim growth in the presence of hydrous fluid (Perchuk and Varlamov, 1995). A characteristic feature of stringers from the Bergen Arcs complex is their great thickness (2h > 90 µm) and the presence of trails of mineral inclusions, most common among which are omphacite and quartz (Fig. 2b). It is likely that the granulite garnet contained large cavities (cracks) filled with minerals of the granulite paragenesis, which were later transformed into garnet stringers with omphacite and quartz inclusions. It is known that the primary granulite garnet (Grt1) did not participate in eclogitization reactions (Austrheim and Griffin, 1985). Consequently, the nucleation and growth of the late eclogitic garnet (Grt2) occurred at its faces and crack walls without their dissolution. In such a case, the starting concentrations changed step-wise at the boundary between the two garnets, which must be reflected in the initial condition of the problem. Proceeding from the morphology of stringer, their homogenization was modeled as diffusion from an extended plate of a limited thickness in infinite space (Lasaga, 1998). A similar problem of the cooling of a plate (dike) of a limited thickness is well known in thermal physics (Carslaw and Jaeger, 1986). The one-dimensional diffusion equation is known as Fick's second law and is expressed as C Ci ------- = D ----------i , 2 t x
2

(4)

where Ci is the concentration of diffusing component i, x is the distance from the center of the stringer, and D(t) is the diffusion coefficient of component i specified as a function of time. Substituting
t

t' =

0

D() ----------- d , 2 h

(5)

where is the variable of integration, and x' = x / h , where h is the half-thickness of the stringer, we obtain a new simplified form of differential Eq. (4), Ci C -------i = ------------- , 2 t' ( x' )
2

(6)

which does not include the variable coefficient D(t). Since the concentrations of only two components, magnesium and iron, change in the transitional zone between the stringer and the core (Figs. 4, 7; Table 5), we will assume that the diffusion process had a binary character. Consequently, only these two cations can be considered using the value #Mg. In order to find the
PETROLOGY Vol. 10 No. 2 2002

ECLOGITES OF THE BERGEN ARCS COMPLEX

109

theoretical solution of the problem, initial and boundary conditions should be specified. The shape of the concentration profile across the stringer (Fig. 4) suggests that diffusion affected only thin marginal zones. Consequently, the central part of the stringer retained its initial composition (C2) identical to that of rim zones. The composition of initial garnet (C1) enclosing the stringer is approximated by the composition of the homogeneous core corresponding to the early granulite metamorphic stage. Owing to the inert behavior of granulite garnet in the eclogite-forming process (Austrheim and Griffin, 1985), the initial composition at the stringerhost boundary should change step-wise. Thus, the initial condition for the diffusion problem is written as C ( x', t ' = 0 ) = ( x' ) ; ( x' ) = C 1 , x' 1 , (7) ( x' ) = C 2 , x' < 1 , where the function (x) describes the initial distribution of #Mg in the core (C1) and the stringer (C2) of a thickness of 2h. The isolation of the diffusion zone at the stringer host boundary prevents the input of isomorph components from outside (from fluid or coexisting minerals). Thus, we assume that the system was closed. The general solution of the problem of diffusion from an extended plate of a limited thickness (2h) in infinite space is

Mg, formula unit 2.0 y = 1.0938x + 2.5242 R = 0.998

1.5

1.0

0.5 0.5

1.0

1.5

2.0 Fe, formula unit

Fig. 7. Correlation of atomic iron and magnesium contents along the line of microprobe profile through the diffusion zone between a stringer and enclosing garnet. There is a strong negative linear dependence, R = 0.998 obtained from 56 analyses (Table 5).

1 C ( x', t ' ) = ------------2 t'

( x' ) exp ------------------- ( ) d , 4t'
2

where is the variable of integration. Integration and substitution of the initial condition (Eq. (7)) yield the desired particular solution: C2 C1 1 + x' 1 x' C ( x', t ' ) = C 1 + ----------------- erf ------------ + erf ------------ , (8) 2 t' 2 t' 2 where erf is the error function of the form
z

2 2 erf ( z ) = ------ exp ( ) d ,

0

where is the variable of integration. Equation (8) describes a change in the initial chemical profile across the boundary between a stringer and host garnet at the arbitrary thermal history of the rock. Results of Modeling The initial conditions (Eq. (7)) included the #Mg values of the core (C1 = 67) and stringer (C2 = 39) and a stringer width of 2h = 94 µm. The sequential stages of diffusion homogenization of the stringer were calculated by Eq. (5) at various t' values. Figure 8 demonstrates that the best consistency of the calculated and
PETROLOGY Vol. 10 No. 2 2002

measured microprobe profiles is attained at t' = 0.03, while the stringer homogenizes completely at t' = 5. The estimate of the degree of garnet homogenization (t' = 0.03) allows us to evaluate the timing of the P T rock evolution. For this purpose, we used the approach developed in our recent publications (Perchuk et al., 1998; Perchuk and Philippot, 2000; etc.). This approach is based on the numeric solution of integral Eq. (5) and determination of t' for a series of thermal histories with identical temperature ranges and common functional time dependences. It was assumed that temperature changed linearly during decompression (Fig. 9). In addition to time, Eq. (5) contains the diffusion coefficient (D) and the stringer size (h). At the consideration of a binary process, it is convenient to use the coefficient of coupled diffusion (interdiffusion) D(FeMg) defined by the expression (Barrer et al., 1963) D
FeMg

D Fe D Mg ln i = ------------------------------------------ 1 + ------------- , X Mg D Mg + X Fe D Fe ln X i

(9)

where i is the activity coefficient of component i and Di is the diffusion coefficient. The latter is described by the Arrhenius equation Ea D i = D 0 exp -------------- , R T ( t )

110

Table 5. Part of a chemical profile through the stringercore transitional zone in garnet (Fig. 2b), analyses were obtained at an increment of 3 µm Component SiO2 TiO2 Al2O3 Cr2O3 FeO MnO MgO CaO Na2O K2O Total Si Al Cr Ti Fe3+ Fe2+ Mn Mg Ca Na K Total XMg XCa XAlm XSps XPrp XCrs Core 86 40.72 0.03 23.14 0.00 14.39 0.22 14.92 7.11 0.03 0.00 101 2.96 1.99 0.00 0.00 0.05 0.83 0.01 1.62 0.55 0.00 0.00 8.01 0.66 0.18 0.27 0.00 0.53 0.18 87 40.33 0.01 23.03 0.00 13.67 0.22 14.76 7.04 0.02 0.03 99.1 2.97 2.00 0.00 0.00 0.02 0.82 0.01 1.62 0.56 0.00 0.00 8.01 0.66 0.18 0.27 0.00 0.54 0.18 88 40.47 0.03 23.17 0.00 13.83 0.11 14.69 6.97 0.01 0.00 99.3 2.98 2.01 0.00 0.00 0.01 0.84 0.01 1.61 0.55 0.00 0.00 8.01 0.66 0.18 0.28 0.00 0.53 0.18 89 40.65 0.09 23.05 0.03 14.74 0.11 14.58 7.15 0.02 0.00 100 2.97 1.98 0.00 0.00 0.04 0.86 0.01 1.59 0.56 0.00 0.00 8.01 0.65 0.19 0.28 0.00 0.52 0.18 90 40.24 0.04 23.35 0.09 14.78 0.15 13.68 6.80 0.03 0.00 99.2 2.98 2.04 0.01 0.00 0.00 0.91 0.01 1.51 0.54 0.00 0.00 8.00 0.62 0.18 0.31 0.00 0.51 0.18 91 39.99 0.06 22.96 0.00 17.61 0.20 12.95 6.74 0.00 0.00 100 2.96 2.00 0.00 0.00 0.04 1.05 0.01 1.43 0.53 0.00 0.00 8.02 0.58 0.18 0.34 0.00 0.47 0.17 Transitional zone 92 39.41 0.08 22.47 0.03 18.12 0.41 11.77 6.95 0.02 0.00 99.24 2.97 1.99 0.00 0.00 0.03 1.11 0.03 1.32 0.56 0.00 0.00 8.01 0.54 0.19 0.36 0.01 0.43 0.18 93 38.72 0.14 22.35 0.00 19.74 0.40 10.75 6.66 0.00 0.01 98.8 2.95 2.01 0.00 0.01 003 1.23 0.03 1.22 0.54 0.00 0.00 8.02 0.50 0.18 0.40 0.01 0.40 0.18 94 38.92 0.10 22.19 0.00 20.65 0.33 9.75 6.81 0.03 0.00 98.8 2.98 2.00 0.00 0.01 0.01 1.31 0.02 1.11 0.56 0.01 0.00 8.00 0.46 0.19 0.44 0.01 0.37 0.19 95 38.89 0.08 22.10 0.03 21.07 0.63 9.32 6.93 0.03 0.00 99.1 12 O 2.98 2.00 0.00 0.00 0.02 1.33 0.04 1.06 0.57 0.00 0.00 8.01 0.44 0.19 0.44 0.01 0.35 0.19 96 39.07 0.09 22.18 0.00 21.87 0.46 9.17 6.87 0.02 0.02 99.8 2.98 1.99 0.00 0.01 0.02 1.37 0.03 1.04 0.56 0.00 0.00 8.01 0.43 0.19 0.45 0.01 0.34 0.19 97 38.71 0.03 21.94 0.00 22.14 0.45 8.81 6.80 0.09 0.00 99.0 2.98 1.99 0.00 0.00 0.03 1.40 0.03 1.01 0.56 0.01 0.00 8.00 0.42 0.19 0.46 0.01 0.33 0.19 98 38.48 0.22 22.23 0.04 22.33 0.50 8.68 6.85 0.02 0.00 99.3 2.96 2.01 0.00 0.01 0.01 1.42 0.03 0.99 0.56 0.00 0.00 8.01 0.41 0.19 0.47 0.01 0.33 0.19 99 38.32 0.09 21.95 0.02 21.75 0.46 8.39 6.99 0.04 0.00 98.0 2.98 2.01 0.00 0.01 0.00 1.42 0.03 0.97 0.58 0.01 0.00 8.00 0.41 0.19 0.47 0.01 0.32 0.19 100 Stringer 101 102 38.21 0.10 21.98 0.00 22.84 0.56 8.09 7.20 0.05 0.00 99.0 2.96 2.01 0.00 0.01 0.03 1.45 0.04 0.93 0.60 0.01 0.00 8.01 0.39 0.20 0.48 0.01 0.31 0.20 103 38.61 0.04 22.09 0.00 22.82 0.56 8.05 7.13 0.07 0.00 99.4 2.98 2.01 0.00 0.00 0.02 1.46 0.04 0.92 0.59 0.01 0.00 8.01 0.39 0.20 0.48 0.01 0.31 0.20 104 38.38 0.03 21.99 0.00 22.11 0.62 8.03 7.35 0.00 0.00 98.5

38.15 38.14 0.13 0.14 22.14 22.01 0.07 0.00 23.58 22.36 0.49 0.46 8.51 8.24 7.12 7.01 0.05 0.02 0.00 0.04 100.0 2.92 2.00 0.00 0.01 0.07 1.44 0.03 0.97 0.58 0.01 0.00 8.03 0.40 0.19 0.47 0.01 0.31 0.18 98.4 2.96 2.02 0.00 0.01 0.01 1.44 0.03 0.96 0.58 0.00 0.00 8.01 0.40 0.19 0.48 0.01 0.32 0.19

PERCHUK

2.98 2.01 0.00 0.00 0.01 1.43 0.04 0.93 0.61 0.00 0.00 8.01 0.39 0.20 0.47 0.01 0.31 0.20

PETROLOGY Vol. 10 No. 2 2002

ECLOGITES OF THE BERGEN ARCS COMPLEX Mg/(Mg + Fe ), % 70 К1 = 67 1.E + 01 60 5.00 1.00 50 1.E 01 40 К2 = 39 1.E 02 0.03 1.4 m.y. 30 3 2 1 0 1 2 3 x' = x/h 1.E 03 10 t' 1.E + 0
2

111 ()

t' 1.E + 02

T, °C 700

1 Complete homogenization 2 3 4

500

0 t, m.y.

100

1.E +0

t' = 0.03

10

1

100 10 Duration, m.y. (b)

1

10

2

Fig. 8. Sequential stages of stringer homogenization (Eq. (9)) in the coordinates dimensionless distance (x/h) versus Mg/(Mg + Fe2+) ratio (#Mg) of garnet. Initial conditions (t' = 0): C1 = 67%, C2 = 39% and h = 47 µm (h is the initial boundary between the stringer and enclosing garnet). Numerals are values of the variable t'. The curve calculated at t' = 0.03 adequately describes the points of microprobe profiling (circles).

1.E 01

Chronometer closure t ' = 0.03

1.E 02

where D0 is the pre-exponential factor; Ea is the activation energy of diffusion of the component; R is the universal gas constant; T(t) is the temperature as a function of time (Fig. 9, inset) corresponding to the temperature boundaries of the PT path of eclogite metamorphism (Fig. 6). No correction for the nonideality of solid solution (Eq. (9)) is necessary, because i = 1 for the Alm Prp solid solution (Perchuk, 1991). The calculations were based on the recent experimental data on the diffusion of Mg and Fe in garnet (Ganguly et al., 1998): D0 = 4.66 в 109 m2/s and Ea = 254 + 0.53 в P [kbar] kJ/mol for Mg and D0 = 3.5 в 109 m2/s and Ea = 274 + 0.56 в P [kbar] kJ/mol for Fe. The t' value is linearly dependent on the total duration of cooling (decompression) at fixed stringer size (2h = 94 µm) (Fig. 9). Taking into account the known degree of stringer homogenization (t' = 0.03, Fig. 8), we obtain that the period of cooling and decompression from 730°C and 19 kbar (metamorphic peak) to ~500°C and 6 kbar (retrograde stage) lasted 1.4 m.y. Assuming that rock cooling by 230°C and ascent by 45 km (13 kbar) occurred monotonously, we obtain the average rates of these processes as 160°C/m.y. and 3 cm/y, respectively. Since diffusion coefficient shows a complex temperature dependence, the contributions of various stages of thermal history into t' were different. In particular, the time dependence of t' in the course of 1.4 m.y. retroPETROLOGY Vol. 10 No. 2 2002

0.8 m.y. 1.E 03 10
2

10

1

100 Time, m.y.

10

1

10

2

Fig. 9. Determination of the duration of the retrograde metamorphic stage. (a) Dependence of t' on the duration of retrograde metamorphism at linear rock cooling (see inset). The thickness of the stringer is h = 47 µm. The solid lines correspond to DFeMg at XMg = 0.50: 1, experimental data of Gerasimov (1987); 2, Loomis et al. (1985); 3, Chakraborty and Ganguly (1992); and 4, Ganguly et al. (1998). The duration of the retrograde stage, 1.4 m.y. is determined from the degree of homogenization (t' = 0.03) at the lowest DFeMg value (line 4). Other DFeMg values (lines 13) result in a duration of less than 1 m.y. (b) Dependence of t' on the current time value within 1.4 m.y. (Table 6), which demonstrates that efficient mass transfer terminated (closure of the chronometer) at a boundary of 0.8 m.y.

grade evolution demonstrates (Fig. 9b; Table 6) that in the low-temperature region (T < ~600°C), the variable was essentially invariant indicating that mass transfer became negligible. Consequently, the obtained average rates of cooling and ascent are relevant to the high-temperature part of metamorphism (from 730 to about

112

PERCHUK

Concentration

Distance

Concentration Distance Ia Garnet Ib II Garnet

Thin section surface

III

Fig. 10. Main morphological type of stringers (I, II, and III, see text for explanation). The stringer of type I demonstrates an apparent increase in the width of the diffusion zone at various section orientations (cf. Ia and Ib).

600°C), which continued for about 0.8 m.y. (Fig. 9b, Table 6), rather than to the whole retrograde stage. Sources of Errors of the Method Diffusion coefficient. The influence of diffusion coefficient on t' can be assessed from Fig. 9, which displays calculated t' values for all currently available estimates of the interdiffusion coefficients of FeMg in garnet. The diagram demonstrates that the line obtained from the data of Ganguly et al. (1998) corresponds to the maximum duration and minimum rate of the process. Correspondingly, other published data result in shorter duration and higher rates. Temperature of metamorphism. Temperature was determined by the most recent calibration of the Grt Cpx geothermometer of Krogh (2000), who availed himself of a considerable body of experimental and natural data on this exchange equilibrium. For the eclogite studied, this calibration yields lower temperatures (by 30°C or more) than other versions of this thermometer (Ellis and Green, 1974; Powell, 1985; Krogh, 1988).

Higher temperatures results in higher rates of retrograde metamorphism and exhumation of eclogite. Stringer geometry. Back-scattered electron images (e.g., Fig. 2b), which were used to reveal stringers, are two-dimensional reflections of a three-dimensional pattern, which was never studied. However, proceeding from the supposed relation of these patterns to cracks in garnet, several major spatial forms may be suggested that define the linear stringer geometry on the surface of a thin section. In general, it could be a plate or a wedge (Fig. 10). The former case exactly corresponds to the conditions of the problem studied. At limited development of the diffusion process (if diffusion affected only the boundary zone of the stringer) such an approach is also valid for a wedge-like stringer. Note however, that at a higher extent of diffusion, significant underestimation (variant II, Fig. 10) and overestimation of the degree of homogenization (variant III, Fig. 10) are possible. Noteworthy is also the problem of uncertainty in the estimate of stringer thickness (h): the real value of h is obtained only if the section surface is perpendicular to the stringer plane (variant Ia, Fig. 10). In all other

Table 6. Changes in the parameter t' (Eq. (5)) at the linear cooling of a 94-µm-thick stringer from 730 to 500°C in 1.4 m.y.; DFeMg is after Ganguly et al. (1998) T, °C t, m.y. t' в 10
4

730 0.00 0

707 0.14 175

684 0.28 253

661 0.42 286

638 0.56 299

615 0.70 305

592 0.84 307

569 0.98 308
PETROLOGY

546 1.12 308
Vol. 10

523 1.26 308
No. 2

500 1.40 308
2002

ECLOGITES OF THE BERGEN ARCS COMPLEX

113 (a) 10
3

cases, the stringer thickness and the extent of diffusion mass transfer are overestimated (compare variants Ia and Ib in Fig. 10 and inset), which increases the estimated duration of the thermal history of the rock (Eq. (5)). The analysis of the major error sources allows us to conclude that our method yields the maximum estimate of the duration of metamorphism. Correspondingly, the calculated rates of changes in the intensive parameters of metamorphism are minimum estimates. But even such minimum estimates are an order of magnitude higher than the values previously reported for the Bergen Arcs complex (Boundy et al., 1997). DISCUSSION Thermal History and Age of Metamorphism The absolute age estimates of the peak metamorphic stage show a rather significant scatter. In contrast, the dating of the retrograde stage yielded consistent values (Boundy et al., 1996; Chauvet and Dallmeyer, 1992). The correct interpretation of these data requires estimation of the closure temperatures of respective isotopic systems using the Dodson formalism and independent cooling rate values obtained from mineral chronometry (Gerasimov et al., 1998; Perchuk and Philippot, 2000). The ArAr ages of hornblende and phengite, 452 ± 4 and 445 ± 17 Ma, respectively (Boundy et al., 1996; Chauvet and Dallmeyer, 1992), can be correlated with the thermal history of the rock, if the closure temperature of the Ar isotope system in these minerals (Tc) is known. This parameter can be expressed as (Dodson, 1973) E a /R T c = ----------------------------------------------- , 22 ln [ AD 0 R T c / a sE a ] (10)

Tc, °C Ar / Hbl 700

10 10

2

1

600 max min 500

10

0

400 min max Ar / Mu 700 (b) 10
3

10 600 10 max 500 min 400 min max 1 2 10

2

1

0

where R is the universal gas constant, s is the cooling rate, Ea is the activation energy of diffusion, a is the characteristic diffusion length, and A is the geometrical factor. For a plate, A = 8.7 and a is the half-thickness; for a cylinder, A = 27 and a is the radius; and for a sphere, A = 55 and a is the radius. Substituting the diffusion parameters (Ea and D0) of Ar in amphibole (Harrison, 1981) and phengite (Kirschner et al., 1996) into Eq. (8) and taking into account considerable variations in the size of grains and cooling rates of both minerals in nature, we conclude that Tc can vary by hundreds of degrees (Figs. 11a, 11b). This diagram allows us to estimate Tc for the minerals of the Bergen Arcs complex. At a cooling rate of 160°C/m.y. and a grain radius of 0.050.40 mm (Boundy et al., 1992, 1996), the average temperatures of the closure of Ar system in amphibole and phengite are 510590°C and 440520°C, respectively. Thus, the Tc of amphibole falls within the temperature range of the retrograde metamorphic stage (from 730 to 500°C) and is very close to the temperature interval, for which
PETROLOGY Vol. 10 No. 2 2002

3 4 Radius, mm

Fig. 11. Calculated diagrams for the dependence of Ar system closure in (a) amphibole and (b) phengite from the Bergen Arcs eclogite on the cooling rate (numerals at curves) and grain size. Dotted lines show the minimum (min) and maximum (max) sizes of crystals from the eclogite that provide Tc values at a cooling rate of 180°C/m.y. (see text for explanation).

the mineral chronometer provides reliable values (from 730 to 600°C). If the estimated cooling rate (160°C/m.y.) persisted through the whole retrograde stage, the age of eclogitic metamorphism can be strongly constrained. The latter can differ from the Ar Ar age of the amphibole by no more than 0.8 m.y., i.e., it should be equal to the ArAr age of amphibole taking into account the uncertainty (452 ± 4 Ma). In such a case, the UPb age of zircon (450 ± 10 Ma) deter-

114

PERCHUK

mined by SHRIMP (Boundy et al., 1997) seems to be the most reliable. Mechanism of Exhumation A number of researchers believed that the protolith of the Bergen Arcs complex was represented by the Precambrian basement rocks, granulites and metaanorthosites (Jamtveit et al., 1990). It is remarkable that to our best knowledge, there is no evidence on the early retrograde recrystallization of the Bergen Arcs granulites, which could have indicated the ascent of the complex in pre-Caledonian time before the onset of eclogitization. Thus, it can be supposed that during the Caledonian collision epoch, the Bergen Arcs granulites were situated at the base of the crust in the subducted plate at depths of more than 30 km (9 kbar). As will be shown below, this inference is important for the understanding of the mechanism of eclogite movement from depths much greater than the average thickness of the Earth's crust. Exhumation rates of a few centimeters per year, which were inferred from the dynamics of metamorphism of the Bergen Arcs complex, can be attained at two major geodynamic processes. The first of them assumes viscous flow (direct and return) of plastic material (metasediments enriched in micaceous minerals or serpentinites) in an accretion prism and later in a subduction channel initiated by the movement of a submerging plate (Dobretsov and Kirdyashkin, 1994; Hsu, 1971; Cloos, 1982; Cloos and Shreve, 1988; Beaumont et al., 1999; Gerya et al., 2001). Cloos (1982) calculated that only small eclogite bodies, no larger than a few tens of meters, could be transported to the surface in a matrix of sedimentary rocks. Larger blocks of dense eclogite are irrevocably entrained to depths by a descending flow. It is evident that the ascent of a giant block, such as the Bergen Arcs complex, could hardly be explained within this model. Moreover, the initial position of the complex at the base of the subducted plate is not consistent with the fact that eclogite entrapment by the accretion prism, which is proposed in the collision model, should have occurred from the surface of the overlying mantle and (or) from the subducted plate itself. Rapid changes in temperature and pressure are also predicted for the ascent of large blocks of relatively low-density continental rocks (England and Holland, 1979; Perchuk et al., 1992; Chemenda, 1993; Ernst et al., 1997; Ernst, 1999). The three latter publications proposed a scenario, which is most appropriate for the Norwegian Caledonides. According to this scenario, light salic rocks (Bergen Arcs complex ?) split off the head part of the subducted plate (Western Gneiss Region ?) are moved toward the surface along the convergent boundary of the continental plate and the mantle hanging wall. An important additional factor favoring the rapid ascent of eclogite is the serpentinization of the base of the hanging-wall, which greatly reduces

friction between the contacting crustal and mantle rocks (Guillot et al., 1999). In addition, the rate of exhumation could be strongly affected by squeezing the sheet by converging plates (Dobretsov, 2000). The contributions of these effects to the process of high-pressure rock exhumation have not been quantified. On Metamorphic Fluid The leading role of aqueous fluid in high-pressure mineral formation was first demonstrated by Udovkina (1971) on the example of the Marunkeu complex in the Polar Urals. This idea was universally accepted within many years owing mainly to the brilliant studies of the Bergen Arcs complex by Norwegian geologists (Austrheim and Griffin, 1985; Jamtveit et al., 1990; Andersen et al., 1991a, 1991b, 1993; Austrheim and Mork, 1988). These researchers paid significant attention to the problem of the fluid regime of metamorphism. The investigation of fluid inclusions (Andersen et al., 1991b) in syntectonic PlPhnQtzCzoCal veins from the central part of a shear zone in Holsnoy Island (Bergen Arcs complex) revealed primary aqueous salt (3031% NaCl) and carbon dioxide with minor nitrogen (> 5% N2) inclusions. Although the PVT properties of these inclusions did not correspond to the parameters of the eclogite stage because of inclusion decrepitation, the authors argued that the inclusion compositions adequately characterized the fluid phase and its separation owing to high-temperature immiscibility in the presence of NaCl into water-rich and carbon dioxide-rich parts (Shmulovich, 1988). In their opinion, such a fluid composition is not compatible with the absolute predominance of water in the eclogite-stage fluid (Jamtveit et al., 1990). Thus, our inference on moderate water content in the fluid mixture at granulite eclogitization ( a H2 O < 0.4) is supported by independent evidence from fluid inclusions. CONCLUSIONS (1) The investigation of mineral equilibria in the eclogite of the Bergen Arcs complex suggested that the PT parameters of the eclogite stage were about 730°C and 19 kbar and water activity in the fluid (H2ONaCl N2CO2) was no higher than 0.4. The temperature and pressure of the retrograde stage were estimated as 500°C and 6 kbar, respectively. (2) Modeling FeMg interdiffusion at the boundary of a stringer with the host garnet under the conditions of retrograde evolution allowed us to demonstrate that the cooling (from 730 to 600°C) and decompression (from 19 to 12 kbar) of eclogite during the retrograde stage lasted about 0.8 m.y., which provided cooling and ascent rates of 160°C/m.y. and 3 cm/y, respectively. (3) The calculated closure temperatures of the argon isotope system in amphibole from the Bergen Arcs eclogite (510590°C) is within the temperature range
PETROLOGY Vol. 10 No. 2 2002

ECLOGITES OF THE BERGEN ARCS COMPLEX

115

of the retrograde stage. The short duration of the latter (cooling to 600°C took less than one million years) allows us to suggest that the ArAr age of amphibole (448455 Ma, Boundy et al., 1996) corresponds in fact to the peak of eclogite metamorphism. ACKNOWLEDGMENTS The authors thank M. Erambert and H. Austrheim (University of Oslo), who kindly provided samples of eclogite for the investigation; L.L. Perchuk (Moscow State University), S.P. Korikovsky, V.Yu. Gerasimov, and V.I. Mal'kov-skii (IGEM) for helpful comments on the early version of the manuscript; A.V. Girnis (IGEM) for discussion of the results; and F. Blan (University of Paris 7) for assistance in the investigations on a scanning electron microscope. The study was financially supported by the Russian Foundation for Basic Research, project no. 01-05-64585 (A.L. Perchuk); the program of leading scientific school support, project no. 00-15-98519 (L.L. Perchuk); and the program of cooperation of the Russian Academy of Sciences and CNRS (France), project no. 7768. REFERENCES
Andersen, T., Extensional Tectonics in the Caledonides of Southern Norway, An Overview, Tectonophysics, 1997, vol. 285, pp. 333351. Andersen, T., Austrheim, H., and Burke, E.A.J., Fluid Induced Retrogression of Granulites in the Bergen Arcs, W. Norway: Fluid Inclusion Evidence from Amphibolite-Facies Shear Zones, Lithos, 1991a, vol. 27, pp. 2942. Andersen, T., Austrheim, H., and Burke, E.A.J., Mineral FluidMelt Interactions in High-Pressure Shear Zones in the Bergen Arcs Nappe Complex, Caledonides of W. Norway: Implications for the Fluid Regime of Eclogite Facies Metamorphism, Lithos, 1991b, vol. 27, pp. 2942. Andersen, T., Austrheim, H., Burke, E.A.J., and Elevold, S., N2 and CO2 in Deep Crustal Fluids: Evidence from the Caledonides of Norway, Chem. Geol., 1993, vol. 108, pp. 113132. Aranovich, L.Ya. and Pattison, D.R.M., Reassessment of the GarnetClinopyroxene FeMg Exchange Thermometer: I. Evaluation of the Pattison and Newton (1989) Experiments, Contrib. Mineral. Petrol., 1995, vol. 119, pp. 1629. Austrheim, H. and Griffin, W., Shear Deformation and Eclogite Formation within Granulite-Facies Anorthosites of the Bergen Arcs, Western Norway, Chem. Geol., 1985, vol. 50, pp. 267281. Austrheim, H. and Mork, M.B.E., The Lower Continental Crust of Caledonian Mountain Chain: Evidence from Former Deep Crustal Sections in Western Norway, Nor. Geol. Unders. Spec. Publ., 1988, vol. 3, pp. 102113. Austrheim, H. and Robins, B., Reactions Involving Hydration of Orthopyroxene in AnorthositeGabbro, Lithos, 1991, vol. 14, pp. 275281. Austrheim, H., Erambert, M., and Engvik, A.K., Processing of Crust in the Root of the Caledonian Continental Collision Zone: Role of Eclogitization, Tectonophysics, 1997, vol. 273, pp. 129153.
PETROLOGY Vol. 10 No. 2 2002

Barrer, R.M., Bartholomew, R.F., and Rees, L.V.C., Ion Exchange in Porous Crystals. Part II. The Relationship between Self and Exchange Diffusion Coefficients, J. Phys. Chem., 1963, vol. 24, pp. 309317. Beaumont, K., Ellis, S., and Pfiffner, A., Dynamics of Sediment SubductionAccretion at Convergent Margins: ShortTerm Modes, Long-Term Deformation, and Tectonic Implications, J. Geophys. Res., 1999, vol. 104, pp. 1757317601. Boundy, T.M., Essene, E.J., Hall, C.M., et al., Rapid Exhumation of the Lower Crust during ContinentContinent Collision and Extension: Evidence from 40Ar, 39Ar Incremental Heating of Hornblendes and Muscovites, GSE Bull., 1996, vol. 108, pp. 14251437. Boundy, T.M., Fountain, D.M., and Austrheim, H., Structural Development and Petrofabrics of Eclogite Facies Shear Zones, Bergen Arcs, Western Norway: Implications for Deep Crustal Deformational Processes, J. Metamorph. Geol., 1992, vol. 10, pp. 127146. Boundy, T.M., Mezger, K., and Essene, E.J., Temporal and Tectonic Evolution of the GranuliteEclogite Association from the Bergen Arc, Western Norway, Lithos, 1997, vol. 39, pp. 159178. Bryhni, I., Status of Supracrustal Rocks in the Western Gneiss Region, S. Norway, in The Caledonides Geology of Scandinavia, Gayer, R.A., Ed., London: Graham and Trotman, 1989, pp. 221228. Burg, J.-P. and Podladchikov, Y., From Buckling to Asymmetric Folding of the Continental Lithosphere: Numerical Modeling and Application to the Himalayan Syntaxes, in Tectonics of the Nanga Parbat Syntaxis and the Western Himalaya, Khan M.A., Treloar P.J., Searle M.P., and Jan, M.Q, Eds., Geol. Soc. London. Spec. Publ., 2000, vol. 170, pp. 219236. Burton, K.W., Kohn, M.J., Cohen, A.S., and O'Nions, R.K., The Relative Diffusion of Pb, Nd, Sr, and O in Garnet, Earth Planet. Sci. Lett., 1995, vol. 133, pp. 199211. Carslaw, H.S. and Jaeger, J.C., Conduction of Heat in Solids, Oxford: Oxford Univ. Press, 1986. Chakraborty, S. and Ganguly, J., Cation Diffusion in Aluminosilicate Garnets: Experimental Determination in SpessartineAlmandine Diffusion Couples, Evaluation of Effective Binary Diffusion Coefficients, and Applications, Contrib. Mineral. Petrol., 1992, vol. 111, pp. 7486. Chauvet, A. and Dallmeyer, R.D., 40Ar/39Ar Mineral Dates Related to Devonian Extension in the Southwestern Scandinavian Caledonides, Tectonophysics, 1992, vol. 210, pp. 155177. Chemenda, A.I., Subduction of the Lithosphere and Back Arc Dynamics: Insights from Physical Modeling, J. Geophys. Res., 1993, vols. 9899, pp. 1616716185. Chopin, C., Very-High-Pressure Metamorphism in the Western Alps: Implications for Subduction of Continental Crust, Philos. Trans. R. Soc. London, 1987, vol. 321, pp. 183197. Chopin, C. and Sobolev, N.V., Principal Mineralogical Indicators of UHP in Crustal Rocks, Coleman, R.G. and Wang, X., Eds., Cambridge: Cambridge Univ. Press, 1995, pp. 96131. Cliff, R.A., Barnicoat, A.C., and Inger, S., Early Tertiary Eclogite Facies Metamorphism in the Monviso Ophiolite, J. Metamorph. Geol., 1998, vol. 16, pp. 447455. Cloos, M., Flow Melanges: Numerical Modeling and Geologic Constraints on Their Origin in the Franciscan Subduc-

116

PERCHUK Gerasimov, V.Yu., Temperaturnaya evolyutsiya metamorfizma i obratimost' mineral'nykh ravnovesii (The Temperature Evolution of Metamorphism and Reversibility of Mineral Equilibria), Moscow: Nauka, 1992. Gerasimov, V.Yu., Savko, K.A., Lebedev, V.A., and Arakel'yants, M.M., A Thermochronological Approach to the Determination of Gneiss Ages from the Voronezh Crystalline Massif by 40Ar, 39Ar Techniques with the Laser-Aided Sample Selection, Vestn. Voronezh. Univ., Ser. Geol., 1998, no. 5, pp. 5961. Gerya, T.V., Perchuk, A.L., and Schtockhert, B., Numerical Simulation of Subduction Zones: Coupling of Rheology, Heat Transfer, Fluid Budget, and Phase Equilibria, in Deformation Mechanisms, Rheology, and Tectonics, Abstract vol., Faculty of Earth Sciences, Utrecht Univ., Netherlands, 2001, p. 55. Gerya, T.V., Perchuk, L.L., van Reenen, D.D., and Smit, C.A., Two-Dimensional Numerical Modeling of PressureTemperatureTime Paths for the Exhumation of Some Granulite Facies Terrains in the Precambrian, J. Geodynamics, 2000, vol. 30, pp. 1735. Griffin, W.L. and Brueckner, H.K., Caledonian SmNd Ages and a Crustal Origin for Norwegian Eclogites, Nature (London), 1980, vol. 285, pp. 315321. Griffin, W.L. and Brueckner, H.K., REE, RbSr, and SmNd Studies of Norwegian Eclogites, Chem. Geol., 1985, vol. 52, pp. 249271. Guillot, S., Hattori, K.H., and Sigoyer, J., Mantle Wedge Serpentinization and Exhumation of Eclogites: Insight from Eastern Ladakh, Northwestern Himalaya, Geology, 1999, vol. 28, pp. 199202. Harrison, T.M., Diffusion of 40Ar in Hornblende, Contrib. Mineral. Petrol., 1981, vol. 78, pp. 324331. Holland, T.J., Experimental Determination of the Reaction Paragonite = Jadeite + Kyanite + H2O, and Internally Consistent Thermodynamic Data for Part of the System Na2O Al2O3SiO2H2O, with Application to Eclogites and Blueschists, Contrib. Mineral. Petrol., 1979, vol. 68, pp. 293301. Holland, T.J.B. and Powell, R., An Enlarged and Updated Internally Consistent Thermodynamic Data Set with Uncertainties and Correlations: The System K2ONa2OCaO MgOMnOFeOFe2O3Al2O3TiO2SiO2CH2O2, J. Metamorph. Geol., 1990, vol. 8, pp. 89124. Hsu, K.J., Franciscan Melange as a Model for Eugeosynclinal Sedimentation and Underthrusting Tectonics, J. Geophys. Res., 1971, vol. 76, pp. 11621170. Jamtveit, B., Bucher-Numinen, K., and Austrheim, H., Fluid Controlled Eclogitization of Granulates in Deep Crustal Shear Zones, Bergen Arcs, Western Norway, Contrib. Mineral. Petrol., 1990, vol. 104, pp. 184193. Kirschner, D.L., Cosca, M.A., Masson, H., and Hunziker, J.C., Staircase 40Ar/39Ar Spectra of Fine-Grained White Mica: Timing and Duration of Deformation and Empirical Constrains on Argon Diffusion, Geology, 1996, vol. 24, pp. 747750. Korikovsky, S.P., Contrast Models for ProgradeRetrograde Metamorphic Evolution of Phanerozoic Foldbelts in Collision and Subduction Zones, Petrologiya, 1995, vol. 3, no. 1, pp. 4563. Krabbendam, M., Wain, A., and Andersen, T.B., Pre-Caledonian Granulite and Gabbro Enclaves in the Western Gneiss Region, Norway: Indications of Incomplete Transition at High Pressure, Geol. Mag., 2000, vol. 137, pp. 235255.
PETROLOGY Vol. 10 No. 2 2002

tion Complex, California, Bull. Geol. Soc. Am., 1982, vol. 93, pp. 330345. Cloos, M. and Shreve, R.L., SubductionChannel Model of Prism Accretion, Melange Formation, Sediment Subduction, and Subduction Erosion at Convergent Plate Margins: 1. Background and Description, Pure and Applied Geophys., 1988, vol. 128, pp. 455500. Cohen, A.S., O'Nions, R.K., Siegenthaler, R., and Griffin, W.L., Chronology of the PressureTemperature History Recorded by a Granulite Terrain, Contrib. Mineral. Petrol., 1988, vol. 98, pp. 303311. Dobretsov, N.L., Collision Processes in the Paleozoic Folded Area in Asia and Exhumation Mechanisms, Petrologiya, 2000, vol. 8, no. 5, pp. 451476. Dobretsov, N.L. and Kirdyashkin, A.G., Glubinnaya geodinamika (Deep-Seated Geodynamics), Novosibirsk: Nauchno-Issled. Tsentr, Ob"ed. Inst. Geol. Geofiz. Mineral., Sib. Otd. Ross. Akad. Nauk, 1994. Dodson, M.H., Closure Temperature in Cooling Geochronological and Petrological Systems, Contrib. Mineral. Petrol., 1973, vol. 40, pp. 259274. Ellis, D.J. and Green, E.H., An Experimental Study of the Effect of Ca upon GarnetClinopyroxene FeMg Exchange Equilibria, Contrib. Mineral. Petrol., 1974, vol. 71, pp. 1322. Ellis, S., Beaumont, C., and Pfiffner, O.A., Geodynamic Models of Crustal-Scale Episodic Tectonic Accretion and Underplating in Subduction Zones, J. Geophys. Res., 1999, vol. 104, pp. 1516915190. England, P.S. and Holland, T.J.B., Archimedes and the Tauern Eclogites: The Role of Buoyancy in the Preservation of the Exotic Tectonic Blocks, Earth Planet. Sci. Lett., 1979, vol. 44, pp. 287294. Erambert, M. and Austrheim, H., The Effect of Fluid and Deformation on Zoning and Inclusion Patterns in Poly-Metamorphic Garnets, Contrib. Mineral. Petrol., 1993, vol. 115, pp. 204214. Ernst, W.G., Metamorphism, Partial Preservation and Exhumation of Ultrahigh Pressure Belts, Island Arc, 1999, vol. 8, pp. 125153. Ernst, W.G., Maruyama, S., and Wallis, S., Buoyancy Driven, Rapid Exhumation of the Ultrahigh Pressure Phases in Metamorphosed Continental Crust, Proc. Natl. Acad. Sci. USA., 1997, vol. 94, pp. 95329537. Eugster, H.P., Albee, A.L., Bence, A.E., et al., The Two Phase Region and Excess Mixing Properties of ParagoniteMuscovite Crystalline Solutions, J. Petrol., 1972, vol. 13, pp. 147179. Fossen, H. and Ingdahl, S.E., Tectonostratigraphic Position of the Rocks in the Western Extreme of the Major Bergen Arcs (Fanafjell Nappe), Nor. Geol. Tidsskr., 1988, vol. 67, pp. 5966. Ganguly, J., Cheng, W., and Chakraborty, S., Cation Diffusion in Aluminosilicate Garnets: Experimental Determination in PyropeAlmandine Diffusion Couples, Contrib. Mineral. Petrol., 1998, vol. 131, pp. 171180. Gebauer, D., Lappin, M.A., Grunenfelder, M., Wyttenbach, A., The Age and Origin of Some Norwegian Eclogites, Chem. Geol., 1985, vol. 52, pp. 227247. Gerasimov, V.Yu., An Experimental Study of Reciprocal MgFe Diffusion in Garnets, Dokl. Akad. Nauk SSSR, 1987, vol. 295, pp. 684688.

ECLOGITES OF THE BERGEN ARCS COMPLEX Kretz, R., Symbols for Rock-Forming Minerals, Am. Mineral., 1983, vol. 68, pp. 277279. Krogh, E.G., Evidence for a Precambrian ContinentContinent Collision in the Western Norway, Nature (London), 1977, vol. 75, pp. 7784. Krogh, E.G., The GarnetClinopyroxene FeMg Thermometer--Reinterpretaion of Existing Experimental Data, Contrib. Mineral. Petrol., 1988, vol. 99, pp. 4448. Krogh, E.J. and Carswell, D.A., HP and UHP Eclogites and Garnet Peridotites in the Scandinavian Caledonides, in Ultrahigh Pressure Metamorphism, Coleman, R.G. and Wang, X., Eds., Cambridge: Cambridge Univ. Press, 1995, pp. 244298. Krogh, R.E., The GarnetClinopyroxene Fe2+Mg Geothermometer: An Updated Calibration, J. Metamorph. Geol., 2000, vol. 18, pp. 211219. Lambert, I.B. and Wyllie, P.J., Melting of Gabbro (Quartz Eclogite) with Excess Water to 35 kb Pressure, with Geological Applications, J. Geol., 1972, vol. 80, pp. 693708. Lasaga, A.C., Kinetic Theory in the Earth Sciences, Princeton: Princeton Univ. Press, 1998. Loomis, T.P., Ganguly, J., and Elphick, S.C., Experimental Determination of Cation Diffusivities in Aluminosilicate Garnets II. Multicomponent Simulation and Tracer Diffusion Coefficients, Contrib. Mineral. Petrol., 1985, vol. 90, pp. 4551. Maruyama, S., Liou, J.G., and Terabayashi, M., Blueschists and Eclogites in the World and Their Exhumation, Int. Geol. Rev., 1996, vol. 38, pp. 485594. Medaris, L.G. and Wang, H.F., A ThermalTectonic Model for High-Pressure Rocks in the Basal Gneiss Complex of the Western Norway, Lithos, 1986, vol. 19, pp. 299315. Mork, M.B.E. and Mearns, E.W., SmNd Isotopic Systematics of the GabbroEclogite Transition, Lithos, 1986, vol. 19, pp. 255267. Perchuk, A.L., The New Variant of OmphaciteAlbite Quartz Geobarometer with Allowance for Omphacite and Albite Structural States, Dokl. Ross. Akad. Nauk, 1992, vol. 324, pp. 12861289. Perchuk, A.L., Metamorphism of Kyanite Eclogites from the Krasnaya Skala Terrain (Peredovoi Range, Great Caucasus), Petrologiya, 1993, vol. 1, no. 1, pp. 98109. Perchuk, A.L. and Philippot, P., Rapid Cooling and Exhumation of Eclogitic Rocks from the Great Caucasus, Russia, J. Metamorph. Geol., 1997, vol. 15, pp. 299310. Perchuk, A.L. and Philippot, P., Geospeedometry and Timescales of High-Pressure Metamorphism, Int. Geol. Rev., 2000a, vol. 42, pp. 207223. Perchuk, A.L. and Philippot, P., Subduction Spring: A Record in the Yukon Eclogites, Petrologiya, 2000b, vol. 8, no. 1, pp. 321. Perchuk, A.L. and Philippot, P., The New Sensor of Metamorphism Rate, in Petrografiya na rubezhe XXI veka: itogi i perspektivy. Materialy Vtorogo Vserossiiskogo Petrograficheskogo Soveshchaniya (Petrography at the Turn of 21 Century: Results and Prospects. Second All-Russia Conf. on Petrography), Syktyvkar, 2000c, vol. 3, pp. 8587. Perchuk, A.L. and Varlamov, D.V., A New Type of Prograde Heterogeneity, with Reference to the Eclogites of the Great Caucasus, Geokhimiya, 1995, no. 9, pp. 12961310.
PETROLOGY Vol. 10 No. 2 2002

117

Perchuk, A.L., Gerasimov, V.Yu., and Philippot, P., Theoretical Simulation of the Uplift of Eclogites, Petrologiya, 1996, vol. 4, no. 5, pp. 518532. Perchuk, A.L., Yapaskurt, V.O., and Podlesskii, S.K., Formation Conditions and Uplift Dynamics of Eclogites of the Kokchetav Massif in the Sulu-Tyube Mountain Area, Geokhimiya, 1998, no. 10, pp. 115. Perchuk, L.L., Termodinamicheskii rezhim glubinnogo petrogeneza (Thermodynamic Regime of Deep-Seated Petrogenesis), Moscow: Nauka, 1973. Perchuk, L.L., Derivation of Thermodynamically Consistent Set of Geothermometers and Geobarometers for Metamorphic and Magmatic Rocks, in Progress in Metamorphic and Magmatic Petrology, Perchuk, L.L., Ed., Cambridge: Cambridge Univ. Press, 1991, pp. 93112. Perchuk, L.L., Podladchikov, Yu.Yu., and Polyakov, A.N., P T Paths and Geodynamic Modeling Some Metamorphic Processes, J. Metamorph. Geol., 1992, vol. 10, pp. 311319. Perchuk, L.L., Podlesski, K.K., and Aranovich, L.Ya., Thermodynamics of Some Framework Silicates and Their Equilibria: Application to Geothermobarometry, Progress in Metamorphic and Magmatic Petrology, Perchuk, L.L., Ed., Cambridge: Cambridge Univ. Press, 1991, pp. 131164. Powell, R., Regression Diagnostics and Robust Regression in Geothermometer/Geobarometer Calibration: The Garnet Clinopyroxene Geothermometer Revised, J. Metamorph. Geol., 1985, vol. 3, pp. 231243. Roberts, D. and Gee, D.G., An Introduction to the Structure of the Scandinavian Caledonides, Gee, D.G. and Sturt, B.A., Eds., New York: Wiley, 1985, pp. 5568. Salje, E., Heat Capacities and Entropies of Andalusite and Sillimanite: Influence of Fibrolitization on the Phase Diagram of the Al2SiO5 Polymorphs, Am. Mineral., 1986, vol. 71, pp. 13661371. Scharer, U., UPb and RbSr Dating of Polymetamorphic Nappe Terrain: The Caledonian Joton Nappe, S. Norway, Earth Planet. Sci. Lett., 1980, vol. 49, pp. 205218. Schmeling, H., Monz, R., and Rubie, D.C., The Influence of Olivine Metastability on the Dynamics of Subduction, Earth Planet. Sci. Lett., 1999, vol. 165, pp. 5566. Schreyer, W., Ultradeep Metamorphic Rocks: Retrospective View, J. Geophys. Res., 1995, vol. 100, pp. 83538366. Shmulovich, K.I., Dvuokis' ugleroda v vysokotemperaturnykh protsessakh mineraloobrazovaniya (Carbon Dioxide in High-Temperature Mineral-Formation Processes), Moscow: Nauka, 1988. Smith, D.C., Coesite in Clinopyroxene in the Caledonides and Implications for Geodynamics, Nature (London), 1984, vol. 310, pp. 641644. Smith, D.C., Microcoesites and Microdiamonds in Norway: An Overview, Coleman, R.G. and Wang, X., Eds., Cambridge: Cambridge Univ. Press, 1995, pp. 244298. Sobolev, N.V. and Shatskii, V.S., Mineralogical Indicators of Ultrahigh Pressures and High Temperatures in Metamorphic Rocks, in Eklogity i glaukofanovye slantsy v skladchatykh oblastyakh (Eclogites and Blueschists in Fold Areas), Novosibirsk: Nauka, 1989, pp. 83106. Sobolev, N.V. and Shatsky, V.S., Diamond Inclusions in Garnets from Metamorphic Rocks: A New Environment for Diamond Formation, Nature (London), 1990, vol. 343, pp. 742746.

118

PERCHUK Wenberg, O.P. and Milnes, A.G., Interpretation of Kinematic Indicators along the Northeastern Margin of the Bergen Arcs System: A Preliminary Field Study, Nor. Geol. Tidsskr., 1994, vol. 74, pp. 166173. When Continents Collide: Geodynamics and Geochemistry of Ultrahigh-Pressure Rocks, Hacker, B.R. and Liou, J.G., Eds., Kluwer Academ. Publisher, 1998. Zhang, R., Hirajuma, T., and Banno, S., Petrology of UltraHigh-Pressure Rocks from Southern SuLu Region, Eastern China, J. Metamorph. Geol., 1995, vol. 13, pp. 659677.

Terry, M.P., Robinson, P., Hamilton, M.A., and Jercinovich, M.J., Monazite Geochronology of UHP and HP Metamorphism, Deformation, and Exhumation, Nordoyane, Western Gneiss Region, Norway, Am. Mineral., 2000, vol. 85, pp. 16511664. Wain, A., Waters, D., Jephcoat, A., and Olijynk, H., The High-Pressure to Ultrahigh-Pressure Eclogite Transition in the Western Gneiss Region, Norway, Eur. J. Mineral., 2000, vol. 12, pp. 667687.

PETROLOGY

Vol. 10

No. 2

2002