High-pressure/ultrahigh-pressure eclogites from the Hong’an Block, East-Central China: geochemical characterization, isotope disequilibrium and geochronological controversy

O R I G I N A L P A P E R

Bor-ming Jahn Æ Xiaochun Liu Æ Tzen-Fu Yui N. Morin Æ M. Bouhnik-Le Coz

High-pressure/ultrahigh-pressure eclogites from the Hong’an Block,

East-Central China: geochemical characterization, isotope

disequilibrium and geochronological controversy

Received: 2 September 2004 / Accepted: 7 March 2005 / Published online: 16 April 2005
Springer-Verlag 2005

Abstract The Hong’an Block (western Dabieshan) ex-poses a series of HP/UHP metamorphic rocks, with a S-to-N distribution from blueschist–greenschist, kyanite-free, to kyanite- and coesite-bearing eclogites. The available age data are inconclusive that hinder our understanding of the tectonic evolution of the Block. The metamorphic temperatures in the Hong’an Block (Tmeta
700 to 500C) are lower by 50–150C than that

of the Dabie and Sulu terranes. In this work, we undertook new trace element and Sr–Nd–O isotopic analyses on minerals in order to gain more insight into the geochronological problems. The results are as fol-lows: (1) Trace element distribution patterns suggest that garnet and omphacite in many cases are out of chemical equilibrium; and the presence of high-temperature LREE-rich mineral inclusions (e.g., epidote) in garnet and omphacite has contributed to isotope disequilib-rium. (2) Sm–Nd isotope analyses yielded no isochron ages for the Hong’an eclogites. (3) Rb–Sr isotope anal-yses gave mixed results; in some cases, coexisting min-erals are completely out of isotope equilibrium, and in others, isochron relationship is established, yielding ages

from 210 Ma to 225 Ma. The pattern of Rb–Sr isotope disequilibrium appears to be independent of the petro-logical and O-isotope temperatures. (4) In contrast to the unequilibrated Sm–Nd isotopic systems, oxygen isotopes of the eclogite minerals seem to have attained isotope equilibrium or near-equilibrium. Oxygen isotope temperatures are comparable with petrological temper-atures. However, this is an apparent feature due to mass balance constraints. (5) Whole-rock d18O values show a large variation from +10& to 8&, suggesting that their protoliths have undergone very different processes of water–rock interaction. In view of the overall geo-chronological information, we conclude that the HP/ UHP metamorphism in the Hong’an Block took place in the Triassic at about 220–230 Ma, as observed in the Dabie and Sulu terranes. The significance of published Paleozoic dates (450–300 Ma) for the Xiongdian eclogite is not clear. However, any hypotheses advocating two periods of UHP metamorphic events for the same tec-tonic unit or in the same locality are not constrained by the geochronological data.

Introduction

The Qinling-Dabie orogen marks the suture zone be-tween the Sino-Korean and Yangtze cratons. High-pressure (HP) and ultrahigh-High-pressure (UHP) metamor-phic rocks in this orogen extend from eastern Qinling, through Tongbaishan and Dabieshan, across the Tanlu fault, and then northeastwards to the Sulu region (Fig.1). The Hong’an Block, also known as western Dabieshan, exposes a series of high P/T metamorphic rocks, with a S-to-N distribution from blueschist/ blueschist–greenschist, kyanite-free and kyanite- and coesite-bearing eclogite facies rocks. The Block, as compared with other parts of the Qinling–Dabie orogen, has been considered to conserve better archives for the understanding of the tectonic evolution of this orogen

Electronic Supplementary Material Supplementary material is available for this article at http://dx.doi.org/10.1007/s00410-005-0668-5

Editorial Responsibility: J. Hoefs B. Jahn (&) Æ T.-F. Yui

Institute of Earth Sciences, Academia Sinica, Taipei, 115, Taiwan

E-mail: jahn@earth.sinica.edu.tw B. Jahn Æ X. Liu

Department of Geosciences, National Taiwan University, P. O. Box 13-318, Taipei, 10699, Taiwan

X. Liu

Institute of Geomechanics, CAGS, Beijing, 100081, China

N. Morin Æ M. B.-L. Coz

Ge´osciences Rennes, Universite´ de Rennes 1, 35042 Rennes Cedex, France

for three main reasons (e.g., Eide and Liou2000): (1) it is least affected by the thermal and structural overprint imposed on much of the Dabie terrane during the intrusion of voluminous Cretaceous granitoids; (2) HP eclogites are widespread in the Hong’an Block and they often preserve prograde metamorphism, and can be di-rectly linked to the blueschist/blueschist–greenschist rocks; (3) in comparison with the Dabie terrane, the better exposure of blueschist/blueschist–greenschist fa-cies rocks offers opportunities for simultaneous struc-tural and metamorphic analyses.

The collision between the two Precambrian cratons, Yangtze and Sino-Korean, leading to the formation of UHP metamorphic terranes is generally accepted to have taken place in the Triassic (Ames et al.1993,1996; Li et al. 1993, 1994,2000; Eide et al.1994; Chavagnac and Jahn 1996; Rowley et al. 1997; Maruyama et al. 1998; Hacker et al.1998, 2000; Chavagnac et al. 2001; Ayers et al. 2002; Jahn et al. 2003a; Yang et al.2003). However, the published age data of eclogites from the Hong’an Block have caused much controversy. Some investigators argued that the HP eclogites in the north-ern eclogite zone (so-called Huwan shear zone) of the Block could have formed in a Paleozoic subduction zone of an oceanic plate beneath the Sino-Korean craton (Jian et al.1997,2000; Xu et al.2000; Li et al.2001; Sun et al. 2002). Such an interpretation appears to be in conflict with the scenario that the orogen-scale structure

of Dabieshan is a huge antiform with UHP eclogites at the core and HP eclogites on the two limbs (Hacker et al. 1998,2000; Eide and Liou2000). Indeed, the nature and age of the northern HP eclogite zone and its relation to other HP/UHP eclogites have become a highly contro-versial issue in the last few years (Fu et al. 2002; Gao et al. 2002; Jahn and Liu 2002; Jahn et al. 2003b; Ratschbacher et al.2003).

The motivation of the present work is to resolve the apparent conflict and to find a satisfactory explanation for the complex age patterns. During our isotope anal-yses, we have found that isotopic disequilibrium is the major cause of the age problem. In order to better evaluate the isotope behavior, we have also made com-prehensive geochemical and oxygen isotope analyses on whole-rock and constituent minerals. In the end, we will summarize the available meaningful age data and con-clude that the Hong’an Block shared the same meta-morphic evolution as the Dabie and Sulu terranes. The controversial Paleozoic ages reported in the literature for the eclogites from the Xiongdian locality will be discussed.

Geological setting and age controversy

The Hong’an Block is separated from Tongbaishan in the west by the Dawu Fault, and from the Dabie

Fig. 1 Simplified geological map of the Hong’an Block (also known as western Dabieshan) showing different tectonic units and five eclogite zones. The inset shows the location of the Hong’an Block with respect to the Qinling-Dabie-Sulu orogen of China. The metamorphic grades generally increase from SW to NE. Eight sampling localities are shown

terrane by the Shangma Fault (Fig.1). A variety of HP and UHP metamorphic rocks and structures are well preserved and the Cretaceous tectonic and ther-mal overprint is generally weak or absent. Based on the petrological characteristics, six major lithotec-tonic units, comprising five eclogite zones, can be distinguished (Hacker et al. 1996; Liou et al. 1996; Zhong et al. 1999, 2001; Liu et al. 2004a). They are, from north to south, the Nanwan flysch unit, the Balifan tectonic me´lange unit, the Huwan HP unit, the Xinxian UHP unit, the Hong’an HP unit, and the Mulanshan blueschist–greenschist unit (Fig.1). The details about the lithological characteristics of all individual units can be found in Liu et al. (2004a). In contrast to the well-dated Dabie and Sulu terr-anes, geochronological studies in the Hong’an Block are scarce. The earlier Ar–Ar dating work of Eide et al. (1994) suggests that the Hong’an Block is a part of the Dabieshan Triassic HP/UHP metamorphic belt. Since then, several attempts of U–Pb zircon dating on eclogites from the northwestern corner of the Huwan unit have produced a wide range of metamorphic ages from Silurian to Triassic (Jian et al. 1997, 2000; Xu et al. 2000; Li et al. 2001; Sun et al. 2002; Gao et al. 2002). Presence of two distinct Paleozoic U–Pb zircon ages of about 400 and 300 Ma for the Xiongdian eclogite was first reported (TIMS method) and then reconfirmed (SHRIMP method) by Jian and his col-laborators (Jian et al. 1997, 2000). The date of ca. 400 Ma was interpreted as the minimum age for the eclogite facies metamorphism and the 300 Ma as the time of retrograde metamorphism. Meanwhile, Xu et al. (2000) obtained phengite40Ar/39Ar dates of 350– 430 Ma for the same eclogite, and interpreted the dates as a retrograde metamorphic age. More recently, using SHRIMP U–Pb dating, Sun et al. (2002) ob-tained a Carboniferous metamorphic age of 309±3 Ma for the zircon rims of three eclogite sam-ples from the Xiongdian and adjacent Hujiawan localities. The age was interpreted as the time of a HP metamorphic event. In addition, ages ranging from 350 Ma to 440 Ma were also obtained for zircon cores from the same samples. Sun et al. (2002) interpreted these ages as the time of igneous intrusion or low-pressure metamorphism in the southern margin of the Sino-Korean craton. These authors also reported a protolith age of 752±17 Ma and a Triassic meta-morphic age of 232±10 Ma for an eclogite from Xuanhuadian, about 5-km south of Xiongdian. They concluded that two sutures, a Carboniferous and a Triassic, existed on the two sides of the Huwan unit, and insisted that the Carboniferous eclogites were not affected by the Triassic eclogite facies event. However, Gao et al. (2002) simultaneously reported a Triassic SHRIMP zircon age of 216±4 Ma for precisely the same Xiongdian eclogite. It is clear that inspite of using the same SHRIMP U–Pb dating technique on the same eclogite, the results have not always been

consistent. Table 1 Sampling localities, mineral assemblages and peak metamorphic P  T conditions of the dated rocks from the Hong’an Block Unit (zone) Sample No. Location Coordinates Rock type Mineral assemblage Peak P  T eastimates T ( C) P (kbar) Huwan unit (zone I) XHD07BJ01-107 Xiongdian 31 45.14 ¢N114 28.50 ¢E Eclogite Grt + Omp + Amp + Phen + E p + Qtz + C c + Rt 680–700 15–17 QJH01 Qianjinhepeng 31 45.96 ¢N114 40.67 ¢E Eclogite Grt + Omp + Amp + E p + Qtz + R t 610±40 > 1 2 Xinxian unit (zone III) TP03 Tianpu 31 31.96 ¢N114 57.28 ¢E Eclogite Grt + Omp + K y + Gin + phen + Ep + P g + Cs/Qtz + R t 570±30 29±3 P260 Xiongjiazhui 31 27.79 ¢N114 48.83 ¢E Eclogite Grt + Omp + K y + Gln + Phen + Ep + Cs/Qtz + R t 670±60 26±4 Hong’an unit (zone IV) QLP08 Dongyuemiao 31 23.65 ¢N114 45.77 ¢E Eclogite Grt + Omp + Amp + Phen + E p + Pg + Qtz + R t 530±30 20±2 Hong’an unit (zone V) GQ01 Gaoqiao 31 12.96 ¢N114 34.55 ¢E Eclogite Grt + Omp + Amp + Phen + E p + Qtz + R t + Ttn 510±40 20±1 Mulanshan unit BJ01-101 Duanjiagang 31 04.03 ¢N114 19.98 ¢E Glaucophane Schist Grt + Gln + Phen + E p + Pl + Kfs + Qtz + Ttn 350–450 7–10 Hong’an unit BJ01-115 Marble Hill 31 26.25 ¢N119 39.55 ¢E Marble Cc + Mus + Qtz 410±50 5–10 Mineral abbreviations: Amp amphibole, Cc calcite, Cs coesite, Ep epidote, Gln glaucophane, Grt garnet, Kfs K-feldspar, Ky kyanite, Mus muscovite, Omp omphacite, Phen phen gite, Pg paragonite, Pl plagioclase, Qtz quartz, Rt rutile, Ttn = titanite

Petrology and P–Tconditions of analyzed samples

The sampling localities are shown in Fig.1, and the mineral assemblages and peak metamorphic conditions are summarized in Table1. The PT calculations were performed using the garnet–clinopyroxene thermometry (Powell 1985), jadeite–albite–quartz and garnet–clino-pyroxene–phengite barometry (Carswell and Harley 1990; Carswell et al. 1997) and THERMOCALC pro-gram (Powell et al. 1998). For details, see Liu et al. (2004a). Six basaltic eclogites were analyzed in this work. In addition, a glaucophane schist (sample BJ01-101) from Duanjiagang locality (near Mulanshan) in the Mulanshan blueschist–greenschist unit and a marble (sample BJ01-115) from Shihuishan (Marble Hill) in the Hong’an unit were also analyzed. Their petrographic characteristics are given in Appendix 1.

Briefly, the peak metamorphic temperatures of eclogites from the Hong’an Block range from ca. 500 to 700C, which are lower than those recorded in the Dabie and Sulu terranes at corresponding pressures (Fig.2; Zhang and Liou1994; Liou et al.1996, 2004a; Eide and

Liou2000). The eclogites from the Hong’an and Huwan units have often been referred to as ‘‘cold eclogites’’ and no coesite has been identified. Note that the temperature range (500–700C) is roughly at the threshold of the blocking temperatures for two important chronometers involved in study of UHP metamorphic rocks: garnet Sm–Nd and white-mica Rb–Sr systems (see later for discussion).

Analytical procedures

Elemental abundances

Major and trace element abundances were measured using XRF (3080E) at the National Research Center for Geoanalysis, Chinese Academy of Geological Sciences, and ICP-MS (Elan 6100DRC) at the Key Laboratory of Continental Dynamics, Northeast University (Xi’an), respectively. The uncertainties for trace elements vary from <5% to <20% depending on the concentration levels.

Chemical analyses of mineral grains were performed using an ICP-MS (HP-4500) laser ablation technique in Rennes. Mineral grains were analyzed by the line-scan mode with a continuous laser shot (energy 15–10 Hz) for 90 s, with a beam size of 25 lm, moving at a speed of 1 lm/s. A total of five to ten tracks were measured for each individual mineral. Elemental concentrations were obtained by calibration with respect to standards NIST 610-612-614-616. The calibration range followed the concentration levels of individual minerals. An internal standardization was adopted using 29Si, assuming SiO2content equal to 72%, and all elemental

concentrations were calculated and normalized using the ratio of sample SiO2to NIST SiO2. In the course of

the analyses, we assumed the SiO2 content to be 39%

for garnet, 56% for clinopyroxene and 45% for amphibole.

Sr–Nd isotopic analyses

The analytical procedures for Sr–Nd isotopic analyses are similar to those reported earlier (e.g., Jahn et al. 1996; Chavagnac and Jahn 1996). Analytical precision, Sr–Nd isotope standard and normalization values, and blank levels can be found in the footnotes of data tables. The decay constants (k) used are 0.0142 Ga1for 87Rb and 0.00654 Ga1for147Sm. The depleted mantle based model ages (TDM) were calculated assuming a linear

isotopic evolution of the mantle from
Nd(T) = 0 at

4.56 Ga to +10 at the present time.

TDM ¼ 1=k  ln  143_Nd=144 Nd   s  0:51315 147_Sm=144 Nd   s0:2137   0 5 10 15 20 25 30 35 Temperature (0C) Pressure (kbar) 300 400 500 600 700 800 900 Si 3.5 (Phen) EA AM LGR HGR AEC LEC ZEC DEC Jd₅₀ + K y + H 2O Pg Cs Qtz Jd + Qtz_Ab BS GS Hong'an Dabie Sulu

Fig. 2 Metamorphic PT paths of three UHP terranes in the Dabie-Sulu orogen (modified after Zhang and Liou1994; Eide and

Liou 2000; Liu et al. 2004a). Note that the PT conditions

generally decrease from the Sulu through Dabie to Hong’an. The petrogenetic grid is from Liou et al. (1998). The reaction curves for Qtz = Cs, Pg = Jd50+ Ky + H2O, Ab = Jd + Qtz and phengite Si = 3.5 are from Bohlen and Boettcher (1982), Newton (1986), Holland (1980) and Zhang and Liou (1994), respectively. Abbre-viations for facies names: AM amphibolite; BS blueschist; EA epidote amphibolite; GS greenschist; HGR high-pressure granulite; LGR low-pressure granulite; AEC amphibole eclogite; DEC dry eclogite; LEC lawsonite eclogite; ZEC zoisite eclogite

Sm–Nd isochron ages were computed using the ISOPLOT software of Ludwig (1999). Input errors (2r) for age computations are87Rb/86Sr = 2%,87Sr/86Sr = 0.005%,147Sm/144Nd = 0.2%,143Nd/144Nd = 0.005%. These input errors of 0.005% for 87Sr/86Sr and

143

Nd/144Nd were estimated from the reproducibility of long-term measurements on standards and they repre-sent about five times the in-run precision (2rm) reported

in Table 4. The quoted error in calculated ages throughout this paper represents two standard devia-tions (2r).

O isotopic analyses

O-isotope analyses were performed with a MAT 252 mass spectrometer at the Institute of Earth Sciences, Academia Sinica, Taipei. Garnet, omphacite, epidote (zoisite) and amphibole were analyzed by the CO2

la-ser-fluorination method (Sharp 1990). The analytical precision is slightly better than ±0.1&. UWG-2 garnet standard, with a recommended d18O value of +5.8& relative to SMOW (Valley et al.1995), was employed to normalize the daily data. Quartz and phengite, by contrast, were analyzed by the conventional BrF5

method (Clayton and Mayeda, 1963). The precision is better than ±0.2&. The mean d18O-value for NBS-28 standard obtained during the present study was +9.6&.

Geochemical characterization

Bulk chemical compositions of eclogites

Eclogites from the Dabie terrane can be generally sep-arated into three types based on their field occur-rence—Type I are gneiss-hosted enclaves or layers; Type II are interlayers with or enclaves within marble or calc-silicate rocks, and Type III represent members of layered mafic–ultramafic intrusions, such as the Bixiling and Maowu complexes, or simply in association with ultra-mafic rocks with no clear genetic relationship (Wang et al. 1990; Liou et al. 1996; Jahn 1998). The studied samples from the Hong’an Block are of Type I, and their chemical compositions are quite comparable to the same type of eclogites from the Dabie terrane—all of them show characteristic REE patterns with light-REE enrichment; and depletion in Nb (and Ta) with respect to La in spidergrams. These are also typical features observed in the mafic components of Precambrian gneiss terranes, such as amphibolites or basic granulites (Jahn 1990,1998).

The eclogites are mainly of basaltic composition, but they show a wide range of major and trace element abundances (e.g., Jahn 1998; Jahn et al. 2003b). This suggests their multiple origins and derivation from het-erogeneous mantle sources. However, unusual compo-sitions could be an artifact caused by biased analyses

conducted on banded rocks. Eclogites and ultramafic rocks that have recrystallized in UHP metamorphic conditions often show coarse-grained texture with dis-tinct mineral banding of variable scale. It is very difficult to obtain a truly representative bulk composition for the protolith of a banded eclogite. Fine-grained or homo-geneous textured eclogites also occur, but they seem to be minor in comparison with heterogeneous textured facies. Eclogites from the Dabie and Sulu terranes have a wide range of SiO2contents from 36% to 60%, although

the majority (>70% of all types) still have basaltic or gabbroic compositions (SiO2 = 45–52%; Jahn 1998).

Higher silica eclogites (SiO2 ‡ 53%) suggest that their

protoliths are more differentiated, and lower silica ones (£ 45%) implies a ‘‘cumulate’’ nature of their protoliths

Fe+Ti Al Mg High-Fe tholeiite basalt High-Mg tholeiite basalt TH ser ies CA ser ies

Zone III eclogites Zone II eclogites Zone I eclogites Li et al. (2001) Zone IV eclogites Zone V eclogites Metagabbros Andesite Andesite Basalt Monzorite 40 50 60 2 SiO2 Na 2 O+K 2 O 6 10 0 4 8

Zone III eclogites Zone II eclogites Zone I eclogites Li et al. (2001) Zone IV eclogites Zone V eclogites Metagabbros Trachy-basalt Phono-tephrite Trachy-andesite Basaltic andesite Andesite Basalt Tephrite Picro-basalt (a) (b)

Fig. 3 a Total alkalis versus SiO2 (TAS) classification (Middlemost 1994) diagram for magmatic rocks. Most eclogites fall in the field of basalt; whereas eclogites from Xiongdian of Eclogite Zone I (Huwan unit), including those reported by Li et al. (2001), show higher silica contents, and fall in andesite and basaltic andesite fields. b Cation plot of Jensen (1976) for metaigneous rocks. All eclogites fall in high-Fe basalt of the tholeiitic series, or in basalt of the calc-alkaline series. The ‘‘andesites’’ in TAS diagram also plot in the basalt fields

Table 2 Chemical compositions of eclogites from the Hong’an Block

Huwan unit (eclogite zone I) Xinxian unit (eclogite zone II) Xinxian unit

(eclogite zone III)

Sample No. HW01 P073 QJH01 XHD07 BJ01-107 P101 P114 P126 QLP14 P147 P152 TP03 Locality 1 2 3 4 4 5 6 7 8 9 10 11 (Major oxides, in %) SiO2 49.32 47.44 52.64 58.08 55.88 48.26 46.75 47.06 47.83 50.89 47.48 52.17 TiO2 2.04 1.79 1.03 0.38 0.32 2.00 2.71 3.19 2.68 1.15 1.17 1.04 Al2O3 14.69 14.59 16.89 14.26 14.37 14.00 14.20 13.25 14.21 17.02 18.15 16.65 Fe2O3 4.77 3.37 4.73 1.66 8.48 5.83 5.25 4.12 2.56 5.80 6.63 4.57 FeO 8.71 9.39 6.00 6.30 8.73 9.21 11.48 15.79 4.58 5.91 5.96 MnO 0.20 0.22 0.22 0.15 0.14 0.23 0.25 0.28 0.27 0.14 0.25 0.18 MgO 5.90 6.47 5.86 5.05 5.87 5.57 6.13 5.62 4.32 5.06 4.34 6.11 CaO 9.67 10.47 7.72 8.46 9.25 9.77 10.08 9.85 8.73 8.73 10.11 7.37 Na2O 3.10 3.15 3.09 2.72 3.48 4.56 4.00 2.79 2.24 4.84 3.84 3.78 K2O 0.75 1.17 0.70 0.97 0.86 0.17 0.13 0.26 0.32 0.29 0.04 0.54 P2O5 0.08 0.24 0.29 0.08 < D.L. 0.19 0.48 0.63 0.43 0.37 0.25 0.24 CO2 0.35 0.09 0.18 0.35 0.26 0.18 0.12 0.44 0.27 0.18 H2O+ 0.96 1.22 0.90 1.04 1.19 0.72 0.94 0.92 0.76 1.00 0.92 0.94 Total 100.54 99.61 100.25 99.50 99.84 100.29 100.13 99.63 100.26 100.31 99.36 99.73 mg value 47.3 50.8 53.1 56.2 60.4 44.1 46.6 42.3 32.1 50.6 42.0 54.6 (Trace elements, in ppm) Cs 0.3 2.3 0.8 1.5 2.5 0.1 0.05 0.4 0.4 0.4 0.07 0.10 Rb 18.5 49.2 13.9 22.1 20.4 5.1 1.2 10.8 16.9 10.5 1.2 3.27 Sr 364 66.3 378 165 107 295 94 263 152 437 902 374 Ba 289 249 358 708 417 64.1 65.8 196 146 145 26.1 414 Be 0.6 0.8 0.6 0.6 < D.L. 1.2 1.6 1.4 1.1 0.8 0.4 0.4 Nb 3.0 3.6 1.5 2.3 1.2 1.6 9.4 10.8 15.7 2.4 1.7 11.4 Ta 0.20 0.23 0.09 0.23 0.14 0.10 0.59 0.70 1.02 0.14 0.09 0.81 Th 2.2 0.41 0.71 0.20 < D.L. 1.2 1.0 1.1 5.3 1.2 1.0 0.39 U 0.27 0.11 0.14 0.49 0.22 0.18 0.28 0.31 0.91 0.14 0.37 0.13 Pb 9.1 3.8 5.9 11.5 4.3 3.8 2.8 4.2 5.3 5.7 19.5 12.3 Zr 68 98 68 38 26 51 180 212 293 76 61 64 Hf 1.8 2.6 1.9 1.1 0.9 1.5 4.4 5.3 7.2 1.9 1.6 1.9 Y 19.9 32.6 20.4 11.4 10.2 19.4 44.7 56.7 53 18.4 18.4 28.8 V 648 391 224 190 205 577 365 435 184 115 210 429 Sc 51 52 31 32 50 47 55 29 38 37 63 Cr 10.6 110 123 445 474 19.1 162 110 70.4 88 37.3 3.5 Co 50 40 42 30 33 37 45 39 33 34 31 38 Ni 19 32 53 92 102 22 49 40 84 31 12 22 Cu 40 76 48 60 77 21 69 50 288 30 33 91 Zn 110 124 97 68 72 142 157 178 123 142 93 219 Ga 19 20 15 13 14 20 19 22 15 17 19 16 Sn 0.7 1.0 0.6 0.7 < D.L. 0.8 1.3 1.6 0.7 0.9 La 12.7 8.52 8.39 2.93 2.03 6.04 18.9 21.6 31.8 7.56 10.6 8.93 Ce 25.7 20.9 18.9 6.11 4.41 13.1 44.2 51.2 64.9 17.4 22.1 22.2 Pr 3.2 3.2 2.7 0.9 0.7 2.0 6.3 7.6 8.0 2.6 2.9 3.5 Nd 14.3 16.3 13 4.52 3.54 9.49 30.4 36.9 36.7 12.5 14.1 18.9 Sm 3.24 4.56 3.13 1.26 0.96 2.35 7.58 9.5 8.42 3.06 3.24 4.84 Eu 1.06 1.51 1.07 0.54 0.34 0.76 2.19 2.9 2.14 1.01 1.21 2.43 Gd 3.07 4.28 2.79 1.24 1.29 2.12 6.78 8.68 9.61 2.65 3.22 4.52 Tb 0.5 0.76 0.47 0.23 0.23 0.39 1.12 1.46 1.59 0.44 0.49 0.77 Dy 3.33 5.22 3.2 1.66 1.63 2.96 7.48 9.7 9.46 2.99 2.92 5.1 Ho 0.73 1.17 0.73 0.39 0.34 0.72 1.66 2.13 1.94 0.67 0.61 1.07 Er 1.93 3.07 1.99 1.09 1.07 1.92 4.29 5.54 5.01 1.73 1.65 2.52 Tm 0.28 0.46 0.3 0.17 0.16 0.28 0.63 0.8 0.77 0.25 0.26 0.34 Yb 1.95 3.18 2.07 1.23 1.11 1.96 4.26 5.38 4.98 1.71 1.79 2.19 Lu 0.3 0.48 0.32 0.2 0.18 0.3 0.65 0.83 0.79 0.27 0.28 0.33 Nb/Ta 15.0 15.8 17.3 9.8 8.8 16.4 15.8 15.4 15.4 17.1 19.0 14.1 Th/U 8.2 3.7 5.1 0.4 6.7 3.5 3.6 5.9 8.4 2.6 3.0 Zr/Hf 38.6 37.3 36.5 33.3 29.8 34.5 41.2 40.2 40.6 39.0 37.8 34.5 Th/Ta 11.1 1.8 8.3 0.9 12.1 1.7 1.6 5.2 8.4 10.7 0.5 Nb/U 11.1 33.1 10.6 4.6 5.5 9.1 33.4 34.8 17.3 17.1 4.6 87.7 Ba/La 22.8 29.2 42.7 242 205 10.6 3.5 9.1 4.6 19.2 2.5 46.4 La/Nb 4.2 2.3 5.6 1.3 1.7 3.7 2.0 2.0 2.0 3.2 6.2 0.8 Ti/V 19 27 28 12 9 21 45 44 87 60 33 15

For Locality names: 1. Huwan, 2. Qilongshan, 3. Qianjinhepeng, 4.Xiongdian, 5. Longgushi, 6. Xinwu, 7. Yanjiahe, 8. Sanwu, 9. Ta’ ergang, 18. Luwang, 19. Huajiahe, 20. Gaoqiao, 21. Shuaijiahe, 22. Shujiahe, 23. Wangmuguan, 24. Daleiwa, 24. Duanjiagang

Hong’an unit (z. IV)

Hong’an unit (eclogite zone V) Balifan unit (metagabbro) Blueschist

GJH10 P220 P256 P260 QLP08 XZD01 LW05 HJH01 GQ01 P390 SJH05 QJH03 P032 BJ01-101 12 13 14 15 16 17 18 19 20 21 22 23 24 25 42.32 47.82 49.82 49.40 46.72 47.42 48.36 48.96 48.32 48.26 42.27 49.96 44.12 66.74 1.27 0.65 1.16 1.61 2.60 1.90 2.85 1.71 2.86 2.02 3.25 0.51 1.17 0.72 17.64 20.26 17.21 16.40 14.70 15.07 12.48 14.14 14.14 14.31 13.31 14.81 18.13 16.50 9.71 3.43 6.79 2.82 3.69 2.34 4.56 2.07 3.24 2.35 7.82 1.60 5.48 3.33 2.93 6.41 2.87 7.85 10.62 9.11 10.47 11.19 11.30 10.85 10.65 3.86 5.16 0.22 0.16 0.09 0.18 0.26 0.20 0.26 0.23 0.26 0.23 0.20 0.10 0.15 0.13 3.40 5.75 6.80 6.66 5.98 7.40 4.95 6.47 4.68 6.74 6.65 10.15 8.12 0.92 17.43 11.93 9.73 10.14 10.25 11.18 10.56 11.03 9.19 10.63 11.22 16.59 12.29 1.82 1.04 2.43 3.41 2.47 3.28 3.12 3.39 2.68 2.82 2.95 2.07 1.54 1.91 5.86 0.13 0.08 0.07 0.57 0.12 0.12 0.05 0.07 0.93 0.87 0.37 0.05 0.44 2.51 0.66 0.08 0.06 0.25 0.42 0.34 0.55 0.14 0.58 0.33 0.15 0.01 0.09 0.19 0.35 0.53 0.18 0.35 0.18 0.35 0.18 0.26 0.18 0.05 0.09 0.27 0.18 2.20 1.00 1.22 0.60 1.02 1.08 0.64 1.18 1.00 0.96 1.58 0.92 3.14 1.20 99.30 100.53 99.41 99.30 99.84 99.63 99.30 100.13 99.50 100.55 99.63 100.37 100.38 99.92 36.6 54.5 60.0 55.9 45.9 56.6 40.2 49.5 39.5 50.7 42.7 79.1 61.4 37.8 0.02 0.1 0.1 0.5 0.2 0.06 0.04 0.2 2.1 1.6 0.7 0.1 0.1 2.7 1.06 2.0 1.2 15.8 2.4 2.2 0.9 1.2 26.1 34.4 16.1 1.1 7.1 59.6 2977 536 421 236 284 343 377 89.3 225 260 270 177 915 226 27 40.2 13.9 251 62.1 20.5 212 60.9 350 287 234 22.6 177 961 1.3 0.4 0.7 0.9 2.7 0.8 1.2 0.7 1.2 0.7 0.5 0.1 0.5 2.4 9.1 0.3 3.6 4.9 6.8 4.0 7.7 7.5 7.4 3.5 4.3 0.3 2.4 12.0 0.47 0.02 0.23 0.31 0.46 0.27 0.48 0.49 0.47 0.22 0.30 0.03 0.14 0.90 3.3 0.2 1.5 1.2 0.84 0.56 1.1 0.73 1.2 0.41 0.49 0.05 0.53 9.4 0.51 0.08 0.57 0.22 0.27 0.15 0.33 0.31 0.32 0.11 0.13 0.01 0.11 2.08 22.0 3.5 8.6 5.6 28.6 4.6 7.3 1.6 3.1 7.9 7.7 1.3 6.3 19.6 117 10 100 97 146 88 210 99 218 137 49 12 43 311 3.0 0.3 2.7 2.5 3.7 2.3 5.4 2.9 5.5 3.5 1.5 0.51 1.3 8.0 24.5 5.9 23.9 30.2 43 27.8 54.5 31.9 56.6 38 18.3 8.7 12.6 62.3 246 365 332 277 343 277 429 364 402 307 760 158 268 27 20 35 31 40 45 38 50 50 49 42 43 41 30 5.1 35.5 385 186 123 172 36.1 114 85.3 185 14.1 1583 254 5.1 19 37 45 43 48 46 36 37 32 36 60 27 53 2 14 37 144 63 50 91 21 42 35 55 53 128 109 < D.L. 33 27 41 50 46 51 44 72 42 44 145 34 47 < D.L. 86 71 94 109 139 108 174 123 171 106 137 27 64 115 25 16 20 19 25 18 21 20 22 18 19 11 16 23 1.3 0.6 1.3 1.2 1.3 0.8 1.7 1.5 1.6 2.5 34.9 2.74 11.6 12.7 13.5 9.17 21 7.9 21.2 11.8 6.52 1.33 9.27 51.5 68.2 5.76 23.9 28.2 32.5 21.9 48.1 18.3 48.6 27.9 15.4 3.06 19.7 108.4 9.2 0.8 3.3 3.9 4.9 3.3 7.0 2.9 7.2 3.9 2.1 0.6 2.7 13.5 39.9 3.89 15.4 19 24.8 16.5 34.4 14.8 35.1 20.5 10.8 3.48 12.8 54.8 7.36 1.00 3.85 4.95 6.56 4.26 8.67 4.27 8.75 5.47 2.91 1.24 2.86 12.1 2.14 0.48 1.54 1.60 2.03 1.36 2.59 1.34 2.65 1.98 1.10 0.54 1.02 3.39 5.86 0.88 3.5 4.5 6.2 3.9 8.0 4.1 8.2 6.1 3.3 1.5 2.7 11.3 0.75 0.14 0.59 0.77 1.08 0.68 1.35 0.76 1.41 1.04 0.54 0.28 0.39 1.79 4.31 0.98 3.89 4.98 7.3 4.58 9.1 5.26 9.46 6.42 3.23 1.68 2.18 11.15 0.84 0.22 0.86 1.11 1.6 1 1.98 1.16 2.07 1.38 0.68 0.35 0.43 2.12 2.19 0.57 2.1 2.9 4.14 2.62 5.27 3.01 5.43 3.64 1.7 0.85 1.09 5.92 0.31 0.08 0.29 0.42 0.6 0.37 0.78 0.44 0.8 0.56 0.26 0.12 0.17 0.83 2.12 0.6 1.79 2.82 4.11 2.57 5.2 3.01 5.42 3.63 1.61 0.75 1.07 5.45 0.32 0.09 0.26 0.44 0.62 0.38 0.8 0.46 0.83 0.57 0.25 0.11 0.16 0.86 19.3 17.0 15.7 15.8 14.7 14.8 16.0 15.2 15.7 16.0 14.2 10.7 17.1 13.4 6.5 2.9 2.6 5.6 3.1 3.7 3.3 2.4 3.7 3.7 3.8 5.0 4.8 4.5 39.4 31.3 37.7 38.6 39.4 38.9 39.1 34.7 39.4 39.1 32.8 24.3 32.7 38.8 7.0 11.5 6.4 4.0 1.8 2.1 2.3 1.5 2.5 1.9 1.6 1.7 3.8 10.5 17.7 4.4 6.3 22.3 25.0 26.6 23.2 24.0 23.1 32.1 32.8 32.0 21.7 5.8 0.8 14.7 1.2 19.8 4.6 2.2 10.1 7.7 16.5 24.3 35.9 17.0 19.1 18.7 3.9 8.1 3.2 2.6 2.0 2.3 2.7 1.1 2.9 3.3 1.5 4.2 3.9 4.3 31 11 21 35 45 41 40 28 43 39 26 19 26 159

or a portion of rock rich in low SiO2phases (ex. garnet,

epidote, rutile, etc.) from a banded eclogite. Garnetite is a good example of metamorphic differentiation; it can-not be used to discuss the nature of its protolith. It is common to find that the chemical composition of an eclogite cannot be matched by any reasonable magmatic protoliths. In any case, eclogite protoliths cannot be easily classified with the commonly used classification schemes for igneous rocks because most of them, such as

AFM or TAS (Middlemost 1994), involve the use of alkali elements which are known to be mobile in meta-morphic processes and are generally depleted during high-grade metamorphism. The TAS classification dia-gram shown in Fig.3a serves only for reference. Nev-ertheless, allowing a maximum depletion of 0.5% total alkalis as a result of prograde metamorphism, the majority of eclogites from the Hong’an Block are still gabbroic or basaltic in composition.

Rock/Chondrites Rock/Chondrites Rock/Chondrites 1 10 100 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu HW-01 P-073 QJH-01 XHD-07 Huwan Unit (Eclogite Zone I) 1 10 100 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu QLP-08 XZD-01 LW-05 HJH-01 GQ-01 P-390 Hong'an Unit

(Eclogite Zones VI & V)

1 10 100 1000 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu P-101 P-114 P-126 QLP-14 Xinxian Unit (Eclogite Zone II)

1 10 100 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu SJH-05 QJH-03 P-032 Balifan Unit (Metagabbros) 1 10 100 1000 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu P-147 P-152 TP-03 GJH-10 P-220 P-256 P-260 Xinxian Unit

(Eclogite Zone III)

1 10 100 1000

La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Eclogites from all the 4 units and 5 zones (a) (b) (c) (f) (e) (d) Fig. 4 Chondrite-normalized REE distribution patterns of Hong’an eclogites.

Metagabbros from the Balifan Unit are also shown for comparison

In studies of Precambrian komatiites and other metavolcanic rocks, hydrous and carbonic metasoma-tism and low-grade metamorphism often pose severe problems for geochemical characterization of these rocks. In order to overcome such problems, Jensen (1976) proposed a cation plot based on the proportions of the cations (Fe2++Fe3++Ti), Al and Mg, as shown in Figure 3b. The main advantage of this diagram over other classification schemes is that only refractory (Fe, Ti, Al) and most differentiation-diagnostic (Mg) ele-ments are used. Fig. 3b shows that all the Hong’an eclogites fall in the basalt fields of the calc-alkaline and tholeiitic series. Note that the eclogites from Eclogite Zone I plot in the andesite or basaltic andesite fields in

Fig.3a, but not in Fig.3b. Close examination of five eclogite samples from Xiongdian locality (two from Table2, and three from Li et al.2001) reveal that despite their high silica (55–58%) and low TiO2 (0.32–0.38%)

contents, they have MgO = 5.1–5.9%, Cr = 430– 700 ppm, Ni = 88–116 ppm, and LREE-enriched but less than ten times chondritic REE abundances. These features seem to argue that the protolith of the eclogites was basaltic, but subjected to post-magmatic or syn-metamorphic silica-enrichment process, of which the origin is not yet understood. The case is quite similar to the Weihai eclogites. The high silica contents of Weihai eclogites (54–60%) are demonstrably of metasomatic origin (Jahn et al.1996). Their protoliths were originally e Rock/Primiti v e Mantle Rock/Primiti v e Mantle Rock/Primiti v e Mantle 1 10 100

RbBaTh U K NbLa CeSr P NdZrSm EuTi TbDyHo Y Er YbLu HW-01 P-073 QJH-01 XHD-07 Huwan Unit (Eclogite Zone I) 1 10 100

RbBaTh U K NbLa CeSr P NdZrSm EuTi TbDyHo Y Er YbLu QLP-08 XZD-01 LW-05 HJH-01 GQ-01 P-390 Hong'an Unit

(Eclogite Zones IV & V)

1 10 100

RbBaTh U K NbLa CeSr P NdZrSm EuTi TbDyHo Y Er YbLu P-101 P-114 P-126 QLP-14 Xinxian Unit

(Eclogite Zone II)

0.1 1 10 100

RbBaTh U K NbLa CeSr P NdZrSm EuTi TbDyHo Y ErYbLu SJH-05 QJH-03 P-032 Balifan Unit (Metagabbros) 0.1 1 10 100

RbBaTh U K NbLa CeSr P NdZrSm EuTi TbDyHo Y Er YbLu P-147 P-152 TP-03 GJH-10 P-220 P-256 P-260 Xinxian Unit (Eclogite Zone III)

0.1 1 10 100

RbBaTh U K NbLa CeSr P NdZrSm EuTi TbDyHo Y Er YbLu

Eclogites from all the 4 units and 5 zones

(a) (d) (b) (c) (f) (e) Fig. 5 Primitive-mantle-normalized spidergrams of Hong’an eclogites. Nb is generally depleted with respect to La, attesting to their continental affinity

of basaltic composition as inferred from relatively high MgO (6.3–7.8%), Ni (73–200 ppm) and Cr (280– 410 ppm) contents. Andesitic rocks of island arc or continental margin are renowned for their Ni (and Cr) depletion, but this is not the case for Weihai eclogites. Alternatively, the ‘‘andesitic’’ protolith could be a bon-initic magma derived by hydrous melting of metaso-matized harzburgite in a back-arc basin (Tatsumi et al. 2002).

Figure4a–e show REE distribution patterns for eclogites from the four tectonic units and metagabbros from the Balifan Unit. The global feature of all the

eclogites is further shown in Fig.4f. Despite some subtle differences, the REE distribution is characterized by mild fractionation with LREE enrichment. Their abun-dances vary from 100 to 10 times chondrites in La and all patterns are quasi-parallel. Figure5a–5f are spider-grams for the same eclogite samples. As expected, K and Rb commonly show severe depletion, probably during prograde metamorphism in the eclogite facies. Sr shows both depletion and enrichment with respect to Ce, whereas Nb (and Ta, not shown) exhibits systematic depletion relative to La. Most of the eclogites display concomitant Nb and Zr depletion. All these features are 0.001 0.01 0.1 1 10 100 1000 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu XHD07 Garnet e 0.001 0.01 0.1 1 10 100 1000 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu 0.001 0.01 0.1 1 10 100 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu XHD07 Omphacite 0.001 0.01 0.1 1 10 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu 0.1 1 10 100 1000 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu XHD07 Epidote 0.01 0.1 1 10 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Chondrite-normalized ab undance Chondrite-normalized ab undance Chondrite-normalized ab undance Chondrite-normalized ab undance Chondrite-normalized ab undance Chondrite-normalized ab undance QJH01 Garnet QJH01 Omphacite QJH01 Amphibole (a) (b) (c) (f) (e) (d) Fig. 6 REE distribution

patterns of eclogitic

minerals—garnet, omphacite, epidote and amphibole. The analyses were obtained by laser-ablation ICP-MS

typical of Type I gneiss-hosted eclogites documented in the Dabie orogen (Jahn 1998; Jahn et al. 2003b). The protoliths of eclogites represent magmas emplaced in a continental setting; none of them provide evidence for mid-ocean ridge basalt, nor any implication for a tec-tonic suture.

Chemical compositions of constituent minerals

The objective of the mineral composition study is to help understand the problem of Sm–Nd isotopic equilibrium using information on partition coefficients between

individual phases used for isochron age determination. The analytical data of REE and other trace elements (CMP Data Depository) are graphically shown in Fig.6a–6q. Note that concentrations less than 0.1 times chondritic abundances are not precise. They can be ig-nored in the discussion. The principal observations are given below.

Garnet Most garnet analyses display the characteristic patterns with increasing abundance from La to Lu, or showing maxima at middle REE (e.g., Jahn et al. 2003a). However, this rule does not hold for several samples. Garnets from the Hong’an eclogites generally 0.001 0.01 0.1 1 10 100 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu 0.001 0.01 0.1 1 10 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu 1 10 100 1000 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Chondrite-normalized ab undance Chondrite-normalized ab undance Chondrite-normalized ab undance TP03 Garnet TP03 Omphacite TP03 Epidote Chondrite-normalized ab undance Chondrite-normalized ab undance 0.001 0.01 0.1 1 10 100 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu 0.001 0.01 0.1 1 10 100 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu P260 Garnet P260 Omphacite (g) (h) (i) (k) (j) Fig. 6 (Contd.)

contain large amounts of inclusions. Isotope analyses, to be discussed later, show isotope disequilibrium, which is accompanied in part by nonequilibrated REE partition coefficients between garnet and omphacite. As shown in Appendix, garnet from sample XHD07 has numerous inclusions and exhibits weak compositional zoning. The REE patterns (Fig.6a) follow the general rule of distribution, but their concentrations vary by one to two orders of magnitude. Garnets from sample QJH01 contain abundant inclusions dominated by amphibole and epidote. Five track analyses show

sim-ilar characteristic REE patterns, but two are distinctly different, having much higher LREE abundances with positive Eu anomalies (Fig.6d). These patterns can be explained as due to a ‘‘contamination’’ of epidote inclusions. Although epidote was not analyzed in this sample, it likely has similar REE concentrations as those from samples XHD07 and TP03. Using the epi-dote data given here (Figs.6c, i), the proportion of ‘‘contamination’’ can be estimated to be about 5%. Garnets from sample TP03 show very low LREE concentrations but rather typical MREE and HREE 0.01 0.1 1 10 100 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu 0.001 0.01 0.1 1 10 100 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu 0.001 0.01 0.1 1 10 100 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Chondrite-normalized ab undance Chondrite-normalized ab undance Chondrite-normalized ab undance QLP08 Garnet QLP08 Omphacite QLP08 Amphibole _{Chondrite-normalized ab} undance Chondrite-normalized ab undance Chondrite-normalized ab undance 0.01 0.1 1 10 100 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu 0.001 0.01 0.1 1 10 100 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu 0.001 0.01 0.1 1 10 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu GQ01 Garnet GQ01 Omphacite GQ01 Amphibole (l) (m) (n) (o) (p) (q) Fig. 6 (Contd.)

for garnet (Fig.6g). Garnets of sample P260 are fine-grained and contain inclusions of omphacite, epidote and rutile. Their HREE patterns are very uniform but LREE differ by about two orders of magnitude (Fig.6j). The tracks of higher LREE concentrations are likely due to the presence of omphacite and epidote inclusions. Similarly, garnets from sample QLP08 show two distinct REE patterns (Fig.6l). Two tracks display typical garnet patterns, whereas four others have much higher LREE. The contribution of omphacite or epi-dote inclusions is clear. Garnets from sample GQ01 is porphyroblast, often contain tiny inclusions including amphibole, epidote, quartz, titanite and rutile. The REE patterns are complex with large variation in LREE and show large Eu anomalies (Fig.6o) as a re-sult of different inclusions.

Omphacite The literature data show that typical omphacites have hump-shaped REE patterns with maxima at middle REE (Jerde et al. 1993; Snyder et al. 1993; Bocchio et al.2000; Sassi et al. 2000; Jahn et al. 2003a). Omphacite of sample XHD07 is rather homogeneous, and its REE distributions of this sample generally show the typical omphacite patterns except one with high REE abundances (Fig.6b). This exceptional pattern is due to ‘‘contamination’’ by epidote inclusion which was not visible during the la-ser track-scan. The REE data of epidote are illustrated in Fig.6c. Omphacite of sample QJH01 seldom con-tains mineral inclusions and is quite homogeneous in composition. The REE abundances are rather uniform showing typical hump-shaped patterns (Fig.6e). Amphibole from the same sample also possesses nearly identical REE patterns (Fig.6f), arguing that amphi-bole was transformed from omphacite and inherited the REE budget entirely from omphacite. This was also observed for the case of the Maowu eclogite-ultramafic body from the Dabie terrane (Jahn et al. 2003a). Omphacite from sample TP03 is coarser than garnet and sometimes contains inclusions of small garnet and rutile. Due to their low abundances, the REE data are of poor quality but still show hump-shaped patterns (Fig. 6h). However, two patterns re-quire further explanation—the one with high HREE abundances can be explained as due to the presence of garnet inclusion; whereas the other with higher LREE probably indicates inclusions of epidote, whose REE patterns are shown in Fig.6i. The similar features are observed in omphacite of sample P260. The omphacite has very low REE abundances (<1.0 times chondrite); their poor analyses do not clearly show typical om-phacite patterns (Fig.6k). However, the one with the highest HREE is clearly due to garnet because the track contains very high Al and Fe, and low Sr concentration (13 ppm), whereas the two with high LREE are a result of epidote inclusion. Omphacite from sample QLP08 shows a large variation in both LREE and HREE distribution (Fig. 6m). The lowest abundances with hump-shaped patterns are of

inclusion-free omphacite, the others can be explained as due to variable contributions from epidote inclu-sions. Omphacite from sample GQ01 has much in common with that of QLP08 (Fig.6p). Amphibole also shows similar hump-shaped REE patterns with one exception.

Epidote Epidotes from two samples (XHD07 and TP03) show similar REE patterns (Figs.6c, i), with high and gently sloped LREE (100–400 times, except one) and sharply declined HREE towards Lu; all of them show positive Eu anomalies. The average (La/Lu)Nratio

is about 100.

Summary on mineral REE abundances Sm and Nd concentrations in garnet, clinopyroxene, orthopyroxene and amphibole from a variety of rocks have been com-piled from the literature data (Jahn et al. 2003b). More than 70% of garnets from eclogites have Nd concen-trations less than 1 ppm, and their Sm/Nd ratios very from 0.5 to >10. A small number of garnets (7/140) have Sm/Nd ratios less than the chondritic value of ca. 0.325; most of these have exceptionally high Nd (5– 23 ppm), suggesting that the garnets contain significant amounts of LREE-rich inclusions, such as monazite, zoisite/epidote or apatite. ‘‘Clean’’ garnets, like those from the Bixiling complex of the Dabie terrane, tend to have low Nd concentrations (£ 0.5 ppm) and high Sm/ Nd ratios (Jahn et al. 2003b). With respect to ompha-cites, Nd concentrations vary from 0.1 ppm to 40 ppm, with ca. 70% less than 6 ppm. Sm/Nd ratios generally fall in the range of 0.2–0.5, with only few (7/153) exceeding 1.0.

Figure7 shows plots of Sm/Nd vs [Nd] for garnet and omphacite from this study, with the ranges of the same minerals from the Maowu Complex (Jahn et al. 2003a). The Sm–Nd isotope systems in the Maowu rocks produced correct Triassic ages, hence are con-sidered to have reached isotope equilibrium. The gar-net and omphacite data of Maowu eclogites and pyroxenites fall in two distinct and tight areas. Garnet has a range of Nd concentrations from about 0.1 to 1 ppm, and Sm/Nd ratios from 1 to 10. Such ranges appear to be ‘‘normal’’. By contrast, garnets from the six rocks from the Hong’an Block for isotope analyses show Nd concentrations and Sm/Nd ratios quite var-iable, and many of the ratios less than 0.3. This is very unusual when they are compared with the compiled data described above. Omphacite of the Maowu eclogites–pyroxenites has a small range of Sm/Nd ra-tios about 0.4 (Fig.7) despite of a large range of Nd concentrations from 0.7 ppm to 30 ppm. The Hong’an omphacites have Sm/Nd ratios much more variable, from 0.15 (sample P260) to 2 (samples XHD07 and GQ01), and nearly a quarter (11/43) exceeding 1. The unusual concentrations and Sm/Nd ratios in the Hong’an minerals reflect the effect of inclusions and the coexisting minerals are possibly out of chemical equilibrium.

Rb–Sr and Sm–Nd isotope analyses

The results of Rb–Sr and Sm–Nd isotopic analyses on minerals separated from six basaltic eclogites, one glaucophane schist and a marble are presented in Ta-ble 3 and further illustrated in Figs.8,9,10,11,12,13,14 and15. The most striking observation is that none of the Sm–Nd isotope systems provide useful age information, and half of the Rb–Sr systems are equally disturbed. The apparently undisturbed cases are represented by the phengite-based Rb–Sr isochrons of 212±7 (2sigma) Ma for sample TP03 (Fig.10), 225±34 Ma for QLP08

(Fig.12), 216±10 Ma for glaucophane schist (sample BJ01-101; Fig.14), and 210±4 Ma for Marble Hill (sample BJ01-115; Fig.15). The ages of 225–210 Ma are comparable with40Ar/39Ar ages obtained on a variety of rock types from the Hong’an Block (Eide et al. 1994; Webb et al.1999,2001; Hacker et al.2000). They can be interpreted as cooling ages. The meaning of the younger age of 172±23 Ma for sample P260 (Fig.11) is not clear, but is possibly related to retrograde amphibolite facies metamorphism. Note also that the pattern of Rb– Sr isotope disequilibrium appears to be random and is independent of the petrological temperatures.

In the Sm–Nd isotope systems, negative isochron relationship is observed for two eclogites: one from the coesite-bearing unit (P260; Fig.11), and the other from cold eclogite unit (QLP08; Fig.12). Grt–Omp tie-lines yielded highly variable ‘‘isochron ages’’ of 260 Ma (XHD07 and BJ01-107), 420 Ma (QJH01), 1,057 Ma (TP03), and 182 Ma (GQ01), whereas Grt–WR tie-lines gave 346, 250, 820, and 218 Ma, respectively. Quite

Fig. 7 Sm/Nd versus Nd plots for garnet and omphacite from the Hong’an eclogites. The data of the chemically and isotopically equilibrated Maowu complex are shown for reference. The Hong’an data often show wide scatter and incorrect Sm/Nd ratios. The data obtained by the isotope dilution (ID) method using whole-grain dissolution are very different from those obtained by LA-ICP-MS. This clearly indicates the contribution of REE-rich inclusions such as epidote, apatite, or other phases; many of these phases were not analyzed by LA-ICP-MS

Table 3 Rb–Sr and Sm–Nd isotopic compositions of eclogites and blueschist from the Hong’an Block Sample No. Analysis No. [Rb] (ppm) [Sr] (ppm ) 87 Rb/ 86 Sr 87 Sr/ 86 Sr ±2 r m [Sm] (ppm) [Nd] (ppm) 147 Sm/ 144 Nd 143 Nd/ 144 Nd ±2 r m
Nd(0) f (Sm/Nd)
Nd(T) (220 Ma) TDM -1 (Ma) TDM -2 (Ma)
Nd(T) (750 Ma) XHDO7 WR 14563 21.48 180.2 0.34 0.708075 7 1.182 4.166 0.1715 0.512696 4 1.1  0.13 1.8 1637 800 3.5 XHDO7 Grt 14564 2.54 5.11 1.44 0.710584 8 0.229 0.425 0.3258 0.513036 4 7.8 0.66 4.1 XHDO7 Omp 14565 7.24 30.38 0.69 0.708610 7 0.384 1.321 0.1757 0.512781 4 2.8  0.11 3.4 XHDO7 Phen 14566 164.4 309.5 1.54 0.712319 7 BJ01-10 7 W R 14672 RT 21.36 111.9 0.55 0.708793 8 1.05 3.64 0.1747 0.512690 3 1.0  0.11 1.6 1793 814 3.1 BJ01-10 7 Grt 14672 Gt 2.23 3.33 1.93 0.711923 6 0.31 0.48 0.3924 0.513192 22 10.8 0.99 5.3 BJ01-10 7 Omp 14672 Omp 8.68 26.68 0.94 0.709292 8 0.70 1.85 0.2284 0.512922 3 5.5 0.16 4.7 BJ01-10 7 Amp 14672 Amp 4.83 22.97 0.61 0.709012 7 0.56 1.76 0.1935 0.512818 3 3.5  0.02 3.6 BJ01-10 7 E p 14672 Ep 3.60 2918 0.00 0.706868 7 30.95 98.25 0.1905 0.512701 3 1.2  0.03 1.4 BJ01-10 7 Phen 14672 Phen 156.5 327.5 1.38 0.711486 7 0.02 0.06 0.1519 0.512405 33  4.5  0.23  3.3 QJH01-W R 14567 13.21 348.8 0.11 0.705089 7 2.911 11.934 0.1475 0.512385 3  4.9  0.25  3.6 1756 1264  0.2 QJH01-Grt 14568 1.48 55.82 0.08 0.707332 8 0.997 3.475 0.1735 0.512428 3  4.1  0.12  3.4 QJH01-Omp 14569 0.53 134.9 0.01 0.707558 7 1.810 7.192 0.1521 0.512369 3  5.2  0.23  4.0 QJH01-Amp 14570 3.08 144.4 0.06 0.709258 6 2.034 8.074 0.1523 0.512380 3  5.0  0.23  3.8 TP03 WR 14611 11.02 297.0 0.11 0.705051 7 2.83 11.83 0.1446 0.512389 3  4.9  0.26  3.4 1675 1254 0.1 TP03 Grt 14612 0.84 34.5 0.07 0.704817 7 2.03 6.84 0.1798 0.512578 4  1.2  0.09  0.7 TP03 Omp 14613 0.98 46.1 0.06 0.704950 7 0.27 1.08 0.1514 0.512381 4  5.0  0.23  3.7 TP03 Phen 14614 197.7 77.4 7.39 0.726989 7 TP03 Zois 14615 0.65 3819 0.001 0.704733 6 38.83 163.28 0.1438 0.512382 4  5.0  0.27  3.5 P 260 WR 14571 15.2 241.3 0.18 0.705510 8 4.515 17.328 0.1575 0.512573 3  1.3  0.20  0.2 1563 977 2.5 P 260 Grt 14572 0.6 4.31 0.41 0.706658 8 0.162 0.593 0.1652 0.512513 3  2.4  0.16  1.6 P 260 Omp 14573 2.36 103.7 0.07 0.705138 5 1.994 7.592 0.1588 0.512579 5  1.2  0.19  0.1 P 260 Phen 14574 241 76.0 9.16 0.727424 7 QLP08 WR 14607 10.0 142.5 0.20 0.705001 6 6.50 24.32 0.1616 0.512555 3  1.6  0.18  0.6 1738 1012 1.7 QLP08 Grt 14608 1.48 2.08 2.06 0.706312 8 0.09 0.197 0.2798 0.512488 6  2.9 0.42  5.3 QLP08 Omp 14609 1.21 120.3 0.03 0.704314 6 4.82 18.04 0.1616 0.512579 3  1.2  0.18  0.2 QLP08 Phen 14610 305 160.4 5.50 0.721882 7 1.98 7.29 0.1646 0.512564 6  1.4  0.16  0.5 GQ01 W R 14604 0.30 117.7 0.007 0.705917 7 1.37 4.11 0.2023 0.512918 3 5.5 0.03 5.3 3082 487 4.9 GQ01 Grt 14605 0.09 30.5 0.008 0.705803 6 0.24 0.63 0.2338 0.512963 5 6.3 0.19 5.3 GQ01 Omp 14606 0.15 244.2 0.002 0.705760 6 1.82 5.37 0.2053 0.512929 3 5.7 0.04 5.4 BJ01-10 1 W R 14671 RT 69.3 216.3 0.93 0.711825 7 11.68 52.31 0.1350 0.512241 3  7.7  0.31  6.0 1756 1478  1.8 BJ01-10 1 Grt 14671 Gt 4.2 168.6 0.07 0.709485 7 9.86 45.74 0.1304 0.512236 3  7.8  0.34  6.0 BJ01-10 1 Gln 14671 Glau 5.7 133.9 0.12 0.709683 7 5.94 26.97 0.1332 0.512221 3  8.1  0.32  6.4 BJ01-10 1 Phen 14671 Phen 235 19.1 35.9 0.819695 8 0.75 3.50 0.1296 0.512276 12  7.1  0.34  5.2 BJ01-10 1 P l 14671 Plag 4.4 96.4 0.13 0.706872 7 1.63 7.45 0.1326 0.512225 2  8.1  0.33  6.3 BJ01-10 1 Kfs 14671 Kfs 3.7 81.5 0.13 0.707138 7 1.37 6.28 0.1323 0.512230 3  8.0  0.33  6.2 BJ01-11 5 W R 14673 RT 0.78 161.1 0.01 0.706351 7 BJ01-11 5 Mus 14673 Musc 269 1.63 552.5 2.358570 20

visibly, Carboniferous to Silurian ages, as shown to exist by some workers, can be reproduced from such un-equilibrated isotopic systems. In addition, WR data points are always found to lie outside the Grt–Omp tie-lines. This indicates that isotopic compositions of coex-isting minerals were not homogenized during metamor-phism and that WR must be mass-balanced by nonanalyzed accessory phases. The 147Sm/144Nd ratios of the analyzed garnets range from 0.17 to 0.39. This is much lower than the ‘‘ normal ’’ garnet values of 0.5 to 2.0. Besides, two garnets contain high Nd concentrations (3.5 and 6.8 ppm), which likely resulted from the pres-ence of LREE-rich microinclusions (e.g., epidote and monazite). In fact, the laser track analysis described earlier shows that some pure garnet domains could have very high Sm/Nd ratios (Fig. 7). However, such pure domains could not be separately analyzed in the isotope dilution analysis.

The above observations raise a serious question about the validity of the published Sm–Nd mineral isochron ages on the Hong’an Block. Such unequili-brated Sm–Nd systems are rarely encountered for

eclogites from the higher-temperature Dabie and Sulu terranes (e.g. Li et al.,1993,2000; Chavagnac and Jahn 1996; Chavagnac et al. 2001; Jahn et al.2003a) except two localities (Yakou and Yangkou) of the Sulu terrane (Zheng et al. 2002) and a retrograded clinopyroxene garnetite from northeastern Dabie (Xie et al. 2004). In the Hong’an Block, the Xiongdian eclogite (locality identical to our samples XHD07 and BJ01-107; 3145.14¢N, 11428.50¢E) has been subjected to several zircon U–Pb and garnet Sm–Nd geochronological investigations. However, all the published results failed to produce consistent and interpretable ages. Clearly, the isotopic systems (Sm–Nd and U–Pb) are largely out of equilibrium in these ‘‘cold eclogites’’, and some of the apparent concordant zircon dates of 300 to 450 Ma could have been produced by partial recrystallization (Pidgeon 1992; Pidgeon et al. 1998; Hoskin and Black 2000; Li 2003).

Failure of producing meaningful ages

Meaningful Sm–Nd and Rb–Sr isochron ages have been obtained for the Bixiling and Maowu Complexes from 0.706 0.707 0.708 0.709 0.710 0.711 0.712 0.713 0.0 0.5 1.0 1.5 2.0 XHD07 BJ01-107 Phen Grt Omp WR 87 Rb/86Sr XHD07 & BJ01-107 (Xiongdian) (680-700°C, 15-17 kb) 220 Ma (reference) WR Amp Grt Omp Phen Ep 184 Ma 246 Ma 0.5124 0.5125 0.5126 0.5127 0.5128 0.5129 0.5130 0.5131 0.5132 0.5133 0.15 0.20 0.25 0.30 0.35 0.40 XHD07 BJ01-107 147 Sm/144Nd XHD07 & BJ01-107 (Xiongdian) (680-700°C, 15-17 kb) Grt Omp WR WR Grt Omp Ep Amp T = 346 ± 20 Ma I = 0.51230 ± 0.00004 MSWD = 0.3 (i.e. = 0.2%, 0.005%) (4 points: 2 WR, 2 Grt) T = 260 Ma (Grt-Omp tie-line) T = 252 Ma Sr/ Sr 87 86 143 144 Nd/ Nd (a) (b) eNd(T) = +4.5 eNd(T) = +3.5 e(T) = +2.2

Fig. 8 Rb–Sr and Sm–Nd isochron diagrams for sample XHD07 and BJ01-107 0.704 0.705 0.706 0.707 0.708 0.709 0.710 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 87 Rb/86Sr Amp Grt Omp WR QJH01 (Qianjinhepeng) (610°C, ³ 12 kb) 0.51235 0.51237 0.51239 0.51241 0.51243 0.51245 0.14 0.15 0.16 0.17 0.18 147 Sm/144Nd 250 Ma Grt Omp WR QJH01 (Qianjinhepeng) (610°C, ³ 12 kb) Amp 420 Ma Sr/ Sr 87 86 143 144 Nd/ Nd (a) (b)

the Dabie terrane (Chavagnac and Jahn1996; Jahn et al. 2003a). The metamorphic ages of ca. 220 Ma for the Bixiling Complex are in perfect agreement with the SHRIMP U–Pb zircon ages obtained by Cheng et al. (2000) and Li (2003). The Maowu eclogites and py-roxenites appear to have recorded slightly older ages at ca. 230 Ma, and the results are also consistent with those obtained by the conventional (Rowley et al. 1997) as well as SIMS zircon analyses (CAMECA ims-1270, Ayers et al. 2002). Moreover, meaningful Sm–Nd and Rb–Sr isochron ages of ca. 620 Ma have been obtained for the Pan-African coesite-bearing eclogites from Mali (Jahn et al.2001). The Mali eclogites are the oldest UHP eclogites identified so far. Note that all the rocks have been recrystallized in UHP metamorphic conditions with temperatures well over 700C.

Excess Ar in phengitic mica of HP-UHP rocks have been frequently documented (e.g., Li et al. 1994, 2000; Hacker et al.2000; Jahn et al.2001). Similarly, aberrant Sm–Nd mineral isochron ages have also been obtained, particularly for ‘‘low temperature’’ (£ 600C) HP– UHP metamorphic rocks. The most notorious examples

are from the Alps and the Himalayas (e.g., Luais et al. 2001; Tho¨ni 2002). The Hong’an Block adds a new example to the list. The main cause for this chronometric problem is the lack of isotopic equilibrium between garnet and its coexisting minerals. Garnet is the major Al-carrying phase in eclogite and is probably trans-formed mainly from plagioclase, whereas omphacite is derived from magmatic pyroxenes. However, the detail of phase reactions leading to the formation of these two minerals in such metamorphic conditions could be more complicated.

Granting the simplified case, during the ‘‘low tem-perature’’ metamorphic transformation from plagio-clase to garnet, the reconstitution of lattice-forming major elements may not be closely followed by REEs due to their smaller diffusion coefficients. Isotopic equilibrium would not be expected to occur when chemical equilibrium is not reached. In numerous cases, garnet crystals show major element zonation of pro-grade metamorphism (e.g., Liu et al. 2004a). This is clear evidence for nonequilibrated growth zones. In this case, trace element and isotopic equilibrium is probably out of question. 0.700 0.705 0.710 0.715 0.720 0.725 0.730 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 87 Rb/86Sr Phen GrtOmp WR TP03 (Tianpu) (640°C, 29 kb) T = 212 ± 7 Ma I = 0.7047 ± 0.0001 MSWD = 16 (4 points) Ep 0.51235 0.51240 0.51245 0.51250 0.51255 0.51260 0.51265 0.14 0.15 0.16 0.17 0.18 0.19 147 Sm/144Nd Grt Omp WR Ep TP03 (Tianpu) (640°C, 29 kb) 820 Ma 1057 Ma Sr/ Sr 87 86 143 144 Nd/ Nd (a) (b)

Fig. 10 Rb–Sr and Sm–Nd isochron diagrams for sample TP03

0.700 0.705 0.710 0.715 0.720 0.725 0.730 0.0 2.0 4.0 6.0 8.0 10.0 87 Rb/86Sr Phen Grt Omp WR P260 (Xiongjiazhui) (670°C, 26 kb) T = 171 ± 17 Ma I = 0.7052 ± 0.0010 MSWD = 326 (4 points) T = 172 ± 23 Ma I = 0.7050 ± 0.0006 MSWD = 11.8 (3 points, excl Grt) 0.51248 0.51250 0.51252 0.51254 0.51256 0.51258 0.51260 0.155 0.157 0.159 0.161 0.163 0.165 0.167 147 Sm/144Nd Grt Omp WR P260 (Xiongjiazhui) (670°C, 26 kb) Sr/ Sr 87 86 143 144 Nd/ Nd (a) (b)

The phenomenon of Sm–Nd isotopic equilibrium/ disequilibrium is most dependent on the prime factor of temperature, but little on pressure. Other factors, such as the intensity of deformation or crystal-fracturing in garnet (Whitney 1996), remains to be evaluated. The temperature effect leads to the concept of blocking temperature (TB; Dodson1973), which does not have a

unique value but is also influenced by several other factors including mineral grain-size, rate of cooling, duration of metamorphic reaction at temperature higher than TB, presence of fluid phase and its composition,

and penetrative deformation (e.g., Tho¨ni 2002). The literature data indicate that TBvalues for the garnet Sm–

Nd isotope chronometer range from as high as 850C for fast-cooling terranes to as low as 600–650C for slow-cooling terranes (Humphries and Cliff 1982; Jagoutz 1988; Mezger et al.1992; Burton et al.1995; Zhou and Hensen1995; Gu¨nther and Jagoutz1997; Tho¨ni2002). Garnet out of isotopic equilibrium is often accompanied by nonequilibrated trace element patterns. In some cases, Sm/Nd ratios of garnets are so unusual that they are smaller than that of whole-rock samples; whereas in others, all coexisting minerals show little difference in

Sm/Nd ratios so they form a cluster in an isochron diagram, as shown in the glaucophane schist sample (Fig.14). In still other cases, garnet and omphacite may have established equilibrium Sm/Nd partition coeffi-cients but not their Nd isotopic compositions, thus resulting in a negative slope ‘‘futurechron’’ relationship (Figs.11, 12). This has been found in eclogites of Tso Morari of the Himalayas and the Sesia zone of western Alps (Ducheˆne et al.1997; de Sigoyer et al.2000; Luais et al.2001).

Porphyroblastic garnets often contain mineral inclu-sions. In addition to coesite, many of them are REE-rich phases, such as monazite, zoisite, epidote, allanite, tita-nite, apatite, zircon, etc. Except zircon, all these phases are highly enriched in LREE with very low Sm/Nd ra-tios. Thus, a tiny amount of such inclusions could sig-nificantly lower Sm/Nd ratio in garnet, hence, affecting isochron construction based on Grt–Cpx–WR and other minerals. Inclusions must have undergone the same metamorphic PT path as garnet, clinopyroxene and other major phases. If the metamorphic temperature exceeds the TBof the inclusions (moderate TBminerals,

such as epidote, apatite and titanite), the Nd isotopic 0.700 0.705 0.710 0.715 0.720 0.725 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 87 Rb/86Sr Phen Grt Omp WR QLP08 (Dongyuemiao) (530°C, 20 kb) T = 225 ± 34 Ma I = 0.7043 ± 0.0008 MSWD = 24 (3 points, excl Grt) 0.51245 0.51250 0.51255 0.51260 0.51265 0.12 0.14 0.16 0.18 0.20 0.22 0.24 0.26 0.28 0.30 0.32 147 Sm/144Nd Grt Omp WR QLP08 (Dongyuemiao) (530°C, 20 kb) Phen Sr/ Sr 87 86 143 144 Nd/ Nd (a) (b)

Fig. 12 Rb–Sr and Sm–Nd isochron diagrams for sample QLP08

0.70565 0.70570 0.70575 0.70580 0.70585 0.70590 0.70595 0.000 0.002 0.004 0.006 0.008 0.010 87 Rb/86Sr Grt Omp WR GQ01 (Gaoqiao) (510°C, 20 kb) 0.51289 0.51291 0.51293 0.51295 0.51297 0.51299 0.190 0.200 0.210 0.220 0.230 0.240 147 Sm/144Nd Grt Omp WR GQ01 (Gaoqiao) (510°C, 20 kb) 218 Ma 182 Ma Sr/ Sr 87 86 143 144 Nd/ Nd (a) (b)

compositions of inclusions could be expected to attain isotopic equilibrium with host garnet. In this case, a correct isochron age may be obtained, but the range of data spread would be reduced, and hence the statistical error in age increased. This has been most frequently observed in the ‘‘high-temperature’’ eclogites of the Dabie and Sulu terranes. On the other hand, mineral inclusions of very high TB (e.g., zircon and monazite)

may not be reset isotopically to keep pace with growing garnet. In this case, it may produce a disequilibrated isochron relationship.

In conclusion, the failure of producing a correct Sm– Nd isochron age is due to isotope disequilibrium be-tween garnet and its coexisting minerals, including inclusions. The disequilibrium results from two distinct processes: (1) garnet preserves the isotopic composition of its precursor mineral (plagioclase) because the rate of Nd isotope exchange is more sluggish than that of the chemical and mineral phase reconstitution; (2) garnet contains high-temperature refractory LREE-rich inclu-sions, which have not been isotopically re-equilibrated with host garnet, though the garnet may have attained equilibrium with coexisting clinopyroxene and other principal phases. The above disequilibrium does not include the open system behavior occurring during ret-rograde metamorphism with strong influence of hydro-thermal activity, as demonstrated by Xie et al. (2004).

Oxygen isotope information

Oxygen isotopic compositions of eclogite minerals have generally attained isotope equilibrium at their eclogite facies temperatures (e.g., Zheng et el. 2002). The equi-librium is indicated by the fractionation between indi-vidual minerals, as expressed by D values (d18OA 

d18OB). The sequence of 18

O enrichment in eclogite minerals is consistent with the empirically, experimen-tally and theoretically determined values for equilibrium fractionation (Zheng et al.1998and references therein). In fact, calculated oxygen isotope temperatures have been shown to be quite comparable with petrological temperatures (Yui et al.1995; Zheng et al.1998,1999). The analytical data of oxygen isotope compositions are given in Table4. Figure16illustrates oxygen isotope fractionation values (D18

O) between garnet and om-phacite versus d18O of garnet in eclogites from numerous localities including the Hong’an Block. The range of isotope equilibrium fractionation at eclogitic tempera-tures (D18

O = 0–2&) is shown by grey area for refer-ence. Garnet covers a wide range of d18O values from +10 to 10&, but the majority of eclogites (70%) fall in the area of isotope equilibrium. Other rocks with positive fractionation (D‡2.2&) are out of equilibrium, whereas many eclogites (ca. 30%) show negative or re-verse fractionation, resulting in quartz–omphacite iso-tope temperatures lower than 400C (Zheng et al.1999). This phenomenon of disequilibrium has been ascribed to retrograde hydration reactions, in which the responsible 0.700 0.720 0.740 0.760 0.780 0.800 0.820 0.840 0 5 10 15 20 25 30 35 40 WR 87 Rb/86Sr T = 216 ± 10 Ma I = 0.7092 ± 0.0005 MSWD = 27 (4-point; Pl & Kfs excluded)

Grt, Gln Pl, Kfs Phen BJ01-101 0.51220 0.51222 0.51224 0.51226 0.51228 0.51230 0.120 0.125 0.130 0.135 0.140 0.145 0.150 0.155 0.160 147 Sm/144Nd Grt Gln WR Phen Kfs Pl BJ01-101 (Glaucophane schist) (350-450°C, 7-10 kb) 0.706 0.707 0.708 0.709 0.710 0.711 0.712 0 0.2 0.4 0.6 0.8 1.0 Gln WR BJ01-101 (350-450°C, 7-10 kb) Kfs Grt Pl Sr/ Sr 87 86 143 144 Nd/ Nd (a) (b)

Fig. 14 Rb–Sr and Sm–Nd isochron diagrams for sample BJ01-101 (glaucophane schist) 0.500 1.000 1.500 2.000 2.500 0 100 200 300 400 500 600 WR 87 Rb/86Sr BJ01-115 (Marble Hill) (410 ± 50°C, 5-10 kb) Muscovite T = 210 ± 4 Ma I = 0.70631 ± 0.00004 Sr/ Sr 87 86

Fig. 15 Rb–Sr and Sm–Nd isochron diagrams for sample BJ01-115 (marble)

fluids were probably derived from exsolution of dis-solved hydroxyl in UHP metamorphic minerals (Zheng et al. 1999), or the fluids that had equilibrated with low d18O country gneisses (Yui et al. 1997; Zheng et al. 1999). Note that all the six eclogite samples from the Hong’an Block (solid symbols in Fig.16) have D 18

O values close to zero, indicating roughly high-tempera-ture equilibrium between garnet and omphacite, despite the very large range of garnet d18O values (8 to +10&). The large range suggests that the protoliths of eclogites have probably been subjected to different his-tories of water–rock interaction, although other mech-anisms could be used to explain such a large spread (e.g., Zheng et al.1998).

Figure17 provides another look at the oxygen iso-tope fractionation between coexisting phases. The oxy-gen isotope temperatures inferred from the data points of quartz and garnet, or of phengite and garnet appear to agree well with the petrological temperatures given earlier. However, due to the smallD18

O values between garnet and omphacite, the isotopic temperatures are quite imprecise and higher than 800C, in disagreement with the petrological geothermometry.

Table 4 Oxygen isotope data of eclogites and blueschist from the Hong’an Block

Rock type Phase (/) d18O (&) No. analysis D (Qtz - /) TC (Qtz - /) Petro T (C)

XHD07 Eclogite Grt 9.4±0.1 2 3.8 660 680–700 Omp 9.3±0 2 3.9 495 Phen 11.0±0.2 2 2.2 673 Ep 9.8±0.1 2 3.4 553 Qtz 13.2±0.1 2 BJ01-107 Eclogite Qtz 13.4±0.3 2 680–700 QJH01 Eclogite Grt 3.5±0 2 610±40 Omp 3.4±0.1 2 Amp 3.4±0.1 2 Rt – Qtz – TP03 Eclogite Grt 3.3±0 2 4.3 604 640±30 Omp 2.8±0.1 2 3.8 505 Amp 2.9±0 2 3.9 589 Ep 2.8±0.1 2 3.8 508 Phen 0.9±0.3 2 1.9 745 Qtz 1.0±0.1 2 P260 Eclogite Grt 1.4±0.1 2 4.4 594 670±60 Omp 1.3±0 2 4.5 442 Phen 3.4 1 2.4 633 Qtz 5.8±0.1 2 QLP08 Eclogite Grt 7.9±0.1 2 8.2 530±30 Omp 7.3±0.2 2 7.6 Amp 8.2±0 2 8.5 Phen 5.7±0.2 2 6.0 Qtz – GQ01 Eclogite Grt 8.3±0 2 4.3 604 510±40 Omp 8.6±0.2 2 4.0 485 Amp 7.8±0.1 2 4.8 504 Qtz 12.6 1 BJ01-101 Blueschist Qtz 9.6±0.1 2 350–450 -6 -5 -4 -3 -2 -1 0 1 2 3 4 -15 -10 -5 0 5 10 Qinglongshan (Sulu) Other Sulu areas Shuanghe (Dabie) Bixiling (Dabie) Other Dabie areas XHD07 (Hong'an) QJH01 (Hong'an) P260 (Hong'an) GQ01 (Hong'an) QLP08 (Hong'an) TP03 (Hong'an) 18 O (‰) Garnet

Range of isotope equilibrium between omphacite & garnet

∆

18 O (‰) Omp-Grt

δ

Fig. 16 D 18

O (Omp-Grt) versus d18O (Grt) plot showing the equilibrium values of oxygen isotope fractionation. Note that almost all Hong’an eclogites show quasi-equilibrium between garnet and omphacite, and that garnets from different localities have very large range of d18O, suggesting their distinct premeta-morphic histories of water–rock interaction

Note: All analyses were done at the IES, Academia Sinica (TF Yui). Garnet, Omp, Amp, Zoi - by CO2laser-fluorination method. Phengite and Qtz—by conventional BrF5 method. Precision =±0.1(&); results relative to SMOW. UWG-2 garnet standard = + 5.8&; NBS-28 standard = + 9.6&. Mass spectrometer used

=Finnigan MAT 252. Oxygen isotope temperatures calculated using the fractionation factors of Zheng et al. (2003). Factors A(Qtz-Grt) = 3.31, A(Qtz-Omp) = 2.30, A(Qtz-Phen) = 1.97,