503

Geochemical Journal, Vol. 39, pp. 503 to 516, 2005

*Corresponding author (e-mail: shiguanghai@263.net.cn)

Copyright © 2005 by The Geochemical Society of Japan.

Methane (CH4)-bearing fluid inclusions in the Myanmar jadeitite

G. UANGHAI SHI,1,2* PETER TROPPER,3 WENYUAN CUI,4 JUN TAN2 and CHUANGQIU WANG4

1China University of Geosciences, Beijing 100083, China2Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing 100029, China

3Faculty of Geo- and Atmospheric Sciences, Institute of Mineralogy and Petrography, University of Innsbruck,Innrain 52, A-6020 Innsbruck, Austria

4School of Earth and Space Sciences, Peking University, Beijing, 100871, China

(Received November 26, 2004; Accepted May 12, 2005)

A combined hydrogen-carbon-isotope and microthermometric study has been carried out on CH4-bearing fluid inclu-sions in high-pressure jadeitites from the famous jadeite tract Myanmar. Two types of fluid inclusions were found injadeites, large H2O-rich and CH4-poor inclusions and small H2O-poor and CH4-rich inclusions, thus indicating a possibleentrapment of CH4-H2O fluids under unmixing conditions. Microthermometric results yield lower temperature limits forthe entrapment of these fluid inclusions of ca. 300 to 400°C. The bulk composition of the fluid inclusions is mostly H2O(87 to 94 mol.% H2O) and the isotopic composition of methane and water in the inclusions is characterized by δ13C(CH4)values ranging from –30.1 to –25.5‰, and δD(H2O) values ranging from –56.3 to –49.8‰. The stable isotope data wouldbe indicative of an abiogenic mechanism of CH4 formation; the occurrence of the jadeite veins in this paleo-subductionzone thus most likely point to the formation of these CH4-bearing fluid inclusions by abiogenic thermal maturation ofsubducted organic carbon. These data not only provide evidence for cycling of organic carbon in paleo-subduction zonesbut also show that CH4 not only occurs as shallow CH4-rich plumes in accretionary prisms of recent subduction zones butalso occurs in deeper portions of at least the upper 20 km of paleo-subduction zones.

Keywords: fluid inclusions, jadeitite, methanogenesis, methane (CH4), Myanmar

under high-P/low-T conditions (Harlow, 1994; Okay,1997). Due to their often common vein-like texture, whichimplies a possible metasomatic origin, jadeitites can bethought of as petrological recorders of the fluid historyof a subduction environment (Sorensen and Harlow, 1999,2001; Harlow and Sorensen, 2001, 2005).

Fluid inclusion studies in jadeitites therefore provideimportant constraints on the composition of the metamor-phic fluid present during formation of the jadeitites indeep subduction zone environments. In recent studies,Harlow (1986) and Johnson and Harlow (1999) reportaqueous fluid inclusions containing up to 8.7 wt% NaClin coarse-grained jadeites from the Guatemaleanjadeitites. The authors proposed that the fluid is probablyof seawater origin and was entraped during subductionin the minerals. Fluid inclusion studies on high pressurerocks such as eclogites show that fluids in subductionzones are variably saline brines with minor CO2 and N2and rarely contain some CH4 (Sorensen and Barton, 1987;Kastner et al., 1998; Scambelluri and Philippot, 2001).Studies of oceanic rocks and of eclogitized ophiolites in-dicate that prior to subduction, the slab can be consider-ably altered by surficial fluids (Scambelluri and Philippot,2001). Metamorphic processes such as devolatilizationproduce in the initial stages low-salinity, CO2- and CH4-

INTRODUCTION

Jadeite commonly occurs in high-pressure or ultra-high pressure metamorphic rocks such as jadeite-bearingquartzites and gneisses (Schertl et al., 1991), jadeite-K-feldspar bearing schists (Okay, 1997) and occasionallyin jadeitites, which are primarily composed of jadeite.Jadeitites are relatively rare and have been described froma handful of localities such as northwestern Myanmar;Kamuikotan area, Hokkaido, Japan; San Benito County,California; Pay-Yer massif, Polar Urals, Russia; northernnear-Balkash region, Kazakhstan; Montagua Fault andGuatemala (Chhibber, 1934; Iwao, 1953; Morkovkino,1960; Coleman, 1961; Dobretsov and Ponomareva, 1965;McBirney et al., 1967; Takajama, 1986; Kobayashi et al.,1987; Harlow, 1994). Jadetite petrogenesis has been in-terpreted to be either metamorphic or metasomatic(Harlow, 1994; Miyazaki et al., 1998; Radvanec et al.,1998). Jadeiteites commonly occur as lenses, pods andveins or as tectonic blocks within a serpentinizedultramafic matrix such as dunites and are thought to form

504 G. Shi et al.

bearing fluid inclusions (Kastner et al., 1998). At highergrades, CH4 seems to disappear and CO2 forms insteadas reported from fluid inclusions from eclogite-faciesrocks (Xiao et al., 2000; Scambelluri and Philippot, 2001;Yang et al., 2001; Fu et al., 2003a). But recent studiesshow that ultrahigh pressure (UHP) metamorphiceclogites and associated gneisses from the Dabie-Suluterranes, despite most of them having carbonic fluid (CO2-rich) inclusions, also contain CH4-rich fluid inclusionscoexisting with high-salinity brine inclusions (Fu et al.,2001, 2002, 2003b) and organically derived carbonate inwhole-rock and apatite (Zheng et al., 2000, 2003a; Li etal., 2000). CH4 is a common fluid species in hydrother-mal systems in the oceanic crust and commonly formseither by reactions involving magmatic CO2 or duringserpentinization of olivine and/or other mafic phases in-volving CO2-rich fluids (Kelley and Früh-Green, 1999).So far there is only indirect evidence for the presence ofCH4 in subduction zones in shallow (1–3 km depth) CH4-rich plumes emanating from the accretionary prisms inconvergent margins (Craig et al., 1987; Watanabe et al.,1994; Tsunogai et al., 1998). Recent investigations toconstrain the retention and loss of volatile elements suchas CH4 during subduction showed that fluxes of carboninto subduction zones are larger than returned to the sur-face, thus indicating that CH4 could also occur in deeperlevels of subduction zones (see Sadofsky and Bebout,2003; Shaw et al., 2003).

CH4-bearing fluid inclusions in jadeite have not beenreported before the investigations of Shi et al. (2000) onthe jadeitites from the famous Myanmar jadeite Tract andthus could help to put constraints on the flux of CH4 inpaleo-subduction zones. In this paper we describe CH4-bearing fluid inclusions in jadeite crystals from theMyanmar jadeitites, their composition including theδD(H2O) and δC(CH4) isotopes and the implications forcarbon cycling in paleo-subduction zones.

SAMPLE SELECTION AND DESCRIPTION

The outcrops are located in an area in the western partof the Sagaing strike-slip fault belt in the Parkhan (orcalled Hpakan or Pharkan) city, Kachin state, Myanmar(central location at N25°36.9′, E96°18.6′ determined byGPS measurements, Figs. 1A and B). The famous Bur-mese (Myanmar) jadeite occurs in this belt. Primaryjadeite deposits occur as veins cross-cutting theserpentinized peridotite bodies called the Pharkan-Tammaw ultramafic body, which is part of the Indo-Burmarange of ophiolites, and Chhibber (1934) described theseveins as a “dike”. As there is no strong evidence for amagmatic origin, we prefer the term “vein”. The jadeiteveins are almost vertical, strike N-S, and have a width of0.5 m up to 5 meters and a length of about 10–100 me-

ters. According to Chhibber (1934), the boundary to thesurrounding serpentinites is marked by a zone that con-sists of a mixture of chlorite, occasionally with calcite,actinolite, and talc. In contrast, we only observed jadeititeveins with complex sodic- and sodic-calcic amphiboleboundary zones on both sides (Shi et al., 2003). In a fewoutcrops, thin veins of albite, commonly less than 5 mmwide, cross-cut the jadeitite veins and thus obviously

Fig. 1. (A) Geological overview of the jadeitite area located inthe west of Sagaing strike-slip faults belts within the collisionzone between the Indian Plate and the Yangtze Block, modifiedafter Bender (1983); (B) Geological sketch map of the Myanmarjadeite area, the primary jadeitites occur in ultra-mafic rockbodies, modified after Chhibber (1934).

Fluid inclusions in Myanmar jadeitite 505

formed at a later stage. The amphibole boundaries have awidth of 1–50 cm and fragments of the amphibole felscan be found within jadeitite veins due to later deforma-tion. Within or adjacent to the amphibole boundary zones,jadeitite fragments containing kosmochlor or chromianjadeite can often be found as roundish aggregates orblocks wrapped by amphiboles. Outside the ultramaficbodies, high-pressure metamorphic rocks such asphengite-bearing glaucophane schists, and amphibolite-facies rocks such as garnet-bearing amphibolites,diopside-bearing marbles and stilpnomelane-bearingquartzites occur (Shi et al., 2001).

We examined about 150 specimens of jadeitites,jadeite-kosmochlor-bearing jadeitites, and rocks from theamphibole boundary zones (Shi et al., 2003). Throughmicroscopic observations on thin sections of these sam-ples, about 10 samples of undeformed coarse-grainedjadeitites were found containing a sufficient number offluid inclusions.

Polished thin- and thick sections were made for pet-rographic examination, microthermometry using the heat-ing/cooling stage, and micro-Raman spectroscopy to ob-tain the chemical composition of individual fluid inclu-sions. In a second step, rock chips were crushed to obtainjadeite grains which were then prepared for bulk chemi-cal analyses of the fluid inclusions. The fluid inclusionsin the jadeite grains were then decrepitated to obtain theirbulk composition. The bulk chemical methods performedin this study were ion chromatography to obtain the ioniccontent of the fluid inclusions and quadrupole massspectrometry (QMS) to obtain the bulk H2O contents ofthe fluid inclusions. In addition, six samples were alsoused for bulk stable isotope measurements of δD and δ13Cby gas-source mass spectrometry.

ANALYTICAL TECHNIQUES

The chemical composition of the jadeites was deter-mined by electron microprobe analysis (EMPA) in theelectron microprobe laboratory of the Institute of Geol-ogy and Geophysics, Chinese Academy of Sciences(IGGCAS). The EMPA data were obtained with aCAMECA CAMEBAX SX51 system with the analyticalconditions of 15 kV and 12 nA beam current.

Microthermometric investigations of 0.2 mm thickpolished sections were performed with a LINKAMTHMASG600 heating-cooling stage at JICA lab,IGGCAS, with a heating rate of 1 to 5°C/min and a cool-ing rate of 0.5 to 3°C/min.

Micro-Raman spectroscopy was performed with aRenishaw-1000 Micro Raman spectrometer with an ar-gon laser beam (λ = 514.5 nm) and analytical conditionsof 5 mW on thick sections (0.2 mm) of coarse-grainedjadeitites. The detection limits of the method for the dif-

ferent gases are 0.5 mol.%. An objective lens with a mag-nification of 50 times was used.

Quadrupole mass spectrometry, which is used to ob-tain information about additional volatiles in the indi-vidual aqueous inclusions (Sasada et al., 1992; Küsterand Stöckhert, 1997; Röller et al., 2001), was performedto check whether small amounts of additional carbon-bear-ing compounds besides CH4 exist. Double-polished thicksections with a size of 2 × 2 × 0.2 mm were prepared andinserted in the extraction tube and heated up stepwise to800°C within 150 minutes in high vacuum. During heat-ing, individual fluid inclusions decrepitated due to inter-nal overpressure which leads to the formation of smallpressure pulses which are recorded. The released gaseswere continuously analyzed with a QMS (RG202 at JICAlab, IGGCAS) in a rapid scanning mode (50 msec/amu)for the following molecular masses: 16(CH4), 18(H2O),28(CO) and 44(CO2). With this method, the compositionof individual inclusions in terms of their volatile compo-nents, size and temperature of decrepitation can simulta-neously be recorded.

In addition to the QMS measurements, we also usedan ion chromatograph (SHIMADZU, HIC-6A) at JICAlab, IGGCAS to measure the bulk cation and anion con-tents of the fluid inclusions. For this purpose, the selectedjadeitites were crushed and sieved with a 20–80 mesh.One gram of jadeite grains was picked out from each sam-ple and cleaned with highly distilled water until the con-ductivity of the water used was the same as pure water.The grains were then dried and heated up to 1000°C todecrepitate all fluid inclusions. Afterwards 5 ml of ultra-pure water was added to the decrepitated specimens todissolve any soluble ions, which was then analyzed withthe ion chromatograph.

Water was extracted from the fluid inclusions by in-duction heating in a vacuum extraction line. The materialof the line is Pyrex except for the sample tube and thecrucible holder, which are made of silica glass. The sam-ples were put into a Pt crucible and were heated stepwisewith an induction furnace. The temperature was increasedstepwise from room temperature up to about 1400°C. H2Owas cryogenically separated from the other volatile spe-cies that were first condensed by liquid-N2 trap connect-ing to sample tube. H2O was then converted into H2 gasby exposing it to hot uranium at 650°C. H2 was then col-lected with the Topler pump and sealed in a sample tubefor mass spectrometry analysis.

CH4 was extracted from the fluid inclusions usingsealed tube combustion. The samples together with cop-per-oxide threads were put into a Pyrex reaction tube andthen sealed in vacuum. The sealed tubes were then heatedup to 800°C and held within a period of 20 hours to reachcomplete conversion of CH4 into CO2 according to thereaction CH4 + 4CuO = 4Cu + 2H2O + CO2. Subsequently,

506 G. Shi et al.

CO2 was then collected in high vacuum and sealed in asample tube for mass spectrometric analysis. D/H ratiosand 13C/12C were measured with a double collector gassource mass spectrometer (MAT 252 and MAT 251) atIGGCAS, and reported in the δD and δ13C notations withthe references of VSMOW and VPDB, respectively. Ana-lytical uncertainties are about ±3‰ and ±0.5‰ for δDand δ13C, respectively.

PETROGRAPHY AND MINERAL CHEMISTRY

OF THE JADEITITES

Most jadeitites are strongly deformed and fine-grained

Fig. 2. Photomicrograph of the textures in the jadeitites show-ing closely associated undeformed and deformed jadeitite.

Fig. 3. (A) Cathodoluminescence (CL) image illustrating com-plex rhythmical zoning patterns within undeformed jadeite crys-tals; (B) Backscatter electron (BSE) image of the rhythmicchemical zoning in an undeformed jadeite crystal. The lightgrey zones are Ca-rich and the dark grey zones are Ca-poor.The chemical substitution occurs along the exchange vector Ca+ Mg ⇔ Na + Al.

Fig. 4. Photomicrographs of fluid inclusions: (A) The H2O-rich fluid inclusions (FI) show a gaseous bubble (G) in a liquidmatrix (L) and they also display a great variability in size. Somefluid inclusions show a strong elongation along the c-axis. (B)Interspersed among the larger H2O-rich fluid inclusions aresome CH4-rich fluid inclusions.

Fluid inclusions in Myanmar jadeitite 507

and occasionally domains with coarse-grained, less- orundeformed, jadeite crystals occur (Fig. 2). A fewjadeitites reveal a primary vein structure, with largeeuhedral to subhedral grains. Cathodoluminescence im-ages reveal a rhythmic zoning pattern of the coarse-grained jadeite crystals as shown in Fig. 3A, similar tothe observations of Harlow (1994) and Sorensen andHarlow (1998) from the Guatemalan jadeitites. Thejadeitites of this study are monomineralic and albite andanalcite rarely occur as secondary alteration products. Thechemical compositions of jadeite are given in Table 1.Euhedral jadeites are very pure, with jadeite (XJd) con-tents of more than 98 mol.%. Backscatter electron (BSE)images of some of the coarse-grained jadeiteporphyroblasts reveal a chemical zoning pattern as shownin Fig. 3B with rhythmically increasing Ca and Mg con-tents (light grey zones in the image) ranging from 0.01a.p.f.u. to 0.08 a.p.f.u. thus leading to variations in XJdfrom 0.91 to 0.99 (Table 1).

Sample Dc1 Dc2 Dc3 Df1 Df2 Fc1-4 Fc1-6 Fc1-8

SiO2 58.50 57.9 59.01 58.4 57.62 57.59 58.36 58.37TiO2 n.d. n.d. 0.04 n.d. 0.01 n.d. n.d. n.d.Al2O3 26.26 26.29 26.60 26.06 26.25 23.56 23.18 22.67Cr2O3 n.d. 0.06 0.03 0.01 0.05 0.05 0.02 n.d.Fe2O3 n.c. n.c. n.c. n.c. 0.69 n.c. n.c. n.c.FeO 0.03 n.d. 0.07 0.15 n.d. 0.91 0.44 0.76MnO 0.07 0.14 n.d. 0.04 0.11 n.d. 0.04 0.01NiO n.d. n.d. 0.06 0.13 0.04 0.01 n.d. 0.04MgO 0.05 0.05 0.09 0.21 0.04 1.64 1.26 1.60CaO 0.05 0.12 0.12 0.26 n.d. 2.05 1.59 1.96Na2O 14.96 14.60 14.62 15.04 14.94 13.88 14.41 14.13K2O 0.02 0.01 n.d. n.d. n.d. 0.03 0.01 0.02∑ 99.94 99.17 100.64 100.3 99.75 99.72 99.31 99.57

Si 1.97 1.97 1.98 1.96 1.95 1.96 1.99 1.98Ti n.d. n.d. <0.01 n.d. <0.01 n.d. n.d. n.d.Cr n.d. <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 n.d.Al 1.04 1.05 1.05 1.03 1.05 0.94 0.92 0.91Fe3+ n.c. n.c. n.c. n.c. 0.02 n.c. n.c. n.c.Fe2+ <0.01 n.d. <0.01 <0.01 n.d. 0.02 0.01 0.02Mn <0.01 <0.01 n.d. n.d. n.d. n.d. <0.01 n.d.Mg <0.01 <0.01 0.01 0.01 n.d. 0.08 0.06 0.08Ni n.d. n.d. <0.01 <0.01 <0.01 <0.01 n.d. n.d.Ca <0.01 n.d. n.d. 0.01 n.d. 0.08 0.06 0.07Na 0.98 0.96 0.95 0.98 0.98 0.91 0.94 0.92K <0.01 <0.01 n.d. n.d. n.d. n.d. <0.01 n.d.∑cat. 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00

XJd 1.00 0.99 0.99 0.99 0.98 0.91 0.93 0.91

Table 1. Chemical compositions of jadeites (wt.%)

The pyroxene formulae were calculated on the basis of 6 oxygens and 4 cations. Fe3+ was calculated from charge balance considerations. Theanalyses Dc1, Dc2, Dc3 are from euhedral-subhedral coarse-grained jadeites whereas the analyses Df1 and Df2 are from fine-grained jadeites.Fc1-4, Fc1-6 and Fc1-8 are from a zoned jadeite porphyroblast. n.d.: not detected; n.c.: not calculated; XJd = Na/(Na+K+Ca).

Gas content (vol.%) Tm (°C) Th (°C)

20 –4.3 36515 –1.5 31538 –1.8 31828 –3.7 32523 –5.5 32131 –3.0 33020 –3.3 31830 –2.1 32821 –3.1 378

Table 2. Microthermometric results ofindividual fluid inclusions

Tm = melting temperature, Th = homogenization temperature.

OPTICAL DESCRIPTION OF THE FLUID INCLUSIONS

Two-phase, gas/liquid fluid inclusions were only foundin coarse-grained jadeite crystals. Large fluid inclusions(ca. 10 µm), which are about 10% of all inclusions, con-

508 G. Shi et al.

tain small visible bubbles (Figs. 4A and B), while thebubbles of the small ones, especial the very small ones(<3 µm), are not visible under the polarization micro-scope. The fluid inclusions appear to be of primary ori-gin, since they are elongated along the c-axis of the crys-tals, show a random distribution throughout the jadeitegrains, and no occurrence of inclusions along healed frac-tures can be found. The volumetric contents of gaseouscomponents in the larger fluid inclusions ranges from 15to 38 vol.% as shown in Table 2.

CHEMICAL ANALYSIS OF

INDIVIDUAL FLUID INCLUSIONS

MicrothermometryThe investigations were carried out only on fluid in-

clusion in size large enough to permit accurate measure-ments. The fluid inclusions were brought to temperaturesof –180°C and then subsequently heated up. The finalmelting temperatures of fluid inclusions show a bimodaldistribution. Some small fluid inclusions show freezingtemperatures below –180°C, indicating CH4-rich compo-sitions, while most fluid inclusions show melting tem-peratures ranging from –1.5 to –5.5°C, which indicatesH2O-rich fluid inclusions, containing small amounts ofdissolved additional components such as salts (e.g., NaCl,KCl, CaCl2) (Table 2). Homogenization temperatures (L+ V = L) of the fluid inclusions range from 315 to 378°C.Based on the homogenization temperatures and the volu-metric amount of gaseous components it was possible tocalculate the chemical composition of the H2O-rich fluidinclusions in terms of X(CH4) and X(H2O) with the pro-gram BULK by Bakker (2003). The data show for theH2O-rich fluid inclusions calculated X(H2O) of 0.995 andX(CH4) of 0.005. Calculations of X(CH4) and X(H2O)for the CH4-rich fluid inclusions was not possible sincethe CH4 homogenization temperature (L + V = V) couldnot be determined due to the small size which preventedclear observation of CH4 homogenization.

Micro-Raman spectroscopyAbout 30 inclusions were measured with a laser beam

with a diameter of ca. 20 × 20 µm and representative spec-tra of the detected components in the fluid inclusions areshown in Figs. 5A, B and C. According to the Ramanspectra, H2O and CH4 are the only components whichwere identified in these fluid inclusions, other chemicalcompounds such as carbon dioxide (CO2), carbon mon-oxide (CO) or ethane (C2H6) have not been detected. Thespectra clearly indicate two groups of fluid inclusions,H2O-rich/CH4-poor fluid inclusions (Fig. 5A) and CH4-rich/H2O-poor fluid inclusions (Figs. 5B and C). The CH4-rich/H2O-poor fluid inclusions occur only as very smallfluid inclusions.

BULK CHEMICAL ANALYSIS OF THE FLUID INCLUSIONS

Quadrupole mass spectroscopyThe composition of individual fluid inclusions in terms

of their CH4, CO and CO2 contents indicates that the fluidinclusions contain CH4 as the only carbon-bearing spe-cies. The data suggest that decrepitation of fluid inclu-sions is also a function of size since most of the largefluid inclusions decrepitate within a temperature rangeof 190–275°C whereas most of the smaller ones decrepi-tated between 475–550°C.

In order to obtain bulk molar ratios of H2O versus othergaseous components, the crushed and sieved jadeite grains(about 0.3 g weight of 20–40 mesh grains) were put intoa Silex glass tube connected to the QMS, heated up to

Fig. 5. Raman spectra of the fluid inclusions: (A) H2O-richfluid inclusion showing a very small peak of CH4; (B) CH4-rich fluid inclusion showing a small H2O peak; (C) CH4-richfluid inclusion showing nearly no H2O peak.

Fluid inclusions in Myanmar jadeitite 509

about 700°C and held at this temperature for about 10minutes in high vacuum. The decrepitated gases were thencondensed in traps cooled by liquid nitrogen to preventoccurrence of reactions among the different gas species.After the hot tube cooled down to room temperature, liq-uid nitrogen was removed and the total velamen pressureof all gaseous components was recorded. Afterwards, H2Owas frozen by using dry ice slush and subsequently theH2O-absent velamen pressure was then recorded. Themolar ratio of H2O versus the other components in thegas can be calculated by using the ratio of pressure re-duction due to the removal of H2O versus the total vela-men pressure. The results show that the bulk composi-tion of the fluid inclusions is very H2O-rich with molarratios (XH2O) ranging from 87 mol.% to 94 mol.% H2O.

Ion chromatographyThe samples are characterized by very low ionic con-

tents, the most abundant cation being Na+ and the mostabundant anion is Cl–. Due to the very low Cl–, Ca andMg contents, only very small amounts of salts such asNaCl, CaCl2 and MgCl2 can thus be present in the fluidinclusions (Table 3).

Stable isotope analysisResults of hydrogen and carbon isotope analyses are

listed in Table 4. The Raman and QMS measurementssuggest that the δ13C/12C values reported here can be at-tributed to CH4 from the decrepitated fluid inclusions.The isotopic composition of the fluid inclusions from this

study, is characterized by δD(H2O) values ranging from–56.3 to –49.8‰ and δ13C(CH4) values ranging from–30.1 to –25.5‰ (Table 4).

DISCUSSION

P-T constraints on the entrapment of the fluid inclusionsFrom the coexistence of CH4- and H2O-rich, vapor

and liquid-dominated fluid inclusions, it is supposed thatthe fluids were possibly entrapped under immiscible con-ditions (Kelley, 1996). QMS measurements of the bulkchemical composition of the fluid inclusions have shownthat the bulk fluids contain 87–94 mol.% H2O. Experi-mental investigations of CH4-H2O-fluids indicate that attemperatures below 400°C a homogeneous CH4-H2O-fluid will unmix into a CH4- and a H2O-rich fluid, thusputting an upper temperature constraint on the entrapmenttemperature of the fluid inclusions (Holloway, 1984). Therange of the homogenization temperatures of the fluidinclusions of 315–378°C thus yields a lower temperaturelimits for the formation of these fluid inclusions. The realtrapping temperatures are expected to be higher if a pres-sure correction of the homogenization temperature is ap-plied. Huizenga (2001) shows in a thermodynamic evalu-ation of the C-O-H system that CO2-poor fluids are sta-ble only at temperatures below 400 to 450°C, dependingon the fO2 and whether graphite was present. Calculationof isochores, based on the composition of the fluid inclu-sions has not been attempted for the following reasons:(1) the lack of microthermometric data of the CH4-rich/H2O-poor fluid inclusions, and (2) the high probabilityof density modification during uplift, which leads to con-siderable pressure underestimations as shown in manystudies of fluid inclusions in high pressure rocks (seePhilippot and Selverstone, 1992; Selverstone et al., 1992;Sterner and Bodnar, 1989). Despite this discrepancy, fluidinclusions from high-pressure rocks still give insight onthe fluid/rock interactions at high pressure conditions(Philippot and Selverstone, 1992; Selverstone et al.,1992). The determination of the pressures of the forma-tion of the jadeites is hampered by the lack of appropri-ate equilibria and the lack of quartz. Equilibria such asjadeite = albite + nepheline and jadeite + H2O = analcimethus provide only lower pressure limits since the low-Pphases are absent in the rocks. For the temperature rangeof the investigations, application of the analcime break-down reaction according to the calculations of Harlow(1994) yields lower pressure limits of 0.6 to 0.7 GPa be-tween 250 and 400°C for the formation of the Guatemalajadeites. But this pressure is below the lower stability limitof jadeite + H2O relative to analcite at 0.8 to 1.0 GPa(Boettcher and Wyllie, 1969), even slightly lower than alower limit of formation of omphacite + quartz of 0.6 to0.8 GPa at 200 to 300°C in Californian metamorphic rocks

Sample F– Cl– SO42– Na+ K+ Mg2+ Ca2+

Fc3 0.03 0.38 0.19 3.92 0.16 0.05 0.18Fc5 n.d. 0.02 0.02 3.00 0.14 0.05 0.14Fr1 0.07 0.18 0.19 3.21 0.20 0.19 0.22Fc1 0.02 0.20 0.09 4.40 n.d. 0.07 0.08Fc8 n.d. 0.12 0.19 3.58 n.d. 0.02 0.11

Sample δD(H2O)‰ δ13C(CH4)‰

FC1 –51.8 –25.5FC2 –49.8 –26.2FC3 –52.3 –30.1FC5 –52.3 –28.7FC6 –56.3 –28.7FC8 –51.5 –26.1

Table 3. Bulk ionic contents of the fluid inclusions

The contents (µg/g) of ions were measured with detection limits rang-ing from 2.4–3.9 ng/ml (Zhu, 1999); n.d.: not detected.

Table 4. Bulk stable isotope composi-tions of fluid inclusions

510 G. Shi et al.

(Essene and Fyfe, 1967). A recent study on the Myanmaramphibole felses yielded a minimum pressure of 1.0 GPaat temperatures of about 250 to 370°C (Shi et al., 2003),which are consistent with temperatures inferred from thefluid inclusions in this investigation.

Biogenic versus abiogenic formation of CH4 in thejadeitite veins

CH4 is the only hydrocarbon that was found in thefluid inclusions of this study. CH4 occurring in the Earth’scrust is either formed by abiogenic or by biogenic proc-esses. Abiogenic processes involve outgassing of primor-dial CH4 from the mantle, the formation of CH4 duringinorganic chemical reactions involving species such asCO2, H2 or other C-H molecules at temperatures >300°C(Fischer-Tropsch-type reactions) or thermogenic proc-esses (thermochemical decomposition of organic matter).Biogenic processes involve the alteration of carbonthrough bacterial processes (Schoell, 1988). AlthoughCH4-rich fluid inclusions have been frequently observedin nominally fluid-absent rocks such as peridotites andgabbros (Kelley, 1996; Kelley and Früh-Green, 1999; Suet al., 1999; Yang et al., 2001), their presence is not anunambiguous criteria for abiogenic formation and hencecarbon isotopes (δ13C) are a widely used tool in distin-guishing between biogenic and abiogenic processes re-sponsible for the formation of CH4, although an attribu-tion to either one or the other process is still a matter ofdebate (Kelley and Früh-Green, 1999; Whiticar, 1999).

Abiogenic CH4 is characterized by highly enrichedδ13C(CH4) ranging from –25 to –8‰ as indicated by fluidinclusions from plutonic samples from the East PacificRise and the Southwest Indian Ridge (Welhan and Craig,1983; Kelley and Früh-Green, 1999). Abiogenic CH4 isalso accompanied by measurable quantities of hydrocar-bons such as ethane, propane and butane (Sherwood Lollaret al., 1983; Abrajano et al., 1988). On the other hand,biogenic CH4 is characterized by highly depletedδ13C(CH4) with values ranging from –110 to –50‰ andlow contents of additional hydrocarbons (Whiticar, 1999;Charlou et al., 2002). In contrast to CH4 from bacterialprocesses, thermogenic CH4 is enriched in 13C-compoundwith a δ13C(CH4) of –50 to –20‰.

The isotopic composition of CH4 in the fluid inclu-sions of this study is characterized by δ13C(CH4) valuesranging from rom –30.1 to –25.5‰, which could be in-dicative of an abiogenic origin since these data are veryclose to abiogenic CH4 (East Pacific Rise, Welhan andCraig, 1983; Zambales Ophiolite, Abrajano et al., 1988).Although mantle carbon either occurs as CO or as CO2with δ13C values of –7 to –5‰, several studies have shownthat mantle peridotites can contain significant quantitiesof reduced carbon species such as CH4 which show dis-tinctly lighter carbon isotope ratios of –30 to –20‰ (see

Kelley and Früh-Green, 1999 and references therein).Moreover, seemingly primordial CH4-rich fluid inclusionsfrom mantle peridotites are very rare and contain signifi-cant amounts of additional components such as hydro-carbons, H2, N2, H2S and CO2 (Su et al., 1999; Yang etal. , 2001). The formation of CH4 in most mantleperidotites though is probably due to reduction of CO2during water-rock interaction since there is no evidencethat CH4 is present in the upper mantle in other than mi-nor concentrations, because the oxidation state of theupper mantle is too high (Apps and van de Kamp, 1993).In addition, several studies have shown that negative δ13Cvalues from alkali basalts and peridotite xenoliths fromEast China might result from recycled crustal components(e.g., see Kessen and Ringwood, 1989). Yang et al. (2001)found CH4-rich fluid inclusions in peridotite xenolithsfrom the Dabie terrane and their investigation indicatedthat the CH4 most likely formed due to the influence ofCO2-rich aqueous fluids during serpentinisation prior toor during plate subduction. These observations lead anumber of studies to conclude that later-stage abiogenicformation of CH4 instead of degassing of primordial CH4is more common than previously thought (e.g., seeCharlou et al., 2002 and references therein). In addition,recent experimental investigations by Berndt et al. (1996)and Horita and Berndt (1999) have shown that CH4 canbe produced by serpentinization of olivine in the pres-ence of Fe-Ni alloys through reactions involving H2O andCO2 on the reactant side such as forsterite + H2O + CO2= magnetite + serpentine + brucite + CH4 + H2 or forsterite+ H2O = magnetite + serpentine + brucite + H2 where H2reacts with CO2 to form CH4 and H2O according to thereaction CO2 + 4H2 = CH4 + 2H2O (Fischer-Tropsch syn-thesis). This process might be able to explain the occur-rence of abiogenic CH4 in shallow plumes in sedimentlayers above ultramafic rocks on the oceanic floor (Ronaet al., 1992; Charlou and Donval, 1993), fluid inclusionsin olivine-rich drill cores from the oceanic crust (Kelley,1996), CH4-rich gas hydrates in subduction zones (Kastneret al., 1998) and the occurrence of CH4-rich gases ema-nating from ophiolite complexes (Abrajano et al., 1988;Sturchio et al., 1989).

Although the jadeitites occur within strongly altered(serpentinized) peridotites and the isotope signature ob-tained from the fluid inclusions in the jadeitites point toa possible abiogenic origin, fluids which are attributed tothe formation by Fischer-Tropsch synthesis should con-tain significant amounts of H2 (Kelley and Früh-Green,1999; Charlou et al., 2002) which has not been found inthe fluid inclusions of this study. However, H diffusionout of the system can be quite faster even at low tempera-ture. A recent study by McCollom and Seewald (2001),on the other hand, has shown that Fischer-Tropsch-typereactions alone cannot account for large-scale reduction

Fluid inclusions in Myanmar jadeitite 511

of dissolved CO2 to CH4 since the conversion of CO2 intoCH4 in their experiments was highly insufficient due tokinetic reasons despite extremely long run times (2500h). In addition, the presence of a Ni-Fe alloy also seemsto be of crucial importance in the conversion of CO2 intoCH4, although the effectiveness of this alloy over timeseems to decrease strongly (Horita and Berndt, 1999). NoNi-Fe alloys have been found in the adjacent serpentinizedperidotites surrounding the jadeitites in this investigationso far and only magnetite has been found. Several stud-ies also concluded that fluids from peridotite-hosted hy-drothermal systems where incomplete conversion of CO2into CH4 took place, should thus contain large fractionsof formate (McCollom and Sewald, 2001; SherwoodLollar et al., 2002). The absence of Ni-Fe alloys in theperidotites therefore does not seem to indicate abiogenicformation of these fluid inclusions by serpentinization ofthe surrounding ultramafic country rocks.

Thermogenic CH4 is formed as organic-rich sedimentsmove through progressively higher thermal maturity re-gimes during increasing depths of burial. The main proc-ess of formation is thermochemical decomposition of or-ganic matter. During early maturation (<150°C) thermalCH4 is accompanied by other hydrocarbons and non-hy-drocarbon gases, whereas at highest thermal maturities,CH4 alone is formed by either breaking C-C bonds inkerogen, bitumen and oils or from the reaction betweenwater and graphite (Wiese and Kvenvolden, 1993). Theisotopic composition of these thermal gases depends on:(1) the isotopic composition of the source rock, (2) thedegree of fractionation during thermogenesis, and (3) thedegree of fractionation after CH4 formation. Thermal CH4has a range of carbon isotopic values from ca. –50 to –25‰ depending on the degree of fractionation of the light-est and/or heaviest source material (Fuex, 1979). There-fore, depending on the carbon isotope separation betweenthermogenic CH4 and organic matter [δ13C(CH4) –δ13C(Corg)] which ranges between 0 and ca. 30‰, ther-mogenic CH4 might still retain the isotopic signature ofthe precursor organic matter which has on average a δ13Cof –25 ± 1.5‰ (Deines, 1980).

The stable isotopic signature of hydrogen in most casesconsists of at least two sources, a mantle-derived magmaand/or a secondary source (caused by crustal subductionor fluid venting). The hydrogen isotopic compositions ofhydrous minerals or fluid inclusion in high pressure (HP)rocks are rarely homogeneous and reflect local equilib-rium (Scambelluri and Philippot, 2001). It is generallyassumed that fluids in the upper parts of a subduction zonerepresent seawater and δD(H2O) will start changing whenthe water/rock ratio is sufficiently reduced so thatδD(H2O) of the fluid is influenced by isotopic exchangewith hydrous minerals or by hydrous mineral breakdown(Johnson and Harlow, 1999). At high pressure the H iso-

tope signatures of serpentinites spread over a large rangeof values which reflected pre-eclogitic interactions withseawater-derived fluids and with (D-depleted metamor-phic fluids released at fore-arc environments (Früh-Greenet al., 2001). This will lead to a considerable shift toisotopically lighter δD(H2O) values due to the contribu-tion of metamorphic water which has a δD(H2O) of –70to –20‰. The isotopic composition of the fluid inclusionsfrom our study is characterized by δD(H2O) values rang-ing from –56.3 to –49.8‰, suggesting that the water inthe fluid inclusions in the jadeites might reflect a strongcomponent of metamorphic water released by dehydra-tion reactions within the metamorphosd rocks. This dif-fers from the Guatemala jadeites which are isotopicallyheavier with a more distinct seawater signature (Johnsonand Harlow, 1999). Johnson and Harlow (1999) also con-cluded that their isotopic shift to heavier δD(H2O) valueswas due to serpentinization which also was suspected tolead to the formation of H2, which is completely absentin our fluid inclusions.

On the other hand, δD(H2O) values of hydrous miner-als from ultramafic mantle rocks are also in a similar rangefrom –90 to –40‰. Typical signature of meteoric waterwas observed in hydroxyl-bearing minerals from UHPeclogites and gneisses from the Dabie-Sulu terranes, withδD values of –127 to –61‰ for micas, –100 to –72‰ foramphiboles, –75 to –37‰ for epidotes (Rumble and Yui,1998; Zheng et al., 1998, 1999; Fu et al., 1999; Xiao etal., 2002; Li et al., 2004). Compared to fluid inclusionsfrom the eclogites, the δD(H2O) values from our studynot only fall within the δD range of hydroxyl-bearingminerals, but also are very similar to the δD(H2O) ob-tained from the Monviso eclogites from the Western Alps(Nadeau et al., 1993), although both localities show amuch larger variation in δD due to small scaleheterogeneities. Overall, the δD(H2O) data favor a crust-derived fluid with insignificant input from seawater.

The origin of carbon and CH4 in high-pressure rocks andimplications for carbon cycling in subduction zones

Recent investigations to constrain the retention andloss of volatile elements during subduction showed thatfluxes of carbon into subduction zones are larger thanreturned to the surface in arcs (Sadofsky and Bebout,2003; Shaw et al., 2003). If this is true, then large amountsof carbon would be transported into the deeper mantleand could contribute to mantle carbon budgets and thusto carbon (CO2, CH4) fluxes in ocean-island basalts andat mid-oceanic ridges (e.g., see Marty and Tolstikhin,1998; Charlou et al., 2002 and references therein) as wellas in HP and UHP metamorphic rocks from continentalsubduction zones (e.g., see Yang et al., 2001; Zheng etal., 2000, 2003a, 2003b). For instance, several fluid in-clusion studies from the UHP Dabie terrane have shown

512 G. Shi et al.

that organic carbon seems to play an important role inthe carbon budgets of paleo-subduction zones (Yang etal., 2001; Zheng et al., 2000, 2003b). Stable isotope stud-ies on fluid inclusions in peridotites, enclosed in eclogites,revealed 13C-depleted CO2 (down to –25.1‰) present inthis UHP terrane, which is most likely derived from oxi-dation of organic matter during interaction with surfacefluids during prograde UHP metamorphism. On the otherhand, fluid inclusions from eclogites also showed enrich-ment in δ13C (–18.5 to 4.6‰), indicating that decarboni-sation of marbles also contributes to the δ13C signatureof CO2 (Yang et al., 2001). Unfortunately, no such dataon CH4-rich fluid inclusions in paleo-subduction zones,yet exist.

Most of the evidence for the presence of CH4 at deepcrustal levels (10 to 35 km) comes from graphite-bearingmetasedimentary rocks where reactions with graphiticcarbon are the source for CH4 (Burruss, 1993). Still pri-mary CH4-rich fluid inclusions rarely occur in high-grademetamorphic rocks except for some occurrences ingranulites (Hall and Bodnar, 1989; Hurai et al., 2000)and very rare occurrences in eclogite-facies rocks (Klemdet al., 1992, 1995; Janak et al., 1999). Several studiesfrom the Dabie-Sulu terrane have also shown CH4-bear-ing, CO2-rich fluid inclusions in the eclogites and felsicgranulites (Xiao et al., 2000, 2002; Fu et al., 2001, 2002,2003a, 2003b; Yang et al., 2001) and the low δ13C signa-ture in the UHP eclogites and gneisses (Zheng et al., 2000;2003a, 2003b; Li et al., 2000). These studies show thatthe highest-grade fluids are comprised of N2 ± CO2 andCH4-bearing fluid inclusions only occur at the retrogradeportion of the P-T path. Thus no data on the presence ofCH4-rich fluid inclusions during growth of high-P miner-als such as jadeite or omphacite yet exist. Our data thusindicate that CH4-bearing fluids can still occur during theprograde path of high-P rocks such as jadeitites.

Unfortunately, in the literature there is only indirectevidence for the presence of CH4 in subduction zones sofar, namely in CH4-rich plumes emanating from oceanictroughs in convergent margins (Craig et al., 1987;Watanabe et al., 1994; Tsunogai et al., 1998). Theseplumes form in the accretionary prisms of the subductionzones and rise from depths of 1 to 3 km and their isotopicsignatures (δ13C) indicated either CH4 production by mi-crobial activity in shallow sedimentary layers or ther-mogenic production of CH4 from organic sediments(Haggerty, 1991; Tsunogai et al., 1998). Thus the occur-rence of CH4-bearing fluid inclusions in high-P rocks fromsubduction zones helps to put constraints on the stabilitylimits of CH4-bearing fluids as well as constraints on theflux of CH4 in subduction zones. Sadofsky and Bebout(2001, 2003) investigated low-T high-P rock suites fromthe Franciscan Complex and the Western Baja Terrane(Mexico) to study the influence of forearc devolatilization

on geochemical cycling in convergent margins. Theirstudy showed that rocks from the Franciscan Complexcontained carbonate veins with δ13C values very close tomarine carbonates (most between 0 and –10‰) and re-duced graphite with organic-like δ13C (ca. (25‰) thusindicating that most of these rocks have not experiencedsignificant decarbonisation. In cases where decarbonisa-tion took place, shifts to higher δ13C of the reduced car-bon occurred. Sadofsky and Bebout (2001, 2003) showedthat carbon in the Franciscan Complex originated frommarine sediments, thus organic material was the sourceof the carbon.

Assuming that thermogenesis of organic carbon is themost likely mechanism of CH4 production in the fluidinclusions of this study, Fuex (1979) showed that a sourcerock with organic carbon having δ13C of –19‰, whichexperience minimum fractionation of the heaviest sourcematerial, is able to produce CH4 which has a δ13C of –25‰. Considering that average marine organic matter hasa δ13C of –25 ± 1.5‰ (Deines, 1980), formation of CH4from a mixture of organic carbon and carbon whichevolved from carbonates, would allow the formation ofthermogenic CH4 with a δ13C similar to the isotopic val-ues yielded from our fluid inclusions. Figure 6 shows asimple ideal calculation of the evolution of δ13C(CH4) byRayleigh distillation in a closed system when it evolves

Fig. 6. Evolution of δ13C(CH4) (‰ vs. PDB) by Rayleigh dis-tillation in a closed system at temperatures of 200°C, 300°Cand 400°C from graphite (CH4-C; continuous lines) and cal-cite (CH4-cc; stippled lines). The fractionation factors are fromBottinga (1969) and the shaded area indicates the range of bulkδ13C(CH4) of the fluid inclusions from this study.

Fluid inclusions in Myanmar jadeitite 513

either from graphite (starting δ13C = (25‰) or calcite(starting δ13C = 0‰). The calculations and the tempera-tures of formation of these fluid inclusions indicate thatthe stable isotope signatures of the fluid inclusions fromthe jadeitites rather point to an origin from graphite afterhigh fractionation of CH4 out of it instead of forming fromcalcite alone by decarbonisation, which is very similar tothe formation of the CO2-rich fluids in the Dabie-Suluterranes (Zheng et al., 2000; Yang et al., 2001). On theother hand, mixtures of both sources allows δ13C(CH4)values which are between these two end-member sce-narios and possibly close to our obtained data. Althoughwe do not have any isotopic data of carbon (graphite) fromwhole-rock samples in this study, but diopside-bearingmarbles do occur in the vicinity of these veins,fractionation data indicate that the δ13C(CH4) values ofthe fluid inclusions from these rocks from a similar low-P high-T setting may have evolved either by (1) gradualfractionation from an isotopically similar protolith source(organic carbon from marine sediments) or (2) anisotopically heavier protolith such as a mixture of marinesediments and marine carbonates as indicated in Fig. 6.

CONCLUSIONS

The δ13C(CH4) values and the absence of Fe-Ni al-loys in adjacent serpentinites and additional componentsin the fluid inclusions such as hydrocarbons and H2 indi-cate that CH4 production due to biogenic processes suchas bacterial alteration and abiogenic processes such asserpentinization or the release of primordial CH4 fromthe mantle might be the unlikely sources for CH4 in thesefluid inclusions. Nonetheless, if these processes occurredthey must have been subordinate and hence otherabiogenic processes such as thermogenic processes (ther-mal decomposition of organic matter) in subducted or-ganic carbon from marine sediments must be consideredto be of primary importance in the formation of CH4-richfluids in these rocks. These data indicate that in additionto CH4 release in plumes from sediments from accretion-ary prisms in convergent margins, CH4 might also be sta-ble to at least the upper 20 km of a subduction zone wherejadeitite veins formed under low-T and high-P conditions.

Acknowledgments—Thanks to Zhang F. S., Wang L. J., ZhuH. P. and Zhang W. H. for their helps in analyzing the chemicalcompositions of the fluid inclusions. The first author thanksZhu R. X. for his encouragement. Discussions with Kaindl R.and Sadofsky S. J. are also gratefully acknowledged. The au-thors wish to thank Charlou J. L. and Scambelluri M. for theirconstructive and critical reviews which considerably improvedthe manuscript. The constructive reviews of Philippot P.,Hornibrook E. D. and an anonymous reviewer of an earlier draftof this manuscript are greatly appreciated. The editorial com-ments by the journal editor, Zheng Y.-F., were very helpful and

are also gratefully appreciated. This work was supported byNational Science Foundation of China (40221402, 40234045and 40302030).

REFERENCES

Abrajano, T. A., Sturchio, N. C., Bohlke, J. K., Lyon, G. L.,Poreda, R. J. and Stevens, C. M. (1988) Methane-hydrogengas seeps, Zambalas ophiolite, Philippines: deep, or shal-low origin? Chem. Geol. 71, 211–222.

Apps, J. A. and van de Kamp, P. C. (1993) Energy gases ofabiogenic origin in the Earth’s crust. The Future of EnergyGases (Howell, D. G., ed.), U.S. Geol. Surv. Prof. Paper,1570, 81–132.

Bakker, R. (2003) Package FLUIDS 1. computer programs foranalysis of fluid inclusion data and for modeling bulk fluidproperties. Chem. Geol. 194, 3–23.

Bender, F. (1983) Geology of Burma. Gebrüder Bornträger,Berlin, 260 pp.

Berndt, M. E., Allen, D. E. and Seyfried, W. E. (1996) Reduc-tion of CO2 during serpentinization of olivine at 300°C and500 bar. Geology 24, 315–354.

Boettcher, A. L. and Wyllie, P. J. (1969) Phase relationships inthe system NaAlSiO4-SiO2-H2O to 35 kilobars pressure. Am.J. Sci. 267, 875–909.

Bottinga, Y. (1969) Calculated fractionation factors for carbonand hydrogen isotope exchange in the system calcite-car-bon dioxide-methane-hydrogen-water vapor. Geochim.Cosmochim. Acta 33, 49–64.

Burruss, R. C. (1993) Stability and flux of methane in the deepcrust: a review. The Future of Energy Gases (Howell, D.G., ed.), U.S. Geol. Surv. Prof. Paper, 1570, 21–29.

Charlou, J. L. and Donval, J. P. (1993) Hydrothermal ventingbetween 12°N and 26°N along the Mid-Atlantic Ridge. J.Geophys. Res. 98, 9625–9642.

Charlou, J. L., Donval, J. P., Fouquet, Y., Jean-Baptiste, P. andHolm, N. (2002) Geochemistry of high H2 and CH4 ventfluids issuing from ultramafic rocks at the Rainbow hydro-thermal field (36°14′ N, MAR). Chem. Geol. 191, 345–359.

Chhibber, H. L. (1934) The Mineral Resources of Burma.Macmillan London, 320 pp.

Coleman, R. G. (1961) Jadeite deposit of the Clear Creek area,New Idria district, San Benito County, California. J. Pet-rol. 2, 209–247.

Craig, H., Horibe, Y., Farley, K. A., Welhan, J. A., Kim, K. R.and Hey, R. N. (1987) Hydrothermal vents in the MarianaTrough: results of the first ALVIN dives. Eos 68, 1531.

Deines, P. (1980) The isotopic composition of reduced organiccarbon. Handbook of Environmental Geochemistry, 239–406, Elsevier, New York.

Dobretsov, N. L. and Ponomareva, L. G. (1965) Comparativecharacteristics of jadeite and associated rocks from PolarUral and Near-Balkhash region. Academia Scientifica USSR31, 178–243.

Essene, E. J. and Fyfe, W. S. (1967) Omphacite in Californianmetamorphic rocks. Contrib. Mineral. Petrol. 15, 1–23.

Früh-Green, G. L., Scambelluri, M. and Vallis, F. (2001), O-Hisotope ratios of high pressure ultramafic rocks: implica-tions for fluid sources and mobility in the subducted hy-

514 G. Shi et al.

drous mantle. Contrib. Mineral. Petrol. 141, 145–159.Fu, B., Zheng, Y.-F., Wang, Z.-R., Xiao, Y.-L., Gong, B. and

Li, S.-G. (1999) Oxygen and hydrogen isotope geochemistryof gneisses associated with ultrahigh pressure eclogites atShuanghe in the Dabie Mountains. Contrib. Mineral. Pet-rol. 134, 52–66.

Fu, B., Touret, J. L. R. and Zheng, Y.-F. (2001) Fluid inclu-sions in coesite-bearing eclogites and jadeite quartzite atShuanghe, Dabie Shan, China. J. Metamor. Geol. 19, 529–545.

Fu, B., Zheng, Y.-F. and Touret, J. L. R. (2002) Petrological,isotopic and fluid inclusion studies of eclogites fromSujiahe, NW Dabie Shan (China). Chem. Geol. 187, 107–128.

Fu, B., Touret, J. L. R. and Zheng, Y. F. (2003a) Remnants ofpremetamorphic fluid and oxygen isotopic signatures ineclogites and garnet clinopyroxenite from the Dabie-Suluterranes, eastern China. J. Metamor. Geol. 21, 561–578.

Fu, B., Touret, J. L. R., Zheng, Y. F. and Jahn, B. M. (2003b)Fluid inclusions in granulites, granulitized eclogites, andgarnet clinopyroxenites from the Dabie-Sulu terranes, east-ern China. Lithos 70, 293–319.

Fuex, A. N. (1979) The use of stable carbon isotopes in hydro-carbon exploration. J. Geochem. Explor. 7, 155–188.

Haggerty, J. A. (1991) Evidence from fluid seeps atop serpen-tine seamounts in the Mariana Forearc—clues for emplace-ment of seamounts and their relationship to forearc tecton-ics. Mar. Geol. 102, 293–309.

Hall, D. L. and Bodnar, R. J. (1989) Comparison of fluid inclu-sion decrepitation and acoustic emission profiles of West-erly granite and Sioux quartzite. Tectonophysics 168, 283–296.

Harlow, G. E. (1986) Jadeitites and their fluid inclusions fromRio Motagua, Guatemala (abstract). International Mineral.Assoc. General Meeting, 14th, Mineral. Soc. Am. A119.

Harlow, G. E. (1994) Jadeitites, albitites and related rocks fromthe Motagua fault zone, Guatemala. J. Metamor. Geol., 12,49–68.

Harlow, G. E. and Sorensen, S. S. (2001) Jade: Occurrence andmetasomatic origin. Aust. Gemm. 21, 7–10.

Harlow, G. E. and Sorensen, S. S. (2005) Jade (nephrite andjadeitite) and serpentinite: metasomatic connections. Int.Geol. Rev. 47, 113–146.

Holloway, J. R. (1984) Graphite-CH4-H2O-CO2 equilibria atlow-grade metamorphic conditions. Geology 12, 455–458.

Horita, J. and Berndt, M. E. (1999) Abiogenic methane forma-tion and isotopic fractionation under hydrothermal condi-tions. Science 285, 1055–1057.

Huizenga, J. M. (2001) Thermodynamic modelling of C-O-Hfluids. Lithos 55, 101–114.

Hurai, V., Janák, M., Ludhová, L., Horn, E. E., Thomas, R. andMajzlan, J. (2000) Nitrogen-bearing fluids, brines and car-bonate liquids in Variscan migmatites of the Tatra Moun-tains, Western Carpathians—heritage of high-pressure meta-morphism. Eur. J. Mineral. 12, 1283–1300.

Iwao, S. (1953) Albitite and associated jadeite rock from KotakiDistrict, Japan: a study in ceramic raw material. Report ofGeological Survey of Japan, 153 pp.

Janak, M., Hurai, V., Ludhova, L., O’Brien, P. J. and Horn, E.

E. (1999) Dehydration melting and devolatilization duringexhumation of high-grade metapelites: the Tatra Mountains,Western Carpathians. J. Metamor. Geol. 17, 379–395.

Johnson, C. A. and Harlow, G. E. (1999) Guatemala jadeititesand albitites were formed by deuterium-rich serpentinizingfluids deep within a subduction-channel. Geology 27, 629–632.

Kastner, M., Kvenvolden, K. A. and Lorenson, T. D. (1998)Chemistry, isotopic composition, and origin of a methane-hydrogen sulfide hydrate at the Cascadia subduction zone.Earth Planet. Sci. Lett. 156, 173–183.

Kelley, D. S. (1996) Methane-rich fluids in the oceanic crust.J. Geophys. Res. 101, 2943–2962.

Kelley, D. S. and Früh-Green, G. L. (1999) Abiogenic methanein deep-seated mid-ocean ridge environments: insights fromstable isotope analyses. J. Geophys. Res. 104, 10439–10460.

Kessen, S. E. and Ringwood, A. E. (1989) Slab-mantle interac-tions: II. The formation of diamonds. Chem. Geol. 78, 97–118.

Klemd, R., Kerkhof van den, A. M. and Horn, E. E. (1992)High-density CO2-N2 inclusions in eclogite-faciesmetasediments of the Münchberg gneiss complex, SE-Ger-many. Contrib. Mineral. Petrol. 111, 409–419.

Klemd, R., Bröcker, M. and Schramm, J. (1995) Characteriza-tion of amphibolite-facies fluids of Variscan eclogites fromthe Orlica-Snieznik dome (Sudetes, SW Poland). Chem.Geol. 119, 101–113.

Kobayashi, S., Miyake, H. and Shoji, T. (1987) A jadeite rockfrom Oosa-cho, Okayma Prefecture, Southwestern Japan.Mineral. J. 13, 310–327.

Küster, M. and Stöckhert, B. (1997) Density changes of fluidinclusions in high pressure low temperature metamorphicrock from Crete: A thermobarometric approach based onthe creep strength of the host minerals. Lithos 41, 151–167.

Li, X.-P., Zheng, Y.-F., Wu, Y.-B., Chen, F.-K., Gong, B. andLi, Y.-L. (2004) Low-T eclogite in the Dabie terrane ofChina: Petrological and isotopic constraints on fluid activ-ity and radiometric dating. Contrib. Mineral. Petrol. 148,443–470.

Li, Y.-L., Zheng, Y.-F., Gong, B. and Fu, B. (2000) Carbon iso-topes in eclogite and apatite separate from Huangzhen andShima in SE Dabie. Sci. China (D) 43, 449–457.

Marty, B. and Tolstikhin, I. N. (1998) CO2 fluxes from Mid-Ocean Ridges, Arcs and Plumes. Chem. Geol. 145, 233–248.

McBirney, A. R., Aoki, K. I. and Bass, M. (1967) Eclogite andjadeite from the Motagua Fault zone, Guatemala. Am. Min-eral. 52, 908–918.

McCollom, T. M. and Seewald, J. S. (2001) A reassessment ofthe potential for reduction of dissolved CO2 to hydrocar-bons during the serpentinization of olivine. Geochim.Cosmochim. Acta 65, 3769–3778.

Miyazaki, K., Sopaheluwakan, J., Zulkarnain, I. and Wakita,K. (1998) A jadeite-quartz-glaucophane rock fromKarangsambung, Central Java, Indonesia. Island Arc 7, 223–230.

Morkovkino, V. F. (1960) Jadeitites in the hyperbasites of thePolar Urals. Izvestia Akademia Nauk USSR, SeriesGeologia, 4.

Fluid inclusions in Myanmar jadeitite 515

Nadeau, S., Philippot, P. and Pineau, F. (1993) Fluid inclusionsand mineral isotopic compositions (H-C-O) in eclogiticrocks as tracers of local fluid migration during high-pres-sure metamorphism. Earth Planet. Sci. Lett. 114, 431–448.

Okay, A. I. (1997) Jadeite-K-feldspar rocks and jadeitites fromnorthwest Turkey. Mineral. Mag. 61, 835–843.

Philippot, P. and Selverstone, J. (1992) Trace element-richbrines in eclogitic veins: implications for fluid composi-tion and transport during subduction. Contrib. Mineral.Petrol. 106, 417–430.

Radvanec, M., Banno, S. and Ernst, W. G. (1998) Chemicalmicrostructure of jadeite from Pacheco Pass, California. Am.Mineral. 83, 273–279.

Röller, K., Küster, M. and Stöckhert, B., 2001. Quadrupol-mass-spectrometeric (QMS) analysis of fluid inclusions-applica-tion and perspectives. XVI ECROFI European Current Re-search On Fluid Inclusions (Porto) 7, 383–386.

Rona, A., Bougault, H., Charlou, J. L., Appriou, P., Nelsen, T.A., Trefry, J. H., Eberhard, G. L., Barone, A. and Needham,H. D. (1992) Hydrothermal circulation, serpentinization, anddegassing at a rift valley-fracture zone intersection: Mid-Atlantic Ridge near 15°N, 45°W. Geology 20, 783–786.

Rumble, D. and Yui, T.-F. (1998) The Qinglongshan oxygenand hydrogen isotope anomaly near Donghai in JiangsuProvince, China. Geochim. Cosmochim. Acta 62, 3307–3321.

Sadofsky, S. J. and Bebout, G. E. (2001) Paleohydrogeology at5- to 50-kilometer Depths of accretionary Prisms: TheFranciscan Complex, California. Geophys. Res. Lett. 28,2309–2312.

Sadofsky, S. J. and Bebout, G. E. (2003) Record of forearcdevolatilization in low-T, high-P/T metasedimentary suites:Significance for models of convergent margin chemicalcycling. Geochem. Geophys. Geosyst . 4(4), 9003,doi:10.1029/2002GC000412.

Sasada, M., Sawaki, T. and Takeno, N. (1992) Analysis of fluidinclusion gases from geothermal system, using a rapid-scan-ning quadrupole mass spectrometer. Euro. J. Mineral. 4,895–906.

Scambelluri, M. and Philippot, P. (2001) Deep fluids in sub-duction zones. Lithos 55, 213–227.

Schertl, H. P., Schreyer, W. and Chopin, C. (1991) The pyrope-coesite rocks and their country rocks at Parigi, Dora MairaMassif, Western Alps: detailed petrography, mineral chem-istry and PT-path. Contrib. Mineral. Petrol. 108, 1–21.

Schoell, M. (1988) Multiple origins of methane in the Earth.Chem. Geol. 71, 1–10.

Selverstone, J., Franz, G. and Thomas, S. (1992) Fluidheterogeneities in 2 GPa eclogites: implications for elementrecycling during subduction. Contrib. Mineral. Petrol. 112,341–357.

Shaw, A. M., Hilton, D. R., Fischer, T. P., Walker, J. andAlvarado, G. E. (2003) Contrasting He-C relationships inNicaragua and Costa Rica: insights into C cycling throughsubduction zones. Earth Planet. Sci. Lett. 214, 499–513.

Sherwood Lollar, B., Frape, S. K., Weise, M., Fritz, P., Macko,S. A. and Welhan, J. A. (1983) Abiogenic methanogenesisin crystalline rocks. Geochim. Cosmochim. Acta 57, 5087–5097.

Sherwood Lollar, B., Westgate, T. D., Ward, J. A., Slater, G. F.and Lacrampe-Couloume, G. (2002) Abiogenic formationof alkanes in the Earth’s crust as a minor source for globalhydrocarbon reservoirs. Nature 416, 522–524.

Shi, G. H., Cui, W. Y., Wang, C. Q. and Zhang, W. H. (2000)The fluid inclusions in jadeitite from Pharkant area,Myanmar. Chinese Sci. Bull. 45, 1896–1900.

Shi, G. H., Cui, W. Y., Liu, J. and Yu, H. X. (2001) The petrol-ogy of jadeite-bearing serpentinized peridotite and its coun-try rocks from Northwestern Myanmar (Burma). Acta Pet-rol. Sinica 17, 483–490.

Shi, G. H., Cui, W. Y., Tropper, P., Wang, C. Q., Shu, G. M. andYu, H. X. (2003) The petrology of a complex sodic andsodic-calcic amphibole association and its implications forthe metasomatic processes in the jadeitite area in northwest-ern Myanmar, formerly Burma. Contrib. Mineral. Petrol.145, 355–367.

Sorensen, S. S. and Barton, M. D. (1987) Metasomatism andpartial melting in a subduction complex: Catalina schist,southern California. Geology 15, 115–118.

Sorensen, S. S. and Harlow, G. E. (1998) Acathodoluminescence (CL)-guided ion and electron micro-probe tour of jadeitite chemistry and petrogenesis (abstract).Annual Meeting, Geol. Soc. Am. 30th, A60.

Sorensen, S. S. and Harlow, G. E. (1999) The geochemical evo-lution of jadeitite-depositing fluids (abstract). Annual Meet-ing, Geol. Soc. Am. 31th, A101.

Sorensen, S. S. and Harlow, G. E. (2001) The jadeitites ofNansibon, Myanmar: Records of the geochemistry of sub-duction zone fluids (abstract). M. Goldschmidt Conference11th, A3800.

Sterner, S. M. and Bodnar, R. J. (1989) Synthetic fluid inclu-sions: VII. Re-equilibration of fluid inclusions in quartzduring laboratory-simulated burial and uplift. J. Metamor-phic. Geol., 7, 243–260.

Sturchio, N. C., Abrajano, T. A., Murowchik, J. B. andMuehlenbachs, K. (1989) Serpentinization of the Acojemassif, Zambales ophiolite, Philippines: hydrogen and oxy-gen isotope geochemistry. Tectonophysics 168, 101–107.

Su, L., Song, S. G. and Wang, Z. H. (1999) CH4-richfluidinclusions in the Yushigou mantle peridotite and theirimplcations, North Qilian Mountains, China. Chinese Sci.Bull. 44, 1992–1995.

Takayama, M. (1986) Mode of occurrence and significance ofjadeite in the Kamuikotan metamorphic rocks, Hokkaido,Japan. J. Metamor. Geol. 4, 445–454.

Tsunogai, U., Ishibashi, J., Wakita, H. and Gamo, T. (1998)Methane-rich plumes in the Suruga Trough (Japan) and theircarbon isotopic characterization. Earth Planet. Sci. Lett.160, 97–105.

Watanabe, A., Kimura, M., Kasuya, M., Kotake, M. and Katoh,T. (1994) Methane in groundwater used for Japanese agri-culture: its relationship to other physico-chemical proper-ties and possible tropospheric source strength. Geophys. Res.Lett. 21, 41–44.

Welhan, J. A. and Craig, H. (1983) Methane, hydrogen and he-lium in hydrothermal fluids of 21°N on the East Pacific Rise.Hydrothermal Processes at Seafloor Spreading Centers(Rona, A. et al., eds.), 391–409, Plenum, New York.

516 G. Shi et al.

Whiticar, M. J. (1999) Carbon and hydrogen isotope systemat-ics of bacterial formation and oxidation of methane. Chem.Geol. 161, 291–314.

Wiese, K. and Kvenvolden, K. A. (1993) Introduction to mi-crobial and thermal methane. The Future of Energy Gases(Howell, D. G., ed.), U.S. Geol. Surv. Prof. Paper, 1570,13–20.

Xiao, Y., Hoefs, J., van den Kerkhof, A. M., Fiebig, J. andZheng, Y. (2000) Fluid history of UHP metamorphism inDabie Shan, China: a fluid inclusion and oxygen isotopestudy on the coesite-bearing eclogite from Bixiling. Contrib.Mineral. Petrol. 139, 1–16.

Xiao, Y.-L., Hoefs, J., van den Kerkhof, A. M., Simon, K.,Fiebig, J. and Zheng, Y.-F. (2002) Fluid evolution duringHP and UHP metamorphism in Dabie Shan, China: Con-straints from mineral chemistry, fluid inclusions and stableisotopes. J. Petrol. 43, 1505–1527.

Yang, X. Y., Zheng, Y. F., Liu, D. and Dai, J. (2001) Chemicaland carbon isotope compositions of fluid inclusions inperidotite xenoliths and eclogites from eastern China:Geodynamic implications. Phys. Chem. Earth (A) 26, 705–718.

Zheng, Y.-F., Fu, B., Li, Y.-L., Xiao, Y.-L. and Li, S.-G. (1998)Oxygen and hydrogen isotope geochemistry of ultrahighpressure eclogites from the Dabie Mountains and the Suluterrane. Earth Planet. Sci. Lett. 155, 113–129.

Zheng, Y.-F., Fu, B., Xiao, Y.-L., Li, Y.-L. and Gong, B. (1999)Hydrogen and oxygen isotope evidence for fluid(rock in-teractions in the stages of pre- and post-UHP metamorphismin the Dabie Mountains. Lithos 46, 677–693.

Zheng, Y. F., Gong, B., Li, Y. L., Wang, Z. R. and Fu, B. (2000)Carbon concentrations and isotopic ratios of eclogites fromthe Dabie and Sulu terranes in China. Chem. Geol. 168, 291–305.

Zheng, Y. F., Gong, B., Zhao, Z. F., Fu, B. and Li, Y. L. (2003a)Two types of gneisses associated with eclogite at Shuanghein the Dabie terrane: carbon isotope, zircon U-Pb datingand oxygen isotope. Lithos, 70, 321–343.

Zheng, Y. F., Fu, B., Gong, B. and Li, L. (2003b) Stable iso-tope geochemistry of ultrahigh pressure metamorphic rocksfrom the Dabie-Sulu orogen in China: implications forgeodynamics and fluid regime. Earth Sci. Rev. 62, 105–161.

Zhu, H. P. (1999) The determinations of F and Cl from rocksby chromatography. Mineral and Resource Geology 13,317–319 (in Chinese).