| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Petrology | Volume 44 | Number 6 | Pages 1097-1120 | 2003
© Oxford University Press 2003
Mid-crustal Metasomatic Reaction Veins ina Spinel Peridotite
1 INSTITUT FÜR MINERALOGIE, PETROLOGIE UND GEOCHEMIE, EBERHARD-KARLS-UNIVERSITÄt, WILHELMSTRASSE 56, D-72074 TÜBINGEN, GERMANY
2 MINERALOGISCH–PETROGRAPHISCHES INSTITUT, BERNOULLISTRASSE 30, CH-4056 BASEL, SWITZERLAND
E-mail: markl{at}uni-tuebingen.de
RECEIVED NOVEMBER 28, 2001; ACCEPTED DECEMBER 23, 2002
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONREGIONAL GEOLOGY AND SAMPLE...PETROGRAPHYMINERAL AND ISOTOPE GEOCHEMISTRYTHERMOMETRY AND MINERAL...DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
In Central Dronning Maud Land, East Antarctica, rare metre-sized lenses of spinel peridotite are enclosed in high-grade metamorphic rocks. The rocks experienced a medium-P granulite-facies metamorphism at
575 Ma and a low-P amphibolite-facies overprint at 530 Ma. The latter is probably related to extensive granitoid magmatism between 530 and 500 Ma, which produced large volumes (about half of the outcrops today) of granitic to syenitic rocks as well as abundant K-feldspar–quartz pegmatites. One of the spinel peridotite lenses in the Schirmacher Oasis of Central Dronning Maud Land is crosscut by several small (up to 10 cm wide) veins with a characteristic zoned sequence of mineral assemblages, which was formed by reaction of a hydrous, SiO2-saturated fluid or pegmatitic melt with the peridotite. The zoned sequence consists of the following mineral assemblages (from the centre of the vein towards the outer margin): zone 0, plagioclase + quartz; zone 1, green biotite intergrown with zircon + clinoamphibole; zone 2, cummingtonite + dark brown biotite intergrown with rutile + clinoamphibole; zone 3, cummingtonite + light brown biotite + spinel; zone 4, olivine + orthopyroxene + spinel ± clinopyroxene (unaltered peridotite). This sequence was investigated with respect to its conditions of formation, modal mineralogy, mineral chemistry, fluid inclusions, and oxygen and hydrogen isotope compositions of selected minerals. Based on the stability of cummingtonite and on equilibrium calculations in the MgO–SiO2–H2O system and on quartz–biotite oxygen isotope thermometry, the reaction vein formed at 650°C, which is in accord with typical pegmatite crystallization temperatures. The pegmatite of zone 0 is interpreted to have formed in an open fissure whereas, on textural grounds, zone 3 replaces former peridotite. On the basis of mass balance constraints, the boundary between zones 1 and 2 is interpreted to approximately represent the former boundary between peridotite and the open fissure before reaction. Oxygen isotope systematics show that the infiltrating fluid had an isotopic composition of 9–10
SMOW. All minerals of the reaction vein with the exception of the inherited spinel and olivine in the adjacent peridotite are in equilibrium with such a fluid. Spinel in the peridotite is depleted in 18O compared with coexisting olivine, which suggests isotopic disequilibrium. Spinel in zone 3 has a distinctly different isotopic composition compared with that in the peridotite, apparently approaching but not reaching equilibrium. The combination of mineral chemistry and mass balance constraints of the modal mineralogy constrains the volume change during metasomatism and the direction of elemental diffusion. It is indicated that Mg, Cr and Ni always diffused towards the vein, whereas Si, Al, K, Na, H2O and possibly Fe diffused into the peridotite. KEY WORDS: peridotite; metasomatism; pegmatite; diffusion; reaction
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONREGIONAL GEOLOGY AND SAMPLE...PETROGRAPHYMINERAL AND ISOTOPE GEOCHEMISTRYTHERMOMETRY AND MINERAL...DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
When mantle fragments are tectonically emplaced into crustal rocks, two distinctly different chemical systems are juxtaposed: Mg-rich mantle against Si-rich crustal rocks in the most simplified way (e.g. Bucher-Nurminen, 1988, 1990). This juxtaposition typically results in the formation of hydrated ‘black wall assemblages’, which are controlled by addition of hydrous fluids along the tectonic contact and by diffusive processes driven by chemical potential gradients (e.g. Korzhinskii, 1970). Such fluids may also penetrate peridotites along cracks or fissures, resulting in the formation of metasomatic reaction veins with characteristically zoned mineral assemblages perpendicular to the rock–fissure boundary.
Carswell et al. (1974) investigated this type of crust–mantle interaction at temperatures of 700°C in a Norwegian dunite body, where a zoned sequence of olivine 
enstatite anthophyllite tremolite chlorite developed in a vein crosscutting the dunite. Carswell et al. (1974) invoked a fluid in equilibrium with silica-saturated country rock and concluded that the zoned sequence developed as a result of different diffusivity of the components entering the dunite. Qualitatively, Mg was the only component diffusing towards the vein centre, whereas Al, Ca, H2O and Si (in order of increasing mobility) moved away from it. Carswell et al. (1974) were, however, not able to quantify this mobility and did not address the problem of where the original boundary between dunite and open fissure in their specific zoned sequence was. Evans & Trommsdorff (1974) and Pfeifer (1977) quantified various aspects of metamorphism and fluid–rock interaction in ultramafic rocks, but as far as we are aware the work by Carswell et al. (1974) is the only one dealing in more than a purely descriptive way with this vein-like type of metasomatic alteration in ultramafic rocks.
The concept of local equilibrium combined with irreversible thermodynamics was introduced and applied to a variety of rock types and textures by Fisher (1973, 1975, 1977), as well as Joesten (1977), Foster (1983), Ashworth & Sheplev (1997), Markl et al. (1998) and most recently by Abart et al. (2001). However, estimates of absolute or relative diffusivities critically depend on the assemblages in metasomatic reaction sequences. Most of these reaction sequences are under-determined, i.e. they contain too few phases to constrain the chemical potentials of all diffusing species. In this case, only qualitative constraints can be set on how a specific metasomatic sequence formed. Furthermore, the question arises as to where the original boundary between two minerals, two rocks, or between fissure and rock was before the reaction commenced. This question opens the discussion about volume changes and mass transfer during metasomatism and is best addressed by investigating well-constrained geometries such as veins. The aim of the present study is to present a detailed chemical, stable isotope and petrological analysis of metasomatic reaction veins in peridotites and to constrain the chemical and physical parameters during their formation as quantitatively as possible.
|
REGIONAL GEOLOGY AND SAMPLE LOCALITY |
|---|
|
TOPABSTRACTINTRODUCTIONREGIONAL GEOLOGY AND SAMPLE...PETROGRAPHYMINERAL AND ISOTOPE GEOCHEMISTRYTHERMOMETRY AND MINERAL...DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
The area of the present study lies within the East Antarctic craton and has experienced at least two major regional tectonometamorphic events: the
1·1 Ga Grenvillian orogeny and the younger Panafrican or Ross orogeny (500–600 Ma; Jacobs et al., 1998). The Panafrican event is evidenced throughout the area, overprinting structures and metamorphic textures of the Grenvillian orogeny more or less completely (Jacobs et al., 1998). Following the arguments of Dalziel (1991) and Piazolo & Markl (1999), the area that is now Dronning Maud Land would have been the continuation of the Panafrican belt of Laurentia found in Africa (Mozambique Belt), the southern tip of India, in Sri Lanka and Madagascar.
According to the results of Ravich & Solov'ev (1966) and of the German GeoMaud expedition in 1995–1996, rock units in Central Dronning Maud Land (CDML) comprise large areas of mainly undeformed acid and less commonly basic intrusive rocks including anorthosites, gabbros and syenites. Additional rocks encountered in CDML are a metapelitic supracrustal series including Grt–Sil–Bt schists, a likely bimodal metavolcanic sequence, migmatites, Fo–Spl marbles, calcsilicates and rare mantle fragments tectonically emplaced into the metasedimentary rocks. These mantle fragments comprise harzburgites and spinel peridotites and they occur in the Schirmacher Oasis, at the ‘AI rocks’ north of the Humboldt Mountains and in the Dallmann Mountains (Fig. 1). The samples investigated here come from a 5 m x 2 m boulder of spinel peridotite in the northern part of the Schirmacher Oasis, 3 km north of the Indian Antarctic Station. This boulder occurs as several large fractured pieces weathered out from metasedimentary garnet-bearing gneisses and schists (Fig. 2a). It shows a distinct primary layering with black layers rich in chromite, green layers rich in Cr-bearing clinopyroxene and brownish layers rich in olivine and orthopyroxene with minor evenly distributed chromite. The spacing of this layering is irregular, but chromite and clinopyroxene layers generally are rare and do not exceed 1 cm in width. Macroscopically, there are no signs of deformation or metamorphism apart from veins of 1–10 cm width and up to 1 m length (Fig. 2b), which comprise a mineral sequence zoned parallel to their orientation. The veins are not parallel to the primary layering, but generally crosscut the more abundant olivine–orthopyroxene–chromite rock.
|
|
|
PETROGRAPHY |
|---|
|
TOPABSTRACTINTRODUCTIONREGIONAL GEOLOGY AND SAMPLE...PETROGRAPHYMINERAL AND ISOTOPE GEOCHEMISTRYTHERMOMETRY AND MINERAL...DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
The reaction veins consist of the following zoned sequence of mineral assemblages, where zone 0 is a pegmatite in the centre of a vein and zone 4 is unaltered peridotite:
- zone 0, plagioclase + quartz (pegmatitic);
- zone 1, green biotite intergrown with zircon + clinoamphibole;
- zone 2, cummingtonite + dark brown biotite intergrown with rutile + clinoamphibole;
- zone 3, cummingtonite + light brown biotite + spinel (chromite);
- zone 4, olivine + orthopyroxene + spinel (chromite) ± clinopyroxene.
- zone 1, green biotite intergrown with zircon + clinoamphibole;
In spite of the ordered sequence, the veins exhibit variations in thickness and not all of the zones are developed in every sample (Fig. 3). Zone 0 is present only in the largest veins and some small veins may consist of zone 3 only. However, the order of zones in the sequence is never changed. If one or more zones are missing, then zone(s) 0, 1 and, in some cases, also 2 are always the innermost. Some of the larger veins have small ‘satellite veins’ (Fig. 3), which probably represent thin fractures where only little fluid could enter the rock. Zone 0 and zone 2 are coarse grained with biotite flakes of up to 3 cm in some instances. The other zones are medium to fine grained and typically show prismatic amphibole crystals oriented perpendicular to the direction of the vein.
|
Based on their undisturbed geometry, two samples (GM 587 and GM 588) were chosen for detailed studies of the veins. Sample GM 588 shows continuous and sub-parallel reaction zones, but lacks zone 0 (Fig. 3a), which was therefore sampled from GM 587 (Fig. 3b). Sample GM 587 comes from the same vein as GM 588 but
0·5 m away from the latter. Here, the vein is slightly wider, possibly indicating a larger opening of the original fissure. Two thin sections covering different parts of the zoned sequence were prepared from sample GM 588 (Fig. 4). The complete sequence of zones 1–4 is present in thin section GM 588-1 and a traverse was analysed by microprobe. Microscopically, zone 4 consists of an equilibrium texture of olivine, orthopyroxene and spinel. Clinopyroxene was absent in both thin sections. The spinel is brownish in plane-polarized light, indicating a high chromium content. Towards zone 3, olivine and orthopyroxene are replaced by cummingtonite, but spinel remains almost unchanged both in appearance and in modal abundance (Fig. 5, Table 1). Small rims of reddish biotite around spinel grains imply, however, that some reaction has occurred. The needle-like cummingtonite of zone 3 looks identical to the ortho- and clinoamphiboles of zones 2 and 1 but the clinoamphiboles are greenish in zone 3. Zone 2 is characterized by large, reddish brown flakes of biotite with abundant sagenite-like intergrowths of rutile needles. In this zone, zircon is rare or absent. The brown biotite changes into green biotite at the margin of zone 1. Both generations and generally all minerals in the reaction sequence appear to be in equilibrium. For a large biotite flake, the outer (i.e. directed towards the peridotite) brown end of the grain gradually changes colour towards the inner green end. Further towards the centre of the vein, biotite and amphibole are very fine grained. The green biotite contains abundant inclusions of zircon, but no rutile. The innermost zone 0 consists of milky quartz with abundant fluid inclusions with minor, but variable amounts of sodic plagioclase. Representative mineral textures from sample GM 588 are depicted and related to their position in the zonal sequence in Fig. 6.
|
|
|
|
|
MINERAL AND ISOTOPE GEOCHEMISTRY |
|---|
|
TOPABSTRACTINTRODUCTIONREGIONAL GEOLOGY AND SAMPLE...PETROGRAPHYMINERAL AND ISOTOPE GEOCHEMISTRYTHERMOMETRY AND MINERAL...DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Analytical techniques
Minerals were analysed with a JEOL JX8900 electron microprobe at the Institut für Mineralogie, Petrologie und Geochemie at the University of Tübingen. Both natural and synthetic standards were used for calibration. Counting times were 20 s on the peak and 10 s on the background for most elements; F, Cl, Cr and Ni were counted for 30–60 s on the peak. The emission current was 20 nA, and the acceleration voltage was 15 kV. The raw data were corrected using the CITZAF-correction (Armstrong, 1991). To avoid errors caused by evaporation of Na, this element was measured first. Feldspar was generally measured with a slightly defocused beam of
5 µm width.
The oxygen isotope compositions of hand-picked mineral separates were measured at the University of Tübingen, using a method similar to that described by Sharp (1990) and Rumble & Hoerig (1994). Between 0·5 and 2 mg of sample was loaded onto a small Pt sample holder and pumped out to a vacuum of about 10-6 mbar. After prefluorination of the sample chamber overnight, the samples were heated with a CO2-laser in 50 mbar of pure F2. The extracted O2 was collected on a molecular sieve (13X) and subsequently expanded and analysed on an isotope ratio mass spectrometer (Finnigan MAT 252). Oxygen isotope compositions are given in the standard 
-notation, expressed relative to VSMOW in per mil (). Replicate oxygen isotope analyses of the standards used NBS-28 quartz (n = 2) and UWG-2 garnet (n = 6) (Valley et al., 1995) and had an average precision of ±0·07 for 18O. The accuracy of 18O values was better than 0·2 compared with accepted 18O values for NBS-28 of 9·64 and UWG-2 of 5·8.
Hydrogen isotope measurements were performed on hand-picked mineral separates using the method of Vennemann & O'Neil (1993). Hydrogen isotope compositions are given in the standard -notation, expressed relative to VSMOW in per mil (). Values are reported relative to a NBS-30 (Biotite) value of -65 with a precision of ±2.
Mineral compositions
The average mineral compositions used in the calculations are given in Table 2. In the peridotite of zone 4, olivine composition varies from Fo80 to Fo83 with NiO contents between 0·34 and 0·45 wt%, MnO contents around 0·1 wt % and negligible Ti and Ca contents. Orthopyroxene composition is relatively constant around En80Wo1 with 0·1–0·3 wt % Cr2O3, 0·1 wt % NiO, 0·3–0·5 wt % CaO, 0·1–0·15 wt % MnO, 0·05–0·15 wt % TiO2 and 1–2 wt % Al2O3. Spinel has compositions around Chr45–50Spl25Herc25–30 in zone 3 and Chr35–40Spl40Herc20–25 in zone 4 (Fig. 7).
|
|
Throughout the zoned sequence, the compositions of orthoamphibole, clinoamphibole and biotite generally change considerably (Fig. 8). The orthoamphiboles all fall into the cummingtonite field of Leake et al. (1997) and Ghiorso & Evans (2002), although some contain up to 0·5 Al per formula unit (p.f.u.). XFe [here and in the following defined as FeO/(FeO + MgO)] varies between 0·29 and 0·24; the orthoamphiboles contain <0·1 F p.f.u., no measurable Cl, and negligible Ni, Ti and Cr. Clinoamphiboles range in composition between actinolite and tschermakite; however, most of the analyses plot in the magnesiohornblende field of Leake et al. (1997). Their XFe varies between 0·17 and 0·26, Ti is <0·03 atoms p.f.u., and Ni and Cr are negligible. In contrast to orthoamphibole, the clinoamphibole contains
0·03–0·08 Cl and 0·13–0·2 F p.f.u. Biotite is a phlogopite with XFe ranging from 0·15 to 0·28, Ti from 0·08 to 0·14, Cl from 0·04 to 0·06, F from 0·35 to 0·5, and Al from 1·05 to 1·2 p.f.u. Cr and Ni are close to or below 0·01 p.f.u. As expected, F shows strong negative correlation with XFe and positive correlation with Al, whereas Cl shows the opposite behaviour (e.g. Volfinger et al., 1985; Morrison, 1991; Markl & Piazolo, 1998).
|
Change of mineral compositions in the zoned sequence
A quantitative line of measurements through the whole zoned sequence in thin section GM 588-1 revealed that not only mineral assemblages but also mineral compositions change systematically from vein centre to vein margin. This is depicted in Fig. 8a–c for biotite, clinoamphibole and orthoamphibole, and it is related to the position in the thin section shown in Fig. 5.
Biotite composition (Fig. 8a) changes continuously in most of the profile. From vein centre outwards, biotite shows slightly increasing Si, Ni, F and especially Cr and—after some increase in the first 8 mm—decreasing Al, Na, XFe and Cl. The change in slope is at the border between zones 1 and 2. Interestingly, Ti first decreases and after 8 mm increases steadily. The reason for this feature is unknown. Hence, the components that are abundant in the peridotite, namely Mg and Cr, increase towards it. The increase of Si and F is most probably related to crystal chemical effects such as the ones decribed by Volfinger et al. (1985).
Cummingtonite (Fig. 8b) is not present in zone 1 and hence begins not before 8 mm. It is a phase that incorporates less minor and trace elements than biotite. Accordingly, only XFe records significant changes from vein centre to margin, the latter being more Mg rich. Cr and Ni show a very slight increase from internal to external zones, Ca is almost constant with the exception of the very first cummingtonites at 8 mm, but all other elements remain constant or change in an irregular manner.
Olivine, illustrated in Fig. 8c, shows a slight increase in fayalite content of 2% from the contact with zone 3 towards the unaltered peridotite, but at the outermost margin of the thin section, the XFe is the same again as at the boundary with zone 3.
Stable isotope compositions of minerals
The variation in oxygen and hydrogen isotope compositions of hand-picked minerals was measured across the complete section in sample GM 588-1 (Table 3), to match the spatial analyses of the mineral chemistry. In addition, quartz and plagioclase from zone 0 and biotite and amphibole from the margin between zone 0 and 1 were measured in hand-picked mineral grains from sample GM 587. For comparison, the O-isotope compositions of pyroxene and olivine were also measured in sample GM 590, which is completely devoid of any reaction veins and hence, may represent the least altered peridotitic material. The results of the isotope measurements are reported in Table 3 and illustrated in Fig. 9.
|
|
Quartz from zone 0 has a
18O value of 10·4, whereas coexisting plagioclase, biotite and amphibole have values of 7·6, 7·1 and 6·9, respectively. In general, the 18O values of amphibole and biotite in the various reaction zones are uniform at 7 for both, with the exception of amphibole in zone 1, which has a 18O value of 6·0. Pyroxene in zone 4 has values between 5·9 and 6·9, the latter being similar to those for pyroxene from sample GM 590. Olivine from zone 4 has 18O values between 5·6 and 6·2, and higher values are obtained closer to the vein. Measurements from the macroscopically unaltered sample GM 590 show values of 4·8–5·7 for olivine and 6·9–7·2 for orthopyroxene. Spinel, finally, shows very low 18O values of 2·8 in the unaltered peridotite, whereas in zone 3 it attained a value of 4·7. The hydrogen isotope composition of biotite varies between -56 and -48 in zone 1 and between -54 and -38 in zone 2. Amphibole from the same zones has values of -103 and -120, respectively (Table 3).
Compositions of the zones 0–4 in the reaction vein
In principle, there are three ways to determine the chemical composition of the various reaction zones: (1) to separate the zones from each other and then to analyse the composition by X-ray fluorescence; (2) to point-count the mode using a polarization microscope and to combine this mode with average mineral compositional data from each zone; (3) to perform line measurements by microprobe for a statistically relevant number of grains and to calculate the mode and the average composition of each zone from these measurements. We chose to use the third method, because the first method is imprecise owing to the curved boundaries between the zones and the second method is difficult to perform in very fine-grained rocks. In the coarser-grained portions (zones 3 and 4), point-counting using an optical microscope has also been performed and has been used to control the results of method 3. Both methods agreed within 10% relative. The average mole fractions and the compositions of the various zones were then calculated by adding all microprobe analyses weighted by the specific molar volumes of the respective phases and then dividing these sums by the number of analyses. A total of three lines with 1000 point measurements in sample GM 588 were used. The molar volumes used are from Berman (1988) and they are reported in Table 1. The resulting modes and chemical compositions of the various zones and of the average mineral compositions in each zone are reported in Tables 1, 2 and 4, and the chemical compositions of the zones are shown in Fig. 10. Table 4 reports the zone compositions both as wt % and as moles in 1000 cm3 rock.
|
|
|
THERMOMETRY AND MINERAL EQUILIBRIA |
|---|
|
TOPABSTRACTINTRODUCTIONREGIONAL GEOLOGY AND SAMPLE...PETROGRAPHYMINERAL AND ISOTOPE GEOCHEMISTRYTHERMOMETRY AND MINERAL...DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Temperature
As shown in Fig. 11, equilibria in the CMSH system are useful indicators of metamorphic temperatures, but do not serve to constrain pressure. Figure 11 was constructed using the GEOCALC software of Lieberman & Petrakakis (1991) with the thermodynamic database of Berman (1988). Activities of the phases were corrected using ideal site mixing for talc, the formalism of Will & Powell (1992) for tremolite, unit activity for antigorite and the MELTS supplemental calculator (which is available at http://gneiss.geology.washington. edu/~ghiorso/MeltsCALC/) for all other phases. H2O activity was varied in Fig. 11a and b to test its effect on the equilibria. The stability of cummingtonite constrains the temperature to
550–680°C at 4 kbar. At lower H2O activities or higher pressures, this bracket is further reduced. The equilibria depicted in Fig. 11 place an upper pressure limit at 8–10 kbar.
|
Oxygen isotope fractionations between quartz and plagioclase (An60) of
2·8 in zone 0 correspond to equilibration temperatures of 470°C, whereas quartz–biotite and quartz–amphibole fractionations in this zone correspond to temperatures of 650 and 600°C, respectively, using the calibrations of Zheng (1993a, 1993b). These temperatures are in fair to good agreement with temperature estimates based on phase equilibria and on the stability of cummingtonite as detailed above. It is of interest to note, however, that the isotopic compositions of biotite and amphibole are reversed in that biotite generally has higher 18O content compared with coexisting amphibole. This may be related to compositional effects, in particular the F-rich nature of the biotite (e.g. Zheng, 1993a; Fortier et al., 1994) and/or to an analytical artefact in that it was difficult to clearly separate amphibole from pyroxene. Original 18O values of pyroxene are expected to be slightly lower than the amphibole values, given a mantle-derived peridotite origin. That the peridotite may indeed have a mantle origin is indicated by the 18O values of 5·2 for olivine in sample GM 590. It is unlikely that the biotite–amphibole reversal in oxygen isotopic compositions resulted from low-temperature diffusive exchange of oxygen between these phases (with lower closure temperatures for biotite compared with amphibole; Giletti et al., 1978; Fortier & Giletti, 1991), as such an exchange would almost certainly also involve plagioclase and would tend to lower the 18O values of biotite relative to values at its temperature of formation. The quartz–biotite pair from zone 0, however, provides a temperature estimate of 660°C according to the calibration of Zheng (1993b). This temperature estimate perfectly agrees with phase equilibrium constraints.
In the context of regional thermobarometry (Bucher & Ohta, 1993; Markl & Piazolo, 1998; Piazolo & Markl, 1999), the temperature estimates constrain the pressure and possibly even the timing of the metasomatism: conditions of 650°C prevailed during the regional magmatic events at 530–500 Ma. This time span is also characterized by intrusion of abundant pegmatitic veins in the whole area. A pressure of 3–5 kbar for Central Dronning Maud Land was estimated based on phase equilibrium constraints from cordierite-bearing gneisses (Bucher & Ohta, 1993), from charnockitic intrusive rocks (Markl & Piazolo, 1998) and from calcsilicates (Piazolo & Markl, 1999). We therefore suggest that the vein formation took place at 660°C and most probably at 4 kbar pressure between 500 and 530 Ma.
Phase equilibrium constraints on intensive parameters within the reaction vein
Phase equilibria can be used to constrain at least some of the chemical characteristics of the infiltrating fluid. SiO2 activity is an important chemical parameter for a fluid that precipitates quartz (in the centre of the veins) within olivine-bearing mantle-type rocks. Furthermore, as the peridotite is an essentially anhydrous rock, H2O activity is also a parameter that determines the precise mineralogical structure of the reaction vein. Activity–activity diagrams as shown in Fig. 12 may serve to put constraints on such variables. These diagrams were calculated in the same way as Fig. 11.
|
Figure 12a shows phase equilibria in the KMASH system in a log a(K+/H+) vs log aSiO2 diagram at 4 kbar, 650°C and an H2O activity of 0·5. The coexistence of olivine and enstatite fixes SiO2 activity to just below 0·1 and the log a(K+/H+) ratio to values <4 in zone 4. At the border between zones 3 and 4, coexistence of cummingtonite with enstatite again fixes SiO2 activity (at a fixed H2O activity) to a value of
0·16 and the log a(K+/H+) ratio to four. The latter is fixed in zone 3 by the coexistence of cummingtonite, spinel and biotite, but not in zones 0, 1 and 2. The lower limit of a(K+/H+) in these zones, however, is the cummingtonite–spinel–biotite equilibrium at a value of four. SiO2 activity increases to quartz saturation in zone 0 without any fixed point in zones 3, 2 and 1. It can be stated that both SiO2 activity and log a(K+/H+) ratio increase steadily from zone 4 to zone 0. The same method applied to the CMASH system in Fig. 12b provides more constraints on SiO2 activity at constant H2O activity: the olivine–enstatite buffer in zone 4 (aSiO2 = 0·09) is replaced by the enstatite–cummingtonite buffer at the margin of zones 3 and 4 (aSiO2 = 0·13), by an equilibrium among cummingtonite, spinel and tschermakite at the border of zones 2 and 3 (aSiO2 = 0·15) and an equilibrium between tremolite and cummingtonite in zone 2 (aSiO2 = 0·17) before it drastically increases to quartz saturation in zone 0. Figure 12b also implies that the log a(Ca2+/H+2) ratio may increase only slightly from values around six in zone 4 to values around 6·2 at the border of zones 1 and 2. No constraint exists in zone 0.
A quantitative constraint on H2O activity in the reaction zone exists only at the boundary between zones 3 and 4. Here, H2O activity is constrained by the enstatite–cummingtonite–olivine equilibrium to a value of aH2O = 0·3. Unfortunately, suitable fluid inclusions are missing and further suitable phase equilibria are dependent on both SiO2 and H2O activity.
We explicitly state that the above calculations are prone to large errors as a result of the difficulties with activity models for the amphiboles and biotite, because of the compositional variations in these minerals within the zonation, which have not been taken into account, and because of the general lack of a constraint on the activity of H2O (with the exception of the boundary between zones 3 and 4; see below). However, they do provide crude guidelines for a comparison between theoretical phase equilibria and observed textures.
Oxygen and hydrogen isotope composition of the infiltrating fluid
To estimate the oxygen isotopic composition of the fluid responsible for the vein formation, the temperature of vein formation has to be known and it has to be assumed that the fluid has reached equilibrium with the alteration minerals. On the basis of the foregoing discussion it is likely that quartz, biotite and amphibole equilibrated at temperatures of 600–700°C. Furthermore, the relatively constant isotopic composition of biotite in zones 1–3 may be taken as evidence for equilibration with the fluid phase. Water calculated to be in equilibrium with these minerals at temperatures between 600 and 700°C would have 18O values between 8·5 and 10, according to the water–mineral fractionation factors of Zheng (1993a, 1993b). Such 18O values of water are typical for magmatic or metamorphic fluids. The variations in calculated isotopic compositions of olivine suggest that olivine approached equilibrium with the infiltrating fluid only at the very boundary of the reaction vein. The oxygen isotope exchange front does not appear to have penetrated much further into the peridotite. Olivine in sample GM 590 shows values typical for mantle peridotites (Mattey et al., 1994). This is not the case for orthopyroxene however, as its 18O values, even in sample GM 590, are atypical for mantle peridotites. This difference in the isotope compositions of orthopyroxene and olivine may be related to different oxygen diffusivities in these two minerals and/or variations in grain size of the samples analysed. For spinel, the value of 2·8 is typical for mantle peridotite, whereas the value of 4·7 may represent an isotopic composition in equilibrium with the pegmatitic fluid according to the fractionation factors of Zheng (1991) for chromite–hercynite–spinel.
The hydrogen isotope compositions of phlogopite are relatively homogeneous with values between -37 and -58, which would correspond to values of about -27 to -48 for water in equilibrium with phlogopite at temperatures of 650°C (Suzuoki & Epstein, 1974). However, the D values of amphibole are markedly different (-102 to -120) and would correspond to D values of -80 to -99 for water in equilibrium with these amphiboles. The latter assumes, however, that hornblende may be used as a proxy for the amphibole. As such, the hydrogen isotope composition would indicate two distinct alteration fluids for these minerals. An alternative explanation could again be the dependence of the mineral–water fractionation factors on the chemical composition of the hydrous mineral. Differences in octahedrally coordinated cations such as Fe, Mg and Al in micas and amphiboles, and the relatively high fluorine content of phlogopite can cause marked differences in the mica–water fractionation factors (e.g. Suzuoki & Epstein, 1974). Another possibility is that the phlogopite experienced hydrogen isotope exchange with another fluid at a later stage. Further investigations would be needed to distinguish these possibilities. In summary, the infiltrating fluid (or fluids) have originally been out of isotopic equilibrium with the peridotite and, on the basis of the oxygen isotope fractionations, the fluids were probably derived from crystallizing felsic pegmatites such as that found in zone 0, i.e. the centre of the vein.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONREGIONAL GEOLOGY AND SAMPLE...PETROGRAPHYMINERAL AND ISOTOPE GEOCHEMISTRYTHERMOMETRY AND MINERAL...DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Inferences from stable isotopes
The reaction veins discussed here formed by interaction of crustal, SiO2-saturated fluids with a fractured spinel peridotite tectonically entrained into crustal rocks. The mineral assemblage in the centre of the vein, the textures, the oxygen isotope data and the thermometric estimates agree with the interpretation of the fluid as an SiO2-saturated aqueous fluid derived from a pegmatitic melt. This inference is mainly based on the homogeneous oxygen isotope composition of the fluorine-rich phlogopite and amphibole and their fractionation relative to quartz and feldspar in the pegmatitic veins at temperatures between 600 and 700°C. The observation that the fluid was relatively F rich, but Cl poor, as shown by the phlogopite compositions, and the finding of abundant zircon in the phlogopite from zone 1 (Fig. 4) corroborate this interpretation. The small ‘secondary’ cummingtonite veins emanating from the main reaction zone clearly indicate that not the pegmatite itself, but an aqueous, F-rich fluid released from it was responsible for the reaction zones observed. The patchy appearance of zone 0 (see Fig. 3) and its irregular occurrence within the veins indicate crystallized melt pools, whereas the rest of the fractures only or mainly came in contact with the fluids expelled from the melt. This interpretation would allow a connection to be made between the metasomatic reaction veins and granitoid magmatism between 530 and 500 Ma (Jacobs et al., 1998), which occurred at a pressure of
3–5 kbar (Markl & Piazolo, 1998).
Olivine has resisted isotopic exchange with the infiltrating fluid and has reached equilibrium with it only in the innermost portion of zone 4, but it may preserve original, typical mantle-peridotite values further out in the peridotite. This does not appear to be the case for the orthopyroxene analysed so far, as it has relatively high 18O values, in equilibrium with the alteration fluid, out towards zone 4. The differences in behaviour could be related to differences in the oxygen diffusivity between olivine and pyroxene.
Development of the reaction zone
Figure 13 shows a schematic model for the development of the zoned reaction sequence and a schematic section through the reaction vein. The original boundary and the flux direction of the most important components have been indicated. The geometry of the vein implies that it formed by interaction of a fluid with the peridotite along a crack or an open fissure (Fig. 3). The zoned sequence of mineral assemblages that developed around this interface typically indicates transport control from peridotite towards fissure and vice versa (e.g. Joesten, 1977) in a field of chemical potential gradients. The vein geometry implies advective transport of an aqueous fluid along fissures, but diffusion- rather than advection-controlled element transport perpendicular to the fissures into the wall rock. The questions to address so as to understand vein metasomatism quantitatively are the following:
- Which elements moved towards the fissure, which towards the peridotite and which—if any—were immobile?
- Where was the original boundary between peridotite and fissure? Which parts of the reaction zones are replaced peridotite and which are newly grown into the fissure?
- What was the volume and mass change associated with the replacement of the peridotite?
|
Unfortunately, the chemical potential gradients that existed between fluid and peridotite could not be calculated from the observed mineral assemblages, as done previously by Joesten (1977), Markl et al. (1998) and Abart et al. (2001), because the geometry and the assemblages of the specific vein dealt with here are under-determined, and only saturation with more phases than actually present (such as, for example, biotite or cordierite and K-feldspar in the pegmatite) would allow estimates of maximum chemical potentials, which in turn can be linked to estimates for Onsager diffusion coefficients. More reaction zones or more phases in the various reaction zones would be necessary for more quantitative interpretations.
Element gains and losses
The isocon method of Grant (1986a, 1986b) allows estimation of the gains and losses of elements if the composition of the precursor rock is known. Figure 14 shows isocon plots for the whole reaction zone (Fig. 14a) and for the individual mineral bands (zone 1: Fig. 14b; zone 2: Fig. 14c; zone 3: Fig. 14d). SiO2, CaO and TiO2 (as well as K2O, Na2O, F, Cl and H2O, which are not shown, because they are essentially absent in unaltered peridotite) show gains in all zones, implying addition from the fluid. Al2O3 and MnO show gains in some and losses or constant content in other zones, whereas FeO, MgO, NiO and Cr2O3 show losses in all zones. It is noteworthy that in zone 3 (the only zone where we know that it is replaced peridotite) FeO is virtually constant and only MgO, Cr2O3 and NiO show losses, of which, however, only that of MgO is really significant if the uncertainties in mode and bulk composition estimates are taken into account. If parts of zone 2 or even of zone 1 are also peridotite replaced at constant volume, then FeO loss is significant as well.
|
The position of the original boundary and volume constraints
The position of the former boundary is shown as a dashed line in Fig. 13, as it cannot be directly determined from the geometry. However, it is clear from the textures (see Fig. 4) that zone 3 is replaced peridotite including original chromite. Hence, the outermost position possible for the original boundary is the border between zones 2 and 3. The irregular mode of occurrence, the very coarse grain size and the mineral assemblage indicate for zone 0 that it precipitated from the fluid or liquid in the open fissure. Hence, the innermost position possible for the original boundary is between zones 0 and 1, or, in turn, the original boundary must have been situated in or at the borders of zone 1 or 2.
The implications of different positions of the original interface for mass fluxes can be derived from mass balance considerations. This involves calculation of element fluxes. This can be looked at quantitatively by calculating the fluxes of elements in relation to the position of the original boundary and to the volume change during replacement of peridotite by an observed assemblage. The results of such calculations are illustrated in Fig. 15. In these calculations, a standardized zoned sequence was used, in which zone 3 is assigned a volume of 1000 cm3, and correspondingly—according to their thickness ratios in the observed veins—zone 2 and zone 1 have volumes of 700 and 500 cm3, respectively. The volume factor 
is defined as the ratio of the volume of zone x divided by the volume of zone y it was produced from. The curves shown represent the ‘no-flux-curves’ for the particular components, i.e. the curves above which the element flux is in the direction from the fissure towards the peridotite and below which the flux is in the direction from the peridotite towards the fissure. Only the components shown have such ‘inflection points’, where no flux occurs, under the conditions plotted in Fig. 15, i.e. the flux of the other components does not set any limit to the position of the original boundary.
|
The texture of the chromite grains in zone 3 (see Fig. 4) implies not only that this zone represents former peridotite, but also that it underwent only minor volume change during conversion. Chromite modal abundance increases from 10 to 12 modal %, which is within the limits of measurement uncertainty. Volume change can thus be limited by two ‘extreme’ cases: no volume change on the one side and constant chromite content on the other side. These constraints would indicate volume factors
of 1–1·2. We chose a value of 1·1 for the calculations, which were performed in the following way.
(1) The flux from zone 3 into zone 2 was calculated using the equation
 |
). The resulting flux J2-3 is given in Table 4.
(2) Then, successively larger parts of zone 2 were assumed to be replaced peridotite and the volume change 
for this reaction was varied:
 |
(3) In the right parts of the diagrams of Fig. 15, zone 2 was assumed to be completely replaced peridotite and successively larger parts of zone 1 were replaced peridotite. Here, however, another volume factor, 
for the conversion of peridotite into the zone 1 assemblage, must be considered. This volume factor was varied between 0·5 (Fig. 15a) and 1·5 (Fig. 15c):
 |
The hatched areas in the diagrams represent solutions that are unlikely for both volume factor and for the position of the original boundary. The most important constraint is exerted by Al2O3 followed by SiO2: both elements can safely be assumed to have been introduced into the peridotite, and hence the flux direction must be into the peridotite; this means, in this diagram, above the ‘no-flux-curves’ for Al2O3 and SiO2. Unfortunately, this indicates that no constraints on the position of the original boundary additional to the textural arguments presented above can be gained from mass balance arguments. It is, however, possible, to define the values of , and original boundary position that are incompatible with independent mass balance constraints (hatched areas in Fig. 15).
No textural volume constraints exist in zones other than zone 3, but given the observation of approximate (within 20%) volume conservation in zone 3 and given the similar molar volume of the phases in zones 1, 2 and 3, we assume that volume change in zones 2 and 1 also was small and, because we deal with hydration reactions, probably slightly positive.
|
SUMMARY AND CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONREGIONAL GEOLOGY AND SAMPLE...PETROGRAPHYMINERAL AND ISOTOPE GEOCHEMISTRYTHERMOMETRY AND MINERAL...DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
A spinel peridotite lens in the Schirmacher Oasis of Central Dronning Maud Land, Antarctica, shows several veins of up to 10 cm width with a characteristic zoned sequence of mineral assemblages from plagioclase plus quartz in the interior through biotite with clinoamphibole, biotite with cummingtonite and clinoamphibole, cummingtonite with biotite and chromite to unaltered peridotite (olivine + orthopyroxene + chromite). This reaction sequence most probably formed by a reaction of the peridotite with fluids derived from the crystallization of a pegmatitic melt at about 660°C and 4 kbar, and most probably related to the widespread Panafrican magmatism at 530–500 Ma.
Detailed analysis of the mechanism of formation of this type of vein has shown that the formation is related to mass transport of mainly SiO2, K2O, H2O and halogens from the fluid into the peridotite, which in turn releases mainly MgO and smaller quantities of Cr2O3 and NiO. Whether FeO was consumed or released depends on the volume change during the reaction and on the precise location of the original boundary of the fracture before the reaction, and can only be limited by endmember-type scenarios. The mass transport was probably governed by advective and diffusive mechanisms, but the well-developed geometrical pattern of the zoned reaction sequence and the curved rather than stepwise trend of mineral compositions throughout the zoned sequence argue for mainly diffusive control.
This vein is considered to be typical of veins encountered in harzburgites and peridotites. These veins may bear enstatite, talc or an orthoamphibole, in combination with possible tremolite. In most cases, the textures of the zoned sequence are interpreted as being derived from the interaction between an aqueous fluid and the peridotitic wall rock around a fracture. Hence, the veins dealt with here, which are apparently derived from pegmatite–peridotite interaction, are a slightly more complicated version of a general, mineralogically simple feature. In spite of this, the vein is not complex enough to allow derivation of diffusion coefficients as done by Joesten (1977), Markl et al. (1998) or Abart et al. (2001). The succession of zones and the mineral assemblages in the zones do not provide enough Gibbs–Duhem constraints for the large number of components involved in this type of reaction. The uncertainty regarding the position of the original boundary additionally hinders the quantitative derivation of mass fluxes, which would be another constraint on diffusivity. Against intuition, a simple geometry and a very simple (ultramafic) chemical system are not good candidates for derivation of quantitative constraints on diffusion. It appears that the chemical systems involved here—mantle rock vs crustal fluid—are too different, and hence the number of components important for the development of the reaction vein is much larger than the number of minerals and/or zones. Less different chemical systems may be controlled by a lower number and by smaller chemical potential gradients, and may hence involve only few components that are important for the determination of vein geometry and mineralogy. Obviously, these systems or their combination are more useful for quantitative treatment of diffusion-controlled processes.
Nevertheless, the present work improves the quantitative understanding of vein formation in these complicated systems and places limits on original geometry and element mobility. Because the chemical interaction of crustal fluids with ultramafic rocks is relatively common in orogenic settings, this first attempt to quantitatively understand such interactions may be augmented in future.
|
ACKNOWLEDGEMENTS |
|---|
Very thoughtful and constructive reviews by Jamie Connolly, Kåre Kullerud and Volkmar Trommsdorff are most gratefully acknowledged. Jamie Connolly deserves a special acknowledgement for not only raising some points of concern, but also for his help during elimination of these troublesome points. Field work for this investigation was carried out during the German ‘GeoMaud 95/96’ Antarctic expedition of the Bundesanstalt für Geowissenschaften und Rohstoffe (BGR), Hannover. Participation of G.M. on this expedition is gratefully acknowledged, as is financial support by grant Ma2135/2-1 from the German Science Foundation (DFG).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONREGIONAL GEOLOGY AND SAMPLE...PETROGRAPHYMINERAL AND ISOTOPE GEOCHEMISTRYTHERMOMETRY AND MINERAL...DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Abart, R., Schmud, R. & Harlov, D. (2001). Metasomatic coronas around hornblendite xenoliths in granulite-facies marble, Ivrea zone, N Italy: constraints on component mobility. Contributions to Mineralogy and Petrology 141, 473–493.[Web of Science]
Armstrong, J. T. (1991) Quantitative elemental analysis of individual microparticles with electron beam instruments. In: K. F. J. Heinrich and D. E. Newbury (eds) Electron Probe Quantitation. New York: Plenum, pp. 261–315.
Ashworth, J. R. & Sheplev, V. S. (1997). Diffusion modelling of metamorphic layered coronas with stability criterion and consideration of affinity. Geochimica et Cosmochimica Acta 61, 3671–3689.[CrossRef][Web of Science]
Berman, R. G. (1988). Internally-consistent thermodynamic data for minerals in the system Na2O–K2O–CaO–MgO–FeO– Fe2O3–Al2O3–SiO2–TiO2–H2O–CO2. Journal of Petrology 29, 445–522.
Bucher, K. & Ohta, Y. (1993). Granulites and garnet–cordierite gneisses from Dronning Maud Land, Antarctica. Journal of Metamorphic Geology 11, 691–703.[Web of Science]
Bucher-Nurminen, K. (1988). Metamorphism of ultramafic rocks in the Central Scandinavian Caledonides. Norsk Geologisk Undersoekelse Special Publication 3, 86–95.
Bucher-Nurminen, K. (1990). Transfer of mantle fluids to the lower continental crust: constraints from mantle mineralogy and Moho temperature. Chemical Geology 83, 249–261.[CrossRef][Web of Science]
Carswell, D. A., Curtis, C. D. & Kanaris-Sotiriou, R. (1974). Vein metasomatism in peridotite at Kalskaret near Tafjord, South Norway. Journal of Petrology 15, 383–402.
Dalziel, I. W. (1991). Pacific margins of Laurentia and East Antarctica–Australia as a conjugate rift pair: evidence and implications for an Eocambrian supercontinent. Geology 19, 598–601.
Evans, B. W. & Trommsdorff, V. (1974). Stability of enstatite + talc, and CO2-metasomatism of metaperidotite, Val d'Efra, Lepontine Alps. American Journal of Science 274, 274–298.[Abstract]
Fisher, G. W. (1973). Non-equilibrium thermodynamics as a model for diffusion controlled metamorphic processes. American Journal of Science 273, 897–924.[Abstract]
Fisher, G. W. (1975). The thermodynamics of diffusion controlled metamorphic processes. In: Cooper, A. R. & Heuer, A. H. (eds) Mass Transport Phenomena in Ceramics. New York: Plenum, pp. 111–122.
Fisher, G. W. (1977). Non-equilibrium thermodynamics in metamorphism. In: Fraser, D. G. (ed.) Thermodynamics in Geology. Boston, MA: D. Reidel, pp. 381–403.
Fortier, S. M. & Giletti, B.J. (1991). Volume self-diffusion of oxygen in biotite, muscovite, and phlogopite micas. Geochimica et Cosmochimica Acta 55, 1319–1330.[CrossRef][Web of Science]
Fortier, S. M., Lüttge, A., Satir, M. & Metz, P. (1994). Oxygen isotope fractionation between fluorphlogopite and calcite: an experimental investigation of temperature dependence and F-/OH- effects. European Journal of Mineralogy 6, 53–65.
Foster, C. T. (1983). Thermodynamic models of biotite pseudomorphs after staurolite. American Mineralogist 68, 389–397.[Abstract]
Ghiorso, M. S. & Evans, B. W. (2002). Thermodynamics of the amphiboles: Ca–Mg–Fe2+ quadrilateral. American Mineralogist 87, 79–98.
Giletti, B. J., Semet, M. P. & Yund, R. A. (1978). Studies in diffusion—III: oxygen in feldspars: an ion microprobe determination. Geochimica et Cosmochimica Acta 42, 45–57.[CrossRef][Web of Science]
Grant, J. A. (1986a). The isocon diagram; a simple solution to Gresens' equation for metasomatic alteration. Economic Geology 81, 1976–1982.
Grant, J. A. (1986b). Quartz–phlogopite–liquid equilibria and origins of charnockites. American Mineralogist 71, 1071–1075.[Abstract]
Jacobs, J., Fanning, C. M., Henjes-Kunst, F., Olesch, M. & Paech, H.-J. (1998). Continuation of the Mozambique belt into East Antarctica: Grenville-age metamorphism and polyphase Pan-African high-grade events in central Dronning Maud Land. Journal of Geology 106, 384–406.
Joesten, R. (1977). Evolution of mineral assemblage zoning in diffusion metasomatism. Geochimica et Cosmochimica Acta 41, 649–670.[CrossRef][Web of Science]
Korzhinskii, D. S. (1970). The Theory of Metasomatic Zoning. Oxford: Clarendon Press, 162 pp.
Leake, B. E., Woolley, A. R., Arps, C. E., Birch, W. D., Gilbert, M. C., Grice, J. D., Hawthorne, F. C., Kato, A., Kisch, H. J., Krivovichev, V. G., et al. (1997) Nomenclature of amphiboles; report of the Subcommittee on Amphiboles of the International Mineralogical Association, Commission on New Minerals and Mineral Names. American Mineralologist 82, 1019–1037.[Abstract]
Lieberman, J. & Petrakakis, K. (1991). TWEEQU thermobarometry; analysis of uncertainties and applications to granulites from western Alaska and Austria. Canadian Mineralogist 29, 857–887.[Web of Science]
Markl, G. & Piazolo, S. (1998). Halogen-bearing minerals in syenites and high-grade marbles of Dronning Maud Land, Antarctica: monitors of fluid compositional changes during late-magmatic fluid–rock interaction processes. Contributions to Mineralogy and Petrology 132, 246–268.[CrossRef][Web of Science]
Markl, G., Foster, C. T. & Bucher, K. (1998). Diffusion-controlled olivine corona textures in granitic rocks from Lofoten, Norway: calculation of Onsager diffusion coefficients, thermodynamic modelling and petrologic implications. Journal of Metamorphic Geology 16, 607–623.[CrossRef][Web of Science]
Mattey, D. P., Lowry, D. & Macpherson, C. (1994). Oxygen isotope composition of mantle peridotite. Earth and Planetary Science Letters 128, 231–241.[CrossRef][Web of Science]
Morrison, J. (1991). Compositional constraints on the incorporation of Cl into amphiboles. American Mineralogist 76, 1920–1930.[Abstract]
Pfeifer, H.-R. (1977). A model for fluids in metamorphosed ultramafic rocks. Schweizerische Mineralogische und Petrographische Mitteilungen 57, 361–396.
Piazolo, S. & Markl, G. (1999). Humite- and scapolite-bearing assemblages in marbles and calcsilicates of Dronning Maud Land, Antarctica and their implications on Gondwana reconstructions. Journal of Metamorphic Geology 17, 91–107.[CrossRef][Web of Science]
Ravich, M. G. & Solov'ev, D. S. (1966). Geology and Petrology of the Central Part of Queen Maud Land, East Antarctica (in Russian). Leningrad: Nedra, 290 pp.
Rumble, D., III & Hoering, T. C. (1994). Analysis of oxygen and sulfur isotope ratios in oxide and sulfide minerals by spot heating with a carbon dioxide laser in a fluorine atmosphrere. Accounts of Chemical Research 27, 237–241.[CrossRef][Web of Science]
Sharp, Z. D. (1990). A laser-based microanalytical method for the in-situ determination of oxygen isotope ratios of silicates and oxides. Geochimica et Cosmochimica Acta 54, 1353–1357.[CrossRef][Web of Science]
Suzuoki, T. & Epstein, S. (1974). Hydrogen isotope fractionation between OH-bearing minerals and water. Geochimica et Cosmochimica Acta 40, 1229–1240.
Valley, J. W., Kitchen, N., Kohn, M. J., Niendorf, C. R. & Spicuzza, M. J. (1995). UWG-2, a garnet standard for oxygen isotope ratios: strategies for high precision and accuracy with laser heating. Geochimica et Cosmochimica Acta 59, 5223–5231.[CrossRef][Web of Science]
Vennemann, T. W. & O'Neil, J. R. (1993). A simple and inexpensive method of hydrogen isotope and water analyses of minerals and rocks based on zinc reagent. Chemical Geology 103, 227–234.[CrossRef][Web of Science]
Volfinger, M., Robert, J.-L., Vielzeuf, D. & Neiva, A. M. R. (1985). Structural control of the chlorine content of OH-bearing silicates (micas and amphiboles). Geochimica et Cosmochimica Acta 49, 37–48.[CrossRef][Web of Science]
Will, T. M. & Powell, R. (1992). Activity–composition relationships in multicomponent amphiboles: an application of Darken's quadratic formalism. American Mineralogist 77, 954–966.[Abstract]
Zheng, Y.-F. (1991). Calculation of oxygen isotope fraction- ation in metal oxides. Geochimica et Cosmochimica Acta 55, 2299–2307.[CrossRef][Web of Science]
Zheng, Y.-F. (1993a). Calculation of oxygen isotope fractionation in hydroxyl-bearing silicates. Earth and Planetary Science Letters 120, 247–263.[CrossRef][Web of Science]
Zheng, Y.-F. (1993b). Calculation of oxygen isotope fractionation in anhydrous silicate minerals. Geochimica et Cosmochimica Acta 57, 1079–1091.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||