Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.geol.msu.ru/deps/petro/Perchuk_publ/perchuk_geochem08-eng.pdf Äàòà èçìåíåíèÿ: Wed Feb 15 16:07:20 2012 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 01:53:51 2012 Êîäèðîâêà: koi8-r

ISSN 0016-7029, Geochemistry International, 2008, Vol. 46, No. 3, pp. 296­303. © Pleiades Publishing, Ltd., 2008. Original Russian Text © A.L. Perchuk, 2008, published in Geokhimiya, 2008, No. 3, pp. 331­338.

SHORT COMMUNICATIONS

Unusual Inclusions in Garnet from the Diamond-Bearing Gneiss of the Erzgebirge, Germany
A. L. Perchuka,
a b

Institute of Geology of Ore Deposits, Petrography, Mineralogy, and Geochemistry (IGEM), Russian Academy of Sciences, Staromonetnyi per. 35, Moscow, 119017 Russia e-mail: alp@igem.ru b Institute of Experimental Mineralogy, Russian Academy of Sciences, Chernogolovka, Moscow region, 142432 Russia
Received December 22, 2006

DOI: 10.1134/S0016702908030063

INTRODUCTION The É-Ì conditions of diamond-bearing metamorphic complexes [1] correspond to "cold" geotherms with relatively low temperature (Ì) to pressure (É) ratios. It is known that such geotherms may occur only in subduction and collision zones [2], and the collision geotectonic setting is favorable for the formation of diamond-bearing metamorphic complexes [3, 4]. According to formal criteria (mineralogy, structure, and texture), the diamond-bearing rocks of metamorphic complexes are classified as metamorphic. However, some petrological, geochemical, and experimental data of recent years indicated their possible partial or complete melting [5­12]. It should be noted that there are still no reliable criteria for the identification of such melts. The reason is that the exhumation rate of ultrahigh pressure (UHP) complexes is lower than the rate of phase transformations, and metamorphic reactions obliterate, therefore, evidence for melting. On the other hand, the knowledge of the physical state of rocks is very important, because it strongly affects the mechanisms of very important processes, including diamond formation and rock exhumation. Petrological and mineralogical studies played a key role in the discovery of diamondiferous metamorphic complexes and can be used to solve these important problems. This paper reports a discovery of unusual inclusions in garnet from the diamond-bearing garnet­mica gneiss of the Saxonian Erzgebirge. The morphology and phase composition of these inclusions could hardly be explained within the model of solid-phase evolution only. GEOLOGIC SETTING AND PETROLOGY (LITERATURE DATA) The Erzgebirge composes a large anticlinorium in the Variscan basement of the Bohemian Massif in Central Europe [12, 13]. The core of the complex is composed of migmatite-bearing gneisses and schists hosting eclogite and peridotite lenses. Diamond-bearing

garnet­mica gneisses occur as lenses (up to several hundred meters long) in the same sequence [14]. The appearance of these rocks is indistinguishable from that of rocks of the same name widespread in moderate pressure complexes. Diamonds in the gneisses of the Erzgebirge were discovered by chance owing to problems with thin section polishing [15]. Subsequently, in addition to diamond, another lesser known indicator of high pressures was found in these rocks, a polymorphous modification of rutile--titanium dioxide with the structural type of PbO2 [16]. The lower pressure stability limit of this mineral depends on the crystal size: the pressure is lower (about 4 GPa) for a nanophase and 6 GPa for larger crystals [17]. A nanometer-sized plate of this mineral sandwiched between rutile twins in an inclusion in garnet was reported from the Erzgebirge [16]. Garnet and, occasionally, zircon and kyanite are host minerals for diamond crystals. Thus, the main indicators of the deep origin of the rocks are preserved only as inclusions in highly resistant minerals. Diamond occurs mainly in polymineralic inclusions, which may contain quartz, potassium feldspar, phlogopite, phengite, paragonite, rutile, and apatite [12, 14]. Graphite pseudomorphs after diamond and/or individual graphite scales were found among the inclusions. There are also inclusions free of carbon phases. Phengite is partly replaced by retrograde chlorite in decrepitated inclusions [12]. The central and marginal parts of garnet have different compositions. The cores are enriched in Ca and contain Ti, whereas the rims are richer in Mg and free of Ti (table). The compositions of phengite from inclusions and matrix are distinctly different (table). The phengite from inclusions shows a higher silica content related to the high-pressure celadonite endmember. It is believed that the diamond-bearing gneisses of the Erzgebirge were formed under very high, for crustal rocks, É-Ì conditions: >1000°Ê and 4.5 GPa [14].

296


UNUSUAL INCLUSIONS IN GARNET FROM THE DIAMOND-BEARING GNEISS

297

Representative microprobe analyses of garnet and phengite from the diamond-bearing gneiss of the Erzgebirge, Germany Garnet Component center SiO2 TiO2 A12O3 Cr2O FeO MnO MgO CaO Na2O K2O Total Si A1 Cr Ti Fe
3+* 3

Phengite rim near Phn 37.39 0.00 21.31 0.10 27.81 0.43 6.89 5.14 0.17 0.01 99.25 2.94 1.97 0.01 0.00 0.08 1.75 0.03 0.81 0.43 0.03 0.00 8.02 0.32 0.14 Inclusion in Grt 49.94 0.01 26.04 0.10 3.57 0.01 2.95 0.26 0.06 8.54 91.48 11 O atoms 3.52 2.16 0.01 0.00 ­ 0.21 0.00 0.31 0.02 0.01 0.77 7.00 0.32 ­ matrix 45.94 1.48 28.89 0.02 2.11 0.00 2.42 0.14 0.15 10.19 91.34 3.25 2.41 0.00 0.08 ­ 0.12 0.00 0.26 0.01 0.02 0.92 7.06 0.36 ­

rim near Qtz 38.02 0.00 22.24 0.00 27.90 0.44 7.41 5.33 0.01 0.10 101.45 12 O atoms 2.92 2.01 0.00 0.00 0.07 1.72 0.03 0.85 0.44 0.00 0.01 8.04 0.33 0.14

37.54 0.26 21.49 0.00 27.04 0.25 6.82 5.67 0.22 0.00 99.29 2.94 1.99 0.00 0.02 0.06 1.71 0.02 0.80 0.48 0.03 0.00 8.00 0.32 0.16

Fe2+ Mn Mg Ca Na K Total Mg/(Mg + Fe) Ca/(Ca + Mg + Fe)

* Calculated from stoichiometry.

PETROLOGICAL CHARACTERISTICS OF THE DIAMOND-BEARING GNEISS The major minerals of the gneiss are quartz, plagioclase, phengite (sometimes replaced by retrograde biotite), and garnet. The accessory minerals are rutile and zircon. Garnet forms small (up to 3 mm) anhedral grains, which often contain single-phase and polymineralic inclusions of quartz, paragonite, phengite, phlogopite, rutile, and apatite. Diamond (up to 30 µm in diameter) usually associates with these inclusions. Unusual mineral assemblages were observed in one thin section. Two relatively large inclusions (quartz + phengite ± rutile ± apatite assemblage) were observed
GEOCHEMISTRY INTERNATIONAL Vol. 46 No. 3

in the central part of garnet (Fig. 1). Each inclusion contains microcrystals (~3 µm) with high optical relief. Their energy-dispersive spectra corresponded to carbon, and they were interpreted as diamonds. The existence of radial cracks around the polymineralic inclusions (Fig. 1) suggests local stress relaxation. The largest cracks are filled with secondary chlorite and disturb in part the integrity of the inclusions, which is evident from the partial replacement of primary phengite by chlorite in such inclusions (Figs. 1b, 1c). It is remarkable that the inclusions are surrounded by peculiar haloes containing swarms of tiny phases, less than one micrometer in size. The haloes are clearly discernible under an optical microscope, both in plane-polarized
2008


298

PERCHUK ()

300 µm

(b)
Rut Grt Phn Rut Qtz Ap

(c)
Phn Chl Qtz

Chl

Phn Chl Grt

100 µm

100 µm

Fig. 1. Multiphase inclusions in garnet from the diamond-bearing garnet­phengite gneiss of the Erzgebirge, Germany. (a) Decrepitation haloes around polymineralic inclusions. Cross-polarized light. The inset shows the decrepitation halo in plane-polarized light. (b) and (c) back-scattered electron images of the polymineralic inclusions and host garnet shown in Fig. 1a (Qtz, quartz; Phn, phengite; Rut, rutile; and Chl, secondary chlorite). The arrows point to diamonds. Note that the resolution of the scanning electron microscope is not sufficient for the detection of decrepitation haloes. The light dots in quartz and garnet are contamination (of vacuoles?) during thin section preparation (Pb prevails in their energy-dispersive spectra).

and cross-polarized light (Fig. 1). Our attempts to examine the compositions and morphologies of these phases using a CamScan MV2300T scanning electron microscope with an energy-dispersive X-ray (EDX) spectrometer (INCA Energy) (Institute of Experimental Mineralogy, Russian Academy of Sciences) failed,

because the inclusions were indistinguishable from the host mineral. A comparison of integrated analyses of garnet with and without a halo (both over an area of 20 â 20 µm) did not reveal any difference in chemical composition. The inclusions are probably so small that their relative content in the near-surface part of the thin
Vol. 46 No. 3 2008

GEOCHEMISTRY INTERNATIONAL


UNUSUAL INCLUSIONS IN GARNET FROM THE DIAMOND-BEARING GNEISS

299

minerals [18, 19], attention is drawn to a unique lamella composed of the three mentioned phases (Fig. 3). DISCUSSION Arguments for the Partial Melting of the Rock It is well known from the experience of the investigation of fluid and melt inclusions in high-pressure rocks that, when local overpressure develops within them owing to changes in temperature and/or pressure, the inclusions may decrepitate (burst), and part of their material will be lost. The morphology of decrepitation products depends on the physicochemical conditions of the environment, viscoelastic properties of the host mineral and inclusions, and their geometry and size [20]. Well known are decrepitation haloes formed by numerous microscopic phases surrounding the parent inclusion. For instance, Fig. 4. presents spectacular decrepitation haloes around a sulfide melt inclusion in olivine from lherzolite [21] and inclusions of carbon dioxide in olivine from melilitite [22], which are amazingly similar to the haloes around the polymineralic inclusions in garnet from the diamond-bearing gneiss (cf. Figs. 1 and 4). The origin of haloes in garnet is difficult to interpret assuming a solid-phase evolution. Therefore, their morphological similarity to the decrepitation haloes in olivine shown in Fig. 4 is most likely related to a common decrepitation mechanism of their formation. The P-T conditions of the diamond-bearing gneisses are significantly higher than the second critical point in the granite­water system (Fig. 5), above which complete miscibility between fluid and melt is observed [23­25]. Therefore, a fundamental point in the finding
Phl? & Grt

200 µm
Fig. 2. Crystallographically oriented inclusions (indicated by arrows) in a 0.4-mm-thick garnet section. Plane-polarized light. See text and Fig. 3 for further details.

section (X-ray radiation is excited in a few micrometerthick layer) appeared to be below the sensitivity of microprobe analysis. Oriented acicular microcrystals, less than 1­2 µm wide and up to 200 µm long, were observed in the central parts of garnets (including the zones of haloes) (Fig. 2). It is difficult to reliably determine the composition of such thin grains, because the excitation volume (~5 µm in diameter) inevitably involved the host mineral. The integrated EDX patterns allowed us to unambiguously identify only rutile lamellae, whereas phases with high contents of (K, Mg) and Na are supposed to be phlogopite and paragonite, respectively. Given the high interest in oriented lamellae in UHP
Grt

Rut & Grt

Prg? & Grt

10 µm

Fig. 3. Back-scattered electron image of a three-phase acicular inclusion consisting of rutile (Rut) + Na-bearing phase (paragonite?, Prg) + K-bearing phase (phlogopite?, Phl) in garnet (Grt). The insets show integrated energy dispersive spectra of the lamella and host garnet. A phlogopite­paragonite inclusion (dark) occurs below the lamella. GEOCHEMISTRY INTERNATIONAL Vol. 46 No. 3 2008


300

PERCHUK Pressure, GPa 5 70 4 90 3 50 30 10% H2O Erzgebirge

2 Fluid (a) 100 µm 1 Melt

0

400

800

1200 Temperature, °C

Fig. 5. Melting curve in the granite­H2O system [23­25]. The shaded area shows the P-T conditions of the formation of the diamond-bearing gneiss [14]. The phase boundary between melt and fluid disappears above the second critical point. The numbers are water contents in wt %.

(b)

50 µm

Fig. 4. Decrepitation haloes around melt and fluid inclusions (data from the literature) similar to the aureoles surrounding the multiphase inclusions in garnet from the diamond-bearing gneiss. (a) Decrepitation of a sulfide melt inclusion in olivine from lherzolite, Mushugai-Khuduk, Mongolia [21]. (b) Decrepitated fluid inclusion (CO2) in olivine from olivine melilitite, Rhine graben, Germany [22].

of decrepitation haloes is not the presence of a fluid or a melt but the fact of the existence of a water­silicate supercritical liquid as the main mineral-forming medium. Proceeding from the phase diagram (Fig. 5), the inclusion could crystallize from an essentially silicate liquid with a water content of no higher than 10 wt %. The decrepitation halo could be formed during various stages of inclusion evolution. On the one hand, high temperatures and pressures diminish the viscosity of silicic melts [26] and could promote the decrepitation of inclusions already within the field of homogeneous supercritical fluid. In such a case, microinclusions in the decrepitation haloes must be rich in silicate material (Fig. 5). According to another scenario, the fluid in the inclusions could reach the P-T conditions of unmixing; in such a case, the least viscous water-rich phase containing >70% H2O (Fig. 5) had to

escape into the decrepitation halo. Such a liquid will fill the vacuoles of microinclusions in the decrepitation halo, whereas silica-rich melt will remain in the inclusion up to its complete crystallization. A common assumption in both these scenarios is the presence of supercritical liquid at peak P-T parameters, which was probably a diamond-forming medium. The experimental study of a model continental crust composition [9] supported the possibility of its melting with the formation of garnet as a liquidus phase under the P-T conditions of the formation of the diamond-bearing gneisses. The composition of the polymineralic inclusions in garnet is not uniform. While quartz apparently dominates the inclusions surrounded by decrepitation haloes, other small inclusions contain both quartz and significant amounts of mica minerals; such inclusions were regarded by Stockert et al. [12] as relics of C­O­H silicate fluid. The portions of silicic liquid trapped in the garnet crystals probably reflect either its chemical heterogeneity or partial loss of material from the inclusions. If the inclusions in garnet were formed from a supercritical liquid, the matrix of the rock could not also avoid extensive melting at the peak P-T conditions. This is related to the fact that, in addition to the minerals occurring in the inclusions (quartz and phengite), the gneiss matrix contains plagioclase, which depresses the melting temperature of the system. However, it should be admitted that, in contrast to the isolated
Vol. 46 No. 3 2008

GEOCHEMISTRY INTERNATIONAL


UNUSUAL INCLUSIONS IN GARNET FROM THE DIAMOND-BEARING GNEISS

301

inclusions in garnet, evidence for matrix melting was obliterated by metamorphic reactions during ascent from depths of more than 100 km and cannot be unambiguously identified. The polymineralic inclusions surrounded by decrepitation cracks and haloes are either portions of supercritical liquid encapsulated within the growing garnet or products of melting of a mineral association trapped by the garnet during early stages of metamorphic evolution. The inclusions of phengite and rutile indicate the presence of water and titanium in the hypothetical melt. According to experimental data [9, 10], the interaction of such a melt with the host garnet (Grt1) must lead to its extensive recrystallization and the formation of a new titanium-bearing garnet (Grt2). The replacement Grt1 Grt2 could develop not only from the inclusion but also in the reverse direction, from the boundary of garnet with the matrix, if the latter occurred in a molten state. In the case of garnet interaction with Ti-bearing melt, the maximum Ti concentration in garnet is controlled by the Ti partition coefficient between garnet and melt. The presence of Ti in the core parts of all the garnets indicates the possibility of such interaction. Oriented Mineral Inclusions During recent years, much attention is being focused on oriented lamellae in the minerals of UHP rocks. Their formation is usually explained by the decompression-related exsolution of UHP components of the host mineral. For instance, such an interpretation was proposed for the oriented inclusions of sanidine, phlogopite, ilmenite, titanite, and quartz in clinopyroxene; ilmenite and chromite in olivine; pyroxene and rutile in garnet; and coesite in titanite [18, 19, 27­31]. The volume of inclusions formed by such a mechanism (under closed-system conditions) indicates the content of a corresponding UHP component in the host mineral and can be used as a geobarometer. However, there is an alternative mechanism in addition to the well-known epitaxial growth in low-temperature metamorphic processes. For instance, inclusions of sanidine in clinopyroxene and kokchetavite in garnet and clinopyroxene from the garnet­clinopyroxene rocks of the Kokchetav Massif were interpreted as decompression products that developed owing to the interaction of the host mineral with coexisting melt [7, 32]. To our knowledge, the oriented lamellae of rutile and Na and K­Mg phases in garnet from the diamondbearing gneisses of the Erzgebirge were never described. While oriented acicular rutil crystals in garnet are well known in UHP rocks from various regions [19, 29, 31], needlelike inclusions of Na and K­Mg phases are a specific feature of the rocks described here. The lamellae of the Na phase (paragonite?) could, in principle, inherit Na and (ÈÃ)­ from UHP garnet [33-35], and, therefore, similar to rutile, they could crystallize from the garnet under closed-system conditions. However, such a scenario is improbable for the
GEOCHEMISTRY INTERNATIONAL Vol. 46 No. 3

K­Mg lamellae (phlogopite?). It is known from the literature that the large cation Ä+ is incorporated in the garnet structure only at pressures higher than 20 GPa [36], which is several times higher than the pressure of the Erzgebirge complex. Consequently, the source of potassium must lie outside the garnet structure. The finding of a three-phase needle (Fig. 5) suggests that not only K but also Ti and Na were probably scavenged from an external source, which must have occurred either in the rock matrix or in inclusions. Recall that paragonite and phlogopite were often observed in polymineralic inclusions (sometimes with diamond) in the garnet. Thus, the oriented inclusions in garnet from the diamond-bearing gneiss were not exsolved from the host mineral and, consequently, cannot be considered as an indicator of ultrahigh pressures. They could be formed through the interaction of garnet with coexisting liquid occurring in inclusions and/or in the rock matrix during decompression. CONCLUSIONS The mineral assemblage of the matrix of the diamond-bearing gneisses of the Erzgebirge complex was recrystallized during its ascent and, therefore, shows no evidence for UHP conditions. Relevant information is preserved only on micrometer and nanometer levels, in mineral inclusions. Because of this, the products of partial melting of rocks may not occur in forms typical of metamorphic sequences (migmatites and veinlets). Vestiges of melts (high-density supercritical liquids) should be sought primarily in inclusions. Such a localization, unusual for metamorphic rocks, and inevitable crystallization of melt complicate its identification. Decrepitation haloes around multiphase inclusions are considered by us as an important indicator of a liquid, from which diamond crystallized. Investigations with high-resolution instruments will provide insight into the chemical composition of the microinclusions but will not alter the main conclusions of this study. ACKNOWLEDGMENTS The author is grateful to I.P. Solovova for consultations on melt inclusions and photomicrographs of decrepitated inclusions. The manuscript benefited from critical comments by I.P. Solovova, L.L. Perchuk, I.A. Andreeva, and A.I. Gorshkov. A sample from the collection of V.V. Zakharchuk was used in this study. Scanning electron microscope investigations were carried out under the participation of A.N. Nekrasov and K.V. Van. This study was financially supported by the Russian Foundation for Basic Research, project nos. 06-05-65204 and 06-05-64098; the Foundation for the Support of Russian Science; and grant no. NSh-5338.2006.05 of the Program for the Support of Leading Scientific Schools of Russia (supervised by L.L. Perchuk).
2008


302

PERCHUK

REFERENCES
1. C. Chopin, "Ultrahigh-Pressure Metamorphism: Tracing Continental Crust into the Mantle," Earth Planet. Sci. Lett. 212, 10­14 (2003). 2. S. Peacock, "Numerical Simulation of Metamorphic Pressure­Temperature­Time Paths and Fluid Production in Subduction Slabs," Tectonics 9, 1197­1211 (1990). 3. N. L. Dobretsov, V. S. Shatsky, and N. V. Sobolev, "Comparison of the Kokchetav and Dabie Shan Metamorphic Complexes: Coesite- and Diamond-Bearing Rocks and UHP­HP Accretional­Collisional Events," Int. Geol. Rev. 37, 636­656 (1995). 4. W. G. Ernst, S. Maruyama, and S. Wallis, "Buoyancy Driven, Rapid Exhumation of the Ultrahigh Pressure Phases in Metamorphosed Continental Crust," Proc. Nat. Acad. Sci. USA. 94, 9532­9537 (1997). 5. L. L. Perchuk and V. O. Yapaskurt, "Deep-Seated Ultrapotassic Liquids," Geol. Geofiz., No. 12, 1756­1765 (1998). 6. V. S. Shatsky, E. Jagoutz, N. V. Sobolev, et al., "Geochemistry and Age of Ultrahigh Pressure Metamorphic Rocks from the Kokchetav Massif (Northern Kazakhstan)," Contrib. Mineral. Petrol. 137, 185­205 (1999). 7. L. L. Perchuk, O. G. Safonov, V. O. Yapaskurt, and J. M. Barton, "Crystal­Melt Equilibria Involving PotassiumBearing Clinopyroxene as Indicator of Mantle-Derived Ultrahigh-Potassic Liquids: An Analytical Review," Lithos 60, 89­111 (2002). 8. J. Hermann, "Experimental Constraints on Phase Relations in Subducted Continental Crust," Contrib. Mineral. Petrol. 143, 219­235 (2004). 9. A. L. Perchuk, M. M. Burchard, W. V. Maresch, and H.-P. Schertl, "Fluid-Mediated Modification of Garnet Interiors under Ultrahigh-Pressure Conditions," Terra Nova 17, 545­553 (2005). 10. H. M. Lang and J. A. Gilotti, "Partial Melting of Metapelites at Ultrahigh-Pressure Conditions, Greenland Caledonides," J. Metamorph. Geol. 25, 129­147 (2007). 11. A. V. Korsakov and J. Hermann, "Silicate and Carbonate Melt Inclusions Associated with Diamonds in Deeply Subducted Carbonate Rocks," Earth Planet. Sci. Lett. 241, 104­118 (2006). 12. B. StÆckhert, J. Duyster, C. Trepmann, and H.-J. Massonne, "Microdiamond Daughter Crystals Precipitated from Supercritical COH Silicate Fluids Included in Garnet, Erzgebirge, Germany," Geology 29, 391­394 (2001). 13. A. P. Willner, E. Sebazungu, T. V. Gerya, et al., "Numerical Modeling of PT-Paths Related to Rapid Exhumation of High-Pressure Rocks from the Crustal Root in the Variscan Erzgebirge Dome (Saxony/Germany)," J. Geodynamics 33, 281-314 (2002). 14. H.-J. Massonne, "A New Occurrence of Microdiamonds in Quartzofeldspathic Rocks of the Saxonian Erzgebirge, Germany, and Their Metamorphic Evolution," in Proceedings of 7th International Kimberlite Conference, Cape Town, South Africa, 1999 Ed. by P. H. Nixon (Red Roof Design, Cape Town, 1999), Vol. 2, pp. 533­539. 15. H. J. Massonne, H. J. Bernhardt, D. Dettmar, et al., "Simple Identification and Quantification of Microdiamonds in Rock Thin-Sections" Eur. J. Mineral. 10, 497­ 504 (1999).

16. S. L. Hwang, P. Y. Shen, H. T. Chu, and T. F. Yui, "Nanometer-Size -PbO2 Type TiO2 in Garnet: A Thermobarometer for Ultrahighpressure Metamorphism," Science 288, 321­324 (2000). 17. J. S. Olsen, L. Gerward, and J. Z. Jiang, "On the Rutile­ PbO2 Type Phase Boundary of TiO2," J. Phys. Chem. Solids. 60, 229­233 (1999). 18. L. F. Dobrzhinetskaya, H. W. Green, and S. Wang, "Alpe Arami: A Peridotite Massif from Depths of More than 300 km," Science 271, 1841­1845 (1996). 19. R. Y. Zhang, S. M. Zhai, Y. W. Fei, and J. H. Liou, "Titanium Solubility in Coexisting Garnet and Clinopyroxene at Very High Pressure: the Significance of Exsolved Rutile in Garnet," Earth Planet. Sci. Lett. 216, 591­601 (2003). 20. E. Roeder, Fluid inclusions Rev. Mineral. 12 (1984). 21. V. I. Kovalenko, I. P. Solovova, V. B. Naumov, et al., "Mantle Mineral Formation with Participation of Carbon Dioxide­Sulfide­Silicate Fluid," Geokhimiya, No. 3, 289­303 (1985). 22. I. P. Solovova, A. V. Girnis, I. T. Rass, et al., "Composition and Evolution of Fluid-Saturated Calcium-Rich Melts: Inclusions in the Minerals of Olivine Melilitite from Mahlberg, Rhine Graben," Geokhimiya, No. 9, 928­946 (2005) [Geochem. Int. 43, 843­861 (2005)]. 23. J. Hermann, "Experimental Evidence for DiamondFacies Metamorphism in the Dora-Maira Massif," Lithos. 70, 163­182 (2003). 24. H. Bureau and H. Keppler, "Complete Miscibility between Silicate Melts and Hydrous Fluids in the Upper Mantle: Experimental Evidence and Geochemical Implications," Earth Planet. Sci. Lett. 165, 187­196 (1999). 25. W. L. Huang and P. J. Wyllie, "Melting Relations of Muscovite Granite to 35 kbar as a Model for Fusion of Metamorphosed Subducted Oceanic Sediments," Contrib. Mineral. Petrol. 42, 1­14 (1973). 26. E. S. Persikov, "The Viscosity of Magmatic Liquids: Experiment, Generalized Patterns. A Model for Calculation and Prediction. Applications," in Advances in Physical Geochemistry, Ed. by L. Perchuk and I. Kushiro (Springer, New York, 1991), Vol. 9, pp. 28­30. 27. N. V. Sobolev and V. S. Shatsky, "Mineralogical Indicators of Ultrahigh Pressures and High Temperatures in Metamorphic Rocks," in Eclogites and Glaucophane Schists in Fold Areas (Nauka, Novosobirsk, 1989), pp. 83­106 [in Russian]. 28. H. L. M. van Roermund, M. R. Drury, A. Barnhoorn, and A. A. De Ronde, "Non-Silicate Inclusions in Garnet from an Ultra-Deep Orogenic Peridotite," Geol. J. 35, 209­229 (2000). 29. E. D. Mposkos and D. K. Kostopoulos, "Diamond, Former Coesite and Supersilicic Garnet in Metasedimentary Rocks from the Greek Rhodope: A New Ultrahigh-Pressure Metamorphic Province Established," Earth Planet. Sci. Lett. 192, 497­506 (2001). 30. Y. Ogasawara, K. Fukasawa, and S. Maruyama, "Coesite Exsolution from Supersilicic Titanite in UHP Marble from the Kokchetav Massif, Northern Kazakhstan," Am. Mineral. 87, 454­461 (2002). 31. N. Malaspina, J. Hermann, M. Scambelluri, and R. Compagnoni, "Multistage Metasomatism in Ultrahigh-Pressure
Vol. 46 No. 3 2008

GEOCHEMISTRY INTERNATIONAL


UNUSUAL INCLUSIONS IN GARNET FROM THE DIAMOND-BEARING GNEISS Mafic Rocks from the North Dabie Complex (China)," Lithos 90, 19­42 (2006). 32. S. L. Hwang, P. Shen, H. T. Chu, et al., "Kokchetavite: A New Potassium-Feldspar Polymorph from the Kokchetav Ultrahigh-Pressure Terrane," Contrib. Mineral. Petrol. 148, 380­389 (2004). 33. N. V. Sobolev and J. G. Lavrent'ev, "Isomorphic Sodium Admixture in Garnets Formed at High Pressure," Contrib. Mineral. Petrol. 31, 1­12 (1971).

303

34. M. Enami, B. Cong, T. Yoshida, and I. Kawabe, "A Mechanism for Na Incorporation in Garnet: An Example from Garnet in Orthogneiss from the Su-Lu Terrane, Eastern China," Am. Mineral. 80, 475­482 (1995). 35. R. Lu and H. Keppler, "Water Solubility in Pyrope to 100 kbar," Contrib. Mineral. Petrol. 129, 35­42 (1997). 36. W. Wang and E. Takahashi, "Subsolidus and Melting Experiments of a K-Rich Basaltic Composition to 27 GPa: Implication to Potassium in Mantle," Am. Mineral. 84, 357­361 (1999).

GEOCHEMISTRY INTERNATIONAL

Vol. 46

No. 3

2008