Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.geol.msu.ru/deps/petro/Perchuk_publ/DubininaPerchuk_Doklady12_oxygen.pdf Äàòà èçìåíåíèÿ: Tue Apr 2 15:36:52 2013 Äàòà èíäåêñèðîâàíèÿ: Thu Feb 27 23:31:58 2014 Êîäèðîâêà:

ISSN 1028 334X, Doklady Earth Sciences, 2012, Vol. 444, Part 2, pp. 734­737. © Pleiades Publishing, Ltd., 2012. Original Russian Text © E.O. Dubinina, A.L. Perchuk, O.S. Korepanova, 2012, published in Doklady Akademii Nauk, 2012, Vol. 444, No. 5, pp. 534­538.

GEOCHEMISTRY

Oxygen Isotopes Effects due to Dehydration of the Blueschist: the Experimental Data under P­T Conditions of Subduction Zone
E. O. Dubininaa, A. L. Perchukb, c, and O. S. Korepanovab
Presented by Academician V.V. Yarmolyuk February 27, 2012 Received February 20, 2012

DOI: 10.1134/S1028334X12060128

The oxygen isotope composition of minerals from metamorphic rocks carries information on the nature of the protolith, the temperature of the metamorphism peak, and the isotope composition and content of the fluid phase participating in the metamorphic process. Among the main sources of fluid in high grade meta morphism are hydrous minerals, for example, those delivered to the subduction zone with sedimentary and metamorphic rocks of the oceanic slab. The process of dehydration should result in the oxygen isotope shift in the dehydrated rock (restite), the surrounding crustal rocks, and the overlying rocks of the mantle wedge. Currently it is generally accepted that the oxygen isotope effect under metamorphic dehydration is very low (<1) [1]. Moreover, small isotope shifts become poorly reflected against the background of the variable lithology of rocks in metamorphic complexes. More and more rocks with anomalously light or widely ranging oxygen isotope composition are found in granulitic and eclogitic metamorphic complexes. Signif icant oxygen isotope lightening (18O up to ­11) was observed in eclogite and gneiss of the Dabie­Sulu belt in China; moderate oxygen isotope lightening (up to ­4), in eclogite of the Kokchetav massif; and extreme oxygen isotope lightening (18O from ­16 to ­25), in Paleoproterozoic gneiss of the Belomorian Complex in Karelia [2­4]. The authors interpreted these unusual compositions as attributes of the pro tolith, which was hydrothermally altered before meta morphism with participation of meteoric waters under the conditions of a cool paleoclimate. Certainly, it is problematic to explain the reason for preservation of such contrasting isotope characteristics in rocks,
Institute of Geology of Ore Deposits, Petrography, Mineralogy, and Geochemistry, Russian Academy of Sciences, Moscow, Russia b Institute of Experimental Mineralogy, Russian Academy of Sciences, Moscow oblast, Chernogolovka, Russia c Geological Faculty, Moscow State University, Moscow, Russia e mail: delta@igem.ru
a

which underwent several stages of prograde metamor phism and metamorphic recrystallization. However, it is also very difficult to explain the appearance of hydrous fluid with the isotope composition typical of atmospheric precipitates of high latitudes under the conditions of high pressure metamorphism. An alternative explanation for the appearance of anomalous isotope markers in high grade metamor phic rocks is provided by significant isotope shifts at the expense of kinetic fractionation between mole cules of the hydrous fluid and the solid phases [1], in mineral reactions proceeding in the same directions and with higher rates. In this relation, of special inter est are the reactions of dehydration of hydrous miner als, which are one of the main suppliers of hydrous fluid in subduction zones. There are a few experimen tal investigations on the behavior of the 18/16 ratio under dehydration [5, 6], and all of them were per formed at low pressures ("evaporation into a vac uum"). The most detailed investigation of kinetic frac tionation was performed for serpentine at a tempera ture from 300 to 600°C, for which an oxygen isotope lightening of hydrous fluid in relation to residual ser pentine by 27 was obtained [6], and this effect prac tically did not depend on temperature. However, it is difficult to relate these experiments to the subduction environment, since they were carried out at a pressure of <1 bar, and the influence of released hydrous fluid on rock was not studied in them. Experimental and empirical data on the scale and nature (kinetic or equilibrium) of oxygen isotope fractionation resulting from the processes of dehydration at high pressures are currently absent. This study is aimed at experimental estimation of the possible oxygen isotope effect caused by dehydra tion of hydroxyl bearing minerals under P­T param eters (25 kbar, 720­1000°C) of the subduction envi ronment. In contrast to the experiments performed by the method of evaporation into a vacuum [6], the high pressure experiment does not provide the techni

734


OXYGEN ISOTOPES EFFECTS DUE TO DEHYDRATION OF THE BLUESCHIST

735

cal possibility to collect the released hydrous fluid, at least with a concentration sufficient for isotope analysis. We performed the experiment by the "sandwich" method, which models the process of fluid generation from minerals of the "cool" oceanic crust (glau cophane schist) and the interaction of released fluid with the "hot" mantle wedge (olivine). Analysis of the material after the experiment in the zone of glau cophane schist may provide evidence for isotope shifts resulting from dehydration and in the zone represented by olivine, on the effects resulting from the influence of released fluid on the material of the mantle wedge. The condition of the open system in relation to fluid was met, which is an approximation to the conditions of the influence of fluid flow formed due to dehydration of the subducted oceanic crust on mantle matter. MATERIALS AND METHODS The experiment was performed at the Institute of Experimental Mineralogy, Russian Academy of Sci ences. A high pressure piston cylinder apparatus equipped with a saline cell with a diameter of 1.27 cm and graphite heater was used for the experiment. The starting materials were loaded into the Pt capsule by layers: glaucophane schist was placed at the bottom and an olivine layer was placed on it. To realize the condition of the open system in relation to the fluid, the capsule was hermetically welded in its lower part, whereas its upper part was wrapped (without welding). The starting materials loaded in the Pt ampoule were heated for 1 h at 130°C before the experiment in order to remove adsorbed atmospheric water. The ampoule was loaded into the cell vertically and placed into a "cover" of soft ceramics to prevent deformation. The starting materials were represented by glau cophane schist (model analog of the crust) from the Atabashi Complex (Tian Shan) composed of glau cophane (95%) and secondary minerals (albite, cal cite, phengite, chlorite, and winchite), on the one side, and chemically homogeneous olivine (model analog of the mantle) of jeweler quality (#Mg = 0.93, Akhaim quarry, Norway). The chemical analysis of the starting glaucophane schist demonstrated that the concentration of water in it did not exceed 11%. The experiment was performed by the original methodology [7, 8] at a pressure of 25 kbar with a maintained temperature of 1000°C in the upper part of ampoule (Ol zone) and a calculated temperature of 720°C in the lower part (Gls zone). The P­T condi tions at the boundary between the two materials in the ampoule corresponded to the crust/mantle boundary in the Cascadian subduction zone, northwestern part of the United States [9]; i.e., the process of interaction between the cool crust and the hot mantle under the conditions of so called hot subduction was modeled. The duration of the experiment was 168 h. The temperature in the upper part of the ampoule was measured and controlled by WRe5/WRe20 ther
DOKLADY EARTH SCIENCES Vol. 444 Part 2 2012

(a)

Ol

Gls

1 18O, 16 14 12 10 8 6 4 2 1 2

2

3

4 (b)

5 1 mm

6

7

Starting schist

Ol

Gls

Starting olivine

3

4

5

6 7 Zone number

Fig. 1. Results of experiments. (a) Ampoule after experi ment; the boundary between materials is indicated by arrows and a dotted line; numerals denote the numbers of zones distinguished for the oxygen isotope analysis; (b) results of measurements of the oxygen isotope composi tion in distinguished zones; horizontal dashed lines denote the oxygen isotope composition of the starting materials; vertical dashed lines denote the boundaries between the zones; the dashed line between zones 4 and 5 indicates the contact between olivine and glaucophane schist.

mocouples with an accuracy of ±3°C. The tempera ture at the base of the ampoule was calculated by the methodology of [10]. The possibility of calculative tem perature estimations at the bottom of the ampoule (with an accuracy of ±20°C) was checked in the calibration run with two thermocouples located at the levels of the cell corresponding to the position of the bottom and upper part of the ampoule [7]. Oxygen fugacity in runs was controlled by a graphite heater maintained at a level slightly lower than the NNO buffer [11]. After the experiment, the ampoule was cut along by a diamond saw into two equal parts. One part (Fig. 1a) was polished and used for investigations on the micro probe and scanning electron microscope; the other part was divided into 7 fragments separated from each other by quench fractures. These fragments were used for the oxygen isotope analysis.


736

DUBININA et al.

Gl Gl Omp + Q

(1, n = 18); the value 18O = 5.25 ± 0.11 (1, n = 20) was obtained for olivine of San Carlos. As a whole, the error of measurements of experimental materials (1) shown in the diagram (Fig. 1) did not exceed 0.1. MINERAL REACTIONS AND OXYGEN ISOTOPE SHIFTS IN RUN PRODUCTS The zone of glaucophane schist contains alter ations resulting from dehydration decomposition of glaucophane, which may be expressed as the general ized reaction: glaucophane omphacite + SiO2 ± Ca­Na amphibole + hydrous fluid. The content of omphacite and quartz decreases towards the contact, being replaced by Ca­Na amphibole of the winchite composition. Mapping of the spatial distribution of mineral phases in run prod ucts on a microprobe demonstrates that the products of glaucophane decomposition occupy <10% of visual area of the sample (Fig. 2). The formation of the orthopyroxenite layer is registered at the base of the olivine zone, which is controlled by the reaction olivine + SiO2 (in fluid) = orthopyroxene. It is suggested that this layer may play the role of a filter or screen for melts or solutions coming to the mantle wedge from the subducting slab [8]. The oxygen isotope composition of glaucophane schist (Gls) and olivine (Ol) measured before the experiment was 18Gls = 12.62 ± 0.06 (n = 6) and 18Ol = 5.43 ± 0.10 (n = 6). Significant isotope shifts occurred in both materials during the experi ment (Fig. 1). Isotope lightening of Gls and isotope weighting of Ol in relation to the starting compositions is observed only in the near contact zones (4 and 5 in Fig. 1). It is not excluded that these effects are con trolled by partial capture of material from the neigh boring zone. Growth of the 18Gls with synchronous decrease of 18Ol values occurs at a distance from the contact (zones 6 and 7 in Gls and zones 1­3 in Ol); the absolute values of shifts increase with distance from the contact. As a result, the 18Gls value at the "cool" edge of the ampoule exceeds the starting value by 2.4, whereas at the "hot" end the 18l value becomes lower than the starting value by 2.0. DISCUSSION The evolution of the 18 values in both phases towards the isotope equilibrium between glaucophane and olivine at high temperatures, which should range from 1.36 to 1.95 at a temperature from 720 to 1000°C, is observed only in the Ol­Gls contact area (zones 4 and 5) [13, 14]. If only the process of isotope equilibration of two components (olivine and glau cophane) proceeds in experiment, then the values of 18Gls and 18l should tend to their starting compo sitions with distance from the contact zone. However,
DOKLADY EARTH SCIENCES Vol. 444 Part 2 2012

50 m
Fig. 2. Local areas of omphacite (Omp) with quartz (Q) formation after glaucophane by the reaction of dehydra tion in the run products.

METHODOLOGY OF THE OXYGEN ISOTOPE ANALYSIS The oxygen isotope analysis was performed by the method of fluoration using laser heating [12] at the Institute of Geology of Ore Deposits, Petrography, Mineralogy, and Geochemistry, Russian Academy of Sciences, Moscow. Each fragment of run products was preliminarily weighed. Fragments with a weight exceeding 3 mg were divided into pieces (1­2 mg each), which were weighed and analyzed separately. As a result, the mid weight oxygen isotope composition was calculated for each of the experimental zones dis tinguished. Laser heating was performed with a 30W CO2 laser in an atmosphere of bromine pentafluoride. The methodology of heating by defocused laser beam was applied by analysis of the starting glaucophane schist (Gls), which was homogenized by pounding for the experiment. The isotope analysis of Gls by the method of volumetric fluoration was performed for the control; this method provided results identical within the error for both methods (12.62 ± 0.06 at fluoration with laser heating and 12.5 ± 0.3 by the method of volumetric fluoration). Measurements of the 18O/16O ratios in oxygen extracted from the samples were performed on a DELTAplus mass spectrometer (Finnigan) in the regime of double blousing. All 18 values in this study are repre sented in the generally accepted scale V SMOW ():

18 - 16 O O V SMOW s a mp l e 18 O = â 1000. 18 O 16 O V SMOW All measured values were calibrated in the V SMOW scale in relation to the international standards of quartz NBS 28 and garnet UWG 2. Measurements of the international standard of biotite NBS 30 within the calibration provided a value of 18O = 5.09 ± 0.17

18 O 16 O


OXYGEN ISOTOPES EFFECTS DUE TO DEHYDRATION OF THE BLUESCHIST

737

the zones distant from the contact area (zones 6 and 7 in Gls and zones 1­3 in Ol) are characterized by diver gence of the 18Gls and 18l values, which is impos sible in the equilibrium process and results in increase of the oxygen isotope shift 18Gls­18l, but not its decrease towards the glaucophane­olivine isotope equilibrium [13, 14]. Most likely the obtained divergence of the 18Gls and 18l compositions in the experiment results from the process of isotope redistribution between the phases with participation of fluid. The role of the ther mal gradient, which strengthened removal of fluid from the ampoule in addition with possible ther moduffusion, is still not clear in this process. Melts of basaltic composition [15] were characterized by oxy gen isotope partitioning with a gradient of 1.5 per 100°C under the conditions of a closed system within the temperature field of 1380­1530°C. The isotope effects for oxygen obtained in these experiments at the expense of thermodiffusion are comparable with our data by the value. Thermodiffusive isotope partition ing in gases, solutions, and melts is a well known pro cess, but it occurs in homogeneous media [15]. Our experiment was performed in the heterogeneous medium at lower temperatures, and melting was not registered in its products (also due to the presence of a small 2 content in the fluid, which decreases the temperature of the hydrous solidus of the system). This fact does not allow us to consider thermodiffusion as the main factor of the observed isotope effect. We con sider the assumption that divergence of the 18Gls and 18 l values was controlled by the influence of hydrous fluid separated from glaucophane schist on olivine to be more substantiated. The positive isotope shift occurring during glau cophane dehydration provides evidence for the fact that hydrous fluid should have a light oxygen isotope composition. Using the data on the value of the coef ficient of equilibrium fractionation in the glau cophane­water system [14], we may demonstrate that in the case of isotope equilibrium of hydrous fluid with glaucophane schist, the 18 value of fluid should be +14.2 (at = 720°, 103ln(Gl­W)= ­1.6). Most likely fluid with such an oxygen isotope composition cannot lead to isotope lightening in olivine, even in spite of the fact that the olivine­water fractionation in the hot zone of the ampoule should be the opposite (at 1000° 103ln(Fo­W) = ­2.8 [13]). Removal of fluid with such an oxygen isotope composition from glaucophane schist might result in the isotope lighten ing of glaucophane schist. However, this is not observed. Thus, we may suggest that fluid was not in isotope equilibrium with glaucophane schist and had a lighter oxygen composition in comparison with the isotope equilibrium. Preliminary balance calculations demonstrate that the 18 value in hydrous fluid should be negative, and the kinetic isotope effect upon dehydration should be comparable in scale with that obtained under dehydra
DOKLADY EARTH SCIENCES Vol. 444 Part 2 2012

tion into a "vacuum" [6] or exceed it. Further experi mental investigations should clear up the absolute val ues of this effect. However, even at this stage we may make a conclusion on the kinetic nature of the oxygen isotope fractionation with dehydration under the con ditions of high temperatures and pressures. The data obtained originally demonstrate that the reactions of dehydration may result in significant shifts by the oxygen isotope composition in dehydrated rocks, as well as in rocks influenced by released fluid. Most likely dehydration of hydrous minerals under the conditions of a thermal gradient and open system may result in kinetic isotope fractionation between the fluid and the restite rock. This process in the subduction zones may lead to the formation of fluid with a "light" oxygen isotope composition at the expense of decom position of hydrous minerals of the oceanic crust, which may change the isotope markers of rocks of the subducting slab, as well as the material of the mantle wedge. An analogous process may be realized in rocks of granulitic complexes, which results in the appear ance of isotope anomalies without participation of meteoric waters in the protolith history. ACKNOWLEDGMENTS This study was supported by the Russian Founda tion for Basic Research, project nos. 12 05 00860, 09 05 01217, and 12 05 01093. REFERENCES
1. L. P. Baumgartner and J. W. Valley, Rev. Mineral. Geochem. 43, 415 (2001). 2. D. Rumble and T. F. Yui, Geochim. Cosmochim. Acta 62, 3307 (1998). 3. Y. F. Zheng, B. Fu, B. Gong, and L. Li, Earth Sci. Rev. 62, 105 (2003). 4. I. N. Bindeman, A. K. Schmitt, and D. A. D. Evans, Geology 38, 631 (2010). 5. J. P. Girard and S. M. Savin, Geochim. Cosmochim. Acta 60, 469 (1996). 6. R. N. Clayton and T. K. Mayeda, J. Phys. Chem. 113, 2212 (2009). 7. A. L. Perchuk and O. S. Korepanova, Moscow Univ. Geol. Bull. 66 (4), 250 (2011). 8. A. L. Perchuk, O. S. Korepanova, and V. O. Yapaskurt, Dokl. Earth Sci. 437 (1), 393 (2011). 9. E. M. Syracuse, P. E. van Keken, and G. A. Abers, Phys. Earth Plan. Inter. 183, 73 (2010). 10. F. Schilling and B. Wuender, Eur. J. Mineral. 16, 7 (2004). 11. R. B. Merrill and P. J. Wyllie, Geol. Soc. Am. Bull. 86, 555 (1974). 12. Z. D. Sharp, Geochim. Cosmochim. Acta 54, 1353 (1990). 13. Y. F. Zheng, Geochim. Cosmochim. Acta 57, 1079 (1993). 14. Y. F. Zheng, Earth Planet. Sci. Lett. 120, 247 (1993). 15. F. M. Richter, E. B. Watson, R. Mendybaev, et al., Geochim. Cosmochim. Acta 73, 4250 (2009).