Sampling the petrographic fluid phase in subduction (i.e. HP metamorphic) environments is of importance for both constraining thermodynamical equilibria and transfer of elements, as subduction appears to be the key process for the recycling of both continental and oceanic crusts towards the mantle and the overlying crust, and a potential site for the circulation of fluids. Most of the studies on fluids and transfer of elements in metamorphic environments have dealt with dry and very HP rocks such as eclogites, or domains corresponding to relatively low-pressure (LP), such as the greenschist facies. In those studies concerned with metapelites, few have concentrated on blueschists facies rocks, which correspond to intermediate conditions along the subduction plane (~30-60 km). Barr (1990), as a preliminary study, reports predominant aqueous fluids and important variations of salinity, up to saturated brines.
Alpine metapelites at the front of the Schistes lustrés (penninic domain), have experienced a retrograde path which preserved at least some of the high-pressure fluid inclusions, therefore allowing us to analyse the assumed syn-crystallization fluid phase. In these rocks, characterization of the fluid phase is also critical because the prograde evolution of high-pressure low-temperature metamorphism in the metapelites leads to the crystallisation, at the expense of a protolith containing alkali, of alkali-free minerals like carpholite and lawsonite, mainly occurring in mesoscopic quartz (+ carbonates) veins, which are further transformed to alkali-bearing assemblages of the greenschist facies during retrogression. If fluid inclusions are thought to witness the related changes in the fluid phase composition, high alkali concentrations could be expected at high pressure, as proposed by Goffé & Vidal (1992).
Fluid inclusions assemblages were carefully selected in order to be representative of the different stages of the metamorphic evolution. Special care was taken to ensure an accurate location relatively to the tectonic setting at all scales. Peak metamorphism inclusions (isolated fluid inclusions not yet decrepitated, or coherent cluster-derived inclusion sets), transition fluids, evidenced by the healing of quartz with still present (metastable) quartz-hosted needles of carpholite, inclusions of the greenschist segregations, and, finally, inclusions in late (fragile) veins were compared. Analyses were performed using Microthermometry, Raman and Crush-leach techniques. Data reveal that these inclusions are essentially aqueous, therefore justifying the assumption a(H20)=1 in thermodynamical calculations. No CO2 could be detected, a conclusion in good agreement with studies on the stability conditions of lawsonite. The amount of dissolved species is generally small, being marginally higher than 5 weight % equivalent NaCl. Variations of the salinity are therefore small when comparing the successive quartz generations, but the sense of variation is coherent with what could be logically expected: higher salinities are encountered in the blueschist facies quartz than in those of the greenschist facies segregations. However, high-pressure fluids do not exhibit highly concentrated brines as could be expected by the alkali-free high-pressure assemblages.
Perhaps, large ratios of fluid:rock may have induced a considerable dilution of the fluids, as these high-pressure minerals are both very hydrated (12 %) and always associated to syn-metamorphic quartz segregations, whose formation has to be related to a large mass transfer. But geochemical studies led in the area (Henry et al., 1995) do not support such an interpretation. Alternatively, mass transfer towards adjacent terranes can be suspected, but is not conspicuous: no large metasomatic albitization of the overlying rocks, for example, could be evidenced. Rather, alkali certainly have remained in the rock. Chemical analyses and field observations point to the role of buffering of the fluid phase at the rock scale, and to the importance of phyllosilicates as a likely exchange medium.
Phyllosilicates might the sink for alkali cations at high-pressure, instead of the fluid phase as was previously thought. Indeed, in highly substituted phengites equilibrated at high-pressure conditions, interfoliar potassium content is high compared to that in LP (greenschist) alpine phengites. Paragonite, derived from albite during the prograde evolution, may well be stable at high-pressure conditions. In addition, structural observations show that the late albitization and veining of the Schistes lustrés themselves, always intimately relates to the syn- to post-greenschist tectonic evolution, in connection with local tectonic processes (shearing) and associated thermodynamical conditions. Although the fluid phase will probably enhance reaction rates, solid-solid reactions might therefore be the chief controller of the mass transfer of alkali in these metamorphic environments.
Barr, H., Min. Mag. 54, 159-168 (1990).
Goffé & Vidal, O., In Water-rock interaction (Kharaka & Maest eds.) 1499-1502 (Balkema, 1992).
Henry, C., Burkhard, M. & Goffé, B., J. Chem. Geol. (1995).