Fluids in High-Pressure Granulites

Abstract

Properties of fluids in high-pressure granulites were studied in HP granulites (~8.7–11 kbar, ~800–900°С) and syngranulite fluid infiltration-driven HP metasomatites (~11–9 kbar, ~920–850°С) from the Lapland granulite belt of the Fennoscandian Shield. The study involved large-scale mapping of the rocks, microthermometry of mineral-hosted fluid inclusions, multiequilibrium mineral thermobarometry, and calculations of H2O activity based on mineral equilibria. The mafic pyroxene granulites and syngranulite metasomatites (quartz blastomylonites with orthopyroxene, sillimanite, and garnet; veins and vein-like bodies of orthopyroxene–garnet and diopside–scapolite rocks) contain similar assemblages of syngenetic fluid inclusions (which are hosted mostly in quartz and also in garnet, orthopyroxene, and scapolite) of contrasting chemical composition: nearly pure СО2 (distinctly predominant), brines (the dominant salts are CaCl2 and NaCl), and N2 ± H2O. These three types of inclusions coexist in the same generations of early inclusions: rarer primary (p) and predominant primary–secondary (ps). The CO2 inclusions have either high or low densities, and the N2 inclusions are of low density. The brine inclusions show a wide range of total salt contents (up to 30–35 wt%) and variable concentration proportions of the dominant salts: p-inclusions with a salinity of 20 wt% CaCl2 + 10 wt% NaCl; ps-inclusions with a salinity of 5 wt% CaCl2 + 20 wt% NaCl; p- and ps-inclusions with a salinity of 5–23 wt% NaCl eq; and p-inclusions with halite (up to 35 wt% NaCl). In general, CaCl2 is the predominant salt component in the early p- and ps-inclusions of the rocks. Considered together, currently available data (including Sr, Nd, and O isotope systems) on these rocks indicate that the external fluid flow during the origin of granulites was evidently of mantle origin. At the peak P–T parameters, the inclusions were entrapped from a heterogeneous fluid in which immiscible water–salt and CO2-rich fluids, which initially contained N2, coexisted. Data on the chemical composition and salt concentrations of the fluids, $${{a}_{{{{{\text{H}}}_{2}}{\text{O}}}}}$$ = 0.40–0.51, are compared with the theoretically predicted phase state of the fluids and the properties of the coexisting immiscible fluid phases at the estimated P–T parameters of granulite petrogenesis on the basis of numerical models in the H2O–CO2–NaCl and H2O–CO2–CaCl2 ternary systems. The location of the tie-lines and solvus were calculated to subsequently use for the thermodynamic prediction. The paper discusses similarity and the reasons for the difference between the theoretical compositions of the generated fluid phases and the composition of fluid inclusions, geochemical consequences of the heterogenization of granulite fluids (the formation of concentrated alkaline brines and a potentially acidic CO2-rich fluid phase, the values of the mass and volume fractions of these phases depending on variations in the composition of the initial homogeneous fluid, etc.). It follows that an extensive region of the compositions of aqueous fluids with different concentrations of CO2 and Na and Ca chlorides exists at the P–T parameters of HP granulites in which originally homogeneous fluid splits into compositionally contrasting fluid phases with different properties. This region of coexisting immiscible fluids significantly expands with increasing CaCl2 concentration. Hence, the lower crust at the level of the HP granulite facies may be the region where high-temperature immiscible fluids are generated. One of these fluids is a denser phase of alkaline brines, and the other is a less dense potentially acidic phase of H2O–CO2 fluids rich in CO2. Ascending along regional permeable zones, these fluid phases of deep origin can play an important role in magmatic, metamorphic, metasomatic, and ore-forming processes in the middle and upper crust.