Metamorphic CO2 Production in Collisional Orogens: Petrological Constraints from Phase Diagram Modeling of Himalayan, Scapolite-bearing, Calc-silicate Rocks in the NKC(FIMAS(T)-HC system

A reliable quantitative estimate of the metamorphic CO2 flux from collisional orogens is fundamental to our understanding of the deep carbon cycle, but it is still far from being constrained. Among major uncertainties are the poor knowledge of the nature of metamorphic CO2-producing processes and the amount of CO2 potentially released through these reactions. Previous studies of metamorphic decarbonation reactions in metacarbonate rocks mainly used simple model reactions between end-members in simplified model systems. However, fully quantitative modelling of calcsilicate rocks requires an investigation of very complex systems with more than six components. Moreover, scapolite solid solution has been rarely included in previous studies, although this mineral is often a major constituent of calc-silicate rocks. This study focuses on (1) the CO2-producing processes occurring in scapolite-bearing calc-silicate rocks and (2) the discussion of a methodological approach suitable to understand and quantify these processes. In this framework, phase relations and devolatilization reactions in the Na20 K2O Ca0)(Fe0) Mg0 A1203 Si02 (Ti02)H20 0O2 INKC(F)MAS(T)HCJ system are considered, with application to high-grade clinopyroxene + calcite K-feldspar scapolite plagioclase+ zoisite calc-silicate rocks from the Himalaya. The NKC(F)MAS(T)-HC equilibria involving scapolite and plagioclase solid solutions are investigated using (1) isobaric T X(CO2) phase diagram sections and pseudosections and (2) mixed-volatile P T phase diagram projections. This phase diagram approach allowed us to identify scapolite-bearing, CO2-producing, univariant (i.e. isobaric invariant) equilibria that have never been recognized before, and that could not be detected without considering Na Ca solid solutions in the calculations. It is demonstrated that the investigated calc-silicate rocks behaved as a nearly closed, internally buffered, system during prograde metamorphism and that most of the observed key microstructures correspond to isobaric univariant or invariant assemblages. In such a nearly closed system, the fluid was mostly produced during prograde heating at the isobaric invariant points, where abrupt changes in mineral modes also occurred. The proposed phase diagram approach further allows quantitative estimation of the amount and composition of the fluid produced at such isobaric invariant points. On average, 2,5 mol of CO2 (110 g) per 1000 cm(3) of reacting rock were produced during prograde metamorphism of this calc-silicate rock-type. Because similar scapolitebearing calc-silicate rocks are abundant in the Himalayan orogen, it is suggested that this calc-silicate rock-type might have produced large amounts of CO2-rich fluids during Himalayan metamorphism. A preliminary estimate of these amounts at the scale of the whole orogen suggests a total metamorphic CO2 production of (2-7) x 10(17) mol, corresponding to (1-3) x 10(10) Mt of CO2. Integrated over 20 Myr (i.e. the maximum duration of prograde metamorphism), the