Forward modeling of calc-silicate microinclusions and fluid evolution of a graphitic metapelite (NE Sardinia)

The primary microinclusion assemblage anorthite +/- margarite +/- epidote occurs within garnet and rutile in a graphitic metapelite from northeast Sardinia. The existence of the microinclusion assemblage and the observed paragenetic sequence and mineral compositions for the entire rock are in good agreement with phase relations predicted from thermodynamic data for the system CaO-Na2O-K2O-MgO-FeO-Al2O3-SiO2-TiO2-C-O-H. The complexity of the microinclusion phase relations is reduced by the specification of a component saturation hierarchy that permits thermodynamic projection through phases of variable composition. The microinclusion assemblage is predicted to be stable over the pressure and temperature intervals of 6585-6685 bars and 550-575-degrees-C, with a C-O-H fluid of composition X(o) = 0.46-0.57 (ca. 26-45 mol% CO2), where X(o) represents the atomic fraction of total O relative to O + H in the fluid phase. The assemblage anorthite + potassium feldspar + albite + muscovite + quartz, present within other garnets, requires an even more CO2-rich fluid composition with X(o) > 0.9 (ca. 89 mol% CO2). With progressive metamorphism, fluid composition was buffered toward more H2O-rich fluid compositions by the minute calc-silicate domains represented by the microinclusions. The CO2 fluid compositions represent a significant deviation from those generated by simple dehydration in the presence of graphite (X(o) = 1/3) and suggest that an additional source of O was contributed to the fluid, most probably as CO2. Because the buffering capacity of the calc-silicate domains would be overwhelmed by any significant fluid influx, an external source of CO2 is improbable. No internal source of CO2 is preserved, but the calcic character of the inclusions and the high-X(o) fluid compositions could be explained by carbonate in the metapelite protolith. The calc-silicate inclusions record an interval of the metamorphism during which low rates of fluid production and infiltration inhibited bulk equilibration. This interval was terminated by an episode of dehydration that introduced more voluminous quantities of fluid into the rock. These results suggest that fluid from adjacent rocks had not permeated the metapelite when the inclusions formed and provide an argument against the importance of pervasive fluid movement during a stage of the prograde metamorphism.