CO2-H2O immiscibility to depths of > 30 km should be a direct result of the depression of isotherms that occurs during active transport of material down a subduction zone. Fluid phase separation in this environment should have a number of important consequences for the thermal, mechanical, and metamorphic development of subduction zones, yet the effect documenting the extent to which fluid immiscibility controls petrologic processes at high pressure. Particular attention will be paid to determining whether small-scale interlaying of blueschist and greenschist and blueschist and eclogite assemblages results from generation and preservation of fluid heterogeneities and whether such heterogeneities could be caused by fluid phase separation. The approach taken will be (1) to use mineral phase equilibria to calculate a (H20) and a (C02) ratios between adjacent layers both within facies types and across facies boundaries, (2) to determine fluid compositions directly from multiple generations of fluid inclusions, and (3) to model reaction pathways for rocks in both the immiscible and supercritical fluid regimes. Samples to be studied are from coherent high-pressure terranes on the Ile de Groix, in the Tauern Window, and on the island of Sifnos. The results should provide better constraints on the compositions of fluids present at high pressure and result in better models for devolatilization of subducting slabs.
