Initially, my work involved the determination of metamorphic pressure-temperature-fluid conditions in an effort to further our understanding of the role that fluids play in granulite facies metamorphism. The formation of the granulite facies is a topic that has generated some debate, in part because granulite formation plays an important role in the stabilization of continental crust. It has been argued that the infiltration of CO₂ into the lower crust is necessary for granulite formation. Alternatively, the stabilization of mineral assemblages that characterize the granulite facies could result from magmatic processes (e.g., partial melting). One of my earliest studies of the granulite facies involved the use of mineral equilibria to determine H₂O activities in rocks from the Southern Adirondack Mountains. This work confirmed the presence of peak metamorphic low H₂O activities, as calculated values were approximately 0.15.

The presence of reduced peak metamorphic H₂O activities in granulites could be an indication of high CO₂ activities. Thus, in order to test the CO₂-infiltration hypothesis I performed a number of thermodynamic calculations in the C-O-H system; these calculations show that a CO₂-rich fluid is not compatible with oxygen fugacities that are significantly more reducing than the graphite-CO₂ oxygen buffer. A review of the available oxygen fugacity data from granulites shows that at least some granulites did not form via CO₂ infiltration.

Another line of evidence, often cited in support of CO₂-streaming, is the presence CO₂-rich fluid inclusions in rocks from the granulite facies. We tested the hypothesis that fluid inclusions contain samples of the peak metamorphic fluid by comparing the compositions of the fluid found in inclusions with the compositions of metamorphic fluids as inferred from fluid-buffering mineral equilibria. Samples containing mineral equilibria that are inconsistent with CO₂-rich fluids often contained CO₂-rich fluid inclusions. This work shows that, in samples from the Adirondack Mountains, at least some CO₂-rich fluid inclusions do not contain samples of the peak metamorphic fluid. Instead, these inclusions were formed or modified after the peak of granulite facies metamorphism. More recently, I have used the presence of small amounts of carbonate minerals associated with CO₂-rich inclusions in granulites to show that the inclusions do not contain samples of the peak metamorphic fluid.

In general, the results from my work indicate that magmatic processes are important in the formation of Adirondack granulites. Furthermore, it was possible to use the results of the investigations of fluid inclusions to constrain the retrograde portion of the Adirondack P-T path.