MODELING REACTIONS IN DOMAINS OF CONTRASTING COMPOSITION AND MINERAL ASSEMBLAGE IN HIGH-GRADE METAPELITIC ROCKS: AN EXAMPLE FROM THE WILMINGTON COMPLEX, DE-PA PIEDMONT, USA

We explore the origins of cm-scale domains of different bulk composition and mineral assemblage in granulite-facies metapelites from the Wilmington Complex, a Paleozoic volcanic/magmatic arc. The metapelites experienced thermal overprinting at different distances from Silurian gabbros: rocks farther from intrusions are characterized by lower-T phases (bt-sil-qtz) while samples closer to contacts contain less bt and more grt-crd-spn-crn and opx. Srogi et al. (1993, AJS 293) proposed a model of prograde reaction overstepping based on: 1) metastable preservation of reactant minerals; 2) adjacent domains with contrasting silica activity; 3) textural evidence for epitaxy and interface control; and 4) product assemblages in different domains that could not be produced along a single P-T path in a reduced-component-space Schreinemaker’s diagram. We use Theriak-Domino (T-D) to test these ideas and explore models for kinetic overstepping based on reaction affinity (e.g., Pattison et al., 2011, JMG 29). Observed assemblages (plag-ilm-grt-bt-sill-cord-sp) are reproduced on T-D models in samples farther from intrusions and indicate a narrow P-T range at 830-875°C and 570-650 MPa with liquid present. These rocks contain crd-spn-sil without crn and small pockets of pl-kfs-qtz intergrowths that suggest complete reaction of crn, a small degree of partial melting, and possibly closer approach to equilibrium. In rocks closer to intrusion contacts, anhedral crn porphyroblasts surrounded by crd-spn-sil suggest overstepping of crn-breakdown reactions initiated at T ~ 700-750°C. Assemblages in these higher-T rocks have not been reproduced on T-D sections for any whole-rock bulk composition used so far; some models predict minerals (olivine) or compatibilities (opx-spn and opx-crd) not observed in any samples. EDS analyses of entire thin sections and individual compositional domains will provide starting compositions for additional T-D models. Mineral modes and compositions will be used for further testing of model predictions. In addition, we will employ T-D sections showing components (e.g., SiO2), entropy, and reaction affinity to explore the roles that the kinetics of dissolution, nucleation, and diffusion play in generating domains of contrasting mineral assemblages in the high-T rocks.