29 Metamorphism in the Southeast and Beyond

Friday, 11 April 2014
8:00 AM - 10:00 AM
Duck Pond

CHEMICAL POTENTIALS: THE KEY TO UNLOCKING THE CONDITIONS OF GARNET NUCLEATION?

Paper number: 29-1
Program: Theme Sessions
Time: 8:00 AM – 8:20 AM

Authors

  • K.E. McCall Presenting Author
    Virginia Polytechnic Institute and State University, Blacksburg, VA
  • M.J. Caddick
    Virginia Tech, Blacksburg, VA
  • Besim Dragovic
    Virginia Polytechnic Institute and State University, Blacksburg, VA

Abstract

Porphyroblast nucleation has previously been shown to require a departure from thermodynamic equilibrium, but the specific mechanisms controlling this are still poorly understood. Here we examine the chemical potential of individual components that combine to nucleate a garnet crystal, calibrating their relative contributions. We propose that although, as typically assumed, the overall Gibbs free energy of the system can predict the growth of garnet at a given P and T, nucleation does not occur until the chemical potential of each garnet-forming oxide departs from it’s equilibrium (pre-nucleation) state. Previous studies (e.g. Gaidies et al., 2011; Pattison et al., 2011; Pattison and Tinkham, 2009; Waters and Lovegrove, 2002) looked at barriers to nucleation from a cumulative free energy standpoint, rather than this element-by-element basis.

We use published garnet crystal core compositions to determine where in P-T space they nucleated, and thermodynamic modeling to study the energetic consequences of suppressed nucleation. Sample choices were based on areas with a previously reported significant degree of reaction overstepping: two regionally metamorphosed terranes (Pomfret Dome, VT and Central Maine Belt, ME), one contact aureole (Nelson Aureole, B.C.) and one subduction zone (Cycladic Blueschist Unit, Sifnos). Two phase diagrams were calculated for each sample, one allowing all phases to form and a second prohibiting garnet and staurolite. Chemical potential and phase abundance differences between each were then examined. Results routinely show that a deviation in the chemical potentials of all garnet-forming elements does not occur at the equilibrium garnet-in reaction but instead coincides with the P-T of inferred crystal nucleation (based on core composition). Our models produce similar results for a wide variety of metamorphic settings, rock types, and apparent overstep amount, indicating that these deviations likely control the extent of overstep of garnet-forming reactions in many natural samples.