Disequilibrium melt distributions during static recrystallisation

DISEQUILIBRIUM MELT DISTRIBUTIONS DURING STATIC RECRYSTALLISATION N.P. Walte (1), P.D. Bons (2), C.W. Passchier (1), D. Koehn (1), J. Arnold (1) (1) Institute for Earth Sciences, Johannes Gutenberg-University, Mainz, Germany, (2) Institute for Earth Sciences, Eberhard Karls University, Tübingen, Germany (walte@mail.uni-mainz.de) The geometry of melt-filled pores in a partially molten rock strongly controls the permeability, rheology and initial segregation of melt. Current theory for monomineralic aggregates, using only the wetting angle and melt fraction as parameters, predicts a perfectly regular melt framework or equally shaped melt inclusions on grain boundary junctions. However, published melt-present high-temperature experiments with rock forming minerals such as quartz or olivine show considerable deviations from this predicted regular equilibrium melt geometry. Disequilibrium features, such as fully wetted grain boundaries, melt lenses, and large melt patches have been described, and were attributed to surface energy anisotropy of the minerals. This study used static analogue experiments with norcamphor plus ethanol liquid, that allow continuous in-situ observation of the evolving distribution of melt during static recrystallisation. The liquid-crystal surface energy of norcamphor is effectively isotropic. For the experiments an approximately 0.1 mm thin sample of norcamphor plus ethanole was placed between two glass plates and observed with a miroscope. Ethanol was used as a melt analogue because it allows to run experiments at room temperature, avoiding any temperature gradients. The wetting angle is approximately 15°, which is well below 60° and within the range reported for quartz and olivine plus melt experiments. The experiments show that all described disequilibrium features can form during fluid-enhanced static recrystallisation, especially where surrounding grains consume small, few-sided grains. These features are unstable and transient: a fully wetted grain boundary may, for example, evolve to a trapped melt lens. Our experiments demonstrate that the effect of static recrystallisation alone suffices to explain the deviations from predicted ideal melt distributions. There is no need to invoke surface energy anisotropy, although this would enhance the effect. Current wetting-angle-based theory is therefore not sufficient to predict permeability and transport of melt or fluids and the rheology of partially molten rocks.