HR: 11:30h
AN: T41F-05    [PDF]
TI: Does the electrical conductivity of partially molten rocks depend on grain-boundaries?
AU: * Schilling, F R
EM: fsch@gfz-potsdam.de
AF: GeoForschungsZentrum Potsdam, Division 4 Telegrafenberg, Potsdam, D-14473 Germany
AB: In the active orogenic systems of the Central Andes and the Himalayas, extremely high electrical conductivities are observed and attributed to partially molten rocks within the crust. Laboratory experiments in combination with numerical calculations show that the conductivity behaviour of partially molten rocks can be explained by a complete interconnectivity of the melt phase for melt fractions $>$ 20 vol.%. Furthermore, a comparison of laboratory experiments and numerical modelling reveals an excellent agreement between observations and models. This holds true for Hashin-Shtrikman as well as a modified brick-layer model. In these models, only one conductivity mechanism is assumed for both, crystals and melt phase. However, there is experimental evidence that the conductivity mechanism in the solid matrix is dominated by electrons whereas ionic charge transfer dominates the conductivity in the melt phase. At the high frequencies applied in laboratory studies, the effect of charge transfer from crystals to melt will not - or slightly - affect the bulk-conductivity of the composite. However, at lower frequencies as used in field studies, the different charge transfer mechanisms in crystals and melt will lead to a blocking of charge transfer between solid and melt. This enhances the resistivity of partially molten rocks at low frequencies compared to laboratory studies using high frequencies. Therefore, grain-boundaries will influence the electrical conductivity of partially molten rocks. In other words, in partially molten rocks processes at grain-boundaries hinder a direct scaling of laboratory observations to field studies. However, the knowledge of the underlying mechanisms of charge transfer in crystals and melts allow an extrapolation of laboratory observations to field studies, if models are used which take the grain-boundary effects into account. Such a model will be presented and discussed.
DE: 0600 ELECTROMAGNETICS
DE: 0925 Magnetic and electrical methods
DE: 3210 Modeling
DE: 3630 Experimental mineralogy and petrology
DE: 3914 Electrical properties
SC: Tectonophysics [T]
MN: 2003 Fall Meeting