Program > Browse abstracts by speaker > Yang Xiaozhi

Laboratory electrical conductivity measurements of a melt or a partially molten system are key to understanding the electrical structure of Earth's interior, as well as the relevant chemical, physical and dynamical properties. This is owing to the potentially significant role of a melt in enhancing the bulk conductivity. An issue of critical importance is if and how laboratory-yielded data can be scaled to natural environments. The effect of a melt on the bulk conductivity is determined by its distribution and connectivity in the system, which are in turn controlled by edges and corners of grain boundaries that are sensitive to grain sizes of the constitutive minerals. For an ideal system of identical grains, ordered packing and complete grain wetting by melt, the melt effect on the bulk conductivity is independent of mineral grain size. Such ideal conditions are rarely met in the realistic Earth, and the effect of mineral grain size has not been studied.

We experimentally measured the electrical conductivity of a melt-olivine mixture, using a highly conductive and mobile carbonatite melt analog and olivine of contrasting grain sizes. We carefully prepared the samples for textural equilibrium. We demonstrate that, at otherwise similar conditions, the bulk conductivity decreases with increasing olivine grain size. We model the effect of mineral grain sizes on the electrical conductivity of a melt-mineral system. We show that, to generate a fixed bulk conductivity, the effect of partial melt is overestimated when mineral grain size effect is considered. We discuss the implications for the electrically conductive asthenosphere. The grain size effect on the bulk electrical conductivity should be taken into account for any other melt- or fluid-bearing systems in the crust and mantle.