Electrical conductivity of orthopyroxene and plagioclase in the lower crust

The electrical conductivities of lower crustal orthopyroxene and plagioclase, as well as their dependence on water content, were measured at 6–12 kbar and 300–1,000°C on both natural and pre-annealed samples prepared from fresh mafic xenolith granulites. The complex impedance was determined in an end-loaded piston cylinder apparatus by a Solarton-1260 Impedance/Gain Phase analyzer in the frequency range of 0.1–10⁶ Hz. The spectra usually show an arc over the whole frequency range at low temperature and an arc plus a tail in the high and low frequency range, respectively, at high temperature. The arc is due to conduction in the sample interior, while the tails are probably due to electrode effects. Different conduction mechanisms have been identified under dry and hydrous conditions. For the dry orthopyroxene, the activation enthalpy is ~105 kJ/mol, and the conduction is likely due to small polarons, e.g., electrons hopping between Fe²⁺ and Fe³⁺. For the dry plagioclase, the activation enthalpy is ~161 kJ/mol, and the conduction may be related to the mobility of Na⁺. For the hydrous samples, the activation enthalpy is ~81 kJ/mol for orthopyroxene and ~77 kJ/mol for plagioclase, and the electrical conductivity is markedly enhanced, probably due to proton conduction. For each mineral, the conductivity increases with increasing water content, with an exponent of ~1, and the activation enthalpies are nearly independent of water content. Combining these data with our previous work on the conductivity of lower crustal clinopyroxene, the bulk conductivity of lower crustal granulites is modeled, which is usually >~10⁻⁴ S/m in the range of 600–1,000°C. We suggest that the high electrical conductivity in most regions of the lower crust, especially where it consists mostly of granulites, can be explained by the main constitutive minerals, particularly if they contain some water. Contributions from other highly conducting materials such as hydrous fluids, melts, or graphite films are not strictly necessary to explain the observed conductivities.