Dielectric Properties of Liquid-Impregnated Porous Solids

Dielectric spectroscopy is a versatile tool that can often be used as a non-destructive, indirect way of measuring other important properties (such as content of crude oil or degree of curing of a resin). The dielectric properties of materials are thus of interest in several circumstances and not only for materials used for electrical purposes (such as electrical insulation). In this work, the dielectric properties have been studied for porous solids, impregnated with a liquid. This is an often encountered system that is seldom studied in detail due to two intrinsic difficulties: (a) the geometry of the pores is complex and (b) there may be effects due to the interaction between the solid and the fluid at the interface. The studies have been both theoretical and experimental. I have mostly studied two different kinds of idealised systems. The first type consists of sand beads, glued with small amounts of epoxy. The second type consists of polypropylene beads, that have been sintered to form a porous material. These porous materials have been impregnated with salty water or an insulation liquid (a mixture of mono- and dibenzyl toluenes) to which ions have been added. Although the dielectric dispersion of the constituent materials can be neglected in the frequency range used, measurements show that two dielectric relaxations arise in the "bulk" heterogeneous material. At high frequencies, a so-called Maxwell-Wagner relaxation arises. At low frequencies, another relaxation has been found. This relaxation is attributed to diffusion effects in the electrochemical double-layer. For frequencies above the low-frequency relaxation, an interface conductivity was obtained, which increases the strength of the Maxwell-Wagner relaxation. The dependence on temperature and of pore liquid conductivity for this low-frequency relaxation has been studied. Models for porous materials have been examined and further extended in order to be more general. The so-called Grain Consolidation Model has been found to be especially versatile. Even if the models give a simplified picture of reality, their predictions have fair agreement with experimental results. An extension of so-called Differential Effective-Medium theory was found to yield good agreement with experimental data for the Maxwell-Wagner relaxation. The low-frequency relaxation, and the interface conductivity has been found to be related to the microgeometry through a parameter, L, that can be estimated from electrical measurements and measurements of the fluid permeability