Modeling of the permittivity of epoxy nanocomposites

Abstract

Polymer nanocomposites exhibit a number of desirable dielectric properties, like increased resistance to partial discharge, electrical treeing, dc breakdown strength, or a decreased intake of space charges. The permittivity is one of the properties at which a direct impact from the introduction of nanoparticles can be observed. Many applications would benefit from the ability to tailor the permittivity of their insulation. Both a reduction, for e.g. cable systems and GIS spacers, and an increase, for e.g. capacitors, would be desirable if they can be controlled and exploited accordingly. Hence there is an interest in predicting how a filler material would alter the permittivity. There are two main groups of theoretical approaches to the problem of composite permittivities. Effective medium theories, which utilize average fields or polarizabilities and induced dipole moments, and integral methods, which use low concentration formulae and integrate them to higher concentration. When modeling the permittivity of nanocomposites both methods struggle to deliver satisfying results. This exploratory work tries to make estimations about the bulk permittivity by taking into account the host-polymer interface, to which many of the astounding properties of polymer nanocomposites are ascribed. As a starting point for the modeling, nanocomposites are considered which are based on bisphenol-A type epoxy resin with aluminum oxide and magnesium oxide nanofiller. The selected filler materials are spherical or quasi-spherical, to minimize the influence of additional parameters like the aspect ratio.