Abstract
It is well established that dramatic increases in conductivity occur upon the addition of conductive filler materials to highly resistive polymeric matrices in experimental settings. However, the mechanisms responsible for the observed behavior at low filler loadings, below theoretical percolation limits, of even high aspect ratio fillers such as carbon nanotubes (CNT) are not completely understood. In this study, conductive composites were fabricated using CNT bundles dispersed in epoxy resins at diverse loadings, using different dispersion and curing protocols. Based on electron microscopy observation of the CNTs strands distribution in the polymeric matrices and the corresponding electrical conductivities of those specimens, we concluded that no single electron transfer model can accurately explain the conductive behavior for all the loading values. We propose the existence of two different conductive mechanisms; one that exists close to the percolation limit, from 'low loadings' to higher CNT contents (CNT % wt > 0.1) and a second for 'extremely low loadings', near the percolation threshold (CNT % wt < 0.1). The high conductivity observed for composites at low CNT loading values can be explained by the existence of a percolative CNT network that coexists with micron size regions of non-conductive material. In contrast, samples with extremely low CNT loading values, which present no connectivity or close proximity between CNT bundles, show an electrical conductivity characterized by a current/voltage dependence. Data suggests that at these loadings, conduction may occur via a material breakdown mechanism, similar to dielectric breakdown in a capacitor. The lessons learned from the data gathered in here could guide future experimental research aimed to control the conductivity of CNT composites.