Quantitative Analysis of Connectivity and Conductivity in Mesoscale Multiwalled Carbon Nanotube Networks in Polymer Composites

Mesoscale networks of multiwalled carbon nanotubes (CNTs) dispersed in 50/50 PS/PPO (polystyrene/poly(2,6-dimethyl-1,4-phenylene oxide)) are studied. Various dispersion states are formed by several processing methods such as solvent casting (SC) and extrusion (EXT) and followed by compression molding (CM), which alters the topology of the conductive CNT network. This in turn affects the measured electrical conductivity up to 9 orders of magnitude. Quantitative large-area STEM imaging of the representative topology revealed the existence of subtle differences in topology for various dispersion states. The analyses also reveal that strong localized CNT interactions during the SC process result in highly conductive inhomogeneous networks consisting of CNT clusters connected by individual CNTs embedded in the polymer matrix. This leads to a low electrical percolation threshold, so that as compared to a random network a higher local volume fraction than the overall volume fraction of CNTs is present. A dispersion state created by the extrusion process (EXT) results in less conductive and close to randomly oriented networks. The electrical conductivity of the mesoscale networks is further enhanced by compression molding (annealing), which increases the effective CNT CNT contact area per unit volume and hence decreases the contact resistance between CNTs and results in significant improvement in electrical conductivity. This study illustrates how multiscale information can be used to understand and subsequently tailor the macroscopic functional properties of many scientifically and technologically relevant functional materials.