Enhanced thermal conductivity of epoxy/alumina composite through multiscale-disperse packing

Abstract

Inspired by the size of the voids in closest packing structures, we propose to use the combination of spherical particles with different size scales to increase the loading fraction of the fillers in epoxy-based composites. In this study, high loading up to 79 vol% has been achieved with multiscale particle sizes of spherical Al2O3 particles. The highest thermal conductivity of Al2O3-filled liquid epoxy measured by steady-state method is 6.7 W m−1 K−1 at 25°C, which is approximately 23 times higher than the neat epoxy (0.28 W m−1 K−1). Three models based on Maxwell mean-field scheme (MMF), differential effective medium (DEM) and percolation theory model (PTM) were utilized to assess our measured thermal conductivity data. We found that both DEM and PTM models could give good results at high volume fraction regime. We have also observed a considerable reduction (10–15%) of thermal conductivity in our Al2O3-filled cured epoxy samples. We attribute this reduction to the increasing of thermal interfacial resistance between Al2O3 particles and cured epoxy matrix, induced by cure shrinkage during the reaction. Our experiments have demonstrated that systems with multiscale particle sizes exhibit lower viscosity and can be filled with much higher fraction of fillers. We thus expect that higher thermal conductivity (probably >12 W m−1 K−1 based on DEM) can be achieved in future via filling higher thermal conductivity spherical fillers (e.g., AlN, SiC), increasing loading fraction by multiscale-disperse packing and reducing the effect from cure shrinkage.