Phenomenological model of thermal transport in carbon nanotube and hetero-nanotube films

Abstract

Abstract The thermal properties of individual single-walled carbon nanotubes (SWCNTs) have been well documented in the literature following decades of intensive study. However, when SWCNTs form a macroscale assembly, the thermal transport in these complex structures usually not only depends on the properties of the individual tubes, but also is affected and sometimes dominated by inner structural details, e.g. bundles and junctions. In this work, we first performed an experimental measurement of the thermal conductivities of individual SWCNT bundles of different sizes using a suspended micro-thermometer. The results, together with the data that we obtained from a previous work, give a complete experimental understanding of the effect of bundling on the thermal conductivity of SWCNTs. With these quantitative understandings, we propose a phenomenological model to describe the thermal transport in two-dimensional (2D) SWCNT films. The term ‘line density’ is defined to describe the effective thermal transport channels in this complex 2D network. Along with experimentally obtained geometric statistics and film transparency, the thermal conductance of SWCNTs is estimated, and the effects of bundle length, diameter, and contact conductance are systematically discussed. Finally, we extend this model to explain thermal transport in 2D networks of one-dimensional van der Waals heterostructures, which are coaxial hetero-nanotubes we recently synthesized using SWCNTs as the template. This extended model suggests that the contribution of boron nitride nanotubes (BNNTs) to the overall performance of a SWCNT–BNNT heterostructured film depends on the transparency of the original SWCNT film. The increase in the thermal conductance of a highly transparent film is estimated to be larger than that of a less transparent film, which shows a good agreement with our experimental observations and proves the validity of the proposed phenomenological model.