Unique Thermal Conductivity Behavior of Single-Walled Carbon Nanotube−Polystyrene Composites

Abstract

Introduction. Carbon nanotubes have extremely high thermal conductivities; single-walled carbon nanotubes (SWCNTs) have been predicted through molecular dynamics modeling to have a thermal conductivity up to 6600 W/(m K). 1 This value is more than 10 times the value for copper and 2 times more than the highest thermally conductive diamond. An obvious SWCNT application would therefore be a high thermally conductive dispersion for applications such as heat conducting composites 1 or heat transfer fluid suspensions. 2 Using Maxwell's mixing theory, SWCNTs should increase a polymer's thermal conductivity by 50-fold at a volume fraction of ∼1 vol %. Experiments have shown that such improvements do not happen in practice, and improvements, if any, are at least an order of magnitude less than expected. [3][4][5] This lack of improvement is due to the presence of an interfacial resistance to heat transfer between the solid nanotubes and the surrounding matrix termed the Kapitza resistance. This name has been given because Kapitza first measured such interfacial resistances in 1941 with experiments on metal-liquid helium interfaces. 6 A complete description of the cause of this resistance is beyond the scope of this paper, but in short, the Kapitza resistance occurs because of modulus mismatch between the matrix and the inclusion; the greater the mismatch, the greater the resistance. 7,8 The Kapitza resistance becomes more significant with increasing filler surface area and is extremely high in composites with carbon nanotubes because of their high stiffness as well as their high surface area. 9,10 Because of its importance, this type of resistance is still being studied today through experiments and computational models. 11 Duong et al. 3,12 reported Monte Carlo simulation results for heat transfer through carbon nanotube composites that produced conductivity improvements similar to experimental results. Work by Gonnet et al. 1 showed that magnetically aligned carbon nanotube mats (bucky paper) have a relatively high thermal conductivity (42 W/(m K) at ambient temperature). However, after incorporating them in epoxy, the thermal conductivity drops to that of random composites. 1 Although we do not doubt the correctness of the theory, experimentalverificationoftheKapitzaresistanceinpolymer-matrix nanocomposites is not a trivial task. Work by Yodh et al. 5 found a difference in thermal conductivity for composites processed with and without surfactant and attributed the larger increase in thermal conductivity for the surfactant-processed samples to the surfactant layer modifying this interfacial resistance; however, there was no way to know that the dispersion quality was the same. The work by Gonnet et al. 1 described briefly above is certainly persuasive; however, the bucky paper network might have been affected by infusion and curing of the epoxy. One obvious direct comparison would be to measure the thermal conductivity of two polymers with differing moduli filled to identical levels with nanotubes or some other high c...

Temperature dependence of thermal conductivity enhancement in single-walled carbon nanotube/polystyrene composites

“…The T g of the PS and composites studied here was 376 K, as determined by differential scanning calorimetry ͑DSC; TA Instruments Q200͒. However, in contrast to the earlier study using a modulated DSC technique, 28 we found that most samples exhibited an irreversible drop in above T g ͑one exhibited no change͒.…”

“…A recent study of SWCNT/PS composites by Peters et al 28 suggested that a change in interfacial thermal resistance also should occur at the glass transition temperature ͑T g ͒. The T g of the PS and composites studied here was 376 K, as determined by differential scanning calorimetry ͑DSC; TA Instruments Q200͒.…”

Inter-carbon nanotube contact in thermal transport of controlled-morphology polymer nanocomposites

“…This is an interesting result that needs to be validated with experimental measurements. Very recent results by Peters et al [21] indicate that the thermal conductivity of SWNTpolystyrene composites does not increase with the volume fraction of the CNTs as much as one would predict. The reason might very well be the increase of the SWNT-SWNT contact points in the composite, or the value of the thermal boundary resistance.…”

Thermal properties and transition studies of multi-wall carbon nanotube/nylon-6 composites

“…In this case, 5 wt% SWCNTs caused a 30% decrease of the heat capacity. Moreover, one DSC study indicates a similar decrease for polystyrene with 5 wt% SWCNTs [40], but another reports essentially no change for a 2.5 wt% SWCNT/polystyrene composite [41]. A DSC study of 1 wt% SWCNTs in poly(methylmethacrylate) [42] also shows essentially the same heat capacity as the neat polymer.…”