Plasticity tuning of thermal conductivity between nanoparticles

Abstract

We study the effects of uniaxial pressure on the thermal conductivity between two nanoparticles using atomistic simulation. While the system is compressed, we analyze the evolution of contact area, the relative density, and the dislocation density. Lattice thermal conductivity is calculated by non-equilibrium molecular dynamics simulations at several stages of the compression. Despite the increment of dislocation defects, thermal conductivity increases with pressure due to the increase in relative density and contact radius. The behavior of the contact radius is compared with the Johnson–Kendall–Roberts (JKR) model. While there is good agreement at low strain, after significant plasticity, signaled by the emission of dislocations from the contact region, the discrepancy with JKR grows larger with the dislocation density. The results for thermal conductivity show good agreement with previous studies at zero strain, and a theoretical model is used to accurately explain its behavior vs strain-dependent contact radius. Both the Kapitza resistance and thermal resistance decrease with strain but with very different evolution. Simulations of a bulk sample under uniaxial strain were also carried out, allowing for a clear distinction between the role of compressive stress, which increases the conductivity, vs the role of dislocations, which decrease the conductivity. For the NP system, there is the additional role of contact area, which increases with stress and also modifies conductivity. An analytical model with a single free parameter allows for a description of all these effects and matches both our bulk and NP simulation results.