Homogeneous nonequilibrium molecular dynamics method for heat transport and spectral decomposition with many-body potentials

Abstract

The standard equilibrium Green-Kubo and nonequilibrium molecular dynamics (MD) methods for computing thermal transport coefficients in solids typically require relatively long simulation times and large system sizes. To this end, we revisit here the homogeneous nonequilibrium MD method by Evans [Phys. Lett. A 91, 457 (1982)] and generalize it to many-body potentials that are required for more realistic materials modeling. We also propose a method for obtaining spectral conductivity and phonon mean free path from the simulation data. This spectral decomposition method does not require lattice dynamics calculations and can find important applications in spatially complex structures. We benchmark the method by calculating thermal conductivities of three-dimensional silicon, two-dimensional graphene, and a quasi-one-dimensional carbon nanotube and show that the method is about one to two orders of magnitude more efficient than the Green-Kubo method. We apply the spectral decomposition method to examine the long-standing dispute over thermal conductivity convergence vs. divergence in carbon nanotubes.

Thermal conductivity reduction in carbon nanotube by fullerene encapsulation: A molecular dynamics study

“…It can be clearly seen that the average (from 50 independent runs) of the thermal conductivity converges well between 0.6 ns and 1 ns, and we thus calculated the thermal conductivity of the CNP from this time interval, which is κ ≈ 1600 ± 100 W/mK. Using a similar EMD simulation and the same GPUMD code, the thermal conductivity of a (10, 10) SWCNT has been calculated to be κ ≈ 2200 ± 100 W/mK [25]. Therefore, in the limit of infinite length, our EMD simulations predict a reduction of about 30% of the thermal conductivity of a (10, 10) SWCNT with C 60 encapsulation.…”

“…There is a missing entry for EMD simulations of (10, 10) SWCNT, because results from Ref. [25] will be used. higher, but the system length considered was about 20 nm [14] or up to about 80 nm [19] only, and the conclusion is not guaranteed to be valid for longer systems.…”

“…Note that the thermal conductivity calculated in the HNEMD method can be naturally decomposed into components according to the heat current decomposition [25], which is physically equivalent to the decomposition in the Green-Kubo formalism [47]. In the HNEMD simulations, we choose a sufficiently small value of F e = 0.05 µm −1 to ensure the validity of linear response theory, which is the theoretical foundation of the HNEMD method.…”

“…In the HNEMD simulations, we choose a sufficiently small value of F e = 0.05 µm −1 to ensure the validity of linear response theory, which is the theoretical foundation of the HNEMD method. The simulation cell in the HNEMD method is the same as that in the EMD method [25]. It is conventional to examine the cumulative average of the thermal conductivity as a function of time in the production stage [25,[48][49][50][51].…”

“…The simulation cell in the HNEMD method is the same as that in the EMD method [25]. It is conventional to examine the cumulative average of the thermal conductivity as a function of time in the production stage [25,[48][49][50][51]. The results for a C 60 @(10, 10) CNP and a (10, 10) SWCNT are shown in Figs.…”