Stochastic Simulation of Nonequilibrium Heat Conduction in Extended Molecule Junctions

TL;DR

This paper introduces a stochastic nonequilibrium molecular dynamics framework to simulate heat conduction in molecular junctions, aligning well with experimental data and revealing the influence of molecular structure on thermal transport.

Contribution

It presents a novel simulation approach that models realistic heat reservoirs and compares classical and quantum methods, advancing understanding of nanoscale heat conduction in molecular systems.

Findings

Results agree with previous NEGF calculations and experiments.

Heat transport is dominated by low-frequency vibrational modes.

Molecular conjugation affects thermal conductance.

Abstract

Understanding phononic heat transport processes in molecular junctions is a central issue in the developing field of nanoscale heat conduction and manipulation. Here we present a Stochastic Nonequlibrium Molecular Dynamics simulation framework to investigate heat transport processes in molecular junctions in and beyond the linear response regime. We use extended molecular models which filter Markovian heat reservoirs through an intermediate substrate region, to provide a realistic and controllable effective bath spectral density. The results obtained for alkanedithol molecules connecting gold substrates agree with previous nonequilibrium Green's function calculations in frequency domain, and match recent experimental measurements (e.g. thermal conductance around 20 pW/K for alkanedithiols in single molecular junctions) Classical MD simulations using the full molecular forcefield and quantum Landauer-type calculations based on the harmonic part of the same forcefield are compared, and the similarity of the results indicate that heat transport is dominated by modes in the lower frequency range. Heat conductance simulations on polyynes of different lengths illuminates the effects of molecular conjugation on thermal transport.