Parametric Study of the Thermal Conductivity in Binary Solids

Simulated binary systems of atoms interacting through Lennard-Jones potentials have been studied to determine the thermal conductivity of solid samples as a function of variable relative concentration, mass ratio, hard-core atomic diameter, and attractive well depth. In this study, we use an isobaric Monte Carlo simulation to prepare a simulated sample at near-zero pressure and then use an isoenergy Molecular Dynamics simulation to reach equilibrium. Once at equilibrium, the dynamic Green-Kubo approach is taken to calculate the heat current time-dependent autocorrelation function and determine the thermal conductivity of the sample. Our results indicate that the thermal conductivity has a strong dependence on the investigated relative parameters and relative concentrations of the two species. Discrepancies with the inverse temperature dependence at low temperatures have shown to be significant.