Scalability of quantum simulations of thermoelectric superlattice devices

The effects of layer thickness and periodicity on thermoelectric properties of Si/Ge superlattice materials are studied through a non-equilibrium Green’s function (NEGF) computational approach. Results show an independence of Seebeck coefficient for increasing the number of superlattice periods. Additionally, a critical layer thickness is obtainable to achieve a favorable transmission spectrum while maximizing hetero-interface density. These findings are important to researchers who computationally explore thermoelectric properties of nanostructured materials. In particular, these results identify ranges of parameters worthy of further experimental study.

► Simulations of limited superlattice bilayer systems can be used to approximate many-layer behavior.
► Estimation of Seebeck coefficient is independent of number of bilayers.
► Calculation of transmission provides explanation of convergence behavior.
► Problem size scalability is N2; parallel scalability is linear.