Published online by Cambridge University Press: 11 September 2015
Over the last decade, the inclusion of nanofeatures has been demonstrated extensively for improving the performance of thermoelectric materials. The continued approach is to nanofeaturing these materials in a smart manner tailoring their electronic and thermal response. This research provides a computational tool for predicting all parameters in the thermoelectric figure of merit for a Si/Ge superlattice structure as a function of doping and layer thickness. The approach involves coupling a nonequilibrium Green's function electronic and thermal transport model. The phonon description is communicated between the two models to facilitate spatially resolved multiphonon frequency electron–phonon scattering. Findings support the consideration of multiple phonon frequency scattering to accurately predict ZT values. An extrema in ZT as function of both doping and geometry were predicted. Furthermore, the optimal superlattice design was determined to be a Si(2 nm)/Ge(7 nm) with a donor concentration on the order of 1019 cm−3 for operation at 300 K.