抄録
ABSTRACT
Over the last decade, there has been a great enhancement in efficiency of bulk thermoelectric materials by reducing lattice thermal conductivity with internal nanostructures. This includes sintered nanostructures, which has merit in applicability and scalability. While the material development so far has been rather empirical, for further improvement of the bulk nanostructured materials, the need for material design based on prediction of lattice thermal conductivity is growing. This paper reviews recently developed multiscale method to calculate lattice thermal conductivity of bulk crystalline nanostructures. The method seamlessly combines first-principles calculations of interatomic force constants, lattice dynamics calculations of intrinsic phonon transport properties, and Monte Carlo simulation of phonon Boltzmann transport through nanostructures. The method can handle phonon transport through a large system with randomly placed and shaped nanostructures, and thus it is sufficient to quantify how reduction of thermal conductivity of polycrystalline nanostructures depends on mean size and distribution of nano-grains. The results for sintered polycrystalline silicon nanostructure are presented.
