Computational materials physics is playing an increasingly important role in developing a fundamental understanding of the basic properties of materials and in the search and design of new materials for improved technologies. Advanced techniques have been developed to examine structural and electronic properties of materials from first principles. Thermal transport properties, however, have been much less developed despite the important role these properties play in devices. The utility of a non-metallic material for heat spreading or thermoelectric applications is determined by its lattice thermal conductivity, kL. To rigorously understand kL and the mechanisms by which to manipulate thermal transport, accurate representation of the intrinsic anharmonic phonon scattering is critically important.
Here, I will discuss a first principles approach for calculating lattice thermal conductivity that combines a full solution of the Peierls-Boltzmann transport equation with interatomic forces determined from density functional theory. This parameter-free, atomistic approach is: i) quantitatively accurate, ii) predictive, and iii) transferable to a range of systems. I will present highlights from our first principles calculations, including our prediction of the ultra-high kL of cubic BAs, and discuss the intrinsic vibrational properties that determine kL. Further, I will discuss the role of some extrinsic mechanisms for engineering thermal transport properties in materials.