High-throughput computations of phonon-limited electronic transport

The last decade has seen significant developments in our capability to model transport processes limited by phonons such as mobility, conductivity or Seebeck coefficients. A full ab initio treatment of electronic transport within Boltzmann transport formalism and taking into account electron-phonon coupling is now possible and has been demonstrated for several important materials. These fully ab initio computations are however computationally expensive and most importantly require in general a very dense k-point mesh obtained through Wannierization, a process that is not always trivial and challenging to automatize.

I will report on our efforts in bringing phonon-limited electronic transport to a high-throughput level where the transport of hundreds to thousands of materials can be computed automatically. I will present our approach offering a Wannier-free implementation of electron-phonon transport and present how automation offers possibilities to more effectively test the different possible relaxation-time approximations versus the fully iterative solution of the Boltzmann transport equations. Finally, I will show how these techniques can be used in the field of transparent conducting oxides to discover new material and compare the performance of our model to simpler models such as AMSET on selected examples.