Embedded self-energy technique for solving arbitrary nanoelectronic systems using FEAST

This paper presents an efficient computational technique based on embedded self-energy functions, and describes how this technique can be used in the case of electronic structure problems. This technique demonstrates a numerical implementation of the mathematical framework built around the FEAST eigenvalue solver, as proposed in [1]. By taking advantage of the benefits of the finite element method for performing muffin-tin type domain decomposition, this technique naturally removes the non-linearity of the resulting reduced eigenvalue problem. This approach can work for arbitrary atomistic systems of different dimensions. Results for 2D systems are shown, along with plans for future work.