Modeling device-level semiconductors and their interfaces by orbital-free DFT

Orbital-Free Density Functional Theory (OFDFT) is one of most promising methods for large-scale quantum mechanical simulations as it offers a good balance of accuracy and computational cost. Million atom simulations are nowadays achievable with OFDFT because the noninteracting kinetic energy functional is approximated by an explicit functional of the electron density removing the need to employ orbitals and diagonalizations. The commonplace belief is that because of the underlying approximations, OFDFT can only approach metallic systems. However, newly developed functionals allow the quantitative description of semiconductors and semiconducting quantum dots. We present an implementation of OFDFT entirely in Python providing some useful abstractions to deal with molecules and materials. The simulations are extremely computationally efficient, delivering converged electronic structures for million-atom system sizes realizing experimentally observed devices. We present calculations of work functions and Schottky barriers for an array of realistically sized systems that are still completely out of reach of commonly employed Kohn-Sham DFT methods.