GW-BSE Workflows for High-Throughput Study of Ultra-Wide Band Gap Materials

Ultra-wide band gap (UWBG) materials present an exciting area of research in high-power and RF electronics, deep-ultra-violet optoelectronics, and quantum information science. However, the investigation of the fundamental properties of UWBG materials and their heterostructures, such as impact-ionization rates, band gaps, band offsets, and point defect energy levels, is an enormously challenging task due to the vast configuration space associated with these studies. We develop an open-source Python package, pyGWBSE, for performing automated first-principles calculations within the GW-BSE framework to investigate the fundamental properties of UWBG materials and their heterostructures in a high-throughput manner. GW approximation accurately predicts the electronic structure of materials, for instance, it overcomes the bandgap underestimation issue of the more widely used density functional theory. BSE formalism accurately predicts the excitonic properties of materials producing absorption spectra that are directly comparable with experimental observations. We show how pyGWBSE tackles the two main challenges of developing high-throughput GW-BSE frameworks, namely the convergence of the interdependent simulations parameters and the optimization of the computational cost associated with the multi-step simulations. Lastly, we use the pyGWBSE to compare the computed properties of AlN with those measured experimentally and also present the pyGWBSE predicted properties of B_xAl_1-xN alloy structures in the entire x = 0 to 1 range.