Machine learning of nonlocal exchange-correlation functionals

DFT is the cornerstone of modern computational materials science, but its current approximations fall short of the required accuracy and efficiency for predictive calculations of defect properties, band gaps, stability and electrochemical potentials of materials for energy storage and conversion. To address the self-interaction problem, we introduce non-local charge density descriptors that satisfy exact scaling constraints and learn the exact exchange functional. These models are orders of magnitude faster in self-consistent calculations for solids than hybrid functionals but similar in accuracy. We also extend the capability by learning non-local descriptions of screened exchange and correlation energies for periodic solids. This CIDER formalism is implemented in both Gaussian type orbital and plane wave basis code and allow for large-scale production level calculations of semicondictor, catalyst and battery materials at improved accuracy compares to existing approximations.