Exploring DFT+U parameter space with a Bayesian calibration assisted by Markov chain Monte Carlo sampling

Abstract The density-functional theory is widely used to predict the physical properties of materials. However, it usually fails for strongly correlated A popular solution use Hubbard correction treat electronic states. Unfortunately, values U and J parameters are initially unknown, they can vary from one material another. In this semi-empirical study, we explore parameter space a group iron-based compounds simultaneously improve prediction (volume, magnetic moment, bandgap). We Bayesian calibration assisted by Markov chain Monte Carlo sampling three different exchange-correlation functionals (LDA, PBE, PBEsol). found that LDA requires largest correction. PBE has smallest standard deviation its most transferable other compounds. Lastly, predicts lattice reasonably well without