Target State Optimization with Density Functional Theory for Computing K-Edge X-ray Absorption Spectra

This study investigates the performance of Target State Optimization Density Functional Theory (TSO-DFT) in predicting molecular K-edge X-ray Absorption Spectra (XAS). In contrast to Time-Dependent Density Functional Theory (TDDFT), which systematically underestimates core excitation energies by more than 10 eV, TSO-DFT optimizes wave function for every state to account for orbital relaxation effects and can achieve an accuracy of subelectronvolts. We apply TSO-DFT to predict the K-edge XAS for which is critical for accurate predictions of core excitations. TSO-DFT is applied to predict XAS for CO₂, N₂O, carbonyl compounds, and radicals. TSO-DFT succeeds in predicting both the main features and core excitation energies. TSO-DFT is also used to predict the angle-dependent XAS of porphyrin and polarized XAS of the uranyl ion. The calculated XAS agrees quite well with experiments and can give details of electronic structure change that cannot be obtained straightforwardly from experiments alone. TSO-DFT is a promising method for studying molecular XAS and is available in the software Qbics, which can be downloaded free of charge.