On the Performance of Optimally Tuned Range-Separated Hybrid Functionals for X-ray Absorption Modeling.

We investigate the performance of optimally tuned range-separated hybrid functionals (OT-RSH) for modeling X-ray absorption spectra (XAS) of a benchmark set of simple molecules (water, ammonia, methane, hydrogen peroxide, hydrazine, and ethane), using time-dependent density functional theory (TDDFT). Spectra were simulated within the Path Integral based Reflection Principle methodology. Relative intensities, peak positions, and widths were compared with available experimental data. We show that the OT-RSH approach outperforms empirically parametrized functionals in terms of relative peak positions and intensities. Furthermore, we investigate the effect of geometry specific tuning where the range separation parameter is optimized for each geometry. Finally, we propose a simple correction scheme allowing for calculations of XAS on the absolute energy scale using the OT-RSH approach combined with ΔSCF/TDDFT-based calculations of core ionization energies.