Enhancing Excited State Characterization in Transition Metal Oxides Through Advanced Wavefunction Methods

Understanding the nature of excitations in complex materials, particularly in solids and transition metal oxides, is a critical challenge due to the limitations of traditional mean-field methods, which, although often accurate for ground states, offer less reliability for excited states. This study introduces an approach employing selected Configuration Interaction (sCI) and wavefunction optimization techniques to address these challenges, providing a more physically representative depiction of excited states in these systems.

Conventional plane wave methods, even when enhanced with corrections like Hubbard U in Density Functional Theory (DFT), typically yield virtual orbitals that, while not necessarily incorrect, are often "correct for the wrong reasons." These methods lack in capturing the nuanced behavior of excited states, particularly in transition metal compounds. Our approach, utilizing advanced wavefunction methodologies, not only achieves a richer physical representation but also holds significant implications for computational material science. The optimized wavefunctions present a robust foundation for more accurate predictions in various complex behaviors, offering enhanced applicability in other advanced methods such as Diffusion Monte Carlo (DMC) and GW approximations. This synergy underscores the potential for these refined techniques to revolutionize predictive modeling and understanding of electronic excitations in complex materials.