State-Specific Configuration Interaction for Excited States

TL;DR

This paper presents a state-specific configuration interaction method for excited states, offering a systematically improvable and accurate alternative to traditional approaches, especially effective for challenging multireference and low-lying excited states.

Contribution

Introduction of the $ riangle$CI approach, a systematically improvable, state-specific method for excited-state calculations that outperforms standard CI and is competitive with coupled-cluster methods.

Findings

$ riangle$CI is more accurate than standard ground-state CI.

$ riangle$CISD+Q surpasses EOM-CC2 and EOM-CCSD for larger systems.

The method effectively handles multireference and various excited states.

Abstract

We introduce and benchmark a systematically improvable route for excited-state calculations, state-specific configuration interaction ($\Delta$CI), \alert{which is a particular realization of multiconfigurational self-consistent field and multireference configuration interaction.} Starting with a reference built from optimized configuration state functions, separate CI calculations are performed for each targeted state (hence state-specific orbitals and determinants). Accounting for single and double excitations produces the $\Delta$CISD model, which can be improved with second-order Epstein-Nesbet perturbation theory ($\Delta$CISD+EN2) or a posteriori Davidson corrections ($\Delta$CISD+Q). These models were gauged against a vast and diverse set of 294 reference excitation energies. We have found that $\Delta$CI is significantly more accurate than standard ground-state-based CI, whereas close performances were found between $\Delta$CISD and EOM-CC2, and between $\Delta$CISD+EN2 and EOM-CCSD. For larger systems, $\Delta$CISD+Q delivers more accurate results than EOM-CC2 and EOM-CCSD. The $\Delta$CI route can handle challenging multireference problems, singly- and doubly-excited states, from closed- and open-shell species, with overall comparable accuracy, and thus represents a promising alternative to more established methodologies. In its current form, however, it is only reliable for relatively low-lying excited states.