Methods for fast and reliable computation of electronic excitation energies are in short supply, and little is known about their systematic performance. Low-scaling methods, such as time-dependent density functional theory (TD-DFT) are plagued by problems such as underestimating charge transfer states and their unpredictability. Methods based off of the coupled cluster ansatz, such as the equation of motion coupled cluster singles and doubles (EOM-CCSD), are much more reliable than DFT, but their high computational scaling O(N6) prohibits their application to many large systems. We developed and investigated the performance of several low-scaling O(N5) wave function methods for electronic excited states, including P-EOM-MBPT2 and EOM-CC2, testing them against a series of organic molecules with accurate gas-phase experimental data.  We found that the accuracy of P-EOM-MBPT2 outperforms CIS(D) by several tenths of an eV, especially in the case of Rydberg states. Comparisons between EOM-CC2 and P-EOM-MBPT2 provide an excellent illustration of the relative importance of single and double excitations for describing electronic excitations, as well as providing guidelines for selecting low-scaling methods for computing electronic excitation energies.

Equation of motion formalism of second order many-body perturbation theory (EOM-MBPT2) and second-order approximate coupled-cluster (EOM-CC2)

248th National Meeting (2014)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.