The objective of Phase I is to identify the strengths and weaknesses of the various multi-reference coupled-cluster (MRCC) methods that have been proposed for the description of molecular states depending upon near degeneracies and non-dynamic electron correlation. Such effects are encountered in bond breaking, at transition states, for complex open shell systems like transition metal atoms, and for excited states. The applicable MRCC methods include state-universal and state-specific Hilbert space approaches, and valence universal Fock space methods. They will be assessed for formal consistency and numerical performance, and further compared to the best possible single-reference CC results, which defines the state-of-the-art in the field. There is no MRCC method that yet satisfies all preferred formal properties, encouraging new developments. Armed with this MRCC assessment, in Phase II, the most promising approaches will be further generalized and written into the massively parallel ACES III system that currently runs at >80% efficiency on 40,000 processors enabling DoD scientists and others to routinely make such applications to their problems. Currently ACES offers the best and most extensive collection of single-reference CC applications for comparison. The MRCC capability in ACES III will be unique in the field providing a resource in demand throughout DoD.
Benefit: Unique quantum chemistry software for computing highly accurate couple-cluster and multi-reference coupled-cluster wave functions for predicting molecular structures, spectra, and energetics. Software for DoD and government use and for potential commercialization.
Keywords: Coupled Cluster Theory,Quantum Chemistry, Multi-Reference, Excited States, Structures, Spectra, Rate Constants, Plume Detection
