General-model-space state-universal coupled-cluster methods for excited states: diagonal noniterative triple...

Area of Science:

• Quantum Chemistry
• Computational Chemistry
• Theoretical Chemistry

Background:

• The development of accurate computational methods is crucial for understanding molecular properties.
• Coupled-cluster (CC) methods are highly accurate but computationally demanding.
• Handling excited states and symmetry-degenerate states requires advanced CC approaches.

Purpose of the Study:

• To extend the general-model-space, state-universal coupled-cluster singles and doubles (GMS SU CCSD) approach to include perturbative triples.
• To assess the effectiveness of this GMS SU CCSD(T) method for calculating excitation energies, particularly for systems with static and nondynamic correlation.
• To compare the performance of GMS SU CCSD(T) with the (N,N)-CCSD method for excited states and symmetry-breaking issues.

Main Methods:

• Extension of the GMS SU CCSD method to incorporate perturbative triples.
• Application of the GMS SU CCSD(T) and (N,N)-CCSD methods to calculate low-lying excitation energies of C2, N2, and CO molecules.
• Analysis of static and nondynamic correlation effects in spin- and/or space-symmetry degenerate levels using the spin-orbital formalism.

Main Results:

• Triple corrections significantly improve absolute energies but have a modest impact on excitation energies.
• The inclusion of triples can be detrimental to excitation energy accuracy if ground and excited states are not balanced.
• Triple corrections aid in mitigating symmetry-breaking effects in the spin-orbital formalism, though less effectively than (N,N)-CCSD.

Conclusions:

• The GMS SU CCSD(T) method offers improvements in absolute energy calculations.
• Careful consideration of state balancing is necessary when applying triple corrections to excitation energies.
• While helpful for symmetry breaking, GMS SU CCSD(T) is less effective than (N,N)-CCSD in this regard.