Quadratically convergent algorithm for orbital optimization in the orbital-optimized coupled-cluster doubles method...

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Theoretical Chemistry

Background:

• Standard coupled-cluster doubles (CCD) and Møller-Plesset perturbation theory (MP2) methods rely on fixed molecular orbitals (MOs).
• Variational optimization of MOs can lead to more accurate and stable theoretical descriptions of molecular properties.

Purpose of the Study:

• To derive and implement orbital-optimized versions of CCD (OO-CCD) and MP2 (OO-MP2) methods using a Lagrangian-based approach.
• To develop an improved algorithm for the variational optimization of molecular orbitals.
• To compare the performance of OO-CCD and OO-MP2 against standard methods for various molecular systems.

Main Methods:

• A Lagrangian-based approach for deriving the equations for variational MO optimization in CCD and MP2.
• Implementation of an improved orbital optimization algorithm using Newton-Raphson method with MO gradient and Hessian.
• Application of OO-CCD, OO-MP2, and standard MP2, CCD, CCSD, CCSD(T) methods to H(2)O, diatomics, and O(4)(+).

Main Results:

• OO-CCD and CCSD yield nearly identical results for H(2)O and diatomics.
• OO-CCD shows improved vibrational frequency predictions for symmetry-breaking systems like O(4)(+).
• OO-MP2 provides more accurate vibrational frequencies for O(4)(+) compared to MP2, with generally longer bond lengths and smaller frequencies.

Conclusions:

• Orbital-optimized methods (OO-CCD, OO-MP2) offer significant advantages, including easier analytic gradient computation and improved handling of symmetry breaking.
• OO-CCD provides benefits over CCSD, such as simpler one-electron property calculations and gauge-invariant transition dipole moments.
• OO-MP2 shows promise for excited-state properties and offers improved accuracy for challenging molecular systems.