Assessment of Orbital-Optimized Third-Order Moller-Plesset Perturbation Theory and Its Spin-Component and Spin-Opposite Scaled Variants for Thermochemistry and Kinetics

Soydas E., Bozkaya U.

JOURNAL OF CHEMICAL THEORY AND COMPUTATION, vol.9, no.3, pp.1452-1460, 2013 (SCI-Expanded) identifier identifier identifier

  • Publication Type: Article / Article
  • Volume: 9 Issue: 3
  • Publication Date: 2013
  • Doi Number: 10.1021/ct301078q
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.1452-1460
  • Hacettepe University Affiliated: No


An assessment of the OMP3 method and its spin-component and spin-scaled variants for thermochemistry and kinetics is presented. For reaction energies of closed-shell systems, the CCSD, SCS-MP3, and SCS-OMP3 methods show better performances than other considered methods, and no significant improvement is observed due to orbital optimization. For barrier heights, OMP3 and SCS-OMP3 provide the lowest mean absolute deviations. The MP3 method yields considerably higher errors, and the spin scaling approaches do not help to improve upon MP3, but worsen it For radical stabilization energies, the CCSD, OMP3, and SCS-OMP3 methods exhibit noticeably better performances than MP3 and its variants. Our results demonstrate that if the reference wave function suffers from a spin-contamination, then the MP3 methods dramatically fail. On the other hand, the OMP3 method and its variants can tolerate the spin-contamination in the reference wave function. For overall evaluation, we conclude that OMP3 is quite helpful, especially in electronically challenged systems, such as free radicals or transition states where spin contamination dramatically deteriorates the quality of the canonical MP3 and SCS-MP3 methods. Both OMP3 and CCSD methods scale as n(6), where n is the number of basis functions. However, the OMP3 method generally converges in much fewer iterations than CCSD. In practice, OMP3 is several times faster than CCSD in energy computations. Further, the stationary properties of OMP3 make it much more favorable than CCSD in the evaluation of analytic derivatives. For OMP3, the analytic gradient computations are much less expensive than CCSD. For the frequency computation, both methods require the evaluation of the perturbed amplitudes and orbitals. However, in the OMP3 case there is still a significant computational time savings due to simplifications in the analytic Hessian expression owing to the stationary property of OMP3. Hence, the OMP3 method emerges as a very useful tool for computational quantum chemistry.