skip to main content


This content will become publicly available on May 23, 2019

Title: Hybrid Correlation Energy (HyCE): An Approach Based on Separate Evaluations of Internal and External Components

In this work, a novel hybrid correlation energy (HyCE) approach is proposed that determines the total correlation energy via distinct computation of its internal and external components. This approach evolved from two related studies. First, rigorous assessment of the accuracies and size extensivities of a number of electron correlation methods, that include perturbation theory (PT2), coupled-cluster (CC), configuration interaction (CI), and coupled electron pair approximation (CEPA), shows that the CEPA(0) variant of the latter and triples-corrected CC methods consistently perform very similarly. These findings were obtained by comparison to near full CI results for four small molecules and by charting recovered correlation energies for six steadily growing chain systems. Second, by generating valence virtual orbitals (VVOs) and utilizing the CEPA(0) method, we were able to partition total correlation energies into internal (or nondynamic) and external (or dynamic) parts for the aforementioned six chain systems and a benchmark test bed of 36 molecules. When using triple-ζ basis sets it was found that per orbital internal correlation energies were appreciably larger than per orbital external energies and that the former showed far more chemical variation than the latter. Additionally, accumulations of external correlation energies were seen to proceed smoothly, and somewhat linearly,more » as the virtual space is gradually increased. Combination of these two studies led to development of the HyCE approach, whereby the internal and external correlation energies are determined separately by CEPA(0)/VVO and PT2/external calculations, respectively. When applied to the six chain systems and the 36-molecule benchmark test set it was found that HyCE energies followed closely those of triples-corrected CC and CEPA(0) while easily outperforming MP2 and CCSD. Lastly, the success of the HyCE approach is more notable when considering that its cost is only slightly more than MP2 and significantly cheaper than the CC approaches.« less
ORCiD logo [1] ;  [2]
  1. Frederick National Laboratory for Cancer Research sponsored by the National Cancer Institute, Frederick, MD (United States). Advanced Biomedical Computational Sciences Group
  2. Ames Lab. and Iowa State Univ., Ames, IA (United States). Department of Chemistry
Publication Date:
Report Number(s):
Journal ID: ISSN 1089-5639
Grant/Contract Number:
AC02-07CH11358; HHSN 261200800001E
Accepted Manuscript
Journal Name:
Journal of Physical Chemistry. A, Molecules, Spectroscopy, Kinetics, Environment, and General Theory
Additional Journal Information:
Journal Volume: 122; Journal Issue: 23; Journal ID: ISSN 1089-5639
American Chemical Society
Research Org:
Ames Laboratory (AMES), Ames, IA (United States)
Sponsoring Org:
USDOE Office of Science (SC), Basic Energy Sciences (BES) (SC-22). Chemical Sciences, Geosciences & Biosciences Division
Country of Publication:
United States
OSTI Identifier: