Abstract
A systematic study of techniques for treating noncovalent interactions within the computationally efficient density functional theory (DFT) framework is presented through comparison to benchmark-quality evaluations of binding strength compiled for molecular complexes of diverse size and nature. In particular, the efficacy of functionals deliberately crafted to encompass long-range forces, a posteriori DFT+dispersion corrections (DFT-D2 and DFT-D3), and exchange-hole dipole moment (XDM) theory is assessed against a large collection (469 energy points) of reference interaction energies at the CCSD(T) level of theory extrapolated to the estimated complete basis set limit. The established S22 [revised in J. Chem. Phys. 132, 144104 (2010)] and JSCH test sets of minimum-energy structures, as well as collections of dispersion-bound (NBC10) and hydrogen-bonded (HBC6) dissociation curves and a pairwise decomposition of a protein-ligand reaction site (HSG), comprise the chemical systems for this work. From evaluations of accuracy, consistency, and efficiency for PBE-D, BP86-D, B97-D, PBE0-D, B3LYP-D, B970-D, M05-2X, M06-2X, B97X-D, B2PLYP-D, XYG3, and B3LYP-XDM methodologies, it is concluded that distinct, often contrasting, groups of these elicit the best performance within the accessible double-ζ or robust triple-ζ basis set regimes and among hydrogen-bonded or dispersion-dominated complexes. For overall results, M05-2X, B97-D3, and B970-D2 yield superior values in conjunction with aug-cc-pVDZ, for a mean absolute deviation of 0.41 - 0.49 kcal/mol, and B3LYP-D3, B97-D3, ωB97X-D, and B2PLYP-D3 dominate with aug-cc-pVTZ, affording, together with XYG3/6-311+G(3df,2p), a mean absolute deviation of 0.33 - 0.38 kcal/mol.
| Original language | English |
|---|---|
| Article number | 084107 |
| Journal | Journal of Chemical Physics |
| Volume | 134 |
| Issue number | 8 |
| DOIs | |
| State | Published - Feb 28 2011 |
Funding
This work was performed under the auspices of grants provided by the United States National Science Foundation (Grant No. CHE-1011360). The Center for Computational Molecular Science and Technology is funded through a NSF CRIF award (Grant No. CHE-0946869) and by Georgia Institute of Technology. B.G.S. acknowledges support from the Center for Nanophase Materials Sciences, which is sponsored at Oak Ridge National Laboratory by the Scientific User Facilities Division, U.S. Department of Energy. Á.V.-M. acknowledges support from the Department of Energy, Offices of Basic Energy Science and Advanced Scientific Computing Research as part of the SciDAC program. This research used resources supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC05-00OR22725 and advanced computing resources provided by the National Science Foundation. The computations were performed partially on Kraken at the National Institute for Computational Sciences ( http://www.nics.tennessee.edu/ ). The authors wish to thank Professor Stefan Grimme for helpful discussions on the DFT-D3 method and Mr. Edward G. Hohenstein for inspiration regarding a figure.