Density functional theory (DFT) is the primary quantum mechanics-based tool for calculation and prediction of molecular structure and properties. Standard DFT treatments fail to properly describe dispersion interactions such as Van der Waals and H-bonding. This leads to inaccuracies in estimates of properties of systems where dispersion competes significantly with other effects, and thus hinders particularly the application of DFT to biomolecules, drug design, and medical applications.

As one of the more well-recognized practitioners in the field, Stefan Grimme, University of Münster, Germany, recently developed a variant of DFT that includes the effect of dispersion (DFT-D3). He and his co-workers propose a revision of their scheme to account for London dispersion energy. They use a damping function proposed by Becke and Johnson (BJ) and show that it has a closer connection to reality than the traditional zero damping. They used the GMTKN30 molecular energy database (1200 species and 800 reference values) and found that the revised DFT-D3(BJ) approach is better than DFT-D3(zero damping) for energy benchmarks involving medium range correlation effects, such as Diels-Alder reaction energies in DARC.

Avoiding artificial repulsive interatomic interactions at short distances and any need for pair-specific cut-off radii, and at the cost of one additional global fitting parameter, DFT-D3(BJ) performed better than undamped DFT-D3. However, applicability to intramolecular and thermochemical problems remains an open problem for both types of DFT-D3.

• Effect of the damping function in dispersion corrected density functional theory
  S. Grimme, S. Ehrlich, L. Goerigk,
  J. Comp. Chem. 2011, 32, 1456–1465.
  DOI: 10.1002/jcc.21759