grimmedisp

grimmedisp

This keyword specifies the use of Grimme’s third-generation dispersion correction (DFT-D3).

Options

Hint

Always enable DFT-D3 in DFT calculations—especially for systems held together by weak interactions. In practice, it is advisable to apply it in all cases.

If the input file omits the grimmedisp keyword, no DFT-D3 correction is applied.

type
Value bj Becke–Johnson–damped DFT-D3 (DFT-D3(BJ)).
zero Zero-damping DFT-D3 (DFT-D3(0)).
Default None

You must choose either bj or zero; there is no default value.

The bj form usually performs better, but some functionals—such as M06, M06-2X, and M06-HF—are parametrised only for the zero form.

three_body

The optional three-body switch adds the Axilrod–Teller–Muto three-body dispersion term, which can be beneficial for large systems.

tz

When triple-zeta (or larger) basis sets are used, including the three-body term often yields additional accuracy.

Theoretical Background

The DFT-D3 method is a dispersion correction applied to DFT energies. It is based on the pairwise summation of damped dispersion terms. In modern DFT calculations, including the DFT-D3 correction is almost always beneficial; therefore, it is recommended for all calculations.

Input Examples

Examples: Weak Interactions in CH4-C2H6

This example illustrates the importance of applying the DFT-D3 correction when evaluating weak interactions. Both CH4 and C2H6 are non-polar, so the interaction between them is extremely weak. We assess this interaction using the EDA method at the B3LYP-D3(BJ)/def2-SVP level of theory:

The output is as follows:

Thus, the interaction energy between CH4 and C2H6 is −0.62 kcal mol−1. The dispersion component is −0.87 kcal mol−1, making it the dominant contribution. Without the DFT-D3 correction, the interaction energy would be −0.62 − (−0.87) = +0.25 kcal mol−1; in other words, the complex would be unbound. This represents a quantitative error in the interaction energy. Always include DFT-D3 in your DFT calculations.