Correcting Dispersion Corrections with Density-Corrected DFT

Area of Science:

• Computational chemistry
• Quantum chemistry
• Materials science

Background:

• Empirical parametrizations of dispersion corrections in DFT often mix functional and density-driven errors.
• Existing methods rely solely on energy errors, limiting accuracy.

Purpose of the Study:

• Introduce a dual-calibration approach, Density and Dispersion-Corrected DFT (D2C-DFT).
• Improve the parametrization of dispersion interactions by accounting for density delocalization errors.

Main Methods:

• Exclude density-sensitive reactions from training datasets.
• Employ tailored dispersion corrections with Hartree-Fock densities.

Main Results:

• Achieved a significant reduction in both errors and variations.
• Demonstrated improved performance across several semilocal functionals and global hybrids.
• Validated the effectiveness of the D2C-DFT approach.

Conclusions:

• D2C-DFT offers a more robust method for dispersion corrections in DFT.
• Accounting for density delocalization errors leads to more accurate and reliable results.
• The approach enhances the predictive power of DFT for various chemical systems.