Using DFT-D3(BJ) during QM/MM calculations

[Amami Haruka]

When running AIMD, I usually add DFT-D3(BJ) dispersion correction to account for the lack of dispersion when running at the level of BLYP/TZVP-MOLOPT-GTH.

&VDW_POTENTIAL

        POTENTIAL_TYPE PAIR_POTENTIAL

        &PAIR_POTENTIAL

          PARAMETER_FILE_NAME dftd3.dat

          TYPE DFTD3(BJ)

          REFERENCE_FUNCTIONAL BLYP

        &END PAIR_POTENTIAL

      &END VDW_POTENTIAL

However, I wonder if I would still need DFT-D3(BJ) during QM/MM calculation? The dispersion between the QM region and the MM region or between the atoms in the MM region is taken into account by Lennard-Jones potential.

For cp2k, is the dispersion between the atoms in the QM region taken into account by the Lennard-Jones potential, or one would need to add dispersion correction via DFT-D3(BJ)?

Thank you.

[Dawid das]

Hi,

You might want to go through this paper: https://pubs.acs.org/doi/pdf/10.1021/jp3060944.

They perform AIMD for a chromophore of fluorescent proteins which is electrostatically embedded in the protein environment

so they must also account for vdW interaactions between QM and MM subsystems at LJ model.

They compare using both BLYP and BLYP+D3 for QM region, so it should be fine. They also use CP2K.

Best wishes,

Dawid

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/2ae51493-c381-445f-8a9d-62dd07116846n%40googlegroups.com.