Running NVT AIMD simulation with RPBE D3 in air-water interface system.

[zhaoan chen]

Dear all,
I want to study the Air-Water Interface system, so I build a 15* 15* 30 Å water slab with a periodic box of 15* 15* 71.44 Å. The initial condition was obtained from a GROMACS-NVT simulation with the SPC/E force field. Then I want to run a NVT simulation with RPBE-D3 in CP2K, but I can't find how to add D3 correction with RPBE function. Here is my input file and the snapshot at 6900th frame. I find some water molecules running out of the box. Is my input file correct?

I would be very grateful indeed for any help you could give me.
Best wishes.
Zhaoan Chen

My input file setting (section XC) is as follows:

......

   &XC
      FUNCTIONAL_ROUTINE NEW
      &XC_POTENTIAL
        ENERGY XC_FUNCTIONAL
      &END XC_POTENTIAL
      &XC_FUNCTIONAL
        &LIBXC
  FUNCTIONAL GGA_X_RPBE
&END LIBXC
      &END XC_FUNCTIONAL
      &VDW_POTENTIAL
        POTENTIAL_TYPE PAIR_POTENTIAL
        &PAIR_POTENTIAL
          TYPE DFTD3
          CALCULATE_C9_TERM .TRUE.
          REFERENCE_C9_TERM .TRUE.
          LONG_RANGE_CORRECTION .TRUE.
          PARAMETER_FILE_NAME ./dftd3.dat
          REFERENCE_FUNCTIONAL PBE
        &END PAIR_POTENTIAL
      &END VDW_POTENTIAL
    &END XC

.....

[Jürg Hutter]

Hi

some comments on your input and yes, CP2K is not folding back atoms, so they can "move outside" the box.

Don't use &XC_POTENTIAL, it is not what you want.
Most likely your definition of &XC_FUNCTIONAL is incomplete (you only have exchange, no correlation functional).
Don't use LONG_RANGE_CORRECTION: your system is not isotropic (see manual!)
See https://www.chemiebn.uni-bonn.de/pctc/mulliken-center/software/dft-d3/dft-d3
for a list of parametrized functionals for D3.

regards
JH

________________________________________
From: cp...@googlegroups.com <cp...@googlegroups.com> on behalf of zhaoan chen <zz24...@gmail.com>
Sent: Sunday, June 4, 2023 5:20 PM
To: cp2k
Subject: [CP2K:18884] Running NVT AIMD simulation with RPBE D3 in air-water interface system.

--
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<mailto:cp2k+uns...@googlegroups.com>.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/502385d8-dd90-447b-a05e-e8a8b784f80cn%40googlegroups.com<https://groups.google.com/d/msgid/cp2k/502385d8-dd90-447b-a05e-e8a8b784f80cn%40googlegroups.com?utm_medium=email&utm_source=footer>.

[zhaoan chen]

Dear Prof. Dr. Jürg Hutter
Thank you for your reply very much. I'm sorry to reply so late. After Using dftd3 software to calculate my system with rpbe d3, we get the following data. Then
s6 : 1.0000
S8 : 0.5148
rs6 : 0.8720
rs18 : 1.0000
alpha6 : 14.0000
alpha8 : 16.0000
k1-k3 : 16.0000  1.3333  -4.0000
Cutoff : 94.8683 a.u.
CN-Cutoff: : 40.0000 a.u.

Then I changed &XC section of the inp file to the following, and the simulation worked normally.
    &XC
      FUNCTIONAL_ROUTINE NEW

      &XC_FUNCTIONAL
        &LIBXC
  FUNCTIONAL GGA_X_RPBE
&END LIBXC
      &END XC_FUNCTIONAL
      &VDW_POTENTIAL
        POTENTIAL_TYPE PAIR_POTENTIAL
        &PAIR_POTENTIAL
           TYPE DFTD3

  D3_SCALING 1.0000 0.8720 0.5148
  EPS_CN 21.167
  R_CUTOFF 50.20
          PARAMETER_FILE_NAME ./dftd3.dat
          REFERENCE_FUNCTIONAL RPBE

        &END PAIR_POTENTIAL
      &END VDW_POTENTIAL
&END XC

Best wishes.
Zhaoan Chen