[CP2K-user] [CP2K:17501] Re: Carbon Nitride cell optimization not converging

[Xavier Bidault]

Hi,

The main issue is the mixing algorithm. Change to Broyden.
Then you could consider the MOLOPT basis sets.
Also, "c" parameter is a bit small, I would recommend using k-points, al 
least "1 1 2" grid mesh.
NGRIDS 12 is less efficient than NGRIDS 5
Grimme D2 or D3, see the resulting lattice parameters and volume. What's 
best for your goal?

What works here:
...
  &DFT
    BASIS_SET_FILE_NAME BASIS_MOLOPT
    POTENTIAL_FILE_NAME POTENTIAL
    CHARGE 0
    MULTIPLICITY 1
    &MGRID
      CUTOFF 600
      !REL_CUTOFF 60
      NGRIDS 5
    &END MGRID
    &KPOINTS
      SCHEME MONKHORST-PACK 1 1 2
      !SYMMETRY .TRUE.
    &END KPOINTS
    &QS
      METHOD GPW
      EPS_DEFAULT 1.0E-10
    &END QS
    &POISSON
      PERIODIC XYZ
    &END
    &SCF
      EPS_SCF 1.0E-8
      SCF_GUESS ATOMIC
      MAX_SCF 250
     &MIXING
       METHOD BROYDEN_MIXING
     &END MIXING
    &END SCF
    &XC
      &XC_FUNCTIONAL PBE
      &END XC_FUNCTIONAL
      &VDW_POTENTIAL
         POTENTIAL_TYPE PAIR_POTENTIAL
         &PAIR_POTENTIAL
            TYPE DFTD2
            !PARAMETER_FILE_NAME dftd3.dat
            REFERENCE_FUNCTIONAL PBE
            R_CUTOFF [angstrom] 15.0
            !&PRINT_DFTD
            !&END
         &END
      &END VDW_POTENTIAL
    &END XC
  &END DFT
...
    &KIND C
      BASIS_SET DZVP-MOLOPT-GTH
      POTENTIAL GTH-NLCC-PBE-q4
    &END KIND
    &KIND H
      BASIS_SET DZVP-MOLOPT-GTH
      POTENTIAL GTH-NLCC-PBE-q1
    &END KIND
    &KIND N
      BASIS_SET DZVP-MOLOPT-GTH
      POTENTIAL GTH-NLCC-PBE-q5
    &END KIND
...

Best,
Xavier

On Tuesday, August 16, 2022 at 11:39:35 AM UTC-5 shrsi... at gmail.com wrote:

> Dear all,
> I am trying to optimize the crystal structure of carbon nitride  (Unit 
> cell contains 144 atoms: C48H24N72). I have tried various suggestions for 
> Basis sets (SZV, DZVP, TZVP), optimization method (CG BFGS, L-BFGS) but 
> cell-optimization is not converging. Energy is highly fluctuating, 
> ultimately forces on a set of atoms become very high and pressure deviation 
> huge. The initial structure is already optimized in FHI-AIMS, and taken 
> from a reference (
> https://pubs.acs.org/doi/full/10.1021/acs.chemmater.7b00965).
> Particularly, I noticed very high forces on Nitrogen atoms that are 
> involved in Hydrogen-bonding (see attached image of before and after 
> structure)
> I have attached one of the CP2K input file, initial geometry and output 
> file for your reference.
> Thanks in advance.
> Shreya
> Email: shrsi... at gmail.com
>
>

-- 
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+unsubscribe at googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/b39ca7e3-356d-4819-bf28-6d9976c67590n%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20220816/46f2c524/attachment.htm>