Bond breaking during the optimisation

[Nikhil Maroli]

The bonds are breaking during the optimization, Initially, i have tried to optimize the structure with gaussian but it started to break bonds even at correct distances. Anyone can give any suggestion to avoid bond breaking in CP2K and to achieve optimized structure.

Inputs are given below for your information.

&GLOBAL

 PROJECT job2

 RUN_TYPE GEO_OPT

&END GLOBAL

&FORCE_EVAL

 METHOD QS

 &DFT

# Please copy the GTH_BASIS_SETS and POTENTIAL files

# which may have been in cp2k-x.x.x/data or cp2k-x.x.x/tests/QS directory

# to the CP2K executable directory if you use these default files.

   CHARGE 0

   MULTIPLICITY 2

  BASIS_SET_FILE_NAME  GTH_BASIS_SETS

  POTENTIAL_FILE_NAME  POTENTIAL

  &MGRID

   NGRIDS 4

   CUTOFF 30

  &END MGRID

  &XC

   &XC_FUNCTIONAL PBE

   &END XC_FUNCTIONAL

  &END XC

 &END DFT

 &SUBSYS

  &COORD

    Pb         1.94602       -2.12632        1.25157

    Br         3.01219       -0.06256        2.35356

    Pb         2.21556        2.03217        1.09683

    Br        -0.12169        3.23840        1.92449

    Pb        -2.34714        2.39328        0.99984

    Br        -3.57990        0.20869        1.95980

    Pb        -2.65498       -2.14276        1.00547

    Br        -0.34094       -3.29325        1.98965

    Pb         2.06448       -1.33415       -1.21263

    Pb         2.13521        1.12177       -1.15878

    Br         0.06365        1.19694       -0.08612

    Pb        -1.79865        1.47428       -1.68114

    Br        -0.57583       -0.01056       -3.59712

    Pb        -1.98293       -1.29020       -1.66395

    Br         0.01062       -1.18331       -0.10788

    Br        -3.82309       -3.16020       -1.27765

    Br         3.19291       -3.45179       -0.50642

    Br        -3.41601        3.49721       -1.28741

    Cs        -0.28609        0.02294        3.08499

  &END COORD

   &KIND Br

    BASIS_SET DZVP-GTH

    POTENTIAL GTH-PBE-q7

   &END KIND

   &KIND Cs

    BASIS_SET DZVP-GTH

    POTENTIAL # type pseudopotential functional name !

   &END KIND

   &KIND Pb

    BASIS_SET DZVP-GTH

    POTENTIAL # type pseudopotential functional name !

   &END KIND

  &CELL

   ABC 50 50 50 # Provisional Cell Param. Please "Add Unit Cell" !

  &END CELL

 &END SUBSYS

&END FORCE_EVAL

[Patrick Gono]

Dear Nikhil,

I have a couple things to suggest:

1) Your CUTOFF is definitely too small. Please, make sure your model is converged with respect to your CUTOFF by running a series of single-point (ENERGY) calculations, as described in: https://www.cp2k.org/howto:converging_cutoff. In general, the values for CUTOFF should be in the range of (higher) triple digits, 400 - 800 Ry, or even more. Whenever possible, make sure you are converged.

2) You are not using any pseudopotential for your Pb and Cs atoms. Look into the GTH_POTENTIALS file and choose the appropriate one for your choice of basis sets.

3) Are you sure about your cell parameters? If your plan is to study a periodic system, you must specify periodic symmetry and choose appropriate cell parameters. If you plan to simulate an isolated nanoparticle, consider reducing the size of your cell. 50 angstrom in each direction  might be an overkill and needlessly slow down your jobs.

Yours sincerely,

Patrick Gono

--
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/96a565a1-0f85-4692-98fe-68f8d43c59e9%40googlegroups.com.