Warning in connectivity generation

[Alex]

Hi all,

I have a system that contains a molybdenum disulfide (MoS2) sample and a charged (+1) potassium ion. The MoS2 structure itself gets energy-minimized without any issues, but in combination with this ion I get a nonstop stream of GENERATE warnings like this: 

 GENERATE|  Preliminary Number of Bonds generated:                           360

 GENERATE| WARNING in connectivity generation!

 GENERATE| Two molecules/residues named (MOL1) have different  number of atoms.

 GENERATE| Molecule starting at position (13) has Nr. <0> of atoms.

 GENERATE| while the other same molecules have Nr. <162> of atoms!

 GENERATE| Increasing bondparm_factor by 1.05.. An error was found in the generated

 GENERATE| connectivity. Retry...

 GENERATE| Present value of BONDPARM_FACTOR (   Infinity ).

When I replace the potassium ion with e.g. a proton, everything is back to normal.

Here's the DFT section of my input:

  &DFT

    CHARGE 1

    BASIS_SET_FILE_NAME BASIS_SET2

    POTENTIAL_FILE_NAME GTH_POTENTIALS2

    &MGRID

      NGRIDS 5

      CUTOFF 500

    &END MGRID

    &QS

      METHOD GPW

    &END QS

    &SCF

      SCF_GUESS RESTART

      EPS_SCF 5.0E-6

      MAX_SCF 150

      ADDED_MOS  100

      &MIXING

          METHOD BROYDEN_MIXING

          ALPHA    0.1

          NBROYDEN   8

      &END MIXING

    &END SCF

     &XC

        &XC_FUNCTIONAL  PBE

        &END XC_FUNCTIONAL

        &VDW_POTENTIAL

          POTENTIAL_TYPE  PAIR_POTENTIAL

          &PAIR_POTENTIAL

            TYPE  DFTD3

            PARAMETER_FILE_NAME dftd3.dat

           REFERENCE_FUNCTIONAL PBE

           CALCULATE_C9_TERM  T

          &END PAIR_POTENTIAL

        &END VDW_POTENTIAL

      &END XC

  &END DFT

Any suggestions? If you need further info, please let me know. 

Thank you,

Alex

[Marcella Iannuzzi]

Hi Alex, 

the connectivity is usually not needed for a DFT calculation.

Is there any particular reason why you want cp2k to generate the connectivity?

Anyway, the input part that is related to that is the SUBSYS section and not the DFT one.

Regards

Marcella

[Gabriele]

Hi Alex,

It looks like the error comes from the SUBSYS section in https://manual.cp2k.org/trunk/CP2K_INPUT/FORCE_EVAL/SUBSYS/ and not from the DFT section. If you are using just DFT there is no need to generate the connectivity, which is where you get the warning. We can have a closer look if you upload the whole input file. 

Cheers,

Gabriele

On Thursday, November 23, 2017 at 3:02:52 AM UTC+1, Alex wrote:

[Alex]

See, but that's the funny part. I absolutely DO NOT want any topologies created and I did not think there was any need for it in this type of simulations. Other than using a PDB file with coordinates as input as part TOPOLOGY section, I am not asking (not explicitly, anyway) for any connectivity tables. should I just convert to xyz and dump the whole TOPOLOGY section? All suggestions are very welcome.

Thanks,

Alex

p.s. Complete file below:

&GLOBAL
  PROJECT opt
  PRINT_LEVEL MEDIUM
  RUN_TYPE GEO_OPT
&END GLOBAL
 &MOTION
  &GEO_OPT
    MAX_ITER 400
    OPTIMIZER BFGS
  &END
&END MOTION
&FORCE_EVAL
  METHOD QS
  STRESS_TENSOR ANALYTICAL
  &DFT
#    CHARGE 1   

  &SUBSYS
    &CELL
     ABC    22.087 21.859 50.000
    ANGLES    90.000 90.000 90.000
        PERIODIC XYZ
        MULTIPLE_UNIT_CELL 1 1 1
    &END
    &TOPOLOGY
       &CENTER_COORDINATES TRUE
       &END
       COORD_FILE_NAME pore3_potassium4.pdb
       COORD_FILE_FORMAT PDB
       MULTIPLE_UNIT_CELL 1 1 1
    &END
    &KIND S
      BASIS_SET  DZVP-MOLOPT-SR-GTH
      POTENTIAL  GTH-PBE-q6
    &END KIND
    &KIND MO
      BASIS_SET  DZVP-MOLOPT-SR-GTH
      POTENTIAL  GTH-PBE-q14
    &END KIND
    &KIND O
      BASIS_SET  DZVP-MOLOPT-SR-GTH
      POTENTIAL  GTH-PBE-q6
    &END KIND
    &KIND K
      BASIS_SET  DZVP-MOLOPT-SR-GTH
      POTENTIAL  GTH-PBE-q9
    &END KIND
&END SUBSYS
&END

--
You received this message because you are subscribed to a topic in the Google Groups "cp2k" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cp2k/Nsy-57sQOJ4/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cp2k+uns...@googlegroups.com.
To post to this group, send email to cp...@googlegroups.com.
Visit this group at https://groups.google.com/group/cp2k.
For more options, visit https://groups.google.com/d/optout.

[Alex]

Okay, nevermind. Indeed, it was the TOPOLOGY section. By simply invoking it to supply the coordinates, I asked cp2k to create, well, the topology. :)

All fixed by converting coordinates to xyz and sticking them into the input file.

Thanks!

Alex

[Gabriele]

Hi Alex,

You can use the topology section to import the coordinates from an external pdb or xyz file, which may be useful especially for large systems. Try to insert CONNECTIVITY OFF inside the TOPOLOGY section. You should have no issues associated with the generation of the connectivity...

Cheers,

Gabriele

[Alex]

Hi Gabriele,

I wish I looked at the TOPOLOGY section description more carefully -- apparently CONN_FILE_FORMAT (with the alias you mentioned) is indeed the key. Why it sets the default to generating a topology when DFT is used to calculate interactions is a bit beyond me. In any case, thank you!

Alex