Total dipole/Wannier centers print in cp2k 2.3

975 views
Skip to first unread message

Jana

unread,
Jun 27, 2013, 8:28:39 PM6/27/13
to cp...@googlegroups.com
Dear all,
I've been told I can get a total dipole file containing delta dipoles and these can be postprocessed via FT from a CPMD utility (and unit conversion) to obtain vibrational frequencies.

(This is an attempt to get a comparison as I've got reasonable IR spectrum simply from NORMAL_MODES where all the modes fit but for some reason the only one most characteristic frequency is strangely shifted like 500 cm-1 to low frequencies...this happens only in one of my several model systems. I've been trying to improve convergence criteria or basis sets or adding vdw correction but either no help or no convergence even after like more than one week on the wall...)

To get this I should run MD and print out TOTAL_DIPOLE, with reference to centers of atomic charges. I've tried to get dipoles and wannier centres according to what I've found while searching this google group, but I have only managed to get it run with ENERGY calculation, not MD, and as it is just a trial and error I got a bit lost and cannot really see what am I doing wrong. Could you please somebody help?

...
&MOTION
 &MD
    ENSEMBLE NVT
    STEPS 1000
    TIMESTEP 0.5
    TEMPERATURE   300.0
    &THERMOSTAT
       TYPE NOSE
       REGION GLOBAL
       &NOSE
         LENGTH 3
         YOSHIDA 3
         MTS 2
       &END
    &END
  &END MD
&END MOTION

&FORCE_EVAL
  METHOD QS
  &DFT
&LOCALIZE
    METHOD CRAZY
    OPERATOR BERRY
 &END LOCALIZE
&PRINT
 &LOCALIZATION
        &WANNIER_CENTERS MEDIUM
                FILENAME =water2.wannier
                &EACH
                MD 1
                &END EACH
        ADD_LAST
        COMMON_ITERATION_LEVELS 3
        &END

        &MOLECULAR_DIPOLES HIGH
        &END MOLECULAR_DIPOLES

        &TOTAL_DIPOLE MEDIUM
                FILENAME =water2.dip
                &EACH
                MD 1
                &END EACH
        REFERENCE COM
        PERIODIC TRUE
        COMMON_ITERATION_LEVELS 3
        &END TOTAL_DIPOLE
 &END
&END PRINT

   BASIS_SET_FILE_NAME GTH_BASIS_SETS
    POTENTIAL_FILE_NAME GTH_POTENTIALS
    &MGRID
      CUTOFF [Ry] 450
    &END MGRID
    &QS
      METHOD GPW
    &END QS
    &SCF
      SCF_GUESS RESTART
      MAX_SCF 100
      EPS_SCF 1.0e-7
        &OUTER_SCF
        MAX_SCF 5
        EPS_SCF 1.0E-7
        &END
    &END SCF
    &XC
      &XC_FUNCTIONAL PBE
      &END XC_FUNCTIONAL
        &XC_GRID
                XC_SMOOTH_RHO NN50
                XC_DERIV NN50_SMOOTH
        &END XC_GRID
    &END XC
&END DFT
...

I've also tried it with no specification of filename or each or common iteration levels, also for nonperiodic case with MT poisson solver but I never get any file or part of the output mentioning dipole or wannier centres. Quite sure the mistake is something really stupid, but I've been just spending few nights on it and still cannot figure it out. 
Any help would be much appreciated!
J.

Jana

unread,
Jun 27, 2013, 8:38:05 PM6/27/13
to cp...@googlegroups.com
(just noticed few typos - 
ADD_LAST NUMERIC and  REFERENCE COAC)

Message has been deleted

JQ

unread,
Jul 4, 2013, 3:15:57 PM7/4/13
to cp...@googlegroups.com
Hi Jana,

I have just done a test run on my pc based on your input. Your input seems working fine that I manage to get the output of delta dipole moment and etc, saved into the files as defined. You may refer to the attached input file and output file (only few MD steps), for your reference. I am using CP2K 2.3 and you can find its details in the output file

JQ
h2o.xyz
md-dipole.inp
md-dipole.out

Wagner Homsi Brandeburgo

unread,
Jul 15, 2013, 8:00:17 AM7/15/13
to cp...@googlegroups.com
Dear Jana, Jing,

probably by now you have already solved the problem, but I was just faced with the same
issue and I've just figured out what happened. Apparently, there was a minor change from 2.3, 2.4 (development version) and 2.4 (trunk version), my guess is that you're using (2.4 trunk). On this version the location of the subsections related to printing dipole moments and wannier centers have changed on the input file. Now, you don't need the LOCALIZATION section anymore, you can just start the subsection TOTAL_DIPOLE inside LOCALIZE%PRINT. 

In your case, just delete the following lines:
 &LOCALIZATION
and
&END (refering to LOCALIZATION)

and move &END LOCALIZE to after &END PRINT

cheers,
Wagner.

ps. if you keep the old sintax the code won't complaint, it'll just run normally without computing it, maybe it's something for the developers to fix.




--
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 post to this group, send email to cp...@googlegroups.com.
Visit this group at http://groups.google.com/group/cp2k.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

Jana

unread,
Jul 15, 2013, 6:10:22 PM7/15/13
to cp...@googlegroups.com
Dear Wagner,

it is sorted, thanks for the remark though.



I am a bit puzzled by the post-processing - give time difference in input file (in atomic units):
 [CAUTION: this is the MD time step times the sampling frequency.]

The sample value in the fourier transform utility was 17. My time step is 0.5 fs but how do I find out what the sampling frequency is? I understand I should find out after how many samples I save one, but how do I get that?

Originally I used the Used time column in *ener file but JQ found out it wasn't it.


Sandeep Kumar Reddy

unread,
Jul 16, 2013, 12:54:21 AM7/16/13
to cp...@googlegroups.com
Hi,
     It should be calculated as follows. 

timestep*dumping_frequency

timestep = 0.5 fs =  20.665 a. u. 

dump_frequency = how often are you calculating the dipole moment and dumping it to the output file. From your input file, i believe it is 1. 

so sampling frequency =  20.665*1 = 20.665   


Hope this helps.

Regards,
Sandeep

PS: You need not calculate dipole moment of the system at every MD step. To save CPU time, you can calculate dipole moment at every 5th MD step.  In this case, the sampling frequency is 103.325 (20.665*5). 







Reply all
Reply to author
Forward
0 new messages