[Pw_forum] Phonons with DFPT and DF2 functional

Florian Altvater altvater at berkeley.edu
Fri Nov 7 09:53:06 CET 2014


Thanks for the clarification!

Florian

Am 07.11.2014 um 00:05 schrieb Paolo Giannozzi:
> On Thu, 2014-11-06 at 13:32 -0800, Florian Altvater wrote:
>
>> just to clarify, do Grimme-D2 or the Tkatchenko-Scheffler correction
>> fall under the "vdW of any kind" category
> yes
>
>> or are they implemented in 5.1 for DFPT?
> no. Any new develoment in the self-consistent code that contributes 
> an energy term has to be explicitly added to the DFPT calculation. 
>
>> Also, I don't feel comfortable enough with the code yet to be able to
>> find that information in the source, but I am sure it is in there
>> somewhere. 
> in the header of the phonon.f90 code (although in a format whose logic
> I never remember). Alternatively: if you know a key variable related
> to some new development, and this variable is never used in the phonon
> code, it is almost certain that the new development hasn't percolated
> to the phonon code.
>
> Paolo
>
>> If you want to give me any pointers, so that I could
>> potentially answer those questions myself in the future, I would greatly
>> appreciate it.
>>
>> Thanks,
>> Florian
>>
>>
>> Am 05.11.2014 um 11:06 schrieb Paolo Giannozzi:
>>> On Wed, 2014-11-05 at 03:51 -0800, Florian Altvater wrote:
>>>> Hi,
>>>> I am interested in looking at effects of using van der Waals functionals
>>>> like vdW-DF2 on the phonon dispersion of organic crystals. Preliminary
>>>> results indicate that there might be some issues with using DFPT and DF2
>>>> together. More specifically, DFPT gives a couple "very imaginary"
>>>> frequencies (below -100cm^-1) for Gamma-point phonons (with and without
>>>> acoustic sum rule), while finite displacement results seem to be just
>>>> fine, using the same convergence parameters.
>>>>
>>>> So before I dig deeper into eventual other issues at play here, I wanted
>>>> to ask if you know of any incompatibilities when using DFPT and vdW-DF2?
>>> DFPT + vdW (any kind) is not currently implemented. The next version
>>> (out in a few days) will contain a check on this
>>>
>>> Paolo
>>>
>>>> Either inherently embedded in the underlying theory (which I could then
>>>> dig into first), or just in the way it is implemented in QE.
>>>>
>>>> Thanks for your help,
>>>> Florian
>>>> _______________________________________________
>>>> Pw_forum mailing list
>>>> Pw_forum at pwscf.org
>>>> http://pwscf.org/mailman/listinfo/pw_forum
>> ---
>> Florian Altvater
>> PhD candidate
>> Neaton group
>> UC Berkeley/LBNL
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://pwscf.org/mailman/listinfo/pw_forum




More information about the users mailing list