[CP2K-user] Solid state DFT and vibrational analysis

[Matt W]

CP2K is a perfectly fine force / energy calculator. Lack of k-points and 
symmetry might make things cumbersome. There is / used to be a wrapper for 
phonopy which can be used to drive the phonon calculations. Lev 
Kantorovich's Tetr / Lev00 also have some functionality.

Have a look at this:

https://www.cp2k.org/exercises:2017_uzh_cmest:phonon_calculation

it might be useful.

Matt

On Tuesday, April 30, 2019 at 2:51:00 AM UTC+1, bobs your uncle wrote:
>
> Hi folks,
>
> Just a general question regarding vibrational analysis in an organic 
> crystal using CP2K (just solid state DFT ignoring AIMD for now).
> I thought about using more specialized software like quantum espresso - 
> although I would prefer to use Crystal which I don't have access to 
> currently - but the possibility of using CP2K for both solid state DFT, and 
> later AIMD was more appealing to me for obvious reasons.
> So first off, would you recommend CP2K for vibrational analysis of 
> periodic systems? It would seem to me that it is uncommon, as I am having 
> issues finding examples in the literature, and there aren't any guides 
> either.
>
> To get to grips with CP2K, I first completed some of the tutorials as 
> recommended before moving onto a "toy" system, where I have been attempting 
> to reinvent the wheel with vibrational analysis. 
> Of course, I would really like to know if anyone could advise me on the 
> "right" way to do this. 
> Currently, I start with a cell optimisation on an experimental crystal 
> structure - following the sodium chloride guide on the CP2K website - then 
> reoptimize the geometry with tighter convergence threshholds etc. 
> I take that geometry and perform a vibrational analysis, and have 
> succeeded in producing a complete set of real harmonics this way - which is 
> a start - but I haven't been able to simulate some of the low frequency 
> "terahertz" modes that have been reproduced in the literature (although for 
> my system, the papers I can find don't talk at all about the software or 
> process followed).
> No one in my immediate academic circle is trained in these calculations, 
> so I would really appreciate some of your wisdom if you could share it!
>
> Sorry, that was a bit of a read, but I'll put input files up if you are 
> interested.
>
> Cheers
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20190430/38d5fc9d/attachment.htm>