[CP2K-user] [CP2K:18222] MOF cell opt with KEEP_SYMMETRY FALSE

Corrado Bacchiocchi bacchio at gmail.com
Tue Dec 20 10:49:51 UTC 2022


Dear Matthias,
thanks a lot for the information and the suggestion.
Interestingly, the CELL_OPT with KEEP_SYMMETRY FALSE does not change the 
angles that remain perfectly at 90 deg at all steps, so, I am afraid that by 
applying the constraint KEEP_ANGLES might not going to help in this case, 
since they are already not changing, but I will try and let you know. It 
might be a useful information to the community.
Thanks again!
Corrado


On Tuesday, December 20, 2022 at 11:24:42 AM UTC+1 Matthias Krack wrote:

> Hi Corrado
>
>  
>
> CP2K does not take advantage of symmetry for better computational 
> efficiency. The constraints are rather meant to preserve or enforce a 
> certain symmetry during structure relaxations, because the symmetric 
> structure is not the exact minimum or it distorts slightly due to numerical 
> noise. The latter is the reason why it can make sense to use the constraint 
> KEEP_ANGLES which will keep all cell angles exactly at 90 degree, because 
> CP2K/Quickstep is faster for orthorhombic cells. Therefore, based on the 
> information which you provided for your system, I would apply only the 
> constraint KEEP_ANGLES in the runs with small molecules adsorbed.
>
>  
>
> HTH
>
>  
>
> Matthias
>
>  
>
> *From: *cp... at googlegroups.com <cp... at googlegroups.com> on behalf of 
> Corrado Bacchiocchi <bac... at gmail.com>
> *Date: *Tuesday, 20 December 2022 at 09:54
> *To: *cp2k <cp... at googlegroups.com>
> *Subject: *[CP2K:18220] MOF cell opt with KEEP_SYMMETRY FALSE
>
> Dear Colleagues,
>
> I ask you to kindly share your experience about the following question.
>
>  
>
> We have performed on our MOF structure the "Converge CUTOFF and 
> REL_CUTOFF" procedure, as described in the very helpful page:
>
>  
>
> https://www.cp2k.org/howto:converging_cutoff
>
>  
>
> and found that, with  NGRIDS 4, to reach a 1.E-7 difference with respect 
> to the limiting value of the energy, the minimum settings are:
>
>       
>       CUTOFF 1800
>       REL_CUTOFF 80
>
>  
>
> We have then successfully performed CELL_OPT on a 2x2x2 supercell of our 
> MOF with KEEP_SYMMETRY TRUE.
>
> When we use this optimized structure for a subsequent CELL_OPT with 
> KEEP_SYMMETRY FALSE the BFGS algorithm is not able to further decrease the 
> energy. The structure remains essentially unmodified across hundreds of 
> steps and the optimization does not complete.
>
> I attach the input file and the guess structure and I kindly ask the 
> following:
>
> - is it essential, for any publication quality structure, that 
> KEEP_SYMMETRY is FALSE or is it acceptable to have this constraint in place?
>
> - more in general, for any subsequent calculation, e.g. interaction of 
> small molecules on the surface of the MOF, Nudged Elastic Band energy path 
> investigations, etc., is it possible/acceptable to KEEP_SYMMETRY TRUE?
>
>  
>
> Thank you very much for your kind help
>
> Have a nice Holiday Season!
>
> Corrado
>
> -- 
> 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... at googlegroups.com.
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/cp2k/14f06834-79c7-4ac8-922c-d1d6f008d27fn%40googlegroups.com 
> <https://groups.google.com/d/msgid/cp2k/14f06834-79c7-4ac8-922c-d1d6f008d27fn%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
>

-- 
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+unsubscribe at googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/d5f1d6ed-fabb-47ff-bb26-7564ab762a6fn%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20221220/55fef0eb/attachment-0001.htm>


More information about the CP2K-user mailing list