[CP2K-user] Difference between DZVP-MOLOPT-GTH and DZVP-MOLOPT-SR-GTH-q18?

[Travis]

Hi,

Specifically the ones you listed, yes there would be a tangible difference 
between SR and non-SR variants. SR basis sets have greatly reduced 
diffusivity to the basis functions, making them faster to reach 
convergence. They come at a small cost to accuracy in terms of basis set 
incompleteness errors. However, DZVP-MOLOPT-SR-GTH and 
DZVP-MOLOPT-SR-GTH-q18 mean the same thing. The basis set format for CP2K 
is as follows,

Element symbol  Name of the basis set  Alias names

So DZVP-MOLOPT-SR-GTH-q18 is the alias for the basis set 
DZVP-MOLOPT-SR-GTH. q18 refers to the number of valence electrons. The 
convention in the POTENTIAL file is just the opposite,

Ni GTH-PBE-q18 GTH-PBE

-T

On Wednesday, August 28, 2019 at 3:00:07 PM UTC-3, Stephen Vicchio wrote:
>
> All, 
>
> I am new to CP2K and still setting up my calculations. I noticed something 
> the other day in the following talk that is confusing me (on slide 12): 
> https://www.cp2k.org/_media/events:2015_cecam_tutorial:ling_basis_pseudo.pdf
> . 
>
> When selecting a MOLOPT basis set, is there a difference between Ni 
> DZVP-MOLOPT-GTH and DZVP-MOLOPT-SR-GTH-q18? The presentation mentions if I 
> call DZVP-MOLOPT-SR-GTH-q18 that there will be 18 valence electrons 
> included in the pseudopotential; is that correct? If I call DZVP-MOLOPT-GTH 
> there would be no valence electrons in the pseudopotential? I've tried 
> searching online for an answer, but I haven't been able to find anything 
> helpful.
>
> Also, I've read the original paper from VandeVondele and Hutter (2007). 
> Does anyone have any other recommended readings on the MOLOPT basis sets? 
>
> Thank you! 
>
> Stephen 
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20190828/82df086d/attachment.htm>