[CP2K-user] [CP2K:21472] Re: Clarification on k-point sampling, slab passivation, and convergence in CP2K

[FS]

Hi
If possible, I would add on these questions as I had overlapping/similar 
ones.

A. SCF Convergence:

a-What are the typical variables in cp2k that make sense to play around 
with, with trailing convergence? Are there some rule of thumbs that work 
well for values/ratios for ALPHA and NBROYDEN in the SCF/MIXING block? 

b-I notice that i run into convergence issues with the SCCS solvation (and 
to an lesser extend the gCP correction, guess noise). Any tips for 
parameters for the SCCS block here?

&SCCS on
&ANDREUSSI
RHO_MAX 0.001
RHO_MIN 0.0001    
&END ANDREUSSI
      DIELECTRIC_CONSTANT 40
      EPS_SCF 1.0e-3
      EPS_SCCS 3.0E-6
      MAX_ITER 300
    &END SCCS

c-As I come from non-periodic calcs: fermi smearing for convergence, what 
are acceptable temperatures to use? 

B. Adding to the Fermi Smearing:

Similar to molecular systems with strong static electron correlation (SEC) 
I looked into in the past, it might make sense for me to use FT-DFT as 
described in Grimme's FOD papers at smear temps of 20000K*%HFexchange + 
5000K in geometry optimizations, freqs and single points to account for 
strong SEC effects. Is there anything in cp2k that would make FT_DFT behave 
differently/unexpectedly for these calcs as in programs like ORCA?

C. Is there like a good practice reference I could look into for periodic 
calcs, similar to 
https://onlinelibrary.wiley.com/doi/10.1002/anie.202205735 for mostly 
non-periodic calcs?

All the best
usman ul muazzam schrieb am Freitag, 16. Mai 2025 um 13:14:43 UTC+2:

> Dear CP2K Users and Developers,
>
> I am trying to model *Ga and O adatom diffusion barriers* on substrates 
> like *diamond and sapphire* using CP2K (periodic DFT + NEB), and I have a 
> few questions regarding best practices:
>
> *1. k-point Sampling:*
> Some CP2K publications seem to use *only Gamma-point*, even for periodic 
> systems. In contrast, codes like Quantum ESPRESSO require k-point 
> convergence for accuracy.
>
>    - 
>    
>    Is k-point convergence generally *not needed* in CP2K.
>    - 
>    
>    When should one use explicit *k-point grids* in CP2K.
>    
> *2. Slab Model Passivation:*
> In plane-wave codes, bottom surface passivation (e.g., with hydrogen) is 
> common in slab models to avoid spurious states. But I rarely see this in 
> CP2K-based surface studies.
>
>    - 
>    
>    Is *bottom-layer passivation not necessary* in CP2K.
>    
> *3. Convergence Parameters:*
> Is optimizing *CUTOFF* and *REL_CUTOFF* typically *sufficient* for 
> accurate periodic DFT and NEB calculations in CP2K.
>
>    - 
>    
>    Are other parameters like *basis sets (as here we use Gaussian instead 
>    of plane wave)* or *grid levels* also important to converge.
>    - 
>    
>    Any recommended references for CP2K convergence strategies.
>    
> Any guidance or pointers to best practices would be greatly appreciated.
>
>
> Best Regards,
>
> Usman
>

-- 
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 visit https://groups.google.com/d/msgid/cp2k/f0d94d2a-93d1-41ee-b8d6-764a38035ab5n%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20250516/67d1c755/attachment.htm>