[CP2K-user] [CP2K:21795] Re: About SCF convergence：Can first perform a pre-optimization using a larger EPS_SCF?

[Konstantin Tokarev]

среда, 3 сентября 2025 г. в 17:14:05 UTC+3, Charlotte Lai: 

Dear CP2K community,
While optimizing the structure of a triangular MoS₂ monolayer, I find that 
the SCF convergence cannot be pushed below the originally set EPS_SCF = 1 × 
10⁻⁶. Could experienced users advise whether it is acceptable to first 
carry out a coarse geometry optimization with EPS_SCF = 5 × 10⁻⁶, and then 
restart the calculation on the resulting structure with the tighter 
criterion EPS_SCF = 1 × 10⁻⁶ for a more accurate refinement?
Thank you in advance for any suggestions.


In my opinion, EPS_SCF 1E-6 is already very coarse (I normally use 1E-8). 
If you are having convergence troubles with 1E-6, it likely indicates 
numeric instability in SCF, which, for example, may be caused by basis set 
being too diffuse for the system under investigation.

-- 
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/852fd55c-ca93-4992-ad01-43a8c1ae6851n%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20250903/78650942/attachment-0001.htm>