SCF calculation can not converge

[ce fu]

Dear CP2K community,

      I am a new CP2K's user and recently started to CP2K for calculating a periodic system, like alloy slab. I need to run a geometry optimization, (before this opt step , I had tried use same structure without vacuum ,it can be converged )but when I added vacuum slab to run, it couldn’t to converge .I have ask some people to get some solution for this problem, someone told me to use SURFACE_DIPOLE_CORRECTION parameter in the &DFT keyword ,decline ALPHA 0.4 to 0.3 ,and add BETA 0.5 in the &MIXING keyword. Unfortunately, it also couldn’t get converged.

       Now I don’t know how to fix this problem ,to be a toddler ,I hope some expert can tell me what I should do or learn .   I will be very glad for any response.

[Fabian Ducry]

Hi,

You should further decrease ALPHA and increase ELECTRONIC_TEMPERATURE in &SMEAR. Only activate SURFACE_DIPOLE_CORRECTION once the SCF is converged without it and you can use a restart wavefunction.

With ELECTRONIC_TEMPERATURE 3000 and ALPHA 0.05 the first SCF converges.

Cheers

--
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...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/241189ca-b375-495e-8a14-3e93b2bde597n%40googlegroups.com.

[ce fu]

Hi, 

I am very glad to receive your reply.Now it is running ,and to see the first SCF result.

I have a question about &MIXING section , what should I understand ALPHA and BETA parameter? I want to know what  will be  controlled by this parameter?

My friend told me for my metal system the ALPHA parameter is not  better lower than 0.3 and should add the BETA 0.5 .  I don't why.

Best regrads!