Energy changes abnormally in OT method during the GEO_OPT

[Xiaokun]

Hi，everyone！

I am new one to CP2K and I am doing a Geo-optimization in a MOF. In the input file, I use the OT method, constrain the most atomes obtained from the experiment and relax the NO3- in the MOF cell. I want to find the resonable positions of the NO3- , cause the experiment cannot sure where NO3- it is.  And there seems to some errors in my output file. The energy changes abnomally in a scf. Did I set some keywords wrong？Can anyone give me some suggestions about my claculation? Attached are my input and output files. Thank you.

Best wishes.

Xiao-kun

 

[Patrick Gono]

Dear Xiao-kun,

Looking at your output, I noticed the value of the convergence parameter to be "stuck" in some of the SCF loops. I have encountered this issue while using the FULL_ALL preconditioner. Switching to FULL_SINGLE or FULL_SINGLE_INVERSE solved the problem for me, although I did not dig deeper to understand what the actual underlying issue was that caused this behaviour.

In my case, the system was semi-metallic under PBE, with a very low band gap. Even though I used a tiny value for ENERGY_GAP, and a near-optimized starting geometry, some SCF loops turned pathological very much like in your case. Switching to FULL_SINGLE apparently resolved the issue.

I am also open to any further suggestions or explanations by other experts in the community, though.

Yours sincerely,

Patrick Gono

--
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/bc268d58-7cab-4c7c-83ba-71bf8d74726f%40googlegroups.com.

[Xiaokun]

Dear Patrick Gono,

      

Thanks for your suggestions. I will try your recommendation. 

Yours sincerely,

Xiao-kun

在 2019年11月10日星期日 UTC+8上午1:52:13，Patrick Gono写道：

To unsubscribe from this group and stop receiving emails from it, send an email to cp...@googlegroups.com.

[Xiaokun]

Dear Travis,

thanks for your reply, I will have a try.

yours,

Xiao-kun

在 2019年11月10日星期日 UTC+8上午4:35:22，Travis写道：

Hi,

For high spin materials like this, I recommend switching away from the OT method. Sometimes the combination of high spin and use of GGAs makes your system half-metallic.

  &SCF
   EPS_SCF 1e-06
   MAX_SCF 200
   SCF_GUESS RESTART
   ADDED_MOS  200
   &OUTER_SCF  F
   &END OUTER_SCF
   &SMEAR  T
    METHOD  FERMI_DIRAC
    ELECTRONIC_TEMPERATURE     3.0000000000000000E+02
   &END SMEAR
   &MIXING  T
    METHOD  BROYDEN_MIXING
    ALPHA     4.0000000000000002E-01
    BETA     1.5000000000000000E+00
    NMIXING  5
    NBUFFER  8
   &END MIXING
  &END SCF

-T

[Xiaokun]

Hi Travis,

I have tried your recommended DIAG method for my system. But unfortunately it failed. Even though I changed  EPS_SCF 1e-05 and MAX_SCF 400, it's hard to converge. The energy change seems to be very large (1E0~1E-1) and the 'Convergence' also is a large number during the SCF WAVEFUNCTION OPTIMIZATION progress.  How can I solve this problem?

Yours,

Xiao-kun