[CP2K-user] SCF does not converge for a single atom of cerium

[Gabriel Reynald]

Hi Jonathan,

Did you try to use smearing? It's very useful for metallic systems. See the 
final of the tutorial of Energy and Force here 
https://www.cp2k.org/howto:static_calculation. There's shows how to use it.
Em segunda-feira, 29 de março de 2021 às 07:02:56 UTC-3, Matt W escreveu:

> Couple of comments. This is actually a rather difficult calculation to get 
> correct - the full solid state calcs will probably me much easier to 
> converge.
> (i) your grid cuttoff is probably far to low for a q30 pseudo. You could 
> also try uing GAPW method (METHOD GAPW in the QS section)
> (ii) Your cell is probably too small for the MT poisson solver try 
> increasing the size of the cell. WAVELET solver may be more efficient for 
> atomic calcs.
> (iii) ALPHA might be too big - try reducing it. Maybe then restart and 
> increase a little, or try OT method.
> Matt
>
> On Monday, March 22, 2021 at 2:08:15 PM UTC jona... at gmail.com wrote:
>
>>
>> Dear cp2k developers,
>>  I am trying to test the calculation using a single atom of cerium. The 
>> coordinates should be ok, I am using a single atom at the origin of a box 
>> of 5 A of side length. Also both basis set and pseudopotential should be 
>> ok, I have found them in the data directory. I do not understand why the 
>> SCF does not converge. Surely, there is an error of mine but I cannot see 
>> it. I have put the muliplicity to 3, because in its electronic 
>> configuration Ce has 2 unpaired. electrons. This is my input.
>> &GLOBAL
>>   PROJECT Ce
>>   RUN_TYPE ENERGY
>>   PRINT_LEVEL LOW
>> &END GLOBAL
>> &FORCE_EVAL
>>   METHOD QS
>>   &DFT
>>     BASIS_SET_FILE_NAME BASIS_MOLOPT_UCL
>>     POTENTIAL_FILE_NAME GTH_POTENTIALS
>>     CHARGE 0
>>     MULTIPLICITY 3
>>     &MGRID
>>       CUTOFF 550
>>       REL_CUTOFF 80
>>     &END MGRID
>>     &QS
>>       EPS_DEFAULT 1.0E-12
>>       EXTRAPOLATION PS
>>       EXTRAPOLATION_ORDER 3    
>>     &END QS
>>    &SCF
>>       SCF_GUESS RESTART
>>       EPS_SCF 1.0E-7
>>       MAX_SCF 500
>>       &MIXING T
>>         METHOD BROYDEN_MIXING
>>         ALPHA 0.4
>>         NBROYDEN 8
>>       &END MIXING
>>       
>>       &DIAGONALIZATION
>>         ALGORITHM STANDARD
>>       &END DIAGONALIZATION
>>     &END SCF
>>    &XC
>>        DENSITY_CUTOFF     1.0000000000000000E-10
>>        GRADIENT_CUTOFF     1.0000000000000000E-10
>>        TAU_CUTOFF     1.0000000000000000E-10
>>        &XC_FUNCTIONAL  NO_SHORTCUT
>>         &PBE
>>             PARAMETRIZATION ORIG
>>         &END PBE
>>        &END XC_FUNCTIONAL
>>      &END XC  
>>     &POISSON
>>       PERIODIC NONE
>>       POISSON_SOLVER MT
>>     &END POISSON
>>   &END DFT
>>   &SUBSYS
>>     &CELL
>>       ABC 5. 5. 5.
>>       PERIODIC NONE
>>     &END CELL
>>     &COORD
>>     Ce 0.00 0.00 0.00
>>     &END COORD
>> &KIND Ce
>>       ELEMENT   Ce
>>       BASIS_SET DZVP-MOLOPT-SR-GTH-q30
>>       POTENTIAL GTH-PBE-q30
>>     &END KIND
>> &END SUBSYS
>> &END FORCE_EVAL
>> Have a nice day,
>> Jonathan
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20210413/b495df18/attachment.htm>