[CP2K-user] [CP2K:19051] Re: Convergence problem transition metal oxide

[Diego López]

Hi Marcella,

Thanks for your answer. I think you are right because non spin-polarized 
calculations are converging but the spin-polarized does not, so I assume 
there is a problem with magnetization. I am using &BS section to describe 
the valence electrons of Co. Do you think I should move to MAGNETIZATION 
keyword?

Regards,
Diego

El lunes, 3 de julio de 2023 a las 12:55:27 UTC+2, Marcella Iannuzzi 
escribió:

> Hi Diego
>
> It is hard to guess what is going on without more information. 
> Probably you need to explore more values of the U+J parameters or a 
> different way to initialise the magnetization.
>
> Regards
> Marcella
>
>
> On Monday, July 3, 2023 at 12:46:13 PM UTC+2 Diego López wrote:
>
>> Hello Marcella,
>>
>> Thank you for your answer. I have tried also with bigger supercells 
>> (3x3x3 and 5x5x5) but there is not convergence. Also I tried to use 
>> different sets of U - J values (2 eV seems to be wide used in bibliography) 
>> but the result is the same. Any further recommendations?
>>
>> Regards,
>> Diego
>>
>> El lunes, 3 de julio de 2023 a las 11:36:33 UTC+2, Marcella Iannuzzi 
>> escribió:
>>
>>> Dear Diego
>>>
>>> A box of  5.938 5.945 5.938 is very small and not properly representing 
>>> the bulk system, if only Gamma point is considered. 
>>> Often for transition metal oxides DFT+U turns out to help, probably you 
>>> can find in the literature whether this is the case of the oxides you are 
>>> interested in. 
>>>
>>> Regards
>>> Marcella
>>>
>>> On Monday, July 3, 2023 at 10:15:37 AM UTC+2 Diego López wrote:
>>>
>>>> Dear CP2K users,
>>>>
>>>> I am currently working with a series of transition metal oxides and 
>>>> CP2K is working fine but when I try to converge a calculation on the mixed 
>>>> valence Cu oxide (Cu3O4) my calculations are not converging. I have read 
>>>> that this oxide is a tricky system for DFT, so I would be really grateful 
>>>> if you could give me some suggestions to solve this problem. I attatch the 
>>>> input file that I am using. I tried to use different mixing, 
>>>> diagonalization and smearing methods with the same problem.
>>>>
>>>> &GLOBAL
>>>>   PROJECT Co3O4
>>>>   RUN_TYPE ENERGY
>>>>   PRINT_LEVEL MEDIUM
>>>> &END GLOBAL
>>>>
>>>> &FORCE_EVAL
>>>>   METHOD Quickstep
>>>>   &DFT
>>>>     BASIS_SET_FILE_NAME  BASIS_MOLOPT_UZH
>>>>     POTENTIAL_FILE_NAME  POTENTIAL_UZH
>>>>     UKS
>>>>     MULTIPLICITY 1
>>>>     &POISSON
>>>>       PERIODIC XYZ
>>>>     &END POISSON
>>>>     &XC
>>>>       &XC_FUNCTIONAL PBE
>>>>        &PBE
>>>>         PARAMETRIZATION PBESOL
>>>>        &END
>>>>       &END XC_FUNCTIONAL
>>>>     &END XC
>>>>     &MGRID
>>>>       NGRIDS 4
>>>>       CUTOFF 800
>>>>       REL_CUTOFF 50
>>>>     &END MGRID
>>>>     &QS
>>>>       METHOD GPW
>>>>       EXTRAPOLATION ASPC
>>>>       EPS_DEFAULT 1E-014
>>>>     &END QS
>>>>     &SCF
>>>>       SCF_GUESS ATOMIC
>>>>       EPS_SCF 1.0E-7
>>>>       ADDED_MOS 20
>>>>       MAX_SCF 20
>>>>       &OUTER_SCF
>>>>        MAX_SCF 300
>>>>        EPS_SCF 1.0E-7
>>>>       &END
>>>>       &SMEAR ON
>>>>         METHOD FERMI_DIRAC
>>>>         ELECTRONIC_TEMPERATURE [K] 3000
>>>>       &END SMEAR
>>>>       &DIAGONALIZATION
>>>>         ALGORITHM STANDARD
>>>>       &END DIAGONALIZATION
>>>>       &MIXING
>>>>         METHOD BROYDEN_MIXING
>>>>         ALPHA 0.02
>>>>         BETA 0.01
>>>>       &END MIXING
>>>>      &END SCF
>>>>   &END DFT
>>>>   &SUBSYS
>>>>     &CELL
>>>>       ABC 5.938 5.945 5.938
>>>>       ALPHA_BETA_GAMMA 60 60 60
>>>>       PERIODIC XYZ
>>>>     &END CELL
>>>>     &TOPOLOGY
>>>>       COORD_FILE_FORMAT xyz
>>>>       COORD_FILE_NAME  ./Co3O4-2.xyz
>>>>     &END
>>>>     &KIND O
>>>>       ELEMENT O
>>>>       BASIS_SET DZVP-MOLOPT-PBE-GTH-q6
>>>>       POTENTIAL GTH-PBE-q6
>>>>       &BS
>>>>       &ALPHA
>>>>       N     2
>>>>       L     1
>>>>       NEL  +2
>>>>       &END ALPHA
>>>>       &BETA
>>>>       N     2
>>>>       L     1
>>>>       NEL  +2
>>>>      &END BETA
>>>>      &END BS
>>>>     &END KIND
>>>>     &KIND Co
>>>>       ELEMENT Co
>>>>       BASIS_SET DZVP-MOLOPT-PBE-GTH-q17
>>>>       POTENTIAL GTH-PBE-q17
>>>>      &BS
>>>>       &ALPHA
>>>>        N    4  3
>>>>        L    0  2
>>>>        NEL -2 -1
>>>>       &END ALPHA
>>>>       &BETA
>>>>        N    4  3
>>>>        L    0  2
>>>>        NEL -2 -1
>>>>       &END BETA
>>>>       &END BS
>>>>       &DFT_PLUS_U
>>>>        L 2
>>>>        U_MINUS_J [eV] 2.0
>>>>      &END DFT_PLUS_U
>>>>     &END KIND
>>>>     &KIND Co1
>>>>       ELEMENT Co
>>>>       BASIS_SET DZVP-MOLOPT-PBE-GTH-q17
>>>>       POTENTIAL GTH-PBE-q17
>>>>      &BS
>>>>       &ALPHA
>>>>        N    4  3
>>>>        L    0  2
>>>>        NEL -2 -3
>>>>       &END ALPHA
>>>>       &BETA
>>>>        N    4  3
>>>>        L    0  2
>>>>        NEL -2 +3
>>>>       &END BETA
>>>>       &END BS
>>>>       &DFT_PLUS_U
>>>>        L 2
>>>>        U_MINUS_J [eV] 2.0
>>>>      &END DFT_PLUS_U
>>>>     &END KIND
>>>>     &KIND Co2
>>>>       ELEMENT Co
>>>>       BASIS_SET DZVP-MOLOPT-PBE-GTH-q17
>>>>       POTENTIAL GTH-PBE-q17
>>>>      &BS
>>>>       &BETA
>>>>        N    4  3
>>>>        L    0  2
>>>>        NEL -2 -3
>>>>       &END BETA
>>>>       &ALPHA
>>>>        N    4  3
>>>>        L    0  2
>>>>        NEL -2 +3
>>>>       &END ALPHA
>>>>       &END BS
>>>>       &DFT_PLUS_U
>>>>        L 2
>>>>        U_MINUS_J [eV] 2.0
>>>>      &END DFT_PLUS_U
>>>>     &END KIND
>>>>   &END SUBSYS
>>>> &END FORCE_EVAL
>>>>
>>>> Thank you in advanced.
>>>> Best regards,
>>>> Diego.
>>>>
>>>

-- 
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 on the web visit https://groups.google.com/d/msgid/cp2k/b5290a11-81b8-4681-a822-6fef8d7882fen%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20230703/2e76e423/attachment.htm>