[CP2K-user] Band Gap Error for substituted compounds
AFRID MAZHAR SHIRSEKAR
ash... at g.clemson.edu
Tue Oct 27 20:09:17 UTC 2020
Dear Nick,
Thank you for your valuable input, i ran again two sets of calculations
with and without relaxing for 25%hfx and also increased the cutoff radius
to 4 almost half of the lattice constant, but still there was difference of
1ev. My question is, is it fine to not relax the system before calculating
the band gap, since there is an underlying issue in the relaxed
coordinates, will the band gap values be considered reliable?
regards,
afrid
On Tuesday, October 27, 2020 at 12:46:36 PM UTC-4 n... at berkeley.edu wrote:
> There doesn't seem to be any issue with your inputs/outputs.
>
> Are you trying to simulate Sr as a dopant in low concentration? Or are you
> trying to create a (Ba, Sr)TiO3 system with Sr:Ba 1:2? I'm assuming the
> second option.
>
> What I would suggest is printing out all of the unoccupied eigenlevels as
> well as the occupied ones. From there you can see if the Sr substitution is
> creating new states in the gap, or if its part of the VB/CB. What I see in
> your LUMO is a huge shift of 1 eV from the unrelaxed state cell, which is
> strange since the cell parameters are fairly similar. I might also do a
> static calculation on a larger cell to see if that closes the gap; 40 atoms
> is quite small by gamma-point-only standards.
>
> Last thing I would say is that hybrid DFT can exactly reproduce the band
> gap and still get the physics wrong, since the band gap is roughly linear
> with the amount of hfx you mix in to your GGA calculation. For example, you
> have a mixing fraction of 37%, which seems very large (to me) for BaTiO3,
> which has a electronic permittivity in the neighborhood of 6. I would
> expect the fraction to be in the 15-25% range. It's possible that your HF
> contribution is off and you need to fix that before you tune the fraction.
> Ways your HF could be wrong include your truncated PBE0 functional, which
> has a cutoff radius of 3, which is not that large, and so you could be
> disregarding too much of the long range interactions, which is made up for
> by your high mixing fraction, but comes back to bite you when you change
> the system chemistry.
>
> Hope that's enough to think about for now.
>
> -Nick
> On Tuesday, October 27, 2020 at 9:08:44 AM UTC-7 a... at g.clemson.edu
> wrote:
>
>> Dear Nick,
>> Sorry for the troublesome articulation of the message. Yes the points you
>> made are correct
>> PFA,
>> regards,
>> Afrid.
>>
>> On Tuesday, October 27, 2020 at 11:45:44 AM UTC-4 n... at berkeley.edu
>> wrote:
>>
>>> Just to be clear, as I'm having a little trouble following you're
>>> message: the expected outcome is that Sr substitution will decrease the
>>> band gap; you substituted 2 Sr atoms in place of Ba, and this increased the
>>> band gap; when you do a static calculation (no relaxation) the band gap
>>> does decrease, but when you enable relaxation then the band gap increases.
>>>
>>> Are all three of those points correct?
>>>
>>> The next thing I would ask, if that is correct, is to see the output
>>> file.
>>> On Tuesday, October 27, 2020 at 8:02:01 AM UTC-7 a... at g.clemson.edu
>>> wrote:
>>>
>>>> Dear developers,
>>>> I have been trying to study the effect on band gap, when substitutions
>>>> are implemented on BaTiO3 structure. My methodology is that i generate a 2
>>>> x 2 x 2 supercell of BaTiO3 and substitute 2 Sr atoms in the Ba positions,
>>>> and then i conduct cell optimization on the new structure, and then i used
>>>> the converged coordinates file and the unit cell parameters from the
>>>> output to evaluate the band gap.
>>>> The band gap for BaTiO3 is around 3.2 ev, i used 37%hfx to approximate
>>>> the bandgap closer to experimental one for BaTiO3 structure, but when the
>>>> substitutions are implemented the calculation overestimates the band gap by
>>>> 1 ev (4.2ev) which was expected to decrease from the initial value.
>>>> i calculated the bandgap without performing the cellopt and just
>>>> substituted the Sr atoms in the coord file, and the result was in
>>>> accordance with my hypothesis, ie decrease in the band gap.
>>>> I even tried keeping the symmetry during the cellopt calculations, but
>>>> still the result was same. The scf guess parameter is set to atomic
>>>> below i attaching the input files for my calculations
>>>> Thanking you in anticipation
>>>> Sincerely
>>>> Afrid.
>>>>
>>>> &GLOBAL
>>>>
>>>> PROJECT perovskite_bandgap
>>>>
>>>> RUN_TYPE ENERGY
>>>>
>>>> PRINT_LEVEL MEDIUM
>>>>
>>>> &END GLOBAL
>>>>
>>>>
>>>> &FORCE_EVAL
>>>>
>>>> METHOD Quickstep
>>>>
>>>> &DFT
>>>>
>>>> BASIS_SET_FILE_NAME BASIS_file
>>>>
>>>> POTENTIAL_FILE_NAME POTENTIALS_file
>>>>
>>>> BASIS_SET_FILE_NAME BASIS_ADMM_MOLOPT
>>>>
>>>> BASIS_SET_FILE_NAME BASIS_ADMM
>>>>
>>>> WFN_RESTART_FILE_NAME afrid-test-1.restart
>>>>
>>>>
>>>> &PRINT
>>>>
>>>> &MO_CUBES
>>>>
>>>> WRITE_CUBE .FALSE.
>>>>
>>>> NHOMO 1
>>>>
>>>> NLUMO 1
>>>>
>>>> &END
>>>>
>>>> &END
>>>>
>>>> &QS
>>>>
>>>> METHOD GPW
>>>>
>>>> EXTRAPOLATION PS
>>>>
>>>> EXTRAPOLATION_ORDER 3
>>>>
>>>> EPS_DEFAULT 1.0E-10 #E-6
>>>>
>>>> &END QS
>>>>
>>>> &POISSON
>>>>
>>>> PERIODIC XYZ
>>>>
>>>> &END POISSON
>>>>
>>>> &SCF
>>>>
>>>> SCF_GUESS ATOMIC
>>>>
>>>> EPS_SCF 1.0E-6 #E-7
>>>>
>>>> MAX_SCF 30
>>>>
>>>> !EPS_LUMO 1.00000000E-005
>>>>
>>>> !CHOLESKY INVERSE
>>>>
>>>>
>>>>
>>>> &OT
>>>>
>>>> PRECONDITIONER FULL_SINGLE_INVERSE #FULL ALL
>>>>
>>>> MINIMIZER DIIS # minimiser cg
>>>>
>>>> ALGORITHM IRAC #no algorithm before
>>>>
>>>> ENERGY_GAP 0.01
>>>>
>>>> &END
>>>>
>>>> &OUTER_SCF
>>>>
>>>> MAX_SCF 20
>>>>
>>>> EPS_SCF 1e-06
>>>>
>>>> &END OUTER_SCF
>>>>
>>>> &MIXING
>>>>
>>>> METHOD BROYDEN_MIXING
>>>>
>>>> ALPHA 0.2
>>>>
>>>> BETA 1.5
>>>>
>>>> NBROYDEN 8
>>>>
>>>> &END MIXING
>>>>
>>>> &END SCF
>>>>
>>>> &MGRID
>>>>
>>>> CUTOFF 940
>>>>
>>>> REL_CUTOFF 80
>>>>
>>>> NGRIDS 5
>>>>
>>>> PROGRESSION_FACTOR 3
>>>>
>>>> &END MGRID
>>>>
>>>> &XC
>>>>
>>>> &XC_FUNCTIONAL PBE
>>>>
>>>> &PBE
>>>>
>>>> SCALE_X 0.630
>>>>
>>>> SCALE_C 1.0
>>>>
>>>> &END
>>>>
>>>> &END XC_FUNCTIONAL
>>>>
>>>> &VDW_POTENTIAL
>>>>
>>>> POTENTIAL_TYPE pair_potential
>>>>
>>>> &PAIR_POTENTIAL
>>>>
>>>> TYPE DFTD3(BJ)
>>>>
>>>> PARAMETER_FILE_NAME dftd3.dat
>>>>
>>>> REFERENCE_FUNCTIONAL PBE
>>>>
>>>> &END PAIR_POTENTIAL
>>>>
>>>> &END VDW_POTENTIAL
>>>>
>>>> &HF
>>>>
>>>> FRACTION 0.370
>>>>
>>>> &SCREENING
>>>>
>>>> EPS_SCHWARZ 1.0E-7
>>>>
>>>> !SCREEN_ON_INITIAL_P TRUE
>>>>
>>>> &END
>>>>
>>>> &MEMORY
>>>>
>>>> MAX_MEMORY 7500
>>>>
>>>> &END
>>>>
>>>> &INTERACTION_POTENTIAL
>>>>
>>>> POTENTIAL_TYPE TRUNCATED
>>>>
>>>> CUTOFF_RADIUS 3
>>>>
>>>> T_C_G_DATA ./t_c_g.dat
>>>>
>>>> &END
>>>>
>>>> &END
>>>>
>>>> &END XC
>>>>
>>>> &AUXILIARY_DENSITY_MATRIX_METHOD
>>>>
>>>> ADMM_PURIFICATION_METHOD NONE
>>>>
>>>> METHOD BASIS_PROJECTION
>>>>
>>>> &END AUXILIARY_DENSITY_MATRIX_METHOD
>>>>
>>>> &END DFT
>>>>
>>>>
>>>> &SUBSYS
>>>>
>>>> &CELL
>>>>
>>>>
>>>>
>>>> A 8.043 0.000 0.000
>>>>
>>>> B 0.014 7.998 0.000
>>>>
>>>> C 0.014 0.019 7.998
>>>>
>>>>
>>>>
>>>> ALPHA_BETA_GAMMA 89.867 89.900 89.900
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> MULTIPLE_UNIT_CELL 1 1 1
>>>>
>>>> &END CELL
>>>>
>>>>
>>>>
>>>> &TOPOLOGY
>>>>
>>>> COORD_FILE_FORMAT XYZ
>>>>
>>>> COORD_FILE_NAME BaTi03-cellopt-pos-1-converged.xyz
>>>>
>>>> MULTIPLE_UNIT_CELL 1 1 1
>>>>
>>>>
>>>>
>>>> &END TOPOLOGY
>>>>
>>>> &KIND Ba
>>>>
>>>> ELEMENT Ba
>>>>
>>>> BASIS_SET DZVP-MOLOPT-SR-GTH-q10
>>>>
>>>> POTENTIAL GTH-PBE-q10
>>>>
>>>> BASIS_SET AUX_FIT cFIT10
>>>>
>>>> &END KIND
>>>>
>>>> &KIND Ti
>>>>
>>>> ELEMENT Ti
>>>>
>>>> BASIS_SET DZVP-MOLOPT-SR-GTH-q12
>>>>
>>>> POTENTIAL GTH-PBE-q12
>>>>
>>>> BASIS_SET AUX_FIT cFIT10
>>>>
>>>> &END KIND
>>>>
>>>> &KIND O
>>>>
>>>> ELEMENT O
>>>>
>>>> BASIS_SET DZVP-MOLOPT-SR-GTH-q6
>>>>
>>>> POTENTIAL GTH-PBE-q6
>>>>
>>>> BASIS_SET AUX_FIT FIT9
>>>>
>>>> &END KIND
>>>>
>>>> &KIND Sr
>>>>
>>>> ELEMENT Sr
>>>>
>>>> BASIS_SET DZVP-MOLOPT-SR-GTH-q10
>>>>
>>>> POTENTIAL GTH-PBE-q10
>>>>
>>>> BASIS_SET AUX_FIT cFIT10
>>>>
>>>> &END KIND
>>>>
>>>> &KIND Pb
>>>>
>>>> BASIS_SET DZVP-MOLOPT-SR-GTH-q4
>>>>
>>>> POTENTIAL GTH-PBE-q4
>>>>
>>>> BASIS_SET AUX_FIT cFIT9
>>>>
>>>> &END KIND
>>>>
>>>> &KIND Ca
>>>>
>>>> BASIS_SET DZVP-MOLOPT-SR-GTH-q10
>>>>
>>>> POTENTIAL GTH-PBE-q10
>>>>
>>>> BASIS_SET AUX_FIT cFIT10
>>>>
>>>> &END KIND
>>>>
>>>> &END SUBSYS
>>>>
>>>>
>>>> &END FORCE_EVAL
>>>>
>>>>
>>>>
>>>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20201027/b05a4b15/attachment.htm>
More information about the CP2K-user
mailing list