The confusing of the results of DOS calculated by cp2k (density found at/around Fermi energy for insulator)
Marcella Iannuzzi
marci... at gmail.com
Sun Jun 10 07:57:35 UTC 2018
For insulators there is no definition of the Fermi energy.
Your system shows a clear gap, indeed it is an insulator.
In your plot, the zero is most probably set at the energy of the highest
occupied state.
But this choice is arbitrary.
Regards
Marcella
On Saturday, June 9, 2018 at 5:38:57 PM UTC+2, Jannen wrote:
>
> Dear CP2K Users,
>
> I am calculate the DOS for a crystal structure. It is a insulator. The
> density of state was found at Fermi energy level as shown in attached
> picture. I was really confused by the results. The DOS should be 0 at Fermi
> energy level. Even though the plots was broadening using deskinsgroup code,
> but that will just cause a slightly shifting above the Fermi energy level.
> Could anyone help me on the results? Any problem with the input?
>
> Please see the attached settings for DFT section.
>
> &DFT
> BASIS_SET_FILE_NAME BASIS_MOLOPT
> POTENTIAL_FILE_NAME GTH_POTENTIALS
> CHARGE 0
> MULTIPLICITY 1
> UKS TRUE
> &QS
> EPS_DEFAULT 1.0E-14
> &END QS
> &MGRID
> CUTOFF 400
> NGRIDS 4
> REL_CUTOFF 60
> &END MGRID
> &SCF
> SCF_GUESS RESTART
> EPS_SCF 1.0E-07
> MAX_SCF 1000
> &OT TRUE
> PRECONDITIONER FULL_ALL
> MINIMIZER CG
> ENERGY_GAP 0.001
> &END OT
> &PRINT
> &RESTART OFF
> &END RESTART
> &END PRINT
> &END SCF
> &XC
> &XC_FUNCTIONAL PBE
> &END XC_FUNCTIONAL
> &VDW_POTENTIAL
> DISPERSION_FUNCTIONAL PAIR_POTENTIAL
> &PAIR_POTENTIAL
> TYPE DFTD3
> REFERENCE_FUNCTIONAL PBE
> PARAMETER_FILE_NAME dftd3.dat
> R_CUTOFF 10.0
> &END PAIR_POTENTIAL
> &END VDW_POTENTIAL
> &END XC
> &PRINT
> &PDOS
> NLUMO -1
> COMPONENTS TRUE
> &LDOS
> COMPONENTS TRUE
> LIST 1..486
> &END LDOS
> &END
> &END PRINT
> &END DFT
>
> Best regards,
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20180610/da0e6b8b/attachment.htm>
More information about the CP2K-user
mailing list