how to analyse the PDOS in Bader volume
Posted: Thu Sep 01, 2011 4:25 pm
Dear all,
First, many thanks to the Vasp tools developers in UT, I am a big fan of them, especilly the NEB code. Recently, I am studying the charge transfer in a slab. For Mulliken charge result uing CASTEP, I am quite confused that the sum of the charge on each atom is not equal to the the number of the electrons in the supercell. It might be understandable as it is dificult to define the size of the ion, thus some charge might be missed and some might be counted two times. With VASP+Bader charge, it seems better, and the total Bader charge on the atoms is very close to the number of electrons in the supercell. That's a good start, I think. Then I tried to project the DOS onto the Bader volume. I don't know how the code make it. ( Calculate the radius of the atom from the Bader volume then do the projection? ).
Anyway, I got the PDOS finally. However I found a problem that, for specific energy level, the sum of the number of states of each atoms does not equal to the total DOS on that energy level! I think the total DOS can not be wrong for a converged electronic structure calculation. May be something is not proper during the projection. I want to know if the PDOS of one atom is intergrated, is that equal the Bader charge? Also I got the line 'Volume of typ 1 -13%' printed in the OUTCAR, which suggest a overlap during the projection.
Any suggestion will be appreciated.
Best Wishes
Ross
The INCAR is attached:
Startparameter for this run:
PREC = Accurate medium, high low
ISTART = 0 job : 0-new 1-cont 2-samecut
ICHARG = 2 charge: 1-file 2-atom 10-const
ISPIN = 1 spin polarized calculation?
LNONCOLLINEAR = F non collinear calculations
LSORBIT = F spin-orbit coupling
INIWAV = 1 electr: 0-lowe 1-rand 2-diag
Electronic Relaxation
ENCUT = 525.0 eV
GGA = 91
NELM = 200; NELMIN= 3; NELMDL= -15 # of ELM steps
LMAXMIX =4 to d electron, 6 for f electron, using with PAW
EDIFF = 0.1E-04 stopping-criterion for ELM
LREAL = T real-space projection
IALGO = 48 #all empty bands are treated accurately algorithm
NBANDS = 1000
Ionic relaxation
EDIFFG = -0.05 stopping-criterion for IOM
NSW = 0 number of steps for IOM
IBRION = 2 ionic relax: 0-MD 1-quasi-New 2-CG
ISIF = 2 stress and relaxation
ISYM = 2 0-nonsym 1-usesym 2-fastsym
POTIM = 0.40 time-step for ionic-motion
#PSTRESS= 0.0 pullay stress
DOS related values:
EMIN = -30.00; EMAX = 10.00 energy-range for DOS
ISMEAR = 1; SIGMA = 0.20 broadening in eV -4-tet -1-fermi 0-gaus
LORBIT = 11
NEDOS = 1000
ICHAIN = 5
NGX = 12
NGY =18
NGZ = 270
Write flags
LWAVE = F write WAVECAR
LCHARG = F write CHGCAR
First, many thanks to the Vasp tools developers in UT, I am a big fan of them, especilly the NEB code. Recently, I am studying the charge transfer in a slab. For Mulliken charge result uing CASTEP, I am quite confused that the sum of the charge on each atom is not equal to the the number of the electrons in the supercell. It might be understandable as it is dificult to define the size of the ion, thus some charge might be missed and some might be counted two times. With VASP+Bader charge, it seems better, and the total Bader charge on the atoms is very close to the number of electrons in the supercell. That's a good start, I think. Then I tried to project the DOS onto the Bader volume. I don't know how the code make it. ( Calculate the radius of the atom from the Bader volume then do the projection? ).
Anyway, I got the PDOS finally. However I found a problem that, for specific energy level, the sum of the number of states of each atoms does not equal to the total DOS on that energy level! I think the total DOS can not be wrong for a converged electronic structure calculation. May be something is not proper during the projection. I want to know if the PDOS of one atom is intergrated, is that equal the Bader charge? Also I got the line 'Volume of typ 1 -13%' printed in the OUTCAR, which suggest a overlap during the projection.
Any suggestion will be appreciated.
Best Wishes
Ross
The INCAR is attached:
Startparameter for this run:
PREC = Accurate medium, high low
ISTART = 0 job : 0-new 1-cont 2-samecut
ICHARG = 2 charge: 1-file 2-atom 10-const
ISPIN = 1 spin polarized calculation?
LNONCOLLINEAR = F non collinear calculations
LSORBIT = F spin-orbit coupling
INIWAV = 1 electr: 0-lowe 1-rand 2-diag
Electronic Relaxation
ENCUT = 525.0 eV
GGA = 91
NELM = 200; NELMIN= 3; NELMDL= -15 # of ELM steps
LMAXMIX =4 to d electron, 6 for f electron, using with PAW
EDIFF = 0.1E-04 stopping-criterion for ELM
LREAL = T real-space projection
IALGO = 48 #all empty bands are treated accurately algorithm
NBANDS = 1000
Ionic relaxation
EDIFFG = -0.05 stopping-criterion for IOM
NSW = 0 number of steps for IOM
IBRION = 2 ionic relax: 0-MD 1-quasi-New 2-CG
ISIF = 2 stress and relaxation
ISYM = 2 0-nonsym 1-usesym 2-fastsym
POTIM = 0.40 time-step for ionic-motion
#PSTRESS= 0.0 pullay stress
DOS related values:
EMIN = -30.00; EMAX = 10.00 energy-range for DOS
ISMEAR = 1; SIGMA = 0.20 broadening in eV -4-tet -1-fermi 0-gaus
LORBIT = 11
NEDOS = 1000
ICHAIN = 5
NGX = 12
NGY =18
NGZ = 270
Write flags
LWAVE = F write WAVECAR
LCHARG = F write CHGCAR