The ABINIT Users Mailing List ( CLOSED )

[Bernard Amadon <bernard.amadon@cea.fr>]

Dear Peng,

Note that, in our paper PRB_77_155104(2008), we have mostly done LDA+U 
calculations (and not GGA+U).
More importantly, the use of DFT+U requires in this case (and most of 
the time)
to do a spin polarized calculation (nsppol=2, nspden=2).
If you do not spin-polarize, you are not able to localize the f-electron 
of cerium !

Best regards
Bernard Amadon

shipeng a écrit :
> Hello everyone,
>
> I'm learning  to use atompaw & atompaw2abinit , and  wanted to generate a PAW
> PP  rare earth element cerium. The following is my input file with some
> comments :
> Ce 58                                     !definition of atom cerium
> GGA-PBE scalarrelativistic loggrid 2001   !All-electrons calc.:GGA-PBE
> -log.grid with 2001 grids
> 6 5 5 4 0                                 !max. n per angular momentum: !6s2
> 5p6 5d1 4f1
> 5 1 6                                     ! semi core                   5p:
> occ=6
> 5 2 0                                     !                             5d:
> occ=0
> 4 3 0                                     !                             4f:
> occ=0
> 6 0 0                                     !                             6s:
> occ=0
> 0 0 0                                     !End of occupation section
> c                                         !1s: core state
> c                                         !2s: core state
> c                                         !3s: core state
> c                                         !4s: core state
> v                                         !5s valence
> v                                         !6s valence
> c                                         !2p: core state
> c                                         !3p: core state
> c                                         !4p: core state
> v                                         !5p valence
> c                                         !3d: core state
> c                                         !4d: core state
> v                                         !5d valence
> v                                         !4f valence
> 3                                         !max. l for partial wave basis
> 2.45 2.45 2.45 2.45                       !r_paw, r_shape,r_vloc,r_core(a.u.)
> default is rpaw
> n
> y                                         !add an additional p partial wave ?
> yes
> -2.0
> n
> y                                         !add an additional d partial wave ?
> yes
> 2.0
> n
> y                                         !add an additional f partial wave ?
> yes
> 2.3
> n
> custom polynom vanderbiltortho besselshape!PS scheme, RRKJ PW+sinc shape func.
> 4 0 ultrasoft                             !simple bessel vloc
> 2.45                                      ! r_c matching radius for first  s
> partial wave
> 2.45                                      ! r_c matching radius for second s
> partial wave
> 2.45                                      ! r_c matching radius for first  p
> partial wave
> 2.45                                      ! r_c matching radius for second p
> partial wave
> 2.45                                      ! r_c matching radius for first  d
> partial wave
> 2.45                                      ! r_c matching radius for second d
> partial wave
> 1.55                                      ! r_c matching radius for first  f
> partial wave
> 2.45                                      ! r_c matching radius for second f
> partial wave
> 0
>
> although " evale" and all the curves plotted by logderiv* and wfn* files seem
> to be  good. Using the PP file generated by this file, I did some tests such 
> as
> equilibrium lattice constants and bulk modulous of both gamma cerium and ceria
> , the  results are not physical both in GGA and GGA+U scheme. And I could not 
> reproduce  the results in paper[PRB_77_155104(2008)]  ,here is my test abinit
> script:
> # geometry optimization of gamma cerium
> # option for output
> enunit     2
> prtden     1
>
> #spin
> nspinor    1
> nspden     1
> nsppol     1
>
> # PAW-related options
> pawecutdg  40.0
> #usepawu2    1
> #lpawu2      3
> #upawu2     6.1 eV   # U
> #jpawu2      0.7 eV   # J
>
> # parameters controning the convergence
> ecut       16
> ecutsm     0.5
> #nband      20
> occopt     7        # Gaussian smearing
> tsmear     0.005    # for f orbital electron
>
> # option for the choice of calculations
> ixc        11
> nstep      200
> iscf       17
> toldfe     1.0d-10
>
> # cell optimization
> optcell    1
> ionmov     3
> ntime      15
> dilatmx    1.05
>
> # definition of special k-points
> kptopt     1
> ngkpt      6 6 6
> nshiftk    4
> shiftk     0.5 0.5 0.5
>            0.5 0.0 0.0
>            0.0 0.5 0.0
>            0.0 0.0 0.5
>
> #Definition of the unit cell
> acell      3*9.8
> rprim      0.0  0.5  0.5
>            0.5  0.0  0.5
>            0.5  0.5  0.0
> xred       0.0  0.0  0.0
> natom      1
> ntypat     1
> typat      1
> znucl      58
>
> AND I'm deeply frustrated by this, is there anything wrong with my PAW
> generation file or the test file?
> Any comments and advice are welcomed,
>
> thanks,
>
> Peng Shi
>
>