How to output optic matrix element ?

[jzzhao]

Hi, all

I use vasp5.2 to calculate bcc Fe, and I want to obtain optic information with spin-orbit couping. How I can output optic matrix element, i.e. < phi_i | nabla | phi_j >?

I have tried to add some code in "optics.F" aroud the lines 156 ~ 191,

Code: Select all

!-----------------------------------------------------------------------
! calculate < phi_i | nabla | phi_j >
!-----------------------------------------------------------------------
!ad---------------by fwx
      io_begin
      open(91,file="nabij.out",status="unknown")
        open(92,file="Px.dat",status="unknown")
        open(93,file="Py.dat",status="unknown")
        open(94,file="Pz.dat",status="unknown")
      write(92,*) KPOINTS%NKPTS, WDES%NB_TOT
      write(93,*) KPOINTS%NKPTS, WDES%NB_TOT
      write(94,*) KPOINTS%NKPTS, WDES%NB_TOT
      io_end
!ad------------------ by fwx

! loop over all (current) k-points, spin component and cartesian directions
      DO NK=1,KPOINTS%NKPTS
      DO ISP=1,INFO%ISPIN
      !  Do isp = 1,2
        DO IDIR=1,3
! get the matrix element of the nabla-operator in eV/Angstroem units
         CALL NABIJ_SOFT(NABIJ,IDIR,NK,ISP,W,WDES,GRID_SOFT,LATT_CUR)
         CALL NABIJ_AUG_ADD(NABIJ,IDIR,NK,ISP,W,WDES,P,T_INFO)

         NB_TOT=WDES%NB_TOT

         CALLMPI( M_sum_g(WDES%COMM,NABIJ(1,1),NB_TOT*NB_TOT))

!ad-----by fwx-------------
        io_begin
        if(idir==1)then
           write(92,"(I5,4E15.7)")Nk,(WDES%VKPT(I,NK),I=1,3),WDES%WTKPT(NK)
           do i=1,NB_TOT
           do j=1,NB_TOT
             write(92,"(2I5,2E15.7)") i,j,nabij(i,j)
           enddo
           enddo
           write(92,*)
        else if(idir==2)then
           write(93,"(I5,4E15.7)")Nk,(WDES%VKPT(I,NK),I=1,3),WDES%WTKPT(NK)
           do i=1,NB_TOT
           do j=1,NB_TOT
             write(93,"(2I5,2E15.7)") i,j,nabij(i,j)
           enddo
           enddo
           write(93,*)
         else
           write(94,"(I5,4E15.7)")Nk,(WDES%VKPT(I,NK),I=1,3),WDES%WTKPT(NK)
           do i=1,NB_TOT
           do j=1,NB_TOT
             write(94,"(2I5,2E15.7)") i,j,nabij(i,j)
           enddo
           enddo
           write(94,*)
        endif
        io_end
!ad-----by fwx-------------


         SCALE_NABIJ_TO_VIJ=AUTOA    ! hbar/m_e * 1/length in whatever unit
         NABIJ=NABIJ*SCALE_NABIJ_TO_VIJ   ! rescaling to bohr radii/Hartree
! write OPTIC
         io_begin
         IF (IU>=0) THEN
           CALL OUTOPT(NABIJ,IDIR,NK,ISP,W%CELTOT,W%FERTOT,NB_TOT, &
                       INFO%ISPIN,KPOINTS%NKPTS,WDES%VKPT, &
                       WDES%WTKPT,IRECLO,NBVAL,NBCON,NKOPT,NKOFF,IU)
         ENDIF
         io_end

        ENDDO    ! IDIR
       ENDDO     ! ISP
      ENDDO      ! NK

      io_begin
      IF (IU>=0) CLOSE(IU)
!ad-----by fwx-------------

      CLOSE(91)
      CLOSE(92)
      CLOSE(93)
      CLOSE(94)
!d----------by fwx
      io_end

But in my output file "Px.dat, Py.dat, Pz.dat", the matrix elements < phi_i | nabla | phi_j > (i.e. the array nabij(:,:) in above code) has not the same valus with two equivalent k point. My KPOINTS is:

Code: Select all

fixed line  mesh
6
Reciprocal lattice
   0.00000000  0.00000000  0.00000000       0.037
   0.00000000  0.00000000  0.00000000       0.037
   0.00000000  0.00000000  0.00000000       0.037
   0.50000000  0.50000000 -0.50000000       0.037
   0.50000000  0.50000000 -0.50000000       0.037
   0.50000000  0.50000000 -0.50000000       0.037

for example, the first and second k point are equivalent, the gamma point, but their "nabij(:,:)" has not smae values.

Another problem, the "nabij(:,:)" has different values in two same calcuation with the same file INCAR. My INCAR is:

Code: Select all

ENCUT=350
#IBRION =  -1
NSW=0
#ISPIN=2
#MAGMON=0 16*5 6*0
LSORBIT=T
SAXIS= 0 0 1
ALGO=Fast
ICHARG=11
NSW=0
LWAVE=F
LMAXMIX=4
VOSKOWN  =  1
NBANDS=  30
NELMDL=6
#ISMEAR=0
AMIX = 0.2
BMIX = 0.0001
AMIX_MAG = 0.8
BMIX_MAG = 0.0001
#LNABLA = .TRUE.
LOPTICS = .TRUE.
NPAR = 1
#MAXMIX=50

What cause this problem? My output code? my wrong calulation about vasp5.2? or the vasp own problem?

Any suggestions will be gratefull!

jz zhao


<span class='smallblacktext'>[ Edited ]</span>

[support_vasp]

Hi,

We're sorry that we didn’t answer your question. This does not live up to the quality of support that we aim to provide. The team has since expanded. If we can still help with your problem, please ask again in a new post, linking to this one, and we will answer as quickly as possible.

Best wishes,

VASP