Nanosurf Lab

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.
Hartree potential that may serve as an input for the PPM can be produced by various DFT codes. A slab calculations are necessary for a creation of the Hartree potential. The calculated system should be in the lower part of the unit cell, but z coordinates of the top-most layer has to be above zero! At least 10 Ǎ of vacuum above a slab geometry has to be used. 15-20 Ă of Vacuum are recommended. Here we show examples how to produce a hartree potential, for couple of DFT codes:

== Fireball ==

A __fireball.in__ for creation of an __fftpot.xsf__ file with Hartree potential with using a McWEDA functional should look like:

  &OPTION
  basisfile =     'answer.bas'
  lvsfile =       'input.lvs'
  kptpreference = 'input.kpts'
  nstepf     = 1
  icluster   = 0
  itdse      = 0
  iqout      = 1
  ifixcharge = 1     ! 0 if you don't have pre-calculated atomic charges in CHARGES
  iquench    = -1
  &END
  
  &OUTPUT
  iwrtxsf = 1   ! print the important file
  iwrtden = 1   ! print the important file
  &END
  
  &MESH
  ifixg0 = 1            ! 
  g0     = 0.0,0.0,0.0  ! do not shift the position of atoms in fftpot.xsf with respect to the answer.bas
  Ecut = 300.0d0        ! not really necessary, but gives grid sampling approximately 100 pm.
  &END

A __fireball.in__ file for calculations with XC on a grid computations:

  &OPTION
  basisfile     = 'answer.bas'
  lvsfile       = 'input.lvs'
  kptpreference = 'samplek.kpts'
  nstepf     = 1
  icluster   = 0
  itdse      = 0
  iqout      = 1
  ifixcharge = 0
  dt         = 0.5
  iquench    = -1
  iks        = 1
  imcweda    = 0
  idogs      = 0
  bmix       = 0.05
  &END
  
  &OUTPUT
  iwrtxsf = 1   ! print the important file
  iwrtden = 1   ! print the important file
  &END
  
  &MESH
  ifixg0 = 1            ! 
  g0     = 0.0,0.0,0.0  ! do not shift the position of atoms in fftpot.xsf with respect to the answer.bas
  Ecut = 300.0d0        ! not really necessary, but gives grid sampling approximately 100 pm.
  &END

== VASP  ==

[[http://www.vasp.at/]]

A __LOCPOT__ file containing the Hartree potential is produced during standart //VASP// run. One just have to care about a dipole correction (if necessary). The __LOCPOT__ file has to be convert into an *.xsf file, eg. via [[http://theory.chm.tu-dresden.de/~jk/software.html]]

== FHI-AIMS ==


[[http://aimsclub.fhi-berlin.mpg.de/]]

Add following sequention into your __control.in__ file:

  output cube hartree_potential
    cube origin lx ly lz

where lx, ly & lz are shifting the output cube file so the written grid would be at the beginning of the Cartesian system. PPM can now read geometry from a cube file created by //FHI-AIMS// and shift it according to the shift of the cube cell; however for good run of the PPM calculation at least lz has to be properly specified. l = length of lattice vector / 2 - 0.05; But proper shift is used only, when the lattice vector follows some Cartesian direction.
Nanosurf Lab

User Tools

Site Tools

Page Tools
