Output-bandstructure

The output of potential, valence and conduction bands is controlled by this keyword.

!-------------------------------------------------------------------!
$output-bandstructure                                      optional !
 destination-directory                 character           required !
 conduction-band-numbers               integer_array       optional !
 valence-band-numbers                  integer_array       optional !
                                                                    !
 potential                             character           optional !
 built-in-potential                    character           optional !
 electric-field                        character           optional !
 band-gap                              character           optional !
                                                                    !
 output-grid-position                  double_array        optional !
 output-grid-position-octant           character           optional !
$end_output-bandstructure                                  optional !
!-------------------------------------------------------------------!

Syntax:

destination-directory = band_structure/
                      = /MOSFET/band_structure/

Name of directory to which the files should be written. Must exist and directory name has to include the slash (\ for DOS and / for UNIX). Must be the first specifier.

conduction-band-numbers = 1 2 3

Numbers of conduction band edges that are put out (1, ..., max_num_cbbands). The numbering corresponds to the original numbering in the database. If a band is split because of strain, there will be several columns in the output file for all subbands.
1 = Gamma band
2 = L band
3 = X band

If one does not want to print any conduction band, one can put this line into comments or delete it.

 

valence-band-numbers  = 1 2 3

Numbers of valence bands that are put out (1, ..., max_num_vbbands).
1 = heavy hole band
2 = light hole band
3 = split-off hole band

If one does not want to print any valence band, one can put this line into comments or delete it.

 

potential = yes
          = no

Flag whether to put out the electrostatic potential in units of [V].
The electrostatic potential is the solution of the Poisson equation.

 

built-in-potential = yes
                   = no

Flag whether to put out the electrostatic built-in potential in units of [V].
Two built-in potentials are written out:
The               electrostatic built-in potential is the solution of the Poisson equation in equilibrium.
The classical electrostatic built-in potential is the solution of the Poisson equation in equilibrium using only classical densities, i.e. ignoring any quantum mechanical densities. (This potential is used as a start value for the quantum mechanical calculations.)
==> potential_built_in_cl_1D.dat
==> potential_built_in_1D.dat

Additionally, the starting value (initial guess) of the electrostatic potential is written out.
Currently, this initial guess is determined depending on the local doping concentration.
==> potential_initial_guess_1D.dat

 

electric-field = yes
               = no

Flag whether to output the electric field in the file electric_field.fld.
Units: [kV/cm] ( = 1d-5 [V/m] )
The electric field is defined on the material grid points, i.e. on the points lying exactly in the middle of the "physical" grid points where the electrostatic potential is defined.
The electric field is calculated as the negative gradient of the electrostatic potential. (CHECK: Is this correct? Shall we output or call it the displacement D instead?)

 

band-gap = yes  ! (default for 1D simulations)
         = no   ! (default for 2D/3D simulations)

Flag whether to put out the band gap in the file BandGap1D.dat.
Units: eV
The minimum of all conduction band edges (E_c_min), the maximum value of all valence band edges (E_v_max), and the corresponding band gap is part of the output (E_gap_min).
Also, the band gaps for the Gamma (E_gap_Gamma), L (E_gap_L) and X (E_gap_X) band edges are printed out.
If bands are split due to strain, for each split band, the band gap is printed out. They are labeled with _a, _b, _c, _d.
For 1D simulations, all data is contained in this file: band_structure/BandGap1D.dat
For 2D and 3D simulations, for each a separate file is printed out. They are called: E_c_min*, E_v_max*, E_gap_min*, E_gapGamma*, E_gapL*, E_gapX*
For 1D simulations, all data is contained in this file: band_structure/BandGap1D.dat

 

output-grid-position        = 10d0              ! [nm]  1D:  x      =  10 nm
                            = 10d0 20d0         ! [nm]  2D: (x,y)   = (10 nm, 20 nm)
                            = 10d0 20d0 0d0     ! [nm]  3D: (x,y,z) = (10 nm, 20 nm, 0 nm)
Here, one can specify a grid point (i,j,k) where output like band edges, potential, densities, ... are written out.
This also works together with a sweep (e.g. voltage sweep, magnetic field sweep).

output-grid-position-octant = left              ! 1D: octant 1,  i-         (default for 1D)
                            = right             ! 1D: octant 2,  i+

                            = lowerleft         ! 2D: octant 1, (i-,j-)     (default for 2D)
                            = lowerright        ! 2D: octant 2, (i+,j-)
                            = upperleft         ! 2D: octant 3, (i-,j+)
                            = upperright        ! 2D: octant 4, (i+,j+)

                            = bottomlowerleft   ! 3D: octant 1, (i-,j-,k-)  (default for 3D)
                            = bottomlowerright  ! 3D: octant 2, (i+,j-,k-)
                            = bottomupperleft   ! 3D: octant 3, (i-,j+,k-)
                            = bottomupperright  ! 3D: octant 4, (i+,j+,k-)
                            = toplowerleft      ! 3D: octant 5, (i-,j-,k+)
                            = toplowerright     ! 3D: octant 6, (i+,j-,k+)
                            = topupperleft      ! 3D: octant 7, (i-,j+,k+)
                            = topupperright     ! 3D: octant 8, (i+,j+,k+)
This is necessary for properties that are discontinuous at material interfaces.
If there is an interface at the grid point output-grid-position, then one has to specify for which octant the output should refer to.
(In 3D there are 8 octants, in 2D there are four "octants", and in 1D there are two "octants".

Output

Band-edges:

Filename: 

Structure:                          

Remark:

Due to strain the bands with degenerate minima split into several subbands. These subbands are listed in different columns (e.g. in silicon for the X band (band no. 3) if strain is present, the band edges split.).

In 1D simulations, an additional files exists: BandEdges1D.dat
It contains all band edges and all Fermi levels that should be printed out in one output file.

Potential:

Filename: 

Structure:                          

Classical built-in potential for the device

First, the Poisson equation is solved in equilibrium, using on classical densities, i.e. without quantum mechanics.
The resulting electrostatic potential is called the built-in potential of the device for a classical density.

==> potential_built_in_cl_1D.dat

Built-in potential for the device

Then, the Poisson equation is solved again in equilibrium, using either classical or quantum mechanical densities, or a combination of both, depending on the input file).
The resulting electrostatic potential is called the built-in potential of the device.
In case, no quantum mechanical densities are involved, the built-in potential is identical to the classical built-in potential.

==> potential_built_in_1D.dat