The ABINIT Users Mailing List ( CLOSED )

[matthieu verstraete <matthieu.jean.verstraete@gmail.com>]

>> 1)      Despite high k-sampling (8,8,1), highly converged (toldfde=1e-9)
>> scf run, high energy cutoffs, etc. I can’t converge the iscf run to
>> tolwfr=1e-12.  Even for nstep=500 the calculation stops at tolwfr=6.8e-11.
>> Is this ok, or should I expect better convergence?#
are you using nbdbuf to neglect the last few bands in the convergence
criterion? Check the log file for the "res" fields: your bands may all
be converged very well (you can get down to 1.e-20 and more in most
cases, but vacuum is difficult) except the last few. Are you using
iscf 7 or 17?

btw toldfe 1.e-9 and 8x8 k-points is not _that_ tight convergence. It
depends what you look at, that's the real criterion for convergence.

>>
>> 2)      Even without any vacuum the simple tetragonal supercell bands
>> calculation doesn’t quite agree with the bulk (body centered tetragonal)
>> bands.  There is some difference especially around the gamma point and some
>> additional bands (blue in the plot) are found in the supercell case.
In the supercell case you have more atoms per unit cell, and have
folded back the bulk bands into the slab BZ. I trust you have made
sure the path for the band structures is the same in absolute
coordinates: for the usual reduced coordinates, the k-points you plot
would be very different in the 2 unit cells (bulk and slab).


>>
>> 3)      As I increase the layers of vacuum I get some occupied states
>> (green in plot) within the gap.  I don’t understand why these states 
>> appear,
>
> Surface states, probably. neat. I would also like to here if anyone has any
> comments about points 1) and 2). Cheers,
correct. You can plot the wf with cut3d to check. Often they are not
strictly localized on the surface, but also hybridize with bulk
states. This is all physical and gives neat experiments too.


Matthieu


>
> adam
>
>>
>> and for the work function calculation, should I ignore these for the HOMO
>> energy value.
>>
>>
>>
>> Thanks for any suggestions,
>>
>> Luke Thulin
>>
>> Nanoptek Corp
>>
>>
>>
>> I am using Abinit v5.8.4 with the following input files:
>>
>>
>>
>> # Crystalline anatase titania
>>
>> #Geometry optimization parameters, none for this case.
>> ionmov 0
>>
>> #Use to get a vacuum potential reference
>> prt1dm 1
>>
>> #Definition of the unit cell;
>> acell 3.802 3.802 9.68411/2 angstrom    # a,a,c/2
>>
>> #Orient the vacuum above the (001) plane.
>> rprim   1.0            0.0              0.0
>>         0.0            1.0              0.0
>>         0.0            0.0              6.5
>> chkprim 0   # This input variable allows to use non-primitive unit cells.
>>
>> #Definition of the atom types
>> ntypat 2          # There are two types of atoms: Ti and O
>> znucl 22 8
>>
>> #Definition of the atoms
>> natom 18           # There are 18 atoms in our supercell.
>> typat 1 1 2 2 2 2 1 1 2 2 2 2 1 1 2 2 2 2
>> xangst
>>    0.000  0.000  0.0000000000E+00   #layer 1
>>    0.000  1.901  2.4210275000E+00
>>    0.000  0.000 -2.0020593599E+00
>>    0.000  0.000  2.0020593599E+00
>>    0.000  1.901  4.1896814012E-01
>>    0.000  1.901  4.4230868599E+00
>>    1.901  1.901  4.8420550000E+00   #layer 2
>>    1.901  3.802  7.2630825000E+00
>>    1.901  1.901  2.8399956400E+00
>>    1.901  1.901  6.8441143600E+00
>>    1.901  3.802  5.2610231400E+00
>>    1.901  3.802  9.2651418600E+00
>>    0.000  0.000  9.6841100000E+00   #layer 3
>>    0.000  1.901  1.2105137500E+01
>>    0.000  0.000  7.6820506400E+00
>>    0.000  0.000  1.1686169360E+01
>>    0.000  1.901  1.0103078140E+01
>>    0.000  1.901  1.4107196860E+01
>>
>> #Definition of the planewave basis set
>> ecut  24.0         # Maximal kinetic energy cut-off, in Hartree
>> pawecutdg 60.0
>>
>> #Definition of the k-point grid;
>> kptopt 1
>> kptrlatt  8  0  0
>>           0  8  0
>>           0  0  1
>> shiftk 0.5 0.5 0.0
>>
>> #Definition of the SCF procedure
>> iscf 17
>> nstep 150         # Maximal number of SCF cycles
>> toldfe 1.0d-9     # Will stop when, twice in a row, the difference
>>                   # between two consecutive evaluations of energy
>>                   # differ by less than toldfe (in Hartree)
>> #SCF preconditioner
>> iprcel 45         # Inhomogeneous systems work better with RPA method.
>> diemix 0.7        # Default for PAW
>>
>>
>>
>>
>>
>> Band structure calculation input file:
>>
>> # Crystalline anatase titania
>> #
>> # Computation of the band structure.
>> # Abinit automatically looks for density from a previous run, file
>> #Slabi_DEN, proceeds with non-SCF band structure calculation.
>>
>> #The band structure
>> iscf    -2     #Perform a non-scf calculation
>> kptopt  -1     #Use one segment and rely on kptbounds and ndivk
>> nband   90     #The number of bands
>> ndivk   22     #The number of samples for each segment
>>
>> kptbounds    0.0  0.0  0.0  #Gamma
>>              1/2  1/2  0.0  #X
>>
>> tolwfr  1.0d-12       # Tolerance of the wavefunction squared residual
>> enunit  1             # Will output the eigenenergies in eV
>> prteig  1             # Will outupt the eigenenergies to a separate file
>>
>> #Definition of the unit cell;
>> acell 3.802 3.802 9.68411/2 angstrom    # a,a,c/2
>>
>> #Alter rprim, oriented with vacuum above the (001) plane.
>> rprim   1.0            0.0              0.0
>>         0.0            1.0              0.0
>>         0.0            0.0              6.5
>> chkprim 0   # This input variable allows to use non-primitive unit cells.
>>
>> #Definition of the atom types
>> ntypat 2          # There are two types of atoms
>> znucl 22 8
>>
>> #Definition of the atoms
>> natom 18           # There are 18 atoms in our supercell
>> typat 1 1 2 2 2 2 1 1 2 2 2 2 1 1 2 2 2 2
>> xangst
>>    0.000  0.000  0.0000000000E+00   #layer 1
>>    0.000  1.901  2.4210275000E+00
>>    0.000  0.000 -2.0020593599E+00
>>    0.000  0.000  2.0020593599E+00
>>    0.000  1.901  4.1896814012E-01
>>    0.000  1.901  4.4230868599E+00
>>    1.901  1.901  4.8420550000E+00   #layer 2
>>    1.901  3.802  7.2630825000E+00
>>    1.901  1.901  2.8399956400E+00
>>    1.901  1.901  6.8441143600E+00
>>    1.901  3.802  5.2610231400E+00
>>    1.901  3.802  9.2651418600E+00
>>    0.000  0.000  9.6841100000E+00   #layer 3
>>    0.000  1.901  1.2105137500E+01
>>    0.000  0.000  7.6820506400E+00
>>    0.000  0.000  1.1686169360E+01
>>    0.000  1.901  1.0103078140E+01
>>    0.000  1.901  1.4107196860E+01
>>
>> #Definition of the planewave basis set
>> ecut  24.0         # Maximal kinetic energy cut-off, in Hartree
>> pawecutdg 60.0
>>
>> nstep 250         # Maximal number of non-SCF cycles
>>
>>
>>
>>
>>
>>
>



-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dr. Matthieu Verstraete

Universite de Liège
Institut de Physique, Bat. B5
Allée du 6 aout, 17
B- 4000 Sart Tilman, Liège
Belgium

Phone : +32 4 366 37 50
Fax   : +32 4 366 36 29

Mail : matthieu.jean.verstraete@gmail.com