Zero Band Gap for semiconductor

[anveeksh]

Hello all,

I got a 0eV band gap for Co3O4 which is a semiconductor. The literature has 1.96 eV using DFT. Whatever underestimation DFT does, it should show some kind of band gap for semiconducting oxides. Is my path for calculation of band structure wrong or is there any wrong with my input? Please help. Thanks

I have minimized the energy and used optimized structure for my calculation.

#Band Structure Calculation
# Co3O4

ndtset 2

#Dataset 1: Self-Consistent Density Calculation
kptrlatt1 5 -5 5 -5 5 5 -5 -5 5
kpt1 -1.00000000E-01 -2.00000000E-01 0.00000000E+00
-1.00000000E-01 -4.00000000E-01 0.00000000E+00
-2.00000000E-01 -3.00000000E-01 0.00000000E+00
-1.00000000E-01 -3.00000000E-01 1.00000000E-01
-1.00000000E-01 4.00000000E-01 0.00000000E+00
-2.00000000E-01 5.00000000E-01 0.00000000E+00
-1.00000000E-01 5.00000000E-01 1.00000000E-01
-3.00000000E-01 -4.00000000E-01 0.00000000E+00
-2.00000000E-01 -4.00000000E-01 1.00000000E-01
-1.00000000E-01 -4.00000000E-01 2.00000000E-01
-1.00000000E-01 2.00000000E-01 0.00000000E+00
-2.00000000E-01 3.00000000E-01 0.00000000E+00
-3.00000000E-01 4.00000000E-01 0.00000000E+00
-4.00000000E-01 5.00000000E-01 0.00000000E+00
-3.00000000E-01 5.00000000E-01 1.00000000E-01
-2.00000000E-01 5.00000000E-01 2.00000000E-01
-1.00000000E-01 0.00000000E+00 0.00000000E+00
-3.00000000E-01 0.00000000E+00 0.00000000E+00
5.00000000E-01 0.00000000E+00 0.00000000E+00

kptrlen1 7.63647721E+01
ngfft1 96 96 96
nkpt1 19
optforces1 1
shiftk1 5.00000000E-01 5.00000000E-01 5.00000000E-01
tolmxf1 5.0d-9
toldff1 5.0d-10

#Dataset2: Band Structure Calculation
iscf2 -2 #non-self consistent calculation, read densigh from previous calculation
getden2 -1
kptopt2 -5 # Enter the k points manually
nband2 100
ndivsm2 10
kptbounds2
0.0 0.0 0.0 # Gamma point
0.375 0.375 0.75 # K point
0.5 0.5 0.5 # L point
0.625 0.25 0.625 # U point
0.5 0.25 0.75 # U point
0.5 0.25 0.5 # U point
tolwfr2 1.0d-12
enunit2 1
ngfft2 96 96 96

ixc 11
nstep 200
nband 100

wtk 0.04800 0.04800 0.04800 0.09600 0.04800 0.04800
0.04800 0.04800 0.09600 0.09600 0.04800 0.04800
0.04800 0.04800 0.09600 0.04800 0.01600 0.01600
0.00800
ecut 80
pawecutdg 150
diemac 3
ecutsm 0.5

#Spin
nsppol 2
nspden 2
nspinor 1
spinat 0.0 0.0 0.0
0.0 0.0 4.0
0.0 0.0 0.0
0.0 0.0 0.0
0.0 0.0 0.0
0.0 0.0 0.0
0.0 0.0 0.0
0.0 0.0 4.0
0.0 0.0 0.0
0.0 0.0 0.0
0.0 0.0 0.0
0.0 0.0 0.0
0.0 0.0 0.0
0.0 0.0 0.0

#Definition of the unit cell
acell 1.0953162321E+01 1.0953162321E+01 1.0953162321E+01 Bohr

rprim -4.0463729910E-18 7.0710678119E-01 7.0710678119E-01
7.0710678119E-01 -2.4039657124E-18 7.0710678119E-01
7.0710678119E-01 7.0710678119E-01 -7.2715423427E-18
#Definition of the atom types
ntypat 2 # There is only one type of atom
znucl 27 8 # The keyword "znucl" refers to the atomic number of the

#Definition of the atoms
natom 14
typat 1 1 2 1 2 1 2 1 2 2 1 2 2 2
xred 6.2500000000E-01 6.2500000000E-01 6.2500000000E-01
7.6363558336E-17 7.6363767562E-17 1.5963052158E-16
3.8647520629E-01 3.8647520629E-01 3.8647520629E-01
6.2500000000E-01 6.2500000000E-01 1.2500000000E-01
3.8647520629E-01 3.8647520629E-01 8.4057438113E-01
6.2500000000E-01 1.2500000000E-01 6.2500000000E-01
3.8647520629E-01 8.4057438113E-01 3.8647520629E-01
2.5000000000E-01 2.5000000000E-01 2.5000000000E-01
8.6352479371E-01 8.6352479371E-01 4.0942561887E-01
8.6352479371E-01 8.6352479371E-01 8.6352479371E-01
1.2500000000E-01 6.2500000000E-01 6.2500000000E-01
8.4057438113E-01 3.8647520629E-01 3.8647520629E-01
4.0942561887E-01 8.6352479371E-01 8.6352479371E-01
8.6352479371E-01 4.0942561887E-01 8.6352479371E-01

[ilukacevic]

Hi!

1] Is your ground-state scf wave-function converged? Is your structure relaxed?

2] Did you try other types of pseudopotentials?

Igor

[anveeksh]

Hi ilukacevic,

Yes, my structure is relaxed. I performed an optcell 2 calculation to get a relaxed structure. I then took those parameters and gave them as input in my first dataset (SCF). And second dataset is Non-SCF to calculate the band structure.

I tried PAW (generated by abinit) and Norm-conserving Troullier Martins (from online source) psp's till now. Do you think I need to try different psp's? Also, the literature did a spin polarized calculation, also which I tried but in vain.

Please help

Thanks
Anveeksh

[ilukacevic]

Yes, try other pseudos. Maybe you need to include more electrons into your calculations, like semicore electrons of Co.

What do you mean "in vain"? You got the same erroneous results with spin included or your spin calculation failed?

Igor

[anveeksh]

In vain in the sense, I could not see any band gap when I plotted using AbinitBandStructureMaker.py.
Ok. I will try with different psuedo potential files.

I just noticed that when I am trying to plot the BS of the dataset that calculated density of states (iscf = -3) using the AbinitBandStructureMaker.py script there is a band gap of 0.2eV. But when I did the same for (iscf = -2) there is no band gap.

1) How to plot Density of states? Is there any tool (python scripting sort of stuff) in abinit?
2) Is the gap I observed in the case of Density of states the correct band gap?
3) So, the input file I am using for band structure calculation correct?

Thanks
Anveeksh

[ilukacevic]

Hi!

1] Personally, I use xmgrace to plot almost any data not just DOS. It is most customizable. I don't know of any tool to plot DOS. I think that's because it's just too simple.

2] Yes, as a theoretical result. You could compare it with an experiment, although one is never advised to do that. You should compare your result with other theoretical results.

3] It seems correct to me.

Igor