ABINIT Discussion Forums

i need help on necessary correction to this input file for calculation of band structure of MgB2
ndtset 3

#Dataset 1 : GS calculation
kptopt1 0 # Use symmetry and treat only inequivalent points
ngkpt1 5 5 10
nshiftk1 1
shiftk1 0.5 0.5 0.5 # This choice preserves the hexagonal symmetry of the grid,
occopt1 1
nband1 26
prtden1 1
tolvrs1 1.0d-18

#Dataset 2 : Band structure calculation
iscf2 -2
getden2 1
kptopt2 -6
tolwfr2 1.0d-12
nband2 26
ndivk2 15 7 12 15 15 12 # 15, 7, 12, 15, 15 and 12 divisions of 6 segments, delimited
# by 7 points
kptbounds2 0.5 0.5 0.5 #A
0.0 0.0 0.0 #Gamma
0.5 0.5 0.0 #M
0.5 0.0 0.0 #X
0.0 0.0 0.0 #Gamma
0.0 0.0 0.5 #Z
0.5 0.0 0.5 #R

#Dataset 3 : Output of the Fermi surface
prtfsurf3 1 # To output the FS
getwfk3 -1
iscf3 -3 # NSCF calculation, occupancies and Fermi level are calculated once convergence is achieved.
# It should give the same FS as a GS calculation provided that the two
# k-meshes are sufficiently dense.
getden3 1
occopt3 1
nband3 26
kptopt3 1
ngkpt3 8 8 8 # too coarse. 24 24 24 for a reasonable FS
shiftk3 0.0 0.0 0.0 # Shifted grids are not supported by Xcrysden.
tolwfr3 1.0d-12

#COMMON INPUT DATA###########################################################
#
ecut 50
ecutsm 0.5
nstep 100

# TiO2 lattice structure
ntypat 2
znucl 12 5
natom 3
typat 1 2 2

acell 2*1.6325116953E+00 1.8627037820E+00
angdeg 90 90 120
spgroup 191
brvltt -1

# Atomic positions
xred 0.00 0.00 0.00 #Mg1
1/3 2/3 1/2 #B2
2/3 1/3 1/2 #B2


ixc 3
timopt 2

Thanks alot
Bamgbose, M. K.

You don't need the spgroup input because Z = 1 in this structure and it's a P-type space group; so you also don't need the brvltt statement (the cell already is primitive). Your acell entries can't be right, note that the default length here is bohr, and the cif file I found gives the cell lengths as 3.08489, 3.08489, 3.52107 angstrom.

Prof. Josef W. Zwanziger thanks alot

Hi,
i dont know what is wrong in this file
#bandstructure of Aluminium Phosphide

ndtset 3

#dataset 1 ;GS calculation
kptopt1 1
ngkpt1 8 8 8
nshiftk1 4 # Use one copy of grid only (default)
shiftk1 0.0 0.0 0.5 # This gives the usual fcc Monkhorst-Pack grid
0.0 0.5 0.0
0.5 0.0 0.0
0.5 0.5 0.5
occopt1 1
nband1 26
prtden1 1
tolvrs1 1.0d-18

#dataset2: the band structure
iscf2 -2
getden2 -1
kptopt2 -5
tolwfr2 1.0d-12
nband2 26
ndivk2 40 40 40 40 40
kptbounds2
0.0 0.0 0.0 #G
0.5 0.5 0.5 #L
0.5 0.0 0.5 #X
0.0 0.0 0.0 #G
0.5 0.5 0.5 #L
0.5 0.0 0.5 #X
enunit2 1

#Dataset 3 : Output of the Fermi surface
prtfsurf3 1 # To output the FS
getwfk3 -1
iscf3 -3 # NSCF calculation, occupancies and Fermi level are calculated once
# convergence is achieved.
# It should give the same FS as a GS calculation provided that the two
# k-meshes are sufficiently dense.
getden3 1
occopt3 1
nband3 26
kptopt3 1
ngkpt3 8 8 8 # too coarse. 24 24 24 for a reasonable FS
shiftk3 0.0 0.0 0.0 # Shifted grids are not supported by Xcrysden.
tolwfr3 1.0d-12


#Definition of the unit cell
diemac 12
acell 3*10.40
#angdeg 60 60 60

rprim 0 0.5 0.5
0.5 0 0.5
0.5 0.5 0

ntypat 2
znucl 13 15
natom 2
typat 1 2

xred 0 0 0
0.25 0.25 0.25

ecut 50.0
ecutsm 0.5
nstep 50

AlP band gap suppose to be 2.5eV, but i got 1.6eV with this file.

Hi!

I think that your input file is ok. Lower value for band gap is most probably because you use DFT + LDA to calculate the electronic structure. If you ought to stick with it, you'll have to be satisfied with that value.
Otherwise, you'll have to use some other method to get more accurate value.

Igor

what other methods are u referring to.
Thanks

Normal DFT almost always underestimates band gaps, this is a very standard and well-known problem that is discussed at length in any text book on DFT in solids. The underestimation will probably be worse for LDA compared to GGA, but even with GGA it will be noticeable. You have to go beyond single-particle DFT to do better, typically with quasi-particle corrections as implemented by the GW method. This is available in abinit, see the tutorials.

Bamgbose,

What Psp files did you use? Are they Norm conserving Troullier Martin Pseudo potentials?

Thanks