Bandstructure calculation
Posted: Mon Dec 05, 2005 4:40 pm
Dear VASP masters and users,
I am doing a band structure calculation for bulk alpha-alumina in a rhombohedral cell. When I set up the lines of high symmetry in the BZ, for some k-points the eigenvalues get calculated twice (see the KPOINTS setup below) e.g. for (0 0 0) . Furthermore, the eigenvalues for (0.0 0.5 1.0) and (0.0 0.5 0.0) should be identical, as these two k-points differ by a reciprocal lattice vector.
The eigenvalues for these k-points are indeed identical, exept for the uppermost 2-3 bands. I performed calculations for different NBANDS and it is always the uppermost 2-3 bands where the eigenvalues differ ~0.1-0.5, for all bands lower than that they are identical.
I am curious whether this has to do with the way the eigenvalues are converged. Of course I performed the calculation with ICHARG=11 and I was careful to use the same grids in both the static calculation that generated CHGCAR and the band structure calculation.
Has anybody observed something similar and knows the reason for this?
Thank you very much. Best regards, Berit Hinnemann
-----------------------------------------------------------------
Mesh for rhombohedral Al2O3 band structure
10
Line-mode
reciprocal
0.0 0.0 0.0 1 ! Gamma
0.5 0.5 0.5 1 ! Z along line Lambda
0.5 0.5 0.5 1 ! Z
0.0 0.5 1.0 1 ! L along line B' or Y
0.0 0.5 0.0 1 ! L
0.0 0.0 0.0 1 ! Gamma
0.0 0.0 0.0 1 ! Gamma
0.0 0.5 -0.5 1 ! F' along line Sigma or Q'
I am doing a band structure calculation for bulk alpha-alumina in a rhombohedral cell. When I set up the lines of high symmetry in the BZ, for some k-points the eigenvalues get calculated twice (see the KPOINTS setup below) e.g. for (0 0 0) . Furthermore, the eigenvalues for (0.0 0.5 1.0) and (0.0 0.5 0.0) should be identical, as these two k-points differ by a reciprocal lattice vector.
The eigenvalues for these k-points are indeed identical, exept for the uppermost 2-3 bands. I performed calculations for different NBANDS and it is always the uppermost 2-3 bands where the eigenvalues differ ~0.1-0.5, for all bands lower than that they are identical.
I am curious whether this has to do with the way the eigenvalues are converged. Of course I performed the calculation with ICHARG=11 and I was careful to use the same grids in both the static calculation that generated CHGCAR and the band structure calculation.
Has anybody observed something similar and knows the reason for this?
Thank you very much. Best regards, Berit Hinnemann
-----------------------------------------------------------------
Mesh for rhombohedral Al2O3 band structure
10
Line-mode
reciprocal
0.0 0.0 0.0 1 ! Gamma
0.5 0.5 0.5 1 ! Z along line Lambda
0.5 0.5 0.5 1 ! Z
0.0 0.5 1.0 1 ! L along line B' or Y
0.0 0.5 0.0 1 ! L
0.0 0.0 0.0 1 ! Gamma
0.0 0.0 0.0 1 ! Gamma
0.0 0.5 -0.5 1 ! F' along line Sigma or Q'