Bandgap from multiple k-points

[Dr Vitaly V. G. Chaban]

I need to derive a bandgap from multiple kpoint caluclation.

I have done the stages as described in http://cms.mpi.univie.ac.at/vasp-forum/ ... php?4.4709 and got the EIGENVAL file which contains something like that for all of my k-points.


160 160 1 1
0.2142764E+02 0.2003204E-08 0.2003204E-08 0.8543663E-09 0.1000000E-15
1.000000000000000E-004
CAR
unknown system
640 5 404

0.0000000E+00 0.0000000E+00 0.5000000E-01 0.2000000E+00
1 -20.1196
2 -20.0724
3 -20.0565
4 -19.9369
5 -19.9255
6 -19.7069
7 -19.6991
8 -19.3917
...................................
...................................


How can I get a bandgap (single number) from these data?

[tlchan]

There are 5 k-points in your calculations. You can check from OUTCAR the occupancy of all the eigenstates for each k-point. The highest occupied state is the valence band maximum (VBM), and the lowest unoccupied state is the conduction band minimum (CBM). The difference between CBM and VBM gives you the band gap. Since there are 640 electrons in your system, I believe the VBM should be at state 320, while the CBM is at state 321. You may want to check that the actual VBM and CBM are at the k-points sampled in your calculation.