My Community

Sometimes I found that the band occupation on the k point is not 1 or 0 for non-spin polarized calculations, nor 2 or 0 for spin-polarized calculations. Instead it is a fraction of 1 or 2.e.g.

band No. band energies occupation
1273 5.8884 1.14018
1274 5.9042 1.05197
1275 5.9103 1.01722
1276 5.9177 0.97548
1277 5.9215 0.95430
1278 5.9409 0.84589

could anyone suggest what happens here? and where is the true VBM?Thanks!

You are applying smearing, "SIGMA" is the repsonsible INCAR tag. Put it close to 0, like 0.001 and you'll get 1 and 0 occupation only. The question then would be, if this is your real physical picture. And you might run into convergence troubles.

Hth

alex

Thanks alex! I will try to tune the INCAR value and see if it is the case.