My Community

Hi,

I have generated PDOS for a system in which there are clear states in the band gap visible in the OUTCAR, yet they do not show up in the PDOS. I'm not entirely sure why, as I have never experienced this problem before.

I do a self consistent run for a smaller KPOINT mesh, and then increase it and use the following INCAR for PDOS:

ISTART = 1;
ICHARG = 11;
INIWAV = 1;
NEDOS = 3000;
LORBIT = 0;
RWIGS= 2.790 1.400 1.630 1.480 .570;

PREC = Normal;
NELMIN = 5;
NELIM = 40;
EDIFF = 1E-07;
ENCUT = 450;
ISIF = 2; l
LWAVE= TRUE
LCHARG= TRUE
NPAR=1;
LREAL= Auto

NSW = 0;
IBRION = -1;
IWAVPR = 2;
ISYM=0;

LREAL = A
ISMEAR = -5; Tetrahedral nethod w/ Blochl corrections for smearing
SIGMA = 0.1; Broadening in eV -4-tet -1-fermi 0-gaus


Any suggestions as to why this might be happening are gratefully received....

Many thanks

the only reason I can guess is that even NEDOS=3000 is to coarse to capture the gap state peak:
please check from the KS-eigenvalue dispersion of the gap state (in OUTCAR) how broad this peak actually is, and choose NEDOS such that the energy grid on which the DOS is evaluated is more narrow than the peak width. Then please either increase NEDOS even further, or decrease the E-range on which the DOS is calculated (EMIN,EMAX)

Thank you for the reply.

I have performed the calculation restricting the energy range for which the DOS is calculated, and still there is no state appearing. In the OUTCAR the gap state appears at -2.2358 at each of the K-points.

The relevant section of the DOSCAR is included below:

-2.264 0.0000E+00 0.1310E+04
-2.259 0.0000E+00 0.1310E+04
-2.255 0.0000E+00 0.1310E+04
-2.250 0.0000E+00 0.1310E+04
-2.246 0.0000E+00 0.1310E+04
-2.241 0.0000E+00 0.1310E+04
-2.237 0.0000E+00 0.1310E+04
-2.232 0.0000E+00 0.1312E+04
-2.228 0.0000E+00 0.1312E+04
-2.223 0.0000E+00 0.1312E+04
-2.219 0.0000E+00 0.1312E+04
-2.214 0.0000E+00 0.1312E+04
-2.210 0.0000E+00 0.1312E+04

We can see around the energy at which the state should appear the total summed DOS increases, but the state isn't present. With the smearing set at 0.1 should this state not appear on such a fine grid?

Would you suggest I increase the K-points or try and go finer still.

I am still puzzled by this result.

Many thanks again.