Page 1 of 1

Getting Bandgap from DOS calculated by GW method

Posted: Mon May 16, 2011 9:16 am
by felixvasp
Hi everyone,
I am newbie to VASP,especially DOS&BAND structure calculations. I ran the Si diamond bulk DOS calculation using GW method by two steps:

1. ran DFT calculation on Si bulk to get WAVEDER.
INCAR:
System = diamond Si
NBANDS=160
ENCUT=250
ISMEAR = 0; SIGMA = 0.05;
LOPTICS=.TRUE.

2. make another directory, copy POTCAR, KPOINT, POSCAR to the diretory, and ran GW calculations using WAVEDER from step 1 and using INCAR as follows:
INCAR
system = diamond Si
NBANDS=160
ENCUT=250
ISMEAR = 0; SIGMA = 0.05;
LOPTICS=.TRUE.
ALGO=GW
LWAVE=.TRUE.

3. plot DOS from DOSCAR.
But the calculated DOS graph doesn't give the BANDGAP of ~1.2 eV reported by literature but gives a BANDGAP of ~0.5 eV. Is there anything I didn't wrong? but someone reproduced the bandgap of GW by bandstructure. Can DOS produced by GW calculation gives right bandgap or if I did something wrong during producing DOS for GW, how to get the right DOS from GW calculation. Thanks a lot for help.






<span class='smallblacktext'>[ Edited ]</span>

Getting Bandgap from DOS calculated by GW method

Posted: Tue May 17, 2011 9:59 pm
by felixvasp
Anyone has the answer or I need to provide more specific inf.?