Page 1 of 1

ISMEAR for semiconductors

Posted: Wed Apr 26, 2006 3:09 pm
by hu007
Hi everyone,

I want to run optimization and frequency calculations for semiconductors. I need your suggestions for setting the ISMEAR tag. Now, I have two choices. (1) using ISMEAR=-5 for every calculations. (1) using ISMEAR=0 for optimization and frequency calculations and using ISMEAR=-5 for energy calculation. Can anyone compare the computational accuracy and efficiency between these two methods for me? Thanks!

ISMEAR for semiconductors

Posted: Wed May 03, 2006 12:28 pm
by Graciani
Hi,
If your system is a semiconductor you should use ISMEAR=-5 because the forces are always correct, since partial occupancies do not vary. In that way, the total energy and DOS calculations are very accurate.
However, if you have a very large system, you can't use 4 k-points (needed for tetrahedron method, ISMEAR=-5) and you should use ISMEAR=0.