work function and fermi energy

Message

simly · #1 Post by **simly** » Wed Dec 20, 2006 3:28 am

Hi,
I would like to find out calculate work function of any system (for instance are there any settings we should include in the INCAR). Which run would give a more accurate fermi energy - static or relaxation?

E-fermi : 6.2344 XC(G=0): -9.6242 alpha+bet :-12.4761
Also, is it accurate to get the fermi energy based on the first term?

Thanks

#2 Post by **admin** » Wed Dec 20, 2006 2:06 pm

1) the fermi level is only defined to within a constant energy offset in all calculations based on periodic boundary conditions (please lookup in any standard solid state theory textbook)
2) for slab calculations (sufficient vacuum size) the energy in the center of the vacuum can be regarded as energy zero. Set LVTOT=True in INCAR and use the votoav.f tool to calculate the planar-averaged potentials of the LOCPOT-file, and hence the vacuum energy level. All energies have to be given with respect to that reference. (choosing alpha+bet is a very poor approximation to that, except for free atoms).
3) static/relaxation is no criterium for the accuracy of the Fermi level. It changes with ionic relaxation, and the accuracy depends on the quality of the BZ integration (integration method, k-mesh, sometimes even NEDOS)

simly · #3 Post by **simly** » Thu Dec 21, 2006 9:35 am

That helps a lot. Thanks
May I find out where can we obtain vtotav.f tool?

#4 Post by **admin** » Wed Dec 27, 2006 1:59 pm

please contact vaspm3@univie.ac.at