Fermi energy and work function

Message

user · #1 Post by **user** » Sat Apr 14, 2007 4:52 pm

Dear All,

Work function is defined as: \Phi=E(vacuum)-E(Fermi); in VASP calculations, E(Fermi) is set at the top of the valence band, which can be found in OUTCAR.

For metals, it is straightforward to calculate \Phi using E(Fermi) from OUTCAR.

My question is: in semiconductors, how do we define E(Fermi) for the work function? should we instead take it at the middle of the band gap?

Thanks.

#2 Post by **admin** » Mon Apr 16, 2007 12:42 pm

experimentalists usually define it as the middle of the band gap. However, by standard DFT calculations the band gap usually is under-estimated, therefore E-fermi is usually calculated from the hightest occupied eigenvalue in DFT calculations.

lcyin · #3 Post by **lcyin** » Wed Apr 25, 2007 7:17 am

How can we obtain the E(vacuum), does the E(vacuum) equal to zero?

user · #4 Post by **user** » Wed Apr 25, 2007 4:59 pm

Nope. You have to get E(vacuum) from the averaged electrostatic potential, see Hand-ons III: http://cms.mpi.univie.ac.at/vasp-worksh ... tation.htm

lcyin · #5 Post by **lcyin** » Thu Apr 26, 2007 1:52 am

Thanks!

lcyin · #6 Post by **lcyin** » Thu Apr 26, 2007 7:11 am

But I have not the vtotav utility in hand
Can Admin. or other vasp user send it to me?
Thank you.