fermi energy level below highest-occupied band !!!

Message

vaspGG · #1 Post by **vaspGG** » Fri Mar 29, 2013 10:18 am

Hi all,

I was doing DFT+U calculation with the setting NUPDOWN=4 in INCAR (force the eletron to be isolated). However, I got a very strange results: fermi leverl is below the band that the 4 isolated eletrons stay. How can the eletron stay above the fermi level.

outputs of band occupy in OUTCAR:

spin 1:
881 2.7060 1.00000
882 2.8060 1.00000
883 2.9311 1.00000 (highest occupied energy is 2.9)
884 5.3137 0.00000
885 5.3894 0.00000
886 5.4030 0.00000

spin 2:
882 2.8024 1.00000
883 2.9270 1.00000
884 3.6412 1.00000
885 3.8348 1.00000
886 3.9242 1.00000
887 4.1433 1.00000(highest occupied energy is 4.1)
888 5.3802 0.00000
889 5.3881 0.00000
890 5.4003 0.00000
891 5.4062 0.00000
892 5.4135 0.00000

fermi energy:
E-fermi : 3.4193 XC(G=0): -11.1231 alpha+bet :-11.4232
add alpha+bet to get absolut eigen values

It seems that the 4 electrons with spin 2 was out of fermi surface!

Any reply is appreciated.

#2 Post by **admin** » Tue Apr 02, 2013 3:20 pm

Fermi energy is the highest occupied band. In your example EF is the band 887, spin 2.

vaspGG · #3 Post by **vaspGG** » Wed Apr 03, 2013 9:07 am

[quote="'smallblacktext'>[ Edited Wed Apr 03 2013, 09:08AM "]</span>

#4 Post by **admin** » Mon Apr 08, 2013 6:20 pm

if NUPDOWN is set explicitely, the Fermi level is NOT re-adjusted after the spin-down part is calculated, the Fermilevel written in OUTCAR is the Fermi-level of the spin-down channel only. --> you have to calculate the real Fermi-energy by analysing the complete data set (including the spin-up part as well).