symmetrically equivalent structures have different energies

Message

#1 Post by **admin** » Fri Jan 12, 2007 2:46 pm

this should not be the case. please check if VASP found the same symmetry for both systems.

alex · #2 Post by **alex** » Sat Jan 13, 2007 3:45 pm

Hi,

maybe it's triplett and singulett. No, you haven't put ISPIN=2.
Check for occupation on the O atom. Maybe one is 2/3 per p-orbital and the other has 2 fully occupied p-orbitals ...
Happens for O atom in cubic cells ...

alex

scerwin · #3 Post by **scerwin** » Fri Jan 19, 2007 7:57 pm

I think the problem is that your 2 k-points are not the irreducible subset of a complete symmetrized k-point "star".

The proof of this is that by using your INCAR and POSCAR files, but instead asking vasp to automatically generate a 2x2x2 Monkhorst-Pack set, one gets 3 k-points, not 2. When using this properly formed set, changing the coordinates of the O atom as you did leads to only trivial differences in the total energy (10^-6 eV).

Hope this helps.

SE