Choice of exchange correlation

[jerry]

Daer,

When i tried to find out the ground state of a plannar boron nitride crystal (ribbon) with defects, i found that there were several states which are very close to each other in energy (the difference is about 0.00003 eV) but different in the spin polarization. The parameters i chose are as follows:
1 exchange correlation: SGGA(PBE)
2 LCAO basis set: double zeta polarized
3 K point sampling: 1*1*71
Others are remained as the default values. The atk version i use is 12.2

Can someone help me with this?

Thanks!

[Anders Blom]

Help with what?

[jerry]

Do the obtained results  differentiate or idntify the ground state? Given the fact that these states are so close to each other in energy

[Anders Blom]

The ground state is a property of the entire system, in fact the only thing DFT can (formally) calculate is the ground state. The energy levels (the eigenvalues of the one-electron Kohn-Sham orbitals) are something very different than the ground state. What you may be interested in is looking for the HOMO or LUMO, but if it's an extended system you are rather looking for the band gap. For further guidance you will have to post the geometry and a picture of the results, because I don't understand what you are really asking about.

[jerry]

I have sent you a email with my script. Thank you very much for your valuable help!