ABINIT Discussion Forums

Hi,
I wonder whether or not it is possible to do a berry phase calculation for a ferromagnetic system.
I am getting and error that says:
"the number of bands is not the same for every k-point".
How can I assure the number of bans is the same for each k-point?
Thanks!

I suspect it's not the type of spin-order that's a problem here, it's that your system may be metallic. The entire berryphase theory is predicated on the assumption that the system is an insulator, and that you are only treating the occupied bands.

Dear jzwanzig,
Thank you very much for your replay.
My system is not metallic, it is a ferromagnetic insulator.
I am working with nsppol=2, I have 317 valence electrons.
The calculation does not even start, it crashes when checking the number of bands in initberry.F90 (line 462).
I compared dtefieldnband_occ with nband_occ_k and got this:
dtefieldnband_occ: 159
nband_occ_k:158
My latest try was that I modified the number of valence electrons in the pseudopotentials to 316 (which is wrong, but i did it just to check if it could solve be the problem, notice that dtefieldnband_occ+nband_occ_k=317) and it started the calculation.

Dear Professor Zwanziger,
The question is whether or not abinit has the capability to do spin polarized Berry's phase.
For a ferromagnetic system the up nband will be different from the down nband.

Thanks!

Yes, berryopt works for nsppol /= 1 although not for parallel processing.

I know it works for antiferromagnetic (AFM) with nsppol /= 1 (I think even in parallel) but it doesn’t seem to in a ferromagnetic system. The problem is that nband_up and nband_down are different. It doesn’t seem ABINIT has the facility to deal with this case. Is that correct? If not, how do you set nband_up and nband_down differently. It seems to want a single number.

Currently the code requires the occupation for a given band to be identical at every k point, so I'm not sure whether it will work for the ferromagnetic insulator case. Have you tried it with nspinor 2 (rather than nsppol 2)?