Difference in magnetization values

[l.bellarosa]

Hi everybody,

I found a related post, dating November 2012, but nobody replied to that, so I thought that it could be a good idea to repost it.

I have a very simple molecule, NO, in a 18x18x18 ang box. Calculations are gamma-centered and cutoff energy is 415 eV. Optimization criteria are: EDIFF = 0.1E-05, EDIFFG=-0.015. IBRION=2.
I optimized it with ISPIN=2, MAGMOM= 2*3.0. The result is a geometry comparable with experimental data.

According to OUTCAR:
magnetization(N): 0.46; magnetization(O): 0.24. ==> total magnetization: 0.70

But, in the OSZICAR:
magnetization=1.0000

So.. what does it mean exactly the magnetization in the OSZICAR and in the OUTCAR? Why are they different?

Thank you in advance,

Luca

[l.bellarosa]

Hi everybody,

it is still me. I managed to find out why the total magnetization in the OSZICAR is different from the sum of the magnetization projections in the OUTCAR. For this reason, I will reply to myself in case somebody else needs this explanation in the future.
The difference in the values roots in the fact that, when you project the magnetization on the atoms, you use a default value (RWIGS) to define the radius of each atom. For this reason, even if you compute the magnetization for all the atoms, the sum of their "volume" does not fill up the total volume of the unit cell, and so the sum of the projections of the magnetization on each atom is different from the overall magnetization, written in the OSZICAR.

Thanks for everything,

Luca