Meaning and units of 'magnetization (x/y/z)' vs 'orbital moment (x/y/z)' in OUTCAR (LSORBIT=T, LORBMOM=T)

Queries about input and output files, running specific calculations, etc.


Moderators: Moderator, Global Moderator

Post Reply
Message
Author
lorenzor
Newbie
Newbie
Posts: 1
Joined: Thu Apr 16, 2026 6:36 pm

Meaning and units of 'magnetization (x/y/z)' vs 'orbital moment (x/y/z)' in OUTCAR (LSORBIT=T, LORBMOM=T)

#1 Post by lorenzor » Mon Jul 13, 2026 7:00 pm

I am running non-collinear, SOC-enabled calculations with:

- LSORBIT = T
- LORBIT = 11
- LORBMOM = T

With LORBMOM = T, OUTCAR prints an "orbital moment (x/y/z)" table, which I understand to be the orbital magnetic moment projected onto each atom.

OUTCAR also prints a separate "magnetization (x/y/z)" table.

My question is: do these magnetization (x/y/z) values represent the total magnetic moment of an atom (spin + orbital), or the spin magnetic moment only? And what are the units (for both the "orbital moment (x/y/z)" and "magnetization (x/y/z)"?


jonathan_lahnsteiner2
Global Moderator
Global Moderator
Posts: 302
Joined: Fri Jul 01, 2022 2:17 pm

Re: Meaning and units of 'magnetization (x/y/z)' vs 'orbital moment (x/y/z)' in OUTCAR (LSORBIT=T, LORBMOM=T)

#2 Post by jonathan_lahnsteiner2 » Tue Jul 14, 2026 11:21 am

Dear lorenzor,

The magnetization (x/y/z) represents the spin-only magnetic moment. It does not include the orbital contribution.
The orbital moment (x/y/z) represents the orbital-only magnetic moment.
Total Moment: To find the total magnetic moment of an atom, you need to manually sum the two vector components: Mtotal​=Mspin​+Morbital​

Units: Both quantities are reported in Bohr magnetons (μB​).

Please when computing the total magnetic moment note: If you are using the default SAXIS = 0 0 1, you can sum these vectors directly. However, if you have customized SAXIS to a different direction, the spin moments are projected in that rotated spinor frame, while the orbital moments always remain in the default Cartesian frame. You will need to rotate the spin vector back to the Cartesian frame before adding them together.

I hope this is of help otherwise feel free to contact me again.

All the best Jonathan


Post Reply