My Community

In the OSCAR output from a spin-density calculation a number is
given titled mag= ..., thus
5 F= -.30124170E+03 E0= -.30123533E+03 d E =-.154285E-05 mag =0.103053E+01
No where in the VASP write up do I find what units these are. Is this Bohr magnetons, or just the net z-component of the total electon spins in the simulation
cell defined in POSCAR? For example, if I am doing a SDFT calculation for a
vacancy in a cell containing N atoms, the vacancy represents a defect of say
3 electrons, then the magnetization maximum would be 3 up-spins. What units
would VASP report this in under mag=...?

the magnetization as written in OSZICAR is the difference between the spin up and down charge (number of electrons) , as integrated over the complete unit cell. Please check if your k-mesh was sufficient and if magnetic charge density convergence was reached

So, if I pull out an sp2 hybridized carbon from a graphite like
network, since I am leaving 4 electrons in the vacancy, it beats me how I can get approximately one net "up" electron. I will try your suggestion of checking k-points etc. Thanks.

that is strange, indeed! please check your results for convergence.

Dear chandre,

This is slightly off the question about magnetism, but I must ask anyway. Why do you expect to "leave 4 electrons in the vacancy" when you remove a carbon atom from graphite? When you remove the atom you remove all of its electrons, too. Unless you explicitly add electrons back to your cell (using NELECT), your carbon vacancy will be a neutral defect.

Do you not agree?

Dear Scerwin
Your are right. When I remove a carbon, I remove all its electrons.
Of these, the inner 2 electons in the 1s do not enter into the pseudopot.
So we only have 4 electrons (neutral), but there are 4 dangling electrons left. What happens to them is the object of the calculation.