which is a more reasonable way to control unpaired electrons ?

[zongtan_fang1]

Hello,

Just curious, to get a desired total number of unpaired electrons (difference between spin-up and spin-down), one can either try different MAGMOM or set NUPDOWN to the total number, which way is more efficient ?

Does it make sense to combine MAGMOM and NUPDOWN parameters in the INCAR ?

Thanks a lot.

Best,

Zongtang

[alex]

Hello Zongtan,

yes, this is a very powerful way if you want to investiage e.g. a set of different spin states.

Cheers,

Alex

[martin.schlipf]

Just to be clear: MAGMOM sets the initial magnetic moment. The code will modify this value during the SCF if it is energetically favorable to do so. NUPDOWN enforces that the system exhibits a certain magnetic moment.

If you want to enforce a certain spin state, you set NUPDOWN. Then it is beneficial (time to convergence) to specify which atoms have the magnetic moments using MAGMOM.