Spin polarized calculation in CNT

[fly]

APPLIED PHYSICS LETTERS 97, 043115 (2010). In this article, the author demonstrate that hybrid BN–C
nanotubes (BN-CNTs) have diverse electronic and magnetic properties. It is konwn that  the ferromagnetic, anti-ferromagnetic, or ferrimagnetic states must be considered for graphene nanoribbon. But, how can we set the value of " initial_scaled_spin"  in this configure. Who can give me some tips?

[zh]

From the results presented in that paper, the spin polarization is mainly localized around the conjunction region between BN-C, i.e, see their results shown in Fig1.(d) in that paper. The initial magnetic moments of atoms near the conjunction are set as usual. I guess that paper considered the ferromagnetic states. 

[fly]

It means that the ferromagnetic and anti-ferromagnetic are considered for GNR and BN nanoribbons , respectively. For example, when we consider FM states in Fig1.(d) in that paper , we should set FM states in edge atoms of GNR and BN nanoribbons in the conjunction region.  That is to say, when we consider the initial spin configures of BN-CNTs, we can simulate the case of bilayer GNRs (J. Phys. Chem. C 2010, 114, 13098–13105). Is it right ?

[zh]

Yes, it is right. For the FM state, the atoms with large magnetic moment are all set as a same initial spin. For the AFM states of graphene nanoribbons, the initial spins of atoms at one edge are opposite to those of atoms at the other edge.