MAGMOM input for different magnetic states

[sankh]

Hi everybody,

Could someone please tell me how to enter MAGMOM tag in the INCAR for the following cases:

1. Super cell (AB)n, n = 12. Is it OK if I mention in INCAR, MAGMOM = 2 -1.5, where say, each of 12 indential A's has 4 spin-ups and each of 12 idential B has 3 spin-downs (or 2*12 -1.5*12, would be appropriate?)

2. Super cell (AB)nOm, n = 8 and m = 16;
A. MAGMOM = 2 -1.5 (Ignoring magentic orientations of all O atoms). Is it possible to ignore magnetic orientations of some atoms while considering for other atoms in a super cell?

B. MAGMOM = 2 -1.5 1 -1 (here I intend to divid m O atoms in two classes, one with all ups and the other one with all downs)

C. Say I want to consider only 1st, 7th and 23rd atoms magnetic moments in this supercell and for rest I don't need. In this case, shall I specify Magmom = 1*2 7*-1.5 23*-1 in the INCAR (where 1st atom has 4 spin-ups, 7th atom has 3 spin-downs and 23rd atom has 2 spin-downs?

I presume that if one needs to enter the magnetic moments for each atoms in INCAR, then POTCAR will be too large for a super-cell, although in that super cell some indentical atoms can be grouped into one or two or four or so on classes of magnetic orientations and this will reduce the size of the POTCAR.

Well, I have never done before any calculations into magentic states using VASP, so I shall be very grateful if someone could kindly help me out toward understanding the input requirements for initial calculation. Thank you.

Best regards.
Sankh
<span class='smallblacktext'>[ Edited ]</span>

[cmuhich]

Hi,

1: you should use 12*4 12*-3. You should also probably over exagerate the spin. If in reality the spin is 4 spin up and 3 spin down, use maybe 4.5 and 3.5.

2) A. You still have to input the spin, so 8*2 8*-1.5 and 16*0. The oxygens will figure out the correct spin for the O's

B. you would still need to say how many oxygens are spin up and spin down so 8*2 8*-1.5 8*1 8*-1

C. if you consider some atoms to be magnetic, the all are going to end up being magnetizable. In this input your saying 1 atom has spin +2, 7 atoms have spin -1.5 and 23 atoms have spin -1.

[sankh]

Hi Cmuhich,

Thanks a lot for your reply and elaboration. Now I have got a much clear idea about MAGMOM input syntax. However, as you explained, for ex in 8*2 8*-1.5 16*0, the first numbers are the number of indentical atoms having same magentic moments.

Therefore, I am still not sure whether or not it is possible, as well, to enter a specific atom's (say, by the position number of that atom in the super cell) magnetic moment in the INCAR, which is basically my question "C". So, could you be kind enough once again to tell to me whether or not it is possible, and the syntax if possible? I sincerely thank you for this advance.

Regards,
Sankh

[forsdan]

The MAGMOM tag needs one entry assigned per atom. If you don't think there will be any magnetic moment on certain atoms you can assign them to zero. In your case C you can then write

MAGMOM = 2 5*0 -1.5 15*0 -1 0

if you now you have 24 atoms. However, please note that the MAGMOM entries are only an initial guess and that magnetic order of the self-consistent solution can differ from what you assign.

Cheers,
/Dan

[sankh]

Hi Dan,
Thank you very much for further help and clarification. Now I got the full picture.
Cheers,
Sankh