[Wien] Non-collinear magnetism in WIEN2k? Total magnetic moment, spin moment and orbital moment...

[Pablo de la Mora]

Dear Pavel,
    Thank for your answer, I would like to ask you for some fine points:
The question is; The final direction of the spin and orbital moments 
would correspond, in a first approximation, to a result of a 
non-collinear calculation?
In my case one of moments seems to point towards the axis of symmetry, 
but another one points away from it!
In other words, Can these results point towards a non-collinear 
magnetization? And for small angles these results can be regarded as 
correct?

As for the second question then the total moment is the sum of spin and 
orbital ones, is this correct? I ask for this detail since it is 
confusing in the SCF file when it says MAGNETIC MOMENT IN SPHERE and 
TOTAL MAGNETIC MOMENT IN CELL implying that this is the sum of all the 
contributions, s, p, d and f, but also of both spin and orbital. But I 
think this is not the case since the TOTAL MAGNETIC MOMENT IN CELL does 
not take into account the orbital term! If this is the case it should be 
pointed out in the SCF file since it misleads.

       Pablo

Pavel Novak escribió:
> Dear Pablo,
>
> if M is not along the symmetry axis, the spin orbit coupling will cant the
> spin out of the direction prescribed for magnetization in case.inso. Also
> orbital moment will deviate from it. Both spin and orbital moment are
> calculated in LAPWDM, which allows determination of all components of
> these vectors. Note that there is some inconsistency: M direction is
> prescribed in case.inso, but after finishing the calculation, you'll find
> magnetization in different direction. The reason is that you keep along
> the selected direction the exchange field, but the total field is sum of
> the exchange field and 'field' from spin-orbit coupling s<l>.
>
> As to your second question: in LAPWDM you calculate the spin moment from
> selected electrons only (usually d or f), while moment in the sphere is
> the sum from all electrons: you can check it by running LAPWDM for all s,
> p, d and f states.
>
> Pavel
>
>
>  On Mon, 21 Apr 2008, Pablo de la
> Mora wrote:
>
>   
>> When one adds spin-orbit coupling for a magnetic system one has to
>> define a magnetization direction in '<case>.inso':
>>    0.  1.  0.                 direction of magnetization (lattice vectors)
>>
>> The results of the calculation come out in the <case>.scf file. There
>> are three different results, the total magnetic moment, the spin moment
>> and the orbital moment:
>> MAGNETIC MOMENT IN SPHERE   1    =    1.24926
>> SPIN MOMENT:  -0.03321   1.15380   0.00000 PROJECTION ON M  1.15380
>> ORBITAL MOMENT:  0.14357 -1.02102  0.00000 PROJECTION ON M -1.02102
>>
>> There are two kind of questions that I have:
>> 1) Why are these spin and orbital moments given in vector form?
>>     Does this imply a kind of non-colinear magnetism?
>>     If this is the case I would presume that it is valid for small
>> deviations from the magnetization direction, in this case the deviation
>> is of 8 degrees for the orbital moment and 1.5 degrees for the spin
>> moment. For these relatively small angles, are these values correct?
>>
>> 2) In the <case>.scf file :MMI001 gives the 'MAGNETIC MOMENT IN SPHERE',
>> but with spin-orbit coupling the 'SPIN MOMENT' and the 'ORBITAL MOMENT'
>> appear. So what is the total magnetic moment in this case?
>>
>> Adding spin and orbital gives:
>> 0.11036  0.13278  0.00000 PROJECTION ON M 0.13278
>>
>> as it can be seen it is much smaller than the MAGNETIC MOMENT IN SPHERE.
>>
>> I have observed that magnetic moment in sphere and the spin moment are
>> quite symmilar but not equal, Do they have, in principle, the same
>> meanning? Which is the correct one?
>>
>>     Yours
>>
>>                    Pablo de la Mora
>>
>> _______________________________________________
>> Wien mailing list
>> Wien at zeus.theochem.tuwien.ac.at
>> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>>
>>     
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>
>