[Wien] Spin-polarized and spin-orbit

[Gerhard Fecher]

If you include spin-orbit interaction
then only one kind of bands exists (there are no up or down bands anymore)
however, you still can calculate the spin resolved DOS.

Ciao
Gerhard
________________________________________
Von: wien-bounces at zeus.theochem.tuwien.ac.at [wien-bounces at zeus.theochem.tuwien.ac.at] im Auftrag von Lukasz Plucinski [pluto at physics.ucdavis.edu]
Gesendet: Donnerstag, 12. Juni 2008 23:59
An: A Mailing list for WIEN2k users
Betreff: Re: [Wien] Spin-polarized and spin-orbit

Thank you for the quick answer.

I still cannot produce what I need. After creating case.klist_band --
should I run (exactly):

 x lapw1 -band -up
 x lapw1 -band -dn
x lapwso -band -up
x lapw2 -band -so -qtl -up
x lapw2 -band -so -qtl -dn
x spaghetti -so -up
x spaghetti -so -dn

This does not work, and at the moment x spaghetti -so -up gives an error.

I believe I don't understand some simple thing. If you look at Rampe et
al.PRB 57, 14370 (1998) you will see that the same band might have up or
down character for different part of BZ, as a result of band gap created
by spin-orbit due to hybridization. Plotting majority and minority
components separately  seems not possible anymore. But I don't see how
to plot both of them in one figure.

Lukasz



Robert Laskowski wrote:
> lapwso produce spin up and dn projected vectors. The corresponding energy files are the same for up and dn spin chanels.
> After  x lapwso -up  run x lapw2 -so -qtl -up/dn  and then x spaghetti -so -up/dn
>
> regards
>
>         Robert
>
>
> On Thursday 12 June 2008, Lukasz Plucinski wrote:
>
>> Ok, let's say I went through the fcc_Co calculation, as explained in P.
>> Novak's script on spin orbit.  I have used LSDA, 110 magnetization
>> direction, mixing 0.2, and the rest default.Program went successfully
>> through the series of iterations with following pattern:
>>
>> Thu Jun 12 06:58:52 PDT 2008> (x) lapw0
>> Thu Jun 12 06:58:54 PDT 2008> (x) lapw1 -up
>> Thu Jun 12 06:58:56 PDT 2008> (x) lapw1 -dn
>> Thu Jun 12 06:58:57 PDT 2008> (x) lapwso -up
>> Thu Jun 12 06:58:58 PDT 2008> (x) lapw2 -c -up -so
>> Thu Jun 12 06:58:59 PDT 2008> (x) lapw2 -c -dn -so
>> Thu Jun 12 06:59:00 PDT 2008> (x) lcore -up
>> Thu Jun 12 06:59:00 PDT 2008> (x) lcore -dn
>> Thu Jun 12 06:59:01 PDT 2008> (x) mixer
>>
>> Now I've defined k-points for band structure and I would like to plot
>> it. Which series of iterations should I use now ? I tried:
>>
>> Thu Jun 12 07:05:07 PDT 2008> (x) lapw1 -band -up
>> Thu Jun 12 07:05:35 PDT 2008> (x) lapwso -band -up
>> Thu Jun 12 07:06:36 PDT 2008> (x) spaghetti -so
>>
>> and similar, then even managed to plot some picture. But this is not
>> right, since now both up and dn bands must be in one image - some of
>> them suppose to change from up to down, I guess. Could you advise which
>> sequence of command-line commands should I use to get such band structure ?
>>
>> Lukasz
>>
>>
>>
>>
>> Robert Laskowski wrote:
>>
>>> hi,
>>> it is rather ordinary calculation. Because of magnetism you have to go through  initso,
>>> which will be called at the end of init_lapw
>>>
>>> regards
>>>
>>> Robert
>>>
>>> On Thursday 12 June 2008 11:34:12 Lukasz Plucinski wrote:
>>>
>>>
>>>> Most likely this was already covered, however, maybe someone can give me
>>>> quick answer or a reference.
>>>>
>>>> Experiment is MLD on Co/Cu(100). I would like to calculate
>>>> spin-polarized (sp) band structure of fcc Co with spin-orbit (SO).
>>>>
>>>> Is is possible ? Which parameters are most crucial ? How to include SO ?
>>>> -- should I calculate sp first, and then add SO ?
>>>>
>>>> Then I would like to know the symmetry of such bands -- as Prof. Blaha
>>>> mentioned this is included in 'irrep' module, but does it work when both
>>>> sp and SO are included ?
>>>>
>>>> I believe that sp with SO is not an easy task. Are there any fundamental
>>>> limitations in Wien2k ?
>>>>
>>>> Lukasz
>>>> _______________________________________________
>>>> 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
>>
>>
>
>
>
>

_______________________________________________
Wien mailing list
Wien at zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien