[Wien] DOS plot for spin-up and spin-down are different even with the application of spin-orbit coupling. Shouldn't it couple each other. Pls help

Peter Blaha pblaha at theochem.tuwien.ac.at
Wed Apr 11 14:35:37 CEST 2012


Of course you get two different DOS plot for up and dn. Add them together and you get
the total DOS corresponding to your bandstructure.

You also get two different DOS for s and d states of an atom, although you have only ONE band structure
and not one bandstructure for s and another one for d-states ????

So once again:

Without SO: the eigenvalues for spin-up and dn aer "different" and we calculate them separate. Thus you can
plot both, the bands AND the DOS for spin-up and dn separately.

With SO: all eigenvalues are MIXTURES of spin-up and dn, thus we get them only "together" and a normal
bandstructure has both spins included.
Do the "fat band" plots for spin-up and dn for your bandstructure, you will see that some of your
bands have spin-up and others have spin-dn character.
For the DOS we do the same trick. Of course there is only ONE DOS, but we project out spin-up leading to
a spin-up DOS and the same for spin-dn. Thus we get two different DOS curves.

Maybe you understand it then.


Am 11.04.2012 10:11, schrieb Madhav Ghimire:
> Dear Prof. Blaha,
>      Thank you very much for your prompt reply.
> I am bit confused with the statement you provided for DOS (my first question).
> (a) Did you mean that the appearance of DOS plot for spin up and spin down should be coupled to single DOS plot like the band structure. *
> If it is so, I am not getting it. Rather I am getting two separate DOS for spin up and spin down. How can the problem be resolved?*
>   OR,
> (b) Did you mean that we can have both different nature of spin up and spin down DOS.
> *If it it so, then how is it possible as the SOC couples two spin into one in band structure. So shouldn't it be same? Please explain.*
>
> Sorry for the inconvenience
> I will be very grateful to you for explaining this part again.
>
> M. P. Ghimire
>
>
>
> On Wed, Apr 11, 2012 at 5:09 PM, Madhav Ghimire <ghimire.mpg at gmail.com <mailto:ghimire.mpg at gmail.com>> wrote:
>
>     Dear Prof. Blaha,
>          Thank you very much for your prompt reply.
>     I am bit confused with the statement you provided for DOS (my first question).
>     (a) Did you mean that the appearance of DOS plot for spin up and spin down should be coupled to single DOS plot like the band structure. *
>     If it is so, I am not getting it. Rather I am getting two separate DOS for spin up and spin down. How can the problem be resolved?*
>       OR,
>     (b) Did you mean that we can have both different nature of spin up and spin down DOS.
>     *If it it so, then how is it possible as the SOC couples two spin into one in band structure. So shouldn't it be same?* Please explain.
>     For your convenience, I am attaching herewith the DOS plot which are not similar for both spins. (Pls note, y-axis corresponds to DOS (states/ev) and x-axis corresponds to
>     Energy (eV)
>     Sorry for the inconvenience
>
>     I will be very grateful to you for explaining this part again.
>
>     M. P. Ghimire
>
>
>     On Wed, Apr 11, 2012 at 4:34 PM, Madhav Ghimire <ghimire.mpg at gmail.com <mailto:ghimire.mpg at gmail.com>> wrote:
>
>         Dear Prof. Blaha,
>              Thank you very much for your prompt reply.
>         I am bit confused with the statement you provided for DOS (my first question).
>         Did you mean that the appearance of DOS plot for spin up and spin down should have same plot like the band structure.
>           OR, we can have different nature of DOS for spin up and spin down.
>         For your convenience, I am attaching herewith the spin orbit coupled DOS plot whose nature differs in spin up and spin down cases showing the ferromagnetic behaviour.
>         I will be very grateful to you for explaining this part again.
>         Sorry for the inconvenience
>         M. P. Ghimire
>
>
>         On Wed, Apr 11, 2012 at 3:20 PM, Peter Blaha <pblaha at theochem.tuwien.ac.at <mailto:pblaha at theochem.tuwien.ac.at>> wrote:
>
>             SO couples spin-up and dn, thus you get only "one" band structure.
>
>             Still, you could use the "fat-bands" and indicate the spin-up and dn character of each band.
>             In most cases you will still see a clear domination of either spin-up or dn in a single ("mixed")
>             eigenvalue.
>             The same we do for the DOS: we simply project out the spin-up/dn character of each single eigenvalue
>             and plot the corresponding "partial" DOS. It is like an s and p DOS. A single eigenvalue may have both,
>             s and p character, but we project out its s (or p) character.
>
>             No, you cannot "play" with the spin-orbit coupling parameter. This is NOT an external parameter, but
>             clearly defined by quantum mechanics (proportional to dV/dr).
>
>             Am 11.04.2012 06:33, schrieb Madhav Ghimire:
>
>                 Dear wien2k developers and users,
>                      Since few weeks, I am concentrating on the calculations of spin-polarized iridates with inclusion of spin-orbit coupling. The guidance from Prof. Blaha is
>                 greatly appreciated
>                 for identifying the problem with initso_lapw and sending me the proper file. That matter is solved and the scf cycles converges smoothly for serial version (for
>                 parallel version
>                 with 32 CPU, I still observe crash in lapw2 etc) PC.
>                 Now I have one more problem to post. The problem is in DOS plot for spin-up and spin-down cases. While performing the LAPW2 calculation, I observed two different
>                 structure of DOS
>                 for spin-up and spin-down.  Could anyone tell me how the DOS cases differ from that of band structure. If this case is correct I would like to know the reason. If
>                 not correct, how
>                 can I correct. By the way, In band structure, I get the coupled bandstructure for spin-up and spin down (I mean, the up and down bandstructure are same which is
>                 due to coupling of
>                 spin up and spin down bands).
>                 Similarly, for 5d atoms, the U value is small due to which spin-orbit interaction parameter is large (~2 eV). But by using wien2k, I found that it generates by
>                 itself. In such
>                 case, how will we know whether the generated parameters are the required one for our calculations. Is there any way to change the spin-orbit coupling parameter.
>                 Please help me to solve out this problem. I will be very grateful to you
>                 Thank you in advance
>                 With best regards
>                 --
>                 M. P. Ghimire
>                 NIMS, Tsukuba
>
>
>
>                 _________________________________________________
>                 Wien mailing list
>                 Wien at zeus.theochem.tuwien.ac.__at <mailto:Wien at zeus.theochem.tuwien.ac.at>
>                 http://zeus.theochem.tuwien.__ac.at/mailman/listinfo/wien <http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien>
>
>
>             --
>
>                                                   P.Blaha
>             ------------------------------__------------------------------__--------------
>             Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
>             Phone: +43-1-58801-165300 <tel:%2B43-1-58801-165300>             FAX: +43-1-58801-165982 <tel:%2B43-1-58801-165982>
>             Email: blaha at theochem.tuwien.ac.at <mailto:blaha at theochem.tuwien.ac.at>    WWW: http://info.tuwien.ac.at/__theochem/ <http://info.tuwien.ac.at/theochem/>
>             ------------------------------__------------------------------__--------------
>             _________________________________________________
>             Wien mailing list
>             Wien at zeus.theochem.tuwien.ac.__at <mailto:Wien at zeus.theochem.tuwien.ac.at>
>             http://zeus.theochem.tuwien.__ac.at/mailman/listinfo/wien <http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien>
>
>
>
>
>         --
>         M. P. Ghimire
>
>
>
>
>     --
>     M. P. Ghimire
>
>
>
>
> --
> M. P. Ghimire
>
>
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien

-- 

                                       P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300             FAX: +43-1-58801-165982
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------


More information about the Wien mailing list