[Wien] Optics with spin-orbit interaction

[whoming]

Dear P.Blaha and Vidya Ravindran,

           I am still some confused by my result of spin-polarised optic caluculation with SO
coupling of bcc-Fe.
           I have set case.inop to calculate the matrix element 1(Re<x><y>) and 7(Im<x><y>).
           My calculation sequence is almost the the same as Vidya's except that I didn't chage 
TETRA to ROOT in the third step, because the USERGUIDER book suggests TETRA.
          But  I have got totally different case.jointup and case.jointdn, not the same as Dr.Vidya's 
result.
          I have checked that the case.vectorsoup and case.vectorsodn are different, but case.weightup 
and case.weightdn are the same. 
          I have also done x opticc -so -up and also x opticc -so -dn. The case.outmatup/dn files are
different, so after x joint -up/x joint -dn the case.jointup/dn are not equal too.
          According to Prof. P. Blaha's explanation, up/dn should be the same in the SO calculation,
because the mixer of spin up/dn.
          So what should I do to get the correct optical properties in such case? How can I get the 
contribution from spin-up and spin-dn?And how to get the total contribution from both spins? 
          I am really confused. Many thanks for your help!

Yours,
Whoming
----- Original Message ----- 
From: "Peter Blaha" <pblaha at zeus.theochem.tuwien.ac.at>
To: <wien at zeus.theochem.tuwien.ac.at>
Sent: Tuesday, July 01, 2003 3:02 PM
Subject: Re: [Wien] Optics with spin-orbit interaction


> > I am trying to calculate optical properties of a ferromagnetic
> > system using WIEN2k_03 (Release 18/01/2003). I start with Fe as a test
> > case. When I do calculation without SO, the Epsilon_2 spectra for up-spin
> > and down-spin are different.
> >
> > However when I include SO, both *.outputjointup and
> > *.outputjointdn are exactly the same. I am getting the same plasma
> > frequency for up and down-spin. I will try to describe the procedure that
> > I followed:
> >
> > 1. I did the calculations upto self-consistency.
> > 2. Incresed the k-mesh.
> > 3. Replaced TOT by FERMI in the 1st line and TETRA by ROOT in 3rd line
> >    of *.in2c
> > 4. runsp_lapw -so -s lapw1 -e lcore ; to get FERMI weights
> > 5. x opticc -so (-up/-dn) and
> > 6. x joint (-up/-dn)
> >
> > I believe that the above-mentioned procedure is correct. I am
> > unable to understand why I am getting same spectra for up and down spins.
> > Herewith I have attached *.inso, *.inop and *.injoint files. Could you
> > kindly look into those files and tell me why I am getting the
> > same Epsilon_2 spectra for both spin channels?
> 
> Spin-orbit couples spin-up and down! Thus each single eigenstate has a
> mixture of spin-up and dn. When calculating the matrixelements, also
> "crossterms" occur and a strict seperation of up and dn is no longer
> possible. Thus when calculating "-so -up" the -up (or -dn) just tells the code
> that the calculation is spinpolarized, but otherwise it calculates
> the total epsilon.
> 
> Regards
>                                       P.Blaha
> --------------------------------------------------------------------------
> Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
> Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
> Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
> --------------------------------------------------------------------------
> 
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>