[Wien] Problem calculating optical conductivity for Fe

[Sebastian Putz]

Dear WIEN2k users, 

My goal is to calculate the optical conductivity tensor sigma for bcc bulk-Fe
magnetized in the z-direction. 

I ran a spin-polarized SCF cycle including spin-orbit-coupling until
convergence (-ee 0.000001) with 3339 k-points in the irreducible Brillouin-zone
using PBE. So far so good. Then I performed the optics calculation with the
following procedure in order to calculate the complex sigma_xx and sigma_xy
components of the optical conductivity tensor: 

kgen with 13275 k-points in the irreducible BZ 
change TOT to FERMI in case.in2c 
runsp_lapw -so -s lapw1 -e lcore 
opticc -so -up 
joint -up 
kram -up 

Analyzing the file case.sigmakup I see that Re_sigma_xx, Im_sigma_xx and
Im_sigma_xy yield very good results (compared to experiment) but Re_sigma_xy
differs from the expected curve in the following way: There is a constant
offset of approximately 0.6 (in the output units of case.sigmakup). The notion
that this offset is something unphysical is supported by two observations: 

When I manually add 0.6 to all values of Re_sigma_xy (shifting it upwards)...


(1) ...the experiment is reproduced. 
(2) ...the sumrule that the integral d(omega) over Re_sigma_xy(omega) from 0
to +infinity equals zero (omega is the photon frequency) is fulfilled. Without
the manual correction this sumrule diverges. 

Has anyone of you experienced a similar situation or do you have any
suggestions where this "error" in Re_sigma_xy comes from, if it is an error at
all? 

Any help is very much appreciated. 

Best regards, 
Sebastian Putz 

================================== 

Here are the last lines of the case.sumrules file (note that these 3 sumrules
are different from (2) above): 

   135.907310     7.3702   7.1363   1.5408     7.3729   7.1367   1.5405    
3.6399 903.7869  26.3505 
   135.934530     7.3703   7.1364   1.5408     7.3729   7.1367   1.5405    
3.6406 904.3399  26.3560 
   135.961740     7.3703   7.1364   1.5408     7.3730   7.1367   1.5405    
3.6413 904.8932  26.3615 
   135.988950     7.3703   7.1364   1.5408     7.3730   7.1368   1.5405    
3.6421 905.4466  26.3670 
   136.016160     7.3704   7.1365   1.5408     7.3730   7.1368   1.5405    
3.6428 906.0003  26.3725 

The input files are as follows (the energy ranges are rather high because this
is the last of several tries, increasing the energy range step by step) 

===== case.in1(c) ==================== 

WFFIL        (WFPRI, SUPWF) 
 10.00       12    4 (R-MT*K-MAX; MAX L IN WF, V-NMT 
  0.35    6  0      (GLOBAL E-PARAMETER WITH n OTHER CHOICES, global APW/LAPW)
 
 0    0.30      0.000 CONT 0  
 0   -6.50      0.001 STOP 0  
 1    0.50      0.000 CONT 0  
 1   -3.97      0.001 STOP 0  
 2    0.52      0.000 CONT 0  
 2    1.50      0.000 CONT 0  
K-VECTORS FROM UNIT:4  -10.0       15.0    23   emin/emax/nband 


===== case.inso ==================== 

WFFIL 
 4  1  0                      llmax,ipr,kpot  
 -10.0000   5.0000           emin,emax (output energy window) 
   0.  0.  1.                 direction of magnetization (lattice vectors) 
 0                           number of atoms for which RLO is added 
 0 0 0 0 0                    number of atoms for which SO is switch off;
atoms 


===== case.inop ==================== 

99999 1       number of k-points, first k-point  
-5.0 10.0      Emin, Emax for matrix elements 
3             number of choices (columns in *outmat) - 0: MME into case.mme 
1             Re xx 
3             Re zz 
7             Im xy 
OFF           ON/OFF   writes MME to unit 4 


===== case.injoint ==================== 

    1  9999 9999              : LOWER,UPPER and (optional) UPPER-VAL BANDINDEX

   0.0000    0.0010   10.000 : EMIN DE EMAX FOR ENERGYGRID IN ryd  
eV                            : output units  eV / ryd  / cm-1 
     4                        : SWITCH   
     3                        : NUMBER OF COLUMNS 
   0.1  0.1  0.1              : BROADENING (FOR DRUDE MODEL - switch 6,7 -  
ONLY) 

SWITCH: 
   ... 
   4...Im(EPSILON)   
   ... 

===== case.inkram ==================== 

  0.4    Gamma: broadening of interband spectrum 
  0.0    energy shift (scissors operator) 
  0      add intraband contributions? yes/no: 1/0 
 12.60   plasma frequencies  (from joint, opt 6) 
  0.20   Gammas for Drude terms 
================================
Sebastian Putz
Institute of Experimental and Applied Physics (Group Prof. Back)
Institute of Theoretical Physics (Group Prof. Fabian)
University of Regensburg
Universitätsstraße 31
93040 Regensburg, Germany
Tel.: 0941 943 -4931 (or -2032)
E-mail: sebastian.putz at physik.uni-regensburg.de



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20100827/328a4250/attachment.htm>