[Wien] question on irrep (irreducible representation) calculation

[Gavin Abo]

Yes, the "WILL BE IMPLEMENTED" message in your email seems to clearly 
indicate that the more complicated group theory is still not implemented 
in WIEN2k 14.2 [ http://www.wien2k.at/reg_user/limitations/ ].

For a non-spin polarized calculation with spin-orbit, I think you can 
still find whether it is non-symmorphic or not by checking the 
case.outputs file since the symmetry is not changed in the struct file 
by initso [ 
http://www.mail-archive.com/wien%40zeus.theochem.tuwien.ac.at/msg08031.html 
], and you could try the shifting of k-vectors trick [ 
http://www.mail-archive.com/wien%40zeus.theochem.tuwien.ac.at/msg08217.html 
].

On 8/13/2015 3:23 AM, 남태식 wrote:
> Hi. I'm using WIEN2k_14.2 version and have probelm on irrep 
> (irreducible representation) calculation.
> My system is CeFe2Al10 with spacegroup #63 (Cmcm).
> I implemented scf cycle, with spin-orbit coupling (init_lapw and 
> initso_lapw) & spin non-polarized, and plotted band with bandpath 
> Z-R-S-GAMMA-Z-T-Y-GAMMA. When I tried to ! compile irrep calculation 
> by x -irrep -so -p command (I implemented with parallel job), some of 
> the symmetric points-Z,R,T gives the result like below in 
> case.outputirso_1 file
>
>
> ********************************************************************************
>
>
> knum =  1    kname= Z
> k = 0.000000 0.000000 0.500000
>
>  The point group is D2h
>  8 symmetry operations in  8 classes
>  Table 23   on page  39 in Koster  et al [7]
>  Table 31.4 on page 247 in Altmann et al [8]
>
>              E    C2    C2`   C2"    I   IC2   IC2`  IC2"
> & nbsp;G1+   A1g   1     1     1 &! nbsp;   1 1     1     1     1
>  G2+   B3g   1    -1     1    -1     1    -1     1    -1
>  G3+   B1g   1     1    -1    -1     1     1    -1    -1
>  G4+   B2g   1    -1    -1     1     1    -1    -1     1
>  G1-   A1u   1     1     1     1    -1    -1    ! -1    -1
>  G2-   B3u   1    -1     1    -1    -1     1    -1     1
>  G3-   B1u   1     1    -1    -1    -1    -1     1     1
>  G4-   B2u   1    -1    -1     1    -1     1     1    -1
>  --------------------------------------------------------
>  G5+   E1/2g 2     0     0     0     2    ! ; 0     0 0
>  G5-   E1/2u 2     0     0     0    -2     0     0     0
>
>  labeling of IRs can change due to choice of
>  symmetry axes: G2 <-> G3 <-> G4
>
>  Non-symmorphic crystal and k-point at the BZ surface:
>  IR of the space group for this k-point cannot simpl! e
>  be expressed as IR of the corresponding point group
>  times a phase factor, since exp(-i*k(Ri*tj-tj)) not 1
>  for all pair of {Ri|ti} and {Rj|tj}.
>
>
>
>  WILL BE IMPLEMENTED
>
>
> ********* 
> **************************************************************! *********
>
> When I take GAMMA-Y-F0-S-GAMMA-Z-T-G0-R-Z bandpath, there was no 
> result like this one in R,Z,T point. I wonder why I have this kind of 
> result in certain bandpath (Z-R-S-GAMMA-Z-T-Y-GAMMA) although the 
> parity or irrep result implemented on symmetric point shouldn't depend 
> on the bandpath. Isn't the latest version of WIEN2k can't implement 
> irrep calculation on Non-symmorphic crystal? Is this structre 
> non-symmorphic crystal? Importantly, why this discrepancy in irrep 
> result happens between two different bandpaths?
>
> Many thanks!
>
> Nam
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20150813/8fbb3776/attachment.html>