[Wien] monoclinic C

[Peter Blaha]

>  I have a question regarding the bug in monoclinic C. Is the problem below
> coming from that known bug, and is there a smart way around it?
> 
>  I am dealing with space group C2/m, and can run the calculation just fine in
> P2/m. However, then I have too many bands of course, as the primitive volume
> is doubled. Hence only the DOS is correct, not the spaghetti diagram.
> 
> Running x sgroup results in C2/m, unique axis c. First problem shows up in
> xsymmetry (-3 not found).
> The 3-fold axis bothers me of course, considering the monoclinic system. In
> the outputs file I found:
> 
> Cu1        operation #  1     1
> Cu1        operation #  2     -1
> Cu1        operation #  5     2 || z
>   pointgroup is 3 (neg. iatnr!!)
>   axes should be: 3 || z


Are you using a recent version of "symmetry" ? If no, please update, if
yes, send me the struct file (to my private email) together with some 
information about the desired crystal structure.

The error in dstart comes because the input files were not created correctly in symmetry.

PS: The spaghettis are not "wrong" ! They are just represented in a doubled
unitcell resulting in twice as many bands due to backfolding.

                                      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/
--------------------------------------------------------------------------