[Wien] why the oscilation appear when I use supercell

Peter Blaha pblaha at theochem.tuwien.ac.at
Tue Oct 25 09:22:57 CEST 2005


Metallic !! supercells are always more difficult to converge than small
unit cells (or non-metals). There is much more freedom to move charge from
one atom to another.
In particular when using a supercell, but NO symmetry break (by impurities,...)
there is lots of "degeneracy", which however is not forced by symmetry 
constrains, but should come up itself. And this does not happen so easily,
since at EF two states may flip there order all the time).

You will need:  
small mixing parameters (even 0.01 in case.inm)

Use a "broadening method" in case.in2 (TEMP or GAUSS with 0.020; the latter
should be tryed to be reduced to about 0.05 at a later stage )

a better k-mesh ?


> >       I try to calculate the electronic structure of TaS2. The results
> > are very good when i use the unitcell(including one Ta atom and two S
> > atoms)as tructure file.But when i use the supercell (3*3*1
> > including 39 atoms) in structure file ,oscilation appear the
> > selfconsistent iterations.and do not reach a convergent (-cc 0.0001 same
> > as using the unitcell).what is the problem? thank you !!!!


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


More information about the Wien mailing list