Dear Andrei,

Thank you.

With Best Regards, Andrei Buin.



On Thu, Aug 22, 2013 at 6:34 AM, <[email protected]> wrote:

> Dear Andrei:
>
> The calculation of band structure is just an extra calculation
> over selected k points. Technically it is independent on whether
> your previous (self-consistent) calculation was done just with Gamma,
> or with more k-points, be it a single cell or a supercell.
>
> As you pass to a supercell, the BZ is reduced; you have more (folded) bands
> which disperse less, because the k-path becomes shorter.
>
> Take care of how you define the path; there is a potential danger
> to mess up absolute / relative units, and relative with respect
> to single cell lattice constant, or the supercell one.
>
> Best regards
>
> Andrei Postnkov
>
>
> > Let's say i have computed supercell with Gamma point only. Later i
> decided
> > to look for a bandstructure and specify path in bandLines block G-X-M-L
> > ....
> >
> > I receive flat band picture, although it should not be the case. My
> > question is, will Siesta sample at this points defined along different
> > paths or is it completely wrong to start out with Gamma point
> > calculations,
> > and then specify BandLines block ?
> >
> > With Best Regards, Andrei Buin.
> >
>
>

Responder a