[Pw_forum] Questions on display Fermi Surface(continued)

Tone Kokalj tone.kokalj at ijs.si
Wed Jun 8 08:22:06 CEST 2005


On Sun, 2005-06-05 at 12:51 +0000, å¼  洪彬 wrote:
> Thank you for you kind guide first.
>

>     1. In the sample Fermi Surface file, when define "first
>     spanning vector of the bandgrid (i.e. first reciprocal lattice
> vector)" and the two  follows, the data are different from the PWSCF
> format I think, as to the  first reciprocal lattice vector, say, in the
> example file it gives  C(i+j-k), where C is a constant; but in the PWSCF
> format, the first  reciprocal lattice vector should be D(-i-j+k) format ,
> and further more,the  two constants C, D don't have the same value or a
> value I can interprete in  an reasonable way.

The Rh-fcc example was not computed by the PWscf, therefore the
direction of the lattice vectors is different. About the constants: they
are 1/a_0 and not 2pi/a_0 in that file. The contant is not that
important, because xcrysden merely plot the Fermi surface, and by only
scaling the vectors the FS picture is the same.

>     2. I wonder if there is a convenient way to get energy values which
> should be inputted in the XSF Fermi Surface file. You see, in the XSF
> file,  one must get energy for a K-grid in the whole reciprocal unit
> cell, while  calculation performed by PWSCF usually in the Brillouin
> Zone. Do I have to  get it manually one by one for those K_points beyond
> the first Brillouin  Zone?

You will have to do that by yourself with some coding. You need to use the
relations: E(k) = E(-k) and E(k+K) = E(k), hence you can rearrange all
the k-points into the whole reciprocal unit cell (use also nosym
= .true. during the PWscf calculation)

Regards, Tone






More information about the users mailing list