The ABINIT Users Mailing List ( CLOSED )

[Matthieu Verstraete <mjv500@york.ac.uk>]

On Wed, 2 Jan 2008, PGanesh wrote:

> I actually get a somewhat jagged plot. So I was wondering if it could be 
> because I am not entering the q-vectors correctly.
No: for the DOS you need a homogeneous and dense grid, not just the 
special kpoints. The DOS is an integrated quantity. If you are using the 
standard method (along with thermodynamical quantities) the grid is given 
in the anaddb input file (see 
http://www.abinit.org/Infos_v5.5/users/anaddb_help.html#ngrids
http://www.abinit.org/Infos_v5.5/users/anaddb_help.html#nwchan and related 
variables), and needs to be densified. Usually this is cheap and you may 
go up to 30x30x30 or more to get a nice smooth curve. Doing an integration 
over special points only is bad, as there are no weights in the sum and 
they are not dense enough anyway.

> What is the best way to obtain the special q-points for a given space group? 
> Should the entries in crystallographic server be good enough?
should be. Otherwise check the litterature. Any compound with the same 
space group will do of course, but there are different choices for the 
ordering of the kpoint path

M.
--
================================================================
Dr. Matthieu Verstraete                 mailto:mjv500@york.ac.uk
Dept. of Physics, University of York,     tel: +44 1904 43 22 08
Heslington, YO10 5DD York, United Kingdom fax: +44 1904 43 22 14