[Pw_forum] crystal_sg in QE input reg
Dae Kwang Jun
jdaekwang at gmail.com
Fri Jul 1 11:07:50 CEST 2016
Dear Muthu V,
This is what I understand, it may not be correct. I hope that anyone that
notices a mistake corrects it.
You will find that most of the crystals are in Wyckoff positions. You can
find the structure of crystals in websites such as
In the COD website you will find them inside the cif files. The Wyckoff
positions are the ones after _atom_site_calc_flag.
You should specify this. It is given in the website and cif files.
*nat = # independent positions*
The number of atoms listed in the cif file or website should be the number
of independent positions. Independent positions are positions that cannot
be generated by symmetry of the space group. Simply put, the number of
independent positions is equal to the number of lines. For the case of
GaAs, there are two positions listed. One for Ga and one for As.
*ntyp = # of different types of atoms*
The input for this doesn't change because you are using crystal_sg. GaAs
has two types of atoms. Ga and As. Often times the number for nat will be
equal to ntyp, but they are not the same thing.
The usual format for them is
Element Site coordinate
An example for GaAs space group 216
An example for ZnO (wurtzite structure) space group 186
Zn 2b 0
O 2b 0.3825
You only need to specify the coordinate if it is needed. You can check
whether it is needed or not with
Put the space group of the crystal. For space group 186 site 2b, you can
see that the coordinates are(1/3,2/3,z), (2/3,1/3,z+1/2). You only need the
value of z to generate it. So you only need to take the z value of the
For space group 216 site 4a the coordinates are (0.0,0). Notice that it
does not have any x, y, or z included in the coordinate. Thus, we do not
coordinates for this site. In summary, go to the website, check the site,
check the positions, check if the positions have x, y, or z in them. If
they do, you need
to included the x, y, or z positions in ATOMIC_POSITIONS.
On Fri, Jul 1, 2016 at 1:50 PM, Muthu V <muthu.physicsmath at gmail.com> wrote:
> Dear D K Jun
> I'm Muthu V working in DFT field using QE. I'm trying to define my system
> in-term of crystal_sg method in input file. i have gone through your
> discussion in pwscf forum. but still i need some more clarification on
> defining the structure. in this case what about nat and ntyp tages in
> SYSTEM block.
> and when i have specify
> like C 1a
> like C 8g x
> like C 24m x y
> like C 48n x y z
> can you give some example.
> look forward to your mail
> Thank you
> i followed your discussion on
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the users