The ABINIT Users Mailing List ( CLOSED )

[Steven Homolya <Steven.Homolya@spme.monash.edu.au>]

On Fri, 10 Oct 2003 12:57 am, homegchiche@yahoo.fr wrote:
> Dear Abinit Users.
>
> Thank to Steven Homolya, Allan, Douglas C Dr and all abint users for
> responces for my questions. Now I have create my input files for nickel
> with 7 atom and one vacancy. but I don't know if it is complete or no. This
> is my input files calculation for unralaxed system :
>

Your supercell cell consisting of 1x1x2 conventional fcc unit cells is not a 
good choice. The vacancies will be very close along 100 and 010 but 
reasonably far apart along 001. Stick with cubic supercells. For an 8 site 
supercell use 2x2x2 fcc unit cells (that's 32 atoms) and make that supercell 
fcc (i.e. 32/4 = 8 atoms), something like,

rprim
0.0 0.5 0.5
0.5 0.0 0.5
0.5 0.5 0.0

acell 3*13.3

xcart
... put "Cartesian" coordinates of 8 sites here
(or just put 7 sites, and then get rid of vaclst and vacnum lines in the file)
You can use reduced coordinates, but I find them little awkward here.

If you go with Cartesian coords, make sure to choose the 8 (or 7) sites so 
that they are "inequivalent", i.e., you can't go from any one site to any 
other site using primitive vectors.

I think you'll find that an 8 atom supercell is too small (but it's a good 
place to start). To go up to 16 sites use a 32 site bcc structure. Next one 
up from there is 27 sites, with 108 sites in conventional fcc unit cell.

One more thing: Avoid using chkprim 0. A non-primitive unit cell is at best a 
waste of CPU time & RAM, but it usually tells me that I didn't get the 
geometry of the system right.

Steve

-- 
Steven Homolya
School of Physics and Materials Engineering
Monash University VIC 3800
Australia
Tel: INT +61 3 9905 3694
Fax: INT +61 3 9905 3637