Hi Marcos
I think that you procedure is fine. The problem with vesta is that it
needs yo save changes by step. I mean, first Open your file and remove
the symmetry, and get a p1 structure. Then save it as a cif file.
Afyer this read this file and perform the transformation, and save it
as a different cif file. Finally load it and apply the symmetry, if
you want, and get another cif file. If you have further problems send
me your sf and i'll try to check it
Regards
Ricardo
--------------------------------------------
Dr. Ricardo Faccio
Prof. Adj. de Física
Av. Gral. Flores 2124. CC 1157. CP 11800.
Phone: + 598 2 924 9859
Fax: + 598 2 924 1906
Web: http://cryssmat.fq.edu.uy/ricardo/ricardo.htm
---------------------------------------------
El 06/04/2011, a las 10:27, Marcos Veríssimo Alves <marcos.verissimo.al...@gmail.co
m> escribió:
Hi all,
So, great to know there are knowledgeable people on Vesta in the
mailing list. My question is the following: I have a cubic
structure, a perovskite. I have a [sqrt(2) x sqrt(2)] cell of this
structure, with relaxed positions with rotated oxygen octahedra in
the x-y plane only, out of which I would like to create a [2 x 2]
cell. I actually managed to do it (using a xsf file that i created
from the xyz output), but in a way which I would not call the most
elegant. It is also a way that is more prone to errors, than I wish
it were.
The problem is, when one loads an xsf file with the coordinates in
Vesta, it displays a structure with more atoms than the unit cell
actually does, and apparently it is the equivalent of XCrysDen's
conventional cell mode. However, one can choose, in XCrysDen, to
display only the translational asymmetric unit. This is one thing I
don't know how to do in Vesta. For a matter of convenience
(laziness, actually :) ), I have set both PRIMVEC and CONVVEC to the
cubic lattice vectors, in the xsf file.
Although I have managed to create a [2 x 2] cell, starting from the
[sqrt(2) x sqrt(2)] using a transformation matrix (easy to deduce),
I have to eliminate some atoms by setting boundaries to the
fractional coordinates and checking that the number of atoms is
actually the expected number. In the case of a [2 x 2] cell, of
course, we have 4 times as many atoms as in the [1 x 1] cell.
However, I would be more reassured if there were an automatic way to
guarantee that the correct number of atoms would be output, instead
of having to make somewhat extensive checks to ensure this.
Another doubt is about symmetrization of atomic coordinates. If I
choose a space group for Vesta to symmetrize the atoms' coordinates,
it ends up creating extra atoms which I can't seem to eliminate - at
least i don't know the correct way to do it. I have tried the option
"remove duplicate atoms", but it didn't work. Has anyone run into
such problems?
If things are not clear, I'll be glad to provide an xsf file to
illustrate the issues.
Best regards,
Marcos