[Wien] Convergence problem in mbj potential calculation

[Laurence Marks]

To add one small thing; you can remove all the Li and it will still "match"
the XRD. If it is powder data, you can probably remove all the O and it
will still "match".

---
Professor Laurence Marks
"Research is to see what everybody else has seen, and to think what nobody
else has thought", Albert Szent-Gyorgi
http://www.numis.northwestern.edu
Corrosion in 4D http://MURI4D.numis.northwestern.edu
Partner of the CFW 100% gender equity project, www.cfw.org/100-percent
Co-Editor, Acta Cryst A


On Jun 15, 2017 05:52, "pieper" <pieper at ifp.tuwien.ac.at> wrote:

You keep getting back to this match between the (presumably powder-)xrd
you calculated from your .struct file and the experimental one. This
does NOT imply that your structural model describes the physical
situation correctly! With your - in Gerhards words - 'interesting
approach' you should be able to produce any number of different .struct
files, and calculating their powder-xrd find a similar 'match with the
experimental one'.

Did you calculate the xrd with any of the alternative distributions of
Li, Ni and Ti on their sites in your unit cell? Did you try the same
with the many, many other distributions you can generate in larger
supercells? I am prettty sure you will find something that matches your
xrd data. It will make as much sense as what you have now ...

Be aware of the fact that structural models of Wien2k have symmetry
properties  different from your physical situation: in your model the
metal atoms occupy lattice sites regularly, breaking certain symmetries
of the underlying lattice. The actual distribution in your case is
(apparently more or less) random, which overall breaks no symmetry, but
locally breaks every symmetry somewhere.

If the distribution of Li, Ni, and Ti  really is random on one shared
site I know of only one (expensive) way to proceed: try to reduce the
influence of short range periodicity in your models by increasing the
size of the supercell and putting Li, Ni, and Ti at random positions. Do
structural relaxations, look what changes in your (GGA-) calculations,
and keep fingers crossed that things you are interested in converge for
supercells of a size you still can handle.



---
Dr. Martin Pieper
Karl-Franzens University
Institute of Physics
Universitätsplatz 5
A-8010 Graz
Austria
Tel.: +43-(0)316-380-8564
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20170615/e68d76ca/attachment.html>