[Wien] rare earth + Mo system

Laurence Marks L-marks at northwestern.edu
Thu Oct 27 14:39:49 CEST 2011


You have to understand that the scf cycle is not just some black-box process
that you do before you can get any results, rather it tells you some
important things about the physics/chemistry of the problem you are trying
to model. Beyond this you need to think about what you are doing otherwise
what you get will be GIGO (Garbage In, Garbage Out).

1) For rare earths you need to consider carefully how you are treating the f
electrons, either putting them into the core or (better) using an on-site
hybrid or Hubbard U. See the FAQ.

2) Changing the B site changes everything in the chemistry. From what you
say it appears that with Nb/Ta you have an insulator, and when you
substitute in Mo/Ru you have a metal. Metals require a better BZ sampling.
Beyond this, you may be producing an unphysical system which is
energetically unstable, i.e. GIGO. The scf iteratations will converge poorly
for an unphysical system.

3) For many (perhaps all) transition-metal oxides the conventional
functionals (LDA/PBE/WC etc) give results with too much covalent bonding,
not enough ionicity of the metal and oxygen. A Hubbard U or the on-site
hybrid helps, but it does not fully solve everything.

4) Many, many things have changed since Wien05. Wien10 (or better the latest
release) is more accurate, converges more accurately (the older version
could appear to have converged when it had not, what is called a "trap") and
more.

2011/10/27 tripurari sinha <sinha_tp at yahoo.com>

> Dear Prof. Blaha,
>
> I have tried to run the scf cycle for spin-polarized case of a double
> perovskite oxides contains rare earth and Mo5+ (Molybdenum) in B-site using
> Wien 10.
>
> The convergence criteria are: energy convergence 0.0001 Ry and charge 0.001
> e.
>
> After running 40 cycles the results shows that energy is not converged in
> scf.
>
> But if I run the similar system with same crystal structure without
> spin-polarization scf cycle is converged.
>
> Also previously I run the same system in wien05 with spin-polarization and
> it successfully converged.
>
> What is wrong when I am using  wien 10.
>
> Similar problem arises if I replaced Mo5+ by Ru5+ i.e., for magnetic
> materials.
>
> But if Mo is replaced by Nb5+ or Ta5+ (non-magnetic)  it works  nicely.
>
> I am waiting for your reply. Thanking you,
>
> With regards
> T. P. Sinha
>
>
> From: Prof. T. P. Sinha
> Department of Physics
> Bose Institute
> 93/1, Acharya Prafulla Chandra Road
> Kolkata - 700009, India
> Cell No. 09830159422
>
>
>
>
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>
>


-- 
Laurence Marks
Department of Materials Science and Engineering
Northwestern University
www.numis.northwestern.edu
"Research is to see what everybody else has seen, and to think what nobody
else has thought"
Albert Szent-Gyorgi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20111027/fe6877f1/attachment.htm>


More information about the Wien mailing list