[CPMD-list] Energy and lattice constant

[Ari P Seitsonen]

Dear Oleksandr,

  Since apparently no one else has replied anything I'll give a small 
contribution...

  I didn't quite get how large the decrease of the energy is, and how 
much smaller your lattice constant was than in experiments. However, how 
many k points did you use (Gamma only?), and which value did you use for 
the broadening of the occupation numbers (= electronic temperature)? If 
you want you can send your results, the input and a sample output file 
and someone will try to look at the problem.

    Greetings from Zurich,

       apsi

-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
 Ari Paavo Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/
 Tel +41 1 635 44 97 / Fax +41 1 635 68 38 / GSM +41 79 719 09 35      
 Anschrift: Physikalisch Chemisches Institut (PCI), Universität Zürich (UniZh)
 Indirizzo: Winterthurerstraße 190, CH-8057 Zürich
 Address:   Schweiz / Svizzera / Suisse / Svizra / Switzerland

On Fri, 16 Apr 2004, Oleksandr Shepelyev wrote:

> Dear CPMD Users!
> 
> I have successfully implemented the PW91 (GGA) density functional into
> CPMD code. Results for single Fe atom are close to normconserving
> Goedecker PP. Also I found the evident energy minimum at interatomic
> distance close to experimental lattice constant. But calculations for
> 54 and 128 atoms show that decreasing of the lattice constant leads to
> decreasing of energy and the minimum of energy is not evident. For 54
> atoms after passing the point which is close to lattice constant some
> increasing of energy was found, but at the next point the energy starts
> to fall down.
> Does this mean that my lattice is unstable? Or as I found in literature
> I must study my lattice using the experimental lattice constant?
> 
> Also the Total Integrated Spin Density looks too high (33 for 54 atoms,
> 7.6 for 128 atoms but 0. for 1 atom) in my results.
> 
> Sincerely yours,
> Alex Shepelyev
> _______________________________________________
> CPMD-list mailing list
> CPMD-list at cpmd.org
> http://www.cpmd.org/mailman/listinfo/cpmd-list
>