Dear Dr. Wang,

ADF prefers symmetric solutions, which can mean that, for example, p_x
and p_y are both half-filled.
 From the symmetry point of view, this is the optimal solution.

Other DFT codes will usually give occupation numbers 0 and 1 in such a
case, arbitrarily occupying
either p_x or p_y. With the current xc functionals, such a solution may
be lower in energy, and in that
sense energetically preferable to the symmetric solution.

You can force ADF to use also the occupation numbers 0 and 1 by using
the occupations keyword
and impose a lower symmetry. This will show you the energetic effect.

I would say that different scientists may have different opinions on
which solution is "best".

The message below occurs very frequently in atomic create runs. In that
case, there is no problem,
as this just provides some reference energy to which to compare the
various molecular energies.
Only if one wants to compare to atomization energies in the literature,
this issue may need to be
considered.

Best regards,
Stan van Gisbergen

On Aug 23, 2005, at 2:42 AM, Yong Wang ($B2&1J(B) wrote:

> Dear ADFers,
>
> My model system is a FeO(porphyrin)(SH)+ethanol system on the ground
> state. when I optimized the Transition State, I got following warning
> messege, could anyone kindly tell me whether the computational result
> is ok? if not how to overcome this?
>
> Any response will be greatly appreciated!
>
> -----------------------------------------------------------------------
> -------------------------------------
> <Jan15-2004> <03:13:13> SCF converged
> <Jan15-2004> <03:14:09> 25 ErrMat 0.00000123 MaxEl -0.00000013
> <Jan15-2004> <03:14:43> Solutions with partially occupied orbitals
> may not be
> <Jan15-2004> <03:14:43> lowest in energy. You might consider
> lowering the
> <Jan15-2004> <03:14:43> symmetry in the input and explicitly
> specifying integer
> <Jan15-2004> <03:14:43> occupations. In that case always check that
> you obtain
> <Jan15-2004> <03:14:43> an aufbau solution.
> <Jan15-2004> <03:14:45> >>>> COREPS
> <Jan15-2004> <03:15:05> >>>> TOTEN
> <Jan15-2004> <03:21:37> >>>> POPAN
> <Jan15-2004> <03:21:38> >>>> DEBYE
> <Jan15-2004> <03:21:41> >>>> AMETS
> <Jan15-2004> <03:22:14> >>>> POPUL
> <Jan15-2004> <03:22:22> Bond Energy LDA -12.75667154 a.u.
> <Jan15-2004> <03:22:22> Bond Energy LDA -347.12683608 eV
> <Jan15-2004> <03:22:22> + GGA-X -10.99600390 a.u.
> <Jan15-2004> <03:22:22> + GGA-X -299.21661242 eV
> <Jan15-2004> <03:22:22> + GGA-XC -11.79178266 a.u.
> <Jan15-2004> <03:22:22> + GGA-XC -320.87086315 eV
> <Jan15-2004> <03:22:22> NORMAL TERMINATION
> <Jan15-2004> <03:22:23> END
> NORMAL TERMINATION
> -----------------------------------------------------------------------
> ------------------------------------------
>
> Yours, truly
> Yong Wang
> 2005-08-23
> -----------------------------------------------------------------------
> --------------------
> Yong Wang
> Ph.D. student of Prof. Keli Han
> Center for Theoretical and Computational Chemistry
> State Key Laboratory of Atomic and Molecular Reaction Dynamics,
> Dalian Institute of Chemical Physics, Chinese Academy of Sciences
> Dalian 116023, P. R. China
>
> Email: wangyong_at_dicp.ac.cn
> Webpage:
> http://www.beam.dicp.ac.cn/groupmem/personal/ywang/ywang.htm
> -----------------------------------------------------------------------
> --------------------
>
>
>
Dr. S.J.A. van Gisbergen
Scientific Computing & Modelling NV
Theoretical Chemistry, Vrije Universiteit
De Boelelaan 1083
1081 HV Amsterdam
The Netherlands
vangisbergen_at_scm.com
http://www.scm.com

Please note NEW FAX AND TELEPHONE NUMBERS:
T: +31-20-5987626
F: +31-20-5987629
Received on 2005-08-23 10:39:30