Hi List,
I want to calculate ionization energies of some systems. But I'm having
problem with SCF convergence of (n-1) electrons system...( n = no. of
electrons in un-ionized species). I'm stating this problem in a bit
detail (for some reason I dont want to use spin polarised KS equations
for n-1 system):
1.) I tried all available convergence tricks but I cant converge the
density to correct electronic configuration. I f I dont provide
occupation , then scf does converge but to incorrect and broken-aufbau
cofiguration.
Electron smearing also helps but resulting configuration is of no
physical sense. And also restart from thus converged scf does not help
either.
2.) Moreover I tried 'fractional ionization' specifying CHARGE key with
fractional no. of charge. Correspondingly specifying occupations so that
additional 'fraction of electron' goes to orbital one up to HOMO.
In this case SCF did converge to specified configuration. But from this
one can not get ionization potential. I could get a stable convergence
for 'fractionally ionized species' for upto (n-0.4) electrons . I hoped
that with successive restarts I'll reach upto (n-1) electrons but FOMO
(fractionally occupied orbital) does'nt seem to accomodate an electron
larger than 0.6, while retaining an energy one upto HOMO. Beyond this it
only converges to broken-aufbau configuration.
3.) Interestingly (n+1) electron system, despite having more negative
charge does'nt have any problem giving SOMO (singly occupied orbital) in
correct aufbau order.
4.) My species has a transition metal and (n) electron system typically
has one negative charge.
5.) Annoyingly a ristricted-open-shell run with traditional GTO's and
hybrid functional in gaussian gave a converged (n-1) SCF.

Any clue?
Thanks in advance

-- 
Your's truly
Masroor Ahmad
-------------------------------------------------|
Institute f¨¹r Anorganische und Analytische Chemie|
Friedrich-Schiller Universita"t Jena              |
Carl-Zeiss Promenade 10                          |
D-07743 Jena (Germany)                           |
-------------------------------------------------|
Tel: +49(0)3641-644028                           |
Fax: +49(0)3641-644036                           | 
-------------------------------------------------|