single point DFT covengercy problem

[gjgj]

Hi

I am currently using TURBOMOLE60 program and trying to get the ground state of a donor-acceptor dimmer.

Level of calculation : test single point DFT with 6-31G* basis set.

The default guess (Ext.Huck) ends up by a hole containing electronic configuration. A second calculation with altered occupation in the guess ends up with another double excited state.

Other occupation guesses work even worst.

The same calculations were repeated with other levels :

-          the results are correct (real ground state)  at    HF, B3LYP, PBE0, TPSSh  levels, (only with altered order of the EH guess)

-          the results are wrong (excited state) at  B97-D, PBE, TPSS levels (independently of the initial guess)

 Did I miss any particular parameter or procedure which make these calculations work ?

Thank you in advance for your help

[Arnim]

Hello,

you could try to use femi to find the correct occupation. (in the scf submenu of the general menu using define)

If this does not work, you might have to check, if your system is a multi-reference case. Then you have a problem, of course.

Best
Arnim