CCL: Negative Ionization potential

 Hello Morgan and All,
There's nothing intrinsically odd with negative IP's. As commented, it just means that the species is more stable when it loses an electron. Add enough electrons to any molecule and it will at some point have a negative IP...
Sometimes the higher-energy anionic states can be "real" in the sense that they have a finite life-time before the electron escapes. The simplest example would perhaps be the (or rather, a) helium anion.
When doing DFT calculations, many times the higher energy of the anions will be an artifact, though, due to self-interaction error and wrong asymptotic behaviour of the xc potential; you probably have positive occupied orbital energies for your molecules.
You could try to do LC-corrected DFT (not w/ Turbomole, though). Or then a recently proposed methodology where the density (orbitals) of the molecules/anions are computed at HF level, and then the DFT energy is evaluated using the HF orbitals (non-SCF). I have no personal experience with how that works, but the idea seems sound enough for the pragmatic: 
 [1] Kim, Sim, Burke, "Avoiding unbound anions in density functional
     calculations", JCP 134 (2011) 171103.
     http://dx.doi.org/10.1063/1.3590364
 For a good general overview, see:
 [2] Simon, "Molecular anions", JPCA 112 (2008) 6401.
     http://dx.doi.org/10.1021/jp711490b
 Good luck,
     Mikael J.
     http://www.iki.fi/~mpjohans/
 On Tue, 15 Apr 2014, Durand Morgan durand_morgan::yahoo.fr wrote:
 
 Dear All,
  
 Thank you very much for your replies.
 Here is the protocole I used for both M and M-1e- forms of the molecule.
  
 Conformational search (Based on Balloon + BP-86 SVP +BP-86 TZVP)
 BP-86 TZVP frequency analysis to check for a true minimum
 PBE0 def2-TZVP Geometry optimization
 PBE0 def2-TZVPPD Single point energy
Based on these calculations, I compute the adiabatic IP based on the energy difference corrected with the enthalpic and entropic correction.
As suggested, I checked the initial geometry of the molecules leading to negative IP. All of them have a carboxylate in alpha of an amine group. For the lowest energy M+ form, the amine group seems planar or nearly planar. I guess it would be suspect for the M form but I don't know in the case of the M+ form. 
  
 Any ideas ?
  
 Best Regards,
  
 Morgan