ABINIT Discussion Forums

Hi all,
Using the aim processing tool to get the bader charges on the 12 atoms of a benzen molecule, I found that the 6 carbons where charged -0.25 while the 6 hydrogens where charged +0.12. I wonder how this is possible : there would be a net charge of almost 0.8 e- on the molecule.
Any ideas ?

Thanks,

Emmanuel

Ok, since there are no answers, I'll do the follow up myself.
I guess the problem comes from the fact that bader surfaces are computed one at a time, and independently (1 independent aim calculation per atom). Using another program doing them all at once (1), it seems that charges on C atoms are difficult to evaluate in the benzen cycle (strong delocalisation of electrons), and greatly vary from one to the other. Yet the global computation ensures that the sum is zero. So I guess the charge on each C has to be taken as the opposite of the charge on each H (which is much easier to compute).

Any comment is welcome.

Emmanuel

(1) : See http://theory.cm.utexas.edu/vtsttools/bader/ . It uses as imput Gaussian "cube" format, which can be obtained from _DEN files by cut3d. Be carefull though, one has to output the total electron density including core charges (prtden 3). One can also visualize the Bader surfaces obtained with v_sim, which is great to track any non-physical behaviour.

Hello Emmanuel,

this sounds reasonable. I though AIM worked with all atoms at once (provided all the core density files) but perhaps you are right. Certainly a molecule (in paricular with delocalized electrons) is the worst case scenario, where the surfaces have to extend to the edge of the box to collect all of the charge correctly...

Cheers

Matthieu