Re: [Cdk-devel] Re: aromaticity detection problem (?)

[Kai H. <Kai.Hartmann@Uni-Koeln.De>]

Hi Irilenia,

the topic "aromaticity" is a very complicated one. For simplicity, we 
chose the Hueckel rule in the cdk code. In the case of quinone it works 
quite well, as quinone behaves - in the chemical sense - mostly like a 
non-aromate. It should be mentioned however that not all Hueckel rules 
are checked by the code (planarity).

By the way, as it is now only rings with C or N may be aromatic. If I 
remember my time as a student of chemistry correctly, the aromatic 
character of ring systems with O/P/S and other atoms have less aromatic 
than non-aromatic properties. But the question remains: where to draw 
the border line.

To totally confuse any human being, there are quite a few other 
aromaticity concepts :)

Hope that was more of help than it was confusing you,

Kai


> Anyway, I wrote the original code for the aromaticity detection and I 
> remember that originally it detected quinone as aromatic, because it 
> just counted the sp2-centers in the ring. And I also remember that we 
> did quite an extensive rewrite because we agreed that detecting quinone 
> as aromatic is wrong.
> 
> Here is a citation from one of my chemistry books:
> 
> "Quinones can be regarded as oxidation products of diphenols. The loss 
> of 2 H atoms also causes the loss of 2 pi-electrons in the benzene core 
> and thus the loss of the aromatic character."
> 
> Cheers,
> 
> Chris
> 


-- 
Kai Hartmann
CUBIC - Cologne University BioInformatics Center
Institute for Biochemistry
Zu"lpicher Strasse 47
50674 Koeln
Phone: +49-221-470-7719
Fax: +49-221-470-7786