[Memy]

Hello,

I have a pretty stupid question....why the cyanide ion attacks the carbonyl group through it's soft-more polarisable carbon end, instead of attacking through it's hard-N end ?
the cabonyl group is an hard electrophile and it should be attacked by hard nucleophiles, therefore by the hard end of the cyanide ion ( the N end)

Thank you very much

   

[movies]

You might not like this answer, but it's all about the molecular orbitals.  If you draw out the HOMO for the cyanide anion you should find that the MO is mostly centered on carbon.

[Memy]

Hello,

so, in other words the lone pair of C is higher in energy than the lone pair of N and therefore the HOMO of the nucleophile is the lone pair of C that attacks the LUMO of the carbonyl compound?

I have however found that  in Sn1 reactions where CN is involved , when the carbocation is produced, the harder  end of the CN ( N end) attacts the carbocation
while in Sn2 reactions the softer end ( the C end) attacks the carbon from which the leaving group will depart.
In the Sn1 case, electrostatic forces are involved and therefore the carbocation is referred as a hard electrophile ( and the attacking N-end of CN is referred as hard).
In the Sn2 case, the C-X ( X is the leaving group) bond is not very polarized and therefore it is a soft electrophile (the attacking C end of CN is thereby soft).

The fact that the carbonyl group is a hard electrophile should mean that the hard end of CN ( the N end) should attack it, or it does not apply here?

thank you very much

[movies]

Your explanation for hard vs. soft ends of the nitrile is right on.  I'm not sure about defining carbonyl groups as hard electrophiles, however.  They are considerably polarized, but when I think of a real hard electrophile I think about something that is practically ionic, and a carbonyl is defintely covalent.

The definition from this book is "Hard electrophiles have a high-energy LUMO and usually have a positive charge.  Soft electrophiles have a low-energy LUMO but do not necessarily have a positive charge."

[Memy]

So, the LUMO of the carbonyl group ( antibonding Pi orbital) is lowered by the presence of the oxygen atom and therefore it has a quite low LUMO and partial positive character on carbon and can therefore be attacked by both soft and hard electrophiles?

[movies]

Well, there is only one electrophilic position on a carbonyl group, so whatever the nucleophile is, it will have to attack the carbon.  I think the LUMO of a carbonyl is relatively low energy, probably because of the electronegativity of the oxygen.

[Memy]

So, in conclusion, the fact that CN attacks the carbonyl group by its softer C end and not its harder N end does not apply here ( as the carbonyl group is not really a hard electrophile) : the carbon lone pairs are just higher in energy since carbon is less electronegative than nitrogen and therefore constitute the HOMO of the cyanide ion...?  

[movies]

Yes.

In general, nucleophiles and electrophiles will come together in such a way as to maximize the matching of the energies of the frontier MOs; that is to say, with the HOMO of the nucleophile attacking to LUMO of the electrophile.  Hard electrophiles are really the exception to the rule since they present a stronger electrostatic attraction between the two species.  Therefore, the electrophile can in some cases be more attracted to electrons that are not in the HOMO of the nucleophile, but reside on an atom with a greater electron density in general.