[Darkstar]

How do you replace the methyl group of toluene with something else (specifically NH2). It's so stable I can't remember a method to tear it off.

[movies]

Are you trying to make aniline or benzyl amine?

For benzyl amine you could oxidize with KMnO₄ to give benzoic acid, then make the amide with ammonia, then reduce to benzyl amine with LiAlH₄.

I'm not so sure about making aniline from toluene though.  You might be able to oxidize to benzoic acid with KMnO₄, then make the acid chloride with SOCl₂, then add sodium azide and heat the crap out of it to do a Curtius rearrangement.  That'd probably work.  There might be a more direct method though.

[AWK]

Both, Curtius and Hoffman rearrangements work in this case with yields of 60-80 %

[Darkstar]

Yes, I mean aniline. If Curtius and Hoffmann do work with arenes, that's good.

[Mitch]

I do not see how you could take that methyl off toluene

[movies]

Well, once you have the azidoamide of benzoic acid you do the Curtius rearrangement.  In the mechanism you lose a molecule of nitrogen and get a nitrene (like a carbene, but with N).  Then the aryl group migrates and you can push the electrons to make an isocyanate intermediate (so, Ar-N=C=O).  The isocyanate is then hydrolyzed upon work-up and you lose the carbon from the methyl group of the toluene as CO₂.

You can also draw a polar mechanism where the aryl group migrates and displaces molecular nitrogen, but I like the nitrene mechanism better.

BTW, this is almost exactly like an Arndt-Eistert homologation, but with nitrogen instead of carbon.

[anhuiw]

CH3---COOH----CONH2-----NH2