Sciencemadness Discussion Board

Demethylation mechanism

peach - 4-12-2008 at 09:33

Does anyone have an understanding of demethylation?

I've been able to find mechanisms for reactions I'm interested in, although none for demethylation.

Those demethylations that I've seen seem to be using a Friedel Crafts mechanism, more commonly used for adding methyls. For instance, a lot of them seem to use large amounts of Lewis Acid and a weak nucleophile. The Lewis acid is also used in FC, but as a catalyst.

I could do with knowing how demethylating agents respond to the presence of primary / secondary C=C bonds, ketones and animes, and how generally sensitive the work is to moisture and atmospheric oxygen.

Outer - 4-12-2008 at 11:30

What kind of demethylation do you mean?
If you mean demethylation of aromatic methyl (like in toluene), this reaction is impossible in presense of other complex groups (COCH3, NH2 etc.) in molecule.

Other kind of demethylation (OMe, COOMe etc.) is possible by using of different agents.

Klute - 4-12-2008 at 11:53

Are you talking of O-demethylation?

Well, basicly the doublet of the oxygen is protonated or complexed with a Lewis acid, weakening the O-CH3 bond, then the added nucleophile attacks the methyl group, which sends back an electron to the oxygen, forming a new doublet.

For example, if using HBr, H+ protonates the oxygen, and Br- attacks the methyl group, forming MeBr and R-OH.

If using AlX3 and NBu4X, AlX3 complex the doublet, forming RO(+)(Me)-Al(-)X3, and the added X- attacks the methyl group, giving MeX and RO-AlX2. The demethylated product will then have to be hydrolyzed to obtain the phenol.

Nothing too complicated, a simple SN2 if I am not mistaken.

Although during a FC the lewis acid has a catalytic role, it is generally used in stoechiometric amounts, even in excess.


The compatibility of the demethylation system with other functions depends: HX and a double bond will surely not be happy together, lewis acids can cause side reactions with carbonyls, and amines will have to be under salt form, or even protected, as you will be generating MeX, which will methylate a free amine immediatly.

chemrox - 4-12-2008 at 21:10

Another O-demethylation conceived by Rappoport in 1951 uses pyridine HCl. See the Rhodium archive on it because it discusses some other methods.

Nicodem - 5-12-2008 at 01:01

Like Klute said, the O-demethylations are just simple S<sub>N</sub>2 reactions. The nucleophile attacks the oxonium electrophile (that's why acids like BBr3, HBr, H2SO4, AlCl3, LiCl, pyridine hydrochloride, etc., are needed). This is why sometimes a soluble bromide or iodide salt is added. In some cases, in the presence of very strong nucleophiles like the thiophenoxide, no acid is required and the reaction can be done under basic conditions (it is still an S<sub>N</sub>2 mechanism, but on the ether/ester directly instead of the highly electrophilic oxonium ion). A typical example of such a reaction is the Krapcho decarbalkoxylation which originally used KCN for the nucleophile.

N-Demethylations are essenciall identical in mechanism. A quaternary amine needs to be formed in order for the S<sub>N</sub>2 reaction to occur. This is generally done with reagents which form amides which can be easily regenerated to amines while at the same time the reaction forms a strong enough nucleophile. For example, cyanogen bromide (BrCN) is commonly used to equilibrate with R<sub>2</sub>MeNCN<sup>+</sup> + Br<sup>+</sup> which on heating decompose to MeBr an R<sub>2</sub>NCN from which the secondary amine is relatively easily generated by hydrolysis. This is generally know as the von Braun reaction.

The aromatic demethylations involve a retro-Friedel-Crafts mechanism (a completely identical mechanism to the one in Friedel-Crafts reactions, but in reverse). These are usually called Friedel-Crafts transalkylations or Friedel-Crafts disproportionation (depending if the alkyl group is transferred to another aromatic system or to another substrate molecule). But the ease of dealkylation drops in this direction: t-Bu > i-Pr > Et >> Me (the demethylation would thus be very hard).

Klute - 5-12-2008 at 01:44

Thank you for the details, Nicodem!

I was unaware of the N-demethylation via CNBr, good to know.

peach - 6-12-2008 at 13:22

Thanks for the replies everyone, that's an excellent help! Things are making more sense, and I have a mechanism to learn.

This is indeed an O-demethylation of a methyl ether, coming off a phenyl. with a tail that is ortho to the ether.

Polymerization has been an issue we think due to a lone double bond as a primary on the tail it's self - as klute suggests, the double bond seems to be attacked rather than the methyl of interest, with it being so far from the centre of action (well, aromatic stability).

I have seen some papers with demethylation of virtually identical substrates where the tail primary double bond is replaced with something like a carboxylic acid. Is it likely that oxidizing the tail double bond to a ketone would offer some protection for the harsh demethylation conditions Nicoderm mentions? We can also reduce this to an anime. If we have a tail double bond on a phenyl, would it be more stable if it were a keytone or an anime? And anime would loose the double bond, but I'm not up enough on my organic chemistry to know if the remaining single bond would be more resilient to attack than the double oxygen bond.

Klute - 6-12-2008 at 13:50

I don't think you can easily demethylate your o-methoxy amine without complications, but I suggest you check out the litt to see if anything in that area has been done.

In any case, regarding eugenol:

Quote:

Attempts to demethylate eugenol to obtain 4-allylcatechol in one step were made. The demethylating agents used were hydriodic acid, pyridine hydrochloride, anhydrous aluminium chloride, sulphuric acid and potassium hydroxide. However, in each case a polymerized product was obtained.


Sethi Rao & Subba Rao
Ind. J. Chem. 2, 323 (1964)


So options left would be AlI3/PTC, LiCl in DMF, etc