Sciencemadness Discussion Board

N-demethylation of 3° amines

indiangold - 24-6-2011 at 03:35

Hi all ! I am working on this pretty little problem. I have a 3° amine having a benzyl group, a 2° alkyl group and a methyl. I want to remove selectively only the methyl. Any suggestions?

IPN - 24-6-2011 at 06:37

A simple google search resulted in a lot of hits for selective demethylation of amines. I attached one of the articles I came across. I recommend you run some searches too. You can always request articles in the references subforum if you can't access them yourself. :)


A General, Selective, High-Yield N-Demethylation Procedure for Tertiary Amines by Solid Reagents in a Convenient Column Chromatography-like Setup
Org. Lett., 2004, 6 (4), pp 541–544
DOI: 10.1021/ol036319g
Publication Date (Web): January 23, 2004

Attachment: ol036319g.pdf (87kB)
This file has been downloaded 3666 times

To IPN

indiangold - 24-6-2011 at 19:52

Thanks IPN ! I have this article. Basically they use sodium percarbonate to make the N-oxide of tert. amine and then use sodium salt of dichlorocyanuric acid as a Polonowski reagent. I could not reproduce the experiment as I did not have access to this sodium dichlorocyanurate. But thanks again for such prompt response, IPN !

Jor - 25-6-2011 at 03:59

The sodium dichlorocyanurate can be bought in many places as a swimming pool chlorinater. Otherwise you might be able to convert trichloroisocyanuric acid (available almost everywhere as pool chlorinator) to the salt by treatment with base?

Nicodem - 25-6-2011 at 06:42

They use "sodium salt of 4,6-dichloro-2-hydroxy-[1,3,5]triazine", which is related to cyanuric chloride in that it is an acid chloride (the sodium salt of monohydrolysed cyanuric chloride, CAS: 2736-18-7). It has nothing to do chemical property-wise with the sodium dichlorocyanurate (aka "sodium salt of dichlorocyanuric acid", CAS: 2893-78-9) which is not an acid chloride, but an oxidant instead. I don't understand why people avoid checking the structures and draw conclusions based only to the similarities in chemical name spelling. Organic chemistry is about molecular structures, not words.

The Polonowski reaction proceeds with nearly any acid anhydride or acid chloride/base as the activator (just like the mechanistically related Pummerer rearrangement). In the article they use that unusual reagent just because it can be conveniently fixed in the column for a two step flow reaction (it and its products cyanuric acid, its salts, KCl, KHCO3 and NaCl are insoluble in the solvent used). Paradoxically, the sodium dichloroisocyanurate hydrate could potentially be used for the N-oxidation step instead of the sodium percarbonate. This would produce an irony of chemical names as you would have a "sodium dichlorocyanurate / sodium salt of cyanuric dichloride" flow reaction column.

Thus, if the Polonowski reaction is to be applied, which is the best choice as SN2 based dealkylations of benzylmethylalkylamines cause debenzylation in preference to demethylation, then even something as trivial as acetic acid anhydride can be used.
In any case the "sodium salt of 4,6-dichloro-2-hydroxy-[1,3,5]triazine" has limited commercial availability, so this flow trough column method is not very feasible.

The article authors have an unusually sharp scientific humour when describing a reaction runaway:
Quote:
Larger amounts of inducing agent, more than 1.5% relative to the starting amine, caused the reaction to become uncontrollable (which manifested in an esthetically rewarding, fountain-like discharge of the column filling, but was otherwise of little value).

@ nicodem

indiangold - 27-6-2011 at 04:08

dear nicodem, thanks for correction. i really meant the sodium salt of the monohydrolyzed tri-acid chloride of 2,4,6-trihydroxy-1,3,5-triazine, which is cynuric acid. i am fully aware of chemical nomenclature. and when i said 'sodium salt of dichlorocyanuric acid' i did not mean N-chloro compounds, which are oxidants. by the way, i could not control my laugh at the most elegant description 'esthetically rewarding, fountain-like discharge' !!! because i have been myself creator of many a 'not-so-esthetically-rewarding' discharges that have reached the ceiling of my lab !!!
by the way, i was really looking for a method for selective removal of a methyl group in a tert. amine having a methyl, a benzyl and a 2° alkyl group. and that too in a reaction that is amenable to be carried out in a stirred vessel, not a chromatographic-looking column......... For a Polonowski or a von Braun type reaction, the order of the ease of removal of an alkyl group on N atom roughly follows the decreasing order of stability of carbocations, which in my case is unfavorable to methyl

Nicodem - 27-6-2011 at 06:46

Quote: Originally posted by indiangold  
For a Polonowski or a von Braun type reaction, the order of the ease of removal of an alkyl group on N atom roughly follows the decreasing order of stability of carbocations, which in my case is unfavorable to methyl

I don't think you fully understood what I mentioned in regard to selectivity. In a von Braun reaction and its variants, the dealkylation occurs via an SN2 substitution. The benzylation is somewhat more favourable in SN2 substitutions when compared to the methylation, thus you are unlikely to get a useful yield for demethylation. Though harder nucleophiles will tend to give a better selectivity for the demethylation, the nature of the von Braun reaction gives only few choices for the nucleophile. For example, the fluoride and the carboxylate are the hardest potentially still applicable, but I never saw any acyl fluoride or acid anhydride being used for this purpose and I doubt they would work under tolerable conditions. In any case, no matter how hard the nucleophile, you would still get a mixture of debenzylated and demethylated products.

However, like the above article demonstrates, the Polonowski reaction has the opposite selectivity. It is the methyl group that gets oxidized preferentially. Just check their example of dibenzylmethylamine demethylation. They get a 89% yield of dibenzylamine!

PS: If the selectivity of the von Braun type reactions would be based on carbocation stability of the dealkylated group, then demethylation would not be possible at all! Yet only few alkyl groups are more easily removed in this reaction when compared to the methyl. Unfortunately for you, the benzyl is one of them.

@nicodem

indiangold - 27-6-2011 at 20:02

thanks nicodem for information on the intricacies of von Braun reaction. I was not really aware of this. So no other way to remove methyl in favor of benzyl? actually, i have a 2° amine having a methyl and a 2° alkyl group on it. i want to get 1° amine devoid of that methyl group from this 2° amine. that's why all this exercise of attaching one more benzyl and trying to remove the methyl first and then later hoping remove the benzyl by Pd-C/H2 hydrogenolysis...........
by the way, how can i insert a chemdraw drawing from my computer into a message box? i wanted to put the actual reaction here.

Nicodem - 28-6-2011 at 10:54

Quote: Originally posted by indiangold  
actually, i have a 2° amine having a methyl and a 2° alkyl group on it. i want to get 1° amine devoid of that methyl group from this 2° amine. that's why all this exercise of attaching one more benzyl and trying to remove the methyl first and then later hoping remove the benzyl by Pd-C/H2 hydrogenolysis...........

Why didn't you say so in the first place?
Anyway, secondary methylamines can be demethylated by neutral nucleophiles in their protonated form (basic nucleophiles are of no good due to the obvious reason). A very efficient method is the reaction of the amine with benzeneselenol [1]. The selenol is acidic enough to protonate the amines to their corresponding ammonium benzeneselenolate. Though selenolates are very strong nucleophiles, the ammonium salts are generally extremely lousy electrophiles, thus the reaction still requires quite some heating. Of course, not everybody enjoys working with selenols, but at least it seems to work like a charm.

Depending on the rest of the molecule, other conditions can be used. Robust amine substrates can be selectively N-demethylated by heating in slight excess of hydriodic acid until dry and the temperature reaches about 200-230 °C [2]. Needless to say, only few substrates are suitable. The iodide is not the only neutral strong nucleophile. Phosphines are also quite nucleophilic and neutral, therefore you might get N-demethylation by heating an ammonium salt (e.g., a hydrohalide) of your substrate with triphenylphosphine (or preferably tributylphosphine if courageous enough) in DMSO, sulfolane, PEG or other suitable high bp solvent at 150-200 °C. But I have not searched the literature on the topic and don't remember reading much else about secondary amine dealkylations.

If none of these SN2 based dealkylations works, then perhaps some oxidative dealkylation might, but I would expect most would cause the dealkylation of the secondary alkyl. You should review the literature and see what fits best to your substrate.
Quote:
by the way, how can i insert a chemdraw drawing from my computer into a message box? i wanted to put the actual reaction here.

You can either use the attachment function or "Insert an image" function. In any case, use the GIF format so that the scheme is visible in the browser. For the "Insert an image" function you need to have the file uploaded somewhere in order to give the URL (the forum scipics FTP is generally used, UTFSE for instructions).

[1] J. Org. Chem., 1979, 44, 3148–3151. DOI: 10.1021/jo01332a012

[2] http://www.chempensoftware.com/reactions/RXN195.htm (I never saw examples of this being used in a preparative reaction, but I see no reason for not doing so)

indiangold - 28-6-2011 at 19:58

thanks a ton nicodem sir ! i will try the selenol and also triphenylphosphine route and report the finding here shortly. regards - indiangold