Sciencemadness Discussion Board

Decarboxylation / Methylation at alpha carbon of N-protected amino acid

kitkat - 1-11-2011 at 12:38

My first post, but don't be gentle.

A little background first:

I have N-Phthaloylalanine and N-phthaloylvaline via phthalic anhydride and the corresponding amino acids according to the reference:

Zeng et al (2004) Mild and effective N-phthaloylation of amino acids. Amino Acids 27: 183-186

I have read that TCICA (sometimes abbreviated TCCA) can decarboxylate amino acids to nitriles and have some reasonable evidence that this is indeed the case (evolution of CO2 during a small test reaction with DL-alanine and TCICA) via:

Hiegel et al (2004) Conversion of α‐Amino Acids into Nitriles by Oxidative Decarboxylation with Trichloroisocyanuric Acid. Synthetic Communications 34(19): 3449-3453

I have also read that iodobenzene diacetate/iodine with catalytic iron dust (purportedly creating iron salts in situ) can form an acyl iminium ion by oxidative decarboxylation of phthalimide amino acid derivatives which can then be reacted with a nucleophile (in one pot incidentally) to decarboxylate and then functionalize at the alpha carbon:

Fan et al (2008) A one-pot oxidative decarboxylation–Friedel-Crafts reaction of acyclic a-amino acid derivatives activated by the combination of iodobenzene diacetate/iodine and iron dust. Organic & Biomolecular Chemistry 6: 4615-4621

Now my guess is that my N-phthaloyl amino acids can also be decarboxylated by TCICA yielding an acyl iminium ion which I also guess may be subsequently chlorinated at the a-carbon by the TCICA as well. I have not yet performed this reaction but will probably attempt soon time providing.

However, ultimately what I'd like to do is perform a nucleophilic methylation at the alpha carbon (or nucleophilic substitution of the chlorine), but none of the references I've found merely methylate the alpha carbon, instead they use aryl groups or allylTMS (allyltrimethylsilane), and other nucleophiles.

I believe one could do this using a grignard such as MeMgCl or organometallic reagent, but these are not easy to prepare and not easy to handle.

In further reading, I've found that I might be able to utilize dimsyl sodium (sodium methylsulfinylmethylide) prepared with NaH reacted in DMSO and utilize that as the nucleophilic carbanion, but then I have to get rid of the sulfur group (sulfenic acid). However, I don't have any references as to whether dimsyl sodium would work and I'm concerned about the keto groups on the phthaloyl group anyway.

http://www.gaylordchemical.com/uploads/images/pdfs/literatur...

So my question is:
Any other thoughts as to how to methylate acyl iminium ion or any thoughts as to how to proceed after utilizing dimsyl sodium? Or am i stuck with grignard?

Thanks!

PS. basically nucleophilic substitution with a carbanion, but I don't want to prepare a grignard. Maybe this belongs in "Beginnings".

I've attached a picture.

[Edited on 1-11-2011 by kitkat]

nucleophilicsubstitutionacyliminium.png - 8kB

[Edited on 1-11-2011 by kitkat]

497 - 1-11-2011 at 16:43

I'm going to guess that the decarboxylation will not go foreward. If you look at the mechanism of the amino acid decarboxylation by TCCA you will find that it relies on the formation of a chloramine intermediate, which can't form with an N protected amino acid.

AndersHoveland - 1-11-2011 at 17:20

I could see chlorine reacting under strongly acidic conditions (through the formation of a tautomer) to replace the carboxyl group with a chlorine atom.

In any case, a chlorine atom might not be stable on the same carbon atom that is connected to a nitrogen atom. Those two carbonyl groups are electron-withdrawing, which may help with the prospects of stability, but generally what typically happens is that the chlorine atom ionizes off as a chloride ion, leaving the carbon atom double bonded to the nitrogen atom which now has a positive charge (so it has four bonds total). Then a hydroxyl ion (even if the solution is not alkaline) from the ambient water solution replaces the place where the chlorine atom formerly was. After this, there is typically messy polymerisation, as the molecule would essentially be an aldehyde condensed to a biuriet derivitive, in a similar way in how formaldehyde cannot be definitively condensed with merely one molecule of NH3.

Another route you might consider is that diborane or LiAlH4 can reduce carboxyl groups to alcohols. Then oxidize the alchol to an aldehyde, then reduce it two a methyl group using anhydrous hydrogen iodide. This might not be selective enough for your molecule (since the carbonyl groups might also be reduced).

In conclusion, I just do not see any good direct way to do your reaction. But I am not extremely knowledgeable in mainstream organic chemistry, so have hope. Someone else may be able to provide you with the answer you seek.

Perhaps if you have access to ortho-xylene (1,2-Dimethylbenzene) you could oxidize the methyl groups to carbonyl groups, then react it with SCl2 (PCl3 is more common but it is also extremely toxic) to turn the carboxyl groups into
-COCl groups. Then simply react with the desired amine, in your case 2-amino,3-methyl butane to get the cyclised product in moderate yield, the polymerised byproduct could be filtered off.

[Edited on 2-11-2011 by AndersHoveland]

Thanks!

kitkat - 4-11-2011 at 08:59

497 & Anders,
Thank you so much for your excellent replies, sorry it took me so long to come back and respond. Yes, it appears that there will be difficulties with this so I appreciate your having brought those to light for me. I think I was so focused on the goal that I didn't investigate thoroughly enough how this decarboxylation actually happens. I may have to go with another mechanism then, so the search goes on.

I'll probably give it a test shot anyway this weekend just for fun, but I think you are both right.

By the way if anyone else reads this, I've found that TCCA + H2O + base on standard amino acids (not these protected ones yet - have limited amt of those, so not wasting them) is pretty exothermic and tempermental. Although halloween has passed, this reaction can still jump right out of the flask and bite you :)

Thank you both again, I appreciate your responses very much!

[Edited on 4-11-2011 by kitkat]

497 - 4-11-2011 at 22:57

If you can get sodium borohydride you can reduce the amino acid to the amino alcohol. Deoxygenating that should be reasonably simple.

Sedit - 4-11-2011 at 23:43

Since its an N protected amino acid Methylation may be possible with the correct solvent using the Kolbe electrolysis and Sodium acetate. Large excess of Sodium acetate should be used.

The Dimsy ion performs Methylation after thermal elimination IIRC, I do not know if your compound will hold up under those conditions.

kitkat - 7-11-2011 at 16:32

I have some good news. I have the materials necessary to make iodobenzene diacetate, so I will likely give it a shot with the oxidative decarboxylation / nucleophilic substitution reference I posted about originally. To begin with, I'll utilize one of the same nucleophiles used in that journal article and iteratively tune my process until I can get similar results.
Quote: Originally posted by Sedit  
The Dimsy ion performs Methylation after thermal elimination IIRC, I do not know if your compound will hold up under those conditions.


@Sedit: So the thermal elimination should be solved by using the PhI(OAc)2/I2 system.

My big question now is, instead of utilizing NaH (or KH or lithium amide) can I instead use DMSO/KOH? Or will I get an alkoxy rather than methylation? I've read some articles which seem to say "maybe" - again, having to run out the door - will post references in a couple days when I'm back.

Or am I stuck with having to get NaH as well? I really would prefer not to.

Sedit - 7-11-2011 at 17:46

I sent you a PM,

I honestly feel that Lithium Nitride will yield the Dimsyl ion rapidly as well. The equilibrium between the Dimsyl ion and KOH is way out of favor so I think one would really need to use R-OK alkoxides to have a better effect but this is still out of favor meaning that the Dimsyl ion will need to be removed from the equation right after they react in order to have the reaction proceed to the right. Some literature suggest this is possible by thermal elimination so only experimentation would tell.

kitkat - 8-11-2011 at 05:01

Quote: Originally posted by Sedit  
I think one would really need to use R-OK alkoxides to have a better effect but this is still out of favor meaning that the Dimsyl ion will need to be removed from the equation right after they react in order to have the reaction proceed to the right. Some literature suggest this is possible by thermal elimination so only experimentation would tell.


Ah, for some reason I was completely missing your point with regard to thermal elimination.