[Babcock_Hall]

Hello Everyone,

We are trying to make a vinylsulfone with two carboxylic acids for enzyme inhibition studies (compound 2 to 3 in attached scheme).  We can tolerate a final product that has a slow decomposition in water near neutrality in such studies.  While researching deprotection of methyl esters in the third edition of Philip Kocienski's book Protecting Groups, I notinced Schemes 6.9 and 6.10 (p. 399).  In two cases in which there was a single sp³ carbon between the carboxylic acid and the sulfone, there was decarboxylation.  In one case tetramethylammonium acetate n HMPA at 100 degrees did this chemistry, and in the other cesium benzenethiolate in DMF at 85 °C did this chemistry.  It is curious that the conditions are not particularly acidic.  It is also curious that the solvents are dipolar aprotic (I vaguely recall some early bioorganic studies on a derivative of thiamine pyrophosphate, in which decarboxylation was accelerated by solvent, which I can look up if anyone is interested).  In looking over this section of the book, I see other examples of decarboxylations that follow nucleophilic attack on the methyl group of a carboxylate ester in a beta-relationship to a carbonyl group.

The second line of evidence that suggests that decarboxylation might be a problem is that we saw some decarboxylation in a different attempted synthesis of the final compound (1 to 3), one that yielded a mixture of products.  We have an intervening carbon that is sp² hybridized, as opposed to the examples in the book on protecting groups.  We don't have a choice of deprotection conditions chosen yet.  The deprotection must avoid not only decarboxylation but also nucleophilic addition into the vinylsulfone.  My question is how serious the problem of decarboxylation is likely to be in our case.

[pgk]

Neighboring of sulfone group to the carboxylic acid groups, may cause decarboxylation in the same way, as β-keto acids behave.
α-Neighboring of carboxylic acid groups with the double bond, further boosts decarboxylation that might occur at lower temperatures than usual, even near the room temperature.
So, try to remove the methoxy- groups by LiOH mediated hydrolysis that works at room temperature.
LiBr mediated ester hydrolysis at room temperature, is also reported.
If you are lucky enough, decarboxylation might not occur.
1). Lithium hydroxide/aqueous methanol: mild reagent for the hydrolysis of bile acid methyl esters, Steroids, 55(5), 233-237, (1990) http://www.sciencedirect.com/science/article/pii/0039128X90900213
2). Surprise in the Lithium Hydroxide Hydrolysis of a NXO-Compound, 13th Electronic Conference on Synthetic Organic Chemistry, Vienna, (2009) http://www.usc.es/congresos/ecsoc/13/hall_c_BMNPC/c13.pdf
3). A mild hydrolysis of esters mediated by lithium salts, Tetrahedron Letters, 48(14), 2497-2499, (2007)
http://www.sciencedirect.com/science/article/pii/S0040403907003152

[Babcock_Hall]

Thank you very much; I will check those out.

I am concerned however, that the decarboxylation might be happening one step earlier than the deprotection (consider this to be a preliminary assessment).  We took dimethyldibromosuccinate and sodium phenylsulfinate and heated in DMF for an extended period of time in an attempt to make compound 2.  Our synthesis was patterned after one that is in the literature.  The H-1 NMR spectrum of the crude product appears to have an AB pattern in the vicinity of 6.8-6.9 ppm, although the desired product should only have one proton.  One possibility is that phenylsulfinate attacked the methyl group of one of the methyl esters, instead of displacing bromide ion.  Decarboxylation might then have occurred.  We are attempting to predict the NMR spectrum of possible products.  If this preliminary assessment is correct, we will have to change protecting groups or change the synthetic route to compound 2.

After we started our work, we became aware of another route to 2 (Rajakumar and Kannan, J. Chem. Soc.
chem. Comm. 1989, 154-155).  It is a short paper, and there is less characterization than one might like.  Moreover, some of the H-1 data looks wrong.

[pgk]

It sounds a little weak because ΦSO2- > MeO-, as a leaving group.
Besides, simultaneous ΦSO2- mediated dehydrobromination in mixture and/or magnetic non-equivalence, could explain the 1-H NMR spectrum of the crude product.
Evans pKa Table:
MeOH, pka = 15.5
ΦSO2H, pka = 2.1
http://evans.rc.fas.harvard.edu/pdf/evans_pKa_table.pdf

[Babcock_Hall]

We may be talking about two slightly different mechanisms.  I was envisioning nucleophilic attack at the methyl carbon to displace a carboxylate leaving group.

[pgk]

Mechanism No1: ΦSO2- nucleophilic attack to the carbonyl that generates a MeO- leaving group and forms the mixed carboxy-sulfinic anhydride, which is then hydrolyzed during extraction, followed by decarboxylation due to the heat that is generated during anhydride hydrolysis:
Not possible because ΦSO2- > MeO-, as a leaving group.
Mechanism No2: ΦSO2- nucleophilic attack to the methyl carbon to that a carboxylate leaving group, followed by decarboxylation as described above:
Not so possible because electron donor bromide group decrease the pka of the acid but simultaneously electron attracting groups, such as carboxylic and phenylsulfinic increase the pka of the acid and thus, probably:
ΦSO2- > R(Br,CO2Me)CO2-, as a leaving group.

[rolnor]

Maby run the reaction in NMR tube to monitor?

[Babcock_Hall]

If you mean the formation of the vinyl sulfone,  we would have to put the NMR tube into a bath at 60 °C for many hours.  Is that feasible?  If you mean the deprotection, then that might be possible.

[wildfyr]

Yes, NMR can be run at raised temperatures, and spectra taken at time intervals, though of course you have to think about pressure in the tube if the solvent is volatile. It will obviously depend on your NMR model if it has such a feature built in.

You can also just hog the NMR for a day (or night), run it in a tube, and walk it back and forth from the machine and the lab every 20 minutes/1hour/whatever. The reaction will basically stop as it cools to RT right?