[pgk]

After the end of the reaction the compound is recrystallized in acetic acid. Thus, the second equivalent of pyridine is removed as acetate salt. Concerning the selective salt formation, it’s not only a question of pka but is also a question of concentrations; acetic acid is in high excess, during recrystallization.

[Babcock_Hall]

I had not thought about the acidity during the recrystallization; thank you.

[Babcock_Hall]

My students tried a recrystallization from acetic acid, but no crystals came out.  I told them to remove some solvent, and still nothing happened.  Recalling that we intended to remove one equivalent of pyridine at some point (so that we make the carboxylic acid form of our product--see upthread), I suggested pulling a vacuum and hoping that both acetic acid and pyridine would leave.  It might be possible to use NMR integrations to decide whether or not the second pyridine is present.  Such an event would mean that our product would still in the crude form, as a bromide salt.  The original procedure called for a second recrystallization, this one using 1-butanol; therefore, this step might remove the brown color that we see.

Should we keep plugging away at the acetic acid step or should we try a high vacuum in the hopes of removing the excess pyridine that way?

[pgk]

1). According to the publication, the temperature must be ≈ 0oC (or possibly, slightly higher) during recrystallization in the acetic acid, followed by recrystallization in butanol. Did the students work so?
2). Pyridine cannot easily be removed from a pyridinium salt by vacuum distillation, even if being a salt of a weak acid that is supposed to be in equilibrium with the non-ionic form.

[Babcock_Hall]

I believe that they did the synthetic steps (the ones involving diethyl ether) using ice, and they did get the solid initially.  The recrystallization did not specify an amount of solvent or a temperature, unless I am mistaken.  I was worried that if they cooled acetic acid below room temperature, that they would get crystals of acetic acid.  However with the crude product present, perhaps it is possible to go lower in temperature with respect to the recrystallization.

[pgk]

Indeed, the article does not specify the amount of acetic acid, needed and whether glacial acetic acid or not. On the other hand, acetic acid crystals might possibly cause a co-crystallization of the desired product.   

[wildfyr]

I know it's deviating off paper, but have you considered copper sulfate extraction to remove pyridine? Or maybe add a little water to help that recrystallization.

[Babcock_Hall]

I used copper sulfate to remove lutidine or collidine a long time ago, but I had not thought about employing it here.  It might not work, in that the product may be too water-soluble.  As far as having water in the recrystallization, can you explain your rationale?  This is a very old paper, and I have found one error so far (the amount of pyridine); therefore, I am not opposed to deviating from the protocol.  I am not entirely sure how I would.

One trick that has worked for me in the past is ion-exchange chromatography using a gradient of triethylammonium bicarbonate.  I am thinking it might be time to dust off that trick again.

[wildfyr]

If your stuff doesn't crystallize out of acetic acid easily, just add more of a bad solvent to it. Water might not be the right choice, it could be cyclo hexane or something organicky.  I'm advocating trying like 10-50% of something that is an even worse solvent for the product than acetic acid is.

[Babcock_Hall]

As pgk pointed out to us, the acetic acid might be instrumental in removing the second equivalent of pyridine, the one that is ion-paired to the carboxylate group.  However, at this point I would be happy for a relatively pure product in any form.

I recently rediscovered a description of the ethyl ester of our desired compound, which is compound (19) in this paper:  Lubbers et al. "Design, synthesis, and structure–activity relationship studies of new phenolic DNA gyrase inhibitors" Bioorg Med Chem Letters (2007) 17:4708-14.   I have not looked up one of the papers in reference 10 in this paper yet, which might be the actual description of how it was made.  One advantage is that the ethyl ester of bromopyruvate is not as expensive a starting material as the free acid.  One disadvantage (besides having to start over) is that we would need to work out a hydrolysis of the product that does not harm the pyridinium ketone functional group.
EDT
I looked up one paper in reference 10 within Lubbers' paper, the one from Tetrahedron (2004) 60:2937, by Wang et al.  They make the ethyl ester of pyridinium pyruvate, but they move on, apparently without isolating it.

[pgk]

You could try alkaline hydrolysis of the betaine ester with NaOH in DCM/MeOH (1) or K2CO3 in DCM (2) that are very fast operations and work at room temperature. Aldol/Claisen condensation and/or rupture of the quaternary salt do not seem possible during that short time and under these mild alkaline conditions.
1). A simple method for the alkaline hydrolysis of esters, Tetrahedron Letters, (2007), 48(46), 8230-8233
Indian Journal of Chemistry-Section B, (2006), 45B(09), 1729-1733
http://www.sciencedirect.com/science/article/pii/S0040403907018539
2). Isomerization of the Baylis-Hillman adducts using amberlyst-15 as a heterogeneous reusable catalyst: a simple and efficient stereoselective synthesis of (E)-cinnamyl alcohol derivatives, Indian Journal of Chemistry-Section B, (2006), 45B(09), 1729-1733
http://nopr.niscair.res.in/bitstream/123456789/6587/1/IJCB%2045B%287%29%201729-1733.pdf

[Babcock_Hall]

We are confused by the H-1 NMR of our putative product, which is pyridinium pyruvate.  The aromatic region looks reasonable, but the only candidate for the N-CH₂-C(O)R protons is at 5.0 ppm.  In our case R is a carboxylic acid; in the case of the pyridinium ketones I have found so far, R is an aromatic ring of some kind.  When R is aromatic, the chemical shift of this methylene group is often near 6.6 with some variation.  5.0 is slightly downfield of what I would expect if the -CH₂- group were next to a simple alkyl chain, not a ketone.  Thoughts?

[pgk]

Methylene 1H-NMR shift at δ = 5 ppm, seems reasonable because there is an additive effect from α-pyrdinium, α-carbonyl and β-carboxyl groups (Curphy-Morrison Shift Additivity Rules).
Nevertheless, the final determination of the structure can be concluded only by correlation of 1H-NMR, 13C-NMR and IR spectra.

[Babcock_Hall]

For the compound PyrNCH₂C(O)Ph, the chemical shift is 6.60.  For the compound 4CNPyrNCH₂C(O)Ph, the chemical shift is 6.67 ppm (Szafran and coworkers, J Molecular Structure 2004 708:87-95).  For the compound 4-CNPyrN-CH₂C(O)-Me, the chemical shift is 6.13 ppm (Szafran and coworkers, J Molecular Structure 2002 643:55-68).

I checked on page 222 of the 4th edition of Silverstein.  For groups that are β with respect to a CH₂ group, the downfield shift change of a phenyl ring is 0.35 ppm, relative to a β-methylene group.  The downfield shift change of a β-carboxlate ester is 0.5 ppm, relative to a β-CH₂.  In other words both the phenyl group and the carboxylate ester group should shift downfield in the general vicinity of 0.3-0.5 ppm.  IMO 5.0 ppm is too upfield for the proposed structure.  We will attempt to acquire more data, however.

[pgk]

1). How many protons correspond to the integration of the peak at 5 ppm?
2). Is your product a pyridiniumpyruvic acid derivative or is pyridiniumpyruvic acid, which is a zwitterionic betaine (pyridiniumpyruvate)?