[Babcock_Hall]

We plan to synthesize Pyrdinium-CH2C(O)CO₂X (where X could be a proton or a pyridinium ion) by the method of Arthur Green and Raymond Delaby ("Pyridinium and thiazolium salts with multifunctional propyl side chain" Bull Soc Chim France 1955, pp. 697-699).  My student translated this article.  It seems straightforward (solvent is diethyl ether), except for one problem.  The mass of bromopyruvic acid (MW 166.96) is 16.7 g, and the mass of pyridine (MW 79.1) is 1.7 g).  This works out to 100 mmols of bromopyruvic cid, and 21.5 mols of pyridine.  In other words, the bromopyruvic acid is almost 5-fold in excess, which strikes me as odd.

One possibility is that the mass of pyridine is a typo, and there should be 17 grams of pyridine, which would provide just over a twofold excess of pyridine.  A second equivalent of pyridine might be expected to deprotonate the carboxylic acid, which might facilitate the reaction.  Any suggestions on the best way to proceed?

[pgk]

a). Probably, it is a typing error because for the preparation of Girards T and P reagents, equimolar amounts of tertiary amine and haloacetate ester are used.
1). Helvetica Chimica Acta, 19(1), 1095-1107, (1936)
http://onlinelibrary.wiley.com/doi/10.1002/hlca.193601901148/full
2). Organic Syntheses, Coll. Vol. 2, 85, (1943)
http://orgsyn.org/demo.aspx?prep=cv2p0085
b). Pay attention to translation from French to English (if not already done) because French use a different numeration system for decimals.
Seven thousand = 7.000 in French
One and a half = 1,5 in French.

[Babcock_Hall]

For both numbers the authors use commas.  The masses are given for bromopyruvic acid and pyridine are 16,7 g and 1,7 g, respectively.

[kriggy]

Actualy I´ve never seen anyone in europe writing 7 thousand as "7.000 its always 7000).

[Babcock_Hall]

I think that they must have meant 17 grams, not 1.7 grams.  This would allow for a second equivalent to neutralize the carboxylic acid.

[pgk]

Officially, they do so in France, Belgium and in other French speaking countries, as well as in Greece.

[Babcock_Hall]

What they isolated had a single nitrogen by elemental analysis.  Therefore, if they used two equivalents of pyridine (corresponding to 17 grams), one equivalent was discarded in the synthesis.

[pgk]

Then obviously, it was a typing error. But the obtained results indicate that the reaction occurs via preliminary formation of the carboxylate salt, followed by formation of the tertiary amine salt and not through direct betaine formation that would need only one equivalent of pyridine.

[rolnor]

Should not  the product be a quaternary amine?

[pgk]

Yes, right!
....the obtained results indicate that the reaction occurs via preliminary formation of the carboxylate salt, followed by formation of the "quaternary" amine salt and not through direct betaine formation that would need only one equivalent of pyridine.
Thanks Rolnor and very sorry for that serious mistake.

[Babcock_Hall]

This is my attempt to type out the original (which I believe was a scanned image in the form of a pdf file.  I left out the accents and such.  I may have made some typos:
Bromure de N-(-carboxy –cetoethyl)pyridinium (II.7).

L’acide bromopyruvique (16,7 g) est dissous dans 100 cm³ d’ether absolu.  La solution etheree est traitee par du noir Norit et le filtrate est fortement refroidi.  Par ailleurs, 1,7 g de pyridine seche est dissous dans 100 cm³ d’ether anhydreet cette solution est egalement refroidie dans la glace.  Lorsque la temperature des solutions etherees est de 0 °, la solution contenant la pyridine est verse dans l’autre: aussitot un precipite huileux se depose.  Peu apres, il se recrystallize dans l’acide acetique et seche ensuite sous vide sur P₂O₅; il est recristiallise dan l’alcool butylique.  Le produit est conserve au sec.  Obtenu: 15 g.

Analyses pour C₈H₈O₃NBr:

[Babcock_Hall]

The pK_a of pyruvic acid itself is 2.50, and the pK_a is 5.17 for the conjugate acid of pyridine (both values from Williams).  This is our translation of the passage in my previous message:

N-(β-carboxy-β-oxoethyl)pyridinium bromide (II.7)

Bromopyruvic acid (16.7 g) was dissolved in 100 mL absolute ether. The ether solution was treated with Norit® activated carbon and the filtrate was significantly cooled. In addition, 1.7 g dry pyridine was dissolved in 100 mL anhydrous ether and this solution was also cooled in ice. Once the temperature of the ether solutions is 0°, the solution containing the pyridine was poured into the other: immediately, an oily precipitate settled. Shortly after, it forms a white crystalline mass. The compound was rotary evaporated; recrystallized from acetic acid and then dried under vacuum with P2O5; it is recrystallized from butyl alcohol. The product is kept dry. Obtained: 15 g.
Analysis for C₈H₈O₃NBr
Calculated: C % 39.02      H % 3.25   N % 5.69    Br % 32.52
Found:           41.2              3.59           5.98             35.6

[pgk]

It an obvious typing error, as being concluded from your experimental results, as well from the preparation protocols of Girard’s reagents and various betains.

[Babcock_Hall]

I agree that 1.7 grams is a typing error.  It is less clear to me what the correct value should be, with 17.0 and 8.5 grams being the best contenders.  The reactant in the reference from Organic Syntheses is an ester, as opposed to a free acid.  If only one equivalent of pyridine is added, the reaction may be sluggish, because some of the pyridine will exist as a pyridinium salt with the carboxylate group.

That is why I think that the authors intended 17.0 grams (two equivalents).  My best guess is that after the reaction is over (during rotary evaporation or high vacuum), the second pyridinium ion donates its proton back to the carboxylate group, and the now-neutral pyridine leaves under reduced pressure.  This would return the carboxylate group back into its neutral acid form. Something like this must happen when excess triethylammonium bicarbonate is removed from a product.  TEAB is sometimes used in the gradient ion-exchange purification of nucleotides.  In its removal, the triethylammonium ion must give its proton back to bicarbonate, so that both will be volatile. 

[pgk]

The example in the "Organic Syntheses" is indicative to the necessity of one equivalent of tertiary amine per one halogen substituent. In your case, another equivalent of tertiary amine is also required for neutralization of the carboxylic group, in order to avoid the inhibition the formation of the quaternary salt.