HNMR of Acids: Carboxyl OH Proton Chemical Shifts

[CrimpJiggler]

Chemical shifts for carboxyl OH protons usually appear down around 10-12 because they are so deshielded. Deshielding is basically what causes acidity so shouldn't more acidic carboxylic acids have chemical shifts greater than 11? Acetic acid has a peak at around 11.5 ppm:
http://riodb01.ibase.aist.go.jp/sdbs/cgi-bin/cre_frame_disp.cgi?spectrum_type=hnmr&sdbsno=306
Benzoic acid has a peak at around 12 ppm which makes sense since the benzene ring delocalises the charge on the proton and increases its acidity. I can't find a HNMR spectrum for trifluoroacetic acid but here's one for dichloroacetic acid:
http://riodb01.ibase.aist.go.jp/sdbs/cgi-bin/cre_frame_disp.cgi?spectrum_type=hnmr&sdbsno=5238
it contains a peak at around 9.4 ppm. I don't get it. The fluorine atoms are strongly electron withdrawing which is why difluoroacetic acid is more acidic than acetic acid. Why does the peak appear at 9.4 ppm? The proton should be far more deshielded than acetic acid so I would have expected its peak to be higher than 11.5 ppm. What am I missing here?

 

[CrimpJiggler]

Can nobody answer this question?

 

[Ygggdrasil]

I don't know the answer, but here are some speculations. Perhaps hydrogen bonding will affect the chemical shifts. For example, in non-polar aprotic solvents, acetic acid forms a dimer in which the acidic hydrogen forms a hydrogen bond with the carbonyl oxygen of another molecule of acetic acid. I'm not sure if this would occur in CDCl₃ or whether the effect would differ between difluoroacetic acid and acetic acid.

Here's another idea. Because CDCl₃ contains an acidic deuteron, perhaps the protons are exchanging between the dichloroacetic acid and the solvent:

CHF₂COOH + CDCl₃ <--> CHF₂COOD + CHCl₃

If this exchange occurs at rate faster than the NMR timescale, this effect would tend to lower the observed chemical shift of the acidic proton toward the chemical shift of the chloroform proton. If the difluoroacetic acid participates in the exchange more readily than acetic acid, this could perhaps provide some explanation for your observation.

 

[CrimpJiggler]

Thanks for the reply. I considered that hydrogen bonding may play a role but I didn't think about exchange between the deuterium and the acidic proton. That theory would explain the lack of direct proportionality between Ka and chemical shift. Theres a serious lack of information about this on the internet, is this a relatively unresearched area or something? It would be very easy to put that deuterium exchange theory to the test. With a proper lab that is, if I =had an NMR spectrometer and the equipment and chemicals required, I'd setup an experiment myself.

EDIT: Then again, CDCl3 isn't very acidic (the pKa of chloroform is 25, I don't know what differences there are between hydrogen and deuterium acidity though) so I assume its deuterium would only be likely to exchange with an acidic proton of an acid when the acids conjugate base dedeuteriumates (I know that's not the right word, I'll say deneutroprotonate from now on to be more accurate lol) it. Since the strength of the acid is directly proportional to the stability of its conjugate base, the stronger the acid, the less likely its conjugate base is to deneutroprotonate CDCl3. Surely there have been experiments setup to investigate all this. It brings to question the fundamental principles upon which NMR spectroscopy is based.

 

[Mike H]

I find this curious.

Checking my handy Cambridge Isotope Labs NMR solvents data sheet (don't leave home without it! - actually, it's an electronic copy on my laptop), I see that the acidic proton chemical shifts (referenced to TMS) for acetic acid-d₄ and trifluoroacetic acid-d are 11.65 ppm and 11.5 ppm, respectively. Either there is something unusual with difluoroacetic acid or there is something fluky about that database. Given that I know very little about difluoroacetic acid, but have had experience dealing with wonky data from that database in past research projects, I would not say that there is anything fundamentally wrong with NMR spectroscopy, at least not yet. ;)

Looking further at said solvents data sheet, I do see a significant change in the ¹³C carbonyl resonances for acetic acid and TFA - it's ~ 179 ppm for acetic acid and ~ 164 ppm for TFA. That seems to be reasonable enough.

People have been looking at deuterium exchange in small molecules and macromolecules via NMR for a while now. Perhaps not in detail for this particular set of molecules (I mostly do solids NMR of macromolecules nowadays, so I'm not overly familiar with the small molecule literature), but it's out there in the more general sense.