adidx, where is this figure from, particularly the numbers? You said you did not read about this anywhere, so where is it coming from?
I don't like this hydrogen binding argument, intramolecular H-bonding in a 6-membered ring no problem, but a 5-membered ring I think is too unfavorable deviation from the standard angles.
Babcock_Hall, I don't understand why are you making a comparison to the methyl ester of glycine and dismissing my comparison to methylamine? I believe the comparison to a primary amine reveals the inductive effect of the carboxylic group. Your comparison to the methyl ester is of course very useful as well - but I don't think electrostatics (why would the molecule want its ends pulled together? That might be enthalpically ok, but screams against the entropy) or hydrogen bonding play a role. Please prove me wrong, I will take some data for your models.
I have seen this numerous times in textbooks and in conversations with organic physical chemists - the inductive model is quite well discussed. Good you mentioned Gly-Gly as well - so the trend is Gly, Gly-Gly, Gly-OMe - acidity of the ammonium increases. Around the ammonium's pKa's the carboxylate in these molecules exists as: C(O)O-, C(O)NHR, C(O)OMe. The electron density of the carboxylic moiety decreases in the same order (see IR data for the C=O stretch for confirmation) and hence the inductive electron withdrawing capabilities of the group increase - this leads to the destabilization of the ammonium and makes it a progressively stronger acid. Of course the increase in electron density is explained by a) the anionic O- being the best donor for the reasons of delocalization of charge, b) amide nitrogen being a better lone pair donor than ester's oxygen because of the better overlap with the C=O π*.
The increased acidity is similarly explained by the inductive withdrawing effects of the NH2 group.
What do you think?