# Decarboxylation of Salicylic acid

I would like to know how and under what conditions can decarboxylation of salicylic acid (o-hydroxybenzoic acid) occur. I couldn't find any sources for it online. Here are some reasons I think it would happen:

1. The $\ce{COOH}$ group is out of plane of the benzene ring due to the ortho $\ce{OH}$ group, making it easier to decarboxylate.
2. The keto form is a beta keto-acid, which makes it easier to undergo decarboxylation.

From (1):

Salicylic acid decarboxylated to give phenol and carbon dioxide, at a convenient rate for kinetic measurements, at $200-230^\circ$, in a homogeneous solution of benzoic acid and certain other high-boiling solvents.

... there was a considerable amount of interest in the possibility of a bimolecular decarboxylation mechanism involving an attack by a proton on the free acid or its anions. A variety of compounds such as anthracene-9-carboxylic acid, substituted cinnamic mesitoic acid, aminosalicylic acid, and salicylic acid, appeared to represent examples of this type of decarboxylation mechanism. In most cases, pseudo-first-order rate constants were observed experimentally.

... mechanism is also proposed for the decarboxylation of salicylic acid in benzoic acid solution. A competition between an oxygen atom and the terminal ring carbon atom for the proton to give either the undissociated acid or a $\sigma$-complex leading to products is shown by eq. 3.

The effect of substituents on the rate of decarboxylation of salicylic acid has been used to support a bimolecular mechanism. Electron-donating groups in the ortho or para positions have led to an increase in rate. The increased electron density on the ring carbon atom bonded to the carboxyl group would presumably facilitate an attack by a proton at this position.

... An attack by a proton on the terminal ring carbon atom of the salicylate ion would appear, at first glance, to be extremely unfavorable because of the close proximity of a formal negative charge. Any interaction should merely lead to undissociated acid. The hydroxyl group in the ortho position appears to influence the distribution of the charge because of the steric configuration. Resonance and tautomeric forms, eq. 5, which do not involve an extensive shift of atoms, can be drawn which would distribute the charge in a manner favorable to the proposed mechanism, eq. 3.

This unique steric arrangement is not possible with m- or p-hydroxybenzoic acid, the effect of which is borne out by the fact that these two acids did not decarboxylate at a measurable rate in benzoic acid solution at $212^\circ$. Since inductive and resonance effects would be expected to be similar for the ortho and para isomers, the steric effect appears to exert a critical influence.

(1) Kaeding, W. W. Oxidation of Aromatic Acids. IV. Decarboxylation of Salicylic Acids J. Org. Chem. 1964, 29 (9), 2556–2559. DOI: 10.1021/jo01032a016