Rates of Decarboxylation of Substituted Anthranilic Acids in Nitrobenzene Solution

Abstract

The synthesis of a series of meta-and para-substituted anthraililic acids is reported with their rates of decarboxylation in nitrobenzene solution a t 210.5' C. The reaction is second order with respect to anthranilic acid. The rate is increased by electron-releasing substituents and decreased by deuterium in the functional groups. The relative rates are interpreted b y means of the Hainnlett equation and its extensions and it is concluded that the rate-determining step involves attack b y proton froin one anthranilic acid ~nolecule 011 carbon 1 of a second n~olecule.Previous investigations of anthranilic acid decarboxylation have shown that the reaction is first order in boiling aqueous solution but catalyzed by mineral acids (1,2). The aqueous-acid-catalyzed reactioil xvas coilsidered to involve attack by a proton fro111 the mineral acid on carbon 1 of the aromatic ring ( 2 ) and therefore to belong to the class of SE2 aromatic displacenle~lt reactions (3). Decarboxylatioil in the melt, however, must involve shift of a protoil to the ring from one of the functional groups of ailthranilic acid itself. Several possibilities for such a proton transfer are available: it inay occur in the neutral acid or the zwitterion, it inay involve the carboxyl group or the amino group, and it inay be intermolecular or intramolecular. The present investigation is an attempt to decide among these various possibilities.In order to determine whether the reaction is iilterinolecular or intrainolecular it is desirable to know the order of the reaction. In the melt the reaction has been reported to be first order (2, 4) but this is not very i~lfornlative since the con~position of the reaction mixture changes during the decarbosylation fro111 pure anthranilic acid to pure aniline. Since aniline could serve as proton donor it might take over the role of a second ~nolecule of anthranilic acid in an inter~nolecular reaction. An aprotic solvent would be preferable and it must be a high-boiling one since the temperature required for thermal decarboxylation is around 200" C. We have had some success with 1-methylnaphthalene as solveilt but nitrobenzene proved to be inore satisfactory since it is a better solvent for acids. In both these solvents the reaction is second order with respect to anthranilic acid and the reaction is therefore probably intermolecular.In an intermolecular reaction there are several possibilities for the rate-determining step. I t nlay involve ( a ) hydrogen-oxygen bond breaking, (b) hydrogen-nitrogen bond breaking, (c) hydrogen-carbon bond making, or (d) carbon-carbon bond breaking. Stevens, Pepper, and Lounsbury have eliminated possibility (d) by showing that there is no carbon isotope effect in the decarboxylation of anthranilic acid either in aqueous solution or in the melt (2). The three reinaini~lg possibilities may be distinguished by the site of the rate-determining step: in (a) it is the carboxyl group, in (b) it is the amino group, and in (c) it is carbon 1 of the ring. We have attenlpted to de...

Mechanism of decarboxylation of substituted salicylic acids. I. Kinetics in quinoline solution

“…The fist-order decarboxylation of salicylic acids in quinoline contrasts with that of substituted anthranilic acids in nitrobenzene which is second order (10). To determine whether the difference in order is a function of the acids o r of the solvents, salicylic acid was decarboxylated in nitrobenzene at 200 "C. The reaction is second order with respect to salicylic acid with a rate constant of 2 x M-I s-l. Conversely, a number of 4-and 5-substituted anthranilic acids have been found to decarboxylate by a first-order reaction in quinoline (1 3).…”

“…The apparatus for measuring rates of decarboxylation was similar to that previously described (10). The thermostat consisted of a 3 1 round-bottom flask containing a refluxing liquid (n-butyl ether) into the vapors of which dipped the reaction vessel.…”

EFFECT OF CHANGING pH UPON THE RATE AND MECHANISM OF DECARBOXYLATION OF SUBSTITUTED ANTHRANILIC ACIDS

“…T h a t is, HZA* should be more easily formed but less easily decarboxylated, so that the former alternative (eq. [9]) ~vould seem more reasonable for HZ* but the latter (eq. [lo]) would be more liliely for HZA*.…”

“…If it is assumed that the glass electrode iiieasures hydrogen-ion activity and that ionic strength will have little effect on the activity of the neutral species, HA, then it follows that: so that there will be an isotope effect unless kf >> k-,,, and k* >> k-,, but if this were the case eq. [9] would become equivalent to eq. [5] which does not fit the data.…”

“…However, for eq. [9] Since the quantity in parentheses cannot be less than 1, k* must be considerably smaller than k-,. Therefore mechanism [9] and its special case [ I l l require a carboxyl-carbon isotope effect a t pH 4 or greater.…”