Background: We previously reported the 1:1 condensation reaction between ophthalaldehyde
and primary amine in the presence of 1,2,3-1H-benzotriazole and 2-mercaptoethanol as
dual synthetic auxiliaries to provide phthalimidines (2,3-dihydroisoindol-1-ones) under mild reaction
conditions in good isolated yields. However, this reaction was found to proceed via dissymmetrization of
equi-oxidation stage functionalities (carbonyl groups herein), which would give rise to regioisomer(s)
whenever the orientation of substituents in the target molecule phthalimidine was dissymmetrical. In order
to overcome such defects, we focused on phthalide, which possessed identical electronic structure
with that of target molecule phthalimidine. Thus, the condensation between a phthalide derivative and a
primary amine would eliminate a water molecule, but does not accompany with any kind of redox paths.
Therefore, dissymmetrical trends possessed by the phthalide derivative should transfer directly to the resulting
phthalimidine, and as a result the formation of possible regioisomer(s) can completely be prevented.
This strategy has existed as the oldest method of phthalimidine synthesis, however, Lewis acid
catalysts utilized in the past were not appropriate to extend its scope. In this report, we demonstrate an
improved method of phthalimidine synthesis along this line using InCl3 as a catalyst, which would allow
the synthesis of this class to accommodate with the contemporary standard of organic syntheses.
Method: In general, aniline (2.0 mmol, used at once as reagent and solvent) was added to the flask containing
phthalide (1.0 mmol) and InCl3 (an appropriate amount) under Ar atmosphere over Ca. 3 min with stirring,
and the whole mixture was heated at gentle reflux for an appropriate period. After usual work-up, crude
products were purified by column chromatography on silica gel. Isolated components were ascertained by
1H NMR spectral comparison with the authentic samples prepared by our former methods.
Results: Optimization of the reaction condition was explored using phthalide and aniline. Thus, decrement
of yields occurred whenever amounts of InCl3 catalyst were reduced, however, longer reaction periods
improved isolated yields every time. Even in the use of 1 mol% of InCl3 catalyst, reaction would
complete within a few days. As a result, the direct condensation between phthalide and aniline with the
use of a truly catalytic amount of Lewis acid catalyst was made possible for the first time in the history of
phthalimidine syntheses. Extension of this strategy to some primary amines other than aniline was successful,
too. Based on experimental results, our plausible mechanism was proposed, in which the reaction
commenced by nucleophilic attack of an amino group towards 3-position of oxygen-chelated phthalide.
Conclusion: We succeeded in the direct condensation reaction between phthalide and primary amine
in the presence of a catalytic amount of Lewis acid catalyst, InCl3, for the first time. Although successful
examples have thus far been limited to some combination that affords clear reaction mixtures (except
for the catalyst) under high-temperature, solvent-free heating, we believe that we have attained a
tacit basis to prepare a variety of phthalimidine derivatives possessing a variety of substituent patterns.