6751
amount of cuprous chloride in DMF or i-PrOH gave nitrile (2a) in about 10% yield. A signi®cant
improvement (50% yield) was observed when both hydrogen peroxide and ammonium hydroxide
were added dropwise simultaneously to the other reagents.
Yet better results (Table 1, method A) were obtained when starting aldehyde (1a) (0.38 mol)
was also added parallel with hydrogen peroxide (1.4 mol) and ammonium hydroxide (0.44±0.49
mol) to a stirred mixture of copper catalyst (0.01 mol) and 300 ml of i-PrOH at 17±30ꢀC (the
reaction was noticeably exothermic, so some cooling was necessary). Importantly, the addition
rates were selected so as to always maintain low concentrations of all starting reagents and
intermediate azomethine (3a) in the course of the reaction. Speci®cally, ammonium hydroxide
was added for about 3 hours, and the addition of aldehyde (1a) and hydrogen peroxide
commenced almost simultaneously with ammonium hydroxide and was ®nished in 2.5 and
4 hours, respectively. Other aldehydes (1b±h) were converted similarly. After a conventional
work-up, the products were puri®ed by distillation or crystallization.
In method B, gaseous ammonia (2.5±2.8 mol) was passed for 3 hours into a solution of 4 g of
cuprous chloride in 800 ml of i-PrOH, while 1.5 mol of aldehyde (1), and 3.35 mol of 50%
hydrogen peroxide were added dropwise for 2.5 and 4 hours, respectively. Both procedures A and
B were optimized with respect to the conversion of aldehyde (1a) into nitrile (2a) only, which
explains the relatively lower yields of other a,b-unsaturated and aromatic nitriles. Poor yields of
aliphatic nitriles (2g,h) may be due to the slow rate of oxidative dehydrogenation of the
corresponding azomethines (3g,h).
In the case of a,b-unsaturated aldehydes, the formation of small amounts of retro-aldol
by-products was observed by GLC, for instance 6-methyl-5-hepten-2-one from aldehyde (1a) or
benzaldehyde (and benzonitrile) from (1b). Also, small amounts of carboxylic acids and acetone
were formed due to the direct oxidation of starting aldehydes and i-PrOH solvent, respectively.
The requirement for a signi®cant excess of hydrogen peroxide over the aldehyde raised a
legitimate suspicion that the real oxidant is oxygen formed in the copper-catalyzed decomposition
of hydrogen peroxide. However, control experiments starting with aldehyde (1a) and using air or
oxygen instead of hydrogen peroxide gave practically no nitrile (2a), so hydrogen peroxide
remains the most likely actual oxidant. There is not enough data yet to discuss the mechanistic
aspects of the reaction, and even the intermediacy of azomethines (3) has not been proven. The
reaction certainly deserves more detailed study.
Regardless of its mechanism, the reaction of aldehydes with ammonia and hydrogen peroxide
in the presence of copper catalysts presents a new, convenient and ecient method for obtaining
nitriles.
References
1. Jose, B.; Sulatha, M. S.; Pillai, P. M.; Prathapan, S. Synth. Commun. 2000, 30, 1509±1514.
2. Desai, D. G.; Swami, S. S.; Mahale, G. D. Synth. Commun. 2000, 30, 1623±1626.
3. Sampath Kumar, H. M.; Subba Reddy, B. V.; Tirupathi Reddy, P.; Yadav, J. S. Synthesis 1999, 586±587.
4. Chaudhari, S. S.; Akamanchi, K. G. Synth. Commun. 1999, 29, 1741±1745.
5. Wang, E.-Ch.; Lin, G.-J. Tetrahedron Lett. 1998, 39, 4047±4050.
6. Smith, F. R.; Albright, J. A.; Waring, A. M. J. Org. Chem. 1966, 31, 4100±4102.
7. Mochovski, J. Chem. Pap. 1998, 52, 45±51.
8. Miller, M. J.; Loudon, G. M. J. Org. Chem. 1975, 40, 126±127.
9. Katritzky, A. R.; Molina-Buendia, P. J. Chem. Soc., Perkin Trans. 1 1979, 1957±1960.