Reduction of Aniline in Ammonia

Abstract

In a recent paper by Watt, Knowles and Morgan¹ on the course of reduction in liquid ammonia the following statement is made: “In view of the results reported by Wooster² it is surprising that reduction of nitrobenzene by the sodium-methanol combination yielded aniline and not dihydroaniline—”. Unpublished work by me has shown that aniline is reduced by sodium and methanol in liquid ammonia, as would be expected by analogy with the reduction of dimethylaniline³, which Watt, Knowles and Morgan do not mention. If the reduction is carried out rapidly, the product is a dihydroaniline, probably the 2:5-, since acid hydrolysis leads to cyclohexenone. Slow reduction, however, gives chiefly a tetrahydroaniline, probably the 1:2:5:6-; 3:5-dinitrobenzoyl derivative, m.p. 198° (found: C, 53.3; H, 4.6. C₁₃H₁₃O₅N₃ requires C, 53·6; H, 4·5 per cent). Its melting point is depressed by admixture with the derivative from aniline, m.p. 235°, and it cannot be an enamine since it is stable to mineral acid. There is little reduction of dimethylaniline past the dihydro-stage under the conditions used here, namely, excess of alcohol, so that the further reduction of dihydroaniline is probably due to isomerization of the enamine to the imine: