作者:Venkataraman Amarnath、Douglas C. Anthony、Kalyani Amarnath、William M. Valentine、Lawrence A. Wetterau、Doyle G. Graham
DOI:10.1021/jo00024a040
日期:1991.11
The mechanism of Paal-Knorr reaction between a 1,4-dicarbonyl compound and ammonia or a primary amine to form a pyrrole is explored. In aprotic solvents and in aqueous solutions near neutrality, d,l diastereomers of 3,4-dimethyl- and 3,4-diethyl-2,5-hexanediones (1r and 2r) formed pyrroles 1.3-57.0 times faster than the corresponding meso diastereomers (1m and 2m). This contradicts any intermediate, such as the enamine 15, which does not remain saturated at both the 3- and 4-positions through the rate-determining step. The demonstrated stereoisomeric difference in reactivity coupled with the following results support the hemiaminal (9) as the intermediate undergoing cyclization in the rate-limiting step of the Paal-Knorr reaction: (1) The reaction rate was adversely affected by increase in the size of the alkyl substituents on the dione. (2) Racemic 2,3-dimethyl-1,4-diphenyl-1,4-butanedione (3r) was more reactive toward ammonium acetate (2.2:1) and 2-aminoethanol (11.2:1) than the meso isomer (3m), ruling out the involvement of the less substituted enamine 14. (3) The relative rate of pyrrole formation of 1,4-diphenyl-1,4-butanedione (5) and its dimethoxy (6) and dinitro (7) derivatives (1:03:6) does not support cyclization of the imine (11) to the pyrrolinium ion (12). (4) The rates of reaction of 2,2,3,3-tetradeuterio-1,4-diphenyl-1,4-butanedione (5D) and perdeuterio-2,5-hexanedione (4D) were very close to those of unlabeled diketones, indicating the absence of a primary isotope effect in the reaction. (5) Neither the isomerization of the unreacted diastereomers of 1, 2, and 3 nor hydrogen exchange of 4D and 5D was detected during the reaction.