The synthetic study commenced with the preparation of the known
N-alkyl substituted 3, 4, 5, 6, 7, 8-hexahydro-2(1
H)-quinolinone
10 [
12,
13] as the precursor of BE rings (
Scheme 2). Although a principle method for preparation of this crucial type of ene-lacatams has been reported starting from cyclohexone and acrylonitrile by Murahashi
via a two-step sequence [
13], poor yield and using of noble metal catalyst (RuH
2PPh
3) under forcing conditions (sealed tube, 120 ℃) limit its application in a scalable synthesis. Other methodology [
12] involving aza-annulation with imines and various acrylate derivatives were restricted by lack of alkene regioselectivity and poor yields that result from formation of the imines from ketones. In this event, we present a new two-step reaction to produce
10 involving a highly efficient aza-annulation of a
β-enamino ketone and an incidentally discovered enamine tautomerization induced by a facile decarboxylation. First, following the Stille's aza-annulation protocol [
14], reaction of ethyl 2-cyclohexanonecarboxylate
9 with benzyl amine in the presence of the Et
2O·BF
3 followed by exposure of resultant crude enamine to acryloyl chloride gave the
12a smoothly, with the alkene exocyclic relative to the ring generated through annulation. Then hydrolysis of the ester group of
12a provided the crude acid, which without purification, was warmed at 50 ℃ under acidic conditions in one pot to effect a mild decarboxylation, delivering the desired double bond shifting ene-lacatam
10a in 81% yield over two steps. Pleasingly, replacement of benzyl amine to ethyl amine in first step could achieve the almost the same yield of
N-ethyl substituted ene-lacatam
10b by executing the same two-step reaction sequence.