Optimized synthesis of enantiomeric C9-keto-5-phenylmorphans, essential intermediates for novel MOR agonists and antagonists

Abstract

Two modifications of previous procedures enabled the high-yield scalable synthesis of the pharmacologically important enantiomers of a C9-keto-5-phenylmorphan ((1R,5R)- and (1S,5S)-5-(3-methoxyphenyl)-2-methyl-2-azabicyclo[3.3.1]nonan-9-one). The use of sodium hydride rather than sodium amide as the base improved yields in two of the 6 step synthesis. This was initially in the benzylic alkylation to form 4-(dimethylamino)-2-(3-methoxyphenyl)butanenitrile (2), followed by the subsequent Thorpe-Ziegler cyclization with 5-bromovaleronitrile to provide the second of the three rings of the phenylmorphan. The second needed modification provided the third and final ring of the phenylmorphan was the demethylation of the quaternary amine 5-(3-methoxyphenyl)-2,2-dimethyl-9-oxo-2-azabicyclo[3.3.1]nonan-2-ium bromide (5) using refluxing triglyme under inert gas, rather than dry distillation or use of 1-nonanol; neither of these former conditions led to a desirable yield of 5-(3-methoxyphenyl)-2-methyl-2-azabicyclo[3.3.1]nonan-9-one ((±)-6) in a large scale reaction. Several of the intermediate steps leading to the C9-keto-5-phenylmorphan were also improved so that over the 6 steps that included enantiomeric separation of the racemic mixture, an overall yield of 40% was achieved and the desired compounds synthesized in 100 g amounts.