Synthetic approaches to Diebenz[d,f]azonine alkaloid derivatives and analogous systems.

In the course of this work, the area which received most attention was the exploration of new synthetic avenues, leading from medium-ring heterocyclic precursors. A variety of these had become accessible via a ring-destruction approach from fused 1,2,3,4-tetra-hydroisoquinoline derivatives. The hitherto unexplored retrosynthetic relationship between the dibenz[dMazonine and dibenz[d,f]azecine alkaloids and the 3-benzazonine and 3-benzazecine ring systems was the primary area to which the present work was addressed. In particular, derivatives of 2,3,4,5,6,7-hexahydro-9,10-dimethoxy-1H-3-benzazonine and 1,2,3,4,5,6,7,8-octahydro-10,11-dimethoxy-3-benzazecine, were available bearing 7- and 8-hydroxyl groups respectively. These known compounds were converted efficiently, using standard methods, into the corresponding ketones bearing methyl N-carboxylate, cyanamide and methyl groups in the 3-position. These heterocyclic ketones exhibited unusual spectroscopic properties, previously observed only in medium-ring amino-ketones. Analysis of their structures as determined by X-ray crystallography and spectroscopic means, led to the conclusion that in each case, strong transannular interactions were responsible for their observed properties. A series of molecular mechanics calculations were also carried out on two model compounds employing the Allinger MMP1 method, which assisted in analysis of the spectroscopic properties of these ketones. It was envisaged that these ketones might serve as progenitors for dibenz[d,f]azonine and dibenz[d,f]azecine derivatives, using one of the many variations of the Robinson annulation reaction. The results of efforts in this direction are presented in Chapter 2. Using a cyclopentenone annulation procedure proceeding by way of a 1,4-diketone derivative it was hoped that syntheses of protostephanine and Cephalotaxusalkaloids might be developed. However little progress was made beyond the 1,4-diketone stage. Attempts to apply the classical procedures utilising a weakly basic catalyst and 3-buten-2-one or some of its synthetic equivalents, proved fruitless. This was attributed primarily to low acidity of the ketones. The use of stronger bases was incompatible with these annulation reagents, so attention was turned away from the Michael-type alkylations. However, it was found that the ketones mentioned above could be alkylated in very good yield, employing low-temperature enolate generation with lithium di-isopropylamide as base, followed by reaction with an appropriate alkyl halide. These procedures were developed using 3-benzazonin-7-one derivatives and since the best yields were obtained from alkylation of the methyl N-carboxylate derivative, this was generally employed. Using either Wichterle's reagent, 1-iodo-3-chloro-2-butene or Stork's reagent, (E)-trimethyl-(3-iodo-1-methyl-1-propenyl)silane, it was possible to prepare 3-benzazonin-7-one derivatives bearing a masked 3-oxobutyl group in the 6-position. Similarly reaction with 3-iodo-1-propyne yielded derivatives bearing a masked 2-oxopropyl group. Liberation of this latent functionality by accepted procedures, was in each case attended by some problems, which were satisfactorily overcome.