Studies in morphinian chemistry
Synthetic routes from thebaine to l4β-acylaminodeoxydihydrocodeinones
were explored. Each route was eventually frustrated by
our inability to remove reductively either a 4-phenoxy or a 4-(1-phenyltetrazol-
5-yl)oxy group from a number of derivatives.
Routes to the preparation of a thebaine derivative containing
a piperidine nitrogen-carbon(14) bridge were investigated. In the
N-northebaine series no bridged compounds were formed but in the
thebaine series two quaternary ammonium chlorides with the desired
bridge were prepared.
Treatment of the cycloadduct 6β,14β-(N-chloroacetylepoxyimino)-
6,14-dihydro-N-t-butoxycarbonylnorthebaine with sodium ethoxide in
ethanol afforded, unexpectedly, l4β-(2,2-diethoxyethanoylamino)-N-tbutoxycarbonylnorcodeinone.
Similarly, the cycloadduct 6β,14β-
(N-phenylacetylepoxyimino)-6,14-dihydrothebaine was found to react with
high stereoselectivity with methoxide to give l4β-[(S)-2-methoxy-2-
phenylethanoylamino]codeinone. Treatment of the hydroxamic acid
l4β-(N-2-phenylethanoylhydroxyamino)codeinone with toluene-p-sulphonyl
chloride in pyridine followed by methoxide gave a yellow product which
had incorporated one molecule of pyridine. All the foregoing reactions
are believed to proceed via aziridinone (α-lactam) intermediates.
Further studies on model compounds supported the proposed mechanisms.
An alternative, high yielding synthesis of the known a-lactam, l-t-butyl-
3-phenylaziridin-2-one, has been achieved by treating the appropriate
hydroxamic acid with trifluoromethanesulphonic anhydride and triethylamine
at 700 C.
The reaction of the hydroxamic acid l4β-(N-chloroacetylhydroxyamino)-
N-t-butoxycarbonylnorcodeinone ethylene acetal with ethoxide
gave the l4β-aminonorcodeinone derivative. A similar treatment of
the corresponding hydroxamic acid derived from thebaine and also of
model compounds yielded ester products. These reactions are
rationalised in terms of oxazetidinone intermediates.