The predominance of the rearranged product (V) and the absence of conessine (II) in the reaction mixture when ether was used as a solvent points to the intermediate formation of a rapidly rearranging carbonium ion or to the immediate formation of a rearranged item via a concerted cleavage of Cl2-0 and C 13-Cl4 linkages.
Consequently, we thought it would be useful to work out conditions favouring the unimolecular Cl*-0 bond cleavage of the sulphonate.
It was indeed verified that the rearrangement of holarrhenine mesylate (III) is complete without occurrence of O-S bond cleavage, when the reaction is performed in anhydrous alcohols using an excess of sodium borohydride. We think this procedure could be a choice method to convert 12B-hydroxy steroids with a normal skeleton into the corresponding saturated C-nor-D-home derivatives when the reaction with lithium aluminium hydride and the Bamford-Stevens reaction on 12-keto steroids (4) fail.
Recently it was reported (5a) that 12B-tosyloxyconanine (IV, H2 at C3, 5a-H) rearranges in the same way with lithium aluminium hydride-aluminium chloride as a reagent in ether.
Mechanistic studies including deuterium labeling afforded a strong evidence that the rearrangement as outlined above is followed by a hydride shift from Cl8 to Cl3 on the rear side of the molecule, which could be assisted by the electron pair of the nitrogen atom (5b).
Reduction of holarrhenine mesylate (III) with sodium borodeuteride in 0-deuteroethanol or in 0-deutero-i-propanol led to a 97% monodeuterated rearranged product (V, 18a-d) with at least 80% deuterium incorporation at Cl8 (NM%?, MS). These results are in good agreement with the findings of the French authors.
Further details on the experiments and spectroscopic as well as chemical evidence for the suggested structure (V) will be presented soon in a full publication.