Ferrier–Nicholas Cations from C‐3‐Alkynylglycals: Access to C‐3‐Branched Allylic Glycosides and Ring‐Opening Derivatives

Abstract

Hexacarbonyldicobalt–C‐3‐alkynyl‐substituted glycal derivatives, when treated with BF3·OEt2 give rise to Nicholas‐stabilized Ferrier cation intermediates (Ferrier–Nicholas cations) that react with alcohols or C‐nucleophiles to give C‐3‐branched 2,3‐unsaturated glycosides or C‐glycosides, respectively. α‐C‐Glycosides were the sole compounds obtained when allyltrimethylsilane was used as the nucleophile. On the contrary, the reaction of C‐3‐alkynylglycals with heteroaryl or alcohol nucleophiles led to anomeric mixtures in which the β‐anomers prevailed. The presence of the hexacarbonyldicobalt–C‐3‐alkynyl substituent seems to be of key importance in the stereoselectivity of these transformations, since the reaction of C‐3‐alkynylglycals – devoid of the hexacarbonyldicobalt moiety – showed a preferred α‐stereoselectivity. Furthermore, a bis(indolyl) linear compound was obtained from the reaction of a hexacarbonyldicobalt–C‐3‐alkynylglycal with 2 equiv. of indole.

Diversity‐Oriented Synthetic Endeavors of Newly Designed Ferrier and Ferrier‐Nicholas Systems Derived from 1‐C‐Alkynyl‐2‐deoxy‐2‐C‐Methylene Pyranosides

“…From the results in Table , and in accordance with our previous study on the behavior of related Ferrier‐Nicholas systems,[23b] we have postulated a reaction pathway that could accommodate the formation of isolated compounds, 52a – 54 (Scheme ). Accordingly, acid‐activation of the anomeric trimethylsilyloxy group in 15a would produce allylic Ferrier‐Nicholas cation 31 , which could react with adventitious H 2 O either at C‐1 or at C‐2′.…”

“…Reaction at C‐1, would produce the corresponding ketose ( 15a , OH), and thence open‐chain hydroxy‐aldehyde 52a , whereas reaction at C‐2′ would lead to allylic silyloxy (or C‐2′‐hydroxymethyl) glycal 56 . The latter, by activation of its C‐3 substituent, could experience a Ferrier rearrangement to give Ferrier‐Nicholas cation 57 that, in line with our previous observations,[23b] could experience a 1,6‐hydride transfer generating oxocarbenium ion 58 . Hydrolysis of this species, rather than Prins cyclization,[23b] would originate the 3‐deoxy‐6‐hydroxy derivative 59 .…”

“…The latter, by activation of its C‐3 substituent, could experience a Ferrier rearrangement to give Ferrier‐Nicholas cation 57 that, in line with our previous observations,[23b] could experience a 1,6‐hydride transfer generating oxocarbenium ion 58 . Hydrolysis of this species, rather than Prins cyclization,[23b] would originate the 3‐deoxy‐6‐hydroxy derivative 59 . The latter upon acid‐activation would generate Ferrier‐Nicholas cation 60 , from which the formation of 1,6‐anhydro pyranose 53 , or open‐chain 3‐deoxy derivative 54 , could take place either by intramolecular‐glycosylation or hydrolysis, respectively.…”

“…In line with our previous studies on the behavior of Nicholas‐cations, compound 15a displayed a reactivity pattern different to that of their non‐cobaltated analogs 14 . [23b], Thus, reaction of 15a in the presence of InBr 3 yielded open‐chain derivative 52a along with 1,6‐anhydro‐2,3‐dideoxy‐2‐ C ‐methylene pyranose 53 (Table , entry i ). When the reaction was performed with BF 3 · OEt 2 as catalyst (1.2 equiv., –20 °C), three compounds were obtained (Table , entry ii ): compound 52a , an additional open‐chain derivative 54 , and 1,6‐anhydro pyranose 53 , this time as the major isomer.…”

“…Based on our previous experience with pyranosidic C ‐alkynyl and C ‐alkynyl‐hexacarbonyl dicobalt derivatives, we identified two potential sets of Ferrier‐ and Ferrier‐Nicholas‐like derivatives that could behave as pictured in Scheme . In this context, compounds 12 and 13 (Figure ), initially contemplated owing to their similarity to glycals type 4 , were soon discarded because of the complex synthetic route involved in their preparation.…”