Direct and highly diastereoselective synthesis of azaspirocycles by a dysprosium(III) triflate catalyzed aza-Piancatelli rearrangement.

number of elegant approaches to construct azaspirocycles have been developed. The methods differ in the strategy used to address the two main synthetic challenges: construction of the tertiary carbon center bearing the nitrogen atom, and the formation of the spirocyclic ring system. Generally, the synthetic routes rely on a two-step process where the tertiary carbon center and the spirocycle are formed in separate, discrete synthetic steps. The most practical, but also the most difficult and rare approach to construct this challenging azaspirocyclic framework is to combine the construction of the tertiary stereocenter and the formation of the spirocyclic ring system within a single operation. Recently, our group developed a new cascade strategy for the efficient synthesis of trans-4,5-disubstituted cyclopentenones based on the aza-Piancatelli rearrangement. The overall transformation is highly diastereoselective and believed to proceed through a cascade sequence that terminates with a 4p-electrocyclic ring closure to give a pentadienyl cation; a step that is analogous to the Nazarov cyclization. Given the synthetic importance of the azaspirocycle structural motif and the stereospecificity of the 4p electrocyclizations, we hypothesized that 2-furylcarbinols bearing an aminoalkyl side chain at the 5-position of the furan ring would be effective for the general synthesis of functionalized azaspirocycles. Piancatelli et al. demonstrated that 2-furylcarbinols with substituents at the 5-position undergo an intermolecular rearrangement to give the corresponding cyclopentenone containing a tertiary alcohol; often a decrease in both yield and reaction rate was observed. More recently, Wu et al. described the reactivity of 2-furylcarbinols containing a hydroalkyl side chain at the 5-position of the furan ring. In their case the rearrangement led exclusively to the more stable oxa-bicyclic cyclopentenone and they believe the reaction does not proceed through an intramolecular Piancatelli rearrangement. We envisioned a mechanistic scenario that would proceed in an intramolecular fashion, analogous to the proposed azaPiancatelli rearrangement (Scheme 1). By employing the proposed cascade rearrangement, we sought to overcome the challenges associated with the generation of an azaspirocyclic framework by relying on the electrocyclization to construct the sterically congested azaspirocycle (D–E). Herein, we report an efficient and highly diastereoselective synthesis of functionalized azaspirocycles, based on the first example of an intramolecular aza-Piancatelli rearrangement. Furylcarbinol 9 was selected as a model substrate to probe the feasibility of the new intramolecular rearrangement and dysprosium(III) triflate (Dy(OTf)3) was chosen as the catalyst Figure 1. Alkaloids that contain 1-azaspirocycle motifs.