A Simple and Broadly Applicable C−N Bond Forming Dearomatization Protocol Enabled by Bifunctional Amino Reagents

Abstract A C−N bond forming dearomatization protocol with broad scope is outlined. Specifically, bifunctional amino reagents are used for sequential nucleophilic and electrophilic C−N bond formations, with the latter effecting the key dearomatization step. Using this approach, γ‐arylated alcohols are converted to a wide range of differentially protected spirocyclic pyrrolidines in just two or three steps.

Purification of the product by flash column chromatography (33% EtOAc/Hexane) afforded the title compound (0.44 g, 50%) as a colorless oil. Spectroscopic properties were consistent with the data available in the literature. The compound was prepared according to a literature procedure. 27 Spectroscopic properties were consistent with the data available in the literature. 28

2-(Cinnamyloxy)naphthalene 29
The title compound was prepred according to a literature procedure. 29 Spectroscopic properties were consistent with the data available in the literature. 29 30

80
The title compound was prepared according to a literature procedure. 29 Spectroscopic properties were consistent with the data available in the literature. 30

Computational methods
An evaluation of the conformational freedom of the reactant, TS and products was performed using the GMMX routine implemented in PCMODEL v.8.5, 31 and subsequent DFT calculations using Jaguar v.8, 32,33 at the PBE/6-31+G(d) level of theory. 34,35 All subsequent geometry optimisations have been performed in Gaussian09 36 using the PBE0 hybrid functional, 37  The intrinsic reaction coordinate (IRC) scan 43,44 at the PBE-D3/6-31+G(d) level has been performed in Jaguar v.8, with 30 forward points and 8 backward points from an initial analytically calculated Hessian (0.1 Å step size).

Discussion
In the calculated transition state, the C-N-O(Ts) angle is 154.4°, consistent with an SN2-like mechanism. Note that the calculations reported herein did not test the viability of an SET mechanism. Further multireference ab initio calculations would need to be performed to investigate the possibility of an SET mechanism, but these are outside the scope of the current study. 99 In modelling the reaction of 5a-Me to 6a-Me the following approximations have been made to afford a computationally tractable problem. (1) The reactant arene is assumed to be deprotonated, and to be sufficiently stable to exist as the free ion in solution.
(2) Implicit solvation is sufficient to model the interaction between the reacting species and the solvent.
Clearly, this approach does not capture any hydrogen bonding interactions present in the real system. However, we have no reason to suggest that the stabilisation of the reactant and TS will be significantly different, thus the calculated barrier is unlikely to differ considerably if explicit solvent molecules were included in the calculation. Table S1. Energies for the species along the calculated energy pathway quoted in kcal mol -1 and relevant distances in Å.

IRC calculation
The IRC calculation has been performed to confirm that the TS links the reactant and product geometries. The pure GGA PBE functional was used in Jaguar to increase computational efficiency. Figure S1 depicts the relative potential (SCF) energy change over the IRC scan in addition to pertinent bond distances. See Table S2 for SCF and relative energies. The TS is very early along the reaction coordinate (d(C(para)-N) = 2.95 Å, 0.3 Å longer than that computed at the PBE0/6-31+G(d) level of theory), consistent with the highly exergonic reaction.