Organic Letters
yields (14 and 15). More electron-rich aromatic nucleophiles,
such as dimethoxybenzene, underwent smooth coupling with a
variety of halogenated vinyl triflates in good yields (16−18).
There was a minimal decrease in efficiency when performing
the reaction on a 1 g scale with the iodinated vinyl triflate,
giving styrene 18 in 64% yield. Furthermore, cyclooctenyl
triflate 19 was observed to undergo a transannular C−H
insertion, Friedel−Crafts cascade with 4-methylanisole, giving
alkylated arene 20 in 57% yield (Figure 2b). Here two C−C
bonds, a 5,5-fused ring system, and a quaternary carbon center
were all forged in a single step. Notably, all of the reactions
outlined in Figure 2 were performed on the bench and
required neither scrupulous drying of substrates nor catalysts.
We then sought to validate our hypothesis that these readily
accessible organocatalysts were able to tolerate various
functional groups in the context of vinyl cation C−H insertion
reactions. To explore the functional group tolerance, a variety
Figure 4. Urea-catalyzed C−H insertion reactions of β-ketoester-
a
derived vinyl triflates. Isolated yield after column chromatography. 10
equiv of LiH.
candidates because they are readily accessible from corre-
sponding β-ketoesters, they would yield highly functionalized
cyclopentenes, and they are natively heteroatom-rich, provid-
ing a further test of the compatibility of this catalytic system.
Under the standard LiHMDS conditions, we found no
conversion to the desired product, likely due to the
electrophilicity of (vinylogous) esters as well as recent reports
15,16
describing ester functionalization with LiHMDS.
We found that using LiH as the base allowed for productive
transformations. Methyl, halogen substituents, boronic esters,
and methyl ethers were all tolerated, yielding the ester
products 26−29 in 41−64% yield. Notably, under these
basic conditions, acid-sensitive functional groups such as a
methoxymethyl (MOM) ether-protected phenol or tert-butyl
ester remained intact, yielding cyclized products 30 and 31 in
3
8 and 33% yield, respectively. We observed the exclusive
Figure 3. Urea- and lithium-catalyzed C−H insertion reactions.
a
formation of the β,γ-unsaturated products; we attribute this to
the increased stability of these products in comparison with the
α,β-unsaturated products, likely due to allylic strain.
Isolated yield after column chromatography. Yield determined by
b
c
NMR using an internal standard. Catalyst 4. LiHMDS base in
14
d
e
cyclohexane solvent. Catalyst 5. LiOtBu base in 1,2-DCE solvent.
With the information derived from our initial scope studies,
we began our investigation into the mechanism of this
transformation. During our studies of the vinylogous acyl
triflates, we consistently noticed small amounts of olefinic
products in our crude reaction mixtures. Careful purification of
the reaction mixture derived from tolyl triflate 32 provided γ-
lactone 33 in 16% yield (Figure 5a). We attribute the
formation of this byproduct to the intermediate vinyl cation 34
undergoing a 1,5-hydride shift, generating secondary carboca-
tion 35. This putative intermediate can then undergo a facile
1,2-hydride shift to yield secondary carbocation 36 followed by
trapping by the pendant ester, yielding the lactone product.
This overall “rebound”-type mechanism has been proposed by
Stang, Hanack, Olah, Mayr, Caple, and others. To further
support this mechanistic hypothesis, we synthesized propylated
triflate 37. Under the reaction conditions, neither the desired
insertion product nor lactone byproducts were observed
(Figure 5b). We attribute this to the inherent difficulty in
the formation of primary carbocation 38. These mechanistic
findings stand in stark contrast with those of previously
observed vinyl cation C−H insertion reactions of propylated
benzosuberonyl triflates (e.g., 1, Table 1) and previously
were quite pleased to find that a substrate bearing a pyridine
substituent was competent in this transformation, yielding
cyclopentenylpyridine 21 in 61% yield. Substrates bearing
electron-withdrawing substituents, however, resulted in prod-
1
4
ucts with poor olefin isomer ratios.
Upon further
optimization, we discovered that the utilization of LiH allowed
for high-yielding reactions with excellent olefin selectivity for
these substrates (22−24). Moreover LiOtBu was also a
competent base for this transformation, allowing for the
formation of dihydrofuran 25 in 61% yield, via insertion into
an ether tether. To the best of our knowledge, this example
showcases the first heterocycle synthesis from a C−H insertion
reaction of a vinyl cation. Furthermore, the variety of Li bases
used for these transformations highlights the modularity of this
system as well as the importance of both the hydrogen-
bonding catalyst and base.
17
Inspired by the successful synthesis of ester 23, we posited
that we could also form aryl cyclopentene derivatives via C−H
insertion reactions of vinylogous acyl triflates derived from
butylated β-ketoester (Figure 4). These substrates are good
C
Org. Lett. XXXX, XXX, XXX−XXX