D. Enders et al.
to the corresponding primary amine (86%), which in turn
was converted by acylation into a camphanic amide and a p-
bromobenzamide in good yields and without racemisation
(for details on compounds 9–11, see Supporting Informa-
tion).
Based on our empirical data, as well as previously report-
ed studies,[7g,12,13,22] a mechanism is proposed to explain this
multicatalytic system (Scheme 2). Starting from C3,C2-un-
substituted indoles and ortho-alkyne-substituted nitrostyr-
Scheme 3. Indirect mechanistic evidence for the formation of spirocyclic
intermediates.
to effect an expansion from a
six- to a seven-membered ring
(15).[23] The final driving force
towards 16 lies in the indole
rearomatisation and protodeau-
ration step. It is worth noting
that our methodology allows a
three-carbon tether between
the C3 atom of indole and the
alkyne to undergo
a seven-
membered-ring formation by a
formal 7-endo-dig cyclisation.[8d]
This is in contrast to previous
reports, in which an intramolec-
ular Friedel–Crafts attack on a
three-carbon, aliphatic tether
resulted in six-membered rings,
formed by a formal 6-exo-dig
reaction.[12–14] One possible ex-
planation is that the presence
of a fused phenyl group on the
Scheme 2. Proposed mechanism of the sequential organo- and gold-based catalysis.
tether might stabilise the endo-
cyclic olefin better than an exo-
enes, 6c serves as a bifunctional organocatalyst, similar to a
thiourea catalyst, forming a tight transition state TS12,
which is stabilised by multiple hydrogen-bonding interac-
cyclic one, suggesting that thermodynamic factors could be
responsible for the observed regioselectivity.
In conclusion, we have developed a new methodology to
access tetracyclic indole derivatives enantioselectively by
merging organo- and gold-based catalysis through two se-
quential Friedel–Crafts type reactions. It is remarkable that
this one-pot protocol allows insertion of a stereogenic
centre into a highly conjugated, tetracyclic system, without
compromising the stereogenic integrity of the dibenzylic po-
sition. Moreover, a rare 7-endo-dig cyclisation provides
deeper insights into gold catalysis on indoles. As some ques-
tions regarding the exact dependence of the tethering bridge
and the mode of cyclisation remain unsolved, further mech-
anistic studies are currently ongoing in our laboratories.
À
tions with the nitroolefin and the N H of the attacking
indole nucleophile. This locks the reacting partners in fa-
vouring an indole Si-attack and explains the high enantiose-
lectivity and (R)-configuration in the first Friedel–Crafts
type reaction. Subsequently, the C3-substituted intermediate
13 is activated by the p-acidic [AuACTHNUTRGNEUNG(PPh3)]NTf2, owing to the
presence of the auraphilic alkyne. The alkyne activation
allows the C3 of indole to undergo a second Friedel–Crafts
type reaction in a 6-endo-dig fashion, forming a spirocyclic
intermediate 14.[12,13,22] A mechanistic test was conducted by
using our protocol on a C2-substituted indole (Scheme 3).
Indeed, a small amount of the spirocyclic product 18 was
isolated, albeit in low yield, possibly owing to the formation
of a highly congested quartenary centre. Nevertheless, the
successful isolation of this compound provides a plausible
evidence that spirocyclic intermediates, such as 14, could be
involved in the catalytic cycle. With the known character of
gold catalysis to engage in nonclassical cationic-type casca-
des,[7b,l] the intermediate 14 can rearrange through a 1,2-shift
Experimental Section
General procedure: Trifluomethanesulfonic acid (7.1 mL, 0.08 mmol,
0.1 equiv) was added under air to a Schlenk tube, containing a solution
of 6c (0.08 mmol, 0.1 equiv) in CHCl3 (4 mL). Then, NaBArF24
(0.08 mmol, 0.1 equiv) was added to the reaction mixture, which was
13412
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 13409 – 13414