Pd(II)-catalyzed regioselective 2-alkylation of indoles via a norbornene-mediated C-H activation: Mechanism and applications

Article abstract of DOI:10.1021/ja3058138

A palladium-catalyzed direct 2-alkylation reaction of free N-H indoles was developed based on a norbornene-mediated regioselective cascade C-H activation. The detailed reaction mechanism was investigated by NMR spectroscopic analyses, characterization of the key intermediate, deuterium labeling experiments, and kinetic studies. The results indicate that a catalytic cycle operates, in which an N-norbornene type palladacycle is formed as the key intermediate. Oxidative addition of alkyl bromide to the Pd(II) center in this intermediate is the rate-determining step of the reaction. The synthetic utility of this indole 2-alkylation method was demonstrated by its application in natural product total synthesis. A new and general strategy to synthesize Aspidosperma alkaloids was established employing the indole 2-alkylation reaction as the key step, and two structurally different Aspidosperma alkaloids, aspidospermidine and goniomitine, were synthesized in concise routes.

Full text of DOI:10.1021/ja3058138

Article
pubs.acs.org/JACS
Pd(II)-Catalyzed Regioselective 2‑Alkylation of Indoles via a
Norbornene-Mediated C−H Activation: Mechanism and Applications
Lei Jiao, Eberhardt Herdtweck, and Thorsten Bach*
Lehrstuhl fu
̈
r Organische Chemie I and Catalysis Research Center (CRC), Technische Universitat
̈
̈
Munchen, Lichtenbergstrasse 4,
85747 Garching, Germany
*
S Supporting Information
ABSTRACT: A palladium-catalyzed direct 2-alkylation reaction of free
N-H indoles was developed based on a norbornene-mediated
regioselective cascade C−H activation. The detailed reaction mechanism
was investigated by NMR spectroscopic analyses, characterization of the
key intermediate, deuterium labeling experiments, and kinetic studies. The
results indicate that a catalytic cycle operates, in which an N-norbornene
type palladacycle is formed as the key intermediate. Oxidative addition of
alkyl bromide to the Pd(II) center in this intermediate is the rate-
determining step of the reaction. The synthetic utility of this indole 2-
alkylation method was demonstrated by its application in natural product
total synthesis. A new and general strategy to synthesize Aspidosperma alkaloids was established employing the indole 2-alkylation
reaction as the key step, and two structurally different Aspidosperma alkaloids, aspidospermidine and goniomitine, were
synthesized in concise routes.
INTRODUCTION
Scheme 1. Palladium-Catalyzed Direct 2-Alkylation of Indole
and Originally Proposed Reaction Mechanism
■
The direct C−H functionalization of heterocycles has attracted
increasing research interest due to the structural prevalence of
substituted heterocycles in natural products, drugs, and other
1
biologically active molecules. As a consequence, many elegant
methods to regioselectively substitute a heteroaromatic C−H
bond by an aryl or alkenyl group have been established by
2
employing transition-metal catalysis. However, protocols for
the alkylation of a heteroaromatic C−H bond are still limited to
3
date. Therefore, the development of direct C−H alkylation
methods for heterocycles, especially of those which exert
regioselectivities different from the simple Friedel−Crafts
alkylation, is still highly demanded.
As a step toward this goal, we have recently developed a
palladium-catalyzed direct 2-alkylation reaction of free N-H
indoles by invoking a norbornene-mediated cascade C−H
4
activation (Scheme 1). In the presence of a Pd(II) catalyst,
norbornene, and a base, an indole substrate reacts with a
primary alkyl bromide to form a 2-alkyl substituted indole
product regioselectively. The transformation tolerates a wide
variety of functional groups on both components. Both electron
rich and electron deficient indoles have been shown to be
suitable substrates. N,N-Dimethylacetamide (DMA) has
evolved as the optimal solvent and the addition of water has
a positive effect on the reaction outcome. Various Pd(II)
catalysts, such as Pd(OAc) and PdCl (MeCN) , have shown
similar catalytic activity in this reaction and no additional ligand
is required. This reaction provides a straightforward route to 2-
alkylated indole derivatives starting from simple substrates
under mild and practical conditions.
The 2-alkylation process was originally assumed to proceed
according to the mechanism shown in the lower part of Scheme
4
1
. It was speculated that the reaction is initiated by the well-
5
established C3-palladation of indole and that a subsequent
insertion of norbornene into the C3Pd bond occurs to form
6,7
intermediate IN . As discovered by Catellani and co-workers,
1
the Pd(II) center was thought to activate under basic
conditions the proximal C2H bond by electrophilic
2
2
2
palladation to transform intermediate IN to a C3-norbornene
1
Received: June 15, 2012
Published: August 22, 2012
©
2012 American Chemical Society
14563
dx.doi.org/10.1021/ja3058138 | J. Am. Chem. Soc. 2012, 134, 14563−14572

Journal of the American Chemical Society
Article
type palladacycle IN , which after alkylation produces via 2-
norbornene insertion. To identify possible intermediate(s)
involved in the catalytic process, NMR spectroscopic studies of
the reaction mixture were conducted. The model 2-alkylation
reaction of indole with BuBr was investigated. To simplify the
reaction system for NMR analysis, an initial reaction was
performed without addition of the alkylating reagent. A mixture
2
alkyl-3-indolylpalladium species IN the observed 2-alkylindole
3
product together with the Pd(II) catalyst. The combination of
the indole-palladium chemistry and the Catellani reaction
allows for a successful cascade C−H activation cycle that
regioselectively alkylates the less reactive C2 position of indole.
As an extension of this work, we sought to further study the
detailed reaction mechanism of the Pd(II)-catalyzed direct 2-
alkylation of indole. Our research along these lines led to the
conclusion that the reaction does not follow the above-
mentioned proposed mechanism. Rather our studies suggest a
reaction mechanism, in which an N-norbornene type pallada-
8
of PdCl (MeCN) , indole, norbornene, KBr, and K CO in
2
2
2
3
DMA with 0.5 M H O was allowed to react at 70 °C for 1 h.
2
The resulting reaction mixture, after filtration, was directly
1
subjected to H NMR analysis. It was found that a new set of
indole-based aromatic signals were generated (Figure 1a,b, red
cycle, but not a C3-norbornene type palladacycle IN , acts as
2
the key intermediate. Therefore, a revised catalytic cycle is
presented and the rate-determining step in the cycle was
assigned by kinetic studies. In addition to the mechanistic
investigations, we have exploited the synthetic utility of the
indole 2-alkylation reaction by applying it to natural product
total synthesis. Two indole alkaloids, goniomitine and
aspidospermidine, were synthesized using the present reaction
as key step. A new and concise strategy for the synthesis of
Aspidosperma alkaloids has thus been established.
MECHANISTIC STUDIES
■
Motivation for the Mechanistic Study. To further
explore the substrate compatibility of the indole alkylation
reaction, we tested the reaction of N-methylindole (1) and 3-
methylindole (2) with n-BuBr under the standard reaction
conditions. It was found that N-methylindole failed to give any
2
-alkylation product and remained intact after 14 h. In stark
1
Figure 1. 500 MHz H NMR spectra at room temperature in DMA
contrast, 3-methylindole exhibited unexpected reactivity under
identical conditions and produced the 2-alkylation product 3 in
a moderate yield (Scheme 2). Control experiment showed that
this reaction did not proceed in the absence of norbornene.
with 0.5 M H O. Red peaks refer to the newly generated species in the
reaction system. Conditions: (a) indole; (b) indole (0.053 M),
PdCl (MeCN) (0.057 M), norbornene (0.40 M), KBr (0.054 M),
and K CO (0.41 mmol·mL ) after 1 h at 70 °C; (c) indole (0.11 M),
PdCl (MeCN) (0.042 M), norbornene (0.19 M), BuBr (0.23 M), and
K CO (0.39 mmol·mL ) after 50 min at 70 °C; (d) indole (0.20 M),
PdCl (MeCN) (0.024 M), norbornene (0.41 M), BuBr (0.47 M), and
2
2
2
−1
2
3
2
2
−1
Scheme 2. Reactions of N- and 3-Methylindole
2
3
2
2
−1
K CO (0.40 mmol·mL ) after 1.5 h at 70 °C.
2
3
signals). The new signals included four benzene C−H signals
and the C3−H signal at 6.05 ppm, but signals corresponding to
N-H and C2−H were lacking. New aliphatic peaks with similar
intensity also emerged (not shown in the spectrum).
1
Subsequently, a crude H NMR spectrum of the alkylation
reaction mixture in the presence of BuBr was recorded (Figure
1
4
c). To ensure a sufficient intensity of the intermediate peaks,
0 mol % of PdCl (MeCN) catalyst was used. The spectrum
2
2
showed the existence of the same intermediate, which was the
major intermediate species during the alkylation reaction.
Finally, H NMR analysis of the crude reaction mixture with a
1
The previously proposed reaction mechanism (Scheme 1)
could not account for the experimental observations. According
to the proposed mechanism, the N−H bond of indole is not
involved in the reaction; thus, it is expected that 1 undergoes
the same alkylation reaction as free N-H indole. On the other
hand, the proposed mechanism requires C3-palladation of
indole as the initiation step, which is incompatible with a 3-
substituted indole. The successful alkylation of 2 contradicted
this mechanistic interpretation. These results questioned the
proposed catalytic cycle and prompted us to conduct detailed
mechanistic studies on this reaction.
normal catalyst loading (12 mol %) showed the same
intermediate peaks (Figure 1d) albeit with lower intensity.
The NMR analyses seemed to indicate that indole,
norbornene, and the palladium catalyst form a primary
intermediate without participation of the alkylating reagent
(Figure 1b). This intermediate appears to be the major
intermediate throughout the reaction (Figure 1c,d) and seems
to play a role in the following alkylation steps. The NMR
spectra suggested that the intermediate was a norbornene-
embedded C2-palladation product (due to the absence of the
C2−H signal and appearance of new aliphatic signals) different
NMR Spectroscopic Study. The reactions employing
indoles 1 and 2 suggested that the indole alkylation does not
proceed via intermediates arising from a C3-palladation/
from the previously proposed intermediate IN (due to the
2
absence of the N-H signal and the presence of the C3−H
1
4564
dx.doi.org/10.1021/ja3058138 | J. Am. Chem. Soc. 2012, 134, 14563−14572

Journal of the American Chemical Society
Article
signal). Further efforts to separate and characterize this
intermediate were undertaken to elucidate its structure.
Separation and Characterization of the Key Pallada-
cycle Intermediate. Although the primary intermediate could
be directly observed by NMR, attempts to characterize it in
pure form were unsuccessful due to purification problems. After
an aqueous workup procedure and removal of the solvent, a
dark-brown solid was obtained, which proved to be a complex
mixture by NMR analysis. Attempts to purify this product by
recrystallization or column chromatography failed. It seemed
that the observed intermediate aggregated during the workup
procedure, possibly due to weak ligands on the palladium
center (MeCN, DMA, H O, etc.). To solve this problem, we
2
tried to apply a bidentate ligand to form a stable and separable
monomeric complex.
It was found that, after the reaction of indole, norbornene,
Figure 2. ORTEP style plot of compound 5·CDCl in the solid state.
3
Thermal ellipsoids are drawn at the 50% probability level. CDCl is
3
and Pd(OAc) , upon addition of 1,2-bis(diphenylphosphino)-
2
omitted for clarity. Selected bond lengths (Å) and bond angles (°):
Pd1−N2 2.1239(13), Pd1−N3 2.1471(13), Pd1−C1 2.0392(16),
Pd1−C15 1.9774(15); N2−Pd1−N3 78.47(5), C1−Pd1−C15
82.60(6).
ethane (dppe) or 1,10-phenanthroline, new complexes 4 and 5
formed cleanly. These complexes were obtained in pure form
after column chromatography (Scheme 3) and could be
Scheme 3. Synthesis of Palladacycle Complexes
elimination, and subsequent steps to furnish the 2-alkylindole
product, which are similar to the steps in the Catallani
7
reaction. A palladacycle of the IN type was not observed and
2
is probably not involved in the reaction.
Intramolecular Trapping of the Intermediate. Pre-
2
viously we found that the C(sp )−I bond in an aryl iodide
4
could be employed to generate a 2-arylated indole derivative.
Much to our surprise, in a test reaction employing β-
iodostryene (6) as the electrophile, a polycyclic indole
derivative 7 with an embedded norbornene moiety was isolated
(
Scheme 4) instead of the expected 2-alkenylation product.
Scheme 4. Intramolecular Heck-Trapping Reaction
analytically characterized. Two-dimensional NMR spectra and
HRMS analyses established the palladacycle structure of the
two complexes, in which the indole nitrogen atom is bound to
the norbornene moiety and the C3 carbon atom remains
unsubstituted. This structure assignment was unambiguously
confirmed by an X-ray single crystal structure of the 1,10-
9
phenanthroline complex 5 (Figure 2). On the basis of these
observations, the intermediate species observed by NMR was
tentatively assigned structure IN . The structure is in agreement
4
1
with the recorded H NMR spectrum, in which a C3−H signal
was observed but N-H and C2−H signals could not be detected
(
Figure 1).
The results, together with the NMR spectroscopic studies,
The product is believed to be formed by an intramolecular
Heck-type reaction following a normal norbornene-mediated 2-
alkenylation process. A plausible mechanism for its formation is
depicted in Scheme 4. The conversion starts with the formation
of palladacycle IN . Subsequently, intermediate IN and
indicated that the N-norbornene type palladacycle IN , rather
4
than the previously proposed C3-norbornene type palladacycle
IN , was the major palladium species during the catalytic
2
process. This N-norbornene type palladacycle is assumed to be
generated by a cascade C−H activation process involving syn-
aminopalladation of norbornene and subsequent ortho-C−H
activation.
4
4
iodoalkene 6 undergo oxidative addition and reductive
elimination to form the 2-alkenylated species IN . Because of
5
the close proximity of the norbornene−Pd bond and the newly
attached CC bond, an intramolecular carbopalladation
occurs to generate the pentacyclic palladium intermediate
If the in situ generated palladacycle IN was treated with
4
BuBr instead of a bidentate ligand at 70 °C, 2-butylindole was
generated in 53% yield within 0.5 h. This result together with
the results presented in Scheme 2 support a reaction
IN . Subsequent β-hydride elimination leads to product 7.
6
The formation of the intramolecular trapping product 7
provides additional evidence for the involvement of nitrogen-
norbornene type intermediates in the catalytic process. The N-
norbornene type palladium intermediates can be observed at an
mechanism, in which IN acts as the key active palladacycle
4
intermediate in the 2-alkylation process. It was assumed that
IN reacts with alkyl bromide via oxidative addition, reductive
4
1
4565
dx.doi.org/10.1021/ja3058138 | J. Am. Chem. Soc. 2012, 134, 14563−14572

Journal of the American Chemical Society
Article
early stage (i.e., before C2-functionalization) and trapped at a
late stage of the reaction (i.e., after the C2-functionalization
step), implying that this type of intermediates are involved
throughout the catalytic cycle and play a key role in the
Scheme 6. Competition Experiment for Determining a
Primary KIE
reaction. Therefore, this result further supported that IN is the
4
key palladacycle intermediate reacting with alkyl bromide to
achieve the 2-alkylation on indole.
Deuterium Labeling Experiment. At this stage, further
attention was paid to the key palladacycle intermediate IN . A
4
reaction employing 2-deuterated indole was conducted to
probe the reversibility of the norbornene-mediated cascade C−
H activation leading to IN . If the formation of IN was
of the Catellani chemistry, in which the cleavage of the C−H
7d
4
4
bond is also not rate-determining.
involved in an equilibrium, the reverse reaction from IN to
4
A Revised Catalytic Cycle. Taking the results of the above
mechanistic studies together, it is evident that the norbornene-
mediated 2-alkylation of indole does not proceed via the
originally proposed mechanism (Scheme 1), and a revised
catalytic cycle can be drawn (Scheme 7). This catalytic cycle
indole would decrease the C2-deuterium incorporation of the
remaining indole substrate. Alternatively, an irreversible
formation of IN would not decrease the percentage of C2-
4
deuterium incorporation in the unconverted indole substrate.
In a control experiment, treatment of 2-D-indole with a
catalytic amount of PdCl (MeCN) in DMA (0.5 M H O) at
2
2
2
Scheme 7. Revised Catalytic Cycle
7
0 °C in the presence of K CO for 14 h did not result in loss
2 3
of deuterium at carbon atom C2 of indole, indicating that there
is no reversible direct C2-palladation of indole by the Pd(II)
catalyst (Scheme 5). In a second experiment, the reaction of 2-
Scheme 5. Reactions of 2-Deuteroindole
D-indole with butyl bromide was conducted under standard
conditions. The reaction was stopped after 3 h, when 39% of
the 2-butylation product was generated at ca. 70% conversion
1
(
Scheme 5). The recovered indole substrate, as shown by H
NMR, had 96% 2-deuterium incorporation, indicating no loss
of 2-deuterium within the limits of experimental error. This
observation is consistent with an irreversible formation of
consists of two major stages: (1) irreversible formation of
palladacycle IN by norbornene-mediated C−H activation and
4
(2) reaction of IN and the alkyl bromide to complete the C2-
4
palladacycle IN by an norbornene-mediated cascade C−H
alkylation of indole. Stage 1 is supported by the NMR
spectroscopic study, by the synthesis and characterization of the
palladacyclic complexes, and by the deuterium labeling
4
activation.
Kinetic Isotope Effect. By employing 2-deuteroindole as
the substrate, the kinetic isotope effect (KIE) of the 2-alkylation
reaction was measured to gain some insight into the cascade
C−H activation step. Since deuterium labeling experiments
experiments. The palladacycle intermediate IN is believed to
4
be formed by N-palladation of indole (i.e., N−H activation of
indole), syn-aminopalladation of norbornene, and ortho-C−H
palladation via intermediates IN and IN , respectively. Stage 2
proved the formation of palladacycle intermediate IN to be
4
7
8
irreversible, a competition experiment using equal molar
amounts of indole and 2-deuteroindole was expected to
provide a potential primary KIE for the formation of
palladacycle IN (Scheme 6). The k /k value was obtained
of the mechanism is evidenced by the reaction of the in situ
generated palladacycle IN4 with BuBr, the intramolecular
trapping reaction (formation of 7), and by the observed
reactivity of 3-substituted indole in this 2-alkylation reaction.
4
H
D
1
7d
by H NMR analysis of unconverted indole starting material. It
was found that the recovered indole showed a 51% 2-deuterium
incorporation, indicating that almost no primary KIE was
observed for this reaction. The lack of a primary KIE is
consistent with a palladation mechanism, in which the C−H
bond cleavage does not take place during the rate-determining
Similar to the Catellani reaction, the alkylation process is
believed to take place through oxidative addition of the alkyl
bromide to the Pd(II) center to generate Pd(IV) intermediate
IN , reductive elimination on the Pd(IV) center to form IN ,
9
10
and norbornene expulsion to form 2-alkyl-N-palladaindole
IN . Hydrolysis of IN delivers the final 2-alkylindole product
1
1
11
1
0
step. This is in agreement with the ortho-C−H activation step
and regenerates the Pd(II) catalyst.
1
4566
dx.doi.org/10.1021/ja3058138 | J. Am. Chem. Soc. 2012, 134, 14563−14572

Journal of the American Chemical Society
Article
Kinetic Studies. With the reasonable catalytic cycle in
mind, we conducted kinetic studies on a model indole
alkylation reaction (Scheme 8) to probe the concentration
Scheme 8. Model Reaction for Kinetic Studies
Figure 3. Initial rate as a function of [indole]. Conditions: [indole] =
dependence of primary reaction components (indole, BuBr,
norbornene, and palladium catalyst). It was hoped that kinetic
studies would further support the mechanism and would also
provide further insight into the nature of the proposed key
intermediate IN4.
0.1−0.8 M, [PdCl
2
(MeCN)
2
] = 8.0 mM, [BuBr] = 0.40 M,
3
−
1
[
norbornene] = 0.40 M, K CO 0.40 mmol·mL , DMA with 0.5 M
2
H O, 70 °C.
2
The progress of the model reaction was monitored by gas
chromatography, and the obtained reaction time-course
revealed monotonic increase in the concentrations of both
product 8 and overalkylation byproduct 9. The lack of an
induction period enabled us to obtain kinetic parameters by the
11
initial-rate method. Since both product 8 and byproduct 9
were generated from the beginning and an independent
reaction showed that 8 can be converted to 9 under the
12
reaction conditions, we took the total concentration of 8 and
as product concentration. The plot of ([8] + [9]) versus time
9
showed a linear relationship during the initial period and the
initial rate of the 2-alkylation was derived from the slope of this
line.
Concentration Dependence of the Initial Rate on Each
Reaction Component. The order of each component
employed in the alkylation reaction was determined by the
Figure 4. Initial rate as a function of [BuBr]. Conditions: [BuBr] =
(MeCN) ] = 8.0 mM, [indole] = 0.20 M,
norbornene] = 0.40 M, K CO 0.40 mmol·mL , DMA with 0.5 M
0.2−1 M, [PdCl
2
2
−
1
[
2
3
H O, 70 °C.
2
initial-rate method. Since the reaction is heterogeneous (K CO3
2
is only slightly soluble in the reaction media), we first measured
the initial rates of the model reaction with different amounts of
K CO (0.2−0.6 mmol/mL reaction volume, i.e., 1−3 equiv.
2
3
versus indole). It was found that the initial rate did not change
as a function of the K CO loading, indicating that the phase
2
3
transfer of base is not rate-determining (see Supporting
Information for details). This allowed us to gain useful
mechanistic information from the rate-dependence measure-
ments.
The dependence of the alkylation rate on [indole] was found
to be of negative order, indicating that indole has an inhibition
effect on the reaction (Figure 3). The nonlinear dependence fits
to an equation describing off-cycle binding interaction between
indole and the active palladium species (vide infra).
The dependence of the initial rate on the concentration of
both [BuBr] and [PdCl (MeCN) ] was found to be first-order
Figure 5. Initial rate as a function of [Pd]. Conditions:
[PdCl (MeCN) ] = 2−24 mM, [indole] = 0.20 M, [BuBr] = 0.40
M, [norbornene] = 0.40 M, K CO 0.40 mmol·mL , DMA with 0.5
2
2
−1
2
3
2
2
M H O, 70 °C.
(
Figures 4 and 5). However, the rate dependence on [Pd]
2
deviates from first-order at higher [Pd] (16−24 mM, the order
13
of [Pd] was found to be ca. 0.5). The dependence of the
alkylation rate on [norbornene] was found to be zero-order
is not the formation of IN , but the reactions involving IN and
4 4
the alkyl bromide (oxidative addition or reductive elimination).
Although the present kinetic data do not provide direct
evidence to support either step as rate-determining, the
(
Figure 6).
The above-mentioned kinetic results support a rate-
determining step in which both the Pd catalyst and alkyl
bromide are involved and they rule out the C−H activation as
the rate-determining step. Given the revised catalytic cycle
proposed in Scheme 7, it is clear that the rate-determining step
oxidative addition of IN is more likely to be rate-determining
4
based on the following considerations: (1) it is generally
accepted in organopalladium(IV) chemistry that the reductive
14
elimination on a Pd(IV) center is facile; (2) experiments
1
4567
dx.doi.org/10.1021/ja3058138 | J. Am. Chem. Soc. 2012, 134, 14563−14572

Journal of the American Chemical Society
Article
the key intermediate. A rate-determining oxidative addition of
the alkyl bromide to the Pd(II) center in IN is suggested by
4
kinetic studies. To the best of our knowledge, there has so far
not been any clear evidence for a palladaheterocycle in
7
,18
Catellani-type alkylation reactions.
The present reaction
represents a new reaction category among these reactions. As
disclosed in the mechanistic study, the ability of the present
catalytic system to achieve 2-alkylation on a 3-substituted
indole greatly extends its synthetic power (Scheme 2). This
feature enables a new access to 2,3-disubstituted indole
derivatives and was applied to indole alkaloid synthesis.
SYNTHETIC APPLICATIONS
■
Figure 6. Initial rate as a function of [norbornene]. Conditions:
Indole alkaloids compose an important class of natural products
19
[
0
norbornene] = 0.1−0.8 M, [PdCl (MeCN) ] = 8.0 mM, [indole] =
2
2
due to their unique structures and biological activities. To
date, tremendous efforts have been devoted to the total
synthesis of indole alkaloids, and many elegant strategies have
been developed to construct their core structure. The direct
-alkylation reaction of indoles enables a direct access to 2-
−1
.20 M, [BuBr] = 0.40 M, K CO 0.40 mmol·mL , DMA with 0.5 M
2 3
H O, 70 °C.
2
20
showed that bulkier alkyl bromides (e.g., iso-butyl bromide)
exhibit a decreased alkylation reactivity compared with n-butyl
bromide, which is consistent with a mechanism, in which
oxidative addition is rate-determining and reductive elimination
is not.
2
alkylindole derivatives from indole and an alkyl bromide.
Therefore, it was expected that a new and more straightforward
synthetic strategy could be realized by applying this method to
indole alkaloid synthesis.
4
The intriguing initial-rate dependence on [indole] might be
understood by considering indole as an inhibitory ligand.
Negative dependence of a primary substrate in a catalytic
reaction is uncommon, and this type of substrate inhibition
often suggests off-cycle binding between the substrate and the
catalytically active species that decreases the effective catalyst
Total Synthesis of Aspidospermidine. The Aspidosperma
21
family is one of the largest groups of indole alkaloids. As a
representative member of this alkaloid family, aspidospermidine
has attracted attention from the synthetic community since the
22
1
960s due to its unique pentacyclic skeleton with four
contiguous stereocenters. To date, a number of successful
synthetic routes have been established, and the general
15
concentration within the cycle.
In this reaction, it is reasonable to attribute the observed
strategies to construct the multisubstituted dihydroindole
substrate inhibition to the off-cycle binding between IN and
22b,23c,f−h,24j,m−r
4
core include Fischer indole synthesis,
Spengler reaction/rearrangement,
Pictet-
indole. A kinetic model can be tentatively established based on
this assumption and on the revised catalytic cycle (Scheme 9).
23a,b,i,24a
and intramolecular
23d,e,24c,h,k,l
cyclization of an aniline-type intermediate.
With the
indole 2-alkylation method in hand, we planned to develop an
approach to this alkaloid family by a direct coupling of the
indole core and the alkyl subunit.
Scheme 9. Kinetic Model of the Reaction
Retrosynthetic analysis suggested that the AB rings of
aspidospermidine can originate from indole (Scheme 10). For
the construction of the CDE ring system, the E ring was
planned to be created by base-induced cyclization of alcohol 11
Scheme 10. Retrosynthetic Analysis of Aspidospermidine
A rate law derived from this model accounts for the observed
16
kinetic features of this reaction (eq 1), and the nonlinear fit of
−1
the curve shown in Figure 4 gives a K value of 1.3 ± 0.1 M .
The small equilibrium constant suggests that indole acts as a
weak inhibitory ligand in the alkylation reaction.
k[RCH Br][Pd]
K[indole] + 1
2
total
rate = k[IN ][RCH Br] =
4
2
(1)
The decreased order of [Pd] at high catalyst concentration
suggests the active Pd species may aggregate or decompose
during the reaction, as observed in other Pd-catalyzed
1
7
reactions. However, at present it is not possible to give a
clear picture of this process nor to quantify it.
To summarize the mechanistic section, a revised catalytic
cycle for the direct indole 2-alkylation reaction was established
with an emphasis on N-norbornene type intermediate IN as
4
1
4568
dx.doi.org/10.1021/ja3058138 | J. Am. Chem. Soc. 2012, 134, 14563−14572

Journal of the American Chemical Society
Article
after conversion of the hydroxyl group to a good leaving
Scheme 12. Completion of the Synthesis
2
3d,24e
group.
Alcohol 11 was supposed to be obtained from
hemiaminal 13 by an acid-catalyzed rearrangement via cyclic
2
4e,25
iminium ion 12,
with contemporary formation of the C
and D rings. Hemiaminal 13 could be derived from α-allyl-δ-
lactam 14 by functionalization of the CC bond and reduction
of the lactam carbonyl group. Finally, lactam 14 was planned to
be synthesized by base-induced cyclization and α-allylation of
indole derivative 15, which can be easily prepared from indole
and alkyl bromide 16 by the indole 2-alkylation reaction.
The synthesis of aspidospermidine commenced with the 2-
alkylation of indole with bromide 16 under standard conditions
catalyzed skeletal rearrangement of the δ-lactam-derived
hemiaminal intermediate.
26
The present route also represents formal syntheses of
23j
24o
(
±)-vincadifformine
and (±)-quebrachamine,
as both
can be derived from imine 19 in one step by established
(
Scheme 11). The 2-alkylindole product 15 was obtained in
methods.
Total Synthesis of Goniomitine. In addition to the C2-
alkylation of an unsubstituted indole, the indole 2-alkylation
method allows also to install an alkyl group to the C2-position
of a 3-substituted indole (Scheme 2). This discovery greatly
extends the synthetic utility of the present method, and
therefore provides high potential for the total synthesis of
another Aspidosperma alkaloid, goniomitine (20). Isolated in
a
Scheme 11. Synthesis of Key Aminoalcohol 11
1
987, goniomitine is a significantly unique member of the
Aspidosperma alkaloid family as it exhibits a rare
octahydroindolo[1,2-a][1,8]naphthyridin scaffold and a 2-
hydroxyethyl substituent at the C3-position of its indole
3
1
core. Goniomitine was found to display a promising
antiproliferative activity in several tumor cell lines, and four
3
2
total syntheses have been accomplished to date.
Our retrosynthetic analysis suggested a short sequence
33
starting from O-TBS-protected tryptophol by using the direct
-alkylation method (Scheme 13). The aminal functionality was
2
Scheme 13. Retrosynthetic Analysis of Goniomitine
a
9-BBN = 9-borabicyclo[3.3.1]nonane.
6
5% yield. Treatment of intermediate 15 with lithium
hexamethyldisilazide (LiHMDS) and quenching the generated
lactam enolate with allyl bromide afforded α-allyl-δ-lactam 14
2
7
in high yield. A hydroboration-oxidation sequence of
compound 14 gave alcohol 17, which was oxidized to aldehyde
28
1
8 with Dess-Martin periodinane (DMP). This aldehyde
intermediate was transformed to the key aminoalcohol 11 by a
three-step sequence. First, the 2-hydroxyethylamino group was
installed by reductive amination of aldehyde 18 with ethanol-
amine and NaBH . Second, treatment with diisobutylaluminum
4
hydride (DIBAl-H) at low temperature reduced the lactam
29
carbonyl group to give hemiaminal 13, and finally, under mild
acidic conditions, 13 was converted to the key aminoalcohol 11
in a diastereoselective fashion by acid-catalyzed rearrangement.
Following this route, the ABCD ring system was established in
seven steps.
planned to be created by acid-catalyzed cyclization of
hemiaminal intermediate 21, which can be obtained by
reduction of azide 22. The azide functional group was supposed
Aminoalcohol 11 was treated with MsCl in the presence of
27
Et N and then with t-BuOK to promote the intramolecular
to be installed by hydroboration-oxidation and Mitsunobu
3
2
3d
34
cyclization reaction.
In this manner, the E ring of
reaction at the terminal end of the allyl group in lactam 23.
aspidospermidine was created and the pentacyclic system was
Similar to the synthesis of aspidospermidine, α-allyl-δ-lactam
intermediate 23 was planned to be derived from indole
derivative 24. The key step was the 2-alkylation reaction of
tryptophol derivative 25 with alkyl bromide 16.
When we set out to conduct the planned 2-alkylation
reaction of TBS-tryptophol 25 using alkyl bromide 16, it was
found that the reaction under standard conditions was rather
sluggish. The reaction outcome indicated that the 3-substituted
indole derivative was less reactive than indole itself and the
30
established to give imine 19. The final step in the synthesis
23f
was accomplished by reducing imine 19 with NaBH4,
affording aspidospermidine in a moderate overall yield from
tetracyclic intermediate 11 (Scheme 12). The total synthesis of
racemic aspidospermidine was accomplished in nine steps from
indole and bromide 16 and in an overall yield of 15%. The
brevity of this synthetic route is largely attributed to the
application of direct 2-alkylation of indole and the acid-
1
4569
dx.doi.org/10.1021/ja3058138 | J. Am. Chem. Soc. 2012, 134, 14563−14572

Journal of the American Chemical Society
Article
a
original conditions were no longer suitable. Therefore, we
replaced alkyl bromide 16 with the potentially more electro-
philic iodide 26 and conducted an optimization study to
achieve appropriate reaction conditions for the 2-alkylation of
the 3-substituted indole substrate 25 (Table 1).
Scheme 14. Synthesis of Goniomitine
Table 1. Optimization Study for the 2-Alkylation of O-TBS-
a
Tryptophol
a
DIAD = diisopropyl azodicarboxylate.
a
Conditions: 25 (1 equiv), 26 (4 equiv), norbornene (2 equiv),
K CO (4 equiv), 10 mol % Pd(II) catalyst, 0.5 M H O, 60 °C.
2
3
2
b
c
Calculated based on recovered starting material 25. Yield of isolated
transformation was realized by establishing a palladium-
catalyzed norbornene-mediated cascade C−H activation of N-
H indole derivatives. Initial synthetic studies revealed the scope
and limitations of this reaction, and simultaneously triggered
the mechanistic study to elucidate the reaction mechanism.
Combining the information obtained from NMR studies,
intermediate characterization, intramolecular trapping reaction,
isotope labeling experiments, and kinetic studies, the previously
proposed catalytic cycle was revised. The new catalytic cycle
features an N-norbornene type palladacycle as the key
intermediate and the oxidative addition of alkyl bromide to
the Pd(II) center in this intermediate as the rate-determining
step. The mechanistic findings render this reaction a yet
unknown reaction type in the Catellani chemistry. The utility of
this indole 2-alkylation method was demonstrated by total
syntheses of two structurally different Aspidosperma alkaloids,
aspidospermidine and goniomitine. The indole 2-alkylation
reaction was the key step in both syntheses, and the
construction of the molecular skeleton by an acid-catalyzed
rearrangement of a hemiaminal intermediate also contributed
to the conciseness of both routes. A new and general strategy to
synthesize the Aspidosperma alkaloid has been established.
product 24 after column chromatography.
It was found that DMF was more suited than DMA as the
solvent for this reaction, since the latter resulted in minor
amounts of inseparable impurities in the product 24.
Preliminary experiment showed that the reaction did not
reach full conversion after two days and only a low yield of 2,3-
disubstituted indole product 24 was obtained (entry 1).
Assuming undesired Pd(0) formation may be the reason for
the slow conversion, we applied an oxygen atmosphere and
found that both conversion and yield increased (entry 2). A
5e
mixed DMF−DMSO solvent system was found beneficial
(
entry 3) and PdCl as catalyst turned out to be slightly better
2
suited than PdCl (MeCN) (entry 4). Finally, an air
2
2
atmosphere was found to be optimal and the yield of 2,3-
disubstituted indole 24 was increased to 73% in a reaction time
of 26 h (entry 5).
Starting from key intermediate 24, alcohol 27 was
synthesized following a two-step sequence identical to the
preparation of alcohol 17 in the total synthesis of
aspidospermidine (Scheme 14). Mitsunobu reaction of alcohol
34
2
7 with diphenylphosphoryl azide (DPPA) gave azide 22.
ASSOCIATED CONTENT
Supporting Information
■
Treatment of 22 with LiAlH resulted simultaneously in the
4
*
S
reduction of the lactam to the respective hemiaminal and in the
conversion of the azide to a primary amine. Finally, the
generated intermediate 21 was treated under mild acidic
conditions to promote the cyclic aminal formation as well as
the deprotection of the silyl ether. Goniomitine was obtained
diastereoselectively in high yield. Following this sequence, the
total synthesis of racemic goniomitine was accomplished in
seven steps from tryptophol and iodide 26 in an overall yield of
Experimental procedures, spectroscopic data, and NMR spectra
for synthetic intermediates. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
thorsten.bach@ch.tum.de
■
35%.
Notes
The authors declare no competing financial interest.
CONCLUSION
■
ACKNOWLEDGMENTS
L.J. acknowledges the Alexander von Humboldt foundation for
a postdoctoral fellowship.
The regioselective direct 2-alkylation of the indole core by alkyl
halides is highly demanded because 2-alkylindole derivatives
serve as useful synthetic intermediates and building blocks. This
■
1
4570
dx.doi.org/10.1021/ja3058138 | J. Am. Chem. Soc. 2012, 134, 14563−14572

Journal of the American Chemical Society
Article
halophenyl)-1H-pyrrole and norbornene, see: Hulcoop, D. G.;
Lautens, M. Org. Lett. 2007, 9, 1761.
REFERENCES
■
(
1) Pozharskii, A. F.; Soldatenkov, A. T.; Katrizky, A. R. Heterocycles
in Life and Society; Wiley: New York, 1997.
(19) (a) Rahman, A.-ur.; Basha, A. Indole Alkaloids; Harwood:
Amsterdam, 1997. (b) Gribble, G. W. Part B: The Indole Alkaloids;
Second Supplements to the 2nd ed. of Rodd’s Chemistry of Carbon
Compounds; Elsevier: Amsterdam, 1997; Vol. 4, pp 69−164.
(
2) Selected recent reviews on direct CH functionalization of
heterocycles: (a) Lewis, J. C.; Bergman, R. G.; Ellman, J. A. Acc. Chem.
Res. 2008, 41, 1013. (b) Ackermann, L.; Vicente, R.; Kapdi, A. R.
Angew. Chem., Int. Ed. 2009, 48, 9792. (c) Colby, D. A.; Bergman, R.
G.; Ellman, J. A. Chem. Rev. 2010, 110, 624. (d) Fagnou, K. Top. Curr.
Chem. 2010, 292, 35. (e) Beck, E. M.; Gaunt, M. J. Top. Curr. Chem.
(20) Recent reviews on indole alkaloid synthesis: (a) Lewis, S. E.
Tetrahedron 2006, 62, 8655. (b) Edwankar, C. R.; Edwankar, R. V.;
Namjoshi, O. A.; Rallapalli, S. K.; Yang, J.; Cook, J. M. Curr. Opin.
Drug Discovery Dev. 2009, 12, 752. (c) Higuchi, K.; Kawasaki, T. Nat.
Prod. Rep. 2007, 24, 843. (d) Ishikura, M.; Yamada, K. Nat. Prod. Rep.
2
2
4
010, 292, 85. (f) Bouffard, J.; Itami, K. Top. Curr. Chem. 2010, 292,
31. (g) Ackermann, L.; Potukuchi, H. K. Org. Biomol. Chem. 2010, 8,
503. (h) Schnuerch, M.; Dastbaravardeh, N.; Ghobrial, M.; Mrozek,
2
009, 26, 803. (e) Trost, B. M.; Brennan, M. K. Synthesis 2009, 3003.
B.; Mihovilovic, M. D. Curr. Org. Chem. 2011, 15, 2694. (i) Yeung, C.
S.; Dong, V. M. Chem. Rev. 2011, 111, 1215. (j) Ackermann, L. Chem.
Rev. 2011, 111, 1315.
(f) Ishikura, M.; Yamada, K.; Abe, T. Nat. Prod. Rep. 2010, 27, 1630.
(21) Cordell, G. A.; Saxton, J. E. In The Alkaloids; Manske, R. H. F.,
Rodrigo, R. G. A., Eds.; Academic Press: New York, 1981; Vol. 20, pp
132−280.
(22) A structurally closely related natural product, aspidospermine,
was first synthesized in 1963: (a) Stork, G.; Dolfini, J. E. J. Am. Chem.
Soc. 1963, 85, 2972. The first total synthesis of aspidospermidine was
reported in 1966: (b) Kutney, J. P.; Abdurahman, N.; Quesne, P. L.;
Piers, E.; Vlattas, I. J. Am. Chem. Soc. 1966, 88, 3656.
(
(
(
3) Ackermann, L. Chem. Commun. 2010, 46, 4866.
4) Jiao, L.; Bach, T. J. Am. Chem. Soc. 2011, 133, 12990.
5) (a) Itahara, T.; Ikeda, M.; Sakakibara, T. J. Chem. Soc., Perkin
Trans. 1 1983, 1361. (b) Itahara, T.; Kawasaki, K.; Ouseto, F. Synthesis
1
984, 236. (c) Yokoyama, Y.; Matsumoto, T.; Murakami, Y. J. Org.
Chem. 1995, 60, 1486. (d) Jia, C.; Lu, W.; Kitamura, T.; Fujiwara, Y.
Org. Lett. 1999, 1, 2097. (e) Grimster, N. P.; Gauntlett, C.; Godfrey, C.
R. A.; Gaunt, M. J. Angew. Chem., Int. Ed. 2005, 44, 3125. (f) Maehara,
A.; Tsurugi, H.; Satoh, T.; Miura, M. Org. Lett. 2008, 10, 1159.
(23) Syntheses of nonracemic aspidospermidine: (a) Node, M.;
Nagasawa, H.; Fuji, K. J. Org. Chem. 1990, 55, 517. (b) Schultz, A. G.;
Pettus, L. J. Org. Chem. 1997, 62, 6855. (c) Iyengar, R.; Schildknegt,
(
6) (a) Catellani, M.; Chiusoli, G. P. J. Organomet. Chem. 1988, 346,
C27. (b) Catellani, M.; Chiusoli, G. P. J. Organomet. Chem. 1992, 425,
51.
7) For reviews, see: (a) Catellani, M. Synlett 2003, 298. (b) Catellani,
K.; Aube,
Huang, Y.; Rawal, V. H. J. Am. Chem. Soc. 2002, 124, 4628.
e) Marino, J. P.; Rubio, M. B.; Cao, G.; Dios, A. J. Am. Chem. Soc.
002, 124, 13398. (f) Gnecco, D.; Vazquez, E.; Galindo, A.; Teran, J.
L.; Orea, L.; Bernes, S.; Enríquez, R. G. Arkivoc 2003, xi, 185.
g) Iyengar, R.; Schildknegt, K.; Morton, M.; Aube, J. J. Org. Chem.
́
J. Org. Lett. 2000, 2, 1625. (d) Kozmin, S. A.; Iwama, T.;
1
(
(
2
́
́
M. Top. Organomet. Chem. 2005, 14, 21. (c) Catellani, M.; Motti, E.;
Della Ca, N. Acc. Chem. Res. 2008, 41, 1512. (d) Martins, A.;
Mariampillai, B.; Lautens, M. Top. Curr. Chem. 2010, 292, 1.
̀
(
2
005, 70, 10645. (h) Ishikawa, T.; Kudo, K.; Kuroyabu, K.; Uchida, S.;
(
8) The bromide anion exists in the alkylation reaction system as a
Kudoh, T.; Saito, S. J. Org. Chem. 2008, 73, 7498. (i) Suzuki, M.;
Kawamoto, Y.; Sakai, T.; Yamamoto, Y.; Tomioka, K. Org. Lett. 2009,
byproduct. Therefore, in this NMR study, KBr was added to simulate
−
the real reaction system. It was found that Br acted as a ligand for the
1
1, 653. (j) Jones, S. B.; Simmons, B.; Mastracchio, A.; MacMillan, D.
W. C. Nature 2011, 475, 183.
24) Syntheses of racemic aspidospermidine: (a) Harley-Mason, J.;
observed intermediate, see Support Information for details.
(
9) Single crystal X-ray structure determination of compound
(
5
·CDCl : red fragment, C H N Pd·CDCl , M = 615.25; monoclinic,
3
27 23
3
3
r
Kaplan, M. Chem. Commun. 1967, 915. (b) Kutney, J. P.; Abdurahman,
N.; Gletsos, C.; Le Quesne, P.; Piers, E.; Vlattas, I. J. Am. Chem. Soc.
1970, 92, 1727. (c) Wenkert, E.; Hudlicky, T. J. Org. Chem. 1988, 53,
́
space group P2 /c (No. 14), a = 11.6211(5), b = 25.3321(12), c =
1
3
8
=
.6027(4) Å, β = 103.6754(18)°, V = 2460.7(2) Å , Z = 4, λ(Mo Kα)
−1
−3
0.71073 Å, μ = 1.104 mm , ρ
= 1.661 g·cm , T = 123(1) K,
calcd
F(000) = 1240, θmax: 25.37°, R1 = 0.0171 (4373 observed data), wR2
1953. (d) Wenkert, E.; Liu, S. J. Org. Chem. 1994, 59, 7677. (e) Forns,
P.; Diez, A.; Rubiralta, M. J. Org. Chem. 1996, 61, 7882. (f) Callaghan,
O.; Lampard, C.; Kennedy, A. R.; Murphy, J. A. Tetrahedron Lett.
1999, 40, 161. (g) Callaghan, O.; Lampard, C.; Kennedy, A. R.;
Murphy, J. A. J. Chem. Soc., Perkin Trans. 1 1999, 8, 995. (h) Toczko,
M. A.; Heathcock, C. H. J. Org. Chem. 2000, 65, 2642. (i) Patro, B.;
Murphy, J. A. Org. Lett. 2000, 2, 3599. (j) Banwell, M. G.; Smith, J. A.
J. Chem. Soc., Perkin Trans. 1 2002, 2613. (k) Banwell, M. G.; Lupton,
D. W. Org. Biomol. Chem. 2005, 3, 213. (l) Banwell, M. G.; Lupton, D.
W.; Willis, A. C. Aust. J. Chem. 2005, 58, 722. (m) Sharp, L. A.; Zard,
S. Z. Org. Lett. 2006, 8, 831. (n) Pearson, W. H.; Aponick, A. Org. Lett.
=
0.0435 (all 4508 data), GOF = 1.085, 316 parameters, Δρ
=
max/min
−3
0
.32/−0.41 e·Å . For more details see Supporting Information.
CCDC-885398.
10) Martin-Matute, B.; Mateo, C.; Cardenas, D. J.; Echavarren, A.
M. Chem.Eur. J. 2001, 7, 2341.
11) Yatsimirskii, K. B. Kinetic Methods of Analysis, 1st ed.; Pergamon:
Oxford, 1966; Chapter 3.
12) A study on the conversion of 8 to 9 indicated that Pd(II) is
(
(
(
responsible for this transformation but norbornene is not. See
Supporting Information for details.
(
13) A control experiment was conducted at [Pd] = 24 mM, which
2
006, 8, 1661. (o) Coldham, I.; Burrell, A. J. M.; White, L. E.; Adams,
excluded a phase-transfer controlled scenario at high [Pd]. See
Supporting Information for details.
H.; Oram, N. Angew. Chem., Int. Ed. 2007, 46, 6159. (p) Callier-
Dublanchet, A.-C.; Cassayre, J.; Gagosz, F.; Quiclet-Sire, B.; Sharp, L.
A.; Zard, S. Z. Tetrahedron 2008, 64, 4803. (q) Ishikawa, T.; Kudo, K.;
Kuroyabu, K.; Uchida, S.; Kudoh, T.; Saito, S. J. Org. Chem. 2008, 73,
(
̃
14) Reviews on the organopalladium(IV) chemistry: (a) Muniz, K.
Angew. Chem., Int. Ed. 2009, 48, 9412. (b) Xu, L.-M.; Li, B.-J.; Yang, Z.;
Shi, Z.-J. Chem. Soc. Rev. 2010, 39, 712. (c) Sehnal, P.; Taylor, R. J. K.;
Fairlamb, I. J. S. Chem. Rev. 2010, 110, 824. (d) Malinakova, H. C.
Top. Curr. Chem. 2011, 35, 85.
7
2
2
́
498. (r) Sabot, C.; Guerard, K. C.; Canesi, S. Chem. Commun. 2009,
941. (s) De Simone, F.; Gertsch, J.; Waser, J. Angew. Chem., Int. Ed.
010, 49, 5767.
(
15) Baxter, R. D.; Sale, D.; Engle, K. M.; Yu, J.-Q.; Blackmond, D. G.
J. Am. Chem. Soc. 2012, 134, 4600.
16) See Support Information for the detailed mathematic derivation
of the rate-law expression.
17) (a) ten Brink, G.-J.; Arends, I. W. C. E.; Papadogianakis, G.;
(25) Forns, P.; Diez, A.; Rubiralta, M.; Solans, X.; Font-Bardia, M.
(
Tetrahedron 1996, 52, 3563.
(26) Russell, E.; McElvain, S. M. J. Am. Chem. Soc. 1935, 57, 1443.
(27) Takano, S.; Sato, T.; Inomata, K.; Ogasawara, K. J. Chem. Soc.,
Chem. Commun. 1991, 462.
(28) (a) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155.
(b) Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277.
(29) Arai, E.; Tokuyama, H.; Linsell, M. S.; Fukuyama, T.
Tetrahedron Lett. 1998, 39, 71.
(
Sheldon, R. A. Appl. Catal., A 2000, 194−195, 435. (b) ten Brink, G.-
J.; Arends, I. W. C. E.; Sheldon, R. A. Adv. Synth. Catal. 2002, 344, 355.
(
c) Steinhoff, B. A.; Stahl, S. S. J. Am. Chem. Soc. 2006, 128, 4348.
18) A seven-membered palladaheterocycle was proposed as the key
intermediate in the annulative coupling reaction between 1-(2-
(
1
4571
dx.doi.org/10.1021/ja3058138 | J. Am. Chem. Soc. 2012, 134, 14563−14572

Journal of the American Chemical Society
Article
(
30) The transformation from aminoalcohol 11 to imine 19 can also
be achieved by treatment with TsCl/t-BuOK. See Supporting
Information for details.
(
31) Randriambola, L.; Quirion, J. C.; Kanfan, C.; Husson, H. P.
Tetrahedron Lett. 1987, 28, 2123.
32) (a) Ref.27. (b) Morales, C. L.; Pagenkopf, B. L. Org. Lett. 2008,
0, 157. (c) Ref.24s. (d) Mizutani, M.; Inagaki, F.; Nakanishi, T.;
Yanagihara, C.; Tamai, I.; Mukai, C. Org. Lett. 2011, 13, 1796.
33) Kuwano, R.; Kashiwabara, M.; Sato, K.; Ito, T.; Kaneda, K.; Ito,
Y. Tetrahedron: Asymmetry 2006, 17, 521.
34) (a) Lal, B.; Pramanik, B. N.; Manhas, M. S.; Bose, A. K.
Tetrahedron Lett. 1977, 18, 1977. (b) Hughes, D. L. Org. React. 1992,
2, 335. (c) Swamy, K. C. K.; Kumar, N. N. B.; Balaraman, E.; Kumar,
K. V. P. P. Chem. Rev. 2009, 109, 2551.
(
1
(
(
4
1
4572
dx.doi.org/10.1021/ja3058138 | J. Am. Chem. Soc. 2012, 134, 14563−14572