Formal homo-nazarov and other cyclization reactions of activated cyclopropanes

Article abstract of DOI:10.1002/chem.201102583

The Nazarov cyclization of divinyl ketones gives access to cyclopentenones. Replacing one of the vinyl groups by a cyclopropane leads to a formal homo-Nazarov process for the synthesis of cyclohexenones. In contrast to the Nazarov reaction, the cyclization of vinyl-cyclopropyl ketones is a stepwise process, often requiring harsh conditions. Herein, we describe two different approaches for further polarization of the three-membered ring of vinyl-cyclopropyl ketones to allow the formal homo-Nazarov reaction under mild catalytic conditions. In the first approach, the introduction of an ester group α to the carbonyl on the cyclopropane gave a more than tenfold increase in reaction rate, allowing us to extend the scope of the reaction to non-electron-rich aryl donor substituents in the β position to the carbonyl on the cyclopropane. In this case, a proof of principle for asymmetric induction could be achieved using chiral Lewis acid catalysts. In the second approach, heteroatoms, especially nitrogen, were introduced β to the carbonyl on the cyclopropane. In this case, the reaction was especially successful when the vinyl group was replaced by an indole heterocycle. With a free indole, the formal homo-Nazarov cyclization on the C3 position of indole was observed using a copper catalyst. In contrast, a new cyclization reaction on the N1 position was observed with BrAnsted acid catalysts. Both reactions were applied to the synthesis of natural alkaloids. Preliminary investigations on the rationalization of the observed regioselectivity are also reported.

Full text of DOI:10.1002/chem.201102583

FULL PAPER
DOI: 10.1002/chem.201102583
Formal Homo-Nazarov and Other Cyclization Reactions of Activated
Cyclopropanes
Filippo De Simone, Tanguy Saget, Fides Benfatti, Sofia Almeida, and Jꢀrꢁme Waser*[a]
Abstract: The Nazarov cyclization of
divinyl ketones gives access to cyclo-
pentenones. Replacing one of the vinyl
groups by a cyclopropane leads to a
formal homo-Nazarov process for the
synthesis of cyclohexenones. In con-
trast to the Nazarov reaction, the cycli-
zation of vinyl-cyclopropyl ketones is a
stepwise process, often requiring harsh
conditions. Herein, we describe two
different approaches for further polari-
zation of the three-membered ring of
vinyl-cyclopropyl ketones to allow the
formal homo-Nazarov reaction under
mild catalytic conditions. In the first
approach, the introduction of an ester
group a to the carbonyl on the cyclo-
propane gave a more than tenfold in-
crease in reaction rate, allowing us to
extend the scope of the reaction to
non-electron-rich aryl donor substitu-
ents in the b position to the carbonyl
on the cyclopropane. In this case, a
proof of principle for asymmetric in-
duction could be achieved using chiral
Lewis acid catalysts. In the second ap-
proach, heteroatoms, especially nitro-
gen, were introduced b to the carbonyl
on the cyclopropane. In this case, the
reaction was especially successful when
the vinyl group was replaced by an
indole heterocycle. With a free indole,
the formal homo-Nazarov cyclization
on the C3 position of indole was ob-
served using a copper catalyst. In con-
trast, a new cyclization reaction on the
N1 position was observed with Brønst-
ed acid catalysts. Both reactions were
applied to the synthesis of natural alka-
loids. Preliminary investigations on the
rationalization of the observed regiose-
lectivity are also reported.
Keywords: alkaloids · catalysis · cy-
clopropane
·
heterocyclic com-
pounds · regioselectivity
Introduction
was limited by the necessity to use a stoichiometric amount
of a strong Lewis or Brønsted acid.[3] In the last years how-
ever, several groups have reported milder catalytic meth-
ods.[4–5] The key for success was the introduction of an elec-
tron-donating and an electron-withdrawing group into the a
position to the carbonyl (EDG and R3 in Scheme 1). In such
substrates, one of the vinyl groups becomes nucleophilic,
and the other one electrophilic (cross polarization), resulting
in an enhanced reactivity of the divinyl ketone. This ap-
proach was extensively investigated by Frontier and co-
workers.[4k] Since this seminal discovery, several catalytic
versions were reported, together with the first examples of
asymmetric induction.[5]
Starting from the Nazarov cyclization, a possible approach
to access larger ring systems consists in the substitution of a
double bond by an activated cyclopropyl group, resulting in
a homo-Nazarov process (Scheme 1, B). The ring strain of
cyclopropanes makes them excellent precursors of reactive
intermediates in the formation of new carbon–carbon bonds.
Furthermore, the Walsh orbital description of the three-
The continuing progress in medicinal chemistry and chemi-
cal biology requires flexible synthetic approaches for the
generation of libraries of active molecules and their ana-
logues. A broad range of biologically active natural and syn-
thetic compounds displays a complex polycyclic heterocyclic
scaffold as a core structure. Consequently, the search for
new cyclization and cycloaddition reactions is a fundamental
task in organic chemistry.[1]
Pericyclic reactions involving the concerted rearrange-
ment of bonding electrons have the advantage of low activa-
tion barriers, allowing mild reaction conditions. The Nazarov
cyclization, discovered by Nazarov in 1941,[2] is an important
example of an electrocyclic ring closure–a subclass of pericy-
clic reactions–in which five-membered rings are generated
starting from divinyl (or arylvinyl) ketones (Scheme 1, A).
After activation of the carbonyl with a Brønsted or a Lewis
acid, the intermediate pentadienyl cation I undergoes a con-
rotatory ring-closure generating a cyclopentenone after
proton transfer. For several decades, the Nazarov cyclization
membered ring has been used to illustrate the partial
[6]
À
p-character of C C bonds in cyclopropanes. They can con-
sequently be considered as homologues of olefins in several
annulations and cyclization reactions. Nevertheless, there is
strong evidence that all the cyclizations of vinyl-cyclopropyl
ketones reported so far are stepwise processes in contrast to
the concerted Nazarov reaction (Scheme 1, C) and can con-
sequently only formally be compared to an electrocyclic
process.[7] An important consequence of the stepwise mecha-
[a] F. De Simone, T. Saget, Dr. F. Benfatti, S. Almeida, Prof. Dr. J. Waser
Laboratory of Catalysis and Organic Synthesis
Ecole Polytechnique Fꢀdꢀrale de Lausanne, EPFL SB ISIC LCSO
BCH 4306, 1015 Lausanne (Switzerland)
Fax : (+41)21-693-9700
E-mail: jerome.waser@epfl.ch
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201102583.
Chem. Eur. J. 2011, 17, 14527 – 14538
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
14527

strong Lewis or Brønsted acids (Scheme 2).[7a–e] In 1980,
Murphy and Wattanasin showed that an excess of SnCl4 al-
lowed the cyclization of aryl-cyclopropyl ketones for the
synthesis of aryltetralones in good yields (Scheme 2, A).[7a–c]
Scheme 2. Early examples of formal homo-Nazarov cyclizations. PPA=
polyphosphoric acid.
In 1988, Tsuge and co-workers reported the first cyclization
of vinyl-cyclopropyl ketones using an excess of polyphos-
phoric acid at 808C (Scheme 2, B).[7d] The reaction was not
general and several by-products were obtained. In 2008,
Yadav and co-workers reported the cyclization of silylme-
thylcyclopropyl-ketones by using three equivalents of SnCl4
at 808C (Scheme 2, C).[7e] In this work, the scope of the re-
action could be extended to a broad range of heterocyclic
systems, but the utility of the method was limited by the
harsh conditions needed.
Scheme 1. Nazarov cyclization of divinyl ketones and cyclization reac-
tions of vinylcyclopropyl ketones (EDG=electron-donating group, LA=
Lewis Acid, Nu=nucleophile).
nism is that the reaction is usually not anymore conrotatory.
However, as the starting materials and products are identi-
cal in reactions B and C (Scheme 1), it is useful to keep a
“homo-Nazarov” process in mind for retrosynthetic plan-
ning. The term formal should never been omitted, as it
serves as a strong warning sign to indicate a different mech-
anism. Finally, further reactions involving intramolecular
attack of a nucleophile b to the carbonyl on the cyclopro-
pane, although highly useful, should not be called formal
homo-Nazarov reactions, because they lack the pattern of
p bonds characteristic for the reaction (Scheme 1, D).[7g,8]
To enhance the reactivity of cyclopropanes and allow cyc-
lization reactions under mild conditions, it is often required
to add further activating groups onto the ring. In particular,
zwitterionic intermediates can be generated under mild con-
ditions by tuning the donor-acceptor properties of substitu-
ents on the three-membered ring.[9] Despite the ground-
breaking works of Stork, Danishefsky, Corey and co-work-
ers during the 1970s,[10] intramolecular cyclizations involving
activated cyclopropanes have been only rarely investigated.
Until 2008, the formal homo-Nazarov cyclization of vinyl or
aryl cyclopropyl ketones to form cyclohexenones was limit-
ed in scope and required superstoichiometric amounts of
Inspired by the progress in the Nazarov reaction, we de-
cided to develop the first catalytic formal homo-Nazarov
cyclization. We speculated that activation through adequate
polarization of the cyclopropane ring could also lead to
milder catalytic conditions. In our preliminary studies we ex-
amined the effect of an electron-donating aromatic group b
to the carbonyl on the cyclopropane in combination with an
enol ether, an electron-rich aromatic ring, or an allyl silane
as a nucleophilic double bond (Scheme 3).[7f] By using a cat-
alytic amount of para-toluenesulfonic acid in polar solvents,
we were able to access the desired cyclohexenones under
mild conditions. The reaction tolerated variation of the size
and nature of the groups attached to the double bond and
worked with several electron-rich aromatic groups, but no
reaction was observed with a simple phenyl group. Further-
Scheme 3. Catalytic formal homo-Nazarov cyclization.
14528
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 14527 – 14538

Formal Homo-Nazarov and Other Cyclization Reactions
FULL PAPER
more, no asymmetric induction could be obtained when
using chiral Brønsted or Lewis acids.
allow milder reaction conditions. b-Ketoesters are also
known to favor two point-binding interactions with chiral
Lewis acid catalysts, allowing an easier control of the enan-
tioselectivity. In this context, Trauner and co-workers[5b–c]
proposed one of the first examples of a catalytic Nazarov re-
action in 2004 by using a scandium PYBOX complex with
ketoesters. Aggarwal[5a] and Togni[5d] also reported examples
of asymmetric induction for similar substrates using chiral
copper and nickel catalysts respectively. More recently,
Johnson and co-workers have studied extensively Lewis
acid-catalyzed [3+2] annulations of polarized 1,1-diester cy-
clopropanes with aldehydes and they also reported an asym-
metric cycloaddition using chiral Mg PYBOX catalysts.[12]
Starting from dimethyl malonate ester 1, we first devel-
oped a synthetic route towards cyclopropyl b-ketoesters
6b–f (Scheme 4). The synthesis of cyclopropane 6a has al-
To overcome these limitations, we decided to apply two
strategies: 1) The introduction of an ester group on the cy-
clopropane at the a position to the ketone; the electron-
withdrawing ester group was expected to increase the polari-
zation of the cyclopropane. We examined this approach for
the first time in 2009.[11] In 2010, France and co-workers re-
ported the use of the same class of substrates using indium
catalysts, which allowed them to extend significantly the
scope of the reaction.[7h–i] Apart for enhancing the reactivity,
we also thought that this class of substrates will offer unique
opportunity for asymmetric induction, as two-points binding
with chiral Lewis acids will become possible. 2) The substi-
tution of the aromatic group on the cyclopropane by a het-
eroatom. The lone pair of the heteroatom would give the
electron density required for ring-opening under mild condi-
tions. As heteroatoms are omnipresent in bioactive com-
pounds, the obtained products would furthermore be highly
useful for synthetic applications.
Scheme 4. Synthesis of cyclopropane ketoesters 6a–f.
Herein, we present first our preliminary results on the
former strategy, the synthesis and cyclization of b-ketoester
cyclopropanes and the first study of this class of substrates
in asymmetric cyclizations. Concerning the latter approach,
we have reported in 2010 the first example of the formal
homo-Nazarov cyclization of aminocyclopropanes and their
application in the synthesis of natural alkaloids.[7g] We now
present the details of the challenging synthesis of the amino-
cyclopropanes, the full study of the scope of their cyclization
(in particular using ketones derived from electron-rich het-
erocycles), as well as a comparison with an oxycyclopro-
pane. We also discuss more in detail the mechanism of the
cyclization reaction based on experimental and computa-
tional data.
ready been described in our previous work.[7f] A Regitz
diazo transfer,[13] followed by Rh-catalyzed cyclopropanation
of the required olefins afforded the desired 1,2 cyclopropane
diesters 3b–f.[14] In a first approach, the addition of lithiated
dihydropyran 5 on diesters 3b gave 6b as a diasteromeric
mixture in low yield and with the undesired formation of
double addition products. Therefore a more selective syn-
thetic strategy was developed. Hydrolyzing the less sterically
hindered of the two esters would give a carboxylic acid,
which could be deprotonated and form a stable lithium com-
plex. This will serve two roles: deactivating the acid to solve
the regioselectivity problem and stabilize the tetrahedral in-
termediate obtained after addition of an organometallic re-
agent to prevent double addition. In the event, a selective
saponification of the less sterically hindered trans ester
group in 3c–f led to the corresponding acids 4c–f in high
yields without further purification.[15] Gratifyingly, treating
carboxylic acids 4c–f with two equivalents of lithiated dihy-
dropyran 5 led to regioselective formation of the corre-
sponding keto acids in high yields. Finally, methylation of
the free carboxylate gave access to the desired substrates
Results and Discussion
Activation with an ester substituent
The use of b-ketoester cyclopropanes was expected to lead
to more stable enolate intermediates and consequently to
Chem. Eur. J. 2011, 17, 14527 – 14538
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14529

J. Waser et al.
6c–f. Since this strategy afforded the addition products on
the more hindered face of the dicarbonyl compounds, it con-
stitutes a new method to obtain selectively the more steri-
cally hindered trans-cyclopropane ketoesters.
To have an impression on the reactivity of the ketoester
substrates, the reaction in presence of several Brønsted and
Lewis acids was first monitored qualitatively by 1H NMR
spectroscopy (Table 1). b-Ketoester cyclopropane 6b was
sion rate at RT, however partial decomposition was induced
when heated (Table 1, entries 6–11). On the other hand,
stronger Lewis acids such as SnCl4 or AlCl3 gave extensive
decomposition in less than one hour (Table 1, entries 12 and
13). In the search for a milder catalyst, we tested Ni-
AHCTUNGTRENNG(UN ClO4)2·6H2O, which had proven to be an effective Lewis
acid for the opening of donor–acceptor cyclopropanes, both
as anhydrous[12a,16] or as aquo complex.[15] In our hands, only
the aquo complex was active and gave a clean
and complete conversion of 6c after 26 h, with
63% of 7c isolated (Table 1, entry 14). Surpris-
ingly BF3·Et2O, which gave decomposition with
6a (Table 1, entry 15), was the catalyst of choice
for b-ketoesters cyclopropane 6c (Table 1,
entry 16), promoting the cyclization in 83% yield
in a short reaction time.
Table 1. Preliminary screening of conditions for the homo-Nazarov cyclization for ke-
toester substrates.
With the optimized conditions in hand, we also
examined methyl- and chloro substituents on the
benzene ring (Scheme 5). Fortunately, the addi-
tional reactivity given by the b-ketoesters could
compensate even the electron-deficiency of the
p-chlorophenyl group in 6e. To facilitate charac-
terization, the cyclic compounds were decarboxy-
lated in the presence of NaCl in wet DMSO. In a
further experiment, no conversion was observed
when the aromatic group was replaced with a
2-phenylethyl substituent in 6 f. This last result
underlined the importance of the stabilizing
effect of the aromatic group for the reactivity of
the cyclopropane.
We also detected a progressive epimerization
of the cyclopropane during the homo-Nazarov re-
action. Using either cis or trans phenylcyclopro-
pane 6c, we isolated the same mixture of syn-
and anti-cyclic isomers with 1:1.5 diastereoselec-
tivity. This result would be expected if the reac-
tion proceeded through a carbocation or if equili-
bration of the products was occurring under these
conditions.
Entry[a] R1
R2
Catalyst
A
Solvent
t
Conv.
[%]
[b]
1
H
MeO (6a) TsOH (20)
CH3CN 18 h
CH3CN 15 min
CH3CN 19 h
CH2Cl2 19 h
CH2Cl2 19 h
CH2Cl2 36 h
CH2Cl2 26 h
CH2Cl2 18 h
CH2Cl2 20 min
DCE
DCE
CH2Cl2 2.3 h
CH2Cl2 1 h
100
100
0
0
0
2
CO2Me MeO (6b) TsOH (20)
3
4
5
6
7
8
9
10
11
12
13
14
15
16
CO2Me H (6c)
CO2Me H (6c)
CO2Me H (6c)
CO2Me H (6c)
CO2Me H (6c)
CO2Me H (6c)
CO2Me H (6c)
CO2Me H (6c)
CO2Me H (6c)
CO2Me H (6c)
CO2Me H (6c)
CO2Me H (6c)
TsOH (20)
TfOH (5)
H2SO4 (5)
Cu
Sn(OTf)2 (15)
MgI2 (50)
Cu(OTf)2 (15)
Sn(OTf)2 (5)
A
<20
E
<20
<20
E
dec.[c,d]
G
15 min
1.15 h
dec.[c,e]
dec.[c,e]
dec.[c]
MgI2 (15)
SnCl4 (50)
AlCl3 (50)
dec.[c]
Ni
MeO (6a) BF3·Et2O (50)
CO2Me H (6c) BF3·Et2O (10)
A
100 (63)[f]
dec.[c]
H
CH2Cl2 25 min
CH2Cl2 3 h
100 (83)[f]
[a] 6a and 6c only trans, 6b as a mixture cis/trans 3.3:1. [b] Reaction performed with 6
(50–400 mmol) in 0.5–4 mL solvent. Conversion estimated by comparison of integra-
tion of peaks of 6 and 7 in the H NMR spectra of the crude reaction mixture. See the
Supporting Information for details. [c] Decomposition of the starting material was ob-
served. [d] Reaction run at 1108C. [e] Reaction run at 558C. [f] Isolated product yield
after column chromatography. DCE=1,2-dichloroethane; TsOH= para-toluenesul-
fonic acid.
1
Inspired by the recent works in asymmetric
cyclization with chiral Lewis acids of Johnson,[12]
submitted to the catalytic conditions used in our previous
work (Table 1, entry 2).[7f] Gratifyingly, compound 6b dis-
played an enhanced reactivity when compared with ketone
cyclopropane 6a using the standard condition for the formal
homo-Nazarov cyclization (20 mol% para-toluenesulfonic
acid in CH3CN): With b-ketoester 6b, complete conversion
was observed after only 15 min at RT (Table 1, entry 2),
whereas a period of 18 h was required in absence of the
ester group (Table 1, entry 1). Since phenyl-substituted cy-
clopropane 6c gave no conversion in our previous work,[7f]
we next decided to study the activating effect of the ester
group for this substrate. However, no conversion was ob-
served with this substrate using para-toluenesulfonic acid as
catalyst (Table 1, entry 3). Increasing the strength of the
Brønsted acid was unsuccessful (Table 1, entries 4 and 5).
Trauner[5b–c] and Togni,[5d] we then turned to the develop-
ment of an asymmetric version of the homo-Nazarov cycli-
zation. Preliminary studies were realized with substrates 6b
and 6g, because the para-methoxyphenyl group allowed a
rapid ring-opening of the cyclopropane, which had been
shown to be crucial for the dynamic kinetic asymmetric
Scheme 5. Effect of the substitution on the aromatic ring on the cycliza-
tion reaction (preparative scale).
CuACHTUNGTRENNUNG(OTf)2, SnACHTUNGTRENNUNG(OTf)2, and MgI2 showed a very slow conver-
14530
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 14527 – 14538

Formal Homo-Nazarov and Other Cyclization Reactions
FULL PAPER
transformation of racemic cyclopropanes.[12b] The more steri-
cally hindered tert-butyl ester 6g was prepared by addition
of lithiated pyran 5 to the corresponding tert-butyl-methyl
cyclopropane diester.[17] Both methyl- (6b) and tert-butyl
(6g) esters were then tested in the asymmetric reaction to
analyze the steric influence of the ester on the stereoselec-
tivity. Based on the work of Johnson and co-workers,[12]
PYBOX ligands 8 and 9 were tested in association with Mg
salts (Scheme 6). As the mixture of diastereoisomers formed
spermidine (10), strychnine (11), vinblastine (12), and vin-
cristine (13).
Despite the importance of acyliminium cyclization in al-
kaloid synthesis, there are only a few examples on the use of
aminocyclopropanes to generate these intermediates. In the
seminal work of Wenkert and co-workers, aminocyclopro-
panes were opened under acidic conditions to generate the
corresponding lactones, which were used as precursors of
acyliminium intermediates in intermolecular addition reac-
tions on indole derivatives.[18] To the best of our knowledge,
the only example of intramolecular cyclization of aminocy-
clopropanes has been reported by Six and co-workers, who
studied the intramolecular electrophilic aromatic substitu-
tion of simple alkyl cyclopropanes.[19]
Scheme 6. Attempts towards an asymmetric homo-Nazarov cyclization.
TFA=trifluoroacetic acid.
Oxycyclopropanes on the other hand were shown to un-
dergo formal homo-Nazarov cyclization in the presence of a
stoichiometric amount of Lewis acid.[7e] Since the electron-
density of oxycyclopropanes could be compared to that of
aminocyclopropanes substituted with a carbamate group, a
similar reactivity in the catalytic formal homo-Nazarov cycli-
zation could be expected. We speculated that the high elec-
tron-density of both amino- and oxy-cyclopropanes should
allow their use also under catalytic conditions. We conse-
quently decided to examine a range of amino- and oxy-cy-
clopropanes to assess the potential of the reaction.
during the homo-Nazarov reaction made analysis difficult,
the enantiomeric excess (ee) values were measured on the
decarboxylated product 7a. Only 11% ee was observed
when using the methyl ester 6b. A slightly higher value
(25%) could be achieved using tert-butyl ester 6g. Although
the observed enantioselectivity is still very low, these pre-
liminary results constituted the first example of asymmetric
induction in the formal homo-Nazarov reaction. Further
screening of chiral catalysts (especially those successful in
the Nazarov reaction)[5] to increase the selectivity of the re-
action will be part of future work.
As a first attempt, we decided to examine substrate 14,
because it would lead to the tetracyclic scaffold of the aspi-
dosperma alkaloid aspidospermidine (10). The synthesis of
14 started with d-valerolactam (24), which was protected, re-
Introduction of heteroatoms on the cyclopropane
Introduction and substrate synthesis: Although new condi-
tions with ketoesters were successfully developed, the neces-
sary presence of an aromatic stabilizing group still constitut-
ed a strong limitation in terms of application and scope. Re-
placing the aromatic group by a heteroatom would give a
good stabilization of the formed carbocation and at the
same time allow broadening the range of structures accessi-
ble. For example, a nitrogen-substituted cyclopropane would
generate a zwitterionic species stabilized as an acyliminium
intermediate, which could also be considered as the product
of a retro-Mannich reaction, a process difficult in the case of
acyliminium ions. Moreover, the cyclohexylamines obtained
from the aminocyclopropanes are largely represented in the
core structure of several natural alkaloids such as aspido-
Chem. Eur. J. 2011, 17, 14527 – 14538
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14531

J. Waser et al.
duced, and dehydrated to afford enamide 26 (Scheme 7).[20]
The non-optimized conditions of cyclopropanation using
copper-bronze gave a diastereomeric mixture of esters 27. A
sequence of hydrolysis and amide coupling yielded Weinreb
amide 29. In the hydrolysis step, only the less sterically hin-
dered exo-cyclopropane reacted, allowing the isolation of a
single isomer of acid 28.
Scheme 9. Cyclization of indole aminocyclopropane 14. TsOH=toluene-
sulfonic acid.
it was deprotected prior to complete the structural assign-
ment.
With this successful result in hand, we decided to synthe-
size the ethyl-substituted cyclopropane 15 because the ethyl
group is present in aspidospermidine (10) (Scheme 10).
Scheme 7. Synthesis of Weinreb amide 29. Cbz=carbobenzyloxy.
NMM=N-methylmorpholine.
Optimization of the coupling of 2-lithiated-N-methyl
indole (30) with amide 29 was required, because we only ob-
served the formation of enamide 31 derived from cyclopro-
pane ring-opening under standard conditions (Scheme 8).
Scheme 10. Synthesis of ethyl-substituted cyclopropane 15.
Following a modified procedure of Grieco and Kauf-
man,[20] d-valerolactam (24) was alkylated and protected in
one pot. Lactam 34 was then reduced and the resulting half
aminal was dehydrated to afford enamide 35. The fine
tuning of catalyst and reaction time was crucial for the re-
producibility and the yield of the cyclopropanation of 35.
Since the use of Cu/Zn alloy in the cyclopropanation gave
no reproducible results, we decided to test several copper
and rhodium catalysts. The slow addition of a diluted solu-
tion of ethyl diazoacetate was important to avoid extensive
polymerization and copper(I) catalysts allowed a better
yield when compared with rhodium(II) or copper(II) cata-
lysts. Using copper(I) triflate and syringe-pump addition of
an ethyl diazoacetate solution, a diasteromeric mixture
(1.3:1 diastereomeric ratio (d.r.)) of cyclopropane ethyl
esters 36a and 36b was obtained in a reproducible 76%
yield. The isomerization of the endo ester 36b into the more
stable exo form 36a was possible with catalytic BF3·OEt2 in
excellent yield.[20] Ester 36a was then saponified affording
the carboxylic acid 37 as a pure diasteroisomer. The use of
4-(4,6-dimethoxy-1,3,5-triazin-2-yl)-4-methylmorpholinium
chloride (DMTMM) allowed the formation of Weinreb
Scheme 8. Coupling of Weinreb amide 29 with lithiated indole 30 (rsm=
recovered starting material).
We speculated that the cyclopropane-opening was favored
due to the high stability of the N,O-lithium chelate formed
upon addition of the organometallic reagent. The addition
of potassium tert-butoxide could potentially lead to the ex-
change of lithium by potassium and form a weaker chelate,
which should favor a faster elimination of the hydroxyla-
mine and preserve the cyclopropane integrity. Indeed, we
were pleased to obtain the desired product 14 in 67% yield
using this approach.
Using the standard conditions developed for the catalytic
homo-Nazarov cyclization, we were delighted to isolate the
cyclic compound 32 in 90% yield with high diasteroselectivi-
ty for the cis-fused product (Scheme 9). As NMR analysis of
32 was difficult due to the presence of carbamate rotamers,
14532
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 14527 – 14538

Formal Homo-Nazarov and Other Cyclization Reactions
FULL PAPER
amide 38 in good yield without further purification.[21] The
coupling reaction between 2-lithiated-N-methylindole (30)
and Weinreb amide 38 afforded the precursor 15 of the
homo-Nazarov reaction.
In our previous publication we successfully tested the cyc-
lization involving substituted indoles.[7g] Encouraged by
these promising results, a further series of substrates was
synthesized to obtain more information regarding the scope
of heterocycles in the cyclization step. For this purpose, a
range of heterocycles with decreasing nucleophilicity were
targeted as nucleophiles in the cyclization reaction. We fo-
cused our studies on the cyclopropanes derived from Wein-
reb amide 38 because the cyclization products would repre-
sent interesting analogues of natural alkaloids. Combining
the appropriate metalated heterocycle with Weinreb amide
38, we generated 3-indole-, pyrrole-, dihydropyran-, and
As aspidodermidine contains a free indole nitrogen, cycli-
zation of substrate 16 with an unprotected indole would be
more efficient. When using a free indole in the cyclization
however, regioselectivity could become a serious issue, be-
cause cyclization on the nitrogen could also occur. For this
reason cyclopropane 39, bearing a free indole, was first syn-
thesized starting from lithiated N-carboxy indole and the
corresponding Weinreb amide cyclopropane. We were de-
lighted to find that a catalytic amount of para-toluenesulfon-
ic acid in acetonitrile gave exclusively the C3 cyclization
product 40 in quantitative yield (Scheme 11).
thiophene-substituted
carbonylcyclopropanes
17–20
(Scheme 13). The pyrrole was carboxylated on its free nitro-
gen prior to lithiation. For these new substrates, no attempt
was made to optimize the reaction conditions.
Scheme 11. Cyclization of free indole cyclopropane 39.
With these promising preliminary results in hand, we
moved on to the synthesis of indole 16. The use of indole
N-carboxylates pioneered by Katritsky[22] presented several
advantages for our synthetic strategy: the directing group
property of carboxylate favored the selective lithiation on
the C2 position of indole and a protecting-group-free 2-car-
bonyl indole is generated after aqueous work up. Neverthe-
less, this method had been used only with simple substrates
in the past. The lithium indole-N-carboxylate was initially
prepared in situ by a sequence of deprotonation/carboxyla-
tion/lithiation then added at À788C into a solution of Wein-
reb amide 38 affording 16 in low yield. Optimization of ad-
ditives, temperature, stoichiometry, reaction time, and the
use of pre-formed and recrystallized indole-N-carboxylic
acid 41 resulted in an increased yield and reproducible re-
sults on a multigram scale (Scheme 12). In particular, the
use of LiCl as additive and the slow increase of temperature
from À788C to À208C combined with a reversed quench at
08C were crucial to prevent side reactions. This implementa-
tion of the Katritsky method in complex settings allowed a
highly convergent synthesis of indole cyclopropane 16.
Scheme 13. Coupling of 38 with different heterocycles.
To further examine the scope and limitations of the ami-
nocyclopropane cyclization, acyclic carbamates 21 and 22
were then synthesized (Scheme 14). The absence of a rigidi-
fying ring makes this class of compounds particularly chal-
lenging with regard to the cyclization step. The required en-
amide 46 was obtained from a Curtius rearrangement of
acryloyl chloride followed by benzylation.[23] The optimized
Scheme 12. Synthesis of indole cyclopropane 16.
Scheme 14. Synthesis of acyclic cyclopropyl carbamates 21 and 22.
Chem. Eur. J. 2011, 17, 14527 – 14538
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14533

J. Waser et al.
cyclopropanation conditions gave a diastereoisomeric mix-
ture (d.r.=1:1) of aminocyclopropyl ethyl esters 47a and
47b in 69% yield. After isomerization with BF3·OEt2 and
saponification, the resulting acid 48 was converted into
Weinreb amide 49. The right choice of lithiating agent, the
fine tuning of reaction conditions, and the use of N-methyl-
indole 50 or carboxy indole 41 was crucial to give useful
yields of cyclopropanes 21 and 22.[24]
Scheme 16. Cyclization of ethyl-substituted cyclopropane 15.
We then synthesized the cyclic ether cyclopropane sub-
strate 23 starting from dihydropyran (44; Scheme 15). A
high temperature cyclopropanation with copper catalyst and
EDA afforded the desired ester 51 as a single diasteroisom-
were surprised to isolate two different compounds. After re-
moval of the protecting group we were able to identify the
desired product 57 obtained from the attack at the C3 posi-
tion of indole and compound 58 derived from the cyclization
on the N1 position in a 1.6:1 ratio (Scheme 17). To find the
Scheme 15. Synthesis of oxycyclopropane 23. EDC=N’-(3-dimethylami-
nopropyl)-N-ethylcarbodiimide;
TEA=triethylamine.
DMAP=4-dimethylaminopyridine;
Scheme 17. Regio- and diasteroselective cyclization of cyclopropane 16.
er.[25] Equilibration of the diastereoisomers probably did
occur spontaneously through ring-opening at high tempera-
ture. During saponification to afford the carboxylic acid, the
oxycyclopropane ring partially opened, generating a vinyl-
ether side product 53. The mixture of open and closed forms
was used directly in a peptide coupling with N’-(3-dimethy-
laminopropyl)-N-ethylcarbodiimide (EDC) to afford an in-
separable mixture of Weinreb amides 54 and 55. Finally, the
separation of the two forms was possible after the reaction
with the bis-lithiated carboxyindole obtained from 41 allow-
ing the isolation of pure cyclopropane 23 (Scheme 15).[26]
optimum conditions for controlling the regioselectivity of
the cyclization, a screen was carried out; the combination of
À
polar solvent and a soft Lewis acid gave selectively the C C
cyclization compound, whereas a strong Brønsted acid in an
[27]
À
apolar solvent favored C N bond formation (Scheme 17).
One of the potential explanations we had proposed in our
previous work[7g] was based on the hard/soft reactivity of the
generated charged intermediates (Scheme 18). The use of
CuACTHUNTRGENNG(U OTf)2 would generate a soft copper-bound enolate I,
which would promote the attack on the iminium by the
softer C3 indole position with the higher electron-density in
the HOMO orbital. Rearomatization of the cyclization
product II and proton-transfer would lead to 57 and regen-
erate the catalyst. Moreover, the stabilizing effect offered by
Cyclization: Encouraged by the preliminary results obtained
with indole 14 (Scheme 9), we examined the cyclization of
ethyl-substituted cyclopropane 15 required for the synthesis
of aspidospermidine (10). The presence of a quaternary
carbon and the resulting steric hindrance was expected to be
a major challenge for the cyclization process. We were con-
sequently delighted to observe quantitative cyclization with
perfect diastereoselectivity under our standard conditions
(Scheme 16). The high diastereoselectivity shown by the
formal homo-Nazarov cyclization is an important advantage
compared with other synthetic strategies proposed for the
synthesis of aspidosperma alkaloids using aminocyclopro-
panes, which were based on intermolecular addition and
gave low stereoselectivity.[18d]
In a first cyclization attempt for free indole 16, using the
standard conditions for the homo-Nazarov cyclization, we
Scheme 18. Probable intermediates for the cyclization reactions.
14534
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 14527 – 14538

Formal Homo-Nazarov and Other Cyclization Reactions
FULL PAPER
the polar solvent plays a fun-
damental role in favoring an
orbital controlled over
a
charge-controlled cyclization.
On the other hand, when the
aminocyclopropane is activat-
ed with para-toluenesulfonic
acid, the formed enol III
would have a “hard” character,
which would favor the attack
on the formed acyl iminium by
the harder N1 indole position.
Using the non-coordinating
solvent methylene chloride, a
fast charge-controlled reaction
on the harder nitrogen is fur-
ther favored. From the ob-
tained cyclization intermediate
IV, a simple deprotonation and
tautomerization would give 58.
Figure 1. Optimized geometries of the four isomers arising from C3 and N1 cyclization. The energy values
[kcalmolÀ1] reported in parentheses are relative to 57.
The formation of the more stable thermodynamic product
over the kinetic product should also be favored under the
Cu-catalyzed conditions. To support this hypothesis, we ex-
amined interconversion of the two cyclization products
under the reaction conditions. It was possible to convert the
cyclic compound 58 into the regioisomer 57 using copper tri-
flate in acetonitrile (Scheme 17), whereas no interconversion
was observed when 57 was submitted to para-toluenesulfon-
ic acid in dichloromethane. These last results suggested that
twist conformation observed in 59 and 60. Consequently, the
high diastereoselectivity observed could either result from
the higher stability of the cis product, or from the require-
ment of a twist transition state to form the trans product
(Fꢂrst–Plattner rule).[32] Furthermore, the higher thermody-
namic stability of the C3-cyclization cis adduct 57 in respect
to the N1-cyclization cis adduct 58 (DE57À58 =6.59 Kcal
molÀ1) is in agreement with the observed conversion of 58
in 57 through copper catalysis, but not vice-versa.
À
À
À
C N bond formation is under kinetic control, whereas C C
cyclization is thermodynamically more favored. Neverthe-
less, the reaction using the Cu catalyst is probably leading
directly to the C3 cyclization product, because the reaction
time for the conversion of 58 into 57 was much longer than
for the direct cyclization.
The C C cyclization product, after deprotection of the
amine, led to a formal total synthesis of aspidospermidine
(10), since the free amine had been used in the total synthe-
sis reported by Wenkert.[18d] On the other hand, the scaffold
obtained from the N3 attack, corresponded to the core of
the gonioma alkaloid goniomitine. Starting from a protected
tryptophol derivative, we were able to synthesize goniomi-
tine in 13 linear steps and 11% overall yield using this cycli-
zation strategy.[7g]
For a better understanding of the observed cyclization
selectivity, geometry optimization calculations were per-
formed on the four possible cyclization products (see
Figure 1), namely the cis and trans diastereomers arising
from C3 cyclization (compounds 57 and 59, respectively),
and the cis and trans diastereomers deriving from N1 cycli-
zation (compounds 58 and 60, respectively). All geometry
optimization calculations[28] were performed at the
Variation of the heterocyclic part was then tested. The
synthesized 3-methylindole (17), 2-pyrrole (18), 2-pyran
(19), and 2-thiophene (20) derivatives were therefore sub-
mitted to several cyclization conditions (Scheme 19). All the
conditions optimized for the cyclization gave polymerization
DFTACHTUNGTRENNUNG(B3LYP)/6-31+G* level in redundant internal coordi-
nates[29] by using the Gaussian 09 (G09) suite of programs.[30]
The nature of all found critical points (geometries are re-
ported in the Supporting Information) was assessed at the
same level of theory and accuracy by means of frequency
calculations. In all calculations the solvent (acetonitrile)
effect was taken into account at the implicit level using the
SCRF-CPM method (G09 defaults were used).[31]
The cis-fused ring systems 57 and 58 were found to be
substantially lower in energy than their trans counterparts
59 (DE57À59 =11.45 KcalmolÀ1) and 60 (DE58À60 =9.97 Kcal
molÀ1); this can be easily ascribed to the increased stability
of the chair conformation adopted by the piperidine ring in
compounds 57 and 58, compared with the less favorable
Scheme 19. Influence of the heterocycle structure on the cyclization.
Chem. Eur. J. 2011, 17, 14527 – 14538
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14535

J. Waser et al.
Table 3. Selective cyclization of acyclic carbamate 22.
or a-amidoalcohol side products resulting from the attack of
water on the iminium intermediate with 3-methylindole
(17), 2-pyrrole (18), and 2-thiophene (20) derivatives. On
the other hand, dihydropyran derivative 19 gave the cyclic
compound after a brief treatment with catalytic para-tolue-
nesulfonic acid in acetonitrile. Again, the cyclization gave
only the cis diastereoisomer. These results demonstrated
that the substitution pattern of the heterocycle is very im-
portant for the success of the cyclization reaction with imini-
um. For future work, it would be important to further modu-
late the reactivity of the iminium intermediate to extend the
scope of the reaction.
Entry[a]
Catalyst
[mol%]
Solvent
t
Products
[%]
N
ACHTUNGTNER[NUNG min]
1
2
3
TsOH
Cu(OTf)2
TsOH
CH2Cl2
CH3CN
dry CH3CN
15
15
15
65+67
G
65 (70)[b]
66 (75)[c]
[a] Reaction run with 22 (5–30 mg, 12–70 mmol) in 0.5–3 mL solvent at
RT. [b] Yield calculated by NMR spectroscopy, see the Supporting Infor-
mation for details. [c] Isolated product yield.
When we moved to acyclic carbamate 21, the standard op-
À
timized conditions to obtain C C cyclization generated an
aldehyde side product derived from the opening of the cy-
clopropane and hydrolysis of the intermediate iminium
(Table 2, entry 1). To reduce the amount of water, which is
cyclization product was now also possible (Table 3). Never-
theless, we found that the principles discovered in the case
of the more rigid bicyclic cyclopropanes could also be used
in this case. The C-Cyclization product 65 was isolated after
Table 2. Cyclization of acyclic carbamate 21.
only 15 min in 70% yield using CuACHTNUTRGNEUNG(OTf)2 in acetonitrile
(Table 3, entry 2), whereas the combination of para-toluene-
À
sulfonic acid and acetonitrile afforded the C N cyclization
product in 75% yield (Table 3, entry 3). In the latter case,
however, hydrolysis of the formed aminal was observed to
give the corresponding half aminal 66.
Entry[a]
Catalyst
[mol%]
Solvent
t
Observed
products [%]
Despite the report of Yadav on the cyclization of oxycy-
clopropane using a large excess of SnCl4,[7e] cyclization of
the oxycyclopropane derivatives was not efficient under our
conditions. Indeed, all the conditions tested with oxycyclo-
propane 23 gave side products derived from the water
attack on the three-membered ring or, in dry conditions,
non-defined decomposition, confirming the high sensitivity
of this class of compounds observed during the synthesis.
However, the stoichiometric use of BF3·Et2O in combination
with a non-coordinating solvent gave a useful yield of N-cyc-
lization product 68 (Scheme 20). The fact that oxycyclopro-
panes preferred reaction with water or cyclization on nitro-
gen over C-cyclization is in accordance with a potentially
harder oxonium intermediate.
G
A
1
2
3
4
5
6
7
8
TsOH (20)
TsOH (20)
TFA (20)
TFA (20)
TFA (20)
TsOH (20)
TsOH (20)
TsOH (20)
CH2Cl2
10
30
30
90
720
30
180
180
64
CH2Cl2, M.S.
dry CH3CN
dry CH3CN
dry CH3CN
CH3CN
62+63
21+62
63 (89)[b]
63+62
62 (36)[b]
62 (20)[b]
62+dec.
CH3CN
CH2Cl2
[a] Reaction run with 21 (10–50 mg, 20–120 mmol) in 1–5 mL solvent at
RT. [b] Yield calculated by using NMR spectroscopy, see the Supporting
Information for further details. M.S.=molecular sieves.
responsible for the iminium hydrolysis, we carried out the
reaction in presence of molecular sieves (Table 2, entry 2).
After 30 min, the reaction was not complete and we ob-
served the presence of aromatic compound 62 derived from
elimination of benzyl-carbamoylbenzylamide. Fortunately,
reducing the acid strength (trifluoroacetic acid) and using a
more coordinating solvent such as acetonitrile, we could iso-
late the C-cyclization product after 90 min in 63% yield of
the isolated product (89% crude yield by NMR spectrosco-
py, Table 2, entry 4). Careful control of the reaction time
was also important, since the compound 63 could easily aro-
matize (Table 2, entry 5). By using para-toluenesulfonic acid
in acetonitrile, it was also possible to obtain the aromatic
compound 62 in 36% yield (Table 2, entry 6). With longer
reaction times, the yield was lower due to decomposition
(Table 2, entries 7 and 8).
Scheme 20. Cyclization of oxocyclopropane 23.
Conclusion
In summary, we have shown a further expansion of the
scope of cyclization reactions of activated cyclopropanes, in-
cluding the formal homo-Nazarov process. We proposed two
different strategies to modulate the polarization of the
three-membered ring to increase its reactivity under catalyt-
The use of unprotected indole 22 made control over the
reaction even more challenging, as formation of the C N
À
14536
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 14527 – 14538

Formal Homo-Nazarov and Other Cyclization Reactions
FULL PAPER
ic conditions. In the first approach, we introduced an ester
group on the cyclopropane, allowing an easy access to six-
membered rings bearing not only electron-rich, but also
electron-poor aromatic substituents. Moreover, the presence
of an additional carbonyl group allowed a first proof of prin-
ciple for asymmetric control in the cyclization reaction,
probably through a two-point binding interaction with the
chiral Lewis acid catalyst.
[1] a) S. Kobayashi, K. A. Jørgensen, Cycloaddition Reactions in Or-
ganic Synthesis, Wiley-VCH, Weinheim, 2002; b) S. M. Ma, Hand-
book of Cyclization Reactions, Wiley-VCH, Weinheim, 2009.
[2] I. N. Nazarov, I. I. Zaretskaya, Izv. Akad. Nauk SSSR, Ser. Khim
1941, 211.
[3] K. L. Habermas, S. E. Denmark, T. K. Jones, Org. React. (N.Y.,
Engl. Transl.) 1994, 45, 1.
[4] a) S. Giese, F. G. West, Tetrahedron 2000, 56, 10221; b) Y. Wang,
B. D. Schill, A. M. Arif, F. G. West, Org. Lett. 2003, 5, 2747; c) C.
Bee, E. Leclerc, M. A. Tius, Org. Lett. 2003, 5, 4927; d) W. He, X. F.
Sun, A. J. Frontier, J. Am. Chem. Soc. 2003, 125, 14278; e) M. Janka,
W. He, A. J. Frontier, R. Eisenberg, J. Am. Chem. Soc. 2004, 126,
6864; f) W. A. Batson, D. Sethumadhavan, M. A. Tius, Org. Lett.
2005, 7, 2771; g) J. A. Malona, J. M. Colbourne, A. J. Frontier, Org.
Lett. 2006, 8, 5661; h) L. Zhang, S. Wang, J. Am. Chem. Soc. 2006,
128, 1442; i) J. M. Tang, S. Bhunia, S. M. A. Sohel, M. Y. Lin, H. Y.
Liao, S. Datta, A. Das, R. S. Liu, J. Am. Chem. Soc. 2007, 129,
15677; j) I. Walz, A. Bertogg, A. Togni, Eur. J. Org. Chem. 2007,
2650; k) W. He, I. R. Herrick, T. A. Atesin, P. A. Caruana, C. A.
Kellenberger, A. J. Frontier, J. Am. Chem. Soc. 2008, 130, 1003;
l) M. Kawatsura, Y. Higuchi, S. Hayase, M. Nanjo, T. Itoh, Synlett
2008, 8, 1009; m) A. P. Marcus, A. S. Lee, R. L. Davis, D. J. Tantillo,
R. Sarpong, Angew. Chem. 2008, 120, 6479; Angew. Chem. Int. Ed.
2008, 47, 6379; n) M. Amere, J. Blanchet, M. C. Lasne, J. Rouden,
Tetrahedron Lett. 2008, 49, 2541; o) T. Jin, Y. Yamamoto, Org. Lett.
2008, 10, 3137; p) G. Lemiere, V. Gandon, K. Cariou, A. Hours, T.
Fukuyama, A. L. Dhimane, L. Fensterbank, M. Malacria, J. Am.
Chem. Soc. 2009, 131, 2993; q) T. Vaidya, A. C. Atesin, I. R. Her-
rick, A. J. Frontier, R. Eisenberg, Angew. Chem. 2010, 122, 3435;
Angew. Chem. Int. Ed. 2010, 49, 3363; r) P. A. Wender, R. T. Stemm-
ler, L. E. Sirois, J. Am. Chem. Soc. 2010, 132, 2532; for reviews, see:
s) M. A. Tius, Eur. J. Org. Chem. 2005, 2193; t) A. J. Frontier, C.
Collison, Tetrahedron 2005, 61, 7577; u) H. Pellissier, Tetrahedron
2005, 61, 6479; v) W. Nakanishi, F. G. West, Curr. Opin. Drug
Discov. Devel. 2009, 12, 732; w) S. Thompson, A. G. Coyne, P. C.
Knipe, M. D. Smith, Chem. Soc. Rev. 2011, 40, 4217; x) T. Vaidya, R.
Eisenberg, A. J. Frontier, ChemCatChem, 2011, 3, 1531.
In the second approach, we studied the influence of heter-
oatoms on the cyclopropane. The use of aminocyclopro-
panes in particular was key in the application of the meth-
odology in the total synthesis of natural alkaloids. General
synthetic strategies were developed for the synthesis of
donor–acceptor cyclopropanes with indole-, pyrrole-, pyran-,
or thiophene-ketones as electron-withdrawing groups and
cyclic or acyclic carbamates, or ethers as donor groups. This
broad range of substrates allowed us to better assess the
scope and limitation of the methodology, and led to the fol-
lowing conclusions: 1) both the nucleophilicity and the sub-
stitution pattern of the heterocycle were essential for a suc-
cessful cyclization. If the fit is not ideal, reaction with water
or polymerization became favored. 2) By using a free indole
as a nucleophile, the outcome of the cyclization is depen-
dent on the electrophilic character of the formed carbocat-
ionic intermediate. A benzylic cation gave C3 cyclization ex-
clusively, whereas oxonium favored N1 cyclization. For imi-
nium intermediates, it was possible to switch the regioselec-
tivity using a copper catalyst in acetonitrile for C3 cycliza-
tion and para-toluenesulfonic acid in dichloromethane for
N1 cyclization. Both experiments and calculations showed
that C3 cyclization led to the thermodynamic product.
3) Acyclic carbamates could be successfully cyclized, but
they are more sensitive substrates. Mild conditions are re-
quired to prevent a facile aromatization of the product.
However, this last reaction manifold could have potential
for the synthesis of substituted carbazoles.
[5] a) V. K. Aggarwal, A. J. Beffield, Org. Lett. 2003, 5, 5075; b) G. X.
Liang, S. N. Gradl, D. Trauner, Org. Lett. 2003, 5, 4931; c) G. X.
Liang, D. Trauner, J. Am. Chem. Soc. 2004, 126, 9544; d) I. Walz, A.
Togni, Chem. Commun. 2008, 4315; e) M. Rueping, W. Ieawsuwan,
A. P. Antonchick, B. J. Nachtsheim, Angew. Chem. 2007, 119, 2143;
Angew. Chem. Int. Ed. 2007, 46, 2097; f) P. Cao, C. Deng, Y. Y.
Zhou, X. L. Sun, J. C. Zheng, Z. W. Xie, Y. Tang, Angew. Chem.
2010, 122, 4565; Angew. Chem. Int. Ed. 2010, 49, 4463; g) A. K.
Basak, N. Shimada, W. F. Bow, D. A. Vicic, M. A. Tius, J. Am.
Chem. Soc. 2010, 132, 8266.
In conclusion, we have shown the versatility of the formal
homo-Nazarov reaction and other cyclization processes
using keto-ester activated cyclopropanes, aminocyclopro-
panes and oxycyclopropanes. We have extended the scope
of the reaction and have shown the first preliminary exam-
ple of asymmetric induction. Currently, our research in this
transformation is focused on new applications in total syn-
thesis, as well as increasing the enantioselectivity by the use
of other catalysts.
[6] A. D. Walsh, Nature 1947, 159, 712.
[7] a) W. S. Murphy, S. Wattanasin, Tetrahedron Lett. 1980, 21, 1887;
b) W. S. Murphy, S. Wattanasin, J. Chem. Soc. Perkin Trans. 1 1981,
2920; c) W. S. Murphy, S. Wattanasin, J. Chem. Soc. Perkin Trans. 1
1982, 1029; d) O. Tsuge, S. Kanemasa, T. Otsuka, T. Suzuki, Bull.
Chem. Soc. Jpn. 1988, 61, 2897; e) V. K. Yadav, N. V. Kumar, Chem.
Commun. 2008, 3774; f) F. De Simone, J. Andres, R. Torosantucci, J.
Waser, Org. Lett. 2009, 11, 1023; g) F. De Simone, J. Gertsch, J.
Waser, Angew. Chem. 2010, 122, 5903; Angew. Chem. Int. Ed. 2010,
49, 5767; h) D. V. Patil, L. H. Phun, S. France, Org. Lett. 2010, 12,
5684; i) L. H. Phun, D. V. Patil, M. A. Cavitt, S. France, Org. Lett.
2011, 13, 1952; for a review, see: j) F. De Simone, J. Waser, Chimia
2009, 63, 162.
Acknowledgements
The EPFL is acknowledged for financial support. F.B. is a recipient of
IEF Marie Curie fellowship (no. 253274). The following EPFL students
and technical assistants are thanked for the synthesis of starting material,
chiral ligands, or reagents: Letitzia Gullifa, Frederic Steimer, and Valen-
tin Manzanares.
[8] a) D. V. Patil, M. A. Cavitt, P. Grzybowski, S. France, Chem.
Commun. 2011, 47, 10278; b) D. V. Patil, M. A. Cavitt, S. France,
Heterocycles. 2012, 84, DOI: 10.3987/COM-11-S(P)87.
[9] a) H. U. Reissig, R. Zimmer, Chem. Rev. 2003, 103, 1151; b) F.
Gnad, O. Reiser, Chem. Rev. 2003, 103, 1603; c) M. Yu, B. L. Pagen-
kopf, Tetrahedron 2005, 61, 321; d) C. A. Carson, M. A. Kerr, Chem.
Soc. Rev. 2009, 38, 3051; e) F. De Simone, J. Waser, Synthesis 2009,
Chem. Eur. J. 2011, 17, 14527 – 14538
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14537

J. Waser et al.
3353; f) T. P. Lebold, M. A. Kerr, Pure Appl. Chem. 2010, 82, 1797;
g) T. F. Schneider, D. B. Werz, Org. Lett. 2011, 13, 1848.
[10] a) G. Stork, M. Marx, J. Am. Chem. Soc. 1969, 91, 2371; b) G. Stork,
M. Gregson, J. Am. Chem. Soc. 1969, 91, 2373; c) G. Stork, P. A.
Grieco, J. Am. Chem. Soc. 1969, 91, 2407; d) E. J. Corey, R. D. Bal-
anson, Tetrahedron Lett. 1973, 14, 3153; e) S. Danishefsky, Acc.
Chem. Res. 1979, 12, 66.
[24] See the Supporting Information for further details.
[25] In contrast to the case of cyclic enamide, this cyclopropanation was
best performed at a higher temperature and with a fast addition of
carbene precursor. Furthermore, isomerization of the oxycyclopro-
pane was faster.
[26] Yield of 23 is based on the amount of pure 54 in the starting materi-
al.
[11] T. Saget, Masters Thesis, Ecole Polytechnique Fꢀdꢀrale de Lausanne
and Ecole Nationale Supꢀrieure de Chimie de Montpellier, 2009 .
[12] a) P. D. Pohlhaus, S. D. Sanders, A. T. Parsons, W. Li, J. S. Johnson, J.
Am. Chem. Soc. 2008, 130, 8642; b) A. T. Parsons, J. S. Johnson, J.
Am. Chem. Soc. 2009, 131, 14202.
[27] See Ref. [7g] for a more detailed discussion of the optimization of
the cyclization reaction.
[28] B. Schlegel, J. Comput. Chem. 2003, 24, 1514.
[29] C. Peng, P. Y. Ayala, H. B. Schlegel, M. J. Frisch, J. Comput. Chem.
1996, 17, 49.
[13] P. Wyatt, A. Hudson, J. Charmant, A. G. Orpen, H. Phetmung, Org.
Biomol. Chem. 2006, 4, 2218.
[14] F. Gonzꢃlez-Bobes, M. D. B. Fenster, S. Kiau, L. Kolla, S. Kolotu-
chin, M. Soumeillant, Adv. Synth. Catal. 2008, 350, 813.
[15] C. Perreault, S. R. Goudreau, L. E. Zimmer, A. B. Charette, Org.
Lett. 2008, 10, 689.
[16] M. D. Ganton, M. A. Kerr, J. Org. Chem. 2004, 69, 8554.
[17] See the Supporting Information for a detailed procedure.
[18] a) E. Wenkert, T. Hudlicky, H. D. H. Showalter, J. Am. Chem. Soc.
1978, 100, 4893; b) E. Wenkert, Acc. Chem. Res. 1980, 13, 27; c) E.
Wenkert, T. D. J. Halls, L. D. Kwart, G. Magnusson, H. D. H. Show-
alter, Tetrahedron 1981, 37, 4017; d) E. Wenkert, T. Hudlicky, J. Org.
Chem. 1988, 53, 1953.
[30] Gaussian 09, Revision A. 02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V.
Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X.
Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Son-
nenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven,
J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J.
Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J.
Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,
J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A.
Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, ꢄ.
Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, and D. J. Fox,
Gaussian, Inc., Wallingford CT, 2009.
[19] L. Larquetoux, N. Ouhamou, A. Chiaroni, Y. Six, Eur. J. Org. Chem.
2005, 4654.
[20] P. A. Grieco, M. D. Kaufman, J. Org. Chem. 1999, 64, 7586.
[21] The choice of coupling agent was important when considering the
instability of the aminocyclopropanes. The coupling involving
DMTMM has the added advantage of simple purification during
work up without requiring chromatographic purification.
[22] A. R. Katritzky, K. Akutagawa, Tetrahedron Lett. 1985, 26, 5935.
[23] a) G. M. Wieber, L. S. Hegedus, B. Akermark, E. T. Michalson, J.
Org. Chem. 1989, 54, 4649; b) A. Kamatani, L. E. Overman, J. Org.
Chem. 1999, 64, 8743.
[31] J. Tomasi, B. Mennucci, R. Cammi, Chem. Rev. 2005, 105, 2999.
[32] F. De Simone, J. Waser, Synlett 2011, 5, 589.
Received: August 19, 2011
Published online: November 23, 2011
14538
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 14527 – 14538