Enantioselective formal [4+2] cycloadditions to 3-nitroindoles by trienamine catalysis: Synthesis of chiral dihydrocarbazoles

Article abstract of DOI:10.1002/anie.201509693

The first enantioselective formal [4+2] cycloadditions of 3-nitroindoles are presented. By using 3-nitroindoles in combination with an organocatalyst, chiral dihydrocarbazole scaffolds are formed in moderate to good yields (up to 87 %) and enantioselectivities (up to 97 % ee). The reaction was extended to include enantioselective [4+2] cycloadditions of 3-nitrobenzothiophene. The reaction proceeds through a [4+2] cycloaddition/elimination cascade under mild reaction conditions. Furthermore, a diastereoselective reduction of an enantioenriched cycloadduct is presented. The mechanism of the reaction is discussed based on experimental and computational studies. A formal event: Enantioselective formal [4+2] cycloadditions of 3-nitroindoles in the presence of an organocatalyst are presented. Chiral dihydrocarbazole scaffolds are formed in moderate to good yields. The reaction also proceeds with 3-nitrobenzothiophene. The mechanism of the reaction is discussed based on experimental and computational studies.

Full text of DOI:10.1002/anie.201509693

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201509693
German Edition: DOI: 10.1002/ange.201509693
Cycloaddition
Enantioselective Formal [4+2] Cycloadditions to 3-Nitroindoles by
Trienamine Catalysis: Synthesis of Chiral Dihydrocarbazoles
Yang Li+, Fernando Tur+, Rune Pagh Nielsen, Hao Jiang, Frank Jensen, and
Karl Anker Jørgensen*
Abstract: The first enantioselective formal [4+2] cycloaddi-
tions of 3-nitroindoles are presented. By using 3-nitroindoles in
combination with an organocatalyst, chiral dihydrocarbazole
scaffolds are formed in moderate to good yields (up to 87%)
and enantioselectivities (up to 97% ee). The reaction was
extended to include enantioselective [4+2] cycloadditions of 3-
nitrobenzothiophene. The reaction proceeds through a [4+2]
cycloaddition/elimination cascade under mild reaction condi-
tions. Furthermore, a diastereoselective reduction of an enan-
tioenriched cycloadduct is presented. The mechanism of the
reaction is discussed based on experimental and computational
studies.
Figure 1. The 4,9-dihydro-1H-carbazole alkaloid scaffold in natural
products.
in good yields and high enantioselectivities. We demonstrate
that the reaction concept can also be extended to 3-nitro-
benzothiophene.
T
he ubiquity of the indole heterocycle in natural products
Different approaches for the formation of chiral dihydro-
carbazole scaffolds can be envisioned. The most direct
approach would be the cycloaddition of a diene to 2,3-
indolyne (Scheme 1, top). However, 2,3-indolynes and related
and pharmaceuticals underscores the importance of this
privileged motif.[1] Therefore, the preparation of substituted
indole derivatives has received significant attention over the
years.[2] The vast majority of these methods involve the well-
known nucleophilic character of indole.[2] Exploiting the
umpolung[3] reactivity of indole would provide complemen-
tary synthetic methods to access complex indole derivatives.[4]
Recently, electrophilic indoles have been disclosed as prom-
ising electron-deficient alkenes in the asymmetric catalysis
area.[5] In this context, we hypothesized that combining
electrophilic indoles and trienamine catalysis[6] might provide
access to new highly functionalized indole scaffolds which are
difficult to obtain by conventional strategies.
The synthesis of a key heterocyclic moiety, such as
carbazole, has been broadly studied.[7] However, the synthe-
ses of the related tetrahydro-[8] and dihydrocarbazoles[9] have
received less attention because of the limited synthetic
procedures available, even though, the unique dihydrocarba-
zole scaffold constitutes a compound class (dihydrotubingen-
sins A and B; Figure 1) which has been isolated from natural
sources.[10] Consequently, asymmetric organocatalytic proto-
cols providing access to these compounds are highly desirable
for the development of reactions for skeletal diversity.[11]
Herein, we disclose a catalytic enantioselective formal
[4+2] cycloaddition/elimination cascade of 2,4-dienals with 3-
nitroindoles, thus providing chiral dihydrocarbazole scaffolds
Scheme 1. Synthetic strategy for the formation of dihydrocarbazoles.
LG=leaving group.
heteroarynes have, according to the best of our knowledge,
not been generated and applied in catalytic asymmetric
reactions.[12]
We envisioned that an electron-deficient indole[13] might
be a suitable candidate for reacting under organocatalytic
asymmetric conditions with an electron-rich diene function-
ality, which is generated in situ, by using the trienamine
approach.[6] To be able to control the reaction course we
assumed that installing a nitro functionality at the 3-position
of the indole,[5] and use of a hydrogen-bonding organo-
catalyst[14] would direct the dienophile to approach the amino-
activated diene functionality in a stereoselective manner
(Scheme 1, bottom). The strategy was expected to induce
LUMO lowering of the dienophile, thus making it more
reactive with the in situ generated diene. We anticipated that
the role of the catalyst would be threefold: activation of both
[*] Y. Li,[+] Dr. F. Tur,[+] R. P. Nielsen, Dr. H. Jiang, Prof. Dr. F. Jensen,
Prof. Dr. K. A. Jørgensen
Department of Chemistry, Aarhus University
8000 Aarhus C (Denmark)
E-mail: kaj@chem.au.dk
[+] These authors contributed equally to this work.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201509693.
1020
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1020 –1024

Angewandte
Communications
Chemie
substrates, induction of stereoselectivity, and promotion of
results in various solvents and with various bases (entries 6–
the elimination of the nitro group.
13). The best results were obtained in CH2Cl2 with DABCO
as the base, with 4a in 81% ee (entry 6). It was notable that
the elimination of HNO2 was normally slow at 08C. However,
by increasing the temperature to 408C, after completion of
the cycloaddition step, the elimination was accelerated with-
out observing any deleterious effect on the enantioselectivity
of 4a. Moreover, the multiple roles of the additive DABCO
were found to be essential in the reaction, thus preventing
potential self-aggregation of 3c, and increasing conversion
and enantioselectivity (entries 6 and 12). We also tested other
aminocatalysts for the model reaction and used, for example,
the TMS-protected diarylprolinol catalysts, which gave lower
conversions and enantioselectivities, required longer reaction
times, and higher reaction temperatures, thus verifying the
threefold role of 3 (see the Supporting Information).
To study the steric and electronic effects, the reaction of
the 3-nitroindoles 1 with 2a was investigated with various N-
protecting groups in 1 (Scheme 2). From the results, it appears
that the reaction is not dependent on the N-protecting group
in 1, as greater than 95% conversion was observed and the
enantioselectivity was rather consistent (76–81% ee) for all
four N-protecting groups studied.
Towards our aim, the model reaction was envisioned to
proceed by reaction of the 3-nitroindole 1a with (E)-5-
methylhexa-2,4-dienal (2a) in the presence of a hydrogen-
bond directing organocatalyst (3), which should have the
properties mentioned above (Table 1). Furthermore, a base
additive would also be needed as the strategy is based on the
elimination of HNO2 from 1a.[15] Table 1 presents some
representative screening results. The reaction was dependent
on the additive as greater than 95% conversion was obtained
when a base was used, and no reaction took place when, for
example, N,N-diethylacetamide (DEA) was present (entries 1
and 2). A basic additive such as DABCO was required for the
reaction to take place, and in the presence of the squaramide-
based catalyst 3a high conversions were found in both toluene
and CH2Cl2 (entries 2–4). The catalyst 3a gave up to 79% ee
of the cycloaddition product 4a (entry 4), while the thiourea-
based catalyst 3b gave the same enantioselectivity and
increased conversion to greater than 95% (entry 5). The
corresponding urea-based catalyst 3c showed promising
Table 1: Screening and optimization of reaction conditions for the formal
[4+2] cycloaddition to 3-nitroindole 1a.[a]
Scheme 2. Variation of the N-protecting group in the 3-nitroindoles 1.
Conversion of the limiting reagent was measured by 1H NMR analysis
of the crude reaction mixture. The ee values were determined by chiral-
phase UPC2 after reduction of 4 into the alcohol. Boc=tert-butoxycar-
bonyl, Cbz=carboxybenzyl, Ts =4-toluenesulfonyl.
Entry Catalyst Additive Solvent
T [8C] Conv. [%][b] ee [%][c]
Next, a variety of 3-nitro-1H-indole-1-carboxylates (1)
and 2,4-dienals (2) were tested to study the scope of the
reaction (Table 2). It was observed that better yields and
enantioselectivities were obtained when 2,4-dienals other
than 2a were used for the reaction with 1a (products 4e–h).
Notably, 4-phenylhepta-2,4-dienal gave excellent enantiose-
lectivity (97% ee) and a nearly pure diastereoisomer (4h).
Installing a methyl substituent at the 6-position of 1a had
a small effect on the outcome of the reaction (4i). Substrates
with an electron-withdrawing substituent at the 5-position
showed consistent reactivity and stereoselectivity when
reacting with different 2,4-dienals and provided 89–90% ee
(4j–n). During the process of expanding the substrate scope,
we observed that 4-phenylhepta-2,4-dienal gave a better yield
and enantioselectivity than 2a, which we used for optimizing
the reaction conditions. Therefore, we tested the reaction
between 3-nitroindoles, bearing N-Boc and N-Cbz protecting
groups, and 4-phenylhepta-2,4-dienal. As it appears from
Table 2, the N-Boc- and N-Cbz-protected 3-nitroindole gave
nearly the same enantioselectivity.
1
2
3a
3a
3a
3a
3b
3c
3c
3c
3c
DEA[d]
toluene
RT <1
RT >95
–
À72
À55
À79
79
81
79
75
–
–
79
69
76
DABCO toluene
DABCO toluene
DABCO CH2Cl2
DABCO CH2Cl2
DABCO CH2Cl2
DABCO CHCl3
DABCO ClCH2CH2Cl
3[e]
4
0
0
68
85
5
6
7
8
0
0
0
0
>95
>95
>95
>95
<1
9
DEA[d]
DIPEA
Et3N
CH2Cl2
CH2Cl2
CH2Cl2
0
10[e] 3c
0
37
11[e] 3c
0
0
>95
85
12
3c
DABCO[f] CH2Cl2
DABCO CH2Cl2
13[g] 3c
À10
>95
[a] All reactions were performed using 1a (0.05 mmol), 2a (0.1 mmol),
0.2 mL solvent, reaction time: 24 h. Then, the reaction mixture was
heated up to 408C to complete the elimination of HNO2. [b] Conversion
1
of limiting reagent was measured by H NMR analysis of the crude
reaction mixture after 24 h. [c] The ee value was determined by chiral-
phase UPC2 after reduction of 4a into the corresponding alcohol.
[d] DEA=N,N-diethylacetamide. [e] Reaction time: 72 h. [f] 20 mol%
DABCO was used. [g] Reaction time: 48 h. DABCO=1,4-diazobicyclo-
[2.2.2]octane, TFA=trifluoroacetic acid.
Angew. Chem. Int. Ed. 2016, 55, 1020 –1024
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1021

Angewandte
Communications
Chemie
Table 2: Scope of the formal [4+2] cycloaddition of the 3-nitroindoles 1.[a]
the presence of 3c and DABCO, the results presented in
Table 3 show similar results to those obtained for the
reactions employing 1a as the substrate. The highest enantio-
selectivities were found for the products 6c and 6d, with
98%and 95% ee, respectively.
The utility of the enantioenriched cycloadducts 4 is
outlined in Scheme 3. The diastereoselective reduction of
Table 3: Scope of the formal [4+2] cycloaddition of 5.[a]
[a] All reactions were performed using 5 (0.1 mmol), 2 (0.2 mmol), 3c
(0.02 mmol), DABCO (0.1 mmol) in 0.4 mL CH2Cl2. All ee values were
determined by chiral-phase UPC2. The absolute configuration was
assumed to be identical to the cycloadducts obtained for 1 as the
substrate. See the Supporting Information. [b] Product isolation and
determination of the ee value were carried out by converting the
cycloadduct into the alcohol. [c] Product isolation and determination of
the ee value were carried out by converting the cycloadduct into the
Wittig adduct 6c. The d.r. value was determined by 1H NMR analysis of
the crude reaction mixture.
the olefin and cleavage of the Cbz-group in 4p were
performed in a single step, thus affording the cycloadduct 7
in high yield and excellent stereocontrol. Notably, the cyclo-
adduct 7 is similar to the core structure of the natural
compound aristolasicone, which has only been synthesized in
racemic form.[17]
Experimental observations and computational investiga-
tions have provided some insight into the mechanism for the
cycloaddition of 3-nitroindoles and 3-nitrobenzothiophene.
During the optimization of the reaction conditions (Table 1),
we observed an intermediate as a diastereoisomeric mixture
(Scheme 1, lower left). For the reactions described in Tables 2
and 3, and Scheme 2, this intermediate was also observed by
NMR spectroscopy for several of the reactions, while for the 5
the intermediate was observed for 6a,b,d in about a 1:1 d.r.
[a] All reactions were performed using 1 (0.1 mmol), 2 (0.2 mmol), 3c
(0.02 mmol), DABCO (0.1 mmol) in 0.4 mL CH2Cl2. All ee values were
determined by chiral-phase UPC2. Absolute configuration was deter-
mined by X-ray crystallography analysis of 4k and the remaining
structures were assigned by analogy. See the Supporting Information.
[b] Product isolation and determination of the ee value were carried out
by converting the cycloadduct into the alcohol. [c] The ee value was
determined by converting 4h into the alcohol. The d.r. was determined by
1H NMR spectroscopy of the crude reaction mixture. [d] Product
isolation and determination of the ee value were carried out by
converting the cycloadduct into the Wittig adduct 4o.
To evaluate the practicality of the present synthetic
methodology we performed reactions on scales ranging
from 0.1 to 1.0 mmol. These experiments gave the following
results for 4h (d.r. > 20:1): 0.1 mmol: 76% yield, 97% ee;
0.5 mmol: 84% yield, 97% ee and 1.0 mmol: 85% yield, 97%
ee. These scale-up experiments thus demonstrate the reli-
ability of the method.
To our surprise, the reaction proceeded successfully with
3-nitrobenzothiophene (5) as the dienophile[16] under mild
reaction conditions, which are in striking contrast to the only
previous reaction conditions for the [4+2] cycloaddition to 3-
nitrobenzothiophene, where up to 1808C were required.[16a]
For the reaction of 5 with four different types of 2,4-dienals in
Scheme 3. Diastereoselective reduction and cleavage of the Cbz group
in the enantioenriched cycloadduct 4p.
1022
www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1020 –1024

Angewandte
Communications
Chemie
The role of 3c acting also as a hydrogen-bonding donor is
roles of the hydrogen-bonding catalyst 3c: it activates both
supported experimentally, as the installation of competing
hydrogen-bonding functionalities, such as nitro and cyano
groups at the 5-position of 3-nitroindole resulted in no
reaction, even at 408C.
substrates in the addition step and promotes the elimination
of the nitro group in the second step, thus reforming the
double bond.
In conclusion, the first enantioselective formal [4+2]
cycloaddition to 3-nitroindoles is reported to proceed by using
organocatalysis. The combination of 3-nitroindoles and 2,4-
dienals in the presence of an organocatalyst leads to the
formation of chiral dihydrocarbazole scaffolds in moderate to
good yields (up to 87%) and enantioselectivities (up to 97%
ee) under mild reaction conditions. This novel approach is
applicable to the enantioselective cycloaddition reactions for
3-nitrobenzothiophene, thereby providing the cycloaddition
products in up to 98% ee. Synthetic manipulation of the chiral
cycloadducts by a diastereoselective reduction of the olefin in
the product is also presented. Mechanistic studies based on
experimental observations and computational studies point
towards an asynchronous/stepwise addition followed by an
elimination in which the hydrogen-bonding catalyst plays
a pivotal role. The present development provides an alter-
native reaction concept for cycloaddition reactions of 2,3-
indolynes and 2,3-benzothiophynes, as the generation of these
heteroarynes have not yet been possible.
DFT calculations[18] (see the Supporting Information)
were performed to account for the role of the catalyst for the
activation of the dienophile (5 was used to reduce the number
of conformers). The electronic interaction leading to the
reaction is that between the LUMO of 5 and the HOMO of
the trienamine, represented as an s-cis-diene (Figure 2, top
Acknowledgements
This work was financially supported by Aarhus University
and the Carlsberg Foundation. Y.L. acknowledges the Chi-
nese Scholarship Council for a Ph.d. fellowship. F.T. acknowl-
edges the European Commission for a Marie Curie Intra
European Fellowship for Career Development (PIEF-GA-
2013-622413) within the 7th European Community Frame-
work Programme. Magnus E. Jensen is gratefully acknowl-
edged for performing X-ray analysis. Thanks are expressed to
a reviewer for suggesting that we perform computational
studies.
Figure 2. Selected results from the computational studies of the
transition states for the simplified model for the formal cycloaddition
of 5. Values within brackets are those for the uncatalyzed reactions.
left).[19] Calculations have shown that the terminal carbon
atom in the s-cis-diene moiety has the largest HOMO
amplitude and might point towards an asynchronous or
stepwise mechanism.[20] By activating 5 by hydrogen bonding
to 3c the LUMO is calculated to be lowered by 0.84 eV
compared to that of the uncatalyzed system. We used the
transition state for the addition of butadiene as a model for
the s-cis-diene moiety of the trienamine (Figure 2, top right).
The role of the hydrogen-bonding activation of 5 by 3c is
supported by the calculations as the activation energy is
reduced from 79 to 37 kJmolÀ1. Furthermore, the calculations
gave C1-C2 and C3-C4 bond lengths as 1.93 Š and 3.21 Š,
respectively, in the transition state in the presence of 3c (TS I,
Scheme 2, bottom left). The same bond lengths in the absence
of 3c are 1.96 Š and 2.61 Š, respectively. These calculations
indicate that the reaction without catalyst proceeds in a single
step, while the catalyzed reaction occurs in two steps.[20]
The role of DABCO as a base for the elimination step has
also been investigated (Figure 2, bottom right). In the
Keywords: asymmetric catalysis · cycloaddition · heterocycles ·
organocatalysis · synthetic methods
How to cite: Angew. Chem. Int. Ed. 2016, 55, 1020–1024
Angew. Chem. 2016, 128, 1032–1036
[1] a) R. J. Sundberg, The Chemistry of Indoles, Academic Press,
New York, 1970; b) R. J. Sundberg, Indoles, Academic Press, San
Diego, 1996.
[2] For selected reviews, see: a) M. Bandini, A. Eichholzer, Angew.
Chem. Int. Ed. 2009, 48, 9608; Angew. Chem. 2009, 121, 9786;
b) V. Sharma, P. Kumar, D. Pathak, J. Heterocycl. Chem. 2010,
47, 491; c) G. Bartoli, G. Bencivenni, R. Dalpozzo, Chem. Soc.
Rev. 2010, 39, 4449; d) M. Bandini, Org. Biomol. Chem. 2013, 11,
5206; e) R. Dalpozzo, Chem. Soc. Rev. 2015, 44, 742; f) M.
Bandini, A. Melloni, A. Umani-Ronchi, Angew. Chem. Int. Ed.
2004, 43, 550; Angew. Chem. 2004, 116, 560; g) M. Bandini, A.
Melloni, S. Tommasi, A. Umani-Ronchi, Synlett 2005, 1199;
h) T. B. Poulsen, K. A. Jørgensen, Chem. Rev. 2008, 108, 2903;
i) S. L. You, Q. Cai, M. Zeng, Chem. Soc. Rev. 2009, 38, 2190.
[3] D. Seebach, Angew. Chem. Int. Ed. Engl. 1979, 18, 239; Angew.
Chem. 1979, 91, 259.
À
presence of 3c, the bond length of the breaking C3 N bond
is calculated to be 2.61 Š in the transition state (TS II), and is
in sharp contrast to the value of 2.02 Š for the uncatalyzed
À
À
reaction. The C1 H 1.19 Š (1.56 Š) and N H 1.66 Š (1.18 Š)
bond lengths show similar changes in the presence and
absence of 3c (values in brackets are for the uncatalyzed
reaction). The catalyst lowers the activation energy from
98 kJmolÀ1 to 36 kJmolÀ1. These results support the multiple
Angew. Chem. Int. Ed. 2016, 55, 1020 –1024
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1023

Angewandte
Communications
Chemie
[4] C. C. J. Loh, D. Enders, Angew. Chem. Int. Ed. 2012, 51, 46;
Angew. Chem. 2012, 124, 46.
[12] For selected reviews, see: a) A. E. Goetz, N. K. Garg, J. Org.
Chem. 2014, 79, 846; b) A. E. Goetz, T. K. Shah, N. K. Garg,
Chem. Commun. 2015, 51, 34.
[13] For selected examples, see: a) B. Biolatto, M. Kneeteman, P.
Mancini, Tetrahedron Lett. 1999, 40, 3343; b) B. Biolatto, M.
Kneeteman, E. Paredes, P. Mancini, J. Org. Chem. 2001, 66, 3906.
[14] For the seminal work, see: J. P. Malerich, K. Hagihara, V. H.
Rawal, J. Am. Chem. Soc. 2008, 130, 14416.
[5] For recent examples using 3-nitroindoles in asymmetric catalysis,
see: a) A. Awata, T. Arai, Angew. Chem. Int. Ed. 2014, 53, 10462;
Angew. Chem. 2014, 126, 10630; b) J.-Q. Zhao, M.-Q. Zhou, Z.-J.
Wu, Z.-H. Wang, D.-F. Yue, X.-Y. Xu, X.-M. Zhang, W.-C. Yuan,
Org. Lett. 2015, 17, 2238.
[6] For selected reviews, see: a) K. L. Jensen, G. Dickmeiss, H.
Jiang, Ł. Albrecht, K. A. Jørgensen, Acc. Chem. Res. 2012, 45,
248; b) J.-L. Li, T.-Y. Liu, Y.-C. Chen, Acc. Chem. Res. 2012, 45,
1491; c) E. Arceo, P. Melchiorre, Angew. Chem. Int. Ed. 2012, 51,
5290; Angew. Chem. 2012, 124, 5384; d) I. D. Jurberg, I.
Chatterjee, R. Tannert, P. Melchiorre, Chem. Commun. 2013,
49, 4869; e) I. Kumar, P. Ramaraju, N. A. Mir, Org. Biomol.
Chem. 2013, 11, 709.
[15] a) L. A. Kaplan, N. E. Burlinson, J. Org. Chem. 1972, 37, 3932;
b) D. Seebach, M. S. Hoekstra, G. Protschuk, Angew. Chem. Int.
Ed. Engl. 1977, 16, 321; Angew. Chem. 1977, 89, 334; c) P.
Bakuzis, M. L. F. Bakuzis, T. F. Weinartner, Tetrahedron Lett.
1978, 19, 2371; d) F. Terrier, R. Goumont, M.-J. Pouet, J.-C.
ˇ
HallØ, J. Chem. Soc. Perkin Trans. 2 1995, 1629; e) S. Marchalín,
B. Baumlovµ, P. Baran, H. Oulyadi, A. Daïch, J. Org. Chem.
2006, 71, 9114; f) R. López, M. Zalacain, C. Palomo, Chem. Eur.
J. 2011, 17, 2450.
[7] a) A. W. Schmidt, K. R. Reddy, H.-J. Knçlker, Chem. Rev. 2012,
112, 3193; b) H.-J. Knçlker, K. R. Reddy, Chem. Rev. 2002, 102,
4303.
[16] a) C. D. Della Rosa, P. Mancini, M. Kneeteman, A. F. Lopez
Baena, M. A. Suligoy, L. R. Domingo, J. Mol. Struct. 2015, 1079,
47; for other examples, see: b) D. N. Reinhoudt, C. G. Kouwen-
hoven, Tetrahedron Lett. 1974, 15, 2503; c) G. W. Gribble, E. T.
Pelkey, F. L. Switzer, Synlett 1998, 10, 1061.
[8] For enantioselective methods, see: a) Y. Liu, M. Nappi, E.
Arceo, S. Vera, P. Melchiorre, J. Am. Chem. Soc. 2011, 133,
15212; for racemic methods, see: b) V. Pirovano, L. Decataldo,
E. Rossi, R. Vicente, Chem. Commun. 2013, 49, 3594;
c) Y. A. M. Mohamed, F. Inagaki, R. Takahashi, C. Mukai,
Tetrahedron 2011, 67, 5133; d) H. Fuwa, M. Sasaki, Chem.
Commun. 2007, 2876; e) J. Barluenga, F. J. FaÇanµs, R. Sanz, Y.
Fernµndez, Chem. Eur. J. 2002, 8, 2034.
[9] a) C. Exon, T. Gallagher, P. Magnus, J. Chem. Soc. Chem.
Commun. 1982, 613; b) S. F. Vice, H. Nandin de Carvalho, N. G.
Taylor, G. I. Dmitrienko, Tetrahedron Lett. 1989, 30, 7289; c) N.
Kuroda, R. Takahashi, K. Yoshinaga, C. Mukai, Org. Lett. 2006,
8, 1843; d) F. Inagaki, M. Mizutani, N. Kuroda, C. Mukai, J. Org.
Chem. 2009, 74, 6402; e) D. Gagnon, C. Spino, J. Org. Chem.
2009, 74, 6035; f) E. lvarez, P. García-García, M. A. Fernµn-
dez-Rodríguez, R. Sanz, J. Org. Chem. 2013, 78, 9758.
[10] H. L. Sings, G. H. Harris, A. W. Dombrowski, J. Nat. Prod. 2001,
64, 836.
[17] R. Güller, M. Dobler, H.-J. Borschberg, Helv. Chim. Acta 1991,
74, 1636.
[18] a) J.-D. Chai, M. Head-Gordon, Phys. Chem. Chem. Phys. 2008,
10, 6615; b) F. Jensen, J. Chem. Theory Comput. 2014, 10, 1074.
[19] Z.-J. Jia, H. Jiang, J.-L. Li, B. Gschwend, Q.-Z. Li, X. Yin, J.
Grouleff, Y.-C. Chen, K. A. Jørgensen, J. Am. Chem. Soc. 2011,
133, 5053.
[20] Computational work by Houk et al. indicates that for extended
systems a stepwise mechanism might be operating: A. Dieck-
mann, M. Breugst, K. N. Houk, J. Am. Chem. Soc. 2013, 135,
3237.
Received: October 16, 2015
[11] M. D. Burke, S. L. Schreiber, Angew. Chem. Int. Ed. 2004, 43, 46;
Angew. Chem. 2004, 116, 48.
Revised: November 16, 2015
Published online: December 4, 2015
1024
www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1020 –1024