Biomimetic cationic polyannulation reaction catalyzed by Bi(OTf) 3: Cyclization of 1,6-dienes, 1,6,10-trienes, and aryl polyenes

Article abstract of DOI:10.1021/ol201028f

Nonactivated trienes and aryltrienes were cyclized into polycyclic compounds in good to excellent yields under bismuth triflate catalysis in a biomimetic fashion. The reaction showed broad applicability and allowed for the formation of functionalized bicyclic to tetracyclic structures from simple precursors in one pot. For some specific substrates, the cyclization was followed by a methyl shift as encountered in terpenoid biosynthesis.

Full text of DOI:10.1021/ol201028f

ORGANIC
LETTERS
2011
Vol. 13, No. 13
3320–3323
Biomimetic Cationic Polyannulation
Reaction Catalyzed by Bi(OTf)3:
Cyclization of 1,6-Dienes, 1,6,10-Trienes,
and Aryl Polyenes
~
Julien Godeau, Sandra Olivero, Sylvain Antoniotti, and Elisabet Dunach*
ꢀ
^
ꢀ
Laboratoire de Chimie des Molecules Bioactives et des Aromes, Universite de Nice
Sophia Antipolis - CNRS UMR 6001, Institut de Chimie de Nice, Parc Valrose,
06108 Nice Cedex 2, France
dunach@unice.fr
Received April 20, 2011
ABSTRACT
Nonactivated trienes and aryltrienes were cyclized into polycyclic compounds in good to excellent yields under bismuth triflate catalysis in a
biomimetic fashion. The reaction showed broad applicability and allowed for the formation of functionalized bicyclic to tetracyclic structures from
simple precursors in one pot. For some specific substrates, the cyclization was followed by a methyl shift as encountered in terpenoid
biosynthesis.
Polyene cyclization reactions, aimed at mimicking ter-
penoid cyclase-catalyzed biosynthetic pathways to cyclic
terpenoids in plants,1 have been a rewarding matter of
interest primarily devoted to the understanding of cation-π
based cyclization mechanisms, rearrangements, and stereo-
chemistry.2 Such cationic-based cyclizations were further
used as a powerful tool for the construction of polyannu-
lated compounds in steroid syntheses, promoted by
the stoichiometric use of BF3ÀEtO2, SnCl4,3 MeAlCl2/
Me2AlCl,4 or Et2SBrÀSbCl5Br,5 for example. Strong
protic acids such as FSO3H6 or ClSO3H7 could also be
used as stoichiometric promoters in such cyclizations. The
first enantioselective polyene cyclization was reported in
1999, with the design of a reagent guided by the concept of
Lewis acid assisted Bronsted acid (LBA), and applied to
the synthesis of the benchmark odorant (À)-ambrox
through the cyclization of (E,E)-homofarnesol promoted
by 1 equiv of SnCl4 associated with monomethylated (R)-
BINOL.8 Several applications of chiral LBAs in the med-
iation of polyene cyclization followed.9
To our knowledge, the metal-catalyzed cyclization of
nonconjugated trienes or higher order unsaturated sub-
strates has not yet been described.10 To date, metal-catalyzed
reactions involving cationic-type polyene cyclizations have
(1) Wendt, K. U.; Schulz, G. E.; Corey, E. J.; Liu, D. R. Angew.
Chem., Int. Ed. 2000, 39, 2812–2833.
(2) (a) Stork, G.; Burgstahler, A. W. J. Am. Chem. Soc. 1955, 77,
5068–5077. (b) Stadler, P. A.; Eschenmoser, A.; Schinz, H.; Stork, G.
Helv. Chim. Acta 1957, 40, 2191–2198.
(6) Vlad, P. F. Pure Appl. Chem. 1993, 65, 1329–1336.
(7) Barrero, A. F.; Altarejos, J.; Alvarez-Manzaneda, E. J.; Ramos,
J. M.; Salido, S. J. Org. Chem. 1996, 61, 2215–18.
(8) Ishihara, K.; Nakamura, S.; Yamamoto, H. J. Am. Chem. Soc.
1999, 121, 4906–4907.
(9) (a) Ishihara, K.; Ishibashi, H.; Yamamoto, H. J. Am. Chem. Soc.
2002, 124, 3647–3655. (b) Ishihara, K.; Ishibashi, H.; Yamamoto, H.
J. Am. Chem. Soc. 2001, 123, 1505–1506. (c) Nakamura, S.; Ishihara, K.;
Yamamoto, H. J. Am. Chem. Soc. 2000, 122, 8131–8140. (d) Kumazawa,
K.; Ishihara, K.; Yamamoto, H. Org. Lett. 2004, 6, 2551–2554.
(3) (a) Johnson, W. S.; Semmelhack, M. F.; Sultanbawa, M. U. S.;
Dolak, L. A. J. Am. Chem. Soc. 1968, 90, 2994–2996. (b) Johnson, W. S.;
Li, T.-T.; Harbert, C. A.; Bartlett, W. R.; Herrin, T. R.; Staskun, B.;
Rich, D. H. J. Am. Chem. Soc. 1970, 92, 4461–4463. (c) Bartlett, P. A.;
Johnson, W. S. J. Am. Chem. Soc. 1973, 95, 7501–7502.
(4) (a) Corey, E. J.; Wood, H. B. J. Am. Chem. Soc. 1996, 118, 11982–
11983. (b) Surendra, K.; Corey, E. J. J. Am. Chem. Soc. 2009, 131,
13928–13929.
(5) Snyder, S. A.; Treitler, D. S. Angew. Chem., Int. Ed. 2009, 48,
7899–7903.
r
10.1021/ol201028f
Published on Web 06/01/2011
2011 American Chemical Society

only been reported for transformations involving the ring
closure of non-conjugated dienic systems in which the
cyclization was initiated by an activated function (epoxide,
keto ester, or other reactive functional groups).11
Whereas catalytic cycloisomerization processes invol-
ving enynes or diynes with organometallic complexes
(mainly Pd, Au, Ru) have been extensively reported in
recent years,12 the catalytic cycloisomerization of non-
conjugated olefins is rare because of the lower reactivity
of the double bonds, in particular in highly substituted and
bulky olefinic systems.
could be isolated from the reaction of 1a under milder
conditions (see the Supporting Information).
A screening of several metallic triflates and triflimidates,
including SnIV, ScIII, AlIII, FeIII, or CuII derivatives,
revealed that commercially available Bi(OTf)3, a hydrated
salt,20 afforded the best activity and selectivity in terms of
yield and catalytic activity for the tandem cyclization/
rearrangement of 1a affording 1b in up to 78% yield.
Scheme 1. Double Cylization of (E)-1a Catalyzed by Bi(OTf)3
We have been interested in the preparation of Lewis
superacids such as metallic triflates and triflimidates13 and
their use as catalysts for the functionalization of non-
activated olefins,14 R-amidoalkylation of carbon-centered
nucleophiles,15 FriedelÀCrafts-type allylations,16 as well
as 1,6-diene cycloisomerisations.17
During our previous studies on 1,6-diene cycloisomeri-
zations, Sn(NTf2)4, used at 2À5 mol %, was identified as
an efficient catalyst, both in terms of yield and substrate
scope, over a series of metallic triflates and triflimidates.17
In the present contribution, we have extended our studies
to cationic cyclization of non-conjugated trienes such as
diethyl 2-geranyl-2-prenylmalonate 1a, chosen as a model
substrate (Scheme 1). Sn(NTf2)4, used at 5 mol %, resulted
in a moderate 42% yield of the bicyclic product 1b.
Interestingly, a double cyclization occurred, presumably
via intermediates AÀD with the additional 1,2-methyl shift
from the bridgehead carbon of intermediate B toward the
vicinal position as in C. This rearrangement is similar to
whatisobservedinterpenoidchemistry, inparticularin the
1,2-methyl shift of eudesmyl carbocation18 and in the
It is unusual to observe BiIII compounds exhibiting a
π-Lewis character,21 while they are commonly used as
σ-Lewis acids.22 The efficient results obtained with Bi(OTf)3
for the cyclization of 1a, and its apparent superiority over
Sn(NTf2)4, prompted us to evaluate its catalytic activity in
polyene cycloisomerizations. Particular attention was paid
to the structural requirements in the substrate choice, the
influence of a gem-diester moiety, or the possibility of
involving an aryl group in the cyclization. Bi(OTf)3 dis-
played good efficiency for this cyclization and functiona-
lized polycycles were obtained in good yields (Table 1); its
additional interest with respect to Sn(NTf2)4 being the use
of a less toxic Lewis acid.23 In the series of polyunsaturated
compounds examined, a remarkable efficiency in the
synthesis of bicyclic (entries 1À5), tricyclic (entries 6À8),
and tetracyclic stuctures (entry 9) was attained.
overstoichiometric BF3 OEt2-mediated rearrangement in
the labdaneseries.19 Elimination compounds fromA and B
3
(10) During revision of this manuscript, an example of electrophilic
platinum-catalyzed polyene cyclization appeared: Sokol, J. G.; Korapala,
ꢀ
C. S.; White, P. S.; Becker, J. J.; Gagne, M. R. Angew. Chem., Int. Ed. 2011,
DOI: 10.1002/anie.201100463.
(11) (a) Mullen, C. A.; Campbell, A. N.; Gagne, M. R. Angew. Chem.,
Int. Ed. 2008, 47, 6011–6014. (b) Mullen, C. A.; Gagne, M. R. J. Am.
Chem. Soc. 2007, 129, 11880–11881. (c) Zhao, Y.-J.; Tan, L.-J. S.; Li, B.;
Li, S.-M.; Loh, T.-P. Chem. Commun. 2009, 3738–3740. (d) Carpenter,
N. E.; Kucera, D. J.; Overman, L. E. J. Org. Chem. 1989, 54, 5846–8. (e)
Abelman, M. M.; Overman, L. E. J. Am. Chem. Soc. 1988, 110, 2328–9.
ꢀ
(f) Domingo, V.; Dieguez, H. R.; Morales, C. P.; Arteaga, J. F.; Quilez
del Moral, J. F.; Barrero, A. F. Synthesis 2010, 67–72. (g) Zhao, Y.-J.; Li,
B.; Tan, L.-J. S.; Shen, Z.-L.; Loh, T.-P. J. Am. Chem. Soc. 2010, 132,
10242–10244.
(12) (a) Toullec, P. Y.; Blarre, T.; Michelet, V. Org. Lett. 2009, 11,
2888–2891. (b) Sethofer, S. G.; Mayer, T.; Toste, F. D. J. Am. Chem.
Soc. 2010, 132, 8276–8277.
Changing the prenyl side chain of 1a into a crotyl side
chain such as in 2a, still associated with the geranyl moiety,
resulted similarly in a double cyclization/methyl migration
and led to 2b (entry 2). The methyl rearrangement as
observed in the cycloisomerization of 1a also occurred
with analogous structures 3a and 4a. The replacement of a
~
(13) (a) Favier, I.; Dunach, E. Tetrahedron Lett. 2003, 44, 2031–2032.
~
(b) Antoniotti, S.; Dunach, E. Chem. Commun. 2008, 993–995. (c)
~
Antoniotti, S.; Dalla, V.; Dunach, E. Angew. Chem., Int. Ed. 2010, 49,
7860–7888.
~
(14) Antoniotti, S.; Poulain-Martini, S.; Dunach, E. Synlett 2010,
2973–2988.
(15) Ben Othman, R.; Affani, R.; Tranchant, M.-J.; Antoniotti, S.;
~
Dalla, V.; Dunach, E. Angew. Chem., Int. Ed. 2010, 49, 776–780.
(16) Vece, V.; Ricci, J.; Poulain-Martini, S.; Nava, P.; Carissan, Y.;
(20) Gaspard-Iloughmane, H.; Le Roux, C. Eur. J. Org. Chem. 2004,
2517–2532.
Humbel, S.; Dunach, E. Eur. J. Org. Chem. 2010, 6239–6248.
~
(17) Grau, F.; Heumann, A.; Dunach, E. Angew. Chem., Int. Ed.
2006, 45, 7285–7289.
(21) (a) Komeyama, K.; Takahashi, K.; Takaki, K. Org. Lett. 2008,
10, 5119–5122. (b) Kimihiro, K.; Natsuko, S.; Motoyoshi, M.; Ken, T.
Angew. Chem., Int. Ed. 2009, 48, 9875–9878.
(18) Christianson, D. W. Chem. Rev. 2006, 106, 3412–3442.
(19) George, J. H.; McArdle, M.; Baldwin, J. E.; Adlington, R. M.
Tetrahedron 2010, 66, 6321–6330.
(22) Leonard, N. M.; Wieland, L. C.; Mohan, R. S. Tetrahedron
2002, 58, 8373–8397.
(23) Le Roux, C.; Dubac, J. Synlett 2002, 181–200.
Org. Lett., Vol. 13, No. 13, 2011
3321

charge to the ring by a bond for which free rotation is
possible. Free rotation indeed allows for a perfect align-
ment of the unoccupied atomic orbital of the carbocation
and the σ-orbital of the CÀH bond of the migrating
substitutent. If free rotation is not possible, as with inter-
mediate E from substrate 5a, where the charge is formed
within the 7-membered ring, the migration does not occur
and a proton is simply eliminated leading to 5b (Scheme 2).
a
Table 1. Polyene Cyclizations Catalyzed by Bi(OTf)3
Scheme 2. Cyclization of (E)-5a Does Not Feature the Methyl
Shift
Interestingly, in the presence of a remote aryl group, a
tandem diene or triene cyclization followed by a FriedelÀ
Crafts-type cyclization occurred, affording the corre-
sponding arene polycycles in excellent yields of 90À99%
(entries 6À9).
Compounds 6bÀ8b were obtained in good yields (90%)
within short reaction times. The presence of a quaternary
carbon on the tether, often advocated to facilitate the
cyclization process through a ThorpeÀIngold effect,24
was not necessary, since 7a was efficiently converted into
7b in good yield and short reaction time.
Substrate 9a, presenting the structural features to under-
go the methyl shift described above, indeed led to the
rearranged polycycle 9b in excellent yield (99%) and with
total stereoselectivity inthe creation of three CÀC bondsin
one single operation involving a methyl shift and several
proton shifts. The structure of 9b was also confirmed by
X-ray spectroscopy (see the Supporting Information).
Following the StorkÀEschenmoser paradigm, E-olefins
led to trans-fused rings with total selectivities, products 5b,
6b, and 9b were obtained as single isomers, and structures
were confirmed by X-ray crystallography. For substrates
3a, 4a, 7a, and 8a, the remote asymmetric carbon did not
influence the creation of the additional asymmetric center
and moderate dr’s were obtained (6/4 to 7/3).
From a mechanistic viewpoint, the catalytic role of the
metal ion versus the activity of TfOH formed by hydrolysis
from residual water is still under debate.25 The possibility
of Bi(OTf)3 forming hydrates or undergoing hydrolysis to
free acid or hybrid species cannot be excluded. In the
presence of 2,6-di-tert-butylpyridine as proton scavenger,
a General conditions: substrate (1 mmol), Bi(OTf)3 (5 mol %), in
anhydrous refluxing nitromethane (5 mL). b GCÀFID yields determined
with hexadecane as standard, isolated yields after column chromato-
graphy in parentheses. c Structure confirmed by X-ray spectroscopy.
˚
or in the presence of molecular sieves 3 A to trap residual
water, the cycloisomerization was inhibited. On the other
prenyl by a methallyl unit as in 5a led to the trans-fused
bicyclo[5.4.0] compound 5b in 63% yield without rear-
rangement, as confirmed by X-ray analysis (see the Sup-
porting Information). According to our results, the
requirements for such migration are (1) the formation of
intermediates featuring a methyl group at the position 3
relative to the carbocation as in structure B (Scheme 1) and
(2) connection of the carbon atom bearing the positive
(24) Beesley, R. M.; Ingold, C. K.; Thorpe, J. F. J. Chem. Soc. Trans.
1915, 107, 1080–1106.
(25) (a) Kobayashi, S.; Nagayama, S.; Busujima, T. J. Am. Chem.
Soc. 1998, 120, 8287–8288. (b) Wabnitz, T. C.; Yu, J.-Q.; Spencer, J. B.
Chem.;Eur. J. 2004, 10, 484–493. (c) Williams, D. B. G.; Lawton, M.
J. Mol. Catal. A: Chem. 2010, 317, 68–71. (d) Rosenfeld, D. C.; Shekhar,
S.; Takemiya, A.; Utsunomiya, M.; Hartwig, J. F. Org. Lett. 2006, 8,
4179–4182.
3322
Org. Lett., Vol. 13, No. 13, 2011

hand, running the reaction with 1a in non-distilled
CH3NO2 resulted in a slower reaction yielding 76% of 1b
after 1 h. However, and interestingly, when running the
cyclization of 1a with TfOH as the catalyst (5 mol %,
conditions of Table 1), complete conversion was attained
after 3 h and the yield of 1b was of 48% because of
degradation of the starting material in the presence of the
strong Bronsted acid.
We propose that the catalytic cycloisomerization pro-
cess might proceed via a Lewis acid assisted Bronsted acid-
type activation (LBA),9 in which water molecules coordi-
nated to Bi(OTf)3 would present a strongly enhanced but
controlled acidity due to the presence of BiIII. However,
low steady state concentrations of TfOH formed upon
Bi(OTf)3 hydrolysis acting as catalyst could not be dis-
carded. This feature of bismuth triflate has to be related to
its structure. Triflates and triflimidates are indeed counter-
ions with a particular behavior,13c different from what is
observed with other counterions such as halides, for
example. Likewise, structural data on hydrated bismuth
triflate, although scarce, show that triflate anions and
water molecules are located at a similar distance from the
bismuth center, which undoubtedly resulted in increased
acidity of water molecules.26
decreases and the yields and the selectivities drop. The
catalytic system therefore needs the fine-tuning between
Bi(OTf)3 and H2O, provided by bismuth triflate, for high
efficiency. In the particular case of highly substituted
olefins, the tandem cycloisomerizations reported here
could be run in good yields simply with commercially
available Bi(OTf)3 as catalyst.
In summary, we describe herein efficient catalytic poly-
ene cyclization system leading to differently functionalized
polycycles in good to excellent yields, achieved through the
use of a simple, nontoxic and readily available bismuth
triflate catalyst, used in 5 mol %. This constitutes the first
example of nonconjugated and nonfunctionalized polyene
cyclization run with a catalytic amount of a Lewis acid.
Tandem catalytic cyclization-FriedelÀCrafts reactions
were also efficiently performed. In several cases, selective
methyl rearrangements involving 1,2-methyl shifts were
observed in catalytic Lewis acid catalysis. As compared to
other cationic-driven polyene cyclizations, no particular
function but the olefin seems to be necessary to initiate the
reaction. The cyclizations proceed smoothly in 63À99%
yields, involving the formation of up to three cycles and
three new CÀC bonds in a single operation.
Acknowledgment. Supported by the French Agence
Nationale de la Recherche (project ANR07-CP2D-CA-
SAL-04-01), the University of Nice - Sophia Antipolis and
the CNRS.
Theoretical studies of a related cycloisomerization of
bis-homoallylic alcohols catalyzed by Al(OTf)3 have
shown that the coordination of the hydroxyl group of
the substrate to the catalyst resulted in a strong enhance-
ment of the acidity of the former.27 If more water is added,
the efficiency of the catalytic cycloisomerization system
Supporting Information Available. Detailed procedure
and analytical data (1H and 13C NMR, MS and HRMS)
for 1bÀ9b and additional X-ray spectra for 5b, 8b, 9b.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(26) Louer, M.; Le Roux, C.; Dubac, J. Chem. Mater. 1997, 9, 3012–
3016.
(27) Coulombel, L.; Rajzmann, M.; Pons, J.-M.; Olivero, S.;
~
Dunach, E. Chem.;Eur. J. 2006, 12, 6356–6365.
Org. Lett., Vol. 13, No. 13, 2011
3323