Enantioselective polyene cyclization via organo-SOMO catalysis

Article abstract of DOI:10.1021/ja100185p

The first organocatalytic enantioselective radical polycyclization has been accomplished using singly occupied molecular orbital (SOMO) catalysis. The presented strategy relies on a selective single-electron oxidation of chiral enamines formed by condensation of polyenals with an imidazolidinone catalyst employing a suitable copper(II) oxidant. The reaction proceeds under mildly acidic conditions at room temperature and shows compatibility with an array of electron-poor as well as electron-rich functional groups. Upon termination by radical arylation followed by subsequent oxidation and rearomatization, a range of polycyclic aldehydes were accessed (12 examples, 54?77% yield, 85?93% ee). The enantioselective formation of up to six new carbocycles in a single catalyst-controlled cascade is described. Evidence for a radical-based cascade mechanism is indicated by a series of experimental results.

Full text of DOI:10.1021/ja100185p

Published on Web 03/24/2010
Enantioselective Polyene Cyclization via Organo-SOMO Catalysis
Sebastian Rendler and David W. C. MacMillan*
Merck Center for Catalysis at Princeton UniVersity, Princeton, New Jersey 08544
Received January 19, 2010; E-mail: dmacmill@princeton.edu
Scheme 1. Enantioselective Polycyclization via SOMO Catalysis
Inspired by the biosynthetic origins of steroids and terpenoidal
natural products, synthetic organic chemistry has witnessed con-
siderable efforts in the field of biomimetic polyene cyclization.1
Most importantly, cationic polycyclizations, originally pioneered
by van Tamelen, Johnson, and Goldsmith, have emerged as a
valuable tool for the generation of diverse polycylic skeletons.2,3
While substrate- or auxiliary-mediated induction has traditionally
been employed to render such cascade reactions stereoselective,
recent advances by Yamamoto, Ishihara, and Gagne´ have shown
that catalytic enantiocontrol can also be exploited in these cation-
driven bond constructions.4 In contrast, radical-mediated polyene
cyclizations, as originally proposed (and dismissed) by Breslow as
a biosynthetic pathway,5 have received considerably less attention.1,6
Moreover, while auxiliary-based cascade radical cyclizations have
been reported,6 enantioselective approaches via catalysis remain
largely elusive.7,8 With the objective of developing a general method
for asymmetric catalytic polyene cyclizations, we reasoned that the
application of our previously reported SOMO activation strategy9
should allow us to establish a powerful cascade reaction10 via a
single electron transfer-induced open-shell pathway.5,6e,f,11 In this
communication, we report the successful execution of these ideas
and a new application of SOMO catalysis to the enantioselective
construction of multiple C-C bonds and contiguous stereocenters
in the context of steroidal and terpenoidal architecture.
We hypothesized that a suitably functionalized aldehyde (e.g., 1)
bearing the requisite degree of tethered unsaturation should condense
with imidazolidinone catalyst 2 to give R-imino radical intermediate
3 upon oxidation with an appropriate metal oxidant (Scheme 1). At
this stage, we expected radical cation 3 to engage in a series of 6-endo-
trig radical cyclizations terminated by a suitable arene to give
cyclohexadienyl radical 4. A second oxidation step would then furnish
the corresponding cyclohexadienyl cation, which upon rearomatization
and liberation of the catalyst would deliver the requisite pentacycle
5.12 By analogy to our previous SOMO catalysis studies,9g we
presumed that catalyst 2 would favor the R-imino radical geometry
shown for 3, in which the polyene chain is oriented away from the
bulky tert-butyl substituent. At the same time, the aryl moiety on
the catalyst framework should effectively shield the Si face, leaving
the Re face exposed to addition across the proximal trisubstituted
alkene. As a key design element, the catalyst-induced enantio- and
diastereocontrol arising in the production of the first carbocycle should
be structurally relayed in subsequent ring formations. Moreover, we
recognized that the electronic properties of the tethered polyene would
play a pivotal role in partitioning the putative single-electron pathway
toward cascade ring construction, in lieu of nonproductive mechanisms
such as (i) radical oxidation or (ii) nonregioselective alkene addition.
To this end, polyolefins such as 1 that incorporate an alternating
sequence of polarity-inverted CdC bonds (acrylonitrile and isobutene
moieties) were chosen as suitable substrates that would electronically
match nucleophilic olefins with electron-deficient radical intermediates
(and vice versa). Moreover, the selection of trisubstituted nucleophilic
olefins was expected to favor the desired 6-endo-trig mechanism over
a competing 5-exo-trig mode, as radical additions to fully substituted
carbons are strongly disfavored.13 Similarly, the use of unsaturated
nitriles should enforce 6-endo regiocontrol, while functional conversion
of CtN to a range of common steroidal substituents (H, CHO, Me)
is known to be operationally trivial.6f,14
Our proposed polycyclization strategy was first evaluated using
bicyclization substrate 6, imidazolidinone 2a, and a series of metal
oxidants. Surprisingly, our initial survey revealed that strong oxidants
that had been successful in previous SOMO activation studies {e.g.,
cerium(IV) ammonium nitrate (CAN)9a-d,f and [Fe(phen)3](PF6)39g
}
did not generate any of the desired tricyclic product 7. In both cases,
products derived from premature oxidation of the tertiary alkyl radical
intermediate or capture of the radical by external nucleophiles were
predominant.9d In contrast, the use of Cu(OTf)2 in acetonitrile with
sodium trifluoroacetate (NaTFA) as a base furnished a small amount
of the desired tricycle 7, albeit with low enantioselectivity (Table 1,
entry 1).15,16 Subsequent optimization studies revealed that slow
addition of the oxidant (via syringe pump) greatly improved the yield
of desired tricyclic product 7, presumably as a result of the kinetic
preference for a unimolecular cascade process versus an interrupted
pathway that involves bimolecular radical oxidation. While it is possible
that the carbocation intermediates from an interrupted step may still
9
10.1021/ja100185p  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 5027–5029 5027

C O M M U N I C A T I O N S
Table 1. Reaction Optimization for Enantioselective Bicyclizationa
Table 2. Scope Studies in Enantioselective Bi- and
, ,
Tricyclizationa b c
a Reactions were peformed on a 0.20 mmol scale using 2.5 equiv of
Cu(OTf)2 and 2.0 equiv of NaTFA. b Using 3.0 equiv of TFA. c Isolated
yield. d Determined by chiral HPLC analysis. e With slow addition of
oxidant and base as a solution in MeCN or i-PrCN. f A 3:2 i-PrCN/
DME mixture (0.08 M).
participate in a productive cyclization pathway, it is clear that ꢀ-H
elimination remains the dominant mechanism in this case.17
A
subsequent evaluation of catalyst architecture revealed markedly
improved enantioselection with catalyst 2b, in which the benzyl group
of catalyst 2a has been replaced with the extended shielding of a (1-
naphthalene)methyl moiety (74% ee; entry 4). Moreover, a survey of
reaction media revealed a pronounced effect on the enantiomeric excess
in that a 3:2 mixture of isobutyronitrile with 1,2-dimethoxyethane
(DME) gave rise to synthetically useful results (70% yield, 87% ee;
entry 6).
We next evaluated the generality of our newly developed organo-
catalytic protocol in a series of cascade bicyclizations (Table 2). A
wide array of arenes readily acted as terminating groups; in addition
to electron-neutral phenyl rings (7, 70% yield, 87% ee), electron-rich
and -poor aryl aldehydes also participated, furnishing 8-12 in good
yields and comparable enantioselectivity (65-77% yield, 85-90%
ee).18 While the reaction yields for all of the examined aryl terminators
were similar, we observed that electron-deficient arenes were successful
over a larger range of reaction concentrations, a result that supports
the matched interaction of a nucleophilic trialkyl radical with an
electron-poor terminator in the final SRAr cyclization.19 We presume
that in such cases, electron-withdrawing substituents increase the
substitution rate and stabilize the newly formed cyclohexadienyl radical
intermediate. Notably, when meta-substituted arenes were subjected
to these cascade conditions, cyclization at the aryl 2-position was
predominant (affording products 10 and 12), again emphasizing the
radical nature of this process (2-4:1 ortho/para selectivity, 65-75%
yield, 88-90% ee).9g,19 Indeed, the prevalent formation of ortho-
substituted product 12 was in sharp contrast to the literature precedent
for para selectivity when cationic conditions were employed with
similar polyene substrates.2d It is important to note that all of the
products of this survey were obtained as single diastereomers.20
We next tested the capacity of our SOMO catalysis strategy to
accomplish polyene tricyclization. On the basis of the design
principles outlined above, we assumed that aldehydic substrates
incorporating an alternating sequence of electron-rich and electron-
poor olefinic acceptors (isobutene and acrylonitrile) would be
suitably matched to allow intramolecular radical chain propagation
prior to coupling and termination with a suitable π-rich aryl
ring.9g,19 To our delight, a range of tricyclization substrates readily
furnished the desired products 13-15 in moderate to good yields
and useful enantioselectivities (54-71% yield, 86-92% ee; Table
a Conditions: Slow addition (7 h) of Cu(OTf)2 (2.5 equiv), NaTFA
(2.0 equiv), and TFA (3.0 equiv) in i-PrCN (2 parts) to aldehyde and
catalyst (30 mol %) in 1:2 i-PrCN/DME to give a 0.08 M solution with
subsequent stirring for 17 h at room temperature. b Isolated yield. c ee
was determined by chiral HPLC analysis. d Ortho/para mixture (4:1 for
10; 2:1 for 12). e ee of the ortho product; 91% ee for the para
regioisomer of 12.
2).21 It should be noted that electron-rich benzene (13) and indole
(14) terminators were tolerated using these mild oxidative condi-
tions. Moreover, the successful generation of tricyclic aldehyde 15
revealed that π-nucleophilic olefins such as 1,1-disubstituted
styrenes can also serve as suitable terminating groups.
As outlined in Scheme 2, we finally sought to probe the
limitations of this radical cyclization strategy via application to
extended ring systems such as tetra-, penta-, and hexacycles.
Remarkably, for the SOMO-catalyzed tetracyclization of trienal 1,
we observed levels of reaction efficiency and enantiocontrol similar
to those obtained in bicycle and tricycle formation (pentacycle 5:
56% yield, 92% ee, single diastereomer). Moreover, extension of
this multi-bond-forming reaction to tetraene cascade ring construc-
tion cleanly afforded pentacyclization product 16 in almost identical
yield and enantiomeric excess (63% yield, 93% ee). Perhaps most
notably, this new SOMO-polyene cyclization concept was success-
fully extended to the enantioselective22 production of the hexacy-
clization adduct 17 as a single diastereomer in 62% yield
(corresponding to an average yield of 92% per bond formed). In
the course of this cascade bond construction, a total of 11 contiguous
9
5028 J. AM. CHEM. SOC. VOL. 132, NO. 14, 2010

C O M M U N I C A T I O N S
Scheme 2. Extended Ring Systems by Organo-SOMO Catalysisa
1993, 115, 497. (e) Corey, E. J.; Lin, S. J. Am. Chem. Soc. 1996, 118, 8765.
(f) Bogensta¨tter, M.; Limberg, A.; Overman, L. E.; Tomasi, A. L. J. Am. Chem.
Soc. 1999, 121, 12206. For a reagent-controlled enantioselective example,
see: (g) Zhao, Y.-J.; Loh, T.-P. J. Am. Chem. Soc. 2008, 130, 10024.
(4) (a) Ishihara, K.; Nakamura, S.; Yamamoto, H. J. Am. Chem. Soc. 1999, 121,
4906. (b) Ishihara, K.; Ishibashi, H.; Yamamoto, H. J. Am. Chem. Soc. 2001,
123, 1505. (c) Ishibashi, H.; Ishihara, K.; Yamamoto, H. J. Am. Chem. Soc.
2004, 126, 11122. (d) Sakakura, A.; Ukai, A.; Ishihara, K. Nature 2007,
445, 900. (e) Mullen, C. A.; Campbell, M. A.; Gagne´, M. R. Angew. Chem.,
Int. Ed. 2008, 47, 6011.
(5) (a) Breslow, R.; Barrett, E.; Mohaosi, E. Tetrahedron Lett. 1962, 3, 1207.
(b) Breslow, R.; Olin, S. S.; Groves, J. T. Tetrahedron Lett. 1968, 9, 1837.
For a related example, see: (c) Lallemand, J. Y.; Julia, M.; Mansuy, D.
Tetrahedron Lett. 1973, 14, 4461.
(6) For a review, see: (a) Dhimane, A.-L.; Fensterbank, L.; Malacria, M. In
Radicals in Organic Synthesis; Renaud, P., Sibi, M. P., Eds.; Wiley-VCH:
Weinheim, Germany, 2001; p 350. For selected nonenantioselective examples,
see: (b) Chen, L.; Gill, G. B.; Pattenden, G. Tetrahedron Lett. 1994, 35, 2593.
(c) Handa, S.; Nair, P. S.; Pattenden, G. HelV. Chim. Acta 2000, 83, 2629.
(d) Dombroski, M. A.; Kates, S. A.; Snider, B. B. J. Am. Chem. Soc. 1990,
112, 2759. (e) Zoretic, P. A.; Weng, X.; Caspar, M. L.; Davis, D. G.
Tetrahedron Lett. 1991, 32, 4819. (f) Zoretic, P. A.; Fang, H.; Ribeiro, A. A.
J. Org. Chem. 1998, 63, 7213. For selected substrate- or auxiliary-controlled
examples, see: (g) Heinemann, C.; Demuth, M. J. Am. Chem. Soc. 1997,
119, 1129. (h) Heinemann, C.; Demuth, M. J. Am. Chem. Soc. 1999, 121,
4894. (i) Barrero, A. F.; Herrador, M. M.; Qu´ılez del Moral, J. F.; Arteaga,
P.; Arteaga, J. F.; Piedra, M.; Sa´nchez, E. A. Org. Lett. 2005, 7, 2301.
(7) For examples of enantioselective atom-transfer radical bicyclizations employing
chiral Lewis acids, see: (a) Yang, D.; Gu, S.; Yan, Y.-L.; Zhao, H.-W.; Zhu,
N.-Y. Angew. Chem., Int. Ed. 2002, 41, 3014. (b) Yang, D.; Zheng, B.-F.;
Gao, Q.; Shen, G.; Zhu, N.-Y. Angew. Chem., Int. Ed. 2006, 45, 255.
(8) For reviews of catalytic enantioselective radical (cascade) reactions, see: (a)
Godineau, E.; Landais, Y. Chem.sEur. J. 2009, 15, 3044. (b) Miyabe, H.;
Takemoto, Y. Chem.sEur. J. 2007, 13, 7280. (c) Sibi, M. P.; Manyem, S.;
Zimmermann, J. Chem. ReV. 2003, 103, 3263.
a Conditions: See Table 2, footnotes a-c. b Determined by chiral
supercritical fluid chromatography (SFC). c Determination of enantiomeric
excess in this case was not possible because of the sparing solubility of the
polycycle in HPLC or SFC solvents: [R]D ) -25.3 (c ) 0.68, CHCl3);
also see ref 22.
(9) (a) Beeson, T. D.; Mastracchio, A.; Hong, J.-B.; Ashton, K.; MacMillan,
D. W. C. Science 2007, 316, 582. (b) Yang, H.-Y.; Hong, J.-B.; MacMillan,
D. W. C. J. Am. Chem. Soc. 2007, 129, 7004. (c) Kim, H.; MacMillan,
D. W. C. J. Am. Chem. Soc. 2008, 130, 398. (d) Graham, T. H.; Jones,
C. M.; Jui, N. T.; MacMillan, D. W. C. J. Am. Chem. Soc. 2008, 130, 16494.
(e) Amatore, M.; Beeson, T. D.; Brown, S. P.; MacMillan, D. W. C. Angew.
Chem., Int. Ed. 2009, 48, 5121. (f) Wilson, J. E.; Casarez, A. D.; MacMillan,
D. W. C. J. Am. Chem. Soc. 2009, 131, 11332. (g) Conrad, J. C.; Kong, J.;
Laforteza, B. N.; MacMillan, D. W. C. J. Am. Chem. Soc. 2009, 131, 11640.
(10) (a) Walji, A. M.; MacMillan, D. W. C. Synlett 2007, 1477. (b) Enders, D.;
Grondal, C.; Hu¨ttl, M. R. M. Angew. Chem., Int. Ed. 2007, 46, 1570. (c)
Tietze, L. F. Chem. ReV. 1996, 96, 115.
stereocenters, of which five are all-carbon quaternary centers, were
formed from a simple acyclic starting material under the influence
of imidazolidinone 2b. It is instructive to consider that the last
stereogenic center formed in the course of this cascade resides 11.7
Å from the catalyst binding point,23 thus showcasing the stereoin-
duction efficiency of this new SOMO cyclization sequence.
In summary, we have developed the first catalytic enantioselec-
tive cyclization strategy for accessing steroidal and terpenoidal
frameworks using organocatalysis. This strategy represents an
ambient-temperature protocol, which is unprecendented in SOMO
activation catalysis with respect to carbon-carbon bond formation.
Future work will be devoted to the application of this new
technology to the synthesis of complex natural products and
pharmaceutically relevant entities.
(11) Snider, B. B. Chem. ReV. 1996, 96, 339.
(12) This mechanism is in agreement with our previously described intramolecular
R-arylation of aldehydes (see ref 9g).
(13) (a) Curran, D. P.; Giese, B.; Porter, N. A. Stereochemistry of Radical
Reactions; VCH: Weinheim, Germany, 1996; p 71. (b) Spellmeyer, D. C.;
Houk, K. N. J. Org. Chem. 1987, 52, 959. (c) Beckwith, A. J. L.
Tetrahedron 1981, 37, 3073.
(14) Tobisu, M.; Nakamura, R.; Kita, Y.; Chatani, N. J. Am. Chem. Soc. 2009,
131, 3714.
(15) (a) Jenkins, C. L.; Kochi, J. K. J. Am. Chem. Soc. 1972, 94, 843. (b) Jenkins,
C. L.; Kochi, J. K. J. Am. Chem. Soc. 1972, 94, 856.
(16) Additon of both base (NaTFA or NaHCO3) and TFA was found to be necessary
for high conversion. Replacement of Cu(OTf)2/NaTFA by Cu(TFA)2 ·H2O/
NaOTf showed inferior reactivity, presumably because of the kinetic inacces-
sibility of the active oxidant [(i-PrCN)4Cu]X2 (outer-sphere complex; X ) OTf,
TFA) from the inner-sphere complex Cu(TFA)2 (see ref 15a).
(17) Under the reaction conditions, alkene protonation followed by subsequent
cationic cyclization is not a feasible process. Under typical cationic cyclization
conditions (see ref 4d), cationic cyclization could be achieved in a separate
step.
Acknowledgment. Financial support was provided by NIGMS
(R01 01 GM093213-01) and kind gifts from Merck and Amgen.
S.R. thanks the Deutsche Forschungsgemeinschaft (DFG) for a
postdoctoral fellowship (Re 2882/1-1).
Supporting Information Available: Experimental procedures,
syntheses of starting materials, X-ray crystallographic proof of the
absolute configuration of 8, spectral data for all new compounds, and
crystallographic data (CIF). This material is available free of charge
via the Internet at http://pubs.acs.org.
(18) Structural assignments of 8 and 13 as well as the absolute configuration of 8
(after chemical derivatization) were secured by X-ray crystallographic analysis
(see the Supporting Information for details).
(19) Tiecco, M.; Testaferri, L. In ReactiVe Intermediates, Volume 3; Abramo-
vitch, R. A., Ed.; Plenum Press: New York, 1983; p 61.
(20) General procedure for the enantioselective polyene cyclization: An oven-dried
vial equipped with a magnetic stir bar and a rubber septum was charged with
a solution of catalyst 2b ·TFA (0.060 mmol, 0.30 equiv) and the polyenal (0.200
mmol, 1.00 equiv) in 1:2 i-PrCN/DME (1.5 mL) under an Ar atmosphere. To
this mixture, a solution of Cu(OTf)2 (0.500 mmol, 2.50 equiv), NaTFA (0.400
mmol, 2.00 equiv), and TFA (0.400 mmol, 2.00 equiv) in i-PrCN (1.0 mL)
was slowly added over 7 h using a syringe pump at room temperature. After
stirring had been continued for a further 17 h, the light-green solution was
subjected to aqueous workup and purified by flash chromatography to afford
the cyclization product.
References
(1) (a) Yoder, R. A.; Johnston, J. N. Chem. ReV. 2005, 105, 4730.
(2) (a) van Tamelen, E. E. Acc. Chem. Res. 1968, 1, 111. (b) van Tamelen,
E. E.; McCormick, J. P. J. Am. Chem. Soc. 1969, 91, 1847. (c) Johnson,
W. S.; Kinnel, R. B. J. Am. Chem. Soc. 1966, 88, 3861. (d) Johnson, W. S.;
Semmelhack, M. F.; Sultanbawa, M. U. S.; Dolak, L. A. J. Am. Chem. Soc.
1968, 90, 2994. (e) Goldsmith, D. J.; Phillips, C. F. J. Am. Chem. Soc.
1969, 91, 5862.
(3) For an early review, see: (a) Bartlett, P. A. In Asymmetric Synthesis, Volume
3; Morrison, J. D., Ed.; Academic Press: Orlando, FL, 1984; p 341. For
selected examples of substrate- or auxiliary-controlled enantioselective syn-
thesis, see: (b) Bartlett, P. A.; Johnson, W. S. J. Am. Chem. Soc. 1973, 95,
7501. (c) Johnson, W. S.; Brinkmeyer, R. S.; Kapoor, V. M.; Yarnell, T. M.
J. Am. Chem. Soc. 1977, 99, 8341. (d) Johnson, W. S.; Fletcher, V. R.;
Chenera, B.; Bartlett, W. R.; Tham, F. S.; Kullnig, R. K. J. Am. Chem. Soc.
(21) Further evidence against a radical polar crossover reaction mechanism was
obtained by the examination of a mismatched tricyclization substrate consisting
of two trisubstituted electron-rich alkene moieties: significantly lower reaction
efficiency (25% yield of the desired tetracycle) and only partial conversion
were observed.
(22) We assume levels of enantiocontrol similar to those observed for 5 and 16.
(23) Estimated by semiempirical PM3 calculations on 17 using Gaussian 03.
JA100185P
9
J. AM. CHEM. SOC. VOL. 132, NO. 14, 2010 5029