Enantioselective synthesis of polycyclic carbocycles via an alkynylation-allylation-cyclization strategy

Article abstract of DOI:10.1021/ol5028333

A new general three-stage strategy to access polycyclic ring systems bearing all-carbon quaternary centers with high enantioselectivity is reported. The required starting materials were readily accessed in racemic form through the α-alkynylation of ketoes

Full text of DOI:10.1021/ol5028333

Letter
pubs.acs.org/OrgLett
Enantioselective Synthesis of Polycyclic Carbocycles via an
Alkynylation−Allylation−Cyclization Strategy
Maria Victoria Vita,† Pascal Mieville,‡ and Jerome Waser*,†
†Laboratory of Catalysis and Organic Synthesis, Ecole Polytechnique Fed
Lausanne, Switzerland
‡Institute of Chemical Sciences and Engineering (ISIC), Ecole Polytechnique Fed
́ ́
erale de Lausanne (EPFL), 1015 Lausanne,
́ ́
erale de Lausanne, EPFL SB ISIC LCSO, BCH 4306, 1015
Switzerland
S
* Supporting Information
ABSTRACT: A new general three-stage strategy to access
polycyclic ring systems bearing all-carbon quaternary centers
with high enantioselectivity is reported. The required starting
materials were readily accessed in racemic form through the α-
alkynylation of ketoesters with EBX (EthynylBenziodoXolone)
hypervalent iodine reagents. A Pd-catalyzed asymmetric
decarboxylation allylation was then achieved in high yields
and enantioselectivities with Trost’s biphosphine ligands.
Finally, transition-metal catalyzed cyclization of the obtained
chiral enynes gave access to fused and spiro polycyclic ring systems constituting the core of many bioactive natural products.
Scheme 1. Our Synthetic Strategy To Access Polycyclic Ring
Systems
he fascinating structure of natural products is the result of
Tmillions of years of evolution in interaction with biological
targets. Particularly striking is the high occurrence of complex
polycyclic ring systems with numerous stereocenters. An
interesting substructure present in natural products is con-
stituted by a benzene ring fused to at least two further saturated
carbocycles. Both fused and spiro ring systems can be found in
important bioactive compounds including steroids such as
estradiol or alkaloids such as buprenorphine.1 The presence of
all-carbon quaternary centers embedded in the polycyclic core of
many of these molecules represents a formidable challenge for
synthetic chemistry.2 In this context, we envisioned that ketone 1
bearing a 1,5-enyne substructure around an all-carbon quaternary
center would be an ideal precursor for both fused and spiro ring
systems (Scheme 1). Addition of an allyl Grignard reagent,
followed by olefin ring-closing metathesis (RCM), would give an
easy access to fused ring systems. The 1,5-enyne itself can be used
to access spiro or rearranged ring systems based on transition-
metal catalysis.3 Whereas RCM is now a mature method,
cycloisomerization reactions to access spiro ring systems have
been much less investigated. Indeed, there are only two examples
of such transformations so far, both occurring on unsubstituted
alkane rings.3h,k
this report only one example leading to all-carbon quaternary
centers in a noncyclic product was presented. In 2013, the same
group reported a kinetic resolution of racemic propargylic esters
using chiral ligands.5b
Clearly, none of the reported methods to access 1,5-enynes
appeared promising in the context of our synthetic goal, and we
decided to develop a new strategy based on the enantioselective
decarboxylative allylation of ketoesters (Tsuji−Trost reaction).6
This very successful approach to access all-carbon quaternary
centers on carbocycles was first introduced by Segusa7 and Tsuji8
and then further developed into enantioselective methods by
Nevertheless, the main challenge with the proposed strategy
resided in the enantioselective synthesis of enyne 1. Most
synthetic methods known to access 1,5-enynes take advantage of
the reactivity of a propargylic cation generated through either
catalytic or stoichiometric Lewis acid activation, allowing access
only to racemic material.4 Morken et al. showed recently that a
Pd-catalyzed enantiospecific cross-coupling reaction gave 1,5-
enynes with high enantiomeric excess starting from enantioen-
riched activated propargylic alcohols and allyl boronic esters.5a In
Received: September 25, 2014
Published: October 15, 2014
© 2014 American Chemical Society
5768
dx.doi.org/10.1021/ol5028333 | Org. Lett. 2014, 16, 5768−5771

Organic Letters
Letter
Stoltz9 and Trost.10,11 Nevertheless, alkynyl substituents in the α-
position of the ketoesters have never been reported. In this class
of substrates, the α-alkynyl group would be in conjugation with
the formed enolate I and will change its electronic properties. It
also leads to a major challenge in regioselectivity, as attack at both
the α- (pathway a) or γ-position (pathway b) would be
possible.12 Moreover, α-alkynyl ketoesters are difficult to access
because both enolates and acetylides are nucleophiles, and
connecting them requires therefore an Umpolung of the
reactivity. To solve the latter synthetic hurdle, our group has
used hypervalent iodine reagents for the α-functionalization of
carbonyl compounds.13 We have in particular developed an
efficient α-ethynylation of ketoesters under mild conditions
using EBX (EthynylBenziodoXolone) reagents.13a,b
Herein, we would like to present the successful implementa-
tion of the outlined three-stage strategy, including: (1) an
improved method for the alkynylation of ketoesters at room
temperature in up to quantitative yield, (2) the first example of
enantioselective decarboxylative allylation at the propargylic
position proceeding in up to quantitative yield and with an
enantiomeric excess higher than 90% on 5-, 6-, and 7-membered
rings, and (3) the use of the obtained enynes to access fused and
spiro tricyclic ring systems via an RCM or cycloisomerization
strategy, as well as the unanticipated formation of a rearranged
[6,6,5,3] ring system.
Table 1. Optimization of the Pd-Catalyzed Decarboxylative
Allylation
a
b
entry
reaction conditions
[Pd2(dba)3], THF
L
ee
1
5a
5a
5b
5b
5c
5d
5e
5d
5e
5e
5e
0
0
0
2
Pd(cinnamyl)Cp (6), THF
6, THF
3
4
6, Et2O
12
59
76
74
70
79
86
93
5
6, Et2O
6
6, Et2O
7
6, Et2O
8
6, MTBE
9
6, MTBE
10
11
5 mol % 6/5 mol % L, MTBE, 0.1 M
2 mol % 6/2.5 mol % L, MTBE, 0.1 M
c
a
All the reaction mixtures were degassed three times and stirred at rt
for 8−12 h using substrate 2a (0.01 mmol), the indicated solvent
(0.033 M), Pd source (20 mol %), and ligand (44 mol %) unless stated
otherwise. All the reactions gave full conversion to the desired product
b
1a exclusively (yield >90% by 1H NMR). The ee was determined by
c
HPLC using Chiralcel columns. Substrate 2b was used.
Preliminary studies showed that the decarboxylation−
allylation sequence proceeded only in very low yield with
ketoesters bearing a free acetylene obtained using our previously
published alkynylation method.13a We consequently first
investigated the possibility to access protected acetylenes in
the α-alkynylation of β-keto esters with hypervalent iodine
reagents. Using diazabicycloundecene (DBU) or tetramethyl-
guanidine (TMG) as a base, the desired alkynes 2 could be
obtained in up to quantitative yield at rt using TIPS-EBX (4a)
and TPDPS-EBX (4b) starting from the corresponding
ketoesters 3 (eq 1).14
reaction conditions led to the best results at rt with a 5 mol %
palladium loading and a concentration of 0.1 M in MTBE (entry
10).16 In the case of ketoester 2b bearing a sterically more
hindered methallyl substituent the enantioselectivity reached a
remarkably high 93% ee even with only 2 mol % palladium (entry
11).17 This result is impressive, as only moderate enantiose-
lectivity has been reported in the past when using indanone
substrates and acyclic allyl groups: 76−84% ee with Trost
ligands10 and 71−80% ee with phenyloxazoline (PHOX)
ligands.9d
Preliminary studies on the decarboxylative allylation were then
conducted using substrate 2a under the reaction conditions
previously reported by Stoltz.9d The substrate was converted
fully to the desired allylation product 1a, but no enantiomeric
excess was obtained using the PHOX ligand 5a (Table 1, entry
1). A switch in the Pd source to Pd(cinnamyl)Cp (6), an
established clean precursor of Pd(0) phosphine complexes,15 did
not improve the result (entry 2). Keeping Pd(cinnamyl)Cp (6)
as the Pd source, we then turned our attention to Trost ligand 5b
which had been applied previously in the decarboxylation of allyl
ketoesters.10b Also in this case full conversion to 1a was observed,
but with no enantioselectivity (entry 3). However, changing the
solvent to Et2O gave for the first time a measurable
enantioselectivity of 12% (entry 4). A screening of other Trost
ligands in Et2O led to better results with an encouraging 59% ee
for 5c, 76% ee for 5d, and 74% ee for 5e (entries 5−7). Changing
the solvent to MTBE gave the desired product in 70% ee with 5d
and 79% ee with 5e (entries 8 and 9). Final optimization of the
Several indanone derivatives were then subjected to the
optimized reaction conditions on a 0.12 mmol scale. 1,5-Enyne
1a was obtained in 89% ee using indanone 2a (Scheme 2).
Compound 2b yielded product 1b in 93% ee and quantitative
yield. A different silyl group (TBDPS) on the acetylene could
also be used to give enyne 1c. Without a silyl protecting group,
the desired product 1d was obtained in 41% yield and 87% ee.18
A
high tolerance toward the electronic properties and the position
of the substituents on the aromatic ring was observed. Neutral
indanone 1e was obtained in quantitative yield and 94% ee.
Bromo and chloro substituents were also well tolerated
(products 1f and 1g), but a lower enantiomeric excess was
observed in the case of trifluoromethylated product 1h.
Modification of the substituent on the allyl group was also
possible: both chloroallyl enyne 1i and enyne 1j bearing a
5769
dx.doi.org/10.1021/ol5028333 | Org. Lett. 2014, 16, 5768−5771

Organic Letters
Letter
a
Scheme 2. Scope of the Pd-Catalyzed Decarboxylative Allylation
a
Isolated yields after column chromatography using the conditions of entry 11 in Table 1 on 0.12 mmol scale are reported. b Using Pd(cinnamyl)Cp
c
(6) (5 mol %) and ligand 5e (6.5 mol %). The (S,S)-enantiomer of ligand 5e was used.
protected alcohol were obtained in high enantiomeric excess and
yield. When changing to tetralones, ee’s higher than 96% were
observed (products 1k−l). A lower enantioselectivity was
obtained for enyne 1m with a simple allyl substituent. Fused
heterocycle 1n and cycloheptanone 1o could also be accessed in
good yields and 90% ee.
Scheme 4. Au- and Pt-Catalyzed Cycloisomerizations
In order to access fused ring systems, an allyl Grignard was
then added to ketone 1b and the corresponding alcohol 7b was
obtained in quantitative yield and with 4.5:1 diastereoselectivity
(Scheme 3). Fused [6,5,6] tricyclic compound 8b was then
Scheme 3. Synthesis of Fused Ring Systems
products, could be used in this case to access the spiro carbocyclic
compounds. Diene 11a and its nonconjugated isomer 12a were
obtained in 88% yield using PPh3AuClO4 as the catalyst (Scheme
4B). In the case of tetralone 10, cyclization was much slower and
could not be achieved using PPh3AuClO4 as the catalyst. Higher
activity was displayed by PPh3AuPF6, which gave the desired
spiro tricyclic ring system in 64% yield.
In conclusion, we have reported a new three-stage strategy to
access polycyclic ring systems efficiently and in high
enantiopurity. The first step of our approach was the synthesis
of α-alkynylated β-ketoesters in high yields using EBX
hypervalent iodine reagents to achieve the Umpolung of
acetylide. The cornerstone of the strategy was the second step:
the first efficient enantioselective synthesis of 1,5-enynes bearing
an all-carbon propargylic quaternary center achieved through a
Pd-catalyzed decarboxylative allylation starting from racemic
ketoesters. The Pd-catalyzed decarboxylation allylation gave high
enantioselectivity by using the DACH-naphthyl Trost ligand,
which had never been used for the allylation of propargylic
positions before. In the third stage, the chiral 1,5-enynes obtained
were finally converted into the desired tricyclic [6,5,6] and
[6,6,6] fused and spiro ring systems using ring-closing metathesis
and cycloisomerization reactions, respectively. Furthermore, the
unanticipated synthesis of a new [6,6,5,3] ring system could also
synthesized in 83% yield from the syn isomer of 7b via RCM
using a Grubbs II ruthenium catalyst. Using the same synthetic
sequence, [6,6,6] tricyclic compound 8k was obtained in 65%
overall yield.19
To study new cycloisomerization reactions, free acetylene 1d
and 10 were synthesized in gram scale (9.9 and 3.4 mmol) in a
one-pot asymmetric allylation−silyl deprotection sequence from
ketoester 2b and 2k in 83% yield/95% ee and 64% yield/97% ee
respectively (Scheme 4). When the synthesis of spiro compound
11a was attempted using PtCl2 as the catalyst with enyne 1d as
reported by Kozmin et al.,3i cyclopropane product 9 was
obtained instead in 52% yield (Scheme 4A).20 The formation of 9
can be explained by the intermediacy of a platinum-carbenoid,
followed by a 1,2-shift and rearomatization. The first two steps of
this sequence have been reported by Toste et al., but using a gold
catalyst.3h
Fortunately and somewhat surprisingly, we found that gold
catalysis, which usually favors formation of the cyclopropane
5770
dx.doi.org/10.1021/ol5028333 | Org. Lett. 2014, 16, 5768−5771

Organic Letters
Letter
Angew. Chem., Int. Ed. 2005, 44, 6924. (c) Streuff, J.; White, D. E.; Virgil,
S. C.; Stoltz, B. M. Nat. Chem. 2010, 2, 192. (d) Behenna, D. C.; Mohr, J.
T.; Sherden, N. H.; Marinescu, S. C.; Harned, A. M.; Tani, K.; Seto, M.;
Ma, S.; Novak, Z.; Krout, M. R.; McFadden, R. M.; Roizen, J. L.; Enquist,
J. A.; White, D. E.; Levine, S. R.; Petrova, K. V.; Iwashita, A.; Virgil, S. C.;
Stoltz, B. M. Chem.Eur. J. 2011, 17, 14199. (e) Keith, J. A.; Behenna,
D. C.; Sherden, N.; Mohr, J. T.; Ma, S.; Marinescu, S. C.; Nielsen, R. J.;
Oxgaard, J.; Stoltz, B. M.; Goddard, W. A. J. Am. Chem. Soc. 2012, 134,
19050.
(10) (a) Trost, B. M.; Xu, J. J. Am. Chem. Soc. 2005, 127, 2846.
(b) Trost, B. M.; Xu, J.; Schmidt, T. J. Am. Chem. Soc. 2009, 131, 18343.
(c) Trost, B. M.; Schaffner, B.; Osipov, M.; Wilton, D. A. A. Angew.
Chem., Int. Ed. 2011, 50, 3548.
(11) Selected other reports: (a) Fournier, J.; Lozano, O.; Menozzi, C.;
Arseniyadis, S.; Cossy, J. Angew. Chem., Int. Ed. 2013, 52, 1257.
(b) Gartshore, C. J.; Lupton, D. W. Angew. Chem., Int. Ed. 2013, 52,
4113. (c) Li, Z.; Zhang, S.; Wu, S.; Shen, X.; Zou, L.; Wang, F.; Li, X.;
Peng, F.; Zhang, H.; Shao, Z. Angew. Chem., Int. Ed. 2013, 52, 4117.
(12) To the best of our knowledge, this challenging regioselectivity
question has never been addressed in the case of α-alkynyl ketoesters.
For a related study on the racemic allylation of α-vinyl ketoesters, see:
Waetzig, S. R.; Rayabarapu, D. K.; Weaver, J. D.; Tunge, J. A. Angew.
Chem., Int. Ed. 2006, 45, 4977.
be achieved with a platinum catalyst. The stage is now ready for
using the strategy in the synthesis of natural and synthetic
bioactive compounds and for extending it to other types of
carbocyclic or heterocyclic scaffolds.
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental procedures and analytical data for all new
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: jerome.waser@epfl.ch.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank F. Hoffmann-La Roche Ltd. for an unrestricted
research grant and the Swiss National Science Foundation
(SNSF, Grant Number 200020_134550) for financial support.
(13) (a) Fernan
2010, 16, 9457. (b) Fernan
Waser, J. Adv. Synth. Catal. 2013, 355, 1631. (c) Vita, M. V.; Waser, J.
Org. Lett. 2013, 15, 3246.
́
dez Gonzal
́
ez, D.; Brand, J. P.; Waser, J. Chem.Eur. J.
́
dez Gonzal
́
ez, D.; Brand, J. P.; Mondiere, R.;
̀
REFERENCES
■
(14) For some substrates, the use of the more stable TPDPS protecting
group was needed to prevent desilylation. Currently, desilylation cannot
be prevented in the absence of the benzene ring. See Supporting
Information (SI) for further details.
(15) MacKinnon, A. L.; Baird, M. C. J. Organomet. Chem. 2003, 683,
114.
(1) Heel, R. C.; Brogden, R. N.; Speight, T. M.; Avery, G. S. Drugs
1979, 17, 81.
(2) (a) Douglas, C. J.; Overman, L. E. Proc. Natl. Acad. Sci. U.S.A. 2004,
101, 5363. (b) Christoffers, J.; Baro, A. Quaternary Stereocenters:
Challenges and Solutions for Organic Synthesis; John Wiley & Sons: 2006;
p 359. (c) Trost, B. M.; Jiang, C. H. Synthesis 2006, 369. (d) Cozzi, P. G.;
Hilgraf, R.; Zimmermann, N. Eur. J. Org. Chem. 2007, 5969.
(3) Reviews: (a) Aubert, C.; Buisine, O.; Malacria, M. Chem. Rev. 2002,
102, 813. (b) Zhang, L.; Sun, J.; Kozmin, S. A. Adv. Synth. Catal. 2006,
(16) See SI for a full list of tested conditions.
(17) The absolute stereochemistry was determined using Trost’s
mandelate ester method: (a) Trost, B. M.; Belletire, J. L.; Godleski, S.;
McDougal, P. G.; Balkovec, J. M.; Baldwin, J. J.; Christy, M. E.;
Ponticello, G. S.; Varga, S. L.; Springer, J. P. J. Org. Chem. 1986, 51, 2370.
(b) Trost, B. M.; Schroeder, G. M. Chem.Eur. J. 2005, 11, 174. The
obtained configuration is in accordance with the prediction model
developed by Trost et al. for this type of ligands (ref 10; see Figure S1 in
SI).
(18) We speculate that interactions between the free alkyne and the Pd
catalyst slow down the reaction in this case. Lower ee’s were observed
with aryl-substituted alkynes. Alkyl-substituted alkynes cannot be
synthesized using EBX reagents.
(19) Overall yield of 8k from ketone 1k based on 81% isolated yield of
the cis isomer of 7k. In the case of the [6.6.6] ring system, ring closing
metathesis of the minor trans diastereomer of 7k was also successful. See
SI.
(20) Interestingly, 9 was optically active. Further studies will be
required to determine the enantiomeric excess and absolute
configuration of this compound.
348, 2271. (c) Jimenez-Nunez, E. s.; Echavarren, A. M. Chem. Rev. 2008,
́
́
̃
108, 3326. (d) Michelet, V.; Toullec, P. Y.; Genet, J. P. Angew. Chem., Int.
Ed. 2008, 47, 4268. (e) Watson, I. D. G.; Toste, F. D. Chem. Sci. 2012, 3,
2899. Selected examples: (f) Mamane, V.; Gress, T.; Krause, H.;
Furstner, A. J. Am. Chem. Soc. 2004, 126, 8654. (g) Harrak, Y.;
̈
̀
Blaszykowski, C.; Bernard, M.; Cariou, K.; Mainetti, E.; Mouries, V.;
Dhimane, A. L.; Fensterbank, L.; Malacria, M. J. Am. Chem. Soc. 2004,
126, 8656. (h) Luzung, M. R.; Markham, J. P.; Toste, F. D. J. Am. Chem.
Soc. 2004, 126, 10858. (i) Sun, J.; Conley, M. P.; Zhang, L.; Kozmin, S.
A. J. Am. Chem. Soc. 2006, 128, 9705. (j) Buzas, A.; Gagosz, F. J. Am.
Chem. Soc. 2006, 128, 12614. (k) Fukamizu, K.; Miyake, Y.;
Nishibayashi, Y. Angew. Chem., Int. Ed. 2009, 48, 2534.
(4) Selected examples: (a) Crowe, W. E.; Sammakia, T.; Schreiber, S. L.
J. Am. Chem. Soc. 1986, 108, 3128. (b) Mukaiyama, T.; Inubushi, A.;
Hayashi, M. Chem. Lett. 1987, 1975. (c) Luzung, M. R.; Toste, F. D. J.
Am. Chem. Soc. 2003, 125, 15760. (d) Georgy, M.; Boucard, V.;
Campagne, J. M. J. Am. Chem. Soc. 2005, 127, 14180. (e) Kabalka, G. W.;
Yao, M. L.; Borella, S. J. Am. Chem. Soc. 2006, 128, 11320. (f) Kuninobu,
Y.; Ishii, E.; Takai, K. Angew. Chem., Int. Ed. 2007, 46, 3296. (g) Lepore,
A. J.; Pinkerton, D. M.; Ashfeld, B. L. Adv. Synth. Catal. 2013, 355, 1500.
(5) (a) Ardolino, M. J.; Morken, J. P. J. Am. Chem. Soc. 2012, 134, 8770.
(b) Ardolino, M. J.; Eno, M. S.; Morken, J. P. Adv. Synth. Catal. 2013,
355, 3413.
(6) Selected reviews: (a) Mohr, J. T.; Stoltz, B. M. Chem.Asian J.
2007, 2, 1476. (b) Weaver, J. D.; Recio, A., III; Grenning, A. J.; Tunge, J.
A. Chem. Rev. 2011, 111, 1846. (c) Hong, A. Y.; Stoltz, B. M. Eur. J. Org.
Chem. 2013, 2745.
(7) Tsuda, T.; Chujo, Y.; Nishi, S.; Tawara, K.; Saegusa, T. J. Am. Chem.
Soc. 1980, 102, 6381.
(8) (a) Shimizu, I.; Yamada, T.; Tsuji, J. Tetrahedron Lett. 1980, 21,
3199. (b) Shimizu, I.; Tsuji, J. J. Am. Chem. Soc. 1982, 104, 5844.
(9) (a) Behenna, D. C.; Stoltz, B. M. J. Am. Chem. Soc. 2004, 126,
15044. (b) Mohr, J. T.; Behenna, D. C.; Harned, A. M.; Stoltz, B. M.
5771
dx.doi.org/10.1021/ol5028333 | Org. Lett. 2014, 16, 5768−5771