Rh(I)-catalyzed formal [5 + 1]/[2 + 2 + 1] cycloaddition of 1-yne-vinylcyclopropanes and two CO units: One-step construction of multifunctional angular tricyclic 5/5/6 compounds

Article abstract of DOI:10.1021/ja110039h

A novel Rh(I)-catalyzed formal [5 + 1]/[2 + 2 + 1] cycloaddition of 1-yne-vinylcyclopropanes and two CO units for the construction of multifunctional angular tricyclic 5/5/6 skeletons with one or two adjacent bridgehead quaternary all-carbon stereocenters in one step has been developed. Preliminary density functional theory calculations have been carried out to investigate the reaction mechanism and the substituent effects.

Full text of DOI:10.1021/ja110039h

COMMUNICATION
pubs.acs.org/JACS
Rh(I)-Catalyzed Formal [5 þ 1]/[2 þ 2 þ 1] Cycloaddition of 1-Yne-
vinylcyclopropanes and Two CO Units: One-Step Construction of
Multifunctional Angular Tricyclic 5/5/6 Compounds
Mu Lin, Feng Li, Lei Jiao, and Zhi-Xiang Yu*
Beijing National Laboratory for Molecular Sciences (BNLMS), Key Laboratory of Bioorganic Chemistry and Molecular Engineering of
the Ministry of Education, College of Chemistry, Peking University, Beijing 100871, China
S Supporting Information
b
Scheme 1. Natural Products Containing Angular Tricyclic
Systems
ABSTRACT: A novel Rh(I)-catalyzed formal [5 þ 1]/[2 þ
2 þ 1] cycloaddition of 1-yne-vinylcyclopropanes and two
CO units for the construction of multifunctional angular
tricyclic 5/5/6 skeletons with one or two adjacent bridge-
head quaternary all-carbon stereocenters in one step has
been developed. Preliminary density functional theory
calculations have been carried out to investigate the reaction
mechanism and the substituent effects.
preeminent goal of organic synthesis is to reach structural
complexity in step- and atom-economical manners, prefer-
A
Scheme 2. Previously Reported Syntheses of Angular
Tricyclic Systems and Our New Work
ably using a single-step reaction with all atoms embedded in the
desired products.1,2 This concept and its practice are influencing
today’s science of synthesis. Tricyclic and polycyclic carbocyclic
ring skeletons are widely found in natural products with sig-
nificant biological properties, such as the 5/5/5 systems of
angular triquinanes and the 5/5/6 systems of the lycopodiaceae
family (Scheme 1).3 One of the challenges in the synthesis of
these natural products and their analogues for biological and
medicinal studies is usually related to finding one or a series of
reactions to construct these ring systems. The widely applied
approaches for constructing the tricyclic carbocyclic skeletons
involve either building each ring independently, single-step
building of two rings first and then the third ring, or building
one ring first and then the other two rings in one step. An ideal
way to build a tricyclic skeleton that would have step- and atom-
economical characters would be to build the three rings in one
step from readily available starting materials.
(1-yne-VCPs) and two CO units for the synthesis of multi-
functional angular tricyclic 5/5/6 compounds (Scheme 2c).
We previously developed a synthetically useful [(3 þ 2) þ 1]
cycloaddition8 that converts 1-yne-VCP derivatives and CO into
bicyclic cyclohexenones in the presence of a catalytic amount of
[Rh(CO)2Cl]2. It was found that all 1-yne-VCPs with terminal
alkynes gave only [(3 þ 2) þ 1] cycloadducts, no matter whether
a higher or lower CO pressure was applied. For 1-yne-VCP 1a,
which has an internal alkyne, the [(3 þ 2) þ 1] cycloadduct was
obtained exclusively when a CO pressure of 1 atm was applied
(Table 1, entry 1). However, to our surprise, when a CO pressure
of 0.2 atm was applied, the reaction of 1a afforded, in addition
to the expected [(3 þ 2) þ 1] cycloadduct 3a in 39% yield,
the angular tricyclic 5/5/6 compound 2a in 32% yield and
the tricyclic 5/5/5 compound 4a in 4% yield (Table 1,
entry 2). Single-crystal X-ray diffraction analysis confirmed the
Modern transition-metal-catalyzed cycloadditions4 represent
powerful tools enabling quick and efficient access to mono-,
bicyclic, and sometimes tricyclic structures5 in one step. How-
ever, few methods have been developed for the one-step synth-
esis of compounds with angular tricyclic skeletons (such as those
embedded in angular triquinanes and the lycopodiaceae family).
Previously, the construction of angular tricyclic compounds
using transition-metal-catalyzed cycloadditions has involved
Pauson-Khand reactions of substrates that already have one
five-membered ring (Scheme 2a,b).6,7 We envisioned that a one-
step construction of an angular tricyclic 5/5/6 skeleton would be
the most efficient strategy to achieve this goal. Here we report an
unexpected and unprecedented novel Rh(I)-catalyzed formal [5
þ 1]/[2 þ 2 þ 1] cycloaddition of 1-yne-vinylcyclopropanes
Received: November 9, 2010
Published: January 20, 2011
r
2011 American Chemical Society
1690
dx.doi.org/10.1021/ja110039h J. Am. Chem. Soc. 2011, 133, 1690–1693
|

Journal of the American Chemical Society
COMMUNICATION
Table 1. Optimization Studies
yield (%)a
entry
catalyst (5 mol %)
solvent
CO pressure (atm)
temp (°C)
t (h)
2a
3a
4a
1
2
3
4
5
6
7
8
9
[Rh(CO)2Cl]2
[Rh(CO)2Cl]2
[Rh(CO)2Cl]2
[Rh(CO)2Cl]2
[Rh(CO)2Cl]2
[Rh(CO)2Cl]2
[Rh(COD)Cl]2
[Rh(dppp)Cl]2
toluene
toluene
dioxane
DCE
1
80
80
80
80
80
80
80
80
80
7
19
13.5
1
-
32
48
61
32
9
85
-
4
0.2
0.2
0.2
1
39
49
-
-
-
-
-
32
31
DCE
1
DCE
4
3
64
DCE
0.2
0.2
0.2
17.5
24
24
45
21
DCE
no reaction
no reaction
b
[Rh(dppp)]SbF6
DCE
a Isolated yield after column chromatography. b Using 10 mol % catalyst.
tricyclo[7.3.0.04,9]dodecane skeleton of 2a. Apparently, this unique
angular tricyclic 5/5/6 compound originated from the cyclo-
addition of 1a and two CO units, and the trace amount of the
tricyclic 5/5/5 byproduct 4a came from the cycloaddition of 1a
and one CO molecule. In view of the fact that this “dicarbonylative
cycloaddition” process9 is very rare in transition-metal-catalyzed
cycloaddition chemistry and more importantly that the angular
5/5/6 tricyclic structure of 2a could be reached in one step from
easily prepared 1-yne-VCP 1a, we were eager to optimize and
develop this reaction to a useful synthetic method in the
laboratory.
Therefore, we employed 1-yne-VCP 1a as the model substrate
and studied the influence of CO pressure, solvent, and catalyst on
the reaction (Table 1). It was found that higher CO pressure
usually facilitated the [(3 þ 2) þ 1] cycloaddition and inhibited
the formation of the angular tricyclic cycloadduct 2a (entries 1, 2,
and 4-6). 1,2-Dichloroethane (DCE) was found to be an
optimal solvent in comparison with toluene and 1,4-dioxane. A
brief screening of different Rh(I) complexes indicated that
[Rh(CO)2Cl]2 was the best catalyst among those examined
(entries 4 and 7-9).
chlorine-substituted alkyl chain, which could facilitate further
modification of the cycloadduct (entry 4). A TMS-substituted
1-yne-VCP could be employed to give a moderate yield; this
could serve as a surrogate of the nonsubstituted 1-yne-VCP, as
the TMS group in the product can be easily removed afterward
(entry 6). We also attempted to forge an angular tricyclic 6/5/6
skeleton using this formal [5 þ 1]/[2 þ 2 þ 1] reaction of 1s and
CO (see ref 10). However, in this case only [(3 þ 2) þ 1]
product instead of the tricyclic 6/5/6 compound was obtained.10
Further investigations of this [5 þ 1]/[2 þ 2 þ 1] reaction
showed that 1-yne-VCP substrates with CR substituents are also
compatible and afford C4-substituted tricyclic products, albeit in
moderate yields and accompanied by remarkable amounts of [(3
þ 2) þ 1] products (Table 3). Significantly, these reactions allow
the construction of angular tricyclic skeletons with two adjacent
bridgehead quaternary all-carbon stereocenters (C4 and C9),
which is a real challenge in cycloaddition reactions. This feature
further enhances the synthetic utility of the present cycloaddi-
tion.
We refer to the present reaction as a formal [5 þ 1]/[2 þ 2 þ
1] cycloaddition since we thought that this reaction could
start from [5 þ 1] reaction11 of VCP and CO, which would be
followed by a Pauson-Khand reaction12 of the in situ-generated
alkene, alkyne, and CO (Scheme 3a). However, a control
experiment showed that an independently synthesized yne-β,γ-
cyclohexenone, 5, could not afford the cycloaddition product 2a
under the optimal reaction conditions: 96% of 5 was recovered
after 1 h, the same amount of time as required for the standard
reaction in entry 4 of Table 1 (Scheme 3b). This result clearly
indicated that this cycloaddition reaction does not proceed via a
cascade [5 þ 1]/[2 þ 2 þ 1] mechanism, even though after 24 h,
a trace amount of 2a was found in the reaction system. To present
this reaction in an easily memorized way, we still name this
reaction as a formal [5 þ 1]/[2 þ 2 þ 1] cycloaddition.
Two possible pathways (see Figure S1 in the Supporting
Information) were proposed to account for this reaction. Prelimin-
ary density functional theory (DFT) calculations suggested that the
Various 1-yne-VCP substrates were submitted to the optimal
reaction conditions (5 mol % [Rh(CO)2Cl]2 as catalyst, DCE as
solvent, 80 °C). It was found that the yields of angular tricyclic 5/
5/6 cycloadducts were generally good to excellent and that the
[(3 þ 2) þ 1] cycloadducts were generated as byproducts in
minor amounts (Table 2). In all cases except when 1k was used as
the substrate, the tricyclic 5/5/5 byproducts (of 4a type) were
not observed. Nitrogen-, oxygen-, and gem-diester-tethered sub-
strates could afford hetero- and carbotricyclic diones. The steric
bulkiness of the alkyne substitution plays an important role in this
i
t
reaction: bulky groups, such as Pr, cyclohexyl (Cy), and Bu, led
to higher yields of the desired tricyclic cycloadducts (entries 2,3,5
and 8-11). Less bulky alkyne substituents such as Me resulted in
diminished reaction yields (Table 1, entry 4, and Table 2, entry 7).
A chloroalkyl substituent on the 1-yne-VCP substrate could
also be tolerated, producing the angular tricyclic product with a
1691
dx.doi.org/10.1021/ja110039h |J. Am. Chem. Soc. 2011, 133, 1690–1693

Journal of the American Chemical Society
COMMUNICATION
Table 2. Rh(I)-Catalyzed Formal [5 þ 1]/[2 þ 2 þ 1]
Table 3. Exploration of Substrate Scope for Formation of
Two Adjacent Bridgehead Quaternary All-Carbon Stereo-
centers
Cycloaddition
yield (%)a
yield (%)a
entry
substrate
X = TsN
CO pressure (atm) t (h)
2
3
entry
substrate
X = TsN
CO pressure (atm) t (h)
2
3
1
2
3
4
5
6
1b R = Et
0.5
1
2.5
4
58
72
71
60
61
39
4
1
2
3
4
1m
1n
1o
1p
R = Me
0.2
1
4
36
31
32
39
61
46
59
57
1c R = iPr
1d R = tBu
1e R = (CH2)3Cl
1f R = Cy
1g R = TMS
X = C(CO2Me)2
1h R = Me
1i R = iPr
1j R = tBu
X = O
9
26
29
14
6
R = iPr
R = tBu
R = Cy
1.5
1.5
1.5
1
3.5
3.5
2
1
0.5
1
1
X = C(CO2Me)2
0.2
5
5
6
1q
R = tBu
X = O
R = iPr
1
1
1.5
1.5
55
57
18
27
7
8
9
1
1
1
1
47
91
74
19
-
-
1r
1
a Isolated yield after column chromatography.
1.5
b
10
11
1k R = iPr
1
1
1.5
2
87
77
-
1l R = Cy
-
a Isolated yield after column chromatography. b A tricyclic 5/5/5 product
was obtained in 11% yield (see the Supporting Information for details).
pathway shown in Figure 1 is reasonable (see the details in the
Supporting Information). In this pathway, the 1-yne-VCP and
Rh(I) catalyst undergo complexation, cyclopropane cleavage, and
alkyne insertion, generating intermediate B.13 Insertion of CO into
the Rh-C(sp2) bond gives rhodacycloheptenone C. If a direct elimina-
tion occurs on this intermediate, the [(3 þ 2) þ 1] cycloadduct
could be generated. Alternatively, if insertion of the alkene CRdCβ
double bond into the Rh-C(dO) bond in intermediate C takes
place, a tricyclic rhodacyclohexane D could be generated. Next,
another CO molecule could enter this process, and the final 5/5/6
product could be achieved through CO insertion and reductive
elimination (via intermediates E and F, respectively).
Figure 1. Catalytic cycle supported by DFT calculations.
A rationalization of why the angular tricyclic 5/5/6 products
are usually accompanied by the [(3 þ 2) þ 1] cycloadducts and
how the steric bulkiness of the alkyne substituent R of the 1-yne-
VCP influences the results of these two competitive pathways are
proposed (see Figure 2 and Figure S2). DFT calculations
indicated that no matter whether R is small or big, the activation
enthalpies (from IN4 to TS4-[(3 þ 2) þ 1]) required for the [(3
þ 2) þ 1] pathway are almost the same (∼23.0 kcal/mol).
However, as R becomes bigger, the activation enthalpies (from
IN4 to TS4) required for the formal [5 þ 1]/[2 þ 2 þ 1]
pathway are reduced from 24.6 to 22.8 to 21.8 kcal/mol, making
the formal [5 þ 1]/[2 þ 2 þ 1] pathway gradually preferred over
the [(3 þ 2) þ 1] pathway. A cursory examination of IN4, TS4, and
TS4-[(3 þ 2) þ 1] shows that when a bulky R group is introduced
in the 1-yne-VCP, the alkene group in IN4 coordinates to the Rh
atom much more strongly. This stronger coordination makes alkene
insertion into the Rh-C(dO) bond easier, and consequently, the
formal [5 þ 1]/[2 þ 2 þ 1] cycloaddition becomes favored.
Scheme 3. Preliminary Mechanistic Study
In summary, a novel Rh(I)-catalyzed formal [5 þ 1]/[2 þ 2 þ 1]
cycloaddition of 1-yne-VCPs and two CO units has been devel-
oped, yielding multifunctional angular tricyclic 5/5/6 diones in
1692
dx.doi.org/10.1021/ja110039h |J. Am. Chem. Soc. 2011, 133, 1690–1693

Journal of the American Chemical Society
COMMUNICATION
(d) Gabriele, B.; Mancuso, R.; Salerno, G.; Plastina, P. J. Org. Chem.
2008, 73, 756.
(10) We also attempted the Rh(I)-catalyzed formal [5 þ 1]/[2 þ
2 þ 1] reaction of a 1-yne-VCP substrate with an elongated tether.
Unfortunately, the reaction gave a [(3 þ 2) þ 1] product rather than a
tricyclic 6/5/6 cycloadduct. For more examples, see the Supporting
Information.
Figure 2. DFT study showing how the R group influences the
selectivity.
generally good to excellent yields and providing a one-step,
efficient, versatile, and diastereoselective approach to carbo- and
heteroangular tricyclic 5/5/6 compounds with one or two bridge-
head quaternary all-carbon stereocenters. Preliminary DFT cal-
culations have been carried out to investigate the reaction
mechanism and the substituent effects. Further studies of this
reaction (scope, mechanism, and application) are underway.
(11) Selected examples of transition-metal-catalyzed [5 þ 1] cy-
cloaddition reactions: (a) Murakami, M.; Itami, K.; Ubukata, M.; Tsuji,
I.; Ito, Y. J. Org. Chem. 1998, 63, 4. (b) Kamitani, A.; Chatani, N.;
Morimoto, T.; Murai, S. J. Org. Chem. 2000, 65, 9230. (c) Taber,
D. F.; Kanai, K.; Jiang, Q.; Bui, G. J. Am. Chem. Soc. 2000, 122, 6807.
(d) Kurahashi, T.; de Meijere, A. Synlett 2005, 2619. (e) Brancour, C.;
Fukuyama, T.; Ohta, Y.; Ryu, I.; Dhimane, A.-L.; Fensterbank, L.;
Malacria, M. Chem. Commun. 2010, 46, 5470.
(12) Selected reviews of the Pauson-Khand reaction: (a) Brummond,
K. M.; Kent, J. L. Tetrahedron 2000, 56, 3263. (b) Gibson, S. E.;
Stevenazi, A. Angew. Chem., Int. Ed. 2003, 42, 1800. (c) Shibata, T.
Adv. Synth. Catal. 2006, 348, 2328.
’ ASSOCIATED CONTENT
S
Supporting Information. Experimental and computa-
b
tional details; crystallographic data (CIF). This material is
available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
(13) (a) Yu, Z.-X.; Wender, P. A.; Houk, K. N. J. Am. Chem. Soc.
2004, 126, 9154. (b) Yu, Z.-X.; Cheong, P. H.-Y.; Liu, P.; Legault, C. Y.;
Wender, P. A.; Houk, K. N. J. Am. Chem. Soc. 2008, 130, 2378.
Corresponding Author
yuzx@pku.edu.cn
’ ACKNOWLEDGMENT
We thank the Natural Science Foundation of China (20825205,
21072013) and the National Basic Research Program of China
(973 Programs-2011CB808600, 2010CB833203) for financial
support.
’ REFERENCES
(1) (a) Wender, P. A.; Miller, B. L. Nature 2009, 460, 197.
(b) Wender, P. A.; Verma, V. A.; Paxton, T. J.; Pillow, T. H. Acc. Chem.
Res. 2008, 41, 40.
(2) (a) Trost, B. M. Science 1991, 254, 1471. (b) Trost, B. M. Angew.
Chem., Int. Ed. Engl. 1995, 34, 259. (c) Trost, B. M. Acc. Chem. Res. 2002,
35, 695.
(3) Selected review: Mehta, G. Chem. Rev. 1997, 97, 671.
(4) Selected reviews of transition-metal-catalyzed cycloaddition
reactions: (a) Lautens, M.; Klute, W.; Tam, W. Chem. Rev. 1996, 96,
49. (b) Reissig, H.-U.; Zimmer, R. Chem. Rev. 2003, 103, 1151.
(c) Rubin, M.; Rubina, M.; Gevorgyan, V. Chem. Rev. 2007, 107,
3117. (d) Yu, Z.-X.; Wang, Y.; Wang, Y. Chem.;Asian J. 2010, 5, 1072.
(5) Selected reports on tricyclic systems: (a) Ojima, I.; Lee, S.-Y. J.
Am. Chem. Soc. 2000, 122, 2385. (b) Bennacer, B.; Fujiwara, M.; Ojima,
I. Org. Lett. 2004, 6, 3589. (c) Bennacer, B.; Fujiwara, B.; Lee, S.-Y.;
Ojima, I. J. Am. Chem. Soc. 2005, 127, 17756. (d) Montero-Campillo,
M. M.; Rodríguez-Otero, J.; Cabaleiro-Lago, E. J. Phys. Chem. A 2008,
112, 2423. (e) Taber, D. F.; Guo, P.; Guo, N. J. Am. Chem. Soc. 2010,
132, 11179.
(6) Schore, N. E.; Knudsen, M. J. J. Org. Chem. 1987, 52, 569.
(7) Magnus, P.; Fielding, M. R.; Wells, C.; Lynch, V. Tetrahedron
Lett. 2002, 43, 947. However, the reaction in this report was confined to
the synthesis of angular tricyclic 5/5/6 N-heterocycles.
(8) Jiao, L.; Lin, M.; Zhuo, L.-G.; Yu, Z.-X. Org. Lett. 2010, 12, 2528.
(9) Selected reports on dicarbonylative cycloadditions: (a) Gabriele,
B.; Mancuso, R.; Salerno, G.; Veltri, L. Chem. Commun. 2005, 271.
(b) Huang, Q.; Hua, R. Chem.;Eur. J. 2007, 13, 8333. (c) Luo,
Q.; Wang, C.; Zhang, W.-X.; Xi, Z. Chem. Commun. 2008, 1593.
1693
dx.doi.org/10.1021/ja110039h |J. Am. Chem. Soc. 2011, 133, 1690–1693