Tandem Rh(I)-catalyzed [(5+2)+1] cycloaddition/aldol reaction for the construction of linear triquinane skeleton: Total syntheses of (±)-hirsutene and (±)-1-desoxyhypnophilin

Article abstract of DOI:10.1021/ja7100449

A tandem reaction involving a Rh(I)-catalyzed two-component [(5+2)+1] cycloaddition and an aldol condensation has been developed to construct the tricyclo[6.3.0.0^2,6]undecane skeleton and its heteroatom-imbedded analogues. Meanwhile, this method has been successfully applied to natural product synthesis for the first time. The present strategy enables a straightforward approach to the natural linear triquinane skeleton, as demonstrated by concise and step economical syntheses of hirsutene and 1-desoxyhypnophilin, whereby the linear triquinane core is diastereoselectively established in one manipulation with correct placement of all stereocenters, including two quarternary centers. This first application of the Rh-(I)-catalyzed [(5+2)+1] cycloaddition in natural product synthesis highlights the efficiency of this methodology for constructing complex fused ring systems.

Full text of DOI:10.1021/ja7100449

Published on Web 03/12/2008
Tandem Rh(I)-Catalyzed [(5+2)+1] Cycloaddition/Aldol
Reaction for the Construction of Linear Triquinane Skeleton:
Total Syntheses of (()-Hirsutene and
(()-1-Desoxyhypnophilin
Lei Jiao, Changxia Yuan, and Zhi-Xiang Yu*
Beijing National Laboratory for Molecular Sciences (BNLMS), Key Laboratory of Bioorganic
Chemistry and Molecular Engineering of Ministry of Education, College of Chemistry, Peking
UniVersity, Beijing 100871, People’s Republic of China
Received November 6, 2007; E-mail: yuzx@pku.edu.cn
Abstract: A tandem reaction involving a Rh(I)-catalyzed two-component [(5+2)+1] cycloaddition and an
aldol condensation has been developed to construct the tricyclo[6.3.0.02,6]undecane skeleton and its
heteroatom-imbedded analogues. Meanwhile, this method has been successfully applied to natural product
synthesis for the first time. The present strategy enables a straightforward approach to the natural linear
triquinane skeleton, as demonstrated by concise and step economical syntheses of hirsutene and 1-desoxy-
hypnophilin, whereby the linear triquinane core is diastereoselectively established in one manipulation with
correct placement of all stereocenters, including two quarternary centers. This first application of the Rh-
(I)-catalyzed [(5+2)+1] cycloaddition in natural product synthesis highlights the efficiency of this methodology
for constructing complex fused ring systems.
Scheme 1. Rh(I)-Catalyzed Two-Component [(5+2)+1]
Cycloaddition Developed in the Yu Group
Introduction
Transition-metal-catalyzed cycloaddition reactions are actively
pursued to tackle synthetic challenges and to achieve step
economical synthesis of complex cyclic compounds.1,2 Among
them, designing transition-metal-catalyzed high-order cycload-
dtions to construct medium-sized rings is especially attractive
since the traditional cyclization strategies to these ring systems
are often difficult or inhibited due to unfavorable entropic factors
and transannular interactions.2b To date, despite numerous
transition-metal catalyzed high-order cycloadditions to access
complex structures have been developed, only a few have been
utilized in natural product synthesis.3 We recently reported a
Rh(I)-catalyzed two-component [(5+2)+1] cycloaddition of
ene-vinylcyclopropanes (ene-VCPs) with CO for efficient
construction of bicyclic cyclooctenones (Scheme 1).4 This
method has potential application in the construction of complex
fused ring systems with eight-membered carbocycle imbedded
as a key skeleton, which are widely found in many natural
products of biological and therapeutical significances (e.g., taxol
and pleuromutilin). In addition to fused eight-membered car-
bocycles, we wish to report that this [(5+2)+1] cycloaddition
methodology could also access naturally existing tricyclo-
[6.3.0.02,6]undecane skeleton, the linear triquinanes.
Linear triquinane natural products are a subset of polyquinanes,
which constitute an important class of sesquiterpenoids (Figure
1).5 Linear triquinane natural products are isolated from plants,
microbes, and marine organisms, and they have been attracting
continuous attention from synthetic chemists due to their
promising biological activity and novel architecture.5 Because
of this, numerous strategies have been developed to construct
the linear triquinane skeleton5a since its discovery in 1966.6
However, most of the previous achievements require relatively
long routes to sequentially establish three five-membered rings
(1) (a) Wender, P. A.; Bi, F. C.; Gamber, G. G.; Gosselin, F.; Hubbard, R. D.;
Scanio, M. J. C.; Sun, R.; Williams, T. J.; Zhang, L. Pure Appl. Chem.
2002, 74, 25. (b) Wender, P. A.; Handy, S. T.; Wright, D. L. Chem. Ind.
(London) 1997, 765. (c) Wender, P. A.; Miller, B. L. Organic Synthesis:
Theory and Applications; Hudlicky, T., Ed.; JAI: Greenwich, 1993; Vol.
2, pp 27-66.
(2) (a) Lautens, M.; Klute, W.; Tam, W. Chem. ReV. 1996, 96, 49. (b) Yet, L.
Chem. ReV. 2000, 100, 2963. (c) Murakami, M. Angew. Chem., Int. Ed.
2003, 42, 718.
(3) Total synthesis utilizing transition-metal catalyzed high-order cyclo-
additions: (a) Wender, P. A.; Ihle, N. C. J. Am. Chem. Soc. 1988, 110,
5904. (b) Wender, P. A.; Fuji, M.; Husfeld, C. O.; Love, J. A. Org. Lett.
1999, 1, 137. (c) Wender, P. A.; Zhang, L. Org. Lett. 2000, 2, 2323. (d)
Ashfeld, B. L.; Martin, S. F. Org. Lett. 2005, 7, 4535. (e) Wender, P. A.;
Croatt, M. P.; Witulski, B. Tetrahedron 2006, 62, 7505. (f) Trost, B. M.;
Hu, Y.; Horne, D. B. J. Am. Chem. Soc. 2007, 129, 11781.
(4) Wang, Y.; Wang, J.; Su, J.; Huang, F.; Jiao, L.; Liang, Y.; Yang, D.; Zhang,
S.; Wender, P. A.; Yu, Z.-X. J. Am. Chem. Soc. 2007, 129, 10060. We
suggest here a new nomenclature for Rh(I)-catalyzed cycloaddition reac-
tions: for three-component cycloaddition of vinylcyclopropane, alkyne, and
CO, this is referred to as [5+2+1] cycloaddition; for two-component
cycloaddition of ene-VCP and CO, this is referred to as [(5+2)+1]
cycloaddition, where (5+2) part comes from one molecule, ene-VCP.
(5) Reviews: (a) Singh, V.; Thomas, B. Tetrahedron 1998, 54, 3647. (b) Mehta,
G.; Srikrishna, A. Chem. ReV. 1997, 97, 671.
(6) Comer, F. W.; Trotter, J. J. Chem. Soc. Phys. Org. 1966, 1, 11.
9
10.1021/ja7100449 CCC: $40.75 © 2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 4421-4430
4421

A R T I C L E S
Jiao et al.
Figure 1. Linear triquinane natural products.
Scheme 2. Tandem Rh(I)-Catalyzed [5+2+1] Cycloaddition/Aldol Reactions
of the triquinane system. In 2000, Wender and co-workers
reported an elegant three-component [5+2+1] cycloaddition of
VCP, alkyne, and CO (Scheme 2a).7 In most cases, the final
products obtained are diquinanes, which are formed via an aldol
condensation of the first-step-generated [5+2+1] cycloadducts.
Inspired by this discovery, we envisioned that an aldol reaction,
in tandem with a Rh(I)-catalyzed two-component [(5+2)+1]
cycloaddition,4 could establish the linear triquinane architecture
in a novel way (Scheme 2b). If this strategy can be realized,
then we can achieve a more concise and efficient approach to
this type of natural products. Therefore, two representative linear
triquinanes, hirsutene (1) and 1-desoxyhypnophilin (2), were
selected as our synthetic targets. Herein, we present our
endeavors toward the development of a tandem Rh(I)-catalyzed
two-component [(5+2)+1] cycloaddition/aldol reaction as a
general method to construct triquinane skeleton and its analo-
gous, together with its application to the total syntheses of
hirsutene and 1-desoxyhypnophilin.
lecular alkene functionality, the most important issue with which
we were concerned is whether the Rh(I)-catalyzed [(5+2)+1]
cycloaddition could proceed on this new alkoxyl-ene-VCP
system. Especially important is that introducing an additional
five-membered ring may disfavor the desired aldol condensation
due to possible ring strain in the transition state, making the
designed tandem reaction failed. Moreover, because two ad-
ditional chiral centers are generated in our designed tandem
process, it is also a critical challenge whether the correct relative
stereochemistry of linear triquinane (cis-anti-cis) could be
established via this approach.
With the above questions in mind, we started our initial
attempt on the designed tandem two-component [(5+2)+1]/
aldol reaction. Methoxyethyloxy-ene-VCP 7, which has the same
alkoxy group as the substrates in Wender’s three-component
[5+2+1]/aldol reaction,7 was synthesized and selected as the
first model substrate (Scheme 3). To our disappointment, when
ene-VCP 7 was exposed to the reaction conditions of Wender’s
three-component [5+2+1]/aldol reaction (1 atm CO, 60 °C,
dioxane as solvent), a complex mixture was obtained and no
desired cycloadduct could be detected (Scheme 4). Then the
previously optimized conditions for the two-component [(5+2)+1]
reaction (0.2 atm CO, 80 °C, dioxane as solvent)4 were used.
To our delight, the desired tricyclic cycloadduct 9 was isolated
in 16% yield as a single diastereomer (Scheme 4), though most
of the starting material became a mixture of unidentified
products. The relative configuration of cycloadduct 9 was
determined by analogy to compound 15d, which has been
confirmed to adopt the natural linear triquinane cis-anti-cis
configuration by NOE experiment (see further for details).
Although the yield is not satisfactory, this preliminary result
confirmed that the designed tandem two-component [(5+2)+1]/
aldol reaction is feasible for triquinane skeleton construction.
Results and Discussion
Initial Attempts in Developing the Tandem Two-Compo-
nent [(5+2)+1]/Aldol Reaction. Our strategy to construct the
linear triquinane architecture relies on the development of a two-
component version of the tandem [5+2+1]/aldol reaction.
However, this design will encounter additional challenges
compared to the original three-component tandem [5+2+1]/
aldol reaction. In Wender’s system, the substrates are vinylcy-
clopropanes, activated alkynes, and CO; whereas in our designed
system, the substrates are ene-VCPs and CO (Scheme 2). Since
the more reactive electron-deficient alkyne component in three-
component [5+2+1]/aldol reaction is replaced by an intramo-
(7) For a tandem three-component [5+2+1]/aldol reaction to produce diquinanes,
see: Wender, P. A.; Gamber, G. G.; Hubbard, R. D.; Zhang, L. J. Am.
Chem. Soc. 2002, 124, 2876.
9
4422 J. AM. CHEM. SOC. VOL. 130, NO. 13, 2008

Triquinane Synthesis
A R T I C L E S
Scheme 3. Synthesis of Alkoxy-Ene-VCP Substrates as Cycloaddition Precursors
Scheme 4. Initial Exploration of the Tandem Two-Component [(5+2)+1]/Aldol Reaction
Given that the O-substituent in the VCP moiety may play an
important role in the tandem reaction, we conducted optimization
on ene-VCP substrates. Ethoxy-ene-VCP 8 was synthesized
(Scheme 3) and tested under the cycloaddition conditions,
however, this substrate afforded none of the desired cycloadduct,
but a complex mixture (Scheme 4). Then we turned our attention
to siloxy-substituted ene-VCP substrate and synthesized tert-
butyldimethylsiloxy-ene-VCP 11 (Scheme 5). It was found that,
upon exposure to the cycloaddition reaction conditions, ene-
VCP 11 underwent the desired tandem [(5+2)+1]/aldol reaction
to furnish tricyclic cycloadduct 9 in 19% yield (Scheme 5).
Although alkoxy-ene-VCP 7 and siloxy-ene-VCP 11 gave
similar results, we preferred siloxy-ene-VCPs for further
investigation due to their easier preparation.
oxalate 12, which then underwent an elimination reaction in
the presence of 4-(N,N-dimethylamino)pyridine (DMAP) in hot
toluene to give tricyclic enone 13, a key intermediate in total
synthesis of hirsutene by Iyoda and co-workers.8 Therefore, the
elaboration of 9 to 13 represents a formal synthesis of hirsutene.
However, the low efficiency of the crucial tandem [(5+2)+1]/
aldol reaction led to a low total yield of this route, and
optimization studies to improve the yield of triquinane cycload-
duct were required. Unfortunately, efforts in this line did not
result in satisfactory improvement on the yield of cycloadduct
9. Therefore, we decided to investigate this tandem reaction in
detail by conducting more model reactions to identify factors
affecting the reaction yield. This investigation is also worthwhile
for studying the relative stereochemistry of the [(5+2)+1]/aldol
cycloaddition products.
Before we further describe our development of the tandem
reaction, we want to point out that cycloadduct 9 can be
converted to hirsutene after several manipulations (Scheme 6).
Acylation of compound 9 with methyl oxalyl chloride afforded
(8) Iyoda, M.; Kushida, T.; Kitami, S.; Oda, M. J. Chem. Soc., Chem. Commun.
1986, 1049.
9
J. AM. CHEM. SOC. VOL. 130, NO. 13, 2008 4423

A R T I C L E S
Jiao et al.
Scheme 5. Synthesis of Siloxy-Ene-VCP Substrate 11 and Its Cycloaddition Reaction
Scheme 6. Conversion of Tricyclic Ketone 9 to Hirsutene
Model Reaction Study on the Tandem Two-Component
[(5+2)+1]/Aldol Reactions. Several new siloxy-ene-VCP
substrates were designed to examine the possible factors
affecting the efficiency of the reaction. First, to test the influence
of the tether group between the alkene functionality and the
VCP moiety, siloxy-ene-VCP substrates 14a and 14b with
tosylamide and geminal diester as tether groups were synthesized
as model substrates.9 To our disappointment, when siloxy-ene-
VCPs 14a and 14b were employed in the tandem reaction, there
was no significant improvement on the reaction yields (Table
1, entries 1 and 2): both tosylamide and geminal-diester-tethered
siloxy-ene-VCPs gave low yields of cycloaddition products, and
most of the starting material was transformed to unidentified
byproducts. This indicates that different tether groups have
minor effect on the reaction, and we should concentrate our
efforts on other aspects to improve the reaction yield.
3-5). More excitingly, NOE experiment on product 15d
suggested that the stereochemistry of the cycloadduct was in
agreement with the cis-anti-cis configuration of natural linear
triquinanes. This assignment also confirms the configuration of
other cycloadducts produced in the tandem [(5+2)+1]/aldol
reaction. Interestingly, siloxy-ene-VCP 14f with an elongated
tether could also undergo the tandem [(5+2)+1]/aldol reaction
to construct a 6,5,5-tricyclic skeleton diastereoselectively (Table
1, entry 6). It is notable that, when (E)-siloxy-ene-VCP 14g
was employed, the cycloaddition-step-generated trans-fused
[(5+2)+1] cycloadduct4 could not undergo the followed aldol
condensation and the final product was bicyclic cyclooctadione
15g (Table 1, entry 7). This suggests that the cycloaddition-
step-generated fused 5-8 bicyclic compound have to adopt a
cis configuration for the success of the tandem [(5+2)+1]/aldol
reaction.
The above results show that the two factors, the geometry of
VCP moiety (cis or trans) and the substitution pattern of VCP
(methyl or H substitution), are very important in determining
the stereochemistry of the [(5+2)+1] cycloaddition step, which
in turn affects the possibility of the following aldol condensation.
Here, we present our preliminary rationalization for the stereo-
chemistry of the tandem [(5+2)+1]/aldol reaction (Scheme 7).
The [(5+2)+1] cycloaddition is proposed to proceed through
VCP cleavage, alkene insertion, CO insertion, and reductive
elimination (Scheme 7a).4,10,11 The 5-8 ring-fusion stereochem-
istry is determined by the irreversible alkene insertion step,
giving either a cis or a trans 5-8 bicyclic product or both. For
Z-ene-VCP substrate A, the cis alkene-insertion transition state
TS1-cis is favored over the trans alkene-insertion transition state
TS1-trans, independent of the R substituent. Therefore, cis-
We turned our attention to the substitution pattern on the
vinylcyclopropane moiety. In the linear triquinane natural
product, there is a methyl group attached to the R-position of
the exo terminal alkene. We envisioned that if a methyl
substituted siloxy-ene-VCP were employed, then it would give
a tricyclic cycloadduct with correct placement of the methyl
group to build a quaternary carbon center, and this would avoid
further manipulation for methyl installation. However, introduc-
ing a methyl group in the VCP moiety of ene-VCP will give
either a cis or a trans-fused bicyclic cycloadduct, depending on
the geometry of the CdC bond in VCP moiety: in previous
[(5+2)+1] methodology development, we found that (E)-methyl
substituted ene-VCPs give trans-[(5+2)+1] cycloadducts, while
their Z counterparts produce cis-[(5+2)+1] cycloadducts.4
Therefore, in order to achieve the desired cis-[(5+2)+1]
cycloaddition, we synthesized (Z)-siloxy-ene-VCPs 14c-e as
model substrates.9 To our excitement, the reactions of ene-VCPs
14c-e produced tricyclic cycloadducts 15c-e in acceptable
yields with only one diastereomer generated (Table 1, entries
(10) Yu, Z.-X.; Wender, P. A.; Houk, K. N. J. Am. Chem. Soc. 2004, 126, 9154.
(11) The configuration of the CdC bond in VCP determines the configuration
of the π-allyl-rhodium species (Scheme 7). For Z-ene-VCP substrate A,
where the C1 and C4 are in cis relationship, π-allyl-rhodium intermediates
and transition states (TS1-cis and TS1-trans) adopt configurations that C1
and C4 are in cis relationship. Alternatively, for E-ene-VCP substrate B,
where the C1 and C4 are in trans relationship, π-allyl-rhodium intermediates
and transition states (TS2-cis and TS2-trans) adopt configurations that C1
and C4 are in trans relationship.
(9) Procedures for the synthesis of ene-VCPs 14a-g are given in Supporting
Information.
9
4424 J. AM. CHEM. SOC. VOL. 130, NO. 13, 2008

Triquinane Synthesis
A R T I C L E S
Table 1. Model Tandem Two-Component [(5+2)+1]/Aldol
trans alken-insertion transition state. When R ) H, the R-C4
repulsion is minor so that TS2-cis is still favored over TS2-
trans, leading to the cis 5-8 bicyclic products. However, when
R ) Me, the R4-C4 steric repulsion is severe, making TS2-
cis disfavored compared with TS2-trans. Consequently, trans
5-8 bicyclic products are formed. The above analysis is
supported by DFT calculations and detailed investigations are
ongoing in our group. The stereochemistry of the aldol
condensation following the [(5+2)+1] cycloaddition is less
complicated (Scheme 7b). For cis cycloadduct, the cis-anti-cis
condensation is favored over the cis-syn-cis condensation
because of less steric repulsion encountered in the former
transition state. However, for trans 5-8 cycloadduct, the
following aldol condensation could not occur, possibly due to
unfavorable ring strain.
Reactionsa
Through the model study, we obtained valuable information
for the crucial reaction of the total synthesis work. First, the
designed tandem two-component [(5+2)+1]/aldol reaction is
feasible when siloxy-ene-VCPs are employed as substrates.
Second, methyl substituted siloxy-ene-VCP substrate has been
proved to be better than nonsubstituted siloxy-ene-VCP sub-
strate. Third, this tandem reaction could produce the desired
diastereomer of the tricyclic cycloadduct, making it applicable
to the synthesis of linear triquinane natural products. Encouraged
by these positive conclusions, we started our work on the total
syntheses of hirsutene and 1-desoxyhypnophilin.
Retrosynthetic Analysis. On the basis of the model study,
we did the retrosynthetic analysis of the two natural products
(Scheme 8). Hirsutene can be obtained from Wittig reaction of
norketone 16 according to literature procedure.12a Norketone
16 is supposed to be derived from the key intermediate 18
through radical deoxygenation reaction. Alternatively, 1-des-
oxyhypnophilin is also envisioned to generate from the key
intermediate 18 via Wittig reaction, allylic oxidation, dienone
formation, and regioselective epoxidation.13 On the basis of the
previous model reaction study, cycloadduct 18 could be directly
obtained from the tandem Rh(I)-catalyzed two-component
[(5+2)+1] cycloaddition/aldol reaction of (Z)-siloxy-ene-VCP
(12) Hirsutene synthesis using the Wittig reaction of norketone 9: (a) Sternbach,
D. D.; Ensinger, C. L. J. Org. Chem. 1990, 55, 2725. Previous report of
hirsutene synthesis: (b) Singh, V.; Vedantham, P.; Sahu, P. K. Tetrahedron
2004, 60, 8161. (c) Banwell, M. G.; Edwards, A. J.; Harfoot, G. J.; Jolliffe,
K. A. Tetrahedron 2004, 60, 535. (d) Wang, J.-C.; Krische, M. J. Angew.
Chem. Int. Ed. 2003, 42, 5855. (e) Lee, H.-Y.; Kim, Y. J. Am. Chem. Soc.
2003, 125, 10156. (f) Singh, V.; Vedantham, P.; Sahu, P. K. Tetrahedron
Lett. 2002, 43, 519. (g) Anger, T.; Graalmann, O.; Schro¨der, H.; Gerke,
R.; Kaiser, U.; Fitjer, L.; Noltemeyer, M. Tetrahedron 1998, 54, 10713.
(h) Oppolzer, W.; Robyr, C. Tetrahedron 1994, 50, 415. (i) Toyota, M.;
Nishikawa, Y.; Motoki, K.; Yoshida, N.; Fukumoto, K. Tetrahedron 1993,
49, 11189. (j) Ramig, K.; Kuzemko, M. A.; McNamara, K.; Cohen, T. D.
J. Org. Chem. 1992, 57, 1968. (k) Hua, D. H.; Venkstaraman, S.; Ostrander,
R. A.; Sinai, G.-Z.; McCann, P. J.; Coulter, M. J.; Xu, M. R. J. Org. Chem.
1988, 53, 507. (l) Disanayaka, B. W.; Weedon, A. C. J. Org. Chem. 1987,
52, 2905. (m) Mehta, G.; Murthy, A. N.; Reddy, D. S.; Reddy, A. V. J.
Am. Chem. Soc. 1986, 108, 3443. (n) Curran, D. P.; Rekiewicz, D. M. J.
Am. Chem. Soc. 1985, 107, 1448. (o) Curran, D. P.; Rakiewicz, D. M.
Tetrahedron 1985, 41, 3943. (p) Hua, D. H.; Sinai-Zingde, G.; Venkat-
araman, S. J. Am. Chem. Soc. 1985, 107, 4088. (q) Magnus, P.; Quagliato,
D. J. Org. Chem. 1985, 50, 1621. (r) Funk, R. L.; Bolton, G. L. J. Org.
Chem. 1984, 49, 5021. (s) Little, R. D.; Higby, R. G.; Moeller, K. D. J.
Org. Chem. 1983, 48, 3139. (t) Wender, P. A.; Howbert, J. J. Tetrahedron
Lett. 1982, 23, 3983. (u) Little, R. D.; Muller, G. W. J. Am. Chem. Soc.
1981, 103, 2744. (v) Mehta, G.; Reddy, A. V. J. Chem. Soc., Chem.
Commun. 1981, 756. (w) Hudlicky, T.; Kutchan, T. M.; Wilson, S. R.;
Mao, D. T. J. Am. Chem. Soc. 1980, 102, 6351. (x) Hudlicky, T.; Koszyk,
F. J.; Kutchan, T. M.; Sheth, J. P. J. Org. Chem. 1980, 45, 5020. (y) Tatsuta,
K.; Akimoto, K.; Kinoshita, M. J. Am. Chem. Soc. 1979, 101, 6116.
(13) 1-Desoxyhypnophilin synthesis: (a) Harrowven, D. C.; Lucas, M. C.;
Howes, P. D. Tetrahedron Lett. 2000, 41, 8985. (b) Harrowven, D. C.;
Lucas, M. C.; Howes, P. D. Tetrahedron 2001, 57, 9157.
a E ) COOEt. b Isolated yield. c A trans-fused cyclooctanedione 15c′ was
also isolated in 26% yield (see Supporting Information for details). d Based
on recovered starting material. e A degredation product, enone 15f′, was
also isolated in 24% yield (see Supporting Information for details). Relative
stereochemistry of 15f was determined by X-ray crystallographic analysis.
f Single diastereomer.
fused 5-8 bicyclic products will be formed. The preference of
the cis alkene-insertion transition state over its trans counterpart
can be well understood by considering the ring strains: the
alkene-insertion transition states prefer to have the forming five-
membered ring (C3-C4-C5-C6-C7) to be cis-fused with the
Rh-C2-C3-C7-C8 ring to minimize the bicyclic ring strain.
For E-ene-VCP substrate B, the alkene-insertion transition states
still prefer to have the forming five-membered ring (C3-C4-
C5-C6-C7) to be cis-fused with the Rh-C2-C3-C7-C8 ring
to minimize the bicyclic ring strain. However, in this case, the
cis alkene-insertion transition state suffers from an additional
R-C4 steric repulsion, whereas this repulsion is absent in the
9
J. AM. CHEM. SOC. VOL. 130, NO. 13, 2008 4425

A R T I C L E S
Jiao et al.
Scheme 7. Rationale for the Stereochemical Outcome of the Tandem [(5+2)+1]/Aldol Reaction
Scheme 8. Retrosynthetic Analysis
19 with CO. In this complexity-increasing reaction, the linear
triquinane skeleton can be established and appropriate functional
groups can be in place for further modifications. The definitive
substrate for the tandem reaction, (Z)-siloxy-ene-VCP 19, could
be obtained from dimethylhexenal 3 through Z-selective Hor-
ner-Wadsworth-Emmons (HWE) olefination, silylenol etheri-
fication, and cyclopropanation reactions. The advantage of the
present design is that most of the complexity increasing
9
4426 J. AM. CHEM. SOC. VOL. 130, NO. 13, 2008

Triquinane Synthesis
A R T I C L E S
Scheme 9. Synthesis of the Key Tricyclic Intermediate 18
Scheme 10. Synthesis of (()-Hirsutene
steps are condensed into a tandem process, while the remaining
steps require only simple and ordinary conversions.
through a radical deoxygenation. Since a ketone functionality
exists in the substrate, the preparation of the deoxygenation
precursor has to avoid the use of strong basic conditions.
Initially, O-phenyl carbonothioate 22 was selected as the
precursor for radical deoxygenation.15 However, the reaction
of tricyclic ketone 18 with phenyl chlorothionoformate in the
presence of DMAP failed to give thioxoester 22 (Scheme 10).
Then methyl oxalate derivative, which was reported to be
suitable for radical deoxygenation of a tertiary alcohol,16 was
taken into consideration. Gratifyingly, the reaction of methyl
oxalyl chloride with intermediate 18 afforded the deoxygenation
precursor 23 in 92% yield. We then conducted the tributylstan-
nane-mediated deoxygenation of oxalate 23 at different tem-
peratures and various stannane loadings (toluene as solvent),
but the yields of norketone 16 always ranged from 30% to 40%.
Fortunately, we finally found that adding the stannane reagent
and AIBN to a refluxing toluene solution of substrate 23 in
one portion (instead of gradually heating a solution of stannane,
AIBN, and the substrate from room temperature to over
100 °C) can improve the isolated yields of norketone 16 to 66-
68% (Scheme 10).
Synthesis of Hirsutene. The first step of the total synthesis
was to prepare (Z)-siloxy-ene-VCP 19 that can be further
elaborated to finish the total syntheses of two target molecules.
The synthesis began with HWE reaction of dimethylhexenal 3
(Scheme 9), which occurred under Still-Jin-modified condi-
tions14 in 87% yield to afford (Z)-R,â-unsaturated ketone 20 as
the major isomer. Treatment of (Z)-20 with TBSOTf and Et3N
in dry ether afforded silylenol ether (Z)-21 quantitatively, which
was then converted to pure (Z)-ene-VCP 19 in 86% yield
through chemoselective cyclopropanation of the more electron
rich olefin in compound 21.
With (Z)-siloxy-ene-VCP 19 in hand, we began to explore
the key reaction of the synthesis. Gratifyingly, exposure of ene-
VCP 19 to [(5+2)+1] cycloaddition conditions4 led to a smooth
reaction, affording the desired tricyclic ketone 18 in 60-62%
yield as a single diastereomer (Scheme 9), together with a small
amount of bicyclic cyclooctadione 15g (10%). It was found that
using low concentration of ene-VCP 19 did not improve the
reaction yield (see the table in Scheme 9).
Finally, norketone 16 was transformed by Wittig reaction to
hirsutene in 63% yield according to the reported procedure12a
In order to finish the hirsutene synthesis, the hydroxyl group
in the key intermediate 18 has to be replaced by hydrogen to
furnish norketone 16, and such transformation can be achieved
(15) Pettus, T. R. R.; Inoue, M.; Chen, X.-T.; Danishefsky, S. J. J. Am. Chem.
Soc. 2000, 122, 6160.
(16) Dolan, S. C.; MacMillan, J. Chem. Commun. 1985, 1588.
(14) Yu, W.; Su, M.; Jin, Z. Tetrahedron Lett. 1999, 40, 6725.
9
J. AM. CHEM. SOC. VOL. 130, NO. 13, 2008 4427

A R T I C L E S
Jiao et al.
Scheme 11. Synthesis of (()-1-Desoxyhypnophilin
1
(Scheme 10). The H and 13C NMR spectra of the synthetic
anhydrous dioxane (8 mL) was degassed by bubbling CO/N2 (1:4 V/V)
for 5 min. The catalyst [Rh(CO)2Cl]2 (6.0 mg, 15 µmol) was added in
one potion and a light yellow solution formed, which was further
bubbled by the above gas for 5 min. The solution was heated to 80 °C
in an oil bath with stirring under a positive pressure of the mixture
gas. After 48 h, TLC indicated the disappearance of the starting material.
The resulting brown reaction mixture was cooled to room temperature
and hydrolyzed by adding 1% HCl in EtOH (0.3 mL) and water (0.1
mL) and stirred at rt for 20 min. Solvent was evaporated and the residue
was purified by flash column chromatography on silica gel (5 g, eluted
with petroleum ether/ethyl acetate 5:1 to 3:1) to afford tricyclic
compound 18 (28.2 mg, 62%) as a pale yellow oil. 1H NMR (300 MHz,
CDCl3): δ 0.90 (s, 3H), 0.96 (s, 3H), 1.07 (s, 3H), 1.22 (dd, J ) 9.6
and 11.7 Hz, 1H), 1.37-1.56 (m, 2H), 1.60-1.74 (m, 3H), 1.85-1.97
(m, 2H), 2.07-2.16 (m, 1H), 2.25-2.37 (m, 1H), 2.47-2.58 (m, 2H),
2.77 (dd, J ) 9.9 and 19.2 Hz, 1H). 13C NMR (75.5 MHz, CDCl3): δ
12.6, 26.5, 29.0, 32.1, 35.2, 40.3, 41.6, 42.1, 44.6, 48.6, 49.2, 60.0,
89.5, 221.8. MS (EI, 70 eV): m/z 222 (M+, 26), 207 (5.0), 189 (7.0),
165 (42), 149 (14), 113 (100), 95 (20). IR (neat): V 2950, 2863, 1728,
1463, 1365, 1261, 1216 cm-1. HRMS Calcd for C14H22O2: 222.1620.
Found: 222.1621.
hirsutene are in agreement with the reported data,12 confirming
that the tandem [(5+2)+1]/aldol reaction indeed produced the
linear triquinane skeleton with correct relative configuration.
The present synthesis of hirsutene is concise and efficient,
requiring 8 steps with 11% overall yield from commercially
available 2,2-dimethylpent-4-enal, while a recently reported
synthesis from the same starting material requires 13 steps.12d
Synthesis of 1-Desoxyhypnophilin. Synthesis of 1-desoxy-
hypnophilin from the common key intermediate 18 requires
more manipulations (Scheme 11). Wittig olefination of inter-
mediate 18 occurred in 85% yield to furnish hydroxyl hirsutene
17. Allylic oxidation of compound 17 by t-BuOOH (TBHP) in
the presence of SeO2 provided a mixture of allylic alcohol 24
(83%) and hydroxyl enone 25 (16%). Since enone 25 is the
desired precursor for dienone formation, oxidation of alcohol
24 was carried out to transform it to hydroxyl enone 25. To
our surprise, when submitted to Swern oxidation conditions,
compound 24 gave only 26% of the normal oxidation product
25 and most of compound 24 was directly converted to dienone
26. This unexpected oxidation-dehydration reaction might be
caused by excess Swern oxidant present in the reaction system.
The remaining hydroxyl enone 25 was dehydrated by refluxing
in benzene with iodine to afford dienone 26 in 83% yield.
Therefore, through oxidation and dehydration, compound 17
can be transformed to dienone 26 in 74% overall yield.
cis-anti-cis-1,4,4-Trimethyltricyclo[6.3.0.02,6]undecan-11-one (16).
To a stirred solution of tricyclic ketone 18 (55.0 mg, 0.247 mmol) and
DMAP (46.5 mg, 0.381 mmol) in 4 mL of dry CH2Cl2 was added
methyl chlorooxalate (63.8 mg, 0.521 mmol) dropwise. After stirred
at room temperature for 1 h, the resulting solution was washed with
water, dried over Na2SO4, and concentrated. The residue was purified
by flash column chromatography on silica gel (5 g, eluted with
petroleum/acetate 6:1) to give oxalate 23 (70.3 mg, 92%) as a colorless
oil. A solution of oxalate 23 (20.9 mg, 0.0678 mmol) in 1.4 mL of
degassed toluene was heated to reflux in an oil bath under argon. A
solution of n-Bu3SnH (84 mg, 0.289 mmol) and AIBN (4.6 mg) in
degassed toluene (1.0 mL) was added in one potion. The reaction
mixture was stirred for another 45 min under reflux. The resulting
mixture was cooled to room temperature and evaporated. The crude
product was purified by flash column chromatography on silica gel (2
g, eluted with petroleum ether/ethyl acetate 20:1) to afford hirsutene
norketone 16 (9.2 mg, 66%) as a colorless oil, which solidifies on
The final step for 1-desoxyhypnophilin synthesis was the
epoxidation of the dienone intermediate. According to a recently
reported procedure,13b we achieved regioselective epoxidation
of dienone 26 in 73% yield (94% based on recovered starting
1
material), giving rise to racemic 1-desoxyhypnophilin. The H
and 13C NMR spectra of the synthetic 1-desoxyhypnophilin are
identical to those reported.13 This successful synthesis further
demonstrated the efficiency of the crucial tandem [(5+2)+1]/
aldol reactions, since the first synthesis of 1-desoxyhypnophilin
completes in 15 steps,13 whereas we finish its second total
synthesis in 9 steps with 13% overall yield.
1
standing. H NMR (300 MHz, CDCl3): δ 0.90 (s, 3H), 0.94 (s, 3H),
0.98-1.02 (m, 1H), 1.04 (s, 3H), 1.18 (dd, J ) 11.6 Hz, 1H), 1.32-
1.48 (m, 2H), 1.56-1.76 (m, 3H), 2.00 (dddd, J ) 6.2, 8.4, 9.8, and
13.1 Hz, 1H), 2.19-2.45 (m, 3H), 2.52 (dquintet, J ) 3.4 and 9.0 Hz,
1H), 2.80 (dt, J ) 10.6 and 8.7 Hz, 1H). 13C NMR (75.5 MHz,
CDCl3): δ 17.2, 22.3, 26.5, 29.2, 29.7, 34.2, 37.5, 41.1, 41.8, 43.3,
46.6, 48.8, 59.3, 224.8. The spectroscopic data was identical to that
reported.12c
(()-Hirsutene (1).12a To a solution of KOBut (37.9 mg, 0.338 mmol)
in tBuOH (0.5 mL) and benzene (2.2 mL) was added at room
temperature under argon methyltriphenyl-phosphonium bromide (134
Experimental Section
Selected experimental procedures for the preparation of 18, 16, 1
(hirsutene), 17, 24, 25, 26, and 2 (1-desoxyhypnophilin) appear below.
Full experimental details for all compounds are given in the Supporting
Information.
cis-anti-cis-8-Hydroxy-1,4,4-trimethyltricyclo[6.3.0.02,6]undecan-
11-one (18). A solution of ene-VCP 19 (62.7 mg, 0.203 mmol) in
9
4428 J. AM. CHEM. SOC. VOL. 130, NO. 13, 2008

Triquinane Synthesis
A R T I C L E S
mg, 0.375 mmol) in one portion and the resulting yellow solution was
stirred at room temperature for 30 min. A solution of hirsutene
norketone 16 (13.4 mg, 0.065 mmol) in dry benzene (1 mL) was added,
and the reaction mixture was brought to reflux for 1 h in a 100 °C oil
bath. The resulting mixture was cooled, poured into water, and extracted
with petroleum ether. The combined extract was dried over MgSO4
and evaporated. Flash column chromatography on neutral alumina
provided crude hirsutene as a colorless oil (containing unidentified
nonpolar impurities). Another column chromatography on AgNO3
impregnated silica gel (eluting with petroleum ether and then petroleum
ether/acetone 50:1 to 25:1) afforded pure (()-hirsutene (8.3 mg, 63%)
as a colorless oil. Spectroscopic data for hirsutene: 1H NMR (300 MHz,
CDCl3): δ 0.92 (s, 3H), 0.95 (s, 3H), 0.99-1.04 (m, 1H), 1.05 (s,
3H), 1.21 (t, J ) 11.7 Hz, 1H), 1.40-1.48 (m, 4H), 1.64 (ddd, J )
2.0, 8.4, and 10.6 Hz, 1H), 1.69-1.78 (m, 1H), 2.11-2.19 (m, 1H),
2.43-2.49 (m, 2H), 2.50-2.65 (m, 2H), 4.77 (s, 1H), 4.82 (s, 1H).
13C NMR (75.5 MHz, CDCl3): δ 23.2, 26.8, 27.2, 29.7, 30.9, 38.6,
40.9, 41.9, 44.3, 49.0, 49.9, 53.4, 56.0, 103.5, 162.9. The spectroscopic
data was identical to that reported.12a
2.65 (m, 1H), 2.61 (d, J ) 18.1 Hz, 1H), 5.22 (s, 1H), 6.03 (s, 1H).
13C NMR (75.5 MHz, CDCl3): δ 17.6, 26.5, 29.1, 39.9, 41.9, 42.8,
44.9, 49.1, 49.3, 52.8, 54.6, 87.0, 116.9, 155.4, 204.5. MS (EI, 70
eV): m/z 234 (M+, 14), 216 (6.0), 201 (6.0), 192 (7.0), 177 (6.0), 149
(5.0), 124 (100). IR (neat): V 2952, 2935, 2866, 1726, 1635, 1465,
1384, 1267, 1116 cm-1 HRMS calcd. for C15H22O2: 234.1620.
.
Found: 234.1620.
11-Methylene-1,4,4-trimethyltricyclo[6.3.0.02,6]undec-8-en-10-
one (26). To a stirred solution of oxalyl chloride (132 mg, 1.04 mmol)
in dry CH2Cl2 (3 mL) at -78 °C was added a solution of DMSO (152
mg, 1.95 mmol) in CH2Cl2 (1 mL) dropwise. The solution was stirred
at -78 °C for 30 min and a solution of diol 24 (21.2 mg, 90 mmol) in
CH2Cl2 (1 mL) was added. After stirred at -78 °C for 30 min, the
reaction mixture was allowed to warm to -30 °C and further stirred
for 1.5 h. The reaction was recooled to -78 °C and neat Et3N (253
mg, 2.50 mmol) was added. The solution was stirred for 2 h and allowed
to warm to room temperature. The reaction mixture was washed with
aqueous HCl (2 M) and saturated NaHCO3 and dried over MgSO4.
Evaporation of the solvent gave the crude product as a brown oil. Flash
column chromatography on silica gel (5 g, eluted with petroleum ether/
ethyl acetate 20:1 to 3:1) afforded dienone 26 (10.1 mg, 52%) as a
pale-yellow oil, and hydroxy enone 25 (5.4 mg, 26%) as a white solid.
Spectroscopic data for 26: 1H NMR (300 MHz, CDCl3): δ 0.95 (s,
3H), 1.13 (s, 3H), 1.16 (s, 3H), 1.22-1.29 (m, 1H), 1.57 (d, J ) 9.0
Hz, 2H), 1.81 (dd, J ) 6.5 and 12.3 Hz, 1H), 2.24-2.34 (m, 1H),
2.36-2.46 (m, 1H), 2.71-2.84 (m, 2H), 5.15 (s, 1H), 5.88 (s, 1H),
5.89 (d, J ) 1.4 Hz, 1H). 13C NMR (50 MHz, CDCl3): δ 23.5, 27.4,
29.0, 32.6, 40.2, 44.1, 44.9, 48.1, 49.6, 51.7, 112.9, 123.1, 154.2, 189.9,
197.8. The spectroscopic data is identical to that reported.13b
cis-anti-cis-8-Hydroxy-11-methylene-1,4,4-trimethyltricyclo-
[6.3.0.02,6]undecane (17). To a solution of KOBut (92.8 mg, 0.828
t
mmol) in BuOH (1 mL) and benzene (4.5 mL) was added at room
temperature under argon methyltriphenyl-phosphonium bromide (247
mg, 0.692 mmol) in one portion and the resulting yellow solution was
stirred at room temperature for 30 min. A solution of hydroxy ketone
18 (28.0 mg, 0.126 mmol) in dry benzene (0.5 mL) was added and the
reaction mixture was brought to reflux for 2 h in a 100 °C oil bath.
The resulting mixture was cooled, poured into water, and extracted
with petroleum ether. The combined extract was dried over MgSO4
and evaporated. Flash column chromatography on silica gel (6 g, eluent
with petroleum ether/ethyl acetate 10:1) provided hydroxy alkene 17
Dehydration of â-Hydroxyl ketone 25 to Form Dienone 26. To a
solution of hydroxy enone 25 (4.3 mg, 0.018 mmol) in benzene (2
mL) was added a small piece of iodine (2.9 mg, 0.012 mmol). The
resulting solution was heated to reflux under stirring for 4 h and allowed
to cool to room temperature. The reaction mixture was washed with
aqueous Na2S2O3 solution to remove iodine. The organic phase was
dried over MgSO4 and evaporated to give the crude product as a brown
oil. Flash column chromatography on silica gel (2 g, eluted with
petroleum ether/ethyl acetate 20:1 to 10:1) afforded dienone 26 (3.3
mg, 83%) as a colorless oil.
1
(23.4 mg, 85%) as a colorless oil. H NMR (300 MHz, CDCl3): δ
0.91 (s, 3H), 0.96 (s, 3H), 1.07 (s, 3H), 1.22 (dd, J ) 8.4 and 12.3 Hz,
1H), 1.25 (s, 1H), 1.39 (ddd, J ) 2.0, 7.8, and 12.0 Hz, 1H), 1.50 (dd,
J ) 5.3 and 8.4 Hz, 1H), 1.54-1.61 (m, 1H), 1.63-1.71 (m, 2H),
1.89 (ddd, J ) 5.2, 8.8, and 12.8 Hz, 1H), 1.99 (dd, J ) 9.2 and 13.7
Hz, 1H), 2.30-2.65 (m, 4H), 4.81 (m, 2H). 13C NMR (75.5 MHz,
CDCl3): δ 18.0, 27.2, 29.0, 29.5, 35.9, 40.0, 41.4, 42.7, 45.0, 49.1,
53.4, 55.5, 92.3, 105.2, 161.1. MS (EI, 70 eV): m/z 220 (M+, 18), 205
(11), 187 (6.0), 177 (12), 151 (12), 124 (10), 111 (100). IR (neat): V
2950, 2864, 1650, 1463, 1365, 1069 cm-1. HRMS Calcd. for
C15H24O: 220.1827. Found: 220.1827.
(()-1-Desoxyhypnophilin (2).13b To a stirred solution of dienone
26 (10.5 mg, 0.0486 mmol) in THF (1.4 mL) and water (1.4 mL) at 0
°C was sequentially added NaHCO3 (69 mg, 0.82 mmol) and H2O2
(30% aqueous solution, 0.14 mL, 1.2 mmol) and the reaction mixture
was stirred at 0 °C for 5 h. The reaction mixture was extracted with
ether and the combined extract was dried over MgSO4. Evaporation of
the solvent gave the crude product as a colorless oil. Flash column
chromatography on silica gel (5 g, eluted with petroleum ether/ethyl
acetate 20:1 to 10:1) afforded (()-1-desoxyhypnophilin (8.2 mg, 73%)
as a colorless oil, together with the recovered dienone 26 (2.4 mg, 23%).
Spectroscopic data for 1-desoxyhypnophilin: 1H NMR (300 MHz,
CDCl3): δ 0.92 (s, 3H), 1.12 (s, 3H), 1.16 (s, 3H), 1.17-1.26 (m,
1H), 1.48-1.56 (m, 2H), 1.80 (ddd, J ) 1.4, 7.6, and 12.3 Hz, 1H),
1.99 (d, J ) 8.7 Hz, 2H), 2.39 (dt, J ) 11.5 and 9.2 Hz, 1H), 2.73 (tq,
cis-anti-cis-8,10-Dihydroxy-11-methylene-1,4,4-trimethyltricyclo-
[6.3.0.02,6] undecane (24). To a stirred solution of hydroxy alkene 17
(38.3 mg, 0.174 mmol) in CH2Cl2 (3 mL) was sequentially added SeO2
t
(13.0 mg, 0.117 mmol) and BuOOH (65% aqueous solution, 70 mg,
0.50 mmol). The resulting solution was stirred at room temperature
for 2 h and TLC indicated the reaction was complete. The reaction
mixture was poured into water, extracted with CH2Cl2, dried over
MgSO4, and concentrated. Flash column chromatography on silica gel
(5 g, eluted with petroleum ether/ethyl acetate 6:1 to 3:1) afforded diol
24 (34.2 mg, 83%) as a colorless oil and hydroxy enone 25 (6.3 mg,
16%) as a white solid. Spectroscopic data for 24: 1H NMR (300 MHz,
CDCl3): δ 0.88 (s, 3H), 1.07 (s, 3H), 1.09 (s, 3H), 1.11-1.16 (m,
1H), 1.26 (s, 2H), 1.36-1.51 (m, 3H), 1.62-1.70 (m, 1H), 1.86 (dd, J
) 2.5 and 14.0 Hz, 1H), 2.02-2.09 (m, 1H), 2.11 (dd, J ) 5.4 and
14.0 Hz, 1H), 2.32-2.44 (m, 2H), 4.55 (br s, 1H), 4.98 (d, J ) 1.0
Hz, 1H), 5.18 (d, J ) 1.0 Hz, 1H). 13C NMR (75.5 MHz, CDCl3): δ
19.3, 27.5, 29.6, 29.7, 39.2, 41.7, 41.9, 43.6, 45.0, 48.5, 54.7, 75.7,
92.2, 108.9, 165.8. MS (EI, 70 eV): m/z 236 (M+, 4.0), 218 (10), 203
(11), 179 (22), 166 (28), 126 (74), 122 (53), 109 (66), 95 (42). IR
(neat): V 3383, 2950, 2935, 2864, 1465, 1284, 1115, 1070 cm-1. HRMS
Calcd. for C15H24O2: 236.1776. Found: 236.1784. Spectroscopic data
for 25: 1H NMR (300 MHz, CDCl3): 0.92 (s, 3H), 1.09 (s, 3H), 1.14
(s, 3H), 1.23-1.29 (m, 3H), 1.45-1.52 (m, 1H), 1.58-1.77 (m, 3H),
1.98 (dd, J ) 8.9 and 14.0 Hz, 1H), 2.46 (d, J ) 18.1 Hz, 1H), 2.52-
J ) 8.6 and 11.1 Hz, 1H), 3.44 (s, 1H), 5.27 (s, 1H), 6.05 (s, 1H). 13
C
NMR (75.5 MHz, CDCl3): δ 17.5, 27.2, 28.9, 30.0, 39.1, 40.0, 42.5,
46.5, 49.5, 49.8, 61.0, 120.0, 153.2, 198.1. The spectroscopic data is
identical to that reported.13b
Conclusion
In summary, a tandem reaction involving a Rh(I)-catalyzed
two-component [(5+2)+1] cycloaddition and an aldol conden-
sation has been realized and successfully applied to the syntheses
of (()-hirsutene and (()-1-desoxyhypnophilin. The present
strategy enables a concise and step economical route to natural
linear triquinanes, whereby their core structure is diastereo-
9
J. AM. CHEM. SOC. VOL. 130, NO. 13, 2008 4429

A R T I C L E S
Jiao et al.
selectively established in a single step with correct placement
of all stereocenters, including two quarternary centers. This first
application of the Rh(I)-catalyzed [(5+2)+1] cycloaddition in
natural product synthesis highlights the efficiency of this
methodology for constructing complex fused ring systems.
Further applications of the two-component [(5+2)+1] cycload-
dition and the present tandem reaction are actively pursued in
our group.
0240203 and 20672005). Professors Paul A. Wender of Stanford
University and K. N. Houk of UCLA are highly appreciated
for their supports to ZXY and the theoretic and synthetic organic
chemistry lab at PKU.
Supporting Information Available: Experimental procedures,
spectroscopic data, and NMR spectra for synthetic intermediates.
This material is available free of charge via the Internet at http://
pubs.acs.org.
Acknowledgment. We are indebted to generous financial
support from Peking University (through a new faculty startup
fund) and the Natural Science Foundation of China (9800445,
JA7100449
9
4430 J. AM. CHEM. SOC. VOL. 130, NO. 13, 2008