Syntheses of seven-membered rings: Ruthenium-catalyzed intramolecular [5+2] cycloadditions

Article abstract of DOI:10.1002/chem.200401065

The Ru-catalyzed intramolecular [5+2] cycloaddition of cyclopropylenynes is investigated with respect to the regio- and diastereoselectivity as well as the functional group compatibility of the reaction. Evidence for the mechanism as occurring through a ruthenacyclopentene intermediate is elucidated from 1) the study of the diastereoselectivity of the cycloaddition; 2) the effect of variation of substituents on the regioselectivity of cyclopropyl bond cleavage in 1,2-trans- and 1,2-cis-disubstituted cyclopropanes and 3) examples that clearly do not involve ruthenacyclohexene as intermediates as products still incorporate the cyclo propyl moiety. The scope and limitations of the Ru-catalyzed cycloaddition are discussed and compared with the Rh-catalyzed reaction. The potential power of this methodology towards natural product total synthesis is demonstrated by the formation of several polycyclic systems with the chosen reaction conditions and readily available cyclopropylenyne substrates.

Full text of DOI:10.1002/chem.200401065

FULL PAPER
Syntheses of Seven-Membered Rings: Ruthenium-Catalyzed Intramolecular
[5+2] Cycloadditions
Barry M. Trost,* Hong C. Shen, Daniel B. Horne, F. Dean Toste,
Bernhard G. Steinmetz, and Christopher Koradin[a]
Abstract: The Ru-catalyzed intramo-
lecular [5+2] cycloaddition of cyclopro-
pylenynes is investigated with respect
to the regio- and diastereoselectivity as
well as the functional group compati-
bility of the reaction. Evidence for the
on the regioselectivity of cyclopropyl
bond cleavage in 1,2-trans- and 1,2-cis-
disubstituted cyclopropanes and 3) ex-
amples that clearly do not involve ruth-
enacyclohexene as intermediates as
products still incorporate the cyclo-
propyl moiety. The scope and limita-
tions of the Ru-catalyzed cycloaddition
are discussed and compared with the
Rh-catalyzed reaction. The potential
power of this methodology towards
natural product total synthesis is dem-
onstrated by the formation of several
polycyclic systems with the chosen re-
action conditions and readily available
cyclopropylenyne substrates.
mechanism as occurring through
a
ruthenacyclopentene intermediate is
elucidated from 1) the study of the dia-
stereoselectivity of the cycloaddition;
2) the effect of variation of substituents
Keywords:
diastereoselectivity
homogeneous catalysis · ruthenium
cycloaddition
·
·
·
enynes
Introduction
bonds. Exceptional control over the direction of cyclopro-
pane bond cleavage was exercised through the selection of
substituents and/or catalyst. Inspired by the mechanistic
pathways of the ruthenium catalyzed Alder–ene reaction
[Eq. (2)][4] and the Rh-catalyzed [5+2] cycloaddition, we en-
visioned that the Ru-catalyzed [5+2] cycloaddition reaction
[Eq. (3)] may proceed and thereby generate various mecha-
nistic questions and synthetic opportunities.[5]
The development of new chemical reactions involving the
serial formation and cleavage of multiple C C bonds leads
ꢀ
to powerful synthetic tools in organic synthesis. As such, cy-
cloaddition reactions catalyzed by transition metals[1] have
been of continuing interest to us due to their efficiency in
constructing complicated structures from much simpler
starting materials in an atom-economical fashion.[2] Wend-
erꢀs group pioneered the development of both inter- and in-
tramolecular Rh-catalyzed [5+2] cycloadditions of cyclo-
propyl enynes [Eq. (1)].[3] This reaction involves the formal
cleavage of two bonds and the formation of three new
[a] Prof. B. M. Trost, H. C. Shen, D. B. Horne, F. D. Toste,
B. G. Steinmetz, C. Koradin
In this article we report the scope of Ru-catalyzed intra-
molecular [5+2] cycloadditions regarding the functional
group tolerance of the reaction, and the length and atom
type of the tether between alkyne and vinylcyclopropane.
To probe the reaction mechanism, we systematically studied
the effects of substituents on the regio- and diastereoselec-
tivity of the Ru-catalyzed reaction. The utility of this meth-
Department of Chemistry
Stanford University
Stanford, CA 94305-5080 (USA)
Fax : (+1)650-725-0002
E-mail: bmtrost@stanford.edu
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
Chem. Eur. J. 2005, 11, 2577 – 2590
DOI: 10.1002/chem.200401065
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2577

odology is reflected by the synthesis of the core structures
of several biologically important natural products (Figure 1)
that contain a polyhydroazulene moiety, such as rameswara-
lide 1,[6] dilatriol 2,[7] and grayanotoxin 3,[8] which represent
we envisioned that it may also catalyze the [5+2] cycloaddi-
tion of vinylcyclopropyl alkynes, primarily due to the possi-
bility of generating an intermediate ruthenacyclopentene
which may then undergo ring expansion to form the seven-
membered ring [Eq. (3)].
We were pleased to observe that this Ru catalyst can
indeed catalyze the desired intramolecular [5+2] cycloaddi-
tion of a broad range of vinylcyclopropyl alkyne substrates
under ambient conditions (room temperature in acetone or
DMF) (Table 1). All reactions were run with 5–10% [CpRu-
(CH3CN)3PF6] (4), 0.1–0.2m in acetone, at room tempera-
ture, unless otherwise noted. Other Ru complexes also ex-
Figure 1. Polycyclic natural products containing seven-membered rings.
Table 1. Scope of substrates in the Ru-catalyzed [5+2] cycloadditions.
demanding synthetic challenges. The opportunity to apply
the Ru-catalyzed [5+2] cycloaddition should offer short,
atom-economical routes to such targets. The model systems
as shown in Equation (4) raise a number of reactivity and
selectivity issues that are addressed herein.[5]
Entry
Substrate
Product
Yield [%]
1
83
5a
5b
5c
6a
6b
6c
2
3
82
82
Results and Discussion
Substrate synthesis: We initiated our investigation of the
Ru-catalyzed [5+2] cycloaddition by preparation of the vi-
nylcyclopropyl alkyne substrates. Various methods were uti-
lized to access the range of substrates described herein and
are illustrated in detail in the Supporting Information. A re-
curring theme in many of the synthetic sequences was the
utilization of palladium catalyzed allylic alkylation as a key
step for substrate formation. The synthesis of substrate 17d
is illustrated by this technique (Scheme 1).
4
87
83
5d
5e
6d
6e
5[a]
6[b]
95
84
Scope and limitations: Cyclopentadienyltris(acetonitrile)
ruthenium(ii) hexafluorophosphate 4,[9] readily available on
a multigram scale following a procedure recently developed
by our group,[9c] was found to catalyze the intramolecular
Alder–ene reaction of enynes.[4d] Based on this observation,
5 f
5g
6 f
6g
7[c]
8[d]
75
75
5h
6h
6i
Scheme 1. Preparation of vinylcyclopropyl alkyne 17d: a) CH2I2, Et2Zn,
CH2Cl2, 70%. b) PCC, CH2Cl2, 44%. c) (EtO)2P(O)CH2CO2Et, nBuLi,
THF, 88%. d) DIBAL-H, CH2Cl2, 94%. e) nBuLi, THF; ClCO2Me,
76%. PCC=pyridinium chlorochromate, DIBAL-H=diisobutylalumini-
um hydride.
9[e]
5i
2578
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 2577 – 2590

Seven-Membered Rings
Table 1. (Continued)
FULL PAPER
Table 1. (Continued)
Entry
Substrate
Product
Yield [%]
88
Entry
Substrate
Product
Yield [%]
84
17
10
5p
5q
6p
5j
6j
18[h]
67
11[f]
75
73
6q
5e
6k
19
20
21
22
73
72
71
74
12[g]
5r
5s
5t
6r
6s
6t
5 f
6l
6m
13
82 (3.7:1)
5m
6m’
5u
5v
6u
6v
14
77
5n
6n
23
68 (5:1)
6v’
[a] Reaction performed in 1,2-dichloroethane. [b] Reaction performed in
1,2-dichloroethane and 25% catalyst. [c] Reaction performed in the pres-
ence of water in 1,2-dichloroethane. [d] Reaction run in 1,2-dichloro-
ethane with 30% catalyst. [e] Reaction performed at 508C. [f] Reaction
run in dichloromethane with 10% [Ru(cod)(cot)] and HPF6 (aq.) at RT.
[g] Reaction run in dichloromethane with 10% [Ru(ind)2] and HPF6
(aq.) at RT. [h] Reaction run with 20% catalyst at 508C. See Experimen-
tal Section for General Procedure for Ruthenium Catalyzed [5+2] Cy-
cloadditions.
6o
6o’
6o
15
82 (6.2:1)
5o
amined for catalytic activity are [Ru(cod)(cot)], [Ru(h5-
C8H11)2H]BF4, [Ru(h5-C8H11)(CH3CN)3]BF4, and [(bis-inden-
yl)Ru]. In an attempt to “activate” (bis-indenyl)Ru, it was
treated with fluoroboronic acid in acetonitrile to generate in
situ [(indenyl)Ru(CH3CN)3]BF4. The intermolecular reac-
tion between 1-alkoxyvinylcyclopropanes and alkynes was
also examined and found to proceed, albeit only in moder-
ate (50–70%) yields.[10]
16
78 (1:14)
5o’
Table 1 shows that the catalyst is compatible with a varie-
ty of functional groups and substitution patterns on the
alkyne, alkene and cyclopropane. Importantly, a number of
6o’
Chem. Eur. J. 2005, 11, 2577 – 2590
www.chemeurj.org
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2579

B. M. Trost et al.
trisubstituted alkenes react cleanly at room temperature to
afford cycloadducts without any observable trace of olefin
isomerization (entries 4–8). Bicyclic carbocycles, and O- and
N-containing heterocycles can be obtained in good to excel-
lent yields.
For a few sterically encumbered substrates, the reaction
gave better yields when conducted in dichloroethane (en-
tries 5–8), but required increased catalyst loading (entries 6
and 8). If other highly coordinatively unsaturated ruthenium
complexes were used, such as 10% [Ru(cod)(cot)] or [Ru-
(ind)2] in the presence of aqueous HPF6 (entries 11 and 12),
the reaction cleanly afforded the seven-membered 1,3-
dienes, which were presumably derived by isomerization of
the initially formed 1,4-dienes. As shown in entries 19–22,
the cycloaddition of alkoxy group substituted cyclopropanes
form useful cycloheptenone derivatives in good yields.
Entry 23 shows a ketal converted in situ to a ketone fol-
lowed by isomerization of the olefin to form a more stabi-
lized conjugated diene 6v’.
Scheme 3. Mechanism B for transition metal-catalyzed [5+2] cycloaddi-
tions.
There are two plausible mechanisms for the transition
metal catalyzed [5+2] cycloaddition.[3] Mechanism A in-
volves metallacyclopentene formation followed by ring ex-
pansion to a metallacyclooctadiene (Scheme 2). Subsequent
servations. Clearly, a simple enyne system with an appropri-
ate linker could form a ruthenacyclopentene.
In addition, the contrasting results obtained in entries 15
and 16 can be explained by an intermediate ruthenacyclo-
pentene as follows (Scheme 4). The reaction of E-olefin 5o
Scheme 4. Proposed ruthenacyclopentenes derived from trans- and cis-
olefins 5o and 5o’.
Scheme 2. Mechanism A for transition metal-catalyzed [5+2] cycloaddi-
tions.
afforded a 6.2:1 mixture of products favoring cycloaddition
product 6o. Conversely, the reaction of the Z-olefin 5o’ af-
forded a 78% yield of products as a 14:1 mixture in favor of
the 1,4-diene 6o’ which arises from b-hydride elimination.
Examining the ruthenacyclopentene intermediates derived
from E- and Z-olefins, it is apparent that the group that is
trans in the starting olefin is placed in a pseudoequatorial
position on the convex face of the metallacycle intermedi-
ate. The pseudoequatorial group is geometrically better
suited for interaction with the ruthenium center, a situation
similar to what was recently observed in b-hydride elimina-
tion in titanacyclopentenes.[11]
reductive elimination affords the seven-membered ring
product. In contrast, in mechanism B cleavage of the cyclo-
propane occurs first to form a metallacyclohexene, which
undergoes addition to the alkyne to form a metallacyclooc-
tadiene (Scheme 3). So far mechanism B has been suggested
for the Rh-catalyzed [5+2] cycloaddition in a computational
study by Wender and Houk,[3n] whereas the Ru-catalyzed re-
action most likely proceeds according to mechanism A de-
picted in Scheme 2 according to our study.
The evidence to support mechanism A is first drawn by
comparison with the Ru-catalyzed Alder–ene reaction, in
which a ruthenacyclopentene nicely accommodates all ob-
By invoking the same ruthenacycle intermediate, it is also
understandable why substrate 5b’ bearing a cis-alkene was
far less reactive (Scheme 5) than its trans-counterpart 5b
2580
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 2577 – 2590

Seven-Membered Rings
FULL PAPER
of the [5+2] cycloaddition. In particular, we are interested
in how the relative stereochemistry of substituents in the
tether affects the relative stereochemistry of the newly cre-
ated stereogenic center at the bridgehead carbon atom. To
shed light on this issue, we prepared a range of cycloaddi-
tion precursors with one or more substituents in the carbon
tether and subjected them to the Ru-catalyzed [5+2] cyclo-
addition conditions. The results are shown in Table 2. All re-
actions were run with 5–10% [CpRu(CH3CN)3]PF6 (4), 0.1–
0.2m in acetone, at room temperature, unless otherwise
noted.
Scheme 5. Cycloaddition of cis-olefin 5b’.
Table 2. Diastereoselectivity of the Ru-catalyzed [5+2] cycloadditions.[a]
Entry
Substrate
Product
Yield [%][b]
(entry 2, Table 1). The steric interaction between the Cp
ligand and the cyclopropyl group in intermediate 7a likely
disfavors the formation of the intermediate that leads to cy-
cloadduct 6b.
Substrates 5w–5z (Figure 2) are not reactive even at 508C
for extended periods (such as 14 h), a feature that appears
to be related to the linker length between the alkyne and
the vinyl group. Furthermore, no reaction was observed with
compounds 5w and 5x at elevated temperature even in the
presence of a stoichoimetric amount of ruthenium complex
4, ruling out the possibility of forming ruthenenacyclohex-
ene, an irreversible process in mechanism B. Three examples
(entries 13, 15 and 16 in Table 1) led to products containing
1
92 (dr 3.1:1)
5aa
5aa
6aa
6aa
2[c]
73 (dr 5.1:1)
3
75 (dr >20:1)
5bb
6bb
4
86 (dr >10:1)
86 (dr >20:1)
81 (dr >20:1)
5cc
6cc
5
Figure 2. Failed substrates for Ru-catalyzed [5+2] cycloaddition.
5dd
6dd
6[d]
the vinylcyclopropyl moiety. These observations suggest that
mechanism B is not likely to be operational in these cases.
Presumably, the 6,5-bicyclic system of the ruthenacyclo-
pentene intermediate is considerably more difficult to form
than the 5,5-bicyclic system. However, a substrate (entry 18,
Table 1) bearing a trimethylsilyl alkyne as well as a nitrogen
atom in the linker was successfully converted to the desired
6,7-fused bicyclic compound in good yield. Substrate 5zz is
also unreactive under the standard reaction conditions. The
ketone carbonyl group may interact with Ru and inhibit its
catalytic ability in the [5+2] cycloaddition reaction. A more
detailed explanation of the poor reactivity seen with some
substrates bearing proximal carbonyl groups is given later in
the manuscript (see Figure 4, and the corresponding text).
5ee
5 ff
6ee
6 ff
7[e]
70 (dr >20:1)[e]
8
72 (dr >20:1)
6gg
6gg’
5gg
Diastereoselectivity of Ru-catalyzed [5+2] intramolecular
cycloadditions: One major concern in the endeavors to syn-
thesize 5,7-fused bicyclic structures is the diastereoselectivity
Chem. Eur. J. 2005, 11, 2577 – 2590
www.chemeurj.org
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2581

B. M. Trost et al.
Table 2. (Continued)
important role for the coordination of the alkene to Ru than
the back bonding of the d orbital of the metal to the alkene.
The observation that the angular hydrogen is anti to the ho-
moallylic OR substituent in the major diasteromeric cyclo-
adduct is in agreement with the Stork/Houk–Jꢂger “inside
alkoxy” model (Scheme 6).[12] In the electrophilic coordina-
Entry
Substrate
Product
Yield [%][b]
9
65 (dr >20:1)
5hh
6hh
10
90 (dr >20:1)
5ii
6ii
Scheme 6. Mechanistic rationale for diastereoselectivity.
11
70 (dr >20:1)
tion of the alkene with the cationic Ru catalyst to form ruth-
enacyclopentene, the allylic electron-donor substituent
should stabilize the transition state whereas the electron-
withdrawing substituent such as OR should destabilize the
transition state. If the s*CO orbital overlaps with the alkene
p orbital, the alkene becomes less prone to donate electrons
to cationic Ru and slows down the reaction. Conversely, if
the s*CO orbital is orthogonal to the alkene p orbital
(alkoxy group is “inside”), the overlap of s*CO orbital with
the p orbital of the alkene is minimized. Meanwhile, the
electron-donating sC–R’ or sC–C will stabilize the transition
state. Therefore, the most reactive conformation should
place the allylic alkoxy group inside (Scheme 6) so that the
s-donating effect of the alkene p bond to the Ru species is
maximized. This reactive conformation will then lead to the
formation of the diastereomer with the angular hydrogen
and alkoxy group in a trans relationship.
5jj
6jj
[a] Reaction was run with 0.1–0.2m substrate in acetone at RT from 0.5–
6 h unless otherwise noted. [b] Isolated yield and diastereomeric ratio
were determined by 1H NMR. [c] Reaction was run with 0.1m substrate
in DMF at RT for 6 h. [d] Reaction was run in dichloromethane at ꢀ80
to 158C over 2.5 h. [e] 65% product shown +5% desilylated product.
In all cases, good to excellent yields of the desired hydroa-
zulene products were obtained despite the presence, in all
the substrates, of an ionizable functional group at the allylic
position. Modest to excellent diastereoselectivities were ob-
served. In all cases investigated so far, the bridgehead hy-
drogen atom and the hydroxyl group show a trans relation-
ship, which was revealed by NOE experiments and in one
case, by X-ray crystallography (Figure 3).
Regioselectivity of Ru-catalyzed [5+2] intramolecular cyclo-
additions: To further explore the scope and mechanistic de-
tails of Ru-catalyzed intramolecular [5+2] cycloadditions,
we have extensively studied the influence of a substituent R
on the cleavage of disubstituted cyclopropanes (Scheme 7).
For trans disubstituted cyclopropane 8 or cis-disubstituted
ꢀ
cyclopropane 9, either the more- or the less-substituted C C
bond of the cyclopropane can be cleaved to form two possi-
ble regioisomeric products. We chose the same substrate
system already explored by the Wender group as a probe to
compare the Rh- and Ru-catalyzed [5+2] cycloadditions.
The Wender group found that regioselective cleavage of
either the more-substituted or the less-substituted s bond of
the cyclopropane could be achieved depending on the
choice of catalyst.[3k] However, no rationale is provided to
explain this interesting observation. Our results reveal sever-
al divergent mechanistic aspects for the [5+2] cycloadditions
catalyzed by Ru versus Rh.
Figure 3. X-ray and NOE data of representative compounds.
The proposed mechanism (Scheme 2) of the cycloaddition
involves the coordination of RuII to both alkene and alkyne
followed by the ruthenacyclopentene formation in which the
RuII is oxidized to RuIV. Presumably, the s-donating ability
of the alkene p bond to cationic Ru species plays a more
The regioselectivity of this cycloaddition reaction was ex-
plored with a variety of trans- and cis-1,2-disubstituted cy-
clopropanes, summarized in Tables 3 and 4, respectively. In
these reactions, Ru-catalyst 4 displays excellent compatibili-
2582
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 2577 – 2590

Seven-Membered Rings
FULL PAPER
Table 4. Regioselectivity of cycloaddition of cis substrates.
Entry[a]
R
18:19[b]
Yield [%]
1
2
3
4
5
6
7
CO2CH3 (17a)
CN (17b)
CH2OTIPS (17c)
CH3 (17d)
COCH3 (17e)
CꢁCH (17 f)
CHO (17g)
18a:19a > 20:1
18b:19b >20:1
18c:19c >20:1
18d:19d >20:1
18e:19e 2:1
NA
87
81
85
87
93
NR
82
15l:16l[c] 1:12
[a] All reactions were performed with 10% catalyst using 0.1–0.2m sub-
strate in acetone. [b] Ratio determined by proton NMR. [c] See
Scheme 9.
In order to test the possibility of constructing nine-mem-
bered rings by an analogous [7+2] approach, vinylogous al-
dehyde 14d and vinylogous ester 14e were examined (en-
tries 9 and 10, Table 3). However, only seven-membered
ring compounds were obtained with moderate regioselectivi-
ty. Vinylogous aldehyde 14d exhibits a dramatically dimin-
ished bias for migration of the more substituted cyclopropyl
carbon compared to the aldehyde and shows the same pref-
erence as for the vinylogous ester 14e.
Scheme 7. Regioselectivity of the Ru-catalyzed intramolecular [5+2] cy-
cloaddition.
Table 3. Regioselectivity of the cycloaddition of trans substrates.
The [5+2] cycloaddition of trans-cyclopropyl alkyne 14 f
proceeds under mild conditions with a regioselectivity of 2.5
to 1 (entry 11, Table 3). Presumably the [5+2] cycloaddition
is so facile that the formation of a ruthenium vinylidene[13]
structure is negligible. However, cis-cyclopropyl alkyne 17 f
failed to react even at elevated temperature (entry 6,
Table 4).
The use of [In(OTf)3] as an additive had an intriguing
impact on the regioselectivity of the reaction of substrate
14b. This substrate gave a mixture of two regioisomers fa-
voring the formation of 15b in the absence of any additive
(entry 5, Table 3), but favored the other regioisomer 16b in
the presence of 10 mol% [In(OTf)3] (entry 6, Table 3). Pre-
sumably, the coordination of [In(OTf)3] to the ketone car-
bonyl in 14b facilitated the cleavage of the more substituted
Entry[a]
R
15:16[b]
Yield [%]
1
CO2CH3 (14a)
CO2CH3 (14a)
CO2CH3 (14a)
CO2CH3 (14a)
COCH3 (14b)
COCH3 (14b)
COOH (14c)
COOH (14c)
15a:16a 1:2
15a:16a 1:2.5
15a:16a 1:2.3
15a:16a 1:2
15b:16b 2:1
15b:16b 1:1.2
15c:16c 1:3
NA
15d:16d 1:1.6
15e:16e 1:2.5
15 f:16 f 1:2.5
15g:16g 1.5:1
15h:16h 3:1
15h:16h 2:1
15i:16i 1.6:1
15j:16j 1:1.9
15k:16k 1:1
15l:16l 1:15
90
88
80
78
83
88
78
0
82
87
85
90
81
88
71
87
80
78
2[c]
3[d]
4[e]
5
6[d]
7
8[f]
9
(E)-CH=CH-CHO (14d)
(E)-CH=CH-CO2Et (14e)
CꢁCH (14 f)
10
11
12
13
14[c]
15
16
17
18
CH2OTBS (14g)
CH2OTIPS (14h)
CH2OTIPS (14h)
CH2O-4-Br-Bz (14i)
CN (14j)
ꢀ
C C bond of the cyclopropane due to the electronic activa-
ꢀ
tion of the C C bond proximal to the ketone.
Regarding diastereoselectivity, trans cyclopropanes give
trans products, and cis-cyclopropanes, except aldehyde 17g
(Table 4, entry 7), give cis products (the cis or trans geome-
try of the products refers to the relationship of the angular
hydrogen and the R group, see Scheme 7).
SO2Ph (14k)
CHO (14l)
[a] All reactions performed with 10% catalyst by using 0.1–0.2m sub-
strate in acetone unless otherwise noted. [b] Ratio determined by proton
NMR. [c] Reaction performed in DMF. [d] Reaction performed in the
presence of 10–15% In(OTf)3. [e] Reaction performed in the presence of
10% HMPA. [f] Reaction performed in the presence of 10% Bu4NOH.
To rationalize the regioselectivity of the cycloaddition of
disubstituted cyclopropanes, both electronic and steric fac-
tors have to be considered (Scheme 8). If R is an electron-
withdrawing group, electronically, path b should be favored
resulting from the s-bond activation by the electron-with-
drawing group. Sterically, path a is favored particularly for
cis-cyclopropanes, since Rcis is distal to the Cp ligand in in-
termediate 21 but proximal to Cp in intermediate 24. For
trans-cyclopropanes, path a should still be favored but only
ty with substrates containing various functional groups such
as esters, carboxylic acids, aldehydes, ketones, amides, enals,
sulfonamides, sulfones. The Ru catalyst is able to catalyze
the [5+2] cycloaddition of many substrates at room temper-
ature in a short period of time (30 min to 2 h).
Chem. Eur. J. 2005, 11, 2577 – 2590
www.chemeurj.org
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2583

B. M. Trost et al.
nylcyclopropane (Scheme 9).
The regiochemical preference
of this reaction can be ex-
plained by the presence of an
electron withdrawing group
which should stabilize the in-
termediate cyclooctene 25:26,
as well as weaken the carbon–
carbon bond bearing the alde-
hyde. To the extent that the re-
giochemistry is determined by
the relative rate of cyclopro-
pane ring opening, this ex-
plains the regiochemical pref-
erence for cleavage of the cy-
clopropane
carbon–carbon
bond substituted with the alde-
hyde. Once again, metallacy-
clopentene formation is rever-
Scheme 8. Proposed mechanistic rationale for the regio- and diastereoselectivity of cyclopropane ring opening
for disubstituted cyclopropanes.
slightly over path b, because the steric interaction between
Rtrans and Cp in intermediate 21 is slightly more severe than
that in intermediate 24. Therefore, trans cyclopropanes bear-
ing an electron-withdrawing R group, such as an ester,
ketone, carboxylic acid, enal, a,b-unsaturated ester, cyanide
or sulfone (entries 1–10, 16–17 in Table 3) give a mixture of
two diastereomers. This result indicates a small to almost
negligible bias in cleavage of the more substituted cyclo-
propyl carbon bond despite the large differences of the
steric size of the substituents. In contrast, cis-cyclopropanes
bearing an electron-withdrawing R group such as ester or
cyano (entries 1 and 2, Table 4) give one diastereomer
Scheme 9. RuII-catalyzed regioselective [5+2] cycloadditions of 14l and
17g.
almost exclusively.
If R is not an electron-withdrawing group, path a appears
to be favored for both cis- and trans-cyclopropanes. For cis-
cyclopropane substrate 17c (entry 3, Table 4), the steric re-
pulsion between Cp and Rcis is much more severe in inter-
mediate 24 than that in 21, so path a is preferred and pro-
ceeds to give product 18c. In comparison, for trans-cyclopro-
pane substrate 14g and 14h (entries 12 and 13 in Table 3),
the steric repulsion between Cp and Rtrans is only moderately
more severe in intermediate 24 than that in 21 so both cy-
cloheptadiene regioisomers were obtained. Furthermore, the
migration of the less substituted cyclopropyl carbon is fa-
vored, in contrast to the case of the trans-cyclopropylester
14a, where the migration of the more substituted cycloprop-
yl carbon is preferred. Therefore, in cases where R is not an
electron-withdrawing group, steric effects appear to domi-
nate. This conclusion is further supported by the comparison
of entries 12 and 13 in Table 3. As the steric size of R group
increases from tert-butyldimethylsilyloxy methylene to triiso-
propylsilyloxy methylene, the product derived from the less
sible and the product ratios are determined by the relative
ꢀ
rate of C C bond breaking to from the metallacyclooctenes.
The diastereoselectivity of the reaction can be explained
by considering the structure of proposed intermediate ruthe-
nacycles 29 and 30 (Scheme 10). These ruthenacycles can
readily equilibrate via an O-bound ruthenium(+4) eno-
late.[13] To the extent that this equilibrium is rapid relative to
the rate of reductive elimination, the diastereoselectivity of
the reaction will depend on the relative rates of reductive
elimination. The remarkable preference to form the trans-
product 16l (>10:1 dr) regardless of the geometry of the
starting cyclopropane suggests that the rate of reductive
elimination of 30 is greater than that of 29. This behavior is
unique to aldehyde-substituted cyclopropanes. For example,
replacing the aldehyde with an acetyl group has a dramatic
effect on the course of the reaction. Although, the possibili-
ty of equilibration of the intermediate ruthenacycloctadienes
via an O-bound ruthenium enolate remains, the acetyl sub-
stituted-cyclopropanes behave like the siloxymethyl-substi-
tuted cyclopropanes (Scheme 8) rather than the formyl-sub-
stituted cyclopropanes (Scheme 9). This suggests that, in
ꢀ
substituted C C bond cleavage of the cyclopropane is in-
creased from 1.5:1 to 3:1.
If R is a formyl group, polyhydroazulene 16l is formed in-
dependent of the relative stereochemistry of the starting vi-
2584
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 2577 – 2590

Seven-Membered Rings
FULL PAPER
gioselectivity in the cis series
with the Ru catalyst (entry 1,
Table 4).
Construction of tricyclic com-
pounds containing
a seven-
membered ring: In order to
showcase the above described
methodology, we also utilized
Ru-catalyzed [5+2] cycloaddi-
tion to construct the core skel-
eton of the natural products
depicted in Figure 1. In all
cases good to excellent yields
Scheme 10. Proposed mechanism to account for selective formation of 16l.
and diastereoselectivities were
observed in the key reaction
despite the steric encumbrance
order for the equilibration of 29 and 30 to occur, the rela-
tively small steric size of the aldehyde is also necessary. Cal-
culation (MM2) of the relative energies of the product alde-
hydes (16l and 19g) shows that 16l is 1.8 kcalmolꢀ1 (95:5
ratio) more stable than diastereomer (19g). The difference
may primarily be due to placement of the aldehyde into a
pseudoaxial position in 19g compared to the pseudoequato-
rial position in 16l. This energy difference is presumably re-
flected in the transition state for reductive elimination and
therefore in the difference in the relative rates of reductive
elimination.
of the cycloaddition substrates increasing from 1,1,2-trisub-
stituted, to 1,2,3-trisubstituted, to tetrasubstituted cyclopro-
panes (entries 2–10 in Table 5). The relative stereochemistry
of the angular substituents was assigned by NOE studies.
The mechanism involving a ruthenacyclopentene again ac-
counts for the observed stereochemistry (Scheme 11). Coor-
dination of the double bond results in a dihedral angle of 08
between Hb and Hc, since the double bond is formed be-
tween the corresponding carbon atoms. This leads to the
propagation of the stereochemistry in a 1,4-manner. It is
particular striking that tetrasubstituted substrates such as
42i or trisubstituted substrates such as 42g, h, and j can be
employed, invoking a secondary carbon–ruthenium bond.
An interesting regioselectivity question arises with ketone
42h. Previous work in entries 5 and 6 of Table 3 showed that
an acetyl substituent in a 1,2-disubstituted cyclopropane pro-
duced nearly equal amounts of the two regioisomers. On the
other hand, in the case of 42h in which both cyclopropyl
bonds are substituted, a selectivity in favor of 43h was ob-
served, accompanied by a small amount of 43h’ (84% yield,
43/43h’ 6:1). This observation suggests that electronic fac-
tors can dominate when steric effects are equivalent, but the
Three aspects of our observations should be noted. First,
the aldehyde substrate gave different results compared to
other substrates (Scheme 9), perhaps deriving from the com-
ꢀ
bination of both electronic and steric effects. Second, the C
C bond energy of cyclopropyl appears to be important in
the trans series. All substituents, in particular the electron-
withdrawing substituents, show significant migration of the
more substituted cyclopropyl carbon. The steric effects
appear to be quite variable for these cases. They seem to
play a more significant role in entries 12 and 13 compared
with entries 1 and 15 in Table 3. Third, the steric effects
seem to dominate for the cis substrates. Excellent regiose-
lectivity for migration of the less substituted carbon was ob-
served in this series apart from the examples with formyl or
acetyl substituents.
There are several significant differences between the Rh
and the Ru systems. First, aldehyde substrates 14l and 17g
do not lead to the same cycloaddition product with Rh cat-
alysis. With [Rh(CO)2Cl]2, the cycloaddition of 14l gave 16l
with the angular hydrogen trans to the aldehyde group. Con-
versely, 17g afforded a diastereomer of 16l with the angular
hydrogen cis to the aldehyde group. Second, the siloxy sub-
strate 14g shows the contrasting regioselectivity with the
two different catalytic systems, leading with Wilkinsonꢀs cat-
alyst to exclusive formation of 15g, whereas the Ru complex
4 leads to an equimolar mixture of 15g and 16g. Third, the
ester substituent influences the regioselectivity of the Rh-
catalyzed reaction by favoring the cleavage of the more sub-
ꢀ
stituted C C bond of cyclopropane, but does not effect re-
Scheme 11. Rationale for diastereoselectivity of cycloaddition of 42i.
Chem. Eur. J. 2005, 11, 2577 – 2590
www.chemeurj.org
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2585

B. M. Trost et al.
Yield [%]
Table 5. Ru-catalyzed intramolecular [5+2] cycloadditions of vinylcyclo- Table 5. (Continued)
propyl alkynes.[a]
Entry
Substrate
Product
Entry
Substrate
Product
Yield [%]
72
11
85
(dr >15:1)
1
(dr >20:1)
42j
43j
42a
43a
72
12
13
(dr >20:1)
81
2
3
(dr >20:1)
42k
43k
42b
43b
84
(dr 10:1)
80
(dr >20:1)
42l
43l
42c
42d
42e
43c
43d
43e
43 f
91
(dr 10:1)
14
81
4
5
(dr >20:1)
42m
43m
69
(dr >10:1)
electronic effects are not overwhelming. Interestingly, the
addition of indium triflate as a cocatalyst further increased
the selectivity to favor 43h (83% yield) such that none of
isomer 43h’ was detected. Apparently, Lewis acid complexa-
tion of the carbonyl group activates the adjacent cyclopropyl
bond towards cleavage.
Substrate 42k, derived from ketoglutamic acid (entry 12
in Table 5), does not react at room temperature but cycload-
dition occurs once heated to 508C. The terminal alkyne of
42k was methylated accompanied with simultaneous diaster-
eoselective methylation of the a carbon of the lactam[14] to
form 42l. In contrast to substrate 42k, this substrate
smoothly afforded tricyclic product 43l in 84% yield and
10:1 diastereoselectivity at room temperature (entry 13,
Table 4). This observation suggests that reversible formation
of a vinylidene–ruthenium complex by the reaction of the
terminal alkyne with the Ru catalyst slows down but does
not inhibit the cycloaddition pathway when terminal alkynes
are used as substrates.[17]
The diastereoselectivity of the cycloaddition presumably
results from the stability and reactivity of ruthenacyclopen-
tane intermediates 46 and 47 (Scheme 12). Apparently, in-
termediate 47 is disfavored relative to intermediate 46, an
explanation similar to the rationale provided in Scheme 6
can be given. The considerably higher activation energy for
the transformation of 42kb to 47, compared with that of
42ka to 46, dictates the stereochemical outcome of the reac-
tion.
81
6
7
(dr >20:1)
42 f
42g
93
(dr >20:1)
43g
43h
43h’
84
8
(43h:43h’
6:1)
42h
78
9[b]
(dr >20:1)
42h
42i
43h
43i
85
10
Varying the placement of the cyclohexyl ring provides an
entry to alternative tricycles. For example, dienyne 42m was
exposed to Ru catalyst 4 at room temperature in acetone to
(dr >20:1)
2586
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 2577 – 2590

Seven-Membered Rings
FULL PAPER
In contrast to substrates 42k
and l, substrates 42p–s bearing
a basic nitrogen atom failed to
react even at elevated temper-
ature or in the presence of a
Lewis acid such as MeAlCl2
(Figure 5). This lack of reactiv-
ity most likely results from the
basicity of the tertiary amine
which leads to a Lewis acid–
base interaction with the Ru
catalyst,
a rationale already
discussed to account for the
lack of reactivity of a suitably
juxtaposed carbonyl group.
Scheme 12. Rationale for the diastereoselectivity of the cycloaddition of 42k.
give a 91% yield of the triene 43m (dr 10:1) within 4 h
(entry 14, Table 5). The use of a conjugated enyne for the
acetylene segment provides an entry to 1,3-dienes and thus
sets the stage for further elaboration by means of Diels–
Alder reactions.
In sharp contrast to the facile cycloaddition of 42h, which
was complete within 4 h, aldehyde 42n and ketone 42o
failed to react even at elevated temperature (508C), longer
reaction time (24 h), and increased catalyst loading
(25 mol%). Previous work clearly established that the func-
tionality present in this substrate is compatible with the cy-
cloaddition reaction, hence we examined the reduced forms
of ketone 42o (entries 3–6 in Table 5). Gratifyingly, alcohols
42c and 42d, and also the more sterically demanding silyl
ethers 42e and 42 f reacted normally to give tricyclic com-
pounds with excellent diastereoselectivity. One explanation
for these differences in reactivity can be explained by biden-
date coordination of the carbonyl group and the alkene with
ruthenium; a coordination possible in 42n and 42o, but not
in 42c–f, thus resulting in decreased propensity to form the
ruthenacycle intermediate. An alternative explanation
(Figure 4), is that the carbonyl group in ruthenacyclopen-
tene intermediates 46 and 47 may occupy one coordination
site of the Ru catalyst, thus the free coordination site re-
quired for the ring expansion to occur is blocked.
Figure 5. Unreactive substrates for the Ru-catalyzed [5+2] cycloaddition.
Substrates 42t and 42u were unreactive towards cycload-
dition. This presumably is due to the steric demand of the
substrates which precludes formation of the ruthenacyclo-
pentene intermediate. The sensitivity of this family of Ru
catalysts to steric hindrance has been noted in other reac-
tions, notably enyne additions.[17]
Conclusion
Ru-catalyzed intramolecular [5+2] cycloadditions proceed
under very mild conditions and usually the reactions are
complete at room temperature within a few hours. The reac-
tions generally show excellent chemoselectivity. High dia-
stereoselectivities can be achieved in the intramolecular Ru-
catalyzed [5+2] cycloaddition of cyclopropylenynes. A sys-
tematic study of 1,2-trans- and -cis-disubstituted cyclopropy-
lenynes provides tremendous insight towards the under-
standing of the reaction mechanism. Based on our observa-
tions, a mechanism involving a ruthenacyclopentene inter-
mediate is proposed. A ruthenacyclopentene intermediate
nicely explains all the results, particularly regarding the re-
activity of substrates and the stereochemistry of the cycload-
ducts.
A further extrapolation of this methodology led to the
synthesis of 5,7,6-fused, 5,5,7-fused, and 6,5,7-fused ring sys-
tems with high yields and diastereoselectivities. The synthe-
sis of 5,7,6-fused ring systems also serves as a model study
to some biologically important natural products. The high
diastereoselectivity of the reaction is remarkable, likely re-
Figure 4. Ruthenacyclopentene intermediates that failed to undergo cy-
clopropyl ring opening.
Chem. Eur. J. 2005, 11, 2577 – 2590
www.chemeurj.org
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2587

B. M. Trost et al.
127.4, 127.2, 52.4, 46.4, 31.1, 30.7, 30.3, 25.3, 21.5, ꢀ1.7; HRMS (EI+):
m/z: calcd for C20H29NO2SSi: 375.1688; found: 375.1689 [M]+.
sulting from the additional ring on the tether between the
alkene and alkyne, which serves to rigidify the substrate to
favor one particular reactive conformation. Given the reac-
tivity of this catalytic system and the breath of its scope, this
atom economical process holds much promise to enhance
the efficiency of the total synthesis of polycyclic natural
products containing seven-membered rings.
8-Trimethylsilyl-2,3,3a,4,6,7-hexahydro-1H-azulen-5-one (6r) (Table 1,
entry 19): A solution of 5r (74.2 mg, 0.296 mmol) in acetone (2.2 mL)
was added under argon to a test-tube with [CpRu(CH3CN)3]PF6 (9.6 mg,
0.022 mmol). The reaction was stirred at room temperature for 10 h. The
reaction mixture was concentrated in vacuo and the residue purified by
flash chromatography (silica gel, petroleum ether/diethyl ether 10:1) to
yield 6r (48.4 mg, 0.217 mmol, 73%) as a clear light yellow oil. Rf =0.68
(petroleum ether/diethyl ether 1:1); IR (film): n˜
=
2955, 2851, 1707,
1430, 1319, 1284, 1260, 1204, 1138, 1073, 804 cmꢀ1
;
1H NMR (300 MHz,
CDCl3): d = 2.52–2.28 (m, 6H), 1.99 (m, 2H), 1.90 (dd, J=13.1, 10.3 Hz,
1H), 1.36–1.25 (m, 4H), 0.07 (s, 9H); 13C NMR (75 MHz, CDCl3): d =
210.1, 145.2, 132.4, 41.9, 40.2, 39.8, 36.5, 36.4, 31.1, 21.2, 0.1; elemental
analysis calcd (%) for C13H22OSi: C 70.21, H 9.97; found: C 70.45, H
10.05.
Experimental Section
Selected experimental procedures for preparation of 6h, 6l, 6q, 6r, 6ee,
6gg, 6ii, 16j, 16l, 18b, 18c, 43d, and 43i appear below. Full experimental
details for all substrates and cycloaddition products reported herein are
given in the Supporting Information.
1-Methyl-8-(trimethylsilyl)-1,2,3,3a,6,7-hexahydroazulene-1,3-diol (6ee)
(Table 1, entry 19): [CpRu(CH3CN)3]PF6 (5 mol%, 52 mg, 0.118 mmol) at
ꢀ788C was added to a solution of 5ee (600 mg, 2.38 mmol) in dichloro-
methane (12 mL). The solution was warmed to 158C over 2.5 h. Without
workup, the reaction mixture was purified by chromatography (silica gel,
petroleum ether/diethyl ether 1:1) to give 6ee (485 mg, 1.92 mmol, 81%)
as a white solid. M.p. 1378C; IR (film): n˜ = 3277brm, 2919w, 1390w,
General procedure: Ruthenium catalyzed [5+2] cycloadditions: 10%
[CpRu(CH3CN)3]PF6 was added to a oven-dried test tube and the flask
purged with Ar three times. A solution of eneyne-cyclopropane in freshly
distilled acetone was added via cannula and the solution stirred under Ar
at RT until TLC showed the reaction was complete. The solvent was re-
moved in vacuo, and the residue purified by flash chromatography on
silica gel.
1246w, 1072w, 920w, 838w cmꢀ1 1H NMR (300 MHz, CDCl3): d = 5.72–
;
5.60 (m, 1H), 5.55–5.50 (m, 1H), 4.11 (brs, 1H), 3.77 (brs, 1H), 2.60–
2.30 (m, 3H), 2.30–2.10 (m, 2H), 2.06–1.88 (m, 1H), 1.97 (dd, J=13.6,
2.8 Hz, 1H), 1.72 (dd, J=13.6, 4.0 Hz, 1H), 1.39 (s, 3H), 0.16 (s, 9H);
13C NMR (75 MHz, CDCl3): d = 162.7, 138.4, 132.2, 125.8, 79.6, 74.1,
51.0, 49.5, 31.8, 28.5, 27.6, 1.3; elemental analysis calcd (%) for
C14H24O2Si: C 66.61, H 9.58; found: C 66.70, H 9.44. See Supporting In-
formation for X-ray data.
4-Phenyl-8-trimethylsilyl-3,3a,6,7-tetrahydro-1H-azulene-2,2-dicarboxylic
dimethyl ester (6h) (Table 1, entry 8): Malonate 5h (10.1 mg,
0.025 mmol) and [CpRu(CH3CN)3]PF6 (3.1 mg, 0.0076 mmol) were dis-
solved in dichloroethane (0.25 mL) and stirred under argon at room tem-
perature for 24 h. The solvent was removed in vacuo and the residue pu-
rified by flash chromatography (silica gel, petroleum ether/diethyl ether
3:1) to afford cycloadduct 6h (7.6 mg, 0.019 mmol, 75%) as a pale yellow
oil. Rf =0.43 (petroleum ether/diethyl ether 3:1); IR (film): n˜ = 2953,
(1R*,3R*,10R*)-[3-(4-Methoxy-benzyloxy)-1,4-dimethyl-1,2,3,5,6,8a-hex-
ahydro-azulen-1-yloxy]-trimethylsilyl ether (6gg) and (1R*,3R*,10R*)-3-
(4-methoxy-benzyloxy)-1,4-dimethyl-1,2,3,5,6,8a-hexahydroazulen-1-ol
(6gg’) (Table 2, entry 8): [CpRu(CH3CN)3]PF6 (25 mg, 0.057 mmol) at
RT was added to cyclopropane 5gg (0.44 g, 1.14 mmol) in distilled ace-
tone (2 mL). The solution was stirred for 1.5 h. After removal of the sol-
vent, the residue was separated by flash chromatography eluting with 5
2180, 1738, 1629, 1435, 1251, 1198, 1163, 1077, 838, 734, 703 cmꢀ1
;
1H NMR (300 MHz, CDCl3): d = 7.20 (m, 5H), 5.61 (m, 1H), 4.34 (m,
1H), 3.72 (s, 3H), 3.62 (s, 3H), 3.01 (d, J=15.9 Hz, 2H), 2.59 (m, 2H),
2.32 (m, 2H), 2.18 (m, 2H), 0.16 (s, 9H); 13C NMR (75 MHz, CDCl3):
d= 171.8, 171.7, 152.3, 145.3, 140.2, 135.6, 131.1, 127.8, 127.4, 125.9, 58.2,
52.7, 52.6, 43.1, 41.7, 40.7, 28.8, 28.5, ꢀ0.3; elemental analysis calcd (%)
for C23H30O4Si: C 69.31, H 7.59; found: C 69.17, H 7.50.
!
20% diethyl ether in petroleum ether to afford 6gg (140 mg,
0.45 mmol, 39%) and 6gg’ (144 mg, 0.37 mmol, 33%) both as a colorless
oil. For 6gg: IR (film): n˜ =2960m, 2930m, 1613m, 1514s, 1442m, 1375m,
8-tert-Butyl-4-trimethylsilyl-3,4,5,6-tetrahydro-1H-azulene-2,2-dicarboxyl-
ic dimethyl ester (6l) (Table 1, entry 12):
A solution of [Ru(ind)2]
1302m, 1249s, 1173s, 1150m, 1108s, 1074m, 1036s, 889, 840s, 764m cmꢀ1
;
(0.8 mg, 2.4ꢃ10ꢀ3 mmol) in CH2Cl2 (0.2 mL) under argon was treated
with 0.5 mL of a 5.5m solution of HPF6 in water (3ꢃ10ꢀ3 mmol). The
yellow solution immediately turned orange-brown. This solution was stir-
red 10 min, before malonate 5 f (9.1 mg, 0.024 mmol) was added and the
solution stirred at room temperature for 12 h. The solvent was removed
in vacuo and the residue purified by flash chromatography (silica gel, pe-
troleum ether/diethyl ether 3:1) to afford cycloadduct 6l (6.6 mg,
0.017 mmol, 73%) as a colorless oil. Rf =0.55 (petroleum ether/diethyl
ether 3:1); IR (film): n˜ = 2956, 2871, 2180, 1738, 1436, 1363, 1251, 1200,
1H NMR (500 MHz, CDCl3): d = 7.27 (d, J=8.5 Hz, 2H), 6.90 (d, J=
8.5 Hz, 2H), 5.89 (m, 1H), 5.63 (m, 1H), 4.53 (d, J=11.5 Hz, 1H), 4.41
(m, 1H), 4.35 (d, J=11.5 Hz, 1H), 3.82 (s, 3H), 3.55 (m, 1H), 3.31 (s,
1H), 2.58 (m, 1H), 2.40–2.15 (m, 4H), 1.98 (m, 1H), 1.84 (s, 3H), 1.46 (s,
3H); 13C NMR (125 MHz, CDCl3): d = 158.9, 139.4, 135.3, 131.0, 129.0,
128.7, 128.6, 122.3, 78.0, 69.9, 55.2, 54.3, 52.1, 44.5, 32.9, 28.9, 25.9, 20.7;
HRMS: m/z: calcd for C22H31O3Si: 371.2042; found: 371.2041 [MꢀCH3]+.
For 6gg’: IR (film): n˜ = 3496b, 2959m, 2929m, 1613m, 1586w, 1514s,
1438w, 1303w, 1249s, 1174m, 1116m, 1089m, 1034s, 923w, 825m cmꢀ1
;
1
1029, 845, 761 cmꢀ1; H NMR (300 MHz, CDCl3): d = 5.25 (m, 1H), 3.73
1H NMR (500 MHz, CDCl3): d = 7.27 (d, J=8.5 Hz, 2H), 6.88 (d, J=
8.5 Hz, 2H), 5.86 (dd, J=1.5, 11.0 Hz, 1H), 5.82 (m, 1H), 4.53 (d, J=
11.0 Hz, 1H), 4.46 (d, J=4.5 Hz, 1H), 4.40 (d, J=11.0 Hz, 1H), 3.82 (d,
J=0.5 Hz, 3H), 3.55 (s, 1H), 3.34 (d, J=2.0 Hz, 1H), 2.52 (m, 1H), 2.32
(m, 1H), 2.20 (m, 2H), 2.00 (dd, J=6.0, 13.0 Hz, 1H), 1.75 (d, J=2.0 Hz,
3H), 1.59 (dd,=4.5, 14.5 Hz, 1H), 1.34 (s, 3H); 13C NMR (125 MHz,
CDCl3): d = 159.2, 138.4, 138.1, 129.9, 129.8, 128.1, 113.7, 94.1, 79.2, 78.6,
69.8, 55.2, 51.5, 43.0, 32.8, 25.9, 24.7, 21.4; HRMS: m/z: calcd for
C20H26O3: 314.1882; found: 314.1880.
(s, 6H), 3.39 (t, J=7.2 Hz, 2H), 2.87 (s, 2H), 2.40 (m, 1H), 2.08 (m, 2H),
1.91 (m, 2H), 1.01 (s, 9H), 0.09 (s, 9H); 13C NMR (75 MHz, CDCl3): d =
170.6, 170.5, 149.7, 149.1, 147.2, 119.4, 57.3, 57.2, 52.7, 52.6, 36.8, 36.7,
31.8, 30.2, 29.4, 22.4, 1.0, ꢀ0.1; elemental analysis calcd (%) for
C21H34O4Si: C 66.62, H 9.05; found: C 66.40, H 8.99.
N-(4-Methylbenzenesulfonyl)-5-(trimethyl-silyl)-2,3,4,6,7,9a-hexahydro-
1H-cyclohepta[c]pyridine (6q) (Table 1, entry 18): A solution of vinylcy-
clopropane 5q (18 mg, 0.048 mmol) in acetone (0.4 mL) was added to a
test tube containing [CpRu(CH3CN)3]PF6 (4 mg, 0.009 mmol) and the re-
sulting orange solution stirred at 508C for 2 h. The reaction mixture was
concentrated in vacuo and purified by chromatography eluting with pe-
troleum ether/diethyl ether 6:1 to afford 6q (12 mg, 67%) as a clear film.
(1S*,3S*,9R*)-7-(2-Hydroxyethyl)-3-(4-methoxybenzyloxy)-1,4-dimethyl-
1,2,3,5,6,8a-hexahydro-azul-en-1-ol (6ii) (Table 2, entry 10): [CpRu-
(CH3CN)3]PF6 (22 mg, 0.051 mmol) was added under argon at RT to
enyne 5ii (220 mg, 0.512 mmol) in distilled acetone (3 mL). The reaction
was stirred for 2 h and directly purified by flash chromatography (silica
gel, petroleum ether/diethyl ether 20:1!10:1) to yield 6ii (183 mg,
0.461 mmol, 90%) as a single diastereomer as a pale yellow oil. IR
IR (film): n˜
= ;
2955, 2923, 2852, 1599, 1346, 1248, 1161, 1093 cmꢀ1
1H NMR (500 MHz, CDCl3): d = 7.68 (d, J=8.0 Hz, 2H), 7.27 (d, J=
8.0 Hz, 2H), 5.33 (m, 1H), 5.18 (m, 1H), 4.82 (m, 1H), 3.89 (m, 1H),
3.09 (m, 3H), 2.43 (s, 3H), 2.16 (m, 1H), 2.05 (m, 2H), 0.09 (s, 9H);
13C NMR (125 MHz, CDCl3): d = 143.9, 139.3, 136.5, 135.2, 129.6, 128.5,
(film): n˜
= 3407b, 2958s, 2926s, 2856m, 1728s, 1613w, 1514m, 1464m,
2588
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 2577 – 2590

Seven-Membered Rings
FULL PAPER
1378w, 1282m, 1249s, 1121m, 1073m, 1037m, 861w, 821w, 773w, 741w
led acetone (0.6 mL) was degassed with argon for 5 min before [CpRu-
(CH3CN)3]PF6 (4; 11 mg, 0.025 mmol) was added. The resultant yellow
solution was stirred at RT. for 1 h. Flash chromatography afforded 16l
(56 mg, 0.19 mmol, 78%) as colorless oil. It was a mixture of two isomers
with the ratio of 15:1. The relative stereochemistry was determined by
NOE between the angular proton and a-proton of aldehyde. COSY spec-
tra also confirmed the structure of this compound. IR (film): n˜ = 2955m,
2850w, 2723w, 1734s, 1437m, 1274s, 1202m, 1163m, 1078m, 953w, 885w,
cmꢀ1 1H NMR (500 MHz, CDCl3): d = 7.29 (d, J=9.0 Hz, 2H), 6.88 (d,
;
J=9.0 Hz, 2H), 5.52 (s, 1H), 4.46 (t, J=7.0 Hz, 1H), 4.46 (d, J=11.0 Hz,
1H), 4.39 (d, J=11.0 Hz, 1H), 3.82 (s, 3H), 3.86 (m, 2H), 3.50 (m, 1H),
2.47 (t, J=11.5 Hz, 1H), 2.35 (m, 4H), 2.08 (m, 2H), 1.78 (d, J=2.0 Hz,
3H), 1.72 (dd, J=8.0, 13.0 Hz, 1H), 1.46 (s, 3H); 13C NMR (125 MHz,
CDCl3): d = 159.2, 141.1, 136.9, 130.4, 129.8, 125.5, 124.3, 113.7, 78.5,
70.4, 60.1, 55.3, 51.1, 45.8, 42.6, 33.1, 30.3, 29.5, 25.7, 22.0; HRMS: m/z:
calcd for C22H30O4: 358.2144; found: 358.2143 [M]+.
822w, 804w, 749w cmꢀ1 1H NMR (300 MHz, CDCl3): d = 9.67 (d, J=
;
1.8 Hz, 1H), 5.70 (dtd, J=2.7, 6.0, 10.2 Hz, 1H), 5.61 (dt, J=2.2, 10.5 Hz,
1H), 3.74 (s, 3H), 3.72 (s, 3H), 3.60 (m, 1H), 3.17 (m, 1H), 3.06 (d, J=
17.2 Hz, 1H), 2.91 (d, J=1.8, 17.2 Hz, 1H), 2.66 (m, 2H), 2.32 (dt, J=
5.7, 15.3 Hz, 1H), 2.04 (t, J=12.4 Hz, 1H), 1.66 (s, 3H), 1.08 (m, 1H);
13C NMR (75 MHz, CDCl3): d = 202.8, 172.0, 171.8, 138.4, 134.5, 127.7,
124.3, 58.0, 55.0, 52.9, 52.8, 41.2, 41.0, 39.7, 26.0, 19.9; HRMS: m/z: calcd
for C16H20O5: 292.1311; found: 292.1106.
trans-6-Cyano-8-methyl-3,3a,6,7-tetrahydro-1H-azulene-2,2-dicarboxylic
dimethyl ester (15j) and trans-6-cyano-8-methyl-3,3a,6,7-tetrahydro-1H-
azulene-2,2-dicarboxylic dimethyl ester (16j) (Table 3, entry 16): Malo-
nate ester 14j (15 mg, 0.052 mmol) in distilled acetone (0.3 mL) was de-
gassed by argon before the addition of [CpRu(CH3CN)3]PF6 (2 mg,
0.005 mmol). The resulting brown solution was stirred at RT for 3 h and
directly purified by flash chromatography (silica gel, petroleum ether/di-
ethyl ether 20:1!10:1) to afford 15j and 16j as an inseparable mixture
of regioisomers (13 mg, 0.045 mmol, 87%, dr 1:1.9) as a pale yellow oil.
IR (film): n˜ = 2921s, 2851s, 2234w, 1733s, 1456m, 1435m, 1270m, 1201m,
trans-2,2-Bis(methoxycarbonyl)-7-formyl-8-methyl-1,2,3,3a,6,7-hexahy-
droazulene (16l):
A solution of cis-vinylcyclopropane 17g (20 mg,
0.068 mmol) in acetone (0.7 mL) was added to a test tube containing
[CpRu(CH3CN)3]PF6 (3 mg, 0.007 mmol) and the resulting orange so-
lution stirred at room temperature for 30 min. The reaction mixture was
concentrated in vacuo and purified by chromatography eluting with di-
ethyl ether/petroleum ether 1:1 to afford aldehyde 16l (16 mg, 82%) as a
12:1 mixture of diastereomers. The ratio of diastereomers was deter-
mined by 1H NMR integration of the aldehydic proton: for the major dia-
stereomer a doublet at 9.67 ppm (1H) and for the minor diastereomer a
doublet at 9.62 ppm (1H). IR (film): n˜ = 2954, 2851, 1732, 1434, 1273,
1165m, 1072w, 949w, 844w cmꢀ1 1H NMR (500 MHz, CDCl3) for major
;
isomer 16j: d = 5.68 (m, 1H), 5.60 (d, J=10.5 Hz, 1H), 3.772 (s, 3H),
3.766 (s, 3H), 3.62 (m, 1H), 3.05 (d, J=17.0 Hz, 1H), 2.95 (d, J=17.0 Hz,
1H), 2.74 (m, 1H), 2.61 (m, 1H), 2.48 (m, 1H), 2.30 (m, 1H), 2.04 (t, J=
1
13.0 Hz, 1H), 1.88 (s, 3H); H NMR (500 MHz, CDCl3) for minor isomer
15j: d = 5.72 (dd, J=11.0 Hz, 1H), 5.60 (d, J=10.5 Hz, 1H), 3.768 (s,
3H), 3.766 (s, 3H), 3.56 (m, 1H), 3.05 (d, J=17.0 Hz, 1H), 2.95 (d, J=
17.0 Hz, 1H), 2.74 (m, 2H), 2.64 (m, 1H), 2.20 (m, 1H), 2.05 (t, J=
11.0 Hz, 1H), 1.82 (s, 3H); 13C NMR (75 MHz, CDCl3) for major isomer
16j: d = 171.7, 171.4, 137.3, 133.4, 126.4, 123.2, 120.4, 58.1, 52.9, 41.3,
39.6, 39.2, 33.8, 32.0, 29.7, 22.7; 13C NMR (75 MHz, CDCl3) for minor
isomer 15j: d = 171.8, 171.5, 139.4, 135.6, 127.5, 125.5, 122.7, 58.1, 53.0,
41.2, 39.6, 39.7, 34.9, 30.0, 29.4, 22.0; HRMS: m/z: calcd for C16H19O4N:
289.1314; found: 289.1313.
1200, 1163, 1078, 952, 886 cmꢀ1 1H NMR (500 MHz, CDCl3): d = 9.67
;
(d, J=1.8 Hz, 1H), 5.70 (dtd, J=10.2, 6.0, 2.7 Hz, 1H), 5.61 (dt, J=10.2,
2.2 Hz, 1H), 3.74 (s, 3H), 3.72 (s, 3H), 3.60 (m, 1H), 3.17 (m, 1H), 3.06
(d, J=17.2 Hz, 1H), 2.91 (dd, J=17.2, 1.8 Hz, 1H), 2.66 (m, 2H), 2.32
(m, 1H), 2.04 (t, J=12.4, Hz, 1H), 1.66 (s, 3H), 1.08 (m, 21H); 13C NMR
(125 MHz, CDCl3): d = 202.8, 172.0, 171.8, 138.4, 134.5, 127.7, 124.3,
58.0, 55.0, 52.9, 52.8, 41.2, 41.0, 39.7, 26.0, 19.9; HRMS: m/z: calcd for
C16H19O5: 291.1232; found: 291.1258 [MꢀH]+. Additional signals for
cis-6-Cyano-8-methyl-3,3a,6,7-tetrahydro-1H-azulene-2,2-dicarboxylic di-
methyl ester (18b) (Table 4, entry 2): Ester 17b (21 mg, 0.073 mmol) in
distilled acetone (0.8 mL) was degassed by argon before the addition of
[CpRu(CH3CN)3]PF6 (3.2 mg, 0.007 mmol). The resulting red solution
was stirred at RT for 4 h. Flash chromatography eluting with 5!25% di-
ethyl ether in petroleum ether afforded 18b (17 mg, 0.059 mmol, 81%)
as a single diastereomer as a pale yellow oil. IR (film): n˜ = 2954w, 2855w,
minor diastereomer: 1H NMR (500 MHz, CDCl3): d
= 9.62 (d, J=
1.1 Hz, 1H), 5.75 (m, 2H), 3.71 (s, 3H), 3.32 (m, 1H), 2.99 (d, J=
16.3 Hz, 1H), 2.83 (brd, J=16.3 Hz, 1H), 2.16 (m, 1H), 1.40 (s, 3H).
5-Hydroxy-10-methyl-3,3a,5,6,7,8,8a,9-octahydro-1H-benzo[f]azulene-
2,2-dicarboxylic dimethyl ester (43d) (Table 5, entry 4): [CpRu-
(CH3CN)3]PF6 (1.0 mg, 0.0024 mmol) at RT was added to 42d (8.0 mg,
0.024 mmol) in distilled acetone (0.2 mL). The solution was stirred for
4 h. Without further work-up, flash chromatography eluting with 20!
50% diethyl ether in petroleum ether to afford 43d (5.6 mg, 0.017 mmol,
2238w, 1783s, 1435m, 1258s, 1204s, 1165s, 1062m, 756w, 668w cmꢀ1
;
1H NMR (500 MHz, CDCl3): d = 5.61 (dt, J=2.5, 11.5 Hz, 1H), 5.50
(dq, J=4.0, 11.5 Hz, 1H), 3.76 (s, 3H), 3.75 (s, 3H), 3.66 (bm, 1H), 3.33
(m, 1H), 2.98 (d, J=16.5 Hz, 1H), 2.93 (d, J=13.0 Hz, 1H), 2.85 (m, J=
17.0 Hz, 1H), 2.73 (ddd, J=2.0, 8.5, 13.0 Hz, 1H), 2.24 (dt, J=1.5,
11.5 Hz, 1H), 1.96 (dd, J=11.0, 13.0 Hz, 1H), 1.73 (s, 3H); 13C NMR
(125 MHz, CDCl3): d = 171.61, 171.55, 138.9, 134.4, 126.4, 122.9, 120.8,
58.1, 52.9, 52.8, 41.5, 39.9, 38.9, 36.0, 29.7, 20.8; HRMS: m/z: calcd for
C16H19O4N: 289.1314; found: 289.1320.
81%) as a colorless oil. IR (film): n˜
=
3408b, 2926s, 2855m, 1736s,
1434m, 1268s, 1203s, 1162m, 1121m, 1169m cmꢀ1
;
1H NMR (500 MHz,
CDCl3): d = 5.34 (brs, 1H), 4.10 (dd, J=2.5, 3.0 Hz, 1H), 3.75 (s, 6H),
3.56 (m, 1H), 2.93 (d, J=1.0 Hz, 2H), 2.84 (d, J=14.5 Hz, 1H), 2.73
(ddd, J=1.0, 9.0, 13.0 Hz, 1H), 2.65 (m, 1H), 1.97 (dd, J=9.0, 13.0 Hz,
1H), 1.87 (m, 1H), 1.80 (dd, J=4.0, 14.0 Hz, 1H), 1.73 (d, J=1.5 Hz,
3H), 1.58 (m, 1H), 1.53 (m, 2H), 1.44 (m, 1H), 1.32 (m, 2H); 13C NMR
(125 MHz, CDCl3): d = 172.0, 143.9, 137.4, 128.0, 126.0, 76.1, 58.9, 52.74,
52.66, 42.3, 38.6, 38.5, 37.8, 34.7, 34.6, 33.3, 30.3, 29.7; HRMS: m/z: calcd
for C19H26O5: 334.1780; found: 334.1778.
2,2-Bis(methoxycarbonyl)-8-methyl-6-(triisopropylsilyloxymethyl)-
1,2,3,3a,6,7-hexahydroazulene (18c) (Table 4, entry 3): A solution of vi-
nylcyclopropane 17c (33 mg, 0.073 mmol) in acetone (0.7 mL) was added
to a test tube containing [CpRu(CH3CN)3]PF6 (3 mg, 0.007 mmol) and
the resulting orange solution stirred at room temperature for 2 h. The re-
action mixture was concentrated in vacuo and purified by chromatogra-
phy eluting with 5% diethyl ether/petroleum ether to afford 18c (28 mg,
85%) as a colorless liquid. IR (film): n˜ = 2944, 2865, 1738, 1434, 1256,
4-Methyl-5-oxo-3,4a,5,6,7,8,8a,10a-octahydro-1H-benz[f]azulene-2,2-di-
carboxylic acid dimethyl ester (43h) (Table 5, entry 9): [CpRu-
(CH3CN)3]PF6 (1.5 mg, 0.0036 mmol) and In(OTf)3 (9 mg, 0.018 mmol)
was added under argon at RT to a solution of 42h (12 mg, 0.036 mmol)
in acetone (0.5 mL). The mixture was stirred for 4 h and purified by flash
chromatography eluting with 5!20% diethyl ether in petroleum ether
without workup, to afford 43h (9.3 mg, 0.028 mmol, 78%) as a pale
yellow oil. IR (film): n˜ = 3008w, 2953m, 2923m, 2852w, 1736s, 1689s,
1437m, 1329w, 1287m, 1246m, 1210s, 1123w, 1066m, 1031w, 967w, 899w
1202, 1164, 1111, 1063 cmꢀ1 1H NMR (500 MHz, CDCl3): d = 5.47 (m,
;
1H), 5.43 (m, 1H), 3.75 (s, 3H), 3.74 (s, 3H), 3.57 (dd, J=9.3, 6.0 Hz,
1H), 3.55 (m, 1H), 3.47 (dd, J=9.3, 7.5 Hz, 1H), 3.02 (d, J=16.3 Hz,
1H), 2.89 (dd, J=16.3, 2.1 Hz, 1H), 2.68 (ddd, J=12.4, 8.5, 1.8 Hz, 1H),
2.44 (t, J=12.4 Hz, 1H), 2.30 (m, 1H), 1.99 (ddd, J=13.0, 2.0, 1.7 Hz,
1H), 1.95 (dd, J=13.0, 11.0 Hz, 1H), 1.75 (s, 3H), 1.08 (m, 21H);
13C NMR (125 MHz, CDCl3): d = 172.0, 136.6, 131.1, 130.6, 129.0, 68.1,
58.5, 52.7, 52.6, 42.0, 40.2, 39.5, 38.8, 37.6, 20.9, 18.0, 12.0; HRMS: m/z:
calcd for C25H42O5Si: 450.2802; found: 450.2837.
cmꢀ1 1H NMR (500 MHz, CDCl3): d = 5.76 (dt, J=10.0, 2.5 Hz, 1H),
;
5.52 (ddd, J=2.5, 6.5, 9.5 Hz, 1H), 3.76 (s, 3H), 3.75 (s, 3H), 3.55 (brs,
1H), 3.50 (m, 1H), 3.24 (brs, 1H), 3.10 (d, J=17.5 Hz, 1H), 2.97 (dd, J=
1.5, 17.5 Hz, 1H), 2.61 (ddd, J=2.0, 8.0, 12.5 Hz, 1H), 2.39 (m, 2H), 2.14
(t, J=12.5 Hz, 1H), 2.08 (m, 2H), 1.90 (m, 1H), 1.84 (m, 1H), 1.56 (s,
trans-2,2-Bis(methoxycarbonyl)-7-formyl-8-methyl-1,2,3,3a,6,7-hexahy-
droazulene (16l) (Scheme 9): Compound 14l (72 mg, 0.25 mmol) in distil-
Chem. Eur. J. 2005, 11, 2577 – 2590
www.chemeurj.org
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2589

B. M. Trost et al.
3H); 13C NMR (125 MHz, CDCl3): d = 210.5, 172.3, 171.8, 135.2, 134.9,
129.9, 124.6, 59.1, 57.8, 52.8, 52.7, 42.3, 40.7, 40.6, 39.71, 39.65, 30.0, 24.1,
22.3; HRMS: m/z: calcd for C19H24O5: 332.1624; found: 332.1622.
Love, J. Am. Chem. Soc. 1998, 120, 1940–1941; c) P. A. Wender,
C. O. Husfeld, E. Langkopf, J.A. Love, N. Pleuss, Tetrahedron 1998,
54, 7203–7220; d) P. A. Wender, D. Sperandio, J. Org. Chem. 1998,
63, 4164–4165; e) P. A. Wender, F. Glorius, C. O. Husfeld, E. Lang-
kopf, J. A. Love, J. Am. Chem. Soc. 1999, 121, 5348–5349; f) P. A.
Wender, M. Fuji, C. O. Husfeld, J. A. Love, Org. Lett. 1999, 1, 137–
139; g) P. A. Wender, H. Rieck, M. Fuji, J. Am. Chem. Soc. 1998,
120, 10976–10977; h) P. A. Wender, A. J. Dyckman, C. O. Husfeld,
M. J. C. Scanio, Org. Lett. 2000, 2, 1609–1611; i) P. A. Wender, C. M.
Barzilay, A. J. Dyckman, J. Am. Chem. Soc. 2001, 123, 179–180;
j) P. A. Wender, T. M. Pedersen, M. J. C. Scanio, J. Am. Chem. Soc.
2002, 124, 15154–15155; k) P. A. Wender, A. J. Dyckman, Org. Lett.
1999, 1, 2089–2092; l) P. A. Wender, T. J. Williams, Angew. Chem.
2002, 114, 4732–4735; Angew. Chem. Int. Ed. 2002, 41, 4550–4553;
m) P. A. Wender, G. G. Gamber, M. J. C. Scanio, Angew. Chem.
2001, 113, 4013–4015; Angew. Chem. Int. Ed. 2001, 40, 3895–3897;
n) Z.-X. Yu, P. A. Wender, K. N. Houk, J. Am. Chem. Soc. 2004,
126, 9154–9155.
8a-(tert-Butyldimethylsilyloxymethyl)-4-methyl-3,4a,5,6,7,8,8a,10a-octa-
hydro-1H-benzo[f]azulene-2,2-dicarboxylic dimethyl ester (43i) (Table 5,
entry 10): [CpRu(CH3CN)3]PF6 (2 mg, 0.004 mmol) was added to malo-
nate ester 42i (19 mg, 0.047 mmol) in acetone (0.5 mL). The resulting red
solution was stirred at RT for 3 h. Without workup, flash chromatography
of the reaction mixture eluting with 5!10% diethyl ether in petroleum
ether afforded a yellow oil isolated as a single diastereomer of the tricy-
clic compound 43i (16 mg, 0.040 mmol, 85%). IR (film): n˜
= 2930s,
2857s, 1768s, 1462w, 1435w, 1258s, 1197m, 1162m, 1108m, 1076m, 1007w,
1
939w, 837m, 775m, 667w cmꢀ1; H NMR (500 MHz, CDCl3): d = 5.63 (d,
J=11.5 Hz, 1H), 5.36 (d, J=11.0 Hz, 1H), 3.77 (s, 3H), 3.75 (s, 3H), 3.35
(d, J=9.5 Hz, 1H), 3.24 (m, J=11.0 Hz, 1H), 2.99 (d, J=17.5 Hz, 1H),
2.88 (d, J=17.5 Hz, 1H), 2.53 (dd, J=6.5, 14.0 Hz, 1H), 1.98 (m, 2H),
1.70 (m, J=8.0 Hz, 2H), 1.64 (s, 3H), 1.60 (m, 1H), 1.51 (m, 2H), 1.22
(m, 4H), 0.89 (s, 9H), ꢀ0.004 (s, 6H); 13C NMR (125 MHz, CDCl3): d =
172.7, 172.5, 134.7, 130.1, 130.0, 128.3, 66.4, 56.8, 52.79, 52.75, 47.9, 45.3,
42.7, 41.6, 39.7, 36.4, 30.1, 27.3, 25.9, 22.5, 22.2, 18.3, ꢀ5.59, ꢀ5.72;
HRMS: m/z: calcd for C22H33O5Si: 405.2097; found: 405.2103; elemental
analysis calcd (%) for C26H42O5Si: C 67.49, H 9.16; found: C 67.35, H
9.33.
[4] a) B. M. Trost, L. Drause, M. Portnoy, J. Am. Chem. Soc. 1997, 119,
11319–11320; b) B. M. Trost, G. D. Probst, A. Schoop, J. Am.
Chem. Soc. 1998, 120, 9228–9236; c) B. M. Trost, M. Portnoy, H.
Kurihara, J. Am. Chem. Soc. 1997, 119, 836–837; d) B. M. Trost,
F. D. Toste, Tetrahedron Lett. 1999, 40, 7739–7743.
[5] Preliminary reports of portions of this work have appreared, see:
a) B. M. Trost, F. D. Toste, H. Shen, J. Am. Chem. Soc. 2000, 122,
2379–2380; b) B. M. Trost; H. C. Shen, Angew. Chem. 2001, 113,
2375–2378; Angew. Chem. Int. Ed. 2001, 40, 2313–2316; c) B. M.
Trost, H. C. Shen, Org. Lett. 2000, 2, 2523–2525; d) B. M. Trost,
H. C. Shen, T. Schulz, C. Koradin, H. Schirok, Org. Lett. 2003, 5,
4149–4151.
[6] P. Ramesh, N. Srinivasa, Y. Venkateswarlu, M.V. R. Reddy, D. J.
Faulkner, Tetrahedron Lett. 1998, 39, 8217–8220.
[7] G. R. Pettit, R. H. Ode, C. L. Herald, R. B. Von Dreele, C. Michel,
J. Am. Chem. Soc. 1976, 98, 4677–4678.
[8] K. Nakanishi, T. Goto, S. Ito, S. Natori, S. Nozoe, Natural Product
Chemistry, Vol. 1, Academic Press, New York, 1974, p. 285.
[9] a) R. A. Zelonka, M. C. Baird, J. Organomet. Chem. 1972, 44, 383–
389; b) T. P. Gill, K. R. Mann, Organometallics 1982, 1, 485–488;
c) B. M. Trost, C. M. Older, Organometallics 2002, 21, 2544–2546.
[10] D. B. Horne, PhD Thesis, Stanford University (USA), 2004.
[11] S. J. Sturla, N. M. Kablaoui, S. L. Buchwald, J. Am. Chem. Soc. 1999,
121, 1976–1977.
[12] a) J. Haller, T. Strassner, K. N. Houk, J. Am. Chem. Soc. 1997, 119,
8031–8034; b) K. N. Houk, S. R. Moses, Y.-D. Wu, N. G. Rondan, V.
Jꢂger, R. Schohe, F. R. Fronczek, J. Am. Chem. Soc. 1984, 106,
3880–3882.
[13] B. M. Trost, G. Dyker, R. Kulawiec, J. Am. Chem. Soc. 1990, 112,
7809–7811.
[14] A. J. Mancuso, D. Swern, Synthesis 1981, 165–185.
[15] P. A. Jacobi, K. Lee, J. Am. Chem. Soc. 2000, 122, 4295–4303.
[16] B. M. Trost, B. Vidal, M. Tommen, Chem. Eur. J. 1999, 5, 1055–
1069, and references therein.
Acknowledgements
We thank the National Science Foundation and the National Institutes of
Health, General Medical Sciences (GM33049), for their generous support
of our research. Fellowship support from the Stanford Graduate Fellow-
ship (H.C.S. and F.D.T.) and Eli Lilly and Co. (D.B.H.), is gratefully ac-
knowledged. We also thank Professor Paul Wender and his group for
many helpful discussions and comparison spectra. Mass spectra were pro-
vided by the Mass Spectrometry Regional Center of the University of
California-San Francisco supported by the NIH Division of Research Re-
sources.
[1] For reviews on metal-catalyzed cycloadditions, see: a) M. Lautens,
W. Klute, W. Tam, Chem. Rev. 1996, 96, 49–92; b) L. S. Hegedus,
Coord. Chem. Rev. 1997, 161, 129–255; c) C. P. Dell, Contemp. Org.
Synth. 1997, 4, 87–117; d) H.-W. Frꢄhauf, Chem. Rev. 1997, 97, 523–
596; e) M.-H. Filippini, J. Rodriguez, Chem. Rev. 1999, 99, 27–76;
f) L. Yet, Chem. Rev. 2000, 100, 2963–3008; g) I. Ojima, M. Tzamar-
iousdaki, Z. Li, R. J. Donovan, Chem. Rev. 1996, 96, 635–662;
h) J.A. Love, PhD Thesis, Stanford University (USA), 2001; i) R. C.
Hartley, S. T. Caldwell, J. Chem. Soc. Perkin Trans. 1 2000, 477–501;
j) P. A. Wender, F. C. Bi, G. G. Gamber, F. Gosselin, R. D. Hubbard,
M. J. C. Scanio, R. Sun, T. J. Williams, L. Zhang, Pure Appl. Chem.
2002, 74, 25–31.
[17] B. M. Trost, H. C. Shen, A. B. Pinkerton, Chem. Eur. J. 2002, 8,
2341–2349.
[2] a) B. M. Trost, Angew. Chem. 1995, 107, 285–307; Angew. Chem.
Int. Ed. Engl. 1995, 34, 259–281; b) B. M. Trost, M. J. Krische, Syn-
lett 1998, 1–16.
[3] a) P. A. Wender, H. Takahashi, B. Witulski, J. Am. Chem. Soc. 1995,
117, 4720–4721; b) P. A. Wender, C. O. Husfeld, E. Langkopf, J.A.
Received: October 19, 2004
Published online: February 28, 2005
2590
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 2577 – 2590