Development, synthetic scope, and mechanistic studies of the palladium-catalyzed cycloisomerization of functionalized 1,6-dienes in the presence of silane

Article abstract of DOI:10.1021/ja001730+

A 1:1 mixture of the π-allyl palladium complex (η³-C₃H₅)Pd(CI)PCy₃ (1a) and NaB[3,5-C₆H₃-(CF₃)₂]₄ in the presence of HSiEt₃ catalyzed the cycloisomerization of diethyl diallylmalonate (2b) to form 4,4-dicarbomethoxy-1,2-dimethylcyclopentane (3b) in 98% yield with 98% isomeric purity. The procedure tolerated a range of functionality including esters, ketones, sulfones, protected alcohols, and substitution at the allylic and terminal olefinic carbon atoms. Cycloisomerization of 2b obeyed zero-order kinetics to > 3 half-lives with initial formation of 1,1-dicarboethoxy-4-methyl-3-methylenecyclopentane (4b), followed by secondary isomerization to 3b. Deuterium labeling studies revealed that the conversion of 2b to 4b was accompanied by significant H/D exchange, consistent with an addition/elimination pathway coupled with facile H/D exchange of the Pd-H(D) intermediates with free silane.

Full text of DOI:10.1021/ja001730+

J. Am. Chem. Soc. 2000, 122, 10017-10026
10017
Development, Synthetic Scope, and Mechanistic Studies of the
Palladium-Catalyzed Cycloisomerization of Functionalized 1,6-Dienes
in the Presence of Silane
Philip Kisanga and Ross A. Widenhoefer*
Contribution from the P. M. Gross Chemical Laboratory, Duke UniVersity,
Durham, North Carolina 27708-0346
ReceiVed May 18, 2000
Abstract: A 1:1 mixture of the π-allyl palladium complex (η3-C3H5)Pd(Cl)PCy3 (1a) and NaB[3,5-C6H3-
(CF3)2]4 in the presence of HSiEt3 catalyzed the cycloisomerization of diethyl diallylmalonate (2b) to form
4,4-dicarbomethoxy-1,2-dimethylcyclopentane (3b) in 98% yield with 98% isomeric purity. The procedure
tolerated a range of functionality including esters, ketones, sulfones, protected alcohols, and substitution at the
allylic and terminal olefinic carbon atoms. Cycloisomerization of 2b obeyed zero-order kinetics to >3 half-
lives with initial formation of 1,1-dicarboethoxy-4-methyl-3-methylenecyclopentane (4b), followed by secondary
isomerization to 3b. Deuterium labeling studies revealed that the conversion of 2b to 4b was accompanied by
significant H/D exchange, consistent with an addition/elimination pathway coupled with facile H/D exchange
of the Pd-H(D) intermediates with free silane.
Introduction
The transition metal-catalyzed cycloisomerization of enynes
is an effective method for the synthesis of functionalized
cyclopentane derivatives.1,2 The most highly developed enyne
cycloisomerization procedures use palladium (II) salts in
conjunction with phosphine or bidentate nitrogen ligands as
catalysts to form either dialkylidene cyclopentanes or alkenyl
alkylidene cyclopentanes (eq 1).2,3 Other transition metal
complexes that have been used as enyne cycloisomerization
catalysts include ruthenium complexes such as RuClH(CO)-
(PPh3)34 and CpRu(COD)Cl,5 polymer-bound nickel complexes,6
and low-valent titanocene complexes.7 Related transformations
include the cycloisomerization/rearrangement of 1,6- or 1,7-
enynes to form alkenylcycloalkenes catalyzed by the dimeric
ruthenium complex [RuCl2(CO)3]2 (eq 2),8 and the Rh(I)-
catalyzed cycloisomerization of dienynes9 or vinylcyclopropyl
alkynes10 to form more complex polycyclic products.
The development of diene cycloisomerization has lagged
behind enyne cycloisomerization because most enyne cycloi-
somerization catalysts display little reactivity toward dienes.
However, neutral scandocene11 and cationic zirconocene12
complexes catalyze the cycloisomerization of unfunctionalized
dienes (eq 3), although the synthetic utility of these procedures
is limited by the high oxophilicity and excessive air and moisture
sensitivity of the catalysts. Similarly, early attempts to catalyze
diene cycloisomerization with Pd or Rh salts required forcing
conditions in an acidic medium.13 However, selective diene
cycloisomerization under mild conditions has been demonstrated
recently. For example, Ru(I),14 Ni(II),15,16 or Ti(II)17 complexes
(1) (a) Ojima, I.; Tzamarioudaki, M.; Li, Z.; Donovan, R. J. Chem. ReV.
1996, 96, 635. (b) Lautens, M.; Klute, W.; Tam, W. Chem. ReV. 1996, 96,
49. (c) Trost, B. M. Science 1991, 254, 1471. (d) Trost, B. M. Angew.
Chem., Int. Ed. Engl. 1995, 34, 259. (e) Negishi, E.-i.; Coperet, C.; Ma, S.;
Liou, S.-Y.; Liu, F. Chem. ReV. 1996, 96, 365.
(2) (a) Trost, B. M.; Lautens, M. J. Am. Chem. Soc. 1985, 107, 1781.
(b) Trost, B. M. Acc. Chem. Res. 1990, 23, 34. (c) Trost, B. M.; Krische,
M. J. Synlett 1998, 1.
(3) (a) Trost, B. M.; Lautens, M.; Chan, C.; Jebaratnam, D. J.; Mueller,
T. J. Am. Chem. Soc. 1991, 113, 636. (b) Trost, B. M.; Gelling, O. J.
Tetrahedron Lett. 1993, 34, 8233. (c) Trost, B. M.; Tanoury, G. J. J. Am.
Chem. Soc. 1988, 110, 1636.
(4) Nishida, M.; Adachi, N.; Onozuka, K.; Matsumura, H.; Mori, M. J.
Org. Chem. 1998, 63, 9158.
(5) Trost, B. M.; Toste, F. D. J. Am. Chem. Soc. 2000, 122, 714.
(6) (a) Trost, B. M.; Tour, J. M. J. Am. Chem. Soc. 1987, 109, 5268. (b)
Trost, B. M.; Tour, J. M. J. Am. Chem. Soc. 1988, 110, 5231.
(7) (a) Sturla, S. J.; Kablaoui, N. M.; Buchwald, S. L. J. Am. Chem.
Soc. 1999, 121, 1976.
(10) Wender, P. A.; Dyckman, A. J.; Husfeld, C. O.; Kadereit, D.; Love,
J. A.; Rieck, H. J. Am. Chem. Soc. 1999, 121, 10442. (b) Wender, P. A.;
Sperandio J. Org. Chem. 1998, 63, 4164.
(11) Piers, W. E.; Shapiro, P. J.; Bunel, E. E.; Bercaw, J. E. Synlett 1990,
74.
(12) Christoffers, J.; Bergman, R. G. J. Am. Chem. Soc. 1996, 118, 4715.
(13) (a) Grigg, R.; Mitchell, T. R. B.; Ramasubbu, A. J. Chem. Soc.,
Chem. Commun. 1980, 27. (b) Grigg, R.; Malone, J. F.; Mitchell, T. R. B.;
Ramasubbu, A.; Scott, R. M. J. Chem. Soc., Perkin Trans. 1 1984, 1745.
(c) Grigg, R.; Mitchell, T. R. B.; Ramasubbu, A. J. Chem. Soc., Chem.
Commun. 1979, 669.
(8) Chatani, N.; Morimoto, T.; Muto, T.; Murai, S. J. Am. Chem. Soc.
1994, 116, 6049.
(9) Chatani, N.; Kataoka, K.; Murai, S.; Furukawa, N.; Seki, Y. J. Am.
Chem. Soc. 1998, 120, 9104.
(14) Yamamoto, Y.; Ohkoshi, N.; Kameda, M.; Itoh, K. J. Org. Chem.
1999, 64, 2178.
(15) Radetich, B.; RajanBabu, T. V. J. Am. Chem. Soc. 1998, 120, 8007.
(16) Bogdanovic, B. AdV. Organomet. Chem. 1979, 17, 105.
10.1021/ja001730+ CCC: $19.00 © 2000 American Chemical Society
Published on Web 09/27/2000

10018 J. Am. Chem. Soc., Vol. 122, No. 41, 2000
Kisanga and Widenhoefer
catalyze the selective cycloisomerization of 1,6-dienes to form
methylenecyclopentanes in good yield (eq 4). In contrast, a
general and selective catalyst for the isomerization of 1,6-dienes
to cyclopentenes has not been identified.18,19 Here we report
the development, synthetic scope, and mechanistic studies of a
palladium-catalyzed, silane-promoted procedure for the selective
conversion of 1,6-dienes to 1,2-disubstituted cyclopentenes
under mild conditions.20
Scheme 1
Table 1. Cycloisomerization of 2a in the Presence of Silane (1.5
equiv) and a Catalytic 1:1 Mixture of 1a and NaBAr4 (5 mol %) at
25 °C in CH2Cl2 as a Function of Silane
entry
silane
HSiEt3
HSiMe2Ph
HSiMe2Et
HSiMe2t-Bu
time
yield (%)a
isomer ratiob
Results
1
2
3
4
20 min
20 min
10 min
12 h
86
74
95
92
48:1
22:1
52:1
14:1
Identification of an Active Catalyst System. We recently
reported several related procedures for the cyclization/hydrosi-
lylation of functionalized dienes catalyzed by cationic palladium
phenanthroline21 and pyridine-oxazoline complexes.22 These
complexes were initially targeted as diene cyclization/hydrosi-
lylation catalysts because of their high activity as catalysts for
olefin dimerization,23 ethylene/CO copolymerization,24 and
olefin hydrosilylation.25 Likewise, the cationic π-allyl palladium
complex [(η3-C3H5)Pd(OEt2)PCy3]+ [BAr4]- [Ar ) 3,5-C6H3-
(CF3)2] (1) catalyzes the dimerization of methyl acrylate,26 and
was therefore considered as a potential cyclization/hydrosily-
lation catalyst. However, complex 1 failed to catalyze diene
cyclization/hydrosilylation and instead catalyzed the highly
selective cycloisomerization of 1,6-dienes to 1,2-dimethylcy-
clopentenes. For example, treatment of dimethyl diallylmalonate
(2a), with an excess of triethylsilane (1.5 equiv) and a catalytic
amount of 1 [generated in situ from a 1:1 mixture of (η3-C3H5)-
Pd(Me)PCy3 and HBAr4‚OEt2]26 (5 mol %) formed an initially
colorless solution which darkened rapidly after 15 min at room
temperature. Evaporation of solvent and chromatography of the
residue gave 4,4-dicarbomethoxy-1,2-dimethylcyclopentene (3a)
as the exclusive product in 89% yield as a 49:1 mixture of
isomers (Scheme 1).
a Yield refers to isolated material of >95% purity. b Determined by
capillary GC.
Scheme 2
Cationic palladium phenanthroline and pyridine-oxazoline
catalysts used in diene cyclization/hydrosilylation were gener-
ated in situ either from protonation of dimethyl precatalysts
(N-N)PdMe2 with HBAr4‚OEt2 or by halide abstraction from
methyl chloride precatalysts (N-N)Pd(Cl)Me with NaBAr4.21,22
In general, the latter method produced cleaner reaction mixtures
and also benefited from the enhanced thermal stability and
greater availability of the precatalysts. In a similar manner,
halide abstraction from the palladium chloride complex (η3-C3H5)-
Pd(Cl)PCy3 (1a) with NaBAr4 generated an active diene cyclo-
isomerization catalyst.27 For example, reaction of 2a with HSiEt3
and a catalytic mixture of 1a and NaBAr4 for 20 min at room
temperature led to complete consumption of the starting material
and isolation of 3a in 86% yield as 48:1 mixture of isomers
(Table 1, entry 1).
Although not consumed in the reaction, silane was crucial
for efficient and selective palladium-catalyzed diene cyclo-
isomerization. For example, treatment of diethyl diallylmalonate
(2b) with a catalytic mixture of 1a and NaBAr4 in the absence
of HSiEt3 led to 78% completion after 16 h at room temperature
to form a 1.8:3.2:1.0 mixture of 3b, 1,1-dicarboethoxy-3-methyl-
4-methylenecyclopentane (4b), and 4,4-dicarboethoxy-1,5-hep-
tadiene (5b), which together accounted for >90% of the
products (Scheme 2). Products 4b and 5b were identified by
gas chromatography-mass spectrometry (GCMS) and compari-
son to authentic samples. Use of substoichiometric amounts of
HSiEt3 led to inefficient and unselective cycloisomerization.
Dimethylphenylsilane and dimethylethylsilane also served as
(17) Okamoto, S.; Livinghouse, T. J. Am. Chem. Soc. 2000, 122, 1223.
(18) Rajanbabu has reported 74% selectivity (91% yield) for the
conversion of 2a to 3a catalyzed by a 1:2:2 mixture of [(η3-C3H5)PdCl]2,
AgOTf, and P(4-C6H4OMe)3.15
(19) Other examples of palladium-catalyzed diene cycloisomerization:
(a) Heumann, A.; Moukhliss, M. Synlett 1998, 1211. (b) Schmitz, E.; Heuck,
U.; Habisch, D. J. Prakt. Chem. 1976, 318, 471. (b) Schmitz, E.; Urban,
R.; Heuck, U.; Zimmerman, G.; Grundemann, E. J. Prakt. Chem. 1976,
318, 185.
(20) A preliminary account has appeared: Widenhoefer, R. A.; Perch,
N. S. Org. Lett. 1999, 1, 1103.
(21) (a) Widenhoefer, R. A.; DeCarli, M. A. J. Am. Chem. Soc. 1998,
120, 3805. (b) Stengone, C. N.; Widenhoefer, R. A. Tetrahedron Lett. 1999,
40, 1451. (c) Widenhoefer, R. A.; Stengone, C. N. J. Org. Chem. 1999, 64,
8681. (d) Widenhoefer, R. A.; Vadehra, A. Tetrahedron Lett. 1999, 40,
8499. (e) Pei, T.; Widenhoefer, R. A. Org. Lett. 2000, 2, 1469.
(22) (a) Perch, N. S.; Widenhoefer, R. A. J. Am. Chem. Soc. 1999, 121,
6960. (b) Perch, N. S.; Pei, T.; Widenhoefer, R. A. J. Org. Chem. 2000,
65, 3836.
(23) (a) Rix, F. C.; Brookhart, M. J. Am. Chem. Soc. 1995, 117, 1137.
(b) Rix, F. C.; Brookhart, M.; White, P. S. J. Am. Chem. Soc. 1996, 118,
2436.
(24) Brookhart, M.; Wagner, M. I.; Balavoine, G. G. A.; Haddou, H. A.
J. Am. Chem. Soc. 1994, 116, 3641.
(27) The catalyst generated from (η3-C3H5)Pd(Me)PCy3 and HBAr4‚OEt2
must differ slightly from that generated from 1a and NaBAr4, because no
ether is present in NaBAr4 used in the latter protocol. In this latter case,
the vacant coordination site on palladium presumably is occupied by some
other weakly coordinated ligand such as solvent, water, or silane.
(25) LaPointe, A. M.; Rix, F. C.; Brookhart, M. J. Am. Chem. Soc. 1997,
119, 906.
(26) DiRenzo, G. M.; White, P. S.; Brookhart, M. J. Am. Chem. Soc.
1996, 118, 6225.

Cycloisomerization of Functionalized 1,6-Dienes
J. Am. Chem. Soc., Vol. 122, No. 41, 2000 10019
Table 2. Synthesis of π-Allyl Palladium Chloride Precatalysts
1a-1o
Table 3. Palladium-Catalyzed (5 mol %) Cycloisomerization of 2a
in the Presence of HSiEt3 (1.2-1.5 equiv) at 25 °C in CH2Cl2 as a
Function of Palladium Precatalyst
R1
R2
L
precatalyst
yield (%)
entry
precatalyst
time
yield (%)a
isomer ratiob
H
H
H
H
Me
H
P(Cy)3
1a
1b
1c
1d
1e
1f
1g
1h
1i
1j
1k
1l
1m
1n
1o
86a
79
70
70
72
88
84a
90b
83
95
81
78
85
78
88
1
2
3
4
5
6
7
8
9
10
11
12
13
14
1b
1c
1d
1e
1f
1g
1h
1i
1j
1k
1l
1m
1n
1o
20 min
15 min
50 min
10 min
10 min
1 h
12 h
30 min
12 h
1.5 h
12 h
12 h
12 h
12 h
94
91
95
98
88
74
86
91
87
86
NR
NR
NR
NR
26:1
44:1
31:1
>50:1
>50:1
2.6:1
0.5:1
2.1:1
0.6:1
1.1:1
-
Ph
Me
H
H
P(cyclopentyl)3
P(i-Pr)3
P(n-Bu)3
PPh3
P(o-tolyl)3
P(OPh)3
P(Oi-Pr)3
P(t-Bu)3
pyridine
P(t-Bu)2(2-C6H4Ph)
P(Cy)2(2-C6H4Ph)
-
-
-
a Yield refers to isolated material of >95% purity. b Determined by
capillary GC, refers to the ratio of 3a to the sum of all other isomers.
a Previously reported in ref 26. b Previously reported in ref 48.
crude reaction mixtures of the cycloisomerization of 2a using
catalysts 1a-1d revealed no detectable formation of byproducts
other than small amounts of hexaethyldisiloxane. For example,
neither allyl triethylsilane nor any higher molecular weight
organosilanes were detected in the isomerization of 2a in the
presence of HSiEt3 catalyzed by 1a/NaBAr4 (10 mol %).
Similarly, neither R-methylstyrene nor any hydrogenation,
hydrosilylation, or dimerization products of R-methylstyrene
were detected in the reaction of 2a with HSiEt3 catalyzed by
1b/NaBAr4 (10 mol %).
effective promoters for the cycloisomerization of 2a (Table 1,
entries 2 and 3), whereas dimethyl-tert-butylsilane led to a
dramatic drop in reaction rate although yield and selectivity
remained high (Table 1, entry 4).
Effect of Precatalyst Structure on Cycloisomerization. The
efficiency and selectivity of transition metal-catalyzed processes
can be strongly affected by the nature of the ancillary ligands
on the metal complex. Because of this, the efficiency of
palladium-catalyzed cycloisomerization was probed as a function
of the ligands on the π-allyl palladium precatalyst. The requisite
π-allyl palladium complexes 1a - 1o were prepared by reaction
of the π-allyl palladium chloride dimer with the appropriate
ligand in ether at room temperature (Table 2). Concentration
of solvent followed by filtration gave the desired complexes as
off-white or pale yellow solids,28 and new complexes were
characterized by spectroscopy and elemental analysis. 1H NMR
of all new phosphine-substituted complexes displayed five
resonances corresponding to the unique protons of the static
π-allyl ligand. In contrast, the 1H NMR spectra of the phosphite-
substituted complexes 1j and 1k displayed a quintet at δ 5.0 (J
) 10.2 Hz) corresponding to the central allylic hydrogen atom
and a broad resonance between δ 4.5 and 2.0 corresponding to
the four time-averaged terminal allylic protons. A process
involving π-σ-π interconversion of the allyl ligand coupled
with rotation about the C-C single bond of the σ-allyl
intermediate is sufficient to exchange all four terminal allylic
protons.29
In contrast to the relative insensitivity of cycloisomerization
with respect to the allyl ligand, the rate, yield, and selectivity
of palladium-catalyzed cycloisomerization was highly sensitive
to the nature of the phosphine ligand. Specifically, only
complexes that possessed a tri(secondary alkyl) phosphine such
as 1a [L ) PCy3], 1e [L ) P(cyclopentyl)3], and 1f [L ) P(i-
Pr)3] catalyzed the selective conversion of 2a to 3a (Table 3,
entries 4 and 5). In contrast, use of palladium precatalysts that
possessed a P(n-Bu)3 (1g), triarylphosphine (1h, 1i), or phosphite
(1j, 1k) ligand led to a precipitous drop in selectivity for 3a
although yields remained high (Table 3, entries 6-10). Pal-
ladium complexes that possessed a P(t-Bu)3 (1l), pyridine (1m),
or P(2-C6H4Ph)R2 [R ) t-Bu (1n), Cy (1o)] ligand failed to
catalyze the cycloisomerization of 2a (Table 3, entries 11-14).
Scope of Cycloisomerization. A range of functionalized 1,6-
dienes underwent cycloisomerization in the presence of HSiEt3
(1.2 equiv) and a catalytic 1:1 mixture of NaBAr4 and either
1a or 1e (5 mol %) to form 1,2-disubstituted cyclopentenes in
good yield with high selectivity (Table 4). For example, dienes
which possessed homoallylic ester, ketone, acetoxy, or pivaloyl
groups (2b-2d, and 7-13) underwent palladium-catalyzed
cycloisomerization within 20 min at room temperature to form
the corresponding 1,2-disubstituted cyclopentenes 3b-3d, and
14-20 in >70% yield with high selectivity (Table 4, entries
1-13). The sulfonyl-substituted dienes 21 and 22 also cyclized
to form 1,2-disubstituted cyclopentenes 23 and 24, albeit with
somewhat diminished selectivity (Table 4, entries 14 and 15).
Dienes that possessed only one homoallylic substituent did not
undergo efficient cycloisomerization, presumably because of the
absence of a kinetic Thorpe-Ingold effect.30 In addition, dienes
that did not possess at least one homoallylic oxygenated group
The rate, yield, and selectivity of cycloisomerization was
relatively insensitive to substitution on the palladium-bound allyl
group. For example, complexes that possessed an internal phenyl
(1b) or methyl group (1c) or a terminal methyl group (1d) on
the allyl ligand all served as effective precatalysts for the
cycloisomerization of 2a to form 3a in >90% yield with >95%
isomeric purity (Table 3, entries 1-3). GCMS analysis of the
(28) Complex 1g was isolated as a viscous, pale yellow oil and was
>95% pure as determined by NMR spectroscopy, but had been previously
reported as a white solid.26 This difference presumably stems from the
isomeric purity of the P(n-Bu)3 used in the respective syntheses of 1g.
(29) (a) Vrieze, K. In Dynamic Nuclear Magnetic Resonance Spectros-
copy; Jackman, L. M.; Cotton, F. A., Eds.; Academic: New York, 1975; p
441. (b) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. Principles
and Applications of Organotransition Metal Chemistry; University Science
Books: Mill Valley, CA, 1987; p 177.

10020 J. Am. Chem. Soc., Vol. 122, No. 41, 2000
Kisanga and Widenhoefer
Table 4. Cycloisomerization of 4,4-Disubstituted 1,6-Dienes
Catalyzed by a 1:1 Mixture of NaBAr4 and 1a or 1e (5 mol %) in
the Presence of HSiEt3 (1.2-1.5 equiv) at 25 °C
Table 5. Cycloisomerization of Allylic-Substituted 1,6-Dienes
Catalyzed by a 1:1 Mixture of NaBAr4 and 1a or 1e (5 mol %) in
the Presence of HSiEt3 (1.2-1.5 equiv) at 25 °C
a A ) 1a/NaBAr4 in CH2Cl2; B ) 1e/NaBAr4 in 1,2-dichloroethane.
b Yield refers to isolated material of >95% purity. c Determined by
capillary GC.
Table 6. Palladium-Catalyzed (5 mol %) Cycloisomerization of 31
in the Presence of HSiEt3 in CH2Cl2 as a Function of Precatalyst,
Temperature, and [HSiEt3]
temp
(°C)
[HSiEt3] yield
(M)
entry precatalyst
time
(%)a 32:33b
1
2
3
4
5
6
7
1a
1e
1f
1a
1a
1a
1a
25
25
25
25
70c
25
25
20 min
20 min
20 min
12 h
5 min
12 h
0.075
0.075
0.075
0.075
0.075
0.20
84
86
84
88
89
77
83
3:1
1:1
1:1
1:1
1:1
6:1
31:1
a A ) 1a/NaBAr4 in CH2Cl2; B ) 1e/NaBAr4 in 1,2-dichloroethane.
b Yield refers to isolated material of >95% purity. c Determined by
capillary GC.
12 h
0.40
failed to undergo efficient cycloisomerization, as observed in
palladium-catalyzed diene cyclization/hydrosilylation.21,22
a Yield refers to isolated material of >95% purity. b Determined by
capillary GC. c Solvent ) 1,2-dichloroethane.
Palladium-catalyzed diene cycloisomerization also tolerated
allylic substitution. For example, dienes which possessed an
allylic methyl (25a, 25b) or phenyl (26) group isomerized to
form the corresponding cyclopentenes (27-28) in good yield
and with good selectivity (Table 5, entries 1-3). Isomerization
of dienes 25 and 26 led to no detectable formation of the
isomeric cyclopentenes in which the double bond occupied the
R,â-position relative to the quaternary carbon atom, even though
these carbocycles would also possess a tetrasubstituted olefin.
Dienes which possessed disubstitution at an allylic carbon atom
such as the gem-dimethyl-substituted 29 also underwent efficient
palladium-catalyzed cycloisomerization forming cyclopentene
30 in 82% yield (Table 5, entry 4).
Under the conditions described above, palladium-catalyzed
cycloisomerization of dienes which possessed an internal olefin
gave poor selectivity. For example, treatment of 4,4-dicarbo-
methoxy-1,6-octadiene (31) with a 1:1 mixture of 1a and
NaBAr4 in the presence of HSiEt3 (1.5 equiv) led to the
formation of a 3:1 mixture of the expected cyclopentene 32 and
1,1-dicarbomethoxy-3-ethylidene-4-methylcyclopentane (33) in
84% combined yield (Table 6, entry 1). Use of precatalysts 1e
or 1f, or increasing the reaction time or temperature did not
lead to a significant increase in the selectivity for cyclopentene
32 (Table 6, entries 2-5). In contrast, the selectivity for
cyclopentene 32 increased with increasing silane concentration
(Table 6, entries 6 and 7). For example, in the presence of an
8-fold excess of silane, isomerization of 31 at room temperature
overnight led to the isolation of 32 in 83% yield with >96%
selectivity (Table 6, entry 7).
In contrast to the highly selective cycloisomerization of 1,6-
dienes to form cyclopentenes, palladium-catalyzed cycloisomer-
ization of 1,7-dienes did not lead to the selective formation of
cyclohexenes. For example, reaction of 4,4-dicarboethoxy-1,7-
octadiene (34) with triethylsilane and a 1:1 mixture of 1a/
NaBAr4 at room temperature overnight led to the isolation of
an 82:7:6:1 mixture of cyclopentene 32 and three unidentified
isomers in 82% combined yield (eq 5). Repeated chromatog-
raphy of the mixture led to the isolation of pure 32 in 46%
yield, which was identical with the sample of 32 isolated from
cycloisomerization of diene 31.31 The predominant formation
of 32 from 34 presumably results from initial isomerization to
31 followed by cyclization (see below).
(30) (a) DeTar, D. F.; Luthra, N. P. J. Am. Chem. Soc. 1980, 102, 4505.
(b) Kirby, A. J. AdV. Phys. Org. Chem. 1980, 17, 208. (c) Eliel, E. L.
Stereochemistry of Carbon Compounds; McGraw-Hill: New York, 1962;
p 106.
(31) Other substrates that failed to undergo effective palladium-catalyzed
cycloisomerization include terminally disubstituted and internally substituted
1,6-dienes, and 1,6-enynes and diynes.

Cycloisomerization of Functionalized 1,6-Dienes
J. Am. Chem. Soc., Vol. 122, No. 41, 2000 10021
Intermediates in the Conversion of 2 to 3. Transition metal-
catalyzed cycloisomerization of a 1,6-diene typically forms the
corresponding methylenecyclopentane as the primary product.11-19
Furthermore, the cycloisomerization of diene 31 in the presence
of 1.5 equiv of silane produced considerable amounts of
ethylenecyclopentane 33 (Table 6, entries 1-5). Therefore, it
appeared likely that methylene cyclopentanes were general inter-
mediates in the palladium-catalyzed cycloisomerization of 1,6-
dienes. In accord with this hypothesis, GC analysis of the
cyclization of 2a in the presence of HSiEt3 (1.2 equiv) and 1a/
NaBAr4 (5 mol %) at room temperature after 10 min revealed
that 84% of 2a had been consumed with the formation of a
19:1 mixture of 4a:3a (eq 6). After 20 min, 2a had been
completely consumed with the exclusive (>98%) formation of
3a.
Figure 1. Concentration versus time plot for the conversion of 2b
(0.05 M) to 3b, 4b, and 5b in the presence of HSiEt3 (0.075 M) and a
catalytic mixture of 1a/NaBAr4 (2.5 mM) in CH2Cl2 at 0 °C.
consistent with the relative rates of the isomerization of 2b to
4b and 4b to 3b under reaction conditions.
In contrast to methylenecyclopentane 4b, the isomeric cyclo-
pentene 3,3-dicarbomethoxy-1,5-dimethylcyclopentene (6b) was
neither observed in the conversion of 2b to 3b, nor did an
authentic sample of 6b isomerize to 3b under reaction conditions
(eq 7). In addition, both the conversion of 2b to 3b and 4b to
3b were accompanied by rapid darkening of the reaction mixture
upon consumption of 4b. The resulting dark solutions were
catalytically inactive and addition of either 2b or 4b to these
dark solutions led to no detectable formation of 3b.
To gain more detailed information concerning the formation
and consumption of intermediates in the conversion of 2 to 3,
the reaction of 2b (0.05 M), HSiEt3 (0.095 M), and a 1:1 mixture
of 1a and NaBAr4 (2.5 mM) at 0 °C was monitored periodically
by GC. A plot of [2b] versus time was linear to >3 half-lives
Deuterium Scrambling in the Conversion of 2 to 3. Our
attempts to probe the mechanism of diene cycloisomerization
by spectroscopic analysis of catalytic mixtures proved largely
unsuccessful, partly because of the large excess of silane
(relative to catalyst) required for efficient cycloisomerization,
which obscured key resonances. Therefore, we turned to
deuterium-labeling studies to gain information on the mechanism
of palladium-catalyzed cycloisomerization. In one experiment,
reaction of 2b (0.5 M), DSiEt3 (0.95 M), and a 1:1 mixture of
1a/NaBAr4 (2.5 mM) at 0 °C was monitored periodically by
GC. Little change was observed in the concentration of 2b early
in the reaction, and then after ∼30 min the rate of reaction
increased and followed zero-order decay to ∼90% conversion
with a zero-order rate constant of k ) 8.1 × 10-5 s-1 (t1/2
)
103 min) (Figure S1).32 After 150 min, 72% of 2b had reacted
without formation of detectable quantities of 3b. Rather, the
products consisted of a 12:1 mixture of 4b and 5b, which
accounted for >97% of the products. The relative concentration
of 4b reached a maximum of 80% after 225 min, and then
decreased rapidly to <3% after 255 min with the concomitant
formation of cyclopentene 3b (Figure 1). A plot of [4b] versus
time from 225 to 255 min (Figure S2) provided a lower limit
for the rate of disappearance of 4b of k g 5 × 10-4 s-1 (t1/2 e
17 min), which is g 6 times faster than the initial disappearance
of 2b. In comparison, the relative concentration of intermediate
5b rose to a maximum of ∼4% after ∼100 min, remained
relatively constant until approximately 240 min, and then
disappeared during the next 20 min (Figure 1).
The isomerization of methylenecyclopentane 4b to cyclo-
pentene 3b under reaction conditions occurred at a significant
rate only after the concentration of 2b had decreased to
approximately 10% (e 5 mM) of its original concentration,
which suggests that the isomerization of 4b was inhibited by
excess 2b. In accord with this hypothesis, treatment of an
authentic sample of 4b with triethylsilane (1.2 equiv) and a
catalytic mixture of 1e/NaBAr4 at room temperature in the
absence of 2b led to immediate darkening of the solution and
complete isomerization within 1 min with the isolation of 3b
in 93% yield as a 60:1 mixture of isomers (eq 7). This
(Figure S3). The zero-order rate constant (k ) 8.8 × 10-5 s-1
,
t1/2 ) 95 min) obtained from the linear portion of the slope
was not significantly (<10%) different from the rate obtained
in the presence of HSiEt3.
A preparative scale isomerization of 2b in the presence of
DSiEt3 and 1a/NaBAr4 led to the isolation of carbocycle 3b-d
2
in 92% yield (eq 8). Analysis of 3b-d by 13C and H NMR
spectroscopy was consistent with deuterium incorporation
exclusively into the exocyclic methyl groups with the formation
of multiple isotopomers. For example, the 13C NMR spectrum
displayed a singlet at δ 13.5 and a 1:1:1 triplet at δ 13.2 (J )
19 Hz, isotopic shift ) 270 ppb), corresponding to exocyclic
CH3 and CH2D groups, along with several resonances upfield
2
of the 1:1:1 triplet. Similarly, the H NMR spectrum of 3b-d
displayed a ∼1:3:2 ratio of resonances at δ ∼1.64, 1.61, and
1.58 (isotopic shift ≈ 30 ppb). Mass spectral analysis of 3b-d
established a 66:25:7:2 ratio of d0, d1, d2, and d3 isotopomers
with an average of 0.45 deuterium atoms per molecule (eq 8).
(32) Additional silane had little effect on the reaction rate. For example,
a 4-fold increase in silane concentration ([HSiEt3] ) 0.38 M) in the
isomerization of 2 at 0 °C led to <15% increase in the reaction rate (k )
9.2 × 10-5 s-1).
isomerization of 4b to 3b is ∼10 times faster than conversion
of 2b to 3b under comparable conditions (Table 3, entry 4),

10022 J. Am. Chem. Soc., Vol. 122, No. 41, 2000
Kisanga and Widenhoefer
Scheme 3
Several additional labeling experiments were performed with
use of deuterated dienes and HSiEt3. For example, cycloisomer-
ization of 4,4-dicarbomethoxy-2,6-dideuterio-1,6-heptadiene
(2a-2,6-d2) in the presence of HSiEt3 led to isolation of 4,4-
dicarbomethoxy-1,2-dimethylcyclopentene (3a-d) in 92% yield
as a 23:38:28:11 mixture of d0-d3 isotopomers (1.28 D/mole-
cule). Similarly, palladium-catalyzed isomerization of 4,4-
dicarbomethoxy-1,1,7,7-tetradeutero-1,6-heptadiene (2a-1,1,7,7-
d4) in the presence of HSiEt3 led to the isolation of 3a-d in
82% yield as a 3:17:27:30:15:2 mixture of d1-d6 isotopomers
(3.25 D/molecule).
significant H/D exchange in both the unreacted diene and in
the methylenecyclopentane 4a-d (Scheme 3).
Deuterium Scrambling in the Conversion of 4 to 3. The
isomerization of methylenecyclopentane 4 to cyclopentene 3
was also accompanied by isotopic scrambling. For example,
treatment of 4b with DSiEt3 (1.2 equiv) and 1a/NaBAr4 (5 mol
%) at room temperature for 10 min led to the isolation of 3b as
a 72:23:4:1 mixture of d0-d3 isotopomers which contained an
average of 0.34 deuterium atoms per molecule of 3b (eq 9). A
second experiment established that all H/D exchange occurred
prior to formation of 3. Specifically, a 1:1 mixture of diene 2b
and cyclopentene 3a were reacted with DSiEt3 (1.2 equiv) and
a 1:1 mixture of 1a/NaBAr4 (5 mol %) for 20 min at room
temperature to form a 1:1 mixture of 3a and 3b. GCMS analysis
of the reaction mixture after 40 min (approximately twice the
time required for complete conversion of 2b to 3b) revealed
no significant (<2%) deuterium incorporation into recovered
3a.
Deuterium Scrambling in the Conversion of 2 to 4. The
H/D exchange detected in the conversion of 2 to 3 represented
the cumulative scrambling that had occurred in all the steps
leading up to the formation of 3. Therefore, H/D exchange was
studied in each step independently. In one experiment, cycliza-
tion of a 1:1 mixture of 2a-2,6-d2 and 2b in the presence of
HSiEt3 (1.2 equiv) and 1a/NaBAr4 (5 mol %) at 0 °C was
monitored periodically by GCMS analysis (Table 7). After 13%
conversion (30 min), unreacted dienes 2a-2,6-d2 and 2b and
carbocycle 4b had undergone ∼15% H/D exchange, consisting
of 83% 2a-d2, 86% 2b-d0, and 91% 4b-d0, respectively (Table
7). In contrast, carbocycle 4a had lost a considerable amount
of deuterium and consisted predominantly (84%) of the 4a-d1
isotopomer (Table 7). GCMS analysis of the reaction at 70%
conversion revealed considerable isotopic scrambling of all
species in the reaction mixture (Table 7).
Discussion
Possible Mechanisms for the Conversion of 2 to 4.
Mechanisms involving both intramolecular carbometalation and
reductive cyclization have been postulated for the cycloisomer-
ization of enynes and dienes.1-19 These two mechanisms were
also considered in the conversion of 2 to 4. For example,
coordination of an olefin of 2 to a palladium hydride species
(LnPd-H) would generate palladium hydride olefin intermediate
I, which could undergo â-migratory insertion to form palladium
alkyl olefin intermediate II (Scheme 4, right cycle). Coordination
and insertion of the pendant olefin into the Pd-C bond would
generate palladium cyclopentylmethyl intermediate III, which
could undergo â-hydride elimination to form 4 and regenerate
the palladium hydride species. In the second mechanism,
coordination of diene 2 to a palladium complex (LnPd) followed
by reductive cyclization would form the palladacyclopentene
intermediate IV (Scheme 4, left cycle). â-Hydride elimination
from IV would form the palladium alkyl hydride complex V,
which could undergo reductive elimination to form 4.
In a second experiment, GCMS analysis of the isomerization
of the tetradeuteride 2a-1,1,7,7-d4 in the presence of HSiEt3 (1.2
equiv) and 1a/NaBAr4 (5 mol %) at 40% conversion revealed
Table 7. Isotopic Composition of 2a, 2b, 4a, and 4b Formed in
the Partial Conversion of a 1:1 Mixture of 2a-2,6-d2 and 2b to 4a
and 4b in the Presence of HSiEt3 (1.2 equiv) and a 1:1 Mixture of
1a and NaBAr4 (5 mol %) in CH2Cl2 at 0 °C
The considerable H/D exchange observed in the cyclization
of 2 to 4 is consistent with the intermediacy of palladium hydride
complexes that undergo facile H/D exchange with free silane.
Because both the carbometalation and oxidative cyclization
mechanisms invoke palladium-hydride intermediates, this ob-
servation does not distinguish between the two pathways.
However, the isotopic composition of the dienes recovered from
reaction of 2a-2,6-d2/2b with HSiEt3 is not in accord with H/D
exchange via the oxidative cyclization mechanism. For example,

Cycloisomerization of Functionalized 1,6-Dienes
J. Am. Chem. Soc., Vol. 122, No. 41, 2000 10023
Scheme 4
Scheme 5
Scheme 6
even if all the steps in the oxidative cyclization mechanism
preceding reductive elimination were reversible,33 and H/D
exchange of intermediate V with silane were fast, only the two
internal olefinic H/D atoms of the diene would undergo
exchange (Scheme 4). Therefore, the recovery of significant
amounts of 2a-d3 (9%) and 2b-d3 (4%) from the reaction of
2a-2,6-d2/2b with HSiEt3 at 70% conversion (Table 7) and the
significant scrambling of the terminal olefinic deuterium atoms
of 2a-1,1,7,7-d4 recovered from the reaction with HSiEt3
(Scheme 3) are not in accord with H/D exchange via the
oxidative cyclization mechanism.34
The isotopic composition of the dienes recovered from
reaction of 2a-2,6-d2/2b or 2a-1,1,7,7-d4 with HSiEt3 is con-
sistent with exchange via an addition/elimination pathway. For
example, reversible 1,2-hydrometalation of 2a-2,6-d2 via the
palladium (primary alkyl) intermediate VI-d2 would lead to H/D
exchange of the internal olefinic positions with formation of
2a-d1 and LnPd-D (Scheme 5, path a). Reversible 2,1-deute-
riometalation of 2a-2,6-d2 with LnPd-D released in the trans-
formation above could lead to formation of 2a-d3 via palladium
(secondary alkyl) intermediate II-d3 (Scheme 5, path b).35 An
analogous series of insertion/elimination reactions would also
exchange the olefinic H/D atoms of diene 2a-1,1,7,7-d4.
Unfortunately, the addition/elimination process that leads to
exchange of the olefinic H/D atoms of 2 is not necessarily
related to the process that converts 2 to 4, as the two events
could occur in separate reaction manifolds.36 Nevertheless, these
experiments support the carbometalation pathway by providing
evidence for palladium alkyl complex II, a key intermediate in
the carbometalation mechanism.
An H/D exchange process that is clearly tied to the cycloi-
somerization of 2 to 4 is the significant loss of deuterium in
the conversion of 2a-2,6-d2 to 4a in the presence of HSiEt3 early
in the reaction (Table 7). The carbometalation mechanism
accounts for this behavior provided that H/D exchange between
palladium deuteride intermediates and silane early in the reaction
is fast and essentially complete. For example, successive
hydrometalation and carbometalation of 2a-2,6-d2 would form
palladium alkyl intermediate III-d2, which could undergo
â-deuteride elimination to form 4a-d1 and LnPd-D (Scheme 6).
The palladium-deuteride intermediate could then undergo H/D
(33) Grubbs, R. H.; Miyashita, A. J. Organomet. Chem. 1978, 161, 371.
(b) Grubbs, R. H.; Miyashita, A. J. Am. Chem. Soc. 1978, 100, 1301. (c)
Grubbs, R. H.; Miyashita, A.; Liu, M.; Burk, P. J. Am. Chem. Soc. 1978,
100, 2418. (d) McLain, S. J.; Wood, C. D.; Schrock, R. R. J. Am. Chem.
Soc. 1979, 101, 4558. (e) Knight, K. S.; Wang, D.; Waymouth, R. M.;
Ziller, J. J. Am. Chem. Soc. 1994, 116, 1845. (f) Taber, D. F.; Louey, J. P.;
Lim, J. A. Tetrahedron Lett. 1993, 34, 2243.
(34) For the observed scrambling to result from the reductive cyclization
mechanism, a reversible ring contraction of palladium alkyl hydride inter-
mediate V involving a palladacyclobutane intermediate would be required.
For examples of ring contraction from metallacyclopentanes see: (a) Yang,
G. K.; Bergman, R. G. Organometallics 1985, 4, 129. (b) Smith, G.; Milan,
S. J.; Schrock, R. R. J. Organomet. Chem. 1980, 202, 269. (c) Schrock, R.
R.; McLain, S.; Sanch, J. Pure Appl. Chem. 1980, 52, 729 (d) McLain, S.
J.; Sancho, J.; Schrock, R. R. J. Am. Chem. Soc. 1980, 102, 5610.
(35) Both 1,2- and 2,1- insertion of R-olefins into the Pd-C bonds of
cationic palladium(II) alkyl complexes have been documented with the
regioselectivity determined by the steric and electronic nature of the
olefin: Ittel, S. D.; Johnson, L, K.; Brookhart, M. Chem. ReV. 2000, 100,
1169.
(36) For example, the palladium alkyl hydride intermediate V or a
palladium-hydride impurity could be responsible for the observed H/D
exchange. For such an example see: Thorn, M. G.; Hill, J. E.; Waratuke,
S. A.; Johnson, E. S.; Fanwick, P. E.; Rothwell, I. P. J. Am. Chem. Soc.
1997, 119, 8630.

10024 J. Am. Chem. Soc., Vol. 122, No. 41, 2000
Kisanga and Widenhoefer
Scheme 7
kinetics). This kinetic behavior also appears consistent with the
failure of methylenecyclopentane 4b to isomerize to 3b in the
presence of significant concentrations (g5 mM) of 2b. Specif-
ically, a monosubstituted olefin of diene 2b should coordinate
more strongly with palladium than does the 1,1-disubstituted
olefin of 4b.42 Therefore, when the concentration of 2b is
sufficiently high (g5 mM), the bulk of the catalyst is tied up
as a catalyst-substrate adduct and unavailable for the isomer-
ization of 4b to 3b. Possible candidates for the catalytic resting
state in the carbometalation pathway are the palladium alkyl
olefin intermediates II or III, either of which could be further
stabilized by coordination of a pendant carbonyl group (Scheme
4).35
The dramatic increase in the rate of diene cycloisomerization
upon addition of silane points to activation of the precatalyst
by silane, perhaps by hydride donation to palladium. For
example, the induction period observed in the cyclization of
2b in the presence of DSiEt3 is consistent with H(D)-Si bond
cleavage during catalyst activation, and deuterium-labeling
experiments established the intermediacy of Pd-H(D) com-
plexes. In addition, organosilanes, including triethylsilane, have
been employed as hydride donors in the palladium-catalyzed
reductive cyclization of enynes and diynes.43 However, if
palladium hydride complexes are formed by hydride donation
from silane, the mechanism of this transfer remains unclear.
For example, the direct reaction of a silane with a cationic
palladium π-allyl complex to form a palladium-hydride complex
seems unlikely.44,45 Furthermore, no direct or indirect evidence
for the formation of palladium hydride complexes was obtained
from spectroscopic analysis of mixtures of 1a/NaBAr4, HSiEt3,
and 2a, or from product analysis of the crude reaction mixtures.
Several experimental observations suggest that excess silane
also serves to stabilize the active palladium catalyst, which
facilitates the secondary isomerization of alkylidene cyclopen-
tane to cyclopentene. For example, poor selectivity in the
cycloisomerization of 2 was observed when <1 equiv of silane
(relative to diene) was used. Similarly, the selectivity for the
formation of cyclopentene 32 from the cycloisomerization of
31 increased with increasing silane concentration, but not with
increasing reaction time or temperature (Table 6). Triethylsilane
has been shown to bind to the cationic palladium (II) fragment
[(phen)Pd(SiEt3)]+ and is readily displaced by an olefin.25 In a
similar manner, excess silane could perhaps stabilize the
palladium-hydride catalyst responsible for secondary isomer-
ization by serving as a weakly bonded ligand.
exchange with free silane to regenerate LnPd-H, which would
then attack a second molecule of 2a-2,6-d2 (Scheme 6).37,38 The
overall process would result in the loss of one deuterium atom
per 2a-2,6-d2 f 4a-d1 cycle. However, the likely conditions of
incomplete H/D exchange of the Pd-D intermediate with HSiEt3
or exchange of the olefinic H/D atoms of the diene prior to
cyclization would result in the loss of <1 deuterium atom in
the conversion of 2a-2,6-d2 to 4a, as was observed experimen-
tally (Table 7).
Possible Mechanisms for the Conversion of 4 to 3. Olefin
isomerization catalyzed by soluble transition metal complexes
has been studied extensively,39 and mechanisms involving
addition/elimination40 or 1,3-hydrogen migration have been
established.41 In a similar manner, the isomerization of 4 to 3
could occur via initial hydrometalation of the exocyclic meth-
ylene group of 4 to the palladium cyclopentyl intermediate VI
followed by â-elimination of the tertiary hydride to form 3
(Scheme 7). Alternatively, oxidative addition of tertiary allylic
hydrogen of 4 could form the palladium π-allyl hydride
intermediate VII followed by reductive elimination to form 3
and regenerate LnPd (Scheme 7). Either of these pathways could
account for the scrambling observed in the conversion of 4b to
3b in the presence of DSiEt3. However, because H/D exchange
in the conversion of 2 to 4 occurred via an addition/elimination
pathway, it is also likely that an addition/elimination pathway
is responsible for H/D exchange and isomerization of 4 to 3.
Kinetics and Role of Silane. The zero-order disappearance
of 2b under reaction conditions is consistent with the formation
of a catalyst-substrate adduct which undergoes turnover-
limiting unimolecular reaction to form product 4b (saturation
Conclusions
(37) Under the reaction conditions, LnPd-H would also be produced
from the concurrent cyclization of 2b.
A 1:1 mixture of (η3-C3H5)Pd(Cl)PR3 [R ) Cy (1a),
cyclopentyl (1e), or i-Pr (1f)] and NaB[3,5-C6H3(CF3)2]4 in the
presence of a stoichiometric amount of HSiEt3 catalyzed the
cycloisomerization of functionalized 1,6-dienes to form 1,2-
disubstituted cyclopentenes in good yield with high selectivity.
Effective cycloisomerization required both a tri(secondary alkyl)
(38) The loss of deuterium in the conversion of 2a-2, 6-d2 to 4a can
also be accounted for by the reductive cyclization mechanism. For example,
â-deuteride elimination from palladacyclopentane IV-d2 would form the
palladium deuteride V-d2. H/D exchange of the deuteride ligand of V-d2
with free silane could generate V-d1, and subsequent reductive elimination
would form 4-d1.
(39) Parshall, G. W.; Ittel, S. D. Homogeneous Catalysis, 2nd ed.; John
Wiley & Sons: New York, 1992; pp 10-23.
(42) Hegedus, L. S. Transition Metals in the Synthesis of Complex Or-
ganic Molecules; University Science Books: Mill Valley, CA, 1994; p 50.
(43) (a) Trost, B. M.; Rise, F. J. Am. Chem. Soc. 1987, 109, 3161. (b)
Trost, B. M.; Lee, D. C. J. Am. Chem. Soc. 1988, 110, 7255.
(44) A palladium-hydride complex could be generated through metathesis
of the H-Si bond of the silane with the Pd-C σ-bond of the allyl ligand
or by oxidative addition of the silane H-Si bond. However, metathesis of
a silane with an electrophilic late-transition metal alkyl complex typically
forms a M-Si complex,25,45 whereas it seems unlikely that a palladium (IV)
alkyl hydride complex would be stable with respect to C-H reductive
elimination.
(40) Orchin, M. AdV. Catal. 1966, 16, 1. (b) Andrieux, J.; Barton, D. H.
R.; Patin, J. J. Chem. Soc., Perkin Trans. 1 1977, 359. (c) Tolman, C. A.
J. Am. Chem. Soc. 1972, 94, 2994. (d) Tolman, C. A.; McKinney, R. J.;
Seidel, W. C.; Druliner, J. D.; Stevens, W. R. AdV. Catal. 1985, 33, 1. (e)
Bingham, D.; Webster, D. E.; Wells, P. B. J. Chem. Soc., Dalton Trans.
1974, 1514. (f) Bingham, D.; Webster, D. E.; Wells, P. B. J. Chem. Soc.,
Dalton Trans. 1974, 1519. (f) Cramer, R. J. Am. Chem. Soc. 1966, 88,
2272. (g) Cramer, R. Acc. Chem. Res. 1968, 1, 186.
(41) (a) Bingham, D.; Hudson, B.; Webster, D. E.; Wells, P. B. J. Chem.
Soc., Dalton Trans. 1974, 1521. (b) Tuner, M.; Jouanne, J. V.; Brauer,
H.-D.; Kelm, H. J. Mol. Catal. 1979, 5, 425. (c) Tuner, M.; Jouanne, J. V.;
Brauer, H.-D.; Kelm, H. J. Mol. Catal. 1979, 5, 433. (d) Tuner, M.; Jouanne,
J. V.; Brauer, H.-D.; Kelm, H. J. Mol. Catal. 1979, 5, 447.
(45) (a) Burger, P.; Bergman, R. G. J. Am. Chem. Soc. 1993, 115, 10462.
(b) Brookhart, M.; Grant, B. E. J. Am. Chem. Soc. 1993, 115, 2151. (c)
Marciniec, B.; Pietraszuk, C. Organometallics 1997, 16, 4320.

Cycloisomerization of Functionalized 1,6-Dienes
J. Am. Chem. Soc., Vol. 122, No. 41, 2000 10025
PCy3,46 1a,47 1g,48 and 1h,48 were prepared by a modified literature
procedure (see Supporting Information).26,46,47 NaBAr4 and HBAr4‚OEt2
[Ar ) 3,5-C6H3(CF3)2]4 were prepared using known procedures.48
Authentic samples of methylenecyclopentanes 4a, 4b, and 33 were
synthesized by the method of Urabe et al.49
phosphine ligand on the palladium precatalyst and a stoichio-
metric amount of silane. The procedure tolerated a range of
functionality including esters, ketones, sulfones, and protected
alcohols. However, dienes which possessed only one homoal-
lylic substituent or did not possess at least one homoallylic
oxygenated group failed to undergo efficient cycloisomerization.
Cycloisomerization also tolerated substitution at the allylic and
terminal olefinic position of the diene, although a 6-fold excess
of silane was required for good selectivity in the latter case.
Cyclopentenes. 4,4-Dicarbomethoxy-1,2-dimethylcyclopentene
(3a). Dimethyl diallylmalonate (2a) (100 mg, 0.47 mmol) and HSiEt3
(80 mg, 0.70 mmol) were added sequentially via syringe to a solution
of 1a (10 mg, 0.022 mmol) and NaBAr4 (24 mg, 0.024 mmol) in CH2-
Cl2 (8 mL) at 0 °C. The resulting yellow solution was stirred at room
temperature for 20 min to form a dark brown solution. Solvent and
silane were evaporated under vacuum, and the oily residue was
chromatographed (hexane:EtOAc, 12:1) to give 3a (89 mg, 89%) as a
Palladium-catalyzed cycloisomerization of diethyl diallyl-
malonate (2b) obeyed zero-order kinetics to >3 half-lives with
initial formation of 1,1-dicarboethoxy-4-methyl-3-methylene-
cyclopentane (4b) along with traces (e4%) of 4,4-dicarbo-
ethoxy-1,5-heptadiene (5b). The zero-order disappearance of 2b
was indicative of the rapid formation of a substrate-catalyst
adduct which underwent turnover-limiting intramolecular rear-
rangement to form 4b. The relative concentration of 4b increased
to ∼80% of the reaction mixture and then underwent secondary
isomerization g6 times faster than the initial disappearance of
2b to form 4,4-dicarbomethoxy-1,2-dimethylcyclopentene (3b).
The unusual accumulation/consumption behavior of 4b appeared
to result from both the saturation kinetics and the greater
coordinating ability of 2b relative to 4b. Specifically, in the
presence of significant concentration of 2b (g5 mM), rapid and
quantitative formation of the catalyst-substrate adduct rendered
the active catalyst unavailable for isomerization of 4b to 3b.
1
colorless oil. H NMR: δ 3.69 (s, 6 H), 2.92 (s, 4 H), 1.56 (s, 6 H).
13C{1H} NMR: δ 172.9, 127.9, 57.0, 45.8, 13.1. IR (neat, cm-1): 1731
(CdO). Anal. Calcd (found) for C11H16O4: H, 7.60 (7.33); C, 62.26
(61.99).
The procedure given for the synthesis of 3a was applied to the
synthesis of all cyclopentenes found in Table 4 unless otherwise stated.
1
1,2-Dimethyl-4-phenyl-4-benzyloxycyclopentene (14). H NMR:
δ 7.60-7.09 (m, 10 H), 3.30 (d, J ) 15.2 Hz, 2 H), 2.70 (d, J ) 15.2
Hz, 2 H), 1.60 (s, 6 H). 13C{1H} NMR: δ 201.9, 146.7, 136.2, 132.0,
130.3, 129.1, 128.4, 128.2, 126.5, 125.6, 60.4, 49.9, 13.7. IR (neat,
cm-1): 1669 (CdO). Anal. Calcd (found) for C20H20O: H, 7.30 (7.32);
C, 86.91 (86.62).
4-Acetoxymethyl-1,2-dimethyl-4-phenylcyclopentene (15). 1H
NMR: δ 7.26 (m, 5 H), 4.10 (m, 2 H), 2.72 (d, J ) 14.9 Hz, 2 H),
7.25 (d, J ) 14.6 Hz, 2 H), 1.94 (s, 3 H), 1.62 (s, 6 H). 13C{1H} NMR:
δ 170.5, 146.4, 128.4, 127.4, 126.2, 125.3, 71.0, 47.2, 47.1, 20.2, 13.1.
IR (neat, cm-1): 1748 (CdO). Anal. Calcd (found) for C16H20O2: H,
8.25 (8.12); C, 78.65 (78.43).
4-Trimethylacetoxymethyl-1,2-dimethyl-4-phenylcyclopentene (16).
1H NMR: δ 7.25 (m, 5 H), 4.04 (s, 2 H), 2.74 (d, J ) 14.8 Hz, 2 H),
2.55 (d, J ) 14.6 Hz, 2 H), 1.62 (s, 6 H), 1.07 (s, 9 H). 13C{1H} NMR:
δ 178.7, 147.4, 129.3, 128.2, 127.2, 126.1, 72.3, 48.2, 48.1, 39.1, 27.3,
14.0. IR (neat, cm-1): 1729 (CdO). Anal. Calcd (found) for
C19H26O2: H, 9.15 (8.91); C, 79.68 (79.42).
4,4-Bis(acetoxymethyl)-1,2-dimethylcyclopentene (17). 1H NMR:
δ 3.99 (s, 2 H), 2.15 (s, 2 H), 2.04 (s, 3 H), 1.55 (s, 3 H). 13C{1H}
NMR: δ 170.5, 127.9, 66.7, 43.7, 42.0, 20.2, 12.9. IR (neat, cm-1):
1741 (CdO). Anal. Calcd (found) for C13H20O4: H, 8.39 (8.67); C,
64.98 (64.96).
4,4-Bis(trimethylacetoxymethyl)-1,2-dimethylcyclopentene (18).
1H NMR: δ 3.97 (s, 4 H), 2.17 (s, 4 H), 1.55 (s, 4 H), 1.17 (s, 18 H).
13C{1H} NMR: δ 178.5, 128.8, 67.7, 44.9, 43.3, 39.1, 27.4, 13.8. IR
(neat, cm-1): 1730 (CdO). Anal. Calcd (found) for C19H32O4: H, 9.95
(9.84); C, 70.32 (70.35).
4-Carbomethoxy-1,2-dimethyl-4-phenylcyclopentene (19). 1H
NMR: δ 7.27 (m, 5 H), 3.62 (s, 3 H), 3.28 (dd, J ) 0.7, 14.3 Hz, 2
H), 2.70 (dd, J ) 0.8, 15.3 Hz, 2 H), 1.63 (s, 6 H). 13C{1H} NMR: δ
177.1, 155.6, 129.4, 128.5, 126.8, 126.6, 56.7, 52.6, 48.4, 13.9. IR (neat,
cm-1): 1730 (CdO). Anal. Calcd (found) for C15H18O2: H, 7.88 (8.09);
C, 78.22 (78.07).
4-Acetyl-4-carbomethoxy-1,2-dimethylcyclopentene (20). 1H
NMR: δ 3.62 (s, 3 H), 2.83 (m, 4 H), 2.13 (s, 3 H), 1.56 (s, 6 H).
13C{1H} NMR: δ 202.4, 173.1, 127.3, 62.9, 51.9, 43.6, 25.1, 12.6. IR
(neat, cm-1): 1738 (CdO). Anal. Calcd (found) for C11H16O3: H, 8.22
(8.00); C, 67.32 (67.34).
4-Carbomethoxy-4-methylsulfonyl-1,2-dimethylcyclopentene (23).
10:1 Mixture of diastereomers. 1H NMR: δ 3.81 (s, 3 H), 3.08 (s, 4 H),
3.00 (s, 3 H), 1.56 (s, 6 H). 13C{1H} NMR: δ 170.3, 128.3, 74.5, 53.9,
43.1, 38.0, 13.5. IR (neat, cm-1): 1748 (CdO), 1306, 1121 (SdO).
Anal. Calcd (found) for C10H16O4S: H, 6.94 (7.00); C, 51.71 (51.16).
1,2-Dimethyl-4-phenyl-4-phenylsulfonylcyclopentene (24). 1H
NMR: δ 7.26-7.54 (aromatic region, 10 H), 3.22 (ABq, J ) 16.4 Hz,
Deuterium-labeling experiments involving 2b with DSiEt3 or
deuterated dienes 2a-2,6-d2 or 2a-1,1,7,7-d2 with HSiEt3
revealed that conversion of 2 to 4 led to considerable exchange
of the olefinic H/D atoms of recovered diene 2 as well as the
exocyclic H/D atoms of carbocycle 4. These data were consistent
with H/D exchange via an addition/elimination pathway coupled
with rapid H/D exchange of the Pd-H(D) intermediates with
free silane. Although H/D exchange did not necessarily occur
within the same reaction manifold as the conversion of 2 to 4,
these experiments provided support for the palladium alkyl
complex II, a key intermediate in the carbometalation pathway.
Furthermore, the carbometalation mechanism was consistent
with all of our observations concerning isotopic exchange in
the conversion of 2 to 4 and 4 to 3.
Silane served dual roles in the palladium-catalyzed cyclo-
isomerization of 1,6-dienes. First, silane activated the π-allyl
palladium precatalyst, perhaps via hydride donation, to generate
an active species which catalyzed the cycloisomerization of the
diene to the alkylidene cyclopentane. Second, silane stabilized
the active palladium hydride catalyst which facilitated the
secondary isomerization of the alkylidene cyclopentane to the
cyclopentene.
Experimental Section
General Methods. All reactions were performed under an atmo-
sphere of nitrogen using standard Schlenk techniques. NMR were
1
obtained at 300 MHz for H and 75 MHz for 13C in CDCl3 unless
otherwise noted. Gas chromatography was performed on a Hewlett-
Packard 5890 gas chromatograph equipped with a 25 m poly-
(dimethylsiloxane) capillary column. Flash chromatography was per-
formed employing 200-400 mesh silica gel (EM). Elemental analyses
were performed by E+R Microanalytical Laboratories (Parsippany, NJ).
Methylene chloride and 1,2-dichloroethane (DCE) were distilled from
CaH2 under nitrogen. Dimethyl and diethyl diallylmalonate (Lancaster)
and silanes (Aldrich) were used as received. Deuterated dienes 2a-2,6-
d2, 2a-1,1,7,7-d4, and 2a-3,3,5,5-d4 were synthesized by standard
procedures (see Supporting Information) and were >95% isotopically
pure as determined by NMR and GC analysis. Palladium π-allyl
chloride complexes including the known complexes (η3-C3H5)Pd(Me)-
(46) Hayashi, Y.; Matsumoto, K.; Nakamura, Y.; Isobe, K. J. Chem.
Soc., Dalton Trans. 1989, 1519.
(47) Powell, J.; Shaw, B. L. Inorg. Phys. Theor. 1967, 1839.
(48) Brookhart, M.; Grant, B.; Volpe, A. F. Organometallics 1992, 11,
3920.
(49) Urabe, H.; Hata, T.; Sato, F. Tetrahedron Lett. 1995, 36, 4261.

10026 J. Am. Chem. Soc., Vol. 122, No. 41, 2000
Kisanga and Widenhoefer
4 H), 1.40 (s, 6 H). 13C{1H} NMR: δ 137.5, 136.7, 133.4, 130.4, 129.9,
2129.3, 128.5, 128.4, 128.0, 75.8, 46.4, 13.3. IR (neat, cm-1): 1295,
1137, (SdO). HRMS Calcd (found) for C19H20O2S (M+): 312.1184
(312.1186).
13C{1H} NMR: δ 173.2, 134.1, 127.5, 57.4, 52.9, 46.1, 43.4, 21.2,
13.3, 12.6. IR (neat, cm-1): 1737 (CdO).
For 33. H NMR: δ 5.16 (q, J ) 4.4 Hz, 1 H), 3.68 (s, 6 H), 2.95
(d, J ) 17.2 Hz, 1 H), 2.80 (qd, J ) 2.0, 17.6 Hz, 1 H), 2.46 (m, 2 H),
1.65 (m, 1 H), 1.55 (m, 3 H), 1.01 (d, J ) 6 Hz, 3 H). 13C{1H} NMR:
δ 172.7, 144.0, 115.3, 58.4, 52.8, 42.6, 37.3, 18.1, 14.6. IR (neat, cm-1):
1734 (CdO).
1
1
4,4-Dicarboethoxy-1,2,3-trimethylcyclopentene (27b). H NMR:
δ 4.18 (m, 4 H), 3.32 (q, J ) 3.9 Hz, 1 H), 3.19 (d, J ) 16.6 Hz, 1 H),
2.51 (d, J ) 16.8 Hz, 1 H), 1.56 (s, 6 H), 1.22 (t, J ) 7.1 Hz, 3 H),
1.21 (t, J ) 7.1 Hz, 1 H), 0.91 (d, J ) 7.1 Hz, 3 H). 13C{1H} NMR:
δ 172.8, 171.0, 133.4, 127.1, 62.6, 61.3, 61.2, 48.9, 44.1, 14.3, 14.2,
13.7, 12.0. IR (neat, cm-1): 1720 (CdO). Anal. Calcd (found) for
C14H22O4: H, 8.72 (8.93); C, 66.12 (66.31).
Detection of Intermediates in the Isomerization of 2b. A solution
of 2b (100 mg, 0.47 mmol), HSiEt3 (110 mg, 0.94 mmol), 1a (10 mg,
0.022 mmol), NaBAr4 (24 mg, 0.024 mmol), and naphthalene (21 mg)
in CH2Cl2 (10 mL) was stirred at 0 °C and monitored periodically by
GC analysis. Relative concentrations of 2b, 3b, 4b, and 5b were
determined from the area of the respective peaks relative to the
naphthalene peak in the GC. Intermediates 4b and 5b were identified
by comparison to authentic samples. A plot of [2b] versus time was
linear to >3 half-lives from which a zero-order rate constant of k )
8.1 × 10-5 s-1 was obtained (Figure S1).
4,4-Dicarbomethoxy-1,2-dimethyl-3-phenylcyclopentene (28). 1H
NMR: δ 7.25 (m, 3 H), 7.10 (d, J ) 6.8 Hz, 2 H), 4.56 (s, 1 H), 3.72
(s, 3 H), 3.28 (d, J ) 16.8 Hz, 1 H), 3.09 (s, 3 H), 2.62 (d, J ) 17.0
Hz, 1 H), 1.72 (s, 3 H), 1.44 (s, 3 H). 13C{1H} NMR: δ 173.2, 170.3,
139.2, 132.0, 130.2, 129.4, 128.2, 127.3, 64.1, 61.8, 53.1, 44.9, 13.8,
12.8. IR (neat, cm-1): 1737 (CdO). HRMS (EI) Calcd (found) for
C17H19O4 (M - H+): 287.1283 (287.1288).
Isomerization of a 1:1 Mixture of 2a-2,6-d2 and 2b. A solution of
2a-2,6-d2 (50 mg, 0.47 mmol), 2b (50 mg, 0.47 mmol), HSiEt3 (80
mg, 0.70 mmol), 1a (10 mg, 0.022 mmol), and NaBAr4 (24 mg, 0.024
mmol), and naphthalene (21 mg) in CH2Cl2 (10 mL) was stirred at 0
°C and analyzed periodically by GC. Relative concentrations of 2a-
2,6-d2 and 2b were determined from the area of the respective peaks
relative to the naphthalene peak in the GC. Deuterium incorporation
was determined from GCMS of the respective aliquots.
4,4-Dicarbomethoxy-1,2,3,3-tetramethylcyclopentene (30). 1H
NMR: δ 3.67 (s, 6 H), 2.72 (s, 2 H), 1.59 (s, 3 H), 1.49 (s, 3 H), 1.05
(s, 6 H). 13C{1H} NMR: δ 171.9, 136.1, 126.3, 66.4, 52.2, 52.0, 43.1,
22.4, 14.0, 9.8. IR (neat, cm-1): 1733 (CdO). Anal. Calcd (found) for
C13H20O4: H, 8.39 (8.46); C, 64.98 (64.93).
4,4-Dicarbomethoxy-1-ethyl-2-methylcyclopentene (32) and 1,1-
Dicarbomethoxy-3-ethylidene-4-methylcyclopentane (33). A solution
of diene 31 (100 mg, 0.47 mmol), HSiEt3 (80 mg, 0.70 mmol), 1a (10
mg, 0.022 mmol), and NaBAr4 (24 mg, 0.024 mmol) in CH2Cl2 (8
mL) was stirred at 0 °C for 10 min to form a dark brown solution.
Evaporation of solvent and chromatography (hexane:EtOAc, 12:1) gave
a 3:1 mixture of 32 and 33 (84 mg, 84%) as a colorless oil. Carbocycle
33 was identified by comparison of the spectroscopy with that of an
authentic sample. In a separate reaction, a solution of diene 31 (100
mg, 0.47 mmol), HSiEt3 (440 mg, 3.8 mmol), 1a (10 mg, 0.022 mmol),
and NaBAr4 (24 mg, 0.024 mmol) in CH2Cl2 (8 mL) was stirred at
room temperature for 12 h to form a dark brown solution. Evaporation
of solvent and chromatography (hexane:EtOAc, 12:1) gave 32 (84 mg,
84%) as a colorless oil as a 33:1 mixture of 32:33. Anal. Calcd (found)
for C12H18O4: H, 8.02 (7.88); C, 63.70 (63.57).
Acknowledgment is made to the National Institutes of Health
(GM59830-01) and the donors of the Petroleum Research Fund
(33131-G1), administered by the American Chemical Society
for support of this research. R.W. thanks the Camille and Henry
Dreyfus Foundation and DuPont for young faculty awards and
the Alfred P. Sloan Foundation for a Research Fellowship. We
thank Nicholas Perch for performing some initial experiments.
Supporting Information Available: Experimental proce-
dures, spectroscopic, and analytical data for relevant compounds
(PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
1
For 32. H NMR: δ 3.69 (s, 6 H), 2.92 (overlapping region, 4 H),
2.00 (q, J ) 1 Hz, 2 H), 1.56 (s, 3 H), 0.92 (t, J ) 7.6 Hz, 3 H).
JA001730+