Regioselective catalytic transformations involving β-silyl-substituted (η3-allyl)palladium complexes: An efficient route to functionalized allylsilanes

Article abstract of DOI:10.1021/jo991172l

Various alkyl derivatives of 1-(trimethylsilanyl)but-3-en-2-ol acetate (1a-e) undergo regioselective palladium-catalyzed nucleophilic substitution via β-silyl-substituted (η³-allyl)palladium intermediates. With external nucleophiles, such as malonates and enolates, the nucleophilic substitution occurs with complete allylic rearrangement, providing functionalized allylsilanes as building blocks of high synthetic potential. Internal nucleophiles, such as disilanes and NaBPh₄, afford bisallylic disilanes and (allylsilyl)benzene derivatives with good regioselectivity. For both types of nucleophiles, the double bond geometry of the resulting allylsilane is selectively trans. The β-silyl-substituted (η³-allyl)palladium intermediates of the reaction were also isolated. The ¹H NMR studies indicate selective formation of the syn-isomer of the key (η³- allyl)palladium intermediates, which explains the high trans-selectivity of the double bond formation in the allylsilane products. According to the ¹³C NMR studies, the β-silyl functionality exerts deshielding effects on the nearest allylic terminal carbon (C3), which can be ascribed to hyperconjugative interactions between the silyl functionality and the allylpalladium moiety. It was concluded that, together with the steric effects of the silyl group, these electronic interactions are responsible for the high regioselectivity of the nucleophilic attack in the catalytic process.

Full text of DOI:10.1021/jo991172l

J . Org. Chem. 1999, 64, 9547-9556
9547
Regioselective Ca ta lytic Tr a n sfor m a tion s In volvin g
â-Silyl-Su bstitu ted (η3-Allyl)p a lla d iu m Com p lexes: An Efficien t
Rou te to F u n ction a lized Allylsila n es
Istva´n Macsa´ri, Eike Hupe, and Ka´lma´n J . Szabo´*
Arrhenius Laboratory, Department of Organic Chemistry, University of Stockholm,
S-106 91 Stockholm, Sweden
Received J uly 22, 1999
Various alkyl derivatives of 1-(trimethylsilanyl)but-3-en-2-ol acetate (1a -e) undergo regioselective
palladium-catalyzed nucleophilic substitution via â-silyl-substituted (η3-allyl)palladium intermedi-
ates. With external nucleophiles, such as malonates and enolates, the nucleophilic substitution
occurs with complete allylic rearrangement, providing functionalized allylsilanes as building blocks
of high synthetic potential. Internal nucleophiles, such as disilanes and NaBPh4, afford bisallylic
disilanes and (allylsilyl)benzene derivatives with good regioselectivity. For both types of nucleophiles,
the double bond geometry of the resulting allylsilane is selectively trans. The â-silyl-substituted
1
(η3-allyl)palladium intermediates of the reaction were also isolated. The H NMR studies indicate
selective formation of the syn-isomer of the key (η3-allyl)palladium intermediates, which explains
the high trans-selectivity of the double bond formation in the allylsilane products. According to
the 13C NMR studies, the â-silyl functionality exerts deshielding effects on the nearest allylic
terminal carbon (C3), which can be ascribed to hyperconjugative interactions between the silyl
functionality and the allylpalladium moiety. It was concluded that, together with the steric effects
of the silyl group, these electronic interactions are responsible for the high regioselectivity of the
nucleophilic attack in the catalytic process.
In tr od u ction
is the nucleophilic attack of the silylating reagent on the
(η3-allyl)palladium intermediate of the reaction.12-14 Al-
ternatively, the catalytic process can be started from an
appropriate silyl precursor, which undergoes oxidative
addition to Pd(0), providing a â-silyl-substituted (η3-allyl)-
palladium intermediate (Scheme 1). Subsequently, this
intermediate is attacked by the nucleophile, to give a
functionalized allylic silane. The main advantage of this
latter method is the employment of the electronic and
steric effects of the â-silyl functionality to control the
regiochemistry of the nucleophilic attack on the (η3-allyl)-
palladium intermediate.
Functionalized allylsilanes have proven to be an ex-
ceedingly useful class of organometallic reagents, and
they continue to show enormous potential in regio- and
stereocontrolled C-C bond formation reactions.1,2 Ac-
cordingly, allylsilanes provide building blocks of unprec-
edented synthetic potential, which are also frequently
used precursors in natural product synthesis.3-8 Although
many excellent ways of making allylsilanes are already
known,9-11 the great importance of allylsilane chemistry
for organic synthesis urged a continuing search for new
methods for efficient regio- and stereocontrolled prepara-
tion of functionalized allylsilanes.
However, there are few methods available on catalytic
allylations proceeding through â-silyl-substituted (η3-
allyl)palladium intermediates. Apart from our studies15
on the alkylation of 2-acetoxy-5-(dimethylphenylsilanyl)-
cyclohexene, we have only found one example in the
literature of this type of catalytic procedure involving a
single reaction of 1-(trimethylsilanyl)but-3-en-2-ol acetate
(1a ) with the sodium malonate nucleophile.16 Besides, in
these previously conducted studies only malonate nucleo-
philes were employed with just two different silyl sub-
strates, and therefore, the synthetic scope of the palla-
dium-catalyzed procedures proceeding through â-silyl-
substituted (η3-allyl)palladium intermediates remained
largely unexplored. Our recent results15 on the structure,
properties, and reactivity of â-silyl-substituted (η3-allyl)-
palladium complexes indicated that the steric and elec-
tronic effects of the silyl group enhance the regio- and
stereoselectivity of the nucleophilic attack, which prompted
Allylpalladium chemistry has been successfully used
for the preparation of allylsilanes.11-14 Usually the key
step in the palladium-catalyzed synthesis of allylsilanes
* To whom correspondence should be addressed. E-mail:
kalman@organ.su.se.
(1) Fleming, I.; Dunogues, J .; Smithers, R. Org. React. 1989, 37, 57.
(2) Colvin, E. W. Silicon Reagents in Organic Synthesis; Academic
Press: New York, 1988.
(3) Fleming, I.; Barbero, A.; Walter, D. Chem. Rev. 1997, 97, 2063.
(4) Langkopf, E.; Schinzer, D. Chem. Rev. 1995, 95, 1375.
(5) Masse, C. E.; Panek, J . S. Chem. Rev. 1995, 95, 1293.
(6) Panek, J . S. In Heteroatom Stabilized Carbanion Equivalents;
Fleming, I., Trost, B., Eds.; Pergamon: New York, 1991; Vol. 1, p 579.
(7) Hosomi, A. Acc. Chem. Res. 1988, 21, 200.
(8) Schinzer, D. Synthesis 1988, 263.
(9) Fleming, I.; Higgins, D.; Lawrence, N. J .; Thomas, A. P. J . Chem.
Soc., Perkin Trans. 1 1992, 3331.
(10) Sarkar, T. K. Synthesis 1990, 969 and 1101.
(11) Horn, K. A. Chem. Rev. 1995, 95, 1317.
(12) Tsuji, Y.; Funato, M.; Ozawa, M.; Ogiyama, H.; Kajita, S.;
Kawamura, T. J . Org. Chem. 1996, 61, 5779.
(13) Trost, B. M.; Yoshida, J .-I.; Lautens, M. J . Am. Chem. Soc. 1983,
105, 4494.
(14) Matsumoto, Y.; Ohno, A.; Hayashi, T. Organometallics 1993,
12, 4051.
(15) Macsa´ri, I.; Szabo´, K. J . Organometallics 1999, 18, 701.
(16) Chaptal, N.; Colovray-Gotteland, V.; Grandjean, C.; Cazes, B.;
Gore´, J . Tetrahedron Lett. 1991, 32, 1795.
10.1021/jo991172l CCC: $18.00 © 1999 American Chemical Society
Published on Web 11/25/1999

9548 J . Org. Chem., Vol. 64, No. 26, 1999
Macsa´ri et al.
Sch em e 1
Sch em e 2
us to further explore the synthetic utility of the pal-
ladium-catalyzed allylic substitution reactions proceeding
through this intermediate.
giving 1a -e, which are stable compounds under nona-
cidic conditions.
Usually palladium-catalyzed allylic substitution reac-
tions of allylic acetates require a high reaction temper-
ature or a relatively long reaction time20 since the
oxidative addition of allylic acetates to Pd(0) is relatively
slow. However, 1a -e undergo unusually facile oxidative
addition to Pd(0) (vide supra), which enables one to
employ mild reaction conditions in the silylallylation
reactions. Using the literature procedures,17,18 acetates
1d and 1e are obtained as 1:1 diastereomeric mixtures.
When these compounds undergo allylic substitution
reactions leading to diastereomeric allylsilanes, the ratio
of the diastereomers in the products is usually the same
as in the starting acetates.
In this study we present our results on the reactivity
of several different types of nucleophiles with 1a -e in
palladium-catalyzed allylic substitution reactions. In
particular, we discuss the level of regioselectivity as a
function of the direction of the nucleophilic attack
(Scheme 1), as well as the influence of the substitution
pattern of 1 on the rate of the catalytic substitution
reaction. To gain greater insight into the mechanistic
details, we prepared the â-silyl-substituted (η3-allyl)-
palladium key intermediates of the reaction and studied
their properties and their reactivities.
Ca ta lytic Allylic Su bstitu tion Rea ction s by Ex-
ter n a l Nu cleop h iles. Palladium-catalyzed allylic sub-
stitution reactions of 1 (Scheme 2 and Table 1) by
malonates, enoxyborates, and cyclopentadienyl-sodium
proceed smoothly under mild reaction conditions, provid-
ing δ-functionalized allylsilanes in good yield. All reac-
tions proceed with complete allylic rearrangement, and
the double bond produced has exclusively a trans geom-
etry.
Rea ction s w ith Ma lon a te Nu cleop h iles (En tr ies
1-7). Diethyl malonate can be easily allylsilylated under
mild conditions in THF at 45 °C. This reaction proceeds
with very high selectivity and with a good yield for all
alkyl-substituted derivatives of 1. The reaction times
depend on the substitution pattern of 1, as well as on
Resu lts a n d Discu ssion
The substrates of silylallylation reactions, 1a-e (Scheme
2), could be easily obtained by simple literature proce-
dures.17,18 The precursors of 1a -e (1-(trimethylsilanyl)-
but-3-en-2-ol derivatives) are often used for the prepa-
ration of conjugated dienes since these silylbutenol
derivatives easily undergo Peterson elimination19 by loss
of R3SiOH. Indeed, when we attempted to replace the
hydroxy group of the silylbutenol derivatives by better
leaving groups, such as Cl, OCOOR, or OCOCF3 func-
tionalities, the Peterson type of elimination reactions
were dominating. On the other hand, the silylbutenol
derivatives could be acylated under mild conditions,
(17) Sato, S.; Matsuda, I.; Izumi, Y. J . Organomet. Chem. 1988, 344,
71.
(18) Wilson, S. R.; Price, M. F. J . Am. Chem. Soc. 1982, 104, 1124.
(19) Peterson, D. J . J . Organomet. Chem. 1967, 9, 373.
(20) Tsuji, J . Palladium Reagents and Catalysis: Innovations in
Organic Synthesis; Wiley: Chichester, 1995; Chapters 3 and 4.

Catalytic Transformations Involving Pd Complexes
J . Org. Chem., Vol. 64, No. 26, 1999 9549
Ta ble 1. P a lla d iu m -Ca ta lyzed Allylic Su bstitu tion of 1a -e
a
Unless otherwise stated 5 mol % palladium catalyst was used in THF. Method A, Pd(PPh3)4; B, Pd(dba)2, dppe, 1:1; C, Pd(dba)2,
P(OMe)3, 1:1; D, 17, MAH, 1:3 in CH2Cl2; E, Pd(OAc)2, MAH, 1:3 in CH2Cl2; F, 2 mol % Pd(dba)2, Pd(PPh3)4, 1:2. For details see the
b
Experimental Section. Isolated yield. c 1- and 2-cyclopentadienyl isomers form in a ratio of 1:1.
the applied ancillary ligand. The catalytic reactions are
complete in 1-2 h for 1c and 1e; however, a longer
reaction time is required when the attacked carbon bears
an alkyl substituent (R1 ) CH3). It was also found that
the alkylation reaction proceeds faster in the presence
of P(OCH3)3 and dppe ligands than with a PPh3 ligand.
En oxybor a tes a s Nu cleop h iles (En tr ies 8-16).
Enoxyborates were introduced by Negishi and co-
workers21-23 as versatile enolate type nucleophiles in
allylpalladium chemistry. Enoxyborate reagents are eas-
ily obtained by treating potassium enolates with trialky-
lboranes, such as triethylborane. These reagents are
particularly useful for the preparation of keto- and
formyl-functionalized allylsilanes from 1. The silylally-
lation process is remarkably fast for substrates that are
unsubstituted at the δ-carbon (R1 ) H). The presence of
an alkyl group at the R-position (R2 ) C5H11) does not
influence the rate of alkylation. Allylic substitution of 1a ,
1b, and 1d with enoxyborates is usually complete in 30
min at 0 °C. It is interesting to note that the typical
reaction conditions for palladium-catalyzed alkylation of
enoxyborates usually involve stirring for 12 h at ambient
temperature.22 The alkylation of enoxyborates failed for
allylic acetates 1c and 1e bearing an alkyl substituent
at the δ-position (R1 ) CH3). Under the applied alkylation
conditions, 1c and 1e undergo AcOH elimination, giving
silyl-substituted 1,3-pentadienes.
Silyla llyla tion of Cyclop en ta d ien e (En tr ies 17
a n d 18). We have briefly investigated the application of
cyclopentadienyl-sodium as a nucleophile24 in palladium-
(21) Negishi, E.-i.; Matsushita, H.; Chatterjee, S.; J ohn, R. A. J . Org.
Chem. 1982, 47, 3188.
(22) Negishi, E.-i.; Luo, F.-T. J . Org. Chem. 1983, 48, 2427.
(23) Negishi, E.-i.; J ohn, R. A. J . Org. Chem. 1983, 48, 4098.
(24) Fiaud, J . C.; Malleron, J . L. Tetrahedron Lett. 1980, 21, 4437.

9550 J . Org. Chem., Vol. 64, No. 26, 1999
Macsa´ri et al.
Sch em e 3
catalyzed allylic substitution of 1b and 1d . The rate of
alkylation is rather high at rt, and a good yield of the
cyclopentadienyl products 4a and 4b is obtained when
the δ-carbon is unsubstituted. However, the yield is poor
when 1c is used as a substrate due to the extensive
formation of trimethylpenta-1,3-dienylsilane. Although
the regioselectivity and trans-selectivity of the catalytic
reaction are excellent, the 1- and 2-cyclopentadienyl
isomers form in a ratio of 1:1.
Ca ta lytic Allylic Su bstitu tion Rea ction s by In -
ter n a l Nu cleop h iles. Palladium-catalyzed allylic sub-
stitution reactions of allylic acetates with disilanes 13a ,b
and NaBPh4 (14) involve coordination of the nucleophile
to palladium followed by cis attack on the allyl moiety
(see Scheme 1).12,25 Reagents 13 and 14 react with 1 in a
relatively good yield, leading to the E-isomers of 5 and
6; however, the regioselectivity is usually lower than with
external nucleophiles.
Rea ction s w ith Disila n es 13a ,b (En tr ies 19-26).
Bisallylic disilanes could be obtained from 1a -e under
mild conditions using a modified version of the pal-
ladium-catalyzed silylation procedure by Tsuji and co-
workers.12 According to the original procedure12 allylic
acetates are reacted with disilanes 13a ,b in the presence
of CF3COOH at ambient temperature or in the presence
of LiCl at 100 °C. However, these harsh reaction condi-
tions are not suitable for the preparation of 5. A further
problem is that phosphine ligands cannot be used in the
silylation reaction (vide supra), and in the absence of
PPh3 or other π-acceptor ligands, amorphous Pd(0) is
precipitated after a couple of catalytic cycles. However,
we have found that the catalytic activity of Pd(0) can be
maintained when the reaction is conducted under neutral
conditions at 30 °C in CH2Cl2 in the presence of maleic
anhydride (MAH).26 Relatively good yields can be ob-
tained with Pd(OAc)2 as the catalyst; however, the yields
are improved when Pd(OAc)2 is replaced by a catalytic
amount of the allylpalladium intermediates of the reac-
tion. A complete control of the regioselectivity can be
achieved for substrates 1a , 1d , and 1e; however, a
considerable amount (20%) of 1,2-isomer is formed when
1c is applied as substrate.
â-silyl functionality is present in the intermediary (η3-
allyl)palladium complex, useful levels of the regioselec-
tivity can be achieved. Thus, palladium-catalyzed phen-
ylation of 1d and 1e with NaBPh4 gives only δ-substituted
products 6e and 6f. The level of the regioselectivity is
still acceptable for phenylation of 1b (6a :6b ) 6:1);
however, the regioselectivity is poor when 1c is employed
as a substrate. The phenylation reaction required a
higher temperature (60 °C) than the other allylic sub-
stitution reactions of 1 because of the sluggish phenyl
transfer in the presence of the PPh3 ligand (vide supra).
Str u ctu r e a n d P r op er ties of th e (η3-Allyl)p a lla -
d iu m In ter m ed ia tes. Compounds 1a , 1b, and 1c react
smoothly with Pd(dba)2 in the presence of LiCl at room
temperature, giving allylpalladium complexes 15a -c
(Scheme 3). A complete conversion could be achieved in
2 h without desilylation of the allylic acetates or the
products. At room temperature the rate of addition is
much faster than the oxidative addition of other allylic
acetates without a silyl functionality.20 Since the oxida-
tive addition could be carried out without desilylation,
the driving force of the reaction is apparently the forma-
tion of an exceptionally stable (η3-allyl)palladium com-
plex.
Air-stable complexes 15a -c could easily be purified by
chromatography. The chloride ligand of 15a -c could be
exchanged to PPh3 (16a -c) and OAc (17a -c) ligands,
generating the active intermediates of the external
(entries 1-18) and the internal (entries 19-30) nucleo-
philic attacks, respectively. The phosphine and acetate
complexes are fairly stable at room temperature or at a
slightly elevated temperature (up to 40-50 °C), but
decompose when the solvent is evaporated.
The 1H NMR spectrum of 15a -c shows that only a
single isomer is formed in the oxidative addition process.
3
The relatively large J H2-H3 coupling constant (11.2 Hz)
indicates the selective formation of the syn-isomers.
Exchange of the chloride ligand to phosphine does not
change the configuration of the complex. The high
configurational stability of 15a -c and 16a -c explains
the excellent trans-selectivity of the double bond forma-
tion. The 13C NMR spectra of 15 and 16 show some
interesting features. Comparison of the 13C NMR shifts
of the allylic terminal carbons in 1828 and 15a ,b reveals
the perturbation effects of the SiMe2R3 group on the
electronic structure of the allylic moiety (Figure 1). The
chemical shift difference between the terminal carbons
(∆δt) increases in the presence of the silyl substituent.
P h en yla tion w ith Na BP h 4 (En tr ies 27-30). Allylic
acetates undergo a substitution reaction with NaBPh4
(14) in the presence of a catalytic amount of Pd(PPh3)4,
providing allylbenzene derivatives with rather poor re-
gioselectivity.27 However, for certain substrates, when a
(25) Kurosawa, H.; Ogoshi, S.; Kawasaki, Y.; Murai, S.; Miyoshi,
M.; Ikeda, I. J . J . Am. Chem. Soc. 1990, 112, 2813.
(26) Kurosawa, H.; Kajimaru, H.; Ogoshi, S.; Yoneda, H.; Miki, K.;
Kasai, N.; Murai, S.; Ikeda, I. J . Am. Chem. Soc. 1992, 114, 8417.
(27) Legros, J .-Y.; Fiaud, J .-C. Tetrahedron Lett. 1990, 31, 7453.
(28) A° kermark, B.; Krakenberger, B.; Hansson, S.; Vitagliano, A.
Organometallics 1987, 6, 620.

Catalytic Transformations Involving Pd Complexes
J . Org. Chem., Vol. 64, No. 26, 1999 9551
F igu r e 1. 13C NMR shifts (ppm) in CDCl3 of the allylic carbons. ∆δt denotes the chemical shift differences between C3 and C1
(∆δt ) δC3 - δC1).
F igu r e 2. Reactivity of (η3-allyl)palladium complexes 16b and 17b,c toward nucleophiles 13a and 14.
Furthermore, the chemical shift of the allylic terminus
closer to the silyl functionality (C3) in 15a ,b is observed
at a lower field than the corresponding 13C shift in 18,
indicating a deshielding effect of the â-SiMe2R3 group on
the C3 allylic position. The asymmetrizing effect of the
silyl substituent on the electronic structure of the allylic
moiety is also reflected by the 13C shifts of 15c. The
presence of the silyl functionality leads to a difference of
6.7 ppm in the Ct shift values, and the closer allylic
terminus (C3) is less shielded than the remote one. The
same trend applies to the phosphine complexes. The ∆δt
values of 16a ,b and 1928 differ even more than the
corresponding values in the chloro complexes 15a ,b and
18. Similarly, the perturbation effects of the silyl func-
tionality on the allylic terminal carbons are stronger in
the phosphine complex 16c (∆δt ) 10.4 ppm) than in the
chloro complex 15c (∆δt ) 6.7 ppm). There is a small but
significant difference between the electronic effects of the
SiMe3 and SiMe2Ph functionalities. Since the Ph group
is more electron withdrawing than the methyl group,
replacement of a SiMe3 with a SiMe2Ph functionality is
expected to lead to some deshielding effect at C3. On the
contrary, in the presence of the SiMe3 (15a and 16a )
group, δ(C3) is larger by 1.5-1.7 ppm than for the SiMe2-
Ph functionality, indicating that the electronic interac-
tions between the silyl functionality and allylpalladium
moiety cannot simply be explained on the basis of
inductive effects. The above differences in the 13C chemi-
cal shift values are also in line with our recent observa-
tions15 on the â-effects of the silyl group occurring in
stereodefined cyclic (η3-allyl)palladium complexes.
Nu cleop h ilic Atta ck on â-Silyl Su bstitu ted (η3-
Allyl)p a lla d iu m Com p lexes. The reactivity of 16a -c
and 17a -c was studied with both external and internal
nucleophiles. The outcome of the reaction of 16a -c with
malonates is the same as in the corresponding catalytic
processes. However, the reaction of the allylpalladium
complexes with 13 and 14 revealed some interesting
mechanistic features. When 1b is reacted with 14, Pd-
(PPh3)4 is used as the catalyst, suggesting that the key
intermediate of the reaction would be the phosphine
complex 16b. However, complex 16b does not react with
NaBPh4 at 0 °C or ambient temperature. On the other
hand, the acetate complex 17b reacts immediately with
NaBPh4 at 0 °C, giving 6a and 6b (Figure 2). Since, 17b
is supposed to be the primary product of the oxidative
addition of 1b to palladium, it is assumed that the key
intermediate of the phenylation process is the acetate
complex, while formation of 16b inhibits the catalytic
cycle. The most probable explanation for the inhibitory
effect of PPh3 is that phosphine ligands coordinate too
strongly to Pd, and accordingly, the phenyl transfer from
-
BPh4 to palladium is hindered. The high reaction
temperature (60 °C) employed in the catalytic reactions
is required because of this inhibitory effect of the PPh3
ligand. On the other hand, the PPh3 ligand is required

9552 J . Org. Chem., Vol. 64, No. 26, 1999
Macsa´ri et al.
Sch em e 4
also ref 15). Systematic changes in the δ(Ct) values
(Figure 1) can also be ascribed to this effect. As the
palladium atom has a considerable shielding effect on the
allylic carbons,31 partial deshielding of C3 arising from
the presence of the â-silyl substituent is indicative of
weakening of the Pd-C3 interactions. Accordingly, varia-
tion of the 13C NMR shifts in 15 and 16 indicates
significant hyperconjugative interactions between the
C-Si bond and the allylmetal π-system.
The regioselectivity of the nucleophilic attack is influ-
enced by three important effects arising from the hyper-
conjugative interactions:
to maintain the catalytic cycle by complexing Pd(0)
formed in the reductive elimination step.
(1) The hyperconjugation is extended to the C1-C4
fragment and the silicon atom (Scheme 4). A nucleophilic
attack at the C3 position would interrupt this conjuga-
tion, so the C1 terminus is preferentially attacked.
Besides, the π-lobe in the LUMO at C3 is partially
occupied because of the interaction with the filled σ(C-
Si) MO, which effectively hinders the charge transfer
from the lone pair of an attacking nucleophile.
Phosphine ligands also inhibit the nucleophilic attack
by disilanes 13. Phosphine complexes 16a -c do not
undergo silylation with 13a ,b, probably because the silyl
coordination is also hindered in the presence of a PPh3
ligand (Figure 2). On the other hand, disilanes 13 react
readily with the acetate complexes 17a -c. However, a
complete conversion of 17a -c with 13 requires at least
0.5 h at 0 °C. This indicates that acetate complexes react
considerably more slowly with disilanes than with the
NaBPh4 reagent. The regioisomer ratio in the stoichio-
metric reaction of 17c with hexamethyldisilane (13a ) is
9:1 (5c:5d ), which is better than the isomer ratio (4:1) in
the corresponding catalytic process (entry 22). It seems
that the low reaction temperature (0 °C) considerably
improves the isomer ratio of the internal attack with
preference to the C1 attack of the (η3-allyl)palladium
intermediate. On the other hand, the catalytic silylation
and phenylation reactions proceed slowly at 0 °C, and at
low temperature, the starting material 1 cannot be
completely converted to the corresponding products. The
chloro complexes 15a -c also react with 13a ; however,
the reaction is very slow at ambient or slightly elevated
temperatures. For example, the reaction of 15c with 13a
could not be completed within 24 h at 30 °C. Chloride is
also a strongly coordinating ligand, and probably, the silyl
coordination is hindered in the presence of chloride
ligands.
Regioselectivity of th e Nu cleop h ilic Atta ck . Ap-
parently, the steric effects of the silyl functionality are
important factors for the regiochemical outcome of the
external nucleophilic attack. However, the 13C NMR
studies (Figure 1) and the previous theoretical studies15
have shown that the silyl functionality has considerable
electronic effects on the allylic moiety. It is well estab-
lished that σ(C-Si) orbitals readily conjugate with low-
lying unfilled π-levels.29,30 The interactions between σ(C-
Si) and the pπ-orbital of carbocations form the basis of
the well-known â-silicon effect.29,30 The effects of â-silicon
substituents in (η3-allyl)palladium complexes can also be
ascribed to the electronic interactions between the un-
filled LUMO of the allylpalladium fragment (dπ*) and the
high-lying σ(C-Si) orbital (Scheme 4). Delocalization of
electrons from the bonding σ(C-Si) orbital leads to
weakening of the C-Si bond. Since the dπ* is Pd-Ct
antibonding, the allylpalladium bonding is also weak-
ened. The σ(C-Si) orbital directly overlaps with the
π-lobe of C3, and accordingly, the Pd-C3 bonding is
weakened to a larger extent than the Pd-C1 bonding (see
(2) For an external attack (entries 1-18), the steric
effects of the bulky silyl group also facilitate the nucleo-
philic attack at the C1 position of the allyl. Previously
conducted theoretical studies have shown15 that the
hyperconjugative interactions stabilize such conforma-
tions where the silyl group is trans to palladium and the
C-Si bond is perpendicular to the plane of the allyl
moiety (Scheme 4). Accordingly, the electronic interac-
tions restrict the conformation of the silyl functionality
so that the steric effects of SiMe2R3 can direct the
external nucleophiles to the less substituted allylic posi-
tion most effectively.
(3) The nucleophilic attack on the allyl moiety involves
Pd-C bond breaking,32 which is easier for the Pd-C3
bond weakened by the hyperconjugative interactions.
However, this interaction would facilitate the 1,2-attack
of the nucleophile to a greater extent.
The electronic (1) and the steric (2) effects work in the
same direction for an external nucleophilic attack on the
(η3-allyl)palladium intermediates. However, for an inter-
nal attack, the steric effect of the â-silyl substituent does
not influence the regioselectivity, and the subtle balance
between the opposing effects (1 and 3) determines the
level of the regioselectivity. Apart from entry 22, the
silylation reactions provide useful levels of regioselectiv-
ity. However, phenylation by NaBPh4 requires the pres-
ence of a bulky alkyl group at the R-position to proceed
with high regioselectivity (entries 29 and 30). The sto-
ichiometric reaction (Figure 2) between 17b and NaBPh4
suggests that the internal attack is very fast and the
coordination of Ph to palladium leads to immediate cis-
migration. However, the silyl migration is considerably
slower than the phenyl migration process. Thus, before
the cis-migration takes place, the various silyl-coordi-
nated (η3-allyl)palladium complexes can equilibrate to a
greater extent than the phenyl coordinated ones, induc-
ing a more selective nucleophilic attack in the silylation
process. Furthermore, the stoichiometric reaction of 17c
with 13a (Figure 2) at low temperature (0 °C) proceeds
slower but with higher regioselectivity than the corre-
sponding catalytic reactions at 30 °C, which also indicates
that a slow cis-migration process involves larger dif-
(29) Ibrahim, M. R.; J orgensen, W. L. J . Am. Chem. Soc. 1989, 111,
819.
(30) Wierschke, S. G.; Chandrasekhar, J .; J orgensen, W. L. J . Am.
Chem. Soc. 1985, 107, 1496.
(31) Schindler, M. J . Am. Chem. Soc. 1987, 109, 1020.
(32) Szabo´, K. J . Organometallics 1998, 17, 1677.

Catalytic Transformations Involving Pd Complexes
J . Org. Chem., Vol. 64, No. 26, 1999 9553
Sch em e 5
ferentiation between the C1 and C3 carbons of the (η3-
allyl)palladium intermediates.
Hosomi-Sakurai cyclization of 20b is currently under
investigation.
Syn th etic Utility of F u n ction a lized Allylsila n es
2-6. Products of the above silylallylation reactions are
useful building blocks of high synthetic potential. Cyclo-
pentadienyl derivatives 4a ,b can be used for the prepara-
tion of new types of cyclopentadienyl ligands to modulate
the structure and reactivity of different transition metal
based metallocenes.33 Bisallylic disilanes 5a -h are useful
precursors for the preparation of regio- and stereodefined
homoallylic silanes.34 Our method can also be used for
the preparation of bisallylic silanes with different silyl
groups (entries 20, 24, and 26). Because of the different
reactivity1 of the SiMe3 and SiMe2Ph functionalities,
these silyl groups can be selectively functionalized.
(Allylsilyl)benzene derivatives are often used as synthetic
intermediates,9 and the above method offers a simple
access to some new derivatives.
Con clu sion s
In this study we have shown that palladium-catalyzed
allylic substitution involving â-substituted (η3-allyl)-
palladium intermediates is a versatile method for the
preparation of functionalized allylsilanes. An excellent
regioselectivity was obtained when external nucleophiles,
such as malonates and enoxyborates, were employed.
Still useful levels of regiochemistry can be achieved using
internal nucleophiles such as disilanes and NaBPh4. For
both types of nucleophiles, the double bond geometry in
the resulting allylsilane is selectively trans. NMR studies
of the key intermediates 15 and 16 revealed that hyper-
conjugative interactions occur between the â-silyl sub-
stituent and the allylpalladium moiety, and that these
interactions are also dependent on the electronic effects
of the ancillary ligand employed. It was concluded that
these electronic effects enhance the regioselectivity of the
nucleophilic attack on the (η3-allyl)palladium intermedi-
ates of the catalytic reactions.
Allylsilanes functionalized by keto, formyl, and carboxy
groups are particularly important synthetic inter-
mediates,18,35-39 which are frequently used as precursors
in natural product synthesis. Accordingly, among the
difunctional allylsilanes obtained using the above tech-
nology (Scheme 2), 2 and 3 are certainly the most useful
synthetic precursors. These compounds possess a carbo-
nyl functionality, which can be converted by both nu-
cleophilic and reductive reagents without affecting the
silyl functionality. On the other hand, the allylsilyl
functionality can be transformed by electrophiles without
affecting the carbonyl group. Furthermore, these com-
pounds can also be valuable precursors for the prepara-
tion of stereodefined isocyclic compounds and sesquiter-
pene analogues.18,39 This can be illustrated by the synthesis
sequence shown in Scheme 5. Aldehyde 3e is converted
to 20a ,b by a Grignard reaction followed by oxidation.
When 20a ,b are treated with Lewis acids, an intramo-
lecular Hosomi-Sakurai reaction40 takes place, affording
isocyclic compounds 21a ,b. When R ) CH3, the intramo-
lecular cyclization of 20 provides selectively the cis-
derivative 21b. The mechanism of this stereoselective
Exp er im en ta l Section
The starting materials were purchased from Aldrich or
Lancaster. All solvents were freshly distilled prior to use. All
the reactions were conducted under an argon atmosphere by
employing standard manifold techniques. All 1H and 13C NMR
spectra were recorded in CDCl3 solutions at 300 and 75 MHz,
respectively. The chemicals shifts (ppm) are obtained using
CDCl3 as internal standard (7.26 ppm, 1H; 77.0 ppm, 13C). For
chromatography Merck silica gel 60 (230-400 mesh) was used.
Gen er a l P r oced u r e for Ca t a lyt ic Alk yla t ion w it h
Sod iu m Ma lon a tes (En tr ies 1-7). Meth od A. Dialkyl
malonate (1.92 mmol) in THF (2.6 mL) was slowly added to a
slurry of NaH (1.47 mmol, 0.059 g, 60% suspension in mineral
oil) in THF (2.6 mL). After the resulting clear solution was
transferred via cannula to a mixture of acetate 1 (0.37 mmol)
and Pd(PPh3)4 (0.018 mmol, 5 mol %, 0.021 g) in THF (0.9 mL),
the combined mixture was stirred for the temperatures and
times listed in Table 1. After the reaction was complete, the
reaction mixture was diluted by ether (30 mL), filtered through
Celite, followed by extraction with saturated NH4Cl solution
and brine, and dried over anhydrous MgSO4. Removal of the
solvent in vacuo and purification of the residual yellow oil by
chromatography (pentane-ether, 9:1) gave the product as a
colorless oil.
(33) Barlow, S.; O’Hare, D. Chem. Rev. 1997, 97, 637.
(34) Wickham, G.; Kitching, W. Organometallics 1983, 2, 541.
(35) Castan˜o, A. M.; Ba¨ckvall, J .-E. J . Am. Chem. Soc. 1995, 117,
560.
(36) Castan˜o, A. M.; Persson, B. A.; Ba¨ckvall, J .-E. Chem. Eur. J .
1997, 3, 482.
(37) Block, M. H.; Cane, D. E. J . Org. Chem. 1988, 53, 4923.
(38) Wilson, S. R.; Augelli-Szafran, C. E. Tetrahedron 1988, 44, 3983.
(39) Wilson, S. R.; Price, M. F. J . Org. Chem. 1984, 49, 722.
(40) Hosomi, A.; Shirahata, A.; Sakurai, H. Tetrahedron Lett. 1978,
3043.
Meth od B. Pd(dba)2 (0.018 mmol, 5 mol %, 0.011 g) and
dppe (0.018 mmol, 0.007 g) were used as the catalyst.
Meth od C. Pd(dba)2 (0.018 mmol, 5 mol %, 0.011 g) and
P(OMe)3 (0.04 mmol, 0.005 g) were employed as the catalysts.
Otherwise the method was the same as method A.

9554 J . Org. Chem., Vol. 64, No. 26, 1999
Macsa´ri et al.
(E)-2-(5-(Tr im eth ylsila n yl)p en t-3-en -2-yl)m a lon ic Acid
Dieth yl Ester (2a ). H NMR: δ 5.48 (m, 1H), 5.17 (m, 1H),
1H), 5.16 (m, 1H), 2.12 (d, J ) 7.5 Hz, 2H), 1.68 (d, J ) 7.8
Hz, 2H), 0.98 (s, 6H), 0.27 (s, 6H). 13C NMR: δ 206.8, 139.0,
133.9, 129.3, 128.1, 130.3, 123.7, 46.4, 41.0, 22.3, 21.5, -3.0.
Anal. Calcd for C16H24OSi: C, 73.79; H, 9.29. Found: C, 73.65;
H, 9.47.
(E)-2,2-Dim eth yl-6-(tr im eth ylsila n yl)h ex-4-en a l (3e).
1H NMR: δ 9.48 (s, 1H), 5.43 (m, 1H), 5.13 (m, 1H), 2.14 (d, J
) 7.2 Hz, 2H), 1.43 (d, J ) 7.8 Hz, 2H), 1.03 (s, 6H), -0.02 (s,
9H). 13C NMR: δ 206.5, 130.8, 122.7, 46.6, 41.1, 23.4, 21.6,
-1.4. Anal. Calcd for C11H21OSi: C, 66.94; H, 10.72. Found:
C, 66.74; H, 10.91.
1
4.15 (m, 4H), 3.20 (d, J ) 9.1 Hz, 1H), 2.88 (m, 1H), 1.39 (dd,
J ) 8.1 and 1.1 Hz, 2H), 1.23 (m, 6H), 1.05 (d, J ) 6.7 Hz,
3H), -0.04 (s, 9H). 13C NMR: δ 168.7, 168.6, 130.1, 128.0, 61.5,
61.4, 58.9, 37.9, 23.1, 19.4, 14.6, -1.6. Anal. Calcd for C15H28O4-
Si: C, 59.96; H, 9.39. Found: C, 59.96; H, 9.50.
(E)-2-(4-(Tr im eth ylsila n yl)n on -2-en yl)m a lon ic Acid Di-
eth yl Ester (2b). 1H NMR: δ 5.33 (m, 1H), 5.15 (m, 1H), 4.18
(m, 4H), 3.35 (t, J ) 7.6 Hz, 1H), 2.60 (dt, J ) 7.7 and 0.9 Hz,
2H), 1.39 (m, 1H), 1.26 (t, J ) 6.9 Hz, 6H), 1.22 (br m, 8H),
0.87 (t, J ) 6.6 Hz, 3H), -0.07 (s, 9H). 13C NMR: δ 169.5,
135.8, 123.2, 61.6, 53.2, 33.5, 32.5, 32.1, 29.3, 29.0, 23.3, 14.5,
14.4, -2.9. Anal. Calcd for C19H36O4Si: C, 64.00; H, 10.17.
Found: C, 64.04; H, 10.26.
(E )-2-(4′-(Tr im e t h ylsila n yl)n on -2′-e n yl)cycloh e xa -
n on e (3f). 1H NMR: δ 5.18 (m, 2H), 2.54-2.20 (br m, 3H),
2.18-1.91 (br m, 4H), 1.89-1.78 (m, 1H), 1.76-1.54 (br m,
2H, H5), 1.48-1.06 (br m, 10H), 0.87 (t, J ) 7.0 Hz, 3H), -0.06
(s, 9H). 13C NMR: δ 213.3, 134.2, 125.5, 51.5, 42.3, 33.5, 33.4,
33.2, 28.3, 25.1, 23.0, 14.5, -2.8. HRMS (EI): m/z calcd for
(E)-2-Meth yl-2-(4-(tr im eth ylsila n yl)n on -2-en yl)m a lon -
1
ic Acid Dim eth yl Ester (2c). H NMR: δ 5.31 (m, 1H), 5.07
C
18H34OSi (M+) 294.2379, found 294.2378.
(m, 1H), 3.71 (s, 6H), 2.59 (dd, J )7.5 and 0.8 Hz, 2H), 1.39 (s,
3H), 1.25 (br m, 9H), 0.87 (m, 3H), -0.05 (s, 9H). 13C NMR: δ
173.0, 137.6, 121.3, 54.3, 52.8, 39.7, 33.7, 32.0, 29.4, 29.0, 22.9,
20.1, 14.5, -2.9. Anal. Calcd for C18H34O4Si: C, 63.11; H, 10.00.
Found: C, 63.20; H, 10.14.
(E)-6-(Tr im eth ylsilan yl)-1-ph en ylu n dec-4-en -1-on e (3g).
1H NMR: δ 7.95 (m, 2H) 7.49 (m, 3H), 5.27 (m, 2H, H4), 3.01
(t, J ) 7.8 Hz, 2H), 2.45 (m, 2H), 1.38 (m, 1H), 1.33-1.11 (m,
8H), 0.86 (t, J ) 6.8 Hz, 3H), -0.07 (s, 9H). 13C NMR: δ 200.3,
137.5, 133.2, 128.9, 128.4, 133.4, 126.4, 39.5, 33.4, 32.1, 29.4,
29.1, 28.1, 22.9, 14.5, -2.8. Anal. Calcd for C20H32OSi: C,
75.89; H, 10.19. Found: C, 75.74; H, 10.36.
(E)-2-(5-(Tr im eth ylsila n yl)d ec-3-en -2-yl)m a lon ic Acid
Dieth yl Ester (2d ). Formed as a 1:1 mixture of diastereomers
1
from a 1:1 diastereomeric mixture of 1e. H NMR: δ 5.26 (m,
1H), 5.07 (m, 1H), 4.16 (m, 4H), 3.23 (t, J ) 7.5 Hz, 1H), 2.90
(m, 1H), 1.4-1.1 (br m, 15H), 1.06 (dd, J ) 6.6 and 1.8 Hz,
3H), 0.87 (t, J ) 6.6 Hz, 3H), -0.06 (s, 9H). 13C NMR: δ 168.9,
168.7, 133.7, 133.4, 129.5, 129.4, 61.5, 61.4, 59.0, 58.8, 38.0,
37.8, 33.5, 33.2, 32.1, 32.0, 29.4, 29.2, 29.1, 28.9, 22.9, 19.6,
19.4, 14.5, -2.9. Anal. Calcd for C20H38O4Si: C, 64.82; H, 10.33.
Found: C, 64.76; H, 10.46.
(E)-4-Meth yl-6-(tr im eth ylsilan yl)tr idec-6-en -3-on e (3h ).
Formed as a 1:1 mixture of diastereomers from a diastereo-
1
meric mixture of 1d . H NMR: δ 5.21 (m, 1H), 5.10 (m, 1H),
2.54 (m, 1H), 2.44 (dq, J ) 7.2 and 1.5 Hz, 2H), 2.32 (m, 3H),
2.03 (m, 1H), 1.42-1.32 (m, 1H), 1.30-1.08 (m, 8H), 1.04 (d,
J ) 7.0 Hz, 3H), 1.02 (t, J ) 7.2 Hz, 3H), 0.86 (t, J ) 6.8 Hz,
3H), -0.07 (s, 9H). 13C NMR: δ 215.1, 134.7, 124.9, 47.0, 46.9,
37.0, 36.8, 34.8, 34.7, 33.5, 33.4, 32.1, 32.0, 29.4, 29.1, 22.9,
16.5, 16.4, 14.5, 8.0, -3.1. Anal. Calcd for C17H34OSi: C, 72.27;
H, 12.13. Found: C, 72.10; H, 12.31.
Gen er a l P r oced u r e for Ca t a lyt ic Alk yla t ion w it h
En oxybor a tes (En tr ies 8-16). The corresponding ketone
(1.84 mmol) in THF (0.9 mL) was slowly added to a slurry of
KH (1.68 mmol, 0.225 g, 30% suspension in mineral oil) in THF
(0.8 mL) at 0 °C and stirred at rt until the gas evaluation had
finished. Triethylborane (1.84 mmol, 1.84 mL, 1 M solution
in THF) was slowly added to the suspension of potassium
enolate formed, and the mixture was stirred for a further 5
min. The resulting clear, pale yellow solution was transferred
via cannula to the THF (1.5 mL) solution of acetate 1 (0.51
mmol) and Pd(PPh3)4 (0.026 mmol, 5 mol %, 0.031 g) and then
stirred for the temperatures and times given in Table 1. After
the reaction mixture was diluted by ether (30 mL), it was
extracted with saturated NH4Cl solution and brine, followed
by drying over anhydrous MgSO4. Subsequently, the solvent
was removed, and the residual yellow oil was purified by
chromatography.
(E)-2-(4′-(Dim et h ylp h en ylsilyl)b u t -2′-en yl)cycloh exa -
n on e (3a ). 1H NMR: δ 7.49 (m, 2H), 7.34 (m, 3H), 5.39 (m,
1H), 5.20 (m, 1H), 2.52-2.14 (br m, 4H), 2.08-1.96 (m, 3H),
1.90 (m, 1H), 1.82 (m, 1H), 1.66 (d, J ) 7.7 Hz, 2H), 1.74-
1.52 (br m, 2H), 1.37-1.22 (m, 1H), 0.27 (s, 6H). 13C NMR: δ
213.3, 139.2, 133.9, 129.2, 128.0, 127.8, 127.1, 51.4, 42.3, 33.5,
33.1, 28.3, 25.2, 22.2, -2.9. Anal. Calcd for C18H26OSi: C,
75.46; H, 9.15. Found: C, 75.29; H, 9.40.
(E)-2,2-Dim eth yl-6-(tr im eth ylsila n yl)u n d ec-4-en a l (3i).
1H NMR: δ 9.47 (s, 1H), 5.25 (m, 1H), 5.11 (m, 1H), 2.17 (d, J
) 7.4 Hz, 2H), 1.40 (m, 1H), 1.34-1.14 (m, 8H), 1.03 (s, 6H),
0.87 (t, J ) 6.0 Hz, 3H), -0.05 (s, 9H). 13C NMR: δ 206.8,
136.6, 122.1, 46.5, 41.2, 33.7, 32.0, 29.4, 29.0, 23.0, 21.6, 21.5,
14.5, -2.8. Anal. Calcd for C16H32OSi: C, 71.56; H, 12.01.
Found: C, 71.34; H, 12.09.
Gen er a l P r oced u r e for Ca ta lytic Allylic Su bstitu tion
w ith Cyclop en ta d ien yl-sod iu m (En tr ies 17 a n d 18). A
mixture of acetate 1 (0.37 mmol), Pd(dba)2 (0.018 mmol, 5 mol
%, 0.011 g), and dppe (0.018 mmol, 0.007 g) in THF (0.9 mL)
was treated with cyclopentadienyl-sodium in THF (4 mL),
prepared from freshly distilled cyclopentadiene (1.1 mmol,
0.073 g) and NaH (0.74 mmol, 0.03 g). After this reaction
mixture was stirred at rt for 2 h, it was diluted with ether (20
mL) and then washed with saturated NaHCO3 solution and
brine. The organic layer was dried over Na2SO4, concentrated,
and chromatographed (pentane) to afford the air-sensitive
product as a colorless oil.
(E)-1- a n d 2-(4′-(Dim et h ylp h en ylsilyl)b u t -2′-en yl)-
cyclop en ta -1,3-d ien e (4a ). Formed as a 1:1 mixture. 1H
NMR: δ 7.53 (m, 2H), 7.37 (m, 3H), 6.42, 6.26, 6.11, 5.97 (m,
3H), 5.54-5.34 (br m, 2H), 3.08 (dd, J ) 15.0 and 6.7 Hz, 2H),
2.97 (s, 1H), 2.90 (s, 1H), 1.72 (dd, J ) 7.8 and 2.8 Hz, 2H),
0.30 (s, 6H). 13C NMR: δ 149.2, 146.4, 134.9, 133.9, 132.7,
131.1, 126.8, 126.4, 139.1, 133.9, 129.1, 128.0, 133.9, 127.9,
127.3, 127.2, 43.7, 41.7, 34.8, 33.9, 22.2, 22.1, -2.7, -2.8. MS
(EI): m/z (rel intens) 254 (M+, 46), 150 (6), 135 (100), 119 (3),
91 (9), 75 (23).
(E)-2-Met h yl-2-(4′-(d im et h ylp h en ylsilyl)b u t -2′-en yl)-
cycloh exa n on e (3b). 1H NMR: δ 7.49 (m, 2H), 7.34 (m, 3H),
5.39 (m, 1H), 5.14 (m, 1H), 2.47-2.01 (br m, 4H), 1.90-1.56
(m, 1H), 1.67 (d, J ) 7.8 Hz, 2H), 1.54-1.38 (br m, 2H, H5),
1.37-1.22 (m, 1H), 0.99 (s, 3H), 0.26 (s, 6H). 13C NMR: δ 216.0,
133.9, 129.3, 128.1, 129.6, 124.2, 41.1, 39.2, 38.8, 27.7, 23.0,
22.3, 21.4, -3.0. Anal. Calcd for C19H28OSi: C, 75.94; H, 9.39.
Found: C, 75.71; H, 9.56.
(E)-4-Meth yl-8-(dim eth ylph en ylsilyl)oct-6-en -3-on e (3c).
1H NMR: δ 7.49 (m, 2H), 7.34 (m, 3H), 5.40 (m, 1H), 5.15 (m,
1H), 2.49 (m, 1H), 2.40 (dq, J ) 7.5 and 3.3 Hz, 2H), 2.29 (m,
3H,), 2.01 (m, 1H), 1.65 (d, J ) 7.8 Hz, 2H), 1.02 (t, J ) 7.1
Hz, 3H), 1.01 (d, J ) 6.9 Hz, 3H), 0.26 (s, 6H). 13C NMR: δ
215.1, 139.1, 133.9, 129.3, 128.1, 128.5, 126.6, 46.2, 36.7, 34.7,
22.1, 16.4, 8.0, -3.1. Anal. Calcd for C17H26OSi: C, 74.39; H,
9.55. Found: C, 74.06; H, 9.68.
(E)-1- a n d 2-(4′-(Tr im eth ylsila n yl)n on -2′-en yl)cyclo-
p en ta -1,3-d ien e (4b). Formed as a 1:1 mixture. H NMR: δ
1
6.43, 6.27, 6.15, 6.01 (m, 3H), 5.29 (m, 2H), 3.10 (dd, J ) 18.0
and 5.7 Hz, 2H), 2.96 (m, 1H), 2.88 (m, 1H), 1.43 (m, 1H), 1.35-
1.13 (br m, 8H), 0.88 (t, J ) 7.1 Hz, 3H), -0.03 (s, 9H). 13C
NMR: δ 149.6, 146.7, 135.0, 133.4, 132.7, 131.0, 126.7, 126.3,
133.8, 133.1, 126.3, 125.6, 43.7, 41.7, 34.9, 34.0, 33.5, 32.2, 29.5,
29.3, 23.1, 14.6, -2.5, -2.6. MS (EI): m/z (rel intens) 262 (M+,
24), 247 (2), 191 (23), 189 (22), 131 (16), 117 (100), 91 (21), 73
(84).
(E)-2,2-Dim eth yl-6-(dim eth ylph en ylsilyl)h ex-4-en al (3d).
1H NMR: δ 9.42 (s, 1H), 7.49 (m, 2H), 7.35 (m, 3H), 5.42 (m,

Catalytic Transformations Involving Pd Complexes
J . Org. Chem., Vol. 64, No. 26, 1999 9555
Gen er a l P r oced u r e for Ca ta lytic Allylic Su bstitu tion
w ith Disila n es (En tr ies 19-26). Meth od D (See Ta ble 1).
A mixture of the chloride complex 15 (0.025 mmol) and AgOAc
(0.1 mmol, 0.017 g) in CH2Cl2 (0.7 mL) was stirred for 1 h at
0 °C. After the white AgCl precipitate was removed, the yellow
solution was added to the mixture of the corresponding acetate
1 (0.5 mmol) and maleic anhydride (0.075 mmol, 0.0074 g) in
CH2Cl2 solution (4 mL). To this reaction mixture was added
disilane 13 (1.0 mmol) in CH2Cl2 (4 mL) by syringe pump for
3 h at 30 °C; this was followed by stirring for an additional 14
h. After the solvent was removed, the residue was purified by
column chromatography (pentane) to yield the disilylated
product as a colorless oil.
2.74 (d, J ) 11.6 Hz, 1H), 1.29 (dd, J ) 13.2 and 4.0 Hz, 1H),
1.16 (m, 1H), 0.05 (s, 9H). 13C NMR: δ 108.8, 87.5, 57.4, 24.5,
-1.4. MS (CI): m/z (rel intens) 539 (M+, 7), 466 (63), 393 (35),
237 (26), 149 (36), 111 (88), 73 (100).
Bis(µ-ch lor o)bis[(1,2,3-η)-3-m eth yldim eth ylph en ylsilyl]-
p a lla d iu m (15b). 1H NMR: δ 7.48-7.32 (m, 5H), 4.99 (m,
1H), 3.99 (m, 1H), 3.75 (d, J ) 6.6 Hz, 1H), 2.66 (d, J ) 11.7
Hz, 1H), 1.52 (dd, J ) 13.2 and 4.2 Hz, 1H), 1.34 (m, 1H), 0.31
(s, 6H). 13C NMR: δ 137.6, 133.7, 129.6, 128.1, 108.9, 86.3,
57.4, 23.6, -3.0. MS (CI): m/z (rel intens) 661 (M+ - 1, 7),
189 (85), 167 (31), 151 (37), 149 (100), 135 (18), 77 (14), 75
(64).
Bis(µ-ch lor o)bis[(1,2,3-η)-1-m eth yl-3-m eth yltr im eth yl-
silyl]p a lla d iu m (15c). 1H NMR: δ 5.02 (t, J ) 11.3 Hz, 1H),
3.86 (dt, J ) 11.3 and 4.2 Hz, 1H), 3.62 (m, 1H), 1.26 (d, J )
6.3 Hz, 3H), 1.20 (d, J ) 4.5 Hz, 1H), 1.15 (m, 1H), 0.03 (s,
9H). 13C NMR: δ 110.9, 82.1, 75.4, 23.8, 18.2, -1.5. MS (CI):
m/z (rel intens) 565 (M+, 4), 491 (20), 418 (8), 265 (32), 213,
(44), 125 (49), 73(100).
Meth od E is identical to m eth od D, with one exception,
that Pd(OAc)2 (0.025 mmol, 5 mol %, 0.0056 g) was used as
the catalyst.
(E)-1,4-Bis(tr im eth ylsila n yl)n on -2-en e (5e). 1H NMR: δ
5.06 (br m, 2H), 1.43-1.26 (m, 14H), -0.04 (s, 18H). 13C
NMR: δ 130.3, 124.0, 33.4, 32.1, 29.4, 29.3, 23.2, 23.0, 14.5,
-1.5, -2.7. MS (EI): m/z (rel intens) 270 (M+ - 1, 29), 196
(26), 154 (13), 126 (11), 122 (28), 73 (100).
P r ep a r a tion of (η3-Allyl)p a lla d iu m P h osp h in e Com -
p lexes. Chloride complex 15 (0.05 mmol) was dissolved in
CDCl3 (0.2 mL), mixed with AgBF4 (0.11 mmol, 0.022 g) in
the presence of PPh3 (0.22 mmol, 0.058 g) at 0 °C, and then
stirred for 10 min. The AgCl precipitate was removed, and the
resulting yellow solution was used without further purification.
Bis(tr ip h en ylp h osp h in e)[(1,2,3-η)-3-m eth yltr im eth yl-
(E)-1-(Dim eth ylp h en ylsilyl)-4-(tr im eth ylsila n yl)n on -2-
1
en e (5f). H NMR: δ 7.53-7.33 (m, 5H), 1.68 (d, J ) 6.6 Hz,
2H), 1.35-1.25 (m, 9H), 0.89 (t, J ) 6.9 Hz, 3H), 0.27 (s, 6H),
-0.08 (s, 9H). 13C NMR: δ 134.0, 132.6, 129.1, 128.0, 131.2,
123.2, 23.0, 22.2, 14.5, 33.5, 32.1, 29.3, -2.8, -2.9. Anal. Calcd
for C20H36Si2: C, 72.21; H, 10.91. Found: C, 72.03; H, 10.99.
(E)-2,5-Bis(tr im eth ylsila n yl)d ec-3-en e (5g). Formed as
a 1:1 mixture of diastereoisomers from a 1:1 diastereomeric
mixture of 1e. 1H NMR: δ 5.22 (dd, J ) 15.1 and 8.1 Hz, 1H),
4.96 (dd, J ) 15.1 and 9.0 Hz, 1H), 1.55-1.16 (m, 10H) 1.02
(d, J ) 7.2 Hz, 3H), 0.88 (m, 3H), -0.04 (s, 18H). 13C NMR: δ
131.0, 127.8, 127.6, 33.6, 33.5, 32.1, 29.5, 29.4, 29.3, 26.8, 26.7,
23.1, 14.8, 14.7, 14.6, -2.5, -2.8. MS (EI): m/z (rel intens)
284 (M+ - 1, 25), 211 (12), 210 (40), 137 (61), 93 (13), 73(100).
(E)-2-(Dim eth ylp h en ylsilyl)-5-(tr im eth ylsila n yl)d ec-3-
en e (5h ). Formed as a 1:1 mixture of diastereoisomers from a
1:1 diastereomeric mixture of 1e. 1H NMR: δ 5.22 (m, 1H),
4.98 (m, 1H), 1.77 (m, 1H), 1.40-1.16 (m, 9H) 1.04 (d, J ) 7.1
Hz, 3H), 0.90 (m, 3H), 0.25 (s, 6H), -0.07 (s, 9H). 13C NMR:
δ 134.2, 130.5, 128.9, 127.8, 131.4, 128.6, 26.3, 23.2, 15.0, 14.7,
33.6, 32.1, 29.4, 29.3, -2.6, -4.0. HRMS (EI): m/z calcd for
silyl]p a lla d iu m Tetr a flu or obor a te (16a ). 1H NMR:
δ
7.41-7.12 (br m, 30H), 5.54 (m, 1H), 4.51 (m, 1H), 3.41 (m,
1H), 3.24 (m, 1H), 0.92 (t, J ) 13.1 Hz, 1H), 0.21 (m, 1H), -0.23
(s, 9H). 13C NMR: δ 134.3-129.0 (6 Ph), 118.5, 107.9, 72.0,
23.5, -1.7.
Bis(t r ip h e n ylp h osp h in e )[(1,2,3-η)-3-m e t h yld im e t h -
ylp h en ylsilyl]p a lla d iu m Tet r a flu or ob or a t e (16b ). 1H
NMR: δ 7.46-7.08 (br m, 35H), 5.44 (m, 1H), 4.46 (m, 1H),
3.37 (m, 1H), 3.24 (m, 1H), 1.09 (tm, J ) 12.3 Hz, 1H), 0.46
(m, 1H), 0.01 (s, 6H). 13C NMR: δ 136.6-128 (6 Ph), 118.7,
106.2, 71.8, 22.4, -3.6.
Bis(tr ip h en ylp h osp h in e)[(1,2,3-η)-1-m eth yl-3-m eth yl-
t r im et h ylsilyl]p a lla d iu m Tet r a flu or obor a t e (16c). 1H
NMR: δ 7.41-7.12 (br m, 30 H), 5.29 (m, 1H), 4.45 (m, 1H),
4.16 (m, 1H), 0.90 (m, 3H) 0.74 (m, 1H), 0.05 (m, 1H), -0.21
(s, 9H). 13C NMR: δ 134.1-129.0 (6 Ph), 119.1, 99.8, 89.4, 22.5,
16.8, -1.7.
P r ep a r a tion of 20a ,b. The corresponding alkenylmagne-
sium bromide (4.0 mmol) was added dropwise to 3e (2.0 mmol)
in THF (4 mL) at 0 °C. After the reaction mixture was stirred
for 12 h at rt, it was quenched with saturated NH4Cl solution,
diluted with ether (20 mL), and washed with brine. The
organic layer was dried over anhydrous MgSO4 and concen-
trated. The residue was oxidized without further purification
using the procedure described in the literature by Ratcliffe and
Rodehorst.41 After chromatography with pentane-ether, 20:
1, the product was obtained as a colorless oil.
C
21H38Si2 (M+) 346.2512, found 346.2507.
P h en yla tion w ith Sod iu m Tetr a p h en ylbor a te (En tr ies
27-30). A mixture of Pd(dba)2 (0.01 mmol, 2 mol %, 0.006 g),
PPh3 (0.02 mmol, 0.005 g), and the acetate 1 (0.5 mmol) in
THF (0.5 mL) was stirred for 15 min at 25 °C. This mixture
was transferred via cannula to the THF solution (0.5 mL) of
NaBPh4 (0.27 mmol, 0.094 g), and the resulting reaction
mixture was stirred for 3 h at 60 °C. The final product was
obtained by chromatography (pentane) as a colorless oil.
(E)-4-(Tr im et h ylsila n yl)-1-p h en yln on -2-en e (6e). 1H
NMR: δ 7.33-7.17 (m, 5H), 5.42-5.20 (m, 2H), 3.35 (d, J )
5.4 Hz, 2H), 1.53-1.20 (m, 9H), 0.88 (t, J ) 6.6 Hz, 3H), -0.04
(s, 9H). 13C NMR: δ 141.9, 133.7, 128.7, 128.5, 126.6, 125.9,
39.8, 33.5, 32.2, 29.6, 29.3, 23.1, 14.6, -2.5. Anal. Calcd for
(E)-4,4-Dim et h yl-8-(t r im et h ylsila n yl)oct a -1,6-d ien -3-
1
on e (20a ). H NMR: δ 6.79 (dd, J ) 17.0 and 10.2 Hz, 1H),
6.31 (dd, J ) 17.0 and 2.2 Hz, 1H), 5.62 (dd, J ) 10.2 and 2.1
Hz, 1H), 5.41 (m, 1H), 5.09 (m, 1H), 2.18 (d, J ) 7.2 Hz, 2H),
1.39 (d, J ) 8.0 Hz, 2H), 1.11 (s, 6H), -0.04 (s, 9H). 13C NMR:
δ 203.8, 131.3, 130.3, 128.2, 123.6, 47.2, 43.1, 24.1, 23.3, -1.4.
MS (EI): m/z (rel intens) 224 (M+, 7), 209 (5), 170 (75), 155
(40), 127 (100), 75 (36).
C
18H30Si: C, 78.75; H, 11.02. Found: C, 78.57; H, 11.25.
(E)-5-(Tr im eth ylsilan yl)-2-ph en yldec-3-en e (6f). Formed
as a 1:1 mixture of diastereomers from a 1:1 diastereomeric
1
mixture of 1e. H NMR: δ 7.32-7.15 (m, 5H), 5.44-5.23 (m,
2H), 3.44 (m, 1H), 1.45-1.19 (m, 12H), 0.86 (t, J ) 6.6 Hz,
3H), -0.06 (s, 9H). 13C NMR: δ 147.3, 133.2, 128.6, 126.1,
131.0, 127.5, 42.8, 33.3, 32.0, 29.4, 29.1, 23.0, 22.1, 14.5, -2.7.
Anal. Calcd for C19H32Si: C, 79.09; H, 11.18. Found: C, 78.93;
H, 11.35.
(2E,7E)-5,5-Dim eth yl-9-(tr im eth ylsilan yl)n on a-2,7-dien -
1
4-on e (20b). H NMR: δ 6.79 (m, 1H), 6.5 (dq, J ) 15.0 and
1.7 Hz, 1H), 5.41 (m, 1H), 5.09 (m, 1H), 2.20 (dd, J ) 7.2 and
0.9 Hz, 2H), 1.87 (dd, J ) 6.9 and 1.8 Hz, 3H), 1.39 (dd, J )
8.1 and 0.8 Hz, 2H), 1.09 (s, 6H), -0.04 (s, 9H). 13C NMR: δ
204.0, 142.7, 130.2, 126.5, 123.9, 46.8, 43.2, 24.2, 23.2, 18.6,
-1.6. MS (EI): m/z (rel intens) 238 (M+, 5), 170 (61), 127 (78),
75 (31), 73 (100).
P r ep a r a tion of (η3-Allyl)p a lla d iu m Ch lor id es. A mix-
ture of the corresponding acetate (2.68 mmol), Pd(dba)2 (2.40
mmol, 1.38 g), and LiCl (10.72 mmol, 0.45 g) was stirred in
THF (15 mL) for 2 h at room temperature. After the solvent
was removed in vacuo, the residue was purified by chroma-
tography (CH2Cl2), affording the allylpalladium complexes as
yellow crystals.
2,2-Dim eth yl-4-vin ylcycloh exan on e (21a). BF3‚Et2O (0.61
mmol, 0.08 mL) was added to ketone 20 (0.36 mmol) in dry
CH2Cl2 (2.4 mL) at -78 °C, and this mixture was stirred for
Bis(µ-ch lor o)bis[(1,2,3-η)-3-m eth yltr im eth ylsilyl]p a lla -
d iu m (15a ). 1H NMR δ 5.12 (dt, J ) 11.2 and 6.4 Hz, 1H),
4.08 (dt, J ) 11.2 and 4.8 Hz, 1H), 3.82 (d, J ) 6.4 Hz, 1H),
(41) Ratcliffe, R.; Rodehorst, R. J . Org. Chem. 1970, 35, 400.

9556 J . Org. Chem., Vol. 64, No. 26, 1999
Macsa´ri et al.
1
20 min. Subsequently, the reaction mixture was allowed to
warm to rt, and then it was stirred for an additional 3 h. After
the resultant yellow solution was diluted with CH2Cl2 (10 mL)
and extracted with saturated NaHCO3 solution and brine, it
was dried over anhydrous MgSO4 and concentrated. Column
chromatography with pentane-ether, 9:1, afforded the product
ether, 20:1) as a colorless oil. H NMR: δ 5.81 (m, 1H), 5.07
(m, 1H), 5.03 (dt, J ) 11.2 and 1.3 Hz, 1H), 2.86 (m, 1H), 2.85
(dd, J ) 13.9 and 5.8 Hz, 1H), 2.31 (m, 1H), 2.11 (dd, J ) 13.9
and 2.7 Hz, 1H), 1.70 (q, J ) 13.6 Hz, 1H), 1.62 (ddd, J )
13.6, 4.4 and 1.5 Hz, 1H), 1.21 (s, 3H), 1.06 (s, 3H), 0.79 (d, J
) 6.8 Hz, 3H). 13C NMR: δ 216.1, 141.1, 114.8, 45.5, 45.0, 40.4,
39.9, 36.6, 26.6, 25.8, 13.7. MS (EI): m/z (rel intens) 166 (M+,
10), 151 (12), 138 (34), 110 (23), 95 (61), 81 (31), 67 (100).
1
as a colorless oil. H NMR: δ 5.75 (m, 1H), 5.05 (dt, J ) 17.0
and 1.5 Hz, 1H), 4.97 (dt, J ) 10.0 and 1.5 Hz, 1H), 2.62 (m,
2H), 2.28 (ddd, J ) 14.8, 4.7 and 2.6 Hz, 1H), 2.05 (m, 1H),
1.78 (dt, J ) 13.6 and 4.4 Hz, 1H), 1.52 (ddd, J ) 12.8, 4.5
and 1.6 Hz, 1H), 1.46 (q, J ) 13 Hz, 1H), 1.21 (s, 3H), 1.05 (s,
3H). 13C NMR: δ 215.9, 142.3, 113.7, 47.1, 45.1, 37.8, 37.0,
33.2, 26.1, 25.8. MS (EI): m/z (rel intens) 152 (M+, 17), 123
(6), 110 (17), 97 (50), 79 (59), 67 (100).
Ack n ow led gm en t. This work was supported by the
Swedish Natural Science Research Council (NFR). E.H.
expresses his thanks for a bilateral scholarship from the
Swedish Institute (SI).
2,2,5-Tr im eth yl-4-vin ylcycloh exa n on e (21b). TiCl4 (0.21
mmol, 0.21 mL, 1.0 M solution in CH2Cl2) was added slowly
to the CH2Cl2 (1.0 mL) solution of ketone 20 (0.18 mmol) at
-78 °C, and this mixture was stirred for 30 min. After the
red reaction mixture was diluted with ether (15 mL), it was
extracted with brine and then dried over anhydrous MgSO4.
The product was obtained after chromatography (pentane-
Su p p or tin g In for m a tion Ava ila ble: NMR peak assign-
ment and determination of the stereochemistry for 21b and
13C NMR spectra for compounds 2a , 2c,d , 3c, 3e,f, 3i, 4a , 5e,
5g,h , 6e, 15a -c, 20a ,b, and 21a ,b. This material is available
free of charge via the Internet at http://pubs.acs.org.
J O991172L