Palladium-catalyzed elimination reaction of acyclic (E)-allylic acetates: The stereochemistry elucidated by syn-effect

Article abstract of DOI:10.1246/cl.2005.256

The stereochemistry of the elimination reaction of acyclic (E)-allylic acetates into the corresponding 1,3-dienes catalyzed by palladium in the presence of a base was investigated. The Z/E ratios of the resulting 1,3-dienes varied with the δ-substituents of the starting allylic acetates. This phenomenon was discussed based on the concept of a syn-effect, which is most primarily rationalized by a σ → π* interaction. Copyright

Full text of DOI:10.1246/cl.2005.256

2
56
Chemistry Letters Vol.34, No.2 (2005)
Palladium-Catalyzed Elimination Reaction of Acyclic (E)-Allylic Acetates:
The Stereochemistry Elucidated by ‘‘Syn-Effect’’
ꢀ
ꢀ
Hiroyuki Takenaka, Yutaka Ukaji, and Katsuhiko Inomata
Division of Material Sciences, Graduate School of Natural Science and Technology, Kanazawa University,
Kakuma, Kanazawa 920-1192
(Received October 18, 2004; CL-041226)
OAc
[Pd(0)] (0.05 equiv.)
Base (3.0 equiv.)
The stereochemistry of the elimination reaction of acyclic
E)-allylic acetates into the corresponding 1,3-dienes catalyzed
γ
γ
α
β
β
CH3CH2
δ
δ
(
Ph
α
Ph
by palladium in the presence of a base was investigated. The
Z=E ratios of the resulting 1,3-dienes varied with the ꢀ-substitu-
ents of the starting allylic acetates. This phenomenon was
discussed based on the concept of a ‘‘syn-effect,’’ which is most
Solvent, rt, t h
CH CH2
3
2a
1
a
Table 1. Elimination reaction of (E)-allylic acetate 1a into 2a
Z=Ea
Entry
[Pd(0)]
Base
Solvent t/h Yield/%
ꢀ
primarily rationalized by a ꢁ ! ꢂ interaction.
1
2
3
4
5
6
7
8
9
[Pd(PPh3)4]
[Pd(dppb)2]
[Pd(dppp)2]
[Pd(dppe)2]
[Pd(dppe)2]
[Pd(dppe)2]
DBU
DBU
DBU
DBU
Et3N
THF
THF
THF
THF
THF
THF
THF
Et2O
90
8
22
6
20
20
68
19
42
61
79
87
71
77
52
71
26
11/89
23/77
38/62
57/43
46/54
48/52
27/73
42/58
18/82
The palladium-catalyzed elimination reaction of allylic
compounds is a useful method for the preparation of 1,3-dienes
1
as versatile synthetic intermediates. In the case of allylic ace-
i
Pr2NEt
tate, it is generally thought that allylpalladium complexes are in-
itially formed by oxidative addition of Pd(0) followed by ꢃ-
elimination of HPdOAc from the ꢂ-allylpalladium complexes
to regenerate Pd(0) with liberation of HOAc. The elimination
is often carried out in the presence of a suitable base to capture
HOAc. The pathway of the elimination is reported to be varied
[Pd(dppe)2] TMEDA
[Pd(dppe)2]
[Pd(dppe)2]
DBU
DBU
toluene 68
a
1
The ratios were determined by 400 MHz H NMR spectra. Only the
6
stereochemistry of Cꢅ–C double bond is shown.
ꢀ
2
depending on the employed conditions.
[
Pd(dppe)2] (0.05 equiv.)
DBU (3.0 equiv.)
OAc
Previously we investigated the stereochemistry of the desul-
fonylation reaction of ꢄ,ꢄ-dialkylated (E)-allylic sulfones with a
base and found that the sterically unfavorable (Z)-dienes were
Ph
2a (Z/E = 54/46)
CH CH2
3
Ph
CH3CH2
THF, rt, 6 h, 63%
3
3
predominantly formed. The result was rationalized by ‘‘confor-
4
The reaction pathway of the present elimination could be
elucidated as follows (Scheme 1). When the reaction was carried
out using palladium complex with monodentate ligand PPh3
(Entry 1) or in less polar, namely less coordinatable, solvents
(Entries 8, 9), syn-ꢃ-elimination via ꢁ-allylpalladium complex
F might proceed predominantly without participation of the base
to give (E)-diene selectively, because of facile generation of a
vacant coordination site on the Pd(II) center. Such an agostic in-
teraction might be difficult for the palladium species coordinated
by a bidentate ligand with small bite angle to avoid syn-ꢃ-hy-
,5
mational acidity’’ that essentially implies a ‘‘syn-effect.’’ We
proposed that the ‘‘syn-effect’’ is primarily caused by a ꢁ !
ꢀ
3,5b–e
ꢂ interaction.
Herein, we described that the sterically un-
favorable (Z)-dienes were also produced in the elimination reac-
tion of acyclic (E)-allylic acetates catalyzed by palladium under
the specific conditions utilizing a base, and the stereochemistry
was elucidated by the concept of the ‘‘syn-effect.’’
First the palladium-catalyzed elimination reaction of ꢀ-ethyl
substituted (E)-allylic acetates 1a was examined under various
conditions, and the results are summarized in Table 1. When
8
dride elimination (Entries 2–4). The influence of the base on
[
THF at rt, the elimination proceeded to give the 1,3-diene 2a
Pd(PPh3)4] was used as a catalyst in the presence of DBU in
the proton elimination would contrast with the syn-ꢃ-elimination
(Entries 4–7). Unprecedented Z-preference using DBU (Entry 4)
might be ascribed to the participation of the base in the transition
6
in the preference of (E)-isomer (Entry 1). By the use of biden-
tate ligands, Z-selectivities were increased (Entries 2–4). Among
9
state such as D.
ꢁ
the bidentate ligands, dppe, whose bite angle 86 is smaller than
7
As mentioned above, we previously found that the sterically
unfavorable (Z)-dienes were predominantly formed in the desul-
fonylation reaction of (E)-allylic sulfones with the base and the
result was rationalized by the ‘‘conformational acidity’’ that es-
others, showed Z-preference (Z=E ¼ 57=43) (Entry 4). It was
confirmed that in the absence of a palladium catalyst, no elimi-
nation proceeds even in refluxing THF. Among the bases exam-
ined, DBU realized higher Z-selectivity (Entries 4–7). In less
polar solvents, Et2O and toluene, Z-selectivity was decreased
3
sentially implies the ‘‘syn-effect.’’ In the present elimination,
Z-preference might be also explained based on the concept of
0
(
Entries 8, 9).
A regioisomeric (E)-allylic acetate 3 was also subjected to
the ‘‘syn-effect’’ in E2 -elimination from ꢁ-complex D. That
is, at the transition state of deprotonation, the eclipsed conforma-
tions G and H might be predominant due to hyperconjugation of
the developing anion generated by the interaction of Hꢀ(s) with
base(s), in both of which the developing anion is aligned with the
the elimination reaction under the same conditions as those in
Entry 4. The corresponding 1,3-diene 2a was obtained with al-
most the same Z=E ratio of 54/46. This result suggested that
the reaction proceeded via the same ꢂ-allylpalladium intermedi-
ate (B, Scheme 1, R = CH3CH2).
ꢀ
ꢂ -orbital and other conformations could be neglected. Between
the conformations G and H, the CC eclipsed syn-conformation
Copyright Ó 2005 The Chemical Society of Japan

Chemistry Letters Vol.34, No.2 (2005)
257
Pd+Ln' OAc
PdL
H
PdLn
Ph
n
BnO{ > BnS{ > CH3{ > CH3CH2{
> (CH3)2CH{ > Ph{ > (CH3)3C{
H
H
H
H
H
Ph
Ph
H
H
H
H
Ph
R
H
H
H
OAc
R
R
In the case of ꢀ-benzyloxy substituted (E)-allylic acetate 1f
AcO
A
B
C
(
R ¼ BnO), CH eclipsed form H is unfavorable due to low donor
OAc
11
Base
ability of C–O bond. Thus exclusive formation of (Z)-2f via
conformation G might have been observed. The bulkiness of
Ph and (CH3)3C groups might exclude the conformation G to
+
Pd Ln'
OAc
H
OAc
H
L Pd
Ln'Pd
n'
H
H
Ph
H
H
H
Ph
H
H
H
R
H
H
H
12,13
give (E)-dienes 2d, 2e selectively.
R
R
H
Base
H
anti-
D
E
F
In conclusion, the stereochemical outcome in the elimina-
tion reaction of acyclic (E)-allylic acetates to the corresponding
dienes by the use of [Pd(dppe)2] as a catalyst in the presence of
E2'-elimination
β
-elimination
syn-
β
-elimination
H
H
H
H
R
Ph
Ph
0
DBU was elucidated by E2 -elimination, and unprecedented
Z-preference could be well rationalized by the ‘‘syn-effect’’ in
the transition state of deprotonation, which arose from a ꢁ !
H
H
H
R
H
(
Z)-2
(E)-2
ꢀ
ꢂ interaction. It is noteworthy that the highest Z-selectivity
was observed for the benzyloxy substituent among the examined
substrates.
Scheme 1.
*
+
*
+
π
π
π
Pd Ln'
Pd Ln'
π
H
H
References and Notes
σ
H
σ
H
Ph
Base
Base
H
Base
H
α
α
1
a) J. Tsuji, ‘‘Palladium Reagents and Catalysts,’’ John Wiley & Sons, Ltd., Chi-
chester (1995), pp 356–363. b) I. Shimizu, ‘‘Handbook of Organopalladium
Chemistry for Organic Synthesis,’’ ed. by E. Negishi, John Wiley & Sons,
Inc., New York (2002), Vol. 2, Chap. V. 2.5.1.
The reaction pathway of elimination reaction of cyclic allylic compounds to the
corresponding 1,3-dienes has been discussed, and not only syn-ꢃ-elimination but
also anti-elimination was proposed to be taken place, especially in the presence
of a base. a) E. Keinan, S. Kumar, V. Dangur, and J. Vaya, J. Am. Chem. Soc.,
116, 11151 (1994). b) P. G. Andersson and S. Schab, Organometallics, 14, 1
(1995). c) J. M. Takacs, E. C. Lawson, and F. Clement, J. Am. Chem. Soc.,
Ph
δ
γ
β
H
δ
γ
H
β
H
σ
R
σ
H
G
R
H
2
G might be preferred to CH eclipsed form H, because hypercon-
jugative electron donation by C–H bond is more effective than
ꢀ
that by C–C bond.10 If this is the case of the present elimination,
the degree of the ‘‘syn-effect,’’ which depends on the ꢀ-substitu-
ents R of (E)-allylic acetates 1, might be similar to that of the
previous desulfonylation reaction of allylic sulfones. Then the
elimination reaction of various (E)-allylic acetates was exam-
ined by the use of [Pd(dppe)2] in the presence of DBU in THF
and the results are summarized in Table 2.
1
19, 5956 (1997).
3
4
5
A. Shibayama, T. Nakamura, T. Asada, T. Shintani, Y. Ukaji, H. Kinoshita, and
K. Inomata, Bull. Chem. Soc. Jpn., 70, 381 (1997).
The ‘‘syn-effect’’ is herein defined as an effect which stabilizes the syn-confor-
mation against the steric hindrance.
Related studies on the ‘‘syn-effect’’; a) T. Hirata, Y. Sasada, T. Ohtani, T. Asada,
H. Kinoshita, H. Senda, and K. Inomata, Bull. Chem. Soc. Jpn., 65, 75 (1992). b)
T. Nakamura, S. K. Guha, Y. Ohta, D. Abe, Y. Ukaji, and K. Inomata, Bull.
Chem. Soc. Jpn., 75, 2031 (2002). c) S. K. Guha, A. Shibayama, D. Abe, Y.
Ukaji, and K. Inomata, Chem. Lett., 32, 778 (2003). d) S. K. Guha, Y. Ukaji,
and K. Inomata, Chem. Lett., 32, 1158 (2003). e) S. K. Guha, A. Shibayama,
D. Abe, M. Sakaguchi, Y. Ukaji, and K. Inomata, Bull. Chem. Soc. Jpn., 77,
OAc [Pd(dppe) ] (0.05 equiv.)
2
γ
α
γ
β
β
DBU (3.0 equiv.)
δ
R
δ
α
Ph
Ph
2
147 (2004). See also references cited therein.
THF, rt, t h
1
R
2
6
In the present reaction, stereochemistry of Cꢄ–Cꢃ double bond was E in all
cases. Herein, only the stereochemistry of Cꢅ–Cꢀ double bond was discussed.
W. L. Steffen and G. J. Palenik, Inorg. Chem., 15, 2432 (1976).
Table 2. Elimination reaction of (E)-allylic acetates 1 into 2
7
8
ꢁ
ꢁ
7
The angles {Cl–Pd–Cl, P–Pd–P} in the complexes [PdCl2(dppe)] {94 , 86 },
14
a
Entry
R
1
t/h
Yield/%
Z=E
ꢁ
ꢁ
7
ꢁ
ꢁ
[
PdCl2(dppp)] {91 , 91 }, and [PdCl2(dppb)] {90 , 94 } might correlate with
the syn-ꢃ-elimination; that is, the phosphine ligand that forms the complex hav-
ing larger Cl–Pd–Cl and smaller P–Pd–P angles decelerates syn-ꢃ-elimination
because ꢃ-agostic interaction would occur in the geometry in which the angle
1
2
3
4
5
CH3CH2
CH3
a
b
c
d
e
f
6
5
7
10
4
5
87
80
77
60
98
71
89
57/43
64/36
17/83
<1/99
6/94
(CH3)2CH
(CH3)3C
Ph
H–Pd–C is small, ca. 63 based on calculation.
ꢁ
15
9
1
The anti-ꢃ-elimination via E could be ruled out because (E)-diene is anticipated
to be produced selectively due to steric repulsion between substituents.
a) T. Laube and H. U. Stilz, J. Am. Chem. Soc., 109, 5876 (1987). b) T. Laube
and T.-K. Ha, J. Am. Chem. Soc., 110, 5511 (1988). c) P. R. Rablen, R. W.
Hoffmann, D. A. Hrovat, and W. T. Borden, J. Chem. Soc., Perkin Trans. 2,
0
6
7
BnO
BnS
95/5
80/20
g
3
1
11 Y. Apeloig, P. v. R. Schleyer, and J. A. Pople, J. Am. Chem. Soc., 99, 5901
999, 1719.
a
1
The ratios were determined by 400 MHz H NMR spectra. Only
6
(1977).
In the cases of 1a, 1b, 1f, 1g, it is also possible to stabilize the syn-conformation
the stereochemistry of Cꢅ–C double bond is shown.
ꢀ
1
2
0
at the transition state of E2 -elimination by 6ꢂ-electron homoaromaticity involv-
As expected, Z-selectivity with respect to the ꢀ-alkyl sub-
stituents decreased along with their bulkiness; CH3– >
CH3CH2– > (CH3)2CH– > (CH3)3C– (Entries 1–4). In the
case of ꢀ-Ph substituent, high E-selectivity was observed (Entry
ing the developing charge at the ꢀ-position and a pseudo p-orbital of the neigh-
0
boring CH2 (R ¼ CH2R ), or a lone pair of electrons in a p-orbital of the neigh-
0
3,5
boring hetero atom X (R ¼ XR ), respectively.
3
,5
1
3
The order of BnS group varied depending on the reactions. This might be re-
ꢀ
ꢀ
lated to the energy level of ꢂ in the substrates, because ꢁC{S ! ꢂ interaction
5
). ꢀ-BnO group showed the highest Z-selectivity (Entry 6),
could also work in the transition state; see Ref. 5e and references cited therein.
ꢀ
12
When the energy level of ꢂ is not close enough to that of ꢁC{S, 6ꢂ-electron
homoaromaticity might relatively more contribute to the ‘‘syn-effect’’ of the
while ꢀ-BnS substituted acetate 1g afforded rather high Z-pref-
erence of diene 2g (Entry 7). The relative degree of the ‘‘syn-ef-
fect’’ depending on the ꢀ-substituents R of (E)-allylic acetates 1
for elimination reaction was similar to our previous observation
BnS group.
1
4
5
V. D. Makhaev, Z. M. Dzhabieva, S. V. Konovalikhin, O. A. D’yachenko, and
G. P. Belov, Russ. J. Coord. Chem., 22, 563 (1996).
N. Koga, S. Obara, K. Kitaura, and K. Morokuma, J. Am. Chem. Soc., 107, 7109
(1985).
1
3
on the desulfonylation reaction of allylic sulfones as follows;
Published on the web (Advance View) January 22, 2005; DOI 10.1246/cl.2005.256