Regiocontrol in palladium-catalysed allylic alkylation by addition of lithium iodide

Article abstract of DOI:10.1039/a707652e

Regioselectivity in the palladium-catalysed allylic alkylation of 1-arylprop-2-enyl acetates [ArCH(OAc)CH=CH₂] or (E)-3-phenylprop-2-enyI acetate (PhCH=CHCH₂OAc) with sodium enolates of soft carbon nucleophiles is controlled by addition of a catalytic amount of lithium iodide to give lienar products [(E)-ArCH=CHCH₂Nu] exclusively; their branch isomers [ArCH(Nu)CH=CH₂] were not detected.

Full text of DOI:10.1039/a707652e

View Article Online / Journal Homepage / Table of Contents for this issue
Regiocontrol in palladium-catalysed allylic alkylation by addition of lithium
iodide
Motoi Kawatsura, Yasuhiro Uozumi and Tamio Hayashi*
Department of Chemistry, Faculty of Science, Kyoto University, Sakyo, Kyoto 606-01, Japan
Regioselectivity in the palladium-catalysed allylic alkylation
of 1-arylprop-2-enyl acetates [ArCH(OAc)CHNCH2] or
(E)-3-phenylprop-2-enyl acetate (PhCHNCHCH2OAc) with
sodium enolates of soft carbon nucleophiles is controlled by
addition of a catalytic amount of lithium iodide to give lienar
products [(E)-ArCHNCHCH2Nu] exclusively; their branch
isomers [ArCH(Nu)CHNCH2] were not detected.
carried out in the presence of 10 mol% of lithium iodide gave a
quantitative yield of linear isomer 5a with 100% regioselectiv-
ity (entry 2). The regioselectivity was not strongly affected by
the addition of lithium fluoride, chloride or bromide, branch
isomer 6a being formed with about 20% regioselectivity
(entries 3–5). High linear selectivity was also observed in the
reaction in the presence of sodium iodide (entry 6), indicating
that the iodide anion is important for control of the re-
gioselectivity. The amount of lithium iodide additive can be
decreased to 2 mol%, which is the same amount as that of the
palladium catalyst (entry 7). Use of the palladium catalyst
generated from [PdI(p-C3H5)]2 instead of [PdCl(p-C3H5)]2
showed the same linear selectivity in the absence of additional
iodide anion (entry 8).
In synthetic organic chemistry, palladium-catalysed allylic
alkylation of allyl esters is a useful reaction for the formation of
carbon–carbon bonds.1 One of the challenging problems in
catalytic allylic alkylation is control of the regiochemistry in the
reaction that proceeds through unsymmetrically substituted
p-allylpalladium
intermediates.
For
example,
the
p-allylpalladium complex containing one substituent at the C-1
position usually produces both linear and branch isomers, the
ratio being dependent on the substituents, nucleophiles and
reaction conditions.1–3 Here we report exclusive formation of
the linear isomer in a palladium-catalysed allylic alkylation,
which is realized by addition of a catalytic amount of lithium
iodide.4
The selectivity in forming the linear isomer in the presence of
lithium iodide was also observed in the reaction of 1-arylprop-
2-enyl acetates 2 and 3 and 3-phenylprop-2-enyl acetate 4 with
the sodium salt of dimethyl methylmalonate (entries 9–13) and
Ar
Ar
[Pd]/P
In the presence of 2 mol% of the palladium catalyst generated
by mixing [PdCl(p-C3H5)]2 with PPh3 (2 equiv. to Pd), allyl
acetates 1–4 were allowed to react with soft carbon nucleophiles
in THF at 0 °C (Scheme 1). The regioselectivity of the reaction
was found to be dramatically changed by the addition of lithium
iodide. Thus, the reaction of 1-phenylprop-2-enyl acetate 1 with
the sodium salt of dimethyl methylmalonate in the absence of
lithium iodide gave a 96% yield of alkylation product;
consisting of linear isomer (E)-5a and branch isomer 6a in a
ratio of 77:23 (entry 1, Table 1). On the other hand, the reaction
Ar
Nu
+
Na–Nu
(LiX)
OAc
Nu
6a–e (branch)
1 Ar = Ph
2 Ar = 4-MeOC6H4
3 Ar = 4-ClC6H4
5a–e (linear)
a Ar = Ph, Nu = CMe(CO2Me)2
b Ar = 4-MeOC6H4, Nu = CMe(CO2Me2)2
c Ar = 4-ClC6H4, Nu = CMe(CO2Me)2
d Ar = Ph, Nu = CH(CO2Me)2
Ph
OAc
4
e Ar = Ph, Nu = CH(COMe)CO2Me
Scheme 1
Table 1 Effects of lithium salts on allylic alkylation of allyl acetates 1–4 catalysed by palladium–phosphine complexesa
Allyl
ester
Phosphine
ligandb
Yield (%)c
5 + 6
Ratiod
5:6
Entry
MX/equiv.
Nu
t/h
1
2
3
4
5
1
1
1
1
1
1
1
1
2
2
3
3
4
1
1
1
1
1
1
none
PPh3
PPh3
PPh3
PPh3
PPh3
PPh3
PPh3
PPh3
PPh3
PPh3
PPh3
PPh3
PPh3
PPh3
PPh3
PPh3
PPh3
dppe
dppe
CMe(CO2Me)2
CMe(CO2Me)2
CMe(CO2Me)2
CMe(CO2Me)2
CMe(CO2Me)2
CMe(CO2Me)2
CMe(CO2Me)2
CMe(CO2Me)2
CMe(CO2Me)2
CMe(CO2Me)2
CMe(CO2Me)2
CMe(CO2Me)2
CMe(CO2Me)2
CH(CO2Me)2
12
24
24
24
24
24
24
24
12
24
12
24
24
12
12
12
12
12
12
96
99
89
96
90
93
97
96
98
92
96
89
99
94
91
95
89
92
61
77:23
100:0
78:22
82:18
80:20
98:2
100:0
100:0
72:28
100:0
86:14
100:0
100:0
83:17
100:0
92:8
LiI (0.1)
LiF (0.1)
LiCl (0.1)
LiBr (0.1)
NaI (0.1)
LiI (0.02)
none
none
LiI (0.1)
none
LiI (0.1)
LiI (0.1)
none
LiI (0.1)
none
LiI (0.1)
none
LiI (0.1)
6
7
8e
9
10
11
12
13f
14
15
16
17
18f
19f
CH(CO2Me)2
CH(COMe)CO2Me
CH(COMe)CO2Me
CMe(CO2Me)2
CMe(CO2Me)2
100:0
89:11
89:11
a All reactions were carried out in THF under nitrogen: THF (1.0 ml), allylic acetate (0.20 mmol), NaNu (0.40 mmol), [PdCl(p-C3H5)]2 (0.002 mmol) and
phosphine ligand at 0 °C. b The ratio of Pd:Phosphine = 1:2. c Isolated yield by silica gel column chromatography. d The ratio was determined by 1H NMR
analysis of the products. e [PdI(p-C3H5)]2 was used. f Carried out at 20 °C.
Chem. Commun., 1998
217

View Article Online
2
field than that for 7 by 6.5 ppm and the chemical shift for C-1
of 8 appears at higher field than that for 7 by 3.1 ppm. The
difference in the chemical shifts may support the idea that the
C-3 carbon of the iodide complex 8 undergoes the nucleophilic
attack giving linear isomer 5a more selectively than that of
7.6
Ph
1
3
Ph
Na–Nu
Ph
Nu
+
Pd
THF, 0 °C
Nu
6a
X
PPh3
5a
7 X = Cl
8 X = I
X = Cl 5a:6a = 80:20
X = I 5a:6a = 100:0
This work was supported by the ‘Research for the Future’
Program, the Japan Society for the Promotion of Science.
Nu = CMe(CO2Me)2
Scheme 2
in the reactions with dimethyl malonate and methyl acetoacetate
(entries 14–17).
Footnotes and References
* E-mail: thayashi@th1.orgchem.kychem.kyoto-u.ac.jp
It is noteworthy that the addition of lithium iodide is not
effective for the reaction catalysed by a palladium–bisphosph-
ine complex. Thus, the reaction of 1 with dimethyl methyl-
malonate in the presence of a palladium catalyst prepared from
[PdCl(p-C3H5)]2 and 1,2-bis(diphenylphosphino)ethane (dppe)
gave a mixture of 5a and 5b in a ratio of 89:11, irrespective of
the addition of lithium iodide (entries 18 and 19). These results
suggest that, in the reaction catalysed by triphenylphosphine–
palladium, the iodide coordinates to the p-allylpalladium
intermediate to form the intermediate PdI(p-allyl)(PPh3), and
that the iodide on the palladium controls the regioselectivity of
the nucleophilic attack. Using the dppe ligand, the intermediate
will be the cationic complex [Pd(p-allyl)(dppe)]+X2, where
iodide is not directly bonded to palladium.
† Selected data for 7: dH(CDCl3, 240 °C) 2.88 (d, J 6.8, 1 H, syn-H on C-3),
2.97 (d, J 11.7, 1 H, anti-H on C-3), 5.37 (dd, JH–H 13.2, JH–P 10.1, 1 H, H
on C-1), 6.08 (ddd, J 6.8, 11.7, 13.2, 1 H, H on C-2), 7.36–7.88 (m, 20 H);
dP(CDCl3, 240 °C) 24.2 (s); dC(CDCl3, 240 °C) 58.2 (C-3), 99.6 (JC–P
26.9, C-1), 111.4 (JC–P 5.2, C-2). For 8: dH(CDCl3, 240 °C) 3.47 (d, J 6.8,
1 H, syn-H on C-3), 3.14 (d, J 12.2, 1 H, anti-H on C-3), 5.21 (dd, JH–H 13.0,
JH–P 10.5, 1 H, H on C-1), 6.08 (ddd, J 6.8, 12.2, 13.0, 1 H, H on C-2),
7.26–7.63 (m, 20 H); dP(CDCl3, 240 °C) 27.9 (s); dC(CDCl3, 240 °C) 64.7
(C-3), 96.5 (JC–P 29.0, C-1), 111.0 (JC–P 5.2, C-2).
1 For reviews on catalytic allylic substitutions: J. Tsuji, Palladium
Reagents and Catalyst, Wiley, Chichester, 1995, pp. 290–340; J. A.
Davies, in Comprehensive Organometallic Chemistry II, ed. E. W. Abel,
F. G. A. Stone and G. Wilkinson, Pergamon, Oxford, 1955, vol. 9, ch. 6;
S. A. Godleski, in Comprehensive Organic Synthesis, ed. B. M. Trost and
I. Fleming, Pergamon, Oxford, 1991, vol. 4, p. 585; G. C. Frost,
J. Holwarth and J. M. J. Williams, Tetrahedron: Asymmetry, 1992, 3,
1089; G. Consiglio and M. Waymouth, Chem. Rev., 1989, 89, 257.
2 For examples: M. P. T. Sjogren, S. Hansson, B. Åkermark and
A. Vitagliano, Organometallics, 1994, 13, 1963; T. Hayashi, K. Kishi,
A. Yamamoto and Y. Ito, Tetrahedron Lett., 1990, 31, 1743; B. M. Trost,
M. J. Krische, R. Radinov and G. Zanoni, J. Am. Chem. Soc., 1996, 118,
6297.
Palladium complex PdCl[p-(1-phenyl)allyl](PPh3) 7 and its
iodide analogue 8 were prepared by mixing [PdX[p-(1-phenyl)-
allyl]]2 (X = Cl and I)5 with PPh3 (1 equiv. to Pd) and were
1
characterized by 31P, H and 13C NMR spectroscopy.† Both
have the substituted carbon (C-1) of the p-allyl trans to the
phosphorus atom of PPh3 and the unsubstituted carbon (C-3) cis
to phosphorus, as determined by the large coupling constants
(J = 10.1 Hz in 7 and 10.5 Hz in 8) between the C-1 proton and
phosphorus, and no coupling between the C-3 protons and
phosphorus. Stoichiometric reaction of chloride complex 7 with
the sodium enolate of dimethyl methylmalonate in THF at 0 °C
gave 5a and 6a in a ratio of 80:20, while the reaction of iodide
complex 8 gave 5a with 100% regioselectivity (Scheme 2).
These selectivities are in good agreement with those observed in
the catalytic reactions, demonstrating that the iodide ligand
bonded to the p-allylpalladium intermediate controls the
regioselectivity. Comparing the 13C NMR spectra of 7 and 8, the
chemical shift for C-3 of the p-allyl group of 8 appears at lower
3 T. Hayashi, M. Kawatsura and Y. Uozumi, Chem. Commun., 1997,
561.
4 An anion effect on the enantioselectivity of palladium-catalysed
asymmetric allylic amination has been reported: U. Burckhardt, M.
Baumann and A. Togni, Tetrahedron: Asymmetry, 1997, 8, 155.
5 P. R. Auburn, P. B. Mackenzie and B. Bosnich, J. Am. Chem. Soc., 1985,
107, 2033. [PdI(p-C3H5)]2 was prepared by mixing [PdCl(p-C3H5)]2
with LiI in THF.
6 B. Åkermark, B. Krakenberger, S. Hansson and A. Vitagliano, Organo-
metallics, 1987, 6, 620.
Received in Cambridge, UK. 23rd October 1997; 7/07652E
218
Chem. Commun., 1998