Enantioselective allylic substitution catalyzed by an iridium complex: Remarkable effects of the counter cation

Article abstract of DOI:10.1039/a907680h

Allylic substitution of a methyl carbonate of cinnamyl alcohol with the anion of dimethyl malonate gave the branched olefin with high enantioselectivity in the presence of (S)-1,1'-binaphthyl-2,2'-diyl phenyl phosphite when a combination of lithium and zinc was used as counter cation.

Full text of DOI:10.1039/a907680h

Enantioselective allylic substitution catalyzed by an iridium complex:
remarkable effects of the counter cation
a
a
b
a
Kaoru Fuji,* Naosumi Kinoshita, † Kiyoshi Tanaka and Takeo Kawabata
a
Institute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011, Japan. E-mail: fuji@scl.kyoto-v.ac.jp
School of Pharmaceutical Sciences, University of Shizuoka, 52-1 Yada, Shizuoka 422-8526, Japan
b
Received (in Cambridge, UK) 22nd September 1999, Accepted 4th October 1999
Allylic substitution of a methyl carbonate of cinnamyl
alcohol with the anion of dimethyl malonate gave the
branched olefin with high enantioselectivity in the presence
of (S)-1,1A-binaphthyl-2,2A-diyl phenyl phosphite when a
combination of lithium and zinc was used as counter
cation.
alkylation with soft carbon nucleophiles catalyzed by an iridium
1
4
5
6
complex. Chiral phosphites 1, 2 and 3 have been used as
ligands. Diethylzinc was used as a base because it is known to
give a greater ee than other bases in allylic alkylations catalyzed
7
by a palladium–(R)-BINAP complex. The results are shown in
Table 1. Both electron-releasing and -withdrawing groups on
the phenyl rings of the ligands gave lower ees (entries 2 and 3)
Unlike palladium-catalyzed allylic alkylations, iridium-cata-
lyzed alkylations take place at the more substituted allylic
2 2
than the parent phosphite (S)-1 (entry 1). CH Cl is the solvent
of choice (compare entry 1 with entries 4–6). Allylic methyl
carbonate 4b gave a better yield than the corresponding acetate
4a in a shorter reaction time (entries 1 and 7).
The effect of the counter cation was investigated using 4b,
and the results are shown in Table 2. A conspicuous feature of
these results is that the proportion of (R)-5 to (S)-5 increases as
1
terminus. Recently, asymmetric versions of this reaction have
2
been reported, where chiral phosphinooxazolines or phospho-
3
rous amidites were used as ligands. We report here enantio-
selective allylic substitutions catalyzed by an iridium complex
of chiral aryl phosphite (S)-1.
the ionic character of the malonate anion increases (entries 1–3).
t
Thus, (R)-5 was obtained in 36% ee by using Bu P
4
base, which
is known to generate naked anions.8 Metals, such as lithium
and zinc, that strongly coordinate with anion gave (S)-5 as the
major enantiomer. The intermediacy of a p-allyl–iridium
complex has been suggested for iridium complex-catalyzed
,9
1
allylic alkylation with triphenyl phosphite as ligand. On the
other hand, the involvement of a s-allyl–iridium complex is
3
indicated for the alkylation of secondary allylic acetates.
Although there is an apparent divergence of mechanistic
considerations, the p-allyl–iridium complex is more likely than
the s-complex when aryl phosphites are used as ligands. The
predominant formation of (S)-5 with lithium and/or zinc as the
We chose 4a as a standard substrate to avoid ambiguities
arising from a chiral centre (Scheme 1), although the ester of an
isomeric secondary alcohol, 1-phenylprop-2-en-1-ol, can also
be a substrate for the same reaction. It has been reported that the
iridium-catalyzed allylic substitution of acetates of secondary
alcohols proceeds with 70–80% retention of configuration.
Aryl phosphites are known to be efficient ligands for allylic
3
†
Present address: Pharmaceutical Production Department, Tokushima
Second Factory, Otsuka Pharmaceutical Co., Ltd., Hiraishi Ebisuno 224-18,
Kawauchi-cho, Tokushima 771-0182, Japan.
Scheme 1
Table 1 Ir-catalysed allylic alkylation of 4a and 4b with dimethyl malonate using ZnEt
2
as baseab
Ratio of 5 : 6
Reaction
time/h
Combined yield of
5 and 6 (%)
Ee of
(S)-5 (%)
GLCc
1H NMR
d
Entry
Substrate
Ligand
Solvent
1
2
3
4
5
6
7
8
4a
4a
4a
4a
4a
4a
4b
4b
1
2
3
1
1
1
1
1
CH
CH
CH
toluene
THF
2
2
2
Cl
Cl
Cl
2
2
2
35
94
47
47
91
48
7
58 (64)e
57
92 : 8
92 : 8
94 : 6
82 : 18
84 : 16
86 : 14
96 : 4
—
—
76
61
32
67
20
54
71
55
91 : 9
96 : 4
81 : 19
85 : 15
81 : 19
95 : 5
85 : 15
62
57 (58)
40 (49)
13 (29)
e
e
e
MeCN
CH
2
Cl
2
88
44 (65)
e
THF
68
a
General procedure (entry 1): A solution of (S)-1 (81.7 mg, 0.2 mmol), [Ir(COD)Cl]
dry CH Cl (1 ml) was stirred for 45 min under argon. To this solution was added a solution prepared by stirring dimethyl malonate (133 mg, 1.0 mmol) and
Et Zn (1.0 M in hexane, 1.0 ml) in CH Cl (3 ml) under argon for 1 h at room temperature. After stirring for the indicated time, usual extractive work-up
with EtOAc followed by preparative TLC (EtOAc–hexane 1+6) gave 58 mg of a mixture of 5 and 6. All reactions were performed at room temperature.
2
(33.6 mg, 0.05 mmol) and cinnamyl acetate (83.1 ml, 0.5 mmol) in
2
2
2
2
2
b
c
Determined by GLC on a Chirasil-DEX CB column at 150 °C. d Determined by HPLC on CHIRALCEL OJ-R (MeOH–H
e
2
O = 75+25). The number in
parenthesis indicates the yield based on the consumed starting material.
Chem. Commun., 1999, 2289–2290
This journal is © The Royal Society of Chemistry 1999
2289

counter cation may be explained by the transition state, shown
in Fig. 1. The bulky phosphite ligand should be located trans to
the more-substituted allylic terminus to avoid steric repulsion.
Nucleophilic attack anti to the phosphite ligand gives (S)-5. As
the anion becomes harder, it favours initial attack at iridium
metal, which is harder than the allylic carbon. Intramolecular
migration of malonate from iridium to the allylic carbon
proceeds to give (R)-5.
Table 2 Effect of the counter cation on the allylic alkylation of the carbonate
4b catalyzed by [Ir(COD)Cl]
–(S)-1 in THF at room temperaturea
2
Reaction Combined
Ratio of Ee of
Another feature is that the ee of (S)-5 is remarkably increased
to 96% when lithium and zinc are present in the reaction
medium (entry 7). It is premature to give a detailed rationaliza-
tion for this extraordinary increase in ee. However, it is possible
b
c
Entry Base
time/h
yield of 5+6 5 and 6 5 (%) Config.
1
2
3
4
5
6
7
a
ButP
KH
NaH
LiH
LDA
BuLi
4
24
7
87
90
96
14 (74)
85 (88)
71 (97)
99
93+7
90+10
95+5
46+54
94+6
92+8
93+7
36
18
6
22
17
24
96
R
R
R
S
S
S
S
3
that lithium chloride formed in situ may play a role. Work is
24
92
30
30
3
d
d
d
currently underway along these lines.
BuLi/ZnCl
2
b
Notes and references
Two equivalents of dimethyl malonate and base were used. Determined
1
c
2
by H NMR. Determined by HPLC on a CHIRALCEL OJ-R (MeOH–H O
=
consumed starting material.
1
R. Takeuchi and M. Kashio, J. Am. Chem. Soc., 1998, 120, 8647; Angew.
Chem., Int. Ed. Engl., 1997, 36, 263.
d
75+25). The number in parenthesis indicates the yield based on the
2
3
4
5
J. R. Janssen and G. Helmchen, Tetrahedron Lett., 1997, 38, 8025.
B. Bartels and G. Helmchen, Chem. Commun., 1999, 741.
P. H. Dussault and K. R. Woller, J. Org. Chem., 1997, 62, 1556.
M. J. Baker and P. G. Pringle, J. Chem. Soc., Chem. Commun., 1991,
1
292.
6
7
8
Easily prepared by a procedure similar to that for 1.
K. Fuji, N. Kinoshita and K. Tanaka, Chem. Commun., 1999, 1895.
R. Schwesinger and H. Schlemper, Angew. Chem., Int. Ed. Engl., 1987,
2
6, 1167.
9
T. Pietzonka and D. Seebach, Chem. Ber., 1991, 124, 1837.
Fig. 1 Transition state structure.
Communication 9/07680H
2290
Chem. Commun., 1999, 2289–2290