Enantioselective copper-catalyzed SN2' substitution with Grignard reagents

Article abstract of DOI:10.1055/s-2001-14626

Cinnamyl chlorides undergo selective S_N2' allylic substitution by Grignard reagents using catalytic amount (1 mol%) of CuCN and 1-2 mol% trivalent phosphorus ligand, in dichloromethane. With chiral phosphorus ligands derived from TADDOL ee's up

Full text of DOI:10.1055/s-2001-14626

LETTER
927
Enantioselective Copper-Catalyzed S 2’ Substitution with Grignard Reagents
N
Alexandre Alexakis, Christophe Malan, Louise Lea, Cyril Benhaim, Xavier Fournioux
Chimie Organique, Sciences II, University of Geneva, Quai Ernest Ansermet, CH-1211, Geneva, Switzerland
Fax +41 22 3287396; E-mail: Alexandre.Alexakis@chiorg.unige.ch
Received 10 January 2001
ported,7-13 but catalytic procedures employing chiral
ligands on copper are scarce and only moderate ee’s are
obtained. Van Koten, Bäckvall et al. recently reported are-
Abstract: Cinnamyl chlorides undergo selective S 2’allylic substi-
tution by Grignard reagents using catalytic amount (1 mol%) of
CuCN and 1-2 mol% trivalent phosphorus ligand, in dichlo-
N
romethane. With chiral phosphorus ligands derived from TADDOL nethiolatocopper(I) complexes which acheive ee’s of up
ee’s up to 73% could be obtained.
to 64% in the catalytic S 2' reaction between allylic ace-
Knochel et al. have developed
chiral ferrocenyl amine ligands, reporting ee’s of up to
N
tates and n-BuMgI.1
4-16
Key words: chiral phosphorus ligands, copper, asymmetric cataly-
sis, S 2’, TADDOL
N
9
8% in the catalytic S 2' reaction between allylic chlo-
N
17,18
rides and highly hindered diorganozinc reagents.
Allylic compounds have been of synthetic, mechanistic Our efforts so far have focused on the enantioselective
and biological importance for over 50 years. Catalytic copper(I)-catalyzed conjugate addition reaction on
asymmetric allylic substitutions are therefore potentially enones19 and other Michael acceptors,
20,21
during the
useful methods for the preparation of a wide range of course of which we have reported the use of several chiral
chiral molecules. The copper(I)-catalyzed allylic substitu- phosphorus ligands.2
2-25
We have recently turned our at-
tion reaction has generated a great deal of interest in re- tention to the application of such ligands to asymmetric
cent years and several methods have been developed for allylic substitution and report herein the results of our in-
the control of both regio- and stereochemistry in this reac- vestigation.
tion. An advantage is that a broad range of organometallic
To establish optimum conditions required for regioselec-
tivity (Scheme 1) we first examined the reaction of cin-
namyl chloride 1 with simple Grignard reagents in the
presence of a copper(I)-catalyst and triethylphosphite.
compounds, organolithium, Grignard and organozinc re-
agents can be used in these allylic substitutions.1 Cop-
-6
per(I)-promoted asymmetric -substitution (S 2’) of
N
allylic substrates with a chiral leaving group has been re-
Various conditions were investigated (Table 1): EtMgBr
Et
EtMgBr (1.2 eq),CuCN (1mol%)
+
Ph
Cl
Ph
Ph
Et
*
Ligand (1mol% bi-, 2 mol% monodentate)
1
2
3
CH2Cl2, –78 °C, 1 h
Scheme 1
Table 1 Copper-Catalyzed Substitution of Cinnamyl Chloride with Ethyl Grignard Reagents in the Presence of Triethyl-
phosphite
EtMgCl (2.8 M in THF). EtMgBr (3.0 M in Et O). Conversion >95% in all cases. *Cinnamyl acetate
2
Normal addition refers to slow addition of the Grignard reagent (over 20 min) to the reaction mixture at 78 °C. Inverse
addition refers slow addition of cinnamyl chloride (over 20 min) to the reaction mixture at 78 °C.
Synlett 2001 SI, 927–930 ISSN 0936-5214 © Thieme Stuttgart · New York

9
28
A. Alexakis et al.
LETTER
Table 2 Some Chiral Phosphorus Ligands Screened; Reaction Conditions as Scheme 1
awith CuBr, b with CuCl
Table 3 CuCN-Catalyzed Substitution of Cinnamyl Halides with EtMgBr in the Presence of Ligand 4
Reactions carried out at 78 °C. EtMgBr (3M in Et O) was added over 20 min unless otherwise stated.
2
*
Grignard addition over 40 min
was shown to give a more favorable regioselectivity than showed a remarkably increased asymmetric induction
EtMgCl; reaction temperature and the choice leaving over the other ligands achieving an ee of 61% in prelimi-
group had a significant effect on product distribution with nary studies.
lower temperatures ( 80 °C) favoring the S 2' product 2
N
Optimization of reaction conditions afforded a maximum
ee, in the case of ligand 4 of 73% (Table 3). It is evident
and cinnamyl chloride giving a favorable S 2' : S 2 ratio
26
N
N
while cinnamyl acetate favored the S 2 product 3; solvent
N
that the choice of solvent, which has a significant effect on
regioselectivity, is critical to enantioselectivity. The opti-
mum enantioselectivity favors a 1:1 ratio of copper(I)-cat-
alyst to ligand and is most efficient when just 1 mol% is
employed (entry 3). The rate of addition of the Grignard
effects were shown to be of importance with CH Cl being
2
2
the favored solvent; and finally the choice of copper salt
was critical to the product distribution, copper(I) cyanide
giving the best results (S 2' : S 2 = 97:3).
N
N
A total of 29 chiral phosphorus ligands were screened, reagent was also shown to be of significance, when the
many of which gave little or no asymmetric induction. Grignard reagent was added over 40 min the ee was in-
However some ligands, particularly: (i) those derived creased by 10% (entry 4).
from ( )-TADDOL and (ii) those bearing two points of
The choice of leaving group is critical to regio- and enan-
attachment on the lateral chain (e.g. an amino alcohol or
tioselectivity: the use of cinnamyl bromide caused a
oxazoline moiety) gave good results (Table 2). One
significant drop in enantioselectivity and an unfavorable
ligand, PO -( )-TADDOL-( )-N-methylephedrine
4
3
product distribution (entry 5), while cinnamyl acetate fa-
Synlett 2001, SI, 927–930 ISSN 0936-5214 © Thieme Stuttgart · New York

LETTER
Enantioselective Copper-Catalyzed S 2’Substitution with Grignard Reagents
929
N
Et
EtMgBr (1.2 eq), CH2Cl2
c-Hex
CuCN (1 mol%), 4 (1 mol%), –78 °C, 1 h
c-Hex
OAc
14
15
exclusively SN2'
100% conv., 0% ee
Scheme 2
Table 4 CuCN-Catalyzed Substitution of Cinnamyl Chloride with various Grignard Reagents in the Presence of Ligand 4
Reactions carried out in CH Cl (0.4 M with respect to substrate), in the presence of 1 mol% of CuCN and 1 mol% of chiral
2
2
ligand, at 78 °C. Grignard reagents were added over 40 min as a solution (typically 3M) in diethyl ether.
a
Reaction at 72 °C
vors the S 2 product (see Table 1, entry 11). The saturated reaction. Furthermore the use of different Grignard re-
N
allylic acetate 14 gave excellent regioselectivity but no agents revealed that there is no single set of reaction con-
asymmetric induction under these conditions (Scheme 2). ditions applicable to all cases. Moderate ee’s (47-57%)
have been achieved with a range of primary and second-
Having confirmed the optimum conditions for the reac-
ary aliphatic Grignard reagents and an ee of 21% has been
tion in question new Grignard reagents were introduced in
observed for the reaction with an aromatic Grignard re-
order to test the scope of the reaction (Table 4). Ethyl
agent. Highly sterically hindered Grignard reagents are
magnesium bromide is clearly preferable to other Grig-
not suitable substrates for the reaction resulting in poor, if
nard salts (entries 2 and 3) giving a greatly increased
any observed enantioselectivity. The reaction has been ex-
enantiomeric excess. Primary and secondary aliphatic
tended to other allylic chlorides and preliminary results
Grignard reagents react with moderate enantioselectivity
suggest that the regio- and enantioselectivity are of a sim-
under these conditions (entries 4-6), however poor enanti-
oselectivity is observed with sterically demanding Grig-
ilar range (e.g. p-methoxycinnamyl chloride gives 54%
ee, 96:4 S 2' :S 2). Further investigations into the scope
N
N
nards reagents such as neopentyl magnesium bromide
entry 7).
and limitations of the reaction are currently underway.
(
Finally the methodology was extended to aromatic Grig-
nard reagents. Several aromatic Grignard reagents were
tested but sufficient separation by chiral GC was not ob-
Acknowledgement
o
The authors thank the Swiss National Research Foundation N . 20-
tained in the majority of cases. Application of 2-methox- 53967.98 and COST action D12/0022/99 for financial support.
yphenylmagnesium bromide, which did not undergo
reaction at 78 °C, yielded a favorable product distribu-
tion at 72 °C and an ee of 21%, which has not yet been
References and Notes
optimized. Further investigations into the application of
aromatic Grignard reagents are currently underway.
(1) Commercon, A.; Bourgain, M.; Delaumeny, M.; Normant, J.
F.; Villieras, J. Tetrahedron Lett. 1975, 16, 3837.
(
2) Barsantai, P.; Caló, V.; Lopez, L.; Marchese, G.; Naso, F.;
Pesce, G. J. Chem. Soc. Chem. Commun. 1978, 1085.
In conclusion, we have reported a new chiral ligand for
the enantioselective copper(I)-catalyzed allylic substitu-
tion reaction, achieving the highest ee (73%) yet reported
with a Grignard reagent. It is clear from our results that the
enantioselectivity of the reaction, which is truly catalytic
requiring only 1 mol% of both the ligand and copper(I)
catalyst, is highly dependent on the reaction conditions.
The mode and time of addition of the Grignard reagent is
critical to the enantioselectivity, as is the temperature of
(3) Tseng, C. C.; Paisley, S. D.; Goering, H. L. J. Org. Chem.
1986, 51, 2884.
(
(
(
4) Bäckvall, J.-E.; Sellen, M.; Grant, B. J. Am. Chem. Soc. 1990,
12, 6615.
5) Bäckvall, J.-E.; Persson, E. S. M.; Bombrun, A. J. Org. Chem.
994, 59, 4126.
6) Persson, E. S. M.; van Klaveren, M.; Grove, D. M.; Bäckvall,
J.-E. Chem.-Eur. J. 1995, 1, 351.
1
1
(7) Denmark, S. E.; Marble, L. K. J. Org. Chem. 1990, 55, 1984.
Synlett 2001, SI, 927–930 ISSN 0936-5214 © Thieme Stuttgart · New York

9
30
A. Alexakis et al.
LETTER
(26) General procedure; To a well dried (two cycles of heating
(
(
8) Caló, V.; Fiandanese, V.; Nacci, A. Trends Org. Chem. 1993,
4
, 479.
9) Caló, V.; Fiandanese, V.; Nacci, A.; Scilimati, A. Tetrahedron
994, 50, 7283.
10) Caló, V.; Nacci, A.; Fiandanese, V. Tetrahedron 1996, 52,
0799.
11) Gais, H.-J.; Müller, H.; Bund, J.; Scommoda, M.; Brandt, J.;
Raabe, G. J. Am. Chem. Soc. 1995, 117, 2453.
12) Rakotoarisoa, H.; Perez, R. G.; Mangeney, P.; Alexakis, A.
Organometallics 1996, 15, 1957.
13) Alexakis, A.; Mangeney, P.; Ghribi, A.; Marek, I.; Sedrani,
R.; Guir, C.; Normant, J. F. Pure Appl. Chem. 1988, 60, 49.
14) Meuzelaar, G. J.; Karlström, A. S. E.; van Klaveren, M.;
Persson, E. S. M.; del Villar, A.; van Koten, G.; Bäckvall, J.-
E. Tetrahedron 2000, 56, 2895.
under vacuum) 25 mL two-necked flask equipped with a
stirrer bar and internal thermometer and purged with argon
was added copper(I) cyanide (3.6 mg, 1 mol%) and the chiral
1
(
(
(
(
(
ligand, PO ( )-TADDOL-( )-N-methylephedrine (27 mg, 1
3
1
mol%). Dichloromethane (8 mL) was added and the mixture
was stirred at room temperature for 20 min. Cinnamyl chloride
(0.56 mL, 4.0 mmol) was introduced dropwise and the
reaction mixture was stirred at room temperature for a further
5 min before being cooled to 78 °C (internal temperature)
via an acetone-dry ice cold bath. Ethyl magnesium bromide
(3.0 M in diethyl ether, 1.6 mL, 4.8 mmol) in dichloromethane
(2 mL) was added over 40 min via a syringe pump, whilst the
internal temperature was maintained at 78 °C. Once addition
was complete the reaction mixture was left at 78 °C for a
further hour at which point gas chromatography of an aliquot
showed that all the starting material had been converted. The
reaction was quenched by addition of aqueous hydrochloric
acid (1 N, 10 mL). Diethyl ether (40 mL) was added and the
aqueous phase was separated and extracted further with
diethyl ether (3 10 mL). The combined organic fractions
were washed with brine (20 mL), dried over anhydrous
sodium sulfate, filtered and reduced in vacuo. The oily residue
was purified by flash column chromatography (silica gel,
(
15) van Klaveren, M.; Persson, E. S. M.; del Villar, A.; Grove, D.
M.; Bäckvall, J.-E.; van Koten, G. Tetrahedron Lett. 1995, 36,
3059.
(16) Karlström, A. S. E.; Huerta, F. F.; Meuzelaar, G. J.; Bäckvall,
J.-E. Accompanying paper in this issue.
(
(
(
17) Dübner, F.; Knochel, P. Angew. Chem. Int. Ed. 1999, 38, 379.
18) Dübner, F.; Knochel, P. Tetrahedron Lett. 2000, 41, 9233.
19) Alexakis, A.; Frutos, J. C.; Mangeney, P. Tetrahedron:
Asymmetry 1993, 4, 2427.
(20) Alexakis, A.; Vastra, J.; Mangeney, P. Tetrahedron Lett.
eluant = cyclohexane, R = 0.5) to yield the product (561 mg,
f
1997, 38, 7745.
3.84 mmol, 96%) as a mixture of S 2' and S 2 regioisomers
N
N
(
(
21) Alexakis, A.; Benhaim, C. Org. Lett. 2000, 2, 2579.
22) Alexakis, A.; Burton, J.; Vastra, J.; Mangeney, P.
Tetrahedron: Asymmetry 1997, 8, 3987.
(96:4). Gas chromatography on a chiral stationary phase
(Chiraldex G-TA) showed that the S 2' product had an ee of
N
73%.
(
(
(
23) Alexakis, A.; Vastra, J.; Burton, J.; Benhaim, C.; Mangeney,
P. Tetrahedron Lett. 1998, 39, 4821.
24) Alexakis, A.; Benhaim, C.; Fournioux, X.; van den Heuvel,
A.; Levêque, J.-M.; March, S.; Rosset, S. Synlett 1999, 1811.
25) Alexakis, A.; Burton, J.; Vastra, J.; Benhaim, C.; Fournioux,
X.; van den Heuvel, A.; Levêque, J.-M.; Mazé, F.; Rosset, S.
Eur. J. Org. Chem. 2000, 4011.
Article Identifier:
1437-2096,E;2001,0,SI,0927,0930,ftx,en;Y00601ST.pdf
Synlett 2001, SI, 927–930 ISSN 0936-5214 © Thieme Stuttgart · New York