Dynamic kinetic asymmetric transformation in copper catalyzed allylic alkylation

Article abstract of DOI:10.1039/b907722g

The first dynamic kinetic asymmetric transformation in copper catalyzed allylic alkylation is reported, with enantioselectivities up to 92%.

Full text of DOI:10.1039/b907722g

View Article Online / Journal Homepage / Table of Contents for this issue
This article was published as part of the
2009 ‘Catalysis in Organic
Synthesis’ web theme issue
Showcasing high quality research in organic chemistry
Please see our website
(http://www.rsc.org/chemcomm/organicwebtheme2009)
to access the other papers in this issue.

COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Dynamic kinetic asymmetric transformation in copper catalyzed allylic
alkylationwz
Jean-Baptiste Langlois and Alexandre Alexakis*
Received (in Cambridge, UK) 16th April 2009, Accepted 8th May 2009
First published as an Advance Article on the web 4th June 2009
DOI: 10.1039/b907722g
The first dynamic kinetic asymmetric transformation in copper
catalyzed allylic alkylation is reported, with enantioselectivities
up to 92%.
elimination reaction (entry 6). Clearly, the enantioselectivity is
intimately linked to the leaving group ability of X (Br c Cl c
OP(O)Ph2 4 OCO2Me 4 OAc). This result prompted us to
further optimize the process using model substrate 5.
Although various ligands were tested in this reaction, L1
was found to be the most effective (entry 1, Table 2). The use
of ligand L2 (Fig. 1) led to a mismatch situation (18% ee; entry 2).9
It is worth noting that the absolute configuration of the
adduct remains the same. A similar situation was observed
with ligands L3 and L4, which afforded product 7 in 32 and
76% ee, respectively (entries 3 and 4). Ligand L5, having a
naphthylethylamine group, gave a comparable result to L1
and L4 (entry 5). Tropos ligand10 L6 and Simplephos ligand11
L7 failed to improve this result and poor enantioselectivities
were obtained (entries 6 and 7). Low reactivity was observed
with Taniaphos ligand L8 (entry 8). Almost no enantioinduction
was obtained using carbene ligand precursor L9 (entry 9).12
Keeping L1 as ligand, different solvents were screened.
Coordinating solvents, such as diethyl ether (Et2O) and
tetrahydrofuran (THF), led to the formation of product 7 as
a racemate (entries 10 and 11). Dichloromethane appeared to
be the solvent of choice for this reaction (entry 1). Lowering
the reaction concentration from 0.18 M to 0.1 M enhanced the
enantioselectivity to 86% ee (entry 12). A concentration of
0.07 M did not further improve this result (entry 13), suggesting
that a precise quantity of CH2Cl2 was needed to reduce the
At the end of the 90s, Trost introduced the concept of dynamic
kinetic asymmetric transformation (DYKAT) in allylic
substitution.1 This represents one of the most efficient
methodologies for the transformation of a racemic substrate
in an enantio-enriched product.2 The concept was applied to
palladium catalyzed asymmetric allylic alkylation (AAA)
using stabilized nucleophiles.3 However, to the best of our
knowledge, only
a few examples were described using
non-stabilized nucleophiles. Some advances in this field were
achieved in nickel catalysis. Consiglio and co-workers
described the alkylation of cyclic allylic ethers by EtMgBr
affording quantitatively alkylated products in high enantio-
selectivity (up to 93% ee).4,5 Recently Fu and co-workers
reported the alkylation of acyclic allylic chlorides with
organozinc reagents with ee up to 98%.6 Surprisingly,
although copper is one of the most efficient transition metals
to form enantioselectively an allylic C–C bond using
non-stabilized nucleophiles,7 no examples of DYKAT in
copper catalysis have been reported to date. In 2007, Backvall
¨
described a copper catalyzed racemisation of enantiopure
allylic acetates, demonstrating that a DYKAT was envisageable
in this chemistry.8 In this context it was interesting to develop
an efficient dynamic process in copper catalysis. We report
herein our investigations in this field.
Table 1 Screening of different leaving groupsa
We started our study with the reaction of cyclohex-1-enyl-3-
acetate 1 and phenethylmagnesium bromide in CH2Cl2
(Table 1). Using copper(I) thiophene-2-carboxylate (CuTC,
5 mol%) and L1 (5.5 mol%) at ꢀ78 1C, the product was
obtained quantitatively, albeit in racemic form (entry 1,
Table 1). Other leaving groups such as carbonate 2 and
phosphonite 3 were tested, resulting in low but measurable
enantiomeric excesses (entries 2 and 3). A large improvement
in enantioselectivity was attained with chloride 4 and bromide
5, which provided the product (S)-7 with 53 and 78% ee,
respectively (entries 4 and 5). The formation of the product
was not observed with triflate 6, probably due to a competitive
Entry
X
Substrate t/h Conversion (%)b ee (%)c
1
2
3
4
5
6
OAc
1
2
3
4
5
6
15
15
3
2
1
83
83
499
499
499
0d
0
OCO2Me
OP(O)Ph2
Cl
Br
OTf
4 (S)
8 (S)
53 (S)
78 (S)
d
Department of Organic Chemistry, Quai Ernest Ansermet, 30,
1211 Geneva 4, Switzerland. E-mail: alexandre.alexakis@unige.ch;
Fax: (+0041) 22-379-3215; Tel: (+0041) 22-379-6522
w This article is part of a ChemComm ‘Catalysis in Organic Synthesis’
web-theme issue showcasing high quality research in organic chemistry.
Please see our website (http://www.rsc.org/chemcomm/organic
webtheme2009) to access the other papers in this issue.
1
—
a
Reaction conditions: Racemic substrate (0.5 mmol) was added to a
solution of CuTC and L1 in dry CH2Cl2 (2 ml). The reaction mixture
was cooled to ꢀ78 1C and the Grignard reagent (1 M in Et2O) was
b
added dropwise. Conversion relative to the formation of 7,
c
determined by GC-MS. Determined by GC on chiral stationary
d
phase. The formation of 7 was not observed.
z Electronic supplementary information (ESI) available: Experimental
details. See DOI: 10.1039/b907722g
ꢁc
This journal is The Royal Society of Chemistry 2009
3868 | Chem. Commun., 2009, 3868–3870

Table 2 Screening of chiral ligands and solventsa
Table 3 Scope of the processa
Entry
Ligands
Solvent
Conversion (%)b
ee (%)c
1
2
3
4
5
6
7
8
L1
L2
L3
L4
L5
L6
L7
L8
L9e
L1
L1
L1
L1
L1h
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
Et2O
499
499
499
499
499
499
499
8
499
499
499
499
499
499
78 (S)
18 (S)
32 (S)
76 (S)
78 (R)
4 (S)
8 (S)
n.d.d
2 (R)
0
Entry Substrate R2
Product Yield (%)b ee (%)d
9
10
11
12
13
14
1
2
3
4
5
6
7
8
9
10
5
5
5
5
5
5
5
8
9
10
PhCH2CH2-
Et-
n-Bu-
t-BuOBu-
Cy-
t-Bu-
7
95
79c
81c
91
98
80c
97
93
98
95
92 (S)
90 (R)
92 (R)
90 (S)
70 (S)
50 (R)
0e
44 (S)
38 (S)
18 (S)
THF
0
11
12
13
14
15
16
f
CH2Cl2g
CH2Cl2f
CH2Cl2
86 (S)
86 (S)
92 (S)
a
Reaction conditions: Racemic substrate 5 (0.5 mmol) was added to a
solution of CuTC (5 mol%) and chiral ligand (5.5 mol%) in dry
solvent (2 ml). The reaction mixture was cooled to ꢀ78 1C and the
Grignard reagent (1 M in Et2O, 1.5 equiv.) was added dropwise.
Ph-
PhCH2CH2- 17
PhCH2CH2- 18
PhCH2CH2- 19
b
c
Conversion determined by GC-MS. Determined by GC on a chiral
a
d
stationary phase. Not determined. Carbene ligand was formed
e
Reaction conditions: Racemic substrate (0.5 mmol) was added to a
solution of CuTC and L1 in dry CH2Cl2 (4 ml). The reaction
mixture was cooled to ꢀ78 1C and the Grignard reagent was added
f
in situ by addition of n-BuLi. 0.1
g
M in 5. 0.07 M in 5.
h
7.5 mol% catalyst loading.
b
c
dropwise. Yield of isolated product. Isolated yield of the
corresponding diastereomeric epoxides obtained by treatment with
d
mCPBA. Determined by GC on a chiral stationary phase after
derivatization of a sample of the isolated product into the corresponding
e
epoxides. Determined by SFC on chiral stationary phase.
16 as a racemate (entry 7). Other cyclic substrates were
surveyed. Cyclopent-1-enyl-3-bromide
1-enyl-3-bromide 9 afforded products 17 and 18 in 44 and
38% ee, respectively (entries and 9). 18% ee was
8 and cyclohept-
8
also obtained with 1-methylcyclohex-1-enyl-6-bromide 19
(entry 10).
Based on these results, we propose a plausible reaction
mechanism (Scheme 1). The first step is the ionization of the
substrate by insertion of the metal into the allylic terminus
(oxidative addition) and the last is the reductive elimination
leading to a mixture of products 23 and ent-23. During these
two steps, the s-allyl species 20 and 21, stemming from the
Fig. 1 Chiral ligands.
possible coordination of ether (contained in the Grignard
solution) to the metal.13
The nature of the copper source does not have a notable
influence on the selectivity of the reaction except for copper
cyanide and copper triflate which gave lower ee’s (see Supporting
Informationz).14
Increasing the catalyst loading from 5 mol% to 7.5 mol%
avoided an erosion of the enantiomeric excess during the
course of the reaction and 92% ee was attained (entry 14,
Table 2; see Supporting Informationz). The use of organozinc
reagents,15 instead of Grignard reagents, led to the same
enantiomeric excess (see Supporting Informationz).
With these conditions in hand, the scope of Grignard
reagents was studied in order to test the generality of this
method (Table 3).y The process was found to be very efficient
with primary alkyl Grignard reagents which always gave
excellent ee’s ranging from 90 to 92% (entries 1 to 4,
Table 3). A lower enantioselectivity was observed with
secondary and tertiary alkyl Grignard reagents. Indeed,
3-cyclohexylcyclohex-1-ene 14 and 3-tert-butylcyclohex-1-ene
15 were isolated with 70 and 50% ee, respectively (entries
5 and 6). The use of phenyl Grignard reagent afforded product
Scheme 1 Proposed mechanism for the DYKAT of racemic substrate
5 (R = alkyl or aryl group, k = equilibrium constant).
ꢁc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 3868–3870 | 3869

ionization of 5 and ent-5, are in equilibrium via meso-p-allyl
complex 22.16 In the case of low reactive allylic acetates,
Cu(III) intermediates are accepted to have a very short
lifetime. Consequently, the ionization step is rate limiting
(k1, k2 { k7, k6) and virtually no p-allyl equilibration takes
place.17 In the present case, allylic bromides are very reactive
and Cu(III) intermediates presumably equilibrate via p-allyl
intermediates of type 22. Hence the reductive elimination
step became rate determining (k1, k2 c k7, k6). Control
experiments showed that the starting material 5 is racemic
throughout the reaction (see Supporting Informationz), which
is consistent with a dynamic process (k1 = k2). Ruling out the
hypothesis of kinetic resolution implies that the enantio-
discrimination is due to a difference in the rate of the reductive
elimination.18 This difference displaces the equilibration of the
intermediates through the formation of the product with the
higher rate of elimination (i.e. if k6 4 k7 the product will be
enriched in 23). This is consistent with the racemates obtained
using coordinating solvents (entries 8 and 9, Table 2) where
the reductive elimination is faster.19 The same observation was
made with the use of bulky R groups such as secondary
and tertiary alkyl groups, which provided hindered allyl-
intermediates favouring the elimination step (entries 5 and 6,
Table 3).
(26), 91 (100), 65 (34), 53 (22). HRMS (ESI) calcd for C14H18 [M+
186.1409, found 186.1407.
]
1 (a) B. M. Trost and F. D. Toste, J. Am. Chem. Soc., 1999, 121,
3543–3544; (b) B. M. Trost, R. C. Bunt, R. C. Lemoine and
T. L. Calkins, J. Am. Chem. Soc., 2000, 122, 5968–5976.
2 (a) K. Faber, Chem.–Eur. J., 2001, 7, 5004–5010; (b) J. Steinreiber,
K. Faber and H. Griengl, Chem.–Eur. J., 2008, 14,
8060–8072.
3 (a) B. M. Trost, Chem. Pharm. Bull., 2002, 50, 1–14;
(b) B. M. Trost and R. C. Bunt, J. Am. Chem. Soc., 1994, 116,
4089–4090; (c) B. M. Trost, D. E. Patterson and E. J. Hembre,
J. Am. Chem. Soc., 1999, 121, 10834–10835.
4 Ni catalyzed allylic alkylation: (a) G. Consiglio, O. Piccolo,
L. Roncetti and F. Morandini, Tetrahedron, 1986, 42,
2043–2053; (b) G. Consiglio and R. M. Waymouth, Chem. Rev.,
1989, 89, 257–276; (c) A. F. Indolese and G. Consiglio,
Organometallics, 1994, 13, 2230–2234.
5 Metal catalyzed allylic arylation: Ni catalysis: (a) K. G. Chung,
Y. Miyake and S. Uemura, J. Chem. Soc., Perkin Trans. 1, 2000,
15–18; (b) K. G. Chung, Y. Miyake and S. Uemura, J. Chem. Soc.,
Perkin Trans. 1, 2000, 2725–2729; Ir catalysis: (c) D. Polet,
X. Rathgeb, C. A. Falciola, J.-B. Langlois, S. El Hajjaji and
A. Alexakis, Chem.–Eur. J., 2009, 15, 1205–1216.
6 S. Son and G. C. Fu, J. Am. Chem. Soc., 2008, 130,
2756–2757.
7 For recent reviews, see: (a) A. Alexakis, J. E. Backvall, N. Krause,
¨
O. Pamies and M. Dieguez, Chem. Rev., 2008, 108, 2796–2823;
´
(b) B. L. Feringa, A. J. Minnaard, K. Geurts, T. Den Hartog and
S. R. Harutyunyan, Chem. Rev., 2008, 108, 2824–2852;
(c) A. Alexakis and C. A. Falciola, Eur. J. Org. Chem., 2008,
3765–3780; (d) A. Alexakis, C. Malan, L. Lea, K. Tissot-Croset,
D. Polet and C. A. Falciola, Chimia, 2006, 60, 124–130. See also:
(e) A. Alexakis and C. A. Falciola, Chem.–Eur. J., 2008, 14,
10615–10627; (f) M. A. Kacprzynski, T. L. May, S. A. Kazane
and A. H. Hoveyda, Angew. Chem., Int. Ed., 2007, 46, 4554–4558;
(g) C. A. Falciola, K. Tissot-Croset and A. Alexakis, Angew.
Chem., Int. Ed., 2006, 46, 5995–5998; (h) F. Lopez, A. W. Van
Zijl, A. J. Minnaard and B. L. Feringa, Chem. Commun., 2006,
409–411.
In conclusion, despite a lack of generality, we have disclosed
the first dynamic kinetic asymmetric transformation in copper
catalyzed asymmetric allylic alkylation. This concept was
applied to the alkylation of cyclohex-1-enyl-3-bromide 5 and
was very efficient with primary alkyl Grignard reagents
(ee’s up to 92%). Work is in progress to widen the scope
and the efficiency of the process.
The authors thank the Swiss National Research Foundation
(grant No. 200020-113332) and COST action D40 (SER
contract No. C07.0097) for financial support, as well as BASF
for the generous gift of chiral amines.
8 J. Norinder and J. E. Ba
4094–4102.
9 B. L. Feringa, M. Pineschi, L. A. Arnold, R. Imbos and A. H.
M. De Vries, Angew. Chem., Int. Ed. Engl., 1997, 36, 2620–2623.
10 A. Alexakis, D. Polet, S. Rosset and S. March, J. Org. Chem.,
2004, 69, 5660–5667.
¨
ckvall, Chem.–Eur. J., 2007, 13,
11 L. Palais, I. S. Mikhel, C. Bournaud, L. Micouin, C. A. Falciola,
M. Vuagnoux-d’Augustin, S. Rosset, G. Bernardinelli and
A. Alexakis, Angew. Chem., Int. Ed., 2007, 46, 7462–7465.
12 (a) A. Alexakis, C. L. Winn, F. Guillen, J. Pitkowicz, S. Roaland
and P. Mangeney, Adv. Synth. Catal., 2003, 345, 345–348;
(b) S. Tominaga, Y. Oi, T. Kato, D. K. An and S. Okamoto,
Tetrahedron Lett., 2004, 45, 5585–5588.
13 The concentration of the reaction mixture could be increased to 0.2 M
if the ether solution of the Grignard reagent was concentrated
beforehand in vacuo and redissolved in dichloromethane. No
change in terms of enantioselectivity was observed. This informa-
tion is particularly important in the case of scaling-up of the
reaction.
Notes and references
y Representative procedure for copper catalyzed allylic alkylation
(entry 1, Table 3): In a flame-dried Schlenk tube under argon
atmosphere, CuTC (7.2 mg, 0.038 mmol, 0.075 equiv.) and L1
(22.2 mg, 0.041 mmol, 0.083 equiv.) were dissolved in dry CH2Cl2
(4 ml) and the solution was stirred for 10 min at room temperature.
Then the substrate 5 (80 mg, 0.5 mmol, 1 equiv.) was added and the
solution was cooled to ꢀ78 1C. After 10 min at this temperature, the
phenethylmagnesium bromide solution (1
M in Et2O, 0.6 ml,
0.6 mmol, 1.2 equiv.) was added dropwise and the reaction mixture
was stirred for 1 h. The reaction was quenched with an aqueous
solution of 1 M HCl (15 ml) and extracted with Et2O (15 ml). The
organic layer was washed with 1 M HCl (15 ml) and brine (15 ml),
dried over Na2SO4, filtered and concentrated in vacuo. The crude
mixture was purified on a silica gel chromatography column (pentane)
to afford product 7 (88 mg, 95%) as a colourless liquid. 7: The
enantiomeric excess was determined by GC on a chiral stationary
phase (Hydrodex B3P column, Method: 60-30-1-140-20-170-5, RT:
102.17 (S), 102.76 (R) min). The enantiomeric excess could also be
determined after derivatization into the corresponding diastereomeric
epoxides (Hydrodex TBDM column, Method: 60-0-1-170-5, RT:
14 L. H. Goering, S. D. Paisley and C. C. Tseng, J. Org. Chem., 1986,
51, 2884–2891.
15 P. Knochel and R. D. Singer, Chem. Rev., 1993, 93, 2117–2188.
16 (a) E. Nakamura and S. Mori, Angew. Chem., Int. Ed., 2000, 39,
3750–3771; (b) M. Yamanaka, S. Kato and E. Nakamura, J. Am.
Chem. Soc., 2004, 126, 6287–6293; (c) E. R. Bartholomew,
S. H. Bertz, S. Cope, M. Murphy and C. A. Ogle, J. Am. Chem.
Soc., 2008, 130, 11244–11245.
17 L. H. Goering, S. S. Kantner and E. P. Seitz, J. Org. Chem., 1985,
50, 5495–5499.
97.99, 99.01, 102.71, 104.25 min). [a]25 = ꢀ0.87 (c = 1.2 in CHCl3,
D
92% ee). 1H NMR (400 MHz, CDCl3, 25 1C): d = 1.34–1.37 (m, 1H),
1.59–1.81 (m, 4H), 1.89 (m, 1H), 2.06 (m, 2H), 2.18 (m, 1H), 2.74
(m, 2H), 5.71–5.76 (m, 2H), 7.25–7.35 (m, 5H) ppm; 13C NMR (100 MHz,
CDCl3, 25 1C): d = 21.6, 25.5, 29.2, 33.4, 34.9, 38.4, 125.8, 127.3,
128.4, 128.5, 131.9, 143 ppm. IR (CHCl3): 71.9, 1453, 1493, 2856,
2923, 3023 cmꢀ1. MS (EI mode) m/z %: 186 (28), 143 (4), 129 (4), 104
18 A referee suggested that ‘‘the two enantiomeric bromides may
interconvert rapidly under the reaction conditions’’. Although this
seems unlikely at ꢀ78 1C, this possibility cannot be ruled out.
19 (a) J. E. Backvall, M. Sellen and B. Grant, J. Am. Chem. Soc.,
¨
1990, 112, 6615–6621; (b) J. E. Backvall and E. S. M. Persson, Acta
¨
Chem. Scand., 1995, 49, 899–906.
ꢁc
This journal is The Royal Society of Chemistry 2009
3870 | Chem. Commun., 2009, 3868–3870