Copper-catalyzed enantioselective 1,4-addition to α,β- unsaturated aldehydes

Article abstract of DOI:10.1021/ol100441r

Figure presented The first asymmetric Cu-catalyzed conjugate addition of dialkylzinc zinc reagents to a large scope of enals in presence of phosphoramidite, SimplePhos, or (R)-BINAP ligands with enantiomeric excesses up to 90% is reported. Moreover, ACA of Grignard reagents afforded moderate to good 1,4-regioselectivities with enantioselectivities up to 90%.

Full text of DOI:10.1021/ol100441r

ORGANIC
LETTERS
2010
Vol. 12, No. 9
1988-1991
Copper-Catalyzed Enantioselective
1,4-Addition to r,ꢀ-Unsaturated
Aldehydes
Lae¨titia Palais, Lucille Babel, Adrien Quintard, Se´bastien Belot, and
Alexandre Alexakis*
Department of Organic Chemistry, UniVersity of GeneVa, 30 quai Ernest Ansermet,
1211 GeneVe, Switzerland
alexandre.alexakis@unige.ch
Received February 23, 2010
ABSTRACT
The first asymmetric Cu-catalyzed conjugate addition of dialkylzinc zinc reagents to a large scope of enals in presence of phosphoramidite,
SimplePhos, or (R)-BINAP ligands with enantiomeric excesses up to 90% is reported. Moreover, ACA of Grignard reagents afforded moderate
to good 1,4-regioselectivities with enantioselectivities up to 90%.
The asymmetric Cu-catalyzed conjugate addition (ACA) of
organometallic reagents to Michael acceptors is among the
most important methodologies to enantioselectively create
C-C bonds.1 In this field, a large variety of R,ꢀ-unsaturated
compounds such as R,ꢀ-carbonyl derivatives, nitroalkenes,
sulfones, etc. have been used with success. The only
substrates that have not been reported are R,ꢀ-unsaturated
aldehydes. This is not surprising, as these are much more
challenging substrates due of their high reactivity, which
allow the undesired direct carbonyl attack (1,2-addition). An
interesting example has been reported by Marshall, in 2005,
where the use of diorganozinc reagents allowed this reaction
in racemic and diastereoselective versions.2 In the same year,
the enantioselective copper-free 1,4-addition of organozinc
reagents to R,ꢀ-unsaturated aldehydes was described by
Bra¨se using [2.2]-paracyclophaneketimine ligands.3 Excellent
enantioselectivities could be attained but with low regiose-
lectivities, with the 1,2-adduct representing 1/3 to 1/2 of the
converted product.
Other asymmetric 1,4-additions to enals using Rh4 or Pd5
catalysis were also reported by Miyaura, Hayashi, and
Carreira. In all of these chiral systems, the 1,4-addition is
limited to enals bearing an aromatic group. Moreover, for
Cu and Cu-free catalysis, only diorganozinc reagents were
used as nucleophiles. In Cu catalysis, alternative indirect
ways have been described by Hoveyda with amides6 and by
Feringa with thioesters,7 where the resulting adducts could
be converted to aldehydes in a further step. Herein, we report
the first successful enantioselective copper-catalyzed conju-
gate addition to several R,ꢀ-unsaturated aldehydes, with
diorganozinc and Grignard reagents.
(4) As a reference containing the results obtained for Rh-catalyzed ACA
to enals, see: (a) Itooka, R.; Iguchi, Y.; Miyaura, N. J. Org. Chem. 2003,
68, 6000–6004. (b) Tokunaga, N.; Hayashi, T. Tetrahedron: Asymmetry
2006, 17, 607–613. (c) Paquin, J. F.; Defieber, C.; Stephenson, C. R. J.;
Carreira, E. M. J. Am. Chem. Soc. 2005, 127, 10850–10851.
(5) As a reference containing the results obtained for Pd-catalyzed ACA
to enals, see: Nishikata, T.; Yamamoto, Y.; Miyaura, N. Chem. Commun.
2004, 1822–1823.
(1) For reviews, see: (a) Alexakis, A.; Benhaim, C. Eur. J. Org. Chem.
2002, 3221–3223. (b) Alexakis, A.; Backvall, J. E.; Krause, N.; Pamies,
O.; Dieguez, M. Chem. ReV. 2008, 108, 2796–2823. (c) Harutyunyan, S. R.;
den Hartog, T.; Geurts, K.; Minnard, A. J.; Feringa, B. L. Chem. ReV. 2008,
108, 2824–2852.
(2) Marshall, J. A.; Herold, M.; Eidam, H. S.; Eidam, P. Org. Lett. 2006,
8, 5505–5508.
(3) (a) Bra¨se, S.; Ho¨fener, S. Angew. Chem., Int. Ed 2005, 44, 7879–
7881. (b) Ay, S.; Nieger, M.; Bra¨se, S. Chem.sEur. J. 2008, 14, 11539–
11556. (c) See also in racemic version: Jones, P.; Reddy, C. K.; Knochel,
P. Tetrahedron 1998, 54, 1471–1490.
(6) Hird, A. W.; Hoveyda, A. H. Angew. Chem., Int. Ed. 2003, 42, 1276–
1279.
(7) Des Mazery, R.; Pullez, M.; Lopez, F.; Harutyunyan, S. R.; Minnaard,
A. J.; Feringa, B. L. J. Am. Chem. Soc. 2005, 127, 9966–9967.
10.1021/ol100441r  2010 American Chemical Society
Published on Web 04/01/2010

We initially investigated the ACA of dialkylzinc reagents
because their low reactivity could probably prevent or slow
the formation of the undesired 1,2-adduct. The first attempt
was done with Et2Zn and trans-2-decenal S1, in the presence
of 5 mol % of CuTC (copper thiophene carboxylate) and 10
mol % of phosphoramidite-type ligand L1 in Et2O at 0 °C
(Scheme 1). After only 3 h, the corresponding adduct A1
Table 1. Optimization of the Chiral Ligands
entry
ligand
convn (%)a
1,4/aldola
ee (%)b
1
2
3
4
5
6
7
8
9
L1
L2
L3
L4
L5
L6
L7
L8
L9c
>99
99
88
>99
99
93
99
94
59
>99
50
57
79
34
64
60
73
50
72
99:1
95:5
94:6
>99
88:12
93:7
99:1
87:13
Scheme 1. First Attempt of ACA of Et2Zn toS1
a Determined by 1H NMR. b Determined by chiral GC. c Performed with
5.25 mol % of L9.
was obtained with full conversion but with only 33% ee.
No 1,2-adduct was observed, but aldol product was formed
by reaction of the resulting zinc enolate with the highly
reactive substrate S1. To prevent the formation of this
byproduct, we optimized the sources of copper salt and the
solvents in presence of L1.
Performing the reaction in dichloromethane or in toluene,
with CuTC, did not improve the ratio of 1,4/aldol product,
even if the enantiomeric excess increased (55% in CH2Cl2
and 52% in toluene). However, when the reaction was done
with copper(II) triflate, as copper salt, and L1 in toluene,
the corresponding 1,4-adduct was obtained with full conver-
sion and 50% ee, without a trace of the aldol byproduct.
With these conditions in hand, we screened various chiral
ligands (Figure 1 and Table 1).
aldol product were observed, the enantioselectivity was
clearly better: 79% ee with L3 and 73% ee with L7.
Moreover, bidentate ligand (R)-BINAP L9 showed an
interesting result9 for the conjugate addition of Et2Zn to S1,
with 72% ee but only 87% of regioselectivity in favor of
the 1,4-adduct (entry 9). Further refinements in the reaction
conditions (Cu salt, solvent, temperature) allowed us to find
two distinct sets, depending on the ligand used. For mono-
dentate ligands L3 or L7, Cu(OTf)2 in toluene at 0 °C was
best, whereas for (R)-BINAP, CuTC in ether at -20 °C was
better. The enals used in this study were either commer-
cially available or easily prepared according to the literature
(Figure 2).
Figure 2. Cyclic and acyclic R,ꢀ-unsaturated aldehydes used.
As shown in Table 2, the ACA proceeds with excellent
conversions. With few exceptions, L7 generally affords better
1,4-selectivities (92-100%), although the enantioselectivities
are usually slightly lower. With L3, ee’s up to 90% could
be attained (entry 5). For the cyclic enal S8, the 1,2-product
was observed with L3 and L7, even if the 1,4-adduct was
obtained with 82% ee with L7 (entry 10). Finally, the
addition of Me2Zn was tested with acyclic enal S1 and ligand
L3. Both total conversion and 1,4-regioselectivity are
observed, affording the corresponding adduct B1 with 60%
ee (entry 11).
Figure 1. Chiral ligands tested with diorganozinc reagents.
Among the monodentate chiral ligands tested, two inter-
esting results were obtained, one with phosphoramidite ligand
L3, bearing ortho-phenyl groups on the chiral binaphthol
part, and SimplePhos8 L7 with a 2-naphthyl group on the
chiral amine (entries 3 and 7). Even if some traces of the
(8) (a) Palais, L.; Mikhel, I. S.; Bournaud, C.; Micouin, L.; Falciola,
C. A.; Vuagnoux-d’Augustin, M.; Rosset, S.; Bernardinelli, G.; Alexakis,
A. Angew. Chem., Int. Ed. 2007, 46, 7462–7465. (b) Palais, L.; Alexakis,
A. Chem.sEur. J. 2009, 15, 10473–10485.
(9) BINAP has been shown to be a good ligand for ACA with Grignard
reagents, not with dialkyl zinc reagents. See: (a) Wang, S. Y.; Ji, S. J.;
Loh, T. P. J. Am. Chem. Soc. 2007, 129, 276–277. (b) Alexakis, A.; Burton,
J.; Vastra, J.; Mangeney, P. Tetrahedron: Asymmetry 1997, 8, 3987–3990.
Org. Lett., Vol. 12, No. 9, 2010
1989

Dialkylzinc reagents offer some advantages because of their low
reactivity and their high tolerance toward functional groups on the
substrate, but also on the organozinc itself.10 However, their
commercial diversity remains limited, and this is why the use of
Grignard reagents was envisaged for this reaction. First, we
screened several chiral mono- and bidentate ligands for the ACA
of EtMgBr to trans-2-decenal S1 catalyzed by CuTC (Figure 3
and Table 4).
Table 2. ACA of R′2Zn to Various Enals with L3 and L7
entry L*
R
adduct convn (%)a 1,4/1,2/aldola ee (%)b
1
2
3
4
5
6
L3 C7H15
L7 C7H15
L3 i-Pr
L7 i-Pr
L3 c-Hex
L7 c-Hex
L3 Ph
L7 Ph
L3 S8
L7 S8
L3 C7H15
A1
A1
A5
A5
A6
A6
A7
A7
A8
A8
B1
88
99
95:0:5
93:0:7
79, R
73, R
51, S
60, S
90, S
81, S
82, S
71, S
60d
100
100
100
100
81
77:11:12
100:0:0
100:0:0
100:0:0
92:0:8
7c
8c
9c
10c
11c
71
92:0:8
100
100
100
53:47:0
22:78:0
100:0:0
82e
60, R
a Determined by 1H NMR. b Determined by chiral GC. c Reaction
performed over 24 h. d Enantiomeric excess of the major diastereoisomer,
dr 68:32. e Enantiomeric excess of the major diastereoisomer, dr 70:30.
Figure 3. Chiral ligands tested with Grignard reagents.
The reaction was performed at low temperature (-78 °C)
due to the high reactivity of both the Grignard reagent and
the enal itself. Not unexpectedly, large amounts of 1,2-
adducts were obtained, as well as some aldol products.
However, (R)-BINAP L9 and (R)-tol-BINAP L10 gave the
best regio- and enantioselectivities, up to 92% (entries 3 and
4). Other mono- and bidentate phosphorus ligands afforded
the corresponding adduct with lower ee’s and 1,4-regiose-
lectivities. Finally, N-heterocyclic carbene precursor L1411
underwent mainly 1,2-addition (entry 8).
In a second screening, the 1,4-addition of dialkylzinc reagents
was performed using 5 mol % of CuTC and 5.25 mol % of
(R)-BINAP L9 (Table 3). Interestingly, in all cases, no traces
Table 3. ACA of R2Zn to Various Enals with L9
entry
R
adduct
convn (%)a
ee (%)b
Table 4. ACA of EtMgBr to Enal S1 with Various Chiral
Ligands
1
2
3
4
C7H15
n-Bu
i-Bu
i-Pr
c-Hex
Ph
C7H15
n-Bu
i-Pr
c-Hex
Ph
A1
A2
A4
A5
A6
A7
B1
B2
B5
B6
B7
100 (75)d
100 (60)e
100
75, R
64, R
70, R
27, S
52, S
44, S
68, R
76, R
70, S
60, S
64, S
100
5
100 (50)d
100 (81)d
85 (79)d
87
6c
7c
8c
9c
10c
11c
entry
L*
convn (%)a
1,4/1,2/aldola
ee (%)b
1c
2c
3
4
5
6
7
8
L1
L7
L9
L10
L11
L12
L13
L14
100
100
98
99
98
96
98
97
24:76:0
21:79:0
62:38:0
32:42:26
21:59:20
2:97:1
50
44
89
92
100
87 (61)d
50
a Determined by 1H NMR. b Determined by chiral GC. c Reaction
performed at 0 °C over 16 h. d Yield in parentheses. e Crude yield.
68
ndd
88
16:64:20
8:89:3
ndd
of 1,2-adduct or aldol formation were observed for the addition
of Et2Zn or Me2Zn. However, the enantioselectivities are usually
lower than with phosphoramidites L3 and L7, except for the
addition of Me2Zn, where the ee’s were better. With this reagent,
the reaction was performed at higher temperature (0 °C) due
to its lower reactivity. Tol-BINAP did not improve these results,
and neither did the addition of trimethylchlorosilane (TMSCl).
a Determined by 1H NMR. b Determined by chiral GC. c Reaction
performed with 10 mol % of ligand at -30 °C. d Not determined.
(10) Knochel, P.; Singer, R. D. Chem. ReV. 1993, 93, 2117–2188.
(11) Wencel, J.; Mauduit, M.; He´non, H.; Kehrli, S.; Alexakis, A.
Aldrichimica Acta 2009, 42, 43–50.
1990
Org. Lett., Vol. 12, No. 9, 2010

After defining the best ligand, some optimizations were
attempted (solvent, temperature, Cu salt)12 with L9 and L10
to increase the 1,4-regioselectivity. The key point was the
beneficial effect of added TMSCl. TMSCl is well-known to
increase the rate of the conjugate addition of cuprates13 and to
favor the 1,4-addition on enals.14 However, TMSCl had never
been used in ACA with Grignard reagents, although its
detrimental effect was noticed with Et2Zn.1a As we had
expected, the addition of TMSCl (1.3 equiv) gave a promising
result for 1,4-regioselectivity, which increased from 32 to 85%
with L10. Moreover, the ee of 90% was not altered. With these
new experimental conditions in hand, we screened the addition
of EtMgBr and MeMgBr to a large scope of acyclic R,ꢀ-
unsaturated aldehydes (Table 5).
Table 6. ACA of Various R′MgBr to S9 and S10 with L10
entry
R′
adduct
convn (%)a
1,4/1,2a
ee (%)b
1
2
3
4
5
6
7
8
9
10
(CH2)2Ph
i-Bu
i-Pr
c-Hex
Ph
n-Bu
(CH2)2Ph
i-Pr
ent-A3
ent-A4
ent-A5
ent-A6
ent-A7
ent-B2
ent-B3
ent-B5
ent-B6
ent-B7
100 (50)c
100
100
100
88
67:33
36:64
67:33
53:47
33:67
36:74
45:55
63:37
71:29
26:74
88, S
80, S
74, R
86, R
80, R
88, S
74, S
65, R
89, R
48, R
100
100
100
c-Hex
Ph
100 (60)c
96
Table 5. ACA of Et and MeMgBr to Various Enals with L10
a Determined by 1H NMR. b Determined by chiral GC or SFC. c Isolated
yield in parentheses.
The small alkyl chains of S9 and S10 have a profound effect.
Although the enantioselectivities remain moderate to good, the
1,4-regioselectivities are lower, particularly for the addition of
primary and secondary alkyl Grignard reagents to substrate S9
(with an ethyl group in the 4-position). The two exceptions were
observed for the addition of phenethylMgBr, which gave 88%
ee and a ratio 1,4/1,2 of 67:33 (entry 1), whereas the addition
of EtMgBr to S3 gave only 83% ee but 77% of 1,4-adduct
(Table 5, entry 3). For the addition of a phenyl group, a large
improvement was also obtained from 53 to 80% ee (Table 5
entry 7 vs Table 6 entry 5), even if the regioselectivity was not
perfect. Concerning the addition of Grignard reagents to crotonal-
dehyde S10, the best result was observed for the addition of
c-HexMgBr, which gave 89% ee and 71% of 1,4-adduct (entry
9). This result is better than the addition of MeMgBr to enal S6
bearing a cyclohexyl group (Table 5, entry 12).
entry
R
adduct
convn (%)a
1,4/1,2a
ee (%)b
1
2
3
4
5
6
7
8
C7H15
n-Bu
(CH2)2Ph
i-Bu
i-Pr
c-Hex
Ph
C7H15
n-Bu
(CH2)2Ph
i-Pr
A1
A2
A3
A4
A5
A6
A7
B1
B2
B3
B5
B6
B7
100 (75)c
100 (46)c
100 (59)c
100 (44)c
100
85:15
60:40
77:23
71:29
36:64
63:37
20:80
65:35
40:60
48:52
43:57
63:37
10:90
90, R
90, R
83, R
90, R
74, S
80, S
53, S
81, R
86, R
88, R
84, S
80, R
nd
100 (51)c
100
100 (40)c
100
9
10
11
12
13
97
86
c-Hex
100 (49)c
100
Ph
a Determined by 1H NMR. b Determined by chiral GC. c Isolated yield
in parentheses.
Finally, a brief test with R3Al reagents showed that the increased
Lewis acidity of these reagents was deleterious, as both 1,2 and
aldol products were obtained in significant amounts.
The best ee’s (up to 90%) were obtained with enals bearing
a primary alkyl group in the ꢀ-position (S1, S2, S3, and S4).
However, despite the improvement provided by TMSCl, the
regioselectivity was moderate, ranging from 71 to 85% for
primary alkyl chains (entries 1-4). With MeMgBr, both the
regio- and enantioselectivities were lower, with the notable
exception of enal S3, bearing a phenethyl group, which gave
88% ee (entry 10). It could be noticed that adduct B3 is a floral
fragrance called Citralis.
Finally, we screened the behavior of various Grignard
reagents to trans-2-pentenal S9 and crotonaldehyde S10 in order
to get the enantiomers of the adducts obtained from the addition
of EtMgBr or MeMgBr to various enals (Table 6).
To conclude, we have described the first copper-catalyzed
conjugate addition of dialkylzinc and Grignard reagents to
a large variety of R,ꢀ-unsaturated aldehydes with moderate
to excellent 1,4-regioselectivities and ee’s up to 90%. Key
to the success with Grignard reagents was the addition of
TMSCl. Further work is in progress to improve these results
through screening of new ligands.
Acknowledgment. We thank the Swiss National Research
Foundation (Grant No. 200020-126663) and COST action
D40 (SER Contract No. C07.0097) for financial support, as
well as Pr. E. P. Ku¨ndig and Audrey Mercier, from
University of Geneva, for the generous gift of ligand L3.
Supporting Information Available: Experimental pro-
cedures and full spectroscopic data for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(12) See Supporting Information for more details.
(13) (a) Corey, E. J.; Boaz, N. W. Tetrahedron Lett. 1985, 26,
6015-6018; 6019-6022. (b) Alexakis, A.; Berlan, J.; Besace, Y. Tetra-
hedron Lett. 1986, 27, 1047–1050. (c) Nakamura, E.; Matsuzawa, S.;
Horiguchi, Y.; Kuwajima, I. Tetrahedron Lett. 1986, 27, 4029–4032.
(14) Chuit, C.; Foulon, J. P.; Normant, J. F. Tetrahedron 1980, 36, 2305–
2310.
OL100441R
Org. Lett., Vol. 12, No. 9, 2010
1991