Aminohydroxyphosphine ligand for the copper-catalyzed enantioselective conjugate addition of organozinc reagents

Article abstract of DOI:10.1021/ol0618306

An alanine-derived aminohydroxyphosphine ligand was developed for copper-catalyzed asymmetric conjugate addition of organozinc reagents to α,β-unsaturated carbonyl compounds. This new tridentate ligand induces consistently high enantioselectivity in react

Full text of DOI:10.1021/ol0618306

ORGANIC
LETTERS
2006
Vol. 8, No. 18
4153-4155
Aminohydroxyphosphine Ligand for the
Copper-Catalyzed Enantioselective
Conjugate Addition of Organozinc
Reagents
Alakananda Hajra, Naohiko Yoshikai, and Eiichi Nakamura*
Department of Chemistry, The UniVersity of Tokyo, Bunkyo-ku, Tokyo 113-0033, Japan
nakamura@chem.s.u-tokyo.ac.jp
Received July 25, 2006
ABSTRACT
An alanine-derived aminohydroxyphosphine ligand was developed for copper-catalyzed asymmetric conjugate addition of organozinc reagents
to -unsaturated carbonyl compounds. This new tridentate ligand induces consistently high enantioselectivity in reactions of a variety of
acyclic substrates. Theoretical mechanistic analysis suggests that the C C bond formation takes place through a highly ordered transition
r,â
−
state by the coordination of the phosphorus and nitrogen atoms to the copper(III) and zinc(II) atoms, respectively, and of the oxygen anion
to both the metal centers.
Copper-catalyzed conjugate addition of a main group
organometallic reagent to an enone is an illustrative example
of the intricacy of multimetallic catalysis.1 It is this complex-
ity that has hampered the development of catalytic enantio-
selective additions until recently.2,3 A common feature of
the chiral ligands developed in the past is their multiligation
ability,2 but its function remained rather unclear. We sug-
gested some time ago that an effective ligand must effectively
coordinate to the copper(III) center on one hand and to the
main group metal center on the other.4 We report here a new
tridentate ligand 1 for the enantioselective conjugate addition
of an organozinc reagent to a variety of R,â-unsaturated
carbonyl compounds (Scheme 1b) developed on the basis
of this guiding principle. The ligand 1 was synthesized in
one step by addition of lithiated triphenylphosphine to N,N-
dibenzylalaninal (Scheme 1a).5 The modular synthesis is a
Scheme 1
useful feature of this ligand design. A working model of the
ligand acceleration of the reaction is also provided (Figure
1).
We recently reported a Ni-catalyzed cross-coupling reac-
tion of a Grignard reagent where a hydroxyphosphine ligand
was employed to promote synergy between nickel and
magnesium atoms.6 On the basis of these experiments and a
series of theoretical studies on organocopper reactions,7 we
modeled computationally C-C bond forming transition states
(1) Multimetallic Catalysts in Organic Synthesis; Shibasaki, M., Yama-
moto, Y., Eds.; Wiley-VCH: Weinheim, 2004.
(2) (a) Alexakis, A.; Benhaim, C. Eur. J. Org. Chem. 2002, 3221-3236.
(b) Feringa, B. L.; Naasz, R.; Imbos, R.; Arnold, L. A. In Modern
Organocopper Chemistry; Krause, N., Ed.; Wiley-VCH: Weinheim, 2002;
pp 224-258. (c) Krause, N.; Hoffmann-Ro¨der, A. Synthesis 2001, 171-
196.
10.1021/ol0618306 CCC: $33.50
© 2006 American Chemical Society
Published on Web 08/08/2006

Table 1. Enantioselective Conjugate Addition to
trans-3-Nonen-2-onea
entry ligand
Cu salt
solvent time/h yield/%
ee/%
1
1
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)
Cu(OAc)2
CuI
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
toluene
Et2O
3
11
3
91
80
90
87
86
90
85
55
15
90
85
78
39
45
98 (S)
-3
96
23
30
88
87
66
3
95
96
93
71
56
2
2
3b
4
1 + 2
3
4
1
1
1
1
1
1
1
1
1
3
5c
6
3
2
7
2
8
THF
12
12
3
9
DME
10
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
11
12
13
14
5
Figure 1. Schematic representations of a C-C bond formation
TS (A) and a π-complex (B) in the conjugate addition reaction
based on density functional calculations of a simplified model (R3
) Me or Cl, R4 ) H, R5 ) Me).
12
12
12
CuBr
CuBr‚Me2S CH2Cl2
a Conditions: 3 mol % of Cu salt, 3.6 mol % of ligand, 1.5 equiv of
Et2Zn, 0 °C. b A 9:1 mixture of 1 and 2 was used. c Ligand 4 was used as
a diastereomeric mixture (86:14).
(TSs) of the organocopper conjugate addition (cf. Figure 1)
and conceived a variety of possible ligand structures that
form rigid Cu/Zn bimetallic complexes. After experimental
screening of several practical candidates, the aminohydroxy-
phosphine 1 emerged as a promising ligand.
sensitive to the countercation of the copper source. Oxygen
anions are uniformly better than halide anions (entries 10-
14). The slower reaction rate and the inferior enantioselec-
tivity by CuBr‚Me2S than by CuBr suggest that Me2S
competes with the ligand 1. On the basis of these results,
we employed Cu(OTf)2 and CH2Cl2 at the synthetically
convenient temperature of 0 °C in the following studies, and
some other choices may also be effective.
Representative results of the screening of the ligand and
the reaction conditions as studied for the addition of Et2Zn
to trans-3-nonen-2-one are summarized in Table 1. The
ligand 1 gave an enantioselectivity of up to 98% ee at 0 °C
(entries 1, 6, 7, and 10-12), whereas its diastereomer 2 gave
virtually no selectivity with a much slower reaction rate
(entry 2). This is a practical merit of the use of 1 because a
90:10 mixture of diastereomers 1 and 2 as synthesized
(Scheme 1a) gives essentially the same enantioselectivity as
pure 1 (entry 3). Ligands derived from bulkier aminoalde-
hydes gave lower selectivity (entries 4 and 5). Choice of
solvent is important.2 Noncoordinating solvents are superior
to coordinating solvents including bidentate 1,2-dimethoxy-
ethane (DME, entries 6-9). The enantioselectivity is also
Table 2 illustrates representative results of the addition of
Me2Zn and Et2Zn to a variety of R,â-unsaturated carbonyl
compounds. The substrate scope of the present catalytic
system has proven to be very wide. Acyclic enones reacted
in uniformly high enantioselectivity (g98% ee) and good
yield (79-91%); they are known as poor substrates for this
class of reactions, and enantioselectivities of >95% are
rare.2,3a,b,d-h,k,l The reaction was applicable to chalcone (entry
1) and aliphatic enones with â-aryl (entries 2-7) and alkyl
(entries 8-11) substituents. The electronic (entries 3-5) and
steric (entries 9 and 11) properties of the â-substituents did
not affect the selectivity. Furyl and thienyl substituents did
not show any adverse effect on the reaction (entries 6 and
7). The two organozinc reagents showed the same level of
selectivity of 98% ee: Me2Zn was less reactive but still gave
good yields (entries 2, 3, 8, and 9).3f The reaction can be
carried out on as much as a 10-mmol scale without significant
decrease of enantioselectivity (entry 10). The reaction of a
trisubstituted cyclic enone proceeded in excellent selectivity
of >98% ee (entry 12).3d Cyclic enones are poorer substrates
for the ligand 1 suggesting in turn that s-cis conformers take
part in the reaction of acyclic enones: 2-cyclohexenone was
only moderately selective (82% ee, entry 13), and 2-cyclo-
heptenone was poorly selective (16% ee, see Supporting
Information). R,â-Unsaturated esters are unreactive, but the
corresponding imide took part in the reaction smoothly and
gave excellent selectivity of 98.7% ee (entry 14).3e
(3) (a) Mizutani, H.; Degrado, S. J.; Hoveyda, A. H. J. Am. Chem. Soc.
2002, 124, 779-781. (b) Shintani, R.; Fu, G. C. Org. Lett. 2002, 4, 3699-
3702. (c) Alexakis, A.; Benhaim, C.; Rosset, S.; Humam, M. J. Am. Chem.
Soc. 2002, 124, 5262-5263. (d) Degrado, S. J.; Mizutani, H.; Hoveyda, A.
H. J. Am. Chem. Soc. 2002, 124, 13362-13363. (e) Hird, A. W.; Hoveyda,
A. H. Angew. Chem., Int. Ed. 2003, 42, 1276-1279. (f) Cesati, R. R.; De
Armas, J.; Hoveyda, A. H. J. Am. Chem. Soc. 2004, 126, 96-101. (g)
Duncan, A. P.; Leighton, J. L. Org. Lett. 2004, 6, 4117-4119. (h) Morimoto,
T.; Mochizuki, N.; Suzuki, M. Tetrahedron Lett. 2004, 45, 5717-5722. (i)
Lo´pez, F.; Harutyunyan, S. R.; Minnaard, A. J.; Feringa, B. L. J. Am. Chem.
Soc. 2004, 126, 12784-12785. (j) Feringa, B. L.; Badorrey, R.; Pena, D.;
Harutyunyan, S. R.; Minnaard, A. J. Proc. Natl. Acad. Sci. U.S.A. 2004,
101, 5834-5838. (k) Takahashi, Y.; Yamamoto, Y.; Katagiri, K.; Danjo,
H.; Yamaguchi, K.; Imamoto, T. J. Org. Chem. 2005, 70, 9009-9012. (l)
Ito, K.; Eno, S.; Saito, B.; Katsuki, T. Tetrahedron Lett. 2005, 46, 3981-
3985. (m) d’Augustin, M.; Palais, L.; Alexakis, A. Angew. Chem., Int. Ed.
2005, 44, 1376-1378. (n) Lo´pez, F.; Harutyunyan, S. R.; Meetsma, A.;
Minnaard, A. J.; Feringa, B. L. Angew. Chem., Int. Ed. 2005, 44, 2752-
2756. (o) Brown, M. K.; Degrado, S. J.; Hoveyda, A. H. Angew. Chem.,
Int. Ed. 2005, 44, 5306-5310. (p) Mazery, R. D.; Pullez, M.; Lo´pez, F.;
Harutyunyan, S. R.; Minnaard, A. J.; Feringa, B. L. J. Am. Chem. Soc.
2005, 127, 9966-9967. (q) Hird, A. W.; Hoveyda, A. H. J. Am. Chem.
Soc. 2005, 127, 14988-14989. (r) Fillion, E.; Wilsily, A. J. Am. Chem.
Soc. 2006, 128, 2774-2775.
(4) Nakamura, E.; Yamanaka, M.; Mori, S. J. Am. Chem. Soc. 2000,
122, 1826-1827.
(5) Reetz, M. T. Chem. ReV. 1999, 99, 1121-1162.
(8) (a) Yamanaka, M.; Nakamura, E. Organometallics 2001, 20, 5675-
5681. (b) Canisius, J.; Gerold, A.; Krause, N. Angew. Chem., Int. Ed. 1999,
38, 1644-1646. (c) Frantz, D. E.; Singleton, D. A.; Snyder, J. P. J. Am.
Chem. Soc. 1997, 119, 3383-3384. (d) Woodward, S. Chem. Soc. ReV.
2000, 29, 393-401.
(6) Yoshikai, N.; Mashima, H.; Nakamura, E. J. Am. Chem. Soc. 2005,
127, 17978-17979.
(7) Nakamura, E.; Mori, S. Angew. Chem., Int. Ed. 2000, 39, 3750-
3771.
4154
Org. Lett., Vol. 8, No. 18, 2006

Being aware of the complexity of the conjugate addition
of lithium organocuprates7-9 and, especially, that of copper-
catalyzed addition of organozinc and magnesium reagents,10
we consider it premature to discuss details of the mechanism
of the whole catalytic cycle. Therefore, we only discuss the
C-C bond forming stage of the reaction, which most likely
comprises the step where the enantioselectivity is determined.
Figure 1 shows a TS model of the addition of a heterocuprate
complex of MeCu and 1 to an s-cis-acrolein (A, focused on
the substrate) based on a computational modeling of a
simplified model (R3 ) Me or Cl, R4 ) H, R5 ) Me;
B3LYP/6-31G(d) and equivalent basis sets).11 The structure
of the π-complex (B, focused on the Cu atom) preceding
the TS is essentially the same as the TS structure.
Table 2. Enantioselective Conjugate Addition to
R,â-Unsaturated Carbonyl Compoundsa
The first important common structural feature is that the
Lewis acidic zinc atom and the nucleophilic copper atom
are properly located to effect ligand acceleration of the
conjugate addition. The central oxygen anion plays the role
of a counteranion for the zinc atom and that of a basic ligand
to the copper atom to form a reactive cuprate complex.
S-trans-Acrolein where the molecule is in its extended
conformation does not fit into this rigid coordination
environment. The large difference between 1 and 2 can be
rationalized on the basis of the positioning of the R4 methyl
group that controls the relative orientation between the nearby
benzene group and the nitrogen substituent.
The second important feature is that the zinc(II) geometry
is tetrahedral and the copper(III) geometry is square-
pyramidal (cf. B): one Me group, a â-carbon atom, a
phosphorus atom, and the zinc enolate occupy the equatorial
positions, and the alkoxide coordinates from the apical site.
The activation energy of reductive elimination is quite low
(ca. 11 kcal/mol) because of the relatively weak donating
ability of the ligands.12
a The reaction was carried out on a 1-mmol scale as described in Scheme
1b unless otherwise noted. b Isolated yields. c The selectivity was determined
by chiral HPLC (entries 1-5), GLC (entries 6-12 and 14), or NMR analysis
after derivatization to diastereomeric ketals (entry 13). d 2 equiv of Me2Zn
was used. e The reaction was carried out on a 10-mmol scale. f 5 mol % of
Cu(OTf)2 and 6 mol % of 1 were used.
(9) (a) Yoshikai, N.; Yamashita, T.; Nakamura, E. Angew. Chem., Int.
Ed. 2005, 44, 4721-4723. (b) Yoshikai, N.; Yamashita, T.; Nakamura, E.
Chem. Asian J., published online (DOI: 10.1002/asia.200600034).
(10) (a) Kitamura, M.; Miki, T.; Nakano, K.; Noyori, R. Bull. Chem.
Soc. Jpn. 2000, 73, 999-1014. (b) Nakano, K.; Bessho, Y.; Kitamura, M.
Chem. Lett. 2003, 32, 224-225. (c) Gallo, E.; Ragaini, F.; Bilello, L.;
Cenini, S.; Gennari, C.; Piarulli, U. J. Organomet. Chem. 2004, 689, 2169-
2176. (d) Pfretzschner, T.; Kleemann, L.; Janza, B.; Harms, K.; Schrader,
T. Chem.-Eur. J. 2004, 10, 6048-6057. (e) Harutyunyan, S. R.; Lo´pez,
F.; Browne, W. R.; Correa, A.; Pen˜a, D.; Badorrey, R.; Meetsma, A.;
Minnaard, A. J.; Feringa, B. L. J. Am. Chem. Soc. 2006, 128, 9103-9118.
(11) Yamanaka, M.; Inagaki, A.; Nakamura, E. J. Comput. Chem. 2003,
24, 1401-1409.
(12) A square planar geometry has been described for the intermediate
of the conjugate addition of a lithium organocuprate (Me2CuLi),8a and the
four ligands on the copper atom consist of two Me groups, a â-carbon atom,
and a lithium enolate moiety. These σ-donating ligands stabilize the high
oxidation state of the copper atom, and therefore, the activation energy of
reductive elimination is rather high (158a-188b kcal/mol with 4,4-dimeth-
ylcyclohexenone).
Acknowledgment. We thank the Ministry of Education,
Culture, Sports, Science and Technology of Japan for a
Grant-in-Aid for Scientific Research (S) and for the 21st
Century COE Program for Frontiers in Fundamental Chem-
istry. A.H. thanks the Japan Society for Promotion of Science
for a fellowship.
Supporting Information Available: Details of experi-
mental and computational procedures, spectral data for new
compounds, and Cartesian coordinates for optimized struc-
tures (PDF). This material is available free of charge via
the Internet at http://pubs.acs.org.
OL0618306
Org. Lett., Vol. 8, No. 18, 2006
4155