A novel example of chiral counteranion induced enantioselective metal catalysis: The importance of ion-pairing in copper-catalyzed olefin aziridination and cyclopropanation

Article abstract of DOI:10.1021/ol000303y

(equation presented) A new ion-pairing route to achieve asymmetric catalysis has been observed in the copper-catalyzed aziridination of styrene with a chiral counteranion. Structural studies suggest that enantioinduction occurs via ion-pairing of the cati

Full text of DOI:10.1021/ol000303y

ORGANIC
LETTERS
2000
Vol. 2, No. 26
4165-4168
A Novel Example of Chiral Counteranion
Induced Enantioselective Metal
Catalysis: The Importance of
Ion-Pairing in Copper-Catalyzed Olefin
Aziridination and Cyclopropanation
David B. Llewellyn, Dan Adamson, and Bruce A. Arndtsen*
Department of Chemistry, McGill UniVersity, 801 Sherbrooke Street West,
Montreal, Quebec H3A 2K6, Canada
bruce_arndtsen@maclan.mcgill.ca
Received October 6, 2000
ABSTRACT
A new ion-pairing route to achieve asymmetric catalysis has been observed in the copper-catalyzed aziridination of styrene with a chiral
counteranion. Structural studies suggest that enantioinduction occurs via ion-pairing of the cationic copper catalyst in the chiral pocket
created by the anion. The degree of asymmetric induction can be tuned with features that affect ion-pairing, such as achiral and chiral ligands,
temperature, and solvent polarity.
Asymmetric synthesis using chiral transition metal complexes
represents one of the more important methods to generate
optically active molecules.1 Traditionally, chirality has been
incorporated into metal catalysis through the design and
synthesis of chiral or pro-chiral ligands and their coordination
to the metal center.2 However, asymmetric induction is an
extremely sensitive phenomenon, with high enantioselec-
tivities requiring only very small differences in transition
state energies.3 This implies that it may be possible to use
more subtle features than metal-ligand interactions to effect
chirality in metal-catalyzed reactions, such as fragments
weakly associated outside the metal coordination sphere
(Figure 1).4-6 Formally nonbonding interactions provide an
attractive potential source of asymmetry in catalysis since
they can, in principle, incorporate chirality into active metal
catalysts without significantly perturbing reactivity. In ad-
dition, weakly associated fragments are easily modified,
allowing them to be readily scanned to achieve high levels
of enantioselectivity.
Since many metal catalysts are cationic, the use of chiral
counteranions may provide a method to effect asymmetric
induction.7,8 There is emerging evidence that noncoordinating
or weakly coordinating counteranions can have dramatic
(1) Ojima, I. Catalytic Asymmetric Synthesis; VCH Publishers: New
York, 1993. (b) Cervinka, O. EnantioselectiVe Reactions in Organic
Chemistry; Ellis Horwood: London, 1995.
(2) Seyden-Penne. Chiral Auxilaries and Ligands in Asymmetric Syn-
thesis; Wiley: New York, 1995.
(3) For example, 3 kcal/mol ) ca. 95% ee at 25 °C.
Figure 1. A nonbonding approach to chiral induction.
10.1021/ol000303y CCC: $19.00 © 2000 American Chemical Society
Published on Web 12/02/2000

effects on selectivity in metal-catalyzed reactions.9 While
this suggests that significant ionic interactions can occur
during catalysis, to our knowledge, the counteranion itself
has never been used to achieve asymmetric induction.10
We report below the first example of an enantioselective
metal-catalyzed reaction where the sole source of asym-
metry is a chiral counteranion. These results demonstrate that
ionic interactions in metal catalysis are not only important
but can be utilized to influence chirality in reaction pro-
ducts.
with either the coordinating ability or size of the anion. In
fact, the trend changes with each ligand employed. In
addition, the ee’s are independent of X- in the more polar
acetonitrile solvent. This demonstrates that strong ion-pairing
does exist in Cu+ X- in nonpolar solvents, suggesting it
would be a reasonable system to explore the effect of a chiral
anion on enantioselectivity.
The chiral counteranion prepared, 3-, contains two bi-
naphthol units bound to a tetrahedral boron center (Scheme
1).14 The chiral binaphthols generate a single C2 symmetric
The systems examined for counteranion influence on chiral
induction are copper-catalyzed nitrene11 and carbene12 trans-
fer reactions to prochiral R-olefins. Prior to exploring the
influence of chiral counteranions on these reactions, the
degree of ion-pairing in Cu+ X- during catalysis was
determined. As illustrated in Table 1, variation of the weakly
Scheme 1
Table 1. Counteranion Influence on CuX-Catalyzed Styrene
Aziridinationa
enantiomer upon complexation to boron and therefore require
no subsequent resolution. The silver salt of (R)-3- can be
readily generated by the reaction of 2 equiv of R-binaphthol
ligand
X
% ee (C6H6)b
% ee (MeCN)b
(R)-1
(R)-1
(R)-1
(R)-1
(S)-2
(S)-2
(S)-2
(S)-2
OTf
ClO4
Cl
PF6
OTf
ClO4
Cl
1 (S)
5 (S)
28 (S)
28 (S)
28 (S)
28 (S)
2 (R)
2 (R)
2 (R)
2 (R)
15
with H2BBr‚SMe2 in dichloromethane followed by slow
addition to Ag2CO3 in acetonitrile. Ag+ (R)-3- provides a
useful precursor for incorporating 3- into metal halides.
The reaction of Ag+ (R)-3- with CuCl in acetonitrile,
followed by precipitation with diethyl ether, provides Cu+
(R)-3-‚4NCMe as a white solid in 79% yield. The S-
enantiomer, Cu+ (S)-3-, can be similarly prepared from
S-binaphthol.
17 (S)
33 (S)
66 (R)
57 (R)
26 (R)
33 (R)
PF6
a [styrene]/[PhINTs] ) 5. b Major enantiomer in parentheses, determined
by HPLC on a (S,S) Whelk-O 1 column (Regis).11c
The crystallization of Cu+ (R)-3- with four acetonitrile
molecules suggests that the copper complex exists as the
+
ionic 18-electron complex Cu(NCMe)4 (R)-3-. However,
coordinating counteranion (X ) OTf, ClO4, PF6) in the CuX-
catalyzed aziridination of styrene with PhINTs13 in the
presence of chiral bis(oxazoline) ligands 1 and 2 leads to
changes in ee’s of over 30% in benzene solvent. This anion
influence on enantioselectivity displays no consistent trend
X-ray structural analysis reveals two isomeric complexes
(9) For representative examples, see: (a) Evans, D. A.; Kozlowski, M.
C.; Murry, J. A.; Burgey, C. S.; Campos, K. R.; Connell, B. T.; Staples, R.
J. J. Am. Chem. Soc. 1999, 121, 669-685. (b) Lanza, G.; Fragala, I. L.;
Marks, T. J. J. Am. Chem. Soc. 1998, 120, 8257-8258. (c) Johannsen, M.;
Jorgensen, K. A. J. Chem. Soc., Perkin Trans. 1997, 1183-1185. (d) Trost,
B. M.; Bunt, R. C. J. Am. Chem. Soc. 1998, 120, 70-79. (e) Bayersdorfer,
R.; Ganter, B.; Englert, U.; Keim, W.; Vogt, D. J. Organomet. Chem. 1998,
552, 187-194.
(10) Buriak, M.; Osborn, J. A. Organometallics 1996, 15, 3161-3169.
(11) (a) Jacobsen, E. N. In ComprehensiVe Asymmetric Catalysis;
Jacobsen, E. N., Pfaltz, A., Yamamoto, H., Eds.; Springer: Berlin, 1999;
p 607. (b) Evans, D. A.; Faul, M. M.; Bilodeau, M. T.; Anderson, B. A.;
Barnes, D. M. J. Am. Chem. Soc. 1993, 115, 5328-5329. (c) Li, Z.; Quan,
P. W.; Jacobsen, E. N. J. Am. Chem. Soc. 1995, 117, 5889-5890.
(12) (a) Zollinger, H. Diazo Chemistry II, VCH: Weinheim, 1995. (b)
Evans, D. A.; Woerpel, K. A.; Hinman, M. M.; Faul, M. M. J. Am. Chem.
Soc. 1991, 113, 726-728.
(4) Sato, K. Kadowaki, K. Soai, Angew. Chem., Int. Ed. 2000, 39, 1510-
1512.
(5) Chiral phase transfer catalysts have been shown to induce high ee’s:
(a) Hiyama, T.; Mishima, T.; Sawada, H.; Nozaki, H. J. Am. Chem. Soc.
1975, 97, 1626-1627. (b) Hughes, D. L.; Dolling, U. H.; Ryan, K. M.;
Schoenewaldt, E. F.; Grabowski, E. J. J. J. Org. Chem. 1987, 52, 4745-
4752.
(6) For chiral environment approaches to catalysis: (a) Walsh, P. J.;
Balsells, J. J. Am. Chem. Soc. 2000, 122, 1802-1803 and references therein.
(b) Brunkan, N. M.; Gagne, M. R. J. Am. Chem. Soc. 2000, 122, 6217.
(7) Solution ion-pairing with a chiral anion has been recently reported
to resolve a chiral iron complex: Lacour, J.; Jodry, J. J.; Ginglinger, C.;
Torche-Haldimann, S. Angew. Chem., Int. Ed. 1998, 37, 2379-2380
(8) For recent examples of the use of chiral counterions, see: (a) Owen,
D. J.; Schuster, G. B. J. Am. Chem. Soc. 1996, 118, 259-260. (b) Owen,
D. J.; VanDerveer, D.; Schuster, G. B. J. Am. Chem. Soc. 1998, 120, 1705-
1717. (c) Schlitzer, D. S.; Novak, B. M. J. Am. Chem. Soc. 1998, 120,
2196-2197. (d) Chang; S.; Galvin, J. M.; Jacobsen, E. N. J. Am. Chem.
Soc. 1994, 116, 6937.
(13) Yamada, T.; Yamamoto, M.; Okawara. Chem. Lett. 1975, 361-
362.
(14) Ishihara, M.; Miyata, K.; Hattori, T.; Tada, H.; Yamamoto. J. Am.
Chem. Soc. 1994, 116, 10520-10524.
(15) Kaufmann, R.; Boese. Angew. Chem., Int. Ed. Engl. 1990, 29, 545-
546.
4166
Org. Lett., Vol. 2, No. 26, 2000

cocrystallizing in a 1:1 ratio: the tetrakis(acetonitrile) copper-
(I) salt as well as a second isomer in which the anion, (R)-
Table 2: Aziridination and Cyclopropanation of Styrene with
Cu+ 3- a
+
3-, is coordinated through an oxygen atom to a Cu(NCMe)3
fragment (Cu-O, 2.16 Å; Figure 2).16 Nevertheless, the 1H
entry
L
X
A
solvent
% eeb
22 (S)
yield, %
1
2
3
4
5
6
(R)-1 (S)-3 NTs
(R)-1 (R)-3 NTs
(S)-2 (S)-3 NTs
(S)-2 (R)-3 NTs
C6H6
C6H6
C6H6
C6H6
75
85
nd
nd
12
20
34
34
86
88
97
87
43
41
90
78
24 (S)
13 (R)
12 (R)
(S)-1 (S)-3 CHCO2Et C6H6
6 (R)/21 (R)c
28 (S)/10 (R)c
6 (R)/15 (R)c
8 (R)/12 (R)c
7 (R)
(S)-1 (R)-3 CHCO2Et C6H6
7d (S)-1 (S)-3 CHCO2Et C6H6
8d (S)-1 (R)-3 CHCO2Et C6H6
9
10
11
12
13
14
15
16
none (R)-3 NTs
none (S)-3 NTs
none (R)-3 NTs
none (R)-3 NTs
C6H6
Figure 2. Crystal structure of Cu(NCMe)3((R)-3-) (hydrogen atoms
and Cu(NCMe)4+ (R)-3- omitted for clarity). Selected bond lengths
[Å]: Cu(1)-O(3) 2.165(2), Cu(1)-N(1) 2.083(6), Cu(1)-N(2)
1.916(4), Cu(1)-N(3) 1.972(4).
C6H6
7 (S)
CH2Cl2
4 (R)
CH3CN <1
4
5
6
7
(R)-3 NTs
(R)-3 NTs
(R)-3 NTs
(R)-3 NTs
C6H6
C6H6
C6H6
C6H6
10 (R)
10 (R)
2 (R)
<1
NMR (CD2Cl2) of Cu+ (R)-3- reveals that the weakly
coordinating anion is completely ionized to form the tetrakis-
(acetonitrile)copper salt in solution, with a single resonance
for the four copper-bound CH3CN molecules (δ 1.48) and
equivalent binaphthol units within (R)-3-.17
a [styrene]/[PhINTs] ) 5 at 0 °C. Aziridination: 3 mol % of Cu+3-,
3.1 mol % of L, analyzed by HPLC on an (S,S) Whelk-O 1 column (Regis).
Cyclopropanation: 1 mol % of Cu+ 3-, 1.1 mol % of L, analyzed by GC
on a CP Chirasil-Dex column.12b b Major enantiomer at C1 in parentheses.
c ee(cis)/ee(trans). d 25 °C.
With the chiral copper salt in hand, the potential influence
of the chiral counteranion on enantioselectivity was explored.
Variation of (R)- or (S)-3- in the copper-catalyzed aziridi-
nation of styrene with chiral bis(oxazoline) ligands 1 and 2
at 0 °C reveals that the chirality of the counteranion has a
minor influence on the enantioselectivity of the reaction (ca.
2%; Table 2, entries 1-4). This is despite the differences in
both the solubility and color of these catalysts at the
beginning of the reaction. On the other hand, the analogous
cyclopropanation of styrene shows a dramatic anion influence
on enantioinduction. The reaction of ethyldiazoacetate and
styrene in the presence of Cu+ (R)-3- or (R)-3- and ligand
1 at 0 °C leads to the formation of trans- and cis-
cyclopropanes in a 1.4 and 2.2 cis/trans ratio, respectively
(entries 5 and 6). Examination of the cis-isomer of the
cyclopropane reveals that modification of the anion chirality
from (R) to (S) not only influences enantioselectivity (28%
to 6%) but actually changes the preferred enantiomer
generated! The effect of ion-pairing in Cu+ 3- is very
temperature sensitive, and performing the same reaction at
25 °C results in a complete loss of the anion’s effect on ee’s
and catalyst activity (entries 7 and 8). This demonstrates that,
under the correct conditions, the chirality of the counteranion
can indeed influence the asymmetric environment about the
metal-ligand complex.
The ability of the chiral counteranion by itself, without
the need of a chiral ligand, to directly affect asymmetry in
olefin aziridination is illustrated in the reaction of PhINTs
and styrene with Cu+ (R)-3- (entry 9). In contrast to other
copper salts, analysis of the aziridine product shows that Cu+-
(R)-3-, solely through the influence of ion-pairing interac-
tions, produces aziridine in 7% enantiomeric excess. When
the (S) enantiomer of counteranion 3- is employed, the
opposite enantiomer of the aziridine is obtained in 7%
enantiomeric excess (entry 10). To our knowledge, this
represents the first example of an enantioselective transition
metal catalyzed reaction where the sole source of chirality
is the counteranion.
(16) Crystallographic data for Cu+ (R)-3-: Mw ) 1614.40, monoclinic,
space group P2(1), a ) 11.113(4), b ) 25.109(7), c ) 14.869(3), b )
104.84(3), V ) 4011(2) Å3, Z ) 2, Fcalcd ) 1.337 mg m-3, T ) 293(2) K,
m ) 1.173 mm-1; R (Rw) ) 0.071 (0.1887). Crystallographic data for 8:
Mw ) 1023.52, monoclinic, space group P2(1), a ) 12.7853(3), b )
15.7945(3), c ) 13.2045(2), b ) 101.9080(10), V ) 2609.10(9) Å3, Z ) 2,
Fcalcd ) 1.318 mg m-3, T ) 293(2), m ) 2.164 mm-1; R (Rw) ) 0.0968
(0.2461). Full details of the crystallographic analysis are described in the
Supporting Information.
As anticipated, the enantioinduction obtained from the
chiral counteranion is very sensitive to factors which affect
the degree and nature of ion-pairing during catalysis. Thus,
changes in solvent polarity from benzene (E ) 2.3) to
(17) The symmetry in (R)-3- could also arise from an equilibrium
coordination to copper. However, the addition of 10 equiv of CH3CN, or 2
equiv of ligand 1, which would perturb any coordination, has almost no
1
effect upon the H NMR of (R)-3-.
Org. Lett., Vol. 2, No. 26, 2000
4167

methylene chloride (E ) 9.1) to acetonitrile (E ) 38.8) results
in a gradual diminishment of ee’s (entries 9-12), consistent
with their effect on ion-pairing.18 The cation/anion interac-
tions can be further tuned by changes in the copper
coordination sphere. The addition of 2,2′-bipyridine (4) or
1,10-phenanthroline (5) results in an actual enhancement of
enantioselectivity to 10% (entries 13 and 14). This increase
in % ee upon addition of a bidentate ligand argues against
the coordination of (R)-3- to the copper center during
catalysis (vide infra) and instead likely arises from ligand
coordination changing the preferred orientation of (R)-3-
about the metal center. Increasing the steric bulk of the ligand
(entries 15 and 16) appears to inhibit tight ion-pairing and
leads to an almost complete loss of chiral induction from
the counteranion.
structure clearly shows that, under the conditions employed
for catalysis, the borate counteranion 3- cannot compete with
the bidentate ligand and styrene for coordination to the
copper center. Interestingly, the counteranion does appear
to have a long-range ion-pairing interaction through one
oxygen of the binaphthol fragment (Cu-O, 2.55 Å in 8,
compared to 2.16 Å in Cu+ (R)-3-). The net effect of this
interaction is to place the copper cation within a chiral pocket
created by two separate binaphthol fragments on the boron
center. This ion-pairing induces the coordination of only a
single styrene enantioface in 8, in which the phenyl group
is directed away from the bulky binaphthol rings. While a
direct extrapolation of this solid state structure to the
catalytically active species in solution cannot be made, this
does suggest that strong ion-pairing in Cu+ (R)-3- can lead
to a preferred orientation of the catalyst in the chiral
environment of the counteranion.
In conclusion, a new method to induce asymmetry into
metal catalysis has been observed via the use a chiral
counteranion with an achiral metal complex. This illustrates
that weak forces such as ion-pairing can indeed be employed
to affect enantioselectivity in catalysis. The significant
influence observed with the chiral counteranion in concert
with chiral ligands suggests that the further design of these
anions can lead to even higher levels of ion-pairing induced
enantioinduction. Considering the flexibility in counteranion
structures, the tunability of their influence on the metal center
(i.e., solvent polarity, temperature, and ligand effects), and
their ease of incorporation into catalysts, such cation/anion
interactions would appear prime candidates for exploration
in the future development of enantioselective metal-catalyzed
processes. Efforts in these directions are currently under
investigation in our laboratory.
All experimental evidence suggests that ion-pairing be-
tween Cu+ and 3- is the source of enantioselectivity.19 To
obtain further insight into the nature of this cation/anion
interaction, the structure of complex Cu+ (R)-3- was
examined under the actual catalysis conditions. The addition
of Cu+ (R)-3- to bipyridine and 5 equiv of styrene in CH2-
Cl2, which models the conditions employed above for styrene
aziridination, followed by the slow diffusion of pentane
results in the crystallization of the copper salt as a yellow
solid. X-ray structural analysis of this complex reveals it to
be (bipy)Cu(H2CdCHC6H5)+ (R)-3- (8) (Figure 3).16 This
Acknowledgment. We thank Dr. Anne-Marie Lebuis for
determination of the crystal structures of Cu+ (R)-3- and 8
and NSERC (Canada) and FCAR (Quebec) for their financial
support of this research. D.B.L. thanks NSERC for a graduate
fellowship.
Supporting Information Available: Synthesis and spec-
troscopic and X-ray structural data on complexes Cu+ (R)-
3- and 8. This material is available free of charge via the
Internet at http://pubs.acs.org.
OL000303Y
(18) Loupy, B. Tchoubar. Salt Effects in Organic and Organometallic
Chemistry, 1st ed.; VCH: Weinheim, 1992.
(19) Control experiments demonstrate that enantioselectivity remains
continuous throughout catalysis, while 11B NMR analysis shows that (R)-
3- decomposes slowly and only at the end of the reaction, demonstrating
that asymmetric induction is due to an intact (R)-3-. We thank the reviews
for this suggestion.
Figure 3. Crystal structure of 8 (hydrogen atoms omitted for
clarity). Selected bond lengths [Å] and angles [deg]: Cu-O(3)
2.546(5), Cu-C(11) 2.022(10), Cu-C(12) 2.021(9); N(1)-Cu-O(3)
84.0(2), N(2)-Cu-O(3) 91.3(2), C(11)-Cu-O(3) 105.5 (3), C(12)-
Cu-O(3) 108.2(3).
4168
Org. Lett., Vol. 2, No. 26, 2000