Redox-configurable ambidextrous catalysis: Structural and mechanistic insight

Article abstract of DOI:10.1039/c5sc02144h

A ligand capable of adopting two pseudo-enantiomeric helically chiral states when bound to copper has been applied as an asymmetric catalyst in the Michael addition of malonate substrates to nitrostyrenes. The absolute configuration of the helically chiral ligand is inverted upon oxidation/reduction of the copper center. In this way, the handedness of the Michael addition product (R/S) can be selected based on the handedness of the catalyst (Λ/Δ). Exciton coupled circular dichroism (ECCD) was used to identify which of the two pseudo-enantiomeric forms the catalyst adopted after reduction/oxidation, with additional support from X-ray crystallographic data. The synthesis of the ligand was achieved in five steps with an overall 61% yield. Enantiomeric excesses of the Michael addition products of up to 72% (S) and 70% (R) were obtained in acetonitrile. The ability to choose the handedness of the product based on the chiral state of the catalyst has been demonstrated with several different solvents, bases, nitrostyrene/malonate substrates, and prochiral malonate substrates. A combination of molecular modelling, crystal structure and kinetic data suggest that one urea moiety of the catalyst ligand likely binds the nitrostyrene substrate while blocking the Re face of the nitrostyrene in the transition state.

Full text of DOI:10.1039/c5sc02144h

Chemical
Science
View Article Online
View Journal | View Issue
EDGE ARTICLE
Redox-configurable ambidextrous catalysis:
structural and mechanistic insight†
Cite this: Chem. Sci., 2015, 6, 5904
Shahab Mortezaei, Noelle R. Catarineu, Xueyou Duan, Chunhua Hu
*
and James W. Canary
A ligand capable of adopting two pseudo-enantiomeric helically chiral states when bound to copper has
been applied as an asymmetric catalyst in the Michael addition of malonate substrates to nitrostyrenes.
The absolute configuration of the helically chiral ligand is inverted upon oxidation/reduction of the
copper center. In this way, the handedness of the Michael addition product (R/S) can be selected
based on the handedness of the catalyst (L/D). Exciton coupled circular dichroism (ECCD) was used
to identify which of the two pseudo-enantiomeric forms the catalyst adopted after reduction/
oxidation, with additional support from X-ray crystallographic data. The synthesis of the ligand was
achieved in five steps with an overall 61% yield. Enantiomeric excesses of the Michael addition
products of up to 72% (S) and 70% (R) were obtained in acetonitrile. The ability to choose the
handedness of the product based on the chiral state of the catalyst has been demonstrated with
several different solvents, bases, nitrostyrene/malonate substrates, and prochiral malonate substrates.
A combination of molecular modelling, crystal structure and kinetic data suggest that one urea
moiety of the catalyst ligand likely binds the nitrostyrene substrate while blocking the Re face of the
nitrostyrene in the transition state.
Received 13th June 2015
Accepted 13th July 2015
DOI: 10.1039/c5sc02144h
www.rsc.org/chemicalscience
dynamic catalyst. Here we provide additional characterization,
scope of reaction studies, and mechanistic insight into the
asymmetric ambidextrous catalyst.15
Introduction
The ability to dynamically control catalysts, either in situ or
before reaction, provides the potential to attain complex target
compounds in far fewer steps. The ability of one catalyst to
mediate multiple transformations reduces waste and cuts costs
associated with other static catalysts. Some initial work in this
area has employed catalytic scaffolds such as ferrocenes,
strained alkenes, rotaxanes, calixarene, and other building
blocks to design dynamic catalysts.1–13 So far, most designed
catalysts tend to use heat, light, or redox state as a means of
catalyst control. Scaffolds that are capable of triggered chirality
inversion have been employed to design several molecular
machines with different functionalities. More recently, our
group14 and that of Feringa3,14 have reported catalysts that
utilize these helically chiral scaffolds to control the enantiose-
lectivity of asymmetric conjugate addition reactions. Although
much progress has been made developing dynamic catalysts,
most of these examples have been limited to proof of concept
studies. The efficient synthesis and straightforward use of the
ambidextrous catalyst lend themselves well to more compre-
hensive and in-depth studies not usually associated with a
Design
A copper complex capable of redox-triggered inversion of helical
chirality was discovered several years ago (Fig. 1).16–18 Exciton
coupled circular dichroism (ECCD) was used as an optical
readout to differentiate between the two possible pseudo-
enantiomeric states. Chromophores were attached to the scaf-
fold to monitor ECCD properties at different wavelengths. Aer
establishing that chromophores could be appended to the
complex without destroying the readout, we sought to attach a
catalytic group to the scaffold in hopes of designing a catalyst
that possessed some form of dynamic control.
Department of Chemistry, New York University, New York, New York 10003, USA.
E-mail: canary@nyu.edu
† Electronic supplementary information (ESI) available: Control studies. CCDC
1405638. For ESI and crystallographic data in CIF or other electronic format see
DOI: 10.1039/c5sc02144h
Fig. 1 Helical inversion of a methionine-based copper complex.16
5904 | Chem. Sci., 2015, 6, 5904–5912
This journal is © The Royal Society of Chemistry 2015

View Article Online
Edge Article
Chemical Science
Fig. 2 Model used for design of a helically chiral scaffold with
appended urea groups.
With a chiral scaffold already in hand, we looked to the eld
of asymmetric catalysis for inspiration. Several chiral scaffolds
have already been appended with either organocatalytic or
metal-containing catalytic moieties to create new asymmetric
catalysts.19–21 We reasoned that a catalyst attached to an asym-
metric scaffold capable of inverting its chirality would be
capable of catalyzing the formation of either enantiomer
product depending on the enantiomeric state of the scaffold. Of
the organocatalysts discovered, the urea and thiourea subset
have proven to be some of the most versatile and robust. (Thio)
urea catalysts have been employed in a diverse set of reactions,
such as cycloadditions, Michael additions, and condensa-
tions.22–24 The mechanism of each of these reactions most oen
features the (thio)urea group activating a substrate by hydrogen-
bonding in a Lewis acidic interaction.25,26 The compatibility of
urea groups with the chiral scaffold led to their selection for the
ambidextrous catalyst.
The asymmetric conjugate addition of malonate esters to
nitroalkenes has been performed several times with a diverse
array of urea-containing organocatalysts.27–31 The heavily
studied nitrostyrene conjugate addition was chosen as a model
reaction for the ambidextrous catalyst.
Molecular modeling studies were used to identify the ideal
position for the placement of the catalytic urea group. A model
of the catalyst was built using X-ray data from a related complex
(Fig. 2).32 Placement of the urea groups on the 6-position of the
quinoline chromophores appeared to project the urea groups
away from the copper scaffold and lower the chance of the urea
groups hindering the chirality inversion process. Additionally, a
chiral cle forms from the near perpendicular orientation of the
urea groups. We hypothesized this asymmetric pocket could be
a potential binding site for prochiral substrates.
Scheme 1 Synthetic route used to produce complex D-1.
to produce urea 5 in high yield. The methyl group of urea 5 was
then oxidized to an aldehyde using SeO2 in 97% yield. The
reductive amination of aldehyde 6 with L-methioninol was
performed using 2-picoline borane as the reducing agent to give
di-alkylated ligand 7 in 94% yield.34 Four equivalents of alde-
hyde 6 were necessary to di-alkylate the amine fully due to
unwanted reduction of the aldehyde to the corresponding
alcohol. The alcohol byproduct can be isolated and re-oxidized
with SeO2 to regenerate aldehyde 6. The ve-step sequence to
ligand 7 proceeded in 61% yield overall. The copper complex (D-
1) was then obtained aer addition of copper(II) perchlorate and
ltration.
Switching and readout
Before testing its catalytic properties, the ability to recongure
the complex was tested. As with previously reported chiral
molecular switches, UV and CD studies are crucial in deter-
mining whether or not the complex is capable of inverting
helicity upon chemical oxidation or reduction of the copper
center.35 Spectra were obtained in acetonitrile and methanol
due to the low level of absorption of these solvents in the UV
region of interest (Fig. 3).
Electronic spectra of both D-1 and the Cu(II) complex in
Fig. 116 appear to be similar qualitatively. However, complex D-1
displays additional features that likely arise from the presence
Results and discussion
Catalyst synthesis
The synthesis of ligand 7 began with the palladium-catalyzed of additional aromatic substituents that may absorb in the UV
Miyaura borylation of commercially available 6-bromo-2-meth- wavelength region. The absorption spectrum of D-1 [Cu(II)]
ylquinoline (Scheme 1). A Suzuki coupling reaction was then showed a attened peak suggesting transitions of similar
performed with 2-bromoaniline using conditions that are intensity near 240 and 250 nm. The transition near 240 nm is
1
amenable to the presence of heterocycles and amines to likely due to the Bb transition with the dipole oriented in the
produce amine 4 in 68% overall yield for two steps.33 Amine 4 longitudinal direction. This transition gives rise to an exciton
was then reacted with 3,5-bis(triuoromethyl)phenyl isocyanate couplet in the CD spectrum with a trough at 247 nm, null near
This journal is © The Royal Society of Chemistry 2015
Chem. Sci., 2015, 6, 5904–5912 | 5905

View Article Online
Chemical Science
Edge Article
Scheme 2 Preliminary catalyst results.
give opposite stereoselectivity. Initial model studies were per-
formed in toluene due to the higher affinity observed between
urea and anionic substrates in nonpolar solvents (Scheme 2).
Triethylamine was used as a co-catalyst with 10 mol%
loading with 5 mol% loading of D-1. The product nitroalkane
(10) was obtained in 55% yield as opposed to 13% yield without
D-1. The Michael adduct (S)-10 was obtained with 24% ee for the
(S)-product, affirming that the catalyst was imparting enantio-
selectivity to the reaction, albeit with low efficiency. These
results compelled us to begin screening solvents with D-1 and
L-1 to determine if reversal of enantioselectivity could be ach-
ieved under other conditions.
Solvent effects
Fig. 3 Redox triggered switching between D-1 and L-1. CD (L molꢀ1
cmꢀ1) and UV (L molꢀ1 cmꢀ1) of D/L-1 (59 mM, acetonitrile).
The ability of the catalyst to function ambidextrously in other
solvents was tested by screening D-1 and L-1 for enantiose-
lectivity in a variety of solvents. Additionally, the solvent screen
238 nm, and a peak near 230 nm. The trough at 258 nm is likely
due to a p–p* transition involving the quinoline and attached
phenyl ring. The L-1 (Cu(I)) compound shows a broad peak
near 248 nm in the absorbance spectrum, and associated peak,
null, and trough at 252, 239, and 232 nm, respectively in the CD
spectrum.36
Table 1 Effects of solvent on D/L-1a
Overall, the CD spectra of the D-1 and L-1 complexes
exhibited signicant mirror image ECCD character, consistent
with inversion of the asymmetric orientation of the chromo-
phores upon change in oxidation state. Chemical reduction of
D-1 was performed using L-ascorbic acid to attain L-1. However,
L-1 may also be obtained by addition of Cu(CH3CN)4PF6 to
ligand 7. CD/UV spectra of the copper(I) complex obtained by
direct formation using a copper(I) precursor are identical to
spectra of the copper(I) complex obtained by reduction of the
copper(II) complex. CD studies of D-1 suggest that ligand reor-
ganization takes place with inversion of helicity as
hypothesized.
D-1
L-1
Solvent
% ee of (S)-10b
% yieldc
% ee of (R)-10
% yield
Toluene
THF
24
48
72
30
46
51
0
22
3
20
55
33
55
40
44
30
81
46
87
35
51
57
70
68
74
60
7
44
0
5
33
78
40
34
43
30
73
52
78
35
MeCN
CHCl3
CH2Cl2
Hexane
EtOAc
DCE
Dioxane
MeOH
Initial catalytic performance
a
All reactions were performed using b-nitrostyrene 8 (0.34 mmol, 1
Aer UV/CD studies of D-1 conrmed operational and control- equiv.), diethylmalonate 9 (0.68 mmol, 2 equiv.), and NEt3 (0.034
mmol, 0.1 equiv.) in solvent (1 mL) with 5 mol% catalyst (D-1 or L-1)
lable helicity inversion, the catalytic activity of D-1 was tested. A
model reaction inspired by Takemoto's work was identied to
b
at room temperature for 24 h. Determined by chiral HPLC analysis.
c
Isolated yields.
test the hypothesis that the catalyst could be recongured to
5906 | Chem. Sci., 2015, 6, 5904–5912
This journal is © The Royal Society of Chemistry 2015

View Article Online
Edge Article
Chemical Science
Table 3 Base screen for basicity and steric effectsa
D-1
was performed to test the effects of different solvents on the
enantiomeric excess and yield of the reaction.
Reversal of enantioselectivity occurred in all solvents in
which the catalyst is able to achieve at least a moderate enan-
tiomeric excess. Acetonitrile provided the best combination of
yield and enantioselectivity (Table 1).
The enantioselectivity is likely dependent on both the sol-
vent's ability to stabilize binding of nitrostyrene to urea and
allow an asymmetrical conformation of the catalyst in both
oxidation states. As expected, reactions performed in nonpolar
solvents outperformed polar solvents in stereoselectivity. This
difference is presumably due to decreased substrate-urea
binding in polar solvents. Yields were generally similar in all
solvents tested, with the exception of dioxane and ethyl acetate.
These solvents likely decrease the rate of competitive anionic
% eeb
% yield of (S)-10c
pKa
d
Base
DBU
NEt3
DIEA
34
72
70
20
55
38
24.33
18.82
18.80
53
33
18.64
66
43
18.62
DABCO
66
57
44
21
18.29
17.95
polymerization of nitrostyrene catalyzed by the exogenous base. DMAP
Base effects
40
10
17
8
15.8
Subsequent screens were performed testing the role of the base
in the catalyzed reaction. Multiple bases were tested in both
oxidation states (Table 2). With all bases tested, reversal of
enantioselectivity was achieved. Stronger bases, such as DBU,
gave both low yield and low stereoselectivity. This nding is
expected due to the base's catalysis of the anionic polymeriza-
tion of nitrostyrene (8).37 Low yields were also obtained with
DMAP. Insoluble polymer side products formed almost imme-
diately aer addition of these bases.
Hindered bases with moderate conjugate acid pKa values,
such as DIPEA and NEt3, performed the best from this screen.38
These results led us to do a more comprehensive base screen
performed in only one of the catalyst states varying steric effects
and basicity (Table 3). Indeed, the best results were obtained
using bases that had high steric hindrance and a moderately
high basicity as indicated by the pKa of the conjugate acid in
acetonitrile.
Collidine
14.77
0
21
61
13–14
TBAF
20
?
a
All reactions were performed using b-nitrostyrene 8 (0.34 mmol, 1
equiv.), diethylmalonate 9 (0.68 mmol, 2 equiv.), and base (0.034
mmol, 0.1 equiv.) in MeCN (1 mL) with 5 mol% catalyst (D-1) at room
b
c
temperature for 24 h. Determined by chiral HPLC analysis. Isolated
yields. d pKa of conjugate acid in acetonitrile.38
that the low nucleophilicity would decrease the polymerization
side reaction, but low stereoselectivity was afforded.
Substrate scope
Sterically hindered bases, such as collidine and 2,6-di-tert-
butyl-4-methylpyridine, provide low yields and stereo-
selectivities whereas bases that are stronger and less sterically
hindered, like DMAP and DABCO, afford higher ee values but
still suffer from low yields. TBAF was tested as a base in hopes
We examined the catalyst against a series of different malonate
and nitrostyrene substrates (Table 4). The Michael addition
reaction with diethyl malonate proceeded with various substit-
uents bound to the nitrostyrene. The yields seemed to be
dependent on the stability of the nitrostyrene to base-catalyzed
anionic polymerization as can be seen from the 97–98% yield
obtained for the dimethoxy substrate (entry 3, Table 4). The
stereoselectivities for the substituted nitrostyrenes were lower
in some cases, but a clear pattern has yet to emerge. Different
malonates were also tested. The more sterically hindered dii-
sopropyl malonate (entry 7, Table 4) gave lower yields than
diethyl malonate. In nearly all cases, the enantioselectivity was
reversed with a change in copper oxidation state, with the
exception being acetylacetone (entry 8, Table 4). The use of acac
reduced the D-1 to L-1 yielding (R)-10h in 60–70% ee. Control
studies performed without nitrostyrene show that in the pres-
Table 2 Base screen for different complex oxidation statesa
D-1
L-1
Base
% eeb
% yield of (S)-10c
% ee
% yield of (R)-10
NEt3
72
70
66
34
57
55
38
44
40
21
70
56
30
7
40
32
48
34
32
DIPEA
DABCO
DBU
DMAP
31
a
All reactions were performed using b-nitrostyrene 8 (0.34 mmol, 1
equiv.), diethylmalonate 9 (0.68 mmol, 2 equiv.), and base (0.034 ence of NEt3 and acac in acetonitrile, D-1 is reduced to a
mmol, 0.1 equiv.) in MeCN (1 mL) with 5 mol% catalyst (D-1 or L-1)
complex with a circular dichroism spectrum identical to that of
L-1. This behaviour was seen with other ketone containing 1,3-
b
at room temperature for 24 h. Determined by chiral HPLC analysis.
c
Isolated yields.
dicarbonyl substrates as well. It is unclear exactly which side
This journal is © The Royal Society of Chemistry 2015
Chem. Sci., 2015, 6, 5904–5912 | 5907

View Article Online
Chemical Science
Edge Article
Table 4 Enantioselective reaction of D/L-1 with different substratesa
D-1
L-1
Entry
R1
R2
% eeb
% yield of (S)-adductc
% ee
% yield of (R)-adduct
1 (10a)
2 (10b)
3 (10c)
4 (10d)
5 (10e)
6 (10f)
7 (10g)
8 (10h)
4-MeC6H4
4-OMeC6H4
2,3-(OMe)2C6H3
4-BrC6H4
4-OHC6H4
Ph
OEt
OEt
OEt
OEt
OEt
OMe
i-PrO
Me
70
60
48
24
40
57
73
60 (R)
67
46
97
45
51
35
30
80 (R)
67
73
57
72
53
64
65
70
55
20
98
48
44
77
30
78
Ph
Ph
a
All reactions were performed using nitrostyrene (0.34 mmol, 1 equiv.), malonate (0.68 mmol, 2 equiv.), and NEt3 (0.034 mmol, 0.1 equiv.) in MeCN
(1 mL) with 5 mol% catalyst (D-1 or L-1) at room temperature for 3–24 h. b Determined by chiral HPLC analysis. Isolated yields.
c
reaction results in the reduction of the Cu(II), but literature enantioselective inversion (Deei) is the sum of the two ee's
suggests
a
copper-catalyzed homocoupling of dicarbonyl obtained for the two diastereomers produced using the catalyst
in its two different oxidation states. The magnitudes of Deei
compound is possible.39
In the studies shown previously, the malonate compounds were inconsistent, as the product 10j (entry 2, Table 5) was
lacked a prochiral center so that the products formed were obtained with similar enantiomeric excess in both catalyst
enantiomers. Reaction of prochiral malonates with nitrostyrene oxidation states (Deei ¼ 46 for d1, 50 for d2) while product 10k
using the ambidextrous catalyst could produce both diastereo- (entry 3, Table 5) was obtained with higher enantioselectivity
meric and enantiomeric products. Four prochiral malonates (77/72%) in the D-1 state than in the L-1 state (20/24%). In the
were tested (Table 5) with nitrostyrene 8 using D-1 and L-1. case of 10k, if the enantioselectivity were the same in the two
Moderate yields (39–54%) were obtained. Again, enantiose- catalyst oxidation states, the Deei for each pair of diastereomers
lectivity was reversed upon a change in oxidation state of the would be ꢁ140–150. The experimental Dee's (97% for d1, 96%
copper center. In these experiments, the magnitude of for d2) are well below the values expected from enantiomeric
Table 5 Stereoselective reaction of D/L-1 with prochiral malonatesa
D-1
L-1
Entry
R1
R2
% eeb d1/d2
% yield of (S)-adductc
% ee d1/d2
% yield of (R)-adduct
1 (10i)
2 (10j)
3 (10k)
4 (10l)
OMe
OBz
OEt
Ot-Bu
OEt
Ot-Bu
Ot-Bu
60/73
26/30
77/72
37/70
54 (dr ꢁ 70 : 30)d
41 (dr ꢁ 50 : 50)
46 (dr ꢁ 50 : 50)
51 (dr ꢁ 67 : 32)
64/64
20/20
20/24
50/40
40 (dr ꢁ 68 : 32)
40 (dr ꢁ 50 : 50)
39 (dr ꢁ 50 : 50)
42 (dr ꢁ 50 : 50)
OBz
a
All reactions were carried out using b-nitrostyrene (0.34 mmol, 1 equiv. malonate (0.68 mmol, 2 equiv.)), and NEt3 (0.034 mmol, 0.1 equiv.) in
b
MeCN (1 mL) with 5 mol% catalyst (D-1 or L-1) at room temperature for 3–24 h. Determined by chiral HPLC analysis (d1 ¼ diastereomer 1).
c
Isolated yield. d (dr ¼ d1 : d2).
5908 | Chem. Sci., 2015, 6, 5904–5912
This journal is © The Royal Society of Chemistry 2015

View Article Online
Edge Article
Chemical Science
transition states for unknown reasons. It should be noted that
no inversion of diastereoselectivity was obtained, yet inversion
of enantioselectivity was achieved for these compounds.
Effects of copper salt
Various copper salts were premixed in situ with ligand 7 to
determine if isolation of the complex prior to use as a catalyst
was necessary and to test the effects of different counter ions on
the Michael addition reaction (Table 6).
The effect of metal source on the Cu(II) catalyst was tested
since spatial modulation of the complex could occur if different
ions were to bind to the copper atom at its open coordination
site.36 The absence of coordinative unsaturation in the Cu(I)
catalyst discouraged us from screening the synthesis of this
catalyst with various counter ions. The isolation of every copper
complex was undesirable from a practical standpoint, so we rst
tested whether the catalyst complex could be formed in situ
before addition of the reagents. Experiments using Cu(II)
perchlorate hexahydrate (entry 1, Table 6) as the copper source
showed that the reaction occurred with slightly lower ee (67%)
and higher yield (64%) than when the complex was isolated
prior to being employed as a catalyst. The lower enantiose-
lectivity was attributed to the presence of water molecules that
originate from the hygroscopic copper perchlorate. The addi-
˚
tion of 4 A molecular sieves remedied this issue (entry 2, Table
6). Other copper salts, such as CuCl2 and Cu(NO3)2, have a
noticeable impact on the reaction, providing product (S)-10 with
lower enantioselectivity. This result can be attributed either to
urea-anion binding or copper-anion coordination. Urea-con-
taining anion sensors have proven capable of binding anions
Fig. 4 Pseudo first order rate studies performed in the presence of
excess diethyl malonate (top) and excess nitrostyrene (bottom) (mal ¼
9, ns ¼ 8c).
ꢀ
such as Clꢀ and NO3 which could limit catalyst efficiency by
malonate, the plot of ln([mal/mal0]) versus time exhibited a
linear correlation, which indicates the reaction is rst-order
blocking nitrostyrene binding. Chloride is also known to coor-
dinate with high affinity to Cu(II), which could cause a distur-
bance in catalyst conguration that lowers the ee. Anions that
with respect to nitrostyrene (Fig. 4, top). A rate of 2.2 ꢂ 10ꢀ4
M
ꢀ
bind weakly to copper and urea, such as BF4 and ClO4ꢀ, yiel-
sꢀ1 was observed. The same studies were performed in excess
nitrostyrene and indicated the reaction was also rst-order with
respect to diethyl malonate (Fig. 4, bottom). A similar rate of 1.2
ꢂ 10ꢀ4 M sꢀ1 was observed.
ded higher stereoselectivities (entries 2 and 4, Table 6).
Kinetics
We were also interested in whether one or both urea groups
were active in the transition state of the reaction. The rate of the
reaction using free ligand as catalyst was compared with the
rate using D-1 (Fig. 5). The free ligand is much less rigid and the
low energy conformations tend to place the urea groups much
farther apart than in the metal complexes. Therefore, if both
urea groups are active in the transition state, then use of the
exible free ligand may result in a slower rate of reaction.
However, the energetic cost for the urea moieties to achieve the
right orientation for reaction would not be high, so the rate may
not be greatly affected. The yields and reaction rates observed
using free ligand [kobs ¼ (1.9 ꢃ 0.1) ꢂ 10ꢀ3 Mꢀ1 sꢀ1] and D-1
[kobs ¼ (5.0 ꢃ 0.1) ꢂ 10ꢀ4 Mꢀ1 sꢀ1] were of the same order of
magnitude with both compounds. This may suggest that only
one urea group is likely active in the transition state of the
Michael addition. Indeed, literature examples of bifunctional
urea catalysts do not always show large rate enhancement for
The kinetic parameters of the reaction were examined using
pseudo-rst-order rate studies.40 Nitrostyrene 8c was used due
to its resistance to base catalyzed degradation. When the reac-
tion using D-1 was carried out in the presence of excess diethyl
Table 6 Effects of different copper salts on catalysisa
Entry
Cu2+ salt
% eeb
% yield of (S)-10c
1
Cu(ClO4)2$6H2O
Cu(ClO4)2$6H2O
CuCl2$2H2O
Cu(BF4)2
Cu(NO3)2$H2O
67
72
20
70
37
64
63
55
50
57
2d
3d
4d
5d
a
All reactions were carried out using b-nitrostyrene 8 (0.34 mmol, 1
equiv.), diethyl malonate 9 (0.68 mmol, 2 equiv.), and NEt3 (0.034
mmol, 0.1 equiv.) in MeCN (1 mL) with 5 mol% catalyst (formed in
b
situ) at room temperature for 3–24 h. Determined by chiral HPLC
analysis. c Isolated yields. d 4 A MS added.
˚
This journal is © The Royal Society of Chemistry 2015
Chem. Sci., 2015, 6, 5904–5912 | 5909

View Article Online
Chemical Science
Edge Article
Fig. 5 Reaction kinetics monitoring formation of product (S)-10 in the
presence of free ligand 7 and D-1.
covalent linkage of two reactive moieties.28 This data instigated
further examination into the mechanism of the catalyst.
Crystal structure and stereochemical rationale
Initial modeling studies were performed by building a space
lling model by modication of previously obtained X-ray data
from a related complex. The model suggested that the urea
groups position themselves nearly perpendicular to one
another, creating a chiral cle. The two oxidation states of the
catalyst produce mirror image chiral substrate binding cles
that lead to enantiomeric transition states. The previous
kinetics data studies show rst order reaction in diethyl malo-
nate, nitrostyrene, and base. This would be consistent with all
of the substrates binding concurrently in the chiral cle, or with
deprotonation of the diethyl malonate followed by attack of the
cle-bound nitrostyrene. The model suggests that nitrostyrene
is comfortable bound in the cle, with little extra room for
concomitant interactions with the diethyl malonate. To
examine the structure of the catalyst more closely, a crystal
suitable for X-ray diffraction was obtained by slow diffusion of
water into a concentrated solution of D-1 in acetonitrile.
The crystal structure of D-1 exists as a dimer with a bis(m-
alkoxo)dicopper(II) bridge connecting the monomers (Fig. 6).
Two quinoline nitrogens and oxygen from deprotonated
hydroxyl group form the trigonal plane with the central copper
ion. A second oxygen atom from the adjacent dimer and tertiary
nitrogen occupy apical positions. Previous crystallographic
structures of similar Cu(II) complexes suggest that complexes
with coordinated alcohol are monomeric while the deproto-
nated alkoxo ligands form dimers.41 It is unclear as to whether
the active catalyst is the monomer or the dimer, and this is the
subject of ongoing study. This issue does not appear to be of
great signicance regarding the catalytic mechanism, as the
space lling model (Fig. 6, top) contains a chiral cle that is
similar to that hypothesized earlier in the modeling study
(Fig. 2).
Fig. 6 The space filling (top) and stick (bottom) molecular structure of
D-1. The disordered groups, perchlorate ions, acetonitrile molecules,
and hydrogen atoms are omitted for clarity.
In the solid state, the dimers are organized into supramo-
lecular polymers via intermolecular hydrogen bonding between
urea groups. It is well known that urea catalysts usually suffer
from decreased activity due to self-quenching through dimer-
ization or oligomerization.42 However, in this catalyst solid state
structure, some urea groups remain free for maximum activity.
Two hydrogen bonds are present between urea groups of the
adjacent dimers, but two are not hydrogen bonded. Noticeably,
a chiral cle was also formed, providing a chiral environment
allowing asymmetric reaction.
Unfortunately, binding studies with L-1 in the presence of
both nitrostyrene and diethyl malonate using NMR spectros-
copy were unsuccessful. Although the urea N–H protons of the
free ligand could be accurately assigned, these N–H protons
5910 | Chem. Sci., 2015, 6, 5904–5912
This journal is © The Royal Society of Chemistry 2015

View Article Online
Edge Article
Chemical Science
upon the addition of nitrostyrene or diethylmalonate gave
inconsistent peak shis. Addition of NEt3 also caused peak
shis that led to inconclusive results. Additionally, UV and CD
studies were conducted to test for shis in the absorption peaks
upon addition of substrates, but no shis occurred in either
spectrum with the addition of either substrate.
Previously designed (thio)urea catalysts that perform the
same or similar reactions are thought to bind and activate
nitrostyrene by hydrogen bonding of the urea N–H to the two
oxygen atoms of the nitro group. This same mechanistic ratio-
nale can be applied to D-1/L-1 with the nitrostyrene binding in
the chiral cle created by the urea groups. The chirality of the
cle then becomes the major determining factor of asymmetric
induction. Fig. 7 shows the nitrostyrene bound to the catalyst in
a geometry where the approach of diethyl malonate from the Re
face of the nitrostyrene is blocked by the catalyst. This allows for
approach exclusively at the Si face, which is less sterically
encumbered by the scaffold of the catalyst.
This mechanism takes into account that only one urea is
involved in activation of the nitrostyrene substrate and the rst-
order dependence on both nitrostyrene and diethyl malonate.
Aer C–C bond formation, the resulting nitroalkane anion is
protonated by the conjugate acid of NEt3, regenerating the
catalytic NEt3.
Fig. 8 Practical handling the ambidextrous catalyst allows for both in
situ catalyst reduction and one pot methods.
energy value obtained from the enantiomeric excess may be
associated with the lower rigidity found in the urea portion of
the catalyst.
An appealing aspect of the ambidextrous catalyst is the ease
of handling. The catalyst was obtained as the copper complex by
simple ltration as the Cu2+ or Cu+ salt and used aerwards.
Additionally, the air stable Cu2+ complex (D-1) was stored and
reduced using L-ascorbic acid to isolate L-1. Model reactions
showed nearly identical yields and ee values resulted whether
the copper(I) catalyst was synthesized using a Cu+ precursor or
was formed by reduction of the copper(II) catalyst.
It has been shown that the complex itself does not have to be
isolated because the complexation was completed in situ in
acetonitrile with either Cu(ClO4)2 or Cu(CH3CN)4PF6 and the
substrates added aerwards (Fig. 8). When the desired reaction
was to be completed in a different solvent, complexation was
performed in acetonitrile, and the acetonitrile was then
removed in vacuo. The in situ formation of the copper(II)
The enantioselectivity of the reaction may be limited by the
conformational bias in the molecule that accounts for the
helical dissymmetry of the two urea-containing arms. The
difference in energies of the pseudoenantiomeric conforma-
tions of a closely related complex was estimated to be 1.5–2.5
kcal molꢀ1
. The ambidextrous catalyst should have a similar
43
difference in energies for the pseudoenantiomeric states. A DG
value of 1.1 kcal molꢀ1 was calculated from the enantiomeric
excess obtained using D-1 at room temperature. The lower free
˚
complex was performed in the presence of 4 A molecular sieves
to absorb the water present in the hygroscopic Cu2+ salts.
Conclusion
A redox recongurable copper complex was synthesized that
was capable of catalyzing the formation of an asymmetric
nitrostyrene conjugate addition. The stereoselectivity could be
modulated to afford either R or S product depending on the
oxidation state of the copper center. The catalyst was designed
by appending urea groups to a recongurable scaffold that was
capable of inverting helicity upon oxidation/reduction of the
copper center. The urea groups were the only active species
inuencing catalysis while the copper ion was used to control
the handedness of the scaffold's helical structure. The ability to
select which enantiomer of product was formed persists with
most solvents, copper salts, and bases tested. A broad scope of
symmetrical malonates, prochiral malonates, and nitrostyrene
substrates were amenable to the synthesis of both enantiomers
of product. Kinetic studies, crystal structure analysis, and
molecular modeling suggest a mechanism in which one urea
group binds and activates a nitrostyrene molecule in a chiral
Fig. 7 Model of catalyst showing hindered attack at Re face of
hydrogen bonded nitrostyrene.
This journal is © The Royal Society of Chemistry 2015
Chem. Sci., 2015, 6, 5904–5912 | 5911

View Article Online
Chemical Science
Edge Article
cle during the stereochemistry-determining step. Malonate 19 A. G. Doyle and E. N. Jacobsen, Chem. Rev., 2007, 107, 5713–
attack likely occurs on just one face of the nitrostyrene, 5743.
affording the conjugate addition product. The mirror image 20 S. Bertelsen and K. A. Jorgensen, Chem. Soc. Rev., 2009, 38,
chiral cle is formed upon catalyst reconguration leading to a 2178–2189.
nearly mirror image transition state. The handling and usability 21 P. I. Dalko, Comprehensive Enantioselective Organocatalysis:
of the catalyst was versatile and allowed several options for
isolation or in situ use.
Catalysts, Reactions, and Applications, Wiley-VCH,
Weinheim, 2013.
22 S. J. Connon, Synlett, 2009, 354–376.
23 Z. Zhang and P. R. Schreiner, Chem. Soc. Rev., 2009, 38, 1187–
1198.
Acknowledgements
We are grateful to the National Science Foundation (CHE-
1412568) for nancial support of this work.
24 Y. Takemoto, Org. Biomol. Chem., 2005, 3, 4299–4306.
25 A. Hamza, G. Schubert, T. Soos and I. Papai, J. Am. Chem.
Soc., 2006, 128, 13151–13160.
26 P. H. Cheong, C. Y. Legault, J. M. Um, N. Celebi-Olcum and
K. N. Houk, Chem. Rev., 2011, 111, 5042–5137.
27 P. S. Hynes, P. A. Stupple and D. J. Dixon, Org. Lett., 2008, 10,
1389–1391.
28 T. Okino, Y. Hoashi, T. Furukawa, X. N. Xu and Y. Takemoto,
J. Am. Chem. Soc., 2005, 127, 119–125.
29 S. H. McCooey and S. J. Connon, Angew. Chem., 2005, 44,
6367–6370.
30 Y. Takemoto, Chem. Pharm. Bull., 2010, 58, 593–601.
31 T. Okino, Y. Hoashi and Y. Takemoto, J. Am. Chem. Soc.,
2003, 125, 12672–12673.
Notes and references
1 C. S. Slone, C. A. Mirkin, G. P. A. Yap, I. A. Guzei and
A. L. Rheingold, J. Am. Chem. Soc., 1997, 119, 10743–10753.
2 E. M. Broderick, N. Guo, C. S. Vogel, C. L. Xu, J. Sutter,
J. T. Miller, K. Meyer, P. Mehrkhodavandi and
P. L. Diaconescu, J. Am. Chem. Soc., 2011, 133, 9278–9281.
3 J. B. Wang and B. L. Feringa, Science, 2011, 331, 1429–1432.
4 J. E. Beves, V. Blanco, B. A. Blight, R. Carrillo, D. M. D'Souza,
D. Howgego, D. A. Leigh, A. M. Slawin and M. D. Symes, J.
Am. Chem. Soc., 2014, 136, 2094–2100.
5 V. Blanco, D. A. Leigh, V. Marcos, J. A. Morales-Serna and
A. L. Nussbaumer, J. Am. Chem. Soc., 2014, 136, 4905–4908.
6 O. B. Berryman, A. C. Sather, A. Lledo and J. Rebek, Angew.
Chem., Int. Ed., 2011, 50, 9400–9403.
32 D. Das, Z. Dai, A. Holmes and J. W. Canary, Chirality, 2008,
20, 585–591.
33 N. Kudo, M. Perseghini and G. C. Fu, Angew. Chem., Int. Ed.,
2006, 45, 1282–1284.
7 H. J. Yoon, J. Kuwabara, J. H. Kim and C. A. Mirkin, Science,
2010, 330, 66–69.
8 C. K. A. Gregson, V. C. Gibson, N. J. Long, E. L. Marshall,
P. J. Oxford and A. J. P. White, J. Am. Chem. Soc., 2006, 128,
7410–7411.
9 M. P. Sibi, J. J. Shay, M. Liu and C. P. Jasperse, J. Am. Chem.
Soc., 1998, 120, 6615–6616.
10 G. Zanoni, F. Castronovo, M. Franzini, G. Vidari and
E. Giannini, Chem. Soc. Rev., 2003, 32, 115–129.
11 T. M. Penning and J. M. Jez, Chem. Rev., 2001, 101, 3027–
3046.
12 T. H. Tanaka and M. Hayashi, Synthesis, 2008, 21, 3361.
13 M. Raynal, P. Ballester, A. Vidal-Ferran and P. W. van
Leeuwen, Chem. Soc. Rev., 2014, 43, 1734–1787.
14 M. Vlatkovic, L. Bernardi, E. Otten and B. L. Feringa, Chem.
Commun., 2014, 50, 7773–7775.
34 Y. Y. Kawase, T. Kutsuma, T. Kataoka, T. Ueda, K. Iwakuma,
T. Nakata and T. Yokomatsu, Synthesis, 2010, 10, 1673–1677.
35 J. W. Canary, S. Mortezaei and J. Liang, Coord. Chem. Rev.,
2010, 254, 2249–2266.
36 S. Zahn and J. W. Canary, J. Am. Chem. Soc., 2002, 124, 9204–
9211.
37 M. E. Carter, J. L. Nash, J. W. Drueke, J. W. Schwietert and
G. B. Butler, J. Polym. Sci., Part A: Polym. Chem., 1978, 937.
38 I. Kaljurand, A. Kutt, L. Soovali, T. Rodima, V. Maemets,
I. Leito and I. A. Koppel, J. Org. Chem., 2005, 70, 1019–1028.
39 H. Q. Do, T. V. Hung and O. Daugulis, Organometallics, 2012,
31, 7816–7818.
40 D. A. Evans, S. Mito and D. Seidel, J. Am. Chem. Soc., 2007,
129, 11583–11592.
41 A. E. S. Holmes, S. A. Simpson and J. W. Canary, Monatsh.
Chem., 2005, 136, 461.
15 S. Mortezaei, N. R. Catarineu and J. W. Canary, J. Am. Chem.
Soc., 2012, 134, 8054–8057.
42 H. S. Rho, S. H. Oh, J. W. Lee, J. Y. Lee, J. Chin and C. E. Song,
Chem. Commun., 2008, 1208–1210.
16 S. Zahn and J. W. Canary, Science, 2000, 288, 1404–1407.
17 H. S. Barcena, B. Liu, M. V. Mirkin and J. W. Canary, Inorg.
Chem., 2005, 44, 7652–7660.
43 J. W. Canary, C. S. Allen, J. M. Castagnetto, Y. H. Chiu,
P. J. Toscano and Y. H. Wang, Inorg. Chem., 1998, 37,
6255–6262.
18 H. S. Barcena, A. E. Holmes, S. Zahn and J. W. Canary, Org.
Lett., 2003, 5, 709–711.
5912 | Chem. Sci., 2015, 6, 5904–5912
This journal is © The Royal Society of Chemistry 2015