Shell cross-linked micelle-based nanoreactors for the substrate-selective hydrolytic kinetic resolution of epoxides

Article abstract of DOI:10.1021/ja206644d

Shell cross-linked micelles (SCMs) containing Co(III)-salen cores were prepared from amphiphilic poly(2-oxazoline) triblock copolymers. The catalytic activity of these nanoreactors for the hydrolytic kinetic resolution of various terminal epoxides was investigated. The SCM catalysts showed high catalytic efficiency and, more significantly, substrate selectivity based on the hydrophobic nature of the epoxide. Moreover, because of the nanoscale particle size and the high stability, the catalyst could be recovered easily by ultrafiltration and reused with high activity for eight cycles.

Full text of DOI:10.1021/ja206644d

COMMUNICATION
pubs.acs.org/JACS
Shell Cross-Linked Micelle-Based Nanoreactors for the
Substrate-Selective Hydrolytic Kinetic Resolution of Epoxides
Yu Liu,† Yu Wang, Yufeng Wang, Jie Lu, Victor Pi~nꢀon, III, and Marcus Weck*
Molecular Design Institute and Department of Chemistry, New York University, New York, New York 10003-6688, United States
S Supporting Information
b
ABSTRACT: Shell cross-linked micelles (SCMs) contain-
ing Co(III)Àsalen cores were prepared from amphiphilic
poly(2-oxazoline) triblock copolymers. The catalytic activ-
ity of these nanoreactors for the hydrolytic kinetic resolution
of various terminal epoxides was investigated. The SCM
catalysts showed high catalytic efficiency and, more signifi-
cantly, substrate selectivity based on the hydrophobic nature
of the epoxide. Moreover, because of the nanoscale particle
size and the high stability, the catalyst could be recovered
easily by ultrafiltration and reused with high activity for eight
cycles.
ubstrate selectivity remains one of the most intriguing
research targets in catalysis. Nature has perfected substrate-
S
selective catalysis, as exhibited by many enzymatic systems that
can interact specifically with certain substrates on the basis of
differences in size/shape, charge, and/or hydrophobicity/hydro-
philicity. The development of substrate selectivity in synthetic
systems is still at an early stage.1 Several catalytic systems have
been reported to achieve size/shape selectivity, including metalÀ
organic frameworks (MOFs),2,3 zeolite-based catalysts,4,5 silica
nanoparticle-based catalysts,6 polyoxometalates (POMs),7 and
supramolecular nanocapsules.8,9 Synthetic systems can be en-
dowed with hydrophilic/hydrophobic-based substrate selec-
tivity through the use of biphasic media such as ionic liquids
and fluorous and supercritical-fluid biphasic reaction media. The
challenge is to realize substrate selectivity of nonbiological
catalytic systems in homogeneous media, such as aqueous envi-
ronments. Examples in the literature include the use of poly-
mer-based hydrogels,10a micellar structures,10b and hydrophobic
reagents.11
Figure 1. Schematic representation of the synthesis of poly(2-oxa-
zoline) SCMs with Co(III)Àsalen-functionalized cores.
bimolecular mechanism. Furthermore, the cross-linked shell
layer should provide an external shielding against metal leaching
or catalyst poisoning, which could benefit catalyst recovery and
recycling. To examine these hypotheses, we synthesized poly
(2-oxazoline)-based SCMs containing Co(III)Àsalen complexes
in the core and studied their use as catalysts for the hydrolytic
kinetic resolution (HKR) of terminal epoxides (Figure 1). The
catalysis results demonstrate that this novel support system is not
only highly active but also displays unusual substrate selectivity.
Our SCM support structure is based on the work of Weberskirch
and co-workers, who reported the synthesis and activity of
Co(III)Àsalen micellar catalysts.21 The Weberskirch system uses
conventional non-cross-linked micelles. Our design includes a
cross-linkable layer decorated with cinnamate side chains. Cin-
namates can undergo cross-linking via UV-activated [2 + 2]
cycloaddition. The cinnamate-functionalized oxazoline mono-
mer was synthesized in five steps in 69% overall yield. The
hydrophilic and hydrophobic blocks were constructed from
methyl 2-oxazoline and methyl 3-(oxazol-2-yl)pentanoate,22
respectively. Poly(2-oxazoline) triblock copolymers were synthe-
sized via microwave-irradiated cationic living polymerization23,24
using methyl triflate as the initiator. The polymerization was
Shell cross-linked micelles (SCMs) are unique and attractive
functional nanoparticles for use in aqueous environments.12 The
micellar structure of SCMs is stabilized by a covalently cross-
linked shell layer, rendering SCMs advantageous over conven-
tional micelles. SCMs show high stability toward external stimuli
such as temperature and solvents and have much lower critical
micelle concentrations.13À17 SCMs have been reported as key
materials in biomedical science, particularly in drug delivery and
bioimaging.18À20 We view SCMs as a unique support structure
for catalysis. The fixed microenvironment composed of the
hydrophobic core, the cross-linked shell, and the hydrophilic
corona has the potential to allow for programmed substrate
selectivity. Additionally, catalysts confined in the core domain of
SCMs can be forced into highly dense arrangements, leading to
enhanced reaction rates for catalytic reactions that follow a
Received: July 21, 2011
Published: August 16, 2011
r
2011 American Chemical Society
14260
dx.doi.org/10.1021/ja206644d J. Am. Chem. Soc. 2011, 133, 14260–14263
|

Journal of the American Chemical Society
COMMUNICATION
Table 1. HKR Tests of 3 with Various Terminal Epoxidesa
loading
(mol %)b
time
(h)
ee
(%)c
conv.
(%)d
entry
R
1
2
3
4
5
6
7
8
CH2Cl
0.05
0.05
0.05
0.05
0.05
0.05
0.15
0.05
0.05
24
24
8
5
3
5
4
CH2OÀallyl
n-butyl
>99
>99
>99
>99
95
52
52
50
51
53
<1
53
Figure 2. (A) AFM and (B) SEM images of 3.
c-hexyl
15
5
CH2OPh
monitored by gel-permeation chromatography (GPC). Clear
shifts of the GPC traces after each block formation indicated the
stepwise growth of the block copolymer [see the Supporting
Information (SI)]. The triblock copolymer was subjected to
deprotection followed by esterification to introduce the salen
ligand to repeating units of the hydrophobic block. The obtained
CH2OBn
8
Ph
24
24
24
CH2O(CH2CH2O)3CH3
<1
9e n-dodecyl
>99
a Reactions were carried out with 7 mmol of epoxide in 10 mL of water at
room temperature. b Catalyst loading based on cobalt. c Determined by
chiral GC or HPLC analyses. d Determined by GC or HPLC analyses
with chlorobenzene as an internal standard. e 1,2-Hexanediol was added
in a 1:1 ratio relative to the epoxide.
triblock copolymer 1 was characterized by H NMR and 13C
1
NMR spectroscopy and GPC. In CDCl3, the 1H NMR spectrum
of 1 showed degrees of polymerization of the individual blocks of
x:y:z:w = 39:3.3:2.2:3.4 (see the SI). GPC analysis revealed a
number-average molecular weight (Mn) of 6300 and a polydis-
persity index of 1.45. Micelle formation was induced by dissol-
ving 1 in methanol and was proved by 1H NMR spectroscopy in
MeOD (see the SI). Only signals of the hydrophilic and
cinnamate blocks were detected. The disappearance of the
signals of the hydrophobic salen block is attributed to the
motional restriction of the micelle core.25,26
or less with 0.01À0.04 mol % Co(III)Àsalen supported on other
polymer or oligomer systems,27À30 demonstrating the substrate
inactivity of 3 toward these epoxides. Increasing the number of
carbons in the side chain of the target epoxides resulted in a
remarkable rate acceleration. The resolution of epoxyhexane was
complete in 8 h (entry 3). For the more sterically hindered
vinylcyclohexane oxide, the reaction was finished within 15 h
(entry 4). Aromatic epoxides, such as phenyl glycidyl ether and
benzyl glycidyl ether (entries 5 and 6) were fully resolved in 5 and
8 h, respectively. When the catalyst loading was increased to
0.15%, even the conjugated compound styrene oxide reached
95% ee after 24 h (entry 7). These results indicate that the HKR
using 3 exhibits a substrate selectivity based on hydrophobicity:
the more hydrophobic the epoxide, the better is the diffusion of
the epoxide into the hydrophobic core of the SCM where the
catalytic active centers are located.
To verify the reason for the substrate selectivity, water-soluble
(2,5,8,11-tetraoxadodecyl)oxirane was synthesized and subjected
to the HKR conditions (entry 8). No resolved product was formed
after 24 h, indicating that the hydrophilic epoxide remained in the
water solution instead of interacting with the SCM catalyst. We
then explored the HKR of epoxytetradecane, a very hydrophobic
epoxide, using 3. Surprisingly, almost no resolution occurred
(ee < 2%) after 24 h. Epoxytetradecane should be able to permeate
easily into the hydrophobic SCM core. We hypothesized that the
hydrolyzed product, the resolved diol, is too hydrophobic to
diffuse through the SCM into the outside water solution,
resulting in blockage of the catalytic cycle. To test this hypoth-
esis, 1,2-hexanediol was added to the reaction mixture in a 1:1
ratio relative to epoxytetradecane to increase the water solubility
of the product. Under these conditions, epoxytetradecane could
be resolved completely and selectively in 24 h (entry 9). In all of
the HKR experiments, although large amounts of water were
applied, the product conversions were consistent with the ee
values; reactions with high ee values showed ∼50% conversions,
and those with low ee were accompanied by poor conversions,
which demonstrates the high enantioselectivity of 3.
Cross-linking of 2 using UV irradiation was conducted in
degassed methanol. The cross-linking progress was monitored by
UV spectroscopy. The disappearance of the absorption at
270 nm indicated the full cross-linking of the cinnamates (see
the SI).26 The formation of the cross-linked shell was also proved
by dynamic light scattering (DLS) analysis. In the nonselective
solvent dichloromethane, only one signal for the SCMs was
detected, corresponding to a hydrodynamic radius of 17.7 nm.
This is in agreement with the hydrodynamic radius of 17.5 nm
measured in methanol. In contrast, without cross-linking, the
dichloromethane solution of 2 showed some large-size signals
(51 and 2700 nm), suggesting the dissociation of micelles and the
formation of large aggregates (see the SI).
The SCMs were metalated with Co(II) acetate under an inert
atmosphere and then oxidized in air to produce micelle Co-
(III)Àsalen catalyst 3. The cobalt content was determined by
inductively coupled plasma mass spectrometry (ICP-MS) to be
0.24%, which corresponds to an 85% metalation yield. Catalyst 3
was characterized by DLS, atomic force microscopy (AFM), and
scanning electron microscopy (SEM). The hydrodynamic radius
determined via DLS was 18.2 nm, which is consistent with the
radii of 20 ( 3 nm obtained by AFM (Figure 2A) and 24 ( 6 nm
obtained by SEM (Figure 2B).
The catalytic activity and selectivity of 3 were examined for the
HKR of a series of epoxides in water (Table 1). During the HKR
of terminal epoxides using 3, we observed substrate selectivity.
With a catalyst loading of 0.05 mol %, 3 showed poor catalytic
activities toward less hydrophobic as well as small-sized epoxides.
After 24 h, epichlorohydrin (entry 1) was less than 5% resolved
with 5% ee. During this time frame, allyl glycidyl ether was 4%
resolved with only 3% ee (entry 2). We, and other groups, have
reported that these epoxides can be resolved completely in 12 h
14261
dx.doi.org/10.1021/ja206644d |J. Am. Chem. Soc. 2011, 133, 14260–14263

Journal of the American Chemical Society
COMMUNICATION
Scheme 1. One-Pot Competitive HKR Reactions To Test the
Substrate Selectivity of 3: HKRs Using 1:1 Ratios of (A)
Epichlorohydrin and Epoxyhexane and (B) Allyl Glycidyl
Ether and Phenyl Glycidyl Ether
Table 2. Recycling of 3 in the HKR of Epoxyhexane in Watera
cycleb
time (h)
ee (%)c
conv. (%)d
1
2
8
>99
>99
>99
>99
98
52
53
54
54
55
53
54
55
8.5
10
3
4
10.5
12
5
6e
7e
8e
11
>99
>99
99
11
12
a Conditions: 0.05 mol % catalyst loading in 10 mL of water at room
temperature. b The SCM catalyst was recovered by passing the reaction
solution through an ultrafiltration membrane with a molecular-weight
cutoff of 30 000. c Determined by chiral GC analyses. d Determined by
GC analyses with chlorobenzene as an internal standard. e During these
cycles, acetic acid (0.01 equiv relative to epoxyhexane) was added.
the catalyst stability and activity in the recycling tests showed that
the SCM catalyst surpasses the conventional micellar catalyst21
and is comparable to the most recyclable homogeneous Co(III)À
salen catalysts.31,32 However, the SCM catalyst holds the advan-
tage of facile product separation and catalyst recovery.
In conclusion, we have synthesized a new class of supported
Co(III)Àsalen complexes using shell cross-linked micelles as a
support structure. These new supported Co(III)Àsalen catalyst
complexes have high catalytic efficiencies and a very unique
substrate selectivity. The high support stability resulted in out-
standing recycling properties of the system.
Our attention was then focused on the role played by the
cross-linked shell layer in this substrate selectivity of the SCM
catalyst. Non-cross-linked micellar catalysts obtained by direct
metalation and oxidation of 2 were applied as controls. Although,
at the same catalyst loading, the non-cross-linked catalyst showed
a slightly higher catalytic activity by resolving epoxyhexane in 5 h,
it exhibited almost no substrate selectivity by producing a 90% ee
of epichlorohydrin in 48 h. In contrast, 3 afforded only a 7.7% ee
of epichlorohydrin during the same reaction time (see the SI).
The main difference between 2 and 3 is the fixed coreÀshell
structure of 3, which we suggest limits the substrate permeability.
To examine further the substrate selectivity of 3, a one-pot
competitive HKR experiment involving epoxyhexane and epi-
chlorohydrin was performed (Scheme 1A). When epoxyhexane
was fully resolved after 10 h of reaction, epichlorohydrin had
reached only 21% ee. A similar result was obtained for the
competitive reaction of phenyl glycidyl ether and allyl glycidyl
ether (Scheme 1B): the resolution of phenyl glycidyl ether was
complete in 50 h, whereas only 15% ee of allyl glycidyl ether was
resolved. To our knowledge, the SCM catalyst is the first HKR
catalytic system with substrate selectivity based on the hydro-
phobicity of epoxides.
We reasoned that SCM-supported catalysts stabilized by the
cross-linked layer might have better recyclability than their
conventional non-cross-linked analogues. We investigated the
recovery and recycling of 3 using the HKR of epoxyhexane
(Table 2). Because of its nanoscale size, the SCM catalyst could
be recovered easily by passing the reaction solution through an
ultrafiltration membrane with a molecular-weight cutoff of
30 000. For cycles 2À5, the recovered catalyst was used directly
without reactivation. In the first five cycles, while a slow decrease
in reaction rate was observed, the SCM catalysts were highly
active, resolving epoxyhexane in 8À12 h. In cycles 6À8, acetic
acid (0.01 equiv relative to epoxyhexane) was added to the
reaction mixture to regenerate the Co(III)Àsalen complexes
in situ with acetate as the counterion. Under these conditions, the
catalyst maintained the high catalytic efficiency by completing
the HKR in 11À12 h. For all of the HKR cycles, the catalyst
loadings were kept at 0.05 mol %. Cobalt leaching was investi-
gated by ICP-MS analyses of the filtrates for cycles 4 and 7. In
both samples, the Co content was below the analytical detection
limit (<1 ppm), proving the high stability of 3. Comparisons of
’ ASSOCIATED CONTENT
S
Supporting Information. Synthesesand characterizations
b
of the cinnamate-containing monomer and triblock copolymer 1,
synthetic procedures and characterization of SCM catalyst 3,
experimental procedures for HKR tests and the catalyst recycling
experiment, and crystallographic data (CIF). This material is
available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
marcus.weck@nyu.edu
Present Addresses
†Department of Chemistry, Northern Michigan University,
Marquette, Michigan, 49855, United States
’ ACKNOWLEDGMENT
We are thankful to the Department of Energy, Office of Basic
Energy Sciences, through Catalysis Contract DEFG02-03ER15459
for financial support of this work. We thank the NYU Molecular
Design Institute for purchasing the Bruker SMART APEXII
diffractometer and Dr. Chunhua Hu for his help with the
structure determination.
’ REFERENCES
(1) Das, S.; Brudvig, G. W.; Crabtree, R. H. Chem. Commun.
2008, 413–424.
(2) Ma, L. Q.; Lin, W. B. Top. Curr. Chem. 2010, 293, 175–205.
14262
dx.doi.org/10.1021/ja206644d |J. Am. Chem. Soc. 2011, 133, 14260–14263

Journal of the American Chemical Society
COMMUNICATION
(3) Czaja, A. U.; Trukhan, N.; Muller, U. Chem. Soc. Rev. 2009, 38,
1284–1293.
(4) Smit, B.; Maesen, T. L. M. Nature 2008, 451, 671–678.
(5) Corbin, D. R.; Herron, N. J. Mol. Catal. A: Chem. 1994, 86,
343–369.
(6) Anisur, R. M.; Shin, J.; Choi, H. H.; Yeo, K. M.; Kang, E. J.; Lee,
I. S. J. Mater. Chem. 2010, 20, 10615–10621.
(7) Mizuno, N.; Uchida, S.; Kamata, K.; Ishimoto, R.; Nojima, S.;
Yonehara, K.; Sumida, Y. Angew. Chem., Int. Ed. 2010, 49, 9972–9976.
(8) Liu, S. M.; Gan, H. Y.; Hermann, A. T.; Rick, S. W.; Gibb, B. C.
Nat. Chem. 2010, 2, 847–852.
(9) Wang, D. H.; Zhang, B. G.; He, C.; Wu, P. Y.; Duan, C. Y. Chem.
Commun. 2010, 46, 4728–4730.
(10) (a) Haimov, A.; Neumann, R. J. Am. Chem. Soc. 2006,
128, 15697–15700. (b) Manabe, K.; Iimura, S.; Sun, X.-M.; Kobayashi,
S. J. Am. Chem. Soc. 2002, 124, 11971–11978.
(11) Biscoe, M. R.; Breslow, R. J. Am. Chem. Soc. 2005, 127, 10812–
10813.
(12) Thurmond, K. B.; Kowalewski, T.; Wooley, K. L. J. Am. Chem.
Soc. 1996, 118, 7239–7240.
(13) O’Reilly, R. K.; Hawker, C. J.; Wooley, K. L. Chem. Soc. Rev.
2006, 35, 1068–1083.
(14) Read, E. S.; Armes, S. P. Chem. Commun. 2007, 3021–3035.
(15) Li, Y. T.; Lokitz, B. S.; McCormick, C. L. Macromolecules 2006,
39, 81–89.
(16) Ding, J. F.; Liu, G. J. Macromolecules 1998, 31, 6554–6558.
(17) Ievins, A. D.; Wang, X. F.; Moughton, A. O.; Skey, J.; O’Reilly,
R. K. Macromolecules 2008, 41, 2998–3006.
(18) Nystrom, A. M.; Wooley, K. L. Acc. Chem. Res. [Online early
access]. DOI: 10.1021/ar200097k. Published Online: June 15, 2011.
(19) Koo, A. N.; Lee, H. J.; Kim, S. E.; Chang, J. H.; Park, C.; Kim, C.;
Park, J. H.; Lee, S. C. Chem. Commun. 2008, 6570–6572.
(20) Li, L.-Y.; He, W.-D.; Li, J.; Zhang, B.-Y.; Pan, T.-T.; Sun, X.-L.;
Ding, Z.-L. Biomacromolecules 2010, 11, 1882–1890.
(21) Rossbach, B. M.; Leopold, K.; Weberskirch, R. Angew. Chem.,
Int. Ed. 2006, 45, 1309–1312.
(22) Zarka, M. T.; Nuyken, O.; Weberskirch, R. Chem.—Eur. J.
2003, 9, 3228–3234.
(23) Wiesbrock, F.; Hoogenboom, R.; Abeln, C. H.; Schubert, U. S.
Macromol. Rapid Commun. 2004, 25, 1895–1899.
(24) Wiesbrock, F.; Hoogenboom, R.; Leenen, M.; van Nispen, S.;
van der Loop, M.; Abeln, C. H.; van den Berg, A. M. J.; Schubert, U. S.
Macromolecules 2005, 38, 7957–7966.
(25) Zandomeneghi, G.; Luisi, P. L.; Mannina, L.; Segre, A. Helv.
Chim. Acta 2001, 84, 3710–3725.
(26) Sandholzer, M.; Bichler, S.; Stelzer, F.; Slugovc, C. J. Polym. Sci.,
Part A: Polym. Chem. 2008, 46, 2402–2413.
(27) Schaus, S. E.; Brandes, B. D.; Larrow, J. F.; Tokunaga, M.;
Hansen, K. B.; Gould, A. E.; Furrow, M. E.; Jacobsen, E. N. J. Am. Chem.
Soc. 2002, 124, 1307–1315.
(28) Zheng, X. L.; Jones, C. W.; Weck, M. J. Am. Chem. Soc. 2007,
129, 1105–1112.
(29) Goyal, P.; Zheng, X. L.; Weck, M. Adv. Synth. Catal. 2008,
350, 1816–1822.
(30) Liu, Y.; Rawlston, J.; Swann, A. T.; Takatani, T.; Sherrill, C. D.;
Ludovice, P. J.; Weck, M. Chem. Sci. 2011, 2, 429–438.
(31) Kureshy, R. I.; Prathap, K. J.; Agrawal, S.; Kumar, M.; Khan,
N. U. H.; Abdi, S. H. R.; Bajaj, H. C. Eur. J. Org. Chem. 2009, 2863–2871.
(32) Kim, G. J.; Lee, H.; Kim, S. J. Tetrahedron Lett. 2003,
44, 5005–5008.
14263
dx.doi.org/10.1021/ja206644d |J. Am. Chem. Soc. 2011, 133, 14260–14263