Polyoxometalate complexes for oxidative kinetic resolution of secondary alcohols: Unique effects of chiral environment, immobilization and aggregation

Article abstract of DOI:10.1039/c4dt00742e

In this paper, the chiral surfactants bearing two long alkyl chains with hydroxyl groups at their terminals were synthesized and employed to encapsulate a catalytically efficient polyoxometalate through electrostatic interaction. The obtained chiral surfactant-encapsulated polyoxometalate complexes, in which a defined chiral microenvironment surrounds the inorganic cluster, were covalently immobilized into the silica matrix via a sol-gel process. Kinetic resolution of racemic aromatic alcohols was selected as the model reaction to evaluate the chiral supramolecular hybrid catalysts. Up to 89% enantiomeric excess was obtained by varying the reaction conditions. Importantly, the change of loading values of the chiral surfactant-encapsulated polyoxometalates leads to mutative inner microstructures ranging from uniform dispersion to subsequent formation of nanocrystalline domains in the silica matrix. Such a structural evolution differentiates the density and stability of the chiral microenvironment, resulting in a regular change of enantioselectivity of the prepared asymmetric catalysts. Moreover, the fixation of the chiral microenvironment surrounding the polyoxometalates by covalent immobilization was proved to have a promoting effect on enantioselectivity. The present research uncovers the unique effect of immobilization on the kinetic resolution. The strategy helps to understand the influencing factors of enantioselectivity, and provides a convenient and efficient approach for the construction of supramolecular asymmetric catalysts based on chiral surfactant-encapsulated polyoxometalate complexes. the Partner Organisations 2014.

Full text of DOI:10.1039/c4dt00742e

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Polyoxometalate complexes for oxidative kinetic
resolution of secondary alcohols: unique effects of
chiral environment, immobilization and
aggregation†
Cite this: Dalton Trans., 2014, 43,
9177
Lei Shi, Yizhan Wang, Bao Li* and Lixin Wu*
In this paper, the chiral surfactants bearing two long alkyl chains with hydroxyl groups at their terminals
were synthesized and employed to encapsulate a catalytically efficient polyoxometalate through electro-
static interaction. The obtained chiral surfactant-encapsulated polyoxometalate complexes, in which a
defined chiral microenvironment surrounds the inorganic cluster, were covalently immobilized into the
silica matrix via a sol–gel process. Kinetic resolution of racemic aromatic alcohols was selected as the
model reaction to evaluate the chiral supramolecular hybrid catalysts. Up to 89% enantiomeric excess
was obtained by varying the reaction conditions. Importantly, the change of loading values of the chiral
surfactant-encapsulated polyoxometalates leads to mutative inner microstructures ranging from uniform
dispersion to subsequent formation of nanocrystalline domains in the silica matrix. Such a structural evol-
ution differentiates the density and stability of the chiral microenvironment, resulting in a regular change
of enantioselectivity of the prepared asymmetric catalysts. Moreover, the fixation of the chiral micro-
environment surrounding the polyoxometalates by covalent immobilization was proved to have a promot-
ing effect on enantioselectivity. The present research uncovers the unique effect of immobilization on the
kinetic resolution. The strategy helps to understand the influencing factors of enantioselectivity, and pro-
vides a convenient and efficient approach for the construction of supramolecular asymmetric catalysts
based on chiral surfactant-encapsulated polyoxometalate complexes.
Received 12th March 2014,
Accepted 11th April 2014
DOI: 10.1039/c4dt00742e
www.rsc.org/dalton
under catalytic conditions, two valuable methods are adopted.
One is to directly graft a chiral organic unit into POMs chemi-
Introduction
Polyoxometalates (POMs) are a class of nano-sized inorganic cally. But in most cases, the POMs possessing highly efficient
clusters presenting abundant potential in the fields of chem- catalytic properties are hard to graft a chiral group covalently
istry, nanotechnology, and materials science.1 Owing to their due to the structural limit, so that most of them are not suit-
intrinsic redox and acidic properties, POMs have also been able for this strategy. Even so, the known compounds could
used in catalysis, especially in organic synthesis and molecular still not provide a high enantioselectivity in asymmetric cataly-
transformation.2 On the other hand, chiral compounds are sis.7 The other method is the encapsulation of chiral cations to
becoming more important in fundamental research and indus- achiral POM clusters. For example, chiral dendritic surfactants
trial preparation such as drug production.3 Therefore, it is of were used to replace the counterions of {PO4[WO(O2)2]4}3−
practical significance to combine the high catalytic efficiency clusters for asymmetric oxidation of sulfides and epoxidation
of POMs with asymmetric catalysis.4 Much effort has been of olefins though the enantiomeric excess (ee) was not high.8
made to prepare chiral POMs and their derivatives in targeting A subsequent improvement largely increased the ee value of
the purpose,5 though it is less successful in making stable oxidized thioethers by using chiral amphiphilic cations
chiral POMs in solutions because of the rapid racemization of bearing two stereocenters to enwrap a POM cluster.9 However,
chiral architectures.6 To obtain stable chiral POM structures some new challenges are still inevitable: (1) the kinetic resolu-
tion in the catalytic process of chiral POM complexes has not
yet been dealt with; (2) the key factors influencing the enantio-
selectivity are not studied in detail; and (3) the efficient separ-
ation of chiral POM catalysts is not involved in the reactions.
State Key Laboratory of Supramolecular Structure and Materials, College of
Chemistry, Jilin University, Changchun 130012, China. E-mail: wulx@jlu.edu.cn;
Fax: +86-431-8519-3421; Tel: +86-431-8516-8481
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
Because the encapsulation of organic cations to POMs
is derived from the electrostatic interaction, all polyanion
c4dt00742e
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 9177–9188 | 9177

View Article Online
Paper
Dalton Transactions
clusters are propitious to this route in principle. The formed
surfactant-encapsulated POM (SEP) complexes exhibit core–
shell structure similar to reverse micelles, endowing them with
a good solubility in weakly polar organic solvents.10 Organic
substrates can go through the hydrophobic layer to POMs, con-
ducting the catalytic reaction. The diverse self-assembling
morphologies of SEP complexes such as lamellar, nanotube,
nanocone and so on are obtained under controllable con-
ditions,11 which could proceed or block the catalytic process.
Through covalent immobilization into the silica, the assem-
blies could be disassembled and discrete SEPs disperse in the
solid matrix due to the stronger chemical binding force in
comparison to weak intermolecular interaction, leading to
more efficient oxidation of alkenes, alcohols, and sulfides.12
Besides, these immobilized catalysts, in which the POMs are
fixed in a hydrophobic microenvironment, still have effective
capability as occurred in a homogeneous system while they
can be separated conveniently. Significantly, the microenviron-
ment around achiral POMs has been applied for the improve-
ment of asymmetric catalysis. Due to multiple catalytic sites of
the POMs, the coverage density of chiral surfactants was
proved to be an important factor for increasing the enantio-
selectivity.9 Notably, the chiral surfactant cations that
surround the surface of POM clusters could make the chiral
sites separate from the catalytic center, which largely decreases
the efficiency of the chiral microenvironment.13 To identify
this speculation, it is necessary to develop a new approach that
can stabilize a chiral microenvironment. We envision that the
flowing nature of chiral surfactants could be greatly sup-
pressed by covalent immobilization of POM complexes into
silica through reaction of end-groups of surfactants, which is
beneficial for stabilizing the chiral microenvironment and
thus promoting enantioselectivity.
Scheme 1 Schematic representation for the preparation of chiral POM
complex catalysts (R-CSHC as an example) and the loading in silica for
the kinetic resolution reaction of racemic secondary alcohols.
(98%) and N-ethyl-N-isopropylpropan-2-amine (99%) were
obtained from J&K Scientific Ltd. (R)-1-Phenylethanamine
(99%, 99% ee), (S)-1-phenylethanamine (98%, 98% ee), cumyl
hydroperoxide (80–85 wt%), tert-butyl hydroperoxide (70 wt%
in H2O), 1-penylethanol (98%), 1-phenylpropanol (97%) and
benzoin (99.5%) were purchased from Aladdin. Hydrogen
peroxide (30% aqueous solution) and TEOS were of analytical
grade and the product of Beijing Chemical Reagents Company.
TEOS was distilled before use. All the solvents used in catalytic
reactions were of analytical grade. Doubly distilled water was
used in the experiments.
Synthesis of R- and S-DOHPA·Br
We herein designed and synthesized new cationic surfac-
tants that bear chiral head and hydroxyl groups at hydrophobic The R-DOHPA·Br and S-DOHPA·Br were synthesized following
ends to encapsulate the POM with highly efficient catalytic the same procedures and confirmed by FT-IR and 1H NMR
activity. The obtained chiral SEP (CSEP) complexes are then co- spectra and ESI-MS. By taking R-DOHPA·Br as an example:
valently immobilized into the silica matrix through a sol–gel N-ethyl-N-isopropylpropan-2-amine (1.84 mL, 10.5 mmol) and
process, giving a series of chiral supramolecular hybrid cata- 11-bromoundecan-1-ol (2.65 g, 10.5 mmol) were added into
lysts (CSHCs) with different loading contents. The kinetic the acetonitrile solution (5 mL) that dissolves (R)-1-phenyletha-
resolution of CSHCs for aromatic secondary alcohols, which namine (0.5 mL, 3.9 mmol). After stirring for 96 h at 40 °C, the
is meaningful for the preparation of chiral aromatic reaction mixture was cooled to room temperature and the
alcohol derivatives, especially in pharmaceutical synthesis,14 is solvent was removed under reduced pressure. The residue was
selected as a model reaction to evaluate enantioselectivity extracted with diethyl ether, washed with water and then dried
(Scheme 1). Up to 89% ee of the remaining substrates is rea- over anhydrous MgSO4. The crude R-DOHPA·Br was dissolved
lized under the optimized reaction conditions. Considering in ethanol, and protonated with hydrochloric acid. The formed
the simplicity and controllability of this approach, these fixed light yellow oily product was then separated in a yield of
CSHCs represent a promising type of chiral POM catalysts in 39.0%. 1H NMR (500 MHz, CD3CN, 25 °C, TMS, ppm): δ =
asymmetric organic synthesis and biological mimic systems.
1.12–2.50 (m, 36H), 1.77 (d, 3H), 2.56 (t, 2H), 2.69 (s, 1H), 2.90
(s, 1H), 3.11 (broad, 2H), 3.50 (broad, 4H), 4.37 (t, 1H),
7.46–7.80 (m, 5H), and 11.52 (broad, 1H). ESI-MS (m/z): 462.3
[M+]. IR (KBr pellet): ν = 3408, 2925, 2854, 2644, 1460, 1385,
1059, 765, and 704 cm–1. Similarly, S-DOHPA·Br (0.75 g) was
prepared in a yield of 41.8%. ESI-MS (m/z): 462.1 [M+]. IR (KBr
Experimental
Materials
Na12[WZn3(H2O)2(ZnW9O34)2]·46H2O was synthesized accord- pellet): ν = 3408, 2925, 2854, 2644, 1460, 1386, 1059, 766, and
ing to the published procedure.16 11-Bromoundecan-1-ol 704 cm–1.
9178 | Dalton Trans., 2014, 43, 9177–9188
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Preparation of R- and S-CSEP
bottom flask. As an example, R-CSHC-62 (0.4 μmol) was added.
The reaction mixture was stirred under a speed of 500 r min−1
at 50 °C and the reaction was monitored by TLC and HPLC.
After the reaction was complete, the catalysts were recovered by
filtration or centrifugation. The filtrate was evaporated under
reduced pressure, and for analysis purposes, a small fraction
of the products was separated by preparative silica-gel TLC.
The ee value was determined by chiral HPLC on a chiralcel
OD-H column, eluting with hexane–isopropanol (95 : 5) at a
flow rate of 1 mL min−1 through a UV detector (254 nm) at
40 °C.
R-CSEP and S-CSEP were prepared following the same pro-
cedures and confirmed by FT-IR and H NMR spectra and EA
1
and TGA measurements. For the preparation of R-CSEP, to a
mixed solution of water–ethanol (3 : 1 in v/v) of Zn5W19 (0.12 g,
0.02 mmol) was added dropwise a mixed solution of water–
ethanol (3 : 1 in v/v) of R-DOHPA·Br (0.1 g, 0.2 mmol), where
the initial molar ratio of Zn5W19 and R-DOHPA·Br was con-
trolled at 1 : 10. After stirring for 2 h at 50 °C, the formed pre-
cipitate was collected by centrifugation and washed several
times with deionized water. The dryness of the crude product
under vacuum gives a pure yellow R-CSEP (0.13 g) in a yield of
74.4%. 1H NMR (500 MHz, CD3CN, 25 °C, TMS, ppm): δ =
1.26–1.50 (m, 39H), 2.48 (broad, 6H), 3.50 (t, 4H), 3.98 (broad,
1H), 7.27–7.42 (m, 5H). IR (KBr pellets): ν = 3408, 2926, 2854,
2661, 1630, 1460, 1385, 1312, 1211, 1057, 924, 878, 766, 704,
555, and 445 cm–1. Elemental analysis (%) calcd for R-CSEP
(C360H672N12Zn5W19O92, 10 461.12): C 41.33, H 6.47, N 1.61;
found: C 41.69, H 6.30, N 1.65, corresponding to the chemical
formula (R-CSEP)12WZn3(ZnW9O34)2. By assuming that the
organic component has completely decomposed and all the in-
organic residuals are WO3 and ZnO at 800 °C, the measured
residue of 44.6 wt% in total from TGA is in agreement with the
calculated value of 45.6 wt% according to the R-CSEP formula.
Similarly, S-CSEP (0.12 g) was obtained in a yield of 68.7%.
IR (KBr pellets): ν = 3417, 2926, 2854, 2662, 1628, 1460, 1385,
Typical catalytic procedure for kinetic resolution of benzoin
Racemic benzoin (0.2 mmol) and 30% H2O2 (2.5 equiv.) were
added into CH3CN (1 mL) in a 5 mL round bottom flask. Then
R-CSHC-86 (0.4 μmol) was added. The reaction mixture was
stirred at a speed of 500 r min−1 at 50 °C and the reaction was
examined by TLC and HPLC. After the reaction was complete,
the catalyst was recovered by filtration or centrifugation. The
filtrate was evaporated under reduced pressure, and for analy-
sis purposes, a small fraction of the products was separated by
preparative silica-gel TLC. The ee value was determined by
chiral HPLC on a chiralcel OD-H column, eluting with hexane–
isopropanol (97.5 : 2.5) at a flow rate of 0.8 mL min−1 through
a UV detector (254 nm) at 30 °C.
Measurements
1312, 1208, 1057, 924, 879, 766, 704, 557, and 459 cm−1
.
1H NMR spectra were recorded on a Bruker Ultra-Shield TM
500 MHz spectrometer using tetramethylsilane as an internal
standard (δ = 0 ppm). FT-IR spectra were collected on a Bruker
Vertex 80v spectrometer equipped with a deuterated triglycine
sulfate detector (32 scans) at a resolution of 4 cm−1. UV-Vis
and CD spectra were recorded on a Bio-Logic MOS-450 spectro-
polarimeter. Organic EA was carried out on a Vario micro cube
from Elementar. ESI-MS was carried out on a Thermo Fisher
ITQ1100 ion trap gas chromatography-mass spectrometry. TGA
was performed on a TA Instruments Q500 thermal analysis
system. DLS measurements were performed using a Zetasizer
NanoZS (Malvern Instruments). XPS spectra are acquired on
an ESCALAB-250 spectrometer with a monochromic X-ray
source (Al Kα line, 1486.6 eV). HR-TEM, SAED, and EDX were
carried out on a JEM-2100F from JEOL Ltd with an accelerating
voltage of 200 kV without staining. XRD analysis was per-
formed with a Rigaku SmartLab (3) X-ray diffractometer with
Cu-Kα radiation with a wavelength of 1.54 Å operated at 40 kV
and 30 mA in Bragg–Brentano geometry. N2 adsorption iso-
therms were measured with a Quantachrome Instruments
Autosorb 1 analyzer. The samples were degassed at 80 °C
under vacuum for more than 54 h prior to the measurements.
The specific surface areas were evaluated with the BET method
in the P/P0 range 0.05–0.35. Pore size distributions were calcu-
lated from the desorption branch of the isotherms by using
the BJH method. The ee values were determined by HPLC ana-
lysis using a SHIMADZU LC-20A equipped with a chiralcel
Elemental analysis (%) calcd for S-CSEP (C360H672N12Zn5-
W19O92, 10 461.12): C 41.33, H 6.47, N 1.61; found: C 40.91,
H 6.43, N 1.72, in good agreement with the chemical formula
of (S-CSEP)12WZn3(ZnW9O34)2.
Preparation of CSHCs
By taking R-CSHC-30 as an example, to a mixture of ethanol
(60 mL) and water (1 mL) dissolving R-CSEP (0.1 g) was added
TEOS (0.9 g). The solution was stirred vigorously, and the
resulting sol was aged for about two weeks at room tempera-
ture. After a full evaporation of solvent, a transparent solid gel
was obtained. During this process, R-CSEPs were immobilized
into the silica matrix and the product was characterized by
FT-IR, EA and TGA measurements, confirming that the
loading value of R-CSEP in the silica matrix was 30 wt%.
R-CSHC-46, R-CSHC-62, and R-CSHC-86 were prepared and
characterized according to similar procedures, except that the
molar ratios of R-CSEP and TEOS were controlled to meet the
corresponding loading values. For the S-CSHC-58, S-CSEP
instead of R-CSEP was immobilized into the silica matrix with
the loading value of 58 wt%. R-CSHC-0 and R-DOHPA/SiO2
were prepared by immobilizing nothing and pure R-DOHPA·
Brs into the silica matrix, respectively.
Typical catalytic procedure for kinetic resolution of
1-phenylethanol
Racemic 1-phenylethanol (0.2 mmol) and 30% H2O2 OD-H column (4.6 × 250 mm) obtained from Daicel Chemical
(2.5 equiv.) were added into CH3CN (1 mL) in a 5 mL round Industries Ltd.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 9177–9188 | 9179

View Article Online
Paper
Dalton Transactions
Results and discussion
Preparation and structural characterization of CSEPs
In order to construct an efficient asymmetric catalyst based on
POMs, we select
(ZnW9O34)2]·46H2O (Zn5W19), with high catalytic activity. R-Di-
(11-hydroxy-undecyl)-1-phenylethyl ammonium bromide
a known polyanion, Na12[WZn3(H2O)2-
(R-DOHPA·Br) and S-di(11-hydroxyundecyl)-1-phenylethyl-
ammonium bromide (S-DOHPA·Br) which bear a chiral head
and two long alkyl chains with a hydroxyl terminal group are
synthesized to serve as the chiral source. The reason why the
chiral center is designed close to the ionic head is that the
chiral sites could attach tightly around the POM surface in the
formed complex. On the other hand, the hydroxyl groups at
the hydrophobic ends can assist in the immobilization in the
silica matrix by reacting with Si–OH groups hydrolyzed from
tetraethyl orthosilicate (TEOS). The expected structures of
R-DOHPA·Br and S-DOHPA·Br are confirmed by 1H NMR
(Fig. 1), UV-Vis (Fig. S1†), and FT-IR (Fig. S2 and Table S1†)
spectra. The electrospray ionization mass spectrum (ESI-MS)
(Fig. S3 and S4†) gives the weights of 462.3 and 462.1 m/z for
[R-DOHPA]+ and [S-DOHPA]+, which are in good agreement
with the calculated molecular weight, 462.7 m/z, from their for-
mulae. The circular dichroism (CD) spectra of R-DOHPA·Br
and S-DOHPA·Br (Fig. 2a) are mirror symmetric, demonstrat-
ing that they are enantiopure compounds. The CSEP electro-
static complexes are prepared by simply mixing Zn5W19
polyanion with DOHPA·Brs in the mixture solvent of water and
ethanol. The corresponding R-CSEP and S-CSEP are no longer
soluble in water but readily dissolve in organic solvents such
as ethanol, acetone and acetonitrile, implying reverse micelle-
like structure of CSEPs with a hydrophilic inorganic Zn5W19
core and a hydrophobic chiral organic shell.
Fig. 2 CD spectra of (a) chiral organic cations, R-DOHPA·Br and
S-DOHPA·Br at a concentration of 3 × 10−4 mol L−1 in CH3CN, and
(b) their POM complexes, R-CSEP and S-CSEP at a concentration of
2.5 × 10−5 mol L−1 in CH3CN.
(DMSO) (Fig. S5†), the proton signal on chiral carbon (Hb) of
[R-DOHPA]+ moves from 4.37 to 3.97 ppm in CD3CN after the
encapsulation (Fig. 1 and Table S2†). Accompanied by proton
signals of the [R-DOHPA]+ head group shifting to higher field,
the peak broadening also confirms the existence of electro-
static interaction between [R-DOHPA]+ and Zn5W19 polyanion.
FT-IR (Fig. S6†) and UV-Vis spectra (Fig. S7†) of R-CSEP and
S-CSEP are almost identical. The elemental analysis (EA)
(Table S3†) and thermogravimetric analysis (TGA) (Fig. S8†)
demonstrate that all counterions of the Zn5W19 core are
replaced by [R-DOHPA]+ and [S-DOHPA]+ in complexes.
The estimated molecular formulae of both CSEPs are
in perfect agreement with the anticipated structure:
(C30H56NO2)12WZn3(ZnW9O34)2, suggesting that one polyanion
core is densely covered by twelve DOHPA surfactants on its
surface due to charge neutralization.
To understand the chirality of CSEPs, the chiroptical activi-
ties of DOHPA·Brs and CSEPs are evaluated by using UV-Vis
and CD spectra. The CD signals of R-DOHPA·Br and
S-DOHPA·Br units are comparable to the virgin ones but shift
slightly to long wavelength after their encapsulation on POMs,
suggesting the influence of the electrostatic interaction on
their chiroptical activities. A broad peak assigned to Zn5W19 in
the range between 250 and 330 nm (centered at 264 nm)
appears in UV-Vis spectra of R- and S-CSEP (Fig. S7†). Corre-
spondingly, mirror symmetric CD signals at around 300 nm
are observed in CD spectra of R- and S-CSEP complexes,
further confirming the influence of chirality of DOHPAs on
POM clusters. The signal can be definitely assigned to the
chiral induction of the organic cation to the O→W charge-
transfer band of the Zn5W19 cluster, according to our previous
results.9
Characteristic vibration bands attributable to Zn5W19 at 924
(vas WvOd), 878 (vas W–Ob–W), and 766 cm−1 (vas W–Oc–W)
are observed in the FT-IR spectrum of CSEPs (Fig. S2 and
Table S1†), which ascertain the well retained framework struc-
ture of the cluster. Similar to the case of dimethyl sulfoxide
We also study the self-assembling behavior of R-CSEP both
in solution and in the solid state by dynamic light scattering
(DLS), X-ray diffraction (XRD) and high resolution trans-
mission electron microscopy (HR-TEM). It is found that
R-CSEP complexes self-aggregate in organic solvents such as
ethanol, acetonitrile and acetone within variable tens of nano-
meters in diameter (Fig. S9–S11†). Their self-assembly in
ethanol is further confirmed by HR-TEM (Fig. S12†). The
spheroids in average diameter of ca. 110 nm, which is consist-
Fig. 1 1H NMR spectra of (a) R-DOHPA·Br and (b) R-CSEP complex in
CD3CN.
9180 | Dalton Trans., 2014, 43, 9177–9188
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Fig. 3 (a) XRD patterns of freshly prepared solid Zn5W19 and R-CSEP
complex prepared from ethanol solution; and (b) wide angle XRD
patterns of R-CSHC-0, R-CSHC-30, R-CSHC-46, R-CSHC-62 and
R-CSHC-86.
Fig. 4 HR-TEM images of (a) Zn5W19 prepared from acetonitrile solu-
tion, and (b) R-CSEP complex prepared from ethanol solution. The circle
area in (a) marks a single Zn5W19 cluster domain with distinguishable
lattice fringes. The scale bar is 2 nm.
ent with the DLS result, were found predominant. In the com-
parison of the rich diffraction peaks of Zn5W19 in the crystal-
line state, the low degree of order can be observed for CSEP
complexes in the solid state. As illustrated by the XRD
pattern,15 only a weak peak belonging to the diffraction of
CSEP complexes emerges. The calculated d-value is about
2.0 nm, much smaller than that of the ideal layer structure,
which is rationally consistent with the disordered structure of
CSEPs. The other two broad peaks at the wide angle region
could be assigned to the collective reflection of Zn5W19 at the
corresponding region (Fig. 3a). Consistently, the detailed
observation of the HR-TEM images indicates that R-CSEP com-
plexes within aggregates are in a considerably disorganized
state. Considering the competitive factors existing in DOHPAs,
it is not surprising to get the disordered aggregation structure.
Following the previous understanding, the alkyl chains in
organic units should assist ordered assembly of SEPs due to
the hydrophobic interaction. However, the terminal hydroxyl
groups used for connecting to the silica matrix in the following
sol–gel process play a destructive role in the assembly of com-
plexes, as reported in other similar systems.12 It is noteworthy
that the single POM cluster is detectable in the HR-TEM image
where the lattice fringe becomes distinguishable, which helps
us to further discern the detailed aggregation morphology.
The size of the single Zn5W19 cluster domain is found to be
in good agreement with reported data,16 and moreover,
the observed lattice distances in single Zn5W19 and CSEPs are
calculated to be identical in 0.29 nm (Fig. 4a and 4b).
prepared similarly except that the molar ratios of R-CSEP and
TEOS are controlled to meet the expected loading content. For
S-CSHC-58, S-CSEP is immobilized with the loading value of
58 wt%. As the contrast catalysts, silica R-CSHC-0 and
R-DOHPA·Br-doped silica R-DOHPA/SiO2 are prepared. The
corresponding FT-IR and X-ray photoelectron spectroscopic
(XPS) measurements further confirm the compositions and
structures of the prepared CSHCs.
Besides the characteristic vibrations at 2926 and 2854 cm−1
attributable to antisymmetric and symmetric CH2 stretching of
the organic component, the bands at 925 (vas WvOd), 872 (vas
W–Ob–W) and 775 cm−1 (vas W–Oc–W) can be well assigned to
the vibrations of inorganic clusters. In addition to the above
vibration absorptions attributed to CSEP, two additional bands
emerge in IR spectra of all CSHCs (Fig. S13†). A band centered
at 1080 cm−1 (v Si–O–Si) suggests the formation of a silica
network, while as expected, the other band at 1163 cm−1 (v Si–
O–C) confirms the covalent linkage of hydroxyl groups at the
periphery of CSEPs and the silica network. The regular change
in relative intensity of the characteristic vibration bands
belonging to POMs and the silica network suggests the
increased loadings of CSEPs. The actual loading values of
CSEPs in all CSHCs are also examined by EA and TGA
measurements (Table S4 and Fig. S14–S18†). For the represen-
tative R-CSHC-62 catalyst, the existence of elements C (1s:
284.65, 286.10, 287.83, and 288.86 eV), N (1s: 401.98 eV), and
W (5p3/2: 37.01 eV and 4p3/2: 427.00 eV) in the X-ray photo-
electron spectroscopic (XPS) measurement further confirms
Composition and structure of CSHCs
To conduct the enantioselectivity of the designed approach for the incorporation of CSEP complexes into the silica matrix
kinetic resolution, a series of CSEP-carried silica are prepared (Fig. S19†). The simulation for binding energy peaks belonging
in different doping contents. By taking R-CSHC-30 as an to the C element can be divided into four types according to
example, it means that the loading of the R-CSEP complex in their chemical environment (C–C, C–N, C–O–H and C–O–Si).
the silica matrix is 30 wt%. During the preparation of the The peak of the C–O–Si bond demonstrates successful conden-
CSHC, the calculated amount of complexes was added to the sation of the Si–OH group hydrolyzed from TEOS and terminal
TEOS solution in proportion. The actual loading value is deter- hydroxyl groups from the R-CSEP complex. Through the com-
mined by the elemental analysis. Considering that the nitro- parison of relative peak areas, we found that ca. 36.4% of OH
gen element is only derived from CSEP, we calculate the groups on CSEP are covalently linked to the silica (Fig. S20†).
loading value of the complex, which is then confirmed by the That means more than four –OH groups of one CSEP are co-
thermogravimetric analysis independently (Table S4 and valently linked to silica backbones on average. The incomplete
Fig. S14–S18†). R-CSHC-46, R-CSHC-62, and R-CSHC-86 are immobilization of CSEP complexes could facilitate the mass
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 9177–9188 | 9181

View Article Online
Paper
Dalton Transactions
Fig. 6 CD spectrum of R-CSEP complexes in acetonitrile and
R-CSHC-62 in the solid state.
Fig. 5 Nitrogen-sorption isotherms of R-CSHC-0, R-CSHC-30 and
R-CSHC-46.
could be attributed to the stabilization of the chiral micro-
environment around the POM clusters. The enhancement of
induced chirality of POMs by covalent immobilization should
contribute to the following kinetic resolution reactions, as
expected.
transfer as occurred in the homogeneous catalytic process.
In addition, the surface element contents of R-CSHC-62 and
R-CSHC-86 are detected, and the molar ratios of Si/W, Si/C and
Si/N in the latter case are all lower than those of the former
(Table S5†), suggesting that a reduced amount of Si element
but an increased content of W, C and N elements occurs. This
implies that more CSEP complexes start to locate at the surface
area of silica as their loading content increases.
Kinetic resolution of racemic secondary alcohols
We select the oxidative kinetic resolution of secondary alcohols
as the model reaction and some commonly used racemic sec-
ondary alcohols are used to evaluate the enantioselectivity of
the CSHCs. Considering that water is the only byproduct, 30%
H2O2 is employed as the oxidant. The reaction is performed
under mild conditions (50 °C and stirring at 500 rpm) and
monitored by high performance liquid chromatography
(HPLC) equipped with a chiral column. The composition of
products is examined by 1H NMR and HPLC to identify the
process by which the substrates are oxidized to the corres-
ponding ketone or other possible by-products (Fig. S25–S27†).
The catalytic activity of the prepared POM complexes in
silica for the oxidation of racemic 1-phenylethanol under the
same reaction condition is listed in Table 1. Lower conversions
and none of asymmetric oxidation are found when pure silica
Fig. 5 displays the nitrogen-sorption isotherms of CSEP-
loaded silica. The corresponding pore size distribution plots of
CSHCs are listed in the ESI (Fig. S21–S23†). The Brunauer–
Emmett–Teller (BET) surface area of R-CSHC-0 is estimated to
be 410 m2 g−1 with a total pore volume of 0.34 cm3 g−1, which
is well consistent with the reported data.17 After the CSEPs are
loaded into the silica matrix, both the BET surface area and
the total pore volume start to decrease (Table S6†), as com-
monly reported for the doping silica prepared by the sol–gel
method.18 Thus for the samples of R-CSHC-30 and R-CSHC-46,
the BET surface areas decrease greatly to 154 and 77 m2 g−1
,
respectively. The effort to get the nitrogen-sorption isotherms
of both R-CSHC-62 and R-CSHC-86 does not succeed,
suggesting even much lower surface areas, as expected.19 In
order to ensure the efficiency and consistency for all sample
entries during the catalytic process, the CSHCs are carefully
ground and their representative size among hundreds of
nanometers is maintained, as shown in the TEM image
(Fig. S24†). Besides that, based on the characteristics of sol–gel
materials, the wide pore size distribution ranging from 1 to
6 nm makes the CSHCs sufficiently allow mass transfer of
organic substrates and hence contact with the CSEP catalytic
center.12,17
In order to figure out the induced chirality of the POMs
after being immobilized into the silica, we examined the chir-
optical activity of the R-CSHC-62 as an example (Fig. 6). The
CD signal of R-DOHPA shifts to a long wavelength slightly,
suggesting the influence of the covalent immobilization on its
chiroptical activity. Importantly, the induced CD signal of
POM clusters at around 300 nm is slightly enhanced, which
Table 1 Summary of catalytic activities of different catalysts for the oxi-
dation of racemic 1-phenylethanola
Entry
Sample
t [h]
Conv.b [%]
eec
3.3 (R)
1.2 (R)
0.5 (R)
1
2
3
4
5
6
7
8
R-CSHC-0
R-DOHPA/SiO2
Zn5W19
72
72
72
72
64
64
64
64
2.9
6.3
39.3
25.6
80.2
74.5
75.1
71.0
R-CSEP
5.0 (R)
R-CSHC-30
R-CSHC-46
R-CSHC-62
R-CSHC-86
15.4 (R)
23.9 (R)
34.1 (R)
38.4 (R)
a All reactions are carried out in acetonitrile (1 mL) with racemic
1-phenylethanol (0.2 mmol), 30% H2O2 (0.3 mmol) and the above
catalysts (1 μmol) at 50 °C. b The conversions are determined by HPLC
based on the crude reaction mixture. c ee values are calculated from
HPLC analysis on
a chiralcel OD-H column, and the absolute
configuration is determined by comparison of the HPLC results with
the data in the literature.20
9182 | Dalton Trans., 2014, 43, 9177–9188
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
R-CSHC-0 and organic cation incorporated silica R-DOHPA/
SiO2 are used in the oxidation reaction. A pure Zn5W19 cluster
exhibits efficient catalytic activity, suggesting its intrinsic cata-
lytic ability to the oxidation of secondary alcohols. But it does
not show any obvious chiral catalytic properties though a very
small ee value appears in the results.
Table 2 Summary of catalytic activities of R-CSHC-62 for the oxidation
of racemic 1-phenylethanol under different reaction conditionsa
Entry Solvent
T [°C] Oxidantb t [h] Conv.c [%] eed
1
2
3
4
5
6
7
8
DMSO
50
50
50
50
1.5 (A)
1.5 (A)
1.5 (A)
1.5 (A)
1.5 (A)
1.5 (A)
1.5 (A)
1.5 (A)
2.5 (A)
2.5 (A)
5.0 (A)
2.5 (B)
2.5 (C)
64
64
64
64
64
64
64
72
70
48
24
84
96
5.1
15.2
31.8
66.5
71.7
73.8
89.1
24.9
63.9
64.3
60.1
15.7
17.7
6.7 (R)
3.4 (R)
3.1 (R)
Ethanol
Acetone
Toluene
Interestingly, the R-CSEP complex exhibits a lower conver-
sion but a higher enantioselectivity in comparison to pure
Zn5W19 (Table 1, entries 3 and 4). Apparently, this change
comes from the existence of a chiral microenvironment of
organic components surrounding the POM surface. The elec-
trostatic encapsulation confirms the synergetic catalysis of the
chiral group and the catalytic center in the R-CSEP complex.9
The main reason for the moderate conversion using Zn5W19 as
a catalyst should be the limited solubility of the naked cluster.
Meanwhile, the self-aggregation of the R-CSEP complex leads
to a restricted catalytic process, which may be responsible for
the even lower conversion. Upon chemical immobilization of
the R-CSEP complex in the silica matrix, the resulting hybrids
accelerate the reaction more than 3 times compared with the
complex alone within a similar reaction time (Table 1, entries
4 and 5). By comparing with the first four entries, the catalytic
result of the POM complex doping in silica in entry 5 implies
that: (1) CSEP aggregations in solution have broken down, pro-
viding a good dispersion of CSEP in the silica matrix; (2) the
mass transfer between the POM catalytic center and substrate
molecules is well maintained. The POM complexes incorpor-
ated into silica matrices also suggest their advantage in
improving the reaction speed besides their easy recycling
nature as a heterogeneous catalyst. More importantly, the ee
value is raised to 15%, demonstrating the stabilization of the
chiral microenvironment and the ensuing effect on the kinetic
resolution. To examine the influence of loading content on the
catalytic efficiency and enantioselectivity, we prepare a series
of immobilized catalysts, R-CSHC-30, R-CSHC-46, R-CSHC-62
and R-CSHC-86. The catalytic reaction conditions are kept the
same and the total amount of R-CSEP complexes in different
catalytic reactions is controlled to be identical. As a result,
though the pore volume decrease of silica under high loading
of POM complexes may restrain the conversion of the sub-
strate, the increased surface density of POM complexes on
silica grains can still, in large part, compensate for the
decrease of conversion. Therefore, the catalytic capability of
the hybrids decreases slowly with increasing the loading of
POM complexes. On the other hand, one can see a quick
increase of the ee value after the chemical immobilization of
chiral POM complexes in silica. With increasing the loading of
POM complexes, the ee value increases quickly in the kinetic
resolution of secondary alcohols. The influence of loading
values on the enantioselectivity will be further discussed in
the part on catalytic mechanism. Considering that the blank
silica matrix R-CSHC-0 displays non-catalytic efficiency in the
reaction, when adding the same amount of CSEP complex, the
catalytic POMs in the reaction are identical, although
the content of the silica matrix is different. That is, when
we discuss the influence of loading values on the enantio-
3.9 (R)
Cyclohexane 50
17.6 (R)
32.8 (R)
9.8 (R)
Acetonitrile
Acetonitrile
Acetonitrile
Acetonitrile
Acetonitrile
Acetonitrile
Acetonitrile
Acetonitrile
50
60
40
40
50
50
50
50
4.0 (R)
9
22.9 (R)
37.8 (R)
35.1 (R)
10.8 (R)
16.0 (R)
10
11
12
13
a All reactions are carried out with racemic 1-phenylethanol (0.2 mmol)
and R-CSHC-62 (1 μmol) in acetonitrile (1 mL). b A: 30% H2O2, B:
cumyl hydroperoxide, C: tert-butyl hydroperoxide. c The conversions
are determined by HPLC based on the crude reaction mixture; d ee
values are calculated from HPLC analysis equipped with a chiralcel
OD-H column and the absolute configuration is determined by
comparison of the HPLC results with the data in the literature.20
selectivity, it is talking about the influence of morphological
evolution resulting from loading values in CSHCs.
We use R-CSHC-62 to study the influence of experimental
conditions on kinetic resolution. The solvents used in the cata-
lytic reactions are found to play an important role on both the
conversion and enantioselectivity. The reactions in strongly
polar DMSO and ethanol lead to very low conversion and ee
value (Table 2, entries 1 and 2). As the polarity decreases, the
conversion increases to about 32% for the reaction in acetone
while the ee value is still the same (Table 2, entry 3). Further
decreasing the polarity, the conversion for the reaction in
toluene and cyclohexane raises up to over 66%, but only the
latter shows an obvious increase of the ee value (Table 2,
entries 4 and 5). When the reaction occurs in acetonitrile, both
the conversion and the ee value display an abrupt increase,
and up to 33% ee value is obtained. As shown in Fig. 1 and
Fig. S5,† we use the proton chemical shifts of R-DOHPA·Br and
R-CSEP to examine the effect of environmental change on the
POM surface. The proton signal of chiral carbon (Hb) in
R-CSEP shows an up-field chemical shift corresponding to the
isolated R-DOHPA·Br component in both DMSO-d6 and
CD3CN. It is known that strong electrostatic interaction leads
to an up-field chemical shift while weak interaction between
organic cations and inorganic anions makes the chemical shift
down-field.21 About two-fold chemical shift for R-CSEP after its
encapsulation in CD3CN to that in DMSO-d6 (Table S2†)
suggests a stronger electrostatic interaction within the R-CSEP
complex in weakly polar solvents. These closer chiral groups to
the POM should be apparently of benefit to the large increase
of enantioselectivity and can well explain the first five entries
in Table 2. For entry 6, it is an exception and hard to explain
just following the above regularity, but the present result
should be quite rational when considering the incomplete
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 9177–9188 | 9183

View Article Online
Paper
Dalton Transactions
fixation of the organic component on the silica matrix as well benzoin within 72 h, suggesting obvious oxidation selectivity
as the solubility and existing state of the substrate. in secondary alcohols. It is reported that more favorable inter-
The temperature indicates another important factor to action between the substrate and the catalyst gives higher
largely influence the conversion and enantioselectivity in the enantioselectivity.23 Similarly, we attribute this catalytic differ-
catalytic reaction. By raising the temperature from 50 °C to ence to the additional interaction between benzene rings of
60 °C, the yield of the oxidized product increases quite much chiral organic cations around the POM surface and the sub-
whereas a great decrease of the ee value occurs (Table 2, entry strate molecules containing different aromatic groups. As an
7). The result indicates that the higher temperature promotes example, the second benzene ring of benzoin, in comparison
catalytic oxidation on one side, but weakens the chiral environ- with 1-phenylpropanol, is favorable to yield additional inter-
ment attaching on the POM surface on the other side due to action with chiral cations that cover POM. This enhanced
the decreased interaction between organic and inorganic absorption and getting closer to the chiral environment
groups. When the reaction is conducted at 40 °C, the asym- around the catalytic surface can be the main driving force
metric oxidation reaction could not proceed efficiently leading to
a higher enantioselectivity. R-CSHC-62 and
(Table 2, entry 8). Because a prompt decrease of conversion is S-CSHC-58 display comparable ee values (60% and 52% ee) at
accompanied by the reaction, this almost non-enantio- the same reaction time (Table 3, entries 2 and 3), confirming
selectivity can be apparently attributed to the lower reaction that the enantioselectivity of the CSHCs definitely originate
speed at this temperature. For the concentration of the from the chiral microenvironment around POMs. For the
oxidant, under the same conditions, more hydrogen peroxide asymmetric catalysts with different doping contents of CSEP,
can raise both the conversion and the ee value rapidly R-CSHC-86 gives up to 56% and 89% ee values for the oxi-
(Table 2, entries 8 and 9). When the oxidant increases up to 5 dation of 1-phenylethanol and benzoin, respectively. Com-
fold of molar equivalence of the substrate, the reaction can be pared with other R-CSHCs employed in the kinetic resolution,
carried out in a much shorter time with similar chiral products the synthesized asymmetric catalyst that contains the higher
(Table 2, entries 10 and 11). According to the reported mech- content of R-CSEP complex shows a corresponding higher
anism in the oxidation of alcohols by POMs, the molar ratio of enantioselectivity.
the substrates to H2O2 is 1 : 1.22 The excess oxidant implies
A kinetic study is carried out to track the course of the
that more catalytic sites on POMs are activated at the same resolution process by employing the R-CSHC-86 as the catalyst
time, which leads to a higher possibility for one substrate (Fig. 7). The kinetic profile for the oxidation of the racemic
molecule to be efficiently influenced by the chiral environ- benzoin illustrates the relationship of the reaction conversion
ment. But when the activated POMs are saturated, further and the corresponding ee value. According to the proceeding
increasing the amount of oxidant becomes no longer effective. of the reaction, plot ln (Ct/C0) versus reaction time t is con-
We evaluate the peroxide, under the same reaction conditions, structed, as depicted in Fig. S28,† where C0 and Ct represent
cumyl hydroperoxide and tert-butyl hydroperoxide which only
give very low conversions and ee values even though the reac-
tion extends up to 96 h (Table 2, entries 12 and 13). As sum-
marized in Table 3, the present catalytic system demonstrates
a general route for selective oxidation of racemic 1-phenyletha-
nol and can be expanded to some other secondary alcohols. By
using R-CSHC-62 as the catalyst, we investigate optically active
remaining substrates. Only 3.4% ee is obtained for racemic
1-phenylpropanol while up to 60% ee value is acquired for
Table 3 Summary of the catalytic activities of CSHCs with different
CSEP loading contents for different aromatic alcohols under the opti-
mized reaction conditionsa
Entry Catalyst
Substrate
t [h] Conv.b [%] eec
1
2
3
4
5
R-CSHC-62 1-Phenylpropanol 72
60.1
85.6
80.6
86.4
92.1
3.4 (R)
60.0 (R)
51.6 (S)
55.6 (R)
89.4 (R)
R-CSHC-62 Benzoin
S-CSHC-58 Benzoin
R-CSHC-86 1-Phenylethanol
R-CSHC-86 Benzoin
72
72
48
72
a All reactions are carried out in acetonitrile (1 mL) with racemic
substrate (0.2 mmol), catalyst (1 μmol), 30% H2O2 (0.5 mmol) at 50 °C.
b The conversions are determined by HPLC based on the crude
Fig. 7 (a) Time profile for the oxidation of racemic benzoin (0.2 mmol)
reaction mixture. c ee values are determined by HPLC analysis on a with 30% H2O2 (0.5 mmol) using R-CSHC-86 (1 μmol) as the asymmetric
chiralcel OD-H column and the absolute configuration is determined
catalyst in acetonitrile (1 mL) at 50 °C, and (b) the kinetic resolution
process for the oxidation reaction of the racemic benzoin.
by comparison of the HPLC results with the data in the literature.24
9184 | Dalton Trans., 2014, 43, 9177–9188
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
the concentrations of benzoin at the initial and intermediate moting enantioselectivity. However, the difficulty in quantitati-
states at time t, respectively. A linear relationship indicates vely evaluating the density and stability of the chiral
that the reaction obeys pseudo-first-order kinetics, consistent environment is still a key challenge in understanding the
with the result reported for a homogeneous catalytic system.25 process. Based on the results on the kinetic resolution and
On the other hand, during the kinetic resolution process, both characterization of R-CSHC, we try to provide some evidence
R-benzoin and S-benzoin transfer into the product benzil. The for the abovementioned hypothesis.
ee of the remaining substrates is dominated by the relative
HR-TEM, XRD, selected area electron diffraction (SEAD)
reaction rate of the enantiomers, KR and KS. A 10% ee value is and energy dispersive X-ray spectroscopy (EDX) are carried out
obtained at the early stage of the reaction (10 h), but this value to detect the inner structure of R-CSHCs with different R-CSEP
increases greatly to 89% after 72 h of reaction with conversion loadings. The POM core in the CSEP complex has been
rising up to 92%. The efficient kinetic resolution suggests that reported to be in a dark dot like domain in TEM images and
KS is faster than KR, leading to the excess of R-benzoin. The the lattice fringes could be observed in HR-TEM.26 The fine
rapid growth of the ee value in the catalytic process illustrates structure with sizes of 1 to 2 nm (Fig. 8), as observed in
that the enantioselectivity of R-CSHC-86 toward benzoin is R-CSHC-30 and R-CSHC-46, is consistent with that of Zn5W19
efficient.
clusters,16 confirming the incorporation of POMs into hybrid
Because of CSEP fixing in silica, the CSHC catalysts can be silica. No obvious aggregation of POMs is observed in
recovered by a simple filtration or centrifugation and the recov- R-CSHC-30 and R-CSHC-46, suggesting that the R-CSEP com-
ered CSHCs can be easily reused. FT-IR and XRD results for plexes are isolated rather than aggregated in the silica matrix.
the catalysts after 3 cycles (Fig. S29–S30†) suggest that the The broad reflection centered at 2θ = 21–22° in the XRD
structures of recovered CSHCs are well maintained though the
enantioselectivity of the recovered CSHCs gradually decreases
in the next catalytic process (Table S7†). TGA of the recovered
CSHCs is carried out to compare its composition with that of
the freshly prepared CSHCs (Fig. S31†). About 5% of outflow is
estimated from the weight loss of CSEP in the recovered CSHC.
Because of the much higher content of CSEP in R-CSHC-62, it
is possible that some of POM complexes are not tightly bound
to the silica matrix covalently due to only 36.4% of OH groups
in the binding state. As analysed in the front part, the reduced
content in the silica matrix can weaken the chiral density of
the remaining complexes after the reaction, leading to the
decreased ee values in the later catalysis.
Mechanism of kinetic resolution
It is obvious that the formed chiral environment of organic
components attached on POMs holds the kinetic resolution
process. A synergetic catalysis within the chiral POM complex
has been detected, in which the POM cluster serves as the cata-
lyst while chiral organic cations on its surface contribute to
the certain enantioselectivity.9 Therefore, the proposed cataly-
tic route is that the substrate molecules penetrate the organic
layer and reach the POM surface, during which time they have
to adopt a preferential steric conformation in the limited
chiral space to accomplish the oxidation on the POM catalytic
center. In other words, variable enantioselectivity is the direct
result of differentiation of the chiral environment. Apparently,
two main factors involving the chiral environment are the
chiral density and the chiral solidity. On the one hand, due to
the multiple catalytic sites on the POM cluster, relatively high
chiral density of the organic layer provides more probability of
substrates to be compressed in the ambience. On the other
hand, the intrinsic fluid nature of the chiral cations rather
than fixing at the POM surface may result in an incompletely
filled chiral layer, which may decrease the opportunity of sub-
Fig. 8 HR-TEM images of (a) R-CSHC-0, (b) R-CSHC-30, (c) R-CSHC-46,
(d) R-CSHC-62, (e) R-CSHC-86, and (f) the corresponding magnified
strates to reproduce the chiral environment. Therefore, the
stabilization of the chiral environment becomes crucial in pro-
image of the area in white box of (e). Inset in (e) is the corresponding
electron diffraction pattern. The scale bars in all images are 2 nm.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 9177–9188 | 9185

View Article Online
Paper
Dalton Transactions
patterns displays the characteristic peak of the sol–gel amor-
phous carrier.27 The high content of the R-CSEP complex in
silica results in the appearance of weak diffraction at small
angle (Fig. S32†). Unexpectedly, small isolated nanocrystalline
domains with distinguishable lattice fringes emerge in the
R-CSHC-62 while the dark dots continue to be dominant. More-
over, the nanocrystalline domains in R-CSHC-86 extend into a
larger area, indicative of further crystallization of the R-CSEP
complex in the silica matrix. This could be confirmed by the
EDX result (Fig. S33†). Correspondingly, a broad diffraction
peak (centered at 2θ = 30.5°) appears and becomes more and
more evident as the loading value of the R-CSEP complex
increases (Fig. 3b). This could be assigned to the peak collec-
tion of pure Zn5W19 diffractions in the same region. The diffr-
action broadening of the corresponding peaks can be ascribed
to the formation of nanocrystalline domains.28 What is more,
the calculated d-value (0.29 nm) from XRD data is in good
agreement with both the lattice distance of pure Zn5W19
(Fig. 4a) and that of R-CSHC-86 (Fig. 8f) measured in the
HR-TEM image. The result further confirms the structural
integrity of the Zn5W19 clusters during the whole preparation
process. It is noted that the size of nanocrystalline domains in
the R-CSHC-62 and R-CSHC-86 remains nano-sized and the
orientation of the crystalline domains is disordered, as con-
firmed by SAED and HR-TEM characterization (inset of
Fig. 8e). The coincident evidence between the XRD pattern and
the HR-TEM image points to a morphological evolution from a
uniform dispersion of R-CSEP complexes to the formation of
nanocrystalline domains in the R-CSHCs.
The structural characterization of the POM complex in
silica provides a slot to understand the contribution of the
chiral environment on the enantioselectivity and the key
factors in the kinetic resolution process. As observed in the
HR-TEM images, the formation of consecutive Zn5W19 nano-
crystalline domains in the matrix indicates the existence of
POM aggregations and the tight cluster’s stacking. The aggre-
gation implies that the partial surface area of single POM
cluster, at which the chiral cations locate originally, starts to
become occupied by adjacent POM clusters, leading to the
rearrangement of those organic components. With this under-
standing, we could figure out a correlation between surface
chiral density around POM and the enantioselectivity
(Scheme 2). To simplify the model, we assume that the POM
cluster can be compressed into a cubic shape with 6 identical
faces. Thus, the total chiral organic cations equally disperse
on other faces when one face of a POM is covered by a face of
another POM upon two clusters forming an aggregated par-
ticle. Considering the huge possibilities when different
numbers of complexes aggregate together, we only analyze the
surface cation density at the densest stacking state. The
surface density (Dcoc) of chiral cations (the average number of
chiral organic cations on one face of a POM in an aggregated
particle) for nanocrystalline particles is calculated and listed
in Table S8.† The original surface density is 2 chiral organic
cations on each face for the isolated CSEP complex that owns
12 organic cations. For the nanocrystalline domain containing
Scheme 2 The calculated surface density (Dcoc
) of chiral organic
cations as a function of the number (Ncc) of R-CSEP chiral complexes in
the formed nanocrystalline domains. The insets are the corresponding
schematic evolution of the complexes in the formation of nanocrystal-
line domains, in which we just depict some faces of the POM cluster
aggregations.
3 CSEP complexes, the surface chiral density increases to 2.6
chiral cations per POM face. In the case of a nanocrystalline
domain containing 8 CSEP complexes, the surface density of
the chiral cation reaches 4.0 chiral cations per POM face, twice
of that in the isolated CSEP complex. From this analysis, the
aggregation is obviously favourable for the increase of chiral
product. Furthermore, according to total surface area based on
the reported crystal data, one face area of an isolated Zn5W19
is calculated to be ca. 2.03 nm2.16 Considering the structural
similarity to the present chiral cation, we use dioctadecyldi-
methyl ammonium with double chains and a known cross-
sectional area (0.57 nm2) to evaluate the surface occupation of
the chiral cations.29 As a result, the surface area (Scoc) of chiral
cations on one face is 0.57 × Dcoc = 1.14 nm2, much less than
the maximum of a single face of Zn5W19. This lower density
chiral environment well explains the lower efficiency of asym-
metric catalysis under lower CSEP complex loading in silica.
With increasing the aggregation number, the density of the
chiral cation on Zn5W19 increases, well explaining why the
CSEP complex aggregation in the silica matrix can lead to an
enriched chiral microenvironment and thus high enantio-
selectivity in the kinetic resolution. However, when the
number of complexes in one nanocrystalline particle reaches
8, the chiral surface area Scoc will increase up to 2.28 nm2, a
bit larger than a single face area of an isolated Zn5W19 cluster.
This implies that the POM surface could not accommodate so
many organic chiral cations. Because of the non-densest stack-
ing in the practical aggregation systems, the compressibility of
organic cations and the electrostatic repulsion between POMs,
the number of complexes in nanocrystalline particles could be
possibly more than the predicted amount. But in any case, as
observed in the TEM image, the number of complexes in
aggregation should be in a limited range.
Due to the steric restriction of the enriched chiral environ-
ment resulting from the decreasing surface area of the CSEP
complex in aggregations, the fluidity of the rearranged chiral
organic cations around the POM clusters is forced to weaken
synchronously, leaving a stiffer chiral layer. Notably, the above
9186 | Dalton Trans., 2014, 43, 9177–9188
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
discussion focuses only on CSEP complex-immobilized hybrid
catalysts rather than on their bulk state in the solution. Actu-
ally, a more beneficial stabilization of the chiral microenviron-
ment in solution even provides a much higher chiral density
without a stabilization strategy. But it is hard to obtain high
enantioselectivity in the kinetic resolution.9 Though we could
not quantitatively evaluate the influence of the stabilization of
the chiral environment, the obvious promoting effect implies
that the fixation of the chiral layer around POMs should be
another critical factor for the chiral resolution of secondary
alcohols.
15697; N. Dupré, P. Rémy, K. Micoine, C. Boglio,
S. Thorimbert, E. Lacôte, B. Hasenknopf and M. Malacria,
Chem. – Eur. J., 2010, 16, 7256; F. M. Toma, A. Sartorel,
M. lurlo, M. Carraro, P. Parisse, C. Maccato, S. Rapino,
B. R. Gonzalez, H. Amenitsch, T. D. Ros, L. Casalis,
A. Goldoni, M. Marcaccio, G. Scorrano, G. Scoles,
F. Paolucci, M. Prato and M. Bonchio, Nat. Chem., 2010, 2,
826; O. A. Kholdeeva, B. G. Donoeva, T. A. Trubitsina, G. Al-
Kadamany and U. Kortz, Eur. J. Inorg. Chem., 2009, 5134;
Y. V. Geletii, B. Botar, P. Kögerler, D. A. Hillesheim,
D. G. Musaev and C. L. Hill, Angew. Chem., Int. Ed., 2008,
47, 3896.
3 R. Noyori, Angew. Chem., Int. Ed., 2002, 41, 2008; Y. Huang,
T. Liu, J. Lin, J. Lü, Z. Lin and R. Cao, Inorg. Chem., 2011,
50, 2191.
4 B. Hasenknopf, K. Micoine, E. Lacôte, S. Thorimbert,
M. Malacria and R. Thouvenot, Eur. J. Inorg. Chem., 2008,
5001; D. Du, L. Yan, Z. Su, S. Li, Y. Lan and E. Wang,
Coord. Chem. Rev., 2013, 257, 702.
5 J. F. Garvey and M. T. Pope, Inorg. Chem., 1978, 17, 1115;
H. An, E. Wang, D. Xiao, Y. Li, Z. Su and L. Xu, Angew.
Chem., Int. Ed., 2006, 45, 904; H. Tan, Y. Li, Z. Zhang,
C. Qin, X. Wang, E. Wang and Z. Su, J. Am. Chem. Soc.,
2007, 129, 10066; Y. Hou, X. Fang and C. L. Hill, Chem. –
Eur. J., 2007, 13, 9442; U. Kortz, M. G. Savelieff, F. Y. Abou
Ghali, L. M. Khalil, S. A. Maalouf and D. I. Sinno, Angew.
Chem., Int. Ed., 2002, 41, 4070; K. Micoine, B. Hasenknopf,
S. Thorimbert, E. Lacôte and M. Malacria, Angew. Chem.,
Int. Ed., 2009, 48, 3466; J. Zhang, J. Hao, Y. Wei, F. Xiao,
P. Yin and L. Wang, J. Am. Chem. Soc., 2010, 132, 14;
F. Xiao, J. Hao, J. Zhang, C. Lv, P. Yin, L. Wang and Y. Wei,
J. Am. Chem. Soc., 2010, 132, 5956.
6 X. Fang, T. M. Anderson and C. L. Hill, Angew. Chem., Int.
Ed., 2005, 44, 3540.
7 M. Carraro, G. Modugno, A. Sartorel, G. Scorrano and
M. Bonchio, Eur. J. Inorg. Chem., 2009, 5164.
8 C. Jahier, M. Cantuel, N. D. McClenaghan, T. Buffeteau,
D. Cavagnat, F. Agbossou, M. Carraro, M. Bonchio and
S. Nlate, Chem. – Eur. J., 2009, 15, 8703; C. Jahier and
S. Nlate, Eur. J. Inorg. Chem., 2012, 833.
9 Y. Wang, H. Li, W. Qi, Y. Yang, Y. Yan, B. Li and L. Wu,
J. Mater. Chem., 2012, 22, 9181.
Conclusions
In conclusion, we have systematically studied the kinetic
resolution of aromatic secondary alcohols by preparing newly
designed chiral POM complex-incorporated silica hybrids. This
kind of CSHC hybrids that possess defined chiral microenvir-
onment surrounding POM clusters exhibit catalytic properties
like those of heterogeneous asymmetric catalysts. The kinetic
resolution process is found to be closely related to the chiral
density and the stabilization of the chiral layer of CSEP com-
plexes that has covalently immobilized into the silica matrix.
Besides fixation in silica, the existing state of the POM
complex in CSHCs is also considered to be responsible for the
corresponding promotion of the enantioselectivity in the
kinetic resolution. A possible mechanism for understanding
the influence of the immobilized catalysts on the enantio-
selectivity is proposed based on the characterization results of
HR-TEM, XRD and SAED. The covalent immobilization and
existing state change of the POM complex in CSHCs are proved
to co-contribute to a more stable and enriched chiral micro-
environment. The present research provides a convenient and
effective approach to prepare asymmetric catalysts and
uncovers valuable principles for developing efficient supra-
molecular asymmetric catalysts based on achiral POMs
bearing chiral organic counterions.
Acknowledgements
This work was financially supported by the National Basic
Research Program (2013CB834503), NSFC (91227110,
21221063), Ministry of Education of China (20120061110047),
111-Project (B06009) for the collaboration with WVU, and
Open Project of SKL of Polymer Physics & Chemistry, CAS.
10 C. Venturello and R. D’Aloisio, J. Org. Chem., 1988, 53,
1553.
11 Y. Yang, Y. Wang, H. Li, W. Li and L. Wu, Chem. – Eur. J.,
2010, 16, 8062; Y. Yan, H. Wang, B. Li, G. Hou, Z. Yin,
L. Wu and V. W. W. Yam, Angew. Chem., Int. Ed., 2010, 49,
9233; A. Nisar, J. Zhuang and X. Wang, Adv. Mater., 2011,
23, 1130; A. Nisar, Y. Lu, J. Zhuang and X. Wang, Angew.
Chem., Int. Ed., 2011, 123, 3245.
Notes and references
1 M. T. Pope and A. Müller, Angew. Chem., Int. Ed. Engl., 12 W. Qi, H. Li and L. Wu, Adv. Mater., 2007, 19, 1983; W. Qi,
1991, 30, 34; special issue on “polyoxometalates”, Chem.
Rev., 1998, 98, 1.
Y. Wang, W. Li and L. Wu, Chem. – Eur. J., 2010, 16,
1068.
2 K. Kamata, K. Yonehara, Y. Sumida, K. Yamaguchi, 13 H. Li, H. Sun, W. Qi, M. Xu and L. Wu, Angew. Chem., Int.
S. Hikichi and N. Mizuno, Science, 2003, 300, 964;
A. Haimov and R. Neumann, J. Am. Chem. Soc., 2006, 128,
Ed., 2007, 46, 1300; Y. Yan and L. Wu, Isr. J. Chem., 2011,
51, 181.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 9177–9188 | 9187

View Article Online
Paper
Dalton Transactions
14 S. D. Rychnovsky, T. L. McLernon and H. Rajapakse, J. Org.
Chem., 1996, 61, 1194; D. C. Ebner, R. M. Trend, C. Genet,
L. Shi and L. Wu, Angew. Chem., Int. Ed., 2013, 125,
4675.
M. J. McGrath, P. O’Brien and B. M. Stoltz, Angew. Chem., 22 A. Haimov and R. Neumann, J. Am. Chem. Soc., 2006, 128,
Int. Ed., 2008, 47, 6367.
15697.
15 S. B. Yoon, J. Y. Kim and J. Yu, Chem. Commun., 2001, 559.
16 C. M. Tourné, G. F. Tourné and F. Zonnevijlle, J. Chem.
Soc., Dalton Trans., 1991, 143.
17 L. L. Hench and J. K. West, Chem. Rev., 1990, 90, 33;
B. Yang, Y. Zhang, E. Drabarek, P. R. F. Barnes and V. Luca,
Chem. Mater., 2007, 19, 5664.
23 F. D. Furia, G. Licini, G. Modena and R. Motterle, J. Org.
Chem., 1996, 61, 5175.
24 C. Bolm, M. Kesselgruber, A. Grenz, N. Hermanns and
J. P. Hildebrand, New J. Chem., 2001, 25, 13;
M. C. Fragnelli, P. Hoyos, D. Romano, R. Gandolfi,
A. R. Alcántara and F. Molinari, Tetrahedron, 2012, 68, 523.
18 S. Cheng, Y. Wei, Q. Feng, K. Qiu, J. Pang, S. A. Jansen, 25 L. Ni, J. Patscheider, K. K. Baldridge and G. R. Patzke,
R. Yin and K. Ong, Chem. Mater., 2003, 15, 1560; F. Shi, Chem. – Eur. J., 2012, 18, 13293.
Q. Zhang, D. Li and Y. Deng, Chem. – Eur. J., 2005, 11, 26 H. Nair, J. T. Miller, E. A. Stach and C. D. Baertsch,
5279. J. Catal., 2010, 270, 40.
19 A. González-Campo, B. Boury, F. Teixidor and R. Núñez, 27 S. C. Nunes, V. de Zea Bermudez, M. M. Silva, M. J. Smith,
Chem. Mater., 2006, 18, 4344; D. K. Yi, S. T. Selvan,
S. S. Lee, G. C. Papaefthymiou, D. Kundaliya and J. Y. Ying,
J. Am. Chem. Soc., 2005, 127, 4990.
D. Ostrovskii, R. A. Sá Ferreira, L. D. Carlos, J. Rocha,
A. Gonçalves and E. Fortunato, J. Mater. Chem., 2007, 17,
4239.
20 S. Zeror, J. Collin, J.-C. Fiaud and L. Aribi-Zouioueche, 28 W. Ma, C. Yang, X. Gong, K. Lee and A. J. Heeger, Adv.
J. Mol. Catal. A: Chem., 2006, 256, 85. Funct. Mater., 2005, 15, 1617.
21 W. Qi, H. Li and L. Wu, J. Phys. Chem. B, 2008, 112, 8257; 29 K. Okuyama, Y. Soboi, N. Iijima, K. Hirabayashi,
H. Li, Y. Yang, Y. Wang, W. Li, L. Bi and L. Wu, Chem.
Commun., 2010, 46, 3750; Y. Wang, H. Li, C. Wu, Y. Yang,
T. Kunitake and T. Kajiyama, Bull. Chem. Soc. Jpn., 1988,
61, 1485.
9188 | Dalton Trans., 2014, 43, 9177–9188
This journal is © The Royal Society of Chemistry 2014