Enantioselective epoxidation of β-methylstyrene catalyzed by immobilized Mn(salen) catalysts in different mesoporous silica supports

Article abstract of DOI:10.1016/j.jcat.2008.03.013

Mesoporous silica-supported chiral Mn(salen) catalysts were prepared and evaluated in the heterogeneous asymmetric epoxidation of β-methylstyrene with NaClO as an oxidant. Homogeneous and immobilized Mn(salen) catalysts exhibit similar cis/trans ratios an

Full text of DOI:10.1016/j.jcat.2008.03.013

Journal of Catalysis 256 (2008) 226–236
Contents lists available at ScienceDirect
Journal of Catalysis
www.elsevier.com/locate/jcat
Enantioselective epoxidation of β-methylstyrene catalyzed by immobilized
Mn(salen) catalysts in different mesoporous silica supports
Haidong Zhang a, Yi Meng Wang b, Lei Zhang a, Gijsbert Gerritsen b, Hendrikus C.L. Abbenhuis b,
Rutger A. van Santen b,∗, Can Li a,∗
a
State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China
Schuit Institute of Catalysis, Eindhoven University of Technology, 5600 MB Eindhoven, The Netherlands
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Mesoporous silica-supported chiral Mn(salen) catalysts were prepared and evaluated in the heteroge-
neous asymmetric epoxidation of β-methylstyrene with NaClO as an oxidant. Homogeneous and immo-
bilized Mn(salen) catalysts exhibit similar cis/trans ratios and ee values when trans-substrate is used but
different cis/trans ratios and ee values when cis-substrate is used. The production of trans-epoxides is
favored in the heterogeneous asymmetric epoxidation of cis-β-methylstyrene, whereas the production of
cis-epoxides is favored in homogeneous reaction. In addition, the cis/trans ratio and ee value of trans-
epoxides produced from cis-β-methylstyrene change sequentially with changes in support materials, but
the ee value of cis-epoxides does not. The textural properties of immobilized Mn(salen) catalysts can be
sequentially adjusted by changing pore dimension and channel length, after which the motion restriction
and confinement effect in nanochannels of immobilized Mn(salen) catalysts can be adjusted sequentially.
Our results reflect that the collapse step of trans-intermediates is considerably affected by the confine-
ment effect, whereas the rotation step of cis-intermediates is greatly influenced by motion restriction.
The motion restriction in nanochannels increases the likelihood of rotation for the C–C single bond of
cis-intermediates and favors the production of trans-epoxides.
Received 21 January 2008
Revised 16 March 2008
Accepted 17 March 2008
Available online 29 April 2008
Keywords:
Chiral catalysis
Mn(salen)
Asymmetric epoxidation
Motion restriction
Confinement effect
Mesoporous
Immobilization
© 2008 Published by Elsevier Inc.
1. Introduction
immobilization strategy [8,23–25]. Various materials (e.g., alumina,
mesoporous silica materials, polymers, carbon materials, zeolites)
can be used as the supports of immobilized Mn(salen) catalysts
Chiral Mn(salen) complex-catalyzed enantioselective epoxida-
tion reactions of unfunctionalized olefins are of great importance
in asymmetric catalysis [1–3]. In homogeneous systems, recycling
of catalysts and separation of products and catalysts are usu-
ally costly multistep processes. Because of the inherent advan-
tages of heterogeneity, including easy separation and recycling,
heterogeneous asymmetric epoxidation catalyzed by immobilized
Mn(salen) catalysts is of significant interest [4–9]. However, the
nanochannels of the supports may exert an unexpected effect on
asymmetric catalytic performance; such an effect due to hetero-
geneity remains an open issue.
Heterogeneous Mn(salen) catalysts can be immobilized with
linkage groups connected to the Mn atoms [10–13] or salen lig-
ands [14–22]. Support is a crucial factor influencing the catalytic
behavior of immobilized Mn(salen) catalysts at every level; thus,
choosing the proper support is critical to formulating an effective
[6–9,23–25]. Among these supports, the mesoporous silica materi-
als, particularly SBA-15 and MCM-41 materials, have received much
attention because of their large surface areas, uniform hexagonally
ordered two-dimensional mesoporous channels, high hydrothermal
stability, and tunable pore dimension and channel structure, allow-
ing ready diffusion of reactants to the active sites located in the
nanopores. More importantly, the nanopores of these mesoporous
supports can influence the mechanism of the heterogeneous epoxi-
dation reaction [25,26], resulting in a different product distribution
than seen in homogeneous reactions. Piaggio et al. found that in
the epoxidation of (Z)-stilbene catalyzed by MCM-41-immobilized
Mn(salen) catalyst with PhIO as an oxidant, the cis/trans ratio was
much higher for heterogeneous reactions than for homogeneous
reactions [27], due to retardation of the rotation of the radical in-
termediates by the MCM-41 nanopores in heterogeneous reactions.
The lifetime of the radical intermediates, which is greatly affected
by their diffusion behaviors in the nanopores of supports, is an-
other possible factor influencing the rotation step [28].
Corresponding authors. Faxes: +31 40 2455054, +86 411 84694447.
*
The mechanism of nanopore-induced changes in product dis-
tribution remains unclear, however. In the present work, we im-
mobilized Mn(salen) catalysts in the nanochannels of different
E-mail addresses: r.a.v.santen@tue.nl (R.A. van Santen), canli@dicp.ac.cn (C. Li).
URLs: http://www.catalysis.nl (R.A. van Santen), http://www.canli.dicp.ac.cn
(C. Li).
0021-9517/$ – see front matter © 2008 Published by Elsevier Inc.
doi:10.1016/j.jcat.2008.03.013

H. Zhang et al. / Journal of Catalysis 256 (2008) 226–236
227
mesoporous silica materials with varying pore dimensions and
channel structures. The heterogeneous Mn(salen) catalysts were
immobilized in two different strategies, in which linkage groups
were connected to either Mn atoms or salen ligands. All immo-
bilized Mn(salen) catalysts were evaluated in the epoxidation of
β-methylstyrene with NaClO as the oxidant to study the effect
of different nanochannels on the reaction. We found that a nar-
row pore and a long channel increase the selectivity to trans-
epoxides in the epoxidation of cis-β-methylstyrene, consistent with
a decreased mobility of reaction intermediates. The motion re-
striction in narrow pores and long channels can increase the
likelihood of C–C single-bond rotation of cis-intermediate and fa-
vors the production of trans-epoxides. In addition, the ee value
for trans-epoxides varies with changing support, but that for cis-
epoxides does not, indicating a considerable confinement effect of
nanopores on the collapse process for trans-intermediates but a
negligible effect for cis-intermediates.
Scheme 1. Preparation of 3-aminopropyl-siloxy group-modified supports.
◦
was added. This mixture was refluxed at 35 C for 20 h and then
◦
hydrothermally treated in an autoclave at 100 C for 48 h. The mix-
2. Experimental
ture was then filtered to collect the white solid product, which was
washed with copious amounts of water and dried at room temper-
Tetraethyl orthosilicate (TEOS, ABCR, low metal impurities,
tetraethoxysilane 99+%), EO20–PO70–EO20 (Pluronic P123, Aldrich),
ammonia (25 wt%, Acros), NH4F (Acros), trimethyl benzene (TMB,
◦
ature. Finally, the white powder was calcined at 540 C for 4 h to
remove the template agent.
Merck), n-decane (Merck), 3-aminopropyl-triethoxysilane (APTES)
2.1.2. Synthesis of long-axis SBA-15 support
ꢀ
(ABCR),
(R,R)-(-)N,N -bis(3,5-di-tert-butylsalicylidene)-1,2-cyclo-
The synthesis procedure of long-axis SBA-15 (LS) supports was
similar to that described by Choi et al. [30]. First, 22.79 g of P123
was added to 700 ml of aqueous HCl solution (0.43 M). The re-
sulting mixture was stirred at room temperature to obtain a clear
solution. Then the solution was balanced at 35 C, followed by the
addition of 52.5 ml of TEOS. This mixture was refluxed at 35 C
hexanediaminomangenese(III) chloride (Jacobsen catalyst, Acros,
98%), dichloromethane (CH2Cl2, Acros), 3-tert-butyl-2-hydroxybenz-
aldehyde (Aldrich, 96%), 3,5-di-tert-butyl-2-hydroxybenzaldehyde
(Aldrich, 99%), (1R,2R)-(-)-1,2-diaminocyclohexane (Aldrich, 98%),
paraformaldehyde (Aldrich), diethyl ether (Et2O, Acros), NaHCO3
(Merck), Na2SO4 (Merck), KOH (Merck), MgSO4 (Acros) concen-
trated hydrochloric acid (HCl, Acros, 37%), Mn(OAc)2·4H2O (Acros),
LiCl (Acros), (1S,2S)-trans-β-methylstyrene oxide (Aldrich, 98%),
(1R,2R)-trans-β-methylstyrene oxide (Aldrich, 97%), trans-β-me-
thylstyrene (Aldrich, 99%), cis-β-methylstyrene (TCI, stabilized with
TBC, > 98% GC), 4-phenylpyridine N-oxide (PPNO, Acros), and m-
chloroperoxybenzoic acid (m-CPBA, Acros) were obtained from
commercial sources and used as received. Toluene (Acros) used
in the synthesis of 3-aminopropyl-siloxy group modified supports
was dried before use.
◦
◦
◦
for 24 h, and then treated in an autoclave at 100 C for 24 h. The
resulting white solid product was then separated by filtration and
washed with a mixture of H2O, ethanol, and HCl, and then dried
at room temperature. The dry white powder was then calcined at
◦
550 C for 4 h to remove the template agent.
2.1.3. Synthesis of short-axis SBA-15 support
To synthesize the short-axis SBA-15 (SS) support, first, 5.04 g
of P123 was added to 175 ml of aqueous HCl solution (1.07 M)
under stirring at room temperature. When a clear solution was
obtained, n-decane was added to obtain a mixture with an n-
decane/P123 molar ratio of 237. The resulting mixture was then
stirred overnight. After stirring, the clear separation of a water-rich
layer and a decane-rich layer was achieved after at least 8 h. The
upper decane-rich layer was carefully removed, and 0.06 g of NH4F
was added to the water-rich layer. The water-rich layer was then
The content of Mn in the immobilized catalysts was deter-
mined by inductively coupled plasma-atomic emission spectrome-
try (ICP-AES) on an Varian Vista spectrometer. The 1H NMR spectra
were recorded on a Bruker Mercury 400 instrument (400 MHz,
◦
CDCl3, 25 C). N2 adsorption–desorption analysis was done at 77 K
on a Micromeritics TriStar 3000 instrument. The small-angle X-ray
diffraction (SXRD) patterns were collected on a Rigaku D/max-
2500/PC X-ray diffractometer with CuKα radiation. FTIR spectra
were recorded on a Perkin–Elmer Spectrum One FTIR spectrom-
eter with a FR-DTGS detector, using the KBr pellet method at
a resolution of 4 cm−1. UV–vis DRS spectra were obtained on
a Shimadzu 2010 instrument. High-resolution scanning electron
microscopy (HRSEM) was done with a FEI QUANTA 200F instru-
ment or a Philips XL30 FEG instrument. Transmission electron
microscopy (TEM) was performed using an FEI Tecnai 20 Twin in-
strument.
◦
stirred at 40 C for 15 min, followed by the addition of TEOS to
achieve a TEOS/P123 molar ratio of 41.4. This mixture was refluxed
◦
at 40 C for 20 h and then hydrothermally treated in an autoclave
◦
at 100 C for 51 h. The mixture was then filtered to collect the
white solid product, which was washed with copious amounts of
◦
water and dried at 60 C in vacuum. Finally, the white powder was
◦
calcined at 550 C for 4 h to remove the template agent.
2.2. Synthesis of 3-aminopropyl-siloxy group-modified supports
The preparation of 3-aminopropyl-siloxy group-modified sup-
port is demonstrated in Scheme 1. First, 2 g of support was treated
by 200 ml of HCl (2 M) at 100 C for 4 h. The support was
2.1. Synthesis of support materials
◦
2.1.1. Synthesis of mesostructured cellular foams
then filtered and washed with copious amounts of water until the
Synthesis of the mesostructured cellular foam (MCF) supports
was done as reported previously [29]. First, 16.67 g of P123 was
added to 400 ml of an aqueous 1.67 M HCl solution under stirring
at room temperature. Once a clear solution was obtained, 0.1875 g
of NH4F and 12.5 g of TMB were successively added. The mix-
◦
used filtrate was neutral, and then dried at 40 C under vacuum
◦
overnight. The dried support was dehydrated at 150 C at 0.1 Torr
for 4 h, followed by the addition of 50 ml of dry toluene and 6.9 ml
◦
of APTES. The resulting mixture was refluxed at 110 C under the
◦
ture was stirred at 35 C for 3 h, after which 35.53 g of TEOS
protection of argon for 18 h, and filtered. The solid support was

228
H. Zhang et al. / Journal of Catalysis 256 (2008) 226–236
Scheme 2. Preparation of immobilized Mn(salen) catalysts with the linkage groups axially connected to Mn atoms.
washed successively with toluene, acetone, and methanol and then
dried at room temperature under vacuum for 4 h.
(3TB5CMB: (1R,2R)-(-)-1,2-diaminocyclohexane: 3,5-di-tert-butyl-
2-hydroxybenzaldehyde = 1:2:3). The resulting mixture was stirred
◦
at 25 C for 48 h. The excess of 3,5-di-tert-butyl-2-hydroxybenzal-
dehyde relative to 3TB5CMB yielded a statistical 6:1 ratio of L2
to L1. Then 2 g of 3-aminopropyl-siloxy group-modified support
was added into the mixture with 20 ml of CH2Cl2 and stirred
2.3. Preparation of immobilized Mn(salen) catalysts with the linkage
groups axially connected to Mn atoms
◦
for 18 h at 25 C. The selective capture of L1 and L2 was al-
The preparation of immobilized Mn(salen) catalysts with the
linkage groups axially connected to Mn atoms is schematically
illustrated in Scheme 2. First, 2 g of 3-aminopropyl-siloxy group-
lowed, whereas the soluble L3 was washed away by the CH2Cl2
from the support. After the extraction treatment by CH2Cl2 in a
Soxhlet extractor for 24 h, C2 was predominant on the supports.
Incorporation of a minor amount of L1 in the catalysts had no
apparent deleterious effect on catalyst reactivity or enantioselec-
tivity [32]. Then the immobilization of MnIII salen complexes was
accomplished by refluxing an ethanolic solution of 0.3637 g of
ꢀ
modified support and 0.1 g of (R,R)-(-)N,N -bis(3,5-di-tert-butyl-
salicylidene)-1,2-cyclohexanediaminomangenese(III) chloride (Ja-
cobsen catalyst, Acros, 98%) were added to 40 ml of CH2Cl2 under
◦
stirring at 20 C and keep stirring for 16 h. The resulting light-
brown solid products were separated by filtration, washed with
CH2Cl2, and then extracted by 200 ml of CH2Cl2 in a Soxhlet ex-
tractor for 24 h. The catalysts were extracted twice to ensure good
removal of physically adsorbed Mn(salen) complexes. The extracted
solid products were then dried at room temperature in vacuum to
remove CH2Cl2. These catalysts are designated Salen-Mn-LS, Salen-
Mn-SS, and Salen-Mn-MCF.
◦
Mn(OAc)2·4H2O at 90 C for 1 h, followed by another 30 min at
◦
90 C after the addition of 0.09538 g of LiCl. The resulting solid
products were then washed with CH2Cl2 and CH3OH to obtain
light-brown catalysts, designated Mn-Salen-LS, Mn-Salen-SS, and
Mn-Salen-MCF.
2.5. Asymmetric epoxidation reaction tests of
cis/trans-β-methylstyrene
2.4. Preparation of heterogeneous Jacobsen catalysts via a linkage
connected to salen ligand
The chiral Mn(III) salen complex 1 was prepared in the proce-
dures as illustrated in Scheme 3.
Racemic cis-β-methylstyrene oxides were prepared by the
epoxidation of 1 e.q. of cis-βmethylstyrene with 1.1 e.q. of m-
CPBA in CH2Cl2 at 0 C for 24 h [33]. The NaClO solution was
◦
2.4.1. Synthesis of
prepared as described by Jacobsen et al. [3]. In a typical run of an
epoxidation reaction test, a reaction mixture of 0.118 g of cis/trans-
β-methylstyrene, 0.0175 g of PPNO, 68 μl of n-decane, 5 ml of
CH2Cl2, catalyst (0.0125 g of Jacobsen catalyst for homogeneous
runs or 0.2 g of immobilized Mn(salen) catalyst for heterogeneous
runs), and 5.5 ml of 0.55 M NaClO (pH 11.3) [3] was stirred at
3-tert-butyl-5-(chloromethyl)-2-hydroxybenzaldehyde
To synthesize 3-tert-butyl-5-(chloromethyl)-2-hydroxybenzalde-
hyde (3TB5CMB; C1 in Scheme 3) [31], first, 30.0 g of 3-tert-butyl-
2-hydroxybenzaldehyde and 11.16
g of paraformaldehyde were
added into 112.2 ml of concentrated hydrochloric acid. The re-
◦
sulting mixture was stirred at 25 C for 48 h and then repeatedly
◦
20 C until the conversion of cis/trans-β-methylstyrene was stable.
extracted with diethyl ether. The diethyl ether layer was repeat-
edly washed with saturated aqueous NaHCO3 and brine and then
dried over Na2SO4. A yellow-white crystalline solid (3TB5CMB) was
obtained after the evaporation of the solvent in vacuum. 1H NMR
(CDCl3, 400 MHz): δ (ppm) 1.43 (s, 9H), 4.58 (s, 2H), 7.42 (d, 1H),
The reaction process was monitored by chiral GC. The conver-
sion of cis/trans-β-methylstyrene, the chemical selectivities of the
epoxides, and the ee values of both the cis- and trans-epoxides
were calculated according the chiral GC analysis results obtained
on a Carlo Erba 6000 Vega series 2 instrument equipped with a
Lipodex E capillary column (25 m, 0.25 mm inner diameter). An
internal standard method was used with n-decane as the internal
standard in all epoxidation reaction tests and calibration runs. In
the calibration runs, (1S,2S)-trans-β-methylstyrene oxide, (1R,2R)-
trans-β-methylstyrene oxide, and racemic cis-β-methylstyrene ox-
ides were used to set up the calibration curves. Analysis of the
products with an Angilent 5975 GC-MS instrument found that ben-
zaldehyde was the main byproduct.
7.52 (d, 1H), 9.87 (s, 1H), 11.86 (s, 1H). FT-IR (KBr): 2967, 2910,
−1
2862, 1650, 1611, 1436, 1268, 1235, 1209, 1162, 979, 772, 697 cm
.
2.4.2. Synthesis of immobilized Mn(salen) catalysts with linkage group
connected to salen ligand
This synthesis was based on a procedure reported by Allen
and Jacobsen [32]. First, 0.2127 g of 3TB5CMB, 0.7030 g of 3,5-
di-tert-butyl-2-hydroxybenzaldehyde, and 0.2284
g of (1R,2R)-
(-)-1,2-diaminocyclohexane were added into 10 ml of CH2Cl2

H. Zhang et al. / Journal of Catalysis 256 (2008) 226–236
229
Fig. 2. Small-angle X-ray diffraction patterns of MCF, SS and LS.
are shown in Fig. 2. In the SXRD pattern of the LS material, three
◦
◦
◦
peaks, centered at 0.92 , 1.54 , and 1.78 , were indexed as (100),
(110), and (200) diffractions. The SXRD pattern of the SS material
exhibited a similar periodic structure. The SXRD patterns of the
LS and SS materials confirm a two-dimensional hexagonal symme-
◦
◦
try (p6mm). Two diffraction peaks at 0.61 and 0.86 were seen in
the SXRD pattern of MCF material, suggesting a different periodic
structure from that of LS or SS.
According to the HRSEM and TEM images in Fig. 3, the MCF ma-
terial is composed of large, uniform spherical cells interconnected
by uniform windows to create a 3-D pore system. The SS parti-
cles exhibit hexagonally ordered channels <300 nm long located
along the short axis of these particles. The LS support has hexago-
nally ordered channels growing along the long axis of the particles
with length much greater than 1000 nm. The HRSEM and TEM re-
sults of the support materials are in line with the SXRD results
shown in Fig. 2. The characteristic pore dimensions and channel
structures of these three different support materials remained in-
tact after the heterogenization of Mn(salen) complexes, as shown
by the TEM images of the immobilized catalysts in Fig. 4.
Table 1 gives the textural and chemical properties of support
materials and immobilized catalysts. The manganese content of
the immobilized catalysts was determined by ICP-AES. Both im-
mobilization strategies led to a specific Mn loading. In compari-
son, pure siliceous support materials without modification of the
3-aminopropyl-siloxy group were first physically mixed with Jacob-
sen catalyst or a mixture of L1, L2, and L3 in CH2Cl2, then treated
with Mn(OAc)2 and LiCl. All of the resulting mixed solid products
were then extracted by CH2Cl2 in a Soxhlet extractor. These mate-
rials exhibited no activity in the asymmetric epoxidation reaction
tests of cis/trans-β-methylstyrene according to the GC-MS results.
After three uses, no evident change in the textural properties and
manganese contents of the LS immobilized catalysts was observed.
The pore diameter and pore volume of the three support mate-
rials decreased in the following sequence: MCF > SS > LS. For
the immobilized catalysts, with the 3-aminopropyl-siloxy group
connected to either the Mn atom or the salen ligand, the pore
diameter and pore volume also decrease in the sequence MCF >
Fig. 1. N2 adsorption–desorption isotherms of (a) MCF; (b) SS; (c) LS. The insets are
pore size distributions by adsorption branch of N2 adsorption–desorption isotherms.
3. Results and discussion
3.1. Characterizations of supports and catalysts
3.1.1. Textural and chemical properties of support materials and
immobilized catalysts
Fig. 1 shows the N2 adsorption–desorption isotherms of the
MCF, SS, and LS support materials. All of the support materials
exhibited representative type IV isotherms [34], with hysteresis
loops typical of ordered mesoporous materials with narrow pore
size distribution centered around 29.1 nm for MCF, 14.6 nm for
SS, and 7.5 nm for LS. The SXRD patterns of the support materials

230
H. Zhang et al. / Journal of Catalysis 256 (2008) 226–236
Fig. 3. HRSEM and TEM images of MCF, SS and LS.
SS > LS, as did the channel length of the support materials or the
immobilized catalysts (Figs. 3 and 4).
tic band overlapped by this O–H bending vibration band [22]. The
−1
bands centered at around 1635 cm
in the spectra of the immo-
bilized Mn(salen) catalysts shifted slightly to lower wavenumbers
and increased in intensity, also suggesting immobilization of the
Mn(salen) complex on the support materials. Similar results were
obtained for all immobilized catalysts, indicating that the different
immobilization strategies shown in Schemes 2 and 3 can immo-
bilize the Mn(salen) complex on all three support materials via a
tethering linkage group connected to the salen ligand or to the Mn
atom.
3.1.2. FTIR studies on immobilized catalysts
Fig. 5 shows the FTIR spectra of the support materials and
immobilized Mn(salen) catalysts. The bands at 2857, 2922, and
2963 cm−1, which were absent in the spectra of support ma-
terials but present in the spectra of all immobilized Mn(salen)
catalysts, are attributed to ν(CH2) of the propyl arm of the sily-
lating agent, confirming the immobilization of Mn(salen) complex
via a tethering linkage group [13,19,22]. In addition, the appear-
−1
ance of the band at 3014 cm
in the spectra of all immobilized
3.1.3. UV–vis DRS studies on support materials and immobilized
catalysts
Mn(salen) catalysts can be attributed to the presence of tert-butyl
groups [22], confirming that the salen ligands with the structure as
shown in Schemes 2 and 3 were grafted onto the surface of sup-
port materials. Furthermore, new characteristic bands of Mn(salen)
centered at 1565, 1550, and 1535 cm
tra of all immobilized Mn(salen) catalysts, indicating that the chiral
Mn(salen) complex was immobilized on the support [13,22]. Each
support material exhibited a broad band at around 1635 cm
due to the O–H bending vibrations of adsorbed water. The C=N
groups in the immobilized Mn(salen) catalysts, which participated
in coordination with manganese ions, should give a characteris-
Fig. 6 shows the UV–vis DRS spectra of the LS immobilized
catalyst on LS, and Jacobsen catalyst from Acros. No band can
be seen in the spectrum of LS. The spectra of the Jacobsen cat-
alyst and the mechanically mixed LS + Jacobsen catalyst exhib-
ited similar features. The bands at 256 and 296 nm are from
the π → π∗ excitation, and the band at 323 nm is related to
the n → π∗ excitation of salen ligand. The band at 447 nm is
due to the charge-transfer (CT) transition from the salen ligand
to the metal [10]. The band at 515 nm is attributed to the d–d
transition of Mn(salen) complex [22]. The band at 248 nm was
−1
can be observed in spec-
−1

H. Zhang et al. / Journal of Catalysis 256 (2008) 226–236
231
Fig. 4. TEM images of the immobilized catalysts with 3-aminopropyl-siloxy group connected to Mn atom (Salen-Mn-MCF, Salen-Mn-SS and Salen-Mn-LS) or salen ligand
(Mn-Salen-MCF, Mn-Salen-SS and Mn-Salen-LS).
detected in the spectrum of S and should be assigned to the C–
N group in S1. As shown in Fig. 6a, the spectrum of Salen-Mn-LS
Table 1
shows five bands similar to those for the Jacobsen catalyst and
the LS + Jacobsen catalyst. The bands for Salen-Mn-LS centered at
239, 288, 316, 423, and 513 nm exhibited a blue shift compared
with those for the Jacobsen catalyst and LS + Jacobsen catalyst,
reflecting the interaction between the Mn(salen) complex and the
LS support. When the linkage group was connected to salen lig-
and as shown in Scheme 3, the immobilized salen ligand C2 gave
five bands centered at 225, 261, 288, 328 and 416 nm but no d–
d transition band at 515 nm. The differences in the spectra of
C2 and the Jacobsen catalyst are due to the absence of the Mn
atom in the C2 complex. The spectrum of the Mn-Salen-LS cat-
alyst showed bands at 244, 292, 321, and 421 nm and a d–d
transition band at 516 nm, confirming the production of immo-
bilized Mn(salen) complex on the support. The minor differences
in the spectra of the Mn-Salen-LS and Salen-Mn-LS catalysts re-
flect the different immobilization strategies for these two catalysts.
Similar results were obtained for SS and MCF support materi-
als.
Textural and chemical properties of support materials and immobilized catalysts
Support materials and
immobilized catalysts
Pore
diameter
(nm)
Specific
area
Pore
Content
of Mn
(mg/g)
volume
(m2/g)
(cm3/g)
MCF
SS
LS
29.1
14.6
7.5
28.2
14.0
6.5
27.8
13.7
6.2
533
690
1055
406
573
832
389
568
875
799
811
2.59
1.42
1.11
–
–
–
Mn-Salen-MCF
Mn-Salen-SS
Mn-Salen-LS
Salen-Mn-MCF
Salen-Mn-SS
Salen-Mn-LS
Recycled Mn-Salen-LSa
Recycled Salen-Mn-LSa
1.76
0.89
0.58
1.52
0.96
0.62
0.55
0.57
0.252
0.194
0.236
0.089
0.079
0.076
0.240
0.080
6.1
6.0
The specific areas were calculated following BET method and the data of pore di-
ameter and pore volume were evaluated by BJH method from adsorption isotherm
plots.
a
Recycled for 2 times; in each recycle, the catalyst was separated by filtration and
sequentially washed by distilled water, ethanol, acetonitrile and dichloromethane
and then dried at room temperature in vacuum.

232
H. Zhang et al. / Journal of Catalysis 256 (2008) 226–236
Fig. 5. FT-IR spectra of the immobilized catalysts with 3-aminopropyl-siloxy group connected to Mn atom (Salen-Mn-MCF, Salen-Mn-SS and Salen-Mn-LS) or salen ligand
(Mn-Salen-MCF, Mn-Salen-SS and Mn-Salen-LS).
3.2. Enantioselective epoxidation of cis/trans-β-methylstyrene with
NaClO as oxygen donor
same sequence as found for the production of cis-epoxides. When
the trans-substrate was used in the homogeneous and heteroge-
neous reactions, only trans-epoxides were detected, and the ee
value for trans-epoxides were constant and low (entries 6, 7, and 8,
Table 2). The homogeneous Mn(salen) catalyzed asymmetric epox-
idation of trans-olefins are always poorly enantioselective accord-
ing to the side-on approach mechanism [37] and recent results of
density functional theory (DFT) and quantummechanics/molecular
mechanics (QM/MM) calculations [39]. This is consistent with the
poor enantioselectivity for the immobilized Mn(salen) catalyst-
catalyzed asymmetric epoxidation of trans-β-methylstyrene shown
in Tables 2 and 3.
To characterize the effect of different immobilization strate-
gies on product distribution in heterogeneous reactions, Table 3
presents the catalytic data for the axially immobilized Jacobsen
catalysts (i.e., Salen-Mn-LS, Salen-Mn-SS, and Salen-Mn-MCF) in
the asymmetric epoxidation of β-methylstyrene. The table re-
ports increased trans-epoxide production with cis-substrate. The
cis-epoxide production and the ee values for the trans-epoxides
also decreased in the following sequence: homo->MCF > SS > LS
(entries 2, 3, and 4, Table 3). When trans-substrate was used,
only trans-epoxides were detected, and the ee values changed only
slightly (entries 6, 7, and 8, Table 3).
Table 3 shows much lower chemical selectivities for heteroge-
neous reactions catalyzed by the axially immobilized Mn(salen)
catalysts compared with the homogeneous (entries 1 and 5) and
heterogeneous reactions catalyzed by the catalysts with linkage
groups connected to the salen ligand (entries 2, 3, 4, 6, 7, and
8, Table 2). The recycled Salen-Mn-LS and Mn-Salen-LS catalysts
(Table 1, entries 10 and 11) exhibited almost the same activity
and ee values but slightly lower chemical selectivity compared
with the fresh catalysts. The decomposition reaction of four β-
methylstyrene epoxides catalyzed by Salen-Mn-LS (Table 4) ex-
hibited no activity in the long run. It is known that changes in
the cis/trans ratio and ee values of trans-epoxides do not result
from the further reaction after the formation of β-methylstyrene
epoxides. This suggests that the heterogeneity, resulting from the
confinement effect of nanopores and the motion restriction in
nanochannels on heterogeneous chiral induction, is the main rea-
son for the varying ee values in our cases [5–7,24,25].
Table 2 characterizes the catalytic performance of the homoge-
neous Jacobsen catalyst and the immobilized Mn(salen) catalysts
(i.e., Mn-Salen-LS, Mn-Salen-SS and Mn-Salen-MCF) in the asym-
metric epoxidation of β-methylstyrene in CH2Cl2 with NaClO as
an oxidant. The commercial Jacobsen catalyst exhibited almost the
same activity, cis/trans ratio of epoxides, and enantioselectivity (en-
tries 1 and 5) as were reported by Zhang et al. [35] also with
NaClO as an oxidant. After the Mn(salen) complex was immobi-
lized in the nanopores of previous supports, the β-methylstyrene
conversion, chemical selectivity, and ee values of the epoxides
changed. As shown in Tables 2 and 3, the immobilized Mn(salen)
catalysts exhibited significantly lower conversion of substrate and
slightly lower TOF values compared with the homogeneous cata-
lyst. This decreased conversion is common in heterogeneous asym-
metric catalyzed by epoxidation immobilized Mn(salen) catalysts
[4–8,11–22]. In a homogeneous reaction system, the content of ho-
mogeneous Mn(salen) catalyst is normally not below 4 mol% [3,
36] against substrate, but was 2 mol% in the present study. The
content of immobilized Mn(salen) complex in heterogeneous reac-
tions was significantly lower than that of homogeneous Mn(salen)
catalyst in homogeneous reactions, leading to an evident decrease
in substrate conversion. When the concentration of Mn(salen) in
homogeneous reactions decreased to a level close to that in het-
erogeneous reactions (about 0.04 mol% against substrate), the con-
version of either cis-β-methylstyrene or trans-β-methylstyrene also
decreased dramatically, and the corresponding facial TOF values
dropped to levels comparable to those for heterogeneous reactions
−3 −1
−3 −1
s for trans-
(1.18 × 10
s
for cis-substrate and 2.06 × 10
substrate).
The cis/trans ratio and the ee values differed between heteroge-
neous and homogeneous reactions. In the homogeneous reaction,
the cis-substrate led to more cis-epoxides (entry 1, Table 2), in
agreement with previous reports [3,35–38]. With cis-substrate in
the heterogeneous reactions, the production of cis-epoxides de-
creased in following sequence with changing support materials:
homo->MCFs > SS > LS (entries 2, 3 and 4, Table 2). The change in
support had no distinct effect on the ee values of the cis-epoxides.
The corresponding ee value for trans-epoxides decreases in the

H. Zhang et al. / Journal of Catalysis 256 (2008) 226–236
233
Scheme 3. Preparation of heterogeneous Jacobsen catalysts via a linkage connected to salen ligand.
In the heterogeneous asymmetric epoxidation of cis-β-methyl-
styrene catalyzed by immobilized Mn(salen) catalysts, the radical
intermediate with a sp3 C–C single bond will either collapse to
produce cis-epoxides (the cis-route) or first rotate and then col-
lapse to produce trans-epoxides (the trans-route), as illustrated in
Scheme 4, similar to the reaction in homogeneous asymmetric
in CH2Cl2 with NaClO as an oxygen donor [37]. The ammonium
salts extend the lifetime of radical intermediates, thus permitting
free rotation of the sp3 C–C single bond, and then selectively col-
lapse to trans-epoxides. According to the reaction results given in
Tables 2 and 3, the nanopore wall may play the same role in
heterogeneous reactions as ammonium salts do in the homoge-
neous reaction, resulting in increased trans-epoxide production in
heterogeneous reactions. Due to the different pore sizes and chan-
nel lengths of these three support materials, the motion restriction
in the nanochannels of these materials changes in the sequence
homo-<MCF < SS < LS, and the production of trans-epoxides with
cis-substrate increases in the same sequence.
◦
epoxidation [13,36–42]. At 20 C, ring closure is the slower pro-
cess and thus the rate-determining step [36]. The cis/trans ratio
◦
can be reversed by cooling the reaction mixture to −78 C, which
slows the sp3 C–C single-bond rotation responsible for cis/trans iso-
merization sufficiently so that the ring-closure step becomes the
faster step [43]. In addition, Chang et al. found that quaternary am-
monium salts can induce increased production of trans-epoxides
in the asymmetric epoxidation catalyzed by Mn(salen) complexes
The optimal pore size of the support is crucial to achieving
enhanced enantioselectivity in nanopores [24,25]. In the asym-

234
H. Zhang et al. / Journal of Catalysis 256 (2008) 226–236
Fig. 6. UV–vis DRS spectra of LS support material, Jacobsen catalyst, mechanically mixed Jacobsen catalyst (Jacobsen catalyst + LS) and (a) Salen-Mn-LS and S1 and
(b) Mn-Salen-LS, C2 and S1.
Table 2
Asymmetric epoxidation of β-methylstyrene catalyzed by homogeneous Jacobsen catalyst and immobilized Mn(salen) catalysts (Mn-Salen-LS, Mn-Salen-SS and Mn-Salen-
MCF)
Entry
Substrate
Catalyst
t
(h)
Conv.
(%)
Epoxides sel. (%)
ee (%)
Epoxides config.
Facial TOF
−3 −1 d
Totala
trans-b
cis-b
trans-
cis-
trans-
cis-
(×10
s
)
1
2
3
4
5
6
7
8
cis-
Homo-c
6
48
48
24
6
48
48
24
99.5
22.2
17.8
19.7
100
66.3
31.9
29.5
79.8
75.0
77.1
68.2
100
97.4
79.6
61.6
15.9
22.8
43.1
56.0
100
100
100
100
84.1
77.2
56.9
44.0
n.d.
n.d.
n.d.
n.d.
70.0
42.8
24.5
6.2
14.9
11.3
10.2
7.6
83.4
78.5
78.3
80.5
–
–
–
–
1S,2S
1S,2S
1S,2S
1S,2S
1R,2R
1R,2R
1R,2R
1R,2R
1R,2S
1R,2S
1R,2S
1R,2S
–
–
–
–
1.84
1.15
1.08
1.21
2.31
1.98
1.74
1.62
Mn-Salen-MCF
Mn-Salen-SS
Mn-Salen-LS
Homo-c
Mn-Salen-MCF
Mn-Salen-SS
Mn-Salen-LS
trans-
◦
Reaction temperature is 20 C. The blank runs with the mixtures of cis/trans-methylstyrene, all supports, dichloromethane, n-decane and 0.55 M NaClO exhibit that the
physical adsorption of cis/trans-methylstyrene by SBA-15 is negligible.
a
The molecular proportion of all epoxides in the products.
The molecular proportion of cis- or trans-epoxide in the summary of all epoxides.
b
c
Jacobsen catalyst from Acros.
d
Facial turnover frequency (TOF) is calculated by the expression [product]/[catalyst] × time (s−1) and according to the reaction data at 6 h; n.d., not detected.
Table 3
Asymmetric epoxidation of β-methylstyrene catalyzed by homogeneous Jacobsen catalyst and axially immobilized Mn(salen) catalysts (Salen-Mn-LS, Salen-Mn-SS and Salen-
Mn-MCF)
Entry
Substrate
Catalyst
t
(h)
Conv.
(%)
Epoxides sel. (%)
ee (%)
Epoxides config.
Facial TOF
−3 −1 d
Totala
trans-b
cis-b
trans-
cis-
trans-
cis-
(×10
s
)
1
2
3
4
5
6
7
8
cis-
Homo-c
6
48
48
72
6
48
48
72
99.5
24.8
13.7
13.8
100
41.1
18.2
19.7
79.8
39.4
20.0
27.5
100
88.0
55.0
61.9
15.9
33.8
42.7
48.6
100
100
100
100
84.1
66.2
57.3
51.4
n.d.
n.d.
n.d.
n.d.
70.0
21.9
12.4
13.3
14.9
14.0
11.0
11.4
83.4
79.4
77.3
77.9
–
–
–
–
1S,2S
1S,2S
1S,2S
1S,2S
1R,2R
1R,2R
1R,2R
1R,2R
1R,2S
1R,2S
1R,2S
1R,2S
–
–
–
–
1.84
0.82
0.78
0.89
2.31
1.68
1.41
1.20
Salen-Mn-MCF
Salen-Mn-SS
Salen-Mn-LS
Homo-c
Salen-Mn-MCF
Salen-Mn-SS
Salen-Mn-LS
trans-
◦
Reaction temperature is 20 C. The blank runs with the mixtures of cis/trans-methylstyrene, all support materials, dichloromethane, n-decane and 0.55 M NaClO exhibit that
the physical adsorption of cis/trans-methylstyrene by SBA-15 is negligible.
a
The molecular proportion of all epoxides in the products.
The molecular proportion of cis- or trans-epoxide in the summary of all epoxides.
b
c
Jacobsen catalyst from Acros.
d
Facial turnover frequency (TOF) is calculated by the expression [product]/[catalyst] × time (s−1) and according to the reaction data at 6 h. n.d., not detected.
metric epoxidation of 6-cyano-2,2-dimethylchromene catalyzed by
an immobilized Jacobsen catalyst on different supports via the
phenoxyl group [44], the ee value reached its maximum with
an optimized pore size of 6.2 nm available in a SBA-15 sup-
port. A larger pore size of 7.6 nm led to lower ee values. DFT
and QM/MM computational studies on the epoxidation reaction
of cis/trans-β-methylstyrene catalyzed by anchored oxo-MnV-salen
into MCM-41 channels suggested that the effect of the MCM-41
channel is steric rather than electronic and that the effect of chan-
nel confinement depends strongly on the channel size [39]. In the

H. Zhang et al. / Journal of Catalysis 256 (2008) 226–236
235
Scheme 4. Illustration of the mechanism for the asymmetric epoxidation of cis-β-methylstyrene catalyzed by immobilized Mn(salen) catalysts [13,36–42].
Table 4
plies that in the epoxidaton of cis-β-methylstyrene, the rota-
tion step from cis-intermediate to trans-intermediates is not re-
versible.
Decomposition of β-methylstyrene epoxides catalyzed by Salen-Mn-LS
Epoxides
t
Relative ratio for epoxides
(h)
(1S,2R):(1S,2S):(1R,2R):(1R,2S)
cis-1S,2R
cis-1R,2S
trans-1S,2S
trans-1R,2R
0
24
48
72
96
1:1.02:1.02:1.09
1:1.03:1.03:1.06
1:1.00:1.00:1.04
1:1.01:0.99:1.05
1:0.99:0.99:1.07
4. Conclusion
As demonstrated in Scheme 5, motion restriction and con-
finement effects in the nanochannels of support materials were
found to affect the cis/trans ratio and enantioselectivity of trans-
epoxides in heterogeneous asymmetric epoxidation of noncyclic
olefins, such as cis-β-methylstyrene. The motion restriction in
the nanochannels of immobilized Mn(salen) catalysts increased
the likelihood of rotation for the sp3 C–C single bond of cis-
intermediate and resulted in more trans-epoxides in the hetero-
geneous reactions compared with the homogeneous reaction. Con-
sistent with this increased trans-epoxide production, the enantios-
electivity of trans-epoxides decreased in the heterogeneous reac-
tions due to the significant confinement effect on the ring-closure
step of the trans-intermediates. In contrast, the confinement effect
of nanochannels on the collapse process of the cis-intermediates
was negligible, and thus the ee values of the cis-epoxides hardly
changed. Motion restriction and confinement effects can be tuned
by changing the pore dimension and channel length of support
materials, after which the production distribution of heteroge-
neous asymmetric epoxidation of cis-β-methylstyrene can be ad-
justed sequentially. Our findings also demonstrate that the rota-
tion step from cis-intermediate to trans-intermediates is hardly
reversible.
Reaction conditions are exactly as same as that for the asymmetric epoxidation
tests.
present work, the ee values of trans-epoxides produced from
cis-substrate in heterogeneous reactions decreased sequentially,
but trans-1S, 2S epoxide was always the main product. The LS
support had the smallest pore size, related to the lowest ee
value for trans-epoxides. The more restricted environment of the
nanochannels led to the lower enantioselectivity of the trans-
epoxides produced from cis-substrate. However, in contrast to the
varying ee values of the trans-epoxides, the restricted environ-
ment of the nanochannels in the heterogeneous systems exhib-
ited no apparent effect on the enantioselectivity for cis-epoxides
(entries 1, 2, 3, and 4, Tables
2 and 3). Our results suggest
that the confinement effect of nanochannels on the collapse pro-
cess is considerable for trans-intermediates but negligible for cis-
intermediates.
Our findings also suggest possible increased production of
trans-1R, 2R epoxide in the trans-route in nanochannels, which
could explain the sequentially decreased ee values of trans-
epoxides in heterogeneous reactions shown in Tables 2 and 3.
In nanochannels, more trans-1R, 2R epoxide can be produced
with
a confinement effect on the ring-closure step of trans-
Acknowledgments
intermediates, so that the ee values of trans-epoxides will de-
crease but with a configuration of trans-1S, 2S. Consistent with
the tuning pore size of these three support materials, the ee
value of trans-epoxides produced from cis-substrate in heteroge-
neous reactions was seen to decrease in the following sequence:
homo->MCF > SS > LS. A more homogeneous system-like envi-
ronment, like that in MCF, benefits both the enantioselectivity of
trans-epoxides and the production of cis-epoxides. For the epox-
idation of trans-β-methylstyrene, only trans-epoxides were pro-
duced, and the heterogeneous and homogeneous reactions ex-
hibited similar product distributions (Tables 2 and 3). This im-
This work was supported by the Programme for Strategic Scien-
tific Alliances (PSA) between China and the Netherlands (04-PSA-
M-01, 2004CB720607, NSFC 20773123). The authors thank Pei Yuan,
Chongqing University of Medical Sciences for his help with the
FTIR tests; Dr. Xianliang Li from the Chongqing Entry-Exit Inspec-
tion and Quarantine Bureau for his help with the ICP-AES tests; Dr.
Song Xiang, Hong Kong University for the helpful discussions; and
Dr. Xin Zhan, University of Kentucky for his help with the English
and overall manuscript preparation.

236
H. Zhang et al. / Journal of Catalysis 256 (2008) 226–236
Scheme 5. Illustration of the confinement effect and motion restriction of nanochannels on heterogeneous asymmetric epoxidation.
References
[22] L.L. Lou, K. Yu, F. Ding, X.J. Peng, M.M. Dong, C. Zhang, S.X. Liu, J. Catal. 249
(2007) 102.
[23] F. Cozzi, Adv. Synth. Catal. 348 (2006) 1367.
[1] W. Zhang, J.L. Loebach, S.R. Wilson, E.N. Jacobsen, J. Am. Chem. Soc. 112 (1990)
[24] J.M. Thomas, R. Raja, D.W. Lewis, Angew. Chem. Int. Ed. 44 (2005) 6456.
[25] C. Li, H.D. Zhang, D.M. Jiang, Q.H. Yang, Chem. Commun. (2007) 547.
[26] M.E. Davis, Nature 417 (2002) 813.
[27] P. Piaggio, D. McMorn, P. Murphy, d. Bethell, P.C. Bullmanpage, F.E. Hancock,
C. Sly, O.J. Kerton, G.J. Hutchings, J. Chem. Soc. Perkin Trans. 2 (2000) 2008.
[28] P.O. Norrby, J. Am. Chem. Soc. 117 (1995) 11035.
2801.
[2] R. Irie, K. Noda, Y. Ito, N. Matsumoto, T. Katsuki, Tetrahedron Lett. 31 (1990)
7345.
[3] E.N. Jacobsen, W. Zhang, A.R. Muci, J.R. Ecker, L. Deng, J. Am. Chem. Soc. 113
(1991) 7063.
[4] C. Bianchini, P. Barbaro, Top. Catal. 19 (2002) 17.
[5] C.-E. Song, S.-G. Lee, Chem. Rev. 102 (2002) 3495.
[6] P. Mcmorn, G.J. Hutchings, Chem. Soc. Rev. 33 (2004) 108.
[7] C. Li, Catal. Rev. Sci. Eng. 46 (2004) 419.
[29] D. Zhao, J. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D. Stucky,
Science 279 (1998) 548.
[30] M. Choi, W. Heo, F. Kleitz, R. Ryoo, Chem. Commun. (2003) 1340.
[31] F. Minutolo, D. Pini, A. Petri, P. Salvadori, Tetrahedron Asymm. 7 (1996) 2293.
[32] D. Allen, E.N. Jacobsen, J. Am. Chem. Soc. 121 (1999) 4147.
[33] P. Fristrup, B.B. Dideriksen, D. Tanner, P.O. Norrby, J Am. Chem. Soc. 127 (2005)
13672.
[8] A. Corma, Catal. Rev. Sci. Eng. 46 (2004) 369.
[9] M. Heitbaum, F. Glorius, I. Escher, Angew. Chem. Int. Ed. 45 (2006) 4732.
[10] D.P. Serrano, J. Aguado, Appl. Catal. A 335 (2008) 172.
[11] S. Xiang, Y.L. Zhang, Q. Xin, C. Li, Chem. Commun. (2002) 2696.
[12] H.D. Zhang, S. Xiang, C. Li, Chem. Commun. (2005) 1209.
[13] H.D. Zhang, Y.M. Zhang, C. Li, J. Catal. 238 (2006) 369.
[14] T.S. Reger, K.D. Janda, J. Am. Chem. Soc. 122 (2000) 6929.
[15] D.W. Park, S.D. Choi, S.J. Choi, C.Y. Lee, G.J. Kim, Catal. Lett. 78 (2002) 145.
[16] V.D. Chaube, S. Shylesh, A.P. Singh, J. Mol. Catal. A Chem. 241 (2005) 79.
[17] M. Cardoso, A.R. Silva, B. de Castro, C. Freire, Appl. Catal. A 285 (2005) 110.
[18] I.K. Biernacka, A.R. Silva, A.P. Carvalho, J. Pires, C. Freire, Langmuir 21 (2005)
10825.
[19] R.I. Kureshy, I. Ahmad, N.H. Khan, S.H.R. Abdi, K. Pathak, R.V. Jasra, J. Catal. 238
(2006) 134.
[20] M. Holbach, M. Weck, J. Org. Chem. 71 (2006) 1825.
[21] B.M. Choudary, T. Ramani, H. Maheswaran, L. Prashant, K.V.S. Ranganath, K.V.
Kumar, Adv. Synth. Catal. 348 (2006) 493.
[34] K.S.W. Sing, D.H. Everett, R.A.W. Haul, L. Moscou, R.A. Pierotti, J. Rouquerol,
T. Siemieniewska, Pure Appl. Chem. 57 (1985) 603.
[35] W. Zhang, N.H. Lee, E.N. Jacobsen, J. Am. Chem. Soc. 116 (1994) 425.
[36] W. Zhang, E.N. Jacobsen, J. Org. Chem. 56 (1991) 2296.
[37] S. Chang, J. M Galvin, E.N. Jacobsen, J. Am. Chem. Soc. 116 (1994) 6937.
[38] W. Adam, K.J. Roschmann, C.R. Saha-Möller, D. Seebach, J. Am. Chem. Soc. 124
(2002) 5068.
[39] K. Malek, C. Li, R.A. Van Santen, J. Mol. Catal. A Chem. 271 (2007) 98.
[40] M. Palucki, P.J. Pospisil, W. Zhang, E.N. Jacobsen, J. Am. Chem. Soc. 116 (1994)
9333.
[41] S. Chang, N.H. Lee, E.N. Jacobsen, J. Org. Chem. 58 (1993) 6939.
[42] K. Malek, A.P.J. Jansen, C. Li, R.A. Van Santen, J. Catal. 246 (2007) 127.
[43] H.H. Lau, Angew. Chem. 73 (1961) 423.
[44] H.D. Zhang, C. Li, Tetrahedron 62 (2006) 6640.