Catalytic performance of H-β nanozeolite microspheres in one-pot dynamic kinetic resolution of aromatic sec-alcohols

Article abstract of DOI:10.1016/j.jcat.2011.12.021

In this paper, uniform H-β nanozeolite microsphere (β-ZMS) prepared by polymerization-induced colloid aggregation, combined with an immobilized lipase B from Candida antarctica (commercially available Novozym 435), is applied as a racemization catalyst to the one-pot dynamic kinetic resolution (DKR) of aromatic sec-alcohols. The DKRs of various aromatic sec-alcohols can be carried out with high selectivity and conversion under optimum conditions. Importantly, it is found that, compared to commercial zeolites β (C-β), β-ZMSs display high selectivity for small sec-alcohol molecules because their shorter microporous channels reduce the diffusion time of substrate/product in the three-dimensional framework, and thereby decrease the possibility of secondary reactions. However, the β-ZMSs show a high racemization rate in the DKR of large sec-alcohol molecules, such as racemic 1-(1-naphthyl)-ethanol, thanks to their abundant accessible active sites. This observation will not only benefit the performance of acidic zeolite catalysts in one-pot DKR systems but also provide a new perspective on the application of nanozeolites in catalysis.

Full text of DOI:10.1016/j.jcat.2011.12.021

Journal of Catalysis 288 (2012) 24–32
Contents lists available at SciVerse ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Catalytic performance of H-b nanozeolite microspheres in one-pot dynamic
kinetic resolution of aromatic sec-alcohols
⇑
Xiang Li, Yi Shi, Zhoujun Wang, Yahong Zhang , Yi Tang
Department of Chemistry, Shanghai Key Laboratory of Molecular Catalysis Innovative Materials, Fudan University, Shanghai 200433, People’s Republic of China
a r t i c l e i n f o
a b s t r a c t
Article history:
In this paper, uniform H-b nanozeolite microsphere (b-ZMS) prepared by polymerization-induced colloid
aggregation, combined with an immobilized lipase B from Candida antarctica (commercially available
NovozymÒ 435), is applied as a racemization catalyst to the one-pot dynamic kinetic resolution (DKR)
of aromatic sec-alcohols. The DKRs of various aromatic sec-alcohols can be carried out with high selectiv-
ity and conversion under optimum conditions. Importantly, it is found that, compared to commercial zeo-
lites b (C-b), b-ZMSs display high selectivity for small sec-alcohol molecules because their shorter
microporous channels reduce the diffusion time of substrate/product in the three-dimensional frame-
work, and thereby decrease the possibility of secondary reactions. However, the b-ZMSs show a high
racemization rate in the DKR of large sec-alcohol molecules, such as racemic 1-(1-naphthyl)-ethanol,
thanks to their abundant accessible active sites. This observation will not only benefit the performance
of acidic zeolite catalysts in one-pot DKR systems but also provide a new perspective on the application
of nanozeolites in catalysis.
Received 3 November 2011
Revised 22 December 2011
Accepted 22 December 2011
Available online 17 February 2012
Keywords:
Nanozeolite microsphere
Dynamic kinetic resolution
Short diffusion channel
Large external surface
Accessible acidic site
Ó 2012 Elsevier Inc. All rights reserved.
1. Introduction
netic resolution (KR) of racemates, transformation starting from
the limited chiral pool of natural resources, and stereoselective
Zeolite materials are always of scientific and technological
interest because of their unique porous crystal structures with high
shape selectivity, acid catalytic activity, and thermal/hydrothermal
stability. The applications of zeolite materials generally involve ion
exchange, adsorption, and catalysis, while many of these benefit
from their highly ordered microporous inherence [1–4]. In recent
years, nanozeolites have been paid considerable attention because
the reduction of particle size from the micrometer to the nanome-
ter scale could lead to a great difference in material properties. The
smaller crystal size can endow them not only with large external
surface area and more accessible active sites, but also with short
diffusion paths for the guest molecules, compared with conven-
tional zeolites with micrometer size. Therefore, nanozeolite mate-
rials are expected to achieve better performance in conventional
zeolite applications such as heterogeneous catalysis, molecular
separation, and ion exchange [5–15].
reaction using (mostly) noble metal complexes with chiral ligands
or biocatalysts [16–18]. Among these, the KR of racemates is still
the most common and important industrial approach to synthesiz-
ing enantiomerically pure compounds, although asymmetric catal-
ysis has undergone tremendous advances in the past few decades
and many studies of the chiral synthesis of aromatic sec-alcohols
are reported [17]. However, the classical KR process can transform
one of the enantiomers in a racemic mixture to the desired product
by using biocatalysts or organometallic compounds as chiral cata-
lysts, while the other is unchanged. Therefore, it must face the lim-
itation of maximum 50% yield. Alternatively, dynamic kinetic
resolution (DKR), which associates a racemization step with the
KR process, offers the prospect of up to 100% yield of the desired
stereoisomer from a racemic starting material. This process has be-
come not only an alternative to classic KR, but also a new proce-
dure for asymmetric synthesis [19–23].
Today, obtaining optically pure enantiomers is an extraordinary
absorbing and challenging task. Optically pure aromatic sec-alco-
hols, as important chiral building blocks, are one of the very impor-
tant target compounds in the fine chemical industry [16]. So far,
three methods of obtaining a single enantiomer are available: ki-
In DKRs of aromatic sec-alcohols, because lipase is the most
widely used as a KR catalyst, the racemization catalyst should be
available when the lipase is used with high activity [24]. Up to
now, several racemization catalysts, such as homogeneous metal
complexes, supported metal catalysts, and solid acids/bases, have
been employed to match the lipase-catalyzed KR process. At pres-
ent, the powerful combination of metal catalysis with enzyme
catalysis, involving a hydrogen-transfer mechanism, has become
⇑
Corresponding author. Fax: +86 21 65641740.
E-mail address: zhangyh@fudan.edu.cn (Y. Zhang).
0021-9517/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.12.021

X. Li et al. / Journal of Catalysis 288 (2012) 24–32
25
the subject of spectacular development and has been well re-
2. Experimental
viewed with an increasing number of publications [19–23,25–
27]. Acidic heterogeneous catalysts such as the acid resins Amber-
lyst [28] and Deloxan [29] and H-zeolite [30–37] provide an alter-
native route for racemization, which involves the elimination and
readdition of hydroxyl groups of aromatic alcohols [38]. In 2003,
a biphasic system was constructed by Jacobs et al. [30,31] to carry
out DKR of sec-alcohols. In this biphasic system, racemization cat-
alyzed by H-zeolite was accomplished in the aqueous phase, while
the KR process induced by immobilized lipase B from Candida ant-
arctica (commercially available NovozymÒ 435) was performed in
the organic phase. Recently, Jaenicke et al. [32] used hydrophobic
zeolite b containing a low concentration of Zr or Al associated with
NovozymÒ 435 to achieve one-pot DKRs of sec-alcohols in organic
solvent. In this case, the product can be prepared at >85% yield and
>97% enantiomeric excess (ee) in 2.5 h at 60 °C when racemic 1-
phenylethanol (rac-1-PE) was adopted as the starting material. In
addition, Masters et al. [33] investigated the DKR performance of
zeolite H-b compartmentalized by a layer-by-layer (LBL) assembly
of polyelectrolyte with the free lipase. In the same area, both the
Iborra group [34,35] and the Thomas group [36] reported a similar
excellent DKR system of rac-1-PE using NovozymÒ 435 and acidic
zeolite as catalysts, and an excellent ionic liquid/supercritical car-
bon dioxide biphasic system provided a 98.0% yield and 97.3% ee
for product at 50 °C under 100 bar. However, as indicated by Pellis-
sier [24], the use of acid zeolites as catalysts could generate several
secondary products, namely via dehydration to styrene (eventually
followed by oligomerization) and coking, which is responsible for
both yield reduction and catalyst deactivation. In particular, the
formation of heavy products will become more serious in zeolite
H-b due to its large three-dimensional channel. Zeolite H-ZSM-5
with an intermediate pore size was used by Lemos et al. [37] to
prevent the secondary products in the DKR of rac-1-PE. Unfortu-
nately, the formation of heavy products related to the presence
of vinyl acetate caused the yield to be limited to 17% [24].
It is imaginable that the small crystal size and short micropo-
rous channels of nanozeolite will shorten the diffusion pathway
for guest molecules in the three-dimensional network and de-
crease the formation of by-products (such like ether and styrene)
in one-pot DKR systems. However, the difficult manipulation of
nanozeolite particles is an obstacle in most of their applications,
whereas the conventional calcination procedures inevitably
cause their aggregation and thereby the lessening of accessible
active surface. Recently, a polymerization-induced colloid aggre-
gation (PICA) method was adopted by our group to prepare uni-
form nanozeolite microsphere (ZMS) via copolymerization of
nanozeolite and urea–formaldehyde (UF) resin [39,40]. After
the polymer is removed, the spherical ZMSs are proved not only
to maintain the surface properties of their parent nanozeolites
well but also to possess enriched secondary mesopores (large
than 30 nm) between the nanozeolites, which will solve the dif-
ficulties in collection and handling of nanozeolites, and provide a
hierarchical porous structure to further facilitate their practical
application.
2.1. Materials
Tetraethylammonium hydroxide (TEAOH, 25 wt.%) and silica
(Aerosil 400) were purchased from Sinopharm Chemical and Yixing
Dahua Chemical Reagent Company, respectively. Hydrochloric acid
(HCl, 36–38 wt.%), aluminum foil (99.5%), urea ((NH2)2CO), formal-
dehyde (HCHO, 37 wt.%), and anhydrous ethanol were obtained
from Shanghai Chemical Reagent Company. The substrates of
sec-alcohols, rac-1-PE (Fluka, >98%), (S)-1-phenylethanol (S-1-PE,
Acros, >99%, S:R > 99.5:0.5 ee), 1-phenyl-1-propanol (Fluka, 98%),
1-(p-tolyl)ethanol (Fluka, 97%), 1-(4-bromophenyl)-ethanol (Acros,
98%), 1-indanol (Aldrich, 99%), rac-1-NPE (Aldrich, >99%), and S-1-
(1-naphthyl)-ethanol (S-1-NPE, Aldrich,>99%, S:R > 99:1 ee) were
purchased from Fluka, Acros, and Aldrich. Vinyl octanoate (VO)
(98%) was obtained from ABCR. Cyclohexane (AR), n-hexane (AR),
toluene (>99%), and chloroform (AR) were from Shanghai Dahe
Chemical Reagent Company. Decane (>99%) was obtained from
Acros. Commercial zeolite b (C-b, SiO2/Al2O3 = 25) catalyst was
obtained from Nankai Catalyst Company. NovozymÒ 435 was
purchased from Aldrich. Hexane (HPLC), isopropanol (HPLC), and
acetonitrile (HPLC) came from Sinopharm Chemical Reagent
Company, and n-butylamine (P99%) was obtained from Aladdin
reagent company, while 2,6-di-tert-butylpyridine (P97%) was
from Aldrich.
2.2. Preparation of Catalysts
2.2.1. Synthesis of colloidal zeolite b nanoparticles [39]
Colloidal zeolite b nanoparticles with a SiO2/Al2O3 ratio of 32
were hydrothermally synthesized by the following procedure with
a
mole ratio of SiO2:Al2O3:(TEA)2O:H2O = 1:0.033:0.135:6.8.
Typically, 3.4 g of fumed silica was dissolved with 6.0 g of
25 wt.% TEAOH aqueous solution and stirred for 12 h (Solution I).
Then, 0.101 g of aluminum foil was dissolved in 3.0 g of 25 wt.%
TEAOH, and the mixture was stirred to get a clear solution
(Solution II). Solution II was added to Solution I with violent
stirring. After continuous stirring for another 1 day, the obtained
mixture was transferred into a Teflon-lined autoclave at 140 °C
for 14 days to get the colloidal solution of nanozeolite b. The colloi-
dal solution of nanozeolite b was collected without any treatment.
2.2.2. Preparation of b-ZMS [39,40]
The above-obtained colloidal solution was used directly to pre-
pare b-ZMS by the PICA method. Typically, 500
l
L of ethanol,
L of HCl
500 L of urea solution in water (0.425 g/mL), and 100
l
l
solution (6 mol/L) were added to the suspension containing
80 mg of zeolite nanocrystals and the resultant mixture was stirred
for uniformity. Afterward, 380 lL of formaldehyde solution
(38 wt.%) was added under stirring and continually mixed at ambi-
ent temperature (25 °C). The total volume was kept at 4 mL. The
product was recovered by filtration, washed with distilled water
and ethanol, and dried at 70 °C. Finally, the product was calcined
in air at 550 °C for 4 h at a heating rate of 1 °C/min to remove
the UF polymer and the organic templates in the nanozeolites.
Herein, H-b nanozeolite microsphere (b-ZMS) prepared by the
PICA method, combined with NovozymÒ 435, was employed to
carry out the DKR process of rac-1-PE and its analogs. The one-
pot DKR process of various aromatic sec-alcohols can perform
quickly and highly selectively under optimum conditions. The cat-
alytic performance of b-ZMSs in the DKR system is evaluated and
discussed, and it is concluded that the short reaction channels
and abundant accessible active sites of b-ZMSs are indeed helpful
in improving the performance of acidic zeolite catalysts in one-
pot DKR systems, and thus, b-ZMSs display distinct advantages in
reaction selectivity and rate for small and large substrate mole-
cules, respectively.
2.3. Catalyst characterization
Scanning electron microscopic (SEM) studies of b-ZMS and C-b
were performed on a Philips XL 30 with accelerating voltage
20 kV. Their N2 sorption isotherms were measured using
Micromeritics ASPS 2010 system at liquid nitrogen temperature.
The crystalline types of zeolite materials were characterized by
a

26
X. Li et al. / Journal of Catalysis 288 (2012) 24–32
X-ray diffraction (XRD) on a Bruker AXS D8 diffractometer with Cu
3. Results and discussion
Ka
radiation at 40 kV and 40 mA.
The number and strength of acidic sites in each zeolite sample
3.1. One-pot DKR process of sec-alcohols
were determined by chemical titration on a potentiometric titra-
tion meter (ZDJ-5, Shanghai Leici Instrument Factory) using n-
butylamine and 2,6-di-tert-butylpyridine following a reported pro-
cedure [41–44]. The solid (0.10 g) was suspended in 25 ml of ace-
tonitrile and agitated for 3 h. Then, the suspension was titrated
with 0.026 mol/L n-butylamine or 2,6-di-tert-butylpyridine in ace-
tonitrile at a rate of 0.1 mL/min. The electrode potential variation
was measured with a manual continuous titration model using a
double junction electrode. The addition continued until no further
change of mV was recorded.
A one-pot DKR process mainly consists of two tandem pro-
cesses: (1) stereoselective KR and (2) chemical racemization of
the off-isomer. This means that the two different reactions cata-
lyzed by two catalysts have to synergistically proceed under the
same reaction conditions and at the same time, that is, the reaction
rates of the two tandem reactions should match each other
throughout the whole reaction under defined reaction conditions.
Thus, the reaction conditions such as solvent, temperature, and ra-
tio and amount of two catalysts will influence not only the cooper-
ation of the two reactions, but also the whole DKR reaction rate
and selectivity. Here, rac-1-PE is first used as a probe molecule to
monitor an optimum reaction condition, where VO is used as an
acyl donor of a transesterification process.
2.4. Catalytic testing
2.4.1. Racemization reaction
The racemization of enantiopure sec-alcohols was carried out in
a glass reactor with stirring at a desired temperature. Typically, the
enantiopure sec-alcohols (0.10 mmol) and b-ZMS/C-b (4.0–8.0 mg)
were added in various organic solvents (2 mL). The reaction mix-
ture was immediately dipped into a water bath for a desired time.
3.1.1. Effect of solvents
The solvents play an important role both in b-ZMS-catalyzed
racemization and in NovozymÒ 435-catalyzed transesterification
[23]. The racemization of S-1-PE and the transesterification of
rac-1-PE were evaluated in various organic solvents with dielectric
constants from 2 to 8. As shown in Table 1, both racemization and
transesterification reactions display very poor reaction activity in
tetrahydrofuran (THF), which has the highest polarity among the
five solvents. Also, the racemization reaction shows high conver-
sion but low selectivity with increasing polarity of the solvents.
However, the transesterification reaction presents a contrary ten-
dency with the increase in the solvent polarity, that is, lower polar-
ity leads to a higher reaction rate. For example, chloroform, with
higher polarity, causes a very high racemization rate and low selec-
tivity, but the transesterification reaction almost cannot occur in it.
Therefore, cyclohexane, with medium polarity, seems to be a good
candidate for a DKR system. Furthermore, the DKR results of rac-1-
PE in the different solvents also prove that the best results (conver-
sion and yield) are obtained by using cyclohexane as a solvent.
Therefore, cyclohexane is employed as the solvent in the following
one-pot DKRs of sec-alcohols.
2.4.2. KR process (transesterification)
The KR process of racemic sec-alcohols took place in a glass
reactor with stirring at
a desired temperature. Generally,
0.10 mmol racemic sec-alcohols and 0.20 mmol VO were added
into organic solvents (2.0 mL), and then, NovozymÒ 435 (4 mg)
was added into the reaction mixture. The reaction mixture was
immediately dipped into a water bath for a desired time.
2.4.3. One-pot DKR of aromatic sec-alcohols
The one-pot DKR of racemic sec-alcohols was carried out in a
glass reactor with stirring at a desired temperature. Typically,
0.10 mmol racemic sec-alcohols and 0.20 mmol VO were first
added into organic solvents (2.0 mL). Then, NovozymÒ 435 (4–
32 mg) and b-ZMS or C-b (2–8 mg) were added into the reaction
mixture. The reaction mixture was immediately dipped into a
water bath for a desired time.
Yields and ee of products (eep) and ee of substrates (ees), except
for rac-1-NPE, were monitored using decane as an internal stan-
dard by gas chromatography (GC, HP 5890) on a CP-CHIRASIL-
3.1.2. Temperature effects
To find the optimum reaction temperature for the DKR of rac-1-
PE, the reactions are carried out at temperatures ranging from 30
to 80 °C, and the results are listed in Table 2. Obviously, the tem-
perature has a great effect on the racemization of S-1-PE catalyzed
by the acidic b-ZMS. The rate of racemization is usually higher at
higher temperatures, as expected. However, the increase of by-
product at higher temperatures, such as styrene, decreases the
selectivity of the reaction. For example, the side reaction mainly
occurs at 80 °C, leading to a very low selectivity of 5% at ees% = 0.
However, the decrease of ees becomes very slow at 30 °C, indicat-
ing that almost no racemization takes place during the reaction. At
50 and 60 °C, the racemization reaction shows good conversion
and selectivity. For transesterification catalyzed by NovozymÒ
435, the reaction displays the same trend with increasing temper-
ature as that of racemization. Finally, the temperature is fixed at
50 °C according to their one-pot DKR results under different tem-
peratures. At 50 °C, a selectivity of 93% and a conversion of 72%
for the product can be obtained within 30 min.
DEX CB chiral column (25 m ꢀ 0.25
lm, CP7502) with FID detec-
tion. For rac-1-NPE, the result of DKR was determined by high-per-
formance liquid chromatography using a Dicel Chiracel OJ-H
column. The side products, such as styrene and aromatic ether,
were determined by using a Finnigan Voyager gas chromato-
graph–mass spectrometer equipped with the same chiral column.
The ees and conversion of substrates, as well as the selectivity,
yield, and eep, of the (R)-ester, were calculated by the following
equations:
½SubstrateꢁðS;tÞ ꢂ ½Substrateꢁ
ees ¼
ðR;tÞ ꢀ 100%
ð1Þ
ð2Þ
½SubstrateꢁðS;tÞ þ ½SubstrateꢁðR;tÞ
½Substrateꢁt
½Substrateꢁ0
Conv: ¼ 1 ꢂ
Selec: ¼
ꢀ 100%
½Productꢁt
ꢀ 100%
ð3Þ
ð4Þ
ð5Þ
½Substrateꢁ0 ꢀ Conv:
Y: ¼ Conv: ꢀ Selec:
½ProductꢁðR;tÞ ꢂ ½Productꢁ
3.1.3. Amount and ratio of the two catalysts in a one-pot DKR process
In a one-pot DKR process, the amount and ratio of the catalysts
also are studied, and the results are listed in Table 3. When the
amount of b-ZMS is kept constant (Table 3, entries 1–5), the ee
of the substrate increases dramatically with increasing amounts
eep ¼
ðS;tÞ ꢀ 100%
½ProductꢁðR;tÞ þ ½ProductꢁðS;tÞ

X. Li et al. / Journal of Catalysis 288 (2012) 24–32
27
Table 1
The effects of solvents on racemization of S-1-PE catalyzed by b-ZMS, transesterification of rac-1-PE catalyzed by NovozymÒ 435, and DKR of rac-1-PE catalyzed by b-ZMS and
NovozymÒ 435.
Solvent
Racemizationa
Dielectric constantb
1.8
Transesterificationd
c
t (min)
ees (%)
Selec. (%)
t (min)
Conv. (%)
90
eep (%)
>99
Hexane
30
90
30
45
30
45
30
360
61
19
33
6
28
2
92
90
89
83
83
78
78
–
30
Cyclohexane
Toluene
2.1
2.3
30
30
89
75
>99
>99
Chloroform
THF
4.9
7.6
2
30
360
9.5
–
>99
–
>99
Dielectric constant
t (min)
ees (%)
Conv. (%)
Selec. (%)
Y.(%)
eep(%)
DKRe
Hexane
1.8
2.1
2.3
30
60
30
60
30
60
25
31
18
16
10
12.5
68
87
72
92
67
94
95
83
93
83
78
70
66
73
67
76
53
66
>99
>99
>99
>99
>99
97
Cyclohexane
Toluene
Note: The volumes of solvents used in all the reactions are 2 mL, and all the reaction temperatures are fixed at 50 °C.
a
4 mg of b-ZMS and 0.10 mmol of S-1-PE.
Data from the Handbook of Solvents, Linneng Cheng, 3rd ed., Chemical Industry Press, 2002.
b
c
The ee of the initial S-1-PE (ees) is 100%, and it will decrease with the ongoing reaction.
d
4 mg of NovozymÒ 435, 0.10 mmol of rac-1-PE, and 0.20 mmol VO.
e
4 mg of b-ZMS and 4 mg of NovozymÒ 435, 0.10 mmol of rac-1-PE and 0.20 mmol of VO.
Table 2
The effects of temperature on racemization of S-1-PE catalyzed by b-ZMS, transesterification of rac-1-PE catalyzed by NovozymÒ 435, and DKR of rac-1-PE catalyzed by b-ZMS and
NovozymÒ 435.
T (°C)
Racemizationa
ees (%)
Transesterificationb
Conv. (%)
DKRc
Selec. (%)
eep (%)
ees (%)
Conv. (%)
Selec. (%)
eep (%)
30
50
60
80
>99
33
0
–
89
70
5
13
89
92
>99
>99
>99
97
15
18
10
0
13
72
91
100
93
75
>99
>99
>99
95
0
100
>99
26
Note: The reaction solvent is cyclohexane (2 mL) and all the reaction times are 30 min in water bath under vigorous stirring.
a
4 mg of b-ZMS and 0.10 mmol S-1-PE.
b
4 mg of NovozymÒ 435, 0.10 mmol of rac-1-PE, and 0.20 mmol of VO.
c
4 mg of b-ZMS, 4 mg of NovozymÒ 435, 0.10 mmol of rac-1-PE and 0.20 mmol of VO.
Table 3
The DKR results of rac-1-PE over different amounts of NovozymÒ 435 and b-ZMS.a
Entryb
NovozymÒ 435 (mg)
b-ZMS (mg)
NovozymÒ 435/b-ZMS
ees (%)
Conv. (%)
Selec. (%)
Y.(%)
eep (%)
1
2
3
4
5
6
7
8
9
4
6
8
10
12
2
4
12
16
4
4
4
4
4
1
2
6
8
1.0
1.5
2.0
2.5
3.0
2.0
2.0
2.0
2.0
16
28
37
49
58
66
60
26
14
91
96
96
97
97
65
89
93
97
83
83
87
87
87
96
93
88
85
75
79
84
85
85
62
83
81
82
>99
>99
>99
>99
>99
>99
>99
>99
>99
a
Various amounts of b-ZMS and NovozymÒ 435 were added into 2 mL cyclohexane containing 0.10 mmol rac-1-PE and 0.20 mmol VO at 50 °C
in a water bath under vigorous stirring.
b
The reaction times of entries 1–7 and 8, 9 are 1.0 h and 30 min, respectively.
of NovozymÒ 435, indicating the enhancement of the transesterifi-
cation rate over that of racemization. The selectivity and
conversion of the reaction reach an equilibrium at NovozymÒ
435/b-ZMS (weight ratio) P2.0 (Table 3, entries 3–5). This means
that the two catalysts determine the DKR process jointly and there
will be an optimum ratio of the two catalysts. When the ratio of the
two catalysts is anchored at 2.0, the conversion of the DKR process
increases but its selectivity decreases with the increase in the cat-
alyst amounts (Table 3, entries 3, 6–9). The reaction reaches an
optimum conversion and selectivity at entry 7 (4 mg NovozymÒ
435 and 2 mg b-ZMS). It is found that the by-products mainly form
in the path of the racemization catalyzed by b-ZMS, so that the
selectivity has a negative correlation with the amount of b-ZMS.
3.1.4. Addition strategy of the two catalysts in the one-pot DKR process
Considering that the existence of the b-ZMS can result in by-
products, a strategy is adopted that in which after the NovozymÒ
435 is firstly added into the solution for a period, after which the

28
X. Li et al. / Journal of Catalysis 288 (2012) 24–32
b-ZMS is added into the reaction solution to further improve the
reaction selectivity. The experimental details and results are listed
in Table 4. According to Table 4, the selectivity of the DKR process
continuously increases, whereas its conversion keeps remains sta-
ble from mode 1 to mode 4. NovozymÒ 435 partly transfers R-1-PE
to R-ester in advance, and the resulted resulting low concentration
of the substrate limits the production of the side product catalyzed
by the acidic b-ZMS. Furthermore, when the amount of the cata-
lysts is increased, this trend becomes clearer (M5 in Table 4 and
entry 3 in Table 3), indicating that the two-step experimental pro-
cess can well improve the reaction selectivity and avoid the side
reactions caused by the existence of large amounts of acidic cata-
lyst in the one one-step process.
Under such optimum condition, the DKR of rac-1-PE is highly
selective and fast at 50 °C. That is, the DKR of rac-1-PE under such
reaction conditions can reach nearly 90% yield and >99% ee of
product within 1 h, which is faster than all previous reports of
one-pot DKR of rac-1-PE driven by the acidic catalysts in organic
solvent [30–33,37] and its selectivity also is higher than for the
previous one-pot DKR of rac-1-PE catalyzed by H-zeolite and lipase
when b-ZMSs are used to replace the conversional H-b zeolite.
3.2. Catalytic performance of b-ZMS in the one-pot DKR of aromatic
sec-alcohols
The ZMSs prepared by PICA maintain the surface properties of
their nanoparticles well [39,40], and their secondary mesopore
architectures resulting from the stacking of nanozeolite particles
are expected to lower the diffusion limitation of substrate mole-
cules. Therefore, the catalytic performance of b-ZSMs is systemi-
cally evaluated in the one-pot DKR of the various aromatic sec-
alcohols, and C-b is used as reference. The scope of the DKR system,
as well as the detailed reaction conditions and results, are listed in
Table 5. Because the DKR processes of the different sec-alcohols
possess different peculiarities due to their different molecular
structures and activities, as shown in Table 5, the reaction condi-
tions are adjusted to optimize the selectivity and yield of products
(R-esters) with a high ee value. For example, DKR of 1-indanol is
carried out at a high catalyst ratio (NovozymÒ 435/b-ZMS = 4.0)
in order to obtain the expected selectivity and yield of the product.
The structure of 1-indanol can easily form 1-indene with strong
conjugation with acid sites on zeolite, and so the amount of the
Table 4
The DKR results of rac-1-PE under different reaction strategies.a
Mode
t(t1/t2)b (min)
ees (%)
Conv. (%)
Selec. (%)
Y.(%)
eep (%)
M1
M2
M3
M4
M5c
0/60
60
70
78
80
27
48
89
86
87
86
91
98
93
94
95
97
93
91
83
80
83
84
85
89
>99
>99
>99
>99
>99
>99
10/60
20/60
30/60
15/30
15/45
a
2 mg of b-ZMS and 4 mg of NovozymÒ 435 were added in 2 mL cyclohexane
solution containing 0.10 mmol rac-1-PE and 0.20 mmol VO at 50 °C in water bath
under vigorous stirring.
b
t1 and t2 in t refer the time of transesterification in the presence of only Nov-
ozymÒ 435 and that of the DKR, respectively.
c
4 mg of b-ZMS and 8 mg of NovozymÒ 435 were used.
Table 5
The DKR processes of various aromatic sec-alcohols catalyzed by b-ZMS/C-b and NovozymÒ 435.a
Sub.
Cat.
t (t1/t2)b (min)
Results
Substrates
ees (%)
Products (R-ester)
By-products
Conv. (%)
Selec. (%)
Y. (%)
eep (%)
Selec.(RAC@C) (%)
Selec.(RAOAR) (%)
C-b
b-ZMS
15/45
15/45
52
48
94
98
80
91
75
89
>99
>99
15
7
5
2
C-b
15/45
15/45
<5
<5
99
99
80
87
79
86
>99
>99
16
10
4
3
b-ZMS
C-b
b-ZMS
15/20
15/20
83
12
80
97
95
91
76
88
>99
>99
5
9
0
0
C-b
b-ZMS
30/75
30/75
>99
74
71
86
94
96
66
83
>99
>99
6
4
0
0
C-b
b-ZMS
30/30
30/30
25
<5
85
97
84
82
72
80
>99
>99
16
20
0
0
C-b
b-ZMS
90/105
90/105
>99
>99
60
95
96
95
58
90
>99
>99
4
5
0
0
a
4 mg of b-ZMS (or C-b) and 8 mg of NovozymÒ 435 were added into 2 mL of cyclohexane containing 0.10 mmol racemic sec-alcohols and 0.20 mmol VO at 50 °C in a water
bath under vigorous stirring.
b
t1 and t2 in t refer the time of KR in the presence of only NovozymÒ 435 and that of DKR, respectively.
c
8 mg of NovozymÒ 435 was replaced with 16 mg of it in the DKR of racemic 1-indanol.
d
8 mg of b-ZMS (or C-b) and 32 mg of NovozymÒ 435 were used in the DKR of rac-1-NPE.

X. Li et al. / Journal of Catalysis 288 (2012) 24–32
29
racemization catalyst has to be limited. For rac-1-NPE with a large
molecular size, a large amount of catalyst and long reaction time
have to be used due to its slow transesterification and racemiza-
tion only occurring on the external surface of zeolite.
Table 6 presents the detailed textural properties of the two different
kinds of acidic catalysts. It is clear that, although the microporosity
of b-ZMS is slightly lower than that of C-b, the former possesses
much larger external surface area and mesopore volume than the
latter, which provides short micropore channels and open acid sites
for b-ZMS. Therefore, it can be proposed that the shorter diffusion
routes and more accessible active sites can promote the racemiza-
tion performance of b-ZMS. In contrast, for C-b, the less accessible
external active sites and long diffusion routes will cause a low con-
version rate or less selectivity (more side reactions).
It is very obvious that b-ZMS displays a clear-cut difference
from C-b in all the DKR processes (Table 5). In detail, for substrates
with a small molecular size, such as rac-1-PE and 1-phenyl-1-pro-
panol (rac-1-PP), the two catalysts possess similar conversion, but
the DKR processes catalyzed by b-ZMS show higher selectivity than
those driven by C-b. Moreover, the similar ees values after 1 h of
reaction indicate that the two zeolite catalysts possess the same
racemization rate in their DKRs. However, for the substrates with
a large molecular size, such as 1-(p-tolyl) ethanol, 1-(4-bromophe-
nyl)-ethanol, 1-indanol, and rac-1-NPE, the catalysts display al-
most the same selectivity but different conversions and yields.
That is, the DKR processes catalyzed by b-ZMS possess higher con-
version/yield than those catalyzed by C-b within the same reaction
time. For example, the yield in the DKR of rac-1-NPE decreases
from 90% to 58% when the racemization catalyst is C-b instead of
b-ZMS. In addition, the low ees values for the DKR catalyzed by
b-ZMS indicate a much higher racemization rate of b-ZMS than
C-b in these DKR processes of bulky molecules. These indicate that
the racemization plays a crucial factor in influencing the conver-
sion and selectivity of the one-pot DKRs.
To investigate the differences between b-ZMS and C-b in these
DKR reactions, the morphologies and structures of the two
catalysts, as well as their physical and chemical properties, are
characterized. The observation of SEM in Fig. 1 shows that
compared to C-b (Fig. 1A), b-ZMS (Fig. 1B) possesses more regular
morphology and smaller particle size. Their XRD patterns (Fig. 1C)
confirm their BEA frameworks. Moreover, their N2 sorption results
(Fig. 1D) show that b-ZMS provides a much larger hysteresis loop at
a high relative pressure (p/p0) different from C-b, corresponding to a
large mesoporosity with a diameter ranging from 15 to 80 nm
(Fig. 1D-c, inset, Barret–Joyner–Halenda model). Furthermore,
Furthermore, the amounts and strengths of acid sites of these
two catalysts on both total and external surface are measured by
nonaqueous potentiometric titration. For potentiometric titration
using n-butylamine, it is suggested as a criterion that the initial
electrode potential (Ei) indicates the maximum acid strength, and
the value of meq amine/g solid at the beginning of the plateau indi-
cates the total acid amount [41]. The acid strength of the sample
could be classified according to the following scale [41–44]: the
amount of n-butylamine titration at E > 100 mV is corresponding
to very strong sites, that at 0 < E < 100 mV to strong sites, that at
-100 < E < 0 mV to weak sites, and that at E < -100 mV to very weak
sites. Fig. 2A shows the titration curves of b-ZMS and C-b with n-
butylamine. The two catalysts display similar numbers of strong
acid sites (E > 0 mV), but b-ZMS possesses more weak acidic sites
than C-b, owing to the abundant structure defects brought about
by its small size. However, when n-butylamine is replaced with
2,6-di-tert-butylpyridine (2,6-DTBP), the two titration curves show
extraordinary differences (Fig. 2B). Although there is no criterion
for nonaqueous potentiometric titration of 2,6-DTBP as in the case
of n-butylamine, the number of acidic sites can be estimated from
the plateau and change trend of their titration curves. Because it
is difficult for 2,6-DTBP (kinetic diameter: 7.9 Å) to enter the
micropores of zeolite, it can only interact with the acidic sites on
the external surface. Its exhausted amount during the titration
thus shows the number of acidic sites on the external surface of
Fig. 1. SEM images of (A) C-b and (B) b-ZMS. XRD patterns (C) and N2 sorption isotherms (D) of C-b (a) and b-ZMS (b), as well as pore size distribution of b-ZMS (D-c). The
scale bars in the insets of (A) and (B) are 1 m.
l

30
X. Li et al. / Journal of Catalysis 288 (2012) 24–32
Table 6
The physical and textural properties of b-ZMS and C-b catalysts.
Zeolite
type
SiO2/
Al2O3
BET surface areaa
(m2/g)
Surface area of micropore
Surface area of external surface Pore volume of microporeb Pore volume of mesoporeb
(m2/g)
(m2/g)
(m3/g)
(m3/g)
C-b
b-ZMS
25
32
490
608
448
372
42
236
0.21
0.17
0.03
0.66
a
Micropore area and external surface area of b-ZMS calculated by t-plot method.
Micropore and mesopore volumes of b-ZMS obtained by t-plot method and desorption data using BJH model between 1.7 and 300 nm width, respectively.
b
Fig. 2. The total (A) and external (B) acidity measurements of b-ZMS (a) and C-b (b) by potentiometric titration in acetonitrile using n-butylamine (A) and 2,6-di-tert-
butylpyridine (B). The insets display the detailed total and external acidic properties of the two catalysts.
the catalysts. From Fig. 2B, the two catalysts possess the same sur-
face acidic strength, but the total number of external surface acid
sites of b-ZMS is at least five times more than that of C-b, which
is consistent with their external surface area ratio (Table 6). Obvi-
ously, it is the difference of numbers of acidic sites on the external
surface that determines the different DKR performances of the two
catalysts (Table 5). Therefore, it can be proposed that the small
substrate molecules can easily diffuse into the micropores of b-
zeolite, and so the two catalysts possess the same conversion ow-
ing to their similar numbers of strong acid sites. However, the
shorter diffusion route of b-ZMS can reduce the formation of by-
products or coking owing to short contact time between active site
and substrates, and so promote the selectivity of the DKR process.
For the large substrate molecules, diffusion into the micropores be-
comes slow and difficult. In this case, the abundant external sur-
face active sites of b-ZMS will greatly accelerate the racemization
of substrates and improve their conversion. Because it is difficult
for the substrates with large size to diffuse into micropores, or they
can diffuse only into a thin layer of the catalyst surface, the differ-
ence of selectivity driven by two catalysts will become negligible
compared to their conversion. In addition, it is notable that, except
for olefin, the by-products of diphenyl ether appear in the DKR
reaction of small substrates, but there is only olefin in the DKR pro-
cesses of large substrates. This can also be attributed to the long
and deep diffusion of small substrates and the difficult and thin dif-
fusion of large ones in the micropores of zeolites b.
S-1-PE, both C-b and b-ZMS give similar curves for the kinetic rate.
However, the selectivity of racemization catalyzed by b-ZMS grad-
ually becomes higher than that catalyzed by C-b, although the
racemization reactions catalyzed by them show the same selectiv-
ity in the first 10 min (Fig. 3A). The selectivity of racemization cat-
alyzed by b-ZMS reaches 85%, while that driven by C-b only is 73%
at 45 min. Because of shorter micropores and more pore opening of
zeolite nanocrystals in b-ZMSs, the substrate/product can easily
diffuse inside/outside the catalyst, decreasing the formation prob-
ability of by-product. However, for C-b, the long diffusion routes
will cause long contact between substrate and active sites and lead
to more by-products. Lower selectivity thus is observed with the
reaction time. Similarly to those of C-b and b-ZMS, the racemiza-
tion of S-1-PE catalyzed by b-ZMS-DTBP also shows higher selec-
tivity than that driven by C-b-DTBP, although both C-b-DTBP and
b-ZMS-DTBP display lower racemization rates compared to C-b
and b-ZMS because of the inhibition of acidic site on their surfaces
(Fig. 3C). Moreover, this advantage of b-ZMS-DTBP on selectivity
becomes more visible than that of b-ZMS (Fig. 3A). Because all
the substrates have to diffuse into micropores of b-ZMS-DTBP/C-
b-DTBP for racemization, the shorter microporous channel of b-
ZMS-DTBP will show more distinct superiority of selectivity. This
also indicates that the differences in racemization selectivity be-
tween C-b and b-ZMS can be assigned to the differences of diffu-
sion length of the substrates in them. For S-1-NPE, b-ZMS
presents a very high racemization rate compared to C-b (Fig. 3B).
For example, the ee of substrate catalyzed by b-ZMS decreases to
16% within 60 min, but that catalyzed by C-b just becomes 82%.
It is difficult for the S-1-NPE to diffuse into the micropores of
zeolite b, and the abundant accessible external acidic sites of b-
ZMS will accelerate the racemization of S-1-NPE on the external
surface of the catalyst. This racemization only occurring on the
external surface can also be confirmed that the racemizations of
S-1-NPE on b-ZMS-DTBP and C-b-DTBP are almost inhibited
Based on the above-mentioned extrapolation, the two sub-
strates with the smallest and largest size, that is, S-1-PE and
S-1-PNE, are employed to study the performance of two different
catalysts in racemization. At the same time, the other two catalysts
(C-b-DTBP and b-ZMS-DTBP), which are obtained by inhibiting the
acidic sites on the surfaces of C-b and b-ZMS with 2,6-DTBP, are
employed to further study the role of b-ZMS in the racemization
of small and bulky sec-alcohols. In detail, in the racemization of

X. Li et al. / Journal of Catalysis 288 (2012) 24–32
31
Fig. 3. The racemization of S-1-PE (A) and S-1-NPE (B) catalyzed by C-b and b-ZMS and the racemization of S-1-PE/S-1-NPE (C) catalyzed by C-b-DTBP and b-ZMS-DTBP
saturated with 2,6-DTBP in 2 mL cyclohexane containing 0.10 mmol substrates at 50 °C in a water bath under vigorous stirring. The amounts of all the catalysts are 4 mg in
the racemization of 1-PE and 8 mg in the racemization of S-1-NPE, respectively.
(Fig. 3B). These phenomena demonstrate again that the external
surface on nanozeolite is a crucial factor in the kinetic reaction
rates of bulk molecules, while the short diffusion channel influ-
ences the reaction selectivity of small molecules.
Finally, the reuses of both catalysts in one-pot DKR of rac-1-PE
is evaluated, and the results are listed in Fig. 4. After the first run,
the reaction solution was removed, and the catalysts were washed
three times with cyclohexane. Then, a fresh cyclohexane (2 mL)
solution containing 0.1 mmol rac-1-PE and 0.20 mmol VO was
added to repeat the above reaction process. As observed from
Fig. 4, b-ZMS shows better reaction stability than C-b, and it attains
a stable yield in the first three runs at least. The C-b after four runs
of DKRs loses its activity, while the yield in the fifth cycle catalyzed
by b-ZMS is still comparable with that in the first cycle catalyzed
by C-b. On one hand, the high reaction selectivity of b-ZMS avoids
the formation of by-products or coke that could block the microp-
ores, and so prolongs its lifetime. On the other hand, b-ZMS with
large external surface area possesses higher tolerance to carbon
Fig. 4. Reuse performance of the catalysts for DKR of rac-1-PE. Reaction conditions:
4 mg of b-ZMS (C-b) and 8 mg of NovozymÒ 435 were added into 2 mL cyclohexane
containing 0.10 mmol rac-1-PE and 0.20 mmol VO, and the reaction was carried out
for 1 h at 50 °C in a water bath under vigorous stirring.
deposits, and this will further improve its reaction stability.
4. Conclusions
(Fig. 3C). Moreover, there are only surface reactions on the cata-
lysts, and no diffusion process of substrate/product is needed;
the two catalysts thus show similar selectivity of racemization
b-ZMS with a hierarchical structure is proved to be a highly
effective and selective racemization catalyst in the one-pot DKR
of aromatic sec-alcohols, associated with transesterification

32
X. Li et al. / Journal of Catalysis 288 (2012) 24–32
catalyst NovozymÒ 435. Many other parameters such as catalyst
amounts, their relative ratios, and different strategies of their
manipulation are investigated to reach the best match of the two
catalysts, for which the one-pot DKR of aromatic sec-alcohol could
be carried out with a high reaction rate and selectivity. Thanks to
its more externally accessible acid sites and short microporous
channel, b-ZMS displays a higher racemization rate for the large
substrates but a higher selectivity of products for the small sub-
strates in their racemization, compared to C-b. This will favor the
performance of acidic zeolite catalysts in one-pot DKR systems
and also provide a new view of the application of nanozeolite in
catalysis.
[13] A.G. Dong, Y.J. Wang, Y. Tang, Y.H. Zhang, N. Ren, Z. Gao, Adv. Mater. 14 (2002)
1506.
[14] L.M. Huang, Z.B. Wang, J.Y. Sun, L. Miao, Q.Z. Li, Y.S. Yan, D.Y. Zhao, J. Am. Chem.
Soc. 122 (2000) 3530.
[15] B.J. Zhang, S.A. Davis, S. Mann, Chem. Mater. 14 (2002) 1369.
[16] M. Breuer, K. Ditrich, T. Habicher, B. Hauer, M. Kebeler, R. Stürmer, T. Zelinski,
Angew. Chem. Int. Ed. 43 (2004) 788.
[17] M. Heitbaum, F. Glorius, I. Escher, Angew. Chem. Int. Ed. 45 (2006) 4732.
[18] H. Pellissier, Tetrahedron 63 (2007) 9267.
[19] F.F. Huerta, A.B.E. Minindis, J.-E. Bäckvall, Chem. Soc. Rev. 30 (2001) 321.
[20] H. Pellissier, Tetrahedron 59 (2003) 8291.
[21] H. Pellissier, Tetrahedron 64 (2008) 1563.
[22] H. Pellissier, Tetrahedron 67 (2011) 3769.
[23] Y. Ahn, S.-B. Ko, M.-J. Kim, J. Park, Coord. Chem. Rev. 252 (2008) 647.
[24] A. Ghanem, Tetrahedron 63 (2007) 1721.
[25] A. Parvulescu, J. Janssens, J. Vanderleyden, D. De Vos, Top. Catal. 53 (2010) 931.
[26] J.H. Lee, K. Han, M.-J. Kim, J. Park, Eur. J. Org. Chem. 6 (2010) 999.
[27] H. Pellissier, Adv. Synth. Catal. 353 (2011) 659.
[28] S.Y. Kalliney, M.V. Ruggeri, WO Patent 08196, 1991.
[29] D.W. House, US Patent 5476964, 1995.
Acknowledgments
[30] S. Wuyts, K. De Temmerman, D. De Vos, P. Jacobs, Chem. Commun. 2003
(1928).
[31] S. Wuyts, K. De Temmerman, D.E. De Vos, P.A. Jacobs, Chem. Eur. J. 11 (2005)
386.
[32] Y.Z. Zhu, K.-L. Fow, G.-K. Chuah, S. Jaenicke, Chem. Eur. J. 13 (2007) 541.
[33] A.F. Fois, A. Yap, A.F. Masters, Th. Maschmeyer, Top. Catal. 48 (2008) 153.
[34] P. Lozano, T. De Diego, M. Larnicol, M. Vaultier, J.L. Iborra, Biotechnol. Lett. 28
(2006) 1559.
[35] P. Lozano, T. De Diego, C. Mira, K. Montague, M.V. Vaultierb, J.L. Iborra, Green
Chem. 11 (2009) 538.
[36] K. Benaissi, M. Poliakoff, N.R. Thomas, Green Chem. 11 (2009) 617.
[37] L. Costa, A. Coelho, F. Lemos, F.R. Ribeiro, J.M.S. Cabral, Appl. Catal. A Gen. 354
(2009) 33.
[38] D. Klomp, T. Maschmeyer, U. Hanefeld, J.A. Peters Chem, Eur. J. 10 (2004) 2088.
[39] Y.J. Kang, W. Shan, J.Y. Wu, Y.H. Zhang, X.Y. Wang, W.L. Yang, Y. Tang, Chem.
Mater. 18 (2006) 1861.
[40] Y. Shi, X. Li, J.K. Hu, J.H. Lu, Y.C. Ma, Y.H. Zhang, Y. Tang, J. Mater. Chem. 21
(2011) 16223.
[41] K.N. Rao, K.M. Reddy, N. Lingaiah, I. Suryana-rayana, P.S. Sai Prasad, Appl. Catal.
A Gen. 300 (2006) 139.
This work was supported by NSFC (20721063, 30828010,
20873025, and 20890122), STCSM (10QH1400300, 08DZ2270500,
and 09DZ2271500), and 973 and 863 programs (2009CB930400,
2009CB623506, and 2009AA033701).
References
[1] D.R. Rolison, Chem. Rev. 90 (1990) 867.
[2] M.E. Davis, R.F. Lobo, Chem. Mater. 4 (1992) 756.
[3] M.E. Davis, Nature 417 (2002) 813.
[4] C.S. Cundy, P.A. Cox, Chem. Rev. 103 (2003) 663.
[5] A. Stein, Adv. Mater. 15 (2003) 763.
[6] F. Xu, Y.J. Wang, X.D. Wang, Y.H. Zhang, Y. Tang, P.Y. Yang, Adv. Mater. 15
(2003) 1751.
[7] F. Schüth, W. Schmidt, Adv. Mater. 14 (2002) 629.
[8] M. Smaihi, E. Gavilan, J.-O. Durand, V.P.J. Valtchev, Mater. Chem. 14 (2004)
1347.
[9] L. Tosheva, V.P. Valtchev, Chem. Mater. 17 (2005) 2494.
[10] V. Valtchev, S. Mintova, Micropor. Mesopor. Mater. 43 (2001) 41.
[11] X.D. Wang, W.L. Yang, Y. Tang, Y.J. Wang, S.K. Fu, Z. Gao, Chem. Commun. 21
(2000) 2161.
[12] K.H. Rhodes, S.A. Davis, F. Caruso, B.J. Zhang, S. Mann, Chem. Mater. 12 (2000)
2832.
[42] A.S. Khder, E.A. EI-Sharkawy, S.A. El-Hakam, A.I. Ahmed, Catal. Commun. 9
(2008) 769.
[43] A.S. Khder, A.I. Ahmed, Appl. Catal. A Gen. 354 (2009) 153.
[44] E.A. EI-Sharkawy, A.S. Khder, A.I. Ahmed, Micropor. Mesopor. Mater. 102
(2007) 128.