Microporous and Mesoporous Materials 247 (2017) 1e8

Contents lists available at ScienceDirect

Microporous and Mesoporous Materials

journal homepage: www.elsevier.com/locate/micromeso

Mesoporous silica-based nanotubes loaded Pd nanoparticles: Effect of

framework compositions on the performance in heterogeneous
catalysis
Jing Sun, Hua Wang, Xue Gao, Xinli Zhu, Qingfeng Ge, Xiao Liu*, Jinyu Han
Key Laboratory for Green Chemical Technology of Ministry of Education, School of Chemical Engineering and Technology, Tianjin University, Tianjin 300072,
China

a r t i c l e i n f o a b s t r a c t

Article history: Well-shaped mesoporous silica-based nanotubes with different framework compositions, such as
Received 9 October 2016 organic groups (ethylene, phenylene), carbon/silica hybrids, could be controllably synthesized with the
Received in revised form inner diameter of less than 10 nm and the surface areas of about 400e900 m2 g
1. Through an
22 March 2017
impregnation-reduction process, palladium (Pd) nanoparticles have been uniformly dispersed in the
Accepted 23 March 2017
Available online 25 March 2017
channels of these different nanotubes, which were confirmed by electron microscopy analysis. The
catalytic performance of these silica-based nanotubes loaded Pd nanoparticles was evaluated by the
aerobic oxidation of benzyl alcohol and enantioselective hydrogenation of a,b-unsaturated carboxylic
Keywords:
Organosilica nanotubes
acid, respectively. Due to one-dimensional nanotube structures and tunable hydrophilic/hydrophobic
Carbon/silica hybrid properties, the organosilica nanotubes supported Pd nanoparticles could afford >99% conversion of
Pd nanoparticles benzyl alcohol and 89% selectivity of benzaldehyde within 2 h. Importantly, carbon/silica hybrid nano-
Benzyl alcohol oxidation tubes may have a positive effect on the reaction, which gave about 95% selectivity of benzaldehyde. The
Asymmetric hydrogenation catalysts could be reused without an obvious decrease in both conversion and selectivity. Furthermore,
the silica-based nanotubes supported Pd nanoparticles were also efficient for asymmetric hydrogenation
of a,b-unsaturated carboxylic acids, which showed a highest conversion of 99% with 48%
enantioselectivity.
© 2017 Elsevier Inc. All rights reserved.

1. Introduction Recently, organosilica nanotubes have demonstrated their po-

One-dimensional nanotubes perform an important role as the
heterogeneous supports/catalysts owing to their unique physico-
chemical properties, such as high specific areas, mechanical sta-
bilities and hollow structures facilitating easy diffusion of reactants
and products [1e4]. Up to date, carbon, silica, metal oxide, or even
metal with tubular morphology have been reported [5e8]. Orga-
nosilica nanotubes represent one of the most interesting materials
for catalysis because they not only possess the ability for the
adjustment of the hydrophobic/hydrophilic properties, but also can
be further functionalized to derive new active sites easily. More-
over, organosilica nanotubes with bridging organic groups such as
ethylene and phenylene could be facilely synthesized by using soft-
template assembly method [9].
* Corresponding author.
E-mail address: liuxiao71@tju.edu.cn (X. Liu).
tential in catalysis. For example, Guo et al. reported that ethylene-
or phenylene- organosilica nanotubes doped with sulfonic acids in
the framework, which could efficiently catalyze the esterification of
palmitic acid and transesterification of yellow horn seed oil with
methanol [10]. Organosilica nanotubes containing bipyridine
groups in their walls were found to act as efficient solid ligands to
immobilize molecular catalyst for water oxidation reaction [11].
Palladium (Pd) nanoparticles dispersed on various supports
have been extensively studied not only for the oxidation reactions
of fine chemicals but also for the hydrogenation reactions of olefins
[12e14]. Recently, carbon and pure silica nanotubes were used to
encapsulate Pd nanoparticles for oxidation of benzyl alcohol and
Suzuki-Miyaura coupling, respectively [15,16]. The heterogeneous
Pd catalysts immobilized on nanotubes showed favorable activities
due to the nano-confinement effects and higher accessibility of
reactants. However, the effects of the nanotube framework com-
positions on the performance in the heterogeneous catalysis have
been seldom systematically studied.

http://dx.doi.org/10.1016/j.micromeso.2017.03.040
1387-1811/© 2017 Elsevier Inc. All rights reserved.
2 J. Sun et al. / Microporous and Mesoporous Materials 247 (2017) 1e8
Herein, by using silica-based nanotubes with the doping of
ethylene, phenylene or carbon in the frameworks as the supports,
we have constructed the heterogeneous catalysts containing Pd
nanoparticles through the impregnation-reduction method. Pd
nanoparticles are uniformly dispersed in the channels of the
nanotubes with different framework compositions, which were
applied in the oxidation of benzyl alcohol and enantioselective
hydrogenation of a,b-unsaturated carboxylic acid, respectively
(Scheme 1). The effects of the silica-based nanotubes with ethylene,
phenylene, carbon or pure silicas in the frameworks on the catalytic
performance were investigated.
2. Experimental section
2.1. Materials
1,2-bis(trimethoxysilyl)ethane (BTME), 1,4-bis(triethoxysilyl)
benzene (BTEB), tetraethyl orthosilicate (TEOS) and potassium
chloride were purchased from J&K Scientific Ltd. Hydrochloric acid
(37%), ethanol, 1,4-dioxane, toluene and acetone were obtained
from Real&Lead Chemical Co. LTD. PdCl2 (>99.9%), Pluronic P123, a-
phenylcinnamic acid (99%) and cinchonidine (CD) were received
from SigmaeAldrich and all reagents were used without further
purification. Oxygen and hydrogen with purity of 99.999% were
used in the oxidation and asymmetric hydrogenation, respectively.
2.2. Synthesis of silica-based nanotubes with different framework
compositions
The organosilica nanotubes with ethylene or phenylene groups
in the framework were synthesized according to the method we
reported previously with minor modifications [9]. Typically, 0.55 g
of P123 and 1.75 g of KCl were completely dissolved in 135 mL of
HCl solution (2 mol L
1) at 38  C. After that, 3.50 mmol of 1,2-
bis(trimethoxysilyl)ethane (BTME) was added into the solution
under vigorous stirring. The mixture continually stirred for 6 min
and remained quiescent for 24 h. Then the resultant mixture was
transferred to a PTFE reactor and kept at 100  C for an additional
24 h. Subsequently, the solid product was obtained by the filtration
and air drying overnight at room temperature. Finally, 200 mL of
ethanol containing 1.5 g of concentrated aqueous HCl solution was
used to extract the surfactants from 1.0 g of synthesized material at
60  C for 24 h and the surfactants-free samples were denoted as E-
SNT. The synthetic process of organosilica nanotubes containing
phenylene groups in the walls (named as B-SNT) was almost the
same as that of E-SNT except for the replacement of the precursor
with 1,4-bis(triethoxysilyl)benzene (BTEB).
Scheme 1. Illustration of Pd supported ethylene, phenylene bridged or carbon doped organosilica nanotubes and common reaction process.
To obtain the carbon-doped silica nanotubes E-CS-NT and B-CS-
NT, the surfactants-free E-SNT and B-SNT were carbonized at 800  C
in nitrogen atmosphere for 2 h, respectively (CS refers to carbon/
silica hybrid). Besides, the pure silica nanotubes (SNT) and SBA-15
were also synthesized as the references of the supports [17,18].
2.3. Preparation of immobilized Pd nanoparticles using nanotubes
as the supports
All of the catalysts supported Pd nanoparticles were prepared by
using an impregnation-reduction process. For example, 150 mg of
E-SNT were immersed into 10 mL ethanol solution containing
13 mg PdCl2 while the content of Pd immobilized was 5 wt%. After
ultrasonic treatment for 1 h, the suspension was stirred at an
ambient temperature for 12 h. Then, the solvents were evaporated
under a negative pressure slowly and the divalent Pd was reduced
at 150  C in H2 for 5 h to obtain the heterogeneous catalysts with
the immobilized Pd nanoparticles.
2.4. Oxidation of benzyl alcohol
The oxidation of benzyl alcohol was carried out in a 25 mL
round-bottom three-necked flask with a magnetic stirring and a
condenser. Since toluene may be a product of the oxidation reac-
tion, we selected xylene as the solvent and undecane as an internal
standard for the product analysis [19,20]. The powdery catalyst
(typically 45 mg) was added into the flask charged with 0.6 g benzyl
alcohol and 12.5 mL xylene as well as relevant amount of undecane.
When the temperature raised to the set value, an O2 flow was
bubbled into the mixture to trigger the reaction under vigorous
stirring. Samples of the reaction mixture were taken out at
appropriate time intervals and analyzed by gas chromatograph
(Bruker BR-GC-456) equipped with a flame ionization detector
(FID) after solid removal.
2.5. Enantioselective hydrogenation of a,b-unsaturated carboxylic
acid
Hydrogenation was carried out in a magnetically stirred auto-
clave under 0.1 MPa H2 pressure at room temperature. The catalyst
(20 mg) was pre-treated in a H2 flow with modifier CD (0.02 mmol)
in 4 mL 1,4-dioxane containing 2.5% (v/v) of water for 30 min. After
that, 1 mmol of a-phenylcinnamic acid in 2 mL solvent was added
to the mixture. The mixture was stirred immediately after adjusting
the hydrogen pressure to 0.1 MPa. After 10 h, the mixture was
neutralized with a diluted HCl solution, extracted with ethyl ether
and analyzed by proton nuclear magnetic resonance (1H NMR,
 J. Sun et al. / Microporous and Mesoporous Materials 247 (2017) 1e8
Bruker DRX400 MHz type spectrometer) for conversion and high
performance liquid chromatography (HPLC, 6890 Agilent Co.)
equipped with a chiral column (Daicel Chiralcel OJ-H) for enan-
tiomeric excess (ee). The ee (%) of the S isomers was expressed
according to ee (%) ¼ ([S]
[R])  100/([S] þ [R]).
2.6. Characterization
The transmission electron microscopy (TEM) images were taken
on a FEI Tecnai G2 F20 electron microscope operated at an accel-
eration voltage of 200 kV. The nitrogen adsorption-desorption
isotherms were measured on Tristar 3000 (Micrometrics) after
the samples were outgassed at 120  C for 3 h. The Brunauer-
Emmett-Teller (BET) specific surface areas were calculated using
adsorption data at the relative pressure range of P/P0 ¼ 0.05e0.25.
Pore size distributions were calculated from adsorption branch
using the Barrett-Joyner-Halenda (BJH) method. The X-ray diffrac-
tion (XRD) patterns were recorded on a Rigaku D/MAX 2500
powder diffraction system using Cu-Ka radiation of 0.15406 nm
wavelength over the 2q range of 15e90 with a scan speed of 5 /
min at room temperature. Raman scattered spectrum was taken on
Thermo Electron Corporation DXR microscope with 532 nm exci-
tation wavelength. Thermogravimetry analysis (TGA) was
measured by Shimadzu TA-50 with a heating rate 10  C/min in air.
X-ray photoelectron spectroscopy (XPS) was acquired using a
PHI1600 X-ray photoelectron spectroscopy (Perkin Elmer).
3. Results and discussion
3.1. Morphology and structure characterization of silica-based
nanotubes loaded Pd nanoparticles
3.1.1. Structure and composition information of nanotubes with
different framework compositions
The TEM images of E-SNT, B-SNT in Fig. 1a and b clearly show the
nanotube structure with inner diameters of less than 10 nm and
lengths of several hundred nanometers. Different types of meso-
porous materials could be synthesized through the soft-template
assembly using P123 as surfactants in acidic solutions. The
Fig. 1. TEM images of (a) E-SNT, (b) B-SNT, (c) E-CS-NT, (d) B-CS-NT, (e) SNT and (f) SBA-15.
3
micelle shapes, such as spherical, rodlike and cylindrical hexago-
nally packed may be formed under different conditions. In our
system, the P123 concentration was much lower than that for the
synthesis of SBA-15, which was beneficial to rodlike micelle for-
mation. Therefore, the nanotube structures are obtained instead of
SBA-15 under diluted P123 solutions. After calcination of E-SNT and
B-SNT in N2, the nanotube morphology was maintained although
narrower diameters were observed and the nanotube lengths
became shorter (Fig. 1c and d). The pure silica nanotubes (SNT) and
SBA-15 were also synthesized and confirmed by TEM analysis in
Fig. 1e and f. All the nitrogen adsorption-desorption isotherms
present typical type IV isotherms, in which two hysteresis loops
indicate the establishment of mesoporous structure and stacking
pores by nanotubes, respectively, displayed at medium and high
relative pressure (Fig. S1). The Brunauer-Emmett-Teller (BET) sur-
face area and pore size determined by Barrett-Joyner-Halenda (BJH)
method with adsorption branch were showed in Table S1. It was
found that the BET surface areas of materials without calcination
were approximately the same (870 m2 g
1 and 910 m2 g
1 for E-
SNT and B-SNT, respectively). However, they reduced to about
400 m2 g
1 after calcination. Likewise, a notable decrease in pore
size from E-SNT and B-SNT to the corresponding E-CS-NT and B-CS-
NT (from 7.3 nm and 6.0 nm to 4.5 nm and 4.0 nm, respectively) was
observed because of the shrinkage of nanotubes during carbon-
ization. The BET surface area and pore size of SNT and SBA-15 were
also calculated from the nitrogen adsorption-desorption experi-
ments (820 m2 g
1 and 9.4 nm for SNT, while 830 m2 g
1 and
6.8 nm for SBA-15).
The mesoporous hybrids with carbon-silica composites could be
obtained through the calcination of mesoporous organosilicas with
phenylene in the walls [21]. Herein, organosilica nanotubes with
ethylene or phenylene in the walls were calcined in N2 to obtain
carbon-silica nanotubes. Raman spectra were recorded to testify
the presence of carbon in E-CS-NT and B-CS-NT showed in Fig. S2a.
The signals of the G peak at 1602 cm
1 and D peak at 1342 cm
1
indicate the formation of carbon after a high temperature treat-
ment, although there is a minor difference in the relative intensity
(IG:ID) caused by the various crystalline and amorphous carbon
ratios [22]. In order to quantify the actual quantity of carbon in
4 J. Sun et al. / Microporous and Mesoporous Materials 247 (2017) 1e8
nanotubes, a certain amount of samples were temperature pro-
grammed in oxygen atmosphere and two independent weight loss
curves were obtained in Fig. S2b. Regardless of the loss caused by
desorption of water and alcohol under 100  C, approximately 10%
and 20% of the mass were calculated by the sharp decline ranging
from 520  C to 700  C, indicating the carbon contents in E-CS-NT
and B-CS-NT are 10% and 20%, respectively.
3.1.2. Pd nanoparticles encapsulated in the channels of different
nanotubes
Through the impregnation-reduction method, the silica-based
materials loaded Pd nanoparticles were obtained and character-
ized by TEM showed in Fig. S3. Spherical Pd nanoparticles are
mainly dispersed uniformly in the inner channels of B-SNT, E-SNT,
E-CS-NT and B-CS-NT, with the mean size of around 2e4 nm. On the
other hand, relatively larger Pd nanoparticles with the average size
of about 5 nm were obtained for SNT and SBA-15 as the supports.
The results that the sizes of Pd nanoparticles anchored on orga-
nosilica or carbon-silica nanotubes are smaller than those on pure
silica nanotubes and SBA-15 indicate that the existence of organic
groups or carbon in the walls has a positive effect on the formation
of smaller metal nanoparticles. In addition, we also loaded Pd
nanoparticles on the amorphous silica by using the same method
and the mean size of Pd nanoparticles is over 10 nm (Fig. S4). This
result demonstrates that the nanotube structure has the confine-
ment effect on the Pd agglomeration due to the location of metal
particles inside the nanotubes.
As for the nitrogen adsorption-desorption isotherms (Fig. 2a),
similar results showing that type IV isotherms with two hysteresis
loops were obtained after the encapsulation of Pd nanoparticles in
the nanotubes, which indicates that the process of impregnation
and reduction did not destroy the nanotube structure. However, the
BET surface areas decreased by an average of about 100 m2 g
1
while the pore size also showed varying-degree decline to 4e7 nm
for the nanotubes loaded Pd nanoparticles (Table 1). The XRD
patterns of various Pd-supported nanotubes were shown in Fig. 2b.
Four distinct peaks at 2q ¼ 40.0 , 46.3 , 68.2 and 82.0 corre-
sponding to the (111), (200), (220) and (311) reflections of Pd metal,
respectively, could be observed for all samples. The average sizes of
nanoparticles supported on nanotubes calculated by Scherrer's
equation with a spherical model for approximation are listed in
Table 1, which was matched well with TEM characterization. In
addition, the XPS characterization was carried out to know the Pd
valence states within different catalysts and the spectra were
showed in Fig. S5. All the catalysts possess both metallic and
oxidized states simultaneously with the binding energy of Pd0 and
Fig. 2. (a) Nitrogen adsorption-desorption isotherms and (b) XRD patterns of the various samples: Pd/E-SNT (black solid square), Pd/B-SNT (red solid round), Pd/E-CS-NT (blue solid
triangle), Pd/B-CS-NT (green hollow round), Pd/SNT (pink solid inverted triangle) and Pd/SBA-15 (cyan hollow square). (For interpretation of the references to colour in this figure
legend, the reader is referred to the web version of this article.)
Pd2þ observed at 335.5 eV and 336.6 eV, respectively, corre-
sponding to Pd 3d5/2 orbit [23]. According to the spectra fitting and
peak area calculation, the Pd0 contents were in the range of 50e62%
summarized in Table 1.
3.2. Heterogeneous catalytic performance
3.2.1. Oxidation of benzyl alcohol
The selective oxidation of alcohols to corresponding carbonyl
aldehydes is one of the most predominant transformation reactions
especially for the production of benzaldehydes, which is widely
used in pharmaceuticals, perfumes and foods (Scheme 2a) [24e26].
Pd nanoparticles dispersed on various supports have been proved
to be efficient heterogeneous catalysts for the oxidation reaction of
alcohols [27e29]. At the same time, benzyl alcohol is the most
common used substrates in the laboratory-scale reaction to eval-
uate the performance of catalysts [30,31].
3.2.1.1. Effect of different framework compositions. The effects of the
nanotube structure and constitution of the supports on the benzyl
alcohol oxidation reaction were investigated by using atmospheric
pressure of O2 over Pd/E-SNT, Pd/B-SNT, Pd/SNT and Pd/SBA-15,
respectively. Among these, catalysts with nanotube structure
could give a higher conversion (98e99%) of benzyl alcohol within
2 h than Pd/SBA-15 (92%) (Fig. S6, Table 2). Fig. 3a shows the re-
action kinetics of various Pd nanoparticles loaded on silica-based
nanotubes, which was evaluated by the yield of benzaldehyde.
Within the first hour, all the reactions proceeded smoothly. After
the peak of the maximum yield of benzaldehyde, a slight drop could
be observed, which may be caused by the further oxidation of
benzaldehydes to benzoic acids [24,32]. The maximum of benzal-
dehyde yield was about 87% catalyzed by Pd/E-SNT within 2 h,
which was higher than that over pure silica nanotubes (77%) and
SBA-15 (70%). These results implied that the nanotube structures
and compositions of the supports could influence the reaction ac-
tivity notably. Owing to the nanotube structure which would make
more active sites exposed, nanotube-based catalysts have the ad-
vantages to make the reactants approach to the active sites much
more easily compared to the hexagonally structured mesoporous
SBA-15. On the other hand, the introduction of ethylene or phe-
nylene in the frameworks makes the supports possess more hy-
drophobic property than that of pure silica nanotubes, which could
be beneficial for benzyl alcohol to diffuse to the active sites easily.
Furthermore, taking the effect of by-product water into consider-
ation, the gradual increasing amount of water adsorbed on the
surface of pure silicas may prohibit both of the access of reactants
 J. Sun et al. / Microporous and Mesoporous Materials 247 (2017) 1e8 5

Table 1
Physicochemical properties of various Pd-supported catalysts.

Catalyst BET surface area (m2 g
1)a Pore size (nm)b Pd cluster size from TEM (nm)c Pd cluster size from XRD (nm)d Pd0 (%)e

Pd/E-SNT 830 6.0 4.0 4.3 62

Pd/B-SNT 690 6.1 2.0 2.7 50
Pd/E-CS-NT 330 4.2 3.1 4.2 50
Pd/B-CS-NT 310 4.0 3.5 3.9 58
Pd/SNT 730 7.5 4.9 4.6 60
Pd/SBA-15 640 6.0 5.6 5.3 56
a
Surface area was determined using the Brunauer-Emmett-Teller (BET) model.
b
Pore size was estimated using Barrett-Joyner-Halenda (BJH) method with adsorption branch.
c
Pd cluster size was measured through the TEM images counting about 300 nanoparticles.
d
Pd cluster size was calculated by Scherrer's equation from XRD.
e
Pd0 content was determined by XPS characterization.

(a)

(b)

Scheme 2. Reaction equations for (a) the oxidation of benzyl alcohol and (b) enantioselective hydrogenation of a,b-unsaturated carboxylic acid.

Table 2
Catalytic performance of various Pd supported catalysts towards oxidation of benzyl alcohol.a

Catalyst Conversion of benzyl alcohol (%)b Maximum yield of benzaldehyde (%) TOF/minc TONd

Pd/E-SNT 98 87 16 228
Pd/B-SNT 98 85 16 224
Pd/E-CS-NT 99 87 15 228
Pd/B-CS-NT 99 94 15 246
Pd/SNT 99 77 16 202
Pd/SBA-15 92 70 9 184
Pd/CMK-3e 55 44 4 192
a
Reaction conditions: 45 mg catalyst containing 5 wt% Pd nanoparticles; 0.6 g benzyl alcohol; 100  C; O2 pressure, 0.1 MPa.
b
The conversion was obtained within two reaction hours.
c
TOF was calculated from the initial reaction rate by using the conversion of benzyl alcohol.
d
TON ¼ the maximum yield of benzaldehyde/the amount of catalyst.
e
Reaction was carried out at 80  C and data were collected within 1 h from Ref. [33].

and desorption of products. For organosilicas, the by-product of
water could be easily desorbed from the surface, which ensures the
reaction higher activity. It should also be noted that the turnover
frequency (TOF) of Pd/E-SNT and Pd/B-SNT was about 16 min
1
(Table 2), 4 times higher than that of Pd/CMK-3 [33].
3.2.1.2. Effect of carbon doped in silica nanotubes. In order to further
investigate the effects of framework compositions on the reaction
activities, two kinds of carbon-doped silica nanotubes E-CS-NT and
B-CS-NT with different carbon containings were used as the sup-
ports for the selective oxidation of benzyl alcohol. From the results
presented in Table 2 and Fig. 3b, the conversion of benzyl alcohol
was as high as 99% within 2 h and the yield of benzaldehyde
increased smoothly with the reaction time. More importantly, the
benzaldehyde yield over Pd/B-CS-NT approached to the maximum
of 94% while Pd/E-CS-NT showed the maximum of 87%, which was
higher than that of Pd/SNT (77%). It has been reported that for-
mation of alkoxy intermediates can help the improvement of
benzaldehyde selectivity [20]. The alkoxy intermediates may be
easier to be generated due to the faster electron transfer between
the reactants and the carbon-doped catalyst, which contributed to
the higher benzaldehydes production. Identically, a decline after
3 h could be observed, caused by the further oxidation of
benzaldehydes.
3.2.1.3. Effect of reaction temperature. Since temperature is a crucial
factor for oxidation of benzyl alcohol, we also investigated the ef-
fect of the reaction temperature by using Pd/E-SNT and the results
6 J. Sun et al. / Microporous and Mesoporous Materials 247 (2017) 1e8

Fig. 3. Catalytic activity of (a) Pd/E-SNT (black solid square), Pd/B-SNT (red solid round), Pd/SNT (blue solid triangle), Pd/SBA-15 (pink solid inverted triangle) and (b) Pd/E-CS-NT
(black solid square), Pd/B-CS-NT (red solid round) towards benzaldehyde yield in the oxidation of benzyl alcohol. Reaction conditions: 45 mg catalyst containing 5 wt% Pd
nanoparticles; 0.6 g benzyl alcohol; 100  C; O2 pressure, 0.1 MPa. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this
article.)

were shown in Fig. S7. When the reaction temperature was set to
60  C, the conversion of benzyl alcohol was only about 10% within
5 h. Almost no benzaldehyde was detected. After elevating tem-
perature to 80  C, we found a dramatic outcome with 67% con-
version of benzyl alcohol and 81% selectivity of benzaldehyde
within 1 h. The conversion was gradually increased to 95% after 5 h.
Continually heated to 100  C, the reaction was almost finished
within 1 h reaching to nearly 100% conversion of benzyl alcohol and
87% selectivity of benzaldehyde. With increasing reaction time, the
selectivity gradually decreased to 81%, which may be caused by the
consumption of benzaldehydes to benzoic acids.
3.2.1.4. Reusability of the catalysts. The stability of the catalysts is of
importance concerning the applications in industry [34,35]. Here,
the stability of nanotube-based catalysts was investigated by using
Pd/E-SNT at 100  C under atmospheric pressure of O2. After the
reaction was completed and cooled to room temperature, the solid
catalyst was recovered by centrifugation and washed thoroughly
before vacuum-dried at 80  C. The recovered Pd/E-SNT was used for
another new reaction and the results were shown in Fig. 4a. After 4
times of consecutive reuse, the catalyst still exhibited excellent
conversion (>99%) with the selectivity of benzaldehyde of nearly
90%. After the fifth reaction, Pd/E-SNT was characterized by TEM
and the nanotube structure was kept from Fig. 4b. Moreover, the
distribution of Pd nanoparticles in the E-SNT was uniformly
dispersed and the average size has almost no change, which
benefited from the nano-confinement of Pd nanoparticles in the
tubes and avoiding the agglomeration of catalyst during the reac-
tion. In addition, the XPS characterization of Pd/E-SNT after reac-
tion was carried out to investigate the Pd valence state and the
results were shown in Fig. 5. Before reaction, the amount of metallic
Pd in Pd/E-SNT was about 62% of the total Pd species. After the fifth
reaction, no remarkable change (58% metallic Pd) could be found,
which indicated that Pd nanoparticles are unlikely to be oxidized
during the oxidation process.
3.2.2. Enantioselective hydrogenation of a,b-unsaturated carboxylic
acid
Chiral carboxylic acid is a key intermediate to produce phar-
maceuticals, fragrances, and flavors. Among the methods to obtain
chiral carboxylic acids, asymmetric hydrogenation of a,b-unsatu-
rated carboxylic acids is one of the most convenient approaches
[36]. The alkaloid-modified Pd nanoparticles supported on TiO2,
Al2O3, active carbon and carbon nanotubes [37e40] are the most
used and studied solid catalysts for this reaction. With the silica-
based nanotubes loaded Pd nanoparticles possessing different
framework compositions in hand, we investigated the performance
of asymmetric hydrogenation of a,b-unsaturated carboxylic acids
(Scheme 2b).
As shown in Table 3, the a-phenylcinnamic acid conversion
reached to 99% within 10 h at room temperature and H2 pressure of
0.1 MPa when Pd was encapsulated in the B-SNT, E-SNT and E-CS-

Fig. 4. (a) The reusability of Pd/E-SNT catalyst in the oxidation of benzyl alcohol; (b) TEM image of Pd/E-SNT catalyst after five recycles of catalysts. Reaction conditions: 45 mg
catalyst containing 5 wt% Pd nanoparticles; 0.6 g benzyl alcohol; 100  C; O2 pressure, 0.1 MPa; 2 h.
 J. Sun et al. / Microporous and Mesoporous Materials 247 (2017) 1e8
Fig. 5. XPS spectra of Pd/E-SNT (a) before and (b) after reaction.
Table 3
Catalytic performance of various Pd supported catalysts towards enantioselective
hydrogenation of a,b-unsaturated carboxylic acid.a
Entry Catalysts Conversion of a,b-unsaturated carboxylic acid (%)
1 Pd/E-SNT >99
2 Pd/B-SNT >99
3 Pd/E-CS-NT >99
4 Pd/SNT 46
5 Pd/CNTb e 51
a
Unless otherwise stated, the Pd loading is 5 wt%. Reaction conditions: H2
pressure, 0.1 MPa; room temperature; 10 h; 20 mg catalyst, 37.2 mg a-phenyl-
cinnamic acid, 6 mg CD, 6 mL 1,4-Dioxaneþwater (39:1).
b
Data reprinted from previous work by Guan [40]. CNT refers to carbon
nanotubes.
NT, respectively. However, within the same reaction time, pure
silica nanotube SNT supported Pd only gave the a-phenylcinnamic
acid conversion of 46%. It was reported that the fast adsorption of
chiral modifiers and reactants onto the supports is the main reason
for the high activity and enantioselectivity [40]. For the silica
nanotubes with ethylene, phenylene or carbon in the framework,
the hydrophobicity may help the preferential adsorption of PCA
and CD to the Pd active centre, which enhanced the reaction ac-
tivity. On the contrary, pure silica nanotubes with the hydrophilic
property did not favor the fast diffusion of the organic reactants and
showed the lowest activity.
For the enantioselectivity (ee), Pd/B-SNT could exhibit the
highest value of 48%, higher than other three catalysts Pd/E-SNT
(40%), Pd/E-CS-NT (30%) and Pd/SNT (33%). One possibility for the
higher enantioselectivity of Pd/B-SNT is that the phenylene groups
in the nanotube framework have stronger p-p interaction with the
chiral modifier cinchonidine, which may induce the favorable steric
configuration for the asymmetric reaction. The ee obtained by Pd/
B-SNT was almost similar with that by Pd nanoparticles sup-
ported on carbon nanotubes (51%) under the same reaction
conditions.
4. Conclusions
Well-shaped mesoporous silica-based nanotubes with different
framework compositions including ethylene, phenylene and car-
bon/silica hybrid have been used to encapsulate Pd nanoparticles
through an impregnation-reduction method. TEM analysis showed
Pd nanoparticles were uniformly dispersed in the channels of the
nanotubes with the average size of 2e4 nm, which was also
confirmed by XRD from Scherrer's equation. The synthesized cat-
alysts with different framework compositions exhibited an
7
excellent catalytic activity towards the selective oxidation of benzyl
alcohol and enantioselective hydrogenation of a,b-unsaturated
carboxylic acid, benefiting from the peculiar nanotube morphology
ee (%) and surface hydrophobic/hydrophilic properties. Furthermore, the
40 doped carbon may have a positive influence on the oxidation of
48 benzyl alcohol, which exhibited the highest yield of benzaldehyde.
30 This study provides a series of silica-based nanotubes as the sup-
33
ports to investigate the relationship between the framework
compositions and the reaction activities in heterogeneous catalysis.
Acknowledgements
We acknowledge the National Natural Science Foundation of
China (No. 21276191, U1662109), the Specialized Research Fund for
the Doctoral Program of Higher Education of China (No.
20120032120083), the Natural Science Foundation of Tianjin, China
(No. 16JCQNJC06200) for financial support.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.micromeso.2017.03.040.
References
[1] D.S. Su, S. Perathoner, G. Centi, Chem. Rev. 113 (2013) 5782e5816.
[2] T. Fujigaya, N. Nakashima, Adv. Mater. 25 (2013) 1666e1681.
[3] Y.-Z. Long, M.-M. Li, C. Gu, M. Wan, J.-L. Duvail, Z. Liu, Z. Fan, Prog. Polym. Sci.
36 (2011) 1415e1442.
[4] W. Chen, X. Pan, X.H. Bao, J. Am. Chem. Soc. 129 (2007) 7421e7426.
[5] Q. Wang, X. Wang, Z. Chai, W. Hu, Chem. Soc. Rev. 42 (2013) 8821e8834.
[6] S. Ding, N. Liu, X. Li, L. Peng, X. Guo, W. Ding, Langmuir 26 (2010) 4572e4575.
[7] P. Roy, S. Berger, P. Schmuki, Angew. Chem. Int. Ed. Engl. 50 (2011)
2904e2939.
[8] Z. Huang, D. Raciti, S. Yu, L. Zhang, L. Deng, J. He, Y. Liu, N.M. Khashab, C. Wang,
J. Gong, Z. Nie, J. Am. Chem. Soc. 138 (2016) 6332e6335.
[9] X. Liu, X. Li, Z. Guan, J. Liu, J. Zhao, Y. Yang, Q. Yang, Chem. Commun. 47 (2011)
8073e8075.
[10] X. Zhang, F. Su, D. Song, S. An, B. Lu, Y. Guo, Appl. Catal. B Environ. 163 (2015)
50e62.
[11] X. Liu, Y. Goto, Y. Maegawa, T. Ohsuna, S. Inagaki, Apl Mater. 2 (2014) 113308.
[12] A.N. Campbell, S.S. Stahl, Acc. Chem. Res. 45 (2012) 851e863.
[13] Q.-A. Chen, Z.-S. Ye, Y. Duan, Y.-G. Zhou, Chem. Soc. Rev. 42 (2013) 497e511.
[14] Y. Wang, J. Yao, H. Li, D. Su, M. Antonietti, J. Am. Chem. Soc. 133 (2011)
2362e2365.
[15] V.M. Shinde, E. Skupien, M. Makkee, Catal. Sci. Technol. 5 (2015) 4144e4153.
[16] G. Park, S. Lee, S.J. Son, S. Shin, Green Chem. 15 (2013) 3468e3473.
[17] Y. Niu, H. Wang, X. Zhu, Z. Song, X. Xie, X. Liu, J. Han, Q. Ge, Microporous
Mesoporous Mater. 198 (2014) 215e222.
[18] J. Yang, W. Chen, X. Ran, W. Wang, J. Fan, W.-x. Zhang, RSC Adv. 4 (2014)
20069e20076.
[19] Y. Chen, H. Lim, Q. Tang, Y. Gao, T. Sun, Q. Yan, Y. Yang, Appl. Catal. A Gen. 380
(2010) 55e65.
8 J. Sun et al. / Microporous and Mesoporous Materials 247 (2017) 1e8

[20] A. Savara, C.E. Chan-Thaw, I. Rossetti, A. Villa, L. Prati, ChemCatChem 6 (2014) [31] F. Su, S.C. Mathew, G. Lipner, X. Fu, M. Antonietti, S. Blechert, X. Wang, J. Am,
3464e3473. Chem. Soc. 132 (2010) 16299e16301.
[21] J. Pang, V.T. John, D.A. Loy, Z. Yang, Y. Lu, Adv. Mater. 17 (2005) 704e707. [32] A. Villa, N. Janjic, P. Spontoni, D. Wang, D.S. Su, L. Prati, Appl. Catal. A Gen. 364
[22] M.A. Pimenta, G. Dresselhaus, M.S. Dresselhaus, L.G. Cancado, A. Jorio, R. Saito, (2009) 221e228.
Phys. Chem. Chem. Phys. 9 (2007) 1276e1291. [33] T. Harada, S. Ikeda, M. Miyazaki, T. Sakata, H. Mori, M. Matsumura, J. Mol.
[23] B. Vinayan, K. Sethupathi, S. Ramaprabhu, Int. J. Hydrogen Energy 38 (2013) Catal. A Chem. 268 (2007) 59e64.
2240e2250. [34] M. Gruttadauria, F. Giacalone, R. Noto, Chem. Soc. Rev. 37 (2008) 1666e1688.
[24] C.D. Pina, E. Falletta, M. Rossi, Chem. Soc. Rev. 41 (2012) 350e369. [35] M. Lamblin, L. Nassar-Hardy, J.-C. Hierso, E. Fouquet, F.-X. Felpin, Adv. Synth.
[25] B.N. Zope, D.D. Hibbitts, M. Neurock, R.J. Davis, Science 330 (2010) 74e78. Catal. 352 (2010) 33e79.
[26] J.M. Hoover, S.S. Stahl, J. Am. Chem. Soc. 133 (2011) 16901e16910. [36] T. Mallat, E. Orglmeister, A. Baiker, Chem. Rev. 107 (2007) 4863e4890.
[27] P. Zhang, Y. Gong, H. Li, Z. Chen, Y. Wang, Nat. Commun. 4 (2013) 1593. [37] Y. Nitta, K. Kobiro, Chem. Lett. 10 (1996) 897e898.
[28] D.I. Enache, J.K. Edwards, P. Landon, B. Solsona-Espriu, A.F. Carley, [38] G. Szo} llo
}si, B. Herma n, K. Felfo
€ldi, F. Fülo
€p, M. Barto
k, J. Mol. Catal. A Chem.
A.A. Herzing, M. Watanabe, C.J. Kiely, D.W. Knight, G.J. Hutchings, Science 311 290 (2008) 54e59.
(2006) 362e365. [39] Y. Nitta, J. Watanabe, T. Okuyama, T. Sugimura, J. Catal. 236 (2005) 164e167.
[29] M.S. Sigman, D.R. Jensen, Acc. Chem. Res. 39 (2006) 221e229. [40] Z. Guan, S. Lu, C. Li, J. Catal. 311 (2014) 1e5.
[30] W. Hou, N. Dehm, R. Scott, J. Catal. 253 (2008) 22e27.