One-Pot Synthesis of Iron-Containing Nanoreactors with Controllable Catalytic Activity Based on Multichannel Mesoporous Silica

Article abstract of DOI:10.1002/cctc.201500868

We have constructed Fe-substituted silicate nanotube reactors (Fe-NTRs) through a newly developed one-step solubilization (OSS) method in which cetyltrimethylammonium bromide and P123 are used as structure-directing and shape-controlling agents, respectively, and ferrocene is utilized as iron source by solubilizing inside the hydrophobic inner core of the surfactant micelles. The Fe-NTRs showed regular morphology, highly ordered multichannel mesopores, controllable sizes with lengths changing from 150nm to 1.2μm, diameters of approximately 130nm and pore sizes of approximately 2.7nm, and iron contents in a range from 0.66 to 1.59wt %. The OSS strategy is also applicable for other metal-functioned silicate nanoreactors. The catalytic activity of Fe-NTRs in the direct hydroxylation of phenol can be essentially manipulated by the controllable sizes. The release behavior of reactants reveals that the diffusion control derived from the lengths of Fe-NTR should be responsible for the resultant reactivity.

Full text of DOI:10.1002/cctc.201500868

DOI: 10.1002/cctc.201500868
Full Papers
One-Pot Synthesis of Iron-Containing Nanoreactors with
Controllable Catalytic Activity Based on Multichannel
Mesoporous Silica
[
a]
Haiqing Wang, Xiaoming Li, Cuirong Xiong, Shuying Gao, Jun Wang,* and Yan Kong*
We have constructed Fe-substituted silicate nanotube reactors
130 nm and pore sizes of approximately 2.7 nm, and iron con-
tents in a range from 0.66 to 1.59 wt%. The OSS strategy is
also applicable for other metal-functioned silicate nanoreac-
tors. The catalytic activity of Fe–NTRs in the direct hydroxyl-
ation of phenol can be essentially manipulated by the control-
lable sizes. The release behavior of reactants reveals that the
diffusion control derived from the lengths of Fe–NTR should
be responsible for the resultant reactivity.
(
(
Fe–NTRs) through a newly developed one-step solubilization
OSS) method in which cetyltrimethylammonium bromide and
P123 are used as structure-directing and shape-controlling
agents, respectively, and ferrocene is utilized as iron source by
solubilizing inside the hydrophobic inner core of the surfactant
micelles. The Fe–NTRs showed regular morphology, highly or-
dered multichannel mesopores, controllable sizes with lengths
changing from 150 nm to 1.2 mm, diameters of approximately
Introduction
[
1,14]
Nanoreactors have received a lot of attention in recent years
owing to their multifunctionality in various catalysis applica-
maintaining the ordered mesostructures.
Recent studied in-
dicated that the tubular particles with multiple nanochannels
can be used as an ideal nanoreactor for nanocatalysis, because
each array of nanochannels can be regarded as a reactor and
the nanochannel can exert more enhanced confinement
[
1,2]
tions.
In principle, the construction of nanoreactors require
elaborate active sites and a controllable nanostructure. Their
catalytic performance (e.g., conversion, yield, selectivity, and
product distribution) can be manipulated in a controllable
manner by tuning the local nanostructures, such as morpholo-
[
9,15,16]
effect.
However, the effective control of multichannel
silica nanotubes by means of direct synthesis was poorly de-
[
3,4–9]
[17]
gy, size, or composition
due to the confined space, particle
fined.
[
10]
size, and morphology-dependent reaction kinetics.
To
Previous works have revealed that the Fe, Ti, and Zr-contain-
ing silicate catalysts are of academic and industrial interest for
their versatility in many important reactions, such as alkylation,
cracking, isomerization, aromatics oxidation, Fenton reaction,
achieve these essential catalytic structures with a high degree
of control, the nanosynthesis techniques of catalysts are cru-
cial, however, they remain technically challenging.
[
7,8,16,18]
Mesoporous silica nanoparticles (MSNs) have been exten-
sively studied owing to their excellent texture characteristics
such as large pore volume, high surface area, well-ordered
pore size (2–50 nm), flexible mesostructure, diverse surface
Friedel–Crafts acylation, and Fischer–Tropsch synthesis.
However, the shape-controlled synthesis of the abovemen-
tioned catalysts combined with ordered mesopores is rarely
achieved. The syntheses of Fe, Ti, and Zr-containing nanoreac-
tors are totally different from that of the recently obtained
[5,11]
functionality, and good biocompatibility.
These properties
[
19]
of MSNs make them attractive as hosts for nanocatalysis. At
present, pure MSNs with various morphologies including
sphere, helix, fiber, rod/tube, hollow, crystal, and other special
copper-containing nanotube reactor,
which can be con-
structed by controlling the simultaneous hydrolysis of copper
and silicon sources by the formation of copper–ammonia com-
plex aqueous solution. Apparently, the synthetic approach
cannot be widely adapted for other inorganic transition metal
sources because the precipitates are usually formed under al-
kaline conditions. Moreover, the inorganic metal salts have re-
vealed more adverse effects on the original shape of surfactant
micelles and caused the assembly of silica monomers by exotic
electrostatic interactions, which inevitably destructed the mor-
phology and mesopores of the resultant metal-containing sili-
[12,13]
morphologies have been obtained.
To further enhance the
inherent properties and endow MSNs with multifunctionality, it
is important to achieve morphology and size control while
[
a] Dr. H. Wang, X. Li, C. Xiong, S. Gao, Dr. Prof. J. Wang, Dr. Prof. Y. Kong
State Key Laboratory of Materials-Oriented Chemical Engineering
College of Chemistry and Chemical Engineering
Nanjing Tech University
[
20,21]
Nanjing 210009 (P.R. China)
Fax: (+86)25-83587860
E-mail: kongy36@njtech.edu.cn
cate materials.
Taken together, the fabrication of mesopo-
rous silicate nanoreactors requires (i) an effective and flexible
strategy for the controllable synthesis of materials; (ii) a suit-
able silicon and metal source; (iii) precisely controlled synthesis
parameters. The one-step synthesis of a real mesoporous sili-
junwang@njut.edu.cn
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201500868.
ChemCatChem 2015, 7, 3855 – 3864
3855
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
cate nanoreactor with transition-metal functionality and regu-
lar multichannel tubular morphology has been rarely reported
owing to the limitations in the design of the support and
active site.
Table 1. The iron content in samples with different addition amount of
CTAB or TEOS.
[a]
Sample
Iron content
wt%]
Sample
Iron content
[wt%]
[
Herein, we report on the development of a new one-step
solubilization (OSS) process to construct iron-doped multichan-
nel mesoporous silica nanotube reactors (Fe–NTRs) with con-
trollable nanostructures. The broader applicability of OSS strat-
egy was demonstrated by using other metallocenes as metal
sources. The effects of iron contents and sizes of Fe–NTRs on
the catalytic activity were investigated in the direct hydroxyl-
ation of phenol with H O as oxidant.
0
0
0
.2, 0.02, 2.0
.4, 0.02, 2.0
.8, 0.02, 2.0
0.50
0.65
0.81
0.4, 0.06, 1.0
0.4, 0.06, 2.0
0.4, 0.06, 3.0
0.70
0.65
0.58
[a] CTAB (g), P123 (g), TEOS (mL)
inside the hydrophobic core of CTAB micelles. In contrast, the
increased amount of tetraethylorthosilicate (TEOS) led to the
decrease of iron content from 0.70 to 0.58 wt%. The latter
should result from the presence of ethanol in the solution
after the addition of TEOS. Ethanol increases the solubility of
ferrocene and leaches the organo-metal complex outside the
micelles, resulting in the lower iron content. This result indi-
cates that in the OSS procedure the iron species were incorpo-
rated into the mesopores of materials by the solubilization of
ferrocene molecules inside the hydrophobic core of CTAB mi-
celles. The FTIR spectra (see Figure 3) and structural parameter
results (see Table 2) further support the OSS strategy, which
will be discussed below.
2
2
Results and Discussion
OSS strategy: in situ incorporation of iron into mesoporous
silica
Our proposed new OSS strategy can be described as the pro-
cess given in Figure 1. The original mixture consisted of cetyl-
Mesostructure
The small-angle XRD patterns of the parent mesoporous silica
and Fe–NTR samples (5R1, 5R2, 5R3, 5R4, and 5R5 with varied
length; 10, 15, and 20R3 with altered iron contents, see Experi-
mental Section for details) are shown in Figure 2a and 2b, re-
spectively. The sharp (100) reflection peak at 2q=2.0 to 2.58
demonstrates the characteristics of a typical mesoporous struc-
ture. Three other weak peaks, indexed to the (110), (200), and
Figure 1. Illustration of OSS strategy.
[
22]
trimethylammonium bromide (CTAB), PEO PPO PEO (P123,
(210) reflections,
confirmed the presence of a highly or-
2
0
70
20
[
23]
M ꢀ5800), and ferrocene in ammonia. The ferrocene mole-
dered P6mm 2D hexagonal mesophase.
Compared with
n
cules were preferentially solubilized inside the hydrophobic
core of SDA micelles, because of their insolubility in water. A
small amount of P123 was used as an additive for the control
over the size and morphology of Fe–NTRs. The micelles solubi-
lizing ferrocene were stacked into hexagonal array (Step I).
After the addition of silica source, the hydrolyzed silicate spe-
cies were assembled into inorganic silica framework around
these micellar arrays, as shown in Step II. Then, the precursor
of organ-iron compound was confined in the nanochannels of
silicate. During the removal of the surfactants by calcination at
pure samples, the typical hexagonal structure of Fe–NTRs was
clearly observed, which indicates that by using this new OSS
strategy, the highly ordered mesopore structure was well re-
tained in the iron-containing Fe–NTR samples. All the diffrac-
tion peaks of Fe–NTRs shift to lower angles, indicating the en-
[
4,8]
larged structure units.
This observation is an evidence of
the ferrocene molecules solubilizing into the hydrophobic
+
region of CTA micelles. Sample 5R5 show relatively low re-
flections suggesting the reduced ordering degree of its meso-
[
7]
structure. For the Fe–NTRs with altered iron content (Fig-
ure 2b), the peaks of 20R3 shifted to higher angles, which may
be caused by the formation of extra-framework iron spe-
5
508C for 6 h (Step III), ferrocene molecules were decomposed
into inorganic iron species. Finally, the iron-incorporated sili-
cate materials with regular morphology and ordered meso-
pores were achieved.
[
7,20]
cies.
The amount of ferrocene that can be solubilized by
CTAB micelles is limited. For sample 25R3 with over 2.5 wt%
iron content, only one broad (100) peak (Supporting Informa-
tion, Figure S1) could be observed, and the (110) and (200) re-
flections disappeared. No reflections for 40R-3 (Figure S1) were
detected suggesting a complete loss of mesopores. These ob-
servations indicate that higher input amount of ferrocene can
lead to a reduced order degree of mesostructure of Fe–NTRs.
To verify the OSS strategy, that is, the solubilization of ferro-
cene inside the hydrophobic core of surfactant micelles, we
conducted control experiments. As shown in Table 1, the iron
contents exhibited an upward trend from 0.50 to 0.81 wt%
with increasing amount of CTAB from 0.2 to 0.8 g. This finding
should be ascribed to the improved solubilization of ferrocene
ChemCatChem 2015, 7, 3855 – 3864
www.chemcatchem.org
3856
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
À1
Fe–NTRs. The characteristic bands of CÀH at 2921 cm and
À1
2
852 cm disappeared in calcined 5R2, suggesting the total
[
7]
removal of template after calcination at 5508C for 6 h. The
complete decomposition of ferrocene to inorganic iron species
was confirmed by the disappearance of C=C vibration band at
À1
1
384 cm .
N physisorption
2
The structure parameters of the obtained materials were evalu-
ated by nitrogen physisorption isotherms. As shown in
Figure 4, all Fe–NTRs exhibited nitrogen adsorption-desorption
Figure 2. XRD patterns of a) pure silica, samples 5R1, 5R2, 5R3, 5R4, and 5R5;
b) 10, 15, and 20R.
In contrast, the mesoporous structure of Fe–NTRs were highly
ordered if a moderate amount of ferrocene was used.
FTIR spectra of ferrocene, as-synthesized pure sample (with-
out ferrocene), as-synthesized 5R2 and calcined 5R2 are
shown in Figure 3. Compared with the spectra of ferrocene
and as-synthesized pure sample, the remarkable peak at
À1
1
380 cm should be assigned to the stretching vibration of
[
24]
C=C in ferrocene, indicating the incorporation of ferrocene
molecules into the materials during the synthesis procedure.
This further supports our proposed OSS synthetic process of
Figure 4. Nitrogen adsorption–desorption isotherms of Fe–NTRs.
isothermal curves similar to that of the pure mesoporous silica.
Their type IV isotherms together with a sharp inflection at rela-
tive P/P between 0.2 and 0.4, indicate the presence of uniform
0
cylindrical mesopores and very narrow pore-size distribu-
[
25]
tions. The second capillary condensation step above 0.95 in
the adsorption branch was observed more apparently from
5
R5 to 5R1 (Figure 4a), which is attributed to the interparticle
[
26]
voids, indicating the smaller particle size.
Sample 5R5
showed relatively low amount of physisorbed nitrogen indicat-
ing a reduced ordering degree of mesopores, in agreement
with the XRD pattern of 5R5. These observations indicated
that the ordered multichannel structure was well preserved in
the iron-containing samples, which accorded well with the re-
sults of XRD. Compared with pure sample synthesized without
Figure 3. FTIR spectra of a) ferrocene, b) as-synthesized silica (without ferro-
cene), c) as-synthesized 5R2, and d) calcined 5R2.
ChemCatChem 2015, 7, 3855 – 3864
www.chemcatchem.org
3857
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
ferrocene, the capillary condensation step in Fe–NTRs shifted
toward a relatively higher relative pressure, which indicates
[
7]
a bigger pore size. Sample 5R3 showed an increase of ap-
proximately 0.3 nm in the pore size, as listed in Table 2, which
Table 2. Structural properties of Fe–NTRs 5R1, 5R2, 5R3, 5R4, and 5R5.
Sample
D
100
a
0
D
p
W
t
S
BET
2
Total pore volume
À1
3
À1
[nm] [nm] [nm] [nm] [m g
]
[cm g
]
pure silica
3.65
3.98
3.98
3.91
3.91
3.88
4.22
4.60
4.60
4.51
4.51
4.47
2.37
2.71
2.71
2.70
2.71
2.70
1.85
1.89
1.89
1.81
1.81
1.77
985
965
949
917
901
619
0.89
1.34
1.15
1.12
1.03
0.66
5
5
5
5
5
R1
R2
R3
R4
R5
Figure 5. SEM images of Fe–NTRs: a) 5R (i.e., 5R2), b) 10R, c) 15R, and
d) 20R.
should be due to the introduction of ferrocene inside the hy-
drophobic core of CTAB micelles. These findings demonstrated
the correctness of the new OSS strategy. Some structural pa-
rameters of samples 10R, 15R, and 20R are listed in Table 3.
flected in Figure S1), the tubular morphology was till main-
tained (in Figure S2b), indicating that the overloading of iron
precursor only destruct the local ordering of CTAB micelles.
These observations demonstrated that the morphology of Fe–
NTRs can be well-preserved with the OSS strategy.
Table 3. Structural properties of Fe–NTRs 10R, 15R, and 20R.
Samples 5R1, 5R2, 5R3, 5R4, and 5R5 were obtained by al-
tering the concentration of shape-controlling additive. The
effect of the concentration of P123 on the particle morphology
and size was investigated with SEM. As shown in Figure 6, the
tube morphology of 5Rn samples are found to be uniform
throughout the products. With a decrease in [P123] from
Sample
D
100
a
[nm]
0
D
[nm]
p
w
[nm]
t
S
BET
2
Total pore volume
À1
3
À1
[nm]
[m g
]
[cm g
]
1
1
2
0R
5R
0R
3.91
3.98
3.65
4.51
4.59
4.22
2.62
2.57
2.12
1.89
2.02
2.10
908
859
842
0.83
0.76
0.70
À1
À2
2
.59”10 mm to 3.45”10 mm, it is clear that the lengths of
For samples 15R3 and 20R3, the pore sizes, the specific area,
and the total pore volume were reduced with the increase of
iron content, which should be ascribed to the formation of
samples 5R1 to 5R5 were increased from 150 nm to 300 nm,
600 nm, 800 nm, and 1.2 mm. Considering the lower order
degree of mesopores in 5R5, as reflected in XRD pattern and
nitrogen physisorption isotherm, it was suggested that local
destruction effect of ferrocene should be more evident with
the increased length of samples. Samples have a similar outer
diameter of approximately 130 nm. The aforementioned results
indicated that the shape-controlling additive P123 endowed
the Fe–NTRs with controllable sizes. The sizes and the chang-
ing trend of Fe–NTRs with varied [P123] were summarized in
Figure 6e. The P123 micelles had a hydrodynamic radius of ap-
proximately 9.8 nm, whereas all Fe–NTRs displayed a pore size
of approximately 2.7 nm, which indicates that the P123 may
not enter into the micelles. Moreover, the break-up of P123 mi-
celles usually happens at high ionic surfactant concentra-
[4,7]
iron oxide on the channel surface of materials.
The results of aforementioned characterizations show that
by using the OSS method, the iron species were successfully
incorporated into the mesopores of Fe–NTRs. The ordered mul-
tichannel mesopore structure of the original samples can be
well-preserved in the iron-containing samples (Fe–NTRs).
Morphology
The SEM images of samples 5R (i.e., 5R2), 10R, 15R, and 20R
with variations in iron contents are revealed in Figure 5. Note
that if an inorganic iron source such as FeCl was used, a bronz-
3
[
27]
ing precipitation of ferric hydroxide (Fe(OH) ) was quickly gen-
tions. We suggest that the hexagonal mesostructure of Fe–
NTRs was predominantly constructed by the strong electrostat-
ic interactions between negatively charged silicates and posi-
3
erated. The resultant silicate sample showed monolithic irregu-
lar morphology, not shown here. In contrast, it is apparent that
the resultant particles in Figure 5 are well-monodispersed and
exhibit regularly tubular morphology. The tubes have an aver-
age size of 300 nm in length and 130 nm in diameter. Com-
pared with the pure sample (Figure S2a) synthesized with the
same amount of CTAB and P123, these samples showed similar
size, indicating that the addition of ferrocene exhibited negligi-
ble effects on the morphology of Fe–NTRs. Note that although
the hexagonal mesophase of sample 40R was collapsed (as re-
+
[28]
tively charged CTA micelles.
The P123 molecules may
reside at the outer surface of micelles of CTAB. If the molar
ratio of P123/CTAB was further increased, more end-face cover-
age of P123 molecules will ultimately decrease the length of
micelles and realize the controllable formation of Fe–NTRs.
Therefore, the P123 molecules should act as a shape-control-
ling additive to modulate the morphology and size of Fe–
NTRs. Elemental mapping by energy-dispersive X-ray spectros-
ChemCatChem 2015, 7, 3855 – 3864
www.chemcatchem.org
3858
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 7. TEM images of the representative Fe–NTRs: a) 5R1, b) 5R3, c) 5R5,
and d) 10R. The insets in (b) give the top view (top) and the SAED pattern
(bottom).
onstrates the highly ordered 2D-hexagonal channel arrange-
ment of materials. The well-ordered multichannel structure of
5
R1, 5R3, and 10R was clearly visualized in a side view (Fig-
ure 7a, b, and d), and the mesostructure of 5R5 appears slight-
ly disordered (Figure 7c), according well with the results of
XRD and N physisorption. The pore sizes were found to be in
2
the range of 2.6–2.8 nm, in good agreement with the Dp re-
sults calculated from the BJH model using desorption data
(
Table 2). Each hexagonal unit of the paralleled pore structures
can be regarded as a nanochannel reactor. In contrast, the
mesopores of 25R3 with higher iron content (TEM, Figure S3)
are less ordered in hexagonal arrays, which accords with the
XRD result. These observations indicate that the Fe–NTRs with
OSS process can maintain the tubelike morphology as well as
the ordered hexagonal mesostructure of the parent silica. Note
that iron oxide particles are clearly observed with the direct
calcination of ferrocene. However, no iron oxide particles or
aggregations are detected outside or inside the channels of
Fe–NTR. The iron species are highly dispersed into the chan-
nels, which should ascribe to the confinement of silicate nano-
channel, indicating the introduced iron active centers are
highly accessible to guest molecules.
Figure 6. SEM images of Fe–NTRs: a) 5R1, b) 5R3, c) 5R4, d) 5R5, e) the size
variation of Fe–NTRs with [P123], f,g) EDX elemental mapping of aggregated
5
R3 without ultrasonic treatment (SEM image (g), scale bar is 1 mm).
copy (EDX) shows that the elements Si, O, and Fe are homoge-
neously distributed throughout the whole particle (Figure 6 g).
To the best of our knowledge, this is the first report for the
controllable preparation of metal-containing mesoporous sili-
cate nanotube nanoreactor (Fe–NTRs) by a one-step method.
Taken together, in this new procedure, the OSS strategy pre-
served the ordered multichannel structure and original tube
morphology of particles. In addition, the shape-controlling ad-
ditive P123 provided an effective and flexible route for the pre-
cise design of nanoreactors.
To demonstrate the applicability of the OSS strategy, titano-
cene dichloride and zirconocene dichloride were used for pre-
paring Ti and Zr-containing mesoporous silicate, whereby
a molar ratio of metallocene/silicon=0.05 was used in the syn-
thesis gel. The four well-defined reflection peaks in XRD pat-
tern (Figure S4a and b) indicate the presence of highly ordered
mesoporous channels in the Ti and Zr-containing reactors. The
Ti and Zr contents are 0.051 wt% and 0.047 wt%, respectively.
UV/Vis diffuse reflectance (Figure S4c and d) spectroscopy sug-
[
29]
[30]
gests the formation of the framework Ti and Zr species in
the mesoporous silicate reactors. As shown in Figure S5, the
TEM images of Ti– and Zr–NTR illustrate a highly ordered mul-
tichannel mesoporous structure and regular tube morphology.
The result indicates that this new OSS strategy is broadly appli-
cable to the preparation of various metal-containing silicate
nanoreactors for practical applications. Potentially, this method
can endow nanoreactor with two different metals within one
High-resolution transmission electron microscopy (HRTEM)
TEM images of the representative Fe–NTRs: 5R1, 5R3, 5R5,
and 10R are depicted in Figure 7. The hexagonal honeycomb-
like mesostructure of Fe–NTR was well-defined as seen in the
top view of 5R3 (Figure 7b, upper inset). The selected-area
electron diffraction pattern (SAED, Figure 7b, lower inset) dem-
ChemCatChem 2015, 7, 3855 – 3864
www.chemcatchem.org
3859
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
assembly, which will attract special attention because of poten-
tial synergic effects of different metals for a variety of ad-
persed on the channel surface of the samples. In Figure 8b, for
the samples 15R and 20R, if the content of iron reached more
than 1.18 wt% (in Table 4), the extra broad absorption band
[
31]
vanced applications.
The states of iron species
Table 4. Catalytic activity of Fe–NTRs: 5R, 10R, 15R, and 20R in the
phenol hydroxylation to dihydroxybenzene.
UV/Vis spectroscopy is a very effective method for the charac-
terization of coordinated environment of iron species in silicate
materials. The UV/Vis diffuse reflectance spectra in the wave-
length range of 230–650 nm of some samples are shown in
Figure 8. Note that no absorption band is present for the
[a]
[b]
[c]
[d]
[e]
Sample
Iron content
[wt%]
X
S
[%]
Y
Cat./Hyd.
[%]
[%]
5
10R
R
0.66
0.98
1.18
1.59
–
42.4
51.3
29.4
15.2
9.8
57.3
47.6
30.1
23.5
18.6
24.3
24.4
8.8
3.6
1.8
1.52
1.61
1.43
1.64
1.70
1
2
5R
0R
Fe
2
O
3
[
[
a] ICP results. [b] Phenol conversion. [c] Dihydroxybenzene selectivity.
d] Yield. [e] Molar ratio of catechol/hydroquinone.
between 600 and 650 were detected indicating that the exis-
[
6]
tence of aggregated Fe O particles. However, no characteris-
2
3
tic peaks of iron oxides were found in high-angle XRD patterns
not shown here), suggesting that the introduced iron species
by OSS procedure should be incorporated into the framework
(
[
4,7]
of silica or highly dispersed on the wall of nanoreactor.
This
result should be ascribed to the fact that the interactions be-
tween ferrocene molecules were weakened when being solubi-
lized inside the hydrophobic core of CTAB micelles. As a result,
the dispersion degree of iron species was improved.
X-ray photoelectron spectroscopy (XPS)
The calcined Fe–NTRs were characterized by XPS technique. In
Figure 9 the Fe2p XPS spectra of the representative 5R and
Figure 8. UV/Vis absorption spectra for the Fe–NTRs: a) pure silica, 5R1, 5R2,
5
R3, 5R4, and 5R5; b) 10R, 15R, and 20R.
parent mesoporous silica. A broad band between 230 nm and
00 nm centered at 255 nm, assigned to the low-energy d –p
3
P
P
charge-transfer transitions between tetrahedral oxygen ligands
3
+
[7,32]
and central Fe ion, was observed for all samples.
band demonstrated that the bonds of FeÀOÀSi were formed
and most of iron atoms existed in the framework of silicate.
A shoulder band at around 330 nm indicates the presence of
This
Figure 9. XPS spectra of Fe2p Fe–NTRs: a) 5R and b) 20R.
[
20]
[33]
polyferrate (FeÀOÀFe)n on the silicate surface.
The broad
20R samples are given. Two peaks assigned to Fe2p3/2
band between 400 and 550 nm should be ascribed to highly
(ꢀ712 eV) and Fe2p1 (ꢀ725.0 eV) are clearly observed for
/2
[7]
dispersed extra-framework iron oligomer. The observations
from the UV/Vis spectra indicate that for Fe–NTRs with iron
content of 0.6 wt%, most iron atoms are incorporated into the
framework of Fe–NTRs in the form of isolated iron species and
only very few extra-framework iron species were highly dis-
the iron-containing samples, which indicate the presence of
3
+ [33–35]
Fe
.
The other peaks are attributed to shake-up satellites
[
35]
of the 2p3/2 and 2p1/2 peaks. The binding energy of the pure
iron oxide is approximately 710.8 eV and 724.2 eV, respective-
[
7]
ly. In contrast, the Fe2p3/2 and Fe2p1/2 of all Fe-containing
ChemCatChem 2015, 7, 3855 – 3864
www.chemcatchem.org
3860
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[
7]
samples show a little positive shift of 0.8–1.0 eV in binding
energy, which may be attributed to the higher electronegativi-
to side-reaction such as the deep oxidation of phenol. As
a result, the selectivity to dihydroxybenzene was dropped with
the increasing amount of extra-framework iron species. Al-
though the existence of extra-framework iron is evident, the
catalytic activity showed in 15R and 20R can demonstrate that
extra-framework iron species were highly dispersed in agree-
ment with the result of high-angle XRD. The aforementioned
results indicate that the iron species with tetrahedral coordina-
tion in the framework of Fe–NTR were effective for the phenol
hydroxylation reaction. In contrast, the oligomeric extra-frame-
work iron results in the lower conversion rate of phenol and
dihydroxybenzene selectivity.
[7]
ty of silicon than iron. The result indicates the incorporation
[35]
of iron atoms into the silicate framework, which accords well
with the UV/Vis spectroscopy observations. The two peaks of
2
0R became broader and more intensive, indicating an in-
[7,35]
creasing iron content in the samples.
Moreover, the bulk
iron content and the surface iron content in sample 20R
1.59 wt%) are 0.017 and 0.009 determined by inductively cou-
(
pled plasma mass spectrometry (ICP–MS) and XPS, respectively.
As for the 5R (0.65 wt%) with lower iron loading amount, the
bulk iron content (molar ratio of Fe/Si=0.007) is higher than
0
.002 of surface iron content obtained from XPS. This observa-
tion indicates, with the increasing amount of iron loading,
more extra-framework iron species were produced. In the case
of lower iron loading amount, most of iron species are incor-
porated into the silicate framework of Fe–NTRs. These findings
further support the conclusion obtained from UV/Vis spectros-
copy.
Effects of morphology
The effects of the morphology of Fe–NTRs on the catalytic ac-
tivity in phenol hydroxylation were studied by using 5R1, 5R2,
5R3, 5R4, and 5R5 with different length as catalysts. As shown
in Table 5, the conversion rate of phenol first went up from
Catalytic performance
Table 5. Catalytic activity of Fe–NTRs: 5R1, 5R3, 5R4, and 5R5.
The catalytic activity of the produced Fe–NTRs was examined
by the one-step hydroxylation of phenol to dihydroxybenzene
with H O as an oxidant.
Samples
Length
mm]
Iron content
[wt%]
X
[%]
S
[%]
Y
[%]
Cat./Hyd.
[
2
2
5
5
5
5
R1
R3
R4
R5
0.15
0.60
0.80
1.20
0.68
0.65
0.65
0.65
35.5
48.8
50.1
40.6
61.8
55.6
50.5
47.2
21.9
27.1
25.3
19.2
1.55
1.49
1.48
1.42
Effects of iron content
The effects of iron content were evaluated based on the Fe–
NTRs 5R, 10R, 15R, and 20R. The results were summarized in
Table 4. Pure mesoporous silica nanotube showed no catalytic
activity because of the absence of active sites. If using the
iron-containing samples, the catalytic activity was remarkably
improved, which indicates that the iron species should be re-
sponsible for the phenol hydroxylation reaction. From the cata-
lytic activity of Fe–NTRs with the same morphology, we found
that with the increasing loading amount of iron, the conver-
sion rate of phenol increases first from 48.8% of 5R to 51.3%
of 10R, and then decreases to approximately 29.4% of 15R
and 15.2% of 20R. The selectivity to dihydroxybenzene de-
creased gradually from 55.6% of 5R to 47.6% of 10R, 30.1% of
35.5% to 50.1%, as the lengths of samples 5R1 to 5R4 in-
creased from 150 nm to 800 nm, and then decreased to 40.6%
of 5R5 with 1.2 mm in length. In contrast, the selectivity to di-
hydroxybenzene of samples from 5R1 to 5R4 decreased grad-
ually from 61.8% to 47.2%. The Fe–NTR 5R3 with 600 nm in
length exhibited the highest yield of 27.1% with 48.8% con-
version of phenol and 55.6% selectivity to dihydroxybenzene
under the reaction condition.
It is generally known that the performance of a catalyst is
often related to the amount of active sites, morphology, and
[
4,6,7,36]
structural parameters of the support.
As for 5R5, the de-
1
5R, and 23.5% of 20R. Free Fe O nanoparticles, often used
creased degree of order of mesostructure, as reflected by XRD
2
3
as reference catalysts for similar reactions, show relatively low
activities. For the 5R and10R samples, their catalytic activities
are highly related to the amount of iron. As revealed by UV/Vis
spectroscopy, most of iron species in Fe–NTRs 5R and 10R
were incorporated in the silicate framework existing in the
form of tetra-coordinated iron. This observation indicates that
the framework iron species should act as active sites in phenol
hydroxylation. However, despite the improved iron content,
patterns and N physisorption isotherms, should be the reason
2
for its lower catalytic activity. In contrast, for samples 5R1 to
5R4, the total pore volume and BET specific surface area de-
3
À1
2
À1
crease from 1.34 to 1.03 cm g and from 965 to 901 m g
(Table 2), respectively, with the increasing length of Fe–NTRs.
The four samples were similar in the iron content and the va-
lence state of iron species. The catalytic activity should drop
according to the present analysis. Therefore, the unexpected
dependence of the catalytic activity of materials on the meso-
structure suggests that the altered lengths of Fe–NTRs should
be responsible for the resultant catalytic activity. It is well-
known that mass transfer is an important factor in heterogene-
ous catalysis, and closely related to the particle sizes. The inner
channels of mesoporous silica should be relatively hydrophilic
1
5R and 20R showed a dramatic decrease in phenol conver-
sion and dihydroxybenzene selectivity. The oligomeric iron of
extra-framework was detected in UV/Vis spectra for samples
1
5R and 20R. The reduced conversion rate of phenol should
be ascribed to the fact that the accessibility to the framework
iron was limited by the extra-framework oligomer of iron spe-
cies. Moreover, the extra-framework iron species contributed
[
12]
owing to the presence of abundant silanol groups. The sol-
ChemCatChem 2015, 7, 3855 – 3864
www.chemcatchem.org
3861
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Table 6. Comparison of the catalytic performance of iron-containing mesoporous silica-based catalysts in phenol hydroxylation to dihydroxybenzene.
[
a]
[c]
Catalysts
Fe/Si
Morphology
Reaction conditions
Phenol/H
2
O
2
X
S
Y
Ref.
[%]
[%]
[%]
[
b]
Fe–NTRs (5R3)
0.007
0.091
0.02
tube
20 /258C/5 h/20 mg
3:2
3:1
1:1
3:1
48.8
25.3
20.1
ꢀ20.0
55.6
78.4
73.3
ꢀ90.0
27.1
20.0
14.8
18.0
[
b]
8
2
Fe
Fe-Si
irregular
irregular
sphere
20 /208C/3 h/50 mg
[20]
[7]
[32]
[
b]
10 /258C/1.5 h/100 mg
[b]
Fe–MCM-41 NPs(120)
0.008
20 /608C/3 h/50 mg
(
120 nm)
[
a] Molar ratio of Fe/Si, [b] Weight ratio of phenol/catalyst. [c] Molar ratio of phenol/H O .
2 2
vent water is expected to be competing with the reactant
To be more precise on this point, we added the references
about other iron-containing mesoporous silica-based catalysts
in phenol hydroxylation to dihydroxybenzene for comparison.
In Table 6, some typical results reported in the literature are
summarized. Compared with the irregular Fe-containing cata-
[37]
phenol for the adsorption on the active sites. We suggested
that the increasing length of nanoreactor inevitably enhanced
the inner diffusion time of reactants. As a result, the effective
absorbance of phenol onto the active centers was promoted,
and the conversion rate of phenol was enhanced accordingly.
This result is consistent with the finding of Somorjai and co-
[
20]
[7]
lysts, such as 8Fe and 2Fe–Si, although Fe–NTR 5R3 has
lower iron content, it exhibited a relatively higher catalytic per-
formance. This finding suggests that the regular tube morphol-
ogy as well as the parallel multichannel structure of Fe–NTRs
plays an essential role for the enhanced catalytic performance
of the nanoreactors. Notably, compared with Fe–MCM-41
NPs(120) that has spherical morphology with 120 nm in diame-
ter and similar iron content to Fe–NTR, the catalytic activity of
[
38]
workers that the reactivity can be enhanced by increasing
the residence time of the reactants in fixed-bed plug flow reac-
tor. Moreover, the release profiles of phenol loaded Fe–NTRs
were studied in test tubes. Release profiles of typical 5R1 and
5
R4 with different lengths are shown in Figure 10. 5R1 and
5
R3 obtained at room temperature is higher than that of Fe–
MCM-41 NPs(120) at 608C. This observation indicates that the
longer channel should be favorable to the enhancement of
conversion rate phenol hydroxylation reaction, which is in
agreement with the experimental result.
Stability of Fe–NTRs
The test of the reusability of Fe–NTRs was conducted with
sample 5R3 as representative. We found that the conversion of
phenol decreased from 48.8% (fresh catalyst), 41.5% (first recy-
cle) to 31.3% (second recycle). Simultaneously, the Fe contents
determined by ICP–MS after each cycle of reaction were de-
creased from 0.65, 0.52 to 0.47 wt%. The deactivation of the
Fe–NTR 5R3 in phenol hydroxylation could be partially attrib-
uted to the gradual leaching of iron species. The selectivity to
dihydroxybenzene is in the range of 52.3% to 58.5%. The
result indicated that the iron species incorporated in the sili-
cate framework of Fe–NTR are more resistant to the leaching
and necessary for the stability of the iron-containing silicate
Figure 10. Release profiles of typical 5R1 and 5R4 with different lengths.
5
R4 exhibit the similar maximum release amounts of 0.033 g
phenol at 7 h and 11 h, respectively. These findings suggested
the diffusion properties of reactants can be well tailored by
finely controlling the sizes of nanoreactors and the longer Fe–
NTR showed advantages in promoting the presence time of re-
actants in nanoreactor. This could result in different catalytic
behaviors. Longer diffusion time inside the channels of Fe–
NTRs was unfavorable to the selectivity to dihydroxybenzene,
due to the existence of small amount of iron oxide species.
Moreover, the Fe–NTRs showed a potential regulation for cate-
chol/hydroquinone ratio, which decreased with the lengths of
Fe–NTRs, because the longer channels should be more favor-
able for the diffusion of hydroquinone. These findings exhibit-
ed that the catalytic activity of catalysts as well as the product
distribution can be well-regulated by fine control over size and
morphology of materials, suggesting a manipulable feature of
multichannel nanoreactor for practical applications.
[
20]
catalyst. The multichannel structure and the tubular mor-
phology of the Fe–NTRs were still preserved after reuse, as
shown in Figure S6.
Conclusions
We developed a novel one-step solubilization (OSS) strategy
for constructing Fe-containing multichannel mesoporous sili-
cate nanotube reactors (Fe–NTRs). cetyltrimethylammonium
bromide (CTAB) was used as structure-directing agent to form
mesopores of Fe–NTRs. The ferrocene was chosen as the metal
precursor, which was introduced into Fe–NTRs by solubilizing
+
inside the hydrophobic core of CTA . PEO PPO PEO (P123,
20
70
20
ChemCatChem 2015, 7, 3855 – 3864
www.chemcatchem.org
3862
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
M ꢀ5800) was applied as a shape-controlling additive to mod-
Characterization
n
ulate the morphology of nanotubes. The ordered multichannel
mesopore structure and regular tube morphology are well-pre-
served in Fe–NTRs. The sizes of Fe–NTRs are manipulable. For
the Fe–NTRs with lower iron loading amount, most of iron spe-
cies are incorporated into the framework of mesoporous silica.
With the increasing amount of iron loading, more extra-frame-
work iron species were produced. The iron species with tetra-
hedral coordination in the framework of Fe–NTR were favor-
able and effective for the phenol hydroxylation reaction. In
contrast, the oligomeric extra-framework iron results in lower
conversion rate of phenol and dihydroxybenzene selectivity.
The catalytic activity of Fe–NTRs can be well-regulated by the
fine control of size and morphology of nanoreactors. The OSS
strategy is a simple, controllable, and efficient process, and
also applicable to the preparation of Ti and Zr-containing sili-
cate tube nanoreactors. The present strategy of integrating the
functions of the tube morphology, multichannel mesopores,
controllable sizes, and active sites reveals a simple and flexible
access to the design and synthesis of a novel class of function-
al silicate nanoreactors.
The XRD patterns of the samples were collected with a diffractome-
ter (RigakuD/Max-RAX) equipped with a rotating anode and CuKa
radiation (l=0. 154178 nm). N physisorption isotherms were mea-
2
sured on a Micromeritics ASAP-2020 analyzer at 77 K. Before the
measurements, calcined samples were degassed in vacuum at
2
008C for 3 h. Surface areas were calculated using the BET equa-
tion and pore size distributions were obtained by the Barrett-
Joyner-Halenda (BJH) method using desorption branch data. FTIR
spectra were recorded with Bruker VECTOR22 FTIR spectrometer in
À1
the range of 4000–400 cm . Field-emission scanning electron mi-
croscopy was performed on a Hitachi S4800 Field Emission Scan-
ning Electron Microscopy. HRTEM images were recorded on a JEM-
2010 EX microscope operated at an accelerating voltage of 200 kV.
The samples were crushed in A.R. grade ethanol and the resulting
suspension was allowed to dry on carbon film supported on
copper grids. The iron content in the samples was determined
using a Jarrell-Ash 1100 Inductively Coupling Plasma mass spec-
trometry (ICP–MS). The samples were completely dissolved in hy-
drofluoric acid before analysis. Diffuse reflectance UV/Vis spectra
were recorded by a PerkinElmer Lambda 35 spectrophotometer in
the range of 200-800 nm with BaSO as reference. The XPS were
4
conducted on PHI 5000 Versa Probe X-ray photoelectron spectrom-
eter equipped with AlKa radiation (1486.6 eV). The C1s peak at
2
84.6 eV was used as the reference for binding energies.
Experimental Section
Catalytic test
Chemicals
The direct oxidation of phenol to dihydroxybenzene was per-
formed according to the following procedure. catalyst (0.02 g) and
phenol (0.4 g) were added to deionized water 10 mL in a 50 mL
CTAB, (C H (CH ) NBr, 99%) was purchased from J&K Chemical
1
6
33
3 3
Ltd. PEO PPO PEO (P123, M ꢀ5800) was purchased from
2
0
70
20
n
Sigma–Aldrich Co. TEOS (AR), ferrocene (C H Fe, AR) and phenol
1
0
10
flask at RT. Then, H O2 (0.26 mL (30 wt%) was dripped into the
2
(
C H OH, AR) were purchased from Sinopharm Chemical Reagent
6 5
above mixture with a magnetic stirring. The molar ratio of phenol/
H O was 3:2. After all H O was added, the reaction mixture was
Co., Ltd. Ammonia solution (25 wt%) and hydrogen peroxide
H O , 30 wt%) were purchased from Nanjing Chemical Reagent
2
2
2
2
(
2
2
stirred for an additional 5 h. After high speed centrifugation, the
supernatant was obtained and analyzed by an Agilent 1100 HPLC
equipped with a reversed phase C18 column. The used catalyst
was filtrated, dried and calcined at 5008C before being recycled.
Co., Ltd. Titanocene dichloride (C H Cl Ti, 97%) and zirconocene
1
0
10
2
dichloride (C H Cl Zr, 98%) were purchased from Aladdin Industri-
1
0
10
2
al Co.
Diffusion properties
Preparation
Phenol loading: Samples of 0.1 g representative Fe–NTRs (5R1 and
All chemicals were used as received without further purification.
5
R4) were added to phenol (15 mL, 0.2 g) aqueous solution at
Fe–NTRs with varied iron contents: In a typical run, CTAB, P123,
and ferrocene were dissolved in concentrated ammonia (100 mL,
room temperature. Then the test tubes were sealed to prevent
water from evaporation. After the dispersion of Fe–NTRs by ultra-
sound, the mixture was vigorously stirred for 12 h. The phenol-
loaded Fe–NTRs was obtained by centrifugation at 5000 rpm for
10 min, washed with 45 mL deionized water and then dried in air
at 358C. Phenol-loaded Fe–NTRs powder (0.06 g) was immersed
into 12 mL water. After ultrasonic treatment (1 min), the mixture
was stirred at a constant rate of 300 rpm. A 2 mL volume of the
suspension was extracted at given time intervals and centrifuged
at 10000 rpm for UV/Vis analysis at the wavelength of 210 nm. The
release profile was plotted using the average data from three inde-
pendent experiments.
NH , 25 wt%) with a stirring of 300 rpm at 408C. The concentra-
3
tions of CTAB and P123 in the mixed solution were 10.98 mm and
À1
1
.73”10 mm, respectively. After stirring for 1 h, TEOS (2 mL) was
dripped into the mixed solution. Then, the mixture was continu-
ously stirred for an additional 3 h. The resulting products were col-
lected by filtration and dried at RT. The organic components were
removed by calcination at 5008C for 5 h. By altering the input
amount of ferrocene, a series of Fe–NTRs with varied iron contents
were obtained. They are designated as xR: 5R, 10R, 15R, and 20R,
respectively, where x is 100 times of the calculated molar ratio of
ferrocene/silicon used in the synthesis gel.
Fe–NTRs with controllable sizes: Fe–NTRs with controllable sizes
were achieved by altering the input amount of P123 under the
above conditions. These samples were labeled as 5Rn: 5R1 (2.59”
Acknowledgements
À1
À1
À1
This work was supported by the National Natural Science Foun-
dations of China (No. 21276125, 21476108), the Program for Sci-
entific Innovation Research of College Graduates in Jiangsu Prov-
1
0
mm P123), 5R2 (1.73”10 mm P123), 5R3 (1.04”10 mm
À2
À2
P123), 5R4 (6.90”10 mm P123), and 5R5 (3.45”10 mm P123). n
from 1 to 5 represents an increasing length of Fe–NTRs.
ChemCatChem 2015, 7, 3855 – 3864
www.chemcatchem.org
3863
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
ince (No. CXZZ13_0443), the Key Program of National Natural
Science Foundation of China (No. 21136005) and the Project of
Priority Academic Program Development of Jiangsu Higher Edu-
cation Institutions (PAPD).
[16] J. Kim, C. Anand, S. N. Talapaneni, J. You, S. S. Aldeyab, E. Kim, A. Vinu,
Angew. Chem. Int. Ed. 2012, 51, 2859–2863; Angew. Chem. 2012, 124,
2913–2917.
[
[
17] a) X. L. Huang, L. L. Li, T. L. Liu, N. J. Hao, H. Y. Liu, D. Chen, F. Q. Tang,
ACS Nano 2011, 5, 5390–5399; b) T. Yu, A. Malugin, H. Ghandehari, ACS
Nano 2011, 5, 5717–5728.
18] a) Y. M. Liu, J. Xu, L. He, Y. Cao, H. Y. He, D. Y. Zhao, J. H. Zhuang, K. N.
Fan, J. Phys. Chem. C 2008, 112, 16575–16583; b) Z. Sun, B. Sun, M.
Qiao, J. Wei, Q. Yue, C. Wang, Y. Deng, S. Kaliaguine, D. Zhao, J. Am.
Chem. Soc. 2012, 134, 17653–17660; c) C. Liu, J. Li, J. Qi, J. Wang, R.
Luo, J. Shen, X. Sun, W. Han, L. Wang, ACS Appl. Mater. Interfaces 2014,
Keywords: iron
· mesoporous materials · nanotubes ·
surfactants · synthetic methods
[
[
1] B. Louis, G. Laugel, P. Pale, M. M. Pereira, ChemCatChem 2011, 3, 1263–
272.
2] a) Y. Yang, X. Liu, X. B. Li, J. Zhao, S. Y. Bai, J. Liu, Q. H. Yang, Angew.
Chem. Int. Ed. 2012, 51, 9164–9168; Angew. Chem. 2012, 124, 9298–
6, 13167–13173; d) K. Kandel, J. W. Anderegg, N. C. Nelson, U. Chaud-
1
hary, I. I. Slowing, J. Catal. 2014, 314, 142–148; e) M. Popova, A. Ristic,
K. Lazar, D. Maucec, M. Vassileva, N. N. Tusar, ChemCatChem 2013, 5,
986–993; f) H. Xin, J. Zhao, X. Li, J. Tang, Q. Yang, Microporous Mesopo-
9
302; b) Z. M. Cui, Z. Chen, C. Y. Cao, L. Jiang, W. G. Song, Chem.
rous Mater. 2014, 190, 54–62; g) Y. Zhao, Y. Zhang, J. Chen, J. Li, K. Liew,
M. R. Bin Nordin, ChemCatChem 2012, 4, 265–272.
Commun. 2013, 49, 2332–2334; c) S. Crossley, J. Faria, M. Shen, D. E. Re-
sasco, Science 2010, 327, 68–72; d) D. J. Cole-Hamilton, Science 2010,
[
[
19] H. Q. Wang, S. Y. Gao, J. Du, J. Chen, Y. Kong, J. Wang, Appl. Catal. A
3
27, 41–42.
2015, 504, 228–237.
[
3] a) S. J. Ding, J. S. Chen, G. G. Qi, X. N. Duan, Z. Y. Wang, E. P. Giannelis,
L. A. Archer, X. W. Lou, J. Am. Chem. Soc. 2011, 133, 21–23; b) G. Zhang,
J. Long, X. Wang, Z. Zhang, W. Dai, P. Liu, Z. Li, L. Wu, X. Fu, Langmuir
20] C. Wu, Y. Kong, F. Gao, Y. Wu, Y. N. Lu, J. Wang, L. Dong, Microporous
Mesoporous Mater. 2008, 113, 163–170.
[
[
21] Y. Yang, C. J. Sun, Y. Ren, S. J. Hao, D. Q. Jiang, Sci. Rep. 2014, 4, 4540.
22] C. Bosch-Navarro, E. Coronado, C. Marti-Gastaldo, P. Amoros, Nanoscale
2
2
010, 26, 1362–1371; c) A. Stein, B. J. Melde, R. C. Schroden, Adv. Mater.
000, 12, 1403–1419.
2
014, 6, 7981–7990.
23] K. Mori, T. Yamaguchi, S. Ikurumi, H. Yamashita, Chem. Commun. 2013,
9, 10468–10470.
24] T. Zhang, P. Gao, C. Gao, H. Yang, Y. Zhao, Chinese J. Catal. 2010, 31,
01–704.
[
[
[
[
[
[
4] H. Q. Wang, J. Chen, Y. Wu, X. Y. Xu, J. Wang, Y. Kong, RSC Adv. 2014, 4,
0832–50839.
5] C. X. Jiang, K. Hara, A. Fukuoka, Angew. Chem. Int. Ed. 2013, 52, 6265–
268; Angew. Chem. 2013, 125, 6385–6388.
6] A. I. Carrillo, E. Serrano, R. Luque, J. Garcia-Martinez, Appl. Catal. A 2013,
53, 383–390.
7] X. Liang, R. Yang, G. Li, C. Hu, Microporous Mesoporous Mater. 2013, 182,
2–72.
8] J. F. Lin, B. H. Zhao, Y. Cao, H. Xu, S. H. Ma, M. Y. Guo, D. R. Qiao, Y. Cao,
Appl. Catal. A 2014, 478, 175–185.
9] J. Fan, W. Q. Shui, P. Y. Yang, X. Y. Wang, Y. M. Xu, H. H. Wang, X. Chen,
D. Y. Zhao, Chem. Eur. J. 2005, 11, 5391–5396.
[
[
[
5
4
6
7
25] A. Popat, J. Liu, Q. Hu, M. Kennedy, B. Peters, G. Q. Lu, S. Z. Qiao, Nano-
scale 2012, 4, 970–975.
4
[
[
26] T. W. Kim, P. W. Chung, V. S. Y. Lin, Chem. Mater. 2010, 22, 5093–5104.
27] a) K. SchillØn, J. Jansson, D. Lof, T. Costa, J. Phys. Chem. B 2008, 112,
6
5551–5562; b) Y. Han, H. Yu, H. Du, W. Jiang, J. Am. Chem. Soc. 2010,
1
32, 1144–1150.
[
[
28] K. Suzuki, K. Ikari, H. Imai, J. Am. Chem. Soc. 2004, 126, 462–463.
29] R. Peng, C. M. Wu, J. Baltrusaitis, N. M. Dimitrijevic, T. Rajh, R. T. Koodali,
Chem. Commun. 2013, 49, 3221–3223.
[
[
10] a) J. J. Xie, Y. Y. Duan, S. A. Che, Adv. Funct. Mater. 2012, 22, 3784–3792;
b) V. V. Pushkarev, K. An, S. Alayoglu, S. K. Beaumont, G. A. Somorjai, J.
Catal. 2012, 292, 64–72; c) A. Zürner, J. Kirstein, M. Doblinger, C.
Brauchle, T. Bein, Nature 2007, 450, 705–708; d) A. Nisar, Y. Lu, J.
Zhuang, X. Wang, Angew. Chem. Int. Ed. 2011, 50, 3187–3192; Angew.
Chem. 2011, 123, 3245–3250.
11] a) T. Suteewong, H. Sai, R. Hovden, D. Muller, M. S. Bradbury, S. M.
Gruner, U. Wiesner, Science 2013, 340, 337–341; b) M. Egi, K. Sugiyama,
M. Saneto, R. Hanada, K. Kato, S. Akai, Angew. Chem. Int. Ed. 2013, 52,
[
[
30] A. Ramanathan, B. Subramaniam, R. Maheswari, U. Hanefeld, Micropo-
rous Mesoporous Mater. 2013, 167, 207–212.
31] a) L. Shi, X. Q. Zhu, Y. Su, W. Z. Weng, H. Feng, X. D. Yi, Z. X. Liu, H. L.
Wan, J. Catal. 2013, 307, 316–326; b) C. H. Cui, H. H. Li, J. W. Yu, M. R.
Gao, S. H. Yu, Angew. Chem. Int. Ed. 2010, 49, 9149–9152; Angew. Chem.
2
010, 122, 9335–9338.
32] Y. Jiang, K. Lin, Y. Zhang, J. Liu, G. Li, J. Sun, X. Xu, Appl. Catal. A 2012,
45, 172–179.
[
4
3
654–3658; Angew. Chem. 2013, 125, 3742–3746; c) V. S. Y. Lin, D. R.
[
[
33] N. S. Sanjini, S. Velmathi, RSC Adv. 2014, 4, 15381–15388.
34] H. J. Kim, J. E. Ahn, S. Haam, Y. G. Shul, S. Y. Song, T. Tatsumi, J. Mater.
Chem. 2006, 16, 1617–1621.
Radu, M. K. Han, W. H. Deng, S. Kuroki, B. H. Shanks, M. Pruski, J. Am.
Chem. Soc. 2002, 124, 9040–9041.
[
[
12] S. H. Wu, C. Y. Mou, H. P. Lin, Chem. Soc. Rev. 2013, 42, 3862–3875.
13] a) F. Tang, L. Li, D. Chen, Adv. Mater. 2012, 24, 1504–1534; b) S. Che, Z.
Liu, T. Ohsuna, K. Sakamoto, O. Terasaki, T. Tatsumi, Nature 2004, 429,
[
[
35] J. V. Coelho, M. S. Guedes, R. G. Prado, J. Tronto, J. D. Ardisson, M. C.
Pereira, L. C. A. Oliyeira, Appl. Catal. B 2014, 144, 792–799.
36] F. Hayashi, K. I. Ishizu, M. Iwamoto, Eur. J. Inorg. Chem. 2010, 2235–
2
81–284.
2243.
[
[
14] T. W. Kim, M. J. Kim, F. Kleitz, M. M. Nair, R. Guillet-Nicolas, K. E. Jeong,
H. J. Chae, C. U. Kim, S. Y. Jeong, ChemCatChem 2012, 4, 687–697.
15] a) A. N. Khlobystov, ACS Nano 2011, 5, 9306–9312; b) K. S. Ha, G. Kwak,
K. W. Jun, J. Hwang, J. Lee, Chem. Commun. 2013, 49, 5141–5143;
c) Z. Y. Zhao, J. F. Tian, Z. X. Wu, J. Liu, D. Y. Zhao, W. Shen, L. Z. He, J.
Mater. Chem. B 2013, 1, 4719–4722; d) K. Kageyama, J. Tamazawa, T.
Aida, Science 1999, 285, 2113–2115.
[
[
37] R. Kumar, P. Mukherjee, A. Bhaumik, Catal. Today 1999, 49, 185–191.
38] E. Gross, J. H. C. Liu, F. D. Toste, G. A. Somorjai, Nat. Chem. 2012, 4, 947–
952.
Received: August 3, 2015
Published online on November 3, 2015
ChemCatChem 2015, 7, 3855 – 3864
www.chemcatchem.org
3864
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim