An iron-based micropore-enriched silica catalyst:: In situ confining of Fe2O3 in the mesopores and its improved catalytic properties

Article abstract of DOI:10.1039/c6ra13146h

Surface exposed catalytic active species are thought to be responsible for overall catalytic activity and selectivity. In this paper, controllable contents of iron oxides were in situ introduced into the inner surface of anionic surfactant-templated mesop

Full text of DOI:10.1039/c6ra13146h

RSC Advances
View Article Online
View Journal | View Issue
PAPER
An iron-based micropore-enriched silica catalyst:
in situ confining of Fe2O3 in the mesopores and its
improved catalytic properties†
Cite this: RSC Adv., 2016, 6, 76064
a
a
a
a
Saifu Long,a Shijian Zhou,ab Fu Yang, Kangchao Lu, Tao Xi and Yan Kong
*
Surface exposed catalytic active species are thought to be responsible for overall catalytic activity and
selectivity. In this paper, controllable contents of iron oxides were in situ introduced into the inner
surface of anionic surfactant-templated mesoporous silica (AMS) by the metal-modified anionic
surfactant templating route, in which anionic micelles decorated with different amounts of Fe2+ were
direct utilized to assemble with organosilicate. The characterization results demonstrated that, after
thermal treatments, Fe2O3 species were in situ formed and highly-dispersed in the channels of AMS, and
the pore size was accordingly tailored to the microporous level (<2 nm). In each AMS catalyst, Fe2O3
showed better dispersion and stability compared with the post-impregnated method. Moreover, the
catalytic performances of phenol hydroxylation on the as-synthesized catalysts were significantly
enhanced, and the optimal catalyst (Fe/Si ¼ 3.5 wt%) gave a phenol conversion of 44.3% with
a selectivity of 82.6% to dihydroxybenzene. By comparison with some reference mesoporous silica
catalysts (0.15Fe/MCM-41 and 0.15Fe/SBA-15), the iron-based micropore-enriched silica catalyst (Fe/
AMS) gave an obviously higher selectivity to dihydroxybenzene. Finally, the optimal catalyst was
examined for at least 5 runs in phenol hydroxylation.
Received 20th May 2016
Accepted 3rd August 2016
DOI: 10.1039/c6ra13146h
www.rsc.org/advances
large pore volume, which are regarded as the best choice of
support to disperse the active sites.12 In this case, to introduce
1. Introduction
the active sites in the mesoporous silica materials, some tradi-
tional synthesis methods including thermal decomposition,13
ultrasonic treatment,14 chemical vapor deposition15 and organic
metal modication16 have been developed. Liu et al.17 intro-
duced well-dispersed metal oxides into the pore of SBA-15
through incipient wetness impregnation and thermal trans-
formation. Gu et al.18 reported that highly-dispersed copper
species were successfully coated onto the inner surface of the
rod-like mesoporous SBA-15 silica by using the hydrophobic
core of a surfactant micelle as a carrier. However, there are still
some disadvantages existed in these processes, such as
uncontrolled growth, uneven distribution, low loading and
complicated synthesis process, which would be detrimental to
their practical application.
On the other hand, even with high surface areas and well-
dispersed catalytic active sites, the large pore size of meso-
porous silica would make some large molecule byproducts form
in the pore channels and then diffuse across the channels,
leading to the low selectivity of target products.19 However, it is
noticeable that microporous materials, such as TS-1,20 Cu-ZSM-
5, Fe-ZSM-5 (ref. 21) etc., show better selectivity to the small
molecule products, which should be attributed to the shape
selectivity of microporous size.22,23 Nevertheless, microporous
materials generally exhibit the reduced surface area, which
could lead to the decreased number of exposed active sites.24
In recent years, extremely well dispersed metal oxides on solid
supports are of great interest for current and future sustainable
technologies such as their use as adsorbents,1 in medical
imaging,2 chemical sensing,3 electronic devices4 etc. Among
them, owing to the low price, rich abundance, environmentally-
friendly properties and remarkable catalytic performances,
iron-based catalysts have received great attention in phenol
hydroxylation,5 the Fenton reaction,6 Michael additions,7
isomerization8 etc. In addition, researchers found that the dis-
persity, microenvironment and surface interaction of iron
oxides on the supports have a great inuence on the catalytic
performance in these reactions.
For enhancing the dispersity of catalytic active sites, new
opportunities have gradually emerged since the discovery of
mesoporous silica materials.9–11 Mesoporous silica materials
with various pore structures (e.g., hexagonal, cubic, and lamella)
are well-known for their high surface area, regular structure and
aState Key Laboratory of Materials-Oriented Chemical Engineering, Nanjing Tech
University, Nanjing 210009, China. E-mail: kongy36@njtech.edu.cn; Fax: +86-025-
83587860; Tel: +86-025-83587860
bJiangsu National Synergetic Innovation Center for Advanced Materials (SICAM),
Nanjing Tech University, Nanjing 210009, Jiangsu, China
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c6ra13146h
76064 | RSC Adv., 2016, 6, 76064–76074
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
Therefore, in order to obtain the relevant advantages of both respectively. Finally, the obtained dried samples were calcined
ꢁ
ꢁ
mesoporous materials and microporous materials, great efforts at 550 C for 6 h in dry air stream with a heating rate of 1 C
have been devoted to modifying the pore surface of mesoporous minꢀ1. The nal samples are designated as xFe/AMS (where x
materials based on its large surface area and pore channels.25 represents the molar ratio of Fe2+/Na-Nlg, which is 0.05, 0.10,
Gao et al.26 prepared hierarchically porous ZSM-5/SBA-15 0.15 and 0.20). The pure AMS was prepared based on the above
materials, which possessed a high density of active sites by similar procedure.
connement of sub-nanosized Pt particles within small-pore
The comparative post-impregnated samples were prepared
zeolites and exhibited high catalytic performance for the by the classical wet impregnation method.27 In brief, 0.1 g of
hydrogenation of 1,3-butadiene and cyclooctadiene at room pre-synthesized pure AMS powder, MCM-41 powder and SBA-15
temperature. However, the hierarchically porous materials powder were respectively dispersed into the alcoholic solution
possess a bimodal pore size distribution and complicated of FeCl2$4H2O (2.53 wt%), and followed by rotary evaporation
synthesis process, which would be unacceptable in practical until complete dryness. The resultant various samples were
application. Therefore, the simple method for the synthesis of then dried at 60 ꢁC overnight and then calcined at 550 ꢁC for 6 h
silica catalysts with high surface area, micropore size distribu- in dry air stream with a heating rate of 1 ꢁC minꢀ1. The obtained
tion and well-dispersed catalytic active sites is still urgent for us products were designed as 0.15Fe/AMS(p), 0.15Fe/MCM-41(p)
to develop.
and 0.15Fe/SBA-15(p), accordingly.
Herein, we made Fe2+-modied anionic micelles as the new
template for in situ inducing Fe2+ into the channel of AMS
during the assemble process. The Fe2O3 species were directly
formed and highly-dispersed inside the channels with a stable
2.3. Characterizations
state by the thermal process, and the pore size was accordingly The UV-visible absorption (UV-vis) spectra of micelles were
tailored into the microporous level (<2 nm). Subsequently, all analyzed by a UV-vis spectrophotometer (BLV-GHX-V, Bilang
the prepared samples were efficient characterized by UV-vis, FT- Corporation, Shanghai, China).
IR, H2-TPR, XPS, XRD, N2 adsorption and TEM. The catalytic
Fourier transform infrared (FT-IR) spectra from the samples
performance and stability of catalysts were investigated by the were recorded in the range of 400–4000 cmꢀ1, with powders
reaction of phenol hydroxylation.
dispersed on the pure KBr on Bruker VECTOR22 resolution
(Bruker, Germany).
Diffused Reection UV-vis (DRUV-vis) spectra of samples
were obtained in the range of 200–800 nm by Lambda 950
spectrophotometer (PerkinElmer, Waltham, MA, USA).
Hydrogen temperature programmed reduction (H2-TPR)
measurements were performed utilizing a xed-bed reactor
under a ow of 10% H2$Ar gas mixture and a heating rate of 10
ꢁC minꢀ1 from room temperature to 750 ꢁC. Before the TPR
analysis, the carbonates and hydrates impurities were removed
by owing argon over the catalyst at a ow rate of 30 mL minꢀ1
at 300 ꢁC for 1 h and the system was then cooled to room
temperature. The amount of H2 uptake during the reduction
was measured continuously using a thermal conductivity
detector (TCD).
2. Experimental section
2.1. Materials
Sodium N-lauroylglutamate (Na-Nlg, >93%) was purchased from
PengYuan (GuangZhou) Chemical Co., Ltd. Ferrous chloride
tetrahydrate (FeCl2$4H2O) was purchased from Aladdin. 3-
Aminopropyltriethoxysilane (APTES) was purchased from
TCI(Shanghai) Development Co., Ltd. Sodium hydroxide
(NaOH), phenol, hydrogen peroxide (H2O2, 30 wt%) and tet-
raethoxysilane (TEOS) were purchased from Sinopharm Chem-
ical Reagent Co., Ltd. All the reagents except for H2O2 and Na-
Nlg were analytical grade.
The X-ray photoelectron spectra (XPS) were performed on
a PHI 5000 Versa Probe X-ray photoelectron spectrometer
2.2. Preparation of catalysts
The typical synthesized process was depicted as follows: 3.70 g equipped with Al Ka radiation (1486.6 eV) (Ulvac-PHI, Chiga-
of Na-Nlg was dissolved in a mixed solution containing 230 mL saki, Japan). The C 1s peak at 284.6 eV was used as the reference
of deionized water and 30 mL of NaOH solution (0.5 mol Lꢀ1
)
for binding energies.
under vigorous stirring at room temperature. The pH of
The powder X-ray diffraction (XRD) patterns of the samples
resulting solution was adjusted to 8 with dilute hydrochloric were collected with Smartlab TM 9 KW (Rigaku Corporation,
acid (1 mol Lꢀ1). Then, different molar amounts of FeCl2$4H2O Tokyo, Japan) equipped with a rotating anode and Cu Ka radi-
were added in above obtained solutions, respectively. Aer that, ation (l ¼ 0.154178 nm).
5.24 mL of APTES was added in the above mixture under
The N2 (77.4 K) adsorption–desorption measurements were
vigorous stirring for 30 s to form a tawny suspension. Subse- carried out with BELSORP-MINI volumetric adsorption analyzer
quently, 8.91 mL of TEOS was added into the suspension with (BEL Japan, Osaka, Japan) in a relative pressure range P/P0 from
continuous and intensive stirring for 10 min. Next, the resulting 0.01 to 0.99. The annealed samples were outgassed in vacuum at
mixture was aged for 48 h at 42 ꢁC. A nal molar composition of 150 ꢁC for 5 h before measurements. The specic surface areas
mixture was 1.0 (Na-Nlg) : x (FeCl2$4H2O) : 3 (APTES) : 4 and distribution of pore size were calculated using the Bru-
(TEOS) : 1750 (H2O). The synthesized samples were centrifuged nauer–Emmet–Teller (BET) and Barrett–Joyner–Halenda (BJH)
and washed with deionized water and ethanol for three times, methods, respectively.
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 76064–76074 | 76065

View Article Online
RSC Advances
Paper
High-resolution transmission electron microscopy (HRTEM)
n(H2O2) ¼ r ꢂ V ꢂ wt%/M ¼ 1.13 g mLꢀ1 ꢂ 2.1 mL ꢂ 30%/
images were recorded on a JEM-2010 EX microscope (JEOL, 34.01 ¼ 0.021 where n, r, V, wt% and M represent the density,
Tokyo, Japan), which was operated at an accelerating voltage of the volume, the weight concentration and the molecular weight
200 kV. The samples were crushed in A.R. grade ethanol and the of H2O2, respectively.
resulting suspension was allowed to dry on carbon lm sup-
The molar ratio of n(H2O2)/n(phenol) is about 2.
ported on copper grids.
In order to investigate the recycling properties of 0.15Fe/AMS
Field emission scanning electron microscopy (FE-SEM) was and 0.15Fe/AMS(p), the two catalysts were separated from the
performed on a Hitachi S4800 Field Emission Scanning Elec- reaction mixture by centrifugation and used again in a fresh
tron Microscopy (Hitachi, Tokyo, Japan).
reaction respectively.
The iron content in the samples was determined using a PE
Optima 2000DV (PerkinElmer, Waltham, MA, USA) Inductively
Coupling Plasma emission spectrometer (ICP). The samples
were completely dissolved in hydrouoric acid before analysis.
3. Results and discussion
3.1 Illustration for the synthetic method
The micropore-enriched silica composite with Fe2O3 inside the
channel was prepared by a modied anionic micelle templating
assemble route (SꢀX+Iꢀ). Typically, in this pathway (Scheme 1),
2.4. Catalysis tests
The phenol hydroxylation was carried out in 50 mL of three- a novel anionic template Na-Nlg exhibits excellent chelating
necked round-bottom ask equipped with a magnetic stirrer capability due to a large number of carboxyl functional groups
and reux condenser. The reaction mixture consists of 0.22 g of and is able to self-assemble to form regular anionic micelles in
catalyst, 4.29 g of phenol, 0.36 g of sodium dodecyl sulfate and the proper solutions. During the assemble process, Fe2+ would
2.1 mL of H2O2 (30 wt%), and the reaction was carried out under be anchored on the micelles by coordination with carboxyl
reux at room temperature for 2 h. Then, the product distri- groups as metal-modied micelle template (Na-Nlg, Sꢀ). In the
butions were analyzed by an HPLC 100 (Wufeng Scientic following step, the head of metal-modied micelle template
Instruments Co., Ltd, Shanghai, China). The conversion of could match amino groups of the co-structure directing agent
reactant and selectivity of products were calculated by external (CSDA) (APS, X+). Then, silicate oligomers (Iꢀ) deposited on
standard method. In addition, the molar ratio of n(phenol)/ CSDA matched functional micelles by the co-hydrolysis with
n(H2O2) is dened as follows:
CSDA. Finally, aer calcination, Fe2O3 species were in situ
n(phenol) ¼ m/M ¼ 4.29/94.11 ¼ 0.045 where n, m and M formed and highly-dispersed inside the channels of AMS
refer to the molar number, the mass and the molecular weight material. Obviously, the formation of metal-modied anionic
of phenol, respectively.
template would be the key factor during this synthesis process.
Scheme 1 Metal-modified micelle templating method (SꢀX+Iꢀ) for the synthesis of silica composite.
76066 | RSC Adv., 2016, 6, 76064–76074
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
vibrations of Si–O–Si, respectively, while another absorbed peak
at ca. 961 cmꢀ1 is typically corresponding to the stretching
vibration of Si–OH surface groups, these bands are the feature
absorption of silica. As for the uncalcined As-pure AMS and As-
0.15Fe/AMS, several characteristic peaks appear at ca. 2923,
1588 and 1402 cmꢀ1 ascribed to –CH2–, C]N and C–H of the
surfactant. Moreover, as compared to the As-pure AMS, an
obvious shi from 1558 cmꢀ1 to 1581 cmꢀ1 for the absorbed
peak of C]N is observed in the sample of As-0.15Fe/AMS,29
which is the strong evidence to favor the coordination of Fe2+
with the group of C]N on the amino acid head.
3.3. Characterizations for Fe2O3 state
Fig. 3 displays the wide-angle XRD patterns of series samples
directly synthesized and post-impregnated sample of 0.15Fe/
AMS(p). All the curves exhibit a broad diffraction peak from
20ꢁ to 30ꢁ ascribed to amorphous silica. Specically, as for the
post-impregnated sample of 0.15Fe/AMS(p), several diffraction
peaks at 24ꢁ, 33ꢁ, 36ꢁ, 49ꢁ, 54ꢁ, 62ꢁ and 64ꢁ are detected and
indexed to (012), (104), (110), (024), (116), (214) and (300) crystal
faces of Fe2O3 (JCPDS 88-0315). However, no diffraction peak of
Fe2O3 crystalline phase is detected in the curves of series xFe/
AMS samples, suggesting that high-dispersion and amor-
phous status of Fe2O3 in these samples.
For better understanding the nature and the coordination
environment of iron species in different samples, Diffused
Reection UV-vis spectra (DRUV-vis) of xFe/AMS were examined
and their results are presented in Fig. 4. As observation, the
curves of xFe/AMS are divided into four sub-curves with maxima
at ca. 218, 258, 360, and 503 nm. Typically, the absorption band
at 218 nm is attributed to the characteristic absorption of silica.
The second absorption band at 258 nm is associated with the
dp–pp charge transfer (CT) between Fe and surface O of silica,30
indicating that some iron atoms may directly link to siliceous
matrix. The result should be due to the special templating
method in which Fe2+ are enriched on the anionic micelles and
then directly match structure directing agents or siliceous
oligomers, resulting in the presence of surface-linked Fe species
aer calcination. For comparison, a DRUV-vis spectra of pure
Fig. 1 UV-vis spectra and the digital photo (inset) of different solutions
of (a) FeCl2 solution, (b) NLG solution and (c) mixed solution of NLG
and FeCl2.
3.2 Characterizations for the interaction between Na-Nlg
and Fe2+
In order to verify the formation of metal-modied anionic
micelle template, the chelating evidence of Fe2+ on anionic
surfactant of Na-Nlg is examined by UV-vis spectroscopy. The
UV-vis spectra and digital photos of different solutions of FeCl2,
Na-Nlg and their mixture are shown in Fig. 1. As compared with
single FeCl2 solution and single Na-Nlg solution, the mixture
solution exhibits a newly emerging band centered at ca. 336 nm,
which indicates the existence of interaction between Na-Nlg and
Fe2+.28 Furthermore, in the given digital photo, the mixture (c)
displays an obvious change in contrast to the colorless single
FeCl2 solution (a) and single Na-Nlg solution (b), which strongly
supports the coordination of Fe2+ on the Na-Nlg.
For further analyzing the interaction between introduced
Fe2+ and surfactant of Nlg-Na, the FT-IR spectra of uncalcined
As-pure AMS, uncalcined As-0.15Fe/AMS and 0.15Fe/AMS are
also examined and shown in Fig. 2. As observed, for the sample
of 0.15Fe/AMS, the peaks at ca. 1081 cmꢀ1 and 791 cmꢀ1 are
attributed to the asymmetric and symmetric stretching
Fig. 2 FT-IR spectra of 0.15Fe/AMS, As-pure AMS and As-0.15Fe/AMS. Fig. 3 Wide-angle XRD patterns of xFe/AMS and 0.15Fe/AMS(p).
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 76064–76074 | 76067

View Article Online
RSC Advances
Paper
Fig. 4 DRUV-vis spectra of xFe/AMS and pure iron oxides.
iron oxides has been presented in Fig. 4. As expected, the curve when the surface linked Fe atoms reach saturation, the exces-
shows little or no absorbance in the range 200–400 nm, which sive Fe atoms would aggregate into clusters or particles. In this
demonstrates the absence of surface-link Fe species. However, case, the presence of other two absorption bands at ca. 360 nm
Table 1 State and relative distribution of iron species in xFe/AMSa
Surface-linked Fe
% area
Fe2O3 clusters
% area
Fe2O3 particle
% area
Samples
Feb (wt%)
Fe (wt%)
Fe (wt%)
Fe (wt%)
0.05Fe/AMS
0.10Fe/AMS
0.15Fe/AMS
0.20Fe/AMS
0.92
1.62
2.53
2.89
12.5
15.1
13.4
13.3
0.12
0.24
0.34
0.38
75.7
75.4
78.1
73.7
0.69
1.22
2.01
2.12
11.8
10.4
8.5
0.11
0.16
0.18
0.36
13.0
a
It is calculated from DRUV-vis spectra of xFe/AMS. b It is determined by ICP-AES.
76068 | RSC Adv., 2016, 6, 76064–76074
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
Fig. 5 H2-TPR profiles of xFe/AMS and 0.15Fe/AMS(p).
Fig. 7 Low-angle XRD patterns of xFe/AMS and pure AMS.
and 503 nm are attributed to octahedral Fe3+ in oligomeric
Fe2O3 clusters and larger Fe2O3 particles, respectively, as can be
seen in the spectra of pure iron oxides.31–33 In order to analyze
these results with great certainty, the further detailed infor-
mation for the metal content and status distribution of different
iron species are summarized in Table 1. As observed in Table 1,
from the calculated results of relative percentage of band areas,
the more incorporation of Fe2O3 does not lead to the consecu-
tive enhancement of absorption of oligomeric Fe2O3 clusters,
which indicates the saturation of Fe2O3 clusters on the pore
wall. However, the dominated absorbed peak of oligomeric
Fe2O3 clusters indicates that most of the Fe2O3 species are
existed as the oligomeric Fe2O3 clusters status in all the
samples. These results demonstrate that Fe2O3 are well-
dispersed on the pore wall with ne oligomeric Fe2O3 clusters.
In order to verify the chemical environments of iron oxide of
xFe/AMS and 0.15Fe/AMS(p), H2-TPR experiments are performed
and their results are displayed in Fig. 5. It is known that supported
metal oxides on inert supports have a complicated reduction
behavior due to difference of particle size, dispersity and surface
interaction. In this case, the reduction prole of the iron species
loaded on the support would be the result of the competition
between two factors: the faster reduction of smaller size particles
due to the increase in the surface/volume ratio, and the slower
reduction of these smaller particles owing to the higher interac-
tion with the support.32 As shown in Fig. 5, the curve of 0.15Fe/
AMS(p) exhibits two reduction peaks with the maximum
temperature at ca. 382 ꢁC and 646 ꢁC, which can be related to the
Fe3+ / Fe3+/2+ and Fe3+/2+ / Fe0 processes, respectively.34
Fig. 6 XPS spectra of the sample 0.15Fe/AMS: (a) Si 2p, (b) O 1s and (c) Fe 2p.
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 76064–76074 | 76069

View Article Online
RSC Advances
Paper
content, the Fe2+ species are stabilized on the pore wall with high
resistance to the reduction to metallic state.32 On the other hand,
the enhanced intensity of reduction peak (Fe3+ / Fe2+) suggests
gradually increasing Fe2O3 clusters are embedded into the
channel of mesoporous silica with the increase introduction of
iron. Therefore, the TPR results further conrm high dispersion
of Fe2O3 clusters in the channel of mesoporous silica, which is in
good agreement with DRUV-vis results.
The surface composition and chemical state of 0.15Fe/AMS
sample are further analyzed by XPS spectra. The Si 2p spectrum
of 0.15Fe/AMS is shown in Fig. 6(a). This spectrum is composed
of a single peak centered at ca. 103.1 eV, which is characteristic of
silicates. The spectrum of O 1s of 0.15Fe/AMS (Fig. 6(b)) is also
composed of a single peak centered at ca. 532.6 eV corresponding
to oxygen in the SiO2. While other peaks of oxygen attributed to
iron oxides such as Fe2O3 (at ca. 530 eV) are not observed. The
result is probably due to that their low signal is masked by the
intense signal corresponding to the oxygen of the SiO2.36 The Fe
2p spectrum of 0.15Fe/AMS is shown in Fig. 6(c). The binding
energy of Fe 2p3/2 at 711.9 eV, accompanied by a satellite peak at
720 eV, and the binding energy of 2p1/2 at 724.3 eV, are indicative
of the presence of Fe3+ in iron oxide environment.37 According to
the literature, the binding energy value of pure iron oxide pres-
ents at 710.6 eV, but the binding energy value of Fe 2p3/2 of
0.15Fe/AMS shis toward a higher value (711.9 eV), indicating
that Fe2O3 have a strong interaction with silica wall.38 Meanwhile,
the presence of the peak at 715.9 eV between the peak of Fe 2p3/2
and its satellite peak could be attributed to Fe linked to surface O
atoms on the silica wall,39 which is consistent with DRUV-vis
result. To explain this phenomenon, as compared with Fe3+ in
iron oxide environment, a decrease in the crystal eld energy of
the Fe3+ ions located at the surface should be the reason for the
presence of this peak. Thus, with a decrease in the coordination,
the surface-linked Fe3+ would be surrounded by a lower electron
density, which requires the binding energy (715.9 eV) to be used
to produce a photoelectron. Taken together, these results
demonstrate Fe2O3 species are highly-dispersed and stabilized in
the channel of silica in the form of clusters.
Fig. 8 N2 adsorption/desorption isotherms (a) and pore size distri-
bution (b) of xFe/AMS and pure AMS.
However, as for the directly synthesized samples, the reduction
peaks attributed to Fe3+ / Fe3+/2+ move towards lower tempera-
ture of 361 ꢁC. This lower reduction temperature should be
attributed to the higher surface/volume ratio in the smaller metal
particles,35 indicating the presence of highly-dispersed Fe2O3
clusters in these samples. Moreover, it is interesting to note that
3.4. Characterizations for the structure of samples
the sample 0.05Fe/AMS shows a reduction peak at ca. 519 ꢁC, The presence of Fe2O3 clusters on the pore walls of mesoporous
which is assigned to the Fe3+/2+ / Fe2+ process. At low iron silica would inuence mesostructure ordering and pore size of
Table 2 Textural parameters of all the samples
Fe/Sia (wt%)
SBETb (m2 gꢀ1
)
Vp (m3 gꢀ1
)
Dp (nm)
a0d (nm)
d100 (nm)
dwe (nm)
c
Samples
Pure AMS
0.05Fe/AMS
0.10Fe/AMS
0.15Fe/AMS
—
312
518
499
434
399
156
490
872
0.32
0.23
0.20
0.19
0.15
0.11
1.39
0.58
3.29
1.72
1.72
1.72
1.67
2.60
5.40
2.40
5.42
4.21
4.25
4.29
4.56
4.32
10.00
4.03
4.69
3.65
3.68
3.72
3.95
3.74
8.66
3.49
2.13
2.49
2.53
2.57
2.89
1.72
4.60
1.63
0.92
1.62
2.53
2.89
3.50
3.50
3.50
0.20Fe/AMS
0.15Fe/AMS(p)
0.15Fe/SBA-15(p)
0.15Fe/MCM-41(p)
a
It is determined by ICP-AES. b It is calculated by BET method. c It is calculated by BJH method using desorption data. d a0 ¼ 2 ꢂ d100/O3, which
e
represents unit cell parameter. dw ¼ a0 ꢀ Dp, which represents pore wall thickness.
76070 | RSC Adv., 2016, 6, 76064–76074
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
Fig. 9 Representative TEM images of 0.15Fe/AMS (a) and EDX pattern of selected area (b).
xFe/AMS samples. As observed in Fig. 7, pure AMS exhibits the specic surface areas for the samples decrease slightly
a (100) diffraction peak at 2q z 1.9ꢁ with (110) and (200) from 518 to 399 m2 gꢀ1, suggesting that iron species incor-
diffraction peaks. The (100) diffraction peak for the pure AMS porated into the pore surface of mesoporous materials do not
exhibits a stronger intensity and low angle of peak position, damage the mesostructure completely. In addition, it is very
indicating the well-ordered mesostructure and large pore size of interesting that the specic surface areas of xFe/AMS is obvi-
pure AMS. However, series of xFe/AMS exhibit a weak intensity ously higher than that of pure AMS, which may be due to the
(100) diffraction peak at ca. 2q z 2.4ꢁ, which suggests the increase number of pores under the same pore volume.
presence of mesostructure in xFe/AMS. The result is due to that Besides, the pore size distribution in Fig. 8(b) shows that the
xFe/AMS are prepared by modifying the pore surface of meso- most probable pore size of xFe/AMS is at ca. 1.70 nm and the
porous AMS. Additionally, the intensity of (100) diffraction peak more incorporation of Fe2O3 do not lead to the gradual drop of
of xFe/AMS is approximately invariable with the increase of iron pore size, which is due to that the aggregated Fe2O3 particles
content, indicating that the gradual formation of Fe2O3 clusters migrate out of the channel of silica during the thermal treat-
in the channel of mesoporous silica have little effect on the ment process. In addition, it is noticeable that pore wall
regularity of mesostructure.
thickness (dw) increases from 2.13 nm to 2.89 nm with the
The textural characteristics of the prepared samples are increment of iron in the samples. It suggests that the
displayed in N2-adsorption/desorption isotherms (Fig. 8(a)). As increasing thickness may be assigned to the occupation of
observed from the curve, the pure AMS shows the typical IV Fe2O3 species on the pore wall.
isotherm with an obvious capillary condensation step, indi-
The structural characteristics of samples and distribution of
cating the mesostructure of pure AMS. However, the isotherms iron species in the channel are further examined by high-
of xFe/AMS samples exhibit type I-like curve, which are typical resolution transmission electron microscopy (TEM). The
for microporous materials. Generally, a type I isotherm would ordered and parallel channels are observed in the representa-
have to level off below a relative pressure of about 0.1 for the tive TEM images of 0.15Fe/AMS (Fig. 9(a)). The measured pore
materials to be exclusively microporous. However, the diameter from the TEM images is ca. 1.68 nm, which is close to
isotherms for xFe/AMS samples do not level off below the N2 adsorption/desorption results. This result reveals that the
relative pressure of about 0.1, the samples are likely to exhibit porous structure is not destroyed aer in situ introducing Fe2O3
an appreciable number of mesopores with pore sizes close to species. More importantly, visual Fe2O3 particles fail to be
the micropore range.40,41 The presence of micropore is attrib- detected from TEM images, however, as shown in Fig. 9(b), the
uted to the coverage of Fe2O3 clusters on the pore walls of energy-dispersive X-ray (EDX) analysis suggests that 0.15Fe/AMS
mesoporous silica. On the other hand, the more detailed contains Si, O and Fe elements. In this case, Fe2O3 should be
textural parameters of samples are further summarized in highly dispersed in the silica composite, which is consistent
Table 2. With the contents of irons increased from 0.05 to 0.20, with XPS and N2 adsorption/desorption results.
Fig. 10 (a) SEM image of 0.15Fe/AMS sample and the corresponding element mappings of the sample: (b) Si, (c) O, (d) Fe.
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 76064–76074 | 76071

View Article Online
RSC Advances
Paper
Table 3 The catalytic activities of all the samples for phenol hydroxylationa
Samples
Pore sizeb (nm)
Fe/Sic (wt%)
Fe/Sid (wt%)
X(Ph)e/%
S(CAT)f/%
S(HQ)g/%
Ref.
0.05Fe/AMS
0.1Fe/AMS
1.72
1.72
1.72
1.67
2.40
5.40
2.60
0.54
0.50
1.11
2.23
3.50
4.67
3.50
3.50
3.50
—
0.92
1.62
2.53
2.89
2.52
2.61
2.50
—
29.8
34.8
44.3
43.6
38.6
40.1
36.7
29
44.3
49.4
52.2
50.1
32.8
26.1
31.4
50.2
60.5
28.4
29.7
30.4
30.3
18.6
19.4
19.6
49.8
39.5
This work
0.15Fe/AMS
0.20Fe/AMS
0.15Fe/MCM-41(p)h
0.15Fe/SBA-15(p)
0.15Fe/AMS(p)
TS-1
44
21
Fe-ZSM-5
2.00
1.54
32.9
a
Reaction conditions: molar ratio, phenol/H2O2z2 : 1, time, 2 h, reaction temperature, room temperature. b It is calculated by BJH method using
desorption data. c The initial Fe/Si mass ratios in the synthesis gel. d Fe/Si mass ratios are determined by ICP-AES. e The conversion of phenol. f The
selectivity of catechol. g The selectivity of hydroquinone. h The post synthesized sample.
To further understand the distribution of iron species on the that extra Fe2O3 can accelerate the decomposition of H2O2 and
surface of silica, elemental mapping from SEM measurement is yield high concentration of HO$, leading to the excessive
conducted and the results are shown in Fig. 10. As observed, oxidation.42 Therefore, the 0.15Fe/AMS is regarded as optimal
silica composite prepared by the metal-modied micelle tem- catalyst in phenol hydroxylation, which exhibits high catalytic
plating method contains elements of Si, O, and Fe. Moreover, it activities, giving the phenol conversion at 44.3% with selectivity
is also clear that the iron element is uniformly dispersed on of 82.6% to hydroquinone.
silica composite, which could conrm the high dispersion of
To better understanding the catalytic performances of xFe/
iron species in the silica composite and is consistent with XPS AMS, catalytic performance of 0.15Fe/MCM-41(p), 0.15Fe/SBA-
and EDX results.
15(p), 0.15Fe/AMS(p), reported TS-1 and Fe-ZSM-5 are also
investigated, and their catalytic results are summarized in Table
3. As observed in Table 3, as compared with 0.15Fe/MCM-41(p),
0.15Fe/SBA-15(p) and 0.15Fe/AMS(p), 0.15Fe/AMS shows
3.5. Catalytic tests
In the present work, the catalytic performance of xFe/AMS is
evaluated by direct oxidation of phenol at room temperature
and the corresponding catalytic results are summarized in
Table 3. According to recent reports,5 the pure AMS fails to show
any activity and the bulk Fe2O3 performed poor activity in this
reaction. While in Table 3, series samples of xFe/AMS exhibit
high conversion and selectivity, which indicates that highly-
dispersed Fe2O3 clusters could be more effective as an active
site. In addition, by the gradually growth of the Fe/Si mass ratio
from 0.92 to 2.89, the catalytic performances of 0.05Fe/AMS,
0.10Fe/AMS and 0.15Fe/AMS are improved correspondingly.
According to the DRUV-vis and XPS results, this should be
attributed to the increased number of exposed active sites in
each catalyst. However, the catalyst of 0.20Fe/AMS gives
a decreasing trend for catalytic activities. The result is due to
Scheme 2 Depiction for the phenol hydroxylation catalyzed by xFe/ Fig. 11 Recycling test of 0.15Fe/AMS and 0.15Fe/AMS(p): (a) the
AMS catalysts.
conversion of phenol and (b) mass ratios of Fe/Si.
76072 | RSC Adv., 2016, 6, 76064–76074
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
a highest conversion and selectivity. Base on the results of exhibited highest catalytic activities in phenol hydroxylation,
textual properties, the high conversion should be due to the receiving the phenol conversion of 44.3% with hydroquinone
presence of highly-dispersed Fe2O3 clusters in 0.15Fe/AMS, selectivity of 82.6%. The extremely-dispersed catalytic active
which results from the metal-modied anionic surfactant sites and massive micropores should be responsible for high
templating route. On the other hand, it is worth noting that the catalytic activities. In addition, due to better chemical stability
order of pore size is Fe-ZSM-5 (0.5 nm) < TS-1 (0.54 nm) < of the catalyst, 0.15Fe/AMS exhibited superior catalytic stability.
0.15Fe/AMS (1.72 nm) < 0.15Fe/MCM-41(p) (2.40 nm) < 0.15Fe/
AMS(p) (2.60 nm) < 0.15Fe/SBA-15(p) (5.40 nm). And the
molecular size of tar is much bigger than that of benzoquinone
Acknowledgements
(0.54 nm) and dihydroxybenzene (0.56 nm). Therefore, for the This work was supported by the National Natural Science
microporous materials of Fe-ZSM-5 and TS-1, due to the shape Foundations of China (No. 21276125, 21476108, 20876077) and
selectivity,43 the relatively large molecule by-products of tar the Project of Priority Academic Program Development of
could hardly be formed in the channels, leading to the high Jiangsu Higher Education Institutions (PAPD).
selectivity of dihydroxybenzene. Similarly, compared with
0.15Fe/MCM-41(p), 0.15Fe/SBA-15(p) and 0.15Fe/AMS(p), the
micropore-enriched 0.15Fe/AMS should exhibit a high selec-
Notes and references
tivity of dihydroxybenzene. Nevertheless, even with high selec-
tivity to dihydroxybenzene, the low surface area of Fe-ZSM-5 and
TS-1 would lead to the decreased number of exposed active sites
in the catalysts.24 Besides, their extremely small micro-channel
could hinder the reactant molecules diffusing across the
channel. In contrast, the prepared micropore-enriched catalysts
of xFe-AMS show larger pore channels, which can ensure
that more of reactants diffuse into the channels. Therefore,
compared with Fe-ZSM-5 and TS-1, xFe/AMS catalysts exhibit
a high conversion in phenol hydroxylation.
In summary, as depicted in Scheme 2, the prepared
micropore-enriched Fe/AMS catalysts show better dispersion of
iron oxides, which would increase the number of exposed active
sites. Moreover, the presence of more protable pore channels
of Fe/AMS catalysts could make more of reactants diffuse into
the channels. Consequently, compared with other catalysts,
micropore-enriched Fe/AMS catalysts exhibit a high conversion
and selectivity in phenol hydroxylation.
1 J. Kou, C. Lu, W. Sun, L. Zhang and Z. Xu, ACS Sustainable
Chem. Eng., 2015, 3, 3053–3061.
2 C. Corot, P. Robert, J. M. Idee and M. Port, Adv. Drug Delivery
Rev., 2006, 58, 1471–1504.
3 X. Liu, R. Wang, T. Zhang, Y. He, J. Tu and X. Li, Sens.
Actuators, B, 2010, 150, 442–448.
4 K. Zilberberg, J. Meyer and T. Riedl, J. Mater. Chem. C, 2013,
1, 4796.
5 C. Wu, Y. Kong, F. Gao, Y. Wu, Y. Lu, J. Wang and L. Dong,
Microporous Mesoporous Mater., 2008, 113, 163–170.
6 L. Guo, F. Chen, X. Fan, W. Cai and J. Zhang, Appl. Catal., B,
2010, 96, 162–168.
7 S. Sobhani, M. Bazrafshan, A. A. Delluei and Z. P. Parizi, Appl.
Catal., A, 2013, 454, 145–151.
8 X. Xu, T. Liu, P. Xie, Y. Yue, C. Miao, W. Hua and Z. Gao,
Catal. Commun., 2014, 54, 77–80.
9 F. Yang, S. Gao, C. Xiong, S. Long, X. Li, T. Xi and Y. Kong,
RSC Adv., 2015, 5, 72099–72106.
In addition, to evaluate the catalytic stability of optimal 10 H. Wang, X. Li, C. Xiong, S. Gao, J. Wang and Y. Kong,
catalyst, the 0.15Fe/AMS and 0.15Fe/AMS(p) catalysts are tested
ChemCatChem, 2015, 7, 3855–3864.
by running the reaction successively ve times. As showed in 11 J. Fan, X. Jiang, H. Min, D. Li, X. Ran, L. Zou, Y. Sun, W. Li,
Fig. 11 and Table S1,† aer recycled in 5 times, 0.15Fe/
AMS(p) exhibits the low iron concentration and catalytic
J. Yang, W. Teng, G. Li and D. Zhao, J. Mater. Chem. A, 2014,
2, 10654.
´
performances. But for 0.15Fe/AMS catalyst, the iron concen- 12 L. Chmielarz, P. Kustrowski, R. Dziembaj, P. Cool and
tration is still as high as 1.94 wt% and the catalytic phenol E. F. Vansant, Appl. Catal., B, 2006, 62, 369–380.
conversion is above 37.7%. This result should be attributed to 13 Y. Li, J. Sun, J. Wang, L. Zhang, L. Gao and X. Wu, J. Porous
the strong interact between Fe2O3 species and pore wall of
mesoporous silica and is in accord with the above result.
Mater., 2011, 19, 389–396.
14 S. Zhu, Z. Zhou, D. Zhang and H. Wang, Microporous
Mesoporous Mater., 2006, 95, 257–264.
15 T. Hisatomi, M. Otani, K. Nakajima, K. Teramura, Y. Kako,
D. Lu, T. Takata, J. N. Kondo and K. Domen, Chem. Mater.,
2010, 22, 3854–3861.
4. Conclusions
In this work, controllable contents of surface iron oxides were in
situ introduced into AMS by metal-modied anionic surfactant 16 A.-L. Radu, C. Damian, V. Fruth, T.-V. Iordache, A. Zaharia,
templating route. The structure properties, iron oxide status,
chemical stability and catalytic properties of different catalysts
H. Iovu and A. Sarbu, Microporous Mesoporous Mater.,
2014, 198, 281–290.
for phenol hydroxylation have been studied. The results indi- 17 C.-H. Liu, N.-C. Lai, S.-C. Liou, M.-W. Chu, C.-H. Chen and
cated that Fe2O3 species from template-induced method show
better dispersion and stability compared to that of post-
C.-M. Yang, Microporous Mesoporous Mater., 2013, 179, 40–
47.
impregnated one. As unexpected, massive micropores (ca. 1.7 18 J. Gu, Y. Huang, S. P. Elangovan, Y. Li, W. Zhao, I. Toshio,
nm) were directly formed due to the occupation of irons species
in mesochannel. In this case, the optimal 0.15Fe/AMS catalyst
Y. Yamazaki and J. Shi, J. Phys. Chem. C, 2011, 115, 21211–
21217.
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 76064–76074 | 76073

View Article Online
RSC Advances
Paper
´
19 S. Kulawong, S. Prayoonpokarach, A. Neramittagapong and 32 N. I. Cuello, V. R. Elıas, C. E. Rodriguez Torres, M. E. Crivello,
J. Wittayakun, J. Ind. Eng. Chem., 2011, 17, 346–351.
20 T. Yokoi, P. Wu, Y. Guo and T. Tatsumi, Catal. Commun.,
2003, 4, 11–15.
21 A. L. Villa, C. A. Caro and C. M. d. Correa, J. Mol. Catal. A:
Chem., 2005, 228, 233–240.
M. I. Oliva and G. A. Eimer, Microporous Mesoporous Mater.,
2015, 203, 106–115.
33 Q. Zhang, Y. Li, D. An and Y. Wang, Appl. Catal., A, 2009, 356,
103–111.
34 B. Coq, M. Mauvezin, G. Delahay and S. Kieger, J. Catal.,
2000, 195, 298–303.
35 X. Shi, H. He and L. Xie, Chin. J. Catal., 2015, 36, 649–656.
22 A. Thangaraj, R. Kumar and P. Ratnasamy, J. Catal., 1991,
131, 294–297.
23 A. Tuel, S. Moussa-Khouzami, Y. B. Taarit and C. Naccache, J. 36 A. De Stefanis, S. Kaciulis and L. Pandol, Microporous
Mol. Catal. A: Chem., 1991, 68, 45–52. Mesoporous Mater., 2007, 99, 140–148.
24 M. Kruk, T. Asefa, M. Jaroniec and G.-A. Ozin, J. Am. Chem. 37 J. L. Gu, X. Dong, S. P. Elangovan, Y. Li, W. Zhao, T. Iijima,
Soc., 2002, 124, 6383–6392.
Y. Yamazaki and J. L. Shi, J. Solid State Chem., 2012, 186,
25 L. Wu, Z. Li, D. Han, J. Wu and D. Zhang, Fuel Process.
Technol., 2015, 134, 449–455.
208–216.
38 T. Yamashita and P. Hayes, Appl. Surf. Sci., 2008, 254, 2441–
26 D. Gao, A. Zheng, X. Zhang, H. Sun, X. Dai, Y. Yang, H. Wang,
2449.
Y. Qin, S. Xu and A. Duan, Nanoscale, 2015, 7, 10918–10924. 39 A. P. Grosvenor, B. A. Kobe, M. C. Biesinger and
27 M. A. Karakassides, D. Gournis, A. B. Bourlinos, N. S. McIntyre, Surf. Interface Anal., 2004, 36, 1564–1574.
P. N. Trikalitis and T. Bakas, J. Mater. Chem., 2003, 13, 40 Y.-D. Chiang, H.-Y. Lian, S.-Y. Leo, S.-G. Wang, Y. Yamauchi
871–876. and K. C. W. Wu, J. Phys. Chem. C, 2011, 115, 13158–13165.
28 X. J. Lin, A. Z. Zhong, Y. B. Sun, X. Zhang, W. G. Song, 41 H. Nan, Q. Cai, Q. Sun, J. Qi, Y. Liu, H. Zhou, Y. Yu, W. Shi,
R. W. Lu, A. M. Cao and L. J. Wan, Chem. Commun., 2015,
51, 7482–7485.
T. Wang, C. Xu and H. Li, Microporous Mesoporous Mater.,
2011, 142, 585–591.
29 K. Niu, D. Shi, W. Dong, M. Chen and Z. Ni, J. Colloid 42 X. Liang, R. Yang, G. Li and C. Hu, Microporous Mesoporous
Interface Sci., 2011, 362, 74–80. Mater., 2013, 182, 62–72.
30 A. Gervasini, C. Messi, P. Carniti, A. Ponti, N. Ravasio and 43 T. Okuhara, Appl. Catal., A, 2003, 256, 213–224.
F. Zaccheria, J. Catal., 2009, 262, 224–234.
31 A. I. Carrillo, E. Serrano, R. Luque and J. Garcıa-Martınez,
44 P.-Y. Chao, S.-T. Tsai, T.-C. Tsai, J. Mao and X.-W. Guo, Top.
Catal., 2009, 52, 185–192.
´
´
Appl. Catal., A, 2013, 453, 383–390.
76074 | RSC Adv., 2016, 6, 76064–76074
This journal is © The Royal Society of Chemistry 2016