Structure Modification Function of g-C3N4 for Al2O3 in the in Situ Hydrothermal Process for Enhanced Photocatalytic Activity

Article abstract of DOI:10.1002/chem.201500224

Heterojunctions of g-C₃N₄/Al₂O₃ (g-C₃N₄=graphitic carbon nitride) are constructed by an in situ one-pot hydrothermal route based on the development of photoactive γ-Al₂O₃

Full text of DOI:10.1002/chem.201500224

DOI: 10.1002/chem.201500224
Full Paper
&
Heterojunctions
Structure Modification Function of g-C3N4 for Al2O3 in the In Situ
Hydrothermal Process for Enhanced Photocatalytic Activity**
Fa-tang Li,*[a] Shao-jia Liu,[a] Ya-bin Xue,[a] Xiao-jing Wang,[a] Ying-juan Hao,[a] Jun Zhao,[a] Rui-
hong Liu,[a] and Dishun Zhao[b]
Abstract: Heterojunctions of g-C3N4/Al2O3 (g-C3N4 =graphitic
carbon nitride) are constructed by an in situ one-pot hydro-
thermal route based on the development of photoactive g-
Al2O3 semiconductor with a mesoporous structure and
a high surface area (188 m2gÀ1) acting as electron acceptor.
A structure modification function of g-C3N4 for Al2O3 in the
hydrothermal process is found, which can be attributed to
the coordination between unoccupied orbitals of the Al ions
and lone-pair electrons of the N atoms. The as-synthesized
heterojunctions exhibit much higher photocatalytic activity
than pure g-C3N4. The hydrogen generation rate and the re-
action rate constant for the degradation of methyl orange
over 50% g-C3N4/Al2O3 under visible-light irradiation (l>
420 nm) are 2.5 and 7.3 times, respectively, higher than
those over pristine g-C3N4. The enhanced activity of the het-
erojunctions is attributed to their large specific surface areas,
their close contact, and the high interfacial areas between
the components as well as their excellent adsorption per-
formance, and efficient charge transfer ability.
Introduction
conduction bands (CBs) of the coupled semiconductors, which
cannot only enhance the separation efficiency of photo-gener-
ated charge carriers, but also endow the composites new un-
expected properties due to synergistic effect.[10g]
Ever since Wang et al.[1] reported its hydrogen production abili-
ty in 2009, graphitic carbon nitride (g-C3N4) has received in-
creasing attention in the fields of photocatalysis, electrochem-
istry, and photoelectrochemistry due to its fascinating two-di-
mensional (2D) structure and notable characters, such as low
cost, earth-abundant elements, non-toxicity, thermal and
chemical stability, and narrow band gap of about 2.70 eV.[2]
Nevertheless, the recombination rate of photoinduced elec-
tron–hole pairs of g-C3N4 is still fast, which needs to be further
slowed down for an enhancement of the quantum efficiency.[3]
Thus, some methods have been exploited, such as ion
doping,[4] noble-metal nanoparticle deposition,[5] morphology
or pore structure design,[6] acid or alkali treatment,[7] dye sensi-
tization,[8] conjugated polymer coupling,[9] and heterojunction
fabrication.[3,10] In particular, the construction of hybrids by
combining g-C3N4 with other semiconductors is an effective
strategy. The basic design thought for composites is based on
the fact that the photo-induced electrons on the lowest unoc-
cupied molecular orbital (LUMO) of g-C3N4 can migrate to the
On the other hand, searching for appropriate earth-abun-
dant, inexpensive, and nontoxic semiconductors is indispensa-
ble for practical large-scale application in industry.[11] Al2O3 can
meet the requirements based on its ultraviolet-light-response
ability and its active role as electron acceptor.[12] In construct-
ing g-C3N4/Al2O3 hybrids, most recently, our group found that
a-Al2O3-containing amorphous matter obtained through high-
temperature combustion is poor of hydroxyl group on its sur-
face and thus cannot form effective heterojunction with g-
[13]
C3N4. After treating g-C3N4 with aqueous ammonia solution
to graft hydroxyl groups, the effective g-C3N4/a-Al2O3 compo-
sites can be obtained by an ultrasonic dispersion route.[13] In
addition, compared to a-Al2O3, the g phase of Al2O3 has
a more disorder porous structure and a higher specific surface
area and the g-C3N4/g-Al2O3 composite has also been prepared
by using ultrasonic dispersion.[14] However, the simple combi-
nation of g-C3N4 and Al2O3 through ultrasonic dispersion re-
sults in limited contact areas between the two components.
Hence, more effective g-C3N4/Al2O3 should be further devel-
oped.
[a] Prof. F.-t. Li, S.-j. Liu, Y.-b. Xue, Dr. X.-j. Wang, Dr. Y.-j. Hao, Dr. J. Zhao,
R.-h. Liu
In situ fabrication of heterojunctions is a potential tech-
nique, which can supply more contact sites and interactions
between/among components. g-C3N4-based heterojunctions
College of Science
Hebei University of Science and Technology
Shijiazhuang 050018 (P.R. China)
E-mail: lifatang@126.com
constructed through an in situ route, such as ZnFe2O4/g-
[b] Prof. D. Zhao
[10h]
C3N4
and g-C3N4/(BiO)2CO3,[15] have exhibited their superiori-
School of Chemical and Pharmaceutical Engineering
Hebei University of Science and Technology
Shijiazhuang 050018 (P.R. China)
ty in enhancing the photocatalytic performance. In this work,
we employ a hydrothermal route to develop a mesoporous
photoactive g-Al2O3 phase with a high specific surface area
[**] g-C3N4 =graphitic carbon nitride.
Chem. Eur. J. 2015, 21, 10149 – 10159
10149
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
(SBET) and then construct g-C3N4/Al2O3 heterojunctions through
an in situ hydrothermal synthesis for the first time. The defect
site on Al2O3 is designed to trap electrons from g-C3N4 to im-
prove the separation efficiency of charge carriers of g-C3N4. In-
terestingly, a structure modification role of g-C3N4 for Al2O3 in
the hydrothermal process is found and the resultant hybrid
with an appropriate g-C3N4 ratio possesses a nearly equal SBET
value to pure g-C3N4 in spite of the low SBET value of g-C3N4.
Compared with bare g-C3N4, the composite exhibits much
higher photocatalytic ability for hydrogen generation and deg-
radation of methyl orange (MO) under visible light. In addition,
compared to pristine g-C3N4, the hybrids show weaker photo-
luminescence intensity, which is different from other g-C3N4-
based composites. This research not only offers a new method
for developing inexpensive photoactive mesoporous Al2O3 and
finds a structure adjustment effect of g-C3N4 in designing het-
erojunctions through a hydrothermal process, but also pro-
vides a new insight into defect chemistry concerning the pho-
toluminescence behavior of heterojunctions composed of g-
C3N4 and semiconductors with ample defect sites.
(JCPDS card No. 87-1526). The XRD pattern of g-C3N4 treated
hydrothermal in the absence of Al(NO3)3 is also presented in
Figure 1. It is obvious that there is no change for the peak po-
sitions in comparison with the untreated g-C3N4. However, for
the g-C3N4/Al2O3 hybrids, there are two changes compared to
pure g-C3N4. Firstly, the peaks at 13.1 and 27.48 of pure g-C3N4
shift to 14.2 and 27.68, respectively, in the g-C3N4/Al2O3 hybrids,
demonstrating that the corresponding interlamellar spacings
become narrower in the presence of Al ions. It is speculated
that the unoccupied 3p or 3d orbitals of the Al ions can coor-
dinate with the lone-pair electrons on the
N
atoms.[16]
Scheme 1 shows the graphic representation of g-C3N4 with Al
for shorting the (100) spacing. Secondly, the intensity of the
Results and Discussion
XRD analysis
Figure 1 shows the X-ray diffraction (XRD) patterns of the as-
prepared pure g-C3N4, pure Al2O3, and the g-C3N4/Al2O3 compo-
sites with different mass ratios. It can be seen from the pattern
of Al2O3 that there are two peaks at 45.7 and 67.18, corre-
sponding to the (400) and (440) diffraction planes of cubic g-
Al2O3, respectively (JCPDS card No. 10-0425); meanwhile, the
broad peak shape implies its low crystallinity, which also indi-
cates the existence of defect. Furthermore, the raw material of
Al(NO3)3 or other intermediate crystal products cannot be de-
tected, indicating the transformation of Al(NO3)3 to Al2O3
through the hydrothermal and calcination processes is com-
plete. As for the pure g-C3N4 sample obtained through pyroly-
sis of melamine without the hydrothermal process, a distinct
diffraction peak located at 27.48 and an obscure peak at 13.18
are indexed as (002) and (100) planes of g-C3N4, respectively
Scheme 1. Graphic representation of g-C3N4 in the presence of Al.
peak at 14.28 for the hybrids is much stronger than that of
pure g-C3N4. This result indicates that the Al element benefits
the growth of (100) planes, which is also attributed to the co-
ordination effect. In addition, it can be seen from the patterns
of the g-C3N4/Al2O3 composites with high g-C3N4 contents, that
the two peaks of the Al2O3 phase cannot be observed due to
its weak crystallinity. On the other hand, there are two very
weak peaks at 37.6 and 49.48 emerging in the 60 and 70% g-
C3N4/Al2O3 composites. The former peak corresponds to the
(311) planes of Al2O3 and the latter is to be further investigat-
ed, which shows that g-C3N4 can change the growth of the
Al2O3 planes.
Morphology, surface areas, and microstructures
To compare the morphology and to verify the defect, transmis-
sion electron microscopy (TEM) and high-resolution (HR) TEM
images of the commercial g-Al2O3 and the as-prepared Al2O3
are shown in Figure 2. It can be seen from Figures 2a and
b that the commercial g-Al2O3 shows a regular particle aggre-
gation, whereas the as-prepared Al2O3 exhibits a short rod-like
structure. It is clear that there is lattice fringe in the HRTEM
image of commercial g-Al2O3 shown in Figure 2c, and the in-
terplanar spacing is about 0.46 nm, corresponding to the (111)
plane of g-Al2O3. For comparison, it is hard to find obvious lat-
tice fringe in the image of the as-prepared Al2O3 sample, verify-
Figure 1. XRD patterns of Al2O3, g-C3N4 before and after hydrothermal pro-
cess, and g-C3N4/Al2O3 composites with different g-C3N4 contents.
Chem. Eur. J. 2015, 21, 10149 – 10159
www.chemeurj.org
10150
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 2. TEM images of a) commercial g-Al2O3 and b) as-prepared Al2O3,
and HRTEM images of c) commercial g-Al2O3 and d) as-prepared Al2O3. Scale
bars=50 nm for 2a, 2b, and 10 nm for 2c, 2d.
ing its low-crystallinity or amorphous-similar structure, which is
the base of abundant defects.
The scanning electron microscopy (SEM) image of the as-pre-
pared Al2O3 is displayed in Figure 3a. It is obvious that Al2O3 ex-
hibits a cotton-like morphology, implying its amorphous-similar
structure and high specific surface area. Compared with bare
Al2O3, the sheet-like g-C3N4 can be observed in the 50% g-C3N4/
Al2O3 composite (Figure 3b). It is evident that g-C3N4 is not
coated on the surface of Al2O3 simply, but embodied in the
cotton-like Al2O3. In addition, the EDS pattern shown in Fig-
ure 3c and the corresponding elemental mapping images show
that the elements C, N, O, and Al can be found in the sample,
providing the direct solid evidence for the combination of g-
C3N4 and Al2O3. From the TEM image of the 50% g-C3N4/Al2O3
hybrid (Figure 3d), it is seen that the strips of g-C3N4 are com-
bined with Al2O3. Figure 3e shows the HRTEM image of the
composite. By measuring the lattice fringe, the interplanar spac-
ing is about 0.198 nm, corresponding to the (400) plane of g-
Al2O3. However, the main field is composed of a cloud without
lattice fringe, which is consistent with Figure 2d.
Figure 3. SEM images of a) pure Al2O3 and b) the 50% g-C3N4/Al2O3 sample.
c) EDS spectrum and corresponding elemental mapping images of 50% g-
C3N4/Al2O3. d) TEM and e) HRTEM images of 50% g-C3N4/Al2O3.
The TEM images of Pt-50% g-C3N4/Al2O3 after Pt particle
deposition through in situ light irradiation, are exhibited in
Figure 4. The dark dots are Pt metal particles, which prefer de-
positing on Al2O3 rather than doing on striped g-C3N4.
Figure 5 exhibits the N2 adsorption/desorption isotherms
and the corresponding pore size distribution curves (inset) of
the as-prepared Al2O3 and g-C3N4/Al2O3 samples. The isotherms
belong to the IV Brunauer type[17] with a H3 hysteresis loop, in-
dicating the existence of mesopores (2–50 nm). All the data of
the obtained specific surface areas (SBET), the average pore di-
ameters, and the total pore volumes are summarized in
Table 1. The pure Al2O3 has a high SBET value of 188.6 m2gÀ1
and the SBET value of pure g-C3N4 is only 14.0 m2gÀ1. However,
after introduction of g-C3N4, the decrease in the SBET value is
not proportionate to the g-C3N4 mass ratio in the hybrids. For
example, there is a very slightly reduction for the 30% g-C3N4/
Al2O3 sample from 188.6 to 185.0 m2gÀ1. Furthermore, the aver-
age pore diameter and the pore volume become larger. These
results indicate that the addition of g-C3N4 can induce
a change in the growth and structure of Al2O3, which can also
be proven by the pore size distribution shown in the inset of
Figure 5, suggesting that the hybrid has a broader pore size
Chem. Eur. J. 2015, 21, 10149 – 10159
www.chemeurj.org
10151
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ports on bulk g-C3N4-based hybrids with a high surface area
over 100 m2gÀ1. In this research, such high SBET values of these
composites supply a firm foundation for enhanced adsorption
capacity and photocatalytic activity.
Figure 6a depicts the X-ray photoelectron spectroscopy (XPS)
survey spectra of bare Al2O3 and 50% g-C3N4/Al2O3. The C1s
peak in the Al2O3 is due to the adventitious elemental carbon.
Compared to the survey spectrum of Al2O3, the 50% g-C3N4/
Al2O3 sample exhibits a stronger C1s peak and the new N1s
peak appears owning to the introduction of g-C3N4. At the
same time, the peak intensities of the O1s, Al2s, and Al2p orbi-
tals decrease because of the reduction of the Al2O3 mass ratio.
Figures 6b–e show the high-res-
Figure 4. TEM images of 50% g-C3N4/Al2O3 after Pt metal particle deposition.
olution XPS spectra of the Al2p,
O1s, N1s, and C1s orbitals of re-
lated samples, respectively. The
Al2p peak at 73.88 eV can be
seen for pristine Al2O3, which is
close to the references values
(73.85[18] and 73.90 eV[19]). After
combination with g-C3N4, this
peak shifts toward a higher bind-
ing energy of 74.48 eV (Fig-
ure 6b); meanwhile, the binding
energy of O1s at 530.78 eV in
pure Al2O3 also shifts to a higher
value of 531.38 eV (Figure 6c).
The two shifts towards higher
binding energies of Al2O3 indi-
cate that g-C3N4 is closely cov-
ered on the surface of Al2O3 and
there is a chemical interaction
between them. Furthermore, the
peak at 398.85 eV in pure g-C3N4,
typically a characteristic peak of
g-C3N4 that can be attributed to
sp2-hybridized N atoms (C=NÀ
C),[20] moves to a lower binding
energy of 398.23 eV in the 50%
g-C3N4/Al2O3 hybrid. For C1s
spectra, only one peak located at
284.68 eV is observed for the
pure Al2O3 sample, whereas
there are two peaks at 288.10
and 284.60 eV observed for the
g-C3N4 sample. The peak at
around 284.6 eV is derived from
contaminated carbon,[21] whereas
Figure 5. N2 adsorption/desorption isotherms and pore size distribution curves (inset) of a) the as-synthesized
Al2O3, b) 30%g-C3N4/Al2O3, c) 40%g-C3N4/Al2O3, d) 50%g-C3N4/Al2O3, e) 60%g-C3N4/Al2O3, and f) 70%g-C3N4/Al2O3
composites.
the peak at 288.10 eV is identi-
fied as sp2-bonded C atoms (NÀ
C=N) involved in C3N4 triazine
cycles.[21,22] The g-C3N4/Al2O3
than pure Al2O3. This result indicates that g-C3N4 has a structure
adjustment role for Al2O3 in the hydrothermal process, which is
related to the coordination between the Al ions and the N
atoms. To the best of our knowledge, there is no report on g-
C3N4 acting as structure-directing agent and there are few re-
composite exhibits two C1s peaks and there is also a shift from
288.10 to 287.58 eV, showing the existence of graphitic carbon
nitride and the effect of Al2O3 on g-C3N4. From the XPS results,
it can be concluded that a strong chemical interaction exists
between the components Al2O3 and g-C3N4.
Chem. Eur. J. 2015, 21, 10149 – 10159
www.chemeurj.org
10152
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
sured by UV/Vis diffuse reflec-
tance spectroscopy (DRS). It can
be seen from Figure 7 that the
commercial g-Al2O3 does not
show any absorbance in the
Table 1. SBET values and pore parameters of the samples and their photocatalytic performances for MO remov-
al.
Sample
SBET
[m2 gÀ1
Average pore
Pore volume Fitted
equation
k
[hÀ1
Correlation
coefficient (R)
]
diameter [nm] [cm3 gÀ1
]
]
g-C3N4
Al2O3
14.0
188.6
185.0
160.3
147.8
147.0
113.0
60.2
7.8
6.8
9.3
8.5
9.4
10.1
11.4
5.6
0.04
0.32
0.43
0.34
0.35
0.37
0.32
0.20
y=0.0623x+0.0121 0.0623 0.9981
range
of
l=200–800 nm,
–
–
–
whereas the as-prepared Al2O3
can respond to the ultraviolet
light, which can be attributed to
its disorder amorphous fractio-
n.[12a] The result also indicates
the photoactive function of the
as-synthesized g-Al2O3 through
a hydrothermal process. Com-
pared to pure g-C3N4, the ab-
sorption band edges of the hy-
brids are slightly blue shifted
due to the existence of Al2O3.
However, in the UV region
below l=400 nm, some compo-
sites display stronger optical ab-
sorption than the pure g-C3N4 al-
though the absorption of the
component Al2O3 is still the
weakest. This phenomenon is
consistent with a previous report
on g-C3N4/TiO2,[23] indicating that
there should be a synergistic
effect between g-C3N4 and Al2O3
to enhance the photoabsorp-
tion.
30% g-C3N4/Al2O3
40% g-C3N4/Al2O3
50% g-C3N4/Al2O3
60% g-C3N4/Al2O3
70% g-C3N4/Al2O3
g-C3N4/Al2O3 (2:1)[a]
y=0.0984x+0.0715 0.0984 0.9996
y=0.1348x+0.0619 0.1348 0.9995
y=0.4550x+0.0471 0.4550 0.9982
y=0.3219x+0.0492 0.3219 0.9974
y=0.2028x+0.0481 0.2028 0.9978
y=0.2456x+0.0377 0.2456 0.9937
[a] For reference.
Figure 8a shows the photoca-
talytic H2 production over the
as-prepared
photocatalysts
under visible-light illumination
(l>420 nm). No hydrogen could
be detected without either pho-
tocatalyst or light irradiation. All
the in situ Pt-deposited samples
exhibit photocatalytic activity for
hydrogen generation. The H2
evolution amount is nonlinearly
increased with the irradiation
time, especially in the initial
three hours, because the mass
of deposited Pt is gradually be-
coming more and more under
the light illumination, thus lead-
ing to an enhanced electrons
transfer ability step by step. The
average photocatalytic hydrogen
production rates in 6 h of the
samples are shown in Figure 8b.
Figure 6. a) XPS survey spectra of bare Al2O3 and the 50% g-C3N4/Al2O3 composite, and high-resolution XPS spec-
tra of b) Al2p, c) O1s, d) N1s, and e) C1s of the different samples.
Optical absorption and photocatalytic hydrogen generation
performances
All the hybrids display higher photocatalytic activity than
pure g-C3N4. The 50% g-C3N4/Al2O3 sample reaches the highest
hydrogen evolution with a rate of 52.10 mmolhÀ1, which is
2.5 times higher than that of pristine g-C3N4.
The optical absorption of pristine Al2O3, g-C3N4, and the g-
C3N4/Al2O3 composites with different g-C3N4 contents was mea-
Chem. Eur. J. 2015, 21, 10149 – 10159
www.chemeurj.org
10153
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 9. Transient photocurrent responses of g-C3N4 and a g-C3N4/Al2O3
hybrid electrodes with light on-and-off cycles.
Figure 7. UV/Vis DRS results of the as-prepared Al2O3, commercial g-Al2O3, g-
C3N4, and the g-C3N4/Al2O3 hybrids.
electrons and holes in the hybrids, which can be ascribed to
the match of energy level between g-C3N4 and Al2O3.
To further understand the charge carrier transfer behavior,
the PL measurement was conducted with an excitation wave-
length of l=290 nm. There are three abnormal PL phenomena
observed from Figure 10a. Firstly, the intensity order of the
pure g-C3N4 and the heterojunctions are contrary to their pho-
tocatalytic hydrogen activities. That is, the photocatalytic activ-
ity increases with the PL intensity, which conflicts with the pre-
vious papers on g-C3N4-based hybrids.[3,10] The reason is that
the PL intensity is determined by two factors. Generally, The PL
signal is derived from the radioactive recombination of photo-
induced electrons and holes; hence, a high PL emission inten-
sity indicates a high recombination rate of electron–hole
pair,[25] thus a low quantum efficiency. On the other hand,
photo-generated electrons can be bound by the defects of the
photocatalysts to form free or binding excitons so that the PL
signal can easily occur, and the greater the number of defects,
the stronger the PL peak intensity, the higher is the photocata-
lytic activity.[26] For the investigated g-C3N4/Al2O3 hybrids, de-
fects play a dominant role for the PL signal and the sample
with stronger PL intensity exhibits the higher photocatalytic
ability due to the existence of abundant defect sites in g-Al2O3.
As for the reported g-C3N4-based composites, because the em-
ployed semiconductors are crystals without or with little de-
fects, their PL behavior is determined by the separation of the
charge carriers and the sample with the stronger PL intensity
exhibits the lower photocatalytic ability.
Figure 8. a) Photocatalytic H2 production amount versus time and b) average
H2 evolution rate in 6 h over the photocatalysts.
Secondly, the main emission wavelength of pure g-C3N4 at
l=460 nm is blue shifted to l=440 nm for the composites. To
investigate the reason, the PL test of hydrothermal-treated g-
C3N4 was carried out. It is seen that its main wavelength is still
at l=460 nm and its intensity is lower than that of original g-
C3N4, demonstrating that hydrothermal process cannot change
the emission wavelength but can improve the crystallinity of
g-C3N4 and thus decrease the PL intensity. Therefore, the shift
of the emission wavelength for the composites is due to the
interaction between Al2O3 and g-C3N4.
The transient photocurrent is widely considered as the
direct evidence for demonstrating the electron and hole sepa-
ration in heterostructured photocatalysts.[23,24] Figure 9 shows
the transient photocurrent versus time curves of the 50% g-
C3N4/Al2O3 sample and pure g-C3N4 electrodes under visible-
light illumination (l>420 nm) in an on-and-off cycle mode. It
is obvious that the photocurrent response exhibits a good re-
producibility for both samples through five intermittent irradia-
tion on-off cycles, indicating that the electrodes are stable. The
photocurrent value of the 50% g-C3N4/Al2O3 electrode is about
2.3 times higher than that of the pristine g-C3N4 electrode. This
indicates an enhanced separation efficiency of photo-induced
Thirdly, under the excitation wavelength of l=290 nm, pure
Al2O3 exhibits a broad emission peak from l=325 to 375 nm.
However, this peak is not detected completely in the g-C3N4/
Chem. Eur. J. 2015, 21, 10149 – 10159
www.chemeurj.org
10154
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 11. EPR signals of the as-prepared and commercial g-Al2O3.
can confirm the presence of solid-state defect,[27] which can act
as acceptors for photo-generated electrons from g-C3N4.
Photocatalytic performance for the degradation of methyl
orange
Figure 12a shows the photocatalytic decomposition of solu-
tions of the refractory organic dye MO. It can be seen that
only 0.3% MO is removed in 2 h in the absence of photocata-
lysts, demonstrating that MO is a relative stable dye and the
self-photolysis of MO can be neglected under visible-light illu-
mination. With pure g-C3N4, the degradation efficiency is 12.9.
Compare value below 47.9% in 2 h over 50% g-C3N4/Al2O3, ex-
hibiting its poor photocatalytic ability. Over hydrothermal-
treated g-C3N4, the removal yield of MO is decreased to 9.4%
(not shown in Figure 12a), indicating that the high-tempera-
ture hydrothermal process decreases the photocatalytic activity
of g-C3N4. When the heterojunctions are employed, the adsorp-
tion capacity for MO is increased from 1.2% (bare g-C3N4) to
4.7–6.9% and then the photocatalytic degradation efficiency is
greatly enhanced. To quantitatively reveal the photocatalytic
activities of these samples, the following pseudo-first-order
equation [Eq. (1)] is employed,[12a] in which k is the reaction
rate constant, C0 represents the initial concentration of MO
(10 mgLÀ1), C1 is the concentration after 30 min adsorption,
and C is the MO concentration at time t.
Figure 10. a) Photoluminescence spectra of various samples (excitation
wavelength: l=290 nm) and schematic diagram for PL occurrence of
b) pure Al2O3 and g-C3N4, and c) g-C3N4/Al2O3 heterojunctions.
Al2O3 hybrids. Based on the PL signals, the recombination pro-
cesses of photo-induced electrons and holes are proposed as
Figures 10b and c for pure Al2O3, pure g-C3N4, and g-C3N4/
Al2O3 hybrids. Accepting the photo energy, the electrons on
the valence band (VB) of Al2O3 or the highest occupied molec-
ular orbital (HOMO) of g-C3N4 can be excited to the CB or
LUMO of them, and then some of the electrons migrate back
to the VB directly or through the lower defect energy levels.
For the g-C3N4/Al2O3 composites, the electrons on the LUMO of
g-C3N4 would combine with holes, or transfer to the CB of
Al2O3. The electrons on Al2O3 excited from the VB or trans-
ferred from g-C3N4 would move to other sites or to the HOMO
of g-C3N4, but not to the VB of Al2O3.
ln C0=C ¼ kt þ ln C0=C1
ð1Þ
The time-course variation of ln(C0/C) of all samples is illus-
trated in Figure 12b. The results of the calculated rate con-
stants (ks) and the correlation coefficients (Rs) are listed in
Table 1. From Figures 12a and b as well as Table 1, it is clear
that the pure g-C3N4 has a low rate constant of only 0.0623 hÀ1
and that the g-C3N4/Al2O3 hybrids show much higher photoca-
talytic performances. Especially for 50% g-C3N4/Al2O3, the reac-
tion rate constant (0.4550 hÀ1) is 7.3 times that of pure g-C3N4.
For comparison, the photocatalytic activity of a 2:1 g-C3N4/
Al2O3 heterojunction obtained through ultrasonic dispersion[14]
is 3.9 times that of pure g-C3N4. This indicates that the one-pot
fabrication of the heterojuncions through a hydrothermal pro-
cess is a more efficient route, which can be ascribed to the
To further investigate the existence of defects, electron para-
magnetic resonance (EPR) measurements were performed. It is
clear that the as-prepared g-Al2O3 shows obvious triplet g
value signals (g=1.987, 2.004, and 2.024) (Figure 11), whereas
the commercial g-Al2O3 does not. This triplet g value signals
Chem. Eur. J. 2015, 21, 10149 – 10159
www.chemeurj.org
10155
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 13. a) Cycle runs of the 50% g-C3N4/Al2O3 heterojunction for the
photocatalytic degradation of MO under visible-light irradiation
(l>420 nm). b) XRD patterns of 50% g-C3N4/Al2O3 before and after use.
To investigate the reactive species accounting for the MO
degradation, scavenger experiments were conducted in the
presence of ammonium oxalate monohydrate (AO, 2 mm, h+
CÀ
scavenger), 1,4-benzoquinone (BQ, 1 mm, O2 scavenger), or
[28]
C
isopropanol (iPrOH, 10 mm, OH scavenger). As shown in Fig-
ure 14a, it is obvious that the addition of only 1 mm BQ most
acutely depressed the photodegradation of MO, which demon-
CÀ
strates that O2 is the dominant species responsible for the
MO degradation. Meanwhile, AO and iPrOH can also lower the
degradation efficiency of MO, indicating that h+ and OH are
C
the active species at the same time. Because the HOMO poten-
tial of g-C3N4 is 1.4 eV versus a normal hydrogen electrode
Figure 12. Time-course variation of a) C/C0 and b) ln(C0/C) of the degradation
of MO, and c) photocatalytic degradation of 4-chlorophenol (4-CP) over vari-
ous samples under visible-light illumination.
(NHE),[29] which is less than the potentials of OH /H2O
C
À
(2.27 eV)[30] and OH /OH (1.99 eV),[31] this implies that the
C
holes on g-C3N4 cannot oxidize H2O or OHÀ to form OH radi-
C
C
close contact of g-C3N4 and Al2O3 through the in-situ process
and the higher surface areas.
cals. Therefore, the OH radicals involved in the photocatalytic
CÀ
reaction are derived from O2 based on the fact that the
The stability and re-usability of the 50% g-C3N4/Al2O3 com-
posite were evaluated by a cycling degradation experiment
and the results are shown in Figure 13a. It can be seen that
the sample does not display an obvious loss in its photocata-
lytic degradation activity after four runs, demonstrating that
the g-C3N4/Al2O3 hybrids are stable and the interaction be-
tween g-C3N4 and Al2O3 is strong due to their in situ formation
process. As shown in Figure 13b, compared with unused g-
C3N4/Al2O3, the used photocatalyst exhibits a weaker g-C3N4
peak intensity, whereas the Al2O3 almost remains the same.
This result demonstrates that there is slight g-C3N4 loss in the
photocatalytic process; meanwhile, there is also loss of fine g-
C3N4/Al2O3 nanoparticles in the recycling process. As a result,
the activity is decreased after four cycle runs.
LUMO potential of g-C3N4 at À1.3 eV[29] is more negative than
that of E (O2/O2 ) (À0.28 eV vs. NHE).[30]
o
CÀ
C
CÀ
To further observe the existence of OH and O2 radicals di-
rectly, the 2-hydroxyterephthalic acid fluorescence and nitro-
blue tetrazolium techniques were used based on the facts that
C
OH radicals can react with terephthalic acid (TA) to form the
high-fluorescent compound TAOH,[32] and NBT has a maximum
CÀ
absorption peak at l=259 nm whereas the product of O2
and NBT does not.[33] To perform TAOH and NBT tests, TA was
dissolved in 2.0 mm NaOH to obtain a 0.5 mm solution and
NBT was dissolved in water to receive a 2.5”10À5 m solution.
Then, 0.1 g of the photocatalyst were added to 100 mL of the
solutions, and the mixtures were irradiated under visible light.
The fluorescence and ultraviolet-visible absorption intensities
Chem. Eur. J. 2015, 21, 10149 – 10159
www.chemeurj.org
10156
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 15. Proposed mechanism for a) the photocatalytic H2 generation and
b) the MO degradation over g-C3N4/Al2O3 heterojunctions under visible-light
irradiation. TEOA=triethanolamine.
electron–hole pairs and improved the photocatalytic activity.
Then the electrons on the LUMO of g-C3N4 or the defect sites
of Al2O3 reduce H+ to produce H2. As for the MO degradation,
electrons on the two positions can reduce O2 adsorbed on the
Figure 14. a) C/C0 variation of the MO degradation over various samples in
the presence of scavengers, time-course variation of b) the 2-hydroxytereph-
thalic acid fluorescence (TAOH) intensity, and c) transformation percentage
of nitroblue tetrazolium (NBT) over the 50% g-C3N4/Al2O3 hybrid.
CÀ
CÀ
cals then react with H2O to produce OH radicals. Finally, MO
surface of the photocatalysts to generate O2 . The O2 radi-
C
+
CÀ
C
can be degraded by the left h , O2 , and OH radicals.
To observe the universality of the as-prepared heterojunc-
tions, colorless 4-chlorophenol was selected to evaluate the
photocatalytic activity. Figure 12c displays the decrease of the
4-CP concentration with the reaction time. After 4 h of irradia-
tion, the 4-CP removal is 51.3 and 36.5% over g-C3N4 and 50%
g-C3N4/Al2O3, respectively. The corresponding calculated first-
order reaction rate constants are 0.1704 and 0.1052 hÀ1 (not
shown), respectively, which means that the photocatalytic ac-
tivity of 50% g-C3N4/Al2O3 is 1.62 times that of pure g-C3N4 in
the degradation of the colorless organic pollutant 4-CP.
of the aqueous solutions were measured on a Hitachi F-4600
fluorescence spectrophotometer (excitation wavelength: l=
315 nm) and Evolution 220 UV/Vis spectrophotometer, respec-
tively. The TAOH experiment results shown in Figure 14b illus-
trate that the peak intensity of TAOH over 50% g-C3N4/Al2O3 is
increased with the irradiation time, demonstrating the exis-
C
tence of OH radicals. Meanwhile, the NBT concentration is dra-
matically decreased with the time as shown in Figure 14c, im-
CÀ
plying that there are abundant O2 radicals reacting with NBT,
which is consistent with the scavenger result.
On the basis of the photocatalytic H2 generation and MO
degradation data and tests, the potential routes of the photo-
induced electron transfer are proposed as shown in Figure 15.
Under visible-light illumination, electrons are migrated from
the HOMO of g-C3N4 to its LUMO; afterwards, some electrons
move to the defect sites of amorphous Al2O3 due to the sink
effect. This leads to an enhanced separation efficiency of the
Conclusion
We have successfully synthesized photoactive g-Al2O3 and fab-
ricated mesoporous g-C3N4/Al2O3 heterojunctions through an
in situ one-pot hydrothermal route and a following calcination
treatment. A structure adjustment role of g-C3N4 for Al2O3 in
the hydrothermal process is found, which is beneficial to keep
Chem. Eur. J. 2015, 21, 10149 – 10159
www.chemeurj.org
10157
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
system by using a 300 W Xe lamp with l=420 nm cut-off filter as
visible-light source. In a typical experiment, powders of the photo-
catalyst (0.1 g) were dispersed in a mixed solution (100 mL ) con-
taining deionized water (70 mL), triethanolamine (25 mL), and an
aqueous solution of H2PtCl6·6H2O (5 mL). Herein, TEOA is taken as
sacrificial reagent to capture the photo-induced holes. The de-
signed in situ photodeposited Pt metal quality serving as cocata-
lyst to transfer electrons is 0.001 g, which accounts for 1.0 wt% of
the added catalysts. After stirring the suspension for 30 min, the
light was turned on to trigger reaction. Hydrogen evolved was an-
alyzed by an on-line gas chromatograph equipped with a thermal
conductivity detector (TCD) every 1 h. In the photocatalytic pro-
cess, the solution temperature was kept at 308C by using a thermo-
static water bath. There was no hydrogen generated without pho-
tocatalyst or light irradiation, demonstrating that hydrogen was
produced through a photocatalytic reaction.
the high specific surface area of the composites. The photolu-
minescence behavior of pure g-C3N4 and the g-C3N4/Al2O3 com-
posites are also investigated, which is different with other g-
C3N4-based heterojunctions, showing that defects play an im-
portant role. The defect site of the as-synthesized g-Al2O3 is
confirmed by EPR technique. This work not only prepares
a defect photoactive Al2O3 with a high surface area through
a hydrothermal method, but also presents the structure modi-
fication function of g-C3N4 for Al2O3 in the hydrothermal pro-
cess for the first time and thus provides new insight into fabri-
cating high efficient heterojunctions.
Experimental Section
The photocatalytic degradation performance towards pollutants
was evaluated by the removal of methyl orange (MO), a typical re-
fractory azo dye, and colorless 4-chlorophenol (4-CP). A 350 W Xe
lamp with a l=420 nm cutoff filter was employed as the visible-
light source, with a total light intensity of 0.40 WcmÀ2 in the range
of l=420–1064 nm measured by using a Newport 842-PE optical
power/energy meter. In each experiment, an aliquot of the sample
(0.10 g) was suspended in aqueous solutions of MO or 4-CP
(100 mL, 10 mgLÀ1). After being stirred for 30 min to achieve the
adsorption/desorption equilibrium of MO on the catalyst, the light
was turned on to begin the photocatalytic reaction. The MO con-
centration was analyzed by using a Thermo Scientific Evolution
220 UV/Vis spectrophotometer and the value at the maximal ab-
sorbance wavelength was noted to plot the removal efficiency of
MO and thus, measure the photodegradation ability of the sam-
ples. The concentration of 4-CP was collected by measuring the
absorbance at l=230 nm. For the cycle experiments, the nano-
powders were separated from the MO solution by centrifugation
at 13000 rminÀ1 for 5 min and subsequently dried at 110 8C for 2 h
for the next photocatalytic reaction cycle.
Catalysts preparation: All chemicals and commercial g-Al2O3 as
reference were procured from Aladdin (Shanghai, China) and used
as received. g-C3N4 was prepared through the thermal condensa-
tion of melamine. Melamine (5.0 g) was added in a semicovered
crucible and heated to 5508C for 2 h at a rate of 108CminÀ1. g-
C3N4/Al2O3 hybrids were synthesized through a hydrothermal route
and a following calcination. The designed total weight of the one-
pot heterojunction was 0.30 g. In a typical procedure for 50% g-
C3N4/Al2O3, Al(NO3)3·9H2O (1.103 g) was dissolved in distilled water
(30 mL) and then as-prepared g-C3N4 (0.150 g) was dispersed in the
solution under ultrasonication for 20 min. Afterwards, 1m NaOH
solution was added dropwise to the mixture till the pH reached 8–
9. After being magnetically stirred for 2 h, the mixed suspension
was transferred into a 50 mL Teflon-lined stainless autoclave and
heated at 1408C for 24 h. Then the obtained product was collect-
ed, thoroughly washed with deionized water for three times and
air dried at 808C for 3 h. Finally, the resultant solid particles were
calcined at 4008C for 2 h to obtain g-C3N4/Al2O3 composites, which
are labeled as x% g-C3N4/Al2O3 (x=mass fraction of g-C3N4 in the
hybrids). By using the aforementioned process, pristine Al2O3 was
synthesized without the addition of g-C3N4. For comparison, g-
C3N4/Al2O3 (2:1) was prepared through chemisorption according to
reference [14], in which Al2O3 was obtained by precipitation syn-
thesis.
Acknowledgements
This work was financially supported by the National Natural
Science Foundation of China (No. 21376061), the Program for
New Century Excellent Talents in University (No. NCET-12-
0686), the Natural Science Foundation for Distinguished Young
Scholar of Hebei Province (B2015208010), and the Scientific Re-
search Foundation for High-Level Talent in University of Hebei
Province (GCC2014057).
Catalysts characterization: The XRD patterns of the samples were
obtained on a Rigaku D/MAX 2500 X-ray diffractometer. The SEM
and element mapping images were acquired from a HITACHI
S4800 instrument equipped with energy dispersive X-ray spectrom-
eters (EDS, Bruker Quantax). TEM and HRTEM tests for the mor-
phology observation were recorded on a JEOL JEM-2010 micro-
scope. N2 adsorption/desorption measurements were carried out
on a Micromeritics Tristar II 3020 apparatus. XPS spectra were ach-
ieved by using a PHI 1600 ESCA XPS system. The UV/Vis DRS were
collected by using a Thermo Scientific Evolution 220 spectropho-
tometer. EPR measurements were carried out on a Bruker EMX-8/
2.7 X-band EPR spectrometer for detecting the oxygen vacancy of
samples.
Keywords: aluminum
·
catalysis
·
ceramics
·
electron
transport · photochemistry
[1] X. C. Wang, K. Maeda, A. Thomas, K. Takanabe, G. Xin, J. M. Carlsson, K.
Domen, M. Antonietti, Nat. Mater. 2009, 8, 76–80.
Photoelectrochemical measurements were performed on an elec-
trochemical workstation (CHI-660E, China) by using a conventional
three-electrode quartz cells. An aqueous solution containing 0.2m
Na2S and 0.04m Na2SO3 was employed as the electrolyte. The
counter and the reference electrodes were platinum wire and Ag/
AgCl (saturated KCl), respectively. The as-obtained photocatalyst
film electrodes deposited on cleaned 1.5 cm”1.0 cm fluoride-tin
oxide (FTO) glass served as the working electrode.
[2] a) Y. Zheng, J. Liu, J. Liang, M. Jaroniec, S. Z. Qiao, Energy Environ. Sci.
2012, 5, 6717–6731; b) J. J. Zhu, P. Xiao, H. L. Li, S. A. C. Carabineiro,
ACS Appl. Mater. Interfaces 2014, 6, 16449–16465; c) J. Chen, S. H. Shen,
P. H. Guo, P. Wu, L. J. Guo, J. Mater. Chem. A 2014, 2, 4605–4612; d) H. P.
Li, J. Y. Liu, W. G. Hou, N. Du, R. J. Zhang, X. T. Tao, Appl. Catal. B 2014,
160–161, 89–97.
[3] M. Xu, L. Han, S. J. Dong, ACS Appl. Mater. Interfaces 2013, 5, 12533–
12540.
Photocatalytic activity tests: Photocatalytic hydrogen evolution
was carried out in a PerfectLight Labsolar-III gas-closed circulation
[4] a) C. Chang, Y. Fu, M. Hu, C. Y. Wang, G. Q. Shan, L. Y. Zhu, Appl. Catal. B
2013, 142–143, 553–560; b) S. Zhang, J. Li, M. Zeng, J. Li, J. Xu, X.
Chem. Eur. J. 2015, 21, 10149 – 10159
www.chemeurj.org
10158
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Wang, Chem. Eur. J. 2014, 20, 9805–9812; c) S. Z. Hu, L. Ma, J. G. You,
F. Y. Li, Z. P. Fan, G. Lu, D. Liu, J. Z. Gui, Appl. Surf. Sci. 2014, 311, 164–
171.
[16] D. Ghosh, G. Periyasamy, S. K. Pati, J. Phys. Chem. C 2014, 118, 15487–
15494.
[17] S. Brunauer, P. H. Emmet, E. Teller, J. Am. Chem. Soc. 1940, 62, 1723–
1732.
[18] C. D. Wagner, D. E. Passoja, H. F. Hillery, T. G. Kinisky, H. A. Six, W. T.
Jansen, J. A. Taylor, J. Vac. Sci. Technol. 1982, 21, 933–944.
[19] L. P. H. Jeurgens, W. G. Sloof, F. D. Tichelaar, E. J. Mittemeijera, Surf. Sci.
2002, 506, 313–332.
[20] Q. Xiang, J. Yu, M. Jaroniec, J. Phys. Chem. C 2011, 115, 7355–7363.
[21] N. Yang, G. Q. Li, W. L. Wang, X. L. Yang, W. F. Zhang, J. Phys. Chem.
Solids 2011, 72, 1319–1324.
[22] L. Q. Ye, J. Y. Liu, Z. Jiang, T. Y. Peng, L. Zan, Appl. Catal. B 2013, 142–
143, 1–7.
[23] K. Dai, L. H. Lu, C. H. Liang, Q. Liu, G. P. Zhu, Appl. Catal. B 2014, 156–
157, 331–340.
[24] D. M. Chen, K. W. Wang, D. G. Xiang, R. L. Zong, W. Q. Yao, Y. F. Zhu,
Appl. Catal. B 2014, 147, 554–561.
[25] N. Q. Wu, J. Wang, D. N. Tafen, H. Wang, J. G. Zheng, J. P. Lewis, X. G.
Liu, S. S. Leonard, A. Manivannan, J. Am. Chem. Soc. 2010, 132, 6679–
6685.
[26] a) L. Q. Jing, X. J. Sun, B. F. Xin, B. Q. Wang, W. M. Cai, H. G. Fu, J. Solid
State Chem. 2004, 177, 1213; b) Y. H. Ni, Y. Zhu, X. Ma, Dalton Trans.
2011, 40, 3689–3694.
[27] a) S. Fukuzawa, K. M. Sancier, T. Kwan, J. Catal. 1968, 11, 364–369; b) Y.
Wang, C. Feng, M. Zhang, J. Yang, Z. Zhang, Appl. Catal. B 2010, 100,
84–90.
[28] a) S. X. Ge, L. Z. Zhang, Environ. Sci. Technol. 2011, 45, 3027–3033;
b) J. H. Kou, Z. S. Li, Y. P. Yuan, H. T. Zhang, Y. Wang, Z. G. Zou, Environ.
Sci. Technol. 2009, 43, 2919–2924; c) Y. X. Chen, S. Y. Yang, K. Wang, L. P.
Lou, J. Photochem. Photobiol. A 2005, 172, 47–54.
[29] F. Z. Su, S. C. Mathew, G. Lipner, X. Z. Fu, M. Antonietti, S. Blechert, X. C.
Wang, J. Am. Chem. Soc. 2010, 132, 16299–16301.
[30] A. Fujishima, X. T. Zhang, C. R. Chim. 2006, 9, 750–760.
[31] S. Kim, W. Choi, Environ. Sci. Technol. 2002, 36, 2019–2025.
[32] K. I. Ishibashi, A. Fujishima, T. Watanabe, K. Hashimoto, Electrochem.
Commun. 2000, 2, 207–210.
[5] a) N. Y. Cheng, J. Q. Tian, Q. Liu, C. J. Ge, A. H. Qusti, A. M. Asiri, A. O. Al-
Youbi, X. P. Sun, ACS Appl. Mater. Interfaces 2013, 5, 6815–6819; b) D. S.
Deng, Y. Yang, Y. T. Gong, Y. Li, X. Xu, Y. Wang, Green Chem. 2013, 15,
2525–2531.
[6] a) X. Li, A. J. Ward, A. F. Masters, T. Maschmeyer, Chem. Eur. J. 2014, 20,
7345–7350; b) J. Liu, J. H. Huang, H. Zhou, M. Antonietti, ACS Appl.
Mater. Interfaces 2014, 6, 8434–8440; c) M. Woznica, N. Chaoui, S. Taab-
ache, S. Blechert, Chem. Eur. J. 2014, 20, 14624–14628; d) M. Baar, S.
Blechert, Chem. Eur. J. 2015, 21, 526–530.
[7] a) C. Liu, L. Q. Jing, L. M. He, Y. B. Luan, C. M. Li, Chem. Commun. 2014,
50, 1999–2001; b) T. Sano, S. Tsutsui, K. Koike, T. Hirakawa, Y. Teramoto,
N. Negishi, K. Takeuchi, J. Mater. Chem. A 2013, 1, 6489–6496.
[8] S. X. Min, G. X. Lu, J. Phys. Chem. C 2012, 116, 19644–19652.
[9] D. S. Wang, H. T. Sun, Q. Z. Luo, X. L. Yang, R. Yin, Appl. Catal. B 2014,
156–157, 323–330.
[10] a) X. J. Wang, W. Y. Yang, F. T. Li, Y. B. Xue, R. H. Liu, Y. J. Hao, Ind. Eng.
Chem. Res. 2013, 52, 17140–17150; b) X. Chen, X. Huang, Z. Yi, Chem.
Eur. J. 2014, 20, 17590–17596; c) H. Katsumata, T. Sakai, T. Suzuki, S.
Kaneco, Ind. Eng. Chem. Res. 2014, 53, 8018–8025; d) H. Xu, J. Yan, Y.
Xu, Y. Song, H. Li, J. Xia, C. Huang, H. Wan, Appl. Catal. B 2013, 129,
182–193; e) X. J. Wang, Q. Wang, F. T. Li, W. Y. Yang, Y. Zhao, Y. J. Hao,
S. J. Liu, Chem. Eng. J. 2013, 234, 361–371; f) Y. P. Zang, L. P. Li, X. G. Li,
R. Lin, G. S. Li, Chem. Eng. J. 2014, 246, 277–286; g) Y. F. Chen, W. X.
Huang, D. L. He, Y. Situ, H. Huang, ACS Appl. Mater. Interfaces 2014, 6,
14405–14414; h) S. W. Zhang, J. X. Li, M. Y. Zeng, G. X. Zhao, J. Z. Xu,
W. P. Hu, X. K. Wang, ACS Appl. Mater. Interfaces 2013, 5, 12735–12743.
[11] J. R. Ran, J. Zhang, J. G. Yu, M. Jaroniec, S. Z. Qiao, Chem. Soc. Rev. 2014,
43, 7787–7812.
[12] a) F. T. Li, Y. Zhao, Y. J. Hao, X. J. Wang, R. H. Liu, D. S. Zhao, D. M. Chen,
J. Hazard. Mater. 2012, 239, 118–127; b) F. Tzompantzi, Y. PiÇaa, A. Man-
tilla, O. Aguilar-Martínez, F. Galindo-Hernµndez, X. Bokhimi, A. Barrera,
Catal. Today 2014, 220–222, 49–55; c) P. P. Hankare, R. P. Patil, A.
Jadhav, K. M. Garadkar, R. Sasikala, Appl. Catal. B 2011, 107, 333–339.
[13] F. T. Li, Y. Zhao, Q. Wang, X. J. Wang, Y. J. Hao, R. H. Liu, D. S. Zhao, J.
Hazard. Mater. 2015, 283, 371–381.
[33] X. L. Xu, X. Duan, Z. G. Yi, Z. W. Zhou, X. M. Fan, Y. Wang, Catal.
Commun. 2010, 12, 169–172.
[14] F. T. Li, Y. B. Xue, B. Li, Y. J. Hao, X. J. Wang, R. H. Liu, J. Zhao, Ind. Eng.
Chem. Res. 2014, 53, 19540–19549.
Received: January 18, 2015
Published online on June 3, 2015
[15] W. D. Zhang, Y. J. Sun, F. Dong, W. Zhang, S. Duan, Q. Zhang, Dalton
Trans. 2014, 43, 12026–12036.
Chem. Eur. J. 2015, 21, 10149 – 10159
www.chemeurj.org
10159
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim