Template-free preparation and properties of mesoporous g-C 3N4/TiO2 nanocomposite photocatalyst

Article abstract of DOI:10.1039/c3ce42513d

A novel and facile template-free method was presented to fabricate a network structured mesoporous g-C₃N₄/TiO₂ nanocomposite with enhanced visible-light photocatalytic activity. The protonation effect accompanied with TiO₂ nanoparticle growth is believed to be a probable mechanism for the formation of such a porous structure of the g-C₃N₄ matrix. The Royal Society of Chemistry 2014.

Full text of DOI:10.1039/c3ce42513d

CrystEngComm
View Article Online
View Journal | View Issue
COMMUNICATION
Template-free preparation and properties of
mesoporous g-C N /TiO nanocomposite
photocatalyst†
3
4
2
Cite this: CrystEngComm, 2014, 16,
868
1
a
a
ab
a
a
Jianchao Shen, Hui Yang, Qianhong Shen,* Yu Feng and Qifeng Cai
Received 9th December 2013,
Accepted 3rd January 2014
DOI: 10.1039/c3ce42513d
www.rsc.org/crystengcomm
A novel and facile template-free method was presented to fabri-
without templates is of particular interest. Recently, it was
reported that the direct protonation of the base functionali-
ties in g-C N can make bulk g-C N be broken up from its
3 4 3 4
cate
composite with enhanced visible-light photocatalytic activity. The
protonation effect accompanied with TiO nanoparticle growth is
believed to be a probable mechanism for the formation of such a
porous structure of the g-C matrix.
a
network structured mesoporous g-C
3
N
4
/TiO
2
nano-
2
defects, thereby forming the relatively fine fragment parti-
6
cles. This strategy has provided a convenient route to raise
3
N
4
3 4
the specific surface area of bulk g-C N .
On the other hand, to increase the quantum yield of
g-C N , many methods have been developed to inhibit the
3 4
Graphitic carbon nitride (g-C N ), as a novel visible light
3
4
driven metal-free photocatalyst, has stimulated intensive
interest in the photocatalytic field due to its high stability,
recombination of photogenerated electron–hole pairs, such
7
as doping with foreign elements, and coupling with
1
appealing electronic structure and medium band gap. How-
8
9
10
graphene, noble metal or other semiconductors. Among
them, coupling g-C with other components is a promising
ever, the photocatalytic activity of the bulk g-C N is limited
3
4
3 4
N
by its low specific surface area and poor quantum yield, so a num-
strategy for improving the charge separation in photo-
catalysis, especially when the nanojunction structure is formed
in the composites. This type of structure will promote
more efficient interparticle electron transfer due to the
size effect of particles and intimate contact between hetero-
2
ber of strategies have been employed to modify bulk g-C N .
3
4
To increase the specific surface area and provide more
active sites for adsorption and photocatalytic reaction, vari-
ous soft or hard templates are usually employed to introduce
3
pores in g-C N . The pores can be fabricated after the
10
3
4
phase nanoparticles.
removal of templates, but this extra template removal step
may induce an adverse impact on the preparation and prop-
erties of as-prepared g-C N . For example, the most used
In this communication, we report a novel and facile chem-
ical method to fabricate a mesoporous g-C N /TiO nano-
3 4 2
composite based on the protonation of base functionalities
as well as the in situ growth of TiO2 nanoparticles on the
3
4
hard templates of porous silica are usually removed by dis-
solving it in NH HF , HF or NaOH aqueous solutions, which
4
2
3 4 3 4 2
g-C N surface. The as-prepared g-C N /TiO nanocomposite
4
are not environmentally friendly; while for the soft tem-
plates, such as a variety of surfactants and amphiphilic block
polymers, it is difficult to avoid the increase of carbon con-
tent in the resulting materials as well as the generation of
more potential recombination centers in semiconductors,
exhibits efficient interparticle electron transfer, achieving the
enhanced quantum efficiency. More interestingly, the proton-
ation effect induces formation of unexpected network struc-
tured mesopores in the bulk g-C
3
N
4
rather than finely
6
fragmented g-C N particles. To the best of our knowledge,
3
4
5
lowering the photocatalytic activity. Therefore, the develop-
this work is the first report of such a phenomenon and also
ment of new procedures for synthesizing porous g-C N
3
4
provides a new idea for preparing mesoporous g-C
3 4
N based
materials through a template-free route.
a
State Key Laboratory of Silicon Materials, Department of Materials Science &
Engineering, Zhejiang University, Hangzhou 310027, PR China.
To understand the formation mechanism of mesoporous
g-C /TiO (denoted as CN–Ti), the pure g-C modified in
the ethanol–water solution (denoted as CN–H) and the H SO
4
3
N
4
2
3 4
N
E-mail: s_qianhong@163.com; Fax: +86 571 87953313; Tel: +86 0571 87953313
b
2
Zhejiang California International NanoSystems Institute, Zhejiang University,
solution (denoted as CN–S when the solution has the same
pH value as that of the g-C /TiO preparation system, and
Hangzhou 310027, PR China
N
4
2
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c3ce42513d
3
CN–SS when the solution has an even lower pH value) were
1868 | CrystEngComm, 2014, 16, 1868–1872
This journal is © The Royal Society of Chemistry 2014

View Article Online
CrystEngComm
Communication
also prepared by a similar procedure for reference. The for-
mation mechanisms of various modified g-C N samples are
illustrated in Scheme 1.
dosage in the reaction system, the porosity of g-C N will be
3
4
increased more obviously, further confirming the promotion
effect of protonation on the formation of a mesoporous
structure (Fig. S1d, ESI†).
3
4
Under the hydrothermal conditions, the water has higher
reaction activity than itself in a normal state due to the
increase of its self-ionization constant, leading to the proton-
ation of base functionalities in g-C N (Case I in Scheme 1).
However, when introducing Ti(SO4)2 as the precursor of
2
TiO (Case III in Scheme 1), a network porous structure
appears in the g-C N matrix, which is quite different from
3
4
3 4
As shown in Fig. 1a, the diffraction peak of the pure g-C N
that of the CN–S sample although they have the same prepa-
ration condition of the pH value (Fig. 2a). At the beginning of
the hydrothermal reaction, the intense protonation occurs in
the presence of H SO , which is generated from the hydroly-
3
4
without modification (denoted as CN) at 2θ = 13.1°, which
represents the in-plane structural packing motif, disappears
after the hydrothermal reaction, indicating that the proton-
2
4
6
ation occurs. Moreover, the diffraction peak of the CN sam-
4 2
sis of Ti(SO ) , leading to the formation of some little pores
ple at 2θ = 27.1° shifts to higher angles in the XRD pattern of
the CN–H sample due to modified interlayer organization,
accompanied by compression of the average interlayer dis-
and microcracks, thereby increasing the defects on the sur-
face of g-C N . Meanwhile, in a low concentration aqueous
3
4
4
+
2+
Ti(SO
4 2 2 2 4
) solution, Ti exists as an octahedral [Ti(OH) (OH ) ]
1
1
13
tance towards smaller d values (Fig. 1b). Owing to the pro-
monomer or an edge-sharing dimer. The g-C N is favor-
able to the adsorption of [Ti(OH) (OH ) ] species on the
2 2 4
3
4
2
+
tonation effect, the g-C N decomposes to clusters/oligomers
3
4
6
of condensed triazine rings. As a result, parts of the clusters/
oligomers with good water solubility are removed by decom-
position under hydrothermal conditions, forming some spo-
formed g-C
bridging nitrogens via hydrogen bonding, thus providing
the nucleating sites for the growth of TiO nanocrystals.
As hydrothermal reaction proceeds, the solution is in a
3 4
N surface defects of terminal amino groups or
1
4
2
1
2
radic pores and cracks on the surface of the CN–H sample
2
+
(
Fig. S1b, ESI†). Furthermore, when modifying the g-C N in
saturated state, and the [Ti(OH) (OH ) ] species in solution
2 2 4
3
4
the H
2
SO
4
solution (Case II in Scheme 1), which has the
/TiO preparation system,
are unstable and prone to combine together through
oxolation or olation at the nucleating sites to form original
nuclei, and then the nuclei grow further into anatase type
same pH value as that of the g-C
3
N
4
2
the number of water-soluble clusters/oligomers may increase
due to the more intense protonation effect at a higher proton
concentration, so the porosity of the CN–S sample increased
1
3
crystallites (Fig. 1 and 2). The increasing particle size of
TiO nanoparticles at the formed defects, such as little pores
and microcracks, may result in the increase of tensile stress
on the surface of g-C , and thus forming more cracks.
2
(Fig. S1c, ESI†). Actually, when further increasing the H SO
2 4
3 4
N
Therefore, more protons will diffuse into the inner layer of
g-C N through the new formed cracks, inducing a new
3
4
round of the protonation and decomposition. Finally, the
network mesoporous structure is formed, and the specific
2
−1
2
−1
surface area is increased from 3.53 m
g
to 21.21 m g
(
Fig. S2, ESI†). According to the results of particle size distri-
bution (Fig. S3, ESI†), the increase of the specific surface area
is mainly caused by the network mesoporous structure rather
than g-C
layered structure of g-C
acting as a visible-light sensitizer, is well kept after the
3
N
4
particle refinement. Furthermore, the typical
, as well as the melem structure
3 4
N
1
5
intense protonation, and the obtained CN–Ti sample
3 4
Scheme 1 Formation mechanism of various modified g-C N samples.
Fig. 2 Electron microscope images of the as-prepared CN–Ti samples.
(a) SEM image, and the inset is the corresponding image of low magni-
Fig. 1 XRD patterns of various as-prepared samples of CN, CN–H,
CN–Ti and TiO in the range of (a) 2θ = 10°–70°; (b) 2θ = 24°–30°.
fication. (b) TEM image, and the inset is the corresponding image of
high resolution.
2
This journal is © The Royal Society of Chemistry 2014
CrystEngComm, 2014, 16, 1868–1872 | 1869

View Article Online
Communication
CrystEngComm
contains two fundamental components which are g-C N and
3
4
anatase TiO2 (Fig. 1 and S4, ESI†). These two components
have formed a close interface (Fig. 2b), which is favorable for
the highly efficient interparticle electron transfer.
Fig. 3a shows UV-vis diffuse reflectance spectra of
as-prepared samples. It can be seen that the CN sample exhibits
an intrinsic semiconductor-like absorption in the blue region
of the visible spectrum with a band edge wavelength around
4
58 nm. The UV-vis diffuse reflectance spectrum of the CN–H
sample indicates that the optical band gap and the semicon-
ductor properties are maintained, while the band edge wave-
length blue shifts from 458 nm to 451 nm. After combining
g-C N with TiO , the optical absorption of the CN–Ti sample
3
4
2
decreased slightly in the region of 350–450 nm, while it
enhanced in the region of ultraviolet light below 330 nm. This
phenomenon can be contributed to the optical response of
TiO2 and g-C N as well as the combination of these two
3
4
Fig.
4
Photocatalysis process analysis of as-prepared samples.
semiconductors. Furthermore, the band gaps determined
(a) Photocatalytic degradation and corresponding kinetics of MO over
1
/2
from the (αhν) versus photon-energy plots are about 2.71 eV
CN), 2.75 eV (CN–H), 2.75 eV (CH–Ti), 3.18 eV (TiO ),
P25 particles and as-prepared samples of TiO , CN, CN–H, CN–Ti under
visible-light irradiation. The inset shows the proposed electron–hole
2
(
2
3 4 2
separation process of the g-C N /TiO nanojunction under visible-light
respectively. Fig. 3b illustrates PL emission spectra of
as-prepared samples monitored at an excitation wavelength of
irradiation. (b) UV-vis spectra of MO after absorption by the CN, CN–H
and CN–Ti samples in the dark for 2 h. The inset displays the corre-
sponding surface structure of CN–H and CN–Ti samples.
3
25 nm. The CN sample shows a typical PL spectrum with the
emission peak at 475 nm. However, the PL curve of the CN–H
sample is split into two peaks with the emission peak around
4
40 nm and 460 nm, and the fluorescence intensity increased
apparent first-order rate constant. It can be seen that the
compared with the CN sample. This phenomenon is believed
2
commercial P25 and as-prepared TiO show rather weak
to be a result of the protonation of base functionalities in
activities as they work only with ultraviolet irradiation, while
6
g-C
3
N
4
. After introducing TiO
2
nanoparticles on the surface
the CN–Ti sample exhibits an excellent photocatalytic degra-
−
1
of g-C
3
N
4
, the fluorescence intensity of the CN–Ti sample
dation rate of 0.234 h , which is about 29.3, 13.8, 11.7 and
significantly reduced, indicating that the recombination of
photo-generated charge carriers is greatly inhibited due to the
11.4 times as high as that of the P25, TiO , CN and CN–H
2
sample, respectively. As discussed above, the CN–Ti sample
shows a much higher specific surface area than that of other
samples and becomes positively charged due to protonation,
thereby inducing a higher adsorption capacity to negatively
charged MO molecules (Fig. 4b).
Fig. 5a illustrates UV-vis absorption spectra of the MO
solution photocatalytically degraded by the CN–Ti sample for
different time intervals under visible light. It can be seen that
3 4 2
efficient electron transfer between g-C N and TiO . This
result shows good agreement with other heterojunction
1
0
semiconductors.
The photocatalytic performance of as-prepared samples is
evaluated by methyl orange (MO) photocatalytic degradation
under visible-light (λ > 420 nm) irradiation. The correspond-
0
ing first-order kinetics plot by the equation ln(C /C) = kt is
shown in Fig. 4a, where C and C are the MO concentrations
0
of the solution at times 0 and t, respectively, and k is the
Fig. 5 Photocatalysis process analysis of the CN–Ti sample. (a) UV-vis
spectra of MO solution photocatalytic degraded by the CN–Ti sample
for different time intervals under visible light. The insets show the color
change of the corresponding resulting solution and peak fitting of a
typical UV-vis absorption spectrum of the resulting solution after 2 h
irradiation. (b) The HPLC chromatograms of the MO solution photo-
catalytically degraded by the CN–Ti sample for different time intervals
under visible light.
Fig. 3 Optical property of as-prepared samples. (a) UV-vis diffuse
reflectance spectra of various samples. The inset is the band gap of as-
1/2
obtained samples determined from the (αhν) versus photon-energy
plots. (b) PL emission spectra of various samples monitored at an exci-
tation wavelength of 325 nm.
1870 | CrystEngComm, 2014, 16, 1868–1872
This journal is © The Royal Society of Chemistry 2014

View Article Online
CrystEngComm
Communication
the characteristic absorption peak of the initial MO solution
at 464 nm, which is associated with the azo bond (–NN–),
Conclusions
1
6
shows a large decrease in intensity simultaneously with a sig-
nificant blue shift during the photocatalysis process,
resulting in the distinct color change from orange to light
yellow-green. By the means of peak fitting, the broadened
absorption peak can be further divided into three peaks
located at 464 nm, 438 nm and 380 nm (the inset of Fig. 5a).
This suggests that some intermediates may be generated
and the color of the resulting solution comes from the
combination of the degraded MO chromophore ring and the
intermediates. The resulting solutions from photocatalytic
degradation are further analyzed by using HPLC, as shown in
Fig. 5b. There can be found a fast intensity decrease of the
MO peak at 8.5 min and the appearance of two new peaks
emerged at retention times of 6.7 (IM1) and 4.3 (IM2) min.
This result further confirms the generation of intermediates
during the photocatalytic degradation.
In summary, we developed a novel and facile template-free
method to fabricate the mesoporous g-C N based material.
3
4
Protonation of base functionalities lets g-C N decompose
3
4
into clusters/oligomers of condensed triazine rings, some of
which exhibit good water solubility and can be removed
under hydrothermal conditions. Moreover, the in situ growth
2 3 4
of TiO nanoparticles on the surface of g-C N not only forms
a g-C N /TiO nanojunction with a more efficient inter-
3
4
2
particle electron transfer but also promotes further proton-
ation, thereby increasing the porosity of the g-C matrix.
The as-prepared mesoporous g-C N /TiO nanocomposite
3 4
N
3
4
2
combined the advantages of visible-light response, efficient
charge separation and high adsorption ability for a contami-
nant molecule, thus exhibiting highly efficient photocatalytic
activity under visible light. Studies focused on the perfor-
mance optimization of this functionalized g-C
3 4
N and cou-
Indeed, the N-demethylation of MO will occur due to the
attack of hydroxyl radicals on the N,N-dimethyl groups, gen-
erating aminoazobenzene-4′-sulfonic acid sodium salt,
pling g-C N with other semiconductors are ongoing.
3
4
1
7
Acknowledgements
which corresponds to the IM1 peak in the HPLC chromato-
gram. Therefore, as the photocatalytic reaction proceeds, the
IM1 peak increases gradually with the decrease of MO in the
initial 3 hours, resulting in the blue shift of the main absorp-
tion peak position from 464 nm to 438 nm in UV-vis spectra.
Subsequently, the aminoazobenzene-4′-sulfonic acid sodium
salt is further oxidized by hydroxyl radicals, producing
p-nitroaniline due to the cleavage of an azo-bond. The
p-nitroaniline has a characteristic UV-vis absorption peak at
We would like to thank the Key Technologies R&D Program
of China under grant no. 2013BAJ10B05, the Zhejiang Provin-
cial Natural Science Foundation of China under grant no.
LQ12E02008 and the Fundamental Research Funds for the
Central Universities under grant no. ZJUR011104.
Notes and references
1
8
3
80 nm, which is in good agreement with the analysis of
peak fitting. Considering the molecule polarity and the
response characteristic under the detection wavelength of
HPLC, the small peak at 4.5 min, which obviously increased
with the decrease of the IM1 peak after photocatalytic reac-
tion for 6 h, may attributed to p-nitroaniline. Hence, it can
be speculated that the possible photocatalytic degradation
pathway for MO in this reaction system involves N-demethylation
of MO to generate aminoazobenzene-4′-sulfonic acid sodium
salt, followed by cleavage of the azo-bond to yield kinds of
1 (a) X. Wang, K. Maeda, A. Thomas, K. Takanabe, G. Xin,
J. M. Carlsson, K. Domen and M. Antonietti, Nat. Mater.,
2009, 8, 76; (b) X. H. Li, X. Wang and M. Antonietti, Chem.
Sci., 2012, 3, 2170.
2 (a) X. Bai, L. Wang, R. Zong and Y. Zhu, J. Phys. Chem. C,
2013, 117, 9952; (b) Y. Wang, Z. Wang, S. Muhammad and
J. He, CrystEngComm, 2012, 14, 5065.
3 (a) F. Su, S. C. Mathew, G. Lipner, X. Fu, M. Antonietti,
S. Blechert and X. Wang, J. Am. Chem. Soc., 2010, 132, 16299;
(b) S. Yang, W. Zhou, C. Ge, X. Liu, Y. Fang and Z. Li,
RSC Adv., 2013, 3, 5631; (c) H. Yan, Chem. Commun.,
2012, 48, 3430.
4 (a) F. Goettmann, A. Fischer, M. Antonietti and A. Thomas,
Angew. Chem., Int. Ed., 2006, 45, 4467; (b) X. Wang,
K. Maeda, X. Chen, K. Takanabe, K. Domen, Y. Hou, X. Fu
and M. Antonietti, J. Am. Chem. Soc., 2009, 131, 1680.
5 Y. Wang, X. C. Wang, M. Antonietti and Y. J. Zhang,
ChemSusChem, 2010, 3, 435.
1
9
intermediates.
Based on above discussion and analysis, the highly effi-
cient photocatalytic activity is mainly ascribed to the follow-
ing two aspects: on the one hand, owing to the delocalized
3 4 2
conjugated π structure of g-C N , its coupling with TiO hav-
ing a wide band gap is an appropriate approach to achieve
1
0
an improved charge separation (the inset of Fig. 4a), lead-
ing to efficient generation of hydroxyl radicals. These radicals
are the key for an efficient photocatalytic reaction in the case
2
0
of g-C
3
N
4
based photocatalyst. On the other hand, the net-
6 Y. Zhang, A. Thomas, M. Antonietti and X. Wang, J. Am.
Chem. Soc., 2009, 131, 50.
work mesoporous structure in the CN–Ti sample will improve
the adsorption capacity for the target pollutants and provide
7 (a) J. Li, B. Shen, Z. Hong, B. Lin, B. Gao and Y. Chen, Chem.
Commun., 2012, 48, 12017; (b) Y. J. Zhang, T. Mori, J. H. Ye
and M. Antonietti, J. Am. Chem. Soc., 2010, 132, 6294.
8 Q. Xiang, J. Yu and M. Jaroniec, J. Phys. Chem. C,
2011, 115, 7355.
more active sites to accelerate the photocatalytic degrada-
3
tion. Therefore, the as-prepared mesoporous g-C
3
N
4
/TiO
2
nanocomposite exhibits highly efficient photocatalytic activity
under visible light.
This journal is © The Royal Society of Chemistry 2014
CrystEngComm, 2014, 16, 1868–1872 | 1871

View Article Online
Communication
CrystEngComm
9
L. Ge, C. Han, J. Liu and Y. Li, Appl. Catal., A,
011, 409–410, 215.
14 T. Saplinova, V. Bakumov, T. Gmeiner, J. Wagler, M. Schwarz
and E. Kroke, Z. Anorg. Allg. Chem., 2009, 635, 2480.
15 D. Mitoraj and H. Kisch, Angew. Chem., Int. Ed.,
2008, 47, 9975.
16 M. Hou, F. Li, X. Liu, X. Wang and H. Wan, J. Hazard.
Mater., 2007, 145, 305.
2
1
1
1
1
0 (a) S. Kumar, T. Surendar, A. Baruah and V. Shanker,
J. Mater. Chem. A, 2013, 1, 5333; (b) S. C. Yan, S. B. Lv,
Z. S. Li and Z. G. Zou, Dalton Trans., 2010, 39, 1488.
1 A. Thomas, A. Fischer, F. Goettmann, M. Antonietti,
J. Müller, R. Schlögl and J. M. Carlsson, J. Mater. Chem.,
17 C. Galindo, P. Jacques and A. Kalt, J. Photochem. Photobiol.,
A, 2000, 130, 35.
2
008, 18, 4893.
2 M. Shahbaz, S. Urano, P. R. LeBreton, M. A. Rossman,
18 C. Y. Chiu, P. J. Chung, K. U. Lao, C. W. Liao and
M. H. Huang, J. Phys. Chem. C, 2012, 116, 23757.
19 J. R. Darwent and A. Lepre, J. Chem. Soc., Faraday Trans. 2,
1986, 82, 1457.
R. S. Hosmane and N. J. Leonard, J. Am. Chem. Soc.,
1
984, 106, 2805.
3 Y. Zheng, E. Shi, Z. Chen, W. Li and X. Hu, J. Mater. Chem.,
001, 11, 1547.
2
20 S. C. Yan, Z. S. Li and Z. G. Zou, Langmuir, 2009, 25, 10397.
1872 | CrystEngComm, 2014, 16, 1868–1872
This journal is © The Royal Society of Chemistry 2014