One-step synthesis of mesoporous Pt-Nb2O5 nanocomposites with enhanced photocatalytic hydrogen production activity

Article abstract of DOI:10.1039/c6ra19118e

Mesoporous Pt-Nb₂O₅ photocatalysts with high photocatalytic hydrogen production activity were synthesized by a one-step method. The mesoporous structure of the as-made samples was characterized by X-ray diffraction, N₂ adsorption-desorption isotherms and transmission electron microscopy. Pt-Nb₂O₅ exhibited higher hydrogen production activity than pure Nb₂O₅ and 0.8-Pt-D-M-Nb exhibited the best photocatalytic activity. In addition, Pt-Nb₂O₅ synthesized by a one-step method exhibited higher photocatalytic hydrogen production activity than that prepared by an incipient wetness impregnation method; 0.8-Pt-D-M-Nb exhibits a H₂ evolution rate of 9.79 mmol h^-1 g^-1, while 0.8-Pt-I-M-Nb₂O₅ exhibits a H₂ evolution rate of 7.73 mmol h^-1 g^-1. The enhanced photocatalytic activity of Pt-Nb₂O₅ was predominantly attributed to the efficient separation of photoinduced electrons and holes.

Full text of DOI:10.1039/c6ra19118e

RSC Advances
View Article Online
View Journal | View Issue
PAPER
One-step synthesis of mesoporous Pt–Nb O
2
5
nanocomposites with enhanced photocatalytic
Cite this: RSC Adv., 2016, 6, 96809
hydrogen production activity†
Hualei Zhang, Qianying Lin, Shangbo Ning, Yangen Zhou, Huaxiang Lin,* Jinlin Long,
Zizhong Zhang and Xuxu Wang*
Mesoporous Pt–Nb
synthesized by
characterized by X-ray diffraction, N
microscopy. Pt–Nb exhibited higher hydrogen production activity than pure Nb
M–Nb exhibited the best photocatalytic activity. In addition, Pt–Nb synthesized by a one-step
2
O
5
photocatalysts with high photocatalytic hydrogen production activity were
one-step method. The mesoporous structure of the as-made samples was
adsorption–desorption isotherms and transmission electron
and 0.8-Pt–D–
a
2
2
O
5
2 5
O
2 5
O
method exhibited higher photocatalytic hydrogen production activity than that prepared by an incipient
Received 28th July 2016
Accepted 30th September 2016
ꢀ1 ꢀ1
wetness impregnation method; 0.8-Pt–D–M–Nb exhibits a H
while 0.8-Pt–I–M–Nb exhibits evolution rate of 7.73 mmol
photocatalytic activity of Pt–Nb was predominantly attributed to the efficient separation of
photoinduced electrons and holes.
2
evolution rate of 9.79 mmol h
g ,
ꢀ1
ꢀ1
2
O
5
a
H
2
h
g . The enhanced
DOI: 10.1039/c6ra19118e
2 5
O
www.rsc.org/advances
photoinduced electron–hole pairs, which ultimately leads to
excellent photocatalytic activity. Mesoporous materials have
tunable structures, a tailored framework, high surface areas and
1
. Introduction
Aer Honda and Fujishima’s discovery of photocatalytic H
2
1
evolution over a TiO
2
electrode in 1972, heterogeneous semi- alternative pore shapes, which shorten the transfer path of
conductor photocatalysis as a promising environmentally photogenerated carriers from the bulk to the surface of the
friendly technology for the conversion of solar energy into photocatalyst, and therefore decrease the electron–hole
6–8
chemical energy has attracted augmenting attention. In the past combination rate. It has been reported that the photocatalytic
0 years, tremendous efforts have been devoted to investigate activity of mesoporous Nb was 20 times higher than that of
4
2 5
O
4
various inorganic and organic systems as photocatalysts for the unstructured one.
water splitting. Some semiconductor materials such as Nb
ZnO, a-Fe
catalytic H
Among these semiconductor materials, niobium oxide as an smallest H
n-type photocatalyst with a wide band gap (3.1–3.5 eV) has loading onto the semiconductor to enhance the photocatalytic
2
2
O
5
,
Another method to improve the photocatalytic activity is
, and ZnS have been applied to photo- loading the semiconductor with metal particles. Among the
3
2
O
3
, WO
3,4
many tested metals, Pt has the highest work function and the
2
production in aqueous solution.
evolution over-potential and is quite suitable for
2
9
become widely prevalent because of its unique physical prop-
H
2
production capacity. Platinum nanoparticles loaded onto
erties such as a high refractive index, good chemical and TiO
thermal stability and good photocatalytic characteristics. But efficiency of the photochemical reaction process.
2
surfaces have been demonstrated to enhance the quantum
5
10,11
It is sug-
the high recombination rate of photo-induced hole–electron gested that the Pt serves as a trapping centre for electrons
pairs hampers the photocatalytic efficiency. Many efforts have generated in light activated TiO . There are also some reports
been made to improve the photocatalytic activity of niobium about Pt doped Nb . Some groups loaded Pt particles onto
oxide by controlling the morphology, surface structure and mesoporous Nb by photo-deposition to study the photo-
surface modication.
chemical hydrogen production. The results show that aer
2 5
Meanwhile, constructing mesoporous semiconductors is loading with Pt particles, the photocatalytic activity of Nb O
2
O
2 5
O
2 5
5,12
one of the effective methods to prompt the separation of was greatly improved.
However, many methods for loading Pt on semiconductors
mostly involve a multi-step synthesis, such as thermal impreg-
nation and chemical and photo-deposition reduction
Research Institute of Photocatalysis, State Key Laboratory of Photocatalysis on Energy
and Environment, Fuzhou University, Fuzhou 350002, PR China. E-mail: lhx@fzu.edu.
cn
13–16
methods.
These methods can load Pt onto semiconductor
surfaces well but have some shortcomings. The thermal
impregnation method is simple, convenient, and easily
†
Electronic supplementary information (ESI) available. See DOI:
0.1039/c6ra19118e
1
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 96809–96815 | 96809

View Article Online
RSC Advances
Paper
operated but the particle size of Pt is non-uniform and the refers to the controlled volume ratio of Pt (x ¼ 0.1%, 0.3%,
activity of the catalyst needs to be improved. The chemical 0.50%, 0.80%, 1.0%, 1.5%), which would also be abbreviated to
reduction method oen needs chemical agents in the reduction 0.1-Pt–D–M–Nb, 0.3-Pt–D–M–Nb, 0.5-Pt–D–M–Nb, 0.8-Pt–D–M–
process, such as formaldehyde and hydrazine or sodium boro- Nb, 1.0-Pt–D–M–Nb, and 1.5-Pt–D–M–Nb.
hydride, which are not environmentally friendly. Photo-
(2) Incipient wetness impregnation method. 0.4 g of meso-
and 20 ml ethanol was added into a special vase.
method for preparing Pt co-catalysts, but the distribution of Pt Then the HPtCl6 solution was added and the Pt content was
nanoparticles on the surface of the semiconductors is non- 0.8% wt relative to the mesoporous Nb . Aer ultrasonication
deposition is an efficient and environmentally friendly porous Nb
2 5
O
17
2 5
O
18
uniform. Recently, Pt has been loaded onto TiO
2
with uniform for 30 min, the ethanol was removed by a rotary evaporation
size and good distribution by a single-step method and Pt/TiO2 method. The obtained buff solid was rstly calcined at 350 C
exhibits higher photocatalytic activity than that prepared by an for 6 h and then at 450 C for 6 h. The Pt nanoparticle loading
ꢁ
ꢁ
incipient wetness impregnation method (IWI) and PCD was denoted as 0.8-Pt–I–M–Nb.
16
methods.
In this work, Pt was loaded onto mesoporous Nb
2 5
O by a one-
2.3. Sample characterization
step method. Compared with the traditional method, this
method was simple, convenient and it was easier to obtain
a uniform size and high distribution of Pt on mesoporous
Nb O . Pt–Nb O exhibited higher photocatalytic H evolution
The samples were characterized by XRD recorded on a Bruker
D8 Advance X-ray diffractometer equipped with a Cu Ka X-ray
˚
source (l ¼ 1.5406 A). The nitrogen adsorption–desorption
2
5
2
5
2
isotherms were measured on a Micromeritics apparatus model
activity than pure Nb O . In addition, Pt–Nb O also exhibited
2
5
2 5
ASAP 2020 using the Brunauer–Emmett–Teller (BET) method at
much higher photocatalytic activity than that prepared by an
incipient wetness impregnation method (IWI). Finally,
a possible mechanism for enhancement of the photocatalytic
7
2
7 K aer the sample had been degassed in a ow of N at 160
ꢁ
C for 4 h. The BET surface area was calculated from the linear
part of the BET plot (P/P
0
¼ 0.05–0.30). The average pore size
2 5
activity in Pt–Nb O was also discussed.
and the pore size distribution plots were estimated using the
adsorption branch of the isotherm and the Barrett–Joyner–
Halenda (BJH) formula. Transmission electron microscopy
2. Experimental
(
2
TEM) and HRTEM images were obtained on a JEOL model JEM
010 EX microscope at an accelerating voltage of 200 kV. X-ray
photoelectron spectroscopy was carried out on an ESCALAB
50 XPS system with a mono-chromatized Al Ka X-ray source (15
2.1. Materials
Pluronic P-123, supplied by Aladdin, was used as a template.
Anhydrous niobium chloride supplied by Alfa Aesar was used as
the metal source. Anhydrous ethanol, used as the solvent, was
obtained from the Nanjing Chemical Reagent Company.
Chloroplatinic acid was purchased from Aladdin. It was
2
kV, 200 W, 500 mm, pass energy ¼ 20 eV). All binding energies
were referenced to the C 1s peak at 284.58 eV of surface
adventitious carbon. Photo-luminescence (PL) measurements
were obtained on an Edinburgh Analytical Instrument PLS920
FL spectrophotometer and a 320 nm He–Cd laser was used as an
excitation light source.
ꢀ1
prepared as a solution with a Pt content of 7.53 g L . DMPO is
an abbreviation of 5,5-dimethyl-1-pyrroline-N-oxide, supplied
by Aladdin, and was used as the trapping agent.
2.2. Synthesis of the mesoporous photo-catalysts
2.4. Photocatalytic activity measurements
Mesoporous Nb O was synthesized by versatile evaporation The photocatalytic methanol reforming reactions to generate
2
5
19
induced self-assembly (EISA). 0.6 g of Pluronic P-123 was
2
H were carried out in a closed gas circulation system. The
dissolved in 20 ml anhydrous ethanol followed by addition of reaction cell was equipped with an inner irradiation quartz
2
.3 mmol NbCl
5
ꢁ
under vigorous stirring. The as-prepared sol reaction vessel and a cooling water jacket to keep the temper-
ꢁ
was kept at 40 C for 2 days and then a gel was obtained. ature at 10 C. A scheme of the photo-reactor is shown in
Subsequently, the gel was calcined rstly at 350 C for 6 h and Fig. S1.† 0.1 g of photocatalyst powder was suspended in 165 ml
then at 450 C for 6 h. The as-prepared white powder was named distilled H O in the inner quartz cell (Fig. S1†) and stirred at
ꢁ
ꢁ
2
M–Nb.
a speed of 550 rpm. Then 5 ml methanol was added to the
The mesoporous Nb O loaded with Pt nanoparticles was suspension as an oxidizable reducing reagent. A 165 W high-
2
5
synthesized by two different methods.
pressure mercury lamp was used to provide a UV light source.
(1) One-step method. 0.6 g of Pluronic P-123 was dissolved in The electronic spectrum of the lamp is shown in Fig. S2.† Prior
anhydrous ethanol followed by addition of 2.3 mmol NbCl
5
to irradiation, the reaction system was evacuated using
6
under continuous stirring. Then the HPtCl solution was added a mechanical pump and then lled with 101 kPa of high-purity
drop wise to the above solution. The as-prepared sol was kept at Ar (>99.99%). This process was repeated three times in order to
ꢁ
4
0 C for 2 days and then a gel was obtained. Subsequently, the remove O from the system. At given irradiation time intervals,
2
ꢁ
ꢁ
gel was calcined rstly at 350 C for 6 h and then at 450 C for 6 the generated H was circulated using a micro diaphragm gas
2
h. Thereby, a series of Nb O compounds loaded with different pump in the system and its amount was determined using an
2
5
Pt contents were prepared by adjusting the amount of HPtCl
6
.
online gas chromatograph (GC112A, Shanghai Precision
The prepared samples were denoted as x-Pt–D–M–Nb, where x Scientic Instrument Co, Ltd, TCD, Ar carrier).
96810 | RSC Adv., 2016, 6, 96809–96815
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
ꢁ
ꢁ
peaks at 25 and 55 , which is in accordance with Lin and
3
. Results and discussion
4,12
Chen’s report.
phase by a lack of long-range order.
Those broad peaks indicate an amorphous
3.1. Sample characterization
XRD analysis was carried out to investigate the mesoporous
Fig. 2 shows nitrogen adsorption–desorption isotherms
structure and phase structures of the catalysts. In the wide- and BJH pore size distribution plots of M–Nb, 0.8-Pt–D–M–Nb
angle XRD patterns (Fig. 1b), two broad peaks were observed and 0.8-Pt–I–M–Nb. From the N
2
adsorption–desorption
ꢁ
at 25 and 55 which can be indexed to the orthorhombic crys- isotherms, the samples (M–Nb, 0.8-Pt–D–M–Nb and 0.8-Pt–I–
tallized phase (Pbam 55, JCPDS no. #71-0336), indicating that M–Nb) showed the characteristics of mesoporous materials.
the three samples were in a low crystalline phase because of the But their size distributions did exhibit some differences,
low calcination temperatures. No platinum species were where the most probable aperture distribution of M–Nb was at
detected for 0.8-Pt–D–M–Nb and 0.8-Pt–I–M–Nb owing to the 5.37 nm and that of 0.8-Pt–D–M–Nb was at 6.22 nm. However,
small content of Pt.
the samples of 0.8-Pt–I–M–Nb possess an average pore diam-
eter of ca. 7.55 nm. In addition, the pore size distribution of
0
.8-Pt–I–M–Nb was wider than those of M–Nb and 0.8 Pt–D–M–
Nb. These results indicated that loading of Pt onto meso-
porous Nb by the incipient wetness impregnation method
3
.2. Pore structure
The pore structures of these materials were analysed by small-
angle XRD, nitrogen adsorption–desorption isotherms, TEM
and HRTEM. In Fig. 1a, the low-angle XRD patterns of M–Nb
and 0.8-Pt–D–M–Nb samples showed well-resolved peaks at 1.2
and 2.0 , indicating an ordered 2D-hexagonal mesoporous
structure (space group P6mm). The obvious low-angle X-ray
diffraction of 0.8-Pt–D–M–Nb showed that loading of plat-
inum on Nb
mesoporous structure. However, the low-angle XRD patterns of
0
2 5
O
more seriously damaged the mesoporous structure of Nb
than by the one-step method.
2 5
O
ꢁ
More structural information about Nb O , Pt–D–M–Nb and
2
5
ꢁ
Pt–D–M–Nb was provided by TEM analysis. The TEM image in
Fig. 3a reveals that Nb showed wormhole-like mesoporous
2 5
O
structures with a diameter of 4–6 nm in the pore channels and
the pores were well-ordered hexagonal arrays of mesoporous
symmetry. The TEM images in Fig. 3b of 0.8-Pt–D–M–Nb also
revealed well ordered two-dimensional pores. Compared with
M–Nb, the structure of 0.8-Pt–D–M–Nb has little changed.
However, the TEM of 0.8-Pt–I–M–Nb in Fig. 3c showed that
the wormhole-like mesoporous structures were damaged to
2 5
O by the one-step method did not break the
ꢁ ꢁ
.8-Pt–I–M–Nb have lower peaks at 1.2 and 2.0 . This indicated
that the mesoporous structure may have been damaged a little
by platinum during the impregnation process. In Fig. 1b, we can
see the M–Nb, 0.8-Pt–D–M–Nb and 0.8-Pt–I–M–Nb wide-angle
XRD patterns. All of them have two similar, extremely broad
Fig. 2 The N
2
adsorption–desorption isotherm (a) and Barret–Joy-
Fig. 1 The small-angle XRD (a) and wide-angle XRD patterns (b) of M– ner–Halenda (BJH) pore size distribution plots (b) of (red) M–Nb
2 5
O ,
Nb and the 0.8-Pt–D–M–Nb and 0.8-Pt–I–M–Nb samples.
(black) 0.8-Pt–D–M–Nb, and (blue) Pt–I–M–Nb.
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 96809–96815 | 96811

View Article Online
RSC Advances
Paper
Aer loading a small amount (0.1%) of Pt, the activity for
hydrogen production is greatly increased. In addition, with the
increase in the content of Pt, the hydrogen production activity
also gradually improved. Aer reaching a maximum, the
production activity declined with the increase in the content of
16
Pt. This phenomenon was similar to Sreethawong’s work,
which reported that there is a certain optimum content of Pt for
each synthesis method. In this work, the optimum amount of Pt
2
loading is 0.8% wt, which exhibits the H evolution rate 9.79
ꢀ
1
ꢀ1
mmol h
According to previous suggestions,
increase in the content of Pt, a mesoporous wall of Nb O may
g .
1
6,21
with the initial
2
5
carry more Pt nanoclusters that facilitate the separation of
photo-generated electron–hole pairs, and lead to an increase
in the photocatalytic activity. However, excess metal particles
loaded on the semiconductor may form a recombination
centre for electron–hole pairs. Despite the activities, the
stability of the photo-catalyst plays an important role in its
practical application. Thus, a stability test of 0.8-Pt–D–M–Nb
Fig. 3 TEM images obtained of M–Nb (a), 0.8-Pt–D–M–Nb (b), and
0
2 5
.8-Pt–I–M–Nb O (c) and HRTEM images of 0.8-Pt–D–M–Nb (d).
for H evolution under UV light irradiation was carried out 5
2
times in recycle experiments. As shown in Fig. S4,† the
hydrogen evolution rate is linear as a function of time and
remains almost unchanged aer 5 cycles of UV irradiation,
indicating the sufficient stability of 0.8-Pt–D–M–Nb for pho-
tocatalytic hydrogen evolution.
some extent. From Fig. 3d, HRTEM analysis of 0.8-Pt–D–M–Nb
showed that there are many black spots that were distributed
uniformly on Nb
2 5
O . Their lattice space is 0.22 nm which is well
matched with the (111) lattice spacing of the face-centre Pt
20
In addition to the content of Pt, the same Pt loading
content on Nb O prepared by different methods also
cubic phase, indicating that the Pt nanoparticles were
deposited uniformly on the Nb . However, the size of the Pt
2
5
2 5
O
exhibited different photocatalytic activity. As shown in Fig. 5,
particles was larger than that of those in Pt–D–M–Nb and the
distribution of Pt was non-uniform as shown in Fig. S3.† These
results suggest that the Pt particle loading on Nb O by the one-
ꢀ1
0
g
7
.8-Pt–D–M–Nb exhibits a H
2
evolution rate of 9.79 mmol h
exhibits a H evolution rate of
. The preparation methods have great
ꢀ1
, while 0.8-Pt–I–M–Nb
2
O
5
2
2
5
ꢀ
1
ꢀ1
.73 mmol h
g
step method has little inuence on the mesoporous structure of
Nb and the Pt particle loading on Nb by the impregna-
tion method to a certain extent damaged the mesoporous
structure of Nb
effects on the surface distribution, size of the Pt particles and
the chemical states of the Pt particles, which may inuence
the photocatalytic activity of Pt modied catalysts. As re-
2
O
5
2 5
O
2 5
O .
ported, Pt particles loaded on TiO by a Pt–citrate sol are very
2

nely dispersed and exhibit higher photocatalytic activity,
3
.3. Photocatalytic activity
but those that are prepared by an irradiation deposition
method agglomerated into large particles and show lower
The photocatalytic activities of M–Nb and Pt–D–M–Nb were
examined by the photocatalytic hydrogen production upon UV-
light irradiation as shown in Fig. 4. As can be seen, M–Nb
exhibits little photocatalytic activity for hydrogen production.
2
2
photocatalytic activity. They suggested that Pt particles that
used a one step method may deposit on the surface of TiO
2
Fig. 4 The rates of H
different Pt contents of 0, 0.1, 0.3, 0.5, 0.8, 1.5, and 2.0 wt%.
2
production of mesoporous Pt–D–M–Nb with Fig. 5 The rates of H
and 0.8-Pt–I–M–Nb.
2
production of mesoporous 0.8-Pt–D–M–Nb
96812 | RSC Adv., 2016, 6, 96809–96815
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
through the mesoporous network instead of on the outer and
1
6
near the outer surface of the TiO supporter. In this case,
2
the Pt particles prepared by the one step method were obvi-
ously embedded in the network instead of on the outer and
near the surface of the Nb O aggregates (TEM images (b and
2 5
d)). However, compared with the one-step method, the Pt
nanoparticles prepared by the impregnation method precip-
2 5
itated on the Nb O surface, and the sizes were larger than
those prepared by the one step method (from the TEM
images).
3.3.1. XPS analysis of the different catalyst loading
methods. To further characterize the differences in Pt–D–M–Nb
and Pt–I–M–Nb, the chemical states of Pt–D–M–Nb and Pt–L–
M–Nb were investigated by XPS analysis. As shown in Fig. S5,†
only C, Nb, O and Pt were detected in the samples of Pt–D–M–
Nb and Pt–I–M–Nb. The C 1s peak at around 285.0 eV is
ascribed to the carbon signal from within the instrument and
was used for calibration. The O 1s in Fig. 6a at around 531.39–
531.55 eV is characteristic of O–Nb in all samples, and the O 1s
peak at 529.82–530.06 eV is due to the surface hydroxyl groups
23
(
O–H) on the samples. The Nb 3d5/2 and Nb 3d3/2 XPS in Fig. 6b
show two peaks at 207.01 and 209.74 eV, respectively, which is
24
2 5
in good agreement with those of Nb O . No additional diffu-
sive electronic states were observed from the spectra of the Nb
elements of the M–Nb sample. But the binding energies of Nb
reveal small chemical shis due to the addition of the Pt
element. The binding energy of Nb on Pt–D–M–Nb shis from
207.01 eV to 206.81 eV and that of Pt–I–M–Nb shis from 207.01
eV to 206.66 eV. These shis were ascribed to the interaction of
the Nb species with Pt particles and the size of Pt particles. The
larger the Pt particles, the more the binding energy of Nb
decreases.
The Pt 4f XPS spectrum of 0.8-Pt–D–M–Nb and 0.8-Pt–I–M–
Nb is shown in Fig. 6c. It showed that both the Pt spectra of
0
.8-Pt–D–M–Nb and 0.8-Pt–I–M–Nb exhibit three chemical
states. The peaks at 70.0 eV and 71.3 eV can be attributed to
0
2+
Pt . The peak at 73.1 eV was particular to Pt and the peak at
4
+
25
7
4.8 eV was due to Pt . However, compared to 0.8-Pt–I–M–
Nb, the number of oxidized states of Pt and Pt in 0.8–Pt–D–
M–Nb is fewer than in 0.8-Pt–I–M–Nb. Generally, Pt is bene-
2
+
4+
0
Fig. 6 X-ray photoelectron spectra (XPS) of (a) O 1s and (b) Nb 3d of
M–Nb, 0.8-Pt–D–M–Nb and 0.8-Pt–I–M–Nb; (c) Pt 4f spectrum of

cial to the separation of electron–hole pairs but too many
2
+
4+
oxidized states of Pt and Pt may form recombination
0
.8-Pt–D–M–Nb and 0.8-Pt–I–M–Nb. The orange peak is assigned
0
2+
4+
centres of electron–hole pairs, which reduces the generation of to Pt , the blue peak to Pt , and the pink peak to Pt
radicals. According to the photocatalytic hydrogen production
.
of M–Nb, 0.8-Pt–D–M–Nb and 0.8-Pt–I–M–Nb (Fig. 5), 0.8-Pt–
Nb O , indicating a quenching of PL emissions due to the
D–M–Nb may exhibit a higher separation rate of electron–hole
pairs than that of 0.8-Pt–I–M–Nb and M–Nb. Here, the uo-
rescence spectra of M–Nb, 0.8-Pt–D–M–Nb and 0.8-Pt–I–M–Nb
were further tested. Fluorescence spectra are a helpful method
to disclose the efficiency of charge carrier trapping, migration
2 5
incorporation of Pt on the Nb O surface. In addition, their
2
5
photoluminescence (PL) spectrum intensity decreases in the
order 0.8-Pt–D–M–Nb < 0.8-Pt–I–M–Nb < M–Nb, indicating the
highest separation of photo-induced electron–hole pairs in the
case of 0.8-Pt–D–M–Nb.
2
9
and transfer in semiconductors. As shown in Fig. 7, all the
spectra revealed two emission bands located at ꢂ400 and ꢂ475
nm. The two signals are attributed to the surface irradiative
recombination and bulk irradiative recombination below the
3.3.2. ESR analysis of the different catalyst loading
methods. According to previously-reported work, the photo-
catalytic H production reaction process is a complex process.
2
3
0
The process involves several elemental steps such as separation
and interfacial transfer of photogenerated carriers, electron–
reduction, hole–oxidization, and sequential radical reactions, in
conductive band and free holes at the valance band edge.
However, the intensity of both the emission bands in 0.8-Pt–
D–M–Nb and 0.8-Pt–I–M–Nb was much lower than that in
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 96809–96815 | 96813

View Article Online
RSC Advances
Paper
observation is highly consistent with the results reported by Jie
27
ꢀ
Xu . Fig. 8a shows the EPR signal of cO and CH OHc in the
2
2
suspension of 0.8-Pt–D–M–Nb, 0.8-Pt–I–M–Nb and -M–Nb in
a DMPO/methanol system. Two sets of lines were assigned
ꢀ
unambiguously to the DMPO–cO
2
2
and DMPO–CH OHc
25
adducts, respectively. The EPR signal intensity of both the
ꢀ
DMPO–cO
2
and DMPO–CH
2
OH adducts decreases in the order
of 0.8-Pt–D–M–Nb > 0.8-Pt–I–M–Nb > M–Nb, which indicates
ꢀ
the greater abilities for generating cO
2
radicals of 0.8-Pt–D–M–
Nb than those of 0.8-Pt–I–M–Nb and Nb O . The EPR signal of
2
5
the DMPO–cOH adducts in the irradiated catalyst/DMPO/water
system is shown in Fig. 8b. The four characteristic peaks with
intensity of 1 : 2 : 2 : 1 were assigned to DMPO–cOH. It can be
Fig. 7 Fluorescence spectra of M–Nb, 0.8-Pt–D–M–Nb and 0.8-Pt– seen that the EPR signal intensity of the DMPO–cOH adducts
I–M–Nb.
decreases in the order of 0.8-Pt–D–M–Nb > 0.8-Pt–I–M–Nb > M–
26
Nb. In Gu’s report, the intensity of cOH reects the hole
ꢀ
transfer rate of a photocatalyst and the signal of DMPO–cO
2
2ꢀ
which the generation of radicals such as cOH and cO has could judge the electron transfer rate. The nano-particles of Pt
a very signicant task in the photocatalytic process. Analysis of on the support will capture the photo-generated electrons of the
26
these radicals will help us to better understand the co-catalyst photocatalyst and decrease the recombination of the photo-
1
0
effect of Pt in mesoporous Nb
2
O
5
. Thus, we compared the exited electron–hole pairs. The above results indicate that
ꢀ
generation capability of cO
2
and cOH radicals of 0.8-Pt–D–M– the interfacial transfer rate has been improved by Pt NPs and
Nb and 0.8-Pt–I–M–Nb in the dark and under UV light irradia- that the Pt NPs prepared by the one step method exhibit
tion by the DMPO spin-trapping EPR technique. There were no a higher interfacial transfer rate than that prepared by the
signals in the dark, conrming that the radicals were generated impregnation method.
by catalyst irradiation under UV light (not shown here). This
3.3.3. Possible mechanism. According to the photo-
catalytic activity and analysis of the ESR and PL, a possible
mechanism for photocatalytic hydrogen evolution was
proposed as illustrated in Scheme 1. When under UV light
2 5
irradiation, the mesoporous Nb O generates photo-excited
electrons and holes. Electrons and holes may experience two
major processes. On the one hand, the electrons and holes
migrate to the surface of the photocatalyst and then recom-
bine. On the other hand, the photo-electrons are captured by
the Pt particles and the photo-generated holes oxidize CH OH
3
to CO . The latter process prompts the separation of the photo-
2
excited electrons and holes and may enhance the photo-
catalytic activity. Pt has the highest work function and the
2 2
smallest H evolution over-potential which is benecial to H
2
0,28
production.
2ꢀ
Fig. 8 (A) EPR signals of the DMPO–cCH
2
OH and cO spin adducts in
the sample–methanol suspension; (B) EPR signals of the DMPO – cOH Scheme 1 Schematic of photocatalytic hydrogen evolution over Pt–
spin adducts in the sample – water suspension. Nb under UV light irradiation.
2 5
O
96814 | RSC Adv., 2016, 6, 96809–96815
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
1
1 S. H. wang, M. C. Lee and W. Choi, Appl. Catal., B, 2003, 46,
4
. Conclusions
49–63.
Mesoporous Nb
2
O
5
with a uniform and narrow pore size 12 H. Y. Lina, H. C. Yangb and W. L. Wang, Catal. Today, 2011,
distribution was prepared. The loading of platinum onto the
mesoporous Nb greatly improved the photocatalytic H
evolution activity. Pt–Nb prepared by a one step method
174, 106–113.
2
O
5
2
13 C. Zhang, H. He and K. i. Tanaka, Appl. Catal., B, 2006, 65,
37–43.
2 5
O
provided superior and more stable photocatalytic activity than 14 J. Yu, L. Qi and M. Jaroniec, J. Phys. Chem. C, 2010, 114,
that prepared by an impregnation method. The main reason
13118–13125.
may be ascribed to the fact that the Pt particles loaded onto 15 X. J. Zheng, L. F. Wei, Z. H. Zhang, Q. J. Jiang, Y. J. Wei, B. Xie
Nb O by the one step method are more uniform and smaller in
and M. B. Wei, Int. J. Hydrogen Energy, 2009, 34, 9033–9041.
size, which was benecial to interfacial transfer and produced 16 T. Sreethawong and S. Yoshikawa, Int. J. Hydrogen Energy,
2
5
more radicals.
2006, 31, 786–796.
1
1
1
2
2
2
2
7 F. Zhang, J. Chen, X. Zhang, W. Gao, R. Jin, N. Guan and
Y. Li, Langmuir, 2004, 20, 9329–9334.
8 J. J. Zou, H. He, L. Cui and H. Y. Du, Int. J. Hydrogen Energy,
Acknowledgements
2007, 32, 1762–1770.
This work was nancially supported by the NSFC (Grant No.
9 P. Yang, D. Zhao, D. I. Margolese, B. F. Chmelka and
G. D. Stucky, Chem. Mater., 1999, 11, 2813–2826.
0 A. A. Ismail and D. W. Bahnemann, Green Chem., 2011, 13,
21373051, and U1305242), the Science and Technology project
of the Education Office of Fujian Province of P. R. China
(
(
JA12017), and the National Basic Research Program of China
973 Program, No. 2012CB722607).
4
28.
1 H. Courbon and J. M. Herrmann, J. Phys. Chem., 1984, 88,
210–5214.
2 K. E. Karakitsou and K. A. X. E. Verykios, J. Catal., 1992, 134,
5.
5
Notes and references
1
1
2
A. Fujishima and K. Honda, Nature, 1972, 238, 37–38.
J. Liu, Y. Liu, N. L. Y. Han, Z. Kang and Y. Lifshitz, Science,
3 S. F. Ho, S. Contarini and J. W. Rabalais, J. Phys. Chem., 1987,
91, 10.
2015, 347, 969–974.
24 S. C. S. F. Ho and J. W. Rabalais, J. Phys. Chem., 1987, 91, 10.
3
4
H. Ahmad, S. K. Kamarudin and L. J. Minggu, Renewable 25 F. B. Li and X. Z. Li, Chemosphere, 2002, 48, 1103–1111.
Sustainable Energy Rev., 2015, 43, 599–610.
26 Q. Gu, J. Long and H. Zhuang, Phys. Chem. Chem. Phys., 2014,
16, 12521–12534.
X. Chen, T. Yu and X. Fan, Appl. Surf. Sci., 2007, 253, 8500–
8
506.
2
7 J. Xu, S. He, H. Zhang, J. Huang, H. Lin, X. Wang and J. Long,
J. Mater. Chem. A, 2015, 3, 24261–24271.
5
6
7
8
Y. H. Pai and S. Y. Fang, J. Power Sources, 2013, 230, 321–326.
M. E. Davis, Nature, 2002, 417, 813–821.
A. Corma, Chem. Rev., 1997, 97, 2373–2420.
28 H. Huang, C. Wang, J. Huang, X. Wang, Y. Du and P. Yang,
Nanoscale, 2014, 6, 7274–7280.
J. N. Kondo, M. Uchida, K. Nakajima, D. L. Lu, M. Hara and 29 S. B. Ning, L. Y. Ding, Z. G. Lin, Q. Y. Lin, H. L. Zhang,
K. Domen, Chem. Mater., 2004, 16, 4304–4310.
T. Maruyama and T. Nishimoto, Ind. Eng. Chem. Res., 1991,
H. X. Lin, J. L. Long and X. X. Wang, Appl. Catal., B, 2016,
185, 203–212.
9
3
0, 1634–1638.
0 X. Wang, J. C. Yu and H. Y. Yip, Chemistry, 2005, 11, 2997–
004.
30 J. Q. Yan, G. J. Wu, N. J. Guan and L. D. Li, Appl. Catal., B,
1
2014, 152, 280–288.
3
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 96809–96815 | 96815