Heterojunctions between amorphous and crystalline niobium oxide with enhanced photoactivity for selective aerobic oxidation of benzylamine to imine under visible light

Article abstract of DOI:10.1039/c5ta03214h

The formation of heterojunctions between two crystals with different band gap structures, acting as a tunnel for the unidirectional transfer of photogenerated charges, is an efficient strategy to enhance the photocatalytic performance of semiconductor pho

Full text of DOI:10.1039/c5ta03214h

Journal of
Materials Chemistry A
View Article Online
View Journal
PAPER
Heterojunctions between amorphous and
crystalline niobium oxide with enhanced
photoactivity for selective aerobic oxidation of
benzylamine to imine under visible light†
Cite this: DOI: 10.1039/c5ta03214h
b
b
Yantao Zhang,ab Linjuan Pei, Zhanfeng Zheng, Yong Yuan,c Tengfeng Xie,d
*
Juan Yang, Shuai Chen, Junwen Wang, Eric R. Waclawikf and Huaiyong Zhuf
e
b
a
*
The formation of heterojunctions between two crystals with different band gap structures, acting as a
tunnel for the unidirectional transfer of photogenerated charges, is an efficient strategy to enhance the
photocatalytic performance of semiconductor photocatalysts. Considering that surface complex
photocatalysts also exhibit charge separation and recombination processes, the heterojunctions may
also promote the visible-light-response photoactivity of any surface complex catalysts by influencing the
transfer of photogenerated electrons. Herein, Nb2O5 microfibers, with a high surface area of interfaces
between an amorphous phase and a crystalline phase, were designed and synthesised by the calcination
of hydrogen-form niobate while controlling the crystallisation. The photoactivity of these microfibres
towards selective aerobic oxidation reactions was investigated. As predicted, the Nb2O5 microfibres
containing heterojunctions exhibited the highest photoactivity. This could be due to the band gap
difference between the amorphous phase and the crystalline phase that allows electron transfer
unidirectionally, which decreased the recombination rate and improved the efficiency.
Received 5th May 2015
Accepted 20th July 2015
DOI: 10.1039/c5ta03214h
www.rsc.org/MaterialsA
reported that the red-shi of the light absorption on Nb2O5 is
attributed to the formation of a surface complex between the
1 Introduction
Nb2O5 surface and adsorbed amine by forming Nb-amide
species.4 The surface complex can potentially adsorb light and
inject an electron from the N localised on the amide nitrogen
into the conduction band of Nb2O5 under light illumination
using longer wavelengths, which is the so-called ligand-to-metal
charge transfer (LMCT) mechanism. Amine itself behaves as a
light sensitizer and reactant, and the photocatalyst surface self-
regenerates aer a reaction cycle. This phenomenon is also
observed in TiO2 (band gap ꢀ3.2 eV) and surface adsorbates
that can absorb visible light and initiate reactions.13–16 These
ndings are important because the visible region of the solar
spectrum is more abundant than the UV region, which account
for 43% and 4% of the solar energy incident on the surface of
the earth.17 In general, shiing the light absorption of wide
band gap semiconductors (e.g. TiO2) to the visible range is
difficult. Oen, lattice doping with metal and non-metal
elements is applied for this purpose; however, the loss of the
dopant element during recycling is a big concern.18–20 Herein,
the formation of a surface complex may provide an option to
achieve high photoactivity in the visible light range.
Imines are important intermediates for the synthesis of ne
chemicals and pharmaceuticals.1 Recently, green processes
using light and molecular oxygen to drive the direct oxidation of
amines to imines have attracted great attention. Many mate-
6
rials, including metal oxides (e.g. TiO2,2,3 Nb2O5,4,5 BiVO4
,
hydrotalcite7) and plasmonic metals (e.g. Au–Pd8 and Ag9), have
shown high activity and good selectivity for amine to imine
conversion. Specically, the oxidation of amine on Nb2O5, a
wide band gap semiconductor (ꢀ3.2 eV), has shown activity
under visible light illumination up to 450 nm which is much
narrower than its band gap adsorption at 390 nm.4,10–12 It is
aCollege of Chemistry and Chemical Engineering, Taiyuan University of Technology,
Taiyuan 030024, China. E-mail: wangjunwen@tyut.edu.cn
bState Key Laboratory of Coal Conversion, Institute of Coal Chemistry, Chinese
Academy of Sciences, Taiyuan 030001, China. E-mail: zfzheng@sxicc.ac.cn
cXi'an Thermal Power Research Institute Co. Ltd, Xi'an 710032, China
dCollege of Chemistry, Jilin University, Changchun 130012, China
eDepartment of Physics and Chemistry, Henan Polytechnic University, Jiaozuo 454003,
China
In a semiconductor photocatalytic system, the phase junc-
tion between two semiconductors acts as an efficient electron
and hole separator. It hinders the recombination process and
improves the photoactivity.21–27 The effect of a surface complex
fSchool of Chemistry, Physics and Mechanical Engineering Queensland University of
Technology, Brisbane, QLD 4000, Australia
† Electronic supplementary information (ESI) available: Output spectrum of LED
light, normalised photocatalytic activity using the same surface area. See DOI:
10.1039/c5ta03214h
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
˚
in conjunction with the heterojunction may yield a photo- diffractometer with Cu-Ka radiation (l ¼ 1.5418 A) and oper-
catalyst with high activity by suppressing recombination on the ating in a 2q range of 10–63ꢁ at a scanning rate of 4ꢁ minꢂ1
.
catalyst surface. To form a heterojunction, there needs to be a Fourier transform infrared spectroscopy (FTIR) was performed
band gap match between the two phases. In a previous report, using a Nicolet Magna-IR 550-II spectrometer with KBr pellets.
we prepared niobate bres by a hydrothermal method and Nitrogen sorption isotherms were measured by a volumetric
found that the cations in the structure are exchangeable with method on an automatic adsorption instrument (Micro-
other cations.28 Therefore, exchange with H+ cations and meritics, TriStar II 3020 analyser) at liquid nitrogen tempera-
deliberate control of the calcination temperature can release ture (77 K). The specic surface area was calculated by the
water from the structure and develop a route to prepare Nb2O5 Brunauer–Emmett–Teller (BET) method from the data in a P/P0
with phase junction structures and thus tune the photocatalytic range between 0.05 and 0.2. To investigate the light absorption
activity. This procedure is comparable to the phase trans- and emission behaviour of the samples, the UV-Vis diffuse
formation that occurs during calcination of protonated reectance (UV-Vis-DR) spectra of the samples were measured
titanate.24
on a Shimadzu UV-3600 spectrometer. Thermogravimetric/
In the present study, in order to study the structural inu- differential thermal (TG/DTA) analysis was carried out on a
ence on the photoactivity, niobium oxide (Nb2O5) microbres Rigaku TG in air with a heating rate of 10 ꢁC minꢂ1. Raman
were prepared by the calcination of hydrogen-form niobate spectra were measured on a Horiba LabRAM HR800 spec-
(H-niobate) at temperatures ranging from 400 to 600 ꢁC to serve trometer equipped with a 514 nm argon laser. The electron spin
as photocatalysts for the selective oxidation of benzylamine to resonance (ESR) spectra were measured using a JEOL JES-RE2X
the corresponding imine. It was found that the sample obtained electron spin resonance spectrometer, conducted at room
aer being calcined at 500 ꢁC generates a mixed phase of temperature with the addition of 5,5-dimethyl-1-pyrroline
amorphous and crystalline Nb2O5. This form of Nb2O5 micro- N-oxide (DMPO, BePharm). X-ray photoelectron spectroscopy
bres exhibited much higher activity than the pure amorphous (XPS) was carried out on a Thermo ESCALAB 250 spectrometer
phase and crystalline phase and all demonstrated high selec- at room temperature by using an Al Ka X-ray source (hn ¼
tivity to imine.
1486.6 eV). The C 1s peak at 284.6 eV was used as a reference for
the calibration of the binding energy scale. Transient photo-
voltage technique (TPV) characterisation was conducted in air
atmosphere and at room temperature on a home-made instru-
ment. The sample was excited with a laser radiation pulse
(wavelength of 355 nm and pulse width of 5 ns) from a third-
harmonic Nd:YAG laser (Polaris II, New Wave Research, Inc.).
The TPV signal was registered by using a 500 MHz digital
phosphor oscilloscope (TDS 5054, Tektronix).
2 Experimental
2.1 Catalyst preparation
All chemicals were from Sinopharm Chemical Reagent Co. Ltd
unless otherwise stated. Hydrogen-form niobate (H-niobate)
microbres were prepared by an alkaline hydrothermal treat-
ment of Nb2O5 powders and a subsequent ion exchange
process. In a typical synthesis, 1.0 g of Nb2O5 (AR, Aladdin) was
dispersed in a 60 mL aqueous solution of 10 M NaOH and
stirred continuously for 4 h. The obtained mixture was auto-
claved at 140 ꢁC for 6 h to yield sodium niobate solids. The white
precipitate was recovered by centrifugation and subsequently
washed with deionised water to remove adsorbed NaOH
(designated as NN). Aerwards, the collected solids were
washed with an aqueous solution of 0.1 M HCl to exchange the
Na+ in the structure with H+. This was followed by washing with
deionised water again and then solid recovery using centrifu-
gation. The sample was dried in air at 100 ꢁC for 12 h
(H-niobate, designated as HN). To prepare niobium oxide
microbres, H-niobate was calcined at 400, 500, and 600 ꢁC for
3 h (designated as HN-400, HN-500, and HN-600, respectively).
2.3 Evaluation of photocatalytic activity
Typically, the photocatalytic selective amine oxidation reactions
were carried out at 60 ꢁC in an oxygen atmosphere for 10 h. The
reaction system consisted of a catalyst (50 mg) and benzylamine
(0.2 mmol) in acetonitrile (5 mL). The incident light source was
a 100 W COB white LED. The liquid products were analysed
using a Shimadzu 2014C GC with a WondaCap 5 column. The
single coloured LEDs (blue and green) were also used as irra-
diation sources to investigate the inuence of light wavelength
on the photocatalytic activity (output spectra shown in Fig. S1†).
3 Results and discussion
3.1 Physicochemical properties of Nb2O5 microbres
2.2 Characterisation
The X-ray diffraction (XRD) patterns of the as-obtained sodium
The transmission electron microscopy (TEM) studies on the niobate (labelled as NN), H-niobate (labelled as HN) and
samples were carried out on a JEM-2100F microscope operating niobium oxide microbres (H-niobate calcined between 400
at 200 kV. The sample powder was dispersed in ethanol by and 600 ꢁC, labelled as HN-400, HN-500, and HN-600) are
sonication; drops of the suspension were applied onto a copper shown in Fig. 1a. Sodium niobate, a hydrated Na2Nb2O6 phase,
micro-grid-supported transparent carbon foil and dried in air. can be indexed using a space group C2/c with parameters a ¼
The scanning electron microscopy (SEM) images were obtained 1.705 nm, b ¼ 5.029 nm, c ¼ 1.654 nm and b ¼ 113.9ꢁ, which is
for the morphology study using a JSM-7001F instrument. X-Ray consistent with the previous report.28–31 H-niobate and HN-400
diffraction (XRD) patterns were recorded on a MiniFlex II exhibited similar patterns which are obviously in the
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry A
amorphous state. This could be due to the disordering of the the relationship between the absorption coefficient (a) and the
crystal structure during the proton exchange process. According optical band gap energy (Eg) near the band edge follows the
to the Hard–So Acid–Base (HSAB) theory, hard acids bind formula: ahn ¼ A(hn ꢂ Eg)1/2, where hn and A are the photon
strongly to hard bases and so acids bind strongly to so energy and a constant, respectively.38,39 The band gap energy of
bases.32,33 H+ and Na+ cations are both hard acids and bind Nb2O5 can be obtained by extrapolating the linear part of the
tightly with H2O, the hard base, in the interlayer of metal oxides. plots to the photon energy axis. The calculated band gaps are
Hydrated H+ and Na+ have radii of 0.28 and 0.358 nm,34 3.43, 3.27, 3.14 and 3.07 eV for H-niobate, HN-400, HN-500, and
regardless of the fact that H+ and Na+ cations have small radii of HN-600, respectively. This difference could be due to the
ꢂ0.018 and 0.102 nm.35,36 Hydrated H+ is a smaller cation and quantum size effect.40,41 They absorb UV light, reect visible
thus possesses higher priority in the ion exchange compared to light and thus exhibit white colour. The Raman spectra of the
hydrated Na+. Meanwhile, hydrated H+ should have a stronger microbres, shown in Fig. 1d, also prove that the microbres
interaction with the layer compared to hydrated Na+ due to the have a different crystal structure. The spectrum of HN-600
small size and may cause the collapse of the layer structure. The exhibits Raman bands at 692, 314, 230 and 141 cmꢂ1, whichꢂis1
periodic crystal structure is destroyed and the amorphous similar to the previous reports.28,42 The major band at 692 cm
structure of the HN is attained. For HN-500, broad diffraction is the characteristic band for the structure consisting of NbO6
peaks attributed to the amorphous phase and weak diffraction octahedra-sharing corners.43 The Raman spectrum of HN-500 is
peaks (001), (100), and (101) attributed to hexagonal phase similar to that of HN-600, but with a much weaker intensity due
Nb2O5 were observed (matched to PDF Card no. 28-0317). For to the different crystallinity. The spectrum of HN-400 exhibited
HN-600, all the characteristic diffraction peaks of hexagonal a broad peak at 656 cmꢂ1, which is a big shi of the 692 cmꢂ1
Nb2O5 were well-developed which indicates a better crystal- band for HN-600. This is indicative of the distorted crystal
linity. The thermogravimetric/differential thermal analysis structure in HN-400.43 Valence band X-ray photoelectron spec-
(TG/DTA) curves for the H-niobate powders are presented in troscopy (VB-XPS) is routinely used to measure the valence band
Fig. ꢁ1b. Most of the H2O evolved from the structure at T < edge position of a semiconductor.44 From the VB XPS spectra of
350 C, which yielded an ꢀ8% mass loss. Only an ꢀ1% mass HN-400 and HN-600 (Fig. 1e), the edges of the maximum energy
ꢁ
loss was observed in the range between 350 and 600 C, which are the same. The UV-Vis spectra of HN-400 and HN-600
means that a niobium oxide product formed. The only endo- revealed that the band gap difference between them is about
thermic peak at 540 ꢁC in the DTA curve should be attributed to 0.2 eV. Therefore, the minimum energy of the conduction band
the phase crystallisation from the amorphous phase to crystal- edge of HN-600 should be lower about 0.2 eV than that of
line phase.37 The H-niobate and niobium oxide microbres HN-400 (Fig. 1f).
exhibited slightly different UV-Vis absorption, as is shown in the
The scanning electron microscopy (SEM) images of the
UV-Vis spectra (Fig. 1c). As for a direct band gap semiconductor, samples are shown in Fig. 2. As can be seen, the H-niobate
sample exhibits a brous morphology with the mean diameter
distribution from 0.5 to 2 mm. Upon heating from 400 to 600 ꢁC,
the long bres were slightly broken into shorter bres. HN-400
is in an amorphous phase and it gradually transformed into a
crystalline phase as shown by TEM analysis (Fig. S2†). Interfaces
between the crystalline phase and amorphous phase are clearly
Fig. 1 (a) XRD patterns, (b) TGA-DTA curves, (c) UV-Vis spectra, (d)
Raman spectra and (e) valence-band XPS spectra of the microfibres. (f)
Schematic diagram of the DOS of the amorphous and crystalline Fig. 2 SEM images of H-niobate and Nb2O5 microfibres. The scale
Nb2O5.
bars are all 1 mm.
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
difference between the three catalysts lies in the phase content:
HN-400 and HN-600 are in a pure amorphous phase and crys-
talline Nb2O5 phase, respectively, while HN-500 is in a mixed
phase with heterojunctions between the two. In the catalytic
cycle, an electron is transferred to the conduction band of
Nb2O5; meanwhile, a hole forms on the surface complex. If the
electron is not transferred to an electron acceptor, it recom-
bines with the surface complex to make a null cycle.14 Therefore,
the heterojunction may suppress the hindered recombination
process and give a high activity analogous to the recombination
process in a semiconductor. It is noted that the recombination
process only takes place on the surface, while it happens both in
the bulk and on the surface in a semiconductor system.
Fig. 3 TEM (a), SAED (inset of a) and HRTEM (b) patterns of a HN-500
microfibre.
HN-500 exhibited good recyclability and there is little
difference in both relative activity and selectivity aer 5
successive cycles (Fig. 5a). The photocatalyst before and aer
the catalytic reactions was characterised using various tech-
niques including XRD and UV-Vis spectroscopy (Fig. 5b and c).
It was found that there is no obvious structure change aer the
reaction according to the XRD patterns of the sample aer 1
cycle, 3 cycles and 5 cycles. Aer the reaction, the surface of HN-
500 was adsorbed with reactants and caused a red-shi in the
UV-Vis absorption, and the absorption in the visible range is
almost the same for the three reused samples. The vibrations
ꢀ
ꢀ
seen in HN-500 (Fig. 3). The fringes of (0001), (1011) and (1011)
planes attributed to hexagonal Nb2O5 are indexed to the
HRTEM images of HN-500. It is noted that the HN-600 is a
single crystal microbre as shown by TEM observation.
3.2 Photocatalytic performances of Nb2O5 microbres
The photocatalytic performances of H-niobate and niobium
oxide microbres for the selective oxidation of benzylamine are
shown in Fig. 4. No conversion was observed in any of the
reactions in the absence of light irradiation or O2, which
conrms that the oxidation of benzylamine is a light-driven and
dioxygen oxidation reaction. These photocatalysts exhibited
very high selectivity (>95%) for the oxidation of benzylamine to
the corresponding imine. When the same amounts of catalysts
were added, the conversion rate of HN-500 was the highest at
90% (60 ꢁC, 10 h) among the catalysts investigated whose
conversion rates were 58% (H-niobate), 80% (HN-400), and 65%
(HN-600) and the selectivity was all above 95%. These micro-
bres had a similar specic surface area at 5.5, 6.7, 5.8, and
5.3 m2 gꢂ1 for HN, HN-400, HN-500, and HN-600, respectively.
Aer the specic area of the samples was normalised (using the
surface area sample HN-500), HN-500 still exhibited the highest
conversion rate at 90%, while HN, HN-40 and HN-600 exhibited
a similar rate at 63%, 70%, and 72%, respectively (Fig. S3†). The
dNH from benzylamine were observed (Fig. S4†), and this may
2
indicate that benzylamine is non-dissociatively adsorbed on the
oxide surfaces. Niobium oxide does not absorb light of wave-
lengths larger than 404 nm (the crystalline phase HN-600 has a
band gap of 3.07 eV, Fig. 5c). A white-light LED with a wave-
length range from 410 nm to 750 nm was used. Therefore, it is
reasonable to attribute the activity to the ligand-to-metal charge
transfer (LMCT) mechanism. This behaviour has been reported
previously by Lang et al. who found that anatase TiO2 could
catalyse benzylamine oxidation under visible light illumina-
tion.2,3 It is believed that the absorbed benzylamine acts as a
photosensitiser under visible light illumination, one electron
Fig. 4 Performances of H-niobate and Nb2O5 microfibers for the
selective oxidation of benzylamine under visible light. Reaction Fig. 5 Recyclability and stability of HN-500 in the benzylamine
conditions: catalyst (50 mg), benzylamine (0.2 mmol), acetonitrile (5 oxidation reaction. (a) Conversion and selectivity in five successive
mL), O2 (1 atm), 60 ꢁC, 10 h, light intensity (0.6 W cmꢂ2, white LED). No cycles. (b) XRD patterns and (c) UV-Vis spectra of HN-500 after 1, 3 and
product was observed in the absence of light.
5 reaction cycles. Each cycle was conducted at 80 ꢁC for 8 h.
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry A
increased gradually when the reaction temperature was
increased from 40 to 80 ꢁC. Generally, the recombination rate in
a semiconductor photocatalyst increases with temperature,
leading to a reduction in the reaction rate.47,48 The result
observed here further conrms that the complex-route works
following a different mechanism. The apparent reaction rate
constants (k) of the oxidation of benzylamine on Nb2O5 micro-
bres at 40, 60 and 80 ꢁC under visible light irradiation are 3.0 ꢃ
10ꢂ3 mol Lꢂ1 sꢂ1, 3.9 ꢃ 10ꢂ3 mol Lꢂ1 sꢂ1 and 5.1 ꢃ 10ꢂ3 mol
Lꢂ1 sꢂ1, calculated from the slope by plotting the curve of
conversion vs. the reaction time. Benzylamine was converted to
N-benzylbenzaldimine with a very high selectivity (>95%)
regardless of the reaction temperature. Benzaldehyde was also
detected in the product during the reaction which may indicate
that benzaldehyde is the intermediate product and converts to
imine through a condensation reaction with another benzyl-
amine molecule.3 The estimated apparent activation energy (Ea)
Fig. 6 TPV of HN-400, HN-500 and HN-600 before and after (sample
name with suffix “-BA”) benzylamine adsorption.
transfers to a single Ti site, and thus the oxidation reaction
occurs.
for the oxidation of benzylamine using HN-500 is 12.2 kJ molꢂ1
,
The transient photovoltage (TPV) techniques can provide
direct information about the separation and transfer of photo-
generated charge carriers in a semiconductor. To understand
the transfer properties of the photogenerated electrons, the TPV
spectra of the niobium oxide bers were performed under
355 nm light irradiation and the spectra are shown in Fig. 6. The
negative response of photovoltage implies that photogenerated
electrons accumulated at the surface of the sample and a
positive response indicates that photogenerated holes accu-
mulated at the surface.45,46 As shown in Fig. 6, the TPV response
of HN-400 is negative at the initial period upon light irradiation
and it begins to increase and nally a positive TPV response is
obtained. This means that photogenerated electrons rstly
accumulated at the surface and then photogenerated holes
accumulated at the surface. The TPV response of HN-600 is
negative, indicating that photogenerated electrons accumulate
at the surface of crystalline Nb2O5. While for HN-500, the mixed
phase sample, a much stronger negative signal was observed
compared to the amorphous and crystalline sample (HN-400
and HN-600), which indicates that the heterojunction helps to
improve the charge separation.
As for the sample aer benzylamine adsorption (sample
name with the suffix “-BA”), HN-500-BA, with the heterojunction
structure, exhibits a stronger positive response compared to
HN-600-BA and HN-400-BA. This means that photogenerated
holes accumulated at the surface for the benzylamine adsorbed
sample, which is consistent with the LMCT mechanism. The
separated electron and hole pair originated from the surface
complex is dominant compared to the pair formed due to the
Nb2O5 bandgap excitation. The TPV response conrms the
result that the photogenerated charges have a better separation
performance in the sample with the heterojunction structure.
calculated from the slope of the curve by plotting the reaction
rate against 1/T. It was reported that the apparent activation
energy for the selective oxidation of benzylamine to N-benzyl-
benzaldimine on the Au–Pd@ZrO2 catalyst in the dark and
under visible light irradiation was 131.5 and 118.4 kJ molꢂ1
,
respectively.8 Therefore, the photocatalytic reaction for the
oxidation of benzamine on Nb2O5 microbres has smaller
activation energy and can proceed readily. The reaction rate
increase with the increasing temperature may indicate that the
reaction is a dynamics controlled process, which is due to the
increased collision frequency of molecules. In the photo-
catalytic reports, researchers seldom study the temperature
effect.49–51 However, both light and heat can speed up the
reaction rate and it is important to note that the reaction rate
can be greatly improved with a slight increase in the reaction
temperature.
The change of either the wavelength or the intensity of the
incident light affects the photocatalytic activity of HN-500
(Fig. 8). A higher conversion rate was observed under irradiation
of higher intensity. To understand the mechanism, we also
studied the dependence of the conversion rate on the irradia-
tion wavelength. Two coloured LED lights [blue (425–510 nm,
peaked at 459 nm), green (475–600 nm, peaked at 525 nm), with
3.3 Photocatalytic mechanism of Nb2O5 with
heterojunctions
The reaction kinetic data obtained at various temperatures (40,
ꢁ
60 and 80 C) for the oxidation of benzylamine using HN-500
under visible light irradiation are presented in Fig. 7. It exhibits
Fig. 7 The reaction profiles in the selective oxidation of benzylamine
a typical zero-order kinetic reaction character. The reaction rate over HN-500 microfibres at different temperatures (40, 60 and 80 ꢁC).
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
butylamine was used, no imine production was observed in our
study. These studies demonstrate that the presence of the
benzene ring is important to realise the high conversion during
amine oxidation. It is reported that N or C radicals are gener-
ated during the reaction process. Therefore, the benzene ring
may play an important role in stabilising the adjacent C
radical.2,5 It is expected that the intermediate product should
also be benzaldehyde when an aromatic amine, with more than
one C between the benzene ring and the –NH2 group was used.
As expected, when phenethylamine was selected for selective
oxidation and the production was found to be 1-phenyl-N-(2-
phenylethyl)methanimine which is clearly a condensation
product of benzaldehyde and phenethylamine.
Based on the above analysis and the previous reports,2,3,53,54
we tentatively propose a mechanism for this reaction, as sche-
matically shown in Fig. 9. When the surface complex is irradi-
ated under light, the photogenerated electron injects to the
conduction band of Nb2O5, while the hole is le on the surface
complex. The hole cannot transfer and the electron can transfer
Fig. 8 Light intensity influence of selective oxidation of benzylamine
using HN-500 microfibres conducted at 80 ꢁC for 8 h. Conversion of
benzylamine ( ), selectivity to imine ( ) and benzaldehyde ( ).
output spectra shown in Fig. S1†] were used for the wavelength
experiments. It was found that HN-500 was active under blue
light but did not exhibit any photoactivity under green light.
Given that HN-500 has a bandgap of ꢀ3.2 eV, the irradiation
with wavelength l > 425 nm cannot initiate the photocatalytic
reaction. Therefore, the catalyst active site under irradiation
should be another photocatalytic component – the surface
complex. To conrm the radical component formed during
reaction, DMPO was introduced into the photocatalytic system
and the ESR signal of DMPO-OO(H) was clearly seen (partially
decomposed to DMPO-OH, Fig. S5†), which indicates the pres-
ence of O2cꢂ radicals.52
To understand the substituent inuence on the reaction, the
catalytic oxidation was conducted on derivative compounds
with electron donation groups (–CH3 and CH3O–) and with the
electron withdrawing group (–Cl). The results are summarised
in Table 1. The substituents on the benzene ring appear to have
only a slight inuence on the reaction rate and a minor inu-
ence on the product selectivity. When a straight chain amine-
Table 1 The influence of substituents on the benzene ring on the
reaction rate and selectivity (80 ꢁC, 8 h)a
Reactant
Product
Conv. (%)
99.98
Sel. (%)
93.15
99.61
91.97
96.12
97.43
93.59
89.85
—
quant.
—
84.26
80.50
Fig. 9 Schematic illustration of (a) the heterojunction structure and (b)
the proposed benzylamine oxidation mechanism of niobium oxide
photocatalysts.
a
Conv. ¼ conversion, sel. ¼ selection, quant. ¼ quantitative.
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry A
between the two phase. Due to the existence of the hetero- Department of Human Resources and Social Security of Shanxi
junction, the net result is that electrons transfer to the crystal- Province for Returned Scholars (No. 2013-35), the Shanxi
line phase with the narrower band gap, which suppresses the Science and Technology Department (No. 2015081044), and the
recombination process. In a reaction cycle, rstly, the reactant Hundred Talents Program of the Chinese Academy of Sciences.
benzylamine is adsorbed on the surface of Nb2O5 bres to form
a surface charge transfer complex. Secondly, holes and elec-
trons would be generated on the surface charge transfer
complex under light illumination. The photogenerated hole
Notes and references
induces H+ abstraction from the N of benzylamine and the
photogenerated electron transfers to the conduction band of
Nb2O5, resulting in the activation of molecular O2 to O2cꢂ. The
nitrogen-centred radical would rearrange via intramolecular H-
abstraction to form a more stable carbon-centred radical.55,56
Thirdly, O2cꢂ adsorbed on the surface of catalysts forms an
intermediate possessing a ve-member-ring structure with the
carbon radical and Nb5+ with the termination of the radicals.
Fourthly, the cleavage of the C–N bond of benzylamine and the
O–O bond of oxygen occurs to form benzaldehyde and a Nb
bonded three-member-ring structure. The produced benzalde-
hyde then further follows a condensation reaction with
unreacted benzylamine to yield imine. Finally, the photo-
catalytic cycle are completed by the regeneration of Nb5+ sites
on the surface of Nb2O5 bres through desorption of the NH2–
OH molecule.
1 R. A. Sheldon and J. K. Kochi, in Metal-catalyzed Oxidations of
Organic Compounds, ed. R. A. S. K. Kochi, Academic Press,
1981, pp. 17–32.
2 X. Lang, H. Ji, C. Chen, W. Ma and J. Zhao, Angew. Chem., Int.
Ed., 2011, 50, 3934–3937.
3 X. Lang, W. Ma, Y. Zhao, C. Chen, H. Ji and J. Zhao, Chem.–
Eur. J., 2012, 18, 2624–2631.
4 S. Furukawa, Y. Ohno, T. Shishido, K. Teramura and
T. Tanaka, ACS Catal., 2011, 1, 1150–1153.
5 S. Furukawa, Y. Ohno, T. Shishido, K. Teramura and
T. Tanaka, J. Phys. Chem. C, 2013, 117, 442–450.
6 B. Yuan, R. Chong, B. Zhang, J. Li, Y. Liu and C. Li, Chem.
Commun., 2014, 50, 15593–15596.
7 X.-J. Yang, B. Chen, X.-B. Li, L.-Q. Zheng, L.-Z. Wu and
C.-H. Tung, Chem. Commun., 2014, 50, 6664.
8 S. Sarina, H. Zhu, E. R. Jaatinen, Q. Xiao, H. Liu, J. Jia,
C. Chen and J. Zhao, J. Am. Chem. Soc., 2013, 135, 5793–5801.
9 Z. Zheng, C. Chen, A. Bo, F. S. Zavahir, E. R. Waclawik,
J. Zhao, D. Yang and H. Zhu, ChemCatChem, 2014, 6, 1210–
1214.
Herein, the Nb2O5–amine surface complex structure is
signicant towards a green synthesis of imine: (1) it can induce
a red-shi of the light absorption and improve the light
absorption efficiency; (2) it is a self-regeneration process and
the active sites can always be “fresh” with the progress of the 10 J. He, Y. Hu, Z. Wang, W. Lu, S. Yang, G. Wu, Y. Wang,
catalytic reaction; (3) it results in high product selectivity. This
process is different from a semiconductor photocatalytic
S. Wang, H. Gu and J. Wang, J. Mater. Chem. C, 2014, 2,
8185–8190.
process where the selectivity is very difficult to control due to the 11 Y. Zhao, C. Eley, J. Hu, J. S. Foord, L. Ye, H. He and
high oxidation potential of OHc radicals in aqueous photo-
catalytic systems.2,18,57–60
S. C. Tsang, Angew. Chem., Int. Ed., 2012, 51, 3846–3849.
12 X. Li, N. Kikugawa and J. Ye, Adv. Mater., 2008, 20, 3816–
3819.
13 G. Zhang, G. Kim and W. Choi, Energy Environ. Sci., 2014, 7,
4 Conclusions
954.
In summary, the amorphous and crystallite phases of Nb2O5 14 S. Kim and W. Choi, J. Phys. Chem. B, 2005, 109, 5143–5149.
microbres were prepared by the facile calcination of hydrogen 15 S. Higashimoto, N. Kitao, N. Yoshida, T. Sakura, M. Azuma,
niobate at deliberately controlled temperatures ranging from
H. Ohue and Y. Sakata, J. Catal., 2009, 266, 279–285.
400 to 600 C. The Nb2O5 microbres calcined at 500 C con- 16 R. Li, H. Kobayashi, J. Guo and J. Fan, J. Phys. Chem. C, 2011,
tained heterojunctions between the amorphous phase and the 115, 23408–23416.
crystallised phase produced during the crystallisation process. 17 Z. Zou, J. Ye, K. Sayama and H. Arakawa, Nature, 2001, 414,
It has been found that the surface complexes were formed aer
625–627.
amine adsorption on Nb2O5, resulting in visible-light photo- 18 A. L. Linsebigler, G. Lu and J. T. Yates, Chem. Rev., 1995, 95,
catalytic activity and enabling direct electron transfer to the 735–758.
conduction band. The catalyst was self-regenerated with the 19 R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki and Y. Taga,
progress of the reaction and exhibited good stability. The het- Science, 2001, 293, 269–271.
erojunctions effectively improved the photogenerated electron 20 M. Anpo and P. V. Kamat, Environmentally Benign
ꢁ
ꢁ
and hole separation efficiency and presented a high conversion
rate and product selectivity for the selective oxidation of ben-
zylamine to imine under visible light irradiation.
Photocatalysts: Applications of Titanium Oxide-based
Materials, Springer, London, 2010.
21 D. C. Hurum, A. G. Agrios, K. A. Gray, T. Rajh and
M. C. Thurnauer, J. Phys. Chem. B, 2003, 107, 4545–4549.
22 D. Yang, H. Liu, Z. Zheng, Y. Yuan, J. Zhao, E. R. Waclawik,
X. Ke and H. Zhu, J. Am. Chem. Soc., 2009, 131, 17885–17893.
Acknowledgements
This work was supported by the National Natural Science 23 J. Zhang, Q. Xu, Z. Feng, M. Li and C. Li, Angew. Chem., Int.
Foundation of China (No. 51401233), the Advanced Programs of
Ed., 2008, 47, 1766–1769.
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
24 Z. Zheng, H. Liu, J. Ye, J. Zhao, E. R. Waclawik and H. Zhu, J. 42 S. A. O'Neill, I. P. Parkin, R. J. H. Clark, A. Mills and N. Elliott,
Mol. Catal. A: Chem., 2010, 316, 75–82. J. Mater. Chem., 2003, 13, 2952.
25 T. Kawahara, Y. Konishi, H. Tada, N. Tohge, J. Nishii and 43 J.-M. Jehng and I. E. Wachs, Catal. Today, 1990, 8, 37–55.
S. Ito, Angew. Chem., Int. Ed., 2002, 41, 2811–2813.
26 L. Li, P. A. Salvador and G. S. Rohrer, Nanoscale, 2014, 6, 24–
42.
44 X. Chen, L. Liu, P. Y. Yu and S. S. Mao, Science, 2011, 331,
746–750.
45 L. Zhang, T. Jiang, S. Li, Y. Lu, L. Wang, X. Zhang, D. Wang
and T. Xie, Dalton Trans., 2013, 12998–13003.
27 L. Zhang, D. Jing, X. She, H. Liu, D. Yang, Y. Lu, J. Li,
Z. Zheng and L. Guo, J. Mater. Chem. A, 2014, 2, 2071–2078. 46 H. Zhu, X. Chen, Z. Zheng, X. Ke, E. Jaatinen, J. Zhao, C. Guo,
28 H. Zhu, Z. Zheng, X. Gao, Y. Huang, Z. Yan, J. Zou, H. Yin, T. Xie and D. Wang, Chem. Commun., 2009, 7524–7526.
Q. Zou, S. H. Kable, J. Zhao, Y. Xi, W. N. Martens and 47 Z. Zheng, J. Zhao, H. Liu, J. Liu, A. Bo and H. Zhu,
R. L. Frost, J. Am. Chem. Soc., 2006, 128, 2373–2384. ChemCatChem, 2013, 5, 2382–2388.
29 A. J. Paula, M. A. Zaghete, E. Longo and J. A. Varela, Chem.– 48 D. Jiang, W. Wang, E. Gao, L. Zhang and S. Sun, J. Phys.
Eur. J., 2008, 2008, 1300–1308. Chem. C, 2013, 117, 24242–24249.
30 A. Yu, J. Qian, L. Liu, H. Pan and X. Zhou, Appl. Surf. Sci., 49 A. Yamamoto, Y. Mizuno, K. Teramura, T. Shishido and
2012, 258, 3490–3496. T. Tanaka, Catal. Sci. Technol., 2013, 3, 1771.
31 J. H. Jung, C.-Y. Chen, W.-W. Wu, J.-I. Hong, B. K. Yun, 50 T. A. Westrich, K. A. Dahlberg, M. Kaviany and
Y. Zhou, N. Lee, W. Jo, L.-J. Chen, L.-J. Chou and
Z. L. Wang, J. Phys. Chem. C, 2012, 116, 22261–22265.
32 R. G. Pearson, Coord. Chem. Rev., 1990, 100, 403–425.
33 R. G. Pearson, J. Am. Chem. Soc., 1963, 85, 3533–3539.
34 A. G. Volkov, S. Paula and D. W. Deamer, Bioelectrochem.
Bioenerg., 1997, 42, 153–160.
35 N. Li, L. Zhang, Y. Chen, M. Fang, J. Zhang and H. Wang,
Adv. Funct. Mater., 2012, 22, 835–841.
36 R. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr.,
Theor. Gen. Crystallogr., 1976, 32, 751–767.
J. W. Schwank, J. Phys. Chem. C, 2011, 115, 16537–16543.
51 M. Janus, E. Kusiak-Nejman and A. W. Morawski, React.
Kinet., Mech. Catal., 2012, 106, 289–295.
52 F. Althoff, K. Benzing, P. Comba, C. McRoberts, D. R. Boyd,
S. Greiner and F. Keppler, Nat. Commun., 2014, 5, 4205.
53 M. Zhang, Q. Wang, C. Chen, L. Zang, W. Ma and J. Zhao,
Angew. Chem., Int. Ed., 2009, 48, 6081–6084.
54 Q. Wang, M. Zhang, C. Chen, W. Ma and J. Zhao, Angew.
Chem., Int. Ed., 2010, 49, 7976–7979.
55 C. L. Hawkins, D. I. Pattison and M. J. Davies, Amino Acids,
2003, 25, 259–274.
37 L. Pan, Y. Wang, X. Wang, H. Qu, J. Zhao, Y. Li and
A. Gavrilyuk, Phys. Chem. Chem. Phys., 2014, 16, 20828– 56 C. L. Hawkins and M. J. Davies, Free Radical Biol. Med., 2005,
20833. 39, 900–912.
38 J. Yan, G. Wu, N. Guan and L. Li, Appl. Catal., B, 2014, 152– 57 V. Augugliaro, T. Caronna, V. Loddo, G. Marci, G. Palmisano,
153, 280–288.
L. Palmisano and S. Yurdakal, Chem.–Eur. J., 2008, 14, 4640–
4646.
58 S. Yurdakal, G. Palmisano, V. Loddo, V. Augugliaro and
L. Palmisano, J. Am. Chem. Soc., 2008, 130, 1568–1569.
39 K. Yoshimura, T. Miki, S. Iwama and S. Tanemura, Thin Solid
Films, 1996, 281–282, 235–238.
40 M. Kouhnavard, S. Ikeda, N. A. Ludin, N. B. Ahmad
Khairudin, B. V. Ghaffari, M. A. Mat-Teridi, M. A. Ibrahim, 59 R. Kydd, W. Y. Teoh, J. Scott, D. Ferri and R. Amal,
S. Sepeai and K. Sopian, Renewable Sustainable Energy Rev.,
2014, 37, 397–407.
ChemCatChem, 2009, 1, 286–294.
60 K. Takanabe and K. Domen, ChemCatChem, 2012, 4, 1485–
41 N. Satoh, T. Nakashima, K. Kamikura and K. Yamamoto,
Nat. Nanotechnol., 2008, 3, 106–111.
1497.
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2015