Applied Surface Science 391 (2017) 184–193

Contents lists available at ScienceDirect

Applied Surface Science

journal homepage: www.elsevier.com/locate/apsusc

Enhanced visible light activity on direct contact Z-scheme

g-C3 N4 -TiO2 photocatalyst
Juan Li a,b,c , Min Zhang a,b , Qiuye Li a,b,∗ , Jianjun Yang a,b,∗
a
National & Local Joint Engineering Research Center for Applied Technology of Hybrid Nanomaterials, Henan University, Kaifeng, 475004, China
b
Collaborative Innovation Center of Nano Functional Materials and Applications of Henan Province, Henan University, Kaifeng, 475004, China
c
College of Chemistry and Chemical Engineering, Henan University, Kaifeng, 475004, China

a r t i c l e i n f o a b s t r a c t

Article history: Direct contact Z-scheme g-C3 N4 -TiO2 nanocomposites without an electron mediator are prepared via
Received 10 May 2016 simple annealing the mixture of bulk g-C3 N4 and nanotube titanic acid (NTA) in air at 600 ◦ C for 2 h. In
Received in revised form 22 June 2016 the process of annealing, the bulk g-C3 N4 transformed to ultra-thin g-C3 N4 nanosheets, and NTA con-
Accepted 25 June 2016
verted to a novel anatase TiO2 , then the two components formed a close interaction. The XPS result
Available online 27 June 2016
reveals that some amount of nitrogen is doped into this novel-TiO2 , and g-C3 N4 nanosheets exist in the
composites. The results of XRD, TEM and TG indicate that the thickness of g-C3 N4 nanosheets is very
Keywords:
thin. The ESR spectrum shows the existence of Ti3+ and single-electron-trapped oxygen vacancy in the
g-C3 N4 nanosheets
Z-scheme
30%g-C3 N4 -TiO2 composites. In photocatalytic activity test, the 30%g-C3 N4 -TiO2 nanocomposites showed
Nanotube titanic acid (NTA) an excellent photo-oxidation activity of propylene under visible light irradiation (␭≥ 420 nm), and the
Photo-oxidation of propylene removal efficiency of propylene reached as high as 56.6%, and the activity kept nearly 82% after four
consecutive recycles. Photoluminescence (PL) result using terephthalic acid (TA) as a probe molecule
indicated that the g-C3 N4 -TiO2 nanocomposites displayed a Z-sheme photocatalytic reaction system and
this should be the main reason for the high photocatalytic activity. A possible photocatalytic mechanism
was proposed on the basis of PL result and transient photocurrent-time curves.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction Graphite-like carbon nitride (denoted as g-C3 N4 ) with layered

On account of environment protection, water purification and
utilization solar light for hydrogen production from water splitting,
photocatalysis has become the research focus of numerous schol-
ars in the world. Among the multitudinous photocatalysts, TiO2
has been known as a promising photocatalyst because of non-toxic,
good stability and strong oxidation ability [1,2]. While the intrinsic
imperfection of broad band-gap (3.2 eV for anatase and 3.0 eV for
rutile) resulting in only having activity under UV light irradiation
and the high recombination rate of photoinduced electrons-holes
seriously hindered its practical application [3–5]. Hence it is an
urgent problem to explore photocatalysts possessing high activity
under visible light in practical application. Many modified meth-
ods for TiO2 have been developed to overcome this problem, such
as metal ion doping [6–10], nonmetal doping [11–17], and coupling
with narrow band-gap semiconductors [4,18–20].
∗ Corresponding authors at: National & Local Joint Engineering Research Center for
Applied Technology of Hybrid Nanomaterials, Henan University, Kaifeng, 475004,
China.
E-mail addresses: qiuyeli@henu.edu.cn (Q. Li), yangjianjun@henu.edu.cn
(J. Yang).
two-dimensional planar structure is the most stable crystal form
under ambient conditions and has been a hotspot since Wang et al.
reported g-C3 N4 could produce hydrogen from water under visible
light in 2009 [21]. Although bulk g-C3 N4 has the relative narrow
band-gap (2.7 eV) and excellent visible light response, it is seriously
confined by high photoinduced electrons-holes recombination rate
and small specific surface area (ca 10 m2 /g) in practical application
[22,23]. While g-C3 N4 nanosheets possesses the advantages includ-
ing large specific surface area for providing abundant reactive sites
and short bulk diffusion length for reducing the recombination
probability of photoexcited charge carriers. The g-C3 N4 nanosheets
are mainly obtained by thermal oxidation etching of bulk g-C3 N4
in air [24] or long-time ultrasound exfoliation bulk g-C3 N4 in sol-
vents [25]. For photocatalytic reaction, except for increasing the
surface area of the material, improving the separation efficiency
is also important. Coupling of two different semiconductors with
an appropriate energy band to form the Z-scheme heterojunction
structure is a good way to solve this problem. And the Z-scheme
multi-components photocatalysts are mainly classified into two
kinds: PS-C-PS systems [26] and PS–PS systems [27–33]. The struc-
ture of Z-scheme multi-components photocatalysts can facilitate
the separation of photo-generated carriers and provide the stronger

http://dx.doi.org/10.1016/j.apsusc.2016.06.145
0169-4332/© 2016 Elsevier B.V. All rights reserved.
 J. Li et al. / Applied Surface Science 391 (2017) 184–193 185

(002)
(100)
27.8

g-C3N4 nanosheets

Intensity (a.u.)

(101)
20% g-C3N4-TiO2

(200)
(004)

(105)
(211)

(204)

(215)

(224)
(116)
(220)
30% g-C3N4-TiO2

27.4 50% g-C3N4-TiO2

10 20 30 40 50 60 70 80 90
2 Theta (degree)
Fig. 1. XRD patterns of the prepared photocatalysts.

Fig. 2. TEM images of a) nanotube titanic acid (NTA), b) g-C3N4 nanosheets, and c) 30% g-C3 N4 -TiO2 , and High-resolution TEM image of d) 30% g-C3 N4 -TiO2 .

redox ability. Yu et al. prepared a direct g-C3 N4 /TiO2 Z scheme
photocatalyst without an electron mediator by a facile calcina-
tion route utilizing cost-affordable P25 and urea as feedstock and
its enchanced photocatalytic activities was evaluated by photocat-
alytic oxidation of HCHO in air under UV light irradiation [19]. Li
et al. prepared Ti3+ self-doped TiO2 /g-C3 N4 heterojunctions with
high photocatalytic performance under LED light irradiation by cal-
cination the mixture of H2 Ti3 O7 and melamine [5]. According to the
reported literature, generally speaking, g-C3 N4 was excited by vis-
ible light to produce the charge carriers, while TiO2 was only used
186 J. Li et al. / Applied Surface Science 391 (2017) 184–193
Fig. 3. TG curves for the bulk g-C3 N4, NTA, 20%g-C3 N4 -NTA mixture and g-C3 N4 -TiO2
nanocomposites.
as a supporter in the nanocomposites and usually had no visible
light absorption.
In the previous work of our group, we have clearly clarified
the formation mechanism of nanotube titanic acid (NTA) and
also demonstrated that the novel-TiO2 formed by NTA dehydra-
tion contained large amounts of single-electron-trapped oxygen
vacancy (denoted as V0 • ) and had a certain extent visible light
absorption but had no visible light activity [34]. Therefore we con-
sidered coupling the novel-TiO2 based on NTA as precursor with
g-C3 N4 nanosheets to form direct Z-scheme photocatalyst which
not only could obviously promote the transport of charge carriers
between the novel-TiO2 and g-C3 N4 nanosheets thus effectively
decreased the photoinduced electrons-holes recombination rate
but also could realize nitrogen element doping to boost the visible
light absorption of novel-TiO2 .
In this work, a series of g-C3 N4 -TiO2 direct contact Z-scheme
nanocomposites were performed via simple calcining the mixture
of bulk g-C3 N4 and NTA with different mass percentage at 600 ◦ C
for 2 h. Then their morphology, composition, microstructure and
photocatalytic activity for propylene oxidation removal under visi-
ble light (␭ ≥ 420 nm) were investigated. The results demonstrated
that the composite samples showed significantly enhanced photo-
catalytic activities for propylene photo-oxidation. The main reason
for the excellent photoactivity was due to the high separation effi-
ciency of the photo-excited charge carriers in the direct Z-scheme
heterojunction system, and the higher oxidation ability of O2 −
formed in the CB of g-C3 N4 . At last, on the basis of experiment
results, a possible direct Z-scheme photocatalytic mechanism for
the photo-oxidation of propylene over g-C3 N4 -TiO2 photocatalyst
was proposed.
2. Experimental
2.1. Preparation of C3 N4 -TiO2 nanocomposites
The NTA powder was prepared by alkaline hydrothermal treat-
ment of the commercial P25 TiO2 powder according to our previous
method [3]. Namely, 5 g P25 TiO2 powder was dispersed in aqueous
NaOH (300 mL, 10 M) under magnetic stirring, and then the white
suspension was transferred into Teflon-lined autoclaves and heated
at 150 ◦ C for 24 h. The sediment was washed repeatedly by deion-
ized water until neutral, and then put into hydrochloric acid (0.1 M,
500 mL) for ion-exchange for 7 h, and at last, the suspension was
washed by deionized water until without Cl− existing, followed by
filtration and dried at room temperature under vacuum, and as a
result, the obtained product was NTA.
The bulk g-C3 N4 was prepared by heating melamine at 550 ◦ C in
air for 2 h in a tubular furnace and the heating rate was 5 ◦ C min−1 .
The composites g-C3 N4 -TiO2 was prepared as follows: A certain
amount of bulk g-C3 N4 and NTA powder were ground for 30 min
thoroughly, and then heated at 600 ◦ C for 2 h with a heating rate of
5 ◦ C min−1 in air. The obtained g-C3 N4 -TiO2 nanocomposites were
denoted as x%g-C3 N4 -TiO2 (x% was the mass percentage of bulk
g-C3 N4 in the mixture of bulk g-C3 N4 and NTA powder).
As the reference sample, the pure g-C3 N4 nanosheets, was pre-
pared by annealling of the bulk g-C3 N4 at 600 ◦ C for 2 h in air with
the heating rate of 5 ◦ C min−1 . Another reference sample, NTA-600
(namely novel-TiO2 ) was prepared by annealing NTA under the
same condition with that of pure g-C3 N4 nanosheets.
2.2. Characterization methods
X-ray diffraction (XRD) patterns were recorded on a Ger-
man Bruker D8-ADVANCE diffractometer with Cu K␣ radiation
(␭=0.15418 nm) operated at 40 kV and 40 mA. Transmission elec-
tron microscope (TEM) and high-resolution TEM images were taken
with JEM-2100 operated at 200 kV. Thermogravimetric analysis
(TG) was performed in TGA/SDTA851e thermal analysis instrument
with the heating rate of 10 ◦ C min−1 . The nitrogen adsorption-
desorption isothermals and BET specific surface areas were carried
out on a Quadrasorb SI instrument. The UV–vis absorption spectra
were obtained on a Shimadzu UV-2600 UV/Vis spectrophotome-
ter (reference: BaSO4 ). X-ray photoelectron spectroscopy (XPS)
measurement was performed on a Thermo ESCALAB 250Xi spec-
trometer with an Al-K␣ achromatic X-ray source (voltage 15 kV,
current 10 mA). The C1 s binding energy of hydrocarbon (284.8 eV)
was used as the standard for the correction of charging shift.
The photocurrent-time curve was obtained on CHI600E electricity
chemical analysis instrument in a standard three-electrode photo-
electrochemical cell with 0.1 M NaSO4 as electrolyte under different
visible light irradiation (␭ ≥ 420 nm). Electron spin resonance (ESR)
spectra were obtained on a Bruker EMX-10/12 spectrometer oper-
ated in the X band (ca. 9.80 kHz) with a 100 kHz field modulation,
a 1.0 G amplitude modulation and a 20.4 mW microwave power. G
values were measured with diphenyl picryl hydrazyl (g = 2.0036) as
the reference. The fluorescence spectra were measured by F-7000
FL Spectrophotometer with the excitation wavelength of 315 nm
using terephthalic acid (TA) as a probe molecule to capture the
hydroxyl radicals (• OH) under different visible light irradiation
(␭ ≥ 420 nm).
2.3. Evaluation of photocatalytic activity
The photocatalytic activity of g-C3 N4 -TiO2 nanocomposites
were evaluated by propylene oxidation removal. The photocatalytic
reactor was made of a flat quartz tube, and one 300 W xenon lamp
equipped with a glass optical filter to cut off the short wavelength
components (␭ ≥ 420 nm) as the visible light source was located
outside the reactor [3]. The catalyst was spread on both sides of
surface-roughened glass plate. Feed gas of 600 ppm V (ca. 12 ␮mol)
propylene was made up of pure C3 H6 and air, and was stored
in a high pressure steel cylinder. The change of C3 H6 concentra-
tion before and after reaction was determined by chromatographic
method with hydrogen flame detector.
 J. Li et al. / Applied Surface Science 391 (2017) 184–193 187

Fig. 4. The nitrogen adsorption-desorption isotherms and the corresponding pore size distributions (inset) of the prepared photocatalysts.

NTA
NTA-600
20%g-C3N4-TiO2
Absorbance (a.u.)

30%g-C3N4-TiO2
50%g-C3N4-TiO2
Absorbance (a.u.)

400 500

Wavelength (nm)

200 300 400 500 600 700 800

Wavelength (nm)
Fig. 5. UV–vis diffuse reflectance spectra of the photocatalysts.

3. Results and discussion
3.1. Phase structure and morphology
To identify and determine the phase structures of prepared pho-
tocatalyst samples, XRD was performed. As shown in Fig. 1, g-C3 N4
nanosheets have two main diffraction peaks at 2␪ = 27.8◦ and 16.1◦ ,
which corresponded to (002) interlayer stacking of the aromatic
rings and (100) plane diffractions of g-C3 N4 . The wide peak of 16.1◦
indicated that g-C3 N4 nanosheets were disordered and very thin
which was caused by the decreased planar sizes of the layers in
the thermal oxidation process in air (seen in Fig. 2b). While for the
composites of g-C3 N4 nanosheets and TiO2 , XRD peaks at about
2␪ = 25.4◦ , 37.8◦ , 48.1◦ , 54.0◦ , 55.1◦ , 62.7◦ , 68.8◦ , 70.4◦ , 75.1◦ and
82.7◦ can be clearly observed, which are respectively correspond-
ing to (101), (004), (200), (105), (211), (204), (116), (220), (215) and
(224) crystal faces of anatase TiO2 (PDF card 21-1272, JCPDS). This is
in accordance with our previous finding that NTA is dehydrated to
188 J. Li et al. / Applied Surface Science 391 (2017) 184–193

30%g-C3N4-TiO2 30%g-C3N4-TiO2
O1s A B
C1s

Ti2p
Intensity (a.u.)

Intensity (a.u.)
Ti2s

C1s N1s

0 200 400 600 800 1000 1200 282 284 286 288 290

Binding Energy (eV) Binding Energy (eV)

30%g-C3N4-TiO2 C Ti 2p D
N1s 400.3
N-(C)3
Intensity (a.u.)
Intensity (a.u.)

399.5
3
N-sp C
397.5
Ti-N

NTA-600

30%g-C3N4-TiO2

394 396 398 400 402 456 458 460 462 464 466 468
Binding energy (eV) Binding Energy (eV)

Fig. 6. (A) XPS survey spectrum, high-resolution (B) C1s, (C) N1 s and (D) Ti 2p XPS spectra of 30%g-C3 N4 -TiO2 samples.

form anatase TiO2 when it is annealed above 400 ◦ C [3,34]. No obvi-
ous g-C3 N4 nanosheets diffraction peaks can be observed except for
a small peak at 2␪ = 27.4◦ in 50%g-C3 N4 -TiO2 sample which indi-
cates that the content of g-C3 N4 nanosheets in the composites is
very little (seen in Fig. 3), so the g-C3 N4 nanosheets have almost
no effect on the anatase TiO2 phase structure. Meanwhile, the peak
corresponding to (002) crystal faces of g-C3 N4 nanosheets shifted
from 27.8◦ to 27.4◦ , indicating that there was some close interaction
between g-C3 N4 nanosheets and TiO2 .
Fig. 2 showed TEM images of NTA, g-C3 N4 nanosheets, 30%g-
C3 N4 -TiO2 , and high-resolution TEM image of 30%g-C3 N4 -TiO2 ,
respectively. It is seen that NTA is composed of uniform nanotubes
with the length of several hundreds of nanometers, while the g-
C3 N4 nanosheets are making up of very thin nanosheets which
tend to bend and its edges are wrinkled as a result of minimizing
the surface energy of the sheets. What is more, many micropores
can be obvserved in g-C3 N4 thin nanosheets, which were caused by
bulk g-C3 N4 decomposition into some gaseous products in the ther-
mal oxidation etching [24]. For the optimum photocatalytic activity
composites of 30%g-C3 N4 -TiO2 (seen in Fig. 2c and d), it contains
a few g-C3 N4 nanosheets and novel-TiO2 nanoparticles with the
diameter of about 20 nm. The lattice fringe of 0.31 nm and 0.35 nm
are corresponding to (002) facet of g-C3 N4 and (101) facet of TiO2
respectively, indicating the existence of the two components.
3.2. TG analysis
The amount of g-C3 N4 in the composite samples was character-
ized by thermal gravimetric analysis. For the samples of g-C3 N4 ,
NTA and the physical mixture (uncaclined) of g-C3 N4 and NTA, the
TG experiments (seen in Fig. 3) were carried out in air atomosphere
from room temperature to 850 ◦ C at a heating rate of 10 ◦ C min−1 .
It can be seen that NTA has a weight loss of about 15% because of
the dehydration of intralayer and interlayer OH groups [34], while
the pure g-C3 N4 was stable before 520 ◦ C then disintegrated dra-
matically and finally decomposed completely at 740 ◦ C. The sample
of 20%g-C3 N4 -NTA-mixture had a weight loss of about 35% and
the final decomposed temperature reduced about 100 ◦ C, which
resulted from the reason that the g-C3 N4 could be catalytically oxi-
dized by the coupled TiO2 [35,36]. For the samples of g-C3 N4 -TiO2
nanocomposites, there is only very little weight loss (ca. 2%) at the
temperature range of 25–850 ◦ C, which indicates that the content
of g-C3 N4 nanosheets is very low. That is to say, in the process of the
thermal oxidation etching of bulk g-C3 N4 to g-C3 N4 nanosheets, the
 J. Li et al. / Applied Surface Science 391 (2017) 184–193 189

60
A 56.6%

The removal rate of propylene (%)

50

40.8%
40

30 28.1%

21.1%
20

10

1.5% 2.8%
0 0
NTA NTA-600 g-C3 N4 nanosheets

2
2

iO
iO

iO
iO

4 -T
4 -T

4 -T

4 -T

3N
3 N

3 N
3N

C
C

C
C

g-
g-

g-
g-

%
%
%

50
40
20

30

70
B 1st 2nd 3rd 4th
60
The removal rate of propylene (%)

50

40

30

20

10

0 2 4 6 8

Time/ h

Fig. 7. (A) Comparison of visible light photocatalytic propylene activity of as prepared photocatalysts, (B) Stability test of the sample 30%g-C3 N4 -TiO2 for photocatalytic
degradation propylene under visible light irradiation.

mass of g-C3 N4 decreased a lot and became to a lighter structure, (inset). It can be seen that g-C3 N4 nanosheets mainly contains
which was in accordance with the TEM results. slitlike pores concentrated at 3–5 nm and mesopores at 6–12 nm,
while the pore sizes of g-C3 N4 -TiO2 composites are almost larger
mesopores located at 5–30 nm.
3.3. Nitrogen adsorption
From Table 1, it can be seen that g-C3 N4 nanosheets has the
biggest BET specific surface areas, average pore diameter and pore
The BET specific surface areas and pore size distributions of pure
volume. While increasing with the content of bulk g-C3 N4 , the BET
g-C3 N4 nanosheets and the g-C3 N4 -TiO2 composites were inves-
specific surface areas, average pore diameter and pore volume of g-
tigated by nitrogen adsorption-desorption testing. It can be seen
C3 N4 -TiO2 composites present the feature of rising firstly and then
from Fig. 4 that all the as-prepared photocatalysts have isotherms of
decreasing. The reason was as follows: Due to the specific surface
type IV according to the Brunauer-Deming-Deming-Teller (BDDT)
areas of NTA (312.3 m2 /g) is much larger than that of bulk g-C3 N4
classification, indicating the presence of mesopores.
(10.3 m2 /g), so it is generally speaking that the specific surface areas
For all the prepared photocatalysts, the shapes of the hystere-
of composites will reduce with increasing with the content of bulk
sis loops are the type H3 coming from layered structure materials
g-C3 N4 . However, with the increase of the content of bulk g-C3 N4 ,
at the relative pressure range of 0.5-1.0, indicating the presence of
more gas products such as NH3 , CO2 etc. would be released during
slitlike pores. The pore size distributions curves calculated accord-
the thermal oxidation etching, which might dash the layers of g-
ing to Barrett-Joyner-Halenda (BJH) method were shown in Fig. 4
190 J. Li et al. / Applied Surface Science 391 (2017) 184–193

3.6
A B
3.5
e
3.4
Current density (µ A/cm2)

3.3
d

Intensity (a.u.)
3.2
g=1.978
c
3.1

3.0
b
2.9 g=2.001
a
2.8

2.7

2.6
0 100 200 300 400 500 600 700 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5
Time (sec) g

Fig. 8. (A) Photocurrent-time curves of the catalyst samples a: NTA-600, b: g-C3 N4 nanosheets, c: 50% g-C3 N4 -TiO2 , d: 20% g-C3 N4 -TiO2 , e: 30% g-C3 N4 -TiO2 . (B) ESR spectra
of 30%g-C3 N4 -TiO2 measured at room temperature in atmosphere.

Table 1
Physical properties of the as-prepared photocatalysts.

Sample SBET /(m2 /g) Average Pore Diameter/(nm) Pore Volume/(cm3 /g)

g-C3 N4 nanosheets 90.1 31.4 0.333

NTA-600 77.4 30.9 0.408
20%g-C3 N4 -TiO2 66.2 27.2 0.349
30%g-C3 N4 -TiO2 75.5 27.4 0.443
50%g-C3 N4 -TiO2 52.9 26.8 0.297

C3 N4 to form more larger pores and then the BET surface areas,
average pore diameter and pore volume of composites augmented.
Based on the two parameters, and as a result, the largest BET surface
area was obtained on the sample of 30%g-C3 N4 -TiO2 composites.
3.4. UV–vis diffuse reflectance spectrum
Fig. 5 showed the UV–vis diffuse reflectance spectra of the NTA,
NTA-600 and g-C3 N4 -TiO2 composites with different composition.
It can be seen that the absorption edges of g-C3 N4 -TiO2 nanocom-
posites red-shifted apparently to visible light region as compared
with NTA and NTA-600. According to the results of calculation,
the band gaps of 20%, 30% and 50% g-C3 N4 -TiO2 composites are
3.08, 3.06 and 3.04 eV respectively. This indicated that g-C3 N4 could
act as a photo-sensitizer and effectively extend the absorption of
TiO2 to visible light range. The additional visible light absorption at
400–470 nm comparing with NTA-600 attributed to the existence
of g-C3 N4 , and the absorption intensity of g-C3 N4 -TiO2 composites
strengthened with the increase of the content of g-C3 N4 .
As shown from Fig. 6B, two peaks centered at 284.8 and 288.9 eV
could be observed. The one at 284.8 eV is assigned to contamina-
tive hydrocarbon on the sample surface, and another corresponds
to sp2 C atoms in C3 N3 triazine cycles, demonstrating that C3 N4
existences in the as-synthesized nanocomposites. The high reso-
lution N1 s spectra of 30%g-C3 N4 -TiO2 (seen in Fig. 6C) showed
that three peaks located at 397.5, 399.5 and 400.2 eV appeared.
They are respectively corresponding to Ti-N [16,37], N-sp3 C and
N-(C)3 , and the latter two peaks indicate the existence of g-C3 N4
in the composites [38,39]. According to our previous work [17],
the formation of SETOV in TiO2 matrix (seen in Fig. 8B) results in
visible light response, while the doped nitrogen facilitates to pre-
vent photoinduced electrons and holes from recombination. The
two peaks in the high-resolution XPS Ti 2p spectrum of 30%g-
C3 N4 -TiO2 in Fig. 6D were respectively Ti 2p3/2 and Ti 2p1/2 peaks
and the slightly shifted to higher binding energy comparing with
that of NTA-600 indicated that it attributed to the close interaction
between g-C3 N4 and TiO2 [4]. Such an interaction may facilitate the
charge separation between g-C3 N4 and TiO2 , thereby improving the
photocatalytic activity.

3.5. XPS analysis

XPS testing was employed to study the elemental composition
and identify the chemical state of elements in g-C3 N4 -TiO2 hybrid
composites.
Fig. 6A showed that the XPS survey spectrum of 30%g-C3 N4 -TiO2
contains the four elements of Ti, O, C, N with sharp photoelectron
peaks, but the intensity of N peak was very low which suggested
that the content of N element was very low. The reason was that
during thermal oxidation etching of bulk g-C3 N4 , most of C atoms
and N atoms were lost to form the very thin g-C3 N4 nanosheets
[24].
3.6. The photocatalytic activity, photocurrent curve and ESR
spectra
The photocatalytic activity of these samples was tested
by photo-oxidation propylene under visible light irradiation
(␭ ≥ 420 nm). As shown from Fig. 7A, the photo-oxidation propy-
lene activity of hybrid composites is much better than that of pure
g-C3 N4 nanosheets and NTA-600. As the increase of the amount of
g-C3 N4 , the removal efficiency of propylene presents the feature of
rising firstly and then decreasing. This result is consistent with that
of BET surface areas, average pore diameter, pore volume.
 J. Li et al. / Applied Surface Science 391 (2017) 184–193 191

100 100

A NTA-600 B
20%g-C3N4-TiO2
80 30%g-C3N4-TiO2

Fluorescence intensity (a.u.)

80

50%g-C3N4-TiO2
Fluorescence intensity

60 60min 60 g-C3N4 nanosheet

50min
40min
30min 40
40
20min
10min
0 min
20 20

0 0

360 380 400 420 440 460 480 500 520 540 0 10 20 30 40 50 60
Wavelength (nm) Irradiation time (min)

Fig. 9. (A) PL spectral changes observed during visible light irradiation for 30%g-C3 N4 -TiO2 sample in 2 × 10−3 M NaOH solution with presence of 5 × 10−4 M terephthalic
acid (excitation wave: 315 nm), (B) Comparison of PL intensity@425 nm against irradiation time for as-prepared photocatalyts.

Fig. 10. The schematic diagram for the photocatalytic mechanism in g-C3 N4 -TiO2 photocatalysts was proposed under visible light irradiation.

The stability of the 30%g-C3 N4 -TiO2 composite was evalutated,
and as shown in Fig. 7B. The removal rate of propylene was 56.6%,
48.3%, 46.4% and 46.2% when it is subjected to 4-cycle consecutive
photocatalytic tests over a period of up to 8 h. This indicates that the
g-C3 N4 -TiO2 nanocomposites possess a very good stability in long-
term-run photocatalytic reactions because the solid–solid contact
interface connected by strong chemical bonds [40].
To reveal the difference of photocatalytic ability in g-C3 N4 -
TiO2 hybrid composites compared with g-C3 N4 nanosheets and
NTA-600, we perform the photocurrent curve test and ESR spectra
test. The photocurrent curve is obtained via recording the tran-
sient photocurrent density immediately increases or decreases
upon turning-on or turning-off the visible light source (␭ ≥ 420 nm)
which can indicate the strength of the electron transport ability.
While the ESR spectra can demonstrate the existence of radical
species containing unpaired electron.
As shown from Fig. 8A, we can see that the photocurrent density
of hybrid composites is obviously larger than that of pure g-C3 N4
nanosheets and NTA-600, which indicates the photo-generated
charge carriers of g-C3 N4 -TiO2 nanocomposites possess much
higher separation efficiency because the thin g-C3 N4 nanosheets
provides a large surface area and short bulk diffusion length along
the in-plane direction [24]. And 30% g-C3 N4 -TiO2 composite has the
strongest photocurrent in the composites which is due to the syn-
ergy effect of the biggest BET surface areas, average pore diameter,
pore volume and the highest charge separation efficiency. That is
consistent with the photoactalytic activity (seen in Fig. 7A).
Fig. 8B indicates that there are two sharp signal peaks corre-
sponding to Ti3+ (g = 1.978) and single-electron-trapped oxygen
vacancies (SETOV, g = 2.001) in 30% g-C3 N4 -TiO2 composite. That
confirms well to our previous work [3] that the novel anatase TiO2
containing a large amount of SETOV can be obtained from the dehy-
dration of NTA. In addition, the released NH3 during the thermal
oxidation etching of bulk g-C3 N4 will reduce Ti4+ to Ti3+ in TiO2 , so
the Ti3+ signal peak can be observed in 30% g-C3 N4 -TiO2 composite.
3.7. Photocatalytic mechanism of g-C3 N4 -TiO2 composites
To understand the photocatalytic mechanism, the hydroxyl rad-
icals (• OH) formed on the surface of g-C3 N4 nanosheets, NTA-600
and g-C3 N4 -TiO2 composite samples under visible-light irradiation
were detected by photoluminescence (PL) test using terephthalic
acid (TA) as a probe molecule which readily reacted with • OH to
produce a highly fluorescent product, 2-hydroxyterephthalic acid
(TA-OH) [20].
From Fig. 9A, it can be seen that the concentration of hydroxyl
radicals (• OH) on the surface of 30%g-C3 N4 -TiO2 as the optimum
activity sample grew sharply with the increase of visible light irra-
192 J. Li et al. / Applied Surface Science 391 (2017) 184–193
diation time. While Fig. 9B demonstrated that a gradual augment in
PL intensity of all samples at about 425 nm was observed with the
increase of irradiation time. This result was due to that the poten-
tial of valence band (VB) of g-C3 N4 (+1.4 eV) was higher than that
of OH− /• OH couple (+2.8 eV), so the holes could not reacted with
OH− to form • OH radicals. While the holes in VB of TiO2 could
oxidize OH− to produce • OH radicals on account of the potential
of VB of TiO2 (+2.9 eV) was lower than that of OH− /• OH couple
(+2.8 eV). According to the above analysis, g-C3 N4 -TiO2 composites
were direct contact Z-scheme structure (photocatalytic mechanism
seen in Fig. 10) instead of conventional heterojunction structure.
Furthermore, g-C3 N4 -TiO2 composites could produce higher con-
centration of • OH radicals than that of NTA-600 in any irradiation
time (seen in Fig. 9B), which could be explained that the equilib-
rium between the CB of electrons and holes of VB of TiO2 was broken
because of the prompt charge transfer and separation of Z-scheme
structure and so more holes were produced in the VB of TiO2 to
form more • OH radicals.
Based on the above results and discussion, the proposed
photocatalytic mechanism for photocatalytic degradation of propy-
lene on g-C3 N4 -TiO2 composites was shown in Fig. 10. For
NTA-600, under visible light irradiation, the recombination of
photo-generated electrons and holes is rapid and only a small part
of them could participated in the photocatalytic reaction which
resulted in the very low activity. While for the direct contact Z-
scheme structure of g-C3 N4 -TiO2 composite, under visible light
irradiation the electrons jumped from the VB of TiO2 to sub-band
(formed by large amounts of oxygen vacancy) and then jumped to
the CB of TiO2 . According to the work of Wang et al. [17], doped N is
able to avoid photogenerated e− -h+ from recombination, making it
feasible for photogenerated electrons to transfer to the catalyst sur-
face and facilitating the visible light photo-oxidation of propylene.
Subsequently, the electrons in the CB of TiO2 migrated to combine
with the holes in VB of g-C3 N4 under the action of internal elec-
tric field. The electrons kept in CB of g-C3 N4 and holes in VB of
TiO2 possess the much higher redox ability than those in the con-
ventional heterojucntion system, so its photocatlytic activity was
better. And this is the advantage of the Z-scheme system between
the semiconductors.
4. Conclusions
A series of direct contact Z-scheme g-C3 N4 -TiO2 nanocompos-
ites were prepared via simple annealing the mixture of nanotube
titanic acid and bulk g-C3 N4 . In the thermal treament, nanotube
titanic acid dehydrated to generate the novel-TiO2 while bulk g-
C3 N4 was oxidized etching by the oxygen in the air to form very
thin g-C3 N4 nanosheets which could provide large specific surface
area and short bulk diffusion length to facilitate the transport of car-
riers. And meanwhile a small quantity of nitrogen was doped into
the novel-TiO2 , consequently the direct contact Z-scheme hetero-
junction was formed. Doped N is apt to stablize oxygen vacancy to
cut the route of electrons jumping from sub-band to the VB of TiO2
which is in favor of the charge transfer and separation. The photo-
catalyst of 30%g-C3 N4 -TiO2 demonstrated the best photocatalytic
activity for propylene oxidation under visible light irradiation. The
removal efficiency of propylene reached as high as 56.6%, and the
activity kept nearly 82% after four consecutive recycles. This could
be contributed to the synergy effect of the biggest BET surface areas,
average pore diameter, pore volume and the excellent charge sep-
aration efficiency of the direct contact Z-scheme heterojunction
system composed of the very thin g-C3 N4 nanosheets and novel-
TiO2 nanoparticle.
Acknowledgments
This work was supported by the National Natural Science Foun-
dation of China (No. 21471047), Program for Science & Technology
Innovation Talents (15HASTIT043) and Innovative Research Team
(16IRTSTHN015) from the University of Henan Province, and Pro-
gram for Changjiang Scholars and Innovative Research Team in
University from the Ministry of Education of China (PCS IRT1126).
References
[1] M.R. Hoffmann, S.T. Martin, W. Choi, D.W. Bahneman, Environmental
applications of semiconductor photocatalysis, Chem. Rev. 95 (1995) 69–96.
[2] X. Chen, S.S. Mao, Titanium dioxide nanomaterials: synthesis, properties,
modifications, and applications, Chem. Rev. 107 (2007) 2891–2959.
[3] H. Li, F. Ren, J. Liu, Q. Wang, Q. Li, J. Yang, Y. Wang, Endowing
single-electron-trapped oxygen vacancy self-modified titanium dioxide with
visible-light photocatalytic activity by grafting Fe(III) nanocluster, Appl. Catal.
B: Environ. 172–173 (2015) 37–45.
[4] Q. Li, L. Zong, Y. Xing, X. Wang, L. Yu, J. Yang, Preparation of g-C3 N4 /TiO2
nanocomposites and investigation of their photocatalytic activity, Sci. Adv.
Mater. 5 (2013) 1316–1322.
[5] K. Li, S. Gao, Q. Wang, H. Xu, Z. Wang, B. Huang, Y. Dai, J. Lu, In-situ-reduced
synthesis of Ti3+ self-doped TiO2 /g-C3 N4 heterojunctions with high
photocatalytic performance under LED light irradiation, ACS Appl. Mater.
Interfaces 7 (2015) 9023–9030.
[6] Z.-L. Shi, C. Du, S.-H. Yao, Preparation and photocatalytic activity of cerium
doped anatase titanium dioxide coated magnetite composite, J. Taiwan Inst.
Chem. Eng. 42 (2011) 652–657.
[7] J. Fan, E.-z. Liu, L. Tian, X.-y. Hu, Q. He, T. Sun, Synergistic effect of N and Ni2+
on nanotitania in photocatalytic reduction of CO2 , J. Environ. Eng. 137 (2011)
171–176.
[8] L. Sun, X. Zhao, X. Cheng, H. Sun, L. Yanlu, P. Li, W. Fan, Synergistic effects in
La/N codoped TiO2 anatase (101) surface correlated with enhanced
visible-light photocatalytic activity, Langmuir 28 (2012) 5882–5891.
[9] A. Escudero, F. Langenhorst, Chromium incorporation into TiO2 at high
pressure, J. Solid State Chem. 190 (2012) 61–67.
[10] K. Bhattacharyya, J. Majeed, K.K. Dey, P. Ayyub, A.K. Tyagi, S.R. Bharadwaj,
Effect of Mo-incorporation in the TiO2 lattice: a mechanistic basis for
photocatalytic dye degradation, J. Phys. Chem. C 118 (2014)
15946–15962.
[11] Y. Wang, C. Feng, M. Zhang, J. Yang, Z. Zhang, Enhanced visible light
photocatalytic activity of N-doped TiO2 in relation to single-electron-trapped
oxygen vacancy and doped-nitrogen, Appl. Catal. B: Environ. 100 (2010)
84–90.
[12] J. Yu, Q. Li, S. Liu, M. Jaroniec, Ionic-liquid-assisted synthesis of uniform
fluorinated B/C-codoped TiO2 nanocrystals and their enhanced visible-light
photocatalytic activity, Chem. Eur. J. 19 (2013) 2433–2441.
[13] P. Simon, B. Pignon, B. Miao, S. Coste-Leconte, Y. Leconte, S. Marguet, P. Jegou,
B. Bouchet-Fabre, C.c. Reynaud, N. Herlin-Boime, N-doped titanium monoxide
nanoparticles with TiO rock-salt structure, low energy band gap, and visible
light activity, Chem. Mater. 22 (2010) 3704–3711.
[14] V.c. Štengl, V. Houškova, S. Bakardjieva, N. Murafa, Photocatalytic activity of
boron-modified titania under UV and visible-light illumination, ACS Appl.
Mater. Interfaces 2 (2010) 575–580.
[15] X. Wang, M. Blackford, K. Prince, R.A. Caruso, Preparation of boron-doped
porous titania networks containing gold nanoparticles with enhanced
visible-light photocatalytic activity, ACS Appl. Mater. Interfaces 4 (2012)
476–482.
[16] R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki, Y. Taga, Visible-light photocatalysis
in nitrogen-doped titanium oxides, Science 293 (2001) 269–271.
[17] Y. Wang, C. Feng, M. Zhang, J. Yang, Z. Zhang, Visible light active N-doped TiO2
prepared from different precursors: origin of the visible light absorption and
photoactivity, Appl. Catal. B: Environ. 104 (2011) 268–274.
[18] W.-K. Jo, T.S. Natarajan, Influence of TiO2 morphology on the photocatalytic
efficiency of direct Z-scheme g-C3 N4 /TiO2 photocatalysts for isoniazid
degradation, Chem. Eng. J. 281 (2015) 549–565.
[19] J. Yu, S. Wang, J. Low, W. Xiao, Enhanced photocatalytic performance of direct
Z-scheme g-C3 N4 /TiO2 photocatalyst for decomposition of formaldehyde in
air, Phys. Chem. Chem. Phys. 15 (2013) 16883–16890.
[20] S. Kumar, A. Baruah, S. Tonda, B. Kumar, V. Shanker, B. Sreedhar,
Cost-effective and eco-friendly synthesis of novel and stable N-doped
ZnO/g-C3 N4 core-shell nanoplates with excellent visible-light responsive
photocatalysis, Nanoscale 6 (2014) 4830–4842.
[21] X. Wang, K. Maeda, A. Thomas, K. Takanabe, G. Xin, J.M. Carlsson, K. Domen,
M. Antonietti, A metal-free polymeric photocatalyst for hydrogen production
from water under visible light, Nat. Mater. 8 (2009) 76–80.
[22] X. Wang, S. Blechert, M. Antonietti, Polymeric graphitic carbon nitride for
heterogeneous photocatalysis, ACS Catal. 2 (2012) 1596–1606.
[23] S.C. Yan, Z.S. Li, Z.G. Zou, Photodegradation performance of g-C3 N4 fabricated
by directly heating melamine, Langmuir 25 (2009) 10397–10401.
 J. Li et al. / Applied Surface Science 391 (2017) 184–193
[24] P. Niu, L. Zhang, G. Liu, H.-M. Cheng, Graphene-like carbon nitride nanosheets
for improved photocatalytic activities, Adv. Funct. Mater. 22 (2012)
4763–4770.
[25] S. Yang, Y. Gong, J. Zhang, L. Zhan, L. Ma, Z. Fang, R. Vajtai, X. Wang, P.M.
Ajayan, Exfoliated graphitic carbon nitride nanosheets as efficient catalysts
for hydrogen evolution under visible light, Adv. Mater. 25 (2013)
2452–2456.
[26] J. Fu, S. Cao, J. Yu, Dual Z-scheme charge transfer in TiO2 -Ag-Cu2 O composite
for enhanced photocatalytic hydrogen generation, J. Materiom. 1 (2015)
124–133.
[27] D. Xu, B. Cheng, S. Cao, J. Yu, Enhanced photocatalytic activity and stability of
Z-scheme Ag2 CrO4 -GO composite photocatalysts for organic pollutant
degradation, Appl. Catal. B: Environ. 164 (2015) 380–388.
[28] W. Yu, D. Xu, T. Peng, Enhanced photocatalytic activity of g-C3 N4 for selective
CO2 reduction to CH3 OH via facile coupling of ZnO: a direct Z-scheme
mechanism, J. Mater. Chem. A 3 (2015) 19936–19947.
[29] Y. Liu, R. Wang, Z. Yang, H. Du, Y. Jiang, C. Shen, K. Liang, A. Xu, Enhanced
visible-light photocatalytic activity of Z-scheme graphitic carbon
nitride/oxygen vacancy-rich zinc oxide hybrid photocatalysts, Chin. J. Catal.
36 (2015) 2135–2144.
[30] J. Lv, K. Dai, J. Zhang, L. Geng, C. Liang, Q. Liu, G. Zhu, C. Chen, Facile synthesis
of Z-scheme graphitic-C3 N4 /Bi2 MoO6 nanocomposite for enhanced visible
photocatalytic properties, Appl. Surf. Sci. 358 (2015) 377–384.
[31] J. Jin, J. Yu, D. Guo, C. Cui, W. Ho, A hierarchical Z-Scheme CdS-WO3
photocatalyst with enhanced CO2 reduction activity, Small 11 (2015)
5262–5271.
[32] D. Xiao, K. Dai, Y. Qu, Y. Yin, H. Chen, Hydrothermal synthesis of
␣-Fe2 O3 /g-C3 N4 composite and its efficient photocatalytic reduction of Cr(VI)
under visible light, Appl. Surf. Sci. 358 (2015) 181–187.
193
[33] W. Chen, T. Liu, T. Huang, X. Liu, J. Zhu, G. Duan, X. Yang, In situ fabrication of
novel Z-scheme Bi2 WO6 quantum dots/g-C3 N4 ultrathin nanosheets
heterostructures with improved photocatalytic activity, Appl. Surf. Sci. 355
(2015) 379–387.
[34] M. Zhang, Z. Jin, J. Zhang, X. Guo, J. Yang, W. Li, X. Wang, Z. Zhang, Effect of
annealing temperature on morphology, structure and photocatalytic behavior
of nanotubed H2 Ti2 O4 (OH)2 , J. Mol. Catal. A: Chem. 217 (2004)
203–210.
[35] Y. He, J. Cai, T. Li, Y. Wu, H. Lin, L. Zhao, M. Luo, Efficient degradation of RhB
over GdVO4 /g-C3 N4 composites under visible-light irradiation, Chem. Eng. J.
215-216 (2013) 721–730.
[36] T. Li, L. Zhao, Y. He, J. Cai, M. Luo, J. Lin, Synthesis of g-C3 N4 /SmVO4 composite
photocatalyst with improved visible light photocatalytic activities in RhB
degradation, Appl. Catal. B: Environ. 129 (2013) 255–263.
[37] A. Ghicov, J.M. Macak, H. Tsuchiya, J. Kunze, V. Haeublein, L. Frey, P. Schmuki,
Ion implantation and annealing for an efficient N-doping of TiO2 nanotubes,
Nano Lett. 6 (2006) 1080–1082.
[38] Z. Sheng, L. Shao, J. Chen, W. Bao, F. Wang, X. Xia, Catalyst-free synthesis of
nitrogen-doped graphene via thermal annealing graphite oxide with
melamine and its excellent electrocatalysis, ACS Nano 5 (2011) 4350–4358.
[39] Y. Li, H. Zhang, P. Liu, D. Wang, Y. Li, H. Zhao, Cross-linked g-C3 N4 /rGO
nanocomposites with tunable band structure and enhanced visible light
photocatalytic activity, Small 9 (2013) 3336–3344.
[40] P. Zhou, J. Yu, M. Jaroniec, All-solid-state Z-scheme photocatalytic systems,
Adv. Mater. 26 (2014) 4920–4935.