Journal of Hazardous Materials 410 (2021) 124594

Contents lists available at ScienceDirect

Journal of Hazardous Materials

journal homepage: [Link]/locate/jhazmat

Research paper

One-step synthesis of nitrogen-defective graphitic carbon nitride for

improving photocatalytic hydrogen evolution
Qing Yang a, Wenbin Yang b, *, Fangfang He b, Kewei Liu a, Hongmei Cao a, Hongjian Yan a, *
a
College of Chemistry, Sichuan University, No.24 South Section 1, Yihuan Road, Chengdu 610065, China
b
State Key Laboratory of Environment-friendly Energy Materials, Southwest University of Science and Technology, 59 Qinglong Road, Mianyang 621010, Sichuan,
China

A R T I C L E I N F O A B S T R A C T

Editor: Dr. Daniel C.W. Tsang Defects are significant for graphitic carbon nitride (g-C3N4, CN) in photocatalytic applications. Defective CN not
only accelerate charge separation but also extend spectral response. Engineering carbon or nitrogen defects in CN
Keywords: has been achieved by variety of strategies, but it is still a long-term interest to develop a simple and controllable
Carbon nitride route for engineering defects in CN. Herein, we present tuning the nitrogen defects in CN by either changing the
Nitrogen defects
heating rate or prolonging the pyrolysis time during polymerization melamine sulfate. It was found that either
One-step synthesis
lower heating rate or longer pyrolysis time lead to the formation of more N vacancies and suspended terminal
Photocatalytic H2 evolution
amino. As a result, an optimal photocatalytic H2 yield rate (λ > 420 nm) of 905 μmol g− 1 h− 1 was reached, which
was 2 times higher than that of CN prepared with a heating rate of 1 ◦ C/min and pyrolysis at 600 ◦ C for 4 h.

1. Introduction so on, (Yan et al., 2009; Zhang et al., 2012; Han et al., 2016) it is urgency
to extend visible-light spectral response, suppress charge recombination,
Graphitic carbon nitride (g-C3N4, CN), which is similar to graphene create more active sites. Until now, a variety of strategies, such as
with two-dimensional (2D) conjugated lamellar structure, and consists nanostructure engineering (morphology controlling, surface modifica­
of heptazine rings connected by van der Waals interaction and hydrogen tion, tuning the porosity), (Hong et al., 2012; Jun et al., 2013) con­
bond, is a metal-free n-type semiconductor. CN was usually prepared by structing heterojunctions with other semiconductors, (Li et al., 2013a,
thermally polymerizing precursors like dicyandiamide, urea, thiourea, 2013b; Xu et al., 2013a, 2013b), controlling crystallinity, heteroatom
etc. (Cao et al., 2015; Lee et al., 2017; Thomas et al., 2008; Wang et al., doping (B, S, P, Fe, Zn, Cu, etc) or copolymerization, (Tonda et al., 2014;
2009a, 2009b; Zhang et al., 2010) Owing to its unique features such as Guo et al., 2016; Li et al., 2016; Ran et al., 2015; Liang et al., 2015), etc,
energy band gap (2.7 eV) for absorbing visible light with wavelength as a common strategy of tuning its photocatalytic performance, have
longer than 450 nm, (Wang et al., 2009a, 2009b) suitable band levels for been attempted to either extend visible-light response range or tune its
water oxidation and reduction, (Liu et al., 2015) easily-modified mo­ electronic structure. For example, macroscopic three-dimensional (3D)
lecular structure, good stability, CN has gathered extensive interest as CN monolith showed higher hydrogen evolution rate than CN powder
efficient metal-free photocatalyst for water splitting, degradation of (2D nanosheet) (Hao et al., 2018). The formation of heterojunction
pollutants, organic synthesis, and CO2 reduction, etc. (Niu et al., 2012a, between CN and other semiconductors (such as TiO2, ZnS, P3HT, co­
2012b; Zhang et al., 2015; Fu et al., 2018; Ye et al., 2013; Mamb et al., valent organic frameworks, etc) can accelerate charges’ mobility, lead­
2016; Dong et al., 2018; Barrio et al., 2018; Hong et al., 2019). ing to higher hydrogen evolution rate than that of pure CN (Chen et al.,
Considering that the practical use of unmodified CN without any 2014; Bai et al., 2015; Luo et al., 2019a, 2019b). Besides, another gen­
optimization on its chemical structure and photoelectric properties as eral method to tune band structure and extend spectral response is
photocatalyst is still faced with various obstacles, such as narrow visible- introducing other atoms in semiconductor. Although elements like P, S,
light response region (with wavelength up to 460 nm, whereas the O, I, Fe, etc were successfully doped in melon structure and could
visible light between 400 and 700 nm is ca. 43% in the solar spectrum), effectively enhance visible-light absorption (Jiang et al., 2017), only
facile recombination of photogenerated charges, low surface area, and little band gap of melon units was changed (Li et al., 2012; Wang et al.,

* Corresponding authors.
E-mail addresses: yangwenbin@[Link] (W. Yang), hjyan@[Link] (H. Yan).

[Link]
Received 1 September 2020; Received in revised form 22 October 2020; Accepted 13 November 2020
Available online 24 November 2020
0304-3894/© 2020 Elsevier B.V. All rights reserved.
Q. Yang et al. Journal of Hazardous Materials 410 (2021) 124594

2015). The CN modified by these strategies are believed to be valid for during the polymerization process was suppressed by applying mel­
enhancing photocatalytic performance. amine sulfate as precursor (Yan et al., 2012). During heating, the mel­
The surface structure and optical properties of CN have a vital in­ amine sulfate decomposes gradually and releases H2SO4. The H2SO4 can
fluence on its activity. Some studies indicated that the defects/vacancies intercalate in the interlayer space of CN (Xu et al., 2013a, 2013b), and
within structure serve as active sites, play a positive role in tuning en­ act as strong oxidant at high temperature. Herein, we report a facile
ergy band gap and electronic properties (Tay et al., 2015; Niu et al., one-step strategy to fabricate CN and tune its nitrogen defects simulta­
2018). Defect engineering in semiconductors is considered com­ neously by either changing the heating rate or prolonging the pyrolysis
plementally as another vital approach to optimize the activity. As a time during polymerization melamine sulfate. It was found that either
typical example of defect engineering, black hydrogenated TiO2 with lower heating rate or longer pyrolysis time lead to more defects, and
disordered nanophase and many oxygen vacancies exhibited excellent consequently a red shift of light absorption edge of CN. As a result, the
visible-light response with a prominent red shift of light-absorption edge photocatalytic activity of CN with more defects was enhanced, which
from 375 to 1100 nm and high photocatalytic activity for H2 evolution was 2 times higher than that of CN prepared by higher heating rate and
under visible-light irradiation (Chen et al., 2011). Whereas unmodified shorter calcine time.
TiO2 has actually no response to visible light. Although higher crystal­
linity of most semiconductors has been proven to be beneficial for 2. Experimental
optimizing photocatalytic performance, partial structural distortion can
result in the change of electronic affinity located in different domain, 2.1. The synthesis of defective CN
accelerating charge mobility and extending spectral response range (Ou
et al., 2017). Consequently, in recent years, defect engineering has been The melamine sulfate was prepared according to our previous work
popular as an effective strategy to modulate the photocatalytic perfor­ (Yan et al., 2012). 10 g melamine sulfate was placed into a semiclosed
mance of semiconductor photocatalysts. Defects or amorphous structure quartz tube which was then put in a cannular muffle furnace. The
in CN also plays a vital role in tuning its electronic structure and spectral melamine sulfate was firstly heated to 380 ◦ C within 70 min, then to
response, which consequently improves its photocatalytic activity (Yu 600 ◦ C with different heating rate (0.01–1 ◦ C/min) and finally held at
et al., 2016; Sun et al., 2017). When nitrogen vacancies were introduced 600 ◦ C for 4 h. The sample was labeled as CN-x, where x represents the
into CN, its light-absorption range can be extended to 850 nm and en­ heating rate (◦ C/min). For fabricating samples with prolonged pyrolysis
ergy band gap could significantly narrowed from 2.7 eV to 2.0 eV. Ho­ time, melamine sulfate was heated to 600 ◦ C with a heating rate of
mogeneous self-modification of defects in CN could fully modify 1 ◦ C/min, at which it was held for different pyrolysis time (40–365 h).
electronic structure, whereas surface doping of heteroatom only caused The sample was labeled as CN-Ty, where y represents the pyrolysis time
partial localized state in band structure (Niu et al., 2014). (h). During whole thermal polymerization process, an argon flow was
In the past few years, various methods have been developed to introduced into the quartz tube.
construct defects in CN, including acid or alkali-assisted method, (Xue
et al., 2019) hydrothermal method, (Guo et al., 2013) thermal poly­ 2.2. Characterization
merization, (Cao et al., 2014; Papailias et al., 2015) vapor-assisted
method, (Mo et al., 2019) templating method, (Zhang et al., 2019) X-ray diffraction (XRD) analysis of CN was conducted by DX2700
reduction or oxidation, etc. In the primary stage, defect engineering for with Cu Kα radiation (λ = 1.5418 Å, 40 mA, 40 kV). The morphology
CN was focused on controlling polymerization temperature or time images of photocatalyst was obtained from transmission electron mi­
during the direct condensation of untreated precursor. However, the croscopy (TEM, Tecnai G2 F20 S-TWIN) and atomic force microscope
surface area was too low (12 m2/g) to show prospective in developing (AFM, Dimension Icon Scan Asyst). X-ray photoelectron spectroscopy
highly efficient photocatalysts. Recently, we have successfully fabri­ (XPS, Thermo electron, USA) and Fourier transform infrared (FT-IR,
cated CN nanosheet (CNS) with tunable defects and enhanced surface Perkin-Elmer spectrometer) were used to analyze the element state and
area (up to 190 m2/g) by utilizing the balance between the oxidation molecular structure, respectively. The Brunauer–Emmett–Teller (BET)
and re-polymerization during thermal exfoliation bulk CN. In compari­ surface area and pore size were recorded by using nitrogen adsorp­
son with bulk CN (1.2 mmol h− 1 g− 1), CNS exhibited a better H2 evo­ tion–desorption curve (Micromeritics). The unpaired electron was
lution rate of 2.0 mmol h− 1 g− 1 (Luo et al., 2020). Nitrogen defect observed by room-temperature electron paramagnetic resonance (EPR,
generally contained amino groups in imperfect triazine ring, nitrogen Bruker Emx Pro). The spectral response was tested by using UV–vis
vacancies, cyano groups. The -C˭N-C- of triazine structure is disrupted to diffuse reflectance spectra (DRS, UV3600, Shimadzu) with the reference
form imperfect structure (Hong et al., 2013; Medina et al., 2019). Acid of BaSO4. In addition, the separation and the lifetime of photogenerated
or alkali could be used to treat the precursor in order to accelerate the charge were measured by using the photoluminescence emission spectra
disruption of chemical bond. During polymerization process, molten (PL, Hitachi F-7000) and the time-resolved PL spectra, respectively.
KOH released OH- to react with amino group of the intermediates and The determination on photocurrent, electrochemical impedance
caused the generation of cyano groups in heptazine ring (Xue et al., spectroscopy (EIS) and Mott–Schottky curve (1000HZ) were processed
2019; Yu et al., 2017). Recently, photocarving strategy was used to by electrochemical workstation (CHI660E). The process of sample
control defect concentration and distribution, but it need two-step preparation for the electrochemical test is as follows: 40 mg sample
synthesis (Yang et al., 2019). Other strategies like vapor-assisted (Se) powder was dispersed into the mixed solution of 150 μL deionized water
route or colloidal crystal templating method are more complex. Not only and 80 μL isopropyl alcohol. Then 10 μL naphthol was added into the
the final products need Se vapor treatment or SiO2 demould, but also the formed suspension and mixed ultrasonically. 40 μL suspension was
toxicity and high cost of Se vapor were inconsistent with the concept of taken to form uniformly 1.0 × 1.0 cm film on the surface of FTO
"Green chemistry" (Mo et al., 2019; Zhang et al., 2019). Most strategies conductive glass. At last, the FTO glass was dried for 5 h at 70 ◦ C. The
introducing defects is based on two-step or multiple-step synthesis of photoelectrochemical properties were measured in a three electrodes
CN. It is difficult to control defect concentration and distribution by model, utilizing the prepared thin film as working electrode, Ag/AgCl
these methods (Schmidt-Mende et al., 2007). Considering CN is electrode as the reference electrode, Pt piece as the counter electrode,
nitride-based photocatalyst, the effect of nitrogen vacancies in adjusting and 0.1 M Na2SO4 as electrolyte solution.
optical and photocatalytic properties is still needed to be explored.
Therefore, it is still a long-term interest to develop a simple and 2.3. Photocatalytic test
controllable route for defects engineering in CN.
In our previous work, we found that the sublimation of melamine The whole photocatalytic reaction was carried out in a airtight and

2
Q. Yang et al. Journal of Hazardous Materials 410 (2021) 124594

constant-temperature glass photoreactor with gas circulation system. originated from the out-of-plane bending vibrations of tri-s-triazine
0.1 g photocatalyst and specified amounts of H2PtCl6 (3 wt% Pt co- units, whilst the obvious peaks located between 1000 and 1750 cm− 1
catalyst) were dispersed in 200 mL aqueous triethanolamine solution were consistent with stretching vibration of N-C˭N or N-(C)3 in the CN
(TEOA, 10 vol%). Prior to photocatalytic H2 test, Pt co-catalyst was heterorings of polymeric “melon” units (Wang et al., 2009a, 2009b). The
loaded on CN via in situ photoreduction H2PtCl6 by irradiation the so­ weak absorption peaks at 3000–3600 cm− 1, which are ascribed to the
lution 30 min with full-spectrum using a 300 W Xe lamp as light source. stretching vibration of –NH, indicated there are amino groups in these
Photocatalytic hydrogen evolution was proceeded under visible light samples.
irradiation (300 W Xe lamp equipped with optical cutoff filter, λ > 420 To further reveal the effects of heating rate and pyrolysis time on the
nm). The amount of photocatalytic hydrogen evolution was determined elemental composition and structural change of CN, the samples were
every one hour by gas chromatograph (GC-112AT). performed by XPS and organic element analysis (OEA). Both XPS survey
scan (Fig. S2a, and b) and OEA (Table S1) indicated no S element
3. Results and discussion remained in these samples. The experimental C/N value (ca. 0.66) is
lower than theoretical value of 0.75, indicating there are residual amino
The defective CN were prepared by one-step thermally polymeriza­ groups. Ninhyfrin test is a general method to detect amino groups in
tion of melamine sulfate according to the procedure shown in Scheme 1. protein. Previously work indicated amino group in imperfect triazine
We adopted two distinct routes to tune the defects in CN. One is structure can react with ninhyfrin to form "Ruhemann’s purple" (Wu
changing the heating rate, the other is prolonging the pyrolysis time at et al., 2016). We also adopted ninhyfrin test to determine amino group
600 ◦ C. in our samples. However, there is no color change. Additionally, we also
The XRD patterns of these samples prepared via both routes (Fig. 1) applied ninhyfrin test to determining melamine and no chromogenic
exhibited all characteristic peaks of g-C3N4. The diffraction peak at 13.0◦ reaction occurred. The result is consistent with the reported literature,
and 27.7◦ is assigned to the periodical in-plane structure with the hep­ which showed both melamine and cyanuric could not react with
tazine unit ((100) facet) and the interlayer-stacking of CN aromatic unit ninhydrin, even after hydrolysis (Field et al., 2010). Therefore, the
((002) facet), respectively (Lotsch et al., 2007; Liao et al., 2019). amino groups in our samples are suspended terminal amino located at
However, the peaks became more broaden and the intensity became perfect triazine ring. In addition, the ratio of H/C and O/N in CN-x were
weaker gradually with decreasing the heating rate. Furthermore, the gradually increased with increasing the heating rate (Table S1), which
diffraction peak of lateral (002) plane shifted to lower 2θ, slightly from manifested that more suspended terminal amino (-NHx) species were
27.7◦ to 27.1◦ (Fig. 1b), indicating that the inter-layer stacking distance produced and the samples were oxidized gradually during slowly ther­
was increased from 0.33 to 0.34 nm (calculated from Bragg’ law) or the mal polymerization process, respectively. Subtracted N in suspended
stacking of CN sheets along the direction of (002) became less denser. By terminal amino group (calculated according to the content of H), the
contrast, the peaks of CN samples prepared by prolonging the pyrolysis C/N molar ratio in CN-0.01 (3.00:3.57) is higher than that in CN-1
time at 600 ◦ C have negligible distinction (Fig. 1d), except CN-T365 (2.98:4.17), indicating some nitrogen atoms in carbon-nitrogen aro­
exhibited weaken peaks with lower diffraction angles. The difference matic ring were lost to form vacancies (Niu et al., 2014). To further
of XRD patterns between the samples prepared by the two distinct routes confirm the formation of nitrogen defects, N1s and C1s high resolution
may depend on how the defective CN form. For CN-x, due to interca­ XPS spectra were collected. The N1s XPS spectra can be deconvoluted
lation and oxidation of H2SO4 during polymerization process, when into three typical peaks at 399.01, 400.03, 401.55 eV, respectively
heated slowly, melamine sulfate decomposes slowly above 360 ◦ C and (Fig. 2a and b). These binding energy can be assigned to C-N-C(N2C),
releases H2SO4 which can intercalate into the interlayer space of CN and N-(C)3(N3C), C-N-H, respectively (Dementjev et al., 2000). With either
act as oxidant, leading to the slightly increased interlayer distance and reducing the heating rate or prolonging the pyrolysis time, the peak-area
the formation of disordered CN (Xu et al., 2013a, 2013b). In contrast, ratio of N2C/N3C was reduced from 2.02 to 1.75, or 2.14–1.79, respec­
when heated with higher rate, H2SO4 decomposes too rapidly to exert tively, revealing the loss of some nitrogen atoms located at N2C sites and
the intercalation and oxidation effect. For CN-Ty, as the hydrogen-bond the formation of N2C vacancies. In the C1s spectra, a major peak at
interaction and van der waals’ force was too weak to resist against the 288.05 eV originating from the N-C˭N (sp2-hybridized carbon) in the
continues thermal oxidation, the packing of CN aromatic conjugated heptazine ring and a small peak at 284.70 eV assigning to adventitious
system was loosened and the aromatic units were oxidized, vanished carbon could be observed (Fig. 2c, d). Furthermore, one peak at
gradually (Lotsch et al., 2007; Liao et al., 2019). Thus, the structure of 288.63 eV could be observed, although for samples prepared at higher
CN under long-time thermal exfoliation at 600 ◦ C, such as CN-T365, was heating rate, such as CN-1, CN-0.1, CN-0.07 and CN-T40, this peak was
transformed into more disordered phase. In addition, all samples very weak. Generally, there is only one carbon species (C-(N)3) located
showed the same FT-IR spectra (Fig. S1), indicating that the funda­ in the framework of defect-free CN (Wild et al., 2002). Therefore, the
mental structure of CN was remained. The typical peak at 813 cm− 1 was additional peak at higher binding energy of 288.63 eV should be related

Scheme 1. Schematic illustration of the procedure for preparing defective CN by thermally polymerization of melamine sulfate.

3
Q. Yang et al. Journal of Hazardous Materials 410 (2021) 124594

(a) (b)
CN-0.01 27.1° CN-0.01

CN-0.03 CN-0.03
27.3°
CN-0.05 CN-0.05

Intensity(a.u.)
Intensity(a.u.)

CN-0.07 CN-0.07

CN-0.1 CN-0.1
27.7°

(002)
(100)
CN-1 CN-1

10 20 30 40 50 60 20 25 30 35
2 Theta/degree
2 Theta/degree

(c) (d)
27.4°
CN-T365
CN-T365

CN-T125

Intensity(a.u.)
Intensity(a.u.)

CN-T125

CN-T75
CN-T75

CN-T65
CN-T65
27.7° CN-T40
CN-T40

10 20 30 40 50 60 70 80 20 25 30 35
2 Theta/degree 2 Theta/degree

Fig. 1. XRD patterns of CN-x (a) and CN-Ty (c). (b) and (d) shows the patterns between 20◦ and 35◦ in (a) and (c), respectively.

(a) (b)
N1s N1s
CN-1 CN-T365
CN-0.1
CN-T125
CN-0.07
CN-T75
CN-0.05
CN-T65
CN-0.03
CN-T40
CN-0.01
396 398 400 402 404
396 398 400 402 404

(c) (d)
C1s CN-1 C1s CN-T365
CN-0.1
CN-T125
CN-0.07
CN-T75
CN-0.05

CN-0.03 CN-T65

CN-0.01 CN-T40
282 284 286 288 290 292 282 284 286 288 290 292

Fig. 2. The XPS N1s spectra of CN-x(a) and CN-Ty(b), the XPS C1s spectra of CN-x(c) and CN-y(d).

to sp2-hybridized carbon bonding to amino groups (Wu et al., 2016). The The loss of nitrogen caused the formation of carbon radicals with
presence of O (Fig. S2c, d and Table S1) may be caused by the absorbed unpaired electrons in the structure (Barklie et al., 2001). The unpaired
H2O molecular, or the oxidation of terminal C atom (Zheng et al., 2014). electron can be detected to assess the degree of defects. Therefore, the

4
Q. Yang et al. Journal of Hazardous Materials 410 (2021) 124594

direct evidence on the presence of nitrogen defects was further provided samples was slightly changed from yellow to orange as either the
by the EPR spectra (Fig. 3a, b). The apparent symmetrical signal peak heating rate decreased or the pyrolysis time prolonged. The UV–vis DRS
located at a g value of 2.0038, which was consistent with the unpaired (Fig. 6a, b) revealed all the samples has absorption between 250 and
electron in the aromatic C atom, originating from the change of nitrogen 400 nm, which originates from π-π transition in the CN conjugated ar­
species or some disordered structure in CN, could be obviously observed omatic ring (Köhler et al., 2015). With either decreasing the heating rate
in the EPR spectra (Sun et al., 2017). The intensity of the EPR signal or prolonging the pyrolysis time, the absorption edge red-shift gradually
increased with either lowering heating rate or prolonging pyrolysis time from 430 to 550 nm nearly, illustrating that the sample with more de­
at 600 ◦ C, which indicated the degree of nitrogen defects were enhanced fects was more photo-responsive. Notably, a new absorption shoulder at
in the framework of CN. As mentioned above, H2SO4 could be released around 500 nm (2.48 eV) occurred, which is consistent with the darker
slowly, intercalate into CN and oxidize it under low heating rate, giving color of the samples (Scheme 1). The absorption shoulder should be
more chances to form nitrogen vacancies. This accords with the results derived from the defects generated during the process of either
of the above C1s and N1s spectra. In the case prolonging heating time at decreasing heating rate or prolonging pyrolysis time (Wu et al., 2014).
600 ◦ C, the defects was generated by pyrolysis CN, that is, the amount of Correspondingly, the optical band gap (Eg) calculated by Kubelka–Munk
nitrogen defects was closely associated with the prolonging pyrolysis function was narrowed from 2.86 to 2.55 eV for CN-x, and 2.87–2.53 eV
time. Therefore, the EPR intensity for CN-T365 was relatively stronger. for CN-Ty, respectively. The intrinsic Eg was narrowed as a consequence
During the long-time enough heating process, it is difficult for the of more nitrogen vacancies, due to the partial overlap of conduction
hydrogen bonding and intermolecular forces to struggle against the band, as the existence of a gradually increased amount of nitrogen de­
continuous thermal oxidation. Therefore, the stacking of heptazine units fects (Niu et al., 2012a, 2012b). That is, more defective sites generated
in the polymer structure of CN could be gradually loosened and in the structure of CN can optimize the photon-response ability and
decomposed, forming disordered phase, simultaneously more nitrogen capture more incident photon.
defects and terminal -NHx groups. The separation efficiency of photogenerated electron-holes can be
TEM and AFM images (Fig. S3) showed that all the samples exhibited determined by PL spectra. Generally, the higher emission intensities
irregular sheet-like structure with stacked pattern. It is hardly to observe suggested a increased probability of charge recombination. The PL
the specific lattice spacing in HRTEM (Fig. S3 g, h), indicating disor­ spectra of the samples (Fig. 7a, b) were collected under the excitation of
dered structure of CN. The AFM showed the thickness of these samples 279 nm. As displayed in Fig. 7a, b, there was a maximum PL peak
was mainly distributed between 3.8 and 4.5 nm (Fig. 4a-d and Fig. S4), centered at around 490 nm. As either the heating rate decreased or the
indicating the samples were equivalently composed of approximate 2–3 pyrolysis time prolonged, the intensity of fluorescence emission peak
layers of CN (the thickness of single layer of CN is ca. 1.6 nm). The showed the downward trend, which meant a greater restraint to charge
thickness of CN-x was increased slightly with decreasing the heating recombination in CN with more defects. Moreover, it can be seen intu­
rate, (Fig. 4k-n) which should be ascribed to the intercalating effect of itively that with the decrease of PL intensity, the peak position was
H2SO4. By contrast, the thickness of CN-Ty was gradually reduced due to slightly red-shifted to 500 nm, in agreement with the narrowing optical
long-time thermally exfoliation (Fig. S3i-p). The porosity and BET sur­ Eg. Due to the generation of nitrogen vacancies, some extra electrons
face area of CN-x and CN-Ty (Table S2) were obtained by the N2 remained on the missing sites have a tendency to capture holes, conse­
adsorption–desorption measurement. The N2 adsorption/desorption quently gave rise to the decrease of photogenerated electron-hole
isotherms were agreement with type Ⅳ (Fig. 5) (Kruk et al., 2001). The recombination.
BET surface area was changed from 47.0 to 26.5 m2/g with decreasing The experiment of visible-light-irradiated hydrogen evolution reac­
the heating rate (Table S2), which might be ascribed to structural tion was used to evaluate the photocatalytic activity of CN-x and CN-Ty.
decomposition under the oxidation of H2SO4 and the slightly increased Comparably, the samples with more defects exhibited relatively higher
thickness from 4.2 to 4.7 nm. On the contrary, with prolonging the hydrogen yield. As shown in Fig. 8a, the lowest H2 evolution rate
pyrolysis time, the BET surface area was slightly increased from 37.3 to (450 μmol g− 1 h− 1) was observed over CN-1. However, as the heating
46.2 m2/g, which can be interpreted by the sufficient thermally exfoli­ rate decreased, the photocatalytic activity presented the increasing
ation under the high-temperature Ar atmosphere. trend and the highest H2 yield rate of 850 μmol g− 1 h− 1 was observed for
The driving on photocatalytic reaction depended on many factors. CN-0.01, CN-0.05, and CN-0.03. The positive change of activity could be
Crucially, visible-light spectral response and photogenerated charge attributed to the enhanced light-absorption ability and improved charge
separation are indispensable for realizing photocatalytic water splitting. separation efficiency derived from N defects. Similarly, Fig. 8b indicated
The change in the visible-light-harvesting range and optical band that the photocatalytic activity of CN-Ty was enhanced from
structure of CN-x, CN-Ty was tested by UV–vis DRS. The color of the 456 μmol g− 1 h− 1 for CN-T40 to 906 μmol g− 1 h− 1 for CN-T125, as the

(a) (b) CN-1

CN-1
CN-0.1 CN-T40
CN-T65
CN-0.07 CN-T75
CN-0.05 CN-T125
CN-0.03 CN-T365
Intensity(a.u.)

Intensity(a.u.)

CN-0.01

2.03 2.02 2.01 2.00 1.99 1.98 2.03 2.02 2.01 2.00 1.99 1.98
g Value g Value

Fig. 3. The EPR spectra of CN-x(a) and CN-Ty(b) respectively.

5
Q. Yang et al. Journal of Hazardous Materials 410 (2021) 124594

Fig. 4. The AFM image of CN-1(a) and CN-0.01(b), and (c) and (d) shows the thickness curves along the line of CN-1 and CN-0.01, respectively.

(a) 250 (b) 200

CN-1
CN-T40
CN-0.1
Absorbed volume(cm /g,STP)

CN-T65
200 CN-0.07
Absorbed volume(cm /g,STP)

160 CN-T75
CN-0.05
CN-T125
CN-0.03
3

150 CN-T365
CN-0.01
3

120

100
80

50 CN-1
40

0
CN-0.01 0

0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
Relative pressure(P/P0) Relative pressure(P/P0)

Fig. 5. N2 adsorption–desorption isotherm of CN-x (a), and CN-Ty (b).

pyrolysis time prolonged. However, the activity of CN-T365 with the lower H2 evolution rate (Wu et al., 2014; Samsudin et al., 2015; Li et al.,
excessive defects was slightly decreased, which may be related with the 2014). Fig. 8c showed the yield of hydrogen was no obvious decline after
created defect energy level under the abundant N vacancies. Some 4 cycles, revealing the high photocatalytic stability of the tested sample
defect energy band may generated with much low reduction driving after introducing N defects.
force and could trap partial photogenerated electrons, resulting in the The charge transfer properties of CN-x and CN-Ty were further

6
Q. Yang et al. Journal of Hazardous Materials 410 (2021) 124594

(b)
(a) CN-0.01
CN-0.03
CN-0.05

1/2
(αhv)1/2
CN-0.07

(αhv)
CN-0.1
CN-1

Absorbance(a.u.)
Absorbance(a.u.)

2.55eV 2.86eV 2.53eV 2.87eV

2.0 2.5 3.0 3.5 4.0 2.0 2.5 3.0 3.5 4.0
hv(eV) CN-1 hv(eV)
CN-T40
CN-T65
CN-T75
CN-T125

200 300 400 500 600 700 800 200 300 400 500 600 700 800
Wavelength(nm) Wavelength(nm)

Fig. 6. The UV–vis DRS of CN-x (a) and CN-Ty (b). The inset shows the plots of (αhν)1/2 vs photon energy.

(a) (b)

CN-1 CN-1
CN-0.1 CN-T40

Intensity(a.u.)
CN-0.07 CN-T65
Intensity(a.u.)

CN-0.05 CN-T75
CN-0.03 CN-T125
CN-0.01

300 400 500 600 700 800 300 400 500 600 700 800
Wavelength(nm) Wavelength(nm)

Fig. 7. The PL spectra of CN-x (a) and CN-Ty (b).

(a) 1000
(b) 1000 (c) 3.5

3.0
The H 2 evolution rate(μmol/g/h)
The H 2 evolution rate(μmol/g/h)

800
The H 2 evolution rate(μmol/g/h)

800
2.5

2.0
600 600
1.5

1.0
400 400
0.5

200 0.0
200
CN-1 CN-0.1 CN-0.07 CN-0.05 CN-0.03 CN-0.01 CN-T1 CN-T2 CN-T3 CN-T4 CN-T5 0 4 8 12 16
Time(h)

Fig. 8. The photocatalytic H2 evolution rate of CN-x (a), and CN-Ty (b). (c) shows the time course of H2 evolution over 16 h for CN-0.05 under visible-light
irradiation.

studied by electrochemical impedance spectra (EIS) and photocurrent abnormal phenomenon was observed that electronic impedance was
test, combining with the chemical-physical properties. EIS Nyquist plot increased a little as well as the photocurrent response was a little un­
(Fig. 9a, b) showed the size of arc radius of CN-x, CN-Ty was gradually stable and decreased slightly, which is consistent with the decline of
reduced with either decreasing the heating rate or prolonging the py­ activity.
rolysis time, manifesting charge transfer was accelerated as the defects Another significant factor related with the photocatalytic H2 evolu­
increased. Besides, Fig. 9c, d reflected that the photocurrent response tion reaction over photocatalyst was its band structure. Therefore, the
was increased gradually with either decreasing the heating rate or impedance test was employed to obtain the position of the conduction
prolonging the pyrolysis time, confirming that superior photogenerated band (CB) and valence band (VB) of CN-x, CN-Ty. The impedance test
charge separation efficiency was exhibited as defects increased. These was regarded as an available tool to measure the flat band (Efb) poten­
variation of photoelectric property over defect-rich CN could positively tial. Fig. 10a and b presents the Mott-Schottky plot of CN-x and CN-Ty,
influence the photocatalytic activity. However, except for CN-T365, respectively. The positive slope of the curves reflected the n-type

7
Q. Yang et al. Journal of Hazardous Materials 410 (2021) 124594

(a) 2.0 (b) 5

CN-T40
CN-T65
1.6 4 CN-T75
CN-T125
CN-T365

1.2 3

Z''(KΩ)
Z''(KΩ)

0.8 CN-1 2
CN-0.1
CN-0.07
0.4 CN-0.05 1
CN-0.03
CN-0.01
0.0 0
0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
Z'(KΩ) Z'(KΩ)

(c) 6
CN-1 (d) CN-T40
CN-0.1 2.0 CN-T65
CN-0.07 CN-T75
Light on Light off CN-0.05 Light on Light off CN-T125
CN-0.03 1.6
Current(μA cm )

4 CN-T365
-2

Photocurrent(μA cm )
CN-0.01

-2
1.2

2
0.8

0.4
0
0.0

50 100 150 200

50 100 150 200
Time(s) Time(s)

Fig. 9. The EIS spectra of CN-x (a) and CN-Ty (b). (c), and (d) shows the photocurrent response under the UV–vis light irradiation of CN-x and CN-Ty, respectively.

semiconductor feature of these samples (Yang et al., 2017). The inter­ formula of average life time: τav = (τ21A1 + τ22A2)/(τ1A1 + τ2A2), A is the
cept of the curve on the x axis was recorded as the value of Efb, which relative contribution of the decayed charge. The τav of CN-0.01
was gradually decreased from − 0.45 eV to − 0.56 eV for CN-x (28.57 μs) was slightly longer than that of CN-1 (27.39 μs), which
(Fig. 10a). The slightly negative shift in the flat potential signified that accelerated the charge transfer more efficiently with decreasing the
the more active and unstable electron was retained in the CB of samples heating rate. For CN-Ty (Fig. 11b), with prolonging the pyrolysis time,
with more defects, thus drive water splitting with ease. The Efb (vs. NHE) the relative percentage of short-lived carriers (< 5 μs) was decreased
was calculated in accordance with the formula of Efb (vs. NHE) = Efb(vs. from 35.98% to 34.53% and that of long-lived carriers (> 30 μs) was
Ag/AgCl) + EAgCl + 0.059 × pH, where EAgCl and pH is around 0.197 V, increased from 64.02% to 65.47% (Table S3), which certainly provided
7.0 respectively (Huang et al., 2019). Thus, combining with the VB XPS more possibility for the charge carriers to participate in the redox re­
spectra and Efb (vs. NHE) (Fig. 10c), the position of VB was determined at action. The changed relative percentage of charge carriers and fluores­
2.18–2.36 eV for CN-x. By further combining with the Eg obtained from cence intensity were related with localized states of the band gap with
UV–vis DRS, the position of CB was determined at − 0.27 ~ − 0.67 eV. nitrogen defects. In accordance with the above results, it was proposed
The narrowed bandgap in CN-x could be ascribed to a decrease in CB that in the localized electronic states, the trapping sites were provided
position. Analogously, VB position and CB position of CN-Ty were ob­ for photogenerated charge carriers, causing the retard of charge
tained. The band structure of CN-x, CN-Ty was depicted in Fig. 10e. The recombination as a result of lowering electron-hole spatial overlap, thus
energy band of available photocatalysts need to be over than that could drive the photocatalytic reaction.
required to drive water splitting(1.23 eV) and VB, CB are positioned
appropriately. Generally, the more negative position of CB than the 4. Conclusion
reduction potential of H+/H2 (0 eV, vs NHE) and the more positive of VB
than the oxidation potential of O2/H2O (1.23 eV, vs NHE), could be In summary, nitrogen defects in CN was successfully tuned by
more probably to drive thermodynamically uphill photocatalytic water changing either heating rate or pyrolysis time during thermal poly­
splitting reaction. As shown in Fig. 10e, the position of CB was gradually merization of melamine sulfate. The structural defect was produced due
lower in CN samples with more defects, revealing that photo-induced to the intercalation and oxidation effect of H2SO4, and the structural
electron located in the VB was easily excited to the CB, leading to an order was broken under the continues thermal exfoliation. The N defects
optimal H2 yield rate. increase with decreasing the heating rate or prolonging the pyrolysis
Photocatalytic activity was closely associated with the life time of time. The formation of nitrogen defects efficiently prevented the fast
photogenerated charge. The time-resolved fluorescence decay spectra recombination of electrons and holes and thus extend visible-light
was utilized to detect decay time (τ1, τ2) of charge. The fluorescence response. As a result, an optimal H2 yield rate of 905 μmol g− 1 h− 1
decay curve (Fig. 11a) showed that τ1 and τ2 (Table S3) were increased (λ > 420 nm) was observed over defect-abundant CN. This simple and
from 3.60 and 29.20 μs for CN-1–4.97 and 30.63 μs for CN-0.01, controllable method to tune defects in CN is promising strategy for
respectively, as well as the relative percentage of the long-lived charge developing photocatalysts with high performance.
was increased from 66.54% to 69.77%. According to the calculated

8
Q. Yang et al. Journal of Hazardous Materials 410 (2021) 124594

(a) 2.0 (b) 2.0

CN-1 CN-T40
CN-0.1 CN-T65
1.6 CN-0.07 1.6 CN-T75
CN-0.05 CN-T125
CN-0.03

Cs (10 F cm )
-2 10 -2 4
Cs (10 F cm )

1.2 CN-0.01
4

1.2
-2
10

0.8
0.8
-2

0.4 CN-T125:-0.53eV
0.4

0.0 CN-0.03:-0.56eV CN-1:-0.45eV CN-T40:-0.4eV

0.0
-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4
Potential(V vs. Ag/AgCl) Potential(V vs. Ag/AgCl)

(c) (d)
CN-0.01 CN-T125
2.37eV
2.49eV
CN-0.03 CN-T75
2.35eV 2.62eV
Intensity(a.u.)

Intensity(a.u.)
CN-0.05
CN-T65
2.39eV
2.43eV
CN-0.07
CN-T40
2.49eV
2.21eV
CN-0.1
2.32eV CN-1
2.34eV
CN-1
2.34eV

20 18 16 14 12 10 8 6 4 2 0 20 18 16 14 12 10 8 6 4 2 0

Binding energy(eV) Banding energy(eV)

(e)
-2

-1 -0.67eV 0.79eV
-0.41eV -0.27eV -0.31eV -0.33eV
-0.28eV -0.26eV -0.19eV
Energy(eV, vs. NHE)

-0.14eV
+
0 H /H2

Eg 2.86eV 2.65eV 2.63eV 2.62eV 2.58eV 2.55eV 2.87eV 2.69eV 2.56eV 2.53eV
1
O2/H2O

2
2.08eV
2.19eV 2.24eV 2.24eV
2.35eV 2.36eV 2.18eV 2.36eV 2.42eV 2.35eV
3 CN-1 CN-0.1 CN-0.07 CN-0.05 CN-0.03 CN-0.01 CN-T40 CN-T65 CN-T75 CN-T125

Fig. 10. The Mott–Schottky curves of CN-x(a) and CN-Ty(b). The VB-XPS spectra of CN-x(c) and CN-Ty(d). The band structure alignments for all samples(e).

(a) CN-1
(b)
CN-T40
CN-0.1 CN-T65
CN-0.07 CN-T75
CN-0.05 CN-T125
CN-0.03
CN-0.01
Intensity(a.u.)

Intensity(a.u.)

0 50 100 150 200 0 50 100 150 200

Time/μs Time(μs)

Fig. 11. Time-resolved fluorescence lifetime spectra of CN-x(a), CN-Ty(b).

9
Q. Yang et al. Journal of Hazardous Materials 410 (2021) 124594

CRediT authorship contribution statement Hong, Y.Z., Liu, E.L., Shi, J.Y., Lin, X., Sheng, L.Z., Zhang, M., Wang, L.Y., Chen, J.H.,
2019. A direct one-step synthesis of ultrathin g-C3N4 nanosheets from thiourea for
boosting solar photocatalytic H2 evolution. Int J. Hydrog. Energy 44, 7194–7204.
Qing Yang: Conceptualization, Methodology, Writing-Original Draft Hong, Z.H., Shen, B., Chen, Y.L., Lin, B.Z., Gao, Bifen, 2013. Enhancement of
preparation. Wenbin Yang: Experimental design, and Funding Acqui­ photocatalytic H2 evolution over nitrogen-deficient graphitic carbon nitride.
sition. Fangfang He: Software, Formal Analysis. Kewei Liu: Validation. J. Mater. Chem. A 1, 11754–11761.
Huang, J.N., Cao, Y.H., Wang, H.J., Yu, H., Peng, F., Zou, H.B., Liu, Z.L., 2019. Revealing
Hongmei Cao: Validation. Hongjian Yan: Experimental design, Su­ active-site structure of porous nitrogen-defected carbon nitride for highly effective
pervision, Funding Acquisition, Resources, Project Administration. photocatalytic hydrogen evolution. Chem. Eng. J. 373, 687–699.
Jiang, L.B., Yuan, X.Z., Zeng, G.M., Chen, X.H., Wu, Z.B., Liang, J., Zhang, J., Wang, H.,
Wang, H., 2017. Phosphorus- and sulfur-codoped g-C3N4: facile preparation,
Declaration of Competing Interest mechanism insight, and application as efficient photocatalyst for tetracycline and
methyl orange degradation under visible light irradiation. ACS Sustain. Chem. Eng.
5, 5831–5841.
The authors declare that they have no known competing financial Jun, Y.S., Lee, E.Z., Wang, X.C., Hong, W.H., Stucky, G.D., Thomas, A., 2013. From
interests or personal relationships that could have appeared to influence melamine-cyanuric acid supramolecular aggregates to carbon nitride hollow
spheres. Adv. Funct. Mater. 23, 3661–3667.
the work reported in this paper.
Köhler, et al., 2015. Electronic Processes in Organic Semiconductors. Wiley-VCH,
Germany.
Acknowledgments Kruk, M., Jaroniec, M., 2001. Gas adsorption characterization of ordered organic-
inorganic nanocomposite materials. Chem. Mater. 13, 3169–3183.
Lee, H.L., Sofer, Z., Mazánek, V., Luxa, J., Chua, C.K., Pumera, M., 2017. Graphitic
We are grateful for financial support from the National Natural Sci­ carbon nitride: effects of various precursors on the structural, morphological and
ence Foundation of China (22072100), NSAF of China (No. U1730114), electrochemical sensing properties. Appl. Mater. Today 8, 150–162.
Li, J.H., Shen, B., Hong, Z.H., Li, B.Z., Gao, B.F., Chen, Y.L., 2012. A facile approach to
the Open Project of State Key Laboratory of Environment-friendly En­ synthesize novel oxygen-doped g-C3N4 with superior visible-light photoreactivity.
ergy Materials (No. 18kfhg09), and the Fundamental Research Funds for Chem. Commun. 48, 12017–12019.
the Central Universities (No. 20826041D4117). We also thank Sichuan Li, S., Zhang, L.J., Jiang, T.F., Chen, L.P., Lin, Y.H., Wang, D.J., Xie, T.F., 2014.
Construction of shallow surface states through light Ni doping for high-efficiency
University Analytical & Testing Centre for the analysis of UV-Vis, PL, FT-
photocatalytic hydrogen production of CdS nanocrystals. Chem. Eur. J. 20, 311–316.
IR measurements. Li, T.T., Zhao, L.H., He, Y.M., Cai, J., Luo, M.F., Lin, J.J., 2013b. Synthesis of g-C3N4/
SmVO4 composite photocatalyst with improved visible light photocatalytic activities
in RhB degradation. Appl. Catal. B Environ. 129, 255–263.
Appendix A. Supporting information Li, Y.B., Zhang, H.M., Liu, P.R., Wang, D., Li, Y., Zhao, H.J., 2013a. Cross-linked g-C3N4/
rGO nanocomposites with tunable band structure and enhanced visible light
Supplementary data associated with this article can be found in the photocatalytic activity. Small 9, 3336–3344.
Li, Z., Kong, C., Lu, G.X., 2016. Visible photocatalytic water splitting and photocatalytic
online version at doi:10.1016/[Link].2020.124594.
two-electron oxygen formation over Cu- and Fe-Doped g-C3N4. J. Phys. Chem. C 120,
56–63.
References Liang, Q.H., Li, Z., Yu, X.L., Huang, Z.H., Kang, F.Y., Yang, Q.H., 2015. Macroscopic 3D
porous graphitic carbon nitride monolith for enhanced photocatalytic hydrogen
evolution. Adv. Mater. 27, 4634–4639.
Bai, X.J., Sun, C.P., Wu, S.L., Zhu, Y.F., 2015. Enhancement of photocatalytic
Liao, J.F., Wu, W.Q., Zhong, J.X., Jiang, Y., Wang, L.Z., Kuang, D.B., 2019. Enhanced
performance via a P3HT-g-C3N4 heterojunction. J. Mater. Chem. A 3, 2741–2747.
efficacy of defect passivation and charge extraction for efficient perovskite
Barklie, R.C., 2001. Characterisation of defects in amorphous carbon by electron
photovoltaics with a small open circuit voltage loss. J. Mater. Chem. A 7,
paramagnetic resonance. Diam. Relat. Mater. 10, 174–181.
9025–9033.
Barrio, J., Lin, L.H., Wang, X.C., Shalom, M., 2018. Design of a unique energy-band
Liu, J., Liu, Y., Liu, N.Y., Han, Y.Z., Zhang, X., Huang, H., Lifshitz, Y., Lee, S.T., Zhong, J.,
structure and morphology in a carbon nitride photocatalyst for improved charge
Kang, Z.H., 2015. Metal-free efficient photocatalyst for stable visible water splitting
separation and hydrogen production. ACS Sustain. Chem. Eng. 6, 519–530.
via a two-electron pathway. Science 347, 970–974.
Cao, S.W., Low, J.X., Yu, J.G., Jaroniec, M., 2015. Polymeric photocatalysts based on
Lotsch, B.V., Döblinger, M., Sehnert, J., Seyfarth, L., Senker, J., Oeckler, O.,
graphitic carbon nitride. Adv. Mater. 27, 2150–2176.
Schnick, W., 2007. Unmasking melon by a complementary approach employing
Cao, Y.Q., Zhang, Z.Z., Long, J.L., Liang, J., Lin, H., Lin, H.X., Wang, X.X., 2014. Vacuum
electron diffraction, solid-state NMR spectroscopy, and theoretical calculations-
heat-treatment of carbon nitride for enhancing photocatalytic hydrogen evolution.
structural characterization of a carbon nitride polymer. Chem. Eur. J. 13,
Mater. Chem. A 2, 17797–17807.
4968–4969.
Chen, X.B., Liu, L., Yu, P.Y., Mao, S.S., 2011. Increasing solar absorption for
Luo, M.L., Gong, H.H., Yang, W.B., He, F.F., Cao, Y., Zhang, Y.F., Zhang, Y., Liu, K.,
photocatalysis with black hydrogenated titanium dioxide nanocrystals. Science 331,
Cao, H.M., Yan, H.J., 2019a. Composite of g-C3N4 and poly(3-hexylthiophene)
746–749.
prepared by polymerizing thiophene-3-acetic acid treated g-C3N4 and 3-hexylthio­
Chen, Y.F., Huang, W.X., He, D.L., Situ, Y., Huang, H., 2014. Construction of
phene for enhanced photocatalytic hydrogen production. Int. J. Hydrog. Energy 44,
heterostructured g-C3N4/Ag/TiO2 microspheres with enhanced photocatalysis
7108–7117.
performance under visible-light irradiation. ACS Appl. Mater. Interfaces 6,
Luo, M.L., Yang, Q., Liu, K.W., Cao, H.M., Yan, H.J., 2019b. Boosting photocatalytic H2
14405–14414.
evolution on g-C3N4 by modifying covalent organic frameworks (COFs). Chem.
Dementjev, A.P., de Graaf, A., van de Sanden, M.C.M., Maslakov, K.I., Naumkin, A.V.,
Commun. 55, 5829–5832.
Serov, A.A., 2000. X-Ray photoelectron spectroscopy reference data for
Luo, M.L., Yang, Q., Yang, W.B., Wang, J.H., He, F.F., Liu, K.W., Cao, H.M., Yan, H.J.,
identification of the C3N4 phase in carbon-nitrogen films. Diam. Relat. Mater. 9,
2020. Defects engineering leads to enhanced photocatalytic H2 evolution on
1904–1907.
graphitic carbon nitride-covalent organic framework nanosheet composite. Small 16,
Dong, H., Guo, X.T., Yang, C., Ouyang, Z.Z., 2018. Synthesis of g-C3N4 by different
2001100.
precursors under burning explosion effect and its photocatalytic degradation for
Mamba, G., Mishra, A.K., 2016. Graphitic carbon nitride (g-C3N4) nanocomposites: a new
tylosin. Appl. Catal. B Environ. 230, 65–76.
and exciting generation of visible light driven photocatalysts for environmental
Field, A., Field, J., 2010. Melamine and cyanuric acid do not interfere with bradford and
pollution remediation. Appl. Catal. B Environ. 198, 347–377.
ninhydrin assays for protein determination. Food Chem. 121, 912–917.
Medina, J.O., Wang, Z.P., Silva, R.C., Gomez, A.M., Wang, F., Yao, X.D., Terrones, M.,
Fu, J.W., Yu, J.G., Jiang, C.J., Cheng, B., 2018. g-C3N4-Based heterostructured
Endo, M., 2019. Defect engineering and surface functionalization of nanocarbons for
photocatalysts. Adv. Energy Mater. 8, 1701503.
metal-free catalysis. Adv. Mater. 31, 1805717.
Guo, H.L., Su, P., Kang, X.F., Ning, S.K., 2013. Synthesis and characterization of nitrogen-
Mo, R., Li, J., Tang, Y.H., Li, H.X., Zhong, J.X., 2019. Introduction of nitrogen defects
doped graphene hydrogels by hydrothermal route with urea as reducing-doping
into a graphitic carbon nitride framework by selenium vapor treatment for enhanced
agents. J. Mater. Chem. A 1, 2248–2255.
photocatalytic hydrogen production. Appl. Surf. Sci. 476, 552–559.
Guo, Y.F., Li, J., Yuan, Y.P., Li, L., Zhang, M.Y., Zhou, C.Y., Lin, Z.Q., 2016. A rapid
Niu, P., Liu, G., Cheng, H.M., 2012a. Nitrogen vacancy-promoted photocatalytic activity
microwave-assisted thermolysis route to highly crystalline carbon nitrides for
of graphitic carbon nitride. J. Phys. Chem. C 116, 11013–11018.
efficient hydrogen generation. Angew. Chem. Int. Ed. 55, 14693–14697.
Niu, P., Zhang, L.L., Liu, G., Cheng, H.M., 2012b. Graphene-like carbon nitride
Han, Q., Wang, B., Gao, J., Cheng, Z.H., Zhao, Y., Zhang, Z.P., Qu, L.T., 2016. Atomically
nanosheets for improved photocatalytic activities. Adv. Funct. Mater. 22,
thin mesoporous nanomesh of graphitic C3N4 for high-efficiency photocatalytic
4763–4770.
hydrogen evolution. ACS Nano 10, 2745–2751.
Niu, P., Yin, L.C., Yang, Y.Q., Liu, G., Cheng, H.M., 2014. Increasing the visible light
Hao, X.Q., Zhou, J., Cui, Z.W., Wang, Y.C., Wang, Y., Zou, Z.G., 2018. Zn-vacancy
absorption of graphitic carbon nitride (melon) photocatalysts by homogeneous self-
mediated electron-hole separation in ZnS/g-C3N4 heterojunction for efficient visible-
modification with nitrogen vacancies. Adv. Mater. 26, 8046–8052.
light photocatalytic hydrogen production. Appl. Catal. B Environ. 229, 41–51.
Niu, P., Qiao, M., Li, Y.F., Huang, L., Zhai, T.Y., 2018. Distinctive defects engineering in
Hong, J.D., Xia, X.Y., Wang, Y.S., Xu, R., 2012. Mesoporous carbon nitride with in situ
graphitic carbon nitride for greatly extended visible light photocatalytic hydrogen
sulfur doping for enhanced photocatalytic hydrogen evolution from water under
evolution. Appl. Catal. B Environ. 44, 73–81.
visible light. J. Mater. Chem. 22, 15006–15012.

10
Q. Yang et al. Journal of Hazardous Materials 410 (2021) 124594

Ou, H.H., Lin, L.H., Zheng, Y., Yang, P.J., Fang, Y.X., Wang, X.C., 2017. Tri-s-triazine- Xu, M., Han, L., Dong, S.J., 2013a. Facile fabrication of highly efficient g-C3N4/Ag2O
based crystalline carbon nitride nanosheets for an improved hydrogen evolution. heterostructured photocatalysts with enhanced visible-light photocatalytic activity.
Adv. Mater. 29, 1700008. ACS Appl. Mater. Interfaces 23, 12533–12540.
Papailias, I., Giannakopoulou, T., Todorova, N., Demotikali, D., Vaimakis, T., Trapalis, Xu, J., Zhang, L.W., Shi, R., Zhu, Y.F., 2013b. Chemical exfoliation of graphitic carbon
2015. Effect of processing temperature on structure and photocatalytic properties of nitride for efficient heterogeneous photocatalysis. J. Mater. Chem. A 1,
g-C3N4. Appl. Surf. Sci. 358, 278–286. 14766–14772.
Ran, J.R., Ma, T.Y., Gao, G.P., Du, X.W., Qiao, S.Z., 2015. Porous P-doped graphitic Xue, Y.J., Guo, Y.C., Liang, Z.Q., Cui, H.Z., Tian, J., 2019. Porous g-C3N4 with nitrogen
carbon nitride nanosheets for synergistically enhanced visible-light photocatalytic defects and cyano groups for excellent photocatalytic nitrogen fixation without co-
H2 production. Energy Environ. Sci. 8, 3708–3717. catalysts. J. Colloid Interface Sci. 556, 206–213.
Samsudin, E.M., Hamid, S.B.A., Juan, J.C., Basirun, W.J., Kandjani, A.E., 2015. Surface Xu, S.M., Yan, H.J., Chen, Y., 2012. Synthesis of graphitic carbon nitride by directly
modification of mixed-phase hydrogenated TiO2 and corresponding photocatalytic heating sulfuric acid treated melamine for enhanced photocatalytic H2 production
response. Appl. Surf. Sci. 359, 883–896. from water under visible light. Int. J. Hydrog. Energy 37, 125–133.
Schmidt-Mende, L., MacManus-Driscoll, J.L., 2007. ZnO nanostructures, defects, and Yan, S.C., Li, Z.S., Zou, Z.G., 2009. Photodegradation performance of g-C3N4 fabricated
devices. Mater. Today 10, 40–48. by directly heating melamine. Langmuir 25, 10397–10401.
Sun, N., Liang, Y., Ma, X.J., Chen, F., 2017. Reduced oxygenated g-C3N4 with abundant Yang, P., Ou, H., Fang, Y., Wang, X.C., 2017. A facile steam reforming strategy to
nitrogen vacancies for visible-light photocatalytic applications. Chem. Eur. J. 23, delaminate layered carbon nitride semiconductors for photoredox catalysis. Angew.
15466–15473. Chem. Int. Ed. 129, 4050–4054.
Tay, Q.L., Kanhere, P., Ng, C.F., Chen, S., Chakraborty, S., Huan, A.C.H., Sum, T.C., Yang, P.J., Wang, L., Zhang, H.Y., Wang, R.R., Titirici, M.M., Wang, X.C., 2019.
Ahuja, R., Chen, Z., 2015. Defect engineered g-C3N4 for efficient visible light Photocarving nitrogen vacancies in a polymeric carbon nitride for metal-free oxygen
photocatalytic hydrogen production. Chem. Mater. 27, 4930–4933. synthesis. Appl. Catal. B Environ. 256, 117794.
Thomas, A., Fischer, A., Goettmann, F., Antonietti, M., Muller, J.O., Schloglb, R., Ye, L.Q., Liu, J.Y., Jiang, Z., Peng, T.Y., Zan, L., 2013. Facets coupling of BiOBr-g-C3N4
Carlsson, J.M., 2008. Graphitic carbon nitride materials: variation of structure and composite photocatalyst for enhanced visible-light-driven photocatalytic activity.
morphology and their use as metal-free catalysts. J. Mater. Chem. 18, 4893–4908. Appl. Catal. B Environ. 142, 1–7.
Tonda, S., Kumar, S., Kandula, S., Shanker, V., 2014. Fe-doped and -mediated graphitic Yu, H.J., Shang, L., Bian, T., Shi, R., Waterhouse, G.I.N., Zhao, Y.F., Zhou, C., Wu, L.Z.,
carbon nitride nanosheets for enhanced photocatalytic performance under natural Tung, C.H., Zhang, T.R., 2016. Nitrogen-doped porous carbon nanosheets templated
sunlight. J. Mater. Chem. A 2, 6772–6780. from g-C3N4 as metal-free electrocatalysts for efficient oxygen reduction reaction.
Wang, Li, Q., Liu, B.S., Cheng, B., Ho, W.K., Yu, J.G., 2015. Sulfur-doped g-C3N4 with Adv. Mater. 28, 5080–5086.
enhanced photocatalytic CO2-reduction performance. Appl. Catal. B Environ. 176, Yu, H.J., Shi, R., Zhao, Y.X., Bian, T., Zhao, Y.F., Zhou, C., Waterhouse, G.I.N., Wu, L.Z.,
44–52. Tung, C.H., Zhang, T.R., 2017. Alkali-assisted synthesis of nitrogen deficient
Wang, X.C., Chen, X.F., Thomas, A., Fu, X.Z., Antonietti, M., 2009a. Metal-containing graphitic carbon nitride with tunable band structures for efficient visible-light-
carbon nitride compounds: a new functional organic-metal hybrid material. Adv. driven hydrogen evolution. Adv. Mater. 29, 1605148.
Mater. 21, 1609–1612. Zhang, J.H., Chen, Y., Wang, X.C., 2015. Two-dimensional covalent carbon nitride
Wang, X.C., Maeda, K., Thomas, A., Takanabe, K., Xin, G., Carlsson, J.M., Domen, K., nanosheets: synthesis, functionalization, and applications. Energy Environ. Sci. 8,
Antonietti, M., 2009b. A metal-free polymeric photocatalyst for hydrogen 3092–3108.
production from water under visible light. Nat. Mater. 8, 76–80. Zhang, J.S., Chen, X.F., Takanabe, K., Maeda, K., Domen, K., Epping, J.D., Fu, X.Z.,
Wild, E., Raymundo-Piñero, D., Cazorla-Amorósa, A., Linares-Solano, J., Find, U., Antonietti, M., Wang, X.C., 2010. Synthesis of a carbon nitride structure for visible-
Wild, R., Schlögl, 2002. Structural characterization of N-containing activated carbon light catalysis by copolymerization. Angew. Chem. Int. Ed. 49, 441–444.
fibers prepared from a low softening point petroleum pitch and a melamine resin. Zhang, S., Hu, C., Jia, H.H., Zhanga, L.L., Li, F., 2019. Facile synthesis of nitrogen-
Carbon 40, 597–608. deficient mesoporous graphitic carbon nitride for highly efficient photocatalytic
Wu, P., Wang, J.R., Zhao, J., Guo, L.J., Osterloh, F.E., 2014. Structure defects in g-C3N4 performance. Appl. Surf. Sci. 478, 304–312.
limit visible light driven hydrogen evolution and photovoltage. J. Mater. Chem. A 2, Zhang, Y.W., Liu, J.H., Wu, G., Chen, W., 2012. Porous graphitic carbon nitride
20338–20344. synthesized via direct polymerization of urea for efficient sunlight-driven
Wu, W.T., Zhang, J.Q., Fan, W.Y., Li, Z.T., Wang, L.Z., Li, X.M., Wang, Y., Wang, R.Q., photocatalytic hydrogen production. Nanoscale 4, 5300–5303.
Zhang, J.T., Wu, M.B., Zeng, H.B., 2016. Remedying defects in carbon nitride to Zheng, Y., Lin, L.H., Ye, X.J., Guo, F.S., Wang, X.C., 2014. Helical graphitic carbon
improve both photooxidation and H2 generation efficiencies. ACS Catal. 6, nitrides with photocatalytic and optical activities. Angew. Chem. Int. Ed. 53,
3365–3371. 11926–11930.

11