Switching charge transfer process of carbon nitride and bismuth vanadate by anchoring silver nanoparticle toward cocatalyst free water reduction

Journal of Colloid and Interface Science 529 (2018) 375–384

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science journal homepage: www.elsevier.com/locate/jcis

Switching charge transfer process of carbon nitride and bismuth vanadate by anchoring silver nanoparticle toward cocatalyst free water reduction Meiting Song, Yuhang Wu, Xiaojing Wang, Mengqing Liu, Yiguo Su ⇑ College of Chemistry and Chemical Engineering, Inner Mongolia University, Hohhot, Inner Mongolia 010021, PR China

g r a p h i c a l a b s t r a c t A ternary catalyst was constructed which exhibit superior H2 production rate with co-catalyst free under visible light irradiation.

a r t i c l e

i n f o

Article history: Received 24 April 2018 Revised 13 June 2018 Accepted 14 June 2018 Available online 18 June 2018 Keywords: g-C3N4/Ag/BiVO4 Cocatalyst free Water splitting Z-scheme photocatalysis

⇑ Corresponding author. E-mail address: [email protected] (Y. Su). https://doi.org/10.1016/j.jcis.2018.06.029 0021-9797/Ó 2018 Elsevier Inc. All rights reserved.

a b s t r a c t With the aim of exploring and modulating the interfacial charge kinetics, a ternary g-C3N4/Ag/BiVO4 was constructed with excellent photocatalytic performance and preferable stability toward H2 evolution in absence of cocatalyst. Both density functional theory (DFT) and experimental results implied that the type II g-C3N4/BiVO4 composite can be switched to Z-scheme via Ag nanoparticles as the electron shuttle. The optimal photocatalytic H2 yield rate achieved for g-C3N4/Ag/BiVO4 was 57.4 mmol
g 1
h 1, being far surpassed the H2 harvest rate of g-C3N4/BiVO4, Ag/g-C3N4 and g-C3N4, which is 2.9, 14.8 and 1.7 mmol
g 1
h 1, respectively. The apparent quantum efficiency of g-C3N4/Ag/BiVO4 photocatalyst was also determined to be 1.23%. Besides, the photocatalytic performance of g-C3N4/Ag/BiVO4 well preserved over 5 runs in 50 h. The improved H2 production performance is considered as the consequence of promoted segregation of photoexcited charge carriers and SPR effects of Ag nanoparticles. In combination with photocurrent measurement, examination of active species and DFT calculation, it is found that Ag nanoparticles as an electron mediator can highly promote the Z-scheme carrier migration that electrons come from conduction band of BiVO4 will quickly assemble with the photo-induced holes from valence band of g-C3N4, leaving electrons in the conduction band of g-C3N4 and holes in valence band of BiVO4 that could greatly enhance the charge separation efficiency. Ó 2018 Elsevier Inc. All rights reserved.

376

M. Song et al. / Journal of Colloid and Interface Science 529 (2018) 375–384

1. Introduction Semiconductor photocatalysis for hydrogen production provides an important environmental benign strategy toward sustainable energy conversion [1–5]. To meet the flexible industrial processing ability, the development of efficient photocatalytic system with high solar energy conversion efficiency is demanded. However, the bottlenecks of photocatalysts, such as low visible light absorption and poor quantum efficiency, are still of hindrance for practical applications, in spite of great efforts have been dedicated, including band gap engineering, surface modification and semiconductor heterojunction [6–8]. Recently, heterostructured materials by combining two semiconductors with band alignment are found effective photocatalysts to improve catalytic activity. However, traditional heterojunctions meet weakened reduction and oxidation ability of photoinduced charge carriers, which fall short of the requirement of high redox potential of H+/H2 for hydrogen production [9]. The Z-scheme photocatalytic systems possessing excellent redox driving force and charge separation efficiency have garnered much research attention for enhanced photocatalytic performance [10,11]. Using a redox shuttle or electron mediator, a Z-scheme charge transfer can be modulated to achieve spatial isolation of photoinduced charge carriers for more positive valence band potential and more negative conduction band potential [12]. Hence, it’s highly desirable to modulate the interfacial charge kinetics to develop Z-scheme photocatalysts for improving photocatalytic activity. Among widely investigated oxides, nitrides, and oxyhalides, graphite-like carbon nitrides carries a more negative conduction band potential ( 1.1 V vs. NHE) with a strong redox capability, which has been widely applied for photocatalytic hydrogen production [13–16]. However, it still has some drawbacks, like the high rate of recombination for photoexcited charge carriers and strong reliance on co-catalysts. Building g-C3N4 based heterostructure can provide an effective way for overcoming these disadvantages. A great deal of g-C3N4 based Z-scheme heterojunctions, including g-C3N4/TiO2, g-C3N4/SnS2, g-C3N4/Ag3PO4, and so forth were developed [17–20]. For instance, the nanojunction of g-C3N4/TiO2 resulted in the constitution of a settled electric field that can gather photoinduced electrons and holes in diverse face region and prolong the longevity about charge carriers with strong reductive and oxidative ability [10]. Since the band gap construction of g-C3N4 matches well with that of BiVO4, the junction of g-C3N4 and BiVO4 may predict spatial charge separation and improved photocatalytic reactivity. However, it is still controversial that the charge migration between g-C3N4 and BiVO4 demonstrates type II or Z-scheme transfer process [21–25]. Moreover, binary component heterojunctions still have drawbacks of slower charge transfer rate and inefficient charge separation ratio. To solve these problems, semiconductor-metal-semiconductor (S-M-S) ternary component systems have been developed. As a model S-M-S system, CdS-Au-TiO2 possessed a fast vertical electron transfer of CdS ? Au ? TiO2, resulting in high photocatalytic activity toward methylviologen (MV2+) reduction, which improved by a factor of 1.6, 1.8 and 2.3 times in comparison to Au/TiO2, CdS/TiO2 and bare TiO2 [9,26]. Usually, noble metallic nanoparticles were adopted as the electron shuttle to construct S-M-S system for promotion of photoreactivity. Having these in mind, the construction of ternary g-C3N4-M-BiVO4 system predicts efficient charge separation and the subsequent improved photocatalytic performance. Though some studies of g-C3N4-M-BiVO4 photocatalytic systems have been reported about degradating organic pollutants and oxidizing NO [8,12], it is still highly necessary to explore the interfacial charge kinetics and its application in water reduction. In this work, using g-C3N4/Ag/BiVO4 ternary photocatalytic system as model, we found that the type II g-C3N4/BiVO4 composite

can be switched to Z-scheme via Ag nanoparticles as electron shuttle. The coupling effects of g-C3N4 and BiVO4 nanoparticles with Ag modification were systematically investigated in attempt to shed light on the interfacial charge kinetics and possible reasons for significant enhancement of photocatalytic performance toward hydrogen evolution. 2. Experiment 2.1. Sample synthesis Synthesis of Ag/g-C3N4 photocatalyst: All reagents used were of analytical grade purity none of re-distillation. Water used was ultrapure water. The synthetic procedure for Ag/g-C3N4 catalyst can be briefly depicted as follows: g-C3N4 was obtained by calcining melamine. In a typical synthesis, 2.38 g melamine (CP, Sinopharm Group chemical reagent Co. Ltd) was ground for 30 min and placed in a semiclosed alumina crucible with a lid and heated at a speed of 5 °C/min to 550 °C for 4 h, cooling and regrinding to obtain yellow powder of g-C3N4, and the g-C3N4 sample was used directly without any further washing treatment. For the second step, different quality (1.00 g, 0.98 g, 0.96 g, 0.94 g, 0.92 g, 0.90 g) of g-C3N4 was blended with the different quality (0 g, 0.03 g, 0.06 g, 0.09 g, 0.12 g, 0.15 g) of AgNO3 (AR, Wind ship in tianjin chemical reagent Co. Ltd), and ground for 30 min, then kept at 550 °C for 4 h. Finally, they were allowed to cool to ambient temperature and reground to reach different Ag/g-C3N4 composite photocatalyst (Ag mass ratios 0%, 2%, 4%, 6%, 8%, 10%). They were marked as g-C3N4, AC1, AC2, AC3, AC4 and AC5, respectively. Synthesis of BiVO4 photocatalyst: 5 mmol Bi(NO3)3
5H2O (AR, Wind ship in tianjin chemical reagent Co. Ltd), was dissolved in 20 mL nitric acid (2 M) to obtain solution A. Meanwhile, 5 mmol of NH4VO3 (AR, Adamas Reagent Co. Ltd) was dissolved in 10 mL NaOH solution (2 M) to acquire solution B and was injected drop by drop into solution A under intense stirring to achieve a steady suspension. The suspension liquid was regulated to adjust pH value of 7 using 1 M NaOH solution with energetic stirring for 30 min. Then it was poured into 100 mL Teflon-lined stainless steel autoclaves and reacted at 180 °C for 12 h. The achieved outcome was washed with ultra-pure grade water for several times and maintained at 60 °C to obtain BiVO4. Synthesis of g-C3N4/Ag/BiVO4 photocatalyst: Different quality of Ag/g-C3N4 (with Ag mass ratio of 8%) was mixed with the different quality of the BiVO4 to synthesize the ternary g-C3N4/Ag/BiVO4 heterojunction (with mass ratios of 9:1, 7:3, 5:5, 3:7, 1:9). The compound was calcined at 400 °C at a speed of 5 °C /min for 4 h to acquire different ratio compound semiconductor photocatalytic materials of g-C3N4/Ag/BiVO4. They were marked as ACB1, ACB2, ACB3, ACB4 and ACB5, respectively. For the purpose of comparison, the g-C3N4/BiVO4 (with mass ratio of 9:1) photocatalyst was prepared and marked as CB1. 2.2. Sample characteristic analysis Phase purity of all catalysts was measured by X-ray diffraction (XRD) using Panalytical X-ray diffractometer with a copper target at 40 kV and 40 mA. The diffraction patterns were kept track in the scope of 2h = 5–80° at a scan speed of 1° min 1. Particle dimension and morphologies of the catalyst were tested using transmission electron microscopy (TEM) on a FEI TECNAI F20 S-Twin with an acceleration voltage of 200 kV. The Fourier transform infrared (FT-IR) spectra of all catalysts were taken on a Perkin–Elmer IR spectrophotometer at a resolution of 4 cm 1 using the KBr pellet skill. The UV–Vis absorption spectra were recorded on a UVIKON XL/XS diffuse reflectance spectrophotometer in the wavelength

M. Song et al. / Journal of Colloid and Interface Science 529 (2018) 375–384

scope of 190–800 nm. The reference was BaSO4 power. X-ray photoelectron spectroscopy (XPS) analyses were characterized by an ESCALab220i-XL with a homochromous Al Ka and charge corrector. The binding energies of all the elements were referenced to C1s (284.6 eV). XPS measurements were carried out on a Thermo ESCALAB 250 with Al Ka (1486.6 eV) line at 150 W. Electron Paramagnetic Resonance spectra on superoxide radicals and hydroxyl radicals were tested in dark or visible light (sample, 4 mg; DMPO, 0.22 M) methyl alcohol volume, 2.0 mL and aqueous solution volume, respectively. All calculations were performed using CASTEP code based on density functional theory. The Perdew Burke Ernzerh (PBE) parametrization of the generalized gradient approximation (GGA) was adopted for the exchange and correlation functional. Plane wave functions are used as basis sets. The electron wave function was expanded in plane waves up to a cutoff energy of 420 eV in case of ultrasoft potential. The geometries for all the compositions are optimized using a conjugated gradient technique in a direct minimization of the Khon-Sham energy functional. The lattice parameters, as well as the atomic positions relaxations were carried out until the forces and total energy converged within 0.015 eV/Å and 10 5 eV, respectively. The 3  1  1 unit supercell of g-C3N4, Ag, BiVO4 has been built. 2.3. Photoelectrochemical performance The electrochemical measurement was conducted by an Autolab model AUT302N.-FRA32M.V three-electrode cell electrochemical workstation, which includes working electrode, the counter electrode (Pt wire) and reference electrode (Ag/AgCl). 0.1 g photocatalyst was confected into turbid liquid with conducting binder (0.5 mL C2H5OH and 30lL Nafion solution from Aladdin Industrial Corporation). The suspension was ultrasonic-assisting for 30 min and vigorous stirring for 12 h. Then, the suspension was spined and coated on glass electrode (FTO glass 1 cm  1 cm, the area of the glass electrode was controlled 1 cm  1 cm through the hollow out of the insulating tape, and the area of the hollow place was 1  1 cm) and dried at 60 °C for 30 min for the working electrode. The working electrode loading capacity of sample is 5.0 mg
cm 2. Every part of the cell was filled with the same stock of electrolyte solution with an original concentration of 0.2 M of Na2SO4. The electrolyte was purged with high purity N2 for 10 min prior to use. N2 gas was gently flowed on top of the electrolyte interface during measurements. The illumination source was a LED lamp (LDCNW-White (Neutral), 4100 K, 700 Ma, 690 lm). 2.4. Photocatalytic experiments Photocatalytic water splitting was performed in a quartz with cover. For H2 evolution, 0.1 g catalyst was diffused in a mixed solution of 20 mL methanol and 80 mL ultrapure water in the flask to form a mixture. Before light, the reaction system was cleansed with Ar gas and withdrawn many times by a vacuum pump to guarantee the reactor in a vacuum condition. Then the mixed liquor was stirred and irradiated with a 300 W xenon lamp (k  420 nm) for lack of another cocatalyst. The H2 evolution speed was recorded by an online gas chromatograph (GC2014C, TCD), Ar was used as the carrier. The apparent quantum efficiency (AQE) was measured over g-C3N4/Ag/BiVO4 nanostructure under monochromatic light irradiation provided by a 300 W Xe lamp with the band-pass filter of 420 nm. At k = 420 nm, the average incident light intensity was determined to be 30 mW
cm 2 by an irradiatometer (CELNP2000, Beijing CEAULIGHT), and the irradiation area was 24.6 cm2. After irradiation of 10 h, the amount of H2 generation of g-C3N4/Ag/BiVO4 was 574.33 lmol.

377

3. Results and discussion The interfacial charge transfer between two semiconductors highly depends on the electronic band potentials and Fermi levels. When a semiconductor contacts with another one, electrons will flow between the surfaces because of the Fermi level difference in the two semiconductors [27]. To clearly clarify the origin of the interfacial charge isolation between g-C3N4 and BiVO4, the work functions were calculated for g-C3N4 (0 0 1) and BiVO4 (0 0 1) surfaces, as illustrated in Fig. 1a and b. By aligning the Fermi level to vacuum energy level, the work functions for g-C3N4 (0 0 1) plane and BiVO4 (0 0 1) plane were counted to be 4.025 eV and 7.161 eV, respectively, which is close to previous reported results [28]. The adsorption of surface species, like methanol, water, and triethanolamine, can up/down shift the Fermi levels, which has great impact on the interfacial charge transfer type. Thereby, we calculated the work functions of BiVO4 (0 0 1) and g-C3N4 (0 0 1) surfaces to confirm how methanol adsorption affects the Fermi levels of BiVO4 and g-C3N4. As Fig. 1c and d, by adsorbing methanol on BiVO4 and g-C3N4 surfaces, the Fermi levels of both BiVO4 and g-C3N4 upshifted to 6.018 eV and 3.565 eV, respectively. Besides Fermi level, the bandgap and energy levels are also important factors affecting the charge transfer process. Fig. 1e illustrates the optical bandgap performances for g-C3N4 and BiVO4. Since BiVO4 and g-C3N4 display an indirect optical transition [25,29], the optical bandgap for BiVO4 and g-C3N4 was counted to be 2.45 eV and 2.54 eV, respectively. The conduction band edge location of a semiconductor can be estimated according to Mulliken electronegativity theory (Fig. S1). The conduction band edge positions of g-C3N4 and BiVO4 were counted to be 1.04 V and 0.33 V, respectively. And therefore, the valence band edge potentials of g-C3N4 and BiVO4 were estimated to be 1.50 V and 2.78 V, as shown in Fig. 1f. In the light of the relation between EAVL (AVL means absolute vacuum level) and ENHE (NHE means normal hydrogen electrode), the Fermi levels of g-C3N4 and BiVO4 were estimated to be 0.475 V/ 0.935 V and 2.661 V/1.518 V before/ after methanol adsorption, respectively. Since the Fermi level of g-C3N4 is higher than that of BiVO4, the electrons will flow from g-C3N4 to BiVO4 [30]. Then, BiVO4 is negatively charged and g-C3N4 is positively charged, which can establish an internal electronic field directed from g-C3N4 to BiVO4 when an equalized Fermi level acquires. Consequently, the photo stimulated electrons will shift from the conduction band of g-C3N4 to the conduction band of BiVO4, while photogenerated holes will flow from the valence band of BiVO4 to the valence band of g-C3N4 in the presence of visible light. The photogenerated charges in g-C3N4/BiVO4 composite display type II charge transfer process, which can spatially separate the charge carriers but is not favorable for hydrogen production due to weakening redox potential of the composite. So as to switch the charge transfer of g-C3N4/BiVO4 from type II to Z-scheme, the S-M-S system was established in attempt to attempt to shed light on the interfacial charge kinetics and possible reasons for significant enhancement of photocatalytic performance toward hydrogen production. Fig. 2a shows the XRD patterns of the as-obtained photocatalysts. Two diffraction peaks of g-C3N4 appeared. The strongest diffraction peak is readily indexed to (0 0 2) plane of graphitic carbon nitride, which is a typical feature of sandwich heaping peak of aromatic types. The weak peak at small angle corresponds to the (1 0 0) face of g-C3N4[31]. The main diffraction peaks had no changes with the improvement of the Ag loaded content, but the diffraction peaks of Ag were observed at high Ag content (Fig. S2). It is showed that the featured peaks of tetragonal g-C3N4 weakened in the Ag/g-C3N4 samples, suggesting the crystallinity of g-C3N4 was prohibited by incorporating into Ag. As for BiVO4, it is seen that whole characteristic peaks can be well

378

M. Song et al. / Journal of Colloid and Interface Science 529 (2018) 375–384

Fig. 1. Electrostatic potential of g-C3N4 (0 0 1) face (a) and BiVO4 (0 0 1) face (b). Electrostatic potential of g-C3N4 (0 0 1) face (c) and BiVO4 (0 0 1) face (d) adsorbed with methanol in vacuum. Optical bandgap of g-C3N4 and BiVO4 (e). Bandgap alignment diagram of g-C3N4 and BiVO4 (f).

Fig. 2. XRD patterns of the as-obtained samples (a). TEM (b) and HRTEM images (c, d) over g-C3N4/Ag/BiVO4 composite.

M. Song et al. / Journal of Colloid and Interface Science 529 (2018) 375–384

indexed into a signal phase of monoclinic BiVO4 (JCPDS Card No. 14-0688). The narrow and sharp diffraction peaks indicate fine crystalline nature and large particle size of BiVO4. When BiVO4 coupled with g-C3N4, the main typical diffraction peaks of g-C3N4/BiVO4 altered inconspicuously. Moreover, the addition of Ag nanoparticles also led to no apparent changes of the diffraction peaks. A weak diffraction peak at 2h = 27.3° belonging to g-C3N4 and two diffraction peaks of Ag nanoparticles could be found (Fig. S3), confirming the coexistence of g-C3N4 and Ag in g-C3N4/ Ag/BiVO4. The interfacial contact feature of g-C3N4/Ag/BiVO4 was also affirmed by TEM observations. Fig. 2b reveals the TEM graph of g-C3N4/Ag/BiVO4. As illustrated in Fig. 2b, three types of materials were observed in g-C3N4/Ag/BiVO4, which corresponded to g-C3N4, Ag and BiVO4, respectively. This result indicates that g-C3N4 may act as a support to bind Ag and BiVO4 nanoparticles for fine interfacial contact in g-C3N4/Ag/BiVO4. High resolution TEM (HRTEM) images suggest the crystalline nature of Ag and BiVO4 nanoparticles. The space between adjacent lattice fringes

379

was 0.467 nm, which is identical to that of (0 1 1) plane for BiVO4 (Fig. 2c). Another lattice flat face space was 0.236 nm, which is consistent with 0.237 nm of (1 1 1) plane of Ag. Qualitative analysis of all elements of the as-obtained Ag/g-C3N4/BiVO4 sample was performed by energy dispersive X-ray energy spectrum (EDS) (Fig. S4) and the results of EDS were consistent with the results from the XRD. The EDS confirmed the main elemental components are Ag, Bi, V, C, N and O. And the atomic percentages of Ag, Bi, V, C, N and O in g-C3N4/Ag/BiVO4 from the energy-dispersive X-ray spectra of SEM are 5.22, 6.39, 1.55, 33.84, 51.03 and 1.97 (Table S1), which are close to that of theoretical value. In order to clarify the surface constituent, the oxidation states and chemical surrounding of g-C3N4/Ag/BiVO4 composite, XPS analysis was carried (Fig. 3). Fig. 3a illustrates the whole scanning XPS spectrum of g-C3N4/Ag/BiVO4 (ACB1 sample) photocatalyst. The main elements are Ag, Bi, V, C, N and O, which agree with that of EDS data (Fig. S4). Fig. 3b–f make clear that the HR-XPS spectra of the g-C3N4/Ag/BiVO4 sample for Bi, O, Ag, C and N, respectively.

Fig. 3. Whole scanning XPS spectrum over the g-C3N4/Ag/BiVO4 photocatalyst (a). High-resolution regional XPS spectra of Bi 4f (b), O 1s (c), Ag 3d (d), C 1s (e) and N 1s (f).

380

M. Song et al. / Journal of Colloid and Interface Science 529 (2018) 375–384

As shown in Fig. 3b, the binding energies of Bi 4f7/2 and Bi 4f5/2 are 158.6 and 163.9 eV, respectively, indicating that the chemical state of Bi is Bi (III) [32]. The O1s spectrum (Fig. 3c) can be well- regenerative by three Lorentzian-Gaussian lines locating at 529.2, 531.9 and 530.5 eV, respectively. At the same time, the peak of 529.2 eV could be attributable to Bi-O in (Bi2O2)2+ of BiVO4 layered structure. The symmetrical peaks at 531.9 and 530.5 eV are corresponding to carboxylic oxygen and hydroxyl oxygen [33], respectively. Fig. 3d shows the XPS spectra of Ag0, and the binding energy of Ag0 (3d3/2 and 3d5/2 orbitals) locates at 374.0 and 368.0 eV, respectively. The binding energy in the XPS spectra (Fig. 3e) was the characteristic of C 1s at 284.60 eV. The peak located at 288 eV is attributed to the sp2 hybridization of (C-(N)3) [34]. The broadened peak (397–402 eV) in Fig. 3f shows that the N 1s peak located at 398.5 eV is ascribed to the sp2 of (C@NAC) [35]. The binding energies located at 400.1 eV and 401 eV are in consistent with the reported results of g-C3N4[36]. All of these outcomes presented the insight that the g-C3N4/Ag/BiVO4 catalyst was successfully prepared and the surface chemistry analysis in the local structure of the as-prepared samples can be further identified by FT-IR spectra (Fig. S5). Generally, the electronic structure and the band edge levels have great impact on the interfacial charge migration as well as photocatalytic properties of the heterojunction materials. The optical absorption feature of the as-obtained photocatalysts was explored by UV–visible diffuse reflectance spectra (DRS), as shown in Fig. 4a. It’s found that pristine g-C3N4 exhibited an absorption edge at 460 nm and the absorption edge of pure BiVO4 is about 525 nm. By modification with noble Ag nanoparticles, Ag/g-C3N4 composite photocatalyst exhibited a forceful and broad absorption band tail in the visible light region in comparison to pristine g-C3N4. It was obvious for the absorption was gradually enhanced with the increase of Ag nanoparticles (Fig. S6). The absorption edge

of g-C3N4/BiVO4 moved to low energy levels. The junction of Ag/g-C3N4 and BiVO4 caused a red shift of the absorption band edge and distinctly enhanced band tail absorption in the region of 600–800 nm (Fig. S7). Besides visible light absorption capability, the electronic band potentials of g-C3N4 and BiVO4 were also tested by Mott-Schottky plots. As is seen in Fig. 4b, typical Mott-Schottky curves of g-C3N4 and BiVO4 in dark condition reveal n-type characteristic semiconductor feature because of positive slope of the Mott-Schottky plots [37]. Moreover, the charge carrier density of g-C3N4 and BiVO4 are of equal order of magnitude from MottSchottky equation, which expects fine interfacial charge carrier between g-C3N4 and BiVO4. The flat-band potential was also obtained from the Mott-Schottky data. The flat-band potentials were determined to be 1.70 V and 0.33 V versus Ag/AgCl electrode for g-C3N4 and BiVO4, respectively. Association with the relationship of redox potential between Ag/AgCl electrode and normal hydrogen electrode (NHE), the flat-band potential of g-C3N4 and BiVO4 is estimated to be 1.38 V and 0.02 V versus NHE. As well depicted in previous literatures, the flat-band potential is verged on the conduction band of semiconductors [38]. Thereby, the conduction band edge of g-C3N4 and BiVO4 was estimated to be 1.38 V and 0.02 V versus NHE, which is close to the Mulliken electronegativity theory prediction and reported results [39]. Consequently, the valence band of g-C3N4 and BiVO4 was determined to be 1.16 V and 2.43 V on the basis of the band gap energy (Fig. 1e). The band structure match between g-C3N4 and BiVO4 predicts efficient charge carrier separation and promoted photocatalytic hydrogen production performance. As shown in Fig. 4c, all as-prepared photocatalysts exhibit photoreactivity toward hydrogen production under visible light irradiation. Photocatalytic hydrogen yield for the as-prepared photocatalysts was analyzed by gas chromatography in initial 10 h irradiation. With improving visible light exposure time, the hydrogen evolution amounts steadily

Fig. 4. UV–vis diffuse reflectance spectra of the as-obtained samples (a). Mott-Schottky plots over g-C3N4 and BiVO4 (b). Photocatalytic property of the as-prepared photocatalysts: Time course of hydrogen evolution at the absence of cocatalysts under vis-light (k  420 nm) irradiation (c). The cycling experiment of the as-prepared g-C3N4/Ag/BiVO4.

M. Song et al. / Journal of Colloid and Interface Science 529 (2018) 375–384

increased. In the light of electronic structure analyses, the junction of g-C3N4 with BiVO4 shows type II interfacial charge transfer with hydrogen evolution rate of 2.9 mmol
g 1
h 1, which is close to original g-C3N4 (1.7 mmol
g 1
h 1). Based on literature results [40], Ag nanoparticles on g-C3N4 surfaces can efficient gather the photoexcited electrons from the conduction band of g-C3N4 and thus promote charge carrier separation, which displayed optimal hydrogen production rate of 14.8 mmol
g 1
h 1 (Fig. S8). Moreover, using Ag nanoparticles as ‘‘electron linker”, the optimal photocatalytic hydrogen production rate reached to 57.4 mmol
g 1
h 1 under k  420 nm light illumination (Fig. S9), being far surpassed the hydrogen production by g-C3N4/BiVO4, Ag/g-C3N4 and pristine g-C3N4, respectively. Besides photocatalytic activity, the stability is also a key factor for a catalyst. To access the stability of g-C3N4/Ag/BiVO4, the reaction system was evacuated several times for recycling tests. As illustrated in Fig. 4d, the photocatalytic properties of g-C3N4/Ag/ BiVO4 after five runs retained 97.7%. Meanwhile, XRD patterns of g-C3N4/Ag/BiVO4 before and after photocatalytic reaction was also tested to confirm the stability of the photocatalyst. As shown in Fig. S10, the strong diffraction peaks were well matched with the standard data JCPDS file (No. 14-0688) of the BiVO4 (Fig. S10) and no apparent changes occurred after five runs photocatalytic test although the crystallinity decreased compared with before the reaction. The diffraction peaks intensity of Ag and the g-C3N4 before and after the reaction were almost the same. It indicated that the g-C3N4/Ag/BiVO4 maintained good crystallinity of the initial catalyst. All the results suggested g-C3N4/Ag/BiVO4 photocatalyst is highly stable. The apparent quantum efficiency (AQE) of g-C3N4/Ag/BiVO4 photocatalyst was also determined to be 1.23% (Fig. S11), which is higher than any other specimens, such as g-C3N4, Ag/ g-C3N4 and g-C3N4/ BiVO4. In order to explore the photocatalytic potential of the catalyst g-C3N4/Ag/BiVO4, the photocatalytic oxygen evolution performance

381

was also explored under visible light irradiation (k  420 nm) without any sacrificial agents (Fig. S12). Excitingly, the catalysts can produce oxygen without sacrificial agents and the rate of photocatalytic oxygen production reached to 27.36 mmol
g 1
h 1. This catalyst not only produces hydrogen but also produces oxygen with or without sacrificial agent. Since g-C3N4/Ag/BiVO4 photocatalyst possesses optimal photocatalytic hydrogen production activity, it is quite necessary to clarify the underlying mechanism for the highly enhanced photocatalytic property of the ternary g-C3N4/Ag/BiVO4. Photoelectrochemical analysis was tested in order to prove the favorable junction effects of g-C3N4, Ag and BiVO4 for light collection and migration of photo-induced electron-hole pairs under white light LED irradiation. In Fig. 5a, the photocurrent response of the asprepared samples to white light LED irradiation is rapid. Apparently, g-C3N4/Ag/BiVO4 sample exhibited stronger photocurrent intensity, which is about 7.23, 3.95, 2.17, and 1.70 times than that of the g-C3N4, Ag/g-C3N4, g-C3N4/BiVO4, and BiVO4, respectively. This result is further testified by the photocurrent responses under single-wavelength excitation (Fig. 5b). It is seen that original g-C3N4 exhibited photocurrent response below 450 nm, which agree with the band gap energy of g-C3N4. BiVO4 showed an obvious photocurrent response below 500 nm, with a weak one up to 520 nm. For g-C3N4/BiVO4, an obvious enhanced photocurrent response appeared below at 420 nm. As for Ag/g-C3N4, evident photocurrent signal appeared at 475 nm, while the weak photocurrent signal extended to 550 nm (Fig. S13). In comparison to g-C3N4, the additional photocurrent signal in Ag/g-C3N4 is likely to originate to Ag nanoparticle, which could be ascribed to the SPR effect of Ag [41]. As expected, the g-C3N4/Ag/BiVO4 sample displayed strongest photocurrent signal up to 550 nm, which threshold wavelength is close to that of Ag/g-C3N4. Based on the above photocurrent response results, it’s expected that the junction of g-C3N4 and BiVO4 can accelerate the separation of the photoinduced

Fig. 5. Comparison of transient photocurrent responses under white LED light irradiation. (LED LDCNW-White (Neutral), 4100 K, 700 Ma, 690 lm, [Na2SO4] = 0.2 M) (a), transient photocurrent responses at different single-wavelength excitation (b), EIS Nyquist plots (c) and photoluminescence spectra (d) of the photocatalysts.

382

M. Song et al. / Journal of Colloid and Interface Science 529 (2018) 375–384

charge carriers. Moreover, further addition of Ag nanoparticles to form ternary g-C3N4/Ag/BiVO4 composite can drastically influence the interfacial charge carrier transfer behavior. The improved charge carrier separation and transfer of g-C3N4/Ag/BiVO4 can also be certified by the electrochemical impedance spectroscopy (EIS) analysis. As shown in Fig. 5c, the Nyquist plot data were in accordance with an equivalent circuit model [42,43]. Rct for the g-C3N4, Ag/g-C3N4, BiVO4, g-C3N4/BiVO4 and g-C3N4/Ag/BiVO4 samples are 152, 150, 136, 140 and 40.3 KX, respectively. The smallest Rct over g-C3N4/Ag/BiVO4 composite manifested the increase over the electrical conductivity of the heterojunction. The EIS and photocurrent density information confirmed the effective charge separation and the enhancement of electrical conductivity and the instantaneous charge transfer speed of the g-C3N4/Ag/BiVO4 photocatalyst, which possibly boost the photocatalytic property. The improved interfacial charge separation was further demonstrated by photoluminescence spectra of the photocatalysts. Basically, BiVO4 displays very weak luminescent property due to low V-V separation and efficient energy transfer from V-O species to quenching centers [44], as illustrated in Fig. 5d. However, g-C3N4 shows strong visible light emission within the scope of 400–700 nm. The intensity variation of the visible light emission for g-C3N4 can qualitatively predict the separation behavior of the photo-stimulated charge carriers. The luminescence intensity of g-C3N4/BiVO4 is lower than that of pure g-C3N4. Besides, the emission intensity of g-C3N4/Ag/BiVO4 sample was significantly suppressed, which is compatible to Ag/ g-C3N4. Accordingly, for g-C3N4/Ag/BiVO4 and Ag/g-C3N4, the recombination of photoexcited charge carriers can be greatly prevented. Undoubtedly, the interfacial charge movement leads to the efficient charge separation, leading us to predict that Ag nanopartilces are likely play important roles in modulating the interfacial charge transfer process as well as visible light responsive behavior.

In order to get further information of the impacts of Ag nanoparticles on the interfacial charge separation process, photodegradation of rhodamine B (Rh B) was taken as a typical reaction to testify the primary active species using xenon lamp of 500 W in the photochemical reactions reactor (XPA-7). The content of the remnant RhB in the aqueous solution was tested by a Varian UV–vis spectrophotometer (UVIKON XL/XS). During Rh B photodegradation process, certain types of radical species scavengers were carefully added. In brief, Rh B photodegradation process was modified by adding superoxide radical scavenger (1,4benzoquinone (BQ)), hole scavenger (ammonium oxalate (AO)), hydroxyl radical trapping agent (tert-butyl alcohol (TBA)). The effects of scavengers [45] on Rh B degradation over g-C3N4/BiVO4 and g-C3N4/Ag/BiVO4 with visible light illumination are exhibited in Fig. 6a and b. In Fig. 6a, Rh B degradation process was clearly prohibited in present of BQ, AO and TBA. The result show that superoxide active species, direct hole oxidation and hydroxyl radical species were involved in the Rh B degradation process. Since g-C3N4/BiVO4 shows type II interfacial charge process, the above results seem to be contrary to the theoretical prediction since the conduction band edge potential of BiVO4 is about 0.33 V versus NHE and the valence band edge potential of g-C3N4 is 1.50 V versus NHE, which are unable to generate superoxide active species and hydroxyl radical species. A possible explanation, that is, the interfacial charge movement between g-C3N4 and BiVO4 is inefficient, which can be verified by photocatalytic property of g-C3N4/BiVO4 to hydrogen evolution and Rh B degradation (Fig. S14). For g-C3N4/Ag/BiVO4, it’s noticed that the photocatalytic activity toward Rh B degradation was greatly inhibited when BQ was added. Meanwhile, photodegradation process was also decreased by adding AO and TBA scavengers, suggesting photogenerated holes and hydroxyl radical species also taken part in Rh B degradation process. The above observation gives evidence that superoxide

Fig. 6. Active species capture experiment on Rh B degradation through the as-obtained g-C3N4/BiVO4 (a) and g-C3N4/Ag/BiVO4 (b). O2 radical active species text by EPR in vis-light over g-C3N4/BiVO4 and g-C3N4/Ag/BiVO4 (c). OH radical active species text by EPR in vis-light over g-C3N4/Ag/BiVO4 (d).

M. Song et al. / Journal of Colloid and Interface Science 529 (2018) 375–384

active species played predominantly act in degrading Rh B for gC3N4/Ag/BiVO4 sample. Thereby, it’s supposed that the capability to generate superoxide active species in g-C3N4/Ag/BiVO4 is much stronger than that in g-C3N4/BiVO4, which can be further proved through electron paramagnetic resonance (EPR) measurement. The superoxide radical species was determined by EPR measurement using 5,5-Dimethyl-1-pyrroline-N-oxide (DMPO) as capture agent. With variation of irradiation time, characteristic EPR signals belonging to DMPO-O2 were detected for g-C3N4/BiVO4 and g-C3N4/Ag/BiVO4. The signal intensity was gradually increased (Figs. S15 and S16) [46]. This observation predicts that superoxide radical species can be driven in both g-C3N4/BiVO4 and g-C3N4/Ag/ BiVO4. Most importantly, upon irradiation for 12 min, the intensity of DMPO-O2 signal for g-C3N4/Ag/BiVO4 is more distinct than that for g-C3N4/BiVO4. As discussed above, the conduction band potential of g-C3N4 locates at about 1.04 V versus NHE, which is more negative than that of O2/ O2 (-0.33 V versus NHE) [41]. Therefore, the production of superoxide active species is potential for adsorption oxygen by taking photo stimulated electrons in the conduction band of g-C3N4. Briefly, stronger capability to generate superoxide active species implies higher reductive force and more photogenerated electrons in the conduction band of g-C3N4. The accumulation of electrons in the conduction band of g-C3N4 in g-C3N4/Ag/BiVO4 may demonstrate an alternative interfacial charge transfer approach in comparison to g-C3N4/BiVO4. To achieve photogenerated electron accumulation in the conduction band of g-C3N4, the interfacial charge migration over g-C3N4/Ag/BiVO4 was supposed to behoove Z-scheme charge transfer process by servicing Ag as electron shuttle. Having all the above information in mind, a more reasonable explanation for the highly improved photocatalytic property toward hydrogen evolution of g-C3N4/Ag/BiVO4 heterojunction photocatalysts can be illustrated in Fig. 7. With visible light, incident photons could be absorbed by g-C3N4 and BiVO4 and photoinduced electrons are excited in the conduction band of g-C3N4 and BiVO4, respectively. As the Schottky barrier was formed when the metal contacted to n-type semiconductor, the band energy at the interface was bent. In the case of metal contact with different semiconductors, band bending was different (Fig. S17). When the metal work function WB is larger than the semiconductor work function Wn (Fig. S17a), the band energy of the semiconductor is bent upwards to form a Schottky barrier, and the surface barrier needed to be overcome when electrons were transferred from the metal to the semiconductor. When the metal work function WB is smaller than the semiconductor work function Wn (Fig. S17b), the band energy of the semiconductor is bent downwards to form a Schottky barrier anti-blocking layer, which was favorable for the transfer of electrons from the semiconductor to the metal. By aligning the Fermi level to vacuum energy level, the work functions for Ag (0 0 1) plane were counted to be 4.227 eV (Fig. S18). The band energy of the BiVO4 was bent downwards

383

and the band energy of the g-C3N4 was bent upwards. So sectional Ag nanoparticles located at the interfacial position could serve as electron mediator to contact g-C3N4 and BiVO4, leading to the result that photo-induced electrons from the conduction band of BiVO4 could readily integrate with the photo-induced holes from the valence band of g-C3N4 using Ag electron shuttle, leaving photogenerated electrons in the conduction band of g-C3N4 and photogenerated holes in the valence band of BiVO4. This Z-scheme charge transfer process could greatly enhance the charge separation efficiency as same as reduction and oxidation driving energy. On the other hand, Ag nanoparticles also absorb incident photons to produce electrons and holes in accordance with SPR effect. And the photogenerated electrons in Ag nanoparticles can infused into the conduction band of g-C3N4 to advance photocatalytic property and extend the visible light absorption region. 4. Conclusions A ternary g-C3N4/Ag/BiVO4 was constructed with the aim of exploring and modulating the interfacial charge kinetics which exhibit superior photocatalytic activities and favorable stability toward H2 evolution in the absence of cocatalyst under vis-light irradiation. Compared with other literatures [47–52], the mechanism was explained by density functional theory and experiments for the first time. Mechanism study implied that the type II g-C3N4/ BiVO4 composite can be switched to Z-scheme via Ag nanoparticles as electron shuttle. The optimal photocatalytic activity toward hydrogen production of g-C3N4/Ag/BiVO4 was far surpassed the H2 production rate over g-C3N4/BiVO4, Ag/g-C3N4 and g-C3N4. The promoted photocatalytic H2 production property is considered as the consequence of enhanced separation of charge carriers and SPR effects of Ag nanoparticles. In combination with photocurrent measurement, examination of active species and density functional theory calculation, it is found that Ag nanoparticles as an electron mediator can highly promoted the Z-scheme charge carrier migration that electrons from the conduction band of BiVO4 can efficiently assemble with the photo-induced holes from the valence band of g-C3N4, leaving photogenerated electrons in the conduction band of g-C3N4 and photogenerated holes in the valence band of BiVO4 that could greatly enhance the charge separation efficiency as same as reduction and oxidation driving force. This work provides theoretical and experimental guidance for the construction of ternary Z-scheme catalysts in the future, and it also applies to the use of other metals as electron shuttles. Acknowledgement This work is financially supported by the National Natural Science Foundation of China (Grants 21563021), the Inner Mongolia Natural Science Foundation (grant no. 2016JQ01) and Program for Young Talents of Science and Technology in Universities of Inner Mongolia Autonomous Region (NJYT-18-A01). Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.jcis.2018.06.029. References

Fig. 7. Scheme for the energy band levels and the underling mechanism for g-C3N4/Ag/BiVO4.

[1] Y. Zhou, Y. Zhang, M. Lin, J. Long, Z. Zhang, H. Lin, J.C.S. Wu, X. Wang, Monolayered Bi2WO6 nanosheets mimicking heterojunction interface with open surfaces for photocatalysis, Nat. Commun. 6 (2015) 8340–8348. [2] H. Sudrajat, S. Hartuti, Structural properties and catalytic activity of a novel ternary CuO/gC3N4/Bi2O3 photocatalyst, J. Colloid Interf. Sci. 524 (2018) 227– 235.

384

M. Song et al. / Journal of Colloid and Interface Science 529 (2018) 375–384

[3] J. Guo, K. Wang, X. Wang, Photocatalytic reduction of CO2 with H2O vapor under visible light over Ce doped ZnFe2O4, Catal. Sci. Technol. 7 (2017) 6013– 6025. [4] M. Karimi-Nazarabad, E.K. Goharshadi, Highly efficient photocatalytic and photoelectrocatalytic activity of solar light driven WO3/g-C3N4 nanocomposite, Sol. Energy Mater. Sol. Cells 160 (2017) 484–493. [5] G. Yang, H. Yin, W. Liu, Y. Yang, Q. Zou, L. Luo, H. Li, Y. Huo, H. Li, Synergistic Ag/TiO2-N photocatalytic system and its enhanced antibacterial activity towards Acinetobacter baumannii, Appl. Catal. B: Environ. 224 (2018) 175– 182. [6] B. Modak, S.K. Ghosh, An efficient strategy for controlled band gap engineering of KTaO3, J. Phys. Chem. C 120 (2016) 6920–6929. [7] M. Liu, X. Qiu, M. Miyauchi, K. Hashimoto, Energy-level matching of Fe (III) ions grafted at surface and doped in bulk for efficient visible-light photocatalysts, J. Am. Chem. Soc. 135 (2013) 10064–10072. [8] F. Chen, Q. Yang, Y. Wang, J. Zhao, D. Wang, X. Li, Z. Guo, H. Wang, Y. Deng, C. Niu, G. Zeng, Appl. Catal. B: Environ. 205 (2017) 133–147. [9] H. Wang, L. Zhang, Z. Chen, J. Hu, S. Li, Z. Wang, J. Liu, X. Wang, Semiconductor heterojunction photocatalysts: design, construction, and photocatalytic performances, Chem. Soc. Rev. 43 (2014) 5234–5244. [10] J. Liu, B. Cheng, J. Yu, A new understanding of the photocatalytic mechanism of the direct Z-scheme g-C3N4/TiO2 heterostructure, PCCP 18 (2016) 31175– 31183. [11] N. Srinivasan, E. Sakai, M. Miyauchi, Balanced excitation between two semiconductors in bulk heterojunction Z-scheme system for overall water splitting, ACS Catal. 6 (2016) 2197–2200. [12] M. Ou, S. Wan, Q. Zhong, S. Zhang, Y. Song, L. Guo, W. Cai, Y. Xu, Hierarchical Zscheme photocatalyst of g-C3N4@Ag/BiVO4 (040) with enhanced visible-lightinduced photocatalytic oxidation performance, Appl. Catal. B: Environ. 221 (2018) 97–107. [13] W.J. Ong, L.L. Tan, Y.H. Ng, S.T. Yong, S.P. Chai, Graphitic carbon nitride g-C3N4-based photocatalysts for artificial photosynthesis and environmental remediation: are we a step closer to achieving sustainability, Chem. Rev. 116 (2016) 7159–7329. [14] S. Yuan, Q. Zhao, B. Xu, S. Liu, J. Wang, J. Xie, M. Zhang, T. Ohno, A new precursor to synthesize g-C3N4 with superior visible light absorption for photocatalytic application, Catal. Sci. Technol. 7 (2017) 1826–1830. [15] M. Wang, M. Shen, L. Zhang, J. Tian, X. Jin, Y. Zhou, J. Shi, 2D–2D MnO2/g-C3N4 heterojunction photocatalyst: in-situ synthesis and enhanced CO2 reduction activity, Carbon 120 (2017) 23–31. [16] H. Ou, P. Yang, L. Lin, M. Anpo, X. Wang, Carbonitride aerogels for the photoredox conversion of water, Angew. Chem. Int. Ed. 56 (2017) 10905– 10910. [17] G. Li, Z. Lian, W. Wang, D. Zhang, H. Li, Nanotube-confinement induced sizecontrollable g-C3N4 quantum dots modified single-crystalline TiO2 nanotube arrays for stable synergetic photoelectrocatalysis, Nano Energy 19 (2016) 446– 454. [18] J.C. Wang, H.C. Yao, Z.Y. Fan, L. Zhang, J.S. Wang, S.Q. Zang, Z.J. Li, Indirect Zscheme BiOI/g-C3N4 photocatalysts with enhanced photoreduction CO2 activity under visible light irradiation, ACS Appl. Mater. Interfaces 8 (2016) 3765–3775. [19] T. Di, B. Zhu, B. Cheng, J. Yu, J. Xu, A direct Z-scheme g-C3N4/SnS2 photocatalyst with superior visible-light CO2 reduction performance, J. Catal. 352 (2017) 532–541. [20] Y. He, L. Zhang, B. Teng, M. Fan, New application of Z-scheme Ag3PO4/g-C3N4 composite in converting CO2 to fuel, Environ. Sci. Technol. 49 (2015) 649–656. [21] R. Sun, Q. Shi, M. Zhang, L. Xie, J. Chen, X. Yang, M. Chen, W. Zhao, Enhanced photocatalytic oxidation of toluene with a coral-like direct Z-scheme BiVO4/gC3N4 photocatalyst, J. Alloys Compd. 714 (2017) 619–626. [22] Y. Wang, J. Sun, J. Li, X. Zhao, Electrospinning preparation of nanostructured gC3N4/BiVO4 composite films with an enhanced photoelectrochemical performance, Langmuir 33 (2017) 4694–4701. [23] S. Samanta, S. Khilari, D. Pradhan, R. Srivastava, An efficient, visible light driven, selective oxidation of aromatic alcohols and amines with O2 using BiVO4/g-C3N4 nanocomposite: a systematic and comprehensive study toward the development of a photocatalytic process, ACS Sustain. Chem. Eng. 5 (2017) 2562–2577. [24] N. Tian, H. Huang, Y. He, Y. Guo, T. Zhang, Y. Zhang, Mediator-free direct Zscheme photocatalytic system: BiVO4/g-C3N4 organic-inorganic hybrid photocatalyst with highly efficient visible-light-induced photocatalytic activity, Dalton Trans. 44 (2015) 4297–4307. [25] J. Zhao, J. Yan, H. Jia, S. Zhong, X. Zhang, L. Xu, BiVO4/g-C3N4 composite visiblelight photocatalyst for effective elimination of aqueous organic pollutants, J. Mol. Catal. A: Chem. 424 (2016) 162–170. [26] H. Tada, T. Mitsui, T. Kiyonaga, T. Akita, K. Tanka, All-solid-state Z-scheme in CdS–Au–TiO2 three-component nanojunction system, Nat. Mater. 5 (2006) 782–786. [27] Z.F. Huang, J. Song, X. Wang, L. Pan, K. Li, X. Zhang, L. Wang, J.J. Zou, Switching charge transfer of C3N4/W18O49 from type-II to Z-scheme by interfacial band bending for highly efficient photocatalytic hydrogen evolution, Nano Energy 40 (2017) 308–316.

[28] J. Zhang, F. Ren, M. Deng, Y. Wang, Enhanced visible-light photocatalytic activity of a g-C3N4/BiVO4 nanocomposite: a first-principles study, PCCP 17 (2015) 10218–10266. [29] S.P. Adhikari, Z.D. Hood, K.L. More, V.W. Chen, A. Lachgar, A visible-light-active heterojunction with enhanced photocatalytic hydrogen generation, ChemSusChem 9 (2016) 1869–1879. [30] Z. Zhang, J.T. Yates Jr., Band bending in semiconductors: chemical and physical consequences at surfaces and interfaces, Chem. Rev. 112 (2012) 5520–5551. [31] S. Thaweesak, M. Lyu, P. Peerakiatkhajohn, T. Butburee, B. Luo, H. Chen, L. Wang, Two-dimensional g-C3N4/Ca2Nb2TaO10 nanosheet composites for efficient visible light photocatalytic hydrogen evolution, Appl. Catal. B: Environ. 202 (2017) 184–190. [32] M. Liu, L. Lv, X. Du, J. Lang, Y. Su, Y. Zhao, X. Wang, Photo-synergistic promoted in situ generation of Bi0-BiSbO4 nanostructures as an efficient catalyst for nitrobenzene reduction, RSC Adv. 5 (2015) 103013–103018. [33] Z. Ai, W. Ho, S. Lee, L. Zhang, Efficient photocatalytic removal of NO in indoor air with hierarchical bismuth oxybromide nanoplate microspheres under visible light, Environ. Sci. Technol. 43 (2009) 4143–4150. [34] Y. Yang, W. Guo, Y. Guo, Y. Zhao, X. Yuan, Y. Guo, Fabrication of Z-scheme plasmonic photocatalyst Ag@AgBr/g-C3N4 with enhanced visible-light photocatalytic activity, J. Hazard. Mater. 271 (2015) 150–159. [35] S.C. Yan, Z.S. Li, Z.G. Zou, Photodegradation of rhodamine B and methyl orange over boron-doped g-C3N4 under visible light irradiation, Langmuir 26 (2010) 3894–3901. [36] H. Ji, F. Chang, X. Hu, Q. Wei, J. Shen, Photocatalytic degradation of 2, 4, 6-trichlorophenol over g-C3N4 under visible light irradiation, Chem. Eng. J. 218 (2013) 183–190. [37] C.N. Ri, S.G. Kim, K.S. Ju, H.S. Ryo, C.H. Mun, U.H. Kim, The synthesis of a Bi2MoO6/Bi4V2O11 heterojunction photocatalyst with enhanced visible-lightdriven photocatalytic activity, RSC Adv. 8 (2018) 5433–5440. [38] C. Ye, J.X. Li, Z.J. Li, X.B. Li, X.B. Fan, L.P. Zhang, B. Chen, C.H. Tung, L.Z. Wu, Enhanced driving force and charge separation efficiency of protonated g-C3N4 for photocatalytic O2 evolution, ACS Catal. 5 (2015) 6973–6979. [39] T. Giannakopoulou, I. Papailias, N. Todorova, N. Boukos, Y. Liu, J. Yu, C. Trapalis, Tailoring the energy band gap and edges’ potentials of g-C3N4/TiO2 composite photocatalysts for NOx removal, Chem. Eng. J. 310 (2017) 571–580. [40] Y. Fu, T. Huang, L. Zhang, J. Zhu, X. Wang, Nanoscale 7 (2015) 13723–13733. [41] X. Xin, J. Lang, T. Wang, Y. Su, Y. Zhao, X. Wang, Construction of novel ternary component photocatalyst Sr0.25H1.5Ta2O6
H2O coupled with g-C3N4 and Ag toward efficient visible light photocatalytic activity for environmental remediation, Appl. Catal. B: Environ. 181 (2016) 197–209. [42] Y.H. Chang, F.Y. Wu, T.Y. Chen, C.L. Hsu, C.H. Chen, F. Wiryo, K.H. Wei, C.Y. Chiang, L.J. Li, Three-dimensional molybdenum sulfide sponges for electrocatalytic water splitting, Small 10 (2014) 895–900. [43] J.J. Duan, S. Chen, B.A. Chambers, G.G. Andersson, S.Z. Qiao, 3D WS2 nanolayers@heteroatom-doped graphene films as hydrogen evolution catalyst electrodes, Adv. Mater. 27 (2015) 4234–4241. [44] Z. Hou, P. Yang, C. Li, L. Wang, H. Lian, Z. Quan, J. Lin, Preparation and luminescence properties of YVO4: Ln and Y(V, P)O4: Ln (Ln = Eu3+, Sm3+, Dy3+) nanofibers and microbelts by sol gel/electrospinning process, Chem. Mater. 20 (2008) 6686–6696. [45] L.L. Chen, D.L. Jiang, T. He, Z.D. Wu, M. Chen, In-situ ion exchange synthesis of hierarchical AgI/BiOI microsphere photocatalyst with enhanced photocatalytic properties, CrystEngComm 15 (2013) 7556–7563. [46] H. Lin, L. Ding, Z. Pei, Y. Zhou, J. Long, W. Deng, X. Wang, Au deposited BiOCl with different facets: on determination of the facet-induced transfer preference of charge carriers and the different plasmonic activity, Appl. Catal. B: Environ. 160–161 (2014) 98–105. [47] H.Y. Si, C.J. Mao, J.Y. Zhou, X.F. Rong, Q.X. Deng, S.L. Chen, J.J. Zhao, X.G. Sun, Y. M. Shen, W.J. Feng, P. Gao, J. Zhang, Z-scheme Ag3PO4/graphdiyne/g-C3N4 composites: enhanced photocatalytic O2 generation benefiting from dual roles of graphdiyne, Carbon 132 (2018) 598–605. [48] H.B. Yu, B.B. Huang, H. Wang, X.Z. Yuan, L.B. Jiang, Z.B. Wu, J. Zhang, G.M. Zeng, Facile construction of novel direct solid-state Z-scheme AgI/BiOBr photocatalysts for highly effective removal of ciprofloxacin under visible light exposure: mineralization efficiency and mechanisms, J. Colloid Interface Sci. 522 (2018) 82–94. [49] Y. Lin, S.H. Wu, X. Lia, X. Wu, C.P. Yang, G.M. Zeng, Y.R. Peng, Q. Zhou, L. Lu, Microstructure and performance of Z-scheme photocatalyst of silver phosphate modified by MWCNTs and Cr-doped SrTiO3 for malachite green degradation, Appl. Catal. B 227 (2018) 557–570. [50] J.B. Pan, J.J. Liu, S.l. Zuo, U.A. Khan, Y.C. Yu, B.S. Li, Structure of Z-scheme CdS/ CQDs/BiOCl heterojunction with enhanced photocatalytic activity for environmental pollutant elimination, Appl. Surf. Sci. 444 (2018) 177–186. [51] X.Q. Liu, L. Cai, Novel indirect Z-scheme photocatalyst of Ag nanoparticles and polymer polypyrrole co-modified BiOBr for photocatalytic decomposition of organic pollutants, Appl. Surf. Sci. 445 (2018) 242–254. [52] L.B. Jiang, X.Z. Yuan, G.M. Zeng, J. Liang, X.H. Chen, H.B. Yua, H. Wang, Z.B. Wu, J. Zhang, T. Xiong, In-situ synthesis of direct solid-state dual Z-scheme WO3/gC3N4/Bi2O3 photocatalyst for the degradation of refractory pollutant, Appl. Catal. B 227 (2018) 376–385.