LIU Mu-yao ，WANG Jian-yun ，DUAN Lian,* ，LIU Xian ，ZHANG Lei

（1. School of Chemistry and Chemical Engineering, Taiyuan University of Technology, Taiyuan 030024, China；2. Department of Chemistry, Taiyuan Normal University, Jinzhong 030619, China；3. Shenzhen Huace International Certification Co., Ltd, Shenzhen 518101, China）

Abstract: Recently, a new carbon nitride (C3N5) photocatalyst has attracted much attention due to its excellent light harvesting and unique 2D structure. However, high recombination rates of electron-hole pairs of bulk C3N5 serious affect the photocatalytic performance. Herein, nickel oxide (NiO) modified C3N5 p-n junctions photocatalyst was synthesized by a facile hydrothermal method. Results indicated that the 9-Ni/C3N5 nanosheet photocatalyst showed excellent hydrogen production efficiency under visible light. The hydrogen production rate reached 357 μmol/(g?h), which was 107-fold higher than that of pristine C3N5. The high catalytic performace was attributed to the 9-Ni/C3N5 p-n junctions which could efficiently promote photogenerated electron-hole pair separation and thus promote the hydrogen evolution reaction.

Key words: hydrogen evolution；C3N5 nanosheet；nickel oxide；photocatalysis

Graphitic carbon nitride (g-C3N4), the most popular metal-free semiconductor, has been widely used for sunlight-driven water splitting[1?3], carbon dioxide photoreduction[4,5]and organic pollutant photodegradation[6,7]. However, the somewhat wide bandgap of g-C3N4means that it can absorb only the ultraviolet and blue fraction of the solar spectrum (λ<450 nm), which has seriously limited its photocatalytic performance[8,9]. It has been found that the bandgap can be reduced significantly by increasing the N/C ratio[10].Thereby, it is quite necessary to develop N-rich carbon nitride materials.

g-C3N5, as a new carbon nitride photocatalyst with high nitrogen content and narrower bandgap, showed extraordinary properties in the field of photocatalysis[11]. The C3N5framework contains heptazine moieties bridged together by azo linkage(?N=N?). The presence of azo linkage extends the π conjugated network due to overlap between theporbitals on N atoms constituting the azo bond and π system of heptazine motif, which resulted in the reduction of the electronic bandgap[12]. Like most of semiconductors, g-C3N5also suffers the innate drawback of carrier recombination and low specific surface area of bulk g-C3N5[13]. In the previous reports,various hybrid composites have been developed to improve the catalytic activity of C3N5. For example, the CeTi2O6/g-C3N5heterojunction exhibited outstanding photocatalytic response under the visible light towards the degradation of endocrine rupture material 2,4-dichlorophenol (2,4-DCP) than its single component[14].Nitrogen vacancies g-C3N5/BiOBr composites also exhibited excellent PEC NRR performance without the addition of noble metals[15].

Nickel (II) oxide (NiO), as a non-noble metalptype semiconductor monoxide, plays an important role in the photocatalytic hydrogen production process due to it can offer more active sites for hydrogenreleasing[16?21]. Moreover, NiO can promote the separation of photogenerated electron-hole pairs by formation ofp-njunctions with othern-type semiconductors. Recently, NiO modified g-C3N4nonnoble metal heterojunction photocatalyst exhibited enhanced phototcatalytic performance for hydrogen production[22?24]. Therefore, it is possible to design NiO/C3N5p-njunctions photocatalysts with desirable performance. To the best of our knowledge, there is no report available for NiO/g-C3N5composite towards the photocatalytic hydrogen evolution.

In the present work, we aimed to construct a NiO modified C3N5non-noble metal photocatalyst with enhanced hydrogen evolution performance. Varying amounts of Ni modified C3N5samples were prepared and characterized. Based on the experimental results, it can be found that the migration and separation of photogenerated electron-hole pairs by formation ofp-njunctions is beneficial to the photocatalytic hydrogen evolution performance. Furthermore, the possible photocatalytic hydrogen evolution mechanism was discussed.

1 Experimental

1.1 Preparation of bulk C3N5

All the reagents in this work were of analytical grade and used as received without any purification.Bulk C3N5had been prepared expediently from the thermal polymerization method[25]. In brief, 8 g 3-amino-1,2,4-triazole (3-AT) and 8 g NH4Cl was grounded to form a homogeneous solid mixture. The mixture was heated at 550 °C for 3 h with a ramping rate of 15 °C/min, and cooled naturally to room temperature. Finally, the obtained chocolate brown color product was dried at 70 °C in a vacuum oven for 5 h.

1.2 Preparation of C3N5 nanosheets

The C3N5nanosheets were obtained according to the literature with a few modifications[26]. Simply, 1.0 g bulk C3N5was added into round-bottom flask containing 20 mL 3.0 mol/L nitric acid solution. The solution was stirred continuously for 24 h till it turned yellow. Then product was diluted with 1.0 L deionized water, collected by suction filtration with membrane and dried at 70 °C for 4 h. The obtained powder was named as C3N5nanosheets.

1.3 Preparation of 9-Ni/C3N5 composite

The 9-Ni/C3N5sample was prepared by a hydrothermal method. Firstly, 50 mg C3N5nanosheets powder was dispersed in 80 mL of distilled water with vigorous ultrasound for 1.0 h. Thereafter, a desired amount of NiCl2?6H2O was added with continuous stirring, and then adjust the pH value 12.0 using 28%ammonia. After 0.5 h, the suspension was transferred to a 100 mL Teflon-lined stainless-steel autoclave and maintained at 150 °C for 12 h. Subsequently, the powder product was centrifugated, washed and dried at 70 °C in the vacuum drying oven. The final product was obtained after heat treatment at 300 °C for 4 h in air. The samples were named asx-Ni/C3N5, wherexis the weight ratio of NiCl2?6H2O (0, 3%, 5%, 9% and 18%) to the composite.

1.4 Photocatalytic H2 generation

Photocatalytic H2generation experiments were measured in a lab solar H2-evolution system. Xe lamp with an AM 1.5 G filter (CEL-HXF300, 300 W,λ≥420 nm) was used as a simulated solar light and the light density was 160 mW/cm2. In a typical measurement, 50 mg NiO/C3N5composite was dispersed by ultrasound into a 100 mL triethanolamine(TEOA) solution (15%) and the system was kept at?0.1 MPa and 5 °C. GC-7900 gas chromatograph was employed to detect the H2on line after every 1 h.

2 Results and discussion

2.1 Characterizations of NiO/C3N5 composite

The structure and morphology were observed by scanning electron microscopy (SEM) and transmission electron microscopy (TEM), as shown in Figure1.

From Figure1(a), it can be seen that C3N5shows an obvious sheet-like structure by acidifying. After modified with NiO, the dense multilayer structure was formed (Figure1(b)), this structure was also verified by high magnification TEM image (Figure1(c)). From Figure1(d), we can observe that some dark nanoparticles dispersed uniformly on the multilayer of 9-Ni/C3N5and the average diameter of the black spots is about 5?6 nm, which demonstrate that NiO particles were successfully decorated on the surface of C3N5.The elemental mapping on the selected area shown in Figure1(e)?(f) demonstrates a uniformly distribution of C, N and Ni atoms on the surface of the NiO/C3N5composite.

XRD patterns of the as-prepared samples are presented in Figure2(a). All samples show strong diffraction peak at 28.1°, which is assigned to the interlayer stacking of graphitic-like structure, indicating that the lattice structure of C3N5remains unchanged after NiO modification, which is beneficial for photocatalytic properties of NiO/C3N5. Moreover, all of the NiO/C3N5samples exhibit a little blue-shift to about 27.1°?27.6°, suggesting that parts of Ni may implant into the lattice of C3N5. No obvious peak was detected after loading NiO species onto pure C3N5, probably because of low crystallinity and amorphous structure.

To gain an insight into the surface functional group of as prepared samples, the FT-IR analysis was carried out. As exhibited in Figure2(b), there is an obvious peak at about 808 cm?1, which can be assigned to the bending vibration of triazine units[27]. In addition,the band ranged from 1200 to 1600 cm?1belongs to the stretching in aromatic C?N. The broad peak at 3000?3600 cm?1is attributed to the stretching vibration of N?H or O?H groups. The bare C3N5and Ni/C3N5nanocomposites show similar absorption bands,indicating that the structure of C3N5remains unchanged after NiO modification.

To analysis the composition and chemical state of constituent elements in 9-Ni/C3N5sample, XPS measurements were conducted. Referring to Figure3(a),the XPS survey spectra of 9-Ni/C3N5sample not only exhibits the peaks of C 1speak and N 1speak, but also exhibits a relatively weak O 1sand Ni 2ppeak,indicating the well combination of C3N5and NiO. In the C 1sspectrum (Figure3(b)), the binding energies at 285 and 288.25 eV can be attributed to C?C and N?C=N peaks, respectively[28]. There are three peaks for the N 1sXPS peak in Figure3(c), which are ascribed to the amino groups (400.7 eV), N(C)3(399.7 eV)and C?N=C (398.6 eV), respectively[3]. The peak of Ni spectra (Figure3(d)) centered at 872.7 and 855.4 eV is related to Ni 2p1/2and Ni 2p3/2, respectively[29?31]. The two relatively weak satellite peaks located at 879.9 and 862.2 eV belong to the shake-up types of Ni 2p1/2and Ni 2p3/2. This result was consistent to previous literature.

2.2 Photocatalytic H2-production

The photocatalytic H2-production activity of thex-Ni/C3N5nanosheets was evaluated using triethanolamine as an electron donor. As shown in Figure4(a), as a controlled experiment, negligible H2was detected without either photocatalyst or irradiation.Trace H2evolution was observed for bare C3N5, while Ni-modified C3N5nanosheets displayed better H2evolution rate than bare C3N5. After loading different Ni proportions (3%, 5%, 9% or 18%) on C3N5nanosheets, the photocatalytic performance of NiO/C3N5has been remarkably improved, and the highest amount of H2evolution of 9-Ni/C3N5can reach 357 μmol/(g?h) in the first 4 h reaction. The excellent photocatalytic H2-production activity is probably due to the separation of photogenerated electron-hole pairs by formation ofp-njunctions, which can be confirmed by PL (Figure5(a)). The PL emission intensity of 9-Ni/C3N5was the lowest among all the samples,indicating 9-Ni/C3N5has the highest separation efficiency of electrons and holes. Moreover, the BET surface area of 9-Ni/C3N5was the largest among all the samples (the BET surface area of 9-Ni/C3N5was found to be 24.6425 m2/g, which was larger than bulk C3N5(20.7077 m2/g), 3-Ni/C3N5(22.1024 m2/g), 5-Ni/C3N5(24.4835 m2/g) and 18-Ni/C3N5(9.3544 m2/g)), which was benefit to catalytic hydrogen evolution reaction.

In order to demonstrate the stability and durability of 9-Ni/C3N5, the recycling experiments were performed under similar experimental conditions, and each test cycle time is 4 h. As shown in Figure4(b), the results show that the amount of H2produced was retained by about 82% after four cycles, indicating the high stability properties of 9-Ni/C3N5composite.

2.3 Mechanism analysis

According to the results of mentioned above, the enhanced H2-production performance of 9-Ni/C3N5photocatalyst mainly attributes to three reasons: (1)NiO nanoparticles offer more active sites for hydrogenreleasing, (2) broaden the photo light absorption region and (3) the effective separation of photogenerated electron-hole pairs by formation ofp-njunctions,which were verified further by the following experiments.

Photoluminescence (PL) spectroscopy is usually employed to investigate the optical properties and charge-separation efficiency. As shown in Figure5(a),PL intensity of NiO/C3N5gradually increases first, and then decreases with the increase of Ni content. These results indicate that the NiO nanoparticles on C3N5are able to effectively promote the transfer of charge and thus inhibit photogenerated electron-hole pairs recombination. The PL emission intensity of 9-Ni/C3N5was the lowest among all the samples, indicating 9-Ni/C3N5has the highest separation efficiency of electrons and holes. The time-resolved PL (TRPL)decay profiles of C3N5and 9-Ni/C3N5in Figure5(b)showed that the charge carrier lifetime in 9-Ni/C3N5(2.386 ns) was longer than that of C3N5(2.006 ns). The result further indicated that 9-Ni/C3N5has excellent separation efficiency of electrons and holes.

The optical absorption properties of the samples were tested by UV-vis absorption spectroscopy. As shown in Figure6(a), both C3N5and 9-Ni/C3N5exhibit clear absorption in the region of visible light. Their band-gap energies are 1.95 and 1.78 eV, respectively(Figure6(b)), which agree with previous reports well.Furthermore, 9-Ni/C3N5can harvest visible light more efficiently for above catalytic reaction, which may be due to a broader tail (bathochromic/red shift) in the absorption spectrum. The band positions were confirmed by Mott-Schottky plots, as depicted in Figure6(c). The flat band or CB potential of C3N5was?0.67 V.

From the obtained CB and band gap potential, it is easy to estimate the VB position of C3N5, which was calculated to be 1.28 V. The positive slopes could prove then-type semiconductors properties of C3N5,which was favorable to formp-njunction withp-type NiO nanoparticles. Thep-njunction will promote the separation of electron-hole pairs. This result was further clarified by the electrochemical impedance spectroscopy (EIS) plot (Figure6(d)). 9-Ni/C3N5has the smaller semicircle diameter compared to pure C3N5in higher frequency region, which indicated enhanced charge-carrier transfer ability of 9-Ni/C3N5[32].

Based on the results and analysis above, a possible photocatalytic H2-production mechanism over 9-Ni/C3N5photocatalyst was summarized in Figure7.

Under visible light irradiation, only C3N5can easily absorb visible light and generate photo-induced electron-hole pairs. Photogenerated electrons in the CB of C3N5migrate to the more negative potential CB of NiO (ECB= ?0.5 V vs NHE) for proton reduction, and photogenerated holes in the VB of C3N4were consumed by the sacrificial electron donor TEOA. In a word, the photo-induced electron-hole pairs can be efficient separation by constructing an inner electric field of NiO/C3N5p-nheterojunction.

3 Conclusions

In summary, we successfully synthesized 9-Ni/C3N5p-njunctions photocatalyst through a facile hydrothermal method. This photocatalyst showed excellent hydrogen production efficiency under visible light. The hydrogen production rate reached 357 μmol/(g?h), which was 107-fold higher than that of pristine C3N5. This mainly attributed to the NiO modification being able to promote photoinduced electron-hole pair separation, thus promote the hydrogen evolution reaction. This work has provided a feasible strategy to design non-noble metal modified carbon nitride for high-efficient solar conversion.