Z-Scheme LaCoO3/g-C3N4 for Efficient Full-Spectrum Light-Simulated Solar Photocatalytic Hydrogen Generation.

Photocatalytic decomposition of water is the most attractive method for the sustainable production of hydrogen, but the development of a highly active and low-cost catalyst remains a major challenge. Here, we report the preparation of LaCoO3/g-C3N4 nanosheets and the utilization of LaCoO3 instead of noble metals to improve the photocatalytic activity for the production of hydrogen. First, LaCoO3 was successfully prepared by the sol-gel method, and then a series of highly efficient Z-scheme LaCoO3/g-C3N4 heterojunction photocatalysts were synthesized by the solvothermal method. Various characterization techniques (X-ray diffraction (XRD), Fourier transform infrared (FT-IR), scanning electron microscopy (SEM), transmission electron microscopy (TEM), energy-dispersive spectroscopy (EDS), X-ray photoelectron spectroscopy (XPS), ultraviolet-visible (UV-vis) diffuse reflectance spectroscopy (DRS), photoluminescence (PL), transient photocurrent response test, electron paramagnetic resonance (EPR)) confirm that the heterostructure and interfacial interaction had been formed between LaCoO3 nanoparticles and g-C3N4 nanosheets. In the photocatalytic water splitting test, LaCoO3/g-C3N4-20 wt % exhibited the highest photocatalytic activity of 1046.15 μmol h-1 g-1, which is 3.5 and 1.4 times higher than those of LaCoO3 and g-C3N4, respectively. This work leads to an inexpensive and efficient LaCoO3/g-C3N4 photocatalysis system for water splitting or other photocatalytic applications.

Introduction

With increasingly serious environmental and energy issues, it has become urgent to explore environmentally friendly, economical, and renewable alternative energy sources to promote future sustainable development. Hydrogen, as a nonpolluting, high-energy-density energy source, is expected to replace fossil energy as a clean energy source in the future.[1−4] Light-driven photocatalytic water splitting is one of the most attractive methods for the sustainable production of hydrogen. Among the widely studied photocatalysts, graphitic carbon nitride (g-C3N4) has attracted considerable interest. Because it has sp2 hybridized carbon and nitrogen atoms arranged in six-member stacked rings[5−9] and semiconductor characteristics with a band gap of ∼2.7 eV, it has very promising properties such as high chemical stability and a fast charge transfer rate.[10−14] However, g-C3N4 also has a narrow absorption band and a high recombination rate of photoelectron–hole pairs.[15−19] To overcome these disadvantages, researchers at home and abroad have developed a large number of different methods. In recent studies, a series of new attempts have been made, such as Bi2W2O9/g-C3N4,[20] AgBr/P-g-C3N4,[12][12] g-C3N4 with SnO2,[21][21] MoS2/Ag/g-C3N4,[22][22] WO3/g-C3N4,[23][23] BiOI/g-C3N4,[24] Co3(PO4)(2)/g-C3N4,[25] and MoS2/g-C3N4.[26] Although these reports indicate that the construction of heterostructures with energy levels in a staggered gap is conducive to the separation of charge carriers, they still need to solve the problems of low catalytic efficiency and high cost. Designing and constructing heterojunction photocatalysts on g-C3N4 is still a challenge, and discovering a material with a suitably positioned energy level, light activity, light stability, low cost, and easy preparation is the goal of many scientists.

Perovskite-type semiconductor materials are promising photocatalysts due to their tunable band gaps, strong resistance to photocorrosion, and sufficient oxygen vacancies.[27−29] Because of these characteristics, they are very promising materials for solar cells and photocatalytic reactions and have attracted considerable attention from researchers.[28,30−32] The formation of a heterojunction with g-C3N4 was expected to improve the photocatalytic performance of g-C3N4 and perovskite. This strategy reduces the recombination of electrons and holes, enhances charge separation, and improves the overall photocatalytic activity. Cai et al.[33] reported the synthesis of graphitic-C3N4-hybridized N-doped La2Ti2O7 two-dimensional layered composites for photocatalytic H2 evolution. The optimal Pt-g-C3N4/NLTO with 10 wt % g-C3N4 shows a H2 evolution rate of 430 μmol h–1 g–1. Xu et al.[34] reported a LaFeO3/g-C3N4 heterojunction with Pt (3 wt %) as a cocatalyst for photohydrolytic hydrogen production. The optimal material for hydrogen production is the 5%-LaFeO3/g-C3N4 heterojunction nanomaterial, and the maximum hydrogen release rate of the composite material is 158 μmol g–1 h–1 under visible light irradiation from a 300 W xenon lamp. Chen et al.[35] reported a Z-scheme two-dimensional (2D)/2D MnIn2S4/g-C3N4 architecture toward treatment of pharmaceutical wastewater and hydrogen evolution. The result shows a H2 evolution rate of 200.8 μmol h–1 g–1. Previous reports indicate that perovskite and g-C3N4 heterojunction photocatalytic hydrogen production efficiency is low and that the materials are mostly doped with precious metals. Although this doping has increased the hydrogen production efficiency, the cost of the materials has also increased significantly. Some exploration of the combination of g-C3N4 and perovskite for hydrogen production by photocatalytic hydrolysis is necessary, which requires further research. Therefore, we explored and prepared a high-perovskite-doped LaCoO3/g-C3N4 heterojunction to improve photocatalytic efficiency.

In this study, the sol–gel method was used to synthesize LaCoO3. Using mechanical agitation and solvothermal methods, LaCoO3 was evenly loaded on g-C3N4 nanosheets, which catalyzed water decomposition to produce hydrogen under full-spectrum irradiation. A new type of Z-scheme heterojunction photocatalyst was designed and prepared. The formation of the Z-scheme heterostructure improves the separation efficiency of photogenerated electron–hole pairs and improves the photocatalytic activity. The photocatalytic mechanism of the LaCoO3/g-C3N4 heterojunction photocatalyst is proposed. By forming a Z-scheme heterojunction structure, the photocatalytic activity is improved due to the high separation efficiency of electron–hole pairs. The photocatalytic hydrogen production rate is as high as 1046.15 μmol g–1 h–1, and the optimal composite rate is approximately 3.5 and 1.4 times those of bare LaCoO3 and g-C3N4, respectively. To the best of our knowledge, this is the first study to employ the solvothermal method to hybridize the high-content layered perovskite photocatalysts LaCoO3 and g-C3N4. Furthermore, the full spectrum of xenon lamps is used to simulate sunlight across the full waveband for the photocatalysis experiments.

Experimental Section

Materials

Lanthanum nitrate hexahydrate (purity > 99.99%), cobalt nitrate hexahydrate (purity > 99.99%), citric acid (purity > 99%), and urea (purity > 99%) were used. All other reagents used in this study were analytically pure and used without further purification. Deionized water was used for all of the experiments.

Catalyst Preparation

Preparation of g-C3N4

g-C3N4 was prepared by a facile thermal oxidation method with direct heating of urea. First, urea was placed in a ceramic crucible and dried at 80 °C in a blast-drying oven for 10 h. Then, it was heated to 550 °C in a muffle furnace for 1.74 h at a heating rate of 5 °C min–1 in air and kept at this high temperature for 3 h. Yellow powders were obtained and ground after the products were cooled to room temperature.

Preparation of LaCoO3

The obtained solid was calcined at 400 °C for 4 h and then at 700 °C for 4 h in a muffle furnace. The black LaCoO3 powders were collected after the machine was cooled naturally.

Preparation of the LaCoO3/g-C3N4 Composites

LaCoO3/g-C3N4 with different weight ratios were prepared by the solvothermal method. Typically, certain amounts of LaCoO3 and g-C3N4 were separately dispersed into 10 mL of absolute ethanol and ultrasonically processed for 30 min. Then, the LaCoO3 suspension was slowly added to the g-C3N4 ethanol solution under stirring and again ultrasonicated for 30 min. Subsequently, the mixed solution was stirred at ambient temperature for 12 h and transferred to a hydrothermal kettle to be heated at 120 °C for 6 h. Finally, the resulting mixture was centrifuged and dried at 80 °C for 12 h. Samples of g-C3N4/LaCoO3 nanosheets with different LaCoO3 contents of 70, 75, 80, 85, and 90 wt % were prepared.

Characterization

The crystal structures and phase composition of the obtained samples were determined by means of a Shimadzu Maxima X-ray diffractometer (XRD 7000) with Cu Kα radiation at a current of 30 mA and a voltage of 40 kV. The surface morphologies and microstructures of the samples were characterized using scanning electron microscopy (SEM, Hitachi S-4800) and transmission electron microscopy (TEM, FEI Talos F200S) in conjunction with energy-dispersive spectroscopy (EDS). X-ray photoelectron spectroscopy (XPS) analysis was performed with a ThermoFisher electron spectrometer (XPS, Escalab 250Xi) equipped with a monochromatized microfocused Al Kα X-ray source, and the binding energy was referenced to the C 1s peak at 284.6 eV. Fourier transform infrared (FT-IR) spectroscopy was carried out using an infrared spectrometer (Shimadzu IRAffinity-1). Using a UV-2600 spectrophotometer (Shimadzu, Japan; BaSO4 as the reflection standard), the optical properties of the prepared samples were tested by ultraviolet–visible (UV–vis) diffuse reflectance spectroscopy (DRS). Photoluminescence (PL) spectroscopy was performed at room temperature on an FS5 fluorescence spectrophotometer with a 500 W xenon lamp light source and an excitation wavelength of 300 nm. In a standard three-electrode system, we use a CHI660E electrochemical workstation and a platinum mesh as the counter electrode and a silver/silver chloride electrode as the reference electrode. 5 mg of LaCoO3, g-C3N4, and LaCoO3/g-C3N4-20% wt % is, respectively, dissolved in 800 μL of distilled water, 200 μL of isopropanol, and 30 μL of Nafion mixed solution, ultrasonic treatment, drops on the glassy carbon electrode as a working electrode. Further, 0.1 mol L–1 Na2SO4 solution was used as the electrolyte. The width of the excitation and emission slit is 5 nm. The H2 yield was detected on a Shimadzu gas chromatograph (GC-2014C) by manual injection. The electron spin resonance spectrum is measured by an electron paramagnetic resonance (EPR) spectrometer (Bruker E500).

Photocatalytic Experiment

The photocatalytic activity of the as-prepared samples was evaluated by hydrogen production by light-driven water splitting. In this article, all water splitting experiments were performed at room temperature with magnetic stirring during light irradiation. The light source was a 300 W xenon lamp (PLS-SXE300) using full-spectrum (250 < λ < 1200 nm) simulated sunlight to perform the photocatalysis experiments. The specific operation parameters were as follows. First, 0.03 g of the prepared catalyst was placed in a custom-made quartz round-bottom flask and sonicated to suspend it in 30 mL of 10% CH3OH aqueous solution in the dark, and nitrogen gas was passed through the solution for 30 min to remove dissolved oxygen. A gas chromatograph equipped with a molecular sieve column and a TCD detector was used to monitor the H2 precipitation rate every hour. We used a manual injection method using a gas injector. It is worth mentioning that each material was subjected to three photocatalytic experiments to eliminate unexpected factors and enhance the reliability of the experimental data.

Results and Discussion

Morphology and Component Analysis

With a smooth and flat surface, g-C3N4 is an excellent substrate, as shown in Figure a. It can be seen from Figure b that pure LaCoO3 is a collection of many nanoscale ellipsoids, and each ellipsoidal columnar particle grows closely together to form a two-dimensional nanolayer structure. From Figure c,d, we can observe that the LaCoO3 nanoparticle layer was divided into small pieces that grew vertically on the g-C3N4 nanosheets. This structure may contribute to the effective separation of electron and hole pairs and the migration of electrons and holes, respectively.

Figure 1

SEM images of (a) g-C3N4, (b) LaCoO3, and (c, d) LaCoO3/g-C3N4-20 wt %.

Furthermore, the material structure and nanoparticle size were verified by TEM and high-resolution TEM (HRTEM). In Figure a, LaCoO3 shows an obvious ellipsoid structure with a distribution from 90 to 130 nm. From the HRTEM image in Figure b,c, we can observe that the interplanar space between adjacent lattice fringes is approximately 0.27 nm, which is consistent with (104). As seen in Figure d,e, g-C3N4 shows an obvious lamellar structure, which is consistent with the above observation. In Figure f,g, lattice plane separation can be observed at the interfaces between LaCoO3 and g-C3N4. We can observe that the plane spacing between adjacent lattice stripes is approximately 0.2703 nm, consistent with (104). In addition, the selected area electron diffraction (SAED) patterns of the LaCoO3/g-C3N4-20 wt % composite reflected a single crystal structure with a few irregular bright spots, which means that the original structure of LaCoO3 changed due to the combination with g-C3N4. Energy-dispersive X-ray spectroscopy (EDS) confirmed the elemental composition and distribution of LaCoO3/20 wt % g-C3N4. The above microscopy images also prove the strong coupling between LaCoO3 and g-C3N4 (Figure ).

Figure 2

(a) TEM image and (b, c) HRTEM images of LaCoO3; (d) TEM images and (e) HRTEM image of g-C3N4; and (f) TEM image, (g) HRTEM images, and (h) SAED patterns of the LaCoO3/g-C3N4-20 wt % composite.

Figure 3

EDS elemental mapping analysis of the LaCoO3/g-C3N4-20 wt % composite.

Phase Structure Analyses

The crystalline structures of pure LaCoO3, g-C3N4, and g-C3N4/LaCoO3 with different weight ratios were characterized by X-ray diffraction (XRD) and are shown in Figure . The X-ray powder diffraction characteristic curve of g-C3N4 shows two main peaks at 13.1 and 27.7°. The stronger peak at 27.7° could be attributed to the (002) facet of the layered g-C3N4, and the weaker peak at 13.1° could be attributed to an in-plane (100) facet (JCPDS card no. 87-1526).[36−38] Each XRD diffraction peak of LaCoO3 is in good agreement with the crystal phase of JCPDS No. 84-0848, and there are no other diffraction peaks, indicating that the as-prepared sample is bare LaCoO3.[39] Interestingly, the peak intensity of LaCoO3 decreased significantly after the combination of LaCoO3 and g-C3N4, but the characteristic peak of g-C3N4 cannot be clearly observed in the conjugate, which might be ascribed to the low crystallinity of g-C3N4 and the small amount of g-C3N4 in the hybrid.[39,40]

Figure 4

XRD patterns of g-C3N4; LaCoO3; and 10,15, 20, 25, and 30 wt % g-C3N4/LaCoO3 powders.

FT-IR Result Analysis

The FT-IR results of pure LaCoO3, g-C3N4, and g-C3N4/LaCoO3 with different weight ratios are shown in Figure . For pure g-C3N4, the characteristic peaks at 1508.4, 1541.2, and 1558.6 cm–1 were consistent with aromatic C–N stretching vibration modes. The peaks at approximately 3649.48 and 3736.3 cm–1 were attributed to the stretching and bending vibrations of N–H, respectively, which are derived from the uncondensed terminal amino group. In addition, the breathing vibration of s-triazine to g-C3N4 gave rise to a characteristic absorption peak at 810 cm–1. The characteristic peaks of LaCoO3 at 598 cm–1 were related to the bending and tensile vibration of Co–O and corresponded to the perovskite structure. With an increase in g-C3N4 content in the LaCoO3 samples, the peak of the composite gradually became sharper. Moreover, compared with that of g-C3N4, the spectrum of the LaCoO3/g-C3N4 sample showed a blue shift at 810 cm–1 (tritriazine unit). This indicates that the number of hydrogen bonds in the g-C3N4 structure was reduced, which would promote the transport of charge carriers in the carbon nitride layer;[41] this was more significant for LaCoO3/g-C3N4.

Figure 5

(a) FT-IR spectra of g-C3N4; LaCoO3; and 15, 20, and 25 wt % LaCoO3/g-C3N4. (b) FT-IR spectra of 10, 15, 20, 25, and 30 wt % LaCoO3/g-C3N4.

XPS Analysis

To confirm the element chemical state, the surface chemical composition, the interaction between LaCoO3 and g-C3N4, and the XPS spectra were studied, and the results are depicted in Figure . All of the binding energies were calibrated relative to the C 1s peak at 284.6 eV. The LaCoO3/20 wt %-g-C3N4 composite is composed of La, Co, O, C, and N. The high-resolution C 1s spectrum of pure g-C3N4 could be deconvoluted into two main peaks located at approximately 284.6 and 288.0 eV. The C 1s peak at 284.6 eV is attributed to the C–C bond in the adventitious carbon, and the peak at 288.0 eV is attributed to C=N bonds of the sp2-type.[42,43] The high-resolution N 1s spectrum of the g-C3N4 sample shown in Figure c could be deconvoluted into three main peaks. We assigned the N 1s peak at the lowest binding energy (398.6 eV) to the C–N bond, assigned the center peak (399.3 eV) to the sp2-type C=N bond,[44] and determined the peak with the largest binding energy at 401.0 eV to be derived from nitrogen surrounded by three carbon atoms in the amorphous CN network. It is worth mentioning that the N–(C)3 bond peak in the spectrum of LaCoO3/20 wt % g-C3N4 red-shifts by 0.4 eV, which could indicate an increase in electron density after g-C3N4 is combined with LaCoO3.

Figure 6

(a) XPS spectrum survey scan for LaCoO3/g-C3N4-20 wt % and pure LaCoO3,g-C3N4; high-resolution spectra of (b) C 1s and (c) N 1s for LaCoO3/g-C3N4-20 wt % and pure g-C3N4; and high-resolution XPS spectra of (d) O 1s, (e) La 3d, and (f) Co 2p for LaCoO3/g-C3N4-20 wt % and pure LaCoO3.

The high-resolution O 1s XPS spectrum presents two primary features at approximately 528.7 and 531.3 eV. The spectra of the pure LaCoO3 samples can be deconvoluted into three peaks: that with the lowest binding energy (approximately 528.7 eV) is due to lattice oxygen atoms on the surface, the next one (approximately 530.9 eV) is ascribed to the hydroxyl oxygen, and the peak with the largest binding energy (approximately 532.3 eV) is ascribed to surface-adsorbed oxygen. After synthesizing LaCoO3 and the g-C3N4 complex, the adsorbed oxygen content was reduced to an almost undetectable level. The high-resolution O 1s spectrum of LaCoO3/g-C3N4 presents two primary features at 528.7 and 531.3 eV, which are attributed to lattice oxygen atoms on the surface and hydroxyl oxygen, respectively. It is worth mentioning that the hydroxyl oxygen bond peak position in the spectrum of LaCoO3/20 wt % g-C3N4 increases by 0.4 eV, which could indicate that the chemical environment had been changed after combining with LaCoO3. The typical high-resolution XPS La 3d spectra of the LaCoO3/(20 wt %) g-C3N4 sample show two shoulder peaks with shake-up features located at 830–840 and 850–857 eV, which could be deconvoluted into two noticeable shoulder peaks located at 833.7 and 837.1 eV and at 850.1 and 853.8 eV, respectively, which were assigned to the binding energies of La 3d5/2 and La 3d3/2, respectively, confirming the presence of La3+ in the crystal structure.[45−47] For the high-resolution Co 2p spectrum of the LaCoO3/g-C3N4-20 wt % sample (Figure f), there were two main peaks located at 780.2 and 795.3 eV, which were attributed to typical Co3+ with a typical shake-up structure at 787.2 eV.[45,48] In brief, the XPS results further demonstrated that the LaCoO3/g-C3N4 composite was successfully obtained and the main electron transfer route was from LaCoO3 to g-C3N4 and g-C3N4 linked with LaCoO3 cocatalysts via a chemically bound interface rather than a physical contact.

UV–Vis Diffuse Reflectance Spectra

As shown in Figure a, UV–vis DRS of g-C3N4, LaCoO3, and LaCoO3/20 wt % g-C3N4 composite were measured by UV–vis diffuse reflectance spectroscopy in the range of 220–800 nm. It was obvious that g-C3N4 had strong absorption to UV and its visible light spectrum absorption edge was at about 470 nm, which was in agreement with known reports. LaCoO3/g-C3N4 composite material has similar absorption characteristics as g-C3N4. For bare LaCoO3, it is worth noting that almost all ranges of light can be absorbed, which shows excellent photoelectric properties in LaCoO3. LaCoO3/g-C3N4 composites have similar absorption characteristics as g-C3N4. It is worth noting that compared with bare g-C3N4, the absorption edge of LaCoO3/g-C3N4 composite material shows a red shift, which indicates that LaCoO3/g-C3N4 composite material can absorb more visible light by moving to a lower-energy region. The band gap energies (Egs) of semiconductors are determined according to the Kubelka–Munk equation as followsHerein, α, h, ν, A, Eg, and n represent the absorption coefficient, Planck’s constant, incident light frequency, a constant, band gap energy, and an integer, respectively. The value of n mainly depends on the electronic transition structure of different semiconductors. If the band gap is a direct transition, then n = 1/2; if the band gap is an indirect transition, then n = 2. Figure b shows the calculated detailed band gap Eg values of g-C3N4, LaCoO3, and LaCoO3/g-C3N4-20 wt %, which are about 3.06, 2.84, and 2.66 eV, respectively. It can be seen that the Eg value obtained by the composite material is smaller. In addition, the VB and CB potentials of the semiconductor can also be obtained from the following empirical formulawhere ECB and EVB represent CB and VB edge potentials of the semiconductor, respectively, and χ is the electronegativity of the semiconductor. According to the refs (49, 50)., the χ values of bare LaCoO3 and g-C3N4 could be obtained, which are about 4.63 and 5.64 eV, respectively. Ee is the energy of free electrons with the hydrogen scale (∼4.5 eV vs normal hydrogen electrode (NHE)) and Eg is the band gap energy of the semiconductor.

Figure 7

(a) UV–vis DRS of LaCoO3, g-C3N4, and LaCoO3/g-C3N4 composites and (b) plots of the band gap energies (Egs) for g-C3N4 and LaCoO3/g-C3N4-20 wt %. (c) Plots of the band gap energies (Egs) for LaCoO3.

Photocurrent Response Test

The separation and migration efficiency of photogenerated charge carriers can be evaluated by photoluminescence (PL) spectroscopy. A higher PL intensity indicates higher photogenerated carrier recombination, and a lower intensity indicates lower recombination. The results of LaCoO3, g-C3N4, and LaCoO3/g-C3N4 composites with different weight ratios were excited by 300 nm light at room temperature, as shown in Figure a. It can be clearly seen that pure g-C3N4 has a strong emission peak, but pure LaCoO3 has almost no fluorescence response signal under the excitation of 300 nm wavelength light. The fluorescence response signal of the complex of LaCoO3 and g-C3N4 is significantly lower than that of pure g-C3N4, indicating that the recombination rate of the material carrier after the composite is reduced, and the fluorescence response signal of LaCoO3/20 wt % g-C3N4 is the lowest, which indicates that It has the lowest carrier recombination rate, which is very beneficial for photocatalytic hydrogen production. Moreover, the fluorescence band edge of LaCoO3/g-C3N4 composites blue-shifts compared with bare g-C3N4; it is evident that there existed an interaction between LaCoO3 and g-C3N4. In addition, compared with bare g-C3N4, the fluorescence band edge of LaCoO3/g-C3N4 composite materials undergoes a blue shift, indicating that there is an interaction between LaCoO3 and g-C3N4. The analysis of the photoluminescence spectrum revealed that the recombination of charge carriers in LaCoO3/g-C3N4 composite material is greatly suppressed, which is mainly due to the fact that the photogenerated electrons in the LaCoO3 conduction band tend to migrate down to the valence band in g-C3N4 after being irradiated by full-wavelength light, forming a Z-type heterojunction structure, thereby effectively preventing the direct recombination of electron–hole pairs.

Figure 8

(a) PL spectra of LaCoO3, g-C3N4, and LaCoO3/g-C3N4 composites and (b) photocurrent spectra of LaCoO3, g-C3N4, and LaCoO3/g-C3N4-20 wt %.

The photocurrent response spectrum of the prepared LaCoO3/g-C3N4-20% wt % composite was tested and compared with those of bare LaCoO3 and g-C3N4 to further clarify the separation and migration efficiency of photogenerated carriers. The result is shown in Figure b. It can be seen that the LaCoO3, g-C3N4, and LaCoO3/g-C3N4-20% wt % composite material keeps the photocurrent intensity constant when the lamp is turned on. When the lamp is turned off, the photocurrent intensity quickly decreases to zero, showing a fast and stable photocurrent response. Among them, LaCoO3/g-C3N4-20% wt % composite exhibits the strongest transient photocurrent response, with a photocurrent density of 0.017 mA cm–2, which is higher than that of LaCoO3 (0.015 mA cm–2) and g-C3N4 (0.012 mA cm–2). The photocurrent was formed due to the diffusion of photogenerated electrons to the back contact, while holes are absorbed by hole acceptors in the electrolyte. Therefore, the enhanced photocurrent response indicates that the LaCoO3/g-C3N4-20 wt % composite has a higher photogenerated electron–hole separation efficiency.

Photocatalytic Performance Evaluation

First, to evaluate the influence of light and the catalyst on the experiment, a blank control experiment was conducted in the absence of full-spectrum light or photocatalyst. Light-driven charge separation plays an important role in improving photocatalytic efficiency. Under these experimental conditions, the photolytic efficiency without a photocatalyst is negligible, which indicates that water containing 10% methanol is stable under visible light irradiation. To rule out the possibility of methanol reforming under the full spectrum, an anhydrous control experiment was carried out. Briefly, 30 mL of methanol and LaCoO3/20%g-C3N4 catalyst were added for photocatalysis experiments, and the results proved that there was no hydrogen generation. This indicates that the role of methanol in the experiment is that of a hole sacrificial agent, which consumes holes to promote photocatalytic hydrogen production. Then, the photocatalytic preparation of synthesized LaCoO3, g-C3N4, and LaCoO3/g-C3N4 nanoparticles with different mass ratios was tested in an aqueous solution containing 10% methanol by volume as a sacrificial reagent under full-spectrum (250–1200 nm) irradiation from a 300 W Xe lamp. Gas chromatography was used to detect the H2 emission, as shown in Figure a,b. We conducted three parallel tests on all samples and calculated the average H2 escape rate, which can eliminate unexpected factors and obtain more convincing data. Obviously, the produced LaCoO3, g-C3N4, and LaCoO3/g-C3N4 composite materials with different mass ratios have different levels of hydrogen production by photolysis. The H2 evolution rates for pure LaCoO3 and g-C3N4 were 298.17 and 732.65 μmol h–1 g–1, respectively. The LaCoO3/g-C3N4-20 wt % composite material had the largest photocatalytic activity, with an average hydrogen production per hour of 1046.15 μmol h–1 g–1, which is 3.5 and 1.4 times those of LaCoO3 and g-C3N4, respectively. As shown in Figure c, even after four consecutive cycles, the photocatalytic efficiency did not show a significant loss, which indicates that the heterojunction photocatalyst is highly stable during the photocatalytic hydrolysis process under full-spectrum irradiation of a xenon lamp.

Figure 9

(a) Photocatalytic H2 evolution rate as a function of irradiation time of LaCoO3, g-C3N4, and LaCoO3/g-C3N4 composites for water splitting under full-spectrum light irradiation. (b) Average hydrogen production rate of different materials per hour. (c) Long-term stability test of LaCoO3/20% g-C3N4 for 20 h.

Photocatalytic Mechanism Discussion

For a deeper understanding of the mechanism, the electron paramagnetic resonance (EPR) spectra are further adopted. As shown in Figure a, after 10 min of exposure to xenon lamp full-spectrum light, both the DMPO-radical •OH signals of LaCoO3 and LaCoO3/g-C3N4-20% can be observed, while no obvious DMPO-•OH signal is observed for pure g-C3N4. The absence of DMPO-•OH signal in g-C3N4 is because •OH cannot be produced by the holes in the VB of g-C3N4 (EVB = 1.67 eV, EOH–/θ = 1.99 eV vs NHE). The observation of DMPO-•OH signal of the LaCoO3/g-C3N4-20 wt % composite suggests that the photogenerated holes still stay in the VB of LaCoO3 and do not transfer to the VB of g-C3N4. In Figure b, after 10 min of exposure to xenon lamp full-spectrum light, BMPO-+O2– signals are observed for g-C3N4 and LaCoO3/g-C3N4-20 wt % composite samples in an aqueous suspension, whereas a very weak BMPO-+O2– signal is observed for LaCoO3. The results indicate that the photogenerated electrons in g-C3N4 and LaCoO3/g-C3N4-20 wt % composite samples have enough reduction ability to reduce O2 to form superoxide radical anions (+O2–) (EOθ•O2 = −0.33 eV vs NHE) (Figure ).

Figure 10

Spin-trapping EPR spectra recorded with bare g-C3N4, LaCoO3, and LaCoO3/g-C3N4-20 wt % composites in (a) aqueous dispersion (for BMPO-O2–) and (b) aqueous dispersion (for DMPO-OH) under xenon lamp full-spectrum light irradiation.

Figure 11

Z-scheme photocatalytic mechanism of LaCoO3/g-C3N4 heterojunction.

The EPR results show that photogenerated electrons and holes exist in the CB of g-C3N4 and VB of LaCoO3, respectively, and the charge transfer does not follow the conventional type II heterojunction mechanism. Therefore, it is reasonable to propose that the charge transfer path is a Z-scheme mechanism; both LaCoO3 and g-C3N4 as semiconductors have a forbidden bandwidth. When the photon energy is equal to or greater than the forbidden bandwidth, photogenerated electrons transition from the VB of g-C3N4 and LaCoO3 to the CB and generate holes in the valence band (eq ). Then, the electrons in LaCoO3 recombine rapidly with the holes in g-C3N4 through the heterostructure (eq ). Water and holes combine to generate hydrogen ions and oxygen (eq ), and hydrogen ions further combine with electrons in the conduction band of g-C3N4 to generate hydrogen (eq ). The specific process of photocatalysis is as follows

Conclusions

In summary, a highly efficient light-driven Z-type LaCoO3/g-C3N4 heterostructure nanocomposite with different weight contents of g-C3N4 has been successfully prepared by a solvothermal method and applied to photocatalytic hydrogen production. Compared with LaCoO3 and g-C3N4 alone under light irradiation, the obtained LaCoO3/g-C3N4 photocatalyst showed excellent hydrogen evolution efficiency. LaCoO3/g-C3N4-20 wt % exhibited the highest photocatalytic activity of 1046.15 μmol h–1 g–1, which is 3.5 and 1.4 times higher than those of LaCoO3 and g-C3N4, respectively. The enhanced photocatalytic activity was ascribed to the formation of a Z-scheme LaCoO3/g-C3N4 heterostructure, which possessed higher separation and transfer efficiencies of the photogenerated electron–hole pairs. Therefore, for LaCoO3/g-C3N4, effective photocatalytic water decomposition (UV–vis–near-infrared (NIR)) in the full spectrum can be achieved. This work leads to an inexpensive and effective LaCoO3/g-C3N4 photocatalytic system for water splitting or other photocatalytic applications.