Facile construction of CuFe₂O₄/g-C₃N₄ photocatalyst for enhanced visible-light hydrogen evolution

Abstract

A new type of CuFe₂O₄/g-C₃N₄ composite photocatalyst has been prepared by a facile one-pot calcination approach using urea and a CuFe₂O₄ gel as precursors. In this composite, CuFe₂O₄ has been finely dispersed in g-C₃N₄ matrix, resulting in a much enhanced efficiency of the CuFe₂O₄/g-C₃N₄ heterostructure in photocatalytic H₂ production by water splitting, under visible light. A peak H₂ production rate of 76 μmol h⁻¹ has been obtained, which is about 3 times that on pure g-C₃N₄ obtained from urea. This remarkable improvement can be attributed to the enhanced visible light absorbance, enlarged surface area and promoted charge carrier separation and transfer.

---

1. Introduction

Driven by increasingly severe energy demands and climate crises, researchers are now devoting themselves to the development of artificial photosynthesis, typically, water splitting and CO₂ reduction by sunlight, to generate chemical fuels (H₂, CH₄, and CH₃OH). The success of TiO₂ (ref. 1) in photocatalysis has served as a prelude to the photoelectrochemical splitting of water. Unfortunately, this classical photocatalyst with wide bandgap and low solar-energy utilization efficiency encounters a “bottleneck” to meet the requirements for practical applications. Recently, the development of photocatalysts has been stimulated and accelerated by various strategies, such as band engineering by heterogeneous doping, sensitization by incorporating efficient visible-light converters (e.g. metal–organic dyes), and invention/discovery of narrow band-gap semiconductors.^2,3 However, developing visible-light-driven photocatalysts with further enhanced solar-to-fuel conversion efficiency, much improved stability, and optimized cost-effectiveness for future applications is still a great challenge.

Sparked by the flourishing investigation on two-dimensional graphene-based materials, graphitic carbon nitride (g-C₃N₄) with π-conjugated structure was introduced as a metal-free photocatalyst for highly efficient hydrogen evolution under visible light.⁴ This easily prepared photocatalyst has good physicochemical stability, as well as appealing electronic structure combined with a medium band gap (2.7 eV). However, pure g-C₃N₄ suffers from shortcomings such as rapid recombination of photo-generated electron–hole pairs, small specific surface area and low visible light utilization efficiency.⁵ Doping,⁶ copolymerization,⁷ hetero-composition^8–10 and other strategies¹¹ have been exploited to optimize the electronic structure of g-C₃N₄ and enhance its photocatalytic activity and efficiency. Among all the strategies, g-C₃N₄ based nanocomposites show highly promising potential with many merits such as synergistic catalytic effect, strengthened light absorption, efficient charge separation and transportation.¹² A number of suitable candidates (MoO₃,¹³ ZnO,¹⁴ WO₃,¹⁵ and ironoxides^16,17) have been coupled with g-C₃N₄.

Spinel ferrites, MFe₂O₄ (M = Co, Ni, Mn, Cu, and Zn), have recently received increasing attention in catalysis, owing to their environmental compatibility and stability, as well as their easy synthesis and low cost. Among them, copper ferrite (CuFe₂O₄) has a unique valence shell electronic configuration of (3d¹⁰ 4s¹), is environmentally benign, and has therefore been used in various fields,¹⁸ especially in electronics,¹⁹ and catalysis.^20,21 However, suffering from low surface area (usually lower than 10 m² g⁻¹), agglomeration and irregular morphology, CuFe₂O₄, obtained by solid reaction, sol–gel combustion and post-precipitation methods, exhibits limited catalytic activity and applicability. Constructing heterogeneous nanocomposites is an effective strategy to improve its photocatalytic performance; e.g. CuFe₂O₄/TiO₂,²² CuFe₂O₄/CdS,²³ CuFe₂O₄/graphene oxide²⁴ and other composites have been reported. In 2009, Yang et al.²⁵ first reported the potential of CuFe₂O₄ in photocatalytic water splitting under visible light; however, subsequent studies on this topic have been rare because of the great difficulty in avoiding phase separation during its synthesis, and enlarging its surface area.

CuFe₂O₄ is a plausible candidate for coupling with g-C₃N₄ because of the perfect matching of their conduction band (CB) and valence band (VB) levels. However, in the known trials of MFe₂O₄/g-C₃N₄ (M = Co,²⁶ Ni,²⁷ Zn,²⁸ and Cu²⁹), g-C₃N₄ obtained from cyanamide or melamine with low surface area (below 10 m² g⁻¹) is normally used as the matrix to load large amounts of MFe₂O₄. To date, the photocatalytic water splitting ability of the CuFe₂O₄/g-C₃N₄ composite has not been reported. Herein, we developed a facile one-pot calcination method to construct the CuFe₂O₄/g-C₃N₄ composite, adding trace amounts of CuFe₂O₄ to modify g-C₃N₄ derived from urea. For the first time, we have found considerably enhanced photocatalytic activity, together with excellent stability of the CuFe₂O₄/g-C₃N₄ composite in hydrogen evolution by water splitting under visible light, with Pt as co-catalyst.

2. Experimental section

2.1 Catalyst synthesis

2.1.1 Chemicals and reagents. Urea (98%), copper(II) nitrate trihydrate (99%), and iron(III) nitrate nonahydrate (98%) were purchased from Sinopharm Chemical Reagent Co. Ltd.; L-glutamic acid (>99%) was purchased from TCI Co. Ltd.; triethanolamine (>97%) and chloroplatinic acid hexahydrate (>37.5%, Pt basis) were purchased from Sigma-Aldrich Co. LLC. All the chemicals were used as received.

2.1.2 Synthesis of the CuFe₂O₄ precursor. The CuFe₂O₄ precursor was prepared via a sol–gel method. In a typical procedure, a metal salt solution containing Fe(NO₃)₃·9H₂O and Cu(NO₃)₂·3H₂O was mixed with 0.3 M L-glutamic acid solution under vigorous magnetic stirring, in which the molar ratio of total metals (Cu²⁺ + Fe³⁺) and L-glutamic acid was 2 : 9. After reaction for 1 h at 80 °C, the gel was dried at 130 °C for 6 h. The resultant dark brown powder was collected through grinding, and named as CuFe.

2.1.3 Fabrication of the CuFe₂O₄/g-C₃N₄ composite. A one-pot calcination method has been developed here to synthesize the CuFe₂O₄/g-C₃N₄ composite. A typical procedure is as follows: 20 g urea and different amounts of CuFe as the CuFe₂O₄ precursor were mixed homogeneously by grinding in an agate mortar; the powder was then put into an alumina crucible and calcined for 2 h at 550 °C. The resultant material was denoted as xCuFe–CN (x represents the amount of CuFe₂O₄ precursor, and x = 1 mg, 3 mg, 5 mg, 10 mg, and 30 mg), whereas pure g-C₃N₄ was named as CN. To further confirm the chemical state of Cu and Fe in the composite, sample 200CuFe–CN with high CuFe₂O₄ content was prepared using 200 mg CuFe₂O₄ precursor and 20 g urea.

2.2 Characterization

X-ray diffraction measurements (XRD) were conducted using a Rigaku D/Max 2200PC diffractometer with Cu Kα radiation (λ = 1.5418 Å) in the 2θ range of 10–70°. Fourier transform infrared (FTIR) spectra were obtained on a Nicolet iS10 FTIR spectrometer. Inductively Coupled Plasma Emission Spectroscopy (ICP-AES) was performed using Thermofisher X series 2. Transmission electron microscopic (TEM) imaging was carried out using a JEOL 200CX electron microscope operated at 200 kV, together with the energy dispersive X-ray spectrum (EDS) obtained from an attached Oxford Link ISIS energy-dispersive spectrometer fixed on a JEM-2010 electron microscope. FE-SEM (Field Emission Scanning Electron Microscopy) images were obtained on a Magellan 400, together with the scanning transmission electron microscopy (STEM) images obtained on a JEOL-2010F electron microscope. Nitrogen adsorption–desorption isotherms at 77 K were recorded on a Micromeritics TriStar 3000 instrument. X-ray photoelectron spectroscopy (XPS) was collected on a Thermo Scientific ESCALAB 250 spectrometer with Al Kα radiation as the excitation source. Binding energies for the high-resolution spectra were calibrated by setting C 1s to 284.8 eV. The UV-Vis diffuse reflectance absorption spectra were obtained, using a UV-3600 Shimadzu spectroscope. Photoluminescence spectra (PL) were obtained using a fluorescence spectrometer (Shimadzu RF-5301PC) at room temperature, with excitation by incident light of 380 nm.

2.3 Electrochemical analysis

Electrochemical measurements were conducted on a CHI 760 electrochemical workstation in a standard three-electrode cell, using a Pt plate and an Ag/AgCl electrode as counter electrode and reference electrode, respectively. Working electrodes were obtained by electrophoretic deposition of photocatalyst suspensions (20 mg sample powder and 40 mg iodine in 40 mL acetone solution) onto FTO conductive glass with a fixed area of ca. 1 cm². Two parallel FTO glasses were placed in the abovementioned solution and kept for 5 min with a 10 V bias under potentiostatic control. All the working electrodes were then annealed at 150 °C for 1 h. A 0.2 M Na₂SO₄ (50 mL) aqueous solution was chosen as the supporting electrolyte. The visible light was generated by a 300 W Xe lamp with a 420 nm cut-off filter, which was chopped manually. The bias voltage for obtaining Nyquist plots of these electrodes was set at −0.4 V versus Ag/AgCl electrode in the dark.

2.4 Photocatalytic H₂ evolution

The visible-light photocatalytic H₂ evolution reactions were carried out in a quartz reaction vessel connected to a closed gas circulation and evacuation system, using a 300 W Xe lamp, equipped with an optical UV-IR cut-off filter (780 nm > λ > 420 nm), as the visible light source. The photocatalyst (0.1 g) was loaded with 3 wt% Pt by an in situ photodeposition method using H₂PtCl₆ as the precursor, and suspended in an aqueous solution (100 mL) containing triethanolamine (TEA, 10 vol%) as the sacrificial electron donor. The reactant solution was then thoroughly degassed under flowing N₂ for 15 min, prior to the irradiation H₂-evolution experiment. The reaction's solution temperature was maintained at 20 °C during the photocatalytic reaction, by a flow of cooling water. The evolved gas was analyzed every 1 h by gas chromatography, along with a thermal conductive detector (TCD), using nitrogen as the carrier gas. After reaction for 3 h, the system was evacuated again. For stability testing, a 16 h recycling experiment was performed with intermittent evacuation every 4 h. The apparent quantum yield (AQY) for H₂ evolution was measured under the same photoreaction conditions but with a 420 nm band-pass filter, using a Coherent FieldMaxII-TO spectroradiometer to measure the average intensity of irradiation. The AQY was calculated as follows:

AQY (%) = number of evolved H₂ molecules × 2 × 100/number of incident photons

3. Results and discussion

3.1 Catalyst characterizations

ICP analysis was first performed to assay the as-prepared materials. The results confirmed the successful introduction of CuFe₂O₄ into all the composite samples. The corresponding weight fraction of CuFe₂O₄ in all samples was calculated based on the Fe content, as shown in Table 1. The net amount of g-C₃N₄ in the xCuFe–CN composites obtained from 20 g urea and x mg CuFe precursor was calculated by the yield and weight fraction of g-C₃N₄ (Table S1†), and the actual molar ratio of Cu/Fe was determined to be ca. 1 : 2 (Table S1†).

Table 1 CuFe₂O₄ content and surface area of the as-prepared CN and xCuFe–CN samples

Sample	CN	1CuFe–CN	3CuFe–CN	5CuFe–CN	10CuFe–CN	30CuFe–CN
CuFe2O4/wt%	—	0.11	0.41	0.49	0.96	2.65
SBET/m^2 g^−1	81	100	117	115	99	95

---

In Fig. 1a, the XRD patterns of all samples maintain the typical graphite-like structure of g-C₃N₄. The distinct peak at 27.48° belongs to the characteristic (002) interlayer-stacking reflection of aromatic systems, whereas the other pronounced peak at 12.78° originating from the (100) plane represents the in-plane structural packing motif.³⁰ No peaks of ferric species can be identified. This reveals that CuFe₂O₄ loading does not change the bulk structure and chemical skeleton of g-C₃N₄. The location of the (002) peak up-shifts from 27.48° (CN) to 27.58° (30CuFe–CN) (Fig. S1a†), indicating the decrease of the corresponding interlayer distance from 0.324 to 0.323 nm, i.e., the gradual structural compaction and thickening of the g-C₃N₄. For gel CuFe, typical diffraction peaks at 18.36°, 29.98°, 34.69°, 35.66°, 37.21°, 43.84°, 58.14° and 62.33° could be indexed to the (101), (112), (103), (211), (202), (220), (321) and (224) planes of the tetragonal spinel structure of CuFe₂O₄ (JCPDS no.34-0425) (Fig. 1b), respectively. As can be observed in Fig. S1b,† with a large dosage of CuFe, sample 200CuFe–CN shows detectable diffraction signals of CuFe₂O₄ in the composite.

Fig. 1 XRD patterns of (a) CN and xCuFe–CN samples, and (b) CuFe (inserted lines correspond to typical diffraction peaks of tetragonal spinel structure CuFe₂O₄, JCPDS no.34-0425).

XPS was performed to analyze the surface chemical nature of the as-synthesized composite. As shown in Fig. 2a, aside from the three classic signals belonging to C, N, and O of pure carbon nitride, Cu and Fe can also be found in the sample 3CuFe–CN (Fig. 2a), which further confirms the incorporation of CuFe₂O₄ into g-C₃N₄. The corresponding high-resolution spectra of C 1s, N 1s, O 1s, Cu 2p and Fe 2p are given. The C 1s spectra can be fitted with two peaks (Fig. 2d). The dominating peak at 288.1 eV is ascribed to sp²-bonded carbon (N–C=N) in N-containing aromatic rings, whereas the peak at 284.8 eV originates from sp² C–C bonds in carbon species. The N 1s spectra (Fig. 2e) can be deconvoluted into four sub-bands centered at 398.6, 399.9, 401.2 and 404.2 eV, which are attributed to sp² nitrogen (C=N–C) involved in triazine rings, the tertiary nitrogen bonded to carbon atoms in the form of N–(C)₃, the amino functional groups (C–N–H), and the charging effects or positive charge localization in heterocycles, respectively. In the O 1s spectrum of 3CuFe–CN (Fig. 2f), three peaks at 531.5, 532.2 and 533.9 eV correspond to OH⁻, adsorbed H₂O and oxygen species,³¹ which can also be found in the O 1s spectrum of pure g-C₃N₄. In addition, the peak at 530.0 eV belongs to the spinel structure of CuFe₂O₄.³² For the trace amounts of Fe and Cu in the sample 3CuFe–CN, the Fe 2p and Cu 2p peaks are very weak, and can hardly be deconvoluted (Fig. 2b and c).

Fig. 2 XPS spectra of CN and 3CuFe–CN: (a) survey, (b) Fe 2p, (c) Cu 2p, (d) C 1s, (e) N 1s and (f) O 1s.

The XPS spectra of the CuFe₂O₄ precursor (CuFe) and 200CuFe–CN are shown in Fig. S2.† The Fe 2p_3/2 spectra of both samples possess two characteristic peaks at about 710.6 eV and 712.5 eV, corresponding to Fe²⁺ and Fe³⁺.³¹ The Cu 2p_3/2 spectrum of 200CuFe–CN has two peaks at 934 and 932.8 eV, which could be attributed to Cu²⁺ and Cu⁺, respectively. For CuFe, its Cu 2p_3/2 spectrum only has a main peak of Cu²⁺, whereas the 200CuFe–CN Cu 2p_3/2 spectrum shows both Cu²⁺ and Cu⁺ peaks, indicating the obvious transformation of Cu²⁺ into Cu⁺ during the formation of the CuFe₂O₄/g-C₃N₄ composite.²⁹ This surface-rich Cu⁺ may benefit the electron transfer and the stable hydrogen production activity under light irradiation.^33,34

Both CN and 3CuFe–CN samples exhibit layered and curved morphologies, as shown in Fig. 3 (TEM) and S3† (SEM). For 3CuFe–CN, platelet-like fragments can be observed, and EDS analysis shows no obvious aggregation of CuFe₂O₄, indicating its high dispersion in the g-C₃N₄ matrix. CuFe₂O₄ precursor particles (Fig. 3a) are irregular in shape, agglomerated and well crystallized as proved by XRD. STEM and HRTEM analyses were conducted on 200CuFe–CN. As can be observed in Fig. S4,† fine CuFe₂O₄ species with particle sizes of about 15–25 nm are uniformly dispersed on the g-C₃N₄ layer.

Fig. 3 Typical TEM images of (a) CuFe, (b) CN, (c and d) 3CuFe–CN.

The FT-IR spectra obtained in the range of 400–4000 cm⁻¹ show that the composite basically maintains the classic carbon nitride structure. As displayed in Fig. 4a, a series of strong bands at 1200–1650 cm⁻¹ correspond to the stretching modes of aromatic C–N heterocycles. The most intense band located at 809 cm⁻¹ represents the out-of-plane bending vibration of tri-s-triazine rings and the broad bands at about 3000–3500 cm⁻¹ originate from the adsorbed H₂O molecules and the terminal amino groups in the g-C₃N₄ structure. For CuFe, the absorption bands in the region 400–600 cm⁻¹ indicate the presence of copper ferrites (Fig. 4b).³⁵ The band at ∼566 cm⁻¹ can be attributed to intrinsic stretching vibrations of Fe cations on tetrahedral sites, whereas the band at ∼430 cm⁻¹ refers to Cu cations on octahedral sites. The bands at 3437 cm⁻¹ and 1630 cm⁻¹ are related to the adsorbed H₂O molecules. The bands at 1465 cm⁻¹, 1350 cm⁻¹ and 1049 cm⁻¹ are ascribed to CO₃²⁻ and HCO₃⁻ species, characteristic of the surface complexed carbonates.³⁶ For 10CuFe–CN and 30CuFe–CN samples, the typical signals of CuFe₂O₄ can be detected, further reflecting the existence of CuFe₂O₄ species in the composite.

Fig. 4 (a) FT-IR spectra and (b) the corresponding high-resolution FT-IR spectra of CN and 3CuFe–CN samples.

3.2 Photocatalytic mechanism on the as-synthesized CuFe₂O₄/g-C₃N₄ composite

The photoreduction activities of the obtained CuFe₂O₄/g-C₃N₄ nanocomposite and pure g-C₃N₄ are presented in Fig. 5a and S5.† While pure CuFe₂O₄ is inactive in photocatalytic water reduction, the composite materials show enhanced hydrogen evolution performance, and 3CuFe–CN shows the best photoreduction performance with a H₂ evolution rate of 76 μmol h⁻¹, which is about 3 times that of the pure CN obtained from urea. The corresponding apparent quantum yield (AQY) value on 3CuFe–CN at 420 nm has been calculated to be 3.3%, which is much higher than that (1.3%) on CN.³⁷ Furthermore, the 3CuFe–CN composite also has excellent stability in H₂ evolution under irradiation, and there is no significant deactivation during 4 consecutive runs (16 h) (Fig. 5b).

Fig. 5 (a) Photocatalytic H₂ evolution on pure CN and xCuFe–CN samples under visible light irradiation (λ > 420 nm). (b) Stability test of H₂ evolution on 3CuFe–CN.

It is acknowledged that the surface area, energy levels of photo-induced electrons, bandgap width of the photocatalyst and separation efficiency of charge carriers, can all determine the hydrogen evolution activity of semiconductor photocatalysts. For this CuFe₂O₄/g-C₃N₄ nanocomposite, the incorporation of CuFe₂O₄ has greatly increased the surface area of g-C₃N₄, though the surface area of the CuFe₂O₄ precursor is only 8 m² g⁻¹, as can be observed in Table 1 and indicated by the N₂ adsorption–desorption isotherms (Fig. S6†). This can be attributed to the gaseous CO₂ and H₂O generated and released from the decomposition of the CuFe precursor during the polymerization of urea. The high surface area of this composite means an increased amount of catalytically active sites and consequently, enhanced photocatalytic activity. However, when more than 5 mg of CuFe precursor is dosed in 20 g urea, the surface area of the composite decreases due to the increased weight fraction and enlarged particle size of CuFe₂O₄, resulting in the decreased photocatalytic efficiency of the composite.

The UV-Vis diffuse reflectance absorption spectra shown in Fig. 6a demonstrate that all composite samples exhibit significantly improved light harvesting ability above 450 nm, compared to the pure CN sample. The results also reflect the gradual decrease of their bandgap from 2.89 eV (CN) to 2.82 eV (30CuFe–CN) (Table S2†); the bandgap of CuFe is 1.42 eV. The intrinsic HOMO energy levels of CN and 3CuFe–CN were estimated to be 2.46 eV and 2.25 eV, respectively (Fig. S7c†).³⁸ Considering the bandgap of CN (2.89 eV) and 3CuFe–CN (2.86 eV), the LUMO energy level of 3CuFe–CN is about 0.18 eV higher than that of CN. Therefore, the photo-induced electrons located at the intrinsic LUMO level of 3CuFe–CN have much stronger reducibility than those of CN, which is greatly helpful to improve the photocatalytic performance. Photoluminescence (PL) spectra were obtained to further understand the photophysical behavior of photoexcited charge carriers in CN and xCuFe–CN samples (Fig. 6b). Both pure and composited g-C₃N₄ samples possess the representative dual-band PL spectra. Typically, the band at shorter wavelength (∼440 nm) is ascribed to the intrinsic LUMO to HOMO emission of g-C₃N₄, corresponding to its bandgap, whereas the other shoulder band at ∼463 nm results from charge transfer emission.³⁸ The intensity of PL bands decreases significantly after the incorporation of CuFe₂O₄, suggesting the decreased luminous recombination probability of photo-induced charge carriers; i.e., the improved charge separation and more charge carriers being involved in photocatalytic reactions.³⁹

Fig. 6 (a) UV-Vis diffuse reflectance absorption spectra, (b) PL spectra (excitation wavelength: 380 nm) of CN and xCuFe–CN samples.

Identified by EIS spectra (Fig. 7a), the obvious radius decrease of the semicircular Nyquist plots of 3CuFe–CN in the dark indicates the accelerated interface transport of charge carriers in the composite and the improved re-localization of electrons on the surface terminal sites, which is believed to be beneficial for elevating its photocatalytic reactivity.⁴⁰ Compared with CN, 3CuFe–CN shows a remarkable enhancement in the photocurrent intensity (I_ph) (Fig. 7b), which indicates the greatly improved transfer ability of the charge carriers in CuFe₂O₄/g-C₃N₄.

Fig. 7 (a) Electrochemical impedance spectroscopy (EIS) Nyquist plots in the dark, and (b) periodic on/off photocurrent response under visible-light irradiation of CN and 3CuFe–CN.

The proposed photocatalytic mechanism on the as-prepared CuFe₂O₄/g-C₃N₄ composite is shown in Scheme 1. As reported in most of the literature about g-C₃N₄, Pt, a good hydrogen evolution co-catalyst that could trap photogenerated electrons in semiconductors, was also loaded onto g-C₃N₄ and the CuFe₂O₄/g-C₃N₄ composite. More importantly, in this heterogeneous composite system, CuFe₂O₄ and g-C₃N₄ are proposed to be uniformly in contact with each other, which may lead to significant synergetic catalytic effects in the photocatalytic reaction.¹²

Scheme 1 Schematic of the proposed photocatalytic mechanism on the as-prepared CuFe₂O₄/g-C₃N₄ composite.

Under visible light irradiation, electron and hole pairs are generated and charge transfer occurs. The E_CB of CuFe₂O₄ (−1.21 eV) is more negative than that of g-C₃N₄ (−0.43 eV) and the E_VB of CuFe₂O₄ (+0.21 eV) is more positive than that of g-C₃N₄ (+2.46 eV).²² Therefore, the photogenerated electrons in CuFe₂O₄ have a tendency to diffuse to the CB of g-C₃N₄ via the interface, whereas the photogenerated holes in g-C₃N₄ have an opposite transfer direction, to be injected into the VB of CuFe₂O₄. Typically, the electrons on the CB of g-C₃N₄ will transfer to Pt nanoparticles, wherein the activated H₂O can be reduced to H₂; the holes on the VB of CuFe₂O₄ will migrate to TEA and oxidize it to TEA⁺. Compared with the recently reported MgFe₂O₄/g-C₃N₄ composite⁴¹ used in photocatalytic hydrogen generation, which unavoidably has the issue of MgFe₂O₄ being the recombination center of charge carriers, herein the slight amount and tiny CuFe₂O₄ particles highly dispersed in the composite offer much more rapid separation paths for charge carriers. This composite system with CuFe₂O₄ and Pt as hole and electron traps, respectively, can greatly improve the separation, transfer and surface reactivity of photogenerated charge carriers, thus remarkably enhancing the photocatalytic activity of g-C₃N₄.

4. Conclusions

We have successfully constructed a new type of CuFe₂O₄/g-C₃N₄ composite photocatalyst by a simple one-pot method using coppery-ferric gel and urea as the precursors of CuFe₂O₄ and g-C₃N₄, respectively. Slight amounts of CuFe₂O₄ species can be highly dispersed in the g-C₃N₄ matrix, which significantly elevates the visible light absorbance, surface area, and separation/transfer ability of the charge carriers of g-C₃N₄, thus remarkably enhancing its H₂ evolution performance under visible light. The H₂ evolution rate on CuFe₂O₄/g-C₃N₄ reaches 76 μmol h⁻¹, which is about 3 times that on the pure g-C₃N₄ obtained from urea. The perfect matching of the energy levels between g-C₃N₄ and CuFe₂O₄, together with the high dispersion of slight amounts of uniformly dispersed, tiny CuFe₂O₄ nanoparticles in the composite, enables effective and rapid separation of charge carriers. In this heterogeneous composite, CuFe₂O₄ acts as a hole trap during the photocatalytic reaction. This study demonstrates a new cost-effective approach for constructing heterogeneous composite catalysts of spinel ferrites and g-C₃N₄ for hydrogen evolution under visible light. The synergetic catalysis effect of the composite may provide a new route for designing promising photocatalysts for solar fuel production.

Acknowledgements

This study was financially supported by the National Key Basic Research Program of China (2013CB933200), the National 863 plans projects (2012AA062703), the National Natural Science Foundation of China (21177137), and Youth Innovation Promotion Association of CAS (2012200).

Notes and references

1. A. Fujishima and K. Honda, Nature, 1972, 238, 474.
2. G. Zhang, G. Kim and W. Choi, Energy Environ. Sci., 2014, 7, 954.
3. J. Ran, J. Zhang, J. Yu and S. Z. Qiao, Chem. Soc. Rev., 2014, 43, 7787.
4. X. Wang, K. Maeda and M. Antonietti, Nat. Mater., 2009, 8, 76.
5. X. Wang, S. Blechert and M. Antonietti, ACS Catal., 2012, 2, 1596.
6. X. Wang, A. Thomas and M. Antonietti, Adv. Mater., 2009, 21, 1609.
7. J. Zhang, X. Chen and K. Takanabe, Angew. Chem., Int. Ed. Engl., 2010, 49, 441.
8. S. Nayak, L. Mohapatra and K. Parida, J. Mater. Chem. A, 2015, 3, 18622.
9. S. Pany and K. Parida, Phys. Chem. Chem. Phys., 2015, 17, 8070.
10. S. Samanta, S. Martha and K. Parida, ChemCatChem, 2014, 6, 1453.
11. S. Martha, A. Nashim and K. Parida, J. Mater. Chem. A, 2013, 1, 7816.
12. J. Shi, Chem. Rev., 2013, 113, 2139.
13. L. Huang, H. Xu and H. Li, Appl. Surf. Sci., 2013, 283, 25.
14. W. Liu, M. Wang and S. Chen, Chem. Eng. J., 2012, 209, 386.
15. Z. Jin, N. Murakami and T. Ohno, Appl. Catal., B, 2014, 150–151, 479.
16. K. Sivula, F. Formal and M. Grätzel, ChemSusChem, 2011, 4, 432.
17. J. Xue, S. Ma and X. Liu, RSC Adv., 2015, 5, 58738.
18. D. Wang and D. Astruc, Chem. Rev., 2014, 114, 6949.
19. W. Zhang, B. Quan and Y. Piao, ACS Appl. Mater. Interfaces, 2015, 7, 2404.
20. P. Jing, J. Li and Q. Liu, J. Hazard. Mater., 2015, 284, 163.
21. B. Li, M. Li and D. Zhao, J. Mater. Chem. A, 2013, 1, 6742.
22. A. Kezzim, N. Nasrallah and M. Trari, Energy Convers. Manage., 2011, 52, 2800.
23. N. Nasrallah, M. Kebir and Z. Koudri, J. Hazard. Mater., 2011, 185, 1398.
24. Y. Fu, Q. Chen and X. Wang, Ind. Eng. Chem. Res., 2012, 51, 11700.
25. H. Yang, J. Yan and Z. Lu, J. Alloys Compd., 2009, 476, 715.
26. S. Huang, Y. Xu and H. Li, Colloids Surf. A, 2015, 478, 71.
27. H. Ji, X. Jing and H. Li, RSC Adv., 2015, 5, 57960.
28. S. Zhang, J. Li and X. Wang, ACS Appl. Mater. Interfaces, 2013, 5, 12735.
29. Y. Yao, F. Lu and S. Wang, J. Hazard. Mater., 2015, 297, 224.
30. H. Xu, J. Yan and H. Li, Nanoscale, 2014, 6, 1406.
31. T. Lin, G. Seshadri and J. Kelber, Appl. Surf. Sci., 1997, 119, 83.
32. E. Kester and B. Gillot, J. Phys. Chem. Solids, 1998, 59, 1259.
33. K. Lalitha, G. Sadanandam, V. Kumari and N. Hebalkar, J. Phys. Chem. C, 2010, 114, 22181.
34. L. Clarizia, D. Spasiano and D. Dionysiou, Int. J. Hydrogen Energy, 2014, 39, 16812.
35. Y. Wang, H. Zhao and G. Zhao, Appl. Catal., B, 2014, 147, 534.
36. K. Zaharieva, V. Rives and M. Milanova, Mater. Chem. Phys., 2015, 15, 271.
37. X. Zhang, L. Yu and X. Li, ACS Catal., 2014, 4, 162.
38. X. Fan, R. Cheng and J. Shi, ACS Catal., 2015, 5, 5008.
39. M. Li, L. Zhang and J. Shi, J. Mater. Chem. A, 2015, 3, 5189.
40. V. Etacheri, M. Seery and S. Pillai, Chem. Mater., 2010, 22, 3843.
41. J. Chen, D. Zhao and S. Shen, ACS Appl. Mater. Interfaces, 2015, 7, 18843.

---

Footnote

† Electronic supplementary information (ESI) available: Analytical data. See DOI: 10.1039/c5ra27221a

---