Nitrogen-doped CeO_x nanoparticles modified graphitic carbon nitride for enhanced photocatalytic hydrogen production

Abstract

Nitrogen-doped CeO_x nanoparticles (N-CeO_x NPs) were directly formed on graphitic carbon nitride (g-C₃N₄) via a facile one-pot method by annealing a mixture of Ce(NO₃)₃ and melamine as CeO_x and g-C₃N₄ precursors, respectively. Nitrogen was in situ doped into CeO_x under NH₃ atmosphere released by the decomposition of melamine during the annealing process. The physical and photophysical properties of N-CeO_x NPs modified g-C₃N₄ photocatalysts were investigated to reveal the effects of N-CeO_x NPs on the photocatalytic activities of g-C₃N₄. It was found that the one-pot annealing method was favorable for forming intimate interfacial contact between N-CeO_x and g-C₃N₄. The visible light photocatalytic hydrogen production activity over g-C₃N₄ was enhanced by ca. 1.2 times with the N-CeO_x NPs modification. The significant enhancement in photocatalytic performance for the N-CeO_x/g-C₃N₄ heterojunction should be mainly because of the promoted charge transfer and charge separation between N-CeO_x NPs and g-C₃N₄ resulting from the intimate interfacial contact and Type II band alignment. In addition, the improved visible light absorption of N-CeO_x NPs induced by nitrogen doping could be another reason for the enhanced photocatalytic activity of the N-CeO_x/g-C₃N₄ heterojunction.

---

Introduction

Photocatalytic water splitting to produce hydrogen over semiconductors by solar irradiation has received considerable attention due to its potential applications to alleviate the global energy crisis and environmental pollution. The development of efficient photocatalysts utilizing visible light (∼43% of the solar spectrum) is of vital importance to fulfill the practical applications of this technique. H₂ production under visible light irradiation has been demonstrated with a myriad of semiconductor photocatalysts,^1–3 including oxides, sulfides, and oxynitrides etc. However, some intrinsic drawbacks, such as poor optical absorption ability (e.g., TiO₂,⁴ ZnO⁵) for wide band gap oxides, poor photochemical stability for sulfides (e.g., CdS⁶), and complicated preparation processes for oxynitrides (e.g., GaN : ZnO⁷), still significantly limit their practical application for photocatalytic solar-hydrogen conversion. To address these limitations, novel photocatalytic materials with efficient visible light absorption and good photochemical stability, as well as facile preparation methods must be developed for practical application of this state-of-the-art technique.

Since the first demonstration by Wang et al.,⁸ the novel metal-free polymer n-type semiconductor prepared by simply annealing cheap and abundant precursors (urea, melamine, dicyandiamide, and cyanamide), namely layered carbon nitride with a graphitic structure (g-C₃N₄), has attracted extensive interest because of its proper band gap (E_g ∼2.7 eV) and excellent photocatalytic stability.⁹ However, because of poor optical absorption above 460 nm and the fast recombination of photogenerated electron–hole pairs, the photocatalytic efficiency of pure g-C₃N₄ is relatively low. To enhance its photocatalytic performance, different methods have been exploited, including creating porous structures,¹⁰ doping with metal¹¹ or non-metal ions,¹² combining with other semiconductors¹³ and loading co-catalysts.¹⁴ Among them, combining g-C₃N₄ with other semiconductors to form semiconductor/g-C₃N₄ composite photocatalysts has been the most widely studied method. Several nanocomposites such as BiOBr/g-C₃N₄,¹³ TiO₂/g-C₃N₄,¹⁵ ZnO/g-C₃N₄,¹⁶ Fe₂O₃/g-C₃N₄,¹⁷ and Ag₂O/g-C₃N₄ ¹⁸ have been developed for improved charge separation and/or enhanced visible light absorption, leading to higher photocatalytic activities compared to pure g-C₃N₄.

As a reactive rare earth oxide with outstanding catalytic properties, CeO_x has been widely used in the solar cells, photocatalytic degradation of dye pollutants and hydrogen evolution.^19,20 To date, various types of heterojunctions formed between CeO_x and other semiconductors (e.g., TiO₂,²¹ Cu₂O,²² BiVO₄,²³ and CdS,.²⁴) have been developed with enhanced photocatalytic performances because of the improved charge separation ability. However, in these CeO_x/semiconductor composite photocatalysts, CeO_x can only act as a Type II band alignment counterpart to improve the charge separation, but not as a visible light sensitizer because of its large band gap (E_g ∼2.9 eV) limiting its optical absorption in the visible light region. Nitrogen doping is a widely accepted approach to enhance the visible light response of CeO_x as well as other wide band gap metal oxides.^4,25,26 Mao et al.²⁵ synthesized nitrogen doped CeO_x (N-CeO_x) through a wet-chemical route; the optical property of CeO_x was considerably improved and the visible-light photocatalytic performance for methylene blue decomposition was enhanced. Jorge et al.²⁶ prepared porous nitrogen doped CeO_x with effective visible light-active performance for the decomposition of acetaldehyde by annealing mesoporous nanocrystalline ceria under NH₃ at 550–700 °C. It should be noted that g-C₃N₄ can be easily synthesized from various simple precursors via a polycondensation reaction at 500–600 °C with the release of NH₃.^27,28 It becomes highly possible to dope nitrogen into the lattice of the semiconductors to form N-semiconductor/g-C₃N₄ heterojunctions by heating the mixture precursors of semiconductors and g-C₃N₄.^29–31 However, as far as we know, N-CeO_x/semiconductor composite photocatalysts with N-CeO_x acting as an effective visible-light sensitizer have not yet been reported.

In the present study, we have successfully prepared a series of nitrogen doped CeO_x nanoparticles (N-CeO_x NPs) modified g-C₃N₄ photocatalysts via a facile one-pot method by simply heating a mixture of Ce(NO₃)₃ and melamine as CeO_x and g-C₃N₄ precursors at 550 °C. The effect of N-CeO_x NPs on the optical absorption and charge carrier separation, as well as the enhanced photocatalytic activities of g-C₃N₄, was thoroughly investigated.

Experimental details

Synthesis procedure

All chemicals in the present study were of analytical grade and used as received without further purification. N-CeO_x NPs modified g-C₃N₄ photocatalysts were prepared via a facile one-pot method as illustrated in Fig. S1.† Typically, 3.0 g of melamine and V mL (V = 1, 3, 5, 7, 10, and 15 mL) of 0.031 g mL⁻¹ Ce(NO₃)₃·6H₂O aqueous solution were added to 200 mL of distilled water and heated at 90 °C for 1 h with stirring; the temperature was then increased to 100 °C for complete water evaporation. The resulting mixture was placed in an alumina crucible with a cover, and annealed at 550 °C under Ar flow (30 mL min⁻¹) for 4 h with a ramping rate of 20 °C min⁻¹. g-C₃N₄ modified with the different amounts of N-CeO_x NPs by adding V mL of 0.031 g mL⁻¹ Ce(NO₃)₃·6H₂O aqueous solution is denoted as Ce-CN-V (V = 1, 3, 5, 7, 10, 15 mL).

Pure g-C₃N₄ and CeO_x were prepared by the same process without adding Ce(NO₃)₃ or melamine. Ce-CN-M composite as reference was prepared by the mechanical mixing of CeO_x and pure g-C₃N₄. Ce-CN-T composite was also prepared as the reference by a two-step method, namely, annealing the grounded mixture of melamine and pure CeO_x following the same heating process as described above. To be consistent with the loading amount of CeO_x in Ce-CN-5, CeO_x used for preparing Ce-CN-M and Ce-CN-T was obtained by heating and then annealing 5 mL of 0.031 g mL⁻¹ Ce(NO₃)₃·6H₂O aqueous solution at 550 °C. In addition, the reference samples of Ce-CN-M₂ and Ce-CN-T₂ were also prepared following the same preparation procedure as that for Ce-CN-M and Ce-CN-T, respectively, using 15 mL of Ce(NO₃)₃·6H₂O aqueous solution as the CeO_x precursors to be consistent with the loading amount of CeO_x in Ce-CN-15.

Characterization

The transmission electron microscopy (TEM) images were obtained from a FEI Tecnai G2 F30 transmission electron microscope at an accelerating voltage of 300 kV. The X-ray diffraction (XRD) patterns were obtained from a PANalytical X'pert MPD Pro diffractometer operated at 40 kV and 40 mA using Ni-filtered Cu-Kα irradiation (Wavelength 1.5406 Å). UV-vis absorption spectra (UV-vis) were measured on a HITACHI U4100 instrument equipped with a labsphere diffuse reflectance accessory. Photoluminescence spectra (PL, including steady and time-resolved fluorescence emission) were carried out at room temperature on a PTI QM-4 fluorescence spectrophotometer. X-ray photoelectron spectroscopy (XPS) data were obtained on a Kratos Axis-Ultra DLD instrument with a monochromatized Al Kα line source (150 W). All the binding energies were referenced to the C 1s peak at 284.8 eV.

Photocatalytic activity evaluation

Photocatalytic hydrogen evolution was performed in a gas-closed circulation system with a 230 mL Pyrex cell as the photoreactor. A 300 W Xe lamp was used as the light source, and the UV part of the irradiated light was removed by a 420 nm cut-off filter. The evolved hydrogen was analyzed on a thermal conductivity detector (TCD) gas chromatograph (NaX zeolite column, N₂ as a carrier gas) every 60 min. In a typical experiment, the photocatalyst powder (50 mg) was dispersed in 200 mL of 10 vol% triethanolamine (TEOA) aqueous solution with stirring. Pt (1 wt%) was photodeposited in situ on the photocatalysts from the precursor of H₂PtCl₆·6H₂O. Nitrogen was purged through the reactor for 15 min before the photocatalytic reaction to remove oxygen. The temperature for all the photocatalytic reactions was kept at 35 ± 0.5 °C by thermostatic water bath.

Results and discussion

Fig. 1 shows the TEM images of the as-prepared Ce-CN-V (V = 1, 5, 10) photocatalysts. g-C₃N₄ photocatalyst exhibits a layered structure, as shown in Fig. 1A–C, which offered a good substrate for loading N-CeO_x NPs. As demonstrated in Fig. 1A, for the Ce-CN-1 photocatalyst, N-CeO_x NPs were evenly dispersed on g-C₃N₄ with a size range of 7–17 nm (See Fig. S2 in ESI† for the particle size distribution). When the volume of the Ce(NO₃)₃ precursor solution was increased to 5 mL, more N-CeO_x NPs were found on the surface of g-C₃N₄ (Fig. 1B). Despite aggregation, no bulk CeO_x was found on g-C₃N₄, and the particle size distribution of Ce-CN-5 was almost the same as that of Ce-CN-1, as shown in Fig. S2.† Further increasing the volume of Ce(NO₃)₃ precursor solution gave rise to more aggregates of N-CeO_x NPs, covering larger areas of the g-C₃N₄ surface, as observed for Ce-CN-10 (Fig. 1C). It was reported that the shading effect induced by nanoparticle aggregation was detrimental for the photocatalytic hydrogen production reaction due to covering of the active species on the surface of the photocatalyst.^32,33 As shown in Fig. 1D, the fringe spacings of N-CeO_x in Ce-CN-5 were measured to be 3.18 Å and 1.96 Å, corresponding to the (111) and (220) lattice planes of cubic CeO_x, respectively, which are slightly larger than the reported values of the (111) and (220) lattice planes of cubic CeO_x (3.12 Å and 1.91 Å, JCPDS 004-0593). This phenomenon is very likely because nitrogen was introduced into the crystal lattice of CeO_x, and thus the lattice parameters of nitrogen doped CeO_x in Ce-CN-V nanocomposites became larger than those of pure CeO_x due to the substitution of 4-coordinated O²⁻ (ionic radius: 1.38 Å) by N³⁻ (ionic radius: 1.46 Å) in CeO_x.³⁴

Fig. 1 TEM images of Ce-CN-V photocatalysts: (A) V = 1 mL, (B) V = 5 mL, (C) V = 10 mL. (D) The magnified TEM image of Ce-CN-5.

X-ray diffraction (XRD) patterns for g-C₃N₄, CeO_x and Ce-CN-V photocatalysts were acquired to investigate their crystal structures. As shown in Fig. 2A, the (100) peak at 2θ = 13.0° corresponding to the d value of 0.676 nm is related to an in-plane structural packing motif of g-C₃N₄, such as the hole-to-hole distance of the nitride pores in the crystal.³⁵ Both pure g-C₃N₄ and Ce-CN-V nanocomposites show a strong characteristic (002) peak at 2θ = 27.4°, corresponding to the interlayer-stacking of the conjugated aromatic system with a stacking distance of 0.326 nm, indicating that the as-prepared nanocomposites possess layered structures of g-C₃N₄. It was also found that with increased N-CeO_x amounts in the nanocomposites, the relative intensity of both (100) and (002) peaks were significantly decreased; this was possibly due to the deposition of N-CeO_x on the surface of g-C₃N₄.¹³ Here, the weakening of the (100) peak could be attributed to a slight distortion of the nitride pores, caused by N-CeO_x modification, which changed the hole to hole distance.^35,36 For the (002) peak, the weakening could result from a decrease in structural correlation length induced by the introduction of small sized N-CeO_x NPs.^37,38

Fig. 2 (A) XRD patterns of g-C₃N₄, CeOx and Ce-CN-V photocatalysts with various volumes of Ce(NO₃)₃ precusor solution; (B) XRD patterns of CeOx and Ce-CN-V (V = 7, 10, and 15 mL) in the range of 2θ = 25°–50°. (◆) and (●) denote the characteristic peaks of g-C₃N₄ and CeO_x, respectively.

The XRD peaks of pure CeO_x were in good agreement with the cubic phase of CeO_x (JCPDS 004-0593). For Ce-CN-V nanocomposites, with increase in the volume of Ce(NO₃)₃ precursor solution, the CeO_x diffraction peaks gradually appeared and became stronger, indicating the presence of more N-CeO_x NPs in the nanocomposites. However, as shown in Fig. 2B, the XRD peaks at 2θ = 28.6°, 33.1°, and 47.5° for CeO_x slightly shifted to lower angles for Ce-CN-V, which should be attributed to the substitution of 4-coordinated O²⁻ (ionic radius: 1.38 Å) by N³⁻ (ionic radius: 1.46 Å) in CeO_x.³⁴ This is understandable because the nitrogen doping had occurred in CeO_x NPs obtained by annealing Ce(NO₃)₃ precursor in the NH₃ atmosphere^25,26 created by heating melamine at 550 °C, as described in Fig. S1 and S3 in ESI†. In addition, when compared with the reference samples of Ce-CN-T₂ and Ce-CN-M₂, these three XRD peaks also slightly shifted to lower angles for Ce-CN-15, as shown in Fig. S4 in ESI†. These results could help in distinguishing the N-CeO_x NPs modified g-C₃N₄ from CeO_x/g-C₃N₄ composites, which again indicates that the one-pot method was favorable for nitrogen doping into the lattice of CeO_x, as compared to the mechanical mixing or two-step method.

As determined by the XPS analysis (Fig. S5†), the actual weight ratios of Ce in Ce-CN-V (V = 1, 3, 5, 7, 10, and 15 mL) were determined to be 0.95, 2.45, 4.22, 5.88, 8.52 and 12.82 wt%, respectively. The weight contents for all the elements (C, N, O, Ce) in g-C₃N₄ and Ce-CN-V were also calculated, and are listed in Table S1 in ESI†. Ce 3d XPS spectra of pure CeO_x (Fig. 3a) and Ce-CN-15 (Fig. 3b) were further compared to provide evidence that nitrogen was indeed introduced into the crystal lattice of CeO_x. As shown in Fig. 3, because of the highly non-stoichiometric nature of CeO_x, both 3+ and 4+ states are present in CeO_x and Ce-CN-5. It should be noted that the Ce 3d binding energies (both Ce³⁺ and Ce⁴⁺) of Ce-CN-5 are slightly lower than those of pure CeO_x. For example, the binding energies of Ce⁴⁺ 3d_3/2, Ce⁴⁺ 3d_5/2, Ce³⁺ 3d_3/2 and Ce³⁺ 3d_5/2 for pure CeO_x are centered at ca. 916.8, 898.1, 900.7 and 882.3 eV, respectively; while those for Ce-CN-5 are located at 916.4, 897.9, 900.4 and 882.0 eV, respectively. Such lower shifts in the Ce 3d binding energies for Ce-CN-5 should be attributed to an increase in electron density around the Ce cation (Ce-N > Ce-O), which was induced by the replacement of oxygen atoms by nitrogen atoms with lower electronegativity.^25,39

Fig. 3 Ce 3d XPS spectra of (a) CeO_x and (b) Ce-CN-15 photocatalysts.

Fig. 4 shows UV-vis absorption spectra obtained to evaluate the optical absorption properties of the as-prepared photocatalysts. As calculated by the Kubelka–Munk method,^40,41 the absorption edges of pure g-C₃N₄ and CeO_x at around 460 nm and 430 nm correspond to the band gaps of 2.70 eV and 2.90 eV, respectively. The narrowed band gap and band position fine-tuning of CeO_x by nitrogen doping was also confirmed by DFT calculation (see Computational Methods and Fig. S6 in ESI†). By forming nanocomposites, the absorption edge showed a gradual red shift to 490 nm for Ce-CN-15 with increasing N-CeO_x loading amounts. It was also found that the light absorption in the range of λ > 500 nm was obviously enhanced, which was mainly due to the nitrogen doping-induced significant increase of the light absorption above 500 nm.^25,26,34 The enhanced visible light absorption was beneficial for the photocatalytic performance, which is widely reported.^17,18 However, for the Ce-CN-V (V = 1, 3, 5, 7, 10, and 15 mL) photocatalysts, the increasing N-CeO_x loading amounts, which could gradually enhance the visible light absorption, would give rise to increasing serious shading effects,^32,33,41 and thus they offset the benefit of the light absorption of N-CeO_x. As a result, this negative shading effect induced by the excessive loading of N-CeO_x NPs leads to decreased photocatalytic activity for hydrogen production, which is verified by the hydrogen production evaluation of the Ce-CN-V photocatalysts in the later sections.

Fig. 4 UV-vis diffuse reflectance spectra of the as-prepared photocatalysts. (a) pure CeO_x, (b) g-C₃N₄ and Ce-CN-V photocatalysts with various V: (c) V = 1 mL, (d) V = 3 mL, (e) V = 5 mL, (f) V = 7 mL, (g) V = 10 mL, (h) V = 15 mL.

It can be also noted that the visible light absorption of Ce-CN-5 was significantly enhanced compared to Ce-CN-T or Ce-CN-M (See Fig. S7 in ESI†), especially in the range of λ > 500 nm. This should be attributed to the fact that nitrogen was hardly introduced into the lattice of CeO_x by the two-step or mechanical method, and thus the visible light absorption ability of Ce-CN-5 with N-CeO_x modification was considerably stronger than that of Ce-CN-T or Ce-CN-M with only CeO_x modification.

The photoluminescence (PL) emission spectra are very useful for understanding the behavior of photo-generated charge carriers in photocatalysts because PL emission results from the irradiative recombination of electrons and holes.⁴²Fig. 5A shows the PL spectra for the as-prepared Ce-CN-V photocatalysts. All the photocatalysts exhibited a broad emission peak centered at around 460 nm, and a tail extending to 700 nm. The emission could be attributed to the band-band PL phenomenon related to the recombination of photoinduced electrons and holes in g-C₃N₄.^41,43 The PL emission intensity exhibits the highest value for pure g-C₃N₄ and monotonously decreases with increasing N-CeO_x content. This indicates that the recombination of photogenerated charge carriers is inhibited in the Ce-CN-V photocatalysts because of the efficient separation of the photo-induced electron hole pairs in the N-CeO_x/g-C₃N₄ heterojunctions.

Fig. 5 (A) Photoluminescence (PL) spectra of the Ce-CN-V photocatalysts. (a) g-C₃N₄, (b) Ce-CN-1, (c) Ce-CN-3, (d) Ce-CN-5, (e) Ce-CN-7, (f) Ce-CN-10, (g) Ce-CN-15. (B) PL spectra of (a) g-C₃N₄, (b) Ce-CN-M, (c) Ce-CN-T, (d) Ce-CN-5.

To prove that the N-CeO_x/g-C₃N₄ heterojunctions prepared by the one-pot annealing method are superior for charge carrier separation, the PL spectra of the reference samples (Ce-CN-T, Ce-CN-M) were also obtained for comparison, as shown in Fig. 5B. For the Ce-CN-T photocatalyst prepared by the two-step method, the PL intensity was stronger than that of Ce-CN-5, indicating poorer charge separation in Ce-CN-T than in Ce-Ce-5. This phenomenon should be due to the larger size of CeO_x nanoparticles, and thus the poorer interfacial contact between CeO_x and g-C₃N₄ (as evidenced in Fig. S8C, D in ESI†), which consequently results in less efficient charge separation in Ce-CN-T. Moreover, for the mechanical mixture sample of Ce-CN-M, the PL emission intensity was significantly stronger relative to Ce-CN-T, indicating very poor charge separation between CeO_x and g-C₃N₄ in Ce-CN-M. This is reasonable because of the serious aggregation of CeO_x nanoparticles and loose interfacial contact between CeO_x and g-C₃N₄ (Fig. S8E, F in ESI†) obtained by simply mechanical mixing the CeO_x and g-C₃N₄.

The charge separation in composite photocatalysts is usually related to the band alignment induced charge transfer at the interfaces of the two components (e.g., type II band alignment).⁴¹ In the present work, the valence band maximum (VBM) edge potentials of CeO_x and g-C₃N₄ were measured by valence band X-ray photoelectron spectroscopy (VB XPS) to estimate the band alignment for the Ce-CN-V photocatalysts. As shown in Fig. 6, the peak assigned to Ce 4f can be observed at ∼1.5 eV. It was reported that Ce 4f was present for Ce₂O₃ but was absent for pure CeO₂,⁴⁴ confirming the presence of Ce³⁺. The band stretching from ca. 3 to 8 eV is assigned to the O 2p orbitals.⁴⁴ For CeO_x, the valence band is derived from the O 2p orbitals, and the band gap of 2.9 eV is derived from the O 2p–Ce 4f transition;^24,44 thus, the VBM edge of the as-prepared CeO_x is estimated to be approximately 2.24 eV below the Fermi band (E_f) from the VB XPS spectra. For g-C₃N₄, the VBM edge is estimated to be 2.36 eV from Fig. 6, which is in accordance with our previous reports.^41,43 According to the equation E_CB = E_VB − E_g,^41,45 the E_CB of CeO_x and g-C₃N₄ were calculated to be 0.66 eV and 0.32 eV above E_f, respectively. Based on the above deduction, as illustrated in Fig. 6, the E_CB difference of 0.34 eV could facilitate the electron transfer from the conduction band of CeO_x to that of g-C₃N₄, while the E_VB difference of 0.12 eV will favor hole transfer from the valence band of g-C₃N₄ to that of CeO_x. It should be noted that because of nitrogen doping in N-CeO_x, the real E_g value for the N-CeO_x must be smaller (dotted line in Fig. 6), as suggested by the improved visible light absorption and DFT calculation results, leading to enhanced visible light response of the nanocomposite photocatalysts of N-CeO_x/g-C₃N₄ (e.g., the as-prepared Ce-CN-V heterojunctions).

Fig. 6 VB XPS spectra for (a): CeO_x and (b): g-C₃N₄. The inset is the energy band alignment of the CeO_x/g-C₃N₄ heterojunction. (Solid lines: band position for pure CeO_x (red) and g-C₃N₄ (black); black dotted lines near the CB and VB of CeO_x: estimated band position for N-CeO_x nanoparticles). E_f: Fermi level; E_VB: valance band energy level; E_CB: conduction band energy level.

To further confirm the efficient charge separation between N-CeO_x and g-C₃N₄ induced by Type II band alignment and intimate interfacial contact, time-resolved fluorescence emission decay spectra of Ce-CN-5, with g-C₃N₄, Ce-CN-M, and Ce-CN-T as references, were measured at room temperature, as shown in Fig. 7. The kinetic parameters for these as-prepared samples are summarized in Table 1. The average lifetimes of the carriers (τ_avg) were determined to be 4.94 ns, 4.99 ns, 9.85 ns and 21.10 ns for g-C₃N₄, Ce-CN-M, Ce-CN-T and Ce-CN-5, respectively (see decay kinetics calculation in the ESI†), indicating that the charge recombination in g-C₃N₄ was effectively inhibited when modified with N-CeO_x (or CeO_x) nanoparticles.^41,43,45 In particular, the N-CeO_x/g-C₃N₄ heterojunction prepared by the one-pot annealing method (Ce-CN-5) showed significantly prolonged carrier lifetime when compared to the reference samples of g-C₃N₄, Ce-CN-M and Ce-CN-T, which again confirmed the efficient charge transfer and separation between N-CeO_x and g-C₃N₄, due to the intrinsic straddling (Type II) band alignment and the intimate interfacial contact.

Fig. 7 Time-resolved fluorescence emission decay curves for g-C₃N₄, Ce-CN-5, Ce-CN-T and Ce-CN-M at room temperature. Observation wavelength for all the samples was 470 nm and the excitation wavelength was 337 nm. Solid lines represent the kinetic fit using tri-exponential decay analysis.

Table 1 Kinetic parameters of the emission decay analysis of the as-prepared samples

Samples	A 1	τ 1 (ns)	A 2	τ 2 (ns)	A 3	τ 3 (ns)	τ avg (ns)	χ ^2
g-C3N4	0.1109	1.425	0.8667	0.033	0.0224	8.558	4.94	0.9571
Ce-CN-M	0.1274	7.929	0.4567	1.410	0.4159	0.343	4.99	0.9606
Ce-CN-T	0.2987	4.122	0.6196	1.207	0.0817	18.69	9.85	1.054
Ce-CN-5	0.2488	2.269	0.1195	25.80	0.6317	0.295	21.10	0.9335

---

The photocatalytic activity of N-CeO_x NPs modified g-C₃N₄ photocatalysts (Ce-CN-V, V = 1, 3, 5, 7, 10, and 15 mL) was evaluated by the hydrogen production reaction under visible light (λ > 420 nm). Pt (1 wt%) was used as the co-catalyst, providing hydrogen evolution sites on the as-prepared photocatalyst for efficient photocatalytic hydrogen production (Fig. S9 in ESI†).^8,46–48

As shown in Fig. 8A, all the samples are photocatalytically active, and no hydrogen was detected in the absence of either photocatalyst or light irradiation. The rate of hydrogen evolution over pure g-C₃N₄ was 134.5 μmol h⁻¹ g⁻¹. As N-CeO_x content increased, the photocatalytic activity of Ce-CN-V was significantly improved, with the highest hydrogen evolution rate (292.5 μmol h⁻¹ g⁻¹) achieved for Ce-CN-5, which was about 2.2 times that of pure g-C₃N₄. However, the photocatalytic activity decreased as the N-CeO_x content of Ce-CN-V was further increased (V = 7, 10, and 15 mL).

Fig. 8 (A) Photocatalytic H₂ production under visible-light irradiation over g-C₃N₄, Ce-CN-V (V = 1, 3, 5, 7, 10, and 15 mL), Ce-CN-M and Ce-CN-T photocatalysts. T represents CeO_x/g-C₃N₄ (Ce-CN-T) prepared by the two-step method and M represents the mechanical mixture of CeO_x and g-C₃N₄ (Ce-CN-M). (B) Stable hydrogen evolution over a Ce-CN-5 sample under visible light irradiation. The reactor was flushed with N₂ every five hours to evacuate the H₂ produced.

To clarify the mechanism for photocatalytic H₂ production over Ce-CN-V, it is necessary to clarify how the N-CeO_x NPs influence the activity of g-C₃N₄; which is based on three effects resulting from N-CeO_x NPs modification: (1) efficient charge separation between g-C₃N₄ and N-CeO_x induced by the formed Type II band alignment,^15,16 (2) N-CeO_x acting as visible light sensitizers,^17,18 (3) the negative shading effect of the excessive N-CeO_x.^32,33,41 For the Ce-CN-V (V = 1, 3, and 5 mL) photocatalysts, as shown from the UV-vis and PL spectra in Fig. 4 and 5A, both the visible light absorption and charge separation efficiency were improved with increasing loading amounts of N-CeO_x NPs, leading to a gradual increase in photocatalytic activity for hydrogen evolution. In addition, it is noteworthy that both the optical absorption enhancement and the shading effect were related to the N-CeO_x content; thus, larger the N-CeO_x content, the stronger are the two effects. As seen from Fig. 4, the visible light absorption increased with the increase in the N-CeO_x loading amounts, while the N-CeO_x NPs aggregated as observed in Fig. 1C, leading to the negative shading effect of aggregated N-CeO_x NPs. Thus, for Ce-CN-V (V = 7, 10, and 15 mL), although the light absorption ability was improved by increasing the N-CeO_x loading amounts, the negative shading effect induced by the aggregation of N-CeO_x NPs became more significant, leading to a decrease in photocatalytic performance. Therefore, the three factors mentioned above balanced and competed with each other, eventually resulting in the highest photocatalytic performance for the Ce-CN-5 photocatalyst with optimized N-CeO_x loading.

To investigate the photocatalytic stability, the photocatalytic experiment for hydrogen production was run for 5 cycles. As shown in Fig. 8B, Ce-CN-5 shows good stability for photocatalytic hydrogen production, with the hydrogen production rate remaining almost unchanged during the 35 h reaction.

Hydrogen production over the reference samples of Ce-CN-M (mechanical mixture) and Ce-CN-T (two-step prepared) was also tested to prove the photocatalytic superiority of the N-CeO_x/g-C₃N₄ heterojunction formed by the one-pot annealing method. As shown in Fig. 8A, the reference Ce-CN-T sample shows lower photocatalytic activity (201.7 μmol h⁻¹ g⁻¹) than the Ce-CN-5 sample (292.5 μmol h⁻¹ g⁻¹). This may be due to the limited visible light absorption ability of CeO_x (E_g ∼2.9 eV) without nitrogen doping as well as the considerably larger CeO_x nanoparticles hindering the charge transfer and separation between g-C₃N₄ and CeO_x in Ce-CN-T (as evidenced by the PL properties in Fig. 5 and 7). Moreover, the Ce-CN-M reference sample, with a hydrogen production rate of only 142.5 μmol h⁻¹ g⁻¹, shows considerably lower photocatalytic activity than Ce-CN-T, but close to that of pure g-C₃N₄. This may be because the mechanical method did not form CeO_x/g-C₃N₄ heterojunctions with intimate interfacial contact because of the serious aggregation of CeO_x nanoparticles, resulting in a very low efficiency of charge transfer and separation between the two components.

To prove the existence and function of Pt nanoparticles as hydrogen evolution sites, TEM images for Pt loaded Ce-CN-5 after the photocatalytic hydrogen production reaction were obtained, as shown in Fig. S9 in ESI†. Pt NPs with an average size of ca. 10 nm were only formed on the surface of g-C₃N₄ for both pure g-C₃N₄ and Ce-CN-5, indicating that Pt NPs were preferentially reduced by the CB electrons from g-C₃N₄. Therefore, the Pt NPs acted as the reduction active sites on g-C₃N₄ to entrap photoinduced electrons for photocatalytic hydrogen production, which has been evidenced by previous studies.^8,43,46–48

As analyzed above, a possible photocatalytic mechanism of the as-prepared N-CeO_x NPs modified g-C₃N₄ nanocomposites for photocatalytic hydrogen production under visible light was proposed, and it is illustrated in Fig. 9. Upon visible light irradiation, both N-CeO_x and g-C₃N₄ were photo-excited to produce CB electrons and VB holes. Because of the VB and CB energy level difference between the two components, the CB electrons in N-CeO_x could easily transfer to the CB of g-C₃N₄, which could then get accumulated at Pt reduction active sites for the H₂ production reaction, while the VB holes in g-C₃N₄ tended to transfer to the VB of N-CeO_x to react with the sacrificial reagent in the aqueous solution. In this Pt/N-CeO_x/g-C₃N₄ system, N-CeO_x acted as a sensitizer improving the light absorption, and the Type II band alignment formed between N-CeO_x and g-C₃N₄ efficiently separated the electron–hole pairs. Thus, the enhanced photocatalytic efficiency for hydrogen production under visible light was obtained over N-CeO_x NPs modified g-C₃N₄ when compared to pure g-C₃N₄.

Fig. 9 Schematic of band structure and charge transfer process in N-CeO_x NPs/g-C₃N₄ for photocatalytic hydrogen generation.

Conclusions

Nitrogen-doped CeO_x nanoparticles modified g-C₃N₄ photocatalysts were successfully prepared via a one-pot annealing method. The photocatalytic activity for hydrogen production over g-C₃N₄ under visible light was significantly improved by nitrogen-doped CeO_x NPs modification. The photoactivity was enhanced by ∼1.2 times, and the hydrogen evolution rate increased from 134.5 μmol h⁻¹ g⁻¹ for pure g-C₃N₄ to 292.5 μmol h⁻¹ g⁻¹ for the optimized sample of nitrogen-doped CeO_x NPs modified g-C₃N₄ prepared by the one-pot method. It was believed that the significantly enhanced photocatalytic activity was mainly attributed to the improved visible light absorption with nitrogen-doped CeO_x NPs acting as sensitizers, as well as the Type II band alignment-induced efficient charge separation between nitrogen-doped CeO_x NPs and g-C₃N₄ with intimate interfacial contact. The present study presents a new facile method for nitrogen doping and heterojunction formation, which might provide novel opportunities for exploring nanostructured heterojunctions for photo(electro)catalytic and optoelectronic applications.

Acknowledgements

The authors gratefully acknowledge the financial supports from the National Natural Science Foundation of China (no. 51102194, no. 51323011, no. 51121092), the Natural Science Foundation of Shaanxi Province (no. 2014KW07-02), the Natural Science Foundation of Jiangsu Province (no. BK20141212) and the Nano Research Program of Suzhou City (no. ZXG201442, ZXG2013003). One of the authors (S. Shen) was supported by the Foundation for the Author of National Excellent Doctoral Dissertation of China (no. 201335) and the “Fundamental Research Funds for the Central Universities.”

Notes and references

1. K. Akihiko and M. Yugo, Chem. Soc. Rev., 2009, 38, 253–278.
2. K. Zhang and L. J. Guo, Catal. Sci. Technol., 2013, 3, 1672–1690.
3. X. Chen, S. Shen, L. Guo and S. S. Mao, Chem. Rev., 2010, 110, 6503–6570.
4. L. L. Amy, G. Q. Lu and T. Y. John, Chem. Rev., 1995, 95, 735–758.
5. P. Li, Z. Wei, T. Wu, Q. Peng and Y. D. Li, J. Am. Chem. Soc., 2011, 133, 5660–5663.
6. J. Fang, L. Xu, Z. Zhang, Y. Yuan, S. Cao, Z. Wang, L. S. Yin, Y. S. Liao and C. Xue, ACS Appl. Mater. Interfaces, 2013, 5, 8088–8092.
7. M. Kazuhiko, T. Tsuyoshi, H. Michikazu, S. Nobuo, I. Yasunobu, K. Hisayoshi and D. Kazunari, J. Am. Chem. Soc., 2005, 127, 8286–8287.
8. X. C. Wang, K. Maeda, A. Thomas, K. Takanabe, G. Xin, J. M. Carlsson, K. Domen and M. Antonietti, Nat. Mater., 2009, 8, 76–80.
9. X. C. Wang, B. Siegfried and A. Markus, ACS Catal., 2012, 2, 1596–1606.
10. J. Liang, Y. Zheng, J. Chen, J. D. Hulicova, M. Jaroniec and S. Z. Qiao, Angew. Chem., Int. Ed., 2012, 51, 3892–3896.
11. X. F. Chen, J. S. Zhang, X. Z. Fu, M. Antonietti and X. C. Wang, J. Am. Chem. Soc., 2009, 131, 11658–11659.
12. G. Liu, P. Niu, C. Sun, C. S. Smith, Z. G. Chen, G. M. Lu and H. M. Cheng, J. Am. Chem. Soc., 2010, 132, 11642–11648.
13. J. Fu, Y. L. Tian, B. B. Chang, F. N. Xi and X. P. Dong, J. Mater. Chem., 2012, 22, 21159–21166.
14. K. Maeda, X. Wang, Y. Nishihara, D. Lu, M. Antonietti and K. Domen, J. Phys. Chem. C, 2009, 113, 4940–4947.
15. H. Yan and H. Yang, J. Alloys Compd., 2011, 509, 26–29.
16. Y. J. Wang, R. Shi, J. Lin and Y. F. Zhu, Energy Environ. Sci., 2011, 4, 2922–2929.
17. S. Ye, L. G. Qiu, Y. P. Yuan, Y. J. Zhu, J. Xia and J. F. Zhu, J. Mater. Chem. A, 2013, 1, 3008–3015.
18. M. Xu, L. Han and S. Dong, ACS Appl. Mater. Interfaces, 2013, 5, 12533–12540.
19. A. Primo, T. Marino, A. Corma, R. Molinari and H. Garcia, J. Am. Chem. Soc., 2011, 133, 6930–6933.
20. S. Kundu, J. Ciston, S. D. Senanayake, D. A. Arena, E. Fujita, D. Stacchiola, L. Barrio, R. M. Navarro, J. L. G. Fierro and J. A. Rodriguez, J. Phys. Chem. C, 2012, 116, 14062–14070.
21. Y. Wang, B. Li, C. Zhang, L. Cui, S. Kang, X. Li and L. Zhou, Appl. Catal., B, 2013, 130, 277–284.
22. S. Hu, F. Zhou, L. Wang and J. Zhang, Catal. Commun., 2011, 12, 794–797.
23. N. Wetchakun, S. Chaiwichain, B. Inceesungvorn, K. Pingmuang, S. Phanichphant, A. I. Minett and J. Chen, ACS Appl. Mater. Interfaces, 2012, 4, 3718–3723.
24. W. Li, S. Xie, M. Li, X. Ouyang, G. Cui, X. Lu and Y. Tong, J. Mater. Chem. A, 2013, 1, 4190–4193.
25. C. Mao, Y. Zhao, X. Qiu, J. Zhu and C. Burda, Phys. Chem. Chem. Phys., 2008, 10, 5633–5638.
26. A. B. Jorge, Y. Sakatani, C. Boissiere, C. Roberts, G. Sauthier, J. Fraxedas, C. Sanchez and A. Fuertes, J. Mater. Chem., 2012, 22, 3220–3226.
27. G. Zhang, J. Zhang, M. Zhang and X. Wang, J. Mater. Chem., 2012, 22, 8083–8091.
28. S. C. Yan, Z. S. Li and Z. G. Zou, Langmuir, 2009, 25, 10397–10401.
29. G. Li, N. Yang, W. Wang and W. F. Zhang, J. Phys. Chem. C, 2009, 113, 14829–14833.
30. Q. Li, B. Yue, H. Iwai, T. Kako and J. H. Ye, J. Phys. Chem. C, 2010, 114, 4100–4105.
31. Y. Liu, G. Chen, C. Zhou, Y. Hu, D. Fu, J. Liu and Q. Wang, J. Hazard. Mater., 2011, 190, 75–80.
32. M. Y. Wang, L. Sun, Z. Q. Lin, J. H. Cai, K. P. Xie and C. J. Lin, Energy Environ. Sci., 2013, 6, 1211–1220.
33. Y. D. Hou, A. B. Laursen, J. S. Zhang, G. G. Zhang, Y. S. Zhu, X. C. Wang, S. Dahl and I. Chorkendorff, Angew. Chem., Int. Ed., 2013, 52, 3621–3625.
34. D. Sun, M. Gu, R. Li, S. Yin, X. Song, B. Zhao, C. Li, J. Li, Z. Feng and T. Sato, Appl. Surf. Sci., 2013, 280, 693–697.
35. A. Thomas, A. Fischer, F. Goettmann, M. Antonietti, J. O. Müller, R. Schlögl and J. M. Carlsson, J. Mater. Chem., 2008, 18, 4893–4908.
36. Y. J. Zhang, A. Thomas, M. Antonietti and X. C. Wang, J. Am. Chem. Soc., 2008, 131, 50–51.
37. Y. B. Li, H. M. Zhang, P. R. Liu, D. Wang, Y. Li and H. J. Zhao, Small, 2013, 9, 3336–3344.
38. X. H. Li, X. C. Wang and M. Antonietti, ACS Catal., 2012, 2, 2082–2086.
39. R. Li, C. Fu, C. Q. Li, S. Yin, Y. Zhang and T. Sato, J. Ceram. Soc. Jpn., 2010, 118, 555–557.
40. V. Kumar, S. K. Sharma, T. P. Sharma and V. Singh, Opt. Mater., 1999, 12, 115–119.
41. J. Chen, S. Shen, P. Guo, M. Wang, P. Wu, X. Wang and L. Guo, Appl. Catal., B, 2014, 152–153, 335–341.
42. S. Shen, P. Guo, L. Zhao, Y. Du and L. Guo, J. Solid State Chem., 2011, 184, 2250–2256.
43. J. Chen, S. Shen, P. Guo, P. Wu and L. Guo, J. Mater. Chem. A, 2014, 2, 4605–4612.
44. D. R. Mullins, S. H. Overbury and D. R. Huntley, Surf. Sci., 1998, 409, 307–319.
45. J. Zhang, M. Zhang, R. Q. Sun and X. C. Wang, Angew. Chem., Int. Ed., 2012, 124, 10292–10296.
46. J. Zhang, Q. Xu, Z. Feng, M. Li and C. Li, Angew. Chem., Int. Ed., 2008, 47, 1766–1769.
47. P. Gao, J. C. Liu, S. Lee, T. Zhang and D. D. Sun, J. Mater. Chem., 2012, 22, 2292–2298.
48. S. R. Lingampalli, U. K. Gautam and C. N. R. Rao, Energy Environ. Sci., 2013, 6, 3589–3594.

---

Footnote

† Electronic supplementary information (ESI) available: The illustrations for the fabrication and formation process of Ce-CN-V photocatalysts, particle size distribution histogram, XPS, DFT calculation results, UV-vis, PL spectra, TEM images and decay kinetics for the as-prepared photocatalysts. See DOI: 10.1039/c4gc01683a

---