Noble-metal-free Ni3C as co-catalyst on LaNiO3 with enhanced photocatalytic activity

Han Zhang, Zhaoming Luo, Yalun Liu, Yuwen Jiang*

School of Physical Science and Technology, Southwest University, Chongqing 400715, China

*Corresponding author E-mail address: ywj623@swu.edu.cn.

Graphical abstract

Schematic illustration of photocatalytic H2 evolution and charge transfer process in the LaNiO3-Ni3C composites photocatalyst under light irradiation. Besides, with loading Ni3C NPs from 1wt% to 6wt%, the variation of charge-separation performances, H2-evolution kinetics, bandgaps, and photocatalytic H2 evolution rate are compared together.

Highlights

It is the first time transition metal carbide is used as a cocatalyst with the oxide semiconductor. · The H2 evolution rate of LaNiO3 is improved about 5 times by loading proper Ni3C content, which is 2 times as that of loading 1wt%Pt. ·

3. Loading Ni3C can notably improve the charge-separation performance and the H2 evolution kinetics. ·

4. Restriction of Ni3C content can suppress corrosion from the reductive carbonaceous precursor in the loading process.

Abstract

In this study, Ni3C nanoparticles (NPs) were loaded on LaNiO3 NPs as a co-catalyst during the photocatalytic H2 evolution. The evolution rate of the LaNiO3-3wt%Ni3C composite (4093.9 μmolg-1 h - 1 ) is about 5 times and 2 times as that of pristine LaNiO3 NPs (819.2 μmolg-1 h -1 ), and 1wt% Pt loaded LaNiO3 NPs (1927 μmolg-1 h -1 ). The effects of co-catalyze with Ni3C on light absorption, electron-hole spatial separation, and surface H2-evolution kinetics are investigated. Loading with Ni3C can enhance the spatial separation of electron-hole pairs, and improve the H2-evolution kinetics. Further investigation shows that surface decomposition of LaNiO3 NPs can only be observed when the Ni3C content is over 3wt%, caused by the reductive carbon source needed in the loading process. This work indicates that transition metal carbide can be a promising co-catalyst candidate for oxides in photocatalytic applications without concerning corrosion from the reductive carbonaceous precursors in the loading process.

Keywords: Photocatalytic hydrogen evolution; Noble-metal-free Ni3C co-catalysts; Electron-hole spatial separation; H2 evolution kinetics; reductive corrosion.

1. Introduction

Hydrogen production by photocatalytic water splitting represents one of the most promising approaches for energy storage and clean fuel generation in a post-fossil age. An ideal photocatalyst for water splitting requires absorption of light in a broad range, transporting the photoinduced electron-hole pairs with good spatial separation, and reducing water into hydrogen at the surface with high activity, which is hard to be archived on one material [1, 2]. Therefore, constructing a composite photocatalyst, composing a semiconductor part for light absorption, and a co-catalyst for water reduction becomes a widely applied strategy for designing water-splitting H2 evolution photocatalyst [2, 3]. So far, noblemetal materials are the best candidates for the co-catalyst, due to the large work functions of which a Schottky barrier would be formed at their interface when a noble metal particle makes contact with a semiconductor [4]. The Schottky barrier can facilitate the separation of the electrons and holes, then enhance the photocatalytic efficiency [5, 6]. On the other hand, with the excellent conductivity and the proper hydrogen binding energy at the surface, noble-metal materials can remarkably reduce the overpotential for hydrogen evolution, showing excellent surface activity for hydrogen evolution from water [7, 8]. However, the high cost and low earth abundance limit its usage in large-scale applications. It is urgently demanded to develop cost-efficient co-catalysts for photocatalytic water-splitting.

Remarkable advances have been made regarding the use of various transition metal compounds as cocatalysts for photocatalytic water splitting, such as transition metal chalcogenides [9-11], transition me borides [12, 13] and transition metal carbides [7, 14, 15]. These materials generally have good conductivity, high activity, and earth abundance, which make them promising substitutes for noble-metal materials. Among these candidates, transition metal carbides (TMCs) are highlighted by several advantages, such as noble-metal-like electronic configuration, environment-friendly, wide pH applicability, and, more importantly, outstanding hydrogen evolution reaction (HER) activity and stability [16, 17]. However, in most researches, the TMCs are combined with sulfide semiconductor or g-C3N4 [18, 19]. Naturally, the performance of TMCs as co-catalyst with other kinds of photocatalyst draws our great attention. Especially the performance with oxide semiconductors interests us most, which is the most promising candidate for large-scale photocatalytic H2 evolution application for its outstanding photocatalytic stability, the high abundance, and the environment-friendly [20-22].

In this study, we selected Ni3C as a co-catalyst loaded on LaNiO3 nanoparticles (NPs). Loading 3wt% Ni3C on LaNiO3 NPs, the H2 production rate is increased about 5 times and 2 times compared with pristine LaNiO3 nanoparticle and 1wt% Pt loaded LaNiO3 nanoparticle. To the best of our knowledge, this is the first time metal carbide is used as a co-catalyst with oxide semiconductors for H2 production. The structure, composition, and morphology of the Ni3C loaded LaNiO3 NPs are characterized by XRD, XPS, SEM, and HRTEM. The effects of loading different contents of Ni3C on the photocatalytic progress are investigated by UV-vis, EIS test, photocurrent test, electrocatalytic hydrogen evolution test, and Tafel analysis. We find that the charge-separation performance and H2 evolution kinetics can be continuously improved by increasing Ni3C contents from 1wt% to 3wt%, then dramatically deteriorated with further increasing Ni3C contents from 4wt% to 6wt%. With the detailed characterization of the Ni3C overloaded LaNiO3 NPs, the deterioration can be ascribed to the surface decomposition of LaNiO3 NPs, caused by the reductive carbon source needed in the chemically Ni3C loading process. This work indicates that using TMC as co-catalyst with oxide photocatalyst can be a promising strategy for designing low cost, high efficient, and long sustainability H2 evolution photocatalyst. The corrosion from the reductive carbonous precursors in the loading process can be suppressed by restraining the TMC content.

2. Experimental section

Lanthanum nitrate (La(NO3)3·6H2O) was purchased from Aladdin. Nickel acetate (C4H6O4Ni·4H2O), Diethylene glycol (DEG), and Chloroplatin acid (0.1g/ml) were purchased from the Beijing Chemical Reagent (Beijing, China). Oleylamine (OLA) was purchased from Macklin. All reagents were of analytical grade and used without further purifications.

2.1. Synthesis of LaNiO3 

NPs Firstly, 0.4 mmol of La(NO3)3·6H2O and 0.4 mmol of C4H6O4Ni·4H2O (Ni(Ac)2·4H2O) were dissolved in 10 ml of DEG. The mixed solution was heated up to 130 oC under stirring until a clear solution was obtained. After a proper amount of water was added, the solution was further heated up to 180 oC and kept this temperature for 5 h. The emerging suspension was cooled and precipitated by adding ethanol and distilled water. Finally, the crystalline LaNiO3 was synthesized by annealing the resulting precursor powder in air at 700 oC for 12 h.

2.2. Synthesis of LaNiO3-Ni3C composites

In a typical synthesis process, 100 mg of the as-synthesized LaNiO3 sample, proper content of Ni(Ac)2·4H2O, and 10 ml of OLA were mixed into a 100 ml three-neck round-bottom flask and stirred magnetically for 30 min. Then, it was heated to 250 °C in an atmosphere of inert gas for 2h. After cooling to room temperature, 30 ml of acetone was added into the mixture. The same centrifugation and isolation procedure was repeated several times to purify the composites.

The pure Ni3C NPs were synthesized by the same method without adding the LaNiO3 sample.

2.3. Synthesis of LaNiO3-1wt%Pt composites

80 mg of the as-synthesized LaNiO3 sample, 14 μl of the chloroplatin acid solution with a concentration of 0.1 g/ml were dispersed in 0.4 ml of deionized water. The mixed solution was sonicated for 30 minutes and dried at 90 oC . Then, the obtained powder was placed in a tube furnace and heated at 180 oC for 2 hours and cooled naturally.

2.4. Photocatalytic experiments procedures

Photocatalytic reactions were performed in a gas-closed circulation system (CEL-SPH2N) with top irradiation. 40 mg of the as-prepared photocatalyst was dispersed in 100 ml of an aqueous solution containing sodium sulfide (Na2S, 0.25 M) and sodium sulphite (Na2SO3, 0.25 M) in a quartz reactor. A 300 W Xe lamp (λ>200 nm) was employed as the light source, and the temperature was maintained at 6 °C. The amount of H2 was recorded by online gas chromatography (GC7920, TCD detector, N2 carrier).

2.5. Characterization

X-ray powder diffraction (XRD) patterns of the prepared photocatalysts were collected by PANalytical X’Pert Powder diffractometer using Cu Kα radiation (λ =0.154 nm). The transmission electron microscopic (TEM) and high-resolution transmission electron microscopic (HRTEM) images of the prepared photocatalysts were obtained by a field emission FEI Tecnai G2 F30 apparatus. The UV-vis absorption spectra were obtained on a Shimadzu UV-2600 spectrometer. X-ray photoelectron spectroscopy (XPS) data was performed with a VG ESCALAB250 surface analysis system. All the binding energies were calibrated using the C 1s level at 284.8 eV as the reference. Electrochemical tests (EIS, Photocurrent, Electrocatalytic hydrogen evolution, Tafel analysis) were conducted using a CHI700E Electrochemical Analyzer with a standard three-electrode system.

Please refer to supplementary information for detailed sample preparation methods and characterization procedures.

3. Results and discussion

3.1. loading 3wt% Ni3C on LaNiO3 samples

The structures, compositions, the morphology and light absorption of the composites

Fig. 1 depicts the XRD patterns of as-obtained pure Ni3C, LaNiO3, and LaNiO3-Ni3C composites. For Ni3C in Fig. 1A, all XRD peaks can be indexed as a Hexagonal Ni3C phase (JCPDS No. 06-0697) [7]. The strong and sharp peaks at 39.5°, 41.9°, 44.9°, 58.6°, 71.2°, and 78.2° correspond to the (110), (006), (113), (116), (300), and (119) crystal planes of the Ni3C lattice. As shown in Fig. 1B, the typical diffraction peaks of pristine LaNiO3 are observed at 2θ = 23.1°, 32.7°, 40.5°, 47.1°, 53.4°, 58.5°, 68.6°, and 78.4° which are attributed to the (101), (110), (021), (202), (211), (122), (220) and, (312) crystal planes of perovskite-type LaNiO3 (JCPDS Card No. 34-1181) [23], respectively. In the XRD pattern of the pristine LaNiO3 samples, no other miscellaneous peaks are observed, indicating the high purity of the obtained sample. The diffraction peaks of LaNiO3-3wt%Ni3C composites are similar to those of pristine LaNiO3 (Fig. 1B). It is noteworthy that no characteristic diffraction peaks of Ni3C are observed in the XRD pattern of LaNiO3-3wt%Ni3C composite, indicating the small contents and the homogeneous dispersion of the Ni3C in the composites.

Fig. 1. (A) The XRD pattern of pure Ni3C. (B) The XRD patterns of LaNiO3 samples and LaNiO3- 3wt%Ni3C composites.

XPS technology is further employed to characterize the surface composition and chemical state of various elements in the LaNiO3-3wt% Ni3C sample. As displayed in the inset in Fig. 2A, the fully scanned spectrum indicates the existence of La, Ni, and O and C elements in the LaNiO3-3wt% Ni3C composite. The C 1s XPS spectrum is shown in Fig. 2A. The peak appearing at 283.2 eV should be assigned to the Ni-C bond in Ni3C, which indicates that Ni3C NPs are attached to the surface of LaNiO3 NPs [15, 24]. The peak at 284.6 eV can be assigned to the sp2 graphite component (C-C) of the C in the Ni3C NPs [15, 25]. The peaks at 286.3 eV and 288.0 eV can be assigned to the C-O, and C=O functional groups, respectively [26], both of which are usually quite weak compared to the C-C peak and considered coming from the slightly surface oxidation in the Ni3C NPs [27]. However, the C-O peak in Fig. 2A is comparable to the C-C peak and is much stronger than the weak C=O peak. We suggest that the strong C-O peak in Fig. 2A may come from the surface interaction between Ni3C NPs and the LaNiO3 NPs [7, 14, 24], while the C=O peak should come from the surface oxidation of the Ni3C NPs [14, 26]. Fig. 2B shows the XPS spectrum of O 1s. The two peaks with the binding energies of 528.5 eV and 531.7 eV should come from the lattice oxygen and the oxygen vacancies, respectively [27, 28]. The peak at 534 eV can be assigned to the O-C group, which may be too strong to be ascribed to the surface oxidation of Ni3C [19]. Considering the prominent C-O peak (286.3 eV) in the C 1s spectrum, we believe the prominent O-C peak should also be ascribed to the surface contact between Ni3C NPs and the LaNiO3 NPs. In Fig. 2C, the four peaks at 834.3 eV, 837.7 eV, and 850.7 eV, 854.4 eV, could be assigned to the binding energies of La 3d5/2 and La 3d3/2, respectively, which suggests that the valence state of La element is +3 [29, 30]. Fig. 2D shows the Ni 2p XPS spectrum. Interestingly, it is not easy to determine the Ni 2p3/2 core level spectrum by XPS analysis due to Ni 2p3/2 peaks is strongly overlapped with the La 3d3/2 peaks [31, 32]. Thus, the Ni 2p peaks and parts of the La 3d3/2 peaks are both shown in Fig. 2D. The five peaks appeared at 856.3 eV, 862.0 eV, and 866.7 eV, 872.3 eV, 879.0 eV could be assigned to the binding energies of Ni 2p3/2, and Ni 2p1/2, respectively. The Ni 2p3/2 peak at 856.3 eV indicates the chemical state of Ni is +3 [33, 34]. It should be mentioned, the valance state of Ni in Ni3C is +1, which has the corresponding peaks at about 855 eV and 861 eV [14, 26, 35]. It is difficult to separate these peaks in the spectrum for the low content (3wt%). All the XPS results further confirmed that the Ni3C NPs were successfully loaded on the surface of LaNiO3 NPs.

Fig. 2. XPS spectra of LaNiO3-3wt%Ni3C composites: (A) C 1s, (B) O 1s, (C) La 3d, (D) Ni 2p. The inset of (A) is the XPS survey spectra of LaNiO3-3wt%Ni3C composites.

The morphology and nanostructures of the pristine LaNiO3 NPs and the LaNiO3-3wt%Ni3C composite were characterized by SEM, TEM, and HRTEM measurements. Fig. 3A and 3B show the SEM and TEM images of the pristine LaNiO3 sample, in which small nanoparticles with a cube-like shape and a dimension of about 10-20 nm are piled up. In Fig. 3C, the HRTEM image of LaNiO3 NPs shows the lattice fringes with interplanar spacings of 0.273 nm, which is corresponding to the (110) crystal plane of perovskite structure LaNiO3 [36]. The corresponding selected area electron diffraction (SAED) pattern (inset of Fig. 3C) presents well-defined rings, confirming the polycrystalline characteristic of the prepared LaNiO3 sample. Rings corresponding to the (110), (202), and (122) crystal plane of the LaNiO3 are labeled [37]. Fig. 3D and 3E show the SEM and TEM images of the LaNiO3-3wt%Ni3C composite, in which stacked nanoparticles with a dimension of about 10-20 nm are also observed. It is noticed that the nanoparticles of the LaNiO3-3wt%Ni3C composite show smoother edges than those of pristine LaNiO3, which may come from the Ostwald ripening process occurring in the synthesis of LaNiO3- 3wt%Ni3C composites. Part of the nanoparticle surface may be dissolved into the oleylamine solution and reformed into a more, thermodynamically favored, isotropic shape [7, 14, 38]. Fig. 3F shows the HRTEM image of the LaNiO3-3wt%Ni3C composites. Lattice fringes with interplanar spacings of 0.384 nm are observed, which is corresponding to the (101) crystal plane of the perovskite LaNiO3 structure [39]. The HRTEM image also clearly shows lattice fringes with interplanar spacings of 0.158 nm and 0.202 nm in some small particles (marked with cycles), which is corresponding to the (116) crystal plane and (113) crystal plane of the hexagonal Ni3C structure, respectively [40]. The HRTEM image indicates the intimate contact between Ni3C NPs and LaNiO3 NPs, and the sample is not just a mixture of the former and the latter.

Fig. 3. (A-C) The SEM, TEM, and HRTEM images of LaNiO3 NPs. The inset of (C) is SAED pattern of LaNiO3 NPs. (D-F) The SEM, TEM and HRTEM images of LaNiO3-3wt%Ni3C composite.

Fig. 4A shows the energy-dispersive X-ray spectrum (EDX) of the LaNiO3-3wt%Ni3C composites. La, Ni, O, and C elements were detected, confirming the existence of Ni3C NPs on the LaNiO3 NPs. The corresponding SEM elemental mapping (Fig. 4B–F) could confirm the co-existence of La, Ni, O, and C elements in the LaNiO3-3wt%Ni3C composite, which further confirms the uniform dispersion of Ni3C NPs on the surface of LaNiO3 particles.

Fig. 4. (A) EDX spectrum of LaNiO3-3wt%Ni3C composites and (B–F) the corresponding elemental mapping of LaNiO3-3wt%Ni3C composite: (B) SEM image of the mapping area, (C) C, (D) O, (E) La, (F) Ni.

Fig. S1A illustrated the typical UV-Vis diffuse reflectance spectra (DRS) of the as-obtained pristine LaNiO3, LaNiO3-1wt%Pt, and LaNiO3-3wt%Ni3C composites. The LaNiO3 sample and LaNiO3-1wt%Pt composite show nearly identical spectra with a broad light absorption range from 200 nm to 800 nm [18]. The LaNiO3-3wt%Ni3C composites show a similar absorption spectrum except for the slightly improved absorption at the long-wavelength end, which may be ascribed to the interaction between LaNiO3 and Ni3C, as reported [18, 24, 41]. Furthermore, the bandgaps of the pristine LaNiO3, LaNiO3-1wt%Pt, and LaNiO3-3wt%Ni3C composites are calculated by Kubelka–Munk method [42, 43], which is 1.88 eV, 1.91 eV and 1.92 eV, respectively (Fig. S1B). It seems the loading with Pt or Ni3C NPs could barely influence on the bandgap of LaNiO3.

The charge-separation performances

EIS tests were employed to study the separation efficiency of electron-hole pairs and the charge transfer resistance at solid/electrolyte interfaces. The Nyquist plots of the samples are shown in Fig. 5A.

The semicircle radius of the pristine LaNiO3 sample should be larger than 15000, which is reduced to nearly 11000 by loading with 1wt% Pt. While the semicircle radius of the LaNiO3-3wt%Ni3C samples is about 1500 (enlarged image is shown in the inset in Fig. 5A). This result suggests that the loading of 3wt% Ni3C co-catalysts could effectively decrease the interface charge transfer resistance and improve the separation efficiency of photo-generated charge carriers.

The transient photocurrent response tests were conducted to characterize the charge transferring in the sample. The transient photocurrent response (I–t curves) of pristine LaNiO3, LaNiO3-1wt%Pt, and LaNiO3-3wt%Ni3C composites were measured for several light on-off cycles and shown in Fig. 5B. The average photocurrent intensity is improved from about 0.26 μA/cm2 to nearly 0.34 μA/cm2 by loading with 1wt% Pt. While the photocurrent intensity is markedly improved to about 0.58 μA/cm2 by loading with 3wt% Ni3C, suggesting that a higher separation efficiency of the photo-generated carriers is achieved after loading Ni3C on the surface of LaNiO3 samples. Therefore, more photo-generated electrons could be transferred to the surface for hydrogen evolution.

The above results demonstrated that Ni3C NP is a better co-catalyst than Pt on the improvement of the charge-separation performances of LaNiO3 NPs.

Fig. 5. (A) Nyquist plots of different working electrodes in 0.25 M Na2S and Na2SO3 mixed aqueous solution in the dark. (B) Transient photocurrent responses (I-t curves) of LaNiO3, LaNiO3-1wt%Pt and LaNiO3-3wt%Ni3C composite in 0.1 M Na2SO4 aqueous solution under Xe lamp irradiation with (λ > 200 nm) at 0.1 V vs. Ag/AgCl. (C) Polarization curves of the composite measured at a scan rate of 5 mVs-1 in 0.25 M Na2S and Na2SO3 mixed aqueous solution. (D) Tafel slops of LaNiO3, LaNiO3-1wt%Pt, and LaNiO3-3wt%Ni3C composite measured in 0.25 M Na2S and Na2SO3 mixed aqueous solution.

H2-evolution kinetics

The electrocatalytic hydrogen evolution test of LaNiO3, LaNiO3-1wt%Pt, and LaNiO3-3wt%Ni3C samples were carried out in the potential range of -2.5 to 0.5 V vs. RHE with 0.25 M Na2S and Na2SO3 mixed aqueous solution (pH=12) as an electrolyte solution. The corresponding polarization curves are displayed in Fig. 5C. The LaNiO3-3wt%Ni3C sample exhibits the lowest overpotential than others, indicating that using Ni3C NPs as co-catalysts could decrease the onset potential, thereby improving the H2-evolution kinetics.

The hydrogen evolution activities of the samples were also evaluated by the Tafel analysis in the potential range of -3.5 to 1.5 V vs. RHE with 0.25 M Na2S and Na2SO3 mixed aqueous solution (pH=12). Fig. 5D shows the corresponding Tafel slope values of the composites, which refers to the overpotential required to raise the current density by one order of magnitude [44-46]. The Tafel slope of LaNiO3- 3wt%Ni3C composite (342 mV/dec) is lower than that of LaNiO3-1wt%Pt samples (462 mV/dec) and that of LaNiO3 samples (542 mV/dec), suggesting its faster electron and hole mass transfer during the decomposition of water [47]. The above results indicate that the catalytic H2-evolution kinetic of LaNiO3 can be substantially enhanced by loading with the Ni3C NPs.

The activities and stabilities for photocatalytic H2 evolution

The photocatalytic H2-evolution performance of LaNiO3, LaNiO3-1wt%Pt, and LaNiO3-3wt%Ni3C are shown in Fig. 6. The time courses of H2 production over the photocatalysts (Fig. 6A) all show a linear increase in the rates of H2 evolution during the photocatalytic reactions. To further compare the H2- evolution activity of different samples, the average H2-evolution rates of all photocatalyst during the whole reaction time were calculated and displayed in Fig. 6B. As shown in Fig. 6B, the H2 evolution rate of LaNiO3-3wt% Ni3C composite is 4093.9 μmolg -1 h -1 , which is about 5 times as that of the pristine LaNiO3 sample (819.2 μmolg-1 h -1 ) and about 2 times as that of LaNiO3-1wt%Pt composite (1927.0 μmolg-1 h -1 ). Loading a suitable amount of Ni3C co-catalysts on the LaNiO3 sample could significantly improve the photocatalytic H2-evolution activity of LaNiO3. The photocatalytic activities of other LaNiO3 host material are listed in table. S1 for reference.

The cycling tests were also performed to confirm the long-term stability and reproducibility of the LaNiO3-3wt%Ni3C composite for H2 evolution. Fig. 6C shows the H2-evolution reaction time courses of LaNiO3-3wt%Ni3C composite in every reaction cycle. After 30 hours, the hydrogen production performance of the LaNiO3-3wt%Ni3C composite decreased slightly, which indicates excellent stability and reproducibility of the LaNiO3-3wt%Ni3C composite for H2 evolution. The XRD pattern of the LaNiO3-3wt%Ni3C samples after 30 h H2 evolution is shown in Fig. S2 with the XRD pattern of the original sample for comparing. After the 30 h reaction, the intensity of the XRD peaks reduces a little, and the signal-to-noise ratio (SNR) of the XRD pattern becomes worse, both of which may come from the subtle variation of the crystallization due to the slight photocorrosion. On the other hand, no new miscellaneous peaks are observed in the after reaction XRD pattern, which indicates that the decrease of photocatalytic activity may be mainly ascribed to a slow fall-off of the Ni3C NPs co-catalyst from the LaNiO3 NPs.

Fig. 6. H2 evolution performance of the samples in 0.25 M Na2S and Na2SO3 mixed aqueous solution under Xe lamp Journal.irradiation (λ > 200 nm): (A) Reaction time courses. (B) the average rate of H2 evolution over different photocatalysts: (a) LaNiO3, (b) LaNiO3-1wt%Pt, and (c) LaNiO3-3wt%Ni3C. (C) Repeated time courses of photocatalytic H2 evolution over LaNiO3-3wt%Ni3C samples.

3.2. The variation of Ni3C contents in LaNiO3-Ni3C composite

It is not hard to predict that some kinds of damage will occur to the oxide photocatalyst when loading TMC co-catalyst on it by chemical method. Because the carbonaceous precursors usually can be reductive. Therefore, we synthesized a series of LaNiO3-Ni3C composites with the Ni3C content increasing from 1wt% to 6wt% to investigate the corresponding influence on both the structures, compositions, the morphology of the composite, and the photocatalytic H2 evolution reaction progress. The structures, compositions, the morphology and light absorption of the LaNiO3-Ni3C composites

Fig. 7 shows the XRD patterns of the LaNiO3-Ni3C composites, from 1wt% to 6wt% Ni3C. The XRD patterns are almost unchanged, with the Ni3C content lessthan 4wt%. On the contrary, a new peak appears at 43.3o , which becomes more evident as the content of Ni3C increases from 4wt% to 6wt%. This peak could be assigned to the (012) crystal-plane of NiO (JCPDS 89-7390) [23, 36]. Another new peak appears at 31.3o in the XRD pattern of LaNiO3-6wt% composite, which could be assigned to the (103) crystalplane of La2NiO4 (JCPDS 34-0314) [23, 48]. The XRD analysis indicates that vice products like NiO and La2NiO4 may be formed with the increase Ni3C content over 3wt%.

Fig. 7 the XRD patterns of the LaNiO3-Ni3C composites, from 1wt% to 6wt%. The LaNiO3-6wt%Ni3C composites are further characterized by XPS analysis. The C 1s XPS spectrum of the LaNiO3-6wt%Ni3C composite is shown in Fig. 8A. Compared with the LaNiO3-3wt%Ni3C composite sample (Fig. 2A), the Ni-C bond peak (283.2 eV) is enhanced, while the C-O bond peak (286.0 eV) is significantly reduced. The increase of Ni3C content can explain the enhancement of the Ni-C bond. While the reduction of the C-O bond peak should indicate a reduction of interaction between the Ni3C NPs and LaNiO3. Fig. 8B shows the XPS spectrum of O 1s of the LaNiO3-6wt%Ni3C composite. Compared with the LaNiO3-3wt%Ni3C composite, a new peak (529.4 eV) is observed, the lattice oxygen peak (528.4 eV) is enhanced, and the O-C peak (534 eV) is dismissed. The new peak at 529.4 eV could be attributed to the O-Ni bond, as reported in researches on NiO [23, 36, 49], which confirms the formation of NiO as depicted by the XRD analysis. The disappearing of the O-C peak (534 eV) is consistent with the significant reduction of the C-O peak (286.0 eV) in the C 1s spectrum. Simultaneously, the enhancement of the lattice oxygen peak (528.4 eV) may be ascribed to the reduction of interaction between the Ni3C NPs and LaNiO3 NPs, for which more lattice oxygen can be exposed on the LaNiO3 surface. In Fig. 8C, the four peaks at 833.5 eV, 837.3 eV, and 850.2 eV, 854.2 eV, could be assigned to the binding energies of La 3d5/2 and La 3d3/2, respectively, which are all moved to lower energy, compared with the LaNiO3-3wt%Ni3C composite. Especially the 833.5 eV peak has moved 0.8 eV from 834.3 eV to 833.5 eV. These shifts may confirm the formation of La2NiO4 in the composite, as depicted by XRD analysis. Because the shift of XPS peaks of a cation to lower energy usually indicates the reduction of valance state. In La2NiO4, La has a mixed valance state of +2 and +3, which is lower than the +3 valance state in LaNiO3. Fig. 8D shows the Ni 2p spectrum of LaNiO3-6wt%Ni3C samples. Compared with the LaNiO3-3wt%Ni3C composite, the peak at about 854.0 eV is significantly enhanced, while the peak at 856.3 eV is reduced remarkably. It should be noticed that Ni element with valance state +1 and +2 can both contribute to the peak at about 854.0 eV [19, 35, 38]. Therefore the enhancement of the 854.0 eV peak may come from the increased Ni3C content and the newly formed NiO in the composite. Meanwhile, the reduction of the 856.3 eV peak may suggest a decrease of +3 Ni in the composite.

The formation of NiO, the La 3d chemical state shifting, the reduction of interaction between LaNiO3 and Ni3C NPs, and the +3 Ni decrease all indicate that decomposition of LaNiO3 occurred at the surface with loading too much Ni3C NPs.

Fig. 8. XPS spectra of LaNiO3-6wt%Ni3C composites: (A) C 1s, (B) O 1s, (C) La 3d, (D) Ni 2p.

Fig. 9A and 9B show the SEM and TEM image of LaNiO3-6wt%Ni3C composites. The SEM and TEM images show that cubic-like nanoparticles in LaNiO3-3wt%Ni3C composites (Fig. 3D) have begun to fuse into larger particles with a dimension of 30-55 nm with a sphere-like shape. Also, small particles with a dimension below 10 nm appear on the surface of the larger particles in the LaNiO3-6wt%Ni3C composite. In the HRTEM image of LaNiO3-6wt%Ni3C composite (Fig. 9C), lattice fringes with interplanar spacings of 0.272 nm are observed on the large particles, which is corresponding to the (110) crystal plane of perovskite LaNiO3. Meanwhile, lattice fringes with interplanar spacings of 0.241 nm are observed on the surface-attached small particle, which is corresponding to the (111) crystal plane of cubic NiO [36]. Moreover, Ni3C NPs are also observed in the HRTEM image, showing the lattice fringes with the interplanar spacings of 0.158 nm (marked with cycle), which is corresponding to the (116) crystal plane of hexagonal Ni3C [40]. The SEM, TEM, and HRTEM analysis of LaNiO3-6wt%Ni3C composite not only confirms the formation of NiO but also depicts a significant shape reform of the LaNiO3 NPs,compared with the LaNiO3-3wt%Ni3C composite, which can not be explained by Ostwald ripening for the same heating process in the synthesis of 3wt% and 6wt% sample. Instead, the decomposition and recrystallization occurred at the surface of LaNiO3 NPs can be a more convincing explanation.

The surface decomposition of LaNiO3 NPs, along with the formation of NiO and La2NiO4, are revealed by XRD, XPS, and morphology analysis when more than 3wt%% Ni3C NPs are loaded. A possible explanation is that the acetate, which is from the nickel acetate in precursors as the carbon source in the Ni3C assembling process, shows reduction activity in the oleylamine solution. With loading too much Ni3C NPs, the concentration of acetate may be high enough to trigger the reduction and the followed disintegration of LaNiO3 at the surface [7, 15, 38, 40].

Fig. 9. (A-C) The SEM, TEM, and HRTEM images of LaNiO3-6wt%Ni3C composites.

Fig. S3 illustrated the UV-Vis absorption spectra of the LaNiO3-Ni3C composites, from 1wt% to 6wt%. The spectra shape is almost identical, while the absorption long-wave end variated, which may come from the variation of LaNiO3-Ni3C interaction with different Ni3C content. The bandgaps of the LaNiO3- Ni3C composites are also calculated by Kubelka–Munk method, as shown in Fig. S3. B. The bandgaps of the 1wt%, 2wt%, 3wt%, 4wt%, 5wt% and 6wt% LaNiO3-Ni3C composites are 1.90 eV, 1.89 eV, 1.92 eV, 1.99 eV, 1.98 eV and 1.95 eV, respectively. It seems that the bandgap of the LaNiO3-Ni3C composites variated subtly until loading Ni3C less than 3wt%, and enlargement of less than 0.07 eV can be observed with loading more Ni3C than 3w%

The charge-separation performances of the LaNiO3-Ni3C composites

The EIS tests of the LaNiO3-Ni3C composites were also conducted, and the corresponding Nyquist plots are shown in Fig. 10A. First, the semicircle radius decreases from 3100 to 1500 as the Ni3C content increase from 1wt% to 3wt% (the inset of Fig. 10A). However, the semicircle radius of the LaNiO3- 4wt%Ni3C composite shows a dramatic increase to about 8900. The composites with 5wt% and 6wt% Ni3C show semicircle radiuses larger than 25000 and 30000, respectively, which are even larger than that of pristine LaNiO3.

The LaNiO3-Ni3C composites are also analyzed by the transient photocurrent response test and shown in Fig. 10B. The average photocurrent intensity of the LaNiO3-Ni3C composites, from 1wt% to 6wt%, is about 0.23 μA/cm2 , 0.42 μA/cm2 , 0.61 μA/cm2 , 0.35 μA/cm2 , 0.2 μA/cm2 , and 0.18 μA/cm2 , respectively. The average photocurrent intensity of the LaNiO3-Ni3C composites increased gradually at first, reached the maximum at 3wt% Ni3C content, and then decreased significantly from 4wt% Ni3C. The average photocurrent intensity of the LaNiO3-5wt%Ni3C and LaNiO3-6wt%Ni3C samples are below 0.25 μA/cm2 , which is even lower than that of the pristine LaNiO3 sample (0.26 μA/cm2 , Fig. 5B).

Both characterizations suggest that 3wt% is likely to be the most appropriate Ni3C loading ratio for charge-separation performances.

H2-evolution kinetics of the LaNiO3-Ni3C composites

The electrocatalytic hydrogen evolution test of LaNiO3-Ni3C composites was carried out in the potential range of -2.0 to 0.5 V vs. RHE with 0.25 M Na2S and Na2SO3 mixed aqueous solution (pH=12). The corresponding polarization curves are displayed in Fig. 10C. The electrocatalytic H2-evolution overpotential decreases with loading more Ni3C until 3wt%, which shows the lowest overpotential than others. Then the overpotential increases dramatically with loading more Ni3C from 4wt% to 6wt%. Notably, 4wt%, 5wt%, and 6wt% Ni3C samples all show an overpotential higher than the pristine LaNiO3 samples.

The hydrogen evolution activities of LaNiO3-Ni3C composites were also evaluated by the Tafel analysis in the potential range of -3.5 to 1.5 V vs. RHE with 0.25 M Na2S and Na2SO3 mixed aqueous solution (pH=12). The corresponding Tafel slopes are displayed in Fig. 10D, which show similar variation as electrocatalytic H2-evolution overpotential. First, the Tafel slope decreases with loading Ni3C from 1wt% to 3wt% (1wt% 414 mV/dec, 2wt% 392 mV/dec, and 3wt% 342 mV/dec). Then increase with loading Ni3C from 4wt% to 6wt% (4wt% 532 mV/dec, 5wt% 596 mV/dec and 6wt% 645 mV/dec). Again the Tafel slope of the LaNiO3-5wt%Ni3C (596 mV/dec) and the LaNiO3-6wt%Ni3C (645 mV/dec) is higher than that of pristine LaNiO3 (542 mV/dec, Fig. 5D).

Both results suggest that loading 3wt% Ni3C can accelerate the electron and hole mass transfer during the decomposition of water mostly, which is likely to be the most appropriate Ni3C content for improving the H2-evolution kinetic of LaNiO3.

Fig. 10. (A) Nyquist plots of different working electrodes in 0.25 M Na2S and Na2SO3 mixed aqueous solution in the dark. (B) Transient photocurrent responses (I–t curves) of LaNiO3-xwt%Ni3C composite in 0.1 M Na2SO4 aqueous solution under Xe lamp irradiation with (λ > 200 nm) at 0.1 V vs. Ag/AgCl. (C) Polarization curves of the LaNiO3-xwt%Ni3C composites were measured at a scan rate of 5 mVs-1 in 0.25 M Na2S and Na2SO3 mixed aqueous solution. (D)Tafel plots of LaNiO3-xwt%Ni3C composites in 0.25 M Na2S and Na2SO3 mixed aqueous solution (x=1, 2, 3, 4, 5, 6).

The activities for photocatalytic H2 evolution of the LaNiO3-Ni3C composites

The time courses of H2 production over the LaNiO3-Ni3C composites (Fig. 11A) all show a linear increase in the rates of H2 evolution during the photocatalytic reactions. The H2 production rate of LaNiO3-Ni3C composites gradually increases with loading Ni3C NPs from 1wt% to 3wt%, and then gradually decreases with loading Ni3C NPs from 4wt% to 6wt%. To further compare the H2-evolution activity, the average H2-evolution rates of all samples during the whole reaction time were calculated and displayed in Fig. 11B. As shown in Fig. 11B, the H2 evolution rate increases dramatically with loading 1wt% Ni3C NPs (from 819.2 μmolg-1 h -1 for pristine LaNiO3, as shown in Fig. 6B, to 2732.6 μmolg-1 h - 1 ) and reaches the max with loading 3wt% Ni3C NPs (4093.9 μmolg-1 h -1 ). Then the rate decrease to 2981.2 μmolg-1 h -1 with loading 4wt% Ni3C. Significant decreases of H2 evolution rate are observed for LaNiO3-5wt%Ni3C and LaNiO3-6wt%Ni3C composites, the values of which are merely about 1200 μmolg-1 h -1 .

Fig. 11. (A) Reaction time courses. (B) the average rate of H2 evolution over different photocatalysts in aqueous solution: (d) LaNiO3-1wt%Ni3C, (e) LaNiO3-2wt%Ni3C, (c) LaNiO3-3wt%Ni3C, (f) LaNiO3-4wt%Ni3C, (g) LaNiO3-5wt%Ni3C and (h) LaNiO3-6wt%Ni3C.