In recent years, solar-driven photoelectrochemical (PEC) water splitting, which is an efficient, environmentally friendly, and sustainable technique, has attracted considerable attention [1, 2]. Improvements in the sunlight-absorption properties, conversion efficiencies, and catalytic activities of photoelectrode materials are crucial in the use of PEC for transforming solar energy into chemical energy [3, 4]. Various photoelectrode materials such as TiO₂ [5, 6], WO₃ [7, 8], and α-Fe₂O₃ [9, 10] have been widely studied and used for efficient production of hydrogen at the photocathodes of PEC cells. The use of BiVO₄ in PEC water splitting is increasing because of its small bandgap (E_g = 2.4–2.5 eV), which enables significant visible-light absorption, and its high valence band, which provides an overpotential that is sufficient for water oxidation [11-13]. However, weak electron–hole separation and a poor transport yield restrict the overall quantum efficiency in PEC processes when pure BiVO₄ is used as the photoanode [14].

To address these limitations, diverse strategies for modification of BiVO₄, such as doping [15, 16], addition of a cocatalyst [17, 18], and construction of heterojunction structures [19, 20] have been developed to improve the PEC properties of BiVO₄ photoanodes. Transition-metal (oxy)hydroxides such as Ni(OH)₂ [21], FeOOH [22], and NiFe layered double-hydroxide (NiFe-LDH) materials [23, 24] have been widely explored for use in electrochemical oxygen evolution reactions (OERs). An operando Mössbauer spectroscopic study showed that Fe⁴⁺ was formed in the electrochemical OER process and the active sites were Fe⁴⁺ species generated within an Fe-doped NiOOH lattice [25], suggesting that Fe plays a critical role. In situ surface-enhanced Raman spectroscopy showed that NiOO^- generated from Ni(Fe)OOH acted as an electrochemical OER precursor [26]. It is therefore believed that doping with Fe or Ni is an effective method for enhancing PEC activity.

In this work, a BiVO₄ film was prepared by electrochemical deposition, and used as a precursor in hydrothermal production of Ni(OH)₂/BiVO₄, Fe(OH)₂/BiVO₄, and NiFe-LDH/BiVO₄ films [27]. The NiFe-LDH/BiVO₄ composite gave the best PEC performance. The photocurrent response of NiFe-LDH/BiVO₄ in the high-potential region (> 0.5 V vs Ag/AgCl) was better than that of pure BiVO₄. Hydrogen production of 1260 μmol was achieved in 3 h under visible-light illumination; this is 2.5 times that produced with the pristine BiVO₄ photoanode. Ni can facilitate efficient hole transfer to the surface and Fe increases the light-absorption capacity. NiFe-LDH nanoparticles are therefore not only an excellent electrochemical catalyst, but also an excellent PEC catalyst, and these nanoparticles have potential applications in solar water splitting.

Scheme 1 shows a schematic diagram of the synthesis of the pure BiVO₄, Ni(OH)₂/BiVO₄, Fe(OH)₂/BiVO₄, and NiFe-LDH/BiVO₄ photoanodes. A BiVO₄ electrode was synthesized by a previously reported method [28]. In short, after electrodeposition of a BiOI film, vanadium was added, followed by calcination; excess V₂O₅ was removed by soaking, to give pure BiVO₄. The composite photoanodes were prepared by a traditional hydrothermal method. An aqueous solution (30 mL) containing Ni(NO₃)₂·6H₂O (0.1047 g), Fe(NO₃)₃·9H₂O (0.1454 g), and urea (0.2162 g) was transferred to a 50-mL Teflon-lined autoclave in which BiVO₄ was placed with the BiVO₄ side facing the wall of the autoclave; the mixture was heated at 120 ℃ for 6 h. The formed composite was thoroughly rinsed with purified water and dried at room temperature before use.

The morphologies of the photoelectrodes were examined using scanning electron microscopy (SEM) and high-resolution transmission electron microscopy (HRTEM); the images are shown in Fig. 1. The surface of the bare BiVO₄ photoanode consisted of nanoparticles of uniform size. Ni was loaded on the BiVO₄ surface in the form of nanospheres. After Fe deposition, the BiVO₄ particle size decreased, and NiFe-LDH nanoparticles were wrapped around the BiVO₄ particle surfaces. Fig. 1(e) shows that the NiFe-LDH nanoparticles were deposited on the BiVO₄ surface. The interplanar distances were estimated to be about 0.363 and 0.318 nm, corresponding to the (511) and (421) planes, respectively, of the NiFe-LDH nanoparticles. Fig. S1 confirms that a composite film was prepared.

The crystallographic structure of all the prepared photocatalysts were determined using X-ray diffraction (XRD; Fig. 2). The results confirm that all the BiVO₄ photoanodes have a single monoclinic scheelite structure. However, diffraction peaks corresponding to Ni(OH)₂, Fe(OH)₂, and NiFe-LDH nanoparticles were not observed. This may be because the contents of Ni(OH)₂, Fe(OH)₂, and NiFe-LDH nanoparticles were low, possibly below the XRD detection limit.

Ultraviolet-visible (UV-vis) spectroscopy was used to determine the bandgap values; the spectra are shown in Fig. 3. The absorption for the NiFe-LDH/BiVO₄ photoanode was slightly higher than that for the BiVO₄ photoanode. This can be attributed to a larger absorption efficiency and narrower bandgap in the case of the NiFe-LDH nanoparticles. Fig. S2 shows the photoluminescence (PL) emission spectra of the samples at an excitation wavelength of 350 nm. The PL intensity for the NiFe-LDH/BiVO₄ film was lower than that for the pure BiVO₄ film, showing that the electrons and holes in the NiFe-LDH/BiVO₄ film were easily separated.

Linear current-voltage curves for the BiVO₄, Ni(OH)₂/BiVO₄, Fe(OH)₂/BiVO₄, and NiFe-LDH/BiVO₄ photoanodes were recorded (Fig. 4). The photocurrent densities of the Ni(OH)₂/BiVO₄ and Fe(OH)₂/BiVO₄ films were 0.81 and 1.98 mA, respectively, at 0.6 V (vs Ag/AgCl). The BiVO₄ film coated with NiFe-LDH nanoparticles gave the highest photocurrent density, i.e., 2.49 mA, at 0.6 V (vs Ag/AgCl), and its photoactivity was 217% higher than that of the bare BiVO₄ film. Fig. 4(b) shows that the onset potentials of the BiVO₄ and NiFe-LDH/BiVO₄ photoanodes in water oxidation were 1.2 and 0.8 V (vs Ag/AgCl), respectively. The clear cathodic shift in its onset potential and overpotential in water oxidation compared with those for pure BiVO₄ indicate that the NiFe-LDH nanoparticles had a catalytic effect. Photogenerated holes accumulate on the BiVO₄ photoanode surface, resulting in significant surface charge recombination. Loading with NiFe-LDH nanoparticles could reduce the amount of accumulated holes, and this would prevent recombination during water oxidation; this is supported by the much less pronounced photocurrent transients (Fig. S3).

Electrochemical impedance spectroscopy (EIS) was used to evaluate the charge-transfer kinetics at the semiconductor-electrolyte interface (Fig. 5). The Nyquist plots for the BiVO₄ and NiFe-LDH/BiVO₄ electrodes were consistent with the linear sweep voltammetry results, with and without light irradiation. The semicircle for NiFe-LDH/BiVO₄ was smaller than that for the BiVO₄ electrode under both dark and illuminated conditions. This indicates that NiFe-LDH/BiVO₄ gave highly efficient electron and hole transport. Further analysis to clarify the mechanism of the PEC water-splitting reaction was performed by obtaining Motty-Schottky curves, shown in Fig. S4. Fitting of the linear parts of the plots showed that the carrier density in the BiVO₄ electrode was increased by surface modification with NiFe-LDH nanoparticles, and superior PEC hydrogen activity was achieved as a result of effective separation and transfer of charge carriers.

Fig. 6 shows that hydrogen generation with the NiFe-LDH/BiVO₄ photoanode under visible-light illumination for 3 h was higher than that with the pure BiVO₄ electrode. The reactor was filled with flowing argon gas for 30 min to exclude air from the system. The PEC reaction was then started with the assistance of light from a light-emitting diode lamp (CEL-LED100). A 1-mL injector was used to extract the evolved hydrogen for analysis with a gas chromatography (GC-9560) online system at set intervals of 30 min during light irradiation. Fig. 6 shows that the NiFe-LDH/BiVO₄ photoanode gave an excellent PEC water-splitting performance. Hydrogen production was 1260 μmol in 3 h under light illumination; this is 2.5 times that produced with the pristine BiVO₄ photoanode. Furthermore, the NiFe-LDH/BiVO₄ photoanode showed excellent stability compared with the bare BiVO₄ photoanode. NiFe-LDH/BiVO₄ therefore has potential applications in the development of new solar water-splitting techniques.

A possible mechanism for the reaction at the NiFe-LDH/BiVO₄ photoelectrode is proposed on the basis of the above results and previously reported results (Scheme 2). First, the NiFe-LDH nanoparticles promote visible-light absorption by NiFe/BiVO₄. The NiFe-LDH nanoparticles tend to shift to the positive potential region and generate positive charges, and then accept electrons from the semiconductor to revert to their original state. This is similar to cocatalysis with Ag and Bi [29, 30]. The transfer of holes from the semiconductor to the metal suppresses recombination of the photogenerated electrons and holes, and this increases the lifetimes of the carriers.

In summary, a high-performance BiVO₄ photoanode with a NiFe-LDH nanoparticle cocatalyst was developed for PEC water splitting. The photocurrent density and hydrogen production achieved with the optimized NiFe-LDH/BiVO₄ were respectively approximately 217% and 2.5 times the values achieved with pristine BiVO₄. The NiFe-LDH nanoparticles facilitate efficient hole transfer to the surface and enhance the light-absorption capacity. The optimized NiFe-LDH/BiVO₄ photoanode also showed excellent stability in PEC water splitting.