Surface modification by ligand growth strategy for dense copper bismuth film as photocathode to enhance hydrogen production activity

Abstract

Hydrogen production from photoelectrochemical (PEC) water splitting has been regarded as a promising way to utilize renewable and endless solar energy. However, semiconductor film grown on photoelectrode suffers from numerous challenges, leading to the poor PEC performance. Herein, a straightforward sol-gel method with the ligand-induced growth strategy was employed to obtain dense and homogeneous copper bismuthate photocathodes for PEC hydrogen evolution reaction. By various characterizations, it was found that the nucleation and surface growth of CuBi₂O₄ layer induced by 2-methoxyethanol ligand (2-CuBi₂O₄) demonstrated a decent crystallinity and coverage, as well as a large grain size and a low oxygen vacancy concentration, leading to the good ability of light absorption and carrier migration. Consequently, under simulated sunlight irradiation (AM1.5G, 100 mW/cm²), the 2-CuBi₂O₄ photocathode achieved an enhanced photocurrent density of −1.34 mA·cm⁻² at 0.4 V versus the reversible hydrogen electrode and a promising applied bias photon-to-current efficiency of 0.586%. This surface modification by ligand growth strategy will shed light on the future design of advanced photoelectrodes for PEC water splitting.

Graphical abstract

Keywords

copper bismuthate / photocathode / ligand growth strategy / dense film / PEC

1 Introduction

Hydrogen (H₂), due to its high energy density and zero-pollution end product (H₂O), has been regarded as one of the most promising candidates to address the energy crisis caused by excessive carbon footsteps [1–4]. Currently, the primary sources of hydrogen come from steam methane reforming (gray hydrogen) and coal vaporization (black hydrogen), which triggers heavy greenhouse gas emissions [5–8]. It is thus a wide summons for “green hydrogen” via the renewable and carbon-free energy sources (e.g., solar, wind, and tide). Among the technology for H₂ generation, photoelectrochemical (PEC) hydrogen production deciphers the predicament of sluggish hydrogen evolution rates for photochemical (PC) water splitting and laborious reliance on electricity for electrochemical (EC) water splitting, respectively [9–12]. Yet, the response ability of semiconductor for the light absorption, charge transport, and transfer critically remains a problem for efficient PEC performance.

Many metal oxide semiconductors, including TiO₂ [13], Fe₂O₃ [14], ZnO [15], WO₃ [16], BiVO₄ [17], and CuBi₂O₄ [18], have been explored for PEC water splitting with a favorable PC performance. However, most photoelectrodes reported for PEC water splitting so far still endure numerous challenges, such as the growth of non-dense catalysts that are particularly vulnerable to intrinsic crystal defects (point defects, pinholes, and grain boundaries), and photogenerated electron traps readily inducing the recombination of photogenerated carriers, which lead to a low PEC performance [19]. Moreover, the porous microstructure of the semiconductor grown on the photoelectrode would lead to the low coverage surface, which was unfavorable for charge transfer. Therefore, the rational design of the dense microstructure of the photoelectrodes is crucial to boost PEC activity. Recently, Li and coworkers [11] proposed a molecular coordination engineering strategy to regulate the nucleation and growth of lead chromate photoanodes, improving the charge separation efficiency from 47% to 90%. Decorated with Co-Pi co-catalyst, a photocurrent density of photoanode reached 3.15 mA/cm² at 1.23 V versus the reversible hydrogen electrode (RHE) under simulated AM1.5G illumination and an applied bias photo-to-current efficiency (ABPE) of 0.82% [11].

As a prospective photocathode material, spinel-type copper bismuth (CuBi₂O₄) has attracted a lot of attention due to its composition of affordable, non-toxic, and earth-abundant elements [20–22]. Eminently, the small bandgap (1.6–1.8 eV) and the suitable band-edge position of CuBi₂O₄ are favorable for water reduction to hydrogen [23–25]. Notably, the high internal photovoltage of CuBi₂O₄ denotes a high theoretical photocurrent density. However, the reported photocurrent density to date is vastly diminished than expected. In addition, most studies have neglected the grown process and nucleation mechanism of catalysts, which is directly related to the performance of the semiconductor materials due to the structure-activity relationship. For instance, the conventional sol-gel methods for CuBi₂O₄ film growth employing ethylene glycol [26–29], ethanol [30], or acetic acid with polyvinyl pyrrolidone [31] as the inducer lead to the decent control of nucleation and growth, which makes the photoelectrode film inhomogeneous with poor substrate contact and small grain size. Lee and coworkers [32] designed an all-bismuth-based oxide tandem cell for solar overall water splitting, where the optimum CuBi₂O₄ photocathode exhibited a decent PEC hydrogen evolution performance. However, the exploration of the growth mechanism of semiconductor film is missing inside.

Herein, a ligand growth strategy was developed by employing 2-methoxyethanol (2-Me) and nitric acid to deposit a dense crystalline CuBi₂O₄ photocathode (2-CuBi₂O₄) with large grain size on fluorine-doped tin oxide (FTO) to enhance the PEC hydrogen evolution reaction. Incorporated with various spectroscopic characterizations, the 2-CuBi₂O₄ film layers grown by the ligand strategy displayed a better crystallinity and denseness, demonstrating a higher light trapping ability and a faster charge migration rate. Additionally, the low oxygen vacancy (O_V) defect concentration suppressed photogenerated carrier recombination. For PEC hydrogen evolution without any scavenger additions, the resultant 2-CuBi₂O₄ photocathode achieved a high photocurrent density of −1.34 mA/cm² at 0.4 V (versus RHE) and a high ABPE of 0.586% under simulated AM1.5G (100 mW/cm²), indicating that the ligand growth strategy is a promising method for fabricating premium photocathode film.

2 Materials and methods

2.1 Materials

Except for the 2-methoxyethanol (C₃H₈O₂, AR) which were purchased from Aladdin, the ethylene glycol (C₂H₆O₂, AR), nitric acid (HNO₃, AR), bismuth (III) nitrate (Bi(NO₃)₃·5H₂O, AR), copper (II) nitrate (Cu(NO₃)₂·3H₂O, AR), ethanol (C₂H₆O, AR), sodium hydroxide (NaOH, AR) were purchased from Sinopharm Chemical Reagent Co., Ltd. All raw materials were used without any further purification.

2.2 Photocathode preparation

2.2.1 Synthesis of 2-CuBi₂O₄ film photocathode

First, 4 mmol of Bi(NO₃)₃·5H₂O was dissolved in 1 mL of concentrated nitric acid and 9 mL of 2-Me, and the mixture solution was stirred until complete dissolution. Subsequently, 2 mmol of Cu(NO₃)₂·3H₂O was added before continued stirring to finally obtain a light blue clear solution. Meanwhile, a piece of FTO (1 cm × 2.5 cm) was cleaned by repeated ultrasound with ethanol and deionized water, and the above solution was dropped onto the surface of FTO with an amount of 5 μL/cm². Then, the pre-heating process was performed at 250 °C for 1 h on a heating plate and naturally cooled to room temperature. Finally, the FTO was placed in a muffle furnace and heated at 550 °C for 2 h.

2.2.2 Preparation of E-CuBi₂O₄ film photocathode

The procedure for the preparation of E-CuBi₂O₄ film photocathode was the same as that described in Section 2.2.1, except for replacing concentrated nitric acid and 2-Me with 10 mL of ethylene glycol.

In the above two processes, after the pre-heating process, the CuBi₂O₄ photocathodes were referred to as Pre-2-CuBi₂O₄ and Pre-E-CuBi₂O₄, respectively.

2.3 Characterizations of materials

The X-ray diffraction (XRD) data patterns were measured for the crystal structure on an X’Pert PRO diffractometer with Cu Kα radiation. The characteristic peak of (211) was employed to determine crystallite size (D) using Scherrer equation (Eq. (1)) [33]:

(1)D=\frac{K\lambda }{\beta \mathrm{c}\mathrm{o}\mathrm{s}\theta }\times 100\mathrm{\%},

where K is shape factor (0.94), λ is X-ray wavelength (0.15406 nm), β is full width at half maxima (FWHM) in rad units, and θ is the Bragg angle of XRD pattern. The surface morphology of the photoelectrodes was observed by scanning electron microscope (SEM, JEOL JSM-7900) coupled with X-ray energy dispersive spectroscopy (EDS, Oxford Ultim Max 65). Raman spectra were collected on HORIBA LabRAM Spectrometer with a laser of 532 nm. The surface chemical composition state was tested through X-ray photoelectron spectroscopy (XPS) by an XSAM800. The ultraviolet-visible diffuse reflection (UV-Vis) spectra were observed by a double-beam UV-Vis spectrophotometer (Lambda 750 S) using Ba₂SO₄ as the reflectance standard. Photoluminescence spectra were collected by a FLS1000.

2.4 Electrochemical measurements

All EC measurements were performed on a CHI660 EC workstation. A three-electrode system was carried out with 0.5 mol/L NaOH electrolyte, where Ag/AgCl (saturated KCl solution) electrode and Pt plate electrode were used as a reference and counter electrode, respectively. The potential verse Ag/AgCl electrode was translated verse RHE through Nernst equation (Eq. (2)):

(2)E_{\mathrm{R}\mathrm{H}\mathrm{E}}= E_{\mathrm{A}\mathrm{g}/\mathrm{A}\mathrm{g}\mathrm{C}\mathrm{l}}+ 0.059\times \mathrm{p}\mathrm{H}+0.209,

where pH was determined as 13.5. EC impedance spectroscopy (EIS) was examined from 10⁻¹ to 10⁶ Hz with a voltage amplitude of 0.005 V under open-circuit. The double-layer capacitance values were obtained via cyclic voltammetry (CV) curves with different scan rates in the non-Faraday region.

2.5 Photoelectrochemical hydrogen evolution reaction test

The PEC hydrogen evolution reaction was performed by the above three-electrode system on a CHI660 EC workstation. The working area was 1 cm × 1 cm at room temperature. Before the test, Ar was purged into the chamber to remove the air for 30 min. The linear sweep voltammetry (LSV) curves were determined at a scan rate of 50 mV/s with a range from 1.2 to 0.3 V (versus RHE). Unless obviously mentioned, all PEC tests were performed under illumination of simulated AM1.5G (100 mW/cm²). The gaseous H₂ product was detected by a gas chromatograph (GC-2014AT) equipped with a thermal conductivity detector (TCD) for analysis. Mott–Schottky analysis was conducted for the semiconductor type and flat-band potential. For the straight line of the Mott–Schottky curve, Eq. (3) was applied:

(3)\frac{1}{{C_{\mathrm{s}}}^2}=\frac{2}{A_{\mathrm{s}}^2\mathrm{e}{N}_{\mathrm{A}}{\epsilon }_0{\epsilon }_{\mathrm{r}}}\left(V_{\mathrm{a}\mathrm{p}\mathrm{p}\mathrm{l}\mathrm{i}\mathrm{e}\mathrm{d}}-V_{\mathrm{f}\mathrm{b}}-\frac{kT}{\mathrm{e}}\right),

where C_s is the interfacial capacitance, A_s is the surface area of the film, N_A is the acceptor density, ε₀ is the permittivity of vacuum, ε_r is the dielectric constant (80), V_applied is the applied potential, V_fb is the flat-band potential, k is the Boltzmann constant, and T is the temperature. The ABPE was calculated by Eq. (4) [34]:

(4)\mathrm{A}\mathrm{B}\mathrm{P}\mathrm{E}\left(\mathrm{\%}\right)= \frac{J\times \left(1.23-V_{\mathrm{a}\mathrm{p}\mathrm{p}\mathrm{l}\mathrm{i}\mathrm{e}\mathrm{d}}\right)}{P_{\mathrm{l}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}}}\times 100 \mathrm{\%},

where J is the photocurrent density, V_applied is the applied potential, and P_light is the light intensity (100 mW/cm²).

3 Results and discussion

3.1 Fabrication and characterization of CuBi₂O₄ film photocathode

The ligand coordination engineering strategy for surface modification was introduced to grow CuBi₂O₄ photoelectrode films with well-stitched grain boundary morphology (please refer to Section 2 for details.). Compared with the CuBi₂O₄ photoelectrode induced by the most used ligand of ethylene glycol (E-CuBi₂O₄), the introduction of 2-Me prompted delightful nucleation growth of Bi (III) and Cu (II) elements. The SEM images shown in Fig.1(a) and Fig.1(b) manifested that the surface of 2-CuBi₂O₄ film demonstrated a higher coverage and a larger grain size, which was in contrast to that of the E-CuBi₂O₄ film interspersed with pores of about 100 nm (Fig.1(d) and Fig.1(e)). In addition, the growth thicknesses of both two CuBi₂O₄ films were similarly about 450 nm (Fig.1(c) and Fig.1(f)), indicating that the difference of the HER activity between 2-CuBi₂O₄ and E-CuBi₂O₄ during subsequent PEC test was probably independent of the loading mass of film. Furthermore, the energy dispersive spectrum (EDS) revealed the uniform distribution of each element of the 2-CuBi₂O₄ film on the substrate FTO (Fig.1(g)), while the atomic ratio of Cu to Bi element was nearly 1/2 (Tab.1). XRD spectroscopy (Fig.2(a)) in conjunction with Raman spectra (Fig.2(b)) were conducted, suggesting that both 2-CuBi₂O₄ and E-CuBi₂O₄ could be well indexed to crystallographic copper bismuth and no impurity phase was detected except for the signal from FTO. Typically, the signal of XRD peak at 28.0 degree was attributed to the (211) plane of CuBi₂O₄. It was worth noting that the signal intensity of the CuBi₂O₄ film grown with 2-Me ligand was closely three times that of CuBi₂O₄ grown with EG ligand. Moreover, a smaller FWHM of XRD peak corresponding to 2-CuBi₂O₄ indicated its higher crystallinity. According to the semi-quantification analysis via Scherrer equation (Tab.2), the crystalline size of 2-CuBi₂O₄ with a higher crystallinity was larger than that of E-CuBi₂O₄. The large crystalline size could ensure a high coverage on the surface of CuBi₂O₄ film, which was in accord with the statistics of grain coverage on substrate FTO (Fig.2(c)), which could also promote full absorption of light by semiconductor. Meanwhile, there were four characteristic peaks with Raman spectrum located at 126, 261, 405, and 588 cm⁻¹, corresponding to CuBi₂O₄ species. In addition, the 2-CuBi₂O₄ film exhibited a larger grain size with a superior homogeneity (Fig.2(d)), which might promote the light absorption, charge migration, and transportation.

3.2 Optical properties and electrochemical tests

The light absorption properties of semiconductor materials play an important role for subsequent PEC hydrogen evolution activity. Based on it, the ultraviolet-visible (UV-Vis) diffuse reflection spectra were conducted to establish the intrinsic optical properties of 2-CuBi₂O₄ and E-CuBi₂O₄. As shown in Fig.3(a), both the two CuBi₂O₄ films exhibited distinct absorption peaks in the visible region with approximate positions of the absorption peaks, revealing that both two samples could response to a wide visible absorption range. Notably, the strengthened light absorption signal was observed for the 2-CuBi₂O₄ film, indicating the more photos absorption under light illumination, which was favorable for PEC reaction. The intrinsic optical properties of light collection could be attributed to the fact that the surface of 2-CuBi₂O₄ film featured a higher coverage and a larger grain size compared to that of E-CuBi₂O₄, mitigating the depletion of illumination. In addition, the bandgap of the 2-CuBi₂O₄ and E-CuBi₂O₄ films were calculated as 1.75 and 1.70 eV, respectively (Fig.3(b)), which were consistent with the theoretical values (1.6–1.8 eV) [35]. The intrinsic optical properties of light collection could be attributed to the fact that the surface of 2-CuBi₂O₄ film featured a higher coverage compared to that of E-CuBi₂O₄, mitigating the depletion of illumination.

The flat-band potential and semiconductor type of two samples were determined by Mott–Schottky curves (Fig.3(c)). The curves for the two CuBi₂O₄ photocathodes displayed a negative slope, indicating that they were p-type semiconductors [36]. Accordingly, the flat-band potentials of 2-CuBi₂O₄ and E-CuBi₂O₄ were 1.30 and 1.36 V (versus RHE), respectively. Based on the previous reports, which suggested that the valance band of the p-type semiconductors exhibited a potential of +0.2 V higher than the flat-band potential [37,38], the calculated valence band potentials of 2-CuBi₂O₄ and E-CuBi₂O₄ were 1.50 and 1.56 V (versus RHE), respectively. Summarily, the band structure of two semiconductors was manifested in Fig.3(d), expressing the analogical energy band structure with band-edge positions. It was worth noting that the conduction band potentials were both lower than the theoretical potential of 0 V (versus RHE) for hydrogen production, thus the separated photo-generated electrons had a sufficient redox ability for driving the HER.

The PEC performance depends heavily on the EC properties of the photoelectrodes. Thus, the EIS was used to ascertain the difference between the conductivity of photoelectrode under dark and light condition. According to Fig.4(a), CuBi₂O₄ is not able to be utilized as the catalyst directly for EC water splitting because of its poor electrical conductivity without light stimulation. By contrast, the inducing of light contributed to the sharp decrease of the EC impedance of both the two CuBi₂O₄ photoelectrodes, which was on account of the enormous number of photo-generated carriers promoting the charge transfer. Nevertheless, 2-CuBi₂O₄ exhibited a lower EC impedance when exposed to light, indicating its enhanced conductivity and carrier movement over that of E-CuBi₂O₄. In addition, to account for the variation in active sites between the two photoelectrodes, the EC active specific area (ECSA) was measured, which was proportional to the double capacitance (C_dl) (Fig.4(b)–Fig.4(d)). The fitting curves demonstrated that the C_dl for E-CuBi₂O₄ was 9.096 μF/cm², which was higher than the C_dl for 2-CuBi₂O₄ of 8.101 μF/cm². This was contributed to the more pore structure and smaller grain size of E-CuBi₂O₄. However, the actual PEC performance of the semiconductor photoelectrode is not necessarily resulted from the high ECSA. In fact, the intrinsic activity of the catalytic site also authorities the PEC hydrogen evolution performance.

3.3 Origin of PEC performance and mechanism of crystal growth

XPS tests were used to probe the electronic structure and elemental states of CuBi₂O₄ samples. The XPS survey spectra shown in Fig.5(a) demonstrated the presence of Cu, Bi, and O elements, which was consistent with the EDS results. Concerning the high resolution XPS spectra of O 1s (Fig.5(b)) for the 2-CuBi₂O₄ and E-CuBi₂O₄ samples, it was found that the lattice oxygen (O_L) and the oxygen vacancy (O_V) were detected for both samples at 529.7 and 531.1 eV [11,39], respectively, while the signal of O_V for 2-CuBi₂O₄ film was much lower than that of E-CuBi₂O₄. This was further verified by electron paramagnetic resonance (EPR) spectroscopy shown in the inset of Fig.5(b), which demonstrated that the oxygen vacancy concentration of E-CuBi₂O₄ was much higher than that of 2-CuBi₂O₄. Since oxygen vacancy in semiconductor materials tended to work as the recombination centers for photogenerated electrons and holes, a lower oxygen vacancy content would lead to a longer lifetime of photogenerated charges, which was favorable for photocatalytic reactions [40,41]. Fig.5(c) conveyed the XPS fine spectra of Bi 4f for the two samples, which exhibited comparable characteristic peak signals, indicating that the electronic structures of the Bi elements were similar in both the samples. A small shift of about 0.1 eV toward the high binding energy of the corresponding characteristic peaks for 2-CuBi₂O₄ was observed, due to the slump oxygen vacancy contents. Likewise, the XPS fine spectra of Cu 2p for the 2-CuBi₂O₄ and E-CuBi₂O₄ samples (Fig.5(d)) demonstrated analogous characteristic peak signals with nearly no shift, revealing that the electronic structures of the Cu elements in the two samples were not significantly divergent. These results from XPS analysis indicated the CuBi₂O₄ grown with 2-Me had a low oxygen vacancy concentration, which ensured the migration but not recombination of photogenerated carriers. Subsequently, photoluminescence (PL) spectra were detected to probe the photo-responded properties of the two CuBi₂O₄. As shown in Fig.5(e), the steady-state PL spectra of both samples exhibited an emission peak at 690 nm, which was in good agreement with their band gap values. Furthermore, a time-resolved PL spectra were also tested (Fig.5(f)), demonstrating that 2-CuBi₂O₄ exhibited a longer carrier lifetime than E-CuBi₂O₄, while the longer carrier lifetime implied a lower recombination rate of photogenerated electron-hole pairs. These results suggested that 2-CuBi₂O₄ absorbed more incident light and produced more photogenerated electron-hole pairs with a lower oxygen vacancy concentration than E-CuBi₂O₄, and the photogenerated electron-hole pair recombination of 2-CuBi₂O₄ was more mitigatory, which were all favorable for PEC hydrogen evolution. Notably, the 2-CuBi₂O₄ photocathode exhibited better PEC properties, which might be inseparable from the dense nucleation growth process due to the structure-activity relationship.

Apparently, 2-CuBi₂O₄ exhibited better PEC properties, which might be inseparable from the dense nucleation growth process. The process of photocathode preparation consisted of two heating steps. The first pre-heating step was conducted on a heating plate (the samples were labeled as “Pre-2-CuBi₂O₄” and “Pre-E-CuBi₂O₄”, respectively.) and the second calcination step took place in a muffle furnace. The intermediate products in different growth procedures were investigated by XRD and Raman. As shown in Fig.6(a), there were no obvious diffraction peaks except for those peaks belonging to the substrate FTO for the Pre-2-CuBi₂O₄ and Pre-E-CuBi₂O₄ samples, indicating that the temperature for orderly crystal growth was not reached during this procedure. Additionally, the second high-temperature calcination process resulted in the appearance of sharp XRD peaks belonging to CuBi₂O₄, e.g., (211) plane at 28.0 degree, etc. Subsequently, the as-obtained samples were detected by Raman spectrum (Fig.6(b)). Surprisingly, the characteristic peaks located at 1330 and 1600 cm⁻¹ appeared in the Raman signals of the Pre-E-CuBi₂O₄ sample, which were attributed to the D and G peaks, respectively, while the former corresponding to the defective carbon and the latter corresponding to the graphite carbon for in-plane stretching vibrations of the sp₂ hybridization of C atom [42–44]. These results proved that the surface of Pre-E-CuBi₂O₄ contained a large amount of carbon footprint after the pre-heating in the first step, which significantly affected the subsequent lattice growth of CuBi₂O₄ [45]. On the other hand, the Pre-2-CuBi₂O₄ sample did not show obvious D and G peak signals, indicating its excellent eventual nucleation and growth to a dense surface, which had been displayed in SEM results. To further verify that the surface of Pre-E-CuBi₂O₄ included a vast carbon footprint, the elemental constituent of two samples were also tested by EDS, as shown in Tab.3. Typically, the content of C element for the Pre-E-CuBi₂O₄ sample was excessive, while there was no C element on the Pre-2-CuBi₂O₄ sample, which was consistent with the Raman results.

The above results proved that there were enormous carbon-based materials on the surface of the EG-induced grown CuBi₂O₄ after the pre-heating step, which seriously restricted the orderly growth of the crystals during the subsequent calcination procedure. Moreover, during heating in the furnace, these carbon-based products were decomposed toughly, which might be the reason for the pore structures, small grain size, and low surface coverage of E-CuBi₂O₄ samples [46]. Meanwhile, this procedure also led to the deterioration of E-CuBi₂O₄ crystallinity and the enhanced oxygen vacancy concentration, eventually culminating in the inferior PEC performance. In contrast, concentrated nitric acid and 2-Me were used as inducer reagents in the preparation of 2-CuBi₂O₄ film photocathode, which were comfortable to evaporate and decompose. This favorable ligand growth strategy thus slightly affected the nucleation and crystallization process of 2-CuBi₂O₄, leading to the ultimate generation of large grain size, high surface coverage, low oxygen vacancy content, and excellent PEC properties and hydrogen production performance.

3.4 Photoelectrochemical performance

The potential of 2-CuBi₂O₄ as a photocathode for PEC hydrogen evolution reaction was investigated. To eliminate the effect of concentrated nitric acid introduced during the growth on the performance of CuBi₂O₄ samples, the LSV curves of hydrogen evolution performance were tested, while the control samples using ethylene glycol (E-CuBi₂O₄) as an inducer without/with adding the concentrated nitric acid (N-E-CuBi₂O₄) were prepared for photocathode. As shown in Fig.7(a), the photocurrent density of N-E-CuBi₂O₄ was slightly lower than that of E-CuBi₂O₄, demonstrating that the presence of concentrated nitric acid did not greatly affect the subsequent PEC tests. Fig.7(b) displayed the LSV curves corresponding to the various samples induced by different concentrations of 2-Me ligand (C_x-2-CuBi₂O₄, where x was the concentrations radio of the precursors for Cu (II) salts to Bi (III) salts). Since the thickness of the prepared CuBi₂O₄ films was correlated with the concentration of the precursors, the thickness of the 2-CuBi₂O₄ film photocathode was thinner as the lower precursor concentration, giving rise to the smaller photocurrent density. As the precursor concentration increased, the photocurrent density increased and was eventually in nearly invariance [35]. Fig.7(c) showed the LSV curves of E-CuBi₂O₄ and 2-CuBi₂O₄ under dark and light conditions. When testing under dark condition, both two photocathodes exhibited weak current signals, which was consistent with the EIS results. In sharp contrast, once the light was introduced, the photogenerated charge with strong redox ability could drive the water reduction to hydrogen evolution and facilitate the enhanced photocurrent densities for 2-CuBi₂O₄ and E-CuBi₂O₄ photocathodes, certifying the significance of illumination. Notably, the photocurrent density of 2-CuBi₂O₄ was significantly larger than that of E-CuBi₂O₄ at the same potential. In addition, the photocurrent density of 2-CuBi₂O₄ at 0.4 V (versus RHE) was about −1.34 mA/cm², approximately twice that of E-CuBi₂O₄ (−0.57 mA/cm²), which revealed that the ligand-induced strategy was promising for the design of photocathode (Tab.4). Likewise, the ABPE of the samples was also calculated, and as shown in Fig.7(d), the highest ABPE of 2-CuBi₂O₄ film photocathode reached an advanced value of 0.586%, whereafter, during the photocurrent stability test (Fig.7(c)), the photocurrent intensity of 2-CuBi₂O₄ was always higher than that of E-CuBi₂O₄ with a decent stability. These results indicated that the ligand-induced growth of dense and highly crystalline film photocathodes exhibited excellent optical and PEC properties, which could be attributed to the better light absorption and carrier transfer. To verify the actual hydrogen production capability of 2-CuBi₂O₄ film photocathode, a real-time test of chronoamperometry curve at 0.4 V (vs. RHE) for hydrogen production was performed (Fig.7(f)), demonstrating that the amount of hydrogen production for 2-CuBi₂O₄ photoelectrode was 0.173, 0.166, and 0.165 µmol during the 1st, 2nd, and 3rd hours, respectively, which suggested the decent actual potential for 2-CuBi₂O₄.

4 Conclusions

In summary, the surface modification by the ligand growth strategy was exploited to prepare the compact 2-CuBi₂O₄ film photocathode on FTO with a large grain size for the enhancement of PEC hydrogen evolution reaction. Due to few carbon residues on the substrate surface during growth, the as-synthetic dense 2-CuBi₂O₄ film layers disclosed a better crystallinity, a more sufficient light harvest ability, and a faster charge migration, together with suppressed photogenerated carrier recombination. The 2-CuBi₂O₄ photocathode prepared with ligand growth strategy exhibited a high photocurrent density of −1.34 mA/cm² at 0.4 V (versus RHE under simulated AM1.5G) and a promising ABPE of 0.586%. This work demonstrated the prominence of the dense crystalline photocathode via ligand-induced growth strategy and advanced the development of PEC water splitting.

References

[1] Zhou P, Navid I A, Ma Y. . Solar-to-hydrogen efficiency of more than 9% in photocatalytic water splitting. Nature, 2023, 613(7942): 66–70

[2] Huang C, Zhou Q, Duan D. . The rapid self-reconstruction of Fe-modified Ni hydroxysulfide for efficient and stable large-current-density water/seawater oxidation. Energy & Environmental Science, 2022, 15(11): 4647–4658

[3] Song H, Luo S, Huang H. . Solar-driven hydrogen production: Recent advances, challenges, and future perspectives. ACS Energy Letters, 2022, 7(3): 1043–1065

[4] ZhouBSun S. Approaching the commercial threshold of solar water splitting toward hydrogen by III-nitrides nanowires. Frontiers in Energy, 2023, online, https://doi.org/10.1007/s11708-023-0870-z

[5] Guerra O J, Eichman J, Kurtz J. . Cost competitiveness of electrolytic hydrogen. Joule, 2019, 3(10): 2425–2443

[6] Terlouw T, Bauer C, McKenna R. . Large-scale hydrogen production via water electrolysis: A techno-economic and environmental assessment. Energy & Environmental Science, 2022, 15(9): 3583–3602

[7] Hosseini S E, Wahid M A. Hydrogen production from renewable and sustainable energy resources: Promising green energy carrier for clean development. Renewable & Sustainable Energy Reviews, 2016, 57: 850–866

[8] Zhou Q, Liao L, Zhou H. . Innovative strategies in design of transition metal-based catalysts for large-current-density alkaline water/seawater electrolysis. Materials Today Physics, 2022, 26: 100727

[9] Rashid R T, Chen Y, Liu X. . Tunable green syngas generation from CO₂ and H₂O with sunlight as the only energy input. Proceedings of the National Academy of Sciences of the United States of America, 2022, 119(26): e2121174119

[10] Yu L, Wu L, McElhenny B. . Ultrafast room-temperature synthesis of porous S-doped Ni/Fe (oxy)hydroxide electrodes for oxygen evolution catalysis in seawater splitting. Energy & Environmental Science, 2020, 13(10): 3439–3446

[11] Zhou H, Zhang D, Gong X. . A dual-ligand strategy to regulate the nucleation and growth of lead chromate photoanodes for photoelectrochemical water splitting. Advanced Materials, 2022, 34(29): 2110610

[12] Wang Z, Sheng B, Chen Y. . Photocatalytic syngas production from bio-derived glycerol and water on AuIn-decorated GaN nanowires supported by Si wafer. Green Chemistry, 2023, 25(1): 288–295

[13] Li H, Liu H, Wang F. . Hot electron assisted photoelectrochemical water splitting from Au-decorated ZnO@TiO₂ nanorods array. Nano Research, 2022, 15(7): 5824–5830

[14] Masoumi Z, Tayebi M, Kolaei M. . Efficient and stable core-shell α–Fe₂O₃/WS₂/WO_x photoanode for oxygen evolution reaction to enhance photoelectrochemical water splitting. Applied Catalysis B: Environmental, 2022, 313: 121447

[15] Bagal I V, Jun S, Choi M. . Investigation of charge carrier dynamics in beaded ZnO nanowire decorated with SnS₂/IrO_x cocatalysts for enhanced photoelectrochemical water splitting. Applied Surface Science, 2023, 613: 156091

[16] Kong H, Yang H, Park J S. . Spatial control of oxygen vacancy concentration in monoclinic WO₃ photoanodes for enhanced solar water splitting. Advanced Functional Materials, 2022, 32(36): 2204106

[17] Li Y, Dai X, Bu Y. . Photoelectrochemical performance improving mechanism: Hybridization appearing at the energy band of BiVO₄ photoanode by doped quantum layers modification. Small, 2022, 18(21): 2200454

[18] Mary A S, Murugan C, Pandikumar A. Uplifting the charge carrier separation and migration in Co-doped CuBi₂O₄/TiO₂ p-n heterojunction photocathode for enhanced photoelectrocatalytic water splitting. Journal of Colloid and Interface Science, 2022, 608: 2482–2492

[19] Shin D, Saparov B, Mitzi D B. Defect engineering in multinary earth-abundant chalcogenide photovoltaic materials. Advanced Energy Materials, 2017, 7(11): 1602366

[20] Malyi O I, Zunger A. False metals, real insulators, and degenerate gapped metals. Applied Physics Reviews, 2020, 7(4): 041310

[21] Li C, He J, Xiao Y. . Earth-abundant Cu-based metal oxide photocathodes for photoelectrochemical water splitting. Energy & Environmental Science, 2020, 13(10): 3269–3306

[22] Alizadeh A, Roudgar-Amoli M, Bonyad-Shekalgourabi S M. Dye sensitized solar cells go beyond using perovskite and spinel inorganic materials: A review. Renewable & Sustainable Energy Reviews, 2022, 157: 112047

[23] Wu T, Du Y, Dai L. . A direct Z-scheme AgBr/CuBi₂O₄ photocathode for ultrasensitive detection of ciprofloxacin and ofloxacin by controlling the release of luminol in self-powered microfluidic photoelectrochemical aptasensors. Analytical Chemistry, 2022, 94(30): 10651–10658

[24] Reddy D A, Kim Y, Varma P. . Inverse opal CuBi₂O₄ photocathodes for robust photoelectrochemical water splitting. ACS Applied Energy Materials, 2022, 5(5): 6050–6058

[25] Cao H, Zhang Z, Zhang M. . The effect of defects in tin-based perovskites and their photovoltaic devices. Materials Today Physics, 2021, 21: 100513

[26] Kang D, Hill J C, Park Y. . Photoelectrochemical properties and photostabilities of high surface area CuBi₂O₄ and Ag-doped CuBi₂O₄ photocathodes. Chemistry of Materials, 2016, 28(12): 4331–4340

[27] Oh W D, Lua S K, Dong Z. . Rational design of hierarchically-structured CuBi₂O₄ composites by deliberate manipulation of the nucleation and growth kinetics of CuBi₂O₄ for environmental applications. Nanoscale, 2016, 8(4): 2046–2054

[28] Shen X, Zhu Z, Zhang H. . Novel sphere-like copper bismuth oxide fabricated via ethylene glycol-introduced solvothermal method with improved adsorptive and photocatalytic performance in sulfamethazine removal. Environmental Science and Pollution Research International, 2022, 29(31): 47159–47173

[29] Xu Y, Jian J, Li F. . Porous CuBi₂O₄ photocathodes with rationally engineered morphology and composition towards high-efficiency photoelectrochemical performance. Journal of Materials Chemistry. A, Materials for Energy and Sustainability, 2019, 7(38): 21997–22004

[30] Wang Y, Hu J, Liu S. . Influence of grain size on photoelectrocatalytic performance of CuBi₂O₄ photocathodes. International Journal of Hydrogen Energy, 2022, 47(89): 37774–37782

[31] Yang Y, He A, Yang M. . Selective electroreduction of CO₂ to ethanol over a highly stable catalyst derived from polyaniline/CuBi₂O₄. Catalysis Science & Technology, 2021, 11(17): 5908–5916

[32] Kim J H, Adishev A, Kim J. . All-bismuth-based oxide tandem cell for solar overall water splitting. ACS Applied Energy Materials, 2018, 1(12): 6694–6699

[33] Jaiswal M K, Carrow J K, Gentry J L. . Vacancy-driven gelation using defect-rich nanoassemblies of 2D transition metal dichalcogenides and polymeric binder for biomedical applications. Advanced Materials, 2017, 29(36): 1702037

[34] Mohamed H S H, Rabia M, Zhou X G. . Phase-junction Ag/TiO₂ nanocomposite as photocathode for H₂ generation. Journal of Materials Science and Technology, 2021, 83: 179–187

[35] Seo G, Kim B, Hwang S W. . High-performance bulky crystalline copper bismuthate photocathode for enhanced solar water splitting. Nano Energy, 2021, 80: 105568

[36] Li M, Tian X, Zou X. . Promoting photoelectrochemical hydrogen evolution activity of CuBi₂O₄ photocathode through ramping rate control. International Journal of Hydrogen Energy, 2020, 45(30): 15121–15128

[37] Guo F, Shi W, Wang H. . Facile fabrication of a CoO/g-C₃N₄ p–n heterojunction with enhanced photocatalytic activity and stability for tetracycline degradation under visible light. Catalysis Science & Technology, 2017, 7(15): 3325–3331

[38] Li X, Yu J, Low J. . Engineering heterogeneous semiconductors for solar water splitting. Journal of Materials Chemistry. A, Materials for Energy and Sustainability, 2015, 3(6): 2485–2534

[39] Wan Z, Mao Q, Chen Q. Proton-dependent photocatalytic dehalogenation activities caused by oxygen vacancies of In₂O₃. Chemical Engineering Journal, 2021, 403: 126389

[40] Lindberg A E, Wang W, Zhang S. . Can a PbCrO₄ photoanode perform as well as isoelectronic BiVO₄?. ACS Applied Energy Materials, 2020, 3(9): 8658–8666

[41] Freysoldt C, Grabowski B, Hickel T. . First-principles calculations for point defects in solids. Reviews of Modern Physics, 2014, 86(1): 253–305

[42] Kedar L, Bond C E, Muirhead D. Carbon ordering in an aseismic shear zone: Implications for Raman geothermometry and strain tracking. Earth and Planetary Science Letters, 2020, 549: 116536

[43] Varanasi C, Petry J, Brunke L. . Growth of high-quality carbon nanotubes on free-standing diamond substrates. Carbon, 2010, 48(9): 2442–2446

[44] Pang Y X, Li X, Zhang X. . The synthesis of carbon-based quantum dots: A supercritical fluid approach and perspective. Materials Today Physics, 2022, 27: 100752

[45] Bourque A J, Rutledge G C. Heterogeneous nucleation of an n-alkane on graphene-like materials. European Polymer Journal, 2018, 104: 64–71

[46] Khaleghi A, Sadrameli S M, Manteghian M. Thermodynamic and kinetics investigation of homogeneous and heterogeneous nucleation. Reviews in Inorganic Chemistry, 2020, 40(4): 167–192

[47] Wang F, Chemseddine A, Abdi F F. . Spray pyrolysis of CuBi₂O₄ photocathodes: improved solution chemistry for highly homogeneous thin films. Journal of Materials Chemistry. A, Materials for Energy and Sustainability, 2017, 5(25): 12838–12847

[48] Song A, Plate P, Chemseddine A. . Cu:NiO as a hole-selective back contact to improve the photoelectrochemical performance of CuBi₂O₄ thin film photocathodes. Journal of Materials Chemistry. A, Materials for Energy and Sustainability, 2019, 7(15): 9183–9194

[49] Kang D, Hill J C, Park Y. . Photoelectrochemical properties and photostabilities of high surface area CuBi₂O₄ and Ag-doped CuBi₂O₄ photocathodes. Chemistry of Materials, 2016, 28(12): 4331–4340

[50] Hahn N T, Holmberg V C, Korgel B A. . Electrochemical synthesis and characterization of p-CuBi₂O₄ thin film photocathodes. Journal of Physical Chemistry C, 2012, 116(10): 6459–6466

[51] Berglund S P, Abdi F F, Bogdanoff P. . Comprehensive evaluation of CuBi₂O₄ as a photocathode material for photoelectrochemical water splitting. Chemistry of Materials, 2016, 28(12): 4231–4242

[52] Li J, Griep M, Choi Y. . Photoelectrochemical overall water splitting with textured CuBi₂O₄ as a photocathode. Chemical Communications, 2018, 54(27): 3331–3334

[53] Choi Y H, Yang K D, Kim D H. . p-Type CuBi₂O₄ thin films prepared by flux-mediated one-pot solution process with improved structural and photoelectrochemical characteristics. Materials Letters, 2017, 188: 192–196

RIGHTS & PERMISSIONS

Higher Education Press 2023