Electrodeposition of Copper Oxides as Cost-Effective Heterojunction Photoelectrode Materials for Solar Water Splitting

Abstract

Photoelectrocatalytic hydrogen production is crucial to reducing greenhouse gas emissions for carbon neutrality and meeting energy demands. Pivotal advances in photoelectrochemical (PEC) water splitting have been achieved by increasing solar light absorption. P-type Cu-based metal oxide materials have a wide range of energy band gaps and outstanding band edges for PEC water splitting. In this study, we first prepared Cu₂O thin films using electrodeposition and fabricated a heterojunction structure of CuO/Cu₂O by controlling annealing temperatures. The surface morphological, optical, and electrochemical properties were characterized using various analytical tools. X-ray and Raman spectroscopic approaches were used to verify the heterojunction of CuO/Cu₂O, while surface analyses revealed surface roughness changes in thin films as the annealing temperatures increased. Electrochemical impedance spectroscopic measurements in conjunction with the Mott–Schottky analysis confirm that the CuO/Cu₂O heterojunction thin film can boost photocurrent generation (1.03 mA/cm² at 0 V vs. RHE) via enhanced light absorption, a higher carrier density, and a higher flat band potential than CuO and Cu₂O thin films (0.92 and 0.08 mA/cm², respectively).

1. Introduction

Solar energy is one of the most viable renewable energy resources to meet global energy demand and minimize environmental pollution and global warming [1]. Effective solar energy harvesting has been a central direction in improving photoconversion efficiencies [2,3,4]. Among solar energy-driven approaches for producing green hydrogen, photoelectrochemical (PEC) water splitting has attracted significant attention [5]. Briefly, it generates electron-hole pairs via a semiconductor photoelectrode by absorbing sunlight and separating the charges that go to the photoelectrodes. Then, the reduction and oxidation processes of water on the electrodes by charge carriers occur, producing hydrogen and oxygen [6]. To date, many photoelectrode materials for PEC water splitting have been designed, including TiO₂ [7,8], Fe₂O₃ [9,10], BiVO₄ [11,12], Si [13,14], Cu₂O [15,16], and CuO [17,18]. Photoelectrodes for PEC water splitting must have a suitable energy band gap [5]. For example, the energy of the conduction band (CB) should be lower than that of reduced water, while that of the valence band (VB) should be higher than that of oxidized water. In particular, p-type semiconducting copper oxides (i.e., Cu₂O and CuO) have a low energy band gap (2.0–2.5 eV [19,20,21] and 1.3–2.1 eV [20,22,23], respectively). It is also known that they are a potential electrode material for photovoltaic applications [24] because of their cost-effectiveness, non-toxicity, excellent carrier mobility [25], and high absorption coefficient. Similarly, these advantages make copper oxides suitable as an efficient photoelectrode for PEC water splitting, as shown in Table S1 [26,27,28,29,30,31,32,33,34,35,36,37] and Table S2 [38,39,40]. One crucial obstacle to copper oxides may be their high electron-hole recombination rate, lowering photoconversion efficiencies. One of the effective approaches to address the issue is to employ a heterostructure configuration with a non-copper-containing semiconductor material, for example, Fe₂O₃/CuO [41,42], CuO/ZnO [43], TiO₂/CuO [44], WO₃/Cu₂O [45], InGaN/Cu₂O [46], and Cu₂O/g-C₃N₄/Cu₂O [47]. Although the heterojunction structures reduce the electron-hole recombination rate, they also increase the complexity of devices. Accordingly, heterostructured CuO/Cu₂O, which has been proposed as the heterojunction configuration, can be prepared using simple thermal oxidation from Cu₂O (Cu₂O + 1/2O₂ → 2CuO) [28]. In this work, we reported a facile method to prepare a CuO/Cu₂O heterojunction structure for PEC water splitting. Cu₂O thin films were prepared by electrodeposition, and thermal treatment was performed [28,40,48]. Simple and eco-friendly electrodeposition by controlling annealing temperatures resulted in micron thick copper oxide layers (~1.4 μm thick) compared to physical methods (i.e., sputtering and pulsed laser deposition). Controlling thin-film thicknesses is one of the essential design factors to improve electron-hole separation [49,50]. The heterojunction configuration demonstrated improved PEC performance compared to CuO and Cu₂O, as confirmed by the analysis of X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), ultraviolet-visible (UV-Vis) spectroscopy, electrochemical impedance spectroscopy (EIS), and Raman spectroscopy. In particular, copper oxide thin films were characterized to understand the change in carrier densities and flat-band potentials.

2. Experimental Details

2.1. Fabrication of Copper Oxide Thin Films

Cu₂O thin films were prepared by electrodeposition on fluorine-doped tin oxide glass substrates (FTO, ~400 nm and 1 cm × 2 cm). The substrates were first cleaned with an ultrasonicator with acetone, isopropyl alcohol (IPA), ethanol, and deionized (DI) water for 5 min and dried in an oven. In this study, all chemicals and reagents were used without further purification. Electrodeposition was performed using a three-electrode system with a potentiostat (SP-150e, BioLogic, Seyssinet-Pariset, France). A cleaned FTO substrate, a coiled platinum wire (0.5 mm diameter), and Ag/AgCl (3 M NaCl) were used as working, counter, and reference electrodes, respectively. Copper sulfate (Sigma-Aldrich, Burlington, MA, USA, >98%) and 3M lactic acid (Sigma-Aldrich, Burlington, MA, USA, >85%) mixed in 40 mL of DI water were used after adjusting pH = 9 with 5 M NaOH. The chronoamperometric (CA) method was applied at −0.36 V (V vs. Ag/AgCl) for 20 min to deposit Cu₂O thin films on an FTO glass substrate at ~50 °C and with magnetic stirring at 600 rpm. We then carried out thermal oxidation at different temperatures for 1 h on a hot plate using as-prepared Cu₂O thin films after washing with DI water and drying with nitrogen. After conducting the annealing temperature effects, we found that heterostructured CuO/Cu₂O/FTO and CuO/FTO thin films can be prepared at 350 and 500 °C, respectively (Figure 1). For simplicity, Cu₂O, CuO/Cu₂O, and CuO thin films correspond to Cu₂O/FTO prepared by electrodeposition without annealing, a heterostructured configuration of CuO/Cu₂O/FTO annealed at 350 °C and CuO/FTO annealed at 500 °C.

2.2. Materials Characterization

The surface morphologies of three thin films (Cu₂O, CuO/Cu₂O, and CuO deposited on an FTO-coated glass substrate) were observed by scanning electron microscopy (SEM, Zeiss Gemini 450, Jena, Germany) operated at an acceleration voltage of 10 keV. Ultraviolet-visible spectroscopy (UV-VIS, SEMSO-3000, AUBOTECH, Taipei, Taiwan) was used to examine the optical properties of the thin films. X-ray diffraction (XRD, Brucker D8 DISCOVER, Billerica, MA, USA) with Cu K_α radiation (λ = 0.154184 nm) and Raman microscopy (CL Technology UniDRON-S, New Taipei, Taiwan) with a 532 nm excitation wavelength were applied to investigate the transformation of Cu₂O to CuO. An X-ray photoelectron spectrometer (XPS, Thermo Scientific K-Alpha, Waltham, MA, USA) was used with monochromatic Al K_α radiation and a flood gun to examine the chemical valences of the thin films. All electrochemical measurements were performed using a standard rectangular quartz cuvette of ~20 mL with the three-electrode system. A leakless miniature Ag/AgCl reference electrode (ET072-1, eDAQ, 3.4 M KCl) was used after calibration using the Ag/AgCl (3 M NaCl) reference electrode to convert the potentials to the reversible hydrogen electrode (RHE) scale [51]. While 0.2 M Na₂SO₄ (pH = 6.2) was used as an electrolyte, an active area of thin films of 1.0 cm² was set using a chemical-tolerant 3 M tape (Electroplating Tape 470, 3 M). For the Mott–Schottky analysis, the frequency was set at 1k Hz and an amplitude of 10 mV. For PEC measurements, potentials were applied from 0 to 0.5 V (vs. RHE) at the scan rate of 5 mV/s. A solar simulator with an AM 1.5 G filter (Enlitech SS-X, Enlitech, Kaohsiung, Taiwan) and a light intensity of 100 mW/cm² was used as a light source after accurate calibration. The EIS was also adopted to ensure the rate of charge-transfer reactions between the electrode surface and the electrolyte. The Nyquist plots were obtained by applying a static potential of 0.3 V vs. Ag/AgCl with the 5 mV sinusoidal wave from 0.1 to 1 M Hz without and with light illumination.

3. Results and Discussion

3.1. Morphology and Optical Properties of Copper Oxide Thin Films

The surface morphologies of the thin films were examined by SEM and AFM, as shown in Figure 2. After 1-h thermal treatment at 350 and 500 °C, it was observed that the grain sizes and shapes changed, and the surfaces became denser than pristine Cu₂O thin films. The cross-sectional view of the Cu₂O thin film indicates a thickness of ~1.40 μm. Due to the difference in density (Cu₂O: 6.0 g/cm³ and CuO: 6.31 g/cm³) [52] and the thermal expansion of Cu₂O [53] after thermal oxidation (Cu₂O + 1/2O₂ → 2CuO), a slight increase in thickness was observed, similar to previous reports [28]. As summarized in Table S3, energy-dispersive X-ray spectroscopy (EDS) was also employed to examine the chemical composition of the thin films before and after annealing at 350 and 500 °C. It verifies the successful fabrication of Cu₂O using electrodeposition and the complete transformation of Cu₂O to CuO after annealing at 500 °C for 1 h. Additionally, the EDS analysis shows a heterostructure of CuO/Cu₂O ([Cu]/[O] = 1.52). The SEM analysis illustrates that the annealing of metal oxides [54] led to the agglomeration of adjacent grains [55].

A similar trend was also observed using AFM (Figure 2d–f), which is attributed to the random distribution of grains and phase changes [56]. It was found that the root mean square (RMS) roughness of the thin films increased from 21.6 nm (Cu₂O) to 33.9 nm (CuO/Cu₂O at 350 °C) and 46.5 nm (CuO at 500 °C) (Cu₂O < CuO/Cu₂O < CuO). We assume that the controlling of thin film thicknesses using the electrolyte temperature affects RMS roughness [40]. Furthermore, since the light absorption properties of thin films play an essential role in PEC performance, we measured the absorbance of Cu₂O, CuO/Cu₂O, and CuO thin films using a UV-Vis spectrometer. Figure 3a shows the typical absorbance of the three thin films, demonstrating that the Cu₂O thin film has an absorption edge in the range of 500 to 650 nm. It shows that the absorption edge of thermally treated thin films (CuO/Cu₂O and CuO) was red-shifted. In addition, we measured the band gaps of these samples using the Tauc plot analysis [57] using ${\left(\alpha h\nu \right)}^n=B\left(h\nu -E_g\right)$, where α is the absorption coefficient, h is the Plank constant, v is the photon energy, B is a constant, and E_g is the band gap. The copper oxide thin films were analyzed with a direct transition (n = 2). As Figure 3b and Table S4 show, the band gaps measured (Cu₂O: 2.35 eV, CuO/Cu₂O: 1.81 eV, CuO: 1.45 eV) are in good agreement with a previous study [28]. CuO/Cu₂O and CuO exhibited a higher band gap than Cu₂O, denoting that they can expand the absorption range of visible light compared to Cu₂O. Consequently, it could improve PEC performance [58].

3.2. Characterization of Copper Oxide Thin Films Using XRD, Raman Spectroscopy, and XPS

Figure 4a demonstrates the XRD pattern for the CuO/Cu₂O thin film, while Figure S1 shows those for Cu₂O and CuO thin films. For the Cu₂O thin film, the peaks at 36.6° and 42.5° are attributed to the (111) and (200) planes [59,60] (JCPDS file No. 99-0041). The XRD pattern of Cu₂O illustrates that the thin films exhibit the (200) plane as the preferential growth direction [61]. Cu₂O and CuO diffraction peaks were observed in the thin film annealed at 350 °C, while the thin film annealed at 500 °C exhibited the pure CuO phase (JCPDS file No. 72-0269) [62]. The distinctive diffraction peak at 35.6° indicates that the thin film has a good crystal structure of CuO after annealing at 500 °C. The XRD analysis may support that the thermal oxidation process at 350 °C produced heterostructured CuO/Cu₂O similar to the previous studies [28,63]. The particle sizes of the thin films were evaluated using the Scherrer–Debye equation [64], $D=\frac{K\lambda }{\beta \cos \theta }$, where D is the average crystal particle size, K is a constant, λ is the wavelength of X-ray, β is the full width at half maximum (FWHM), and θ is the Bragg angle. The Cu₂O and CuO thin films were determined as 22.3 and 29.1 nm, respectively, indicating the improvement of the homogeneity of the thin films and the enhancement of the electrical properties [65]. As shown in Figure 4b, we also performed Raman spectroscopic measurements in the spectral region of 100 to 800 cm⁻¹ to complement the XRD and SEM analyses. Single-phase Cu₂O has only one Raman active mode (F_2g), while single-phase CuO has three Raman active modes (A_g + 2B_g) [66]. The Raman bands of the Cu₂O thin film at 218, 278, and 623 cm⁻¹ are contributed from the second-order Raman-allowed modes of single-phase Cu₂O [67], while those of the CuO thin film at 287, 338, and 614 cm⁻¹ are in good agreement with single-phase CuO, as reported in the literature [68]. As the annealing temperatures increased, Cu₂O peaks decreased, and CuO peaks appeared gradually. The Raman spectrum of the CuO/Cu₂O thin film shows that the combined peaks of Cu₂O and CuO (209, 279, 329, and 614 cm⁻¹) are assigned to the A_g and B_g modes [66]. As illustrated in Figure 4c,d, XPS was also used to analyze the elemental compositions of copper oxides. Figure 4c shows a representative XPS survey spectrum of the CuO/Cu₂O thin film [69]. All XPS spectra were calibrated with the C 1s photoemission peak of adventitious carbon at 284.5 eV [70]. Figure S2a,b shows the XPS spectra of Cu 2p of the Cu₂O and CuO thin films that align with those reported in the literature [71,72]. To examine the heterostructure of the CuO/Cu₂O thin film, we carried out the deconvolution of its XPS spectrum using XPSPEAK 4.1. Two main peaks and satellite peaks were observed in the spectrum. The deconvolution process identified that the peaks at 932.91 and 952.74 eV are related to Cu⁺ (Cu₂O), which are assigned to Cu 2p_3/2 and Cu 2p_1/2. By contrast, those at 934.27 and 954.12 eV are attributed to Cu 2p_3/2 and Cu 2p_1/2, confirming the presence of Cu²⁺ (CuO). As summarized in Figure S2c, we observed a slightly broadened peak for Cu 2p_3/2 of CuO/Cu₂O compared to CuO. Based on the deconvolution of the CuO/Cu₂O thin film, we manifested that the sample is a heterojunction (CuO:Cu₂O = 56.4%:43.6%), which qualitatively agrees with the analysis using SEM/EDS (Table S3). As reported [73,74], metal oxide thin films fabricated via simple annealing could generate cracks inside thin films due to thermal stress, enabling X-ray beams to irradiate the FTO surface. Thus, the peak at 496.08 eV can be assigned to Sn 3d originating from FTO [75]. We assumed that the CuO/Cu₂O thin film prepared by electrodeposition could become a heterojunction configuration as reported [28] according to the characterization of the SEM/EDS, XRD, XPS, and Raman spectroscopic analyses.

3.3. Electrochemical Measurements of Copper Oxide Thin Films

The Mott–Schottky technique was applied to examine changes in electrochemical interfaces between the electrolyte and photocathodes prepared in this study. Carrier densities and flat band potentials can be accurately measured using the Mott–Schottky equation for p-type semiconductors [76,77].

$$\frac{1}{C^2}=-\frac{2}{\epsilon {\epsilon }_{0}e{N}_A{A}^2}\left(V-V_{fb}-\frac{k_{B}T}{e}\right)$$

where C is the interfacial capacitance, $\epsilon$ is the dielectric constant of the semiconductor electrode, ${\epsilon }_0$ is the vacuum permittivity, e is the electron charge, V_fb is the flat band potential, N_A is the major carrier density, A is the electrode area, V is the applied voltage, k_B is the Boltzmann constant, and T is the absolute temperature. Carrier densities of p-type copper oxides (N_A) can be determined from a straight line by plotting 1/C² vs. V, while the flat band potential (V_fb) can be obtained from the intercept with the x-axis. While carrier densities affect charge transfer rates [28,78,79], the flat band potential is associated with a driving force to separate electron-hole pairs in the space charge region, which can be used as one of the key factors to predict PEC performance. For example, a shift of the flat band to a more positive potential denotes improved electron transfer, decreasing the recombination of electron-hole pairs and boosting PEC efficiencies [80]. Figure 5a–c demonstrates Mott–Schottky plots for Cu₂O, CuO/Cu₂O, and CuO thin films. The three thin films exhibit p-type semiconductor characteristics with a negative slope. ε of 7.6 for Cu₂O [81] and 10.26 [36] for CuO/Cu₂O and CuO thin films were used to calculate carrier densities. It resulted in a N_A value of 1.79 × 10¹⁶, 6.91 × 10¹⁷, and 6.26 × 10¹⁷ cm^–3, respectively, in the order of Cu₂O < CuO < CuO/Cu₂O. Moreover, we also successfully confirmed the same trend of flat band potentials of 0.46, 1.04, and 0.89 V, respectively. According to the Mott–Schottky analysis, the improved PEC performance of the heterojunction structure of CuO/Cu₂O can be rationally anticipated [28]. Figure 5d,e shows Nyquist plots of three thin films without and with illumination. It exhibited that the radius of the semicircle decreased with light illumination to the working electrode. It denotes that illumination accelerated the rate of the charge transfer reaction between the photoelectrode surface and the electrolyte, and accordingly, the carrier density increased. As shown in Figure 5e, EIS data were then analyzed by fitting with an equivalent electrical circuit model to examine transfer resistance on different photoelectrode materials. As reported, R_s is the series resistance, R_sc is the resistance on the depletion layer, and R_ct is the charge transfer resistance, while C_sc is the capacitance related to the depletion layer of a semiconductor and C_H is the Helmholtz layer capacitance [82]. It is anticipated that the heterostructured electrode exhibits the lowest charge transfer resistance (R_ct) due to its smallest diameter, leading to the highest charge transfer reaction rate [28]. Figure 5f illustrates representative photocurrent densities of three electrode materials under AM1.5G illumination. For the Cu₂O thin film, a photocurrent of 0.08 mA/cm² at 0 V (vs. RHE) was observed, while for the CuO/Cu₂O and CuO thin films, the photocurrent increased to 1.03 and 0.92 mA/cm², respectively. It aligns with Mott–Schottky and EIS-based analyses.

Table 1. Experimental values of carrier density (N_A), flat band potential (V_fb), and photocurrent density for p-type Cu₂O, CuO/Cu₂O, and CuO thin films.

Thin Film	NA (cm−3)	Vfb (V)	Photocurrent Density (mA/cm2)
Cu2O	1.79 × 1016	0.46	0.08
CuO/Cu2O	6.91 × 1017	1.04	1.03
CuO	6.26 × 1017	0.89	0.92

compiles carrier densities, flat band potentials, and photocurrent densities of Cu₂O, CuO/Cu₂O, and CuO thin films. It is noted that PEC performance can be affected by too high annealing temperatures, altering the electrical property of FTO-based thin films [83]. In summary, it manifests that a higher carrier density, a higher flat band potential, and lower charge transfer resistance of the heterojunction CuO/Cu₂O thin film could be highly correlated with the PEC performance compared to the Cu₂O and CuO thin films [28].

4. Conclusions

Cu₂O thin films were effectively fabricated using electrodeposition, followed by controlling annealing temperatures. The thickness and morphology of thin films were analyzed using SEM, while the surface roughness was investigated using AFM. XRD, XPS, and Raman spectroscopic measurements confirmed the heterojunction structure of CuO/Cu₂O after annealing at 350 °C. As the thermal oxidation temperatures increased, the light absorbance edge of the thin films was red-shifted, demonstrating a better absorber in the visible region and enhancing photocurrent densities in the order of Cu₂O < CuO < CuO/Cu₂O. In particular, the Mott–Schottky and EIS measurements confirmed that the heterojunction structure has a higher carrier concentration, a larger flat band potential, and a lower charge transfer resistance than the Cu₂O and CuO thin films. Accordingly, the CuO/Cu₂O thin film annealed at 350 °C achieved a boosted photocurrent density of 1.03 mA/cm² at 0 V (vs. RHE). We demonstrated that the heterojunction configuration of copper oxide thin films (CuO/Cu₂O) via the cost-effective electrodeposition method could improve PEC performance by reducing the electro-hole pair recombination and increasing carrier densities.