Solution-Processed Synthesis of Copper Oxide (CuxO) Thin Films for Efficient Photocatalytic Solar Water Splitting

This article reports a solution-processed synthesis of copper oxide (CuxO) to be used as a potential photocathode for solar hydrogen production in the solar water-splitting system. CuxO thin films were synthesized through the reduction of copper iodide (CuI) thin films by sodium hydroxide (NaOH), which were deposited by the spin coating method from CuI solution in a polar aprotic solvent (acetonitrile). The phase and crystalline quality of the synthesized CuxO thin films prepared at various annealing temperatures were investigated using various techniques. The X-ray diffraction and energy dispersive X-ray spectroscopy studies confirm the presence of Cu2O, CuO/Cu2O mixed phase, and pure CuO phase at annealing temperatures of 250, 300, and 350 °C, respectively. It is revealed from the experimental findings that the synthesized CuxO thin films with an annealing temperature of 350 °C possess the highest crystallinity, smooth surface morphology, and higher carrier density. The highest photocurrent density of −19.12 mA/cm2 at −1 V versus RHE was achieved in the photoelectrochemical solar hydrogen production system with the use of the CuxO photocathode annealed at a temperature of 350 °C. Therefore, it can be concluded that CuxO synthesized by the spin coating method through the acetonitrile solvent route can be used as an efficient photocathode in the solar water-splitting system.


INTRODUCTION
Energy consumption is increasing rapidly because of the increase in population and industrialization in the world. Production of hydrogen (H 2 ) as chemical energy using the abundant solar power can be an excellent resolution to this growing energy crisis. 1 A simple technique to produce H 2 is water splitting in a photoelectrochemical (PEC) cell. 1,2 In the PEC system, the photoelectrode is irradiated by sunlight, light is absorbed in the photoelectrode, and electron−hole pairs are generated, which are driven into the solution by the electric field at the photoelectrode/water junction to produce H 2 and oxygen (O 2 ) in the PEC water-splitting system. 3 Efficient photoelectrode materials are selected based on some fundamental properties such as a high absorption coefficient, excellent charge transport capability, photocorrosion stability in the electrolyte solution, and low kinetic overpotentials. 4,5 However, there is no single material that meets all the criteria for an efficient photoelectrode. Therefore, finding suitable materials for the absorption of a full portion of incident photons, developing active catalysts, and realizing the appropriate interface design between the catalyst, photoabsorber, and electrolyte to minimize losses are the main challenges for an efficient PEC system. 5 A number of photocatalysts including titanium oxide (TiO 2 ), 6−10 zinc oxide (ZnO), 11−13 tungsten oxide (WO 3 ), 14−16 reduced graphene oxide-polypyrrole (rGO@ PPy), 17 iron oxide (Fe 2 O 3 ), 2,18 bismuth vanadium tetra-oxide (BiVO 4 ), 19 cuprous oxide (Cu 2 O), 1,20−23 and the more active catalyst cupric oxide (CuO) 24−28 have been investigated for perceiving a high-efficiency PEC water-splitting system. Among these catalysts, copper oxides (CuO and Cu 2 O) have recently attracted special attention because they are abundant in the crust of the earth, low-cost and nontoxic. 29 Moreover, copper oxide photocathodes can be prepared using electrodeposition, 21,23 spin coating, 30−32 RF-magnetron sputtering, 28 and spray pyrolysis techniques. 33 Depending on the techniques and criteria of preparation, the direct band gap of the p-type Cu 2 O and CuO has been observed to be nearly 2.0−2.5 and 1.3−1.7 eV, respectively. 34 The band alignment of the Cu 2 O and CuO is favorable for the generation of the H 2 fuel in the PEC system. 33,35,36 The theoretical solar-to-hydrogen conversion efficiency of both Cu 2 O and CuO is >18%. 33,35 However, the practical Cu x O photoelectrodes suffer from the poor-stability problem in the aqueous solution because its redox potential lies within the band gap. 1,37−40 Surface passivation of the photocathodes through the use of conformal coating could be one possible way to enhance the stability of the Cu x O photoelectrodes. 1,21,23,37,38 Low efficiency of the Cu x O photocathodes is another critical issue. The unfavorable minority carrier diffusion length and slow carrier transfer rate at the photoelectrode surface are the main issues for the practical copper oxide (Cu 2 O, CuO) photocathodes that are responsible for the low efficiency of the devices. 29,33 However, the performance of the copper oxide photocathode can be enhanced by improving the crystal quality and surface morphology of the material. To address this problem, researchers have applied different techniques such as doping 41 and high-temperature treatment 42 to copper oxide for the fabrication of an efficient PEC system. So far, the electrodeposited Cu 2 S-coated Cu 2 O NW photoelectrode has been reported to produce a photocurrent of −1.0 mA/cm 2 at −0.6 V versus Ag/AgCl. 21 A photocurrent of 0.49 mA/cm 2 at −0.25 V versus Ag/AgCl has been reported for the electrodeposited Cu 2 O electrode with NiFe-layered double hydroxides. 23 Son et al. also reported a photocurrent density of −5.2 mA/cm 2 at 0 V versus RHE for the electrodeposited Cu 2 O photocathode using CuO/NiO as a back protected layer. 37 Lim et al. reported a photocurrent of −0.28 and −0.35 mA/cm 2 at 0.05 V versus RHE for the sol−gel deposited Cu 2 O and CuO photoelectrode, respectively. 24 The RF-magnetron-sputtered CuO photocathode yielded a photocurrent of −3.1 mA/cm 2 at 0 V (RHE). 28 A solution-processed CuO nanoparticle films provided a photocurrent of −1.2 mA/cm 2 at −0.7 V versus Ag/AgCl. 30 Baran 31 A photocurrent of −24 mA/cm 2 at 0.25 V versus RHE has also been reported for the spray-pyrolyzed tenorite CuO thin films. 33 In addition, the solution-processed synthesis of Cu x O thin films by the spin coating method could produce potential candidates as efficient photoelectrodes for application in solar hydrogen production as it reduces cost and fabrication complexity. Recently, a simple one-step synthesis of p-type Cu x O thin films via in situ reaction of a spin-coated CuI thin film in aqueous NaOH solution has been reported for the fabrication of efficient thin-film transistors. 32 The Cu x O thin films synthesized by the same route have been used as a hole transport layer for the perovskite solar cells. 43 Moreover, the deposition of CuI thin films by the spin coating method from a CuI solution dissolved in an acetonitrile solvent has already been reported. 44 However, there are few reports on the application of CuO photocathodes in PEC solar water splitting synthesized from the reduction of spin-coated CuI thin films.
In this article, we demonstrate the facile one-step synthesis of Cu x O thin films through the reduction of CuI thin films prepared from a CuI solution dissolved in acetonitrile by the spin coating method for the efficient application in the solar water-splitting system. The synthesized Cu x O thin films as a photocathode exhibited excellent performance in the PEC solar water-splitting system with better reproducibility. It is revealed from the findings that the acetonitrile-assisted synthesized Cu x O thin film could be a deserving candidate as a photocathode in the PEC system for solar H 2 production by water splitting.  The crystallite size (D) of the synthesized Cu x O thin films was calculated using the Debye−Scherrer formula and is shown in Table 1. It is found that the crystallite size increases with annealing temperature and the Cu x O thin film annealed at 350°C showed the highest crystallinity.
Thermogravimetric analysis (TGA) and differential thermal analysis (DTA) are the popular methods to investigate the thermal stability of the semiconducting materials. 46−49 Therefore, TGA and DTA of the Cu x O thin film (preannealed at 120°C ) were carried out to determine the proper annealing temperature. The TGA−DTA spectra were taken by scanning the temperature from 30 to 500°C at 20°C/min under a continuous flow of nitrogen gas at 20 mL/min. Figure 1b delineates the TGA−DTA spectra of the copper oxide sample. About 22% weight loss appeared from temperature 273−400°C with no weight loss found except for this range. Such a huge weight loss reveals the transformation of the Cu 2 O phase to the CuO phase and the evaporation of other volatile elements. Further weight loss was not observed above 400°C, indicating the thermal stability of Cu x O from 400 to 500°C. Therefore, the appropriate annealing temperature is >400°C under a nitrogen atmosphere. Simultaneous thermal decomposition and improved crystallinity of the prepared sample appear from the DTA spectra. It also predicts that an optimum crystallinity of the sample could be obtained at a temperature of >400°C.   at different temperatures. The surface of the Cu x O thin film annealed at 250°C appears to be covered almost uniformly with grains of ∼0.9 μm. Several void spaces were also observed in the film. The surface morphology of the film significantly changed for the films annealing at 300°C. The grains of the 300°C-annealed copper oxide were found to be randomly distributed. The increase in void spaces, grain boundaries, and the surface roughness of the films might be due to the phase transition of the copper oxide from Cu 2 O to CuO and evaporation of the volatile elements. 50 The nonuniformity of grains and a large number of grain boundaries in the photocathode increase the probability of charge scattering and carrier recombination rate, which leads to the performance degradation of the photocathode in the PEC system. The nucleation rate was found maximum at 350°C and ∼0.34 μm grains were observed with a few amounts of grain boundaries. As shown in Figure 2c, the surface of the Cu x O film annealed at 350°C is very compact and smoother with reduced void spaces than compared to that of the films annealed at other temperatures. Improved surface morphology of the photocathode is favorable for the PEC performance, 51 which resists the recombination of the charge carriers and benefits in the enhancement of the charge transport property. 52,53 Figure 2d−f depict the energy dispersive X-ray spectroscopy (EDX) spectra used to evaluate the chemical composition of the Cu x O thin films annealed at a temperature from 250 to 350°C

Surface Morphological Study of Cu x O Thin Films.
. The atomic ratio of "Cu" and "O" in the synthesized Cu x O thin films is listed in Table 2, which indicates that Cu x O annealed at 250°C was entirely in the Cu 2 O phase whereas the film annealed at 300°C was CuO with a small amount of the Cu 2 O phase and the film annealed at 350°C was purely CuO phase. The EDX results are consistent with the XRD analysis.  Figure 3. The absorption spectra of the Cu x O thin films are depicted in Figure 3a. The absorbance increases with increasing annealing temperature. The Cu x O thin film annealed at 350°C showed better absorbance than other films that may arise because of large grain size, less void spaces, a smooth surface, and compactness of the film. The band gap of Cu x O thin films was calculated using the Tauc plot, 44 as shown in Figure 3b. It is reported that both Cu 2 O and CuO are p-type semiconductors having direct and indirect optical band gap, respectively. 54−57 It was revealed from the XRD and EDX studies that the films annealed at 250, 300, and 350°C were the pure Cu 2 O, Cu 2 O/CuO mixed, and pure CuO phase, respectively. Therefore, the direct band gap where C is the differential capacitance, ε is the dielectric constant of the semiconductor, ε 0 is the permittivity of free space, N D is the density of dopants, V is the applied potential, V fb is the flat band potential, k is the Boltzmann constant, and T is the absolute temperature. The negative slope of the MS plot confirms the p-type conductivity of the Cu x O thin films.  59 The schematic diagram of the two-electrode PEC water-splitting system using the synthesized Cu x O films as the working electrode (WE) and Pt as the counter electrode (CE) is also shown in the inset of Figure 4b. When the light is incident on the Cu x O photocathode, the electron−hole pairs will be generated. The protons will be reduced at the Cu x O/ electrolyte interface by the CB electrons. Besides, oxidation of oxygen will occur at the CE because of the transfer of photogenerated holes from the Cu x O photocathode to the CE.
2.5. Photocatalytic Performance of Cu x O Thin Films. The current−voltage characteristics of the Cu x O photocathodes in the three-electrode PEC water-splitting system are depicted in Figure 5a. The plot shows the photocurrent density of the best performing Cu x O photocathodes among a number of samples with better reproducibility. First, the experimental potential was measured with respect to the Ag/ AgCl reference electrode (RE). Then, the potential versus Ag/ AgCl was converted into the potential versus RHE (reversible hydrogen electrode) scale with the help of the following formula where, E RHE is the potential converted into the RHE scale, E Ag/AgCl is the experimentally measured potential with respect to the Ag/AgCl RE, pH is the pH of the electrolyte, and E Ag/AgCl 0 = 0.1976 V at 25°C. At an applied potential of −0.5 V versus RHE (onset potential), H 2 production was initiated from the photocathode annealed at 250°C (production of bubbles), and the production of bubbles increased with increasing applied bias potential; the onset potential of −0.47 V versus RHE and −0.4 V versus RHE was observed for the photocathodes annealed at 300 and 350°C, respectively. The onset potential followed the variation of the flat band potential in a reversible way; a low onset potential is desirable for an efficient photocathode. The synthesized photocathode annealed at 250, 300, and 350°C produced a photocurrent density of −12.50, −2.13, and −19.12 mA/cm 2 , respectively, at −1 V versus RHE. As it was confirmed from the XRD and EDX results that Cu x O thin films   Figure 5a that the photocathode annealed at 300°C shows low photocurrent density than the other photocathodes. The presence of mixed copper oxide phases (CuO/Cu 2 O) in the film annealed at 300°C might be the reason for the low photocurrent density because the mixed (nonstoichiometric) copper oxide phases induce more defects (trapping centers) 60 and reduced the carrier mobility. The CuO photocathode produced a higher photocurrent than the Cu 2 O photocathode. The higher crystallite size, uniform surface morphology, and higher carrier density of the CuO photocathode helped to produce higher photocurrent than the Cu 2 O photocathode, 33,61 which is consistent with the XRD, SEM, and MS plot analysis discussed in the previous sections. The dark current was found significantly very low compared with the light current, which indicates that the corrosion rate of our synthesized Cu x O photoelectrode in aqueous solution is low.
The PEC performance of the synthesized Cu x O photoelectrodes was investigated by calculating the applied bias photon-to-current efficiency (ABPE) in a two-electrode system using the following formula 62 where V b is the applied potential between the WE and CE in a two-electrode configuration and its corresponding photocurrent density is J ph and P total is the total light (100 mW/ cm 2 ) incident on the photocathode. As illustrated in Figure 5b, the optimum ABPE for the Cu 2 O, CuO/Cu 2 O, and CuO photocathodes is 4.69% (at 0.9 V vs Pt), 0.79% (at 0.9 V vs Pt), and 7.06% (at 0.85 V vs Pt), respectively. The observed PEC performance of the synthesized photocathodes is much h i g h e r t h a n t h e p r e v i o u s l y r e p o r t e d pe r f o r mance. 1,19−22,24,27,28,30,31,33,63−65 Table 3 shows a comparison of Cu x O-based photocathode performance with that of our synthesized Cu x O photocathode. The objective of the present work was to synthesize an efficient Cu x O photocathode for the PEC system. Because the stability of the Cu x O photocathodes is an important issue, the stability of the photocurrent of the synthesized photocathodes was observed under 100 mW/cm 2 sun light conditions at −1 V versus Pt using the two-electrode system and is shown in Figure 6. Till now, the photostability of the Cu x O photocathodes is far from that of the typical photoelectrodes (WO 3 and α-Fe 2 O 3 ); the fast photocorrosion deactivated the Cu x O photocathodes. 63,64,71 It has been demonstrated that the copper oxides spontaneously decompose to metal Cu under illumination; Cu accumulated on the surface and deactivated the inner copper oxide after a certain period. 63,71 It is observed from the Figure 6 that the photocathodes which were annealed at 250 (Cu 2 O), 300 (CuO/Cu 2 O), and 350°C (CuO) retained about ∼85% of their initial photocurrent for ∼30, ∼34, and ∼39 min, respectively. After this period, the    Figure 7 shows the preparation steps of the copper oxide photocathode. First, the CuI thin films were prepared on commercialized indiumdoped tin oxide (ITO)-coated glass substrates from a copper iodide (CuI) solution of 30 mg/mL in acetonitrile following a recipe reported in previous work. 44 The ITO-coated glass substrates were cleaned using an ultrasonic vibrator where acetone, isopropanol, and distilled water were used in sequence for cleaning before the deposition of the films. CuI thin films were deposited on the ITO glass substrates by spin coating at a speed of 2000 rpm for 30 s following a preannealing process at 150°C for 10 min. The process was repeated five times to deposit thicker films. Then, the Cu x O films were produced through dipping of the CuI thin films in aqueous sodium hydroxide (NaOH) solution (0.25 mol L −1 ). The Cu x O thin films were then rinsed in methanol and dried in air to eliminate the undesired surplus. Because thermal treatment is an essential factor in the one-step solution process for the phase transformation of copper oxide, the synthesized photocathodes were finally annealed at temperatures of 250, 300, and 350°C for 1 h. The thicknesses of the Cu x O thin films were in the range of 650−700 nm.
4.3. Characterization. Structural properties of the copper oxide thin films were investigated using XRD with Cu Kα radiation having wavelength 1.540598 Å using the EMMA Xray diffractometer (GBC Corporation, Australia). The surface morphology of the films was observed using a scanning electron microscope from EVO 18, Carl Zeiss, UK. The elemental composition of the films was measured by EDX using a Team EDS system (EDAX, AMETEK, USA; beam voltage: 15 kV). The Bruker Dektak XTL thickness profiler was used to measure the thickness of the films. The optical properties of the films were investigated using UV−visible spectrophotometry using a T60 UV−visible spectrophotometer (PG Instruments) with 1 nm increment in wavelength. Capacitance−voltage (C−V) measurement was done using an impedance analyzer (model: Agilent 4192 A). All the films were characterized at room temperature (∼27°C) and a relative humidity of ∼50%.
4.4. PEC Measurement. The PEC performance of the synthesized Cu x O photocathodes was explored under 1.5 AM (100 mW/cm 2 ) simulated sun illumination using a Solar Light 16S-Series 150−300 W UVB solar simulator. The measurement was carried out in a well-recognized three-electrode PEC system, where platinum (Pt) as the CE, saturated Ag/AgCl as the RE and Cu x O photocathode as the WE having an active area of 1.5 cm 2 were immersed in a 0.025 M H 2 SO 4 aqueous electrolyte solution. For all samples, current versus potential measurements were recorded at a scanning rate of 10 mV s −1 using a Squidstat plus potentiostat from Admiral Instruments, USA. The Cu x O photocathode was installed perpendicularly to the direction of light illumination and illuminated from the front side because such alignment is convenient to utilize the performance of the photoelectrode properly. 2