Key Strategies on Cu₂O Photocathodes toward Practical Photoelectrochemical Water Splitting

Abstract

Cuprous oxide (Cu₂O) has been intensively in the limelight as a promising photocathode material for photoelectrochemical (PEC) water splitting. The state-of-the-art Cu₂O photocathode consists of a back contact layer for transporting the holes, an overlayer for accelerating charge separation, a protection layer for prohibiting the photocorrosion, and a hydrogen evolution reaction (HER) catalyst for reducing the overpotential of HER, as well as a Cu₂O layer for absorbing sunlight. In this review, the fundamentals and recent research progress on these components of efficient and durable Cu₂O photocathodes are analyzed in detail. Furthermore, key strategies on the development of Cu₂O photocathodes for the practical PEC water-splitting system are suggested. It provides the specific guidelines on the future research direction for the practical application of a PEC water-splitting system based on Cu₂O photocathodes.

1. Introduction

1.1. Basic Principles of Photoelectrochemical Water Splitting

Photoelectrochemical (PEC) water splitting has been regarded as an ideal technology for generating CO₂-free hydrogen because it does not emit CO₂ as a by-product during PEC operation, contrary to traditional hydrogen-production technologies based on fossil fuel. Figure 1 illustrates an operational principle of the PEC water-splitting system with two semiconductor electrodes. The overall PEC water-splitting reaction includes three steps [1,2,3,4]: (1) Electron-hole pairs are generated when sunlight is absorbed by the semiconductor in the water. (2) After separating the generated electron-hole pairs by the internal electric field in the depletion layer or the external bias, electrons in the conduction band (CB) of the semiconductor and holes in the valence band (VB) of the semiconductor drift to the semiconductor/water interface. (3) Water splits into hydrogen and oxygen by electrons and holes, respectively, at the semiconductor/water interfaces via the following reactions.

2H⁺ + 2e⁻ → H₂ (Hydrogen evolution reaction, HER)

(1)

2H₂O + 4h⁺ → 4H⁺ + O₂ (Oxygen evolution reaction, OER)

(2)

2H₂O → 2H₂ + O₂ (Overall water-splitting reaction)

(3)

Semiconductors are a key component in the PEC water-splitting system because it absorbs sunlight, and moreover water-splitting reactions occurs at its interface. There are two criteria of a semiconductor for overall PEC water splitting according to the operational principle of PEC water splitting, as illustrated in Figure 1 [1,5,6,7,8]. First, its CB energy level (E_CB) should be lower than the HER potential (0 versus reversible hydrogen electrode, RHE), while its VB energy level (E_VB) should be higher than the OER potential (1.23 V versus RHE).

E_CB < 0 V versus RHE, E_VB > 1.23 V versus RHE

(4)

Second, its band gap (E_g) should theoretically be larger than 1.23 eV because the difference between the HER potential and OER potential is 1.23 eV. Practically, an Eg of 1.5~1.8 eV is necessary due to the HER/OER overpotentials (η_HER and η_OER) [9,10].

E_g > 1.23 eV + η_HER + η_OER

(5)

Among various semiconductors, few wide-band-gap semiconductors, such as TiO₂ [11], ZnO [12], SrTiO₃ [13], and Nb₂O₅ [14], meet these criteria, as shown in Figure 2a. Photoelectrodes with these materials can complete the entire PEC water-splitting reactions using only sunlight without external bias. However, they can absorb only ultraviolet light below 400 nm due to their band gaps above 3.0 eV, thereby utilizing the limited solar energy. Hence, PEC photoelectrodes based on these wide-band-gap semiconductors show a low solar-to-hydrogen (STH) efficiency below 3% (Figure 2b) [15]. Therefore, semiconductors with a short band gap are necessary to utilize the sufficient solar energy for efficient PEC water splitting.

Intrinsic characteristics of short-band-gap semiconductors determine the charge transport direction. Figure 3 shows the energy levels of the semiconductor and electrolyte before and after contacting each other, depending on the intrinsic characteristics of the semiconductor. As illustrated in Figure 3, the downward band bending is formed when n-type semiconductors contact the electrolyte, while the upward band bending is formed when p-type semiconductors contact the electrolyte, due to the Fermi level equilibration between the semiconductor and electrolyte. Due to this band bending, holes in the n-type semiconductor easily move to the semiconductor/electrolyte interface, whereas electrons in the p-type semiconductor quickly move to the semiconductor/electrolyte interface. Therefore, in general, n-type semiconductors such as BiVO₄ [19,20,21], WO₃ [22,23,24], Fe₂O₃ [25,26,27], and Ag₃PO₄ [28] have been used as photoanodes for generating oxygen, while p-type semiconductors such as CuO [29,30,31], Cu₂O [17,32,33,34], Sb₂Se₃ [35,36,37], LaFeO₃ [18,38,39], ternary copper oxide [40], and transitional metal dichalcogenides [41] have been used as photocathodes for generating hydrogen.

1.2. Cuprous Oxide Photocathode

Cuprous oxide (Cu₂O) is a remarkable photocathode material for the efficient and economical PEC water-splitting system. Its intrinsic characteristics and energy level are suitable for generating hydrogen via water reduction reactions at the interface. As shown in Figure 2, its E_CB is much lower than the HER potential, which is advantageous for transporting electrons to the semiconductor/water interface. In addition, it is possible to utilize the visible light due to its band gap (2 eV). Theoretically, it enables Cu₂O photocathodes to produce a high photocurrent density up to −14.7 mA cm⁻², corresponding to a solar-to-hydrogen (STH) efficiency of 18% [15]. Furthermore, it is cheap and earth-abundant. However, it is still challenging to achieve an efficient and durable PEC performance using only a Cu₂O photoelectrode. Hence, the state-of-the-art Cu₂O photocathode consists of a back contact layer, overlayer for the heterostructure, protection layer, and HER catalysts, as well as a Cu₂O light absorber, as illustrated in Figure 4 [32,42,43,44].

In this review paper, the fundamentals of each component in the Cu₂O photocathode mentioned above are briefly introduced. In addition, recent research progress on these components for the efficient and durable Cu₂O photocathode is summarized. Further, recent efforts in practical PEC water splitting based on Cu₂O photocathodes are reviewed. Finally, the future outlook and research directions on the Cu₂O photocathodes toward practical PEC water splitting are discussed. It will give comprehensive guidelines and insights on the efficient and durable Cu₂O photocathode to researchers focusing on the practical PEC water-splitting system.

2. Fundamentals and Research Progress

2.1. Cu₂O Light Absorber

Cu₂O is a key component of Cu₂O photocathodes because it not only generates electron-hole pairs by absorbing sunlight but also transports the generated charges. Therefore, its optical and charge transport characteristics are a significant parameter affecting the PEC performance of Cu₂O photocathodes. Based on the absorption coefficient, thick film above 1 μm is necessary to absorb sufficient light using the Cu₂O light absorber [45,46,47]. However, it generally has a limited minority carriers (electrons) diffusion length below 200 nm [45,48,49]. In this regard, the generated electrons in the thick Cu₂O film are easily recombined before reaching the water interface, irrespective of enough light utilization. On the other hand, the light utilization is not sufficient in the thin Cu₂O film, even though it is advantageous for efficient electron transport. Therefore, it is necessary to overcome this mismatch for achieving the high PEC performance of Cu₂O photocathodes.

The introduction of nanostructure is an effective strategy to enhance the light utilization by the light-trapping effect. Cu₂O nanowire structures have been widely adopted in the Cu₂O photocathode because of the simple fabrication process. In general, the metallic Cu substrate is transformed into the copper hydroxide (Cu(OH)₂) nanowire electrode by the chemical reactions. Subsequently, it is converted into the Cu₂O nanowire electrode via the thermal treatment. Hsu et al. fabricated Cu₂O micro/nanostructured photocathodes by chemical oxidation and subsequent thermal reduction under a N₂ atmosphere, as shown in Figure 5a [50]. They found that the temperature of thermal reduction is crucial to obtain pure and well-structured Cu₂O nanowire photocathodes. However, their devices had flower-like Cu₂O structures with micro scales (2~3 μm) on Cu₂O nanowires, thereby showing the limited PEC performance, corresponding to an STH efficiency of 1.97%. Salehmin et al. also demonstrated that the coverage of Cu₂O microflowers on the Cu₂O nanowires results in the reduced PEC performance because it interrupts the light penetration into the nanowire structure and it prolongs the electron transport length [51]. They fabricated vertically aligned Cu₂O nanowire photocathodes by controlling the ageing time of nanowire growth. It showed a better PEC performance, resulting from the improved charge transport and prolonged light penetration path, compared to one in the work by Hsu et al. [50] (Figure 5b).

Luo et al. developed high-quality Cu₂O nanowire photocathodes with excellent electronic and photonic properties by an anodization process in a strong alkaline solution and an annealing process under an Ar atmosphere [52]. The pure Cu₂O nanowire photocathode was successfully completed using the Cu sputtered fluorine-doped tin oxide (FTO) substrate with a thickness of Cu above 1.5 μm. Such a thick Cu layer continuously provides the Cu source for the transformation of Cu(OH)₂ into pure Cu₂O during the annealing process. In addition, they electrodeposited the Cu₂O thin layer on the Cu₂O nanowire photocathode as a blocking layer for preventing the charge recombination. As shown in Figure 5c, the improved incident-photon-to-current efficiency (IPCE) in the longwave length region (500~600 nm) was observed in their Cu₂O nanowire photocathodes, due to the light-trapping effect by the nanowire structure. As a result, it showed a remarkably improved PEC performance compared to the planar Cu₂O photocathodes. A current density was achieved up to −8 mA cm⁻² at the HER potential, corresponding to an STH efficiency of 10%, combined with an aluminum-doped zinc oxide (AZO) overlayer and ruthenium oxide (RuO_x) HER catalysts. Interestingly, the diameter of the Cu₂O nanowire in this literature is within the minority carrier diffusion length of Cu₂O. Therefore, the nanowire structure is beneficial for better charge transport, as well as improved light utilization.

Recently, various template-assisted Cu₂O nanostructure photocathodes have been also introduced. Zhao et al. developed pyramidal Cu₂O photocathodes assisted with the pyramidal silicon (Si) template [53]. It was fabricated by electrodepositing Cu₂O film on the gold (Au) layered pyramidal Si template. The pyramidal structure exhibited a significantly reduced reflectance in the longwave length region above 500 nm (Figure 5d). This means that it is effective to enhance the light utilization by the light-trapping effect. Cu₂O micro pillar photocathodes was successfully fabricated by Yoo’s group [54]. They used the nickel (Ni) micro pillars with heights of 16 μm and diameters of 1~1.5 μm as a template for synthesizing Cu₂O micro pillar photocathodes. Contrary to other Cu₂O nanowire photocathodes, it was aligned well perpendicularly to the substrate, as shown in Figure 5e. It provides more active sites for water reduction reactions, which is favorable to improve the PEC performance. Inverse-opal is a typical 3D nanostructure for improving the light utilization in photonic devices. Wu et al. recently developed Cu₂O inverse-opal photocathodes using polystyrene (PS) microspheres as a colloidal template [55]. The inverse-opal Cu₂O photocathode was successfully fabricated by removing the PS microsphere template after depositing a Cu₂O film by the electrochemical deposition method. The light absorption characteristics of the inverse-opal Cu₂O structure were dependent on the size of the PS microsphere template (Figure 5f). Based on this observation, they successfully optimized the inverse-opal structure for efficient Cu₂O photocathodes.

The poor charge transport characteristics of Cu₂O should also be improved for developing the efficient Cu₂O photocathode. It is well-known that cation doping is generally effective to enhance the p-type conductivity of Cu₂O electrodes, depending on the fabrication method [56,57,58]. Hence, many groups have tried to apply various cations as a dopant to the Cu₂O photocathode. The Ni-doped Cu₂O photocathode is representative of these efforts [59,60]. As illustrated in Figure 6a, in the Ni-doped Cu₂O photocathode, the band gap became narrow due to an acceptor impurity from Ni dopants compared to the pristine Cu₂O photocathode [60]. It triggered not only accelerated charge separation but also extended light absorption. Furthermore, Ni dopants did not induce the structural distortion of the Cu₂O photocathode because the ionic radius of Ni²⁺ (0.72 Å) is similar to that of Cu⁺ (0.77 Å). It promotes charge transport, thereby improving PEC performance. Alkaline ions (Li⁺, Na⁺, and K⁺) were also used as a cation dopant by Chen et al. [61]. Although these were all effective to improve charge transport, Li dopants were the most effective to improve the PEC performance (Figure 6b) because lithium(Li)-doped Cu₂O photocathodes had less defects compared to Na-doped and K-doped Cu₂O photocathodes. It is mainly due to the different ionic radii of alkaline ions: that of Li⁺ (0.76 Å) is the closest to that of Cu⁺ compared to others (Na⁺: 1.02 Å, and K⁺: 1.38 Å). Silver(Ag)-doped and iron(Fe)-doped Cu₂O photocathodes were also introduced by Upadhyay et al. [62,63]. They showed the improved PEC performances, compared to those of undoped Cu₂O photocathodes (Figure 6c,d). It was mainly due to the enhanced charge transport and conductivity by the doping effect. However, the excessive cation doping had negative influences on the PEC performance because it caused a non-homogeneous Cu₂O photocathode with many defect sites. Therefore, it is important to find the optimal doping level for the improved PEC performance of Cu₂O photocathodes.

The grain boundary of Cu₂O photocathodes also affects the charge-transport characteristics. It is advantageous for the efficient Cu₂O photocathode to alleviate the grain boundaries in the Cu₂O because they act as a recombination center of charge, reducing the PEC performance. It is controllable to modify the fabrication process. Baek et al. introduced a thin antimony (Sb)-incorporated Cu₂O (Cu₂O:Sb) seed layer to grow the highly-oriented Cu₂O photocathode using electrodeposition [64]. The Cu₂O:Sb seed layer was synthesized by electrodeposition in the Sb-ions-added copper sulfate aqueous solution. The Sb ions retard lateral diffusion of Cu ions in the solution, resulting in the vertically-oriented Cu₂O seed layer. It facilitates the growth of highly-oriented Cu₂O photocathode with less grain boundaries during the electrodeposition process (Figure 7a). Qin et al. recently compared the characteristics of electrodeposited Cu₂O photocathodes and magnetron-sputtered Cu₂O photocathodes [65]. They demonstrated that the magnetron sputtering is more feasible to fabricate Cu₂O photocathodes with less grain boundaries than electrodeposition (Figure 7b). These works reported the improved PEC performance of a Cu₂O photocathode by the mitigation of grain boundaries in the Cu₂O photocathode, as shown in Figure 7.

2.2. Back Contact Layer

The back contact layer is responsible for forwarding generated holes in the Cu₂O layer to the external load via the conductive substrate. In general, the thin metal film is employed as a back contact layer; thus, the Cu₂O/metal interface is critical to promote the migration of holes. The characteristic of the Cu₂O/metal interface is entirely dependent on the work function of Cu₂O (φ_Cu2O) and metal (φ_metal). When the φ_metal is smaller than φ_Cu2O, Schottky contact is formed by the Fermi level equilibration, thereby preventing the hole transportation into the conductive substrate (Figure 8a). On the other hand, the φ_metal is larger than φ_Cu2O, Ohmic contact is formed by the Fermi level equilibration, facilitating the accelerated migration of holes to the conductive substrate (Figure 8b). Hence, the metal should have a larger work function than that of Cu₂O (4.84 eV) [66] for fitting into the back contact layer for the Cu₂O photocathode.

Table 1. Available metals as a back contact layer of the Cu₂O photocathode and their work functions [66,67,68,69].

Metal	Au	Pt	Cu	Ni	Co	Pd	Ir
Work function (eV)	5.1~5.2	5.35~5.65	5.1	5.15	5.0	5.12~5.30	5.25~5.27

shows metals with a large work function. These are available as a back contact layer for Cu₂O photocathodes because their work functions are larger than that of Cu₂O, leading to the Ohmic contact for hole migrations.

A Au back contact layer has been frequently used for Cu₂O photocathodes [17,32,33,70]. The Cu₂O photocathode with a Au back contact layer shows a reliable and outstanding PEC performance, due to its excellent hole collection characteristic. Nevertheless, it is disadvantageous for the large-scale application and tandem configuration with a short-band-gap semiconductor, because it is expensive and opaque, respectively. To solve this limitation, Dias et al. developed transparent Cu₂O photocathodes with a thin Au back contact layer [17]. They found that a thin Au layer below 5 nm is sufficient for collecting holes, similar to a thick Au layer. It also enables the usage of Au to be reduced. The Cu back contact layer is a promising candidate for replacing the expensive Au back contact layer because it is earth-abundant and low-cost [52,71]. Cu₂O photocathodes with a Cu back contact layer generally showed similar PEC performances with Cu₂O photocathodes with a Au back contact layer.

Non-metallic materials have been explored as an alternative to the metallic back contact layer for Cu₂O photocathodes. The main goal for these efforts is to develop the affordable Cu₂O photocathode avoiding the usage of precious components. Nickel oxide (NiO) is a well-known material as a hole selective layer in perovskite solar cells [72,73,74]. It is also possible to utilize the back contact layer of Cu₂O photocathodes, because it has a larger work function (5.0 eV) than that of Cu₂O. As shown in Figure 9a, it forms the Ohmic contact with Cu₂O for smoothly migrating holes, while the fluorine-doped tin oxide (FTO) substrate forms the Schottky contact with Cu₂O for interrupting the hole migration due to a shallow band barrier [75]. Son et al. reported the copper nickel mixed oxide (CuO/NiO) hole selective layer for the Cu₂O photocathode [76]. They fabricated the CuO/NiO thin layer by a sequential metallic Cu/Ni sputtering and annealing process. It efficiently blocks the charge recombination at the interface between the Cu₂O layer and the conductive substrates due to its huge energy barrier, resulting in an improved PEC performance (Figure 9b). Furthermore, it is quite transparent, which is beneficial for using the Cu₂O photocathode as a top absorber in the tandem configuration. Copper thiocyanate (CuSCN) was applied as a back contact layer of a Cu₂O photocathode in the work by Pan et al. [77]. Although the hole transport from Cu₂O into CuSCN is difficult due to the VB offset in terms of energy level, solution-processed CuSCN thin film facilitates the smooth hole transport from Cu₂O into conductive substrates by the band-tail states existence, as illustrated in Figure 9c. Moreover, the huge barrier generated by the large CB band offset effectively prevents the charge recombination at the back contact interface. Zhou et al. suggested iron oxide hydroxide (FeOOH) as a hole transfer layer in the Cu₂O photocathode [78]. Electrodeposited FeOOH thin film promotes the extraction of holes from Cu₂O into conductive substrates due to its energy level (Figure 9d). Hence, the Cu₂O photocathodes with the FeOOH back contact layer showed not only an enhanced PEC performance but also an improved stability.

2.3. Overlayer

A heterostructured Cu₂O photocathode with a semiconductor overlayer is a promising strategy to improve the overall water-splitting performance of Cu₂O photocathodes because the semiconductor overlayer renders the accelerated electrons/holes’ separation and promotes electron transport from Cu₂O into the water interface. However, the mechanism is slightly different, depending on the intrinsic characteristics of the semiconductor overlayer (n-type or p-type). When the p-n heterostructure is formed with an n-type semiconductor overlayer in the Cu₂O photocathode, the built-in voltage (V_BI) created from the difference of E_F and E_F,redox further increases compared to the single Cu₂O photocathode (Figure 10a,b) [79]. It promotes the electrons/holes separation and transport. On the other hand, the staircase-type energy levels are created, when the p-p heterostructure is formed with p-type semiconductor overlayer, which has a relatively lower CB and VB than Cu₂O (Figure 10c) [80]. It accelerates the electron transport into the water interface. Therefore, the n-type semiconductor-overlayered Cu₂O photocathodes show a remarkable early onset potential, as well as an improved photocurrent density, whereas the p-type semiconductor-overlayered Cu₂O photocathodes primarily show an improved photocurrent density.

Table 2. Onset potential and photocurrent density at the HER potential of heterostructured Cu₂O photocathodes with n-type or p-type semiconductor overlayers. Some devices consist of a TiO₂ protection layer and HER catalysts, as well as overlayers.

Overlayer	Device	Onset Potential (V versus RHE)	Photocurrent Density (mA/cm2, 0 V versus RHE)	Ref.
TiO2	Cu2O/TiO2	0.0 (versus Ag/AgCl)	−0.8 (−1.0 V versus Ag/AgCl)	[81]
TiO2	Cu2O	0.37	−0.52	[82]
	Cu2O/TiO2	0.42	−1.40
	Cu2O/TiO2/NiFe	0.50	−2.60
	Cu2O/TiO2/rGO/NiFe	0.54	−3.71
ZnO	Cu2O/ZnO/TiO2/Pt	0.65	−4.00	[83]
AZO	Cu2O/ZnO	0.50	−1.60	[84]
	Cu2O/AZO	0.63	−2.90
Ga2O3	Cu2O	0.50	-	[86]
	Cu2O/Ga2O3/Pt	0.90	−4.00
	Cu2O/Ga2O3/TiO2/Pt	1.00	−6.50
Ga2O3	Cu2O/AZO/TiO2/RuOx	0.50	−8.00	[87]
	Cu2O/Ga2O3/TiO2/RuOx	1.00	−9.60
CuO	Cu2O	0.45	−0.21	[89]
	Cu2O/CuO	0.80	−2.47
CuO	Cu2O	0.0 (versus Ag/AgCl)	−0.06 (−0.3 V versus Ag/AgCl)	[90]
	Cu2O/CuO	0.1 (versus Ag/AgCl)	−0.26 (−0.3 V versus Ag/AgCl)
CuO	Cu2O/CuO	0.60	−2.70	[91]
CuO	Cu2O	0.50	−0.15	[92]
	Cu2O/CuO	0.50	−1.20

summarizes the onset potential and photocurrent density at the HER potential of heterostructured Cu₂O photocathodes with n-type or p-type semiconductor overlayers. In the case of p-n heterostructures, Cu₂O photocathodes with a TiO₂ overlayer show an improved onset potential (earlier onset potential) compared to the single Cu₂O photocathode. The Cu₂O/TiO₂ heterojunction reinforces the band bending, resulting in an improved PEC performance [81,82]. However, TiO₂ has been widely applied to the protection layer of Cu₂O photocathodes, rather than the overlayer for the heterojunction effect, because it is an intrinsically stable oxide in the aqueous solution. Although the ZnO overlayer is also effective to improve the charge transport by the formation of heterojunction with Cu₂O, the aluminum-doped zinc oxide (AZO) overlayer is more efficient for the heterojunction effect because it is more conductive compared to the ZnO overlayer [83,84]. Minami et al. reported that the gallium oxide (Ga₂O₃)/Cu₂O heterostructure improves the photovoltage of Cu₂O-based solar cells due to the decreased defect levels at the interfaces [85]. Inspired by this work, Li et al. introduced the Ga₂O₃ overlayer in the Cu₂O photocathode for improving the PEC performance [86]. The improved photovoltage by the Ga₂O₃/Cu₂O heterostructure leads to a remarkable enhanced onset potential in the Cu₂O photocathodes. Pan et al. further improved the PEC performance of Ga₂O₃ overlayered Cu₂O photocathodes with the Cu₂O nanowire and ruthenium oxide (RuO_x) HER catalysts [87]. Their devices also showed the improved onset potential, approximately 1.0 V versus RHE. It is advantageous for improving the PEC performance of an unbiased water-splitting system with the Cu₂O photocathode and photoanode/photovoltaic [88]. Consequently, they succeeded in demonstrating the unbiased all-oxide solar water-splitting system using a Ga₂O₃ heterostructured Cu₂O photocathode and transparent BiVO₄ photoanode with an operating current density of 2.5 mA cm⁻¹, corresponding to an STH efficiency of 3%.

On the contrary to the n-type semiconductor overlayers, most researchers have used a cupric oxide (CuO) overlayer to form the p-p heterostructure for the Cu₂O photocathode because it is easily prepared using a pure Cu₂O photocathode via a thermal oxidation process [89,90,91,92]. The electrons are easily moved to the water interface by the staircase-type energy level due to the energy levels of CuO and Cu₂O, as illustrated in Figure 10c. In addition, the light utilization is also enhanced in the CuO/Cu₂O heterostructured photocathode, due to a narrow band gap of CuO (1.3~1.7 eV) [93,94]. Hence, the improvement of photocurrent density is prominent, rather than the enhancement of onset potential, in the CuO-overlayered Cu₂O photocathode. The optimization of the CuO/Cu₂O heterostructured photocathode has continuously been explored by several groups. Du et al. derived the optimal annealing temperature (650 °C) for fabricating the highly efficient CuO/Cu₂O photocathode [90]. Jeong et al. found the optimal thickness of CuO for promoting the heterojunction effect [92]. The CuO overlayer with a thickness of approximately 90 nm is optimal for improving the performance of the Cu₂O photocathode by the CuO/Cu₂O heterostructure, because it is close to the minority carrier diffusion length of CuO.

2.4. Protection Layer

The main challenge of the Cu₂O photocathode is its poor stability against water. It is easily degraded in the aqueous solution within a few minutes, thereby losing its PEC characteristics [32]. Figure 11 illustrates the stability change in the semiconductor in the water. In general, the semiconductor is readily oxidized in the water when its oxidation potential is smaller than the OER potential, while it is simply reduced in the water when its reduction potential is larger than the HER potential, instead of aiding in water-splitting reactions [95]. Cu₂O meets these two conditions: It is reduced into the metallic Cu in the potential window of 0.3~0.4 V versus RHE, whereas it is oxidized into CuO or copper hydroxide (Cu(OH)₂) in the potential window of 0.6~1.05 V versus RHE (Figure 11) [96]. Hence, it is extremely unstable in water. This is a reason why the protection layer is an essential component in the Cu₂O photocathode for durable PEC water splitting.

Many researchers have used a TiO₂ thin film as a protection layer for the Cu₂O photocathode because it is intrinsically stable in water. Its reduction potential is larger than the HER potential; thus, it is robust to corrosion in the aqueous solution. In addition, its CB is beneficial for transporting electrons into the water interface [95]. The TiO₂ protection layer is generally deposited on the Cu₂O photocathodes by atomic layer deposition (ALD), which is favorable to deposit the homogeneous thin layer with the thickness of nanometer scale. The benchmark stability of Cu₂O photocathodes was recorded by Prof. Grätzel’s group using an ALD-deposited amorphous TiO₂ protection layer in 2018 [87]. Their devices showed a remarkable stability beyond 100 h, as shown in Figure 12a. However, it was gradually degraded after PEC operation for 100 h. This means that the amorphous TiO₂ protection layer is not sufficient for completely protecting the Cu₂O photocathode.

Many efforts have been made to improve its protection capability. Azevedo et al. carried out low-temperature steam treatment on the amorphous TiO₂-protected Cu₂O photocathode [97]. Although the crystallinity of TiO₂ was not changed, the surface of the TiO₂ became much smoother after the steam treatment (Figure 12b). It cured any defects and cracks in the TiO₂ protection layer, resulting in an excellent durability with 10% of PEC performance loss over more than 50 h. The crystallization of the amorphous TiO₂ protection layer is also one of these efforts because the crystalline TiO₂ is more robust than the amorphous one [96,98]. Nishikawa et al. succeeded in applying the crystalline TiO₂ protection layer on the Cu₂O photocathodes using the solution process assisted with excimer laser irradiation [99]. As shown in Figure 12c, the mild laser irradiation with continuous shots induced the crystallization of TiO₂, especially the rutile phase. It was demonstrated that it efficiently prevents the redox reaction of Cu₂O, facilitating the stability enhancement of Cu₂O photocathodes. Furthermore, the structural modification of the TiO₂ protective layer enables the long-term stability of the Cu₂O photocathode to improve. A thick TiO₂ protection layer above 100 nm is beneficial to improve the stability, but it is prone to show a decreased PEC performance due to the disturbance of electron transport. Kim et al. solved this mismatch by introducing a metallic nano filament to reinforce the electron transport in the thick TiO₂ protection layer [100]. Consequently, the Cu₂O photocathode showed an excellent stability with a considerable PEC performance for 100 h (Figure 12d).

A conductive layer also assists to enhance the long-term stability of Cu₂O photocathodes via the fast electron transfer from the Cu₂O into the water interface. In this regard, some groups used the metallic layer, such as Au and Ag, for protecting the degradation of Cu₂O photocathodes [101,102]. Its thickness should be considered because it can block the light irradiation to the Cu₂O because of its opacity, leading to the decreased PEC performance. A thin metallic layer with tens of nanometer is advantageous to improve the stability without decreasing the PEC performance. Carbon-based material has been explored as a conductive protection layer for Cu₂O photocathodes. Kunturu et al. developed CuO heterostructured Cu₂O photocathodes with a 15 nm carbon layer [103]. The thin carbon layer suppresses the photocorrosion of Cu₂O photocathodes by facilitating fast electron transfer to the surface. Das et al. tried to use a graphene layer, which is more conductive than pure carbon, as a protection layer [104]. They deposited the graphene protective layer on the Cu₂O photocathode by the chemical vapor deposition (CVD) method. The CVD-fabricated graphene layer had several microcracks, resulting in the degradation of Cu₂O. They solved this shortcoming by introducing a thin TiO₂ layer below 10 nm. It patches the microcrack in the graphene layer, thereby efficiently inhibiting the photocorrosion of Cu₂O. Titanium nitride (TiN) was also used as a protection layer in the work by Diao et al., because it is highly conductive and corrosion-resistant [105]. They controlled the thickness of TiN by adjusting the cycle of ALD deposition to trade off the PEC performance and the stability. Finally, the Cu₂O photocathodes with the ultra-thin TiN protection layer (8 nm) showed 100% stability during the PEC operation for 1 hr, without critical deterioration of the PEC performance.

Figure 12. Efforts for improving the stability of Cu₂O photocathodes using a TiO₂ protection layer and its modification. (a) Long-term stability of Cu₂O nanowire photocathodes with a Ga₂O₃ overlayer and RuOx HER catalysts. Reprinted from [87] with permission from Macmillan Publisher Ltd., Springer Nature, copyright 2018. (b) TiO₂ surfaces before/after steam treatment and the durability of TiO₂-protected Cu₂O photocathodes assisted with low-temperature steam treatment. Reprinted from [97] with permission from the Royal Society of Chemistry, copyright 2014. (c) Crystallization of the TiO₂ protection layer induced by laser irradiation and its effect on the stability of Cu₂O photocathodes. Reprinted from [99] with permission from Elsevier B.V., copyright 2015. (d) Stability of Cu₂O photocathodes with the modified TiO₂ protection layer using a metallic nano filament design. Reprinted from [100] with permission from Wiley-VCH GmbH, copyright 2021.

Figure 12. Efforts for improving the stability of Cu₂O photocathodes using a TiO₂ protection layer and its modification. (a) Long-term stability of Cu₂O nanowire photocathodes with a Ga₂O₃ overlayer and RuOx HER catalysts. Reprinted from [87] with permission from Macmillan Publisher Ltd., Springer Nature, copyright 2018. (b) TiO₂ surfaces before/after steam treatment and the durability of TiO₂-protected Cu₂O photocathodes assisted with low-temperature steam treatment. Reprinted from [97] with permission from the Royal Society of Chemistry, copyright 2014. (c) Crystallization of the TiO₂ protection layer induced by laser irradiation and its effect on the stability of Cu₂O photocathodes. Reprinted from [99] with permission from Elsevier B.V., copyright 2015. (d) Stability of Cu₂O photocathodes with the modified TiO₂ protection layer using a metallic nano filament design. Reprinted from [100] with permission from Wiley-VCH GmbH, copyright 2021.

Recently, organic materials have also received a lot of attention as a protective layer for Cu₂O photocathodes. Li et al. introduced the compact polymer layer for protecting Cu₂O photocathodes [106]. Polyurethane acrylate (PUA) thin film was used to cover the Cu₂O by a solution process using a viscous urethane acrylate monomer. Their device showed a 98% photo-durability during PEC operation under continuous light illumination for 6 h. Zhang et al. developed phenylethynyl copper (Ph-C≡C-Cu)-protected Cu₂O photocathodes [107]. The Ph-C≡C-Cu layer was a self-assembled monolayer fabricated by a photothermal method. Hence, the conformal Ph-C≡C-Cu protection layer was successfully applied to the Cu₂O photocathode. As a result, they constructed a stable unbiased water-splitting system operating for 5 h using the Ph-C≡C-Cu-protected Cu₂O photocathode with the NiOOH/FeOOH-layered BiVO₄ photoanode. These protection layers based on organic materials have hydrophobic characteristics, which can inhibit the corrosion of Cu₂O by avoiding contact with the aqueous solution. Hence, it is thoroughly effective to prolong the stability of Cu₂O photocathodes.

2.5. Co-Catalysts

In general, HER catalysts have been employed as a co-catalyst on the surface of the protection layer of Cu₂O photocathodes to reduce the overpotential of the water reduction reaction, leading to the improved PEC performance of Cu₂O photocathodes. It is well-known that platinum (Pt) is the most promising HER catalyst because of its highly intrinsic activity in the water reduction reaction and low hydrogen adsorption energy [108,109]. Thus, it has been actively adopted for improving the PEC performance of Cu₂O photocathodes [32,43,110,111,112,113]. Nevertheless, it is unfavorable for long-term stability because it detaches from the Cu₂O photocathodes during PEC operation. It is mainly attributed to the weak bonding of the Pt and TiO₂ protection layer [96]. RuO_x, which was suggested by Tilley et al. [114], is an alternative HER catalyst to Pt for the durable PEC operation of Cu₂O photocathodes. It is more robust than Pt due to the strong bonding with the TiO₂ protection layer. Hence, the RuO_x-photoelectrodeposited Cu₂O photocathode showed a better stability with a competitive PEC performance, compared to the Pt-catalyzed Cu₂O photocathode. However, these components (Pt and RuO_x) are not suitable for the economical PEC water-splitting system because they are a noble metal. Therefore, many researchers have pursued the application of inexpensive HER catalysts with low-cost materials to the Cu₂O photocathodes for obtaining the economic feasibility of the PEC water-splitting system.

Ni-based materials are a promising candidate as a low-cost HER catalyst. Lin et al. deposited the NiO/nickel hydroxide (Ni(OH)₂) composite on the Cu₂O nanowire photocathode by the solution process and sequential annealing process [115]. They found that the NiO/Ni(OH)₂ composites allow for an improved charge transfer by retarding charge recombination via the synergy effect of NiO and Ni(OH)₂. Hence, the NiO/Ni(OH)₂ composite-decorated Cu₂O nanowire photocathode showed an enhanced PEC performance. Recently, Jian et al. developed a highly efficient NiO_x composite (mixture of Ni and NiO) HER catalyst fabricated by vacuum evaporation deposition and heat treatment [116]. As shown in Figure 13a, it showed a remarkable HER activity (Tafel slope of 35.9 mV dec⁻¹), which was very close to the HER activity of Pt (Tafel slope of 32.5 mV dec⁻¹). Thus, the NiO_x-catalyzed Cu₂O photocathode showed a positive shift of onset potential from 0.2 V to 0.6 V versus RHE. Furthermore, it showed quite a stable PEC performance for 90 min. This means that the NiO_x composite plays a role as not only an HER catalyst but also as a protection layer for Cu₂O photocathodes. The electrodeposited Ni-Mo (Molybdenum) alloy, which was suggested in the work by Morales-Guio et al. [117], also significantly reduced the HER overpotential in an alkaline aqueous solution. The Ni-Mo-decorated Cu₂O photocathode produced a high photocurrent density of −6.3 mA cm⁻² at 0 V versus RHE without any dark currents in the basic solution (Figure 13b). Qi et al. successfully applied Ni-Fe (Iron)-layered double hydroxide (LDH) to the Cu₂O photocathode [118]. The electrodeposited Ni-Fe LDH catalysts boosted the electron transfer into the water interface by the appropriate band alignment with Cu₂O. Consequently, the Ni-Fe LDH co-catalyzed Cu₂O photocathode showed a seven-fold increase in PEC performance with remarkable stability for 40 h compared to a bare Cu₂O photocathode (Figure 13c).

Transition metal sulfides have been considered as a good HER catalyst for Cu₂O photocathodes, due to their high conductivities. Morales-Guio et al. deposited molybdenum sulfide (MoS_x) co-catalysts on the TiO₂-protected Cu₂O photocathode by photoelectrodeposition [119]. The MoS_x-loaded Cu₂O photocathode showed an outstanding PEC performance with a photocurrent density of −5.7 mA cm⁻² at the HER potential, corresponding to an STH efficiency of 7%, in the strong acidic electrolyte (pH 1), as shown in Figure 14a. Although the condition of the electrolyte was extremely harsh, their devices survived with a reliable PEC performance for 10 h. Chen et al. developed nickel sulfide (NiS)-combining aluminum (Al) nanoparticles HER catalysts for the Cu₂O photocathode [120]. NiS HER catalysts were deposited by a successive ionic layer adsorption reaction (SILAR) method. They demonstrated that the NiS attracts more protons (H⁺) into the surface of Cu₂O photocathodes and forms enriched electron conditions by the Ni-H bonds, resulting in a reduced HER overpotential. As a result, the Al nanoparticles NiS-co-catalyzed Cu₂O photocathode showed an improved PEC performance, along with the plasmonic effect by the Al nanoparticles (Figure 14b).

Transition metal phosphides have also been explored as low-cost HER catalysts for Cu₂O photocathodes because they have an excellent catalytic activity for HER. Cobalt phosphide (CoP) catalysts were successfully adopted for the Cu₂O photocathode in the work by Stern et al. [121]. Nanoflower-like CoP catalysts showed an extremely low HER overpotential of approximately 97 mV in a strong acidic condition to reach a current density of 10 mA cm⁻². Consequently, their CoP-decorated Cu₂O photocathodes showed a considerable photocurrent density of −5.3 mA cm⁻² at the HER potential in the strong acidic electrolyte, which is a favorable condition for the water reduction reaction (Figure 14c). Chhetri et al. developed nickel phosphide (NiP)-decorated Cu₂O photocathodes [122]. They successfully deposited NiP catalysts on the CuO heterostructured Cu₂O photocathode by the pulse plating electrodeposition. Their devices showed a noticeably improved PEC performance by the fast electron transfer into the water interface due to the NiP catalysts compared to the bare Cu₂O photocathodes (Figure 14d).

3. Outlook and Future Research Directions

Four aspects should be considered for the practical PEC water-splitting system: efficiency, stability, cost, and mass production. In terms of efficiency and cost, Cu₂O is a frontrunner photocathode material for the practical PEC water-splitting system, because the state-of-the-art Cu₂O photocathode shows a high STH efficiency above 10%, which is a benchmark for the commercialization. In addition, it is an earth-abundant material. However, it is necessary to further improve its stability above 100 h for the durable PEC water-splitting operation. Furthermore, upscaling is essential for the mass production of hydrogen via the PEC water-splitting system based on Cu₂O photocathodes. Therefore, future research on the Cu₂O photocathode should move forward based on strategies as below.

• Cu₂O absorber: The further improvement of electron transport capability will be necessary. To this end, doping is the most efficient strategy. In the case of this strategy, the ionic radius of the dopant should be similar to that of Cu⁺ ions for reducing the defects on the Cu₂O film. In addition, the doping level should be optimized for improving the electron transport in the Cu₂O photocathodes, because the excessive doping has a negative influence on the PEC performance. The fabrication of a high-quality Cu₂O film with less defects or grain boundaries is also advantageous to improve the electron transport in the Cu₂O photocathode. Furthermore, the development of transparent Cu₂O photocathodes with efficient PEC performance paves the way for developing the efficient PEC-PEC or PEC-PV water-splitting system with short-band-gap materials;
• Back contact layer: The development of an alternative back contact layer to the expensive Au back contact layer is a main goal of this component. In the case of metal, its work function should be higher than that of Cu₂O. Moreover, control of the opacity is necessary for the development of transparent Cu₂O photocathodes. A semiconductor with a huge energy barrier is a good option because it efficiently hinders the electron recombination at the interface. In this case, the suitable deposition method of Cu₂O should be considered on the semiconductor-based back contact layer;
• Overlayer: In the case of n-type overlayers, the created photovoltage in contact with Cu₂O should be considered because it motivates the charge separation in the p-n junction with Cu₂O. In the case of p-type overlayers, the exploration on the alternative material to the CuO overlayer with the proper energy level for enhancing the electron transfer into the water interface is a good strategy for the future research direction;
• Protection layer: Although the amorphous TiO₂ protection layer is highly efficient for protecting a Cu₂O photocathode against the corrosion, it is still not sufficient due to its pinholes or defects. Hence, the reduction in pinholes or defect of the amorphous TiO₂ protection layer is useful for further improvement of its protection capability. The crystallization method of the TiO₂ protection layer without damaging the Cu₂O photocathode is also feasible to improve the stability of the Cu₂O photocathode without a decreased PEC performance. The technique to form the hydrophobic surface on the Cu₂O photocathode is a promising strategy to improve the stability of Cu₂O photocathodes;
• Co-catalysts: The development of non-noble HER catalysts and the alleviation of noble components in HER catalysts is essential for the low-cost PEC water-splitting system. Although various HER catalysts have recently been developed [123], the deposition method should be considered for successfully applying to the Cu₂O photocathode. Furthermore, the bonding of the HER catalyst with a Cu₂O photocathode should be concerned for the durable Cu₂O photocathode because it is directly related to the stability;
• Upscaling: The reported high PEC performance of Cu₂O photocathodes is normally based on a small scale below 1 cm². In general, it is significantly reduced in the large-scale Cu₂O photocathodes [124]. Therefore, the research on maintaining its high PEC performance in the large-scale Cu₂O photocathodes is necessary, such as a novel design. Although a few groups have recently reported their works on the large-scale Cu₂O photocathode [124,125], more vigorous efforts on this are still essential for realizing the mass production of hydrogen via the PEC water-splitting system based on the Cu₂O photocathode in the future.

4. Conclusions

Cu₂O photocathodes have rapidly advanced for the practical PEC water-splitting system. Although some challenges remain to be overcome, such as the stability and the upscaling, the commercialization of the PEC water-splitting system using Cu₂O photocathodes is remarkably optimistic in the future, due to its potential for a high PEC performance and economic feasibility. Key strategies on each component in the Cu₂O photocathode, as suggested in this review paper, provide a shortcut for realizing this optimistic prospect. Furthermore, it will be the cornerstone of the successful entrance of a practical PEC water-splitting system into the eco-friendly hydrogen-fuel-based economy in the near future.