With the increasing energy consumption and the depletion of fossil fuels, it is urgent to develop clean, earth-abundant, and renewable energy resources to effectively address energy issues. Photoelectrochemical (PEC) water splitting has been found to be a promising approach for directly converting solar energy into sustainable and environmentally friendly hydrogen fuel.12.3.4.–5 Over the past four decades, even though great achievements have been obtained, the solar-to-hydrogen (STH) efficiency is still too low for practical applications.6 The key challenge for efficient solar hydrogen generation is to explore suitable photoelectrode materials that meet the following criteria: (1) broad range of solar spectrum absorption, (2) high photochemical stability, (3) efficient use of photogenerated electrons and holes, (4) suitable band edge positions, (5) low overpotential, and (6) low cost.78.9.–10 Typically, a single semiconductor is difficult for achieving all of the above requirements. For example, is stable in aqueous electrolytes under light illumination, but its large bandgap (3.0 eV) limits its light utilization to the UV region that accounts for only of the solar energy on earth, resulting in low STH efficiency.11,12 On the other hand, although multijunction silicon or heterojunction III–V semiconductors as photoelectrodes considerably increase the visible light conversion efficiency, they increase the cost while shortening the lifetime of the whole devices due to their photocorrosion in aqueous electrolytes.1314.–15 Therefore, the number of reliable and reproducible semiconductors for solar water splitting is limited. Combining the advantages of different semiconductors to form heterojunction systems or integrating state-of-the-art semiconductors with photovoltaic (PV) devices to form tandem cells may open up more opportunities, which have been extensively investigated during the past decade.1617.18.19.–20 Especially, self-driven monolithic and PEC tandem systems have attracted great attention due to the low cost when compared to the use of only PV devices for water splitting.222.214.171.124.–26 Nevertheless, the obtained STH and stability for the reported systems are still too low for the 10% target toward practical applications, even though many unique and novel material configurations have been revealed in recent years.27126.96.36.199.32.–33
In this review, we concisely summarize the recent progress and challenges in the unassisted PEC water splitting system, which has been considered as the ultimate solar water splitting process, compared to many reported PEC systems, which still rely on the applied external bias. We pay particular attention to nanostructure engineering of the photoelectrodes and the concepts of PEC configuration design for efficient water splitting. The review will start with some basic principles of unassisted PEC water splitting, followed by the design of photoelectrode materials based on nanostructure engineering and surface modifications. Then, the recent progress of PEC system configurations will be summarized. Finally, a brief overview including the challenges and perspectives of this research field will be given. We believe that this comprehensive review would not only provide important information about the recent progress of unassisted PEC water splitting systems but also could inspire readers to apply the discussed strategies to further improve the PEC solar fuel production performance.
Principles of Unassisted Photoelectrochemical Systems
Figure 1(a) presents the basic principle of a conventional PEC water splitting system. Under light illumination, electrons () will be excited from the valence band (VB) to the conduction band (CB), where holes () are left in the VB of the working electrode. Then, the electrons transport to the counter electrode and participate in the hydrogen-evolution half-reaction (HER). Meanwhile the holes transfer to the surface of the working electrode and are involved in the oxygen-evolution half-reaction (OER).34,35 Typically, overall water splitting can be achieved by applying an external bias if the semiconductor itself in the working electrode is suitable only for OER or HER. For example, some metal oxides, such as , , and , can be used as working electrodes for water splitting with a proper bias, even though their CB positions are unable to reduce water themselves.3637.38.39.–40
According to the general mechanism in PEC production, the selection of light absorbers and device configurations is important in driving effective water splitting reactions. A Z-scheme concept has been developed through a two-step excitation mechanism induced by coupling two different light absorbers, which is inspired by natural photosynthesis in green plants.4142.43.–44 The charge transfer in a Z-scheme system is different from that in the general p-n heterojunction configuration. For p-n heterojunctions, photogenerated electrons in photosystem II (PS II) with higher CB are transferred to photosystem I (PS I) with lower CB to be involved in HER, while holes are transferred to the opposite direction for OER due to the built-in electrical potential in the interfaces [Fig. 1(b)]. By contrast, photogenerated holes in the VB of PS I are promoted in OER, while electrons in the CB of PS II are used for HER in the Z-scheme configuration [Fig. 1(c)]. Thus, the Z-scheme system demonstrates the strong oxidized ability of PS I and the strong reduced ability of PS II. The advantage of the Z-scheme water splitting system is to utilize a wider range of visible light because the energy required to activate each photocatalyst can be reduced as compared to the one-step water splitting system.4546.–47 Successful overall water splitting via two-step photoexcitation under visible light illumination has been reported by combining several semiconductors.4849.50.–51 Especially, the artificial photosynthesis (AP) concept has emerged as an attractive approach to mimic the natural photosynthesis process by utilizing suitable semiconductors to efficiently drive the thermodynamically uphill water splitting reaction under sunlight illumination.5253.54.–55
Recent Progress and Challenges of Unassisted Photoelectrochemical Devices
In this section, recent efforts on the development of novel nanostructured design concepts and surface modification strategies for an unassisted PEC water splitting system will be briefly presented and discussed in terms of the improvement of performance.
Nanostructured Semiconductor Design
In general, PEC performance of the devices depends on the efficient light utilization, photogenerated carrier transport/separation, and stability in the PEC system. Nanostructured semiconductors have been revealed to be promising for PEC water splitting due to their large surface area and size-dependent properties (e.g., tuned bandgap energy, shortened carrier collection pathways, and efficient charge transfer).56 Therefore, the selection of a suitable semiconductor with excellent nanoarchitecture is of importance for PEC system construction. n-type semiconductors (e.g., , , , and ) are commonly developed as photoanodes or top absorbers in tandem devices due to their low VB energy positions with more positive VB potential for water oxidation reaction.5758.59.60.–61 However, when these semiconductors are used as photoelectrodes, the randomly oriented bulk nanoparticles (NPs) in photoelectrodes might decrease the PEC performance with a high charge recombination rate due to a large number of charge trap sites. To overcome this issue, well oriented one-dimensional (1-D),62188.8.131.52.67.–68 two-dimensional (2-D),6970.71.72.–73 and three-dimensional (3-D)7475.–76 nanostructures have been developed in recent years. The 1-D nanostructures including nanowires,7778.–79 nanorods,8081.–82 and nanotubes8384.85.–86 provide both efficient light harvesting and superior charge transport for PEC water splitting. Meanwhile, 2-D nanostructures have also shown interesting properties in various photochemical fields.8788.–89 In particular, metal oxide and chalcogenide 2-D nanosheets are unique nanostructures with high-specific surface area and excellent crystallinity, which is beneficial for efficient charge separation and migration in water splitting.9091.92.93.–94 More recently, researchers have integrated different functional materials of 1-D or 2-D structures to construct 3-D hierarchical nanostructures in order to increase large surface areas for light harvesting without inhibiting charge transfer and separation.9596.97.98.–99 Hence, it is a promising nanostructure model for efficient water splitting due to broader light absorption, rapid charge migration and separation.
On the other hand, p-type semiconductors such as boron-doped Si,100 ,101 ,102 and 14 exhibit small bandgaps, thereby a significant proportion of visible light can be harvested. However, when p-type materials are employed to PEC water splitting, severe challenges such as large overpotential, instability, and low absorption coefficient have been observed.103 Thus, recent strategies have been focused on improving the photostability through loading suitable protective layers or coupling with n-type semiconductors to form heterojunctions.104 For example, has been reported as an effective protective layer for enhancing the photostability and photoactivity of p-type photoelectrodes.105106.–107 Such a heterojunction concept by coupling different photoanode and photocathode layers would be a promising approach for unassisted PEC water splitting systems. For designing such systems, the nanostructured morphology and crystallinity of semiconductors with suitable CB and VB positions for water splitting and the interfacial properties in the heterojunction should be carefully considered. Moreover, the fabrication process of the photoelectrodes in a PEC system should be optimized by exploring simple, cost-effective, and environmentally friendly techniques. Figure 2 represents the engineering strategies that are critical to meet all requirements for efficient PEC water splitting. The recent reported PEC systems, which have been developed based on these strategies, will be further presented and discussed in Sec. 3.3.
Surface Modification of Photoelectrodes
In addition to nanostructure engineering, the surface modification of photoelectrodes is also important for achieving high PEC performance. As the surfaces of the photoelectrode materials are active sites for redox reaction, the surface states are closely related to photon energy absorption properties and surface charge transfer ability of the semiconductors. Doping is one common strategy to improve the conductivity of semiconductors and extend the lifetime of charge carriers by reducing recombination of the photogenerated electron–hole pairs.108184.108.40.206.113.114.115.–116 In addition, loading HER or OER cocatalysts on the surface of the semiconductor is also an efficient approach to reduce the charge recombination at a lower applied potential and increase the charge transfer for the water splitting process.117 Cocatalysts should be optimized to offer highly active sites to separate the photogenerated charges for HER or OER half reactions, avoiding the photogenerated electrons and holes with high redox ability to react with the semiconductor itself, which improves the stability. Noble metals (e.g., Pt) and noble metal oxides (e.g., , ) are usually used as effective HER and OER cocatalysts, respectively.118,119 Nevertheless, it is not favorable to scale up their utilization in PEC devices due to their cost and rarity. Thus, noble metal-free cocatalysts, such as cobalt compounds,120220.127.116.11.125.–126 nickel-based materials,127,128 and chalcogenides129130.131.–132 for photoelectrode surface modification, have been developed.
In addition, plasmonic NP coupling is another strategy to utilize the optical properties of nanometals for better light harvesting.133134.–135 Typically, plasmonic metal NPs can be employed to capture light and transfer the excitation energy through the localized surface plasmon to a combined semiconductor with large bandgap. In general, when plasmonic metal nanostructures are contacted with semiconductors, they should be able to alleviate charge recombination and enhance light absorption, resulting in higher PEC performance than their pure semiconductor counterparts.13618.104.22.168.22.214.171.124.145.–146 For instance, Zhang et al.97 studied PEC performance of Au NPs deposited on the bilayered structure photoanode and reported a STH efficiency of 0.71% achieved, around which is one of the highest values reported in -based photoanodes. The excellent PEC water splitting activity in visible light can be ascribed to the generated energetic hot of Au NPs that were injected into the CB of through the excitation and decay of surface plasmons. In another study, the utilization of a plasmonic Au nanohole array combined with nanorod array photoanode led to significant enhancement of the photocurrent density147 and a plasmon-induced resonant energy transfer (PIRET) effect was proposed to explain this phenomenon. In addition to photoanodes, CuO nanowires decorated with Ag NPs as photocathodes were also studied in which the surface plasmon energy transfer from Ag NPs contributed to enhanced photocurrent with a STH efficiency of 2.16% achieved.148 The use of plasmonic Ag NPs deposited on photoanodes can also further improve the photocurrent density by LSPR property of Ag NPs.149,150
Recent works on nanostructure engineering and surface modification for innovative PEC system design are summarized in Table 1. It is apparent that many efforts have focused on controlling nanostructures, doping with suitable elements, and loading cocatalysts in heterojunction photoelectrodes. However, the high photocurrent and STH efficiency achievements in most of the systems are limited and external bias is generally required for overall water splitting. Notably, multilayered PEC devices composed of nanostructured metal oxide semiconductors, perovskite solar cells (PSCs), and Si photoelectrodes can exhibit high STH up to 10%, whereas such PEC devices are expensive, complicated, and unstable in aqueous electrolyte under light illumination. In order to overcome these drawbacks, the selection of excellent crystalline photoelectrode materials and suitable electrolytes is essentially important, and the overall PEC system should be considered by designing the configuration of multilayered photoelectrodes with protective layers. Such conditions would be the key factors in developing unassisted PEC systems for overall water splitting.
PEC devices for solar water splitting.
|Photoelectrode system||Cocatalysts||Photocurrent (mA cm−2)||Light source||STH/IPCE (%)||References|
|Pt||120 at||11 suns,||Ref. 24|
|3jn-a-Si||OER: Co–Bi, HER: NiMoZn||at||1 sun, AM 1.5 G||(wireless) and 4.7% (wired)||Ref. 151|
|Si pillars||at||,||Ref. 130|
|Pt||at||1 sun, AM 1.5 G||Ref. 152|
|Pt||at||, AM 1.5 G||Ref. 153|
|AlGaAs/Si||black||20.1, unassisted||Ref. 154|
|W-doped||Co-Pi||4 at||1 sun, AM 1.5 G||Ref. 155|
|on 1D ZnO nanorods||Co-Pi||3 at||, AM 1.5 G||Ref. 68|
|nanorod||6.7 at||1 sun, AM 1.5 G||Ref. 156|
|Pt/Ru||15 at||Ref. 157|
|nanotree||HER: Pt OER:||—||Ref. 79|
|−AuNPs||Au||2.6 at||Ref. 158|
|5.76 at||AM 1.5 G||at 400 nm and||Ref. 95|
|Pt-doped||Co-Pi||4.32 at||1 sun, AM 1.5 G||at 400 nm and||Ref. 159|
|Pt||at||AM 1.5 G||between 350 and 480 nm,||Ref. 160|
|at||AM 1.5 G||Ref. 131|
|HER: OER: Co-Pi||1, unassisted||Ref. 161|
|Au at||—||at||AM 1.5 G||at 340 nm||Ref. 43|
|—||at||AM 1.5 G||at||Ref. 162|
|Pt||at||300 W Xe lamp||at||Ref. 163|
|Pt||8.1 at||AM 1.5 G||at 500 nm and||Ref. 164|
|with NiFe DLH/Ni foam||NiFe DLH||10, unassisted||AM 1.5 G||Ref. 165|
|Ni||12 at||1 sun, AM 1.5 G||—||Ref. 166|
Artificial Photosynthesis Photoelectrochemical System Design
Tandem cell configuration
Tandem cell configuration is a promising approach to address the issues of single or heterojunction PEC devices for unassisted solar water splitting. In the early stage, tandem PEC devices consisting of a semiconductor and a dye-sensitized solar cell (DSSC) were applied for water splitting. For instance, Kim et al.167 revealed a tandem cell composed of a photoelectrode connected with a DSSC for unassisted water splitting where the maximum STH efficiency of 0.35% was obtained. Subsequently, a and photoanode with a DSSC in tandem configuration was developed, as illustrated in Fig. 3.168 Without external bias, the achieved STH efficiencies were 3.10% and 1.17% for the and tandem systems, respectively. The low STH efficiency in this tandem configuration is due to the mismatched band energy configuration and interfacial charge recombination between photoelectrodes and the combined DSSCs. Especially, charge recombination occurred between the injected electrons from the semiconductors and the oxidized dye (), inhibiting the efficient redox reaction with water in the system.
More recently, an efficient wireless monolithic tandem device consisting of a bipolar configuration with a highly transparent -sensitised mesoporous film and a porphyrin-dye-based photoelectrode was reported.169 An extraordinarily high STH efficiency of 5.7% and evolutions were revealed without any applied bias. The porphyrin-dye-sensitized photoanode with a cobalt electrolyte generated sufficient bias to drive the water splitting reaction forward in this monolithic tandem configuration. Although the efficient design for AP is controversial, the combination of abundant metal oxide-based photoelectrodes with DSSCs has been explored as a promising PEC device configuration for unassisted solar water splitting. Nevertheless, the DSSC component generates low voltages that are not sufficient for water splitting, and more attention has been paid to recently developed PSCs due to their much higher (0.9 to 1.5 V).170 For instance, a dual artificial-leaf-type tandem PEC device composed of a robust cobalt carbonate (Co-Ci)-catalyzed Mo-doped photoanode and a -based PSCs (Fig. 4) can deliver a photocurrent density of at 1.23 V versus RHE for the unique and stable photoanode and a STH efficiency of 3.0% for the wireless device.171 Gratzel et al. reported a novel PSCs-assisted PEC water splitting system composed of organohalide -based PSCs as the external power supply and NiFe-layered double hydroxides as the catalytic electrodes, which achieved a photocurrent density of with a STH efficiency of 12.3%.165 Although this device configuration exhibited a high efficiency for PEC water splitting, the electrodes were not directly settled in photoreactions. Consequently, the hydrogen generation occurred by a water electrolysis process that power was generated through the externally wired two PSCs in series. Meanwhile, a monolithic tandem design with only one PSC/Ni layer without an external wire connection was developed and a high photocurrent density of was achieved (Fig. 5).166 Overall, the device complexity and stability issue of the dye and organohalide perovskite materials in the DSSCs and PSCs, and efficient photon utilization should be considered in these design concepts.
On the other hand, coupling a PEC device with a PV cell in tandem was developed to extend the light absorption range and improve the photocurrent density in unassisted solar water splitting. Ding et al.172 achieved a STH efficiency up to 2.5% in an unassisted water splitting system using an earth-abundant photoanode and a Si-solar-cell-based photocathode (Fig. 6). The parallel irradiation mode exhibited higher efficiency than the tandem illumination one because the former system offered sufficient driving force for water splitting. More recently, an overall STH efficiency of 0.91% was achieved on a system composed of a photoanode and a Si photocathode with and as cocatalysts (Fig. 7).173
Despite continuing efforts on searching various photoanode materials to combine with solar cells, the high current densities are accompanied with the challenges of photocorrosion or chemical instability under oxidative conditions. A stand-alone unassisted PEC device would be an ultimate consideration for evolutions in water splitting. Accordingly, simpler concepts using dual-absorber tandem devices have been explored with acceptable STH efficiencies. For example, a p-n junction photoelectrode comprising p-type Cu-Ti-O and n-type nanotube arrays with efficient spectral photoresponse (380 to 885 nm) and antiphotocorrosion properties can result in a significantly enhanced photocurrent density of and a photoconversion efficiency of 0.30% under global AM 1.5 G illumination (Fig. 8).174 Similarly, a Mn-oxide– coated triple junction-amorphous-Si (Tj-a-Si) was used for monolithic self-driven water splitting with a high STH efficiency of 3.25% and good long-term stability under solar-simulated light illumination.175
In a parallel study, Reece et al.151 reported a Tj-a-Si system modified with cocatalysts for water splitting. The wireless configuration of a Tj-a-Si with Co-oxygen-evolving complex and NiMoZn as OER and HER cocatalysts presented a high STH efficiency of 4.7% under AM 1.5 G simulated sunlight illumination. It has been confirmed that this design can generate electrons and holes with enough energy to split water () in a simply engineered configuration. However, the high cost of the photoelectrode materials and the complexity of the fabrication process would be critical limiting factors in commercializing this water splitting system.
Heterojunction components of photoelectrochemical device design
Most studies have been focused on improving the performance of individual semiconductors to achieve the targeted STH efficiency. However, the performance of the whole PEC system and the interaction between the individual components for solar energy conversion should also be explored. Liu et al.79 reported a fully integrated nanotree PEC system consisting of cocatalysts loaded Si and nanowires for direct solar water splitting (Fig. 9). An STH efficiency of 0.12% was achieved in this unassisted PEC water splitting system. However, this nanotree PEC system still needs to improve the current density and maximize the STH efficiency via excellent interfacial engineering and surface modifications. More recently, a tandem PEC system consisting of a n-p homojunction of Si microwire arrays coated with thin films presented a current density of and a STH efficiency of 0.6% without external bias (Fig. 10).176 It is well known that the use of Si photoelectrodes in a PEC water splitting system reduces the cost competitiveness, thus novel systems with low cost materials have continuously been explored.
Another promising concept for unassisted PEC water splitting systems is a standalone Z-scheme configuration. To obtain high performance, we need to consider the main factors, in particular, the design of the fully integrated system, the functions of the individual materials, and the optimization of the interfaces between the individual components.177 One typical example is the design of W/Mo-doped and dual n-type photoelectrode system with the presence of redox mediators ( or ),178 which highlights the challenging nature of designing redox mediator-free Z-scheme systems via metal doping and surface modification for efficient water splitting. Likewise, Yang et al.179 studied the surface modification with different catalysts in order to improve the overall water splitting efficiency and stability of the system (Fig. 11). They designed a system comprising a CdS nanorod photoanode and a ZnS/CdSe nanosheet photocathode for self-driven PEC water splitting, which delivered a maximum STH efficiency of 0.17%, comparable to natural photosynthesis. Later, a system composed of dual nanowire photoelectrodes of p-InGaN/Si-InGaN was developed, where a STH efficiency of 2% at 0.6 V versus RHE and a high open circuit potential were achieved under AM 1.5 G 1 sun illumination (Fig. 12).180 Nevertheless, this design requires expensive and rare cocatalysts to drive the solar water splitting reaction. Therefore, other active and cheaper semiconductors and cocatalysts for overall water splitting should be explored in future studies.
Summary and Perspectives
In this review article, we provided an overview of a library of PEC system designs toward unassisted solar water splitting. It is clear that the design of efficient unassisted PEC devices relies not only on material design but also the integration of PEC configuration. In particular, the development of semiconductors via novel nanostructure engineering, surface modification with exotic element doping or cocatalyst loading, and innovative system design based on heterojunction configurations are important strategies for improving light harvesting, charge separation, and surface reaction kinetics. This brief review would deliver useful information for further construction and development of efficient unassisted PEC water splitting systems with high STH efficiency, long-term stability, and low cost. For example, the stability and poor kinetics of water splitting issues in the wireless integrated tandem system with dual light absorber configuration should be addressed by employing earth abundant materials, protective layers, and morphology engineering techniques. On the other hand, novel strategies to inhibit photocorrosion of the photocathode materials are urgently needed for efficient water splitting in a long-term period. Additionally, the tandem system combining semiconductors with a PV device has a conflict in terms of instability in aqueous electrolyte, high cost and complexity. Using DSSCs or PSCs in tandem systems for unassisted water splitting demonstrates low cost and comparable STH efficiencies, but the efficient light absorption and stable organic dyes for DSSCs and the stability of perovskites should be considered when equipped in a PEC system.
Although numerous achievements have been obtained in recent years, it is still highly challenging to develop an efficient unassisted PEC water splitting system for practical applications. Currently, the tentative STH goal for practical applications of the unassisted PEC water splitting device is 10%. It is expected that some novel concepts of fully integrated PEC devices put forward the breakthrough in solar hydrogen energy conversion and long-term stability. Notably, the Z-scheme concept for unassisted PEC water splitting systems is promising for effective photon utilization and efficient solar water splitting. However, there still remains significant room for the improvement of this system for practical applications. It is also important to explore new active and low cost cocatalysts that can efficiently promote the forward reactions for water splitting. To this end, in-depth understanding of the mechanisms of unassisted PEC water splitting and the strategies to boost charge transfer and separation for redox reaction is critically important to develop an efficient solar water splitting system for more sustainable solar energy utilization.
The authors would like to acknowledge financial support from the Australian Research Council through its DP and FF programs. P.P. acknowledges support from the University of Queensland for International Scholarships (UQI). S.W. acknowledges support from IPRS and UQ Centennial Scholarships.
Piangjai Peerakiatkhajohn received her MSc degree from the Joint Graduate School of Energy and Environment (JGSEE), King Mongkut’s University of Technology Thonburi in 2011. Currently, she is a PhD student under the supervision of Prof. Lianzhou Wang at School of Chemical Engineering, The University of Queensland (UQ), Australia. Her research interest focuses on the design and development of novel photoelectrochemical system for efficient solar energy conversion.
Jung-Ho Yun received his PhD degree in chemical engineering from the University of New South Wales in 2012. Before his PhD, he worked at NANOPAC, Korea, for energy/environmental applications using functional nanomaterials between 2003 and 2007. Currently, he is working as a postdoctoral research fellow in Prof. Lianzhou Wang's Group, The University of Queensland. His research focuses on novel nanostructured photoelectrodes for photoelectrochemical system and solar cells.
Songcan Wang received his BEng. (2011) and MEng. (2014) from the Central South University (CSU), China. Currently, he is a PhD student under the supervision of Prof. Lianzhou Wang at School of Chemical Engineering, The University of Queensland (UQ), Australia. His research interests are mainly focused on the design and development of high performance photoanodes for photoelectrochemical energy storage and conversion.
Lianzhou Wang is currently a professor in the School of Chemical Engineering and director of the Nanomaterials Centre, the University of Queensland (UQ), Australia. He received his PhD degree from Shanghai Institute of Ceramics, Chinese Academy of Sciences, in 1999. Before joining UQ in 2004, he worked at two national institutes (NIMS and AIST) of Japan for 5 years. His research interests include the design and application of semiconducting nanomaterials in renewable energy conversion/storage systems, including photocatalysts and low-cost solar cells.