Rational design of semiconductor-based photocatalysts for advanced photocatalytic hydrogen production: the case of cadmium chalcogenides

Abstract

Semiconductor-based photocatalytic hydrogen (H₂) production from water has recently received considerable attention because of its enormous potential for solving worldwide ever-increasing energy crises and environmental issues. Among various semiconductors, cadmium chalcogenides (CdX, X = S, Se, Te), with a variety of superior properties including appropriate bandgaps for visible-light absorption, proper conduction band edge potentials for water reduction and abundant reserves on the earth, have fuelled great interest in exploring their photocatalytic properties for solar-driven H₂ production from water. This review article summarizes the recent progress in developing CdX-based photocatalyst systems for the production of H₂ from solar water splitting. The basic mechanistic fundamentals of CdX photocatalysts and various strategies to enhance the activity and stability of this class of photocatalysts are introduced in detail. Finally, some key scientific issues and prospective directions in this area of research are also discussed.

---

Bin Zhang

Bin Zhang received his PhD degree from the University of Science and Technology of China in 2007. He carried out postdoctoral research at the University of Pennsylvania (July 2007–July 2008) and worked as an Alexander von Humboldt fellow at the Max Planck Institute of Colloids and Interfaces (August 2008–July 2009). Currently, he is a professor in the chemistry department at Tianjin University and the Collaborative Innovation Center of Chemical Science and Engineering (Tianjin). He mainly focuses on the development of chemical transformation strategies to prepare porous and ultrathin nanomaterials and their hybrids for energy catalytic applications.

---

1. Introduction

Rising global energy demands, rapidly diminishing reservoirs of traditional fossil fuels and ever-increasing environmental concerns have motivated the significant requirements for clean and renewable energy sources.^1–5 Hydrogen (H₂), with its relative abundance of sources of generation, high density of energy (122 kJ g⁻¹) and environmental friendliness, is regarded as an attractive and alternative energy carrier without CO₂ emissions.^6–8 Currently, H₂ used in industry is mainly obtained from non-renewable fossil sources by means of steam reforming or coal gasification processes. However, the production processes based on these pathways are usually complex, expensive, environmentally hazardous and unsustainable, not to mention the generation of greenhouse gases (CO₂) as a by-product. Recently, splitting water for H₂ production using solar energy has received much attention due to the fact that it can convert sustainable solar energy into storable renewable chemical fuel (H₂), which provides a promising solution to the worldwide ever-increasing energy crises and environmental issues. Approaches for solar water splitting mainly include thermochemical water splitting, photobiological water splitting, photochemical or photocatalytic water splitting and photoelectrochemical or photoelectrocatalytic water splitting. Among these approaches, photochemical water reduction for H₂ production using semiconductor-based photocatalyst systems, which can utilize solar energy to drive a thermodynamic uphill reaction to generate molecular H₂, is the focus of considerable attention.

Since the pioneering discovery of photoelectrochemical water splitting on a TiO₂ photoanode by Honda and Fujishima in 1972,⁹ the technology of semiconductor-based photocatalytic water splitting has undergone considerable development and numerous active semiconductor-based photocatalyst systems have been explored in this important field. In general, the photocatalytic materials for water splitting are required to perform at least two fundamental functions: light harvesting of the maximal possible part of the solar energy spectrum and a catalytic function of efficient water decomposition into O₂ and H₂. Thus, to efficiently utilize solar energy for photocatalytic H₂ production from water, an ideal semiconductor material should meet at least three requirements: an appropriate bandgap for solar energy absorption, proper conduction band position for water reduction, and good stability under the required reaction conditions.

Generally, H₂ and/or O₂ generation from water using solar energy and semiconductor materials can be achieved by means of two different kinds of systems: photoelectrochemical water splitting systems (including single-photoelectrode and dual-photoelectrode systems) and photocatalytic water splitting systems (including half-reaction and overall water splitting systems). In brief, the basic principle of photoelectrochemical water splitting is based on the conversion of solar energy into electricity, which is then used for water electrolysis, within a photoelectrochemical cell involving two electrodes immersed in an electrolyte solution. In this case, at least one electrode is made of a semiconductor which is able to efficiently absorb the sunlight and generate photoexcited charge carriers. There are three options for the arrangement of photoelectrodes in the assembly of photoelectrochemical cells for the water splitting reaction: (i) an n-type semiconductor as the photoanode and a metal counter electrode (usually Pt); (ii) a p-type semiconductor as the photocathode and a metal counter electrode (usually Pt); (iii) an n-type semiconductor as the photoanode and a p-type semiconductor as the photocathode. In photoanodes made of n-type semiconductors, photogenerated holes accumulate on the surface of the n-type semiconductor and can be used to oxidize water to generate O₂ (if its valance band edge is more positive than the O₂ evolution potential), while electrons are transferred to the counter electrode via the back contact and an external circuit, and used for the H₂ evolution reaction (case i). On the other hand, photocathodes made of p-type semiconductors can utilize photogenerated electrons for H₂ evolution when their conduction band edge is more negative than the H₂ evolution potential; simultaneously, the holes are transferred over the external circuit to the counter electrode for O₂ evolution (case ii). Alternatively, a photoanode and a photocathode can be connected in tandem to generate a dual-photoelectrode system (case iii). With the help of the external electric field, which facilitates the charge carrier separation and greatly reduces the recombination possibility of charge carriers, photoelectrochemical systems usually exhibit relatively high efficiency for water splitting.

Compared to photoelectrochemical systems, the main advantages of photocatalytic water splitting systems are their simple operation and low cost. Photocatalytic water splitting systems can be divided into two types: single photocatalyst systems and Z-scheme systems. A semiconductor material can work as a photocatalyst for photocatalytic H₂ and/or O₂ production from water if it meets all the requirements (e.g. band gap, band edge positions). Note that a single photocatalyst overall water splitting system under visible light is one of the greatest challenges in the field of water splitting. In most cases, water reduction or oxidation half-reactions are studied in this type of system in the presence of appropriate sacrificial reagents for H₂ or O₂ generation. Alternatively, two semiconductors can be connected (in series with reversible redox shuttles in some cases) to construct Z-scheme systems, which can achieve overall water splitting.

Concerning the mechanisms of the reactions, the principle of photocatalytic water splitting is similar to that of photoelectrochemical water splitting. The essential difference between the two consists of the location of the sites of the surface reactions. In the photoelectrochemical process, H₂- and O₂-evolution reactions take place at two spatially separated electrodes. In the photocatalytic process, both oxidation and reduction occur on the surface of the photocatalyst, which exhibits the functions of both anode and cathode. The practical difference between photocatalytic and photoelectrochemical water splitting is that the latter results in both O₂ and H₂ evolving separately, while in the former, a mixture of both gases is evolved. Moreover, as the photoelectrochemical reactions on photoelectrodes are driven by photoexcited minority carriers, the material selection for constructing the photoanode and photocathode is limited not only by their band gap and band edge positions but also their intrinsic conductivity characteristics (n-type or p-type). Differing from the photoelectrochemical water splitting process, both n-type and p-type semiconductors can work as photocatalysts for photocatalytic water splitting when their band gap and band edge positions meet the requirements. This review is focused on the development of photocatalytic systems for H₂ generation.

Among the various promising systems under investigation, cadmium chalcogenides (CdX, X = S, Se, Te) are regarded as good candidates for photocatalytic H₂ production because of their following superior characteristics:^10–12 (1) CdX are visible-light-responsive semiconductors, which enables them to absorb solar energy across a broader spectrum than UV-absorbing semiconductors; (2) CdX nanocrystals possess excellent light-harvesting characteristics, with molar absorptivities of about 10⁵–10⁷ M⁻¹ cm⁻¹; (3) their bandgaps, redox potentials and absorption spectra, which largely determine their photocatalytic properties, can be tuned by controlling a series of structural parameters such as composition, size, shape and morphology; (4) they can be surface-functionalized with a rich variety of ligands, which not only enable them to possess either hydrophilic or hydrophobic properties, but also provides selectivity of functional groups for interaction with other materials. One can see that CdX materials, which have been extensively studied for photocatalytic H₂ generation and which we extensively refer to in this article, are indeed potential abundant and useful water splitting materials, even though there are environmental and toxicity concerns about potentially harmful release of cadmium.

During the past four decades, significant interest and considerable efforts have been directed to the development of CdX-based materials as photocatalysts for visible-light-driven H₂ generation due to their unique physical, chemical and optical properties. To date, numerous successful achievements have demonstrated their significant applications in photocatalytic H₂ production, while recent great progress in materials science and nanotechnology has also greatly accelerated the evolution of this field. Up till now, many excellent earlier and recent review articles have been published by several research groups to summarize the achievements and advances in the field of semiconductor-based photocatalytic water splitting.^13–27 However, there is no one review focusing on CdX-based photocatalysis for H₂ production. Therefore, a timely review to systematically summarize the related progress and advances in CdX-based photocatalytic H₂-evolution systems is highly desirable. In this review article, we will first describe the basic mechanistic fundamentals of CdX-based photocatalysts. We will then discuss the various approaches to improve the photocatalytic performance of various kinds of CdX-based photocatalysts. The synthetic procedures, structural features and photoactivity enhancement mechanism are discussed based on some representative and important examples. In particular, the focus will be on the relationship between the structural characteristics and features of photocatalysts and their photocatalytic performance. Finally, some key scientific issues as well as prospective directions are also discussed.

2. The mechanistic fundamentals of CdX-based photocatalysts

Semiconductor-based photocatalysis is defined as the chemical reaction induced by photoirradiation in the presence of one or more semiconductors as the catalysts, or more specifically, the photocatalysts. Basically, as illustrated in Fig. 1a, semiconductor-based photocatalysis for the splitting of water into H₂ and O₂ involves the following three major steps: (i) absorption of solar energy by the semiconductor and generation of mobile charge carriers; (ii) photogenerated charge separation and diffusion from the photogenerated center to the semiconductor's surface; (iii) charge utilization for water reduction or oxidation on the photocatalyst surface. The first two steps (i and ii) relate to the photon absorption and charge carrier transfer in the semiconductor, which are analogous to the photovoltaic process, and the last step (iii) is carrier-induced surface reactions. In a semiconductor-based photocatalysis process, the intrinsic properties of the semiconductor materials fundamentally determine the overall efficiency of the photocatalytic reaction.¹³ Generally, the light-absorbing capability of semiconductor materials largely depends on their band gap. Semiconductors with a wider band gap can only utilize the parts of the solar energy spectrum in the UV or near UV regions, while those with a narrower band gap can respond to visible light. From the perspective of solar energy utilization, those visible-light-responding semiconductor materials are more promising photocatalysts as they can utilize solar energy more efficiently. Theoretically, the minimum band gap of semiconductor materials for the water splitting reaction appears to be 1.23 eV according to the redox potential of H⁺/H₂ and O₂/H₂O redox pairs. In addition to this minimum thermodynamic requirement, the additional kinetic overpotential associated with the electron transfer processes and gas evolution step due to energy losses should be taken into account.²⁸ Therefore, a much larger band gap (usually >2.0 eV) of the semiconductor is often required for practical photocatalytic systems. Besides the band gap, the positions of the valence band (VB) and conduction band (CB) of a semiconductor, which determine the reduction and oxidation power of photogenerated carriers, are also important for the process of photocatalytic water splitting. Specifically, for H₂ production from photocataltic water-splitting, the bottom of the CB should be higher (more negative) than the hydrogen evolution potential (+0 V vs. normal hydrogen electrode at pH 0), while the top of the VB should be lower (more positive) than the water oxidation potential (+1.23 V vs. normal hydrogen electrode at pH 0) for water-splitting O₂ production.^13,16 Among the CdX semiconductors, CdS and CdSe are n-type semiconductors with direct bandgaps of ∼2.5 and ∼1.7 eV, respectively, while CdTe is a p-type semiconductor with a direct bandgap energy of ∼1.4 eV (Fig. 1b). From a thermodynamics perspective, CdX semiconductors are promising candidates for photocatalytic water reduction because they possess more negative CB potentials compared to the reduction potential of water. In addition, their electronic band structure and the positions of the CB and VB can be finely tuned by modifying their crystal structure, composition and morphology.²⁹ Numerous research results have demonstrated that precise engineering of a material’s composition and structure can lead to improved light-harvesting efficiency or/and enhanced reduction/oxidation abilities of the photogenerated carriers, which finally results in an improved photocatalytic activity.

Fig. 1 (a) Illustration of the main processes in semiconductor-based photocatalytic water splitting for H₂ and O₂ evolution. (b) Relationship between band structure of CdS, CdSe and CdTe semiconductors and redox potentials of water splitting (pH = 0).

After excited charges are generated upon light absorption by a semiconductor, efficient charge separation and fast charge transport are fundamentally important for semiconductor-based photocatalysis. During step (ii), the recombination of the photogenerated charge carriers inside the CdX photocatalyst competes with the charge separation/migration process, which greatly affects the overall efficiency of the photocatalytic reaction. Charge recombination in step (ii) can occur via different mechanisms, including radiative and nonradiative recombinatiom.¹⁵ It is well recognized that the crystal structure, crystallinity, and particle size of semiconductor materials strongly affect step (ii). Currently, considerable efforts have been made to increase the separation efficiency and to reduce the recombination probability of photoexcited charge carriers in CdX-based photocatalysts. One popular strategy is structure and morphology control, that is, to develop novel CdX-based photocatalysts with controlled structures and morphologies. Another promising option is to integrate the CdX with other materials to form heterogeneous photocatalysts. When carefully designed, some photocatalytic behaviours, including the direction of mobility, the diffusion distance, and the recombination rate of photogenerated charge carriers in these CdX-based heterogeneous photocatalyst systems comprising two or more materials or phases can be well-controlled. Beside these two strategies, another alternative strategy for the extraction of photogenerated electrons from the CdX semiconductor to reduce the recombination probability of photogenerated carriers is the hybridization of CdX with carbon-based materials (such as carbon nanotubes or graphene). Taking the advantages of carbon-based materials, this kind of CdX-based composite can exhibit improved photoactivity.

It should be emphasized that, even if the separated photogenerated charge carriers possess thermodynamically sufficient potential for H₂- and O₂-evolution reactions, they may recombine if there are no suitable active/reaction sites available on the photocatalyst surface. Therefore, the surface chemical reactions in step (iii), which are driven by photogenerated carriers, are crucial for achieving high efficiency in charge utilization. Water molecules in contact with the semiconductor surface can be reduced to form H₂ by photoelectrons from the CB and oxidized to form O₂ by photoholes from the VB, if the energetic positioning of the bands matches the redox potentials of water reduction and oxidation. It is well recognized that loading appropriate cocatalysts on the surface of photocatalysts can promote or accelerate the surface chemical reactions.^17,18,30 On one hand, cocatalysts can serve as the active reaction sites where H₂- or O₂-evolution reactions occur. On the other hand, cocatalysts can facilitate the charge separation and reduce the overpotential and activation energy of H₂- or O₂-evolution reactions. Additionally, cocatalysts can enhance the photostability of semiconductor photocatalysts. So far, many kinds of materials, including metals, inorganic compounds, hydrogenases and molecular complexes, have been employed as cocatalysts to promote or accelerate the surface redox reactions of CdX-based photocatalysts. Some of them display high efficiency and good robustness to significantly enhance the photocatalytic H₂ evolution activity of CdX-based photocatalyst systems.

Apart from the above mentioned aspects, photocorrosion, which results from hole-driven oxidation reactions, is another important issue for CdX-based photocatalysts. Taking CdS as an example, S²⁻ in the CdS rather than a water molecule is oxidized by photogenerated holes, accompanied with elution of Cd²⁺ ions in aqueous media when exposed long-term to visible light,^12,16,31 according to eqn (1). Photocorrosion is a critical drawback of CdX photocatalysts, which seriously limits their practical applications in photocatalysis. Sacrificial electron donors are often needed to consume the photongenerated holes or provide electrons for proton reduction when using CdX semiconductors for photocatalytic H₂ production from water. Frequently used hole scavenger materials in CdX-based photocatalytic H₂-evolution systems include lactic acid, a mixture of sulphide and sulphite, ascorbic acid, tertiary amines, triethanolamine and alcohols.

CdS + 2h⁺ → Cd²⁺ + S (1)

Taking these mechanistic fundamentals into account, an efficient and stable photocatalyst system based on CdX semiconductors should achieve high efficiency in visible-light harvesting, facilitate the photogenerated charge carrier separation and transfer and effectively prevent charge recombination, whilst efficiently suppressing photocorrosion, and possess good photostability. Synthetic advancement makes it possible to design and synthesize the desired nanostructures and nanomaterials with a particular band alignment favouring an expected charge separation. During the last decades, extensive efforts have been made to enhance the photocatalytic activity and stability of CdX semiconductors. The important and representative strategies will be discussed in detail in the following section based on some selected examples.

3. Strategies for enhancing the photocatalytic performance of CdX

3.1 Modifying the crystal structure and morphology of CdX

Bulk CdX materials alone demonstrate low photoactivity and poor stability for H₂ photogeneration, mainly due to the rapid surface/bulk recombination of photogenerated electrons and holes and large H₂ production activation energy or overpotential. Compared to their bulk counterparts, CdX materials on the nanoscale show more potential for photocatalytic applications owing to their unique properties and advantages. Firstly, the decrease in particle size can usually directly result in an increase in the surface area and active reaction sites, both of which are important for surface chemical reactions. Secondly, reducing the particle size of a photocatalyst can shorten the diffusion pathway of the photogenerated charge carriers, leading to decreased recombination probability. Thirdly, making semiconductor materials in the form of nanoparticles allows us to utilize the quantum confinement effect for both opening their bandgap and for adjustment of their energy levels. Furthermore, investigations have revealed that the separation and migration of photogenerated charge carriers are strongly affected by the crystal and structural features of the materials (e.g. crystallinity) and a series of surface properties including particle size, surface area, surface structure, active reaction sites, etc.¹³ By tuning these parameters, the properties of a semiconductor material such as the bandgap and the CB and VB positions can, in principle, be tailored to enhance its activity and stability in photocatalytic H₂ production applications.^32–42

3.1.1 Modifying the crystal structure of CdX. The crystal structure of the semiconductor plays an essential role in determining its photocatalytic properties. It is well recognized that good crystallinity can lead to significant improvement of the charge carrier separation and migration properties, whereas lattice defects in semiconductors usually operate as trapping and recombination centers for the photogenerated charge carriers, which is not conducive for the photocatalysis. Bao et al. reported a facile thermolysis procedure of a Cd–thiourea complex to synthesize phase-controlled CdS nanocrystals for photocatalytic H₂ production.³² This study showed that the photocatalytic activity for H₂ production over CdS had a direct relationship with phase structure and phase composition, and the pure hexagonal CdS had the highest photocatalytic activity among the different phases of CdS photocatalysts with Pt as the cocatalyst. Li and co-workers demonstrated that the photocatalytic activity of the Pt/CdS system for H₂ evolution can be notably enhanced by suitable photoetching.³³ They clarified that the selective removal of grain boundary defects, which acted as charge recombination centers for photogenerated charge carriers, can decrease the recombination probability of the charge, thus increasing the photocatalytic activity. In a recent example, Wang et al. reported the fabrication of facet-engineered CdS nanocrystals by controlling the synthesis kinetics.³⁴ Growth rate control of the {0001} facets (r₁) and {101̄1} facets (r_1′) of CdS nanocrystals was achieved by simply employing a syringe pump, which enables fine tuning of the crystal shape from nanocones, to nanofrustums, and further to nanoplates (Fig. 2a). Study of their photocatalytic activities revealed that the CdS nanoplates, with the largest {0001} facets, exhibited the highest photocatalytic activity for H₂ production in comparison to the nanocones and nanofrustums (Fig. 2b).

Fig. 2 (a) Schematic illustration of the formation of a CdS nanocone, a CdS nanofrustum, and a CdS nanoplate, simply controlled by adjusting Cd²⁺ injection rate. (b) Average photocatalytic H₂ production rate of the various CdS nanostructures. Inset in (b) shows corresponding specific activities that were normalized with respect to their calculated surface areas. Reproduced with permission from ref. 34. Copyright 2015 American Chemical Society.

3.1.2 Modifying the size and morphology of CdX. Besides the crystal structure, size and morphology are also important surface properties affecting the photocatalytic properties of CdX materials. For example, Zhu and co-workers synthesized various CdS nanostructures with controlled morphologies (nanorods and nanospheres) and compared their photocatalytic activity for H₂ production.³⁵ They found that the one-dimensional CdS nanorods (∼10 nm in diameter and ∼200 nm in length), which could be prepared by using diethylenetriamine (DETA) as a template and coordination agent, exhibited higher photocatalytic H₂ generation activity than the other CdS nanostructures. They attributed the favorable photocatalytic performance of these CdS nanorods to their anisotropic characteristic and good dispersion, which led to excellent separation ability of photogenerated charge carriers. In another similar example, Osterloh et al. have demonstrated that CdSe nanoribbons show enhanced photocatalytic activities toward H₂ production when compared to bulk CdSe counterparts.³⁶ Recently, we synthesized ultrathin CdS nanosheets and studied their photocatalyic activity toward H₂ generation.³⁷ In this case, the ultrathin CdS nanosheets with a thickness of ∼4 nm were prepared through an ultrasonication-induced aqueous exfoliation approach using CdS–DETA hybrid nanosheets as starting materials (Fig. 3a). Fig. 3b shows the transmission electron microscopy (TEM) image of the ultrathin CdS nanosheets. Study of their photocatalytic activity demonstrated that these ultrathin CdS nanosheets exhibited a much higher H₂ production rate than CdS nanoparticles and CdS aggregates, as well as the pristine CdS–DETA hybrid nanosheets. The enhancement in photocatalytic activity could be ascribed to their unique ultrathin morphology and single-crystal-like structural feature which can not only provide a huge surface area and more active sites, but also lead to the shift of the CB to a higher position and facilitate the separation of the photogenerated charge carriers. Very recently, our group described the synthesis of 3D hierarchical ultrathin-branched CdS nanowire arrays (3DHU-CdS) and their application for photocatalytic H₂ evolution.³⁸ In this study, the 3DHU-CdS with an adjustable branch size could be synthesized by means of a facile chemical transformation strategy using 2D ZnS–amine inorganic–organic hybrid nanosheets as the precursor (Fig. 3c and d). We found that the obtained 3DHU-CdS exhibited enhanced photocatalytic activity for visible-light-driven H₂ generation from water due to their unique hierarchical and ultrathin structural features. Peng et al. described flower-like CdSe architectures composed of ultrathin nanosheets for enhanced visible-light-driven photocatalytic H₂ production.³⁹

Fig. 3 (a) Schematic illustration of the preparation of L-cysteine-stabilized ultrathin CdS nanosheets. (b) TEM image of ultrathin CdS nanosheet and the corresponding colloidal CdS dispersion displaying the Tyndall effect. (c) Schematic illustration of the preparation of 3D hierarchical ultrathin-branched CdS nanowire arrays. (d) TEM image of the 3D hierarchical ultrathin-branched CdS nanowire arrays. (a, b) Reproduced with permission from ref. 37. Copyright 2013, The Royal Society of Chemistry. (c, d) Reproduced with permission from ref. 38. Copyright 2015, The Royal Society of Chemistry.

Several research groups have investigated the size-dependent performance of CdX materials in the photocatalytic H₂ evolution. The work by Sathish et al. showed that CdS particles on the nanometer scale could display a higher photocatalytic activity for H₂ generation compared to bulk CdS counterparts.⁴⁰ In a recent example, Osterloh and co-workers provided the first quantitative analysis of quantum-size-controlled photocatalytic H₂ evolution at semiconductor (CdSe)–solution interfaces.⁴¹ Based on their results, it was found that the H₂ evolution rate from illuminated suspended CdSe quantum dots (QDs) in aqueous Na₂SO₃ solution was dependent on the size of the CdSe nanocrystals. The study by Grigioni et al. also demonstrated the size-dependent performance of CdSe QDs in the photocatalytic evolution of H₂ under visible light irradiation.⁴² They found that 2.8 nm-sized CdSe samples exhibited the highest H₂ production rate due to their good visible light harvesting capability, suitable conduction band position and low probability of photogenerated charge carrier recombination.

It must be realized that despite numerous evidences demonstrating the feasibility of photocatalytic water splitting using pure CdX semiconductor nanomaterials, their H₂ evolution rates and the quantum efficiencies (QEs) are still too low to be able to efficiently utilize solar energy, even in the presence of sacrificial agents. Therefore, many research efforts have been directed to improve their photocatalytic properties by modifying CdX with other materials through a series of approaches to further modify their activity and stability.

3.2 Forming CdX-based solid solutions

Forming a solid solution is an efficient composition-engineering strategy to modify the photocatalytic properties of CdX semiconductor materials. Compared to the pristine CdX materials, solid solutions possess some unique advantages, such as tunable absorption properties in the visible region of the solar spectrum, controllable band structure and better electrical conductivity with the help of inserted metals (M), leading to an enhanced visible-light response and improved photoactivity.

The Cd_xZn_1−xS (0 < x < 1) solid solution, formed by combining semiconductor ZnS with a wide bandgap and semiconductor CdS with a narrow bandgap, has attracted much research interest for visible-light-driven photocatalysis applications.^43–51 For example, our group have synthesized nanoporous single-crystal-like Cd_xZn_1−xS nanosheets with tunable pore size and composition by the cation-exchange reaction of inorganic–organic hybrid ZnS–DETA nanosheets with Cd²⁺ ions and demonstrated their high photocatalytic activity for H₂ production.⁴³Fig. 4a displays the typical scanning electron microscopy (SEM) image of the porous Cd_0.5Zn_0.5S nanosheets. We demonstrated that the porous Cd_0.5Zn_0.5S nanosheets showed a stable H₂ production rate of about 0.5 mmol h⁻¹ per 0.3 g catalyst, which is about 2.5 times than that of Cd_0.5Zn_0.5S nanorods (Fig. 4b). In another example, Guo, Yao and co-workers have synthesized a Cd_xZn_1−xS solid solution with nano-twin structures through a precipitate-hydrothermal method and evaluated their photocatalytic activity for H₂ evolution.⁴⁴ Their study results revealed that such Cd_xZn_1−xS nanocrystal photocatalysts showed a higher photocatalytic H₂ production activity compared to the pristine CdS counterpart. Among them, Cd_0.5Zn_0.5S photocatalysts showed the highest activity with a H₂ evolution rate of 1.79 mmol h⁻¹ per 0.1 g catalyst and an apparent QE of 43% at 425 nm.

Fig. 4 (a) SEM image of porous Cd_0.5Zn_0.5S nanosheets. (b) Time course of evolved H₂ under visible-light irradiation of Cd_0.5Zn_0.5S porous nanosheets (★) and Cd_0.5Zn_0.5S nanorods (●). (c) TEM image of hollow Cd_0.33Zn_0.67Se nanoframes. (d) Time course of evolved H₂ under visible-light irradiation of hollow Cd_0.33Zn_0.67Se nanoframes (●), hollow Cd_0.25Zn_0.75Se nanoframes (■), and Cd_0.33Zn_0.67Se nanoparticles (▲). (a, b) Reproduced with permission from ref. 43. Copyright 2012, John Wiley and Sons. (c, d) Reproduced with permission from ref. 56. Copyright 2012, John Wiley and Sons.

Investigations demonstrate that the photocatalytic performance of the Cd_xZn_1−xS solid solution can be further improved by doping with some transition metal ions, such as Cu,^45,46 Ni,⁴⁷ Sn,⁴⁸ and La.⁴⁹ For example, Liu et al. have successfully synthesized a Cu doped Cd_0.1Zn_0.9S photocatalyst by a coprecipitation method, and an apparent QE of about 9.6% can be achieved by the Cd_0.1Cu_0.01Zn_0.89S photocatalyst without any cocatalyst at 420 nm.⁴⁵ Chen and co-workers reported an efficient strategy for the enhancement of photocatalytic H₂ evolution in which La was successfully doped into the depletion layer of the Cd_0.6Zn_0.4S photocatalyst via a facial solvothermal method.⁴⁹ Apart from metal doping, non-metal doping is another effective approach to modify the photocatalytic properties of semiconductors. One successful case is doping various nonmetal anions (such as C, N, S, etc.) onto TiO₂ to modify its optical and photocatalytic properties,^52,53 which should open up interesting possibilities to modify the photocatalytic properties of CdX semiconductors by appropriate nonmetal ion doping.

Except for Cd_xZn_1−xS, other CdX-based solid solutions, such as Cd_xMn_1−xS,⁵⁴ CdIn₂S₄,⁵⁵ and Cd_xZn_1−xSe,⁵⁶ are also found to possess efficient photocatalytic activity. For example, Ikeue and co-workers have successfully fabricated a Cd_xMn_1−xS photocatalyst through a hydrothermal method.⁵⁴ They found that the high photocatalytic activity for H₂ evolution of this Cd_xMn_1−xS photocatalyst was caused by its low-crystalline character. Moreover, they demonstrated that this Cd_xMn_1−xS photocatalyst showed better tolerance to photocorrosion compared to its pure CdS counterpart. The study by Kale et al. showed efficient photocatalytic H₂ production over nanostructured CdIn₂S₄.⁵⁵ As an example of the study on Cd_xZn_1−xSe solid solutions, our group synthesized hollow Cd_xZn_1−xSe nanoframes via a selective cation-exchange strategy and demonstrated their composition-dependent photocatalytic activity for H₂ production.⁵⁶Fig. 4c shows the typical TEM image of the hollow Cd_0.33Zn_0.67Se nanoframes. It was found that the Cd_0.33Zn_0.67Se nanoframes showed an obviously higher H₂ production rate than those of Cd_0.25Zn_0.75Se nanoframes and Cd_0.33Zn_0.67Se nanoparticles (see Fig. 4d).

Apart from the component (the choice of inserted metals (M)), the composition (the ratio of M to Cd) of the solid solution can also influence the band energy level and thus the photocatalyic properties.^57,58 For example, Yu and Gong et al. have revealed the bandgap effect of composition-tunable Cd_xZn_1−xS solid solutions on their superior photocatalytic H₂ evolution.⁵⁷ They found that the Zn/Cd molar ratio in these Cd_xZn_1−xS solid solutions could greatly affect the band structure and positions and the final photocatalytic activity of the resultant solid solution photocatalysts. Among various Cd_xZn_1−xS solid solutions, the Zn_0.5Cd_0.5S sample presented the highest H₂-production rate (7.42 mmol h⁻¹ g⁻¹), which was about 24 and 54 times higher than those of the pure CdS and ZnS samples, respectively. The activity of Zn_0.5Cd_0.5S was even higher than that of the optimal Pt-loaded CdS sample (see Fig. 5).

Fig. 5 Comparison of the photocatalytic activities of Zn_1−xCd_xS samples with different compositions and a Pt (0.9 wt%)/CdS sample for the photocatalytic H₂-production from 0.44 M Na₂S and 0.31 M Na₂SO₃ mixed aqueous solution under visible-light irradiation. The inset is the conduction and valence band edge potentials of Zn_1−xCd_xS samples with different compositions. Reproduced with permission from ref. 57. Copyright 2013 American Chemical Society.

3.3 Loading cocatalysts onto CdX

As discussed above, loading appropriate cocatalysts on the semiconductor photocatalyst can effectively promote or accelerate the surface chemical reactions.^17,18,30 Generally, the cocatalysts for semiconductor-based photocatalytic water splitting can be divided into two categories according to their different functions, that is, H₂-evolution cocatalysts and O₂-evolution cocatalysts. In this section, we focus on the development of the combination of H₂-evolution cocatalysts with CdX-based photocatalysts for photocatalytic H₂ generation.

3.3.1 Metal cocatalysts. Noble metals, such as Pt,^{32,33,40,59–78} Au,^63,74 Ru,^76,79 Pd,^74,80 and Rh,^81,82 have been widely used as efficient H₂-evolution cocatalysts in combination with CdX semiconductor photocatalysts by forming metal/CdX heterogeneous photocatalysts, with some representative examples summarized in Table 1. Among them, Pt, with its large work function and low overpotential for H₂ evolution reaction, is one of the most effective and most investigated H₂-evolution cocatalysts. Excellent H₂ production rates and QEs from various Pt/CdX heterogeneous photocatalysts have been reported by many research groups. In a typical example, Bao et al. demonstrated that the loading of monodisperse Pt nanoparticles (3–5 nm) onto porous CdS nanostructures could remarkably improve their photocatalytic activity for H₂ evolution.⁵⁹ Study of the photocatalytic performance showed that the pure nanoporous CdS nanostructures exhibited a H₂ evolution rate of 0.13 mmol h⁻¹ g⁻¹ from an aqueous solution containing Na₂S and Na₂SO₃ as sacrificial reagents under visible-light irradiation, and the H₂ evolution rate increased to 3.67 mmol h⁻¹ g⁻¹ over (0.5%) Pt/CdS and further to 27.3 mmol h⁻¹ g⁻¹ over (13%) Pt/CdS, with an apparent QE reaching ∼60% at 420 nm. Amirav et al. designed and synthesized a three-component, well-defined nanoheterostructure composed of a Pt-tipped CdS rod with an embedded CdSe seed.⁶⁰ In this photocatalyst system, the photogenerated holes were three-dimensionally confined to the CdSe, whereas the photogenerated electrons were transferred to the Pt tip. The ultimate result was the efficient separation of photogenerated electrons and holes, the decrease of the charge recombination probability, and the suppression of photocorrosion of CdS. They found that the activity for H₂ production could be significantly increased compared to that of unseeded nanorods by tuning the length of the nanorod heterostructure and the seed size. Study of their photocatalytic activity revealed that the three-component photocatalytic system was not only highly active for H₂ evolution with an apparent QE of 20% at 450 nm, but also active under orange light illumination, and demonstrated improved stability compared to Pt-tipped CdS nanorods without a CdSe seed. The Li group have presented a successful example by co-loading of Pt as the reduction cocatalyst and PdS as the oxidation cocatalyst on the CdS photocatalyst.⁶¹ The optimal Pt(0.3 wt%)–PdS(0.13 wt%)/CdS photocatalyst system could achieve an improved QE of 93% for H₂ production in the presence of Na₂S and Na₂SO₃ as sacrificial agents under visible-light irradiation, and also exhibited excellent stability during the long-term photocatalytic reaction. This QE was close to that of the primary courses of natural photosynthesis. The detailed study indicated that the cocatalysts can effectively prohibit the recombination of electrons and holes and promote the efficient utilization of charges at the shallow trap states of CdS. In a recent important advance, Yu, Jiang and co-workers described a ternary ZnS–(CdS/M) (M = Au, Pd, Pt) nano-archittecture for highly efficient photocatalytic H₂ production.⁷⁴ The ternary ZnS–(CdS/M) hybrids were prepared by the post-synthetic modification of binary multi-node sheath ZnS–CdS heteronanorods, which were transformed from single component ZnS nanorods via sequential cation exchange. In this nano-architecture, the delivery of photogenerated electrons from the CdS node sheath not only to the metal surface but also to the exposed ZnS stem occurred, promoting the separation of electron and hole carriers, thereby leading to the performance improvement of photocatalytic H₂ evolution (see Fig. 6). Tongying et al. reported CdSe/CdS/Pt double heterojunction nanowires as an efficient photocatalyst system for H₂ generation.⁷⁵

Fig. 6 (a, b) HAADF-STEM images of the ZnS–(CdS/Au) hetero-nanorods. (c) Schematic representation of the photocatalytic generation of H₂ from the hetero-nanorods. Reproduced with permission from ref. 74. Copyright 2015, John Wiley and Sons.

Table 1 CdX-based photocatalysts containing inorganic cocatalysts for photocatalytic H₂ production

Photocatalyst	Cocatalyst	Light source^a	Aqueous reaction solution^b	H2 evolution			Ref.
				Activity (μmol h^−1 g^−1)	QE (%)	Enhancement factor^c
a Xe: xenon lamp, Hg: mercury lamp. b TEOA: triethanolamine, AA: ascorbic acid. c Enhanced factor is calculated from the activity enhancement of photocatalysts loaded with the optimal amount of cocatalysts over photocatalysts without the loading of cocatalysts.
Nanoporous CdS nanostructures	13 wt% Pt	λ ≥ 420 nm (Xe)	0.35 M Na2SO3 + 0.25 M Na2S	27 333	60.34 (420 nm)	205	59
CdSe-seeded CdS nanorods	Pt	300 W Xe	Isopropanol	40 000	20 (450 nm)		60
CdS	0.3 wt% Pt + 0.13 wt% PdS	λ > 420 nm (Xe)	0.5 M Na2SO3 + 0.5 M Na2S	29 233	93 (420 nm)	381	61
CdS	1 wt% Pt	λ > 420 nm (Xe)	0.1 M Na2SO3 + 0.1 M Na2S	13 800	22 (420 nm)		62
CdSe nanorods	Pt	λ ≥ 420 nm (Xe)	0.35 M Na2SO3 + 0.25 M Na2S	36 250	0.63 (454 nm)		63
Multi-armed CdS nanorods	0.23 wt% Pt	λ ≥ 420 nm (Xe)	10 v% lactic acid	24 200	51 (420 nm)	60.5	64
Single crystal CdS nanowires	0.3 wt% Pt	λ ≥ 420 nm (Xe)	10 v% lactic acid	18 625	61.7 (420 nm)	74.5	65
Porous CdS nanosheet-based flowers	0.5 wt% Pt	λ > 420 nm (Xe)	10 v% lactic acid	9374	24.7 (420 nm)		66
Mixed-phase CdS	0.3 wt% Pt	λ > 400 nm (Hg–Xe)	0.02 M Na2SO3 + 0.1 M Na2S	668		2	67
CdS	1 wt% Pt	UV	0.35 M Na2SO3 + 0.24 M Na2S	6000		8.82	68
CdS	1 wt% Pt	Overall spectrum	0.3 M Na2SO3 + 0.2 M Na2S	13 860	1.3 (overall spectrum)	3.3	69
CdS nanorods	0.05 wt% Pt	λ > 420 nm (Hg)	Formic acid	4603.2	13.9 (>420 nm)	21	70
CdS	1.5 wt% Pt	λ > 420 nm (Hg)	Formic acid	1127.5		16.3	33
CdS nanowires	1 wt% Pt	λ ≥ 420 nm (Hg)	0.1 M Na2S + 0.02 M Na2SO3	∼60			71
CdS	3 wt% Pt	λ ≥ 420 nm (Hg)	0.25 M Na2S + 0.35 M Na2SO3	83.3			32
CdS	2 wt% Pt	λ > 430 nm (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	5625	24.1 (420 nm)	5	72
CdS nanowire arrays	10 wt% Pt	λ ≥ 420 nm (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	2600	7.2 (420 nm)		73
CdS	1 wt% Pt	UV	0.35 M Na2SO3 + 0.24 M Na2S	14 150		19.4	40
ZnS–CdS heteronanorods	Pt	300 W Xe	0.25 M Na2SO3 + 0.35 M Na2S	5050		7.5	74
CdSe/CdS core/shell nanowires	∼30 wt% Pt	450 W Xe	0.1 M Na2SO3 + 0.1 M Na2S	434.29		5.4	75
CdS–ZnO–CdO	0.5 wt% Ru	150 W Xe	0.1 M Na2S + 0.04 M Na2SO3	∼75		50	76
CdS/Al–HMS	0.07 wt% Ru	λ ≥ 420 nm (Xe)	Formic acid	825.9	1.2 (420 nm)		79
CdS	4 wt% Ni	λ > 420 nm (Xe)	(NH4)2SO3	25 848	26.8 (420 nm)		83
CdS nanorods	Ni	447 nm (LED)	Ethanol	63 000	53 (447 nm)		84
CdS	5 wt% Ni	λ > 400 nm (Xe)	50 v% lactic acid	30 048		∼6.6	85
CdS–titanate	1.2 wt% Ni	λ ≥ 420 nm (Xe)	20 v% ethanol	11 038	21 (420 nm)	77.7	88
CdS	32 mol% NiO	500 W Phoenix tungsten halogen lamp	0.25 M Na2SO3 + 0.35 M Na2S	745	6.02		90
CdS	13.2 mol% Ni2O3	λ > 400 nm (Xe)	30 v% methanol	4456		4.1	91
CdS	1 mol% NiOx	λ > 400 nm (Xe)	30 v% methanol	5908	8.6 (400 nm)	117	92
TiO2/CdS	2.1 wt% CoOx	λ > 400 nm (Xe)	0.125 M Na2S + 0.175 M Na2SO3	660		7	93
CdS	3 mol% Co3O4	λ ≥ 420 nm (Xe)	0.5 M Na2S + 0.5 M Na2SO3	236		32.4	94
CdS	3.5 mol% MoO3	λ > nm (Xe)	0.2 M Na2S + 1.5 M Na2SO3	5250	28.86 (420 nm)	87.5	95
CdS	23 mol% Ni(OH)2	λ > 420 nm (Xe)	25 v% TEOA	5085	28 (420 nm)	145	96
CdS/RGO	1.0 wt% Ni(OH)2	λ > 420 nm (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	4731		10	97
Zn0.8Cd0.2S	0.6 mol% Ni(OH)2	λ ≥ 420 nm (Xe)	20 v% TEOA	7160	29.5 (420 nm)	25.5	98
CdS nanorods	6.8 mol% Co(OH)2	500 W Xe	25 v% methanol	61		41	99
CdS	0.2 wt% MoS2	λ > 420 nm (Xe)	10 v% lactic acid	∼5400		36	101
CdSe	0.5 wt% MoS2	λ ≥ 420 nm (Xe)	0.1 M Na2SO3 + 0.1 M Na2S	890		3.7	102
CdS	0.4 wt% RGO + 2 wt% MoS2	500 W UV lamp	10 v% lactic acid	6857		71	103
Zn0.2Cd0.8S	3 wt% MoS2	λ > 420 nm (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	420		210	104
Zn0.3Cd0.7S	0.6 wt% MoS2	λ > 420 nm (Xe)	10 v% lactic acid	1196		7	106
CdS	15 wt% MoS2	λ > 420 nm (Xe)	0.02 M Na2SO3 + 0.1 M Na2S	4770		10	107
CdS	∼11 mol% WS2	λ > 420 nm (Xe)	10 v% lactic acid	1984		16	109
	∼7 mol% MoS2	λ > 420 nm (Xe)	10 v% lactic acid	1472		12	109
CdS	1.0 wt% WS2	λ > 420 nm (Xe)	10 v% lactic acid	∼4200	5.0 (420 nm)	28	110
CdSe/CdS	MoS3	450 nm (Hg–Xe)	TEOA	100 000	10 (450 nm)		111
CdS	1.2 mol% NiS	λ ≥ 420 nm (Xe)	30 vol% lactic acid	7267	51.3 (420 nm)	34	112
CdS nanorods	5 mol% NiS	λ > 420 nm (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	1131	6.1 (420 nm)	20.6	113
Zn0.5 Cd0.5 S/RGO	3 mol% NiS	AM1.5	0.25 M Na2SO3 + 0.35 M Na2S	7514	31.3 (420 nm)	1.8	114
Cd0.5Zn0.5S	0.025 mol% NiS	λ ≥ 430 nm (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	1400	33.9 (425 nm)	3.5	115
CdLa2S4	2 wt% NiS2	λ ≥ 420 nm (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	2500	1.6 (420 nm)	3.1	118
CdS nanorods	3 mol% CuS	500 W (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	332		3.5	119
Zn0.5Cd0.5S	CuS	λ ≥ 420 nm (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	4638.5	20.9 (420 nm)	7	120
Zn0.8Cd0.2S	3 wt% CuS	500 W (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	2792	36.7 (420 nm)	4.4	121
CdS	20 mol% SrS	λ ≥ 430 nm (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	615	2.85 (420 nm)	2	122
CdS	∼41.6 mol% TiS2	λ > 399 nm (Xe)	3.5 v% benzyl alcohol	1000		2.9	123
CdS	∼41.6 mol%TaS2	λ > 399 nm (Xe)	3.5 v% benzyl alcohol	2320		6.9	123
CdS nanorods	0.5 wt% Ni2P	λ > 420 nm (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	1 200 000	41 (450 nm)	22	125
CdS nanorods	16.7 wt% MoP	λ > 420 nm (Xe)	10 v% lactic acid	163 200	5.8 (450 nm)	9.2	126
CdS nanorods	0.44 wt% Cu3P	λ > 420 nm (Xe)	1.25 M Na2S + 1.75 M Na2SO3	200 000	25 (450 nm)	6.6	127
CdS nanorods	30 wt% Fe2P	λ > 420 nm (Xe)	0.5 M AA	186 000	15 (450 nm)	31	128
CdS	10 wt% WC	λ ≥ 420 nm (Hg-arc)	0.1 M Na2S + 0.02 M Na2SO3	1400		23	130

---

For the cocatalyst loading onto the CdX semiconductors, a series of factors, such as the loading conditions, geometrical configuration of the semiconductors, and the size and composition of the cocatalysts, can greatly influence the final effect of the composite catalysts. For example, Xu and co-workers compared the photocatalytic activities of various Pt/CdS photocatalyst systems obtained from different Pt reduction environments for H₂ production and investigated the effect of loading conditions for Pt nanoparticles on the surface of the CdS photocatalyst.⁶² In this study, Pt was deposited on the CdS photocatalyst by means of a photoreduction method. They found that the solution environment during Pt photoreduction played an essential role on the activity of the final Pt/CdS photocatalyst system. Study of the photocatalytic activity for H₂ production from various Pt/CdS photocatalyst systems revealed that the sample with Pt reduced in NaOH alkaline solution exhibited a markedly higher H₂ evolution activity (13 mmol h⁻¹ g⁻¹) than that of samples obtained in acidic or neutral solution (<0.5 mmol h⁻¹ g⁻¹) in the presence of Na₂S and Na₂SO₃ as sacrificial reagents. Song et al. investigated geometrical (single- or double-tipped) and compositional (Pt or Au) variations of active metal components in a well-defined CdSe nanorod system.⁶³ In this case, single Pt-tipped, single Au-tipped, double Pt-tipped, and double Au-tipped CdSe nanorods could be fabricated by controlling the component and the concentration of Pt (or Au) precursors. Study of their photocatalytic activities revealed that these catalysts exhibited significant dependency of the catalytic activity on the catalyst geometry and the choice of the noble metal tips. They found that the Pt-tipped catalysts showed higher photocatalytic activity than Au-tipped catalysts. Moreover, the H₂ evolution amount of the single Pt-tipped CdSe nanorods was about ∼50% larger than that of the double Pt-tipped structure, although the loading amount of Pt in the former was only half of that in the latter (see Fig. 7).

Fig. 7 (a, b) Typical TEM images of single (a) and double (b) Pt-tipped CdSe nanorods. (c) Time course of H₂ evolution by various photocatalysts: single Pt-tipped CdSe (◆), double Pt-tipped CdSe (▲), double Au-tipped CdSe (●), and single-Au tipped CdSe (■). Reproduced with permission from ref. 63. Copyright 2012 American Chemical Society.

Recently, several earth-abundant transition metal H₂-evolution cocatalysts, such as Ni^83–88 and Co,⁸⁹ have been used as substitutes for noble metals to enhance the activity of CdX photocatalysts. For example, Wang et al. reported enhanced activity for photocatalytic H₂ generation in Ni/CdS nanocomposite photocatalysts.⁸³ In their study, Ni nanoparticles with a high degree of crystallization and an average size of about 10 nm were prepared via chemical reduction of a NiCl₂ precursor by hydrazine in an aqueous solution followed by loading on CdS nanorods by photo-induced deposition. They found that the loading of Ni nanoparticles could increase the specific surface area of CdS nanorods. Upon visible light irradiation, the Ni/CdS photocatalyst with a Ni content of 4 wt% could provide a H₂ evolution rate of 25.8 mmol h⁻¹ g⁻¹ from aqueous (NH₄)₂SO₃ solution with a QE of 26.8% at 420 nm. Moreover, the Ni/CdS nanocomposite photocatalyst could retain high stability and activity even after 20 h of photocatalytic reaction. Feldmann and co-workers also demonstrated efficient photocatalytic H₂ production on Ni-decorated CdS nanorods.⁸⁴ In this study, the Ni nanoparticles (2–8 nm) were loaded on the surface of the cysteine stabilized CdS nanorods through in situ photodeposition. They employed a hydroxyl anion/radical redox couple (˙OH/⁻OH), which possess good mobility and can react with CdS and the sacrificial agent (ethanol) to assist in the efficient transfer of the photogenerated holes from the CdS to ethanol (see Fig. 8). As a result, a significant enhancement in the H₂ generation rate could be achieved, with an apparent QE of 53% at 447 nm and a remarkable H₂ formation rate of 63 mmol g⁻¹ h⁻¹ under optimal conditions. Moreover, the obtained photocatalyst exhibited an excellent long-term stability (over 220 h) for photocatalytic H₂ production. The work by Chen et al. showed that metallic Ni can be deposited on CdS by a hydrothermal reduction method with ethylene glycol.⁸⁵ The optimal Ni (5 wt%)/CdS photocatalyst gave a H₂ evolution rate of around 3 mmol h⁻¹, which was even higher than that of the Pt(0.5 wt%)/CdS system under the same reaction conditions. More recently, Zhukovskyi et al. demonstrated efficient photocatalytic H₂ generation from Ni nanoparticle-decorated CdS nanosheets.⁸⁶ As an example of loading a Co cocatalyst onto CdX, Wu and co-workers demonstrated a robust catalyst system formed in situ from CdTe QDs and CoCl₂·6H₂O in aqueous ascorbic acid solution for photocatalytic H₂ generation.⁸⁹ In this case, the Co²⁺ ions, which were directly bonded to the surface of the CdTe QDs, could be photoreduced partially or completely under visible-light irradiation to form a hybrid Co_h–CdTe catalyst system, which could achieve a high H₂ evolution rate of 25 μmol h⁻¹ mg⁻¹ with a turnover number (TON) of 86 250 with respect to the CdTe QDs with 21 h irradiation.

Fig. 8 (a) Schematic of the photocatalytic generation of H₂ from Ni-decorated CdS nanorods under the proposed hole shuttle mechanism. (b) HAADF-STEM image of the Ni-decorated CdS nanorods. Reproduced with permission from ref. 84. Copyright 2014, Nature Publishing Group.

3.3.2 Transition metal oxides and hydroxides. Some earth-abundant transition metal oxides, such as NiO,⁹⁰ Ni₂O₃,⁹¹ NiO_x,⁹² CoO_x,⁹³ Co₃O₄,⁹⁴ and MoO₃,⁹⁵ are also well-known as highly active cocatalysts after being integrated with CdX materials for photocatalytic H₂ production, as summarized in Table 1.

For example, Chen et al. showed that in situ photodeposition of NiO_x on CdS could enhance its photocatalytic H₂ generation activity. They found that the environment of deposition could affect the photocatalytic activity of the resultant NiO_x–CdS photocatalyst.⁹² The NiO_x–CdS photocatalyst formed in alkaline solution showed the highest activity of photocatalytic H₂ production due to the high dispersion and tight deposition of NiO_x on CdS, and better hole transfer under alkaline conditions, reaching 590.8 μmol h⁻¹, which was 117 times higher than that of pure CdS. Yan et al. prepared a CoO_x-loaded TiO₂/CdS semiconductor composite by a simple solvothermal method.⁹³ Study of their photocatalytic behaviour revealed that CoO_x could act as an efficient cocatalyst to enhance the photocatalytic activity of TiO₂/CdS and its content exhibited a significant influence on the activity. The optimal CoO_x loading content was determined to be 2.1 wt%, and the average rate of H₂ evolution was 660 μmol g⁻¹ h⁻¹ in the presence of Na₂S and Na₂SO₃ as sacrificial agents, which is about 7 times higher than those of the TiO₂/CdS composite and CdS. The work by Yuan et al. showed that Co₃O₄ can serve as an efficient H₂-evolution cocatalyst in combination with CdS semiconductor.⁹⁴

Several metal hydroxides^96–99 and layered double hydroxides (LDH)¹⁰⁰ have also been developed as efficient and noble metal-free cocatalysts for photocatalytic H₂ production. Jaroniec, Yu and co-workers found highly efficient photocatalytic H₂ production on a Ni(OH)₂-loaded CdS photocatalyst. The optimal Ni(OH)₂ (0.23 mol%)/CdS photocatalyst could achieve a visible-light H₂-production rate as high as 5085 μmol h⁻¹ g⁻¹ and a QE of 28% at 420 nm.⁹⁶ The H₂-production rate on the Ni(OH)₂-loaded CdS photocatalyst was 145 and 1.3 times higher than those of pure CdS and Pt (1 wt%)/CdS photocatalyst, respectively. Yan et al. demonstrated the use of Ni(OH)₂ as an efficient cocatalyst on a CdS/reduced graphene oxide (CdS/RGO) composite photocatalyst for visible light-driven H₂ production.⁹⁷ The optimal Ni(OH)₂ (1.0 wt%)/CdS/RGO photocatalyst could give a H₂-production rate of 4731 μmol h⁻¹ g⁻¹, which is nearly 10 times higher than that of the CdS/RGO photocatalyst under the same conditions. The work by Zhang et al. showed that Co(OH)₂-modified CdS nanorods exhibited a visible-light H₂-production rate of 61 μmol h⁻¹ g⁻¹, which exceeded that of pure CdS by a factor of 41 (see Fig. 9).⁹⁹ The outstanding H₂-evolution activity of Co(OH)₂ and the interface formed between the Co(OH)₂ and CdS, which facilitates the electron transfer from CdS to Co(OH)₂, were thought to be the main reasons for the enhanced photocatalytic activity. Hwang and co-workers prepared strongly coupled Zn–Cr–LDH–CdS nanohybrids through the electrostatically derived self-assembly of cationic Zn–Cr–LDH nanosheets and cationic CdS QDs with anionic linkers.¹⁰⁰ They found that the hybridization of CdS with Zn–Cr–LDH leads to significant enhancement of the photocatlaytic activity of CdS for visible-light-induced H₂ generation.

Fig. 9 (a) TEM image and (b) HRTEM image of the Co(OH)₂-modified CdS nanorod. Reproduced with permission from ref. 99. Copyright 2014, American Chemical Society.

3.3.3 Transition metal sulphides, carbides and phosphides. To date, many transition metal sulphides,^101–123 phosphides^124–129 and carbides¹³⁰ have been employed as efficient H₂-evolution cocatalysts to enhance the photocatalytic performance of CdX photocatalysts, as summarized in Table 1.

MoS₂ is a well-known cocatalyst for photocatalytic H₂ production, which has been widely investigated on a variety of semiconductor photocatalysts. Li and co-workers studied the role of MoS₂ as a cocatalyst loaded on CdS for photocatalytic H₂ production.¹⁰¹ In this case, the MoS₂/CdS photocatalyst system with highly dispersed MoS₂ on CdS was prepared by impregnating CdS with an aqueous solution of (NH₄)₂MoS₄, followed by treatment in a H₂S flow at high temperature. This study showed that the activity of CdS can be enormously increased by loading MoS₂ as a cocatalyst, and the activity of the MoS₂/CdS photocatalyst system could be even higher than that of Pt/CdS. In another recent example, Zhang and co-workers prepared MS₂–CdS (M = W or Mo) nanohybrids, with single-layer MS₂ nanosheets with a lateral size of 4–10 nm selectively grown on the Cd-rich (0001) surface of wurtzite CdS nanocrystals via a facile one-pot wet-chemical method, for photocatalytic H₂ production¹⁰⁹ (Fig. 10a). The photocatalytic performance of the WS₂–CdS and MoS₂–CdS nanohybrids towards H₂ production under visible-light irradiation (λ > 420 nm) was found to be about 16 and 12 times higher than that of pure CdS, respectively. The enhancement in activity was attributed to the large number of active sites of the single-layer MS₂ nanosheets and the inherent p/n heterjunction formed between MS₂ and CdS (Fig. 10b). The study by Zong et al. also demonstrated that the loading of WS₂ onto CdS could significantly enhance the evolution of H₂ from lactic acid solution under visible-light irradiation.¹¹⁰ Alivisatos and co-workers described the coating of amorphous MoS₃ on CdSe-seeded CdS nanorods by means of a facile one-step thermal process.¹¹¹ This study showed that the maximum H₂-evolution rate could reach 100 mmol h⁻¹ g⁻¹ with a QE of 10% at 450 nm.

Fig. 10 (a) HR-TEM image of MoS₂–CdS nanohybrids. (b) Schematic illustration of the photocatalytic process of MS₂–CdS nanohybrids in lactic acid solution. Reproduced with permission from ref. 109. Copyright 2015, John Wiley and Sons.

Besides MoS₂, MoS₃ and WS₂, some other transition metal sulphides, such as NiS,^112–117 NiS₂,¹¹⁸ CuS^119–121 and CoS,^112,117 can also serve as efficient H₂-evolution cocatalysts to boost the photocatalytic activity of CdX photocatalysts. By using a simple hydrothermal route, Xu and co-workers have prepared a NiS/CdS nanocomposite, with ultrafine NiS nanoparticles deposited on CdS nanoparticles, as a highly active photocatalyst for H₂ production.¹¹² The optimized NiS (1.2 mol%)/CdS photocatalyst could reach a visible-light-driven H₂-evolution rate of 7.27 mmol h⁻¹ g⁻¹ using lactic acid as a sacrificial agent with a QE of 51.3% at 420 nm, which is even higher than that of the Pt (1 wt%)/CdS photocatalyst. They proposed that the loading of the NiS cocatalyst could not only efficiently promote the separation and transport of photogenerated electrons from the CB of CdS to the NiS nanoparticles but could also accelerate the electrochemical adsorption and desorption kinetics for H₂ evolution through the following reactions (eqn (2) and (3)):

NiS + e⁻ + H⁺ →\; HNiS (2)

HNiS + e⁻ + H⁺ → NiS + H₂ (3)

The study by Xue and co-workers showed that the loading of NiS₂ co-catalyst resulted in remarkable enhancement for H₂ production over the CdLa₂S₄ photocatalyst under visible light irradiation.¹¹⁸ The optimal hybrid photocatalyst with 2 wt% NiS₂ loading exhibited a H₂ production rate of 2.5 mmol h⁻¹ g⁻¹, which was more than 3 times higher than that of the pristine CdLa₂S₄ photocatalyst. Zhang et al. prepared CuS/CdS nanoposites by a simple hydrothermal and cation exchange method and investigated their photocatalytic acitivity for H₂ produciton.¹¹⁹ They found that the loading of CuS as the cocatalyst could remarkably enhance the photocatalytic performance of CdS. The optimized CuS (3 mol%)/CdS displayed the highest photocatalytic activity, giving an H₂ evolution rate of 332 μmol g⁻¹ h⁻¹, exceeding that of pure CdS by 3.5 times.

Recently, several transition metal phosphides have been developed as electrocatalysts and found to give excellent H₂ evolution reaction activities. More recently, transition metal phosphides have been utilized as highly efficient H₂-evolution cocatalysts to enhance the photocatalytic acitivity of CdX semiconductors.^124–129 The work by Cao et al. has shown that the combination of CdS nanorods as the photocatalyst with water-soluble Ni₂P nanoparticles (5–8 nm) as the cocatalyst could create a highly efficient, stable, nobel-metal-free photocatalytic system for H₂ generation.¹²⁴ Under the optimal conditions, this resulted in a photocatalytic system that could achieve a TON and turnover frequency (TOF) of 26 300 and 2110 h⁻¹, respectively, based on the Ni₂P cocatalyst. Very recently, Du and co-workers have demonstrated an integrated nanostructure composed of CdS nanorods as the photocatalyst and Ni₂P nanoparticles as the cocatalyst for visible light-driven H₂ evolution.¹²⁵ In this case, crystalline Ni₂P nanoparticles were tightly anchored onto the one-dimensional CdS nanorods via an easy solvothermal method, forming a Ni₂P/CdS heterostructure (Fig. 11a and b). Under optimal conditions, the Ni₂P/CdS heterostructured photocatalyst could achieve a H₂ evolution rate of up to ∼1200 mmol h⁻¹ mg⁻¹, giving a TON of ∼3 270 000 and a TOF of 36 400 vs. Ni₂P in 90 h. The apparent QE was ∼41% at 450 nm. The exceptional activity of this photocatalytic system was probably due to fast charge separation between the Ni₂P and the CdS nanorods (Fig. 11c). Based on this study, they further employed other transition metal phosphides, such as MoP,¹²⁶ CuP,¹²⁷ and Fe₂P,¹²⁸ as H₂-evolution cocatalysts combined with CdS semiconductors for photocatalytic H₂ production.

Fig. 11 (a) HRTEM image and (b) high-magnification HRTEM image of Ni₂P/CdS heterostructures. (c) A schematic describing an artificial photocatalytic H₂ production system using CdS NRs as the photosensitizer and Ni₂P as the cocatalyst. Reproduced with permission from ref. 125. Copyright 2015, The Royal Society of Chemistry.

3.3.4 Hydrogenase and artificial molecular cocatalysts. Hydrogenases, one of the rare families of organometallic biomolecules, are unique H₂-activation catalysts as they can efficiently catalyze the reversible reduction of protons to H₂. In recent years, many investigations have indicated that some natural hydrogenases could be utilized as efficient cocatalysts to combine with CdX photocatalysts for visible-light driven H₂ production (Table 2). For example, Brow et al. have investigated the integration of clostridium acetobutylicum [FeFe]-hydrogenase I (CaI) with MPA-capped CdTe nanocrystals (MPA = 3-mercaptopropionic acid) for photocatalytic H₂ evolution under visible-light illumination.¹³¹ In this case, CaI could form stable photocatalytic complexes with MPA-capped CdTe through their molecular assembly with each other, which was mediated by electrostatic interactions, as shown in Fig. 12a. In these enzymatically active complexes, rapid intermolecular photogenerated electron transfer from CdTe to CaI, which acts as a H₂-evolution cocatalyst, could be achieved, leading to efficient H₂ generation. Under optimal conditions, the CdTe–CaI complexes could achieve a single wavelength QE of 9% (λ = 532 nm) and an efficiency of 1.8% under AM 1.5 white light irradiation in the presence of ascorbic acid.

Fig. 12 (a) Schematic illustration of the self-assembly of CaI and MPA capped CdTe nanocrystals, and the light-induced H₂ evolution of the self-assembled complex. (b) Schematic illustration of photocatalytic H₂ production by the CdS : CaI complex. (a) Reproduced with permission from ref. 131. Copyright 2010, American Chemical Society. (b) Reproduced with permission from ref. 132. Copyright 2012, American Chemical Society.

Table 2 CdX-based photocatalysts containing hydrogenase and molecular cocatalysts for photocatalytic H₂ production

Photocatalyst	Cocatalyst^a	Light source^b	Solvent	Sacrificial reagent^c	H2 evolution			Ref.
					TON vs. cocatalyst	TOF vs. cocatalyst	QE (%)
a CaI: clostridium acetobutylicum [FeFe]-hydrogenase I, PAA: poly(acrylic acid), PEI: polyethylenimone, DHLA: dihydrolipoic acid. b TH: tungsten halogen lamp, Hg: mercury lamp, Xe: xenon lamp. c TEOA: triethanolamine, AA: ascorbic acid, TEA: triethylamine.
CdTe	CaI	AM 3.0 (TH)	H2O	AA		25 s^−1	9 (532 nm)	131
CdS	CaI	405 nm (LED)	H2O	AA	∼1 000 000	380 s^−1	20 (405 nm)	132
		AM 1.5 (LED)	H2O	AA		983 s^−1
CdTe	1	λ > 400 nm (Hg)	H2O	AA	505	50 h^−1		133
CdTe	2 + chitosan	410 nm (LED)	CH2OH/H2O	AA	52 800	∼1.4 s^−1		134
CdSe	3	410 nm (LED)	H2O	AA	8781	596 h^−1		135
CdSe	4 + PAA	450 nm (LED)	H2O	AA	27 135	3.6 s^−1	5.07 (450 nm)	136
CdSe	5 + PEI	410 nm (LED)	H2O	AA	10 600			137
CdS	6	λ > 420 nm (Xe)	CH3CN/H2O	TEOA	171		9.1 (420 nm)	138
CdSe/ZnS	9	λ > 400 nm (Xe)	Toluene	TEOA	>151			139
CdTe	10	λ > 400 nm (Xe)	H2O	AA	650	9 min^−1		140
CdS	11	λ > 420 nm (Xe)	H2O	Na2SO3 + Na2S	64 700		29 (420 nm)	141
CdTe	12	λ > 400 nm (Xe)	H2O	AA	14 400	850 h^−1	5.32 (400 nm)	142
CdSe	Ni^2+–DHLA	520 nm (LED)	H2O	AA	>600 000	7000 h^−1	36 (520 nm)	144
CdS	13	λ > 420 nm (Xe)	C2H5OH/H2O	TEA	28 000	311 h^−1	11.5 (420 nm)	145
CdSe	21	520 nm (LED)	H2O	AA	280 000			146

---

Based on this early work, Brown et al. recently further employed CaI as a cocatalyst combined with colloidal CdS nanorods, which possess a larger absorption cross-section and surface area compared to CdTe nanocrystals, for photochemical H₂ generation from water.¹³² The resultant CdS nanorod–CaI photocatalytic complexes could display a TOF of 380–900 s⁻¹ (based on hydrogenase) and a QE of up to 20% under illumination at 405 nm using ascorbic acid as an electron donor (see Fig. 12b). The advances in these investigations not only demonstrate the feasibility of the integration of natural hydrogenase cocatalysts with CdX photocatalysts for a sustainable and economically viable H₂ production, but also provide some deeper insight into the rational design and preparation of other hybrid photocatalytic systems comprising natural hydrogenases and inorganic semiconductors.

Despite hydrogenases being highly active for H₂ evolution, the poor stability and the high cost of separation and purification of the enzymes from natural systems limit their large scale application. Relatively speaking, biomimetic molecular complexes capable of mimicking the active sites of hydrogenases and reproducing the enzymic activity seem to be more promising, as they are more tolerant to O₂ and the structures are tunable. In recent years, many hybrid photocatalytic systems with CdX nanocrystals as photocatalysts and biomimetallic molecular complexes as cocatalysts have been developed for photochemical H₂ generation. These hybrid systems were constructured either with molecular cocatalysts and CdX photocatalysts as two separate components or with molecular catalysts covalently linked to the surface of CdX photocatalysts. This section mainly focuses on the development of the hybrid photocatalytic systems comprising CdX photocatalysts and Fe-,^133–137 Co-,^138–143 Ni-^144–147 and Cu-based¹⁴⁸ biomimetic complexes as molecular cocatalysts, with some representative examples summarized in Table 2.

Since the structure of [FeFe]-hydrogenase ([FeFe]-H₂ase) was resolved in 1998, many dinuclear [FeFe]-H₂ase mimics have been synthesized and employed as H₂ activation catalysts for both electrochemical and photochemical H₂ evolution. The first example of the incorporation of a [FeFe]-H₂ase mimic cocatalyst with a CdX photocatalyst was reported in 2011 by Wu and coworkers.¹³³ In this study, a water soluble [FeFe]-H₂ase mimic (1, Fig. 13) containing a hydrophilic isocyanide ligand in combination with MPA-stabilized CdTe QDs and ascorbate formed an efficient photocatalytic system for H₂ production in fully aqueous solutions at pH 4. The TON and TOF reached up to 505 and 50 h⁻¹ (based on 1) after 10 h of illumination (λ > 400 nm). The researchers proposed that there existed a significant interaction between MPA–CdTe and 1 in the ground state and the oxidative quenching of the excited MPA–CdTe by 1 was the dominant pathway for photocatalytic H₂ evolution. Studies on a similar hybrid system with [Fe₂(CO)₆(μ-adt)CH₂C₆H₅] (2, Fig. 13) as the molecular cocatalyst showed that the activity and stability of the [FeFe]-H₂ase mimic were significantly improved by the addition of polysaccharide chitosan to mimic the protein matrix microenvironment surrounding the active site of [FeFe]-H₂ases.¹³⁴ The 2/MPA–CdTe QD/H₂A system displayed a remarkable TON of 52 800 for photochemical H₂ evolution in MeOH/H₂O at pH 4.5 in the presence of a large amount of chitosan (1.0 g L⁻¹) over 60 h of irradiation with 410 nm LEDs, affording an initial TOF of ∼1.4 s⁻¹ in the first 2 h. The TON of the H₂ evolution and the stability of the system were significantly improved in the presence of chitosan, which was attributed to the fact that (i) the encapsulation of the MPA–CdTe QDs by polycationic chitosan prevented the QDs from aggregating and enhanced the emission quantum efficiency of the QDs and (ii) the intimate interaction of 2 and chitosan rendered the water insoluble catalyst well-dispersed in aqueous solution and improved the stability of the reduced species of 2. These results demonstrated that not only the structure of the active center but also the biological environment surrounding the center in natural [FeFe]-H₂ases should be taken into account when designing and fabricating active H₂ evolution systems based on [FeFe]-H₂ase mimics.

Fig. 13 Structures of [FeFe]-H₂ase mimics 1–3 used as cocatalysts in hybrid systems with CdX nanocrystals. (1) Reproduced with permission from ref. 133. Copyright 2011, John Wiley and Sons. (2) Reproduced with permission from ref. 134. Copyright 2013, Nature Publishing Group. (3) Reproduced with permission from ref. 135. Copyright 2013, The Royal Society of Chemistry.

For most [FeFe]-H₂ase mimics, their poor solubility in water is a critical drawback which limits their application for coupling with water splitting technology. In order to solve this problem, Wu and co-workers developed an effective strategy for interface-directed assembly of a simple [FeFe]-H₂ase mimic, [Fe₂S₂(CO)₆] (3, Fig. 13), onto the surface of CdSe QDs, forming a water-soluble photocatalytic system for H₂ producton.¹³⁵ Under optimal conditions, the 3/MPA–CdSe assembly system was able to reach a TON of 6530 based on 3 in water at pH 4.0 over 82 h of illumination with λ > 400 nm. A higher TON of 8781 was obtained when the same hybrid system was irradiated with 410 nm LEDs under otherwise identical conditions, giving a TOF of 596 h⁻¹ in the first 4 h. The high efficiency and good durability of this system for photocatalytic H₂ production were ascribed to the close contact between the CdSe QDs and 3. In addition to direct anchoring of the [FeFe]-H₂ase mimic to the QDs, the other strategy used by Wu and co-workers was to coanchor the [FeFe]-H₂ase mimic and the QDs to the side chain of a functionalized polymer.¹³⁶ In this case, both [FeFe]-H₂ase mimic 4 (Fig. 14a) and MPA–CdSe QDs were linked to the abundant carboxy groups in poly(acrylic acid) (PAA), leading to a short distance between the light harvester and the catalyst. PAA could enhance the emission of the MPA–CdSe QDs and prevent the MPA–CdSe nanoparticles from aggregating. The PAA–4/MPA–CdSe/H₂A system in water at an initial pH of 4.0 displayed a TON of H₂ evolution of up to 27 135 over 8 h of irradiation with 450 nm LEDs, affording a TOF of 3.6 s⁻¹ in the first half hour and a QE of 5.07%. Similarly, [FeFe]-H₂ase mimic 5 (see Fig. 14b) was anchored on branched polyethylenimone (PEI) to generate a water-soluble [FeFe]-H₂ase mimic, which then was coupled with MPA-capped CdSe QDs to create an efficient artificial photocatalytic system for H₂ evolution.¹³⁷

Fig. 14 (a) Proposed scaffold for the hybrid system of PAA–4/MPA–CdSe QDs. (b) Schematic illustration of the synthetic route of PEI-g-Fe₂S₂ (5) production. (a) Reproduced with permission from ref. 136. Copyright 2013, John Wiley and Sons. (b) Reproduced with permission from ref. 137. Copyright 2015, John Wiley and Sons.

Some molecular cobalt complexes are also employed to combine with CdX nanocrystals for photocatalytic H₂ production.^138–143 For example, by using CdS nanoparticles as the photocatalyst and cobaloxime complexes (6–8, Fig. 15) as molecular cocatalysts, Li and co-workers have demonstrated an efficient hybrid system for efficient photocatalytic H₂ evolution.¹³⁸ In this system, they found the physical adsorption of cobaloxime complexes on the surface of CdS particles facilitated the interfacial electron transfer from photoexcited CdS to cobaloxime complexes. Among various combinations of cobaloxime complexes with CdS, the cobaloxime 6/CdS hybrid system showed the highest activity for H₂ production from a TEOA/H₂O/CH₃CN solution, giving a TON of 171 (based on 6) for H₂ generation over 15 h of irradiation (λ > 420 nm) and a QE of 9.1% at 420 nm. In another example, a cobaloxime complex 9 (Fig. 15) was anchored onto the surface of CdSe/ZnS core/shell QDs via a phosphonate linkage, generating a hybrid system.¹³⁹ In such a system, the ZnS shell could greatly improve the photostability and fluorescence quantum efficiency of the QDs compared with CdSe without the shell. Consequently, the cobaloxime 9/CdSe/ZnS system was capable of catalyzing H₂ generation photochemically in the presence of a proton source (triethylamine hydrochloride) and a sacrificial agent (TEOA) in toluene, giving a TON of over 10 000 H₂ per QD in 10 h. The work by Llobet et al. showed that the integration of a molecular cobalt cocatalyst 10 (Fig. 15) with CdTe QDs could result in efficient light-induced proton reduction to molecular H₂.¹⁴⁰ Chen et al. reported an efficient artificial photocatalytic system, consisting of CdS nanorods as the photocatalyst and Co-based complex 11 (Fig. 15) as a molecular cocatalyst for H₂ generation.¹⁴¹ Also, Sun and co-workers demonstrated that a hybrid system with a coordinative interaction between a Co complex 12 (Fig. 15) and CdTe QDs displayed high activity and improved stability for photochemical H₂ generation form water.¹⁴²

Fig. 15 Structures of Co-based complexes 6–12 used as cocatalysts in hybrid systems with CdX nanocrystals. (6–8) Reproduced with permission from ref. 138. Copyright 2011, Elsevier. (9) Reproduced with permission from ref. 139. Copyright 2012, American Chemical Society. (10) Reproduced with permission from ref. 140. Copyright 2014, American Chemical Society. (11) Reproduced with permission from ref. 141. Copyright 2015, The Royal Society of Chemistry. (12) Reproduced with permission from ref. 142. Copyright 2015, The Royal Society of Chemistry.

In recent years, several molecular nickel complexes have been proved to be effective molecular cocatalysts in conjunction with CdX photocatalysts for photochemical H₂ evolution. In 2012, Krauss and coworkers showed that the integration of an in situ-formed Ni-based complex, Ni²⁺–DHLA (DHLA = dihydrolipoic acid) with DHLA-capped CdSe nanocrystals could create a highly active and superiorly stable hybrid system for visible-light-driven H₂ generation¹⁴⁴ (see Fig. 16). Upon irradiation of this catalytic system with 520 nm LEDs, a TON of more than 600 000 (based on the Ni catalyst) and an initial TOF of 7000 h⁻¹ were obtained in the presence of ascorbic acid as the electron donor, corresponding to a QE of over 36%. Moreover, this catalytic system presented undiminished activity for over 360 h. Very recently, by employing CdS nanosheets (∼4 nm in thickness) as the photocatalyst and Ni-based complexes (13–16, see Fig. 17) as the molecular cocatalysts, our droup demonstrated an efficient noble metal-free artificial photocatalytic system for H₂ production.¹⁴⁵ Among the CdS-complex systems containing various molecular nickel complexes (13–16), the CdS/complex 13 catalyst system exhibited the highest activity for H₂ production, and could generate H₂ with a TON of over 28 000 (based on complex 13) in a TEA/CH₃OH/H₂O solution. We ascribed the high activity of this hybrid photocatalytic system to improvement in charge carrier separation and the efficient electron transfer from the CdS nanosheet to molecular nickel complexes. The work by Das et al. demonstrated Ni-based complexes (17–21, Fig. 17) for robust light-driven H₂ production from water in combination with CdSe QDs as the photocatalyst.¹⁴⁶ Wang et al. reported the integration of Ni–DMSA complexes (DMSA = meso-2,3-dimercaptosuccinic acid) with DMSA capped CdSe/CdS core/shell nanocrystals for efficient photocatalytic H₂ production.¹⁴⁷

Fig. 16 Cartoon illustration of the relevant energies for H₂ evolution (at pH 4.5). AA and dHA indicate ascorbic acid and dehydroascorbic acid, respectively. Reproduced with permission from ref. 144. Copyright 2012, The American Association for the Advancement of Science.

Fig. 17 Structures of Ni-based complexes 13–21 used as cocatalysts in hybrid systems with CdX nanocrystals. (13–16) Reproduced with permission from ref. 145. Copyright 2015, John Wiley and Sons. (17–21) Reproduced with permission from ref. 146. Copyright 2015, American Chemical Society.

Molecular copper complexes used for photochemical H₂ production in combination with CdX photocatalysts were less reported compared with iron, cobalt and nickel complexes. Zhang and co-workers showed that a copper(I) cysteine complex formed by mixing Cu(II) ions with cysteine in aqueous solution could greatly enhance the activity of a CdSe photocatalyst for H₂ evolution under visible light irradiation.¹⁴⁸ Differing from the other molecular cocatalysts mentioned above, the copper(I) cysteine complex functioned as an oxidation cocatalyst, which could accept the photogenerated holes from the VB of CdSe and complete its catalytic cycles with a high reaction rate. Under optimal conditions, the copper(I) cysteine complex could enhance the H₂ evolution rate by as much as 150 times.

3.4 Hybridization of CdX with other semiconductors

3.4.1 Sensitization. Semiconductors with smaller bandgaps can be employed to sensitize those with larger bandgaps by forming semiconductor/semiconductor heterostructures. In the general sensitization mechanism, photogenerated electrons can be transferred from the CB of the smaller bandgap semiconductor to the CB of the larger bandgap semiconductor, leaving photogenerated holes in the VB of the smaller bandgap semiconductor.¹⁴ Thus, photogenerated electrons and holes can be spatially separated from each other, leading to decreased recombination probability and increased electron lifetimes.

The combination of CdX photocatalysts with another wider bandgap semiconductor to form heterogeneous photocatalysts is an effective avenue to improve their photoactivity. Among them, CdS/TiO₂ systems have undergone the most extensive studies.^149–167 For example, Lee et al. combined CdS nanowires with TiO₂ nanoparticles and fabricated a CdS/TiO₂ heterogeneous photocatalyst.¹⁴⁹ They proved that these CdS/TiO₂ heterogeneous photocatalysts exhibited a significantly higher H₂ evolution rate than CdS nanowires alone under the same conditions. Park et al. synthesized a series of ternary hybrid composites composed of CdS, TiO₂ and Pt and systematically investigated the effects of varied combinations of the components on the photocatalytic activity for H₂ generation.¹⁵⁷ The authors found that the separation and transfer behaviours of the photogenerated charge carriers were greatly affected by how the ternary hybrid composites were organized. For achieving high efficiency in charge carrier separation, direct particle-to-particle contact between CdS and TiO₂ was required, which facilitates the photogenerated electron transfer from CdS to TiO₂. Moreover, the activity of the ternary hybrid photocatalysts was found to be extremely sensitive to the loading position of Pt on the CdS–TiO₂ composite. The [CdS–(Pt/TiO₂)] nanocomposite formed by photodeposition of Pt nanoparticles onto TiO₂, followed by the loading of CdS onto the Pt/TiO₂, showed the highest H₂ evolution rate (Fig. 18c).

Fig. 18 Illustrative diagrams of the electron transfer in various hybrid photocatalysts comprising CdS, TiO₂, and Pt. Reproduced with permission from ref. 157. Copyright 2008, The Royal Society of Chemistry.

3.4.2 p–n heterojunctions. The fabrication of semiconductor–semiconductor p–n heterojunction represents another effective strategy to improve the photocatalytic performance. In this case, both n- and p-type semiconductors can be excited to generate electron–hole pairs and photogenerated electron–hole pairs can be efficiently separated owing to the internal electric field in the direction from the n-type semiconductor to the p-type semiconductor.¹⁴ The migration of electrons and holes to different termini may also yield improvement in charge carrier separation, and thus enhancement in photocatalytic activity for hydrogen generation.

By means of a hydrothermal assisted ion-exchange route, Hou et al. have synthesized heterogeneous CdS@TaON with a core–shell structure.¹⁶⁸ In such a structure, both the n-type CdS and p-type TaON can be excited to generate electron–hole pairs under visible-light irradiation. The resultant CdS@TaON core–shell structure could generate p–n heterojunctions between the two semiconductors due to the CB position of CdS being more negative than that of TaON, while the VB position of TaON was more positive than that of CdS. Consequently, the photogenerated electrons could be transported from CdS to TaON while the holes were transported from TaON to CdS. Study of their photocatalytic activity for H₂ evolution from an aqueous solution of Na₂S and Na₂SO₃ showed that the optimized (1 wt%) CdS@TaON heterogeneous photocatalyst could achieve a H₂ production rate of 1.53 mmol h⁻¹ g⁻¹ by loading 0.4 wt% Pt cocatalyst under visible-light illumination, which was much higher than those of TaON (0.045 mmol h⁻¹ g⁻¹) and CdS (0.0675 mmol h⁻¹ g⁻¹) under the same conditions. Further, they found that the rate of H₂ production can be raised to a higher value by coupling the CdS@TaON heterogeneous photocatalyst containing 0.4 wt% Pt with 1 wt% graphene, reaching 3.315 mmol h⁻¹ g⁻¹ with a QE of 31% at 420 nm.

3.4.3 Z-scheme combination. Besides sensitization and p–n heterojunctions, Z-scheme combination of two semiconductors to form heterogeneous structures is another alternative strategy to enhance the photocatalytic performance. Differing from sensitization and p–n heterojunction, photogenerated electrons in Z-scheme combination would migrate to semiconductors with a higher VB and recombine with photogenerated holes there.^169–171 As a result, the strongly reductive electrons stay in semiconductors with a higher CB and the strongly oxidative holes are kept in those with a lower VB, which can keep the electron-driven reduction reaction and hole-driven oxidation reaction occurring in different regions in a single heterogeneous structure and thus could improve the photocatalytic efficiency.

Several CdX-based heterogeneous photocatalysts based on Z-scheme combination have been exploited for efficient photocatalytic H₂ production.^172–180 In a typical example, Tada et al. have successfully prepared an anisotropic CdS–Au–TiO₂ nanojunction, in which CdS, TiO₂ and the electron-transfer medium (Au) were all spatially fixed.¹⁷² As illustrated in Fig. 19, small Au nanoparticles (∼3.4 nm) were tightly anchored onto the antase TiO₂, while CdS was coated on Au nanoparticles, forming Au@CdS/TiO₂. In this three-component CdS–Au–TiO₂ nanojunction, photogenerated electrons could migrate from TiO₂ to Au and to CdS under UV-light irradiation. Consequently, these photoinduced electrons confined in the CB of CdS could drive the reductive reaction, while photoinduced holes kept in the VB of TiO₂ could function in the oxidation reaction. Moreover, this Z-scheme combination can also restrict the photocorrosion of CdS because photogenerated holes in CdS can be recombined with electrons from TiO₂via the electron-transfer medium (Au). Consequently, this three-component photocatalyst system exhibited high photocatalytic activity, far exceeding those of single- and two-component systems. What is more, they found that the H₂ evolution rate can be further accelerated by selective deposition of an appropriate amount of Pt onto the surface of the catalyst. In a recent example, Cha and co-workers demonstrated enhanced H₂ production from a novel DNA-assembled Z-scheme TiO₂–CdS photocatalyst system.¹⁷⁵ In this study, DNA was used to assemble Pt@CdS nanorods and TiO₂ nanoparticles into a well-organized Z-scheme photocatalyst system, which exhibited a significant improvement in H₂ production compared to either single photocatalyst or unassembled, dispersed catalyst mixtures. The work by Zhang et al. demonstrated an efficient Z-scheme photocatalyst system consisting of CdS and WO₃ for visible-light driven H₂ produciton.¹⁷⁶

Fig. 19 Energy and diagram scheme of the CdS–Au–TiO₂ three-component systems. Reproduced with permission from ref. 172. Copyright 2006, Nature Publishing Group.

3.5 Hybridization of CdX with carbon-based materials

Many carbon-related nano-structured materials such as carbon nanotubes (CNTs),^181–187 graphene or its derivatives,^{97,103,114,188–202} and graphitic carbon nitride (g-C₃N₄)^203–209 are also widely used to stabilize and enhance the photoactivity of CdX semiconductors, as summarized in Table 3.

Table 3 Photocatalyst systems containing CdX and carbon-based materials for photocatalytic H₂ produciton

Photocatalyst	Optimized mass fraction of carbon-based material	Cocatalyst	Light source^a	Aqueous reaction solution^b	H2 evolution			Ref.
					Activity (μmol h^−1 g^−1)	QE (%)	Enhancement factor^c
a Xe: xenon lamp. b TEOA: triethanolamine, AA: ascorbic acid. c Enhancement factor refers to the enhancing times of photocatalyst H2 generation rate of photocatalyst system containing CdX and carbon-based materials, compared to pure CdX photocatalyst.
MWCNTs/CdS	10 wt%		λ ≥ 420 nm (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	4977	2.16 (420 nm)	1.8	181
CNTs/Cd0.1Zn0.9S	0.25 wt%		λ ≥ 420 nm (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	1564	7.9 (420 nm)	3.2	183
Zn0.83Cd0.17S/CNTs			300–800 nm (Xe)	0.02 M Na2SO3 + 0.1 M Na2S	6030		1.5	184
CdS/CNTs	4 wt%	0.75 wt% NiS	λ > 420 nm (Xe)	0.25 M Na2SO3 + 1.0 M Na2S	12 130		15	186
Zn0.83Cd0.17S/CNTs	0.25 wt%		500 W Xe lamp	0.02 M Na2SO3 + 0.1 M Na2S	5410		1.3	187
CdS/RGO	1 wt%	0.5 wt% Pt	λ ≥ 420 nm (Xe)	10 vol% lactic acid	56 000	22.5 (420 nm)	45.6	188
CdS/GO	5 wt%		λ ≥ 420 nm (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	3140	4.8 (420 nm)	1.3	189
CdS/RGO	10 wt%		λ ≥ 420 nm (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	4200	10.4 (420 nm)	1.5	190
CdS/RGO	1 wt%		λ ≥ 420 nm (Xe)	0.05 M Na2SO3 + 0.1 M Na2S	700		4.8	191
CdS–MoS2/graphene	1.33 wt%	0.67 wt% MoS2	λ ≥ 420 nm (Xe)	20 vol% lactic acid	9000	28.1 (420 nm)	72	192
CdS–Al2O3/GO	1 wt%		500 W Phoenix tungsten halogen lamp	0.25 M Na2SO3 + 0.35 M Na2S	1750	14 (420 nm)	4.2	193
CdS–ZnO/GO	1 wt%		500 W Phoenix tungsten halogen lamp	0.25 M Na2SO3 + 0.35 M Na2S	3755	30 (420 nm)	9.0	193
CdS–ZnO/RGO	2 wt%		300 W Xe lamp	0.1 M Na2SO3 + 0.1 M Na2S	5100		34.0	194
CdS–TaON/RGO	1 wt%	0.4 wt% Pt	λ ≥ 420 nm (Xe)	0.04 M Na2SO3 + 0.1 M Na2S	3165	31 (420 nm)		195
CdS QDs–ZnIn2S4/RGO	1 wt%	0.4 wt% Pt	λ ≥ 420 nm (Xe)	0.04 M Na2SO3 + 0.1 M Na2S	27 000	56 (420 nm)		196
Zn0.8Cd0.2S/RGO	0.25 wt%		AM 1.5G (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	1824	23.4 (420 nm)	4.5	197
CdS/N-doped graphene	2 wt%		λ ≥ 420 nm (Xe)	0.1 M Na2SO3 + 0.1 M Na2S	1050		5.3	198
CdS/sulfonated graphene	2 wt%		Metal halide lamp, λ ≥ 380 nm	0.07 M Na2SO3 + 0.05 M Na2S	166			199
CdS/cysteine-modified RGO	0.7 wt%	2.1 wt% Pt	λ ≥ 420 nm (Xe)	10 vol% lactic acid	29 861	50.7 (420 nm)		200
CdS/RGO	5 wt%	0.5 wt% Pt	λ ≥ 420 nm (Xe)	1.25 M (NH4)2SO3	29 400		7.3	202
g-C3N4/CdS	1 wt%	1 wt% Pt	λ ≥ 420 nm (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	5303	3.6 (420 nm)	2.5	208
CdS-g-C3N4	88 wt%	0.5 wt% Pt	λ ≥ 420 nm (Xe)	0.1 M AA	4494	8.0 (420 nm)	1.8	204
g-C3N4–CdS	85 wt%	9 wt% NiS	λ ≥ 420 nm (Xe)	10 vol% TEOA	2563		4.4	209
CdS-g-C3N4	2 wt%	0.6 wt% Pt	λ ≥ 420 nm (Xe)	0.25 M Na2SO3 + 0.35 M Na2S	4152	4.3 (420 nm)	2	205

---

3.5.1 CdX/CNT heterogeneous photocatalysts. CNTs, as a representative one-dimensional structure allotrope of carbon, are usually used for hybridising with semiconductors for photocatalytic applications owing to their outstanding unique properties, which are as follows:^210–212 (1) CNTs possess a large electron-storage capacity so that it can effectively accept photogenerated electrons from semiconductors in heterogeneous photocatalysts; (2) CNTs can serve as an excellent electron transporter and efficiently prevent the recombination of photogenerated carriers owing to their unique long-rang π electronic conjugation; (3) the high specific surface area of CNTs (>150 m² g⁻¹) can increase the adsorption of reactants; (4) the CNT, as a semiconductor itself, might function as a photoabsorber to sensitize others.

Peng et al. reported a hydrothermal method for the direct growth of CdS nanoparticles (12–28 nm) on the surface of modified CNTs and successfully obtained CdS/CNT heterogeneous photocatalysts with different contents of CNTs.¹⁸¹ In such a heterostructure, CdS nanoparticles were deposited on the outer surfaces of the CNTs, which resulted in an intimate contact between the CdS nanoparticles and CNTs. The researchers demonstrated that those CdS/CNT heterogeneous photocatalysts with appropriate CNT contents exhibited much higher H₂ evolution rates than CdS alone. The optimal 10 wt% CdS–CNT photocatalyst system showed a H₂-evolution rate as high as 4.97 mmol h⁻¹ g⁻¹, while the H₂ evolution rate over pure CdS nanoparticles was only 2.80 mmol h⁻¹ g⁻¹. Moreover, they found that the presence of CNTs could enhance the stability of the catalyst and suppress the photocorrosion of CdS. Kim and co-workers fabricated ternary nanocomposites of CdS–CNT–M (M represents a transition metal) and studied their photocatalytic activity for H₂ production.¹⁸² In this case, M was first deposited on CNTs, which were pretreated with acid, by means of chemical reduction with NaBH₄, followed by the loading of CdS nanoparticles on the as-obtained CNT/M, forming CdS/CNT/M three-component nanocomposites. The authors compared the activities of a series of CdS/CNT/M catalysts containing various M species and demonstrated that the CdS/CNT/Pt system exhibited the highest rate of H₂ evolution (∼0.825 mmol h⁻¹ g⁻¹), which was obviously higher than that over its CdS/CNT counterpart (see Fig. 20). The study by Yu and co-workers showed that the H₂-evolution activity of a Cd_0.1Zn_0.9S photocatalyst could be improved by loading an appropriate amount of CNTs. The optimal 0.25% CNT/Cd_0.1Zn_0.9S photocatalyst system could achieve a H₂-generation rate of 1.56 mmol h⁻¹ g⁻¹ with a QE of 7.9% at 420 nm.

Fig. 20 Illustration of photocatalytic H₂ production in CdS/CNT/M suspensions under light irradiation. M and D refer to metal catalyst and electron donor (Na₂S and Na₂SO₃), respectively. On the right-hand side, the reported work functions of selected materials are given. Reproduced with permission from ref. 182. Copyright 2011, The Royal Society of Chemistry.

3.5.2 CdX/graphene heterogeneous photocatalysts. Since it was first isolated in 2004 from graphite, graphene has offered exciting new opportunities in various important fields, such as electronics and catalysis, owing to its two-dimensional structure with sp²-bonded carbon atoms arranged in a honeycomb structure. Graphene possesses unique but excellent properties, such as rich surface chemistry, an extremely large surface area (∼2620 m² g⁻¹), superior thermal and electrical conductivity, and great mechanical strength, which make it a good candidate to be bonded with semiconductors or to be used as a support for loading semiconductors for photocatalytic applications. The past several years have witnessed tremendous advances in the development of CdX semiconductor–graphene heterogeneous photocatalysts.^{7,103,114,188–202} Generally, the key roles of graphene in these CdX/graphene photocatalyst systems include: (1) promoting photogenerated charge carrier separation and transfer; (2) inhibiting aggregation and overgrowth of CdX nanocrystals; (3) extending sunlight absorption range; (4) increasing the number of adsorption and active reaction sites.

Gong, Yu and co-workers reported a highly-efficient CdS/graphene heterogeneous photocatalyst for photocatalytic H₂ production from water splitting under visible-light irradiation.¹⁸⁸ This CdS/graphene system was fabricated by means of a facile solvothermal method. The optimal CdS/graphene (1.0 wt%) photocatalyst system could achieve a high H₂-production rate of 56 mmol h⁻¹ g⁻¹ with loading of 0.5 wt% Pt cocatalyst (about 4.87 times higher than that of pure CdS nanoparticles), corresponding to an apparent QE of 22.5% at 420 nm. The authors demonstrated that graphene in the CdS/graphene system could serve as an electron collector and transporter to efficiently lengthen the lifetime of photogenerated charge carriers from the CdS nanoparticles (see Fig. 21), which led to enhanced photocatalytic activity for H₂ generation. The work by Zhang et al. showed that a nanocomposite photocatalyst consisting of reduced graphene oxide and Zn_xCd_1−xS exhibited high activity for photocatalytic H₂ production.¹⁹⁷ Some modified graphene materials have also been incorporated with semiconductors to improve their photocatalytic properties.^198–200 In a typical example, Jia and co-workers coupled CdS with N-doped graphene to construct a highly-efficient photocatalyst system.¹⁹⁸ At an optimal content of N-graphene (2 wt%), the rate of H₂ production over CdS/N-graphene reached 1.05 mmol h⁻¹ g⁻¹, while the H₂ evolution rates over CdS/graphene and the pure CdS counterpart were only 0.495 mmol h⁻¹ g⁻¹ and 0.2 mmol h⁻¹ g⁻¹, respectively. They clarified that graphene doped with N could effectively alter the electron density of states in graphene and enhance the conductivity of graphene. Moreover, the N-graphene could act as a protective layer to effectively prevent photocorrosion of CdS under irradiation during the photocatalytic process. Furthermore, Wang and co-workers compared the photocatalytic activity of CdS/graphene and CdS/CNT heterogeneous photocatalysts under the same conditions.²⁰¹ In this case, CdS nanoparticles (∼35 nm) were loaded on graphene sheets or the outer surfaces of CNTs, forming heterogeneous photocatalysts. They found that the H₂-production rate over the optimal CdS/graphene system (0.7 mmol h⁻¹ g⁻¹) was higher than that of the optimal CdS/CNT counterpart (0.52 mmol h⁻¹ g⁻¹). Noticeably, the H₂-production rates over both the CdS/graphene and CdS/CNT systems were higher than that of the single CdS photocatalyst (0.145 mmol h⁻¹ g⁻¹). The researchers attributed the enhancement in photocatalytic activity over the CdS/graphene photocatalyst system to the stronger interactions or the larger contact interface between CdS and graphene. These results demonstrated that by taking the advantages of carbon-related materials in composites, the photocatalytic activity and stability of semiconductors can be remarkably improved.

Fig. 21 (a) TEM and (b) HRTEM images of graphene decorated with CdS clusters. The inset of (b) shows the SAED pattern. (c) Schematic illustration of the charge separation and transfer in the graphene–CdS photocatalyst system under visible light. Reproduced with permission from ref. 188. Copyright 2011, American Chemical Society.

3.5.3 CdX/g-C₃N₄ heterogeneous photocatalysts. Recently, g-C₃N₄ has shown promising applications in photocatalysis and photoelectrocatalysis due to its unique semiconducting properties and excellent chemical and thermal stability.²¹³ Many studies have demonstrated that it is a good candidate to be bonded with semiconductors or to be used as a support for loading semiconductors for constructing superior heterogeneous photocatalysts. For instance, Ge et al. prepared CdS QDs/g-C₃N₄ nanocomposites via a chemical impregnation method, which showed efficient activity for photocatalytic H₂ production.²⁰³ The optimal loading content of CdS QD was found to be 30 wt%, and the corresponding visible-light-driven H₂ evolution rate from the resultant CdS/g-C₃N₄ photocatalyst system reached 0.173 mmol h⁻¹ g⁻¹ with 1.0 wt% Pt cocatalyst loading, exceeding that of Pt/g-C₃N₄ by more than 9 times. Similarly, Xue and co-workers reported the in situ growth of CdS QDs on g-C₃N₄ nanosheets via a solvothermal method involving dimethyl sulfoxide (DMSO).²⁰⁴ The optimal CdS (12 wt%)/g-C₃N₄ exhibited a 115 times higher H₂-evolution rate than that of pure g-C₃N₄. In a recent example, Yu and co-workers synthesized CdS/g-C₃N₄ core/shell nanowires with different g-C₃N₄ contents by a combined solvothermal and chemisorption method²⁰⁵ (Fig. 22a). The optimal g-C₃N₄ content was determined to be 2 wt%. In this case, g-C₃N₄ could form thin and unclosed shells on the CdS nanowires. In such CdS/g-C₃N₄ core/shell nanowires, the well-matched overlapping band structures between CdS and g-C₃N₄ facilitate the separation of photogenerated charge carriers (Fig. 22b). Consequently, this CdS/g-C₃N₄ photocatalyst system could reach a H₂-production rate of up to 4.152 mmol h⁻¹ g⁻¹ with an apparent QE of up to 4.3% at 420 nm. Moreover, the resultant CdS/g-C₃N₄ photocatalyst system exhibited excellent photostability due to the fact that the corrosive photogenerated holes could be effectively transferred from CdS to g-C₃N₄.

Fig. 22 (a) TEM image of the CdS/g-C₃N₄ core/shell nanowire. (b) Schematic illustration of the energy-level configuration and the photogenerated charge-transfer process between g-C₃N₄ and CdS. Reproduced with permission from ref. 205. Copyright 2013, American Chemical Society.

4. Conclusions and future perspective

The last few decades have witnessed a sustained increase in research attention on semiconductor-based photocatalysis applications. In the present review, we selectively summarized and analyzed the recent progress in photocatalytic water splitting for H₂ production mediated by CdX semiconductors. It is encouraging that the efficiencies of photocatalytic H₂ production with CdX-based photocatalysts have been tremendously improved, benefiting from the great advances in materials science and nanotechnology during the past few decades.

Bulk CdX semiconductors alone display low photoactivity and bad stability for H₂ production due to the ultrafast surface/bulk recombination of photogenerated charge carriers and the large H₂ evolution overpotential, while researchers have made great achievements in designing and synthesizing CdX materials, especially those of nanoscale size, with controlled size, shape, morphology and crystal structure, to improve their photocatalytic performance. To further address the issues of low activity and bad stability in photocatalysis, current efforts are focused on the modification of CdX with other materials to enhance their photocatalytic activity and stability. Promising strategies, such as forming a solid solution, loading with cocatalysts and hybridization with other semiconductors or carbon-based materials, have been widely developed for CdX semiconductors to maximize their photoactivity. These strategies can significantly facilitate separation and migration of photogenerated electrons and holes, reduce charge recombination probability, improve photoactivity and prolong the functional life-time of photocatalysts.

Despite this progress, several challenges still need to be addressed to obtain a satisfying H₂ production efficiency from solar water splitting. Firstly, the current state-of-the-art CdX-based photocatalyst systems are still far from being used for sustainable H₂ production from water with both economic and environmental benefits. Secondly, the photocorrosion of CdX under irradiation in the photocatalytic process and the poor photostability are still key problems obstructing large-scale application. Thirdly, for most of the CdX-based heterogeneous photocatalyst systems, the interrelationships among compositions, microstructures and photocatalytic performance give no clear and systematic answers. Additionally, introducing additional electron donors (sacrificial reagents) to achieve efficient H₂ production from water is needed in most of the current photocatalytic systems.

In order to address these challenges and promote the further development of CdX-based photocatalytic H₂ production, further work in this field should be focused on the following aspects: (i) exploiting novel CdX-based photocatalysts with controlled structure, composition and morphology and tuning their light absorption properties, bandgaps and the band-edge positions; (ii) conducting more in-depth theoretical and computational studies to probe the interrelationships among compositions, microstructures and photocatalytic properties; (iii) developing more novel and effective modification strategies to enhance the photocatalytic activity and stability of CdX semiconductors; (iv) exploring more new characterization technologies to uncover the photogenerated charge carrier transfer dynamics and the mechanism of H₂ production in heterogeneous photosynthetic systems, which can purposefully direct the design and synthesis of highly efficient photocatalysts; (v) constructing highly efficient photocatalytic systems to achieve overall water splitting from pure water without needing extra sacrificial reagents and organic solvents. Continued research in this exciting field will surely accelerate the development of photocatalytic water splitting technology and open new opportunities for extension to other semiconductor-based photocatalytic applications, such as pollutant photodegradation, CO₂ photoreduction and organic photosynthesis.

Acknowledgements

We appreciate the National Natural Science Foundation of China (no. 21422104) for financial support.

Notes and references

1. Y. Tachibana, L. Vayssieres and J. R. Durrant, Nat. Photonics, 2012, 6, 511.
2. N. Armaroli and V. Balzani, Angew. Chem., Int. Ed., 2007, 46, 52.
3. J. Barber, Chem. Soc. Rev., 2009, 38, 185.
4. T. R. Cook, D. K. Dogutan, S. Y. Reece, Y. Surendranath, T. S. Teets and D. G. Nocera, Chem. Rev., 2012, 110, 6474.
5. J. Chow, R. J. Kopp and P. R. Portney, Science, 2003, 302, 1528.
6. Z. Han and R. Eisenberg, Acc. Chem. Res., 2014, 47, 2537.
7. F. Wen and C. Li, Acc. Chem. Res., 2013, 46, 2355.
8. L.-Z. Wu, B. Chen, Z.-J. Li and C.-H. Tung, Acc. Chem. Res., 2014, 47, 2177.
9. A. Fujishima and K. Honda, Nature, 1972, 238, 37.
10. M. B. Wilker, K. J. Schnitzenbaumer and G. Dukovic, Isr. J. Chem., 2012, 52, 1002.
11. M. Amelia, C. Lincheneau, S. Silvi and A. Credi, Chem. Soc. Rev., 2012, 41, 5728.
12. Q. Li, X. Li, S. W. Ahmed, A. Al-Ghamdi and J. Yu, Adv. Energy Mater., 2015, 5, 1500010.
13. X. Chen, S. Shen, L. Guo and S. S. Mao, Chem. Rev., 2010, 110, 6503.
14. W. Fan, Q. Zhang and Y. Wang, Phys. Chem. Chem. Phys., 2013, 15, 2632.
15. R. Marschall, Adv. Funct. Mater., 2014, 24, 2421.
16. A. Kudo and Y. Miseki, Chem. Soc. Rev., 2009, 38, 253.
17. J. Ran, J. Zhang, J. Yu, M. Jaroniec and S. Z. Qiao, Chem. Soc. Rev., 2014, 43, 7787.
18. Y. Xu and R. Xu, Appl. Surf. Sci., 2015, 351, 779.
19. X. Chen and W. Shangguan, Front. Energy, 2013, 7, 111.
20. T. Hisatomi, J. Kubota and K. Domen, Chem. Soc. Rev., 2014, 43, 7520.
21. A. Vaneski, J. Schneider, A. S. Susha and A. L. Rogach, J. Photochem. Photobiol., C, 2014, 19, 52.
22. K. Zhang and L. Guo, Catal. Sci. Technol., 2013, 3, 1672.
23. F. E. Osterloh, Chem. Soc. Rev., 2013, 42, 2294.
24. C.-H. Liao, C.-W. Huang and J. C. S. Wu, Catalysts, 2012, 2, 490.
25. M. Wang, K. Han, S. Zhang and L. Sun, Coord. Chem. Rev., 2015, 287, 1.
26. Y. Xu and B. Zhang, Catal. Sci. Technol., 2015, 5, 3084.
27. X. Li, J. Yu, J. Low, Y. Fang, J. Xiao and X. Chen, J. Mater. Chem. A, 2015, 3, 2485.
28. Y. Qu and X. Duan, Chem. Soc. Rev., 2013, 42, 2568.
29. D. Zhang, G. Li, H. Li and Y. Lu, Chem. – Asian J., 2013, 8, 26.
30. J. Yang, D. Wang, H. Han and C. Li, Acc. Chem. Res., 2013, 46, 1900.
31. D. Meisser, R. Memming and B. Kastening, J. Phys. Chem., 1988, 92, 3476.
32. N. Bao, L. Shen, T. Takata, K. Domen, A. Gupta, K. Yanagisawa and C. A. Grimes, J. Phys. Chem. C, 2007, 111, 17527.
33. Y. Li, J. Du, S. Peng, D. Xie, G. Lu and S. Li, Int. J. Hydrogen Energy, 2008, 33, 2007.
34. X. Wang, M. Liu, Z. Zhou and L. Guo, J. Phys. Chem. C, 2015, 119, 20555.
35. J. Wang, B. Li, J. Chen, N. Li, J. Zheng, J. Zhao and Z. Zhu, J. Alloys Compd., 2012, 535, 15.
36. F. A. Frame, E. C. Carroll, D. S. Larsen, M. Sarahan, N. D. Browning and F. E. Osterloh, Chem. Commun., 2008, 2206.
37. Y. Xu, W. Zhao, R. Xu, Y. Shi and B. Zhang, Chem. Commun., 2013, 49, 9803.
38. Y. Huang, Y. Xu, J. Zhang, X. Yin, Y. Guo and B. Zhang, J. Mater. Chem. A, 2015, 3, 19507.
39. Y. Peng, L. Shang, T. Bian, Y. Zhao, C. Zhou, H. Yu, L.-Z. Wu, C.-H. Tung and T. Zhang, Chem. Commun., 2015, 51, 4677.
40. M. Sathish and R. P. Viswanath, Catal. Today, 2007, 29, 421.
41. J. Zhao, M. A. Holmes and F. E. Osterloh, ACS Nano, 2013, 7, 4316.
42. I. Grigioni, M. Bernareggi, G. Sinibaldi, M. V. Dozzi and E. Selli, Appl. Catal., A, 2015 DOI:10.1016/j.apcata.2015.09.021.
43. Y. Yu, J. Zhang, X. Wu, W. Zhao and B. Zhang, Angew. Chem., Int. Ed., 2012, 51, 897.
44. M. Liu, L. Wang, G. Lu, X. Yao and L. Guo, Energy Environ. Sci., 2011, 4, 1372.
45. G. J. Liu, Z. H. Zhou and L. J. Guo, Chem. Phys. Lett., 2011, 509, 43.
46. J. Zhang, Q. Xu, S. Z. Qiao and J. Yu, ChemSusChem, 2013, 6, 2009.
47. X. H. Zhang, D. W. Jing and L. J. Guo, Int. J. Hydrogen Energy, 2010, 35, 7051.
48. M. Kimi, L. Yuliati and M. Shamsuddin, Int. J. Hydrogen Energy, 2011, 36, 9453.
49. Y.-G. Yu, G. Chen, L.-X. Hao, Y.-S. Zhou, Y. Wang, J. Pei, J.-X. Sun and Z.-H. Han, Chem. Commun., 2013, 49, 10142.
50. J. Yu, J. Zhang and M. Jaroniec, Green Chem., 2010, 12, 1611.
51. S. N. Garaje, S. K. Apte, S. D. Naik, J. D. Ambekar, R. S. Sonawane, M. V. Kulkarni, A. Vinu and B. B. Kale, Environ. Sci. Technol., 2013, 47, 6664.
52. J. Yu, Q. Li, S. Liu and M. Jaroniec, Chem. – Eur. J., 2013, 19, 2433.
53. A. Kubacka, M. Fernández-García and G. Colón, Chem. Rev., 2012, 112, 1555.
54. K. Ikeue, S. Shiiba and M. Machida, Chem. Mater., 2010, 22, 743.
55. B. B. Kale, J.-O. Baeg, S. N. Lee, H. Chang, S.-J. Moon and C. W. Lee, Adv. Funct. Mater., 2006, 16, 1349.
56. X. Wu, Y. Yu, Y. Liu, Y. Xu, C. Liu and B. Zhang, Angew. Chem., Int. Ed., 2012, 51, 3211.
57. Q. Li, H. Meng, P. Zhou, Y. Zheng, J. Wang, J. Yu and J. Gong, ACS Catal., 2013, 3, 882.
58. G. Zhang, D. Monllor-Satoca and W. Choi, Catal. Sci. Technol., 2013, 3, 1790.
59. N. Bao, L. Shen, T. Takata and K. Domen, Chem. Mater., 2008, 20, 110.
60. L. Amirav and A. P. Alivisatos, J. Phys. Chem. Lett., 2010, 1, 1051.
61. H. Yan, J. Yang, G. Ma, G. Wu, X. Zong, Z. Lei, J. Shi and C. Li, J. Catal., 2009, 266, 165.
62. Y. Wang, Y. Wang and R. Xu, J. Phys. Chem. C, 2013, 117, 783.
63. J. U. Bang, S. J. Lee, J. S. Jang, W. Choi and H. Song, J. Phys. Chem. Lett., 2012, 3, 3781.
64. J. Yu, Y. Yu and B. Cheng, RSC Adv., 2012, 2, 11829.
65. J. Jin, J. Yu, G. Liu and P. K. Wong, J. Mater. Chem. A, 2013, 1, 10927.
66. Q. Xiang, B. Cheng and J. Yu, Appl. Catal., B, 2013, 138, 299.
67. L. A. Silva, S. Y. Ryu, J. Choi, W. Choi and M. R. Hoffmann, J. Phys. Chem. C, 2008, 112, 12069.
68. M. Sathish, B. Viswanathan and R. P. Viswanath, Int. J. Hydrogen Energy, 2006, 31, 891.
69. B. Girginer, G. Galli, E. Chiellini and N. Bicak, Int. J. Hydrogen Energy, 2009, 34, 1176.
70. Y. X. Li, Y. F. Hu, S. Q. Peng, G. X. Lu and S. B. Li, J. Phys. Chem. C, 2009, 113, 9352.
71. J. S. Jang, U. A. Joshi and J. S. Lee, J. Phys. Chem. C, 2007, 111, 13280.
72. D. W. Jing and L. J. Guo, J. Phys. Chem. B, 2006, 110, 11139.
73. N. Z. Bao, L. M. Shen, T. Takata, D. L. Lu and K. Domen, Chem. Lett., 2006, 35, 318.
74. T.-T. Zhuang, Y. Liu, M. Sun, S.-L. Jiang, M.-W. Zhang, X.-C. Wang, Q. Zhang, J. Jiang and S.-H. Yu, Angew. Chem., Int. Ed., 2015, 54, 11495.
75. P. Tongying, F. Vietmeyer, D. Aleksiuk, G. J. Ferraudi, G. Krylova and M. Kuno, Nanoscale, 2014, 6, 4117.
76. R. M. Navarro, F. del Valle and J. L. G. Fierro, Int. J. Hydrogen Energy, 2008, 33, 4265.
77. S. Saha, G. Das, J. Thote and R. Banerjee, J. Am. Chem. Soc., 2014, 136, 14845.
78. J. Thote, H. B. Aiyappa, A. Deshpande, D. D. Díaz, S. Kurungot and R. Banerjee, Chem. – Eur. J., 2014, 20, 15961.
79. Y. J. Zhang and L. Zhang, Appl. Surf. Sci., 2009, 255, 4863.
80. R. Sasikala, A. R. Shirole, V. Sudarsan, K. G. Girija, R. Rao, C. Sudakar and S. R. Bharadwaj, J. Mater. Chem., 2011, 21, 16566.
81. I. B. Rufus, B. Viswanathan, V. Ramakrishnan and J. C. Kuriacose, J. Photochem. Photobiol., A, 1995, 91, 63.
82. I. B. Rufus, V. Ramakrishnan, B. Viswanathan and J. C. Kuriacose, Langmuir, 1990, 6, 565.
83. H. Wang, W. Chen, J. Zhang, C. Huang and L. Mao, Int. J. Hydrogen Energy, 2015, 40, 340.
84. T. Simon, N. Bouchonville, M. J. Berr, A. r. Vaneski, A. Adrović, D. Volbers, R. Wyrwich, M. Döblinger, A. S. Susha, A. L. Rogach, F. Jäckel, J. K. Stolarczyk and J. Feldmann, Nat. Mater., 2014, 13, 1013.
85. S. Chen, X. Chen, Q. Jiang, J. Yuan, C. Lin and W. Shangguan, Appl. Surf. Sci., 2014, 316, 590.
86. M. Zhukovskyi, P. Tongying, H. Yashan, Y. Wang and M. Kuno, ACS Catal., 2015, 5, 6615.
87. S. Cao, C.-J. Wang, X.-J. Lv, Y. Chen and W.-F. Fu, Appl. Catal., B, 2015, 162, 381.
88. C.-T. Dinh, M.-H. Pham, F. Kleitz and T.-O. Do, J. Mater. Chem. A, 2013, 1, 13308.
89. Z.-J. Li, X.-B. Li, J.-J. Wang, S. Yu, C.-B. Li, C.-H. Tung and L.-Z. Wu, Energy Environ. Sci., 2013, 6, 465.
90. Z. Khan, M. Khannam, N. Vinothkumar, M. De and M. Qureshi, J. Mater. Chem., 2012, 22, 12090.
91. X. P. Chen, W. Chen, P. B. Lin, Y. Yang, H. Y. Gao, J. Yuan and W. F. Shangguan, Catal. Commun., 2013, 36, 104.
92. X. Chen, W. Chen, H. Gao, Y. Yang and W. Shangguan, Appl. Catal., B, 2014, 152–153, 68.
93. Z. Yan, H. Wu, A. Han, X. Yu and P. Du, Int. J. Hydrogen Energy, 2014, 39, 13353.
94. J. Yuan, J. Wen, Q. Gao, S. Chen, J. Li, X. Li and Y. Fang, Dalton Trans., 2015, 44, 1680.
95. Z. Shen, G. Chen, Y. Yu, Q. Wang, C. Zhou, L. Hao, Y. Li, L. He and R. Mu, J. Mater. Chem., 2012, 22, 19646.
96. J. Ran, J. Yu and M. Jaroniec, Green Chem., 2011, 13, 2708.
97. Z. Yan, X. Yu, A. Han, P. Xu and P. Du, J. Phys. Chem. C, 2014, 118, 22896.
98. J. Ran, J. Zhang, J. Yu and S. Z. Qiao, ChemSusChem, 2014, 7, 3426.
99. L. J. Zhang, R. Zheng, S. Li, B. K. Liu, D. J. Wang, L. L. Wang and T. F. Xie, ACS Appl. Mater. Interfaces, 2014, 6, 13406.
100. J. M. Lee, J. L. Gunjakar, Y. Ham, I. Y. Kim, K. Domen and S.-J. Hwang, Chem. – Eur. J., 2014, 20, 17004.
101. X. Zong, H. Yang, G. Wu, G. Ma, F. Wen, L. Wang and C. Li, J. Am. Chem. Soc., 2008, 130, 7176.
102. D. Cao-Thang, P. Minh-Hao, F. Kleitz and D. Trong-On, J. Mater. Chem. A, 2013, 1, 13308.
103. T. Jia, A. Kolpin, C. Ma, R. C.-T. Chan, W.-M. Kwok and S. C. E. Tsang, Chem. Commun., 2014, 50, 1185.
104. M. Nguyen, P. D. Tran, S. S. Pramana, R. L. Lee, S. K. Batabyal, N. Mathews, L. H. Wong and M. Graetzel, Nanoscale, 2013, 5, 1479.
105. J. Zhang, Z. Zhu and X. Feng, Chem. – Eur. J., 2014, 20, 10632.
106. Y. Lu, D. Wang, P. Yang, Y. Du and C. Lu, Catal. Sci. Technol., 2014, 4, 2650.
107. J. Xiong, Y. Liu, D. Wang, S. Liang, W. Wu and L. Wu, J. Mater. Chem. A, 2015, 3, 12631.
108. R. Sasikala, A. P. Gaikwad, O. D. Jayakumar, K. G. Girija, R. Rao, A. K. Tyagi and S. R. Bharadwaj, Colloids Surf., A, 2015, 481, 485.
109. J. Chen, X.-J. Wu, L. Yin, B. Li, X. Hong, Z. Fan, B. Chen, C. Xue and H. Zhang, Angew. Chem., Int. Ed., 2015, 54, 1210.
110. X. Zong, J. Han, G. Ma, H. Yan, G. Wu and C. Li, J. Phys. Chem. C, 2011, 115, 12202.
111. M. L. Tang, D. C. Grauer, B. Lassalle-Kaiser, V. K. Yachandra, L. Amirav, J. Yano, J. R. Long and A. P. Alivisatos, Angew. Chem., Int. Ed., 2011, 50, 10203.
112. W. Zhang, Y. Wang, Z. Wang, Z. Zhong and R. Xu, Chem. Commun., 2010, 46, 7631.
113. J. Zhang, S. Z. Qiao, L. Qi and J. Yu, Phys. Chem. Chem. Phys., 2013, 15, 12088.
114. J. Zhang, L. Qi, J. Ran, J. Yu and S. Z. Qiao, Adv. Energy Mater., 2014, 4, 1301925.
115. N. Li, B. Zhou, P. Guo, J. Zhou and D. Jing, Int. J. Hydrogen Energy, 2013, 38, 11268.
116. S. Devi, P. Korake, S. N. Achary and N. M. Gupta, Int. J. Hydrogen Energy, 2014, 39, 19424.
117. J. Wang, B. Li, J. Chen, N. Li, J. Zheng, J. Zhao and Z. Zhu, Appl. Surf. Sci., 2012, 259, 118.
118. Y.-P. Yuan, S.-W. Cao, L.-S. Yin, L. Xu and C. Xue, Int. J. Hydrogen Energy, 2013, 38, 7218.
119. L. J. Zhang, T. F. Xie, D. J. Wang, S. Li, L. L. Wang, L. P. Chen and Y. C. Lu, Int. J. Hydrogen Energy, 2013, 38, 11811.
120. J. Zhang, Q. L. Xu, S. Z. Qiao and J. G. Yu, ChemSusChem, 2013, 6, 2009.
121. L. J. Zhang, T. F. Jiang, S. Li, Y. C. Lu, L. L. Wang, X. Q. Zhang, D. J. Wang and T. F. Xie, Dalton Trans., 2013, 42, 12998.
122. H. Liu, K. Zhang, D. Jing, G. Liu and L. Guo, Int. J. Hydrogen Energy, 2010, 35, 7080.
123. U. Gupta, B. G. Rao, U. Maitra, B. E. Prasad and C. N. R. Rao, Chem. – Asian J., 2014, 9, 1311.
124. S. Cao, Y. Chen, C.-J. Wang, P. He and W.-F. Fu, Chem. Commun., 2014, 50, 10427.
125. Z. Sun, H. Zheng, J. Li and P. Du, Energy Environ. Sci., 2015, 8, 2668.
126. Q. Yue, Y. Wan, Z. Sun, X. Wu, Y. Yuan and P. Du, J. Mater. Chem. A, 2015, 3, 16941.
127. Z. Sun, Q. Yue, J. Li, J. Xu, H. Zheng and P. Du, J. Mater. Chem. A, 2015, 3, 10243.
128. Z. Sun, H. Chen, Q. Huang and P. Du, Catal. Sci. Technol., 2015, 5, 4964.
129. S. Cao, Y. Chen, C.-C. Hou, X.-J. Lv and W.-F. Fu, J. Mater. Chem. A, 2015, 3, 6096.
130. J. S. Jang, D. J. Ham, N. Lakshminarasimhan, W. Y. Choi and J. S. Lee, Appl. Catal., A, 2008, 346, 149.
131. K. A. Brown, S. Dayal, X. Ai, G. Rumbles and P. W. King, J. Am. Chem. Soc., 2010, 132, 9672.
132. K. A. Brown, M. B. Wilker, M. Boehm, G. Dukovic and P. W. King, J. Am. Chem. Soc., 2012, 134, 5627.
133. F. Wang, W. G. Wang, X. J. Wang, H. Y. Wang, C. H. Tung and L. Z. Wu, Angew. Chem., Int. Ed., 2011, 50, 3193.
134. J.-X. Jian, Q. Liu, Z.-J. Li, F. Wang, X.-B. Li, C.-B. Li, B. Liu, Q.-Y. Meng, B. Chen, K. Feng, C.-H. Tung and L.-Z. Wu, Nat. Commun., 2013, 4, 2695.
135. C.-B. Li, Z.-J. Li, S. Yu, G.-X. Wang, F. Wang, Q.-Y. Meng, B. Chen, K. Feng, C.-H. Tung and L.-Z. Wu, Energy Environ. Sci., 2013, 6, 2597.
136. F. Wang, W.-J. Liang, J.-X. Jian, C.-B. Li, B. Chen, C.-H. Tung and L.-Z. Wu, Angew. Chem., Int. Ed., 2013, 52, 8134.
137. W.-J. Liang, F. Wang, M. Wen, J.-X. Jian, X.-Z. Wang, B. Chen, C.-H. Tung and L.-Z. Wu, Chem. – Eur. J., 2015, 21, 3187.
138. F. Y. Wen, J. H. Yang, X. Zong, B. J. Ma, D. G. Wang and C. Li, J. Catal., 2011, 281, 318.
139. J. Huang, K. L. Mulfort, P. Du and L. Chen, J. Am. Chem. Soc., 2012, 134, 16472.
140. C. Gimbert-Suriñach, J. Albero, T. Stoll, J. Fortage, M.-N. Collomb, A. Deronzier, E. Palomares and A. Llobet, J. Am. Chem. Soc., 2014, 136, 7655.
141. H. Chen, Z. Sun, S. Ye, D. Lu and P. Du, J. Mater. Chem. A, 2015, 3, 15729.
142. K. Han, M. Wang, S. Zhang, S. Wu, Y. Yang and L. Sun, Chem. Commun., 2015, 51, 7008.
143. A. Das, Z. Han, M. G. Haghighi and R. Eisenberg, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, 16716.
144. Z. Han, F. Qiu, R. Eisenberg, P. L. Holland and T. D. Krauss, Science, 2012, 338, 1321.
145. Y. Xu, X. Yin, Y. Huang, P. Du and B. Zhang, Chem. – Eur. J., 2015, 21, 4571.
146. A. Das, Z. Han, W. W. Brennessel, P. L. Holland and R. Eisenberg, ACS Catal., 2015, 5, 1397.
147. P. Wang, J. Zhang, H. He, X. Xu and Y. Jin, Nanoscale, 2015, 7, 5767.
148. Y. Peng, L. Shang, Y. Cao, G. I. N. Waterhouse, C. Zhou, T. Bian, L.-Z. Wu, C.-H. Tung and T. Zhang, Chem. Commun., 2015, 51, 12556.
149. J. S. Jang, H. G. Kim, U. A. Joshi, J. W. Jang and J. S. Lee, Int. J. Hydrogen Energy, 2008, 53, 5975.
150. H. Fujii, M. Ohtaki, K. Eguchi and H. Arai, J. Mol. Catal. A: Chem., 1998, 129, 61.
151. J. S. Jang, S. M. Ji, S. W. Bae, H. C. Son and J. S. Lee, J. Photochem. Photobiol., A, 2007, 188, 112.
152. J. S. Jang, H. G. Kim, P. H. Borse and J. S. Lee, Int. J. Hydrogen Energy, 2007, 32, 4786.
153. J. S. Jang, W. Li, S. H. Oh and J. S. Lee, Chem. Phys. Lett., 2006, 425, 278.
154. J. S. Jang, S. H. Choi, H. G. Kim and J. S. Lee, J. Phys. Chem. C, 2008, 112, 17200.
155. Y. J. Zhang, W. Yan, Y. P. Wu and Z. H. Wang, Mater. Lett., 2008, 62, 3846.
156. C. L. Li, J. A. Yuan, B. Y. Han, L. Jiang and W. F. Shangguan, Int. J. Hydrogen Energy, 2010, 35, 7073.
157. H. Park, W. Choi and M. R. Hoffmann, J. Mater. Chem., 2008, 18, 2379.
158. L. Qi, J. Yu and M. Jaroniec, Phys. Chem. Chem. Phys., 2011, 13, 8915.
159. V. M. Daskalaki, M. Antoniadou, G. L. Puma, D. I. Kondarides and P. Lianos, Environ. Sci. Technol., 2010, 44, 7200.
160. D. He, M. Chen, F. Teng, G. Li, H. Shi, J. Wang, M. Xu, T. Lu, X. Ji, Y. Lv and Y. Zhu, Superlattices Microstruct., 2012, 51, 799.
161. S. K. Parayil, J. Baltrusaitis, C.-M. Wu and R. T. Koodali, Int. J. Hydrogen Energy, 2013, 38, 2656.
162. X. Cui, G. Jiang, M. Zhu, Z. Zhao, L. Du, Y. Weng, C. Xu, D. Zhang, Q. Zhang, Y. Wei, A. Duan, J. Liu and J. Gao, Int. J. Hydrogen Energy, 2013, 38, 9065.
163. M. Khatamian, M. S. Oskoui, M. Haghighi and M. Darbandi, Int. J. Energy Res., 2014, 38, 1712.
164. J. Fang, L. Xu, Z. Zhang, Y. Yuan, S. Cao, Z. Wang, L. Yin, Y. Liao and C. Xue, ACS Appl. Mater. Interfaces, 2013, 5, 8088.
165. H. N. Kim, T. W. Kim, I. Y. Kim and S. Hwang, Adv. Funct. Mater., 2011, 21, 3111.
166. D. Jing and L. Guo, J. Phys. Chem. C, 2007, 111, 13437.
167. S. Qian, C. Wang, W. Liu, Y. Zhu, W. Yao and X. Lu, J. Mater. Chem., 2011, 21, 4945.
168. J. Hou, Z. Wang, W. Kan, S. Jiao, H. Zhu and R. V. Kumar, J. Mater. Chem., 2012, 22, 7291.
169. Y.-P. Yuan, L.-W. Ruan, J. Barber, S. C. J. Loo and C. Xue, Energy Environ. Sci., 2014, 7, 3934.
170. K. Maeda, ACS Catal., 2013, 3, 1486.
171. P. Zhou, J. Yu and M. Jaroniec, Adv. Mater., 2014, 26, 4920.
172. H. Tada, T. Mitsui, T. Kiyonaga, T. Akita and K. Tanaka, Nat. Mater., 2006, 5, 782.
173. L. Ding, H. Zhou, S. Lou, J. Ding, D. Zhang, H. Zhu and T. Fan, Int. J. Hydrogen Energy, 2013, 38, 8244.
174. H. Zhu, B. Yang, J. Xu, Z. Fu, M. Wen, T. Guo, S. Fu, J. Zuo and S. Zhang, Appl. Catal., B, 2009, 90, 463.
175. K. Ma, O. Yehezkeli, D. W. Domaille, H. H. Funke and J. N. Cha, Angew. Chem., Int. Ed., 2015, 54, 11490.
176. L. J. Zhang, S. Li, B. K. Liu, D. J. Wang and T. F. Xie, ACS Catal., 2014, 4, 3724.
177. X. Wang, G. Liu, Z. Chen, F. Li, L. Wang, G. Lu and H. Cheng, Chem. Commun., 2009, 3452.
178. X. Wang, G. Liu, L. Wang, Z.-G. Chen, G. Q. Lu and H.-M. Cheng, Adv. Energy Mater., 2012, 2, 42.
179. H. J. Yun, H. Lee, N. D. Kim, D. M. Lee, S. Yu and J. Yi, ACS Nano, 2011, 5, 4084.
180. Z. B. Yu, Y. P. Xie, G. Liu, G. Q. Lu, X. L. Ma and H. M. Cheng, J. Mater. Chem. A, 2013, 1, 2773.
181. T. Peng, P. Zeng, D. Ke, X. Liu and X. Zhang, Energy Fuels, 2011, 25, 2203.
182. Y. Kim and H. Park, Energy Environ. Sci., 2011, 4, 685.
183. J. G. Yu, B. Yang and B. Cheng, Nanoscale, 2012, 4, 2670.
184. L. Wang, Z. Yao, F. Jia, B. Chen and Z. Jiang, Dalton Trans., 2013, 42, 9976.
185. G. Khan, S. K. Choi, S. Kim, S. K. Lim, J. S. Jang and H. Park, Appl. Catal., B, 2013, 142–143, 647.
186. X. Wang, M. Liu, Q. Chen, K. Zhang, J. Chen, M. Wang, P. Guo and L. Guo, Int. J. Hydrogen Energy, 2013, 38, 13091.
187. Z. Yao, L. Wang, Y. Zhang, Z. Yu and Z. Jiang, Int. J. Hydrogen Energy, 2014, 39, 15380.
188. Q. Li, B. Guo, J. Yu, J. Ran, B. Zhang, H. Yan and J. R. Gong, J. Am. Chem. Soc., 2011, 133, 10878.
189. T. Y. Peng, K. Li, P. Zeng, Q. G. Zhang and X. G. Zhang, J. Phys. Chem. C, 2012, 116, 22720.
190. P. Zeng, Q. G. Zhang, T. Y. Peng and X. H. Zhang, Phys. Chem. Chem. Phys., 2011, 13, 21496.
191. A. H. Ye, W. Q. Fan, Q. H. Zhang, W. P. Deng and Y. Wang, Catal. Sci. Technol., 2012, 2, 969.
192. K. Chang, Z. W. Mei, T. Wang, Q. Kang, S. X. Ouyang and J. H. Ye, ACS Nano, 2014, 8, 7078.
193. Z. Khan, T. R. Chetia, A. K. Vardhaman, D. Barpuzary, C. V. Sastri and M. Qureshi, RSC Adv., 2012, 2, 12122.
194. X. W. Wang, L. C. Yin and G. Liu, Chem. Commun., 2014, 50, 3460.
195. J. G. Hou, Z. Wang, W. B. Kan, S. Q. Jiao, H. M. Zhu and R. V. Kumar, J. Mater. Chem., 2012, 22, 7291.
196. J. G. Hou, C. Yang, H. J. Cheng, Z. Wang, S. Q. Jiao and H. M. Zhu, Phys. Chem. Chem. Phys., 2013, 15, 15660.
197. J. Zhang, J. G. Yu, M. Jaroniec and J. R. Gong, Nano Lett., 2012, 12, 4584.
198. L. Jia, D. Wang, Y. Huang, A. Xu and H. Yu, J. Phys. Chem. C, 2011, 115, 11466.
199. X. J. Lv, W. F. Fu, H. X. Chang, H. Zhang, J. S. Cheng, G. J. Zhang, Y. Song, C. Y. Hu and J. H. Li, J. Mater. Chem., 2012, 22, 1539.
200. S. S. Zhang, H. J. Wang, M. S. Yeung, Y. P. Fang, H. Yu and F. Peng, Int. J. Hydrogen Energy, 2013, 38, 7241.
201. A. Ye, W. Fan, Q. Zhang, W. Deng and Y. Wang, Catal. Sci. Technol., 2012, 2, 969.
202. Y. Hong, P. Shi, P. Wang and W. Yao, Int. J. Hydrogen Energy, 2015, 40, 7045.
203. L. Ge, F. Zuo, J. Liu, Q. Ma, C. Wang, D. Sun, L. Bartels and P. Feng, J. Phys. Chem. C, 2012, 116, 13708.
204. S.-W. Cao, Y.-P. Yuan, J. Fang, M. M. Shahjamali, F. Y. C. Boey, J. Barber, S. C. J. Loo and C. Xue, Int. J. Hydrogen Energy, 2013, 38, 1258.
205. J. Zhang, Y. Wang, J. Jin, J. Zhang, Z. Lin, F. Huang and J. Yu, ACS Appl. Mater. Interfaces, 2013, 5, 10317.
206. W. Li, C. Feng, S. Dai, J. Yue, F. Hua and H. Hou, Appl. Catal., B, 2015, 168–169, 465.
207. F. Jiang, T. Yan, H. Chen, A. Sun, C. Xu and X. Wang, Appl. Surf. Sci., 2014, 295, 164.
208. H. Yu, F. Chen, F. Chen and X. Wang, Appl. Surf. Sci., 2015, 358, 385.
209. J. Yuan, J. Wen, Y. Zhong, X. Li, Y. Fang, S. Zhang and W. Liu, J. Mater. Chem. A, 2015, 3, 18244.
210. L. Wang, H. Liu, R. M. Konik, J. A. Misewich and S. S. Wong, Chem. Soc. Rev., 2013, 42, 8134.
211. Y. Yan, J. Miao, Z. Yang, F.-X. Xiao, H. B. Yang, B. Liu and Y. Yang, Chem. Soc. Rev., 2015, 44, 3295.
212. Z. Yang, J. Ren, Z. Zhang, X. Chen, G. Guan, L. Qin, Y. Zhang and H. Peng, Chem. Rev., 2015, 115, 5159.
213. S. Cao and J. Yu, J. Phys. Chem. Lett., 2014, 5, 2101.

---