Nanocomposite Marvels: Unveiling Breakthroughs in Photocatalytic Water Splitting for Enhanced Hydrogen Evolution

Abstract

An overview of the significant innovations in photocatalysts for H₂ development, photocatalyst selection criteria, and photocatalytic modifications to improve the photocatalytic activity was examined in this Review, as well as mechanisms and thermodynamics. A variety of semiconductors have been examined in a structured fashion, such as TiO₂-, g-C₃N₄-, graphene-, sulfide-, oxide-, nitride-, oxysulfide-, oxynitrides, and cocatalyst-based photocatalysts. The techniques for enhancing the compatibility of metals and nonmetals is discussed in order to boost photoactivity within visible light irradiation. In particular, further deliberation has been carried out on the development of heterojunctions, such as type I, type II, and type III, along with Z-systems, and S-scheme systems. It is important to thoroughly investigate these issues in the sense of visible light irradiations to enhance the efficacy of photocatalytic action. In fact, another advancement in this area may include hiring mediators including grapheme oxide and metals to establish indirect Z-scheme montages with a correct band adjustment. The potential consideration of reaction chemology, mass transfer, kinetics of reactions, restriction of light diffusion, and the process and selection of suitable light and photoreactor also will optimize sustainable hydrogen output efficiency and selectivity.

1. Introduction

Due to the exponential increase in global energy consumption and degradation of the environment due to fossil fuels, it is very necessary to enhance renewable and sustainable resources. Recently, solar energy has gained prominence as a clean and ecofriendly source of power. There exist various methods to harness solar energy, with one approach being the utilization of photocatalytic processes to split water and generate hydrogen via the conversion of solar energy. As hydrogen is a developing energy bearer, it has a high energy density without carbon content and moreover is readily transportable. Hydrogen can also discharge energy by straight ignition or in a hydrogen power module with the main side product being water. Of late, the obstacle to the acknowledgment of commercialized water splitting is the necessity to deliver hydrogen production at a focused expense contrasted with the contemporary commercial hydrogen generation procedure, including changing nonrenewable energy sources. Types of solar water splitting systems are as follows: photovoltaic electrolysis (PV-E), photocatalysis, and photoelectrochemical cells (PE-Cs).¹ Photocatalytic systems are much cheaper and have a low operational cost out of the three methods of conversion. The cost of hydrogen production hinges on certain assumptions, particularly that future photocatalytic systems will achieve a solar-to-hydrogen conversion efficiency of up to 10%, a significant improvement compared to the approximately 1% efficiency observed in small-scale trials.^2,3 Given the critical role of the solar-to-hydrogen conversion efficiency in determining overall hydrogen production costs, the primary objective of this endeavor is to develop and build a highly efficient water-splitting technique.⁴

Materials in this environment the must display the properties needed to successfully achieve specific applications. In order to produce hydrogen, the photocatalytical materials require particular parameters. They are (i) adequate band edge location, (ii) a narrow energy gap band, (iii) enhanced charge separation, (iv) improved resistance to recombination, and (v) effective interfacial interlinkages.^5,6 Under these circumstances the synthesis of materials with changeable band gap energies and band edge placement has been given the main emphasis in order to generate the species needed for redox and mostly to absorb the entire energy of the sun, which is the ultraviolet– visible–near-infrared (UV–vis–NIR) sun spectrum.^7,8 Though experts have acknowledged that photocatalytic components such as TiO₂ seem to be UV-powered, in later years photocatalytic materials illuminated by the sun will become more powerful in their photocatalytic usage design and implementation.⁹

Scheme 1. Solar Energy Spectrum Diagram with Wavelength Distribution.

The industrial-scale development of the Earth’s abundant light energy source to perform photocatalytic reactions in the restraint and transferable environments. We can note that solar energy falling on earth contains 5% of UV, 45% visible and 50% NIR energies.¹⁰ To design a photocatalyst, the band edge position of the valence band and conduction band is very important. To design a full-sunlight-driven photocatalyst, it is very daring and rousing to control both its band edge location and band gap energy.⁸ Combining components with certain characteristics that can capture UV–vis–NIR energy can achieve this.¹¹ Besides the basic requirements of photocatalytic activity, there is a further component known as the electron storage material (ESM) is essential to build a photocatalyst to catalyze a reaction under obscure circumstances.^12,13

The energy requirements of the reaction have limited the semiconductor band gap. The most suitable band gap for splitting of water is 1.9–2.3 eV.¹⁴ However, in photocatalytic processes, many of the narrow band gap semiconductors are not durable. Broad range semiconductor gaps like TiO₂ are durable but can only absorb UV light, resulting in reduced solar to hydrogen conversion rates.^15,16 One approach to this problem is to combine two different materials to produce a catalyst based on a heterojunction.

Localized surface plasmon resonance (LSPR) is formed on nanostructures decorated with a metal surface, such as Au, Ag, Al, or Cu.¹⁷⁻¹⁹ Au, Ag, Al, and Cu nanoparticles (NPs) can be used as photosensitizers to tune broad absorption bands from ultraviolet to near-infrared wavelengths through LSPR to suppress the recombination of electron–hole pairs for the application of the photodetector. Thus, LSPR interacted with the piezo potential effect to catalyze the HER.^17−19,190

The photocatalyst is the foundation stone of the photocatalysis process. The transition metal oxide and sulfide compose several hundred photocatalyst compounds. In 2009, a new age of free metal photocatalysts started with the creation of polymer carbon nitride in contrast to metal systems.²⁰ Metal-untied carbon nitrides, which include carbon and nitrogen, are earth-rich and affordable, whereas the majority of the metal photocatalyst is photocorrosive as a result of long-term light radiation. Polymeric carbon nitrides show excessive enlarged stability regardless of light irradiation and the pH of the solution.²¹

The overall water splitting by the photocatalyst is a thermodynamic process consisting of two half reactions to generate hydrogen and oxygen, requiring +237.2 kJ/mol free energy and 1.23 V standard reduction potential.^22,23

There are myriad reviews in regards to overall water splitting dependent on particulate photocatalysts that have been promulgated. Here we are reviewing recent important developments in photocatalysts for water splitting hydrogen production.

2. Mechanism for Splitting Water and Evolving Hydrogen through Photocatalysis

Photocatalytic water splitting is modified by a photosynthetic procedure that may be portrayed from the perspective of the photocatalyst as a synthetic response enhanced via photoillumination. The photosynthetic process in plants is a crucial mechanism involving two photochemical processes and multiple intermediate enzyme reduction reactions. At the point when daylight falls over the plant, chlorophyll, the primary donor in photosystem II, is excited.²⁴ It is a huge marvel in the photocatalytic procedure, assuming crucial jobs in regards to the determination of appropriate photocatalysts for upgraded hydrogen evolution advancement. For the most part, the following advances are associated with the photocatalytic procedure: (1) light collection, (2) charge partition, (3) transport of electrons and holes outside the photocatalyst, and (4) reduction response.

Figure 1.

(a) The mechanism of charge separation in natural photosynthesis. Reprinted with permission from ref (24). Copyright 2020 Elsevier. (b) The phenomenon of charge recombination in the process of photocatalysis. Reprinted with permission from ref (25). Copyright 2015 Royal Society of Chemistry.

The theory for the function of the photocatalytic splitting of water is composed of three components: a photocatalyst, a light source, and a water source. When exposed to light, holes generated in the photocatalyst oxidize the water in to O₂ and H⁺, whereas H⁺ is reduced to H₂ by photogenerated electrons. However, if electrons and holes do not necessarily reach the base, they are recombined and heat is released. In the design of photocatalysts, all phases involved in the photocatalytic process, including photon assimilation, photorecognized separation of charge, charging diffusion/transport, mass transmission, and catalytic responses on the operative location of the catalyst, are crucial to comprehend. The process of photocatalytic water splitting depends greatly on the characteristics of the photocatalyst, the cocatalyst, reaction conditions, and reactant adsorption.

Figure 2.

Schematic of important phases for the hydrogen production in the process of photocatalytic water separation. Reprinted with permission from ref (26). Copyright 2018 Royal Society of Chemistry.

Regarding the time scale, photoexcited charges rapidly reposition themselves within femto- to picoseconds to their respective bands and subsequently transport to the catalytic surface within a nano- or microsecond time frame for oxidation and reduction reactions.²⁶

2.1. The Thermodynamic Aspects of the Water Splitting Process

Energetic electrons are driven to achieve internal equilibrium rather than reside near the band gap, which is primarily because they have a shorter relaxation time within the conduction band. This state of inner equilibrium in electrons is referred to as quasi-balance states, and it corresponds to the potential of electrons and holes at quasi-Fermi levels.²⁷

1

2

3

Here E_c and E_v are the energy level positions of the conduction band minimum and valence band maximum, respectively; k_B is the Boltzmann constant; N_c and N_v are the effective density of states of the conduction band and valence band, respectively; and n and p are concentrations of charge carriers.

The thermodynamic photoreactive driving force is proportional to the electron and hole population differential, as shown in eq 3. Figure 3(a) indicates that the reduction potential and oxidation potential disparity is referred to as the Gibbs energy transition. Water splitting is a reaction that requires energy, as illustrated in Figure 3(b). If electrons are thermodynamic (ΔH = 0), ΔG is null, such that a photocatalytic reaction can be achieved without any availability of net force. This demonstrates that heat is not an impulse for electron–hole pair generation. However, the yield of photocatalytic water splitting has been reported to increase with temperature.

Figure 3.

(a) The electronic structure of photocatalytic semiconductor and associated Gibbs energy change. (b) Splitting of water as mounting reaction. Reprinted with permission from ref (28). Copyright 2018 Wiley.

3. Photocatalyst Selection for Splitting of Water for Hydrogen Evolution

3.1. Photocatalyst Recapitulation and Selection

The main element determining the choice of photocatalyst for water splitting is optimizing efficiency. In the development of photocatalysts, the following frameworks must be contemplated: valence and conduction band position, light harvesting capacities, migration of charge carriers, surface area, and photostability.^29,30 However, photocatalysts with a single component cannot meet the desired photocatalytic characteristics. The standard titanium dioxide (TiO₂) photocatalyst, for instance, suffers from a recombination of the load and a wide band gap; likewise, pure CdS has a lower band gap than TiO₂, but its efficiency is limited because of the recombination of the charges (electrons and holes).³¹ After research on the synthesis of hydrogen from water under solar radiation, some more semiconductors such as ZnO, Ta₂O₅, CdSe, Fe₂O₃, SnO_2, SrTiO3, WO₂, WO₃, CuO, Cu₂O, and so forth have also been noted.^32,33

Figure 4.

(a) Electronic band arrangement of certain semiconductor materials at a pH level of 7. (b) Methods for photocatalytic hydrogen production involving both thermodynamic and kinetic principles. Reprinted with permission from ref (34). Copyright 2004 Elsevier.

3.2. Types of Photocatalysts

3.2.1. Titanium Dioxide (TiO₂)

Titanium dioxide (TiO₂) nanomaterials exhibis a robust energy band configuration,^32,35,36 along with high photochemical stability.³⁶⁻³⁸ These energized radicals and ions then engage in reactions and break down dye molecules, resulting in the production of CO₂ and water. However, TiO₂’s wider bandgap (3.2 eV) limits its ability to efficiently absorb natural light radiation. Therefore, it becomes necessary to modify the TiO₂ structure to enhance its photocatalytic efficiency. Numerous efforts have been undertaken to improve TiO₂’s photocatalytic capabilities, including methods such as cation or anion doping, the addition of extra layers, and the creation of heterostructural hybrids. TiO₂ is an inorganic semiconductor with the well-known form of the n type. It can then be used to shape a p–n heterojunction arrangement with a p-type semiconductor like p-Si, which accelerates movement of the electrons from p-Si to active positions.

3.2.2. Other Metal Oxides

Several types of metal oxides are commonly used as water splitting photocatalyst materials, such as, for instance, ZnO, WO₃, Fe₂O₃, CuO and so forth.³⁹ ZnO contributes to fast photogenerated recombination of charges and thus poor performance induced by its photocatalytic activity. WO₃, on the contrary, is a stable O₂ generation photocatalyst in light radiation; notably, however, no H₂ evolution is noticed at speed because of its narrow conduction band. Meantime, Fe₂O₃ is not very stable and has the same problem as WO₃. Copper oxide (CuO), having a low bandgap of 1.4 eV, is among the principal p-type semiconductors and has been used for photoconducting and photochemical purposes in solar cells.⁴⁰⁻⁴² CuO is commonly used as a photocatalyst sensitizer, as when CuO is used alone as a photocatalyst it is unable to achieve sufficient photocatalytic activity.^43,44

3.2.3. Metal Nitrides

Metal nitride is known as a clear potential H₂ production, O₂ analysis, and CO₂ removal photocatalyst and has exclusion from clear processing to contaminants. The electronic configuration of this visible-light photocatalyst exhibits a limited surface area and an elevated rate of charge recombination.⁴⁵ For this reason, it is essential to enhance the photocatalytic characteristics of g-C₃N₄ in a visible light source with various functional groups in order to overcome these anomalies.^46,47 β-Ge₃N₄is a metal nitride that acts photocatalytically and is likely to function as a possible photocatalyst when RuO₂ is spread on a nitride surface as a photocatalyst for the evolution of H₂ and O₂.⁴⁸

3.2.4. Metal Sulfides

Metal sulfides are also known as forward-looking photocatalyst candidates that respond clearly to light. The sulfide valence band generally consists of S 3p orbitals, leading to a more adverse and smaller range than metal oxides. Yet, when it comes to photocatalysts, the question of photocorrosion is a typical problem for most metal sulfides. ZnS is a convenient photocatalyst for the production of H₂ ,but it corresponds to the UV light in its 3.6 eV bandgap. In the absence of sacrificial donors, it experiences photochemical decomposition into components.⁴⁹ However, it is unstable at negative conduction band potentials because of its corrosion behavior. The construction of a heterojunction with an appropriate electronic interface arrangement is, therefore, certainly an effective means of recognizing the separation of photoinducing charge carriers. WS₂ and CuInS₂ have been reported as metal sulfide photocatalysts, but their H₂ evolution efficiency is quite low when used as single materials.

3.2.5. Oxysulfides and Oxynitrides

In comparison to oxide families, two hardly investigated materials contain metal oxynitrides and oxysulfides. They have a wide range of adjustable band gaps from semiconductors to metals. As a result, their conductivity properties vary by structure. These are more covalent than metal oxides and thus, in comparison to metals, are not easily oxidized. Due to their robust half reactions in H₂/O₂ and the absorption of O in the visible region, they were considered to be very good photocatalysts in water splitting is due to their substitution with low-negativity N or S atoms.⁵⁰ The valence bands of oxysulfides like Sm₂–Ti₂S₂O₅ and oxynitrides like TaON consist primarily of blended N 2p (S 3p) and O 2p orbitals, although there are vacant N or S d orbitals as basic elements of the valence band.

A photoelectrochemical (PEC) cell, constructed using Al-doped La₅Ti₂Cu_0.9Ag_0.1S₅O₇ (Al-LTCA) and BaTaO₂N (BTON) photoelectrodes fabricated via particle-transfer (PT) methods, demonstrates the potential for spontaneous overall water splitting under visible light, accompanied by a notably high Faradaic efficiency.¹⁸⁴ The favorable attributes of the Al-LTCA photocathode, such as its long absorption edge wavelength and positive onset potential, contribute to its effectiveness in spontaneous PEC water splitting when combined with photoanodes possessing extended absorption edges, such as BTON, BaNbO₂N, and Ta₃N₅. Despite these advantages, the observed photocurrent in the PEC cell falls below expectations. A primary factor contributing to this lower photocurrent is attributed to the use of a suboptimal electrolyte, as the pH 11 aqueous sulfate solution utilized in the PEC cell is not ideal for both photoelectrodes.

It is worth noting that while an Al-LTCA photocathode may exhibit a relatively high photocurrent in a mildly alkaline aqueous sulfate solution, this photocurrent significantly decreases in strongly alkaline aqueous phosphate solutions. In contrast, a BTON photoanode demonstrates a comparatively substantial photocurrent in strongly alkaline aqueous phosphate solutions.

La₅Ti₂CuS₅O₇ is used with the goal of fashioning photocatalysts responsive to visible light and boasting extended absorption edge wavelengths. Through a solid-state reaction, La₅Ti₂Cu(S_1–xSe_x)₅O₇ (LTCS_1–xSe_xO) solid solutions were synthesized across the composition spectrum of 0 ≤ x ≤ 1, and their physical characteristics, as well as their efficacy in the photocatalytic H₂ evolution reaction in aqueous solutions, were scrutinized.¹⁸⁵

Analyzing the structural refinements disclosed that the LTCS_1–xSe_xO series adhered to Vegard’s law, with lattice constants and cell volumes exhibiting systematic changes. As the Se^2– content increased, the absorption edge of LTCS_1–xSe_xO consistently shifted toward longer wavelengths, reaching a peak at 820 nm for LTCSeO (x = 1). Notably, LTCS_1–xSe_xO with lower Se^2– concentrations displayed comparatively heightened H₂ evolution. For instance, the sample with x = 0.2 produced H₂ under visible light with wavelengths surpassing 640 nm, a feat not achieved by LTCSO.^186,187

3.2.6. Cocatalyst

Finding a photocatalyst that satisfies all criteria for efficient hydrogen and oxygen production is challenging. An effective approach is critical to the development of integrated photocatalyst cocatalysts with enhanced conversion efficiencies that can separate photogenerated electron–hole pairs.⁵¹⁻⁵⁵ Noble metals are used in particular as efficient cocatalysts in the manufacture of photocatalytic responses.^56,57 The most commonly used cocatalysts are noble metals (Pt, Pd, Au, Ru, Rh, Co, Ni, Mo, and W) and the most common metal oxides (NiO and RuO₂).^58,59 Non-noble metal cocatalysts like alloys, phosphides, nitrides, sulfides, and carbides have been studied to increase the photocatalytic generation of hydrogen.⁶⁰⁻⁶⁴ The photocatalyst packs the precious metal onto the surface, and photogenerated electrons migrating to the photocatalyst’s surface are captured by the noble metal cocatalyst due to its lower Fermi energy level compared to that of the semiconductor photocatalyst.^65,66

Essential for maximizing efficiency in the process of photocatalytic and photoelectrochemical (PEC) water splitting, the presence of suitable cocatalysts is paramount.^67,68 These cocatalysts play a crucial role by providing active sites for reduction or oxidation, catalyzing surface reactions through the reduction of activation energies, capturing charge carriers, and preventing the recombination of photogenerated electrons and holes.⁶⁹ Elements like noble metals, transition metal oxides, and sulfides can function as cocatalysts for either reduction or oxidation in photocatalytic reactions.¹⁹¹

In essence, a well-designed photocatalyst should encompass three fundamental functionalities: light harvesting (as a semiconductor) and dual cocatalysts for both reduction and oxidation reactions. The effectiveness of a cocatalyst relies on its compatibility with semiconductors in terms of energy levels and electronic structures. This entails having a harmonious lattice and electronic structures with suitable Fermi levels or band gaps, facilitating the right direction of charge transport processes between the semiconductor and cocatalysts driven by the built-in electric field at the interface.

Comparatively, the water oxidation half reaction poses greater challenges, both thermodynamically and kinetically, making it the rate-determining step for most water splitting photocatalysts. To overcome these challenges and achieve highly efficient overall water splitting, the search for more efficient cocatalysts for water oxidation is imperative. Notably, oxidation cocatalysts play a crucial role in safeguarding light-harvesting semiconductors, such as CdS, from photocorrosion, which is a critical issue, particularly for semiconductors with narrow band gaps like oxynitride and oxysulfide.¹⁹²

4. Strategies for Improving the Photoactivity of Photocatalysts

4.1. Surface Refinement

Multiple methods are utilized on the semiconductor material surface to enhance the separation of charge carriers; these are band gap engineering, regulation of defects, surface plasmon resonance (SPR) impact, and surface refinement.⁷⁰ A pure semiconductor has no sink to capture electrons, so the charges generated will automatically recombine, making it feasible for photocatalytic processing to use just a limited portion of the carrier. The arrangement of the band gap and the presence of surface defects can modify the response to visible light and the process of separating charge carriers.⁷¹ The difference between the higher work function of the metal and that of the semiconductor allows the development of a Schottky barrier, resulting in increased H₂ production efficiency. The efficacy of a semiconductor is improved through the transfer of electrons from the semiconductor conduction band into the metal, and the hole of the semiconductor in the valence band is left behind. This occurs when the Fermi level of metals is less than the semiconductor conduction band energy.⁷² The heavy SPR influence of noble metals such as Au, Ag, and Pt improves visible light absorption and the isolation of charge carriers. The SPR effect causes hot photogenerated electrons under visible light radiation and is transferred to the conduction band of a semiconductor. Eventually, semiconductor electrons are used on the semiconductor surface to reduce H⁺ for H₂ evolution.⁷³ Quantum dots and organic dyes enhance the absorption of visible light and enable the semiconductor to utilize visible light for the generation of electrons in the conduction band, thereby facilitating efficient hydrogen evolution.⁷⁴ Noble metals are frequently employed to enhance visible light absorption through the SPR effect and to isolate charge carriers within a semiconductor. By and large, plasmonic metals loaded into a semiconductor may act to increase the visible radiation absorption of a semiconductor as an electron sink and through photorefinement. Au supplies SPR effects, while silver metal promotes charge carrier isolation.^75,76 Therefore, maximizing the H₂ evolution by the synergistic functioning of two different metals is more likely in comparison with the single metal loading process. Enhancements made to a semiconductor with an inappropriate band structure are impractical, as they cannot be utilized, and the reactions for oxidation and reduction remain ineffective. In these conditions, the formation of heterojunction semiconductors integrated with surface refinement is a promising way of maximizing water splitting H₂ evolution.

Figure 5.

Various surface treatments aimed at achieving visible light absorption and the separation of charge carriers can be categorized as follows: (a) the semiconductor alone, (b) a semiconductor with a Schottky contact to a metal, (c) a semiconductor modified with plasmonic metals, and (d) a semiconductor modified with quantum dots. Reprinted with permission from ref (75). Copyright 2020 Elsevier.

4.2. Heterojunction Emergence

A semiconductor may also improve its photocatalytic activity by creating a heterotransition, which could improve visible light assimilation, bending the band and development of inner electric field and thus dramatically improving photocatalytic behavior by preferential band arrangement and structural charge isolation.⁷⁷ Semiconductors have the capability to form three distinct types of heterojunction structures depending on their band alignment: spanning gap, shifted gap, and fractured gap.^78,79 In the case of a staggered gap heterojunction, semiconductor II has a less favorable conduction band compared to semiconductor I, while semiconductor I has a more favorable valence band than semiconductor II.⁸⁰

Figure 6.

Scheme presenting several forms of heterojunctions for the production of photocatalytic hydrogen: (a) mounting gap heterojunction, (b) lurched gapheterojunction, and (c) fragmented gapheterojunction. Reprinted with permission from ref (78). Copyright 2017 Wiley.

4.3. Z-Scheme Heterojunction System

With the Z-scheme, a proper shuttle redox mediator satisfies the blend of both photocatalysts. The shuttle redox mediator is a pair of acceptors and donors that helps the Z-scheme with two dissimilar photocatalyst varieties that are not in direct physical contact.⁸¹⁻⁸³ Such redox mediators are IO^3–/I^–, NO^2–/NO^3–, Fe²⁺/Fe³⁺, and Co²⁺/Co³⁺. When the light is irradiated, a response in the photosystem for both H₂ and O₂ occurs. Redox mediators play a key role in transferring electrons from O₂ to H₂-generating photocatalysts. However, the concurrent evolutionary responses for H₂ and O₂ are difficult because of the reverse reaction.⁸⁴ The reverse reaction happens when the acceptor and the donor are able to react with light-produced electrons, thereby decreasing the photocatalyst’s quantities of agitated electrons and holes. Semiconductor I and semiconductor II at the interface include a photocatalytic Z-scheme arrangement utilizing a solid electron mediator.^85,86 Hence, a Z-scheme system builds a heterojunction system to obtain more adverse conduction band and more favorable valence band depending on the band positions of the conduction band and valence band for competent oxidation and reduction reactions. The efficiency of the Z-scheme system using mediators and metal loading, which acts as a Schottky bridge for the charge carrier separation, will further improve. Due to their excellent visible light harvesting ability and the band positionin,g oxynitride semiconductors reduce H⁺ and show good stability in aqueous solutions, making them promising candidates for H₂ evolution photocatalysts in Z-scheme systems.

4.4. S-Scheme Photocatalyst Systems

Confronting the challenges outlined above, researchers have actively sought to propose an illustrative mechanism that effectively facilitates the charge transfer process in heterojunction photocatalysts. The forefront contender in this endeavor is the S-scheme heterojunction, which not only addresses the limitations inherent in the type II heterojunction system but also provides a streamlined mechanism to unravel the intricacies of charge transfer pathways in heterojunction photocatalysts. A pivotal moment in this trajectory occurred in 2019 when Yu and collaborators introduced the S-scheme heterojunction. This groundbreaking concept successfully surmounted the challenges associated with compromising the redox ability in type-II heterojunctions while upholding commendable activity levels.

As visualized in Figure 7a and b, the S-scheme heterojunction system ingeniously merges two n-type semiconductors with meticulously staggered band configurations. The oxidation photocatalyst is strategically positioned with a more positively inclined valence band, while the reduction photocatalyst boasts a conduction band with a more negative orientation. In the absence of light, electrons from the reduction photocatalyst (RP) autonomously migrate and accumulate on the adjacent oxidation photocatalyst (OP), as shown in Figure 7a. This results in positive charges being left on the RP due to the higher Fermi level of the latter. Upon achieving Fermi level equilibrium between RP and OP, the region near OP becomes negatively charged, gaining electrons, while the area near RP becomes positively charged. Consequently, this establishes an interfacial electric field (IEF) at the interface, directing the flow from RP to OP.¹⁸⁹

Figure 7.

(a) S-scheme system in the dark. (b) S-scheme system under light irradiation. Reprinted with permission from ref (189). Copyright 2022 Elsevier.

Under the influence of light, as portrayed in Figure 7b, this IEF orchestrates the guided accumulation of photogenerated electrons and holes on the RP and OP, respectively. This orchestrated spatial separation of photogenerated electrons and holes sets the S-scheme heterojunction photocatalysts apart from their type II counterparts, showcasing their optimized redox capability.¹⁸⁹ The unique attributes of the S-scheme heterojunction, promising heightened photocatalytic performance, have spurred extensive research exploring various combinations of oxidation photocatalysts (OPs) and reduction photocatalysts (RPs).

4.5. Photocatalysts Using Titanium Dioxide as a Base Material

Since TiO₂ is nontoxic and cost-effective and exhibits high photostability, it serves as an excellent semiconductor material for photocatalysis.⁸⁷ However, despite these advantages, TiO₂ has three primary limitations. First, the recombination of charge carriers hinders surface reactions, leading to a low energy conversion efficiency. Furthermore, backward reactions arise immediately as hydrogen and oxygen are mixed to create H₂O. Third, TiO₂ is a UV-active catalyst and cannot utilize the entire spectrum of the solar system. TiO₂ has a wider band gap, and it is UV irradiation-active. Because visible light is a large part of solar irradiation, this limits the photocatalyst’s H₂ evolution application. There are several modifications to increase the H₂ production efficiency of TiO₂. Several approaches to improving the efficiency of TiO₂ include adding a scavenger, adding carbonate salts, dye sensitization, as well as constructing composites with TiO₂. The donor of the electron in sacrificial form will help minimize recombination of the charge carrier, as the photogenerated hole in the valence band would react with additional electrons, boosing the isolation of electron–holes and increasing quantum efficiency.⁸⁸⁻⁹⁰ However, this approach has a drawback, as electrons provided by sacrificial agents can be depleted throughout the entire reaction.⁹¹⁻⁹⁵ To mitigate this issue, the addition of carbonate salts is employed to limit the reverse reaction, thereby promoting hydrogen production.⁹⁶⁻¹⁰¹

4.6. Metal and Nonmetal Doping in TiO₂

The use of metal and nonmetallic doping elements in TiO₂ principally helps to modify its characteristics to keep electrons and holes from recombining. Metal doping mainly changes the energy level of the band gap, making it active with visible light. The constraint on charge recombination resulting from nonmetal doping arises from the formation of oxygen vacancies.¹⁰² For instance, within TiO₂ doping, various metals and nonmetals, such as platinum, gold, copper, vanadium, carbon nanotubes (CNTs), iron, tin, carbon, nitrogen, and sulfur, have been referenced.^103,104 TiO₂’s band gap decreased to 2.2 eV with nitrogen and sulfur doping, thus retaining the correct band gap for redox reaction startup.¹⁰⁵ The hydrothermally synthesized bromine- and chlorine-doped TiO₂ leads to extra narrow band gaps for more visible light absorption.¹⁰⁶

For metal doping, sol–gel techniques, photoreduction processes, and hydrothermal processes are typical methods.¹⁰⁷ The introduction of copper-doped TiO₂ resulted in a heightened rate of hydrogen evolution in comparison to unadulterated TiO₂. This enhancement can be ascribed to the effective separation and transfer of charges, facilitated by the Schottky interaction between copper and TiO₂. Furthermore, TiO₂ not only possesses limited active regions for hydrogen evolution but also contains active sites where metals function as cocatalysts to facilitate the reduction reaction of hydrogen evolution. In another case, C/Pt/TiO₂ synthesis reveals that the efficient transfer of charges increased the photocatalytic activity and C as a cocatalyst led to surface enhancement that induced more active areas.¹⁰⁸

The surface plasmon resonance effect is another technique generated by doping metals like (Ag, Au, and Cu) onto the TiO₂. The SPR phenomenon arises when metal nanoparticles are exposed to radiation at their plasmonic resonance frequency, resulting in the creation of a potent local electric field in the vicinity of the metal surface.¹⁰⁹ This produces moving electrons and holes to facilitate the photocatalytic reactions. The progress of H₂ and Ag-TiO₂ under UV as well as visible light irradiation, along with their impressive performance attributed to the establishment of the Schottky barrier, is noteworthy, and effective TiO₂-Ag and Ag SPR-influenced electron transport has been reported.¹¹⁰ Due to the enriched TiO₂ visible light excitement via the Au SPR effect, more electrons were transferred from Au metal on to the surface of the TiO₂, and the production of H₂ by doped Au-TiO₂ deemed threefold higher than that of Pt-TiO₂.¹¹¹Table 1 summarizes the recent study on photocatalytic splitting of hydrogen production by means of modified TiO₂ photocatalysts for metals and nonmetals.

Table 1. Metals and Nonmetal Doping in TiO₂ for Photocatalytic H₂ Evolution.

metal/nonmetal doping	source of light	H2 production	refs
Pt/TiO2	250 W iron halogenide mercury arc lamp with 37 mW cm–2 intensity	18.6 mmol g–1 h–1	(61)
Au/TiO2	250 W Iron halogenide mercury arc lamp with 37 mW cm–2 intensity	13.3 mmol g–1 h–1	(61)
Co/TiO2	300 Xe lamp	1080 μmol h–1	(75)
NiO/TiO2	300 W Xe lamp with 25 mW cm–2 intensity	337 μmol h–1	(76)
Ta-TiO2, Pt/Cr	visible light greater than 420 nm	11.7 μmol h–1	(77)
TiO2/ Pt-CdS	500 W Hg lamp (visible light)	3.7 mL	(78)
Fe/TiO2	500 W halogen lamp with 368 mWm–2 intensity	4.9 mL	(79)
RhB-Co/TiO2	300 W (O3-less Xe lamp)	227 μmol h–1	(80)
Ag2O/TiO2	solar light irradiation with 19.03 mW cm–2 intensity	67 μmol h–1	(81)
Ag/TiO2	250 W (Fe halogenide Hg lamp)	1.17 mmol g–1 h–1	(61)
TiO2-N	150 W xenon lamp solar	28 μmol h–1	(82)
TiO2/Ru	16 W black-light tubes with 0.8 mW cm–2 intensity	71.67 μmol h–1	(83)
TiO2/Rh	16 W black-light tubes with 0.8 mW cm–2 intensity	10.2 μmol h–1	(84)
CuS/TiO2	500 W Xe lamp	111 μmol h–1	(85)
V-TiO2	visible light irradiation	100 μmol h–1	(86)

Figure 8.

(a) TiO₂ band gap structure. (b) TiO2 doping with nonmetals (N and S). (c) Metal doping (W and Mn) in TiO₂. Reprinted with permission from ref (112). Copyright 2017 Elsevier.

5. Metal Oxide-Based Photocatalysts and Their Improvements

5.1. Zinc Oxide (ZnO)

ZnO has a 3.37 eV direct wide band gap and is an exciting 60 meV binding energy.^113,114 Although ZnO may not exhibit the same level of photocatalytic effectiveness as TiO₂, it is still acknowledged for its notable attributes, including high electron mobility, excellent thermal stability, cost-effectiveness, nontoxic nature, robust oxidation capabilities, significant surface-to-volume ratio, environmentally friendly characteristics, and the potential to form well-defined crystals. As a result, ZnO remains a promising semiconductor.^115−118 Due to its limited active site and the possibility of photocorrosion, the performance of ZnO is restricted.^119−124 Composites formed using another material with pure ZnO can increase the photocatalytic hydrogen production. Federal doping of metals such as Au will further boost the photocatalytic function of ZnO.^125−128 Photocorrosion is not present in gold, and it can be firmly grounded as a noble metal on surfaces. Furthermore, it demonstrates a characteristic plasmonic effect on its surface within the visible spectrum due to the collective stimulation of electrons within the gold nanostructure.^129,130

Figure 9.

Schematic diagram for the charge transfer process in RGO/ZnO nanocomposites. Reprinted with permission from ref (75). Copyright 2020 Elsevier.

Regarding hydrogen generation, ZnO-f proved to be the most consistent due to its morphology and heightened efficiency in absorbing light. The photocatalytic performance is further improved by loading Au on ZnO-f. Likewise, CuO-Cu₂O/ZnO has been reported for hydrogen production with improved photocatalytic activity via an interfacial charge carrier for Z-scheme systems.^131−135 A ZnO/CdS photocatalyst exhibited improved production of H₂. Enhanced photocatalytic activity is a result of the Z-scheme heterojunction formed with floral structures, which effectively separates charge carriers.

5.2. Tungsten Trioxide (WO₃)

WO₃’s energy band gap is relatively low, i.e. 2.6 eV, making it a reactive photocatalyst under visible light.^136,137 WO₃ has a valence band potential that is near the valence band potential of TiO₂, but WO₃ does not work efficiently because the conduction band level of WO₃ is small and limits its ability to react effectively to the surface redox.^138,139 A downside of lowering the conduction band of WO₃ so that it promotes swift recombination of the electron–hole pair produced, ultimately resulting in reduced efficiency of photocatalytic activity.¹⁴⁰ Photocatalytic reactions have been limited by the WO₃ photocatalyst. Researchers are currently focused on and dedicating considerable effort to enhancing the photocatalytic performance of WO₃. This involves making modifications like incorporating metals into WO₃ and establishing connections between WO₃ and various other semiconductors.¹⁴¹ Several composites of WO₃, like Pt/Au/WO₃, CdS/WO₃, G-C₃N₄/WO₃, and TiO₂/WO₂, have been explored for improved photocatalytic water splitting.^142−145

One significant approach to boosting photocatalytic activity is achieved by altering WO₃ through the incorporation of TiO₂. When WO₃ is mixed with TiO₂, WO₃ takes on the role of an electron acceptor, which aids in the transfer of electrons from TiO₂ to WO₃. This extends the interfacial charge duration, leading to improved hydrogen production. Additionally, when modified WO₃ is paired with TiO₂, the surface of the photocatalyst becomes more acidic due to the presence of monolayer of WO_x particles on TiO₂. This arrangement leads to an increased absorption of hydroxyl groups on the surface, subsequently enhancing the rate of reduction reactions for hydrogen evolution.¹⁴⁴ WO₃ is a successful photocatalyst that can be joined to more specifically. Many ultraviolet semiconductors, notably AgCl, ZnO, and TiO₂, are triggered under visible light with quicker separation of the load carriers and an adjustable band configuration for a more favorable valence band and more unfavorable conductive band. Thus, photocatalytic H₂ evolution by WO₃ is notably increased under visible light and can be used in the application of solar energy.

Table 2. ZnO-Based Photocatalysts for H₂ Evolution.

materials	source of light	H2 production	refs
Cu/ZnO	Xe lamp with cutoff >420 nm	1932 μmol h–1	(118)
Pt/TiO2–ZnO	400 W mercury arc lamp	2150 μmol g–1 h–1	(119)
Au/ZnO-f	UV–visible irradiation (λ= 300 nm)	427 μmol g–1 h–1	(10)
Graphene/ZnO	UV light	89 μmol g–1	(120)
r-GO/ZnO	300 W xenon lamp	610 μmol g–1 h–1	(121)
Cu2O@ZnO	300 W Hg lamp	236.3 μmol	(122)
ZnO/CdS	300 W halogen lamp with 135 mW cm–2 intensity	6.18 μmol h–1	(123)

5.3. Iron Trioxide (Fe₂O₃)

Hematite is yet another excellent solar water separation photocatalyst with a corresponding band gap, which can be filled with charge excitation and separation in visible light irradiation. Fe₂O₃ is a leading photocatalyst for H₂ production, having 12.7–16.8% conversion efficiency with bad gap 2.0 eV.^146,147 Fe₂O₃’s poor photocatalytic efficiency is due to its high resistance and small diffusion lengths of excitation from 2 to 20 nm.^148−151 The small aperture diffusion distance results in Fe₂O₃ absorbing light only up to 20 nm long, with less hydrogen evolution. Fe₂O₃ moderation improves the performance of photocatalytic water splitting. For example, composite TiO₂ and WO₃ formations and doping with Mo, Cr, Sn, Au, and Pt can be implemented to modify Fe₂O₃.¹⁵² In another case, a heterojunction of grading Fe₂O₃/TiO₂ exhibited a H₂ evolution rate of 217.6 μmol h^–1 due to the synergistic and morphological compositional effects.^153,154 The α-Fe₂O₃/Zn_0.4Cd_0.6S Z-scheme heterostructure exhibited impaired charge recombination and demonstrated efficient H₂ evolution in visible light with a quantum efficacy of 11.2%.¹⁵⁵ Similarly, the electron transfer from Fe₂O₃’s CB to Cu₂O’s VB through rGO as a mediator encouraged H₂ production in the case of a rGO-Cu₂O/Fe₂O₃ composite. Due to the substantial transfer of electrons from CdSe and ZnS to Fe₂O₃, electron–hole pair recombination is decreased, which leads to effective H₂ production. Z-scheme α-Fe₂O₃/g-C₃N₄ has shown higher efficiency in hydrogen evolution under visible light than pure Fe₂O₃ due to its enhanced visible light activity, α-Fe₂O₃ facet,¹¹⁰ quantum size outcomes, increased facet area, and efficient transmission and electron detachment.¹³³ From these we can infer that Fe₂O₃ is an encouraging visible binary light semiconductor that can be accelerated through the coupling of heterojunction with other semiconductive systems and can be used in many applications of solar energy.

5.4. Cupric Oxide and Cuprous Oxide (CuO and Cu₂O)

As photocatalysts for hydrogen generation, CuO and Cu₂O are more negative than the hydrogen generation potential −0.2 V and −0.7 V, respectively, since the conduction band edges of both materials are more negative.^156−,158 CuO is a p-type semiconductor that has a 1.2–1.3 eV indirect band gap, meaning that it can theoretically generate a photocurrent up to 35 mA cm^–2. The Cu₂O is a semiconductor p-type with 2 eV band gap and hence it has a theoretical photocurrent of 14.7 mA cm^–2 in normal air mass 1.5 radiation.¹³⁵ It can absorb visible light in comparison with conventional photocatalysts like TiO₂, which only absorbs UV light. In contrast to Cu₂O, the utilization of CuO for the purpose of producing H₂ through photocatalytic water splitting is significantly limited.^159−161 However, CuO has been used as a hydrogen production cocatalyst as the active material in solar cells, as well as in practical nanostructures and nanocomposites in many papers.^162−166 Although both CuO and Cu₂O photocathodes may undergo corrosion and experience degradation in photocurrent over time, the CuO photocathode appears to exhibit greater stability compared to Cu₂O.^167−170

6. Metal Nitride-Based Photocatalysts and Their Improvement

6.1. Graphitic Carbon Nitride (g-C₃N₄)

Graphitic carbon nitride (g-C₃N₄) possesses distinctive characteristics as a metal-free polymer and an n-type semiconductor, exhibiting exceptional attributes in terms of electrical, optical, structural, and physiochemical properties.¹⁷¹ Generally, g-C₃N₄ has an optical wavelength of 460 nm and a band gap 2.7 eV, which means that it works with visible light.^172,173 Additionally, the other attributes of g-C₃N₄ include special facet properties, nonpoisonous character, abundance, and flexibility that lead to splitting of water with solar irradiation.^174,175 Bulk g-C₃N₄ has, however, a low photocatalytic efficiency because of disadvantages of high rates electron–hole pair recombination, the small g-C₃N₄ size (∼10m²/g), a tiny agile facet for connection, low surface response kinetics, insufficient visible absorption (less than 460 nm), mild oxidation, grain boundary implications, and poor charge mobility.¹⁷⁶ Various modifications have been implemented to limit these, including strategies like adjusting the band gap, employing micro- or nanoscale engineering, adopting bionic approaches, integrating cocatalysts, and enhancing surface properties.^177,178 The electrons produced in the presence of light are driven by a large thermodynamic driving power, suggesting that they have a high H₂ evolution potential in order to be capable of reducing various small molecules including H₂O, CO₂, and O₂. Attaining the correct electronic band alignment for g-C₃N₄ is crucial for its widespread application in the realm of photocatalytic water splitting. While the structure of g-C₃N₄ inherently supports a photocatalytic response, its performance in this regard is significantly hampered by the substantial challenges it faces. Substantial research on g-C₃N₄-based photocatalyst modifications for photocatalytic activities may improve the photocatalyst use of g-C₃N₄. The effectiveness of g-C₃N₄ can be enhanced through the introduction of metals and nonmetals and the creation of heterojunctions. The incorporation of dopant metals and nonmetals into active semiconductor photocatalysts is essential for modifying g-C₃N₄, as it can modify the band gap responsiveness to light and the reduction in band potential.^179,180

N doping in g-C₃N₄ shows enhanced visible light absorption because of an impurity band forming in the vicinity of the valence band, and the synergistic effects of N doping facilitated the development and growing transfer of charges for enhanced hydrogen evolution.^181,182 Doping g-C₃N₄ with O reduced the optical band gap from 0.21 to 2.49 eV, which necessitates the consumption of complete photons by the natural sunlight. The doped material demonstrated approximately four times more photocatalytic activity as compared to pure g-C₃N₄ under identical circumstances. During the photocatalytic response, no evolution of nitrogen was detected, proving the stable H₂ production of O-doped g-C₃N₄. Utilizing metal doping serves as a viable approach for optimizing the band gap, as it substantially enhances light absorption, reduces the band gap, accelerates the movement of charges, and prolongs the lifespan of charge carriers. This results in an improved efficiency for photocatalytic processes. Pt doping to g-C₃N₄ contributes considerably to electrotrapping because of the Schottky barrier that allows efficient separation of charges. For an expanded region of the unique porous vacancy g-C₃N₄ with an enlarged surface region, the evolution of H₂ improved 13.5× as the capacity of the charge separation was enhanced. The adding of Pt as a cocatalyst enhanced photocatalytic activity because of the fact that the Schottky barrier was created to separate photogenerated electron–holes pairs. Pt and Co now have greater ability to boost photocatalytic operation in the codoping of g-C₃N₄.

Incorporating both metals and nonmetals into g-C₃N₄ can enhance the H₂ production, quantum yield, and sustainability. In summary, the combination of metal and nonmetal doping, along with the refinement of lone metal surfaces, when paired with organic dyes or quantum dots, has a notable impact on trapping and facilitating the transport of photogenerated electrons and holes across the semiconductor’s surface under visible light irradiation. The use of nonmetals can alter the valence band position within the semiconductor’s band gap, reducing the band gap energy and enabling greater involvement of visible light, while metals hold the promise of confining electrons through the Schottky barrier. As a result of the change in optical and electronic characteristics, doping g-C₃N₄ with metals and other nonmetals will increase the photocatalytic activity. Besides metallic and nonmetallic doping, the combination with semiconductors is also seen as an important approach for increasing the production of photocatalytic H₂. A photocatalyzer of water in hydrogen and oxygen, coupled with nanoparticles of RuO₂, is noted to be β-Ge₃N₄, a typical metallic nitride with a d¹⁰ electronic configuration.

7. Metal Sulfide-Based Photocatalysts and Their Improvement

7.1. Zinc Sulfide (ZnS)

ZnS nanostructures have been shown to provide strong photocatalysts, for example, halogenated benzene derivatives for photoreductive dehalogensis, CO₂ photoredox, photocatalysis, and photocatalytic water splitting to produce H₂. Moreover, ZnS has many advantages, including good transportation characteristics (reducing scattering and recombination of carriers), intrinsic n-type semiconductor properties, improved thermal stability, lofty electronic movement, nontoxic character, insolubility in water, and low cost. Photocatalysts with long wavelengths are highly desirable for efficient solar light collection. Indeed, a truly effective visible-light-initiated photocatalyst should not only operate efficiently under visible light but also demonstrate stability when exposed to sunlight irradiation. In developing visible-light-active photocatalysts that can use as much solar light as possible efficiently, many changes have been made. The synthesized of ZnS nanostructures include (a) bare ZnS nanoparticles of various morphology, including nanospheres, nanorods, nanotubes, and nanoflowers; (b) nanocomposites containing metal, nonmetal, and dye components with ZnS nanoparticles; and (c) ZnS nanostructures. ZnS is a quasi-conductive photocatalyst with a direct wide range with outstanding chemical stability against hydrolysis and oxidation if the size of the particles is only a meager few nanometers. However, due to a high charge, its photocatalytic efficiency is very low, and the rate of recovery and photocorrosion are not stable under irradiation. A conventional method of preparing efficient photocatalysts and light-responsive photocatalysts was loading a foreign substance into nonoxide ultraviolet-active photocatalysts. ZnS doped with noble metals has demonstrated its efficiency as a photocatalyst for the generation of hydrogen through water splitting under visible light irradiation.

7.2. Cadmium Sulfide (CdS)

The energy band gap of cadmium sulfide is 2.4 eV, which is ideal for the solar spectrum. Owing to photocorrosion, CdS is not stable in aqueous solutions. Pure CdS’s photocatalytic behavior is not that efficient. Many researchers focus primarily on the preparation of materials in which different morphologies and crystal CdS synthesis are used in order to address the problem; that is, a material change is used to enhance the photoactivity of the water splitting photocatalytic system. Recent research efforts have also been directed toward modifying CdS with metal sulfides such as MoS₂ and WS₂. This approach offers a promising avenue for enhancing the performance of CdS photocatalysis, providing a potential alternative to noble metal cocatalysts.

CdS electrons from the conduction band were moved to the Co facet, and the holes remained in the CdS valence band, resulting in efficient separation of charge. Moreover, the enhancement in hydrogen (H₂) evolution was attributed to its relatively low affinity for H⁺ ions, which led to a reduction of H⁺ ions to H₂ by the electrons present on the surface of cobalt (Co). In a separate experiment, composites consisting of CdS and NiSe nanorods exhibited improved photocatalytic performance and greater stability over a period of up to 30 hours. The Ni-loaded CdS nanorods was prepared and applied for the photocatalytic hydrogen (H₂) evolution under visible light irradiation. The Ni-loaded CdS photocatalysts delivered a synergic effect for photocatalytic water splitting H₂ evolution. The synergic effect was due to the formation of a NiO thin layer over the metallic Ni surface (Ni@NiO). The Ni@NiO in Ni@NiO-CdS acted as cocatalyst and trapped the photoexcited electrons from CdS for the reduction of protons (H⁺) to produce H₂ at surface-interface between the Ni@NiO and CdS nanorods, resulting in the enhanced photoactivity for H₂ evolution.

Figure 10.

(a) Band position alignment and (b) charge transfer process for photocatalytic water splitting in the Ni@NiO-CdS heterostructure. Reprinted with permission from ref (183). Copyright 2021 Elsevier.

8. Conclusion

This Review delves into the noteworthy advancements in photocatalysts for hydrogen (H₂) production, examining criteria for photocatalyst selection and various modifications to enhance the photocatalytic activity. The exploration encompasses diverse semiconductors, such as TiO₂, g-C₃N₄,¹⁸⁸ graphene, sulfides, oxides, nitrides, oxysulfides, oxynitrides, and cocatalyst-based photocatalysts. The discussion extends to techniques for improving the compatibility of metals and nonmetals to augment photoactivity under visible light irradiation.

A particular focus is placed on the development of heterojunctions, including type I, II, and III heterojunctions, as well as Z-scheme and S-scheme systems. Emphasis is placed on a comprehensive investigation of these aspects in the context of visible light radiation to enhance the efficacy of photocatalytic processes. Another notable advancement involves the incorporation of mediators, such as graphene oxide and metals, to establish indirect Z-scheme configurations with precise band adjustments. Consideration of reaction chemistry, mass transfer, kinetics, light diffusion limitations, and process intricacies and the selection of suitable light and photoreactors all contribute to optimizing sustainable hydrogen output efficiency and selectivity.

The authors declare no competing financial interest.