Single-Atom Transition Metal Photocatalysts for Hydrogen Evolution Reactions

Abstract

Hydrogen is one of the potential fuels that is easily stored in ammonia compounds and reacts with oxygen in an environmentally friendly manner, producing water and transferring a significant amount of heat for powering mechanical facilities or transportation. Recently, single-atom photocatalysts have attracted significant attention owing to their ability to produce clean fuels or reduce gaseous pollution, thereby contributing to the preservation of our planet. Utilizing metals composed of a single atom on a semiconductor platform can improve the active sites, thereby increasing the efficiency of the hydrogen evolution reaction. This review focuses on the use of single-atom transition metals as photocatalysts in a solar-powered water-splitting system that produces hydrogen gas. The approach to synthesis, reaction mechanism, and current performance of these materials is exhaustively discussed. In addition, the main challenges and improvement strategies are highlighted.

1. Introduction

The impending depletion and negative environmental effects of fossil fuels necessitate a vast supply of renewable energy sources for human life and industry. Hydrogen (H₂) is considered as a potential renewable energy source owing to its abundance, high energy density, and environmental friendliness [1,2,3]. In addition, hydrogen plays an important role in numerous applications, including fuel, oil refinery, fertilizer, and metal refining. Figure 1 illustrates solar hydrogen production and its applications. Using solar energy, the photocatalyst absorbs photons to produce excited electrons that can combine with a proton to produce hydrogen gas. This approach is sustainable and can satisfy the demand for a large-scale, low-cost hydrogen generation system [4,5]. H₂ possesses the lightest weight, the strongest reaction with oxygen that produces water, and a tremendous amount of heat energy, allowing it to operate machinery or generate electricity via a converter [6]. The majority of hydrogen gas is produced by methane steam reforming or oil reforming, but water splitting is significantly less expensive, requires fewer facilities, consumes renewable solar energy, and emits non-toxic by-products [7,8,9]. Therefore, the development of this sustainable energy pathway is the optimal choice for the energy economy of the future.

Protons (H⁺) receive an electron to produce hydrogen, resulting in the formation of hydrogen gas and formation of a bubble on the surface of materials [10]. This reaction depends on the redox potential of H⁺/H₂, and the active proton source depends on the pH of the electrolyte, where the protons are from the free H⁺ in acidic media or the H₂O molecule in basic media [11,12,13,14]. For example, in the acidic media (pH < 5), the protons are directly absorbed on the surface of material, receive the electrons, and are reduced to hydrogen. Meanwhile, in basic media (pH > 8), the protons are separated from water molecules on the surface of active material. Therefore, the water splitting potential depends on the pH level; the HER potential is 0 V vs. standard hydrogen electrode (SHE) at pH = 0, while it is −0.83 V vs. SHE at pH = 14 [15,16]. This change can be calculated as a function of pH; thus, to simplify the reaction, we can mention the behaviour of the adsorbed H⁺ (namely, H_ads). Accordingly, the adsorption ability determines the performance of the hydrogen evolution reaction (HER). The main active material in single-atom photocatalysts is the light harvester, a semiconductor, which absorbs visible light or UV light and then excites the formation of electron-hole pairs, producing an electron source for HER [17]. Excited electrons with higher energy than the band gap of the semiconductor will transfer to the conduction band, while the holes remain in the valence band or join in an oxidation reaction [18]. By selecting a semiconductor with a conduction band close to the H⁺/H₂ reduction level of ~4.5 eV (or 0 V vs. SHE), protons can receive electrons from host materials with relative ease. This is how conventional photocatalysts function [19,20].

Recently, single-metal atom (SMA) photocatalysts have attracted considerable interest [21,22,23]. The term of SMA photocatalyst indicates the decoration of a very tiny metal particle in atomic scale on the host photocatalyst (semiconductors). The single metal functions as a co-catalyst on the surface of semiconductors, increasing the number of active sites, receiving electrons from the light-harvesting process, and efficiently reducing the amount of precious metal. Typically, precious metals such as Pt and Au have a low work function that is close to the H⁺/H₂ reduction level; as a result, they have been utilized as hydrogen catalysts in electrochemical and photoelectrochemical hydrogen generation [24]. Due to the Schottky contact between the metal atom and semiconductor host material, the electron can easily transfer to a lower potential and is able to donate electrons to H_abs, thereby forming hydrogen molecules [25,26]. Common semiconductor hosts include perovskites, TiO₂, and NiO, as well as graphitic carbon nitride (g-C₃N₄) [27]. Meanwhile, to reduce the expense of precious metals, numerous types of abundant transition metals such as Pt, Au, Ni, V, and Fe are used to produce SMA photocatalysts [28,29,30,31].

In this review, the use of transition metal as a single-atom photocatalyst is examined. In addition, the fundamentals and predictions based on theoretical calculations are discussed, and the evaluation, recent developments, and future prospects are discussed.

2. Fundamentals

Typically, the SMA photocatalyst consists of a single metal atom or tiny metal nanoparticle anchored on a host material, as shown in Figure 2a. The host material is semiconductor-type material, which itself can act as a photocatalyst or light harvesting material [21]. As mentioned above, the requirement of active material for HER should have a conduction band close to ~4.5 eV. However, it is hard to tune a band gap of semiconductors as designed. For metal, the conduction band and valence band are overlapped or have no band gap. Therefore, the electron and hole are easily formed in normal conditions. As per previous reports, the work function of metal can be understood by the “activity descriptors” [32]. Thus, Pt with the work function of ~5.5 eV seems to be the best element to effectively perform HER [24]. In fact, many other factors of transition metal affect HER performance, which will be discussed later. To operate the HER, the transition metal nano particle should be anchored on semiconductor material, as described in the schematic in Figure 2a; the host material acts as the light harvester, then gives the generated electrons to metal NPs. The tiny metal NPs receive the electron and perform the HER, producing H₂ gas. Due to the natural catalytic activity and the increasing surface area of metal nanoparticles, photocatalytic properties of SMA can enhance outstanding photocatalytic properties of bare semiconductor material [33].

To fully comprehend the mechanism of hydrogen absorption and reduction on the surface of a transition metal, the density functional theory (DFT) calculation can be used to gain insight into the electron structure, particularly the d-orbital of the transition metal, which determines the conducting electron energy [10,36,37,38]. After adsorption on the material surface, the proton forms a bonding state that can be described as a 1s-d hydrogen-transition metal bond. Hammer et al. reported that the density of one-electron states (DOS) of four typical metals in this group, including Pt, Au, Ni, and Cu, were calculated to predict the chemisorbed hydrogen on their surfaces [39]. Au and Cu have electrons in their H 1s-d antibonding states, while those of Ni and Pt are empty. It indicates that metals interact more strongly with H_ads of Ni and Pt than with H_ads of Au and Cu. However, the H 1s-d bonding in both these transition metals shows a strong intensity, indicating they can highly absorb the hydrogen atom. Therefore, the position and occupation of d-orbitals determine the hydrogen adsorption properties of metal surfaces. In HER, the bond-breaking barrier and the electron transfer process are also essential [40]. Metals appear to have the lowest bond-breaking barrier, allowing them to readily transfer electrons and activate hydrogen in HER. Thus, the use of metal atoms such as Pt, Au, Pd, Co, Ni, and Cu in photocatalyst systems will reduce the massive electron transfer effort of the semiconductor, resulting in more efficient hydrogen production.

The ability to absorb hydrogen can be described by a physical quantity called Gibbs free energy ($\Delta G_{H*}$) [41,42]. The value of $\Delta G_{H*}$ is more negative, indicating that hydrogen bonds to the surface more strongly. The greater the positive value, the more difficult it is to absorb hydrogen. Thus, the material with the closest value to zero for $\Delta G_{H*}$ is the best material for interacting with hydrogen atoms with the lowest activation energy, as it can absorb and release the proton and hydrogen gas, respectively, with ease. Figure 2b depicts the typical free energy diagram of Ni metal at various sites [34]. The Ni 1-zigzag edge has the lowest $\Delta G_{H*}$ value of −0.01 eV, indicating HER’s high activity. In contrast, the other coordinated sites exhibit either strong or very weak hydrogen bonds, making hydrogen absorption difficult. By repeating the calculation with other transition metals, such as Mn, Fe, Co, Ni, Cu, and Pd, Gao et al. obtained a clear picture of these materials in which, if the right coordinated sites are exposed in the bonding with photocatalyst active sites, they are both highly active in HER [34]. Figure 2c displays the volcano plots of Gibbs free energy for the majority of transition metals (Pt, Au, Pd, Rh, Ir, Ni, Co, Mo, Nb, W, Ag), which were summarized by Yao et al. using previously reported data [35,43,44,45,46]. It is easy to evaluate transition metals for hydrogen adsorption based on this graph. However, as stated previously, the Gibbs free energy of a metallic surface is highly dependent on its coordinated sites. Therefore, a graph with numerous coordinated sites must be completely filled to provide an overview of transition metal in HER.

3. Single-Atom Metal Photocatalyst Preparation

To synthesize single-atom photocatalysts, numerous techniques can be used, which can be categorized into three approaches: Absorption, deposition, and refining method (as shown in Figure 3). In the absorption method, metallic salts were absorbed by a structure, such as a metal organic framework (MOF), a polymer structure, or a mixture of precursors of host materials. Zhang et al., for instance, fabricated a few types of SMAs, including Cu, Co, Ni, Fe, Mn, Zn, and Pt, by diffusing these ions into a MIL-125-Ti, which is an MOF structure of Ti with terephthalic acid, and then reducing this structure to SMA on TiO₂ host structure as a photocatalyst for hydrogen production [47]. Cha et al. used TiO₂ dissolved in Pt, Pd, and Au chlorides [48]. The absorption of these ions on the surface of TiO₂ produces 2–5 nm-sized particles that are considered atomic-scale catalysts. The atomic deposition technique permits a single atom to exist on the surface of materials within seconds. Pan et al. investigated the electrostatic deposition method for SMA Ni particle on ZnIn₂S₄, as an example [49]. Due to the large surface area of ZnIn₂S₄, Ni²⁺ ions are readily attracted to the material via electronic attraction. By hydrothermally reducing Ni²⁺, the SMA Ni particle was uniformly decorated on the surface of the host material. Meanwhile, Zhou et al. employed a three-electrode system and a solution of H₂PtCl₆ as a Pt source for electrodeposition on Ni/NiO/Ag nanowire [50]. The amount of Pt particles can be manipulated by adjusting the time and current/voltage of the electrodes system. In the final step of the refining method, metal ions were reduced through a natural decomposition process under light/dark conditions (with light-sensitive compounds of Pt, Ag, and Au) or through the use of reduction agents such as hydrogen or calcination with a reducing agent. Vile et al. applied sodium borohydride to C₃N₄ nanosheets to reduce Ni²⁺ [51]. Additionally, Zhang et al. used a high annealing temperature of ~850 °C to decompose cobalt phthalocyanine in a composition containing carbon black in the presence of nitrogen [52]. Moreover, Zhou et al. used the self-decomposition of IrCl₃ in the absence of light to produce single-atom Ir on TiO₂ nanotubes [53].

Table 1. Summarization of synthesis approaches for SMA photocatalyst.

Single-Atom Metal	Host Materials	Synthesis Method	Ref
Cu, Co, Ni, Fe, Mn, Zn, Pt	TiO2	Absorption on MOF and calcination	[47]
Ni	TiO2	Refining by hydrothermal	[54]
Ni	ZnIn2S4	Deposition: electrostatic deposition + hydrothermal	[49]
Pd, Pt, Au	TiO2	Absorption: Immersing in salts	[48]
Ni	C3N4	Refining by sodium borohydride	[51]
Co	N-Carbon	Refining by calcination	[52]
Pt	Ni/NiO on Ag NWs	Electrodeposition	[50]
V	C3N4	Refining by calcination	[55]
Ir	TiO2	Deposition in dark	[53]
Pt	Graphene	Deposition: Atomic layer deposition	[56]
Cu	BN	Refining by calcination	[57]

provides a summary of additional methods. Approaching the synthesis of SMA photocatalysts may not be a simple process, but it is not an impediment to obtaining the desired photocatalysts.

4. Single-Atom Photocatalysts

4.1. TiO₂ Base

TiO₂ is a well-known photocatalyst with a wide band gap of ~3.2 eV that absorbs ultraviolet light intensely (UV) [58,59]. In addition, it has been used globally for numerous applications, including batteries, ceramic compounds, colorants, and solar cells, owing to its low cost, low toxicity to the environment, and ease of processing [60,61,62]. As a hydrogen photocatalyst, TiO₂ shows a low efficiency due to the ultrafast recombination of excited electron–hole pairs [63]. Correspondingly, the morphology and size of TiO₂ can be altered to improve its catalytic performance [62,64]. However, to reach the large-scale production requirement, there is still a gap to overcome. Numerous strategies aim to modify the TiO₂ surface, including the use of semiconductors with a smaller band gap and the combination of metal particles and graphene, among others. Both received a tremendous performance boost from TiO₂. In particular, the use of SMA on photocatalysts opens a new route for functionalizing the TiO₂ surface with tiny metal particles.

Figure 4a,b demonstrates that Yi et al. synthesized single-atom Co on N-doped graphene composited with TiO₂ nanobelts, resulting in a significant improvement [65]. The rate of H₂ generation could reach ~677 μmolh⁻¹ g⁻¹, which is comparable to the performance of Pt-decorated TiO₂ samples (~741 μmolh⁻¹ g⁻¹). Meanwhile, the hydrogen production efficiency of unmodified TiO₂ and N-doped graphene/TiO₂ is poor and low, respectively. Using a similar approach, Cha et al. compared the effect of using the noble metals Pt, Pd, and Au on the surface of TiO₂, as depicted in Figure 4c,d [48]. The size of tiny noble metals was between 2 and 5 nm. By adjusting the concentration of noble metal chlorides, it is possible to observe the effect of SMA on TiO₂ host material, which demonstrates that a single atom can enhance the performance of photocatalysts by increasing the number of active sites, without obstructing the light to the host material compared to larger nanoparticles. The best performance as an SMA photocatalyst was recorded for Pd, which has a hydrogen production rate of ~600 μLh⁻¹. Zhou et al. deposited Ir on TiO₂ nanotubes using a dark deposition technique, loading the nanotubes with Ir nanoparticles measuring ~2.5 nm in size [53]. The SMA on TiO₂ photocatalyst performs significantly better than the conventionally prepared co-catalyst TiO₂, achieving a high turnover frequency of ~4 × 10⁶ h⁻¹. Zhang et al. reported the SMA Cu on TiO₂, as depicted in Figure 5, after selecting a material with a lower cost [47]. The trapped Cu²⁺ ion in MIL125-Ti was easily converted to SMA Cu/TiO₂ via high-temperature calcination. By comparing various types of transition metals, such as Co, Ni, Fe, Mn, Zn, and Pt, the author concluded that single atom Cu (~2 nm) on TiO₂ possesses a high H₂ evolution rate of ~100 mmolh⁻¹ g⁻¹, which is higher than that of SMA Pt on TiO₂ (~80 mmolh⁻¹ g⁻¹). The significant improvement of these SMA photocatalysts can be attributed to two factors. First, the new approach using MIL-125-Ti demonstrates that an MOF can absorb other ions efficiently if their ionic size and repulsion/attraction force are suitable. In this work, Cu is abundantly absorbed by the network, resulting in the formation of 1–2 atomic-sized uniform single atoms. Cu is a well-known metal that is utilized in HERs effectively due to its compatibility with the H⁺/H₂ redox potential. The disintegration of MIL-125-Ti also generates a large number of Ti vacancies, thereby increasing the specific surface area and exposing sites for Cu atoms. Therefore, the use of SMA in conjunction with a highly porous host material can increase the efficiency of hydrogen generation.

4.2. g-C₃N₄ Base

Graphitic carbon nitride (g-C₃N₄), a two-dimensional layered material that is a member of the graphene-like family, is an efficient metal-free photocatalyst material [66]. In contrast to graphene, which has a zero band gap, g-C₃N₄ has a band gap of ~2–3 eV [67,68]. Consequently, g-C₃N₄ has a broad light absorption band. In addition, g-C₃N₄ has a simple synthesis method, high physical and chemical stability, and an active band gap in relation to the redox potential of water-splitting reaction [69]. Therefore, g-C₃N₄ materials can be utilized as photocatalysts for HER. However, due to its low conductivity and rapid recombination of excited charge carriers, g-C₃N₄ is ineffective as a photocatalyst [70]. Accordingly, surface engineering of g-C₃N₄ is required to overcome the limitation and exploit the semiconducting properties effectively. Due to the abundance of nitrogen atoms in its structure, the surface of g-C₃N₄ can be easily modified by covalent or noncovalent bonding with a variety of functional groups [71]. During its early use as a photocatalyst, the metal decoration in the surface of g-C₃N₄ has garnered considerable attraction [72,73]. The emergence of SMA properties on the host photocatalyst material g-C₃N₄ can have a substantial impact on their hydrogen production performance. Cao et al. utilized a single atom of noble Pt metal on g-C₃N₄ as an effective photocatalyst in HER, as depicted in Figure 6a–b [74]. The Pt atoms are widely distributed on the host surface materials as individual atoms, thereby enhancing the active site and photocatalytic properties. Pt is used not only to increase the active surface area but also to improve the electronic conducting property, resulting in outstanding catalytic properties. The hydrogen production rate of SA Pt on g-C₃N₄ is significantly greater than that of bare g-C₃N₄ and greater than 13 times that of Pt nanoparticles of larger size (a few nanometers) on g-C₃N₄. In addition to V, Co, Cu, and Fe, other transition metals are used in the single atomic concept of g-C₃N₄. To display a wider graph of SMA on g-C₃N₄, however, the synthesis approach still has some limitations. Li et al. fabricated SMA Co metal on P-doped g-C₃N₄ (Co/P/CN) using hexahydrate triethanolamine as a reducing agent during the calcination process in a microwave reactor, as depicted in Figure 6c,d [75]. The rate of H₂ production was measured at ~3730 μmolh⁻¹ g⁻¹. The addition of Co and P atoms to g-C₃N₄ increases its active surface area. Consequently, the presence of P-N and Co-N bindings causes a change in the band gap structure, which is reduced from ~2.78 eV for bare g-C₃N₄ to ~2.58 eV for Co/P/CN, as depicted in Figure 6c. Therefore, the photogenerated hole–electron pairs could easily separate and contribute to the HER. Similarly, Wang et al. reduced V ion during the preparation of g-C₃N₄, resulting in a single atom of V on g-C₃N₄ photocatalyst (SAVCN) for hydrogen evolution reaction (Figure 6e,f) [55]. Additionally, the simulations demonstrated that the V on SAVCN decreases the $\Delta G_{H*}$ from −1.11 to −0.34 eV, which is closer to 0 eV and therefore better for hydrogen adsorption. Under blue and green LED lights, the H₂ production capacity of the SAVCN catalyst was measured and found to be ~5.0 and 3.0 mmolh⁻¹ g⁻¹, respectively. Consequently, the SMA improved the light absorption, hydrogen adsorption, and band gap structure of the host material to facilitate the HER.

4.3. Other Host Materials

In addition to TiO₂ and g-C₃N₄, which are the traditional photocatalyst host materials, there are numerous other semiconductor materials whose band gap matches the requirements for a photocatalyst material [76]. However, their use is contingent on their influence on low-cost production, simple processing, and environmentally friendly properties, which are not adequately accounted for by their benefits and drawbacks. Graphene quantum dots, for instance, are one of the most promising candidates; however, large-scale production and uniformity are still being evaluated [23]. Due to their layered structure, photo-stability, low band gap of ~2.5 eV, and low toxicity, ZnIn₂S₄ semiconductor materials have recently attracted the attention of researchers [77,78,79]. To date, Pan et al. investigated the SMA Ni on ZnIn₂S₄ materials for HER, as shown in Figure 7a,b [49]. Preparing a ZnIn₂S₄ semiconductor with sulphur vacancies and synergizing it with SMA Ni on the surface enabled the photocatalytic evolution of hydrogen. The evolution rate with 0.9 wt.% Ni on ZnIn₂S₄ was approximately 89.4 μmolh⁻¹. Shi et al. fabricated SMA Pt on hexagonal ZnIn₂S₄ [80]. Calculating a new $\Delta G_{H*}$ at −0.23 eV for the third hydrogen atoms demonstrated that the presence of Pt increases the hydrogen adsorption of the photocatalyst. Under solar light, the rate of hydrogen production was approximately 30 mmolg⁻¹ h⁻¹, and under visible light, it was greater than 16 mmolg⁻¹ h⁻¹. These outcomes suggest ZnIn₂S₄ bearing SMA coating is also a possible photocatalyst for hydrogen production.

Organometal halide perovskites (OHP) are well-known in the solar cell industry due to their tuneable band gap, broad light absorption band, long carrier diffusion length, and simple synthesis [83]. In particular, the photocatalytic splitting of hydroiodic acid (HI) in aqueous solution promotes the solar-powered production of hydrogen [84,85,86,87,88,89,90]. Zhou et al. identified Cs₂SnI₆ perovskites combined with SMA Pt as a promising candidate for the treatment of HER [91]. The Pt-I₃ sites were proposed to have a strong metal-support interaction effect, enhancing the photocatalytic performance of HER. Wu et al. also utilized a single Pt atom on a formamidinium (FA)-based lead perovskite containing Br and I atoms [81]. The FA perovskites have a small band gap and halide mixture, which increases their stability and light absorption, as illustrated in Figure 7c. Pt on FA perovskites exhibited a $\Delta G_{H*}$ of ~0.04 eV, which is suitable for the adsorption of hydrogen atoms and the subsequent release of hydrogen gas. Approximately 700 μmolh⁻¹ of hydrogen was produced. From light-harvesting materials, OHP with SMA could be a candidate for hydrogen production and other photocatalytic applications.

In addition to conventional semiconductors, the 2D structure, high stability, and catalytic properties of covalent organic frameworks (COFs) make them attractive for photocatalyst systems [92,93,94,95,96,97,98,99,100]. High catalytic performance is exhibited by COFs containing transition metals when reducing the toxic organic chemical CO₂ or producing hydrogen. For example, Zhong et al. synthesized single Ni site COFs for CO₂ and H₂ production [95]. The COFs were produced using a solvothermal method involving 1,3,5-triformylphloroglucinol and 5,5′-diamino-2,2′-bipyridine, resulting in a material that is stable and more selective for CO₂ reduction than HER. In COFs, the selected metal and framework play a greater role in determining their catalytic properties. Dong et al. used a Pt single atom in β-ketoenamine-linked COFs (Pt-SA-TPa-1-COFs), as shown in Figure 7d,e [82]. The 2D Pt-SA-TPa-1-COFs exhibit numerous active sites and a low band gap of ~2.0 eV, resulting in a strong absorption of visible light from ~550 nm to the near IR band. Consequently, the calculation of $\Delta G_{H*}$ reveals a low energy of –0.092 eV, activating the hydrogen adsorption property and encouraging the release of H₂ following production. Pt-SA-TPa-1-COFs can generate hydrogen at a rate of 719 μmolh⁻¹ g⁻¹. With an SMA on the surface, the development of semiconductors is accelerating. Accordingly, numerous semiconductor laboratories are employing this approach to identify the most effective catalyst activator for OER, HER, CO₂ reduction, and organic reduction. They can be metal oxide material, MOF, metal sulphides, metal selenides, etc. [50,101,102,103,104,105,106].

To evaluate the performance of different materials,

Table 2. Comparison of various kinds of catalyst with/without single transition metal atom for hydrogen evolution reaction.

Metal Atom	Host Materials	Photocurrent (μA cm–2)	H2 Evolution Rate (mmol g−1·h−1)	Reference
Co	TiO2 nanobelt	~83	~0.677	[65]
-		~28	~0.0217
Co	TiO2 nanosheets (NSs)	n/a	~2.9	[107]
Ni		n/a	~1.0
-		n/a	~0.04
Cu	TiO2 (derived from MIL125-Ti)	~3.0	~101.7	[47]
Pt		n/a	~95.0
-		~1.0	~4.2
Pt	g-C3N4 NSs	n/a	~0.042	[74]
-		n/a	~0.001
Co	P-doped C3N4 ultrathin NSs	6.0	~3.7	[75]
-		4.0	~0.4
Pd	Carbon deficient g-C3N4	n/a	~2.8	[108]
-		n/a	~0.115
V	Polymeric C3N4	4.6	~5.0	[55]
-		1.7	~1.5
Ni	Sulfur-vacancy-enrich ZnI2S4	~22.5	~0.089	[49]
-		~10.0	~0.04
Pt	ZnI2S4 NSs	~17.0	0.35	[80]
-		~4.0	0.02
-	CuInS2/ZnIn2S4	~0.75	~0.34	[109]
Pt	FAPbBr3−xIx	~12.0	~0.7	[81]
-		~3.0	~0.05
-	Cs3Bi0.6Sb1.4I9	~20	n/a	[84]
Pt	TpPa COF	n/a	~0.72	[82]
-		n/a	0.015
-	TP-BDDA COF	n/a	0.32	[96]
-	TP-DTP COF	n/a	0.03

shows the summary of the single atom photocatalyst for hydrogen evolution reaction with various metals and host materials. It can be clearly recognized that in each host material, the different transition metal shows the different behaviour. In general, the presence of transition metal can improve the catalytic properties of host material. This is because free electron can be easily generated and transferred from host material to give it to H_ads. Pt is the most common metal for enhancing catalytic performance, due to its natural catalytic properties [74]. However, the abundance and price of Pt limits its commerciality. Therefore, Cu, Ni, and Co are the next candidates for Pt replacement [47,107]. Moreover, the host materials in nano scale, such as nanosheets, nanobelts, or particles, are generally utilized. It is worth noting that a high surface area is a prerequisite for a light harvester. In Table 2, the TiO₂ is derived from MIL125-Ti structure, which can achieve a high surface area from the collapse of this framework. Therefore, it can boost the hydrogen production up to ~100 mmol g⁻¹ h⁻¹ [47]. Moreover, the use of Cu in TiO₂ shows a better performance in comparison to Pt. This indicates that the suitable structure and metal can generate a synergistic effect for their catalytic properties. Therefore, host materials in the right form with metals are promising to be a future active material for hydrogen production.

5. Conclusions and Outlook

The use of single-atom photocatalyst has reached a milestone in terms of its physical and chemical properties, thereby enhancing the performance of the catalyst in HER. In addition, it allows any type of light harvester to become a photocatalyst for hydrogen production, CO₂ reduction, oxygen evolution reaction (OER), or organic oxidation. In order to obtain the superior photocatalyst, various heterogeneous or homogeneous methods are utilized. Various in/ex situ experimental/computational techniques, such as transmission electron microscopy, X-ray photoelectron spectroscopy, X-ray absorption spectroscopy, and DFT calculation, gradually reveal the structure and electronic properties. The mechanism of the photocatalyst is partially revealed. Recorded hydrogen production rates range from hundreds of micro to millimoles per hour. Accordingly, these are significant steps toward the future of producing clean, environmentally friendly fuels.

Despite numerous advancements, the SMA photocatalyst still faces a number of obstacles and requires additional research in areas such as selectivity, metal coordination investigation in practical experiments, stability, and controllable processes. For instance, computational research readily identified the transition metals (Mn, Fe, Co, Ni, Cu, Pd, and Pt) with distinct Gibbs free energy coordination for hydrogen absorption; however, there is still a gap where experimental work can determine the behaviour of transition metals with various coordination [34]. In addition, computational work is necessary to predict and comprehend the behaviour of single-metal atoms on various types of materials. It has been suggested that OHP are highly stable; however, there is a lack of experimental evidence demonstrating how long the material can maintain continuous photocatalytic performance. The combination of light-harvesting materials with varying absorption ranges can be an effective strategy for using solar light to drive photocatalytic reactions.

In addition, the OER reaction is also a boost-up key for hydrogen evolution reaction due to its counter reaction to HER [110,111]. As the balancing system, the generated electron–hole couple should be effectively separated and join in HER and OER. Therefore, the rate of the overall system is determined based on the rate of HER and OER. Unlike HER, OER works based on oxidation by the valence band of semiconductors [112,113]. Recently, the single atom photocatalyst also shows a great improvement for OERs; therefore, with the same concept, SMA photocatalysts are promising for practical and sustainable environmental devices [114,115]. Solar driven HER and OER by two separated electrodes will be the ideal system, efficiently producing H₂ and O₂. However, they both face the same hurdles to overcome: full investigation, material preparation, and theoretical predictions. The rapid rise of single atoms to the forefront of photocatalyst research demonstrates a promising future for the use of a brand new catalyst material to address a variety of issues, including energy, the environment, and healthcare.