Strategies to Enhance Interfacial Spatial Charge Separation for High‐Efficiency Photocatalytic Overall Water‐Splitting: A Review

Developing efficient photocatalysts for overall water‐splitting (OWS) has garnered considerable attention due to their potential in renewable energy conversion and storage. Enhancing the efficiency of interfacial spatial charge separation poses a key challenge in this field, as it plays a momentous role in the photocatalytic process. In this review article, a range of strategies aimed at improving interfacial spatial charge separation in photocatalysts for realizing high‐efficiency OWS are explored. To provide a comprehensive understanding, first the fundamentals of photocatalytic water‐splitting are introduced and the main bottlenecks in the reaction process, along with various charge separation and transfer mechanisms are identified. Subsequently, recent advancements and efforts in designing spatial charge separation at the interfaces of 0D, 1D, 2D, and 3D nanostructured photocatalysts are discussed. Finally, a summary is presented and a long‐term outlook for spatial charge separation in the field of OWS is offered. By consolidating the current state‐of‐the‐art research, this review highlights the key challenges and favorable circumstances for future advancements in the pursuit of efficient OWS.


Introduction
7][8][9][10] Solar energy, in particular, can be harnessed through solar cells or stored as chemical fuels, representing an enticing solution for energy storage and utilization.16][17][18] Modulating the bandgap for light absorption in the visible region is helpful in generating more electron and hole pairs.Most of these photogenerated charge-carriers recombine in bulk, and a small fraction reaches the catalyst surface for the catalytic reaction.Minimizing the recombination of these charge carriers is essential for promoting photocatalytic efficiency. [19]The charge-separation process involves the transfer of photoexcited electrons from the conduction band (CB) and holes from the valence band (VB) to the surface of the photocatalyst.If the electrons and holes recombine before participating in the chemical reactions, the efficiency of photocatalytic OWS will be significantly reduced.Therefore, enhancing charge separation has become a major research focus in the field of photocatalysis.[22][23] Several previous review articles have delved deeper into the OWS literature and reviewed the advancement in the field.26][27][28][29] Wang et al. summarized and proposed standard testing protocols in particulate OWS. [30]Bie et al. highlighted the issues and limitations of OWS such as unfavorable thermodynamics, backward reaction, dissolved oxygen and slow reaction kinetics. [31]Song and coworkers focused on the solar-to-hydrogen (STH) conversion efficiency, durability, and economic and environmental viability of the photocatalytic OWS system. [32]Lakhera et al. discussed the progress made in achieving high apparent quantum yield (AQY) and STH conversion efficiency of the photocatalysts. [2]Ng et al. underlined the challenges ahead for photocatalytic OWS, particularly the energy dissipation due to scattering effect in the reactor and fast charge recombination rate that limits the STH efficiency and hinders its commercial viability. [33]Furthermore, Juan Corredor et al. provided an integrated overview of the various photocatalytic, heterogeneous, homogeneous, and hybrid systems. [34]Several other excellent review articles discussed various aspects of OWS such as materials design, [35][36][37][38][39][40][41][42][43][44] cocatalysts, [45] band engineering, [46] photocatalytic reaction engineering, [47] large-scale solar hydrogen production, [48] and charge-separation mechanism, [49,50] to name a few.Efficient charge separation at the interface of the photocatalysts is essential for achieving higher STH conversion efficiency, and the interfacial contact area between the photocatalysts depends on the surface morphology.A comprehensive review of morphology-dependent spatial charge-carrier separation for OWS has been scarcely reported.
In this review, we focus on exploring the effect of interfacial spatial charge separation in photocatalytic OWS.We discuss various methods to enhance interfacial charge separation and provide an explanation of the principles, thermodynamics, and mechanisms underlying these methods.Specifically, we reviewed spatial charge separation and transportation in different nanostructures such as 0D/1D nanostructures, 2D/2D nanosheets, hybrid nanostructures (0D/1D/2D/3D), and 3D nanostructures.Furthermore, we also reviewed the concept of facet-dependent spatial charge separation.By summarizing and evaluating these findings, we aim to contribute to a deeper understanding of interfacial processes involved in photocatalytic OWS and offer insights into potential strategies for improving overall efficiency.

Photocatalytic OWS 2.1. Principle, Thermodynamics, and Mechanism
The photocatalytic OWS process involves using a semiconductortype catalyst that is excited by photons carrying an energy which is equivalent to or larger than semiconductor's bandgap energy.The catalyst undergoes three main steps in photocatalytic reactions: first, a photon with enough energy is absorbed, exciting a negatively charged electron from the VB which migrates to the CB, putting behind a positively charged hole in the VB.Second, the photogenerated electron and hole pairs separate and move toward the surface redox sites.Finally, the electronhole pairs accumulating on the redox sites are used in the photocatalytic hydrogen and oxygen production reactions shown in Figure 1a.Even though splitting water into hydrogen and oxygen has essential applications in energy production, it is also a thermodynamically unfavorable reaction.The Gibbs free energy change of this reaction determines whether it is spontaneous or not, and it is influenced by the enthalpy and entropy changes that occur during the process.The change in the Gibbs free energy for this reaction is given in Equation ( 1) and (2) [51,52] ΔG ¼ ΔH À TΔS where ΔH (þ285.8kJ mol À1 ) is the change in enthalpy, ΔS (þ163.3J K À1 mol À1 at 298 K) is the change in entropy, and lastly, T is the temperature (K).
As ΔG is positive, this reaction is thermodynamically unfavorable, requiring energy input to proceed [52] H The half-reactions are described as shown in Equation ( 4) and ( 5) To function as a water-splitting photocatalyst, a semiconductor must have a bandgap greater than or equal to 1.23 eV, the thermodynamic necessity for water splitting.However, the position of the conduction and VBs of the photocatalyst is equally vital.Specifically, the CB minimum is expected to have a more negative value than the reduction potential of H þ /H 2 (0 V vs normal hydrogen electrode), while the top of the VB must be more positive than the oxidation potential of O 2 /H 2 O (1.23 V vs normal hydrogen electrode). [29,45]This requirement ensures that the excited electrons can be transferred to the H þ /H 2 redox couple to produce hydrogen and that the holes can be transferred to the O 2 /H 2 O redox couple to produce oxygen during water splitting.Thus, the position of the conduction and VBs with the bandgap value is a critical factor for the suitability of a photocatalyst for water splitting shown in Figure 1b.

Quantum Efficiency and STH Conversion Efficiency
The commonly used measure of photocatalytic activity is the reaction rate with the unit of mol h À1 g cat

À1
. [24] However, the reaction rate does not always reflect the actual photocatalytic activity of the catalyst, as it can be influenced by factors such as the mass of the photocatalyst and illumination conditions.To overcome these limitations, alternative measures of photocatalytic activity have been proposed, including the apparent quantum yield (AQY) and STH conversion efficiency. [26,53]The AQY is defined as the ratio of number of reacted electrons (or holes) to the number of photons incident on the semiconductor.The STH conversion efficiency is defined as the ratio of output energy as H 2 to the incident light energy. [53]In one-step photocatalysts, two electrons are consumed to produce a molecule of H 2 , and in Z-scheme photocatalysts four electrons are consumed to produce a molecule of H 2 .The mathematical expressions for AQY and STH conversion efficiency are given in the following equations. [53,54]he AQY for H where the number of incident photons (N) is The STH conversion efficiency is calculated as These measures offer a more accurate evaluation of photocatalytic activity and are greatly used for photocatalytic hydrogen production or pure water-splitting research.[57][58] TOF shows how frequently a reaction is taking place on the active sites of the catalyst. [58]TON describes the maximum number of products evolved on an active site.TON and TOF are shown in Equation ( 15). [55]F ¼ dðTONÞ=dt (15)

Bottlenecks of Photocatalytic Water Splitting
In photocatalytic OWS reactions, several crucial limiting factors emerge.One of these factors pertains to the light-absorption ability, which is directly linked to the bandgap of the semiconductor photocatalyst.The quantity of surface-active sites on a photocatalyst serves as a determining factor in its photocatalytic activity.Furthermore, the stability of a photocatalyst significantly influences the scalability of the photocatalytic water-splitting reaction.Lastly, the OWS reaction is greatly hindered by high recombination rate of charge carriers, which are considered the primary bottleneck.This section aims to elaborate on these limiting factors. [59]

Light Absorption
Light absorption by a semiconductor material is critical for forming electron-hole pairs and subsequent photocatalytic water-splitting activity.Some of the photocatalysts and their band positions are displayed in Figure 2. The wavelengths at which a photocatalyst absorbs light and its band positions are essential factors to consider.[62][63] Therefore, it is crucial to develop strategies to enhance light absorption, such as doping the semiconductor [64,65] coupling the photocatalyst with a sensitizer [65,66] and creating defect sites, [67] such as oxygen vacancy sites, lattice defects, and inter-band states, [68,69] and utilization of up-conversion process. [2]These methods can improve light absorption and enhance the semiconductor's photocatalytic activity, ultimately leading to more efficient water splitting.
Several previous review articles have extensively discussed these strategies in detail and the readers are encouraged to refer to these articles for further insight. [2,24,25,29,34,36]

Surface Active Sites
The number of active sites on a catalyst's surface is typically smaller than the total available surface sites due to the limited catalytic activity of only a fraction of the catalyst surface. [70]atalytic surfaces often exhibit nonuniform activity, with specific arrangements of surface atoms or distinct chemical compositions determining the active sites. [71]Generally, a larger surface area promotes an increase in active surface sites. [72][75] For instance, Wu et al. reported the exfoliation of g-C 3 N 4 and the formation of defect sites through the irradiation of an aqueous suspension of catalyst with a femtosecond pulsed laser.The cyano (ÀC≡N) defects created by these laser pulses not only facilitate the anchoring of Pt atoms but also provide more active sites for surface reactions, thereby significantly suppressing the backward reaction of water splitting. [76]In another study, Bei-Bei Xu et al. designed a system with Co single atoms and PtCo alloy nanoparticles supported on g-C 3 N 4 nanosheets (CNN) for hydrogen and oxygen evolution reactions. [77]he Co single atoms and PtCo alloy nanoparticles served as highly active sites for hydrogen and oxygen evolution reactions, respectively.Similar observations have also been reported regarding the role of active sites in other photocatalytic applications, such as photocatalytic CO 2 reduction, [78] N 2 fixation, [79] and oxidation of benzylamine. [80]

Stability of the Photocatalysts
Photocatalytic stability poses a significant challenge in watersplitting reactions, particularly for long-term and large-scale applications.This is attributed to the degradation of photocatalysts under harsh reaction conditions and the corrosive nature of reactants.Over time, various degradation mechanisms occur during the water-splitting process, including photo-corrosion, surface contamination, and agglomeration, which progressively reduce the photocatalytic activity and efficiency of the catalyst. [81]mong these degradation mechanisms, photo-corrosion of the photocatalysts is a prominent process in acidic or alkaline solutions.It involves the oxidation or reduction of the photocatalyst surface. [82]For instance, bare CdS photocatalyst suffers from prompt recombination of photogenerated charge carriers and photocatalytic instability due to CdS photo-corrosion.The restrained photocatalytic stability of CdS is predominantly a consequence of photo-corrosion resulting from the oxidization of photogenerated holes.This instability arises from the slow hole-transfer rate, a key efficiency-restraining factor in CdS-based structures.Combined with the low oxidation kinetics of water or sacrificial agents, this leads to the accumulation of photogenerated holes in CdS and subsequent photo-corrosion. [83]Certain metal oxides, including Cu 2 O and ZnO, also exhibit limited stability in aqueous solutions. [84,85]When exposed to light irradiation, photogenerated holes become trapped on the surface of ZnO, facilitating the formation of O 2 molecules and the rapid release of Zn 2þ ions into the aqueous solution.Likewise, Cu 2 O undergoes oxidation to CuO under similar conditions.These photogenerated holes are also responsible for the decomposition of oxynitride materials and the oxidation of nitrogen anions into N 2 . [84]imilarly, perovskite materials, such as lead halide perovskite, [86] exhibit extreme sensitivity to various environmental factors, further complicating their practical use.The degradation of perovskites generally starts from the surface or from the interface and eventually spreads to the bulk catalyst.Additionally, the corrosion pathways can be influenced by the interfaces between charge extraction layers and perovskite materials. [87]nother factor contributing to reduced performance is surface contamination, caused by adsorbed species or impurities, which can obstruct active sites and alter the electronic structure of the photocatalyst.Furthermore, agglomeration of photocatalyst Figure 2. Bandgap energies and band positions for oxide, nitrate, and chalcogenide photocatalysts.Reproduced with permission. [171]Copyright 2021, Wiley-VCH GmbH.
particles decreases the surface area and adversely affects photocatalytic activity and efficiency.To address these issues, it is crucial to develop stable and long-lasting catalysts through surface modifications, such as applying protective coatings.

Mass-Transfer Limitations
Efficient mass transfer of reactants and products to and from the surface of the photocatalyst is a crucial factor in achieving high efficiency during photocatalytic reactions. [88]Heterogeneous photocatalytic reactions involve four steps for the transfer of active species and targeted reactants. [89]First, active radicals such as superoxide and hydroxyl radicals are generated.Second, the targeted reactive materials (e.g., organic substances in photodecomposition reactions) adsorb onto the surface of the photocatalyst, interacting with the active radicals.Third, oxidation products desorb from the catalyst surface into the reaction solution.Lastly, photogenerated reactive species desorb from the photocatalyst surface, shaping the desired materials.Consequently, limitations in mass transfer play a critical role in the overall photocatalytic activity.To enhance the mass transfer between reactive species and reactants, several strategies can be employed.[94][95] Furthermore, the structural design of the photocatalyst itself is crucial.Implementing porous photocatalysts or photocatalysts with nanowire arrays can increase the exposure of active sites, [92,[96][97][98][99] resulting in improved mass transfer and heightened photocatalytic activity.

Separation of Charge Carriers
The process of photocatalytic water-splitting begins with the absorption of light by the semiconductor, resulting in the formation of electron-hole pairs that participate in redox reactions.However, not all of these charge pairs are utilized due to recombination in both the bulk and on the surface of the photocatalyst.As the electrons and holes migrate to the surface, recombination may occur in bulk, and only a portion of the pairs that reach the redox-active sites are used in oxygen and hydrogen evolution reactions. [60,61][112] The various charge-separation mechanisms are discussed later.

Various Charge-Separation and Transfer Mechanisms
In a photocatalytic reaction, multiple steps take place, and one of the most significant factors that negatively impact photocatalytic activity is charge-carrier recombination.When a photon is absorbed, photogenerated electron and hole pairs can either travel to the surface of the photocatalyst and trigger redox reactions or recombine, leading to wasted energy.The Coulomb force, which relies on the magnitudes of electron and hole charges and their distance, drives recombination, as shown in Equation ( 16). [22]c ¼ k q e q h r 2 (16)   where F c is the Coulombic force, k is the Coulomb constant, q e and q h are the charge magnitudes of the electron and the hole, respectively, and r is the distance between the centers of the charges.Due to the exceedingly large Coulomb constant, electron-hole pairs recombine quickly, making it challenging to prevent recombination.Achieving efficient electron-hole pair separation relies heavily on catalyst design.Several strategies, including doping, cocatalyst use, and semiconductor heterojunction formation, have enhanced photocatalytic activity. Aong these strategies, forming a semiconductor heterojunction is particularly effective because it increases the distance between the electron and hole centers, decreasing the Coulombic force that drives recombination.[14,17,72,113,114] 4.1.Type I Heterojunction In a type I heterojunction system, the CB and the VB of the photocatalysts I (PC I) are lower and higher, respectively, compared to the CB and VB bands of the photocatalyst II (PC II).Consequently, when illuminated, electrons and holes gather up on the CB and VB of PC II.As both electrons and holes accumulate within the same photocatalyst, effective separation of electron-hole pairs is hindered as displayed in Figure 3a.
Additionally, since both reduction and oxidation reactions occur on the semiconductor with the lower redox potential, the redox ability of the heterojunction photocatalyst diminishes importantly. [17,22]

Type II Heterojunction
In a type II heterojunction system, both CB and VB levels of PC I are higher than that of PC II [22,115,116] .As a result, photogenerated electrons migrate to through PC II, while photogenerated holes drift to PC I under illumination, facilitating spatial separation of electron-hole pairs.However, similar to type I heterojunctions, the redox ability of the type II heterojunction system is weakened due to reduction occurring in PC II with a lower CB level and oxidation occurring in PC I with a higher VB level as shown in Figure 3b. [115]Consequently, this system fails to utilize the high redox abilities of PC I and PC II.

Z-Scheme Heterojunction
The Z-scheme photocatalyst system, which mimics artificial photosynthesis, is considered a promising strategy for enhancing photocatalytic activity.A Z-scheme heterojunction can manifest in three distinct configurations: the traditional Z-scheme, the all-solid-state Z-scheme, and the direct Z-scheme heterojunction.The original Z-scheme concept, attributed to Bard, [117] involves the interconnection of two semiconductors via a shuttle redox mediator, such as Fe 3þ/ Fe 2þ , IO 3À/ I À , or I 3À/ I À . [115,117]hotogenerated holes situated in the VB of the first photocatalyst (PC I) interact with electron donors (D), like Fe 3þ , leading to the generation of corresponding electron acceptors (A), such as Fe 2þ .Simultaneously, photogenerated electrons located in the CB of the second photocatalyst (PC II) engage with electron acceptors, resulting in the creation of electron donors.Consequently, this process leads to the accumulation of photogenerated electrons in the CB of PC I and holes in the VB of PC II.These charge accumulations facilitate reduction and oxidation reactions, respectively.In the second generation, termed the all-solid-state Z-scheme mechanism, upon irradiation, photogenerated electrons within the CB of PC II migrate toward a solid conductor, subsequently transferring to the VB of PC I. [115] In the third generation, designated as the direct Z-scheme mechanism, photogenerated electrons originating from the CB of PC II directly transit to the holes present in PC I through their contact interface.This process results in the buildup of electrons on PC I and holes on PC II.The concept of a direct Z-scheme photocatalyst was initially introduced by Kudo et al. [60] for OWS, by Wang et al. [118b] for sacrificial hydrogen evolution, and was later explored by Yu et al. [118c] for photocatalytic formaldehyde degradation. [118]The direct Z-scheme heterojunction system resembles of the type II heterojunction system; however, their electron and hole migration mechanisms differ significantly.The Z-scheme system incorporates an electron-hole migration pathway resembling the letter "Z."As previously mentioned in the context of type II heterojunctions, electrons accumulate on the less negative CB, while holes accumulate on the less positive VB.Consequently, the high redox potentials of PC I and PC II are not utilized in photocatalytic reactions.Conversely, in Z-scheme heterojunctions, electrons accumulate on the more negative CB, and holes accumulate on the more positive VB, allowing for the utilization of the high redox abilities of PC I and PC II as shown in Figure 3c,d. [113,114,119,120]Moreover, the migration of electrons to the more negative CB and more positive VB reduces the electrostatic attraction between electrons and holes.However, compared to type II heterojunction, available charge-carrier number in Z-scheme shrinks due to interface electron-hole recombination at the junction. [121,122]4.Metal-Semiconductor Barrier (Schottky) Creating a metal-semiconductor (M-S) junction is an effective approach to establishing a space-charge-separation region known as the Schottky barrier. [17]At the interface of the two semiconductors, electrons flow from the higher to the lower Fermi level, aligning the Fermi energy levels. [115,119]In the common case where the work function of the metals is higher than the work function of n-type semiconductors, the Fermi energy levels were adjusted by the electron flow from the semiconductor into the metal. [123,124]The formation of the Schottky barrier brings about accumulation of photogenerated electrons on the metal and accumulation of holes on the semiconductor as depicted in Figure 3e.Moreover, the Schottky barrier acts as an electron trap which prevents charge recombination and thereby enhances the photocatalytic activity of the photocatalyst. [17]5.S-Scheme Heterojunction The effectiveness of the type II heterojunction system is limited by its weak redox capability, primarily due to the accumulation of electrons in the CB of PC II and holes in the VB of PC I.This leads to a diminished redox potential, posing a challenge for efficient charge transfer.Although the Z-scheme systems (traditional Z-scheme, all-solid-state Z-Scheme) offer advantages over the type II configuration, they also introduce a higher likelihood of alternative pathways for charge transfer, beyond the initially proposed mechanisms.The involvement of redox mediators can further complicate matters, potentially consuming the pool of high-energy charge carriers.Additionally, the stability of these mediators is often pH-dependent and can influence their reliability.Within the all-solid-state Z-scheme setup, the efficiency of charge transfer through the solid conductor is contingent upon the extent of surface contact between the two semiconductors and the solid conductor itself.
In light of these considerations, Junwei Fu et al. introduced a new concept known as the step-scheme (S-scheme)staggered heterojunction system. [125]The S-scheme photocatalytic system comprises two essential components: an oxidation photocatalyst (OP) and a reduction photocatalyst (RP). [126]The CB and VB positions of RP are higher than that of OP while the work function is smaller than the work function of OP.When RP and OP come into contact, electrons transfer from RP to OP; and therefore, an electron depletion layer near the interface of RP was created spontaneously.Simultaneously, an electron accumulation layer comes into being near the interface of OP.Consequently, OP becomes negatively charged, while RP becomes positively charged.Although, the Fermi energy levels are inclined to preserve the same level on both sides when the interfaces of semiconductors are in contact, however, due to the bending of the bands, electrons from the CB of OP and holes in the VB of RP recombine.This characteristic curved shape gives the S-scheme heterojunction its name.The photogenerated electrons on the OP tend to recombine with the holes on RP due to the Coulombic force between electrons and holes.Hence, useless electrons and holes are phased out by the recombination, while the potent electrons in the CB of RP and the holes in the VB of OP are secured for engaging in photocatalytic reactions as illustrated in Figure 3f.[129] Despite its resemblance to the type II heterojunction, the S-scheme system distinguishes itself by the accumulation of electrons in the CB of the RP and holes in the VB of the OP.This unique arrangement results in significantly heightened redox ability, setting it apart from the type II heterojunction where such accumulations typically lead to weaker redox capabilities.

Spatial Charge-Carriers Separation
Fabricating heterostructures has emerged as a crucial approach to enhance the separation of photogenerated electron and hole pairs in photocatalytic hydrogen production.When two semiconductors with aligned conduction and valence energy levels come into contact, the resulting heterojunction at the interface promotes spatial charge separation, thereby reducing the recombination rate of photogenerated electrons and holes. [130]Various morphologies have been developed to achieve spatially separated charge carriers, including 0D/1D nanostructures; 2D/2D nanosheets; hybrid nanostructures combining 0D, 1D, and 2D components; 3D nanostructures, and single photocatalysts with distinct redox facets.Additionally, cocatalysts have been employed to facilitate the oxygen and hydrogen evolution reactions, directing electrons and holes toward the redox surfaces to enhance charge separation.Catalysts' structural and surface engineering plays a vital role in enhancing their photocatalytic performance.Different strategies have been employed to control photocatalysts' structure and surface characteristics.For instance, the synthesis method can be modified to obtain AgVO 3 nanowires, [131,132] nanobelts, [133] and nanoparticles. [134]he following sections will discuss some of these structures and their corresponding charge-transfer mechanisms.

0D/1D Nanostructures
[137][138] In comparison to nanoparticles and bulk materials, 0D/1D nano-structural materials offer distinct advantages as potential photocatalysts. [139]To illustrate this, Liu et al. synthesized a 1D CdS/TiO 2 core-shell nanocomposite photocatalyst, emphasizing the remarkable benefits of its 1D geometry.These advantages include fast and efficient electron transport over long distances, enhanced light absorption and scattering, a larger specific surface area, and increased pore volume. [140]n a similar way, Dong et al. fabricated a 1D Ta 3 N 5 nanorod/ BaTaO 2 N (BTON) nanoparticle hybrid structure using a one-step ammonia thermal route and investigated the charge-separation dynamics within this intriguing 0D/1D nanostructure. [130]igure 4d illustrates the spatial separation of photogenerated electron and hole pairs between the 1D Ta 3 N 5 nanorods and 0D BTON nanoparticles, elucidating the occurrence of O 2 evolution on BTON and H 2 evolution on Ta 3 N 5 .Impressively, the Ta 3 N 5 /BTON photocatalyst demonstrated robust production rates of 4.8 μmol h À1 for H 2 and 2.4 μmol h À1 for O 2 , while exhibiting remarkable stability over five cycles (Figure 4a-c). [130]imilarly, Li et al. prepared 1D gallium nitride (GaN) nanorods via the top-down etching of n-GaN films with selfassembled silica spheres as the mask, paying particular attention to their nonpolar and polar surfaces. [141]Since GaN nanorods had surfaces with different polarities, the spatially separated OER and HER cocatalysts were synthesized by photo-deposition method.MnO x and CoO x were deposited by photooxidation deposition method and noble metals such as Ag, Rh, and Au were deposited by photoreduction deposition method.The presence of surface polarity served as a catalyst for the spatial separation, resulting in a photogenerated charge-separation efficiency enhancement from approximately 8% to over 80%.Furthermore, by strategically depositing reduction and oxidation cocatalysts in spatial arrangements (Figure 4e,f,g,h), the researchers succeeded in further augmenting the quantum efficiency from 0.9% to an impressive 6.9%, offering a captivating glimpse into how the 1D structure of GaN actively contributes to the spatial segregation of electrons and holes.This segregation  and d) a schematic illustration depicts the spatial separation of photogenerated electrons and holes between the 1D Ta 3 N 5 nanorod and 0D BTON nanoparticle.Reproduced with permission. [130]Copyright 2019, Wiley-VCH GmbH.e) A top-view SEM image of Rh-CoO x /GaN system, showcasing the surface morphology of the material.f ) A cross-section view of the SEM image.g) The photocatalytic OWS activity of GaN nanorod arrays without and with the cocatalysts Rh and CoO x , and h) a schematic illustration of the spatial separation of electrons and holes on GaN nanorod arrays.Reproduced with permission. [141]Copyright 2020, Wiley-VCH GmbH.
allows for the selective deposition of oxygen evolution cocatalysts on oxygen evolution sites, while hydrogen evolution cocatalysts are exclusively deposited on hydrogen evolution sites.This extraordinary selectivity is achieved by harnessing the 1D structure adorned with 0D particles (Table 1).

2D/2D Nanosheets
The 2D/2D nanosheets have gained significant attention in photocatalytic OWS area due to their uncommon electronic and optical properties, as well as the synergistic effects resulting from the heterojunctions they form.The coupling of 2D ultrathin nanosheets, such as TiO 2 /g-C 3 N 4 , [142] rGO/Bi 2 WO 6 , [143] and BiOCl/g-C 3 N 4 [144]   heterostructures, has shown enhanced photocatalytic activity.This can be associated with the increased surface area available for contact with reactants and other semiconductors, shorter transmission paths for photoinduced charges, and effective separation of photogenerated charge carriers.Recently, Che et al. reported a 2D C 3 N 4 -based in-plane heterostructure (C ring /g-C 3 N 4 ) to be used as a photocatalyst in OWS reaction. [145]This heterostructure can realize spatially separated charge-carrier transfer swiftly by increasing the charge diffusion length and charge-carrier life time by the help of the local in-plane π-conjugated electric field shown in Figure 5a.In Figure 5b, C ring /g-C 3 N 4 photocatalyst showed H 2 and O 2 production rates of 11.13 and 5.52 μmol h À1 , respectively, in which H 2 production rate was 10 times higher than that of single g-C 3 N 4 .The increase in photocatalytic activity was attributed to the in-plane electron and hole migration to C ring /g-C 3 N 4 from g-C 3 N 4 explained in Figure 5a-c).Moreover, C ring /g-C 3 N 4 photocatalyst had a noteworthy quantum yield of 5% at 420 nm wavelength.
Likewise, Zhu et al. synthesized black phosphorus (BP) and BiVO 4 nanosheets separately and achieved a 2D/2D heterostructure by adding BiVO 4 to a BP/N-methyl-2-pyrrolidone dispersion for self-assembly of the heterostructure. [146]The high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) image of the BP/BiVO 4 (BPB) nanostructure is shown in Figure 5d, while Figure 5e explains the chargetransfer mechanism between BP and BiVO 4 nanosheets.The Z-scheme charge-transfer mechanism between these materials facilitated spatial separation of the photogenerated charges, with BP acting as the hydrogen evolution catalyst and BiVO 4 as the oxygen evolution catalyst.By optimizing the catalyst amounts, stable hydrogen and oxygen production from overall water splitting was attained for three cycles (Figure 5f,g).
In another study, She et al. synthesized an α-Fe 2 O 3 /g-C 3 N 4 (FC) hybrid material by in situ production of carbon nitride. [147]he α-Fe 2 O 3 nanosheets, synthesized via a hydrothermal method at 180 °C for 12 h, exhibited a 2D structure.Melamine was added to an aqueous dispersion of α-Fe 2 O 3 in a crucible under and was heated to 550 °C for 4 h and α-Fe 2 O 3 /multilayered g-C 3 N 4 was obtained.This α-Fe 2 O 3 /multilayered g-C 3 N 4 was further calcined at 550 °C for a certain time to get final α-Fe 2 O 3 /2D g-C 3 N 4 .Figure 5h illustrates the 2D/2D structure of the α-Fe 2 O 3 / g-C 3 N 4 hybrid material.The hybrid material demonstrated inadequate recombination of photogenerated charge carries, as evidenced by photoluminescence (PL) intensity measurements.Additionally, DMPO spin-trapping ESR spectra revealed that CNNs did not produce OH radicals, while α-Fe 2 O 3 nanosheets H 2 and O 2 evolution from overall photocatalytic seawater splitting.
did not produce superoxide.The electron-transfer mechanism of the hybrid structure is depicted in Figure 5i,j, indicating that CNNs functioned as hydrogen evolution catalysts, while α-Fe 2 O 3 nanosheets acted as oxygen evolution catalysts.The synergistic effect between these 2D/2D materials contributed to increased quantum efficiency and overall photocatalytic activity for water splitting.Similarly, Zhao et al. reported the synthesis of 2D/2D CNNs/ B-doped CNNs through electrostatic self-assembly. [148]Boron doping induced a negative shift in both the conduction and VB positions of g-C 3 N 4 .Upon contact, a Z-scheme heterojunction was formed between these two semiconductors, with CNN functioning as a hydrogen evolution photocatalyst and boron-doped g-C 3 N 4 nanosheets (BCNN) as an oxygen evolution photocatalyst.The highest amounts of O 2 and H 2 produced from OWS were observed at a mass ratio of CNN:BCNN of 1:1.The synthesis of ultrathin nanostructures and the strong interfacial interaction, along with the Z-scheme heterojunction, facilitated efficient charge-carrier separation and transfer, resulting in an STH and AQY efficiency of 1.16% and 11.90% under 1 sun illumination, respectively.
In addition to experimental studies, theoretical studies about 2D/2D structures in OWS reaction were published. [149,150]omputational calculations showed that 2D structures are promising capability of hydrogen production from pure water splitting (Table 2).

3D Nanostructures
The 3D bulk structures, such as complex hierarchical structures, nanoparticles, and microspheres, have been specifically designed for spatial charge separation.Among these structures, 3D hollow structures possess unique architectural characteristics that enable efficient spatial charge separation.Moreover, they have a positive impact on enhancing the light-absorption properties of photocatalysts and improving their efficacy in solar energy conversion applications.This enhancement can be attributed to the occurrence of multiple light reflections. [151,152]dditionally, the utilization of 0D nanoparticles as the basic building units of hollow spheres helps reduce the diffusion length of photogenerated carriers.Another advantage of Reproduced with permission. [145]Copyright 2017, American Chemical Society.d) HAADF-STEM image of the BPB heterostructure, providing a detailed view of the structure.e) The Z-scheme photocatalytic water splitting using BPB under visible-light irradiation.f ) The effect of the ratio of BP in BPB/Co 3 O 4 on photocatalytic OWS under visible light (λ > 420 nm), highlighting the impact of composition on the photocatalytic performance.g) The photocatalytic stability test on BPB/Co 3 O 4 under visible light (λ > 420 nm).Reproduced with permission. [146]Copyright 2017, Wiley-VCH GmbH.The α-Fe 2 O 3 /g-C 3 N 4 (FC) hybrids: h) an HRTEM image of the FC hybrid.i) Photocatalytic OWS over FC, depicting H 2 and O 2 evolution rates using RuO 2 /FC/Pt under λ > 400 nm light irradiation.j) The Z-scheme mechanism in FC photocatalyst, illustrating the charge-transfer and separation processes involved in the photocatalytic water-splitting mechanism.Reproduced with permission. [147]Copyright 2017, Wiley-VCH GmbH.
employing hollow spheres as photocatalysts is their ability to fabricate heterojunctions, which further enhances charge-transfer and separation efficiency.For example, Wei et al. fabricated hollow multishelled structures (HoMSs) of La-and Rh-co-doped SrTiO 3 (STO:La/Rh), coupled with BiVO 4 nanosheets, by mixing and sintering (STO:La/Rh) and BiVO 4 nanosheets with the molar ratio of 1:1.The calcination was carried in a furnace under 400 °C for 4 h under Ar atmosphere.The transmission electron microscopy (TEM) image of the final photocatalysts are shown in Figure 6a. [153]The hollow-shelled structure enhanced the lightabsorption capacity and charge-separation efficiency, as depicted in Figure 6c,d.Furthermore, the amalgamation of STO:La/Rh HoMSs and BiVO 4 (BVO) nanosheets demonstrated a Z-scheme heterojunction and enabled OWS.A cocatalyst of 1 wt% Pt was used.The steady evolution of hydrogen and oxygen was observed under visible light and simulated sunlight, as illustrated in Figure 6b.The STH efficiency of the STO:La/Rh HoMS-BVO structure with a double shell was calculated as 0.08%.
In another study, Wei et al. constructed hollow HoMSs of SrTiO 3 /TiO 2 to enhance light-absorption properties and improve solar energy conversion applications. [153]SrTiO 3 /TiO 2 (ST) heterogeneous HoMSs were obtained through hydrothermal crystallization of SrTiO 3 on hollow TiO 2 surface of the multishells.This resulted in the full coverage of SrTiO 3 on the TiO 2 surface and the formation of ST heterojunctions, as depicted in Figure 6e.The type II heterojunction between SrTiO 3 and TiO 2 reduced the recombination of photogenerated electron and hole pairs, facilitating their spatial separation, as shown in Figure 6f-h.Another example for 3D structures is ZnTiO 3Àx N y hollow spheres with spatially separated hydrogen and oxygen evolution cocatalyst, reported by Wei et al., displayed in Figure 6i. [154]Pt/ZnTiO 3Àx N y /RhO x photocatalyst showed stoichiometric H 2 /O 2 production ratio from pure water splitting and ZnTiO 3Àx N y with spatially separated cocatalyst showed 7.0 and 3.5 μmol h À1 H 2 and O 2 evolution rate, respectively, which is tremendously higher compared to that of Pt/RhO x /ZnTiO 3Àx N y photocatalyst as illustrated in Figure 6k,l.PL-intensity results proofed that charge recombination of spatially separated photocatalyst diminished as shown in Figure 6j and boosted number of charge carriers improved pure water-splitting activity.
Pan et al. also explained that spatially separating the hydrogen evolution and oxygen evolution sites using CdS and g-C 3 N 4 , respectively, improved charge-carrier separation. [155]They synthesized a MnO x /g-C 3 N 4 /CdS/Pt hollow core-shell heterojunction, using a continuous chemical annealing and photoreduction method.This unique morphology promoted the separation and transport of photogenerated electron and holes through the appropriate potential, increasing the active sites for hydrogen evolution and oxygen evolution reactions.It resulted in improved photocatalytic efficiency and stability (Table 3).

Facet-Dependent Charge Separation
Spatial charge separation can occur on different redox facets of a single material.Some of the examples of such materials include TiO 2 , [156,157] BiVO 4 , [158,159] and SrTiO 3 . [160]The presence of distinct redox facets aids in the separation of photogenerated electron and hole pairs, thereby curtailing charge recombination and enhancing the photocatalytic activity of the single crystal.Additionally, the deposition of cocatalysts for oxygen evolution and hydrogen evolution on these distinct redox facets can further enhance the photocatalytic activity.For instance, Mu et al. published an OWS study of SrTiO 3 nanocrystals with different number of the exposed facets. [160]The SrTiO 3 nanocrystals were prepared by hydrothermal synthesis method.They reported that reduction and oxidation active sites can be spatially separated only on the anisotropic facets.Therefore, photo-deposition of OEC and HEC is spatially separated only on 18-facet SrTiO 3 nanocrystals with spatially separated and Co 3 O 4 cocatalysts (Figure 7a).Similarly, Yu et al. synthesized a single-crystalline Bi 2 YO 4 Cl with redox facets to enhance photocarrier separation for hydrogen production from pure water splitting. [161]The platelike single crystals with exposed (001) and ( 100) facets, possessing different energy states, created an internal built-in electric field capable of separating photocarriers and directing their flow toward the corresponding HER and OER facets (Figure 7d-f ).Bi 2 YO 4 Cl achieved an apparent quantum efficiency of 7.51% for O 2 and 2.52% for H 2 evolution rates at 420 AE 20 nm, owing to its facet-aided charge-separation mechanism.Furthermore, Wan et al. reported a single PbTiO 3 crystal with HER and OER facets attuned by oxygen vacancies and ferroelectric polarization. [162]It was elucidated that intrinsic defects, particularly oxygen vacancies, can shield the polarization.Figure 7g,h illustrates how polarization aids in the separation of charge carriers to the HER and OEC facets.Figure 7i depicts the photodeposition of HEC and OEC on different redox facets of the single crystal.The authors reported an apparent quantum yield of 0.025% at 435 nm using PbTiO 3 /Rh/Cr 2 O 3 , demonstrating H 2 and O 2 evolution from OWS.The photocatalytic OWS performances of STO:La/Rh-BVO photocatalyst system.c) Fluorescence spectra of 3 Â 10 À3 m terephthalic acid solution produced by single photocatalyst and heterojunction systems after light irradiation with λ > 400 nm and d) the proposed Z-scheme charge-transfer mechanism for the composite, shedding light on the underlying processes.Reproduced with permission. [153]opyright 2021, Wiley-VCH GmbH.e) The TEM-energy-dispersive X-ray (EDX) mapping of a slice of SrTiO 3 /TiO 2 (ST) HoMs after a 12 h hydrothermal reaction.f ) Photoluminescence (PL) spectra, highlighting the emission properties of the samples.g) Photocatalytic OWS performances of ST HoMSs samples with different reaction times.h) The possible band structure diagram for ST HoMSs.Reproduced with permission. [172]Copyright 2019, Wiley-VCH GmbH.i) Charge-transfer and separation mechanism on Pt/ZnTiO 3Àx N y /RhO x photocatalyst is illustrated.j) The PL spectra of the samples, aiding in the understanding of their luminescent characteristics.k) The OWS measurements of spatially separated OEC and HEC on Pt/ZnTiO 3-x N y /RhO x and l) the OWS measurements of Pt/ZnTiO 3Àx N y /RhO x with OEC and HEC on the same surface (nonspatially separated OEC and HEC).Reproduced with permission. [154]Copyright 2021, Wiley-VCH GmbH.
In another report, Takata et al. reported a photocatalyst of Al-doped strontium titanate (SrTiO 3 : Al) with OER and HER facets, exhibiting an almost unity quantum efficiency. [163]Selective photo-deposition of the hydrogen evolution catalyst, Rh/Cr 2 O 3 , and the oxygen evolution catalyst, CoOOH, on different crystal facets of SrTiO 3 : Al particles spatially separated the hydrogen and oxygen evolution reactions.This spatial separation of HER and OER facets advances forward charge transfer while inhibiting the backward charge transfer, so that the quantum efficiency for photocatalytic OWS is maximized.The H 2 and O 2 evolution rates by photocatalytic OWS on SrTiO 3 :Al loaded with different cocatalysts were measured.Between SrTiO 3 :Al loaded with Rh/Cr 2 O 3 through two-step photo-deposition and SrTiO 3 :Al loaded with Rh/Cr 2 O 3 /CoOOH through three-step photodeposition, and SrTiO 3 :Al loaded with Rh/Cr 2 O 3 through the co-impregnation method, the second one showed the best performance.These results demonstrate that spatially separated HEC and OEC on different facets of a single crystal are crucial for achieving high efficiency and activity in photocatalysis (Table 4).

Hybrid Nanostructures (0D/1D/2D/3D)
Hybrid nanostructures are composed of a combination of materials in various dimensions, including 0D, 1D, 2D, and 3D.This combination of materials aids in the separation of photogenerated charge carriers.Specifically, the incorporation of 0D materials as electron mediators helps to reduce the recombination of separated electrons and holes.For instance, Wang et al. synthesized a Z-scheme heterostructure coupling sulfur-deficient ZnIn 2 S 4 (ZIS) and oxygen-deficient WO 3 (WO) using 2D surface-carbonized wood (C-wood) layer as an electron-transfer bridge, depicted in Figure 8a,b. [164]They used pine wafers which are carbonized by treating on the alcohol flame (C-wood) for 2 min, as electron-transfer layer and then Pt/ZIS and CoO x /WO were spin-coated on C-wood.The rates of H 2 and O 2 evolution were approximately 169.2 and 82.5 μmol h À1 , respectively, as depicted in Figure 8c.The STH efficiency was calculated to be 1.52%.The Z-scheme facilitated charge transfer between ZIS and WO, resulting in the spatial separation of charge carriers.Additionally, C-wood functioned as an electron-transfer bridge, enhancing both charge transfer and separation, as illustrated in Figure 8d.Furthermore, the quantum yield efficiency (AQY) of ZIS-WO/C-wood (ZWC-wood) was determined to be 22.63% at 380 nm.
In another study by Gogoi et al., a step-scheme CdS/MnO x -BiVO 4 (CMB) photocatalyst was prepared, where hydrogen and oxygen evolution cocatalysts were spatially deposited on a 3D decahedron BiVO 4 surface. [165]The 0D MnO x and 1D CdS nanowires were used for this purpose, as shown in Figure 8e-g.The CMB photocatalyst exhibited significantly higher OWS activity compared to the individual materials.The production rates of H 2 and O 2 were measured to be 1.01 and 0.51 mmol g À1 h À1 , respectively, with an AQY of 11.3%, as illustrated in Figure 8h.Neither BiVO 4 nor CdS alone exhibited hydrogen or oxygen production from pure water without sacrificial agents.The enhanced OWS activity of the composite material was attributed to the existence of the oxygen vacancies and effective charge-separation properties and transfer kinetics.[196]   380 nm α-Fe 2 O 3 /RP 3D microstructure 0.32 0.14 S-scheme 0.011% [197]  Similarly, Guan et al. synthesized a 2D NiBHT material deposited on Al-doped SrTiO 3 . [166]Even though single SrTiO 3 : Al did not show any OWS activity, after deposition of NiBHT and CoO x which were employed as HEC and OEC, respectively, NiBHT/SrTiO 3 :Al/CoO x showed around 7.8 μmol h À1 H 2 and 3.9 μmol h À1 O 2 evolution from OWS as demonstrated in Figure 8i.Charge-transfer mechanism displayed in Figure 8l; photogenerated electrons accumulated on NiBHT and holes migrated to CoO x through SrTiO 3 :Al photocatalyst.Spatially separation of electrons and holes helped to boost the activity as evidenced in Figure 8j,k.The apparent quantum efficiency of NiBHT/SrTiO 3 :Al/CoO x heterostructure was calculated as 6.5% at 350 nm.
In addition, Dai et al. reported a composite material consisting of g-C 3 N 4 /ITO/Co-BiVO 4 (CICB). [167]In this composite, g-C 3 N 4 served as a hydrogen evolution catalyst, Co-doped BiVO 4 acted as an oxygen evolution catalyst, and ITO served as conductive mediators.The OWS activity of the CICB composite, without using any sacrificial agents, was calculated to be four times higher than that of g-C 3 N 4 /Co-BiVO 4 .Under full arc irradiation, the H 2 and O 2 evolution rates were measured to be 95.41 and 40.23 μmol g À1 h À1 , respectively.The STH efficiency of the CICB composite was found to be 0.028%.Importantly, the individual hydrogen and oxygen evolution rates from Co-BiVO 4 and g-C 3 N 4 alone were lower than those of the g-C 3 N 4 /Co-BiVO 4 composite.This demonstrates the positive effect of spatially separated charge carriers on OWS, as well as the benefits of using an electron mediator for the spatial separation of electron-hole pairs.Another example of hybrid structure is BiVO 4 /carbon dots (CDs)/CdS (BCC), as reported by Wu et al.In this case, a Z-scheme photocatalyst, termed BCC (BiVO 4 /CDs/CdS), was developed, where CDs served as solid-state electron mediators. [168]This hybrid photocatalyst exhibited outstanding photocatalytic activity and stability for OWS under visible-light irradiation.Among the different compositions, BCC50 (BiVO 4 /CdS = 50%, mass ratio) showed the uppermost photocatalytic OWS activity, with a H 2 evolution rate of 1.24 μmol h À1 and an O 2 evolution rate of 0.61 μmol h À1 .Copyright 2016, The Royal Society of Chemistry.d) 3D particle morphology and crystal orientation of Bi 2 YO 4 Cl, e) selectively deposition of OEC and HEC in Bi 2 YO 4 Cl, and f ) the photocatalytic H 2 and O 2 production rates of Bi 2 YO 4 Cl photocatalysts synthesized by different methods under visible-light irradiation (λ ≥ 420 nm).Reproduced with permission. [161]Copyright 2023, American Chemical Society.g) Schematic illustration ferroelectric polarization effect on electron and hole separation on PbTiO 3 and h) schematic image of single-domain PbTiO 3 before and after photochemical deposition of Au and MnO x .i) Finally, H 2 and O 2 evolution rate of PbTiO 3 /Rh/Cr 2 O 3 under simulated solar irradiation.Reproduced with permission. [162]Copyright 2021, Wiley-VCH GmbH.e) A schematic diagram of BiVO 4 /carbon dots (CDs)/CdS (BCC) heterostructure for OWS.f ) The proposed S-scheme chargetransfer mechanism for this heterostructure under visible-light irradiation, h) gas evolution rates for H 2 and O 2 from different photocatalysts, g) accompanied by an FESEM image of CBM.Reproduced with permission. [165]Copyright 2021, American Chemical Society.i) SEM image of NiBHT/SrTiO 3 :Al photocatalyst is presented, while j,k) H 2 and O 2 evolution rates by OWS using different hybrid structures.Finally, l) the charge-transfer and transfer mechanism on NiBHT/SrTiO 3 :Al/CoO x hybrid photocatalyst system.Reproduced with permission. [166]Copyright 2022, American Chemical Society.

Summary and Long-Term Outlook
In summary, this review article focuses on the importance of interfacial spatial charge separation in photocatalytic OWS.Photocatalytic OWS, which was first discovered by Fujishima and Honda in 1972, [169] has undergone extensive research in the last five decades, leading to significant advancements in our understanding of light-matter interaction and practical applications.Despite the progress made, several challenges persist, resulting in lower-than-anticipated efficiencies in AQY and STH.These challenges include limitations in light absorption, sluggish charge transfer, rapid charge recombination, slow diffusional transport, and complications associated with mass transport.Among these challenges, efficient charge separation at the catalyst interface is a key process for OWS.The special charge separation relies on several critical factors.These factors encompass having an appropriate energy band structure to facilitate efficient charge separation and transfer, implementing surface modification or passivation techniques to optimize defect and trap center functionalities, and fine-tuning morphology to create heterostructures.The primary consideration in designing a photocatalytic system with spatial charge separation pertains to determining the energy band structure and the relative positions of the CB edge and VB edge of the photocatalysts with reference to the reversible hydrogen electrode.Only when the photocatalysts exhibit the fitting energy band structure suitable for OWS can other strategies be viably contemplated.Another important strategy is to choose cocatalysts with the fitting Fermi energy level, compatible lattice, and electronic structure for the primary catalysts. [5,170]Meticulous control over defect sites and trap centers holds the promise of enhancing chargeseparation efficacy.Furthermore, it's crucial to form heterostructures that facilitate intimate surface contact over a large area, enhancing efficient charge separation.To improve the spatial charge-separation process, different architectures have been developed, including 1D/1D, 2D/2D, 1D/2D, 0D/2D, 0D/1D, 0D/3D, 2D/3D, and facet-dependent architectures.Each of these architectures possesses unique properties that can be tailored to enhance spatial charge separation and improve the overall efficiency of photocatalytic water splitting.The 0D structures like quantum dots and nanoparticles offer unique properties that can facilitate efficient spatial charge separation.In contrast, 1D structures like nanowires can improve charge-transfer efficiency due to their high aspect ratio, leading to increased interface contact area for charge separation.For instance, the 1D gallium nitride nanorods coated with spatially separated 0D hydrogen and oxygen evolution cocatalysts serve as a good example of 0D/1D spatial charge separation. [141]However, achieving a uniform and controlled distribution of 0D nanoparticles over 1D nanostructures remains a challenge.The 2D structures such as layered materials and graphene, due to their large surface area, offer enhanced charge mobility and improved light absorption.Novel strategies, such as electrostatic self-assembly used by Zhao et al. to prepare 2D/2D g-C 3 N 4 /BCNNs, can be employed to enhance STH efficiency. [148]Combining 1D and 2D structures in 1D/2D or 2D/3D architectures can further enhance spatial charge-separation and OWS efficiency.For example, a hollow nanosphere can be decorated with hydrogen and oxygen evolution catalysts on the inner and outer surfaces, effectively segregating the reduction and oxidation sites.This approach not only provides spatial charge separation but also has the potential to suppress the OWS back reaction.Moreover, 0D/2D and 0D/1D architectures can enhance surface area and control charge-transfer properties.Facet engineering is an attractive method for achieving spatial charge separation without relying on external charge-separation strategies.Anisotropic charge separation at crystal facets, as demonstrated by Wan et al.
for PbTiO 3 single crystals with spatially separated HER and OER sites, presents another potential route for achieving high photocatalytic efficiency. [162]he development of these architectures for photocatalysis is still in the early stages, offering potential for further optimization and improvement.The future of spatial charge separation in photocatalysis is promising, and the ongoing development of new architectures will continue to be an active area of research in the coming years.Despite the challenges, the potential for photocatalytic water-splitting remains promising, and further research in this field is crucial to address the challenges and opportunities for developing efficient photocatalysts for renewable energy conversion and storage.

Figure 1 .
Figure 1.a) Illustration of the process of photocatalytic overall water splitting (OWS) on a semiconductor photocatalyst, which involves i) excitement of the photocatalyst by light, ii) separation and transportation of charge carriers, and iii) H 2 and O 2 evolution reactions at the surface of the catalyst.b) The energy band diagram shows photocatalytic water splitting over a semiconductor type photocatalyst.

Figure 3 .
Figure 3. Schematic illustration of a) Type I heterojunction, b) Type II heterojunction, c) direct Z-scheme heterojunction, d) Z-scheme heterojunction with electron mediator, e) Schottky junction, and f ) S-scheme heterojunction.

Figure 4 .
Figure 4. a) An high resolution scanning electron microscopy (HR-SEM) image of the Ta 3 N 5 /BaTaO 2 N (BTON) sample, providing a detailed view of its morphology and structure.b) The surface potentials in the dark and under the irradiation of λ = 450 nm.Five cycles of OWS activity of c) the Ta 3 N 5 /BTON heterostructure and d) a schematic illustration depicts the spatial separation of photogenerated electrons and holes between the 1D Ta 3 N 5 nanorod and 0D BTON nanoparticle.Reproduced with permission.[130]Copyright 2019, Wiley-VCH GmbH.e) A top-view SEM image of Rh-CoO x /GaN system, showcasing the surface morphology of the material.f ) A cross-section view of the SEM image.g) The photocatalytic OWS activity of GaN nanorod arrays without and with the cocatalysts Rh and CoO x , and h) a schematic illustration of the spatial separation of electrons and holes on GaN nanorod arrays.Reproduced with permission.[141]Copyright 2020, Wiley-VCH GmbH.

Figure 5 .
Figure 5. a) A schematic representation charge-transfer and separation mechanism on C ring /g-C 3 N 4 in plane heterostructure.b) The OWS stability performance of C ring /g-C 3 N 4 nanosheets (CNN).c) A schematic diagram of the charge-transfer mechanism in the C ring /CNN heterostructure.Reproduced with permission.[145]Copyright 2017, American Chemical Society.d) HAADF-STEM image of the BPB heterostructure, providing a detailed view of the structure.e) The Z-scheme photocatalytic water splitting using BPB under visible-light irradiation.f ) The effect of the ratio of BP in BPB/Co 3 O 4 on photocatalytic OWS under visible light (λ > 420 nm), highlighting the impact of composition on the photocatalytic performance.g) The photocatalytic stability test on BPB/Co 3 O 4 under visible light (λ > 420 nm).Reproduced with permission.[146]Copyright 2017, Wiley-VCH GmbH.The α-Fe 2 O 3 /g-C 3 N 4 (FC) hybrids: h) an HRTEM image of the FC hybrid.i) Photocatalytic OWS over FC, depicting H 2 and O 2 evolution rates using RuO 2 /FC/Pt under λ > 400 nm light irradiation.j) The Z-scheme mechanism in FC photocatalyst, illustrating the charge-transfer and separation processes involved in the photocatalytic water-splitting mechanism.Reproduced with permission.[147]Copyright 2017, Wiley-VCH GmbH.

Figure 6 .
Figure 6.Various characterizations and performance evaluations of different photocatalytic systems.a) A TEM image of the La-and Rh-co-doped SrTiO 3 (STO:La/Rh)-graphene-BiVO 4 (BVO) composite.b) The photocatalytic OWS performances of STO:La/Rh-BVO photocatalyst system.c) Fluorescence spectra of 3 Â 10 À3 m terephthalic acid solution produced by single photocatalyst and heterojunction systems after light irradiation with λ > 400 nm andd) the proposed Z-scheme charge-transfer mechanism for the composite, shedding light on the underlying processes.Reproduced with permission.[153]Copyright 2021, Wiley-VCH GmbH.e) The TEM-energy-dispersive X-ray (EDX) mapping of a slice of SrTiO 3 /TiO 2 (ST) HoMs after a 12 h hydrothermal reaction.f ) Photoluminescence (PL) spectra, highlighting the emission properties of the samples.g) Photocatalytic OWS performances of ST HoMSs samples with different reaction times.h) The possible band structure diagram for ST HoMSs.Reproduced with permission.[172]Copyright 2019, Wiley-VCH GmbH.i) Charge-transfer and separation mechanism on Pt/ZnTiO 3Àx N y /RhO x photocatalyst is illustrated.j) The PL spectra of the samples, aiding in the understanding of their luminescent characteristics.k) The OWS measurements of spatially separated OEC and HEC on Pt/ZnTiO 3-x N y /RhO x and l) the OWS measurements of Pt/ZnTiO 3Àx N y /RhO x with OEC and HEC on the same surface (nonspatially separated OEC and HEC).Reproduced with permission.[154]Copyright 2021, Wiley-VCH GmbH.

Figure 7 .
Figure 7. a,b) TEM images of CoO x and Pt deposited 18-and 6-facet SrTiO 3 , respectively.c) The H 2 and O 2 evolution rates on 18-facet SrTiO 3 loaded with CoO x and Pt.Reproduced with permission.[160]Copyright 2016, The Royal Society of Chemistry.d) 3D particle morphology and crystal orientation of Bi 2 YO 4 Cl, e) selectively deposition of OEC and HEC in Bi 2 YO 4 Cl, and f ) the photocatalytic H 2 and O 2 production rates of Bi 2 YO 4 Cl photocatalysts synthesized by different methods under visible-light irradiation (λ ≥ 420 nm).Reproduced with permission.[161]Copyright 2023, American Chemical Society.g) Schematic illustration ferroelectric polarization effect on electron and hole separation on PbTiO 3 and h) schematic image of single-domain PbTiO 3 before and after photochemical deposition of Au and MnO x .i) Finally, H 2 and O 2 evolution rate of PbTiO 3 /Rh/Cr 2 O 3 under simulated solar irradiation.Reproduced with permission.[162]Copyright 2021, American Chemical Society.

Figure 8 .
Figure 8. a) SEM and TEM analyses of ZnIn 2 S 4 (ZIS)/WO 3 (WO)/C-wood photocatalyst.c) Comparison of H 2 and O 2 evolution rates by OWS over ZIS/WO/C and ZIS-WO/C-wood (ZWC-wood)and d) charge-transfer and separation mechanism.Reproduced with permission.[164]Copyright 2021, Wiley-VCH GmbH.e) A schematic diagram of BiVO 4 /carbon dots (CDs)/CdS (BCC) heterostructure for OWS.f ) The proposed S-scheme chargetransfer mechanism for this heterostructure under visible-light irradiation, h) gas evolution rates for H 2 and O 2 from different photocatalysts, g) accompanied by an FESEM image of CBM.Reproduced with permission.[165]Copyright 2021, American Chemical Society.i) SEM image of NiBHT/SrTiO 3 :Al photocatalyst is presented, while j,k) H 2 and O 2 evolution rates by OWS using different hybrid structures.Finally, l) the charge-transfer and transfer mechanism on NiBHT/SrTiO 3 :Al/CoO x hybrid photocatalyst system.Reproduced with permission.[166]Copyright 2022, American Chemical Society.

Table 1 .
The comparison of spatial charge separation over 0D/1D nanostructures for OWS. a)

Table 2 .
The comparison of spatial charge separation over 2D/2D nanostructures for OWS.

Table 3 .
The comparison of spatial charge separation over 3D nanostructures for OWS.

Table 4 .
The comparison spatial charge separation over the redox facets for OWS.

Table 5 .
The comparison of spatial charge separation over hybrid nanostructures for OWS.