English

• 1. Zinc oxide for photocatalysis

  Jianzhong Ma, * Junli Liu, * Guanjie Huang, Xiaoting Zhou

  1.1 Status

  As one of the most important semiconductors, zinc oxide (ZnO) has become a versatile and technologically interesting material [1]. Especially, the nanosized ZnO exhibits remarkable surface reactivity and photosensitivity, stemming from its elevated surface-to-volume ratio. Typically, these materials can efficiently generate reactive oxygen species (ROS) upon photon exposure. Their ability to produce instantaneous hydroxyl radicals contributes to the efficient degradation of organic pollutants, pesticides, heavy ions and other chemical bonds in organic compounds, as well as biological cell structures. These photocatalytic reactions ultimately yield simpler and non-hazardous byproducts, highlighting the potential of ZnO in various environmental and biological applications. Besides, ZnO has strong antibacterial, anti-corrosion, and self-cleaning abilities. It could be widely used in the field of coatings. Meanwhile, the wide bandgap and extremely high binding energy of ZnO demonstrate its great potential for application in piezoelectric materials. In particular, nano ZnO arrays exhibit extremely high specific surface area, and extensive research has arranged them on flexible substrates to prepare high-performance flexible piezoelectric devices.

  1.2 Current and future challenges

  Despite significant advances, ZnO still has some limitations in photocatalysis applications. One major challenge, the swift recombination rates and inadequate corrosion resilience, hinder the viable application of electron-hole pairs. Besides, the comparatively modest band gap (approximately 3.37 eV) poses another significant hurdle that necessitates future attention and advancements. Furthermore, ZnO has strong dissolubility, which is not conducive to long-term activity. Additionally, ZnO nanoparticles are prone to aggregation, which greatly hinders their applications for coatings. As research continues, improving the photocatalysis activity and the dispersibility of ZnO nanoparticles and endowing them with sustained release properties is still challenging, which has great research significance.

  1.3 Advances in science and technology to meet challenges

  In this paper, four reported methods to promote the photocatalytic efficiency of ZnO under visible light are summarized. (1) Doping other elements or creating defects in zinc oxide; (2) Adjusting the nanostructure morphology of zinc oxide; (3) Combining visible light-active materials with zinc oxide; (4) Forming heterojunctions between zinc oxide and other materials.

  In previse research, doping aluminum and cerium in ZnO films greatly improves its photocatalytic activity due to the creation of additional reaction sites for the adsorption of reactant molecules [2]. However, the doping process may lead to changes in the crystal lattice of ZnO, and the type and concentration of doped elements need to be precisely controlled. In order to overcome such defects, the nano-morphologic structure of ZnO can be adjusted. For example, Fe-doped urchin-like ZnO nanoparticles were prepared by our group via a combination of precipitation and calcination, and the obtained ZnO indicated greatly improved photocatalytic activity [3]. Besides, more environmentally friendly and energy-saving methods can be used to improve the photocatalytic activity of ZnO. As reported, a three-phase composite photocatalyst such as ZnO/ZnS/Cds was synthesized to enhance its photocatalytic activity and stability by improving the rapid separation of photogenerated electrons and holes [4]. Furthermore, ZnO-MXene hybrid was synthesized by the ultrasonic oscillation method, and its photocatalytic performance was significantly improved due to the increased surface area of MXene nanosheets [5]. Additionally, ZnO and CuInS₂ formed a core-shell structure heterojunction, and the charge density difference at the interface indicated obvious charge transfer, which greatly improved the light absorption and carrier transport efficiency [6]. In a word, the heterojunction photocatalyst has a simple structure and high conversion efficiency.

  However, ZnO nanoparticles with high surface area are easy to aggregate, which leads to decreased activity. Although the physical method can improve the dispersion of ZnO nanoparticles, it cannot fundamentally solve their tendency to aggregate. Therefore, some novel methods to improve the dispersibility of ZnO nanoparticles have attracted great attention. For example, dispersing ZnO into graphene oxide could improve its dispersibility [7]. The principle is to combine the negatively charged groups (-OH, -COOH, OR) on the surface of graphene oxide with ZnO. Meanwhile, the modification of graphene oxide with ZnO enhances its interlayer spacing, endows the coating with strong hydrophobicity and corrosion resistance, and enhances its potential application in the field of architectural coatings. However, the production capacity demand for coatings is high, and the extensive preparation of graphene oxide will generate a large amount of wastewater and exhaust gas, which contains a large number of organic compounds and heavy metal ions, causing serious pollution to the environment. Therefore, the search for natural and green template materials is of great significance. In 2024, Wei and colleagues proposed a strategy of combining ZnO nanomaterials with lignin to prepare composite coatings with graded micro/nanostructures [8], as shown in Fig. 1. The -OH and -OR present in lignin combine with ZnO achieve good dispersibility and stability. Meanwhile, compared with oxidized graphene, lignin can endow coatings with UV resistance and hydrophobicity, especially with a maximum contact angle of up to 160°. In addition, lignin is a natural and renewable aromatic polymer with advantages such as environmental friendliness and low cost.

  Figure 1

  

  Figure 1.  Diagram of the superhydrophobic mechanism of modified lignin. Reproduced with permission [8]. Copyright 2024, American Chemical Society.

  DownLoad: Full-Size Img PowerPoint

  While improving the dispersibility of ZnO nanoparticles, their photocatalytic performance could also be improved, mainly in the field of organic synthesis and pollutant degradation. In 2023, the Hwang team loaded ZnO into Ni/Sn ZIF material and used it as a catalyst to selectively convert glycerol into 1, 2-propanediol with a conversion rate of up to 94.2% [9]. This strategy enables ZnO nanoparticles uniformly dispersed in the MOF pore structure, effectively improving the instability and photocatalytic performance of the excited state of ZnO caused by the recombination of photogenerated electrons and holes. However, the production cost of MOF materials is relatively high. Based on this, researchers attempted to disperse ZnO nanoparticles using inexpensive carriers such as metal oxides. In 2024, Monai et al. dispersed ZnO nanoparticles into ZrO₂ through a simple solvothermal synthesis method, demonstrating better catalytic activity in CO₂ hydrogenation to methanol, with a methanol product selectivity of up to 81% [10]. Compared to MOFs, the preparation method of ZnO/ZrO₂ is simpler, consumes less energy, and effectively saves production costs.

  1.4 Concluding remarks and prospects

  At present, research on ZnO nanomaterials has been very extensive, mainly due to their low production cost and excellent physical and chemical properties. However, the industrial application of ZnO nanomaterials is still in the basic stage. Firstly, different morphologies of ZnO nanomaterials have a significant impact on their properties. Therefore, in different applications, ZnO requires changing the solvent medium, synthesis temperature, and adding additional additives to control its morphology, which is not conducive to industrial mass production. Besides, individual ZnO nanomaterials are difficult to fit a large number of practical applications and usually require modification by composite with other functional materials. In addition, ZnO nanomaterials have high surface energy and surface activity, making it difficult to maintain long-term stability, and they can bind with trace amounts of water to cause aggregation, which requires hydrophobic modification, resulting in complex production processes and limited application scope. Therefore, finding a more convenient method to control the morphology of ZnO nanoparticles and combining it with other advanced nanomaterials will further promote the industrial application of ZnO.

  2. Ultrathin two-dimensional titanium dioxide materials

  Francesco Parrino*, Riccardo Ceccato, Leonardo Palmisano

  2.1 Status

  The "graphene-like" features of ultrathin 2D titanium dioxide (TiO₂) morphology allow the coexistence of high specific surface area, enhanced activity and optimized semiconductor-electrolyte contact with outstanding optoelectronic and electrochemical properties deriving from the fast 2D electron-transport and the confined thickness. Most of the preparation methods of 2D TiO₂ usually provide mixtures of nanosheets with varying thicknesses (from a few to even a few thousands of nanometers). This is especially true for top-down methods, which also require specific layered bulk materials as precursors. More recently, molecular self-assembly approaches in solution have enabled the synthesis of high-quality ultrathin 2D TiO₂ nanosheets. Topochemical conversion has been rarely proposed, but the field is promising due to the exceptional quality of the materials and the appealing features of the technique [11]. The present contribution reports some recent achievements and the main challenges related to the production of ultrathin 2D TiO₂ materials and the main advanced technological applications.

  2.2 Current and future challenges

  The most important challenge related to technological applications of ultrathin 2D TiO₂ materials is the appropriate control of fabrication procedures at suitable scales. However, a deep understanding of the growth processes and the possibility of engineering physicochemical characteristics is required and the topic is the subject of recent investigations.

  The main problem of molecular self-assembly approaches is the difficulty in avoiding phase separation and controlling the assembly process of templates and precursors into ordered mesoporous structures. This is particularly true for catalytic applications, where the presence of open mesostructures and the stabilization of metal clusters of suitable size without aggregation phenomena during thermal treatments is crucial. Another challenge currently being addressed is the control of phase growth and the possibility of creating high-quality heterojunctions in 2D TiO₂-based materials [12].

  Recently, several studies have focused on the role of defects and facets engineering in ultrathin 2D TiO₂ layers, and this is a critical topic for catalytic and solar energy conversion processes [13]. There is also a need to gain a thorough understanding of the surface chemistry and liquid-TiO₂ interface, especially for photoelectrochemical and battery applications, where the coating layer is in direct contact with the electrolyte. In this case, the rich knowledge gained for TiO₂ single crystals can be very useful.

  Finally, single-layer TiO₂ sheets, the ultimate goal in the field of ultrathin layers, have been reported to be stable based on theoretical calculations, but experimental insights that will provide significant advances in the field of solar energy conversion remain to be revealed [14]. Despite recent efforts and interesting results reported in the literature, further investigations are still needed.

  2.3 Advances in science and technology to meet challenges

  Recently, the growth mechanism of ultrathin 2D TiO₂ onto Cu(001) has been highlighted [15]. The authors were able to reveal the transition from the quasi-hexagonal to lepidocrocite-like structure. Similar studies have been reported on Ag(100), W(100), Pt(110), and Pt(111). Fundamental advances in understanding the deposition of ultrathin 2D TiO₂ layers on metals are relevant for improving the power conversion efficiency of solar cells, for designing electrode materials, or for preserving the magnetic properties of deposited species on metals by means of a properly designed decoupling layer.

  The interaction between metals and ultrathin 2D TiO₂ layers is crucial also for catalytic applications. Recent metal centers (Pt, Pd, Au) deposited on 2D ultrathin layers catalysed the hydrogenation of 4-nitrostyrene [16]. A Pt single atom-layer has been deposited onto mesoporous ultrathin (7 nm) TiO₂ with a high surface area (139 m²/g). The catalyst showed 100% selectivity towards the hydrogenated compounds, good stability, and poisoning resistance. Selective catalytic reduction is the most widely used technology to reduce nitrogen oxide emissions in coal-fired power plants and the gas flow downstream of diesel engines. One of the most active catalysts is Cu-SAPO-34, which, however, has poor resistance to low-temperature hydrothermal aging. To face this problem an ultrathin layer of TiO₂ was deposited onto this catalyst by means of atom layer deposition. The nano-coating did not significantly affect the activity of the catalyst, but dramatically increased its stability in liquid water at 80 ℃ for 24 h [17]. Ultrathin 2D TiO₂ has been successfully used for CO₂ activation to alcohols [18]. The authors obtained an excellent yield (463.9 mol/g_cat) in 8 h with selectivity values up to 98%. This outstanding result has been ascribed to the large contact area and to the high defect density, which enabled CO₂ adsorption and activation. Also in the field of heterogeneous photocatalysis ultrathin 2D TiO₂ materials have been applied. For instance, Gan et al. [19] obtained 2D/2D TiO₂/g-C₃N₄ heterojunctions, and used this material for efficient removal of levofloxacin. Similarly, Zhu et al. [20] prepared TiO₂ (B) nanosheets hierarchically anchored onto a MOF material (NH₂₋MIL-53(Al)) for photocatalytic water treatment. However, it is important to note that, after several decades of intensive research efforts in the field of water remediation, unfortunately, photocatalysis alone has not yet produced commercially available solutions on a large scale.

  The gap between laboratory and industrial requirements could be filled by using robust, cheap and reliable photocatalytic materials coupled with other already applied advanced oxidation treatments such as ozonation [21] or with processes such as membrane separation [22]. In fact, the intensification of efficiency and the advantages deriving from the synergistic action of different technologies could be the key to a new spring for photocatalytic applications.

  Regarding the issue of controlling the phase growth of ultrathin 2D TiO₂ materials, recently Meng et al. [23] synthesized 2D anatase-rutile lateral heterojunctions showing the possibility of tuning the oxygen vacancy degree. Anatase/rutile nanosheets were grown perpendicularly on Ti meshes and used for photoelectrochemical water splitting. In the same year, Li et al. [24] revealed, by in situ transmission electron microscopy, a tunable phase transition behaviour of ultrathin 2D TiO₂ layers from amorphous to rutile, through the anatase phase, induced by both electron irradiation and temperature.

  Recent research interest in the phase transition behaviour of atomically thin 2D materials reveals the important role of their surface chemistry. Xie et al. [25] reported on the phase transition behaviour and mechanism of 2D TiO₂ (B) nanosheets via water molecule-mediated removal of surface ligands. The resulting ultrathin porous layer showed improved electronic and ionic conductivity. Hydrogen bonding among TiO₂ precursors was proposed to guide 2D assembly, providing easy electrolyte penetration and high surface area. The material showed one of the best performances as an anode in lithium-ion batteries, due to the high Li⁺ diffusion rate and outstanding electrochemical characteristics [26].

  These results have raised further interest in the application of 2D TiO₂ nanosheets as components of complex materials for energy applications. Li et al. [27] prepared layered nanocomposites composed of alternating nanolayers of TiO₂ and amorphous carbon that exhibited superior pseudocapacitive behaviour. An ultrathin amorphous TiO₂ layer was deposited via atomic layer deposition (ALD) on lithium cobalt oxide to stabilize the cathode-electrolyte interface and reduce the extent of unwanted phase transitions of this important cathode material during the rapid diffusion of Li⁺ in lithium-ion batteries [28].

  Ultrathin amorphous 2D TiO₂ has also shown excellent characteristics as an anode material in lithium batteries, exhibiting high specific capacity and excellent performance. It is important to note that the isotropic nature of the 2D anode greatly reduces the damage associated with battery volume gradients [29].

  Further recent applications of 2D TiO₂ nanolayers range from biotechnology to hydrogen storage and advanced electronics. For instance, Capek et al. [30] demonstrated a strong improvement in the cellular interaction with biomedical Ti prostheses by coating them with an ultrathin layer of TiO₂ by atomic layer deposition. In the field of hydrogen storage, it has been demonstrated that ultrathin 2D TiO₂ enriched in oxygen vacancy dramatically improved the stability and the storage kinetics of MgH₂ deposited in the 2D confined region [31]. In the field of advanced electronic applications, it has been shown that the insulator-to-metal transition (IMT), an important feature of quantum materials, can be achieved by depositing VO₂ layers onto ultrathin rutile TiO₂ nanosheets [32].

  More recently, a high mobility modulation has been discovered in ultra-thin amorphous TiO₂ transistors. These ultra-thin TiO₂ film transistors (u-TFTs) possess ultra-sharp sub-threshold swing and excellent on-off ratio. These characteristics make TFTs attractive and promising alternatives to conventional transistors [33].

  2.4 Concluding remarks and prospects

  The recent interest in 2D TiO₂ nanolayers has led to exciting recent applications in different fields ranging from solar energy conversion, catalysis, batteries, hydrogen storage, and advanced electronics. However, the success of technology can only be assessed once the barrier of the scale-up phase has been overcome. In this regard, it is important to leverage the knowledge gained in the preparation of 2D TiO₂ by physical method depositions.

  Most of them are already industrially applied for the production of large-area layers on glasses and ceramics. In this field, many issues have emerged during the scale-up process regarding the optimization of parameters and the quality and durability of the layers. This is the real challenge in the use of 2D TiO₂ layers. Notably, fundamental research is still required to highlight the outstanding features of TiO₂ in the graphene-like form. This may even revitalize traditional heterogeneous photocatalysis by suggesting new advanced applications.

  3. Two-dimensional carbon-based materials for electrocatalytic nitrogen reduction

  Boon-Junn Ng, * Lutfi Kurnianditia Putri

  3.1 Status

  Electrocatalytic nitrogen reduction reaction (NRR) represents a sustainable and eco-conscious approach for converting dinitrogen (N₂) into ammonia (NH₃) under ambient conditions, offering a viable alternative to the energy-intensive Haber-Bosch process. Inspired by the natural nitrogen reduction mechanisms in nitrogenases and the use of molecular catalysts, emerging electrochemical methods utilize electrocatalysts to produce ammonia via an intricate six-electron, six-proton transfer reaction with multiple stages and various reaction intermediates. Transition metals (TMs) have historically dominated the field, leveraging their capabilities to partially inject d-orbital electrons into the empty π* orbital of molecular N₂, thereby reducing the bond order of the N≡N triple bond [34]. Despite the promising kinetic aspects of TM-based electrocatalysts, practical applications of electrochemical NRR remain an uphill battle as ammonia production yield and Faradaic efficiency (FE) have yet to reach appreciable scales. Moreover, TM-based materials present several technical hurdles: (1) The catalytic sites of TMs are prone to poisoning by species in the aqueous solution such as H⁺ and OH^-, significantly reducing the selectivity towards N₂ adsorption, (2) TMs often exhibit strong binding with reaction intermediates, resulting in a high energy barrier for NH₃ desorption, and (3) low corrosion resistance in acidic and alkaline conditions.

  For years, two-dimensional (2D) carbon-based materials such as graphene and polymeric carbon nitride have been prominent candidates for a wide range of energy and environmental applications accredited to their distinctive electronic and physicochemical properties. Notably, 2D nanocarbon materials confer several intrinsic features that make them suitable for electrocatalytic NRR compared to conventional metal-based counterparts, including an adaptable 'soft' polymeric structure that allows customization of the molecular backbone for electronic tunability, a large specific surface area and uniformly exposed lattice planes associated with high porosity [35]. In addition, TMs typically exhibit weak binding with N₂ molecules and significant adsorption of H⁺ ions, leading to poor NRR activation and low Faradaic efficiency [36]. As such, hydrogen evolution reaction (HER) often competes with NRR due to the lower reaction potential and the stronger adsorption kinetics, particularly for TM-based materials. On the flip side, the low selectivity towards HER and unique composition tunability of organic frameworks make carbon-based materials worthy of exploration for NRR with many tantalizing possibilities [34]. Furthermore, 2D materials enable the creation of simple active sites with uniform coordination numbers, maximizing active site density owing to their atomic thickness and ultra-high specific surface area.

  The inception of graphene in 2004 has sparked a surge of interest in this 2D carbon allotrope ascribed to its ballistic conduction and high electron mobility with minimal scattering at room temperature [37]. An ideal graphene sheet consists of a single-atomic layer with a hexagonal close-packed carbon network, where each carbon atom is covalently bonded to three other carbon atoms via σ bonds (Fig. 2). Moreover, graphene provides a versatile platform with a large surface area to support the uniform dispersion of nanoparticles, making it suitable for use in composite materials. Similar to graphene, polymeric carbon nitride, often known as graphitic carbon nitride (g-C₃N₄), is a layered 2D extension of interconnected heptazine (tri-s-triazine) units linked by tertiary amine groups (Fig. 2). Carbon nitride has emerged as a focal material in catalysis due to its tunable structural defects and surface terminal groups which are pivotal in catalytic activation. Besides, carbon nitride possesses two-fold coordinated N atoms and large periodic vacancies, providing sufficient space for the chemisorption of gas phase N₂ [38]. Furthermore, the presence of empty sp² hybridized orbitals in carbon nitride facilitates strong binding with N₂ through π-backdonation, thereby weakening the N≡N triple bond and enhancing the reduction of N₂ [34]. As of recent, heteroatom-rich carbons such as nitrogenated holey graphene (C₂ N) have gained attention as promising NRR catalysts attributed to the regulated local distribution of electron density towards N₂ adsorption. The high concentration of pyrazinic nitrogen and terminal nitrile groups in C₂ N facilitate the cleavage of N≡N via strong-weak electron polarization on the adsorbed N₂ with a low energy barrier [39]. Graphdiyne (GDY), another novel carbon allotrope featuring sp/sp²-cohybridized carbon atoms, renders highly conjugated large π-structures with significant charge distribution inhomogeneity [40]. Recent advancements and density functional theory (DFT) calculations underscore the potential of GDY as an effective catalyst for NRR.

  Figure 2

  

  Figure 2.  Schematic overview of 2D carbon-based electrocatalysts with modification strategies toward efficient N₂ reduction reaction.

  DownLoad: Full-Size Img PowerPoint

  3.2 Current and future challenges

  Despite significant advances, 2D carbon-based materials still encounter several limitations in their applications for electrocatalytic NRR. One major challenge is their relatively low intrinsic activity compared to conventional TM-based catalysts [41]. While carbon-based materials offer high surface area and tunable electronic properties, achieving effective activation of N₂ molecules with high selectivity and minimized competing HER via precise active sites remains formidable due to inherent energetic barriers. Additionally, carbon materials can suffer from poor stability under electrochemical conditions. Furthermore, these stringent requirements are exacerbated by the limited mobility of charge carriers that are typical in organic materials. Taking carbon nitride as a primary example, despite its idealized depiction in literature as a crystalline heptazine structure (C₃N₄) linked by tertiary amine groups, research suggests that common synthesis methods often result in carbon nitride with residual hydrogen (CN_xH_y) [42]. This hydrogen appears as an uncondensed melon with secondary amine bridges, rendering the material amorphous or semi-crystalline [43]. The structural imperfection leads to reduced electron conductivity, further hindering its performance in electrocatalytic applications. On the other hand, undoped graphene is intrinsically inert due to its strongly bound unpaired electrons in the delocalized π-system, which hinder its adsorptivity and reactivity [44]. As research continues, addressing these challenges is essential for the practical implementation of 2D carbon-based materials in sustainable ammonia synthesis.

  3.3 Advances in science and technology to meet challenges

  In light of the aforementioned challenges, significant interest has been directed toward modifying 2D carbon-based materials attributed to the unique atomic-level design flexibility of their 'soft' organic backbones, which allow for extensive structural and electronic tunability. The primary focus revolves around three catalyst design approaches: (1) Vacancy (defect) engineering, (2) p-block heteroatom doping and (3) incorporation of single-atom catalysts. Defect engineering stands out as a promising modification strategy for creating active sites in carbon materials. The introduction of intrinsic defects and impurity phases into the carbon lattice causes a redistribution of extra electrons to adjacent atoms via the delocalized π-electron network, thereby regulating the energy barrier for N₂ activation [45]. Lv et al. demonstrated the crucial role of nitrogen vacancies in polymeric carbon nitride (PCN–NV) for electrocatalytic NRR [46]. The presence of N vacancies induced the chemisorption of N₂ molecules through the formation of a dinuclear end-on coordinated structure (Figs. 3a and b). Owing to the modulated π-electron delocalization, electrons on the adjacent carbon atoms are transferred to the adsorbed N₂ via an electron back-donation process, which drastically activates the N≡N triple bond (Fig. 3c).

  Figure 3

  

  Figure 3.  N₂ adsorption geometry on polymeric carbon nitride with nitrogen vacancy (PCN–NV) shown in (a) top view, and (b) side view. (c) Charge density difference of the N₂-adsorbed PCN–NV. Reproduced with permission [46]. Copyright 2018, Wiley. (d) Schematic illustration of NRR for boron-doped graphene (BG). Reproduced with permission [47]. Copyright 2018, Cell Press.

  DownLoad: Full-Size Img PowerPoint

  As most pristine carbon materials are inert towards electrocatalysis, the activity of metal-free carbon-based electrocatalysts is attributed to heteroatom doping, particularly with p-block elements that possess strong electronic affinity and modulate the charge density of adjacent carbon atoms [34]. Heteroatoms such as B, N, O and P which differ in size and electronegativity than C atoms could effectively regulate the charge and spin density, thereby altering the charge distribution and the adsorption competition between N₂ and H⁺ [45]. In a seminal work by Yu et al., B-doped graphene (BG) was synthesized by thermal reduction of graphene oxide (GO) with H₃BO₃ under H₂/Ar gas [47]. Due to the electron-deficient nature of B dopant sites in BG, it provides excellent positively charged active sites for the formation of B-N bonds and the subsequent reaction for NH₃ production (Fig. 3d). In this context, the catalytic sites of BG act as Lewis acids which are ideal for binding with N₂ (a weak Lewis base) and inhibiting the adsorption of H⁺. As a result, the NRR Faradaic efficiency was significantly enhanced with suppressed HER. In another work, Zou et al. developed a Cl-doped ultrathin GDY with a suppressed kinetic barrier for the first proton-coupled electron transfer to N₂ which is the rate determining rate for NRR [48]. Atomic interface regulation is another effective modification strategy to enhance the electrocatalytic performance of NRR. Gu et al. incorporated tungsten (W) single-atom catalysts (SACs) onto GO, followed by carbonization to yield W-SACs supported on the N-doped carbon layer with O, N coordination [49]. Owing to the unique local coordination of W-N₂O₂ and the higher electronegativity of O atoms, electron localization was enhanced around W atoms. As such, W SACs are conductive to polarization and activation of N₂ molecules, resulting in enhanced NRR activity.

  3.4 Concluding remarks and prospects

  In conclusion, 2D carbon-based materials represent a pivotal platform for electrocatalytic NRR due to their unique tunability of electronic, structural and catalytic sites. Challenges associated with their intrinsic low catalytic activity and Faradaic efficiency could be mitigated through modification strategies such as heteroatom doping, vacancy engineering and single-atom catalyst incorporation. These approaches effectively modulate charge distribution and surface reactivity, thereby enhancing catalytic performance and selectivity. However, electrocatalytic NRR is still in its early stages of development. Despite significant progress in utilizing 2D carbon-based materials for electrocatalytic NRR, understanding of the intricate relationships between structure, properties and activity remains limited. Further research efforts are crucial to unravel the underlying reaction mechanisms of complex multistep reactions and address the physicochemical bottlenecks that govern the efficiency and selectivity of electrocatalytic NRR.

  4. Violet phosphorus for photocatalysis

  Huaxing Li, Rongjie Li, Gang Liu*

  4.1 Status

  As a unique class of metal-free 2D layered materials, black phosphorus (BP) has aroused intense research interest in the scientific community over the past decade [50]. The biggest challenge of BP lies in its poor stability under ambient conditions, which greatly hinders further exploration of its full potential. Very recently, another phosphorus allotrope, violet phosphorus (VP), also known as Hittorf's phosphorus, has been rediscovered with remarkable intrinsic properties, such as an increasing band gap with decreasing layer number from bulk (1.7 eV) to monolayer (2.5 eV) [51], highly anisotropic hole mobility (3000–7000 cm² V⁻¹ s⁻¹) [51], prominent chemical [52] and thermal [53] stability, which have raised the immediate interest of exploring its intriguing potential in heterogeneous photocatalysis [54–59], optoelectronics [60], tribology [61], and cancer therapy [62]. In particular, the emerging research of VP-based photocatalysts in solar-driven photocatalytic hydrogen evolution (PHE) is attracting growing attention [54–59]. PHE is one of the most promising means producing green hydrogen through a renewable source and may play a key role in achieving the ultimate goal of carbon neutrality. The physicochemical properties of VP are related to its unique geometric structure. As shown in Fig. 4a, a monolayer of VP consists of two-layer pentagonal tubes mutually perpendicular connected by covalent bonds, while a bilayer in Fig. 4b with an interlayer spacing of d = 11 Å is linked by van der Waals forces. The crystallographic structure of VP is monoclinic with a space group of P2/n (13) and lattice constants of a = 9.210 Å, b = 9.128 Å, c = 21.893 Å, β = 97.776° [53].

  Figure 4

  

  Figure 4.  Schematic structure of VP. (a) Top view of a monolayer. The unit cell is marked by red solid and dash lines. (b) Side view of a bilayer. Violet balls represent P atoms.

  DownLoad: Full-Size Img PowerPoint

  4.2 Current and future challenges

  Ultrathin 2D materials usually suffer from insufficient solar-light absorption, weak redox capacity, and fast charge-carrier recombination in particular. In the development of VP-based photocatalysts for PHE, various approaches, such as constructing heterojunctions, doping, and coupling cocatalysts, have been devised to tackle the above challenges. The controllable synthesis of ultrathin VP is essential for developing high-performance photocatalytic systems. Currently, few-layer VP nanosheets (NSs) and VP quantum dots (QDs) are obtained from VP bulk crystal by top-down methods, such as sonication-assisted liquid exfoliation. Controllable fabrication of few-layer VP NSs and VP QDs at large scale, especially monolayer VP, is still challenging.

  4.3 Advances in science and technology to meet challenges

  Although VP holds merits as mentioned above, further modification of VP is indispensable to boost its photocatalytic performance. A few approaches have been devised to modify VP, such as constructing heterojunctions, doping, and anchoring cocatalysts. Zhang and coworkers first reported the PHE of bulk VP [54]. In the presence of a sacrificial donor of methanol and a cocatalyst of 1.0 wt% Pt nanoparticles (NPs), VP exhibited a PHE rate of (675 ± 109) µmol h⁻¹ g⁻¹. In our laboratory, we immobilized single Ni atoms onto few-layer VP NSs by photoreduction and explored their PHE performance with the assistance of triethanolamine [55]. As a unique cocatalyst in heterogeneous photocatalysis, single-atom catalysts (SACs) can provide active sites, enhance photon absorption, facilitate charge transport, and hold tunable coordination microenvironment [63]. The as-formed Ni-P₄ moieties were capable of reducing the work function and band gap, facilitating proton adsorption thermodynamics, and enhancing photoexcited charge separation, thereby cooperatively leading to a H₂ evolution rate of 377.8 µmol h⁻¹ g⁻¹ which is an 11-fold increase with respect to that VP NSs. Using a facile ball-milling method, we prepared a heterojunction of few-layer VP NSs and CdS NPs [59]. The prepared heterojunction was verified to follow an S-scheme in terms of photogenerated charge transfer mode across the interface. Directed by an interfacial internal electric field (IEF), such S-scheme heterojunction usually manifests higher charge transfer efficiency as well as stronger redox capacity in comparison with the conventional type Ⅱ counterpart [64,65]. The 5 wt% VP/CdS S-scheme heterojunction exhibited the highest PHE rate of 6.6 mmol h⁻¹ g⁻¹ with lactic acid donor. To further promote photoactivity, we immobilized Pd SAs onto the heterojunction by a one-pot ball-milling method. The 1 wt% Pd-5 wt% VP/CdS hierarchical composite presented a remarkable PHE rate of 82.5 mmol h⁻¹ g⁻¹, approximately 54-fold higher than that of pristine CdS NPs, alongside an apparent quantum efficiency (AQE) of 25.7% at 420 nm. Notably, the Pd SAs together with the VP/CdS S-scheme heterojunction synergistically realized ultrafast electron transport by 2.2 ps. Likewise, Wang, Zhang, and coworkers combined VP QDs and graphitic carbon nitride (g-C₃N₄) to form a VP/g-C₃N₄ heterostructure by an ultrasonic pulse excitation method [58]. In the presence of 0.5 wt% Rh and methanol, a PHE rate of 7.70 mmol h⁻¹ g⁻¹ was achieved which was over 9.2-fold larger than that of bare g-C₃N₄, in addition to an AQY of 11.68% at 400 nm. The underlying photocatalytic mechanism was ascribed to the interfacial P-N bond-mediated direct Z-scheme charge-transfer mechanism. Zhang and coworkers substitutionally doped Sb into VP crystal while the atomic ratio between P and Sb is 76 to 3.27, and examined the doping effects on PHE [57]. With the assistance of 1.5 wt% Pt and ascorbic acid, the H₂ evolution rate of Sb-doped VP was 1473 µmol h⁻¹ g⁻¹ which is nearly 4 times greater than that of VP. The observed photoactivity is attributed to enhanced light absorption, favorable hydrogen adsorption-desorption thermodynamics and superior PHE kinetics.

  4.4 Concluding remarks and prospects

  In summary, the promising potential of VP in PHE has been systematically investigated in the past two years, albeit largely remains unexplored. To prepare VP-based photocatalysts, various methodologies like constructing heterojunctions, doping, and coupling cocatalysts are capable of tackling the grand challenges including charge carrier recombination, a major factor affecting the photoactivity in general. Despite the progress made in VP-based photocatalysts for PHE, there are still some obstacles to overcome, such as long-term photochemical instability. Of considerable importance is the development of facile and sustainable protocols for the large-scale synthesis of VP, in particular single-layer VP NSs and single-layer VP QDs. We can conclude that VP will appeal to the photocatalytic research community in areas including but not limited to PHE, CO₂ photoreduction, etc.

  5. Two-dimensional metal- and covalent-organic framework electrocatalysts

  Yang Wang, Nikolay Kornienko*

  5.1 Status

  Metal-organic frameworks (MOFs), covalent organic frameworks (COFs) and analogous two-dimensional systems are a fascinating class of catalytic material [66]. This material class is typically comprised of organic linkers connected by metal-based nodes, metal sites, or even simple covalent linkages. Thus, these systems offer multiple routes of tunability through modulating the chemical character of the nodes and linkers, thereby changing their electronic, chemical and mechanical properties. In the case that the connectivity is weaker in one axis (such as van der Waals forces between layers vs. covalent bonds within the layers), MOFs and COFs can readily be grown and exfoliated into a 2D morphology with thicknesses down to a single layer [67].

  The unique potential of 2D MOFs and COFs in electrocatalysis stems from their capacity to combine favorable molecular aspects like well-defined and modular active sites with the strengths of heterogeneous catalysts such as direct electronic connection to an electrode and stability through heterogenization [68]. 2D MOFs and COFs can be further tuned through the design of their pores and consequently, the environment around their catalytic site. These sites can function akin to a catalytic pocket of an enzyme to modulate reaction pathways by facilitating the transport of reactants/products into/out of an active site and stabilizing select intermediates [69].

  Thus far, 2D MOFs and COFs have been successfully used as electrocatalysts for simple electrocatalytic reactions like water oxidation [70], CO₂ reduction [71] and hydrogen evolution [72]. This is largely enabled by their function as heterogenized molecular catalysts with high active site exposure [73]. However, these examples only begin to touch on the possibilities of these materials in electrocatalysis. That being said, there are several key areas to pursue to continue the maturation of electrocatalytic technologies centered on 2D MOFs and COFs and these are summarized in Fig. 5.

  Figure 5

  

  Figure 5.  Open avenues to pursue leveraging the unique properties of 2D MOFs and COFs towards next-generation electrocatalytic systems.

  DownLoad: Full-Size Img PowerPoint

  5.2 Current and future challenges

  Developing controllable synthetic strategies, either via bottom-up or town-down approaches, is of paramount importance in taking full advantage of 2D MOFs and COFs. Downsizing the material thickness can enhance the exposure of active sites to the electrolyte and therefore enables a higher catalyst utilization efficiency and consequently, higher mass-normalized activity of the catalyst. However, adverse effects like the re-stacking of 2D layers or their tendency to roll up into a scroll-like morphology have to be simultaneously accounted for and minimized. The benefits of 2D structuring were demonstrated in exfoliated 2D MOFs for water oxidation as higher currents and lower onset potentials were achieved for the exfoliated Cu-based MOF layers relative to the bulk material [74]. In addition, the dimension and layer number of the catalyst have been shown to have a significant effect on the catalyst selectivity when 2D MOFs were used for CO₂ reduction with Cu-based MOFs [75,76]. This may be related to the effects of the layer thickness on the oxidation state, restructuring, and local environment of the active sites. Finally, the assembly of the catalysts on the electrode must be precisely controlled as this will play a determining role in facilitating charge transfer to the catalyst, their active site exposure, and potentially electronic communication of the 2D MOF/COF with the electrode surface.

  An active area for the exploration of 2D MOF and COF catalysts is the orthogonal electronic tuning of the active sites via the organic connecting linkers and interactions with the electrode. Linkers can be used to impart electron donating/withdrawing inductive effects towards the active sites to boost their activity, as evidenced by Co-based COF catalysts used for CO₂ reduction [77]. Similarly, donor-acceptor motifs within Fe-based COFs substantially enhanced their activity toward oxygen reduction [78]. While these studies used electronic interactions within the catalyst to modulate catalytic activity, using the electrode itself to do so is seldom used and can potentially be exploited for further improvement of functionality. In the case where there is a high degree of electronic connectivity between the catalyst and a conductive electrode, the electronic structures of the two become homogenized, as shown by seminal studies in graphite-conjugated catalysts [79]. This has recently been extended to Fe-porphyrin catalysts loaded onto Ni-electrodes to drastically change their selectivity in CO₂ reduction [80]. The utilization of this concept stands to be an orthogonal strategy towards modulating catalyst properties. As with the point above, the synthesis and particularly assembly of 2D MOFs and COFs onto electrodes is key to attain this.

  Due to the tunable chemistry of MOF/COF pores, these materials can be leveraged for precisely controlling the environments around a catalyst. This is akin to an enzyme's structure that features various substrate channels that selectively feed its active site. For example, COFs functionalized with sulfonic acid groups promoted proton transport via the functional groups and O₂ permeation to Pt catalyst surfaces to drastically increase the rates of O₂ reduction on their surfaces [81]. Alternatively, 2D MOF/COF overlayers can be utilized to block particular reactants from reaching a catalyst surface and consequently change selectivity patterns. This was exemplified with COFs that suppressed proton transport but enabled N₂ permeation to boost the selectivity of the underlying catalyst for N₂ reduction to NH₃ [82]. A similar suppression of proton flux, coupled with the enhancement of cation transport increased conferred by COF overlayers boosted the CO₂ reduction capacity of Cu catalysts in acidic electrolytes [83]. These are a few initial functional examples in this direction, though a comprehensive understanding of how to universally apply this strategy to a diverse array of catalyst surfaces and targeted electrocatalytic reactions remains to be developed.

  While heavily researched reactions within the electrocatalysis field such as water electrolysis and CO₂ reduction are also carried out with 2D MOF and COF catalysts, there is much room to grow this reaction scope. For example, electrochemical coupling reactions such as C–N, C-S and C-P coupling to form societally valuable chemicals are now garnering attention from the community [84,85]. The unique advantage of using 2D MOFs and COFs for these reactions is that their active sites and reaction environments can be particularly tuned to promote them. For example, bi-metallic active sites can be incorporated to separately bind CO₂ and NO₃^- and subsequently couple them [86]. Alternatively, designing 2D MOFs/COFs for facilitating selective mass transport towards an active site, as discussed above, can concentrate select reactants and intermediates on/near catalytic sites to similarly promote catalytic coupling [87]. While this is promising, there is much room to grow toward generating true design principles for constructing 2D MOFs and COFs for this purpose.

  5.3 Advances in science and technology to meet challenges

  While proof-of-concept systems have emerged with promise for translating to practical technologies (electrolzers, fuel cells), there are still many gaps to fill. Real-world systems need to operate at high currents, often above 1 A/cm², and 2D MOF and COF catalysts are seldom tested at these current densities. In addition, harsher conditions in the case of alkaline water electrolysis (elevated temperatures, concentrated electrolytes) are used while lab-scale experiments often operate under milder conditions [88]. Further, stability measurements should not be performed over hours but rather over hundreds of hours, and potentially with accelerated degradation tests. Coupled with this is the importance of verifying the actual active state of the catalysts [89]. In particular, MOFs and COFs can readily reconstruct to, for example, metal oxides/oxyhydroxides that are the "real" catalysts for water oxidation [90]. Similarly, Cu-based single-site materials, including MOFs and COFs, reversible reconstruct to Cu clusters under reducing conditions [91]. Therefore, we stress the importance of thorough characterization of the 2D MOFs and COFs employed before, during (e.g., in situ) [92] and after catalytic testing.

  In situ techniques [93] should also be developed to elucidate catalytic processes, particularly for reactions that are not yet fully understood like electrochemical coupling reactions [94,95]. As each technique offers a unique set of insights [96], several complementary techniques are ideally used together to help put together a comprehensive reaction mechanism and list of structure-activity relationships. The experimental data would further be strengthened through combination with computational modelling. Here, machine-learning approaches can come into play once a large enough data set is available [97].

  5.4 Concluding remarks and prospects

  In all, the use of 2D MOFs and COFs is on a promising trajectory with a continually increasing number of proof-of-concept studies. Despite this, there is still much to be done to close both performance gaps in translating these materials to practical applications and knowledge gaps in the exploration of new catalytic chemistry. We encourage the community to push ahead in both directions, particularly where the unique inherent properties of 2D MOFs and COFs can be leveraged for enhanced activity.

  6. Covalent organic frameworks for heterogeneous photocatalysis

  Shan-Shan Zhu, Zhenwei Zhang, Xiaoming Liu*

  6.1 Status

  Currently, fossil fuels such as coal, oil and natural gas still dominate the global energy supply, and the overuse of non-renewable energy sources has caused energy shortages and environmental pollution problems. Sunlight is a clean and sustainable resource, and the conversion of solar energy into chemical energy utilizing photocatalytic technology is one of the effective ways to solve the above dilemma. As a "medium" for absorbing sunlight, photocatalysts play an important role in the energy conversion process and directly determine the performance of photocatalysis. Therefore, the design and synthesis of novel photocatalysts are imminent.

  6.2 Current and future challenges

  COFs are new crystalline organic porous polymers connected by strong covalent bonds, which mainly consist of lightweight elements. In 2005, Yaghi and co-workers reported boron-containing COFs for the first time, which led to a boom in COFs research [98]. COFs have a number of fascinating properties: (1) Structural designability which permits precise integration of functional organic units into extended ordered frameworks in a bottom-up manner for targeted structural design. (2) The abundant channels and large surface area provide rich catalytic sites for the adsorption and reaction of reagents and intermediates, resulting in efficient mass transfer and high photocatalytic activity. (3) Excellent stability is an important prerequisite for the long-term reaction of photocatalysts. COFs linked by covalent bonds have excellent chemical and thermal stability. (4) Extended and conjugated structures in the in-plane and stacking directions allow for rapid charge carrier migration. These intriguing intrinsic properties give COF-based photocatalysts great potential for applications in energy conversion. In fact, COFs have been exploited as heterogeneous photocatalysts in many fields such as hydrogen (H₂) evolution, H₂O₂ synthesis, carbon dioxide (CO₂) reduction, organic conversion and pollutant degradation (Fig. 6a) [99].

  Figure 6

  

  Figure 6.  (a) Redox-active COFs for photocatalysis reactions. (b) Schematic structure of COF-JLU35 for photocatalytic water splitting for hydrogen production. Reproduced with permission [102]. Copyright 2023, American Chemical Society. (c, d) Schematic structure of TAPD-(OMe)₂ COF and Hz-TP-BT-COF for photocatalytic H₂O₂ synthesis. Reproduced with permission [104]. Copyright 2020, American Chemical Society. Reproduced with permission [106]. Copyright 2020, Springer Nature. Reproduced with permission [54]. Copyright 2023, Springer Nature. (e) Schematic structure of EPCo-COF for photocatalytic CO₂ reduction. Reproduced with permission [107]. Copyright 2024, American Chemical Society. (f) Schematic structure of NiCN COF for photocatalytic organic synthesis. Reproduced with permission [108]. Copyright 2023, Wiley.

  DownLoad: Full-Size Img PowerPoint

  Hydrogen has a high combustion value with an energy density of 142 kJ/g (about 3 times that of petrol, 3.9 times that of alcohol, and 4.5 times that of coke), and is non-polluting and easy to store and transport [100]. In 2014, the first COF-based photocatalysts for artificial photocatalytic hydrogen generation were reported by Lotsch et al. [101]. Subsequently, the application of COFs-based photocatalysts in this field is increasing with time. Liu et al. successfully constructed three-component donor-π-acceptor (TCDA) COFs (COF-JLU35 and COF-JLU36, Fig. 6b) for photocatalytic hydrogen evolution. Compared with the two-component COFs, TCDA-COFs are easier to regulate the band structure and optoelectronic properties, which contribute to promoting exciton dissociation and enhancing charge carrier transfer. Under visible light (420–780 nm) irradiation, the sp² carbon-linked COF-JLU35 showed an impressive HER of 70.8 ± 1.9 mmol g h^-1, which was significantly better than that of the imine COF-JLU36 (23.6 ± 1.1 mmol g^-1 h^-1) [102].

  6.3 Advances in science and technology to meet challenges

  As another potential energy carrier, H₂O₂ has been employed in energy research as it emits only water (H₂O) and oxygen (O₂) after decomposition. The energy density of an aqueous solution of H₂O₂ (2.1 MJ/kg for a 60% aqueous solution of H₂O₂) is comparable to that of compressed H₂. The output potential of an H₂O₂ fuel cell can be as high as 1.09 V theoretically, which is slightly lower than that of hydrogen (H₂, 1.23 V vs. NHE) or methanol (1.21 V vs. NHE) fuel cell, but comparable [103]. The application of COFs in photocatalytic H₂O₂ production was first reported by Van Der Voort and co-workers in 2020 [104]. The energy band structures of the highly crystalline COFs (TAPD-(Me)₂ and TAPD-(OMe)₂, Fig. 6c) obtained by the Schiff base reaction can satisfy the requirements for the conversion of O₂ to H₂O₂ (0.68 V vs. NHE), showing the potential of both COFs to generate H₂O₂. Under visible light (420–700 nm) irradiation, the H₂O₂ generation rates of TAPD-(Me)₂ and TAPD-(OMe)₂ in H₂O/EtOH mixture (9:1) were 97 and 91 mmol g^-1 h^-1, respectively. Subsequently, Chen et al. synthesized s-heptazine-based COFs with spatially separated redox centers (HEP-TAPB COF and HEP TAPT-COF) for efficient overall photosynthetic H₂O₂ production [105]. For HEP-TAPB-COF, the embedded s-heptazine and phenyl serve as reaction sites for O₂ reduction and H₂O oxidation, respectively. In HEP-TAPT-COF, both s-heptazine and triazine portions are O₂ reduction reaction sites, and phenyl is the H₂O oxidation center. As a result of the dual O₂ reduction centers, HEP-TAPT-COF exhibited a much higher photocatalytic H₂O₂ production rate (87.50 mmol/h) than that of HEP-TAPB-COF (49.50 mmol/h). Rapid complexation of photogenerated electrons and holes, mass transfer of reactants and accessibility of catalytic sites seriously affect the efficiency in photocatalysis. With the aim of solving this problem, Jiang et al. reported COFs with unique D-A hexavalent frameworks (Hz-TP-BT-COF with non-conjugated alkane linkage, Im-TP-BT-COF with partial π-conjugated imine linkage, and sp²c-TP-BT-COF with full π-conjugated vinyl linkage, Fig. 6d), and proposed structural design elements for the catalysts closely related to photogenerated carriers, charge transport, and substance transfer [106]. Incorporating the electron-rich hexaphenylbenzophenanthrene with the electron-deficient benzothiadiazole unit enables simultaneous carrier generation and catalytic site activation. The combination of densely distributed catalytic sites and tightly stacked columnar π-units serves as a charge supply chain and abundant water oxidation and oxygen reduction centers, while the one-dimensional nano-channels modified with atoms accepting hydrogen bonds can be used as docking sites to transport water and oxygen to the catalytic centers in a timely manner. In the absence of metal and sacrificial donors, Hz-TP-BT-COF exhibited high yield (5.7 mmol g^-1 h^-1), apparent quantum efficiency (17.5% at 420 nm) and turnover frequency (4.2 h^-1) in water and air.

  Carbon dioxide (CO₂), as an important component of air, is a widely available, cheap and easy-to-obtain, non-toxic, stable, renewable and recyclable "carbon one" (C1) resource. Photocatalytic CO₂ reduction to high-value chemical fuels as a sustainable catalytic conversion approach to reduce greenhouse gas emissions. Functional covalent organic frameworks (COFs), named EPM-COF (M = Co, Ni, Cu, Fig. 6e), were constructed by utilizing perfluorinated metallophthalocyanine (MPcF16) and organic biomolecule compound ellagic acid (EA) as building blocks. EPCo-COF with cobalt metal active sites exhibited an enhanced CO production rate of 17.7 mmol g^-1 h^-1 in the photocatalytic CO₂ reduction reaction, as well as an outstanding CO selectivity of 97.8%, which was attributed to the enhanced EA electron donating ability of EPCo-COF-AT under alkaline conditions [107]. Furthermore, the inherent properties of COFs make them present in organic synthesis. Conjugated COF–CN (Fig. 6f) with pyrene and bipyridine structural units were synthesized by Knoevenagel condensation reaction and then metallized to obtain photosensitive NiCN catalysts with long-range conjugated structures, which were applied in efficient photocatalytic borylation and trifluoromethylation of halogenated aromatics by Lin et al. [108]. The potential of COF involved in the photocatalytic C–N, C-S, and C–O coupling reactions also further demonstrated the potential as a photocatalyst.

  6.4 Concluding remarks and prospects

  As emerging crystalline porous organic polymers, COFs have the potential to replace traditional inorganic semiconductors, and hence are considered as ideal candidates for advanced photocatalysts. From the basic principle of photocatalysis, the conversion of solar to chemical energy involves three key steps: light absorption, light-induced charge separation and migration and surface reaction, and each one of them affects the overall efficiency of the photocatalytic system. Based on this principle, a number of strategies for enhancing photocatalytic performance have been developed. Construction of electron donor-acceptor systems, modification of functional groups, doping and hybridization with other semiconductors are common means to improve the performance of COF-based photocatalysts. Additionally, hydrophilicity, crystallinity, degree of conjugation, and specific surface area also affect the catalytic performance of COFs.

  The investigation of COFs is still in its nascent stages, with several challenges yet to be addressed due to the low catalytic efficiency, poor product selectivity and limited organic reaction range of COF-based catalysts. COFs with large surface area, high crystallinity, robust stability and affinity for hydrophilicity were designed and synthesized to meet practical applications. However, the effects of pore size and specific surface area on the catalytic process have not been fully explored, and their potential influence mechanisms remain to be explored. Besides, the investigations about the effects of exciton dynamics and hydrophobicity/hydrophilicity on the photocatalytic activity of COFs are still insufficient. Standardized comparative metrics should be established to evaluate the photocatalytic performance of different research groups and facilitate the exchange of data and results. In-depth studies of the catalytic mechanism of COF-based catalysts are very limited, and currently, only a few studies provide detailed information on the exact location of the active sites and well-defined reaction pathways. Finally, exploration of the structure-performance correlation of COF-based photocatalysts is of particular interest. We believe that COF-based photocatalysts offer great opportunities to address environmental and energy issues.

  7. Graphitic carbon nitride for photocatalysis

  Nur Atika Nikma Dahlan, Siang-Piao Chai*

  7.1 Status

  The broad carbon nitride family (C_xN_y) has been of interest in the last few decades for its great potential in photocatalytic energy conversion and environmental applications. Theoretical calculations predicted various crystalline structures of C_xN_y with distinct properties including CN, C₂ N, C₃ N, C₃N₂, C₃ N₃, C₃N₄, C₃N₅, C₃N₆, C₃N₇, C₄ N, and C₅N. Among the allotropes, C₃N₄ is the most explored experimentally thanks to its facile synthesis from widely available and non-toxic nitrogen-rich precursors. C₃N₄ itself exists in seven possible phases including α-C₃N₄, β-C₃N₄, cubic C₃N₄, pseudocubic C₃N₄, g-h-triazine, g-o-triazine, and g-h-heptazine [109]. The last three graphitic forms are the most stable with visible light activity, rendering their popularity in the field of photocatalysis upon its first encouraging study by Wang et al. in 2009 [110]. Consistent with the widely acceptable designation, g-C₃N₄ herein will refer to the g-h-heptazine structure having a narrow indirect band gap of ~2.7 eV. This representative 2D photocatalyst is composed of sheet-like π-conjugated heptazine units with intraplanar covalent bonds and interplanar weak van der Waals forces. Its polymeric structure presented g-C₃N₄ with charming properties such as high chemical and thermal stability, high tunability, and suitable band structure for a wide range of photocatalytic actions including but not limited to water splitting, carbon dioxide reduction, disinfection, pollutant removal, and organic synthesis.

  7.2 Current and future challenges

  Despite its numerous merits, the photoactivity of pure g-C₃N₄ remains limited attributable to its high rate of photogenerated charge recombination, suboptimal structural crystallinity, low surface area, and sluggish internal electron transport. Research on g-C₃N₄ seems to be approached from a broad range of perspectives, yet the critical idea behind the many strategies to improve its photoactivity boils down to stimulating the effectiveness of charge separation. Often, synergistic effects that alleviate the other limiting factors would manifest alongside this main idea, further boosting the photocatalytic performance. Upon absorption of light, photogenerated charges are produced at a time scale of femtoseconds. Their recombination occurs in less than a microsecond but reaction at surface active sites takes place at a time scale of milliseconds. Suppression of this charge recombination by means of effective separation will directly result in a higher supply of potent charges available for the desirable redox reactions.

  The simplest yet proven approaches to tailor charge dynamics in g-C₃N₄ are internal restructuring via doping, defects engineering, and functionalization. These strategies modulate the electronic arrangement of g-C₃N₄ at the molecular level, at the cost of distortion of its neat planar structure. Defect sites may form and care has to be taken considering their double-edged sword nature, whereby, their presence in controlled amount may aid in trapping charges for reduced recombination rate, but their excessive amount will turn them into charge recombination centers with the opposite influence. The role of these defects at the molecular level is still not fully understood requiring more research in the future. Another promising venture in enhancing charge dynamics is the construction of an internal electric field via heterojunction formation. This strategy brings forward vast and diverse types of materials for potential coupling with g-C₃N₄ such as the conventional TiO₂, carbon-based materials, MXenes, transition metal chalcogenides, metal oxides, metal/covalent organic frameworks (M/COFs), and layered double hydroxides (LDHs). An effective heterojunction requires a proper band structure alignment and minimal interfacial defects, which have motivated current exploration in types of materials that best complement g-C₃N₄ and 2D/2D heterojunction being perceived to be superior for its ability to form intimate contact while maintaining high surface area. As a result, very few have studied the effect of the dimension of the complementary material on charge transfer properties in the formed g-C₃N₄-based heterojunction. There are also limited studies conducting in-depth exploration of both semiconductors in a heterojunction which may have regrettably forsaken their full potential. Aside from that, the design of high-performing metal-free g-C₃N₄-based heterojunctions remains a formidable challenge.

  The conventional and facile high-temperature calcination synthesis method produces g-C₃N₄ with abundant dangling bonds due to its sluggish deamination kinetics, forming a high concentration of defects in the structure. In comparison, a highly crystalline g-C₃N₄ forms a more condensed structure with narrower π-π stacking space serving as efficient charge transport channels for enhanced separation capability (Fig. 7). The first precedents of highly crystalline g-C₃N₄ obtained via molten salt-assisted synthesis method offers better characterization opportunities and precise modification of the polymer. Research on crystalline g-C₃N₄ and its modification is gaining significant interest with future challenges in the field being its complex synthesis incurring high cost.

  Figure 7

  

  Figure 7.  Structural effect of (a) g-C₃N₄ and (b) crystalline g-C₃N₄ on charge transfer and recombination behaviour due to the presence of defect states.

  DownLoad: Full-Size Img PowerPoint

  As an organic semiconductor, g-C₃N₄ innately possesses high exciton binding energy (Ex_BE) viz. the minimum energy needed to dissociate excitons into free electrons and holes. High Ex_BE limits the exciton diffusion length in organic semiconductors (L_D, 5–20 nm) [111], subsequently resulting in soaring charge recombination rate relative to that in inorganic semiconductors, which then translates to low incident photon-to-current efficiency (IPCE) and poor photoactivity. Lowering Ex_BE for improved IPCE and photoactivity remains a pivotal challenge for g-C₃N₄ as an organic photocatalyst.

  7.3 Advances in science and technology to meet challenges

  Extensive research on g-C₃N₄ has primarily focused on its synthesis and applications. Currently, the scientific community is advancing forward by delving deeper into its molecular fundamentals, considering the initial photoexcitation process and the subsequent charge dynamics. Studies on dielectric environment engineering (e.g., doping and heterojunction formation) remain substantial except with greater emphasis on improving g-C₃N₄ intrinsically with precise modification control. As such, publications on the crystalline type of g-C₃N₄, isotype heterojunction/homojunction, single-atom catalysts impregnation, structural asymmetry/polarization engineering, and exciton dissociation studies have become more apparent in the past 3 years (Fig. 8).

  Figure 8

  

  Figure 8.  Schematic illustration of recent strategies to improve charge dynamics and exciton dissociation efficiency in g-C₃N₄.

  DownLoad: Full-Size Img PowerPoint

  The moderate enhancement of photoactivity presented by defects modulation in g-C₃N₄ and limited opportunity in moving forward with this strategy have forged a fresh research trajectory with emphasis on crystalline g-C₃N₄. There is already a rapidly increasing number of research on crystalline g-C₃N₄ with diverse synthetic strategies being developed, from single-component salt-assisted methods as opposed to the complex multicomponent ones, templated methods with anodic aluminium oxide (AAO), Ni-foam, and poly(dimethylsiloxane) as the templates, and two-step calcination, to a more benign microwave-assisted method. As with the defects-rich g-C₃N₄, these crystalline g-C₃N₄ are subjected to various modifications among which K-doping is prominent mainly due to its presence in K-based molten salts used during syntheses. The one-pot molten salt method also poses a structural problem where it favours the formation of polytriazine imide (PTI) rather than the more photocatalytically active tri-s-triazine skeleton. To address this issue, Yuan et al. [112] re-arranged the synthesis sequence to first form P-doped g-C₃N₄ and subjected it to molten-salt post-treatment to obtain high crystallinity P, K-codoped g-C₃N₄ with surface-grafted cyano groups. The photocatalyst yielded an 81-fold improvement in H₂O₂ evolution as compared to the pristine material, attributing the enhancement to (1) bandgap narrowing and acceleration of inter-planar charge migration by K-doping and (2) promoted intra-planar charge transfer by P-doping and crystalline structure. Another notable progress for crystalline g-C₃N₄ is its amalgamation with the emerging S-scheme heterojunction concept. In contrary to the well-known type Ⅱ, the S-scheme mechanism promotes charge separation while maintaining the maximum redox potentials in the photocatalyst system. Coupling two crystalline phases of g-C₃N₄, namely the triazine (TCN) and heptazine (HCN) phases, could form an S-scheme homojunction [113]. The difference in work functions of each crystal phase contributes to the formation of the interfacial electric field upon contact. In the darkness, electron flow from HCN to TCN, and the reversed was observed under light conditions. The homojunction demonstrated 2.09- and 9.45-fold enhancement in CO and CH₄ yields respectively by CO₂ photoreduction in the absence of any sacrificial agents and cocatalysts, as compared to that of bulk g-C₃N₄. On the other hand, Li et al. [65] reported the immobilization of single Pd atoms on g-C₃N₄/CdS S-scheme hybrid to boost charge separation. Single-atom catalysts (SACs) are atomically dispersed metal sites in a catalyst system exhibiting high activity thanks to their maximized surface-active sites and the absence of inactive internal atoms. In the hybrid, electrons in CB of CdS recombined with holes in VB of g-C₃N₄ following the S-scheme mechanism. The remaining electrons then rapidly migrated to Pd atoms via Pd-N and Pd-S bonds, where they reduced H into H₂. Pd SAs herein acted as an electron reservoir that improved the H₂ yield by 20-fold relative to just the hybrid without Pd SA, due to ultrafast charge transfer. Metal atoms bonded to four nitrogen atoms (M-N₄) in g-C₃N₄ establish asymmetry in the distribution of π-electrons, providing a stronger driving force for charge migration. π-electrons are the delocalized electrons arising from π-orbital overlaps forming the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) in organic semiconductors, such as g-C₃N₄. The redistribution of π-electrons therefore, can also alter the optical properties as a result of the modified band structure. Such asymmetry can also be manifested by functionalization, as observed in the P, K-codoped g-C₃N₄ with surface-grafted cyano groups by Yuan et al. [112].

  From the excitonic perspective, the high Ex_BE nature of g-C₃N₄ is a critical factor causing subpar utilization of absorbed photons. Experimentally, Ex_BE can be evaluated from temperature-dependent photoluminescence (TDPL) analysis. Strategies adopted to lower Ex_BE include localized polarization via strain engineering, construction of molecular donor-acceptor units, and molecular junction formation. A facile ultrasonic treatment was reported to induce strain gradients/curvatures in g-C₃N₄ nanosheets, effectively reducing Ex_BE from 52 meV in pristine g-C₃N₄ to 34 meV in the curved structure, due to strong flexoelectric polarizations (vertical charge redistribution) [114]. Meanwhile, a glucosamine-regulated g-C₃N₄ donor-acceptor (GM-CN) system had an Ex_BE of 119.62 meV compared to 208.21 meV for the pristine g-C₃N₄ synthesized from the same approach [115]. Glucosamine introduced carboxylic rings into the g-C₃N₄ structure to serve as the electron acceptor from electron-rich N. The reduction of Ex_BE in modified catalysts manifested into promoted charge separation and photoactivities. The curved g-C₃N₄ gave a 2.9-fold and 2.1-fold increase in H₂ and H₂O₂ yields respectively, while GM-CN gave a 1.34-fold enhancement in xylonic acid production. Aside from improved charge dynamics discussed earlier, the HCN/TCN junction is also advantageous to minimizing Ex_BE. The lowest value of 56 meV was obtained for HCN/TCN as opposed to 83, 79, and 70 meV for cyanuric acid-melamine based g-C₃N₄ (CMCN), HCN, and TCN respectively [116]. The decrease in Ex_BE evidently enriched carrier density and in concert with the construction of a strong internal electric field, the HCN/TCN junction ultimately gave a remarkable 38.5-fold enhancement in photocatalytic H₂O₂ yield relative to that of CMCN.

  7.4 Concluding remarks and prospects

  g-C₃N₄ has emerged as a promising and environmentally benign photocatalyst in numerous applications. However, its full potential remains largely untapped. Extensive research has been carried out to better understand this material at the fundamental level and various modification strategies were exploited to address its shortcomings, resulting in inspiring enhancement in its performance. Current results still fall short of commercial standards but surely, there is still much room for exploration with g-C₃N₄ to push its practical limits. Future research will likely continue to focus on advanced modification strategies extending to the synergistic approaches that combine multiple molecular-level design strategies, or complex catalytic systems (e.g., piezo-photocatalysis, photothermal catalysis, photocatalytic membrane reactor).

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Jing Guo: Writing – original draft. Jianzhong Ma: Writing – review & editing. Junli Liu: Writing – review & editing. Guanjie Huang: Writing – original draft. Xiaoting Zhou: Writing – original draft. Francesco Parrino: Writing – review & editing. Riccardo Ceccato: Writing – original draft. Leonardo Palmisano: Writing – original draft. Boon-Junn Ng: Writing – review & editing. Lutfi Kurnianditia Putri: Writing – original draft. Huaxing Li: Writing – original draft. Rongjie Li: Writing – original draft. Gang Liu: Writing – review & editing. Yang Wang: Writing – original draft. Nikolay Kornienko: Writing – review & editing. Zhenwei Zhang: Writing – original draft. Xiaoming Liu: Writing – review & editing. Nur Atika Nikma Dahlan: Writing – original draft. Siang-Piao Chai: Writing – review & editing. Jianmin Ma: Writing – review & editing.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (Nos. 52272290, 21972030, 52073119, and 52373210), the Natural Science Foundation of Jilin Province (No. 20230101029JC), the Fundamental Research Program of Shanxi Province (No. 202303021212159), and the Monash University Malaysia–ASEAN grant (No. ASE-000010). N. Kornienko and Y. Wang acknowledge the University of Bonn Institute of Inorganic Chemistry.

  
  
• 1. [1]

     Y. Yang, M. Wu, X. Zhu, et al., Chin. Chem. Lett. 30 (2019) 2065–2088. doi: 10.1016/j.cclet.2019.11.001

  2. [2]

     S. Karakaya, L. Kaba, Surf. Interfaces 44 (2024) 103655. doi: 10.1016/j.surfin.2023.103655

  3. [3]

     A. Hui, J. Ma, J. Liu, et al., J. Alloys Compd. 696 (2017) 639–647. doi: 10.1016/j.jallcom.2016.10.319

  4. [4]

     K. He, Int. J. Hydrogen Energy 51 (2024) 30–40. doi: 10.1016/j.ijhydene.2023.08.050

  5. [5]

     X. Liu, C. Chen, Mater. Lett. 261 (2020) 127127. doi: 10.1016/j.matlet.2019.127127

  6. [6]

     F. Qiao, W. Liu, J. Yang, et al., Int. J. Hydrogen Energy 53 (2024) 840–847. doi: 10.1016/j.ijhydene.2023.11.292

  7. [7]

     Z. Zhou, S. Pourhashem, Z. Wang, et al., Chem. Eng. J. 439 (2022) 135765. doi: 10.1016/j.cej.2022.135765

  8. [8]

     D. Wei, S. Liang, S. Lv, et al., ACS Appl. Polym. 6 (2024) 5485–5495. doi: 10.1021/acsapm.4c00679

  9. [9]

     A.S. Nimbalkar, K.R. Oh, S.J. Han, et al., ChemSusChem 17 (2024) e202301315. doi: 10.1002/cssc.202301315

  10. [10]

      X. Zhang, X. Yu, R.G. Mendes, et al., Angew. Chem. Int. Ed. 63 (2024) e202416899.

  11. [11]

      M. Zarattini, C. Dun, L.H. Isherwood, et al., J. Mater. Chem. A 10 (2022) 13884–13894. doi: 10.1039/d1ta06695a

  12. [12]

      M. Li, H. Zhang, Z. Zhao, et al., Acc. Mater. Res. 4 (2023) 4–15. doi: 10.1021/accountsmr.2c00172

  13. [13]

      Y. Chen, L. Soler, C. Cazorla, et al., Nat. Commun. 14 (2023) 6165. doi: 10.1038/s41467-023-41976-2

  14. [14]

      L. Sheng, T. Liao, L. Kou, et al., Mater. Today Energy 3 (2017) 32–39. doi: 10.1016/j.mtener.2016.12.004

  15. [15]

      A.L. Sorrentino, G. Serrano, L. Poggini, et al., J. Phys. Chem. C 125 (2021) 10621–10630. doi: 10.1021/acs.jpcc.1c01098

  16. [16]

      L. Duan, C.T. Hung, J. Wang, et al., Angew. Chem. Int. Ed. 61 (2022) e202211307. doi: 10.1002/anie.202211307

  17. [17]

      H. Tian, W. Xi, Y. Zhang, et al., Sci. China Technol. Sci. 65 (2022) 2325–2336. doi: 10.1007/s11431-022-2149-0

  18. [18]

      C. Yin, X. Li, S. Sun, et al., Chem. Commun. 60 (2024) 3531–3534. doi: 10.1039/d4cc00068d

  19. [19]

      W. Gan, J. Guo, X. Fu, et al., Sep. Purif. Technol. 306 (2023) 122578. doi: 10.1016/j.seppur.2022.122578

  20. [20]

      C. Zhu, H. Yao, T. Sun, et al., Chem. Eng. J. 460 (2023) 141849. doi: 10.1016/j.cej.2023.141849

  21. [21]

      G. Camera-Roda, V. Loddo, L. Palmisano, et al., Appl. Catal. B: Environ. 253 (2019) 69–76. doi: 10.1016/j.apcatb.2019.04.048

  22. [22]

      G. Camera-Roda, V. Loddo, L. Palmisano, et al., Chem. Eng. J. 310 (2017) 352–359. doi: 10.1016/j.cej.2016.06.019

  23. [23]

      M. Meng, L. Yang, J. Yang, et al., J. Colloid Interface Sci. 648 (2023) 56–65. doi: 10.1016/j.jcis.2023.05.193

  24. [24]

      Z. Li, Y. Zhang, B. Zhang, et al., J. Phys. Chem. C 127 (2023) 3640–3646. doi: 10.1021/acs.jpcc.2c08789

  25. [25]

      S. Xie, L. Fan, Y. Chen, et al., Dalton Trans. 52 (2023) 15590–15596. doi: 10.1039/d3dt02752j

  26. [26]

      Y. Qi, X. Zeng, L. Xiao, et al., Sci. China Mater. 65 (2022) 3017–3024. doi: 10.1007/s40843-022-2097-2

  27. [27]

      Y. Li, M.S. Chen, J. Cheng, et al., Langmuir 36 (2020) 2255–2263. doi: 10.1021/acs.langmuir.9b03889

  28. [28]

      L. Gao, X. Jin, Z. Li, et al., Materials 17 (2024) 3036. doi: 10.3390/ma17123036

  29. [29]

      C. Zhongda, Z. Wenqi, L. Zhihao, et al., JUSTC 53 (2023) 0605-1-0605-10.

  30. [30]

      J. Capek, M. Sepúlveda, J. Bacova, et al., ACS Appl. Mater. Interfaces 16 (2024) 5627–5636. doi: 10.1021/acsami.3c17074

  31. [31]

      L. Ren, W. Zhu, Y. Li, et al., Nano-Micro Lett. 14 (2022) 144. doi: 10.1007/s40820-022-00891-9

  32. [32]

      D.J. Lahneman, T. Slusar, D.B. Beringer, et al., npj Quantum Mater. 7 (2022) 72. doi: 10.1038/s41535-022-00479-x

  33. [33]

      N. Tiwale, A. Subramanian, Z. Dai, et al., Commun. Mater. 1 (2020) 94. doi: 10.1038/s43246-020-00096-w

  34. [34]

      C. Liu, A. Tian, Q. Li, et al., Adv. Funct. Mater. 33 (2023) 2210759. doi: 10.1002/adfm.202210759

  35. [35]

      C. Feng, Z.P. Wu, K.W. Huang, et al., Adv. Mater. 34 (2022) 2200180. doi: 10.1002/adma.202200180

  36. [36]

      G. Qing, R. Ghazfar, S.T. Jackowski, et al., Chem. Rev. 120 (2020) 5437–5516. doi: 10.1021/acs.chemrev.9b00659

  37. [37]

      K.S. Novoselov, A.K. Geim, S.V. Morozov, et al., Science 306 (2004) 666–669. doi: 10.1126/science.1102896

  38. [38]

      L. Xia, H. Wang, Y. Zhao, J. Mater. Chem. A 9 (2021) 20615–20625. doi: 10.1039/d1ta04614d

  39. [39]

      W. Zhang, S. Zhan, Q. Qin, et al., Small 18 (2022) 2204116. doi: 10.1002/smll.202204116

  40. [40]

      J. Hu, H. Zou, F. Li, et al., Energy Fuel. 37 (2023) 3501–3522. doi: 10.1021/acs.energyfuels.2c04028

  41. [41]

      Z. Huang, M. Rafiq, A.R. Woldu, et al., Coord. Chem. Rev. 478 (2023) 214981. doi: 10.1016/j.ccr.2022.214981

  42. [42]

      Y. Wang, A. Vogel, M. Sachs, et al., Nat. Energy 4 (2019) 746–760. doi: 10.1038/s41560-019-0456-5

  43. [43]

      B.V. Lotsch, M. Döblinger, J. Sehnert, et al., Chem. Eur. J. 13 (2007) 4969–4980. doi: 10.1002/chem.200601759

  44. [44]

      L.K. Putri, W.J. Ong, W.S. Chang, et al., Appl. Surf. Sci. 358 (2015) 2–14. doi: 10.1016/j.apsusc.2015.08.177

  45. [45]

      S. Zhao, X. Lu, L. Wang, et al., Adv. Mater. 31 (2019) 1805367. doi: 10.1002/adma.201805367

  46. [46]

      C. Lv, Y. Qian, C. Yan, et al., Angew. Chem. Int. Ed. 57 (2018) 10246–10250. doi: 10.1002/anie.201806386

  47. [47]

      X. Yu, P. Han, Z. Wei, et al., Joule 2 (2018) 1610–1622. doi: 10.1016/j.joule.2018.06.007

  48. [48]

      H. Zou, W. Rong, B. Long, et al., ACS Catal. 9 (2019) 10649–10655. doi: 10.1021/acscatal.9b02794

  49. [49]

      Y. Gu, B. Xi, W. Tian, et al., Adv. Mater. 33 (2021) 2100429. doi: 10.1002/adma.202100429

  50. [50]

      L. Ding, P. Shao, Y. Yin, et al., Adv. Funct. Mater. 34 (2024) 2316612. doi: 10.1002/adfm.202316612

  51. [51]

      G. Schusteritsch, M. Uhrin, C.J. Pickard, Nano Lett. 16 (2016) 2975–2980. doi: 10.1021/acs.nanolett.5b05068

  52. [52]

      A. Fali, M. Snure, Y. Abate, Appl. Phys. Lett. 118 (2021) 163105. doi: 10.1063/5.0045090

  53. [53]

      L. Zhang, H. Huang, B. Zhang, et al., Angew. Chem. Int. Ed. 59 (2020) 1074–1080. doi: 10.1002/anie.201912761

  54. [54]

      M. Gu, L. Zhang, S. Mao, et al., Chem. Commun. 58 (2022) 12811–12814. doi: 10.1039/d2cc04461g

  55. [55]

      H. Li, L. Zhang, R. Li, et al., Nan. Today 51 (2023) 101885. doi: 10.1016/j.nantod.2023.101885

  56. [56]

      X. Wang, C. Xu, Z. Wang, et al., J. Mater. Chem. A 11 (2023) 10883–10890. doi: 10.1039/d3ta02021e

  57. [57]

      X. Zhao, M. Gu, R. Zhai, et al., Small 19 (2023) 2302859. doi: 10.1002/smll.202302859

  58. [58]

      X. Wang, Y. Wang, M. Ma, et al., Small 20 (2024) 2311841. doi: 10.1002/smll.202311841

  59. [59]

      H. Li, R. Li, Y. Jing, et al., ACS Catal. 14 (2024) 7308–7320. doi: 10.1021/acscatal.4c00758

  60. [60]

      X. Liu, S. Wang, Z. Di, et al., Adv. Sci. 10 (2023) 2301851. doi: 10.1002/advs.202301851

  61. [61]

      Y. Zhang, D. Zhang, Y. Wang, et al., ACS Appl. Nano Mater. 4 (2021) 9932–9937. doi: 10.1021/acsanm.1c02593

  62. [62]

      C. Zhao, X. Han, S. Wang, et al., Adv. Healthc. Mater. 12 (2023) 2201995. doi: 10.1002/adhm.202201995

  63. [63]

      H. Li, R. Li, G. Liu, et al., Adv. Mater. 36 (2024) 2301307. doi: 10.1002/adma.202301307

  64. [64]

      B. Zhu, J. Sun, Y. Zhao, et al., Adv. Mater. 36 (2024) 2310600. doi: 10.1002/adma.202310600

  65. [65]

      R. Li, H. Li, X. Zhang, et al., Adv. Funct. Mater. 34 (2024) 2402797. doi: 10.1002/adfm.202402797

  66. [66]

      H. Furukawa, K.E. Cordova, M. O'Keeffe, et al., Science 341 (2013) 1230444. doi: 10.1126/science.1230444

  67. [67]

      G. Chakraborty, I.H. Park, R. Medishetty, et al., Chem. Rev. 121 (2021) 3751–3891. doi: 10.1021/acs.chemrev.0c01049

  68. [68]

      K. Kuruvinashetti, J. Li, Y. Zhang, et al., Chem. Phys. Rev. 3 (2022) 021306. doi: 10.1063/5.0090147

  69. [69]

      J. Li, Y. Zhang, N. Kornienko, New J. Chem. 44 (2020) 4246–4252. doi: 10.1039/c9nj05892c

  70. [70]

      S. Gutiérrez-Tarriño, J.L. Olloqui-Sariego, J.J. Calvente, et al., J. Am. Chem. Soc. 142 (2020) 19198–19208. doi: 10.1021/jacs.0c08882

  71. [71]

      H. Gu, G. Shi, L. Zhong, et al., J. Am. Chem. Soc. 144 (2022) 21502–21511. doi: 10.1021/jacs.2c07601

  72. [72]

      Y. Sun, Z. Xue, Q. Liu, et al., Nat. Commun. 12 (2021) 1369. doi: 10.1038/s41467-021-21595-5

  73. [73]

      N. Heidary, T.G.A.A. Harris, K.H. Ly, et al., Physiol. Plant. 166 (2019) 460–471. doi: 10.1111/ppl.12935

  74. [74]

      P. Wu, S. Geng, X. Wang, et al., Angew. Chem. Int. Ed. 63 (2024) e202402969. doi: 10.1002/anie.202402969

  75. [75]

      D. Yang, S. Zuo, H. Yang, et al., Adv. Mater. 34 (2022) 2107293. doi: 10.1002/adma.202107293

  76. [76]

      D. Yang, S. Zuo, H. Yang, et al., Angew. Chem. Int. Ed. 59 (2020) 18954–18959. doi: 10.1002/anie.202006899

  77. [77]

      C.S. Diercks, S. Lin, N. Kornienko, et al., J. Am. Chem. Soc. 140 (2018) 1116–1122. doi: 10.1021/jacs.7b11940

  78. [78]

      Y. Wang, M. Wang, T. Chen, et al., Angew. Chem. Int. Ed. 62 (2023) e202308070. doi: 10.1002/anie.202308070

  79. [79]

      M.N. Jackson, Y. Surendranath, Acc. Chem. Res. 52 (2019) 3432–3441. doi: 10.1021/acs.accounts.9b00439

  80. [80]

      M. Abdinejad, A. Farzi, R. Möller-Gulland, et al., Nat. Catal. 7 (2023) 1109–1119. doi: 10.4121/9247f0c0-aa90-4407-a81f-afe5809fe2bb.

  81. [81]

      Q. Zhang, S. Dong, P. Shao, et al., Science 378 (2022) 181–186. doi: 10.1126/science.abm6304

  82. [82]

      S. Liu, T. Qian, M. Wang, et al., Nat. Catal. 4 (2021) 322–331. doi: 10.1038/s41929-021-00599-w

  83. [83]

      Y. Zhao, L. Hao, A. Ozden, et al., Nat. Synth. 2 (2023) 403–412. doi: 10.1007/s42000-023-00453-7

  84. [84]

      J. Li, Y. Zhang, K. Kuruvinashetti, et al., Nat. Rev. Chem. 6 (2022) 303–319. doi: 10.1038/s41570-022-00379-5

  85. [85]

      J. Li, H. Heidarpour, G. Gao, et al., Nat. Synth. 3 (2024) 809–824. doi: 10.1038/s44160-024-00530-8

  86. [86]

      X. Zhang, X. Zhu, S. Bo, et al., Nat. Commun. 13 (2022) 5337. doi: 10.1038/s41467-022-33066-6

  87. [87]

      C.S. Gerke, Y. Xu, Y. Yang, et al., J. Am. Chem. Soc. 145 (2023) 26144–26151. doi: 10.1021/jacs.3c08359

  88. [88]

      M.F. Lagadec, A. Grimaud, Nat. Mater. 19 (2020) 1140–1150. doi: 10.1038/s41563-020-0788-3

  89. [89]

      W. Zheng, L.Y.S. Lee, ACS Energy Lett. 6 (2021) 2838–2843. doi: 10.1021/acsenergylett.1c01350

  90. [90]

      W. Zheng, M. Liu, L.Y.S. Lee, ACS Catal. 10 (2020) 81–92. doi: 10.1021/acscatal.9b03790

  91. [91]

      C.E. Creissen, M. Fontecave, Nat. Commun. 13 (2022) 2280. doi: 10.1038/s41467-022-30027-x

  92. [92]

      N. Kornienko, Nanoscale 13 (2021) 1507–1514. doi: 10.1039/d0nr07508f

  93. [93]

      N. Heidary, K.H. Ly, N. Kornienko, Nano Lett. 19 (2019) 4817–4826. doi: 10.1021/acs.nanolett.9b01582

  94. [94]

      J. Li, N. Kornienko, Chem. Sci. 13 (2022) 3957–3964. doi: 10.1039/D1SC06590D

  95. [95]

      J. Li, H. Al-Mahayni, D. Chartrand, et al., Nat. Synth. 2 (2023) 757–765. doi: 10.1038/s44160-023-00303-9

  96. [96]

      Y. Zhu, J. Wang, H. Chu, et al., ACS Energy Lett. 5 (2020) 1281–1291. doi: 10.1021/acsenergylett.0c00305

  97. [97]

      M. Suvarna, J. Pérez-Ramírez, Nat. Catal. 7 (2024) 624–635. doi: 10.1038/s41929-024-01150-3

  98. [98]

      A.P. Côté, A.I. Benin, N.W. Ockwig, et al., Science 310 (2005) 1166–1170. doi: 10.1126/science.1120411

  99. [99]

      K. Geng, T. He, R. Liu, et al., Chem. Rev. 120 (2020) 8814–8933. doi: 10.1021/acs.chemrev.9b00550

  100. [100]

       X. Zou, Y. Zhang, Chem. Soc. Rev. 44 (2015) 5148–5180. doi: 10.1039/C4CS00448E

  101. [101]

       L. Stegbauer, K. Schwinghammer, B.V. Lotsch, Chem. Sci. 5 (2014) 2789–2793. doi: 10.1039/C4SC00016A

  102. [102]

       Z. Li, T. Deng, S. Ma, et al., J. Am. Chem. Soc. 145 (2023) 8364–8374. doi: 10.1021/jacs.2c11893

  103. [103]

       S.S. Zhu, Z. Zhang, Z. Li, et al., Chem. Catal. 4 (2024) 100963.

  104. [104]

       C. Krishnaraj, H. Sekhar Jena, L. Bourda, et al., J. Am. Chem. Soc. 142 (2020) 20107–20116. doi: 10.1021/jacs.0c09684

  105. [105]

       D. Chen, W. Chen, Y. Wu, et al., Angew. Chem. Int. Ed. 62 (2023) e202217479. doi: 10.1002/anie.202217479

  106. [106]

       R. Liu, Y. Chen, H. Yu, et al., Nat. Catal. 7 (2024) 195–206. doi: 10.1038/s41929-023-01102-3

  107. [107]

       W. Lin, F. Lin, J. Lin, et al., J. Am. Chem. Soc. 146 (2024) 16229–16236. doi: 10.1021/jacs.4c04185

  108. [108]

       Y. Fan, D.W. Kang, S. Labalme, et al., Angew. Chem. Int. Ed. 62 (2023) e202218908. doi: 10.1002/anie.202218908

  109. [109]

       Y. Luo, Y. Yan, S. Zheng, et al., J. Mater. Chem. A 7 (2019) 901–924. doi: 10.1039/c8ta08464e

  110. [110]

       X. Wang, K. Maeda, A. Thomas, et al., Nat. Mater. 8 (2009) 76–80. doi: 10.1038/nmat2317

  111. [111]

       Y. Chen, C. Yan, J. Dong, et al., Adv. Funct. Mater. 31 (2021) 2104099. doi: 10.1002/adfm.202104099

  112. [112]

       J. Yuan, N. Tian, Z. Zhu, et al., Chem. Eng. J. 467 (2023) 143379. doi: 10.1016/j.cej.2023.143379

  113. [113]

       F. Li, X. Yue, Y. Liao, et al., Nat. Commun. 14 (2023) 3901. doi: 10.1038/s41467-023-39578-z

  114. [114]

       H. Tan, W. Si, W. Peng, et al., Nano Lett. 23 (2023) 10571–10578. doi: 10.1021/acs.nanolett.3c03466

  115. [115]

       J. Liu, R. Zou, H. Zhang, et al., Appl. Catal. B: Environ. Energy 357 (2024) 124312. doi: 10.1016/j.apcatb.2024.124312

  116. [116]

       Y. Zhang, Q. Cao, A. Meng, et al., Adv. Mater. 35 (2023) 2306831. doi: 10.1002/adma.202306831

• 

  Figure 1  Diagram of the superhydrophobic mechanism of modified lignin. Reproduced with permission [8]. Copyright 2024, American Chemical Society.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Schematic overview of 2D carbon-based electrocatalysts with modification strategies toward efficient N₂ reduction reaction.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  N₂ adsorption geometry on polymeric carbon nitride with nitrogen vacancy (PCN–NV) shown in (a) top view, and (b) side view. (c) Charge density difference of the N₂-adsorbed PCN–NV. Reproduced with permission [46]. Copyright 2018, Wiley. (d) Schematic illustration of NRR for boron-doped graphene (BG). Reproduced with permission [47]. Copyright 2018, Cell Press.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Schematic structure of VP. (a) Top view of a monolayer. The unit cell is marked by red solid and dash lines. (b) Side view of a bilayer. Violet balls represent P atoms.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Open avenues to pursue leveraging the unique properties of 2D MOFs and COFs towards next-generation electrocatalytic systems.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  (a) Redox-active COFs for photocatalysis reactions. (b) Schematic structure of COF-JLU35 for photocatalytic water splitting for hydrogen production. Reproduced with permission [102]. Copyright 2023, American Chemical Society. (c, d) Schematic structure of TAPD-(OMe)₂ COF and Hz-TP-BT-COF for photocatalytic H₂O₂ synthesis. Reproduced with permission [104]. Copyright 2020, American Chemical Society. Reproduced with permission [106]. Copyright 2020, Springer Nature. Reproduced with permission [54]. Copyright 2023, Springer Nature. (e) Schematic structure of EPCo-COF for photocatalytic CO₂ reduction. Reproduced with permission [107]. Copyright 2024, American Chemical Society. (f) Schematic structure of NiCN COF for photocatalytic organic synthesis. Reproduced with permission [108]. Copyright 2023, Wiley.

  下载: 全尺寸图片 幻灯片
  

  Figure 7  Structural effect of (a) g-C₃N₄ and (b) crystalline g-C₃N₄ on charge transfer and recombination behaviour due to the presence of defect states.

  下载: 全尺寸图片 幻灯片
  

  Figure 8  Schematic illustration of recent strategies to improve charge dynamics and exciton dissociation efficiency in g-C₃N₄.

  下载: 全尺寸图片 幻灯片
•