English

• 1. Introduction

  The worldwide demand for energy is expected to continue increasing in the coming decades. The extensive burning of fossil fuels over the past centuries has resulted in the release of substantial amounts of greenhouse gases, which has aggravated global warming and the ecological crisis. It is projected that by the end of the 21^st century, the anthropogenic CO₂ levels will reach 950 ppm [1]. If atmospheric CO₂ concentrations surpass 600 ppm, it will cause a devastating catastrophe for the earth, with a sea level rise of 0.4 m. Numerous technologies, such as carbon storage, carbon capture, and carbon utilization have been developed to reduce atmospheric CO₂ concentrations. Converting CO₂ into high-value-added chemicals, such as fuels or other manufactured products, has emerged as a promising solution to combat global warming and the energy crisis [2,3]. Photocatalytic technology has been extensively employed for the treatment of pollutants in water or air, and has demonstrated significant efficacy in the degradation of emerging pollutants [4-7]. This precisely utilizes visible light to induce the production of more active substances, which facilitates the separation of photogenerated carriers for the removal of pollutants [8,9]. Carbon dioxide photocatalytic reduction (CO₂-PR) technology, relying on economical reactants (e.g., H₂O) and low energy requirements, has been recognized as a prospective approach to alleviate the energy crisis and global warming problem [10,11].

  Tungsten oxide-based (W_xO_y-based) materials have garnered significant attention in CO₂-PR research. Typical W_xO_y materials, such as non-stoichiometric W₁₈O₄₉ and WO₃·H₂O, possess abundant defect sites that contain an excess of electrons compared to other defective semiconductors. These photocatalysts exhibit the localized surface plasmon resonance (LSPR) effect, which involves exciting the excess electrons [12]. The LSPR effect in WO_3-x extends its photoresponse to the near-infrared range, thereby enhancing the photoconversion efficiency and charge separation efficiency [13]. As a result, W_xO_y-based materials are able to exhibit excellent catalytic performance in CO₂-PR. However, photocatalysts are often unable to effectively realize the photoconversion of CO₂ because of the limitations of their own structures and properties. Scientists have proposed diverse methods, such as defect engineering [14], structure control [15], morphology control [16], and the introduction of other metals or metal oxides to construct composite semiconductors and boost photocatalytic activity.

  While few reviews have delved into the improvement of CO₂-PR performance through modification of W_xO_y-based catalysts, this paper aims to review recent research progress in this field. Specifically, we focus on different product perspectives, analyze and compare the influencing factors of CO₂-PR activity based on variations in photocatalytic systems across different studies, summarize and forecast the application of W_xO_y-based catalysts in the realm of CO₂-PR. We firmly believe that this review will facilitate future design and study of CO₂-PR, as well as encourage further exploration of W_xO_y-based photocatalytic systems for a wide range of applications.

  2. Basic principle of CO₂-PR

  2.1 Models of CO₂ adsorption and activation on photocatalysts

  CO₂ is thermodynamically stable due to its centrosymmetric linear molecular structure and fully oxidized chemical state (ΔH_f⁰ = −393.51 kJ/mol), characterized by an O═C═O linear structure with σ and π bonds [17]. The bond energy of C═O (750 kJ/mol) is significantly higher than that of C—O (327 kJ/mol), C—H (411 kJ/mol), and C—C (336 kJ/mol) bonds [18]. CO₂-PR is a highly endothermic process with an energy barrier that must be overcome to cleave the C═O bond and excite CO₂-PR.

  In the original CO₂ adsorption activation theory, an ideal model was proposed based on metal/metal oxide photocatalysts (Figs. S1a and b in Supporting information) [19,20]. Typically, photocatalysts possessing mesoporous structures offer larger surface areas and pore volumes, thereby facilitating an increased number of active sites for CO₂ adsorption [21,22]. Moreover, CO₂ functions as an acidic oxide, and its carbon atoms, acting as Lewis acids, can coordinate with Lewis base sites on the photocatalyst surface, which are represented by oxygen atoms acting as Lewis bases [23]. This coordination results in the distortion of CO₂’s linear structure and consequent generation of partially charged CO₂ δ species [19]. Throughout the structural transformation from linear to bent conformation, the lowest unoccupied molecular orbital (LUMO) of the CO₂ molecule is downshifted, thereby promoting charge transfer between the catalyst and CO₂^•− species [24].

  The CO₂-PR pathway and its product selectivity depend strongly on the surface chemistry of CO₂ adsorption on the surface of the photo-/co-catalyst and desorption of the intermediates, which has been comprehensively summarized in other reviews (Fig. S1c in Supporting information). The crucial step in the reaction is the activation of the first C—O bond within the CO₂ molecule, and efficient electron transfer to CO₂ facilitates its activation [25,26]. It has been extensively shown that incorporating catalytically active sites comprising metals and oxides facilitates CO₂ activation and the stabilization of intermediates. In general, there are several approaches to stimulate CO₂ activation:

  (1) Defect engineering, such as introducing non-stoichiometric tungsten oxide (WO_3-x), which is rich in surface oxygen vacancies (V_O), can lead to the generation of LSPR in the visible and infrared regions. This unique characteristic enhances WO₃’s light-harvesting ability, resulting in improved carrier separation [13]. Defects also provide additional adsorption sites for CO₂ activation, thereby reducing reaction barriers and modifying the product distribution [27].

  (2) Modulating surface basic sites offers an alternative approach for CO₂ adsorption and activation [28]. Alkaline modification of W_xO_y-based catalysts’ surface enhances CO₂ adsorption performance. The reaction between the acidic CO₂ and the alkaline surface of the photocatalyst results in the formation of intermediates, such as bicarbonate, favoring the activation and subsequent reduction of CO₂ molecules.

  (3) The decoration of metal co-catalysts has been demonstrated to promote CO₂ activation [29]. When a W_xO_y-based catalyst is excited, the generated electron-hole pairs separate, with the electrons and protons transferred to the catalyst or co-catalyst surface. The electrons are involved in the reduction of CO₂ at the active sites of the W_xO_y-based catalyst or co-catalyst, while the protons participate in the oxidation reaction of water.

  2.2 General reaction pathways for CO₂-PR

  2.2.1   C1 production

  CO₂-PR involves a series of fundamental steps, including electron/proton (e⁻/H⁺) transfer, C═O bond cleavage, intermediate formation, and new bond formation [30].

  The photo-generated electrons in the conduction band (CB) can engage in CO₂-PR to produce renewable hydrocarbon fuels (equation: CO₂ + H₂O → hydrocarbons + O₂) under mild reaction conditions, typically at room temperature and pressure. Conversely, the holes in the valence band (VB) generally take part in the oxidation of water [31]. Furthermore, CO₂-PR exhibits a wide range of potential catalytic products, including CO, HCOOH, CH₂O, CH₃OH, CH₄, C₂H₅OH, and others. The reaction can involve a catalytic process requiring 2–12 electrons (Eqs. 1–8). In the following equations, E⁰ represents the potential of the electrode under standard hydrogen electrode reference conditions.

  (1)

  (2)

  (3)

  (4)

  (5)

  (6)

  (7)

  (8)

  Habisreutinger et al. [32] summarized three potential mechanisms for CO₂-PR to generate C1 products (CO, HCOOH, CH₂O, CH₃OH, and CH₄), as well as some C2 products, including formaldehyde pathway, carbene pathway, and glyoxal pathway. The differentiation among these pathways arises from the various modes of CO₂ adsorption on the surface of the catalysts, encompassing oxygen coordination, carbon coordination, and lateral/mixed coordination. Regardless of the pathway, the reaction commences with the adsorption and activation of CO₂, leading to the formation of charged CO₂^•− species on the catalyst's surface. The discrepancies in these pathways stem from the distinct manners in which CO₂^•− binds to the catalysts. The final product of all three pathways is CH₄, and CH₃OH is produced in both the formaldehyde pathway and the carbene pathway.

  However, Koci et al. [33] found that the distribution of CH₄ and CH₃OH did not align with the formaldehyde pathway alone, as CH₃OH is not an intermediate in the production of CH₄. To address this, Wang et al. [34] proposed a new formyl pathway (Fig. S2 in Supporting information) that integrates the formaldehyde, carbene, and glyoxal pathways. In the formyl pathway, if the catalyst possesses moderate adsorption strength, the C-anchored CO intermediate can accept an e⁻, while a H⁺ attacks the C atom to form CHO. With the assistance of another e⁻/H⁺, CHO can be further converted into CH₂O. Subsequently, the intermediate becomes successively bonded to C by binding with the e⁻/H⁺ until it eventually produces CH₃OH or CH₄.

  2.2.2   C2 production

  C2 production have garnered significant attention due to their higher economic benefits, making the direct generation of C2 production through CO₂-PR a current hot research topic. The synthesis of the C2 product poses a greater challenge compared to the synthesis of the C1 product due to the requirement of a higher concentration of electrons in the active site [35]. The key to further generating C2 products lies in the dimerization of C1 intermediates on the catalyst surface to form C—C bonds. Furthermore, it is crucial for the intermediates to strongly bond to the catalyst surface to generate C2 products instead of early desorption as C1 products, such as CO and CH₂O. Fig. S3 (in Supporting information) depicts three pathways for generating C2 and C₂₊ products.

  2.3 Working parameters

  The CO₂-PR process is influenced by various parameters, including temperature, medium pH, CO₂ partial pressure, light intensity and wavelength, H₂O partial pressure and reaction system. These parameters play a significant role in determining the CO₂-PR conversion and product selectivity. Consequently, this subsection provides a comprehensive analysis of the influence mechanisms associated with different reaction conditions.

  2.3.1   Temperature

  The initiation of a photocatalytic reaction necessitates a driving force that generates the electron-hole (e⁻-h⁺) pair. This driving force is provided by light irradiation in the form of the Gibbs free energy (ΔG). Previous studies have observed that the rate of CO₂-PR increases at higher temperatures due to the augmentation in collision frequency and diffusion rate of the reactant molecules [36]. This implies that higher temperatures are more conducive to enhancing the mass transfer rate of CO₂ from the surface of the photocatalyst, consequently improving the overall reaction rate. Additionally, higher temperatures expedite the reaction rate by reducing the activation energy needed to initiate chemical reactions.

  2.3.2   pH value

  The solvent's pH is a crucial factor in enhancing the rate and efficiency of CO₂-PR. Variations in protons concentration within the reaction medium influence the theoretical reduction potential of CO₂ when it functions as both a reactant and a buffer. Additionally, high pH favors the decomposition of CO₂ into CO₃²⁻ or HCO₃⁻, which possess varied standard redox potential values, affecting their adsorption properties on the photocatalyst surface and consequently leading to different product yields and selectivity. For instance, Das et al. [37] discovered that higher pH levels are correlated with increased CO₂ solubility, elevated CO₂ concentration near the catalyst's surface, and improved CO₂ reduction efficiency. However, excessively high pH levels restrict the supply of surface adsorbed H⁺ to the adsorbed ^•CO₂, thereby hindering further CO₂-PR.

  2.3.3   CO₂ partial pressure

  The photoconversion rate of CO₂ is heavily influenced by the adsorption rate of CO₂ molecules on the photocatalyst's surface. This adsorption process relies on the van der Waals force governing the interaction between the adsorbent (photocatalyst) and the adsorbate (CO₂). This force acts as an adhesion force within the adsorbate-adsorbent system and strengthens with increasing partial pressure of the adsorbate. Consequently, higher partial pressure of CO₂ gas leads to better product yields when CO₂ is in the gas phase. In both gas-phase and liquid-phase systems, the partial pressure of CO₂ (P_CO) regulates the likelihood of adsorption on the photocatalyst's active sites, thereby ensuring higher yields of solar fuels.

  2.3.4   Light intensity and wavelength

  Regarding energy considerations, light intensity and wavelength play pivotal roles in photocatalytic reactions. The energy possessed by the excited e⁻-h⁺ pairs relies on the wavelength of light, while the number of e⁻-h⁺ pairs is contingent upon the intensity of light. Research conducted by Ikreedeegh et al. [38] observed that the CO₂-PR demonstrated heightened activity with increasing light intensity, yet decreased activity when exposed to longer light wavelengths. Higher-power lamps emit a greater quantity of photons and generate a larger number of e⁻-h⁺ pairs. However, even with high-power lamps, photogenerated e⁻ and h⁺ can still undergo rapid recombination, thereby diminishing photocatalytic efficiency [39].

  2.3.5   HER

  The process of H₂O splitting plays a crucial role in CO₂-PR, where H₂O molecules interact directly with excited state e⁻ in the CB and h⁺ in the VB to generate H₂ and O₂ [40]. The recombination of the e⁻-h⁺ pairs can be prevented by h⁺ consuming OH⁻ species [41]. A minimum band gap of 1.23 eV is required for the photocatalytic decomposition of water. In the case of W_xO_y-based catalysts, the bandwidth ranges from 2.5 to 2.8 eV. As a result, the hydrogen evolution reaction (HER) can compete with CO₂-PR due to its higher kinetic feasibility. Furthermore, the slow diffusion of CO₂ molecules on the catalyst surface and the involvement of H₂O in generating H₂ during the reaction can decrease the availability of H⁺ for CO₂ reduction, thereby reducing both the activity and product selectivity of CO₂-PR. Consequently, photocatalysts must exhibit elevated activity for H₂O oxidation while limiting excessive H₂ generation.

  2.3.6   Reaction system

  CO₂-PR is conducted in gas-liquid-solid or gas-solid systems. In gas-liquid-solid systems, co-solvents are added to increase the solubility of CO₂, as well as sacrificial agents to inhibit competing half-reactions, making it easier to detect liquid-phase products, including formic acid. For gas-solid systems, the yield of gaseous products such as CH₄ is relatively high. For smooth CO₂ hydrogenation, a balance of water vapor and CO₂ molecule adsorption on the surface of the material is required, as well as the ability to inhibit the HER [42]. Additionally, without extra sacrificial agents, the surface of the catalyst is exposed to the accumulation of ·OH after the depletion of H⁺, leading to the deactivation of the catalyst. Recently, three-phase photocatalysis has emerged as a promising solution to address mass transfer limitations in CO₂-PR. The efficiency of three-phase photocatalysis lies in its ability to rapidly transport CO₂ to the catalyst surface [43]. Moreover, e⁻-h⁺ pairs play a pivotal role in regulating CO₂-PR, and their effective separation is achieved through external or internal electric fields [44]. Despite these advantages, challenges persist in the photoreduction of insoluble stable gas molecules, including hindered kinetic diffusion, sluggish surface adsorption, and inefficient activation of gaseous reactants at three-phase catalytic interfaces [45].

  To achieve higher yields and superior product selectivity in CO₂-PR, careful selection of all the parameters is necessary. Moreover, there is a need for not only efficient and robust photocatalysts, but also sustainable interfacial photocatalytic systems.

  3. W_xO_y-based materials

  Although W_xO_y-based materials possess appropriate band gaps, they exhibit limited catalytic reactivity, including low sunlight utilization, low surface negative charge density, fast recombination of photogenerated e⁻-h⁺ pairs, and low carrier migration efficiency. Various approaches have been explored to enhance the efficiency of photocatalytic systems, such as morphology engineering, elemental doping, and heterojunction formation [46-48]. The W_xO_y-based photocatalysts for CO₂-PR are showed in Table S1 (Supporting Information). These modifications will be discussed in greater detail below.

  3.1 Defects/phase/facet

  For metal-oxide semiconductors, an alternative and cost-effective method to increase the surface negative charge density and improve CO₂-PR performance is by constructing surface V_O. The presence of V_O serves two important functions: (Ⅰ) It promotes the activation of reactants, leading to the formation of active substances; (Ⅱ) It expands light absorption, thereby enhancing the efficient utilization of solar energy [49]. While CO₂-PR driven by WO₃ is mainly dominated by C1 products (such as CO and CH₄), WO_3-x can enhance the efficiency of C—C coupling and reduce CO₂ to C2 products.

  Although V_O exhibits better ability in increasing the negative charge density on the surface, it is easily inactivated through oxidation due to its typically low concentration on the surface. However, Sun et al. [50] successfully synthesized ultrathin WO₃·0.33H₂O nanotubes with a high level of exposed surface V_O. During the generation of CH₃COOH, the V_O that adsorbed CO₂ could be easily separated from V_O after forming C—C bonds, enabling sustainable utilization of V_O and suppressing V_O inactivation. Moreover, Deng et al. [51] prepared ultrathin WO_3-x nanosheets with oxygen defects. Theoretical calculations confirmed that these defects are crucial in the reduction of CO₂ to C2 hydrocarbons, and V_O promoted the C—C coupling of the reduced C—H intermediates, effectively preventing the complexation of photogenerated e⁻ and h⁺, thereby improving the conversion efficiency of WO_3-x to CO₂.

  The presence of photo-induced V_O is a double-edged sword. While a small amount of V_O enhances the photocatalytic activity, an excessive amount of V_O can disrupt the hexagonal WO₃ structure, leading to a decrease in photocatalytic activity. Wang et al. [52] successfully synthesized V_O-containing samples by treating hexagonal WO₃ in an atmosphere containing H₂ and demonstrated that this V_O generation process is reversible when the catalyst is exposed to an atmosphere containing O₂. The photocatalytic activity significantly increases with the increase in V_O content, reaching its maximum value.

  Stoichiometric WO₃ is comprised of edge-shared octahedral WO₆. It exhibits heterogeneous phases, including monoclinic, tridiagonal, orthorhombic, tetragonal, hexagonal, and cubic phases. The specific phase depends on the angle of lattice tilt and the direction of rotation [53]. In room temperature, monoclinic I has the most stable structure, which is a typical n-type semiconductor with a band gap of 2.62 eV. W₁₈O₄₉ has a unique monoclinic structure and has been widely studied for its ability to exhibit a strong LSPR effect and excellent CO₂-PR performance. Monoclinic and hexagonal crystals are the most promising options among the various crystalline phases. The monoclinic phase is known for its stability, while the hexagonal phase possesses an open-hole structure and abundant intercalation chemistry [54]. Both WO_3-x/GdCrO₃ and WO_3-x-R/GdCrO₃ consist of monoclinic crystals W₁₈O₄₉ and orthorhombic crystals GdCrO₃. The close contact between WO_3-x-R and GdCrO₃ is clearly observed in the HRTEM image (Fig. 1a). The growth of WO_3-x nanosheets with plentiful defect sites on GdCrO₃ along the [−1,0,4] direction is evident. Simultaneously, the crystal surface of WO_3-x can be meticulously crafted to expose a greater number of active sites [55]. At the WO_3-x-R/GdCrO₃ interface, W^5/4+ defect sites are introduced, serving as a bridge for photothermal-induced charge storage, secondary plasmon excitation, and interband transfer, as seen in Fig. 1b. This approach enables efficient charge separation, high redox potential, and sufficient absorption of visible light [56].

  Figure 1

  

  Figure 1.  (a) HETEM of WO_3-x-R/GdCrO₃, and (b) W 4f XPS spectra of the samples. Reprinted with permission [56]. Copyright 2022, Elsevier. (c) CO₂ photoelectrocatalytic performances of WO₃-raw and WO₃-T, and (d) XPS W 4f profiles of WO₃-raw and WO₃-T. Reprinted with permission [57]. Copyright 2020, Elsevier. The aldehyde selectivity of the catalysts is represented by dots, SEM (e) and HAADF-STEM (f) images of WO_3-x-2 synthesized under 2 mL HCl assistance, Raman spectra (g) of WO_3-x-2 during CO₂-PR, and their IR spectra (h) after photocatalysis. Reprinted with permission [58]. Copyright 2020, Elsevier.

  DownLoad: Full-Size Img PowerPoint

  Non-stoichiometric tungsten oxide (WO_3-x) possesses a high density of V_O and shows LSPR in the visible-near-infrared region. The presence of V_O in WO_3-x results in broad absorption in both the visible and infrared regions. Li et al. [57] observed a gradual transition from WO₃ to W₁₈O₄₉ during carbothermal reduction at temperatures higher than 800 ℃. The introduction of W⁵⁺ on the surface of non-stoichiometric WO₃ facilitated the selective conversion of CO₂ to aldehyde with a 35% yield (Fig. 1c). This phase transition led to the presence of low-valent surface components, such as W⁵⁺ and W⁴⁺, alongside the initial W⁶⁺ material in different ratios (Fig. 1d), which assembled the surface structure of non-stoichiometric WO₃-T. Lu et al. [58] performed plasma synthesis of WO₃ with tunable Vo and low-valence W ions. They observed a sea urchin-like structure of WO_3-x-2 (Fig. 1e), with a clear and ordered lattice structure corresponding to the (010) face observed by atomic resolution high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) (Fig. 1f). Raman and IR (Fig. 1, Fig. 1) spectroscopy revealed the presence of “C═C”, “C—C”, adsorbed CO, HOCOO, and adsorbed CO₂ intermediates on the WO_3-x-2 surface, indicating that adjacent V_O provide active sites for C—C coupling to ethylene.

  Surface modification of W_xO_y-based photocatalysts using the characteristics of WO₃, such as the introduction of abundant defect sites and the dominant monoclinic phase structure, can significantly enhance the CO₂ adsorption capacity and provide reactive active sites for the selective production of C1 and C2 products.

  3.2 Doping

  Exogenous ion doping is a simple and effective method to modulate the electronic structure of photocatalysts, altering the valence state of the element and facilitating the introduction of low-valence metal species into WO₃ [59,60]. Furthermore, elemental doping can induce the formation of intermediate energy bands or V_O, optimizing the band structure of the photocatalysts [61]. Importantly, since the CO₂-PR occurs on the catalyst surface, the surface doping of metal ions on the photocatalyst can promote CO₂ adsorption capacity and the transfer of photogenerated carriers [62].

  For instance, Wang et al. [63] synthesized Mo-doped WO₃·0.33H₂O nanorods. As illustrated in Fig. 2a, the electron paramagnetic resonance (EPR) spectra demonstrated a significant increase in the V_O content of Mo-doped WO₃·0.33H₂O, which facilitated the supply and insertion process of H⁺ and improved the hydrogenation process of carbon intermediates for CH₄ generation. This ultimately resulted in an increase in CH₄ yield, as shown in Fig. 2b. And Fig. 2c shows the scheme of Mo-doped WO₃·0.33H₂O CO₂ reduction reaction with H₂O driven by solar energy. Zhang et al. [64] doped Cu⁺ into the W₁₈O₄₉ nanowires. Analysis of the fitted peaks of W 4f revealed an increase in the content of W⁵⁺ after Cu⁺ doping. In addition, the peaks of W 4f shifted towards high binding energies, as depicted in Fig. 2d. The EPR spectrum (Fig. 2e) confirmed the increase in V_O content due to Cu⁺ doping, with the EPR signal of Cu-W₁₈O₄₉–0.005 being significantly higher than that of pure W₁₈O_49. Cu⁺ served as an active site for CO₂ activation, shifting its conduction band to a more negative position and served as an electron transfer center, reducing e⁻-h⁺ pair recombination and enhancing charge transfer.

  Figure 2

  

  Figure 2.  (a) EPR spectra at 300 K for WO and 3% Mo-WO, (b) The yield of CO₂-PR to CH₄ in the first 6 h for WO, 1% Mo-WO, 3%Mo-WO and 5%Mo-WO, and (c) CO₂-PR over Mo-WO with water under solar irradiation. Reprinted with permission [63]. Copyright 2019, Elsevier. (d) W 4f high-resolution XPS spectra and (e) room-temperature EPR spectra of W₁₈O₄₉ and Cu-W₁₈O₄₉–0.005. Reprinted with permission [64]. Copyright 2020, Elsevier.

  DownLoad: Full-Size Img PowerPoint

  Yi et al. [65] discovered that the co-doping of Mo and Cs in reduced WO₃ improved CO₂-PR performance compared to solely modified WO₃. The band gap of CsWO was slightly narrowed by Mo doping, increasing the light absorption capacity of CsWO and inducing a reductive Mo⁵⁺ for the LSPR effect. Therefore, narrowing the band gap of W_xO_y-based materials is also an important strategy for enhancing CO₂-PR performance.

  Doping metal elements into W_xO_y-based photocatalysts serves to modulate the edge positions of VB and CB, while also inhibiting carrier recombination and promoting the formation of V_O. This synergistic interaction with neighboring ligand-unsaturated metal ions, hydroxyl groups, and ions enhances the photocatalytic performance.

  3.3 Composite materials

  Single-component photocatalysts face a challenge in providing both strong redox capacity and a broad-range photoresponse. This is because achieving the former requires a large bandgap, while the latter necessitates a narrow bandgap [66]. Moreover, in single-component nanophotocatalysts, the generated e⁻-h⁺ pairs are often prone to recombination due to strong coulombic interactions. It will result in a significant loss of photogenerated charge [67]. The construction of semiconductor heterojunctions has proven to be an effective strategy. Heterojunctions enable a wide range of photoresponse and promote efficient charge separation, thereby enhancing the strong redox capacity that facilitates CO₂-PR [68].

  3.3.1   Binary composites

  Among the semiconductors with high VB potentials, WO₃ has attracted significant attention from researchers due to its high specific surface area, high visible light activity, and high oxidation potential. However, WO₃ has positive CB position, which is ineffective for CO₂ reduction. Additionally, its narrow band gap leads to rapid recombination of charge carriers, resulting in poor photocatalytic activity. To address these limitations, WO₃ can be combined with semiconductors that have negative VB positions to form composite heterojunctions, thus facilitating efficient separation of e⁻-h⁺ pairs and enhancing the performance of CO₂-PR.

  WO₃ is commonly used as the oxidation component in Z-scheme heterojunction composites due to its positive VB potential and ability to oxidize photogenerated holes. For example, Tahir et al. [69] created Z-scheme WO₃/g-C₃N₄ heterojunction that allowed efficient interaction between the two semiconductors and achieved effective carrier separation. Jiang et al. [70] found that good performance of the CsPbBr₃/WO₃ heterojunction was attributed to the Z-scheme mechanism with a strong internal electric field. Tang et al. [71] designed a three-dimensionally ordered macroporous (3DOM) structure of WO₃ with a slow photon effect. Additionally, they grew g-C₃N₄ thin films on the surface of 3DOM-WO₃ in situ. This approach effectively addresses the limitation caused by the mismatch between the conduction band position of conventional semiconducting materials and the required redox potential for the reaction. Consequently, directional conduction of electrons between the two semiconductor materials is achieved, enhancing the efficiency of visible light trapping. These studies demonstrate the potential of using WO₃ as an efficient visible light-active co-catalyst for constructing heterojunctions for CO₂-PR.

  Yu et al. [72,73] proposed novel S-scheme heterojunction photocatalysts that using a distinct electron transfer mechanism compared to conventional type Ⅱ and Z-scheme heterojunctions. Comparatively, S-scheme heterojunctions consisting of a reduced semiconductor and an oxidized semiconductor exhibit improved charge separation and enhance the kinetic of CO₂-PR. Additionally, they fulfill the thermodynamic requirements for CO₂-PR [74,75]. For instance, the ultrasound-assisted Cs₂AgBiBr₆/Bi₂WO₆ nanocomposites, featuring a staggered-band arrangement of S-scheme heterojunction structures, and the one-dimensional S-scheme heterojunction structure of SiC-W₁₈O₄₉, characterized by a distinctive sea urchin morphology, both demonstrate enhanced CO₂-PR performance compared to individual materials [76,77].

  Recently, the use of 2D/2D interfaces has demonstrated effectiveness in providing larger contact interfaces, reducing carrier transport distances, and increasing the availability of active sites [78,79]. Jiang et al. [80] exploited the synergistic 2D/2D interface (Fig. 3a), Vo-induced defect energy levels (Fig. 3b and c), achieved a significant CO evolution rate increase in the 30V_O-WO_x/BiOCl, which is 4.68 and 3.09 times higher than that of BiOCl and 30WO₃/BiOCl, respectively (Fig. 3d and e). And S-scheme transfer modes of V_O-WO_x/BiOCl composites (Fig. 3f) to achieve faster V_O-WO_x and BiOCl dual transfer modes through carrier separation and improved charge separation and transfer efficiency.

  Figure 3

  

  Figure 3.  (a) HRTEM of the 30V_O-WO_x/BiOCl sample, high-resolution XPS spectra of Bi 4f (b), W 4f (c) of the BiOCl, V_O-WO_x and 30V_O-WO_x/BiOCl samples in the dark and under the light, (d) the CO evaluation of BiOCl, V_O-WO_x. 30WO₃/BiOCl and 30V_O-WO_x /BiOCl, (e) CO production rates of BiOCl, 30WO₃/BiOCl and 30V_O-WO_x/BiOCl, and (f) S-scheme heterojunction composed by BiOCl and V_O-WO_x. Reprinted with permission [80]. Copyright 2022, Elsevier. (g) HRTEM image of the Au/TiO₂/W₁₈O₄₉ plasmonic heterostructure, (h) TEM image of the side view of the Au/TiO₂/W₁₈O₄₉ plasmonic heterostructure, (i) schematic diagram showing the hot electron generation and transfer processes due to the enhanced plasmon coupling between Au and W₁₈O₄₉ under UV–vis–NIR light irradiation, and (j) gas productions and the corresponding CH₄ selectivity through CO₂-PR over the as-synthesized samples under UV–vis–NIR light irradiation: (1) TiO₂ NFs, (2) Au/TiO₂ composite NFs, (3) Au/TiO₂/W₁₈O₄₉ plasmonic heterostructure, (4) W₁₈O₄₉ NWs, (5) W₁₈O₄₉/TiO₂ plasmonic heterostructure. Reprinted with permission [84]. Copyright 2022, Wiley-VCH.

  DownLoad: Full-Size Img PowerPoint

  In addition, a distinctive feature of WO₃ is its ability to have carefully designed V_O states that exhibit broad and strong absorption in the near-infrared (NIR) region [81]. Huang et al. [82] constructed plasma-enhanced S-scheme WO/CdSe-D photocatalysts. The abundant V_O in WO₃ induced LSPR effect, extending the light absorption range of WO/CdSe-D to the NIR region, thus enhancing absorption in the visible and NIR regions. Therefore, the rational utilization of WO₃’s properties offer the potential to develop composite photocatalysts for NIR CO₂-PR.

  Binary semiconductor composites offer several advantages, including improved optical response range, enhanced separation and transport of photogenerated carriers, and increased reactive interfaces and active sites. CB position of W_xO_y enables the effective electron transport and separation of photogenerated carriers when combined with semiconductor materials such as TiO₂ and BiVO₄, which have a negative VB position. As a result, these composites can significantly enhance CO₂-PR activity.

  3.3.2   Ternary composites

  In addition to binary composites, more complex structures, such as ternary catalysts, have also been reported, which can be derived from one or a combination of the structures. Ternary composites have garnered significant attention from researchers due to their ability to facilitate the rapid migration of excited electrons [83].

  Studies have demonstrated that ternary components exhibit superior photocatalytic activity compared to binary components. Jiang et al. [84] proposed the concept of plasma-active “hotspot” confined photocatalysis, where the coexcited LSPR of Au and W₁₈O₄₉ components form metal/non-metal coupled plasma coupling hotspots within a sandwich-like sub-structure of the Au/TiO₂/W₁₈O₄₉ plasma heterostructure (Figs. 3g and h). In comparison to a single-plasma-component system (Au/TiO₂ or W₁₈O₄₉/TiO₂), the two-plasma-component system (Au/TiO₂/W₁₈O₄₉) can generate more thermally excited electrons for CO₂-PR (Fig. 3i) when the LSPR of Au and W₁₈O₄₉ components and the interband leaps of the TiO₂ component are co-excited. This configuration leads to efficient CO₂-PR (Fig. 3j). In the ternary system constructed by Dong et al. [85], the WO₃/CsPbBr₃ Z-scheme heterojunction provided an effective pathway for the transfer of photogenerated charge carriers. This ternary system was further modified with ZIF-67 to enhance charge separation and improve CO₂ adsorption performance. The evaluation of CO₂-PR activities revealed that after 3 h of irradiation, the CO yield of WO₃/CsPbBr₃ was 3.6 times higher than that of bare CsPbBr₃. Further modification with ZIF-67 resulted in the highest CO yield of 99.38 μmol/g, which was 6.8 times higher than that of CsPbBr₃. The porous structure also contributed to a higher specific surface area, increased adsorption sites, and multiple transfer pathways during the photoreduction process. These results highlight the significant contribution of ternary components to the enhancement of CO₂-PR performance.

  While multicomponent catalysts demonstrate clear advantages in improving CO₂-PR activity compared to single-component catalysts, the complexity of their preparation procedures raises the need for further investigation regarding whether the composite between various materials can achieve the theoretical effects. Considering factors such as cost and photocatalytic efficiency, current research on binary composites remains more extensive. However, there are still numerous possibilities awaiting exploration in the realm of ternary composites.

  4. Performance evaluation of W_xO_y-based catalysts

  4.1 Yield

  The metal atoms were incorporated into the W_xO_y-based catalysts through surface modification engineering to enhance the separation and transfer efficiency of CO₂-PR charges. Pd underwent the H₂ spillover effect to form hydrotungsten bronze, which facilitated H₂ dissociation and spillover. WO₃, on the other hand, stored H⁺ and e⁻ as hydrotungsten bronze within its lattice [86]. Furthermore, the Co-modified W₁₈O₄₉ demonstrated efficient carrier separation with positive facilitation. The isolated Co atoms acted as an electron pump, promoting charge separation and transport efficiency, and accelerating CO₂-PR kinetics [87]. Both modifications resulted in increased photocatalytic efficiency for bulk CO production, yielding 7.5 and 21.18 mmol g^–1 h^–1, respectively. Moreover, the study revealed that loading Pd and Au bimetals into TiO₂-WO₃ as CO₂-PR active sites effectively suppressed the complexation of photoexcited e⁻-h⁺ pairs [88].

  The activity of CO₂-PR is influenced by various factors, including temperature, light intensity, reaction time, catalyst loading, H₂O/CO₂ molar feed ratio, among others. Yi et al. [65] observed that the co-doping of Mo and Cs with WO₃ effectively improved CO₂-PR efficiency. However, the samples exhibited the highest activity in producing CO and CH₃OH when the doping concentration of Mo was 5%, and the activity decreased with an increase in the doping concentration to 10%. This decrease may be due to excessive Mo doping, which can cause certain Mo sites to become capture centers, hindering charge separation. WO₃-TiO₂/Cu₂ZnSnS₄ composites deactivated during CO₂-PR due to the formation of residual organics. After continuous irradiation for 5 h, the yields of both reduction products slightly decreased [89]. Furthermore, reaction conditions, such as the concentration of protons from water vapor, can easily affect the yield of CH₄, which resulting in a more pronounced decrease in CH₄ yield compared to CO yield.

  Currently, many photocatalysts primarily produce C1 products. There are several photocatalysts that have shown high yields of CO and CH₄. For example, melamine foam-loaded 1.5 mg WO₃/CsPbBr₃ composites, placed in a 100 mL polytetrafluoroethylene reactor with the addition of 4 mL DI water, reacted under 300 W Xe lamp irradiation for 2 h. This resulted in a high CO yield of 351.83 mmol g^–1 h^–1 (Fig. S4a in Supporting information) [90]. In other experiments, Tahir et al. [69] found that a WO₃/g-C₃N₄ photocatalyst, loaded with 25 wt% WO₃ at a quantity of 150 mg, achieved a CH₄ yield of 226 mmol g^–1 h^–1. The reaction conditions included a temperature of 40 ℃, 20 W Xe HID lamp source irradiation, and a gas-solid H₂O/CO₂ feed ratio of 0.4 (Figs. S4b and c in Supporting information).

  In summary, to realize the optimal activity of tungsten-based catalysts on CO₂-PR, firstly, suitable mono/bimetallic co-catalysts should be found, such as Pd and Co, which can store H⁺ and e⁻. Searching for semiconductor materials that can narrow the band gap of W_xO_y to form heterojunctions for efficient photogenerated e⁻ separation and transport. Meanwhile, the parameters can be tuned in the reaction system to optimize the photocatalytic activity, and the optimal reaction conditions can be achieved by controlling the metal loading, temperature, pressure, saturated vapor pressure, and light intensity. The activity of photocatalysts in promoting CO₂-PR is influenced by several factors, hence the need for multiple optimizations to enhance CO₂-PR efficiency.

  4.2 Selectivity

  The efficiency of photocarrier separation, photoexcitation, and the energy-band structure influence the CO₂-PR activity and product selectivity of different tungsten-based materials [91]. Selectivity is usually calculated in two ways, one is the carbon balance (Eq. 9) and the other is the electron balance (Eq. 10). Assessing product selectivity is crucial in evaluating CO₂-PR activity.

  (9)

  (10)

  During the multistep surface-catalyzed reaction of CO₂ reduction, the adsorption and desorption characteristics of the reactants and intermediates are key kinetic factors influencing product selectivity. Feng et al. [92] proposed a novel “atomic-level single-site” strategy to achieve a CH₄ selectivity of 86.5%, exceeding that of pristine TiO₂ by an order of magnitude. Jiang et al. [84] found that unique bis-heteroactive sites (Au-O-Ti and W-O-Ti) embedded on the plasma-coupled hotspots of the Au/TiO₂/W₁₈O₄₉ plasma heterostructures could adsorb the produced CO and H⁺ and facilitate further reduction of CO to CH₄ through plasma-induced high-energy thermoelectronics.

  Meanwhile, the abundance of active sites on the surface of WO₃ enables the introduction of V_O, which contributes to its unique structural features that promote C—C coupling and selective generation of more C2 products during the CO₂-PR process. Lu et al. [58] demonstrated that WO_3-x-2, with the highest V_O content, exhibited optimal ethylene-generating activity and selectivity in CO₂-PR under full-spectrum light irradiation (89.3% selectivity). This selectivity was 20 times higher than that of WO_3-x-0, which had low V_O content and a thick amorphous surface (17.5% selectivity). The combined effect of UV, visible, and NIR radiation synergistically enhanced the activity and selectivity of plasma WO_3-x-2 in acetylene production (Figs. S4d and e in Supporting information). Shen et al. [93] prepared TiO_2-x and WO_3-y, which formed an effective Z-scheme heterojunction with V_O as the interfacial layer. The synergistic effect of defects and the heterojunction promoted the separation and transport of photogenerated carriers. Furthermore, when loaded with Pd as a co-catalyst, the selectivity of CH₄ was increased to 100%. The presence of the co-catalyst significantly enhanced the selectivity of CO₂-PR products (Fig. S4f in Supporting information).

  The selectivity of reduction products is strongly influenced by the CO₂-PR pathway, particularly in the case of multicarbon products. Structural modulation enables the utilization of abundant defect sites, which in turn induce C—C coupling. This modulation has several effects, such as accelerating the adsorption of CO₂ on the photocatalyst surface, enhancing the desorption of intermediates, introducing metal atoms to increase reactive sites, facilitating the transfer of photogenerated electrons and holes involved in CO₂-PR, constructing special heterojunction structures to improve carrier separation efficiency, and reducing the free energy barriers for the conversion of CO₂ into CH₄, CH₃COOH, and other products. These modifications allow for the selective generation of a greater variety of desired target products in CO₂-PR.

  4.3 Stability

  High interfacial catalytic activity is often accompanied by side reactions, such as photonic corrosion [94] and structural remodeling [95], which can lead to the deactivation of pre-designed catalysts. In the case of CO₂-PR, the selectivity of products and the adsorption of specific intermediates are highly correlated. However, strong adsorption can block the active site and alter the electronic structure of the catalytic surface, ultimately poisoning and deactivating the catalyst [96]. The stability of photocatalysts is crucial for their practical application and determines their ability to be reused without significant loss of activity [97]. Therefore, a fundamental principle in catalyst design is to balance the activity and stability of the catalysts.

  To assess the reusability of photocatalysts, researchers conducted multiple or cyclic degradation tests using the same composite. Yi et al. [65] demonstrated that the 5% Mo-CsWO samples maintained good photocatalytic stability after four testing cycles. XRD and XPS analysis revealed that the structures, surface properties, and compositions of the catalysts remained largely unchanged compared to the pristine sample, indicating excellent structural stability of Mo-CsWO. Similarly, W₁₈O₄₉@Co exhibited high durability in cycling tests, consistently demonstrating significant photocatalytic activity throughout each cycle. Characterization of the catalyzed W₁₈O₄₉@Co showed no noticeable changes in structure or loss of CO content during the reaction [87]. Jiang et al. [70] prepared a CsPbBr₃/WO₃ heterojunction, which showed only a slight 2.6% decrease in activity after four catalytic runs (2 h per cycle), while maintaining the same morphology and phase as the catalysts. This improved stability of CPB NCs in heterojunctions could be attributed to their Z-scheme charge transfer mode, where the photogenerated h⁺ of CPB was consumed by the photogenerated e⁻ of WO₃, preventing the accumulation of photogenerated h⁺ on the CPB side, which is susceptible to photocorrosion (Figs. S4g and h in Supporting information) [98].

  However, certain W_xO_y-based photocatalysts have shown limitations in terms of cycling stability during CO₂-PR. In a study by Zeng et al. [99], the 3% Cu/WO₃ catalyst maintained a stable crystalline structure after cycling experiments. However, a degradation in catalyst performance was observed during the cycling test. Residual products on the used 3% Cu/WO₃ catalyst weakened the adsorption strength of CH₃COOH for CO₂, and the reduction of Cu⁺ sites led to a decrease in its C—C coupling ability, ultimately reducing CO₂-PR activity. Moreover, Shi et al. [100] found that the CO₂-PR process induced an excess increase of W⁵⁺ in Cu₂O/WO₃–001 composites, which would lead to poorer performance of CO₂-PR even after electrochemical regeneration.

  Many W_xO_y-based materials possess remarkable stability in their structures, surface properties, and compositions, which is crucial to prevent degradation of the photocatalyst performance. Firstly, the control of active sites during the doping of metal elements is essential to ensure effective C—C coupling. This helps to avoid undesirable consequences such as oxidative deactivation of the photocatalyst due to insufficient V_O. Moreover, it is important to refrain from using reagent washing photocatalysts that affect the structure and catalytic performance of the photocatalysts. It is also worth noting that residual products remaining on the photocatalyst can influence the adsorption strength of CO₂ on its surface. Therefore, by refining the preparation methods of photocatalysts and enhancing the stability of their surface structures and properties during the compounding process, we can effectively improve the photocatalytic performance in CO₂-PR.

  4.4 AQY

  The activity of the CO₂-PR system is typically evaluated based on the rate of product generation or CO₂ conversion efficiency. However, normalizing the activity by unit photocatalyst mass (i.e., μmol h^–1 g^–1) is not appropriate since the reaction rate cannot be directly compared to the mass of the photocatalyst used. Additionally, it is not advisable to directly compare the reaction rates of photocatalysts reported by different research groups due to variations in experimental conditions. To effectively evaluate the activity of a specific photocatalyst, the measurement of the apparent quantum yield (AQY) in a reactive reduction process can be utilized.

  In the work of Lu et al. [58], the effect of light irradiation on WO_3-x-2 was studied by obtaining the apparent quantum efficiency (AQE) of CO₂-PR under monochromatic light using different filters. The AQE estimates for monochromatic NIR-excitation and UV-excitation were found to be higher than that of monochromatic visible-NIR-excitation, indicating that the LSPR effect with energy band structure excitation enhances CO₂ conversion under UV–visible-NIR radiation. The AQE at 365 nm of the WO_3-x/MoO_3-x heterojunction, prepared by Liu et al. [101], is found to be 3.6 times higher than that of pure MoO_3-x and 2 times higher than that of pure WO_3-x. Abundant V_O formed in the WO_3-x/MoO_3-x heterojunction provide a large number of reduction sites for CO₂-PR, thereby enhancing the CO₂ adsorption and activation ability.

  Currently, the highest reported AQY for W_xO_y-based photocatalysts is observed in the WO₃/g-C₃N₄ heterojunction in CO₂-PR applications. With 150 mg of WO₃/g-C₃N₄ under a light intensity of 20 mW/cm², the QY for CO₂-PR with H₂O at 40 ℃ was estimated as 3.829% for CH₄ and 0.733% for CO at an H₂O/CO₂ feed ratio of 0.40 [69]. These findings suggest that a higher H₂O/CO₂ feed ratio favors more CH₄ generation, while the RWGS reaction promotes CO generation under the same reaction conditions. The enhanced photoreactivity of WO₃/g-C₃N₄ heterojunctions for CO and CH₄ generation can be attributed to their larger exoactive surface area and more efficient carrier separation under visible light irradiation. Due to CO generation requires 2e⁻ while CH₄ requires 8e⁻, it is not possible to effectively analyze the photon flux utilization of CO₂-PR based solely on the generation rate. Therefore, it is more appropriate to calculate AQY when evaluating the performance of catalysts in different reaction systems.

  5. Conclusion and perspective

  Due to the growing interest in renewable resources, particularly solar energy, and the pressing need to convert increasing CO₂ greenhouse gas emissions into organic fuels, the exploration of W_xO_y-based materials has become a prominent area of research in CO₂-PR. This review paper aims to provide a comprehensive overview of the fundamental principles and products of CO₂-PR. Additionally, this paper examines various strategies that have been employed to enhance the photocatalytic efficiency of W_xO_y-based materials for CO₂-PR applications. While a significant body of research has demonstrated the promising potential of W_xO_y-based photocatalysts in CO₂-PR, there are still several challenges need to be addressed before their practical implementation. These challenges can be summarized into three main areas:

  (1) W_xO_y-based photocatalysts continue to face challenges such as limited sunlight absorption and utilization efficiency, structural instability, carrier reorganization, and high redox capacity in the CO₂-PR process. Thus, it is essential to utilize the structural characteristics of W_xO_y which is abundant in V_O, enhance the role of active sites in promoting CO₂-PR process, fabricate additional heterojunctions to enhance both the efficiency and stability of W_xO_y-based materials in CO₂-PR.

  (2) Another important aspect on W_xO_y-based photocatalysts is the selective production of high-value-added products. Currently, most W_xO_y-based photocatalysts for CO₂-PR primarily yield CO and CH₄ as products. The selective generation of high-value-added products through CO₂-PR poses a challenge. Hence, further attention should be directed towards studying W_xO_y-based materials and their role in promoting the generation of C2 products.

  (3) Currently, the majority of studies evaluating the CO₂-PR activity primarily focus on the yield of products while often neglecting the carbon conversion efficiency. However, it should be noted that the reaction rate of CO₂-PR cannot be directly compared to the quality of the photocatalysts. Normalizing the activity to unit mass of the photocatalyst (i.e., μmol g^–1 h^–1) does not provide a proper assessment. As one of the essential factors for effectively assessing photocatalyst activity, calculating the CO₂ conversion efficiency allows us to visualize the catalytic performance exhibited by W_xO_y-based catalysts in CO₂-PR.

  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

  Weidan Meng: Writing – original draft. Yanbo Zhou: Methodology, Project administration, Supervision. Yi Zhou: Conceptualization, Project administration, Resources, Supervision, Validation, Writing – review & editing.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (No. 22376065), the Science and Technology Commission of Shanghai Municipality (No. 22ZR1418600), Shanghai Municipal Science and Technology (No. 20DZ2250400).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.109961.

  
  
• 1. [1]

     M.L. Ding, R.W. Flaig, H.L. Jiang, O.M. Yaghi, Chem. Soc. Rev. 48 (2019) 2783–2828. doi: 10.1039/c8cs00829a

  2. [2]

     D. Voiry, H.S. Shin, K.P. Loh, M. Chhowalla, Nat. Rev. Chem. 2 (2018) 0105. doi: 10.1038/s41570-017-0105

  3. [3]

     C.T. Dinh, T. Burdyny, M.G. Kibria, et al., Science 360 (2018) 783–787. doi: 10.1126/science.aas9100

  4. [4]

     Y. He, M. Gao, Y. Zhou, Y. Zhou, Chemosphere 311 (2023) 136925. doi: 10.1016/j.chemosphere.2022.136925

  5. [5]

     Y. Zhou, J. He, X. Li, J. Lu, Y. Zhou, Chem. Eng. J. 435 (2022) 132434. doi: 10.1016/j.cej.2021.132434

  6. [6]

     J. Lu, Q. Liu, Y. Zhang, Y. Zhou, Y. Zhou, Chin. Chem. Lett. 35 (2024) 109406.. doi: 10.1016/j.cclet.2023.109406

  7. [7]

     Y. Zhou, Y. He, M. Gao, et al., Chin. Chem. Lett. 35 (2024) 108690. doi: 10.1016/j.cclet.2023.108690

  8. [8]

     F. Qin, E. Almatrafi, C. Zhang, et al., Angew. Chem. Int. Ed. 62 (2023) e202300256. doi: 10.1002/anie.202300256

  9. [9]

     S. Lan, C. Yu, F. Sun, et al., Nano Energy 93 (2022) 106792. doi: 10.1016/j.nanoen.2021.106792

  10. [10]

      J. He, X. Wang, S. Lan, et al., Appl. Catal. B: Environ. 317 (2022) 121747. doi: 10.1016/j.apcatb.2022.121747

  11. [11]

      D. Qin, Y. Zhou, W. Wang, et al., J. Mater. Chem. A 8 (2020) 19156–19195. doi: 10.1039/d0ta07460h

  12. [12]

      J. Li, Z. Lou, B. Li, Chin. Chem. Lett. 33 (2022) 1154–1168. doi: 10.1016/j.cclet.2021.07.059

  13. [13]

      J. Huang, J. Wang, Z. Hao, et al., Chin. Chem. Lett. 32 (2021) 3180–3184. doi: 10.1016/j.cclet.2021.03.018

  14. [14]

      I. Hussain, A.A. Jalil, N.S. Hassan, M.Y.S. Hamid, J. Energy Chem. 62 (2021) 377–407. doi: 10.1016/j.jechem.2021.03.040

  15. [15]

      T.L. Wu, L. Liu, M.Y. Pi, D.K. Zhang, S.J. Chen, Appl. Surf. Sci. 377 (2016) 253–261. doi: 10.1016/j.apsusc.2016.03.140

  16. [16]

      R.T. Guo, J. Wang, Z.X. Bi, et al., Chemosphere 295 (2022) 133834. doi: 10.1016/j.chemosphere.2022.133834

  17. [17]

      K. Li, B.S. Peng, T.Y. Peng, ACS Catal. 6 (2016) 7485–7527. doi: 10.1021/acscatal.6b02089

  18. [18]

      R. Zhai, L.H. Zhang, M.Y. Gu, et al., Small 19 (2023) 2207840. doi: 10.1002/smll.202207840

  19. [19]

      V.P. Indrakanti, J.D. Kubicki, H.H. Schobert, Energy Environ. Sci. 2 (2009) 745–758. doi: 10.1039/b822176f

  20. [20]

      L. Liu, W.L. Fan, X. Zhao, et al., Langmuir 28 (2012) 10415–10424. doi: 10.1021/la301679h

  21. [21]

      Z.F. Shen, Q.N. Xia, Y.G. Li, et al., J. CO2 Util. 39 (2020) 101176.. doi: 10.1016/j.jcou.2020.101176

  22. [22]

      X.H. Feng, F.P. Pan, P. Zhang, et al., Chemphotochem 5 (2021) 79–89. doi: 10.1002/cptc.202000181

  23. [23]

      L.M. Wang, W.L. Chen, D.D. Zhang, et al., Chem. Soc. Rev. 48 (2019) 5310–5349. doi: 10.1039/c9cs00163h

  24. [24]

      H.A.E. Omr, M.W. Horn, H. Lee, Catalysts 11 (2021) 418. doi: 10.3390/catal11040418

  25. [25]

      B. Wang, X.H. Wang, L. Lu, et al., ACS Catal. 8 (2018) 516–525. doi: 10.1021/acscatal.7b02952

  26. [26]

      H.J. Yu, J.Y. Li, Y.H. Zhang, et al., Angew. Chem. Int. Ed. 58 (2019) 3880–3884. doi: 10.1002/anie.201813967

  27. [27]

      C.T. Campbell, C.H.F. Peden, Science 309 (2005) 713–714. doi: 10.1126/science.1113955

  28. [28]

      H.J. Freund, M.W. Roberts, Surf. Sci. Rep. 25 (1996) 225–273. doi: 10.1016/S0167-5729(96)00007-6

  29. [29]

      J. Raskó, Catal. Lett. 56 (1998) 11–15. doi: 10.1023/A:1019072021006

  30. [30]

      Y.F. Ji, Y. Luo, J. Am. Chem. Soc. 138 (2016) 15896–15902. doi: 10.1021/jacs.6b05695

  31. [31]

      Y.Q. He, H. Rao, K.P. Song, et al., Adv. Funct. Mater. 29 (2019) 1905153. doi: 10.1002/adfm.201905153

  32. [32]

      S.N. Habisreutinger, L. Schmidt-Mende, J.K. Stolarczyk, Angew. Chem. Int. Ed. 52 (2013) 7372–7408. doi: 10.1002/anie.201207199

  33. [33]

      K. Koci, L. Obalova, O. Solcova, Chem. Process Eng. 31 (2010) 395–407.

  34. [34]

      Y.O. Wang, E.Q. Chen, J.W. Tang, ACS Catal. 12 (2022) 7300–7316. doi: 10.1021/acscatal.2c01012

  35. [35]

      Y. He, Y. Zhou, J. Feng, M. Xing, Environ. Funct. Mater. 1 (2022) 204–217.

  36. [36]

      K. Koci, L. Obalova, D. Placha, Z. Lacny, Collect. Czech. Chem. Commun. 73 (2008) 1192–1204. doi: 10.1135/cccc20081192

  37. [37]

      R. Das, S. Sarkar, R. Kumar, et al., ACS Catal. 12 (2021) 687–697. doi: 10.1109/tencon54134.2021.9707338

  38. [38]

      R.R. Ikreedeegh, M. Tahir, J. CO2 Util. 43 (2021) 101381. doi: 10.1016/j.jcou.2020.101381

  39. [39]

      N. Shehzad, M. Tahir, K. Johari, T. Murugesan, M. Hussain, J. CO2 Util. 26 (2018) 98–122. doi: 10.1016/j.jcou.2018.04.026

  40. [40]

      G.C. Zuo, Y.T. Wang, W.L. Teo, Q.M. Xian, Y.L. Zhao, Appl. Catal. B: Environ. 291 (2021) 120126. doi: 10.1016/j.apcatb.2021.120126

  41. [41]

      P.G. Liu, Z.X. Huang, X.P. Gao, et al., Adv. Mater. 34 (2022) 2200057. doi: 10.1002/adma.202200057

  42. [42]

      Y. Ma, S. Wang, X. Duan, Chem. Eng. J. 455 (2023) 140654. doi: 10.1016/j.cej.2022.140654

  43. [43]

      A. Li, Q. Cao, G.Y. Zhou, et al., Angew. Chem. Int. Ed. 58 (2019) 14549–14555. doi: 10.1002/anie.201908058

  44. [44]

      J.F. Jing, J. Yang, W.L. Li, Z.H. Wu, Y.F. Zhu, Adv. Mater. 34 (2022) 2106807. doi: 10.1002/adma.202106807

  45. [45]

      Y.C. Hao, L.W. Chen, J. Li, et al., Nat. Commun. 12 (2021) 2682. doi: 10.1038/s41467-021-22991-7

  46. [46]

      Z.F. Huang, J. Song, X. Wang, et al., Nano Energy 40 (2017) 308–316. doi: 10.1016/j.nanoen.2017.08.032

  47. [47]

      X.X. Guo, X.Y. Qin, Z.J. Xue, et al., RSC Adv. 6 (2016) 48537–48542. doi: 10.1039/C6RA08551B

  48. [48]

      Y. Liu, Z. Zhang, Y. Fang, et al., Appl. Catal. B: Environ. 252 (2019) 164–173. doi: 10.1016/j.apcatb.2019.04.035

  49. [49]

      X. Xie, Q.U. Hassan, H. Lu, et al., Chin. Chem. Lett. 32 (2021) 2038–2042. doi: 10.1016/j.cclet.2020.10.002

  50. [50]

      S. Sun, M. Watanabe, J. Wu, Q. An, T. Ishihara, J. Am. Chem. Soc. 140 (2018) 6474–6482. doi: 10.1021/jacs.8b03316

  51. [51]

      Y. Deng, J. Li, R. Zhang, et al., Chin. J. Catal. 43 (2022) 1230–1237. doi: 10.1016/S1872-2067(21)63868-8

  52. [52]

      Y. Wang, R. Liu, M. Shi, et al., Chin. Chem. Lett. 34 (2023) 107200. doi: 10.1016/j.cclet.2022.02.006

  53. [53]

      N.N. Vu, S. Kaliaguine, T.O. Do, ChemSusChem 13 (2020) 3967–3991. doi: 10.1002/cssc.202000905

  54. [54]

      D.B. Hemandez-Uresti, D. Sanchez-Martinez, A. Martinez-de la Cruz, S. Sepulveda-Guzman, L.M. Torres-Martinez, Ceram. Int. 40 (2014) 4767–4775. doi: 10.1016/j.ceramint.2013.09.022

  55. [55]

      Y. Deng, Y. Liu, Y. Deng, et al., Chin. Chem. Lett. 35 (2024) 108898. doi: 10.1016/j.cclet.2023.108898

  56. [56]

      J. Li, J. Feng, X. Guo, et al., Appl. Catal. B: Environ. 309 (2022) 121248. doi: 10.1016/j.apcatb.2022.121248

  57. [57]

      T. Li, X. Dong, W. Chen, et al., Appl. Surf. Sci. 526 (2020) 146578. doi: 10.1016/j.apsusc.2020.146578

  58. [58]

      C. Lu, J. Li, J. Yan, et al., Appl. Mater. Today 20 (2020) 100744. doi: 10.1016/j.apmt.2020.100744

  59. [59]

      M. Shi, X. Tong, W. Li, et al., ACS Appl. Mater. Interface. 9 (2017) 34990–35000. doi: 10.1021/acsami.7b10891

  60. [60]

      Z. Xiong, Z. Lei, S. Ma, et al., Appl. Catal. B: Environ. 219 (2017) 412–424. doi: 10.1016/j.apcatb.2017.07.078

  61. [61]

      Z. Zhu, X. Tang, T. Wang, et al., Appl. Catal. B: Environ. 241 (2019) 319–328. doi: 10.1016/j.apcatb.2018.09.009

  62. [62]

      N. Ojha, A. Bajpai, S. Kumar, Catal. Sci. Technol. 9 (2019) 4598–4613. doi: 10.1039/c9cy01185d

  63. [63]

      H. Wang, L. Zhang, K. Wang, X. Sun, W. Wang, Appl. Catal. B: Environ. 243 (2019) 771–779. doi: 10.1016/j.apcatb.2018.11.021

  64. [64]

      M. Zhang, G. Cheng, Y. Wei, et al., J. Colloid Interface Sci. 572 (2020) 306–317. doi: 10.1016/j.jcis.2020.03.090

  65. [65]

      L. Yi, W. Zhao, Y. Huang, et al., Sci. China Mater. 63 (2020) 2206–2214. doi: 10.1007/s40843-019-1263-1

  66. [66]

      L. Wang, X.S. Zheng, L. Chen, Y.J. Xiong, H.X. Xu, Angew. Chem. Int. Ed. 57 (2018) 3454–3458. doi: 10.1002/anie.201710557

  67. [67]

      Y. Xiao, T. Wang, G.H. Qiu, et al., J. Colloid Interface Sci. 577 (2020) 459–470. doi: 10.1016/j.jcis.2020.05.099

  68. [68]

      S.B. Wang, B.Y. Guan, Y. Lu, X.W. Lou, J. Am. Chem. Soc. 139 (2017) 17305–17308. doi: 10.1021/jacs.7b10733

  69. [69]

      B. Tahir, M. Tahir, M.G.M. Nawawi, J. CO2 Util. 41 (2020) 101270. doi: 10.1016/j.jcou.2020.101270

  70. [70]

      X. Jiang, Y. Ding, S. Zheng, et al., ChemSusChem 15 (2022) e202102295. doi: 10.1002/cssc.202102295

  71. [71]

      Z. Tang, C. Wang, W. He, et al., Chin. Chem. Lett. 33 (2022) 939–942. doi: 10.1016/j.cclet.2021.07.020

  72. [72]

      F. Xu, K. Meng, B. Cheng, et al., Nat. Commun. 11 (2020) 4613. doi: 10.1038/s41467-020-18350-7

  73. [73]

      A. Meng, B. Cheng, H. Tan, et al., Appl. Catal. B: Environ. 289 (2021) 120039. doi: 10.1016/j.apcatb.2021.120039

  74. [74]

      S. Karmakar, S. Barman, F.A. Rahimi, T.K. Maji, Energy Environ. Sci. 14 (2021) 2429–2440. doi: 10.1039/d0ee03643a

  75. [75]

      X. Li, B. Kang, F. Dong, et al., Nano Energy 81 (2021) 105671. doi: 10.1016/j.nanoen.2020.105671

  76. [76]

      M. Lin, M. Luo, Y. Liu, et al., Chin. J. Catal. 50 (2023) 239–248. doi: 10.56028/aehssr.4.1.239.2023

  77. [77]

      J. Wang, H. Cheng, D. Wei, Z. Li, Chin. J. Catal. 43 (2022) 2606–2614. doi: 10.1016/S1872-2067(22)64091-9

  78. [78]

      Z. Xing, J. Hu, M. Ma, et al., J. Am. Chem. Soc. 141 (2019) 19715–19727. doi: 10.1021/jacs.9b08651

  79. [79]

      Y. Li, Y. Liu, D. Xing, et al., Appl. Catal. B: Environ. 285 (2021) 119855. doi: 10.1016/j.apcatb.2020.119855

  80. [80]

      H. Jiang, W. Wang, L. Sun, et al., J. Catal. 416 (2022) 1–10. doi: 10.1016/j.jcat.2022.10.015

  81. [81]

      L. Liang, X. Li, Y. Sun, et al., Joule 2 (2018) 1004–1016. doi: 10.1016/j.joule.2018.02.019

  82. [82]

      Y. Huang, K. Dai, J. Zhang, G. Dawson, Chin. J. Catal. 43 (2022) 2539–2547. doi: 10.1016/S1872-2067(21)64024-X

  83. [83]

      C. Zhu, X. Chen, J. Ma, et al., J. Phys. Chem. C 122 (2018) 20444–20458. doi: 10.1021/acs.jpcc.8b06624

  84. [84]

      X. Jiang, J. Huang, Z. Bi, et al., Adv. Mater. 34 (2022) 2109330. doi: 10.1002/adma.202109330

  85. [85]

      Y.J. Dong, Y. Jiang, J.F. Liao, et al., Sci. China Mater. 65 (2022) 1550–1559. doi: 10.1007/s40843-021-1962-9

  86. [86]

      Y.F. Li, N. Soheilnia, M. Greiner, et al., ACS Appl. Mater. Interfaces 11 (2019) 5610–5615. doi: 10.1021/acsami.8b04982

  87. [87]

      H. Zhang, Y. Wang, S. Zuo, et al., J. Am. Chem. Soc. 143 (2021) 2173–2177. doi: 10.1021/jacs.0c08409

  88. [88]

      Z. Zhu, W.R. Huang, C.Y. Chen, R.J. Wu, J. CO2 Util. 28 (2018) 247–254. doi: 10.1016/j.jcou.2018.10.006

  89. [89]

      A. Raza, H. Shen, A.A. Haidry, et al., J. CO2 Util. 37 (2020) 260–271. doi: 10.1016/j.jcou.2019.12.020

  90. [90]

      Q. Chen, X. Lan, K. Chen, Q. Ren, J. Shi, J. Colloid Interface Sci. 616 (2022) 253–260. doi: 10.1016/j.jcis.2022.02.044

  91. [91]

      P.Y. Jia, R.T. Guo, W.G. Pan, et al., Colloid Surface A 570 (2019) 306–316. doi: 10.1016/j.colsurfa.2019.03.045

  92. [92]

      Y. Feng, C. Wang, P. Cui, et al., Adv. Mater. 34 (2022) e2109074. doi: 10.1002/adma.202109074

  93. [93]

      L. Shen, Z. Xie, L. Hou, J. Yang, Q. Li, Energy Fuel. 36 (2022) 11515–11523. doi: 10.1021/acs.energyfuels.2c01366

  94. [94]

      B. Weng, M.Y. Qi, C. Han, Z.R. Tang, Y.J. Xu, ACS Catal. 9 (2019) 4642–4687. doi: 10.1021/acscatal.9b00313

  95. [95]

      C. Kuai, Z. Xu, C. Xi, et al., Nat. Catal. 3 (2020) 743–753. doi: 10.1038/s41929-020-0496-z

  96. [96]

      Z. Wan, G. Zhang, X. Wu, S. Yin, Appl. Catal. B: Environ. 207 (2017) 17–26. doi: 10.1016/j.apcatb.2017.02.014

  97. [97]

      H. Dong, X. Guo, C. Yang, Z. Ouyang, Appl. Catal. B: Environ. 230 (2018) 65–76. doi: 10.1016/j.apcatb.2018.02.044

  98. [98]

      G.F. Samu, C. Janáky, J. Am. Chem. Soc. 142 (2020) 21595–21614. doi: 10.1021/jacs.0c10348

  99. [99]

      D. Zeng, H. Wang, X. Zhu, et al., Chem. Eng. J. 451 (2023) 138801. doi: 10.1016/j.cej.2022.138801

  100. [100]

       W. Shi, X. Guo, C. Cui, et al., Appl. Catal. B: Environ. 243 (2019) 236–242. doi: 10.1016/j.apcatb.2018.09.076

  101. [101]

       Y. Liu, X. Dong, Q. Yuan, et al., Colloid Surface A 621 (2021) 126582. doi: 10.1016/j.colsurfa.2021.126582

• 

  Figure 1  (a) HETEM of WO_3-x-R/GdCrO₃, and (b) W 4f XPS spectra of the samples. Reprinted with permission [56]. Copyright 2022, Elsevier. (c) CO₂ photoelectrocatalytic performances of WO₃-raw and WO₃-T, and (d) XPS W 4f profiles of WO₃-raw and WO₃-T. Reprinted with permission [57]. Copyright 2020, Elsevier. The aldehyde selectivity of the catalysts is represented by dots, SEM (e) and HAADF-STEM (f) images of WO_3-x-2 synthesized under 2 mL HCl assistance, Raman spectra (g) of WO_3-x-2 during CO₂-PR, and their IR spectra (h) after photocatalysis. Reprinted with permission [58]. Copyright 2020, Elsevier.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) EPR spectra at 300 K for WO and 3% Mo-WO, (b) The yield of CO₂-PR to CH₄ in the first 6 h for WO, 1% Mo-WO, 3%Mo-WO and 5%Mo-WO, and (c) CO₂-PR over Mo-WO with water under solar irradiation. Reprinted with permission [63]. Copyright 2019, Elsevier. (d) W 4f high-resolution XPS spectra and (e) room-temperature EPR spectra of W₁₈O₄₉ and Cu-W₁₈O₄₉–0.005. Reprinted with permission [64]. Copyright 2020, Elsevier.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) HRTEM of the 30V_O-WO_x/BiOCl sample, high-resolution XPS spectra of Bi 4f (b), W 4f (c) of the BiOCl, V_O-WO_x and 30V_O-WO_x/BiOCl samples in the dark and under the light, (d) the CO evaluation of BiOCl, V_O-WO_x. 30WO₃/BiOCl and 30V_O-WO_x /BiOCl, (e) CO production rates of BiOCl, 30WO₃/BiOCl and 30V_O-WO_x/BiOCl, and (f) S-scheme heterojunction composed by BiOCl and V_O-WO_x. Reprinted with permission [80]. Copyright 2022, Elsevier. (g) HRTEM image of the Au/TiO₂/W₁₈O₄₉ plasmonic heterostructure, (h) TEM image of the side view of the Au/TiO₂/W₁₈O₄₉ plasmonic heterostructure, (i) schematic diagram showing the hot electron generation and transfer processes due to the enhanced plasmon coupling between Au and W₁₈O₄₉ under UV–vis–NIR light irradiation, and (j) gas productions and the corresponding CH₄ selectivity through CO₂-PR over the as-synthesized samples under UV–vis–NIR light irradiation: (1) TiO₂ NFs, (2) Au/TiO₂ composite NFs, (3) Au/TiO₂/W₁₈O₄₉ plasmonic heterostructure, (4) W₁₈O₄₉ NWs, (5) W₁₈O₄₉/TiO₂ plasmonic heterostructure. Reprinted with permission [84]. Copyright 2022, Wiley-VCH.

  下载: 全尺寸图片 幻灯片
•