English

• 1. Introduction

  In recent years, achieving carbon neutrality has attracted extensive attention in the media, politics, and science [1-3]. As a major greenhouse gas, the effective control of CO₂ emissions is critical to realizing carbon neutrality. If effective strategies to reduce CO₂ emissions are not taken, the steadily increasing concentration of carbon dioxide in the atmosphere will continue to increase the global average temperature, causing a series of natural disasters and extreme weather [3]. However, the increasing hunger for energy by an ever-growing human population is unceasing and there is no doubt that CO₂ emissions will not stagnate [4].

  In nature, CO₂ can be converted into organic matter by plant photosynthesis, providing an effective way to solve the contradiction between the increase of energy consumption and the reduction of CO₂ emissions [5-10]. However, from a thermodynamic point of view, the reduction of CO₂ to CO, methane (CH₄), methanol (CH₃OH), formic acid (CHOOH), or other C2 productions are very unfavorable due to the high bond energy of C=O (750 kJ/mol) [11]. These processes, like the widely studied photocatalytic hydrogen evolution reaction (HER) and nitrogen reduction reaction (NRR) [12-14], require a certain amount of energy. Therefore, it is necessary to develop and use abundant "green" energy to catalyze the conversion of carbon dioxide. Being widely studied, photocatalysis using clean solar energy is one of the most effective methods to catalyze the recovery of CO₂ resources [15, 16]. However, for photoreduction of CO₂ to value-add products, high criteria for photocatalyst design is required: (1) efficient conversion process, (2) good product selectivity, and (3) high stability [17]. As such, the design of photocatalyst is the key factor to achieving CO₂ reutilization via photocatalysis [18].

  With high design criteria, single-atom catalysts (SACs) are the appropriate candidates for CO₂ reduction [19-21]. SACs have many advantages [22]: (1) unexpectedly high activity and selectivity due to its unsaturated coordination and tunable electronic structures; (2) the maximum atom utilization efficiency, resulting in the reduced cost, especially for precious metals; (3) well-defined active sites for mechanistic studies because they can bridge the gap between heterogeneous and homogeneous catalysis. Due to the unique advantages, numerous SACs were synthesized and studied in photocatalysis CO₂ reduction in recent years [18, 23].

  The general catalytic performance of SACs is determined by the selection of carrier materials because when the particle size is reduced to a single atom, the isolated atoms should be anchored on carrier materials that prevent aggregation due to an increase in surface free energy [18, 24, 25]. Therefore, to this end, supports in various carrier materials have been developed for anchoring single atoms, such as metal-containing materials (e.g., TiO₂ [26], FeO_x [27], Mo₂S [28]) and metal-free materials (e.g., graphene oxide [29], carbon nitride [30]).

  Among the numerous carrier materials, two-dimensional (2D) materials are attractive candidates, which is the thickness of only one or a few atomic layers [11, 15, 31]. The unique physical and chemical structure of 2D material allows a higher density of single atoms. In addition, compared to block materials, 2D materials can create more unsaturated coordination centers, have unique electronic structures, and contain higher catalytic activity generally [32-34].

  Combining the above-mentioned advantages of 2D materials and single atom sites, 2D materials confining single atoms (SACs@2D) have the following features [33, 35, 36]: (1) a large specific surface area to increase the metal single atom load and achieve 100% active sites exposure; (2) the energy band structure of 2D materials can be effectively adjusted by single atoms, and thus effectively improve the visible light response and photocatalytic CO₂ reduction performance; (3) the unsaturated coordination structure of SACs@2D benefits the adsorption and activation of CO₂, facilitating higher photocatalytic CO₂ reduction; (4) SACs@2D can be easily analyzed by advanced characterization techniques, providing convenience for theoretical research. Studies have shown that the single atomic sites on 2D materials are excellent CO₂ reduction sites [37-41]. Therefore, SACs@2D for CO₂ photoreduction is worthy of further study.

  Although many reviews have summarized the synthesis, characterization, and application of 2D materials confining single atoms [31, 33, [42-44], a comprehensive overview of SACs@2D for the CO₂ photoreduction is missing, especially how single atoms are supported on 2D materials and the functions of single atoms sites for CO₂ reduction process. Therefore, in this review, the design of SACs@2D is first summarized, providing practical guidance for the design and synthesis of SACs@2D. Secondly, SACs@2D for photocatalytic CO₂ reduction is reviewed to provide a comprehensive understanding of the functions of single atomic sites. Finally, the major challenges and opportunities for further research and application of SACs@2D for CO₂ photoreduction are proposed. Thus, the research status and trends of SACs@2D in CO₂ photoreduction are fully understood, laying a foundation for the further application of SACs@2D for CO₂ photoreduction.

  2. Single atoms supported on 2D nanomaterials

  The key point of SACs synthesis is to anchor isolated atoms on the 2D carrier materials surface properly, because the prerequisite for the practical application of SACs is that the single atoms do not aggregate during the catalytic process, that is, the single atoms locate at the desired location without any significant migration [45]. However, preventing the aggregation of single atoms during preparation and catalysis is a great challenge [46]. Fortunately, there are many great significant progresses for 2D materials that support single atoms, providing theoretical guidance for the rational development of SACs@2D future. Taking typical 2D material g-C₃N₄ as an example [5, 47-50], there are three main kinds of positions for anchoring single atoms: The pore structure, interlayer structure between two layers, and defect sites, as shown in Fig. 1. This section will review the suitable sites of 2D materials for confining single atoms.

  Figure 1

  

  Figure 1.  Potential sites of single atom supported on g-C₃N₄: (a) pore structure; (b) interlayer space; (c) defect sites. Color codes: carbon is brown, oxygen is red, and a confined single atom is green.

  DownLoad: Full-Size Img PowerPoint

  2.1 Single atoms confined into pore structure

  The intrinsic pore structure of some materials is a candidate to confine single atoms. For example, g-C₃N₄ is composed of tri-s-triazine rings, based on sp² hybridization of π conjugated graphite planes by alternating C and N connections [51-53]. The internal pore structure filled with six N atoms with lone-pair electrons is an ideal site for anchoring atoms to prevent them from aggregating.

  In addition to the coordination on the original atoms of the pore structure, the heteroatoms on pore structure of 2D materials are also useful sites to support the single atom. For example, being widely studied, N heteroatom coordination metal single atoms (M-N_x) catalysts have shown their excellent catalytic activity for several catalytic reactions [54, 55]. In addition to N atoms, O or P atoms are also alternative coordination heteroatoms. For example, Gao et al. reported single Co sites coordinated on the partially oxidized graphene nanosheets via a Co-O₃C configuration [22]. Moreover, Chu et al. also synthesized the sing Co atoms by coordinating Co above the void center of g-C₃N₄ by heteroatom P [56]. The strong ligand-metal coordination between single metal atoms and heteroatoms of 2D materials is beneficial for facile manipulation of loading sites.

  One advantage of confining of single atom into the internal pore structure of 2D materials is to improve the light absorption capacity of 2D materials. For example, Xiong et al. synthesized a series of g-C₃N₄ inclusion Pt single atoms photocatalysts, in which Pt single atoms are confined to the "N pots" [57]. By incorporating a trace amount of Pt single atoms in the "N pots" of g-C₃N₄, the light absorption of g-C₃N₄ can be broadened to 900 nm by the metal-to-ligand charge transfer (MLCT). The MLCT mechanism also applies to non-noble metal Cu single atom confined in the internal pore structure. In addition to Pt and Cu atoms being confined to the internal pore structure, Fe and Zn atoms can also be incorporated into the g-C₃N₄ matrix, and it is found that the band gap of g-C₃N₄ becomes narrower with the introduction of Fe atoms [58].

  In addition to light absorption properties, charge transfer properties can also be adjusted by introducing single atoms into the pore structure [59]. For example, Li et al. embedded isolated single Pt atoms on the top of the five-membered rings of g-C₃N₄ network [60]. The dispersion of single Pt atoms on g-C₃N₄ modulated the electronic structure and introduced the intrinsic change of the surface trap states, resulting in a nearly twofold decrease in average recovery lifetime of photon-generated carriers, suggesting that Pt single atoms provide more opportunities for photogenerated electrons to participate in photocatalytic reactions. The charge transfer channel can be constructed by placing the single atoms into the internal pore structure of g-C₃N₄, and the utilization rate of light can be improved.

  Another advantage of confining single atom in the internal pore structure is to optimize the electronic structure and modify energy band levels. For example, M−N/C configurations can optimize the electronic structure and modify the band energy levels, because the electrons tend to transfer from the metal-sites to the N atoms via the strong interaction, resulting in the downshift of the valence band maximum (VBM) level. Inspired by the above phenomena, Su et al. modified the valence band structure of g-C₃N₄ by introducing Pt single atoms via a "high-valence metal single-atom confinement" strategy [61]. The hybridization of Pt 5d and N 2p downward shifts the VBM level of modified g-C₃N₄ by 0.26 V relative to bare g-C₃N₄, triggering a highly efficient photocatalytic water oxidation process.

  Taken together, reasonably confined single atoms in the internal pore structure of 2D materials has the following functions: optimizing light absorption performance, improving charge transfer performance, adjusting band position, and providing the active sites for CO₂ photoreduction.

  2.2 Single atoms confined in layer structure

  The interlayer structure is also a candidate position for confining single atoms, because 2D materials are usually stacked together by weak van der Waals forces, which may prevent metal atoms from clustering [62]. Inspired by this, Cao et al. atomically dispersed Pd single atoms in both bridged sites of adjacent layers and surface-sites of g-C₃N₄ to accelerate charge separation [63]. Theoretical prediction and experiment characterizations reveal that single Pd atoms in bridged sites of adjacent layers provide directional charge-transfer channels. Such a unique structure efficiently boosts electron transport and separation. In addition to single Pd atoms, Cu single atoms can also be confined in the interlayer structure of g-C₃N₄ [64]. For example, Xiao et al. anchored Cu single atoms between g-C₃N₄ planes by intercalation of chlorophyll sodium copper (C₃₄H₃₁CuN₄Na₃O₆, CSC) into a supramolecular precursor. The single Cu atoms in two neighboring layers established Cu-N₄ charge transfer channels, which significantly improve the interlayer transfer of photogenerated charge carriers.

  In addition to being confined in interlayers of the same material, single atoms can also be confined in the interfaces between 2D material and other semiconductor materials. For example, Zhao et al. prepared a series of composite photocatalysts by confining rare earth elements (Lu, Yb, Tm, Er, Ho, Dy, Gd, Eu, Nd, Pr, and La) single atoms onto the interface between g-C₃N₄ and CdS [65]. In these photocatalysts, CdS: Dy³⁺/g-C₃N₄ is the best photocatalyst for CO₂ reduction. Further DFT calculation results show that single Dy atoms are suitable for electron transfer in the CdS/g-C₃N₄ due to the intermediate excited state energy level of Dy³⁺ located between the conduction bands of g-C₃N₄ and CdS. In addition, this strategy is also suitable for the synthesis of SrTiO₃: Er³⁺/g-C₃N₄ composite catalysts [66]. The surface single Er atoms on SrTiO₃ facilitate the construction of charge transport channels between SrTiO₃ and g-C₃N₄, resulting in enhanced interfacial charge transfer and separation.

  In general, the introduction of single atoms can enhance charge transfer in 2D material layers or interface between 2D material and other materials.

  2.3 Single atoms confined into vacancy defects

  Confined single atoms in defect sites, widely used in the synthesis of SACs, is also suitable for the synthesis of SACs@2D. For Example, Cao et al. designed and fabricated a series of single Au atoms supported on CdS with different types of vacancies (Fig. 2) [67]. For the sample of Au single atom supported on CdS with Cd vacancies (Cd_1−xS), the Cd vacancy electron paramagnetic resonance (EPR) signals decreased obviously after supporting single Au atoms, indicating that single Au atoms coverer part of Cd vacancies. Similarly, the authors also synthesized Au single-atom supported on CdS with S vacancies (CdS_1-x) samples. Interestingly, the types of vacancies for anchoring Au single atom highly affect the electron transfer direction.

  Figure 2

  

  Figure 2.  Electron transfer between Au and CdS with different types of vacancies: (a) Cd vacancies; (b) S vacancies. Reprinted with permission [67]. Copyright 2021, Springer Nature.

  DownLoad: Full-Size Img PowerPoint

  A single atom will not only be confined in defects sites, it may also induce the formation of defective structures around it. For example, Zhang et al. found that single Ru atoms supported on MoS₂ induce phase transition of MoS₂ and generation of S vacancies, which efficiently tailors the electronic structure of MoS₂ [68].

  In conclusion, single atoms supported on vacancy defects of 2D materials mainly play roles in regulating the band structure and providing the active site of 2D materials.

  3. Photoreduction of CO₂ by SACs@2D

  Upon identifying the design of SACs@2D, we are now in a position to decode the function of single atoms for CO₂ photoreduction. Generally, it is widely accepted that the photocatalytic CO₂ reactions are triggered by the photogenerated electrons on the surface of a semiconductor. For a SAC@2D catalysts, electrons in valence band of carrier material are excited into its conduction band when the photon energy is equivalent to or higher than its band gap energy. The excited electrons are transferred to the single atom sites to drive the CO₂ photoreduction reaction [69]. However, photocatalytic CO₂ reduction is a multiple proton-coupled electron transfer process with many intermediates, so the different numbers of transferred electrons and protons will form different reduction products. In addition, reduction potential determines whether the reaction can take place, which has strict requirements on the conduction band position of SACs@2D. Furthermore, for the multi-carbon products, the difficulty of C—C coupling is another limiting factor. Up to now, the main products of photocatalytic CO₂ reduction are C1 products, including CO, CH₄, CH₃OH, and HCOOH, and only a few reported products were C2 or C2+ products. [11]. Therefore, the function of SACs@2D for photocatalytic CO₂ reduction will be reviewed in this section, and the typical applications of SACs@2D for photocatalytic CO₂ reduction are listed in Table 1 [22, 35, 67, [70-81].

  Table 1

  

  Table 1.  Summary of various SAC@2D toward CO₂ reduction.

  DownLoad: CSV

  3.1 Noble metals-based SACs@2D

  Due to the advantages of excellent catalytic activity and stability, the function of single sites in noble metals-based SACs@2D is worth studying [27, 82-85]. Therefore, this part mainly introduces the functions of single noble metal sites in CO₂ reduction.

  The noble metals-based single sites determine the selectivity of the CO₂ photoreduction products. For example, Gao et al. investigated Pd/g-C₃N₄ and Pt/g-C₃N₄ for photoreduction of CO₂ by DFT calculations [86]. They found that the desorption of HCOOH from the Pd/g-C₃N₄ catalyst possesses the largest barrier on the entire reaction pathway (0.46 eV), while the desorption energy from the Pt/g-C₃N₄ catalyst is 1.06 eV, suggesting that the formation of HCOOH is unfavorable on the Pt/g-C₃N₄ catalyst. Similarly, Zhou et al. studied the noble metal (Rh, Pd, Ag, Ir, Pt, and Au) single atom supported on defective g-C₃N₄ for CO₂ photoreduction by DFT calculations [87]. They found that the noble metal single atom tends to be trapped at the N-vacancy sites of g-C₃N₄ rather than the surface internal pore structure. The N-vacancy-stabilized Pd single atom on g-C₃N₄ show the lowest energy barrier for HCOOH production.

  In addition to affecting the selectivity of the product, noble metals-based single sites on SAC@2D can also improve the activity of the product. For example, as shown in Fig. 2, Cao et al. proposed a different mechanism for surface electron density modulation using single Au atoms supported CdS [67]. When electrons concentrate on vacancies rather than single Au atoms, the adsorption type of CO₂ changed from physical adsorption to chemical adsorption. The reason for this phenomenon is that the strong hybridization of Au 5d and S 2p orbits accelerates the photo-electrons transfer onto the surface. Therefore, lots of photoelectrons are transferred to the surface to participate in CO₂ reduction reaction and finally achieve efficient carbon dioxide reduction. Moreover, Xiang et al. synthesized Au single atoms on amino group-enriched graphitic carbon nitride through a mild urea reduction method [70]. For this catalyst, the CO and CH₄ yields over this Au single atom photocatalyst are 21.7 and 2.4 µmol/g after 2.5 h of visible-light illumination, which are about 1.97 and 4.15 times those of pure g-C₃N₄. This improved photocatalytic performance is ascribed to the improved intermediate-adsorption ability, narrowed band gap, and decreased carrier-recombination rate after the introduction of Au single atom. Note that the introduced amino group improves its CO₂ affinity and helps improve CO₂ reduction. In addition, Sharma et al. achieved efficient photocatalytic CO₂ reduction to CH₃OH by single Ru atoms supported on porous g-C₃N₄ catalysts with a high yield of 1500 µmol/g (6 h under visible light) [71]. Ru-N/C coordination structure can accelerate charge transfer and increase charge density of active sites, thus realizing the enhanced photocatalytic activity of Ru single atom modified g-C₃N₄ for the aqueous photoreduction of CO₂ to CH₃OH using water as the sacrificial agent.

  To sum up, although the researches of noble metals-based SACs@2D for CO₂ photoreduction have made some progress in improving product yield and selectivity, many problems still need to be addressed. For Example, the application of noble metals-based SACs@2D catalysts in photocatalytic reduction of CO₂ is rarely reported with experimental validation but has been widely studied with other reactions, such as hydrogen production (Pt [57, 60, 88, 89], Pd [90, 91], Ag [59], and Rh [92]), NO oxidation (Pt [93, 94] and Pd [95]), and acetylene hydrogenation (Pd [96] and Au [97]). The main reason why noble metals-based SACs@2D catalysts are rarely used in the photoreduction of CO₂ reaction in experiments is the competitive reaction of H₂ generation [98]. The H₂ production not only affects the utilization efficiency of photogenerated electrons but also limits the yield and selectivity of products such as CH₄. Therefore, in order to realize the application of noble metals-based SACs@2D catalysts in CO₂ reduction reactions, the active sites must be reasonably modified.

  3.2 Transition metal-based SACs@2D

  When catalyst prices are considered, the substitution of cheap transition metals for noble metals is worth studying. Therefore, the function of transition metal single atom in SACs@2D for photoreduction CO₂ is summarized in this part.

  The transition metal single atom with variable redox states is an excellent CO₂ photoreduction site. For example, Wang et al. synthesized Cu modified g-C₃N₄ by C-Cu-N₂ coordination structure via supramolecular preorganization with subsequent condensation [79]. In this photocatalyst, the variable-valence C-Cu-N₂ sites can provide effective charge transfer pathways reduce the energy barrier toward photocatalytic CO₂ reduction, as shown in Eqs. 1–7:

  (1)

  (2)

  (3)

  (4)

  (5)

  (6)

  (7)

  Note that the mechanism mentioned above has a certain universality [99, 100]. For example, Tang et al. demonstrated that single Ti-O species supported on g-C₃N₄ can convert CO₂ into CO by photoreduction [72]. The Ti⁴⁺ can be transformed to Ti³⁺ by the light-generated electrons under the visible light irradiation. Then, the reduced Ti³⁺ changes back to Ti⁴⁺ by one electron transfer to co-catalysts (Co(bpy)₃²⁺) to convert CO₂ into CO.

  Among the various function of transition metal single sites, CO₂ adsorption capacity is one of the most important functions. For example, Cheng et al. reported an unsaturated edge confinement strategy, that is, single Ni sites coordinated on the few layer g-C₃N₄ via a self-limiting method [78]. The obtained sample achieves a comparably high performance of CO₂ photoreduction to CO under visible light illumination (8.6 µmol g⁻¹ h⁻¹), which is 7.8 times that of the unmodified few-layer g-C₃N₄. In this process, further mechanism studies showed that the introduction of unsaturated Ni-N coordination is beneficial to improving CO₂ adsorption capacity and facilitating photogenerated electrons transfer, thereby achieving the comparably high photocatalytic activity for converting CO₂ to CO. Moreover, Gong et al. synthesized g-C₃N₄ supported Co (Fe and Mn) single atom by one-step thermal treatment of Co-MOF (Fe-MOF and Mn-MOF) and urea in the air [101]. The obtained sample exhibits excellent performance for photocatalytic reduction of CO₂ to CO with a high CO evolution rate of 394.4 µmol g⁻¹ h⁻¹ under visible light irradiation with hole sacrificial agent (TEOA), which is over 80 times higher than that of unmodified g-C₃N₄ (4.9 µmol g⁻¹ h⁻¹). Experiment results show that the main reasons for the great improved CO evolution rate are the improved CO₂ absorption ability and charge separation efficiency after the introduction of single Co atoms.

  In addition to charge transfer and CO₂ adsorption, the transition metal single sites also greatly affect the activation of CO₂. For example, Li et al. prepared Cu single atoms supported on crystalline g-C₃N₄ nanorods catalyst by molten salts and reflux method [102]. For this catalyst, photoreduction of CO₂-to-CO is a favorable reaction than CO₂-to-CH₄ over single Cu sits, thus achieving the enhanced photocatalytic CO₂ reduction with near 100% selective CO₂ photoreduction to CO. In addition, Kou et al. introduced single MoN₂ sites into COF to construct Mo-COF materials and found that this active site can reduce CO₂ to C₂ product (C₂H₄) [80]. Further experiments and calculations show that the formation of C₂ is due to the reduction of the energy barrier for the formation of C₂H₄ by the introduction of Mo single atom.

  In general, transition metal sites have multiple functions simultaneously. For example, Zhao et al. reported a single Fe atom supported on g-C₃N₄ catalyst for the CO₂ photoreduction with water as the sacrificial agent [73]. For this catalyst, experiment and DFT calculations results reveals that the function of single Fe atoms as follow: (1) activating CO₂ in the ground state; (2) increasing the rate constant of the limiting step; (3) providing the catalytic active center; (4) lowering the reaction barrier of the C—O bond cleavage in COOH. In other words, the above-mentioned advantages of single Fe atoms together improve the performance of photocatalytic reduction of CO₂.

  In all, great progress has been made in developing the function of transition metal-based SACs@2D, but the role of some elements remains to be further studied. For example, some common single transition metal sites, such as Mn [103] and Zn [104], have been proved to be good choices for CO₂ reduction, but their support in 2D materials needs to be further studied.

  3.3 Rare earth metal-based SACs@2D

  Rare earth metal-based SACs@2D for CO₂ photoreduction is also worthy of study, and the reasons are listed below [105]. Firstly, the rare-earth elements are considered critical materials because of their optical, magnetic properties, and unique electronics, which are widely used in catalysis. Secondly, when rare earth metals are reduced to a single atomic scale, maximum utilization of metal elements can be achieved. In addition, atomic-scale rare-earth metals may have properties not found in their nanoscale homologues. Therefore, in this part, the function of rare earth metal single atom in SACs@2D for photoreduction CO₂ is reviewed.

  The main function of single rare earth metal sites is to construct charge transport channels. For example, Chen et al. designed and synthesized single La atoms modified g-C₃N₄ with La-N charge-transfer bridge [35]. The La-modified g-C₃N₄ realized a high rate (92 µmol g⁻¹ h⁻¹) and CO selectivity (80.3%) in photocatalytic CO₂ reduction. The experimental and DFT calculations results show that the variety of electronic states induced by 4f and 5d orbitals of the rare-earth La single atom and the p-d orbital hybridization of La-N atoms forms a charge transport channel, which is the key active sites for CO₂ activation.

  Another function of the single rare earth element atom in SACs@2D is to build active sites for CO₂ photoreduction. For example, Ji et al. synthesized rare-earth element single atoms (Er) supported on g-C₃N₄ by a novel atom-confinement and coordination strategy [81]. With the increase of the amount of Er precursor, the dispersion density of single atoms can be tuned from the formation of a low-density Er single-atom catalyst with an Er mass loading of 2.5 wt% to a high-density Er single-atom catalyst with an Er mass loading of 20.1 wt%. In these two catalysts, the high-density Er single-atom catalyst shows higher catalytic performance, indicating that the active sites on the catalyst surface were effectively increased.

  In another case, single rare earth metal atoms in SACs@2D have dual functions at the same time. For example, the dual functions of CO₂ molecular activation and charge transfer bridge for rare-earth element single atom systematically studied by Wang's groups [65, 66, 106]. Taking the Dy³⁺ single atom doped in the interface of CdS/g-C₃N₄ heterojunction as an example, this structure combines the characteristics of rare-earth element single atom and the advantages of heterojunction [65]. In this photocatalyst, Dy³⁺ single atom is supported on the surface of CdS, so it also exists at the interface between CdS and g-C₃N₄. On the one hand, the intermediate excited state energy level of Dy³⁺ in the interface of CdS/g-C₃N₄ heterojunction can act as an electron transport bridge. On the other hand, the surface Dy³⁺ single atom changes the density of active sites, which can promote the absorption of CO₂. Under the dual functions of Dy single atom, the optimal catalyst realized an efficient photocatalytic reduction of CO₂ to CH₄. Guided by the same idea, this group synthesized dual functions Er single atom in Zn₂GeO₄/g-C₃N₄ heterojunction [106] and SrTiO₃/g-C₃N₄ heterojunction [66], and these rare-earth element single atom modified catalyst all achieved higher catalytic performance than that of pure g-C₃N₄. In such rare-earth element single atom supported on g-C₃N₄-based heterojunction (Fig. 3), the position of the 4f levels of rare earth elements as electron transport (ET) bridge is very closely related to the conduction band position of the heterojunction materials.

  Figure 3

  

  Figure 3.  The proposed mechanism for single rare-earth element atoms as ET bridges to promote charge separation.

  DownLoad: Full-Size Img PowerPoint

  Over all, both theoretical and experimental studies on the loading of rare-earth elements on 2D materials for the photoreduction of CO₂ are still in the initial stage, which needs to be further studied to show the advantages of rare earth elements.

  4. Conclusions and outlook

  In summary, this review summarizes the design and functions of single atomic sites for CO₂ photoreduction of SACs@2D in recent years. The suitable sites of 2D materials for confining single atoms are internal pore structure, interlayer structure or interface, and vacancy defects or heteroatoms. Moreover, the functions of well-designed single atomic sites in CO₂ photoreduction includes enhanced light absorption, accelerated charge transfer, and improved adsorption and activation of CO₂.

  Although the studies on SACs@2D for CO₂ photoreduction have achieved significant progress, a few research directions still deserve special attention:

  (1) At present, the activity of photoreaction CO₂ over SACs@2D is still at a relatively low level compared to electrocatalysis and thermal catalysis, especially with water as the sacrificial agent. Therefore, the conversion rate of the product needs to be further improved. Similarly, how to improve product selectivity remains an important topic.

  (2) The stability of the single atoms sites needs to be further studied, because single atoms tend to agglomerate into clusters or particles during synthesis and application process. In addition, the presence of other pollutants may also occupy the active site, leading to the loss of catalytic performance of SACs.

  (3) The single atom active site on catalyst surface is a dynamic change process during CO₂ reduction, so more in-situ characterization should be conducted to study the changes of the single atom's physicochemical properties during CO₂ reduction. At present, most conclusions are based on DFT calculation results, so spectroscopic characterization or electron microscopy characterization is needed to give more intuitive evidence.

  (4) At the present stage, most of the products of CO₂ photoreduction are C1 products. Some effective approaches can be taken into consideration to solve the above problems, such as introducing dual single-atom active sites and other reactive groups around the single atom sites, and enhancing the interaction between single atoms and their carrier interface.

  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.

  Acknowledgments

  This work was supported by the Strategic Priority Research Program of the Chinese Academy of Sciences, China (Nos. XDA23010300 and XDA23010000), National Science Foundation of China, China (Nos. 51878644 and 41573138), the National Key Research and Development Program of China, China (No. 2016YFA0203000), and the Plan for "National Youth Talents" of the Organization Department of the Central Committee.

  
  
• 1. [1]

     Z. Jiang, X. Xu, Y. Ma, et al., Nature 586 (2020) 549–554. doi: 10.1038/s41586-020-2738-2

  2. [2]

     J.L. White, M.F. Baruch, J.E. Pander, et al., Chem. Rev. 115 (2015) 12888–12935. doi: 10.1021/acs.chemrev.5b00370

  3. [3]

     C. Hepburn, E. Adlen, J. Beddington, et al., Nature 575 (2019) 87–97. doi: 10.1038/s41586-019-1681-6

  4. [4]

     A. Rafiee, K.R. Khalilpour, D. Milani, M. Panahi, J. Environ. Chem. Eng. 6 (2018) 5771–5794. doi: 10.1016/j.jece.2018.08.065

  5. [5]

     W. Tu, Y. Zhou, Z. Zou, Adv. Mater. 26 (2014) 4607–4626. doi: 10.1002/adma.201400087

  6. [6]

     U. Ulmer, T. Dingle, P.N. Duchesne, et al., Nat. Commun. 10 (2019) 1–12. doi: 10.1038/s41467-018-07882-8

  7. [7]

     O.K. Varghese, M. Paulose, T.J. LaTempa, C.A. Grimes, Nano Lett. 9 (2009) 731–737. doi: 10.1021/nl803258p

  8. [8]

     S. Gong, G. Zhu, R. Wang, et al., Appl. Catal. B: Environ. 297 (2021) 120413. doi: 10.1016/j.apcatb.2021.120413

  9. [9]

     Y. Wang, Z. Zhang, L. Zhang, et al., J. Am. Chem. Soc. 140 (2018) 14595–14598. doi: 10.1021/jacs.8b09344

  10. [10]

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

  11. [11]

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

  12. [12]

      P. Qiu, J. Wang, Z. Liang, et al., Chin. Chem. Lett. 32 (2021) 3501–3504. doi: 10.1016/j.cclet.2021.03.077

  13. [13]

      Z. Liang, Y. Xue, X. Wang, et al., Chem. Eng. J. 421 (2021) 130016. doi: 10.1016/j.cej.2021.130016

  14. [14]

      Z. Liang, X. Meng, Y. Xue, et al., J. Colloid Interface Sci. 598 (2021) 172–180. doi: 10.1016/j.jcis.2021.04.066

  15. [15]

      X. Jiao, K. Zheng, L. Liang, et al., Chem. Soc. Rev. 49 (2020) 6592–6604. doi: 10.1039/d0cs00332h

  16. [16]

      T. Kong, Y. Jiang, Y. Xiong, Chem. Soc. Rev. 49 (2020) 6579–6591. doi: 10.1039/c9cs00920e

  17. [17]

      T. Zhang, W. Lin, Chem. Soc. Rev. 43 (2014) 5982–5993. doi: 10.1039/C4CS00103F

  18. [18]

      Y. Zhang, B. Xia, J. Ran, et al., Adv. Energy Mater. 10 (2020) 1903879. doi: 10.1002/aenm.201903879

  19. [19]

      C. Gao, J. Low, R. Long, et al., Chem. Rev. 120 (2020) 12175–12216. doi: 10.1021/acs.chemrev.9b00840

  20. [20]

      S. Ji, Y. Chen, X. Wang, et al., Chem. Rev. 120 (2020) 11900–11955. doi: 10.1021/acs.chemrev.9b00818

  21. [21]

      Z. Li, S. Ji, Y. Liu, et al., Chem. Rev. 120 (2020) 623–682. doi: 10.1021/acs.chemrev.9b00311

  22. [22]

      C. Gao, S. Chen, Y. Wang, et al., Adv. Mater. 30 (2018) e1704624. doi: 10.1002/adma.201704624

  23. [23]

      Y.B. Lu, Z.H. Zhang, H.M. Wang, Y. Wang, Appl. Catal. B: Environ. 292 (2021) 120162. doi: 10.1016/j.apcatb.2021.120162

  24. [24]

      S. An, G. Zhang, T. Wang, et al., ACS Nano 12 (2018) 9441–9450. doi: 10.1021/acsnano.8b04693

  25. [25]

      A. Wang, J. Li, T. Zhang, Nat. Rev. Chem. 2 (2018) 65–81. doi: 10.1038/s41570-018-0010-1

  26. [26]

      X. Zhou, Acta Physico-Chimica Sinica 37 (2021) 2008064.

  27. [27]

      B. Qiao, A. Wang, X. Yang, et al., Nat. Chem. 3 (2011) 634–641. doi: 10.1038/nchem.1095

  28. [28]

      N. Aguilar, M. Atilhan, S. Aparicio, Appl. Surf. Sci. 534 (2020) 147611. doi: 10.1016/j.apsusc.2020.147611

  29. [29]

      M.B. Gawande, P. Fornasiero, R. Zboril, ACS Catal., 10 (2020) 2231–2259. doi: 10.1021/acscatal.9b04217

  30. [30]

      C. Zupen, E. Vorobyeva, S. Mitchell, et al., Natl. Sci. Rev. 5 (2018) 642–652. doi: 10.1093/nsr/nwy048

  31. [31]

      P. Wang, D.Y. Zhao, L.W. Yin, Energy Environ. Sci. 14 (2021) 1794–1834. doi: 10.1039/d0ee02651d

  32. [32]

      X. Jiao, X. Li, X. Jin, et al., J. Am. Chem. Soc. 139 (2017) 18044–18051. doi: 10.1021/jacs.7b10287

  33. [33]

      Y. Wang, J. Mao, X.G. Meng, et al., Chem. Rev. 119 (2019) 1806–1854. doi: 10.1021/acs.chemrev.8b00501

  34. [34]

      W. Shao, X.D. Zhang, Nanoscale 13 (2021) 7081–7095. doi: 10.1039/d1nr00649e

  35. [35]

      P. Chen, B. Lei, X. Dong, et al., ACS Nano 14 (2020) 15841–15852. doi: 10.1021/acsnano.0c07083

  36. [36]

      Q.S. Wang, D.F. Zhang, Y. Chen, et al., ACS Sustain. Chem. Eng. 7 (2019) 6430–6443. doi: 10.1021/acssuschemeng.8b06273

  37. [37]

      W. Zheng, Y. Wang, L. Shuai, et al., Adv. Funct. Mater. 31 (2021) 2008146. doi: 10.1002/adfm.202008146

  38. [38]

      Y. Zhang, X. Wang, S. Zheng, et al., Adv. Funct. Mater. 31 (2021) 2104377. doi: 10.1002/adfm.202104377

  39. [39]

      X. Wang, S. Feng, W. Lu, et al., Adv. Funct. Mater. 31 (2021) 2104243. doi: 10.1002/adfm.202104243

  40. [40]

      J. Chen, Z. Li, X. Wang, et al., Angew. Chem. Int. Ed. 61 (2022) e202111683.

  41. [41]

      S.F. Ji, Y. Qu, T. Wang, et al., Angew. Chem. Int. Ed. 59 (2020) 10651–10657. doi: 10.1002/anie.202003623

  42. [42]

      X.X. Zhao, K.P. Loh, S.J. Pennycook, J. Phys. Condens. Matter 33 (2021) 063001. doi: 10.1088/1361-648x/abbdb9

  43. [43]

      Y.A. Zhou, J. Li, X.P. Gao, et al., J. Mater. Chem. A 9 (2021) 9979–9999. doi: 10.1039/d1ta00154j

  44. [44]

      X. Li, X. Yang, J. Zhang, et al., ACS Catal. 9 (2019) 2521–2531. doi: 10.1021/acscatal.8b04937

  45. [45]

      J. Jones, H. Xiong, A.T. DeLaRiva, et al., Science 353 (2016) 150–154. doi: 10.1126/science.aaf8800

  46. [46]

      P. Zhou, F. Lv, N. Li, et al., Nano Energy 56 (2019) 127–137. doi: 10.1016/j.nanoen.2018.11.033

  47. [47]

      Z. Wang, Y. Huang, L. Chen, et al., J. Mater. Chem. A 6 (2018) 972–981. doi: 10.1039/C7TA09132J

  48. [48]

      S. Yin, J. Han, T. Zhou, R. Xu, Catal. Sci. Technol. 5 (2015) 5048–5061. doi: 10.1039/C5CY00938C

  49. [49]

      K.S. Lakhi, D.H. Park, K. Al-Bahily, et al., Chem. Soc. Rev. 46 (2017) 72–101. doi: 10.1039/C6CS00532B

  50. [50]

      S. Cao, J. Low, J. Yu, M. Jaroniec, Adv. Mater. 27 (2015) 2150–2176. doi: 10.1002/adma.201500033

  51. [51]

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

  52. [52]

      Y.C. Zhang, N. Afzal, L. Pan, et al., Adv. Sci. 6 (2019) 1900053. doi: 10.1002/advs.201900053

  53. [53]

      Z. Wang, M. Chen, Y. Huang, et al., Appl. Catal. B: Environ. 239 (2018) 352–361. doi: 10.1016/j.apcatb.2018.08.030

  54. [54]

      E. Luo, Y. Chu, J. Liu, et al., Energy Environ. Sci. 14 (2021) 2158–2185. doi: 10.1039/d1ee00142f

  55. [55]

      X.B. Zhu, B. Hu, C.X. Wang, et al., J. Catal. 391 (2020) 1–10. doi: 10.1016/j.jcat.2020.08.016

  56. [56]

      W. Liu, L.L. Cao, W.R. Cheng, et al., Angew. Chem. Int. Ed. 56 (2017) 9312-1317. doi: 10.1002/anie.201704358

  57. [57]

      Y.R. Li, Z.W. Wang, T. Xia, et al., Adv. Mater. 28 (2016) 6959–6965. doi: 10.1002/adma.201601960

  58. [58]

      X. Wang, X. Chen, A. Thomas, et al., Adv. Mater. 21 (2009) 1609–1612. doi: 10.1002/adma.200802627

  59. [59]

      X.H. Jiang, L.S. Zhang, H.Y. Liu, et al., Angew. Chem. Int. Ed. 59 (2020) 23112–23116. doi: 10.1002/anie.202011495

  60. [60]

      X.G. Li, W.T. Bi, L. Zhang, et al., Adv. Mater. 28 (2016) 2427–2431. doi: 10.1002/adma.201505281

  61. [61]

      H. Su, W. Che, F.M. Tang, et al., J. Phys. Chem. C 122 (2018) 21108–21114. doi: 10.1021/acs.jpcc.8b03383

  62. [62]

      Z. Zeng, Y. Su, X. Quan, et al., Nano Energy 69 (2020) 104409. doi: 10.1016/j.nanoen.2019.104409

  63. [63]

      S. Cao, H. Li, T. Tong, et al., Adv. Funct. Mater. 28 (2018) 1802169. doi: 10.1002/adfm.201802169

  64. [64]

      X.D. Xiao, Y.T. Gao, L.P. Zhang, et al., Adv. Mater. 32 (2020) 2003082. doi: 10.1002/adma.202003082

  65. [65]

      Y. Zhao, Z. Han, G. Gao, et al., Adv. Funct. Mater., (2021) 2104976. doi: 10.1002/adfm.202104976

  66. [66]

      Q. Chen, G. Gao, Y. Zhang, et al., J. Mater. Chem. A 9 (2021) 15820–15826. doi: 10.1039/d1ta02926f

  67. [67]

      Y. Cao, L. Guo, M. Dan, et al., Nat. Commun. 12 (2021) 1675. doi: 10.1038/s41467-021-21925-7

  68. [68]

      J.M. Zhang, X.P. Xu, L. Yang, et al., Small Methods 3 (2019) 1900653. doi: 10.1002/smtd.201900653

  69. [69]

      X. Li, J.G. Yu, M. Jaroniec, X.B. Chen, Chem. Rev. 119 (2019) 3962–4179. doi: 10.1021/acs.chemrev.8b00400

  70. [70]

      Y.L. Yang, F. Li, J. Chen, et al., ChemSusChem 13 (2020) 1979–1985. doi: 10.1002/cssc.202000375

  71. [71]

      P. Sharma, S. Kumar, O. Tomanec, et al., Small 17 (2021) 2006478. doi: 10.1002/smll.202006478

  72. [72]

      S.F. Tang, X.P. Yin, G.Y. Wang, et al., Nano Res. 12 (2019) 457–462. doi: 10.1007/s12274-018-2240-4

  73. [73]

      Z. Zhao, W. Liu, Y. Shi, et al., PCCP 23 (2021) 4690–4699. doi: 10.1039/d0cp05570k

  74. [74]

      J. Fu, L. Zhu, K. Jiang, et al., Chem. Eng. J. 415 (2021) 128982. doi: 10.1016/j.cej.2021.128982

  75. [75]

      J. Di, C. Chen, S. Yang, et al., Nat. Commun. 10 (2019) 2840. doi: 10.1038/s41467-019-10392-w

  76. [76]

      X. Ren, L. Shi, Y. Li, et al., ChemCatChem 12 (2020) 3870–3879. doi: 10.1002/cctc.202000546

  77. [77]

      H. Zhang, J. Wei, J. Dong, et al., Angew. Chem. Int. Ed. 55 (2016) 14308–14312.

  78. [78]

      L. Cheng, H. Yin, C. Cai, et al., Small 16 (2020) 2002111. doi: 10.1002/smll.202002111

  79. [79]

      J. Wang, T. Heil, B. Zhu, et al., ACS Nano 14 (2020) 8584–8593. doi: 10.1021/acsnano.0c02940

  80. [80]

      M. Kou, W. Liu, Y. Wang, et al., Appl. Catal. B: Environ. 291 (2021) 120146. doi: 10.1016/j.apcatb.2021.120146

  81. [81]

      S.F. Ji, Y. Qu, T. Wang, et al., Angew. Chem. Int. Ed. 59 (2020) 10651–10657. doi: 10.1002/anie.202003623

  82. [82]

      S. Sun, G. Zhang, N. Gauquelin, et al., Sci. Rep. 3 (2013) 1775. doi: 10.1038/srep01775

  83. [83]

      Y. Ren, Y. Tang, L. Zhang, et al., Nat. Commun. 10 (2019) 4500. doi: 10.1038/s41467-019-12459-0

  84. [84]

      K. Fujiwara, S.E. Pratsinis, Appl. Catal. B: Environ. 226 (2018) 127–134. doi: 10.1016/j.apcatb.2017.12.042

  85. [85]

      N. Daelman, M. Capdevila-Cortada, N. Lopez, Nat. Mater. 18 (2019) 1215–1221. doi: 10.1038/s41563-019-0444-y

  86. [86]

      G.P. Gao, Y. Jiao, E.R. Waclawik, A.J. Du, J. Am. Chem. Soc. 138 (2016) 6292–6297. doi: 10.1021/jacs.6b02692

  87. [87]

      P. Zhou, Y.G. Chao, F. Lv, et al., Sci. Bull. 65 (2020) 720–725. doi: 10.1016/j.scib.2019.12.025

  88. [88]

      Y.J. Cao, D.H. Wang, Y. Lin, et al., ACS Appl. Energy Mater. 1 (2018) 6082–6088. doi: 10.1021/acsaem.8b01143

  89. [89]

      Y. Xue, Y.G. Lei, X.Y. Liu, et al., New J. Chem. 42 (2018) 14083–14086. doi: 10.1039/C8NJ02933D

  90. [90]

      X. Huang, H. Yan, L. Huang, et al., J. Phys. Chem. C 123 (2019) 7922–7930. doi: 10.1021/acs.jpcc.8b07181

  91. [91]

      N. Wang, J. Wang, J. Hu, et al., ACS Appl. Energy Mater. 1 (2018) 2866–2873. doi: 10.1021/acsaem.8b00526

  92. [92]

      Z. Chen, Y. Bu, L. Wang, et al., Appl. Catal. B: Environ. 274 (2020) 119117. doi: 10.1016/j.apcatb.2020.119117

  93. [93]

      M. Ou, S.P. Wan, Q. Zhong, et al., Int. J. Hydrogen Energy 42 (2017) 27043–27054. doi: 10.1016/j.ijhydene.2017.09.047

  94. [94]

      S. Fang, X.R. Zhu, X.K. Liu, et al., Nat. Commun. 11 (2020) 1029. doi: 10.1038/s41467-020-14848-2

  95. [95]

      G. Liu, Y. Huang, H. Lv, et al., Appl. Catal. B: Environ. 284 (2021) 119683. doi: 10.1016/j.apcatb.2020.119683

  96. [96]

      X.H. Huang, Y.J. Xia, Y.J. Cao, et al., Nano Res. 10 (2017) 1302–1312. doi: 10.1007/s12274-016-1416-z

  97. [97]

      Z. Chen, Y. Chen, S. Chao, et al., ACS Catal. 10 (2020) 1865–1870. doi: 10.1021/acscatal.9b05212

  98. [98]

      C. Dong, C. Lian, S. Hu, et al., Nat. Commun. 9 (2018) 1252. doi: 10.1038/s41467-018-03666-2

  99. [99]

      R. Li, W. Zhang, K. Zhou, Adv. Mater. 30 (2018) 1705512. doi: 10.1002/adma.201705512

  100. [100]

       S. Wang, J. Lin, X. Wang, Phys. Chem. Chem. Phys. 16 (2014) 14656–14660. doi: 10.1039/c4cp02173h

  101. [101]

       Y. Gong, B. Shao, J. Mei, et al., Nano Res. 15 (2021) 551–556.

  102. [102]

       Y. Li, B. Li, D. Zhang, et al., ACS Nano 14 (2020) 10552–10561. doi: 10.1021/acsnano.0c04544

  103. [103]

       J. Yang, Z. Wang, J. Jiang, et al., Nano Energy 76 (2020) 105059. doi: 10.1016/j.nanoen.2020.105059

  104. [104]

       X. Cui, X. Dai, A. -. E. Surkus, et al., Chinese J. Catal. 40 (2019) 1679–1685. doi: 10.1016/S1872-2067(19)63316-4

  105. [105]

       Z.M. Migaszewski, A. Galuszka, Crit. Rev. Env. Sci. Technol. 45 (2015) 429–471. doi: 10.1080/10643389.2013.866622

  106. [106]

       Z. Han, Y. Zhao, G. Gao, et al., Small 17 (2021) 2102089. doi: 10.1002/smll.202102089

• 

  Figure 1  Potential sites of single atom supported on g-C₃N₄: (a) pore structure; (b) interlayer space; (c) defect sites. Color codes: carbon is brown, oxygen is red, and a confined single atom is green.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Electron transfer between Au and CdS with different types of vacancies: (a) Cd vacancies; (b) S vacancies. Reprinted with permission [67]. Copyright 2021, Springer Nature.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  The proposed mechanism for single rare-earth element atoms as ET bridges to promote charge separation.

  下载: 全尺寸图片 幻灯片

  Table 1.  Summary of various SAC@2D toward CO₂ reduction.

  下载: 导出CSV
•