1 Introduction

2 Covalent organic frameworks

2.1 Basic information of COFs

2.2 Application of COFs in photocatalysis

3 Basic principles of photocatalytic CO₂reduction

4 COFs for photocatalytic CO₂reduction

5 Conclusion and outlook

Covalent Organic Frameworks (COFs), as a new class of crystalline porous Organic polymer materials, are characterized by periodic permanent porosity, tunable microstructure, highly conjugated structure and high stability. The excellent characteristics endow COFs with more outstanding photocatalytic performance than classical photocatalysts such as TiO₂, cadmium sulfide (CdS), graphitized carbon nitride (g-C₃N₄) and metal-organic frameworks (MOFs), and make them excellent supports for photocatalytic reduction reaction catalysts. Moreover, compared with inorganic nanomaterials, graphite C₃N₄ and MOFs, COFs as photocatalysts for CO₂ reduction show the following advantages: (1) COFs usually have large specific surface area and abundant heteroatoms (N, O, S, etc.), and have higher adsorption capacity and selectivity for CO₂ compared with traditional inorganic semiconductor materials; (2) The framework structure of COFs is connected by covalent bonds, so it has stronger thermal and chemical stability, which makes COFs have excellent potential in the field of photocatalysis; (3) COFs have regular crystal structure and regular molecular arrangement, which reduces the electron recombination phenomenon in the reaction process and improves the electron transfer efficiency; (4) The structure of COFs is easy to design, modify and adjust, so COFs with expected electronic structure can be synthesized to realize or test the researcher's idea; (5) Two-dimensional COFs have a large number of conjugated π electrons, which endows COFs with good light absorption and conductivity^[1].

Recent studies have shown that COFs have shown great potential in the field of photocatalysis, among which the application of COFs in photocatalytic CO₂ reduction is quite promising. In this paper, we will mainly introduce the research progress of novel COFs for photocatalytic CO₂ reduction in recent years, and systematically and thoroughly analyze the relationship between the design concept and function of each work and its reference significance, in order to promote the further development of this field.

In order to cope with the increasingly serious environmental problems and energy crisis caused by fossil fuel combustion, artificial photosynthesis has become one of the most promising methods to solve these problems. Using inexhaustible solar energy as the driving force to reduce CO₂ into high value-added chemical fuels (CO,HCOOH,HCHO,CH₃OH,CH₄, etc.) Is the most attractive method. The exploration of a suitable photocatalyst is essential for the efficient implementation of this catalytic process.

Generally, the photocatalytic process includes three steps: light harvesting, electron-hole separation, and redox reaction (Fig. 2). Furthermore, the photocatalytic Lowest Unoccupied Molecular Orbital reduction reaction includes the following three processes: in the first stage, after the photocatalyst with a suitable band gap absorbs a photon with a certain energy (that is, when the hv>E_g), the ground state electron absorbs the energy and transforms into the excited state, that is, from the LUMO (Lowest Unoccupied Molecular Orbital, LUMO) state.Lowest Occupied Molecular Orbital) to HOMO (Highest Occupied Molecular Orbital), and the transition of electrons leads to the formation of electron-hole pairs (Highest Occupied Molecular Orbital) with certain energy; In the second stage, the photogenerated electrons and holes are separated from each other and transferred to the active sites on the surface of the catalyst for the next reduction and oxidation reactions. In the third stage, the electron – proton is injected into the substrate CO₂ molecule adsorbed at the active site, thereby catalyzing the reduction reaction of the CO₂ molecule on the catalyst surface^[19].

Based on the mechanism of photocatalytic CO₂ reduction, it is generally believed that photocatalysts for efficient photoreduction of CO₂ need to possess several characteristics: (1) appropriate energy band combined with strong light absorption ability to achieve efficient utilization of solar energy; (2) strong chemical adsorption ability for CO₂; (3) sufficient photogenerated electron-hole pairs that can be rapidly separated to ensure the continuity and high efficiency of the reaction under the action of the catalyst; (4) The active site of side reaction can be removed, and the side reaction can be inhibited by rational design to achieve high selectivity.

Combined with the above discussion, it can be seen that the structural characteristics of COFs materials make them have the potential to become excellent photocatalysts, and their rationally designed structures can achieve both suitable energy bands and sufficient active sites, so as to efficiently complete the photocatalytic CO₂ reduction reaction.

It can be seen from the above that many characteristics of COFs make them have the ability and potential to meet the requirements of photocatalytic CO₂ reduction. Therefore, the application research of COFs in the field of photocatalysis has been developed rapidly^[20]. In this paper, the research progress of COFs in the field of photocatalytic CO₂ reduction in recent years will be introduced in chronological order.

In the research of COFs in the field of photocatalytic CO₂ reduction, the first problem to be solved is how to overcome the lack of catalytic active sites of this pure organic material. Here, we can take advantage of the inherent structural features of COFs to further design and synthesize COFs with efficient active sites, which is also a challenging and attractive topic. In 2018, Huang's group used the photoactive 2D triazine-based conjugated COF as the basic framework and introduced the trimethyl chloro (bipyridine) rhenium complex (Re(bpy)(CO)₃Cl) through the post-modification method as the catalytic active site for CO₂ reduction, thus obtaining a heterogeneous COFs photocatalyst for CO₂ photoreduction (hereinafter denoted as Re-COF, Fig. 3)^[21]. The test results show that Re-COF can effectively reduce CO₂ to CO (15 mmol/G) with high selectivity (98%) and durability (> 20 H) under the irradiation of Xe lamp (cut-off wavelength = 420 nm) with TEOA as sacrificial agent, and its activity is better than that of homogeneous Re catalyst. More importantly, this work combined in situ measurement and time-resolved absorption spectroscopy to reveal the key intermediate species for CO₂ reduction.

In addition, by transient electron absorption spectroscopy/X-ray absorption spectroscopy analysis, this work demonstrates that rapid intermolecular charge transfer (ICT) from the excited state COF to the Re component through an energy transfer (ET) process occurs in the COFs structure. This work not only demonstrates the potential of COFs as next-generation photocatalysts for solar energy conversion, but also provides unprecedented insights into the mechanistic origin of light-driven carbon dioxide reduction.

The above studies suggest that metal modification of COFs materials to increase their active sites is an effective way to design efficient photocatalytic CO₂ reduction catalysts for COFs. On this basis, in 2019, Lan's group reported a series of transition-metal-modified COFs that can be used for heterogeneous photocatalytic CO₂ reduction reaction^[22]. A series of DQTP COF-M (M = CO/Ni/Zn) catalysts were obtained by coordinatively loading different kinds of open metal active sites onto the structural framework of COFs, and it was found that the catalytic activity and selectivity of CO₂ reduction products (Co or formic acid) were greatly affected by different metal centers (Fig. 4). Among them, DQTP COF-CO has a higher CO yield of 1.02×10³μmol·h^-1·g^-1, while DQTPCOF-Zn has a higher selectivity for formic acid formation (90% higher than CO). This work highlights the great potential of utilizing stable COFs as substrates and thereby anchoring metal-rich active sites for heterogeneous CO₂ reduction.

The practical application of COFs is not only limited by the catalytic ability, but also the difficulty of large-scale synthesis, which is an urgent problem to be solved. In response to this defect, in 2019, Jiang's team proposed a scalable and general bottom-up method to achieve large-scale (> 100 mg) and high-yield (> 55%) synthesis of ultrathin (< 2.1 nm) imine-based COF nanosheets (NSs) (including COF-366NSs, COF-367NSs, COF-367-CoNSs, TAPB-PDA COFNSs and TAPB-BPDA COFNSs) (Figure 5)^[9]. Among them, ultrathin COF-367-CO NSs are efficient heterogeneous photocatalysts for promoting CO₂-CO conversion, with Co generation rates up to 10 162μmol·g^-1·h^-1 and selectivity up to 78% under visible light irradiation in aqueous solution. This material is comparable to the state-of-the-art visible-light-driven heterogeneous catalysts reported to date.

With the development of COFs photocatalytic CO₂ reduction, a variety of COFs materials have been reported to obtain efficient catalytic performance by increasing the active sites of COFs. For example, in 2019, Wu's group proposed a covalent triazine-based framework (CTF) -based photocatalyst^[23]. The nitrogen-rich triazine group in CTF contributes to the adsorption of CO₂, while its periodic pore structure is beneficial to the regulation of CO₂ and electron mediation. In this work, the authors found that the immobilization of cobalt on CTF could significantly improve its photocatalytic activity, which could reach 44 times of the original CTF activity, and the maximum CO production rate of Co/CTF could reach 50μmol·g^-1. The results show that CTF not only acts as a carrier of active Co species, but also connects the photosensitizer with the cobalt catalytic site, thus allowing efficient transfer of photoexcited electrons.

In 2019, Zou's group designed a COFs material loaded with single Ni site (Ni-TpBpy)^[24]. Under visible light irradiation, this COFs can transfer electrons from the photosensitizer to the Ni site to produce CO while activating the CO₂ molecule (Fig. 6). Among them, Ni-TpBpy showed good activity, and the CO yield rate obtained in 5 H reaction was 4057μmol·g^-1, and the CO selectivity was 96%, which avoided the side reaction of hydrogen production. More importantly, when the partial pressure of CO₂ was reduced to 0.1 atm, the selectivity to CO products using these COFs was still 76%. The theoretical calculation and experimental results show that the synergistic effect of the single nickel catalytic site and TpBpy makes the COFs have good catalytic activity and selectivity. Among them, TpBpy not only adsorbs CO₂ molecules and provides Ni catalytic sites, but also can promote the activation of CO₂ while inhibiting competitive hydrogen production.

As described previously, researchers have developed numerous COFs to improve the photocatalytic CO₂ reduction performance by anchoring monometallic sites through chelating coordination, but often the binding ability of the bipyridyl unit used is not enough to stabilize the active metal center. The inevitable metal leaching phenomenon can be observed in the photocatalytic CO₂ reduction process even in the presence of excess bipyridine units. The porphyrin unit is an ideal alternative to the bipyridyl unit due to its strong chelating ability for metal ions.

In 2020, Wang's group proposed to use porphyrin-tetrastyryl COF (MP-TPE-COF, where M=H₂,Co and Ni; TPE = 4,4 ′, 4 ″, 4- (ethane-1,1,2,2-tetrayl) tetrabenzaldehyde) for photocatalytic CO₂ reduction (Fig. 7)^[25]. Taking advantage of the high performance of the metalloporphyrin unit in the selective adsorption, activation, and conversion of CO₂ and the separation and electron transfer of charge carriers, the CoP-TPE-COF achieved a CO generation rate of 2414μmol·g^-1·h^-1 under visible light irradiation with a selectivity of 61% for H₂, whereas the NiP-TPE-COF achieved a CO generation rate of 525μmol g^-1·h^-1 with a selectivity of 93%. This work provides molecular insights into the mechanism of photocatalytic CO₂ reduction and can be extended to solar energy conversion and other applications of various COFs-based catalytic materials.

At present, most COFs still need to rely on photosensitizers and sacrificial agents to achieve their high photocatalytic efficiency, which greatly reduces the economic value of COFs as photocatalysts. In order to solve the above problems, the exploration of COFs photocatalysts which do not rely on photosensitizers and sacrificial agents has been gradually developed. In 2019, the Lan project was combined into a series of crystalline porphyrin tetrathiafulvalene COF materials that can achieve efficient photocatalytic CO₂ reduction without additional photosensitizers, sacrificial agents, and noble metal co-catalysts (Fig. 8)^[26]. This COFs enable efficient photogenerated electron transfer from the tetrathiafulvalene to the porphyrin via a covalent bond, allowing electron and hole separation and thus CO₂ reduction and water oxidation. In addition, by adjusting the band structure of TTCOFs, it was found that the maximum CO yield obtained by photocatalytic CO₂ reduction of TTCOF-Zn was 12.33 μmol, and the selectivity was about 100%, while water was oxidized to oxygen as a sacrificial agent. This work provides a new idea for the design of efficient artificial crystalline photocatalysts.

In order to achieve more efficient reduction of CO₂ in water, in 2020, Lan's group proposed a strategy to covalently connect COFs with semiconductors, thereby creating a stable organic-inorganic Z-scheme heterojunction for artificial photosynthesis (Fig. 9)^[27]. Based on this strategy, a series of COF-semiconductor Z-scheme photocatalysts were assembled. Such catalysts combine water oxidation active semiconductors (TiO₂,Bi₂WO₆,α-Fe₂O₃) with CO₂ reduction COFs (COF-316/318), exhibiting high CO₂-CO conversion efficiency (69.67μmol·g^-1·h^-1 without the addition of photosensitizers and sacrificial agents). This work is the first to realize the covalently bonded COF/inorganic-semiconductor assembly as a Z-scheme for artificial photosynthesis.

Furthermore, in 2021, Lan's team first proposed a method to load polyoxometalate (POM) in the nanopores of vacant COFs and explored its application in artificial photosynthesis (Fig. 10)^[28]. Among them, TCOF-POM exhibited the highest CO yield (37.25μmol·g^-1·h^-1). The mechanism study showed that CO₂ reduction and H₂O oxidation reactions could occur on POM and COF, respectively, which showed the application prospect of COF-POMs functional materials in the field of photocatalysis.

Changing the morphology and structure of COFs can also significantly affect their catalytic performance, for example, the ultrathin COFs nanosheet structure is conducive to catalytic reactions. In 2020, Cooper's group synthesized a series of CONs (covalent organic framework nanosheet materials) by embedding a single cobalt site and used them to catalyze the CO₂ reduction reaction (Fig. 11)^[29]. Among them, a partially fluorinated and cobalt-loaded CON produced 10.1 μmol of CO under visible light irradiation for more than 6 H (TON = 28.1), and the selectivity reached 76%. Under the irradiation of 420 nm light in the presence of iridium photosensitizer, its external quantum efficiency (EQE) reaches 6.6%. In this case, it seems that the COFs material itself does not act as a photocatalyst, but as a semiconductor carrier to transfer the carrier generated in the photosensitizer to the cobalt center, which is the active site. This study found that the performance of ultrathin CONs is better than that of the corresponding bulk materials in most cases, indicating that this is a general strategy to improve the photocatalytic activity of COFs materials.

In the research of COFs photocatalytic CO₂ reduction, researchers found that not only the number of active sites and light absorption ability have an impact on the photocatalytic performance of COFs materials, but also the crystallinity and porosity can not be ignored. In 2020, Cooper's group combined a molecular catalyst, a rhenium complex [Re(bpy)(CO)₃Cl], with COF to obtain a heterogeneous catalyst with strong visible light absorption ability and high CO₂ binding affinity^[30]. The COF can catalyze the reduction of CO₂ to CO under the irradiation of visible light, the output rate of CO can reach 1040 mmol·g^-1·h^-1, and the selectivity can reach 81%; In the case of adding photosensitizer, the CO efflux rate can reach 1400 mmol·g^-1·h^-1, and the selectivity is 86%. This study shows that the crystallinity of COFs is beneficial to their photocatalytic performance in the reduction of carbon dioxide. Crystallinity and porosity appear to be important in these materials, as amorphous, low-porosity analogs show little photocatalytic activity.

In order to increase the active sites of COFs, various metal ions have been used to modify COFs to improve their photocatalytic performance, but the principle and further development of such methods remain to be explored. Recent studies have shown that not only the type of metal, but also the spin state of metal ions have an impact on its performance. In 2020, Jiang's research group added Co element to the porphyrin center of COF-367-Co, and the Co (II) center in the framework was controllably oxidized to Co (III) in the air without affecting the overall structure (Fig. 12), so it was possible to explore the effect of the spin state transition of Co element on the photocatalytic carbon dioxide reduction performance of COF (its spin state was controlled by the oxidation state)^[31]. In the report, compared with COF-367-CO (Ⅱ) (Co (Ⅱ), S = 1/2), the activity and selectivity of COF-367-Co (Ⅲ) (Co (Ⅲ), S = 0) for the reduction of CO₂ to formic acid in photocatalysis were significantly improved, and the selectivity for the reduction of CO₂ to Co and methane was significantly lower.

Using DFT calculations, the team further explained the reasons for the significant differences in photocatalytic activity and selectivity of the two COFs. All the Photoelectrochemical properties show that COF-367-Co (Ⅲ) has a higher charge separation efficiency than COF-367-Co (Ⅱ), which also explains the improved activity of the former. The experimental results show that the metal spin state in COFs plays a crucial role in the photocatalytic process. It should be noted that this is the first report on the modulation of photocatalytic performance by spin state manipulation.

In the development of the field of COFs photocatalysis, some researchers have also developed new ideas to improve the photocatalytic performance of COFs, that is, to improve the electron-hole pairs in COFs and increase their electron transfer rate at the same time. In 2020, Kong's group used the Schiff base reaction of carbazole-triazinyl D-A monomer to construct a novel donor-acceptor (D-A) COF with suitable band structure, strong visible light harvesting ability and abundant nitrogen sites, namely CT-COF (Fig. 13)^[32]. In this structure, the carbazole group is rich in electrons and has good hole transport ability, while the triazine group has a large electron affinity, so the overall COFs have a high electron mobility. CT-COF, as a metal-free photocatalyst, can reduce CO₂ to CO with gaseous water as electron donor under visible light irradiation without catalyst. Under the same conditions, the production rate of CO (102.7μmol·g^-1·h^-1) of this material is 68.5 times that of g-C₃N₄. In situ Fourier transform (FT) infrared analysis shows that CT-COF can adsorb and activate carbon dioxide and water molecules, and the ^*COOH species may be a key intermediate. DFT calculations showed that the nitrogen atom in the triazine ring may be the active site for photocatalysis.

The effect of the nano-morphology of the catalyst on the photocatalytic performance has also been intensively investigated. In 2021, Lan's group reported a MOF sacrificial in situ acid etching (MSISAE) strategy, which can continuously synthesize COFs-based core-shell, yolk-shell and hollow sphere nanocomposites by adjusting the decomposition rate of MOF core (Fig. 14)^[33]. Among them, the designed morphology of yolk shell and hollow spheres endows COF materials with more exposed active sites, faster electron transfer rate, and stronger light energy utilization. Among the synthesized materials, the NH₂-MIL-125/TiO₂@COF-366-Ni-OH-HAc (yolk-shell) material has higher photocatalytic CO₂-CO conversion efficiency in gas-solid mode. The MSISAE strategy developed in this study enables precise morphology design and control of multi-component hybrid composites based on COFs, paving a new way for the development of multifunctional catalytic reactions with powerful superstructures.

On the basis of the successful introduction of various monometallic ions, the application of bimetallic nanoclusters in COFs has also been successful. In 2021, the Lu research group first loaded the photosensitized N₃-COF confinement effect with ultrafine bimetallic PdIn NCs, and obtained a series of Pd_xIn_y@N₃-COF composites (Fig. 15)^[34]. The test results showed that the PdIn@N₃-COF could be used for photocatalytic CO₂ reduction and water oxidation at the same time, and had good alcohol production performance, with a total yield of 798μmol·g^-1 (methanol 74%, ethanol 26%) in 24 H, which was 9. 7 times higher than that of N₃-COF alone. The excellent performance is related to the bimetallic synergistic effect of Pd_xIn_y@N₃-COF. This is also the first COF-based photocatalyst for photocatalytic reduction of CO₂ and oxidation of water to produce ethanol.

In summary, the design and synthesis of novel COFs and their composites and the latest research progress in the field of photocatalytic CO₂ reduction in recent years are systematically summarized in this paper.Including the introduction of different metal ions to provide active sites by using the structural characteristics of COFs, the increase of photosensitive functional groups to improve the utilization of visible light and other methods to achieve efficient CO₂ photoreduction, and further coupling water oxidation with CO₂ reduction to achieve artificial photosynthesis. Based on the summary of the above work, it is expected to promote the further rapid development of this field. Looking into the future research directions, COFs, as a new class of designable crystalline porous materials linked by covalent bonds with high porosity, large surface area, and low density, still have great potential in the field of photocatalytic CO₂ reduction. However, at present, this potential has not been fully realized. Most of the high catalytic performance of COFs depends on photosensitizers and sacrificial agents, while some of the realization of COFs itself as photosensitizers and catalysts often shows low catalytic efficiency. Among them, the introduction of appropriate photosensitive functional groups to enhance its light absorption capacity, the insertion of metal ions and even the design of bimetallic catalysts based on COFs to enhance its internal electron transfer rate are further explored.The exploration of new synthesis methods to obtain new catalyst structures, the establishment of theoretical models and the in-depth exploration of their catalytic principles have provided ideas and opened up prospects for COFs to design and prepare more efficient photocatalytic CO₂ reduction COFs-based catalysts. Second, using COFs to catalyze the reduction of CO₂ to high value-added products such as C₂₊ (ethylene, ethane, propylene, propane, ethanol, acetic acid, etc.) Is the next development direction. In a word, the application of COFs in the field of photocatalytic CO₂ photoreduction is still in the basic research stage, and how to further realize its practical application remains to be further explored.