Mechanistic Approach towards Designing Covalent Organic Frameworks for Photocatalytic Hydrogen Generation

Abstract

Covalent organic frameworks (COFs) have unique features, including intrinsic porosity, crystallinity, and tunability, making them desirable materials for diverse applications ranging from environmental remediation to energy harvesting. Among these applications, COFs are extensively studied for their photocatalytic hydrogen evolution by converting solar energy into clean and renewable fuel via water splitting. COFs have several advantages over conventional inorganic catalysts, such as tunable band structures, high surface areas, and low cost. However, the research in this field is still in the early stages, and COFs still face some challenges, such as low charge carrier mobility, high exciton binding energy, and poor stability. To overcome these challenges, various design strategies relying on a mechanistic approach have been developed to design and modify COFs for enhanced photocatalytic performance. These include extending the π-conjugation, incorporating heteroatoms or metal complexes, and donor–acceptor (D–A) configuration, which ultimately improves the light absorption charge separation of COFs. Additionally, blending COFs with other functional materials, such as inorganic-organic semiconductors, can create synergistic effects to boost photocatalytic activity. In this review, the design aspects of the fabrication of COFs as effective photocatalysts have been reported.

1. Introduction

Non-renewable fossil fuels, which include coal, oil, and natural gas, currently supply about 87% of the world’s energy [1]. However, unwarranted usage of fossil fuels results in catastrophic environmental damage and resource depletion. In contrast, renewable energy sources like solar energy, wind energy, biomass, geothermal energy, tidal energy, etc., are abundant and have a smaller environmental impact, making them a more secure and reliable alternative energy source. It is worth mentioning that the heat of hydrogen combustion is the highest among all available bio, chemical, and fossil fuels [2]. Hydrogen combustion produces water as a final product, making it an ideal energy source. The production of hydrogen from water splitting through solar energy is considered a sustainable solution to address the expected future energy crisis. Fortunately, both water and solar energy exist in abundance. However, the photochemical splitting of water is a thermodynamically unfavorable process and faces several challenges for practical applications [3]. The most important has been the realization of efficient photocatalysts, which has triggered immense research interest to explore various materials in this line.

Regarding photocatalyst efficiency, it is essential to cross the 10% solar-to-hydrogen (STH) conversion barrier, which is considered a benchmark in the current research scenario [2]. Honda and Fujishima reported the first instance of photocatalytic splitting of water using a TiO₂ electrode, followed by several researchers worldwide in an attempt to achieve higher efficiency [4]. After this seminal work, excellent progress in developing solar energy-driven water splitting has been reported. The majority of the reported catalysts are heterogeneous metal oxides, (oxy) sulfides, oxyhalides, and chalcopyrite [5]. Recently, great progress has been observed in exploiting the metal-organic framework in hydrogen evolution/overall water splitting [6]. However, these catalysts, specially metal-based ones, face two distinct challenges: lack of structural tunability and environmental toxicity. These challenges may be addressed by using organic molecule-based photocatalysts. The journey from inorganic to organic photocatalysts included the use of several nonmetals, including B, C, N, P, and S. Among them, carbon nitride initially attracted great interest as an organic photocatalyst [7]. In 1985, Shozo Yanagida et al. reported the first organic molecule-based Poly(phenylene) polymer for photocatalytic hydrogen generation from water [8]. Since then, several organic photocatalysts, including bipyridine-based linear polymers [9], phenyl copolymers with fluorene [10], and electron-donor acceptor-based copolymers [11,12], etc., have been reported. In 2009, Domen and Antonietti reported the first instance of graphene carbon nitride as a photocatalyst [13]. Since then, several successful examples have been reported by the researcher worldwide [14]. Although excellent progress has been made in modifying carbon nitride by new synthetic routes and post-synthetic modifications, their degree of control and practical application are still challenging. For example, harsh reaction conditions such as several hundred degrees Celsius are required to prepare them, limiting the choice of starting materials and post-functional group engineering. Therefore, new materials such as MXenes [15], metal-based materials [16,17,18], zeolites [19], polymers [20,21], and composite materials [22,23,24,25] emerged as alternative photocatalysts.

Among the emerging materials, covalent organic frameworks (COFs) have great potential as photocatalytic materials for hydrogen generation from water owing to their aromatic framework, tunable microporous structure, and crystallinity [26,27,28]. COFs are organic molecules linked by the covalent bond using thermodynamically controlled dynamic covalent chemistry; typically, they are crystalline and can be synthesized at comparatively lower temperatures, which gives complete control over the selection of monomers. COFs were first reported by Omar Yagi in 2005; since then, COFs have attracted extensive research interest because of their long-range order, crystalline structure, porosity, and freedom for picking the suitable monomer [27]. In recent years, COFs have gained sufficient attention in heterogeneous catalysis [29,30], optoelectronic devices [31], gas storage [32], separation [33,34,35,36,37,38], desalination [39,40,41,42,43], and energy storage [28,44,45]. The main advantages of using COFs for photocatalytic hydrogen evolution reactions are (i) the bandgap of COFs can be tuned by selecting the suitable monomer for visible light absorption [46], (ii) the large surface area provides excellent catalytic sites, and the crystalline and porous structure facilitates fast charge mobility and omits the possibility of charge trapping caused by defects in the materials [47], (iii) the long-range crystalline nature enhances the photogenerated electron-hole separation and decreases the possibility of charge hole recombination and avoids unproductive quenching events [48], (iv) the extended covalently bonded network provided the excellent chemical and thermal stability [44]. COF was reported in 2014 by Lotsch for the hydrogen evolution reaction (HER) [49]. Since then, the number of COF-based photocatalysts has exponentially increased (Figure 1). Several existing reviews cover COFs as photocatalytic materials for hydrogen evolution. Nevertheless, none cover the structural design and mechanical aspects of COFs as photocatalysts for hydrogen evolution. In lieu of simply synthesizing COFs from routine methods and pre-existing starting materials, it is necessary to explore additional key aspects, such as potential new synthetic monomers with the ability of excellent charge separation and migration, by understanding the mechanism of COFs photocatalytic action. Keeping in mind the porous structure and extended network, there is a need to know how the co-catalyst can be embedded in the system and how other heteroatoms impact the charge separation and hydrogen evolution of the photocatalyst. In this review, we have discussed for the first time the current research in the direction of design and mechanistic aspects. In this review, we document the progress in the photocatalytic evolution of hydrogen from water using the COFs in terms of mechanistic aspects, challenges, and future directions.

2. Concept of Photocatalytic Hydrogen Evolution

The photocatalytic splitting of water consists of two half-reactions: (i) hydrogen evolution reaction (HER) and (ii) oxygen evolution reaction (OER). When the photocatalyst absorbs the light photon, an excited-state electron (e⁻) and hole (h⁺) pair is formed and migrates to the catalyst’s surface after separation. The e⁻ reduces the proton to generate hydrogen, and h⁺ oxidizes the water molecule to oxygen. The overall water splitting reaction is thermodynamically unfavored (ΔG = 237 kJ mol⁻¹) [50], so the energy needed (E⁰ = −ΔG/nF) for this reaction is 1.23 eV. The reduction potential for proton (H⁺/H₂) is 0 (pH = 0) (V vs. NHE (normal hydrogen electrode)), and the oxidation potential for water (O₂/H₂O) is 1.23 (pH = 0) (V vs. NHE) [51]. Consequently, overall water splitting at pH = 7 entails −0.41 V for HER and 0.82 V for OER. Thus, for overall water splitting, photocatalysts should have the minimum 1.23 eV bandgap under ideal conditions such as zero overpotential and zero reorganization energy for interfacial charge transfer (Figure 2). The overall water splitting always remains challenging because of unfavorable thermodynamics, slow kinetics, dissolved oxygen, and an undesired backward reaction, and it has been considered the Holy Grail [52]. However, the half-HER reaction is more practicable if a sacrificial reagent is added. The most commonly used sacrificial reagents are triethanolamine (TEOA), ascorbic acid (AA), triethylamine (TEA), lactic acid (LA), and sodium ascorbate (NaAA) (

Table 1. The comparative HER responses to structurally different COFs.

COFs	Band Gap (eV)	Pore Size (nM)	Metal Co-Catalyst	Sacrificial Element	HER Efficiency
Extending π-conjugation
TFPT-COF49	2.8	3.8	Pt	TEOA	1970 μmol h−1 g−1
Nx-COFs56	2.6–2.7	3.5	Pt	TEOA	1703 μmol h−1 g−1
PTP-COF57	2.1	-	Pt	TEOA	83.83 μmol h−1 g−1
TP-EDDA58 TP-BDDA	2.34 2.31	- -	Pt	TEOA	324 ± 10 µmol h−1 g−1 30 ± 5 μmol h−1 g−1
Incorporating heteroatoms
A(B/N/Y)PY-COF59	1.94–1.92	2–2.22	Pt	TEOA	98 µmol h−1 g−1
S-COF, FS-COF and TP-COF60	-	2.90–2.76	Pt	AA	10,100 μmol g−1 h−1 16,300 μmol g−1 h−1 (WS5F) 15,800 μmol h−1 m−2 (a thin film of FS-COF)
sp2c-COF sp2c-COFERDN61	1.90 1.85	1.9	Pt	TEOA	2120 μmol g−1 h−1 1360 μmol g−1 h−1
ODA-COF62	2.00	-	Pt	TEOA	2615 μmol g−1 h−1
TtaTfa/TpaTfa/TtaTpa-COF63	2.73–2.60 (2.22–1.89 H+)	1.74			20.7 ± 2.7 mmol g−1 h−1
v-COF-NS165	1.85	1.7/1.9	Pt	AA	4.4 mmol h−1 g−1
COF-BBT66	2.0	2.3	Pt	AA	48.7 mmol g−1 h−1
Incorporating metal complexes/metals
azine-linked N2-COF68	-	-	molecular cobaloxime	TEOA	782 μmol g−1 h−1
TpDTz COF69	2.07	3.4	NiME cluster	TEOA	941 μmol h−1 g−1
pCOF1070	-	2.3	cobaloxime catalyst	TEOA	163 μmol h−1 g−1
TTR-COF71	2.71	2.95	Au	TEOA	1720 μmol h−1 g−1
NUS-55(Co)72	-	2.9	[Co(bpy)3]Cl2	TEA	2480 µmol g−1 h−1
Cu-salphen-HDCOF-NSs73	1.62	1.5	Cu	TEA	36.99 mmol g−1 h−1
MoS2/TpPa-1-COF74	2.14	-	MoS2	AA	55.85 μmol h−1
CdS-COF75	2.52	-	Pt	LA	3678 μmol h−1 g−1
PY-DHBD-COF76	2.28		Pt	AA	71,160 μmol g−1 h−1
TpBpy-Ni 2%77	1.84	-	Pt	AA	51,300 μmol h−1 g−1
Donor–Acceptor (D–A) configuration
PyTz-COF78	2.20	≈3.2	Pt	AA	2072.4 μmol g−1 h−1
BT-TAPT-COF79	2.35	2.3	Pt	TEOA	949 μmol g−1 h−1
Py-ClTP-BT-COF80	2.36	3.25	Pt	AA	177.50 μmol h−1 g−1
NKCOF-10881	1.8	3.5	Pt	AA	120 μmol h−1
BtCOF15082	2.5	-	Pt	TEOA	750 ± 25 µmol h−1 g−1
BTH-3 COF83	2.02	-	Pt	AA	15.1 mmol h−1 g−1
BTTh-TZ-COF84	2.32	1.35	Pt	AA	5.22 mmol h−1 g−1
PTPA-COF TP-COF85	2.31 2.41	- -	Pt	TEOA	36 μmol h−1 g−1 29.12 mmol h−1 g−1
TM-DMA-COF86	2.07	1.98	Pt	AA	4300 mmol h−1 g−1
TAPFy-PhI COF87	2.21	2.26	Pt	AA	1763 μmol g−1 h−1
TpPa−COF−(CH3)288	2.06	0.48	Pt	NaAA	8.33 mmol g−1 h−1
PyPz-COF89	2.05	2.8	Pt	AA	7542 μmol g−1 h−1
BTT-PDA, BTT-NDA, BTT-AnthDA, and BTT-BPhDA92	2.15 2.16 2.00 2.21	2.48 2.83 3.23 3.23	Pt	AA	2.04 mmol h−1 g−1 5.22 mmol h−1 g−1 4.23 mmol h−1 g−1 3.27 mmol h−1 g−1
COF-JUL35 COF-JUL3693	1.85 2.12	2.7	Pt	A	70.8 ± 1.9 mmol h−1 g−1
BTT-Bpy-COF94	1.41	2.85	Pt	AA	15.8 mmol h−1 g−1
Hetero-structural mixed linkage approach/Miscellaneous approach
NTU-BDA-THAT NTU-BDA-THAT + NH2-Ti3C2Tx95	1.81	1.42 2.09	Pt	AA	1.47 μmol g−1 h−1 14,228.1 μmol g−1 h−1
2D-2D BP/TpPa-1-COF96	-	-	-	-	456.7 µmol h−1 g−1
TSCOFW98	1.95	2.6	Pt	AA	593 mmol h−1 g−1
2D/2D COF/g-C3N499	-	1.7	Pt	TEOA	449.64 μmol h−1
TpPa-1-COF/ZnIn2S4100	-	-	ZnIn2S4	Na2S + Na2SO3	853 μmol g−1 h−1
Tp-nC/BPy2+-COF101	-	-	Pt	AA	34,600 μmol h−1 g−1
TPCBP X-COF X = Butyl103	2.36	2.95–3.71	Pt	TEOA
g-C18N3-COF105	2.42	1.72	Pt	AA	14.6 mol h−1
g-C40N3-COF106	2.36	3.3	Pt	TEOA	206 µmol h−1
g-C54N6-COF107	2.03	2.28	Pt	TEOA	2518.9 μmol h−1 g−1

The high-performance COFs are underlined.

). The role of sacrificial agents is to capture the photogenerated h⁺ by electron donation and promote HER. In most cases, co-catalysts have been used for HER as a Schottky junction, which facilitates the separation between the photogenerated electron (e⁻) and hole (h⁺) pair.

3. Requirements of Photocatalytic Hydrogen Evolution

The main requirements for photocatalytic hydrogen evolution are as follows: (i) the removal of dissolved oxygen in the water because photogenerated electron (e⁻) can reduce the oxygen to superoxide radicals (O2⁻), which can lead to numerous side reactions; (ii) the need for the co-catalyst because, in the absence of the co-catalyst, the photocatalyst showed significantly less efficiency or no efficiency at all; the role of the co-catalyst is to effectively extract all the photogenerated electrons (e⁻) and to provide the active site in addition to reducing the overpotential of the reaction; and (iii) the need for sacrificial agents that react with the photogenerated hole (h⁺) by oxidation half-reactions and ensure oxygen-free conditions.

4. Challenges in Photocatalytic Hydrogen Evolution

Several challenges are associated with the photocatalytic hydrogen evolution reaction caused by splitting the water molecules.

Photocatalyst Efficiency: Many existing photocatalysts suffer from limited absorption of visible light, which accounts for only a small fraction of the solar spectrum. Catalyst Stability: Photocatalysts often experience degradation over time due to photo-corrosion or chemical reactions with reactive intermediates. The stability of photocatalysts is a significant concern as it affects their long-term performance and lifespan. The main target of the current research is to develop photocatalysts with a lifespan of at least 10 years [5]. Charge Carrier Separation and Transfer: The photogenerated electron (e⁻) and hole (h⁺) pair should be effectively separated, and the charge carriers need to be efficiently transferred to the catalytic sites for water splitting. Kinetic Limitations: The reaction kinetics in photocatalytic hydrogen evolution can be slow, limiting the overall hydrogen production rate. The sluggish kinetics can be attributed to slow water oxidation, inefficient proton reduction, and surface charge transfer limitations [53]. Solar to Hydrogen Efficiency (SHE): Solar-to-hydrogen efficiency = (Energy content of produced hydrogen/Energy content of incident solar radiation) × 100%. The energy content of the produced hydrogen is typically determined by measuring the amount of hydrogen gas produced and multiplying it by the lower heating value (LHV) or higher heating value (HHV) of hydrogen. With a few exceptions, researchers are still trying to achieve the target of at least 10% SHE. Co-catalyst Integration: Integrating suitable co-catalysts with the photocatalyst while maintaining efficient charge transfer interfaces is challenging. The development of an effectively coupled photocatalyst and co-catalyst is an active research area. Scalability and Cost: To commercialize the HER, it is essential to develop cost-effective and scalable materials and manufacturing processes. The synthesis and fabrication of efficient and stable photocatalysts using abundant and low-cost materials are crucial challenges. Reactor Design and Engineering: The design of photocatalytic reactors is essential for optimizing light absorption, reactant diffusion, and product separation. Developing efficient reactor configurations and engineering scalable systems that ensure optimal light utilization and mass transfer is a challenge in photocatalytic hydrogen evolution.

The ongoing research aims to overcome these obstacles and improve photocatalytic hydrogen evolution’s efficiency, stability, and practicality for sustainable hydrogen production. The efforts in this direction by utilizing the COFs are exciting and encouraging.

5. Covalent Organic Frameworks (COFs) and Photocatalytic Hydrogen Evolution

Covalent Organic Frameworks (COFs) have gained significant attention in the field of photocatalytic hydrogen evolution due to their unique properties and potential for light harvesting and catalysis [54]. COFs can be pre-designed with specific electronic structures and bandgaps, allowing them to absorb light across a broad range of the solar spectrum. By tuning the chemical composition and connectivity of the organic building blocks, the absorption properties of COFs can be tailored to match the desirable visible solar spectrum for optimal light utilization. The COFs offer opportunities for band engineering by modifying the organic building blocks or introducing guest molecules. This enables tuning of the electronic structure, band positions, and energy levels, which can influence the charge transfer processes and overall photocatalytic performance. COFs offer the advantage of having ordered, extended structures with spatially separated donor and acceptor sites, facilitating the efficient charge separation and transport of photogenerated electron (e⁻) to catalytic sites for hydrogen evolution. COFs can be functionalized with various catalytic sites, such as metal nanoparticles or co-catalysts, to enhance hydrogen evolution efficiency [3]. Incorporating catalytic species within the COF structure provides active sites for the photocatalytic reduction of protons to hydrogen. These catalytic sites can improve the overall reaction kinetics and hydrogen production rates. The stability of COFs is a crucial factor for practical photocatalytic applications. COFs exhibit excellent thermal and chemical stability, making them suitable for long-term photocatalytic reactions [28]. The covalent nature of their framework ensures robustness and resistance to degradation under reaction conditions.

6. Design Principles of COFs for Photocatalytic HER

Several approaches are explored to construct the most efficient COFs for the HER. The main challenge for efficient migration of charge carriers was overcome by building the extended π-conjugated network. To engineer the band gap, several heteroatoms and heterojunctions were successfully explored in the COFs, incorporating the co-catalyst in the porous structure by coordinating them with the electron-rich part of the framework and by a covalent bond. The donor–acceptor approach [55] is also explored to get the maximum efficiency for HER. Several new approaches are also discussed in the miscellaneous section of this review.

6.1. Extending π-Conjugation

The π-conjugation of COFs can affect their light absorption ability, band gap, charge transfer, and conductivity. Generally, extending the π-conjugation of COFs can enhance their visible-light absorption, narrow their band gap, increase their charge separation efficiency, and improve their conductivity.

The era of exploration of COFs in photocatalytic hydrogen evolution began in 2014 when Bettina V. Lotsch reported the hydrazone-linked TFPT–COF (pore size 3.8 nm) having 1,3,5-tris-(4-formyl-phenyl)triazine (TFPT) and 2,5-diethoxy-terephthalohydrazide (DETH) building blocks and Pt as co-catalyst (Figure 3) [49]. The TFPT-COF showed H₂ production of 230 μmol h⁻¹ g⁻¹ when sodium ascorbate was used as the sacrificial agent. In the presence of 10 vol% aqueous triethanolamine (TEOA) solution as a sacrificial donor, H₂ evolved as 1970 μmol h⁻¹ g⁻¹, which was higher than the Pt-modified amorphous melon, g-C₃N₄ [13]. The designing part of the COFs was dependent on the fact that triazine-based molecules are known to have high electron mobilities and good electron-withdrawing character, and the dihedral angle between the phenyl and triazine unit (~7.7°) is smaller than the benzene-centered analog (38.3°), which ultimately facilitated the fabrication of planar COF with an extended π-system and enhanced crystallinity with a band gap of 2.8 eV, which is adequate for HER. Later in 2015, the same research group explored more structural requirements to get better efficiency from the COFs, named Nx-COFs-based photocatalysts [56]. In this report, they concluded that replacing alternate carbon atoms in the central aryl ring of the triphenylaryl platform with nitrogen atoms led to unique alterations in the electronic and steric characteristics of the central ring of Nx-COFs. A series of four azine-linked Nx-COFs was fabricated by gradually replacing the aromatic C-H units with N atoms in the central aryl ring, which led to the reduction in the dihedral angle of the triaryl building block and increased the planarity of COF layers with a four-fold increment in hydrogen evolution with each substitution. The hydrogen evolution was found to be 23, 90, 438, and 1703 μmol h⁻¹ g⁻¹ for N₀, N₁, N₂, and N₃–COF, respectively (Figure 4). The synthesized Nx-COFs absorbed ultraviolet and blue parts of the visible region. They were found to have an optical band gap of ≈2.6–2.7 eV, which was adequate for photocatalytic water splitting. The authors argued that the central aromatic ring becomes more electron deficient with the replacement of each nitrogen atom, which is effective at stabilizing the negative charge generated on the COF and transferring it to the platinum co-catalyst and also increasing the stability of the radical anion, which proved to be HER intermediate theoretically after reductive quenching of Nx-COFs.

In 2017, the same group explored more structural requirements for the COFs as photocatalysts for HER by fabricating COFs with different distributions of nitrogen across the building blocks [57] (Figure 5). The substitution was made on the peripheral rather than the central aryle ring with N atoms.

The synthesized PTP-COFs showed lower crystallinity, porosity, and a different morphology than Nx-COFs, directly concurrent with the hydrogen evolution efficiency (83.83 μmol h⁻¹ g⁻¹). The PTP-COF showed absorption in the visible region with a band gap of 2.1 eV, which was lower than that of Nx-COFs. The poor photocatalytic activity of PTP-COF was found to be associated with the ineffective stabilization of negative charge, as concluded by comparing the vertical radical stabilization energies (RASE) calculated by the PBE0-D3/def2-TZVP level of theory.

In 2017, Arne Thomas and co-workers reported the highly porous and chemically stable acetylene (−C≡C−) and diacetylene (−C≡C−C≡C−) functionalized β−ketoenamine COFs [58]. The two COFs TP-EDDA and TP-BDDA were synthesized by using 1,3,5-triformylphloroglucinol and 4,4′-(ethyne-1,2-diyl)dianiline (EDDA) and 4,4′-(buta-1,3-diyne-1,4-diyl)dianiline (BDDA) as monomers, respectively (Figure 6). The TP-BDDA exhibited hydrogen evolution over 10 h with an average rate of 324 ± 10 µmol h⁻¹ g⁻¹, higher than the TP-EDDA. The band gaps calculated were 2.34 and 2.31 eV for TP-EDDA and TP-BDDA, respectively.

6.2. Incorporating Heteroatoms or Metal Complexes

Incorporating heteroatoms or metal complexes into the COF framework can introduce new energy levels or active sites that can improve their photocatalytic performance.

6.2.1. Incorporating Heteroatoms

In 2018, Bettina V. Lotsch extended their investigation into the structural features and fabricated the A(B/N/Y)PY-COF by using the three tetra-alkyne 1,3,6,8-tetraethinylpyrene building blocks–1,3,6,8-tetrakis(4-ethynylbenzaldehyde)-pyrene (TEBPY), 1,3,6,8-tetrakis(6-ethynylnicotinaldehyde)-pyrene (TENPY), and 1,3,6,8-tetrakis(2-ethynylpyrimidin-5-carbaldehyde)pyrene (TEPPY), which reacted with hydrazine hydrates to yield an azine connected COF (Figure 7) [59]. The diffuse reflectance UV–vis spectra estimated an optical bandgap of 1.94 eV for A-TEBPY-COF, 1.92 eV for A-TENPY-COF, and 1.91 eV for A-TEPPY-COF at an absorption around 600 nm. A new trend was observed by increased nitrogen content as A-TEPPY-COF produced only 6 µmol h⁻¹ g⁻¹ hydrogen, while A-TEBPY- produces hydrogen at the rate of 98 µmol h⁻¹ g⁻¹, A-TENPY-COF with 22 µmol h⁻¹ g⁻¹, concluding that increased nitrogen content can also hamper the HER. The authors concluded that the VB→VB_ex, CB_ex (HOMO→HOMO_ex, LUMO_ex) transition was crucial for photocatalysis, as excited electrons in the CB of A-TEBPY-COF have an increased thermodynamic driving force for HER compared to A-TENPY-COF and A-TEPPY-COF (0.8 eV against 0.5 and 0.4 eV, respectively). A radical cation quenching pathway was predicted for A(B/N/Y)PY-COF because of their electronic-rich nature, which was also proved by the theoretical calculations. The calculations showed that for A-TEBPY-COF the radical cation species is favored by 0.2 eV versus A-TENPY-COF and 0.4 eV versus A-TEPPY-COF, respectively. The HER efficiencies showed a similar pattern in the relative stability of the radical cation: the lesser N content resulted in improved radical cation stability. The formation of excimer was also revealed in the study, as the all-planar linkers of the A(B/N/Y)PY-COF were found to be stacked by a distance of about 3.43 Å and showed an excimer emission band around 670–700 nm.

In 2018, Andrew I. Cooper fabricated three COFs—S-COF, FS-COF, and TP-COF—via a Schiff-base condensation reaction of 1,3,5-triformylphloroglucinol with aromatic diamines 7-diaminodibenzo[b,d]thiophene sulfone (S), 3,9-diamino-benzo[1,2-b:4,5-b′]bis [1]benzothiophene sulfone (FS) and 4,4″-diamino-p-terphenyl (TP) (Figure 8) [60]. The sulphone groups decorated with thiophene linkers increased the overall hydrophilicity of the COFs, which renders them efficient for HER. In the presence of ascorbic acid as a sacrificial electron donor and Pt as a co-catalyst, FS-COF exhibited the highest reported hydrogen evolution at a rate of 10,100 μmol g⁻¹ h⁻¹ under visible light. To show the effect of crystallinity on the HER reaction, the authors also fabricated the amorphous analog of FS-COF, which exhibited a nine times lower (1012 μmol g⁻¹ h⁻¹) activity than FS-COF. The mesoporosity of the FS-COF was further utilized by dye sensitization. WS5F, a less water-soluble dye, showed visible-light HER of 16,300 μmol g⁻¹ h⁻¹, which was attributed to the absorption of significant high photons at higher wavelengths (λ > 420 nm). When a thin film of FS-COF was platinized by drop-casting, the hydrogen evolution increased drastically (15,800 μmol h⁻¹ m⁻²) under solar radiation. The DFT calculations revealed that FS-COF has the smallest optical gap among the four COF materials and the most significant rates of visible photon absorption and photogenerated electron (e⁻) and hole (h⁺) pair. The theoretical calculations also showed favorably located ionization potential (IP)/exciton electron affinity (EA*)/Kohn–Sham valence band maximum (VBM) and electron affinity (EA)/exciton ionization potential (IP*)/conduction band minimum (CBM) levels, to endure a driving force for both redox half-reactions.

In 2019, Donglin Jiang et al. reported the sp² carbon-conjugated COF with C=C linkages [61]. The advantage of this COF was that it was not linked by labile linkages like imine bonds or boron-based linkages, which are sensitive to acidic or basic conditions and sometimes significantly less stable in an aqueous medium (Figure 9). The authors also explained that the synthesized olefinic COFs have fully π-conjugated frameworks that facilitate effective exciton migration over the entire framework, while the COFs having imine-, hydrazone-, or azine-linkages are known to have short range π conjugation and have inadequate competence of exciton migration. The designed sp²c-COF lacks the lone pair available in the imine-, hydrazone-, or azine-linked COFs, which dissipate the excitation energy by involving lone pairs as an integrated electron-transfer pathway to quench the photoexcited states. The sp²c-COF was synthesized by condensation of TFPPy and PDAN, while the three-component polycondensation of TFPPy, PDAN, and ERDN synthesized sp2c-COF_ERDN (heterojunction donor-acceptor). The sp²c-COF_ERDN showed better Hydrogen evolution efficiency (2120 μmol g⁻¹ h⁻¹) than sp²c-COF (1360 μmol g⁻¹ h⁻¹) at ≥420 nm and ≥498 nm radiation, respectively. The Hydrogen evolution efficiency of sp²c-COF decreased when irradiated with longer wavelengths. The authors also compared the efficiency of the fabricated COFs with the imine-linked analogies and found no H₂ evolution, irrespective of the light wavelengths and exposure time. The band gap for sp²c-COF was 1.90 eV, while the band gap for sp²c-COF_ERDN was 1.85 eV. The superiority of sp²c-COF_ERDN was found to be related to the higher HOMO level (−5.66 eV) as compared to sp²c-COF (−5.74 eV), which induced the narrow band gap, which is favorable for exciting electrons to the LUMO.

In 2021, Ruihu Wang et al. reported the oxadiazole-linked COF, which was fabricated by the post-oxidative cyclization of N-acylhydrazone linkage for photocatalytic hydrogen evolution (Figure 10) [62]. Compared to their imine-linked counterparts, N-acylhydrazone-linked COFs are more stable, but there has been a loss in crystallinity and an apparent decline in catalytic activity. The H₂ evolution rate for ODA-COF was 2615 μmol g⁻¹ h⁻¹ with Pt as the co-catalyst and triethanolamine (TEOA) as the electron donor under visible light irradiation (300 W Xe lamp, λ ≥ 420 nm). The optical band gap of ODA-COF was 2.00 eV, much smaller than that of 2.80 eV for H-COF (without oxadiazole linkage), which revealed the increment of the intrinsic carrier concentration in ODA-COF. A paramagnetic absorption signal with a g value of 2.0055 is visible in the electron paramagnetic resonance (EPR) spectroscopy of ODA-COF. The signal becomes stronger upon illumination with visible light, while no apparent EPR signal is observed for H-COF while exposed to light. Overall, these observations concluded the importance and efficiency of oxadiazole-in the COF for photocatalytic hydrogen generation.

Arne Thomas et al. reported the effect of protonation on the imine bond of COFs for HER [63]. Three COFs were synthesized with 2,4,6-Tris(4-aminophenyl)triazine (Tta), Tris(4-formylphenyl)amine (Tfa), 1,3,5-Tris(4-formylphenyl)benzene (Tpa-CHO), and 1,3,5-Tris(4-aminophenyl)benzene (Tpa-NH2) as monomers and named as TtaTfa, TpaTfa, and TtaTpa based on the used building blocks (Figure 11). The protonated TtaTfa with the most robust D-A pair in the backbone exhibited the highest HER rate of 20.7 ± 2.7 mmol g⁻¹ h⁻¹ under visible light irradiation. The band gap of the synthesized COFs decreased significantly after protonation with ascorbic acid; in the case of TtaTfa, the band gap reduced to 1.90 eV from 2.52 eV. The EPR analysis revealed that protonated TtaTfa demonstrated a much higher signal intensity, which suggests a vastly improved charge separation efficiency. This study showed that the protonation of the imine bond not only enhances better light absorption and charge separation but also increases the hydrophilicity of the photocatalyst.

Thiazole is a well-known electron-deficient scaffold and is well exploited in the synthesis of linear conjugated polymers [64]. There are some recent reports in which authors used this core for the synthesis of COF for photocatalytic HER. Tao Zhang et al. reported the benzobisthiazole-vinylene-linked COF synthesized by solid-state condensation between 2,6-dimethyl[1,3]thiazolo[5,4-f][1,3]benzothiazole (DTBT) and benzene-1,3,5-tricarboxaldehyde (BTCA) for v-COF-NS1 (Figure 12) [65]. The pore size and bandgap for synthesized COF were 1.7 nm and 2.27 eV respectively. In the presence of photodeposited Pt nanoparticles the HER reported was 4.4 mmol h⁻¹ g⁻¹. Yanguang Li et al. also utilized the benzobisthiazole core for the synthesis of COF-BBT, but in this report, authors used the 2.6-diaminebenzobisthiazole (DABBT) instead of (DTBT) which was condensed with 1,3,5-triformylphloroglucinol (TFP) to get the COF (Figure 12) [66]. The COF-BBT was found to have good HER at the rate of 48.7 mmol g⁻¹ h⁻¹. The authors claimed the higher hydrophilicity of the synthesized COF by measuring the contact angle, which was ultimately responsible for the higher HER.

6.2.2. Incorporating Metal Complexes/Metals

The metal incorporation in the pores of COF can eliminate or reduce the need for expensive metal for use as a co-catalyst, and it can also provide the insight for mechanistic insight of HER and the fabrication of unistructural COF-based photocatalyst. Metals incorporated into COFs can facilitate efficient charge separation and transfer processes, prevent their recombination with the holes, and enhance the overall charge separation. By incorporating metals with appropriate energy levels, the COF can capture a broader spectrum of light, including visible and near-infrared regions, and by incorporating specific metals, the COF photocatalyst can be designed to favor specific reaction pathways, such as the suppression of side reactions or the promotion of specific HER. Metals can provide structural reinforcement to the COF framework, preventing its degradation during prolonged exposure to reactive species generated during the photocatalytic process. Herein, we are reporting the seminal work done by the researcher in this regard.

In 2019, Bettina V. Lotsch et al. reported the first single-site photocatalytic H₂ evolution from COFs using molecular cobaloxime co-catalysts [67]. In the study, the authors used N₂-COF instead of the most effective N₃-COF because of the easy synthetic route. The cobalt complexes with dimethylglyoxime ligands, known as cobaloximes, have low overpotentials for H₂ evolution and oxygen tolerance (Figure 13) [68]. The optimum hydrogen evolution rate of 782 μmol g⁻¹ h⁻¹ was achieved when HER was performed in 4:1 ACN/H₂O solvent and 60 equivalent of dimethylglyoxime (dmgH₂). The high equivalence of dmgH₂ was justified based on the labile nature of dmgH₂ as a ligand_, as it is believed to undergo exchange with free dimethylglyoxime in solution. To overcome the high use of dmgH₂, authors tried the more stable BF₂-annulated complex of cobaloxime; nevertheless, significantly less production of H₂ was achieved due to the difficulty of cobaloxime to undergo protonation at the oxime functionality as they are covalently linked to boron. To evaluate the bonding behavior of cobaloxime in the COF, the authors used the ¹³C cross-polarization magic angle spinning (CPMAS) NMR spectra and no peaks corresponding to Co were observed, which was further confirmed by the ¹H and IR spectroscopy. The mechanism of action of cobaloxime was also proposed, which involved the step-by-step reduction of the Co^III complex to Co^II, then to Co^I, which is then protonated to form a Co^III hydride intermediate. The authors used the EPR analysis to prove the formation of Co^II but were unsuccessful in confirming the presence of Co^I complex. The H₂ evaluation was linearly dependent on the cobaloxime, which established the single cobalt mechanism for hydrogen generation and denied the possibility of a bimetallic pathway. Intriguingly, utilizing metallic Pt as a co-catalyst results in an H₂ evolution rate three times lower than cobaloxime. This is likely because Pt nanoparticles are poorly distributed and/or photodeposited on the COF surface.

In 2019, they used the same single-site photocatalyst approach for HER using thiazolo[5,4-d]thiazole-bridged COF and nickel-thiolate cluster in water [69]. The thiazolo[5,4-d]thiazole (TzTz) is known to have n-type characteristics, including high oxidative stability and a rigid planar structure, which facilitate the competent intermolecular π–π overlap and can provide high photogenerated electron (e⁻) and hole (h⁺) pair mobility. The TpDTz COF was fabricated by condensing 4,4′-(thiazolo[5,4-d]thiazole-2,5-diyl)dianiline and 1,3,5-triformylphloroglucinol (Figure 14). The authors reported the H₂ evolution at 941 μmol h⁻¹ g⁻¹ more than 70 h in a single run at pH 8.5 using triethanolamine as the sacrificial electron donor. The band gap of the TpDTz COF was found ~2.07 eV, with a significant decrease in the band gap (~2.28 eV) compared to the analog synthesized without the TzTz core. The authors also explored the nature of ligands by using thiourea (TU) and 2-mercaptophenol (MP) as replacement ligands for 2-mercaptoethanol. Still, no hydrogen evolution was observed because of the unfavorable complexation of TU and MP with Ni. According to microkinetic modeling, the rate-limiting phase in the reaction was the outer-sphere electron transfer from the photoexcited COF to the Ni co-catalyst, which takes part in the HER (Figure 14b). In addition, to the new catalytic system, the authors also introduced a new continuous-flow system for direct detection of the H₂ production rate.

In 2020, a different approach for incorporating the co-catalyst in the COFs was employed [70]. The authors reported the azide-functionalized chloro(pyridine)cobaloxime co-catalysts immobilized on a hydrazone-based COF-42 backbone fabricated with the propargyl-containing 2,5-bis(prop-2-yn-1-yloxy)terephthalohydrazide (DPTH) monomer. The covalent bond between COF and co-catalyst was believed to prevent degradation and increase the lifespan of the photocatalytic system (Figure 15). The co-catalyst was introduced in the COF system by post-synthetic modification utilizing Click chemistry, the formation was confirmed by the 2D ¹H–¹H DQ–SQ NMR experiment. Although the designed system was not so efficient, as the H₂ evolution rate was 163 μmol h⁻¹ g⁻¹, this rate was higher than the physically absorbed catalytic system. The longevity of the hybrid system was tested up to 780 h, and it was found that the hydrogen evolution rate was 59 μmol h⁻¹ g⁻¹, while the physisorbed sample reached 35 μmol h⁻¹ g⁻¹ of hydrogen evolution. The authors also explained the reason for the longevity by considering the enhanced re-coordination of the co-catalyst in the COF pores and the additional spatial proximity of the co-catalyst, which also facilitated the charge transfer.

In 2019, Jiaguo Yu and Yan Yu et al. reported the thioether-functionalized covalent organic framework (TTR-COF) for the photocatalytic splitting of seawater for hydrogen evolution. The authors utilized the affinity of sulfur toward gold to immobilize the gold on the COF as a co-catalyst (Figure 16) [71]. The TTR-COF was fabricated by the condensation of 1,3,5-tris(4-formylphenyl)triazine (TFPT) and 2,5-bis(2-(ethylthio)ethoxy)terephthalohydrazide (BETH). The H₂ evolution efficiency for TTR-COF was up to 1720 μmol g⁻¹ in a 4 h reaction and found to be stable for at least 20 h under visible light (λ ≥ 420 nm) and triethanolamine (TEOA) as the sacrificial hole scavenger. The TTR-COF showed maximum absorption at 350–550 nm, which was attributed to the π–π* transitions of the conjugated ring system with a band gap of 2.71 eV. For seawater, the H₂ evolution rate was 141 μmol h⁻¹ g⁻¹ over 20 h light exposure. The binding behavior of Au with respect to other ions was evaluated by fluorescence spectroscopy, and significant quenching was observed by gold due to charge transfer from TTR-COF to Au which was crucial for HER. The H₂ evolution was reduced to 589 μmol g⁻¹ in a 4 h reaction with an increase in Mg²⁺ ions to 250 ppm, proving that Mg²⁺ ions have a detrimental effect on the photocatalytic activity. The bandgap of Mg²⁺ TTR-COF increases compared to TTR-COF, while its band dispersion noticeably shrinks, indicating that the electron transitions from valence bands to conduction bands are particularly challenging, which was in agreement with the observed results.

In 2019, Dan Zhao and co-workers reported the Co-incorporated NUS-55, which was fabricated by the condensation of [2,2′-bipyridine]-5,5′-dicarbohydrazide (BPDC) with 4,4″-diamino-p-terphenyl (TP) [72]. The coordination site present in NUS-55 was treated with Co(NO₃)₂.6H₂O) and [Co(bpy)₃]Cl₂ to make the composite type catalyst NUS-55(Co) (Figure 17). The NUS-55 with impregnated [Co(bpy)₃]Cl₂ showed the visible-light-driven H₂ evolution rate to 2480 µmol g⁻¹ h⁻¹ at irradiation with 450 nm light source. The higher efficiency of the NUS-55 with Co was due to the supplementary H₂ bonding interaction between the BPDC unit and the Co, which ultimately facilitated the photogenerated electron (e⁻) and hole (h⁺) pair separation and hindered the recombination.

Thomas C. W. Mak et al. reported the morphological alteration approach for COF by exfoliating into ultrathin nanosheets (Cu-salphen-HDCOF-NSs) (Figure 18) [73]. The HD-COF was synthesized by the polycondensation of 2,3,6,7,10,11-hexaiminotriphenylene (HATP.6HCl) and 2,6-diformylphenol (DFP). After the cyclization, the HD-COF was treated with Cu(OAc)₂ to get the binuclear Cu-salphen-HDCOF. The Cu-salphen-HDCOF-NSs were obtained by exfoliation of bulk COF materials through facile solvent-assisted liquid sonication. The Cu-salphen-HDCOF-NSs exhibited H₂ production at 36.99 mmol g⁻¹ h⁻¹ in the first 7 h, and the total amount reached 364.89 mmol g⁻¹ over 13 h under visible light, which was lower than the Cu-salphen-HDCOF (27.43 mmol g⁻¹ h⁻¹). The band gap of Cu-salphen-HDCOF-NSs was estimated at 1.62 eV. This study is an excellent example of the structural engineering of the COFs to get high photocatalytic HER.

Feng-Ming Zhang et al. reported the noble metal-free efficient COF-based photocatalytic H₂ evolution system [74]. By in situ growing the TpPa-1-COF in an exfoliated MoS₂ dispersion solution of DMF, the authors created MoS₂/TpPa-1-COF composites using a ketoenamine-based TpPa-1-COF as a model (Figure 19). The MoS₂–3%/TpPa-1-COF composite has a 32-fold higher H₂ evolution rate (55.85 μmol h⁻¹) than pure TpPa-1-COF (1.72 mol h⁻¹). The result of surface photovoltage spectroscopy (SPS) of MoS₂-3%/TpPa-1-COF supported the better separation of the photogenerated electron hole.

Rahul Banerjee et al. reported the doping method for photocatalytic splitting of water [75]. The CdS nanoparticles were placed on the COF matrix; the COF was synthesized by the 1,3,5-triformylphloroglucinol with 2,5-dimethyl-p-phenylenediamine, which was treated with the Cd(OAc)₂.2H₂O under hydrothermal conditions (Figure 20). The CdS-COF (90:10) was found to be active for H₂ evolution at rate of 3678 μmol h⁻¹ g⁻¹, with Pt as co-catalyst by irradiation of λ ≥ 420 nm.

Weiqiao Deng et al. used the unique approach of in situ photo-deposition of platinum clusters on a COF for photocatalytic hydrogen production [76]. The reported COF was fabricated by condensation of 1,4-dihydroxybenzidine (DHBD) and 1,3,6,8-tetra(4-formylphenyl)pyrene (PY-CHO) monomers (Figure 21). After getting the COF, the Pt co-catalyst was photo-deposited in situ by reducing the H₂PtCl₆ in a light reaction. In the presence of a comparatively low quantity (0.5 wt% Pt), the photocatalytic hydrogen evolution rate reached 16,980 μmol g⁻¹ h⁻¹, which was enhanced to 71,160 μmol g⁻¹ h⁻¹ with 3 wt% Pt under visible light irradiation. The uniform distribution of Pt over the COF was further confirmed by the HR-TEM and HAADF-STEM. The authors also proposed that Pt atoms were first adsorbed on the hydroxyl and imine sites on PY-DHBD-COF and further reduced into clusters and metallic particles with increasing concentration through photodeposition. To increase the efficiency of photocatalysis, this work offers a method for creating a microenvironment for co-catalyst deposition on a photocatalyst.

Jin Guo et al. immobilized the nickel ion on the COF for photocatalytic HER. The TpBpy-COF was synthesized solvothermally by the condensation of 1,3,5-triformylphoroglucinol (Tp) and 2,2′-bipyridine-5,5′-diamine [77]. The authors prepared the 2 wt% Ni²⁺ coordinated COFs by the traditional room temperature method and solvothermal method and concluded that the solvothermal method enforces the planarity of trans-form 2,2′-bipyridine moiety for coordination with Ni²⁺ (Figure 22). The band gap of TpBpy-Ni2% was 1.84 eV, which was lower than the TpBpy-COF (2.16 eV). The presence of Ni remarkably enhances the photocatalytic performance (51,300 μmol h⁻¹ g⁻¹), 2.5 times higher than the pristine TpBpy-COF. The Ni coordination to bypyridine of TpBpy-Ni2% facilitates the planar structure by hindering the free bond rotation and enabling an efficient charge transfer process.

6.3. Donor–Acceptor (D-A) Configuration

The D-A configuration refers to the arrangement of electron-donating and electron-accepting moieties within the COF framework, which can create an internal electric field that facilitates charge separation and migration. The D-A configuration can be achieved by using different building blocks or linkages with different electron affinities or introducing heteroatoms or metal complexes into the framework.

In 2020, Yue-Biao Zhang et al. reported the donor–acceptor (D-A) system, PyTz-COF, that was fabricated by the condensation of the electron-rich pyrene (Py) and electron-deficient thiazolo[5,4-d]thiazole (Tz) (Figure 21) [78]. The design strategy was inspired by the work of Lotsch and co-workers [69], in which 2,4,6-trivinylcyclohexane-1,3,5-trione (Tp) subunits were exchanged by more electron-rich 1,3,6,8-tetraphenylpyrene (Py) systems to synthesize the efficient D-A system and an improved electron push-pull interaction, and the overlap of the π-orbitals of the Py and Tz subunits would result in a more effective charge separation. In the fourth hour of HER, the H₂ evolution rate recorded was 2072.4 μmol g⁻¹ h⁻¹, in the presence of Pt as a co-catalyst and ascorbic acid as a hole-capturing reagent under visible light irradiation. The PyTz-COF exhibited an absorption band at 540 nm with a band gap of 2.20 eV. The photoluminescence (PL) spectroscopy and nanosecond transient time profile revealed the superior charge separation ability of PyTz-COF.

In the same year, Yan Geng and Yu-Bin Dong co-workers reported the benzothiadiazole-based BT-TAPT-COF synthesized by the coupling of 1,3,5-tris-(4aminophenyl)triazine (TAPT) and 4,4′-(benzothiadiazole-4,7-diyl)dibenzaldehyde (TP) (Figure 23) [79]. The synthesized COF showed the HER at a rate of 949 μmol g⁻¹ h⁻¹ with Pt as a co-catalyst and ascorbic acid as a sacrificial electron donor under visible light. The band gap calculated for the COF was 2.35 eV, large enough to promote the HER.

Chen et al. used the D-A approach and showed that minor variations of the chemical structure of 2D COF (Py-HTP-BT-COF) via chlorination (Py-ClTP-BT-COF) and fluorination (Py-FTP-BT-COF) could lead to enhanced photocatalytic H₂ evolution (Figure 24) [80]. They fabricated the benzothiadiazole-based COF by using electron-rich 4,4′,4″,4‴-(pyrene-1,3,6,8-tetrayl)tetrabenzaldehyde (Py-CHO) with electron-deficient terphenyl-based diamines (XTP-BT-NH₂) with choro and floro substitution. The hydrogen evolution rate was directly impacted by the substitution as Py-ClTP-BT-COF showed higher HER (177.50 μmol h⁻¹ g⁻¹) than that of Py-FTP-BT-COF (57.50 μmol h⁻¹ g⁻¹) and the non-halogenated Py-HTP-BT-COF (21.56 μmol h⁻¹ g⁻¹). In the absence of Pt co-catalyst, the Py-ClTP-BT-COF showed H2 evolution at a rate of 44.00 μmol h⁻¹ g⁻¹. The optical band gaps calculated were 2.25 eV, 2.36 eV, and 2.34 eV for Py-HTP-BT-COF, Py-ClTP-BT-COF, and Py-FTP-BT-COF, respectively. The authors used transient photocurrent measurements and electrochemical impedance spectroscopy (EIS) for the first time to investigate charge separation and migration. The transient photocurrents of Py-ClTP-BT-COF and Py-FTP-BT-COF were significantly improved, demonstrating the fast photoresponse of Py-ClTP-BT-COF and Py-FTP-BT-COF as compared to HTP-BT-COF. The reduced resistance in halogenated COFs also reduced the charge transfer barrier. The authors also proposed that the electronegativity of halogen also favors the delocalization of π-electrons, which signifies the found result. This study successfully demonstrated that halogen incorporation in the COF skeleton not only promotes efficient charge separation but can also lower the activation energy for HER.

Zhenjie Zhang et al. reported the COFs (NKCOFs), fabricated using electric donor–acceptor moieties (Figure 25) [81]. This study used benzothiadiazole with different functional groups as acceptors and pyrene as donors. The monofloro-substituted benzothiadiazole-based NKCOF-108 exhibited enhanced photocatalytic performances (HER = 120 μmol h⁻¹) at 520 nm among all other synthesized NKCOFs. The band gap of NKCOF-108 was found to be 1.82 eV. The density of states (DOS) of NKCOFs showed that the C atom’s 2p orbital contributed primarily to the HOMO and LUMO levels, proving that the photogenerated electrons result from π-delocalization. Therefore, an efficient method to maximize the photo-redox potentials of COFs is to modify electron acceptors with various functional groups.

Shu Seki et al. reported a series of 2D isoreticular COFs by the condensation reaction of 4,4″–diamino substituted p–terphenyl (Tp) derivatives as the linker with 1,3,5–triformylphloroglucinol (Figure 26) [82]. The BtCOF150 consisting electron-deficient benzothiadiazole, showed good hydrogen evolution efficiency (750 ± 25 µmol h⁻¹ g⁻¹) as compared to other synthesized COFs, with good stability under visible light irradiation. The band gap for BtCOF150 was 2.0 eV, with absorption at 620 nm. The D-A structure, where benzothiadiazole acted as an electron acceptor, and TH acted as the electron donor, was found to be responsible for charge transfer in HER. Because the AA′ stacking significantly enhances exciton migration and charge transfer, it also exhibits better photocatalytic activity, which was correlated with the superior activity of BtCOF150.

Yingjie Zhao et al. reported the three vinylene-linked 2D COFs (BTH-1, 2, 3) containing benzobisthiazoles units as functional groups (Figure 27A) [83]. Under visible light (λ > 420 nm) irradiation, the BTH-3 with benzotrithiophene as the donor and a potent D-A effect had an appealing photocatalytic HER of 15.1 mmol h⁻¹g⁻¹. Even at 600 nm, BTH-3 has good photoactivity due to its narrow band gap (2.02 eV). Strong charge transfer effects may prevent ineffective and spontaneous electron-hole recombination and explain BTH-3’s improved performance. Additionally, electrochemical impedance spectroscopy indicates that BTH-3 had the best electronic conductivity, whereas BTH-2 had the lowest. Recently, same research group reported COF-containing benzotrithiophene moieties with better planarity and π-conjugation (BTTh-TZ-COF) (Figure 27B) [84]. The hydrogen evolution for BTTh-TZ-COF was reported as 5.22 mmol h⁻¹ g⁻¹ compared to TThB-TZ-COF as a counterpart (1.03 mmol h⁻¹ g⁻¹) with adequate gap of 2.32 eV. The authors concluded that the better charge separation and transfer efficiencies originated from the structures themselves and were the main factors for HER. The comparison between the two types of conjugated triazine-based COFs with different connected linkages (imine bond and sp2 -carbon–CN bond) was made by Yingjie Zhao (Figure 27C) [85]. The PTPA-COF was fabricated by the reaction between 2,4,6-tris(4-formylphenyl)-1,3,5-triazine and 2-phenylenediamine, whereas TP-COF was synthesized by the reaction in between 2,4,6-tris(4-formylphenyl)-1,3,5-triazine and 2-phenylacetonitrile. The TP-COF at 450 nm generated the H₂ at the rate of 29.12 mmol h⁻¹ g^−1, which was higher than its imine-linked counterpart (36 μmol h⁻¹ g⁻¹). According to theoretical calculations and fs-TA spectroscopy, cyano-substituted sp2 C=C linkages play a crucial role in high-efficiency charge separation and transfer.

Xiong Chen et al. recently reported the vinylene-linked COFs, which were synthesized by Knoevenagel polymerization with a suitable D-A pair (Figure 28) [86]. The light-harvesting, optical-bandgap, and charge-transfer (CT) characteristics of the COFs were carefully controlled by adjusting the donor moieties in the skeletons from phenyl to 2,5-dimethylbenzene and 3,3′-dimethyl-1,1′-biphenyl. The 2,4,6-trimethyl-1,3,5-triazine (TM) and 3,3′-dimethyl[1,1′-biphenyl]-4,4′-dicarboxaldehyde (DMA) monomers were used to create the TM-DMA-COF, which displayed the most significant D-A interactions, excellent charge-carrier separation and transfer kinetics, and a lower energy barrier for the formation of H₂. As a result, it provided the highest HER, 4300 mmol h⁻¹ g⁻¹.

A new donor–acceptor COFs (TAPFy-PhI and TAPB-PhI) was recently reported Yunsheng Ding et al., in which synthesized by the condensation of 1,3,6,8-tetrakis(4-aminophenyl)pyrene (TAPFy) and 1,3,5-tris(4-aminophenyl)benzene (TAPB) with phthalimide (PhI) (Figure 29) [87]. The pyrene-based COF TAPFy-PhI exhibited a substantial red shift of ~100 nm compared to the TAPB-PhI COF (absorption ranging from 300 to 600 nm) due to more extensive π-conjugation system. TAPFy-PhI and TAPB-PhI COFs found to have band gaps 2.21 and 2.66 eV, respectively. The TAPFy-PhI showed much higher hydrogen evolution rate of 1763 μmol g⁻¹ h⁻¹ as compared to TAPB-PhI (7 μmol g⁻¹ h⁻¹), which was due to the more rigid structure and larger π conjugation of pyrene. The highest hydrogen evolution rate of TAPFy-PhI 2718 μmol g⁻¹ h⁻¹ was obtained under visible-light irradiation (>420 nm) in the presence of 1 wt % Pt.

Xiao-Jun Sun et al. explored a series of ketoneamine-based COFs with different functional groups for photocatalysis [88]. The COFs were fabricated by the condensation reaction of 1,3,5-triformylphloroglucinol (TP) with corresponding diamine (Paraphenylenediamine for TpPa−COF; 2,5-dimethyl-p-phenylenediamine for TpPa−COF−(CH₃)₂; 2-nitro-1,4-phenelynediamine for TpPa−COF−NO₂) (Figure 30). It was found that TpPa−COF−(CH₃)₂ shows the most substantial absorption among all, having a band gap of 2.06 eV. Under visible light irradiation, the TpPa−COF−(CH₃)₂ exhibits a maximal photocatalytic activity of 8.33 mmol g⁻¹ h⁻¹, which was 5.3 times higher than the TpPa−COF and 38 times higher than the TpPa−COF−NO₂. The photocurrent density of TpPa−COF−(CH₃)₂ was found to be the highest among all, which revealed the best separation efficiency. The authors concluded that the electron-donating groups can have a more substantial conjugation effect, which enhances light absorption and strengthens charge carrier mobility.

Fang-You Yan et al. recently reported the Pyrazine-functionalized D-A PyPz-COF synthesized by the condensation of 4,4′,4″,4‴-(pyrene-1,3,6,8-tetrayl)tetraaniline (PyTA) and 4,4′-(pyrazine-2,5-diyl)dibenzaldehyde (PzDA) (Figure 31) [89]. The PyPz-COF contained the eight inward N-sites extend layer by layer along the z-axis, forming an ordered chain of N-sites on pore walls having the size of 2.8 nm. The optical bandgap of PyPz-COF was 2.05 eV. The PyPz-COF showed excellent photocatalysis hydrogen evolution up to 7542 μmol g⁻¹ h⁻¹.

Eunsung Lee et al. recently reported the phenanthroimidazole-based D-A COF for HER [90]. The PIm-COF1 and PIm-COF2 were synthesized via the condensation of 1H-phenanthro[9,10-d] imidazole-5,10-diamine (PIDA) (donor) with different aldehydes (acceptor), such as 1,3,5-benzenetricarboxaldehyde (BTA) and 2,4,6-triformylphloroglucinol (TP), respectively (Figure 32). The PIm-COF1 and PIm-COF2 were found to have the same pore size (2.4 nM). The optical band gap energy of PIm-COF1 and PIm-COF2 was 2.34 and 2.13 eV, respectively. The average rate of the HER was 7417.5 μmolg⁻¹ h⁻¹ for PIm-COF2, which was 14 times higher than that of PIm-COF1 (528.5 μmolg⁻¹ h⁻¹).

In search of efficient D-A COF, researchers used the benzotrithiophene core [91] as an electron-donating counterpart in the synthesis of COFs. Jong-Beom Baek et al. used benzotrithiophene tricarbaldehyde (BTT) and aromatic diamines, benzene-1,4-diamine (PDA), naphthalene-2,6-diamine (NDA), anthracene-2,6-diamine (AnthDA) or [1,1′-biphenyl]-4,4′-diamine (BPhDA) for the synthesis of BTT-PDA, BTT-NDA, BTT-AnthDA, and BTT-BPhDA, respectively (Figure 33A) [92]. The authors found a significant increase in pore size (caluculated by NLDFT) by increasing the aromatic ring in the synthesized COFs from 2.48 to 3.23 nm from BTT-PDA, to BTT-BPhDA. The determined optical band-gaps were 2.15, 2.16, 2.00, and 2.21 eV for the BTT-PDA, BTT-NDA, BTT-AnthDA, and BTT-BPhDA, respectively. The rate of H₂ production increased from phenyl (BTT-PDA; 2.04 mmol g⁻¹ h⁻¹) to naphthalene (BTT-NDA; 5.22 mmol g⁻¹ h⁻¹), and it started to decline in anthracene among the BTT-COFs with various polyaromatic units (BTT-AnthDA; 4.23 mmol g⁻¹ h⁻¹). The reduced rate in BTTAnthDA was found to be caused by the lower band gap and the strong charge stabilization of the anthracene unit. Proper charge stabilization is helpful to prevent charge recombination, but when the charges are firmly localized at particular places, it hinders charge transfers and increases the likelihood of recombination. In addition to the use of benzotrithiophene Xiaoming Liu et al. reported the three-component D-A COFs [93]. The authors used 1,3,5-tris-(4-formyl-phenyl)triazine (TFPT) and benzo[1,2-b:3,4-b′:5,6-b″]trithiophene-2,5,8-tricarbaldehyde (BTT) as nodes, while 1,4-phenyldiacetonitrile (PDAN) or 1,4-phenylenediamine (PDA) as a linker to construct two isomorphic TCDA-COFs COF-JUL35 and COF JUL36 (Figure 33B). Due to the isomorphic framework, the pore size distribution of synthesized TCDA-COFs was comparable to maxima at about 2.7 nm. The calculated optical band gap was 1.85 eV for COF-JLU35, which was smaller than that of COF-JLU36 (2.12 eV). Among both TCDA-COFs JLU35 exhibited the HER of up to 70.8 ± 1.9 mmol g⁻¹ h⁻¹. Bo Wang et al. also utilized the benzotrithiophene core for the synthesis of BTT-Bpy-COF, but they studied the protonated COF for HER (Figure 33C) [94]. The bipyridine-based BTT-Bpy-COF was synthesized by the Knoevenagel condensation of benzotrithiophene-2,5,8-tricarbaldehyde (BTT-CHO) and 2,2′-bipyridine-5,5′-diacetonitrile (BPy-2CN). The band gap calculated for BTT-Bpy-COF was 2.15 eV, which was reduced to 1.41 eV, which caused the absorbance of a broader visible spectrum. The HER performance of BTT-Bpy-COF recorded at the rate of 15.8 mmol g⁻¹ h⁻¹, reached 21.2 mmol g⁻¹ h⁻¹ under simulated sunlight. The authors concluded that the post-synthetic protonation resulted in improved charge-separation efficiency and increased hydrophilicity in the pore channels synergistically.

6.4. Hetero-Structural Mixed Linkage Approach

In addition to the above approaches, several other approaches were adopted, one of which is fabrication of hybrid systems in which COFs are incorporated with the Ti-based superstructure.

Yanli Zhao and co-workers used the hetero-structural mixed linkage approach to enhance the charge separation in the photocatalytic process [95]. Three ß-ketoenamine-linked COFs ordered NTUBDA-HTA, NTU-BDA-DHTA, and NTU-BDA-THTA were fabricated by using the different stoichiometric ratio of benzene-1,4-diamine (BDA) with benzene-1,3,5-tricarbaldehyde derivatives bearing different numbers of hydroxy groups monomers (Figure 34). The NTU-BDA-THTA was found to be superior among others in terms of hydrogen evolution (1.47 μmol g⁻¹ h⁻¹), which is attributed to the increased ß-ketoenamine linkages. The increasing ketoenamine linkages led to the increment in the oxidizing power of photogenerated holes, which enhanced the electron migration in the proton reduction. To further improve the HER COF/NH₂-Ti₃C₂Tx, hetero-hybrids were fabricated by a covalent bond with a sufficient contact interface and efficient charge transfer. The NTU-BDA-THTA with NH₂-Ti₃C₂Tx was found to generate H₂ at a rate of 14,228.1 μmol g⁻¹ h⁻¹. This rational construction of COFs and their integration with other functional materials need to be explored further to get highly efficient photocatalysts.

Yilin Wu et al. recently reported the metal-free photocatalyst with 2D-2D BP/TpPa-1-COF p-n heterojunction by “face-to-face” contact between 2D BP flakes and TpPa-1-COF for efficient HER. The 15% BP/TpPa-1-COF p-n heterojunction exhibited approximately 13-fold higher HER (456.7 µmol h⁻¹ g⁻¹) than pristine TpPa-1-COF and 24-fold higher than 2D bulk phosphorus (BP), respectively (Figure 35) [96]. The authors decided to make the heterojunction of BP with COF because of its excellent p-type semiconductor properties [97]. The Mott-Schottky (MS) charts characterized the n-type and p-type semiconductor properties of 2D BP and TpPa-1-COF, respectively. The charge transfer rate of the heterojunction was 7 × 10⁷ s⁻¹, revealing a more effective charge transfer.

Xin Li, reported the combined TSCOF and WO₃ material by covalent bonds and fabricated an efficient Z-scheme heterostructure photocatalyst for photocatalytic overall water splitting (Figure 36) [98]. The separation of photo-generated electron-hole pairs and utilization efficiency of charge carriers is considerably improved due to the synergistic effect between the built-in electric field in Z scheme heterojunctions and the ultrathin structure of TSCOF. TSCOF has been reported for water reduction, and WO₃ typically led to water oxidation. The hybrid TSCOFW composite photocatalyst outperformed previously reported photocatalysts with its photocatalytic HER of 593 mmol h⁻¹ g⁻¹. The BET surface area of TSCOFW was 83.2 m² g⁻¹ with pore diameters of ≈2.6 nm and a bandgap of 1.95 eV. The authors discovered that, compared to pure TSCOF, the development of Z-scheme heterojunctions in TSCOFW significantly improves the process of charge separation. This is because, in TSCOFW, the electron trapping had a longer lifetime and was more densely distributed along the conduction band edge.

Yangang Wang et al. reported the van der Walls heterojunction 2D/2D COF/g-C₃N₄ heterojunction composite in which COF was synthesized from the condensation of 1,3,5-triformylphloroglucinol (TP) and p phenylenediamine (PDA) (Figure 37) [99]. The COF/CN heterojunction’s design allows the van der Waals forces and internal electric field to significantly speed up the separation of photoinduced e⁻–h⁺ pairs, which enhances the photocatalytic efficiency (HER = 449.64 μmol h⁻¹). The pore size of synthesized COF was 1.7 nm. In the mechanistic aspects, the authors suggest that under visible light irradiation, both CN and COF can be excited, leading to the generation of e⁻–h⁺ pairs. The photogenerated h⁺ in the VB of COF can quickly migrate to the VB of CN, whereas the photoexcited e⁻ on the CB of CN can move quickly to the CB of COF. The photogenerated electron-hole pair can achieve efficient spatial separation due to this type of charge carrier transport channel, which prevents recombination. As a result, the photoexcited h⁺ gathers in the VB of CN and is captured by sacrificial triethanolamine; the photoinduced e⁻ gathers on the surface of COF and combines with hydrogen ions to form hydrogen.

Jinlong Zhang et al. also used the hybridization design approach for the fabrication of organic/inorganic hybrid TpPa-1-COF/ZnIn₂S₄ S-Scheme heterojunctions for hydrogen evolution (Figure 38) [100]. TpPa-1-COF/ZnIn₂S₄ heterojunction was synthesized by the hydrothermal reaction between TpPa-1-COF, Zn(NO³)², In(NO³)³, and L-cysteine followed by ultrasonication. The COF/ZIS-20% sample shows the highest photocatalytic efficiency, reaching 853 μmol g⁻¹ h⁻¹, which was approximately 6.2-fold greater than that of TpPa-1-COF. In the Nyquist plot, the semicircle diameter of the COF/ZIS-20% sample is smaller than that of the TpPa-1-COF and ZnIn₂S₄, which was in support of minimum charge transport resistance further demonstrated that the TpPa-1-COF/ZnIn₂S₄ heterostructure could promote carrier transport and separation. To conclude the S-scheme mechanism rather than the type-II mechanism, authors observed the reaction of •O2− with ZnIn₂S₄ and TpPa-1-COF. It has been found that the •O2− EPR signal was observed only in the presence of ZnIn₂S₄ because the ECBM of TpPa-1-COF was more positive than O2/•O2− and the same was observed in the case of TpPa-1-COF/ZnIn₂S₄, which proved that the photogenerated electrons are localized on the ECBM of ZnIn₂S₄ to participate in the reductive reaction. The same mechanism was also explained with the DFT calculations.

6.5. Miscellaneous Approach

6.5.1. Electron-Transfer Mediators (ETMs)

Jia Guo et al. utilized the concept of electron-transfer mediators (ETMs) incorporated in COF for photocatalytic HER [101]. The ability of electron-transfer mediators (ETMs) to absorb and donate electrons in a dynamic equilibrium and maintain electron-transfer processes to increase reaction efficiency is well recognized [102]. The base COF skeleton was synthesized by the condensation of 5,5′-Diamino-2,2′-bipyridine and triformylphloroglucinol, which was quaternized with dibromo alkanes into cyclic diquats Tp-nC/BPy²⁺-COF (n = 2, 3, 4) (Figure 39). The electrostatic repulsion among positivity-charged BPy²⁺ fixed their position in the framework at a definite place, promoting enhanced electron transfer. The hydrogen evolution efficiency of the 2C/BPy²⁺-COF was the highest among all (34,600 μmol h⁻¹ g⁻¹) up to 48 h. The space charge limited current (SCLC) method revealed the most increased charge mobility for the 2C/BPy²⁺-COF (0.41cm² V⁻¹ s⁻¹).

Recently, Sermet Koyuncu et al. used the quaternization approach, but in different aspects, as they synthesized the viologen-based COFs with different length alkyl chain bridges [103]. They utilized the ability of viologens to undergo reversible redox reactions, which is crucial in the photocatalytic application [104]. The main monomer 4,4′-bis(3,6-dipyridine-9H-carbazo-9-yl)-1,1′-biphenyl (TPCBP) was synthesized using TBCBP and Pyridine-4-boronic acid pinacol ester via Suzuki reaction. The TPCBP was then treated with 1,2-dibromoethane, 1,4-dibromobutane, and 1,6-dibromohexane to get the TPCBP X-COF [X = ethyl (E), butyl (B), and hexyl (H)] (Figure 40). The most efficient COF in terms of HER was found to be TPCBP B-COF (12.276 mmol g⁻¹) with band gap of 2.63 eV. The appropriate bridge length of the alkyl linker in TPCBP X-COF was found to prevent recombination with the TPCBP electroactive structures, resulting in the highest HER efficiency. The TPCBP E-COF, with very short distances between electroactive structures, was more susceptible to charge recombination. On the other hand, TPCBP B-COF, with an ideal bridge length, effectively prevented recombination, making it the most efficient structure. The enhanced photoelectric activity of TPCBP B-COF indicated inhibition of photogenerated electron-hole recombination and favorable charge transfer. In contrast, the flexibility of the hexyl bridge in TPCBP H-COF may have reduced photocatalytic performance by closing the holes, decreasing surface area, and increasing the band gap.

6.5.2. Graphene like COFs

Fan Zhang et al. reported the graphene like g-C_xN_y-COFs, which were synthesized by the Knoevenagel condensation of 2,4,6-trimethyl-1,3,5-triazine (TMTA) and terephthalaldehyde (DFB) or 5′-(4-formylphenyl)-[1,1′:3′,1″-terphenyl]-4,4″-dicarbaldehyde (TFPB) (Figure 41A) [105]. Broad absorption bands with the highest peaks at 450 nm for g-C₁₈N₃-COF and 400 nm for g-C₃₃N₃-COF were found with 2.42 eV and 2.54 eV band gaps, respectively. In comparison to g-C₃₃N₃-COF, which had an average hydrogen evolution rate (HER) of 3.7 mol h⁻¹, g-C₁₈N₃-COF had an HER of 14.6 mol h⁻¹, nearly three times higher cathodic photocurrent of g-C₁₈N₃-COF (~25 μA cm⁻²) demonstrating the more effective separation of charge carriers and superior photocatalytic activity. The longer conjugated backbone of g-C₁₈N₃-COF brings about higher efficiency in the migration and transmission of photogenerated electron-hole pairs. Additionally, crystallinity reduces the recombination effect of the photogenerated charge carriers at the defects. The same group reported COFs with trans-disubstituted C=C linkages via condensation reaction at arylmethyl carbon atoms based on 3,5-dicyano-2,4,6-trimethylpyridine and linear/trigonal aldehyde (i.e., 4,4″-diformyl-p-terphenyl, 4,4′-diformyl-1,1′-biphenyl, or 1,3,5-tris(4-formylphenyl)benzene) monomers (Figure 41B) [106]. The absorption maxima at 500 nm for g-C₄₀N₃-COF was observed with an optical band gap of 2.36 eV. The most active g-C₄₀N₃-COF, with 3 wt% Pt, was chosen, and the reaction was performed for a total of 28 h while being exposed to visible light (>420 nm). Approximately 4400 μmol of hydrogen had evolved overall after 28 h. Later on, they investigated the impact of molecular geometry of the COF for HER [107]. The vinylene-linked COF (g-C₅₄N₆-COF) was synthesized by the Knoevenagel condensation of D_3h-symmetric monomer 2,4,6-tris(4′-formyl-biphenyl-4-yl)-1,3,5-triazine instead of C_2v-symmetric 3,5-dicyano-2,4,6-trimethylpyridine from the previous study (Figure 41C). The g-C₅₄N₆-COF possessed two different types of octupolar subunits centered around the corresponding meta-substituted tricyanobenzene and 1,3,5-triazine moieties. The H₂ evolution efficiency of g-C₅₄N₆-COF was 2518.9 μmol h⁻¹ g⁻¹, two times higher than that of the g-C₅₂N₆-COF (1171.9 μmol h⁻¹ g⁻¹). This work was beneficial for the logical design of polymeric photocatalysts because it expanded the range of vinylene-linked C OFs and demonstrated how changing the geometric symmetry of the constituent parts of a conjugated framework could significantly affect its semiconducting behaviors.

7. Challenges for the Use of COFs as Photocatalysts for Hydrogen Evolution

Despite the significant progress in the development of COF-based photocatalysts for HER, there are still some challenges and constraints that need to be addressed, as follows:

Stability: The stability of COFs under photocatalytic conditions is crucial for their practical applications. However, some COFs may suffer from hydrolysis, oxidation, or photodegradation due to their reversible linkages or sensitive building blocks. Therefore, more robust and stable COFs need to be developed by using irreversible or stable linkages or by protecting or modifying their vulnerable sites.

Scalability: The scalability of COF synthesis is another important factor for its large-scale applications. However, most of the current synthetic methods for COFs are based on solvothermal or ionothermal reactions that require high temperature, pressure, or solvent consumption. Therefore, more facile and green synthetic methods for COFs need to be explored by using low temperature, pressure, or solvent-free conditions.

Overall water splitting: The overall water splitting activity of COFs is rarely known due to the difficulty in achieving efficient water oxidation reactions (WOR). Most of the current COF-based photocatalysts require sacrificial electron donors to achieve HER from water. Therefore, more efforts need to be devoted to developing COF-based photocatalysts that can perform both HER and WOR simultaneously or coupling COFs with suitable WOR catalysts.

8. Conclusions

The successive progression of COFs as photocatalysts for hydrogen evolution from water splitting showcases the immense potential of these materials in the field of renewable energy. Their unique properties, tunable structures, and growing synthetic techniques have pushed COFs to the forefront of photocatalysis research. As we move forward, continued efforts in understanding and improving COF-based photocatalysts will undoubtedly contribute significantly to a sustainable future with efficient hydrogen production from water splitting. This review briefly summarizes the progress in the field of COFs as photocatalysis for HER based on structure tuning and mechanistic aspects. Although COFs are in their infancy stage compared to metallic semiconductors, Xuanhua Li recently reported more than 100% internal quantum efficiency [108]. However, understanding the mechanism involved and the structural needs can surpass the limitations associated with COFs.