Synergistically modulating the active-site density and charge-transfer in covalent organic frameworks for boosting electrocatalytic water splitting

Abstract

Covalent organic frameworks (COFs) have emerged as a promising precise platform for designing metal-free bifunctional electrocatalysts. Herein, we designed and synthesized two bicarbazole-based metal-free COFs (NUST-69 and NUST-70) with different organic linkers, where NUST-70 incorporates a benzothiadiazole (BT) moiety. Notably, NUST-70, featuring multiple N and S active sites of BT and a donor–acceptor (D–A) structure, demonstrates exceptional electrocatalytic activity, particularly for the oxygen evolution reaction (OER), achieving an overpotential of 292 mV at 10 mA cm⁻² and maintaining stability over 48 h of chronoamperometry testing. The enhanced OER performance is attributed to the synergistic modulation of the active-site density and charge-transfer induced by the BT unit, as confirmed by experimental characterization and theoretical calculations. This study highlights the potential of rationally designed D–A type COFs as high-performance metal-free electrocatalysts for advancing clean energy conversion.

---

Gen Zhang

Gen Zhang received his PhD in chemistry from Lanzhou University in 2013. After his postdoctoral research at the University of Manchester (2013–2015) and at Kyoto University (2016–2019), he started independent work at Nanjing University of Science and Technology. The research in his laboratory focuses on functional organic porous materials for energy and environmental applications.

---

Introduction

With the advancement of metal electrocatalysts for both the hydrogen and oxygen evolution reactions (HER and OER), encompassing noble metals (e.g., Pt, IrO₂, RuO₂) and transition metals (e.g., Co, Ni), the design and synthesis of metal-free electrocatalysts have similarly garnered significant interest.^1,2 In this context, metal-free carbon-based materials have gained considerable attention as viable alternatives, owing to their earth abundance, cost-effectiveness, tunable molecular structures, and multifunctional active sites.³ The catalytic performance of these carbon-based materials is predominantly governed by heteroatom doping (e.g., N, S, and P), where the distinct electronegativity, van der Waals radii, valence electron configurations, and electronic structures of these dopants induce significant modifications in the physicochemical properties of adjacent carbon atoms.^4,5 These modifications include altered charge density distributions, adjusted band structures, and enhanced thermal/chemical stability, ultimately dictating the catalytic activity. The reasonable doping of single atoms or multiple heteroatoms is crucial for significantly improving their intrinsic catalytic activity.^6–8 Furthermore, a pronounced synergistic effect emerges when the N–S interatomic distance is maintained below 7.5 Å.^5,9–11 However, due to the lack of atomic-resolution structures, it is hard to give a clear structure–property relationship.

Covalent organic frameworks (COFs), characterized by their covalently linked structures, possess inherent advantages for electrocatalytic applications including permanent porosity, high specific surface area, well-defined one-dimensional channels, and abundant active sites.^12–14 The structural versatility of organic building blocks enables precise control over the configuration and composition of COFs. Among various COFs, donor–acceptor (D–A) type COFs, with their unique electronic structures, have demonstrated significant potential in applications such as catalytic reactions, photothermal therapy, and electronic materials.¹⁵ On one hand, benzothiadiazole, as a state-of-the-art electron acceptor, exhibits exceptional capability in bandgap modulation and significantly enhances charge separation/transfer efficiency in conjugated polymer systems. On the other hand, carbazole, a nitrogen-containing aromatic heterocycle with a rigid fused-ring structure, offers distinct advantages over other nitrogen-containing moieties (e.g., triazine and heptazine).^16,17 Its extended π-conjugation system facilitates enhanced intramolecular charge transfer, contributing to superior optoelectronic properties that have demonstrated exceptional performance in photovoltaic devices,^16,17 gas adsorption systems, and energy storage applications.¹⁸ The strategic incorporation of benzothiadiazole with carbazole to form D–A type porous organic polymers has been scarcely investigated.¹⁹ While some bifunctional HER/OER catalysts have demonstrated promising performance, the development of efficient D–A-type COF catalysts capable of operating in alkaline media remains an outstanding challenge in the field.^5,11,20–22

Herein, we designed and synthesized two aldehyde-functionalized building blocks (BCTB and BCTBZ) with distinct electronic structures through structural modification of the 9,9′-bicarbazole motif. These building blocks were subsequently employed to construct two novel 2D COFs (NUST-69 and NUST-70) via Schiff-base condensation with 1,4-diamino-2,5-dimethylbenzene (DAB-Me) (Fig. 1). Systematic investigation of the micro-environmental effects in the COFs revealed their significant influence on HER and OER electrocatalytic performance. Notably, the synergistic combination of carbazole units and methyl groups in these frameworks facilitates efficient charge transfer during both HER and OER processes, endowing the COFs with remarkable catalytic activity and stability. The N,S-codoped D–A structured NUST-70 exhibited particularly outstanding OER and HER performance in alkaline media, achieving low overpotentials of 292 and 261 mV at a current density of 10 mA cm⁻², respectively. These findings establish a new paradigm for the rational design of metal-free COF based bifunctional electrocatalysts for overall water splitting applications.

Fig. 1 Synthesis routes to NUST-69 and NUST-70.

Experimental section

Synthesis of NUST-69

BCTB (20.0 mg, 0.027 mmol) and DAB-Me (7.3 mg, 0.053 mmol) were weighed into a glass ampoule with mesitylene (1.5 mL) and 1,4-dioxane (0.5 mL). The solution was subjected to ultrasound for 10 minutes, and then 6 M aqueous acetic acid (0.2 mL) was added into the glass ampoule as the catalyst. The glass ampoule was flash frozen at 77 K using a liquid nitrogen bath and degassed by three freeze–pump–thaw cycles, and then was flame sealed. The glass ampoule was placed at 120 °C for 3 days. The solid was isolated by centrifugation and washed with MeOH (3 × 10 mL) and further purified by Soxhlet extraction with N,N-dimethylformamide (DMF). The powder was dried at 80 °C under vacuum overnight to afford NUST-69 as a bright yellow solid (21.8 mg, 86%).

Synthesis of NUST-70

BCTBZ (20.0 mg, 0.020 mmol) and DAB-Me (5.6 mg, 0.041 mmol) were weighed into a glass ampoule with o-dichlorobenzene (1.0 mL) and n-butanol (1.0 mL). The solution was subjected to ultrasound for 10 minutes, and then 6 M aqueous acetic acid (0.2 mL) was added into the glass ampoule as the catalyst. The glass ampoule was flash frozen at 77 K using a liquid nitrogen bath and degassed by three freeze–pump–thaw cycles, and then was flame sealed. The glass ampoule was placed at 120 °C for 3 days. The solid was isolated by centrifugation and washed with MeOH (3 × 10 mL) and further purified by Soxhlet extraction with DMF. The powder was dried at 80 °C under vacuum overnight to afford NUST-70 as a bright yellow solid (21.5 mg, 89%).

Electrochemical test

The electrochemical performance of the synthesized COFs was systematically evaluated using a standard three-electrode configuration, with measurements conducted on a CHI 760E electrochemical workstation (CHI, China). A homogeneous catalyst ink was prepared by thoroughly mixing 2.0 mg of COF material with 8.0 μL of absolute ethanol (AR) and 16 μL of polytetrafluoroethylene (PTFE, 5 wt%) aqueous dispersion. The resulting slurry was uniformly coated onto a pre-treated nickel foam substrate (1 × 1 cm²), which had been sequentially cleaned with 0.1 M HCl and deionized water. The working electrode was subsequently compressed under 5 MPa pressure to ensure optimal catalyst-substrate contact. A Pt sheet and Hg/HgO were considered as the counter electrode and reference electrode respectively. Prior to electrochemical measurements, the 1.0 M KOH electrolyte was oxygen-saturated through continuous O₂ purging to facilitate rapid bubble detachment from the working electrode surface. To optimize the solid–liquid–gas triple-phase boundary at the electrode–electrolyte interface, the electrode was subjected to electrochemical activation to enhance its hydrophilicity before performing linear sweep voltammetry (LSV) analysis.

Theoretical calculations

Density functional theory calculations were performed by using the CP2K package.²³ The PBE functional with Grimme D3 correction was used to describe the system.^24,25 Unrestricted Kohn–Sham DFT was used as the electronic structure method in the framework of the Gaussian and plane waves method.^26,27 Goedecker–Teter–Hutter (GTH) pseudopotentials^28,29 and DZVP-MOLOPT-GTH basis sets were utilized to describe the molecules.²⁶ A plane-wave energy cut-off of 500 Ry was employed. The simulation was carried out in a three-dimensional periodic boundary box of 29.08 × 29.08 × 29.08 ų.

Results and discussion

Synthesis and characterization

Two tetraaldehyde building blocks, 4,4′,4′′,4′′′-([9,9′-bicarbazole]-3,3′,6,6′-tetrayl)tetrabenzaldehyde (BCTB) and 7,7′,7′′,7′′′-([9,9′-bicarbazole]-3,3′,6,6′-tetrayl)tetrakis(benzo[c][1,2,5]thiadiazole-4-carbaldehyde) (BCTBZ), were rationally designed and synthesized from 3,3′,6,6′-tetrabromo-9,9′-dicarbazole (BC-4Br) as the precursor (crystal structure of BCTBZ is shown in Fig. S3 and Table S1). Notably, DAB-Me was selected because the introduction of methyl groups into the COF framework suppresses charge recombination through localized charge redistribution.^30–33 The corresponding 2D COFs were constructed via solvothermal Schiff-base condensation. Specifically, BCTB and DAB-Me were dissolved in a mesitylene/dioxane (3 : 1, v/v) mixture with 6 M acetic acid (10 vol%) as the catalyst, yielding NUST-69 as a yellow powder (86% yield) after 72 h at 120 °C. Similarly, NUST-70 was obtained as a dark red powder (89% yield) using o-dichlorobenzene/n-butanol (1 : 1, v/v) as the solvent under analogous conditions. The formation of imine bonds in the COF skeleton was characterized by the Fourier transform infrared (FT-IR) spectrum (Fig. S4). The absorption peak of the C=O stretching vibration of building blocks BCTB and BCTBZ almost disappears at ∼1695 cm⁻¹. NUST-69 and NUST-70 exhibit stretching vibrations at ∼1599 and 1624 cm⁻¹, respectively, corresponding to the characteristic peak of C=N, indicating successful polymerization and raw material consumption.^33–35 The characteristic signals of C=N at 158 and 152 ppm for NUST-69 and NUST-70, respectively, were clearly observed in ¹³C cross-polarization magic angle spinning nuclear magnetic resonance (CP-MAS NMR) spectra. In addition, the obvious methyl (CH₃) characteristic signal at 18 and 17 ppm further demonstrates that NUST-69 and NUST-70 were successfully synthesized (Fig. S5 and S6).

The crystalline characteristics of the COFs were unambiguously determined by powder X-ray diffraction (PXRD) analysis. As evidenced in Fig. 2a and d, both COF materials exhibit intense diffraction peaks, indicative of their high crystallinity. Remarkably, the experimental PXRD patterns show excellent agreement with the simulated patterns generated by Materials studio, with negligible discrepancies, confirming an AA-stacked sql topology.³⁶ For NUST-69, the predominant diffraction peak at 3.56° is assigned to the (110) crystallographic plane, while the weaker reflections at 5.23°, 7.09°, and 10.67° correspond to the (120), (220), and (330) planes, respectively. Notably, NUST-70 displays nearly identical diffraction features, suggesting the preservation of the same framework architecture despite the structural modification. The unit cell parameters of the two COFs were refined by the Pawley method to obtain: NUST-69, a = 30.87 Å, b = 42.23 Å, c = 4.29 Å, and α = β = γ = 90°, R_wp = 6.56%, R_p = 4.62%; NUST-70, a = 30.27 Å, b = 42.22 Å, c = 4.29 Å, and α = β = γ = 90°, R_wp = 11.38%, R_p = 7.73%.

Fig. 2 Experimental and refined PXRD patterns of (a) NUST-69 and (d) NUST-70, N₂ adsorption–desorption isotherms (77 K) of (b) NUST-69 and (e) NUST-70 (inset: experimental pore size distribution), and structural representations of (c) NUST-69 and (f) NUST-70.

The permanent porosity and specific surface area of the synthesized COFs were systematically investigated through nitrogen adsorption–desorption measurements at 77 K. Prior to analysis, the samples were thermally activated under vacuum at 100 °C for 12 h to remove residual solvents. Both COFs exhibited characteristic type I adsorption isotherms (Fig. 2b and e), with rapid nitrogen uptake observed at low relative pressures (P/P₀ < 0.1), confirming their microporous nature. Brunauer–Emmett–Teller (BET) surface area analysis revealed values of 385 m² g⁻¹ for NUST-69 and 536 m² g⁻¹ for NUST-70. Pore size distribution analysis using nonlocal density functional theory (NLDFT) calculations yielded pore diameters of 2.05 nm and 1.80 nm for NUST-69 and NUST-70, respectively, in excellent agreement with the simulated structures (Fig. 2c and f). Morphological characterization was performed using scanning electron microscopy (SEM) and transmission electron microscopy (TEM) (Fig. 3, S7 and S8). NUST-69 displayed a nanorod morphology with surface-attached nanoparticles, while NUST-70 formed rectangular hollow nanotubes composed of interconnected particles, thereby exposing abundant active sites (Fig. 3a, b and S7). Energy-dispersive X-ray spectroscopy (EDS) mapping confirmed the homogeneous distribution of sulfur species throughout the NUST-70 framework (Fig. 3c). In the high-resolution TEM image, NUST-70 displayed pronounced lattice fringes with an interplanar spacing of 0.47 nm, which were indexed to the (120) crystallographic plane (Fig. 3f and S8c). This observation provides direct evidence for the high crystalline quality of NUST-70. Thermogravimetric analysis (TGA) under a N₂ atmosphere demonstrated exceptional thermal stability, with both COFs maintaining structural integrity up to 500 °C (Fig. S9). To evaluate alkaline stability, the COFs were immersed in 1 M and 12 M KOH solutions for 72 h, with PXRD analysis confirming complete retention of crystallinity (Fig. S10 and S11), highlighting their potential for practical applications in alkaline electrolysis systems.

Fig. 3 SEM images of (a) NUST-69 and (b) NUST-70, (c) elemental mapping images of NUST-70, TEM images of (d) NUST-69 and (e) NUST-70, and (f) HR-TEM images of NUST-70.

Electrocatalytic performance of the COFs

The electrocatalytic water splitting activity of the COF catalysts was systematically evaluated using linear sweep voltammetry (LSV) in O₂-saturated 1 M KOH electrolyte.^37,38 HER and OER measurements were conducted at a scan rate of 5 mV s⁻¹ within potential windows of −0.7 to 0.1 V and 1.0 to 2.2 V vs. RHE, respectively (Fig. 4a and c). The overpotential (η), a key indicator of catalytic efficiency representing the additional potential required to drive the faradaic process, was employed to quantify catalyst performance. Notably, the D-A structured NUST-70 demonstrated superior HER activity with an overpotential of 241 mV at 10 mA cm⁻², outperforming NUST-69 (η = 261 mV) (Fig. S12a). Both COFs exhibited remarkable OER performance, with NUST-70 achieving significantly lower overpotentials (292, 375, and 459 mV at 10, 50, and 100 mA cm⁻², respectively) compared to NUST-69 (331, 416, and 504 mV at corresponding current densities) (Fig. S12b). The enhanced bifunctional activity originates from the synergistic effects between the bicarbazole donor (N-doping source) and benzothiadiazole acceptor, which facilitates efficient charge transfer and provides catalytically active S species, particularly beneficial for the OER process.^10,11 Reaction kinetics were analyzed through Tafel slope measurements derived from LSV curves.³⁹ While both catalysts showed comparable HER kinetics (Tafel slopes of 205.4 and 224.2 mV dec⁻¹ for NUST-69 and NUST-70, respectively) (Fig. 4b), NUST-70 exhibited superior OER kinetics with a lower Tafel slope (63.6 mV dec⁻¹ vs. 65.5 mV dec⁻¹ for NUST-69) (Fig. 4d), indicating more favorable reaction pathways. These results demonstrate that NUST-70 ranks among the best-performing metal-free OER electrocatalysts reported to date (Fig. S18 and Table S3). Further investigation through multi-step chronoamperometry revealed that NUST-70 demonstrated enhanced mass transport characteristics during the OER, as evidenced by larger current responses to identical potential steps (Fig. S13b). This phenomenon underscores the crucial role of the benzothiadiazole-methyl group synergy in promoting both charge transfer and mass transport, ultimately leading to the observed exceptional electrocatalytic water splitting performance.

Fig. 4 (a) and (c) LSV plots of the HER/OER for NUST-69 and NUST-70. (b) and (d) Tafel plots obtained from HER and OER polarization curves.

The electrochemical active surface areas (ECSAs) of the COF catalysts were quantitatively assessed through double-layer capacitance (C_dl) measurements,⁴⁰ as illustrated in Fig. S14. The ECSAs of the two COFs were measured using cyclic voltammetry (CV). The double-layer capacitance values (C_dl) were obtained by fitting half of the current density as a function of scan rate (Fig. S14d). The C_dl values of NUST-69 and NUST-70 in alkaline electrolytes were 8.27 mF cm⁻² and 8.72 mF cm⁻², respectively. The higher C_dl value indicates that the electron acceptor (benzothiadiazole) exposes the COF catalyst to more N- and S-rich active sites, which enhances the catalytic activity. Electrochemical impedance spectroscopy (EIS)⁴¹ analysis (Fig. S14a) revealed lower interfacial charge transfer resistance for NUST-70 (1.90 Ω) compared to NUST-69 (2.14 Ω), demonstrating more efficient electron transfer kinetics and improved electrolyte adsorption/diffusion characteristics at the catalyst–electrolyte interface. Chronopotentiometry measurements conducted at a constant potential of 0.60 V vs. RHE confirmed excellent operational stability, with both catalysts maintaining stable performance over 48 hours of continuous operation in alkaline media (Fig. S15a). The comprehensive electrochemical characterization demonstrates that the bicarbazole-based COF materials exhibit outstanding bifunctional performance, encompassing superior catalytic activity, favorable reaction kinetics, and remarkable operational stability for both HER and OER processes. The superior electrochemical properties, coupled with the structural advantages of metal-free organic frameworks, position NUST-70 as a highly promising candidate for industrial-scale hydrogen production applications.

Density functional theory calculations

The OER reaction mechanisms of NUST-69 and NUST-70 were studied based on first principles calculation using ORCA software.^42,43 As shown in Fig. 5b, we estimated the lowest unoccupied molecular orbital (LUMO) and highest occupied molecular orbital (HOMO) levels of the two COFs. It is worth noting that the LUMO–HOMO band gap of NUST-70 is smaller (2.49 eV) due to the introduction of the electron acceptor benzothiadiazole, demonstrating its high charge transfer efficiency, which is consistent with electrochemical testing. The Ultraviolet-Visible-Near Infrared (UV-Vis-NIR) diffuse reflectance spectra further demonstrate that NUST-70 possesses a broader visible-light absorption range and a distinct red shift relative to NUST-69, indicative of superior charge transfer efficiency (Fig. 5a). Therefore, we speculate that there are six possible adsorption sites for OH⁻ on the surface of NUST-70 (Fig. 5c). Bader charge analysis of the atoms in NUST-70 indicates that the S atom in benzothiazole exhibits the highest positive charge due to its connection to two N atoms, which facilitates the adsorption of reactants, and thus S is the optimal active site for OER catalysis (Table S2). We calculated the adsorption and free energy change of intermediate adsorption on NUST-70 (Fig. S17). Fig. 5d shows the free energy distribution of the OER pathway at U = 0 and 1.23 V, and it is observed that the OOH* adsorption with the highest free energy value is the rate determining step in the NUST-70 OER process.

Fig. 5 (a) The UV-Vis-NIR spectra of NUST-69 and NUST-70, (b) Kohn–Sham LUMOs and HOMOs of model compounds of NUST-69 and NUST-70, (c) theoretically optimized structure of the NUST-70 monolayer (different C and S sites are marked), and (d) free-energy profile for the OER pathway at U = 0 and 1.23 V for the NUST-70.

Conclusions

In summary, we have successfully designed two bicarbazole-based 2D COF electrocatalysts with distinct electronic structures for efficient water splitting. These well-defined porous materials exhibit remarkable structural stability and demonstrate exceptional bifunctional HER/OER activity. Specifically, the benzothiadiazole-integrated NUST-70 COF exhibits a synergistic regulation of active-site density and charge transfer, thereby enhancing its OER performance and achieving a low overpotential of 292 mV. Combined experimental and DFT studies reveal that the improved catalytic efficiency originates from the reduced bandgap and lower activation energy barrier in NUST-70. This work not only presents a novel strategy for developing metal-free electrocatalysts but also highlights the promising potential of COFs in practical electrocatalytic applications.

Author contributions

J. W. conceived, designed, and performed the experiments, and wrote the manuscript. X. G. prepared the revised manuscript. Z. L. and C. C. provided experimental guidance. B. L. and K. Z. conducted DFT calculations and structural simulations. J. S. helped to revise the language of the manuscript. G. Z. directed the research.

Conflicts of interest

The authors declare no competing financial interest.

Data availability

The data supporting this article have been included as part of the SI.

CCDC 2472806 contains the supplementary crystallographic data for this paper.⁴⁴

Supplementary information is available: Experimental procedures and characterization data. See DOI: https://doi.org/10.1039/d5ta04563k.

Acknowledgements

This work was financially supported by the Scientific Research Innovation Capability Support Project for Young Faculty (ZYGXQNJSKYCXNLZCXM-M15), the National Natural Science Foundation of China (22171136 and 22301135), the Natural Science Foundation of Jiangsu Province (BK20220928 and BK20220079), the Medical Innovation and Development Project of Lanzhou University (lzuyxcx-2022-156), and CAMS Innovation Fund for Medical Sciences (CIFMS, 2019-I2M-5-074, 2021-I2M-1-026, 2021-I2M-3-001, 2022-I2M-2-002).

References

1. L. Quan, H. Jiang, G. Mei, Y. Sun and B. You, Chem. Rev., 2024, 124, 3694–3812.
2. M. N. Lakhan, A. Hanan, A. Hussain, I. Ali Soomro, Y. Wang, M. Ahmed, U. Aftab, H. Sun and H. Arandiyan, Chem. Commun., 2024, 60, 5104–5135.
3. Q. Zhai, H. Huang, T. Lawson, Z. Xia, P. Giusto, M. Antonietti, M. Jaroniec, M. Chhowalla, J. B. Baek, Y. Liu, S. Qiao and L. Dai, Adv. Mater., 2024, 36, 2405664.
4. X. Cui, M. Wu, X. Liu, B. He, Y. Zhu, Y. Jiang and Y. Yang, Chem. Soc. Rev., 2024, 53, 1447–1494.
5. K. Qu, Y. Zheng, Y. Jiao, X. Zhang, S. Dai and S. Z. Qiao, Adv. Energy Mater., 2017, 7, 1602068.
6. X. Cui, L. Gao, R. Ma, Z. Wei, C.-H. Lu, Z. Li and Y. Yang, J. Mater. Chem. A, 2021, 9, 20985–21004.
7. H. Luo and J. Liu, Angew. Chem., Int. Ed., 2024, 63, e202410759.
8. C. Hu, R. Paul, Q. Dai and L. Dai, Chem. Soc. Rev., 2021, 50, 11785–11843.
9. N. Gupta, S. Kanungo, R. P. Behere, P. Singh, S. Kanungo, M. Dixit, C. Chakraborty and B. K. Kuila, ACS Appl. Mater. Interfaces, 2023, 15, 29042–29051.
10. Y. Zhao, N. Yang, H. Yao, D. Liu, L. Song, J. Zhu, S. Li, L. Gu, K. Lin and D. Wang, J. Am. Chem. Soc., 2019, 141, 7240–7244.
11. S. Mathew, K.-H. Park, Y. Han, K. N. Hui, O. L. Li and Y.-R. Cho, ACS Appl. Mater. Interfaces, 2023, 15, 46829–46839.
12. K. Geng, T. He, R. Liu, S. Dalapati, K. T. Tan, Z. Li, S. Tao, Y. Gong, Q. Jiang and D. Jiang, Chem. Rev., 2020, 120, 8814–8933.
13. X. Zhao, P. Pachfule and A. Thomas, Chem. Soc. Rev., 2021, 50, 6871–6913.
14. D. Blätte, F. Ortmann and T. Bein, J. Am. Chem. Soc., 2024, 146, 32161–32205.
15. Y. Xia, W. Zhang, S. Yang, L. Wang and G. Yu, Adv. Mater., 2023, 35, 2301190.
16. J. Li and A. C. Grimsdale, Chem. Soc. Rev., 2010, 39, 2399–2410.
17. P. Han and Y. Zhang, Adv. Mater., 2024, 36, 2405630.
18. Q. Chen and B.-H. Han, Macromol. Rapid Commun., 2018, 39, 1800040.
19. S. Kundu, B. Behera, A. Giri, N. Sahab and A. Patra, Chem. Commun., 2021, 57, 6875–6878.
20. A. Sadhukhan, A. Karmakar, K. Koner, S. Karak, R. K. Sharma, A. Roy, P. Sen, K. K. Dey, V. Mahalingam, B. Pathak, S. Kundu and R. Banerjee, Adv. Mater., 2024, 36, 2310938.
21. A. Trabolsi, G. Das, S. S. Roy, F. Abou Ibrahim, A. Merhi, H. N. Dirawi, F. Benyettou, A. K. Das, T. Prakasam, S. Varghese, S. K. Sharma, S. Kirmizialtin, R. Jagannathan, F. Gándara, S. Aouad, M. A. Olson, S. Kundu and B. R. Kaafarani, Angew. Chem., Int. Ed., 2025, 64, e202419836.
22. C. Rong, X. Shen, Y. Wang, L. Thomsen, T. Zhao, Y. Li, X. Lu, R. Amal and C. Zhao, Adv. Mater., 2022, 34, 2110103.
23. J. Hutter, M. Iannuzzi, F. Schiffmann and J. VandeVondele, Wiley Interdiscip. Rev.:Comput. Mol. Sci., 2014, 4, 15–25.
24. J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 1996, 77, 3865–3868.
25. S. Grimme, J. Comput. Chem., 2006, 27, 1787–1799.
26. J. VandeVondele and J. Hutter, J. Chem. Phys., 2007, 127, 114105.
27. J. VandeVondele, M. Krack, F. Mohamed, M. Parrinello, T. Chassaing and J. Hutter, Comput. Phys. Commun., 2005, 167, 103–128.
28. S. Goedecker, M. Teter and J. Hutter, Phys. Rev. B:Condens. Matter Mater. Phys., 1996, 54, 1703–1710.
29. C. Hartwigsen, S. Goedecker and J. Hutter, Phys. Rev. B:Condens. Matter Mater. Phys., 1998, 58, 3641–3662.
30. Z. You, B. Wang, Z. Zhao, Q. Zhang, W. Song, C. Zhang, X. Long and Y. Xia, Adv. Mater., 2023, 35, 2209129.
31. Z. Zhang, C. L. Chen, Y. A. Chen, Y. C. Wei, J. Su, H. Tian and P. T. Chou, Angew. Chem., Int. Ed., 2018, 57, 9880–9884.
32. X. Zhang, Z. Xiao, L. Jiao, H. Wu, Y.-X. Tan, J. Lin, D. Yuan and Y. Wang, Angew. Chem., Int. Ed., 2024, 63, e202408697.
33. J. Wang, Z. Lu, X. Gao, Z. Shan, L. Xu, B. Li, S. Zhang, Y. Fu, J. Su and G. Zhang, ACS Appl. Eng. Mater., 2025, 3, 1561–1570.
34. Z. Gu, Z. Shan, Y. Wang, J. Wang, T. Liu, Z. Yu, J. Su and G. Zhang, Chin. Chem. Lett., 2024, 35, 108356.
35. Z. Gu, J. Wang, Z. Shan, M. Miao, T. Liu, L. Song, G. Wang, X. Ju, J. Su and G. Zhang, J. Mater. Chem. A, 2022, 10, 17624–17632.
36. A. F. M. El-Mahdy, M.-Y. Lai and S.-W. Kuo, J. Mater. Chem. C, 2020, 8, 9520–9528.
37. S. Anantharaj, S. R. Ede, K. Karthick, S. S. Sankar, K. Sangeetha, P. E. Karthik and S. Kundu, Energy Environ. Sci., 2018, 11, 744–771.
38. C. C. L. McCrory, S. Jung, I. M. Ferrer, S. M. Chatman, J. C. Peters and T. F. Jaramillo, J. Am. Chem. Soc., 2015, 137, 4347–4357.
39. C. Costentin, G. Passard and J.-M. Saveant, J. Am. Chem. Soc., 2015, 137, 5461–5467.
40. Y. Yoon, B. Yan and Y. Surendranath, J. Am. Chem. Soc., 2018, 140, 2397–2400.
41. K. Park, B.-Y. Chang and S. Hwang, ACS Omega, 2019, 4, 19307–19313.
42. F. Neese, Wiley Interdiscip. Rev.:Comput. Mol. Sci., 2022, 12, e1606.
43. T. Lu and F. Chen, J. Comput. Chem., 2011, 33, 580–592.
44. J. Wang, X. Gao, B. Li, Z. Lu, C. Chen, K. Zhang, J. Su and G. Zhang, CCDC 2472806: Experimental Crystal Structure Determination, 2025 DOI:10.5517/ccdc.csd.cc2p04yt.

---

Footnote

† J. W. and X. G. contributed equally to this work.

---