Fluorinated covalent organic frameworks for visible-light driven CO₂ reduction

Abstract

The metal-free visible-light-driven CO₂ reduction reaction was achieved by using a fluorine-atom-modified COF, i.e. N3F4-COF, which exhibited over a 5-fold enhancement compared to the pristine COF for syngas production. The activity can be further improved by incorporating a Ru-based photosensitizer, and the syngas ratio could be regulated from 2.57 to 0.14 (CO/H₂) by altering the hydrophilicity of the photosensitizers.

---

Introduction

Carbon dioxide (CO₂) is a prominent greenhouse gas with significant implications for the climate and atmospheric temperatures.^1–3 Currently, the selective conversion of CO₂ into solar fuels and chemicals via photocatalysis is a crucial transformation for both the chemical industry and environmental sustainability.^4–9 The photocatalytic CO₂ reduction reaction (CO₂RR) to produce syngas, an essential industrial feedstock for the Fischer–Tropsch process and methanol synthesis, represents a promising yet challenging approach. This challenge arises from difficulties in controlling the selectivity of CO and H₂ and in enhancing the overall catalytic activity.^10–14 In general, the CO₂RR requires light absorption by a semiconductor to generate photoexcited electron–hole pairs, followed by the separation of these pairs and their subsequent migration to the semiconductor surface to catalyze redox reactions.^15–23 Therefore, developing a strategy to engineer photocatalysts with strong light absorption capabilities, efficient photogenerated charge separation and transport, and high CO₂ adsorption ability is highly desirable.^24–26

Covalent organic frameworks (COFs) are a class of crystalline materials with well-defined and periodic structures that allow for precise chemical design and versatile applications.^27–30 This design flexibility has been exploited in various architectures to enhance photocatalytic activity. For example, anchoring functional moieties to the backbones of the COF can regulate the electron push–pull effect to narrow the bandgap.^31–33 Moreover, designing a COF with a donor–acceptor (D–A) structure is a prevalent strategy to facilitate the separation and migration of photoexcited pairs.^34–37 However, only a few of the COFs can perform the photocatalytic CO₂RR under real metal-free conditions, considering the potential for trace metal residues from the synthesis process.^38–41

Herein, we demonstrate that fluorination is a versatile strategy for regulating COF morphology, narrowing the bandgap, facilitating the separation of photoexcited pairs, and improving CO₂ adsorption ability. Specifically, the fluorine atom-modified COF, i.e. N3F4-COF, demonstrates over a 5-fold enhancement in the metal-free photocatalytic CO₂RR compared to the pristine COF, i.e. N3-COF. Furthermore, activity enhancement and tunable CO/H₂ selectivity can be achieved by adding different Ru-based photosensitizers.

Results and discussion

The 2D COFs, N3-COF and N3F4-COF, were synthesized via metal-free Schiff base condensation, as illustrated in Scheme 1. N3F4-COF was prepared by reacting 2,3,5,6-tetrafluoroterephthalaldehyde (PTA-F4) with 2,4,6-tris(4-aminophenyl)-triazine (TAT) (see the ESI† for further details). The structure and synthesis of N3-COF, which serves as a comparative structure, have been reported in a previous work.⁴² CP/MAS 13C NMR of N3F4-COF (Fig. 1a) shows six peaks in the aromatic benzene ring region (δ = 117 to 138 ppm) and signals corresponding to imine functionalities at ∼154 ppm. Similar chemical shifts are observed for N3-COF (Fig. S1†). Fourier-transform infrared (FTIR) spectra of the materials reveal characteristic stretching bands (Fig. 1b and S2†), with bands at 1579 cm⁻¹ and 1503 cm⁻¹ corresponding to the C=N Schiff base bonds and aromatic triazine C=N bonds, respectively. The band at 1020 cm⁻¹ corresponds to C–F bonds, providing evidence for the incorporation of fluorine atoms into the N3F4-COF structure. The aldehyde band at 1708 cm⁻¹, associated with unreacted terminal groups, is significantly reduced or absent in N3F4-COF, indicating a higher degree of polymerization. Powder X-ray diffraction (PXRD) analysis shows good crystallinity for both COFs (Fig. 1c and S3†). Compared to that in N3-COF, the X-ray photoelectron spectroscopy (XPS) spectrum of C 1s in N3F4-COF displays additional C–F peaks, further indicating the presence of the F atom (Fig. S4b and S5d†). Contact-angle measurement (Fig. 1d and S7†) reveals that both COFs have a contact angle greater than 120°, highlighting their hydrophobic nature.

Scheme 1 Design and synthesis of modified covalent organic frameworks through Schiff base condensation.

Fig. 1 Structural and hydrophobicity characterization. (a) CP/MAS ¹³C NMR spectrum of N3F4-COF. (b) FT-IR spectra of monomers and N3F4-COF. (c) PXRD pattern of N3F4-COF. (d) Contact angle measurement of N3F4-COF.

Scanning electron microscopy (SEM) images (Fig. 2a and b) show that N3F4-COF has a uniform morphology resembling sea anemones, with “tentacles” measuring 100–200 nm in length. In contrast, N3-COF has a smaller particle size (∼1 μm) and an agglomerate structure (Fig. S5†). The morphology change in N3F4-COF is attributed to the introduction of fluorine atoms. Energy dispersive spectroscopy (EDS) mapping images in Fig. 2c show a uniform distribution of C, N, and F, indicating the high elemental uniformity on N3F4-COF. Transmission electron microscopy (TEM) further confirms a uniform surface and good crystallinity (Fig. S9†), attributable to successful condensation.

Fig. 2 Morphological characterization of N3F4-COF. (a) and (b) SEM images, and (c) elemental mapping images.

Surface area analysis using the Brunauer–Emmett–Teller (BET) method (Table 1 and Fig. 3) reveals a significantly larger specific surface area for N3F4-COF (4801.8 m² g⁻¹) compared to N3-COF (253.8 m² g⁻¹). The isotherms (Fig. 3a and b) for both COFs exhibit type I characteristics, indicative of microporous structures with saturation adsorption corresponding to micropore filling. CO₂ uptake capacities (Table 1 and Fig. S10†) show that N3F4-COF has higher CO₂ adsorption abilities than N3-COF at both 273 K and 298 K. Then, using the Clausius–Clapeyron relationship, the isosteric heat of adsorption (Q_st) was calculated from the 273 and 298 K isotherms, with Q_st values of 8.9 kJ mol⁻¹ for N3-COF and 16.2 kJ mol⁻¹ for N3F4-COF. The higher CO₂ affinity of N3F4-COF is attributed to the interaction between fluorine atoms and CO₂, indicating that N3F4-COF, with its optimal electron-withdrawing group content, maximizes the reactivity.⁴³

Table 1 Porosity properties and CO₂ uptake capacities

Polymers	S BET (m^2 g^−1)	CO2 uptake^b (cm^3 g^−1)		Q st (kJ mol^−1)
		273 K	298 K
a Specific surface area calculated by using the BET method. b CO2 uptake capacities at 1 bar. c Isosteric heat of adsorption for CO2.
N3F4-COF	4801.8	9.9	6.5	8.9
N3-COF	253.8	17.2	10.0	16.2

---

Fig. 3 N₂ adsorption–desorption isotherms collected at 77 K for (a) N3F4-COF and (b) N3-COF, respectively; pore size distribution for (c) N3F4-COF and (d) N3-COF, respectively.

UV-vis diffuse reflectance spectra (Fig. 4a) demonstrate that N3F4-COF, enriched with fluorine atoms, exhibits enhanced visible-light absorption ability compared to N3-COF. The bandgaps (E_g) calculated from the Tauc plots are 2.3 eV and 2.6 eV for N3F4-COF and N3-COF, respectively. The conduction band minimum (CBM) potentials derived from the Mott–Schottky plots are −1.06 V and −0.91 V (vs. NHE) for N3F4-COF and N3-COF, respectively (Fig. 4c). The energy band structure diagram (Fig. 4d) indicates that both COFs are suitable for visible-light absorption in the photocatalytic reduction of CO₂ and proton to CO and H₂, respectively. Clearly, the substitution of H atoms with fluorine atoms enhances visible absorbance, resulting in changes to the composition and band edge.⁴⁴

Fig. 4 (a) UV/vis DR spectra and (b) schematic band structure diagram of N3-COF and N3F4-COF; Mott–Schottky plots of (c) N3F4-COF and (d) N3-COF.

Photogenerated charge separation was further investigated through transient photocurrent examined in the presence and absence of light. As illustrated in Fig. 5, the photocurrent density of N3F4-COF is higher than that of N3-COF, suggesting more efficient separation of photogenerated carriers in N3F4-COF. This observation aligns with electrochemical impedance spectroscopy (EIS) measurements, which show that N3F4-COF displays a smaller semicircle in the Nyquist plots, indicating low charge transfer resistance. This enhanced mobility is attributed to the higher fluorine content in N3F4-COF, which form a D (PTA-F4 linker)-A (TAT unit) structure that facilitate the separation and migration of photoexcited carriers.

Fig. 5 (a) Nyquist plots and (b) transient photocurrent response of N3-COF and N3F4-COF.

Collectively, fluorination significantly impacts the morphologies of COFs and improves CO₂ adsorption ability. Additionally, the narrowed bandgap and enhanced separation and migration of photoexcited pairs are achieved through anchoring fluorine atoms into the COF structure.

Photocatalytic experiments were conducted using N3-COF and N3F4-COF as catalysts in 3 mL acetonitrile solution containing 0.5 mL triethanolamine (TEOA), 1 mL water as a proton source, and 0.05 mM BIH as the electron donor, under a CO₂ atmosphere (1.0 atm, 298 K). It should be mentioned that only CO and H₂ were detected as products, and no HCOOH or CH₃OH was observed in the liquid phase by ¹H NMR. N3F4-COF demonstrated a maximum activity of 440 μmol_CO g⁻¹ and 115 μmol_H2 g⁻¹ after 4 h of irradiation with a xenon lamp (300 W, λ ≥ 420 nm) (Fig. 6a). In contrast, N3-COF showed poor activity for producing either CO or H₂. This high activity under metal-free conditions for N3F4-COF should be ascribed to the introduction of fluorine atoms.

Fig. 6 (a) Catalytic performances over N3F4-COF and N3-COF; (b) control experiments by using N3F4-COF in 2 h CO₂ photoreduction; (c) time curve of CO/H₂ formation; (d) durability measurements of N3F4-COF; and [Ru(phen)₃](PF₆)₂ was added after 4th cycle.

As shown in Fig. 6b and c, the addition of [Ru(phen)₃](PF₆)₂ as a photosensitizer enhanced the CO₂RR activity for N3F4-COF, achieving 9400 μmol g⁻¹ CO and 3650 μmol g⁻¹ H₂. This performance is comparable to other catalyst systems used for the photocatalytic CO₂RR to produce syngas (Tables S1 and S2†). In contrast, [Ru(phen)₃](PF₆)₂ alone exhibited a much lower catalytic activity. Control experiments, depicted in Fig. 6b, confirmed that no CO was formed in a N₂ atmosphere, proving that CO originates from CO₂. Furthermore, control exclusion experiments (excluding the light source, water, or electron donors) or replacing the photosensitizer with dyes⁴⁵ (e.g., eosin Y or methylene blue) showed no CO₂RR activity.

Interestingly, when a hydrophobic photosensitizer (Ru(bpy)₃Cl₂·6H₂O) was introduced, N3F4-COF exhibited 88% selectivity towards H₂, while the hydrophilic photosensitizer ([Ru(phen)₃](PF₆)₂) resulted in 72% selectivity towards CO. This unexpected result can be attributed to two factors: (1) the hydrophobic nature of N3F4-COF critically enhances CO₂RR performance by increasing the interfacial CO₂ concentration. Molecular dynamics simulations revealed that the hydrophobic microenvironment prevents hydrogen bond formation on the carbonaceous co-catalyst surface.⁴⁶ (2) A hydrophilic photosensitizer, which dissolves easily in the aqueous phase, may contribute to the formation of “oil-photocatalyst-water” layers,^47,48 potentially alleviating CO₂ solubility limitations and sluggish diffusion in water, thereby accelerating CO₂-to-CO conversion while reducing HER rates.

Then, the long-term stability and reusability of N3F4-COF were also evaluated (Fig. 6d). After 4 h, the photocatalytic activity began to decline, and syngas production became nearly negligible after 8 h. However, the CO/H₂ selectivity remained stable for at least 20 h, indicating that the decomposition process does not significantly affect the CO and H₂ production ratio. To confirm the reason for the decreased activity, the system was evacuated and CO₂ gas was refilled every 2 h. The photocatalytic activity decreased after the 4th cycle but could be restored by adding [Ru(phen)₃](PF₆)₂, suggesting that the decreased activity originated from the decomposition of the photosensitizer. In addition, no significant changes were observed in the XPS spectra of the reused N3F4-COF (Fig. S6†).

Based on the above results, a proposed mechanism for N3F4-COF is shown in Fig. 7a. Upon visible-light excitation, photogenerated electrons transfer from the valence band maximum (VBM) (PTA-F4 linker) to the CBM (triazine unit), facilitating surface reduction reactions. Moreover, [Ru(phen)₃]^2+* is generated under visible light irradiation, then the photogenerated electron is transferred to the triazine unit on the N3F4-COF catalyst. The holes in [Ru(phen)₃]²⁺ and N3F4-COF are consumed by the electron donor, i.e. BIH. Furthermore, in the three-phase interface photocatalysis shown in Fig. 7b, the hydrophilic [Ru(phen)₃](PF₆)₂ dissolves in the aqueous layer, while electronic sacrificial agents (i.e. BIH and TEOA) dissolve in the organic layer, with the hydrophobic N3F4-COF dispersed between these layers. This setup demonstrated that the hydrophobic catalyst plays a critical role in regulating CO₂RR performance by increasing the interfacial CO₂ concentration and alleviating CO₂ mass transfer limitations in the aqueous layer. In short, efficient electron transfer and CO₂ concentration in the aqueous phase are key factors controlling the CO₂RR and HER activity. Higher electron transfer efficiency promotes CO formation, while a hydrophobic microenvironment prevents hydrogen formation.

Fig. 7 (a) Schematic illustration of the charge-transfer process under light irradiation. (b) Possible mechanism of the three-phase photocatalysis for CO₂ reduction over N3F4-COF and inside proposed photoactive electron transfer and reaction pathway in the photoreduction of CO₂.

Conclusions

In summary, fluorine-atom-modified COFs, i.e. N3F4-COF, have been successfully designed and developed as efficient and durable photocatalysts for the CO₂RR to yield syngas. The introduction of fluorine atoms enables regulation of morphology and improves surface area, CO₂ adsorption capacity, visible light absorption, and photoinduced electron–hole separation efficiency. Consequently, N3F4-COF exhibits significantly higher activity in the metal-free photocatalytic CO₂RR compared to N3-COF. Additionally, the activity can be further enhanced by incorporating a Ru-based photosensitizer, and the selectivity between CO and H₂ can be easily tuned by altering the hydrophilicity of the photosensitizers. These findings provide a foundation for designing novel catalysts that could advance CO₂ reduction technologies.

Data availability

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

Author contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

We appreciate the General Research Projects of the Zhejiang Provincial Education Department (Y202454316), the National Natural Science Foundation of China (No. 22005154), and the AI & AI for Science Project of Nanjing University (Fundamental Research Funds for the Central Universities, No. 14380006) for the financial support.

References

1. K. Kleijne, M. A. J. Huijbregts, F. Knobloch, R. Zelm, J. P. Hilbers, H. Coninck and S. V. Hanssen, Nat. Energy, 2024, 9, 1139–1152.
2. M. Meinshausen, N. Meinshausen, W. Hare, S. C. B. Raper, K. Frieler, R. Knutti, D. J. Frame and M. R. Allen, Nature, 2009, 458, 1158–1162.
3. T. R. Karl and K. E. Trenberth, Science, 2003, 302, 1719–1723.
4. X. Ding, W. Liu, J. Zhao, L. Wang and Z. Zou, Adv. Mater., 2024, 2312093.
5. J. Yu, A. Muhetaer, Q. Li and D. Xu, Small, 2024, 20, 2402952.
6. M. R. Hoffmann, S. T. Martin, W. Y. Choi and D. W. Bahnemann, Chem. Rev., 1995, 95, 69–96.
7. F.-Y. Ren, K. Chen, L.-Q. Qiu, J.-M. Chen, D. J. Darensbourg and L.-N. He, Angew. Chem., Int. Ed., 2022, 61, e202200751.
8. X. C. Wang, K. Maeda, A. Thomas, K. Takanabe, G. Xin, J. M. Carlsson, K. Domen and M. Antonietti, Nat. Mater., 2009, 8, 76–80.
9. S. Qi, K. Zhu, T. Xu, H. Zhang, X. Guo, J. Wang, F. Zhang and X. Zong, Adv. Mater., 2024, 36, 2403328.
10. Y.-N. Gong, S. Wang, H.-J. Dong, J.-H. Mei, D.-C. Zhong and T.-B. Lu, Appl. Catal., B, 2024, 357, 124310.
11. Z. B. Fang, T. T. Liu, J. X. Liu, S. Y. Jin, X. P. Wu, X. Q. Gong, K. C. Wang, Q. Yin, T. F. Liu, R. Cao and H. C. Zhou, J. Am. Chem. Soc., 2020, 142, 12515–12523.
12. X. B. Chen, L. Liu, P. Y. Yu and S. S. Mao, Science, 2011, 331, 746–750.
13. J. X. Fan, X. X. Yue, Y. N. Liu, D. Q. Li and J. T. Feng, Chem Catal., 2022, 2, 531–549.
14. W.-J. Wang, K.-H. Chen, Z.-W. Yang, B.-W. Peng and L.-N. He, J. Mater. Chem. A, 2021, 9, 16699–16705.
15. A. L. Linsebigler, G. Lu and J. T. Yates Jr., Chem. Rev., 1995, 95, 735–758.
16. Y.-N. Gong, S.-Q. Zhao, H.-J. Wang, Z.-M. Ge, C. Liao, K.-Y. Tao, D.-C. Zhong, S. Ken and T.-B. Lu, Angew. Chem., Int. Ed., 2024, e202411639.
17. W. Lin, J. Lin, X. Zhang, L. Zhang, R. A. Borse and Y. Wang, J. Am. Chem. Soc., 2023, 145, 18141–18147.
18. S. Yoshino, T. Takayama, Y. Yamaguchi, A. Iwase and A. Kudo, Acc. Chem. Res., 2022, 55, 966–977.
19. H. Tada, M. Fujishima and H. Kobayashi, Chem. Soc. Rev., 2011, 40, 4232–4243.
20. Y.-N. Gong, X. Guan and H.-L. Jiang, Coord. Chem. Rev., 2023, 475, 214889.
21. Y. Yang, H.-Y. Zhang, Y. Wang, L.-H. Shao, L. Fang, H. Dong, M. Lu, L.-Z. Dong, Y.-Q. Lan and F.-M. Zhang, Adv. Mater., 2023, 35, 2304170.
22. H. Li, R. Li, G. Liu, M. Zhai and J. Yu, Adv. Mater., 2023, 35, 2301307.
23. Q. Zhang, S. Gao, Y. Guo, J. Huiyong Wang, X. Wei, H. Su, Z. Zhang and J. Wang Liu, Nat. Commun., 2023, 14, 1147.
24. J. Li, G. Chen, Y. Zhu, Z. Liang, A. Pei, C.-L. Wu, H. Wang, H. R. Lee, K. Liu, S. Chu and Y. Cui, Nat. Catal., 2018, 1, 592–600.
25. J. Li, J. Zhou, X.-H. Wang, C. Guo, R.-H. Li, H. Zhuang, W. Feng, Y. Hua and Y.-Q. Lan, Angew. Chem., Int. Ed., 2024, e202411721.
26. T. H. M. Pham, J. Zhang, M. Li, T. H. Shen, Y. Ko, V. Tileli and W. Luo, Adv. Energy Mater., 2022, 12, 2103663.
27. T. He, Z. Zhao, R. Liu, X. Liu, B. Ni, Y. Wei, Y. Wu, W. Yuan, H. Peng, Z. Jiang and Y. Zhao, J. Am. Chem. Soc., 2023, 145, 6057–6066.
28. W. Zhang, L. Chen, S. Dai, C. Zhao, C. Ma, L. Wei, M. Zhu, S. Y. Chong, H. Yang, L. Liu, Y. Bai, M. Yu, Y. Xu, X.-W. Zhu, Q. Zhu, S. An, R. S. Sprick, M. A. Little, X. Wu, S. Jiang, Y. Wu, Y.-B. Zhang, H. Tian, W.-H. Zhu and A. I. Cooper, Nature, 2022, 604, 72–79.
29. C. Gropp, T. Ma, N. Hanikel and O. M. Yaghi, Science, 2020, 370, eabd6406.
30. K. Sun, O. J. Silveira, Y. Ma, Y. Hasegawa, M. Matsumoto, S. Kera, O. Krejčí, A. S. Foster and S. Kawai, Nat. Chem., 2023, 15, 136–142.
31. C. Li, J. Liu, H. Li, K. Wu, J. Wang and Q. Yang, Nat. Commun., 2022, 13, 2357.
32. K. Wang, Z. Jia, X. Bai, S. E. Hodgkiss, L. Chen, S. Y. Chong, X. Wang, H. Yang, Y. Xu, F. Feng, J. W. Ward and A. I. Cooper, J. Am. Chem. Soc., 2020, 142, 11131–11138.
33. L. Hao, R. Shen, C. Huang, Z. Liang, N. Li, P. Zhang, X. Li, C. Qin and X. Li, Appl. Catal., B, 2023, 330, 122581.
34. J. Yang, S. Ghosh, J. Roeser, A. Acharjya, C. Penschke, Y. Tsutsui, J. Rabeah, T. Wang, S. Y. D. Tameu, M.-Y. Ye, J. Grüneberg, S. Li, C. Li, R. Schomäcker, R. V. D. Krol, S. Seki, P. Saalfrank and A. Thomas, Nat. Commun., 2022, 13, 6317.
35. S. Barman, A. Singh, F. A. Rahimi and T. K. Maji, J. Am. Chem. Soc., 2021, 143, 16284–16292.
36. M. Zhang, P. Huang, J.-P. Liao, M.-Y. Yang, S.-B. Zhang, Y.-F. Liu, M. Lu, S.-L. Li, Y.-P. Cai and Y.-Q. Lan, Angew. Chem., Int. Ed., 2023, 62, e202311999.
37. Y. Wang, W. Hao, H. Liu, R. Chen, Q. Pan, Z. Li and Y. Zhao, Nat. Commun., 2022, 13, 100.
38. J. Yang, Z. Chen, L. Zhang and Q. Zhang, ACS Nano, 2024, 18, 21804–21835.
39. L. Qin, C. Ma, J. Zhang and T. Zhou, Adv. Funct. Mater., 2024, 34, 2401562.
40. S. Biswas, F. A. Rahimi, R. K. Saravanan, A. Dey, J. Chauhan, D. Surendran, S. Nath and T. K. Maji, Chem. Sci., 2024, 15, 16259–16270.
41. F. Meng, J. Wang, M. Chen, Z. Wang, G. Bai and X. Lan, ACS Catal., 2023, 13, 12142–12152.
42. D. Mullangi, S. Shalini, S. Nandi and B. Choksi, J. Mater. Chem. A, 2017, 5, 8376–8384.
43. C. S. Diercks, S. Lin, N. Kornienko, E. A. Kapustin, E. M. Nichols, C. Zhu, Y. Zhao, C. J. Chang and O. M. Yaghi, J. Am. Chem. Soc., 2018, 140, 1116–1122.
44. T. Li, N. Tsubaki and Z. Jin, J. Mater. Sci. Technol., 2024, 169, 82–104.
45. Z. Zhou, H. Yao, Y. Wu, T. Li, N. Tsubaki and Z. Jin, Acta Phys.-Chim. Sin., 2024, 40, 2312010.
46. X. B. Chen, S. H. Shen, L. J. Guo and S. S. Mao, Chem. Rev., 2010, 110, 6503–6570.
47. Q.-S. Huang, C. Chu, Q. Li, Q. Liu, X. Liu, J. Sun, B.-J. Ni and S. Mao, ACS Catal., 2023, 13, 11232–11243.
48. Y. Li, Z. Pei, D. Luan and X. W. D. Lou, J. Am. Chem. Soc., 2024, 146, 3343–3351.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4cy01276c

---