Non-covalent hybridization of phthalocyanine onto polymeric carbon nitride, enabling visible-light CO₂ conversion

Abstract

Cationic imidazolium-functionalized metal phthalocyanine (MPc-Im; M = Fe, Co) was immobilized onto the surface of polymeric carbon nitride (PCN), and the MPc-Im/PCN hybrids were examined as photocatalysts for visible-light CO₂ reduction in a non-aqueous environment. The hybrids produced CO as the main product in the presence of triethanolamine as an electron donor, along with CH₄ formation, as confirmed by tracer experiments with ¹³CO₂.

---

The conversion of carbon dioxide (CO₂) into solar fuels presents a promising strategy to mitigate the global climate crisis driven by excessive greenhouse gas emissions, particularly CO₂.¹ However, this process remains highly challenging due to the thermodynamic stability of the CO₂ molecule, which requires substantial energy input for activation and reduction. Among various CO₂ conversion technologies—such as thermochemical and electrochemical methods—the photocatalytic reduction of CO₂, inspired by the principles of natural photosynthesis, stands out as an attractive strategy.^2,3 Photocatalytic CO₂ reduction allows the direct utilization of solar energy under mild conditions, such as room temperature and atmospheric pressure, offering a sustainable and energy-efficient approach to solar fuel production. CO₂ reduction occurs through multi-proton/electron transfer reactions and is governed by several key processes, including light absorption, charge transfer, redox reactions, and product desorption. Although the efficiency of a photocatalytic system is influenced by a complex interplay of various factors, the photocatalyst remains the most crucial component.

In recent years, hybrid photocatalysts—developed by immobilizing molecular catalysts onto photoresponsive semiconductor supports through diverse anchoring strategies—have emerged as a rational design for efficient CO₂ conversion.^2,4,5 These systems integrate the high reactivity and selectivity of molecular catalysts with the durability of heterogeneous semiconductor materials, leading to enhanced redox reaction efficiency, improved stability, and effective recovery/recyclability of photocatalysts.⁶

Phthalocyanines (Pcs) and related compounds play a crucial role in the development of artificial photosynthetic systems, such as photocatalytic CO₂ reduction and H₂ production, by mimicking the light-harvesting and charge-separation functions essential for photoinduced charge separation and electron transfer events.⁷ Their excellent thermal and photochemical stability, strong absorption in the visible and near-infrared regions, and remarkable synthetic versatility make Pcs ideal components for constructing composite catalysts in photocatalytic reactions. As a result, Pc derivatives have been widely used in the photocatalytic and electrochemical reduction of CO₂, primarily as catalysts and/or photosensitizers. In particular, metal phthalocyanines (MPcs) incorporating transition metals such as Fe, Co, and Ni exhibit superior activity in CO₂ reduction, not only due to their excellent photophysical and chemical properties, but also because of their ability to serve as active centers for multi-electron reactions and provide open coordination sites for CO₂ binding. To date, Pc derivatives have been immobilized onto a wide range of organic and/or inorganic semiconductor materials to fabricate highly efficient hybrid photocatalysts for photo/electrochemical CO₂ reduction and H₂ generation.⁷

Among the semiconductors applicable to the molecule/semiconductor hybrids, polymeric carbon nitride (PCN) has attracted significant interest in recent years as a versatile photocatalytic material owing to a band gap of approximately 2.7 eV that enables visible light absorption, a relatively high surface area that provides more active sites, reduced rates of charge carrier recombination, and enhanced charge mobility due to its extended π-conjugated structure.^8–10 Apart from covalent interactions, MPcs can be immobilized onto PCN through non-covalent interactions such as π–π stacking and electrostatic forces, which help preserve the intrinsic physical and chemical properties of PCN. Although MPcs/PCN hybrid photocatalysts have been widely used for photocatalytic H₂ production, involvement in photocatalytic CO₂ reduction remains limited, with only a few examples reported to date.¹¹ For example, Reisner et al. developed a hybrid photocatalyst comprising mesoporous PCN as the photosensitizer and polymeric cobalt phthalocyanine (CoPPc) as the catalyst for photocatalytic CO₂ reduction.¹² The hybrid system was constructed by in situ polymerization of CoPc onto the PCN, enabling efficient photoinduced electron transfer from the PCN support to the CoPPc catalyst. This hybrid catalyst selectively converts CO₂ to CO with a cobalt-based turnover number (TON) of 90 over 60 hours of irradiation. Although FePcs and CoPcs, among other MPc derivatives, have been extensively explored for electrocatalytic CO₂ reduction,¹³ comparatively less effort has been devoted to developing Pc-based hybrid catalysts that use Pcs as catalytic centers for photocatalytic CO₂ reduction. To address this gap, we focused on designing Pc-based hybrid systems that utilize Pcs as catalytic centers.

Herein, we report a simple and efficient self-assembled photocatalytic system composed of PCN as the photo-absorber and tetracationic, non-noble MPc-Im (M = Fe, Co; Chart 1) as catalysts for the photocatalytic reduction of CO₂. To enhance water solubility and facilitate interaction with the PCN surface, the Fe and CoPcs were functionalized with cationic imidazolium rings, which can provide π–π stacking and electrostatic interactions with the semiconductor support.^14,15 The detail of the synthesis is included in ESI† (see Fig. S1).

Chart 1 Chemical structures of FePc-Im and CoPc-Im.

To ensure the catalytic activity of the MPc-Im for CO₂ reduction, cyclic voltammetry (CV) measurements were conducted. Fig. S2† shows CV curves of the MPc-Im in a N,N-dimethylacetamide (DMA)/TEOA (triethanolamine) mixture containing Et₄NBF₄ (0.1 M) as a supporting electrolyte, measured under an Ar and CO₂ atmosphere. Plain, unfunctionalized Pcs (e.g., FePc and CoPc), which are well-known catalysts for CO₂ reduction, were also tested for comparison. For FePc-Im, the first reduction started to occur at around −1.2–−1.5 V vs. Ag/AgNO₃, which was assigned to the Fe(II)-to-Fe(I) reduction. A clear catalytic current, assignable to CO₂ reduction, was observed at ca. −1.7 V. CoPc-Im exhibited a similar CV profile, with the onset potentials for the Co(II)-to-Co(I) reduction and CO₂ reduction of ca. −1.8 V. Note that the onset potentials of the first reduction wave for FePc-Im and CoPc-Im were more positive than those for plain FePc and CoPc, meaning that FePc-Im and CoPc-Im were more susceptible to reduction than FePc and CoPc. The positive shift in reduction onset potentials likely arises from the electron-withdrawing imidazolium group in the Pc-Im ligand, which stabilizes the reduced metal center and facilitates reduction at more positive potentials. Thus, FePc-Im and CoPc-Im were shown to have CO₂ reduction catalytic activity. The conduction band minimum of PCN prepared from urea has been reported to be −2.0 ± 0.1 V vs. Ag/AgNO₃,⁴ which is more negative than the reduction potentials of FePc-Im and CoPc-Im. So, the electron transfer from PCN to MPc-Im is energetically favorable.

The MPc-Im were non-covalently immobilized onto PCN through sonication-assisted deposition, facilitating effective interaction without altering the structural integrity of either component. PCN was synthesized via a facile thermal polymerization of urea, following a previously published procedure.¹⁶ The specific surface area of the as-prepared PCN was determined to be 47 m² g⁻¹ by means of N₂ gas adsorption at 77 K. The details of the synthesis of PCN and the structural characterization are included in ESI† (Fig. S3). Because the zeta-potential of PCN was measured to be −23.1 mV at pH ∼7, electrostatic attraction between positively charged MPc-Im and the PCN surface can be expected. Actually, no adsorption occurred when plain, neutral FePc and CoPc were used. To investigate the effect of MPc-Im catalyst concentration on CO₂ reduction, we prepared hybrid photocatalysts with varying amounts of MPc-Im. The MPc-Im/PCN heterogeneous catalysts were prepared by mixing 100 mg of PCN with different concentrations of MPc-Im (1.0 × 10⁻⁵ M, 1.0 × 10⁻⁴ M and 3.3 × 10⁻⁴ M) in 10 mL of water, followed by overnight stirring in the dark. Further details about the synthesis and characterization of CoPc-Im/FePc-Im and photocatalysts can be found in the ESI† (see Fig. S4–S9).

The absorption spectra of both MPc-Im (Fig. 1a) show broadened Q bands with a maximum at 665 nm; however, FePc-Im exhibits Q-band splitting, likely due to some type of molecular aggregation in the aqueous solution.¹⁷ Following the successful hybridization of PCN with FePc-Im, the resulting photocatalysts were characterized using UV-vis absorption and FT-IR spectroscopy. Fig. 1b presents the UV-vis diffuse reflectance spectra (DRS) of pristine PCN and MPc-Im/PCN hybrids. Both MPc-Im/PCN hybrids show broadened absorption in the visible region compared to pristine MPc-Im, along with a combined absorption feature from both PCN and MPc-Im, indicating the successful immobilization of MPc-Im onto the surface of PCN.

Fig. 1 a) UV-vis absorption spectra of FePc-Im and CoPc-Im in aqueous solution (3.3 × 10⁻⁴ M). b) UV-vis diffuse reflectance spectra of pristine PCN and MPc-Im-adsorbed PCN (FePc-Im, 32.8 μmol g⁻¹; CoPc-Im, 24.3 μmol g⁻¹).

The amount of MPc-Im adsorbed onto PCN was determined by calculating the absorbance difference between the initial MPc-Im solution and the unadsorbed MPc-Im filtered from the MPc-Im/PCN suspension. The MPc-Im loading on the PCN surface was determined to be 32.8 μmol g⁻¹ (4.9 wt%) for FePc-Im and 24.3 μmol g⁻¹ (3.7 wt%) for CoPc-Im, respectively, when a 3.3 × 10⁻⁴ M Pc solution was used in the preparation of the MPc-Im/PCN photocatalysts. Additional MPc-Im/PCN hybrids were synthesized with lower MPc-Im contents, specifically 0.67 wt% and 0.15 wt%. Further details are provided in Table S1.† FT-IR measurements were conducted to further investigate the adsorption states of MPc-Im on the PCN surface. Fig. S7–S9† display the FT-IR spectra of both MPc-Im powders and MPc-Im/PCN hybrids. In the spectra of MPc-Im powders, distinct bands associated with C=N bending and C=C stretching vibrations were observed within the range of 1340–1740 cm⁻¹.¹⁸ For pristine PCN, a characteristic vibrational peak appeared at 808 cm⁻¹, corresponding to the triazine ring. Additionally, a broad absorption band in the range of 3000–3700 cm⁻¹ was observed, attributed to stretching vibrations of hydroxyl groups of surface adsorbed water molecules and N–H stretching vibrations, along with peaks at 1454, 1313, and 1226 cm⁻¹, which are ascribed to aromatic C–N stretching.¹⁹ Upon hybridization with MPc-Im, the FT-IR spectra of the MPc-Im/PCN composites retained nearly all major peaks of pure PCN, suggesting that its structural framework remained largely intact.^14,20 However, in the MPc-Im/PCN hybrids, slight shifts toward higher wavenumbers were observed in several characteristic PCN peaks around 3000–3700 cm⁻¹, 1228–1450 cm⁻¹, and 808 cm⁻¹. These shifts suggest an interaction between PCN and MPc-Im.²¹ Furthermore, the overlap of some characteristic MPc-Im peaks with the strong absorption bands of PCN suggests that the hybridization between MPc-Im and PCN is primarily governed by noncovalent interactions, such as π–π stacking and electrostatic interactions.²²

Photocatalytic activity of MPc-Im/PCN hybrids was examined in a mixed solution of DMA–TEOA (4 : 1 v/v) under LED irradiation (λ = 405 nm). The DMA–TEOA mixed solution is a valid choice for metal complex/PCN hybrids as the reaction solution.⁶ The gas-phase products were analyzed by means of a gas-chromatogram equipped with a thermal conductive detector (GC-TCD). While unloaded PCN did not show activity (entry 1), immobilization of FePc-Im or CoPc-Im onto PCN resulted in CO production with byproduction of H₂ (Table 1). No liquid phase product could be identified in all cases. This indicates that the immobilized FePc-Im or CoPc-Im acted as catalysts for reduction reactions, in particular for CO₂ reduction into CO. The selectivity to CO was ≥70%, and turnover number for CO formation (TON_CO) greater than 1 was confirmed. However, loading a sufficient amount of Pc catalysts was required to obtain measurable CO formation. The use of Al₂O₃, an insulator, instead of semiconductive PCN did not yield CO formation (entries 8 and 9), indicating that visible light absorption by the PCN component is necessary to drive the CO₂ reduction. No CO formation was observed when physical mixtures of dissolved FePc (or CoPc) and suspended PCN were used (entries 10 and 11). This suggests that intimate contact between catalytic MPc-Im and PCN is necessary for CO₂ reduction. The apparent quantum yields for CO formation over the optimal FePc-Im/PCN and CoPc-Im/PCN hybrids were 0.096 and 0.087% at 400 nm, respectively. These numbers are ∼3 times higher than that reported using the CoPPc/PCN hybrid (0.03% at 400 nm).¹² However, the CO formation rate was degraded with time (Fig. S10†). UV-vis DRS indicated that the characteristic absorption band of MPc-Im were weakened significantly after the reaction (Fig. S11†), indicating that the degradation of activity is attributed to the desorption of MPc-Im from the PCN surface. Therefore, the interface between MPc-Im and PCN was not very strong, and the stability of the hybrids needs to be improved.

Table 1 Results of visible-light CO₂ reduction (λ = 405 nm)^a

Entry	Photocatalyst	Amount of Pc used/μmol g^−1	Product/μmol		CO selectivity/%	TONCO
			CO	H2
a Reaction conditions: catalyst, 10 mg; solution, 10 mL DMA–TEOA mixed solution (4 : 1 v/v); reaction time, 5 h. b Pc was dissolved in the reaction solution prior to reaction.
1	PCN	0	n.d.	0.031	—	—
2	FePc-Im/PCN	2.0	n.d.	0.14	0	—
3	FePc-Im/PCN	8.8	0.10	0.017	86	1.1
4	FePc-Im/PCN	32.8	0.41	0.024	94	1.2
5	CoPc-Im/PCN	2.0	n.d.	0.067	0	—
6	CoPc-Im/PCN	8.8	0.19	0.055	77	1.6
7	CoPc-Im/PCN	24.3	0.38	0.164	70	2.1
8	FePc-Im/Al2O3	8.8	n.d.	0.027	0	—
9	CoPc-Im/Al2O3	8.8	n.d.	<0.01	0	—
10^b	FePc + PCN	33	n.d.	0.015	0	—
11^b	CoPc + PCN	24	n.d.	0.081	0	—

---

To confirm the carbon source for the CO₂ reduction product, ¹³CO₂ reduction was conducted using FePc-Im/PCN and CoPc-Im/PCN hybrids, and the gas-phase products were analyzed by gas-chromatography mass spectroscopy (GC-MS). As shown in Fig. 2, ¹³CO signal with m/z = 29 was observed when ¹³CO₂ was used as the reactant, while ¹²CO (m/z = 28) was the sole product from unlabeled CO₂. Thus, the produced CO during the photocatalytic reactions using FePc-Im/PCN and CoPc-Im/PCN actually originated from CO₂.

Fig. 2 Mass spectra of CO peaks in GC-MS analysis for the gas-phase products of the photocatalytic reaction (λ = 405 nm). Reaction conditions: catalyst, 8 mg (FePc-Im, 24 μmol g⁻¹; CoPc-Im, 32 μmol g⁻¹); solution, 4 mL DMA–TEOA mixed solution (4 : 1 v/v); reaction time, 20 h.

Another interesting finding is that methane was detected during the ¹³CO₂ reduction experiment (Fig. S12†). In principle, methane can be detected by GC-TCD, but it was not possible in this work, probably due to the low production rate. On the other hand, GC-MS has much higher sensitivity for the product detection. Not only ¹³CH₄ but also ¹²CH₄ was detected during the ¹³CO₂ reduction. The ¹²CH₄ produced was probably due to hydrocarbon contaminants in the photocatalyst surface.²³ Anyway, both FePc-Im/PCN and CoPc-Im/PCN was found to have the ability to convert CO₂ into CH₄, which is very rare among the metal complex/semiconductor hybrid photocatalysts ever reported. It is generally known that Pcs work as CO-evolving catalysts in electrochemical and photocatalytic CO₂ reduction. On the other hand, heterogenized Pcs may exhibit the ability to give multi-electron reduction products such as methane and methanol.¹³

In summary, we successfully demonstrated CO₂ reduction into CO with ≥70% selectivity using cationic imidazolium-functionalized Fe(II) and Co(II) phthalocyanine complexes, which were non-covalently bound on the surface of polymeric carbon nitride. Importantly, CO₂-to-CH₄ conversion was confirmed by means of isotope tracer experiments with ¹³CO₂. This work suggests that suitably heterogenized metal phthalocyanines may allow not only CO formation by two-electron reduction of CO₂, but also conversion to more value-added compounds by multi-electron reduction even in a simple hybrid photocatalyst system.

Data availability

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

Author contributions

K. M. and M. I. designed/supervised the project and wrote the manuscript draft. P. K. D. synthesized and characterized Pc derivatives and Pc/PCN catalysts. R. Nakada and R. Nakazato conducted photocatalytic reactions and electrochemical measurements with X. Z. All authors reviewed the manuscript and approved its submission.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by a Grant-in-Aid for Transformative Research Areas (A) “Supra-ceramics” (JP22H05148) (JSPS). The authors thank Prof. Mio Kondo (Science Tokyo) for assistance in GC-MS analysis.

Notes and references

1. J. Wu, Y. Huang, W. Ye and Y. Li, Adv. Sci., 2017, 4, 1700194.
2. S. Sato, T. Arai and T. Morikawa, Inorg. Chem., 2015, 54, 5105–5113.
3. Y. Yamazaki, H. Takeda and O. Ishitani, J. Photochem. Photobiol., C, 2015, 25, 106–137.
4. K. Maeda, Adv. Mater., 2019, 31, e1808205.
5. K. E. Dalle, J. Warnan, J. J. Leung, B. Reuillard, I. S. Karmel and E. Reisner, Chem. Rev., 2019, 119, 2752–2875.
6. A. Nakada, H. Kumagai, M. Robert, O. Ishitani and K. Maeda, Acc. Mater. Res., 2021, 2, 458–470.
7. E. Nikoloudakis, I. Lopez-Duarte, G. Charalambidis, K. Ladomenou, M. Ince and A. G. Coutsolelos, Chem. Soc. Rev., 2022, 51, 6965–7045.
8. K. Maeda, K. Sekizawa and O. Ishitani, Chem. Commun., 2013, 49, 10127–10129.
9. J. Lin, Z. Pan and X. Wang, ACS Sustainable Chem. Eng., 2013, 2, 353–358.
10. R. Kuriki, K. Sekizawa, O. Ishitani and K. Maeda, Angew. Chem., Int. Ed., 2015, 54, 2406–2409.
11. S. A. Balogun, D. Thole, D. Masekela and K. D. Modibane, Adv. Energy Sustainability Res., 2024, 5, 2400169.
12. S. Roy and E. Reisner, Angew. Chem., Int. Ed., 2019, 58, 12180–12184.
13. Z. Yue, C. Ou, N. Ding, L. Tao, J. Zhao and J. Chen, ChemCatChem, 2020, 12, 6103–6130.
14. E. S. Da Silva, N. M. M. Moura, M. G. P. M. S. Neves, A. Coutinho, M. Prieto, C. G. Silva and J. L. Faria, Appl. Catal., B, 2018, 221, 56–69.
15. J. Choi, P. Wagner, R. Jalili, J. Kim, D. R. MacFarlane, G. G. Wallace and D. L. Officer, Adv. Energy Mater., 2018, 8, 1801280.
16. G. Zhang, Z. A. Lan, L. Lin, S. Lin and X. Wang, Chem. Sci., 2016, 7, 3062–3066.
17. M. E. Ragoussi, J. J. Cid, J. H. Yum, G. de la Torre, D. Di Censo, M. Gratzel, M. K. Nazeeruddin and T. Torres, Angew. Chem., Int. Ed., 2012, 51, 4375–4378.
18. A. C. Yuzer, E. Genc, G. Kurtay, G. Yanalak, E. Aslan, E. Harputlu, K. Ocakoglu, I. Hatay Patir and M. Ince, Chem. Commun., 2021, 57, 9196–9199.
19. G. Liu, Y. Wang, Y. Zhou, J. Cao, M. Yuan and H. Lv, J. Colloid Interface Sci., 2021, 594, 658–668.
20. Y. Guo, S. Song, Y. Zheng, R. Li and T. Peng, Dalton Trans., 2016, 45, 14071–14079.
21. L. Lin, C. Hou, X. Zhang, Y. Wang, Y. Chen and T. He, Appl. Catal., B, 2018, 221, 312–319.
22. D. Chen, K. Wang, W. Hong, R. Zong, W. Yao and Y. Zhu, Appl. Catal., B, 2015, 166–167, 366–373.
23. T. Yui, A. Kan, C. Saitoh, K. Koike, T. Ibusuki and O. Ishitani, ACS Appl. Mater. Interfaces, 2011, 3, 2594–2600.

---

Footnotes

† Electronic supplementary information (ESI) available: Additional characterization and reaction data. See DOI: https://doi.org/10.1039/d5cy00676g

‡ Equal contribution.

---