Polyhedral oligomeric silsesquioxane as a recyclable soft template to synthesize mesoporous polymeric carbon nitride with enhanced photocatalytic hydrogen evolution

Abstract

Recyclable template synthesis of polymeric carbon nitride (CN) is rarely reported. Herein, for the first time, octamethyl polyhedral oligomeric silsesquioxane (mPOSS) was used as a recyclable soft template to prepare mesoporous CN with increased specific surface area, photoabsorption, charge separation efficiency, and photocatalytic H₂ evolution rate which is ∼4.5 times that of the bulk CN. mPOSS exhibits high stability and recoverability. This work opens an avenue for recyclable template synthesis of mesoporous CN and may advance its industrial application.

---

Energy and environmental issues induced by industrialization and rapid population growth have become great challenges facing human beings. Largely inspired by the wisdom of nature, semiconductor photocatalysis has been regarded as a promising technology to solve these issues, for which appropriate photocatalysts play the pivotal role.¹ Among a large library of photocatalysts explored, melon-type polymeric carbon nitride (simply marked as CN) has attracted tremendous attention due to its outstanding optical properties, facile synthesis, non-toxicity, extraordinary physicochemical stability,² and applicable energy band structures (with a conduction band edge and valence band edge of ∼−1.2 V and 1.5 V (vs. SHE), respectively)³ for various photocatalytic reactions, such as degradation of organic pollutants,⁴ water splitting,⁵ and reduction of CO₂.⁶ However, the performance of pristine CN is still limited by its low specific surface area (S_B), weak visible light absorption, and fast recombination of photogenerated charge carriers.⁷

Creation of mesopores in CN is one of the most effective ways to simultaneously increase the S_B, the optical absorption, and the charge separation efficiency.^2,8,9 The templating method is used the most in reported work to prepare mesoporous CN,⁷ and templates reported can be classified into hard templates and soft templates. Hard templates mainly include various mesoporous silica^10–12 and calcium carbonate¹³ which must be removed from products by acid etching¹⁴ to achieve mesopores, while soft templates involve a few surfactants¹⁵ and block copolymers¹⁶ which are decomposed in the calcination process to form mesopores. Hitherto, both hard and soft templates reported are unrecyclable, which will substantially increase the synthetic cost of mesoporous CN and prevent their industrial production, let alone that waste fluids from the elimination of hard templates may give rise to environmental pollution¹⁷ and soft templates can lead to detrimental carbon doping in resultant mesoporous CN.⁷ Therefore, it is significant at present to develop recyclable and completely removable templates for cost-effective synthesis of mesoporous CN photocatalysts.

Herein, we, for the first time, report a recyclable soft template, octamethyl polyhedral oligomeric silsesquioxane (mPOSS) that can be used to prepare mesopore-rich CN (denoted as M-CN, Fig. 1a) which exhibits enlarged S_B and enhanced photoabsorption, charge separation, and photocatalytic H₂ evolution. mPOSS possesses a well-defined cage-like molecular structure (Fig. S2†) and comprises irregular particles with sizes from several tens of nanometers to several micrometers (Fig. S3†). When mixed with melamine in DMF and dried by evaporation at 80 °C (S1.2†), melamine blocks could be loaded with small mPOSS particles and encompassed by large particles (Fig. 1a and S4†), as clearly revealed by the energy dispersive X-ray spectroscopy (EDS) spectra and elemental mapping images (Fig. S5†). Fourier-transform infrared (FT-IR) spectra (Fig. S6 and Table S1†) indicate that the molecular structures of melamine and mPOSS are quite stable in the mixing process and hydrogen bonds may form between melamine and mPOSS. Upon calcining the melamine/mPOSS mixture, mPOSS with a sublimation temperature of ∼300 °C¹⁸ acts as the blocking agent to restrain the fusion of melamine and relevant intermediate (e.g., melam) particles, resulting in generation of abundant mesopores in M-CN (Fig. 1a). Survey X-ray photoelectron spectroscopy (XPS) spectra, Si 2p core-level XPS spectra and thermogravimetric analysis indicate residue of a little silica (<5 wt%) in the product (Fig. S7–S9†) which can be completely removed (Fig. S8†) by etching in a 4 M NH₄HF₂ solution (S1.2†). SEM and TEM images of M-CN indeed show lots of pores among aggregated and partially fused small sheet-like particles (Fig. 1b and c), in comparison with those of the bulk CN with few pores (Fig. S10†).

Fig. 1 (a) Schematic illustration of the synthesis and structures of precursors and M-CN; and (b) SEM and (c) TEM images of M-CN.

The pore structure of M-CN was further investigated by N₂ sorption isotherm tests. As shown in Fig. 2a, the isotherms of both CN and M-CN exhibit typical type-IV isotherms featuring type-H3 hysteresis loops, indicating the existence of slit-like mesopores formed by stacking of nanosheets.¹⁹ As expected, M-CN possesses much larger S_B and pore volume (35.5 m² g⁻¹ and 0.091 cm³ g⁻¹) than the bulk CN (8.4 m² g⁻¹ and 0.018 cm³ g⁻¹). The pore-size distribution curve of M-CN shows a wide peak at 10–50 nm and a strong peak at 2–4 nm (see the inset in Fig. 2a), which can be ascribed to the mesopores produced in the sublimation process of mPOSS with different sizes and shapes, in agreement with the SEM and TEM analysis.

Fig. 2 (a) N₂ adsorption–desorption isotherms of CN and M-CN and (inset) corresponding pore size distribution curves obtained by calculating adsorption branches of the isotherms using the Barrett–Joyner–Halenda (BJH) method; and (b) UV-vis diffuse reflectance spectra and (inset) corresponding Tauc plots of samples.

X-ray diffraction (XRD) patterns of M-CN and CN (Fig. S11†) are similar and show two diffraction peaks at ∼13.2° and 27.3°, corresponding to the (210) and (002) planes of melon-type CN with a layered orthorhombic structure,^20,21 respectively. Their interlayer distance is determined as 0.32 nm, according to the (002) peak position. Their FT-IR spectra (Fig. S12†) both show typical absorption bands of melon-type CN, i.e., the absorption bands at 3172, 1830–850, and 810 cm⁻¹, ascribed to the N–H stretching vibration, C=N/C–N stretching vibrations in heptazine rings, and the characteristic breathing mode of the heptazine rings, respectively.²² Moreover, the C 1s and N 1s core-level XPS spectra of CN and M-CN are also similar. In the C 1s spectra (Fig. S13†), peaks at binding energies of 284.6, 286.5, 288.3, and 293.5 eV are assigned to adventitious carbon contaminants,²³ π–π* of the adventitious carbon rings,²⁴ N=C–N bonds, and π excitation of the heptazine rings,²⁵ respectively. In the N 1s spectra (Fig. S14†), peaks centered at 398.5, 400.1, 401.2, and 404.0 eV are attributed to nitrogen in C–N=C, N–H, and N–(C)₃ bonds, and π excitation, respectively.^26,27 Elemental analysis (Table S2†) indicates that the N/C molar ratios of M-CN and CN are both 1.50 (the theoretical value of melon-type CN). These results demonstrate that the generation of massive mesopores in M-CN did not induce detectable variation of the molecular framework, compared with the bulk CN.

The optical absorption performance of CN and M-CN was investigated by UV-vis diffuse reflectance spectroscopy (DRS). As shown in Fig. 2b, M-CN exhibits higher light-harvesting capability than CN across the whole optical spectrum, which should result from the multiple reflections of incident light in the massive mesopores²³ in M-CN. The multiple reflections also induce a slight redshift¹⁷ of the absorption edge of M-CN (446 nm) in comparison with that of CN (453 nm) (Fig. 2b). Tauc plots indicate that the indirect bandgaps (E_g)²⁸ of CN and M-CN are ∼2.67 and 2.60 eV, respectively (see the inset in Fig. 2b), that is, the E_g of M-CN is a little narrower. The remarkably enhanced optical absorption of M-CN can also be reflected by its darker color than that of CN (insets in Fig. 2b).

The charge separation and transfer performance of the samples was also investigated. As shown in Fig. 3a, photoluminescence (PL) spectra of both CN and M-CN show one peak at ∼466 nm, roughly corresponding to their bandgap (∼2.66 eV) emission. Then the lowered PL intensity of M-CN, compared with that of CN, suggests the decreased recombination efficiency of photogenerated charge carriers.¹⁷ Fluorescence lifetimes of the samples were further measured by time-resolved fluorescence spectroscopy. As shown in Fig. 3b, the lifetimes (τ₁, τ₂, and τ₃), achieved by fitting the decay curves to a triple exponential model²⁹ (S1.3†), and the mean lifetime ( where Int_i is the percentage of corresponding τ_i) of M-CN are all shorter than those of CN (see the inset table in Fig. 3b), indicative of rapid transfer of photogenerated electrons and/or holes, decreasing their direct radiative recombination,³⁰ and subsequent non-radiative energy transition of them.³¹ Moreover, M-CN exhibits ∼3.8-fold stronger cathodic photocurrent density than CN (Fig. 3c), also revealing much faster charge separation for M-CN.³² The enhanced charge separation for M-CN arises from its abundant mesoporous structure that shortens the distance and time for the charge carrier transfer from the bulk to the surface.¹⁷

Fig. 3 (a) PL spectra, (b) time-resolved fluorescence spectra, (c) cathodic photocurrent density (j_p), and (d) EPR spectra of CN and M-CN. The inset table in (b) shows the results of fitting the decay curves to a tri-exponential model. Dark current density in (c) was set at 0 for distinct comparison.

Room-temperature electron paramagnetic resonance (EPR) spectroscopy was conducted to roughly characterize electronic structures of the samples (Fig. 3d). Both CN and M-CN show one single Lorentzian line centering at g of 2.0039 in the dark and under light irradiation, corresponding to sp²-caron in the heptazine rings.³³ M-CN exhibits a stronger EPR signal than CN in the dark because of its much larger S_B,³⁴ and the irradiation induced EPR signal enhancement for M-CM is prominently higher than that for CN (Fig. 4d), because the higher photoabsorption performance of M-CN results in generation of more photoexcited holes and electrons,³⁵i.e., greater delocalized electron density than that of CN. The high delocalized electron density favors the charge transfer (i.e., high conductivity) and the photocatalytic performance improvement.

Fig. 4 (a) Photocatalytic H₂ evolution on CN and M-CN under visible light irradiation with Pt as the cocatalyst and triethanolamine as the electron donor; (b) the cyclic experiment of photocatalytic H₂ evolution on M-CN; and (c) XRD patterns and (d) ²⁹Si NMR spectra of fresh and recycled mPOSS. Numbers in (a) are hydrogen evolution rates (μmol g⁻¹ h⁻¹).

Energy band levels of the samples were roughly confirmed. Mott–Schottky plots indicate that both CN and M-CN are n-type semiconductors and their flat-band potentials are −1.11 and −1.25 V (vs. SHE), respectively (Fig. S15†). Their flat-band potentials are approximately equal to their conduction band (CB) minimum (E_CB).³⁶ Then, the valence band (VB) maximum (E_VB = E_g + E_CB) of CN and M-CN is calculated as 1.56 and 1.33 V, respectively. VB-XPS spectra were recorded to determine the energy difference between Fermi levels (E_f) and E_VB. As shown in Fig. S16,† the (E_VB − E_f) values of CN and M-CN are determined as 2.52 and 2.51 eV (or V), respectively, i.e. their E_f of −0.96 and −1.18 V (vs. SHE). The schematic diagram of energy band levels of the samples is shown in Fig. S17,† and apparently, the E_CB of M-CN is higher than that of CN, which means higher reducibility of photogenerated electrons for M-CN,³⁷ favorable for H₂ evolution from water splitting.

The photocatalytic performance of the samples was evaluated via the photocatalytic H₂ evolution reaction. When using triethanolamine (10 vol%) as the electron donor and Pt as the cocatalyst for H₂ evolution on CN and M-CN under visible light irradiation, the optimal Pt loading amount is 3 wt% (Fig. S18†). The H₂ evolution rate (r_H) of M-CN can reach 534 μmol g⁻¹ h⁻¹, that is almost 4.5 times that of CN (118 μmol g⁻¹ h⁻¹) (Fig. 4a), and is comparable to those of most reported template-synthesized mesoporous CN (Table S3†). The turnover number (TON) of M-CN, relative to Pt, reaches ∼21 after 6 hours with a turnover frequency (TOF) of 3.5 h⁻¹. The S_B normalized r_H of M-CN (15.0 μmol m⁻² h⁻¹) is only ∼7% higher than that of CN (14.0 μmol m⁻² h⁻¹), suggesting that S_B is the key factor influencing the photocatalytic activity. The apparent quantum efficiency of M-CN was calculated to be ∼2.0% at 420 nm (Fig. S19†). After four consecutive H₂ production runs, M-CN does not exhibit an obvious photoactivity decrease (Fig. 4b), and the XRD pattern, the FT-IR spectrum, and SEM and TEM images of M-CN after the cycling experiment (Fig. S20†) are similar to those before, revealing the high chemical stability of M-CN. In addition, M-CN with silica residue (before etched with NH₄HF₂) exhibits similar r_H to M-CN (Fig. S21†), indicative of the negligible effect of residual silica on the photocatalytic performance. Omitting the etching process for template removal benefits much the industrial production.

The sublimed mPOSS could be desublimed on the quartz tube wall (close to the outlet) in the tube furnace, signifying that the simple condensation is an effective way to recycle the mPOSS. The XRD pattern of recycled mPOSS is similar to that of fresh mPOSS³⁸ except several observable weak peaks belonging to monoclinic melamine (JCPDS no. 39-1950) (Fig. 4c) coming from the sublimation of feedstock (i.e., melamine) on calcination. The FT-IR spectrum of the recycled mPOSS also shows similar absorption bands to those of the fresh one, apart from the weak absorption bands of melamine (Fig. S22†). Furthermore, ²⁹Si nuclear magnetic resonance (NMR) spectroscopy was performed to characterize the structure of mPOSS. As shown in Fig. 4d, there is only one distinct peak at −74.1 ppm assigned to Si atoms in the cubelike (T³) product³⁹ for both recycled and fresh mPOSS. The SEM image of the recycled mPOSS also shows particles with various shapes and sizes (Fig. S23†), just like that of fresh mPOSS (Fig. S3†). These demonstrate that mPOSS is structurally stable during calcination and condensation processes and is recyclable for M-CN preparation.

Conclusions

In summary, a recyclable soft template, mPOSS is first used to prepare mesopore-rich M-CN with enlarged S_B and enhanced optical absorption, charge separation, and thus photocatalytic H₂ evolution. The H₂ evolution rate of M-CN can reach 4.5-fold that of the bulk CN under visible-light irradiation and its chemical stability is satisfactory. This work opens an avenue for recyclable template synthesis of mesoporous CN and favors the industrial production of the photocatalyst.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This research is supported by the Department of Science & Technology of Shandong Province (Grant 2019JZZY020229).

References

1. L. Jiang, X. Yuan, G. Zeng, J. Liang, Z. Wu and H. Wang, Environ. Sci.: Nano, 2018, 5, 599–615.
2. N. Tian, H. Huang, X. Du, F. Dong and Y. Zhang, J. Mater. Chem. A, 2019, 7, 11584–11612.
3. X. Wang, K. Maeda, A. Thomas, K. Takanabe, G. Xin, J. M. Carlsson, K. Domen and M. Antonietti, Nat. Mater., 2009, 8, 76–80.
4. Z. Liang, Y. Xia, G. Ba, H. Li, Q. Deng and W. Hou, J. Mater. Chem. A, 2019, 7, 25824–25829.
5. S. Yu, J. Li, Y. Zhang, M. Li, F. Dong, T. Zhang and H. Huang, Nano Energy, 2018, 50, 383–392.
6. F. Li, D. Zhang and Q. Xiang, Chem. Commun., 2020, 56, 2443–2446.
7. Z. Yang, Y. Zhang and Z. Schnepp, J. Mater. Chem. A, 2015, 3, 14081–14092.
8. N. Tian, Y. Zhang, X. Li, K. Xiao, X. Du, F. Dong, G. I. N. Waterhouse, T. Zhang and H. Huang, Nano Energy, 2017, 38, 72–81.
9. C. Liu, H. Huang, L. Ye, S. Yu, N. Tian, X. Du, T. Zhang and Y. Zhang, Nano Energy, 2017, 41, 738–748.
10. X. Jin, V. V. Balasubramanian, S. T. Selvan, D. P. Sawant, M. A. Chari, G. Q. Lu and A. Vinu, Angew. Chem., Int. Ed., 2009, 48, 7884–7887.
11. Y.-S. Jun, W. H. Hong, M. Antonietti and A. Thomas, Adv. Mater., 2009, 21, 4270–4274.
12. X. Li, A. F. Masters and T. Maschmeyer, ChemCatChem, 2015, 7, 121–126.
13. J. Wang, C. Zhang, Y. Shen, Z. Zhou, J. Yu, Y. Li, W. Wei, S. Liu and Y. Zhang, J. Mater. Chem. A, 2015, 3, 5126–5131.
14. W. Iqbal, B. Yang, X. Zhao, M. Rauf, M. Waqas, Y. Gong, J. Zhang and Y. Mao, Catal. Sci. Technol., 2018, 8, 4576–4599.
15. Y. Wang, X. Wang, M. Antonietti and Y. Zhang, ChemSusChem, 2010, 3, 435–439.
16. H. Yan, Chem. Commun., 2012, 48, 3430–3432.
17. W. Iqbal, C. Dong, M. Xing, X. Tan and J. Zhang, Catal. Sci. Technol., 2017, 7, 1726–1734.
18. K. J. Shea and D. A. Loy, Chem. Mater., 2001, 13, 3306–3319.
19. S. Liu, J. Chen, D. Xu, X. Zhang and M. Shen, J. Mater. Res., 2018, 33, 1391–1400.
20. Z.-A. Lan, G. Zhang and X. Wang, Appl. Catal., B, 2016, 192, 116–125.
21. F. Fina, S. K. Callear, G. M. Carins and J. T. S. Irvine, Chem. Mater., 2015, 27, 2612–2618.
22. G. Ba, Z. Liang, H. Li, Q. Deng, N. Du and W. Hou, Chem. Eng. J., 2020, 380, 122535.
23. Y. Zheng, L. Lin, X. Ye, F. Guo and X. Wang, Angew. Chem., Int. Ed., 2014, 53, 11926–11930.
24. D. L. Yu, J. L. He, Z. Y. Liu, B. Xu, D. C. Li and Y. J. Tian, J. Mater. Sci., 2008, 43, 689–695.
25. K. He, J. Xie, Z.-Q. Liu, N. Li, X. Chen, J. Hu and X. Li, J. Mater. Chem. A, 2018, 6, 13110–13122.
26. Q. Han, B. Wang, J. Gao, Z. Cheng, Y. Zhao, Z. Zhang and L. Qu, ACS Nano, 2016, 10, 2745–2751.
27. C.-M. Zhu, A. Gao, Y. Wang and Y. Liu, Chem. Commun., 2014, 50, 13889–13892.
28. V. R. Battula, S. Kumar, D. K. Chauhan, S. Samanta and K. Kailasam, Appl. Catal., B, 2019, 244, 313–319.
29. R. Godin, Y. Wang, M. A. Zwijnenburg, J. Tang and J. R. Durrant, J. Am. Chem. Soc., 2017, 139, 5216–5224.
30. L. Ke, P. Li, X. Wu, S. Jiang, M. Luo, Y. Liu, Z. Le, C. Sun and S. Song, Appl. Catal., B, 2017, 205, 319–326.
31. S. Cao, Y. Li, B. Zhu, M. Jaroniec and J. Yu, J. Catal., 2017, 349, 208–217.
32. Y. Wang, X. Liu, J. Liu, B. Han, X. Hu, F. Yang, Z. Xu, Y. Li, S. Jia, Z. Li and Y. Zhao, Angew. Chem., Int. Ed., 2018, 57, 5765–5771.
33. G. Zhang, M. Zhang, X. Ye, X. Qiu, S. Lin and X. Wang, Adv. Mater., 2014, 26, 805–809.
34. G. Zhang, J. Zhang, M. Zhang and X. Wang, J. Mater. Chem., 2012, 22, 8083–8091.
35. W. Ho, Z. Zhang, M. Xu, X. Zhang, X. Wang and Y. Huang, Appl. Catal., B, 2015, 179, 106–112.
36. G. Zhang, Z.-A. Lan, L. Lin, S. Lin and X. Wang, Chem. Sci., 2016, 7, 3062–3066.
37. C. Liu, Y. Zhang, F. Dong, A. H. Reshak, L. Ye, N. Pinna, C. Zeng, T. Zhang and H. Huang, Appl. Catal., B, 2017, 203, 465–474.
38. A. Fina, D. Tabuani, A. Frache and G. Camino, Polymer, 2005, 46, 7855–7866.
39. J. Vince, B. Orel, A. Vilčnik, M. Fir, A. Šurca Vuk, V. Jovanovski and B. Simončič, Langmuir, 2006, 22, 6489–6497.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental section, molecular structure of mPOSS, SEM images and EDS spectra, XRD patterns, partial XPS and FT-IR spectra, TGA curves, Mott–Schottky plots, VB-XPS spectra, schematic diagram of energy band structures, and partial photocatalytic data of samples. See DOI: 10.1039/d0se01388a

---