Switching the photocatalytic activity of g-C₃N₄ by homogenous surface chemical modification with nitrogen residues and vacancies

Abstract

A facile two-step homogenous approach is established to produce and control the nitrogen vacancies on g-C₃N₄ photocatalysts. The g-C₃N₄ undergoes a solvothermal N₂H₄·H₂O reduction inactivation and subsequent thermal reduction process to reactivate and achieve an enhanced photocatalytic activity toward hydrogen evolution.

---

The large-scale generation of hydrogen from photocatalytic water splitting with sunlight could provide a sustainable pathway to address the current global energy and environmental issues. To transform the technology into reality, highly efficient, stable, and cost-effective photocatalysts must be developed. In this context, polymeric graphite phase carbon nitride (g-C₃N₄) has emerged as a promising metal-free visible-light photocatalyst owing to its extraordinary optical characteristics, chemical stability and good thermal solidity.¹ However, the efficiency of g-C₃N₄ is far from satisfaction due to its low practical surface area and high photogenerated charge recombination rate. To improve the photocatalytic activity of g-C₃N₄, various strategies, such as doping with non-metals (B, C, F, O and S etc.)² or metals (Zn, Fe, and Co etc.),³ heterojunctions with other organic or inorganic semiconductors,⁴ exfoliating the g-C₃N₄ into ultrathin nanosheets and morphology controlling have been adopted.⁵ The structural disorders induced by the treatments mentioned above not only increase the number of active centres on surfaces but also provide trapping sites for photo excited carriers and hinder the recombination of electrons and holes which are advantageous for promoting photocatalytic efficiency.^6–8 Moreover, the activities of this organic photocatalyst are also sensitive to the change of non-covalent interactions, such as hydrogen bonding, π–π stacking, and van der Waals forces in bulk and interfaces.⁹

Although the photoactivity of the modified g-C₃N₄ catalysts have been improved with the incorporation of heteroatoms or heterojunctions, the intrinsic molecular structure of g-C₃N₄ was greatly changed and the role of the heteroatoms in the enhancement of g-C₃N₄ is usually too complicated to be fully discussed. Homogeneous modification can play a substantial role in modifying the properties of photocatalysts and explore the possible mechanism of the improvement process.^8,10 Recently, Niu et al. reported a nitrogen defected g-C₃N₄ (melon) by homogeneous hydrogenation demonstrating the underlying role of nitrogen vacancies in modulating optical and photocatalytic properties of g-C₃N₄.¹¹ Nevertheless, the optimized hydrogen treatment temperature window for the g-C₃N₄ with nitrogen vacancies ranges from 520 to 540 °C, which is very close to the thermal decomposition temperature of g-C₃N₄ and easy to lead the serious destruction of the layered structure.^6,11 Therefore, it is necessary to develop a more facile and reliable homogeneous modification method that can effectively control the nitrogen vacancies and photocatalytic performance of g-C₃N₄.

In this study, we for the first time present a two-steps homogenous approach to large-scale production of nitrogen defected g-C₃N₄ samples after a solvothermal and subsequently low temperature thermal reduction process (Fig. 1a). Interestingly, the g-C₃N₄ undergoes a solvothermal hydrazine (N₂H₄·H₂O) reduction inactivation and thermal reduction reactivation process to achieve an enhanced photocatalytic activity toward hydrogen evolution. These results could have important meaning for understanding the possible reactive sites on melon based catalysts and modification strategies making for various applications.

Fig. 1 Schematic (a) and chemical structural formula of the proposed reaction pathway (b) of the solvothermal inactivation and thermal reduction reactivation process.

We selected N₂H₄·H₂O as reducing agent because its widely application in the reduction of graphite oxide (GO), and the perpendicular electron lone pairs on the two pyramidal H₂N can effectively control the reduction of graphene-like materials in both the solution and gas phases.^12,13 Although the nucleophilicity of the terminal amino groups (N1, N2) can be improved by hydrazine substituents,¹³ this chemical approach usually give rise to a high nitrogen impurities (hydrazine derivatives) as depicted in Fig. 1b. With these in mind we further heated our solvothermally treated g-C₃N₄ in reducing atmosphere (H₂) to remove the nitrogen residue and conduct homogeneous modification with nitrogen vacancies.

For simplicity, the solvothermally treated samples were denoted as CN-x, where x (= 1, 2, 3, 4, 5) refers to the concentrations of adding 80% N₂H₄·H₂O aqueous solution of 0.05, 0.1, 0.25, 0.4 and 0.5 mL, respectively. Similarly, the following thermal hydrogen treated samples were denoted as CN-x–H. Compared to the pristine g-C₃N₄ (CN), the C/N molar ratio of CN-x decreases gradually with increased adding amount of N₂H₄·H₂O, from 0.6670 to 0.6582 for CN and CN-5, respectively (as shown in Fig. 2). This clearly proves that N-rich hydrazine derivatives are incorporated into the CN structures. However, a contrary tendency of the C/N molar ratio is detected for CN-x–H samples, around 0.7% of the nitrogen atoms were lost in CN-5–H, suggesting that the loss of nitrogen atoms is closely related to the concentration of N₂H₄·H₂O in solvothermal treatment. Based on previous work by Niu et al.,¹¹ the energy change (ΔE) for removing a lattice N and terminating functional groups by H atoms was followed by N2 < N1 < N3 < N4 shown in Fig. 1b. Thus, N2 and N1 are more easily to be reduced to NH₃ in thermal hydrogen reduction. Upon solvothermal N₂H₄ treatment of CN, N3, N4 may partially converted to N1 or N2 due to the chemical reduction and electrostatic interaction between the positively charged H (in CN) and the negatively charged N (in NHNH₂) and/or between the negatively charged O (unavoidable oxidation of the CN) and a positively charged H (in NHNH₂).¹⁴

Fig. 2 Comparison of C/N molar ratio of CN after solvothermal treatment (black line) and subsequently thermal hydrogen reduction (blue line). Dash lines are the corresponding CN samples before and after thermal hydrogen reduction (CN–H) for reference.

The solvothermal and thermal hydrogen treatments do not change the basic characteristics of the layered the texture structure of CN samples, as indicated by combined analysis of X-ray diffraction (XRD) patterns, Fourier transform infrared (FTIR), X-ray photoelectron (XPS) spectra and C K-edge X-ray absorption near-edge structure (XANES) spectrum (Fig. S1–S4†).

Transmission electron microscopy (TEM) measurements were performed to characterize the morphology and structure of CN, CN-2 and CN-2–H. As shown in Fig. 3a, the CN displays two-dimensional layered and plate-like sheets. Moreover, after solvothermal treatments, some macropores can be observed on the surface of CN-2 and CN-2–H.

Fig. 3 (a) TEM images of CN, CN-2 and CN-2–H samples. The scale bars of the TEM images are 50 nm; (b) room temperature EPR spectra of CN, CN-2 and CN-2–H samples.

The electronic properties of CN, CN-x and CN-x–H samples were further analysed by electron paramagnetic resonance (EPR) technology at room temperature. As shown in Fig. 3b, there is mainly one single Lorentzian line centred at a g value of 2.0035 for the CN sample, originating from the lone pair electrons in sp²-carbon atoms in the aromatic rings, with the π-bonded nanosized clusters on the surface of the semiconductor.¹⁵ Subsequently, a weakened, broadened and superimposed spectra can be detected for the CN-2 materials: one arising from the carbon sites in a typical heptazine g-C₃N₄ and the other one from the introduced hydrazine derivatives groups.^16a To our knowledge, the EPR lines are usually broadened by the anisotropy of the magnetic couplings and the line-broadening in powder spectra of the free radicals with small g-anisotropy is usually caused by the anisotropy of the hyperfine coupling.^16b That is, involved hydrazine derivative groups not only alter the π-electron delocalization in the conjugated system, but also induced the g-anisotropy. In addition, the EPR intensity was greatly strengthened and broadened after the treatment of hydrogen (CN-2–H). Furthermore, the slightly enhanced signals at around 3485 and 3560 G may be attributed to the g-anisotropy which is closely related to the electronic and band structures. Whilst, slightly increased EPR intensities were observed when CN, CN-2 and CN-2–H samples were irradiated with visible light, indicating photochemical generation of radical pairs useable for heterogeneous photocatalytic reactions.¹⁶

After solvothermal N₂H₄ treatment, the incorporation of hydrazine derivatives into CN modifies the π-electron delocalization in the conjugated aromatic system, and thus changes the intrinsic optical/electronic properties of the resulting CN-x and CN-x–H samples. As increasing the N₂H₄·H₂O content, a remarkable redshifts in absorption from about 450 nm to 800 nm are observed for CN-x samples (Fig. S5A†). This would allow maximal utilization of solar photon flux. On the other hand, the continuous red shift happened on the CN-x–H samples (Fig. S5B†), owing to the extension of electron delocalization in the aromatic sheets with enhanced structural connections, somewhat similar to the H-aggregates type intermolecular packing.¹⁷

Fig. S6† gives the photoluminescence (PL) spectra under an excitation wavelength of 365 nm. An obvious fluorescence quenching is observed for CN-x samples (Fig. S6A†), signifying that the radiative charge recombination which happened on the samples surface has been efficiently suppressed. This indicates the electron relocalization on surface terminal sites, or reduced density of charge carrier traps for electron–hole recombination. Simultaneously, for the CN, CN-2 and CN-2–H samples, excepted for the same quenching tendency, a slight blue-shift of fluorescence emission peak from 463.5 to 458.3 eV is detected (Fig. S6B†).

As depicted in Fig. 4a, after solvothermal N₂H₄ treatment, the introduced nitrogen impurities (such as NHNH₂, NHNH, and NNH₂) hindered the separation and transmission of electrons and holes, resulting in sluggish photocatalytic performance. When the adding concentrations exceeding 0.4 mL (CN-3), the CN-x samples were inactivated totally. Subsequently, after the following thermal treatment in H₂ atmosphere, most of the introduced nitrogen residues and some lattice N atoms with comparatively lower ΔE were removed and the delocalization of the π-electrons might be altered simultaneously. At this time, all the CN-x–H samples were not only reactivated but also demonstrated a remarkable improvement in H₂ evolution activity over pristine CN to varying extents. The activity comparison in Fig. 4a inset suggests that the CN-2–H is around 5 times active in photocatalytical generation of H₂ than the pristine CN under visible light irradiation. The above results clearly demonstrate the tow-steps homogeneous approach with controllable nitrogen vacancies can effectively improve the photocatalytic activity of CN samples.

Fig. 4 (a) The rate of H₂ evolution of all the samples. (b) Steady rate of H₂ production by CN-2–H samples as a function of wavelength of the incident light. Ultraviolet-visible absorption spectrum of the catalyst is also shown for comparison. Inset is the stability test of CN-2–H for H₂ production under visible light (red line), the CN (black line) and CN-2 (blue line) are also shown for comparison.

The stability of CN-2–H acting as polymeric catalyst for H₂ evolution was evaluated by six consecutive operating under visible light (>420 nm) (Fig. 4b, inset). The production of H₂ increases steadily with prolonged time of light irradiation. A slightly deactivation with time is observable in the fourth run curves, a fact which we attribute to the decreased concentration of triethanolamine. However, after recycling, the curves start with a higher primary activity. Wavelength-dependent H₂ evolution was performed on the CN-2–H photocatalyst using different long pass optical filters. In Fig. 4b, the trend of H₂ production matches well with the diffuse reflectance spectra, suggesting that the HER is indeed driven by light absorption on the catalyst.

The transient photocurrents of the samples were also evaluated during repeated ON/OFF illumination cycles at 0.6 V vs. Ag/AgCl in 0.2 M Na₂SO₄. As can be seen from the Fig. S7,† upon light irradiation, a typical n-type photocurrent was produced, and when the irradiation was interrupted, the photocurrent rapidly dropped to almost zero, but reverted to the original value once light was switched back on again which is reproducible. As expected, an enhanced photocurrent is generated by the CN-2–H photoelectrode over the bulk CN and CN-2 samples. The tendency of transient photocurrents matches well with the results of photocatalytic performance of CN samples.

In conclusion, a facile two-steps homogeneous approach was introduced to produce and control the nitrogen vacancies on g-C₃N₄ base photocatalysts. The photocatalytic activity of carbon nitride polymer can be easily switched by solvothermal chemical reduction and thermal H₂ treatment process due to the surface nitrogen impurities adsorption and nitrogen vacancies generation. As a consequence, the CN-x–H shows much superior photocatalytic activities compared to the pristine CN in H₂ evolution under visible light irradiation. We believe that the rational design of melon polymers with controlled functional groups through the diversity of organic chemistry will open a new pathway for this promising visible light responding semiconductors. The distinct “plasticity” of g-C₃N₄ makes it a potential candidate in a variety of advanced applications.

Acknowledgements

We thank Prof. Caihao Hong for the valuable discussion on XANES. This work was financially supported by National Natural Science Foundation of China (21373083), SRF for ROCS, SEM, SRFDP, Programme for Professor of Special Appointment (Eastern Scholar) at Shanghai Institutions of Higher Learning and Australian Research Council's Future Fellowships (FT120100913).

Notes and references

1. (a) X. Chen, S. Shen, L. Guo and S. S. Mao, Chem. Rev., 2010, 110, 6503; (b) X. Wang, K. Maeda, A. Thomas, K. Takanabe, G. Xin, J. M. Carlsson, K. Domen and M. Antonietti, Nat. Mater., 2009, 8, 76–80.
2. J. Li, B. Shen, Z. Hong, B. Lin, B. Gao and Y. Chen, Chem. Commun., 2012, 48, 12017–12019.
3. B. Yue, Q. Li, H. Iwai, T. Kako and J. Ye, Sci. Technol. Adv. Mater., 2011, 12, 034401.
4. (a) X. Xu, G. Liu, G. Randorn and J. T. S. Irvine, Int. J. Hydrogen Energy, 2011, 36, 13501–13507; (b) J. Zhang, G. Zhang, X. Chen, S. Lin, L. Möhlmann, G. Dołȩga, G. Lipner, M. Antonietti, S. Blechert and X. Wang, Angew. Chem., Int. Ed., 2012, 51, 3183–3187.
5. (a) J. Xu, L. Zhang, R. Shi and Y. Zhu, J. Mater. Chem. A, 2013, 1, 14766; (b) M. Shalom, S. Inal, C. Fettkenhauer, D. Neher and M. Antonietti, J. Am. Chem. Soc., 2013, 135, 7118–7121.
6. A. Thomas, A. Fischer, F. Goettmann, M. Antonietti, J. Mȕller, R. Schlögl and J. M. Carlsson, J. Mater. Chem., 2008, 18, 4893–4908.
7. (a) X. L. Wang, W. Q. Fang, H. F. Wang, H. Zhang, H. Zhao, Y. Yao and H. G. Yang, J. Mater. Chem. A, 2013, 1, 14089; (b) X. L. Wang, W. Q. Fang, S. Yang, P. Liu, H. Zhao and H. G. Yang, RSC Adv., 2014, 4, 10676.
8. X. B. Chen, L. Liu, P. Y. Yu and S. S. Mao, Science, 2011, 331(10), 746.
9. G. R. Desiraju, Nature, 2001, 412, 397–400.
10. (a) M. K. Nowotny, L. R. Sheppard, T. Bak and J. Nowotny, J. Phys. Chem. C, 2008, 112, 5275–5300.
11. P. Niu, L. C. Yin, Y. Q. Yang, G. Liu and H. M. Cheng, Adv. Mater., 2014, 26, 8046–8052.
12. H. J. Shin, K. K. Kim, A. Benayad, S. M. Yoon, H. K. Park, I. S. Jung, M. H. Jin, H. K. Jeong, J. M. Kim, J. Y. Choi and Y. H. Lee, Adv. Funct. Mater., 2009, 19, 1987–1992.
13. T. Saplinova, V. Bakumov, T. Gmeiner, J. Wagler, M. Schwarz and E. Kroke, Z. Anorg. Allg. Chem., 2009, 635, 2480–2487.
14. (a) M. C. Kim, G. S. Hwang and R. S. Ruoff, J. Chem. Phys., 2009, 131, 064704; (b) J. Zhong, J. Wang, J. G. Zhou, B. H. Mao, C. H. Liu, H. B. Liu, Y. L. Li, T. K. Sham, X. H. Sun and S. D. Wang, J. Phys. Chem. C, 2013, 117, 5931–5936.
15. Y. Zhang, Z. Schnepp, J. Cao, S. Ouyang, Y. Li, J. Ye and S. Liu, Sci. Rep., 2013, 3, 2163.
16. (a) P. Ionita, F. Tuna, M. Andruh, T. Constantinescu and A. T. Balaban, Aust. J. Chem., 2007, 60, 173–179; (b) A. Lund, M. Shiotani and S. Shimada, in Principles and Applications of ESR Spectroscopy, Springer, Netherlands, 2011, pp. 105–164; (c) Y. Cui, Z. Ding, X. Fu and X. Wang, Angew. Chem., Int. Ed., 2012, 51, 11814–11818.
17. G. Zhang, J. Zhang, M. Zhang and X. Wang, J. Mater. Chem., 2012, 22, 8083–8091.

---

Footnote

† Electronic supplementary information (ESI) available: Synthesis, elemental analysis, XRD, FTIR, XPS, etc. See DOI: 10.1039/c5ra00150a

---