Graphitic carbon nitride nanoribbon for enhanced visible-light photocatalytic H₂ production

Abstract

Chemical scissors provide a new vision to manufacture unique carbon nitride nanostructures with improved photocatalytic performance. Herein, graphitic carbon nitride nanoribbon (GCNR), with a typical length of 1.75 μm, width of 210 nm and thickness of 3 nm, was obtained by acid treatment of bulk g-C₃N₄ with HNO₃/H₂SO₄ mixtures. The C/N molar ratio of GCNR (around 0.629) was much smaller than that of pristine g-C₃N₄ (0.758). It was demonstrated that larger amounts of carbon vacancies on the ultra-thin ribbon structures could contribute to improved electron–hole separation efficiency and excellent photocatalytic H₂ production. The average H₂ production rate of GCNR under visible light was 49.4 μmol h⁻¹, which was 20 times that of the original catalyst.

---

Development of highly active and green photocatalysts for environmental protection and energy shortage has attracted lots of attention during recent years.^1,2 Among various photocatalysts, graphitic carbon nitride (g-C₃N₄), as a non-metal organic semiconductor, has been deeply and widely researched due to its low cost, excellent anti-photocorrosion and chemical stability.³ However, bulk g-C₃N₄ directly synthesized from thermal polymerization usually has the disadvantages of low specific area, less active sites and fast electron–hole combination rate, thus hinders their photocatalytic performance.⁴ Meanwhile, serious aggregation and restacking can be caused due to the π–π stacking and van der Waals attraction between g-C₃N₄ nanosheets, and thus limit the transport of H₂O₂ intermediate product which usually poisons g-C₃N₄ nanosheets during the H₂ production process.^3a

Various solutions towards solving these problems of bulk g-C₃N₄ have been developed, which mainly include two approaches: texture modification and control of polymerization degree.⁵ The structure and composition of carbon nitrides could be engineered by doping of non-metal/metal elements, copolymerization or morphology control. Previous works have shown that special structures, i.e., nanoparticle,⁶ nanosheet,⁷ nanorod,⁸ hollow microspheres⁹ and porous materials,¹⁰ can not only endow them with numerous boundaries and thus decrease the aggregation, but more importantly can guide charge movement to accelerate the collection and separation of electron–hole pairs at materials interface for efficient utilization of solar energy to drive relevant reactions.¹¹ Dong et al. prepared high-quality C₃N₄ nanosheet with alkali treatment and inorganic salt assistance method respectively which both show excellent photocatalytic activity.¹² Yang et al. has reported free-standing g-C₃N₄ nanosheets prepared by liquid phase exfoliation have large aspect ratios, high surface area, and stoichiometric N/C ratio.¹³ Further treatment by increasing the amounts of active sites on the g-C₃N₄ structures would also improve the photocatalytic performance.¹⁴ Hong et al. showed that the nitrogen-deficient g-C₃N₄ synthesized by hydrothermal treatment with oxidant owned an enhanced photocatalytic H₂ evolution rate.¹⁵ Therefore, it is very meaningful to make active vacancies on the 2D g-C₃N₄ nanosheet for enhancing photocatalytic performance.

In this work, a facile way was developed to synthesize g-C₃N₄ nanoribbon (GCNR) by acid treatment (H₂SO₄/HNO₃) of pristine g-C₃N₄ (PCN). Meanwhile, g-C₃N₄ was exfoliated into nanoribbons with abundant in-plane carbon vacancies due to strong chemical exfoliation and etching effect. The abundant carbon vacancies may be beneficial for hindering the recombination of electrons and holes when they provide trapping sites for photoinduced holes. The result is photoluminescence and time-resolved fluorescence decay spectra demonstrated that the charge life time and separation of photoinduced electron–hole of GCNR structures have been greatly improved compared to the PCN. Moreover, the hydrogen evolution rate was improved from 2.5 to 49.4 μmol h⁻¹ under visible light.

The volume ratio between H₂SO₄ and HNO₃ was crucial to the formation of nanoribbon structures, and the optimised volume ratio of H₂SO₄ : HNO₃ was 1 : 2. Fig. 1a is the transmission electron microscopy (TEM) images of the GCNR. The nearly transparent feature of the samples indicated the ultra-thin feature. The length distribution of as-prepared GCNR was shown in the inset of Fig. 1, which was evaluated from the SEM images by casually selecting 80 nanoribbons (Fig. S1†). The length ranged from 1 to 3 μm, with the average length of 1.7 μm. The width of nanoribbons is about 210 nm. The pores on the surface of the GCNR may be resulted from the strong irradiation of electron-beam. Atomic force microscopy (AFM) was used to measure the thickness of nanoribbons. Two nanoribbons were chosen randomly and their height profiles were around 2.9 nm and 3.1 nm, respectively, which confirmed the ultrathin feature of the GCNR.¹⁶ Acid treatment, could exfoliate the layer structures, thus reducing the thickness of PCN.¹⁷

Fig. 1 (a) TEM images of the as-prepared PRC; inset of (a) is the size distribution of the sample evaluated from the TEM image. (b and c) AFM image and height profile of the PRCN. High resolution (d) C 1s and (e) N 1s XPS spectra of g-C₃N₄ and GCNR, respectively.

The N₂ adsorption–desorption isotherms of the products were shown in Fig. S2.† The samples exhibited type III isotherms, indicative of mesoporous materials.¹⁴ The Brunauer–Emmett–Teller was 2 times larger than that of PCN (11 m² g⁻¹). The improved specific surface area could facilitate mass transfer during the photocatalytic reaction, possibly enhancing the reaction kinetics.

X-ray diffraction (XRD) was employed to identify the crystal structure and chemical structure of the g-C₃N₄ before and after treatment. In the XRD patterns (Fig. S3†), the diffraction peak (100) around 13.0° was related to the inter-plane structure packing, while (002) peak at 27.2° indicated the periodic interlayer stacking of conjugated aromatic systems.¹⁸ Compared with the PCN, the intensity of GCNR was much lower which indicated the layered g-C₃N₄ has been successfully exfoliated. This was consistent with the AFM analysis. The valance states and chemical compositions were studied via X-ray photoelectron spectroscopy (XPS). The presence of C, N and O elements can be observed in the survey XPS spectra (Fig. S4†). The corresponding C/N molar ratio of GCNR (0.629) was much smaller than that of pristine g-C₃N₄ (0.758), indicating the presence of carbon vacancies in GCNR (Table 1). The formation of carbon vacancies may be resulted from the strong acid etching of the g-C₃N₄ framework. The high-resolution C 1s spectrum was shown in Fig. 2a. The peak centered at 287.8 eV belongs to C–N coordination of g-C₃N₄, whereas the peak at 288.6 eV is assigned to N–C–N and peak at 284.6 eV is C–C coordination in graphite.¹⁹ Interestingly, the peak at 284.6 eV obviously becomes weaken after treatment. The C(C–C)/N_total atomic ratio decreases from 0.15 to 0.04. Meanwhile, the deconvolution of high resolution N 1s XPS profiles show three peaks at 398.8 eV, 399.7 eV, 404.7 eV, representing nitrogen in the form of N=C–N, N–(C)₃ and N–H, respectively.^11,16 It was found that the atomic percentages of N–H and N=C–N bonds decreases slightly may result from the improvement of the groups on the edge after tailored by HNO₃. Fig. S5† shows the high-resolution O 1s spectrum where the peaks located at about 531.4 eV and 532.6 eV corresponded to the –OH and C=O, respectively. The area ratios of the peak –OH at GCNR are obviously higher than that of PCN indicating that a greater more hydroxyl content on the surface.²⁰ The fourier transform infrared (FTIR) spectra confirms the existence of hydroxyl group after acid treatment as shown in Fig. S6.†

Table 1 XPS analysis of PCN and GCNR

	C 1s			N 1s
	Position (eV)	Assignment	C/Ntotal atomic ratio	Position (eV)	Assignment	N/Ntotal atomic ratio	C/N atomic ratio
PCN	284.6	C–C	0.15	398.9	N=C–N	0.69	0.758
	287.8	C–NHX	0.34	400.6	N–(C)3	0.26
	288.6	N–C–N	0.26	404.6	N–H	0.05
GCNR	284.6	C–C	0.04	398.9	N=C–N	0.65	0.629
	287.8	C–NHX	0.37	400.6	N–(C)3	0.27
	288.6	N–C–N	0.22	404.6	N–H	0.08

---

Fig. 2 (a) The evolution process of morphology for tailored g-C₃N₄. (b–f) TEM images of different mixed acid volume ratios: (b) HNO₃ : H₂SO₄ = 1 : 3 (c) HNO₃ : H₂SO₄ = 1 : 2 (d) HNO₃ : H₂SO₄ = 1 : 1 (e) HNO₃ : H₂SO₄ = 2 : 1 (f) HNO₃ : H₂SO₄ = 3 : 1.

The formation mechanism of GCNR was proposed too. Based on the above experimental data and analysis, the key factor to obtain nanoribbon is to control the volume ratio of HNO₃ to H₂SO₄ accurately. Samples with different volume ratio of HNO₃ to H₂SO₄ are prepared as shown in Fig. S7.† In this process, we believe H₂SO₄ is responsible for chemical exfoliation while HNO₃ results in the structure shape. Previous study has demonstrated the bulk g-C₃N₄ can be treated by H₂SO₄ to achieve intercalation and exfoliation.²⁵ After the chemical exfoliation by H₂SO₄. HNO₃ acts as the chemical scissor, carbon atom has a bigger surface area and smaller molecular weight than nitrogen atom, which is prone to break the trizaines (C₃N₃) first due to heptazines unit (C₆N₇) is more energetically stable than triazines unit. At the same time, this selective cutting results in the porous structure formation of g-C₃N₄.²⁶ As the concentration of HNO₃ increases, the structure of g-C₃N₄ can be orientational tailored into nanoribbons. Finally, nanodots can be obtained when the ratio of HNO₃ : H₂SO₄ exceeds 3. As seen in Fig. 2, we can observe that when HNO₃ : H₂SO₄ volume ratio less than 2, porous g-C₃N₄ can be obtained, and that when the concentration of HNO₃ beyonds the threshold, the strong cut would destroy g-C₃N₄ seriously and form g-C₃N₄ ribbons and nanodots.

The resultant GCNR was evaluated for photocatalytic H₂ production under visible light (λ > 420 nm). As shown in Fig. 3a, it was found that the photocatalytic activity of GCNR showed a significant improvement over the PCN. The H₂ evolution rate of GCNR reached 49.3 μmol h⁻¹ was about 19.7 times that of the PCN (2.5 μmol h⁻¹). The apparent quantum yield (AQY) for H₂ evolution was also calculated for the GCNR photocatalysts.²⁷ The AQY values for GCNR at 420 nm is estimated as 2.4%. Table 2 summarizes the photocatalytic water splitting based on C₃N₄ with different morphologies.

Fig. 3 (a) Photocatalytic H₂ evolution performance. (b) UV-visible absorption and Tauc plot inset. (c) PL spectra. (d) Transient photocurrent response. (e) Time-resolved PL spectra. (f) EPR spectra of the pristine g-C₃N₄ (black curve) and GCNR (red curve).

Table 2 Photocatalytic water splitting based on C₃N₄ with different morphologies

Morphology (C3N4)	Mass of catalyst	Reactant solution	Lightsource wavelength	Activity (μmol h^−1)	Quantum efficiency (%)	Reference
Nanosheet	20 mg + 3 wt% Pt	20 mL of 10 vol% triethanolamine aqueous solution	500 W Xe lamp (>420 nm)	230	—	21
Nanosheet	50 mg + 3 wt% Pt	100 mL of 10 vol% triethanolamine aqueous solution	300 W Xe lamp (>420 nm)	93	3.75 (429 nm)	13a
Macroscopic monolith	10 mg + 3 wt% Pt	30 mL of 10 vol% triethanolamine aqueous solution	300 W Xe lamp (>420 nm)	29	—	14
Nanosheet	50 mg + 6 wt% Pt	300 mL of 10 vol% triethanolamine aqueous solution	300 W Xe lamp (>400 nm)	33	—	22
Holey nanosheet	10 mg + 3 wt% Pt	30 mL of 10 vol% triethanolamine aqueous solution	300 W Xe lamp (>420 nm)	82.9	—	23
Mesoporous hexagonal framework	20 mg + 3 wt% Pt	100 mL of 10 vol% triethanolamine aqueous solution	300 W Xe lamp (>420 nm)	243	6.77 (450 nm)	24
Holey nanosheet	50 mg + 3 wt% Pt	210 mL of 14 vol% triethanolamine aqueous solution	300 W Xe lamp (>400 nm)	316.7	—	10
This work	100 mg + 3 wt% Pt	80 mL of 10 vol% triethanolamine aqueous solution	500 W Hg lamp (>420 nm)	49.7	2.4 (420 nm)	—

---

The excellent photocatalytic hydrogen evolution of GCNR can be attributed to the unique morphology and structures. On one hand, the ultrathin nanostructure of GCNR might provide more active sites during the photocatalytic reaction. Then, the existence of more carbon vacancies of GCNR would contribute to remarkably optical absorption properties and abundant active sites. The optical properties of the samples were studied by UV-vis absorption spectra (Fig. 3b). The derived electronic band gaps from the Tauc plots were 2.83 eV for GCNR and 3.0 eV for pristine sample (inset in Fig. 3b).^3a The photoluminescence (PL) (shown in Fig. 3c) further confirmed the above result. Two broad emission peaks at 418 nm and 449 nm were observed for GCNR and PCN, respectively. The blue shift of PL spectra results from the strong quantum confinement effect, which are caused by the decrease of the conjugation length and opposite directions shift of conduction and valence bands.^14,28 The photocurrent responses of the samples with typical on–off cycles under visible light irradiation were further investigated shown in Fig. 3d. The photocurrent density of GCNR is 2.1 times larger than that of the pristine catalyst. The results indicated that the separation rate of the photoexcited charge carriers was promoted. Fig. 3e showed the time-resolved fluorescence decay spectra which indicated the average radiative lifetime. The τ₂ lifetime component improved from 2.7 ns for pristine g-C₃N₄ to 4.2 ns for GCNR. The carbon vacancies may be beneficial for hindering the recombination of electrons and holes when they provide trapping sites for photoinduced holes because the improvement of photocurrent and lifetime of charge carriers. Specifically, the average PL lifetime of GCNR (2.4 ns) was 1.45 times longer than that of pristine g-C₃N₄ (1.65 ns) which was beneficial to improve the transport rate of the photo-generated carriers. This result demonstrated that the separation of electron–hole pairs along the in-plane direction in GCNR nanostructure could be improved due to the anisotropic structure.²⁹ As shown in Fig. 3f, the EPR spectrum of pristine g-C₃N₄ shows a weak similar response at g = 2.0036. Note the GCNR displays a strong symmetrical signal centering indicates the generation of carbon-based radicals and much higher concentration of unpaired electrons which play an important role in the photocatalytic reaction.^14,15 At the same time, the ferromagnetism is generated on the zigzag edges created by the carbon vacancies in the broken honeycomb lattice.^23,30 These unsaturated N atoms could act as the paramagnetic centers and attract photoelectrons from the conduction band of GCNR.³¹

In conclusion, g-C₃N₄ nanoribbon (GCNR) containing a large number of in-plane carbon vacancies was prepared by controlling the HNO₃/H₂SO₄ volume ratio. Photoluminescence and time-resolved fluorescence decay spectra demonstrate GCNR improves the separation of electron–hole and charge lifetime. Compared to the pristine g-C₃N₄, the GCNR shows the hydrogen evolution rates improved from 2.5 μmol h⁻¹ to 49.3 μmol h⁻¹ under visible light as a result of the special structure, abundant carbon vacancies and excellent ability of separation of photoexcited electrons and holes. This work provides a practical method to control the structure and photocatalytic activity of g-C₃N₄, and thus enhance the application of this metal-free, stable and available photocatalyst.

Acknowledgements

This work was supported by NSFC (51572173), STCSM (16JC1402200, 15520720300).

Notes and references

1. J. Zhang, Y. Chen and X. Wang, Energy Environ. Sci., 2015, 8, 3092.
2. X. Dong and F. Cheng, J. Mater. Chem. A, 2015, 3, 23642.
3. (a) J. Liu, Y. Liu, N. Y. Liu, Y. Z. Han, X. Zhang, H. Huang, Y. Lifshitz, S. T. Lee, J. Zhong and Z. Kang, Science, 347, 970; (b) X. D Zhang, X. Xie, H. Wang, J. J Zhang, B. C Pan and Y. Xie, J. Am. Chem. Soc., 2013, 135, 18; (c) X. Zhang, H. Wang, H. Wang, Q. Zhang, J. Xie, Y. Tian, J. Wang and Y. Xie, Adv. Mater., 2014, 26, 4438.
4. (a) S. Cao, J. Low, J. Yu and M. Jaroniec, Adv. Mater., 2015, 27, 2150; (b) Y. Zheng, J. Liu, J. Liang, M. Jaroniec and S. Z. Qiao, Energy Environ. Sci., 2012, 5, 6717; (c) Y. Wang, X. Wang and M. Antonietti, Angew. Chem., Int. Ed., 2012, 51, 68.
5. (a) Y. Zhang, A. Thomas, M. Antonietti and X. C. Wang, J. Am. Chem. Soc., 2009, 131, 50; (b) S. Yan, Z. Li and Z. Zou, Langmuir, 2009, 25, 10397.
6. X. Jin, V. V. Balasubramanian, S. T. Selvan, D. P. Sawant, M. A. Chari, G. Q. Lu and A. Vinu, Angew. Chem., 2009, 121, 8024.
7. Z. Xing, A. M. Asiri, A. O. Al-Youbi and X. Sun, Nanoscale, 2013, 5, 8921.
8. X. Li, X. Wang and M. Antonietti, Chem. Sci., 2012, 3, 2170.
9. D. Zheng, C. Huang and X. Wang, Nanoscale, 2015, 7, 465.
10. P. Yang, J. Zhao, W. Qiao, L. Lia and Z. Zhu, Nanoscale, 2015, 7, 18887.
11. J. Zhang, M. Zhang, C. Yang and X. Wang, Adv. Mater., 2014, 26, 4121.
12. (a) F. Cheng, J. Yan, C. Zhou, B. Chen, P. Li, Z. Chen and X. Dong, J. Colloid Interface Sci., 2016, 468, 103; (b) J. Yan, C. Zhou, P. Li, B. Chen, S. Zhang, X. Dong, F. Xi and J. Liu, Colloids Surf., A, 2016, 508, 257.
13. (a) S. Yang, Y. Gong, J. Zhang, L. Zhan, L. Ma, Z. Fang, R. Vajtai, X. Wang and P. M. Ajayan, Adv. Mater., 2013, 25, 2452; (b) Y. Yuan, W. Xu, L. Yin, S. Cao, Y. Liao, Y. Tang and C. Xue, Int. J. Hydrogen Energy, 2013, 38, 13159; (c) J. Xu, L. Zhang, R. Shi and Y. Zhu, J. Mater. Chem. A, 2013, 1, 14766.
14. Q. Liang, Z. Li, Z. Huang, F. Kang and Q. Yang, Adv. Funct. Mater., 2015, 25, 6885.
15. Z. H. Hong, B. Shen, Y. l. Chen, B. Z. Lin and B. F. Gao, J. Mater. Chem. A, 2013, 1, 11754.
16. (a) M. J. Bojdys, N. Severin, J. P. Rabe, A. I. Cooper, A. Thomas and M. Antonietti, Macromol. Rapid Commun., 2013, 34, 850; (b) X. She, H. Xu, Y. Xu, J. Yan, J. Xia, L. Xu, Y. Song, Y. Jiang, Q. Zhang and H. Li, J. Mater. Chem. A, 2014, 2, 2563.
17. (a) H. Li, B. Sun, L. Sui, D. Qian and M. Chen, Phys. Chem. Chem. Phys., 2015, 17, 3309; (b) J. Sun, Y. Deng, J. Li, G. Wang, P. He, S. Tian, X. Bu, Z. Di, S. Yang, G. Ding and X. Xie, ACS Appl. Mater. Interfaces, 2016, 8, 10226.
18. (a) X. Wang, K. Maeda, A. Thomas, K. Takanabe, G. Xin, J. M. Carlsson, K. A. Domen and M. Antonietti, Nat. Mater., 2009, 8, 76; (b) L. Tian, S. Yang, Y. Yang, J. Li, Y. Deng, S. Tian, P. He, G. Ding, X. Xie and Z. Wang, RSC Adv., 2016, 6, 82648.
19. (a) G. Zhang, J. Zhang, M. Zhang and X. Wang, J. Mater. Chem., 2012, 22, 8083; (b) S. Yang, J. Sun, X. Li, W. Zhou, Z. Wang, P. He, G. Ding, X. Xie, Z. Kang and M. Jiang, J. Mater. Chem. A, 2014, 2, 8660; (c) S. Yang, J. Sun, P. He, X. Deng, Z. Wang, C. Hu, G. Ding and X. Xie, Chem. Mater., 2015, 27, 2004.
20. (a) J. Feng, T. Chen, S. Liu, Q. Zhou, Y. Ren, Y. Lv and Z. Fan, J. Colloid Interface Sci., 2016, 479, 1; (b) S. Yang, C. Zhu, J. Sun, P. He, N. i Yuan, J. Ding, G. Ding and X. Xie, RSC Adv., 2015, 5, 33347; (c) X. Deng, J. Sun, S. Yang, H. Shen, W. Zhou, J. Lu, G. Ding and Z. Wang, Appl. Phys. Lett., 2015, 107, 241905.
21. J. Xu, L. Zhang, R. Shi and Y. Zhu, J. Mater. Chem. A, 2013, 1, 14766.
22. P. Niu, L. Zhang, G. Liu and H. Cheng, Adv. Funct. Mater., 2012, 22, 4763.
23. Q. Liang, Z. Li, Z. Huang, F. Kang and Q. Yang, Adv. Funct. Mater., 2015, 25, 6885.
24. J. Zhang, F. Guo and X. Wang, Adv. Funct. Mater., 2013, 23, 3008.
25. F. Zhao, H. Cheng, Y. Hu, L. Song, Z. Zhang, L. Jiang and L. Qu, Sci. Rep., 2014, 4, 5882.
26. (a) X. Bai, S. Yan, J. Wang, L. Wang, W. Jiang, S. Wu, C. Sun and Y. Zhu, J. Mater. Chem. A, 2014, 2, 17521; (b) M. T. Schumperili, C. Hammond and I. Hermans, Phys. Chem. Chem. Phys., 2012, 14, 11002.
27. L. Yuan, C. Han, M. Yang and Y. Xu, Int. Rev. Phys. Chem., 2016, 35, 1.
28. A. P. Alivisatos, Science, 1996, 271, 933.
29. (a) T. Ma, S. Dai, M. Jaroniec and S. Qiao, Angew. Chem., Int. Ed., 2014, 53, 7281; (b) Y. Tang, P. Wang, J. Yun, R. Amal and Y. H. Ng, J. Mater. Chem. A, 2015, 3, 15876.
30. S. Sun, J. Magn. Magn. Mater., 2013, 344, 39.
31. S. Li, G. H. Dong, R. Hailili, L. Yang, Y. Li, F. Wang, Y. Zeng and C. Wang, Appl. Catal., B, 2016, 190, 26.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra23218c

---