A “pillared” process to construct graphitic carbon nitride based functionalized mesoporous materials

Abstract

We present a novel method to prepare mesoporous materials via in situ self-assembly of graphitic carbon nitride nanosheets and silica nanoparticles. Combining the advantages of g-C₃N₄ nanosheets and mesoporous structure, the as-prepared materials exhibit superior adsorption capabilities for heavy metal ions and organic pollutants.

---

Mesoporous materials have been widely studied in the area of catalysis,¹ energy storage and conversion,² adsorption,^3–5 and biomedicine^6,7 due to their unique features and properties. In general, mesoporous materials are synthesized based on the nanocasting method by employing costly organic molecules as “soft” templates^8–12 or by using mesoporous materials as “hard” templates.¹³ The key drawback of the method is the extra steps required to remove the templates, which not only increases the cost but also results in the production of harmful gases due to the removal of the organic templates by combustion at high temperature. Therefore, a modified procedure for the synthesis of porous materials without using templates is of particular interest. Recently, Yan's group proposed a novel method to prepare three-dimensional (3D) graphene/nanoparticle composites without any templates,¹⁴ and Gao's group obtained 3D graphene/carbon nanotubes composites via a simple cryodesiccation procedure.¹⁵ These studies represent an important step towards the green and facile synthesis of composite materials in the future.

Graphitic carbon nitride (g-C₃N₄), a metal-free polymer semiconductor, possesses a graphite-like structure with weak van der Waals forces between C–N layers.¹⁶ Since its discovery, g-C₃N₄ has been studied in many areas such as catalysis,^17,18 bioimaging,^19–21 and photosynthesis.^22–27 The two-dimensional (2D) g-C₃N₄ nanosheets (CNS), which can be obtained by various methods including thermal oxidation “etching”,²⁸ and liquid phase exfoliation by sonication or strong acid,^19,23,29,30 have exceptional properties with respect to bulk g-C₃N₄. Particularly, the negatively charged CNS may interact efficiently with the nanoparticles (NPs) that are positively charged. Considering the affinity between g-C₃N₄ and graphene, and that the solubility of CNS can be tuned by changing pH value and operation temperature,³⁰ CNS could be used as 2D nanosheets matrix to assemble NPs, leading to 3D CNS/NPs porous materials.

In this study, we demonstrate a novel approach to fabricate mesoporous materials by employing 2D CNS as matrix to capture SiO₂ NPs in liquid phase. The products obtained after simple processing exhibit high surface areas and uniform pore size distribution. This “pillared” strategy may be applied to fabricate various CNS-based functional materials by replacing SiO₂ NPs with other NPs.

The strategy for synthesis of CNS-based mesoporous materials is illustrated in Scheme 1. As described in our previous work,³⁰ bulk g-C₃N₄ was exfoliated to 2D nanosheets and the obtained CNS show pH value-dependent properties. Similar to a process of fishing, the nanoparticles are the “fishes”, and the CNS can be viewed as “fishnets”. The combination of CNS and NPs tend to form a network by self-assembly when the pH value of the system is altered. In this way, the NPs can be captured into the network of nanosheets in a simple one-step process. In principle, any acid resistant NPs homogeneously dispersed in aqueous solution could be embedded into the CNS network to form g-C₃N₄/NPs composite materials.

Scheme 1 Schematic illustration for the preparation of SiO₂/CNS nanocomposites.

The crystal structures of the samples were evaluated using X-ray diffraction. XRD measurements show the typical diffraction patterns for g-C₃N₄ (Fig. 1A). The strong reflection at 27.2° for bulk g-C₃N₄ corresponds to the (002) planes, characteristic of interplanar stacking of graphitic materials.^16,30 In contrast with the bulk g-C₃N₄, a slight shift to 27.8° and a sharp decrease in the overall intensity for the (002) peak can be observed for SiO₂-xN/C₃N₄. These are possibly due to the decreased interlayer distance of g-C₃N₄ and the lower contents of g-C₃N₄ in SiO₂-xN/g-C₃N₄ as compared to the bulk sample. Furthermore, the additional peak, which is assigned to an in-plane structure packing of bulk g-C₃N₄ (13°), shifts negatively (12°) and becomes broadened with increasing SiO₂ contents. During the process of exfoliating bulk g-C₃N₄ into CNS, the crystal structure of the nanosheets substantially remains the same except that the reflection peak becomes less pronounced.²⁹ While after activation by protonation, the peak at 27.4° (bulk-g-C₃N₄) which reflects the characteristic interlayer-stacking of aromatic system shifts to a higher angle.³⁰ The results of XRD patterns are similar to the results that caused by protonation (Fig. S1†). However, the preparation process of the materials here was performed under alkaline condition, and the shifts of XRD peaks should not be attributed to protonation. Furthermore, the protonated g-C₃N₄ can be recovered by calcinations, while the samples in this work are not. Thus, the shift of the peaks could be ascribed to the pillared effect, which was induced by the external stress caused by SiO₂ nanoparticles and decreased the slight undulation of the CNS. The stretched properties of g-C₃N₄ induced by external stress defined as “flatting effect” (Fig. S2†). Moreover, the amorphous SiO₂ peak at about 22° becomes more obvious as the SiO₂ contents increase. These observations confirm that the structures of g-C₃N₄ and SiO₂ NPs were preserved in the samples.

Fig. 1 XRD patterns (A) of bulk g-C₃N₄ (a), SiO₂-xα-g-C₃N₄ composites ((b) x = 2, (c) x = 5, (d) x = 8, (e) x = 11). FT-IR spectra (B) and XPS profile of N 1s (C), C 1s (D) of Si-5α-CN sample.

The structures and compositions of the SiO₂-xα-g-C₃N₄ were further investigated by Fourier transform infrared spectroscopy (FTIR), X-ray photoelectron spectroscopy (XPS), and elemental analysis (EA). The strong FTIR absorption bands of Si-5α-CN shown in Fig. 1B further reveal the typical molecular structure of g-C₃N₄. The absorption bands in the 1200–1600 cm⁻¹ region manifest the typical stretching mode of aromatic CN heterocycles. The sharp band at 806 cm⁻¹ can be attributed to triazine ring mode, which corresponds to condensed CN heterocycles.^23,30 The broad band at 3000–3500 cm⁻¹ may be related to uncondensed terminal amino groups (–NH₂ or =NH groups) and O–H groups from water molecules and SiO₂. In addition, the peak at 1085 cm⁻¹ is attributed to the asymmetric stretching vibrations of Si–O–Si.³¹ Note that the absorption peaks at about 1320 and 1250 cm⁻¹ for Si-5α-CN become less pronounced as compared to the bulk sample. The flatting effect of SiO₂ as described in XRD analysis will reduce the asymmetry of the N–(C)₃ group, leading to the weakened absorption intensity. This can be rationalized by the acidification instead of SiO₂ flatting effect as demonstrated in our previous studies.³⁰ The FTIR results for SiO₂-5α-CN composite are in good agreement with those of bulk g-C₃N₄, implying that the basic unit of g-C₃N₄ was not destroyed.^32–34

The surface chemical states of Si-5α-CN and the bulk g-C₃N₄ were studied by XPS (Fig. 1C and D, S3 and S4†). As shown in Fig. S3,† the sample is comprised of C, N, Si and O elements. To gain insight into the chemical bonding between the carbon and nitrogen atoms in the samples, the high resolution C 1s was further deconvoluted into three Gaussian–Lorentzian peaks. The peak centered at 288.7 eV is assigned to the sp² hybridized carbon in the triazine ring bonded to the –NH₂ group, while the peak at 287.9 eV is attributed to the sp² hybridized carbon bonded to N inside the triazine rings.³⁵ The peak at 284.6 eV is typically ascribed to the signal of standard reference carbon.³⁰ The high resolution XPS spectrum of N 1s could be also fitted with four different peaks. The dominant peak at 398.8 eV (denoted as N1 in Fig. 1D) is commonly attributed to sp² N atoms involved in triazine rings,^35,37 while N2 peak (at 399.6 eV) and N3 peak (at 401 eV) are assigned to bridging N atoms in N–(C)₃ (ref. 37) and N atoms bonded with H atoms,³⁶ respectively. The weak N4 peak (at 405.3 eV) can be attributed to the charging effects or positive charge localization in heterocycles and the cyano-group.^32,37 These assignments are in good agreement with those reported for bulk g-C₃N₄ powder (Fig. S4†), suggesting that the chemical states and the coordination of carbon and nitrogen in the g-C₃N₄ are retained during the “pillared” process by SiO₂ NPs.

Elemental analysis (EA) was performed to determine the C/N mass ratio in the samples (Table S1†). EA of the α sample series give a C/N molar ratio of 0.66, 0.66, 0.74, and 0.77 with increasing SiO₂ contents, indicating excessive amounts of C in Si-8α-CN and Si-11α-CN. This is possibly due to the incomplete hydrolysis of TEOS, leading to the increase of carbon contents with increasing SiO₂ contents. The percentage of g-C₃N₄ in samples changed with the “pillared” SiO₂ contents, and the trend was analyzed and shown in Fig. S3.† Luckily, a linear relationship between the reciprocal of the percentage content of g-C₃N₄ calculated with EA date and the mass ratio of SiO₂ used in the experiments can be observed. The donated quantities of CNS were consistent during the preparation process of the various mesoporous materials, the reciprocal of the percentage content of g-C₃N₄ is related to the SiO₂ contents of the obtained materials (Fig. S5†). Hence, the linear relations mentioned above exhibits the direct ratio relationship between SiO₂ NPs used and the SiO₂ contents of the obtained materials, indicating that essentially all of the SiO₂ NPs were wrapped by CNS.

TEM was performed to further investigate the morphology and microstructure of Si-5α-CN. As shown in Fig. 2 and S6,† the sample with lateral sizes of hundreds of nanometers is observed, which coincides with the CNS without SiO₂ NPs (Fig. S4†). The nature of the structure can be further unraveled by the element mapping of silicon, carbon, nitrogen, and oxygen in the sample (Fig. 2a–d), which indicates a homogeneous dispersion of C, Si, N, and O in Si-5α-CN. Further, the SiO₂ NPs show uniform size distribution, suggesting that the CNS whose solubility changed with the variation of pH value plays a vital role in preventing the agglomeration of SiO₂ NPs during the formation of the mesoporous material.

Fig. 2 TEM images of Si-5α-CN (A and B). Typical scanning transmission electron microscopy (STEM) image and corresponding elemental mapping images of (a) nitrogen, (b) silicon, (c) carbon, and (d) oxygen in the selected area (red rectangle in C).

Combing the results of XRD, XPS, FT-IR and TEM analysis, the typical electronic structure, chemical composition, and tri-s-triazine related bonding of CNS are well established in the obtained materials. Pore structure can be formed when CNS and SiO₂ sol were mixed together, which may impart the materials outstanding physical and chemical properties.

The nitrogen adsorption–desorption isotherms were performed to demonstrate pore structure and pore size distribution of SiO₂-xN/g-C₃N₄. Type-IV curves with a H₃ hysteresis loop (Fig. 3A and C) are observed suggesting that flat type slit pores were formed in all samples. The corresponding pore size distribution curves are shown in Fig. 3B and D. The surface areas, pore volumes, and average pore sizes determined from BET isotherms are summarized in Table S2.† With the increase of silica contents, the sample surface areas increase significantly. The low surface area observed for the sample with low SiO₂ content could be due to the incompletely pillared effect on CNS. In addition, excessive dosage of silica cannot bring a significant enhancement in surface area and can inhibit the function of g-C₃N₄. Therefore, appropriate g-C₃N₄/SiO₂ ratio is the key to achieving superior “pillared” mesoporous materials. Two types of SiO₂ NPs were synthesized to further reveal the relations between nature of SiO₂ NPs and the properties of the obtained materials (size distribution can be derived from Fig. S7†). Compared to the α sample series, the β sample series show smaller pore volume. This can be explained by the fact that smaller silica contributed poorly to column brace effect. The pore sizes of α and β mesoporous materials are inclined to diminish and eventually disappeared as the SiO₂ content increased (Fig. 3B and D). This may be caused by the fact that excessive flat particles arranged in a more disordered way. So SiO₂-5α-CN was selected as the typical example to perform the following property test. These results indicate that the pore size of the samples can be controlled by the size and contents of the silica used. The as-prepared materials combine the advantages of both CNS and mesoporous properties.

Fig. 3 N₂ adsorption–desorption isotherms (A and C) and pore size distributions (B and D) of Si-xN-CN.

It is important to test its thermal stability for the applications of materials. Temperature programmed heating of Si-5α-CN was carried out and the obtained TG-DSC results are shown in Fig. S8.† Almost no weight loss appeared below 400 °C while a 30% weight loss is observed in the temperature ranges from 400 to 630 °C. It can be concluded that the decomposed of CNS was at the beginning of 400 °C compared to the bulk sample at 630 °C, which may be caused by the “flatting effect” discussed above. The TG results reveal that the materials can be potentially used below 400 °C.

To evaluate potential application of the as-prepared materials, the adsorption capacities for the heavy metal ions and organic pollutants malachite green (MG) were investigated. The adsorption efficiencies of the SiO₂-5α-CN, commercial activated carbon and bulk g-C₃N₄ for Cu²⁺ and Cd²⁺ were compared under the same operation conditions. As shown in Fig. 4A, the adsorption efficiency for Cu²⁺ and Cd²⁺ can reach greater than 99.9% for SiO₂-5α-CN, which means that the contents of metal ions can be lowered down to ppb level. It is obvious that the constructed porous material has better adsorption efficiency for heavy metal ions than bulk g-C₃N₄ and activated carbon. UV-visible (UV-vis) absorption spectroscopy was used to test the adsorption performance for MG (617 nm). Fig. 4B and C and S9† show that Si-5α-CN has better adsorption capacity than bulk g-C₃N₄ and activated carbon. Moreover, as shown in Fig. 4D, in the first 10 min, a sharp variation in removal ratios can be observed and deemed as a quick adsorption stage. The subsequent second stage between 10 and 30 min displays a smaller removal ratio (10%). Almost no obvious change can be seen after 30 min. The adsorption capability of Si-5α-CN has also been assessed with the cationic dyes methylene blue (MB) (Fig. S10†). The improved adsorption ability of the as-prepared material can be attributed to the combination of the advantages of g-C₃N₄ and mesoporous structure. In the case of the adsorption of metal ions, the excellent adsorption capacity may be attributed to the strong coordination ability of tri-s-triazine and the uniform nitrogen pore of the sample (Scheme 2). It is reported that g-C₃N₄ exhibits appropriate microstructure, surface terminations with defects, and nitrogen atoms which can be viewed as the active sites for anchoring metal ions.^16,38 Furthermore, the uniform pore structure and high BET surface area of Si-5α-CN are beneficial for pollutants flowing through the pore and being captured by CNS via coordination. In the case of organic pollutants adsorptions, the high adsorption ability of Si-5α-CN may be largely attributed to the π–π conjugation since g-C₃N₄ nanosheet is similar to graphene (Scheme 2). Overall, the ordered porous structure and high BET surface area of the CNS/SiO₂ composites can facilitate the function of each component. Moreover, SiO₂ NPs may be replaced with other substances such as some biological molecules, quantum dots or TiO₂ NPs to integrate the advantages of both g-C₃N₄ and the “pillared” components, resulting in various novel bifunctional materials.

Fig. 4 A representatives adsorption efficiency of metal ions of Si-5α-CN (a), bulk-g-C₃N₄ (b), and activated carbon (c). Absorption spectra (B and C) and adsorption rate (D) of an aqueous solution of MG after treatment with adsorbents at different time intervals.

Scheme 2 Adsorption mechanism of SiO₂/CNS for metal ions and organic pollutants.

In summary, by using “pillared” method without any surfactant and hard template, we have successfully fabricated functional mesoporous SiO₂/CNS composites in which the nature of g-C₃N₄ was maintained as compared to bare CNS. The surface areas and pore volumes of the as-prepared materials can be tuned by altering the size and quantity of SiO₂. Combining the advantages of both g-C₃N₄ and ordered porous structures, the prepared SiO₂/CNS composites exhibit excellent performance in removing inorganic or organic pollutants, and can be stabilized up to 400 °C, which warrant their potential application in energy and environmental fields, such as in catalysis, adsorbent, and so on.

Notes and references

1. A. Corma, Chem. Rev., 1997, 97, 2373–2420.
2. S. H. Joo, S. J. Choi, I. Oh, J. Kwak, Z. Liu, O. Terasaki and R. Ryoo, Nature, 2001, 412, 169–172.
3. A. Vinu, V. Murugesan, O. Tangermann and M. Hartmann, Chem. Mater., 2004, 16, 3056–3065.
4. A. M. Liu, K. Hidajat, S. Kawi and D. Y. Zhao, Chem. Commun., 2000, 1145–1146.
5. Z. Sun, Y. Deng, J. Wei, D. Gu, B. Tu and D. Zhao, Chem. Mater., 2011, 23, 2176–2184.
6. M. Vallet-Regí, L. Ruiz-González, I. Izquierdo-Barba and J. M. González-Calbet, J. Mater. Chem., 2006, 16, 26–31.
7. J. Liu, C. Li and F. Li, J. Mater. Chem., 2011, 21, 7175.
8. S. Tanaka, N. Nishiyama, Y. Egashira and K. Ueyama, Chem. Commun., 2005, 2125–2127.
9. C. Liang, K. Hong, G. A. Guiochon, J. W. Mays and S. Dai, Angew. Chem., Int. Ed., 2004, 43, 5785–5789.
10. Y. Meng, D. Gu, F. Zhang, Y. Shi, H. Yang, Z. Li, C. Yu, B. Tu and D. Zhao, Angew. Chem., Int. Ed., 2005, 44, 7053–7059.
11. F. Zhang, Y. Meng, D. Gu, Y. Yan, C. Yu, B. Tu and D. Zhao, J. Am. Chem. Soc., 2005, 127, 13508–13509.
12. Y. Meng, D. Gu, F. Zhang, Y. Shi, L. Cheng, D. Feng, Z. Wu, Z. Chen, Y. Wan, A. Stein and D. Zhao, Chem. Mater., 2006, 18, 4447–4464.
13. S. Jun, S. H. Joo, R. Ryoo, M. Kruk, M. Jaroniec, Z. Liu, T. Ohsuna and O. Terasaki, J. Am. Chem. Soc., 2000, 122, 10712–10713.
14. W. Chen, S. Li, C. Chen and L. Yan, Adv. Mater., 2011, 23, 5679–5683.
15. H. Sun, Z. Xu and C. Gao, Adv. Mater., 2013, 25, 2554–2560.
16. Y. Wang, X. Wang and M. Antonietti, Angew. Chem., Int. Ed., 2012, 51, 68–89.
17. Y. Wang, J. Zhang, X. Wang, M. Antonietti and H. Li, Angew. Chem., Int. Ed., 2010, 49, 3356–3359.
18. Y. Wang, H. Li, J. Yao, X. Wang and M. Antonietti, Chem. Sci., 2011, 2, 446–450.
19. X. Zhang, X. Xie, H. Wang, J. Zhang, B. Pan and Y. Xie, J. Am. Chem. Soc., 2013, 135, 18–21.
20. X. Zhang, H. Wang, H. Wang, Q. Zhang, J. Xie, Y. Tian, J. Wang and Y. Xie, Adv. Mater., 2014, 26, 4438–4443.
21. T. Ma, Y. Tang, S. Dai and S. Qiao, Small, 2014, 10, 2382–2389.
22. J. Zhang, M. Zhang, L. Lin and X. Wang, Angew. Chem., Int. Ed., 2015, 54, 6297–6301.
23. J. Xu, L. Zhang, R. Shi and Y. Zhu, J. Mater. Chem. A, 2013, 1, 14766–14772.
24. J. Zhang, X. Chen, K. Takanabe, K. Maeda, K. Domen, J. D. Epping, X. Fu, M. Antonietti and X. Wang, Angew. Chem., Int. Ed., 2010, 49, 441–444.
25. F. Su, S. C. Mathew, L. Mohlmann, M. Antonietti, X. Wang and S. Blechert, Angew. Chem., Int. Ed., 2011, 50, 657–660.
26. X. Wang, K. Maeda, A. Thomas, K. Takanabe, G. Xin, J. M. Carlsson, K. Domen and M. Antonitti, Nat. Mater., 2009, 8, 76–80.
27. J. Liu, Y. Liu, N. Liu, X. Zhang, H. Huang, Y. Lifshitz, S. T. Lee, J. Zhong and Z. Kang, Science, 2015, 347, 970–974.
28. P. Niu, L. Zhang, G. Liu and H.-M. Cheng, Adv. Funct. Mater., 2012, 22, 4763–4770.
29. S. Yang, Y. Gong, J. Zhang, L. Zhan, L. Ma, Z. Fang, R. Vajtai, X. Wang and P. M. Ajayan, Adv. Mater., 2013, 25, 2452–2456.
30. X. Du, G. Zou, Z. Wang and X. Wang, Nanoscale, 2015, 7, 8701–8706.
31. H. Jiu, W. Jia, L. Zhang, C. Huang, H. Jiao and J. Chang, J. Porous Mater., 2015, 22, 1511–1518.
32. R. C. Dante, P. Martín-Ramos, A. Correa-Guimaraes and J. Martín-Gil, Mater. Chem. Phys., 2011, 130, 1094–1102.
33. J. Xu, H. T. Wu, X. Wang, B. Xue, Y. X. Li and Y. Cao, Phys. Chem. Chem. Phys., 2013, 15, 4510–4517.
34. M. Sadhukhan and S. Barman, J. Mater. Chem. A, 2013, 1, 2752.
35. A. Vinu, Adv. Funct. Mater., 2008, 18, 816–827.
36. Y. Li, J. Zhang, Q. Wang, Y. Jin, D. Huang, Q. Cui and G. Zou, J. Phys. Chem. B, 2010, 114, 9429–9434.
37. Y. Cui, J. Zhang, G. Zhang, J. Huang, P. Liu, M. Antonietti and X. Wang, J. Mater. Chem., 2011, 21, 13032.
38. X. Wang, X. Chen, A. Thomas, X. Fu and M. Antonietti, Adv. Mater., 2009, 21, 1609–1612.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra26192a

---