Photocatalytic water splitting using semiconductors illuminated with visible light has been considered a green route to utilize the abundant and renewable solar energy. Among the numerous types of photocatalysts used for hydrogen production, graphitic carbon nitride (g-C N ) has shown a great potential for practical applications due to its extraordinary features, such as environmental friendliness, cost-effectiveness, as well as physical and chemical stability. In addition, it has excellent electronic and optical properties. However, the industrial application of g-C N is limited by the high recombination rate of photoinduced electron-hole pairs, low specific surface area, and limited active sites [1,2]. Many approaches have been developed to overcome the deficiencies of g-C N , such as doping with foreign elements [3,4], modification with functional groups , manufacturing different g-C N nanostructures [6–12], and constructing g-C N -based nanocomposites . It is well known that g-C N could be obtained by condensation of various nitrogen containing precursors (melamine, cyanimide, dicyanamide, urea, and thiourea) at a certain temperature (450–650 C). Recent researches have shown that the precursor could maintain its micro-morphology after polymerization at a relatively low temperature (380–450 C). However, the influence of pre-polymerization in the structure and photocatalytic properties of g-C N have not been studied. Herein, we firstly developed a simple pre-polymerization method to obtain porous g-C N nanosheets. This method is facile, environment friendly, and does not need any soft or hard template. Remarkably, the g-C N pre-polymerized at 350 C exhibits dramatically enhanced photocatalytic efficiency for hydrogen evolution under visible light.
Bulk g-C N (CN) was synthesized by direct heating 4g melamine (Me) at 550 C for 4h in air at a ramping rate of 5 C/min. The porous g-C N nanosheets were synthesized by a facile pre-polymerization method. First, 4g Me was heated at 350 C for 2h at a ramping rate of 5 C/min. After cooled down to room temperature, the product was ground into fine powder (designated as Me-350). Finally, 2g Me-350 was allowed to fully polymerize at 550 C for 2h at a ramping rate of 5 C/min. After cooling down to room temperature, the product was collected and designated as CN-350. In order to figure out the optimal pre-polymerization temperature, Me was also heated at 300, 400, and 450 C with the following procedure identical to that of Me-350. Here, the obtained products are marked as Me-300, Me-400, and Me-450, in sequence. The resulting g-C N are designated as CN-300, CN-400, and CN-450, respectively. The crystal structure was analyzed by powder X-ray diffraction (XRD). The chemical structure was investigated with Fourier transform infrared spectroscopy (FT-IR). The morphology of the photocatalysts was observed by field emission scanning electron microscopy (FE-SEM) and transmission electron microscopy (TEM). The photocatalytic activity of the catalysts was evaluated by hydrogen evolution and degradation of organic dyes under visible light irradiation. In the hydrogen evolution test, 50 mg of catalyst was dispersed in a solution containing 10 mL triethanolamine and 90 mL deionized water. Pt (1 wt%) was in-situ photo-deposited on the catalysts using H PtCl as a precursor. The system was evacuated for 5 min to remove air, then irradiated with a 300 W Xeon lamp with a UV cut-off filter (λ 400 nm). An external water cooling circulation system was employed to maintain the temperature of the reactant system at 10 C. The amount of gas evolved was detected by a gas chromatography device. A series of UV-vis filters was applied to test the dependence of hydrogen evolution rate on wavelength.
Figure 1 shows the scheme for the synthesis of bulk g-C N (CN) and the porous g-C N nanosheets (CN-350). CN is synthesized by directly heating Me at 550 C for 4h. For the synthesis of CN-350, Me is first heated at 350 C to pre-polymerize. Then, the pre-polymerized melamine (Me-350) is ground into powder to insure a highly exposed area to the air in the final condensation. Finally, the ground Me-350 is heated to 550 C and kept for 2h to get CN-350. The XRD patterns of Me pre-heated at different temperatures are shown in Figure 2(a). Me-300 displays the same XRD peaks as Me, which means polymerization would not happen at 300 C. The appearance of new peaks in the intermediate products, when Me is heated at 350 C or a higher temperature, is direct evidence of the polymerization process. The intensity of FTIR peaks representing N-H (wave number: 3000–3650) and C-NH bond (wave number: 460–790) for Me-350 (Figure 2(b)) are greatly weakened compared with that of Me. It further indicates that some of the amino groups in Me have been destroyed and the pre-polymerization process would happen, when Me is heated at 350 C. Due to the pre-polymerization process, the ground powder would not aggregate into a macroscopic bulk, but stay in fine powder, when it is heated to 550 C again. Thus, pores are formed, when the pre-polymerized precursor undergoes full polymerization with sufficient exposure to air. The volume of CN-350 sample is about four times larger than that of the CN with the same weight (100 mg), demonstrating that CN-350 becomes fluffy, when a pre-polymerization procedure is introduced before the final condensation process.
The crystal structures of as-synthesized g-C N are analyzed by XRD. No shift of the peaks is observed, indicating that the layered structure of g-C N is largely preserved in the pre-polymerized samples. Two diffraction peaks at 13.1 and 27.3 in the XRD patterns are ascribed to the in-plane trigonal N linkage of tri-s-triazine units and the stacking of the conjugated aromatic structure, respectively. Interestingly, the intensity and full width at half maximum (FWHM) of (002) peak are greatly changed in the pre-polymerized samples as shown in Figure 2(d). CN-350 shows the lowest intensity and highest FWHM among all the samples, indicating that the long-term order in the stacking of graphitic layers of CN-350 has been greatly weakened. The intensity of the peak at 13.1 is also decreased, which is mainly associated with the large amount of in-plan holes in CN-350 (see the subsequent SEM and TEM images). Change in the XRD characteristics clearly demonstrates that the pre-polymerization procedure can modify the structure of g-C3N4 and exfoliate bulk g-C3N4 into porous nanosheets by a bottom-up method.
The morphology and porous structure of the samples are characterized by SEM. As observed in Figure 3, CN accumulates in dense solid agglomerates with the size of several micrometers. CN-300 shares similar micromorphology features as that of CN. However, the surface of CN-350 is mostly corroded after the pre-polymerization treatment. In addition, CN-350 is composed of nanosheets with irregular porous structures, which is caused by the final condensation of Me-350, when it is sufficiently exposed to the air and endures a long full-polymerization period. The diameter of pores varies from tens nanometers to several hundred nanometers. It could be noticed that a small amount of porous bulk g-C3N4 is also formed in CN-350. However, the surface of the bulk material in CN-350 is much more rough than that in CN and CN-300. The size of the bulk is also much smaller (about 1 μm in the in-plane direction). The sheet-like morphology and rough surface could promise a high specific surface area in CN-350. CN-400 and CN-450 also display a porous trait; besides, the amount of pores is decreased with the rising of pre-polymerization temperature. The porous bulk content becomes predominant in CN-400 and CN-450. In other words, the morphology of the as-prepared g-C3N4 turns to porous bulk at a higher pre-polymerization temperature.
TEM is used for further studying the morphology of the as-synthesized samples. As displayed in Figure 4, CN and CN-300 contain dense stacking layers with flat surfaces, whereas CN-350 consists of thin layers with curly edges. The enlarged TEM images of CN and CN-350 in Figure 5 clearly demonstrate that pores with a diameter of 20–50 nm can be induced by the pre-thermal treatment of Me. The porous structure in g-C N nanosheets could greatly enhance the photocatalytic performance due to the increased exposed active edges. The dark nanosheet morphology declares that CN-450 has thicker stacking layers than CN-400 (Figure 4(e)(f)). It is because Me-400 is less stable than Me-450; thus, it undergoes a longer polymerization process and has more mass loss, when fully polymerized at 550 C.
The photocatalytic performance of the pre-polymerized g-C N is evaluated by photocatalytic hydrogen evolution reaction (HER) under visible light irradiation. The photocatalytic H evolution performance of CN and pre-polymerized g-C N is displayed in Figure 5(a). CN-350 can produce 349.8 μmol H in 5h, which is 11 times more than that of CN (31.4 μmol). Owing to the porous structure, increased surface area and possibly more abundant active catalytic sites achieved, CN-350 displays much higher photocatalytic activity than CN. The photocatalytic performance of g-C N pre-polymerized at different temperature is shown in Figure 5(b). The photocatalytic hydrogen evolution rate of g-C N pre-heated at 300, 350, 400, and 450 C is 7.15, 69.69, 53.87, and 37.64 μmol/h, respectively. CN-300 exhibits nearly the same photocatalytic property as CN. It could be ascribed to the fact that melamine is stable at 300 C, thus pre-heating at 300 C has no obvious impact on the structure and photocatalytic performance of g-C N . CN-350 shows the best performance among all the samples, probably due to the longest polymerization time endured by Me-350, when it is fully condensed to the layered graphitic structure. The photocatalytic performance decreases as the pre-polymerization temperature further rises. A higher pre-polymerization temperature leads to a shorter full condensation process later. Longer full polymerization time promises a higher pore volume, a larger specific surface area, and a better photo-catalytic activity. Thus, CN-350 has the best photocatalytic performance among all the samples. The photocatalytic hydrogen evolution performance of CN-350 is superior to most of the recently reported g-C N . The sustainable capability of CN-350 is evaluated by four consecutive operations, as depicted in Figure 5©. It can be observed that CN-350 displays excellent stability for photocatalytic hydrogen evolution, which shows only a deterioration of 5.4% after four cycles of reaction. The CN-350 photocatalyst collected after four cycles of the photocatalytic reaction exhibits the same crystal structure as that before the reaction. All the aforementioned results proved the excellent recycling performance of CN-350. The HER of CN and CN-350 is also evaluated under a series of irradiation wavelengths (350, 400, 420, 450, 500, and 550 nm) in the UV-vis light region. The results show a coincident relationship with the UV-vis absorption spectra, indicating that the absorbed photons are the main driving force for photocatalytic water splitting. The enhanced photocatalytic performance is observed across the whole absorption spectrum in CN-350 catalysts. Interestingly, when the irradiation wavelength reached 450 and 500 nm, the HERs of CN-350 are 2.91 and 0.22 μmol/h, respectively. However, only trace amount of hydrogen can be detected at 450 nm for CN sample and it exhibits no photocatalytic activity at 500 nm. This may be resulted from the multiple-reflection effect and high photoenergy efficiency of CN-350.
A porous g-C N was synthesized by a simple pre-polymerization method. The porous g-C N (CN-350) with a high surface area and a large pore volume exhibited greatly enhanced photo-catalytic performance compared with pristine g-C N . The enhanced photocatalytic activity was primarily attributed to the increased surface area and the porous structure, which could provide abundant active sites and cross-plane diffusion channels to facilitate the mass and charge transportation. Additionally, this unique structure could improve the electron mobility and suppress the recombination of photo-induced carriers. The CN-350 photocatalyst with high yield, high efficiency, and stable activity, holds great promise in utilizing solar energy for fuel generation and other optical applications.