Metal-free C60/CNTs/g-C3N4 ternary heterostructures: synthesis and enhanced visible-light-driven photocatalytic performance

A metal-free C60/CNTs/g-C3N4 nanoheterostructure with excellent visible-light photocatalysis for rhodamine B (Rh B) degradation has been reported. Via a convenient low-temperature solution-phase method, g-C3N4 nanosheets can serve as substrate for dispersion of C60/CNTs. The loading of C60/CNTs onto g-C3N4 nanosheets surfaces significantly enhanced visible-light-driven photocatalytic activity of g-C3N4 catalyst, for oxidation of organic pollutant (Rh B, 100%). Excellent photocatalytic properties of C60/CNTs/g-C3N4 can be predominantly attributed to the intimate interfacial contact among constructing compounds, increased specific surface area and enhanced light adsorption efficiency resulted from C60/CNTs carbon materials. Particularly, the synergistic heterostructure interaction remarkably hinders the electron–hole pairs recombination, giving rise to significantly enhanced photocatalytic performance of C60/CNTs/g-C3N4 in comparison with other counterparts.


Introduction
Photocatalysis employing semiconductor materials has been considered as a promising method for widespread solar energy utilization and environmental remediation [1][2][3]. During the past few decades, many research works on photocatalysis field are mainly focused on the development of metaloxide-based semiconductors, owing to their high thermal and chemical stability [4,5]. Nevertheless, the wide band gap and low quantum efficiency of some metal-oxide-based semiconductors significantly limit their potential applications

. Preparation of C 60 /CNTs
C 60 -decorated single-walled carbon nanotubes (C 60 /CNTs) was prepared by using the Zhang' method [16]: SWCNTs (50 mg) and C 60 (50 mg) were dispersed by ultrasonication in toluene (100 ml) for 1 h, and then stirred at room temperature for 12 h. After volatilization of the toluene, the resultant black powder was washed with ethanol several times and dried in vacuum at 80°C for 10 h (denoted as C 60 /CNTs).

Preparation of C 60 /CNTs/g-C 3 N 4
C 60 /CNTs/g-C 3 N 4 ternary composite was fabricated via a facile hydrothermal method. Especially, C 60 /CNTs (0.05 g), g-C 3 N 4 (0.05 g) powders were added into ethanol solution (15 ml). The suspension was ultrasonicated for 1 h. The mixed solution was then transferred to a 25 ml Teflon-lined stainless steel autoclave reactor and heated at 180°C for 3 h. The C 60 /CNTs/g-C 3 N 4 sample was then collected. The C 60 /g-C 3 N 4 and CNTs/g-C 3 N 4 binary composite was synthesized through a similar method.

Characterization of photocatalysts
The morphology of the samples was studied by SEM (JSM-7800F) and TEM (JEM-2100F). The crystal structure of the samples was tested with XRD (Rigaku, D/max 2500) equipped with Kα radiation. Fourier transform infrared (FTIR) spectra of the samples were recorded on infrared spectrophotometer (America Perkin Elmer, Spectrum One). UV-vis diffuse reflection spectroscopy (DRS) was performed on a scan UV-vis spectrophotometer (SHIMADZU, UV-2550). The X-ray photoelectron spectroscopy (XPS, VGScientific) using 300 W Al Kα radiation as the excitation source was employed for the surface analysis of the prepared C 60 /CNTs/g-C 3 N 4 nanocomposite. The surface areas were studied by the nitrogen adsorption Brunauer-Emmett-Teller (BET) method (BET/BJH Surface Area, 3H-2000PS1). The photoluminescence (PL) spectra of the photocatalysts were measured on a F4500 (Hitachi, Japan) photoluminescence detector with an excitation wavelength of 325 nm.

Photocatalytic activities studies
Photocatalytic activities of the photocatalysts were evaluated by testing the degradation efficiencies of Rh B, using a Xe lamp (500 W), which was equipped with a UV irradiation source filter (λ > 420 nm). In a standard photocatalytic experiment, the photocatalyst (50 mg) was dispersed in a mixed aqueous solution of Rh B (50 ml, 0.01 mmol l −1 ), and the suspension was stirred for 40 min in the dark to ensure that it reached the adsorption-desorption equilibrium. Then, the suspension was illuminated by visible light. At 5 min intervals, a 4.0 ml of the suspension was withdrawn, and centrifuged to remove the catalyst. The concentration of Rh B was analysed by UV-vis spectrophotometry.

Results and discussion
The phase and structure of photocatalysts were studied by XRD studies. As shown in figure 1, the diffraction peaks at 27.5°for g-C 3 N 4 are detected in the pattern of g-C 3 N 4 . The characteristic peaks of the C 60 /g-C 3 N 4 hybrid were respectively attributed to g-C 3 N 4 (27.5°) [18], and C 60 (10.7°, 17.7°, 20.7°a nd 21.7°), indicating the successful formation of the homogeneous hybrid structure [15]. Alternatively, there is also no difference observed between the C 60 /g-C 3 N 4 and C 60 /CNTs/g-C 3 N 4 , probably due to low loading amount and the weak diffraction of CNTs [19,20]. Figure 2 depicts the microstructures of C 60 /CNTs/g-C 3 N 4 composite. As shown in figure 2a, g-C 3 N 4 has layered structure composed of tightly stacked g-C 3 N 4 nanosheets with sizes of the thickness of tens of nanometres. As shown in figure 2b,c, some CNTs with an average diameter of around 10 nm can also be observed, which are attached on the sheet-like g-C 3 N 4 . C 60 molecules with a diameter of approximately 0.7 nm are too small to be directly resolved by TEM observation, which is similar to the previous report [16].
The XPS measurement was performed to determine the chemical composition and valence state of various species. The C 1s (figure 3a) high-resolution spectrum in C 60 /CNTs/g-C 3 N 4 nanocomposite and g-C 3 N 4 both have two distinct peaks at around 282.6 and 286.2 eV. The peak at 286.2 eV is identified as sp 2 -bonded carbon (N-C=N). The peak located at 284.6 eV could be assigned to adventitious carbon, C-C bond from the C 60 and graphitic carbon [16,21]. The high-resolution XPS spectrum for N 1s in the g-C 3 N 4 could be fitted into two peaks at 396.7 and 399.0 eV (figure 3b), corresponding to the pyridinic N species in triazine rings (C-N=C) and tertiary N species or N atoms bonded with H atoms [22]. In the case of C 60 /CNTs/g-C 3 N 4 , the binding energy of N 1s shifted about 0.2 eV, toward higher binding energies, which was caused by the strong interaction between the CNTs and the N atoms in the g-C 3 N 4 [23].
FTIR spectroscopy was further applied to identify the existence of C 60 and g-C 3 N 4 in the C 60 /CNTs/g-C 3 N 4 composite (figure 4a,b). For pure g-C 3 N 4 (figure 4a), the peaks at 1239, 1318, 1410, 1575 and 1633 cm −1 feature typical aromatic C-N heterocycle stretches, and the peaks at 811 cm −1 are corresponding to the tri-s-trizine units of g-C 3 N 4 [23]. It also can be seen in figure 4a that for pure C 60 , the bands at 526, 575, 1180, 1428 and 1540 cm −1 are attributed to the internal modes of the C 60 molecule [16]. For C 60 /CNTs/g-C 3 N 4 , both the characteristic peaks of g-C 3 N 4 and C 60 can be clearly observed, identifying the successful incorporation of C 60 and g-C 3 N 4 in the ternary composite (figure 4b). Figure 5a shows the UV-vis diffuse reflectance spectra (DRS) of the as-fabricated samples. As can be seen, g-C 3 N 4 demonstrated strong absorption ability in visible region. Visible-light harvesting ability of C 60 /g-C 3 N 4 and CNTs/g-C 3 N 4 was substantially enhanced with adsorption edge red shift to 500 and 550 nm, respectively. The result shows that integration of C 60 or CNTs in the nanocomposites could improve the light harvesting efficiency of g-C 3 N 4 . In addition, the loading of C 60 /CNTs in the g-C 3 N 4 leads to a further enhancement in the absorption edge (approx. 530 nm), which reveals the key role of CNTs for promoting the light harvesting and thus leading to the excitation of more electrons over the C 60 /CNTs/g-C 3 N 4 nanocomposite under visible light [23].
The PL spectrum was employed to study the charge-carriers separation/recombination behaviour of photocatalysts. As shown in figure 5b, the PL emission peaks at around 460 nm for the samples reveals that the π-conjugated system of g-C 3 N 4 was not destroyed by the modification treatment [24]. The decrease of PL intensity of C 60 /CNTs/g-C 3 N 4 ternary nanocomposites in comparison with pure   decrease in the BET-specific surface, which may be due to the formation of C 60 clusters in the mesopores of g-C 3 N 4 or CNTs. It should be noted that the loading of CNTs onto g-C 3 N 4 could lead to a significant increase in the BET-specific surface. In addition, the surface area of C 60 /CNTs/g-C 3 N 4 composite is obviously smaller than those of CNTs/g-C 3 N 4 , which may be due to the addition of C 60 . Figure 6a displays the photocatalytic activities of bare g-C 3 N 4 , C 60 , CNTs, C 60 /CNTs, CNTs/g-C 3 N 4 , C 60 /g-C 3 N 4 and C 60 /CNTs/g-C 3 N 4 for the degradation of Rh B solution. Only 27.0%, 6.0% and 23.0% of Rh B were removed by pure g-C 3 N 4 , C 60 , CNTs, respectively, after 60 min irradiation, while 41.0%, 55.0% of Rh B by C 60 /g-C 3 N 4 and CNTs/g-C 3 N 4 , approximately. It means that C 60 or CNTs loading is beneficial for the improvement of photocatalytic activity of the composite sample. Meanwhile, the photocatalytic efficiency of C 60 /CNTs also rose to 72.0%. The photocatalytic efficiency of C 60 /CNTs/g-C 3 N 4 reached 100%, which is remarkably faster than the rate of bare g-C 3 N 4 , C 60 , CNTs, C 60 /CNTs, CNTs/g-C 3 N4 and C 60 /g-C 3 N 4 . It is due to that efficient heterostructure interface among three components that can restrain the recombination of photo-induced charges effectively [25]. In addition, the C 60 /CNTs coating can improve the visible-light absorption efficiency (figure 5a) and BET-specific surface (table 1) of the photocatalyst, which are both beneficial for the ternary composite to photolyse Rh B. In addition, the effect of C 60 amount on the photocatalytic activity of the composites was also investigated (figure 6b). Furthermore, the highest degradation rate was obtained from C 60 (50 wt %)/CNTs/g-C 3 N 4 sample with almost 100% of Rh B removal. This increase may be attributed to the capturing of electrons by the deposited C 60 to hinder the recombination of hole-electron pairs [26,27], whereas the decrease may result from the blocking of visible light by the over-deposited C 60 . In order to further understand the photocatalytic mechanism of the ternary nanocomposite, the trapping experiment of active species was also investigated. As shown in figure 6c, isopropanol (IPA), triethanolamine (TEOA) and p-benzoquinone (BQ) were, respectively, introduced as the scavengers of hydroxyl radicals (·OH), holes (h + ) and superoxide radicals (·O 2− ) to examine the effects of reactive  species on the photocatalytic degradation of Rh B. It can be seen that BQ and IPA led to a remarkable suppression of the degradation rate of Rh B, whereas EDTA-2Na exhibited a weaker restraining effect on the degradation rate. The results confirmed that ·OH and ·O 2− play a more important role than h + in the photocatalytic degradation of Rh B. The stability and reusability of C 60 /CNTs/g-C 3 N 4 for photocatalytic degradation of Rh B were evaluated by repeating the photocatalytic experiments under the same conditions for three cycles. It is found that no obvious deactivation in the photocatalytic activity is detected for the C 60 /CNTs/g-C 3 N 4 composite after continuous visible-light irradiation (figure 6d). Figure 7 shows the SEM image of the C 60 /CNTs/g-C 3 N 4 sample after the photocatalytic reaction. After four catalytic runs, the secondary C 60 /CNTs/g-C 3 N 4 nanostructures were still very intact, clearly showing their stabilities. Thus, it represents a promising kind of composite catalyst for Rh B degradation under visible light.
Based on the above analysis, a schematic illustration of the photocatalysis mechanism of the C 60 /CNTs/g-C 3 N 4 is displayed in figure 8. Under visible-light illumination, photo-induced electrons are excited from the valence band (VB) of g-C 3 N 4 to the conduction band (CB), resulting in electronhole pairs. The CB of g-C 3 N 4 (-1.12 eV versus NHE) is less negative than that of the CNTs (-0.3 eV versus NHE) or C 60 (−0.2 eV versus NHE) [28,29]; therefore, photoelectrons from the CB of g-C 3 N 4 are captured by the CNTs or C 60 . As a result, the photoactivities of CNTs/g-C 3 N 4 , C 60 /g-C 3 N 4 have been  CNTs/g-C 3 N 4 C 60 /CNTs/g-C 3 N 4 50wt% C 60 /CNTs/g-C 3 N 4 25wt% C 60 /CNTs/g-C 3 N 4 75wt% C 60 /CNTs/g-C 3 N 4 C 60  enhanced compared with the bare g-C 3 N 4 . For the C 60 /CNTs/g-C 3 N 4 nanocomposite, there are two possible ways for the transformation of photo-generated charges: (i) the electrons may firstly transfer to C 60 and then to CNTs, due to that, as the strong electron acceptor, the fullerenes can capture the electrons, which are spread and moved along the CNTs and fullerenes, thus facilitating the photogenerated carrier separation; (ii) the electrons in the CB of g-C 3 N 4 may firstly transfer to CNTs, and then to C 60 . Consequently, the carrier recombination can be effectively inhibited. In this way, the excited

Conclusion
In summary, we have demonstrated that C 60 /CNTs/g-C 3 N 4 composite was synthesized by a simple hydrothermal technology. The results showed that the C 60 /CNTs/g-C 3 N 4 nanocomposite exhibited the best photocatalytic activity under visible light, which is almost five times higher than that of pure g-C 3 N 4 . It shows that the ternary heterostructure can effectively hinder the recombination of the photogenerated electron-hole pairs, and thus improve the photocatalytic performance of the C 60 /CNTs/g-C 3 N 4 nanocomposite. In this work, the enhancement of photocatalytic activity of C 60 /CNTs/g-C 3 N 4 could be achieved via the loading C 60 /CNTs, which may provide a promising class of photocatalyst candidates for photocatalytic application.
Data accessibility. The datasets supporting this article have been uploaded as the electronic supplementary material.