Improving Photocatalytic Activity for Formaldehyde Degradation by Encapsulating C60 Fullerenes into Graphite-like C3N4 through the Enhancement of Built-in Electric Fields

The photocatalytic degradation of formaldehyde by graphite-like C3N4 is one of the most attractive and environmentally friendly strategies to address the significant threat to human health posed by indoor air pollutants. Despite its potential, this degradation process still faces issues with suboptimal efficiency, which may be attributed to the rapid recombination of photogenerated excitons and the broad band gap. As a proof of concept, a series of graphite-like C3N4@C60 composites combining graphite-like C3N4 and C60 was developed via an in situ generation strategy. The obtained graphite-like C3N4@C60 composites exhibited a remarkable increase in the photocatalytic degradation efficiency of formaldehyde, of up to 99%, under visible light irradiation, outperforming pure graphite-like C3N4 and C60. This may be due to the composites’ enhanced built-in electric field. Additionally, the proposed composites maintained a formaldehyde removal efficiency of 84% even after six cycles, highlighting their potential for indoor air purification and paving the way for the development of efficient photocatalysts.


Introduction
As the atmospheric environment continues to deteriorate, formaldehyde has emerged as a critical indoor air pollutant that poses a serious threat to human health. It is commonly discharged by industrial activities and building materials, and its impact on human health is irreversible [1,2]. Even at low concentrations, long-term exposure to formaldehyde will extremely damage the human nervous system and respiratory system. Due to growing concern over the harmful effects of formaldehyde on human health, a variety of effective strategies have been developed to eliminate it rapidly. These methods, including adsorption purification [3], thermal catalysis [4,5], biofiltration, and condensation [6,7], are commonly used in indoor air purification. However, most of these approaches involve merely adsorbing formaldehyde onto a filter medium without degrading the pollutants. As a result, these air purifiers have several drawbacks, including low adsorption capacity, the rerelease of formaldehyde into the air, and difficulties in regenerating adsorbents. Thus, it is crucial to prioritize the exploration and development of effective solutions to eliminate formaldehyde at low concentration levels, especially in indoor environments.
Photocatalytic oxidation is regarded as a versatile and promising strategy for air purification credited to its low energy consumption, high efficiency, and lack of secondary pollution; it can effectively mineralize formaldehyde to CO 2

and H 2 O on inexpensive
Molecules 2023, 28, 5815 2 of 11 polymeric semiconductors under mild conditions [2,[8][9][10][11][12][13]. The semiconductor graphitelike C 3 N 4 , which is considered one of the most advanced metal-free photocatalysts, has been widely applied for the photocatalytic degradation of formaldehyde due to its strong oxidation ability, low cost-effectiveness, and long-term stability [14][15][16][17][18]. However, its large band gap and limited light absorption capacity could result in the recombination of its internal excited carriers, which would undermine its photocatalytic efficiency [19,20]. To enhance the separation efficiency of photogenerated excitons, numerous strategies such as semiconductor composites [9], elemental doping [21,22], dye sensitization [23,24], surface engineering [25][26][27][28], and constructing heterojunctions [8] have been adopted to circumvent these obstacles. Despite efforts to enhance the photocatalytic efficiency of graphite-like C 3 N 4 , these strategies are limited by issues such as serious corrosion, high temperatures, and tedious preparation steps. Therefore, there is an utmost urgency to discover and develop a simple and efficient strategy for improving the photocatalytic efficiency of graphite-like C 3 N 4 (Scheme 1).
Photocatalytic oxidation is regarded as a versatile and promising strategy for air purification credited to its low energy consumption, high efficiency, and lack of secondary pollution; it can effectively mineralize formaldehyde to CO2 and H2O on inexpensive polymeric semiconductors under mild conditions [2,[8][9][10][11][12][13]. The semiconductor graphite-like C3N4, which is considered one of the most advanced metal-free photocatalysts, has been widely applied for the photocatalytic degradation of formaldehyde due to its strong oxidation ability, low cost-effectiveness, and long-term stability [14][15][16][17][18]. However, its large band gap and limited light absorption capacity could result in the recombination of its internal excited carriers, which would undermine its photocatalytic efficiency [19,20]. To enhance the separation efficiency of photogenerated excitons, numerous strategies such as semiconductor composites [9], elemental doping [21,22], dye sensitization [23,24], surface engineering [25][26][27][28], and constructing heterojunctions [8] have been adopted to circumvent these obstacles. Despite efforts to enhance the photocatalytic efficiency of graphite-like C3N4, these strategies are limited by issues such as serious corrosion, high temperatures, and tedious preparation steps. Therefore, there is an utmost urgency to discover and develop a simple and efficient strategy for improving the photocatalytic efficiency of graphite-like C3N4 (Scheme 1). Scheme 1. Synthetic routes and notional structures of g-C3N4 and C60 composites.
Fullerenes, such as C60, are distinct forms of carbon with exceptional electronic characteristics [29][30][31][32][33][34]. C60 is considered favorable for efficient electron transfer reduction due to its closed-shell configuration [35,36]. The distinct structure of C60 makes it an outstanding electron acceptor that effectively induces quick photoinduced charge separation while experiencing comparatively slow charge recombination [37][38][39]. Wang et al. have provided a comprehensive overview of the recent notable progress in the realms of hydrogenation and oxidation facilitated by catalytic systems based on graphite-like C3N4 [40,41]. They also discovered that an amalgamation of carbon nitride and carbon nanotubes displayed a synergistic effect during optoelectronic conversion. Consequently, graphite-like C3N4 has become a category of 2D nanomaterials resembling graphite, and its distinct structure offers vast potential for utilization as a metal-free semiconductor that can govern photocatalytic reactions. Huang et al. reported that the electrical performance of covalent Scheme 1. Synthetic routes and notional structures of g-C 3 N 4 and C 60 composites.
Fullerenes, such as C 60 , are distinct forms of carbon with exceptional electronic characteristics [29][30][31][32][33][34]. C 60 is considered favorable for efficient electron transfer reduction due to its closed-shell configuration [35,36]. The distinct structure of C 60 makes it an outstanding electron acceptor that effectively induces quick photoinduced charge separation while experiencing comparatively slow charge recombination [37][38][39]. Wang et al. have provided a comprehensive overview of the recent notable progress in the realms of hydrogenation and oxidation facilitated by catalytic systems based on graphite-like C 3 N 4 [40,41]. They also discovered that an amalgamation of carbon nitride and carbon nanotubes displayed a synergistic effect during optoelectronic conversion. Consequently, graphite-like C 3 N 4 has become a category of 2D nanomaterials resembling graphite, and its distinct structure offers vast potential for utilization as a metal-free semiconductor that can govern photocatalytic reactions. Huang et al. reported that the electrical performance of covalent organic frame-works (COFs) can be enhanced by encapsulating fullerenes into the channels of COFs via a donor-acceptor interaction [42]. The improved electrical conductivity and carrier mobility can promote efficient charge transfer, which is highly desirable in photocatalytic processes.
In light of the above, using an in situ generation approach, we developed a series of photocatalytic composites, denoted as graphite-like C 3 N 4 @C 60 , wherein the guest molecule C 60 was encapsulated into the host graphite-like C3N4 framework. The resulting graphitelike C 3 N 4 /C 60 @6:1 composite induced the formation of a strong built-in electric field, which accelerated charge transport kinetics. In addition, this composite achieved a formaldehyde degradation efficiency of up to 99% under visible-light irradiation, outperforming pure graphite-like C 3 N 4 and C 60 . Our study presents a pioneering approach for designing photocatalysts based on graphite-like C 3 N 4 and achieving efficient solar energy conversion.

Characterization of the Photocatalysts
The as-prepared specimens were characterized by FT-IR spectra to confirm their structures ( Figure 1a). The bands located at~1610 cm −1 were ascribed to C=N stretching vibration, whereas the other strong bands, at~1243 and~1398 cm −1 , were ascribed to C-N stretching vibration, matching the s-triazine ring in the graphite-like C 3 N 4 well. The FT-IR spectrum of the C 60 was relatively weak, but two peaks were observed at~1760 and 1940 cm −1 , corresponding to the C 60 s internal modes. There was no obvious structural variation between the graphite-like C 3 N 4 and the 6 wt% C 60 /graphite-like C 3 N 4 following the deposition of the C 60 . However, the characteristic peaks of the graphite-like C 3 N 4 in the 6 wt% C 60 /graphite-like C 3 N 4 , ranging from~1200 to 1700 cm −1 , were shifted, indicating that there was a weak interaction between the C 60 and the graphite-like C 3 N 4 . This may have facilitated electron transfer and improved the photocatalytic activity of the composites compared to the graphite-like C 3 N 4 . nels of COFs via a donor-acceptor interaction [42]. The improved electrical conductiv and carrier mobility can promote efficient charge transfer, which is highly desirable photocatalytic processes.
In light of the above, using an in situ generation approach, we developed a series photocatalytic composites, denoted as graphite-like C3N4@C60, wherein the guest molecu C60 was encapsulated into the host graphite-like C3N4 framework. The resulting graphi like C3N4/C60@6:1 composite induced the formation of a strong built-in electric field, whi accelerated charge transport kinetics. In addition, this composite achieved a formald hyde degradation efficiency of up to 99% under visible-light irradiation, outperformi pure graphite-like C3N4 and C60. Our study presents a pioneering approach for designi photocatalysts based on graphite-like C3N4 and achieving efficient solar energy conve sion.

Characterization of the Photocatalysts
The as-prepared specimens were characterized by FT-IR spectra to confirm th structures ( Figure 1a). The bands located at ~1610 cm −1 were ascribed to C=N stretchi vibration, whereas the other strong bands, at ~1243 and ~1398 cm −1 , were ascribed to C stretching vibration, matching the s-triazine ring in the graphite-like C3N4 well. The FTspectrum of the C60 was relatively weak, but two peaks were observed at ~1760 and ~19 cm −1 , corresponding to the C60′s internal modes. There was no obvious structural variati between the graphite-like C3N4 and the 6 wt% C60/graphite-like C3N4 following the dep sition of the C60. However, the characteristic peaks of the graphite-like C3N4 in the 6 w C60/graphite-like C3N4, ranging from ~1200 to 1700 cm −1 , were shifted, indicating that the was a weak interaction between the C60 and the graphite-like C3N4. This may have faci tated electron transfer and improved the photocatalytic activity of the composites co pared to the graphite-like C3N4. XPS survey spectra of spectra of C60, g-C3N4, and g-C3N4/C60@6:1; (c) high-resolution XPS spectra C 1s for g-C3N4; (d) high-resolution XPS spectra of N 1s for g-C3N4; (e) high-resolution XPS spec of C 1s for g-C3N4/C60@6:1; and (f) high-resolution XPS spectra of N 1s for g-C3N4/C60@6:1.
X-ray photoelectron spectroscopy (XPS) was conducted to verify the chemical com positions and determine the valence states of different species. The graphite-like C3 Figure 1. (a) FT-IR spectra of C 60 , g-C 3 N 4 , and g-C 3 N 4 /C 60 @6:1, obtained in transmission mode; (b) XPS survey spectra of spectra of C 60 , g-C 3 N 4 , and g-C 3 N 4 /C 60 @6:1; (c) high-resolution XPS spectra of C 1s for g-C 3 N 4 ; (d) high-resolution XPS spectra of N 1s for g-C 3 N 4 ; (e) high-resolution XPS spectra of C 1s for g-C 3 N 4 /C 60 @6:1; and (f) high-resolution XPS spectra of N 1s for g-C 3 N 4 /C 60 @6:1. X-ray photoelectron spectroscopy (XPS) was conducted to verify the chemical compositions and determine the valence states of different species. The graphite-like C 3 N 4 mainly comprised the elements C, N, and O. Among them, the presence of O was attributed to H 2 O and CO 2 absorbed on the surface of the photocatalysts. All peak positions in the XPS spectra of the graphite-like C 3 N 4 and the 6 wt% graphite-like C 3 N 4 /C 60 composite were calibrated using C 1s at 284.6 eV as a reference (Figure 1b). The C 1s XPS spectrum exhibited two peaks, with distinct binding energies, at 284.6 and 288.8 eV, which corresponded to C-C and N-C=N, respectively. The N 1s XPS spectrum of the graphite-like C3N4 exhibited two distinct peaks at 400.5 and 401.7 eV, which might be attributed to C=N-C and N(C) 3 , respectively, upon deconvolution. A comparison of the N 1s XPS spectrum of the graphite-like C 3 N 4 with that of the 6 wt% graphite-like C 3 N 4 /C 60 indicated that the binding energies of N (N-C=N) peak-shifted from 400.5 eV (graphite-like C 3 N 4 ) to 399.8 eV (6 wt% graphite-like C 3 N 4 /C 60 ). This shift suggests that an interaction occurred between the graphite-like C 3 N 4 and the C 60 (Figure 1c-f). The extremely weak peak at 404.8 eV was ascribed to π excitation. Previous studies have demonstrated that a lower binding energy of N 1s for composites is indicative of an increased electronic cloud density around the N atoms of graphite-like C 3 N 4 . This effect is attributed to intermolecular electron diffusion from conjugated polymers to the N sites of graphite-like C 3 N 4 through intermolecular π-π interactions.
The crystallinities for C 60 , graphite-like C 3 N 4 , and graphite-like C 3 N 4 /C 60 @6:1 were characterized with X-ray diffraction. As can be clearly observed in Figure S1, the graphitelike C 3 N 4 and the graphite-like C 3 N 4 /C 60 @6:1 showed similar, prominent strong-intensity peaks at 25.3 • , which correspond to the in-plane repeated units and the structural packing motifs of the aromatic segments, respectively, indicating the successful preparation of the graphite-like C 3 N 4 . The C 60 displayed a cubic phase and could be indexed based on its diffraction pattern, which demonstrated two peaks, at 10.7 and 15.5, attributed to the (111) and (220) crystal planes, respectively. The deposition of C 60 on the graphite-like C 3 N 4 /C 60 @6:1 composite surface had no remarkable effects on the structure of the graphitelike C 3 N 4 , as evidenced by the fact that the positions and intensities of the characteristic peaks at 10.7 • and 15.5 • , corresponding to the (111) and (220) crystal planes, respectively, remained virtually unchanged compared with those of the bare graphite-like C 3 N 4 . These observations indicate that the C 60 was successfully deposited onto the graphite-like C 3 N 4 without changing the crystal structure ( Figure S1). The specific surface areas of the graphitelike C 3 N 4 and the graphite-like C 3 N 4 /C 60 @6:1 were explored using adsorption-desorption isotherms performed at 77 K ( Figures S2 and S3). Accordingly, the BET surface areas of the graphite-like C 3 N 4 and the graphite-like C 3 N 4 /C 60 @6:1 were determined to be 40.0 and 29.6 m 2 /g, respectively. However, the pore volume of graphite-like C 3 N 4 /C 60 @6:1 (0.082 cm 3 /g) was considerably lower than that of graphite-like C 3 N 4 (0.093 cm 3 /g). This difference shows that the pores of the graphite-like C 3 N 4 were blocked or occupied by C 60 nanoparticles. The morphology of the graphite-like C 3 N 4 /C 60 @6:1 composite was similar to that of the graphite-like C 3 N 4 , and there was no specific structure of C 60 in the graphite-like C 3 N 4 /C 60 @6:1 composite because of its limited concentration. Moreover, the presence of dark spots with the lower transmission in the graphite-like C 3 N 4 /C 60 @6:1 composite indicated that there might have been some perturbation or disruption of the C 60 nanoparticles. However, this perturbation did not appear to have affected the overall porous structure of the graphite-like C 3 N 4 . The small sizes of the C 60 nanoparticles may have allowed them to be easily incorporated into the graphite-like C 3 N 4 nanosheets, resulting in a well-dispersed composite ( Figure 2).
UV-vis diffuse reflectance spectroscopy was utilized to characterize the optical properties of the C 60 , the graphite-like C 3 N 4 , and the graphite-like C 3 N 4 /C 60 @6:1. Both the C 60 and the graphite-like C 3 N 4 /C 60 @6:1 composite exhibited wide absorption edges, at 700 nm, ascribing to the intrinsic band gap absorption of the C 60 . (Figure 3a). The graphitelike C 3 N 4 /C 60 @6:1 composite showed obvious red-shift edges after the introduction of C 60 , indicating that the graphite-like C 3 N 4 /C 60 @6:1 composite could broaden the spectrum to the visible region compared to the bare graphite-like C 3 N 4 . Based on the equation E g = 1240/λ, the band gaps of the graphite-like C 3 N 4 /C 60 @6:1 and the graphite-like C 3 N 4 were determined to be 2.62 and 2.85 eV, respectively (Figure 3b). The smaller band gap for the graphite-like C 3 N 4 /C 60 @6:1 indicates that it is more easily excited to generate free holes and electrons. The lowest unoccupied molecular orbital (LUMO) levels of graphite-like C 3 N 4 and graphite-like C 3 N 4 /C 60 @6:1 were calculated using the Mott−Schottky method to be −1.10 and −1.20 V versus those of Ag/AgCl, indicating that the as-prepared graphitelike C 3 N 4 /C 60 @6:1 featured a more sufficient driving force for the reduction of O 2 to O 2 •− (−0.48 V versus Ag/AgCl) compared to the graphite-like C 3 N 4 (Figure 3c,d). UV-vis diffuse reflectance spectroscopy was utilized to characterize the optical properties of the C60, the graphite-like C3N4, and the graphite-like C3N4/C60@6:1. Both the C60 and the graphite-like C3N4/C60@6:1 composite exhibited wide absorption edges, at ~700 nm, ascribing to the intrinsic band gap absorption of the C60. (Figure 3a). The graphite-like C3N4/C60@6:1 composite showed obvious red-shift edges after the introduction of C60, indicating that the graphite-like C3N4/C60@6:1 composite could broaden the spectrum to the visible region compared to the bare graphite-like C3N4. Based on the equation Eg = 1240/λ, the band gaps of the graphite-like C3N4/C60@6:1 and the graphite-like C3N4 were determined to be 2.62 and 2.85 eV, respectively (Figure 3b). The smaller band gap for the graphite-like C3N4/C60@6:1 indicates that it is more easily excited to generate free holes and electrons. The lowest unoccupied molecular orbital (LUMO) levels of graphite-like C3N4 and graphite-like C3N4/C60@6:1 were calculated using the Mott−Schottky method to be −1.10 and −1.20 V versus those of Ag/AgCl, indicating that the as-prepared graphite-like C3N4/C60@6:1 featured a more sufficient driving force for the reduction of O2 to O2 •− (−0.48 V versus Ag/AgCl) compared to the graphite-like C3N4 (Figure 3c,d).  UV-vis diffuse reflectance spectroscopy was utilized to characterize the optical properties of the C60, the graphite-like C3N4, and the graphite-like C3N4/C60@6:1. Both the C60 and the graphite-like C3N4/C60@6:1 composite exhibited wide absorption edges, at ~700 nm, ascribing to the intrinsic band gap absorption of the C60. (Figure 3a). The graphite-like C3N4/C60@6:1 composite showed obvious red-shift edges after the introduction of C60, indicating that the graphite-like C3N4/C60@6:1 composite could broaden the spectrum to the visible region compared to the bare graphite-like C3N4. Based on the equation Eg = 1240/λ, the band gaps of the graphite-like C3N4/C60@6:1 and the graphite-like C3N4 were determined to be 2.62 and 2.85 eV, respectively (Figure 3b). The smaller band gap for the graphite-like C3N4/C60@6:1 indicates that it is more easily excited to generate free holes and electrons. The lowest unoccupied molecular orbital (LUMO) levels of graphite-like C3N4 and graphite-like C3N4/C60@6:1 were calculated using the Mott−Schottky method to be −1.10 and −1.20 V versus those of Ag/AgCl, indicating that the as-prepared graphite-like C3N4/C60@6:1 featured a more sufficient driving force for the reduction of O2 to O2 •− (−0.48 V versus Ag/AgCl) compared to the graphite-like C3N4 (Figure 3c,d).  The photocatalytic oxidation of formaldehyde was employed to assess the photocatalytic activity via the IR multi-gas monitor under UV-vis light irradiation for 1 h at a wavelength range of 310-800 nm and an intensity of 2.9 mW/cm 2 . As displayed in Figure 4a, the photocatalytic performance of the graphite-like C 3 N 4 /C 60 @6:1 composite in HCHO degradation varied with different mass ratios of the composite material. With an increase in the loading of the C 60 on the graphite-like C 3 N 4 , the photocatalytic efficiency increased from 40.2% to 96.4%, indicating that the incorporation of C 60 can remarkably enhance graphite-like C 3 N 4 during photocatalytic HCHO degradation. As the amount of C 60 increases in a graphite-like C 3 N 4 /C 60 composite, it could lead to a decrease in photocatalytic activity due to shading effects that would reduce the amount of light that would reach the graphite-like C 3 N 4 . Under the experimental conditions, the photocatalytic efficiency of the C 60 alone was found to be relatively low (14%). This observation suggests that the incorporation of C 60 into g-C 3 N 4 will enhance photocatalytic activity for HCHO degradation and can act as an efficient photocatalyst. Apart from its photocatalytic activity, photochemical stability is also a crucial consideration for photocatalytic materials. In the case of graphite-like C 3 N 4 /C 60 @6:1, the results of recycling experiments showed that the composite material maintained its photocatalytic activity for the oxidation of the HCHO over multiple cycles without any significant decline in activity (Figure 4b). Furthermore, the control photocatalysis experiment showed that the catalyst was needed for formaldehyde oxidation, and light irradiation alone was not sufficient to degrade the formaldehyde (not shown). The photocatalytic oxidation of formaldehyde was employed to assess the photocatalytic activity via the IR multi-gas monitor under UV-vis light irradiation for 1 h at a wavelength range of 310-800 nm and an intensity of 2.9 mW/cm 2 . As displayed in Figure 4a, the photocatalytic performance of the graphite-like C3N4/C60@6:1 composite in HCHO degradation varied with different mass ratios of the composite material. With an increase in the loading of the C60 on the graphite-like C3N4, the photocatalytic efficiency increased from 40.2% to 96.4%, indicating that the incorporation of C60 can remarkably enhance graphite-like C3N4 during photocatalytic HCHO degradation. As the amount of C60 increases in a graphite-like C3N4/C60 composite, it could lead to a decrease in photocatalytic activity due to shading effects that would reduce the amount of light that would reach the graphite-like C3N4. Under the experimental conditions, the photocatalytic efficiency of the C60 alone was found to be relatively low (14%). This observation suggests that the incorporation of C60 into g-C3N4 will enhance photocatalytic activity for HCHO degradation and can act as an efficient photocatalyst. Apart from its photocatalytic activity, photochemical stability is also a crucial consideration for photocatalytic materials. In the case of graphite-like C3N4/C60@6:1, the results of recycling experiments showed that the composite material maintained its photocatalytic activity for the oxidation of the HCHO over multiple cycles without any significant decline in activity (Figure 4b). Furthermore, the control photocatalysis experiment showed that the catalyst was needed for formaldehyde oxidation, and light irradiation alone was not sufficient to degrade the formaldehyde (not shown). To explore the photocatalytic mechanisms of formaldehyde degradation by graphitelike C3N4/C60@6:1, radical detection was conducted using 5,5-dimethyl-1-pyrroline N-oxide (DMPO) as an active species-trapping agent (Figure 4c). Under visible light and dark (d) ESR spectra of g-C 3 N 4 /C 60 @6:1 for detecting •OH.
To explore the photocatalytic mechanisms of formaldehyde degradation by graphitelike C 3 N 4 /C 60 @6:1, radical detection was conducted using 5,5-dimethyl-1-pyrroline Noxide (DMPO) as an active species-trapping agent (Figure 4c). Under visible light and dark conditions, absolute methanol was employed as a solvent to capture •OH and O 2 •− , respectively. Under the dark conditions, there were no characteristic EPR signals observed. However, when the sample was irradiated with the visible light, characteristic DMPO-•OH (hydroxyl radical) and DMPO-O 2 •− (superoxide radical) peaks appeared. This suggests that the interaction between photoinduced electrons and holes may result in the generation of active •OH and O 2 •− species. Based on the EPR experiments, the main reactive oxygen species involved in the photocatalysis were •OH and O 2 •− . These reactive species were produced when photoinduced charge carriers interacted with the H 2 O and O 2 adsorbed on the surface of the catalyst. When exposed to the visible light, the photocatalyst generated photoinduced electrons, which reacted with the adsorbed O 2 molecules to form O 2 •− . This radical then reacted with the H 2 O to produce •OH. Meanwhile, the surface •OH oxidized formaldehyde into formate species. The valence band holes in the catalyst directly oxidized the H 2 O and/or the −OH to form additional •OH. Eventually, the •OH further oxidized the formate species into H 2 O and CO 2 .
In order to better understand how composites can better enhance photocatalytic activity compared to pure graphite-like C 3 N 4 , a range of photoelectrochemical properties was examined. The EPR signals of the graphite-like C 3 N 4 /C 60 @6:1 composite showed a significant increase when exposed to visible light compared with those of pure graphite-like C 3 N 4 . This indicates that these composites enable more effective production of unpaired electrons and photoinduced charge carrier pairs in graphite-like C 3 N 4 . Thus, charge migration and separation are facilitated through the encapsulation of C 60 into pure graphite-like C 3 N 4 (Figure 5a). The built-in electrical field is a critical factor for driving photogenerated holes and electrons to drift in reverse directions in photocatalysts, which, in turn, dramatically accelerates exciton separation. The BIEF intensities of C 60 , graphite-like C 3 N 4 , and graphite-like C 3 N 4 /C 60 @6:1 were estimated by employing the model established by Kanata, and the results indicated that BIEF strength could be assessed with both surface charge density and surface potential. The surface potentials were determined to be 10.8, 25.7, and 58.7 mV for the C 60 , the graphite-like C 3 N 4 , and the graphite-like C 3 N 4 /C 60 @6:1, respectively. Moreover, the corresponding zeta potentials of the C 60 , the graphite-like C 3 N 4 , and the graphite-like C 3 N 4 /C 60 @6:1 were −6.7, −16.1, and −36.1 V, respectively.
According to the open-circuit potential and zeta potential, the graphite-like C 3 N 4 /C 60 @6:1 (8.4) had the strongest built-in electric field, which exceeded those of the C 60 (0.1) and the graphite-like C 3 N 4 (1.5). This significant increase in the built-in electric field indicates a strong driving force for achieving the efficient separation of excitons ( Figure 5b).
conditions, absolute methanol was employed as a solvent to capture •OH and O2 •− , respectively. Under the dark conditions, there were no characteristic EPR signals observed. However, when the sample was irradiated with the visible light, characteristic DMPO-•OH (hydroxyl radical) and DMPO-O2 •− (superoxide radical) peaks appeared. This suggests that the interaction between photoinduced electrons and holes may result in the generation of active •OH and O2 •− species. Based on the EPR experiments, the main reactive oxygen species involved in the photocatalysis were •OH and O2 •− . These reactive species were produced when photoinduced charge carriers interacted with the H2O and O2 adsorbed on the surface of the catalyst. When exposed to the visible light, the photocatalyst generated photoinduced electrons, which reacted with the adsorbed O2 molecules to form O2 •− . This radical then reacted with the H2O to produce •OH. Meanwhile, the surface •OH oxidized formaldehyde into formate species. The valence band holes in the catalyst directly oxidized the H2O and/or the −OH to form additional •OH. Eventually, the •OH further oxidized the formate species into H2O and CO2.
In order to better understand how composites can better enhance photocatalytic activity compared to pure graphite-like C3N4, a range of photoelectrochemical properties was examined. The EPR signals of the graphite-like C3N4/C60@6:1 composite showed a significant increase when exposed to visible light compared with those of pure graphite-like C3N4. This indicates that these composites enable more effective production of unpaired electrons and photoinduced charge carrier pairs in graphite-like C3N4. Thus, charge migration and separation are facilitated through the encapsulation of C60 into pure graphitelike C3N4 (Figure 5a). The built-in electrical field is a critical factor for driving photogenerated holes and electrons to drift in reverse directions in photocatalysts, which, in turn, dramatically accelerates exciton separation. The BIEF intensities of C60, graphite-like C3N4, and graphite-like C3N4/C60@6:1 were estimated by employing the model established by Kanata, and the results indicated that BIEF strength could be assessed with both surface charge density and surface potential. The surface potentials were determined to be 10.8, 25.7, and 58.7 mV for the C60, the graphite-like C3N4, and the graphite-like C3N4/C60@6:1, respectively. Moreover, the corresponding zeta potentials of the C60, the graphite-like C3N4, and the graphite-like C3N4/C60@6:1 were −6.7, −16.1, and −36.1 V, respectively. According to the open-circuit potential and zeta potential, the graphite-like C3N4/C60@6:1 (8.4) had the strongest built-in electric field, which exceeded those of the C60 (0.1) and the graphite-like C3N4 (1.5). This significant increase in the built-in electric field indicates a strong driving force for achieving the efficient separation of excitons (Figure 5b). Figure 5. (a) EPR spectra for g-C 3 N 4 and g-C 3 N 4 /C 60 @6:1; (b) characterization of the BEF values for C 60 , g-C 3 N 4 , and g-C 3 N 4 /C 60 @6:1; (c) transient photocurrent responses for g-C 3 N 4 and g-C 3 N 4 /C 60 @6:1; (d) Nyquist plots for g-C 3 N 4 and g-C 3 N 4 /C60@6:1; (e) photoluminescence decay traces for g-C 3 N 4 and g-C 3 N 4 /C 60 @6:1; and (f) photoluminescence spectra of g-C 3 N 4 and g-C 3 N 4 /C 60 @6:1. The vital role of encapsulated C 60 in the charge mobility was further explored. As expected, the graphite-like C 3 N 4 /C 60 @6:1 exhibited a higher photocurrent intensity than that of the pure graphite-like C 3 N 4 , which implied accelerated production and separation of photoinduced electron-and-hole pairs at the graphite-like C 3 N 4 /C 60 @6:1 interface (Figure 5c). Electrochemical impedance spectra were employed to explore the transfer of photogenerated charge carriers. Semicircular Nyquist plots showed a remarkable decrease in radii upon the deposition of C 60 on graphite-like C 3 N 4 , indicating a significant enhancement in the rate of charge transfer (Figure 5d). These findings have demonstrated that the electrochemical impedance of graphite-like C 3 N 4 is optimized when it is combined with C 60 .
Transient fluorescence was conducted to illustrate the separation behaviors of photoinduced electron-and-hole pairs by calculating the excited-state lifetimes of graphite-like C 3 N 4 and graphite-like C 3 N 4 /C 60 @6:1. The graphite-like C 3 N 4 /C 60 @6:1 composite demonstrated a longer transient fluorescence lifetime (3.02 ns) compared to the pure graphite-like C 3 N 4 (1.87 ns), indicating that graphite-like C3N4/C60@6:1 has a higher potential for efficiently achieving the photocatalytic degradation of gaseous HCHO (Figure 5e). A significantly lower photoluminescence intensity of the graphite-like C 3 N 4 /C60@6:1 could be observed compared with that of the graphite-like C 3 N 4 , which, in principle, suggests that graphite-like C 3 N 4 /C60@6:1 is more effective in the separation and transfer of photogenerated charge carriers and the suppression of photogenerated exciton recombination (Figure 5f).

Materials
Urea (98%) and C 60 (98%) were supplied by Aladdin. Toluene and other conventional reagents were obtained from d from Beijing HWRK Chem Co., Ltd., (Beijing, China). All solvents and chemicals were used directly without any further purification.

Preparation of Graphite-like C 3 N 4
With reference to the prior literature, 20 g of urea was placed in a crucible and heated to 550 • C for 5 h. The graphite-like C 3 N 4 was obtained as powdery yellow granules and was used directly without further treatment.

Preparation of Graphite-like C 3 N 4 @C 60 Composites
To prepare a mixture of C 60 in the toluene (50 mL), the as-prepared graphite-like C 3 N 4 (200 mg) was added. After stirring, the mixture was treated ultrasonically for 1 h. After the removal of the toluene under vacuum, the residues were rinsed with ethanol and dried to obtain gray graphite-like C 3 N 4 @C 60 composites. Different graphite-like C 3 N 4 @C 60 composites were synthesized, with the weight ratios of C 60 :graphite-like C 3 N 4 ranging from 0 to 8 wt%.

Photocatalytic Degradation of Gaseous Formaldehyde
Approximately 300 mg of composites was evenly dispersed in 30 mL of distilled water, followed by sonication for 30 min. The resulting suspension was transferred into three surface dishes with 6 cm diameters, followed by drying at 60 • C for 12 h. The surface dishes were placed in a reactor, and the reactor was left in a closed system. Then, formaldehyde was injected using a microsyringe and the initial concentration of formaldehyde was set at about 120 ± 20 ppm. The concentrations of formaldehyde were measured online using an IR multi-gas monitor (INNOVA Air Tech Instruments Model 1412) under UV-vis light irradiation for 1 h at a wavelength range of 310-800 nm and an intensity of 2.9 mW/cm 2 .

Conclusions
In summary, various graphite-like C 3 N 4 and C 60 composites were developed via an in situ generation strategy. The efficiency charge separation and photocatalysis for the resultant composites were tuned by simply varying the weight ratios between the C 60 and the graphite-like C 3 N 4 . It was found that the graphite-like C 3 N 4 /C 60 @6:1 composite exhibited the strongest built-in electric field, thus realizing efficient charge separation and rendering superior photocatalytic efficiency for formaldehyde degradation compared to the graphite-like C 3 N 4 and the C 60. Overall, this research has offered valuable insights into the design and development of novel synergistic systems for practical applications, particularly in the field of photocatalysis.