Recent Progress on Fullerene-Based Materials: Synthesis, Properties, Modifications, and Photocatalytic Applications

Abstract

Solar light is an inexpensive energy source making up for energy shortage and solving serious environmental problems. For efficient utilization of solar energy, photocatalytic materials have attracted extensive attention over the last decades. As zero-dimensional carbon nanomaterials, fullerenes (C₆₀, C₇₀, etc.) have been extensively investigated for photocatalytic applications. Due to their unique properties, fullerenes can be used with other semiconductors as photocatalyst enhancers, and also as novel photocatalysts after being dispersed on non-semiconductors. This review summarizes fullerene-based materials (including fullerene/semiconductors and fullerene/non-semiconductors) for photocatalytic applications, such as water splitting, Cr (VI) reduction, pollutant degradation and bacterial disinfection. Firstly, the optical and electronic properties of fullerene are presented. Then, recent advances in the synthesis and photocatalytic mechanisms of fullerene-based photocatalysts are summarized. Furthermore, the effective performances of fullerene-based photocatalysts are discussed, mainly concerning photocatalytic H₂ generation and pollutant removal. Finally, the current challenges and prospects of fullerene-based photocatalysts are proposed. It is expected that this review could bring a better understanding of fullerene-based photocatalysts for water treatment and environmental protection.

1. Introduction

There is no denying that both environmental issues and the energy crisis are becoming serious threats to the sustainable development of human society, with the endless consumption of fossil fuels and the irregular discharge of anthropogenic action [1,2]. In order to solve these problems, industrial development must concentrate on clean energy alternatives, which reduce environmental pollution. As a renewable energy source, solar energy has been an intriguing option. Photocatalysts are an effective route to utilize solar energy for various chemical reactions, including photocatalytic pollutant degradation, disinfection, selective organic synthesis, reduction of CO₂ and H₂ generation. This is an attractive technology which could effectually utilize solar energy, generate clean production (H₂) and remediate the environment. Since the photocatalytic performance of TiO₂ for water splitting was proposed for the first time by Fujishima and Honda in 1970s, much work has been done to study the photocatalytic mechanisms and develop novel photocatalysts [3]. Up to date, numerous appealing photocatalysts have been developed and extensively investigated, such as simple oxides (ZnO) [4], metal chalcogenides (CdS) [5], Ag-based compound (Ag₃PO₄) [6], Bi-based compound (BiVO₄ and Bi₂MoO₆) [7,8], MOFs [9] and g-C₃N₄ [10]. In addition to novel photocatalysts, cocatalysts such as precious metals (Pt), two-dimensional transition metal sulfides (MoS₂, WS₂, etc.) and carbonaceous nanomaterials are also widely developed and applied in the field of photocatalysis [11,12,13].

Since the mid-1990s, carbonaceous nanomaterials have been attracting extensive attention, including fullerene, carbon nanotube (CNT) and graphene [14]. Due to uniquely optical and electrical properties, they have been extensively investigated in photocatalytic applications in the past decades. On one hand, they could enhance the photocatalytic efficiency of other semiconductors after combination. For example, CNTs could induce photocatalytic enhancement via three mechanisms: increasing the surface area, suppressing the recombination of hole(h⁺)-electron(e⁻) pairs and enhancing the adsorption of visible light) [15]. Similar to CNT, graphene covers all three of the mechanisms of photocatalytic enhancement above. On the other hand, carbonaceous nanomaterials display effective photocatalytic performance on their own without combining with other semiconductors and are applied alone as novel photocatalysts in some cases. For example, Luo, et al. [16] proposed a self-photocatalytic activity of multiwalled nanotubes (MWCNTs) in the visible range after highly defective modification. Moreover, modified graphene oxide (GO) with a band gap of 2.4−4.3 eV exhibits effective H₂ generation ability within light illumination (UV or visible), which alone may be regarded as a next-generation photocatalyst [17,18].

Among carbonaceous nanomaterials, fullerene exhibits appealing performances similar to CNT and graphene in the photocatalytic application. In previous studies, extensive attentions have been devoted to exploring the roles that fullerene plays in the photocatalytic processes. It was proven that fullerene can be used not only as a photocatalytic enhancer for other semiconductors but also as a novel photocatalyst itself, after being dispersed on a non-semiconductor support. This is ascribed to its distinct optical, photophysical and photochemical properties. Fullerene, a carbon allotrope, is a kind of zero-dimensional (0D) nanocarbon material discovered by Kroto et al., and it has a closed-cage spherical structure which consists of five-membered and six-membered rings [19]. It is well established that there are various forms of fullerene, such as C₆₀, C₇₀, C₇₆, C₈₂ and C₈₄. Among these forms, the C₆₀ and C₇₀ were more extensively investigated than others. Owing to electron delocalization, fullerenes are used extensively as strong-affinity electron acceptor, and for instance C₆₀ is able to reversibly absorb six electrons [20]. The band gap energy (Eg) of solid fullerenes (such as C₆₀, C₇₀, C₈₄ etc.) are from 1.5 to 1.98 eV between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) [4]. Ascribed to the narrow Eg, fullerenes have intensive absorption of UV light and moderate but extensive adsorption of visible light, which make them appealing options for photocatalytic application. Additionally, fullerenes have been previously reported as an excellent photosensitizer as well, with a high quantum efficiency around 1.0 [21]. Fullerene solutions can induce photochemical reactive oxygen species (ROS) generation via two pathways. Under light irradiation (UV or visible), single oxygen (¹O₂) will be formed in fullerene-toluene solution (pathway II), and superoxide anion radical (O₂^−•) and hydroxyl radical (•OH) can be generated in solvent in the presence of electron donors such as ethylenediamine tetraacetic acid (EDTA) and nicotinamide adenine dinucleotide (NADH) [22]. Typically, ROS is a class of active materials that easily induce chemical reaction, which could play an effective role in photocatalytic application. However, easy aggregation is the main obstacle of fullerene in water treatment applications, which suppresses the photoactivity of fullerene. Namely, when dispersed in water, fullerene tends to form nanoscale aggregates (termed nC₆₀, nC₇₀, etc.) with the quenching of excited states of neighboring fullerene molecules which are brought into close contact via aggregation. For retaining fullerene’s photoactivity in aqueous systems, it is necessary for immobilization of fullerene onto solid supports.

Nowadays, many fullerene-semiconductor materials have been successfully built for photocatalytic applications, such as TiO₂/C₆₀ (C₇₀), ZnO/C₆₀, CdS/C₆₀ and C₃N₄/C₆₀ (C₇₀) [23,24,25]. These photocatalysts have been extensively investigated in photocatalytic pollutant degradation, disinfection and water splitting for H₂ evolution. Note that fullerene can obviously enhance the photocatalytic efficiency. At the same time, a variety of fullerene-support (non-semiconductor) materials were successfully fabricated and used for photodegradation of organic pollutant, photocatalytic organic synthesis and disinfection, such as silica/C₆₀, γ-Al₂O₃/C₆₀, MCM-41/C₇₀ and polysiloxane-supported fullerene derivative [26,27,28]. Apart from high-efficient photocatalytic activity, these photocatalysts not only exhibited more stable than the pristine fullerene in solution but also had superior recyclability.

Previously, Yeh, Cihlář, Chang, Cheng and Teng [13] have reviewed the roles of graphene oxide (GO) in photocatalytic water splitting, which mainly introduces strategies for tuning the electronic structure of GO for photocatalytic water splitting. Gangu, Maddila and Jonnalagadda [15] have reported a review on the MWCNTs mediated semiconducting materials as photocatalysts in water treatment. In another review, Ge, Zhang and Park [14] have discussed recent advances in carbonaceous photocatalysts and the developmental direction for them, such as activated carbon, carbon dots, carbon nanotubes, graphene and fullerene. To our knowledge, no papers have reviewed the fullerene/semiconductor and fullerene/support photocatalysts for wastewater treatment and water splitting. Therefore, the present review provides a comprehensive understanding of fullerene-based photocatalysts, including fullerene/semiconductor photocatalysts and fullerene/support photocatalysts. The optical, photochemical and electronic properties of fullerene are generally presented. Then, recent advances in the synthesis methods and photocatalytic application of fullerene-based photocatalysts are summarized. Meanwhile, the photocatalytic efficiency of these-prepared photocatalysts are discussed in wastewater treatment and water splitting for H₂ evolution, wherein the mechanisms of the fullerene-based photocatalysts are underlined in detail. In the end, the current challenges and prospects of fullerene-based photocatalysts are proposed.

2. Role of Fullerene

2.1. Basic Principles of Semiconducting Photocatalysis

In the photocatalytic procedure of semiconductors, there are three main factors, i.e., light resources, photocatalysts and reaction mediums [14]. The photocatalytic process could be initiated only by the light (i.e., UV, infrared and visible light) with energy equal to or over the band gap energy (Eg) of the photocatalyst. Typically, it could be briefly presented as follows. Upon irradiation by light resource, the electrons in the valence band (VB) could be excited to the conduction band (CB) of the photocatalyst and holes leave in VB, resulting in the separation of photogenerated hole-electron pairs. Immediately, most of them recombine with heat generation while a small fraction can transfer to the semiconductor’s surface to induce redox reactions. Generally, for photocatalytic decontamination, the separated holes and electrons of the semiconductor can react with ambient substances (i.e., H₂O and O₂) to produce free radicals (i.e., •OH, •O₂⁻, HO₂•, H₂O₂). Then, the highly oxidative holes and reactive radicals will intensively degrade organic pollutants into small molecules or inorganic materials through addition/substitution reaction and electron transfer between contaminants and free radicals [29]. Through the processes above, the pollution is mitigated or eliminated. Compared with pollutant degradation, the photocatalytic H₂ generation over semiconductor share some similarities. In detail, the generating processes of the photoinduced charges and formation of partial radicals are identical between pollutant degradation and H₂ generation. The difference in photocatalytic H₂ generation is that photoinduced charges react with H⁺ adsorbed on the photocatalyst or in surroundings to produce H₂ rather than •O₂⁻ [30]. In this process, additives are usually required to facilitate the efficiency of photocatalytic H₂ generation, such as hole scavenger and (or) sacrificial donor.

In the past decades, a number of semiconductors were developed for photocatalysis applications, such as TiO₂, ZnO, CdS, Ag-based semiconductors, Bi-based semiconductors and g-C₃N₄. However, many problems limit the photocatalytic efficiency of the current photocatalysts, including insufficient visible light utilization, wide bandgap, rapid recombination of photoinduced holes and electrons and poor stability. Various strategies were proved to be effective in enhancing the photocatalytic activity, such as morphology control, element doping, heterojunction construction and coupling with carbonaceous nanomaterials [10].

2.2. The Role of Fullerene in Semiconductor/Fullerene Photocatalysts

For effective semiconductor/fullerene photocatalysts, the introduction of fullerene generally enhances the photocatalytic performance through various aspects as follows. For instance, fullerene could capture electrons from CB of semiconductor due to its high-affinity for electrons, which significantly retard the recombination of photoinduced hole-electron pairs [31,32]. As a result, more separated photogenerated holes and electrons could take part in photocatalytic reaction, increasing photocatalytic efficiency. Then, fullerene could enhance the light absorption (both UV and visible light) because it is an excellent photo-response material (300~700 nm), wherein elevated light energy utilization excites more electrons from VB to CB [33,34]. In respect to photo-response characteristics, it must also be mentioned that fullerene could not shift the adsorption edge of pristine photocatalyst unless the introduction of fullerene changed the structure of semiconductor [35,36,37]. In other words, if the fullerene does not change the crystalline structure of semiconductor in the synthesis procedure, which was typically affected by a stronger bonding force other than simply physically blending, it could not change the conduction band (Eg) at all. Undoubtedly, the introduction of fullerene could change the specific surface area BET (Brunauer–Emmett–Teller), while it does not always present the same results of change trend. In previous studies, it increased or decreased the BET of the photocatalyst depending on the specific situation [38,39,40]. So, it was controversial that the fullerene enhances the BET of photocatalyst to contribute to high adsorption ability for reactant.

Fullerene is virtually insoluble in water, but soluble in nonpolar organic solvents, such as toluene and 1,2-dichlorobenzen [41]. Hence, it performs a good stability in water solution. After dissolved in nonpolar solvent, fullerene could be coupled with semiconductor to form stable semiconductor/fullerene composite by various methods, wherein no fullerene is leached out even after the as-prepared composite is repeatedly rinsed with the aforementioned solvent. Additionally, fullerene exhibits superior potentials for stabilizing semiconductors which deeply suffers photocorrosion, such as ZnO, CdS and Ag₃PO₄. For example, a significant increase of stabilization was observed for ZnO/C₆₀ nanocomposite in previous studies [42]. The photogenerated holes of ZnO could easily react with surface oxygen atoms during the photocatalysis process, leading to fast decline of photocatalytic activity. When C₆₀ was covered on ZnO, the activity of surface oxygen atoms was effectively reduced so that photocorrosion effect was effectively inhibited. Similarly, high stabilization of AgPO₄/C₆₀ was also observed, because C₆₀ could obviously suppress the transform of Ag⁺ into Ag of bare Ag₃PO₄ composite [6]. Moreover, Cai, et al. [43] fabricated CdS/C₆₀ nanocomposite in which C₆₀ has shown effectively inhibition of photocorrosion of CdS. In order to estimate the stability of the as-prepared sample, the released Cd²⁺ concentration was determined in remaining solution after three cycles for rhodamine (RhB) degradation. The Cd²⁺ concentration was 381.3 µg/L in solution with naked CdS while it was 51.9 µg/L in solution with 0.4C₆₀/CdS nanocomposite, and the former was 7.3-times of the later. This means that C₆₀ could effectively inhibit the photocorrosion to enhance the stability of CdS.

3. Synthesis of Semiconductor/Fullerene Photocatalysts

A number of fullerene/semiconductors (TiO₂, ZnO, CdS, C₃N₄, etc.) have been fabricated for photocatalytic wastewater treatment and water splitting. It is unquestionable that synthetic process plays an important role in determining the size, morphology and physicochemical characteristic of a photocatalyst. Fullerene/semiconductor photocatalysts can be constructed via a series of synthetic methods, including simple adsorption, hydrothermal/solvothermal synthesis, sol-gel procedure and mechanical ball-milling. These synthetic methods are summarized and illustrated briefly as follows.

3.1. Simple Adsorption Method

Simple adsorption method has been extensively used to fabricate semiconductor/fullerene nanocomposites. This procedure is cost-effective without complicated external condition. It is well established that pure fullerene (C₆₀ or C₇₀) is extremely hydrophobic but dissolves in some organic solvents (benzene, toluene, 1,2-dichlorobenzen etc.), which are mainly used to dissolve fullerene for preparing fullerene/semiconductor photocatalysts [4,44]. Typically, the semiconductor is added into the fullerene solution to form adequately dispersed suspensions, and then the newly generating fullerene/semiconductor nanocomposite is obtained through evaporating the solvent. Many semiconductor/fullerene nanocomposites have been synthesized through this method, such as TiO₂/C₆₀, ZnO/C₆₀, Bi-based oxides/C₆₀ and C₃N₄/C₆₀ [45,46,47]. Note that the simple adsorption method differs from direct mechanical mixture, because no fullerene is leached out when the nanocomposite is added into the aforementioned organic solvent.

3.2. Hydrothermal Synthesis Method

Hydrothermal synthesis is an appealing method for preparing semiconductor/fullerene nanomaterials. In the synthetic procedure, pretreatment of fullerene is imperative and acid-treatment is extensively applied to produce several oxygen-containing active sites on the surface of fullerene. For instance, nitric acid is a common agent to oxide fullerene under reflux condition, further facilitating more efficient combination of fullerene with other semiconductors through hydrothermal procedure. Yu, et al. [48] constructed TiO₂/C₆₀ nanocomposite using this acid-treated C₆₀ via a hydrothermal method. Activated C₆₀ and Ti(OC₄H₉)₄ (titanium source) were mixed into ethanol/water solution (1:2 v/v) and then the reaction mixture solution was transferred into stainless steel autoclave at 180 °C for 10 h. Similarly, TiO₂/C₇₀ was fabricated by a hydrothermal procedure as well [49]. Very recently, several semiconductor/fullerene photocatalysts were obtained by the hydrothermal method, such as CdS/C₆₀, PbMoO₄/C₆₀, BiOCl/C₇₀ and C₃N₄/(C₆₀, C₇₀) [50,51]. For example, Cai, Hu, Zhang, Li and Shen [43] constructed CdS/C₆₀ photocatalyst via a facile one-pot hydrothermal method, wherein a mixture of acid-treated C₆₀, Cd(CH₃COO)₂·2H₂O and _L-cysteine in water was heated at 200 °C for 10 h in autoclave. Moreover, Ma, Zhong, Li, Wang and Peng [39] synthesized BiOCl/C₇₀ photocatalyst via a hydrothermal method and the procedure was presented as follow. Acid-treated C₇₀ and Bi(NO₃)₃∙5H₂O were dissolved in glacial acetic acid and KCl solution was slowly added into, and then the mixture solution was transferred into stainless autoclave maintaining at 180 °C for 24 h.

In addition to nitric acid, meta-chloroperoxybenzoic acid (MCPBA) is an effectively alternative oxidizing agent to pretreat fullerene before hydrothermal method. Typically, fullerene and MCPBA are dissolved into benzene, followed by heating reflux for hours to activated the surface of fullerene. For instance, CoS/C₆₀ nanocomposite was prepared by using the MCPBA-oxidized C₆₀ via hydrothermal method [52]. Namely, MCPBA-oxidized C₆₀, CoCl₂ and Na₂S₂O₃ mixture water solution was heated at 150 °C for 12 h in autoclave, and the CoS/C₆₀ nanocomposite was obtained through filter. Similarly, other semiconductor/fullerene nanocomposites were prepared via hydrothermal method with MCPBA-oxidized C₆₀, including CdSe/C₆₀ and WO₃/C₆₀ [36,53].

3.3. Ball Milling Method

Ball-milling is a facile and eco-friendly method to structure solid–solid composites, which could generate stronger intermolecular interactions than physical blends. Recently, the hybridized MoS₂/C₆₀ nanocomposite was obtained through a planetary ball-milling machine [54]. Mixture of MoS₂ and C₆₀ powders was transferred into a ball-milling jar together with stainless steel balls under Ar atmosphere. After operation at 500 rmp for 48 h, the reactant was Soxhlet extracted by CS₂ to remove the unreacted C₆₀. The strong van der Waals (vdW) interactions contributed to formation of MoS₂/C₆₀ heterostructure rather than a covalent conformation, resulting in elevated photocatalytic activity of pure MoS₂. In addition, a g-C₃N₄/C₆₀ nanocomposite was fabricated via a ball-milling route as well [55]. Additional LiOH was needed as a catalyst before ball-milling process and the detailed synthetic process of the g-C₃N₄/C₆₀ was presented in Figure 1. It was firstly proposed in fullerene chemistry that the covalent bonding forms via a four-membered ring of azetidine between C₆₀ and g-C₃N₄ nanosheets. The LiOH as catalyst breaks π-π carbon bonds of C₆₀ to produce C₆₀ radicals, then ball-milling activates g-C₃N₃ and results in the covalent reaction between g-C₃N₄ and C₆₀ (Figure 1).

3.4. Other Techniques

Sol-gel approach has been used to prepare fullerene-TiO₂-semiconductor ternary photocatalysts. Meng, et al. [56] fabricated CdS-C₆₀/TiO₂ photocatalyst by a sol-gel method. Firstly, NaS₂ solution was dropwise added into oxidized C₆₀ and (CH₃COO)₂Cd·2H₂O mixed ethanol solution and the collected solids were calcinated at 300 °C to obtain CdS-C₆₀ particles. Then, CdS-C₆₀ particles were added into titanium (IV) n-butoxide (TNB) solution with constant stirring and CdS-C₆₀/TiO₂ gels were produced in mixed solution under reflux at 70 °C. Finally, the CdS-C₆₀/TiO₂ nanoparticles were obtained after heat treatment at 400 °C. Bai, Wang, Wang, Yao and Zhu [35] proposed a facile thermal treatment method for fabricating g-C₃N₄/C₆₀ photocatalyst. The procedure was presented as follow: ball-milled C₆₀ and dicyandiamide mixture was transferred into a muffle furnace and held at 550 °C for 4 h. Moreover, Li and Ko [57] successfully prepared MoS₂/C₆₀ nanocomposite by a facile heating treatment procedure. The wetness impregnation method is also an effective method for building semiconductor/fullerene photocatalysts. Apostolopoulou, et al. [58] prepared TiO₂/C₆₀ nanoparticles using 1,2-dichloro-benzene as a solvent via a successive incipient wetness impregnation followed by heating at 180 °C. Similarly, a polyhydroxyfullerene (PHF)/titanium dioxide nanotube was prepared by incipient wetness impregnation [59]. Firstly, fullerene was functionalized by NaOH and H₂O₂ to obtain PHF (or called fullerenol). Then, PHF was added to TiO₂ nanotube solution under a wetness impregnation followed by heating at 400 °C. Hence, this provides a new route for coupling fullerenol with other semiconductors to obtain effective photocatalysts.

4. The Photocatalytic Application of Fullerene/Semiconductor Photocatalysts

The fullerene/semiconductor photocatalysts have been extensively used for photocatalytic wastewater treatment (pollutant degradation, Cr (VI) reduction, disinfection etc.) and water splitting for H₂ generation [49,60,61]. Among them, photocatalytic degradation of organic pollutant is ascribed to decomposition of organic molecule structure and photocatalytic disinfection depends on inactivation of microorganisms. However, photocatalytic Cr (VI) reduction focuses on the transformation from Cr (VI) to Cr (III). It is generally believed that chromium (Cr) is among the sixteen most toxic contaminants due to its carcinogenic and teratogenic effect on human. Hence, the World Health Organization (WHO) and the United State Environmental Protection Agency (USEPA) have set the maximum permissible concentration of Cr in drinking water at 0.05 mg/L and 0.1 mg/L [62]. Note that the reduced Cr (III) is far less toxic and more stable than Cr (VI), so the photocatalytic Cr (VI) reduction is a promising method to reduce the chromium toxicity in water.

Furthermore,

Table 1. Summary of fullerene based photocatalysts for pollutant degradation.

Photocatalyst (Additive Amount)	Synthesis Method (Fullerene Content)	Pollutants	Experimental Conditions (Light Source, Pollutant Concentration and React Time)	Photocatalytic Activity	Enhancement Factor	Reference
TiO2/C60 (1 g/L)	In-situ growth (2.0 wt %)	Methylene blue (MB)	UV irradiation, 1.0 × 10−4 mol/L, 60 min	99%	around 75% for TiO2	[67]
TiO2/C60 (1 g/L)	Ultrasonication–evaporation (1.0 wt %)	RhB	500 W Xe-lamp (>400 nm), 10 mg/L, 150 min	95%	below 5% for TiO2	[68]
TiO2/C70 (1 g/L)	Hydrothermal synthesis (8.5 wt %)	Sulfathiazole	300 W Xenon lamp (>420 nm), 10 mg/mL, 180 min	80%	10% for TiO2	[69]
ZnO/C60 (0.5 g/L)	Simple adsorption (1.5 wt %)	MB	8 W UV lamp (λ = 254 nm), 8 mg/L	k = 0.0569 min−1	3-times than ZnO	[70]
ZnO/C60 (0.83 g/L)	Chemical vapor (16.7 wt %)	Phenol	1500 W xenon lamp simulating solar light, 20 mg/L	k = 0.160 min−1	1.22-times than ZnO	[42]
ZnFe2O4@C60 (1 g/L)	Hydrothermal synthesis	Norfloxacin	Solar irradiation, 20 mL of 50 ppm norfloxacin, 90 min	85%	60% for ZnFe2O4	[71]
WO3@C60	Hydrothermal synthesis (4.0 wt %)	MB	Visible light, 90 min	94%	Inferior degradation efficiency for pure WO3	[53]
ZnAlTi-LDH@C60 (ZnAlTi-LDO) 0.5 g/L	Precipitation (5%)	Bisphenol A (BPA)	300 W xenon lamp simulating visible light, 10 mg/L, 60 min	80%	below 10% for ZnAlTi-LDH	[38]
CdS/C60 (1 g/L)	One-pot hydrothermal method (0.4 wt %)	RhB	300 W xenon lamp (>420 nm), 20 mL, 10 ppm of RhB	k = 0.089 min−1	1.5-times than CdS	[43]
C3N4/C60 (0.6 g/L)	Simple adsorption (1.0 wt %)	RhB	500 W xenon lamp (>420 nm), 50 mL, 1.0 × 10−5 mol l−1 RhB, 60 min	97%	54% for C3N4	[45]
g-C3N4/C60 (0.5 g/L)	Calcination (0.03 wt %)	MB, phenol	500 W xenon lamp (>420 nm), MB (50 mL, 0.01 mM), phenol (50 mL, 5 ppm).	k1 = 1.036 h−1, k2 = 0.093 h−1	3.2- and 2.9-times than C3N4	[35]
Ag3PO4/C60 (0.5 g/L)	Precipitation (2.0 wt %)	Acid red 18 (AR18)	400 W halogen lamp (420–780 nm, 21.5–23.0 mW cm−2), 50 mL, 6.5 × 10−5 mol/L of AR18, 60 min	90%	53% for Ag3PO4	[31]
Ag3PO4/C60 (1 g/L)	Precipitation (5.0 mg/L)	Methyl Orange (MO)	300 W xenon lamp (>420 nm), 10 mg/L	k = 0.453 min−1	k = 0.028 min−1 for Ag3PO4	[72]
PbMoO4/C60 (0.4 g/L)	Hydrothermal synthesis (0.5 wt %)	RhB	18 W low-pressure mercury lamp as the UV light source, 50 mL of RhB (1 × 10−5 M), 2 h	99%	37% for PbMoO4	[50]
Bi2WO6/C60 (1 g/L)	Simple adsorption (1.25 wt %)	MB, RhB	500 W xenon lamp (>420 nm), 1 × 10−5 mol/L RhB or MB (100 mL)	k1 = 0.0099 min−1, k2 = 0.0454 min−1	5.0- and 1.5-times than Bi2WO6	[40]
BiOCl/C70 (1 g/L)	In-situ growth (1.0 wt %)	RhB	500 W xenon lamp (>420 nm), 10 mg/L, 30 min	99.8%	49.7% and 66.4% for BiOCl and P25 (TiO2)	[39]
Bi2TiO4F2/C60	Solvothermal method (1.0 wt %)	RhB	Visible light, 20 ppm RhB, 120 min	93%	65% for Bi2TiO4F2	[73]
CNTs/BiVO4-C60 (2 g/L)	Hydrothermal synthesis (2.5 wt %)	RhB	300 W xenon lamp (>420nm), 100 mL, 0.01 mmol/L RhB, 30 min	96.1%	74.0% for BiVO4	[51]
CNTs/Bi2MoO6-C60 (2 g/L)	Hydrothermal synthesis (2.5 wt %)	RhB	300 W xenon lamp (>420 nm), 100 mL, 0.01 mmol/L RhB, 30 min	88.4%	43.7% for Bi2MoO6	[51]
Ag3PO4/Fe3O4/C60 (1 g/L)	Hydrothermal synthesis (5.0 wt %)	MB	400 W mercury lamp (>420 nm), 50 mL of MB (25 mg/L), 300 min	95%	33% for Ag3PO4	[74]
TiO2/Pt-C60 (1 g/L)	Sol-gel method (7.5 wt %)	MO	8 W halogen lamp (400–790 nm), 50 mL, 1 × 10−5 mol/L of MO	k = 3.67×10−3 min−1	1.58- and 16.4-times than Pt/TiO2 and TiO2	[75]
TiO2/Pd-C60 (1 g/L)	Sol-gel method (21 wt %)	MB	UV lamp box (8 W, 365 nm), 50 mL, 1 × 10−4 mol/L of MB	k = 0.0337 min−1	14-times than TiO2	[76]
Au/TiO2-C60 (1 g/L)	Hydrothermal synthesis (3.25 wt %)	MO	500W tungsten halogen lamp, 20 mL, 10 mg/L of MO, 160 min	95%	47% for TiO2	[77]
TiO2/CdS-C60 (1 g/L)	Sol-gel method (5.0 wt %)	MB	8 W halogen lamp (400–790 nm), 50 mL, 1 × 10−5 mol/L of MB	k = 7.9×10−3 min−1	4.9- and 3.5-times than CdS/TiO2 and TiO2	[56]
TiO2/WO3-C60 (1 g/L)	Sol-gel method (3.0 wt %)	MO	8 W halogen lamp (400–790 nm), 50 mL, 1 × 10−5 mol/L of MO	k = 4.75×10−3 min−1	1.66- and 21.2-times than WO3/TiO2 and TiO2	[78]
TiO2/CD/C60 (1 g/L)	Simple adsorption (1.5%)	MB, 4-chlorophenol (4-CP)	84 W light sources (>420 nm), MB (10 mL, 144 μM), 10 mg/L 4-CP	k1 = 0.014 min−1, k2 = 0.036 min−1	2- and 4.9-times than TiO2	[79]
TiO2/Fullerol (1 g/L)	Wet impregnation	Procion red MX-5B	16 solar UVA lamps (350 nm)	k = 0.0128 min−1	2.6-times than TiO2	[80]
TiO2/Fullerol (1 g/L)	Wet impregnation (1.0 wt %)	Formic acid	Hg lamp (365 nm)	k = 91.0 µmol L−1 min−1	1.3-times than TiO2	[59]
Nb-TiO2/Fullerol (0.5 g/L)	Simple adsorption	4-chlorophenol	300-W Xe arc lamp (>420 nm)	k = 13.9×10−3 min−1	3.3-times than P25	[81]

k means rate constant of photocatalytic degradation, which is calculated by the relationship between−ln(C/C₀) and t (C₀ and C are the concentrations of pollutant in solution at times 0 and t, respectively).

and

Table 2. Summary of fullerene based photocatalysts for photocatalytic H₂ generation.

Photocatalyst (Additive Amount)	Synthesis Method (Fullerene Content)	Experimental Conditions	Photocatalytic Rate of H2 Generation	Enhancement Factor	Reference
CdS/C60 (0.5 g/L)	Hydrothermal synthesis (0.4 wt %)	300 W xenon lamp (>420 nm), 50 mL aqueous solution containing 10 vol% lactic acid and 1 wt % Pt	1.73 mmol h−1 g−1	2.3 Times of pure CdS	[43]
WO3@C60 (0.5 g/L)	Hydrothermal synthesis (4 wt %)	300 W xenon lamp (>420 nm), Triethanolamine (TEA)	154 µmol h−1 g−1	2 times of pure WO3	[53]
MoS2/C60 (0.5 g/L)	Ball milling method (2.8 wt %)	300 W xenon lamp (>420 nm), 20 mL aqueous solution containing 3.5 mg Eosin Y (EY) and 1 mL TEA	6.89 mmol h−1 g−1	9.3 times of ball-milled MoS2	[54]
g-C3N4/C60 (1 g/L)	Ball milling method (12 wt %)	300 W xenon lamp (>420 nm), 100 mL aqueous solution containing 17.5 mg EY and 5 mL TEA	266 µmol h−1 g−1	4.0 times higher than pristine C3N4	[55]
Cr1.3Fe0.7O3-C60 (5 mg/78 mL)	Simple adsorption (3%)	300 W xenon lamp (>420 nm), 78 mL 10 vol% TEA aqueous solution	220.5 µmol h−1 g−1	2 times of the Cr1.3Fe0.7O3	[82]
Fe2O3/C60 (5 mg/78 mL)	Simple adsorption (0.5~1 wt %)	300 W xenon lamp (>420 nm), 78 mL 10 vol% TEA aqueous solution	β-Fe2O3/C60: 1665 µmol h−1 g−1; α-Fe2O3/C60: 202.9 µmol h−1 g−1; γ-Fe2O3/C60: 169.4 µmol h−1 g−1	β-Fe2O3: 169.4 µmol h−1 g−1; α-Fe2O3: 80.6 µmol h−1 g−1; γ-Fe2O3: 252 µmol h−1 g−1; C3N4: 82.7 µmol h−1 g−1	[83]
CdS/TiO2-C60 (50 mg/80 mL)	An ion-exchanged method (0.5 wt %)	Low power UV-LEDs (420 nm), 80 mL solution (0.25 M Na2S, 0.25 M Na2SO3)	120.6 µmol h−1 g−1	8.5 times of CdS/TiO2	[84]
TiO2/C60-d-CNTs (1 g/L)	Hydrothermal synthesis (5 wt %)	300 W xenon lamp (>420 nm), 100 mL 10 vol% TEA aqueous solution	651 µmol h−1 g−1	208 µmol h−1 g−1 for pure TiO2	[85]
g-C3N4/graphene/ C60 (2 g/L)	Wet impregnation	Light-emitting diode (>420 nm), 50 mL solution containing 1 wt‰ Pt and 10 vol% TEA	545 µmol h−1 g−1	50.8 and 4.24 times of graphene/g-C3N4 and C60/g-C3N4	[19]
(2TPABTz)–metal complex/C60	Simple adsorption (2 wt %)	300 W xenon lamp (>420 nm), an aqueous lactic acid (LA)	2TPABTz-Cu/C60: 4.05 mmol h−1 g−1; 2TPABTz-Co/C60: 3.77 mmol h−1 g−1; 2TPABTz-Ru/C60: 6.12 mmol h−1 g−1	2TPABTz-Cu: 4.05 mmol h−1 g−1; 2TPABTz-Co: 3.77 mmol h−1 g−1; 2TPABTz-Ru: 6.12 mmol h−1 g−1; TiO2 (P25): 0.072 mmol h−1 g−1	[86]
WO3/C60@Ni3B/Ni(OH)2 2 g/L	Photo-deposition technique	500 W xenon lamp (>420nm), 100 mL 10 vol% TEA aqueous solution	1.578 mmol h−1 g−1	9.6 times of pure photocatalyst	[87]

summarize the photocatalytic efficiency in pollutant degradation and H₂ generation over all kinds of fullerene/semiconductor photocatalysts, respectively. Next, detailed photocatalytic activity and mechanisms will be discussed for various types of semiconductor/fullerene photocatalysts, which is accompanied with analyses of electron transfer routes, free radical reactions and stability of photocatalysts.

4.1. Fullerene Based TiO₂ Photocatalysts

TiO₂ is the most extensively used photocatalyst due to its easy availability, strong oxidizing ability, and superior photoelectronic properties [63,64]. With a wide band gap (~3.2 eV), TiO₂ could only be excited under UV light, which limits efficient utilization of solar light [65]. Meanwhile, the fast recombination of photoinduced hole-electron pairs restricts the photocatalytic efficiency of TiO₂ [66]. It has been proven that coupling fullerene with TiO₂ is a helpful way to boost the photocatalytic efficiency of pure TiO₂ both under UV light and visible light irradiation.

Oh, et al. [67] prepared TiO₂/C₆₀ photocatalyst by a heat treatment method with 700 °C. It was shown that the TiO₂/C₆₀ exhibited a more significant effect towards MB degradation with an increase of −ln (C/C₀) values than that of the original TiO₂ under UV light illumination. Yu, Ma, Liu and Cheng [48] successfully fabricated mesoporous TiO₂/C₆₀ powders via a hydrothermal method, which demonstrated that C₆₀ molecules could be dispersed as monolayer or few layers onto bimodal mesoporous TiO₂ via covalent bonding. The 0.5 wt % TiO₂/C₆₀ exhibited the best photocatalytic efficiency for acetone decomposition under UV light irradiation and the degradation rate constant (k) was 13.9 × 10⁻³, reaching 3.3-times that of the pure TiO₂. In this UV-light-driven photocatalytic system, the dominant role of C₆₀ is an inhibitor of rapid recombination of photogenerated hole-electron pairs, leading to boost the quantum efficiency of TiO₂. With respect to TiO₂/C₆₀ photocatalyst, the excited electrons will transfer from TiO₂ to C₆₀ because the conduction band potential of TiO₂ (−0.5 V vs. NHE) is more negative than that of C₆₀/C₆₀^•− (−0.2 V vs. NHE). Under UV light irradiation, the photogenerated electrons are excited from the VB of TiO₂ to the CB, leaving holes in the VB. Generally, these holes and electrons incline to fast recombination and only partial carriers take part in redox. However, after C₆₀ is tightly coupled with TiO₂, excited electrons could further transfer to C₆₀ due to its excellent electron adsorption capacity, which effectively inhibits recombination of photoinduced carriers and supplies more carriers participating in photocatalytic reaction. Besides, C₆₀ derivative (C₆₀(CHCOOH)₂) modified TiO₂ nanoparticles fabricated by Mu, Long, Kang and Mu [61] showed superior photocatalytic efficiency in Cr (VI) reduction under UV light illumination. Compared with pristine TiO₂, the C₆₀-derivative-modified TiO₂ nanocomposites exhibited a higher photocatalytic efficiency of 97% for Cr(VI) reduction within 1.5 h UV irradiation. The electron transfer and radical formation procedure is presented as follows, in Equations (1)−(3):

$${\mathrm{TiO}}_2/{\mathrm{C}}_{60}\stackrel{hv}{\to }{\mathrm{C}}_{60}\left({\mathrm{e}}^{-}\right)/{\mathrm{TiO}}_2\left({\mathrm{h}}^{+}\right)$$

(1)

$${\mathrm{TiO}}_2\left({\mathrm{h}}^{+}\right)+{\mathrm{OH}}^{-}\to {\mathrm{TiO}}_2+{}^{•}\mathrm{OH}$$

(2)

$${\mathrm{C}}_{60}\left({\mathrm{e}}^{-}\right)+{\mathrm{O}}_2\to {\mathrm{C}}_{60}+{}^{•}{\mathrm{O}}_2^{-}$$

(3)

In addition to UV-light-driven photocatalytic activity, TiO₂/C₆₀ nanocomposites also exhibit superior photocatalytic capacity under visible light irradiation. In this visible-light-driven photocatalytic system, the introduced C₆₀ could typically enhance the photocatalytic activity in two ways at the same time: one is to increase the visible light adsorption, the other is to prolong the lifetimes of photoinduced carriers for participating redox reaction. For example, an investigation was conducted into the visible-light-induce photocatalytic activity of TiO₂/C₆₀ towards MB degradation [88]. In this study, two crystals of TiO₂ (anatase and rutile) were coupled with C₆₀ to assemble photocatalysts and the rutile-C₆₀ exhibited significantly superior efficiency than pristine rutile under visible illumination. Grandcolas, et al. [89] synthesized C₆₀ modified TiO₂ nanotubes via a simple impregnation method using ethanol and toluene as co-solvents, and the as-prepared sample exhibited superior efficiency in photocatalytic isopropanol degradation under visible light irradiation. More recently, a polycarboxylic acid functionalized fullerene (C₆₀-(COOH)_n) was coupled with TiO₂ to obtain a novel photocatalyst TiO₂/C₆₀ nanocomposite via ultrasonication-evaporation method for the first time [68]. For the as-prepared photocatalysts, the introduction of C₆₀ obviously decreased the aggregation of pure TiO₂ nanocomposites (Figure 2a,b), and the C₆₀ particles were well-dispersed and closely contacted onto the surface of TiO₂ (Figure 2c). Compared with pure TiO₂, the 1 wt % TiO₂/C₆₀ exhibited stronger both UV and visible light absorption, resulting in improving the utilization of light energy (Figure 2d). In order to trace oxidative species involved in the photocatalytic reaction, in situ radical trapping experiments were made, wherein EDTA was used for trapping holes and 1,4-benzoquinone (BQ) was a scavenger for •O₂⁻. In the presence of EDTA, the photocatalytic degradation efficiency to RhB was dramatically retarded, and a similar trend was also observed with BQ addition (Figure 2e). These results meant that the photoinduced h⁺ and •O₂⁻ were involved in the photocatalytic reaction. Under visible light illumination for 150 min, 1 wt % TiO₂/C₆₀ nanocomposite showed 95% degradation efficiency to RhB, which was significantly higher than pristine TiO₂ (Figure 2f). To further check the stability of the TiO₂/C₆₀ photocatalyst, the recovered composite was used for repeatedly photocatalytic degradation experiment towards RhB. After five repeated experiments under visible light irradiation for 150 min, the degradation efficiency decreased from 95% to around 80%. There is no doubt that the long-term stability of photocatalysts is particularly vital to practical application. Therefore, future work could focus on the synthesis method of TiO₂ to improve the stability of TiO₂/C₆₀ photocatalysts. For example, Bastakoti, et al. [90] reported a high-efficiency method to fabricate more stable mesoporous metal oxides (including TiO₂, Ta₂O₅ and Nb₂O₅).

Compared to TiO₂, C₇₀ is a close-shell configuration consisting of 35 bonding molecular orbitals with 70 p-electrons [91]. Similar to C₆₀, C₇₀ has higher electron acceptability and higher efficiency of light harvesting over TiO₂ [92,93]. Thus, C₇₀ is a promising alternative to boost the photocatalytic efficiency of TiO₂. Cho, et al. [94] fabricated both TiO₂/C₆₀ and TiO₂/C₇₀ nanowire to estimate their photocatalytic activity. In this study, TiO₂/C₇₀ showed a significantly stronger absorbance within 400~650 nm and a lower photoluminescence spectra (PL) than TiO₂/C₆₀. This means the C₇₀ displayed better efficiency in boosting visible light absorption and inhibiting recombination of hole-electron pairs than C₆₀. Accordingly, the TiO₂/C₇₀ nanowire displayed higher photocatalytic activity for MB degradation than TiO₂/C₆₀ in the visible light irradiation. Furthermore, Wang, Liu, Dai, Cai, Chen, Yang and Huang [69] assembled TiO₂-C₇₀ hybrids using acid-treated C₇₀, Ti(SO₄)₂ and cetyltrimethylammonium bromide (CTAB) by a hydrothermal method. It was proven that the 8.5 wt % TiO₂-C₇₀ showed the best photodegradation efficiency to sulfathiazole under visible light, which was 4.2 times that of TiO₂ + C₇₀ mixture and 1.6 times that of the corresponding TiO₂-C₆₀ nanocomposite, respectively. The SEM and TEM images of 18 wt % TiO₂-C₇₀ nanocomposite are shown in Figure 3a,b. After C₇₀ introduction, the surface of TiO₂-C₇₀ nanocomposites were uneven, which could increase the BET of the as-prepared samples. The C₇₀ particles were well-dispersed onto the outer boundary of TiO₂ composites, and it was estimated that a monolayer of C₇₀ was covered onto the surface of TiO₂. Additionally, the TiO₂-C₇₀ exhibited better light absorption and higher separation efficiency of hole-electron pairs than those of TiO₂-C₆₀ and pure TiO₂ (Figure 3c,d). It is important to highlight that novel mechanisms of TiO₂/fullerene were proposed for photocatalytic pollutant degradation [69]. From two aspects, UV light and visible light irradiation, the mechanisms are described as follows. Under UV light illumination, electrons are excited from VB to CB of TiO₂, leaving holes in the VB. Then the CB electrons of TiO₂ could rapidly transfer to C₇₀, because the CB potential of TiO₂ (0.5 V vs. NHE) is more negative relative to C₇₀/C₇₀⁻ (0.2 V vs. NHE). In the meantime, the ground-state C₇₀ is excited to a transient-state ¹C₇₀*, then undergoes rapid intersystem crossing (ISC) to a lower lying triplet state ³C₇₀*. In this system, excited electrons can be injected into the three-states transform procedure of C₇₀, resulting in suppression of their falling back to the VB of TiO₂. Hence, this process effectively inhibits the recombination of photoinduced hole-electron pairs, so as to elevate the photocatalytic activity of TiO₂/C₇₀ nanocomposite (Figure 3e). On the other hand, a viewpoint of mid-gap band was proposed for TiO₂/C₇₀ with respect to the visible-light-driven photocatalytic mechanism. It was pointed out that the mid-gap band came into being between TiO₂ and C₇₀ ascribing to the strong chemical boning of these two materials. The electron transfer route was obviously distinct from those TiO₂/fullerene photocatalysts in previous studies. Namely, visible light excites electrons from the VB of TiO₂ to the mid-gap band and then from the mid-band to the CB of TiO₂, leaving holes in the VB (Figure 3e). This procedure significantly prolongs the lifetime of the photoinduced carriers and facilitates the separation of hole-electron pairs for participating in photocatalytic reaction. Accordingly, the effectively separated holes and electrons participate in generation of reactive radical species. The e^⁻ could react with dissolved O₂ to produce •O₂⁻ and h⁺ react with H₂O to produce •OH, then these radical species cause the degradation of sulfathiazole. To research the stability of TiO₂/C₇₀ nanocomposite, the recycled sample was dried for subsequently repeated experiments. After 5 runs (totally 15 h visible light illumination), the degradation efficiency to sulfathiazole over TiO₂/C₇₀ slightly declined, remaining over 90%. This evidenced that the aforementioned hydrothermal method was effective in building stable TiO₂/fullerene photocatalysts.

Oh and Ko [95] fabricated Pt-fullerene/TiO₂ nanocomposites via in-situ growth method using Pt-treated oxidized fullerene and TNB. Firstly, fullerene was oxidized by MCPBA and treated through ion exchange using potassium hexachloroplatinate (IV) (K₂[PtCl₆]), wherein Pt-treated oxidized fullerene was obtained. Then, Pt-fullerene was added into TNB solution (titanium source) for fabricating Pt-fullerene/TiO₂ nanocomposite via a sol-gel method under mild condition (50 °C). The as-prepared sample exhibited elevated performance under UV light and the order of photocatalytic efficiency for MB degradation was: Pt-fullerene/TiO₂ > fullerene/TiO₂ ˃ pristine TiO₂, due to the synergetic effects of Pt, oxidized-fullerene and TiO₂. In this study, it was proposed that Pt-fullerene was homogeneously covered with TiO₂ particles, wherein TiO₂ would be mounted in a 3-dimensional matrix. It was concluded that three factors contributed to the superior photocatalytic activity of Pt-fullerene/TiO₂, including photocatalytic reaction of the supported TiO₂, decomposition of the organo-metallic reaction by the Pt compound and energy transfer effects of fullerene. Through the same method, a number of metal-treated fullerene/TiO₂ composites were prepared for photocatalytic application as well, such as Fe-C₆₀/TiO₂, V-C₆₀/TiO₂ and Pd-C₆₀/TiO₂ [96,97]. For instance, Meng, Zhang, Zhu, Park, Ghosh, Choi and Oh [76] fabricated M-fullerene/TiO₂ (M representing Pt, Y or Pd) composites to compare their photocatalytic efficiency. Among these samples, the Pd-fullerene/TiO₂ showed the best photocatalytic activity for MB decomposition under UV light, due to its better dispersion and larger BET surface over the Pt-fullerene/TiO₂ and Y-fullerene/TiO₂. Further results indicated that the synergistic effects between Pd and fullerene improves the photocatalytic activity of TiO₂, including enhancement of light adsorption by fullerene and Pd as the final electron-acceptor. More recently, Islam, Hangkun, Ting, Zubia, Goos, Bernal, Botez, Narayan, Chan and Noveron [77] prepared AuNPs-TiO₂-C₆₀ composites, wherein the introduction of C₆₀ significantly boosts the photoactivity and photostability of AuNPs-TiO₂. It was reported that C₆₀ played threefold roles in the preparation of AuNPs-TiO₂-C₆₀. The introduction of C₆₀ decreased the size of AuNPs (5 nm) and effectively prevented its agglomeration on the surface of TiO₂, as well as linked AuNPs to TiO₂ surface without any functionalization. The AuNPs-TiO₂-C₆₀ had a broader light adsorption region over pristine TiO₂ and AuNPs-TiO₂ nanocomposites, ranging from 500 to 650 nm. The 4.76% optimal AuNPs-TiO₂-C₆₀ sample showed 95% photodegradation efficiency towards MO after visible light irradiation for 160 min, which was 2 times higher than pristine TiO₂.

Meng and his co-workers assembled a series of semiconductor/fullerene/TiO₂ ternary photocatalysts via sol-gel method, such as TiO₂/CdS/C₆₀, TiO₂/CdSe/C₆₀ and TiO₂/WO₃/C₆₀ nanocomposite [56,78,98]. For example, the TiO₂/CdS/C₆₀ photocatalyst exhibited superior efficiency in photocatalytic pollution degradation and the MO degradation rate (K) of these as-formed nanocomposites was in an order: TiO₂/CdS/C₆₀ > TiO₂/C₆₀ > TiO₂ > TiO₂/CdS. In addition, Lian, Xu, Wang, Zhang, Xiao, Li and Li [84] successfully fabricated C₆₀ decorated TiO₂/CdS mesoporous photocatalyst via an evaporation combined with ion-exchanged method. It is noteworthy that the BET of the TiO₂/CdS/C₆₀ photocatalyst was actually lower than that of the TiO₂/CdS composite, which resulted from the fact that the C₆₀ was inset into the pore of this mesoporous composite. Compared with CdS/TiO₂, the TiO₂/CdS/C₆₀ presented stronger visible light adsorption, lower recombination of photogenerated hole-electron pairs and higher photocurrent density, thus resulting in highly effectively photocatalytic ability for H₂ production (Figure 4a–c). In Na₂S-Na₂SO₃ reaction solution, the H₂ generation rate of the optimal 0.5 wt % TiO₂/CdS/C₆₀ photocatalyst was 6.03µmol h⁻¹ g⁻¹ with 2.0% quantum efficiency (QE) under visible light, which was obviously higher than the rate of TiO₂/CdS (0.71 µmol h⁻¹ g⁻¹). As concluded in this study, C₆₀ could enhance the light adsorption and facilitate the separation of photogenerated hole-electron pairs, as well as serve as H₂ generation site for adsorbing and reducing H⁺ ions. The electron transfer route and reaction mechanism are presented in Figure 4d. Under visible light illumination, electrons in the VB of CdS are excited into the CB firstly. Then the excited electrons in the CB of CdS rapidly transfer into the CB of TiO₂, because the conduction band potential of the former is more negative than that of the later. Finally, the electrons in the CB of TiO₂ transfer to C₆₀, which provide reaction sites for reducing H⁺ to H₂. Meanwhile, the leaving holes in the VB of the CdS are consumed by S²⁻ and SO₃ ²⁻ to facilitate H₂ generation efficiency. Moreover, Chai, Peng, Zhang, Mao, Li and Zhang [85] developed a TiO₂-C₆₀-dCNTs photocatalyst, which was reported to enhance the photocatalytic H₂ production under UV light illumination. At a 5 wt % loading amount of C₆₀, it exhibited the highest H₂ production rate of 651 µmol h⁻¹, which is 2.9 times that of bare TiO₂.

Fullerenol (C₆₀(OH)_x), also called polyhydroxyfullerene (PHF), is a water-soluble fullerene derivative [99,100]. Typically, PHF could be prepared using fullerene via acid hydrolysis or alkali hydrolysis method [101,102,103]. In the earlier time, Krishna and his co-workers found that the addition of PHF in solution could elevate the photocatalytic activity of TiO₂ under UV light illumination [104]. The reaction solution with PHF + TiO₂ showed 2.6 times faster of photocatalytic organic dye degradation and 1.9 times faster of Escherichia coli inactivation than that of solution with TiO₂ alone. While the hydroxylated fullerene (PHF) changes the electronic properties and decreases the electron affinity of fullerene, further studies were conducted by his group to explore the mechanisms of PHF to enhance the photocatalytic activity of TiO₂ [105]. It was proposed that PHF covers onto the surface of TiO₂ by electrostatic interactions in solution. The electron paramagnetic resonance (EPR) results showed that higher production rate of •OH was achieved under UV light after addition of PHF in solution, which contributed to enhancement of TiO₂ photocatalytic activity. However, PHF alone in solution did not generate •OH under UV light, which suggested that synergistic effects come into being between PHF and TiO₂. A hypothesis was proposed that PHF can scavenge the photo-generated electrons from TiO₂ and meanwhile the synergistic effects of PHF and TiO₂ can induce more •OH generation for enhancing the photodegradation activity. Furthermore, Park, et al. [106] proposed a new approach of CT in PHF/TiO₂ system under visible light named as “surface-complex CT” procedure, which had not yet been proposed in TiO₂/C₆₀ system before. Note that fullerol has numerous hydroxyl groups which may link to the surface of TiO₂ through the CT-complex route (Equation (4)). The photocurrents (I_ph) of the PHF/TiO₂-coated electrodes were examined, confirming that the transfer orientation of photogenerated electrons was from PHF to TiO₂. In such surface-complex CT procedure, PHF serves as a photosensitizer in which electrons could be excited into CB of TiO₂ under visible light irradiation, hence elevating the photocatalytic efficiency of TiO₂. The reaction mechanism of radical generation is presented in Equations (5)−(8). Similar CT situations have been proposed in benzene/TiO₂ system and even polycyclic aromatic hydrocarbons (PAH) physical-adsorbed TiO₂ system as well [107,108].

$${\mathrm{C}}_{60}{\left(\mathrm{OH}\right)}_{\mathrm{x}}+\equiv \mathrm{Ti}–\mathrm{OH}\to \equiv \mathrm{Ti}–\mathrm{O}–{\mathrm{C}}_{60}{\left(\mathrm{OH}\right)}_{\mathrm{x}-1}+{\mathrm{H}}_2\mathrm{O}$$

(4)

$$\mathrm{Fullerol}/{\mathrm{TiO}}_2+hv(\mathsf{\lambda }>420\mathrm{nm})\to \left({\mathrm{fullerol}}^{•+}\right)/{\mathrm{TiO}}_2\left({{\mathrm{e}}_{\mathrm{cb}}}^{-}\right)$$

(5)

$$\left({\mathrm{fullerol}}^{•+}\right)/{\mathrm{TiO}}_2\left({{\mathrm{e}}_{\mathrm{cb}}}^{-}\right)\to \mathrm{fullerol}/{\mathrm{TiO}}_2$$

(6)

$$\mathrm{Fullerol}/{\mathrm{TiO}}_2\left({{\mathrm{e}}_{\mathrm{cb}}}^{-}\right)\to {\mathrm{TiO}}_2/\left({\mathrm{fullerol}}^{•-}\right)$$

(7)

$$\left({\mathrm{fullerol}}^{•+}\right)/{\mathrm{TiO}}_2+0.5{\mathrm{H}}_2\mathrm{O}\to \mathrm{fullerol}/{\mathrm{TiO}}_2+{\mathrm{H}}^{+}+0.25{\mathrm{O}}_2$$

(8)

Bai, Krishna, Wang, Moudgil and Koopman [80] assembled a PHF/TiO₂ nanocomposite by a physically mixing method in aqueous suspension and then coated the as-prepared sample onto grout substrate to examine its photocatalytic activity. The nanocomposite coating at a TiO₂/PHF ratio of 0.01 exhibited the best photocatalytic efficiency under UV light irradiation. Accordingly, the 0.01 TiO₂/PHF photocatalyst coating exhibited 2 times higher of Procion red MX-5B photocatalytic degradation efficiency and 3 times higher of photocatalytic spores of Aspergillus niger inactivation than these of bare TiO₂ coating. Similarly, Hamandi, Berhault, Dappozze, Guillard and Kochkar [59] fabricated TiO₂/PHF nanotubes using PHF and TiO₂ nanotubes via wetness impregnation together with heat treated at 400 °C Under UV light illumination, the optimum 1% TiO₂/PHF nanotubes showed a rate constant values (K_exp) of 94.7 µmolL⁻¹ min⁻¹ for photocatalytic degradation towards formic acid, while TiO₂ nanotubes alone exhibited the K_exp of 72.6 µmolL⁻¹ min⁻¹. Moreover, Lim, Monllor-Satoca, Jang, Lee and Choi [81] developed a Nb-TiO₂/fullerol nanocomposite, which proved an elevated visible-light-driven photocatalytic performance. In brief, Nb-doped TiO₂ was fabricated by a sol-gel method then the Nb-TiO₂ was dispersed into fullerol solution buffered at pH 3 with HClO₄. After stirring for 3 h, the filtered solids were dried at 80 °C to acquire brownish particles designated as Nb-TiO₂/fullerol. The Nb-TiO₂/fullerol showed more effectively photocatalytic performance for the reduction of Cr (VI), oxidation of iodide and degradation of 4-chlorophenol than naked TiO₂, Nb−TiO₂ and TiO₂/fullerol under visible light. These results indicated that the synergistic effects between fullerol and Nb improved the photocatalytic activity of TiO₂. It was proved that Nb doping induced vacancies of TiO₂ by ionic substitution of Nb⁵⁺ in Ti⁴⁺ position, which could suppress the photoinduced hole-electron pairs recombination by trapping electrons. Notably, the fullerol significantly enhanced the visible light absorption of Nb-TiO₂ through a surface-complex CT mechanism. Under visible light irradiation, electrons will be excited from HOMO of fullerol to CB of TiO₂ and then from CB to vacancies, which effectively enhances the charge transport and prolongs the lifetime of photoinduced carriers. Another advantage was proposed that the Nb-TiO₂/fullerol showed more highly photochemical stability over typical dye-sensitized-TiO₂.

4.2. Metal Oxides (Except TiO₂)/Fullerene Photocatalyst

In addition to TiO₂, other metal oxides have also been promising materials in photocatalytic application [109,110,111]. Fullerene (C₆₀ and C₇₀) has some advantages to enhance the photocatalytic efficiency of these metal oxides, such as enhancing the light absorption and inhibiting recombination of photogenerated hole-electron pairs. Accordingly, a number of metal-oxide/fullerene photocatalysts have been successfully synthesized and extensively applied in photocatalytic pollutant degradation and H₂ evolution via water splitting, such as ZnO/C₆₀ (C₇₀), WO₃/C₆₀ and SnO₂/C₆₀ [25,112].

Similar to TiO₂, ZnO is an alternative photocatalyst with a band gap of 3.3 eV [113]. Generally, the absorption edge of pristine ZnO locates the near UV region, which usually restricts its photocatalytic efficiency. Meanwhile, the susceptibility to photocorrosion is also another barrier of ZnO for satisfactory photocatalytic performance. Fu, Xu, Zhu and Zhu [70] successfully prepared a C₆₀ hybridized ZnO nanocomposite by a simple absorption method, and the sample with 1.5 wt % C₆₀ exhibited the best photocatalytic sufficiency in MB degradation. The 1.5 wt % ZnO/C₆₀ composite showed 95% photocatalytic degradation sufficiency towards MB under UV light, which was 3-times as high as that of bare ZnO. In such system, the elevated performance was ascribed to the improved light adsorption and a higher separation efficiency of photoinduced hole-electron pairs. During the photocatalysis process, the photogeneration holes could easily react with surface oxygen atom, leading to fast decline of photocatalytic activity of ZnO. When C₆₀ was covered on ZnO, the activity of surface oxygen atoms was effectively reduced so that more holes could participate in photocatalytic reaction. Furthermore, the photocorrosion experiment indicated that the C₆₀-hybridized ZnO nanocomposite did not show obviously decline of photocatalytic sufficiency even after illumination under UV light for 50 h, which was highly superior in stabilization than bare ZnO. Hence, the introduction of C₆₀ effectually suppress photocorrosion of ZnO. Similarly, Hong, et al. [114] successfully prepared a ZnO/C₇₀ nanocomposite via a heat treatment method, which exhibited superior photocatalytic degradation of organic dyes.

Recently, Tahir, Nabi, Rafique and Khalid [53] proposed the elevated photocatalytic efficiency of WO₃/C₆₀ nanocomposite for dye degradation and H₂ evolution. The optimized 4 wt % WO₃/C₆₀ sample showed the best photodegradation ability and the degradation efficiency in MB, RhB and MO under visible light illumination was 93%, 92% and 91%, respectively. Meanwhile, the H₂ evolution rate of the 4 wt % WO₃/C₆₀ was 2-times higher than that of bare WO₃. After coupling with C₆₀, the band gap of WO₃/C₆₀ nanocomposites were lower than that of bare WO₃, which could excite more electrons of semiconductor WO₃ from VB to CB. The BET surface area of these composites was also significantly increased, which could enhance the adsorption reaction. Based on the aforementioned study, Shahzad, Tahir and Sagir [87] constructed a novel heterogeneous photocatalyst WO₃/fullerene@Ni₃B/Ni(OH)₂ for H₂ production. As a co-catalyst, Ni₃B/Ni(OH)₂ was loaded onto WO₃/fullerene thin film by a facile photo-deposition technique. The optimal WO₃/fullerene@1.5%Ni₃B/Ni(OH)₂ presented an outstanding photocatalytic efficiency in H₂ generation, reaching 1578µmol h⁻¹ g⁻¹. In this system, it was proposed that three factors mainly contributed to the superior performance, including inhibition for recombination of photogeneration hole-electron pairs, more active sites for photocatalytic reaction and synergistic effect between nanocomposite and co-catalyst. Even earlier, a ternary photocatalyst WO₃/C₆₀/TiO₂ was successfully prepared via a sol-gel method. Its photocatalytic performance in MO degradation was higher than that of WO₃/C₆₀ or TiO₂/C₆₀, which means these three materials synergistically enhance the photocatalytic activity [78].

In addition, Song, Zhang, Zeng, Wang, Ali and Zeng [83] fabricated a series C₆₀ modified Fe₂O₃ polymorphs (α-, γ- and β-Fe₂O₃) photocatalysts via a simple adsorption method. These as-prepared samples showed superior photocatalytic efficiency in H₂ production and even extremely outstanding effects were observed in the presence of fluorescein. Under visible light irradiation, the photocatalytic capacity was in the order: 1C₆₀/β-Fe₂O₃ > 1C₆₀/γ-Fe₂O₃> γ-Fe₂O₃ > 1C₆₀/α-Fe₂O₃> β-Fe₂O₃ > g-C₃N₄ > α-Fe₂O₃. Behera, Mansingh, Das and Parida [71] proposed a ZnFe₂O₄-fullerene photocatalyst for norfloxacin decomposition and Cr (VI) reduction, wherein fullerene introducing significantly improved the photocatalytic capacity of ZnFe₂O₄. Moreover, Song, Huo, Liao, Zeng, Qin and Zeng [82] successfully prepared a novel photocatalyst Cr_2-xFe_xO₃/C₆₀ via a simple adsorption method. In this study, α-Fe₂O₃ (~2.2eV) and Cr₂O₃ (~3.4eV) were integrated through a sol-gel method in order to construct Cr_2-xFe_xO₃ with a suitable band gap for H₂ generation. While poor electron conducting ability limited its further application, the introducing of C₆₀ was an effective way. The optimal 3%C₆₀/Cr_1.3Fe_0.7O₃ sample presented the H₂ generation rate of 220.5 µmol h⁻¹ g⁻¹, which was about 2-times of the bare Cr_1.3Fe_0.7O₃ composite.

4.3. Metal Sulfide/Fullerene Nanocomposites

Nowadays, metal sulfide semiconductors have attracted extensive attentions in photocatalytic application due to their distinctive optical-electrical characteristic [115,116,117]. CdS is an appealing photocatalyst with narrow bandgap (2.2~2.4 eV) exhibiting superior visible-light respond [118]. While fast recombination of hole-electron pairs and photocorrosion effect are the two main obstacles of naked CdS, which inhibits its photocatalytic efficiency. Coupling with fullerene was proved to be an effective way to boost the photocatalytic performance of CdS. Accordingly, Cai, Hu, Zhang, Li and Shen [43] successfully assembled a CdS/C₆₀ nanocomposite via one-pot hydrothermal synthesis. The as-prepared samples showed better separation efficiency of photoinduced hole-electron pairs and higher photocurrent density than pure CdS (Figure 5a,b). Thus, the improvement of the aforementioned features contributed to a highly photocatalytic activity over CdS/C₆₀ nanocomposite. The optimal H₂ production rate of 0.4 wt % CdS/C₆₀ was 1.73 mmol h⁻¹ g⁻¹ under visible light illumination, which was 2.3-times higher than that of naked CdS (Figure 5c). Its photocatalytic degradation efficiency towards RhB achieved 97% in 40 min (Figure 5d). Furthermore, the photostability of CdS was significantly boosted after CdS combining with C₆₀, and 97.8% of RhB degradation efficiency actually retained after three recycles (Figure 5e). In order to estimate the stability of CdS/C₆₀, the released Cd²⁺ concentration was determined in remaining solution after three cycles for RhB degradation (Figure 5f). The Cd²⁺ concentration was 381.3µg/L in solution with naked CdS while it was 51.9µg/L in solution with 0.4 wt %CdS/C₆₀ nanomaterial, in which the former was 7.3 times of the later. The results above indicated that C₆₀ could effectively inhibit the photocorrosion and boost the stability of CdS. Furthermore, Meng, Peng, Zhu, Oh and Zhang [56] assembled a novel ternary CdS/TiO₂/C₆₀ photocatalyst via a sol-gel method. The introduction of C₆₀ definitely induced 56% increasement of the BET surface of the CdS/TiO₂ composite, which could enhance the adsorption effect. Under the same condition, the MO degradation rate (K) of these nanocomposites was in an order: CdS/TiO₂/C₆₀ > TiO₂/C₆₀ > TiO₂ > CdS/TiO₂. This meant the CdS/TiO₂/C₆₀ composite obtained superior photocatalytic capacity owing to the synergistic reaction of C₆₀, TiO₂ and CdS. It was concluded in this study that the synergistic effects were as follows: (1) C₆₀ could increase the quantum efficiency and charge transfer, as well as enhance the adsorption effect of the ternary photocatalyst; (2) Combining CdS with TiO₂ endows the photocatalyst with a suitable bandgap for visible-light respond and a more effective electron transfer route for generating more •OH and •O₂⁻.

In addition, Meng and co-workers assembled CoS/C₆₀ and AgS/C₆₀ nanocomposites for pollutant decomposition [52,119]. Superior photocatalytic efficiency was obtained in these photocatalysts after the introduction of C₆₀, since C₆₀ is an energy sensitizer that could improve the quantum efficiency and boost charge transfer efficiency. More recently, Guan, Wu, Jiang, Zhu, Guan, Lei, Du, Zeng and Yang [54] fabricated a MoS₂/C₆₀ heterostructure photocatalyst via a ball milling method. The method does not need solvent to dissolve MoS₂ and C₆₀ and can significantly increase the BET surface of product. In this study, it was the first time to propose that a van der Waals heterostructure formed between MoS₂ and C₆₀ through ball milling, detailly wherein C₆₀ nanoparticles bounded onto the edge of the exfoliated MoS₂ nanosheet by non-covalent bond. Noteworthily, the CB minimum of MoS₂/C₆₀ was more negative than that of ball-milled MoS₂ and the VB maximum of the former was more positive than that of the later. Thus, the as-prepared MoS₂/C₆₀ photocatalyst featured more suitable band gap for elevating the H₂ evolution. Under visible light irradiation, the optimal 2.8 wt % MoS₂/C₆₀ sample exhibited the photocatalytic H₂ production rate of 6.89 mmol h⁻¹ g⁻¹ in the presence of EY as a photosensitizer, which was 9.5 times higher than that of ball-milled MoS₂ without C₆₀.

4.4. Bismuth-Based Semiconductor/Fullerene Composites

Bismuth-based semiconductors have been proven promising materials for photocatalytic application, including BiOX (X = Br, Cl and I), Bi₂WO₆, BiVO₄, Bi₂MoO₆, and so on [2,120,121]. Considerable research efforts have been devoted to couple these bismuth-based semiconductors with fullerene (C₆₀ or C₇₀) and enhanced photocatalytic performance could be obtained. For example, Zhu, Xu, Fu, Zhao and Zhu [40] successfully prepared C₆₀ modified Bi₂WO₆ photocatalyst via a simple absorbing process, and 1.25 wt % Bi₂WO₆/C₆₀ displayed 5.0-times the photocatalytic degradation activity towards MB with respect to unmodified Bi₂WO₆. Similarly, Ma, Zhong, Li, Wang and Peng [39] fabricated C₇₀ modified BiOCl by an in-situ preparation procedure and superior photocatalytic performance was observed. Under solar irradiation for 30 min, 49.7% of RhB was degraded over pure BiOCl while 99.8% of RhB could be degraded over 1 wt % BiOCl/C₇₀. In addition, Li, Jiang, Li, Lian, Xiao, Zhu, Zhang and Li [73] successfully developed Bi₂TiO₄F₂/C₆₀ photocatalyst via a solvothermal method, which was a hierarchical microsphere structure. The introduction of C₆₀ can increase the photocurrent of the as-prepared sample, resulting from more efficient mobility efficiency of the charge carriers (Figure 6a). Owing to strong combining and heterojunction formation, the Bi₂TiO₄F₂/C₆₀ nanocomposite showed obviously elevated photocatalytic capacity for degrading RhB relative to bare Bi₂TiO₄F₂ under visible light irradiation (Figure 6b). Meanwhile, the photocatalyst exhibited excellent stabilization as well and highly photocatalytic efficiency of RhB degradation (≈ 80%) was maintained even after eight circles (Figure 6c). The photocatalytic mechanisms of Bi₂TiO₄/C₆₀ nanocomposite are described in Figure 6d. Apart from organic pollutant, bromate (BrO³⁻) also exhibits biotoxicity to aquatic organisms and human since its properties non-biodegradation and accumulation. A strategy for controlling BrO³⁻ pollution is to reduce it to Br^⁻ which is naturally present in surface water bodies. Therefore, Zhao, Liu, Shen and Qu [46] studied the photocatalytic performance of Bi₂MoO₆/C₆₀ for removal BrO³⁻ under visible light. After modification with C₆₀, Bi₂MoO₆/C₆₀ exhibited sharply increase in photocatalytic reduction of BrO³⁻, attributed to the enhanced separation rate of photogenerated electron-hole pairs.

4.5. Carbon Nitride/Fullerene Composites

Graphitic carbon nitride (g-C₃N₄) is an attractive metal-free photocatalyst, which was developed by Wang et al. in 2009 [122]. The pristine g-C₃N₄ features a medium band gap (2.5~2.7 eV) with good visible light response [123]. Currently, this effective organo-photocatalyst has been widely used for pollutant degradation, water splitting and CO₂ reduction [124,125]. However, pristine g-C₃N₄ exhibits insufficient solar absorption and rapid recombination of photogenerated carriers, which limits its photocatalytic efficiency. It has been proven that coupling g-C₃N₄ with fullerene is an effective way to enhance the photocatalytic sufficiency in pollutant degradation and H₂ evolution. Recently, a series of g-C₃N₄/fullerene nanocomposites have been fabricated and they showed elevated photocatalytic efficiency.

Chai, Liao, Song and Zhou [45] prepared g-C₃N₄/C₆₀ nanocomposites via a simple adsorption method. After C₆₀ introduction, g-C₃N₄/C₆₀ nanocomposites enhanced the visible light absorption without changing the absorption edge of g-C₃N₄, as well as lowered the recombination of photogenerated hole-electron pairs. The 1 wt % C₃N₄/C₆₀ showed the highest photodegradation performance towards RhB, which could reach 97% degradation efficiency under visible light after 60 min. The reaction process could be proposed as follows (Equations (9)–(12)):

$${\mathrm{C}}_{60}/{\mathrm{C}}_3{\mathrm{N}}_4\stackrel{hv}{\to } {\mathrm{C}}_{60}\left({\mathrm{e}}^{-}\right)/{\mathrm{C}}_3{\mathrm{N}}_4\left({\mathrm{h}}^{+}\right)$$

(9)

$${\mathrm{C}}_{60}\left({\mathrm{e}}^{-}\right)+{\mathrm{O}}_2\to {\mathrm{C}}_{60}+{}^{•}{\mathrm{O}}_2^{-}$$

(10)

$${}^{•}{{\mathrm{O}}_2}^{-}+2{\mathrm{e}}^{-}+2{\mathrm{H}}^{+}\to {}^{•}\mathrm{OH}+{\mathrm{OH}}^{-}$$

(11)

$$\mathrm{RhB}+{\mathrm{h}}^{+}({}^{•}\mathrm{OH},{}^{•}{{\mathrm{O}}_2}^{-})\to \mathrm{products}$$

(12)

Bai, Wang, Wang, Yao and Zhu [35] introduced C₆₀ into g-C₃N₄ matric via thermal treatment of C₆₀ and dicyandiamide mixture at 550 °C and the obtained nanocomposites exhibited higher photooxidation degradation efficiency towards phenol and MB. Relative to physical blend, this thermal treatment gave rise to strong interface interaction between g-C₃N₄ and C₆₀. The g-C₃N₄/C₆₀ nanocomposite exhibited higher specific surface area than pristine g-C₃N₄, leading to more active sites for catalytic reaction. The photocurrent value of g-C₃N₄/C₆₀ is 4.0-times that of g-C₃N₄, which was highly responsible for enhancing the photocatalytic activity of pristine g-C₃N₄. Moreover, the introduction of C₆₀ decreased the band gap of C₃N₄, wherein the value of g-C₃N₄ and g-C₃N₄/C₆₀ were severally 2.70 eV and 2.58 eV, respectively. The calculation results showed that the valence band maximum (VBM) of C₃N₄/C₆₀ is 0.17V lower than that of g-C₃N₄, which meant a stronger oxidizing capacity. Under visible light, the degradation ability of g-C₃N₄/C₆₀ towards phenol and MB were 2.9- and 3.2-times as high as that of pristine g-C₃N₄, respectively. In addition, a series of C₃N₄/fullerene (C₆₀, C₇₀) photocatalysts were prepared by Ouyang, et al. [126] via a hydrothermal method for disinfection of bacterial pathogens under visible light irradiation. Regarding disinfection of E. coli O157:H7, both C₃N₄/C₆₀ and C₃N₄/C₇₀ hybrids showed stronger bacterial inactivation than pristine C₃N₄ after 4 h irradiation, and the C₃N₄/C₇₀ exhibited the best performance. Note that both •O₂⁻ and •OH were identified as radical species to destruct bacterial cell in the solution under the visible light irradiation.

Coupling g-C₃N₄ with C₆₀ could elevate the photocatalytic ability to H₂ generation as well. For instance, Chen, Chen, Guan, Zhen, Sun, Du, Lu and Yang [55] successfully synthesized a covalent bonding g-C₃N₄/C₆₀ nanocomposite via ball milling with LiOH as catalyst, which was the first time using this method for fabricating semiconductor/C₆₀ nanocomposite. As depicted in XRD image, the lattice structure of g-C₃N₄ nanomaterial was changed after the attachment of C₆₀ component (Figure 7a). It was proven that the covalent bonds were formed in the C₃N₄/C₆₀ nanocomposite and a new peak at 399.5 eV was detected in XPS spectra, which was ascribed to N-C₆₀ bond (Figure 7b). In this study, a new viewpoint was proposed that C₆₀ forms covalent bond with g-C₃N₄ by a four-membered ring of azetidine. However, g-C₃N₄/C₆₀ alone hardly exhibited the photocatalytic capacity in H₂ evolution, thus additional photosensitizer was necessary. Under visible light, the H₂ generation rate of g-C₃N₄/C₆₀ was 266 µmol h⁻¹g⁻¹ using EY as a photosensitizer, which was 4.0 times higher than that of g-C₃N₄ in the same condition (Figure 7c). As depicted in Figure 7d, the mechanisms of g-C₃N₄/C₆₀ nanocomposite are described for photocatalytic H₂ generation. Recently, a novel g-C₃N₄/graphene/C₆₀ composite was successfully prepared and significant enhancement for H₂ evolution ability of the photocatalyst was observed [19]. In the presence of Pt (cocatalyst) and triethanolamine (sacrificial agent), the H₂ evolution rate of the g-C₃N₄/graphene/C₆₀ was 5449.5µmol g⁻¹ within 10 h, which was 50.4 and 4.24 times that of g-C₃N₄/graphene and g-C₃N₄/C₆₀, respectively. It means that C₆₀ and graphene mutually reinforced synergy in H₂ generation of g-C₃N₄, owing to high conductivity of graphene and excellent electron-attracting capacity of C₆₀. Meanwhile, the quantum yield of g-C₃N₄/graphene/C₆₀ reached 7.2% within 72 h.

4.6. Other Semiconductor/Fullerene Photocatalysts

Other semiconductors have also been coupled with fullerene to boost the photocatalytic activity. For example, Dai, Yao, Liu, Mohamed, Chen and Huang [50] successfully fabricated a PbMoO₄-C₆₀ photocatalyst via a hydrothermal method. After introduction of C₆₀, no obvious change was visible in lattice structure of PbMoO₄, but defects were observed on the surface of PbMoO₄ (Figure 8a,b), which could be owed to the decreased crystallinity. As depicted in energy dispersive spectrometry (EDS), a great deal of C element was evenly dispersed on the surface of PbMoO₄-C₆₀ nanocomposite, which was regarded as a layer coating of C₆₀ moiety (Figure 8c,d). Upon the attachment of the C₆₀ moiety, the PbMoO₄-C₆₀ composite displayed obvious enhancement of both UV and visible light absorption (Figure 8e). Meanwhile, the Eg of 5.0 wt % PbMoO₄-C₆₀ (3.08 eV) was narrower than that of pure PbMoO₄ (2.93 eV) (Figure 8f). Therefore, the improvement of optical features and energy band structure contributed to highly photocatalytic efficiency. Song, Yang, Chen and Zhang [72] prepared Ag₃PO₄/C₆₀ photocatalyst via a simple chemical precipitation method. The photodegradation efficiency of MO achieved 93.5% within 8 min of visible light illumination. It is noteworthy that the introduction of C₆₀ significantly enhanced the stabilization of Ag₃PO₄ which was usually susceptible to photo-corrosion. Additionally, some organic semiconductor nanoparticles composing of fullerene exhibited superior photocatalytic performance. For example, Huo and Zeng [86] successfully fabricated a triphenylamine functionalized bithiazole metal complex hybridized C₆₀ photocatalyst. Under visible light irradiation, the photocatalytic H₂ evolution of the as-prepared photocatalyst showed approximately 4–6-times higher than that of the pristine complex without fullerene. In this photocatalytic system, the organic metal nanocomposite worked as two roles which were both a photosensitizer and a photocatalyst. Additionally, Zhang, et al. [127] prepared an organic photocatalyst of fullerene hydrolyzed aluminum phthalocyanine chloride (AlPc/C₆₀) by a reprecipitation method. The photocatalyst showed superior photooxidation degradation of various organic compounds (including N-methyl-2-pyrrolidone (NMP), methanal, and 2-mercaptoethanol). Note that the AlPc/C₆₀ exhibited highly efficiency in complete mineralization towards these organic materials, leading to effective CO₂ generation in reaction solution under visible light irradiation. NMP mineralization experiment was tested in a closed cylindrical reactor containing 10 vol% substrate, wherein the CO₂ generation amount in AlPc/C₆₀ reacting solution reached 3.7 × 10⁻⁷ mol after 24 h irradiation, which was 2.9-times higher than that of the corresponding C₆₀ + AlPc mechanical mixture in solution. It was proposed that this novel photocatalyst based on a biphase structure and features p/n junction-like characteristics.

4.7. Discussions and Conclusions for Photocatalytic Applications of Fullerene/Semiconductor Photocatalysts

Among fullerene-based photocatalysts, the TiO₂/fullerene (C₆₀ and C₇₀) composites have been the most extensively investigated in photocatalytic applications in the past decades. They exhibit efficient performances in wastewater treatment, such as pollutant degradation and disinfection. The synthetic methods are facile and eco-friendly without complicated steps, generally including simple adsorption and hydrothermal synthesis. After fullerene is inserted into TiO₂, it shows to be fairly helpful in enhancing the photocatalytic efficiency of TiO₂. However, there is a small deficiency in these materials, namely insufficient utilization of light energy. In other words, they exhibit excellent absorption of UV light but moderate absorption of visible light, while UV irradiation only accounts for 4% in solar irradiation. Besides, metal sulfide/fullerene nanocomposites are an appealing class of photocatalysts for not only decontamination but water splitting for H₂ generation. They exhibit efficient absorption of visible light together with a moderate stability, such as CdS/C₆₀, MoS₂ and CdS/TiO₂/C₆₀. Compared with the responding pure metal sulfide, they perform significantly boosted efficiency, especially towards H₂ generation. As to Bi-based semiconductor/fullerene photocatalysts, they appear to only be helpful for pollutant degradation but have not displayed effective capacity for photocatalytic H₂ generation. There is no doubt that g-C₃N₄ has always been one of the hot nanomaterials in photocatalytic area since advent, so g-C₃N₄/fullerene photocatalysts are promising options for further photocatalysis. The facile synthetic method makes them attractive materials for photocatalytic applications, such as simple thermal treatment and balling mill. Additionally, it is easy to achieve more intensively oxidation or reduction capacity with tunable bandgap of g-C₃N₄.

To sum up, three crucial characteristics of photocatalysts are needed to be considered for wastewater treatment, such as high-efficiency for removing pollutant, stability in duration and nontoxicity to the environment and humans, respectively. In order to convincingly evaluate a photocatalyst, it is required to concern various properties comprehensively, such as optical absorption, energy band, photocatalytic efficiency, stability, cost and so on. As we all know, photocatalytic applications are currently researched in the laboratory, which seldom involves the time consumption of process and economic efficiency. Future work is imperative to focus on these aspects.

5. Fullerene/Support (Non-Semiconductor) Photocatalysts for Wastewater Treatment

In addition to fullerene/semiconductor photocatalysts, a series of novel fullerene/solid-support photocatalysts have been developed for wastewater treatment. It is well established that pristine fullerene is extremely insoluble in water (solubility of C₆₀ in water < 10⁻⁹ mg/L), but could dissolve in nonpolar organic solvent, such as toluene and 1,2-dichlorobenzen [128,129]. It is worthwhile mentioning that fullerene solution could induce photochemical reactive oxygen species (ROS) generation via two pathways which were defined as type Ⅰ pathway (Equation (13)) and type Ⅱ pathway (Equation (14)), taking C₆₀ as an example as follows [22]. For example, single oxygen ¹O₂ can be produced in fullerene-toluene solution (pathway Ⅱ), and O₂^−• and •OH can be generated in solvent in the present of electron donors such as EDTA and NADH (pathway Ⅰ) under UV light illumination [130,131]. While easily aggregation of pristine fullerene in water impedes ROS production owing to self-quenching mechanisms within the aggregates [132,133]. It has been proven that coupling fullerene with hydrophilic functional groups (namely fullerene derivatives) is a helpful strategy to dissolve fullerene in water together with superior ROS generation, such as polyhydroxyl-fullerene, amine-fullerene and other cationic-fullerenes [134,135,136]. At the earlier time, these water-soluble fullerene derivatives were used as photosensitizers for photodynamic therapy, selective antimicrobial and photooxidation organic synthesis [27,137]. More recently, considerable research efforts have been devoted to wastewater treatment for fullerene-support photocatalysts, including photocatalytic pollutant degradation and disinfection in aqueous solution. Without support combination, aqueous fullerene (nC₆₀) and fullerene derivatives in aqueous solution are easily decomposed due to photolysis and other external conditions, seriously lowering efficacy of C₆₀ as a photocatalyst generating ROS [138,139]. The separation and reutilization are also the barriers of fullerene derivatives used as photocatalysts. Herein, immobilization of fullerene-derivatives on solid support could be a hopeful strategy to fabricate fullerene/solid-support photocatalyst.

$${}^{1}C_{60}\stackrel{h\nu }{\to }{}^{1}C_{60}^{*}\stackrel{ISC}{\to }{}^{3}C_{60}^{*}\stackrel{{}^{3}O_2\to {}^{1}O_2}{\to }{}^{1}C_{60}$$

(13)

$$\begin{array}{r}{}^{1}C_{60}\stackrel{h\nu }{\to }{}^{1}C_{60}^{*}\stackrel{ISC}{\to }{}^{3}C_{60}^{*}\stackrel{e^{-}-donor}{\to }{C}_{60}^{-}•\\ E_{red}=+1.14V\end{array}$$

(14)

Lee, et al. [140] fabricated a series of aminoC₆₀/silica photocatalysts by covalent-bond immobilization of aminoC₆₀ on 3-(2-succinic anhydride) propyl functionalized silica gel. The synthesis route of the aminoC₆₀/silica photocatalysts is presented in Figure 9. In this photocatalytic system, phosphate buffer was required for pollutant degradation and ¹O₂ generated by photochemical procedure was the dominating ROS for photocatalytic activity. The photochemical ¹O₂ generation ability of the as-prepared photocatalysts were estimated using furfuryl alcohol (FFA) as an indicator and immobilized aminoC₆₀ samples exhibited remarkedly higher ¹O₂ generation than water-soluble aminoC₆₀, among which tetrakis aminoC₆₀/silica performed the best (Figure 10a). Accordingly, these aminoC₆₀/silica photocatalysts boosted the photocatalytic oxidation degradation towards pharmaceutical pollutants (including ranitidine and cimetidine) as well as photocatalytic disinfection towards MS-2 bacteriophage upon visible light illumination in contrast to corresponding aminoC₆₀ alone in aqueous solution. In this case, the immobilization method facilitated well dispersion of C₆₀ onto the silica surface, so as to expose more active sites for ROS generation, resulting in enhancement of the photocatalytic efficiency. Additionally, the huge specific surface area of silica significantly enhanced the adsorption of pollutant to the surface of aminoC₆₀/silica for promoting closer contact between ¹O₂ and the pollutant, because the travel distance of ¹O₂ in aqueous is really short within the diffusion length of ∼125 nm over one lifetime [141]. Note that the lifetime of ¹O₂ in water was only 3~4 µs which limits the catalytic performance, so immobilization of aminoC₆₀ could increase the contact time between ¹O₂ and the pollutant from this point of view [142]. In order to further explore the performance of the tetrakis aminoC₆₀/silica photocatalyst, a variety of emerging organic contaminants and endocrine disruptors were involved into photocatalytic degradation experiments [143]. The photodegradation rate of ranitidine and propranolol for amino silica/C₆₀ were 13.987 ± 0.016 h⁻¹ and 10.77 ± 0.019 h⁻¹, which were respectively 31-fold and 75-fold faster than that for aimnoC₆₀ alone. In particular, the silica/aminoC₆₀ was quite effective in trimethoprim degradation while no degradation appeared in aminoC₆₀ aqueous solution, which was also observed in a C₆₀ aminofullerene-magnetite nanocomposite suspension solution [144]. Moreover, at alkaline conditions (pH 10), acetaminophen, bisphenol A, and 4-chlorophenol could also be effectively degraded over the silica/aminoC₆₀ photocatalyst under fluorescent light irradiation. Figure 10b,c compares the photocatalytic efficiency of silica/aminoC₆₀ with other semiconductor photocatalysts including TiO₂, C-TiO₂ and Pt/WO₃. It is shown that silica/aminoC₆₀ exhibits remarkedly higher ¹O₂ generation rate over TiO₂ and C-TiO₂ while it performs lower rate than Pt/WO₃ under fluorescent light or visible light irradiation. Note that the silica/aminoC₆₀ exhibits the best efficiency in pharmaceutical (RA and CM) degradation among these materials under visible light irradiation. Although these photoactive catalysts in the comparisons take effect owing to different ROS generation (e.g., primarily ¹O₂ upon aminoC₆₀, •OH upon TiO₂ and Pt/WO₃), it is indicated that silica/aminoC₆₀ has a potential for application as an alternative environmental photocatalyst.

There is no denying that aforementioned fullerene/solid photocatalysts are involved into complicated synthesis procedure with fullerene derivatives. Moor and Kim [145] used a simpler method to build solid supported C₆₀ photocatalysts via a nucleophilic reaction of a terminal amine onto pristine C_60′s cage, and through this method SiO₂/C₆₀ and polystyrene resin/C₆₀ (PS/C₆₀) were developed. The above two photocatalysts both showed higher ¹O₂ generation rate than nC₆₀ (nanoscale aggregates) in aqueous solution under various illumination conditions, in which the SiO₂/C₆₀ was superior to PS/C₆₀. As a photosensitization catalyst, SiO₂/C₆₀ showed effectively photocatalytic MS2 inactivation, which was ascribed to ¹O₂-mediated oxidization damage effect. In addition, Moor, Valle, Li and Kim [91] successfully fabricated a MCM-41/C₇₀ composite by the same nucleophilic reaction and it was the first time to use C₇₀-solid support photocatalyst for wastewater treatment (Figure 11a). Within photoinactivation experiment towards MS2, the MCM-41/C₇₀ performed obviously higher efficiency than N-TiO₂ nanocomposite, which efficiently induced •OH production in aqueous solution for microbial inactivation (Figure 11b,c). It was confirmed that the as-prepared novel MCM-41/C₇₀ photocatalyst exhibited efficient photodegradation ability to several pharmaceuticals and personal care products (PPCP), including bisphenol A, 17-α-ethynylestradiol and amoxicillin (Figure 11d,e).

The majority of photocatalysts researched for traditional azo-dye decomposition based upon semiconductor composites. Few studies involved into non-semiconducting materials, especially C₆₀/solid support photocatalyst [146]. Whereas, Wakimoto, et al. [147] prepared a SiO₂/C₆₀ powder via a simple adsorption using pristine C₆₀ and silica gel in toluene and it was important to highlight that a novel route for dye photodegradation was proposed for the C₆₀/solid support photocatalyst. Unlike typically semiconductor photocatalysts, the SiO₂/C₆₀ composite exhibited effectively visible-light-driven photocatalytic degradation towards methyl orange in the presence ascorbic acid, while SiO₂/C₆₀ alone without ascorbic acid did not show degradation performance. In this case, ascorbic acid could protonate MO and transform it into quinoid form with strong electron-acceptability. It was proved that both ¹O₂ and O₂^•− species took part in the dye degradation. On one hand, C₆₀ dispersed onto silica surface was excited to undergo the intersystem crossing from the single to triplet state for ¹O₂ generating, then the generated ¹O₂ could attack rich-electron quinoid structure to decompose MO. On the other hand, the electron transferred from ascorbic acid to the excited C₆₀ to forming the C₆₀ radical anion, followed by generation of O₂^•− via O₂ receiving the electron from C₆₀^•−, wherein O₂^•− was an effective specie for dye degradation. It was also confirmed that methyl red could be decomposed by SiO₂/C₆₀ at the same conditions. Likewise, Kyriakopoulos, et al. [148] successfully fabricated a MCM-41/C₆₀ photocatalyst via a dry impregnation method. Coupling C₆₀ with MCM-41 significantly increased the BET of the as-prepared photocatalyst, thus effectively dispersing C₆₀ clusters as well as strengthening its adsorption ability to pollutant. The optimum 3MCM-41/C₆₀ (3 wt % C₆₀) sample showed 74.9% decolorization efficiency in Orange G, which was markedly higher than that of C₆₀ alone. This photocatalyst proved to be remarkably stable, wherein less than 5% photodegradation efficiency was lost after five cycles.

Fullerene-based solid photocatalysts could effectively prevent fullerene aggregations and enhance the photo-stabilization of fullerene alone as well as enhance the pollutant adsorption due to the introduced support, thus leading to increase the photocatalytic efficiency. In the meanwhile, this fullerene-based photocatalyst could perform selected oxidization of pollutant with ¹O₂ that can prevent natural organic matter (NOM) interference, which underline the potential of these materials for wastewater treatment in natural water. Therefore, fullerene-based support photocatalysts are promising materials for environmental applications and further effort will be required to fabricate novel photocatalysts of this type.

6. Conclusions and Perspectives

In summary, the essence of photocatalysis bases on the ROS generation in the presence of light resource irradiation. Fullerenes, including C₆₀ and C₇₀, have been extensively investigated in the photocatalytic application due to their unique optical and photochemical characteristics. Fullerene could be anchored on semiconductors to enhance their photocatalytic activity, and also supported on non-semiconductor solids to fabricate novel fullerene-based photocatalysts due to its self-photocatalytic features. In the present review, fullerene/semiconductor photocatalysts and fullerene-solid support photocatalysts are summarized for wastewater treatment (pollutant degradation, Cr(VI) reduction, disinfection etc.) and water splitting for H₂ generation. A number of synthesis methods have been used to fabricate semiconductor/fullerene photocatalysts, including simple adsorption, hydrothermal synthesis, ball milling, sol-gel and so on. The semiconductors alone usually display limited photocatalytic performance, wherein the fast recombination of photoinduced hole-electron pairs and inefficient light energy utilization are the two main obstacles. Whereas, fullerene could availably enhance the photocatalytic efficiency of semiconductors by retarding the recombination of hole-electron pairs and increasing the light absorption (UV and visible light). In some cases, semiconductor/fullerene photocatalysts display better stabilization than semiconductors alone, and the introduced fullerene increases the BET of the semiconductor for enhancing the pollutant adsorption. The studies manifest that excess fullerene inhibits the photocatalytic ability due to the coverage effect towards excited sites. So, a suitable amount of fullerene is imperative for superior photocatalytic activity of semiconductor/fullerene photocatalysts. Plentiful semiconductors have been coupled with fullerene for wastewater treatment and water splitting, such as TiO₂, ZnO, CdS and C₃N₄. The photocatalytic effect of these photocatalysts are presented and the involved mechanisms are discussed in detail in this review, including the reaction of ROS generation and the transfer route of electron. On the other hand, fullerene-solid support photocatalysts are also discussed in application for wastewater treatment, such as silica/C₆₀, MCM-41/C₇₀ and polysiloxane-supported fullerene photocatalyst. They display excellent photoinduced ROS (mainly ¹O₂) generation in aqueous solution after fullerene was dispersed onto the solid support, which is the direct factor contributing to the photocatalytic reaction. Meanwhile, they effectively enhance the photo-stabilization of fullerene alone as well as enhance the pollutant adsorption. It is noteworthy that these fullerene-based photocatalysts perform selected oxidization of the pollutant, wherein the photoinduced ¹O₂ could prevent natural organic matter (NOM) interference. So, it underlines the potential of these materials for wastewater treatment in natural water.

Although some encouraging properties have been achieved for fullerene-based photocatalysts, the development of fullerene-based photocatalysts still has remaining challenges. (1) The interface contact between the fullerene and semiconductor is not so intimate, leading to limiting the electron transport ability and photostability of semiconductor/fullerene photocatalysts. The majority of semiconductor/fullerene photocatalysts formed by simple adsorption and hydrothermal synthesis, while few studies confirmed the existence of covalent bond or other tight bond in them. (2) The photocatalytic mechanisms of fullerene/semiconductor photocatalysts are partly not clear. Some studies indicated that the electrons transfer from the semiconductor to the fullerene due to the strong electron-accepting ability of fullerene, while others deemed that the electrons transfer in the opposite direction. (3) Not enough research has been done on the novel fullerene/solid (non-semiconductor) photocatalysts, in which the selection of solid support is restricted to silica, MCM-41 and polysiloxane. (4) The majority of fullerene-based photocatalysts are investigated in the treatment of simulated wastewater with the artificial addition of a single pollutant, and actual industrial wastewater is rarely involved.

In our opinion, several directions are worthy of attention for fullerene-based photocatalysts in the future: (1) Innovative strategy should be developed to construct semiconductor/fullerene photocatalysts with efficient performance and high stability. (2) Further works should focus on mechanism studies of the semiconductor/fullerene composite, especially the electron-transfer path which is still in dispute. (3) More attention should be paid to fullerene derivatives, which are promising materials for developing novel fullerene-based photocatalysts. (4) The studying of fullerene-based photocatalysts on their actual performance in natural water, industrial wastewater and multi-polluted wastewater.