A Review on the Progress and Future of TiO₂/Graphene Photocatalysts

Abstract

TiO₂ is seen as a low cost, well-known photocatalyst; nevertheless, its sluggish charge kinetics does limit its applications. To overcome this aspect, one of the recent approaches is the use of its composites with graphene to enhance its photoactivity. Graphene-based materials (nanosheets, quantum dots, etc.) allow for attachment with TiO₂ nanostructures, resulting in synergistic properties and thus increasing the functionality of the resulting composite. The current review aims to present the marked progress recently achieved in the use of TiO₂/graphene composites in the field of photocatalysis. In this respect, we highlight the progress and insights in TiO₂ and graphene composites in photocatalysis, including the basic mechanism of photocatalysis, the possible design strategies of the composites and an overview of how to characterize the graphene in the mixed composites. The use of composites in photocatalysis has also been reviewed, in which the recent literature has opened up more questions related to the reliability, potential, repeatability and connection of photocatalytic mechanisms with the resulting composites. TiO₂/graphene-based composites can be a green light in the future of photocatalysis, targeting pollution remediation, energy generation, etc.

1. Introduction

Photocatalysis involves the direct use of perpetual sunlight as an energy source and therefore has become a spotlight of mainstream researchers. TiO₂ is a promising semiconductor photocatalyst due to its low toxicity, low cost, stability and a favorable bandgap and band edge position for photocatalytic applications. The search for a composite photocatalyst of TiO₂ arises from the limitations of using TiO₂, which include low absorption in the visible range of the solar spectrum, sluggish charge transfer kinetics and, primarily, the inevitable precious metal consumption for obtaining satisfactory performance in practical applications [1,2]. Overcoming the limitations present in photocatalysis over TiO₂, as well as the search for metal-free photocatalysts for the modification of TiO₂ with carbon-based materials, in particular with graphene, has been a focus of current research [1,3,4]. Recently, abundant research towards TiO₂/graphene-based photocatalysts has been witnessed, as evidenced by the increasing number of publications reported in the last three years (2022—12,300, 2021—15,200, 2020—11,900, source: Google scholar, keyword: TiO₂/graphene photocatalyst).

Graphene is an exceptional material with a two-dimensional carbon sheet structure that presents very unique electrical, electronic, optical, thermal and surface features. Combined with a semiconductor, it can serve as a good electron sink or a support to separate and transport the charge effectively in photocatalytic reactions [3,4]. The key lies in its favorable bandgap and unique electrical and electronic properties, which make it a potential macromolecular: (i) photocatalyst, (ii) photosensitizer under visible illumination or (iii) cocatalyst in a composite photocatalyst [5,6]. Its unique surface can also provide copious active sites for redox reactions and to chemisorb molecules [7,8]. Its polar derivative, graphene oxide (GO) or reduced graphene oxide (RGO), has been reported to form unique mixed composites with polymers [9] and semiconductors [10] in view of in-demand applications, including, for instance, water remediation, energy storage devices, photocatalysis, etc. [11]. They possess a variety of oxygen functional groups (carboxyl, hydroxyl, carbonyl and epoxy) which can form chemical bonds with TiO₂, and this results in composites with compatible interfaces and synergistic features [12].

Herein, we have highlighted the critical aspects and key features in TiO₂/graphene photocatalysts, considering also the recent marked interest in this field. We have first included a brief overview of the basics involved in photocatalysis over TiO₂, followed by a discussion of the crucial factors affecting the use of TiO₂ or graphene oxide as a photocatalyst, including the role of graphene in a photocatalytic system. The progress in obtaining and characterizing these composites together with their use in photocatalysis has also been overviewed, resulting in a “how to” guide. This review provides a one-click overview of TiO₂/graphene photocatalysts, including current status, insights, progress, limitations to be addressed and future use.

2. Photocatalysis Over TiO₂: The Basics

TiO₂ is one of the most used semiconductor photocatalysts due to its nontoxicity, stability in aqueous solutions, good quantum efficiencies and suitable bandgap [13]. Under illumination, electrons are promoted to the conduction band (CB) of TiO₂ from the valence band (VB), thereby leaving an electron vacancy, or the hole in the VB, as seen in Figure 1. However, this happens only if the energy of the photon is equal to or greater than the bandgap energy of the photocatalyst, i.e., in our case, TiO₂ [14]. Moreover, for TiO₂ photocatalysts, UV light illumination results in photoexcitation as a result of the bandgap energy of the TiO₂ photocatalyst, which, for anatase is 3.2 eV (λ = 385 nm) and for rutile is 3.0 eV (λ = 410 nm) [15].

Photoexcitation results in the formation of photogenerated electron–hole pairs, and the generated charge pairs can either contribute to a chemical reaction with adsorbed/surface species or undergoes recombination [14,16]. In a typical TiO₂ photocatalyst, the electron–hole pair can then follow the ensuing sequence of chemical reactions. Namely, the CB electron interacts with molecular oxygen and leads to the formation of a superoxide radical anion (O₂^●− ), and the VB hole contributes to the water/hydroxide ion oxidation and hence the formation of a hydroxyl radical (OH^●) [14,16]. These generated radicals can then be further involved in different reactions as a result of photocatalysis over TiO₂, e.g., in reactions such as nitrogen fixation [17], CO₂ reduction [18], hydrogen gas production [19], and the degradation of different pollutants [20]. In the presence of a pollutant (i.e., rhodamine B [21], congo red, methyl orange, or methylene blue dye [22]), VB holes and CB electrons can directly interact with the pollutant species and lead to pollutant mineralization due to water and dioxygen radical formation [23,24].

2.1. Mechanism of Oxidation

In a general photocatalytic reaction, the photogenerated holes oxidize water (commonly the surface of a photocatalyst contains water and is generally referred to as absorbed water). As previously mentioned, the oxidation of water results in the formation of OH^● radicals, which are, by nature, strong oxidizing agents and decompose dye [25]. In the presence of an organic dye and oxygen, there is the formation of some intermediate radicals in organic compounds [26]. This can result in radical chain reactions along with the consumption of oxygen and finally in the decomposition of organic matter into water and carbon dioxide [27,28,29,30].

2.2. Mechanism of Reduction

The reduction process in photocatalysis usually occurs in the presence of air; therefore, it can be simply seen as the reduction of oxygen. The oxygen reduction reaction occurs as an alternative to hydrogen generation reaction [29], and this is due to the fact that oxygen is a relatively easily reducible species. The O₂^●− radicals formed as a result of the interaction between the CB electrons with dissolved oxygen further react with the intermediate organic products formed during the oxidation reaction, thereby either producing peroxide or decomposing to hydrogen peroxide and then to water. Such reduction reactions are more frequent when photocatalytic processes occur in organic-compound-containing media, rather than in pure water. Therefore, the presence of organic matter on the TiO₂ surface leads to the hindrance of charge recombination and hence to promoting photocatalytic activity, and as more organic matter is present, there are more positive holes as well [27,28,29,30].

3. TiO₂—Dependent Factors Affecting Photocatalysis

Photocatalysis is one of the most frequently used green sustainable technologies for environmental remediation [31]. Remarkable research and development has been conducted in semiconductor-photocatalysis-based applications such as solar cells, water splitting, photodynamic therapy, bacterial disinfection and, particularly, for the removal of environmental pollutants [32]. In this context, for a decade, titanium dioxide (titania or TiO₂) has remained one of the most extensively investigated semiconductor photocatalysts, owing to its suitable band gap (anatase TiO₂—3.2 eV), chemical stability and non-toxicity [33]. TiO₂ nanophotocatalysts provide increased effective photocatalytic activity under real conditions due to their higher specific surface area compared to bulk TiO₂ [34].

The overall photocatalytic activity of TiO₂ can be influenced by the physiochemical, chemical and optoelectronic properties of the TiO₂ photocatalysts, which consequently vary depending on the synthesis method, reaction conditions (precursors, reaction medium or solvent, pH, time, reaction temperature, calcination temperature, etc.) and photocatalytic experimental conditions, such as pH, concentration, illumination power, etc. [35,36,37]. The key properties of TiO₂ that can be tuned by the synthesis method include the crystal phase, crystal lattice, exposed crystal facets, uncoordinated surface sites, interface, specific surface area, degree of crystallinity, geometry, size, porosity, composition, lattice defects, etc. [38,39,40]. From these various properties, the crystal phase, morphology, pore size, band gap, active sites, use of sacrificial agent, dopants, cocatalysts and interfacial defects are the leading factors directing the efficiency of the TiO₂ photocatalyst’s photocatalytic activity. Some of these key factors that influence the photocatalysis of TiO₂ are further discussed below.

3.1. Morphology

The morphology of the TiO₂ photocatalyst not only affects the separation of photogenerated charge carriers but also the type of the defects that originate during the synthesis. Nanostructured TiO₂ exhibits an increased number of active sites, as a result of the higher surface-to-volume ratio, and subsequently, such nanostructured photocatalysts that also have a larger specific surface area show improved photocatalytic activity and efficiency. Different TiO₂ nanostructures, such as quantum dots [41], nanoparticles [42], nanotubes [43], nanorods [44], nanosheets [45], nanoflakes [46] and hierarchically porous monoliths [47], can be prepared by sol-gel, hydrothermal/solvothermal, flame hydrolysis, water-in-oil microemulsion and chemical vapor deposition [48]. The morphology and geometry of TiO₂ nanostructures can also modify the optoelectronic properties (e.g., mid-band-gap electronic states) that can shift the absorption spectrum and alter the charge separation/migration.

Recently, Li et al. [46] investigated the impact of different TiO₂ morphologies on photocatalytic hydrogen evolution. Structure engineering via the thermal transformation of protonic titanate as a template was studied for the synthesis of TiO₂ nanoflakes (length: 100−200 nm, width: 50−100 nm), nanorods (length: 3−7 μm, width: 800−1000 nm), nanowires (length: 2−5 μm, width: 80−120 nm) and nanoflowers (diameter: 500 nm) by varying experimental conditions. The following fabrication steps were followed, as schematically represented in Figure 2: (a) nanoflakes by hydrothermal alkalization, the protonation of as-prepared titanium glycolate precursor (TGP) and calcination; (b) nanowires by hydrothermal and reflux alkalization of TGP and by calcination; (c) nanoflowers by hydrothermal alkalization and protonation and by calcination; (d) nanowires by hydrothermal alkalization, protonation and calcination of TiO₂ nanoparticles; and (e) nanoflowers by solvothermal method using tetrabutyl orthotitanate (TBT) in the presence of mixed solvents N,Ndimethylformamide (DMF) and isopropyl alcohol (IPA) and then by calcination. The highest hydrogen evolution rate of 3.63 mmol g⁻¹h⁻¹ was observed for TiO₂ nanoflakes as compared to nanorods, nanowires and nanoflowers. The exceptional photocatalytic hydrogen evolution was attributed to: (i) the negative conduction band position, (ii) fast photoinduced charge carrier separation and (iii) abundant active sites.

3.2. Phase

The three main polymorphs of TiO₂, namely, anatase, rutile and brookite, are stable under ambient and low-pressure conditions, and in addition, their relative stability varies with particle size. For example, rutile in a bulk form is thermodynamically the most stable polymorph of TiO₂, and at the nanoscale, anatase is the most stable phase. Brookite is considered a metastable form [49]. The geometric (crystallographic orientations) and electronic structure of TiO₂ influence the photocatalytic activity, i.e., anatase (indirect band gap, E_g = 3.2 eV) and rutile (direct band gap, E_g = 3.0 eV) show the maximum photoactivity among all phases. Typically, anatase is considered the most promising phase due to its higher value of reduction potential and slow recombination rate of photoexcited electrons and holes, but it also has a limited absorption in the ultraviolet (UV) region of only 4% of the solar spectrum (as a consequence of the wider band gap) [50,51].

3.3. Surface

The surface chemistry, active sites, reactive facets and interfaces are additional crucial factors that define the efficiency of the TiO₂ photocatalysts [52]. The promotion of the proton transfer between the different interfaces improves the surface chemical reactions and hence the efficiency of the photocatalyst [53]. Modification strategies for tailoring the surface properties of TiO₂ photocatalysts through doping [54], co-doping [55], defect creation [56,57], sensitization [56], surface coatings and heterojunction formation using functional nanomaterials (oxides, plasmonic metals, quantum dots, graphene-based nanomaterials, macromolecules, etc.) [52,58,59] have been widely investigated to enhance the photocatalytic performance.

3.4. Light Harvesting

One of the limiting factors for TiO₂ photocatalysis is its narrow photocatalytic region, namely, due to its wider band gap, TiO₂ is UV-active. Further electronic structure modulation in TiO₂ photocatalysts can be performed by forming nanostructures, doping, co-doping, sensitizer doping, semiconductor coupling, heterojunction formation, oxygen vacancies formation, cocatalyst loading and defect engineering to transform it into a visible light active photocatalyst [60,61]. The band gap modification facilitates photoinduced charge separation and migration and improves the photocatalytic activity [14,62,63,64,65]. Wang et al. [66] reported the growth of nanorods and ultrathin nanowires on graphene aerogel for the formation of TiO₂ NR@GA and TiO₂ ultrathin NW@GA composites, where the graphene had a significant influence on the growth of TiO₂. Numerous influencing factors such as phase, morphology, size, band gap and intermediate energy levels were investigated as a function of heterogeneous interfaces.

4. Graphene in Photocatalysis: Benefits and Impacts

Graphene, a dream material, is an atom-thick sheet of sp²-hybridized carbon and has been in continuous use in scientific and engineering communities since its discovery [11,12,13]. One of the most fascinating applications of graphene is in designing heterogeneous semiconductor/graphene composite photocatalysts [14,15]. Heterogeneous photocatalysis is a promising and green process, well-known in the research community [16,17], since Fujishima and Honda discovered water splitting on TiO₂ photoelectrodes. To obtain high efficiencies in photocatalytic processes, photocatalysts and their design are crucial, and most investigated photocatalysts include TiO₂ [18], g-C₃N₄ [19], Fe₂O₃ [20], ZnO [21], SnO₂ [22], and SnS₂ [23], CdS [67] CuO₂ [24], BiVO₄ [68], MoO₃ [25] and MoS₂ [69]. So far, most of the photocatalysts show poor stability and quantum yield and are rarely activated by visible light, posing a major challenge in the design of heterogeneous photocatalysts for practical applications. In this respect, composites based on graphene and semiconductors possess the potential to overcome these limitations, for instance, resulting in increased absorption in the visible range and higher photocatalytic activity [26].

The fundamental properties of a photocatalyst which influence a photocatalytic reaction are exceptional electrical conductivity and a strong electron-accepting ability, work function and chemical, physical and surface properties. These properties depend on the unique structure, interface and surface features, electronic structures, surface chemistry, charge transport, activation ability for certain reactants and molecular adsorption [2,27]. The properties of graphene can be tuned by suitable chemical modification, for instance, by controlling the level of oxidation which influences its electrical conductivity, or by its structural order [28,29]. Below, the intrinsic and tunable features of graphene and its role in a photocatalytic reaction are discussed.

4.1. Photocatalysis-Dependent Properties of Graphene

The performance of graphene and its derivatives varies with the nature of graphene, for example, with the level of oxidation which affects the electrical conductivity and semiconducting/semimetallic nature of graphene, the crystal structure and a variety of functional groups over its surface, and these aspects are discussed in detail below.

4.1.1. Crystal Structure

A crucial factor in photocatalysis is the electronic properties, which, for graphene, are exceptional due to its high-quality two dimensional crystal lattice. The band structure of graphene is rather unique and primarily consists of sp² hybridized carbon, where the valence orbital of the carbon consists of a 2p orbital (2p_z) perpendicular to the basal plane of graphene and three planar δ orbitals at 120° with each other. The hexagonal lattice of graphene is like a honeycomb and possesses two equivalent carbon sublattices per unit cell, referred to as A and B. The π electrons are delocalized in the conjugated network of carbon and are between the two adjacent carbons that are associated with π* (anti-bonding, the lowest unoccupied CB) and π (bonding, the highest occupied VB) bands [2]. These two bands meet at six points, known as neutrality or Dirac points. Due to the graphene’s crystal symmetry, two points, K and K′, which are independent of one another, can show the linear dispersion of orthogonal π and π* states without interactions in graphene. From the point of view of its electronic properties, the contact of the two bands at Dirac points further indicates that intrinsic monolayer graphene is a semimetal or zero-gap semiconductor [2,6]. In GO, an oxidized derivative of graphene, the VB maxima is composed of an O2p orbital rather than the usual π orbital present in graphene, and the CB minima is composed of an antibonding π* orbital [6].

4.1.2. Semimetallic Properties

Pristine graphene is a zero-bandgap, semimetallic material and therefore has a strong ability to accept electrons, which is a key point in promoting the activity of graphene-based photocatalysts. Specifically, in photocatalysis, graphene can accept, store and transfer photogenerated electrons and thereby can serve as an alternative electron sink/reservoir and a conductive support. Thus, graphene can extend the lifetime of photoinduced charge carriers, enhance charge separation and extraction and subsequently promote the photocatalytic activity of graphene-based photocatalysts [70]. These properties of graphene typically depend on an appropriate work function, in addition to the electrical conductivity. The contact barrier between a semiconductor and graphene can be controlled by the graphene’s work function (please note that the latter determines the band alignment of different materials). The work function of graphene increases with the increase in graphene layers and can be as large as that of graphite (≈4.6 eV) [71]. Furthermore, it can be changed by additional methods, such as introducing atomic defects (hole doping) [72] or doping (metal chloride, nitrogen) [73,74]. Moreover, the conductivity of graphene can be affected by the presence of sp³-hybridized carbon, the density of oxygenated groups present in the graphene nanosheets and the vacancy/defects that influence the electron density of states, thereby affecting the conductivity of graphene [75].

4.1.3. Semiconducting Properties

As discussed above, the nature of pristine graphene can be changed by its modification/oxidation. For instance, GO is the oxidized form of graphene, and the bandgap can be adjusted by tuning its level of oxidation. The electronic features of GO depend on its oxidation level, i.e., from pristine graphene, a semimetal, to GO, a semiconductor or an insulator [76]. Jung et al. [75] described the gradual change in GO (an electrical insulator) to graphene (a semimetal) during a thermal reduction process, thus showing an intimate relationship between the electrical properties and chemical structure of GO. The dependence of the GO’s optical properties on its oxidation level was further confirmed [77,78]. Mainly, the presence of oxygen atoms bonded to carbon, the sp³ bonds or/and the other defects are responsible for an energy gap in the density of electron states. The nature of the graphene/GO depends on the value of the bandgap. The RGO shows a gradual decrease in the optical band gap from 3.5 eV to 1 eV as the C/O ratio increases [79].

4.1.4. Electrical Conductivity

As mentioned before, graphene possesses exceptional electrical and electronic features due to the electrons confined in 2D geometrical symmetries. It is evident that semimetallic graphene has a high electrical conductivity, whereas GO is electrically insulating, and that the electrical conductivity can be controlled by the level of oxidation [80,81]. For example, this is indicated by the thermal or chemical reduction of GO, leading to a significant increase in conductivity as a result of the restoration of sp² carbon, as present in the graphitic network. The theoretical value of the charge carrier mobility for graphene is 2 × 10⁵ cm²/Vs [82]. Below, in

Table 1. An overview of the electrical conductivity of different graphene-based films/morphologies.

Material	Electrical Conductivity (S/m)	References
Graphite thin film (~3 μm)	6120	[83]
Graphene (~3 μm)	1750	[83]
RGO thin film (1.5 cm2)	4.21 × 10−5	[84]
GO thin film (1.5 cm2)	4.57 × 10−5	[84]
RGO (by nascent hydrogen)	12,530	[85]
RGO (by hydrazine)	2420	[86]
RGO (by aluminium)	2100	[87]
RGO (by HI)	30,400	[87]
RGO (N2H4 and microwave reduction)	1180	[88]

, the electrical conductivity of different graphene-based materials is listed with the corresponding references.

The electrical properties of graphene can be improved by minimizing the role of different defects, which act as scattering centers and can inhibit the charge transport in graphene-based samples. During sp² carbon restoration, the partial removal of oxygen functionalities can result in the generation of defects which may affect the optoelectronic properties. For instance, RGO with more defects shows lower electrical conductivity as compared to an ideal monolayer of defect-free graphene [89]. The lower conductivity can also influence the charge transfer and separation during photocatalysis.

4.1.5. Surface Features

The properties of graphene vary with the functional groups attached to the graphene surface (e.g., GO has different functional groups: carboxyl, carbonyl, hydroxyl and epoxide [9]) as a result of the oxidation or modification/functionalization of graphene. The same is true for the specific surface area and density of defects in graphene sheets [90,91]. For a monolayer graphene, the theoretical value of the specific surface area is 2630 m²/g [92], whereas for monolayer GO, the estimated surface area is 2391 m²/g [93]. Esmaeili et al. [94] reported the specific surface area of GO as ~53 m²/g. These surface features can play a crucial role in the adsorption of targeted adsorbate and in redox reactions by photogenerated active species (e.g., charge carriers and reactive oxygen species) on the graphene materials, thereby improving the photocatalytic activity.

The unique two-dimensional honeycomb-like hexagonal graphene lattice offers a useful platform for forming composites with different nanostructured semiconductors, including quantum dots, nanowires/nanorods, nanosheets and porous structures, thereby favoring the formation of multifunctional graphene-based composite photocatalysts [95].

4.2. The Role of Graphene in Photocatalysis

In a photocatalytic system, graphene derivatives can act as a photocatalyst, a co-catalyst in most cases and/or as a photosensitizer [96], namely, as a photocatalyst because of its suitable bandgap that can absorb light and generate charge carriers to catalyze different species [97] and as a co-catalyst, owing to its unique electrical features and charge transport properties, thus allowing it to effectively sink/capture electrons from the semiconductor and to reduce photogenerated charge recombination [98]. When acting as a sensitizer, graphene and its derivatives are the main species that absorb light and initiate a photocatalytic process [98,99]. Below, we have further described the fundamental roles of graphene in photocatalysis.

4.2.1. A Photocatalyst Itself

Pristine GO, exhibiting a certain level of oxidation, can be also a semiconductor material and therefore can act as a photocatalyst itself [100], and when its bandgap lies in the range of 2.4–4.3 eV, it can be used for photocatalytic hydrogen generation in water [97]. As previously mentioned, an elevated degree of oxidation leads to an increase in the bandgap of GO, and a well-oxidized GO possesses the suitable CB maxima and VB minima for H₂ and O₂ evolution. For example, Yeh et al. [100] evaluated the photocatalytic properties of GO with different oxidation levels. Only the GO with the highest oxidation degree could generate O₂, and other less oxidized GO, though with a sufficient bandgap, could not generate O₂ under light illumination. Please note that this is related to a typical phenomenon where GO undergoes reduction under light illumination, and therefore, the position of the VB minima can change. The GO with the highest oxidation degree was oxidized enough, and therefore, the VB minima remained positive enough for water oxidation under light illumination [100]. GO can act as a metal-free, eco-friendly and economical photocatalyst material; however, it has the key ability of suppressing its photocorrosion either by modifying GO (e.g., UV irradiation of GO causes its reduction, and after a certain time of exposure, the RGO is stabilized [99]) or by some other method.

4.2.2. A Cocatalyst

Graphene in photocatalysis is also very well-known for its role as a co-catalyst, as a substitute to the traditional costly noble metal co-catalysts. For example, in the case of RGO, when used with TiO₂, its higher work function facilitates the charge transfer from TiO₂ to analyte or to generate reactive oxygen species [101]. Under the illumination of a composite based on RGO and TiO₂, this facilitates the transfer of the photogenerated electron from the CB of TiO₂ to molecular oxygen. Such a process is used, for example, in the photo-oxidation of arsenic, and further investigation into the participating radicals by a competitive radical quencher test verified the simultaneous contribution from both the hydroxyl and the superoxide radicals. Overall, the composite’s results were similar to that of the common Pt/TiO₂ photocatalyst used in As(III) photo-oxidation [101].

4.2.3. A Sensitizer

Graphene can also act as a photosensitizer for visible light absorption in semiconductor-based photocatalysis, as it is an ideal support for charge transport in typical photocatalysts [102] (e.g., with ZnO [103], TiO₂ [99] and ZnWO₄ [99]). This occurs when GO is the main species that absorbs light, without affecting the bandgap of wide bandgap semiconductors such as TiO₂. The key aspect lies in its suitable bandgap for visible light activation, which makes it a potential macromolecular photosensitizer in a composite photocatalyst [10,97,100]. This leads to extended light absorption in the visible region, thereby resulting in promoted photocatalytic activity [96,99].

Therefore, it is clear that, in visible-light-driven photocatalysis, GO can also act as a photosensitizer. For example, Nasir et al. [10] illuminated a TiO₂/GO photocatalyst with a visible light source, where the GO acts as a sensitizer (by transfer of electrons to TiO₂ and by direct interaction of VB holes with water to initiate reactive oxygen species), which, in turn, degrades methylene blue (MB). Tismanar et al. [30] evaluated the visible light activity of GO in a thin film of TiO₂ and GO by degrading MB and imidacloprid. By large-scale DFT calculations, Du et al. [104] validated the direct photo-excitation of the upper VB electrons of graphene to TiO₂ CB under visible light illumination, which was also confirmed experimentally by the wavelength-dependent photocurrents of the similar photoanode. Zheng et al. [99] used GO as a photosensitizer in visible light driven hydrogen generation over TiO₂/GO, and the photoexcited electrons in GO were transferred to TiO₂, thereby reducing water to generate H⁺ and then H₂ gas.

In order to better elaborate and understand the role of GO as a cocatalyst or as a photosensitizer, an example of GQDs and TiO₂ heterojunction is included in Figure 3 [98]. In the case of a cocatalyst, the GQDs separate and transport charges and thereby prohibit charge recombination. The main species that absorbs light is TiO₂, from where the photogenerated electrons are injected into the CB of GQDs. The GQDs accept electrons and transfer them to the electrolyte to generate hydrogen gas. As a photosensitizer, the main species that absorbs light and generates photoexcited charge carriers is GQDs. The photogenerated electrons are then transferred to TiO₂ to further generate hydrogen gas [98]. Please note that the reported bandgap of GQDs in this case is 2.26 eV, and therefore, they act as a cocatalyst under UV illumination and as a photosensitizer for visible illumination.

5. Design of Composite Graphene/TiO₂ Photocatalysts

There are various techniques used for the synthesis of graphene–TiO₂ nanocomposites [1,105,106]. In most, either (i) commercial/pre-synthesized TiO₂ is mixed with graphene or its derivatives by sonication [107] or by simple mixing followed by photoreduction [1,108,109], or (ii) a graphene derivative is used as a support to immobilize and grow TiO₂ nanostructures (e.g., the hydrothermal method [110,111], solvothermal method [106,112], sol gel [113,114], electrodeposition [115,116,117] etc.). However, there is no optimal method that is well-known for the synthesis of these composite structures, and the quest for a more suitable, scalable and refined method is still in progress and the focus of researchers. Recently, Nasir et al. [10] introduced a one-step, UV-induced photocatalytic synthesis of TiO₂/GO composites. The composites obtained could be directly used for visible-light-driven photocatalysis. Below, in

Table 2. An overview of the recent methods for the design of graphene-based/TiO₂ composites.

Ti Precursor	Graphene Derivative	Photocatalyst	Synthesis Method	Morphology		Reference
Commercial P25	Pre-synthesized GO	Photoreduced GO/TiO2	Simple mixing, followed by photoreduction		a.	[3]
TiCl4	RGO seeded on TiO2 nanocrystals	Durian-like mischcrystal TiO2/graphene	Seed-induced hydrothermal approach	Uniform growth of durian-like compact clusters of TiO2 on RGO surface		[8]
Commercial TiO2	Sulphonated RGO GO by Hummer’s method	Edge-sulphonated graphene-decorated TiO2	Ultrasonic self-assembly method		b.	[118]
Commercial TiO2 anatase	Commercial GO	TiO2 nanorod array on a 3D GO framework	Hydrothermal	Uniformly dispersed TiO2 nanorods (l = 200 nm, d = 30 nm) anchored on a 3D GO framework		[119]
TiO2 (P-25)	GO by modified Hummer’s method	TiO2/graphene aerogel	Hydrothermal		c.	[120]
Titanium (IV) isopropoxide (TTIP)	GO by Hummer’s method	1 wt%RGO/S0.05N0.1TiO2 nanocomposite	Solvothermal		d.	[121]
TTIP	GO by modified Hummer’s method	RGO-15%TiO2	Hydrothermal		e.	[92]
TiO2 by microwave-assisted synthesis Precursor: TTIP	GO by Tour’s method, amine-modified GO	GO and amine-modified GO/TiO2	Hydrothermal	No SEM shown		[122]
Ti substrate	RGO	Tungsten-doped TiO2/RGO	Plasma electrolytic oxidation		f.	[123]
Grey TiO2 Precursor: TiO2 (P-25)	Graphite	Grey TiO2/GO	UV induced photolysis		g.	[10]
TiO2 nanotubes by electrochemical anodization	GO	Pt coated, GO wrapped TiO2 nanotubes	Electrophoretic deposition		h.	[124]
Commercial TiO2	GO by Hummer’s method	TiO2/RGO	Hydrothermal		i.	[12]
Commercial TiO2	GQDs by hydrothermal method Precursor: SWCNTs GQDs by hydrothermal method Precursor: Carbon fiber	Graphene quantum dots anchored TiO2	Simple mixing		j.	[125]
Precursor: Tetrabutyl titanate (TBOT)	Commercial Graphene	Black TiO2/graphene composites	Sol gel		k.	[126]
TiO2 (P-25)	GO by modified Hummer’s method	TiO2/GO	Hydrothermal		l.	[127]
Precursor: Titanium (IV) butoxide	Commercial GO	TiO2/GO	Sol-gel	Agglomerated and distributed growth of TiO2 nanoparticles over the GO surface		[128]
TiO2 coating by sol-gel dip coating method Precursor: TTIP	Commercial GO suspension	TiO2/GO TiO2/RGO	Dip-coating Chemical reduction of dip-coated TiO2/GO sample		m.	[129]
Ca-doped TiO2 by hydrothermal method Precursor: TBOT	Exfoliated graphene	Ca2+-doped mixed-phase TiO2/graphene	Electrostatic self-assembly process/ Sonication-assisted mixing		n.	[130]
TiO2 Precursor: TTIP	Few-layer graphene by mechanochemical method	Few-layer graphene TiO2	Mechanochemical synthesis		o.	[131]
TiO2 by sol-gel method Precursor: TIIP	GO by modified Hummer’s method	TiO2 supported-RGO	ultrasonic-assisted solvothermal method		p.	[132]

a. Reprinted with permission from [3] Copyright 2022 Elsevier Ltd. b. Reprinted with permission from [118] Copyright 2022 Elsevier Ltd. c. Reprinted with permission from [120] Copyright 2022 Elsevier Ltd. d. Reprinted with permission from [121] Copyright 2022 BioMed Central Ltd. (Part of Springer Nature). e. Reprinted with permission from [92] Copyright 2022 Elsevier Ltd. f. Reprinted with permission from [123] Copyright 2022 Elsevier Ltd. g. Reprinted with permission from [10]. Copyright 2022. h. Reprinted with permission from [124] Copyright 2022 Elsevier Ltd. i. Reprinted with permission from [12] Copyright 2022 Elsevier Ltd. j. Reprinted with permission from [125] Copyright 2022 MDPI. k. Reprinted with permission from [126] MDPI. l. Reprinted with permission from [127] Copyright 2022 Elsevier Ltd. m. Reprinted with permission from [129] Copyright 2022 Elsevier Ltd. n. Reprinted with permission from [130] Copyright 2022 Elsevier Ltd. o. Reprinted with permission from [131] Copyright 2022 Elsevier Ltd. p. Reprinted with permission from [132] Copyright 2022 Elsevier Ltd.

, we have summarized some of the most recent trends in the design of such graphene-based TiO₂-mixed composites, taking into account the starting Ti material, graphene derivatives and resulting photocatalysts in view of the synthesis method used and of the resulting morphology.

6. Characterizing Composite Photocatalysts of Graphene/TiO₂

Of the existing characterization techniques, the ones that are the most used for confirming the presence and for characterization in view of evaluating graphene in TiO₂/graphene photocatalysts are: (i) Raman spectroscopy for the presence of graphene, graphite and the typical structural characterization of TiO₂ [133,134], (ii) Fourier-transform infrared spectroscopy (FTIR) for molecular groups and chemical bonding (only specific molecules can be detected) [135], (iii) X-ray photoelectron spectroscopy (XPS) for chemical and compositional properties of the material’s surface [136] and (iv) scanning electron microscopy (SEM) and transmission electron microscopy (TEM), coupled with elemental mapping, for evaluating morphology (atomic scale structural information and interactions with the support is also possible with TEM coupled with elemental composition) [137].

6.1. Raman Spectroscopy

Raman spectroscopy is a powerful surface sensitive tool used to study the structure of the composites, giving information related to their structural characterization. For carbon-based materials, it stands as a high-resolution tool enabling the characterization of the lattice structure and the electronic, optical and phonon properties [133,134]. The Raman spectra of anatase TiO₂ typically show four peaks at 148, 398, 514 and 641 cm⁻¹, attributed to Eg₍₁₎, B_1g(1), A1_g + B_1g(2) and E_g(2) vibrations, respectively [138]. In the case of graphite or graphene-based materials, Raman spectra provide detailed structural data related to the G and D bands [133,139]. For example, for GO, the usual peaks observed are: (a) for the D band (disorder/sp³ carbon), usually at ~1350 cm⁻¹, and (b) for the G band (sp² carbon), at ~1580 cm⁻¹ [140], and pure graphite does not show a clearly evident D band [141]. Generally, a higher disorder in the lattice results in broadening of the G and D bands, and a higher relative intensity of the D band compared to the G band (evaluated by their intensity ratio, the I_D/I_G ratio). The Raman spectrum of GO shows not only a higher I_D/I_G ratio but also a blue shift in the peak position of the G band compared to that of graphite [141]. Nasir et al. [10] observed the gradual emergence of the D band with an increase in the GO’s oxidation level, hence revealing an interesting fact, that GO with more oxygenated groups results in a higher I_D/I_G ratio. With the increase in the D band’s intensity, additional characteristic disorder-related peaks appear, i.e., D + D′, 2D and D′ for the more oxidized GO [10]. The D + D′ band (also referred as D + G) appears as a bumpy peak at ~2940 cm⁻¹ and can be observed in highly defective polyaromatic structures [142]. The 2D band appears due to physical defects (i.e., edge defects in graphene sheets), and a D^′ peak appears at ~1620 cm⁻¹ and can be related to any type of defect [10,142]. In contrast to TiO₂/graphite composites, for TiO₂/GO, a more intense and broader G band is observed, and this is typically used to confirm the formation of GO [10]. Cham sa-ard et al. [140] evaluated the difference in the Raman spectra of GO and RGO, and namely, the I_D/I_G ratio for GO and RGO was 1.02 and 1.17 due to the loss of oxygenated functional groups. Shahrezaei et al. [138] used Raman spectroscopy to confirm the presence of a modified form of graphene decorated on TiO₂ multileg nanotubes, and the authors verified the presence of the D and G bands in the composite, which were absent for the pristine nanotubes.

More interestingly, Liu et al. [143] used a TiO₂ nanosheet as a support to directly grow graphene, and Raman spectroscopy was used to observe the growth phenomena as a function of the growth time, as shown in Figure 4. The D and G bands, typical characteristics of carbon materials, became more pronounced after 2 h (180 min), as seen in Figure 4a (inset I), where intense D and G bands can be observed together with the appearance of the 2D and D + G bands, as well as with an increase in the D band. A sharp change in the FWHM of the G band was measured at 0.5 h (30 min) as a result of the complete formation of a graphene monolayer, with no further drastic change [143]. Nasir et al. [10] observed a sequential increase in the I_D/I_G ratio as a function of graphite oxidation (directly dependent on the UV illumination time, which led to a gradual increase in graphite oxidation), as indicated in Figure 4b.

The above are the most critical aspects of how graphene-based materials in a metal oxide composite can be identified by Raman spectroscopy and of what to expect with respect to their different oxidation level, particularly at trace concentrations of graphene-based materials. However, the main and initial fingerprint for the presence and investigation of GO or its modified forms is therefore the appearance of the D and G bands and the I_D/I_G ratio.

6.2. FTIR Spectroscopy

Fourier-transform infrared spectroscopy (FTIR) is typically used for the identification of chemical composition or molecular structure, specifically to identify the chemical bonds and characteristic functional groups present in a material or a composite [144,145]. In graphene-based materials, the extent of oxidation and the presence of oxygenated species can be evaluated from its FTIR spectra. FTIR also provides information with respect to the surface reactive sites of a material and their interactions with certain species [146]. FTIR spectroscopy is typically the second preferred characterization technique used to identify the presence of GO in composites [146,147]. Figure 5 shows a very clear overview of typical FTIR spectra for GO and its composites with TiO₂ [148]. A typical IR spectrum of GO shows a broad band between 3600–3200 cm⁻¹ along with a band at 1400–1200 cm⁻¹ corresponding to –OH groups, at ~1700 cm⁻¹ for carbonyl groups, 1600–1400 cm⁻¹ for the C=C of graphene, 1100–1000 cm⁻¹ for alkoxy groups and ~950 cm⁻¹ for epoxy groups [9]. For TiO₂, bands attributed to –OH groups are present, along with Ti–O/Ti–O–Ti peaks in the ~900–400 cm⁻¹ region [149]. In the case of the GO/TiO₂ composites, overlapped spectra are observed. The main peaks taken into account for the identification of such composite are: (i) –OH groups (3200–3600 cm⁻¹, 1400–1200 cm⁻¹), this is usually observed in pristine anatase and (ii) a Ti ̶ O ̶ C peak, with a blue shift in the wavenumber compared to the Ti–O/Ti–O–Ti peak [148,150]. In TiO₂-RGO composites, the loss of functional groups in the RGO results in less peaks in the spectra compared to the GO composite [148,150]. It is important to note that, in most composites with titanium dioxide, the amount of GO is quite low, and therefore, in the corresponding IR spectra, the functional groups of GO are retained with a significant decrease in peak intensity [148,151]. Sim et al. [152] observed the loss of the C–O peak at 1220 cm⁻¹, ascribed to the interaction of the epoxide/phenolic groups of GO with ̶ OH groups in the Ag/TiO₂ nanotubes, thereby forming Ti–O–C bonds in their composites with GO (the absorption peak for Ti–O–C in combination with Ti–O–Ti vibration appeared at 800 cm⁻¹).

6.3. XPS Spectroscopy

X-ray photoelectron spectroscopy (XPS) is one of the most used characterization techniques, establishing the chemical and compositional properties of surfaces, with a high surface sensitivity. XPS is also widely used for the evaluation of graphene-based materials [153,154] or various graphene-based composites with semiconductor oxides [155,156]. Moreover, up to now, the adventitious carbon peak is typically used for spectra calibration, and more recent reports have shown the drawbacks of such an approach and how to further proceed in order to allow for reliable determination of the chemical states [136,157,158], as well as for accurate peak fitting [157,158].

Typically, the C1s peak of graphene-based materials (graphite, graphene and graphene oxide) consists of the peaks discussed below: (i) for graphite, an sp² peak at 284.1–284.8 eV (showing an asymmetric peak shape) [158,159,160,161], C-OH groups at 285.6 eV and C-O groups at 286.6 eV due to atmospheric oxidation [141], as well as a shake-up feature related to the pi to pi* transition observed around 290.5–291 eV [158,161,162]; and (ii) for graphene/graphene oxide, an sp³ peak at 284.6–285.1 eV (typically 1 eV higher than the sp², [113] C-O at 286.4–286.6 eV (epoxide groups), C=O at 288.2 eV and COOH groups at 289.2 eV [158,163,164]. For this, there is also the presence of adventitious carbon at 284.8 eV.

When evaluating graphene-based composites containing TiO₂, spectra evaluation can also include the Ti2p peak, and when changes occur to the structure of the graphene-based materials, i.e., to the C1s peak position and shape, peak calibration can be performed with the former. In this respect, Nasir et al. [10] obtained GO by the UV illumination of grey TiO₂/graphite nanocomposites, and XPS confirmed the conversion of graphite to GO. Figure 6a–c show the high resolution C1s, Ti2p and O1s peaks of pure graphite powder and of the 0, 4, 8, 24 and 120 h UV illuminated grey TiO₂/graphite composites. The presented XPS peaks clearly validate the changes occurring in the C1s peak (switch from graphite to GO), as well as the decrease in the surface coverage of GO with grey TiO₂ nanoparticles, with increasing the UV illumination time (with more illumination time, an increase in the intensity of the Ti2p and O1s peak is observed, together with a decrease in the C1s peak intensity) [10]. The peak fitting of the pure graphite was in line with data from the literature [158,159,160]. The 24 h UV-illuminated nanocomposite (Figure 6d) confirmed an sp³ contribution at 285.2 eV due to oxidized graphite (together with peaks attributed to C-O at 286.4 eV, for epoxide groups C=O at 288.16 eV and COOH groups at 289.2 eV) [10].

Similarly, Kumari et al. [156] evaluated the chemical state variations of GO/TiO₂ nanocomposites, and in addition to confirming the presence of TiO₂ (Ti2p_3/2 at 458.8 eV), the authors performed detailed investigations of the O1s and C1s peaks. Namely, the peak fittings for the O1s and C1s were confirmed in the case of the O1s peak for the C=O and C-OH functional groups at 531.4 eV and 533.6 eV, respectively, [156,165], and for the C1s peak, there were weaker residual oxygen-containing functional group peaks compared to C=C (confirming the partial reduction of GO during composite formation) [156,166].

Rajender et al. [167] evaluated the changes in the mixed hybrid of TiO₂ and GO quantum dots (QDs) by investigating the C1s and O1s peaks. The fitted peaks (Figure 6e–h) showed the relocation of oxygenated groups after the formation of the hybrid, i.e., the C-O (ether) and COOH functional groups were significantly reduced in the hybrid sample, meaning that some of the oxygen-related functional groups may have been converted to in-plane epoxy groups to facilitate the possible formation of C-O-Ti bonds in the TiO₂-GQD hybrid, observed also in the O1s peak fitting in the peak at 531.7 eV.

Figure 6. High-resolution XPS analysis: (a–c) raw data of Ti2p, C1s and O1s peak of the TiO₂-GO composite (increases in time show oxidation level of GO), (d) C1s spectra of 24 h UV-illuminated composites [10] (Reprinted with permission from [10] Copyright 2022 John Wiley & Sons, Inc.), Composites of GO quantum dots (gQD) and TiO₂ (T16) with the corresponding peak fittings for the (e,f) C1s spectra (GQDs and the T16/GQD composite) and the (g,h) O1s spectra (T16 and T16/GQD) [167] (Reprinted with permission from [167] Copyright 2018 Elsevier Ltd.).

Figure 6. High-resolution XPS analysis: (a–c) raw data of Ti2p, C1s and O1s peak of the TiO₂-GO composite (increases in time show oxidation level of GO), (d) C1s spectra of 24 h UV-illuminated composites [10] (Reprinted with permission from [10] Copyright 2022 John Wiley & Sons, Inc.), Composites of GO quantum dots (gQD) and TiO₂ (T16) with the corresponding peak fittings for the (e,f) C1s spectra (GQDs and the T16/GQD composite) and the (g,h) O1s spectra (T16 and T16/GQD) [167] (Reprinted with permission from [167] Copyright 2018 Elsevier Ltd.).

6.4. SEM and TEM

Scanning electron microscopy (SEM) is extensively used for evaluating the nanostructure and morphology of various materials, both in top-views or in cross-sections, whereas transmission electron microscopy (TEM) enables the detailed evaluation of atomic-scale structural information [137].

For example, SEM is generally used to study and confirm the presence of graphene sheets in composites. The morphology of TiO₂ can vary from 0D nanoparticles, to 1D-nanotubes/nanowires, to 2D-nanosheets. Note that pristine graphite has a thick, layered and non-transparent structure. At a magnification of 5–50 k, many stacks can be found with a much bigger particle size of graphite flakes, whereas GO has a smaller particle size, and only few stacks can be found at 40–300 k (see Figure 7). This is because GO is produced by oxidation, and thereby the exfoliation of graphite structure leads to isolated mono- or multilayer sheets of graphene [168].

The GO can, therefore, be characterized by crumpled, silk-veil-like, thin nanosheets present with the TiO₂ nanostructures. In the case of composites with TiO₂ nanoparticles, well-defined, mono or multilayer sheets of GO decorated with TiO₂ nanoparticles and agglomerates can be clearly observed [169], as seen in Figure 8a. Trapalis et al. [79] observed 2–6 layered sheets of graphene in the TiO₂ nanoparticle/graphene composites (the graphene content was only 1%). When combined with spaced TiO₂ nanotubes [138] (showing distinct spacing in between the tubes compared to the classical close-packed nanotubes [170]), the graphene sheets were found on the top surface and in between the nanotubes as well, as seen in Figure 8b. In Figure 8c,d, SEM images of TiO₂/RGO with 5% RGO concentration have been included together with the as-obtained elemental analysis, confirming the presence of carbon [171].

TEM is an essential tool in the analysis of graphene-based nanosheets for more detailed morphology insight into the nanosheets and for the confirmation of the GO’s nanoscale features. For instance, in Figure 9, very translucent, wrinkled nanosheets of GO can be clearly observed, and the inset in Figure 9b shows that the GO’s sheet thickness is in the nanometer range. Well-dispersed TiO₂ nanorods can be seen on the GO sheet in Figure 9b [172]. The lattice spacing of GO can also be estimated by TEM; for example, Nasir et al. [10] obtained a lattice spacing of 0.35 nm (002) for GO and 0.35 nm (101) for the TiO₂ region.

6.5. UV-Vis Spectroscopy

UV-Vis analysis is a simple analytical method for quantitative analysis and for evaluating the molecular level mechanisms in a chemical compound/material, and it can also estimate band gaps and evaluate the optical properties of the materials [173]. UV-Vis spectroscopy does not give a direct mark for the presence of graphene or GO; however, it proves useful in evaluating the extended range of light absorption in the visible region, which is a usual characteristic of graphene-based semiconductor composites. The typical characteristic features observed for identifying GO are: (i) a shoulder attributed to n−π* plasmon peak, which usually appears at ∼310 nm [174] and (ii) a π−π* plasmon peak between ∼230–300 nm for graphene, or at ∼230 nm for well-oxidized graphene [175], [174]. In graphene-based materials, the π−π* absorption peak corresponds to the sp²/sp³ character. The amount of shift in the position of this peak towards the visible range corresponds to the increase in the sp² character, thereby increasing the growth of the number of sp² layers [175].

Figure 10a includes the UV–Vis spectra of GO and RGO, and the main absorption edge at 226 nm was shifted to 262 nm after reduction. This red shift suggests the restoring of the electronic conjugation in the graphene sheets after reduction [176]. Mixed composites show a red shift towards longer wavelengths and an increase in the visible range absorption compared to pristine TiO₂, as seen in Figure 10b, where TGPPC_x stands for the TiO₂ graphene photocatalyst and x stands for the content of graphene [175]. A modified form of graphene (cyano graphene, cyano platinized graphene), when decorated on the TiO₂ nanotubes, showed a red-shift (ca. 10–20 nm) in the absorption edge and relatively more absorption in the visible range of light than the bare TiO₂ nanotubes [138]. Zhang et al. [177] observed a 30 nm red shift in the absorption edge of a layer-by-layer composite of TiO₂ and graphene. In this case, a strong visible range absorption was also observed, indicating high photocatalytic activity in the UV and visible range. Similarly, Rong et al. [175] observed a red shift in the TiO₂ absorption spectrum, as a function of the graphene content with a decrease in the band gap, especially at a higher graphene content.

7. TiO₂—Graphene Composites: Current Status and Applications

TiO₂ photocatalysis has made several breakthroughs in various practical and industrial applications, primarily in the field of environmental safety (e.g., wastewater treatment, air purification, microorganism inactivation/sterilization, etc.). As previously mentioned, an ideal photocatalyst represents outstanding features such as economical features, environmentally benign features, long-term stability, reproducibility, recyclability and efficient photocatalytic activity [178]. Despite several advantages of TiO₂ photocatalysts, there are some limitations, such as (a) recyclability, which is relevant to separation and recovery, (b) narrow regions of light harvesting due to its wide band gap, (c) faster recombination of photogenerated carriers, (d) low surface area, (e) less adsorption capacity and (f) aggregation issues [179]. Countermeasures usually include, for example, the modification, optimization, design and development of a TiO₂ photocatalyst to improve photocatalytic efficiency, stability, reproducibility and recyclability [126,180,181].

Graphene-based TiO₂ composites have emerged as efficient photocatalysts [1,6,182,183,184,185,186] for the next generation photocatalytic applications [1,105,187,188,189,190,191,192]. The synergistic effect between components in ternary photocatalysts (comprising three components, where each photocatalyst component plays its potential role to maximize the photocatalytic activity) offers significant merits, such as multiple interfaces, improved separation and migration/transfer of charges, reduced recombination of photogenerated charge carriers, prevented agglomeration, high specific area and high stability [193]. Recent progress in the development of TiO₂/graphene-based photocatalysts in various applications, such as wastewater treatment for the removal of pollutants, air purification, water splitting for hydrogen production, dye-sensitized solar cells (DSSCs), photoconversion of CO₂ into renewable fuels, NO_x fixation and deactivation of microorganisms (see also Figure 11), is further discussed in this section.

7.1. Remediation of Water

Typical water contaminants include sewage industrial effluent (e.g., textiles, paper, pharmaceutical industry, etc.) and domestic contaminants (e.g., detergent, pharmaceuticals, pesticides, etc.) [194]. The incorporation of graphene in TiO₂ results in a good adsorbent for the removal of organic pollutants (e.g., dyes, phenolic compounds, chlorinated products, antibiotics, pesticides and crude oil) and microbial inactivation due to electrostatic attraction, π−π interaction, visible light absorption and larger electron mobility [195,196,197,198,199,200]. TiO₂/graphene-based composites have revealed their potential as cost effective, environmentally friendly and sustainable photocatalysts in support of wastewater treatment technology for the degradation of organic pollutants and environmental remediation [200].

An example of such a wastewater treatment application is for xanthate degradation (xahnthane is an environmental pollutant resulting from the mineral industry, i.e., from the mineral flotation process, and is toxic for biota [201]). The photodegradation efficiency of an 18% TiO₂/G composite under simulated light irradiation was reported as 97.03% after 100 min, much higher than for pure TiO₂ (17.88%). This is attributed to an extended photoresponse in the composite material from 200–380 nm to 200–800 nm [127]. Similarly, high photocatalytic degradation was observed for TiO₂/GO composites for picric acid (also known as 2,4,6-trinitrophenol, which is a dangerous nitroaromatic environmental pollutant generated from dye, leather and chemical industries, having low biodegradability and high toxicity [202]) [12]. RGO-nano TiO₂ composite (RGO-nTiO₂) showed 88% degradation efficiency for metachrome yellow dye (or alizarin yellow GG, an azo dye which is textile effluent, highly toxic and carcinogenic [203,204]) under natural sunlight with 80% efficiency, higher than for that of plain TiO₂. The enhanced efficiency was attributed to the superoxide anion radical that played the key role in pollutant oxidation [20]. The Z-scheme photocatalytic composite g-C₃N₄/ RGO/TiO₂ showed an efficient performance for ammonia–nitrogen removal from water. Ammonia–nitrogen is a toxic groundwater pollutant, and its impact is widespread, as its excessive quantity causes malodor in river water that can lead to eutrophication, weakening of the self-depuration capability of water and threats to public health. Heavily polluted waters contain an ammonia–nitrogen level of >15 mg/L, and a minor pollution level is considered to be 8–15 mg/L [205,206]). The photogenerated electrons from the TiO₂ conduction band can transfer to the g-C₃N₄ valence band by the Z-scheme mechanism. These electrons recombine in g-C₃N₄ with photogenerated holes. The g-C₃N₄/RGO/TiO₂ Z-scheme system shows improved redox capacities for NH₃–N oxidation and NO³⁻ reduction (Figure 12) [207].

The CGTC (graphene-oxide-bridged chitin-modified TiO₂/carbon fiber) composite prepared under hydrothermal conditions by chitin-modified and GO-bridged TiO₂/carbon fibers showed 97% photocatalytic degradation of Rhodamine B in 60 min [208]. Yanwen et al. [209] reported the ternary composite N–TiO₂/GO/PP fiber sheet (24.6 cm × 2.7 cm) fabricated by the immobilization of N-doped TiO₂ and GO on polypropylene (PP). Photocatalytic degradation for 9 h under simulated sunlight of 10 mg L⁻¹ roxithromycin (roxithromycin is an antibiotic pollutant from domestic and industrial effluent; its low concentrations may cause high ecological risk after long term exposure [210]) could reach up to 90%, with a degradation rate constant of 0.2299 h⁻¹ in surface water under alkaline conditions (pH 8–9).

7.2. Photoconversion of CO₂

Carbon dioxide (CO₂) emission is a major source of global warming. The photoconversion of CO₂ into useful chemicals, such as CH₃OH, CH₄, CO, etc., is an attractive approach to alleviate the issues of environmental pollution and climate change [211,212,213,214]. Kamal et al. [215] reported the synthesis of plasmonic gold nanoparticles photodeposited on TiO₂ and decorated on N-doped graphene (ANGT_x) to form a heterostructure. The composite photocatalyst showed enhanced photoconversion of CO₂ to methane with high selectivity. The optimized composite delivered the highest electron consumption rate (R_electron) of 742.39 μmol g⁻¹h⁻¹ in 4 h for the reduced products. The outstanding performance of such a photocatalyst was attributed to the synergistic effect between the components [215]. The improved light absorption, higher CO₂ adsorption, faster charge transfer kinetics and lower recombination rate originated from interfacial kinetics between components, and a similar enhancement was also reported for Ni/GO–TiO₂ composites [128].

7.3. Air Purification

Air pollution levels have become increasingly alarming, and outdoor and indoor air purification is becoming indispensable in different urban areas of the world. Air pollution includes organic and inorganic chemical contaminants such as volatile organic compounds (VOCs), carbon monoxide (CO), nitrogen-containing compounds (NO_x), sulfur-containing compounds (SO_x) and pathogens such as bacteria, viruses and fungi [216].

Photocatalytic air purification is the most proficient technology that facilitates the formation of reactive oxygen species (ROS) to decompose air pollutants. Photocatalytic technology offers the complete degradation of air pollutants into nontoxic final products under ambient conditions. It does not need any chemical or external energy source and operates in safe conditions [217]. The utilization of solar light under ambient conditions makes it an ideal, feasible and preferable technique [218]. TiO₂-based photocatalysts facilitate the adsorption of environmental pollutants, owing to their favorable photocatalytic properties applicable to self-cleaning, air-purifying and antimicrobial applications [219,220]. The mixing of TiO₂ with construction materials such as cement could lower the air pollutant concentrations [221]. TiO₂-based photocatalysts can also be applied as coatings on the interiors of buildings to improve air quality [222].

This photocatalytic approach is also effective for the inactivation of airborne microbial pathogens such as Anthrax and against the pandemic of the Coronavirus family [223,224]. Recently, an Ag/TiO₂/RGO ternary composite was also investigated as an anti-bacterial self-cleaning photocatalyst [225]. Additionally, disinfection by titania-based photocatalysts is a relatively inexpensive, non-toxic and safe method.

7.4. Water Splitting

TiO₂-based materials are considered a promising solar-light-driven photocatalyst for water splitting [40]. Edge-sulfonated graphene (RGO-SO₃H) was synthesized through a diazotization reaction coupled with TiO₂ nanoparticles to form an RGO-SO₃H/TiO₂ photocatalyst for a hydrogen evolution reaction (HER). The composite showed a hydrogen production rate of 197.1 mmol h⁻¹g⁻¹, which was higher than that for TiO₂ (5.38 times), RGO/TiO₂ (2.81 times) and SO₃H/TiO₂ (3.40 times). The improved performance of the RGO-SO₃H/TiO₂ photocatalyst is to be attributed to the synergetic effect of covalently functionalized graphene (photoelectron cocatalyst) and sulfonate ions (H⁺-adsorbed active sites) [118].

Moreover, GO sheets anchored onto Fe₂O₃−TiO₂ and V₂O₅−TiO₂ particles exhibited 398.18 and 373.01 μmol h⁻¹ HER rates under solar radiation. The enhanced HER rates were attributed to higher surface areas provided by GO, the lower recombination of photogenerated carriers and faster electron transfer rates [226]. Higher photoelectrochemical activity was also obtained for GO/In₂S₃/TiO₂ nanorods used as photocatalysts for photoelectrochemical water splitting [227]. The improved performance was attributed to the larger surface area, visible light absorption and improved carrier transport because of additional defect states [92,228,229,230].

Put succinctly, binary [231] and ternary [232] graphene/TiO₂-based composites offer affordable photocatalysts with feasibility in processing, responses to visible light, promising photocatalytic activity and lower recombination rates of photogenerated charge carriers. Photostability and photocatalytic performance can be enhanced by heterojunction formation and Z-scheme photocatalysts.

8. Conclusions and Future Outlook

To resolve the limitations of TiO₂ in photocatalysis, TiO₂/graphene composites are well-known and have the potential to overcome increasing energy demand and pollution remediation by means of photocatalysis. Graphene offers advantages of (i) photosensitizing TiO₂ for visible light absorption, which results in an extended range of absorption in the solar spectrum and thereby more photocatalytic activity; and (ii) cocatalysts, assisting in charge transport, prohibiting charge recombination and overcoming the sluggish charge transfer kinetics of TiO₂. Typically, the presence of graphene affects electrical conductivity and electronic features, as well as active sites (due to the good surface area of graphene sheets), and, as a result, the overall photocatalytic activity. However, surface features, the crystal structure, the extent of oxidation and the type of derivative of graphene are crucial for this. In this review, the fundamental properties of graphene and their plausible role in photocatalysis have been elaborated. The most critical aspects on how graphene and its derivatives can be identified in composites by different spectroscopic analysis techniques (such as Raman spectroscopy, FTIR spectroscopy and XPS), and what to expect with respect to their different oxidation levels, particularly at trace concentrations of graphene-based materials in a metal oxide composite, have been elaborated, with a possible ”how to” guide. The current state-of-the-art applications of TiO₂/graphene composites in photocatalysis, including wastewater treatment for the removal of pollutants, air purification, water splitting for hydrogen production and the photoconversion of CO₂ into renewable fuels, have been summarized.

Despite a marked increase in the progress of these mixed composites, several questions still need to be addressed. For instance, there is no optimal method of synthesis of these mixed composites, and the same is valid for obtaining large-scale preparations of photocatalysts as well as for the synthesis of graphene with tunable and controlled degrees of oxidation. The recent progress in the design of these photocatalysts has been reviewed, and a need for an optimal and efficient fabrication method to increase the performance of graphene derivatives and their semiconductor composites stands out. In this regard, recently, a new study [10] on the direct obtaining of visible light active photocatalysts with a controlled degree of oxidation of GO has been reported. The method is new and requires further checks for large-scale implementation; however, it opens up an avenue towards the easy and green obtaining of such mixed composites, which is a step forward towards an ecofriendly design approach for GO. The mechanism of interaction of graphene derivatives with TiO₂ is still unclear or rather vague. There is a need for future efforts in order to evaluate the underlying mechanism in promoted photocatalytic activities supported by experimental and theoretical evidence as well (for instance, the positions of valence and conduction bands, the lifetime of photogenerated charge carriers, also considering the extent of graphene oxidation, etc.). The mineralization, or photocorrosion of the mixed photocatalyst also needs to be addressed. Once well-known, these aspects can lead to designs of highly efficient, reliable and very economical graphene/semiconductor composites that can then be applied in different fields of photocatalysis.

At a glance, the progress in TiO₂/graphene-based photocatalysis has been remarkable. If the limitations are addressed step by step, undoubtedly, such composites will be very useful in photocatalysis, and considerable breakthroughs can be expected in future.