*Review* **Compositing Two-Dimensional Materials with TiO2 for Photocatalysis**

**Yu Ren 1,2, Yuze Dong 1,2, Yaqing Feng 1,2,\* and Jialiang Xu 1,3,\***


Received: 12 November 2018; Accepted: 23 November 2018; Published: 28 November 2018

**Abstract:** Energy shortage and environmental pollution problems boost in recent years. Photocatalytic technology is one of the most effective ways to produce clean energy—hydrogen and degrade pollutants under moderate conditions and thus attracts considerable attentions. TiO2 is considered one of the best photocatalysts because of its well-behaved photo-corrosion resistance and catalytic activity. However, the traditional TiO2 photocatalyst suffers from limitations of ineffective use of sunlight and rapid carrier recombination rate, which severely suppress its applications in photocatalysis. Surface modification and hybridization of TiO2 has been developed as an effective method to improve its photocatalysis activity. Due to superior physical and chemical properties such as high surface area, suitable bandgap, structural stability and high charge mobility, two-dimensional (2D) material is an ideal modifier composited with TiO2 to achieve enhanced photocatalysis process. In this review, we summarized the preparation methods of 2D material/TiO2 hybrid and drilled down into the role of 2D materials in photocatalysis activities.

**Keywords:** photocatalysis; 2D materials; TiO2; composite

### **1. Introduction**

With the massive consumption of fossil energy and serious environmental pollution problems, there is an urgent need for clean energy and more efficient ways to decompose pollutants. Photocatalysis is an advanced technology that uses photon energy to convert chemical reactions occurring under harsh conditions into reactions under mild conditions by appropriate photocatalyst, and thus emerged as recognizable fields such as hydrogen generation [1–4], sewage treatment [5–7], harmful gas removal [8,9], organic pollutant degradation [10–13] and carbon dioxide reduction [14–16].

Since the first report that TiO2 electrode was applied for hydrogen production by Fujishima and Honda in 1972 [17], TiO2 has attracted numerous attention in photocatalysis as a typical n-type semiconductor [18–21]. Being non-toxic, inexpensive, highly stable [22–24], TiO2 is widely investigated in photocatalytic fields. Hoffman proposed the following general mechanism (Table 1) for heterogeneous photocatalysis on TiO2 [25].


**Table 1.** Mechanism for heterogeneous photocatalysis on TiO2.

Where >TiOH represents the primary hydrated surface functionality of TiO2, ecb<sup>−</sup> is a conduction band (CB) electron, etr<sup>−</sup> is a trapped conduction band electron, hVB<sup>+</sup> is a valence band (VB) hole, Red is an electron donor, Ox is an electron acceptor, {>TiIVOH}•<sup>+</sup> is the surface-trapped VB hole (i.e., surface-bound hydroxyl radical), and {>TiIIIOH} is the surface-trapped CB electron. Upon light irradiation, electrons transfer from VB to CB of TiO2, while both electrons and holes can be trapped by primary hydrated surface functionality of TiO2, achieving the separation of photo induced electrons and holes. At the same time, the recombination between electrons and holes exits, which competes with charge-carrier trapping process. The competition has thus a negative effect on later interfacial charge transfer. Deliberating on TiO2 photocatalysis process, some drawbacks exit as following: (1) The wide bandgap of TiO2 (3.2 eV) means that photons with adequate energy can only excite electrons in the VB to the CB of TiO2, which limits its effective use of sunlight (UV region, λ ≤ 387 nm); (2) The recombination of excited electrons and holes is inevitable while time for carrier recombination is much shorter than that for charge transfer. Therefore, the effective function of photoexcitation is suppressed greatly.

Considering the above two factors, the improvement of the photocatalytic efficiency of TiO2 can be obtained through two aspects: the improvement of solar light utilization efficiency and the suppression of recombination of electron and hole pairs. In this text, surface modification and hybridization of TiO2 such as noble metal loading [26–29] and semiconductor heterojunction [30–32] are effective methods to enhance the photocatalytic performance. The Schottky barrier formed at the interface between the noble metal material and TiO2 can effectively promote the separation of photogenerated carriers. Similarly, the heterojunction structure can form a matching energy level at the semiconductor interface to suppress the recombination of photogenerated carriers. However, the opportunities of improvements in photocatalysis performances offered by these attempts are narrow, and thus limited their commercial and efficient application. In the past decade, two-dimensional (2D) materials have attracted more and more attention because of the flexible preparation methods, low price and superior physical and chemical properties. In particular, their high surface area, suitable bandgap, structural stability and high charge mobility [33–36] endow these 2D materials with remarkable performances for applications in photocatalysis [37–41]. When combined with TiO2, not only the utilization of sunlight is improved, but also the matching between energy levels is formed to inhibit the recombination, and the large specific surface area provides support and active sites for the reaction. In this review, we summarize the recent advances of 2D material-TiO2 composites, including synthesis methods, properties, and catalytic behaviors. Furthermore, the photocatalytic mechanism is deliberated in detail to elaborate the role of 2D materials in the photocatalytic processes.

### **2. 2D-Material Modified TiO2**

Based on the mobile dimension of electronics, it can be divided into zero-dimensional (0D) materials, one-dimensional (1D) materials, two-dimensional (2D) material and three-dimensional (3D) materials [36], while 2D materials represent an emerging class of materials that possess sheet-like structures with the thickness of only single or a few atom layers [42]. Compared with the bulk structures, the ultrathin 2D structure exhibits superior properties such as modification of energy level

and larger adjustable surface area. The excellent properties of 2D materials make them widely used in many aspects [43–45]. When composited with TiO2, the synergistic effect of the two can significantly improve the photocatalytic activity and thus 2D materials is ideal for TiO2 photocatalysis.

### *2.1. Graphene Modified TiO2*

Since the first isolation by Geim and Novoselov in 2004, graphene has attracted significant attention [46–49]. Graphene is a 2D honeycomb construction consisting of carbon atoms. The thickness of graphene is only 0.335 nm, which is the thickness of a carbon atom layer. In the sp<sup>2</sup> hybrid distribution form, each carbon atom contributes an unbonded π electron, which can delocalize freely throughout the carbon atom 'net' to form an extended π bond. This construction endows graphene excellent properties such as high charge mobility (200,000 cm2 V−<sup>1</sup> s−1), high thermal conductivity (5000 W m−<sup>1</sup> K−1), and large surface area [35], which is ideal for applications in sensors [50], energy conversion and storage [37], polymer composites [51], drug delivery systems [52], and environmental science [53]. When composited with TiO2, graphene can accept photoinduced electrons from TiO2 and thus greatly enhances the efficiency of carriers' separation [54–58].

### 2.1.1. The Synthesis of Graphene/TiO2 Composites

Graphite oxide and graphene oxide (GO) intermediates are widely used in the process of combining graphene with other materials [59]. The most widely used technique is chemical reduction of GO as shown in Figure 1, which is usually conducted by Hummers' method [60]. Graphite is added to a strongly oxidizing solution such as HNO3, KMnO4, and H2SO4 to prepare graphite oxide and the oxygen-containing groups are introduced into the surface or edge of the graphite during the process. The sheets of graphite oxide were exfoliated to obtain GO. The presence of oxygen-containing groups allows GO to provide more surface modification active sites and larger specific surface areas for synthetic graphene-based composites. GO can be converted to reduced graphene oxide (RGO) by chemical reduction to remove these oxygen-containing group. During this process, the number of oxygen-containing groups on the GO decreases drastically, and the conjugated structure of the graphene base will be effectively restored. The presence of oxygen functionalities in GO allows interactions with the cations and provides reactive sites for the nucleation and growth of nanoparticles, which results in the rapid growth of various graphene-based composites. The preparation methods for graphene/TiO2 composites are divided into ex-situ hybridization and in-situ growth, the difference between which is the process of TiO2 formation.

	- *Reduction method.* Usually, in a reduction method, GO and TiO2 metal salts are mixed as precursors. By controlling the hydrolysis of the precursor, TiO2 crystal nucleus grows on GO,

while GO is reduced to obtain graphene-based TiO2 composite materials [65]. In addition to the chemical reduction method, other commonly used reduction methods are photocatalytic reduction [66] and microwave-assisted chemical reduction [67].


**Figure 1.** Preparation of graphene by chemical reduction of graphene oxide synthesized by Hummers' method. Reprinted with permission from [61]. Copyright 2011, Wiley-VCH.

### 2.1.2. The Role of Graphene in TiO2 Photocatalysis

Due to the large bandgap, the photocatalysis process of pure TiO2 can only be activated under UV light. Thus, the hybridization of graphene and TiO2 is essential to ensure a broad light stimulation process. In graphene/TiO2 system, electrons flow from TiO2 to graphene through interface because of the higher Fermi level of TiO2. Then graphene gains excess negative charges while TiO2 has positive charges, leading to a space charge layer at the interface which is regarded as Schottky junction. The Schottky junction can serve as an electron trap to efficiently capture the photoinduced electrons [73] and thus enhance the photocatalysis activity. Meanwhile, the Schottky barrier also acts as the main obstruction for the electron transport from the graphene to TiO2. Under visible light, electrons on Fermi level of graphene are irradiated and the Schottky barrier has to overcome to ensure the injection of electrons to conduct band of TiO2. In the UV light irradiation process, graphene plays a role as electron acceptor and thus promotes the separation of electron-hole pairs [54] (Figure 2).

Different interface interactions have been extensively studied [55,56]. Compared with 0D-2D Degussa P25 (TiO2)/graphene and 1D-2D TiO2 nanotube/graphene composites, the 2D-2D TiO2 nanosheet/graphene hybrid demonstrates higher photocatalytic activity toward the degradation of rhodamine B and 2,4-dichlorophenol under the UV irradiation [56]. The intimate and uniform contact between the two sheets-like nanomaterials allowed for the rapid injection of photogenerated electrons from the excited TiO2 into graphene across the 2D-2D interface while achieving effective electron-hole pair separation and promoted radical's generation. In another example of RGO–TiO2 hybrid, by having a narrower bandgap, the photo-response range of RGO–TiO2 nanocomposites clearly extended from UV (~390 nm) to visible light (~480 nm), which offered a better utilization of visible light [55]. Raman spectra and other characterization revealed that the narrow bandgap was attributed to the Ti–O–C bond between the two components, and thus caught an intimate interaction between TiO2 nanoparticles and RGO sheets. What's more, the up-conversion photoluminescence (UCPL) effect of RGO assists the light absorption, and enabled the efficient utilization of both UV light and visible light (Figure 3). It is worth to note that the surface area of RGO–TiO2 was smaller than that of pure TiO2 (P25), which revealed that the enhanced photocatalytic activity of RGO-TiO2 was relevant to the improved conductivity and bandgap structure other than their surface area. RGO nanosheet can play a role in both charge transfer and active sites after doping with heteroatoms. TiO2/nitrogen (N) doped reduced graphene oxide (TiO2/NRGO) nanocomposites was applied to photoreduction of CO2 with H2O vapor in the gas-phase under the irradiation of a Xe lamp (the wavelength range of 250–400 nm) [57]. Compared with TiO2, TiO2/NRGO composites exhibited a narrower bandgap due to chemical bonding between TiO2 and the specific sites of N-doped graphene. In the photoreduction of carbon dioxide, the function of nitrogen atoms varied in different chemical environments. The pyridinic-N and pyrrolic-N worked as active sites for CO2 capture and activation while quaternary-N worked as an electron-mobility activation region for the effective transfer of photogenerated electrons from the CB of the TiO2 [57] (Figure 4). The results reveal that the doped atoms can act as basic sites for anchoring target molecular, adjusting the electronic properties and local surface reactivity of graphene.

**Figure 2.** Photocatalytic mechanisms of graphene-TiO2 composite under (**a**) visible light (**b**) UV light. Reprinted with permission from [54]. Copyright 2013, Elsevier.

**Figure 3.** (**a**) Schematic of up-conversion photoluminescence (UCPL) mechanism for reduced graphene oxide (RGO)–TiO2 nanocomposite under visible light (hν∼2.6 eV) irradiation; (**b**,**c**) Schematics of proposed mechanism of Rh. B photodegradation. Reprinted with permission from [55]. Copyright 2017, Springer.

**Figure 4.** Reaction mechanisms for photoreduction of CO2 with H2O over TiO2/NRGO-300 samples. Reprinted with permission from [57]. Copyright 2017, Elsevier. NRGO: nitrogen doped reduced graphene oxide.

Except for dimension factor and bonding interaction between graphene and TiO2, a linkage is introduced to graphene/TiO2 system to achieve better interfacial contact as well. A N-doping Graphene-TiO2 composite nano-capsule for gaseous HCHO degradation was reported [58]. It indicated that wrapping with dopamine on the surface of TiO2 enhanced interfacial contact between TiO2 and melamine-doped graphene (MG) sheets, thus promoting the separation and mobility of photoinduced electrons and holes in TiO2@MG-D. The dopamine acted as bridge between TiO2 and MG, creating numerous migration channels for charges and restraining the recombination of electrons and holes (Figure 5). The introduction of linkage can effectively improve the weak interfacial contact and overcome the long distance of electron transport between the graphene and TiO2, leading to raised separation and mobility of photoinduced electrons and holes and thus higher photocatalytic activity.

**Figure 5.** Schematic illustrations for dopamine bridged Melamine-Graphene/TiO2 nanocapsule and photocatalytic degradation process of HCHO. Reprinted with permission from [58]. Copyright 2018, Elsevier.

Despite of electron accepter and electron storage, graphene can also act as a transport bridge between photocatalysts. For example, in the 2D ternary BiVO4/graphene oxide (GO)/TiO2 system, both the BiVO4 and the TiO2 were connected to GO forming a p-n heterogeneous structure. The CB of BiVO4 was more negative than that of GO and the CB of GO was more negative than that of TiO2; thus, the electrons generated from the CB of BiVO4 can transfer to the GO and then the electron further moved to the conduction band of TiO2 (Figure 6). Therefore, the GO can enhance the effective separation of the photo-generated electron-hole pairs due to its superior electrical conductivity. Meanwhile, the large surface area of the GO is also beneficial for dye attachment [74].

**Figure 6.** Photodegradation mechanism of BiVO4/TiO2/GO photocatalyst. Reprinted with permission from [74]. Copyright 2017, Elsevier.

### *2.2. Graphdiyne Modified TiO2*

Graphdiyne (GD) is a new carbon allotrope in which the benzene rings are conjugated by 1,3-diyne bonds to form a 2D planar network structure and features both sp and sp2 carbon atoms. Since the successful synthesis by Li et al. [75], GD has evoked significant interest in various scientific fields because of unique mechanical, chemical and electrical properties [38,42,76–80]. GD shows potential for photocatalysis with its large surface area as well as high charge mobility. GD features an intrinsic bandgap and exhibits semiconducting property with a measured conductivity of 2.516 × <sup>10</sup>−<sup>4</sup> <sup>S</sup>·m−<sup>1</sup> and was predicted to be the most stable structure among various diacetylenic non-natural carbon allotropes [81]. It also provides highly active sites for catalysis. Furthermore, GD with diacetylene linkage can be chemically bonded with TiO2 [82–85]. Therefore, the TiO2-graphdiyne composites can greatly improve the photocatalytic activity, and thus their application in photocatalysis has been explored recently [83,84,86].

### 2.2.1. The Synthesis of GD/TiO2 Composites

The general preparation of GD film is through a coupling reaction in which hexaethynylbenzene (HEB) acts as precursor and copper foil serves as catalysis. Meanwhile, the copper foil provides a large planar substrate for the directional polymerization growth of the GD film (Figure 7). Despite of film, GD with different morphologies such as nanotube arrays, nanowires, nanowalls and nanosheets have been also prepared for diverse applications [87,88].

**Figure 7.** Preparation of graphdiyne (GD) film.

Ex-situ hydrothermal method is commonly used in preparation of GD/TiO2 composites [83,84,86]. In general, the GD and TiO2 are prepared separately. Then the pre-prepared GD and TiO2 are mixed in H2O/CH3OH solvent. After stirring to obtain a homogeneous suspension, the suspension is placed in Teflon sealed autoclave and heated to combine the TiO2 and GD. Being rinsed and dried, the GD/TiO2 composites are obtained.

### 2.2.2. The Role of GD in TiO2 Photocatalysis

Wang et al. [84] were the first to combine GD with TiO2 for the enhancement of TiO2 photocatalysis. The resultant GD-P25 composites exhibited higher visible light photocatalytic activity than those of the bare P25, P25-CNT (titania-carbon nanotube), and P25-GR (graphene) materials. By changing the weight percent of GD in the hybrid, the photocatalytic activity of P25-GD can be adjusted. It was speculated that the formation of chemical bonds between P25 and GD can effectively decrease the bandgap of P25 and extended its absorbable light range [84]. Namely, electrons in VB of TiO2 can easily migrate to impurity band which is attributed to the insertion of carbon *p*-orbitals into the TiO2 bandgap, and then transfer to CB of TiO2 thus enhancing the photo-response activity. In order to further explore the role of GD, Yang et al. [83] investigated the chemical structures and electronic properties of TiO2-GD and TiO2-GR composites employing first-principles density functional theory (DFT) calculations. The results revealed that for the TiO2 (001)-GR composite, O and atop C atoms could form C–O σ bond, which acted as a charge transfer bridge at the interface between TiO2 and GR. Besides the C–O σ bond, another Ti-C π bond is also formed in TiO2 (001)-GD composite, which makes GD combine with TiO2 tightly and therefore enhances the charge transfer. In addition, calculated Mulliken charge for the surface of TiO2 (001)-GD and TiO2 (001)-GR suggested a stronger electrons' capture ability of former (Figure 8). The calculated results were in accordance with theoretical prediction that TiO2 (001)-GD composites showed the highest photocatalysis performance among 2D carbon-based TiO2 composites, confirming that GD could become a promising competitor in the field of photocatalysis. After that, Dong et al. prepared GD-hybridized nitrogen-doped TiO2 nanosheets with exposed (001) facets (GD-NTNS) [86]. The doped N and incorporated GD efficiently narrowed the bandgap compared with pure TiO2 and widened response range towards light from UV light to 420 nm visible light. The activity of the GD-NTNS photocatalyst presented the most superior performance compared with bare TiO2 nanosheets (TNS) and nitrogen-doped TiO2 nanosheets (NTNS) and GR-NTNS.

**Figure 8.** Plots of electron density difference at the composites interfaces: (**a**) TiO2 (001)-GD; (**b**) TiO2 (001)-GR; (**c**) Mulliken charge of GD or GR (graphene) surface in the composites. Reprinted with permission from [83]. Copyright 2013, American Chemical Society.

The mechanisms of photocatalysis enhancement by introducing GD remain to be understood. In general, with a lower Fermi level than the conduction band minimum of TiO2, GD can be regarded as an electron pool which accept electrons excited from TiO2 [84,89,90] (Figure 9). As a result, it prompts the charge carriers' separation and prevents electron-hole recombination. Moreover, GD can generate an impurity band and thus broaden the visible light absorption in TiO2-GD composites [91–93].

**Figure 9.** Schematic illustration for the possible mechanism of the visible light-driven photocatalytic degradation for the GD-NTNS composites. Reprinted with permission from [86]. Copyright 2018, Springer.

### *2.3. C3N4 Modified TiO2*

Graphitic carbon nitride (g-C3N4) is a 2D polymer material which shows broad application prospects in many fields, given the simple synthesis, rich source, along with unique electronic structure, good thermal stability and chemical stability. Its graphene-like structure is composed of triazine (C3N3) or tri-s-striazine (C6N7) allotropes units (Figure 10). The tri-s-striazine unit structure is more stable and thus draws in extensive studies [34]. Since the first report of g-C3N4 for water decomposition, g-C3N4 has attracted wide attention in photocatalyst [40]. The bandgap of g-C3N4 (2.6–2.7 eV) is moderate and the substantial nitrogen sites and ordered units structure endue g-C3N4 an ideal material to composite with TiO2.

**Figure 10.** Triazine (**a**) and tri-s-striazine (**b**) allotropes units of g-C3N4; (**c**) The synthesis of g-C3N4.

### 2.3.1. The Synthesis of g-C3N4/TiO2 Composites

In general, the synthesis of g-C3N4/TiO2 composites can be also divided into ex-situ method and in-situ method.


**Figure 11.** Vapor deposition process in the preparation of g-C3N4/TiO2 composite. Reprinted with permission from [104]. Copyright 2018, Elsevier.

### 2.3.2. The Role of g-C3N4 in Photocatalysis

With a moderate bandgap of ~2.7 eV, g-C3N4 shows ability of photocatalyst under visible light, in contrast to TiO2, which owns a large bandgap of 3.2 eV (Figure 12). However, because of the rapid recombination of photogenerated electron-hole pairs, the synergistic effect between g-C3N4 and TiO2 plays important roles. In a photocatalyst system of g-C3N4/TiO2 composites, the CB electrons of g-C3N4 transfer to the CB of TiO2 and the VB holes of TiO2 transfer to the VB of g-C3N4, which is a typical Type II system [41]. The electron/hole conduction mechanism can effectively separate electrons and holes, and thus enhances the separation efficiency and inhibit the recombination.

**Figure 12.** Bandgaps of TiO2, monolayer g-C3N4 and bulk g-C3N4.

The structure plays a vital role in enhancing photocatalysis efficiency. g-C3N4 nanosheets (NS)-TiO2 mesocrystals (TMC) composites was prepared by in-situ process [105]. Compared with bulk g-C3N4/TMC composites, the H2 evolution rate of g-C3N4 (NS)/TMC was about six times higher, which was possibly due to a larger surface area of g-C3N4 (NS)/TMC (57.4 m2g−1) than that of bulk g-C3N4/TMC (34.3 m2g−1). What's more, the g-C3N4 nanosheets owned a lower surface defect density, given the surface defects normally is seen as recombination centers for photoinduced electrons and holes. However, surface area is not the unparalleled factor of promoted efficiency of photocatalyst, taking the fact that the surface area of g-C3N4 NS (31 wt%)/TMC (57.4 m2g−1) and g-C3N4 NS (31 wt%)/P25 (52.3 m2g−1) was nearly the same, as the H2 evolution rate of g-C3N4 (NS)/TMC was about 7 times higher. Further research indicated that the tight interface between g-C3N4 NS and TMC facilitated the charge transfer, which is a flexible way to promote solar energy utilization of g-C3N4/TiO2 photocatalyst.

Other structures like core-shell was lucubrated to create high photocatalytic activity towards many dyes [106]. After in-situ calcination and growth of cyanamide on the surface of TiO2, a multiple direction contact structure of TiO2@g-C3N4 hollow core@shell heterojunction photocatalyst (HTCN-1) was synthesized. The g-C3N4 nanosheets grew on the surface of TiO2 caused closer contact between TiO2 and g-C3N4 and a larger interfacial area, as confirmed by XPS analysis [106]. Compared with another core-shell type TiO2@g-C3N4 (C-T) with unidirectional contact structures [107], HTCN-1 possessed higher efficiency in the charge separation and enhanced charge transfer. It demonstrated that multiple direction contact resulted in a large interfacial area, which would provide sufficient channels for efficient and rapid charge transfer (Figure 13) [106]. In another core-shell structure of g-C3N4/TiO2 hybrid, Ag was introduced as interlayers to participate in electrical conduction and bridge the gap between g-C3N4 and TiO2, facilitating the separation of photoexcited charge and reducing the recombination of the photogenerated electron hole (Figure 14) [108]. The surface area of the samples didn't change much upon the introduction of Ag (228.4 m2g−<sup>1</sup> and 210.3 m2g−<sup>1</sup> for Ag/TiO2 microspheres and nonsilver containing TiO2, respectively). It was worth noting that low content of g-C3N4 (2%) in g-C3N4/Ag/TiO2 microspheres had a larger surface area but lower photocatalytic activity than the g-C3N4 (4%)/Ag/TiO2 microsphere sample [108]. The possible reason was that high content of g-C3N4 can generate more electron-hole pairs, leading to a higher photocatalytic activity. However, the g-C3N4 (6%)/Ag/TiO2 microsphere sample showed decreased photocatalytic activity due to reduced surface area, which limited the contact between the catalyst and pollutant and thus lowered the photocatalytic reaction. It reflects that proper surface area is needed to provide both active sites and reaction sites.

The doping of g-C3N4 is another viable way to realize structure modification process. Sulfur was introduced to g-C3N4 nanostructures, and their photocatalytic performance was studied for decomposition of MO dye under visible light. The degradation efficiency over g-C3N4-TiO2 composites (CNT) reached 61% within 90 min, while S-C3N4-TiO2 composites (SCNT) reached nearly 100% within the same period [109]. SEM image showed a more transparent and thinner layer of S-C3N4 compared with g-C3N4 when composited with TiO2, leading to an enhanced visible light absorption capability. On the other hand, unique bar-like structure of SCNT provided a pathway for carriers and isolate photon absorption with carriers' collection in perpendicular directions. Meanwhile, TiO2 nanoparticles were more evenly dispersed on and inside S-C3N4 substrate in SCNT sample, which is beneficial for the interfacial carriers' transportation between S-C3N4 layer and TiO2 particle [109]. Calculations revealed that the modified electronic structure with elevation of CB and VB values owing to doped sulfur, contributed to a higher driving force from CB of S-C3N4 to CB of TiO2 and thus promoted the separation efficiency of electron-hole pairs (Figure 15). The doping of sulfur alternated both the structure and level distribution of C3N4, causing excellent separation efficiency of electron-hole pair when contacted with TiO2.

**Figure 13.** Structure of HTCN-1 (**a**) and C-T (**b**). Reprinted with permission from [106]. Copyright 2018, Elsevier.

**Figure 14.** Photocatalytic mechanism scheme of g-C3N4/Ag/TiO2 microspheres under visible light irradiation (>420 nm). Reprinted with permission from [108]. Copyright 2014, American Chemical Society.

**Figure 15.** Mechanism of fast charge transfer at the interface between (**a**) C3N4-TiO2 and (**b**) S-C3N4/TiO2. Reprinted with permission from [109]. Copyright 2017, Elsevier.

### *2.4. MoS2 Modified TiO2*

2D layered transition metal chalcogenides (TMCs) nanostructures spark a research boom due to its unique physical and chemical properties compared with other 2D materials. The usual formula of TMCs is MX2, while M is transition metal and X is chalcogenide element, namely, S, Se, or Te. Because of the typical 2D structure with high surface-to-volume ratio and missing coordination at edge (Figure 16), TMCs exhibits high chemical sensitivity [36]. Considering its versatile physicochemical properties, TMCs can be applied in catalyst [41], energy storage [39], and biology [110]. Some TMCs such as WS2 [111], TiS2 [112] are also used in TiO2 photocatalysis. Among TMCs, MoS2 show extraordinary potential as semiconductors owing to its thickness dependent bandgap and natural abundance. When bulk MoS2 are stripped into a single layer or several layers of nanosheets, the indirect bandgap (1.3 eV) can be converted to a direct bandgap (1.8 eV) [113] and show excellent performance in photocatalysis after compositing with TiO2 [114]. Besides, its high surface-to-volume ratio makes up for the limitation of the low theoretical specific capacity of TiO2. The synergy between MoS2 and TiO2 endows the TiO2/MoS2 composite superior performance compared to their single material.

**Figure 16.** Structure (**a**) and solution-based preparation (**b**) of 2D layered transition metal chalcogenides (TMCs) nanosheets based on top-down and bottom-up approaches. Reprinted with permission from [36]. Copyright 2018, American Chemical Society.

### 2.4.1. The Synthesis of MoS2/TiO2 Composites

Similar to the synthesis methods of graphene/TiO2 composite, the synthesis of MoS2/TiO2 composites is also divided into ex-situ methods and in-situ methods. For the in-situ method, TiO2 and MoS2 are synthesized separately, then the two are combined by various methods, such as hydrothermal/solvothermal assembly [115,116], mechanical method [117], drop-casting [118], or sol–gel [119], which can be also applied for in-situ methods [120,121]. The ex-situ method is simple and inexpensive, but the two compounds have poor dispersion and show weak interactions. Despite the same process as ex-situ method, there are chemical vapor deposition [122] and co-reduction precipitation [123] in in-situ process. Among them, the hydrothermal method is simple, easy to operate, and has good controllability, and thus is most commonly used in the preparation of MoS2/TiO2 composite materials. The in-situ reduction method uses one of the materials as a substrate and then coats or loads the other material. This involves the molybdenum disulfide as substrate or TiO2 as a substrate. The following paragraphs will discuss the two kinds of composites.


**Figure 17.** (**a**) Schematic illustration and (**b**) TEM image of the ALD TiO2 coating on pristine MoS2. Reprinted with permission from [126]. Copyright 2017, Wiley-VCH.

### 2.4.2. The Role of MoS2 in TiO2 Photocatalysis

During the photocatalysis process, electrons transfer through the interface between TiO2 and MoS2, and therefore the contact between the two is vital for photocatalytic activity. A strategy for construction of 3D semiconductor heterojunction structure by TiO2 and 2D-structured MoS2 is proposed to achieve increase of active sites and decrease of electron-hole pair combination [127,132]. For example, a 3D flower-like N-TiO2-x@MoS2 was obtained by hydrothermal method. Considering that the smooth TiO2 nanosphere shows poor affinity when coated with MoS2 nanosheets, TiO2 was doped with N and

Ti3+. X-ray photoelectron spectroscopy (XPS) shows the existence of electronic interactions between MoS2 and N-TiO2-x and the strong heterostructure effect between the MoS2 nanoflower and N-TiO2-x nanosphere [127]. Another study of 3D TiO2@MoS2 revealed that the formation of Ti-S bonds made TiO2 nanoarrays firmly grasp MoS2, thus affording a marvelous mechanical stability for the integrated architectures [133].

Different phase of MoS2 exhibits various chemical and physical properties when combined with TiO2. MoS2 has two main phases, namely the metallic 1T phase and semiconducting 2H phase. As for 2H phase, the active site with catalytic activity is located at the edge of the MoS2 layers and the basal surface of MoS2 is catalytically inactive [134]. Therefore, the 1T phase of MoS2 with active sites on both edge and basal planes attracts researchers' attention in recent years [118,125,131]. A typical schematic of MoS2/TiO2 composites for photocatalytic hydrogen production is shown in Figure 18. The 1T-MoS2 nanosheets not only provide extra reaction sites on the basal plane, but also play a role in electron delivery. Because of the active site distributing on the edge of 2H-MoS2 nanosheets, the photogenerated electron from TiO2 needs a long-distance move before reacted with H2O. This leaded to a lower diffusion rate compared with 1T-MoS2/TiO2 composites and thus enhanced the separation efficiency of electron-hole pairs. Therefore, the 1T-MoS2/TiO2 composites exhibited excellent photocatalytic activity as the hydrogen production rate of 1T-MoS2/TiO2 was 5 and 8 times higher than those of bare TiO2 and 1T-MoS2/TiO2 [125]. In another research, 1T-MoS2 coated onto TiO2 (001) composite (MST) was synthesized. DFT calculations suggested a closer distance between the interface electrons and MoS2 surface than that of TiO2 [131] (Figure 19). Therefore the photo-induced electrons can easily transfer to the conducting channel of MoS2. Furthermore, the introduction of 1T-MoS2 prolonged the carrier lifetime remarkedly. All the factors led to an enhanced photocatalytic activity.

To further inhibit the recombination of electron-hole pairs, cocatalyst such as graphene is applied to MoS2/TiO2 system [115,135,136]. Xiang et al. employed TiO2/MoS2/graphene composite as photocatalyst [135]. In this system, photo-inducted electrons transfer from VB to CB of TiO2. Then the electrons are further injected into the graphene sheets or MoS2 nanoparticles. What is more, graphene sheets can be seen as electrons transport 'highway' through which electrons move from VB of TiO2 to MoS2 (Figure 20). The cocatalyst of MoS2 and graphene enhances the interfacial charge transfer rate, inhibits the recombination of electron-hole pairs and offers a host of active site for adsorption and reaction. Han et al. constructed 3D MoS2/P25/graphene-aerogel networks. In addition to the above-mentioned advantages, 3D graphene porous architecture has a highly porous ultrafine nanoassembly network structure, excellent electric conductivity, and the maximization of accessible sites [115]. Recently, a 3D double-heterostructured photocatalyst was constructed by connecting a TiO2-MoS2 core-shell nanosheets (NSs) on a graphite fiber (GF@MoS2-TiO2) [136]. Mechanism of photocatalytic decomposition of dyes under both visible light and UV light was discussed (Figure 21). Anatase TiO2 has a wide band gap (2.96 eV), while the band gap of MoS2 is 1.8 eV. Because of the moderate bandgap of MoS2, the electrons can be irradiated from VB to CB of MoS2 and then inject into CB of TiO2 or transfer to graphene through intimate double-heterojunction contact under visible light. Graphene acts as electrons accepter under both circumstance, leading to a high rate of charge separation and thus depress the charge recombination. The contact interfaces and synergy among graphene, TiO2 and MoS2 play an important role in the superior photocatalytic activities.

While the transfer of electrons are paid special attention, the role of capturing the holes are often ignored. To solve this problem, a TiO2/WO3@MoS2 (TWM) hybrid Z-scheme photocatalytic system was structured. TiO2 and WO3 have the appropriate energy level matching to form the Z-scheme, while the position of VB in WO3 is lower than the VB of TiO2, and the CB of WO3 is between the CB and VB of TiO2 [137]. Under UV light irradiation, the VB electrons of all three parts are excited to corresponding CB level. The excited electrons on CB of TiO2 then transfer to CB of MoS2 for H2 evolution, meanwhile the excited electrons on CB of WO3 were inject to the VB of TiO2 (Figure 22). This procedure suppressed the recombination of photoinduced electrons and holes in TiO2, and therefore the photogenerated electrons and holes can be efficiently separated, which further leads to effective photocatalytic activity [137].

**Figure 18.** Schematic illustrating charge-transfer behavior and H2 evolution active sites for (**a**) 1T-MoS2/ TiO2 and (**b**) 2H-MoS2/TiO2. Reprinted with permission from [125]. Copyright 2014, Springer.

**Figure 19.** (**a**) The charge density difference, (**b**) electrostatic potential and differential charge density of the MoS2/TiO2(001) junction; (**c**) Planar-averaged differential electron density Dr(z) for MoS2/TiO2(001); (**d**) Photocatalytic mechanism for 1T-MoS2/TiO2. Reprinted with permission from [131]. Copyright 2017, the Owner Societies.

**Figure 20.** Schematic illustration of the charge transfer in TiO2/MoS2/graphene composites. Reprinted with permission from [135]. Copyright 2012, American Chemical Society.

**Figure 21.** Structure (**a**) and schematic diagram of electron-hole separation mechanism upon UV (**b**) and visible light (**c**) excitation for 3D graphene@MoS2-TiO2 composites. Reprinted with permission from [136]. Copyright 2017, Elsevier.

**Figure 22.** (**a**) Schematic illustration for the growth of MoS2 nanosheets (**b**) Schematic diagram of the photocatalytic H2 generation over the ternary TiO2/WO3@MoS2 heterostructure composite. Reprinted with permission from [137]. Copyright 2017, Elsevier.

### **3. Conclusions**

The coupling between TiO2 and 2D material has proven to be an efficient approach to enhanced photocatalytic activity. Different methods vary the structures and surface contact of the hybrid and thus can modify the carrier separation process. The synergistic effects show that 2D material plays a vital role in photocatalysis when composited with TiO2. First, 2D material can act as electrons accepter or bridge to conduct photoinduced electrons, and therefore represses the recombination of carriers efficiently. Second, the gigantic surface of 2D material provides substantial active sites for substrate capture and reaction, not to mention rapid electrons transfer rate. Third, the 2D material can be decorated to obtain expected properties, for example, non-metal doping to adjust the energy level, specific crystal structure to short the pathway for interfacial charge transfer, and defects or introduced functional group for substrate trapping. What's more, the interfacial heterojunction can adjust energy level to broaden light response range and improve solar utilization. To further enhance the separation efficiency of electron-hole pairs, other photocatalysts are introduced to construct co-catalyst systems among which Z-scheme system can raise the hole trapping rate to some extent, and thus offers a new point to improve the separation of carriers. All factors mentioned above highlight the critical role of 2D material in photocatalyst and the 2D material/TiO2 hybrid is worth to get further insight for a wider range of applications.

**Funding:** This research was funded by the National Natural Science Foundation of China (21761132007, 21773168, and 51503143), the National Key R&D Program of China (2016YFE0114900), Tianjin Natural Science Foundation (16JCQNJC05000), Innovation Foundation of Tianjin University (2016XRX-0017), and Tianjin Science and Technology Innovation Platform Program (No. 14TXGCCX00017).

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
