2D Transition Metal Dichalcogenides and Graphene-Based Ternary Composites for Photocatalytic Hydrogen Evolution and Pollutants Degradation

Photocatalysis have attracted great attention due to their useful applications for sustainable hydrogen evolution and pollutants degradation. Transition metal dichalcogenides (TMDs) such as MoS2 and WS2 have exhibited great potential as cocatalysts to increase the photo-activity of some semiconductors. By combination with graphene (GR), enhanced cocatalysts of TMD/GR hybrids could be synthesized. GR here can act as a conductive electron channel for the transport of the photogenerated electrons, while the TMDs nanosheets in the hybrids can collect electrons and act as active sites for photocatalytic reactions. This mini review will focus on the application of TMD/GR hybrids as cocatalysts for semiconductors in photocatalytic reactions, by which we hope to provide enriched information of TMD/GR as a platform to develop more efficient photocatalysts for solar energy utilization.


Introduction
Since the discovery of the photocatalytic splitting of water on TiO 2 electrodes by Fujishima and Honda in 1972 [1], photocatalysis has attracted great attention for eliminating hazardous pollutants and generating sustainable energy [2]. Semiconductors such as TiO 2 , ZnO, CdS, etc. can act as photocatalysts for the utilization of solar energy [3]. They are however limited in real application by the rapid electron-hole recombination [4][5][6]. Noble metal cocatalysts are usually loaded to enhance the activity of the semiconductor photocatalysts [7]. However, these metals are rare and expensive to apply [8]. The development of highly active and low cost cocatalysts remains a great challenge in the field of photocatalysis.
Transition metal dichalcogenides (TMDs) such as MoS 2 and WS 2 have exhibited excellent activities as cocatalysts for the modification of semiconductors [8,9]. The properties of TMDs can be tailored according to their crystalline structure and the number and stacking sequence of their nanosheets [10][11][12][13][14]. By loading TMDs cocatalysts, semiconductor-semiconductor or metal-semiconductor junctions will form, and more interfaces could be created [15,16]. Charge separation and electron transport can therefore be enhanced, leading to the activity improvement [15]. Furthermore, many kinds of TMDs with different phases were reported to be active for the electrochemical hydrogen evolution reaction (HER), which stems from their exposed and also lower the activation energy and overpotential for photocatalytic H2 evolution [8]. On the other hand, TMDs have special 2D layered structure, and can be used as effective supports for anchor of semiconductor nanoparticles, which could reduce the mobility, provide more active sites, and avoid coalescence and agglomeration of the semiconductors [20][21][22]. Based on the above analysis, TMDs have shown great potential as substitute of noble metal cocatalysts for the synthesis of composite photocatalysts with high activity [8].
Graphene (GR) consists of a single layer and sp 2 -hybridized carbon lattice with excellent electrical (200,000 cm 2 ·V −1 ·s −1 ), thermal, and mechanical properties, and is a novel material that has emerged as a rapidly rising star in the field of material science [23]. The photocatalytic activity of semiconductors can be greatly increased by loading GR as cocatalyst, mainly owing to the effective separation of the electron-hole pairs [24,25].
By combining GR with TMDs, new hybrid cocatalysts could be synthesized with 2D layered structures. GR here can transport the photogenerated electrons rapidly, and the TMDs in the hybrids can accept electrons and act as active sites for H2 evolution or radicals generation. This mini review will focus on the synthesis methods of TMD/GR-based photocatalysts and their applications for photocatalytic H2 evolution and organic pollutants degradation. Based on this review, we hope to offer enriched information of TMD/GR as a platform to fabricate more efficient photocatalysts for solar energy utilization.

TMD/GR-Based Composites for Photocatalytic H2 Evolution
TiO2 is the most frequently used semiconductor for photocatalytic H2 evolution, which can only absorb and utilize UV light due to its large band gap (3.2 eV) [26]. Xiang et al. synthesized ternary composites consisting of TiO2 nanoparticles grown on the MoS2/GR hybrid as enhanced photocatalysts for H2 evolution (Figure 1) [25]. The TiO2-MoS2/GR composites were prepared using a two-step hydrothermal method. As shown in Figure 1, the TiO2 nanoparticles were supported on the 2D MoS2/GR hybrid uniformly with intimate contact. The electrons can therefore transfer rapidly from TiO2 to the MoS2/GR cocatalyst, and the charge recombination can therefore be suppressed. The activity of the ternary composites can be tuned by adjusting the GR percentage of MoS2/GR cocatalysts and the percentage of the MoS2/GR hybrid for the ternary photocatalysts. The optimized TiO2-MoS2/GR composite could obtain a high H2 evolution rate of 165.3 μmol·h −1 and a quantum efficiency of 9.7% at 365 nm.  CdS has a narrow bandgap of 2.3 eV, which is effective for capturing the visible light [36]. Chang et al. used nanosized GR as support for the growth of MoS 2 ; 3D hierarchical CdS-MoS 2 /GR composites with diameters of 100-300 nm were then synthesized with the help of polyvinylpyrrolidone (PVP) (Figure 2) [27]. As shown in Figure 2e, the MoS 2 sheets on nanosized GR have many defect sites and disordered structures due to the low synthesis temperature, CdS nanoparticles can then firmly anchor on these defects and vacancies (Figure 2g-j). After optimizing each component proportion, the highest H 2 evolution rate could be as large as 1.8 mmol/h with an apparent quantum efficiency (AQE) of 28.1% at 420 nm, which was even higher than that of Pt/CdS. They thought that the activities of S atoms in the MoS 2 molecules were different with respect to their different coordination (Figure 3a). Unsaturated S atoms are active for H + adsorption and reduction (Figure 3b), while the saturated S atoms on the basal plane are inert. The nanosized few-layer MoS 2 supported on GR has more exposed edges and unsaturated active S atoms, which is therefore a promising cocatalyst for the activity enhancement of CdS. CdS has a narrow bandgap of 2.3 eV, which is effective for capturing the visible light [36]. Chang et al. used nanosized GR as support for the growth of MoS2; 3D hierarchical CdS-MoS2/GR composites with diameters of 100-300 nm were then synthesized with the help of polyvinylpyrrolidone (PVP) ( Figure 2) [27]. As shown in Figure 2e, the MoS2 sheets on nanosized GR have many defect sites and disordered structures due to the low synthesis temperature, CdS nanoparticles can then firmly anchor on these defects and vacancies (Figure 2g-j). After optimizing each component proportion, the highest H2 evolution rate could be as large as 1.8 mmol/h with an apparent quantum efficiency (AQE) of 28.1% at 420 nm, which was even higher than that of Pt/CdS. They thought that the activities of S atoms in the MoS2 molecules were different with respect to their different coordination ( Figure  3a). Unsaturated S atoms are active for H + adsorption and reduction (Figure 3b), while the saturated S atoms on the basal plane are inert. The nanosized few-layer MoS2 supported on GR has more exposed edges and unsaturated active S atoms, which is therefore a promising cocatalyst for the activity enhancement of CdS.   CdS has a narrow bandgap of 2.3 eV, which is effective for capturing the visible light [36]. Chang et al. used nanosized GR as support for the growth of MoS2; 3D hierarchical CdS-MoS2/GR composites with diameters of 100-300 nm were then synthesized with the help of polyvinylpyrrolidone (PVP) ( Figure 2) [27]. As shown in Figure 2e, the MoS2 sheets on nanosized GR have many defect sites and disordered structures due to the low synthesis temperature, CdS nanoparticles can then firmly anchor on these defects and vacancies (Figure 2g-j). After optimizing each component proportion, the highest H2 evolution rate could be as large as 1.8 mmol/h with an apparent quantum efficiency (AQE) of 28.1% at 420 nm, which was even higher than that of Pt/CdS. They thought that the activities of S atoms in the MoS2 molecules were different with respect to their different coordination ( Figure  3a). Unsaturated S atoms are active for H + adsorption and reduction (Figure 3b), while the saturated S atoms on the basal plane are inert. The nanosized few-layer MoS2 supported on GR has more exposed edges and unsaturated active S atoms, which is therefore a promising cocatalyst for the activity enhancement of CdS.   Xiang et al. synthesized CdS-WS 2 /GR ternary composites for photocatalytic H 2 evolution [32]. The optimized WS 2 /GR content in the ternary CdS-WS 2 /GR composites was determined to be 4.2 wt %. Using 0.35 M Na 2 S/0.25 M Na 2 SO 3 as sacrificial agent, a satisfactory H 2 evolution rate of 1.842 mmol·h −1 ·g −1 could be achieved with an apparent quantum efficiency of 21.2% at 420 nm. The transient photocurrent response was also enhanced by loading the WS 2 /GR cocatalyst, which was promising evidence for the improved charge transport (Figure 4). By loading WS 2 /GR cocatalyst, more active sites will be introduced, and charge separation and interfacial charge transfer could be enhanced, thus leading to the greatly increased photo-activity. Xiang et al. synthesized CdS-WS2/GR ternary composites for photocatalytic H2 evolution [32]. The optimized WS2/GR content in the ternary CdS-WS2/GR composites was determined to be 4.2 wt %. Using 0.35 M Na2S/0.25 M Na2SO3 as sacrificial agent, a satisfactory H2 evolution rate of 1.842 mmol·h −1 ·g −1 could be achieved with an apparent quantum efficiency of 21.2% at 420 nm. The transient photocurrent response was also enhanced by loading the WS2/GR cocatalyst, which was promising evidence for the improved charge transport (Figure 4). By loading WS2/GR cocatalyst, more active sites will be introduced, and charge separation and interfacial charge transfer could be enhanced, thus leading to the greatly increased photo-activity.  During the photocatalytic H2 evolution process, the metal oxide can absorb a photon to create an electron-hole pair with irradiation (Figure 5a,c). Both of the graphene/ graphene •− redox potential  Xiang et al. synthesized CdS-WS2/GR ternary composites for photocatalytic H2 evolution [32]. The optimized WS2/GR content in the ternary CdS-WS2/GR composites was determined to be 4.2 wt %. Using 0.35 M Na2S/0.25 M Na2SO3 as sacrificial agent, a satisfactory H2 evolution rate of 1.842 mmol·h −1 ·g −1 could be achieved with an apparent quantum efficiency of 21.2% at 420 nm. The transient photocurrent response was also enhanced by loading the WS2/GR cocatalyst, which was promising evidence for the improved charge transport (Figure 4). By loading WS2/GR cocatalyst, more active sites will be introduced, and charge separation and interfacial charge transfer could be enhanced, thus leading to the greatly increased photo-activity.  During the photocatalytic H2 evolution process, the metal oxide can absorb a photon to create an electron-hole pair with irradiation (Figure 5a,c). Both of the graphene/ graphene •− redox potential During the photocatalytic H 2 evolution process, the metal oxide can absorb a photon to create an electron-hole pair with irradiation (Figure 5a,c). Both of the graphene/ graphene •− redox potential and conduction band (CB) of quantum-sized MoS 2 are slightly lower than the CB of anatase TiO 2 or CdS (Figure 4b,d). Part of the excited electrons can then directly transfer to active sites of MoS 2 , and another part transfer to active sites of MoS 2 via graphene conducting channel. Graphene here can act as a "highway" for the rapid transport of photo-generated electrons, while MoS 2 nanosheets can accept electrons and act as active sites for H 2 evolution. Therefore, using MoS 2 /GR hybrid as cocatalyst, suppression of charge recombination, improvement of interfacial charge transfer, and an increase in the number of active sites could be achieved, thus leading to the enhanced photo-activity.
It has been reported that sub-10 nm rutile TiO 2 with 1 wt % Pt doping exhibited state-of-the-art activity among TiO 2 -based composites for photocatalytic water splitting. The hydrogen evolution rate could be achieved to 932 mmol·h −1 ·g −1 under visible light (>400 nm) and 1954 mmol·h −1 ·g −1 under simulated solar light [37]. By loading 0.30 wt % of Pt and 0.13 wt % of PdS as cocatalysts on CdS, another CdS-based state-of-the-art material could be synthesized with a quantum efficiency (QE) up to 93% and a hydrogen evolution rate of 8.77 mmol·h −1 [38]. Compared to these state-of-the-art materials, the activities of TMD/GR modified semiconductors are relatively weak, with lower H 2 evolution rates and QEs. Although the TMD/GR cocatalysts are more cost effective, deep studies are still needed to obtain higher efficiencies for real application.

TMD/GR-Based Photocatalysts for Pollutants Degradation
Photocatalysis is also an attractive technology for the degradation of pollutants in water using solar energy [39]. Han et al. used a hydrothermal method to combine the exfoliated MoS 2 , GR, and TiO 2 P25 together [40]. The obtained composite was a novel graphene-based three-dimensional (3D) aerogel embedded with TiO 2 particles and MoS 2 nanosheets ( Figure 6). Porous structure could be observed with pore sizes of about several micrometers (Figure 6b). The Nyquist plots of the samples were also tested, and the final 3D GR-MoS 2 -TiO 2 composite had the smallest cure radius (Figure 7), indicating that the addition of 3D graphene aerogel can decrease the solid state interface layer resistance and the charge transfer resistance. During the application test, the final composite had better adsorption ability for methyl orange (MO) due to the maximization of accessible sites of the 3D interconnected networks (Figure 8a). The 3D photocatalyst was proved to be very effective for the photocatalytic degradation of MO, and nearly no MO was left after 15 min irradiation (Figure 8b). and conduction band (CB) of quantum-sized MoS2 are slightly lower than the CB of anatase TiO2 or CdS (Figure 4b,d). Part of the excited electrons can then directly transfer to active sites of MoS2, and another part transfer to active sites of MoS2 via graphene conducting channel. Graphene here can act as a "highway" for the rapid transport of photo-generated electrons, while MoS2 nanosheets can accept electrons and act as active sites for H2 evolution. Therefore, using MoS2/GR hybrid as cocatalyst, suppression of charge recombination, improvement of interfacial charge transfer, and an increase in the number of active sites could be achieved, thus leading to the enhanced photo-activity. It has been reported that sub-10 nm rutile TiO2 with 1 wt % Pt doping exhibited state-of-the-art activity among TiO2-based composites for photocatalytic water splitting. The hydrogen evolution rate could be achieved to 932 mmol·h −1 ·g −1 under visible light (>400 nm) and 1954 mmol·h −1 ·g −1 under simulated solar light [37]. By loading 0.30 wt % of Pt and 0.13 wt % of PdS as cocatalysts on CdS, another CdS-based state-of-the-art material could be synthesized with a quantum efficiency (QE) up to 93% and a hydrogen evolution rate of 8.77 mmol·h −1 [38]. Compared to these state-of-the-art materials, the activities of TMD/GR modified semiconductors are relatively weak, with lower H2 evolution rates and QEs. Although the TMD/GR cocatalysts are more cost effective, deep studies are still needed to obtain higher efficiencies for real application.

TMD/GR-Based Photocatalysts for Pollutants Degradation
Photocatalysis is also an attractive technology for the degradation of pollutants in water using solar energy [39]. Han et al. used a hydrothermal method to combine the exfoliated MoS2, GR, and TiO2 P25 together [40]. The obtained composite was a novel graphene-based three-dimensional (3D) aerogel embedded with TiO2 particles and MoS2 nanosheets ( Figure 6). Porous structure could be observed with pore sizes of about several micrometers (Figure 6b). The Nyquist plots of the samples were also tested, and the final 3D GR-MoS2-TiO2 composite had the smallest cure radius (Figure 7), indicating that the addition of 3D graphene aerogel can decrease the solid state interface layer resistance and the charge transfer resistance. During the application test, the final composite had better adsorption ability for methyl orange (MO) due to the maximization of accessible sites of the 3D interconnected networks (Figure 8a). The 3D photocatalyst was proved to be very effective for the photocatalytic degradation of MO, and nearly no MO was left after 15 min irradiation (Figure 8b).   Gao et al. fabricated a TiO2-MoS2/GR composite under atmospheric pressure using a simple onepot solvothermal method [34]. Na2MoO4 and thiocarbamide were used as precursors for MoS2, and mixed solvent of (dimethylacetamide (DMAc)/deionized (DI) H2O) was used as reaction media. Under the above conditions, MoS2 quantum dots (QDs) with (100) face exposed could be generated on the surface of TiO2 and GR (Figure 9). Attributed to the small diameter of the MoS2 QDs, more active edge could be created, thus leading to the enhanced photocatalytic activity.  Gao et al. fabricated a TiO2-MoS2/GR composite under atmospheric pressure using a simple onepot solvothermal method [34]. Na2MoO4 and thiocarbamide were used as precursors for MoS2, and mixed solvent of (dimethylacetamide (DMAc)/deionized (DI) H2O) was used as reaction media. Under the above conditions, MoS2 quantum dots (QDs) with (100) face exposed could be generated on the surface of TiO2 and GR (Figure 9). Attributed to the small diameter of the MoS2 QDs, more active edge could be created, thus leading to the enhanced photocatalytic activity.  [34]. Na 2 MoO 4 and thiocarbamide were used as precursors for MoS 2 , and mixed solvent of (dimethylacetamide (DMAc)/deionized (DI) H 2 O) was used as reaction media. Under the above conditions, MoS 2 quantum dots (QDs) with (100) face exposed could be generated on the surface of TiO 2 and GR ( Figure 9). Attributed to the small diameter of the MoS 2 QDs, more active edge could be created, thus leading to the enhanced photocatalytic activity.
Peng et al. synthesized Ag 3 PO 4 -MoS 2 /GR via a simple two-step hydrothermal process [33]. The composite was found to be an effective catalyst for the photo-decomposition of 2,4-dichlorophenol (DCP) under simulated solar light and visible light (λ > 420 nm). They described the major reaction steps involved in this photocatalytic process as follows: ·OH + DCP → CO 2 + H 2 O + other products (6) As shown in the mechanism, electrons and holes could be separated with irradiation (1). The holes could oxidize the DCP molecules adsorbed on the catalyst surface directly (4). They could also react with water (or hydroxyl) to form hydroxyl free radicals (·OH), which are strong oxidants for DCP decomposition (5). The MoS 2 /GR cocatalyst here could act as electron collectors to facilitate the interfacial electron transfer and charge separation. In addition, the MoS 2 /GR cocatalyst could also provide more active sites and allow for the activation of dissolved O 2 for organic degradation [33]. Peng et al. synthesized Ag3PO4-MoS2/GR via a simple two-step hydrothermal process [33]. The composite was found to be an effective catalyst for the photo-decomposition of 2,4-dichlorophenol (DCP) under simulated solar light and visible light (λ > 420 nm). They described the major reaction steps involved in this photocatalytic process as follows: Ag3PO4 (e − ) + MoS2 /GR → Ag3PO4 + MoS2 /GR (e − ) MoS2 /GR (e − ) + O2 → MoS2/ GR + O2 − Ag3PO4 (h + ) + DCP → CO2 + H2O + other products (4) · OH + DCP → CO2 + H2O + other products (6) As shown in the mechanism, electrons and holes could be separated with irradiation (1). The holes could oxidize the DCP molecules adsorbed on the catalyst surface directly (4). They could also react with water (or hydroxyl) to form hydroxyl free radicals (· OH), which are strong oxidants for DCP decomposition (5). The MoS2/GR cocatalyst here could act as electron collectors to facilitate the interfacial electron transfer and charge separation. In addition, the MoS2/GR cocatalyst could also provide more active sites and allow for the activation of dissolved O2 for organic degradation [33].
Using CoS2/GR as cocatalyst, Zhu et al. supported TiO2 nanoparticles on its surface using a facile sonochemical and hydrothermal method [35]. Their photo-activity was then evaluated for the Using CoS 2 /GR as cocatalyst, Zhu et al. supported TiO 2 nanoparticles on its surface using a facile sonochemical and hydrothermal method [35]. Their photo-activity was then evaluated for the degradation of methylene blue (MB) and Texbrite BA-L (TBA) under visible light. Enhanced activity was obtained due to the synergetic effect between TiO 2 and the CoS 2 /GR cocatalyst. The recent progress of TMD/GR based photocatalysts for H 2 evolution and pollutants degradation are summarized and shown in Table 1 for a easier perusal.

Conclusions and Perspective
This mini review focused on the recent developments of the TMD/GR-based composites, including the synthesis methods, the application in photocatalytic H 2 evolution, and the application for organic pollutants degradation. By combination with GR, the TMD/GR hybrids were more effective as cocatalysts for the modification of semiconductors. GR here can act as a conductive electron transport "highway" for the transport of the photogenerated electrons, and the TMDs nanosheets in the hybrids can accept electrons and act as active sites for photocatalytic reactions. Although deep research is still needed for real application, TMD/GR cocatalysts have shown great potential as a platform to fabricate more efficient photocatalysts for solar energy utilization.