Transition Metal Dichalcogenides [MX₂] in Photocatalytic Water Splitting

Abstract

The quest for a clean, renewable and sustainable energy future has been highly sought for by the scientific community over the last four decades. Photocatalytic water splitting is a very promising technology to proffer a solution to present day environmental pollution and energy crises by generating hydrogen fuel through a “green route” without environmental pollution. Transition metal dichalcogenides (TMDCs) have outstanding properties which make them show great potential as effective co-catalysts with photocatalytic materials such as TiO₂, ZnO and CdS for photocatalytic water splitting. Integration of TMDCs with a photocatalyst such as TiO₂ provides novel nanohybrid composite materials with outstanding characteristics. In this review, we present the current state of research in the application of TMDCs in photocatalytic water splitting. Three main aspects which consider their properties, advances in the synthesis routes of layered TMDCs and their composites as well as their photocatalytic performances in the water splitting reaction are discussed. Finally, we raise some challenges and perspectives in their future application as materials for water-splitting photocatalysts.

1. Introduction

Society at large is prospecting for carbon-free fuel alternatives to purvey an unfailing panacea to the present energy plight of the world. Research efforts into the development of alternate clean and renewable energy sources have been at a steady rise in the last few decades. Hydrogen is credited to be an unpolluted, robust, environmentally safe, and emerging fuel, efficient to curb total reliance on petroleum-based fuels. Currently, methane reforming accounts for the largest production of hydrogen fuel, where methane gas is converted to carbon monoxide and hydrogen. This process is energy inefficient and has a negative environmental impact. Consequently, it is desirable generating hydrogen through unexhausted resources such as water and solar power. Solar light-assisted water decomposition is an emerging method for sustainable hydrogen production. Producing hydrogen through solar simulated light, water and photocatalyst is a desired technology not only because water and solar energy are abundant on the earth but also because it is environmentally friendly and economically viable. Hydrogen is attributed to have the maximum energy content per weight in comparison to other combustion fuels, the by-products are environmentally friendly which are water vapour and [1,2,3,4].

To a consumer, similar to natural gas serves many purposes, hydrogen gas could be used to heat homes and as a fuel for automobiles. Environmentally, hydrogen is exceptional among other fuels, being a carbon-free fuel, it will generate zero carbon dioxide during combustion. To the scientists, hydrogen is the most simplified and most abundant element in the world. The hydrogen atom (H) is made of one electron and one proton but as a molecule (H₂), it has a favourable physical feature. In terms of energy, it is the highest carrier of energy, having 2.4 folds higher energy content than that of natural gas. This can be put to use using fuel cell to generate electricity or by combustion to produce heat for homes use. It is also a carbon neutral energy carrier. With many countries till date still dependent on energy from fossil fuel sources which is a major source of environmental pollution, there is a dire need for cleaner energy to meet CO₂ emission reduction targets. The Asian countries are the major players in realising the hydrogen economy. Commitment is a vital key to attain the regulatory CO₂ emission targets, as it is very costly to achieve. Japan has keyed into the Paris climate agreement of 2016 and has made much commitment to be a large-scale hydrogen user for energy. Most countries in Europe have signed an agreement based on the European Green Deal to ensure zero-emissions for cars and vans and to achieve zero net emissions of greenhouse gases in Europe by 2050 [1,5]. Little work has been carried out or reported on hydrogen economy in Africa. In South Africa, the telecommunication industry is already using hydrogen fuel. This is used as support system power for petrol or diesel generators in military, hospitals and mines in case of intermittent power supply.

With water and sunlight being free and abundant in nature, this method of generating hydrogen gas is beneficial, environmentally safe, and renewable avenue which could serve as valuable resource to tackle both the energy plight and environmental pollution. Photocatalytic processes have been considered as one of the most promising methods to split water into hydrogen and oxygen. For an effective photocatalytic process, there is a need to use a sturdy and very active photocatalyst to produce good yields of hydrogen gas. Since Fujishima and Honda [6] used TiO₂ electrodes for photocatalytic water splitting, a large array of other photocatalysts have been studied for solar driven water splitting; for instance, CdS, CuS, C₃N₄, ZnO, BiVO₄, MoS₂ [7,8,9,10,11,12,13]_. So far, numerous semiconductor photocatalysts are being explored and used for photocatalytic water splitting. Photocatalysts can be subdivided into three categories namely: metal oxides, metal chalcogenides, and metal–free photocatalysts. Despite the progress made so far on these photocatalysts, they still have some deficiencies. The metal oxide makes use of ultraviolet region of the spectrum because of their large band gap energy. Some semiconductors favour either water reduction or oxidation activity, hence they are not appropriate for effective water splitting. These semiconductors undergo quick charge co-existence of the electron-hole pair resulting in inefficient charge separation and finally, the active sites must not be covered, but in bulk they are not available to partake in photocatalytic processs [7,14,15].

Transition metal dichalcogenides (TMDCs) is an emerging class of two-dimensional (2-D) materials that have exhibited great potential in photocatalytic applications owing to their intrinsic properties. They are normally used as cocatalysts together with other semiconductor materials and the junction created between them facilitates charge transfer of the photogenerated electrons and holes. TMDCs have the general formula MX₂, where M is a transition metal (Mo, W, Ti, Zr, Hf, Nb, etc.) and X is a chalcogen element (S, Se, or Te). While TMDCs of groups V and VI elements have been extensively studied for applications in photocatalysis, in the open literature, not much information is currently available on the group IV TMDCs ZrS₂ and HfS₂ especially synthesized by colloidal methods as these are very unstable and are easily oxidized in mono or few-layers. This review focuses on the possibility for preparing group IVB TMDCs using colloidal method and the possibility of their oxidation in ambient conditions. Lastly the review portrays group IV TMDC as a possible promoter for semiconductors in photocatalytic water splitting. These group IV TMDCs have only been demonstrated theoretically using Density functional theory (DFT) for water splitting and thus the review also emphasizes the possible consideration of ZrS₂ and HfS₂ for photocatalytic water splitting.

2. Catalysts for Hydrogen Evolution

2.1. Basic Principle of Photocatalytic Water Splitting

The basic principles of photocatalytic water splitting are illustrated in Figure 1. It is expedient that the base of the conduction band minimum (CBM) be more negative than the reduction potential of H⁺/H₂, and the same time, the upper part of the valence band maximum (VBM) be lower than the oxidation potential of H₂O/O₂ and close to 2.0 eV band gap of energy which will be suitable to push the oxidation and reduction of water. Indeed, 1.23 eV is the required free energy needed to achieve water splitting, overpotentials resulting from slow reaction and resistance in the system extend the required band higher. Therefore, intense research by material scientists has been focussed to obtain a photocatalyst whose band gap is very close to 2.0 eV for optimizing solar energy adequately. The catalyst must also be insoluble in aqueous solution. In addition, the catalyst should possess high charge mobility, appropriate Femi levels and finally good photo-corrosion resistance [7,14,16].

In a typical reaction the co-existence of photogenerated electron-hole pairs is important. The fast recombination of the charge carriers reduces the efficacy of the photocatalyst [7,14]. The photocatalytic H_2-evolution systems can be enhanced by lengthening the parts of uncovered surface of photocatalyst and delaying the recombination of electron-hole pairs. The following approaches have been employed to facilitate the process: doping with metals and non-metals; dye sensitization; surface modification; construction of composite with a co-catalyst; doping of composite catalysts with metal and immobilization and stabilization on support materials [7,14,17].

The activity of a photocatalyst depends on many properties which include crystal structure, particle size, band structure, electron affinity, as well as the nature of interaction that exits in a composite whether close or partial. Therefore, to obtain good output from decomposition of water into hydrogen, the structural coupling of the photocatalyst and co-catalyst is important; if the contact is partial the hydrogen yield will be very low. The light absorption of photocatalysts is improved and there is delay in the recombination of the charge carriers when the proper heterojunction is formed [14,16]. Noble metal-based co-catalysts that are commonly used are Ru, Rh, Pt, Pd, Ag and Au. Extensive investigations have shown that these rare and costly metals are very proficient in boosting the activity of the photocatalyst. The most preferred is Pt, it has the advantage of having highest work function and its over potential for H₂ production is the least. Reported research on visible-light irradiation water splitting has shown that Pt loaded on a photocatalyst results in maximum photocatalytic activities for H₂-evolution [17].

Obtaining a high level of quantum efficiency is very important in photocatalytic water splitting with the following methods commonly used to increase quantum efficiency: varying the doping density, defects levels, local excitonic effects and structural responses. The photo response of the catalysts also determines the yield of H₂ liberated. The following methods are used to enhance the light adsorption of a catalyst: by changing factors such as environment or external fields which include electric field, optical fields, magnetic fields, temperature, and pressure [16].

For effective total water decomposition, the reverse reaction of product recombination into water should be prevented. At the same time, the use of sacrificial agents which are electron donors which help in minimizing the co-occurrence of the electron-hole pair is important [18,19]. However, this sacrificial agent may also become oxidized during the reaction and become involved in competitive reduction with the product H₂ formed affecting its production. Sacrificial agents are subdivided into two; (i) sacrificial organic electron donors and (ii) sacrificial inorganic electron donors. Among the sacrificial organic electron donors, methanol has been preferentially used due to its high effectiveness in photocatalytic water splitting as holes collectors. Sulphide, S²⁻ (e.g., Na₂S) and sulphite, SO₃²⁻ (e.g., Na₂SO₃) are commonly also used in photocatalytic HER application as efficient hole acceptors [19]. Thus, a co-catalyst with good activity and selectivity, that can advance H₂ or O_2-evolution and avoid the reverse reactions, as well as effectively separate the photogenerated charges is essential for achieving great productivity in the reaction process. Their use is of important to effectively produce H₂ as well as O_2-evolution. Basso and Urakawa reported that undecorated TiO₂ based material could only generate traces of H₂ without any O₂ evolution but when a co-catalyst was introduced, both H₂ and O₂ production were evident [20]. The most active co-catalysts are reported to be the Rh-Cr promoter such as Rh_2-yCr_yO₃ and core-shell structured Rh-Cr₂O_3. Basso and Urakawa doped gallium oxide with zinc and loaded it with Rh-Cr as co-catalyst. The resulting catalyst produced outstanding results for water splitting with a generation rate of 21 mmolh⁻¹ for H₂ evolution without the use of sacrificial agents. This was one of the best catalytic activities in catalyst modification series. However, the cost and scarcity of these precious metal-based co-catalysts has hampered their usage. Therefore, in the research for precious metal-free co-catalysts, earth abundant and cheap TMDCs have been considered as potential candidates for economical and productive solar driven water splitting.

Photocatalytic water splitting has been reportedly combined with other technologies to achieve high hydrogen production rates. Combining photocatalysis with electrocatalysis has been used in this regard. In photoelectrocatalysis, the photocatalyst is fabricated into an electrode which is initiates electrochemical transformations after light irradiation. In the process, when the electron-hole pairs are generated at the photoanode, electrons are collected by the photoanode and are conducted through an external circuit to the photocathode where they participate in reduction reaction in the presence of an electrocatalyst. The photogenerated holes are consumed by the oxidation of water [21]. Transition metal dichalcogenides have been investigated for photoelectrocatalytic water splitting because of their desirable properties such as narrow band gaps, large surface areas and high surface atoms which are important in minimising the recombination of the charge carriers [22]. An In₂Se₃/MoS₂ heterojunction catalyst was prepared as a photoanode for water splitting and exhibited a significantly higher photocurrent density and higher oxygen evolution amount than the pristine In₂Se₃ and was attributed to more efficient utilization of the photogenerated electron-hole pairs as well as reduced charge transfer resistance resulting from the interaction of the two materials [23]. Hendi et al. [22] prepared a Ag NPs-decorated MoS₂/RGO/NiWO₄ catalyst and found it to demonstrate a higher photoelectrocatalytic water splitting activity to the corresponding pristine and two-component system such as MoS₂, NiWO₄, and MoS₂/NiWO₄. The RGO in the composite catalyst served as an electro mediator for shuttling electrons between MoS₂ and NiWO₄, leading to the accumulation of photogenerated electrons on the conduction band of MoS₂ with high reduction ability of holes on the valence band of NiWO₄ with high oxidation ability. Photocatalysis has also been used in photovoltaic electrochemical cells for water splitting. The photovoltaic electrochemical device consists of a combination of a photovoltaic device as a power source and an electrocatalyst for water splitting. In such a device, the charge carriers are generated in the photovoltaic part and transferred to the electrocatalyst where water is split [24]. A trimetallic NiMoV catalyst was used in the cathode whilst NiO was used in the cathode of a thermally integrated photovoltaic electrolysis device and employed in hydrogen production from water splitting [25]. Photothermal catalysis making use of the synergistic effects of heat energy and photocatalysis has also been used in water splitting to overcome the low efficiency problem in conventional photocatalysis. A multifunctional ZnIn₂S₄ supported Pt catalyst was used for photocatalytic and photothermal photocatalytic water splitting using full spectrum light. In the photothermal reaction, infrared and plasmon thermal effects raised the reaction temperature by about 45 °C on the catalyst surface and led to a hydrogen evolution rate more than twice the conventional photocatalytic reaction [26].

2.2. Transition Metal Dichalcogenides

Among the new photo-responsive materials, much attention is being given to two-dimensional (2D) chalcogenide materials which have attracted great research to explore them in the last few decades [27,28,29,30,31]. TMDC materials represent a new set of sophisticated class of materials that possess layered structure similar to clay structure; this is illustrated in Figure 2, they may have single or few atoms depending on the thickness of the nanomaterial [27,31,32,33]. Figure 2a shows side view of a monolayer TMDC, a monolayer is a three-atom stick. Figure 2b shows the crystal structure of a TMDC in which few layers are upon each other and Figure 2c depicts the sheet like nature of a monolayer TMDC.

The shift to TMDCs emanated about two decades ago, as a result of the exfoliation of graphene, a single-layered carbon material with outstanding properties such as mechanical, thermal, and electrical conductivities. In contrast with graphene, a TMDC’s monolayer is three atoms thick consisting of a layer of transition metal atoms such as Zr, Hf, Ti, Ta, Mo sandwiched between two planes of a chalcogen atoms such as S, Se and Te [31,32]. There are weak van der Waals forces that exist between the layers and covalent bonds within the layers. They possess outstanding electronic properties and large specific surface areas [27,32]. These materials have some fantastic characteristics that they exhibit when exist as mono or few layers. These properties have led to the much research for their application in the field of catalysis. They have the following advantages: (i) they are photosensitive and have the appropriate band gap for water splitting (ii) there is slow recombination of the photogenerated carriers when compared to common photocatalysts and (iii) they have high specific surface areas and many active sites, which become involved during photocatalytic process [14,17,29].

In spite of the above-mentioned merits they still have some drawbacks that have hindered their practical application: (i) they possess high exciton energy which does not favour high photocatalytic activities, (ii) many of the 2D semiconductor are prone to oxidation in ambient air or aqueous solution resulting to flocculation or photo corrosion, (iii) recombination though is minimal when compared to bulk and (iv) many of the 2D semiconductors do not have adequate oxidation and reduction potentials to achieve total water spitting [14,28,29]. Numerous methods have been used over the years to overcome these drawbacks, which include non-metal or metal doping, inducing defects, and forming a junction with semiconductors or metals [7,14,34,35].

The fabrication of heterostructure with TMDC is a facile process which is very flexible and without the need of reducing agents, strong acids or toxic oxidizing agents that are used in modifying a semiconductor catalyst with precious metals or graphene [16]. The fabrication of a proper 2D TMDC nanocomposite entails a right energy alignment between the promoter (TMDC) and the catalyst. This is known as energy band engineering. This provides a favourable avenue for flexible design and upgrade of both the electronic and optoelectronics characteristics. Classification of 2D TMDC-based composites is based on band alignment Z-schemes, Schottky junction systems, and the type I and type II heterostructure systems. The Z-scheme system is a parody of regular photosynthesis. The band alignment of type II is very close to that of Z-scheme except that the direction of electron transfer differs from each other. The Type II system is a p-n junction which made of p catalyst and n-type photocatalyst [14,16]. Figure 3 shows the mechanism of water splitting using heterostructure of a metal oxide (MOX) and TMDC. As the light shines on the TMDC, the photogenerated electrons from the conduction band (CB) of the TMDC transfer electrons to CB of MOX which reduces the H⁺ generated from the H₂O to H₂ and the protons generated from the MOX are transferred immediately to the valence band (VB) of the TMDC thus inhibiting fast recombination of electrons and holes which results in enhanced performance over pristine TMDC or main photocatalyst.

In relation with transition metal dichalcogenide (TMDCs) materials, the monolayer and few-layers of hafnium dichalcogenides have astonishing physical properties speculatively but these are rarely investigated [10,29,36,37]. Results have demonstrated that, the calculated room temperature mobility for HfS₂ and HfSe₂ monolayers, is about 5.3 and 10.3 times, respectively, greater than properly investigated MoS₂ [29,36]. They also exhibit greater sheet current densities much higher than of MoS_2; these remarkable qualities make them valuable materials in FETs application [33,38]_. HfS₂ and ZrS₂ have been proposed to be good photocatalysts based on density functional theory (DFT) calculations for water splitting [39,40]. Notwithstanding this eminence in theoretical projection, correspondingly few experimental results are available, presumably due to the complexity in syntheses methods or as a result of the oxophilicity of the nanomaterials [38,40,41]. Despite few published works, most of the studies use the exfoliation method or chemical vapour deposition (CVD) method for their fabrication [36,38,41,42,43,44]. The thickness and lateral dimension of the nanosheets cannot be manoeuvred using mechanical exfoliation method [41,43,44]. CVD is one of the favourite methods of producing these 2D TMDCs except for TiS₂ [42]. The method is economical, easy to operate and produces product with high quality. However, the CVD method still has some demerits. Ultrathin 2D TMDCs are mostly synthesized on substrates such as mica, sapphire, SiO₂ and hexagonal boron nitride (HBN) which requires them to be conveyed to other substrates for further application [10,32,41,43]. For application purposes it is better these nanosheets are in discrete form than to be produced on substrates [14]. Hence, solution method is still preferred to the CVD method for synthesis of 2D TMDCs. There are three common ways of preparing mono or few-layers TMDCs: (i) exfoliation, (ii) chemical vapour deposition and (iii) wet-chemical/solution based.

2.2.1. Exfoliation

This method of synthesis can be subdivided into two - mechanical and liquid exfoliation. It is employed in the fabrication of mono and few-layers of 2D layered materials. Mechanical exfoliation is a physical process which requires the use of the scotch-tape technique to separate off from the bulk of the material. The adhesive force of the tape helps to break the van der Waals forces that exits within the layers of the material. The isolated layer can undergo successive peeling to give mono and few-layers sample which is then deposited onto a substrate. This produces high quality single-layered nanosheets though with low output while liquid exfoliation is a solution-based method with great output. The first report on exfoliation of TMDCs from bulk samples was published in 2005 by Novoselov et al. Mono layers of MoS₂ and NbSe₂ were isolated from the bulk of the material [45]. Group IVB TMDCs have been prepared using mechanical exfoliation of the bulk. The group IVB TMDCs nanomaterials are easily decomposed in air and thus become oxidized [41,46,47,48,49]. To prevent their oxidation, exfoliation was carried out in a glove box and vacuum transfer chambers or immediate passivation with protective encapsulation layer was used in the synthesis [43,50,51]. Chae et al. fabricated few layers of HfS₂ field effect transistor on a SiO₂/Si substrate inside an integrated vacuum cluster system to prevent ambient oxidation. They realized uniform ambient oxidation of the HfS₂ material, preferentially at defect sites which resulted to thickness enlargement. The oxidized HfS₂ performed poorly as field effect transistor compared to the unoxidized sample [43]. In a similar report Kanazawa et al. synthesized few layer nanosheets from the bulk using scotch tape on single crystal HfS₂, the small piece that was obtained on the tape, was then cleaved numerous times to obtain thin film of average size of thickness on SiO₂/Si/Al₂O₃ substrate. The Al₂O₃ substrate surface was passivated with HMDS before exfoliated HfS₂ was transferred on it [52]. In another similar work by the same group, HfS₂ flakes were exfoliated on AlO₃ substrate, and the exfoliation method used suffered from the difficulty in manipulating the size and thickness of the fabricated TMDCs, as well as not being able to be scaled up for large batch production [46]_.

Liquid exfoliation can be further subdivided into two main types. The one type is simple and does not involve intercalation. The bulk sample is dispersed into the appropriate solvent or surfactant followed by exfoliation through a sonication process. The second type is a two-step process with intercalation preceding the exfoliation process in a solvent [44,53,54,55]. This method entails insertion of alkali metals into the bulk material with compounds such as LiBH₄, n-butyl lithium or organolithium compounds in solvent for 7–14 days at room temperature or at 100 °C for 3–4 days, succeeded by dispersing in an appropriate solvent. Care must be taken to ensure complete exfoliation, if not there is the tendency of generation of metal nanoparticles and Li₂S being precipitated during the process. The lithiated layered material will be recovered through filtration technique and thorough rinsing with solvent such hexane to eliminate lingering impurities of organic residues and alkali metal. Figure 4 shows this illustration.

The intrinsic properties of the dispersing solvent or surfactants is crucial and its role is to break the cohesive energy that exits in the layered material and also determines the exfoliation output [54,56]. The surface tension of the solvent needs to match the surface energy of the dissolved bulk sample. It is necessary to use the right solvent for dispersion of the bulk powder as this prevents re-stacking and aggregation of nanosheets in the solvent. Coleman et al. reported that experimental parameters such as the starting mass of the bulk material, sonication period, centrifugation conditions and nature of solvent determine the concentration, thickness, lateral and broad size of the exfoliated nanosheets produced. An increase in concentration was favoured by increasing the sonication time (200 h) for MoS₂, moreover, the nanomaterial had small lateral sizes and broad size distribution [53]. The solvents that are commonly used are isopropanol (IP), dimethylformamide (DMF), hexane, N-methyl-cyclohexyl-2-pyrolidone (NCHP) and N-methyl-2-pyrrolidone (NMP) [56]. N-methyl-2-pyrrolidone (NMP) is the most suitable solvent for fabrication of MoS₂ but due to its toxicity and difficulty in obtaining free standing nanomaterial after sonication, its application is limited. This has led to use of other solvents such as aqueous solution or volatile solvents. Coleman et al. used this method to prepare few-layers of some TMDCs and metal chalcogenides. For a solvent to be suitable for isolation of MoS₂, it was observed its surface tension should be about 40 mJm⁻² [53]. However, when water is used as a dispersing agent for a material that hydrolyses in water, the nanomaterial becomes oxidized. Traces of water have the tendency to oxidize group IVB TMDCs through hydrolysis, resulting to production of metal oxides (MOx) [41,43,46,49,55]. Figure 5 shows schematic illustration of exfoliation of ZrS₂ from the bulk sample.

Sherrell et al. gave an insight to the oxidation of TiS₂. The group synthesized TiS₂ nanosheet by insertion of alkali ions into TiS₂ powder, afterwards it was exfoliated in deionized water. It was observed that the exfoliated nanosheets were quickly destabilized by oxygen. The TiS₂ nanosheet suspension in water was oxidized to the oxide of titanium. It first generated TiSO species which was the intermediate product, at the same time H₂S gas was liberated to the environment. This was evident, as a colour change was observed in the suspension, it first turned to grey and then to white within 7 days of observation [53]. This method is so popular in the synthesis of single layers of MoSe₂, WS₂, MoS₂, WS₂, TaS₂, TiS₂ and ZrS₂ [56]. The obtained MoS₂ and WS₂ nanosheets using this method were observed to undergo phase transition, it translated from semiconducting (2H) phase to non-semiconducting (1T) phase. A thermal annealing process was needed to reverse it to the semiconducting phase [57]. The use of co-solvents has been developed to enhance the exfoliation process. For this, the use of Hansen Solubility Parameters (HSP) theory must come into play. Using HSP theory, a variety of mixed solvents have been explored to fabricate MoS₂. Zhang et al. exfoliated MoS₂ nanosheets using a co-solvent of water and ethanol. In a similar way MoS₂ was also effectively exfoliated using a mixture of chloroform and acetonitrile [58]. Kaur et al. synthesized few-layers of HfS₂ by dispersing the bulk HfS₂ in N-methyl-cyclohexyl-2-pyrolidone (NCHP) and ultrasonicated the solution. For the exfoliation in NCHP, the sheet formed was more stable in air as the solvent shielded the nanomaterial against ambient oxygen for a few days compared to using N-methyl-2-pyrrolidone and dimethylformamide as solvents [59]. Li et al. synthesized ZrS₂ nanosheets by dispersing the bulk of the powder in IP, the suspension was then sonicated to give few layers of ZrS₂ [47]_. Zeng’s group were the first to report on simplified lithiation using electrochemical lithium intercalation method to produce single layers of MoS₂, WS₂, TaS₂ and TiS₂ by proper adjusting of the amount of lithium intercalated, which was followed by exfoliation in ethanol or water [60]. In another related experiment, the same group executed a systematic study by manipulating circuit parameters such as voltage and current for the for production of few-layered inorganic compounds, such as TiS₂, TaS₂, WSe₂, ZrS₂, NbSe₂, BiTe₃, Sb₂Se₂ and BN. They optimized the parameters and produced high quality NbSe₂ and BN nanosheets [60,61]. Tandem molecular intercalation [TMI] is an improved method of exfoliation, which is a facile single step process that does not involve sonication and is operated under safe and mild conditions. This method makes use of Lewis bases (short and long alkylamines or alkoxides) intercalates, in which short initiator molecules will be the first to intercalate, followed by long primary molecules. Group IV and V TMDCs are better synthesized using weak Lewis bases such as alkylamine while for group VI TMDCs, alkoxide are more appropriate for their synthesis. Single-layered WSe₂ has been exfoliated by intercalation of an alkali ethoxide and alkali hexanoate in dimethyl sulfoxide (DMSO) with agitation lasting for several hours at room temperature. This process is well accepted for the synthesis of TiS₂, ZrS₂, NbS₂ and MoS₂ [62]. The use of surfactants or polymers has not been well explored such as the use of solvents presumably due to the high cost or toxicity of some of these surfactants. Sodium chlorate is the common surfactant that is being used, it helps to coat the sheets in dispersion thus preventing agglomeration [63].

2.2.2. Chemical Vapour Deposition

Chemical vapour deposition (CVD) is a methodology that involves decomposition/or chemical reactions of gaseous precursors through thermally induced means. The product is either formed on a substrate or without a substrate. CVD has generally been employed in the last decade as a bottom–up method for synthesis of various 2D materials, especially the [10,33,64,65,66,67,68,69,70,71]. In the synthesis, an inert gas (e.g., Ar) and H₂ gas are introduced, which help to eliminate oxidation of the material, at the same time reduce formation of impurities [10,29,42,65,70,71]. Figure 6 shows a schematic illustration of a CVD set up.

There are more reports on the use of chemical vapour transport (CVT) on the synthesis of group IVB TMDCs than mere deposition methods (CVD) [66,72,73,74]. CVT requires the use of halogens such as I₂ as the transport gas and is used for bulk single crystal, which takes days for the synthesis to be completed. The synthesized materials are then exfoliated into single or few layers sheets [73]. CVT is also used to synthesize transition metal trichalcogenides (MX₃) which are then subjected to pyrolysis to give MX₂ [42,74]. Wen et al. synthesized ZrS₂ nanoflakes by reacting a stoichiometric ratio of elemental sulphur (S) and ZrCl₄; S was added in excess due to its ease of evaporation at high temperatures. Various temperatures were evaluated, and in each case, the reaction was held for 1 h. ZrS₃ was formed and later decomposed to ZrS₂ on further heating. The optimal temperature was observed at 800 °C with no traces of impurities [71]. In a similar manner with little adjustments, Fu et al. synthesized HfS₂ by using HfCl₄ powder and S powder. Both the metal precursor and S powder were placed upstream of the quartz tubes and heated at a temperature of less than 200 °C while the substrate was inserted in the hot zone at a temperature of about 930 °C. The heating was carried out simultaneously for few minutes and then stopped. The unreacted precursors were immediately eliminated with the aid of magnets, after which the furnace was cooled down naturally [10]. An illustration of the synthesis is shown in Figure 7. Zheng et al. similarly synthesized HfS₂ nanoforest by placing the substrate in the hot zone; and the metal precursor and chalcogen precursors upstream. The heating was carried out at the same time, the temperature was set at 950 °C and 160 °C, respectively, operated for 10 min and then terminated. At the end of the reaction, the furnace was rapidly cooled down [75]. Zhang et al. comparably reported on synthesis of both ZrS₂ and ZrSe₂ by CVD at elevated temperatures above 800 °C. The reaction temperature was sustained for about 20 min and nanostructures of ZrS₂ and ZrSe₂ were grown on substrates [65]. Yan et al. synthesized HfS₂ nanoflakes using S powder and HfCl₄ or HfO₂ powder as precursors. The precursors were placed upstream and the substrate at downstream and the reaction was operated at 900 °C for just 10 min followed by cooling the furnace [29]. Shimazu et al. [66] synthesized a single crystal of ZrS₂ by heating Zr, S₈ and I₂ in a sealed evacuated quartz operated for 3 days at 800–900 °C. TiS₂ was also prepared by Gao et al. using three temperature zones in which Ti/NH₄Cl and S powder were placed at upstream where low temperature was applied, and the substrate placed at downstream operated at 450 °C [76]. CVD is not commonly used for titanium dichalcogenide synthesis, they are often produced through solid-state reaction method. A mixture of titanium and sulphur powders are blended before being transferred and sealed into an ampoule under a vacuum. It is then calcined using muffle furnace for 12 h at 500 °C. Afterwards the temperature is elevated to 800 °C and maintained for another 24 h after which the reaction is stopped and allowed to cool to room temperature [77,78]. Fabrication of high quality 2D ultrathin TMDCs is very difficult, this method does not allow easy manipulation of reaction parameters.

2.2.3. Wet Chemical Synthesis

This method involves the use of surfactants or polymers in the solution during the synthesis process. This method is much easier for fabrication of nanomaterials at low temperatures such as 130 °C and has a better control of kinetic parameters, in contrast with CVD which requires high temperatures of at least 400 °C and may run for several hours before the reaction becomes to completion [79]. The wet chemical synthesis method has four main variations which include colloidal, hydrothermal, sol gel and liquid exfoliation techniques.

(i)

Colloidal Synthesis

Colloidal synthesis provides favourable merits, such as easy to direct and proper grip of the crystallinity, monodispersity and control over the edges of TMDCs [27,57,80,81]. The method has been used to produce Quantum dots, metal nanoparticles and nanomaterials etc. When applied for the synthesis of TMDCs, reaction variables which include reaction time, temperature, nature of the metal precursor, chalcogen precursor and the type of ligand used are crucial in determining the shape and size of the nanomaterial that is formed [79,82]. This synthesis technique can be further subdivided into two; injection and one pot synthesis [79,83,84]. Injection synthesis is often used when either or both the reactant(s) is/are solvent(s). The precursor that is in solid form is first dissolved in an organic solvent (ligand), then purged for about 20 min. The temperature of the system is raised at a controlled rate to a predetermined temperature, at this point the injection of the other precursor is introduced into the hot system by means of a syringe under vigorous stirring. The injection may be rapid or slow depending on the nature of the product required. One-pot synthesis is a non-injection technique which is commonly used when both metal and chalcogen precursors are solids. In this case the reactants are first mixed with the surfactant at room temperature before heating is introduced under inert conditions. Ligands that are commonly used in colloidal synthesis of TMDCs include oleylamine (OLA), 1-hexadecylamine (HAD), oleic acid (OA), oleylalcohol (OYA), trioctylphosphine oxide (TOPO), dodecylamine (DDA), squalene, 1-dodecanethiol (1-DDT), stearic acid, octadecanamine (ODA) and 1-octadecene (ODE) [81,82,83,84]. The use of ligands in colloidal synthesis helps in controlling the morphology during the synthesis of semiconductor nano-/micro- crystals by coordinating to the surface of the growing nanoparticles [84]. This method of synthesis is quite popular because of its simplicity. The use of ligands has eased the synthesis of hierarchical structure-based morphologies such as comb–like, disk–like, dendrite-like, snowflake-like, rod-like, flower–like and urchin–like structures, which show unique properties by combining the features of micrometre and nanometre building blocks in one crystal [84,85].

Colloidal method has been employed in the synthesis of single or few-layers TMDCs. Caution is needed in the synthesis of group IVB TMDCs due to the oxyphilic nature of single or few-layers nanomaterials, to prevent contamination of the nanomaterials with metal oxides. There is a need to strictly avoid water or oxygen in the preparation of the precursor solution. The chlorides of group IVB (TiCl_4, ZrCl₄ and HfCl₄) are commonly used in the synthesis of these nanomaterials. These metal precursors are very hygroscopic; thus, they become hydrolysed while weighing in ambient air. Hence, weighing should be carried out in a glove box. Prabakar et al. [86] reported on the simple colloidal synthesis of hierarchical structures of TiS₂ using hot injection method by dissolving elemental S in non-coordinating ODE and at 300 °C, TiCl₄ was injected and heated up for 15 min. Then the metal precursor was injected into the system at a lower temperature of 150 °C to produce a flake-like structure in contrast with the flower-like structure that was observed at the higher temperature. In 2008, Park et al. [87] synthesized mono layers of nano-disk TiS₂ using OLA in dried form; S powder was dissolved in OLA at 110 °C, the mixture was brought to room temperature, after which titanium was injected and the temperature slowly raised to 215 °C under argon gas, the reaction lasted for 12 h. The sample obtained was vacuum dried. On exposure of the nanomaterial to air, the colour changed from black to brown; however, when it was kept refrigerated under nitrogen, the colour remained unchanged. The thickness of the TiS₂ nano-disk was about 0.6 nm and a lateral size of 50 nm. Increasing the concentration of the chalcogen atom with a rapid rate of temperature increase helped control the lateral size of the nano-disks. When the chalcogen concentration was increased by 100%, decrease in lateral size was observed from 50 nm to 34 nm. At the same time when the concentration of sulphur was increased by a double fold, there was a reduction in the lateral size. However, the synthesized TiS₂ single layer nano-disks were easily destabilised at room temperature, the S atom was displaced by oxygen atom in air. The authors were able to prove this using time independent energy dispersive spectroscopy (EDS) and powder XRD. In a related work, Cheon’s group synthesized TiS₂ by dissolving TiCl₄ in OLA followed by injection of CS₂ into reaction mixture. The reaction was maintained for 15 min at 300 °C. Increasing the concentrations of both metal and chalcogen precursors was so pivotal in controlling the lateral size of nanomaterial formed. When the concentrations were increased to 2.4 times it produced TiS₂ of lateral size of 40 nm. The group changed the S source to S powder, one pot synthesis was deployed to produce TiS₂ nanocrystal. The authors preferred the use of CS₂ over elemental S based on production of highly reactive radicals by S powder which resulted in structural degradation of the nanocrystals produced. Presumably it is better to use chalcogen that contains carbon for the synthesis of group IVB TMDCs as the carbon can react with any traces of metal oxides formed during synthesis. The method was also employed in the synthesis of TiSe₂ by using elemental Se as source of chalcogen [27]. Not much has been reported in the open literature on group IVB, particularly on hafnium dichalcogenides. Few reports are available on colloidal synthesis of HfS₂. Cheon’s group reported on its synthesis, in which the reactants HfCl₄, 1-DDT and OLA were heated up in a reacting flask in an inert atmosphere at 245 °C for a duration of 10 h [88]. In a similar work the group also used CS₂ as a S precursor at a higher temperature of 320 °C [88]. The same group also used a resembling synthesis procedure to synthesize sulphides and selenides nanosheets of the IVB and VB groups. To date, the studies by the Cheon’s group are the only available on ZrS₂ synthesis [27,88,89]. They distilled and degassed OLA prior to use to purify it. However, they did not mention the oxyphilicity and the ease of oxidation of group IVB TMDCs. In 2014, Cheon’s group demonstrated the use of slow decomposing chalcogen precursors (1-DDT) and H₂S gas in the generation of single–layers nanosheets of Group IVB metal sulphides. At high temperatures (over 150 °C) 1-DDT releases S atoms gradually. In the synthesis of HfS₂ nanosheet a combination of HfCl₄, OLA and 1-DDT were heated up in a reacting flask in an inert atmosphere at 245 °C for 10 h. The 1-DDT slowly decomposed to H₂S during the entire synthesis time. Using the same procedure while differing the temperature and time of synthesis, ZrS₂ and TiS₂ nanostructures were also produced. The group also used H₂S gas in a similar set-up; the H₂S was released slowly over 6 h and monolayers of TiS₂, ZrS₂ and HfS₂ were obtained. A high level of precaution needs to be exercised when using H₂S due to its toxicity which can be fatal. The nanocrystals obtained were of a poorer quality compared with the use of 1-DDT [89]. Different S precursors can be used, such as elemental S, 1-DDT, CS₂, diphenylurea, thiourea and thioacetamide. CS₂ and 1-DDT can be systematically introduced into the heating system using hot injection protocol while other precursors that are in solid form can be dissolved in the solvent prior to heating up with the metal precursor. CS₂ is often used for hot injection as a result of its in-situ hydrogen sulphide generation. Metal chloride is commonly used as metal precursor and there is a need to wash the synthesized nanomaterial thoroughly due to lingering impurities such as chloride ions. Similar to group IVB not much work has been reported for group VB. Sekar et al. was the first to report on synthesis of group V TMDCs. The group synthesized NbSe₂ nanowires using one pot synthesis. Into a three-neck reaction flask, a mixture of OLA, DDA and precursors (NbCl₄ and Se) were introduced. The temperature of the reaction was increased to 280 °C under N₂ atmosphere and sustained for 4 h. The reaction vessel was not cooled down and nanomaterials obtained was rinsed with hexane. It was then subjected to heating under N₂ atmosphere at 450 °C for a period of 3 h. The resulting NbSe₂ wires had diameters varying from 2–25 nm. While cooing reaction vessel to room temperature before washing, nanoplates of NbSe₂ were obtained. The nanoplates had lateral dimensions that ranged from 500–1000 nm and thickness between 10–70 nm. The authors also varied the reaction parameters in order to achieve different sizes or morphologies, but no changes were observed [90]. Mansouri and Semagina reported on the synthesis of NbS₂ using a mixture of ligands to produce different morphologies varying from nanosheets, nanospheres, nanohexagons and nanorods. By increasing the time, sulphur precursor and coordinating ligand, nanomaterials with thickness of <2 nm and reduction in lateral dimension was observed. Increased amounts of OA in OLA led to more production of the nanosheets, but better morphology and laterally confined 2D nanostructures was obtained with minimal use of OA. In addition, of great importance is the timing of the reaction, with 0.25 h of reaction, monolayer of NbS₂ nanosheets were formed with OLA as the sole solvent and as the reaction time was increased few-layers of nanosheets were formed [80]. Han et al. synthesized NbS₂ by a one pot synthesis method in which stoichiometric ratios of NbCl₅ and 1-DDT were introduced into three-neck reaction flask containing OLA in a glove box. The use of glove box indicated that the nanomaterial is sensitive to impurities (O₂/H₂O). The reaction was operated for 30 min at elevated temperature of 280 °C after which it was cooled down. The authors do not mention if the reaction was conducted in an inert atmosphere. The synthesized ultrathin triangular NbS₂ nanosheets had thickness of 3.9 nm which was ascribed to represent five layers [91].

A great number of research has been reported on group VIB TMDCs, most especially on MoS₂ which has been effectively illustrated in electrocatalysis for the hydrogen evolution reaction (HER) which is being suggested as possible replacement for platinum as well as its role as proficient co-catalyst in photocatalysis [7,14,92,93]. Mahler et al. [94] synthesized both prismatic 2H-WS₂ and distorted octahedral 1T-WS₂ structures. Hot injection reaction method was employed, the metal precursor was added to OA in a vial, and injection was made into reaction flask containing OLA at 320 °C dropwise for about 0.5 h. Prior to attaining this temperature, CS₂ was injected into the system. This resulted in controlled monolayers by the slow release of the precursors. The addition of hexamethyldisilane (HMDS) after degassing helped in tuning the crystal structure of the nanosheets from prismatic to distorted octahedral structure which was flower-like in shape. In a similar report, Geisenhoff et al. [95] synthesised WSe₂ using the hot injection method whereby tungsten hexacarbonyl was dissolved in combination with TOPO and OA and selenide precursor. The mixture was injected at 330 °C and the process was completed in half hour. The ligand mixture helped to adjust the precursor reactivity and an increased amount of OA limited the metal precursor reactivity, resulting in fewer nucleation and thus bigger nanocrystals were formed. Lin et al. [96] prepared MoS₂ quantum dots by dissolving the single-source precursors, ammonium tetrathiomolybdate ((NH₄)₂MoS₄) in three different capping agents, OLA, OA and ODE. The reaction mixture was purged at 120 °C under vacuum for 2.5 h with stirring. Afterwards the reaction was sustained for 3 h at temperature of 250 °C under N₂ atmosphere before the reaction was quenched. In a related experiment Altavilla et al. prepared both MoS₂ and WS₂ nanosheets using thio-salts of Mo and W by one a pot synthesis method. The single-source precursor OLA was first degassed, and then the temperature was raised to 360 °C under N₂ flow for 0.5 to 15 h. Interestingly, as the reaction time increased so did the number of layers produced [97]. Figure 8 illustrates the steps in the colloidal synthesis of MoS_2.

Antunez et al. in a similar experiment produced Wse₂ nanosheets by injecting di-tertbutyl diselenide (tBu₂Se₂) into a reacting vessel already containing WCl₄ in DDA at temperature above 100 °C under N₂ gas. The reaction lasted for 6 h under strong magnetic stirring at a temperature of 225 °C. WCl₄ is not a suitable metal precursor for the synthesis of group VI TMDCs, due to its ease to hydrolyse in air, thus the sample was introduced into a three-neck round-bottom flask in a glove box. It is well established that WCl₄ is easily reduced in the presence of organics if overheated, hence the temperature during heating must be well controlled. The organic solvent (DDA) was not used as supplied but was deoxygenated and distilled before use [98]. Jung et al. in 2015, using a one pot method synthesized Wse₂ monolayer nanosheets. W(CO)₆ and diphenyl diselenide were dissolved in OA, the system was degassed, and the temperature was raised to 330 °C, the reaction was operated for 12 h. Monolayer nanosheets with lateral dimensions of several nanometres were produced. The group also worked with other surfactants such as OYA and OLA. With OYA few-layers (2–3) with lateral size of few nanometres were obtained while with OLA multi-layers with smaller lateral size were formed [99]. Zhou et al. working with mixed surfactants, explored the influence of mixed solvents to produced different layers of the nanostructures of MoS₂ and WS₂ using injection protocols. The authors used Mo(AC)₂ and W(CO)₆, the surfactants stearic acid, TOPO, ODA and squalene under N₂ atmosphere with stirring. The duration of heating was just 1 h before being quenched. In a similar reaction MoS₂ monolayers were also generated but with a change of surfactants (OA, stearic acid and TOPO). By increasing the concentration of the chalcogen source and decreasing the concentration of the metal as well as increasing the temperature of the reaction multi-layers (3–5) of WS₂ nanosheets were formed. With little modification both thioacetamide and Mo precursor were injected into the reaction system at different times to produce few and multi-layers MoS₂ nanosheets [100].

The mechanism for the formation of single-layered nanosheets of TMDCs is still complex. Three variables have been employed to optimise their formation; firstly, the use of chalcogen precursor that gradually decomposes over a long-time during synthesis or delay injection of reactant for long period and secondly, the nature of the organic solvent and lastly the duration of the reaction have been varied.

(ii)

Hydrothermal or solvothermal synthesis

Hydrothermal or solvothermal method is a versatile and effective synthetic route to produce the nanomaterials with different array of morphologies. Hydrothermal synthesis is one of the most important methods for producing fine powders of metal oxides. The process entails the reactants being dissolved in a solvent, which is then introduced into an autoclave. If the solvent for the reaction is non-aqueous, it is referred as solvothermal; whereas, if the solvent for the reaction media is water, it is termed hydrothermal [101]. Teflon-lined autoclaves are used in this process; they are preferred over glass and quartz autoclaves, since they can tolerate high temperatures and pressures. Furthermore, they support alkaline solutions as well as are resistant to hydrofluoric acid. The flexibility of the method makes it easier to manipulate reaction parameters to produce nanomaterials with desired properties and quality. This method is very appropriate for production of products with different array of shapes in contrast with other methods. It has been extensively used for preparing metal oxide nanoparticles, chalcogenide and phosphide nanomaterials [102,103,104]. Chen and Fan [102] synthesized NiSe₂, NiS₂, CoS₂, FeS₂, MoSe₂ and MoS₂ using hydrothermal synthesis at a low temperature while varying the synthesis parameters. They found that adjusting the reaction variables could extend the method of synthesis to other products. The method is commonly used for the synthesis of group VIB TMDCs and their nano composites [34,105,106,107,108,109]. Huang et al. [105] synthesized MoS₂ nanosheets with a net-like morphology of well linked nanoflakes, the nanoflower material had thickness of a few nanometres for use as capacitor electrode materials using hydrothermal process. The prepared material had a large surface area and good conductivity giving it potential for application in high-performance supercapacitors. Using this synthesis method, Swain et al. [106] synthesized flower-like MoS₂ and composite of MoS₂-CaIn₂S₄ for photocatalysts for hydrogen evolution reaction. The composite catalyst showed good hydrogen evolution performance due to efficient charge separation efficiency. Jang et al. [107] also prepared CdS nano particles using hydrothermal method as well as CdS nano wire/TiO₂ nano particle composite using solvothermal method to produce effective photocatalysts. The composite catalyst configuration resulted in proper charge separation as a result of quick migration of photoelectrons produced from CdS nanowire along the vicinity of TiO₂ nanoparticles resulting in higher generation of hydrogen.

(iii)

Sol-gel method

The sol-gel method is a popular avenue to fabricate metal oxide catalysts such as oxides of Ti and Si. Composites. A variety of materials which includes nanostructures, nanomaterial nanoparticles, glass, ceramics, and nanocomposite are generally fabricated using this method. This process generally takes place in three steps viz: hydrolysis, condensation, and drying. Sol-gel can be sub-divided into two types: aqueous sol-gel and non-aqueous sol-gel method [101,108]. To synthesize these colloids, the common precursors are made of metal alkoxides and alkoxysilanes. The use of tetramethyoxysilane (TMOS), and tetraethoxysilanes (TEOS) is most common. There is a need to first make a homogeneous solution of the alkoxides to be used [108]. This method is very suitable for production of group VIB TMDCs and their nano composite [109,110,111,112,113,114]. The method offers some merits such low synthesis temperatures, high reproducibility, cost effectiveness and products with high purity, high porosity, and large surface area [115]. A one-step direct sol-gel synthesis method was employed to prepare p-type few-layer MoS₂ films in a large volume via deployment of Mo-containing sol-gel including 1% tungsten [116]. It is very functional and good for large production using spin coating deposition method on variety of substrates to produce 2H-MoS₂ thin film having uniform surface areas at moderate temperatures (300–400 °C) followed by annealing. The thin films MoS₂ produced was of good quality and had great electronic properties with a narrow energy band gap of 1.35 eV which is consistent with the material. The product is of n-type semiconductor which find application in electronic devices [111].

3. Application of Transition Metal Dichalcogenides in Photocatalytic Hydrogen Evolution

2D layered transition metal dichalcogenides have received an enormous amount of attention for their magnificent catalytic activities, narrow band gap, crystallinity, excellent properties and their use in catalysis for production of hydrogen gas [7,14,117,118,119]. They are being sought for in recent time as a result of their low cost, earth abundance and excellent catalytic activities as a possible replacement for scarce and expensive noble metals [7,14,118,119]. The transformation that occurred in the material from 3D to 2D paved the way for their novel electronic and mechanical properties. There are many active sites in 2D chalcogenides which participate during catalytic reaction and the possibility of harnessing solar energy by the atoms of the semiconductor. The chalcogenides possess right and adjustable band gap depending on the number of layers they possess. To illustrate this, the band gap of ZrS₂ may be increased from 1.5 to 2.0 eV if the bulk ZrS₂ is exfoliated to single layer of ZrS₂, hence, there is changes in the band gap energy as it moved from an indirect to a direct band gap by reason of quantum confinement [119]. A lot of these 2D chalcogenides have been proven to be good photocatalysts for water splitting, including MoSe₂, MoS₂, WS₂, NiS, NiSe, Wse_2, SnS₂, and ReS₂ [11,28,85,116,117,118]. Some other semiconductors, the likes of HfS₂, ZrS_2, TcSe_2, and TaS₂ are being anticipated as photocatalysts to be considered for decomposition of water [14,39,40].

Composite photocatalysts have been used to improve photocatalytic processes by reason of their excellent configuration. To enhance the properties of these, TMDCs are integrated with other photocatalysts such as TiO₂, ZnO, CdS, CuS, g-C₃N₄ etc. There must be an intimate interface between the semiconductor and the co-catalyst which will accelerate electron transfer within the interface of the semiconductor to the co-catalyst. The design and assembly of an exclusive junction between the photocatalyst and the co-catalyst plays a vital role for enhanced photocatalytic performance [14,17]. There are reports on the utilization of MoS₂ as co-catalyst by several authors to greatly enhance the activity of photocatalysts. Table 1 shows a summary of composite photocatalysts that have been employed in hydrogen evolution reactions. Ma et al. [120] constructed a CdS-MoS₂ hybrid catalyst for photocatalytic production under visible light, by ultrasonication treatment. The presence of MoS₂ greatly boosted the production of hydrogen which was about two times greater than when CdS alone was used. However, due to weak interface between them, the photocatalytic activity decreased (about 35%) after four cycles which was attributed to ease of separation of MoS₂ from CdS. CdS is a well-known photocatalyst, but it has the drawbacks of photo corrosion and great toxicity of Cd to the environment. Chang et al. [121] demonstrated that the number of layers of TMDCs played a key role in a catalytic reaction. In their study for photocatalytic hydrogen evolution reaction using MoS₂ loaded on CdS in the presence of NaS-Na₂SO₃ and lactic acid as sacrificial agents, the yield of hydrogen with lactic acid was higher than with NaS-Na₂SO₃, however the catalyst was more stable in Na₂S-Na₂SO₃ compared to lactic acid. This was expected since lactic acid solution contains abundant H⁺ to Na₂S-Na₂SO₃. The authors do not report on the number of cycles conducted on the catalyst to determine its stability. They revealed that mono, or few layer(s) of MoS₂ produced maximum generation of hydrogen compared to higher numbers (7–122) of layers. The yield of monolayer was the optimum generating 0.00259 mmolh⁻¹ and 0.00201 mmolh⁻¹ H₂, respectively, and this was associated to the following 3 factors: (a) the mono layer displayed a more negative conduction band minimum to H⁺/H₂ potential (b) the edge sites are very pivotal to the hydrogen production and the mono-layer has plethora exposed active sites arising from the unsaturated (S) atom to the other layers (bulk) and (c) delayed rate of recombination of photogenerated carriers was observed in mono-layer compared to bulk which was attributed to good interface contact between mono MoS₂ and CdS. The yield of MoS₂/CdS was 5.89 times higher than Pt/Cds conducted in lactic acid. However, there is possibility of carbon monoxide generation which may lead to Pt catalyst poisoning, thus a reduced production using Pt/Cds in lactic solution. Very recently, Zong et al. [122] fabricated CdS-MoS₂ by conventional impregnation method. The H₂-evolution rate was 36 times higher when 0.2 wt% MoS₂ was loaded on CdS. The enhanced photocatalytic activity was attributed to the tight junction formed between MoS₂ and CdS and the excellent H₂ activation property of MoS₂. In similar work, Zong et al. [123] also fabricated CdS-WS₂ by impregnation method. The loading of 1% weight of WS₂ increased hydrogen evolution rates by 28 times attributed to the interface formed between CdS and WS₂ and the excellent performance of WS₂ as a co-catalyst in catalysing H₂ evolution. CdS is a well-known photocatalyst, but it has the drawbacks of photo corrosion and great toxicity of Cd to the environment. In a recent work, Reddy et al. [124] accounted for the outstanding HER activity of a nanohybrid which was reported to be highest using a ternary catalyst compared to the one available in literature. The group constructed heterostructures by loading few-layered ultrathin MoS₂-WS₂ on CdS and applied the catalyst for photocatalytic water splitting using lactic acid as holes acceptor. The HER performance of 209.79 mmolg⁻¹ h⁻¹ was obtained which was 1.70 folds greater than Cds-MoS (123.31 mmolh⁻¹), 1.70 folds higher than Cds-WS₂ (169.82 mmolh⁻¹), 6 folds higher than CdS-Pt and 83 times higher than bare CdS (2.54 mmolh⁻¹). The 6 wt% MoS₂-WS₂ loading was the most effective which showed robust activity, durability, and stability even up to 2.5 days. The high-level activity was attributed to the following reasons: (i) more actives sites were available by MoS₂-WS₂ cocatalyst in contrast with a single cocatalyst (ii) proper band gap of the nanohybrid (iii) tight heterojunction between the catalyst and cocatalyst which promoted ultrafast electron transfer to the cocatalyst, as well tardy re-joining of the photogenerated carriers. Swain et al. [106] synthesized MoS₂ nanoflowers and used them to decorate CaIn₂S₄ microflowers using double step hydrothermal method. The experiment showed that hybrid catalyst generated higher H₂ production (19 times) compared when the CaIn₂S₄ was used in a visible light irradiation. The catalyst showed a band gap of 2.11 eV which enabled the reaction to be conducted in visible region of light compared to UV-visible light. The MoS₂ provided more edge sites having plethora of unsaturated sulphur ions which facilitate quick capture of the H⁺ ions at the interface of the p-n heterojunction of the catalyst. This activity enhancement was as a result of adequate prevention of quick reoccurrence of electron-hole pair at interface of the hybrid catalyst. This work showed that MoS₂ is a possible replacement to other promoter on a semiconductor catalyst with just 0.50% loading. With increasing loading as low as 1% the output of H₂ production declined due to the black colour of MoS₂ which prevented the transfer of photons. The catalyst also generated H₂ even in the absence of sacrificial agent though with a lower yield. Stability study showed the catalyst to be stable even after 4 cycles. The rate of H₂ production generated in the study was a great improvement over previous studies on CaIn₂S₄ based composite for photocatalytic water splitting as the yield is far higher compared to when other catalysts that were loaded on the semiconductor photocatalyst. Nguyen et al. [125] decorated Zn_xCd_1-xS with MoS₂ via a photo-assisted deposition. The 3% decoration of Zn_0.2Cd_0.8S/MoS₂ gave H₂ evolution 210 times higher than the undecorated photocatalyst. Huang et al. [126] constructed 2D/2D ZnIn₂S₄/MoS₂ nanohybrid using electrostatic self-assembly method. Generally, due to the possibility of shading effect as a result of excess loading of the co-catalyst on the catalyst, low loadings of MoS₂ are required. The 0.75% loading gave the maximum hydrogen yield which was almost 50 times higher than the raw ZnInS₄. Only traces (0.099 mmol h⁻¹) of hydrogen were generated when pristine ZnIn₂S₂ was used for the photocatalytic reaction. However, when MoS₂ was loaded on it, the hydrogen yield increased linearly with increase amount of the promoter until 0.75% after which the yield deceases due to the shading effect. The result was also compared to 0.75% Pt/ZnI₂S_4. The production of H₂ was 2.2 lesser with noble metal as a promoter. The reaction was visible light-driven possibly due to its band gap that is proper for visible light, however the authors did not report on the adsorption edges or band gap of any of the catalyst used. The structure of a catalyst determines its efficiency in a photocatalytic reaction. The high yield could be as a result of the ultrathin structure of both MoS₂ nanosheet and ZnIn₂S₄ nanosheet, and the robust and close contact interface in the hetero layered enhanced the photocatalytic hydrogen generation. The fabrication of ultra-thin 2D/2D structure performs better compared to 0D/2D and 1D/2D counterparts. The catalyst was reported to be stable until fourth cycle of use in which a decrease in the activity was noticed. Pudkon et al [127] synthesized ZnIn₂S₄ using different sulphur sources and loaded it on WS₂ using microwave method of synthesis. The L-cysteine not just being a sulphur source but played the dual purpose as a reducing agent and as capping agent and helped in growth direction for the formation of flower-like nanomaterials. The HER activity of the hierarchical (flower-like) catalyst (145.3 µmolh⁻¹) was far more than the non-hierarchical (81.6 µmolh⁻¹) catalyst. The former provides diverse reflections of the incident light via its shape thus a lengthy lifetime of the incident light in its structures. This can be credited to the surface area as well, surface area is an important factor in a catalytic reactions, the higher the surface the more active site will be available for reactions to take place. The nanomaterial with flower-like morphology had a surface area 1.6 folds higher than non 3D nanomaterials fabricated. The 40%wt loading of WS₂ on flower-like catalyst was most suitable for HER with activity 2 times higher (293.3 µmolh⁻¹) than the pure ZnIn₂S₄ and 6.67 folds than raw WS₂. The close contact contributed to good separation of charge and injection at the interface of the catalysts whereas the composite formed by mere grinding of the two catalysts gave activity of (101.3 µmolh⁻¹) which was even less than activity of untreated flower-like ZnIn₂S₄ alone. Both ZnInS₄ and WS₂ are photocatalysts that absorb within the visible region of solar spectrum. The authors did not report on the band gap of the composite however, the band edge obtained from the UV-absorption showed a slight shift to higher wavelength which is an indication of a better response of the composite to visible light. However, in their report the composite catalyst activity for H₂-evolution was more favoured in UV-visible compared to visible region up to about four folds. This may be accounted for by the band gap of the semiconductor synthesized; it had a band gap of 2.81 eV, close to that of a material that absorbs in the UV region, and the promoter did not significantly improve the response of the composite to visible light. The output of hydrogen obtained was slightly lower than previously reported on ZnInS₄ based heterostructure demonstrated for H₂ production. Photocatalysts that absorb in the visible region are required for visible light driven photocatalysis and they have their band gap very close to 2.0 eV [34]. Zeng et al. [128] in a manner similar to constructed metal-sulphide -metal-selenide hybrid (ZnIn₂S₄-MoSe₂₎ system for HER using a one-pot polyol method. The synthesized ZnIn₂S₂ had hierarchical structures which contributed to the increased HER activity of the untreated catalyst of 1023 µmolh⁻¹ in contrast with previous work in which pristine ZnIn₂S₄ was used for photocatalytic water splitting. The 2 wt% loading on the hierarchical structure ZnIn₂S₄ improved the activity by a multiple of 2.2 times. The stability test on the catalyst showed good stability even after five cycles. The result obtained with MoSe₂ as cocatalyst on a ternary chalcogenide is very comparable to results from previous authors with MoS₂ loaded on the same material. This is an indication the activity of MoSe₂ as a promoter is as well good as MoS₂ in photocatalytic water splitting.

A lot of work has also been reported on the improvement of the photocatalytic activity of TiO₂ with TMDCs. Li et al. [129] synthesized MoS₂/TiO₂ hybrid by hydrothermal treatment and found the MoS₂/TiO₂ hydrogen evolution yield was 2.19 and 3.15 times higher than TiO₂ and MoS₂, respectively. The yield was still good after 4 cycles of use, this as a result of close interaction between the co-catalyst and the photocatalyst. Recently, Zhang et al. [130] fabricated novel mesoporous anatase titania with high photoconductivity and photocatalytic attributes compared to nanocrystal TiO₂ such as P25 using hydrothermal method. MoS₂ nanosheets was loaded on the 3D TiO₂ using same hydrothermal process and evaluated for hydrogen evolution reaction under UV light source for irradiation. The composite showed higher photocatalytic activity for hydrogen evolution compared to undecorated TiO₂ and MoS₂. The mesocrystal of the TiO₂-MoS₂ composite produced an H₂-evolution rate of 0.55 mmolh⁻¹ which was 4 times greater than MoS₂-TiO₂ (P25) hybrid. The performance of MoS₂ as a promoter on the semiconductor in work is an indication of possible replacement over precious metal such as Pt and Pd. The 1% MoS₂ loading resulted in 200 times higher H₂ evolution compared to the bare mesocrystal TiO₂. However, previous reports showed that a maximum loading of about 2% MoS₂ on semiconductor resulted in enhanced photocatalytic water splitting contradict the report in this work where up to 10% of MoS₂ was the optimal loading for high hydrogen production. The main factor was identified as annealing process. Annealing the catalyst at about 160 °C greatly improved its performance towards HER since phases of TMDCs contributes immensely to the catalyst activity towards HER. The annealing caused some phase transition of the MoS₂ from metallic phase to mostly hexagonal phase. Moreover, at annealing temperatures exceeding 240 °C the activity of the composite declined possibly due to the sulphur active sites being oxidized or loss of the metallic phase. The catalyst also showed good stability after five test cycles. In another report, Zhou et al. [100] fabricated MoS₂-TiO₂ heterostructures by hydrothermal method. MoS₂ loaded on TiO₂ nanobelts and the resulting photocatalyst exhibited outstanding H₂-evolution to a value of 1600 µmolh⁻¹ compared with pure TiO₂.

Dong et al. [131] also made TiO_2-MoS₂ heterojunction via annealing impregnation deposition method. The hybrid generated H₂ evolution with activity 7.2 times and 17.6 higher than the bare TiO₂ and MoS₂, respectively, ascribed to that the heterojunction between TiO₂ and MoS₂ quantum dots which could suppress the recombination of electron-hole pairs efficiently and that the hollow TiO₂ nanospheres and MoS₂ quantum dots with high specific surface area and pore diameter could supply plenty of active sites. Moreover, the authors did not experiment on different loading to determine the optimal loading and the percent of co-catalyst used was not reported and more so, the recyclability test was not reported. Feng et al. [132] fabricated quaternary nanocomposite of Mn-Cds-MoS₂-TiO₂ by hydrothermal method. The quaternary nanocomposite gave outstanding H₂-evolution up to a level of 408.27 mmolh⁻¹ which was 30.08 folds that of TiO₂, 5.18 folds of TiO₂-MoS₂ and 2.52 folds for CdS-Mn-TiO₂, respectively. The photocatalyst performance was greatly improved which was attributed mainly to the synergetic effects of CdS-Mn, MoS₂ and TiO₂, forming a Z-scheme system in the CdS-Mn/MoS₂/TiO₂ electrode, which not only accelerates the interfacial charge transfer efficiency but also preserves the strong redox ability of the photogenerated electrons and holes. This is one of the highest activities reported so far for a catalyst for photocatalytic water decomposition. MoS₂ has also been used to improve the photocatalytic properties of ZnO. Yuan et al. [133] loaded MoS₂ on ZnO via hydrothermal route. The loading of MoS₂ as cocatalyst on ZnO elevated its H₂-evolution rate up to 14.8 folds than the pristine ZnO. The optimum loading of 1% MoS₂/ZnO generated 768 µmolh⁻¹ of H₂. The activity enhancement was due to the suppression of the recombination of electron hole pairs of ZnO. Li et al. [134] constructed the nanohybrid of MoS₂-ZnO by dispersion method. The deposition of MoS₂ enhanced the hydrogen evolution 5 times compared to pure ZnO because of the production of more electrons and holes and reducing their recombination. The hydrogen output activity was 27.69 mmolh⁻¹. The method used to prepare to prepare the nanostructure did not produce a thorough interface between the two catalysts thus a reduced activity was noticed, and the recyclability of this catalyst was very difficult. Very recently similar research was conducted by Chang et al. [135] in which they constructed p-n MoS₂-ZnO heterostructures with large surface area using hydrothermal method. The nanohybrid with 0.03 g of MoS₂ yielded the highest activity (145.6 µmolh⁻¹) after which the performance declined with further loading. However, the activity reported was 5.3 times less than what Yuan et al. obtained [133]. Chang’s group [136] also investigated the photocatalytic water splitting activity of MoS₂-SnO₂ core-shell sub microsphere using Na₂S as hole collector, the yield was slightly lower when compared to MoS₂-ZnO (117.2 µmolh⁻¹). Oxides of bismuth are good semiconductors which find application in photocatalytic processes. Khalid et al. [137] constructed nano composite of Bi₂O₃-MoS₂ by hydrothermal method with varying composition from 0–15 wt% of MoS₂. The highest HER was obtained with 11 wt% with activity (10 µmolh⁻¹) which was ten folds higher than raw Bi₂O₃ and pure MoS₂. The ability of the catalyst to absorb much visible light and delayed recombination of electrons and holes led to higher production of H₂.

The use of CeO₂ has not been explored in photocatalytic water splitting until Swain et al. [138] successfully constructed MoS₂-CeO₂ using hydrothermal approach. The best output was when 2 wt% of the promoter was deposited on CeO₂ to form p-n junction with large intimate and close junction with performance of 508.44 µmolh⁻¹ which is far higher (57 times) than unloaded CeO₂. The catalyst still showed good stability even after 3 cycles.

Hollow graphitic carbon nitride (g-C₃N₄) has been a photocatalyst widely used in hydrogen evolution. Both g-C₃N₄ and MoS₂ have comparable layered structures which should help to reduce the lattice mismatch and ease the planar growth of MoS₂ layers on the g-C₃N₄ surface. Hou et al. [139] loaded MoS₂ on mesoporous g-C₃N₄ by impregnation method followed by sulfurization. The nanohybrid played a vital role in boosting the photocatalytic activity to 310 µmolh⁻¹ with 0.5 wt% compared to when pristine g-C₃N₄ (108 µmolh⁻¹) and untreated MoS₂ (0.25 µmolh⁻¹) were used where the activities were much lower. The group [139] in a similar reaction also formed heterostructure of WS₂-C₃N₄ using impregnation system followed by sulfurization. The 3 wt% loading gave the highest activity (20.60. µmolh⁻¹) while with unloaded C₃N₄ the activity was very minimal. In a related research, Ge et al. [140] synthesized composite of MoS₂-g-C₃N₄ by impregnation route. The 0.5 wt% MoS₂-g-C₃N₄ gave the maximum catalytic activity with H₂ evolution of 23.10 µmolh⁻¹ which was 11.3 times higher to unwrapped g-C₃N₄. Wang et al. [141] performed density functional theory calculations on hybrid g-C₃N_4-MoS₂ and found that photo-generated electrons migrate easily from the g-C₃N₄ monolayer to the MoS₂ sheet leading to a high hydrogen evolution activity of the hybrid over g-C₃N₄ or MoS₂ singly. Recently, Liu et al. [142] synthesized 3D/2D nanojunction of flower-like MoS₂ on graphitic carbon nitride using ultrasonication and thermal treatment method. The optimum loading of 5 wt% of MoS₂ was also used by this group to generate 867.6 µmolh⁻¹ of hydrogen which exceeded that of untreated g-C₃N₄ by 2.80 times. The activity observed with 3D MoS₂ and g-C₃N₄ and was reported to be higher than other dimensions of MoS₂ composite with g-C₃N₄. The enhanced HER activity was attributed to tight interface between the catalysts which facilitated higher light harnessing, quicker bond charge separation, faster electron transport and the increased electrical conductivity of the materials. The catalysts also showed good electrocatalytic HER activity with reduced overpotential and high current densities. Zhang et al. [143] used hydrothermal method to prepare heterostructure of nitrogen doped carbon tube molybdenum disulphide (TNCT@MoS₂) for photocatalytic water splitting in the absence of electron donor. The microcomposite displayed increased performance with activity of 120 µmolh⁻¹ which was several times greater than untreated MoS₂. The generation of H₂ without the use of a sacrificial agent which plays the function of acceptance of the hole suggested the micro-heterostructure catalyst to be effective in water splitting. The achieved result was attributed to higher absorption of visible light, creation of more active sites, possibly the power of the xenon lamp used (1000 W); the maximum of power of xenon lamp that was reported in literature was 400 W, as well as the nitrogen dopant which could help in fast electron transfer to the active sites of the MoS₂. The group also observed similar activity when the pH of the water used was varied from 5–11. The gas chromatography used could not detect any O₂ generated during the reaction which was later probed by fluorescence method in which the intermediate OH was picked up. The O₂ produced during reaction was reported to be absorbed by the metal of the cocatalyst to form a peroxide complex.

Gupta and Rao [144] highlighted the use of dye with TMDCs in photocatalytic water splitting. The dyes have capacity to absorb light. Common dyes that are used are noble-metal-centred and metal-free dyes. Noble metal dyes that are often used are Ru(ii) tris bipyridine (Ru(bpy)₂³⁺) and hydrated iridium oxide (IrO₂.nH₂O). The metal free dyes such as Eosin Y or Rose Bengal are commonly used. The noble metal-centred dyes are being reported to be more effective in generation of hydrogen over the metal free dyes because they give more photocurrent upon irradiation with visible light over the metal free dye. They play a dual role as oxidizing catalyst and photosensitizer [145]. Upon excitation it undergoes some transformation to inject an electron to TMDC (MoS₂). In 2009 Zong et al. [146] conducted photocatalytic hydrogen generation in visible light using composite of dye Ru(ii) tris bipyridine (Ru(bpy)₂³⁺) which play the role of an organic photosensitizer and colloidal MoS₂ in presence of ascorbic acid. Upon irradiation in a visible light source, the dye transferred electrons to MoS₂ which has many active sites, which in turn reduce the protons to H₂(g). It was observed an increase in the concentration of the dye used led to a corresponding increase in the amount of H₂(g) formed. It was also noted the amount of the H₂(g) produced increased with temperature employed in the synthesis of colloidal MoS₂ using solvothermal method. When the reaction was performed in a mixture of acetonitrile and methanol solution while ascorbic acid serves as the sacrificial agent, the hybrid catalyst gave rise to 210 mmolh⁻¹ of hydrogen gas, but this got diminished after some hours due to the decomposition of the organic photosensitizer dye. The use of the dye has the drawback of photo-instability.

Other dyes that have been employed as organic photosensitizer is the metal–free dye such as Eosin Y and Rose Bengal. Min and Lu et al. [147] synthesized nano composite of MoS₂ and graphene {G} using hydrothermal method. The nano composite was sensitized by Eosin y dye during photocatalytic reaction using triethanolamine (TEOA) as sacrificial agent, showing an activity of 83.8 µmolh⁻¹. However, when EY-RGO was tested the activity was very low < 0.5 µmolh⁻¹ and activity of 37.7 µmolh⁻¹ with MoS₂/RGO was obtained.

Table 1. Role of TMDC as a good promoter of catalysts in hydrogen evolution reaction.

Catalyst	Synthesis Method	Sacrificial Agent	Light Source	Activity	Ref.
MoS2-TiO2	Hydrothermal	0.35 M Na2SO3 and Na2S	Xe Lamp 300 W	1600 µmolh−1	[100]
MoS2-CaIn2S4	Two-step hydrothermal	0.025M Na2SO3 and Na2S	Xe Lamp 150 W	602 µmolh−1	[106]
MoS2-CdS	Sonication and stirring	Na2SO3 and Na2S	Xe Lamp 300 W	1750 µmolh−1	[120]
MoS2-CdS	Impregnation	Methanol and 10% Lactic acid	Xe Lamp 300 W	532.8 µmolh−1	[121]
MoS2-CdS	Centrifugation	Lactic acid	Xe Lamp 300 W	259 µmolh−1	[122]
WS2-CdS	Impregnation	10% Lactic acid	Xe Lamp 300 W	420 µmolh−1	[123]
CdS-MoS2-WS2	Hydrothermal	10% Lactic acid	-	209.790 µmolh−1	[124]
MoS2-Zn0.2Cd0.8S	Photo deposition	0.25 M Na2SO3 and 0.25 Na2S	Xe Lamp 300 W	2 µmolh−1	[125]
MoS2-ZnIn2S4	Electrostatic self-assembly	Lactic acid	Xe lamp	4974 µmolh−1	[126]
WS2-ZnIn2S4	Micro wave	Na2SO3 and Na2S	Xe lamp 150 W	293.3 µmolh−1	[127]
MoSe2-ZnInS4	One polyol	0.35 M Na2SO3 and 0.25 Na2S	Xe Lamp 300 W	2228 µmolh−1	[128]
MoS2-TiO2	Hydrothermal	Methanol	Xe Lamp 350 W	75 µmolh−1	[129]
MoS2-TiO2	Hydrothermal	10% Lactic acid	-	550 µmolh−1	[130]
MoS2-TiO2	Annealing and impregnation	Triethanolamine	Xe Lamp 350 W	391.1 µmolh−1	[131]
Mn-CdS-MoS2-TiO2	Hydrothermal	Methanol	Xe Lamp 300 W	408.370 µmolh−1	[132]
MoS2-ZnO	Hydrothermal	0.5M Na2SO4	Xe Lamp 1000 W	768 µmolh−1	[133]
MoS2-ZnO	Hydrothermal	0.10M Na2S	Xe Lamp 300 W	27,690 µmolh−1	[134]
MoS2-ZnO	Hydrothermal	0.10 Na2SO4	Xe Lamp 100 W	145.6 µmolh−1	[135]
MoS2-SnO2	Hydrothermal	0.10 Na2SO4	Xe Lamp 400 W	117.2 µmolh−1	[136]
Bi2O3-MoS2	Two steps hydrothermal	-	Xe Lamp 300 W	10 µmolh−1	[137]
MoS2-CeO2	Hydrothermal	Methanol	-	508.44 µmolh−1	[138]
MoS2-g-C3N4	Impregnation and sulfidation	10% lactic acid	Xe Lamp 300 W	108 µmolh−1	[139]
WS2-g-C3N4	Impregnation and sulfidation	10% lactic acid	Xe Lamp 300 W	20.6 µmolh−1	[139]
MoS2-g-C3N4	Impregnation	Na2SO4	Xe Lamp 300 W	23.10 µmolh−1	[140]
MoS2-g-C3N4	Sonication and treatment	Triethanolamine	Xe Lamp 300 W	887.6 µmolh−1	[142]
MoS2-C	Hydrothermal	Na2SO3 and Na2S	Xe Lamp 1000 W	120 µmolh−1	[143]
MoS2-RGO	Hydrothermal	Triethanolamine	Xe Lamp 400 W	42,000 µmolh−1	[147]

Table 1. Role of TMDC as a good promoter of catalysts in hydrogen evolution reaction.

Catalyst	Synthesis Method	Sacrificial Agent	Light Source	Activity	Ref.
MoS2-TiO2	Hydrothermal	0.35 M Na2SO3 and Na2S	Xe Lamp 300 W	1600 µmolh−1	[100]
MoS2-CaIn2S4	Two-step hydrothermal	0.025M Na2SO3 and Na2S	Xe Lamp 150 W	602 µmolh−1	[106]
MoS2-CdS	Sonication and stirring	Na2SO3 and Na2S	Xe Lamp 300 W	1750 µmolh−1	[120]
MoS2-CdS	Impregnation	Methanol and 10% Lactic acid	Xe Lamp 300 W	532.8 µmolh−1	[121]
MoS2-CdS	Centrifugation	Lactic acid	Xe Lamp 300 W	259 µmolh−1	[122]
WS2-CdS	Impregnation	10% Lactic acid	Xe Lamp 300 W	420 µmolh−1	[123]
CdS-MoS2-WS2	Hydrothermal	10% Lactic acid	-	209.790 µmolh−1	[124]
MoS2-Zn0.2Cd0.8S	Photo deposition	0.25 M Na2SO3 and 0.25 Na2S	Xe Lamp 300 W	2 µmolh−1	[125]
MoS2-ZnIn2S4	Electrostatic self-assembly	Lactic acid	Xe lamp	4974 µmolh−1	[126]
WS2-ZnIn2S4	Micro wave	Na2SO3 and Na2S	Xe lamp 150 W	293.3 µmolh−1	[127]
MoSe2-ZnInS4	One polyol	0.35 M Na2SO3 and 0.25 Na2S	Xe Lamp 300 W	2228 µmolh−1	[128]
MoS2-TiO2	Hydrothermal	Methanol	Xe Lamp 350 W	75 µmolh−1	[129]
MoS2-TiO2	Hydrothermal	10% Lactic acid	-	550 µmolh−1	[130]
MoS2-TiO2	Annealing and impregnation	Triethanolamine	Xe Lamp 350 W	391.1 µmolh−1	[131]
Mn-CdS-MoS2-TiO2	Hydrothermal	Methanol	Xe Lamp 300 W	408.370 µmolh−1	[132]
MoS2-ZnO	Hydrothermal	0.5M Na2SO4	Xe Lamp 1000 W	768 µmolh−1	[133]
MoS2-ZnO	Hydrothermal	0.10M Na2S	Xe Lamp 300 W	27,690 µmolh−1	[134]
MoS2-ZnO	Hydrothermal	0.10 Na2SO4	Xe Lamp 100 W	145.6 µmolh−1	[135]
MoS2-SnO2	Hydrothermal	0.10 Na2SO4	Xe Lamp 400 W	117.2 µmolh−1	[136]
Bi2O3-MoS2	Two steps hydrothermal	-	Xe Lamp 300 W	10 µmolh−1	[137]
MoS2-CeO2	Hydrothermal	Methanol	-	508.44 µmolh−1	[138]
MoS2-g-C3N4	Impregnation and sulfidation	10% lactic acid	Xe Lamp 300 W	108 µmolh−1	[139]
WS2-g-C3N4	Impregnation and sulfidation	10% lactic acid	Xe Lamp 300 W	20.6 µmolh−1	[139]
MoS2-g-C3N4	Impregnation	Na2SO4	Xe Lamp 300 W	23.10 µmolh−1	[140]
MoS2-g-C3N4	Sonication and treatment	Triethanolamine	Xe Lamp 300 W	887.6 µmolh−1	[142]
MoS2-C	Hydrothermal	Na2SO3 and Na2S	Xe Lamp 1000 W	120 µmolh−1	[143]
MoS2-RGO	Hydrothermal	Triethanolamine	Xe Lamp 400 W	42,000 µmolh−1	[147]

The presence of graphene also contributed to the enhanced performance of the catalyst. On excitation of the dye with visible light, the dye injected electron (EY⁻¹) into the RGO after which to the active edge sites of the MoS₂ which reduces the proton to H₂. The Eosin was proposed to first move to single state (EY^1*) on excitation, which later descended to low-lying triple state (EY^3*) by inter system crossing and later reduced to (EY⁻¹) in company of sacrificial agent (TEOA). EOSIN Y is preferred over other metal-free dyes due to the prolonged lifetime during excitation state because of bulky (Br) atom in the molecule. High activity was accounted for by the tight p-n junction of the composite. The graphene being heavily doped with the heteroatom nitrogen (15%) boosted electron injection capability of the graphene thus better HER activity. The intensity (power) and kind of lamp used could as well contribute the amount of H₂ produced. The amount of H₂ generated was a magnitude of four-fold higher (42 mmolh⁻¹) when 400 W xenon lamp to when 100 W Halogen lamp was used (10.5 mmolh⁻¹).

Deployment of phase-engineering has greatly help to elevate the catalytic output of MoS₂ by making both the edge sites and basal plane available for HER thus improving on the performance of the catalyst. This method helps to tune the electronic structure of the semi-conductor nanomaterial to a metallic 1T-phase. New research has proved that phase-engineered metallic 1T-MoS₂ was more effective in water splitting over 2H-MoS₂ since both the edge sites and basal plane active sites participated in the catalytic reaction. Maitra’s group [148] demonstrated the use of mono 1T-MoS₂ for photocatalytic water splitting in the presence of dye (EOSIN Y) the yield was 30 mmolh⁻¹ of H₂ which was several folds (600) higher compared to few-layer 2H-MoS₂ (0.05 mmolh⁻¹). With prolonged time (3 h) 1T MoS₂ generated 250 mmolh⁻¹ of H₂. However, when the catalyst was dried, the activity was minimized, and observable reduction was noticed in the performance when annealed sample was applied. The H₂ yield was comparable to that of 1H-MoS₂.This suggested possible transformation to 1H-MoS₂ upon annealing the sample. The group also did further study on the metallic group VI TMDC by working with other chalcogens. When metallic 1T MoSe₂ was tested for HER performance was 900 times higher than 1H-MoSe₂. A similar observation was also recorded for 1T MoTe₂ which showed far better activity when compared to 1H MoTe₂. In related research applying phase engineering the group [144] also produced 1T MoSe₂ which showed high performance for hydrogen evolution using visible light. The reaction was conducted using EOSIN as a sensitizer and TEOA as holes acceptor. It was observed that the HER activity of 7500 mmolh⁻¹ was far higher than for the semiconducting phase (1H-MoSe₂) by almost 100 times, as well more superior to both 2H and 1T- MoSe₂. This was attributed to ultra-fast electron transport to the active sites of the material upon excitation with photon. DFT using theoretical calculation also proved 1T-MoSe₂ to have a lower work function than both 1T and 2-H MoSe₂.

Other methods that are employed to make the basal plane of TMDC active apart from phase engineering include introducing sulphur vacancies and doping with active metal ions charge. Common methods used in introducing chalcogenide deficiencies are contact with argon (Ar) plasma, annealing in H₂ and electrochemical reduction. Doping with active metals such as Cu, Zn, Ag, Pt and Co can increase the active sites and conductivity of TMDC. Recent Investigations have shown that the basal plane of TMDC become active by activating the catalytic activity of Se atoms in the basal plane [149].

4. Conclusions

The need for precious metal as co-catalyst might soon diminish with the development of TMDCs which are highly active and robust co-catalysts with high efficiencies for photocatalytic water splitting without the need for a precious metal is well articulated in this review. The metallic phase MoS₂ works better than the semiconductor phase because both edges and basal sites participate in the catalytic reaction in contrast with the latter in which only the edges become involved in the catalysis. TMDCs can serve as a good promoter to enhance the photocatalytic activity of semiconductor metal oxides. The nanocomposite of CdS with TMDC produced the highest performance towards HER. The use of Cd based catalyst should be discontinued due to the high environmental impact of cadmium. The hydrogen generation increases with the loading of co-catalyst on a photocatalyst until it reaches the optimum loading as result of increase in density of active sites, after which, it drops because of the covering effect which barricades the photocatalyst thus hindering light absorption. Mono or few-layers of TiS₂ prepared by colloidal synthesized are prone to oxidation in ambient conditions which could lead to degradation or oxide of the titanium being formed. ZrS₂ and HfS₂ are also prone to oxidization which affect the true stoichiometry of these nanomaterials; hence they are referred as n type semiconductor, however no study is available of colloidal syntheses of mono or few-layers ZrS₂ and HfS₂ because they are unstable after few days of synthesis. There are still some setbacks preventing the use of TMDC composites for practical application. First, the unseparated interface with the semiconductor and the TMDCs should be well constructed to allow for the charge separation and transportation. This prevents photo-corrosion and agglomeration during photocatalysis. Hence, new proper methods of loading should be developed for practical application. Secondly, some of the synthesis methods of the TMDCs need to be improved. Hydrothermal and sol gel methods are not quite suitable for the synthesis of group IVB TMDCs, however, with CVD the nanomaterials are formed on substrates which hinders their application for photocatalytic reaction in aqueous solution. The use of hydrothermal method seems to be the most facile approach to construct nano heterostructure with high activity. Large scale synthesis is still not possible with the colloidal method. Thirdly, the use of sacrificial agents to suppress the fast recombination of electrons and holes hinders the real application of the catalyst. Till now, it is only DFT that has demonstrated the use of group IVB TMDCs and construction of their composites for water decomposition. There is a need for practical experimental approach to explore these promising nanomaterials similar to MoS₂ and WS₂ have been well explored as cocatalysts.