2D Tungsten Chalcogenides: Synthesis, Properties and Applications

Layered transition metal chalcogenides possess properties that not only open up broad fundamental scientific enquiries but also indicate that a myriad of applications can be developed by using these materials. This is also true for tungsten‐based chalcogenides which can provide an assortment of structural forms with different electronic flairs as well as chemical activity. Such emergence of tungsten based chalcogenides as advanced forms of materials lead several investigators to believe that a tremendous opportunity lies in understanding their fundamental properties, and by utilizing that knowledge the authors may create function specific materials through structural tailoring, defect engineering, chemical modifications as well as by combining them with other layered materials with complementary functionalities. Indeed several current scientific endeavors have indicated that an incredible potential for developing these materials for future applications development in key technology sectors such as energy, electronics, sensors, and catalysis are perhaps viable. This review article is an attempt to capture this essence by providing a summary of key scientific investigations related to various aspects of synthesis, characterization, modifications, and high value applications. Finally, some open questions and a discussion on imminent research needs and directions in developing tungsten based chalcogenide materials for future applications are presented.


DOI: 10.1002/admi.202000002
Their fascinating properties are the result of the layered structure held together by weak van der Waals forces similarly to graphene, however, in TMDCs one layer is much more complex; consisting of a hexagonal plane of transition metals (typically metals of group IV-VII) sandwiched between two planes of chalcogens (S, Se, and Te) by strong covalent bonds (Figure 1). In 2004, the pioneering isolation of graphene sheets [1] gave a tremendous boost to the scientific community scrutinizing similar layered materials which can be separated relatively easily to single-layers or so-called monolayers. Based on their unique electronic transport properties and advantageous band structure these materials are suggested to have a great number of applications in transistors, [2][3][4] solar cells, [5][6][7] optoelectronic devices, [8,9] catalysts, [10] and sensors. [11] As might be anticipated their properties at atomic scale, greatly differ from their bulk counterpart. Recently, monolayers of MoS 2 and WS 2 have been found to exhibit direct semiconducting band gap in the visible spectrum rather than an indirect one that is well-known for their bulk phase. [12] In addition to material thickness, the band gap can be further fine-tuned, implying also beneficial changes in material properties, by doping TMDCs with different chalcogen atoms. As an example, when the thickness of MoS 2 is reduced from bulk to monolayer a significant increase in the band gap can be observed, from ≈1.2 to ≈1.8 eV, [12] accompanied with an indirectto-direct transition, and as expected, single layers also exhibit Layered transition metal chalcogenides possess properties that not only open up broad fundamental scientific enquiries but also indicate that a myriad of applications can be developed by using these materials. This is also true for tungsten-based chalcogenides which can provide an assortment of structural forms with different electronic flairs as well as chemical activity. Such emergence of tungsten based chalcogenides as advanced forms of materials lead several investigators to believe that a tremendous opportunity lies in understanding their fundamental properties, and by utilizing that knowledge the authors may create function specific materials through structural tailoring, defect engineering, chemical modifications as well as by combining them with other layered materials with complementary functionalities. Indeed several current scientific endeavors have indicated that an incredible potential for developing these materials for future applications development in key technology sectors such as energy, electronics, sensors, and catalysis are perhaps viable. This review article is an attempt to capture this essence by providing a summary of key scientific investigations related to various aspects of synthesis, characterization, modifications, and high value applications.
Finally, some open questions and a discussion on imminent research needs and directions in developing tungsten based chalcogenide materials for future applications are presented.

Introduction
In recent years, 2D nanostructures, such as graphene and hexagonal boron nitride, and lately, thin sheets of transition metal dichalcogenides (TMDCs) have gained emerging attention fueled by their excellent properties and potential use in future electrical devices, optoelectronics and associated technologies.

Structure and General Characterization
Typically, the unit cell structure of TMDCs is either trigonal prismatic (hexagonal, 2H or D 3h ) or distorted hexagonal also known as octahedral structure (1T or D 3d ). The 2H unit cell is extended over two chalcogen-metal-chalcogen layers, whereas in 1T there is only one layer in a unit cell. [16] Which of these two lattice structures is more stable depends on the constituent metal and chalcogen atoms. In general, one can say that TMDCs formed of group IV metals are nonmagnetic semiconductors that are only stable in trigonal structures. [17] Albeit, the reversible transition to octahedral unit cell structure is possible, through chemical modification (for example exfoliation), [18,19] it can distort the lattice structure. The significance of the unit cell structure is clearly demonstrated by the semiconducting nature of 2H-MoS 2 compared to the metallic one of 1T-MoS 2 , [20] consequently the unit cell structure affects the attributes of these materials largely. Although the research field related to TMDCs is extremely broad; including a large number of different combinations of metals and chalcogens, taking shape in interesting morphologies varying from 3D to 1D (nanotubes, nanobelts, nanowires, nanoparticles etc.); yet in this review, we are concentrating only on 2D sheets and flakes of tungsten dichalcogenides (WS 2 , WSe 2 , and WTe 2 ). Till today, layered tungsten dichalcogenides (WX 2 ) have not been studied to such an extent as 2D molybdenum dichalcogenides mainly due to their higher melting point and therefore more difficult sublimation of WO 3 (and W) compared to MoO 3 (and Mo), most commonly used reactants for tungsten and molybdenum dichalcogenide synthesis. [21] As the layer thickness of TMDC is reduced, the reflection (002) cease to appear on the XRD pattern, making this technique a convenient tool in verifying the formation of few-layered flakes, though differences between mono-, bi-, and few-layered TMDCs cannot be achieved. [23] Contrary, high resolution scanning transmission microscopy is a quite practical tool to resolve the details of both micro and crystal structure. It is possible even to identify and distinguish 2H and 1T phases. [24] In general, Raman spectroscopy is known to be one of the most useful methods analyzing low-dimensional structures and identifying the number of layers in TMDCs. [25] In a typical spectrum, two characteristic peaks can be identified: i) A 1g , caused by the chalcogen atoms moving of the plane in opposite directions while metal atoms, situated between them, stay stationary, and ii) E 2g 1 attributed to the metal and chalcogen atoms moving in-plane, in opposite directions. If the TMDC comprises of several layers, shift in both peaks occurs (E 2g 1 mode undergoes a redshift, whereas the A 1g , mode shows blue shift), [26] since the extra layers stiffen the out-of-plane movement and help in relaxing the in-plane movement. It has to be noted, that in some reports on WS 2 and WSe 2 only the A 1g mode was found to be thickness dependent while the position of the E 2g 1 mode was stated stationary. [27] Similarly, orthorhombic WTe 2 is an exception; when its thickness is reduced, beside the in-plane A 1 7 mode, a shift for all other vibrational modes is observed. Furthermore, the lack of some Raman peaks, identified in the case of the bulk phase, provide a reliable method to distinguish mono-and bilayer WTe 2 [28] (Figure 2). Differences in Raman spectra are understandable since the vibration modes are completely different for 1H and 1T structures. Further influential factors on the appearance of Raman spectra can be caused by temperature [29,30] carrier concentration and strain in the lattice. Room temperature photoluminescence spectroscopy (PL) is not sensitive enough to determine the quality of monolayer crystals, and therefore low temperature PL is required in order to reveal PL behavior of such materials. [31] A typical PL spectrum of WSe 2 monolayer shows three distinguished features: i) a high energy peak (X 0 ) at 708 nm caused by excition emission, ii) a high energy peak (X − ) at 722 nm corresponding to charge excition emission, and finally iii) a broad emission peak caused by the emission of impurity/defect-trapped excitions. [32,33] Photoluminescence measurements show that similar to MoS 2 both WS 2 and WSe 2 exhibit a transition from indirect to direct band-gap as the thickness of the materials is decreased Figure 1. Crystal structures of single-layer TMDCs: a-d) 1H, trigonal prismatic coordination with hexagonal symmetry; and e-h) 1T, octahedral coordination with tetragonal symmetry. i-k) 2H polytype having two 1H layers in the repeat unit, metal atoms atop chalcogens of the adjacent layer and vice versa, top view (i), side view (j and k). l-n) 1T polytype: two 1T layers in the repeat unit, metal (chalcogen) atoms atop metal (chalcogen) atoms of the adjacent layer, top view (l), side view (m and n). Red and yellow balls represent metal and chalcogen atoms, respectively. Reproduced with permission. [22] Copyright 2015, Elsevier Ltd.

Figure 2.
Raman spectra of a) WS 2 , b) WSe 2 , and c) WTe 2 from bulk to monolayer. (a and b) Reproduced with permission. [27] Copyright 2013, Springer Nature. (c) Reproduced with permission. [28] Copyright 2016, Royal Society of Chemistry. from multilayers to monolayers. The PL intensity is decreasing ≈3 orders of magnitude with the number of layers from 1L to bulk (Figure 3a,d). Apart from the peak of indirect transition (I), two other peaks appear in the spectra at higher energies: one is due to a direct transition (A) and another caused by hot PL (B) (Figure 3b,e). The broadening of these latter peaks suggests an excitonic origin similar to that in MoS 2 . Furthermore, the nearly constant energy difference of ≈400 meV between A and B (Figure 3c,f) suggests limited interlayer hopping in both WS 2 and WSe 2 ultrathin thin films. Ab initio calculations indicate that such thickness independent splitting is due to the giant spin-valley coupling which indeed suppresses interlayer hopping at the K points of valence band edge. [27]

Synthesis of Tungsten Dichalcogenides
In theory, the crystal structure of TMDS enables relatively easy isolation of thin layers since the chalcogenide-metal-chalcogenide layers are attached by weak van der Waals force. Yet, experimental preparation of single layer films has been hindered by the inherent inclination of these materials to form multilayers or nanotubes. Initially, mechanical and chemical exfoliation, [15,34,35] with very low yield, were applied to separate monolayers from the bulk material while in the last few years, the most attention has been devoted to a novel method, namely to thermal-sulfurization of thin layer of corresponding metal oxides. [8,36,37] One of the major problems associated with the preparation of monolayers of TMDCs, is the difficulty in achieving industrial quality monolayers required for large scale production of TMDCs based electrical applications.
In order to exploit the beneficial properties of 2D tungsten dichalcogenides, its layers have to be separated from each other. Unfortunately, the production of large quantities of uniform mono-, bi-, or few-layered materials can be challenging and costly. Currently available large variety of different approaches can be divided into two major categories: top-down exfoliation of layers from bulk material, and bottom-up synthesis of layers from reactants. To fully exploit the unique properties of these materials for future applications, such procedures are required to have control over the crystal size, quality, and thickness. It is generally agreed that the material quality specifications highly depend on the particular application, while some require large quantities others need appropriate well-defined shaped monolayers, and as such the selection of the most suitable approach can be extremely challenging. Nevertheless, in this section we will make an attempt to provide a comprehensive review on the various synthesis methods of 2D tungsten dicalcogenides (Figure 4).
by Scotch tape from the surface of highly ordered pyrolytic graphite (HOPG). [1] It is not surprising that such a simple method was adapted for the cleavage of thin flakes of other weakly bonded layered materials, as well as for tungsten chalcogenides (WS 2 , [29] WSe 2 , [39] WTe 2 [40] ). Although such procedure can provide high quality, thin layers of crystals, it suffers from a great number of drawbacks as it is time-consuming, can produce only low quantity, and has poor reproducibility.
In order to produce large amount of TMDCs nanosheets liquid phase exfoliation assisted by sonication has proven to be a promising method. Throughout this treatment, van der Waals forces keeping together the layers in the bulk, are weakened by chemicals such as solvents like n-methyl-2-pyrrolidone (NMP), [35,41] n-cyclohexyl-2-pyrrolidone (CHP), [42] sodium naphthalenide (NAPH), [43] sodium cholate, [44] (Figure 5) or water mixtures. [45] Intercalation of ions, [41] such as lithium, [23] expanding the layer distance and thereby creating strain in the lattice, is commonly applied as well ( Figure 6). Albeit dispersing bulk WSe 2 and WS 2 in NMP has shown to produce few layered flakes, [35,41] from monolayer to tens of layers, yet as the formed nanosheets are often heterogeneous in quality it limits their use in most of the applications. Ion intercalation applying Figure 4. Strategies for the formation of monolayers of transition metal dichalcogenides including top-down chemical/mechanical exfoliation as well as bottom-up chemical vapor deposition, powder vaporization, metal transformation, chemical vapor transport, pulsed laser, deposition, molecular beam epitaxy, spray pyrolysis, and electrochemical synthesis. Reproduced with permission. [38] Copyright 2015, Annual Reviews. lithium cations is typically done using expensive n-butyllithium as an intercalation agent. [23,41] Once lithium ions have intercalated between the layers, the material is washed with water which on the one hand will remove the excess n-butyllithium (BuLi), on the other hand it will also react with the intercalated lithium resulting in lithium hydroxide and hydrogen gas, eventually pushing the layers apart. Lithium intercalation as such has been shown to be effective to produce mono-and bilayered WS 2 flakes, [23] nevertheless the same method applied to WSe 2 eventuated flakes with thickness of several micrometers. [41] Limitation of this approach lies in its long reaction times and elevated reaction temperatures. [23] Electrochemical lithiation has been found as an efficient route to produce large quantities of single-layer WS 2 nanosheets. [46] The bulk layered material was used as the cathode, whereas lithium foil served as the anode, providing intercalating lithium ions and LiPF 6 was applied as the electrolyte. As soon as current was conducted between the electrodes the lithiation process started, after which the material was washed and exfoliated. While traditional lithium intercalation process takes several days to have fully intercalated product the above described electrochemical process can be accomplished in a few hours. Possibility of its facile scale-up, by increasing the size of the electrodes, can be attributed as a further advantage of this synthesis procedure. Even though lithium intercalation is considered as an effective way to produce high amounts of well-controlled nanosheets, the flammable nature and expensive price of lithium, urge the scientific community to find a replacement for this approach.
Although exfoliation is a conventional method to produce large quantities of nanosheets, yet structural phase transformations from 2H to 1H during intercalation process can make the process unreliable. The nature of intercalation agent has significant impact on the degree of exfoliation as well as on the phase transition of the material. [43] For instance, exfoliation of WS 2 in BuLi the concentration of 1T-WS 2 is ≈10 at.% of, whereas in NAPH it is significantly higher ≈37 at.%. Interestingly, the NAPH exfoliated WS 2 showed room temperature ferromagnetism, which was not detected for BuLi exfoliated samples or for the bulk material.
Other uncommon chemicals to exfoliate bulk WS 2 include supercritical CO 2 [47] but also concentrated chlorosulfonic acid [48] has been suggested to produce few-layered nanosheets. This latter can provide stable and highly concentrated dispersions compared to those in NaDBS, propanol or sulfuric acid, however its major disadvantage is the protonation of the nanosheet surface.

Bottom-Up Synthesis
Other useful routes producing 2D tungsten dichalcogenides are such "bottom-up" methods as chemical vapor deposition (CVD), powder vaporization, metal transformation and chemical vapor transport. These techniques apply high temperatures to facilitate the decomposition of precursors, diffusion and reaction of the participating atoms, and formation of crystalline products. The differences among these methods are only the way of delivery of the precursors.
In CVD, both metal and chalcogen precursors are vapors, which react in the gas phase and deposit on a substrate to form the crystalline product. [21,49,50] Notable examples of substrates include SiO 2 , [21] Si 3 N 4 , [51] sapphire, quartz, and graphene [36,37,50] offering a good selection of surfaces suitable for any particular application. In powder vaporization, oxides of the metals are vaporized and reacted with the chalcogens (either vapors or precursors), which then deposit and grow crystals on supports. In the case of metal transformation, solid metals are first deposited on the substrate, and then exposed to vapors of chalcogens at temperatures typically above 600 °C. Another approach is an application of solid chalcogen film over the metal for solid-solid reactions at high temperatures.
The presence of different molecules seems to play an important role as promoting agents during the formation of 2D tungsten dichalcogenides, although the exact mechanisms remain elusive in many cases. In order to grow highly crystalline monolayers of WSe 2 by selenization of WO 3 , hydrogen gas has shown to have a crucial role as an activating agent. [52] Crystal growth starts by the reduction of WO 3 to volatile WO x which reacts with vaporized Se followed by the formation of WSe 2 layers. Similarly, in the case of WS 2 flake synthesis, a small amount of H 2 has shown to influence the shape of the growing crystals, making those more symmetric by affecting the growth process. [53] Not only H 2 but also additional sulfur has been shown to promote the formation of mono-and few-layered WSe 2 flakes during the CVD synthesis. [54] As evidenced by some research groups CVD process based on organic tungsten and selenium precursors result in thick tungsten dichalcogenide films. [55,56] There are only a few reports of large-area mono-or few-layers of WSe 2 using metalorganic compounds as precursors. [50] In the course of the growth process, pure H 2 as a carrier gas is favored in order to prevent the formation of carbon impurities originating from organometallic compounds. However, it is worth noting here that Kang et al. [57] recently managed to grow 4-inch WS 2 monolayer films on Si wafer in diluted H 2 gas. The key strategy for this procedure was to carefully monitor the concentrations of each chemical throughout the synthesis by regulating the partial pressures.
Another useful route to the large area, triangle crystals WSe 2 growth is the physical vapor deposition (PVD), for which case bulk WSe 2 , instead of some precursor, is used as a raw material to synthesize WSe 2 monolayers on a substrate. [31,58] The growth mechanism shares some similarities with that of the chemical vapor deposition and transport processes, hence the shape of the crystals can be controlled by altering the carrier gas mixture or precursor ratios. As it might be anticipated, parameters, such as flow rate, temperature, pressure and time, moreover the nature of the substrate all has their influences on the quality of the resulting product. [21,[50][51][52] Ultra-narrow WS 2 nanoribbons with layer thickness of one to few was reported by Wang et al. [59] exploiting narrow carbon nanotubes (CNTs) as a growth template. To begin with CNTs are opened, followed by the introduction of organometallic tungsten precursor in H 2 S/H 2 atmosphere at high temperature (800 °C). The layer thickness of nanoribbons is determined by the inner diameter of the CNT used as a template, accordingly monolayered WS 2 nanoribbons are synthesized in single-walled carbon nanotubes (SWCNTs) while bilayered ones are formed in double-walled carbon nanotubes (DWCNTs) having larger inner diameter than the SWCNTs. The length of nanoribbons varied between tens to hundreds of nanometers, whereas the maximum width was found to be around 3 nm. Both armchair and zigzag forms of nanoribbons were detected, latter one being the more stable.
In all above described examples, the growth direction of the crystal layers has been horizontal, parallel to growth substrate. [53] Vertical growth of TMDCs [60] can be achieved by the modification of the metal seed layer thickness on the substrate, since substrates with low nucleation density favor the formation of vertical growth over horizontal. [50] Although there are several other methods to prepare tungsten dichalcogenide nanostructures such as microwave assisted wet chemical approaches, sonochemical and hydrothermal synthesis, the products gained through these methods are not 2D materials but nanowires, [61] nanorods, [62] nanoparticles, [63] nanotubes, [64,65] or layered flower-like structures (Figure 7). [66,67,69] Up to now, there are only a couple of reports describing wet chemical synthesized 2D tungsten dichalcogenides. A simple way of producing WS 2 nanosheets was shown by applying WCl 6 and thioacetamide as precursors in a hydrothermal process (265 °C for 24 h). [30] The as formed few-layered nanosheets had a typical thickness of 1-3 nm and a lateral length of 0.5-1 µm. Another simple, yet unique wet chemical approach was proven to produce either metallic 1T-WS 2 or semiconducting 2H-WS 2 monolayers, depending on the synthesis conditions. [68] The precursor solution includes WCl 4 , oleylamine, and oleic acid, in which carbon disulfide (CS 2 ) serving as a sulfide source and reducing agent is added at 320 °C. At the initial phase of the reaction dithiocarbamate forms, and after thermal decomposition WS 2 forms. Depending on the reactivity of tungsten precursor, controlled by the coordination ligand, the 1T-WS 2 /2H-WS 2 ratio of the product can be adjusted.

Tungsten Disulfide and Tungsten Diselenide
Regarding the band gap of WS 2 and WSe 2 , similarly to MoS 2 monolayers, indirect to direct transition is observed, when the layer thickness is reduced to monolayers. Accordingly, singlelayers of the aforementioned materials are more efficient phonon absorbers and emitters than their bulk counterparts.
The band structures as the function of different layer thickness (Figure 8) clearly reveals that only monolayers own a direct band gap, whereas in double and higher numbers of layers it turns into indirect. To be more exact, the valence band maximum (VBM) and the conduction band minimum (CBM) of the bulk phase are located at the Γ-point and between the Γand the K-points, respectively; on the contrary, both VBM and CBM of single layer shift to the K-point therefore the k-vectors become the same. The transport properties of WSe 2 are strongly influenced by the type of the contact metal and by the preparation method. For example, mechanically exfoliated WSe 2 monolayers were found to be n-type, [84] whereas CVD grown WSe 2 monolayers, contacted using palladium shown p-type behavior whereas with gold or titanium contacts the behavior was found to be ambipolar. [21] Back gated FETs were prepared using the 2D WSe 2 flakes directly grown on Si/SiO 2 substrates. Various contact materials were tested, deposited using e-beam lithography including Pd/Ti (50 nm/0.5 or 1 nm), Au/Ti (50 nm/0.5 or 1 nm), and Ti/Au (5 nm/50 nm). [21] Looking forward we note that the ability to manipulate the properties of such 2D materials by simply altering the material contacts carries large potential for forthcoming electronic and optoelectronic devices.
As for WS 2 , both n and p-type semiconducting behaviors were reported. The crystal seems to be highly sensitive to the surroundings and impurities. Slight doping of reactive sputter deposition grown WS 2 mono and few-layered crystals with N atoms results in a switch from n to p-type behavior. [85] Table 2. Lattice constants (a) and bond lengths (MX) of monolayers, and distance between layers (h) in bulk. [70,78,79,80,81,82] . MX   Heterostructures of WS 2 nanowires and nanoflakes obtained by the sulfurization of WO 3 nanowires showed p-type properties. [86][87][88] In addition, doping WS 2 lattice with Ta atoms changes the electrical properties of the material from semiconducting to metallic. [70] Room temperature ferromagnetic behavior of WS 2 monolayers was detected for both NAPH [43] and DMF [89] exfoliated samples. Interestingly, few-layered WS 2 and WSe 2 were found to have some ferromagnetism when exfoliated in DMF. [90] This peculiar magnetism on the one hand was speculated to be caused by the phase transition from 2H-WS 2 to 1T-WS 2 , on the other hand exfoliation resultant high amount of nanosheet edges and high degree of disorder were also named as inflicting parameters., [91][92][93] Although the reported ferromagnetic values are relatively low (≈10 −3 emu g −1 ), nevertheless such phenomenon was detected neither for bulk nor for WS 2 exfoliated in other solvents.

Tungsten Ditelluride
Tungsten ditelluride differ greatly from other tungsten dichalcogenides since it has the lowest energy when crystallized in distorted 1T lattice structure shown in Figure 9 (also known as Td), having ≈0.075 eV lower energy per unit than that of 2H-WTe 2 . [77] Compared to other TMDCs both bulk and monolayer form of Td-WTe 2 have similar semimetallic band structures, caused by small electron and hole pockets along the Γ-X direction. [94] Theoretical studies of Augustin et al. [95] and Ghosh et al. [96] have demonstrated that Td-WTe 2 has both semimetallic and metallic nature. The first one is due to the partial overlapping of Te 5p and W 5d bands, whereas the latter is attributed to the classical metallic bands. Electrical properties of WTe 2 nanoribbons, when assumed to have hexagonal structure, depends on the arrangement of atoms in the lattice. Similarly to WS 2 nanoribbons, [59] zigzag structure WTe 2 nanoribbon exhibits metallic behavior while atoms forming armchair structure will result in a semiconducting [96] material.
The low density of states at Fermi level (Figures 9 and 10) is responsible for the surprisingly low conductivity of 2D WTe 2 structures, however, high pressure synthesis process, can eventuate low temperature superconductive behavior. [97,98] The large magnetoresistance of single crystal WTe 2 can be reduced at the expense of greater pressure. [97] Moreover, further increasing the pressure a critical value is reached at 10.5 GPa, where magnetoresistance will be turned off completely, generating Figure 9. a) Trigonal prismatic coordinated 2H-WTe 2 structure and b) octahedrally coordinated Td-WTe 2 structure. The electronic band structures of c) WTe 2 in the 2H structure has an indirect 0.70 eV bandgap, whereas in the d) Td structure has a 0.21 eV band overlap in Γ-X and the corresponding density of states e) reaches a minimum, but not zero at the Fermi level. Reproduced with permission. [40] Copyright 2015, Springer Nature. a superconductive material. In situ Hall coefficient measurements revealed that with increasing pressure, while no phase transitions were observed, the number of hole carriers decreased at the same time the number of electron carries increased. WTe 2 follows a dome-shaped superconductivity phase, emerging at 2.5 GPa. [98] Further elevating the pressure, the maximum critical temperature (T c ) of 7 K is attained at around 17 GPa, after which, monotonous decline of the T c is observed with the escalating pressure.
The extraordinary transport properties of WTe 2 manifested at low temperature can be accounted for the approximately similar size of electron and hole pockets at the Fermi level and the resultant perfect balance of electron and hole populations. Accordingly, the extremely large quadratic magnetoresistance reported for single crystals of WTe 2 , showing no signs of saturation up to 60 T, is caused by the temperature dependent charge compensation of the transport. [94,100,101] Obviously such unique magnetoresistance could open up new opportunities in spintronic, and in low-temperature magnetic-field sensing applications as well.

Modification of Properties
Introduction of mechanical strain to the lattice can significantly change the electronic band structure and thus materials properties Application of even small (1-2%) compressive strain to the lattice of WX 2 monolayers can significantly change the band structure. [102] Ab initio calculations revealed that tungsten chalcogenides behave in a similar fashion as their molybdenum based counterparts (Figure 11). [78] Both tensile and shear strains induce a reduced band gap, and upon homogeneous biaxial tensile strain of ≈10% a semiconductor-to-metal transition occurs.
It is worth pointing out, that the electrical, mechanical (Poisson's ratio, in-plane stiffness) and optical properties of monolayer WTe 2 (distorted T-phase) are anisotropic. [77] Introduction of small parallel strain (1%) WTe 2 changes the semimetallic properties to semiconducting, whereas the perpendicular strain has no influence on the properties. Also, the imaginary part of the dielectric function shows different behavior depending on the direction of strain. Promising applications in the field of mechanical sensors [96] is foreseen for WTe 2 nanoribbon owing to its unique ability to change its band gap when defects, wrap, twist or ripple is introduced.

Field-Effect Transistors
Transistors are among the most important applications of semiconductors. Thus, their high structural stability, dangling bond free surface and high carrier mobility make semiconducting TMDCs promising candidates in such devices. Furthermore, their mechanical flexibility and reasonably good optical transparency are additional advantages raised great interest in the (a-c) Reproduced with permission. [16] Copyright 2016, American Chemical Society. Electronic structure of Td-WTe 2 in d) bulk and e) monolayer forms with spin-orbit coupling. Reproduced with permission. [99] Copyright 2015, IOP Publishing.
field. [103] One excellent example is a 2D flexible and transparent thin film transistor (TFT) applying graphene metal electrodes, h-BN (hexagonal boron nitride) gate dielectric and bilayer WSe 2 as a semiconducting channel. [104] The mobility values reported for the WSe 2 based device were ≈100 times better than those of typical amorphous silicon based TFTs, and at the same time its I ON /I OFF ratio was 10 7 . The device was stable under a broad temperature range and more than 88% transparent over the visible spectrum. All these results made it very clear that these materials may revolutionize micro and nanoelectronics and necessitate further efforts to the scrutiny of WS 2 and WSe 2 . [105] Suitability of WS 2 monolayer for a field-effect transistor (FET) application, grown straight onto SiO 2 substrate, was studied by Kang et al., [57] and although the synthesis parameters of the material was not fully optimized, the n-type FET showed relatively high field-effect mobility (18 cm 2 V −1 s −1 ), I ON /I OFF current ratio (10 6 ) and also current saturation. Similar I ON /I OFF ratios were reported for WS 2 -FETs by another group, [106] moreover in this case even higher field-effect mobility values were achieved: 50 cm 2 V −1 s −1 at room temperature; 140 cm 2 V −1 s −1 and 300 cm 2 V −1 s −1 at −205 °C for monolayer and bilayer devices, respectively. [106] The dielectric environment of WS 2 flakes affects the field-effect mobility as demonstrated by Withers et al. [107] comparing devices having SiO 2 and BN/SiO 2 gate dielectrics. The mobility and I ON /I OFF ratio of monolayer WS 2 FETs can be further improved by sandwiching the WS 2 monolayer between h-BN [108] layers. Thereby, mobility values as high as 214 cm 2 V −1 s −1 (at room temperature, with an I ON /I OFF ratio of ≈10 7 ) and 486 cm 2 V −1 s −1 (at −268 °C) could be achieved. The h-BN layers not only provided an ideal substrate for the WS 2 monolayer but also served as protection from doping by air [106] and other impurities. Similarly, in the case of n-type monolayer WSe 2 FET with Ag contacts, an extra layer of Al 2 O 3 deposited on the channel area showed superior behavior (I ON of 205 µA µm −1 and mobility of 202 cm 2 V −1 s −1 ) compared to a device without Al 2 O 3 (I ON of 110 µA µm −1 and mobility of 48 cm 2 V −1 s −1 [75] ). Here, the additional Al 2 O 3 provided highκ dielectric environment and suppressed Coulomb scattering. Although one of the highest values, I ON of 210 µA µm −1 and mobility of 142 cm 2 V −1 s −1 , [84] were reported of WSe 2 -FET with In contacts and Al 2 O 3 gate dielectric yet applying indium has its downsides such as poor adhesion and also low melting point (<160 °C), which makes it a quite unpractical contact material. WSe 2 monolayers enable the construction of both n [75,84] -and p-type [109,110] FETs; and consequently, inverter structures. [111,112] In addition, the better resistance of selenides to oxidation in reference to their sulfide counterparts, make those even more attractive for practical applications. [113] Consequently the above mentioned qualities contributed significantly to the extensive scientific interest toward WSe 2 as a channel material Figure 11. Band gap of monolayer tungsten dichalcogenides as a function strain. xx and yy are uniaxial expansion in x-and y-directions, respectively. xx + yy is homogeneous biaxial expansion in both x-and y-directions. xx − yy is expansion in x-direction and compression in y-direction and yy − xx is expansion in y-direction and compression in x-direction. Reproduced with permission. [78] Copyright 2012, American Chemical Society.
compared to other candidates of the tungsten dichalcogenide family. Yet, WSe 2 based electrical devices suffer from typical challenges, such as the high Schottky barrier formed between WSe 2 and contact metal, that so far hindered its evolution and implementation into commercial products. To overcome this obstacle, two main approaches are considered. On the one hand, contact metals with appropriate work function need to be chosen in preference to the type of the device. Accordingly, materials with high (low) work function in p-type (n-type) FETs are used to minimize the Schottky barrier. On the other hand, doping the contact region may be a further approach to reduce the barrier width. [84,114] Efforts have been devoted to compare the temperature dependent field-effect mobility and hole Hall mobility of few layered (9-15 layers) WSe 2 -FET on p-doped Si. [115] The maximum field-effect mobility of ≈350 cm 2 V −1 s −1 can be reached at room temperature while for Hall mobility for the maximum of 650 cm 2 V −1 s −1 lower temperature is required as at room temperature the latter value drops significantly to ≈200 cm 2 V −1 s −1 .
The tunnel field-effect transistors (TFETs) are other promising candidates of future devices due to their small subthreshold swing and low I OFF which both reduces power consumption. On a flexible and transparent substrate, TFET was built based on two graphene layers separated by WS 2 (acting as a barrier between the graphene contacts). [116] Interestingly, the highest I ON /I OFF ratio (10 6 ) was measured, when the number of layers was four or five. According to the systematic calculations, [71] in which different n-type sources (Mo and W chalcogenides) were combined with p-type drains (Ti, Zr, Hf, V, Nb, Ta chalcogenides), the best combination for vertical TFET was found to be W or Mo telluride and selenide combined with Zr or Hf selenide or sulfide. Simulation results suggest that WTe 2 monolayer TFET might have better performance than TFETs based on other TMDCs. [117] Due to the smaller band gap and the effective mass of electrons and electron holes, the I ON of the device is remarkably higher for WTe 2 than that for WSe 2 (127 µA µm −1 versus 4.6 µA µm −1 ). The performance can be further increased, reaching I ON of 350 µA µm −1 , by doping the WTe 2 lattice; moreover, it should be pointed out that its unique combination of properties makes WTe 2 potentially applicable in both homogenous and heterojunction n-and p-type TFET devices [118] with the highest I ON performance.
The possibility of altering the n-or p-type behavior of WSe 2 -FET just by the selection of the contact material allows the realization of complementary inverters by the insertion of n-type and p-type FETs on the same WSe 2 flake arranged in a double gated transistor geometry. [111,112] Typical contacts in p-type FETs are high work function metals such as Pt and Pd, in contrast, n-type construction can be enabled applying potassium doping the underlapped contact regions [111] (Figure 12) or using Ni instead. [112] Thus, a combination of the two on the same substrate provides the great opportunity as a platform for complementary metal-oxide-semiconductor (CMOS) devices and logic gates. As a result, typically dc voltage gain of WSe 2 based inverters is >12 [52,111] and value as high as ≈25 was reported for structures based on bilayer WSe 2 . [112] I ON /I OFF ratios and mobilities over 10 7 and 200 cm 2 V −1 s −1 , respectively, were accomplished in a single n-and p-type FET device relying on h-BN passivated WSe 2 channels and ionic liquid gated graphene contacts. [119] The work function of graphene was tuned by doping to minimize the height of the Schottky barrier at the graphene/semiconductor interface. Likewise, high carrier mobility as well as electron and electron hole mobility values of 90 and 7 cm 2 V −1 s −1 , respectively, were demonstrated when exploiting the benefits of ion gel, having high ionic concentration, applied as a gate in ambipolar WSe 2 monolayer electric double-layer transistor. [52]

Optoelectronics Devices
Optoelectronic devices able to generate, sense, control, or interact with light including solar cells, light-emitting diodes (LEDs) and photodetectors. Typically, materials having a direct band gap, large electron-hole pair binding energy and strong photoluminescence are promising candidates for such applications. In addition, a wide variety of transparent and flexible components are required for the realization of flexible and transparent optoelectronics, among them flexible, thin 2D TMDCs with typically direct and tunable bandgap are promising candidates. Photons having larger energy than the band gap can create either bound excitons or free carriers depending on their binding energy in the semiconductor, and therefore a direct band gap is demanded to achieve efficient absorption or emission of phonons. In contrast, for indirect bandgap, phonon absorption and emission process is less efficient owing to the required additional phonon participation to conserve the momentum. Since p-n junction is the functional element of many optoelectronic devices thus a considerable amount of efforts have been devoted to studying the interfacial behavior of WSe 2 monolayer by electrostatic tuning. [120][121][122][123] Their abundance coupled with their direct bandgap in the visible range makes TMDCs an appealing light absorbing material for alternative thin film solar cells, including flexible ones suitable to cover buildings and curved surfaces. Further advantages of the WDCs family include photostability under excitation and environment-friendliness. [124,125] Although high energy conversion efficiency of WS 2 solar cells were reported yet despite the technological advances it remains a great challenge to eliminate recombination centers at the surfaces perpendicular to the c-axis. For instance, a stable nanocomposite of 100 nm thick TiO 2 layer sensitized using 5 nm thick WS 2 was investigated as a potential absorber layer for electrochemical solar cells, however as the flow of the photocurrent is blocked at the direction perpendicular to the van der Waals surface of the WS 2 layer, into the electrolyte, the device efficiency is significantly reduced. The efficiency of excitonic solar cell based on WDCs may vary depending on the composition of the applied materials; for, for example, MoS 2 /WS 2 bilayer can attain power conversion efficiencies of 0.4-1.5% [126] while this value for MoS 2 /WSe 2 was showed to be as low as 0.2%. [127] Similarly, the power conversion efficiency of 0.5% was achieved of WSe 2 p-n junction solar cell device [121] in which laterally separated double gate electrodes were used to define the p-n semiconductor interface. Somewhat greater performance, power conversion efficiency as high as 1.5%, could be reached using WSe 2 /MoS 2 hetero p-n junction. [128] LEDs and transistors represent another potential application area for TMDCs applied as active materials. Similarly, to other optoelectronic applications, the main reason for their use is lying in the direct band gap of monolayers allowing much more efficient radiative recombination of electrons and holes that produce photon than in the case of indirect bandgap semiconductors.
Light emitting transistors of WS 2 mono-and bilayers with ionic liquid gate (Figure 13) with ambipolar behavior were demonstrated. [9] Also, an electrically switchable, circularly polarized light source from p-i-n junctions in monolayer and multilayer tungsten diselenide (WSe 2 ) was constructed, [129] in which the electron-hole overlap regulated by the in-plane electric field is assumed to be responsible for the generated circularly polarized electroluminescence. Such findings are expected to pave the way for electrically switchable circularly polarized light sources and broaden the functionality of valley-optoelectronics technology.
The ambipolar conductance inherently present in WSe 2 monolayers, enables p-n junctions to be configured via electrostatic control. Consequently LEDs may be constructed using two local gates to define a p-n junction. [121,123] One major advantage of such construction is that the device relies on a single sheet of WSe 2 . LEDs based on WSe 2 monolayers typically display electroluminescence efficiency between 0.1% -1%, [121,123] considerably less than that of commercial organic counterparts with external quantum efficiency as high as 15-40%.
A very recent study showed that large-area flexible transparent electroluminescent screens can be achieved with monolayers of MoS 2 , WS 2 , MoSe 2 , and WSe 2 by applying an alternating voltage between the gate and the semiconductor. Due to the Schottky contacts the voltage drop is large at the interfaces of semiconductor-metal contacts, which together with the steep energy band bending results in large transient tunneling currents. The induced excess electron and hole populations that are simultaneously present during the transients recombine and produce pulsed light emission (Figure 14). [130] Figure 12. a) False-color scanning electron micrograph and b) schematic drawing of a CMOS inverter on a single WSe 2 flake. c) Transfer characteristics of a WSe 2 p-FET and n-FET at V DS = |1 V| as a function of potassium doping time (1, 2, 3, and 5 min). The transfer characteristics of the p-FET remain the same after doping. d) Calculated contact resistance, R c , as a function of K doping time for FET devices. e) Voltage transfer characteristics of the WSe 2 CMOS inverter at different supply voltages, V DD . b) Differential voltage gains at V DD = 1, 2, and 3 V. Reproduced with permission. [111] Copyright 2014, American Chemical Society.
Few layered tungsten dichalcogenides, owing to their semiconducting nature, are sensitive to light irradiation. For example, our own investigation suggests that exfoliating bulk WS 2 powder in isopropyl alcohol can produce few layered WS 2 flakes; which when drop casted as thin films on pre-patterned electrodes showed ample photosensitivity (Figure 15).
Although, these flakes were found to be photosensitive, however, the performance parameters demanded of an outstanding photodetector such as photo-responsivity was found to be very low (approximately few mWA −1 ) attributed to the highly disordered nature of the film. The disordered nature is also evident from the variation of photocurrent measured in the function of laser intensity (Figure 15e). Such power law dependence is common in disordered semiconductors, and attributed to mid-gap states that play a decisive role in photocarrier relaxation. Even in its very basic exfoliated form, WS 2 materials were proven to be excellent candidates for devices requiring photo-switching as well as photoconduction as seen Figure 13. a,b) Optical microscope images of mono-(ML) and bilayer (BL) WS 2 flakes and c,d) their corresponding light-emitting FET devices with Au electrodes. e) The optical contrast in the R, G, B channels for mono-, bi-, and trilayer flakes. f,g) Schematic drawings of an ionic liquid gated FET in the f) unipolar and g) ambipolar injection regime (i.e., at low and high channel bias, respectively). h) Transfer curves of an ambipolar FET made of 1L WS 2 channel. i) Optical microscopy image of a 1L WS 2 device biased at V SD = 2.9 V at V G = 0.1 V. The red spot is due to electroluminescence. Reproduced with permission. [9] Copyright 2014, American Chemical Society. showing that emission of photons takes place in the proximity of the source contact edge (when an AC voltage is applied between the gate and source). Scale bar shows 10 µm. c) Electro and photoluminescence spectra of monolayer MoSe 2 , WSe 2 , MoS 2 , and WS 2 devices. d) Band diagrams for one cycle of rectangular gate modulation (−6 to 6 V peak to peak) showing electron/hole tunneling to the semiconductor during the transients at voltage increase/decrease and subsequent relaxation with photoemission. e) Photographs of millimeter-scale transparent devices. The area of the electroluminescent surface is 3 mm × 2 mm. Reproduced with permission. [130] Copyright 2018, Springer Nature.  Figure 15. Excellent light matter interaction was found in several other studies including both WS 2 as well as WSe 2 nanomaterials. For example; multilayer WS 2 phototransistor [131] was realized using the gold wire mask method having n-type semiconductor behavior, displayed an electron mobility of 12 cm 2 V −1 s −1 as well as a photosensitive switch with an on/off ratio (defined as I photo /I dark ) of 25. While other devices based on few-layered CVD grown WS 2 films [132] and on multilayers of WS 2 [133] showed responsivity of 2.1 × 10 −2 mA W −1 (about 3 orders of magnitude higher than exfoliated material) and 5.7 AW −1 , respectively. Furthermore, the latter one proved to have a fast response time of <20 ms, and external quantum efficiency of 1118%.
Several studies have also indicated the use of the liquid phase exfoliated materials in nanophotonics. Application of WS 2 based materials obtained via LPE method as an effective broadband saturable absorption material was also demonstrated. [152] In this study it was shown that by integrating WS 2 -based saturable absorbers into a thulium-holmium co-doped fiber ring cavity stable modelocked pulses with a temporal width of ≈1.3 ps at a repetition rate of 34.8 MHz can be generated at 1941 nm. Similarly, it was also demonstrated that not only WS 2 but other tungsten based materials, for example, WTe 2 microflakes [153] can be utilized as a base saturable absorption material for fast mode-lockers capable of generating femtosecond pulses from fiber laser cavities. These studies indicate the importance of TMDCs in the development of passively mode locked ultrafast lasers.
Although most of these studies mainly utilize tungsten based TMDC structures, however, it is becoming clear that 2D tungsten based monochalcogenide can have a very stable structure. For example, very recent density functional studies predict the possibility of stable structures of MoS and WS. [154] From that perspective it is also encouraging to explore other transition metal based mono chalcogenides several applications. For example Jnon et al. have shown [155] that it is possible to generate ultrafast femtosecond fiber lasers via mode locking at 1560 nm using SnSe. According to the authors, the figures of merit obtained using SnSe shows that performance of SnSe as saturable absorption materials is one of the best compared to other 2D-materials.
Phototransistor relying on a single layer of WSe 2 [134] contacted by either Ti or Pd electrodes shed light on the importance of contact material in relation to the performance of these devices. Accordingly, WSe 2 phototransistor with low Schottky-contact (Pd contact) displays excellent photo gain (1.8 × 10 5 ) and detectivity (10 14 jones), but with a slow response time of more than 5 s. In contrast, with high Schottky barrier Ti contacts faster response times are feasible (<23 ms), but in the expense of photo gain and detectivity. Further findings on a phototransistor built on few layers of P-WSe 2 indicate the importance of electrical measurement configuration, corroborated by the considerably large photoconductivity difference observed for four-terminal configuration compared to a two terminal one. For an incident laser power of 248 nW the obtained responsivity and external quantum efficiency for two terminal configuration were found to be 18 A W −1 and 4000%, respectively. This value shows response and external quantum efficiency increase of 370% when using a four terminal configuration. [135] So far, the highest responsivity value of 7 AW −1 under white light illumination was reported on three atomic layers of chemical vapor transport grown WSe 2 photodetectors, [136] surpassing values that of graphene and transition metal dichalcogenides based heterostructures. Recent low temperature investigations of photo conductive behavior revealed considerably weak temperature dependence of responsivity suggesting that tungsten dichalcogenides may find use in photodetectors requiring temperature stability. By reducing the contact resistance and blocking the charge puddle effect through triphenylphosphine (PPh 3 ) and aminopropyltriethoxysilane (APTES) treatment technique the responsivity of a WSe 2 device [137] was boosted as high as 1.27 × 10 6 A W −1 . Furthermore, about twofold enhancement in photoresponse time, from about 420 to 200 ms, was achieved by switching contact metal from Pt to Ti.

Chemical and Biochemical Sensors
TMDCs are sensitive to changes in their surroundings due to their high surface to volume ratio making them promising candidates for sensors. Resistive type gas sensors using WS 2 nanowire-nanoflake hybrid materials as sensing elements were assessed in various redox gases (Figure 16). Since WS 2 Figure 16. a) Resistive sensing of sub-ppm concentrations of H 2 S buffered in synthetic air using a tangled network of WS 2 nanowire-nanoflake hybrid material. b) Sensor sensitivities for different analytes buffered in air. Reproduced with permission. [86] Copyright 2018, Springer Nature.
behaves as a p-type semiconductor thus shows a positive resistance response for reducing gases (H 2 S, NH 3 , H 2 , CO) and negative resistance response for oxidizing ones (NO). Excellent sensitivity (0.043 ppm −1 ) and selectivity for H 2 S gas was demonstrated in air buffer. As revealed by the first principle calculations and XPS analysis, the sensing mechanism for H 2 S is not based on a simple analyte adsorption on and induced carrier doping in the semiconductor but because of a competitive and reversible substitution of oxygen and sulfur atoms in the lattice. Without H 2 S present in the air carrier gas, S is partially replaced with O, which then substituted with S as soon as even traces of H 2 S were introduced in the gas stream. [86] Room temperature NH 3 sensing with a detection limit of 1.4 ppm was demonstrated using WS 2 thin films synthesized by reactive plasma-assisted conversion of WO 3 in the presence of H 2 S. [138] Improvement of gas sensing performance on 2D WS 2 nanosheets functionalized by Ag nanowires was shown. [139] Although pristine WS 2 displayed good response to acetone and NO 2 , it suffered from an incomplete recovery after NO 2 detection, which could be circumvented by the addition of Ag. Multilayer WS 2 nanoflake FET devices were studied for their photo and gas sensing behavior. [133] Photoresponsivity and external quantum efficiency of the devices during NH 3 and ethanol exposure (acting as "n-dopants") were found to increase, whereas in oxidizing atmosphere the opposite occurred.
Apart from resistive devices, also electrochemical sensors based on tungsten dichalcogenides were reported during the past few years. For instance, layered WS 2 -graphene nanocomposite electrochemical sensors were proved to be suitable for the detection catechol, resorcinol and hydroquinone [140] with detection limits of 1 × 10 −7 mol L −1 for hydroquinone and resorcinol, and 2 × 10 −7 mol L −1 for catechol and recoveries of 93.6-104.8%. In another work, a sensitive electrochemical biosensor of DNA with WS 2 -graphene-cithosan composite electrodes was demonstrated capable of detecting the analyte over a large dynamic concentration range from 0.01 to 500 pm with a detection limit of 0.0023 pm. [141] Furthermore, when doping singlelayer WS 2 with species of high electron affinity (such as F or 2,3,5,6-tetrafluoro-7,7,8,8-tetracyanoquinodimethane), [156,157] an enhancement of the photoluminescence peak at ≈2.0 eV is observed due to a change in the concentration of excitons in WS 2 . The intensity of the enhanced emission could be modulated (reduced) by exposing the doped nanoflakes to NH 3 , which acts as an electron donor, with a demonstrated lowest detection limit of 1.25 ppm. [156] Capitalizing on the strong binding affinity of WS 2 toward peptide nucleic acids (PNAs) and on its good fluorescence quenching ability for fluorophore-labeled PNA sequences, a new biosensor was developed lately with a detection limit of 500 pm for DNA. [142] Combination of the technique with nucleic acid amplification technique, sensors for analyzing miRNAs with a detection limit of 300 fm were shown. [143] Furthermore, composites of WS 2 and acetylene black functionalized by DNA probes enable also electrochemical biosensor with very large dynamic concentration range for analysis (0.001 m to 100 pm) with sub-fm detection limit. [144] Interestingly, WS 2 nanosheets were found to have also intrinsic peroxidase-like activity, shown to catalyze the oxidation of peroxidase substrates (e.g., 3,3′,5,5′-tetramethylbenzidine, TMB) in the presence of H 2 O 2 resulting in a change of color. This effect was then exploited in selective and sensitive colorimetric detection of blood glucose (with a detection limit of 2.9 µm) as H 2 O 2 produced in the glucose oxidase-catalyzed reaction could be detected through the oxidation of TMB giving a blue-color. [145]

Catalytic, Electrocatalytic, and Energy Application
Layered transition metal sulfides are often used as catalysts by the petroleum industry in order to increase the hydrogen content of their products achieved by the hydrogenation of aromatic molecules and by the removal of contaminants. Layered transition metals are capable to catalyze hydrocarbon-upgrade reactions even when large amounts of contaminants such as sulfur, nitrogen and metals are present in the reaction. While the catalytic properties and use of layered transition metal sulfides, especially MoS 2 , in catalysis can be read elsewhere, [146,147] yet surprisingly there are hardly any research publications, to the best of our knowledge, published on applying mono-to few layered tungsten dichalcogenides as the catalyst.
Instead, probably due to the great demand for new energy conversion and storage materials and technologies, the electrocatalytic activity of TMDs have been studied, in particular for the hydrogen evolution reaction to replace costly platinum group metals in the electrodes. The first demonstration of the hydrogen evolution reaction on WS 2 nanosheets obtained by a sulfurization of WO 3 nanoparticles (Figure 17) revealed only ≈0.1 V overpotential of H 2 evolution in reference to Pt. [148] The promising results fostered further studies in the field to reveal any peculiarities of WS 2 nanomaterials.
As measured on chemically exfoliated 1T-WS 2 and 2H-WS 2 monolayers, the lowest overpotential and Tafel slope was observed for 1T-WS 2 catalyst, followed by the 2H phase and bulk performing only poor HER activity (Figure 18). [18] The higher activity of the 1T phase was explained by the presence of strained metallic regions in the exfoliated nanosheets thus driving the focus of future quest on 1T-WS 2 with enhanced HER activity. [149] Furthermore, triangular domains of WS 2(1−x) Se 2x monolayers grown by CVD were shown to be performing even better than plain monolayers of WS 2 or WSe 2 . [150] The enhancement is proposed to be caused by the strained and thus activated basal plane of the lattice, which is otherwise rather inert. [150] WSe 2 has also been shown to have measurable steady-state photocurrents in photoelectrochemical HER. Doping the nanoflakes with Pt further increased the activity of H 2 evolution. [151] The activity differences between metallic 1T-WS 2 and semiconducting 2H-WS 2 nanosheets in the photocatalytic hydrogen evolution reaction has also been studied. [68] Addition of 1T-WS 2 to TiO 2 increases the photoactivity more than threefold, however, at the same time addition of 2H-WS 2 decreases the photoactivity more than threefold. The reasons for the very different activities of WS 2 phases can be explained by the rapid transfer of photogenerated electron from TiO 2 to the metallic 1T phase WS 2 and reduction of a proton in the aqueous solution. Whereas in semiconducting 2H phase acts as a light harvesting material and transfers the electron to TiO 2 after photon absorption. Although, 1T-WS 2 is effective co-catalyst in photo catalytic HER, it is worth pointing out that 1T-WS 2 -TiO 2 does not show any photoactivity under the visible light range contrary to 2H-WS 2 -TiO 2 , which exhibits slight H 2 production.
Finally, superacid-treated WS 2 nanoflakes were proved to be suitable as an anode material in the Li-battery application. [48] The resulted device showed three-step charge-discharge behavior, typical for binary transition metal compound-based Li-batteries. The measured first-cycle reversible electrochemical capacity, with 25 mA g −1 current density and voltage hysteresis of 0.93 V, was 470 mAh g −1 , whereas after 50 cycles the capacity declined remarkably to 118 mAh g −1 . The higher first-cycle value was speculated to be caused by intercalation of Li ions between the nanosheets.

Summary and Conclusive Remarks
Monolayer TMDs have recently emerged as intriguing candidates for transistor and optoelectronic devices owing to their excellent properties, such as direct band gap, chemical stability, mechanical flexibility, large in-plane carrier mobility, and photoluminescence. Although some figures may seem lower than that of graphene or ordinary semiconductors (Si, Ge, and GaAs), the application of 2D chalcogenides allow us to overcome specific shortcomings of graphene (need doping to open band gap) and conventional semiconductors (lack of mechanical flexibility and printability). Furthermore, future optical devices could greatly benefit from the low direct band gap of monolayered TMDs that can be tuned by tensile strain or external electric field. Figure 18. a) Polarization curves and b) Tafel slopes for HER over bulk and as-exfoliated WS 2 (as-deposited film of 1T phase, sub-monolayer as-exfoliated film, and 2H phase after annealing at 300 °C). Reproduced with permission. [18] Copyright 2013, Springer Nature. Figure 17. a,b) SEM and (c) TEM images of WS 2 nanosheets synthetized by a sulfurization of WO 3 nanoparticles. d) Polarization curves and e) Tafel plots measured in 0.5 m H 2 SO 4 for various electrocatalysts mounted on glassy carbon disk with the help of Nafion. Reproduced with permission. [148] Copyright 2012, Elsevier B. V.
The progresses made in understanding the key properties of tungsten based chalcogen materials so far is incredible; however, there are several crucial aspects relevant to synthesis, structural modifications, assembly, and application developments that need to be addressed, if we want these materials to be the forerunners in technology innovations. For example, from the synthesis point of view, issues of scalability still lingers. So, one of the key research directions will be to fine tune the production processes reported so far or discover new growth strategies in order to achieve consistency in production of large quantities of high quality uniform mono-, bi-, or few-layered materials. Another open area is exploring the possibility of alloying tungsten-based chalcogens with other elements such as copper for creating ternary materials with substantially different properties, for example, tunable optical/electronic band gaps. Such alloying will also provide a simple way of creating unexplored materials having an in plane heterostructures. The above-mentioned research direction also calls for extensive use of theoretical calculations, predictive modeling, and/or machine learning approaches to narrow down the daunting task of experimental discovery and design of these materials with advanced functionalities. Similarly, defect engineering through physical and/or chemical processes for controllably creating atomic vacancies in these materials in order to develop efficient electrocatalytic and energy materials, sensors etc. could be other directions of investigation. Tremendous opportunity also lies in assembling and investigating vertical heterostructures of layered tungsten based chalcogens with other layered structures in order to create transformative electronic/photo-electronic characteristics for device applications. At this juncture, the prospect for a variety of niche applications using tungsten based chalcogen materials looks extremely promising, however, the fruition of such proposed application will need dedicated and continual research efforts.