|Home | About | Journals | Submit | Contact Us | Français|
The discovery of graphene marks a major event in the physics and chemistry of materials. The amazing properties of this two-dimensional (2D) material have prompted research on other 2D layered materials, of which layered transition metal dichalcogenides (TMDCs) are important members. Single-layer and few-layer TMDCs have been synthesized and characterized. They possess a wide range of properties many of which have not been known hitherto. A typical example of such materials is MoS2. In this article, we briefly present various aspects of layered analogues of graphene as exemplified by TMDCs. The discussion includes not only synthesis and characterization, but also various properties and phenomena exhibited by the TMDCs.
This article is part of the themed issue ‘Fullerenes: past, present and future, celebrating the 30th anniversary of Buckminster Fullerene’.
The discovery of fullerenes by Kroto et al.  initiated a new season of vital interest in low-dimensional materials as well as nanoscience. Carbon nanotubes  which are extended fullerenes gave a further thrust to these areas. The more recent discovery of the novel properties of graphene [3,4], the mother of all graphitic carbons, has particularly enthused the scientific community to explore two-dimensional (2D) materials. Fullerenes and nanotubes of layered inorganic materials such as MoS2 have been investigated extensively [5–7]. There is now a surge in research on layered 2D inorganic materials such as transition metal dichalcogenides (TMDCs). The 2D inorganic analogues of graphene constitute a new class of electronic materials and offer exciting avenues for fundamental and applied research in a wide variety of fields such as spintronics and valleytronics, electronics, optoelectronics, photovoltaics, catalysis, sensing and energy storage [8–11]. Of the TMDCs, MoS2, in particular, has attracted greater attention due to its unique electronic, optical, magnetic and mechanical attributes, with a variety of potential applications . Isolated atomic planes of these 2D materials can be reassembled into designer heterostructures bound by van der Waals interactions, with precise control over the sequence. These van der Waals heterostructures are remarkably complex and exhibit unusual properties and new phenomena [13–16]. Blends of graphene and 2D inorganic layered materials are also found to exhibit novel properties due to synergy . Layered TMDCs with the generic formula MX2 (M=transition metal, X=chalcogen) are van der Waals solids with strong intralayer bonding and weak interlayer bonding. An individual layer of TMDCs constitutes three covalently bonded atomic planes forming a X–M–X sandwich. Quantum confinement of charge carriers and changes in the interlayer coupling and symmetry elements lead to dramatic changes in the electronic structure and thereby the properties of single- and few-layer TMDCs relative to their bulk counterparts . The present article provides a perspective of the recent advances in the area of 2D TMDCs. The aspects covered include synthesis and characterization, electronic structure and properties, Raman spectra, magnetic properties, field-effect transistors, catalysis, Li and Na ion batteries, supercapacitors, valleytronics and trions, giant magnetoresistance (MR) and superconductivity.
Owing to anisotropic bonding, individual layers of TMDCs can be obtained by several physical and chemical methods, the choice of synthesis method usually varying with the relevant application. Important physical methods to synthesize single- or few-layer TMDCs comprise micromechanical exfoliation (scotch-tape technique), liquid-phase exfoliation, laser-thinning and sputtering [4,18]. Mechanical exfoliation provides electronic grade monolayers suitable for high-performance devices and for studies on condensed matter phenomena [11,12,19,20], but is limited to short-scale production. Liquid-phase exfoliation via ultrasonication in a solvent medium provides a route to large-scale production of large-area single- and few layers of a number of layered materials for applications in catalysis and electrochemical storage. Liquid-phase exfoliation has been employed to successfully synthesize mono- and few layers of several materials such as MoS2, MoSe2, MoTe2, WS2, TaSe2, NbSe2 and NiTe2, h-BN, Bi2Te3 . Sonication techniques rely on the solvent (or surfactant) to overcome the cohesive energy between the neighbouring layers of the solute and, therefore, as a crude criterion, the surface energies of the dispersing solvent and the exfoliated material should match [21,22]. More specifically, the dispersion of a solute material in a given solvent depends on the minimization of exfoliation energy through optimizing the balance of solute–solute, solute–solvent and solvent–solvent binding energies. The Hansen solubility parameters are related to dispersion (δD), hydrogen (δH) and polar bonding (δP) contributions to the cohesive energy density. The energy of exfoliation is represented by enthalpy of mixing and the latter is given by ΔHmix=(1−)[(δD,A−δD,B)2+(δP,A−δP,B)2+(δH,A−δH,B)2)], where A, B and denote solute, solvent and volume fraction of solute, respectively . Hence for minimization of ΔHmix, all three Hansen solubility parameters of solvent should match those of solute. After screening a large number of solvents, dimethylformadimide (DMF), isopropyl alcohol (IPA) and N-methyl-2-pyrrolidone (NMP) have been shown to be most suitable solvents for dispersing a variety of TMDCs, owing to combination of suitable surface energies and Hansen solubility parameters . Figure 1a,c shows photographs of dispersions of MoS2, and WS2 in NMP, prepared by liquid-phase exfoliation and the corresponding transmission electron microscope (TEM) images which show transparent ultrathin layers with lateral dimensions spanning hundreds of nanometres. High-resolution TEM (HRTEM) images of single layers obtained by this route show no deviation from the 2H-structure (discussed below) of the bulk counterparts (figure 1c,d). A mixed-solvent strategy was used by Zhou et al.  wherein low-boiling solvent mixtures like ethanol–water in appropriate ratios are used to obtain MoS2 and WS2 monolayers in good yields. The optimum ratios of ethanol to water is determined by the Hansen solubility parameter. A variety of layered crystals including TMDCs, transition metal oxides and BN have been successfully exfoliated in aqueous medium by ultrasonication with sodium cholate as a surfactant to prevent re-aggregation of the dispersed nanosheets. Yao et al.  combined low-energy ball milling with ultrasonication for exfoliation of various inorganic layered materials into nanosheets. Ultrasonication has an undesired effect of breaking the nanosheets and reducing their lateral dimensions. However, the proportion of the moderately large (greater than 1μm) and thin nanosheets can be optimized with a careful choice of the starting mass, sonication time and centrifugation conditions. But the yield of monolayers through liquid-phase exfoliation is not very high. Free-standing nanosheets of VS2  and VSe2  are obtained by refluxing the bulk materials in solvents like formamide. Intercalation of solvent molecules in between the layers causes swelling of bulk VS2 and VSe2, and causes exfoliation to yield dispersions of nanosheets. Laser-thinning of multilayered MoS2 is another reliable approach to obtain single-layer MoS2 with defined shape and size in addition to retaining the semiconducting properties of the pristine material . Irradiating dispersions of bulk TMDCs (MoS2, WS2, MoSe2 and WSe2) in DMF by a KrF excimer laser can also produce single- and few-layer flakes .
Turning to chemical methods, one of the most effective routes to the mass production of fully exfoliated TMDC nanosheets is the lithium intercalation–exfoliation procedure. It involves soaking of the bulk material in hexane solutions of n-butyl lithium (typically 3–4 days at 100°C or one to two weeks at room temperature, for MoS2) followed by exfoliation in ultrasonicated water during which profuse evolution of H2 gas occurs . The yield of single-layer TMDCs is nearly 100%. Single layers of MoS2 , MoSe2 , WS2 , TiS2 , TaS2  have been obtained through this method. An important aspect of this route is the transformation of the thermodynamically stable semiconducting 2H phase of Mo(S,Se)2 and W(S,Se)2 into metastable metallic 1T phases . Both 2H and 1T phases are composed of X–M–X (M=metal, X=chalcogen) sandwich structures. The letters, H and T stand for hexagonal and trigonal, respectively, while the digit indicates the number of X–M–X stacks along the c-axis in a single unit cell. Figure 2c shows the schematic representation of 2H- and 1T-MoS2 viewed down the crystallographic c-axis. 2H-MoS2 has trigonal prismatic coordination of Mo and S atoms (figure 2c). The S atoms of upper plane lie directly over S atoms in the lower plane giving rise to AbA BaB type of stacking sequence (the upper and lower case alphabets denote X and M, respectively). In the 1T phase, the stacking sequence is of AbC AbC type with the S atoms in the upper and lower planes being off-set from each other by 30° such that the Mo atoms lie in the octahedral voids of the S-layers (figure 2c). Consistently, the electron diffraction pattern of 2H-polytpe shows a hexagonal diffraction pattern of spots (figure 2a), whereas that of 1T phase shows extra spots at 30° angular spacing in between the hexagonal spots of the 2H structure (figure 2b). Figure 2d,e shows the HRTEM images of 2H- and 1T monolayers, respectively. It can be seen from magnified HRTEM images (viewed down the c-axis) that for 2H monolayer, each Mo is surrounded by three S atoms (figure 2d, left inset), whereas six S atoms surround a Mo atom in the 1T-structure (figure 2e, left inset) . The metallic 1T phase is induced through charge transfer from Li to the TMDCs which causes a local atomic rearrangement from the 2H to the 1T phase. The 1T phase of nanosheets obtained via Li-intercalation into TMDCs is metastable even after exfoliation through passivation of residual negative charge by water bilayer [9,36,37]. Chhowalla and co-workers [9,36,37] have identified the 2H and 1T phases of MoS2 by carrying out high-resolution scanning transmission electron microscope (STEM) imaging and shown that the samples prepared by Li-intercalation and exfoliation exist as coherent heterostructures of the 1T and 2H phases (figure 2f). Because 1T- and 2H-MoS2 are metallic and semiconducting, respectively, their coexistence represents novel heterojunctions with possible application in molecular electronic devices. High-angle annular dark field (HAADF) imaging in an aberration-corrected STEM has revealed a high concentration of the zigzag-like superlattices identified as strained metallic 1T phase in the exfoliated WS2 nanosheets. The enhanced electrocatalytic hydrogen evolution reaction (HER) activity of WS2 nanosheets has been ascribed to the presence of the tensile strain induced by zigzag-like distortion in the 1T phase . Interestingly, while Mo(S,Se)2 and W(S,Se)2 undergo the transition from the 2H phase to the 1T phase upon Li-intercalation, TaS2 undergoes the reverse transition, i.e. from the 1T phase to the 2H phase .
Despite nearly 100% yield of single-layer TMDCs, the Li-intercalation and exfoliation method suffers from drawbacks such as long lithiation time, difficulty in controlling the degree of lithium insertion and submicrometre flake-size. Zheng et al.  have reported a two-step expansion and intercalation method to produce large (up to 400μm2) single-layer MoS2 flakes. Firstly, bulk MoS2 is reacted with hydrazine (N2H4) under hydrothermal conditions (at 130°C for 48h) where a part of N2H4 rearranges in a redox fashion to N2H5+ upon intercalation. The latter being thermally unstable, decomposes to N2, NH3 and H2 at high temperatures, thereby expanding the MoS2 sheets by greater than 100 times compared with the original volume. Secondly, the expanded MoS2 flakes are intercalated by alkali naphthalenide followed by exfoliation in ultrasonicated water operated at low power to avoid fragmentation of the sheets. Figure 3a–d shows AFM and TEM images of as-exfoliated MoS2 and WS2 monolayers. Sodium naphthalenide has been shown to be most effective for obtaining large-area flakes. High yields of monolayer TiS2, TaS2 and NbS2, as well as few-layer TiSe2, NbSe2 and MoSe2 have been prepared by this route. Jeong et al.  reported a tandem molecular intercalation (TMI) strategy (figure 3e) where a short initiator alkylamine intercalate expands the interlayer gap of TMDCs for an efficient intercalation of another long primary alkylamine. Owing to the difference in chain length, a bilayer arrangement with empty spaces between intercalates is adopted to reduce van der Waals force between them and spontaneous exfoliation occurs finally to yield single-layer TMDCs. The TMI process is advantageous in that it is a simple one-step process at room temperature without the need for harsh exfoliation processes involving ultrasonication or H2 generation. While a relatively weaker Lewis base such as alklyamine is effective for group IV and V TMDCs, a stronger Lewis base such as alkoxide is required for group VI TMDCs. The TEM images in figure 3f–k shows the step-wise TMI process during exfoliation of multilayer WSe2 into single layers. By employing appropriate intercalates, single-layer nanostructures of group IV (TiS2, ZrS2), group V (NbS2) and VI (MoS2, WSe2) have been successfully prepared.
Zeng et al.  report synthesis of nanosheets of various metal chalcogenides such as MoS2, WS2, TiS2, TaS2 and ZrS2 by electrochemical Li-intercalation and subsequent exfoliation in water. Li-intercalation is carried out by controlled galvanostatic discharge at 0.05mA current density in a standard battery test system with the layered bulk material as the cathode and the Li-foil as the anode. During the electrochemical discharge, the Li-foil anode is oxidized, giving out electrons that flow towards cathode. Simultaneously, Li-ions migrate through the electrolyte (LiPF6) from anode to cathode where they intercalate in between the layers, and get reduced to Li-metal by the incoming electrons. Subsequently, the lithiated material is ultrasonicated in water or ethanol during which, intercalated Li-metal reacts with water or ethanol to form Li(OH) and H2 gas. The latter expand the interlayer distance and cause exfoliation to give dispersions mainly containing single layers. Unlike the case of solution-phase Li-intercalation, the authors did not observe any phase-transition from 2H form of bulk to 1T form of monolayers upon electrochemical Li-intercalation of MoS2 . Insufficient insertion of Li leads to ineffective exfoliation while excess Li-insertion leads to the decomposition of the crystal. Zeng et al.  have optimized the cut-off voltage and discharge current needed for the optimum Li-intercalation necessary for preparation of few-layer BN, NbSe2, WSe2, Sb2Se3 and Bi2Te3.
Another widely employed chemical method for fabricating large, highly uniform and electronic grade TMDC monolayers is chemical vapour deposition (CVD). CVD of TMDCs generally employs four strategies (i) sulfurization or selenization of pre-deposited thin films of metal or metal oxides, (ii) vapour-phase reaction of chalcogen and metal-based precursors, (iii) thermolysis of single-source precursors, and (iv) vapour-phase transport of TMDC powders. A pre-deposited thin film of Mo on SiO2 substrate can be converted to a MoS2 thin film by annealing in the presence of elemental sulfur . However, both single- and few-layer MoS2 coexist on the substrates due to the challenge of forming a uniform metal thin film. On the other hand, single-layer MoS2 is obtained if S instead of Mo is preloaded on a Cu surface such that the sulfur source is available only to generate the first MoS2 monolayer . Wafer-scale uniform MoS2 thin films with controlled thickness have been grown by layer-by-layer sulfurization of a MoO3 thin layer. A MoO3 thin film with the desired thickness is first grown on sapphire substrates by thermal evaporation and then reduced to MoO2 or other Mo forms in a H2/Ar environment at 500°C. Subsequent annealing in a sulfur-rich environment at 1000°C leads to the formation of a wafer-scale MoS2 thin layer which can be transferred to arbitrary substrates for electronic device fabrication (figure 4a–c) [44,48]. Large-area WS2 sheets (approx. 1cm2) with controllable thickness have been synthesized using a similar strategy by sulfurization of thermally evaporated WOx thin films .
Continuous films of single- and few-layer MoS2 have been prepared by the vapour-phase reaction of MoO3 and S powders as precursors on SiO2/Si substrates coated with graphene oxide and perylene-3,4,9,10-tetracarboxylate of potassium. The organic coating promotes the growth of MoS2 layers by wetting of the growth surface and lowering the free energy for nucleation . Ternary MoS2xSe2(1−x) nanosheets with tuneable band edge emission have been obtained by Li et al.  by a similar vapour-phase reaction of MoO3 with sulfur and selenium powders (figure 4d and e). Few-layer MoS2 has been grown on a graphitic surface by the thermal reduction of amorphous MoS3 in the presence of reduced graphite oxide at 1000°C under high-vacuum conditions . However, achieving a complete coverage of the substrate was not possible by this method. Three-layer MoS2 sheets can be grown on a variety of insulating substrates by first dip-coating in ammonium thiomolybdate ((NH4)2MoS4) solution followed by annealing under Ar/H2 flow at 500°C and subsequently at 1000°C (figure 4h) . The chemical reaction involved is (NH4)2MoS4+H2→2NH3+2H2S+MoS2. Wu et al.  have reported a vapour–solid (versus) growth method for synthesizing MoS2 monolayer on insulating substrates such as Si wafer and sapphire, by physical vapour transport of MoS2 powder at 900°C (hot zone) under a low pressure of 20 Torr (figure 4f,g). The substrates were placed in a cold zone at 650°C . Feng et al.  have also reported the synthesis of uniformly distributed monolayer MoS2(1−x)Se2x alloys on SiO2/Si substrates through the direct evaporation of MoSe2 and MoS2 powders at a high temperature of approximately 950°C followed by downstream alloying at approximately 650°C. In all the above methods, careful optimization of reaction conditions at several stages in the process is highly desirable to ensure uniformity of the films especially over large area, and defect-free growth for achieving high-quality films.
Other chemical strategies include thermal decomposition of precursors and hydrothermal synthesis. Heating a mixture of molybdic acid or tungstic acid and excess thiourea or selenourea (1:48 ratio) at 773K for 3h yields few-layer samples of MoS2 or MoSe2 and WS2 or WSe2 on a large scale [30,53]. Figure 5a shows the X-ray diffraction patterns of few-layer samples where the (002) reflections are suppressed due to ultrathin nature of few-layer nanosheets. Figure 5b shows a representative HRTEM image of a single-layer MoSe2 synthesized by the above thermal decomposition route. Few-layer MoS2 is obtained by the reaction of MoO3 and KSCN in aqueous solvent at 453K and few-layer MoSe2, by the reaction of molybdic acid and selenium metal in aqueous NaBH4 solution at 453K [30,53]. With the same precursors, few-layer MoS2 and MoSe2 were obtained under microwave conditions by using ethylene glycol as the solvent . Free-standing nanosheets of MoS2 and WS2 can be synthesized by the decomposition of single-source precursors, such as (NH4)2MoS4 and (NH4)2WS4, respectively, in oleylamine at 360°C . Mahler et al.  have recently reported a colloidal synthesis of 2H-WS2 and distorted 1T-WS2 monolayers in oleylamine using WCl6 and CS2 as W- and S-precursors, respectively. 2H-WS2 forms instead of 1T-WS2 when hexamethyldisilazane (HMDS) is added to oleylamine . Figure 5c shows the low- and high-resolution TEM images of 1T- and 2H-WS2 monolayers formed in the absence and presence of HMDS, respectively. Yoo et al.  have recently developed a solution-based protocol called ‘diluted chalcogen continuous influx’ (DCCI) for the synthesis of single-layer nanosheets of group IV metal disulfides. The continuous influx of dilute H2S gas throughout the growth period is a key to obtaining large nanosheets through the exclusive lateral growth processes. 1-Dodecanethiol was used as the sulfur precursor for the slow in situ formation of H2S which reacts with the metal chlorides (MCl4, M=Ti, Zr, Hf) to form large single-layer nanosheets of MS2 . Figure 5d shows TEM and AFM images of single-layer TiS2 nanosheets obtained through above DCCI method.
Bulk 2H-MoS2 is centro-symmetric (P63/mmc) and is ans indirect-band gap semiconductor. The calculated electronic band structures show that bulk 2H-MoS2 has an indirect-band gap (approx. 1.2eV) with the valence band maximum (VBM) and conduction band minimum (CBM) at the Γ point and midpoint along Γ–K path, respectively, in the reciprocal space (or k-space). The Γ point lies at the centre, whereas K points are at the six corners of the hexagonal Brillouin zone in the reciprocal space (figure 14a). Besides, a higher energy direct gap (approx. 1.8eV) exists between CBM and VBM at the K point (figure 6a, inset) [57–60]. The indirect-band gap (along Γ–K) increases with decreasing number of layers, and exceeds the direct-band gap in the limit of monolayer (figure 6c) [57,59,60]. Thus, monolayer MoS2 becomes a direct-band gap semiconductor. The direct-band gap (at the K point) hardly changes with layer thickness because the conduction band states at the K point are primarily composed of strongly localized d-orbitals on Mo atoms, which have minimal interlayer coupling because Mo atoms are sandwiched between two S-planes (S–M–S) in the unit cell. On the other hand, states associated with the indirect-band gap (along Γ–K) originate from a linear combination of d-orbitals on Mo atoms and anti-bonding pz-orbitals on S atoms and hence have strong interlayer coupling and exhibit strong energy dependence on layer thickness [57,59,60]. Confinement of carriers owing to suppressed interlayer hopping in the ultrathin regime results in an increasing indirect-band gap as number of layers decrease. This indirect-to-direct gap crossover from bulk to monolayer is also evident from a sudden increase in photoconductivity at approximately 1.8eV in the monolayer MoS2 as a consequence of the latter being a direct-band gap semiconductor . Monolayer MoS2 (space group, P-6 m2) has time-reversal symmetry [E↑(k)=E↓(−k)] but lacks inversion symmetry [E↑(k)=E↑(−k)] due to which the spin–orbit coupling (SOC) lifts the spin degeneracy of otherwise twofold degenerate valence bands at the K point into two bands with spin-up and spin-down character (figure 14a) [61,62]. Thus two direct excitonic transitions, namely A1 and B1, are allowed at the K point (figure 6a, inset). Consistently, the experimental reflectance and photoluminescence (PL) spectra of monolayer MoS2 reveal two peaks at 670nm and 627nm corresponding to A1 and B1 direct excitonic transitions, respectively (figure 6a and b) [57,63,64]. As a consequence of indirect to direct-band gap transition, PL is significantly enhanced in monolayers of MoS2 relative to its multilayered counterparts (figure 6d). The quantum efficiency of PL in a monolayer MoS2 is about 104 times that of the bulk material [57,60] and the enhanced PL emission from MoS2 monolayer has been attributed to the significant lowering of the intraband relaxation rate [57,60]. On the other hand, PL emission is lost in the metastable metallic 1T-MoS2 monolayers generated by Li-intercalation and the excitonic features emerge only upon annealing up to 300°C when the 2H phase is restored .
Besides the number of layers, the band gap in group VI TMDCs can be tuned by application of electric field or a mechanical strain. For instance, the indirect-band gap of bilayer TMDCs can be driven to zero at an electric field of 2–3Vnm−1 applied perpendicular to the layers, allowing for larger band gap tuneability than that in graphene . Under strain, the band gap of mono- and bi-layer MoS2 decreases and the material undergoes an insulator-to-metal transition [66–69]. Chemical stimuli can also modulate the band gap. MoS2 interacts with electron-donating tetrathiafulvalene (TTF), which forms a radical cation by donating an electron. However, MoS2 being p-type, does not interact with electron-accepting tetracyanoethylene (TCE). First-principles calculations show a large reduction in the band gap of MoS2 upon interaction with TTF molecules . Electron doping can be induced by gate voltage in a field-effect transistor (FET) , though hole-doping is not possible by this means in contrast with graphene which can be doped by both electrons and holes.
Raman spectroscopy has been a powerful analytical tool for determining thickness and stacking of 2D TMDC layered materials to study their mechanical and thermal properties and to directly probe and monitor the charge-doping. Group theory predicts 2H-MoS2 to exhibit four first-order Raman modes, E22g, E1g, E12g and A1g with frequencies at 32, 286, 383 and 408 cm−1, respectively [11,72]. E12g and A1g are the only intense modes which correspond to the in-plane (intralayer) and out-of-plane (interlayer) vibrations, respectively (the inset of figure 7b) . Upon reducing the number of layers, the A1g mode softens and the E12g mode stiffens (figure 7a and b). Within the model of coupled harmonic oscillators, both the A1g and E12g modes are expected to stiffen with increasing number of layers as the interlayer van der Waals interactions increase the effective restoring forces acting on the atoms . While the shift in A1g mode is consistent with the model, the anomalous shift in the E12g mode was ascribed to the dielectric screening of long-range Coulomb interactions which increase with increasing number of layers [73,76]. For 2H-MoSe2, the A1g vibration is the most intense and occurs at a lower frequency than the E12g vibration . The E12g mode stiffens and the A1g mode softens as the number of layers is reduced [77,78]. In bulk 2H-WSe2, the E12g and A1g modes degenerate, giving rise to only a single Raman band. When the dimensionality is decreased, this peak splits into two, with the E12g band at a frequency lower than that of the A1g band and with greater intensity 
Raman bands of mechanically exfoliated monolayer TMDCs show significant temperature dependence. The temperature coefficients (α) of the A1g and E12g modes of MoS2 were found to be −0.0123cm−1K−1 and −0.0132cm−1K−1, respectively [80–82]. The value of α for the A1g mode in monolayer MoSe2 is much smaller (−0.0045×10−2cm−1K−1) than that of monolayer MoS2 due to difference in the strain-phonon modes as revealed by first-principles calculations . When exposed to shockwaves, few-layer MoS2 and other TMDCs undergo morphological changes with reduction in interlayer (002) separation leading to softening of the A1g and E12g bands . Raman spectra of 1T-MoS2 and 1T-MoSe2 prepared by the Li-intercalation and exfoliation method are different from those of respective 2H phases due to change in the symmetry . Figure 7d compares the Raman spectra of 1T-MoS2 with that of the 2H-MoS2. Besides E12g and A1g modes which are blue-shifted with respect to 2H-MoS2, new bands of J1, J2 and J3 appear in the Raman spectrum of 1T-MoS2.
WTe2 occurs in the orthorhombic distorted 1T structure (or Td phase) with a √3×1 superstructure of dimerized zigzag W–W chains running along crystallographic a-axis (figure 15b). First-principles calculations reveal that electronic band structures for bulk and monolayer of Td-WTe2 are rather similar, both being semimetallic in nature . This is in contrast with other TMDCs which exhibit a strong dependence of the electronic structure on the number of layers. There are 33 possible zone-centre Raman active modes for Td-WTe2, of which five distinct Raman bands are observed around 112, 118, 134, 165 and 212cm−1 in bulk Td-WTe2 (figure 7e). The intense bands at 165cm−1 and 212cm−1 are assigned to the A′1 and the A′′1 modes, respectively, based on symmetry analysis and the Raman tensor . In contrast to the 2H-polytypes of group VI TMDCs, the low crystal symmetry of Td-WTe2 leads to mixing of in-plane (along the a–b plane of the unit cell) and out-of-plane (along the c-axis of the unit cell) components of atomic displacements. The A′1 mode involves out-of-plane (z-direction) displacements of Te atoms and in-plane displacements of W atoms while their motion is vice versa in the A′′1 mode. From the layer-dependent Raman spectroscopy, it was found that the intense A′′1 band is most sensitive to the number of layers, exhibiting an up-shift of about 4cm−1 in a 3-layer flake relative to bulk Td-WTe2, whereas the intense A′1 band does not change with the number of layers though both belong to the same A1 symmetry (figure 7f). In the A′1 mode of vibration, the Te atoms of the same plane vibrate in-phase while their motion is out-of-phase in the mode of vibration. Thus, the A′1 mode is plausibly more localized to a layer of WTe2 and shows weaker or no dependence on the number of layers .
ReS2 occurs in a distorted 1T structure with dimerized zigzag Re–Re chains. Bulk ReS2 behaves as electronically and vibrationally decoupled monolayers stacked together . Bulk and monolayer ReS2 have nearly identical band structures both being direct-band gap semiconductors (e.g. approx. 1.5eV). Consequently, in contrast with other TMDCs, the Raman spectrum shows no dependence on the number of layers (figure 7c) and the optical absorption is insensitive to interlayer distance modulated by hydrostatic pressure (dEgap/dP~0.02eV per GPa) . First-principles calculations ascribe the decoupling to Peierls distortion of the 1T structure of ReS2, which prevents ordered stacking and minimizes the interlayer overlap of wave functions. Weak intralayer polarization due to small charge differences between neighbouring Re and S planes is another factor that leads to very weak van der Waals interlayer interaction .
The discovery of defect-induced room temperature ferromagnetism in otherwise non-magnetic graphene has aroused significant attention and led to investigations on layer-dependent magnetism in TMDCs. MoS2 bears striking similarities with graphene such as ferromagnetic ordering along zigzag edges and novel exchange bias phenomenon which makes the former attractive for spintronics applications. Panich et al.  reported electron paramagnetic resonance (EPR) investigations on inorganic fullerene-like MoS2 nanoparticles which indicated the presence of a large density of dangling bonds carrying unpaired electrons. The unsaturated metal centres with partially filled d-orbitals render MoS2 and WS2 clusters (Mo6S12 and W6S12) magnetic . Zhang et al.  report that MoS2 exhibits room-temperature ferromagnetism owing to a high density of prismatic edges containing unsaturated Mo and S atoms. Lithiated MoS2 prepared by soaking the bulk MoS2 in n-butyl lithium also exhibits room-temperature ferromagnetism . Upon exposure to a 2MeV proton beam, MoS2 exhibits room-temperature magnetic ordering and the temperature dependence of magnetization displays ferrimagnetic behaviour with a Curie temperature of 895K. Proton irradiation can induce formation of isolated vacancies, vacancy clusters, edge states and reconstructions of the lattice which may give rise to room-temperature magnetism . Graphene-like MoS2 is ferromagnetic at room temperature with the field-cooled (FC) and the zero-field-cooled (ZFC) magnetization diverging from about 300K [28,89]. Such divergence of the ZFC and FC curves is found in magnetically frustrated systems like spin-glasses where the ferromagnetic and antiferromagnetic domains are randomly distributed. Figure 8a shows temperature-dependent magnetization of few-layer MoS2. Few-layer MoS2 is ferromagnetic at room temperature, as seen from the hysteresis in magnetization versus applied field in the inset of figure 8a. The coexistence of both ferromagnetic and antiferromagnetic interactions in few-layer MoS2 was evident from the exchange bias behaviour in few-layer MoS2 samples at different applied fields (figure 8b). A negative exchange bias of 15.5Oe observed at 1T field decreases with the decreasing applied field and reverses to −15.5Oe at −1T . The existence of ferromagnetism in MoS2 is ascribed partly to the presence of zigzag edges in the magnetic ground state . First-principles calculations show that the MoS2 nanoribbon with armchair-terminated edges is semiconducting with a non-magnetic (spin-unpolarized) ground state, while that with zigzag-terminated edges is metallic with a magnetic (spin-polarized state) ground state . Magnetic force microscopy of few-layer MoS2 nanosheets have further revealed that the nanosheets become non-magnetic beyond a certain thickness .
The relative ease of fabricating complex device structures renders 2D materials attractive for nanoelectronics. Graphene has been widely studied because of its rich physics and very high mobilities. Graphene-based FETs with a charge carrier mobility as high as 106cm2V−1 s−1 have been fabricated previously . However, pristine graphene is a zero-band gap material and a finite band gap is required for several applications including transistors. On the other hand, monolayer MoS2 with a large intrinsic direct-band gap (approx. 1.8eV) is a potential candidate and can complement graphene in applications requiring ultrathin transparent semiconductors, such as optoelectronic and energy harvesting applications [8,94]. An n-type FET based on single-layer MoS2 with HfO2 as the top-gate dielectric has been demonstrated to exhibit high current on/off ratios exceeding 108 and FET mobilities as high as about 200cm2V−1s−1 (figure 9), being comparable with thin silicon films and graphene nanoribbons . Late et al.  have observed hysteresis in back-gated FETs fabricated out of mechanically exfoliated single-layer MoS2, which increases with increasing relative humidity and under illumination. The hysteresis is ascribed to the trapping states induced by surface-absorbed water molecules and the photosensitivity of MoS2 devices. After passivating the device with a 30 nm thick Si3N4 layer, the conductivity of the transistor increases by greater than 100 times, and the hysteresis is almost suppressed. Further, the mobility also increases, plausibly due to the improved contact and suppressed Coulomb scattering . Hao et al.  studied the effect of chemisorbed oxygen and water on conductance values of back-gated MoS2 transistors. Transistors with high-k Al2O3 as the top-gate dielectric exhibit high electron mobility of about 517cm2V−1s−1 and the current on/off ratio, greater than 108 . Min et al.  report that the single-layered MoS2 FET, top-gated with high-k Al2O3, exhibited greater mobility of approximately 170cm2V−1s−1 compared with the double- and triple-layered MoS2 FETs which showed only approximately 25 and approximately 15cm2V−1s−1, respectively. The degradation of performance with MoS2 thickness was ascribed to the increasing dielectric constant with thickness . Flexible, thin-film MoS2 transistors with a current on/off ratio of 105 were fabricated using ion gel gate dielectrics .
Chakraborty et al.  prepared top-gated single-layer MoS2 transistors with a current on/off ratio of about 105 and a mobility of 50cm2V−1s−1 and also carried out Raman studies on these transistors. Although MoS2 is p-type, transistors made of 15nm thick sheets are ambipolar and the measured Hall mobility of holes is greater (86cm2V−1s−1) than that of electrons (44cm2V−1s−1) suggesting that p-type operation is more favoured . Lee et al.  have demonstrated that a FET based on MoS2 prepared by liquid-phase exfoliation shows n-type characteristics with low current on/off ratios of 3–4 and a low mobility of 0.117cm2V−1s−1. Temperature-dependent electrical transport measurements revealed variable range hopping transport at lower temperatures which can be explained by the Coulomb potential of randomly distributed charges at the MoS2–SiO2 interface . Integrated small-signal analogue amplifiers were fabricated by connecting in series two top-gated single-layer MoS2 transistors . A FET fabricated by sandwiching few-layer MoS2 as the semiconducting channel between single-layer graphene and a metal thin film exhibited current on/off ratio greater than 103, and a high current density up to 5000Acm−2 .
Transistors made of mechanically exfoliated MoS2 nanosheets with one- to four-layer thicknesses have been used to detect up to 2ppm levels of NO with 80% decrease in the channel conductance . He et al.  have fabricated flexible transistors with MoS2 as the thin-film channel and reduced graphene oxide (RGO) as the electrode material for sensing NO2. When Pt-decorated MoS2 was used as the channel, the sensitivity increased by about three times, with a detection limit of 2ppb . Devices made of electrochemically reduced MoS2 thin films showed good conductivity and rapid electron transfer and were used for the selective detection of glucose and dopamine . Transistors made of few-layer MoS2 are shown to sense humidity and gases such as NH3 and NO2 (figure 10) . The sensing is based on the change in device-resistivity in the presence of these gases, which occurs due to charge transfer to MoS2 at different applied fields.
Phototransistors have been fabricated of single- and bilayer MoS2 to detect green light and of triple-layer MoS2 to detect red light . MoS2 with thickness ranging from 10 to 60nm was used to detect light in UV-to-near-IR radiation . FETs fabricated using few-layer WS2 prepared by sonication in IPA exhibit ambipolar behaviour, with a high current on/off ratio of approximately 105 and good photosensitivity to visible light . Back-gated FET of ultrathin MoSe2 obtained by mechanical exfoliation showed n-type characteristics with a mobility of 50cm2V−1s−1 and current on/off ratio exceeding 106 . The temperature dependence of FET mobilities in these devices suggests a strong, phonon-dominated scattering. FETs of single-layer WSe2 with chemically doped source/drain contacts and high-κ gate dielectric exhibited ambipolar characteristics, with large hole mobilities in the order of approximately 250cm2V−1s−1 and high current on/off ratios greater than 106 at room temperature [112,113].
MoS2 is a well-known catalyst for the hydrodesulfurization reaction of sulfur-containing hydrocarbons [114–116]. Few-layer MoS2 decorated with Co and Ni nanoparticles has been shown to be a highly efficient catalyst for the hydrodesulfurization of thiophene to butane with a conversion-efficiency of approximately 98% at approximately 375°C . Ni-Fe/MoS2 composites have exhibited good catalytic activity for hydrazine oxidation . Hinnemann et al.  have shown that MoS2 nanoparticles are good catalysts for the HER with an over potential as low as 0.1–0.2V. Theoretical studies have revealed that the edge sites  and vacancies  in MoS2 are catalytically active sites for HER. Enhanced photocatalytic and electrochemical HER activities have been achieved by increasing the surface area of or the density of active edge sites in the MoS2 catalyst [122–126]. With the MoS2/graphene composite, electrocatalytic HER activity was further enhanced  due to a better electrical coupling to the electrode facilitated by graphene . Photocatalytic HER has been carried out with MoS2 as a co-catalyst in the presence of light-absorbing semiconductors such as CdS [129,130], CdSe  and TiO2 , and dyes such as [Ru(bpy)3] and eosin . Incorporation of a small weight per cent of graphene in the composites enhances the activity as graphene can act as an electron sink and promote better charge separation . The charge separation is further improved by forming a p–n junction with p-type MoS2 with n-type N-doped graphene. N-doped graphene composites with MoS2 showed enhanced photocatalytic  and electrocatalytic HER activity  compared with that of MoS2/graphene composites.
The 1T-MoS2 (or 1T-MoSe2) is metallic with t2g (dyx, dzx, dzy) orbitals of Mo being partly filled with two electrons. An extra electron leads to the stable half-filled configuration of t2g orbitals (figure 11a). Therefore, 1T-MoS2 (or 1T-MoSe2) can easily accept an electron from a dye like eosin and catalyse the HER (figure 11b). 1T-MoS2 prepared by Li intercalation and exfoliation indeed showed 600 times higher activity than the few-layer 2H-MoS2 (inset in figure 11c) . 1T-MoSe2 shows even higher HER activity (approx. 62±5mmolg−1h−1) than that of 1T-MoS2 (figure 11c) . The lower work function of MoSe2, and the weaker binding of hydrogen to Se than to S at the edge sites of MoS2 are considered to be responsible for the higher activity of 1T-MoSe2. 1T-WS2 nanosheets prepared by Li-intercalation and exfoliation show high activity for electrocatalytic HER [32,136] which is ascribed to domains of the strained 1T phase. Density functional theory calculations reveal that tensile strain increases the density of states near the Fermi level and facilitates hydrogen binding . 1T-MoS2 also shows better electrocatalytic HER activity than 2H-MoS2; oxidation of MoS2 edges revealed that, in addition to the edges, the basal plane of 1T-MoS2 nanosheets is also active for HER .
Layered TMDCs (MX2; M=Mo, W, Ti, V and X=S, Se) are good host materials for batteries, allowing for rapid intercalation and de-intercalation of guest ions such as lithium or sodium ions between the host layers. The first lithium-ion battery using MoS2 as the anode material was demonstrated by Haering et al. . Chemically exfoliated and restacked MoS2 with an enlarged c-lattice parameter and good surface area exhibits a high charge capacity of 800mAh−1g−1 in the first cycle which is retained till 750mAh−1g−1 after 20 cycles at a current density of 50mAg−1 . Electrodes prepared from hydrothermally synthesized MoS2 nanoflakes and nanoplates exhibited capacities of as high as 1000mAh−1g−1  and 1062mAh−1g−1 , respectively. WS2 nanosheets and nanoflakes also show good reversible capacity [142–144]. Julien et al.  have shown that 0.6mol of Li ions intercalates per mole of WS2. Mesoporous WS2, prepared using SBA-15 as the template, has a high surface area and exhibits a storage capacity of 805mAh−1g−1 at a current density of 100mAg−1 . Owing to its high surface area and chemical stability, graphene has been used as a commercial anode material [147,148]. Composites of TMDCs and graphene have been investigated extensively for Li-ion storage. MoS2/graphene composites exhibited a high reversible capacity of approximately 1290mAh−1g−1 maintained up to 50 cycles at a current density of 100mAg−1. A high specific capacity of 1040mAh−1g−1 was retained even at a higher current density of 1000mAg−1 . MoS2/graphene composites prepared by different routes show high reversible capacity with good capacity retention [150–153]. As shown in figure 12a–c, replacing graphene with N-doped graphene in MoS2/graphene composites further improves the capacity and retention . Increasing vacancies and defects in N-doped graphene appears to promote more Li-ion insertion leading to the observed increase in capacity with an increasing number of cycles (figure 12a) . WS2/graphene composites also showed improved properties such as high specific capacity at high current densities and good capacity-retention compared with the bare WS2 . Amorphous carbon-coated FeS nanosheets synthesized via a solution-phase route exhibited a discharge capacity of approximately 623mAh−1g−1 at a current density of 0.1Ag−1 with good cyclability even at higher discharge rates . David et al.  have reported MoS2/graphene composites as anode materials in Na-ion batteries which showed a stable capacity of 230mAh−1g−1, with a capacity-retention up to several cycles, even at high current densities of 200mAg−1. Graphene or carbon in the composites enhances their electrochemical performance and cyclability by absorbing the mechanical stress during volume changes accompanying the insertion and de-insertion cycles, by increasing the electrical conductivity and by preventing the diffusion of polysulfides into the electrolyte.
Electrodes fabricated with MoS2 samples containing a high density of nanowalls exhibit an electrochemical double layer capacitance comparable with that of electrodes made of carbon nanotube arrays . Edge-oriented MoS2 nanoporous films synthesized by electrochemical anodization of molybdenum metal in the presence of sulfur vapour showed a capacitance up to 12.5mFcm−2 from cyclic voltammetry (CV) measurements at a scan rate of 50mVs−1 and 14.5mFcm−2 from galvanostatic charge-discharge measurements at a current density of 1mAcm−2. These films showed an increase in the capacitance from 2.2 to 10.5mFcm−2 after 10000 testing cycles at a current density of 10mAcm−2 . Supercapacitor performance of composites of MoS2 and RGO of different compositions has been studied . MoS2/RGO (1:2) composite exhibited best results as shown in figure 13. The quasi-rectangular CV curves for MoS2/RGO electrodes resemble those of an ideal supercapacitor, showing a maximum capacitance of 416Fg−1 with MoS2/RGO (1:2) at a scan rate of 5mVs−1 (figure 13a). The current increases and the capacitance decreases with scan rate as shown in figure 13b. Figure 13c shows the charge-discharge curves obtained at a current density of 1Ag−1. The discharge time of MoS2/RGO (1:2) is significantly longer than the other composites, and the nearly symmetrical curve indicates the remarkable charge-storing ability of this composite. The specific capacitance of this composite is calculated to be 249Fg−1 at a current density of 0.3Ag−1. Figure 13d shows the specific capacitance of electrodes made of MoS2, RGO and MoS2/RGO composites obtained at different current densities. Combination of high electronic conductivity of RGO and surface properties of both MoS2 and RGO gives rise to enhanced specific capacitance of MoS2/RGO composites. The synergistic effect is calculated to be 118% in the case of MoS2/RGO (1:2) composite. Composites of MoS2 and graphene synthesized by cysteine-assisted solution-phase synthesis exhibit a maximum specific capacitance of 243Fg−1, which is significantly higher that of the individual components. The electrodes exhibit long-term cyclic stability, with only a 7.7% decrease in specific capacitance after 1000 cycles . Composites of MoS2 and polyaniline (PANI) with different compositions exhibit excellent supercapacitor performance . The MoS2/PANI composites with compositions of 1:1 and 1:6 showed capacitance values of 417 and 567Fg−1, respectively, with improved cyclic stability compared with PANI alone. Nanocomposites of MoS2 and polypyrrole prepared by in situ oxidative polymerization show a specific capacitance of about 700Fg−1 at a scan rate of 10mVs−1 .
While electronic and spintronics devices generally exploit the charge and spin of electrons, respectively, valleytronics relies on the fact that the conduction (valence) bands of some materials have two or more degenerate minima (maxima), separated in the momentum space . To realize a valleytronic device, it is necessary to produce a valley-polarization by controlling the number of carriers in these valleys. Monolayer MoS2 is ideal for valleytronics with conduction and valence band edges having two degenerate valleys at the corners (K and K′ points) of the 2D hexagonal Brillouin zone (figure 14a). The direct-band gaps (approx. 1.9eV) at these two valley points allow for valley-polarization by optical pumping [163–165]. Owing to the broken inversion symmetry, spin–orbit interactions split the valence bands by approximately 160meV and the spin projection, Sz, is well defined along the c-axis of the crystal with the two bands being of spin down (E↓) and spin up (E↑) in character [62,164,166]. This broken spin degeneracy together with time-reversal symmetry leads to inherent coupling of the valley and the spin of the valence bands in monolayer MoS2 leading to the valley-dependent optical selection rule [163,165]. Consequently, inter-band transitions in the vicinity of the K or K′ valleys couple exclusively to left or right circularly polarized light, respectively. The circular component of the band edge luminescence is of the same polarization as that of the circularly polarized excitation (figure 14b) . On the other hand, inversion symmetry is preserved in bilayer MoS2, which in combination with time-reversal symmetry, restores the spin-degeneracy by splitting the fourfold degenerate valence bands into two spin-degenerate valence bands. As valley and spin are decoupled in bilayer MoS2, the valley-dependent selection rule is not allowed and, therefore, negligible circular polarization is exhibited by bilayer MoS2 under the same conditions (figure 14c) .
Tightly bound negative trions, a quasi-particle composed of two electrons and a hole, have been spectroscopically identified in monolayer MoS2 FET . These quasi-particles can be created optically with spin and valley polarized holes. These trions possess a large binding energy (approx. 20meV), nearly an order of magnitude larger than that found in conventional quasi-2D systems like quantum wells. This is due to the significantly enhanced Coulomb interactions in monolayer MoS2, arising from reduced dielectric screening. The existence of tightly bound trions with dynamically controllable hole valley and spin in monolayer MoS2 gives rise to possibilities of novel many-body phenomena and may impact the development of new photonic and optoelectronic devices, such as photovoltaic cells or optical detectors.
In contrast to the regular layered counterparts, distorted layered ditellurides of Mo and W with zigzag metal–metal chains (figure 15a and b) are gaining considerable attention due to their unique properties such as giant MR and pressure-driven superconductivity. Recently, a giant, unidirectional (along W–W chains) positive MR of 452700% at 4.5K in a magnetic field of 14.7T and a record 13 million per cent at 0.53K in a field of 60T (figure 15d), has been reported in single crystals of semimetallic orthorhombic Td-WTe2 (figure 15b) . Even at very high applied magnetic fields, MR in Td-WTe2 does not saturate, which is attributed to a perfect balanced electron-hole resonance. Bulk monoclinic 1T′-MoTe2 with Mo–Mo zigzag chains (figure 15a) is semimetallic and also exhibits giant MR of 16000% in a magnetic field of 14T at 1.8K (figure 15c) .
Pan et al.  have reported pressure-driven dome-shaped superconductivity in Td-WTe2. Superconductivity appears at a pressure of 2.5GPa. The critical temperature (Tc) reaches a maximum of 7K at around 16.8GPa and monotonically decreases with increasing pressure . Kang et al.  also reported superconductivity in Td-WTe2 arising from a suppressed MR state under pressure. With increasing pressure, MR is gradually suppressed and turned off at a critical pressure of 10.5GPa, where superconductivity begins to emerge . Pressure-driven superconductivity has also been reported in single crystals of orthorhombic Td-MoTe2 (below 120K), a low-temperature polymorph of 1T′-MoTe2. A maximum Tc of 8.2K is observed at 11.7GPa .
Besides group VI-metal dichalcogenides, there are other layered metal chalcogenides of considerable interest. Bi2Se3, Bi2Te3 and Sb2Te3 are exotic spin Hall quantum phases known as topological insulators (TIs) wherein gapless, metallic surface states exist within insulating bulk band gap. These Bi- and Sb-based TIs are also the state-of-the-art thermoelectric materials which can reversibly convert waste heat into electricity . Bi2Se3 (Bi2Te3 or Sb2Te3) is an anisotropic layered material containing quintuple layers (QLs) each of approximately 1nm thickness and composed of five covalently bonded atomic planes [Se2-Bi-Se1-Bi-Se2]. These QLs are periodically stacked along the crystallographic c-axis by weak van der Waals interactions. Free-standing five-atom-thick Bi2Se3 single layers can be synthesized via Li-exfoliation of the bulk material . Ultrathin nanodiscs of Bi2Se3 are obtained by heating Bi(NO3)3 and Na2SeO3 in ethylene glycol with a small amount of NH2OH . Few-layer Bi2Se3 nanosheets have been prepared by green, ionothermal synthesis, using a mixture of bismuth acetate and selenourea in an ionic liquid solvent, 1-ethyl-3-methylimidazolium tetrafluoroborate ([EMIM][BF4]). Ionothermal decomposition of single-source precursor, Bi(Se-C6H6N)3 in the same solvent yields few-layer nanodiscs of Bi2Se3 . Mechanical exfoliation  chemical vapour transport  and vapour-phase deposition  have been employed for synthesis of few-layer Bi2Te3 nanosheets. Sb2Te3 nanosheets have been synthesized using SbCl3, Na2TeO3 and hydrazine hydrate in ethylene glycol under microwave heating . Single-crystalline nanobelts and nanosheets of Sb2Te3 have been synthesized by the reaction of SbCl3 and TeO2 with hydrazine hydrate as the reducing agent . A pressure-induced insulator-metal transition has been found in rhombohedral BiI3 (near 1.5GPa), whose high symmetry parent structure is similar to that of Bi2Se3 TI. The calculated Born effective charges suggest the presence of metallic states in the structural vicinity of rhombohedral BiI3 .
The previous sections must clearly bring out the fascinating properties of layered inorganic materials especially when they are made up of single- or few layers. The properties that we described cover a wide range including optical properties, FETs, GMR, catalysis and energy devices. It is interesting that the discovery of single-layer MoS2 has given rise to new areas in physics, namely valleytronics and trions. We believe that these areas will be explored extensively in the coming years. Many new properties are yet to be explored, as evidenced from new discoveries coming out ever so frequently in the last year or two. A typical example is the unusually high magnetoresistance by WTe2. We believe that those who are interested in physics and chemistry of materials can expect to have an exciting period of research in 2D inorganic materials in the years to come.
The article was written jointly by both the authors.
We declare we have no competing interests.
M.K.J. acknowledges CSIR, India for research fellowship.