English

• 1. Introduction

  Transition metal dichalcogenides with the chemical formula MX₂ (M: transition metal element; X: chalcogen) are binary layered compounds with hexagonal lattice structure, where the atoms are bound in the form of M-X bonds in the basal plane and the MX₂ layer is combined with the adjacent layer by van der Waals force [1–4]. Considering its excellent optical [5–7], electrical [8–12], magnetic [13–16], mechanical [17,18] and thermal properties [19–24], it is emerged as a splendid material utilized in energy storage [25–27], catalysis [28–30], semiconductor [31–35], lubrication [36,37] and other fields. For example, if we fabricate flexible electrodes for hydrogen evolution reaction (HER) with transition metal dichalcogenides (TMDs), the mechanical stability of the catalytic system can be enhanced due to the excellent mechanical properties can make the catalyst resist the stress caused by the stretching and rupture of bubbles in the high-current catalytic reaction [38]. The TMDs exist widely in nature, and most of them are air-stable [39–41]. However, considering the defects of primitive TMDs, including inert basal plane and limited electronic transmission efficiency, it is urgent to find effective regulating strategies of MX₂ to develop more potential applications. Fortunately, due to the intrinsic relationship between specific physicochemical properties and atomic structures, the properties and application fields of TMDs can be expanded by chemical modification.

  Oxide modification is an effective method to transform the physicochemical properties of TMDs. In this way, a large material system can be formed, such as variants without changes of intrinsic elements, variants with single oxygen atom doping, heterojunction, and dioxides with interlayered elementary substance. It had been reported that various synthetic strategies, including spontaneous oxidation [42–44], high temperature thermal annealing [45–49], H₂O₂ etching [50–52], electrochemical oxidation [53–56], oxygen plasma treatment [57–60], ultraviolet-ozone (UV-O₃) treatment [61,62], and laser treatment [63–65], were used to treat MX₂ to acquire special oxidized variants with precise atomic structures. For example, Huang et al. [66] prepared the atomic structure of oxidized ReS₂ by UV-O₃ treatment. In their research, except the in-situ oxidation of S atoms, there exist exotic covalently-bonded surface oxygen atoms which distort the atomic structure of single-layer (1 L) ReS₂ slightly. In this way, two of the four crystallographically distinct S atoms in the same unit cell having the bonding lengths which are much is oxygen atop sulfur and the other is substitutional oxygen on the sulfur vacancy (SV). Besides, Tan et al. [67] reported that they fabricated MoS_2−xO_x nanosheets and get the diverse oxygen incorporation extent of MoS_2−xO_x. Due to the enhancement of intrinsic conductivity and the acceleration of the interfacial charge transfer after oxygen doping, the MoS_2−xO_x demonstrated a remarkable degradation activity for antibiotics. The MoS_2−xO_x would dramatically boost the degradation activity of persistent contaminants in the co-catalyst Fenton system, which provides a new method for higher environmental decontamination performance. These studies above have proved that specific oxidation products can be obtained on demand through certain oxidation strategies. However, the recent researches are not sufficient and needs to be summarized systematically.

  In this review, we firstly summarize the atomic structures of TMD oxidized variants. What is more, the mechanisms for corresponding fabrication approaches are provided. Since MoS₂ is a typical kind of TMDs and its oxidation structure has been widely studied in recent years, and TMDs often have similar physical and chemical properties. Based on the researches of MoS₂, the properties of other kinds of oxidation variants of TMDs will be described afterwards. In the end, we provide the outlooks of future applications of the TMD oxidized variants. This review is expected to establish the material system of the TMD oxidized variants, aiming to provide insight into the design potential application of TMD oxidized variants.

  2. Oxidized variants without changes of intrinsic elements

  As we all know, the oxidation process could change the atomic structure of TMDs by forming or breaking chemical bonds. In some cases, the oxidation process only changes the atomic structure, such as forming thinner stacked layers and triangular pits, without introducing new types of elements. The generation of these new structures would bring some novel physicochemical properties and have potential application values.

  2.1 Stacked few-layer (bi/tri-layer) structure

  Transition metal dichalcogenides has the structures featured by stacking layers. Recently, various of preparation researches are reported in order to obtain single or few layers of TMDs [68–71]. Among them, electrochemical oxidation exfoliation is an excellent way which converts multiple layers of TMDs into single layers by partially oxidizing TMDs in an electrochemical environment and the oxidized parts dissolve into the solution.

  Pan et al. [72] reported an electrochemical exfoliation method for stacked few-layer (bi/tri-layer) MoS₂ driven by a lateral inward electrochemical reaction (Fig. 1A). In this electrochemical reaction, water molecules go into interlayered space through the formed two-dimension (2D) channel. Then, the bulk MoS₂ could be electrochemically selectively oxidized and finally the remaining bi/tri-layer MoS₂ forms a stacked structure. By this means, the interlayered MoS₂ reacts as the function: MoS₂ + 8H₂O → Mo⁶⁺ + 2SO₄²⁻ + 16H⁺ + 18e⁻ with the high anodic potential applied to bulk MoS₂ in the solution of sulfate electrolyte. The atomic structure is shown in Fig. 1B, the edge of MoS₂ is etched selectively with the 2D channel created subsequently, leaving a stacked structure in the form of uniform bi/tri-layer MoS₂ layers, which made the remaining MoS₂ layers restacked together with a decreased thickness. This phenomenon results from the oxidation of MoS₂ to Mo⁶⁺ and the dissolution of oxidized MoS₂ in the sulfate electrolyte. The scanning transmission electron microscope (STEM) image (Fig. 1C) proves the formation of this structure: The most part of multilayer MoS₂ get to a decreased thickness. Meanwhile, the bottom part remains the original thickness due to the strong Coulomb interaction between substrate and bottom MoS₂ which inhibits electrochemical reaction. The special atomic structure distinguishes the electrochemical exfoliation from traditional corrosion and etching methods.

  Figure 1

  

  Figure 1.  (A) The schematic of the process driven by a lateral inward electrochemical reaction and water molecules go into bulk phase through the formed 2D channel. (B) The schematic of the thickness of the edge and the basal plane of MoS₂ after exfoliation. (C) The magnified STEM image of the transition zone from bulk to edge. (D) The polarization curves of MoS₂ after different CV cycles treatment (0, 200, and 300 cycles). The dash lines mark the overpotential at 10 and 20 mA/cm². (E) The corresponding Tafel slope plots. Reproduced with permission [72]. Copyright 2021, John Wiley & Sons, Ltd.

  DownLoad: Full-Size Img PowerPoint

  Interestingly, in the process of electrochemical exfoliation, adding some steps can make the product have more properties. In the research of Yu et al. [73], during electrochemical exfoliation of ReS₂, they use large sized molecular intercalants to induce parallel mirror twin boundaries and atomic row dislocation. Through the calculation via density function theory (DFT), they conclude that ferromagnetism is detected due to the combined effects of lattice strain and SVs in the mirror twin boundaries. Meanwhile, the emergence of SVs also enhances the chemical activity of ReS₂ as HER catalysts.

  It is worth noting that electrochemical exfoliation is usually a way to reduce the layer of transition metal chalcogenides. To achieve the above experimental phenomenon, it is necessary to accurately control the partial oxidation between TMD layers. Leong et al. [74] reported that through electrochemical exfoliation with bipolar electrodes, they successfully made WS₂ undergo a phase transition from 2H to 1T, and reduced the multilayer WS₂ to fewer layers. The electrocatalytic performances of WS₂ is also enhanced. Compared with the phenomenon mentioned above, they realize the complete oxidation of interlayered TMD without forming the stacked structure.

  According to the more moderate hydrogen adsorption energy on the sites at edge, this stacked few-layer structure is considered to have an improved HER performance. The exfoliated stacked MoS₂ is applied as catalyst for HER. The polarization curves (Fig. 1D) of the single MoS₂ nanosheet after different CV cycles treatment are recorded in H₂SO₄ based on-chip MoS₂ device. As the bulk MoS₂ electrochemically exfoliated to the stacked bi/tri-layered structure, the overpotential decreases from 347 mV to 127 mV at 10 mA/cm². Meanwhile, the Tafel slope (Fig. 1E) decreases from 326 mV/dec to 199 mV/dec. Taking the charge injection from electrode into account, the electrons in bulk MoS₂ transport in the vertical direction through hopping. A ≈ 0.12 eV gap for electron hopping is formed by the large van der Waals gap. The potential gap for interlayered electron transfer is reduced by the decreased thickness of MoS₂, which enhances HER performance. Furthermore, the expanded edge sites which have few atomic layers inspire more available active sites. Simultaneously, the reduced thickness enhances the hydrogen adsorption energy. Thus, the novel structure of bi/tri-layer MoS₂ fabricated by electrochemical exfoliation without introducing extra elements would have potential applications in the field of HER.

  2.2 TMDs with triangular pits

  Some types of TMD oxide would volatilize easily at high temperature [75,76]. In view of this, a special atomic structure of TMDs with triangular pits could be formed through heating TMDs in the absence of air intervention. The formation of such pits is due to the anisotropic oxidation of the TMD edges.

  Zhou et al. [77] reported the atomic structure of oxidative etching of MoS₂ layers by using a dry oxygen etching reaction where the mechanically exfoliated MoS₂ samples are heated in the center of a tubular furnace with precise modulation at the temperature between 250 ℃ and 385 ℃ without the introduction of carrier or reactive gas. As the figure of MoS₂ atomic structure shown in Fig. 2A, the triangular pits are initiated on MoS₂ layer at natural surface defects of MoS₂. The sizes of triangular pits are attributed to the thickness of MoS₂ layers where the 1 L MoS₂ has the smallest pit sizes. As the heating temperature grows, the pits forming due to the anisotropic oxidation of edges of MoS₂ and the Mo and S atoms are transformed to MoO₃ and SO₂ gas molecules respectively, leaving the oxidized MoS₂ with special ratio of basal plane sites to edge sites. Moreover, the typical atomic force microscope (AFM) images of MoS₂ heated in various temperature (Figs. 2B–E) clarified that the etching process increases with the surrounding temperature grows and the angles of triangular pits are close to 60°. When the surrounding temperature is at 325 ℃, almost no etching is detected by AFM. However, at 345, 355 and 365 ℃, it is easy to find the triangular pits in AFM images and the sides become longer as the temperature grows. Notably, as the temperature increases, the regular triangular pits distort which may attribute to the harsh oxidative etching rate. Lv et al. [78] also fabricated the MoS₂ flakes with triangular pits. They heated up the sample in the chemical vapor deposition (CVD) furnace to 300 ℃ in 10 min, and the sample was etched for 5 min. They also reported no production of MoO_x during the process. Furthermore, they proved that similar oxidation process could be performed on other TMDs, such as MoSe₂ and WS₂.

  Figure 2

  

  Figure 2.  (A) An atomic mechanism showing the formation of triangular pits in MoS₂ layers. A pit in MoS₂ could generally expose various types of edges, including ZZ-Mo and ZZ-S₂. Typical AFM images showing the formation of well oriented equilateral triangular pits on 1 L MoS₂ at various temperatures: (B) 325 ℃, (C) 345 ℃, (D) 355 ℃, and (E) 365 ℃. Raman spectra of n-layer MoS₂ sheets after oxidative etching at (F) 355 ℃ and (G) 365 ℃. Reproduced with permission [77]. Copyright 2013, Tsinghua University Press.

  DownLoad: Full-Size Img PowerPoint

  The Raman spectra of MoS₂ sample heated at 355 (Fig. 2F) and 365 ℃ (Fig. 2G) illustrate that after heating the MoS₂ sample shows the very similar signal with the pristine MoS₂. Although the weak Raman peak shows at ~820 cm⁻¹ by red arrows, considering its appearance in the pristine sample, it could not indicate the existence of MoO₃. In short, the TMDs with triangular pits have the enhanced surface properties attributing to the changes of the ratio of edge sites to basal plane sites, which could be utilized as hydrodesulfurization (HDS) or HER catalysts.

  3. Oxidized variants with single oxygen atom doping

  The oxygen doping is a special kind of oxidation, which mainly embodies in two specific forms of expression: One is the oxygen incorporation of intrinsic SVs generated by artificial synthesis, for example, CVD [79–81]. The other is the oxygen atom bound up atop sulfur atoms, forming new S-O chemical bonds.

  3.1 TMDs with oxygen incorporated into SVs

  The oxygen incorporation of intrinsic SVs could be realized by many methods, including spontaneous oxidation [82], oxygen plasma treatment [83], electrochemical oxidation [84], etc. Although the approaches vary from one to another, the fundamental mechanism remains the same where the oxidation process should be limited in avoid of the complete oxidation of TMDs into trioxide.

  3.1.1   Fabricated by spontaneous oxidation

  Pető et al. [85] fabricate 1 L MoS_2−xO_x through spontaneous oxidation, and the structure diagram is exhibited as Fig. 3A, and atomic-resolution STM images which show the basal plane of a mechanically exfoliated 1 L MoS₂ after one month of exposure in air are shown in Fig. 3B. The fresh 2D MoS₂ crystals prepared had natural spot defect density ranging from 1 × 10¹¹ cm⁻² to 1 × 10¹² cm⁻². After ambient exposure for 1 month, the spot defect density increased to 3 × 10¹²~2 × 10¹³ cm⁻². Considering that such abundant area of MoS₂ basal plane distributes defects, it clearly provides evidence for the establishment of new defect sites in MoS₂ under air conditions. Among the species present under ambient conditions, O is predicted to be the most possible to saturate an S vacancy. STM has been shown to be a slow O substitution reaction. When an S atom is replaced by an O atom, it changes the chemical activity of its neighbors, making them more susceptible to being replaced by O atoms, so individual S atoms are replaced by O atoms one by one, in a sort of chain reaction.

  Figure 3

  

  Figure 3.  (A) Schematic atomic structure of the MoS_2−xO_x solid solution monolayer. (B) Atomic-resolution STM images (5 mV, 2 nA) of an exfoliated MoS₂ single layer after 1 month. (C) The process of chalcogenide atom vacancy formation through oxidation is characterized by a negative oxidation enthalpy (ΔE) for the defect-free 2D MoS₂ basal plane. (D) Kinetic energy barriers calculated by the NEB model for 2D MoS₂ crystals reveal substantially lower barriers for MoS₂ of ~1 eV height that can be overcome even at room temperature on a months-long timescale. (E) Linear sweep voltammogram curves and (F) corresponding Tafel plotsfor Au substrate, MoS₂ single layer, MoS_2−xO_x single layer (1 year old) and Pt substrate. Reproduced with permission [85]. Copyright 2018, London: Nature Publishing Group.

  DownLoad: Full-Size Img PowerPoint

  In summary, we had drawn the observation that the 1 L MoS₂ has the chance to be oxidized in the air. However, the thermodynamic and dynamical processes of MoS₂ oxidation are in need of expounding. As illustrated in Fig. 3C, the surface S atoms reacts with oxygen, yielding SO₂ and forming the SV in the basal plane. The enthalpy was calculated by the formula ΔE = E(reactants) − E(products) = −0.49 eV, which implied that the enthalpy of the final state is lower than for the initial state, drawing the conclusion that thermodynamically favorable for the S atoms to be removed from the MoS₂ basal plane through oxidation, namely, the oxidation of 1 L MoS₂ in the air is thermodynamically spontaneous. Nevertheless, resisted by the principle of kinetics, the substitutional oxidation of MoS₂ faces obstacle. The results of nudged elastic band model calculations (Fig. 3D) clarify that the typical kinetic barrier height is ~1 eV for the reaction pathway proposed. It is worth noting that along with the O substitution of the S vacancies, the energy of the final state was ~4 eV lower compared with the initial state, which clarifies that as a next step, the O substitution process is highly favorable.

  In addition, in line with transition state theory, the dynamic obstacle of ~1 eV for the substitutional oxidation of MoS₂ are hopefully to be conquered in the room condition over a timescale of a month. The polarization curves (Fig. 3E) and corresponding Tafel plots (Fig. 3F) clarify that compared with the reduced pure 2D MoS₂ phase, the 2D MoS_2−xO_x shows a dramatic increase in catalytic HER activity, which attributes to the single O atom incorporating into the MoS₂ basal plane progressively.

  Furthermore, we may doubt if the substitution site is single O atom. As a matter of fact, taking O, O₂, OH saturated S vacancies into account, the only logical imagine, which was agreement with the experimental imagines is the single O atom substitution sites, which are also the thermodynamically most favorable case corresponding to −3.9 eV formation energy.

  It is noteworthy that not all basal plane of TMDs can be spontaneously oxidized in air. The oxidation enthalpy of MoSe₂ basal plane to remove one Se atom by oxidation is ΔE = +0.75 V, which means that it is thermodynamically unfavorable and difficult to spontaneously carry out the process of oxygen substitution of SVs. However, the edge oxidation speed of MoSe₂ is much faster than MoS₂, which means we cannot underrate overall environmental activity of MoSe₂. Namely, the spontaneous oxidation of MoSe₂ may still occur in different parts. Cucinotta et al. [86] reported the spontaneous oxidation process of TiS₂. In their research, the edge of TiS₂ is oxidized by water molecules as the function: TiS₂ + xH₂O → TiS_2−xO_x + xH₂S. The open side of TiS₂ is more easily to be oxidized by water molecules. One of the H atoms is exchanged from water to a neighbor S-up atom, forming a thiol group, leaving a hydroxyl group adsorbed on Ti. In the next process, the H atom in a hydroxyl group moves to the opposite thiol group, forming the H₂S. The final configuration can be characterized by an O atom adsorbed at the edge, replacing one S atom. Actually, the spontaneous oxidation of TMDs could also be influenced by defects and edges. At the initial stage of spontaneous oxidation, the process is relatively slow relating to the intact MoS₂ basal plane. With the process of oxidation, dramatic defects are formed on the basal plane and the edges of TMDs are broken, which greatly speed up the subsequent oxidation process of TMDs. In the research of Gao et al. [87], they study on the spontaneous oxidation at the edges of MoS₂ and WS₂, and the oxidation process of the two is almost identical. Taking MoS₂ as the example (Fig. 4A), fresh 1 L MoS₂ synthesized using CVD had continuous smooth surfaces and were intact under the scale of 10 µm. After stored for 6 months without desiccant (Fig. 4B), it basically kept the original structure with little defects formed at the boundaries. However, after being oxidized at room temperature with desiccant for 1 year (Fig. 4C), the whole crystal was completely degraded and the defects were spread throughout the crystal face. Moreover, at the same time scale of the first six months and the second six months, it is obviously observed that the cracking in the second six months is more serious, which may indicate that when defects occur, the neighboring S atoms will be more easily oxidized, so the cracking speed will be faster.

  Figure 4

  

  Figure 4.  Scanning electron microscopy images of (A) fresh MoS₂, and (B) aged MoS₂ after 6 months without desiccant and (C) after 1 year without desiccant. All the samples were stored in Class 100 cleanroom conditions. Photoluminescence spectra of fresh and aged TMDs. PL intensity map (centered at the A exciton) of fresh MoS₂ (D), and MoS₂ aged for 1 year under ambient conditions (E). The inset optical images show the MoS₂ flakes from which the data were taken. Representative PL spectra of the fresh and 1-year-aged MoS₂ (F). The 1 L MoS₂ films were aged for 1 year under ambient conditions (i.e., room temperature and atmospheric pressure). Reproduced with permission [87]. Copyright 2016, American Chemical Society.

  DownLoad: Full-Size Img PowerPoint

  Photoluminescence (PL) mapping and the optical image of a typical freshly grown 1 L MoS₂ is illustrated in Fig. 4D. Meanwhile, Fig. 4E shows the PL mapping of a typical MoS₂ oxidized for 1 year. The PL spectrum of MoS₂ (Fig. 4F) is dominated by neutral A-exciton emission, along with B-exciton emission which results from the spin-orbital splitting of the valence band, ~0.15 eV away from the A peak. Obviously, compared with the preparation time, the PL intensity on the whole crystal surface of 1 L MoS₂ oxidized for 1 year changes greatly, and PL quenching occurs in some areas, which indicates that the surface of the sample has been greatly damaged by oxidation. The PL intensity of MoS₂ oxidized for one year is also reduced by 2.3 times (Fig. 4F) compared with the preparation time, which means that the composition of MoS₂ has changed dramatically after 1 year in the air, with most of the S atoms being replaced by O atoms. The above study shows that spontaneous oxidation in air can be very significant if the time is long enough.

  3.1.2   Fabricated by oxygen plasma treatment

  Oxygen plasma treatment is explored as the approach to controlling the band structure and thickness of the MoS₂ film, during which the S-O bonds are formed and enhance the character of MoS₂. Considering that if not regulated properly, the oxygen plasma treatment has the chance to oxidize MoS₂ to MoO₃, the low-intensity inductive coupled plasma (ICP) source should be used to minimize the chemical reaction. Comparing with traditional method, the MoS₂ treated by oxygen plasma treatment owns the advantage of smooth surface. Kim et al. [88] reported a precise process to prepare the 1 L MoS₂ with O atom doping (namely MoS_2−xO_x). The prepared few-layer MoS₂ films were transferred onto the Si/SiO₂ substrate. Before plasma treatment, the MoS₂ was cleaned through annealing at 250 ℃ for 3 h at Ar/H₂. Then, the precise regulation of MoS₂ layer was realized by generating an ICP discharge utilizing O₂ gas and varying treating time, a gas flow rate of 30 sccm, at a pressure of 200 mTorr, and a power of 21 W.

  The mechanism of oxygen plasma treatment could be illustrated in Fig. 5A. The primitive MoS₂ was treated with the intervals of 4s. With air exposure, the Mo-S bonds are etched first, which produces abundant S vacancies. Afterwards, the O atoms replace the S vacancies, in which condition the S-O bonds are formed. As the AFM image shown in Fig. 5B, the thickness of 1 L MoS₂ has the 1 nm decrease, which may result from the oxidation or the absorption of H₂O in the air.

  Figure 5

  

  Figure 5.  (A) Illustrations for the atomic states of MoS₂ for showing atomic-level etching depending on the two different plasma treatment processes. (B) An AFM image before and after plasma etching. (C) Mo 3d, (D) S 2p and (E) O 1s XPS spectra of pristine MoS₂ and after plasma treatment. Schematic diagrams showing the structures and energy band diagrams of a MoS₂ FET Energy band diagrams at V_D > 0 and V_G < 0 (off-state) of MoS₂ (F) before and (G) after plasma treatment. Reproduced with permission [88]. Copyright 2016, IOP Publishing.

  DownLoad: Full-Size Img PowerPoint

  As we all know, the O₂ plasma is highly reactive and MoO₃, which is difficult to remove with high melting point of about 800 ℃ and insulative. The chemical reaction associated with the formation of MoO₃ is 2MoS₂ + 7O₂ → 2MoO₃ + 4SO₂. Therefore, it is significant to determine the existence of MoO₃. The X-ray photoelectron spectroscopy (XPS) analysis (Figs. 5C–E) clarified that compared with the primitive MoS₂, the MoS₂ treated after plasma treatment, only the Mo 3d peaks situated at 229.8 and 232.5 eV remain. Without observing the peak at 236 eV, the presence of MoO₃ could be excluded. Additionally, the peaks of 162.8 and 164.0 eV which are corresponded to S 2p_3/2 and 2p_1/2 and 529.8 eV corresponded to O 1s were observed, which provided the evidence of S-O bond formation. Therefore, by oxygen plasma treatment with restriction, the MoS_2−xO_x nanosheet could be fabricated without the formation of MoO₃ nonconductor.

  Similarly, Wang et al. [89] fabricated WSe_2−xO_x by ultraviolet ozone and low power O₂ plasma treatment. They reveal that the oxygen plasma treatment on TMDs is an anisotropic process where the topmost layer is oxidized almost completely while the WSe₂ underneath are affected mildly. This phenomenon matches Fig. 5A well, which proves that most TMDs have similar characteristics.

  In addition, it is crucial to restrict MoO₃ formation in oxygen plasma treatment. Meng et al. [90] reported that after plasma treatment, the PL lifetime of treated MoS₂ decreases from 494 ps to 190 ps due to the transformation to MoO₃, which illustrates the damage the MoO₃ deals to the character of MoS_2−xO_x nanosheet. Furthermore, the plasma power should be limited. As reported by Lin et al. [60], when treated with strong power 100 W, the band gap of MoS₂ would be disordered, which is attributed to the amorphization of crystalline structure.

  To figure out the operating mechanism of the MoS₂ field effect transistors (FETs), the band diagrams before and after oxygen plasma treatment with a positive V_D and a negative V_G are shown in Figs. 5F and G. The primitive MoS₂ before plasma treatment shows off-state at negative V_G, which may result from the large barrier height between drain and source electrodes disturbing the electron flow. By contrast, the MoS₂ after plasma treatment did not show off-state, proving that the highly n-doped state, which is caused by oxygen atom doping, leading to the high field emission current in spite of the negative V_G. The oxygen plasma treatment could also improve the charge transport at contact interfaces. In the research of Wang et al. [89], it is evident by a 100–1000 times improvement in the p-type transport in WSe₂ FET, which means TMDs oxidized by oxygen plasma treatment have promising application prospects in FETs.

  3.1.3   Fabricated by electrochemical methods

  Electrochemical oxidation is also a fast and simple way to prepare MoS₂ with incorporation of oxygen. Usually, this process should be carried on in acidic environment. Deng et al. [91] reported that the MoS₂ is oxidized at the potentials above 0.65 V. The preparation process of oxidized MoS₂ they reported, which is representative in electrochemical oxidation, is shown in Fig. 6A. Firstly, the MoS₂ mesh (MM) was fabricated by the reaction of cleaned Mo mesh and S powder under the condition of 550 ℃ for 20 min. Furthermore, the prepared MoS₂ should be calcined at 350 ℃ with Ar with the intention to remove the excess S as possible. Finally, the MoS₂ mesh was implemented via cyclic voltammetry (CV) scans between +0.273 V vs. RHE and +1.273 V vs. RHE in 0.5 mol/L H₂SO₄ electrolyte by electrochemical oxidation for several cycles. In Fig. 6B, before the voltage increases to +0.873 V vs. RHE, the oxidation peaks of MoS₂ cannot be observed. As the applied voltage grows higher than +0.873 V, the current density increased with the cyclic voltammograms. In addition, the appeared hysteresis loop of CV curves proves the irreversible reaction from Mo⁴⁺ to Mo⁶⁺, implying the incorporation of oxygen into MoS₂ which forms MoS_2−xO_x. In their research, the MoS₂ mesh with oxygen incorporation enhances the character of HER catalysts.

  Figure 6

  

  Figure 6.  (A) The synthesis of MoS₂/Mo mesh (MM), oxygen doped MoS₂/MM by anodic activation and the electrolyser and corresponding CV curve of anodic activation. (B) Cyclic voltammograms within the different range from +0.273 V to +1.473 V vs. RHE. Reproduced with permission [91]. Copyright 2019, Elsevier. (C, D) XPS spectra of MoS₂ on Toray paper recorded at different stages of its life. (1) As prepared after sulfidation. (2a) After initial activity measurements of the HER (CVs between +0.1 and 0.4 V vs. NHE). (2b) Sample 2a after measurements of the HER and subsequent oxidation/deactivation (CVs between +1.4 V and 0.4 V vs. NHE) and removal from the electrolyte at 0.32 V vs. NHE. (3) After measurements of the HER and subsequent oxidation/deactivation (CVs between +1.4 V and 0.4 V vs. NHE) and removal from the electrolyte at 0.4 V vs. NHE. Reproduced with permission [92]. Copyright 2009, The Royal Society of Chemistry.

  DownLoad: Full-Size Img PowerPoint

  Meanwhile, the electrochemical oxidation of MoS₂ also exists in HER process. Bonde et al. [92] reported the change of MoS₂ composition (Figs. 6C and D). Through XPS spectra, compared with the primitive MoS₂ (No. 1), the MoS₂ tested for the HER (No. 2a) sweeping between +0.45 and −0.1 V vs. NHE only appears a peak of SO₄²⁻, which is attributed to the H₂SO₄ condition. When the sweeping section enlarges which is between −0.4 V and +1.4 V vs. NHE (No. 2b), a significance reduction of the Mo 3d, Mo 3p, S 2s and S 2p peaks was observed. Nevertheless, the evidence which proves the existence of MoO₃ was not detected. Considering the sample 2b was pulled out of solution at −0.32 V vs. NHE, which may lead to the dissolution of MoO₃, the sample 3 was pulled out at +0.4 V vs. NHE. This time, the peaks corresponding to MoO₃ appear. The result conforms the theory mentioned above. In their research, the reaction was between two extreme conditions shown as equations as follows, which could be changed as the voltage and concentration of reactant:

  (1)

  (2)

  Thus, the HER catalyst contained with primitive MoS₂ may face the problem of oxidation. Fortunately, the pre-treatment of electrochemical oxidation of MoS₂ into MoS_2−xO_x could handle this and enhance the catalytic performance, which avoid this dilemma. The research of Bonde et al. [92] also remind us that the electrochemical oxidation of MoS₂ should be carried out under restricted condition in avoid of the complete oxidation into MoO₃.

  In summary, the TMD with O atom doping structure could be fabricated by various oxidation strategies. Notably, whatever the oxidation strategy is, it is significant to avoid the transformation of MoS_2−xO_x into MoO₃. The formation of S-O bonds in MoS_2−xO_x greatly enhance the chemical performance and this atomic structure could be widely used in applications of HER and FETs.

  3.2 TMDs with oxygen bound up atop sulfur atoms

  In some special circumstances, the O atom could absorb on top of sulfur, forming a unique atomic structure. It has been reported that the presence of ultraviolet light is indispensable to form this unique S-O bond [93,94]. This oxidation strategy could preserve the primitive MoS₂ structural and electronic properties as the formation of new S-O bonds.

  Ultraviolet-ozone treatment provides a brand-new method to fabricate a special structure of oxidized MoS₂. Similar to other methods, the UV-O₃ treatment, as a way of strong oxidation, could transform MoS₂ into MoO₃, reported by Jung et al. [95]. In their research, the 1 L MoS₂ are treated by UV-O₃ for 5 min at the irradiance of 58 mW/cm². As illustrated in Figs. 7A and B, after 5-min UV-O₃ treatment, both the exfoliated flakes and CVD thin films of MoS₂ become transparent. As discussed above, this indicates that despite of the number of SVs, the MoS₂ could be oxidized to the transparent MoO₃ which conforms to the previous reports. The optical absorbance of MoS₂ thin films (Fig. 7C) clarifies that when the absorbance peaks of the primitive MoS₂ appears which attributes to excitonic transitions (A and B), the MoS₂ thin films treated by 5-min UV-O₃ exposure do not exhibit any peaks throughout the same range of the wavelength. Since light yellow-green MoS₂ thin films are transparent to visible light after 5 min UV-O₃ treatment, it is expected that the band gap of the original 1 L MoS₂ (~1.88 eV) becomes wider after UV-O₃ treatment. This is consistent with the wide band gap of MoO₃. UV-O₃-treated transparent MoS₂ indicates the formation of MoO₃ via 5-min UV-O₃ treatment.

  Figure 7

  

  Figure 7.  Mechanically exfoliated flakes and CVD thin films of MoS₂ monolayers (A) before and (B) after 5-min UV-O₃ treatment (dotted region indicates monolayer). (C) Optical absorbance of CVD thin films before and after 5-min UV-O₃ treatment. Reproduced with permission [95]. Copyright 2019, Springer. (D) X-ray photoelectron spectra corresponding to the S 2p and Mo 3d (E) (5 × 5) supercell of MoS₂ showing the most energetically stable adsorption sites for oxygen: Oxygen on top of sulfur (O_ads) and substitutional oxygen on the SV (O_S). Reproduced with permission [96]. Copyright 2014, American Institute of Physics.

  DownLoad: Full-Size Img PowerPoint

  Notably, Azcatl et al. [96] reported that they found a non-destructive method to treat MoS₂ with UV-O₃ exposure without breaking Mo-S bonds, namely without the formation of MoO₃. The XPS spectra of MoS₂ treated in their research is given in Fig. 7D. Compared with the initial as-exfoliated MoS₂, the additional doublet peaks of the S 2p spectrum at 164.8 eV indicates the oxidation of sulfur. The increase of binding energy position of S 2p feature proves the sulfur is not the state of S²⁻ in MoS₂. However, the position is lower than that of S⁴⁺ reported previously. Considering the XPS spectra also illustrates that no formation of Mo-Mo and Mo-O bonds are detected, the oxidation of sulfur could be deduced as the formation of S-O bonds without the break of Mo-S bonds. The DFT modeling (Fig. 7E) clarifies that there are two different absorption sites of oxygen adatom: Oxygen atop sulfur and the substitution of SVs, which has the formation energies of −0.81 and −1.88 eV, respectively. The DFT result certifies that the formation of S-O bonds is energetically favorable which could occur without Mo-S scission. The result is in conformity with the states of Mo and S observed in XPS spectra. The MoS₂ has special structural characteristics different from other oxidized variants. The unique S-O bonds may arise some potential applications including atomic layer deposition [97] as reported. Huang et al. [66] also oxidized ReS₂ and ReSe₂ with UV light and got the O atom bound up atop S atom. Furthermore, they found that the formation of this structure is reversible. When the ReS₂ or ReSe₂ bound up atop O atom is exposure in laser, the oxidized ReS₂ or ReSe₂ could be transformed into the pristine ones. When studying on the transport properties of FET devices fabricated by ReS₂, they found that after UV treatment the on-state channel current drops by 1–2 orders of magnitude on average compared with the pristine ones. Considering that the reaction process is reversible, this property could be used to realize the switch of FETs under certain conditions.

  4. Oxidized variants in the form of heterojunction

  Obviously, the oxidation process of TMDs could be partially and locally, which forms the TMD-TMO (where TMO refers to transition metal oxides) heterojunction. Considering the heterojunction structure enhances the quantum effect and electron mobility of TMDs, it is valuable to be prepared to apply in the fields such as field effect transistors [98,99], solar cells [100] and catalysts [101–103].

  4.1 The bi-compound heterojunction

  The bi-compound heterojunction is prepared with precise oxidation strategies and usually has the special shape. The unique shape of heterojunction may increase the active sites which improves the performance of products.

  Laser-induced oxidation reaction is a directional, fast process with low activation energy and high selectivity. Considering the laser treatment could oxidize the TMDs locally and precisely which forms a straight channel, it is a useful method to prepare heterojunction of TMDs. Depending on this, Alrasheed et al. [104] investigated the theory of laser treating MoS₂ nanosheets. As shown in Fig. 8A, they treated the MoS₂ nanosheets on SiO₂ substrate with a 532-nm laser. Considering low laser power and ambient condition, the MoS₂ could not be oxidized lacking in energy, which gives rise to the amorphous MoS₂ etching and deposition upon the surface of the flakes. By contrast, when rising the laser power, the oxidized MoS₂ nanoparticles and nanosheets are able to be formed. Furthermore, as the removal of air, the nanosheet oxidation could not be observed. According to this mechanism, the oxidation of MoS₂ induced by laser could be influenced with circumstances including the laser power and the air condition. During i-XPS measurements, the optical images of native and laser treated MoS₂ are illustrated in Fig. 8B. When investigating the S 2s region at around 227 eV, the sulfur content of the native flakes remains which means the surface content is still MoS₂ (Fig. 8C). In the contrast, the region of treated flake grows dark, indicating the transformation of native MoS₂. When investigating the Mo 3d signal at 235 eV (Fig. 8D), only the treated flake is present, which proves the existence of oxidized MoS₂ (MoO_x). The masks of the S_sulfide, Mo_ox, and SiO₂ images with intention to identify regions of interest (ROIs) with minimal overlap resulting from masks were created from the S_sulfide, Mo_ox, and SiO₂ images to identify regions of the laser-treated and the native MoS₂ with the SiO₂ substrate is shown to extract spectra from the unprocessed images (Fig. 8E).

  Figure 8

  

  Figure 8.  (A) Laser-induced thinning conditions mostly result in nanoparticle formation (blue circles) on the MoS₂ surface, however different conditions impact surface chemistry. (1) Low-power laser and the presence of atmospheric oxygen retains the MoS₂ chemical structure. (2) High-power laser and the presence of atmospheric oxygen results in molybdenum oxidation. (3) High-power laser in the presence of vacuum also retains MoS₂ chemical structure. i-XPS measurements on native and laser-treated MoS₂ nanosheets treated in ambient conditions. (B) Optical image of the area containing native and laser-treated nanosheets. (C) Processed image attributed to S_sulfide 2s signal. (D) Processed image attributed to Mo_ox 3d signal. (E) Composite map of masks associated with processed images of Mo_ox (green), S_sulfide (red), and SiO₂ (blue) signals. Yellow, teal, and pink reflect an overlap in masks, while black reflects an absence in masks. (F) i-XPS spectra from the raw images based on pixels from the ROIs defined in the composite map. The Mo 3d and S 2s regions (left) and the O 1s region (right). Reproduced with permission [104]. Copyright 2018, American Chemical Society.

  DownLoad: Full-Size Img PowerPoint

  Spectra for the Mo 3d, S 2s, and O 1s regions which connect with pixels from each mask are offered in Fig. 8F. The native MoS₂ shows the typical Mo 3d and S 2s peaks at 229.8 eV and 227.0 eV respectively which is consistent with the literature value of MoS₂. Nevertheless, sulfur is nearly undetectable with slightly increases in signal intensity which is concluded in the satellite peaks for the shifted Mo 3d peak in the laser-treated MoS₂ flakes, suggesting that the characteristics of sulfide are completely removed. It is worth noting that the spectral features widen at half-width, indicating the presence of a distribution of oxide (Mo⁴⁺ to Mo⁶⁺) species within the spectral envelope. However, in view of the binding energy of MoO₂ is close to that of MoS₂ and MoO₃ is usually found at locations where the binding energy is 0.5 eV to 1.0 eV higher, it is possible that an intermediate oxide component, such as Mo₂O₅, has been shown to be present near the observed fitting peak location.

  Besides, adjusting the treating position of the laser, heterojunction with TMO in the middle of the TMD plane. Wang et al. [105] reported a novel type of lateral TMD-TMO heterostructure. In their research, the metallic NbS₂ as the "canvas" was etched by the laser heating. Due to the existence of oxygen in air, the 3R-NbS₂ with layered structure was oxidized to orthorhombic T-Nb₂O₅ locally, creating a patterned NbS₂-Nb₂O₅-NbS₂ heterostructure which has strong electrical signals in a two-terminal sensor resulting from its perfect interface contacts.

  Thermal Annealing is a relatively simple form of oxidation modulation. Wu et al. [106] fabricated 1 L MoS₂ mesh by thinning the mechanically exfoliated multilayer MoS₂ nanosheets on SiO₂/Si substrates through thermal annealing in air. By means of this method, the above layers of MoS₂ nanosheets were etched, leaving only the bottom layer of MoS₂ with the existence of MoO₃ on the 1 L MoS₂. As illustrated in Figs. 9A–H, the optical images of exfoliated MoS₂ before and after annealing at different temperature ranging from 260 ℃ to 400 ℃ for 1 h clarified that when annealing at relative low temperature (at 260 and 300 ℃), the MoS₂ nanosheets remained their primitive shape. As the temperature grows higher, multilayer MoS₂ nanosheets begin to get thinner (at 330 ℃), and decomposed (at 400 ℃) ultimately. Considering the general accepted fact that MoS₂ would oxidize completely to MoO₃, in their research, they chose 330 ℃ as the annealing temperature avoiding the complete decomposition of MoS₂ nanosheets.

  Figure 9

  

  Figure 9.  Optical images of mechanically exfoliated MoS₂ nanosheets before (A-D) and after (E-H) thermal annealing in air for 1 h at 260 ℃ (E), 300 ℃ (F), 330 ℃ (G) and 400 ℃ (H). AFM height (I) and phase (J) images of MoS₂ nanosheet after thermal annealing for 10 h. AFM height (K) and phase (L) images of the MoS₂ nanosheet after thermal annealing for 15 h. (M) XPS spectrum of thermally annealed MoS₂ nanosheet. Reproduced with permission [106]. Copyright 2013, John Wiley & Sons, Ltd.

  DownLoad: Full-Size Img PowerPoint

  It is worth noting that this preparation process is similar to the research of Zhou et al. [77] mentioned above. However, the reaction product is different. The reason of this phenomenon is that Zhou et al. carries out the heating process without the introduction of carrier or reactive gas, while Wu et al. prepares the chemical product in air which introduces abundant air. Another report from Yao et al. [107] expounds the mechanism of the heating progress. Due to the presence of oxygen, heating the MoS₂ sample in the air causes MoO₃ sites to appear on the surface. Due to the hydrophilicity of these MoO₃ sites, water molecules in the air are adsorbed to the surface of the sample, which further promotes the oxidation process. The same operation in a dry environment can prevent the adsorption of water molecules, making MoO₃ more volatile, so the degradation of MoS₂ in air is significantly slowed down. This report provides a reasonable excuse for the difference of the two researches mentioned above. By virtue of this, we conclude that the introduction of oxygen and water molecules could lead to the oxidation of MoS₂ into MoO₃ through the heating process.

  Taking the double-layer (2 L) MoS₂ into account, after annealing for 10 h, the 2 L MoS₂ nanosheet was thinned to 1 L. The AFM height and phase images (Figs. 9I and J) implied that the MoS₂ nanosheet annealed after 10 h has a smooth surface. It was worth noting that upon the relative smooth surface there existed small triangular pits with less than 200 nm size. Considering this, it was easy to suppose that if lengthening the annealing time, the pits may grow larger, in which case a MoS₂ mesh would form. As expected, when extending the annealing time to 15 h, the AFM height and phase images (Figs. 9K and L) showed the formation of MoS₂ mesh due to high density triangular pits resulting from the anisotropic etching. Notably, compared with MoS₂ nanosheet (Figs. 9K and L), the fabricated MoS₂ mesh owned a rougher surface, which arose from the more formation of MoO₃. As characterized by XPS (Fig. 9M), Mo 3d peaks at 232.2 eV and 235 eV, corresponding to Mo⁶⁺, and Mo 3d peaks at 229.5 eV and 232.5 eV, corresponding to Mo⁴⁺. Unifying with the practical situation, the Mo⁶⁺ are Mo⁴⁺ were contributed by MoO₃ and MoS₂ respectively. This result provided the evidence of the formation of MoO₃. In this way, the formation of triangular pits could be attributed to the evaporation of MoO₃. This research opened up a novel way to make MoS₂ mesh with MoO₃ doping.

  The heterojunction with oxidation variants of TMDs could be used in some fields. For example, Almora et al. [108] reported that when heterojunction with oxidation variants like V₂O₅ and MoO₃ are made into solar cells, they have larger power conversion efficiency than before. And V₂O₅, due to its unique characteristics, is observed that flat solar cell architectures allow a significant increase in passivation of the heterojunction interface. The improvement has been supported by the exploration of the judge of minority carrier lifetimes and the depletion layer features. The more thorough understanding of heterojunction with TMD oxidized variants could result in dramatic improvements for novel solar cells. As a matter of fact, the types of heterojunctions of TMD oxidized variants are myriad, and the structures and the combination of different compounds may produce different properties. Thus, there is still a long way to go to research on the fabrication of heterojunctions of TMD oxidized variants to explore more different types of applications.

  4.2 The multi-compound heterojunction

  In many cases, different oxides may be produced on the same TMD basal surface due to different oxidation sites and degrees [109,110]. Thus, some unique chemical properties may arise due to the wide variety of compounds. The reaction with H₂O₂ reagent is a typical method which produces a certain amount of MoS_xO_y in addition to MoO₃. There are also some special oxidized variants reported in few other reports, which confirm the multi-compound heterojunction with more oxidized variants could emerge.

  As the reagent with strong oxidizing property, H₂O₂ is capable to transform MoS₂ into MoO₃. In light of this, to take H₂O₂ etching into multiple dimension application, we should add some limiting conditions in the process of reaction to decrease the oxidation rate in case of the MoS₂ decomposing to MoO₃ completely. As clarified in Fig. 10A, the mechanism of reaction of MoS₂, proposed by Yun et al. [111], could be described as follows: Firstly, due to the more highly active suspended bonds distributing at the edge of bulk MoS₂. As the process of oxidation, the interlayer distance of MoS₂ is expanded, which gives rise to the penetration of H₂O₂ into the expanded interlayer of MoS₂. Then, the interlayer of MoS₂ is partly oxidized to MoO₃. Meanwhile, the interlayer force gets weaker. Therefore, finally through ultrasound treatment, the atomically thin MoS₂/MoO₃ nanosheets can be prepared. The reaction allowed the equation:

  (3)

  Figure 10

  

  Figure 10.  (A) The schematic of exfoliation mechanism of MoO₃/MoS₂. Reproduced with permission [111]. Copyright 2014, John Wiley & Sons, Ltd. (B) Schematic illustration of MoS₂ exfoliation and hybrid thin film preparation by centrifugation-coating. (C) Mo 3d spectra of MoS₂ oxidized by H₂O₂ (10%) for 15 min to light and to darkness. (D) Illustration of the hybrid MoO₃: MoS₂ synthetic process, and the chemicals structure of MoO₃, MoS₂ and MoO_xS_y. Reproduced with permission [112]. Copyright 2022, Multidisciplinary Digital Publishing Institute.

  DownLoad: Full-Size Img PowerPoint

  Lamkaouane et al. [112] reported that the oxidation rate of MoS₂ could be restricted under darkness condition. However, in most instances, there exists more than one kind of MoS₂ oxides. Lamkaouane et al. proposed a simple method to realize the theory. As illustrated in Fig. 10B, they exfoliated MoS₂ by liquid-phase exfoliation method in distilled water. Then, after a chain of centrifugation and stirring operation, they added H₂O₂ to the MoS₂ dispersion with a volume ratio of 3:1 at ambient temperature. After full of reaction, they transferred the MoO₃/MoS₂ to another centrifuge tube and obtained the MoO₃/MoS₂ thin film through another centrifugation process. As a matter of fact, they prove the existence of MoO_xS_y in the MoO₃ thin film.

  As the XPS spectra shown in Fig. 10C, when using H₂O₂ with the concentration of 10%, after 15 min of oxidation at room conditions, the Mo 3d spectrum corresponds to two doublets: One situated at 232.9 eV and 236.0 eV results from the Mo 3d_5/2 and Mo 3d_3/2 of MoO₃. The other one situated at 231.7 eV and 234.8 eV corresponds to MoO_xS_y. It is obvious that no evidence shows the existence of MoS₂. According to the results, the MoS₂ is completely oxidized to MoO₃. In order to slow down the reaction, they tried to put the samples in the dark during the reaction. This time, after the same reaction time, the doublet situated at 229.0 eV and 233.0 eV appears, which proves the existence of primitive MoS₂. Fig. 10D illustrates the atomic structure of oxidized MoS₂. Due to the partial and different degrees of oxidation, the MoO_xS_y and MoO₃ could be detected in the MoS₂ nanoflakes. In summary, H₂O₂ owns the ability to oxidized MoS₂ to MoO₃ completely. With this method, H₂O₂ etching could be a straight way to prepare pure MoO₃. Meanwhile, to obtain the MoO_xS_yMoO₃−MoS₂ heterojunction, the light condition and the concentration of H₂O₂ should be restricted.

  In addition, in some reports, the researches gave the different reaction equations. For instance, Li et al. [113] gave the equation:

  (4)

  However, Dong et al. [114] found that MoS₂ reacts with H₂O₂ under certain conditions through their own research:

  (5)

  It can be seen that the specific reaction mechanism of MoS₂ and H₂O₂ is very complex, but one consensus could be drawn that the reaction of H₂O₂ and MoS₂ is violent, where a method to slow down the reaction must be taken.

  In conclusion, the heterojunction structures could increase the recombination rate of electrons and holes and enlarge the width of the quantum well of the TMD atomic structure, which may enhance the luminous efficiency and electronic properties. As reported, the TMD heterostructures could be used in light-emitting diodes [115], photodetectors [116], solar cells [117], etc. Furthermore, considering the diversity of the heterojunction structures, endless application prospects are waiting to be explored.

  5. Oxidized variants with uncommon structures

  In addition to the various oxidative variant structures mentioned above, there are some more uncommon atomic structures. For example, TiSe₂ oxidizes in air, forming a special TiO₂ structure with Se atoms sandwiched between layers.

  Sun et al. [118] reported a novel oxidation structure of metallic transition metal diselenides (Fig. 11A). In their research, the whole oxidation process was revealed as a layer-by-layer mode. Firstly, the topmost layer of TiSe₂ be oxidized spontaneously to TiSe_2−xO_x in air condition, which has the similar atomic structure with MoS_2−xO_x mentioned above. However, the TiSe_2−xO_x is unstable in air. The further oxidation starts from the defects and edges which have high chemical activities. Then, the metastable Se-O-Ti bonds are broken gradually, in which circumstance selenium invades into the van der Waals gaps and accumulates into larger bulks which expand the interlayer distance and thus increase the overall height of the oxidized flakes. The enlarged gaps increase the contact area of O₂ and the multilayer flakes which attributes to the complete oxidation of the whole basal plane from top to bottom in the layer-by-layer formation. Finally, the TiSe₂ is oxidized to amorphous TiO₂ with Se nanoparticles inserted in the adjacent layers.

  Figure 11

  

  Figure 11.  (A) Schematics of the oxidation process of 2D TiSe₂. Evolution of TiSe₂ thin flakes in the progress of oxidation under ambient condition. Raman mapping images of (B) as-made TiSe₂ flake, (C) partially oxidized TiSe₂ flake after exposure to air for 5 days and (D) completely oxidized TiSe₂ flake after exposure to air for 10 days, respectively. Insets: corresponding optical images of TiSe₂ flakes. Scale bars: 1 µm. AES mapping images of (E) O, (F) Ti and (G) Se elements on the heavily oxidized TiSe₂ flakes. Scale bars: 2 µm. XPS spectra for binding energies of (H) O 1s, (I) Ti 2p and (J) Se 3d on the surface of partially oxidized sample, respectively. (K) Typical Raman spectra of as-made, partially oxidized and completely oxidized TiSe₂ flakes (from top to bottom), respectively. Reproduced with permission [118]. Copyright 2017, John Wiley and Sons.

  DownLoad: Full-Size Img PowerPoint

  The Raman mapping images of TiSe₂ (Figs. 11B–D) as the time grows show the gradual oxidation process of TiSe₂. It is notably that the blue region becomes larger until it takes part of the whole basal plane, which indicates that the Raman active materials shifting from TiSe₂ to Se. Besides, the auger electron spectroscopy (AES) is utilized to characterize the elemental spatial distribution in the oxidized TiSe₂ flake (Figs. 11E–G). As the images illustrated, the Ti and O elements are homogeneously distributed in the oxidized TiSe₂ flakes, which indicates that the TiSe₂ is oxidized into TiO₂. Meanwhile, the amount of Se detected by AES is very low in the oxidized TiSe₂ flakes, clarifying that Se is stored deeper than the AES detection depth. To further confirm the elementary composition, the XPS spectra of the oxidized TiSe₂ is shown in Figs. 11H–J. The most prominent peak of O 1s locates at ~530–531 eV, which indicates the presence of Se-O-Ti bonds, besides the dominant O 1s peak caused by adsorption (~532 eV). Another small peak at ~529.9 eV is consistent with the O 1s binding energy of TiO₂. The doublet of Ti 2p at ~457.0 eV and ~463.6 eV also proves the existence of Se-O-Ti bonds in addition to those arisen by the primitive TiSe₂ and TiO₂. Furthermore, for Se 3d spectra, the peaks at ~54.8 eV and ~55.8 eV attribute to amorphous selenium and trigonal selenium, respectively, indicating the formation of Se elementary substance. Obviously, in the fully oxidized products, all Ti atoms exists in form of TiO₂ due to no signals of TiSe₂ and the intermediate are detected by XPS. The Raman spectra of partially oxidized and completely oxidized TiSe₂ (Fig. 11K) also shows that two new peaks appear at ~233 and ~253 cm⁻¹ after oxidation, which belong to trigonal selenium and amorphous selenium respectively, and gradually replace the original peaks of TiSe₂. This also indicates that one of the oxidation products is selenium.

  6. Applications of TMD oxidized variants

  The TMDs have been widely used in different fields due to their tunable bandgaps and high catalytic activities. If the oxidation strategies mentioned are well utilized, the TMDs can be appropriately modified to enhance activities of basal plane and electronic transmission efficiency, thereby expanding the applications of TMDs. The typical applications of TMD oxidized variants are shown in Fig. 12.

  Figure 12

  

  Figure 12.  The prospective applications of TMD oxidized variants summarized as examples.

  DownLoad: Full-Size Img PowerPoint

  (1) HER catalysts. Recently, the DFT calculation has proved that due to its proper hydrogen adsorption/desorption strength and good electrochemical stability [119–121], MoS₂ could be a promising HER catalyst if oxidation strategies are utilized to modify its low conductivity and increase active sites [122–126].

  (2) FETs. With the improvement of device requirements, the sizes of transistors are becoming thinner. The silicon transistor has gradually failed to meet the requirement due to its physical limit. Meanwhile, TMDs are about to become promising candidate materials for the next generation of electronic devices due to their ultra-thin bodies and suitable bandgaps [127–129]. The excellent properties of TMDs are owing to their atomic nature and absence of dangling bonds. However, some doping strategies such as ion injection cannot be utilized in TMDs because they damage the atomic thin structure and the quality of devices [130]. Fortunately, oxygen plasma treatment can solve this problem because it allows the TMD to form a gapless band with removed S atoms [89,60], which reduces the contact resistance and provide a route for future device application.

  (3) Solar cells. The TMD monolayers are potential materials to absorb sunlight and can be used to fabricate ultrathin photovoltaic devices due to their semiconductor properties. The researches have demonstrated that 1 nm thick active layers fabricated by TMD monolayers can obtain ~1% power conversion efficiency, which is about 1–3 orders of magnitude higher power density than the existing ultrathin solar cells [131,132]. When oxidation strategies are used to modify TMDs, the power conversion efficiency is better than before [108,117], which may arouse further studies on this field.

  (4) Photodetectors. The vertical stacking of TMDs which forms van der Waals heterostructures has potential optoelectronic properties. This structure could realize efficient electron-hole separation and be applied in the fabrication of photodetectors [116,133,134]. Through modulation strategies, the low absorption capacity of TMD may be conquered and used in the detection of polarized light [135–137].

  7. Conclusion and outlooks

  As promising atomic structures, the oxidized variants have drawn extensive attention in recent years. Nowadays, plenty of researches have been carried out to comprehend the physicochemical properties of the structures and the fabrication methods. In the previous sections, we summarize the typical oxidation structures, including variants without changes of intrinsic elements, variants with single oxygen atom doping, heterojunction, and dioxides with interlayered elementary substance. By virtue of these results, the solid foundation of the further researches would be conducted and the wide application fields of TMD oxidized variants would be explored, such as HER catalysts, FETs and solar batteries. Although dramatic progress has been made, there still exist some obstacles and challenges worthy of further exploration.

  (1) Although current studies have shown that precise TMD oxidation variants can be obtained through some specific ways, the preparation methods of some oxidation variants have strict requirements on reactants and reaction conditions. In these reactions, subtle changes in environmental conditions such as light and temperature may have an enormous impact on the oxidation products. Therefore, more stable and efficient preparation methods need to be explored in future research.

  (2) In recent years, relevant researches have obtained many valuable oxidation variants through previous experience. It is still doubtful whether the structural types of oxidation variants can be expanded. The key to solving this problem lies in theoretical calculation. In the future research, we need to establish the relationship between the intrinsic properties and properties of materials at the atomic scale through theoretical calculations, and strive to efficiently find new materials with practical application value.

  (3) At present, many kinds of TMD oxidation variants have been explored, but most of the researches on the specific properties of these oxidation variants still stay in the laboratory stage, and it is still unclear whether it can also be effective when it is applied to real life. In addition, there is still a long way to go from laboratory synthesis of oxidized variant materials to large-scale production, and more efforts are needed to promote the specific application of TMDs based oxidized variant materials.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  Acknowledgments

  This work was financially supported by the National Natural Science of China (No. 51902101), Natural Science of Jiangsu Province (No. BK20201381), and Science of Nanjing University of Posts and Telecommunications (No. NY219144).

  
  
• 1. [1]

     Y. Wang, L. Zhou, M. Zhong, et al., Nano Res. 15 (2022) 3675–3694. doi: 10.1007/s12274-021-3979-6

  2. [2]

     G.H. Han, D.L. Duong, D.H. Keum, S.J. Yun, Y.H. Lee, Chem. Rev. 118 (2018) 6297–6336. doi: 10.1021/acs.chemrev.7b00618

  3. [3]

     Z. Sofer, D. Sedmidubský, J. Luxa, et al., Chem. Eur. J. 23 (2017) 10177–10186. doi: 10.1002/chem.201701628

  4. [4]

     C. Zhang, Y. Chen, A. Johnson, et al., Nano Lett. 15 (2015) 6494–6500. doi: 10.1021/acs.nanolett.5b01968

  5. [5]

     C. Ruppert, B. Aslan, T.F. Heinz, Nano Lett. 14 (2014) 6231–6236. doi: 10.1021/nl502557g

  6. [6]

     F. Liu, J. Zhou, C. Zhu, Z. Liu, Adv. Funct. Mater. 27 (2017) 1602404. doi: 10.1002/adfm.201602404

  7. [7]

     N.A. Pike, A. Dewandre, B. van Troeye, X. Gonze, M.J. Verstraete, Phys. Rev. Mater. 3 (2019) 074009. doi: 10.1103/PhysRevMaterials.3.074009

  8. [8]

     E. Malic, G. Berghäuser, M. Feierabend, A. Knorr, Ann. Phys. 529 (2017) 1700097. doi: 10.1002/andp.201700097

  9. [9]

     D. Nutting, G.A. Prando, M. Severijnen, et al., Nanoscale 13 (2021) 15853–15858. doi: 10.1039/D1NR04723J

  10. [10]

      N.B. Shinde, B.D. Ryu, K. Meganathan, et al., ACS Appl. Nano Mater. 3 (2020) 7371–7376. doi: 10.1021/acsanm.0c01235

  11. [11]

      C. Fan, T. Li, Z. Wei, et al., Nanoscale 6 (2014) 14652–14656. doi: 10.1039/C4NR05111D

  12. [12]

      S. Wang, D. Zhang, B. Li, et al., Adv. Energy Mater. 8 (2018) 1801345. doi: 10.1002/aenm.201801345

  13. [13]

      V. Caciuc, N. Atodiresei, S. Blügel, Phys. Rev. Mater. 2 (2018) 084001. doi: 10.1103/PhysRevMaterials.2.084001

  14. [14]

      D. Ma, W. Ju, T. Li, et al., Appl. Surf. Sci. 364 (2016) 181–189. doi: 10.1016/j.apsusc.2015.12.142

  15. [15]

      H. Bai, Q. Wu, H. Ai, et al., Adv. Electron. Mater. 8 (2022) 2200209. doi: 10.1002/aelm.202200209

  16. [16]

      G. Wu, Y. Ren, X. He, et al., Phys. Rev. Appl. 13 (2020) 024027. doi: 10.1103/PhysRevApplied.13.024027

  17. [17]

      S. Imani Yengejeh, W. Wen, Y. Wang, Front. Phys. 16 (2020) 13502.

  18. [18]

      N.K. Nepal, L. Yu, Q. Yan, A. Ruzsinszky, Phys. Rev. Mater. 3 (2019) 073601. doi: 10.1103/PhysRevMaterials.3.073601

  19. [19]

      D. Saleta Reig, S. Varghese, R. Farris, et al., Adv. Mater. 34 (2022) 2108352. doi: 10.1002/adma.202108352

  20. [20]

      N. Morell, A. Reserbat-Plantey, I. Tsioutsios, et al., Nano Lett. 16 (2016) 5102–5108. doi: 10.1021/acs.nanolett.6b02038

  21. [21]

      P. Jiang, X. Qian, X. Gu, R. Yang, Adv. Mater. 29 (2017) 1701068. doi: 10.1002/adma.201701068

  22. [22]

      D. Wang, X. Mu, W. Cai, et al., Compos. A: Appl. Sci. Manuf. 121 (2019) 36–44. doi: 10.1016/j.compositesa.2019.03.011

  23. [23]

      G. Zhang, Y.W. Zhang, J. Mater. Chem. C 5 (2017) 7684–7698. doi: 10.1039/C7TC01088E

  24. [24]

      I. Pallecchi, N. Manca, B. Patil, L. Pellegrino, D. Marré, Nano Futures 4 (2020) 032008. doi: 10.1088/2399-1984/ab92f4

  25. [25]

      Q. Yun, L. Li, Z. Hu, et al., Adv. Mater. 32 (2020) 1903826. doi: 10.1002/adma.201903826

  26. [26]

      Q. Yun, Q. Lu, X. Zhang, C. Tan, H. Zhang, Angew. Chem. Int. Ed. 57 (2018) 626–646. doi: 10.1002/anie.201706426

  27. [27]

      X. Zhang, Z. Lai, Q. Ma, H. Zhang, Chem. Soc. Rev. 47 (2018) 3301–3338. doi: 10.1039/C8CS00094H

  28. [28]

      L. Lin, P. Sherrell, Y. Liu, et al., Adv. Energy Mater. 10 (2020) 1903870. doi: 10.1002/aenm.201903870

  29. [29]

      H. Huang, J. Zha, S. Li, C. Tan, Chin. Chem. Lett. 33 (2022) 163–176. doi: 10.1016/j.cclet.2021.06.004

  30. [30]

      T. Chowdhury, E.C. Sadler, T.J. Kempa, Chem. Rev. 120 (2020) 12563–12591. doi: 10.1021/acs.chemrev.0c00505

  31. [31]

      W. Cao, J. Kang, W. Liu, K. Banerjee, IEEE Trans. Electron Dev. 61 (2014) 4282–4290. doi: 10.1109/TED.2014.2365028

  32. [32]

      J. Lindlau, M. Selig, A. Neumann, et al., Nat. Commun. 9 (2018) 2586. doi: 10.1038/s41467-018-04877-3

  33. [33]

      H. Lee, K. Lee, Y. Kim, et al., Nanoscale 11 (2019) 22118–22124. doi: 10.1039/C9NR05065E

  34. [34]

      J. Ding, H. Zhao, X. Zhao, B. Xu, H. Yu, J. Mater. Chem. A 7 (2019) 13511–13521. doi: 10.1039/C9TA04033A

  35. [35]

      M.V. Strikha, A.I. Kurchak, A.N. Morozovska, Phys. Rev. Appl. 13 (2020) 014040. doi: 10.1103/PhysRevApplied.13.014040

  36. [36]

      H. Li, Y. Guo, W. Guo, Friction 10 (2022) 1851–1858. doi: 10.1007/s40544-021-0562-9

  37. [37]

      A. Rapuc, K. Simonovic, T. Huminiuc, A. Cavaleiro, T. Polcar, ACS Appl. Mater. Interfaces 12 (2020) 54191–54202. doi: 10.1021/acsami.0c16789

  38. [38]

      K. Chen, J. Pan, W. Yin, C. Ma, L. Wang, Chin. Chem. Lett. 34 (2023) 108226. doi: 10.1016/j.cclet.2023.108226

  39. [39]

      X. Yang, B. Li, Nanophotonics 9 (2020) 1557–1577. doi: 10.1515/nanoph-2019-0533

  40. [40]

      S.L. Li, K. Tsukagoshi, E. Orgiu, P. Samorì, Chem. Soc. Rev. 45 (2016) 118–151. doi: 10.1039/C5CS00517E

  41. [41]

      J.H. Park, S. Vishwanath, S. Wolf, et al., ACS Appl. Mater. Interfaces 9 (2017) 29255–29264. doi: 10.1021/acsami.7b08244

  42. [42]

      X. Cheng, L. Wang, L. Xie, et al., Chem. Eng. J. 439 (2022) 135757. doi: 10.1016/j.cej.2022.135757

  43. [43]

      S.S. Jo, A. Singh, L. Yang, et al., Nano Lett. 20 (2020) 8592–8599. doi: 10.1021/acs.nanolett.0c03263

  44. [44]

      Z. Luo, Y. Ouyang, H. Zhang, et al., Nat. Commun. 9 (2018) 2120. doi: 10.1038/s41467-018-04501-4

  45. [45]

      S. Kang, S. Kim, S. Jeon, et al., Nano Energy 58 (2019) 57–62. doi: 10.1016/j.nanoen.2019.01.025

  46. [46]

      Q. Lu, Y. Yu, Q. Ma, B. Chen, H. Zhang, Adv. Mater. 28 (2016) 1917–1933. doi: 10.1002/adma.201503270

  47. [47]

      Y. Chen, Z. Liu, J. Li, et al., ACS Nano 14 (2020) 10258–10264. doi: 10.1021/acsnano.0c03624

  48. [48]

      X. Hua, T. Axenie, M.N. Goldaraz, et al., ACS Appl. Mater. Interfaces 14 (2022) 2255–2262. doi: 10.1021/acsami.1c18991

  49. [49]

      S. Susarla, J.A. Hachtel, X. Yang, et al., Adv. Mater. 30 (2018) 1804218. doi: 10.1002/adma.201804218

  50. [50]

      X. Zhou, H. Hao, Y.J. Zhang, et al., Chem 7 (2021) 1626–1638. doi: 10.1016/j.chempr.2021.03.011

  51. [51]

      T. Wang, H. Zhu, J. Zhuo, et al., Anal. Chem. 85 (2013) 10289–10295. doi: 10.1021/ac402114c

  52. [52]

      Y. Shu, L. Zhang, H. Cai, et al., Sens. Actuator. B: Chem. 311 (2020) 127863. doi: 10.1016/j.snb.2020.127863

  53. [53]

      T. Li, H. Li, J. Yuan, et al., Materials 15 (2022) 2654. doi: 10.3390/ma15072654

  54. [54]

      M.Z.M. Nasir, Z. Sofer, M. Pumera, ChemElectroChem 2 (2015) 1713–1718. doi: 10.1002/celc.201500259

  55. [55]

      M. Ramya, P.S. Kumar, G. Rangasamy, et al., Environ. Res. 216 (2023) 114463. doi: 10.1016/j.envres.2022.114463

  56. [56]

      L. Yuwen, F. Xu, B. Xue, et al., Nanoscale 6 (2014) 5762–5769. doi: 10.1039/C3NR06084E

  57. [57]

      K. Czerniak-Łosiewicz, M. Świniarski, A.P. Gertych, et al., ACS Appl. Mater. In-terfaces 14 (2022) 33984–33992. doi: 10.1021/acsami.2c06578

  58. [58]

      P. Zhang, H. Xiang, L. Tao, et al., Nano Energy 57 (2019) 535–541. doi: 10.1016/j.nanoen.2018.12.045

  59. [59]

      C.S. Pang, T.Y.T. Hung, A. Khosravi, et al., Adv. Electron. Mater. 6 (2020) 1901304. doi: 10.1002/aelm.201901304

  60. [60]

      Y.C. Lin, B.M. Bersch, R. Addou, et al., Adv. Mater. Interfaces 7 (2020) 2000422. doi: 10.1002/admi.202000422

  61. [61]

      S. Park, S.Y. Kim, Y. Choi, et al., ACS Appl. Mater. Interfaces 8 (2016) 11189–11193. doi: 10.1021/acsami.6b01568

  62. [62]

      K.C. Kwon, C. Kim, Q.V. Le, et al., ACS Nano 9 (2015) 4146–4155. doi: 10.1021/acsnano.5b01504

  63. [63]

      N. Papadopoulos, J.O. Island, S. Herre, G.A. Steele, IEEE Trans. Electron. Dev. 65 (2018) 4053–4058. doi: 10.1109/TED.2018.2855215

  64. [64]

      H. Chen, X. He, Z. Zhou, et al., J. Nanobiotechnol. 20 (2022) 136. doi: 10.1186/s12951-022-01344-6

  65. [65]

      L. Yuwen, Y. Sun, G. Tan, et al., Nanoscale 10 (2018) 16711–16720. doi: 10.1039/C8NR04111C

  66. [66]

      L. Huang, T. Yang, L.W. Wong, et al., Adv. Funct. Mater. 31 (2021) 2009166. doi: 10.1002/adfm.202009166

  67. [67]

      X. Tan, W. Ding, Z. Jiang, L. Sun, Y. Huang, Nano Res. 15 (2022) 1973–1982. doi: 10.1007/s12274-021-3848-3

  68. [68]

      C. Zhang, D.F. Hu, J.W. Xu, et al., ACS Nano 12 (2018) 12347–12356. doi: 10.1021/acsnano.8b06321

  69. [69]

      U. Maitra, U. Gupta, M. De, et al., Angew. Chem. Int. Ed. 52 (2013) 13057–13061. doi: 10.1002/anie.201306918

  70. [70]

      H. Jiang, D. Ren, H. Wang, et al., Adv. Mater. 27 (2015) 3687–3695. doi: 10.1002/adma.201501059

  71. [71]

      H. Jia, R. Yang, A.E. Nguyen, et al., Nanoscale 8 (2016) 10677–10685. doi: 10.1039/C6NR01118G

  72. [72]

      X. Pan, M. Yan, C. Sun, et al., Adv. Funct. Mater. 31 (2021) 2007840. doi: 10.1002/adfm.202007840

  73. [73]

      W. Yu, Z. Wang, X. Zhao, et al., Adv. Funct. Mater. 30 (2020) 2003057. doi: 10.1002/adfm.202003057

  74. [74]

      S.X. Leong, C.C. Mayorga-Martinez, X. Chia, et al., ACS Appl. Mater. Interfaces 9 (2017) 26350–26356. doi: 10.1021/acsami.7b06898

  75. [75]

      Y. Li, X. Wang, F. Wang, Y. Tian, B. Yang, Vacuum 199 (2022) 110926. doi: 10.1016/j.vacuum.2022.110926

  76. [76]

      W. Li, J. Huang, B. Han, et al., Adv. Sci. 7 (2020) 2001080. doi: 10.1002/advs.202001080

  77. [77]

      H. Zhou, F. Yu, Y. Liu, et al., Nano Res. 6 (2013) 703–711. doi: 10.1007/s12274-013-0346-2

  78. [78]

      D. Lv, H. Wang, D. Zhu, et al., Sci. Bull. 62 (2017) 846–851. doi: 10.1016/j.scib.2017.05.016

  79. [79]

      A. Syari'ati, S. Kumar, A. Zahid, et al., Chem. Commun. 55 (2019) 10384–10387. doi: 10.1039/C9CC01577A

  80. [80]

      X. Ma, J. Zhang, Y. Sun, et al., ACS Appl. Mater. Interfaces 14 (2022) 47288–47299. doi: 10.1021/acsami.2c14200

  81. [81]

      A.W. Robertson, Y.C. Lin, S. Wang, et al., ACS Nano 10 (2016) 10227–10236. doi: 10.1021/acsnano.6b05674

  82. [82]

      V. Paolucci, J. De Santis, V. Ricci, et al., ACS Sens. 7 (2022) 2058–2068. doi: 10.1021/acssensors.2c00887

  83. [83]

      K. Czerniak-Łosiewicz, M. Świniarski, A.P. Gertych, et al., ACS Appl. Mater. In-terfaces 14 (2022) 33984–33992. doi: 10.1021/acsami.2c06578

  84. [84]

      R. Ghosh, M. Singh, L.W. Chang, et al., ACS Nano 16 (2022) 5743–5751. doi: 10.1021/acsnano.1c10772

  85. [85]

      J. Pető, T. Ollár, P. Vancsó, et al., Nat. Chem. 10 (2018) 1246–1251. doi: 10.1038/s41557-018-0136-2

  86. [86]

      C.S. Cucinotta, K. Dolui, H. Pettersson, et al., J. Phys. Chem. C 119 (2015) 15707–15715. doi: 10.1021/acs.jpcc.5b03212

  87. [87]

      J. Gao, B. Li, J. Tan, et al., ACS Nano 10 (2016) 2628–2635. doi: 10.1021/acsnano.5b07677

  88. [88]

      S. Kim, M.S. Choi, D. Qu, et al., 2D Mater 3 (2016) 035002. doi: 10.1088/2053-1583/3/3/035002

  89. [89]

      J. Wang, X. Feng, D. Hedman, et al., Electrochem. Commun. 137 (2022) 107250. doi: 10.1016/j.elecom.2022.107250

  90. [90]

      J. Meng, J. Zhu, S. Jia, X. Ren, Mater. Sci. 26 (2020) 143–146.

  91. [91]

      Y. Deng, Z. Liu, A. Wang, et al., Nano Energy 62 (2019) 338–347. doi: 10.1016/j.nanoen.2019.05.036

  92. [92]

      J. Bonde, P.G. Moses, T.F. Jaramillo, J.K. Nørskov, I. Chorkendorff, Faraday Discuss. 140 (2009) 219–231. doi: 10.1039/B803857K

  93. [93]

      K.M. Freedy, M.G. Sales, P.M. Litwin, et al., Appl. Surf. Sci. 478 (2019) 183–188. doi: 10.1016/j.apsusc.2019.01.222

  94. [94]

      W. Zhang, T. Kanazawa, Y. Miyamoto, Appl. Phys. Express 12 (2019) 065005. doi: 10.7567/1882-0786/ab2199

  95. [95]

      C. Jung, H.I. Yang, W. Choi, Nanoscale Res. Lett. 14 (2019) 278. doi: 10.1186/s11671-019-3119-3

  96. [96]

      A. Azcatl, S. McDonnell, K. Santosh, et al., Appl. Phys. Lett. 104 (2014) 111601. doi: 10.1063/1.4869149

  97. [97]

      H. Liu, K. Xu, X. Zhang, P.D. Ye, Appl. Phys. Lett. 100 (2012) 152115. doi: 10.1063/1.3703595

  98. [98]

      W. Choi, J. Ahn, K.T. Kim, et al., Adv. Mater. 33 (2021) 2103079. doi: 10.1002/adma.202103079

  99. [99]

      J.Y. Lim, M. Kim, Y. Jeong, et al., npj 2D Mater. Appl. 2 (2018) 37. doi: 10.1038/s41699-018-0082-2

  100. [100]

       X. Yu, K. Sivula, Curr. Opin. Electrochem. 2 (2017) 97–103. doi: 10.1016/j.coelec.2017.03.007

  101. [101]

       A. Vijayan, N. Sandhyarani, J. Colloid Interface Sci. 623 (2022) 819–831. doi: 10.1016/j.jcis.2022.05.107

  102. [102]

       S. Seo, S. Kim, H. Choi, et al., Adv. Sci. 6 (2019) 1900301. doi: 10.1002/advs.201900301

  103. [103]

       H.T.T. Nguyen, L.A. Adofo, S.H. Yang, et al., Adv. Funct. Mater. 33 (2023) 2209572. doi: 10.1002/adfm.202209572

  104. [104]

       A. Alrasheed, J.M. Gorham, B.C. Tran Khac, et al., ACS Appl. Mater. Interfaces 10 (2018) 18104–18112. doi: 10.1021/acsami.8b04717

  105. [105]

       B. Wang, H. Luo, X. Wang, et al., Nano Res. 13 (2020) 2035–2043. doi: 10.1007/s12274-020-2872-z

  106. [106]

       J. Wu, H. Li, Z. Yin, et al., Small 9 (2013) 3314–3319. doi: 10.1002/smll.201301542

  107. [107]

       K. Yao, J.D. Femi-Oyetoro, S. Yao, et al., 2D Mater 7 (2019) 015024. doi: 10.1088/2053-1583/ab5971

  108. [108]

       O. Almora, L.G. Gerling, C. Voz, et al., Sol. Energy Mater. Sol. Cells 168 (2017) 221–226. doi: 10.1016/j.solmat.2017.04.042

  109. [109]

       K. Kang, K. Godin, Y.D. Kim, et al., Adv. Mater. 29 (2017) 1603898. doi: 10.1002/adma.201603898

  110. [110]

       H. Li, J. Wu, Z. Yin, H. Zhang, Acc. Chem. Res. 47 (2014) 1067–1075. doi: 10.1021/ar4002312

  111. [111]

       J.M. Yun, Y.J. Noh, C.H. Lee, et al., Small 10 (2014) 2309. doi: 10.1002/smll.201470068

  112. [112]

       H. Lamkaouane, H. Ftouhi, M. Richard-Plouet, et al., Nanomaterials 12 (2022) 3071. doi: 10.3390/nano12173071

  113. [113]

       Q. Li, B. Hu, Q. Yang, et al., Environ. Res. 191 (2020) 110227. doi: 10.1016/j.envres.2020.110227

  114. [114]

       L. Dong, S. Lin, L. Yang, et al., Chem. Commun. 50 (2014) 15936–15939. doi: 10.1039/C4CC07238C

  115. [115]

       F. Withers, O. Del Pozo-Zamudio, A. Mishchenko, et al., Nat. Mater. 14 (2015) 301–306. doi: 10.1038/nmat4205

  116. [116]

       M. Massicotte, P. Schmidt, F. Vialla, et al., Nat. Nanotechnol. 11 (2016) 42–46. doi: 10.1038/nnano.2015.227

  117. [117]

       N.R. Paudel, C. Xiao, Y. Yan, Prog. Photovolt. 23 (2015) 437–442. doi: 10.1002/pip.2446

  118. [118]

       L. Sun, C. Chen, Q. Zhang, et al., Angew. Chem. Int. Ed. 56 (2017) 8981–8985. doi: 10.1002/anie.201612605

  119. [119]

       T.F. Jaramillo, K.P. Jørgensen, J. Bonde, et al., Science 317 (2007) 100–102. doi: 10.1126/science.1141483

  120. [120]

       H. Wang, X. Xiao, S. Liu, et al., J. Am. Chem. Soc. 141 (2019) 18578–18584. doi: 10.1021/jacs.9b09932

  121. [121]

       W. Yin, L. Yuan, H. Huang, et al., Chin. Chem. Lett. 35 (2024) 108351. doi: 10.1016/j.cclet.2023.108351

  122. [122]

       D. Yan, C. Mebrahtu, S. Wang, et al., Angew. Chem. Int. Ed. 62 (2023) e202214333. doi: 10.1002/anie.202214333

  123. [123]

       D. Yan, C. Xia, W. Zhang, et al., Adv. Energy Mater. 12 (2022) 2202317. doi: 10.1002/aenm.202202317

  124. [124]

       P. Fan, Y. He, J. Pan, et al., Chin. Chem. Lett. 35 (2024) 108513. doi: 10.1016/j.cclet.2023.108513

  125. [125]

       C. Chang, L. Wang, L. Xie, et al., Nano Res. 15 (2022) 8613–8635. doi: 10.1007/s12274-022-4507-z

  126. [126]

       M. Liu, H. Li, S. Liu, et al., Nano Res. 15 (2022) 5946–5952. doi: 10.1007/s12274-022-4267-9

  127. [127]

       Y. Sheng, L. Zhang, F. Li, et al., J. Mater. Sci. Technol. 69 (2021) 15–19. doi: 10.1016/j.jmst.2020.05.079

  128. [128]

       M. Wang, W. Wang, Y. Zhang, et al., Chin. Chem. Lett. 32 (2021) 3118–3122. doi: 10.1016/j.cclet.2021.03.048

  129. [129]

       J. Li, J. Pan, W. Yin, et al., Chin. Chem. Lett. 34 (2023) 108049. doi: 10.1016/j.cclet.2022.108049

  130. [130]

       K. Xu, Y. Zhao, Z. Lin, et al., Semiconduct. Sci. Technol. 32 (2017) 124002. doi: 10.1088/1361-6641/aa8ed3

  131. [131]

       M. Bernardi, M. Palummo, J.C. Grossman, Nano Lett. 13 (2013) 3664–3670. doi: 10.1021/nl401544y

  132. [132]

       M. Li, L. Wang, X. Zhang, et al., Chin. Chem. Lett. 34 (2023) 107775. doi: 10.1016/j.cclet.2022.107775

  133. [133]

       J. Yan, X. Yang, X. Liu, et al., Adv. Sci. 10 (2023) 2207022. doi: 10.1002/advs.202207022

  134. [134]

       T. Shen, F. Li, Z. Zhang, L. Xu, J. Qi, ACS Appl. Mater. Interfaces 12 (2020) 54927–54935. doi: 10.1021/acsami.0c14161

  135. [135]

       B. Xu, H. Shen, Y. Xu, et al., J. Alloys Compd. 874 (2021) 159864. doi: 10.1016/j.jallcom.2021.159864

  136. [136]

       Y. Fu, Y. Liu, K. Ma, et al., J. Alloys Compd. 819 (2020) 153063. doi: 10.1016/j.jallcom.2019.153063

  137. [137]

       Z. Li, Z. Hu, J. Peng, et al., Adv. Funct. Mater. 24 (2014) 1821–1830. doi: 10.1002/adfm.201302967

• 

  Figure 1  (A) The schematic of the process driven by a lateral inward electrochemical reaction and water molecules go into bulk phase through the formed 2D channel. (B) The schematic of the thickness of the edge and the basal plane of MoS₂ after exfoliation. (C) The magnified STEM image of the transition zone from bulk to edge. (D) The polarization curves of MoS₂ after different CV cycles treatment (0, 200, and 300 cycles). The dash lines mark the overpotential at 10 and 20 mA/cm². (E) The corresponding Tafel slope plots. Reproduced with permission [72]. Copyright 2021, John Wiley & Sons, Ltd.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (A) An atomic mechanism showing the formation of triangular pits in MoS₂ layers. A pit in MoS₂ could generally expose various types of edges, including ZZ-Mo and ZZ-S₂. Typical AFM images showing the formation of well oriented equilateral triangular pits on 1 L MoS₂ at various temperatures: (B) 325 ℃, (C) 345 ℃, (D) 355 ℃, and (E) 365 ℃. Raman spectra of n-layer MoS₂ sheets after oxidative etching at (F) 355 ℃ and (G) 365 ℃. Reproduced with permission [77]. Copyright 2013, Tsinghua University Press.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (A) Schematic atomic structure of the MoS_2−xO_x solid solution monolayer. (B) Atomic-resolution STM images (5 mV, 2 nA) of an exfoliated MoS₂ single layer after 1 month. (C) The process of chalcogenide atom vacancy formation through oxidation is characterized by a negative oxidation enthalpy (ΔE) for the defect-free 2D MoS₂ basal plane. (D) Kinetic energy barriers calculated by the NEB model for 2D MoS₂ crystals reveal substantially lower barriers for MoS₂ of ~1 eV height that can be overcome even at room temperature on a months-long timescale. (E) Linear sweep voltammogram curves and (F) corresponding Tafel plotsfor Au substrate, MoS₂ single layer, MoS_2−xO_x single layer (1 year old) and Pt substrate. Reproduced with permission [85]. Copyright 2018, London: Nature Publishing Group.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Scanning electron microscopy images of (A) fresh MoS₂, and (B) aged MoS₂ after 6 months without desiccant and (C) after 1 year without desiccant. All the samples were stored in Class 100 cleanroom conditions. Photoluminescence spectra of fresh and aged TMDs. PL intensity map (centered at the A exciton) of fresh MoS₂ (D), and MoS₂ aged for 1 year under ambient conditions (E). The inset optical images show the MoS₂ flakes from which the data were taken. Representative PL spectra of the fresh and 1-year-aged MoS₂ (F). The 1 L MoS₂ films were aged for 1 year under ambient conditions (i.e., room temperature and atmospheric pressure). Reproduced with permission [87]. Copyright 2016, American Chemical Society.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  (A) Illustrations for the atomic states of MoS₂ for showing atomic-level etching depending on the two different plasma treatment processes. (B) An AFM image before and after plasma etching. (C) Mo 3d, (D) S 2p and (E) O 1s XPS spectra of pristine MoS₂ and after plasma treatment. Schematic diagrams showing the structures and energy band diagrams of a MoS₂ FET Energy band diagrams at V_D > 0 and V_G < 0 (off-state) of MoS₂ (F) before and (G) after plasma treatment. Reproduced with permission [88]. Copyright 2016, IOP Publishing.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  (A) The synthesis of MoS₂/Mo mesh (MM), oxygen doped MoS₂/MM by anodic activation and the electrolyser and corresponding CV curve of anodic activation. (B) Cyclic voltammograms within the different range from +0.273 V to +1.473 V vs. RHE. Reproduced with permission [91]. Copyright 2019, Elsevier. (C, D) XPS spectra of MoS₂ on Toray paper recorded at different stages of its life. (1) As prepared after sulfidation. (2a) After initial activity measurements of the HER (CVs between +0.1 and 0.4 V vs. NHE). (2b) Sample 2a after measurements of the HER and subsequent oxidation/deactivation (CVs between +1.4 V and 0.4 V vs. NHE) and removal from the electrolyte at 0.32 V vs. NHE. (3) After measurements of the HER and subsequent oxidation/deactivation (CVs between +1.4 V and 0.4 V vs. NHE) and removal from the electrolyte at 0.4 V vs. NHE. Reproduced with permission [92]. Copyright 2009, The Royal Society of Chemistry.

  下载: 全尺寸图片 幻灯片
  

  Figure 7  Mechanically exfoliated flakes and CVD thin films of MoS₂ monolayers (A) before and (B) after 5-min UV-O₃ treatment (dotted region indicates monolayer). (C) Optical absorbance of CVD thin films before and after 5-min UV-O₃ treatment. Reproduced with permission [95]. Copyright 2019, Springer. (D) X-ray photoelectron spectra corresponding to the S 2p and Mo 3d (E) (5 × 5) supercell of MoS₂ showing the most energetically stable adsorption sites for oxygen: Oxygen on top of sulfur (O_ads) and substitutional oxygen on the SV (O_S). Reproduced with permission [96]. Copyright 2014, American Institute of Physics.

  下载: 全尺寸图片 幻灯片
  

  Figure 8  (A) Laser-induced thinning conditions mostly result in nanoparticle formation (blue circles) on the MoS₂ surface, however different conditions impact surface chemistry. (1) Low-power laser and the presence of atmospheric oxygen retains the MoS₂ chemical structure. (2) High-power laser and the presence of atmospheric oxygen results in molybdenum oxidation. (3) High-power laser in the presence of vacuum also retains MoS₂ chemical structure. i-XPS measurements on native and laser-treated MoS₂ nanosheets treated in ambient conditions. (B) Optical image of the area containing native and laser-treated nanosheets. (C) Processed image attributed to S_sulfide 2s signal. (D) Processed image attributed to Mo_ox 3d signal. (E) Composite map of masks associated with processed images of Mo_ox (green), S_sulfide (red), and SiO₂ (blue) signals. Yellow, teal, and pink reflect an overlap in masks, while black reflects an absence in masks. (F) i-XPS spectra from the raw images based on pixels from the ROIs defined in the composite map. The Mo 3d and S 2s regions (left) and the O 1s region (right). Reproduced with permission [104]. Copyright 2018, American Chemical Society.

  下载: 全尺寸图片 幻灯片
  

  Figure 9  Optical images of mechanically exfoliated MoS₂ nanosheets before (A-D) and after (E-H) thermal annealing in air for 1 h at 260 ℃ (E), 300 ℃ (F), 330 ℃ (G) and 400 ℃ (H). AFM height (I) and phase (J) images of MoS₂ nanosheet after thermal annealing for 10 h. AFM height (K) and phase (L) images of the MoS₂ nanosheet after thermal annealing for 15 h. (M) XPS spectrum of thermally annealed MoS₂ nanosheet. Reproduced with permission [106]. Copyright 2013, John Wiley & Sons, Ltd.

  下载: 全尺寸图片 幻灯片
  

  Figure 10  (A) The schematic of exfoliation mechanism of MoO₃/MoS₂. Reproduced with permission [111]. Copyright 2014, John Wiley & Sons, Ltd. (B) Schematic illustration of MoS₂ exfoliation and hybrid thin film preparation by centrifugation-coating. (C) Mo 3d spectra of MoS₂ oxidized by H₂O₂ (10%) for 15 min to light and to darkness. (D) Illustration of the hybrid MoO₃: MoS₂ synthetic process, and the chemicals structure of MoO₃, MoS₂ and MoO_xS_y. Reproduced with permission [112]. Copyright 2022, Multidisciplinary Digital Publishing Institute.

  下载: 全尺寸图片 幻灯片
  

  Figure 11  (A) Schematics of the oxidation process of 2D TiSe₂. Evolution of TiSe₂ thin flakes in the progress of oxidation under ambient condition. Raman mapping images of (B) as-made TiSe₂ flake, (C) partially oxidized TiSe₂ flake after exposure to air for 5 days and (D) completely oxidized TiSe₂ flake after exposure to air for 10 days, respectively. Insets: corresponding optical images of TiSe₂ flakes. Scale bars: 1 µm. AES mapping images of (E) O, (F) Ti and (G) Se elements on the heavily oxidized TiSe₂ flakes. Scale bars: 2 µm. XPS spectra for binding energies of (H) O 1s, (I) Ti 2p and (J) Se 3d on the surface of partially oxidized sample, respectively. (K) Typical Raman spectra of as-made, partially oxidized and completely oxidized TiSe₂ flakes (from top to bottom), respectively. Reproduced with permission [118]. Copyright 2017, John Wiley and Sons.

  下载: 全尺寸图片 幻灯片
  

  Figure 12  The prospective applications of TMD oxidized variants summarized as examples.

  下载: 全尺寸图片 幻灯片
•