Two-Dimensional MoS2: Structural Properties, Synthesis Methods, and Regulation Strategies toward Oxygen Reduction

Compared with three-dimensional (3D) and other materials, two-dimensional (2D) materials with unique properties such as high specific surface area, structurally adjustable band structure, and electromagnetic properties have attracted wide attention. In recent years, great progress has been made for 2D MoS2 in the field of electrocatalysis, and its exposed unsaturated edges are considered to be active sites of electrocatalytic reactions. In this review, we focus on the latest progress of 2D MoS2 in the oxygen reduction reaction (ORR) that has not received much attention. First, the basic properties of 2D MoS2 and its advantages in the ORR are introduced. Then, the synthesis methods of 2D MoS2 are summarized, and specific strategies for optimizing the performance of 2D MoS2 in ORRs, and the challenges and opportunities faced are discussed. Finally, the future of the 2D MoS2-based ORR catalysts is explored.


Introduction
Due to the high energy conversion efficiency and environmental friendliness, hydrogen fuel cells have been regarded as one of the most important clean energy conversion devices to solve the energy crisis and environmental pollution. However, the main challenge in hydrogen fuel cells is to find a highly efficient catalyst for the oxygen reduction reaction (ORR) [1]. The complex multi-step electron transfer process and the slow mass diffusion of the reaction itself lead to slow ORR kinetics [2]. Hitherto, platinum (Pt)-based catalysts are still considered as the most efficient ORR catalysts, but the high cost and poor durability severely limit their wide applications [3]. Therefore, these disadvantages have triggered intense interest in non-Pt catalysts with both high ORR activity and stability. Currently, a variety of non-platinum metal-based ORR catalysts, such as defective carbon materials [4], heteroatom-doped metal-free carbon [5], transition metal/heteroatomco-doped carbon [6,7], intermetallic compounds [8], and metal nitrides [9], have been synthesized. Among them, two-dimensional (2D) materials with atomic or molecular thickness and infinite planar length exhibit peculiar physicochemical properties, including high specific surface area, adjustable band structure, and electromagnetic properties, which make 2D materials become one of the most important non-Pt-based ORR catalysts [10][11][12][13].

Structural Properties of 2D MoS 2
The properties of 2D MoS 2 are highly dependent on their crystal phase, structure size and chemical composition. In the MoS 2 structure, the coordination of S elements and the stacking state between layers can form different electronic characteristics. The most common crystal structures of MoS 2 are 1T and 2H phases, where the number represents the number of layers of the S-Mo-S layer in a unit cell, and T and H represent the tetragonal crystal system, and the hexagonal crystal system, respectively [44]. As shown in Figure 1A, in the 2H phase, each Mo atom exists in the center of the hexagonal prism coordination structure and covalently bonded to six S atoms, and exhibits semiconductor properties. In the 1T phase, one Mo atom and six S atoms form a twisted octahedral coordination structure, with metallicity [45]. In MoS 2 , the S-Mo-S of each planar unit is regarded as a single layer, which is composed of two layers of S atoms and an intermediate layer of Mo atoms [46]. Due to the weak van der Waals force connection between the layers, the peeling between the layers is easy to occur, which makes it possible to synthesize single-layer MoS 2 [47].
As shown in Figure 1B, the bulk material of MoS 2 is an indirect-bandgap semiconductor, in which the top of the valence band is located at point G and the bottom of the conduction band is located at the midpoint of the axis of symmetry of G-K, while in singlelayer MoS 2 , the top of the valence band and the bottom of the conduction band coincide with point K, indicating that single-layer MoS 2 is a direct-bandgap semiconductor [48]. This indirect-to-direct transition of the bandgap from bulk to the single-layered 2D material is caused by the quantum confinement effect. The electronic properties of 2D MoS 2 are closely related to the d electrons existing in the d orbital between the bonding band and As shown in Figure 1B, the bulk material of MoS2 is an indirect-bandgap semiconductor, in which the top of the valence band is located at point G and the bottom of the conduction band is located at the midpoint of the axis of symmetry of G-K, while in single-layer MoS2, the top of the valence band and the bottom of the conduction band coincide with point K, indicating that single-layer MoS2 is a direct-bandgap semiconductor [48]. This indirect-to-direct transition of the bandgap from bulk to the single-layered 2D material is caused by the quantum confinement effect. The electronic properties of 2D MoS2 are closely related to the d electrons existing in the d orbital between the bonding band and the anti-bonding band of the Mo-S bond. In 2H-MoS2, the non-bonded d orbitals are fully occupied, resulting in semiconductor properties. In 1T-MoS2, the non-bonded d orbitals are partially filled, resulting in higher conductivity. As the d-orbital electron occupancy rate decreases, 2D MoS2 gradually changes from a semiconductor to a conductor, So far, many 2D MoS 2 -based materials have been applied to ORRs. At present, it is generally believed that the active site of 2D MoS 2 in the ORR is the unsaturated Mo site exposed at the edge. Because the Mo site is positively charged due to the possible polarization effect of the surrounding negatively charged S atoms, it can, therefore, be easily combined with negatively charged O atoms [51]. Simultaneously, the vacancy defects on the basal plane also contribute to ORR performance. In general, an excellent ORR electrocatalyst must have favorable stability, small overpotentials, high current densities, and low manufacturing cost [52]. Although MoS 2 has a low cost and high chemical stability, its large inert basal plane and poor electron transport properties limit its catalytic activity. However, it is undeniable that after adjusting the structure of 2D MoS 2 , it still possesses unique advantages and broad development prospects for ORRs. Here are some advantages of 2D MoS 2 in ORRs: (1) The uncoordinated metallic edge centers not only provide a certain intrinsic ORR activity but also give an opportunity for effective functionalization with different groups.
(2) Compared with bulk MoS 2 , nanostructured MoS 2 has a smaller particle size and a larger specific surface area. The small size 2D MoS 2 is more sensitive to the catalytic reactions compared with bulk MoS 2 . At the same time, the large surface area can anchor more active sites [53]. Besides, 2D MoS 2 has a unique open structure, which makes ORR reactants easily access the active site.
(3) 2D MoS 2 with a uniformly exposed lattice plane can supply a much simpler condition as an ideal platform for combing experiments and theoretical results, which is helpful to further understand the origin of ORR activity.

Liquid-Phase-Stripping Method
Liquid-phase stripping is a common method for preparing monolayer and multilayer materials, usually based on ultrasound assistance (in a solvent or aqueous solution) and subsequent centrifugal process [43,54]. As shown in Figure 2, this method is extremely compatible. Since the MoS 2 layers are connected by weak van der Waals forces, the 2D MoS 2 layer can be peeled from bulk by weakening the force between the layers. Recently, an ion intercalation method was created for layered materials that can attract guest molecules between layers and form clathrates. First, intercalation agents such as n-butyllithium [55,56] and iodine bromide are used to increase the interlayer spacing, weaken the interlayer force, and reduce the energy barrier for peeling. Then, the layered material is peeled off by ultrasonic waves [55], thermal shock [57], and other treatments. Ambrosi et al. obtained complete MoS 2 nanosheets under mild conditions by using a variety of Li-based intercalants and explored the effects of different intercalants on the material structure and catalysis performance. Among such Li-based intercalants (tert-butyllithium (t-Bu-Li), n-butyllithium (n-Bu-Li), and methyllithium (Me-Li)), the stripping efficiency of n-Bu-Li and t-Bu-Li is the highest, and the obtained MoS 2 nanosheets have higher catalytic performance [56]. However, this liquid-phase intercalation method is susceptible to changes in the reaction environment, causing structural deformation of some materials [58,59]. Another problem is the re-aggregation of the material during the removal of the intercalant.
To overcome the shortcomings of the intercalation method, such as time-consuming process and product instability, the latest stripping strategy is to place MoS 2 in a solvent with a specific surface energy and then use ultrasonic waves to destroy the material crystallites to produce nanosheets. It was discovered that organic solvents such as N-cyclohexyl-2-pyrrolidone (CHP) [60] and N-methyl-2-pyrrolidone (NMP) [43,61] can effectively strip MoS 2 . With further exploration, it was found that using surfactants such as sodium cholate [62,63] as a stabilizer can have a synergistic effect with the solvent and effectively prevent the nanosheets from re-aggregating ( Figure 3A). For the first time, Peter et al. employed a series of polymers (Polybutadiene-styrene (PBS), Polystyrene (PS), polyvinyl alcohol (PVA), Polyvinyl chloride (PVC), etc.) as stabilizers to replace surfactants, which proved that MoS 2 could be stripped and could remain stable in a mixed solution of a variety of organic solvents and polymers ( Figure 3B,C). More importantly, by a proposed simple model, they predicted that polymer stabilization is effective only when the solubility parameters of the nanosheets, polymer, and solvent are similar [64]. To overcome the shortcomings of the intercalation method, such as time-consuming process and product instability, the latest stripping strategy is to place MoS2 in a solvent with a specific surface energy and then use ultrasonic waves to destroy the material crystallites to produce nanosheets. It was discovered that organic solvents such as N-cyclohexyl-2-pyrrolidone (CHP) [60] and N-methyl-2-pyrrolidone (NMP) [43,61] can effectively strip MoS2. With further exploration, it was found that using surfactants such as sodium cholate [62,63] as a stabilizer can have a synergistic effect with the solvent and effectively prevent the nanosheets from re-aggregating ( Figure 3A). For the first time, Peter et al. employed a series of polymers (Polybutadiene-styrene (PBS), Polystyrene (PS), polyvinyl alcohol (PVA), Polyvinyl chloride (PVC), etc.) as stabilizers to replace surfactants, which proved that MoS2 could be stripped and could remain stable in a mixed solution of a variety of organic solvents and polymers ( Figure 3B,C). More importantly, by a proposed simple model, they predicted that polymer stabilization is effective only when the solubility parameters of the nanosheets, polymer, and solvent are similar [64].

Hydrothermal/Solvothermal Method
The hydrothermal/solvothermal method usually involves transferring the Mo and S precursor to an autoclave and heating at different temperatures. MoS2 nanosheets with different structures can be obtained by adjusting the reactant composition, reaction temperature, time, and pH value [65,66]. This method is easy to control and causes little environmental pollution. In addition, external forces such as microwaves [67,68] and strong

Hydrothermal/Solvothermal Method
The hydrothermal/solvothermal method usually involves transferring the Mo and S precursor to an autoclave and heating at different temperatures. MoS 2 nanosheets with different structures can be obtained by adjusting the reactant composition, reaction temperature, time, and pH value [65,66]. This method is easy to control and causes little environmental pollution. In addition, external forces such as microwaves [67,68] and strong magnetic fields [69] are added based on hydrothermal/solvothermal methods with the deepening of research, which accelerates the synthesis of materials and promotes their combination with other components. As shown in Figure 4A, Hao et al. applied the microwave-assisted hydrothermal synthesis method to obtain flower-structured MoS 2 nanosheets. At the same time, Ni ions were embedded on the MoS 2 nanosheets to form a unique self-supporting structure. This unique structure effectively inhibited the stacking of MoS 2 nanosheets [70].  The self-sacrificing template method is also employed to suppress the re-stacking of nanosheets. For example, taking g-C3N4 as a self-sacrificing template, Huang et al. first obtained a layered MoO3/g-C3N4 precursor with molybdate and g-C3N4 by hydrothermal synthesis and then processing at 900 °C. The g-C3N4 template was decomposed, and MoS2 nanosheets were finally obtained with a porous ultrathin structure ( Figure 4B) [46].

Chemical Vapor Deposition (CVD) Method
Chemical vapor deposition (CVD) is an important synthesis method for 2D materials. Compared with the solvent method and the hydrothermal synthesis method, large-size MoS2 nanosheets can be prepared [71]. However, for large-scale synthesis, MoS2 often forms a rod shape instead of a sheet shape [71][72][73][74], so high-vacuum reaction conditions are required and precious metal materials such as gold are used as the substrate [75][76][77]. The self-sacrificing template method is also employed to suppress the re-stacking of nanosheets. For example, taking g-C 3 N 4 as a self-sacrificing template, Huang et al. first obtained a layered MoO 3 /g-C 3 N 4 precursor with molybdate and g-C 3 N 4 by hydrothermal synthesis and then processing at 900 • C. The g-C 3 N 4 template was decomposed, and MoS 2 nanosheets were finally obtained with a porous ultrathin structure ( Figure 4B) [46].

Chemical Vapor Deposition (CVD) Method
Chemical vapor deposition (CVD) is an important synthesis method for 2D materials. Compared with the solvent method and the hydrothermal synthesis method, large-size MoS 2 nanosheets can be prepared [71]. However, for large-scale synthesis, MoS 2 often Micromachines 2021, 12, 240 7 of 22 forms a rod shape instead of a sheet shape [71][72][73][74], so high-vacuum reaction conditions are required and precious metal materials such as gold are used as the substrate [75][76][77]. The harsh reaction conditions limit the further application of the CVD method for the synthesis of 2D MoS 2 . To overcome such difficulties, simpler synthesis methods have been further explored. Lee et al. took MoO 3 and S powder as reactants to directly synthesize MoS 2 nanosheets on Si/SiO 2 substrates treated with graphene-like molecules ( Figure 5A). Under moderate temperatures, a large area of MoS 2 layers can be directly obtained on the surface of amorphous SiO 2 through the reaction of S vapor and MoO 3 . Without using highly crystalline metal substrates or an ultrahigh-vacuum environment, the cost was greatly reduced [71].
Due to the bottom-up synthesis principle of CVD, the continuous monolayer MoS 2 prepared by CVD is usually atomically neat, and its edges lack saturated atoms as main active sites [78], very unfavorable for the catalytic reaction. To improve this, Tan et al. first prepared porous gold as a template and then synthesized irregularly shaped MoS 2 nanosheets at the holes of the template ( Figure 5B). To further enhance the catalytic activity of MoS 2 nanosheets, a certain out-of-plane strain on the gold template was introduced so that MoS 2 was born on the curved inner surface of the template, which gave MoS 2 a certain degree of strain and improved its hydrogen evolution reaction (HER) catalytic performance [76]. However, this method is still too expensive for mass production; therefore, to reduce the cost, cheaper nanoporous metal templates (such as copper, nickel) could be considered.  Figure 5A). Under moderate temperatures, a large area of MoS2 layers can be directly obtained on the surface of amorphous SiO2 through the reaction of S vapor and MoO3. Without using highly crystalline metal substrates or an ultrahigh-vacuum environment, the cost was greatly reduced [71]. Due to the bottom-up synthesis principle of CVD, the continuous monolayer MoS2 prepared by CVD is usually atomically neat, and its edges lack saturated atoms as main active sites [78], very unfavorable for the catalytic reaction. To improve this, Tan et al. first prepared porous gold as a template and then synthesized irregularly shaped MoS2 nanosheets at the holes of the template ( Figure 5B). To further enhance the catalytic activity of MoS2 nanosheets, a certain out-of-plane strain on the gold template was introduced so that MoS2 was born on the curved inner surface of the template, which gave MoS2 a certain degree of strain and improved its hydrogen evolution reaction (HER) catalytic performance [76]. However, this method is still too expensive for mass production; therefore, to reduce the cost, cheaper nanoporous metal templates (such as copper, nickel) could be considered.

2D MoS 2 for ORRs
For the ORR, there are currently three mainstream reaction mechanisms: (1) dissociation mechanism, where the O-O bond of oxygen is directly broken when adsorbed on the catalyst surface and O* is reduced into OH* and H 2 O* in the next reaction; (2) association mechanism, where OOH* is first produced during the reaction and then split into O* and OH*; and (3) peroxygen mechanism, in which when the two-electron steps reach OOH* and HOOH*, the latter is decomposed into OH*. The formation of adsorbed oxidizing substances (e.g., O*, OH*, OOH*) caused by the aqueous electrolyte and the reaction itself is considered as a decisive process to determine the performance of the catalyst. All the above three mechanisms indicate that the ORR rate is actually limited by the formation of OOH* and the removal of OH*. The binding energy on the surface of the ORR intermediate should not be too strong or too weak [79]. This phenomenon can be represented by a volcano figure [80,81]. The catalyst located near the apex of the volcano ( Figure 6B) theoretically has the best catalytic performance.

2D MoS2 for ORRs
For the ORR, there are currently three mainstream reaction mechanisms: 1) dissociation mechanism, where the O-O bond of oxygen is directly broken when adsorbed on the catalyst surface and O* is reduced into OH* and H2O* in the next reaction; 2) association mechanism, where OOH* is first produced during the reaction and then split into O* and OH*; and 3) peroxygen mechanism, in which when the two-electron steps reach OOH* and HOOH*, the latter is decomposed into OH*. The formation of adsorbed oxidizing substances (e.g., O*, OH*, OOH*) caused by the aqueous electrolyte and the reaction itself is considered as a decisive process to determine the performance of the catalyst. All the above three mechanisms indicate that the ORR rate is actually limited by the formation of OOH* and the removal of OH*. The binding energy on the surface of the ORR intermediate should not be too strong or too weak [79]. This phenomenon can be represented by a volcano figure [80,81]. The catalyst located near the apex of the volcano ( Figure 6B) theoretically has the best catalytic performance. In 2D MoS2, when the size of the S-Mo-S three-layer flake is reduced laterally, the coordination number at the edges and corner atoms would become less, which provides rich active sites for electrocatalysis [82,83]. Through theoretical calculations, it can be found that edge-exposed unsaturated Mo atoms are the active sites of ORRs [84]. These unsaturated sites can effectively regulate the adsorption/desorption of reactants and intermediate products on the surface of the catalyst. Although it has been proved that MoS2 nanosheets exhibit a certain ORR performance, their reaction path is mostly a two-electron path. At the same time, the catalytic performance of 2D MoS2 is lower than that of the commercial Pt/C catalyst and even has a certain gap compared with the popular metalnitrogen-carbon (M-N-C) catalyst. Two main factors lead to the low catalytic performance of MoS2. The first is that pristine 2D MoS2 has low conductivity, which is not conducive to the electron transport in ORR processes and thus limits the reaction rate. The second is that, as a 2D material, although pristine 2D MoS2 has a large specific surface area, its active sites exist only at the edges and defects of the material, and the huge basal plane is inactive, which greatly hinders the reaction. Thus, how to properly adjust the structure of MoS2 to increase the number of active sites and enhance conductivity becomes a key issue in enhancing the catalytic activity of MoS2.

Non-metallic Element Doping
Studies have shown that incorporating non-metallic elements in MoS2 nanosheets can effectively regulate the ORR activity of catalysts due to the changed electronic structures around the doped atoms. It has been demonstrated that the introduction of heteroatoms (O, P, N, etc.) in MoS2 can adjust the electronic structure of materials, reduce the bandgap, In 2D MoS 2 , when the size of the S-Mo-S three-layer flake is reduced laterally, the coordination number at the edges and corner atoms would become less, which provides rich active sites for electrocatalysis [82,83]. Through theoretical calculations, it can be found that edge-exposed unsaturated Mo atoms are the active sites of ORRs [84]. These unsaturated sites can effectively regulate the adsorption/desorption of reactants and intermediate products on the surface of the catalyst. Although it has been proved that MoS 2 nanosheets exhibit a certain ORR performance, their reaction path is mostly a two-electron path. At the same time, the catalytic performance of 2D MoS 2 is lower than that of the commercial Pt/C catalyst and even has a certain gap compared with the popular metal-nitrogen-carbon (M-N-C) catalyst. Two main factors lead to the low catalytic performance of MoS 2 . The first is that pristine 2D MoS 2 has low conductivity, which is not conducive to the electron transport in ORR processes and thus limits the reaction rate. The second is that, as a 2D material, although pristine 2D MoS 2 has a large specific surface area, its active sites exist only at the edges and defects of the material, and the huge basal plane is inactive, which greatly hinders the reaction. Thus, how to properly adjust the structure of MoS 2 to increase the number of active sites and enhance conductivity becomes a key issue in enhancing the catalytic activity of MoS 2 .

Non-Metallic Element Doping
Studies have shown that incorporating non-metallic elements in MoS 2 nanosheets can effectively regulate the ORR activity of catalysts due to the changed electronic structures around the doped atoms. It has been demonstrated that the introduction of heteroatoms (O, P, N, etc.) in MoS 2 can adjust the electronic structure of materials, reduce the bandgap, improve conductivity, and provide catalytically active sites. As presented in Figure 7, Huang et al. used g-C 3 N 4 as a self-sacrificial template, and through hydrothermal and subsequent heat treatment, MoS 2 nanosheets with abundant edge defects were obtained, and then oxygen was introduced into MoS 2 through H 2 O 2 treatments. The sample with the best catalytic performance had an initial potential of 0.94 V and a half-wave electric potential of 0.80 V in 0.1M KOH. It is interesting to note that the ORR selectivity of O-MoS 2 changed from a two-electron path to a four-electron path. They believed that this change was due to the highly electronegative O atom polarizing the unsaturated adjacent Mo atom. The Mo atom at the edge generated an additional positive charge and was preferentially adsorbed by oxygen molecules, accelerating the ORR process [51]. Through density functional theory (DFT) calculations, Xie et al. also reported that compared with the original 2H-MoS 2 nanosheets (1.75 eV), oxygen-bound MoS 2 nanosheets have a (1.30 eV) narrower bandgap ( Figure 7D), indicating that the combination of oxygen and MoS 2 nanosheets may generate more carriers and increase the intrinsic conductivity of the material [85]. This leads to enhanced ORR and HER activity. improve conductivity, and provide catalytically active sites. As presented in Figure 7, Huang et al. used g-C3N4 as a self-sacrificial template, and through hydrothermal and subsequent heat treatment, MoS2 nanosheets with abundant edge defects were obtained, and then oxygen was introduced into MoS2 through H2O2 treatments. The sample with the best catalytic performance had an initial potential of 0.94 V and a half-wave electric potential of 0.80 V in 0.1M KOH. It is interesting to note that the ORR selectivity of O-MoS2 changed from a two-electron path to a four-electron path. They believed that this change was due to the highly electronegative O atom polarizing the unsaturated adjacent Mo atom. The Mo atom at the edge generated an additional positive charge and was preferentially adsorbed by oxygen molecules, accelerating the ORR process [51].  Figure 7D), indicating that the combination of oxygen and MoS2 nanosheets may generate more carriers and increase the intrinsic conductivity of the material [85]. This leads to enhanced ORR and HER activity. In addition, the incorporation of low-electronegativity atoms such as N and P into MoS2 would significantly enhance ORR catalytic activity. However, there are different explanations for performance enhancement. As shown in Figure 8A,B Huang et al. doped P atoms into MoS2 nanosheets, which effectively improved the ORR performance of the MoS2 nanosheets, and explained the enhancement of ORR performance using boundary molecular orbital theory. Compared to the original MoS2 nanosheets, the energy levels of In addition, the incorporation of low-electronegativity atoms such as N and P into MoS 2 would significantly enhance ORR catalytic activity. However, there are different explanations for performance enhancement. As shown in Figure 8A,B Huang et al. doped P atoms into MoS 2 nanosheets, which effectively improved the ORR performance of the MoS 2 nanosheets, and explained the enhancement of ORR performance using boundary molecular orbital theory. Compared to the original MoS 2 nanosheets, the energy levels of highest occupied molecular orbitals (HOMO) and lowest unoccupied molecular orbitals (LUMO) of P-doped MoS 2 nanosheets are higher [46]. In semiconductors, increasing the frontier orbital energy is beneficial to electron donation [86], which makes oxygen adsorption and the formation of the intermediate product OH − easier, accelerating the reaction rate.
Zhang et al. calculated the overpotential of N-or P-doped single-layer MoS 2 in the ORR under acidic conditions and found that replacing S atoms in the MoS 2 single layer with P or N atoms can introduce high spin density into the basal plane of MoS 2 , thereby improving its ability to activate O 2 and making the ORR step more inclined to proceed through the more efficient 4e − pathway [87]. It is worth noting that the ORR performance of P-doped MoS 2 is worse than that of N-doped MoS 2 due to the excessively high adsorption energy of the intermediate product of P-doped MoS 2 . However, this does not mean that P-doped MoS 2 has no application prospects. As shown in Figure 8C, through DFT calculations, Liu et al. investigated the ORR mechanism and key active sites of P-doped single-layer MoS 2 with different structures and the P content in an alkaline environment. It can be seen that double P-doped MoS 2 with a higher P content had better performance than single P-doped MoS 2 . Its catalytic active site is the S2 atom, which is adjacent to the two P atoms, and the two adjacent P atoms lead to a reduction in the charge of the S2 site and provide a stronger hydrogen bond, boosting the adsorption of H 2 O and OH − groups. Simultaneously, compared with single P-doped MoS 2 , double P-doped MoS 2 had an appropriate intermediate product adsorption energy ( Figure 8D). By DFT calculations, the optimal P doping amount of the catalyst for the ORR was 5.5%, close to that obtained in the experiment (4.7%) [88]. This work provides a reasonable explanation for the contribution of P elements to catalytic activity that has long puzzled researchers. ines 2021, 12, 240 12 of 23

Metal Element Doping
Metallic elements can also be added to MoS2 to strengthen ORR catalytic activity. It has been reported that doping Cu, Co, and other transition metals in the S vacancy of MoS2 can effectively enhance ORR catalytic activity. Xiao et al. calculated the theoretical ORR catalytic performance of single-layer MoS2 doped with Co/Ni. It is known that doping

Metal Element Doping
Metallic elements can also be added to MoS 2 to strengthen ORR catalytic activity. It has been reported that doping Cu, Co, and other transition metals in the S vacancy of MoS 2 can effectively enhance ORR catalytic activity. Xiao et al. calculated the theoretical ORR catalytic performance of single-layer MoS 2 doped with Co/Ni. It is known that doping metal atoms into the basal plane of single-layer MoS 2 can adjust the adsorption energy of the reaction intermediates on the catalyst surface, thereby increasing catalysis activity. In Co/Ni-doped single-layer MoS 2 , Co/MoS 2 was similar to the FeN 4 active site in the M-N-C catalyst, and Ni/MoS 2 was similar to the CoN 4 active site. Therefore, the Co/Ni-doped single-layer MoS 2 had higher ORR catalytic activity than the original MoS 2 [89]. In addition, Urbanova et al. and He each incorporated metal elements such as Fe, Mn, Ti, and V into MoS 2 and found a significant improvement in ORR activity ( Figure 9A,B) [90,91]. Although much work has been done on the direction of metal doping in MoS 2 , the choice of metal types is a simple try-out method. Using DFT, Wang et al. systematically explored the ORR performance of single-layer MoS 2 doped with various transition metals. The calculation result showed that transition metal atoms could be embedded in the S vacancy and could significantly change the electronic structure of the material. As shown in Figure 9C,D, they found that Cu-doped MoS 2 had the best theoretical binding strength for ORR intermediates and therefore had the best ORR catalytic activity, indicating its high potential as an efficient single-atom catalyst for ORRs [92]. This work may serve as a useful guidance for further developing metal-doped MoS 2 -based ORR catalysts.
Since metal-S bonds are regarded as reactive sites in metal-doped transition metal dichalcogenides (TMDs), increasing the density of metal atoms on the surface of MoS 2 as much as possible and reducing the metal aggregation during heating can be the key to improving the reactivity of the catalyst, but these are difficult to achieve. The metal loading on the surface of most metal-doped MoS 2 catalysts is very low and difficult to precisely control. Therefore, how to overcome these difficulties is crucial to enhance the ORR performance of metal-doped MoS 2 .  Figure 9A,B) [90,91]. Although much work has been done on the direction of metal doping in MoS2, the choice of metal types is a simple try-out method. Using DFT, Wang et al. systematically explored the ORR performance of single-layer MoS2 doped with various transition metals. The calculation result showed that transition metal atoms could be embedded in the S vacancy and could significantly change the electronic structure of the material. As shown in Figure  9C,D, they found that Cu-doped MoS2 had the best theoretical binding strength for ORR intermediates and therefore had the best ORR catalytic activity, indicating its high potential as an efficient single-atom catalyst for ORRs [92]. This work may serve as a useful guidance for further developing metal-doped MoS2-based ORR catalysts. Since metal-S bonds are regarded as reactive sites in metal-doped transition metal dichalcogenides (TMDs), increasing the density of metal atoms on the surface of MoS2 as much as possible and reducing the metal aggregation during heating can be the key to improving the reactivity of the catalyst, but these are difficult to achieve. The metal loading on the surface of most metal-doped MoS2 catalysts is very low and difficult to precisely control. Therefore, how to overcome these difficulties is crucial to enhance the ORR performance of metal-doped MoS2.

Activating the Basal Plane to Improve the Catalytic Performance of ORRs
To obtain higher ORR performance, many studies have been devoted to optimizing the composition and structure of 2D MoS 2 through post-processing methods. However, these methods will inevitably destroy the stability of the material and reduce conductivity [93,94]. In fact, the huge impact of the large-area inert basal plane of 2D MoS 2 on catalytic performance has been neglected for a long time. Thus, how to effectively activate the basal plane of 2D MoS 2 has become a research hotspot.

Strain Engineering Activates Inert Basal Plane
It has been found that introducing tensile or compressive stress into MoS 2 can affect the ORR performance. The ultrathin flexible structure of 2D MoS 2 makes it easier to generate and maintain strain. Through DFT calculations, Zhao et al. observed that tensile strain would shift the center of the p orbital of chalcogen on the surface of TMDs to move to the Fermi level. This shift can enhance the adsorption of materials for ORR intermediates, thereby effectively promoting the ORR process. At the same time, for 2D MoS 2 , the larger biaxial strain would cause a certain degree of lattice mismatch and then inevitably generate new vacancies on the basal plane. The newly generated basal surface vacancies cooperate with strain and further enhance its catalytic performance [95]. These results indicate that strain engineering is an effective way to adjust the ORR activity of 2D TMDs. Recently, a lot of methods have been developed for introducing strain into the material during synthesis, such as using a template with a certain curvature (nanospheres, porous curved metal) or transferring the material to a patterned substrate. For example, Li et al. transferred MoS 2 to SiO 2 nanocones by means of capillary pressure and added strain into MoS 2 ( Figure 10) [96]. However, the introduction of strain into 2D MoS 2 in the experiment to enhance its ORR performance is still not widely used.

Activating the Basal Plane to Improve the Catalytic Performance of ORRs
To obtain higher ORR performance, many studies have been devoted to optimizing the composition and structure of 2D MoS2 through post-processing methods. However, these methods will inevitably destroy the stability of the material and reduce conductivity [93,94]. In fact, the huge impact of the large-area inert basal plane of 2D MoS2 on catalytic performance has been neglected for a long time. Thus, how to effectively activate the basal plane of 2D MoS2 has become a research hotspot.

Strain Engineering Activates Inert Basal Plane
It has been found that introducing tensile or compressive stress into MoS2 can affect the ORR performance. The ultrathin flexible structure of 2D MoS2 makes it easier to generate and maintain strain. Through DFT calculations, Zhao et al. observed that tensile strain would shift the center of the p orbital of chalcogen on the surface of TMDs to move to the Fermi level. This shift can enhance the adsorption of materials for ORR intermediates, thereby effectively promoting the ORR process. At the same time, for 2D MoS2, the larger biaxial strain would cause a certain degree of lattice mismatch and then inevitably generate new vacancies on the basal plane. The newly generated basal surface vacancies cooperate with strain and further enhance its catalytic performance [95]. These results indicate that strain engineering is an effective way to adjust the ORR activity of 2D TMDs. Recently, a lot of methods have been developed for introducing strain into the material during synthesis, such as using a template with a certain curvature (nanospheres, porous curved metal) or transferring the material to a patterned substrate. For example, Li et al. transferred MoS2 to SiO2 nanocones by means of capillary pressure and added strain into MoS2 ( Figure 10) [96]. However, the introduction of strain into 2D MoS2 in the experiment to enhance its ORR performance is still not widely used.

Stable Phase Change Activated Inert Basal Plane
Two-dimensional MoS2 has different phases, causing various catalytic activities. The 1T phase is a metal phase and has the highest conductivity. However, it is metastable at room temperature and would slowly change into the 2H phase spontaneously. Improving

Stable Phase Change Activated Inert Basal Plane
Two-dimensional MoS 2 has different phases, causing various catalytic activities. The 1T phase is a metal phase and has the highest conductivity. However, it is metastable at room temperature and would slowly change into the 2H phase spontaneously. Improving the performance of 2D MoS 2 by controlling the phase transition has broad research prospects. According to previous reports, the 1T phase can be obtained in a single layer of MoS 2 through different methods such as alkali metal intercalation-exfoliation and plasma hot-electron injection. In addition, the 1T phase can stably exist by doping Re, Tc, Mn, and other metal elements in nanosheets as electron donors [97][98][99][100]. The high electrical conductivity and catalytic activity make 1T-MoS 2 a more effective electrocatalyst than 2H-MoS 2 . As exhibited in Figure 11, Sadighi prepared 1T-MoS 2 nanosheets via the in situ liquid-redox intercalations of sodium ions and solvent exchange exfoliation. The 1T-MoS 2 nanosheets showed an almost equivalent half-wave potential and enhanced the limit diffusion current as compared to Pt/C [101]. The 1T-phase MoS 2 nanosheets prepared by the alkali metal intercalation method can spontaneously transform from the 1T phase to the thermodynamically stable 2H phase in the air [59]. This unstable property hinders catalytic stability. At the same time, the conditions of this preparation method are relatively harsh, and the 1T-phase/2H-phase ratio is not high. As a result, it is necessary to develop methods that can easily synthesize 1T-phase MoS 2 with high stability. the performance of 2D MoS2 by controlling the phase transition has broad research prospects. According to previous reports, the 1T phase can be obtained in a single layer of MoS2 through different methods such as alkali metal intercalation-exfoliation and plasma hot-electron injection. In addition, the 1T phase can stably exist by doping Re, Tc, Mn, and other metal elements in nanosheets as electron donors [97][98][99][100]. The high electrical conductivity and catalytic activity make 1T-MoS2 a more effective electrocatalyst than 2H-MoS2. As exhibited in Figure 11, Sadighi prepared 1T-MoS2 nanosheets via the in situ liquid-redox intercalations of sodium ions and solvent exchange exfoliation. The 1T-MoS2 nanosheets showed an almost equivalent half-wave potential and enhanced the limit diffusion current as compared to Pt/C [101]. The 1T-phase MoS2 nanosheets prepared by the alkali metal intercalation method can spontaneously transform from the 1T phase to the thermodynamically stable 2H phase in the air [59]. This unstable property hinders catalytic stability. At the same time, the conditions of this preparation method are relatively harsh, and the 1T-phase/2H-phase ratio is not high. As a result, it is necessary to develop methods that can easily synthesize 1T-phase MoS2 with high stability.

Compounding Other Materials to Enhance Conductivity
It is well known that the low conductivity and aggregation of nanocrystalline sheets can reduce the catalytic activity of 2D MoS2. To avoid these cases, the combination of 2D MoS2 with materials possessing high conductivity to form a composite has attracted much attention. Among them, compounding 2D MoS2 with carbon materials to form a hybrid structure can effectively boost the conductivity of materials and reduce aggregation [102,103]. As presented in Figure 12A,B, by employing 4-iodophenyl-functionalized monolayer MoS2 as a template, Yuan et al. synthesized a series of MoS2-coupled sandwich-like conjugated microporous polymers (M-CMPs). After direct pyrolysis, the 2D MoS2/Ndoped porous carbon (M-CMPs-T) hybrid with a large specific surface area and layered porous structure was finally obtained. Compared with porous nitrogen-doped carbon without MoS2, 2D porous carbon hybrids exhibited a higher half-wave potential and limit diffusion current in the ORR [104].
Zhao et al. synthesized hybrid nanorods using dopamine and ammonium through self-assembly followed by annealing, while an ultrathin monolayer of MoS2 was formed under the constraints of the carbon skeleton ( Figure 12C,D). The unique tubular structure, large specific surface area, relatively high N dopant content, and oxygen-bonded MoS2 monolayer endowed the MoS2/ nitrogen-doped carbon hybrid nanotubes (C HNT) with high electronic conductivity and abundant ORR active sites. The as-prepared MoS2/C HNT possessed ORR activity close to commercial Pt/C under alkaline conditions [105].

Compounding Other Materials to Enhance Conductivity
It is well known that the low conductivity and aggregation of nanocrystalline sheets can reduce the catalytic activity of 2D MoS 2 . To avoid these cases, the combination of 2D MoS 2 with materials possessing high conductivity to form a composite has attracted much attention. Among them, compounding 2D MoS 2 with carbon materials to form a hybrid structure can effectively boost the conductivity of materials and reduce aggregation [102,103]. As presented in Figure 12A,B, by employing 4-iodophenylfunctionalized monolayer MoS 2 as a template, Yuan et al. synthesized a series of MoS 2coupled sandwich-like conjugated microporous polymers (M-CMPs). After direct pyrolysis, the 2D MoS 2 /N-doped porous carbon (M-CMPs-T) hybrid with a large specific surface area and layered porous structure was finally obtained. Compared with porous nitrogen-doped carbon without MoS 2 , 2D porous carbon hybrids exhibited a higher half-wave potential and limit diffusion current in the ORR [104].
Zhao et al. synthesized hybrid nanorods using dopamine and ammonium through self-assembly followed by annealing, while an ultrathin monolayer of MoS 2 was formed under the constraints of the carbon skeleton ( Figure 12C,D). The unique tubular structure, large specific surface area, relatively high N dopant content, and oxygen-bonded MoS 2 monolayer endowed the MoS 2 / nitrogen-doped carbon hybrid nanotubes (C HNT) with high electronic conductivity and abundant ORR active sites. The as-prepared MoS 2 /C HNT possessed ORR activity close to commercial Pt/C under alkaline conditions [105]. Zhou et al. used hydrothermal and annealing treatment to attach MoS 2 nanosheets to graphene oxide (GO), and the interconnected MoS 2 nanosheets formed a layered structure that completely covered the GO surface. The material revealed excellent ORR performance due to the enhanced conductivity of MoS 2 . At the same time, graphene oxide effectively limited the aggregation of MoS 2 nanosheets [106] (Figure 12E,F). phene oxide (GO), and the interconnected MoS2 nanosheets formed a layered structure that completely covered the GO surface. The material revealed excellent ORR performance due to the enhanced conductivity of MoS2. At the same time, graphene oxide effectively limited the aggregation of MoS2 nanosheets [106] (Figure 12E,F).

2D MoS 2 as a Co-Catalyst
Although 2D MoS 2 can be directly used as an ORR catalyst, its performance is mostly unsatisfactory due to poor conductivity and rare active sites. Thus, to achieve higher catalytic performance, 2D MoS 2 can be used as a co-catalyst through having a synergistic effect with other active substances. As shown in Figure 13A-C, Mao et al. used g-C 3 N 4 as a self-sacrificial template to synthesize ultrathin Ni 3 S 2 /MoS 2 nanosheets with rich heterogeneous interfaces. The heterostructure formed on the two-phase surface of Ni 3 S 2 and MoS 2 generated a large number of unique Mo-Ni-S catalytic active sites. The catalyst revealed higher ORR catalytic activity than the commercial Pt/C [107]. Kwon et al. synthesized an iron phthalocyanine (FePc) and 1T-phase MoS 2 nanoplate heterostructure catalyst through a one-step hydrothermal reaction. Herein, iron phthalocyanine promoted the transition of nanosheets from the 2H phase to the 1T phase, increasing the electrical conductivity. The unique non-planar geometry of the Fe-N 4 active site of FePc further enhanced the ORR catalytic activity of FePc-MoS 2 ( Figure 13D,E) [108]. Our group grew 2D MoS 2 on a N-doped carbon skeleton derived from zeolitic imidazolate frameworks-8 (ZIF-8) and constructed an excellent multifunctional Mo/N/C@MoS 2 catalyst toward HER, ORR and oxygen evolution reaction (OER). Its excellent catalytic performance originated from the synergistic effect between the abundant main active sites on the edges of MoS 2 , the Mo-N phase coupling center, and the N-induced active sites of nanocarbon frameworks. The porous 3D carbon skeleton structure also accelerated the diffusion rate of products and enhanced electrical conductivity [109]. a self-sacrificial template to synthesize ultrathin Ni3S2/MoS2 nanosheets with rich hetero geneous interfaces. The heterostructure formed on the two-phase surface of Ni3S2 an MoS2 generated a large number of unique Mo-Ni-S catalytic active sites. The catalyst re vealed higher ORR catalytic activity than the commercial Pt/C [107]. Kwon et al. synth sized an iron phthalocyanine (FePc) and 1T-phase MoS2 nanoplate heterostructure cataly through a one-step hydrothermal reaction. Herein, iron phthalocyanine promoted th transition of nanosheets from the 2H phase to the 1T phase, increasing the electrical con ductivity. The unique non-planar geometry of the Fe-N4 active site of FePc further en hanced the ORR catalytic activity of FePc-MoS2 ( Figure 13D,E) [108]. Our group grew 2 MoS2 on a N-doped carbon skeleton derived from zeolitic imidazolate frameworks-8 (ZIF 8) and constructed an excellent multifunctional Mo/N/C@MoS2 catalyst toward HER, OR and oxygen evolution reaction (OER). Its excellent catalytic performance originated from the synergistic effect between the abundant main active sites on the edges of MoS2, th Mo-N phase coupling center, and the N-induced active sites of nanocarbon framework The porous 3D carbon skeleton structure also accelerated the diffusion rate of produc and enhanced electrical conductivity [109]. Based on a strategy derived from ZIF-67, Li et al. synthesized a core-double-shell Co 9 S 8 @Co 9 S 8 @MoS 2 heterostructure. The Co@Co 9 S 8 precursor provided sufficient space to limit the growth of MoS 2 , which prevented the aggregation of MoS 2 and improved the electrical conductivity of the material. The core-double-shell Co 9 S 8 @Co 9 S 8 @MoS 2 heterostructure exhibited a synergistic interaction between Co 9 S 8 and MoS 2 and modified electronic structures. Such characteristics made Co 9 S 8 @Co 9 S 8 @MoS 2 exhibit excellent electrocatalytic performance, close to commercial Pt/C [111]. As shown in Figure 13F, Vattikuti et al. deposited silver nanocrystals on the surface of 2D MoS 2 nanosheets by a simple microwave-assisted method. Due to the Ag/MoS 2 heterostructure and the increased specific surface area, the catalyst presented better ORR activity than pristine MoS 2 in alkaline solutions [110].
When used as a co-catalyst, 2D MoS 2 can provide a large-area substrate supporting other active materials (such as precious metals Pt, Ag) and reduce the dosage of active materials. In addition, the heterostructure formed by MoS 2 and other materials can provide good catalytic activity. Although a variety of MoS 2 -based heterostructure catalysts with excellent ORR catalytic performance have been prepared, the exact reaction that occurs on the heterostructure and the internal mechanism of improving the catalytic performance are still not determined. At the same time, the tunable, controllable, and large-scale synthesis of heterostructures is hard to achieve.

Outlook
As a kind of electrocatalytic material with great potential, 2D MoS 2 has shown great application prospects on the HER, but its development in ORRs is stagnant. Most of the 2D MoS 2 materials possess low-electron-transport characteristics and a large area of the unused inert basal plane, which limits their applications. This review summarized the research progress in 2D MoS 2 in ORRs in recent years; focused on the specific works on 2D MoS 2 , from synthesis methods to ORR performance optimization; and identified the key obstacles that 2D MoS 2 faces in ORRs. To promote the ORR performance of such materials, electron transport and the increase in active sites are basic problems to resolve. Some strategies for resolving these problems, such as increasing exposed edges, heteroatom doping, constructing heterostructures, designing composite materials, and stable phase transitions to enhance electron and product transport, etc., were also discussed in this article. These strategies make it possible to develop high-performance 2D MoS 2 -based ORR catalysts. However, despite numerous encouraging outcomes, there are still many issues to be resolved.
(1) Controllable synthesis. The current bottom-up synthesis method requires relatively harsh conditions, and the top-down synthesis method would cause the aggregation of nanosheets, which lowers the quality of 2D MoS 2 . Striving to achieve large-scale synthesis of high-quality 2D MoS 2 materials is still the basic requirement.
(2) At present, there is no consensus on the detailed mechanism of 2D MoS 2 and 2D MoS 2 -based catalysts in the ORR, especially the specific contributions of elements such as O, N, and P. Thus, it is necessary to further analyze the relationship between structure composition and performance to guide future development.
(3) Effective usage of the large-area basal plane with reactive inertness is necessary. The current work mainly focuses on heteroatom doping and material recombination. However, fine control of heteroatom doping has not been well performed. Further exploring the synergy between MoS 2 and other ORR active materials, and the internal mechanism of enhanced ORR activity, is the focus for future research.
(4) The intrinsic properties of MoS 2 can be activated by a stable phase change and strain engineering. Thus, we should find more convenient methods to synthesize long-term stable 1T-phase MoS 2 nanosheets to improve the adverse effects of spontaneous phase transition on the catalytic performance.