Plasma Processing and Treatment of 2D Transition Metal Dichalcogenides: Tuning Properties and Defect Engineering Authors: Saeed Sovizi, Shayan Angizi, Sayed Ali Ahmad Alem, Reyhaneh Goodarzi, Mohammad Reza Rahmani Taji Boyuk, Hajar Ghanbari, Robert Szoszkiewicz, Abdolreza Simchi, Peter Kruse Published: 2023 ABSTRACT Two-dimensional transition metal dichalcogenides (TMDs) offer fascinating opportunities for fundamental nanoscale science and various technological applications. They are a promising platform for next generation optoelectronics and energy harvesting devices due to their exceptional characteristics at the nanoscale, such as tunable bandgap and strong light-matter interactions. The performance of TMD-based devices is mainly governed by the structure, composition, size, defects, and the state of their interfaces. Many properties of TMDs are influenced by the method of synthesis so numerous studies have focused on processing high-quality TMDs with controlled physicochemical properties. Plasma-based methods are cost-effective, well controllable, and scalable techniques that have recently attracted researchers’ interest in the synthesis and modification of 2D TMDs. TMDs’ reactivity toward plasma offers numerous opportunities to modify the surface of TMDs, including functionalization, defect engineering, doping, oxidation, phase engineering, etching, healing, morphological changes, and altering the surface energy. Here we comprehensively review all roles of plasma in the realm of TMDs. The fundamental science behind plasma processing and modification of TMDs and their applications in different fields are presented and discussed. Future perspectives and challenges are highlighted to demonstrate the prominence of TMDs and the importance of surface engineering in next-generation optoelectronic applications. Introduction Within the last decades, nanotechnology revolutionized key industries such as transportation,1,2 biomedical engineering,3 energy storage,4 electronics,5 and optoelectronics.6 The experimental exfoliation of single graphite layers in 20047 was followed by an unprecedented wave of scientific work on alternative two-dimensional (2D) materials such as hexagonal boron nitride (h-BN),8,9 boron carbon nitrides (BCN),10 transition metal dichalcogenides (TMDs),11 transition metal oxides including titania- and perovskite-based oxides,12 2D monoelementary materials (Xenes),13 and 2D transition metal carbides, nitrides, and carbonitrides (MXenes).14Figure 1 gives an overview of the 2D materials families of interest. In most of these materials, covalently bonded atoms within each layer interact only weakly with consecutive layers via van-der-Waals interactions,15−17 facilitating their exfoliation. 2D materials can be categorized in different ways based on their structure or properties. With regard to electrical properties, graphene monolayers have a zero bandgap18 while h-BN has a large bandgap of ∼5.9 eV.9,19 Several 2D TMDs have a bandgap of 1 to 2 electron volts (eV), comparable to conventional semiconducting materials such as Si.20 For decades, bulk TMDs and graphite have been used as solid lubricants21 and as electrodes in some batteries22 and photovoltaic devices.23 Nevertheless, recent advances in the fabrication of graphene and few-layered TMDs enable applications of 2D TMDs in next-generation electrical, photonic, optoelectrical, and energy-harvesting devices.24 Transition Metal Dichalcogenides The general formula for the TMD family is MX2, where M represents a transition metal from groups 4–10 of the periodic table (such as Mo, W, Nb), while X refers to a chalcogen from group 16 of the periodic table (such as S, Se, or Te).71−73 The height of each X-M-X layer is roughly 0.6–0.7 nm.74 By combining different M and X elements, 40 members of this family have been explored so far.74 TMDs are naturally found or artificially synthesized in different crystal structures with unique properties. Among 11 polymorphs,75 TMDs are usually found in three crystal structures of 1T (AbC AbC stacking), 2H (AbA BaB stacking), and 3R (ABA CaC BcB stacking) (Figure 2). The digit shows the number of layers in the unit cell, and the letter stands for the symmetry group: T: trigonal, H: hexagonal, and R: rhombohedral. MoS2 is the most studied member of the TMD family. Its 2H phase is generally regarded as thermodynamically favorable under ambient conditions, featuring trigonal prismatic MoS6 centers. In contrast, its 3R structure is metastable due to its higher formation energy and may undergo undesired side reactions during its degradation. The 1T phase of MoS2, however, is entirely synthetic, with metal atoms being octahedrally coordinated in two sets of degenerate d orbitals known as eg (dz, x2-y2) and t2g (dxy, yz, zx). Each t2g orbital is occupied by one electron, resulting in partially filled bands and conferring the conductive nature of the 1T phase. The 2H polytype of MoS2 exhibits many properties of an n-type semiconductor and demonstrates relatively nonmagnetic semiconducting characteristics due to spin pairing of its 4d electrons.76 Its 3R polytype, belonging to the R3m space group, has a tunable bandgap and a high piezoelectric constant. Compared to the 1T and 3R polytypes, the 2H polytype has attracted significant interest due to its structural tunability and remarkable stability across different synthesis conditions and methods. The synthesis of 1T configurations often encounters challenges related to the instability of the 1T phase and/or undesired coexistence of both 2H and 1T phases within TMD structures.77 As a result, plasma-based methods have emerged as technologically viable approaches that enable researchers to precisely control these interactions, facilitating the development of various TMD polytypes, including the 1T phase, for desired applications.78 Depending on the oxidation state and coordination number of the metallic element, TMDs show various electrical characteristics, for example, HfS2 and 2H-VSe2 are insulators, 2H-MoS2 and 2H-WS2 are semiconductors, and 1T-MoS2, 1T-WS2, 1T-VSe2, and 2H-NbSe2 are conductors.74,79−83 Chemical properties of TMDs also depend on their structures. For instance, 2H-MoS2 has chemically active edges, while the basal plane is chemically inert. In 1T-MoS2, however, it is known that sulfur atoms are active sites in both the edges and basal planes.84 The size-dependent transition of indirect (multilayer)-to-direct (monolayer crystals) bandgap is the most attractive characteristic of TMDs for optoelectronic applications.85 For instance, single layer MoS2 has a direct bandgap of 1.8 eV, strong photoluminescence, acceptable mobility of 700 cm2 V–1 s–1, and a high on/off current ratio of 107−8.86 The small thickness and high surface-to-volume ratio87 as well as mechanical flexibility and lightweight88,89 offer unique properties that make TMDs very attractive for next generation flexible optoelectronic devices such as solar cells and photodetectors.89,90 Moreover, the van der Waals gaps between the TMD layers allow for intercalation and make them attractive for energy storage applications.86 Nevertheless, some critical issues yet need to be addressed, i.e., mass production of large and high-quality single-layer TMDs with large lateral sizes and tailored electronic properties via controlled charge-carrier type and doping level.74,91,92 Surface Modification of TMDs Exploiting modification processes to control the nanoscale properties is pivotal to fabricating functional building blocks for novel applications.93 Over the past decade, particular attention has been paid to the surface modification of TMDs to tune their properties, harnessing the maximum capabilities and broadening the applications.94−96 Apart from economic aspects, the consistency and reproducibility of modification techniques are of paramount importance that may challenge further development. Thus far, various methods including chemical treatments in solution,97,98 electron and ion irradiation,99−101 laser irradiation,102,103 annealing,104−106 ion implantation,107 electrostatic gating,108 and ultraviolet (UV) treatment109 have been employed. For instance, researchers have successfully tuned the electrical properties of MoS2 utilizing He+ ion beams.110 It was shown that the introduced defects could be controlled by altering the ion dose and the beam profile in this process. This resulted in obtaining either insulating, semiconducting, or conducting structures by only adjusting the ion doses. Plasma as a Synthesis and Modification Method Plasma treatment is a powerful modification tool that has recently attracted the attention of researchers for the modification of TMDs. Thanks to their unique capabilities, plasma-based modification methods have gone from a simple deposition methods to a class of approaches to synthesize and manipulate complex nanomaterials and nanostructures.111 The reactive nature of plasma is employed to synthesize and modify a broad range of materials, even with inert chemical surfaces such as ceramics.112 The physical and chemical interactions of plasma components with materials can tune their electrical, chemical, optical, tribological, magnetic, and even mechanical properties.113,114 Such interactions are mostly governed by the plasma gas. Generally, the nature of ionic and neutral components in the plasma significantly depends on the feed gas. For instance, O2 plasma contains negative and positive ions (such as O–, O+, O2+), electrons, free radical molecules, ozone, and UV photons. In contrast, UV photons and byproducts of ozone decomposition are common in UV/O3 treatments.112 Reliability, reproducibility, cost-effectiveness, applicability to different sample geometries and masking techniques as well as scaling-up capabilities for industrial processes have positioned the plasma treatments as widely desired methods for the surface modification of 2D materials.114Figure 3 schematically represents various roles of plasma in the realm of TMDs. Chemical differentiation and stability of 2D materials. 2D materials shown on a green background are stable under ambient environment for at least several weeks. A red background represents 2D materials that are only stable under an inert atmosphere (reactive with air). Orange-shaded 2D materials react with oxygen or water slowly in the case of air exposure and remain intact for only a few hours. Grayed 2D materials show varied behaviors. Some of them can remain intact under ambient environment, while some of them may quickly react with air. The stability conditions of gray-shaded 2D materials are not yet fully explored. Data extracted from refs (12, 14, and 25−70). Schematic 2D and 3D views of common lattice structures and unit cells of TMDs. Schematic illustration of various plasma processes for the fabrication and modification of 2D TMDs. Selected recent advances in modification of 2D TMDs and their optoelectronic properties as well as relevant applications were discussed in a recent mini-review.115 However, plasma processes are not only able to modify the TMDs in a wide variety of ways but also to synthesize such materials in a controlled way with different structures. Due to such capabilities, there is a growing trend for plasma-based techniques to be used as a modification or as synthesis methods for TMDs. Here we present and discuss all aspects of plasma in the context of TMD synthesis and processing in significant detail. At first, the chemistry of plasma is introduced. Then, plasma-based methods developed for the synthesis of TMDs are presented. Following, the science and technology aspects of various modification techniques are discussed, including oxidation with atomic and molecular forms of oxygen, doping, defect engineering, and functionalization via physical adsorption and chemical reactions, as well as etching. New physicochemical properties and applications arising from such modifications are also addressed. Finally, various plasma techniques which have been used for the modification of substrates for the fabrication of TMD-devices are introduced. Since ultraviolet (UV) light and ozone are two common components of oxygen-containing plasma and UV/O3, UV/O3 treatments for modification of TMDs are also considered in this review. Data on proper plasma parameters for specific synthesis and modification treatment of TMDs and the resulting device performance have been summarized in several easy-to-read tables as a reference tool for current and future practitioners of plasma processing. Chemistry of Plasma Plasma was first discovered in the laboratory by Sir William Crookes in the 19th century.116 However, the term “plasma” was introduced much later by Irving Langmuir in 1928117 while working on electric discharges. Plasma is an ionized gas containing substantial amounts of charged particles, including positive and negative ions as well as free electrons.118 A plasma can be artificially generated from gas molecules via various methods such as dielectric barrier discharge (DBD), radio frequency discharge (RFD), and microwave discharge.119,120 At sufficient energy levels, collisions between gas molecules produce free electrons, which are accelerated in the electromagnetic field, resulting in a cascade of electrons and ions.120 There are two types of processes, based on their temperature: (i) high-temperature fully ionized plasma and (ii) low-temperature plasma. The latter can be further classified as thermal or nonthermal. Arcs, plasma torches, and plasma jets are examples of thermal plasma, in which heavy particles (ions and neutral components) have a similar temperature to electrons (103-104 K), i.e., they are in thermal equilibrium. Nonthermal plasmas such as DBD or RFD consist of ionized and neutral gases with temperatures as low as room temperature and electrons at temperatures up to 104 K.120−122 Nonthermal plasmas have two advantages over their thermal counterparts: (i) Since the electron temperature is as high as 104-105 K (1–10 eV), chemical reactions with high activation energies can proceed. For instance, effective synthesis of ozone can only be performed by nonthermal plasma procedures. (ii) The gas temperature in nonthermal plasma is low which removes the need for quenching, minimizes thermal stresses on the plasma chamber and treated interfaces, and reduces possible defects which may be introduced to the treated substrates by thermal plasma processing. That is why nonthermal processes such as DBD and RFD are often used for surface modifications.122,123 The chemistry of plasma is quite complex and includes various single reaction steps. Plasma reactions can be classified into homogeneous and heterogeneous reactions.122 Homogeneous reactions occur in the gas phase due to inelastic collisions between plasma species (Table 1), while heterogeneous reactions happen at the interface between the solid exposed to the plasma and plasma species. Heterogeneous reactions mainly result in erosion of the surface through physical and/or chemical reactions, deposition of plasma species on the surface, a combination of plasma species and surface atoms or already-adsorbed compounds on the surface, and substitution of surface atoms.122 Table 1Different Elementary Reactions in Plasma Gas. Data from ref (123)Involving ElectronsInvolving Heavy Particlese– + X → X* + e–Excitation of atoms and moleculesX+ + Y → X + Y+Asymmetric and symmetric charge transferse– + XY → XY* + e–X+ + X → X + X+e– + X* → X** + e–Stepwise excitationX* + Y → X + Y+ + e–(Associative) Penning ionizationX* + Y → XY+ + e–X* → X + hvSpontaneous de-excitationX* + X* → X + X+ + e–Pair ionizationXY* → XY + hve– + X* → Y + hv + e–Collisional induced de-excitationX* + X → X2+ + e–Hornbeck-Molnar ionizatione– + XY* → XY + hv + e–e– + X → X+ + 2e–IonizationZ+ + XY → ZY+ + XIon–molecule reactione– + XY → XY+ + 2e–e– + XY → XY–e– + X* → X+ + e– + e–Stepwise ionizationX+ + Y– → XYIon–ion recombinatione– + X → X–AttachmentX– + Y → X + Y + e–Collisional detachmente– + X+ → XRecombinationX– + Y → XY + e–Associative detachmente– + XY+ → XYe– + X+ + Z → X + ZThree-body recombinationX* + Y → X + Y*Excitation transfere– + XY+ + Z → XY + Ze– + XY → X(*) + Y + e–Dissociation of molecules with or without excitationXY* + Z → X + Y + ZDissociatione– + XY → X(*) + Y+ + 2e–Dissociative ionization with or without excitationZ(*) + XY → ZY + XChemical reactionsXY + VZ → XV +YZXY + VZ → XYV +Ze– + XY → X– + Y+ + e–Ion-pair generationX + hv → X*Photoexcitation and ionizationX + hv → X+ + e–e– + XY → X– + YDissociative attachmentXY + hv → X + YPhotodissociatione– + XY– → XY + 2e–DetachmentX* + hv → X + hv + hvPhotoinduced emission Different Elementary Reactions in Plasma Gas. Data from ref (123) Nonthermal oxygen or dry-air plasmas are commonly used for surface modifications. Such plasmas modify the surface chemistry by generating reactive atoms, excited atoms, and molecules as well as negative and positive ions (such as O2–, O–, O+, O2+), through dissociative collisions of molecules. Negative ions are important for electronegative gases like oxygen. These ions are formed by dissociative attachment of an electron to an oxygen molecule. Ozone is also another important product of such plasmas. It is a potent oxidizing agent used for various applications such as water treatment and surface modifications. Ozone is formed through three-body collision processes (eq 1):1231in which M is the third collision partner such as N2, O, O2, or even O3, and oxygen atoms are formed through dissociation of O2 by electron impacts. There are also some competitive reactions which reduce the ozone formation such as formation of O2 by recombination of two oxygen atom or combination of one oxygen atom and ozone molecule.124 In order to increase the ozone formation efficiency, a relatively high pressure and low temperature plasma can be utilized. Operating at low temperatures enhances ozone formation due to the fast decay of ozone at high temperatures. High pressure is also favorable for ozone formation through the three-body reaction in eq 1. That is why DBD, which can operate at atmospheric pressure and a low temperature, is preferred among various plasma methods used for ozone generation so far. Moreover, a low concentration of atomic oxygen enhances the ozone formation since eq 1 (by considering M to be O2) is first order in atomic oxygen concentration while the reaction rate for recombination of oxygen atoms is second order with respect to atomic oxygen concentration. As a result, decreasing input energy to achieve a lower O/O2 ratio increases ozone content. On the other hand, since energy density and consequently the dissociation degree are too low in a weak discharge, a considerable part of the input energy (up to 50%) can be dissipated by ions since they do not significantly participate in the ozone formation. Therefore, the input power should be optimized for maximum ozone formation. Using fine mesh metal electrodes in DBD can also increase the ozone formation efficiency by up to 20% in comparison to conventional DBD.124 Another simple method is using a fine-grained material to fill the discharge gap and increase the efficiency through catalysis. Nevertheless, despite numerous attempts to enhance ozone formation, the energy efficiency of ozone generation does not exceed a dozen percent or so of its theoretical value, which is 1200–1226 g kWh–1.125,126 Ozone can be also synthesized through dissociation of dioxygen molecules by UV photons. This process happens in a UV/O3 apparatus where a shortwave, low pressure UV lamp is used to generate UV light with two major peaks at 184.9 and 253.7 nm. The 184.9 nm UV light is absorbed by molecular oxygen and leads to ozone formation by photolysis. The 253.7 nm UV light is absorbed by ozone molecules and leads to oxygen radical formation.127,128 Another common nonthermal plasma for surface modification is CF4 etching plasma. It is used for etching the surfaces for semiconductor chip production and optoelectronic devices. The etching agents in such plasmas are mainly F atoms formed through dissociation of CF4 molecules by direct electron impacts (eq 2). Impact of electrons by F atoms produce excited F atoms and F ions (Table 2). Further chemical reactions of F atoms/ions with surfaces result in the formation of volatile compounds which leave the surface and lead to etching of the material. During etching, ion bombardment transfers the plasma energy to the surface which provides the activation energy for the desorption of reaction products.123 CF4-containing plasma has traditionally been used to etch a wide variety of materials including Si, SiO2, Si3N4, silicides, polymers, organic materials, and metals such as Mo, Nb, Ta, Ti, and W.123 The etching rates of plasma methods highly depend on the target material and its geometry as well as ion energy flux. The latter is greatly influenced by operational conditions such as pressure and power. For instance, higher radio frequency (RF) power usually results in higher etching rates. In this regard, Balachova et al.129 observed that increasing the RF power from 20 to 70 W in CF4 reactive ion etching operated at a frequency of 13.56 MHz and a low pressure (8 × 10–2 mbar) enhanced the etching rate of SiO2 from ∼10 to 50 nm/min. Moreover, they observed a maximum etching rate of 70 nm/min for AZ 5214 photoresist. At each RF power, the etching rate was as follow: AZ 5214 > Si > SiO2 > amorphous hydrogenated carbon a-C:H films.2 Table 2Some Proposed Gas-Phase Reactions for CF4 and C2F6 Plasmas. Data from refs (135 and 136)- Some Proposed Gas-Phase Reactions for CF4 and C2F6 Plasmas. Data from refs (135 and 136) SF6 plasma is also used for surface etching. Similar to CF4 plasma, F radicals are produced through direct electron impacts (eq 3), which can further transform to excited F radicals as well as positive and negative F ions.1303 In the case of a SF6 + O2 mixture, ·SF5 resulting from eq 3 can further react with atomic oxygen to produce more ·F radicals (eq 4).130 SF6-containing plasma can etch various materials such as Si, organic materials, polymers, silicides, Mo, W, Ti, Nb, and Ta.123 Generally, SF6 plasma is more selective toward Si in Si/SiO2 mixtures compared to CF4-containing plasmas. For instance, an etching rate of ∼500 nm/min for Si has been achieved using a SF6 plasma at a pressure of 0.26 mbar, a gas flow of 20 sccm, and a power density of 0.16 W/cm2, which is 30 times higher than the etching rate of SiO2 under the same conditions.131 High etching rates of more than 2.25 μm/min for Si are achievable by using SF6/Ar inductively coupled plasma.1324 C2F6 and CHF3 are other types of plasmas for surface etching and modification of materials such as SiO2 and Si3N4.123 For example, an etching rate of ∼450 nm/min has been obtained for SiO2 by a C2F6 plasma which is higher than the SiO2 etching rate achieved by a CF4 plasma under the same operating conditions.133 Moreover, an etching rate of ∼250 nm/min has been observed for SiO2 by using CHF3, which is comparable to a SiO2 etching rate of C2F6 plasma.134 Generally, various elementary reactions by electron impacts happen in such plasmas to produce reactive species.135−137Tables 2 and 3 summarize some of these reactions. Table 3Some Proposed Gas-Phase Reactions for CHF3 Plasma. Data from ref (135, 137) Some Proposed Gas-Phase Reactions for CHF3 Plasma. Data from ref (135, 137) Among various species in plasma, free electrons gain the highest velocity in the electric field due to their low mass in comparison to other charged particles, resulting in a high chance of collision between electrons and other species. During these collisions, free electrons begin to transmit energy to other species and provide the required energy for plasma reactions including ionization, dissociation, and excitation. As a result, the rate of such reactions depends on the number of electrons with sufficient energy to initiate the reactions. This can be described by the electron energy distribution function (EEDF), which differs for various plasma conditions and discharge systems. However, the EEDF, even for nonequilibrium plasmas, is governed mostly by electron temperature (Te) and described by the Maxwell–Boltzmann distribution (eq 5).1385in which f(ε) is the EEDF or the probability density for an electron with energy of ε, and kB is the Boltzmann’s constant. Since small changes in electric field, electron density or ion density result in complex changes in the EEDF, the plasma properties highly depend on operating parameters.139 For low-temperature plasmas, practical EEDFs can be calculated from measured I-V characteristic of plasmas through eq 6.140,1416in which e is the electron charge, me is the electron mass, and A is the surface collection area used during probing the plasma with a Langmuir probe to obtain the I-V curve. By assuming a Maxwellian distribution, electron density (ne), electron temperature (in Kelvin), and mean energy are obtained by eqs 7–9.140,141789 The mean electron energy for most applicable plasmas is from 1 to 5 eV.138 By considering the Maxwell–Boltzmann distribution, it is possible to calculate the reaction rate constant. This constant can be calculated through the Rutherford formula and using the Thomson model for elementary reaction processes in plasma. For instance, the ionization rate coefficient of a direct ionization process (ionization of a neutral particle by single impact of an electron) can be obtained by eq 10.138,14210in which me is the mass of the electron, Te is the electron temperature in Kelvin, I is the ionization energy, and σ0 is the geometric atomic or molecular cross section. For example, σ0 is about 3 × 10–16 cm–2 and 10–16 cm–2 for argon and molecular nitrogen, respectively.138 Besides direct ionization, neutral species in plasma can be ionized by a stepwise ionization process, especially when the concentration of excited species and plasma density are high. This reaction involves several steps. First, a collision between an electron and a neutral particle produces a highly excited particle. Then, a collision of a free electron with a relatively low energy and the excited particle results in particle ionization. Like direct ionization, the reaction rate coefficient can be obtained by considering the Maxwell–Boltzmann distribution through eq 11.138,14211in which gi and g0 are the statistical weights of an ion and a ground state particle, respectively, E0 is the energy of neutral species, m is the mass of an electron, e is the charge of an electron, and ℏ is the reduced Planck’s constant. By comparing eqs 10 and 11, one may notice that the stepwise ionization is faster than direct ionization due to the high statistical weight of the excited particles. The ratio of stepwise ionization to direct ionization is proportional to (I/kBTe)7/2. This means that when the electron temperature (energy) is considerably lower than the ionization energy, the stepwise ionization is predominant.138 For instance, considering an argon plasma in which an Ar atom needs 15.8 eV for ionization143 and kBTe is around 3 eV (a typical discharge), the ratio of stepwise to direct ionization processes would be 335. More examples of reaction rate coefficients of some proposed reactions in H2/O2 plasma144 are shown in Table 4. Table 4Reaction Rate Coefficients of Various Proposed Reactions in H2/O2 Plasma. Data from ref (144)ReactionReaction Coefficient Reaction Rate Coefficients of Various Proposed Reactions in H2/O2 Plasma. Data from ref (144) Electrical Discharge Plasma Generally, plasmas can be generated by input of electric field energy, thermal energy and radiation or beams of lasers, electrons, and protons as well as UV photons.123 Application of an electric field is the most common method for generating nonthermal plasmas to synthesize and modify materials. Any gas has some electrons and ions which are already formed due to various reasons such as interactions with cosmic rays. By applying an external electric field to a naturally occurring gas, dielectric breakdown occurs, charge carriers accelerate in the electric field, start to impact other particles, transfer their energy to other particles by inelastic collisions, and produce more free electrons, resulting in an avalanche of charged particles. Eventually, a steady-state plasma develops by a balance between the formation of new charged particles and charge carrier losses.120 The discharges can be classified as follows. Direct Current (DC) Discharges They are usually created in closed discharge vessels between interior cathode and anode electrodes. Depending on the applied voltage and discharge current, various types of discharges can be formed. Specifically, at very high currents, an irreversible transition from glow to arc discharge may occur, which can be prevented by using a current limiting resistor in series with the discharge.120,123 In this configuration, the cathode may undergo an etching process which can be used to deposit a thin film on a substrate by cathode sputtering. This process can be made more efficient by applying an external magnetic field parallel to the cathode surface, known as “magnetron discharge.”145 Pulsed DC Discharges Applying a pulsed DC voltage instead of continuous DC voltage increases the operational power and provides more performance controls by changing duty cycles. Moreover, in a DC discharge, plasma chemical reactions in the gas phase may cause variations in the plasma composition between the center of the plasma and its boundary (where favorable surface reactions occur), resulting in inhomogeneous film deposition. However, operating in a pulsed regime while providing a rapid gas exchange in off-cycles minimizes such inhomogeneities in the deposited film.120 Radio Frequency (RF) Discharges They are generated by applying an alternating (AC) voltage in the frequency range of 0.5146 to 100 MHz.123 In these cases, the corresponding wavelengths are between 3 to 600 m which are larger than the reactor dimensions. At low frequencies, the ions accelerate in the field and impact the electrode surface, resulting in formation of secondary electrons similar to DC discharges.123 At high frequencies, ions and electrons cannot reach the electrode surface. In this case, charge carriers are produced by ionization of heavy particles through electron impact. An impedance matching network is used to connect the power supply to the reactor to maximize power transfer and minimize RF power reflection.147 The power coupling to the reactor can be performed in two ways: capacity coupled (CC) discharges and inductively coupled (IC) discharges.123 In CC discharges, the discharge vessel may have inner circular disk-shape electrodes with a few centimeters distance from each other in a dielectric wall (for example quartz or glass).140 Outer electrodes may also be used in some cases. In the latter case, the erosion or contamination of electrodes by chemically reactive species from the plasma are avoided.123 Generally, CC discharges operate with gas pressures of 10–1-103 Pa. By using CC discharges for plasma, it is possible to get an energetic ion current with energies of few hundreds of eV toward the powered electrode.120,123 Its electron densities are in the range of 109-1010 cm–3.123 Plasmas based on CC discharges are used to deposit thin-film materials, etching and sputtering.120 On the other hand, IC discharges are produced by applying an RF electric field through a helical coil surrounding the plasma chamber. RF electric fields can also be generated by a spherical coil in front of the plasma, similar to CC discharges. By varying the magnetic field arising from the RF current in the coil, the plasma-generating voltage is induced. Typical IC plasmas have electron densities of 1011-1012 cm–3. In contrast to CC discharges, IC plasmas have very low ion energies which makes them suitable to be used as surface treatments of semiconductors with minimal structural damage.123 The range of RF frequencies used in IC plasmas is 0.5–50 MHz, usually 13.56 MHz for processing plasma.140,146 Among different operational modes of RF plasma, capacitively coupled electrodeless plasma (CCEP) is of interest for modification of TMDs due to its lesser destructive effects. CCEP is a planar IC plasma which operates in the capacitive discharge mode. Generally, a planar spiral coil transfers the energy from the RF generator to the plasma in both inductive and capacitive modes. The radial potential drop across the two ends of the planar coil produces a capacitively coupled electrostatic field, which is parallel to the sample surface. The mutual induction between the coil and plasma induces an inductively coupled electric field. At a low input power, the radial electrostatic field is much larger than the induced electric field. In this situation, the direct ionization of the input gas occurs by the radial electrostatic field parallel to the sample surface. However, the majority of the generated electrons do not gain enough kinetic energy to initiate further ionization collisions, resulting in a low ionization rate as well as low electron and ion densities. Therefore, the destructive ion bombardment effect is greatly suppressed, which is why CCEP is often referred to as “soft plasma”.141,146,148,149 Microwave Discharges They are produced by electromagnetic waves with frequencies more than 300 MHz,150 resulting in the wavelength being comparable to the size of plasma chamber. Microwave plasmas can operate in a wide range of pressures from 1 to 105 Pa and have high electron densities up to 1013 cm–3 with low ion energies.123 Plasma-Assisted Synthesis of TMDs To date, TMDs have been synthesized via chemical vapor deposition (CVD),151 mechanical exfoliation,152 liquid exfoliation,153 sulfurization,154 atomic layer deposition,155 and electrodeposition.156 Despite significant advances, these methods suffer from various impediments such as high synthesis temperatures and complexity.157,158 The large differences between the melting temperature and vapor pressure of transition metals (Mo, W) and dichalcogenides (S, Se, Te) also hinder direct synthesis methods.159 Plasma-based techniques have recently been employed to synthesize 2D TMDs to overcome these difficulties. The majority of plasma techniques used for TMD thin film synthesis are inspired by what has already been used in the synthesis of graphene, such as the microwave-induced-plasma (MIP)-assisted method,160−163 gas plasma treatment,164,165 plasma-enhanced atomic layer deposition (PEALD),166−169 and plasma-enhanced chemical vapor deposition (PECVD).170−175 The basics of the main methods used for TMDs are schematically shown in Figure 4 and are briefly presented and discussed in the following sections. Schematic representation of different plasma-based methods used for the synthesis of 2D TMDs. Plasma-Enhanced Chemical Vapor Deposition PECVD is an industrial process, which has been widely used for the processing of TMDs due to its good compatibility with fabrication processes in the microelectronics industry.176 PECVD offers a lower operational temperature compared to conventional CVD. This has multiple advantages in the precise synthesis of materials including working on temperature sensitive substrates, reducing thermal stresses, and enhancing film properties. The presence of plasma can reduce impurities such as adventitious carbon, increase substrate affinity for interaction with gaseous precursors by surface activation, and result in a better uniformity and adhesion to substrate.177 In an atmospheric-pressure PECVD technique (Figure 4a as well as Figure 5a,b), MoS2 and WS2 multilayers as well as their heterostructures can directly be grown onto various substrates such as polyethylene terephthalate (PET) and Si at temperatures below 200 °C.178−180 At an operational temperature of 150 °C, it is possible to directly synthesize 1T-MoS2 and 1T-WS2 thin films as well as their heterostructures that cannot be attained via conventional CVD methods.181−183 Schematic illustrations of PECVD of TMDs using (a) a pure transition metal and (b) a metal oxide. Part (b) reprinted with permission from ref (178). Copyright 2018 American Chemical Society. (c) A schematic of the magnetron plasma during the growth of 2D TMDs. Reprinted with permission from ref (190). Copyright 2015 Elsevier. High-angle annular dark-field (HAADF) scanning transmission electron microscopy (STEM) images of Mo1-xWxS2 alloys fabricated by PEALD through (d) a single supercycle composed of 5 cycles for delivering each molybdenum and tungsten precursor and (e) 5 consecutive supercycles composed of one cycle for delivering each molybdenum and tungsten precursor. Scale bars are 2 nm. Insets show the local Fourier transformations, indicating a hexagonal crystal structure. (f and g) STEM brightness profile over the red lines depicted in (d) and (e), respectively. Reprinted with permission from ref (191). Copyright 2022 IOP Publishing. In this process, a chalcogen-containing gas such as H2S, that is serving as a sulfur precursor, is decomposed to its constituent atoms by plasma. The resulting sulfur atoms will then react with the transition metal atoms of the metal oxide to form uniform TMD layers. Byproducts consisting of precursor atoms that are not involved in the main reaction may also be formed. For instance, when using H2S and a transition metal oxide, hydrogen species react with oxygen to form water vapor. In order to create a stable environment for the plasma at a high working pressure, a background of helium gas is commonly used.178 In an alternative approach, a thin layer of metal oxide can be deposited on the substrate before the process; however, the lattice mismatch between the metal oxide and the formed TMD layers along with the fast, high-pressure chalcogenation process may lead to corrugation of the TMD structure.178 It is also possible to deposit a pure transition metal prior to or during the plasma process from a pure metal source.172,184−186 In such cases, using H2 plasma before increasing the temperature during the PECVD process can remove unwanted oxide layers on top of the transition metal layer, improving the quality of the grown TMD layers.187 Recently, PECVD has been utilized to sulfurize already-sputtered amorphous MoS2 films to obtain large-area polycrystalline MoS2 nanosheets. In this procedure, Ar+ ionizes H2S to form H2S+, which is embedded into the MoS2 in an electric field. H2S+ ions further sulfurize the amorphous MoS2 and form polycrystalline MoS2 nanosheets with H2 gas as a byproduct. To enhance the crystallinity, the temperature can be increased from 200 to 400 °C. In comparison, the same procedure without using plasma reduces the crystallinity.188 The PECVD method is also capable of synthesizing vertically aligned TMD layers at low temperatures (300 °C). This vertical alignment is obtained by depositing a metal layer on the substrate prior to the PECVD process and using Ar:H2X (X: chalcogen) as plasma gas.180,189 The impact of gas flow on the morphology of vertically aligned TaS2 layers was studied by Seok et al.189 Remarkably, the authors state that pure H2S is incapable of transforming Ta films to TaS2, while increasing the Ar:H2S ratio to 15:15 resulting in the formation of a few incomplete lateral and amorphous TaS2 films. Further increases in the Ar:H2S ratio to 25:5 resulted in vertically aligned layers. Although the overall sulfur content in the reactor decreased by increasing the Ar flow, more Ar atoms resulted in a higher concentration of reactive sulfur species through their impact with H2S molecules, enhancing sulfur penetration through the Ta layers and making the sulfurization process more efficient. Therefore, the yield of plasma modification strongly depends on the experimental conditions. Plasma-Assisted Physical Vapor Deposition Many current synthesis methods of 2D TMDs are based on CVD approaches.190,192 Besides the advantages of CVD, a variability of film properties over macroscopic areas, high processing temperatures, and highly reactive and often poisonous gas mixtures are some of the disadvantages.193 Physical vapor deposition (PVD) is an alternative method without chemical reactions. In contrast to CVD which involves high temperature and thermo-catalytic control over TMD film nucleation and growth, the higher kinetic energy of arriving species due to the nature of the plasma in plasma-assisted physical vapor deposition (PAPVD) reduces the required growth temperature.190 Despite this great advantage, adjusting the processing condition to reduce defect formation upon ion bombardment is a significant challenge of this process. Modification of pulsed magnetron sputtering deposition (Figure 4b) by adding a tunable auxiliary magnetic field has recently been introduced as a new route for the synthesis of 2D TMDs.190,194 As shown in Figure 5c, a magnetic field with adjustable intensity is employed to control the chemistry of TMDs. Additionally, the magnetic field deflects the plasma flux and controls the density of electrons and ions at the substrate surface. Voevodin et al.190 have proposed that substrate surface recombination is the primary mechanism for the growth of ultrathin MoS2 films. Since there is a direct correlation between the numbers of sulfur vacancies and other structural defects in TMDs with the number of Ar+ ions, the magnetic field effectively controls the spatial distribution of ionized argon species and consequently, the generated defects. The easily scalable magnetron sputtering process provides a direct growth possibility and can be extended for the synthesis of 2D TMD allotropes and other thin nanostructured materials on a molecular scale. A practical way to enhance the quality of the grown TMD films is to increase the mean free path by lowering the carrier and sputtering gas pressure.195 Moreover, utilizing additional targets such as Ag, Ti, and Zr could facilitate implanting metallic nanoparticles within the TMD structure.196,197 Microwave-Induced-Plasma-Assisted Synthesis Microwave-induced-plasma(MIP)-assisted synthesis is a nonequilibrium plasma process in which the concentration of reactive radical, ionic, and excited molecular species is greater than that in the corresponding gas at thermal equilibrium. When solid-state precursors are used, the relative rates of reduction and chalcogenation are different. Under this circumstance, the implementation of MIP can form new structures or induce structural modifications.198 In a typical run, pure elements in powder form are used as precursors and inserted in vacuum-sealed quartz capsules to react and form TMDs (Figure 4c). The flow of the input and output gases is manipulated by placing the quartz ampules inside a bigger quartz tube. During temperature-controlled reactions, the challenges faced with direct synthesis are avoided.159 This method is capable of synthesizing gram-scale layered compounds in a relatively short time (∼1 h).159,199 This capacity makes the MIP-assisted method a competitor to chemical vapor transport (CVT) techniques, which usually need several days to produce the same quantity of these compounds.199,200 Studies have shown that this method enables the production of high-quality TMD flakes with hexagonal shapes and micrometric lateral dimensions having a layer thickness less than 100 nm.159 After exfoliation, 2D TMD nanosheets with acceptable transport and ferroelectric properties usable for magnetic and optoelectronic applications are obtained.199 Recently, Reisinger et al.201 employed this technique to prepare vanadium-based misfit layer compounds with the chemical formula of [(MX)1+δ]m(VX2)n. Herein, M stands for Sn, Pb, Sb, or Bi and X represents a chalcogen atom. By utilizing NH4Cl as a crystallization aid, tube-like and plate-like compounds can be obtained. Plasma-Enhanced Atomic Layer Deposition Atomic layer deposition (ALD) is a powerful technique for the synthesis of high-quality thin films and coatings on complex 3D structures.168,202,203 By introducing low-temperature plasma in the reaction cycle, additional reactivity is provided on the surface.204 In this process, TMD thin films are deposited by using an appropriate precursor and a chalcogen-containing plasma reactant such as H2S (Figure 4d). A high-purity inert carrier gas such as argon enables chemical transfer and gas purge between the sequential exposures to the precursor and the reactants. RF power is applied to ignite plasma during the sulfur precursor pulsing.205 It results in the formation of reactive chalcogen species such as excited chalcogen atoms (*X) and *H2X (in the presence of H2 gas).206 In the case of metal-oxide precursors, the addition of H2 gas enhances the chalcogenation process by acting as a reducing agent.207 Moreover, addition of H2 provides a H-atom-rich environment to remove organic ligands from the TMD surface through hydrogenation.206 The properties of thin films grown by PEALD are controlled by processing parameters including the growth temperature, precursor dosing, extrinsic doping/alloying, purging, and postannealing conditions.168 Recently, PEALD has been used to fabricate TMD alloys such as Mo1–xWxS2 with a high degree of control over the composition range from x = 0 to 1 and atomic ordering from well-mixed to clustered alloy. This method is based on a scheme of supercycles, in which each supercycle is composed of different cycles to deliver molybdenum (n cycles) and tungsten (m cycles) precursors. Fixed numbers of cycles for delivering H2S plasma and Ar purging are also used between precursor cycles. While the ratio of n/(n + m) governs the alloy composition, the total supercycle length (n + m) controls the atomic ordering of the alloy. Moreover, the total number of ALD supercycles N × (n + m) controls the thickness of the alloy.191Figure 5 panels d-g show the HAADF-STEM as well as STEM brightness profiles of two Mo1–xWxS2 alloys fabricated by PEALD with different numbers of molybdenum and tungsten cycles. Molybdenum and tungsten atoms can be distinguished from each other based on their Z contrast under HAADF-STEM imaging. Heavy atoms (tungsten in this case) scatter the electrons more than lighter atoms (molybdenum in this case) and become brighter in HAADF-STEM.191,208 As seen, the sample produced by a longer supercycle length (higher n + m number) has clustered composition, while the metal atoms are well mixed in the sample with a shorter supercycle length (lower n + m number). Since the PEALD-synthesized TMDs have promising quality, they can be used for various applications. So far, various devices such as WS2-based semiconductors with controlled composition,209 edge-rich WS2-based electrocatalysts for hydrogen evolution reaction (HER),210 and MoS2@3D-Ni-foam electrodes for supercapacitors with a high capacitance of 3400 mF/cm2 have been fabricated by PEALD.205 Plasma-Assisted Sulfurization Process In recent years, core/shell nanostructures of transition metal oxide/sulfide have attracted significant attention for the energy industry, particularly solar cells and photocatalytic hydrogen evolution.213−216 Since the thickness and composition of core and shell structures control their properties, precise processing is important.213 Most utilized techniques have various disadvantages such as agglomeration of nanoparticles, low cost-effectiveness, and improper control of the shell thickness, to name but a few. The plasma-assisted sulfurization method has recently been employed to synthesize WO3/WS2 core/shell nanostructured thin films in order to overcome the aforementioned obstacles (Figures 6a,b). In this method, the shell thickness is controlled by adjusting the reactive species of sulfur and hydrogen in an Ar plasma containing H2S. High-resolution TEM studies have shown that the WS2 thickness is controllable via tuning the partial pressures of gases.217 As plasma reduces the sulfurization energy, an intermediate state of W (5+) can be formed during sulfurization (conversion of WO3 (6+) into WS2 (4+)) through oxygen vacancies. Kumar et al.211 have proposed a two-step mechanism including (i) formation of oxygen vacancies in WO3 by reactive species during plasma treatment and (ii) conversion of WO3-x into WS2 by a redox reaction. Herein, the sulfurization temperature controls the thickness of the WS2 shell. It has been shown that the overall process is also controlled by diffusion. The plasma treatment increases the activity of sulfur, reduces the barrier of reaction, and accelerates the overall kinetics of the core/shell formation.218 The core/shell nanostructure has defect-rich surfaces (Figures 6a and b), thus rendering reasonable charge transfer due to substoichiometric phases at their interfaces. Such kind of core/shell nanostructures are promising materials for photocatalysis, solar energy harvesting, and energy storage.211 (a) A TEM image of a core–shell WO3/WS2 synthesized by plasma-assisted sulfurization process. (b) Magnified TEM image of the area depicted by a red circle in (a). Reprinted with permission from ref (211). Copyright 2019 American Chemical Society. (c-e) Cross-sectional TEM images of a PtSe2 film prepared by the plasma-assisted selenization process. (f) Plane-view TEM image of the PtSe2 film. The inset in (f) shows the fast Fourier transform (FFT) image of a ∼30 nm2 area on the PtSe2 film. Reprinted with permission from ref (212). Copyright 2018 John Wiley and Sons. Generally, studies have shown that sulfurization of metallic molybdenum or oxidized molybdenum films by H2S produce 2D MoS2 with much better quality and larger surface area in the presence of plasma in comparison to conventional thermal synthesis. If the plasma-assisted synthesis is carried out under precisely controlled conditions, high-quality films can be obtained at low temperatures.157,219 The limiting factor of MoS2 synthesis from MoOx is the reduction of oxide. Hydrogen radicals in the H2S plasma facilitate the synthesis by facilitating transformation of molybdenum oxide to metallic molybdenum. In fact, the oxidation state of molybdenum plays a pivotal role in such reactions as expected by comparing the enthalpies of these proposed reactions:157121314This also suggests that using molybdenum thin film as a precursor during H2S plasma sulfurization results in better quality MoS2 than using molybdenum oxide precursors which has been practically shown by Perini et al.219 Plasma-Assisted Selenization Process Inductively coupled plasma (ICP) or transformer coupled plasma (TCP) can be employed for the selenization of metal oxides (MOx) at low temperatures. In typical solid–vapor growth processes such as CVD, it is usually difficult to control the number of deposited TMD layers over a large area because of uncontrolled layer formation.220,221 In contrast, predeposition of metal oxides followed by a postselenization process enables controlling the thickness over a larger area.222 This technique also provides an accessible platform for the direct growth of TMDs with a controllable pattern process.223 With a much higher chemical reactivity, Se radicals formed in the plasma environment trigger the selenization via the replacement of existing oxygen by Se atoms while reducing the selenization temperature.224,225 Another promising advantage of this method is the ability to use versatile substrates such as Si, SiO2, glass, and some high temperature-resistant polymers (for example, polyimide films).224 The last type of substrates is of special interests for several applications in transparent and flexible electronics.226 Su et al.212 have employed this technique to fabricate PtSe2 thin films. At first, a Pt thin film is deposited on a substrate by direct current (DC) sputtering in a high purity argon atmosphere. Then, the selenization process is carried out in a vertical furnace, in which the Se powder is placed in a separate tank (Figure 4f) with a different temperature control system. Figures 6c-f show representative TEM images of the synthesized TMD thin films. Other studies have shown that the fabrication of mixed phases of 1T- and 2H-MoSe2/Mo core/shell hierarchical nanostructures at temperatures as low as 400 °C is feasible. At temperatures above 500 °C, a 2H-MoSe2/Mo core/shell structure is formed.227 Due to favorable carrier mobility, flexibility, and high mechanical strength, these 2D TMD layers are useable in electrochemical and photoelectrical applications, for example wearable sensors224 and photodetectors.212 Dielectric Barrier Discharge Recently, a nonthermal plasma method has been used by Zhao et al.228 to synthesize MoS2-based nanocatalysts under ambient conditions. This approach, known as dielectric barrier discharge (DBD), is favorable for forming thinner and smaller MoS2 layers and generates a large number of coordinatively unsaturated active sites.228 The catalysts are synthesized from their salt precursors, which are treated in an atmospheric pressure plasma reactor containing reactive (such as H2S) or inert (such as Ar/H2) gases. Figure 4g shows a schematic of the DBD method for the synthesis of TMDs. The substrate is covered by precursors and then placed between the ground electrode and power electrode/dielectric barrier. A high voltage and an alternating current are applied to the electrodes to create a plasma gas.228,229 This nonthermal procedure creates a number of corner and edge active sites to boost the rate of chemical reactions. For instance, alcohol formation via CO hydrogenation on MoS2-based catalysts exhibits a CO conversion of 27.3% with a selectivity of 68.6% to total alcohols.228 Benefits and Applications of Plasma-Assisted Synthesis of TMDs As mentioned above, employing the plasma methods for the synthesis of TMDs increases reactivity and thus reduces the required synthesis temperature. The methods also yield the opportunity of direct TMD growth on flexible substrates for wearable electronics. Because of the reactive nature of plasma, surface composition and desired defect density can also be controlled by ionic doses, which make it a suitable process to synthesize various morphologies including monolayers, bulk flakes, and core–shell structures for catalytic applications and energy storage.159,167,178,211 The advantages and shortcomings of different plasma-based synthesis methods are summarized in Table 5. Recent studies on the synthesis of 2D TMDs by plasma-based techniques and the fabricated devices are also presented in Table 6. Table 5Advantages and Disadvantages of Plasma-Based Methods Utilized for the Synthesis of 2D TMDsProcessAdvantagesDisadvantagesPECVD• Applicable for direct deposition on various substrates including polymers with a low glass transition temperature for flexible electronics• Toxic byproducts• High compatibility with the fabrication processes of microelectronic devices• Costly equipment• Relatively low-temperature synthesis• Instability against humidity and aging• High deposition rate• Lattice distortion and residual stresses• Applicable with both organic and inorganic precursors• Imperfect rearrangement of the constituent atoms (in the case of prior metal oxide coating on the substrate)176−178,a,b• Transfer-free processPAPVD• Environmental friendliness• Low deposition rate• Economical scalability for square meters coating area• Need for annealing190• Scalability• Absence of toxic precursors and byproductsMIP-assisted• High-quality polycrystalline materials or even large single crystals• Sophisticated equipment159,c• Operability under atmospheric pressure • Rapid and energy-efficiency process • Cost-effectiveness PEALD• Low deposition temperature• Complicated reactor design• High-quality flakes• Limited conformality202,209• High uniformity in large scale• ReproducibilityPlasma assisted sulfurization/ selenization process• Short deposition time at low temperatures• More complexity in the case of using elemental metals157,217,224• High-quality films• No limitation on the substrate (in the case of selenization)DBD plasma• Large-scale production of TMD nanoparticles• Small to medium gas byproducts such as NOx, N2O, CO, O3,228,d,e• Shorter time compared to the conventional thermal methods• Thin and short TMD slabs• Cost-effectiveness• Low-temperature• Vacuum-free ambient conditionaHamedani, Y; Macha, P; Bunning, T.J.; Naik, R.R.; Vasudev, M.C. Chemical vapor deposition - recent advances and applications in optical, solar cells and solid state devices; Intech: Croatia, 2016; chapter 10.bYi, K.; Liu, D.; Chen, X.; Yang, J.; Wei, D.; Liu, Y.; Wei, D. Plasma-enhanced chemical vapor deposition of two-dimensional materials for applications. Acc. Chem. Res.2021, 54, 1011–1022.cHussein, R.; Saifuddin, N. M. A review of microwave-induced plasma for production of high value products from waste glycerol. Res. J. Pharm., Biol. Chem. Sci.2018, 9, 156–184.dYamamoto, T.; Futamura, S. Nonthermal plasma processing for controlling volatile organic compounds. Combust. Sci. Technol. 1998, 133, 117–133.eHu, T.; Jiang, Z.; Fu, Z.; Jiang, Z.-J. A NiFe/NiSe2 heterojunction bifunctional catalyst rich in oxygen vacancies introduced using dielectric barrier discharge plasma for liquid and flexible all-solid-state rechargeable Zn–air batteries. J. Mater. Chem. A2022, 10, 8739–8750. Advantages and Disadvantages of Plasma-Based Methods Utilized for the Synthesis of 2D TMDs Hamedani, Y; Macha, P; Bunning, T.J.; Naik, R.R.; Vasudev, M.C. Chemical vapor deposition - recent advances and applications in optical, solar cells and solid state devices; Intech: Croatia, 2016; chapter 10. Yi, K.; Liu, D.; Chen, X.; Yang, J.; Wei, D.; Liu, Y.; Wei, D. Plasma-enhanced chemical vapor deposition of two-dimensional materials for applications. Acc. Chem. Res.2021, 54, 1011–1022. Hussein, R.; Saifuddin, N. M. A review of microwave-induced plasma for production of high value products from waste glycerol. Res. J. Pharm., Biol. Chem. Sci.2018, 9, 156–184. Yamamoto, T.; Futamura, S. Nonthermal plasma processing for controlling volatile organic compounds. Combust. Sci. Technol. 1998, 133, 117–133. Hu, T.; Jiang, Z.; Fu, Z.; Jiang, Z.-J. A NiFe/NiSe2 heterojunction bifunctional catalyst rich in oxygen vacancies introduced using dielectric barrier discharge plasma for liquid and flexible all-solid-state rechargeable Zn–air batteries. J. Mater. Chem. A2022, 10, 8739–8750. Table 6Processing Conditions, Special Features, Properties and Applications of 2D TMDs Prepared by Plasma-Based MethodsProcessTypeProcessing conditionsImportant featureDevicePerformancePAPVDMoS2MoS2 target, magnetic coil, Ar gasPolycrystalline 3-layered MoS2--Pressure: 2 PaLow incident ion energies (<8 eV) minimized defects190Power: 40 and 60 WCurrent: 0–5 APEALDMoS2Mo(CO)6, H2SStoichiometric and polycrystalline MoS2MoS2@3D Ni-foam as an electrode for supercapacitors3400 mF/cm2 at 3 mA/cm2 with high cyclic stability (>80%) over 4500 charge–discharge cycles205Pressure: 107 PaPower: 100 WTemperature: 200 °CPEALDFeS2, CoS2, NiS2Bis(N,N′-di-tert-butylacetamidinato)iron(II), H2SPure, well-crystalline, and stoichiometric thin films with a thickness of 21 ∼26 nm202--Pressure: 12 PaTime:15 sPEALDMoS2(tBuN)2 (NMe2)2Mo, H2SThickness of 4 nm--Pressure: 20 PaBetter crystallinity on GaN and Al2O3 than oxidized SiaPower: 300 WTemperature: 250 °CPEALDWS2WF6, H2SPolycrystalline layers with a grain size of 5–20 nmb--Power: 100 WTemperature: 300 °CTime: 10 sPEALDWS2bis(tert-butylimido)-bis-(dimethylamido)-tungsten, acetylacetone, H2SMono and few layersc--Pressure: 6.67 PaPower: 500 WTemperature: 250 °CTime: 20 minPEALDHf2SHf(N(CH3)(C2H5))4 and H2SGrain size: ∼27 nm--Power: 1000 WThickness: ∼10–50 nmdTemperature: 150 – 350 °CPlasma-assisted sulfurizationMoS2MoO3, H2SCore–shell MoO2/MoS2 nanoflakese -Pressure: 1.13 × 10–3 PaTemperature: 150 – 550 °CPlasma-assisted sulfurizationWS2WO3, H2S/Ar (10:90, 40 sccm)Activation energy of the sulfurization process: 63.6 ± 1.9 kJ/molCandidate for hydrogen evolution reaction218 Power: 80 WTemperature: 500 °CTime: 1 h,Plasma-assisted sulfurizationWS2WO3, H2S/Ar (10:90)Core/shell WO3-WS2H2 production, Water treatment, solar energy211-Pressure: 20 PaTemperature: 150–550 °CTime: 60 minPlasma-assisted sulfurizationWS2WO3, H2SSuccessful synthesis at temperature as low as 500 °CNH3 detectionDetection limit of 1.4 ppmfPower: 80 WTemperature: 500 °CTime: 60 minPlasma-assisted sulfurizationFeS2F2O3, H2SAs the S:Fe atomic ratio increased from 0 to 2.0, the apparent optical bandgap dropped from 2.2 (hematite) to ∼1 eV (pyrite)Candidate for solar absorberg-Pressure < 1 PaTemperature: 400 °CTime: 90 minPlasma-assisted sulfurizationMoS2MoO3, SMaximum thickness: ∼16 nmh--Pressure: 86.6 PaTemperature: 100 – 600 °CTime: 60 minPlasma-assisted sulfurizationMoS2Mo, H2S5-nm-thick MoS2 layer on the sample treated for 2 s.--Pressure: 20 PaSulfur concentrations decreased with timeiPower: 800 WTemperature: 400 °CTime: 2 s to 30 min,Plasma-assisted sulfurizationMoS2MoO3, H2S:Ar (10:90)Core/shell MoO2/MoS2 nanoflakes with different shell thicknesses (4–16 nm)j -Pressure: 1.1 × 10–3 PaTemperature: 550 °CTime: 60 minPlasma-assisted sulfurizationWS2WO3, H2SThickness of 11 nm217--Pressure: 86.6 PaTemperature: 550 °CPlasma-assisted selenizationMoSe2Mo, Se, N2/H2 (100:40)Wafer Scale 1T- and 2H-MoSe2/Mo core/shell 3D-hierarchical nanostructuresCatalysts227Catalytic activities: 34.7 mV dec–1Pressure: 266.7 PaLow potential: −166 mV at J = 10 mA cm–2Temperature: 220 °CTime: 8 minPECVDMoS2Mo, H2SDouble-layered MoS2172--Power: 600 WTemperature: 150 – 300 °CTime: 30 minPECVDTaSxTa, H2SThickness: 1.5 nm170--Pressure: 1.3–56 PaPower:70 WTemperature: 400 °CTime: 20 minPECVDMoS2Mo, H2S4-in. 4–5 layered MoS2-on-graphene heterostructuresElectronic devices and catalysis177-Temperature: 300 °CA large density of sulfur vacancies and grain boundariesTime: 1.5 hPECVDWS2, MoS2MoO3 and WO3 layersTMD multilayers synthesized directly on flexible substratesWS2/Si and MoS2/Si 2D-3D HeterojunctionLarge and fast photocurrent responses under the illumination of green light (at λ = 532 nm)Pressure: > 1.3 × 104 PaResponsivity of 37 mA/W and 60 mA/W for MoS2/Si and WS2/Si, respectively178Temperature: < 200 °CPECVDMoS2 (1T)Mo, H2SHighly uniform 1T phase of MoS2HER catalystEnhanced catalytic activity compared to 2H MoS2Pressure: 13.3 PaLateral size: 4 in. in diameterRetaining 1T phase after > 1000 cycles of HERs181Power: 550 WThickness: 3–4 layersTemperature: 150 °CInterlayer spacing: 0.62–0.64 nmTime: 90 minPECVDWS2 (1T)W, H2SHighly uniform 1T phase of WS2HER catalystEnhanced catalytic activity compared to untreated planar CVD-grown 1T WS2Pressure: 13.3 PaLateral size: 4 in. in diameterRetaining 1T phase after 700 h exposure to harsh catalytic conditions and > 1000 cycles of HERs182Power: 550 WThickness: 4–5 layersTemperature: 150 °CInterlayer spacing: 0.62–0.63 nmTime: 90 minFlow: 90/10 sccm (H2S/Ar)PECVD(As2S2)nAs, SThickness: 0.43–20 μmDisk-like polarizability sensork-Pressure: 13.3 PaSurface roughness: 3 nmPower: 50 WTemperature: 200 and 380 °CPECVDMoS2Mo, H2SThicknesses: 1.75–1.77 nmFlexible humidity sensorl-Pressure: 13.3 PaPower: 550 WTemperature: 150–300 °CTime:1 hPECVDMoS2S, MoCl5, Ar,Few-layers with nanoparticlesm--Pressure: 26.7–53.3 PaTemperature: 500 °CPECVDMoS2S, MoO3, ArTriangular and hexagonal single crystals of 2D MoS2 with lateral sizes up to 10 μm--Power: 0–200 WPlasma resulted in nonuniform edge structuresnPressure: 83.99 PaTemperature: 750–850 °CFlow: 50 sccmPECVDMoS2-WS2Mo, W, Ar/H2S (10:20 sccm)Vertical heterostructurePhotodetectorPhotoresponsivity: 83.75 mA/W (at λ = 660 nm)oPressure: 6.67 PaPower: 550 WTemperature: 150 °CTime: 1 hPECVDMoS2-WS2Mo, W, Ar/H2S (40:10 sccm),Vertical heterostructureDiodesThe ability to rectify the current180Pressure:13.3 Pa1–2 nm in thicknessTemperature: 300 °CTime:1.5 hPECVDMoS2MoO3, H2S5 nm thick MoS2 synthesized on Ti2O5 switching layermemristor2D MoS2 buffer layer significantly stabilized the switching behaviorpPressure 10.1 × 104 PaPower: 400 WFrequency: 13.56 MHzTemperature: 200 °CPECVDPdSe2Pd, Se, ArLarge-area polycrystalline film with different thicknessesFETIon/Ioff ratio: 104 (4.5 nm thickness)Pressure 9 PaPd film first transforms to intermediate PdSe2-x phase and then to PdSe2Hole mobility: 0.93 cm2 V–1 S1– (11 nm thickness)qPower: 500 WFrequency: 13.56 MHzTemperature: 150–350 °CDBDMoS2Potassium-doped MoS2, H2S/H2 (60 sccm)Crystallite size: 7.7–31.5 nmMoS2-based catalysts for alcohol synthesisHighest CO conversion of 27.3% with the selectivity to total alcohols of 68.6%228Power: 30 WSpecific surface area: 35–63 m2/gTemperature: 180 °CTime: 20 minDBDMoS2h-MoO3, In(NO3)3.xH2O, Ni(NO3)2.6H2O, H2S/Ar (80:20, 40 sccm)In2S3-NiS codecorated MoO3@MoS2 heterojunctions were synthesizedSolar-light induced CO2 photoreductionCH4 and CO yields were 49.11 and 6.19 μmol g–1 h–1, respectivelyrPower: 110 WTemperature: 150 °CTime: 40 minRemote H2 plasmaMoS2[NH4]2MoS4, H2 (100 sccm)MoS2/Carbon nanowall NanocompositesSupercapacitorCapacitance: 33 mF·cm–2Power: 1000 WCharging–discharging rate: 1000 mV·s–1Time: 30 minDurability: > 5000 cyclessaMughal, A. J.; Walter, T. N.; Cooley, K. A.; Bertuch, A.; Mohney, S. E. Effect of substrate on the growth and properties of MoS2 thin films grown by plasma-enhanced atomic layer deposition. J. Vac. Sci. Technol., A2019, 37, 010907.bGroven, B.; Mehta, A. N.; Bender, H.; Smets, Q.; Meersschaut, J.; Franquet, A.; Conard, T.; Nuytten, T.; Verdonck, P.; Vandervorst, W.; et al. Nucleation mechanism during WS2 plasma enhanced atomic layer deposition on amorphous Al2O3 and sapphire substrates. J. Vac. Sci. Technol., A2018, 36, 01A105.cBalasubramanyam, S.; Merkx, M. J.; Verheijen, M. A.; Kessels, W. M.; Mackus, A. J.; Bol, A. A. Area-selective atomic layer deposition of two-dimensional WS2 nanolayers. ACS Mater. Lett.2020, 2, 511–518.dChang, A.; Mao, Y.; Huang, Z.; Hong, H.; Xu, J.; Huang, W.; Chen, S.; Li, C. Low-temperature plasma enhanced atomic layer deposition of large area HfS2 nanocrystal thin films. Chin. Phys. B2020, 29, 038102.eKumar, P.; Singh, M.; Sharma, R. K.; Reddy, G. Reaction mechanism of core–shell MoO2/MoS2 nanoflakes via plasma-assisted sulfurization of MoO3. Mater. Res. Express2016, 3, 055021.fO’Brien, M.; Lee, K.; Morrish, R.; Berner, N. C.; McEvoy, N.; Wolden, C. A.; Duesberg, G. S. Plasma assisted synthesis of WS2 for gas sensing applications. Chem. Phys. Lett.2014, 615, 6–10.gMorrish, R.; Silverstein, R.; Wolden, C. A. Synthesis of stoichiometric FeS2 through plasma-assisted sulfurization of Fe2O3 nanorods. J. Am. Chem. Soc.2012, 134, 17854–17857.hKumar, P.; Singh, M.; Sharma, R. K.; Reddy, G. An experimental study: Role of different ambient on sulfurization of MoO3 into MoS2. J. Alloys Compd.2016, 671, 440–445.iVesel, A.; Zaplotnik, R.; Gaillard, N. Synthesis of MoS2 by treating molybdenum in H2S plasma. Synthesis2018, 417, 421jKumar, P.; Singh, M.; Sharma, R. K.; Reddy, G. A study on role of partial pressure in controlled synthesis of core–shell MoO2/MoS2 nanoflakes. Mater. Chem. Phys.2016, 178, 6–11.kMochalov, L.; Nezhdanov, A.; Logunov, A.; Kudryashov, M.; Krivenkov, I.; Vorotyntsev, A.; Gogova, D.; Mashin, A. Optical emission of two-dimensional arsenic sulfide prepared by plasma. Superlattices Microstruct.2018, 114, 305–313.lAhn, C.; Lee, J.; Kim, H. U.; Bark, H.; Jeon, M.; Ryu, G. H.; Lee, Z.; Yeom, G. Y.; Kim, K.; Jung, J.; et al. Low-temperature synthesis of large-scale molybdenum disulfide thin films directly on a plastic substrate using plasma-enhanced chemical vapor deposition. Adv. Mater.2015, 27, 5223–5229.mBeaudette, C. A.; Held, J. T.; Mkhoyan, K. A.; Kortshagen, U. R. Nonthermal plasma-enhanced chemical vapor deposition of two-dimensional molybdenum disulfide. ACS Omega2020, 5, 21853–21861.nPokhrel, H.; Duncan, J. A.; Woli, Y. B.; Hoang, T. B.; Pollard, S. D. The effect of Ar plasma on the space-confined growth of MoS2 with low-pressure chemical vapor deposition. AIP Adv.2023, 13, 065322.oKanade, C. K.; Seok, H.; Kanade, V. K.; Aydin, K.; Kim, H. U.; Mitta, S. B.; Yoo, W. J.; Kim, T. Low-temperature and large-scale production of a transition metal sulfide vertical heterostructure and its application for photodetectors. ACS Appl. Mater. Interfaces2021, 13, 8710–8717.pJin, S.; Kwon, J. D.; Kim, Y. Statistical analysis of uniform switching characteristics of Ta2O5-based memristors by embedding in situ grown 2D-MoS2 buffer layers. Materials2021, 14, 6275.qZhang, R.; Zhang, Q.; Jia, X.; Wen, S.; Wu, H.; Gong, Y.; Yin, Y.; Lan, C.; Li, C. Thickness-dependent carrier transport of PdSe2 films grown by plasma-assisted metal selenization. Nanotechnology2023, 34, 345707.rZheng, X.; Wang, H.; Wen, J.; Peng, H. In2S3-NiS codecorated MoO3@ MoS2 composites for enhancing the solar-light induced CO2 photoreduction activity. Int. J. Hydrogen Energy2021, 46, 36848–36858.sShin, J. H.; Choi, Y. S.; Park, H. J. Remote plasma-induced synthesis of self-assembled MoS2/carbon nanowall nanocomposites and their application as high-performance active materials for supercapacitors. Nanomaterials2022, 12, 1338. Processing Conditions, Special Features, Properties and Applications of 2D TMDs Prepared by Plasma-Based Methods Mughal, A. J.; Walter, T. N.; Cooley, K. A.; Bertuch, A.; Mohney, S. E. Effect of substrate on the growth and properties of MoS2 thin films grown by plasma-enhanced atomic layer deposition. J. Vac. Sci. Technol., A2019, 37, 010907. Groven, B.; Mehta, A. N.; Bender, H.; Smets, Q.; Meersschaut, J.; Franquet, A.; Conard, T.; Nuytten, T.; Verdonck, P.; Vandervorst, W.; et al. Nucleation mechanism during WS2 plasma enhanced atomic layer deposition on amorphous Al2O3 and sapphire substrates. J. Vac. Sci. Technol., A2018, 36, 01A105. Balasubramanyam, S.; Merkx, M. J.; Verheijen, M. A.; Kessels, W. M.; Mackus, A. J.; Bol, A. A. Area-selective atomic layer deposition of two-dimensional WS2 nanolayers. ACS Mater. Lett.2020, 2, 511–518. Chang, A.; Mao, Y.; Huang, Z.; Hong, H.; Xu, J.; Huang, W.; Chen, S.; Li, C. Low-temperature plasma enhanced atomic layer deposition of large area HfS2 nanocrystal thin films. Chin. Phys. B2020, 29, 038102. Kumar, P.; Singh, M.; Sharma, R. K.; Reddy, G. Reaction mechanism of core–shell MoO2/MoS2 nanoflakes via plasma-assisted sulfurization of MoO3. Mater. Res. Express2016, 3, 055021. O’Brien, M.; Lee, K.; Morrish, R.; Berner, N. C.; McEvoy, N.; Wolden, C. A.; Duesberg, G. S. Plasma assisted synthesis of WS2 for gas sensing applications. Chem. Phys. Lett.2014, 615, 6–10. Morrish, R.; Silverstein, R.; Wolden, C. A. Synthesis of stoichiometric FeS2 through plasma-assisted sulfurization of Fe2O3 nanorods. J. Am. Chem. Soc.2012, 134, 17854–17857. Kumar, P.; Singh, M.; Sharma, R. K.; Reddy, G. An experimental study: Role of different ambient on sulfurization of MoO3 into MoS2. J. Alloys Compd.2016, 671, 440–445. Vesel, A.; Zaplotnik, R.; Gaillard, N. Synthesis of MoS2 by treating molybdenum in H2S plasma. Synthesis2018, 417, 421 Kumar, P.; Singh, M.; Sharma, R. K.; Reddy, G. A study on role of partial pressure in controlled synthesis of core–shell MoO2/MoS2 nanoflakes. Mater. Chem. Phys.2016, 178, 6–11. Mochalov, L.; Nezhdanov, A.; Logunov, A.; Kudryashov, M.; Krivenkov, I.; Vorotyntsev, A.; Gogova, D.; Mashin, A. Optical emission of two-dimensional arsenic sulfide prepared by plasma. Superlattices Microstruct.2018, 114, 305–313. Ahn, C.; Lee, J.; Kim, H. U.; Bark, H.; Jeon, M.; Ryu, G. H.; Lee, Z.; Yeom, G. Y.; Kim, K.; Jung, J.; et al. Low-temperature synthesis of large-scale molybdenum disulfide thin films directly on a plastic substrate using plasma-enhanced chemical vapor deposition. Adv. Mater.2015, 27, 5223–5229. Beaudette, C. A.; Held, J. T.; Mkhoyan, K. A.; Kortshagen, U. R. Nonthermal plasma-enhanced chemical vapor deposition of two-dimensional molybdenum disulfide. ACS Omega2020, 5, 21853–21861. Pokhrel, H.; Duncan, J. A.; Woli, Y. B.; Hoang, T. B.; Pollard, S. D. The effect of Ar plasma on the space-confined growth of MoS2 with low-pressure chemical vapor deposition. AIP Adv.2023, 13, 065322. Kanade, C. K.; Seok, H.; Kanade, V. K.; Aydin, K.; Kim, H. U.; Mitta, S. B.; Yoo, W. J.; Kim, T. Low-temperature and large-scale production of a transition metal sulfide vertical heterostructure and its application for photodetectors. ACS Appl. Mater. Interfaces2021, 13, 8710–8717. Jin, S.; Kwon, J. D.; Kim, Y. Statistical analysis of uniform switching characteristics of Ta2O5-based memristors by embedding in situ grown 2D-MoS2 buffer layers. Materials2021, 14, 6275. Zhang, R.; Zhang, Q.; Jia, X.; Wen, S.; Wu, H.; Gong, Y.; Yin, Y.; Lan, C.; Li, C. Thickness-dependent carrier transport of PdSe2 films grown by plasma-assisted metal selenization. Nanotechnology2023, 34, 345707. Zheng, X.; Wang, H.; Wen, J.; Peng, H. In2S3-NiS codecorated MoO3@ MoS2 composites for enhancing the solar-light induced CO2 photoreduction activity. Int. J. Hydrogen Energy2021, 46, 36848–36858. Shin, J. H.; Choi, Y. S.; Park, H. J. Remote plasma-induced synthesis of self-assembled MoS2/carbon nanowall nanocomposites and their application as high-performance active materials for supercapacitors. Nanomaterials2022, 12, 1338. Functionalization Although 2D materials have revealed their capabilities in a broad range of applications over the past decade, fast-growing technologies have spurred researchers to seek novel approaches that can fully exploit their potential and discover new properties.230 One way to modify and enhance the nanomaterial properties is functionalization.231 By physisorption (noncovalent) or chemisorption (covalent) of adsorbate atoms, molecules, or atomic groups at the surface or edges of 2D materials, the surface chemistry and electronic band structure can be altered.232,233 TMDs, however, tend to be fairly inert. The chalcogen atoms in the basal planes of TMDs have filled valence shells and the metal atom layer is confined within two layers of chalcogen atoms. Functionalization is nevertheless feasible by physisorption on the basal plane or chemisorption at chalcogen atoms or defects. Surface functionalization of TMDs by wet processes to modulate the charge carrier density with dipole interactions has been reported extensively.92 Functionalization with molecules containing hydroxyl, carboxyl, and ammonium groups can be used to improve the solution dispersity/stability and enhance interfacial bonding between TMDs and the matrix in composite materials.91 Nevertheless, solution-based functionalization processes suffer from long processing times and chemical waste issues from the use of large volumes of organic solvents, as well as harsh conditions and small aspect-ratios of the synthesized TMDs.234 Recently, the controllability, reproducibility, and uniformity of the plasma-based functionalization methods have been found to be promising because they yield properties comparable with other techniques.235−237 In this section, we first review plasma functionalization and its impacts on the properties of TMDs. Afterward, the importance of functionalization in the performance of TMD-based devices is elaborated. Oxygen Functionalization The edges and surfaces of TMD sheets provide adsorption sites for oxygen atoms, particularly in harsh environments such as during plasma treatment.238 The unique electronic structure of the edges compared to bulk TMD provides unsaturated bonds and broken symmetry that can result in localized states and energy levels, acting as preferential landing sites for oxygen species. In terms of oxygen radicals adsorbing on to the surface, a TMD’s DOS at the adsorption sites can be influenced by the interaction between the adsorbate and the TMD surface. The oxygen atom/radical can introduce additional electronic states or modify the existing DOS at the adsorption sites. In one of the earliest studies,239 it was demonstrated that surface activation using the UV-ozone treatment functionalizes the surface of MoS2 by forming weak sulfur-oxygen bonds without breaking the Mo-S bonds. The basic concept of UV-ozone treatment arises from the significant enhancement of O3 reactivity (up to 4 orders of magnitude) in the presence of UV. The absorption of UV light (254 nm) by oxygen molecules increases the production efficiency of oxygen radicals.238 The appearance of a doublet peak in the S 2p XPS spectrum at 164.8 eV along with O 1s affirms the oxidation of sulfur atoms, while the absence of Mo-O bonds implies that no interaction between the metal and oxygen atoms occurs (Figure 7a). Spectral calculations have revealed a sulfur to oxygen ratio of 1 to 1.3, which confirms the existence of a monolayer of oxygen atoms on the surface.239 The oxygen atoms can either be placed on the top of sulfur atoms (as a functionalizing agent) or substitute in the sulfur-mono vacancies (as a dopant), as schematically shown in Figure 7b. Since the formation of sulfur vacancies is an energy-consuming process, it is thermodynamically less favorable.240 Computational studies have demonstrated that a lower concentration of O atoms favors fewer states in the bandgap excluding shallow bands at the edges (Figure 7c). With regard to changes in the surface energy affecting the properties, Park et al.241 have studied the deposition of atomically thick dielectrics on MoS2 after UV/O3 treatment. It has been shown that with increasing the UV/O3 exposure time, the formation of S-O bonds (at low powers) or Mo-O bonds (at high powers) increases the surface energy of the material. By controlling the surface energy, they could successfully fabricate a semiconductor capacitor with an electron density of 10+17 cm–3.241 (a) Effect of UV/O3 treatment on the S 2p and Mo 3d XPS spectra of exfoliated MoS2. (b) Supercell of MoS2 indicates the most energetically stable adsorption sites for oxygen. Oads and Os stand for oxygen on the top of sulfur and substitutional oxygen on a sulfur vacancy, respectively. Reprinted with permission from ref (239). Copyright 2014 AIP Publishing. (c) The projected DOS for 2 × 2 × 1 monolayer MoS2 with oxygen chemisorption. Reprinted with permission from ref (242). Copyright 2015 John Wiley and Sons. (d) Raman spectra of as-exfoliated and ozone-treated MoS2 flakes. The peak of the SiO2/Si substrate at 520.7 cm–1 is omitted. No MoO3 peak at 820 cm–1 is seen, while the small negative shift shows an n-type doping. Reprinted with permission from ref (243). Copyright 2015 John Wiley and Sons. (e) A schematic diagram of carrier transport in MoS2 under the synergy of UV illumination and O3 plasma process. (f) Schematic energy band illustration of photoresponse to UV illumination with photon energy above the bandgap. Reprinted with permission from ref (244). Copyright 2017 Elsevier. It is important to differentiate between the processes of functionalization and direct oxidation of TMDs. The objective of functionalization is to create surface/edge oxygen terminated species that can modulate the properties of TMDs for the intended applications. During oxidation, oxygen radicals can remove the chalcogen atoms, leading to a deliberate yet catastrophic alteration in the structure of the TMDs. Using computational tools, activation energies for various scenarios of initial stages during oxidation of TMDs such as MoS2 have been calculated.245 For example, the energy barrier for a dissociative oxygen adsorption on pristine 2H-MoS2, in which the oxygen atoms are dissociated from each other and adsorb on top sulfur atoms (functionalization), is estimated to be ∼1.6 eV. This value can be further decreased to 0.8 eV in the presence of a sulfur vacancy which acts as a favorable adsorption site for one of the oxygen atoms in an oxygen molecule.246 Experimentally, various reaction parameters can be precisely controlled to tune the input energy for functionalizing the surface with oxygen without inducing oxidation. Among these parameters, the type of plasma, temperature, oxygen partial pressure, TMDs surface preparation, and plasma duration are known to be critical.247,248 The excess of oxygen can minimize the diffusion-controlled reactions, leading a higher degree of structural changes. Moreover, relatively high-temperature operation of the plasma or the etched surface of TMDs promotes oxidation by overcoming the energy barrier required to chemically interact with or remove chalcogen atoms from the surface.249 Precise control over the functionalization rate provides an opportunity to engineer complicated structures with a more delicate tool. For example, thin oxide gate integration on MoS2, which is a challenge in MoS2-based transistors due to the lack of surface reactivity, has recently been accomplished by using ultralow-power plasma treatment followed by Al2O3 deposition.250 Low-energy plasma energizes the oxygen atoms to be adsorbed on top of the sulfur layer without surface oxidization while maintaining the MoS2 carrier mobility.250 Other interface engineering approaches to deposit a gate dielectric, such as buffer Y2O3, have also been applied successfully on MoS2-based transistors.251 Although the top gate performance is improved, the capacitance decreases dramatically. Hence, the UV/O3 treatment to modify the surface chemistry is often done just before the deposition of HfO2 and/or Al2O3243,252,253 to overcome the low k-dielectric difficulties. XPS studies have determined that introducing nitrogen gas into the UV/O3 chamber will minimize the oxygen chemisorption. No peak shift or doublet peak (164.8 eV) is observed in S 2s, S 2p, and Mo 3d, meaning that the mobility stays relatively unaffected. Raman spectroscopy has also indicated that prolonging the plasma treatment (>30 s) and/or purging the protecting gas (nitrogen) may induce oxygen and nitrogen doping, respectively (Figure 7d), as indicated by peak shifts at 384 and 402 cm–1.243 Other routes to surface functionalization have also been proposed, such as electrochemical reactions between oxygen ions and TMD surfaces during UV/O3 treatment, which change the electronic properties of TMDs.244 As shown in Figure 7e, UV radiation (hν > Eg) is absorbed by 2D MoS2 to generate electron–hole pairs. The photogenerated holes then move toward the topmost layer and are trapped by O–, resulting in the formation of dioxygen molecules, while the photogenerated electrons remained inside the layer, resulting in a higher carrier density.244,254,255 According to the energy band diagram (Figure 7f), the ionized and neutral oxygen from the plasma arrive at the MoS2 surface and accept electrons from MoS2, thus bending the energy bands upward near the surface. As a result, photogenerated holes move along the bent band and react with negatively charged oxygen ions. It has been demonstrated that oxygen functionalization of TMDs by plasma increases the hydrophilicity of their surfaces, paving the way to utilize them in many applications such as catalysis.256 The surface of 2D MoS2 is mildly hydrophobic, depending on the number of layers and concentration of contamination on the surface.257,258 Moreover, it has been shown that decreasing the temperature during CVD growth of MoS2 reduces its surface free energy and exposes the edges through well-defined crystalline (001) planes, resulting in a more hydrophobic MoS2 layer.259 Vertically aligned and flower-like MoS2 structures also exhibit superhydrophobic behavior.260 After applying O2 plasma or UV/O3 treatment, the hydrophobicity of TMDs is altered in a way that more oxygen is present on the surface of MoS2 and therefore a lower surface energy (smaller contact angle) is attained (Figure 8a). Decreasing the bandgap energy due to oxygen-based disordering also augments the conductivity of MoS2. Measuring the Tafel slope, which is typically defined as the slope of linear portions of the overpotential versus the logarithm of the current, during water dissociation by ozone-treated MoS2 indicates a drop from 185 mV/dec to 54 mV/dec. This enhancement is due to the presence of S-O and interaction with water as an indicator of more reactive accessible surfaces to generate hydrogen.260 (a) Changes in the wettability of MoS2 flowers from Cassie–Baxter state (left) to Wenzel state (middle), and actual angle (right) after oxygen functionalization of sulfur atoms. Reprinted with permission from ref (260). Copyright 2016 American Chemical Society. (b) Raman spectra of monolayer MoS2 illustrating pinning of A1g mode and redshift of E12g mode after O2 plasma treatment at 2 s intervals. (c) PL spectra of pristine and plasma-treated MoS2 monolayers. The reduction in PL intensity was attributed to the generation of sulfur vacancies during the plasma treatment. (d) The emergence of Mo(VI) (3d3/2) at ∼235 eV in the Mo (3d) XPS spectrum after O2 plasma treatment. Reprinted with permission from ref (242). Copyright 2015 John Wiley and Sons. Studies have also determined that adsorption of molecular oxygen on top of the chalcogen layer in 2D TMDs during UV/O3 treatment can drastically change the optical properties. For instance, Xie et al.261 have shown that physisorbed dioxygen molecules induce a direct-to-indirect bandgap transition into the MoS2 monolayer, which quenches the photoluminescence (PL) spectrum. The adsorption process is reversible so that laser irradiation or storing in air restores the optical properties. Although physisorption or chemisorption of oxygen molecules on MoS2 is common, their adsorption mechanisms are not yet well understood. The vacancy-based chemisorption at edges or at the basal plane of MoS2 using Coulomb scattering is one of the proposed mechanisms.259,262 Physisorption occurs through van der Waals interactions of the TMD surface by accepting electrons from adsorbates (without the formation of covalent bonds), resulting in charge transfer while reducing the carrier mobility. Su et al.263 have reported a significant enhancement in the photoluminescence (PL) intensity of MoS2 by employing a prolonged UV-ozone treatment (30 min). Their XPS results have shown formation of S-O bonds without considerable molybdenum oxidation. Tip-enhanced Raman spectroscopy and XPS studies further support that UV-ozone treatment functionalizes the surface to absorb molecules with C–O and O–C=O bonds, which could lessen exciton trapping and improve the PL yield. The substantial improvement in PL of MoS2 has been also achieved by a mild oxygen plasma treatment due to following reasons: (i) adsorption of molecular oxygen induces a strong p-doping effect, leading to a transition from trions to excitons; (ii) excitons’ nonradiative recombination is significantly suppressed at the oxygenated sites of MoS2.264 A careful systematic study on mild O2 plasma treatment of MoS2 has revealed that time optimization of the process is necessary to avoid severe oxidation and damage of the structure. Raman spectroscopy can be used to elucidate whether oxygen atoms were chemically bonded atop the chalcogen layer or incorporated into the MoS2 layers by sulfur substitution (Figure 8b). The thickness-dependent E2g1 (out of plane phonon vibrations of sulfur atoms) Raman mode redshifts as plasma duration increases, while the A1g mode which is sensitive to electron–phonon coupling (in-plane phonon vibration among sulfur atoms) remains unchanged. These findings confirm the functionalization of MoS2 by O atoms covalently bound to the sulfur layer.242,265 This also redshifts the PL spectra due to a decrease in the optical band gap by chemisorption of oxygen on the MoS2 surface (Figure 8c). It is worth mentioning that prolonged plasma treatment (>8 s) causes severe oxidation of MoS2, which results in the appearance of a new peak in the molybdenum XPS spectrum (Figure 8d).242 The formation of MoOx compounds shifts the Fermi level toward the valence band. Oppositely, the formation of S-O bonds upon short exposure (as explained before) pins the Fermi level to the conduction band.249 Theoretical simulations support that the work function of MoO3 is higher than that of MoS2; therefore, upward bending of the band structure occurs when the MoS2 surface is covered by MoO3.266 As a result, a peak shift in S 2p and Mo 3d from 161.8 and 228.9 eV to 161.5 and 228.6 eV, respectively, occurs.249 It has also been shown that the substrate temperature affects the oxidation rate of MoS2 during plasma functionalization.249,267 At room temperature, only S-O bonds are detected, while at elevated temperatures, MoOx is formed (Figure 9a). (a) XPS spectra of Mo 3d, S 2p, and valence band regions for MoS2 after exfoliation and exposure to remote O2 plasma at different substrate temperatures. The plasma exposure time is 5 min. Reprinted with permission from ref (249). Copyright 2016 American Chemical Society. Effect of UV–O3 treatment on the XPS spectra of (b) MoS2 (Mo 3d, S 2p), (c) MoSe2 (Mo 3d, Se 3d), and (d) WSe2 (Se 3d, W 4f). Reprinted with permission from ref (268). Copyright 2015 IOP Publishing under CC BY 3.0 [https://creativecommons.org/licenses/by/3.0/]. (e) Se 3d states are composed of Se 3d3/2 and Se 3d5/2. After a H2 plasma treatment, Se 3d peaks are broadened, and the fitting reveals two groups of components (solid red curve and solid blue curve). (f) Hydrogenation degree of MoSe2 as a function of plasma treatment time. Reprinted with permission from ref (269). Copyright 2017 Royal Society of Chemistry. The oxygen surface treatment has not only been performed on MoS2 but also has been expanded to other TMDs. Generally, the required energy to activate the surface of TMDs depends on their chemical composition. For instance, it has been demonstrated that 10 min UV ozone treatment is enough to fully oxidize the basal plane and edges of CVD-grown monolayer MoS2, whereas this time duration only functionalizes edges or defects of WSe2.270 From the thermodynamic perspective, the adsorption energy of O on the surface of MoS2 is considerably higher than that on WSe2.268 Hence, the desorption barrier of oxygen adatoms on WSe2 is negligible, and the majority will leave the surface after the UV/O3 treatment, even at room temperature.270 Further investigations on the surface chemistry of MoS2, MoSe2, and WSe2 have indicated that after 3 min UV/O3 treatment followed by deposition of HfO2, S-O and Se-O bonds appear in the XPS spectrum while no Mo-O or W-O bonds are observed.268 This trend continues up to 60 min for MoS2 without the formation of Mo-O bonds (Figure 9b). In contrast, exposure of MoSe2 and WSe2 for only 6 min results in the appearance of new peaks related to MoSexOy, SeOx (Se 3d), MoSexOy, MoOx (Mo 3d), Se-O, SeOx (Se 3d), and WOx (W 4f) (Figure 9c,d). Hydrogen Functionalization Besides oxygen, other plasma environments such as hydrogen have been utilized to modify TMDs. Studies have shown that hydrogen-based plasma treatment of MoSe2 is an efficient way to generate the hydrogenated TMD.269 Theoretical calculations have determined that due to the short bond length between chalcogen and hydrogen, hydrogenated TMDs are the thermodynamically most stable phases.271 During processing, H+ interacts with available lone pairs of the chalcogen atoms to form new H-X bonds, while electrically neutral hydrogen radicals and excited hydrogen species are less likely to react with the surface.269,272 Se 3d peaks in MoSe2 (normally located at 55.70 and 54.85 eV) are shifted to lower binding energy due to an increase in electron density around selenium during the formation of Se-H bonds (Figure 9e). The amount of hydrogenation can be as high as 32–80% without destroying the crystallinity of the material (Figure 9f).269 Nitrogen Functionalization Surface functionalization by N2 plasma is a practical technique to prepare efficient, ultrathin, and high k-dielectric metal-oxide-semiconductor field-effect transistors (MOSFETs). This process prevents unwanted reactions between the metal contact and MoS2 (which results in the formation of Schottky barriers) by terminating the surface of MoS2 with N atoms.273 The Raman intensities of the E12g and A1g peaks for MoS2 treated with N2 plasma remain relatively high even after 40 min treatment.274 This finding confirms the superior stability of MoS2 in N2 as compared with O2 plasma. It has also been shown that exposing the TMD surface to N2 plasma produces MOSFETs with remarkably low gate leakage and high dielectric strength due to the formation of AlN/Al2O3.275 Plasma treatment using ammonia has also been introduced as a viable approach to modulate the properties of TMDs using N functionalization. It has recently been demonstrated that mild ammonia plasma treatment induces n-type doping in MoS2 and p-type doping in WS2. Such a process promotes the substitution of S vacancies with N atoms, resulting in p-type doping in MoS2. This significantly reduces the density of defects and enhances carrier mobility, increasing it from 7.05 cm2 V–1 s–1 to 16.52 cm2 V–1 s–1. Conversely, excessive plasma treatment of WS2 leads to a considerable increase in S vacancies, enhancing n-type doping and a subsequent decrease in mobility.276 Moreover, the adsorption of N or O during plasma functionalization are known to perturb the periodic lattice of TMDs which activates acoustic-phonon modes and introduces acoustic-phonon Raman scattering as a practical tool for the detection of the functionalization level.277 The magnetic properties of TMDs can also be altered. Recent studies have shown that nitrogen-functionalization of MoS2 particles via plasma treatment enhances their ferromagnetism. The chemically adsorbed nitrogen ion on sulfur and the adjacent molybdenum atoms form conjugated π bonds and contribute to ferromagnetism.278 Practical Applications Deposition and Growth of Dielectric Materials So far, different dielectric materials such as Al2O3, ZnO, HfO2, ZrO2, and AlN have been deposited on TMDs for various applications, particularly for the fabrication of transistors.279 The deposition can be done as a separate process on a previously functionalized surface or concurrently with the in situ surface functionalization of TMDs. The former includes plasma modification of TMDs to form external bonds such as S-O (O2 plasma) or S-H (hydrogenation) before depositing the dielectric film. The additional chalcogen-O (or H) bonds on the TMD surface act as nucleation sites and result in stronger interactions of the TMD surface with precursors of dielectrics.280 As an example, Cheng et al.281 used UV-ozone treatment with a mixture of H2O/trimethylaluminum (TMA) to form a continuous layer of Al2O3 on MoS2. The formed layer without TMA/H2O treatment exhibited an incomplete coverage due to the rapid growth along the step edge of MoS2 flakes (rather than on the terraces) as a result of the presence of highly reactive dangling bonds. The ozone-treated surface provides many active sites to form a uniform film. Ozone molecules decompose to molecular oxygen and highly reactive oxygen atoms which provide nucleation sites for TMA by weak surface adsorption.281 A similar enhancement in Al2O3 growth by ALD has been also obtained by using water plasma prior to ALD deposition. Water plasma functionalizes the MoS2 surface with hydroxyl groups which act as nucleation sites for the chemisorption of TMA.282 Moreover, thermal ALD of HfO2 on WSe2 results in the growth of segregated islands while the same films are deposited uniformly by PEALD regardless of the growth temperature. This unique characteristic in nucleation and growth of PEALD is not exclusive to HfO2 on WSe2 and can be seen for materials such as HfO2 and Al2O3 on MoS2.283 In general, it has been found that the density of nucleation during PEALD increases with increasing the temperature and the number of applied cycles.284 Device Fabrication The surface of TMDs is commonly smooth without major dangling bonds and functional groups. Therefore, deposition of a subsequent layer, which is essential for field-effect transistors (FETs) and capacitors, will neither be uniform nor of good coverage.275 Therefore, it is important to employ a surface treatment before ALD to functionalize the TMD surface and improve subsequent layer deposition. It has been shown that a controlled functionalization of TMDs is critical because the interfacial bonding between two adjacent layers of the top gate and dielectrics determines the final mobility, current leakage, and charge trap density of the prepared device.285 Using N2 plasma to functionalize few-layered MoS2 flakes results in better deposition of ultrathin ZrO2 with precise control of the film thickness during ALD. This improves the electron tunneling through the dielectric layer even at small biases in top-gate MoS2 MOSFETs.273 The performance of MoS2-based FETs with HfO2 as the dielectric layer (deposited after oxygen functionalization) shows a slight decrease in subthreshold swing (SS) and hysteresis with an increase in transconductance and on-current (ION) after functionalization.284 In addition to FET, the performance of nonvolatile FLASH memories can be improved by depositing three consecutive layers of Al2O3/HfO3/Al2O3 on the top of a remote O2 plasma-treated WSe2 monolayer. It has been reported that the writing speed can be enhanced to 1 μs in this configuration in comparison to traditional top-gate TMD-based FLASH memories with a writing speed of 100 ms.286 Increasing the electron density up to 1017 and minimizing the interface trap density to 1011 eV–2 has also been shown for UV/O3-treated MoS2 with Al2O3 as the dielectric layer. By conducting the HER on MoS2 flowers it has been found that the surface becomes very reactive to generate hydrogen after functionalization by UV/O3 plasma treatment.260Table 7 summarizes the results of several studies showing the effects of plasma treatment on the functionalization and applications of TMDs. Table 7Summary of Plasma Functionalized TMDs and Their ApplicationsFunctionalization atom or groupTMDMethodConditionsCharacteristicsGrown layerDevicePerformanceOxygenMoS2UV/O3Pressure: 9 × 104 PaOxygen adatoms provide the nucleation sites and uniform growth for the subsequent layerAl2O3239--Time: 5 minOxygenMoS2UV/O3Power: 15 and 25 mW/m–2Increased wettability and surface energyAl2O3Al2O3- MoS2 MOS capacitor241Electron density: 10+17 cm–3Time: 1–4 minTrap density: 1011 eV–2OxygenMoS2O2 plasmaPower: 27 WChemically bonded oxygen shifts the E12g Raman mode242---Time: 8 sO2 flow: 80 sccmOxygenMoS2Ultralow energy remote O2 plasmaTime: 30 sKeeping the mobility, benefiting from oxygen physisorbed on the surface without oxidation.Al2O3Top-gate FET0.01 pA/um2 gate leakage at 4.5 MV/cmλ: 253.7 and 184.9 nmIon/Ioff ratio: 108Subthreshold: 101 mV/decMobility: 28 cm2/Vs250OxygenMoS2UV/O3Time: 30 s-HfO2Top-gate FET243Mobility: 46 cm2 V–1 s–1λ: 253.7 and 184.9 nmLeakage current: 0.2 pA mm–2 at 6.5 MV cm–1OxygenMoS2UV/O3-Formation of S-O bonds without breaking Mo-S bondsHfO2/Al2O3Top-gate MOS capacitorHigh k-dielectric compared to Si252OxygenMoS2UV/O3-Functionalization without oxidationHfO2Top-gate MoS2 capacitor and transistorHighly improved dielectic/MoS2 interface253OxygenMoS2O2 Remote plasmaPressure: 600 PaSurface bonding with sulfur (S-O bonds) upon short treatments and oxidation upon long treatments249---Power: 2500 WExcitation range: 1.9–3.2 MHzO2 flow: 130 sccmOxygenVertically grown MoS2UV/O3Time: 2–10 minBandgap reduction-Hydrogen evolution platformsIncreasing the reactive surface areaIncreased conductivityMore efficient hydrogen generation260Improved hydrophilicityOxygenMoS2UV/O3Time: 10 minHigher nucleation siteZnO270--WSe2Less structural defectsOxygenReS2UV/O3Pressure: 9 × 104 Pa time: 15 minThe surface is well functionalized.Al2O3280--Nucleation sites are providedOxygenWSe2Remote O2 plasma-Improvement of interfacial bondingAl2O3/HfO3/Al2O3Nonvolatile flash memoryA large electron memory window and a negligible hole memory windowsEnhanced writing speed to 1 μs286OxygenWSe2O3 plasmaTemperature: 30 °CA noncovalent functionalizing effectAl2O3 (using TMA and ozone)High-quality gate dielectricsa-Low deposition temperatureHydrogenMoS2H2 plasmaPressure: 10 PaThe formation of Se-H bonds equips the surface---Power: 20 WDecreased mobility of MoSe2 from 29 to 9 cm2 V–1 s–1.269RF Frequency: 13.56 MHZH2 flow: 5 sccmNitrogenMoS2N2 remote plasmaTime: 10 minUniform dielectric depositionZrO2/Al2O3MOSFETSmall hysteresisHigh mobility (14 cm2 V–1 s–1)Reduced contact resistanceIncreased current density273NitrogenMoS2N2 remote plasmaRF Power: 50 WIncreasing stability of single-layer MoS2Al2O3/ZrO2274--Temperature: 200 °CN2 flow: 20 sccmNitrogenMoS2N2 plasmaN2/5% H2Providing nucleation sites with the physically adsorbed ionsAl2O3FET285Ion/Ioff ratio: >109Time: 60 sSubthreshold swing: 85 mV/decadeField-effect mobility: 45.03 cm2 V–1 s–1aAzcatl, A.; Wang, Q.; Kim, M. J.; Wallace, R. M. Al2O3 on WSe2 by ozone based atomic layer deposition: nucleation and interface study. APL Mater.2017, 5, 086108. Summary of Plasma Functionalized TMDs and Their Applications Azcatl, A.; Wang, Q.; Kim, M. J.; Wallace, R. M. Al2O3 on WSe2 by ozone based atomic layer deposition: nucleation and interface study. APL Mater.2017, 5, 086108. Defects and Defect Engineering in TMDs Modulation of semiconductor materials characteristics to maximize their functionality is an indispensable process in the electronic and optoelectronic industries. In conventional semiconductor fabrication techniques, impurities (foreign atoms) are introduced to increase the density of charge carriers. As an alternative approach, defect engineering is applied to tune the optoelectronic characteristics through forming anion or cation vacancies for inducing n-doping and p-doping, respectively.287 For instance, creating defects in the basal planes of MoS2 can improve its catalytic activity, engineer the bandgap, and tune the photoluminescence (PL) intensity.288 The appearance of a new peak in the PL spectrum of a WS2 monolayer and improved photoconductivity of WSe2 due to substitutional oxygen defects are other practical examples of the defect engineering.289 It has also been shown that the creation of chalcogen vacancies and the formation of substoichiometric oxides through plasma treatments can change the electronic properties of TMDs.290,291 Moreover, defects can act as recombination centers to facilitate the recombination (mostly nonradiative) of the photoexcited excitons/carriers.291 Generally, there are two broad approaches to introducing defects in 2D TMDs, as described below. In-Situ Methods Defects are generated either unintentionally or intentionally during the growth process during CVT or CVD. The processing parameters including the gas flow rate, temperature, and type of precursors influence the defect generation during synthesis.292−294 For instance, TEM studies on WSe2 layers fabricated by the metal–organic CVD show that increasing the synthesis temperature from 650 to 800 °C considerably decreases the vacancy defects.295 A similar effect of temperature on the defect formation in WS2 has been reported.296 Ex-Situ Methods Ion beam irradiation,297 thermal annealing,298 electron beam irradiation,299 α-particle irradiation,300 ozone treatment,301 laser irradiation,302,303 electrochemical desulfurization,304 proton irradiation,305 and plasma treatment are some examples of ex-situ methods and post-treatments to create defects in 2D TMDs. Chemical reduction via reducing agents or solvents,306,307 vacuum activation,308 and lithium-induced conversion309,310 are other practical methods to induce defects not only in TMDs but also in other materials. Even though it has been demonstrated that mechanical milling and sonication also induce defects in 2D materials including TMDs,311,312 it is not feasible to control the type of defects in these cases. Among various modification methods developed for layered TMDs, plasma processing has recently attracted attention due to its great potential to engineer TMDs and tune their properties by creating a wide variety of defects, as described in the following sections. Types of Plasma-Generated Defects Different types of defects can be presented in 2D TMDs. A categorization of these defects is schematically shown in Figure 10 and explained below. Schematic views of different defects in 2D TMDs. Zero-Dimensional (0D) Defects Vacancies are the most prevalent defects in TMDs. They can be divided into several subgroups. MoS2, for example, can have single sulfur vacancies (VS), double sulfur vacancies (VS2), vacancies of one molybdenum and three sulfurs (VMoS3), vacancies of one molybdenum and six sulfurs (VMoS6), occupation of the molybdenum position by 2 sulfurs (S2Mo), and occupation of VS2 by Mo.313 Defects can be introduced intentionally by controlling the plasma process parameters, but doing so requires an understanding of the underlying thermodynamics and kinetics. Scanning transmission electron microscopy (STEM) can be used to observe single metal or chalcogen vacancies generated during plasma treatments.314 Plasma post-treatment can induce such defects not only in the topmost layer but also several layers beneath the surface. By STEM and EDS analysis, Jo et al.315 found that plasma-induced vacancies can penetrate up to 7 nm of PtSe2 flakes while the Se/Pt atomic ratio almost linearly increases from 1.1 at the surface to 2 at depth of around 7 nm. One-Dimensional (1D) Defects Agglomeration of sulfur vacancies leads to sulfur line vacancies, the formation energy of which depends on the number of vacancies.100,316 Edges and grain boundaries are two important groups of 1D defects in TMDs that significantly affect their catalytic activity.304,317 Microcracks (generally with one layer thickness) are also common 1D defects in 2D TMDs treated with relatively high-energy plasma for a prolonged time. Such cracks are the sign of fragmentation caused by lateral contraction due to loss of atoms and a high concentration of vacancies.288,318,319 Two-Dimensional (2D) Defects Ripples and in-plane strains are formed during plasma-assisted synthesis and/or are deliberately introduced by plasma post-treatments.320,321 Studies have shown that these kinds of defects significantly affect the electronic structure of 2D TMDs.321−323 Folding,324 faults in the layer stacking,325 vertical heterostructures that create lattice mismatch in TMDs,326 and voids327 are also regarded as 2D defects that alter optoelectronic properties. The presence of pits modulates the physicochemical properties due to the generation of new defect states near the Fermi level and an increase in carrier concentration.328 As a result, the surrounding atoms are more reactive toward target molecules, leading to enhanced catalytic activity. These defects also increase surface charge separation efficiency due to the reduced migrating paths of charge carriers.328−330 Importantly, the surface distortions can enhance the thermodynamic stability of 2D TMDs as a result of changes in the bond angle, bond length, coordination number, and interatomic distances.328 Plasma treatment can also induce roughened edges, quasi-ordered chained surface damages, and secondary phases (such as metal oxides). The nature and density of these defects can be controlled by the time and temperature (energy) of the plasma.331Figure 11 shows various possible defects in 2D MoS2, which might be produced by plasma treatment. Various possible defects in plasma-treated TMD layers. (a) A STEM image of N2 plasma-treated MoS2 monolayer containing sulfur vacancies (highlighted by yellow circles). (b) A TEM image of MoS2 monolayer containing lines of sulfur vacancies (marked by yellow and red ellipses). (c) A STEM image of a cracked grain boundary in N2 plasma-treated MoS2 monolayer. (d-i) TEM images of O2 plasma-treated MoS2 nanoplatelets (15 min) containing (d-f) roughened edges and (g-i) surface damages. Parts a and c reprinted with permission from ref (332). Copyright 2019 American Chemical Society. Part b reprinted with permission from ref (333). Copyright 2016 American Chemical Society. Parts d-I reprinted with permission from ref (331). Copyright 2019 Elsevier. Effect of Plasma Conditions Energy Similar to ion bombardment and helium ion milling (HIM), plasma processing can be considered as a subgroup of sputtering techniques334 to induce defects in 2D materials. The input energy can be modulated by the processing conditions, such as type of plasma, temperature, gas, and voltage to affect the metal or chalcogen atoms in the 2D material, resulting in different characteristics, for example, degraded electrical performance or transition from semiconducting to metallic behavior.335,336 To perform selective sputtering of atoms, the input energy should be controlled based on the required energy for atomic ablation. In particular, it is necessary to utilize a low-energy plasma to selectively sputter the chalcogen atoms without displacement of metal atoms.318 For instance, about 20 and 6.9 eV are required to deliver to MoS2 lattice to remove molybdenum and sulfur atoms, respectively.99 This energy is provided by ions during the plasma treatment, and energy transfer from the plasma ions to the target is dependent on the masses of ions and target atoms. Herein, it is recommended to choose an ion source that provides a large difference between the minimum ion energy for shifting the atoms (Eion) for displacing metals and chalcogens (Eionmetal – Eionchalcogen). For instance, helium ions should have at least 17.4 eV (EHeS) energy to displace sulfur atoms in the MoS2 structure while this energy should be increased to 130.2 eV (EHeMo) to make molybdenum atoms leave the structure. However, these energy amounts in the case of argon ions are 7 (EArS) and 24.1 (EArMo) eV. As a result, using low mass ions (helium) in plasma provides a larger window for choosing a plasma energy for selective removal of chalcogen atoms during plasma treatment rather than heavier ions (argon).318 Generally, a higher plasma energy results in a higher possibility to produce defects with larger formation energy (2D defects).337 For instance, as compared with atmospheric-pressure DBD, low-pressure RF plasma treatment results in a higher concentration of sulfur vacancies due to higher fluxes of energized plasma species.321 Another way to induce chalcogen vacancies during plasma treatment without reducing the structural integrity of the TMDs layer is to increase the dose of neutral radicals and reduce the ion dose in the plasma, e.g., by shielding the ions with a mesh electrode.338 Dwell Time An increment in the plasma dwell time results in a higher concentration (density) of various defects including vacancies and pores. STEM studies on O2 plasma-treated TaS2 nanolayers (Figures 12a-e) have shown an increase in pore density with increasing plasma time. Nevertheless, prolonged exposure to plasma enhances the possibility of metal vacancy formation because the neighboring sulfur atoms are sputtered away which reduces the energy for molybdenum defect formation.318,339 Gupta et al.340 also studied the effect of oxygen plasma time on defect-generation and etching in 2D MoS2 layers. They found that increasing the dwell time resulted in the formation of a higher density of single and double sulfur vacancies which are dominant under plasma exposure up to 40 s. However, increasing the exposure time more than 40 s resulted in etching of 1- to 4-layered MoS2 flakes. HAADF-STEM micrographs of (a) pristine and O2 plasma-treated TaS2 nanolayers for (b) 5, (c) 10, and (d) 15 min. The inset in (a) is the simulated HAADF-STEM micrograph of the standard structure of 1T TaS2. (e) Defect density in pristine and plasma-treated TaS2 calculated by atom-by-atom intensity. Reprinted with permission from ref (339). Copyright 2016 John Wiley and Sons. (f) Scanning electron microscopy (SEM) image of a MoS2 monolayer indirectly treated with H2 plasma for 30 min. HAADF-STEM images of (g) pristine and (h) H2 plasma-treated MoS2 monolayers. Single and double sulfur vacancies are marked by yellow and green circles, respectively. Red lines mark the sulfur vacancies along the zigzag direction. The inset in (h) shows a magnified HAADF-STEM image of an area containing single and double sulfur vacancies. (i) Differential energy distribution spectra of Auger electrons from pristine and H2 plasma-treated MoS2 layers. Reprinted with permission from ref (341). Copyright 2021 Springer Nature under CC BY 4.0 [https://creativecommons.org/licenses/by/4.0/]. The effect of plasma dwell time on TMD defects was also studied by Lee et al.341 Only sulfur vacancies were found to form in the beginning of indirect H2 plasma treatment. However, continuing the treatment resulted in the formation of hexagonal-shaped nanocracks along the zigzag edge in MoS2 monolayer (Figure 12f). Such cracks were formed due to relaxation of strain induced by defects. Figure 12 panels g and h show the HAADF-STEM images of their pristine and plasma treated MoS2 layers. Indirect hydrogen plasma treatment increased the sulfur vacancy density and even resulted in the formation of double sulfur vacancies. However, due to the high formation energy for double sulfur vacancies, the authors suggest that such defects were generated by a combination of naturally present sulfur vacancies in the bottom sulfur layer and plasma-generated sulfur vacancy in the top sulfur layer. The formation of sulfur vacancies without molybdenum defects was also confirmed by differential energy distribution spectra of Auger electrons (Figure 12i), in which the intensity of Mo MVV remained intact during plasma treatment, while the S LVV peak decreased with increasing plasma dwell time. Moreover, sulfur vacancies were preferentially aligned along the zigzag orientation, resulting in hexagonal-shaped nanocracks at high sulfur vacancy density. Substrate During ion irradiation of 2D materials, defects can be produced by (i) direct impact of ions, (ii) impact of backscattered ions, and (iii) impact of substrate atoms sputtered from the surface. Plasma-generated ions mainly sputter the TMD layers by direct impact.342 Moreover, relatively high energy ions in plasma which are able to penetrate the full thickness of a TMD layer can further aid defect production in TMD layers by backscattered ions (such as He ions in helium plasma) as well as sputtered atoms originating from the substrate (such as Si or O from SiO2). In addition, the substrate itself may decrease the sputtering yield of direct impact by inhibiting the ablation of chalcogen atoms in the bottom layer. Such opposite effects of the substrate can sometimes cancel each other out, resulting in similar amounts of generated defects in substrate-supported and free-standing TMD layers during ion irradiation or ion plasma treatments.318,342,343 However, most plasma treatments which are used for TMD modification employ low energy ions, minimizing the backscattered ions as well as sputtered ions from the substrate. Therefore, it is expected that the substrate protects the underlying surface of the TMD and results in lower damage to its structure during plasma treatment. Studies have also shown that utilizing a rough substrate such as Si/SiO2 with a “hill and valley” structure is likely to affect the impact of plasmas like H2 plasma on the defect formation in 2D TMDs because more chalcogen vacancies are susceptible to be formed in the TMD layer suspended over substrate valleys.344 Similarly, in comparison to free-standing TMD layers, substrate-supported ones may have lower chemical conversion rates for defect-induced reactions such as oxidation during plasma treatment.327 In addition, relatively strong interactions between MoS2 and SiO2 substrates have been suggested to be the origin of the high stability of MoS2 monolayers during N2 plasma treatment without the formation of wrinkles. Such wrinkles may form due to the strain induced by nitrogen adsorption.274 In this regard, intercalation of water molecules between hydrophilic SiO2 and MoS2 reduces the adhesion force and may result in scrolling of the MoS2 monolayer during plasma treatments. In fact, stains produced due to plasma-induced vacancies or lattice distortions are responsible for this effect.345 MoS2 nanosheets on gold substrates may be more stable under O2 plasma in comparison to their counterparts on SiO2 substrates.340 These studies show that the substrate can play a significant role in defect engineering of TMDs during plasma treatments which needs to be further studied. Properties of Defective TMDs Electronic Properties The defects generated by plasma treatment can alter the electronic structure and properties of 2D TMDs including but not limited to the band structure, charge carrier density, contact resistance, band offset, conductivity, threshold voltage, charge carrier mobility, work function, DOS, and bandgap.318,346 Among 0D defects, chalcogen vacancies are n-doping agents in TMD semiconductors by introducing unpaired electrons into the lattice that form localized electronic states in the bandgap. The electrons in those states participate in the electrical current by hopping between localized states.313,347 Recent studies have shown that hybridization between neighboring chalcogen vacancies induces additional shallow defect levels near the conduction band edge. Despite the fact that the concentration of neighboring vacancies is much lower than single vacancies, they have a large impact on the electronic properties of TMDs.348 While sulfur vacancies act as n-type dopants in TMDs, prolonged plasma treatment may induce abnormal p-doping in TMDs such as WSe2 and MoTe2. This is because excessive chalcogen vacancies and edge defects created by the plasma are susceptible to be filled by oxygen in air, resulting in p-type doping.349 n-Doping by plasma is also reported to elevate the Fermi level and decrease the work function.350−352 In contrast, metal vacancies in chalcogen-rich TMDs lead to p-doped TMDs. Among 1D defects, sulfur line vacancies can form triangular loops and inversion domains that result in stoichiometry changes at the edge of these domains and the formation of electronic states in the bandgap.292 There are also deficiencies in either chalcogen or metal atoms at the grain boundaries in 2D TMDs, which locally changes the electronic structure and results in local n-type or p-type semiconductors, respectively.353,354 However, it is important to note that the n/p doping mechanism is heavily influenced by the specific type of plasma used and its operating conditions. For example, it has been shown that when MoS2 is exposed to soft NH3 plasma for a short duration at a low power, it can result in p-doping of the structure due to substitution of sulfur by nitrogen. Conversely, employing high power and an extended exposure time can lead to n-doping of the structure due to the excessive formation of sulfur vacancies.276 Plasma treatment can thus be utilized to develop p-n junctions.355 Sufficiently high defect concentrations can convert semiconducting MoS2 to a metallic material.356 Shawkat et al.335 have found that defects induced by an Ar plasma in 2D PtSe2 generate midgap energy states and reduce the bandgap energy. With prolonged treatment time, the bandgap energy is gradually reduced, ultimately shifting the electrical behavior from semiconductor to metal. Similarly, Soman et al.357 were able to transform semiconducting 2H MoS2 layers to metallic 1T MoS2 by inducing a high density of sulfur vacancies through H2 plasma treatment. Lateral sliding of the top Te layer in 2H-MoTe2 layers by soft hydrogen plasma has also transformed their structures from the 2H to 1T phase, releasing metal-like electrical behavior and increasing the electron conductance by about seven times.358 Optical Properties In MoS2, as a well-known member of the TMDs family, both the lowest conduction bands and highest valence bands are formed mainly from Mo d-orbitals.359 MoS2 in bulk form has an indirect bandgap of 1.2 eV, between its valence band maximum at the Γ point and the minimum of its conduction band halfway between the Γ and K points. At the same time, there is a larger and direct bandgap at the K point. By decreasing the number of MoS2 layers, the indirect bandgap increases due to quantum confinement effects. When MoS2 reaches monolayer thickness, the direct bandgap at the K point, which has been changed less than 0.1 eV, becomes smaller than the indirect bandgap. As a result, MoS2 shows a transition from direct (∼1.2 eV) to indirect (∼1.8 eV) bandgap by changing from monolayer to bulk form.360−362 It has been shown that ultrathin layers of MoS2 have two prominent absorption peaks at 1.85 and 1.98 eV, known as A1 and B1 excitons, respectively. These direct excitonic transitions are located at the K point of the Brillouin zone, and their energy difference is due to valence band splitting arising from the large spin–orbit interaction. However, the PL of MoS2 usually exhibits a different behavior. Direct excitonic transitions show strong PL emissions in monolayer MoS2 while they are absent in the PL spectrum of bulk MoS2, indicating a very high luminescence quantum efficiency in monolayer MoS2.363,364 Studies have shown that the PL quantum yield of MoS2 increases sharply as the thickness is changed from 6 to 1L, reaching to more than 104-fold because of the indirect to direct bandgap transition.16 As a result, any variation in the band structure of layered TMDs caused by defects would be accompanied by changes in excitonic transitions and thereby optical properties.365 So far, bandgap variation and PL shift by dopants,366 the appearance of PL peak because of bisulfur vacancies,300 and strong PL emission from edges367 have been reported. Studies have shown that plasma-generated defects such as chalcogen vacancies act as an excitation trap to decrease the electron–hole pair interactions, quench the PL spectra, and redshift the PL exciton peaks.318,368−371 The last one arises from the exciton transition between the valence band and shallow bands generated by defects.368 Liu et al.372 showed that vacancies and accompanied dopants produced by O2 plasma in WS2 monolayers reduce the band-edge carrier lifetime. In contrast, Wu et al.373 found that sulfur vacancies formed by mild Ar plasma in ReS2 monolayers increased the defect-related PL emission which was strongly anisotropic under the electric field. Such defect-related excitations had polarization direction along the Re-S atomic chains. Moreover, sulfur vacancy containing ReS2 samples were more sensitive to the electric field in comparison to their pristine counterparts. In addition, Nan et al.264 achieved a strong PL enhancement (thousands-fold) by defect engineering and oxygen bonding through thermal annealing of MoS2 monolayers. Such a PL enhancement occurred at defect sites due to heavy p-doping effect of oxygen bonding and suppression of nonradiative recombination of excitons at defects. Using oxygen plasma treatment instead of thermal annealing resulted in controllable manipulation of MoS2 PL. These studies show that plasma treatments would be promising tools for defect engineering of TMDs toward their usage in optoelectronic applications. Vibrational and Structural Properties Structural defects in the plane or at the edge can function as scattering centers in the Brillouin zone and activate phonon modes,374 providing opportunities to study defects by defect-activated Raman modes.375 Increasing the plasma time and/or energy input results in a higher concentration of chalcogen defects and more lattice distortion. This disturbance in the crystal symmetry results in broadening and intensity reductions of the A1g and E12g peaks. A redshift in the E12g Raman mode indicates n-doping, while the A1g peak remains almost unchanged.318,351,376 On the other hand, p-type doping results in a blueshift of only A1g or both E12g and A1g Raman peaks of 2D TMDs.377,378 Importantly, if the plasma contains species which are capable of substituting the chalcogen vacancies (doping), the change in the peak position would differ. For instance, in contrast to Ar plasma, O2 plasma causes a redshift in E12g due to restructuring and damage by oxygen incorporation in the TMD structure.376 The oxygen incorporation also results in p-type doping, which induces a blueshift of the A1g peak.368,379 The E12g/A1g ratio in Raman spectra of 2D TMDs is another indicator of the extent of defects generated by plasma treatment. A higher ratio indicates a surface mostly composed of basal planes. Lower values correspond to the prevalence of edge-terminated structures as they would result from vertically grown TMDs, which is preferable for higher electrocatalytic activity. Plasma-induced defects were found to yield lower E12g/A1g ratios.380,381 Higher defect densities created by longer plasma treatment time also increase the ILA(M)/ILA(K) intensity ratio in resonance Raman spectra of MoS2, which can be considered as a fingerprint of defects in TMDs.382 Moreover, theoretical and experimental studies have shown that vacancy defects in MoS2 reduce phonon group velocity and localize phonons, resulting in lowered thermal conductivity.383,384 Defects generated by plasma processes can also break inversion symmetries in some TMDs such as PtSe2 and induce the Rashba effect.315 This Rashba effect relates to the structural inversion asymmetry generated by spin-splitting of electronic energy bands due to spin–orbit coupling. Defects caused by plasma can intensify spin–orbit coupling, lifting of spin degeneracy, and the splitting of energy bands near the Fermi level, resulting in spin polarization in momentum space. Therefore, spin and orbital degrees of freedom in spin-based applications will strongly rely on the impact of defects on the spin–orbital splitting mechanism. Chemical Properties In pristine TMDs, the edges are active sites, whereas basal planes are often chemically inert. Therefore, defect engineering (edge geometries, cracking basal planes, chalcogen vacancies, and incorporating foreign atoms) is a practical strategy to enhance the catalytic activity of TMDs.385−389 Chalcogen vacancies that are formed during plasma treatment decrease the kinetic barrier for subsequent doping and chemical reactions, for example, oxidation in the presence of air or water vapor318 or functionalization by organic ligands.390 Manipulating the chemistry of 2D TMDs by plasma-induced defects can also enhance their performance in surface enhanced Raman spectroscopy (SERS). For instance, Jing et al.391 achieved an enhancement factor of 1.83 × 106 and a concentration detection limit of 10–13 M for the analyte rhodamine 6G on the surface of MoTe2 substrates that were treated with a H2 plasma. Such enhancements were attributed to p-doping and lattice distortion of few-layered MoTe2 after the formation of sulfur vacancies during the plasma treatment and further oxygen absorption at the vacancies. Extra holes in the MoTe2 valence band enhanced the recombination of the excited electrons from the affinity level of the analyte and holes in the valence band. This resulted in better charge transfer and SERS signal. Magnetic Properties The effects of plasma-generated defects on the magnetic properties of TMDs have not been studied as thoroughly as on other characteristics. A few theoretical and experimental investigations have determined the plausible effect of defects on the magnetic properties of these nanomaterials.375 For instance, edge ferromagnetism and nonzero net magnetic moments in 2D WS2 and MoS2 have been shown by first-principles calculations and confirmed by experimental results.392,393 While simple single and double chalcogen vacancies or single metal vacancies do not alter the nonmagnetic behavior of the MoS2 structure, theoretical studies have shown that applying strain to such structures or producing more complex vacancy defects such as double metal vacancies or coupled double sulfur and molybdenum vacancies induce ferromagnetism in MoS2.394,395 Ferromagnetic and antiferromagnetic behavior in low-angle and high-angle grain boundaries, respectively, has also been demonstrated.393,396,397 Therefore, plasma treatment with the capability to create a wide variety of defects could be employed as a powerful tool to tune the magnetic properties of layered TMDs. For instance, Duan et al.398 changed the interlayer magnetic interaction of MoS2 bilayers from antiferromagnetism to ferromagnetism by introducing cobalt dopants and sulfur vacancies. Their cobalt-doped MoS2 bilayers showed saturation magnetism of 0.2 emu/cm3 at 300 K. However, introducing sulfur vacancies with Ar plasma treatment in such samples increased the saturation magnetism to 1.53 emu/cm3. Their experimental and theoretical studies revealed that such an enhancement in magnetism was due to the hybridization of a sulfur vacancy-induced impurity band and the Co 3d band, which not only changed the forms of interlayer orbital hybridization between Co atoms in the bottom and top layers but also enhanced the interlayer ferromagnetism. In another study conducted by Wu et al.,399 it was shown that introducing vacancies to nonmagnetic 2D ReS2 flakes by Ar plasma treatment and subsequent thermal annealing at 400 °C induced room-temperature ferromagnetism. Increasing the plasma dwell time from 30 to 90 s was found to enhance the saturation magnetism from 0.021 to 0.106 emu/cm3 due to an increase in the vacancy density. By further annealing, the saturation magnetism could reach up to 1.203 emu/cm3. Such enhancements in magnetism by thermal annealing are due to the movement of cations into the anion sites and formation of antisite defects. Applications of Defective TMDs A summary of results reported in the literature on plasma-generated defects on TMDs is presented in Table 8. Several important applications are briefly explained below. Table 8Characteristics and Applications of Defect-Engineered TMDsDefectsTMDsPlasmaConditionPropertiesApplicationPerformanceFeatureInterlayer deformation or twisting featuresMoS2---FETSub-50 nm multibit memories2-Bit and 3-bit data storage states with long scale retention timesaWSe2PoreTaS2O2Pressure: 1–102 PaPores had S-terminated edgesHERCurrent density: 10 mA cm–1 at 564 mV at 0.24 mg cm–2 TaS2 loadingImproved HER activity339Voltage: 290 VTime: 5–15 minSelenium vacancyPtSe2O2Pressure: 9.3 × 103 Pa-Photoelectrochemical HEROnset potential: 0.29 VMore active sitesbPower: 30 WPhotocurrent: – 32.4 mA/cm2 at 9 VTafel slope: 49 mV/decDoping, rippled layersMoS2SF6, CF4, CHF3, O2-p-typeFET, multibit memory FET, and thin-film photovoltaic cellIon/Ioff ratio: >105-107Good uniformity over large areas335Rippled layersMoS2O2, SF6, CF4, CHF3Pressure: 133.3 PaChanging the band structure and electronic dopingData storage FETWriting/erasing ratio: ∼103 after a 1 h retention test (O2 plasma) and 10 to 100 after 2000 s retention tests (CF4 and CHF3 plasma)Retain trapped charges for a much longer time320Power: 200 WTime: 2 minFlow: 10 sccmRippled layersWSe2O2Pressure: 1.3 PaInducing charge-trapping statesData storage FET-Long charge retention timecPower: 200 WTime: 2 minFlow:10 sccmSulfur vacancyMoSe2N2Pressure: 1.73 × 104 PaDefect states in bandgapHEROverpotential: 0.17 VEnhanced catalytic activitydPower: 100 WIncreased electron concentrationTafel slope: 60 mV/decTime: 30–300 sFrequency: 13.56 MHzN2 flow: 400 sccmSulfur and tin vacanciesSnS2ArPressure: 40 PaNonlinear absorption during laser excitatione---Time: 40 sPower: 100 WFrequency: 13.56 MHzAr flow: 40 sccmSulfur vacancyMoS2Low-energy HePressure: 1.3 × 10–2 PaChange from a semiconducting to metallic-like behavior (creation of midgap states), n-type doping effectsFETIon/Ioff ratio: 1.5Decreased Ion/Ioff ratio318Time: 10 min - 2 hResistance: ∼103 ΩFlux: 1.75 × 1012 ions/cm2·sSulfur vacancyMoS2H2--Fin-FETIon/Ioff ratio: >105Low series resistanceIon: 200 μA/μm at 1 VLow leakage deviceAdjustable threshold voltagefSulfur vacancyMoS2H2Pressure: 17.33 PaEnhanced HER activityHEROverpotential: 183 mVDecreased overpotential from 727 to 183 mVRF Power: 25 WTafel slope: 77.6 mV/decDecreased Tafel slope from 164.5 to 77.6 mV/dec385Frequency: 13.56 MHzCharge transfer resistance: 5.9 ΩTime: 5–35 minThe electro-chemical effective surface area: 1.5 mF/cm2Temperature: 25 °CH2 flow: 20 sccmSulfur vacancyWS2 and MoS2ArPressure: 26.66 PaChanging PL: Distinct peak at 0.1 eV below the A-exciton peakg---Power: 25 WTime: 5–30 s,Nanoscale poresMoS2 and WS2O2Power: 18 WPost thermal annealing in air enlarged the pores with a rapid increase in PL intensity and blueshift the PL peak arising from p-doping329---Time: 1–120 sFlow: 283.17 sccmEdges and vacanciesMoS2O2Time: 5–20 minEnhanced electro-catalytic performanceElectrocatalyst380Tafel slope: 44 mV/dec-Double-layer capacitance: 0.11 mF/cm2Surface defectsMoS2O2/Ar (1:3)Pressure: 500 PaShallow Schottky barrierBilayer MoS2 deviceSchottky barrier height (SBH): 0.33 eV (6s plasma)Reduction in SBH from 2 to 0.33 eV434Time: 4–8 s,Lattice distortion and defectsMoS2O2Pressure:1.3 × 10–3 PaRedshifts of ∼20 nm in the PL emission peaks---RF Power: 300 WReduced intensity of E12g and A1g Raman peaksTime: 10–120 sRaman peak shifting to 381.1 cm–1 (E12g) and 406.3 cm–1 (A1g)368O2 flow: 2 sccmSulfur vacancyMoS2ArPressure: 1.33 Pa-MemristorLearningAn increase in the concentration of movable sulfur vacancies enhances the analogous switching propertieshRF Power: 50–500 WAccuracy for the MNIST handwritten digits data set: 94%Time:10 sDynamic range: 14.8Ar flow: 10 sccmsp3-type defectsMoS2Al2O3 as a sputtering targetPressure: 0.73 PaRedshift PL (n-type doping)Electrical deviceCurrent: ∼40 pA (5 min plasm) and ∼1 nA (after 300 °C annealing)Reduced electrical conductivity370RF Power: 5–100 WTime: 1–60 minSulfur defectsMoS2 and MoSe2simulations-The emergence of dangling bond state slightly below Fermi energyHER-Enhanced catalytic activityiChalcogen defectsMoS2 and WSe2H2Time: 10 minn-type dopingBack-gated MoS2 transistorReduced source/drain series resistance from 640 to 240 kΩ·μmImproved metal/TMD contactjSulfur vacancyMoS2O2 and Ar plasmaRF Power: 5–10 PaFilling localized defect states (formation of Mo–O bonds), p-type dopingFETElectron mobility: ∼5.6–39 cm2 V–1 s–1Improved electron mobility by filling vacancies438Oxygen dopingFrequency: 13.56MHzTime: 5–20 s,Sulfur vacancyMoS2O2Power: 1–3 Wp-doping improves hole transfer rate between CH3NH3PbI3 and MoS2Perovskite-based optoelectronic devicesHole transfer from CH3NH3PbI3 to MoS2 in 320 fs with 83% efficiencyUltrathin hole transport layer346Oxygen doping Time: 1–10 s Flow: 20 sccmStraining and mechanically distortingMoS2H2/O2/N2Pressure: 9.33–16 PaPromoted dielectric deposition439---RF Power: 50 WTemperature: 200 °CFlow: 20 sccmStructural disordersMoS2O2 and ArTime: 5 sEnhancement of Rhodamine 6G (R6G) deposition on plasma-treated MoS2k---O2 or Ar flow: 80 sccmSelenium vacancyWSe2H2Pressure: 18.66 Pan-dopingn-type FETsSubthreshold swing: 66 mV/dec20× improvement in ON current287Power: 5 WTime: 5 sH2 flow: 100 sccmCracksMoS2O2Pressure: 1333 PaFormation of exposed edgesHERTafel slope: 171 V/decade for 20 s plasma treatmentHigher HER activity288Lattice distortionPower: 300 WMo–O bondsTime: 10–30 sSulfur vacancySulfur vacancyMoS2 and WS2CF4 thenPressure: 2 PaControlled n-doping (sulfur vacancy) and p-doping (oxygen incorporation)FETMobility: 0.87 (for p-doped MoS2), 0.13 (for p-doped WS2), and 25 cm2 V–1 s–1 (for n-doped MoS2)412-Oxygen incorporationO2 /ArPower: 50–60 WTime: 10–40 sFlow: 4–30 sccmDeep defectsMoS2O2Power: 40 Wp-type dopingMemory field-effect transistors37 V memory hysteresis window under a + 40/–50 V gateDefects as charge trapping sites to improve the memory capabilitylTime: 6 svoltage pulse for a W/E ratio of ∼102O2 flow: 7 sccmTriangle pitsMoS2O2Power: 200 W,Increased active sitesHEROverpotential: 0.34 V at a current density of 0.5 mA·cm–2Enhanced HER performancemTime: 30 s - 15 minVarious defectsNiSe2/MoS2AirPower: 29.6 W-Composited with carbon fiber paper for HERoverpotential decreased from 143 to 137 mV, and Tafel slope increased from 45 to 50 mV dec–1No significant improvement in catalytic performancenVarious defectsReSe2 ICP Power: 30–120 WIncrease in defect densities decreases the anisotropy of the crystalo---Frequency: 0.5 MHzAr flow: 30 sccmSulfur vacancy and oxideWS2Ar/O2Power: 4 WDecrease in PL IntensityPhotoelectricPhotocurrent at zero gate voltage under 365 nm light: 1.1 μA150-fold increase in photocurrent in vacuum due to formation of sulfur vacancies and possible W oxidespTime: 5 sFlow: 15 sccm (Ar/O2: 50:50)Sulfur vacancyMoS2O2Pressure: 60 PaIncrease in defect density by increasing plasma timeMoS2/hBN/WS2heterostructure proposed for optoelectronic devicesphotoinduced energy transfer rate: 6.58 ns–1 under 405 nm excitation5 times enhancement in photoinduced energy transfer rateqPower: 10 WFrequency: 13.56 MHzTime: 0–72 sSulfur vacancyMoS2Soft ICP NH3 plasma (CCEP)Pressure: 10 Pan-type doping due to sulfur vacancies formed by H etching prevailed p-type doping by N incorporationFETMobility: 3.2 cm2 V–1 s–1Reduced mobilityRF Power: 20 Wthreshold voltage: −13.3 VNegatively shifted threshold voltage276Time: 1–2 minFrequency: 2 MHzFlow: 10 sccmSulfur vacancyMoS2Mild O2 plasmaPressure: 5 Pa100 times enhancement in PL intensity---Power: 5 WChemical adsorption of oxygen264Time: 30 sFrequency: 13.56 MHzLateral sliding of top chalcogen layerMoTe2Soft ICP H2 plasma (CCEP)Pressure: 10 Pa2H-1T’ phase transition (more in ″structural changes″ section)FETLinear Ids–Vds characteristicsOhmic behaviorPower: 50 WReduced temperature dependence of resistance358Frequency: 2 MHzaChen, M.; Liang, X. Multibit memories fabricated through mechanical and plasma induced deformation of layered semiconductors. IEEE 17th International Conference on Nanotechnology. 2017; p 705–706.bChung, C. C.; Yeh, H.; Wu, P. H.; Lin, C. C.; Li, C. S.; Yeh, T. T.; Chou, Y.; Wei, C. Y.; Wen, C. Y.; Chou, Y. C.; et al. Atomic-layer controlled interfacial band engineering at two-dimensional layered PtSe2/Si heterojunctions for efficient photoelectrochemical hydrogen production. ACS Nano2021, 15, 4627–4635.cChen, M.; Wang, Y.; Shepherd, N.; Huard, C.; Zhou, J.; Guo, L.; Lu, W.; Liang, X. Abnormal multiple charge memory states in exfoliated few-layer WSe2 transistors. ACS Nano2017, 11, 1091–1102.dChang, Y.; Chen, C.; Ho, C.; Cheng, C.; Chen, H.; Fu, T.; Huang, Y.; Ke, S.; Du, H.; Lee, K.; et al. Surface electron accumulation and enhanced hydrogen evolution reaction in MoSe2 basal planes. Nano Energy2021, 84, 105922.eDiao, M.; Li, H.; Sun, Y.; Liang, Y.; Yu, Z.; Boukhvalov, D. W.; Huang, Z.; Zhang, C. Enhancing reverse saturable absorption in SnS2 nanosheets by plasma treatment. ACS Appl. Mater. Interfaces2021, 13, 4211–4219.fChen, M. C.; Li, K. S.; Li, L. J.; Lu, A. Y.; Li, M. Y.; Chang, Y. H.; Lin, C. H.; Chen, Y. J.; Hou, Y. F.; Chen, C. C.; et al. TMD FinFET with 4 nm thin body and back gate control for future low power technology. IEEE International Electron Devices Meeting (IEDM) 2015; p 32–34.gChow, P. K.; Jacobs-Gedrim, R. B.; Gao, J.; Lu, T. M.; Yu, B.; Terrones, H.; Koratkar, N. Defect-induced photoluminescence in monolayer semiconducting transition metal dichalcogenides. ACS Nano2015, 9, 1520–1527.hLi, D.; Ryu, B.; Yoon, J.; Li, Z.; Liang, X. Improvement of analogue switching characteristics of MoS2 memristors through plasma treatment. J. Phys. D: Appl. Phys.2020, 53, 135305.iLin, S.H.; Kuo, J.L. Activating and tuning basal planes of MoO2, MoS2, and MoSe2 for hydrogen evolution reaction. Phys. Chem. Chem. Phys.2015, 17, 29305–29310.jLin, C. P.; Lin, C. T.; Liu, P. S.; Yu, M. J.; Hou, T. H. Grain size and plasma doping effects on CVD-based 2D transition metal dichalcogenide. IEEE 16th International Conference on Nanotechnology, 2016; p 501–504.kSun, L.; Hu, H.; Zhan, D.; Yan, J.; Liu, L.; Teguh, J. S.; Yeow, E. K.; Lee, P. S.; Shen, Z. Plasma modified MoS2 nanoflakes for surface enhanced Raman scattering. Small2014, 10, 1090–1095.lZhang, M.; Tong, Y.; Tang, Q.; Liu, Y. Plasma treatment introduced memory properties in MoS2 field-effect transistors. Appl. Phys. Express2015, 9, 014202.mZhang, P.; Xiang, H.; Tao, L.; Dong, H.; Zhou, Y.; Hu, T. S.; Chen, X.; Liu, S.; Wang, S.; Garaj, S. Chemically activated MoS2 for efficient hydrogen production. Nano Energy2019, 57, 535–541.nHuang, Y.; Lv, J.; Huang, J.; Xu, K.; Liu, L. Nanocrystalline NiSe2/MoS2 heterostructures for electrochemical hydrogen evolution reaction. Nanotechnology2021, 32, 175602.o°Yu, J.; Wang, Y.; Jiang, J.; Liang, Y.; Liu, Y.; Zhong, B.; Quan, S.; Zhou, M.; Ni, Z.; Guo, S. Tunable anisotropy in ReS2 flakes achieved by Ar+ ion bombardment probed by polarized Raman spectroscopy. Appl. Phys. Lett.2021, 119, 053104.pCzerniak-Łosiewicz, K.; Świniarski, M.; Gertych, A. P.; Giza, M.; Maj, Z.; Rogala, M.; Kowalczyk, P. J.; Zdrojek, M. Unraveling the mechanism of the 150-fold photocurrent enhancement in plasma-treated 2D TMDs. ACS Appl. Mater. Interfaces2022, 14, 33984–33992.qWang, J.; Liu, H.; Hu, X.; Liu, Y.; Liu, D. Imaging of defect-accelerated energy transfer in MoS2/hBN/WS2 heterostructures. ACS Appl. Mater. Interfaces2022, 14, 8521–8526. Characteristics and Applications of Defect-Engineered TMDs Chen, M.; Liang, X. Multibit memories fabricated through mechanical and plasma induced deformation of layered semiconductors. IEEE 17th International Conference on Nanotechnology. 2017; p 705–706. Chung, C. C.; Yeh, H.; Wu, P. H.; Lin, C. C.; Li, C. S.; Yeh, T. T.; Chou, Y.; Wei, C. Y.; Wen, C. Y.; Chou, Y. C.; et al. Atomic-layer controlled interfacial band engineering at two-dimensional layered PtSe2/Si heterojunctions for efficient photoelectrochemical hydrogen production. ACS Nano2021, 15, 4627–4635. Chen, M.; Wang, Y.; Shepherd, N.; Huard, C.; Zhou, J.; Guo, L.; Lu, W.; Liang, X. Abnormal multiple charge memory states in exfoliated few-layer WSe2 transistors. ACS Nano2017, 11, 1091–1102. Chang, Y.; Chen, C.; Ho, C.; Cheng, C.; Chen, H.; Fu, T.; Huang, Y.; Ke, S.; Du, H.; Lee, K.; et al. Surface electron accumulation and enhanced hydrogen evolution reaction in MoSe2 basal planes. Nano Energy2021, 84, 105922. Diao, M.; Li, H.; Sun, Y.; Liang, Y.; Yu, Z.; Boukhvalov, D. W.; Huang, Z.; Zhang, C. Enhancing reverse saturable absorption in SnS2 nanosheets by plasma treatment. ACS Appl. Mater. Interfaces2021, 13, 4211–4219. Chen, M. C.; Li, K. S.; Li, L. J.; Lu, A. Y.; Li, M. Y.; Chang, Y. H.; Lin, C. H.; Chen, Y. J.; Hou, Y. F.; Chen, C. C.; et al. TMD FinFET with 4 nm thin body and back gate control for future low power technology. IEEE International Electron Devices Meeting (IEDM) 2015; p 32–34. Chow, P. K.; Jacobs-Gedrim, R. B.; Gao, J.; Lu, T. M.; Yu, B.; Terrones, H.; Koratkar, N. Defect-induced photoluminescence in monolayer semiconducting transition metal dichalcogenides. ACS Nano2015, 9, 1520–1527. Li, D.; Ryu, B.; Yoon, J.; Li, Z.; Liang, X. Improvement of analogue switching characteristics of MoS2 memristors through plasma treatment. J. Phys. D: Appl. Phys.2020, 53, 135305. Lin, S.H.; Kuo, J.L. Activating and tuning basal planes of MoO2, MoS2, and MoSe2 for hydrogen evolution reaction. Phys. Chem. Chem. Phys.2015, 17, 29305–29310. Lin, C. P.; Lin, C. T.; Liu, P. S.; Yu, M. J.; Hou, T. H. Grain size and plasma doping effects on CVD-based 2D transition metal dichalcogenide. IEEE 16th International Conference on Nanotechnology, 2016; p 501–504. Sun, L.; Hu, H.; Zhan, D.; Yan, J.; Liu, L.; Teguh, J. S.; Yeow, E. K.; Lee, P. S.; Shen, Z. Plasma modified MoS2 nanoflakes for surface enhanced Raman scattering. Small2014, 10, 1090–1095. Zhang, M.; Tong, Y.; Tang, Q.; Liu, Y. Plasma treatment introduced memory properties in MoS2 field-effect transistors. Appl. Phys. Express2015, 9, 014202. Zhang, P.; Xiang, H.; Tao, L.; Dong, H.; Zhou, Y.; Hu, T. S.; Chen, X.; Liu, S.; Wang, S.; Garaj, S. Chemically activated MoS2 for efficient hydrogen production. Nano Energy2019, 57, 535–541. Huang, Y.; Lv, J.; Huang, J.; Xu, K.; Liu, L. Nanocrystalline NiSe2/MoS2 heterostructures for electrochemical hydrogen evolution reaction. Nanotechnology2021, 32, 175602. °Yu, J.; Wang, Y.; Jiang, J.; Liang, Y.; Liu, Y.; Zhong, B.; Quan, S.; Zhou, M.; Ni, Z.; Guo, S. Tunable anisotropy in ReS2 flakes achieved by Ar+ ion bombardment probed by polarized Raman spectroscopy. Appl. Phys. Lett.2021, 119, 053104. Czerniak-Łosiewicz, K.; Świniarski, M.; Gertych, A. P.; Giza, M.; Maj, Z.; Rogala, M.; Kowalczyk, P. J.; Zdrojek, M. Unraveling the mechanism of the 150-fold photocurrent enhancement in plasma-treated 2D TMDs. ACS Appl. Mater. Interfaces2022, 14, 33984–33992. Wang, J.; Liu, H.; Hu, X.; Liu, Y.; Liu, D. Imaging of defect-accelerated energy transfer in MoS2/hBN/WS2 heterostructures. ACS Appl. Mater. Interfaces2022, 14, 8521–8526. Hydrogen Evolution Reaction (HER) 2D TMDs have unique physicochemical features that have made them attractive as highly efficient HER catalysts. Plasma treatment creates a higher density of defects and edges in 2D TMDs, leading to enhanced catalytic activity. However, care must be taken because excess amounts of defects may decrease the active sites due to etching and/or overdoping.288,321,339,376,385,400,401 For instance, the optimum values for concentration of chalcogen vacancies produced by Ar plasma for best catalytic activity of 1T-MoTe2 and 2H-MoTe2 are found to be 3.18 × 1014 cm–2 and 1.02 × 1014 cm–2, respectively.402 Theoretical studies have shown the sulfur vacancies to reduce the free energy for hydrogen adsorption (ΔGH) by introducing new localized bands near the Fermi level (Figure 13a,b).403 Applying an in-plane tensile strain can also decrease ΔGH by increasing the number of newly introduced bands near the Fermi level, shifting them toward the Fermi level, and/or narrowing the bandgap (Figure 13c,d). As shown in Figure 13e, the more strain is applied to TMDs, the higher the required sulfur vacancy density for achieving the optimal condition of ΔGH = 0 eV would be.403 The improved adsorption of hydrogen atoms on sulfur vacancies results in a more efficient HER. Additionally, the reduced electrical resistivity of plasma-treated TMD nanosheets such as NiSe2 caused by chalcogen vacancies further enhances their electrocatalytic activity.404 It was also found experimentally that the incorporation of oxygen atoms in the structure of 2D TMDs as well as partial phase transformation from 2H to 1T enhances the HER activity.405−407 Meanwhile, the produced oxide species can be reduced during hydrogen evolution, leaving behind a perforated structure that further enhances the catalytic activity.41 Wang et al.408 have reported that O2 plasma treatment of WTe2 nanosheets could result in an up to 4-fold decrease in the Tafel slope by 10 min plasma dwell time, in which the optimum density of defects for best catalytic activity was attained. Tungsten vacancies and a change in the morphology from nanosheet to feather-like occur after 10 min treatment. Under this condition, the highest catalytic activity is achieved, while further plasma treatment collapses the structure with fewer active sites. Similarly, Chen et al.409 could enhance the catalytic activity of vertically aligned 1T-PtSe2 nanosheets by Ar plasma treatment for HER. The argon ion sputtering resulted in islands of amorphous PtSex with a high density of chalcogen vacancies as well as Pt atoms and clusters (detected by XPS and HR-TEM), enhancing the catalytic activity of the overall 3D-structure. (a-c) Calculated band diagrams of single-layer 2H-MoS2 with (a) 0% and (b) 25% sulfur vacancy, and (c) 25% sulfur vacancy under applied strain. (d) Schematic views of strained single layer 2H-MoS2 with sulfur vacancy: top view (upper part) and side view (lower part). (e) ΔGH versus tensile strain (%) for single layer 2H-MoS2 with various sulfur vacancy densities. Reprinted with permission from ref (403). Copyright 2015 Springer Nature. (f) Schematic view of a MoS2-based memory FET. (g) Schematic of O2 plasma-treated MoS2-based FET, representing a rippled layer and charge-trapping sites. (h) Drain-source current (IDS) versus time as an illustration of retention characteristics. The gate voltage and drain-source voltage are 0 and 0.01 V, respectively. The data were recorded for three days and extrapolated to 10 years. Reprinted with permission from ref (320). Copyright 2014 American Chemical Society. (i) I-V curves of an Ar plasma-treated MoTe2-based memristive device with a compliance current of 3 mA over 100 cycles. (j) Raman maps of the sum of the Raman intensity in the range of 100.4–150.1 cm–1 (indication of the 1T′ phase) at the high resistance state (HRS) and low resistance state (LRS) of the memristive device. (k) Raman maps of the sum of the Raman intensities in the range of 211.2–250.9 cm–1 (indication of the 2H phase) at HRS and LRS. The rectangles show the region of interest where the local phase transitions occurred. Reprinted with permission from ref (410). Copyright 2022 Royal Society of Chemistry. Field-Effect Transistors While the formation of sulfur vacancies increases the electron density, the associated reduction of charge carrier mobility makes the application of defective TMDs for FETs challenging. Utilizing a graphene layer,357 h-BN layer,411 or polymer mask412 to reduce the kinetic energy of the plasma ions can effectively generate chalcogen vacancies and n-dope the structure without degradation of carrier mobility arising from structural defects. Yim et al.412 have shown that plasma treatment using a polydimethylsiloxane (PDMS) mask increases the n-type charge carrier density in a few-layer MoS2-based FET from 1.9 × 1011 to 8.1 × 1011 cm–2. Lee et al.351 have reported that plasma treatment induces surface defects in WSe2 layers. As a result, bioreceptors can be immobilized on the surface more effectively, improving the sensitivity of BioFET devices. Recently, Hou et al.413 have employed a mild Ar plasma treatment to generate sulfur vacancies in WS2 monolayers that reduce the effective mass of holes and the work function, while increasing the mobility. The results are reduced turn-on voltage in bottom-electrode FETs, and enhanced conductivity in both top- and bottom-electrode FETs. However, Lee et al.341 observed a contradictory trend. They found that treating MoS2 monolayers by H2 plasma decreased the conductivity of FET devices which even showed insulating behavior at high plasma times. Such degradations were attributed to a convolution of strain and defects as well as formation of nanocracks (Figure 12f) which act as electron–hole scattering centers. Multibit Memory Devices Recently, Chen et al.320 developed a plasma-treated MoS2-based FET with both binary and multibit data storage capability. They have shown that O2 plasma treatment results in doping and rippling of the upper layer(s) of MoS2 (Figure 13f,g), forming charge-trapping sites. The localized states in the bandgap consist of isoelectronic deep-level traps and acceptor/donor types. The gap between the rippled part (charge-trapping sites) and undoped MoS2 can act as a blocking barrier/tunneling layer to retain trapped charges and switch the memory cell states. The FET device exhibits reliable binary data states to retain data for years and 3-bit data states to store data for several days (Figure 13h). Further studies have revealed that employing a shear deformation method along with the plasma treatment can increase the charge-trapping site density by mechanical deformation and enhancing the memory-cell operation.274 Memristive Devices Recently, TMDs with plasma-induced defects have been proposed as candidates for memristive devices.410,414 For instance, He et al.410 fabricated such a device based on Ar plasma treated MoTe2 which showed controllable electric field-induced phase transitions between semiconducting 2H and metallic 1T′ phases. Their devices had cycling endurance of over 105 and data retention times of over 105 s. Figure 13i shows the I-V curves of the device, demonstrating that the transitions between a low resistance state (LRS) to a high resistance state (HRS) occur at about −1 and 1.1 V. Figures 13j and k show the Raman maps at wavenumber ranges correspond to 1T′ and 2H phases for LRS and HRS. As seen, the intensity of Raman features at wavenumber ranges correspond to the 1T′ phase is considerably higher in LRS state than HRS. The intensity of Raman features at wavenumber ranges corresponding to the 2H phase is also slightly higher in the HRS than LRS. Such changes are visible in a portion of the images which indicates that the reversible phase transformations are local. DFT calculations showed that tellurium vacancies act as nucleation sites for the phase transitions and facilitate the phase transition between 2H and 1T′ phases by reducing the energy barrier of the transition.410 Studies have shown that Ar plasma-treated MoS2 memristive devices have good neuromorphic computing capability which can be used for gesture recognition and artificially intelligent electronic skins.415 Lithium-Ion Batteries (LIBs) TMDs are promising anode materials for LIBs due to their abundance and high theoretical capacity. Plasma treatment further improves the performance of TMD-based anodes for energy storage applications because the generated vacancies and defects (and oxygen incorporation in the case of using O2-containing plasma) can act as active sites for lithium ions.416−418 Moreover, O2 plasma produces oxygen-doped TMDs to form metal–O–C bonds in the anode. These bonds stabilize the structural integrity during the cycling and alleviate the volume expansion.419 Supercapacitors 2D materials are interesting choices for supercapacitors due to their high surface-to-volume ratio, large specific surface area, and well-suited electrolyte ion transport path.420,421 Recently, Chang et al.422 used Ar plasma treatment on lamellar NiSe2-butylamine hybrid precursors to directly exfoliate NiSe2 and induce pores and selenium vacancies. Their supercapacitor based on defected NiSe2 had a specific capacitance of 466 F g–1 in 1 M KOH electrolyte with high cycling stability, which was much higher than those of supercapacitors based on NiSe2 particles or perfect NiSe2 nanosheets. This enhancement was due to an increased specific surface area, a reduced ion-transfer distance, increased active site density, and improved conductivity arising from pores and anion vacancies.423−425 In addition to the aforementioned applications, defective and doped TMD structures have been recently proposed for electrochemical sensing of hydrogen peroxide. Luo et al.426 fabricated a MoSe2-based catalyst containing selenium vacancies and nitrogen dopants which were introduced by Ar and N2 plasmas, respectively. The vacancies and dopants produced new active sites, narrowed the bandgap, and enhanced the reactivity of the MoSe2 surface toward ·OH groups, which are intermediate products in the H2O2 reduction process (H2O2 + 2H+ + 2e– → 2H2O). Their sensor fabricated with defective TMD had a H2O2 detection limit as low as 12.6 nmol/L in a wide pH range of 1–13 with good long-term stability and selectivity. Doping In addition to functionalization of TMDs by physisorption or chemisorption of atoms atop TMD surfaces (discussed in section 4), foreign atoms can dope the TMD structure by replacing metal or chalcogen atoms.208,427,428 In addition to conventional solution-based processes,429−431 plasma treatment is a promising technique to modify the electronic structure of 2D TMDs through either in situ or ex-situ doping, as schematically shown in Figures 14a and b. In-situ doping approaches rely on introducing the adatoms to the structure during the fabrication process without the need for any extra steps. Although adatoms may become well-distributed through the structure, the doping processes and mechanisms are not controllable and well-understood.98,432 In contrast, ex-situ processes rely on separate steps to modify the surface and to insert the adatoms into the structure.433 In this section, the plasma doping processes and their impacts on the electronic structure and chemical compositions of 2D TMDs are reviewed. Schematic representations of (a) in situ and (b) ex-situ doping by plasma. (c) Raman spectra of MoS2 after O2 plasma treatments at different conditions. Sample 1 is a pristine MoS2 monolayer. Samples 2, 3, and 4 are treated with O2 plasma for 10, 20, and 40 s. Part (c) reprinted with permission from ref (434). Copyright 2019 IOP Publishing. (d) Drain-source current (IDS) versus drain-source voltage (VDS) at zero gate voltage (VG) for back-gated MoS2 diodes with Ti/Au contacts treated by an O2-containing plasma. The inset shows the schematic of a fabricated diode. IForward/IReverse was 240 at VDS = 1 V and VG = 0. Reprinted with permission from ref (435). Copyright 2013 AIP Publishing. (e and f) XPS spectra of (e) Mo 3d, S 2s, and (f) S 2p core levels for as-exfoliated and O2 plasma treated MoS2 for 300, 600, and 1200 s. Reprinted with permission from ref (248). Copyright 2017 American Chemical Society. Ex-Situ Doping Introducing foreign atoms into the crystal structure by employing additional steps after the synthesis is regarded as “ex-situ doping”. The foreign atoms come from an external environment (here plasma) that alters the surface chemical composition. As the plasma environment is well-controlled (time, frequency, power, etc., are controllable), surface modification can be done accurately by this treatment. A number of different plasma treatments have been utilized to modify TMDs, as briefly explained below. Oxygen As the most dominant heteroatoms found in TMDs, O atoms can be substituted at different positions of a lattice plane such as vacancies, defects, edges, or lattice sites.436 During O2 plasma treatment of TMDs, oxygen atoms can substitute sulfur vacancies to form Mo-O dangling bonds. The sulfur vacancies either exist in the TMD structure437 or are generated during the plasma treatment by bombarding high-energy plasma species. It has been shown that RF oxygen plasma treatment of MoS2, which bombards the surface with oxygen radicals, leads to the incorporation of additional oxygen atoms into the MoS2 structure, resulting in the formation of MoS2-xOx.419 Similar to the case of functionalization by oxygen atoms, the characteristic Raman peaks of oxygen-doped MoS2 monolayers are also red-shifted (Figure 14c).434 The plasma treatment for prolonged times (>40 s) breaks Mo-S bonds on the surface, causing severe oxidation and MoO3 formation. The incorporation of oxygen into 2D TMDs induces a p-type behavior arising from (i) substitution of sulfur by oxygen atom, (ii) charge transfer between the TMD layer and possible formed oxide(s), and (iii) charge transfer between the TMD layer and physisorbed oxygen molecules.438,439 Theoretical calculations have shown that the excess negative charge on oxygen can reach up to 0.903 electrons/dopant atom, indicating p-type doping by oxygen incorporation.440 By employing selected area plasma treatment, the fabrication of FETs441 and back-gated diodes with quasi-Ohmic behavior (Figure 14d), high forward/reverse current ratios and long-term stability at ambient conditions are feasible.435 It has also been shown that the Schottky barrier height (SBH) for holes at drain and source can be modulated by O2 plasma treatment of MoS2-based FETs through oxygen doping. For instance, an O2 plasma treatment broadens the SBH from 0.2 to 0.3 eV (for pristine MoS2) to 0.2–0.8 eV in a MoS2-based FET with Ni/Au contacts. The different oxidation states of molybdenum and the S-O characteristic peak in the XPS spectra (Figures 14e and f) affirm oxygen doping. Meanwhile, the risk of surface oxidation is higher when prolonged treatment is employed.248 A toroidal-magnetic-field (TMF) controlled O2 plasma can be used to decrease the lattice defect density and increase the substitutional ratio because the process significantly increases the concentration of the plasma species to yield a higher oxygen concentration in the lattice.442 In FETs based on pristine TMDs, electrons are the dominant carrier type because of sulfur vacancies and strong Fermi level pinning near the conduction band in metal/TMDs contacts.443 O2 plasma treatment results in partial substitution of sulfur vacancies by oxygen ions; hence, the electron mobility is reduced.444 First-principles calculations determine that increasing the ratio of O2+ ions in the plasma generates shallow acceptor levels (at 42.16 meV) above the valance band maximum and extracts one electron per 4 × 4 × 1 MoS2 lattice with one dopant.442 As a result, the hole mobility and the Ion/Ioff ratio are enhanced due to the filling of the sulfur vacancies by oxygen adatoms. Nevertheless, as mentioned in section 5, plasma treatment can induce a high density of structural defects in the TMD structure, which may degrade its electrical properties. In such cases, surface protection strategies by using self-assembled nanostructures (surface shielding) can be employed to preserve the carrier mobility of 2D TMDs by minimizing structural defects during plasma doping. It has been shown that these techniques are able to control the doping concentration between 2 × 1011 and 8.1 × 1011 cm–2 depending on the type of self-assembled nanostructures and the plasma chemistry.412 Plasma oxygen doping can result in new properties in 2D TMDs or improve their performance in emerging applications. It has been reported that nonradiative transitions and radiative neutral excitons of single layer MoS2 enhance the PL emission. Plasma-treated single layer MoS2@TiO2 core/shell structures exhibit a strong PL enhancement based on reduced chalcogen vacancies, oxygen bonding, and p-doping in the MoS2 lattice,445 which suppresses trion emission and increases neutral exciton emissions (Figure 15a). A reduction in nonradiative recombination of excitons at defect sites (vacancies), which are filled by oxygen atoms, also enhances PL properties.264,446,447 On the other hand, various studies have shown that oxygen doping of TMDs improves their performances in energy conversion applications.376 For instance, enhanced HER kinetics on the surface of MoSe2448 and MoS2449 have been reported. The improved performance arises from defect formation and free energy reduction for the hydrogen adsorption.449 As shown in Figure 15b, the free energy of hydrogen adsorption on pristine MoS2 decreases from 2.18 eV to −0.01 eV for MoS1.96O0.04 monolayers with sulfur vacancies. The enhanced charge transfer also enhances the catalytic activity.450 In organic solar cells (OSCs), improved performance through oxygen doping into the MoS2 structure and changing of the electronic structure has been reported.451 Herein, filling the vacancies by oxygen adatoms reduces deep-donor-defect states in the bandgap (less recombination), making the 2D materials excellent alternative hole transporting layers (HTL) for highly efficient OSCs (Figure 15c). (a) Effect of O2 plasma exposure time on the PL intensity of monolayer MoS2. Reprinted with permission from ref (445). Copyright 2018 American Chemical Society. (b) Gibbs free energy of H atom adsorbed on O atom (O s) and MoS2 (1- x)O2x with sulfur vacancy. Reprinted with permission from ref (449). Copyright 2018 Elsevier. (c) Energy diagram of PTB7:PC71BM sandwiched between ITO (anode) and Al (cathode) in the presence of MoS2 and oxygen-doped MoS2 (as the hole extraction layer). Reprinted with permission from ref (451). Copyright 2014 Royal Society of Chemistry. (d) N 1s XPS spectrum of MoS2 (1) without plasma treatment and after treatment at (2) 30 W and (3) 50 W. Reprinted with permission from ref (452). Copyright 2016 AIP Publishing. (e-f) STEM images show the atomic structure of WS2 (e) without and (f) with employing a N2 plasma treatment. The insets exhibit a schematic of atomic arrangement (sulfur: yellow; tungsten: brown) in the lattice. A selected area electron diffraction (SAED) pattern is shown in the inset of (f). Reprinted with permission from ref (453). Copyright 2019 John Wiley and Sons. Theoretical studies have also shown that oxygen doping by plasma (substituting sulfurs and chemisorbed atop the MoS2 basal plane) can induce ferromagnetism to the MoS2 layers due to overlap between the density of states of oxygen and sulfur as well as oxygen and molybdenum. Such enhancement was experimentally observed by Xie et al.454 They found that a 40 min O2 plasma treatment of MoS2 nanostructures enhanced their weak ferromagnetism with saturated ferromagnetic magnetization of 0.0012 to 0.0343 emu/g at a temperature of 300 K. Nitrogen Nitrogen doping by plasma has been shown to change the electrical properties and band structure of TMDs.452,455 For instance, nitrogen plasma results in p-doping of MoS2 and shifts the Fermi level toward the valence band, increasing the work function of MoS2.456,457 A different study has reported, however, that N2 plasma treatment decreases the work function of MoS2 by reducing the hole concentration. XPS studies have shown the formation of Mo-N bonds notably intensified at higher nitrogen content due to higher input energy and/or prolonged treatment time (Figure 15d). The Mo 3p and 3d peaks as well as the S 2p XPS peaks in those samples are shifted to higher binding energies due to a shift of the Fermi level toward the conduction band.452 N2 plasma treatment of WS2 has been found to repair its sulfur vacancies (Figures 15e and f),453 resulting in an enhanced mobility in nitrogen-doped WS2 FET devices (up to 184.2 cm2 V–1 S–1). Nitrogen doping decreases the electron effective mass; thereby, the carrier mobility of WS2 is improved. Similarly, the covalent bond formation between tungsten and nitrogen after plasma treatment in a nitrogen atmosphere has been shown.458 N2 plasma treatment can also be applied on TMDs with metastable phases (such as 1T MoS2) to dope nitrogen without any phase change. In this situation, a low power and preferably room temperature should be maintained during the plasma treatment to avoid 1T to 2H structural transitions. Such a process can be used to manipulate electronic structures of 1T/2H TMD heterostructures to obtain enhanced catalysts or supercapacitors.459 DFT calculations have revealed that both N+ and N2+ ions forming during N2 plasma treatment can substitute chalcogen atoms in TMDs due to the relatively large negative formation energies of the respective doped structures. However, the formation energies of the sulfur vacancies formed by N+ and N2+ ions in MoS2 have negative and positive values, respectively. Hence N+ is more likely to induce damage to the MoS2 structure. Recently, Zeng et al.460 used TMF modulated N2 plasma to increase the N2+ concentration at low RF power. XPS spectra show the incorporation of N2+ ions within the MoS2 structure with low structural damage. Increasing the RF power of plasma treatment increased the N2+ concentration resulting in more structural damage. Ammonia can be added to N2 plasma to manipulate the chemical composition of TMDs during plasma treatment. For instance, recent studies have shown that the nitrogen content in WSe2 can be elevated up to 5.7 at% by using NH3 plasma treatment.461 Tan et al.276 studied the impact of NH3 plasma process conditions on the electronic behavior of TMD monolayers. A relatively low RF power and short dwell times were found to result in p-doping of MoS2 due to sulfur vacancies being filled by N radicals such as ·NH2, ·NH, and ·N. However, using relatively high plasma power and long duration promotes sulfur vacancy formation and results in n-doping. Similar n-doping was also observed for WSe2 treated with NH3 plasma due to the formation of selenium vacancies introduced by hydrogen species in the ammonia structure, as evidenced by a blueshift in the W 4f and Se 3d XPS spectra.461 Other Heteroatoms Phosphorus, fluorine, chlorine, and carbon have also been introduced as dopant atoms through the plasma treatment of TMDs. The plasma immersion ion implantation (PIII) technique was employed by Neopeal et al.462 to study the feasibility of phosphorus doping of MoS2. Their DFT simulations showed the appearance of DOS at the Fermi level (Ef) which affirmed p-doping. Further electrical measurements also proved the p-doping by incorporation of P atoms. Herein, three possibilities arise (Figure 16a-c): (i) P substitutional doping either by replacing sulfur or occupying an sulfur vacancy; (ii) P-adatom adsorbed on the MoS2 surface; (iii) PH3, PH2+, and PH2+ ions adsorbed on the MoS2 surface.463 The controllability of the PIII technique provides an opportunity for different doping concentrations that are important for device manufacturing. For instance, the authors have demonstrated that the well-controlled p-doping improves the rectification behavior and broadens the working voltage range to ±3 V. Recently, Barbosa et al.464 introduced PH3 to the Ar plasma treatment of NiSe2 to induce selenium vacancies and phosphorus dopants simultaneously. The resulting water-splitting electrocatalyst exhibits a current density of 10 mA/cm2 at 1.62 V. Adding PH3 to Ar plasma results in the generation of PHx (x = 0–3) radicals in the plasma465,466 which provide P atoms for the doping of TMD layers. DOS of MoS2 monolayer with (a) P-substituted, (b) P-adatom, and (c) PH3-adsorbed. For each part, the top panel represents the unimplanted monolayer (1H) MoS2 for comparison. The middle and bottom panels of (a) to (c) depict DOS without and with sulfur vacancies, respectively. Reprinted with permission from ref (463). Copyright 2016 American Chemical Society. (d) IDS-VG characteristics of back-gated MoS2 FETs before and after plasma treatment with different species. For all curves, drain-source voltage is 5 V. Reprinted with permission from ref (435). Copyright 2013 AIP Publishing. (e) Schematic band diagram of a CHF3 plasma-treated MoS2 photovoltaic device with vertically stacked Au/plasma-treated (p-doped) MoS2/untreated (n-type) MoS2/ITO regions. Reprinted with permission from ref (467). Copyright 2014 American Chemical Society. Studies have shown that MoS2 diodes treated with a F-containing plasma (CHF3, CF4, or SF6) exhibit more extreme rectification properties due to doping with fluorine (Figure 16d). A downshift was found in the XPS Mo 3d5/2 and Mo 3d3/2 peaks toward lower binding energies, indicating a Fermi level shift toward the valence band edge, i.e., p-doping.435 Therefore, the process provides a unique opportunity to manufacture reliable p-n junctions on multilayers MoS2 with high photovoltaic responses. As schematically shown in Figure 16e, the band diagram of vertically stacked ITO/untreated n-doped MoS2 over fluorine-treated p-doped MoS2/Au forms a p-n junction with a built-in potential of 0.7 eV that effectively facilitates the separation of photogenerated electrons and holes.467 The rectification behavior of n-type ZnO-MoS2 diodes has also been enhanced by F-containing plasma treatment using SF6. The p-n junction shows a high rectification ratio of 3.4 × 104 and a high saturation current area up to 1.5 V. Such p-n junctions can also act as a piezoelectric semiconductor in photodetectors. It is possible to employ the piezophototronic effect to enhance photodection in these devices by applying an external pressure.468 This effect arises from using the inner-crystal piezoelectric potential (piezopotential) in piezoelectric semiconductors as a gate voltage to control and tune charge carrier transport, recombination, or separation.469 It has been shown that fluorine doping of 2D MoS2 by CHF3 plasma treatment induces compressive strain and lowers the contact resistance as well as field effect mobility, resulting in increased OFF-state current and reduced ON-state current in MoS2-based FETs. Such electrical changes are attributed to induced shallow donor traps in MoS2 through occupying sulfur vacancy sites or interstitial sites near such vacancies by fluorine atoms.470 Fluorine doping by plasma also enhances the catalytic activity of TMDs such as 1T-WTe2. It has been shown that fluorine doping by SF6 plasma manipulates surface charge and hydrogen adsorption performance, resulting in better catalytic activity for HER. The smaller radius of fluorine in comparison to sulfur provides an empty coordination environment, facilitating hydrogen adsorption. The stronger electronegativity of fluorine compared to sulfur also enhances hydrogen adsorption. As a result, the free energy of hydrogen adsorption (ΔGH*) at tungsten sites near to doped fluorine atoms can be as low as 0.18 eV in comparison to ∼1 V in pristine 1T-WTe2. This can reduce the overpotential of the WTe2 monolayer from 0.45 to 0.27 V at a current density of −10 mA cm–2.471 Similarly, Zhang et al.472 decreased the overpotential of MoS2 at current density of −10 mA cm–2 from 0.298 to 0.267 V by simultaneous fluorine doping at edges and creation of etch holes via CHF3 plasma. Carbon doping of TMDs by plasma has been performed to reduce the bandgap of monolayer semiconductors. For instance, while undoped WS2 exhibits unipolar n-type conduction, carbon-doped WS2 through hydrocarbon-containing (such as CH4) plasma treatment is a p-type semiconductor.473 It is believed that carbon doping alters the WS2 electronic structure and redistributes the resonances of electron–phonon and/or hole–phonon–mediated processes. This is inferred by decreasing the intensity of the second-order double-resonance peak involving two longitudinal acoustic phonons [2LA(M) mode] when the CH4 flow is increased. In this case, C, CH, and CH2 radicals from the plasma can act as anion substitutions for chalcogens within the 2D materials. Interestingly, DFT calculations have shown that the most stable structure is MoS2 doped with CH groups rather than carbon atoms.473 ICP has recently been employed to introduce Cl into the MoS2 structure.474 Chlorine atoms reside at sulfur vacancies or substitute sulfur atoms, shift the Fermi-level toward the conduction band minimum, decrease the work function, increase charge carrier density, and intensify the n-type electric behavior of 2D MoS2 sheets. A prolonged plasma time up to a critical duration increases the Cl concentration and the electrical conductivity. However, long plasma treatment can cause structural defects or etch the material with an increase in resistivity.474−476 Kim et al.474 have shown that electron mobility in MoS2-based FETs is enhanced by chlorine doping, with an enhanced photosensitivity in the near-infrared (785 nm) and green (520 nm) regions by factors of 8.59 and 1.94 times, respectively. Recently, H2 plasma has been used for the synthesis of Janus TMD monolayers.477 Trivedi et al.478 used a home-build ICP setup to first remove the top selenium layer from CVD-grown WSe2 or MoSe2 monolayers by H2 plasma. H2 molecules in the plasma were ionized and dissociated into hydrogen radicals and ions. First, hydrogen radicals adsorbed onto the monolayer surface and formed weak bonds with the top chalcogen atoms, resulting in intermediate species. Second, hydrogen ions with high energy bombarded the surface and remove the intermediate species, resulting in formation of HSe+ and H2Se byproducts as well as selenium vacancies. The distance of the monolayer from the plasma tail was adjusted to avoid penetration of hydrogen radicals into the metal layer and formation of vacancies in the bottom selenium layer. Lastly, sulfur atoms of the sulfur precursor located upstream from the sample reacted with the layer and formed WSSe and MoSSe Janus monolayers. Medium annular dark field STEM images clearly showed the position of the chalcogen atoms and the formation of the Janus layers. Another way to produce Janus TMD monolayers is to replace top chalcogen atoms with hydrogen atoms (hydrogen doping) through H2 plasma treatment and then replace the hydrogen atoms by another chalcogen through thermal selenization or sulfurization process. H2 plasma treatment is more likely to preserve underlying structures than other plasmas such as oxygen plasma.479 Successful formation of a Janus MoSH monolayer as an intermediate phase of such reactions and its properties has been recently observed by Wan et al.480 after a controlled remote H2 plasma treatment of a MoS2 monolayer. Their time-of-flight secondary ion mass spectrometry data clearly showed uniform distribution of H ions on only the MoSH top surface. Quantitative analysis of TEM intensity profiles from MoSH also showed intensity drops at sulfur sites similar to sulfur vacancies on the top surface of MoS2 since incident electron beams can easily transmit through H atoms due to their low mass. Combined experimental studies and DFT calculations revealed that Janus MoSH monolayers are metallic in nature with a Fermi level of ∼ −4.11 eV and a high carrier concentration of 2 × 1013 cm–2 which is several orders of magnitude larger than the carrier concentration of intrinsic TMDs. It is estimated that such a Janus monolayer can exhibit a surface plasmon resonance effect in the mid infrared range (>4 μm).480 However, studies have shown that if both hydrogenation and chalcogenization happen at a same time (applying H2 gas and S/Se powders in plasma treatment of TMD monolayer), the intermediate phase is rather a partially substituted Janus monolayer (without H) than a complete MoSH monolayer. In such cases, substitution of top chalcogen atoms in the TMD monolayer by another chalcogen atom is an ongoing process, which increases with longer plasma time until the formation of the Janus layer is complete.481 In-Situ Doping Efficient doping of TMDs can be performed through chemical reactions of precursors during plasma-assisted synthesis processes such as solution-based plasma methods.482,483 Generally, short-lived reactive species including hydroxyl radical (·OH), superoxide (O2–), and hydrogen radicals (·H) as well as long-lived reactive species including ozone and hydrogen peroxide which are formed in aqueous solutions react with the 2D TMDs. Such species are formed when the electrical discharge is carried in the water through the submerged electrodes via mass transfer, vaporization, sputtering, molecular collisions, and excitation/ionization by UV.484 Low-temperature nitrogen doping of MoSe2, WSe2, MoS2, and WS2 with short dwell times has recently been performed in an asymmetric electrical field by electrochemical reactions at the surface of a submerged cathode tip.485 In this process, bulk MX2 is dispersed in an electrochemical cell containing cathode (W), anode (Pt), and electrolyte (such as aqueous NH4 (OH), NH4NO3, or (NH4)2SO4). By applying a voltage, plasma gas is formed around the cathode, which then transfers its energy to the electrolyte via high energy gaseous jets. These jets attack the edges of bulk TMD and disturb the out-of-plane van der Waals attraction forces, leading to exfoliation.485 Moreover, formation of oxygen as well as ·OH and ·H radicals according to eqs 15–17 facilitates exfoliation of the TMDs by penetrating between layers.77 To further enhance exfoliation efficiency, exfoliating agents containing electrolytes (such as KOH) can be used. In this case, plasma generates exfoliating ions (such as K+) which intercalate between TMD layers. Thermally induced stresses generated during the plasma treatment also aid the exfoliation process.486151617Simultaneously, plasma generates NH3+ and NH4+ ions as well as ·NH and ·NH2 radicals which are doping agents for TMD nitrogen doping.77,485 It has been shown that nitrogen doped TMDs fabricated by these methods are suitable catalysts for HER.77,485 For instance, a nitrogen-doped MoS2 exhibited a lower overpotential (164 mV at a current density of 10 mA cm–2) and Tafel slope (71 mV dec–1) compared with exfoliated MoS2 nanosheets and bulk MoS2.485 Zhao et al.228 have synthesized potassium-doped MoS2 by treating Mo-based precursors with an H2S-containing nonthermal plasma using a mixture of manganese nitrate, aluminum nitrate, citric acid, polyethylene glycol, and potassium nitrate as precursors. In another report, niobium-doped MoSe2 has been prepared by plasma-enhanced atomic layer deposition.487 It has been shown that niobium doping gradually improves the gas sensitivity and stability of MoSe2 up to 8% at 3 ppm NO2 concentration. In another study conducted by Cho et al.488 nitrogen-doped MoS2 and WS2 thin films were synthesized by in situ doping during plasma assisted selenization of already-deposited transition metal films. It was found that the addition of Ar gas to a N2/H2S gas mixture enhances the nitrogen doping up to 8.3 and 9.43 at% in WS2 and MoS2, respectively. The Penning excitation effect of Ar facilitated the formation of N2+ and ·N2, which further increased the incorporation of N into the synthesized film. Optical emission spectroscopy of the plasmas showed that N2/H2S plasma had small amount of ·N2 without N2+, while their concentrations in an Ar/N2/H2S plasma was considerable. N2 plasma was utilized as a post-treatment method for nitrogen doping, resulting in nitrogen concentrations in the film below 3 at% due to very low amounts of the aforementioned reactive species. Recently, PECVD has been utilized to synthesize carbon-intercalated MoSe2 nanosheets by using Mo(CO)6 and selenium as precursors. Incorporation of carbon atoms between the MoSe2 layers increases the interlayer spacing. Increasing the PECVD temperature reduces the carbon content, decreases interlayer spacing, and bends the layers, resulting in a morphological transformation from 2D flat layers to hollow spherical nanoparticles. Carbon-intercalated MoSe2 nanosheets with a 2D flat morphology and 10.5 wt% carbon content have shown superior electrochemical performance as cathodes in Na-ion capacitors. These electrodes can deliver a capacity of 227.8 mA h g–1 at 5 A g–1 current density and maintain a reversible capacity of 199.8 mA h g–1 after 1000 cycles at 1 A g–1. Enhanced electrochemical performance is due to the provision of fast diffusion channels for Na+, enhanced electron transfer, faster reaction kinetics, and reduced volume changes of MoSe2 during the capacitor operation.489 Recent reports on plasma doping of TMDs and associated devices are summarized in Table 9. Table 9Summary of Plasma Doping of 2D TMDs and Their Effects on the Performance of DevicesDopantTMDsElectronic dopingConditionsDeviceDevice performanceOxygenMoS2p-typePressure: 0.3 PaFETAmbipolar current transportICP Power: 100 WIon/Ioff ratio: ∼103 andTime: 600 and 1200 sField-effect mobility: 11.5 cm2V–1 s–1 (for electrons) and 7.2 cm2 V–1 s–1 (for holes)248Ar/O2: 20/80OxygenMoS2p-typeToroidal-magnetic-field controlled O2 plasmaFETHole mobility: 115.2 cm2 V–1 s–1Pressure: 12 PaSubthreshold swing: 137 mVdec–1 at room temperature442Power: 10 WFrequency: 13.56 MHzTime: 30 sO2 flow: 60 sccmOxygenMoS2p-typeMild O2 plasmaOptoelectronic devicesStronger PL264Pressure: 5 PaPower: 5 WFrequency: 13.56 MHzOxygenMoS2-Remote O2 plasmaThin-film electrodeImproved contact resistanceaOxygenMoS2p-typeUV/O3 treatmentPhotodetectorSignificant improvement in the visible light responsebTime: 40 minOxygenMoS2-UV/O3 treatmentOrganic solar cellImproved performance including open-circuit voltage, fill factor, power conversion efficiency, and series resistance451Time: 3 minOxygenMoS2p-typeO2 plasmaComplementary metal–oxide–semiconductor inverterPeak voltage gain: 7.48Pressure: 5 PaMaximum static power consumption: 37.7 nW (At supply voltage of 5 V)cPower: 5 WFrequency: 13.56 MHzO2 flow: 60 sccmNitrogenMoS2p-typeRemote N2 PlasmaFETA positive shift in the threshold voltagePressure: 700 PaImprovement in the compressive strain by 1.3%dPower: 100 WFrequency: 13.56 MHzTemperature: 300 °CTime: 2–60 minN2 Flow: 45 sccmNitrogenMoS2n-typeN2 plasmaAg nanoparticles decorated with doped TMDs for sensing of polycyclic aromatic hydrocarbonsEnhanced sensing responseeWS2Pressure: 0.4 PaPower: 40 WFlow: Ar/N2 (2:18 sccm)NitrogenMoSe2p- typeN2 plasma--RF Power: 100 WFrequency: 13.56 MHzTime: 10–60 minTemperature: 200 °CN2 flow: 0.052 sccmfNitrogenWS2p- typeRF N2 plasma--Power density: 7 W/cm2Time: 1–4 sN2 flow: 15 sccmgNitrogenMoS2p- typeToroidal-magnetic-field controlled N2 plasmap-n junctionhRectifying ratio: 103Pressure: 7 PaRF Power: 14 WFrequency: 13.56 MHzFlow: 60 sccmNitrogenMoS2p-typeICP N2 plasmap-n junctionRectifying ratio: 103Pressure: 5–6 PaOptical responsivity: 48.5 A W–1RF Power: 5 WQuantum efficiency: 11 301% (532 nm excitation laser)iFrequency: 13.56 MHzTime: 30 SN2 flow: 60 sccmNitrogenMoS2p-typeSoft ICP NH3 plasma (CCEP)FET276Mobility: 16.52 cm2 V–1 s–1Pressure: 10 PaOff-state current: ∼10–8RF Power: 10 WTime: 1 minFrequency: 2 MHzFlow: 10 sccmNitrogenWS2-N2 PlasmaFETElectron mobility: 184.2 cm2 V–1 s–1Time: 10 minThreshold voltage: 3.8 V453NitrogenMoS2-N2 plasmaN/Ag@MoS2 nanozymes for the colorimetric detection of H2S in aquaculture water.Enhancing Ag deposition, environmental stability and catalytic activity of nanozymejPressure: 100 PaRF Power: 45 WFrequency: 13.56 MHz,NitrogenMoS2-N2 plasmaNitrogen-doped interlayer expanded MoS2/graphene composite for lithium ion batterieskIncreasing the electric conductivity. resulting in discharge capacity of 635.77 mA hg–1 over 400 cycles at current density of 500 mA g–1Pressure: 10 PaPower: 100 WN2 flow: 100 sccmNitrogenMoS2p-typeN2 plasmaPeroxidase enzyme-like catalystLowered Fermi level and enhanced wettability of MoS2 by nitrogen doping resulted in better catalytic activitylNitrogenMoS2p-typeToroidal-magnetic-field controlled N2 plasmaJunction field effect transistormElectron mobility: 7.75 cm2 V–1 s–1Pressure: 5.4 PaIon/Ioff ratio: ∼104Power: 13 WFrequency: 13.56 MHzFluorineMoS2p- typeReactive ion etching with CHF3, CF4 or SF6p-n junctionRectifying ratio: 104 (SF6-treated MoS2)435Pressure: 1.33 PaPower: 100 WFluorineMoS2p-typeSF6, CF4, or CHF3 plasmaPhotovoltaic devicesRectifying ratio: 104Pressure: 1.33 PaPower-conversion efficiency: 2.8%RF Power: 100 WExternal quantum efficiency: 37–78% (300 to 700 nm excitation laser)467Time: 1 minFlow: 10 sccmFluorineMoS2p-typeCHF3 PlasmaFETnThe field-effect mobility: 65.2 cm2 V–1 s –1Ion/Ioff ratio: ∼103FluorineMoS2p-typeSF6 plasmap-n junctionRectification ratio: 3.4 × 104Pressure: 2 PaReverse saturation area: 1.5 VRF power: 80 WExternal quantum efficiency: 52.7% at −2 VTime: 60 sIncreased photocurrent under 23 MPa pressure up to 4 times468PhosphorusMoS2p-typePlasma immersion ion implantation (PH3)FET p-n junctionIncreasing the threshold voltage (for FET)Rectification ratio: 50462PhosphorusMoS2p-typePlasma immersion ion implantation (PH3/He)p-n junction463Rectification ratio: 2 × 104Pressure: 11 PaRF Power: 1000 WDC bias source: 0–5 kVCarbonWS2p-typeCH4 plasmaFET473Electron mobility: ∼0.1 to 1 cm2 V–1 s–1Pressure: 133.32 PaRF Frequency: 13.56 MHzTemperature:400 °CTime: 10 minCH4 flow: 2–8 sccmaLu, Q.; Fang, C.; Liu, Y.; Shao, Y.; Han, G.; Zhang, J.; Hao, Y. Temperature dependence of annealing on the contact resistance of MoS2 with graphene electrodes observed. Jpn. J. Appl. Phys.2018, 57, 04FP07.bYe, J.; Li, X.; Zhao, J.; Mei, X.; Li, Q. A facile way to fabricate high-performance solution-processed n-MoS2/p-MoS2 bilayer photodetectors. Nanoscale Res. Lett.2015, 10, 1–7.cWang, S.; Zeng, X.; Zhou, Y.; Lu, J.; Hu, Y.; Wang, W.; Wang, J.; Xiao, Y.; Wang, X.; Chen, D.; et al. High-performance MoS2 complementary inverter prepared by oxygen plasma doping. ACS Appl. Electron. Mater.2022, 4, 955- 963.dAzcatl, A.; Qin, X.; Prakash, A.; Zhang, C.; Cheng, L.; Wang, Q.; Lu, N.; Kim, M. J.; Kim, J.; Cho, K.; et al. Covalent nitrogen doping and compressive strain in MoS2 by remote N2 plasma exposure. Nano Lett.2016, 16, 5437–5443.eKoklioti, M. A.; Bittencourt, C.; Noirfalise, X.; Saucedo-Orozco, I.; Quintana, M.; Tagmatarchis, N. Nitrogen-doped silver-nanoparticle-decorated transition-metal dichalcogenides as surface-enhanced Raman scattering substrates for sensing polycyclic aromatic hydrocarbons. ACS Appl. Nano Mater.2018, 1, 3625–3635.fXia, Y.; Wang, B.; Zhang, J.; Feng, Y.; Li, B.; Ren, X.; Tian, H.; Xu, J.; Ho, W.; Xu, H.; et al. Hole doping in epitaxial MoSe2 monolayer by nitrogen plasma treatment. 2D Mater.2018, 5, 041005.gdo Nascimento Barbosa, A.; Mendoza, C. A. D.; Figueroa, N. J. S.; Terrones, M.; Júnior, F. L. F. Luminescence enhancement and Raman characterization of defects in WS2 monolayers treated with low-power N2 plasma. Appl. Surf. Sci.2021, 535, 147685.hJin, W.; Zeng, X.; Guo, z.; Zeng, y.; Wang, w.; Zeng, y.; Hu, y.; Ziao, y.; Lu, j.; Lu, j.; et al. Optoelectronic properties of lateral MoS2 p-n Homojunction Implemented by selective p-type doping using nitrogen plasma. J. Phys. D: Appl. Phys. 2020, 53, 405102.iLu, J.; Guo, Z.; Wang, W.; Lu, J.; Hu, Y.; Wang, J.; Xiao, Y.; Wang, X.; Wang, S.; Zhou, Y.; et al. Lateral monolayer MoS2 homojunction devices prepared by nitrogen plasma doping. Nanotechnology2020, 32, 015701.jZhang, S.; Li, P.; Chu, S.; Feng, L.; Li, S.; Fan, J.; Xie, S.; Zhang, Y.; Mao, G.; Wang, H. Nitrogen plasma-mediated deposition of silver onto MoS2 toward robust nanozyme with enhanced catalysis for colorimetric assay of hydrogen sulfide in aquaculture water. Appl. Surf. Sci.2022, 153686.kTang, P.; Jiao, J.; Fan, Q.; Wang, X.; Agrawal, V.; Xu, Q. Interlayer spacing engineering in N doped MoS2 for efficient lithium ion storage. Mater. Chem. Phys.2021, 261, 124166.lFeng, L.; Zhang, L.; Zhang, S.; Chen, X.; Li, P.; Gao, Y.; Xie, S.; Zhang, A.; Wang, H. Plasma-assisted controllable doping of nitrogen into MoS2 nanosheets as efficient nanozymes with enhanced peroxidase-like catalysis activity. ACS Appl. Mater. Interfaces2020, 12, 17547–17556.mWang, W.; Lu, J.; Wan, D.; Zeng, X.; Lu, J.; Xu, T.; Chen, C.; Zhang, T. Improved performance in MoS2 homogeneous junction field effect transistors by optimizing electrodes contact. Mater. Sci. Eng. B2023, 290, 116348.nZhang, X. M.; Tseng, S. H.; Lu, M. Y. Large-area ultraviolet photodetectors based on p-type multilayer MoS2 enabled by plasma doping. Appl. Sci.2019, 9, 1110. Summary of Plasma Doping of 2D TMDs and Their Effects on the Performance of Devices Lu, Q.; Fang, C.; Liu, Y.; Shao, Y.; Han, G.; Zhang, J.; Hao, Y. Temperature dependence of annealing on the contact resistance of MoS2 with graphene electrodes observed. Jpn. J. Appl. Phys.2018, 57, 04FP07. Ye, J.; Li, X.; Zhao, J.; Mei, X.; Li, Q. A facile way to fabricate high-performance solution-processed n-MoS2/p-MoS2 bilayer photodetectors. Nanoscale Res. Lett.2015, 10, 1–7. Wang, S.; Zeng, X.; Zhou, Y.; Lu, J.; Hu, Y.; Wang, W.; Wang, J.; Xiao, Y.; Wang, X.; Chen, D.; et al. High-performance MoS2 complementary inverter prepared by oxygen plasma doping. ACS Appl. Electron. Mater.2022, 4, 955- 963. Azcatl, A.; Qin, X.; Prakash, A.; Zhang, C.; Cheng, L.; Wang, Q.; Lu, N.; Kim, M. J.; Kim, J.; Cho, K.; et al. Covalent nitrogen doping and compressive strain in MoS2 by remote N2 plasma exposure. Nano Lett.2016, 16, 5437–5443. Koklioti, M. A.; Bittencourt, C.; Noirfalise, X.; Saucedo-Orozco, I.; Quintana, M.; Tagmatarchis, N. Nitrogen-doped silver-nanoparticle-decorated transition-metal dichalcogenides as surface-enhanced Raman scattering substrates for sensing polycyclic aromatic hydrocarbons. ACS Appl. Nano Mater.2018, 1, 3625–3635. Xia, Y.; Wang, B.; Zhang, J.; Feng, Y.; Li, B.; Ren, X.; Tian, H.; Xu, J.; Ho, W.; Xu, H.; et al. Hole doping in epitaxial MoSe2 monolayer by nitrogen plasma treatment. 2D Mater.2018, 5, 041005. do Nascimento Barbosa, A.; Mendoza, C. A. D.; Figueroa, N. J. S.; Terrones, M.; Júnior, F. L. F. Luminescence enhancement and Raman characterization of defects in WS2 monolayers treated with low-power N2 plasma. Appl. Surf. Sci.2021, 535, 147685. Jin, W.; Zeng, X.; Guo, z.; Zeng, y.; Wang, w.; Zeng, y.; Hu, y.; Ziao, y.; Lu, j.; Lu, j.; et al. Optoelectronic properties of lateral MoS2 p-n Homojunction Implemented by selective p-type doping using nitrogen plasma. J. Phys. D: Appl. Phys. 2020, 53, 405102. Lu, J.; Guo, Z.; Wang, W.; Lu, J.; Hu, Y.; Wang, J.; Xiao, Y.; Wang, X.; Wang, S.; Zhou, Y.; et al. Lateral monolayer MoS2 homojunction devices prepared by nitrogen plasma doping. Nanotechnology2020, 32, 015701. Zhang, S.; Li, P.; Chu, S.; Feng, L.; Li, S.; Fan, J.; Xie, S.; Zhang, Y.; Mao, G.; Wang, H. Nitrogen plasma-mediated deposition of silver onto MoS2 toward robust nanozyme with enhanced catalysis for colorimetric assay of hydrogen sulfide in aquaculture water. Appl. Surf. Sci.2022, 153686. Tang, P.; Jiao, J.; Fan, Q.; Wang, X.; Agrawal, V.; Xu, Q. Interlayer spacing engineering in N doped MoS2 for efficient lithium ion storage. Mater. Chem. Phys.2021, 261, 124166. Feng, L.; Zhang, L.; Zhang, S.; Chen, X.; Li, P.; Gao, Y.; Xie, S.; Zhang, A.; Wang, H. Plasma-assisted controllable doping of nitrogen into MoS2 nanosheets as efficient nanozymes with enhanced peroxidase-like catalysis activity. ACS Appl. Mater. Interfaces2020, 12, 17547–17556. Wang, W.; Lu, J.; Wan, D.; Zeng, X.; Lu, J.; Xu, T.; Chen, C.; Zhang, T. Improved performance in MoS2 homogeneous junction field effect transistors by optimizing electrodes contact. Mater. Sci. Eng. B2023, 290, 116348. Zhang, X. M.; Tseng, S. H.; Lu, M. Y. Large-area ultraviolet photodetectors based on p-type multilayer MoS2 enabled by plasma doping. Appl. Sci.2019, 9, 1110. Oxidation Among different modification techniques, which tailor the electrical and optical characteristics of TMDs to specific requisites,490−492 surface oxidation is an attractive and practical method to engineer 2D layers.493−496 As discussed in previous sections, oxygen can adsorb on top of the chalcogen slab in a TMD layer (functionalization), induce chalcogen vacancies, and dope the TMD layer by chalcogen substitution. In fact, while the TMD layer itself may lack dangling bonds, the introduction of oxygen atoms can alter the surface electronic structure, induce chalcogen vacancies, and modify the composition of the TMD layer. This interaction can disrupt the chemical bonds between chalcogen atoms and neighboring molybdenum atoms. As a result, chalcogen atoms are extracted from the crystal structure of the TMD, leading to the creation of chalcogen vacancies which alters the stoichiometry of the TMD lattice by reducing the chalcogen content and potentially causing deviations from the ideal atomic arrangement.497 Moreover, by increasing the plasma power or dwell time, the density of oxygen atoms in the TMD structure increases and results in the formation of oxide islands or layers with a different structure than pristine TMD layers. The surface oxidation process is usually utilized to protect TMDs from environmental effects and to control the mobility of charge carriers and conductivity through interface engineering. Surface oxidation of 2D materials can be performed by ALD,498,499 PVD, CVD,500 exposure to high-purity oxygen or high-purity water vapor,501,502 and plasma treatment.503 Many of these methods may not form a uniform layer with suitable crystal structures, which attenuate the performance of TMD-based devices.491,504 For instance, poor covering of a passive layer on the surface of TMDs due to the absence of dangling bonds makes ALD ineffective for the oxidation process.505 On the other hand, PVD and CVD may impose some crystalline defects and undesired mechanical stresses at the interface.499,506 In contrast, plasma treatment provides the opportunity to deposit a (uniform) oxide layer on the TMD surface through controlled plasma processing and enables precise control over the oxidation process. By finely adjusting parameters like flow rate, pressure, power, and treatment time, the rate of oxidation can be meticulously regulated to achieve the desired thickness and quality of the oxide layer. This level of control is often challenging to attain with alternative techniques, such as thermal oxidation, which may lead to uneven oxide growth and the formation of defects. The formed thin oxide layer on TMDs can be beneficial for various purposes including but not limited to subsequent layer deposition to prepare high-performance devices,507 use as sulfide/oxide heterostructures,508−510 as a template for tellurization and sulfurization for 2H/1T heterostructures,511 or as a p-type dopant for underlying TMD layers to modulate their electrical properties.512−514 Oxidation Mechanism So far, the mechanism of plasma oxidation has not been comprehensively elaborated in the literature. DFT studies of HfS2 have revealed that oxidation occurs layer by layer during the plasma treatment.515 As schematically shown in Figures 17a-c, the oxidation process through plasma initiates by doping. Oxygen atoms penetrate the surface (step I) and decrease the energy barrier for the detachment of neighboring sulfur atoms (steps II and III); consequently, the sulfur atoms are oxidized and leave the surface. This process is repeated for the other chalcogen atoms in the vicinity of oxidized site (Figure 17c), resulting in formation of oxides such as HfOx or HfO2 in the case of HfS2.515 Studies have also shown that oxidation of TMDs by UV/O3 treatment has different degrees of homogeneity depending on the chemistry of the exposed TMD flakes. While a uniform layer of MoOx is formed on the surface of MoTe2,516 an inhomogeneous oxide layer is produced on few layer MoS2 after 60 min of treatment.517 It has been shown that the oxidation of MoS2 is initiated from dislocations and other intrinsic defects and then expanded in the planes to form one-dimensional wave (chain)-like oxides. This difference in the oxidation of MoS2 and MoTe2 can be attributed to differences in in-plane chemical bonds. Moreover, since molybdenum oxide layers are thicker than MoS2 or MoTe2 layers, oxidation of the topmost layer of such materials is usually accompanied by an increase in thickness.518 (a-c) Schematic representations of plasma oxidation of TMDs. The structure in the bottom part of (c) is a HfO2/HfS2 heterostructure. Parts (b and c) reprinted with permission from ref (515). Copyright 2018 Royal Society of Chemistry. Raman spectra of (d) air-oxidized and (e) plasma-treated HfSe2 upon intermittent air exposure for different times. Reprinted with permission from ref (522). Copyright 2019 Elsevier. Applications Surface oxidation of TMDs has a critical effect on the performance of electronic devices. In the following sections, some important applications are presented and discussed. Formation of a Protective Layer Due to the better stability and passivation of the oxide layers, plasma oxidation of TMDs enhances their environmental stability.504,515,519,520 The first oxidized layer formed by plasma treatment acts as an obstacle for the penetration of oxygen to underlying layers which remain intact.515,521 For instance, relatively prolonged plasma treatment of WSe2 (up to 10 min) only oxidizes the topmost layer to protect the material against environmental impacts during the preservation or device fabrication.515 It is noteworthy that the protective role of the oxide layer formed by plasma treatment is better than those prepared by other methods. For example, Figure 17 panels d and e show Raman spectra of HfSe2 after oxidation in air and by plasma. As seen, the A1g mode of HfSe2 at 199 cm–1 remained almost unchanged during plasma-treatment. This implies that the air-oxidized sample was less stable than the plasma oxidized one.522 Modification of Optoelectronic Properties Since the optical and electrical properties of TMD layers highly depend on their chemistry, oxidation by plasma treatment can significantly impact these characteristics.515 Many studies have shown that plasma oxidation degrades the optical properties of MoS2.523 As seen in Figure 18a, the PL intensity of MoS2 is severely quenched by increasing the duration of the O2 plasma treatment. Raman spectroscopy has revealed that the intensities of the E2g1 and A1g modes gradually decrease, while a new peak appeared at 225 cm–1, which is attributed to the disordered oxide structure (Figure 18b).524 XPS studies have also shown the emergence of a Mo6+ peak at 236.4 eV,504,523 indicating that PL quenching is related to the molybdenum trioxide formation. In contrast, Kang et al.525 have shown an enhanced PL intensity for monolayer and bilayer MoS2 due to the formation of MoOx during O2 plasma treatment. They reported that oxidation of the top layer forms a MoOx/MoS2 heterostructure, exhibiting an enhanced PL intensity compared with a MoS2 monolayer (by 3.2 times). A multilayered heterostructure consisting of consecutive layers of MoOx and MoS2 could yield an intensified PL spectrum, owing to the interlayer decoupling between MoS2 layers and minimizing nonradiative trion formation. (a) Time-dependent PL spectra of plasma-treated single-layer MoS2. The treatment time increases from the red line (pristine MoS2) to the purple line (t = 6 s). The inset shows the PL intensity of A1 and B1 peaks versus the plasma treatment time. (b) Evolution of MoS2 and oxide product Raman peak intensities of O2 plasma treated MoS2 as a function of plasma exposure time. Reprinted with permission from ref (524). Copyright 2014 American Chemical Society. (c) Representative band diagram of an equilibrated MoS2/MoOx heterojunction, visualizing the trapping of photogenerated holes at the interface. (d) On/off ratio (Ion/off) and the subthreshold swing (SS) of a HfSe2-based FET device as a function of O2 plasma treatment time. Part (d) reprinted with permission from ref (526). Copyright 2017 Royal Society of Chemistry. (e) Cross-sectional TEM image of top contact region of a photovoltaic device based on Au/Pt/WSe2/WOx/graphene heterostructure. WOx layers were generated by plasma oxidation before graphene deposition. The scale bar is 1 nm. (f) Current density versus applied voltage of the photovoltaic device with (solid line) and without (dashed line) oxide layer under one-sun AM1.5G illumination. The amounts of open circuit voltage (VOC), short-circuit current density (JSC), fill factor (FF), and power conversion efficiency (PCE) are extracted from the device with the oxide layer. Reprinted with permission from ref (527). Copyright 2022 American Chemical Society. The formation of oxides alters the electrical properties of TMD layers.528 For instance, the bandgap of a WSe2 monolayer is changed from direct to indirect due to oxidation.529,530 As a result of electron–phonon scattering inside the first Brillouin zone, the radiative recombination is reduced.524 The formation of oxides is also accompanied by lattice distortion that increases the work function compared to MoS2. Therefore, the plasma-treated material becomes an effective medium semiconductor (EMS). The band offset between MoS2 and EMS can be adjusted by the plasma time; thereby, the resistance and carrier mobility are influenced because the band offset serves as a tunnel barrier for the injected conduction electrons of MoS2.531 Electrical characterization of plasma-treated MoS2 can pave the way to determine the nature of formed molybdenum oxides (molybdenum dioxide or molybdenum trioxide).532 It is worth noting that XPS is not sufficient for this purpose because binding energies attributed to Mo 3d and O 1s of MoO2 as well as Mo 3d of MoS2 and O 1s of MoO3 overlap with each other.266,523,533 Yang et al.532 have developed a bottom-gate thin-film transistor with a MoS2 channel. They have shown that O2 plasma treatment degrades the transistor performance due to increased channel resistance, indicating that an insulator layer (MoO3) is formed on the channel surface. They have also argued that notable enhancement in nonradiative recombination could diminish the PL intensity. Another study534 has determined that the photoresponsivity of MoS2 is significantly enhanced (10-fold) by the plasma treatment, owing to the trapping of charge carriers at the lateral interface of MoS2 and molybdenum oxide. As the Fermi level of MoS2 (the donor) is placed above the MoOx (Figure 18c), the energy band at the interface is highly bent. As a result, photogenerated holes are trapped at the interface to hinder radiative recombination, while electrons, as the majority charge carriers, enhance the photocurrent.534,535 Recently, Lv et al.536 have utilized O2 plasma to fabricate WO3-x nanowires on a WSe2 surface. The resulting 1D/2D WO3-x/WSe2 structures exhibit an ultrasensitive SERS effect with a limit of detection as low as 5 × 10–18 M and an enhancement factor as high as 5.0 × 1011 for probing methylene blue molecules. Such an ultrasensitive performance has been attributed to efficient charge transfer arising from the unique structure of oxide nanowires and effective interlayer coupling in the oxide/selenide heterostructure. Similar enhancements have been also reported for the SERS performance of MoOx/MoS2 heterostructures obtained by UV/O3 oxidation of MoS2 nanospheres.537 A Seed Layer for High-k Dielectric Deposition High-k gate dielectric materials play an essential role in the carrier mobility in TMD-based FETs because the Coulomb scattering of charge carriers in TMDs is highly diminished due to the mismatch between the dielectric constants of the two materials.538 The deposited layer can also decrease the hysteresis of the FET devices by hindering moisture absorption.539 As discussed in the section about functionalization, however, the lack of dangling bonds on chemically inert surfaces of TMDs hinders chemisorption of chemical species during dielectric material growth.540−542 Similar to other types of functionalization, plasma oxidation can also prepare TMDs for the deposition of a uniform high-k dielectric layer on TMDs.504,507 Studies have shown that relatively short plasma treatment (<60 s) of MoS2504 and MoSe2507 leads to the formation of a uniform oxide (MoOx) layer, which acts as a nucleation site for ALD of Al2O3. Besides high-k dielectrics, slight oxidation of TMD materials by plasma can also aid in the deposition of metal nanoparticles on TMDs.543 FET and Electronic Devices Depending on the conditions employed during plasma oxidation, the properties of TMD-based FETs may be improved or degraded.504,532 For instance, it has been shown that oxidation of TMDs can enhance the contact interfaces in FETs.544−546 Yang et al.544 have utilized a UV/O3 treatment to oxidize 2D WSe2 selectively at the contact regions in back-gated WSe2-based FETs. By using a 2D h-BN layer for the protection of the channel layer, they have demonstrated that the formation of WOx with a high work function and p-type behavior at the contact enhances the field-effect mobility and on/off current ratio by 4 and 5 orders of magnitude, respectively.544,547,548 This phenomenon originates from selective oxidation of the open contact area of the FET structure, leading to a p-doping effect that improves the device performance. This selective oxidation process only converts the top few layers of WSe2 into WOx, preserving the intrinsic characteristics of the WSe2 channel. Additionally, an h-BN layer protects the channel region, ensuring its integrity over time. The selective UV/ozone treatment induced a high work function WOx, leading to hole doping in the neighboring WSe2 through electron transfer. As a result, the overall device performance was enhanced without compromising the efficiency of carrier transfer due to tunneling resistance.76,548 In another study, p-type doping via formation of an oxide layer on WSe2 at contact areas of a WS2-based MOSFET device through O2 plasma treatment resulted in a low field effect threshold, high mobility, and high Ion/Ioff ratio.549 Similar results have been reported for SF6/O2 plasma treatment of WSe2550 and O2/Ar (1:3) plasma treatment of few layer MoS2.327 An increase in the electrical conductivity of a MoS2-based FET by an order of magnitude was demonstrated as a result of a 6 s O2/Ar plasma treatment. It has been proposed that the plasma treatment forms a substoichiometric oxide layer between the pristine MoS2 and the MoO3-rich insulator, imposing the transient recovery of electrical characteristics. This trend strengthens the possibility of a highly conductive form of MoOx.551−553 Kang et al.526 have employed an O2 plasma to control the active channel thickness in HfSe2 FETs. With the progression of the plasma treatment, the thickness of the insulating HfOx layer is increased, leading to an enhanced Ion/Ioff ratio and a shift in the subthreshold swing to negative values (Figure 18d). However, prolonged plasma treatments degrade the device performance due to the interfacial carrier scattering on the active channel layer. Recently, Pan et al.554 built a p-n junction based on MoS2-MoTe2 via post -O2-plasma treatment. The plasma treatment parameters were chosen to oxidize the MoTe2 surface and p-dope it while preserving the n-type behavior of MoS2 part since MoTe2 is more prone to oxidation than MoS2. Oxidation of TMD layers can also enhance their performance in photovoltaic devices. For instance, Kim et al.527 were able to enhance the power conversion efficiency by more than 5-fold to reach as high as 5.44% by forming an interfacial oxide layer (WOx) on bulk WS2 (as the absorber layer) and graphene (as top electrode) through O2 plasma treatment (Figure 18e,f). This oxide layer in combination with graphene acted as an electron-selective contact and unpinned the Fermi level at the WS2 interface. Oxidation of TMDs has been also used to build TMD-based memristive devices. Jin et al.555 oxidized HfS2 to form native HfOx by O2 plasma and fabricated a flash memory device with a large memory window of more than 60 V, a high on/off current ratio of ∼105, and a long retention time of more than 103 s. They used such a memory device as an artificial synapse in a stimulated artificial neural network which showed a high accuracy of ∼88% in pattern recognition due to enhanced linearity and symmetry in conductance change characteristics in long-term potentiation and depression processes. Such performance can be enhanced in double-floating gate devices with two plasma-generated HfOx/ HfS2 heterostructures with a large memory window of more than 100 V, a high on/off current ratio of ∼107, a long retention time of more than 5 × 103, and a high accuracy of 96.12% in pattern recognition.556 Oxidation of MoS2 channels in MoS2-based transistors by UV/O3 can also enable memristive switching characteristics with potential usage as memtransistors for energy-efficient neuromorphic electronics.557 While most studies used O2 plasma to oxidize the TMD surface for FET devices, Ji et al.558 used XeF2 plasma etching for oxidation of MoTe2. They showed that since MoTe2 is prone to oxidation in air, keeping MoTe2 in air after plasma etching resulted in oxidation due to defects created during plasma treatment. They overcame Fermi-level pinning and reduced contact resistance by producing an oxidized layer at contact areas in their FET devices. Besides FETs, Cho et al.559 have recently oxidized 1T-WS2 surfaces by O2 plasma to develop 1T-WS2/WO3 heterointerfaces for enhanced electrocatalytic activity for HER. Their TEM and AFM phase imaging shows that plasma-formed WO3 films are amorphous and form a patchwork structured WS2/WO3 interface. The hydrophilic surface of WO3 facilitates the hydrogen spillover effect (HSE). In fact, WO3 can adsorb protons due to the low Gibbs free energy of proton adsorption and formation of a hydrogen bronze phase (HxWO3). This material can easily transfer the proton to active sites in WS2 (eq 18), enhancing the overall HER activity of WS2.18Such plasma treatment was observed to decrease the Tafel slope from 136.1 mV/dec for pristine 1T-WS2 to 102.2 mV/dec for plasma-oxidized 1T-WS2. Table 10 lists recent advances in the field of plasma treatment for the oxidation of 2D TMDs. Table 10Recent Studies on the Effect of Plasma Oxidation on the Properties of TMDs.TMDMethodProcess ConditionsDevicePerformanceCommentMoS2O2 plasmaPressure: 33–400 PaFETField-effect mobility (μFET): 3.1 ± 0.6 × 10–3 cm2 V–1 s–1p-dopingPower:100 WReduced mobilityTime: 5–60 sIncreased resistanceLattice distortion caused by MoO3531MoS2O2/Ar plasmaPressure: 500 PaPhotodetectorDrain-source current: 25 μA (at ± 5 V and laser power of 72 μW)Improved photoresponsivity by formation of surface-bound MoOxO2/Ar ratio: 1:31 Order of magnitude increment in field-effect mobilityTime: 2 s1 Order of magnitude increment in drain-source current534MoS2O2/Ar plasmaPressure: 500 PaPhotodetectorThreshold voltage: –37 VImproved conductivity of the channels due to screening effect of 2D MoOx.O2/Ar ratio: 1:3 time: 2–6 sDecreased threshold voltage (by 1.7 times)551Frequency: 13.52 MHzMoS2UV-O3Time: 1–10 minFET—Severe oxidation and disordering at 10 min treatmentIncreased channel resistance due to MoO3327HfSe2O2 plasmaPressure: 66.66 PaFET (hybrid dielectric layer of HfSe2/ HfOx) and photodetectorIon/Ioff ratio: > 105Enhanced Ion/Ioff ratio by 4 orders of magnitudePower: 50 WμFET: 3.04 cm2 V–1 s–140% improvement in field-effect mobilityTime: 2–26 minSubthreshold swing: 4.8 V dec–1Reduced subthreshold swing by > 6 timesO2 flow: 20 sccmMaximum photoresponsivity: 2.64 × 104 mA W1– (at λ = 520 nm)Shifting threshold voltage to positive values526Ohmic behaviorMoTe2O2 plasmaPower: 150 WTransparent FET with MoOx at contact regions (MoOx/Pt/ITO contact)Mobility: 60 cm2 V–1 s–12.5 times enhancement in drain currentTime: 20 sIon/Ioff ratio: ∼105Two times enhancement in the mobilityaO2 flow: 50 sccmTransmittance: 70%HfS2O2 plasmaPressure: 62.66 PaFET (hybrid dielectric layer of HfS2/ HfO2)515Low interface trap density: 6 × 1011 cm–2 eV–1-Power: 20 WHigh Ion/Ioff ratio: > 107O2 flow: 5 sccmμFET: 10 cm2 V–1 s–1Subthreshold swing: 67 mV dec–1ReS2O2 plasmaPressure: 470 mTorrThin-film transistor (TFT)Ion/Ioff ratio: > 104Affected the drain current at the off-state and the recombination of photocarriersbPower: 20 WMobility: 7.6 cm2 V–1 s–1Time: 30 and 60 sPhotoresponsivity: 2.5 × 107 A W1– (at λ = 405 nm)O2 flow: 5 sccmTemporal response with a rise time of 670 msMoSe2O2 plasmaPressure: 10–2 PaFET with MoOx layer as interfacial layer between MoSe2 and Al2O3 as dielectric gateIon/Ioff ratio: 2.88 × 105>250% enhancement in the field-effect mobilityPower: 100 WμFET: 31.57 cm2 V–1 s–13-fold improvement in Ion/IoffTime: 1 minNo degradation in electrical performance after 21 days507O2 flow: 30 sccmCoSe2Ar/O2 plasmaTime: 120 minCoSe2/CoOx heterostructure with rich oxygen vacancies as oxygen reduction reaction (OER) catalystOverpotential: 251 mV at 10 mA cm–2Promoting charge transfer rateTafel slope: 73 mVdec–1Significant improvement in the OER activity595Turnover frequency: 0.02 s–1Charge-transfer resistance: 23 ΩMoS2-HfS2UV-O3-MOSFET with HfOx as semiconductor/insulator interfaceInterface trap density: 5.40 × 1012 eV–1 cm–2A 10-fold interface quality improvementcSubthreshold swing: 120 mV/decWSe2O2 plasmaPressure: 4 Pap-n homojunction photodetector (WOx-WSe2)Photoresponsivity: 2000 mA/WImproved carrier separationdRF Power: 50 WExternal quantum efficiency: 420% (At the laser excitation of 852 nm)Frequency: 13.56 MHzTime: 200 sO2 flow: 30 sccmWSe2O2/He plasmaPressure: 66.66 PaFETIon/Ioff ratio: 9× 105Improved p-branch on/off by the factor of 100Power: 50 WMobility: 7.7 cm2 V–1 s–1Improved contacts in FETFlow (O2/He): 150/50 sccmEnhanced memresistive switching behavioreWSe2O2 plasma-memristive synapses based on WSe2/WO3 heterostructureEmulating two neuron-based synaptic functionsActing as memristive synapsesfEnergy consumption: ∼2.7 pJ/spikeWSe2O2 plasmaPressure: 66.66 PaGraphene/oxidized WSe2/graphene heterostructure for memristive synapsesVersatile synaptic functionsMigration of oxygen vacancies at the interface identifies resistive switchinggPower: 80 WEnergy consumption: ∼16.1 pJ/spike.Time: 20 sO2 flow: 50 sccmMoTe2UV-O3Wavelength: 186/254 nmFETHole concentration: ∼1.0 × 1013 cm–2Increased hole concentration and mobility (by nearly 100 times)Intensity: 19.39 mW/cm2Hole mobility: 101.4 cm2 V–1 s–1Surface charge transfer between MoOx and MoTe2 reduces the Schottky barrier and contact resistancehMoTe2 /MoS2N2O plasmaPressure: 0.27 Pap-n junctionRectification ratio: 104Plasma-induced MoO3 on MoTe2 acts as p-type dopant and sulfur vacancies generated by plasma in MoS2 act as n-type dopantiPower: 10 WFlow: 100 sccmMoS2Remote O2 plasmaTime: 30 sFETMaximum 564.87% enhancement in μFETReduced off-current leakageSmall hysteresisjWS2O2 plasmaPower: 20 Wp-n junction for photodetectorsPhotoresponsivity: ∼7.1 × 104 mA W−1 (at λ = 532 nm)WOx layer formed by plasma act as p-type dopantkTime: 120 saCho, Y.; Park, J. H.; Kim, M.; Jeong, Y.; Ahn, J.; Kim, T.; Choi, H.; Yi, Y.; Im, S. Fully transparent p-MoTe2 2D transistors using ultrathin MoOx/Pt contact media for indium-tin-oxide source/drain. Adv. Funct. Mater.2018, 28, 1801204.bShim, J.; Oh, A.; Kang, D. H.; Oh, S.; Jang, S. K.; Jeon, J.; Jeon, M. H.; Kim, M.; Choi, C.; Lee, J.; et al. High-performance 2D rhenium disulfide (ReS2) transistors and photodetectors by oxygen plasma treatment. Adv. Mater.2016, 28, 6985–6992.cZhang, W.; Kanazawa, T.; Miyamoto, Y. Performance improvement of a p-MoS2/HfS2 van der Waals heterostructure tunneling FET by UV-O3 treatment. Appl. Phys. Express2019, 12, 065005.dMitta, S. B.; Ali, F.; Yang, Z.; Moon, I.; Ahmed, F.; Yoo, T. J.; Lee, B. H.; Yoo, W. J. Gate-modulated ultrasensitive visible and near-Infrared photodetection of oxygen plasma-treated WSe2 lateral pn-homojunctions. ACS Appl. Mater. Interfaces2020, 12, 23261–23271.eLin, Y. C.; Bersch, B. M.; Addou, R.; Xu, K.; Wang, Q.; Smyth, C. M.; Jariwala, B.; Walker, R. C.; Fullerton-Shirey, S. K.; Kim, M. J.; et al. Modification of the electronic transport in atomically thin WSe2 by oxidation. Adv. Mater. Interfaces2020, 7, 2000422.fHe, H. K.; Yang, R.; Huang, H. M.; Yang, F. F.; Wu, Y. Z.; Shaibo, J.; Guo, X. Multigate memristive synapses realized with the lateral heterostructure of 2D WSe2 and WO3. Nanoscale2020, 12, 380–387.gHe, H. K.; Yang, F. F.; Yang, R. Flexible full two-dimensional memristive synapses of graphene/WSe2–XOY/graphene. Phys. Chem. Chem. Phys.2020, 22, 20658–20664.hZheng, X.; Zhang, X.; Wei, Y.; Liu, J.; Yang, H.; Zhang, X.; Wang, S.; Xie, H.; Deng, C.; Gao, Y.; et al. Enormous enhancement in electrical performance of few-layered MoTe2 due to Schottky barrier reduction induced by ultraviolet ozone treatment. Nano Res.2020, 13, 952–958.iXie, Y.; Wu, E.; Fan, S.; Geng, G.; Hu, X.; Xu, L.; Wu, S.; Liu, J.; Zhang, D. Modulation of MoTe2/MoS2 van der Waals heterojunctions for multifunctional devices using N2O plasma with an opposite doping effect. Nanoscale.2021, 13, 7851–7860.jLiu, N.; Baek, J.; Kim, S. M.; Hong, S.; Hong, Y. K.; Kim, Y. S.; Kim, H. S.; Kim, S.; Park, J. Improving the stability of high-performance multilayer MoS2 field-effect transistors. ACS Appl. Mater. Interfaces2017, 9, 42943–42950.kYue, D.; Ju, X.; Hu, T.; Rong, X.; Liu, X.; Liu, X.; Ng, H. K.; Chi, D.; Wang, X. Homogeneous in-plane WSe2 pn junction for advanced optoelectronics devices. Nanoscale2023, 15, 4940–4950. Recent Studies on the Effect of Plasma Oxidation on the Properties of TMDs. Cho, Y.; Park, J. H.; Kim, M.; Jeong, Y.; Ahn, J.; Kim, T.; Choi, H.; Yi, Y.; Im, S. Fully transparent p-MoTe2 2D transistors using ultrathin MoOx/Pt contact media for indium-tin-oxide source/drain. Adv. Funct. Mater.2018, 28, 1801204. Shim, J.; Oh, A.; Kang, D. H.; Oh, S.; Jang, S. K.; Jeon, J.; Jeon, M. H.; Kim, M.; Choi, C.; Lee, J.; et al. High-performance 2D rhenium disulfide (ReS2) transistors and photodetectors by oxygen plasma treatment. Adv. Mater.2016, 28, 6985–6992. Zhang, W.; Kanazawa, T.; Miyamoto, Y. Performance improvement of a p-MoS2/HfS2 van der Waals heterostructure tunneling FET by UV-O3 treatment. Appl. Phys. Express2019, 12, 065005. Mitta, S. B.; Ali, F.; Yang, Z.; Moon, I.; Ahmed, F.; Yoo, T. J.; Lee, B. H.; Yoo, W. J. Gate-modulated ultrasensitive visible and near-Infrared photodetection of oxygen plasma-treated WSe2 lateral pn-homojunctions. ACS Appl. Mater. Interfaces2020, 12, 23261–23271. Lin, Y. C.; Bersch, B. M.; Addou, R.; Xu, K.; Wang, Q.; Smyth, C. M.; Jariwala, B.; Walker, R. C.; Fullerton-Shirey, S. K.; Kim, M. J.; et al. Modification of the electronic transport in atomically thin WSe2 by oxidation. Adv. Mater. Interfaces2020, 7, 2000422. He, H. K.; Yang, R.; Huang, H. M.; Yang, F. F.; Wu, Y. Z.; Shaibo, J.; Guo, X. Multigate memristive synapses realized with the lateral heterostructure of 2D WSe2 and WO3. Nanoscale2020, 12, 380–387. He, H. K.; Yang, F. F.; Yang, R. Flexible full two-dimensional memristive synapses of graphene/WSe2–XOY/graphene. Phys. Chem. Chem. Phys.2020, 22, 20658–20664. Zheng, X.; Zhang, X.; Wei, Y.; Liu, J.; Yang, H.; Zhang, X.; Wang, S.; Xie, H.; Deng, C.; Gao, Y.; et al. Enormous enhancement in electrical performance of few-layered MoTe2 due to Schottky barrier reduction induced by ultraviolet ozone treatment. Nano Res.2020, 13, 952–958. Xie, Y.; Wu, E.; Fan, S.; Geng, G.; Hu, X.; Xu, L.; Wu, S.; Liu, J.; Zhang, D. Modulation of MoTe2/MoS2 van der Waals heterojunctions for multifunctional devices using N2O plasma with an opposite doping effect. Nanoscale.2021, 13, 7851–7860. Liu, N.; Baek, J.; Kim, S. M.; Hong, S.; Hong, Y. K.; Kim, Y. S.; Kim, H. S.; Kim, S.; Park, J. Improving the stability of high-performance multilayer MoS2 field-effect transistors. ACS Appl. Mater. Interfaces2017, 9, 42943–42950. Yue, D.; Ju, X.; Hu, T.; Rong, X.; Liu, X.; Liu, X.; Ng, H. K.; Chi, D.; Wang, X. Homogeneous in-plane WSe2 pn junction for advanced optoelectronics devices. Nanoscale2023, 15, 4940–4950. Structural Changes Plasma treatment may change the structure and morphology of TMDs, including interlayer decoupling, nanoparticle formation, scrolling, and phase engineering (Figure 19). The last one gives the opportunity to fully exploit TMDs by adjusting electronic properties through phase control for opto-electronic and energy-harvesting devices.560−563 Schematic representation of structural changes in TMDs induced by plasma treatment including (a) phase transformation (2H to 1T), (b) interlayer decoupling, (c) nanoparticle and quantum dot formation, and (d) scrolling. Phase transitions in TMDs can proceed by gliding of atomic planes (transition metals and/or chalcogen planes)564 via alkali metals, hydrogen, and molten-metal-assisted intercalations (MMI),565−568 electron-beam irradiation,569 laser irradiation,102 as well as thermal annealing.103 Electron transfer by plasmonic hot-electron excitation,570,571 strain or stress-induced phase transformation,572,573 and electrostatic doping108 have also been realized experimentally or proposed. However, a clean and versatile process with low complexity and a high production rate to induce phase transformation is still challenging. Plasma treatment has been recognized as a controllable method to transform the 2H phase to a 1T structure in TMDs. A mild plasma treatment containing argon or argon/hydrogen gases supplies the required energy for the stretching of M-X bounds to induce the sliding of atomic layers for a phase transformation. The applied energy is usually less than the required energy to generate a considerable concentration of defects. HRTEM studies on Ar/H2 plasma-treated MoS2 have revealed that the transformation rate can be as high as 70% (Figure 20a-c). The transition of 2H to 1T broadens the Raman peaks and decreases their intensity (Figure 20d). It also quenches the PL spectra (Figure 20e) while shifting the metal atom peaks in XPS spectra to lower energies (Figure 20f).574,575 It has also been demonstrated that the transformed structures (1T- and disordered 1T-TMDs) are stable for more than one month in ambient atmosphere. The thermal stability persists even up to 300 °C, making the process attractive for device fabrication. Unlike intercalation-based methods, plasma-induced transformations in TMDs are relatively fast with a reduced risk of intermediate phase formation.358,574 It is worth mentioning that Ar/H2 plasma treatment can be adjusted to etch TMDs controllably, providing the possibility of direct processing of 1T multi(mono)-layer TMDs from a 2H bulk sample.574 Besides Ar/H2 plasma as a post-treatment, using Ar/H2S plasma can control the TMD phase during TMD synthesis by PECVD, allowing for both 1T and 2H phases. This method results in the formation of large-area multilayered TMDs with a thickness of 5–7 nm, which exhibit polycrystalline structures with nanosized grains. Although the 1T phase is inherently metastable, it has been observed that PECVD-grown 1T-MoS2 and 1T-WS2 can remain stable in ambient atmosphere or even after 1000 cycles of hydrogen evolution reaction.181,182 (a) High resolution TEM images of an Ar/H2 plasma-treated MoS2 nanosheet. The 2H, IT, and intermediate phases are illustrated by purple, green, and brown colors, respectively. Sharp diffraction spots are shown in the SAED pattern of the inset. (b) The atomic structure of (b) 2H and (c) 1T phases. Reprinted with permission from ref (574). Copyright 2018 Springer Nature under CC BY 4.0 [https://creativecommons.org/licenses/by/4.0/]. (d) Raman spectra of pristine and Ar plasma-treated MoS2 nanosheets. New peaks at low Raman shifts indicate the presence of 2H/1T mixed phases. (e) Time-dependent PL and (f) XPS spectra of pristine and Ar plasma-treated MoS2 monolayers. Reprinted with permission from ref (575). Copyright 2017 American Chemical Society. (g) High resolution TEM image and elemental maps of nitrogen and platinum doped MoS2 nanosheets fabricated by N2 plasma. (h) HAADF-STEM image of the doped structure and (i) STEM intensity profile along the marked lines in (h). Reprinted with permission form ref (576). Copyright 2022 Royal Society of Chemistry. Phase transformation of 1T MoS2 to the 2H semiconductor phase has recently been reported as well.577 It has been shown that N2 plasma treatment of graphene oxide/1T MoS2 composites restores the 2H structure of MoS2. Since the 1T-MoS2 is unstable, bombarding the 1T-MoS2 surface with high energy N ions provides the activation energy for such a phase transition.577 It is also shown that nitrogen doping by a 40 s plasma treatment with N2 gas of 1.9 sccm, pressure of ∼1.2 Pa, and power of 10 W can increase the prevalence of the 1T phase in a MoS2 1H/1T mixed phase, while no phase transition has been observed in pure (monolayer) 1H-MoS2. The compressive strain caused by nitrogen doping induces the 1H to 1T phase transformation only at the boundaries between these phases. Interestingly, operando Raman studies during operation of the doped MoS2 as a catalyst for HER have shown that nitrogen dopants leave the sample. As a result, the 1T phase caused by nitrogen as well as the initial 1T phase transform to the 1H phase.578 However, it has been observed that nitrogen doping of 1T/2H MoS2 by ICP plasma for 20 min at a frequency of 13.56 MHz, N2 gas of 20 sccm, pressure of 26 Pa, and power of 10 W does not induce any phase transition.459 While the setups used in these studies are not the same, a comparison of the plasma parameters suggests that nitrogen doping by plasma has a lower energy barrier to transform the 1H (monolayer) to the 1T phase in comparison to transforming the 2H (multilayer) to the 1T structure. In optoelectronic and electric devices such as FETs, the interface between 2D TMD semiconductors and the electrodes is usually accompanied by high resistance contacts, which limits the utilization of the nanomaterials. Phase engineering by plasma treatment is a practical method to resolve the problem of high contact resistance for the next-generation electric devices. As an example, studies have shown that 1T MoS2 builds up Ohmic contacts that greatly affect the behavior of MoS2-based FETs.574,575 Oyedele et al.579 prepared seamless Ohmic contacts for PdSe2-based FETs by a phase transition between PdSe2 and Pd17Se12. Such a transition occurs by a different mechanism from what commonly happens in the 2H-1T(T′) transition. During the PdSe2 to Pd17Se12 transition, the high energy ions activated by the plasma generate a high concentration of defects such as selenium vacancies and vacancy lines to provoke the phase change, which is irreversible. Moreover, He et al.580 transformed 2H MoSe2 nanoflowers to 1T ones and doped them with phosphorus atoms via plasma treatment with Na2P2O2 as the dopant precursor. The resulting structure showed a high reversible specific capacity of 510.8 mAh g–1 at a current density of 50 mA g–1 for sodium storage. Both the plasma induced selenium vacancies and the phosphorus dopants were essential for the phase transition. Vacancies promoted phosphorus doping and decreased the phase transition energy barrier by decreasing crystallinity and increasing structural instability. Phosphorus atoms which occupied selenium vacancies manipulated the electronic structure of the Mo 4d states, inducing the phase transition and stabilizing the 1T structure in MoSe2 nanoflowers. In addition, recent studies have shown that phase transitions from 2H to 1T with plasma in some TMD monolayers like WSe2 enhance their valley splitting, the amount of which can be modulated by the percentage of phase transition, indicating the possible role of plasma-treated TMD layers in valleytronics.581,582 The phase transition from 2H to 1T in MoS2 layers is also a practical tool to enhance their performance as catalysts. For instance, Ghanashyam et al.583 used a water-based plasma processing to induce partial 2H to 1T phase transformation into MoS2 nanosheets for enhancing their catalytic activities in HER and OER. Increasing the plasma treatment time up to 90 min increased the 1T phase content up to 28%, resulting in a significant decrease in the Tafel slopes and considerable reduction of the overpotentials for HER and OER. However, a longer plasma treatment time did not enhance the catalytic activity of MoS2 nanosheets.583 In another study, Sun et al.576 used N2 plasma as a post-treatment to dope MoS2 nanosheets synthesized on carbon cloth with nitrogen atoms. Interestingly, they used various transition metal precursors such as H2PtCl6, RhCl3, Na2PdCl4, and RuCl3 during plasma treatment to also dope the MoS2 nanosheets with transition metals. Nitrogen doping by itself resulted in a 62% phase transition from 2H to 1T. Co-doping of MoS2 nanosheets with nitrogen and transition metals enhanced the coverage of the phase transition, which reached up to 87% in the case of 1.2 wt% of platinum dopants. These dopants changed the nature of the chalcogen-transition metal bonds and aided in the sliding of the chalcogen layers within the TMD layers during the plasma treatment, resulting in a phase transition from 2H to 1T. Figure 20g shows the elemental mapping of platinum and nitrogen atoms, which homogeneously doped the nanosheet. The HAADF-STEM image and intensity profiles in Figures 20 h and i also show the successful incorporation of single platinum atoms. Such doping reduced the HER overpotential from 214 to 38 mV (at a current density of 10 mA/cm2) and decreased the Tafel slope from 125 to 39 mV/dec, mainly due to an almost complete phase transition in the MoS2 nanosheets. Plasma treatment can also change the morphology of 2D materials through various phenomena such as amorphization,584 reshaping the platelets,585 layer decoupling,586 particle formation,587 and reducing particle size. Dhall et al.588,589 have employed mild O2 plasma to increase the van der Waals distance between MoS2 layers without structural damages or layer etching. Figure 21a shows the significant increment in height after O2 plasma treatment. Such layer decoupling transforms the indirect bandgap of multi(many)-layered MoS2 to a direct bandgap with a considerable enhancement in the PL emission (Figure 21b), owing to electronic states decoupling in separate layers.588,589 This process also increases the circularly polarized PL, photoexcited carrier lifetime, and charge-neutral doping. The intercalation of oxygen species is responsible for the layer decoupling.590 (a) AFM height profiles of pristine and O2 plasma-treated MoS2 multilayers indicate an increase in the thickness due to interlayer decoupling. Reprinted with permission from ref (588). Copyright 2015 John Wiley and Sons. (b) PL spectra of MoS2 before and after interlayer decoupling. Reprinted with permission from ref (589). Copyright 2016 American Chemical Society. (c) Schematic view of the mechanism of nanocrystal formation on top of MoS2 by soft H2 plasma treatment and subsequent thermal annealing. Reprinted with permission from ref (591). Copyright 2018 John Wiley and Sons. (d) Formation of MoS2 nanoscrolls due to an increase in the plasma treatment time. Reprinted with permission from ref (592). Copyright 2016 John Wiley and Sons. Zhang et al.593 have utilized soft oxygen plasma (CCEP) treatment to produce 2D atomic crystal molecular superlattices (ACMSs) of various CVD-grown and mechanically exfoliated TMD flakes including MoS2, MoSe2, WS2, and ReS2 as well as their heterostructures with thicknesses in the range of 2 to 8 layers. In such a treatment, oxygen plasma intercalation proceeds through two proposed half-reactions:1920van der Waals interactions between oxygen molecules formed in the TMD interlayer space and adjacent TMD monolayers may stabilize the oxygen molecules in the form of an oxygen molecular layer, thereby forming MoS2[O2]x superlattices (Figure 19b). Experimental and DFT simulations have shown that in the beginning of the CCEP treatment, plasma bombardment produces a small number of sulfur vacancies which are further filled by atomic oxygen, resulting in formation of Mo-O bonds and degrading the PL intensity. Thereafter, plasma ionization produces O2+ ions (eq 19), which intercalate between MoS2 layers (eq 20) due to the electrostatic field of the plasma. The electrostatic field should be parallel to the interlayer space as in the case of CCEP to drive the ions between the layers and expand the interlayer spacing. A large increase in interlayer spacing of MoS2 (from 6.05 to 9.27 nm) can be obtained by regulating plasma time. At the optimum time, the effect of interlayer expansion prevails over the effect of Mo-O bonds, resulting in a considerable enhancement in PL intensity. However, oxygen substitution within the MoS2 layers is an ongoing process which becomes predominant after a certain plasma treatment time. In such stages, the crystal quality of MoS2 degrades and its PL intensity drops. Generally, the resulting interlayer gap expansion effectively isolates TMD monolayers and imparts exotic properties to homo- and heterosuperlattices beyond typical capacities of monolayered TMDs such as 100 times higher photocurrent and 100 times stronger photoluminescence.593 Direct bandgap multilayer TMDs are usable in optoelectronic devices such as photodetectors, solid-state lasers, solar cells, LEDs, and FETs.589,593,594 In another study, Wang et al.595 have utilized CoSe2/diethylenetriamine (DETA) hybrid and Ar/O2 plasma treatment to produce exfoliated CoSe2 nanosheets as a catalyst for OER. They have used a solvothermal method to synthesize CoSe2/diethylenetriamine (DETA) hybrids in which DETA resides between CoSe2 layers. The Ar/O2 plasma treatment onto dried powders of the CoSe2/diethylenetriamine (DETA) hybrid has increased the distance between the layers and transformed its structure from cubic to orthorhombic. This phase transformation has been attributed to partial oxidation of CoSe2 nanosheets during Ar/O2 plasma treatment since using pure Ar plasma would have resulted in exfoliated CoSe2 nanosheets without any phase change.595 Likely, the reaction between DETA and plasma species resulted in volatile compounds, resulting in exfoliation by leaving the hybrid structure. It has also been shown that treating bulk TMDs with O2 plasma before the liquid exfoliation increases the chance of exfoliation with larger surface area. This effect has been attributed to interaction of TMD edges and oxygen species in plasma resulting in charge accumulation and yielding a coulombic repulsion at the edges, which helps the exfoliation process. Moreover, oxygen-incorporated edges have better interactions with solvent molecules during exfoliation.596 Cai et al.591 have recently developed a H2 plasma treatment coupled with a thermal annealing process as a robust and scalable dry method to synthesize MoS2 quantum dots on few layer MoS2. The mechanism of the process is schematically shown in Figure 21c. Although the bombarding hydrogen ions do not have sufficient energy to directly etch the top layer, they have enough energy to disintegrate the top layer to form nanoparticles. The thermal annealing process controls the size of nanoparticles (to less than 10 nm) and enhances the quality of the MoS2 layers beneath by removing sulfur vacancies. Other studies have shown that plasma treatment can roll up TMD layers to provide specific electrical and electrochemical properties to be utilized in electrical devices and energy storage applications.592,597 Plasma species can partially remove chalcogen atoms from the top layer and induce relatively large tensile strength, which rolls up the entire layer. As compared with other scrolling methods including liquid-based methods,598−600 electron-beam irradiation,601 and electrostatic deposition under selective gaseous environment,602 this process has the advantages of being scalable, dry, only one step, and highly efficient.592Figure 21d shows the typical formation process of MoS2 nanoscrolls during an Ar plasma treatment. As seen, the plasma treatment separates the grains of TMDs to form nanoscrolls at the grain boundaries and edges.592 When a reactive plasma gas such as oxygen is utilized, the surface of TMDs is oxidized, the bond angles are changed, and defects such as cracks are formed.603 Ultimately, oxidized TMD nanoscrolls (e.g., MoO3 and WO3) are directly produced from 2D TMDs by a single step. The grain size of the initial TMD layer, the density of defects, TMD-substrate interactions, and plasma species affect the density, distribution, and length of the nanoscrolls.592,603 The utilized conditions which researchers used so far to change the structure of TMDs are summarized in Table 11. Table 11Recent Advances in Inducing Structural Changes in TMDs by Plasma TreatmentStructural ChangeTMDsMethodConditionEffect (s)DevicePerformanceFeature2H-1T Phase transitionMoS2Ar/H2 microwave plasmaPower: 40% to the magnetronTransition from semiconductor to metallic behaviorFETNonlinear and linear Ids–Vds curves in 2H and 1T phases, respectivelyOhmic behavior (1T)Frequency: 2.45 GHzNo temperature dependence of resistance574Time: 7.5 minAr/H2: 90/102H-1T Phase transitionMoS2ICP Ar plasmaPressure: 91.99 PaTransition from semiconductor to metallic behaviorFET-Lower contact resistancePower: 20 WEnhanced ON current575Time: 0–40 sFrequency: 13.56 MHzAr flow: 100 sccm2H-1T′ Phase transitionMoTe2Soft ICP H2 plasma (CCEP)Pressure: 10 PaTransition from semiconductor to metallic behaviorFETLinear Ids–Vds characteristicsOhmic behaviorPower: 50 WReduced temperature dependence of resistance358Frequency: 2 MHz2H or 1T synthesisMoS2PECVD with Ar/H2SPressure: 13.3 PaControlling the phase of synthesized large-area TMD by adjusting operational parametersHER catalystOverpotential: 0.35 VEnhanced catalytic activityWS2Power: 550 WTafel slope: 95 mV dec–1 (1T-WS2 at 5 mA cm–2)Stable 1T phase during 1000 cycles of HER181,182Temperature: 150–300 °C2H-1T Phase transitionWS2Ambient solution discharge plasmaDischarge potential: 15 kV38% 2H to 1T phase transitionHER and OER catalystsOverpotential: −0.19 V (HER) and 0.21 V (OER)Specific capacitance of 157.5 F g–1 for the supercapacitor at 1 A g–1Frequency: 25 kHzSupercapacitorTafel slope: 113 (HER) and 124 mV dec–1Enhanced catalytic activityaTemperature: 55 °C[at −10 (HER) and 10 mA cm2 (OER)]2H-1T Phase transition3D WSe2Ar/H2 microwave plasmaPower: 700 WMore plasma time resulted in higher 1T/2H ratioNO2 gas sensorAverage response: 52.24% to 1 ppm of NO2Enhanced response in comparison to untreated sampleFrequency: 2450 MHz56.57% 2H to 1T phase transition (after 60 s plasma)Best performance after 60 s plasma treatmentbTime: 30–120 sFlow: 100/30 sccm (H2/Ar)PdSe2 to Pd17Se12PdSe2ICP Ar plasmaPressure: 0.4 PaTransition from semiconductor to metallic behaviorFETContact resistance: ∼0.75 kΩμmOhmic behaviorPhase transitionPower: 10–50 WSBH: ∼3.3 meVIncreased carrier mobilityTime: 1–40 sReduced SBHFrequency: 13.56 MHzSuperconductivity579Layer decouplingMoS2Remote RF O2 plasmaPressure: 26.66 PaIndirect to direct bandgap transition in multilayers---Power: 20 WEnhanced PL emission intensity588,589Time: 2–3 minLayer decoupling/ phase change/ oxidationCoSe2Ar/O2 plasmaPressure: 0.5–1 PaCubic to orthorhombic phase transformationOER catalystTurnover frequency: 0.02 s–1Higher active surface area and enhanced catalytic activity595Power: 100 WOxides and oxygen vacancies on/in the upper layerTafel slope: 73 mV × dec–1Time: 120 minCharge-transfer resistance: 23 ΩDouble-layer capacitance: 35.0 mF·cm–2Layer decouplingMoS2Soft ICP O2 plasma (CCEP)Pressure: 38 PaIncreased interlayer distance and indirect to direct bandgap transition by oxygen intercalationPhotodetectorPhotocurrent: 48.7 nA at Vds of 1V under 532 nm laser illumination (intercalated MoS2/WS2)100 times increment in photocurrent of MoS2/WS2 heterostructureWS2RF power: 20 W100 times increment in MoS2 PL intensity593MoSe2Time: 60 sMoS2/WS2 heterostructureFrequency: 2 MHzO2 flow: 10 sccmNano scrolling/ oxidationMoS2Air plasmaPressure: 106.7 PaFormation of chalcogen-oxygen bonds---WS2Power: 18 WCrack formationMoSe2Time: 2–3 sScrolling the layers603aPandey, K.; Jeong, H. K. Ambient plasma treated tungsten disulfide for electrochemical energy applications. J. Phys. Chem. Solids2023, 111520.bDuan, Y.; Zhang, S.; Yu, Y.; Qiu, J.; Feng, S. Facile microwave plasma driven 3D-WSe2 2H-1T phase modulation for improving NO2 gas sensing performance. Sens. Actuators, B.2023, 387, 133822. Recent Advances in Inducing Structural Changes in TMDs by Plasma Treatment Pandey, K.; Jeong, H. K. Ambient plasma treated tungsten disulfide for electrochemical energy applications. J. Phys. Chem. Solids2023, 111520. Duan, Y.; Zhang, S.; Yu, Y.; Qiu, J.; Feng, S. Facile microwave plasma driven 3D-WSe2 2H-1T phase modulation for improving NO2 gas sensing performance. Sens. Actuators, B.2023, 387, 133822. Etching Etching is an important step in the fabrication process of TMD-based devices.604−606 Different etching methods have been developed to include laser heating,607 thermal annealing,608 and plasma treatment.148,609,610 The last one has recently attracted significant attention because it overcomes the disadvantages of other methods.611−613 For instance, although etching of monolayer MoS2 can successfully be performed by laser thinning,607 it is vital to apply a precise control on the monotony and dose of laser exposure, which is an obstacle for scaling up.20,607 Thermal annealing of TMD layers without significant degradation of crystalline quality is feasible but in addition to the risk of forming oxides, the process requires extensive time and energy.608,614 Moreover, plasma etching has less chemical waste handling issues and better mask selectivity in comparison to wet chemical etching processes.615,616 To date, different plasma gases such as He plasma,617 Ar plasma,618 a mixture of CF4 with a noble gas plasma,619 and CF4 plasma620,621 have been used for the etching of TMDs. The process involves physical reactions between 2D TMDs and nonreactive plasma species or chemical interactions with reactive components (Figure 22a). The plasma for such processes can be produced by various methods. Among them, reactive ion etching (RIE) is a simple and robust method to etch the 2D TMDs by reactive species produced by plasma. ICP-RIE is another version of this method that uses an RF power supply (usually 13.56 MHz). The main mechanism for etching during the RIE process is through chemical reactions between plasma species and the target substrate. In this process, the feed gas (usually containing halogens such as fluorine and chlorine) enters the plasma chamber, and plasma turns its molecules and atoms into chemically reactive species such as radicals and ions. These species are directed at the surface of the material by the electromagnetic field between the electrodes and adsorb onto the surface. At last, the species react with the target surface atoms, and the products of such reactions leave the surface, resulting in an etched surface.615,622 Schematic representations of (a) layer-by-layer thinning of TMDs during plasma treatment and (b) the transfer-printing process for the fabrication of large-area arrays of prepatterned few-layer MoS2. (c) SEM images of a bulk prestructured MoS2 stamp. Parts (b and c) reprinted with permission from ref (632). Copyright 2013 American Chemical Society. As discussed earlier, O2 plasma has been widely utilized for etching TMDs.623−626 Studies have shown that the etching mechanism is interlayer engraving initiated by removing the chalcogen atoms (because it requires lower energy). Eventually, transition metal atoms are also etched and the entire layer is removed.627 Pie et al.345 reported that at a low RF power (60 W), the oxidation of MoS2 and WS2 is dominant up to 20 s and etching occurs at longer times. At higher RF powers, the oxidation stage of MoS2 becomes shorter but no effect on WS2 is noticeable due to the higher stability of WO3 than MoO3. Like oxidative etching in air where subnanometer oxide residues may reside on chalcogen vacancies,628 oxide residues may be present after O2 plasma etching. In this case, such residues can be removed mechanically (e.g., by an AFM tip in contact-mode),628 or chemically using a wet etching step629 after the plasma etching process. Moreover, partial oxidized TMD layers which may remain after plasma etching can be transformed to complete TMD layers by rechalcogenation.630 Generally, the type and mixing ratio of plasma gases, pressure, as well as input power influence the etching rate. Lee et al.631 used O2/Ar plasma to etch a MoS2 surface. After setting the parameters for layer-by-layer steady state etching, they found that increasing input power or gas pressure monotonically increases the etching rate. However, varying the etching rate versus gas mixing ratio resulted in a nonmonotonic trend in which the etching rate first increases nonlinearly and then decreases by increasing the Ar fraction. They attributed this trend to reaction pathways which include both physical and chemical etching:2122232425 Equations 21–24 happening simultaneously result in metallic molybdenum which can then be only removed by sputtering. However, the authors did not take into account the removal of molybdenum atoms by oxidation (evaporation of MoOx). By considering probabilities and reaction coefficient rates, they built a model that agreed to their experimental results to some extent.631 The chemistry behind this observation can be rationalized as follows: Increasing the Ar content initially enhances the etching process through physical sputtering as well as producing reactive sites for chemical etching pathways. At some point, however, a further increase in Ar content reduces the available oxygen reactive species for chemical etching (eqs 22, 24 and perhaps removal of molybdenum by volatile oxide). When a CF4-containing plasma is utilized, a shorter etching duration is required because the volatile compounds formed by reactions between reactive species coming from CF4 (mostly fluorine radicals) and TMDs accelerate the etching process. In the case of MoS2 (WSe2), the collision of CF4 reactive species with MoS2 (WSe2) forms fluoride compounds such as MoF6 (WF6), SF4, and SF6, which act as etching agents due to their high-vapor-pressure and volatile nature.619,620 However, CF4 plasma can damage the device substrate made of Si, SiO2, and polyimide. Hence, it is necessary to precisely control the process and the patterned channels.619 SF6 may also be utilized as the plasma gas in its pure form633,634 or in a mixture with other gases.635 The progress in etching can be monitored by Raman spectroscopy by evaluating the positions of the Ag1 and E12g peaks.636,637 The positions of these peaks depend on the dielectric screening of the long-range Coulomb interactions in multilayer materials.638,639 It is important to note that the formation of an oxide layer on the top of TMD layers prolongs the etching process. Although CF4 and SF6 are effective in etching TMDs, some impurities such as fluorides or oxides may remain after the process. By employing an S-containing plasma such as H2S, the resulfurization process removes the impurities without affecting the thickness of the layers.274,620 Recent studies have revealed that a combination of Ar + CF4 + O2 plasma gases in an ICP-RIE process can effectively minimize the presence of F residue and maintain the original conductivity of MoS2 while enhancing the etching peel-off ability.640 Such a method is capable of achieving a remarkable thickness reduction from bulk (4L) down to a monolayer with high uniformity and a precise layer-by-layer control on a 4-in. wafer scale MoS2. Among various considered etching radicals (·O, ·F, ·Cl, and ·H), DFT calculations have shown that ·F, ·O, and ·Cl radicals would be useful for etching since the substitution of S atoms by such radicals is energetically favorable. Moreover, while the presence of ·Cl radicals on the MoS2 surface triggers fluorine adsorption, O radicals in a MoS2 structure significantly suppress it, preventing any possible electrical degradation caused by remaining F species after etching. It has been demonstrated that the accurate and clean MoS2 thinning process by using an Ar + CF4 + O2 combinational environment originates from (i) superior etching selectivity of ·F and ·O radicals; (ii) adsorption inhibition activity of ·O radicals which minimizes the formation of F residues; (iii) Ar+ ion bombardment that can effectively eliminate undesired impurities.640 Atomic layer etching (ALE) is an alternative process that can be used to precisely control the number of TMD layers without introducing damages or contaminations. The process consists of two steps including (i) radical-adsorption (e.g., ·Cl) and (ii) reacted-compound desorption by employing a low-energy Ar plasma.611,641 To prepare MoS2 monolayers, Kim et al.641 utilized ·Cl radical adsorption on the surface of MoS2 bilayers for 2 min followed by an Ar plasma treatment (with 20 eV energy) for 2 min. Varghese et al.642 developed a top-down layer engineering approach by utilizing a microwave plasma-assisted etching method. Layer-by-layer physical treatment of 2D materials without involving any chemical reactions and disrupting the original characteristics is feasible. SF6 + N2 plasma etching was recently introduced by Xiao et al.149 as a soft, selective, and highly efficient technique for etching TMD layers. EEDF, electron density and electron temperatures were calculated based on eqs 6 -8. When the input power density was lower than 1.5 mW/cm3, the electron density and temperature were lower than 7 × 108 cm–3 and 0.4 eV, respectively. These values are much lower than those reported for conventional low-temperature plasmas which are typically used for surface treatment. Moreover, their EEDF curves for low input power densities showed a negligible tail at high electron energies (>2 eV). Such a low fraction of electrons which can ionize and dissociate the plasma gas minimizes the structural damage in TMDs because energetic ions and reactive radicals are diminished. The following reactions between the transition metals (M) and the gas phase are proposed:149262728 The attractive features of this process include (i) the formation of NF3 (eq 26), which is a strong oxidant and reacts with both sulfur and molybdenum at the same time; (ii) MoF4, MoF3, F2, and SF4 are highly volatile and leave no residues; (iii) no noticeable effect on the SiO2/Si substrate; (iv) the rate of etching can be adjusted by altering the plasma power density to provide fine and fast etching modes; (v) the roughness remains nearly constant during the process, indicating a soft and uniform etching process with minimum structural damages.149,643 In the case of using only SF6 gas for plasma etching of MoS2, the overall etch chemistry is29 The products of such a reaction leave the surface and result in etched MoS2 layers.470 As a novel method, plasma etching can be used to pattern complex shapes and designs on TMDs.644 Recently, a transfer-printing process was introduced for the fabrication of large-area arrays of prepatterned few-layer MoS2 (Figure 22b).632 First, a photoresist layer with the desired shape is deposited onto the pristine bulk MoS2. Then, a Ti layer is deposited and patterned by lift-off in acetone. In the next step, SF6 plasma-based RIE is utilized to etch exposed MoS2 and transfer the mask shape onto the underlying MoS2. The Ti mask is then etched by HF acid treatment in order to attain a MoS2 stamp. Afterward, O2 plasma is utilized to create electric charges on the SiO2 substrate. Finally, the MoS2 stamp is used to transfer the patterned arrays onto the SiO2 substrate (Figure 22c). Despite the rough surface of the plasma-treated area, the raised pillar mesas conserved by Ti masks are entirely smooth. Therefore, during mechanical printing, conformal contact with the substrate is attained. Recently, a combination of electron beam lithography (EBL) and RIE has been developed to etch and produce nanostructured TMDs with various shapes and sizes (Figures 23a and b).645 In this method, EBL is used to prepare an etching mask with desired shapes, and a plasma with a mixture of CHF3 and Ar gas is utilized in RIE to etch the TMD layers. Using CHF3 may harden the resist layer on top of TMDs during plasma etching due to carbon deposition. On one hand, this effect allows longer times for etching, resulting in better removal of TMDs in unmasked regions. On the other hand, etching of the hardened layer becomes difficult after the process. To overcome this issue, an O2 plasma treatment can be used before removing the resist layer to first remove the hardened layer. Figure 23c shows the confocal Raman map at a peak of 415 cm–1 for the A1g vibration mode of WS2 nanodisk arrays with various pitch sizes, which shows complete etching between the nanodisks. Interestingly, while the WS2 has a negligible second-harmonic generation (SHG) signal due to the absence of broken inversion symmetry, the nanostructuring of WS2 with ELB and IRE enhances the SHG signal (Figure 23d), especially when the pumping wavelength spectrally overlaps with the lattice modes of nanostructured WS2. This method enabled the researchers to build TMD layers with different geometries for nanophotonic devices. For instance, Figure 23 panels e-g show a proof-of-concept device made of WS2 waveguides and microrings. As seen in Figure 23g, light is transmitted through the waveguide and the output is illuminated after exposure of white light at the input site.645 In another study,646 a conical structure of MoS2 and WS2 was attained by employing polystyrene latex nanospheres and SF6/C4F8 plasma etching for enhanced electrocatalytic activity. In this process, a layer of latex nanospheres with a hexagonal close-packed arrangement is deposited onto the TMD surface. Then, an oxygen plasma is applied to shrink the spheres. Finally, SF6/C4F8 plasma treatment creates conical structures, the morphology of which highly depends on the etching time (Figure 23). Short plasma treatment of MoS2 forms cylindrical pillars with close distance arrays, while a prolonged exposure creates truncated nanocone arrays. The transition from cylindrical morphology to the conical structure with a higher aspect ratio occurs at a treatment time of 15 to 30 s. Truncated arrays are obtained after etching for longer times. In contrast, the formation of a conical structure in 2D WS2 is insensitive to the exposure time (Figure 23l-o). Due to the higher reactivity of WS2 crystals with plasma, the dissolution rate is more pronounced than that of MoS2. Table 12 summarizes the recent advances in the usage of plasma treatment for etching TMD layers and other parts of 2D-TMD-based devices. (a-g) Nanostructuring WS2 by EBL and RIE: (a) SEM image of WS2 nanodisk arrays with diameters of 250 nm and pitch size (distance between two nanodisks) of 200 nm. (b) SEM images, (c) confocal Raman maps at peak of 415 cm–1, and (d) quasi-normal incidence reflection spectra of WS2 nanodisk arrays with a diameter of 250 nm and various pitch sizes of (P) of 200 to 600 nm. (e) Top-view SEM image of a WS2 nanophotonic circuit consisting of a microring resonator coupled with two bus waveguides. (f) Side-view SEM image of a WS2 ring resonator coupled with a bus waveguide. (g) Proof-of-concept WS2-based waveguide, in which the light is transmitted through the waveguide after illumination of one side. Reprinted with permission from ref (645). Copyright 2022 John Wiley and Sons. SEM micrographs of SF6/C4F8 plasma-etched MoS2 crystals for (h) 15 ± 1 s (45° tilt angle), (i) 32 ± 1 s (80° tilt angle), (j) 45 ± 1 s (85° tilt angle), and (k) 60 ± 1 s (45° tilt angle). SEM images of SF6/C4F8 plasma-etched WS2 crystals for (l) 16 ± 1 s, (m) 31 ± 1 s, (n) 46 ± 1 s, and (o) 61 ± 1 s (all imaged at 45° tilt angle). Reprinted with permission from ref (646). Copyright 2018 Elsevier under CC BY 4.0 [https://creativecommons.org/licenses/by/4.0/]. Table 12Recent Advances in Using Plasma Treatment for Etching TMDsTMDMethodConditionsEffects other than etchingDeviceEtched areaPerformanceCharacteristicsMoS2SF6Pressure: 2.67 PaImproving the electronic behaviorFETUnderlying MoS2 (to transfer the Ti mask)Ion/Ioff ratio: 105-107-Power: 200 WField-effect mobility (μFET): 6–44 cm2 V–1 s–1SF6 flow: 20 sccmThreshold voltage (Vth): −50 V632MoS2SF6Pressure: 1.33 PaImproving the electronic behaviorFET based biosensorTMDμFET: 48 cm2 V–1 s–1More than 500% improvement in Ion/Ioff ratioPower: 40 WIon/Ioff ratio: 10614% decrement in carrier mobility633SF6 flow:10 sccmSubthreshold swing (SS): 9 V/decVth: −16 VMoS2O2--FETPMMA (removing mask)aμFET: 20 cm2 V–1 s–1-Ion/Ioff ratio: 106MoS2Ar--FET photodetectorTMDμFET: 35 cm2 V–1 s–1-Ion/Ioff ratio: 108Photoresponsivity: 2160 A/W (at λ = 470 nm)bMoS2SF6Pressure: 4 PaImproving the electronic behaviorFETTMDμFET: 50 cm2 V–1 s–1-Power: 15 WIon/Ioff ratio: 105SF6 flow: 30 sccmCurrent density: 38 μA/μmcMoS2SF6--FETTMD (except channel regions)634μFET: 10.4 cm2 V–1 s–1 Ion/Ioff ratio: 108MoS2CHF3Pressure: 1 PaFluorine-doped and p-type MoS2FET photodetectorTMDμFET: 65.2 cm2 V–1 s–1Gate voltage-dependent photoresponsivity arises from p-type dopingdPower: 100 WIon/Ioff ratio: 103Photoresponse time: 37 msMoS2SF6/O2--FETh-BN–MoS2 contact635μFET: 330 cm2 V–1 s–1-Ion/Ioff ratio: 108MoS2ArPressure: 10 Pa-MOSFETTMDμFET: 25.3 cm2 V–1 s–1-Power: 50 WIon/Ioff ratio: 107Ar flow: 100 sccmSS: 86.0 mV/deceMoS2SF6/C4F8Pressure: 2 PaIncreasing the active edge sitesHER catalystTMDPotential shifts of 200 and 100 mV for MoS2 and WS2, respectivelyImproved electrocatalytic activity646Power: 200 W (ICP) and 20 W (RF)Tafel slope: 120 mV/dec (MoS2) and 50 mV/dec (WS2)Time: 15–60 sMoS2Ar-50 eV Ar ions produced single sulfur vacancies and atomic pitsTMDf---100 eV Ar ions etched the topmost layer500 eV Ar ions produced granulated nanostructuresWSe2O2Pressure: 13 PaImproving the electrical propertiesTFTTMDμFET: 0.01 cm2 V–1 s–1Enhancement in Ion/Ioff ratio619O2 flow: 100 sccmIon/Ioff ratio: >104MoSe2SF6/O2--NO2 and NH3 sensorsMoSe2 channelFast response (250 s) Fast recovery (150 s)Limit of detection of NO2: 10 ppbLimit of detection of NH3: 20 ppbgWSe2CHF3/O2--FETMaskIon/Ioff ratio: 105-SS: 180 mV/dechWSe2O2--FETMaskiIon/Ioff ratio: 106-μFET: 2.2 cm2 V–1 s–1WSe2SF6/O2Power: 30 W-FETTMDjIon/Ioff ratio: 104-Flow: SF6/O2 (45: 5 sccm)μFET: 10 cm2 V–1 s–1WSe2Ar-Good electron–hole pair separationPhotodetectorTMDPhotoresponsivity: 110 A/W4 and 3200 times higher photoresponsivity than 2 and 1- layer-based devices, respectivelykSpecific detectivity: 4 × 1011 jones (at λ = 532 nm)Ion/Ioff ratio: 106Photo current (Iph): 39.5 mA·cm–2WSe2CHF3/O2--FETTMDμFET: 12000 cm2/Vs for graphene/WSe2/SiO2-Mobility: 120000 cm2/Vs for graphene encapsulated between WSe2 and h-BNlWSe2O2- FETTMDμFET: 6236 cm2 V–1 s–1 Electrode contact resistance: 0.2–0.5 kΩ·μmmPdSe2SF6/N2Pressure: 2.66 PaSurface roughness: 0.35 nmFETTMDμFET: 88 cm2 V–1 s–1Increased on/off current ratio and threshold voltageFrequency: 14.56 MHzNo structural damageIon/Ioff ratio: 700Enhanced mobilitynFlow: 2:6 sccm (SF6:N2)Vth: 20 VPdSe2O2/ArPressure: 2 PaIncreasing O2 concentration decreased the number of etched layersFETTMDReduced On current by decrement in flake thicknessIncreased p-type conduction due to residual oxidesoICP Power: 500–220 WFlow: 40 sccmWSe2O2RF Power: 700 W3 min plasma etched away WSe2 while hBN capping layer enhanced its stabilityp----Time: 3 and 10 minMoS2Ar/CF4/O2Pressure: 2.67 PaNo damage to the atomic structure--–Maintain original conductivity due to minimizing impurities such as F species640ICP-RIEPower: 120–150 W (top plate) and 10 W (bottom plate)No chemical impurities such as F species due to simultaneous usage of O2 and Ar with CF4Time: 1 s (each cycle)0.3–0.4 nm thickness reduction in each cycleFlow: 10/5/10 sccm (CF4/Ar/ O2)Temperature: 20–25 °CaDumcenco, D.; Ovchinnikov, D.; Sanchez, O. L.; Gillet, P.; Alexander, D. T. L. Large-area MoS2 grown using H2S as the sulfur source. 2D Mater., 2015, 2, 44005.bChee, S. S.; Seo, D.; Kim, H.; Jang, H.; Lee, S.; Moon, S. P.; Lee, K. H.; Kim, S. W.; Choi, H.; Ham, M. H. Lowering the Schottky barrier height by graphene/Ag electrodes for high-mobility MoS2 field-effect transistors. Adv. Mater.2018, 31, 1804422.cKotekar-Patil, D.; Deng, J.; Wong, S. L.; Lau, C. S.; Goh, K. E. J. Single layer MoS2 nanoribbon field effect transistor. Appl. Phys. Lett.2019, 114, 013508.dZhang, X. M.; Tseng, S. H.; Lu, M. Y. Large-area ultraviolet photodetectors based on p-type multilayer MoS2 enabled by plasma doping. Appl. Sci.2019, 9, 1110.eNinomiya, N.; Mori, T.; Uchida, N.; Watanabe, E.; Tsuya, D.; Moriyama, S.; Tanaka, M.; Ando, A. Fabrication of high-k/metal-gate MoS2 field-effect transistor by device isolation process utilizing Ar-plasma etching. Jpn. J. Appl. Phys.2015, 54, 046502.fLu, W.; Birmingham, B.; Zhang, Z. Defect engineering on MoS2 surface with argon ion bombardments and thermal annealing. Appl. Surf. Sci.2020, 532, 147461.gGuo, S.; Yang, D.; Zhang, S.; Dong, Q.; Li, B.; Tran, N.; Li, Z.; Xiong, Y.; Zaghloul, M. E. Development of a cloud-based epidermal MoSe2 device for hazardous gas sensing. Adv. Funct. Mater.2019, 29, 1900138.hMüller, M. R.; Salazar, R.; Fathipour, S.; Xu, H.; Kallis, K.; Künzelmann, U.; Seabaugh, A.; Appenzeller, J.; Knoch, J. Gate-controlled WSe2 transistors using a buried triple-gate structure. Nanoscale Res. Lett.2016, 11, 1–6.iPark, K.; Kim, Y.; Song, J. G.; Kim, S. J.; Wanlee, C.; Ryu, G. H.; Lee, Z.; Park, J.; Kim, H. Uniform, large-area self-limiting layer synthesis of tungsten diselenide. 2D Mater.2016, 3, 14004.jCampbell, P. M.; Tarasov, A.; Joiner, C. A.; Tsai, M. Y.; Pavlidis, G.; Graham, S.; Ready, W. J.; Vogel, E. M. Field-effect transistors based on wafer-scale, highly uniform few-layer p-type WSe2. Nanoscale2016, 8, 2268–2276.kXu, Z. Q.; Zhang, Y.; Wang, Z.; Shen, Y.; Huang, W.; Xia, X.; Yu, W.; Xue, Y.; Sun, L.; Zheng, C.; et al. Atomically thin lateral p-n junction photodetector with large effective detection area. 2D Mater.2016, 3, 041001.lVölkl, T.; Rockinger, T.; Drienovsky, M.; Watanabe, K.; Taniguchi, T.; Weiss, D.; Eroms, J. Magnetotransport in heterostructures of transition metal dichalcogenides and graphene. Phys. Rev. B2017, 125405, 125405.mXu, S.; Han, Y.; Long, G.; Wu, Z.; Chen, X.; Han, T.; Ye, W. Universal low-temperature Ohmic contacts for quantum transport in transition metal dichalcogenides. 2D Mater. 2015, 3, 021007.nDas, T.; Seo, D.; Seo, J. E.; Chang, J. Tunable current transport in PdSe2 via layer-by-layer thickness modulation by mild plasma. Adv. Electron. Mater.2020, 6, 2000008.oHoffman, A. N.; Gu, Y.; Tokash, J.; Woodward, J.; Xiao, K.; Rack, P. D. Layer-by-layer thinning of PdSe2 flakes via plasma induced oxidation and sublimation. ACS Appl. Mater Interfaces2020, 12, 7345–7350.pLu, Z.; Zhu, M.; Liu, Y.; Zhang, G.; Tan, Z.; Li, X.; Xu, S.; Wang, L.; Dou, R.; Wang, B.; et al. Low-temperature synthesis of boron nitride as a large-scale passivation and protection layer for two-dimensional materials and high-performance devices. ACS Appl. Mater. Interfaces2022, 1, 25984–25992. Recent Advances in Using Plasma Treatment for Etching TMDs Dumcenco, D.; Ovchinnikov, D.; Sanchez, O. L.; Gillet, P.; Alexander, D. T. L. Large-area MoS2 grown using H2S as the sulfur source. 2D Mater., 2015, 2, 44005. Chee, S. S.; Seo, D.; Kim, H.; Jang, H.; Lee, S.; Moon, S. P.; Lee, K. H.; Kim, S. W.; Choi, H.; Ham, M. H. Lowering the Schottky barrier height by graphene/Ag electrodes for high-mobility MoS2 field-effect transistors. Adv. Mater.2018, 31, 1804422. Kotekar-Patil, D.; Deng, J.; Wong, S. L.; Lau, C. S.; Goh, K. E. J. Single layer MoS2 nanoribbon field effect transistor. Appl. Phys. Lett.2019, 114, 013508. Zhang, X. M.; Tseng, S. H.; Lu, M. Y. Large-area ultraviolet photodetectors based on p-type multilayer MoS2 enabled by plasma doping. Appl. Sci.2019, 9, 1110. Ninomiya, N.; Mori, T.; Uchida, N.; Watanabe, E.; Tsuya, D.; Moriyama, S.; Tanaka, M.; Ando, A. Fabrication of high-k/metal-gate MoS2 field-effect transistor by device isolation process utilizing Ar-plasma etching. Jpn. J. Appl. Phys.2015, 54, 046502. Lu, W.; Birmingham, B.; Zhang, Z. Defect engineering on MoS2 surface with argon ion bombardments and thermal annealing. Appl. Surf. Sci.2020, 532, 147461. Guo, S.; Yang, D.; Zhang, S.; Dong, Q.; Li, B.; Tran, N.; Li, Z.; Xiong, Y.; Zaghloul, M. E. Development of a cloud-based epidermal MoSe2 device for hazardous gas sensing. Adv. Funct. Mater.2019, 29, 1900138. Müller, M. R.; Salazar, R.; Fathipour, S.; Xu, H.; Kallis, K.; Künzelmann, U.; Seabaugh, A.; Appenzeller, J.; Knoch, J. Gate-controlled WSe2 transistors using a buried triple-gate structure. Nanoscale Res. Lett.2016, 11, 1–6. Park, K.; Kim, Y.; Song, J. G.; Kim, S. J.; Wanlee, C.; Ryu, G. H.; Lee, Z.; Park, J.; Kim, H. Uniform, large-area self-limiting layer synthesis of tungsten diselenide. 2D Mater.2016, 3, 14004. Campbell, P. M.; Tarasov, A.; Joiner, C. A.; Tsai, M. Y.; Pavlidis, G.; Graham, S.; Ready, W. J.; Vogel, E. M. Field-effect transistors based on wafer-scale, highly uniform few-layer p-type WSe2. Nanoscale2016, 8, 2268–2276. Xu, Z. Q.; Zhang, Y.; Wang, Z.; Shen, Y.; Huang, W.; Xia, X.; Yu, W.; Xue, Y.; Sun, L.; Zheng, C.; et al. Atomically thin lateral p-n junction photodetector with large effective detection area. 2D Mater.2016, 3, 041001. Völkl, T.; Rockinger, T.; Drienovsky, M.; Watanabe, K.; Taniguchi, T.; Weiss, D.; Eroms, J. Magnetotransport in heterostructures of transition metal dichalcogenides and graphene. Phys. Rev. B2017, 125405, 125405. Xu, S.; Han, Y.; Long, G.; Wu, Z.; Chen, X.; Han, T.; Ye, W. Universal low-temperature Ohmic contacts for quantum transport in transition metal dichalcogenides. 2D Mater. 2015, 3, 021007. Das, T.; Seo, D.; Seo, J. E.; Chang, J. Tunable current transport in PdSe2 via layer-by-layer thickness modulation by mild plasma. Adv. Electron. Mater.2020, 6, 2000008. Hoffman, A. N.; Gu, Y.; Tokash, J.; Woodward, J.; Xiao, K.; Rack, P. D. Layer-by-layer thinning of PdSe2 flakes via plasma induced oxidation and sublimation. ACS Appl. Mater Interfaces2020, 12, 7345–7350. Lu, Z.; Zhu, M.; Liu, Y.; Zhang, G.; Tan, Z.; Li, X.; Xu, S.; Wang, L.; Dou, R.; Wang, B.; et al. Low-temperature synthesis of boron nitride as a large-scale passivation and protection layer for two-dimensional materials and high-performance devices. ACS Appl. Mater. Interfaces2022, 1, 25984–25992. Defect Healing Chalcogen vacancies are among the most common intrinsic point defects in TMD layers.463,647−650 As mentioned before, these vacancies are generally known as charge scattering centers with deep acceptor states, which can act as efficient electron traps to limit electron mobility.463,647 Chalcogen vacancies are easily formed during processing, for example, mechanical exfoliation, CVD, or electron beam sputtering, due to the high volatility of the chalcogenides, particularly S. It has been shown that the formation of sulfur vacancies creates donor states inside the bandgap and scattering centers with a huge impact on the carrier mobility. While sulfur vacancies can provide more donor states for better conductivity, it is believed that they may act as scattering centers and therefore reduce the charge mobility. This has been proposed as a primary reason for unexpectedly lower mobilities observed for MoS2 monolayers in comparison to their bulk form.292,650−652 For defect healing of 2D TMDs, different methods such as air passivation,649,651 chloride molecular doping,653,654 focused-laser annealing in air, γ-radiation,648 and plasma treatment (CO, NO, NO2, etc.) using different sources have been examined. The concept is to replace the vacancies with foreign atoms to enhance the conductivity and photoconductivity. Lu et al.647 have shown that laser treatment of WSe2 in the ambient condition improves the conductivity by ∼400 times through the passivation of selenium vacancies by O. Yang et al.653 have reported that sulfur vacancies can be filled by chloride molecular doping to reduce the contact resistance in few-layered TMDs such as WS2 and MoS2. Vogl et al.648 have employed γ-radiation to decrease the defect density in monolayer WS2. They have shown that the PL intensity and carrier lifetime increases, owing to the chemisorption of oxygen facilitated by γ-irradiation. Plasma treatment has recently been used as a practical and efficient way to repair structural defects in 2D materials such as graphene,655 WS2,656 HfO2,657 SnS2,658 and MoS2.651 So far, various sources of plasma such as Ar,659 CH4/Ar,660 and CH4/H2661,662 have been utilized for 2D materials treatment. The mechanism of healing is attributed to the etching of photoresist or tape residues on the 2D films along with simultaneous defect generation and healing by plasma species. In fact, etching agents such as hydrogen and argon ions and radicals may generate unwanted vacancies. At the same time, chalcogen or other foreign ions and radicals in plasma can heal the surface by filling both preexisting and plasma-generated defects. This treatment not only removes vacancies in 2D materials but also eliminates other defects such as folded edges655 and holes.663 Studies have revealed that active gases (e.g., oxygen) are more effective at healing defects than a neutral plasma environment (e.g., argon). While an argon treatment on a monolayer MoS2 can lead to a decrease in the carrier mobility due to a high concentration of induced defects and lattice distortion, employing an O2 plasma enhances the mobility by 10-fold through substituting the vacancies with oxygen or chemical bonding of oxygen ions with the matrix.651 It is reported that such chemical reactions between chalcogen vacancies and oxygen atoms, molecules, or ions resulted in enhancement of PL intensity even up to 30-fold.664 H2 plasma treatment of 2D WS2 has also been examined for healing defects via reducing W6+ surface species. The resulting nanomaterial has shown promising performance as electron transport layers (ETLs) in polymer solar cells.665 In another study,657 it has also been reported that UV/O3 plasma treatment of sol–gel-derived TMDs such as HfO2 films can effectively decrease cracks. Consequently, the leakage current in a built HfO2-based electrical device is decreased by more than 3 orders of magnitude. Al-Jawhari et al.666 used a similar method to heal the defects in MoS2 layers and enhance their optical properties for use as a photodetector. During the treatment, oxygen molecules and ions react with the sulfur vacancies and passivate them. Defects, especially chalcogen vacancies, may also form during plasma-based synthesis of other 2D materials on TMD substrates. In these cases, using chalcogen-containing gases such as H2S during the synthesis usually results in healing produced vacancies in the TMD substrates, resulting in better contact with the grown layer.667 Cleaning Residual impurities that remain on the surface from the synthesis process, handling, lithography, post-treatment, and device fabrication can severely degrade the properties of 2D materials and the performance of electronic and optoelectronic devices.668,669 For instance, solvents and polymers used in the transfer process of 2D materials unintentionally act as impurities via surface doping and change the electrical properties.670,671 The impurities can also negatively affect the charge mobility at the electrodes arising from the nonuniform accumulation of charge carriers.662 Due to their large surface area, 2D TMDs are susceptible to accumulating physisorbed impurities which are loosely attached through van der Waals interactions.672 Surface cleaning steps are therefore important to remove the impurities for the fabrication of high-performance devices. Different post-treatment methods have been employed, such as thermal annealing,673,674 applying electrical current,675 contact-mode AFM imaging,676,677 UV-treatment,678 CO2 cluster treatment,679 electrochemical processes,680 centrifugation washing with nonpolar672 and polar hydrocarbon solvents,681 wet chemical treatment,682 and ion sputtering.683 Nevertheless, many of these processes require complicated procedures and may degrade the structure and properties of 2D materials. There is also a limitation in the size or pattern of the 2D materials.662 Recently, plasma662,684−686 and UV/O3687 treatments have been introduced as practical and fast cleaning processes for 2D materials. As schematically shown in Figure 24a, the process enables the removal of contaminations that are adsorbed at the surface or that remain at the contacts between TMDs and electrodes in TMD-based devices.688−692 O2 plasma treatment of MoS2-based FETs (Figure 24b) changes their drain-source current versus drain-gate voltage (IDS-VDG) from nonlinear to linear behavior (Figure 24c). The enhanced drain current at different VDGS (Figure 24d) is attributed to the higher carrier mobility (∼15-fold) after plasma treatment (Figure 24e). If titanium is used as the electrode, oxygen atoms adsorbed by MoS2 during O2 plasma cleaning process may react with the metal to form TiO2, which builds a low conduction band offset with the 2D material and improves carrier injection.693 Chemical analyses have revealed that exposing MoS2 to O2 plasma results in a modification of its Fermi energy. This modification is attributed to the formation of MoOx species, indicating that the observed shift in the Fermi level toward the valence band after O2 plasma treatment is influenced by both hole injection facilitated by MoOx and the removal of residues. Moreover, when Ti is subsequently deposited, it acts as a scavenger for the oxygen species. This scavenging process leads to the reduction of MoOx, resulting in the formation of TiOx. As a result, carrier injection is enhanced, likely due to the presence of a small conduction band offset in TiO2.693 Unwanted physical damages and defect engineering may also happen during O2 plasma cleaning treatment which can degrade the electrical properties of TMDs. Such adverse effects can be avoided by using soft plasma treatments with a low input energy. (a) A schematic representation of plasma surface cleaning of 2D TMDs. (b) A schematic representation of MoS2-based FETs. (c) Drain current versus drain voltage (ID-VD). (d) IDS-VDG curves before and after oxygen plasma treatment. (e) The effect of O2 plasma on the charge carrier mobility. Parts (b-e) reprinted with permission from ref (693). Copyright 2019 IOP Publishing. Surface cleaning of 2D TMDs by oxygen and water vapor may also degrade their electrical behavior and associated devices through surface passivation and/or forming unwanted oxides.694−697 Studies have shown that the implementation of a mild plasma treatment containing nonreactive gas(es) such as argon is effective in removing undesired species to restore electrical characteristics without etching or forming defects.698−700 Another approach could be to implement O2 plasma treatment for a short time. Xia et al.701 have shown that treatment of SnS2 for 5 s with O2 plasma effectively removes residues from the surface and enhances its photoelectric current. However, processing for longer times (e.g., 60 s) induces defects and causes oxygen doping that degrade the photoelectrical characteristics. Growth Substrate Modification Cleaning Surface cleaning of substrates in 2D material-based devices is crucial for the successful deposition of atomic layers and to achieve the desired properties. The presence of any type of contamination on the substrate negatively affects the deposition of the subsequent layers, reduces the adhesion quality, induces voids between the substrate and the upper layer, and forms electrical defects.702,703 So far, various wet-cleaning and dry-cleaning procedures have been utilized in order to attain clean substrates.704−706 Due to safety issues and processability, the dry processes have gained much more attention in recent years.703,706 Low-temperature plasma ions and their byproducts such as UV-light or ozone are effective procedures for decontamination and surface cleaning. These processes have the advantage of being fast, versatile, and effective. Plasma cleans the surface by inducing physical damage (sputtering) or/and chemical reactions between the contaminant and reactive plasma species, turning contaminants into volatile compounds.703,707 Consequently, plasma has been successfully employed for surface cleaning of the substrate before 2D TMD growth and nanodevice fabrication.708−712 Defect Engineering Surface defects can act as preferable nucleation sites for the growth of TMD layers on device substrates.713,714 The density of defects can be tuned via the conditions during plasma treatment. Studies have shown that increasing the defect density on the substrate may result in the formation of more TMD flakes but with smaller lateral sizes.715,716 It has also been reported that single vacancy defects in h-BN substrates facilitate epitaxial (orientational) growth of TMD flakes.715,717 However, the formation of complex defects and a high density of vacancies compromise the stacking orientation at the early stages of TMD growth. For instance, interactions between complex defects in h-BN and the TMD edges form randomly oriented flakes.715,717 Breaking C-C bonds in highly ordered pyrolytic graphite (HOPG) by plasma pretreatment enhances the CVD-growth of TMDs. The small clusters of HOPG arising from broken bonds act as nucleation sites for TMD growth.718 A low-temperature plasma treatment would also roughen the surface of soft and stretchable polymers (such as PDMS) and form wrinkled structures.719,720 By employing these substrates for the growth of 2D TMDs, it is feasible to prepare topological surfaces from the grown layers.721,722 For instance, a photodetector based on plasma-synthesized MoS2 on a PDMS surface which had a wrinkled or buckled interface due to plasma modification of the PDMS film showed an average responsivity of 2.52 μW A–1 under 515.6 nm illumination. Interestingly, the device preserved its photoconductive behavior under tensile strain of lower than 5.72% over more than 1000 stretching cycles.721 In contrast to our understanding of the role of plasma-induced substrate defects in TMD growth, their effects on the performance of TMD-based electric devices are not well analyzed. Generally, due to the high surface-to-volume ratio in 2D materials, their properties are mainly dominated by the surface. Therefore, substrate or interface defects can considerably influence the properties of 2D layers. For instance, as discussed earlier, relatively lower mobilities are achieved for monolayers of TMDs in comparison to their bulk form which have been attributed to trap states originating from vacancies and interface defect states.350,723 Such defects have been also suggested as a key factor for the commonly observed hysteresis in transfer curves of MoS2-based FETs.723,724 On one hand, this hysteresis makes MoS2 as a candidate for memory devises.725 For example, it has been shown that O2 plasma treatment of the h-BN substrate in a 2D SnS2/h-BN heterostructure produces defects in h-BN which act as charge trapping centers and enables synaptic characteristics to be used as artificial synaptic transistors.726 On the other hand, such a hysteresis causes reliability issues for the devices in electronics and optoelectronics.725,727 Studies have shown that the passivation of surface defects in SiO2 substrates by Na+ or various organosilane self-assembled monolayers with different end groups such as −SH and −NH2 can enhance the electron transfer in MoS2-based FETs.728,729 Therefore, not only defect engineering of substrates, but also passivation of their surface defects by plasma functionalization or doping can be considered as interesting tools to further tailor the performance of TMD-based electronic devices. Nitridation Plasma nitridation is a practical tool to enhance the performance of TMD heterostructure-based devices.730 Studies have determined that plasma nitridation of the non-TMD part of the heterostructure may influence the overall electronic properties of the device.731 For instance, Zhang et al.731 have shown that plasma nitridation of GaN in single-layer MoS2/GaN heterostructures with type-II band alignment alters the electronic structure of the MoS2 layer. Herein, the conduction and valence band edges shift upward while augmenting the conduction and valence band offsets. It was shown that this change in the electronic band structure enhances the accumulation capability and photocatalytic activity of the MoS2/GaN heterostructures. Similar results have been reported by Liu et al.732 for multilayer MoS2/Al2O3 heterojunctions with type-I band alignment. Plasma nitridation by NH3 plasma treatment was shown to alter the band structure of Al2O3 instead of TMDs. The plasma nitridation enhances the separation between the valence band maximum and Al 2p core-level energy, which expands the conduction band offset. Haun et al.733 have also observed that plasma nitridation of HfO2 tunes the band alignment in MoS2/HfO2 heterostructures and changes the conduction (valence) band offsets from 2.16 ± 0.1 (2.34 ± 0.1) to 2.41 ± 0.1 (2.09 ± 0.1) eV, respectively. These studies pave the way to integrate such structures in 2D TMD-based electronic devices. In a recent study conducted by Pi et al.,734 nitridation of carbon cloth (CC) by nitrogen plasma enhanced the MoS2 growth by providing nucleation sites and increased electrochemical activity of the CC substrate due to nitrogen-containing groups at its surface. The MoS2/N-doped CC exhibited a significant energy density of 260.94 μWh/cm2 at a power density of 354.48 μW/cm2 and a specific capacitance of 3834.28 mF/cm2 at a current density of 1 mA/cm2. Engineering of Surface Energy To synthesize large-scale TMD layers, the utilized precursors should wet the substrate properly. As explained above, O2 plasma treatment increases the hydrophilicity of silica surfaces to improve their wettability for transition metal sources in solution-based and CVD methods.735−740 The improved hydrophilicity enhances the thinning and coalescence of precursor clusters to form large-area TMD films. Figure 25 shows the influence of O2 plasma treatment on the growth of 2D MoS2 on Si/SiO2 substrates. By prolonging the time of the plasma treatment, the contact angle of Na2MoO4 droplets (used as a precursor) on the surface decreases. Consequently, the nucleation and lateral growth of MoS2 flakes during CVD are promoted.741 O2 plasma also enhances the wettability of (NH4)2MoS4, Na2WO4, or Na2MoO3 solutions on the Si/SiO2 substrate, enhancing the coverage of the resulting TMD film through CVD processes.742−745 A similar method was deployed by Shen et al.746 to increase the hydrophilicity of the Si/SiO2 substrate for uniform CVD growth of doped TMD monolayers by using dopant-containing salt precursors. Raman and PL spectroscopy have determined the formation of single-crystal MoS2 flakes without grain boundaries. Other studies have affirmed that the O2 plasma treatment can enhance the nucleation and growth of TMDs in CVD-based methods due to the enrichment in O-dangling bonds that affects the surface energy.741,747,748 The breaking of chemical bonds on the surface (for example, Si bonds) by plasma also creates new dangling bonds to enhance the nucleation and growth processes.749 Meanwhile, some researchers have argued that prolonging the treatment time may decrease the overall nucleation density of TMDs, for example, WS2750 and MoS2,751 by enlarging the lateral size of the flakes. It has also been claimed that O2 plasma treatment decreases the activation energy for nucleation and growth of TMD flakes.750 Therefore, it is important to optimize the time and power of plasma to avoid the formation of small-area discontinuous layers and/or TMD multilayers.164,752 (a) Effect of O2 plasma treatment on the contact angle of Na2MoO4 droplets on Si/SiO2 substrates. The inset shows the lateral size and surface density of grown MoS2 as a function of treatment time. (b) Raman spectroscopy, (c) PL spectroscopy, and (d) optical imaging indicate the effect of plasma treatment on the quality of synthesized MoS2 flakes. Reprinted with permission from ref (741). Copyright 2019 American Chemical Society. Plasmas other than O2 plasma are being used to alter the surface energy of substrates used in the processing of 2D TMDs. For instance, N2 and CF6 plasma treatments can incorporate nitrogen and fluorine atoms into the surface and reduce the surface energy.753 Consequently, large-area TMD flakes can easily be grown on N2 or CF6-containing plasma-treated substrates such as glasses. UV/O3 treatments can also yield more hydrophilic surfaces and better film coverage.704,754 Generally, large-area 2D TMD films grown on plasma- or UV/O3-treated substrates have shown acceptable electronic properties such as high Ion/Ioff ratios of ∼106 in MoS2-based FETs752,754 with high field-effect mobilities of 15.6 cm2 V–1 s–1 (trilayer film), 8.2 cm2 V–1 s–1 (bilayer film), and 3.6 cm2 V–1 s–1 (monolayer film)752 as well as responsivity of 22 mA/W under white light illumination in MoS2-based photodetectors.743 Besides aforementioned substrate modifications, it has been recently shown that O2 plasma pretreatment of Si or Si/SiO2 substrates prior to CVD growth of TMD monolayers results in in situ etching of such layers during CVD processes. Zhang et al.755 found that by adjusting O2 plasma pretreatment time and CVD parameters, it was possible to generate well-defined etch patterns with zigzag-terminated edges inside CVD-grown TMDs. The O2 plasma pretreatment of the substrate substantially promoted the etching without altering the chemistry of TMD layers, such as by oxygen doping or oxidation. Their experimental and theoretical studies showed that etching evolution starts from the peripheries of the TMD flakes and proceeds to the inside of the flakes. This etching process resulted in the formation of hexagonal etch pits in the case of MoS2 as well as MoSe2 and triangular etch pits in the case of WSe2. Moreover, increasing the plasma pretreatment time resulted in a higher density of etch pits with smaller sizes. In addition, it is observed that O2 plasma pretreatment of a SiO2 substrate before MoS2 exfoliation may form functional groups of −OH at the SiO2 surface,756,757 resulting in p-type doping of the MoS2 layer on top of the substrate and enhancing its PL spectra. Such functional groups act as charge trapping sites which result in a large hysteresis in the transfer curve of MoS2-based FETs.756 Conclusions and Future Perspectives In this article, the role of plasma-based methods in the realm of 2D TMDs for obtaining desired properties and fabricating novel optoelectronic and energy harvesting devices was comprehensively reviewed. The capability of plasma treatment methods to synthesize and modify 2D TMDs was explained in detail. The unique advantages of plasma treatment rely on its ability to be employed for various reactive components at low cost and energy consumption. The scalability and controllability of the process also provide a great opportunity to fabricate and modify TMDs based on industrial needs. It has also been demonstrated that the role of plasma does not end with the synthesis and modification of TMDs, meaning that the technique is a practical tool to modify the substrate and change its structure and properties before the growth of TMD layers for device fabrication. We discussed practical ways to synthesize and modify large area 2D TMDs with desired physicochemical properties through various plasma processing methods. These approaches include (1) surface functionalization of the topmost chalcogen layer by physisorbed or chemisorbed adatoms such as fluorine and nitrogen, (2) induction of various zero, one, or two-dimensional defects, (3) surface doping by replacing chalcogen atoms with other atoms at low concentrations, (4) partial or complete oxidation of the surface by replacing chalcogen atoms with oxygen, (5) change of the phase and morphology of the layered materials, (6) etching of the atomic layers, (7) healing and repairing of the edges, vacancies, and other defects by adding chalcogen or oxygen atoms, and (8) cleaning of the surface from impurities. Obviously, these modifications can extensively influence the physical, electrical, and chemical properties of 2D TMDs. These capabilities introduce plasma processes as promising methods to tune the properties of TMDs beyond the properties reachable by conventional modification techniques such as thermal and solution-based procedures. However, more studies are required to determine the mechanism of plasma modification for successful future implementation of plasma-based methods in industry. In terms of process optimization, it is important to control the plasma environment and related parameters to focus on a specific modification while minimizing unwanted structural changes in the processing materials. For instance, it would be relevant to modulate plasma conditions to achieve desired band structure modifications within the TMD layers without introducing structural damages. Particularly, 2D TMDs are more prone to ion-bombardment damages during low-temperature plasma treatments due to their small size and ultrathin features. In this regard, soft plasma or capacitively coupled electrodeless plasma (CCEP) can be used in capacitive discharge mode with a low input power to obtain various modifications in 2D TMDs without structural damages. The capacitively coupled radial electrostatic field parallel to the sample surface in CCEP minimizes the chance for electrons to gain enough kinetic energy for further ionization processes within the plasma gas, resulting in a low ionization rate and ion density. The electron and ion densities as well as types of reactive species in the plasma are highly dependent on plasma conditions which should be studied for each type of plasma. While the chemistry of various plasmas has been studied during the past decades, many studies which have used plasma as a modification method for TMDs were focused on the outcome of the plasma treatment of TMDs rather than the chemistry and thermodynamics of the plasma that was used. In particular, the role of each reactive species, the energy distribution in the plasma, and the reactions between the plasma species and TMD surfaces has not been sufficiently explored. As a result, studying the electron energy distribution and characterizing the reactive ions with operando optical microscopy methods within the specific plasma conditions used for TMDs can pave the road toward a better understanding of the plasma-modification mechanisms in TMD materials and better control of the molecular-scale interactions. Generally speaking, using plasmas with lower electron temperatures and lower dwell time results in low-energy modifications such as functionalization or formation of single chalcogen vacancies. Increasing the energy of the plasma (i.e., increasing the density of reactive ions) and plasma dwell time increases the possibility of high-energy modifications such as doping, oxidation, 2D defects, and etching. Moreover, the effect of substrates, such as SiO2, on plasma modifications of TMDs and in particular their defect engineering is often neglected and needs to be further studied before large-scale utilization. Various analytical techniques were introduced and reviewed for the purpose of tracing and monitoring the effect of plasma treatment on the thickness, defect formation, incorporation of functional or foreign atoms, and oxidation. It has been shown that plasma treatment affects Raman spectra, PL intensity, light-matter interactions, and electronic band structure of TMDs. It is generally believed that the utilization of plasma in the realm of TMDs will grow significantly once it is possible to overcome challenges related to scalability and controllability in industry. Deepening our understanding of the influence of plasma treatment and its processing parameters on the structural changes and physicochemical properties greatly expands its application for the fabrication of 2D TMD-based devices. Further studies on 2D heterostructures, alloying and metal doping, fabrication of Janus layers, formation and stabilization of new phases, and morphological changes by plasma treatment of TMDs are of high interest for novel and advanced applications.