Exploring p-Type Contact for Monolayer WS2 FETs Using Halogen Doping and Intermediate Layers Authors: D Sharda Devi, Nihar R Mohapatra Published: 2024 ABSTRACT AbstractThis work presents a density functional theory (DFT) study of substitutional and adsorption-based halogen (I or F) doping of WS2-based transistors to enhance their contact properties. Substitutional doping of the WS2 monolayer with halogens results in n-type behavior, while halogen adsorption on the surface of the WS2 monolayer induces p-type behavior. This is attributed to differing directions of charge flow, as supported by the Mulliken analysis. However, due to Fermi-level pinning (FLP) at the WS2-metal interface, the p-type behavior resulting from halogen adsorption is not very prominent. To achieve better p-type contact, intermediate layers of graphene and h-BN are used to mitigate the FLP effect, showing significant improvement. The F-adsorbed WS2-graphene-Pt interface demonstrates excellent p-type contact with a substantial reduction in the hole Schottky barrier height, making it ideal for efficient WS2-based p-type MOS transistors in CMOS technology. Introduction In recent times, transition metal dichalcogenides (TMDs) have garnered significant attention in the semiconductor industry due to their exceptional electronic properties and atomically thin structures. Among these materials, WS2 is particularly noteworthy for exhibiting the highest carrier mobility at room temperature.1,2 Metal oxide semiconductor (MOS) transistors fabricated with WS2 have demonstrated superior electrostatics and device performance.3 However, a major challenge with WS2 is the substantial Schottky barrier height (SBH),4,5 which hinders the achievement of an ideal ohmic metal–semiconductor interface for WS2 MOS transistors. Various strategies have been investigated to mitigate the SBH issue in TMDs through the optimization of metal contacts.6−13 Unfortunately, these efforts have not been successful, leading to the exploration of innovative techniques such as interface geometry engineering, edge contacts,14−16 phase engineering,17,18 and the integration of thin insulating layers.19−21 Although edge contacts can significantly reduce contact resistance, their application in thin 2D materials is limited by the restricted contact area.22 Over the years, numerous methods have been examined to refine the metal-TMD interface, but these have yielded limited success. In traditional bulk CMOS technologies, doping is used to convert the Schottky contacts into ohmic contacts. However, conventional doping techniques such as diffusion or ion implantation are unsuitable for atomically thin 2D semiconductors like WS2. Molecular doping, a technique adopted by industry, offers a promising solution. Yang et al.4 demonstrated successful chlorine (Cl) doping, achieving contact resistance (Rc) values of 700 and 500 Ω μm for MoS2 and WS2, respectively. Ab initio (DFT) simulations were employed to investigate the underlying physics of this observation.23 Despite these efforts, Rc has yet to reach the quantum limit (Rc ≈ h/(2e2kF) = 0.026/ = 30 Ω μm).14,24 Therefore, there remains significant room for improvement in Rc through the use of an appropriate dopant. Notably, bromine (Br) and iodine (I) belong to the halogen group and are positioned adjacent to Cl in the periodic table. Their larger atomic size suggests the potential for further improvement in SBH, with iodine being a more suitable dopant compared to bromine, given bromine’s oxidizing and toxic nature. When molecular doping is discussed in 2D materials, doping can ideally occur through substitution or adsorption, as 2D materials are highly sensitive to their environment. Adsorption can take place on a pristine surface or above a sulfur vacancy, with sulfur vacancies being thermodynamically more stable compared to tungsten vacancies. Numerous adsorption studies have been conducted on both MoS2 and WS2 with various entities such as gaseous molecules, transition metals, hydrogen, graphene, etc.25−29 However, the analyses of substitution and adsorption have not been presented together in a comprehensive manner. This work uses a unified computational setup to provide a comparative understanding of iodine substitution doping and adsorption on a WS2 monolayer. Iodine substitution has already been demonstrated as an efficient n-type dopant compared to chlorine by Santhia et al.30 It was also observed that when MoS2 and WSe2 FETs operate in an iodine gas-filled environment, they show p-type behavior.31,32 This prompted us to investigate the physics of iodine adsorption on pristine WS2 and at sulfur vacancies, and to compare these effects to substitutional doping. It is observed that iodine atom substitution and adsorption over sulfur vacancies result in n-type doping, whereas adsorption on pristine WS2 results in p-type behavior due to charge flow in different directions. However, due to Fermi-level pinning, attaining p-type behavior after interaction with metals is very challenging in WS2.33,34 Achieving efficient p-type contacts is essential for designing an effective PMOS transistor, as both n- and p-type MOS transistors are equally important in CMOS technology. In the literature, different techniques have been used to improve the p-type contacts in WS2 transistors, including the use of semimetal contacts like NiTe2 or ferromagnetic contacts such as VS2,34,35 grown directly over the WS2 layer through vdW epitaxy. While these methods are effective, they utilize non-CMOS-compatible materials. Our objective is to investigate ways to enhance the widely used Au-WS2 contact for p-type conduction, aiming to facilitate CMOS technology fabrication. This approach enables the design of both p- and n-type MOS transistors using the same channel and contact materials. Additionally, various reports could be found in the literature where dopants such as nitrogen atoms,36,37 WOx,38 and CH39 in the channel region are used to enhance p-type conduction. However, when doped WS2 forms contact with metals, due to Fermi-level pinning, the doped WS2 loses its effect; intermediate layers have been introduced in the present work to handle this issue. High-work function metals have been used with these intermediate layers to further enhance the performance of p-type MOS transistors. This research article addresses the problem of p-type contact for WS2 and provides an efficient p-type contact using halogen adsorption with intermediate layers for which the electronic properties of the halogen adsorbed WS2 monolayer, the influence of adsorption on contact properties of WS2–Au interface, and effect of intermediate layers have been explored. Finally, the article concludes by summarizing the significant findings of the present study. Computational Methods The simulations are conducted using the Quantum Atomistic Toolkit (Quantum ATK),40 employing the Generalized Gradient Approximation (GGA) with the Perdew–Burke–Ernzerhof (PBE) semilocal exchange-correlation functional for electronic structure calculations.41 Electronic properties are determined using the Linear Combination of Atomic Orbitals (LCAO) basis set by solving the Kohn–Sham equations. To achieve higher accuracy and efficiency, Pseudodojo pseudopotential is used, which includes multiple projectors, semicore states, and nonlinear core correction. A mesh cutoff of 45 hartree is set, with K-point sampling of 3 × 3 × 1. For atomic-position optimization, the force per atom threshold and stress error tolerance are set at 0.01 eV/Å and 0.1 GPa, respectively. Prior to electronic-structure calculations, all systems are optimized while keeping the lattice parameters constant. Additionally, to address noncovalent interactions, the GGA functional is complemented with a damped van der Waals (vdW) correction using the DFT-D2 method developed by Grimme for improved computational efficiency.42 This vdW method accounts for long-range dispersion interactions, providing a more comprehensive treatment of the vdW forces in the system. Each simulation cell is created by expanding the monolayer WS2 into a 4 × 4 supercell with lattice parameters of at least 10 Å in length. A buffer vacuum space of at least 20 Å is included along the nonperiodic z-axis to prevent unintended interactions between adjacent unit structures. To maintain the metallic properties of the metal crystals, 5 layers of metal are chosen to interface with the monolayer WS2. The first three layers of metals, along with the monolayer WS2, are left free for optimization, while the last 2 metal layers are kept fixed to prevent interactions with the vacuum and surface reconstruction. Due to the difference in lattice parameters between the metals and WS2, lattice strain is introduced into the combined system. Ultimately, the constructed combined supercell had a lattice strain of less than 2% to balance computational efficiency and accuracy. This methodology is partially motivated by the computational method employed in the study of metal-TMD contacts by Kang et al.15 In designing these combined supercells, Au(111) and Pt(111) cleaved surfaces are used as they provide minimal lattice strain. Additionally, combined supercells of WS2 with graphene and h-BN intermediate layers are all designed with strain <2%. The WS2 monolayer is kept stress-free in all interface structures. In the doped WS2, adjacent dopant sites are kept sufficiently far apart (more than 10 Å), reducing mutual interaction and preventing the formation of artificial dopant clusters, i.e., 1 dopant per 4 × 4 supercell. This arrangement enhances stability and guards against lattice distortion. In the case of substitution, the dopant has been substituted in place of the sulfur atom, whereas in adsorption the dopants have been adsorbed over the sulfur atom as it is the most stable position. However, it is important to note that opting for a greater inter-dopant distance or a supercell with lower doping concentration could lead to the creation of a substantially larger commensurate metal–semiconductor supercell, making them impractical for simulation. Results and Discussion Electronic Properties of the Adsorbed and Doped WS2 Monolayer Figure 1 displays the ball and stick model of the WS2 monolayer with (a) I atom adsorption and (b) substitutional doping of an iodine (I) atom by replacing a sulfur (S) atom. We have used a single dopant per 4 × 4 supercell to maintain uniform doping density in the case of both adsorption and substitution. However, iodine is more stable in the I2 form compared to the I atom; hence, a comparative analysis of I2 and I atom adsorption has been depicted in Supplementary Figure S1, where it is observed that I2-adsorption also provides a weaker p-type behavior compared to I atom adsorption. Optimized structure of (a) I-adsorbed WS2 monolayer and (b) I-doped WS2 monolayer. The yellow-colored points represent the sulfur (S) atoms, blue represents the tungsten (W) atoms, and purple color is for iodine(I) atoms. The doped I atoms are encircled by red lines. Figures 2 and 3 show the projected density of states (PDOS) and band structure with a Fermi level alignment for I-adsorbed and I-doped WS2 monolayers, respectively. From the PDOS and band structure, it can be observed that when an I atom is adsorbed on the pristine WS2 monolayer, it exhibits p-type behavior as the Fermi level is closer to the valence band (VB). Conversely, when the same I atom substitutes for an S atom or adsorbs over an S-vacancy, the Fermi level lies closer to the conduction band (CB), indicating n-type behavior. A similar doping analysis has been performed for different TMDs; the results have been included in Supplementary Table S1, where all TMDs show n-type behavior with I-doping and I-adsorption shows p-type behavior. This behavior can be attributed to the different natures of I atom interactions. When an I atom is adsorbed on pristine WS2, it acts as an electronegative atom with seven valence electrons, requiring only one electron to complete its octet and achieve stability. Consequently, it draws electrons from the WS2 monolayer, resulting in p-type behavior. In contrast, when the I atom substitutes an S atom, the extra electron of the I atom, which has seven valence electrons compared to the six of the S atom, does not participate in bond formation. Instead, the I atom loses its charge to the WS2 monolayer, displaying electropositive behavior. Similarly, in the case of I-adsorption over an S-vacancy, the I atom interacts with the WS2 vacancy and loses its charge. This explanation is supported by the Mulliken analysis results for doped WS2 monolayers, as shown in Table 1, which represents the outer shell charge on each atom. PDOS of, (a) I-adsorbed WS2 monolayer, (b) I-doped WS2 monolayer, and (c) I-adsorbed WS2 monolayer at S-vacancy (Svac). I-doped WS2 and I-adsorbed WS2 at Svac show n-type behavior whereas I-adsorbed WS2 shows p-type behavior due to different charge flow directions. Band structure of WS2 monolayer (dark red line) with the intrinsic level in red dash-dot line. The Fermi levels of I-adsorbed WS2 monolayer, I-doped WS2 monolayer and I-adsorbed (at S-vacancy) WS2 monolayer are depicted in green dashed line, blue dashed line and brown dashed line, respectively. Substitution of I atom and adsorption of I atom at S-vacancy gives an n-type monolayer, whereas adsorption of I atom gives a p-type WS2 monolayer. Table 1Mulliken Population Analysis of Pristine and Doped WS2 MonolayersStructureWSIpristine WS213.6416.180 I-doped WS213.6406.1857.027I-adsorbed WS213.6426.1757.123I-ads at Svac WS213.6406.1857.027 Mulliken Population Analysis of Pristine and Doped WS2 Monolayers From Table 1, it is evident that in the case of substitutional doping and I-adsorption with an S-vacancy, the charge on the I atom decreases, resulting in an increased charge on the S atom, which leads to n-type doping. In contrast, for the I-adsorbed case, the charge on the I atom increases while the charge on the S atom decreases, converting the doped WS2 monolayer to p-type. This highlights the different behavior of the I atom in adsorption compared to substitution. A similar interpretation can be obtained using the spin-polarized analysis, the results have been depicted in Supplementary Figure S2. Additionally, to assess the stability of these structures, we calculated the adsorption energy by using eq 1. Here, Eres represents the resultant energy used to evaluate the stability of the final structure, EI-ads/I-dop WS2 denotes the total energy of the I-adsorbed or I-doped WS2 monolayer, and Epristine/Svac WS2 represents the total energy of the pristine WS2 monolayer for adsorption energy calculations. In the case of I-doped and I-adsorbed structures with an S-vacancy, it refers to the total energy of the WS2 monolayer with an S-vacancy. Finally, EI-atom is the total energy of an isolated I atom.43 A negative value of Eres indicates that the structure is physically stable. Using eq 1, the resultant energies are −0.54, −1.58, and −1.59 eV for the I-adsorbed, I-doped, and I-adsorbed on S-vacancy WS2 monolayers, respectively. Since all these values are negative, both adsorption and substitutional doping are feasible methods to tune the electronic properties of the WS2 monolayer. Given that substitutional doping and adsorption of I atom on S-vacancy exhibit similar behavior, only substitutional doping and adsorption on pristine WS2 have been explored further due to their contrasting effects. Influence of Adsorption and Substitution on Contact Properties of the WS2–Au Interface The stability of the interface is evaluated by work of separation (WOS) analysis, given by eq 2. Here, EAu, EI-WS2, and EI-WS2–Au represent the total energies of gold (Au), the iodine-doped WS2 monolayer, and the iodine-doped WS2–Au interface structure, respectively, while A denotes the surface area. The more negative the WOS is, the more stable the interface. The WOS for the I-doped and I-adsorbed WS2–Au interfaces are −7.188 and −7.057 J/m2, respectively. The negative values of the WOS indicate that both the I-doped and I-adsorbed WS2–Au interfaces are stable and practically feasible. From Figure 4, it is evident that the electron density at the I-adsorbed WS2–Au interface is significantly lower compared to that at the I-doped WS2–Au interface, indicating the p-type behavior of the former. Moreover, the electron potential is higher at the I-adsorbed WS2–Au interface than at the I-doped WS2–Au interface. Importantly, neither interface exhibits a tunnel barrier, as their potential lies below the Fermi level (0 eV), as illustrated in Figure 5. Electron density of doped WS2–Au interface structures, (a) comparative view of pristine, I-doped and I-adsorbed WS2–Au interface electron densities. (b) Zoomed image of the interface, highlighted by a blue box in (a). A clear reduction in electron density can be observed in the I-adsorbed WS2–Au interface. Average effective potential of doped WS2–Au interface structures, (a) comparative view of pristine, I-doped, and I-adsorbed WS2–Au average effective potentials, (b) zoomed image of the interface, highlighted by a black box in (a). The interfaces are free of tunnel barriers, as the interface potentials are less than the Fermi level. Table 2 presents the interface Mulliken analysis, revealing that the charge on the I atom in the I-doped interface is slightly lower than in the I-adsorbed interface, although the charges remain comparable. Meanwhile, in Figure 6, the projected density of states (PDOS) analysis of doped WS2–Au interfaces shows that the Fermi level of the I-adsorbed interface aligns closer to the intrinsic Fermi level toward the conduction band (CB), rather than exhibiting clear p-type behavior. The relative positions of the Fermi levels for the I-doped and I-adsorbed WS2–Au interfaces are illustrated in Figure 7. Notably, Figure 7 demonstrates that both Fermi levels reside closer to the CB, indicating that I atom adsorption does not confer an advantage, as discussed previously. Table 2Mulliken Population Analysis of Doped WS2–Au InterfacesStructureWSIWS213.5796.208 I-doped WS213.6166.2077.079I-adsorbed WS213.6166.1997.108 Mulliken Population Analysis of Doped WS2–Au Interfaces PDOS of, (a) I-adsorbed WS2–Au interface and (b) I-doped WS2–Au interface. The p-type behavior of I-adsorption is not prominent in (a) due to Fermi-level pinning. Fermi level alignment of the I-adsorbed WS2–Au interface in green dashed line and the I-doped WS2–Au interface in blue with the WS2 band structure (dark red line) and the intrinsic level (red dash-dot line). For both the I atom-substituted and adsorbed WS2–Au interface, Fermi levels lie closer to the CB, giving n-type behavior due to Fermi-level pinning. These results can be attributed to the interaction of Au layers with the WS2 monolayer, leading to the formation of metal-induced gap states (MIGS) that trap charges and prevent significant shifts in Fermi levels upon doping, thereby causing Fermi level pinning. Consequently, even after I-adsorption, the interface fails to exhibit a p-type behavior. To mitigate the Fermi level pinning issue, intermediate layers have been introduced between the Au layers and WS2 monolayer to shield the effect of Au layers on the WS2 monolayer and preserve the p-type behavior. Intermediate Layer Analysis To screen the effect of Au-layers, a graphene (Gra) monolayer is used as an intermediate layer between Au-layers and the WS2 monolayer, as depicted in Figure 8. From Figure 9, it is clearly evident that the interaction of graphene monolayer with doped WS2 does not change the doping behavior, i.e., the I-doped Gra-WS2 interface shows n-type behavior whereas the I-adsorbed Gra-WS2 interface shows p-type behavior. The Fermi level of WS2-Gra-Au has been depicted in Supplementary Figure S3, where it can be observed that the WS2-Gra-Au interface provides an n-type behavior. Figure 10 illustrates the PDOS for I-adsorbed and I-doped WS2-Gra-Au interfaces. Here, the Iads-WS2-Gra-Au interface displays p-type behavior, with the Fermi level closer to the Valence Band (VB), while Idop-WS2-Gra-Au shows n-type behavior, with the Fermi level closer to the conduction band (CB). This confirms that the graphene monolayer effectively screens the effects of Au-layers. Figure 11 demonstrates a noticeable reduction in electron density at the I-adsorbed WS2–Au interface upon insertion of the graphene layer, indicating an enhanced p-type performance. Additionally, Figure 12 shows an increase in the effective potential for the graphene-inserted I-adsorbed WS2–Au interface, suggesting a reduced barrier for holes. Figure 13 reveals a significant downward shift (471 meV) in the Fermi level of the I-adsorbed WS2–Au interface after graphene monolayer insertion, indicative of a reduced hole Schottky barrier height. Moreover, the Fermi level of the I-adsorbed WS2-Gra-Au lies below the intrinsic Fermi level of monolayer WS2, establishing the interface as p-type. I-adsorbed WS2-Gra-Au interface structure. PDOS of, (a) I-adsorbed WS2-Gra interface and (b) I-doped WS2-Gra interface. The interaction of the graphene layer with doped-WS2 does not change the doping behavior of doped-WS2. PDOS of, (a) I-adsorbed WS2-Gra-Au interface and (b) I-doped WS2-Gra-Au interface. A clear Fermi level depinning can be observed from the above figures by inserting graphene intermediate layers. Electron density of I-adsorbed WS2 interface structures, (a) comparative view of the WS2–Au, I-adsorbed WS2–Au, and I-adsorbed WS2-Gra-Au interfaces, (b) zoomed-in image of the interfaces, highlighted by the black box in (a). A noticeable reduction in electron density is observed for the I-adsorbed WS2-Gra-Au interface. (a) Average effective potential of WS2–Au, IadsWS2–Au, and IadsWS2-Gra-Au interfaces, (b) zoomed image of the interface region highlighted by a black box in (a). The interfaces are free of tunnel barriers as the interface potentials are less than the Fermi level. Fermi level alignment of I-adsorbed WS2–Au interface in green dashed line, I-adsorbed WS2-Gra-Au interface in blue dashed, and intrinsic Fermi level of monolayer WS2 in red dash-dot lines with the band structure of WS2 shown in dark red. The results show a significant downward shift in the Fermi level after graphene insertion due to Fermi-level depinning. Similar to graphene, h-BN also demonstrates Fermi-level depinning and does not change the doping behavior of I-adsorption, as shown in Figure 14. The Fermi levels for I-adsorbed WS2-Gra-Au and I-adsorbed WS2-hBN-Au interfaces are −4.689 eV and −4.682 eV, respectively. We also investigated fluorine (F) adsorption, taking advantage of its higher electronegativity. PDOS of (a) I-adsorbed WS2-Gra interface and (b) I-adsorbed WS2-hBN interface. Both graphene and h-BN layers give p-type behavior on interaction with I-adsorbed WS2 monolayer. As shown in Table 3, the charge on the F atom is significantly higher compared to the I atom in the F-adsorbed WS2-Gra-Au interface, as it draws charge from the W atom. The Fermi level of I-adsorbed WS2-Gra-Au is −4.689 eV, whereas, for F-adsorbed WS2-Gra-Au, it is −4.799 eV, indicating a Fermi level shift of 572 meV toward the valence band (VB) from the I-adsorbed WS2–Au interface, due to the higher electronegativity of the F atom. Additionally, using a monolayer graphene has shown a significant depinning effect. When bilayer graphene (Bi_Gra) is used, the Fermi level shifts even further toward the VB. The Fermi level for the I-adsorbed WS2–Bi_Gra-Au interface is −4.842 eV, and for the F-adsorbed WS2–Bi_Gra-Au interface, it is −5.027 eV. This results in a Fermi level shift of 624 and 809 meV, respectively, from the I-adsorbed WS2–Au interface. For better understanding, Table 4 shows the Fermi levels of WS2–Au interfaces with I atom and F atom adsorption, along with monolayer and bilayer graphene intermediate layers. Further, the analysis of multiple graphene intermediate layers has been included in Supplementary Table S2. Table 3Mulliken Population Analysis of Doped WS2-Gra-Au InterfacesStructureWSIFCWS2-Gra-Au13.6186.183 4.017Iads-WS2-Gra-Au13.6406.1847.124 4.006Fads-WS2-Gra-Au13.6216.185 7.4954.007 Mulliken Population Analysis of Doped WS2-Gra-Au Interfaces Table 4Fermi Levels of WS2–Au Interfaces with Graphene Intermediate LayersWS2–Aumonolayer graphenebilayer grapheneI-adsorbed–4.689 eV–4.842 eVF-adsorbed–4.799 eV–5.027 eV Fermi Levels of WS2–Au Interfaces with Graphene Intermediate Layers Exploring Halogen Adsorption with High Work-Function Metal Up to this point, all of the explored interfaces have utilized Au-layers. However, considering the higher work function of metals like platinum (Pt), according to the Schottky–Mott rule, we would expect to observe p-type behavior. Unfortunately, due to Fermi-level pinning, achieving this has been challenging. In this section, we investigate the impact of halogen adsorption on WS2–Pt interfaces and explore the modification of these interfaces using the graphene intermediate layer to enhance p-type behavior. The Fermi level of the WS2–Pt interface is at −4.588 eV, which is above the intrinsic level of the WS2 monolayer. However, with I adsorption, the Fermi level shifts below the intrinsic level, as shown in Figure 15. From Figure 15, it can also be observed that the insertion of a graphene layer shows significant improvement, especially with F-adsorption. To provide a better understanding, Table 5 summarizes the Fermi levels of the different interfaces. As shown, the Pt metal causes a downward shift in the Fermi level, with the best results obtained for the F-adsorbed WS2-Gra-Pt interface. The F-adsorbed WS2-Gra-Pt interface demonstrated a 917 meV Fermi level shift toward the valence band (VB) compared to the I-adsorbed WS2–Au interface, resulting in a minimal hole Schottky barrier height. This result may be further improved by using multilayer graphene. Fermi level alignment of the WS2–Pt interface with halogen adsorption. The band structure of WS2 and the intrinsic level of WS2 are shown in dark red and red dashed-dotted lines, respectively. The green dashed line shows the F-adsorbed WS2-Gra-Pt Fermi level, and the blue dashed line shows the I-adsorbed WS2-Gra-Pt interface Fermi level. From the results, clear p-type behavior is evident for both I-adsorbed and F-adsorbed WS2-Gra-Pt. Table 5Fermi Levels of Different WS2 InterfacesWS2 interfacesWS2I-ads WS2WS2-GraI-ads WS2- GraF-ads WS2-GraAu–4.159–4.218–4.473–4.689–4.799Pt–4.588–4.676–4.791–4.754–5.135 Fermi Levels of Different WS2 Interfaces Conclusions To summarize, this study offered valuable insights into utilizing substitutional and adsorption-based doping of halogens (I or F) to enhance the contact properties of WS2-based transistors. We identified distinct doping characteristics with different techniques: substitutional doping of halogens yielded n-type behavior, while adsorption of halogens on WS2 resulted in p-type behavior. Given the need for better p-type contacts, we focused our study on halogen adsorption. However, achieving consistent p-type behavior through halogen adsorption proved challenging due to Fermi-level pinning at the WS2-metal interface. To address this issue, we introduced graphene and h-BN intermediate layers, which effectively depinned the Fermi level and significantly enhanced p-type behavior. The F-adsorbed WS2-Gra-Pt interface exhibited excellent p-type contact with a substantial reduction in the hole Schottky barrier height, making it ideal for efficient WS2-based PMOS transistors in CMOS technology. This approach holds great promise for obtaining p-type MOS transistors in 2D materials, which has been a significant challenge for the electronics community. Acknowledgments This work was supported by the Ministry of Education (Government of India) through the Prime Minister’s Research Fellowship (1701246). Supporting Information Available The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.4c08463. Fermi level alignment of the I-adsorbed and I2-adsorbed WS2 monolayer (Figure S1), spin-polarized PDOS of I-doped and I-adsorbed WS2 monolayer (Figure S2), Fermi level alignment of WS2-Gra-Au and I-adsorbed WS2-Gra-Au (Figure S3), Fermi levels of doped TMDs (Table S1) and Fermi levels and distance between WS2 and Au layers with different numbers of graphene intermediate layers (Table S2) (PDF) Fermi level alignment of the I-adsorbed and I2-adsorbed WS2 monolayer (Figure S1), spin-polarized PDOS of I-doped and I-adsorbed WS2 monolayer (Figure S2), Fermi level alignment of WS2-Gra-Au and I-adsorbed WS2-Gra-Au (Figure S3), Fermi levels of doped TMDs (Table S1) and Fermi levels and distance between WS2 and Au layers with different numbers of graphene intermediate layers (Table S2) (PDF) The authors declare no competing financial interest. Supplementary Material ao4c08463_si_001.pdf References Associated Data Supplementary Materials ao4c08463_si_001.pdf