Engineering Efficient p-Type TMD/Metal Contacts Using Fluorographene as a Buffer Layer

DOI: 10.1002/aelm.201600318 structurally stable and largely lack dangling bonds. The production processes of TMDs are currently well established, ranging from top-down exfoliation of the bulk material using mechanical exfoliation, solution-based approaches and the bottom-up synthesis methods using chemical vapor deposition.[7,8] TMDs have gained significant importance as excellent candidates for nanoelectronic applications.[9,10] MoS2 is one of the most commonly studied TMD in this regard, which demonstrates a high mobility (in the range 1–50 cm2 V−1 s−1 at room temperature[11,12]), comparable to that of silicon. In addition, field-effect transistors (FETs) based on MoS2 show low power dissipation[1] and efficient control over switching,[2] leading to widespread research interest in this topic. While these properties are certainly encouraging, one major limitation of such FETs is that the carrier transport in the semiconductor channel is mostly electron-mediated,[13] resulting in n-type FETs (n-FETs). Despite attempts to employ high workfunction metal contacts to obtain hole-based transport, the resulting devices have instead widely shown n-character.[13] This intrinsic behavior of the unmodified MoS2-metal interface hinders the construction of fully integrated circuits,[7] because of the difficulty in obtaining a CMOS (complementary metal oxide semiconductor) device, where the building block of logic gates and digital circuits require both nand p-type MOS architectures. The fabrication of p-FETs based on monolayer MoS2 is challenging[13] because of a particular interfacial phenomenon between the TMD and the metal contact, namely Fermi level pinning.[14,15] It is commonly believed that the interfacial gap states between MoS2 and the metal contacts are responsible for pinning the Fermi level close to the conduction band, even upon using a high work-function metal. These gap states may be surface states (Bardeen’s theory), metal-induced gap states (MIGS) or defect/disorder-induced gap states.[14,15] Guo, Francois and co-workers[16,17] consider the MIGS theory the best candidate to explain the origin of the gap states. A different point of view is assumed by McDonnell et al.,[18,19] wherein they attribute the difficulty in producing hole-based MoS2 devices to the intrinsic low work-function defects present in MoS2, responsible for the variability of electronic properties across the samples. These native defects such as vacancies present in MoS2 and other TMDs like WSe2 may actually result in variations of the TMDs work function, as observed in experiments.[20] P-type transistors based on high work function transition metal dichalcogenide (TMD) monolayers such as MoS2 are to date difficult to produce, owing to the strong Fermi level pinning at the semiconductor/contact metal interfaces. In this work, the potential of halogenated graphenes is demonstrated as a new class of efficient hole injection layers to TMDs such as MoS2 and WSe2 by taking fluorographene (or GF) as a model buffer layer. Using first-principles computations, two commonly obtained GF stoichiometries, C2F and CF, have been studied as buffer layers between MoS2 and Pt. In particular, for high work function TMDs such as MoS2, it has been shown that C2F forms an ohmic contact, while CF leads to a significant p-SBH value. On the other hand, for low work function TMDs such as WSe2, both C2F and CF lead to p-type ohmic contacts. This analysis shows that the ability of these buffer layers to form p-type contacts depends crucially on the charge redistribution at the GF/metal interface, which is dictated by their chemical interaction and equilibrium geometry. The fundamental electronic structures between the different semiconductor/insulator/metal interfaces which are part of this study have also been investigated.