Ab initio simulations of 2D materials, interfaces and devices for beyond CMOS applications

The growing needs of the semiconductor industry are pushing Silicon based transistor devices to their ultimate limits. As scaling continues, novel device concepts are being explored to improve performances and increase transistor densities. One such concept is the Tunnel Field Effect Transistor (TFET), which has been shown to alleviate the power consumption issue and offer steep subthreshold slopes. However, improving device performances not only requires new device concepts, but also demands the exploration of alternate channel materials that can cope with the scaling requirements of the future devices. The current work explores Transition Metal dichalcogenides (TMDs), which are van der Waals (vdW) layered structures. These materials, by their design, offer the promise of ultrathin channels that could potentially resolve scaling issues. In this thesis, we use state-of-the-art simulation techniques and frameworks to explore fundamental properties of materials that ultimately impact on the device performances. First-principles techniques are used to explore the chemical and electronic properties of TMDs of interest, such as MoS2 and WS2. While initially 2D materials were considered for their ultrathin nature and lack of dangling bonds at the surface, as research continued, it turned out that they suffer from various issues that affect their properties in device applications. It has indeed been shown that grown or exfoliated 2D materials constantly suffer from issues pertaining to defects. These defects not only make 2D materials susceptible to react with various kinds of ambient chemical species, they also directly impact on their electronic properties, which are at the heart of any device application. This thesis is divided into different parts dealing with the various stages of material growth and device integration. Density Functional Theory (DFT) is used to study the impact of various point defects, such as chalcogen-related defects, metal-related defects and oxidized defects, in individual materials. Our results show that defects in 2D materials play an important role in determining their chemical reactivity and electronic properties. They can also lead to the natural doping of TMDs. We next present the thermodynamic modelling of the MOCVD growth of 2D materials such as MoS2 and WS2. While investigation of isolated defects in individual materials is rather straightforward, the analysis of the entire growth process in a reaction chamber, over wide temperature and pressure ranges, is a difficult task. This usually involves the occurrence of various convoluted effects, originating from the combined interaction of material defects and chemical species present in the reaction chamber. To assess the equilibrium composition of various materials in the reaction chamber and identify species crucial to the growth process, a Gibbs energy minimization procedure is used, accounting for temperature and pressure conditions which are typically encountered during growth processes. In addition, the interaction/reactivity of 2D materials with gate dielectrics and various supported substrates (used as templates for their growth) is also investigated, using the Gibbs energy minimization scheme. Our simulations reveal that several volatile species are generated during the growth process, which can impact the quality of the grown 2D material or gate dielectric. In addition, simply "contacting" TMDs with high-κ oxides also impacts the quality of the 2D materials, due to the incorporation of substitutional defects. Another important requirement for high performing 2D materials is our ability to successfully and controllably dope them. However, the ultrathin nature of TMDs makes them prone to damage during the doping process. To avoid this, different techniques have been proposed in the literature. In this work, we explore two of these techniques: 1) Doping via electrostatic interaction of Self-Assembled monolayers (SAMs) with the 2D materials and 2) Incorporation of dopants during the growth process by using selected precursors. Our first-principles simulations confirm that it should be possible to alter the electronic properties of TMDs and dope them either by electrostatic interactions with SAMs or by (in-situ) dopant incorporation during the growth process. Lastly, we also investigate the contact resistivity at metal-2D semiconductor junctions, which is an important limiting factor for the device performances. Our theoretical results indicate that there is a fundamental limit to the contact resistivity that can be achieved with 2D materials. We also show that the metal effective masses and fermi energies are crucial to the understanding and improvement of the metal-TMD contacts.status: publishe