Atomic-level Investigation and Proposals to Address Technological Roadblocks and Reliability Challenges in 2D Material Based Nanoelectronic Devices

Abstract

The transistor scaling is witness to many extraordinary inventions during its consecutive miniaturization. The journey began from Dennard’s classical constant field scaling, crossing through the milestones like strain engineering, high ‘k’ gate dielectric, ultrathin body transistor (UTB), silicon on insulator (SOI), and multi-gate 3D architectures, and continues in the form of advanced FinFET technology. However, further downscaling is sensing a dead-end because of the various challenges due to fundamental limitations of silicon, the building material of the transistor. Among these, two significant challenges are mobility degradation due to boundary scattering by surface dangling bonds and loss of gate control due to quantum confinement. To keep downscaling alive, the research community is looking for an alternate material that can mitigate these issues and consist of better fundamental properties from silicon like intrinsic mobility, thermal conductivity, optical response, and mechanical strength. Two-dimension material (2D material) shows great potential for next-generation electronic material and provides multiple avenues for further exploration. The material is one or a few atomic-layer 2D thin sheets of covalently bonded atoms stacked using weak van der Walls (vdW) forces in the third dimension. The lack of surface dangling bond and atomic-scale thickness mitigates the significant challenges of low mobility and inadequate gate control of the silicon material, respectively. Presently, more than 150 materials exist in the 2D material family. Graphene, Transition Metal Dichalcogenides (TMDs), and Phosphorene are well ahead of other family members due to their extraordinary properties, thereby plenty of investigations. Despite these properties, the materials have several roadblocks to their technological application. Opening bandgap and minimizing contact resistance are significant challenges in graphene, and reducing contact resistance and mature growth and reliability are big concerns for the TMDs. Phosphorene, which has hybrid properties of graphene and TMDs, is relatively less explored due to its spontaneous degradation in the ambient environment. Understanding and mitigating its spontaneous ambient degradation is still an open challenge for the electronics and material research communities across the globe. Keeping in mind these limitations, we explore the problems one by one and find their reasonable solutions. Based on DFT investigations, the discussion begins with a proposal for a reliable direct bandgap opening technique in graphene. Graphene possesses zero bandgap due to its highly symmetric hexagonal structures, which touch its π and π* orbitals’ energy states near the Fermi level, known as the Dirac point. Breaking this symmetry by carbon vacancy or Stone-Wales (SW) defects opens the bandgap at the Dirac point. However, the carbon vacancy creates unwanted mid-gap (trap) states, attributed to unbound orbitals of the nearest unsaturated carbon atoms at the vacant site. Moreover, the unsaturated carbon atoms react with ambient gases like oxygen, making graphene unstable. Interestingly, hydrogenation or fluorination of the unsaturated carbon atoms near the vacant site helps prevent the trap states while contributing to promising direct band gaps in graphene. The opened bandgap is tunable in the infrared regime and persists for different sizes and densities of hydrogenated or fluorinated patterns. The proposed approach is thermodynamically favorable as well as stable. The next work demonstrates the contact resistance reduction of graphene with palladium (Pd) by carbon vacancy engineering. The discussion begins with fundamental insights into the Pd-graphene interface and carbon vacancy-assisted contact resistance reduction using Density Functional Theory (DFT), followed by its experimental validation by various processes. Our study reveals significant interaction of Pd with graphene. Their orbitals overlap leads to potential barrier lowering at the interface, which can be reduced further by bringing graphene closer to the bulk Pd using carbon vacancy engineering at the contacts. Thus, the carbon vacancy-assisted barrier modulation reduces contact resistance by increasing carrier transmission probabilities at the interface. The theoretical findings have been emulated experimentally by carbon vacancy engineering at the graphene Field Effect Transistors (FETs). Different contact engineered graphene devices with Pd contacts shows significant contact resistance reduction, measuring as low as ~78 Ω-µm at room temperature. The contact resistance shows a ‘V’ shape curve as a function of defect density. The optimum contact resistance achieved is significantly lower than their pristine counterpart, as predicted by the theoretical estimates. Subsequently, the journey turns towards an atomic level investigation of phosphorene ambient degradation using the first-principles Molecular Dynamics (MD) simulations in the following work. The study reveals that the oxygen molecule dissociates spontaneously over pristine phosphorene in the ambient environment resulting in an exothermic reaction, which is boosted further by increasing partial pressure, temperature, and the presence of oxygen free radicals. The surface reaction is mainly due to lone pair electrons of phosphorous atoms, making the degradation directional and spontaneous under oxygen atoms. Furthermore, water molecules play a vital role in the degradation process by changing the reaction dynamics path of phosphorene-oxygen interaction and reducing activation energy and reaction energy due to its catalyzing action. In addition, phosphorous vacancy acts as an epicenter for oxidation. The oxygen attacks directly over the vacant site and reacts faster than its pristine counterpart. As a result, phosphorene edges resembling extended vacancy are prominent reaction sites that oxidize anisotropically due to different bond angle strains. The edge initiated spontaneous degradation, and rapid oxidation under the free radicals are validated using consistent probing under an optical microscope and Transmission Electron microscope (TEM), respectively. After material exploration, the next work reveals a unique reliability issue in the Phosphorene FETs. Here, we investigate the role of channel excess holes (due to inversion) in phosphorene degradation using the first-principles MD computations and electrical and Raman characterization. The results show that phosphorene degrades faster under negative gate bias (excess hole) than in pristine conditions (unbiased). The rapid degradation is mainly due to the enhanced chemical interaction of oxygen with the available hole in the channel. The computational findings are experimentally verified over phosphorene FETs. Compared to the unbiased condition, the devices show a faster change in drain current and fast decay of all primary Raman peaks in the ambient environment under negative gate bias. At the risk of ambient degradation, phosphorene thin flakes are to be identified quickly using a non-destructive technique like Raman to make their FETs for further exploration. The next work shows that the Raman signature of a low-frequency interlayer out-of-plane phonon mode, known as breathing mode, helps in identifying the thin flake quickly. Further, the work talks about thermal evolution and estimates the first-order temperature coefficient of different breathing modes. All the captured modes show a negative temperature coefficient around -0.002-0.003 cm-1/K across different flake thicknesses. Moreover, a closer look at the thermal evolution reflects that the modes follow three-phonon and four-phonon process dominant scattering phenomena at low and high-temperature ranges. The three-phonon process scattering is dominant below ~100 K, shifting to four-phonon process dominant scattering beyond ~150 K. Besides, the work discusses pristine instrumental error in the Raman shift characterization and suggests a mitigation method using Stokes and Anti-stokes scattering lines. Finally, the last work discusses the interactions of different metals (Au, Cr, Ni, and Pd) with TMDs (MoS2, MoSe2, WS2, and WSe2). The work reflects that Au has a weak interaction with all the TMDs. Thus, it stays more than 2 Å away from the TMDs surfaces. However, other metals show strong chemistry with TMDs. Due to weak interaction, Au offers very few metal-induced gap states (MIGS) in all the TMDs. On the other hand, metals like Cr, Ni, and Pd flood many MIGS in the bandgap region of the TMDs. During interactions, all the metals offer n-type doping to TMDs. Chalcogen vacancy enhances the interaction of the metals with all the TMDs. The vacancy leaves the unbounded orbitals, which bond strongly with the approaching metals. The bonding enhancement reduces the metal-TMDs distances that can be used in contact resistance engineering in their bulk counterparts. Chalcogen interstitial impurity also enhances the bond strength of some metal-TMDs interfaces. Our journey helps in overall technological advancement in the leading 2D materials. The work digs into the leading roadblocks like contact resistance reduction and method of bandgap opening in graphene, understanding the degradation issue of phosphorene at the material and device level, and exploring metal-TMDs interactions for their contact resistance engineering.