| Description |
This dissertation discusses the modeling and benchmarking of a series of emerging field-effect transistor (FET) devices with particular emphasis on terahertz applications. The analysis is carried out on the basis of physical modeling, numerical modeling, and when possible, via fitting of experimental data to these models. The analyzed devices rely heavily on nonconventional physical phenomena such as impact ionization, tunneling, and other quantum effects. First, I discuss the development of a continuous compact direct-current (DC) model, which is capable of describing the current-voltage characteristics of a class of Multiple-Independent-Gate (MIG) silicon FinFETs, namely Dual-Independent-Gate (DIG) FinFETs, over all its biasing design space. This model captures some of the unique features of DIG FinFETs including the dependence of its super-steep subthreshold swing on drain bias, and polarity gate bias. An excellent agreement is shown between the model and measured experimental current-voltage characteristics. Second, I introduce a distributed two-dimensional (2D) model for a graphene symmetric FET (SymFET). This model considers (a) the intra-graphene layer potential distributions and (b) the internal current flows. Our numerical results show that: (i) when the tunneling current is small, due to either a large tunneling thickness or a small coherence length, the voltage distributions along the graphene electrodes have almost zero variations upon including these distributed effects; (ii) when the tunnel current is large, due to either iv a small tunneling thickness or a large coherence length, the local voltage distributions along the graphene electrodes become appreciable, and the device behavior deviates from that predicted by 1D approximation. Third, I discuss the effect of quantum capacitance on the response of graphene-based FETs operating as terahertz detectors. We analyze and identify non-homogeneities, and impurities, which in practice lead to a finite minimum effective charge density as well as to a finite minimum conductivity in graphene, as an effect that can degrade the predicted theoretical performance of graphene-based terahertz devices. Finally, I discuss my work in the context of the state-of-the-art of terahertz detectors by benchmarking the predicted performance of these devices against theoretical limits and experimental results in traditional FETs based on various semiconductor materials. |
