Atom-to-circuit Modeling Strategy for 2d Transistors

Abstract

Two-dimensional materials are now being considered as viable options for CMOS (complementary metal-oxide-semiconductor) technology extension due to their diverse electronic and opto-electronic properties. However, introduction of any new material in the process integration phase of technology development in the semiconductor industry is an expensive and time-consuming affair. It is also difficult to select an appropriate 2D material from the plethora without assessing their performance at circuit level. Thus, first-principles-based multiscale models that enable systematic performance evaluation of emerging 2D materials at device and circuit levels solely from their crystallographic information are in great demand. In this thesis, such an atom-to-circuit modeling framework, addressing three different levels of abstraction (viz. material, device, and circuit), has been demonstrated. Firstly, the model was developed for a van der Waal’s heterostructure (vdWH) based all-2D metal-insulator-semiconductor field-effect transistor (MISFET), comprising of vertically stacked semi-metallic graphene, insulating hexagonal boron nitride (hBN) and semiconducting monolayer molybdenum disulphide (MoS2). Our physics-based compact model demonstrates the effects of band gap opening in graphene due to its sublattice symmetry breaking interactions with underlying hBN layer. This apart, proposed model captures the effects of semiconductor doping and the band gap variation of graphene at device and circuit levels. The model equations were thereafter implemented in a professional circuit simulator using its Verilog-A interface to facilitate design and simulation of integrated circuits. Secondly, the scope of the proposed model was further extended to capture the inertia of the charge carrier in 2D transistors operating at very high frequencies, typically greater than its intrinsic cut-off frequency. Taking phosphorene as a prototypical example, a multiscale model was developed for 2D transistors that can predict the channel-orientation-dependent high-frequency performance of devices and circuits solely from the crystallographic information of their constituent materials. The material-specific parameters obtained from density functional theory calculations were used to develop a continuity equation based non-quasi-static model to gain insight into the high-frequency behaviours. It was found that channel orientation has strong impact on both the low and high frequency transconductance parameters, however it affects only the high-frequency component of transcapacitances. The model was then implemented in industry-standard circuit simulator using the relaxation-time-approximation technique and simulations of analog and digital circuits were carried out to demonstrate its applicability for near cut-off frequency circuit operation. Finally, the idea was also exercised for modeling novel quantum materials like 2D topological insulators and it was shown that the proposed analytical approach could be useful for developing compact models of topological insulator field effect transistors. A Hamiltonian based continuum model was used to unveil the bandgap opening in the edge-state spectra of spatially confined monolayer 1T' molybdenum disulphide (MoS2), molybdenum diselenide (MoSe2), tungsten disulphide (WS2) and tungsten diselenide (WSe2). It was shown that the application of a perpendicular electric field effectuates a topological phase transition in these materials, and it can simultaneously modulate the band gaps of both bulk and edge-states for finite-width samples. The tuneable edge conductances, as obtained from the Landauer-Büttiker formalism, exhibit a monotonous increasing trend with applied electric field for deca nanometer MoS2, whereas the trend is opposite for other cases. The proposed multiscale modeling framework is ‘core’ in nature and various nonideal effects can further be included using suitable pre-correction techniques to establish a full-fledged industry-standard compact model. Since this model effectively bridges between atomistic material modeling tools and industrial design automation tools, it thereby promises solution to the design-technology co-optimization challenges for new materials.