Compact Modeling of Multi-Layered MoS₂ FETs Including Negative Capacitance Effect

Abstract

In this paper, we present a channel thickness dependent analytical model for MoS2 symmetric double-gate FETs including negative capacitance (NC) effect. In the model development, first thickness dependent model of the baseline 2D FET is developed, and later NC effect is included in the model using the Landau-Khalatnikov (L-K) relation. To validate baseline model behavior, density functional theory (DFT) calculations are taken into account to obtain numerical data for the K and Λ valley dependent effective masses and differences in the energy levels of N-layer (N = 1, 2, 3, 4, and 5) MoS2. The calculated layer dependent parameters using DFT theory are further used in a drift-diffusion simulator to obtain electric characteristics of the baseline 2D FET for model validation. The model shows excellent match for drain current and total gate capacitance of baseline FET and NCFET against the numerical simulation.

Modeling the Impact of Phonon Scattering with Strain Effects on the Electrical Properties of MoS2 Field-Effect Transistors

“…This model accounts for the non-idealities of the channel material and considers the impact of Schottky contact. In addition, Nandan et al [ 15 ] proposed a double gate MoS 2 FET which considered the effect of channel thickness. Additionally, Khare et al [ 16 ] reported that the threshold voltage of MoS 2 FET is influenced by the layer’s thickness.…”

Two-dimensional MoSi2N4: An Excellent 2D Semiconductor for Transistors

“…ML-MoSi 2 N 4 is an indirect band gap semiconductor with the experimental band gap value of 1.94 eV. Its elastic constant is ∼ 4 times of ML-MoS 2 , and the carrier mobilities in it are ∼ 4 times and ∼ 4-6 times of ML-MoS 2 (the most widely studied 2D material for FETs application [8], [16]- [28]). Its lattice thermal conductivities [11], [12] are approximately 1.6 times silicon (Si) and much higher than other widely known 2D semiconductors [29]- [31] (ML-MoS 2 , As, Sb, silicene and, ML-BP), but much lower than graphene [29], [32].…”