IEEE Journal of the Electron Devices Society (Jan 2022)
Physics-Based Analytical Channel Charge Model of In<italic><sub>x</sub></italic>Ga<sub>1-<italic>x</italic></sub>As/In<sub>0.5</sub>2Al<sub>0.48</sub>As Quantum-Well Field-Effect Transistors From Subthreshold to Strong Inversion Regimes
Abstract
This paper presents a physics-based analytical channel charge model for indium-rich InxGa1-xAs/In0.52Al0.48As quantum-well (QW) field-effect transistors (FETs) that is applicable from the subthreshold to strong inversion regimes. The model requires only seven physical/geometrical parameters, along with three transition coefficients. In the subthreshold regime, the conduction bands ( $E_{C}$ ) of all regions are flat with finite and symmetrical QW configurations. Since the Fermi–level ( $E_{F}$ ) is located far below $E_{C}$ , the two-dimensional electron-gas density ( $n_{2{-}DEG}$ ) should be minimal and can thus be approximated from Maxwell–Boltzmann statistics. In contrast, the applied gate bias lowers the $E_{C}$ of all structures in the inversion regime, yielding band-bending of an In0.52Al0.48As insulator and InxGa1-xAs QW channel. The dependency of the energy separation between $E_{F}$ and $E_{C}$ on the surface of the InxGa1-xAs QW channel upon $V_{GS}$ enables construction of the charge–voltage behaviors of InxGa1-xAs/In0.52Al0.48As QW FETs. To develop a unified, continuous and differentiable areal channel charge density ( $Q_{ch}$ ) model that is valid from the subthreshold to strong inversion regimes, the previously proposed inversion-layer transition function is further revised with three transition coefficients of $\eta $ , $\alpha $ and $\beta $ in this work. To verify the proposed approach, the results of the proposed model are compared with those of not only the numerically calculated Qch from a one-dimensional (1D) Poisson–Schrödinger solver, but also the measured gate capacitance of a fabricated In0.7Ga0.3As QW metal-insulator-semiconductor FET with large gate length, yielding excellent agreement between the simulated and measured results.
Keywords
