Capacitance Characteristic Optimization of Germanium MOSFETs with Aluminum Oxide by Using a Semiconductor-Device-Simulation-Based Multi-Objective Evolutionary Algorithm Method

Abstract

This paper, for the first time, optimizes the characteristics of capacitance–voltage (C–V) of germanium (Ge) metal-oxide-semiconductor field effect transistors (MOSFETs) with aluminum oxide (Al₂O₃) by using a semiconductor-device-simulation-based multi-objective evolutionary algorithm (MOEA) technique. By solving a set of 2D semiconductor device transport equations, numerical simulation is intensively performed for the optimization of the C–V curve of Ge MOSFET devices. To optimize the capacitance of Ge MOSFETs with respect to the applied voltage, by minimizing the total errors of the C–V curve between the device simulation and a given specification (and experimentally measured data), the thicknesses of Al₂O₃ and GeO₂, the work function of gate electrodes, the distribution range of channel doping, the dielectric constants of Al₂O₃ and GeO₂, and the source/drain doping concentration are considered in the process of optimization. The semiconductor device simulation and the MOEA method are integrated and performed based on a unified optimization framework. According to the sharp variation characteristics of the C–V curve, except for using a residual sum of squares (RSS) (i.e., the sum of squares of residuals) as an objective function, physical key parts of the curve are also considered in the optimization problem. The engineering results of this study indicate that the semiconductor-device-simulation-based MOEA method shows great performance to optimize the parameters, which not only minimize the objective values but also match the curve shape.

Keywords:

• Aluminum oxide
• Design
• Device simulation
• Evolutionary
• Fitting
• Germanium MOSFETS
• Germanium oxide
• Multi-objective
• Optimization
• Process parameters
• Residual sum of squares
• Structure parameters