Inverse modeling of sub-100nm MOSFET with PDE-constrained optimization

Abstract

The inverse modeling of MOSFET aims to extract the process and device parameters of a CMOS technology from electrical test data, such as the I–V curves. Unlike the parameter extraction for compact models, inverse modeling calculates the electrical characteristics with TCAD simulation instead of the analytical formulae of compact models (e.g. BSIM4). The parameters extracted from inverse modeling can be either the process parameters (e.g. Dose, energy, annealing time, etc.), or the device parameters (oxide thickness, peak doping concentration, char. length of Gaussian doping, etc.). Obviously, inverse modeling is an optimization problem to minimize the error between the simulated and the measured electrical characteristics.