Modeling and Optimization of Variability in High-k/Metal-Gate MOSFETs

We present an in-house tool to simulate random dopant fluctuation effects on nano-scale devices with nonuniform channel doping profiles. A novel dopant discretization scheme using Poisson statistics that can achieve self-consistent median parametric values (e.g. average channel concentration) with deterministic device simulations was introduced. This capability was deployed to study the variability contributed from random dopant fluctuation and critical dimension uniformity under different work function shift in advanced transistor structures with non-uniform channels. It is concluded that work function shift can be employed to minimize the total variability.