Simulation of the Time-Dependent Breakdown Characteristics of Heavy-Ion Irradiated Gate Oxides Using a Mean-Reverting Poisson-Gaussian Process

In this work, a compact mathematical representation of the time-dependent breakdown dynamics of heavy-ion irradiated thin oxides subjected to electrical stress is investigated. The model is formulated as a stochastic differential equation for the gate leakage current and consists of one deterministic term and two random terms. The first one deals with the short transient and evolution toward the steady-state conduction level after the triggering of a breakdown event, the second term simulates the noisy behavior of the current between the arrival of events, and the third term represents the opening of successive breakdown paths across the dielectric film. Ion hits are assumed to cause a finite number of latent damaged sites distributed throughout the oxide area, which manifest, as the electrical stress proceeds, in the long-run saturating behavior of the current-time characteristics. A nonhomogeneous Poisson counting process is used to describe the switching from latent to active spot, whereas a Gaussian process is considered for the low-level fluctuations.