Noise Spectroscopy of Dielectric Defects Using Higher‐Order Moments

As pertains to the physical comprehension of defect dynamics in large‐area dielectrics, established methodologies, such as time‐dependent dielectric breakdown (TDDB) and stress‐induced leakage currents (SILC), typically provide insufficient insights. Here, we capitalize on noise spectroscopy to investigate degradation phenomena in 5 nm plasma‐enhanced chemical vapor deposited (PECVD) SiCN films. We leverage kernel methods to estimate the exact statistical distribution of the fluctuation time series, which we subsequently analyze through its moments up to fourth order. The highest‐order moment is identified to be the most representative measure of the defects’ stress response in PECVD SiCN films, at varying voltages and temperatures, constituting a robust and reliable degradation metric.