Advanced Predictive Modeling of Radiation Effects in ReRAM Devices based on electrical characterization augmented by imaging data

In an effort to improve the design of radiation hardened electronic components, this proposal explores the feasibility of creating predictive modeling techniques for nanoscale material properties in advanced integrated electrical devices. This study encompasses the collaborative usage of high resolution Transmission Electron Microscope (HR-TEM) data, circuit design targeted electrical data, and machine learning to generate models informed by the radiation effects observed in the collected data. The study focuses on specific resistive random access memory (ReRAM) material sets with a generalized approach exploring both empirical and compact modeling methodologies that should be applicable to other material systems. The material characterization in conjunction with the electrical data within the the contexts of the developed methodologies allow for the generation of predictive models to determine the performance of the nanoscale devices under radiated conditions. These models allow for a more deterministic approach to the design of radiation hardened circuit design.