Ferroelectric Devices

Our research aims to advance ferroelectric devices as a key platform for next-generation memory and neuromorphic AI hardware. To this end, we conduct research across a broad range of topics, from identifying the physical origins of device degradation to optimizing devices and arrays for practical applications. Specifically, we investigate the physical origins of endurance degradation and retention loss that occur during repeated operation and long-term data storage in ferroelectric devices, and clarify key degradation mechanisms such as trap generation, interface degradation, and depolarization. Based on this understanding, we optimize materials and device structures for FeNAND memory applications to improve retention and memory window. In addition, for neuromorphic applications, we develop devices capable of analog weight control through device-structure optimization and ferroelectric/channel engineering, ultimately aiming to integrate them into NVM arrays for vector–matrix multiplication operations in AI computing.