Photoelectric In-Memory Logic and Computing Achieved in HfO2-Based Ferroelectric Optoelectronic Memcapacitors

We experimentally reported the photoelectric in-memory logic and computing capabilities of HfO2-based ferroelectric optoelectronic memcapacitors (FOMs). By optimizing the annealing process at 600 °C, we achieve a robust noise margin and a substantial photoelectric memory window exceeding 9 fF/
$\mu $
m2, accompanied by an exceptionally low static energy cost below an attojoule per operation. Subsequently, our FOMs gained the capability for reconfigurable photoelectric in-memory Boolean logics of “NAND/NOT” and neuromorphic computing-based hand-written digit image recognition with an accuracy of 92%. These results signify a significant advancement towards energy-efficient sensing-memory-computing electronics, promising next-generation applications in pattern recognition, machine vision, and beyond.