Editor
Researchers have developed hexagonal boron nitride (h-BN)-based metal-insulator-semiconductor (MIS) memristors that operate at attojoule energy levels, aiming to enhance energy efficiency in neuromorphic computing. These memristors are fabricated using metal–organic chemical vapor deposition (MOCVD) to grow a uniform h-BN resistive switching medium directly on highly doped silicon wafers, resulting in high reliability and low variability. Electrical and nanostructural analyses reveal that the h-BN layer enables resistive switching with extremely low high resistance states, while the native SiOx layer on silicon helps suppress excessive current, achieving attojoule-level energy consumption. Additionally, the formation of atomic-scale conductive filaments allows for rapid response times in the nanosecond range and supports multiple resistance states, making these memristors suitable for advanced neuromorphic applications. This advancement has the potential to significantly reduce energy consumption in neuromorphic devices, narrowing the gap between artificial and biological synapses.
ChatGPT
