Boron nitride memristors: from mechanism and device optimization to integrated applications.

Abstract

The relentless scaling of integrated circuits faces significant bottlenecks in conventional memory technologies. This challenge is primarily attributed to the limitations of the von Neumann architecture and the physical constraints of mainstream memory technologies. Non-volatile memristors offer compelling advantages and excellent scalability. They have an inherent ability to emulate synaptic plasticity for neuromorphic computing. Boron nitride (BN) emerges as a highly promising active layer for memristors due to its superior thermal stability, mechanical strength, chemical inertness, atomically smooth surface, and compatibility with complementary metal-oxide-semiconductor (CMOS) processing. This review systematically examines recent advancements in BN memristors, including resistive switching mechanisms, synthesis methods, and the effect of electrode contacts and device architectures on performance. It also highlights key application domains such as memory devices, neuromorphic computing and RF switches. Finally, the review identifies current challenges, including achieving large-area uniformity, precisely controlling filament dynamics, enhancing endurance/retention, and understanding complex switching behaviors. This work provides perspectives on future research directions focused on optimizing material engineering, enabling 3D integration, realizing multi-level storage, and exploring novel heterostructures to inspire the full potential of BN memristors for next-generation electronics.