Enhancing synaptic characteristics using interlayer-inserted hfo2-based memristors by multiphysics modeling

Abstract

Among resistive switching memories (memristor), valence change mechanism (VCM) memristors are seen as efficient memory components for their potential applicability to neuromorphic systems, simple structure, and low-power operation. However, there is a lack of quantitative understanding regarding the mechanism of oxygen vacancy migration in transition metal oxides, filament structure formation, and factors affecting resistance switching. In this paper, through simulation program, the modeling of resistive switching mechanism was demonstrated by inserting various oxides into HfO2-based memristor devices to elucidate the movement of oxygen vacancies. Consequently, it became possible to physically and chemically control the formation of the initial conductive filament and maximize the analog state of resistance switching. Furthermore, by expanding the optimized component configuration through simulations into a 16 × 24 crossbar array structure, it is expected to contribute to future applications in the field of neuromorphic hardware.