Dynamical self-rectifying memristors based on halide perovskite nanocrystals

Abstract

The integration of rectifying effects with resistance switching in a self-rectifying memristor offers the opportunity to suppress the sneak current in high-density crossbar arrays for energy-efficient neuromorphic computing. Here, we report a new type of two-terminal self-rectifying memristor that gets rid of asymmetric complex structures by using CsPbBr₃ perovskite nanocrystals (NCs). The simple metal-insulator-metal (Au/CsPbBr₃ NCs/Au) configuration that eases integration exhibits multiple resistance states that can be precisely controlled by the stimulus properties and dynamical rectifying characteristics dependent on both the bias voltage and bias time. We have extended an earlier proposed theory that predicts electric-potential-distribution-controlled rectification to rationalize all the observed rectifying behavior that are regulated by mobile-ion-induced interfacial electrochemical reactions and found excellent agreement between theory and experiments. Our study thus demonstrates the possibility of constructing controllable self-rectifying memristors without involving asymmetric complex structures, paving a new way for resolving the sneak current issue in crossbar arrays of memristors.