Analysis and Modeling of Bipolar Resistive Switching in 2-D Graphene Electrode- Based Memristor

Abstract

wo-dimensional graphene has attracted considerable interest as an electrode material for the memristor due to its low-voltage operation and high integration density capability. For emerging graphene-electrode-based memristors, early assessment based on the theoretical study becomes increasingly important to identify performance benefits and guide experimental efforts. However, no accurate physics-based model is available for describing the bipolar resistive switching mechanism. Therefore, in this work, we develop a physics-based numerical modeling framework for the TiN/HfO textsubscript X /graphene (GE)-based memristor using the self-consistent solutions of the continuity equation, Poisson’s equation, and Fourier’s equation for Joule heating. The simulated set and reset characteristics of the GE-based memristor show the excellent match with the reported experimental results. Using the developed model, we found that the GE-based memristor could allow a lower set/reset voltage (−0.21/0.18 V), lower set/reset sneak current ( ∼ 57/211 nA), and lower set/reset transition time ( ∼ 1.1/0.74 ns) with more distinct multilevel resistance levels than that for the inert electrode (TiN and Pt)-based memristors. Overall, our work provides a physics-based simulation framework to describe the intricate switching dynamics in graphne-electrode-based memristors and highlights their superior performance compared with memristors with inert electrodes.