Switching characteristics and simulated iodine vacancies distribution of halide perovskite RRAM

Although the simulation research of metal oxide resistive random access memory (RRAM) is ample, the simulation work for halide perovskite RRAM is still absent. An analytic model of halide perovskite RRAM for its switching behaviors is established using kinetic Monte Carlo (KMC) simulation method. The strong conductive filament (CF) is found to be related to macroscopic abrupt switching characteristics by fitting the current-voltage (I-V) characteristics of forming, SET and RESET processes for three current compliance $(I_{cc})$ . Furthermore, the activation energy $(E_{a})$ of halide vacancies is lower than the one of oxygen vacancies and can be tuned by solvent annealing method. Therefore, abrupt and analog switching behaviors are investigated as a dependence on $E_{a}$ based on this model, which could provide the process guideline for the research of halide perovskite RRAM aiming at both data storage and neuromorphic computing.