Inorganic Lead-Free and Bismuth-Based Perovskite Nanoscale-Thick Films for Memristors and Artificial Synapse Applications

Neuromorphic computing systems integrating both analog computing and data storage at the hardware level fundamentally overcome the von Neumann bottleneck. Organolead halide perovskite memristors are considered to be promising candidates to mimic biological synapses due to their highly tunable resistance states. Despite the considerable progress already made, the toxicity and instability of lead severely restrict potential applications. Herein, we successfully demonstrate all-inorganic and lead-free nanoscale Cs3Bi2X9 perovskites films. The Cs3Bi2I9-based memristors exhibited an excellent resistive switching (RS) performance, including ultralow power consumption, high environmental robustness, ultrafast switching speed, stable endurance and retention, large ON/OFF ratio, and multilevel capability. In particular, we propose an electrothermal simulation model and combine various experimental approaches to analyze the RS mechanism. More importantly, essential synaptic functions are emulated by Cs3Bi2I9-based memristors, including paired-pulse facilitation, long-term potentiation/depression, a learning-forgetting-relearning process, spike-voltage-dependent plasticity, spike-frequency-dependent plasticity, and so on. Basic photonic synaptic function plasticity was also investigated in the Cs3Bi2I9 memristors with tunable photophysical properties. Overall, this work is the original systematic study of memristors based on Cs3Bi2X9 perovskite systems and the comprehensive applications of Cs3Bi2I9 memristors in artificial neural networks, opening opportunities for ecofriendly, hardware-oriented neuromorphic computing architectures with perovskites as the functional material.

Automated summary

This study demonstrates the use of lead-free Cs3Bi2X9 perovskite films in memristors, which show excellent resistive switching performance and can emulate various synaptic functions in artificial neural networks. The research provides insights into the potential applications of perovskites in ecofriendly and hardware-oriented neuromorphic computing architectures.