Ferroelectric memristors based on double perovskite Bi2FeCoO6 for synaptic performance and human expression recognition storage

Abstract

This study reports for the first time the application of double perovskite thin-film devices based on the Bi₂FeCoO₆ (BFCO) compound in non-volatile ferroelectric memristors. By spin-coating BFCO onto an N-type silicon (N-Si) substrate, a P-N junction was formed, yielding a thin-film device with ferroelectric properties. The device demonstrated a maximum polarization value of 46.09 μC/cm² and a high switching ratio of 293, along with excellent long-term stability (over 7 days) and high repeatability (1000 cycles). Furthermore, we investigated the synaptic characteristics of the device, including short-term plasticity, paired-pulse facilitation, and long-term potentiation/inhibition behaviors. By designing a confusion matrix recognition scenario with a binary neural network, we validated the potential of double perovskite ferroelectric memristors in intelligent learning applications. Additionally, leveraging the synaptic plasticity of the device, we developed a modal storage memory and recognition system for human emotions. This work not only provides new insights into the development of high-performance double perovskite ferroelectric memristors but also lays the foundation for optimizing synaptic performance in intelligent learning applications.