Electrochemically synthesized multilayer metal–organic coordination polymer films for organic memristors with pattern recognition function

Abstract In recent years, metal–organic coordination polymer (MOCP) films have emerged as promising materials for memristor applications. However, the development of MOCP-based organic memristors remains challenging due to the complexities of synthesis, particularly for layer-by-layer (LbL) systems, which further increase the technological difficulty. In this study, we present a rapid and efficient electrochemical approach for synthesizing LbL MOCP films through a precisely controllable fabrication process. This method not only avoids the introduction of impurities but also enables precise thickness control by adjusting the reaction time, achieving a linear growth accuracy exceeding 90% for single-component films and multi-layered films. The fabricated multi-layered MOCP films exhibit good cycling endurance over 1000 V sweeps and an ON/OFF ratio of 54.4. Through continuous voltage sweeping, the multi-layered polymer-based memristor demonstrates 14 distinct linear conductance states. Moreover, a purely software-based convolutional neural network (CNN), incorporating the measured multistate conductance characteristics of the memristive devices in its simulation, achieves a high recognition accuracy of 98.5% for handwritten digits after training. These findings provide a promising pathway for the scalable fabrication of high-performance memristors with recognition functionality.