Physical Mechanisms and Applications of High-Durability Polymer- Based Artificial Synapses Using P3DT

Organic polymers are regarded as promising candidates for constructing low-cost and flexible neuromorphic computing hardware due to their designable structures, intrinsic flexibility, and compatibility with conventional solution-processing techniques. Such materials hold significant potential for overcoming the limitations of traditional silicon-based hardware in biomimetic intelligent systems. In this study, a highly stable and high-performance memristor was fabricated and characterized based on poly(3-decylthiophene) (P3DT). The device operates at a milliampere-level current, exhibiting good signal-driving capability and circuit compatibility. Electrical measurements confirm that the memristor can successfully emulate various essential synaptic behaviors, demonstrating its potential as a fundamental building block for neuromorphic computing. Furthermore, the device shows remarkable long-term stability, maintaining its resistive switching and conductance modulation characteristics after 150 days of storage in an unencapsulated environment. To validate its applicability, an artificial neural network (ANN) model was constructed based on the conductance modulation behavior of the device and applied to the Modified National Institute of Standards and Technology (MNIST) handwritten digit recognition task. Without extensive optimization, the network achieved a recognition accuracy of 93.8%, effectively demonstrating the feasibility and efficiency of the device in simulating neuromorphic computation. This work provides a reliable device-level solution for the development of practical neuromorphic computing hardware by realizing a memristor that combines high stability with rich bio-inspired functionalities.