2D Material-Based Memristor Arrays for Flexible and Thermally Stable Neuromorphic Applications

The practical implementation of 2D material-based memristors for various applications, such as neuromorphic computing and flexible electronics, faces several challenges in scalability and thermal durability. Here, a large-area, flexible, and thermally durable electrochemical metallization memristor array is demonstrated based on an engineered Al₂O₃/MoS₂/graphene heterostructure. A thin Al₂O₃ layer and graphene serve as dual barriers, effectively suppressing the unintentional metal ion diffusion that typically degrades device reliability under thermal and electrical stress. The fabricated device exhibits excellent uniformity across a 2-inch wafer and maintains stable switching characteristics under mechanical bending down to a 1 mm radius. Finally, simulations based on the device's experimental characteristics confirm its high potential for neuromorphic computing systems, achieving high pattern recognition accuracy. This study provides a comprehensive design strategy for realizing highly reliable, multifunctional memristors for future brain-inspired and flexible electronics.