Organic Iontronic Neuromorphic Adaptive Device

Organic iontronic neuromorphic materials (OINMs) have attracted considerable attention for their mixed iontronic‐electronic conduction, a key mechanism for building brain‐inspired and energy‐efficient neuromorphic systems. However, most existing OINMs‐based artificial neuromorphic devices (ANDs) are limited to a single neuro‐synaptic function, posing a fundamental barrier to the development of multifunctional neuromorphic hardware. Herein, we report adaptive neuronal and synaptic dynamics in a single OINM‐based device through single‐variable modulation of the molecular geometric configuration. Their stacking and interfacial properties are precisely controlled, which in turn govern the device adaptivity. The device reconciles volatile neuronal behavior and non‐volatile synaptic plasticity activated by microsecond‐ and millisecond‐scale pulses, respectively. Furthermore, the interplay between the space‐charge‐limited current (SCLC) and redox mechanisms is elucidated, thereby establishing a clear structure‐property‐function relationship. Leveraging these two adaptive modes, an OINM‐activated transfer learning approach for neuromorphic computing is proposed, which achieves a favorable trade‐off between processing speed and recognition accuracy by freezing two convolutional layers on the CIFAR‐10 dataset. This work not only introduces a new family of D‐A‐type OINMs but also paves the way for hardware that natively adapts its computational timescale to diverse tasks.