Stretchable Synaptic Device with Photonic-Electric Dual Mode for Sign Language Recognition

Abstract

This work presents a photonic-electric dual-mode synaptic transistor based on a stretchable carbon nanotube/ polydimethylsiloxane substrate and poly(3-hexylthiophene) (P3HT) nanofibers. High-crystallinity P3HT nanofiber channels are fabricated via solvent engineering and self-assembly techniques, while the dynamic ion migration characteristics of the device enabled excellent synaptic plasticity under both unstrained and stretched conditions. The device exhibited key synaptic behaviors, including excitatory postsynaptic current, paired-pulse facilitation, and spike-frequency-dependent plasticity. Experimental results demonstrated that the device maintained stable electrical performance under 40% tensile strain (current fluctuation standard deviation σ < 0.57 µA) and successfully enabled real-time monitoring of finger bending postures and gesture recognition. Furthermore, by modulating synaptic weights with optical pulses, the device exhibited broadband photo response (400–650 nm) and high-frequency feature extraction capabilities, mimicking the contour enhancement mechanism of the biological visual system and enabling optical-encoded letter recognition. This study provides new insights for the development of multimodal neuromorphic electronic devices, with promising applications in intelligent electronic skin, soft robotics, and real-time image processing.