POS5-0895
Photo-Switchable Organic Synaptic Transistors for Integrated Reservoir Computing
When and Where
Nov 30, -0001
00:00 - 00:00
Presenter(s)
Yoohyeon Jang (Seoul National University of Science and Technology)
Co-Author(s)
Abstract
Reservoir computing (RC) is a promising neuromorphic computing paradigm for processing time-series data. OEST has attracted attention as a suitable platform for RC implementation due to its rich ion-electron coupling properties; however, instances of dynamically reconfigurable reservoir implementations remain limited. In this study, we introduced photo-responsive azobenzene into P3HT channels to control synaptic dynamics through trans-cis photoisomerization. Molecular structural changes induced by light irradiation modulate short-term plasticity (STP) and long-term plasticity (LTP), enabling the simultaneous implementation of memory and calculation functions within a single device. The integrated RC system constructed based on this approach effectively processed the temporal information of input signals and successfully performed the MNIST handwritten digit recognition task. This study demonstrates the potential of reconfigurable OEST-based RC platforms utilizing photoisomerization and provides a new strategy for the development of next-generation neuromorphic hardware.
