Glycolated-Backbone-Induced Ion Kinetics Control for Implementing Artificial Synapses with Human-Like Temporal Memory Dynamics
발표자
장효익 (서울과학기술대학교)
연구책임자
이은호 (서울과학기술대학교)
공동저자
장효익 (서울과학기술대학교), 김윤희 (경상국립대학교), 이은호 (서울과학기술대학교)
초록
내용
The human brain implements adaptive information processing through gradual memory decay. This phenomenon can be emulated in organic electrochemical transistors (OECTs) via ion back-diffusion-induced conductance attenuation. However, prior OECT-based artificial synapses have been limited to phenomenological demonstrations of synaptic behavior on restricted timescales due to doping state lifetime limitations. In this study, we introduce a backbone glycolation strategy that enhances molecular flexibility and promotes strong ion trapping, thereby substantially extending the persistence of the doped state. Consequently, we successfully implement an artificial synapse that closely follows the Ebbinghaus's forgetting curve on a timescale reaching 105 seconds, directly demonstrating human-like memory decay dynamics. These results establish the potential of backbone glycolation as a key design principle for implementing neuromorphic devices with human-like temporal memory dynamics.