Ion‐Exchange‐Driven Ternary Organic Neuromorphic Devices with High Efficiency and Stability
ABSTRACT Organic neuromorphic platforms based on top‐gate ionic‐liquid architectures rely on precise electrochemical doping at low operating voltages; however, intrinsically unfavorable polymer‐ion interfaces often lead to unstable doping, elevated operating voltages, and poor device‐to‐device uniformity. Herein, we present a solution‐processed ternary organic blend system composed of organic semiconductors and a binary ionic‐liquid system, in which spontaneous ion exchange within the active layer establishes a stabilized ionic environment and facilitates anion access into semicrystalline domains. In situ X‐ray and Raman spectroscopy measurements reveal that this stabilized ionic matrix enables efficient anion penetration and electrochemical doping even under short gate pulses. The resultant ternary‐blend devices operate under a low gate voltage (V GS = −3.5 V) while effectively suppressing rapid de‐doping after bias removal. Notably, the ternary organic blend system exhibits consistent and enhanced electrochemical doping behavior across semiconducting polymers with fundamentally different backbone chemistries, ranging from thiophene‐based donor polymers to donor‐acceptor conjugated systems. In particular, the devices exhibit remarkable electrical and neuromorphic performance, including long retention (∼25h), stable operation over ∼10 3 sequential pulses, and reliable synaptic characteristics under various stimuli. This ternary strategy enables high‐reliability organic neuromorphic systems, achieving 97.98% recognition accuracy in CNN‐based MNIST classification.
Authors
- Liqiang Li (ORCID: https://orcid.org/0000-0001-8399-3957)
- Junyeong Sung
- Young Yong Kim (ORCID: https://orcid.org/0000-0001-6102-9518)
- Sooncheol Kwon (ORCID: https://orcid.org/0000-0003-2799-9558)
- Chandran Balamurugan
- Namsoo Lim (ORCID: https://orcid.org/0000-0001-7497-9440)
- G.Y. Wang
- Yong‐Ryun Jo (ORCID: https://orcid.org/0000-0002-7411-6132)
- Changhoon Lee (ORCID: https://orcid.org/0000-0001-8032-7094)
- Vivek Pratap Singh (ORCID: https://orcid.org/0000-0001-7206-6023)
- Zhongwu Wang (ORCID: https://orcid.org/0000-0002-1331-9796)
- Ji Hoon Shim
- Hyeonryul Lee
- Dongyeop Yang
- Sungmin Lee (ORCID: https://orcid.org/0009-0001-4854-5064)
Institutions
- Pohang University of Science and Technology (KR)
- Tianjin University (CN)
- Gyeongsang National University (KR)
- Dongguk University (KR)
- Gwangju Institute of Science and Technology (KR)
- Korea Foundation for Max Planck POSTECH (KR)
- Maulana Azad National Institute of Technology (IN)
Publication Details
- Journal
- Advanced Science
- Published
- 2026-09-08
- DOI
- https://doi.org/10.1002/advs.77486
- Primary Topic
- Advanced Memory and Neural Computing
- Type
- article
- Field-Weighted Citation Impact
- 0.00