Title | Bifunctional homologous alkali-metal artificial synapse with regenerative ability and mechanism imitation of voltage-gated ion channels |
Authors | Li, Lei Hu, Luodan Liu, Kai Chang, Kuan-Chang Zhang, Rui Lin, Xinnan Zhang, Shengdong Huang, Pei Liu, Heng-Jui Kuo, Tzu-Peng |
Affiliation | Peking Univ, Shenzhen Grad Sch, Sch Elect & Comp Engn, Shenzhen 518055, Peoples R China Natl Chung Hsing Univ, Dept Mat Sci & Engn, Taichung 40227, Taiwan Natl Sun Yat Sen Univ, Dept Phys, Kaohsiung 804, Taiwan Natl Sun Yat Sen Univ, Inst Mat & Optoelect, Kaohsiung 804, Taiwan |
Keywords | ACTION-POTENTIALS PLASTICITY MEMORY MODULATION |
Issue Date | Aug-2021 |
Publisher | MATERIALS HORIZONS |
Abstract | As a key component responsible for information processing in the brain, the development of a bionic synapse possessing digital and analog bifunctionality is vital for the hardware implementation of a neuro-system. Here, inspired by the key role of sodium and potassium in synaptic transmission, the alkali metal element lithium (Li) belonging to the same family is adopted in designing a bifunctional artificial synapse. The incorporation of Li endows the electronic devices with versatile synaptic functions. An artificial neural network based on experimental data exhibits a high performance approaching near-ideal accuracy. In addition, the regenerative ability allows synaptic functional recovery through low-frequency stimuli to be emulated, facilitating the prevention of permanent damage due to intensive neural activities and ensuring the long-term stability of the entire neural system. What is more striking for an Li-based bionic synapse is that it can not only emulate a biological synapse at a behavioral level but realize mechanism emulation based on artificial voltage-gated "ion channels". Concurrent digital and analog features lead to versatile synaptic functions in Li-doped artificial synapses, which operate in a mode similar to the human brain with its two hemispheres excelling at processing imaginative and analytical information, respectively. |
URI | http://hdl.handle.net/20.500.11897/624154 |
ISSN | 2051-6347 |
DOI | 10.1039/d1mh01012c |
Indexed | SCI(E) |
Appears in Collections: | 深圳研究生院待认领 |