ScholarWorks@성균관대학교

초록

Bridging the gap between biological synapses and artificial neural networks requires advanced materials that can precisely emulate dynamic properties. Dion-Jacobson (DJ) perovskite artificial synapses offer a novel platform for modulating synaptic plasticity through crystallographic orientation control. Incorporating formamidinium chloride (FACl) into (PDA)(FA)n-1PbnI3n+1 (n = 2-8) (PDA = propane-1,3-diammonium, FA = formamidinium) results in vertically oriented crystallographic structures, which enhance charge transport efficiency and facilitates precise control of synaptic functions. The devices exhibit key synaptic properties, including paired-pulse facilitation (PPF), long-term potentiation (LTP), and long-term depression (LTD), with superior linearity and symmetry in synaptic weight modulation. These characteristics enable high-performance neuromorphic computing, as demonstrated through artificial neural network (ANN) simulations that achieve 94.47% accuracy in pattern recognition. Additionally, modeling second-language learning mechanisms through synaptic plasticity modulation demonstrates the crucial role of early input in memory formation. These findings highlight the potential of DJ perovskite synapses for advanced neuromorphic applications and fundamental studies on synaptic behavior.

키워드