Metallopolymeric\nMemristor Based Artificial Optoelectronic\nSynapse for Neuromorphic Computing

Abstract

Mimicking the human brain to achieve neuromorphic computing\nholds\npromise in the field of artificial intelligence (AI). Optoelectronic\nsynapses are regarded as the crucial foundation stone in neuromorphic\ncomputing due to their capability to intelligently process optoelectronic\ninput signals. Herein, two donor–acceptor (D–A)-type\nmetallopolymers, <b>P-Cu</b> and <b>P-Zn</b>, containing\nporphyrin moieties are designed and synthesized, which are utilized\nas a resistive switching layer for preparation of memristors. The\nresulting memristors exhibit significantly enhanced electrical characteristics,\ndisplaying a high ON/OFF ratio, a low threshold voltage (<i>V</i>_th), and superior cycle-to-cycle reproducibility. This\nenhancement is attributed to the formation and dissociation of charge\ntransfer (CT) states induced by inserted metal ions. Importantly,\nthe <b>P-Cu</b>-based memristor demonstrates the capability\nto co-modulate optoelectronic signals, effectively emulating versatile\nsynaptic functions of the nervous system. These functions include\nexcitatory postsynaptic current (EPSC), paired-pulse facilitation\n(PPF), short-term plasticity (STP), long-term plasticity (LTP), transition\nfrom short-term memory (STM) to long-term memory (LTM), and learning-experience\nbehavior. Moreover, multiple Boolean logical functions were successfully\nimplemented using the paired stimuli of electrical pulses. The neuromorphic\ncomputing function was also proven through pattern recognition, achieving\na recognition rate of up to 86.08% for handwritten digits. This study\noffers a potent approach for developing multifunctional artificial\nsynaptic devices and artificial neural network platforms and opens\nup the innovative application of metallopolymers in the fields of\noptoelectronics and AI.