Premium
A Ferrite Synaptic Transistor with Topotactic Transformation
Author(s) -
Ge Chen,
Liu Changxiang,
Zhou Qingli,
Zhang Qinghua,
Du Jianyu,
Li Jiankun,
Wang Can,
Gu Lin,
Yang Guozhen,
Jin Kuijuan
Publication year - 2019
Publication title -
advanced materials
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 10.707
H-Index - 527
eISSN - 1521-4095
pISSN - 0935-9648
DOI - 10.1002/adma.201900379
Subject(s) - materials science , transistor , ferrite (magnet) , neuromorphic engineering , synapse , synaptic plasticity , memristor , perovskite (structure) , nanotechnology , optoelectronics , artificial neural network , neuroscience , computer science , electronic engineering , voltage , electrical engineering , artificial intelligence , chemistry , crystallography , composite material , biochemistry , receptor , biology , engineering
Hardware implementation of artificial synaptic devices that emulate the functions of biological synapses is inspired by the biological neuromorphic system and has drawn considerable interest. Here, a three‐terminal ferrite synaptic device based on a topotactic phase transition between crystalline phases is presented. The electrolyte‐gating‐controlled topotactic phase transformation between brownmillerite SrFeO 2.5 and perovskite SrFeO 3− δ is confirmed from the examination of the crystal and electronic structure. A synaptic transistor with electrolyte‐gated ferrite films by harnessing gate‐controllable multilevel conduction states, which originate from many distinct oxygen‐deficient perovskite structures of SrFeO x induced by topotactic phase transformation, is successfully constructed. This three‐terminal artificial synapse can mimic important synaptic functions, such as synaptic plasticity and spike‐timing‐dependent plasticity. Simulations of a neural network consisting of ferrite synaptic transistors indicate that the system offers high classification accuracy. These results provide insight into the potential application of advanced topotactic phase transformation materials for designing artificial synapses with high performance.