Abstract
Abstract
We demonstrate an InGaZnO (IGZO)-based synaptic transistor with a TiO2 buffer layer. The structure of the synaptic transistor with TiO2 inserted between the Ti metal electrode and an IGZO semiconductor channel O2 trapping layer produces a large hysteresis window, which is crucial for achieving synaptic functionality. The Ti/TiO2/IGZO synaptic transistor exhibits reliable synaptic plasticity features such as excitatory post-synaptic current, paired-pulse facilitation, and potentiation and depression, originating from the reversible charge trapping and detrapping in the TiO2 layer. Finally, the pattern recognition accuracy of Modified National Institute of Standards and Technology handwritten digit images was modeled using CrossSim simulation software. The simulation results present a high image recognition accuracy of ∼89%. Therefore, this simple approach using an oxide buffer layer can aid the implementation of high-performance synaptic devices for neuromorphic computing systems.
Funder
Chungbuk National University
Subject
Electrical and Electronic Engineering,Mechanical Engineering,Mechanics of Materials,General Materials Science,General Chemistry,Bioengineering