Bio-inspired synaptic functions from a transparent zinc-tin-oxide-based memristor for neuromorphic engineering

• Published in: Applied surface science Vol. 544; p. 148796
• Main Authors: Ryu, Ji-Ho, Kim, Boram, Hussain, Fayyaz, Mahata, Chandreswar, Ismail, Muhammad, Kim, Yoon, Kim, Sungjun
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01-04-2021
• Subjects: Memristor; Neuromorphic system; Resistive switching; Synaptic device; Transparent memory; Transparent memory; Memristor; Neuromorphic system; Synaptic device; Resistive switching
• ISSN: 0169-4332; 1873-5584
• DOI: 10.1016/j.apsusc.2020.148796

[Display omitted]
•All transparent semiconductor memristor (ITO/ZTO/ITO) was presented.•Potentiaion/depression were demonstrated for artificial synaptic device applications.•Theoretical work can support switching mechanism.•The pattern recognition accuracy was evaluated by system-level neural network simulation.
In this work, ITO/ZTO/ITO transparent resistive memory was fabricated using a fully industrialized sputtering process. We investigate how the electrical properties correspond to biological synaptic characteristics to utilize the device in neuromorphic system. The optical transmittance was over 70% for all wavelengths in the visible region. Especially, it had the highest transparency of 87.1% at 710 nm. Among the diverse electrical measurements conducted on the device, the transient pulse responses with gradual switching is suitable for use in a hardware-based neuromorphic system. The synaptic characteristics in the device include paired-pulse facilitation (PPF) as well as long- and short-term memory plasticity determined by the type of input. The iso-surface charge density, integrated charge density, and Bader charge analysis were conducted to confirm the changes in conductivity of the ITO/ZTO/ITO structure that occurred with and without oxygen vacancies. Moreover, the pattern recognition was carried out by considering the conductance values of the measured data as the synaptic part of neuromorphic engineering.