Loading…

Long-term and short-term plasticity of Ta2O5/HfO2 memristor for hardware neuromorphic application

Here in, we introduce a Pt/Ta2O5/HfO2/TiN memristor with enhanced resistive switching behavior, these conductive effects were induced by inserting a HfO2 layer. We demonstrate that the uniform switching performance of the Pt/Ta2O5/HfO2/TiN device comes from the construction and destruction of oxygen...

Full description

Saved in:
Bibliographic Details
Published in:Journal of alloys and compounds 2021-01, Vol.850, p.156675, Article 156675
Main Authors: Ryu, Ji-Ho, Mahata, Chandreswar, Kim, Sungjun
Format: Article
Language:English
Subjects:
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Here in, we introduce a Pt/Ta2O5/HfO2/TiN memristor with enhanced resistive switching behavior, these conductive effects were induced by inserting a HfO2 layer. We demonstrate that the uniform switching performance of the Pt/Ta2O5/HfO2/TiN device comes from the construction and destruction of oxygen vacancies (ion generation) in the HfO2 film. Low-power response of the analog conductance changes with different dynamic synaptic characteristics were demonstrated, which included paired-pulse depression (PPD), long-term potentiation (LTP), long-term depression (LTD), and spike timing-dependent plasticity (STDP). This was achieved by the proper adjustment of pulse amplitude, width and interval. Furthermore, the pattern recognition accuracy of a system was evaluated which composed in the device by forming a 3-layer neural network (784 × 128 × 10) with Ta2O5/HfO2 based memristor synapses. The experimental research with proposed Pt/Ta2O5/HfO2/TiN memristor provides valuable insight for the optimization of synaptic performances to use in futuristic neuromorphic applications. •With bilayer Pt/Ta2O5/HfO2/TiN memristor enhanced memristive switching properties was obtained as compared to SL memristor devices.•Variety of dynamic synaptic characteristics were demonstrated.•Pt/Ta2O5/HfO2/TiN deposition was confirmed by XPS depth analysis.•3-layer neural network simulation provided the pattern recognition accuracy.
ISSN:0925-8388
1873-4669
DOI:10.1016/j.jallcom.2020.156675