Controllable analog resistive switching and synaptic characteristics in ZrO2/ZTO bilayer memristive device for neuromorphic systems

Muhammad Ismail, Haider Abbas, Changhwan Choi, Sungjun Kim

Research output: Contribution to journalArticlepeer-review

70 Scopus citations

Abstract

The development of artificial synaptic devices is a crucial step for the realization of efficient bio-inspired neuromorphic computing systems. In this work, the bilayer ZrO2/ZTO-based electronic synaptic devices were fabricated for better emulation of the brain's functions for neuromorphic computing. The ZrO2/ZTO switching layer is used to achieve stable and continuous switching for the emulation of the synaptic functions. The growth and rupture of conducting filament can be efficiently controlled by modulating the SET-compliance current (CC, 1 mA to 10 mA with 0.5 mA increment) and the RESET-voltage (1.0 V to 2.0 V, with 0.025 V increment) during the SET- and RESET-process, respectively. The incremental switching characteristics were exploited with proper pulse stimulations to emulate essential synaptic functions. Various pulse measurements were carried out to mimic some of the basic synaptic functions including long-term potentiation (LTP), long-term depression (LTD), spike-rate-dependent plasticity (SRDP), paired-pulse facilitation (PPF), and post-tetanic potentiation (PTP). Finally, the spike-timing-dependent plasticity (STDP) learning behavior was successfully emulated, which demonstrates the feasibility of ZrO2/ZTO-based electronic synaptic device for neuromorphic applications.

Original languageEnglish
Article number147107
JournalApplied Surface Science
Volume529
DOIs
StatePublished - 1 Nov 2020

Keywords

  • Biological synapse
  • Multiple states
  • Neuromorphic computing
  • Resistive switching
  • ZrO/ZTO memristor

Fingerprint

Dive into the research topics of 'Controllable analog resistive switching and synaptic characteristics in ZrO2/ZTO bilayer memristive device for neuromorphic systems'. Together they form a unique fingerprint.

Cite this