4.5 Article

Memristive and artificial synapse performance by using TiOx/Al2O3 interface engineering in MoS2-based metallic filament memory

Journal

Publisher

PERGAMON-ELSEVIER SCIENCE LTD
DOI: 10.1016/j.jpcs.2020.109901

Keywords

Interface engineering; TiOx/Al2O3; MoS2; Memristor; Metallic filament; Artificial synapse; Conductance linearity; Long-term potentiation/depression

Funding

  1. Ministry of Science and Technology (MOST) Taiwan [MOST-107-2221-E-182-041, MOST-108-2221-E-182-026]
  2. Industrial Technology Research Institute (ITRI), Hsinchu, Taiwan
  3. MSSCORPS CO., LTD.

Ask authors/readers for more resources

This study investigates the impact of TiOx/Al2O3 interfacial layer on MoS2-based memristive/artificial synapse characteristics, showing significant improvement with the TiOx layer compared to Al2O3 and MoS2. The TiOx-based memory device demonstrates excellent RESET behavior and low energy consumption, making it suitable for future memory applications in computing.
The MoS2 as a switching material has recently shown promising resistive switching characteristics. In this work, we demonstrate the impact of TiOx/Al2O3 interfacial layer on memristive/artificial synapse characteristics using MoS2-based Al/Cu/TiOx(or)Al2O3/MoS2/TiN structure. Owing to its barrier properties over Cu migration, the memristor with 2 nm-thick TiOx layer shows significant improvement in resistive switching as compared to both Al2O3 interfacial layer and 4 nm-thick MoS2. The memristor shows uniform long P/E endurance of 2 x 10(9) cycles under a low operation current of 200 mu A and the high switching speed of 100 ns is applied. The ex-situ transmission electron microscope image reveals the formation of Cu metallic filament in MoS2 layer after long P/E endurance. At high operation current (1 mA), both the TiOx and Al2O3 interfacial layers show long P/E endurance of >10(9) cycles. In addition, the TiOx based memory device shows excellent RESET voltage controlled gradual RESET behavior without memory loss, enhanced uniform artificial synapse behavior with a maximum long-term potentiation/depression states of 55/500 at a small energy consumption of 5.7 pJ for future memory in computing application.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.5
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available