Full-Inorganic Flexible Ag₂S Memristor with Interface Resistance-Switching for Energy-Efficient Computing

Area of Science:

• Materials Science
• Electronics Engineering
• Computer Science

Background:

• Flexible memristors enable parallel computation for efficient neuromorphic computing in flexible electronics.
• Current flexible memristors often use filamentary mechanisms, leading to high energy consumption during operation.

Purpose of the Study:

• To develop a flexible memristor with a low-energy consumption resistance-switching mechanism.
• To demonstrate the potential of this new memristor for energy-efficient hardware-based neural networks.

Main Methods:

• Fabrication of a silver sulfide (Ag₂S)-based flexible memristor.
• Investigation of an interface resistance-switching (RS) mechanism based on Schottky barrier modification.
• Evaluation of device performance including endurance, retention, and energy consumption.
• Demonstration of image processing using a memristor array.

Main Results:

• The Ag₂S memristor operates via an interface RS mechanism, distinct from filamentary switching.
• Achieved high endurance (10⁵ cycles) and retention (10⁴ s) under bending.
• Demonstrated ultralow switching energy (∼0.2 fJ) due to the absence of filament formation.
• Hardware-based image processing achieved comparable accuracy to software with 2 orders of magnitude lower power consumption compared to filamentary RS.

Conclusions:

• The developed Ag₂S flexible memristor offers a promising pathway for energy-efficient neuromorphic computing.
• The interface resistance-switching mechanism is key to achieving ultralow power consumption in flexible electronic devices.
• This work paves the way for advanced flexible electronics with enhanced computational capabilities.