Reversible Diode with Tunable Band Alignment for Photoelectricity-Induced Artificial Synapse

Area of Science:

• Materials Science
• Nanotechnology
• Electronics

Background:

• The big data era necessitates electronic nanodevices with low energy consumption for analog computing.
• Energy efficiency is a critical challenge in memory and logic circuits.

Purpose of the Study:

• To develop a miniaturized diode with reversible switching states for analog computing.
• To simulate artificial synapses for sensing, memory, and processing applications.

Main Methods:

• Fabrication of an N-n MoS2 homojunction diode.
• Utilizing bandgap renormalization through tunable band alignment (Type-I/Type-II switching).
• Characterization of device performance including rectification, memory retention, and energy consumption.

Main Results:

• Achieved a rectification ratio of 10^3.
• Demonstrated nonvolatile memory retention exceeding 7000 seconds.
• Exhibited ultralow energy consumption for synaptic functions (5 pJ for LTP, 7.8 fJ for LTD).
• Reached a paired pulse facilitation index of 230% and demonstrated optical storage capabilities.

Conclusions:

• The N-n MoS2 homojunction diode effectively simulates artificial synapses.
• The device's tunable band alignment and low energy consumption are suitable for advanced analog computing.
• Potential applications include optical storage and simulating visual cells.