Multifunctional CMOS-integrable and reconfigurable 2D ambipolar tellurene transistors for neuromorphic and in-memory...

Area of Science:

• Materials Science
• Computer Engineering
• Nanotechnology

Background:

• The von Neumann bottleneck limits traditional computing performance.
• Two-dimensional (2D) nanomaterials offer potential for novel computing architectures.
• Existing 2D device structures still have room for innovation beyond the von Neumann architecture.

Purpose of the Study:

• To introduce a CMOS-integrable and reconfigurable ambipolar 2D tellurene (Te) transistor.
• To explore its application in non-von Neumann computing architectures.
• To demonstrate its potential for logic-in-memory computing.

Main Methods:

• Fabrication of a 2D tellurene transistor compatible with CMOS technology.
• Utilizing thermal atomic layer deposition for precise Fermi-level alignment.
• Leveraging the p/n-switchable ambipolar characteristics of tellurene for device functionality.

Main Results:

• Demonstrated a reconfigurable ambipolar 2D Te transistor integrated with CMOS.
• Achieved remarkable synaptic behavior and conventional inverter performance in a single device.
• Successfully constructed diverse complementary logic-in-memory computing circuits.

Conclusions:

• The developed 2D Te transistor enables compact, programmable CMOS inverters with reduced complexity.
• This architecture facilitates diverse complementary logic-in-memory computing applications.
• The study paves the way for revolutionary in-memory computing beyond the von Neumann architecture using 2D nanomaterials.