Cryogenic neuromorphic circuits using gate-controlled negative differential resistance in silicon carbide

Area of Science:

• Materials Science and Engineering
• Solid-State Electronics
• Quantum Computing Hardware

Background:

• Cryogenic electronic circuits are essential for quantum computing but require ultra-low power due to thermal constraints.
• Neuromorphic circuits offer energy efficiency by mimicking biological neurons.
• Existing solutions struggle to meet the demanding requirements of millikelvin environments.

Purpose of the Study:

• To investigate and leverage negative differential resistance (NDR) in silicon carbide (SiC) metal-oxide-semiconductor field-effect transistors (MOSFETs) for cryogenic applications.
• To demonstrate the feasibility of programmable cryogenic spiking neuromorphic circuits.
• To explore the potential of SiC technology for scalable integration in quantum information systems.

Main Methods:

• Fabrication and characterization of SiC MOSFETs exhibiting gate-controlled negative differential resistance (NDR).
• Analysis of the electron-donor impact ionization (EDII) mechanism responsible for NDR.
• Design and demonstration of cryogenic spiking neuromorphic circuits (sensory neurons, logic neurons, integrate-and-fire neurons) using the developed SiC MOSFETs.

Main Results:

• Achieved gate-controlled NDR in SiC MOSFETs with an on/off current ratio exceeding 10⁷.
• Demonstrated full control over NDR behavior via gate voltage.
• Successfully implemented programmable cryogenic spiking neuromorphic circuits with functionality tunable by gate or drain voltages.

Conclusions:

• Gate-controlled NDR in SiC MOSFETs is a viable mechanism for energy-efficient cryogenic electronics.
• Programmable neuromorphic circuits can be realized using this SiC-based NDR effect.
• The manufacturability of SiC technology paves the way for scalable integration in quantum computing and cryogenic sensing.