Defect-engineered hydrogen-terminated diamond optoelectronic synapses for UV-driven neuromorphic computing

Area of Science:

• Materials Science
• Neuroscience
• Computer Engineering

Background:

• Neuromorphic computing offers energy-efficient alternatives to traditional architectures.
• Diamond's properties are ideal for robust computing, but its use in neuromorphic applications is underexplored.
• Hydrogen-terminated diamond (H-diamond) presents a novel platform for exploring these applications.

Purpose of the Study:

• To investigate the potential of H-diamond in optoelectronic synaptic transistors.
• To engineer H-diamond devices for neuromorphic computing applications.
• To demonstrate synaptic behaviors and logic operations using H-diamond.

Main Methods:

• Fabrication of an H-diamond-based optoelectronic synaptic transistor by engineering defects and surface states.
• Characterization of device performance under 365 nm UV illumination.
• Implementation of a spiking neural network using measured device conductance states.

Main Results:

• The device exhibited tunable persistent photoconductivity, enabling emulation of synaptic plasticity.
• Demonstrated essential synaptic behaviors and logic operations.
• Achieved 84.95% recognition accuracy on the Fashion-MNIST dataset using a spiking neural network.

Conclusions:

• H-diamond is a promising material for developing robust, UV-sensitive neuromorphic devices.
• This work establishes a benchmark for diamond-based neuromorphic computing.
• Engineered H-diamond transistors can emulate synaptic functions for advanced information processing.