Performance of GaAs smart pixel components before and after monolithic integration of InGaP LEDs using...

Area of Science:

• Optoelectronics
• Semiconductor device integration
• Smart pixel technology

Background:

• Smart pixel technology merges electronic circuitry with optoelectronic devices.
• Epitaxy-on-Electronics (EoE) allows monolithic integration of optoelectronic devices and electronic circuitry on a common semiconductor substrate.
• This enables novel applications like smart pixel arrays for neural networks.

Purpose of the Study:

• To present the architecture and experimental characterization of optical devices for a novel smart pixel implementation of a neural network.
• To evaluate the performance of detectors and LEDs before and after the Epitaxy-on-Electronics (EoE) process.

Main Methods:

• Monolithic integration of Indium Gallium Phosphide (InGaP) Light Emitting Diodes (LEDs) with Gallium Arsenide (GaAs) electronic circuitry.
• Experimental characterization of optical devices, including detectors and LEDs.
• Evaluation of device performance before and after the Epitaxy-on-Electronics (EoE) process.

Main Results:

• Demonstrated the integration of InGaP LEDs with GaAs circuitry to create smart pixel arrays.
• Measured performance characteristics of detectors and LEDs.
• Identified limitations in the performance of detectors and LEDs for full-scale neural network implementation.

Conclusions:

• The Epitaxy-on-Electronics (EoE) process enables the creation of smart pixel arrays with integrated optoelectronic and electronic components.
• Current limitations in detector and LED performance hinder full-scale neural network applications.
• Ongoing improvements in EoE technology show promise for overcoming these limitations in future implementations.