Dear Toolbox Users and Friends,
Since the research of III-nitride LED and lasers started decades ago, the device design has been largely determined by deliberate device simulation or referring to prior arts.
Recently, the activities of artificial intelligence (AI) through machine learning (ML) such as AlphaGo has prompted us to think about integrating this powerful technique into the LED and laser research to make the research process more efficient with the eventual goal of improving device performance.
Today we have launched an AI-powered tool to facilitate the design of AlGaN deep UV quantum well LEDs and lasers. It is based on numerous sets of device data and our research experiences that we employed to train the 'machine' to learn how to design quantum wells. Polarization and quantum confinement are included.
The tool webpage is below. To use it is straightforward. You can simply input any wavelength. The 'machine' will then start 'thinking' to generate three sets of structures with different compositions and thicknesses. To make the 'machine' think again, you can click the "Predict" button one more time. In addition to deep UV, we are working on quantum wells emitting in other spectral ranges such as near UV and green as well as power devices like HEMTs. Your suggestions and advice are appreciated.
https://polarizationtoolbox.com/DUV-QW
Best Regards,
Polarization Toolbox Team
Advanced Semiconductor Lab
EE of CEMSE at KAUST
https://polarizationtoolbox.com
https://semiconductor.kaust.edu.sa
Related Persons
