Pre-diagnosis of Failure Spots in Orange AlInGaP Light-Emitting Diodes Soaked in Liquid Nitrogen Using Machine Vision and Multiple Optical, Electrical, and Material Characterizations

In this study, Pre-diagnosis of degraded aluminum indium gallium phosphide (AlInGaP) light-emitting diodes (LEDs) that were soaked in liquid nitrogen (LN2) was performed. To visualize the early deterioration of LED emission, a combination of material, electrical, optical, and computer-aidedmachine-vision analysis of LED failure spots on the surface was examined. Results indicate that several small failure spots, which can be identified by the MATLAB processed emission images and surface roughness variation, can be found. Furthermore, LN2 soaking induced some local damages inside the device that penetrate themultiple-quantumwell. The decrease of current and luminescence intensity also reflects the gradual degradation of the device after being soaked in LN2. This integrated analysis and Pre-diagnosis of the degraded device provided early screeningof LEDfailure spots and real-time inspection during operations.