Maximizing the Light Extraction Efficiency for AlGaN-Based DUV-LEDs with Two Optimally Designed Surface Structures under the Guidance of PSO

A novel method of utilizing an intelligent algorithm to guide the light extraction surface structure designing process for deep-ultraviolet light emitting diodes (DUV-LEDs) is proposed and investigated. Two kinds of surface structures based on the truncated pyramid array (TPA) and truncated cone array (TCA) are applied, which are expected to suppress the total internal reflection (TIR) effect and increase the light extraction efficiency (LEE). By addressing particle swarm optimization (PSO), the TPA and TCA microstructures constructed on the sapphire layer of the flip-chip DUV-LEDs are optimized. Compared to the conventional structure design method of parameter sweeping, this algorithm has much higher design efficiency and better optical properties. At the DUV wavelength of 280 nm, as a result, significant increases of 221% and 257% on the LEE are realized over the two forms of optimized surface structures. This approach provides another design path for DUV-LED light extraction structures.