Proximal light field control in red InGaN Micro-LEDs via evolution-algorithm-designed multilayer stacks

This study addresses the coupled challenges of low vertical light extraction efficiency and insufficient color uniformity in red InGaN micro-light-emitting diodes (Micro-LEDs) by implementing an evolutionary-algorithm-driven optimization strategy for the thickness design of multilayer dielectric stacks. Through multi-objective optimization of Si3N4/SiO2 layer thicknesses, photon propagation modes were engineered via wavelength-selective interference, simultaneously enhancing vertical extraction and achieving spectral compression of the red emission for improved color purity. The designed stacks with optimized thickness were validated through Monte Carlo ray-tracing simulations modeling light manipulation over the proximal light field, with fabrication feasibility confirmed via plasma-enhanced chemical vapor deposition growth demonstrating sub-nanometer thickness control. Experimental results demonstrate a 47.52% directional intensity increment within the 70°–90° angular domain, a 6.0% reduction in chromatic dispersion, and robust reproducibility characterized by thickness uniformity <2% and spectral error <6.88%. This framework enables spectro-angular co-regulation of light emission, significantly advancing monolithic Micro-LED performance for high-fidelity augmented reality/virtual reality displays while providing a scalable solution to optical bottlenecks in nitride-based micro-optoelectronics for next-generation displays.