A Taguchi-based performance predictive optimization model to design broadband antireflector

Abstract

To suppress the Fresnel reflection loss and enhance light transmission, subwavelength structured surfaces are used as antireflector. In this work, rigorous coupled wave (RCWA) analysis-based design approach has been adopted to simulate nano-sized moth-eye like structures on silicon substrate for wide-angle, short-wave infrared (SWIR) antireflector. Large reflection loss due to high refractive index of silicon is detrimental to optical performance. The proposed surface relief moth-eye structures introduce gradient refractive index to the surface depending on substrate materials, structure geometry, height, periods. Optimum selection of these parameters during design and fabrication are essential steps for the effective quelling of undesirable reflection from air-surface boundary and enhance transmission. Due to subwavelength nature, maintaining accuracy of all design parameters during fabrication on silicon is challenging. Careful tradeoff is required to fix tolerance of each parameter depending on priority to overall performance. In this work, with help of Taguchi optimization techniques, optimum combination of the structure height, periods, and top surface area of moth-eye structure are selected for antireflector. Analysis of variance (ANOVA) approach has been opted to identify the contribution of individual design parameters to performance. This performance model based on RCWA design, Taguchi optimization techniques and ANOVA analysis acts as a tool to predict the performance trend and fix tolerance of design parameters. For wavelength range (700nm -3000nm), with optimized height 600 nm, period 200 nm, and flat top diameter 70 nm of tapered moth-eye structures, the obtained reflectance is less than 1 % for the incidence angle up-to 45°.