Genetic algorithm optimized microstructure to enhance waveguide light coupling efficiency at normal incidences

Abstract

The recent trend to place well-designed photonic structures on waveguides is capable of effectively enhancing waveguides properties. One typical example is a nanostructure-empowered waveguide targeted for efficient light coupling. However, conceiving the high-freedom structures is a time-consuming and labor-intensive process, where an ineffective workflow limits the development of photonic microstructures. To address this issue, we deploy a genetic algorithm to customize structures in order to improve the coupling coefficients under predetermined situations (i.e., normal incidence combined with two polarizations). Three types of micropatterns are first conceived in the periodical model and then fully characterized on the real waveguide sizes. The simulated data reveal that the 550–1650 nm average coupling efficiencies of structure-enabled waveguides are raised by about 2% in contrast to the bare case, and the lineshapes are also flattened thanks to the grating modifications. In short, our solution underlines the role of an algorithm-developed nanostructure to lift waveguide coupling coefficients. By integrating well-engineering patterns, the waveguide-based probes may find a multitude of usages for weak signal detection and communication systems.