Abstract

Computational methods are essential in the design and analysis of three-dimensional (3D) photonic crystals [1]. These methods predict the physical properties of a particular design before elaborate nanomanufacturing. However, any manufactured crystal device inevitably differs from the design [2,3]. These differences are not included in modelled crystals that we therefore call “utopian”, which are typically assumed to be infinite for computational reasons. Therefore, the inevitable differences between experimental measurements and theoretical predictions can be explained by two sources, the imperfect model and the effects of manufacturing. The central problem is that the current knowledge and methods are not sufficient for attributing the differences to either cause, which hampers a systematic approach to improving device performance and thus device applications. Here, we aim to improve the understanding of manufacturing effects by using the structure of a real photonic crystal as input for computations, thereby including all real manufacturing defects.

© 2023 IEEE