2019/04/17 by Rasmus E. Christiansen, Christiansen, Rasmus E., Fengwen Wang +5 · 1 citation
Computer Science · Physics and Astronomy · #FOS: Physical sciences #Mesoscale and Nanoscale Physics (cond-mat.mes-hall) #Optics (physics.optics) #Photonic Crystals and Applications #Topological Materials and Phenomena #Topological and Geometric Data Analysis #cond-mat.mes-hall #physics.optics
paper · pdf · doi:10.48550/arxiv.1904.08106
7 pages, 5 figures
arxiv created 2019/04/17 · openalex publication_date 2019/04/17 · arxiv updated 2019/04/18 · openalex created_date 2022/07/13 · openalex updated_date 2026/07/28
Designing photonic topological insulators is highly non-trivial because it requires inversion of band symmetries around the band gap, which was so far done using intuition combined with meticulous trial and error. Here we take a completely different approach: we consider the design of photonic topological insulators as an inverse design problem and use topology optimization to maximize the transmission through an edge mode with a sharp bend. Two design domains composed of two different, but initially identical, C6v-symmetric unit cells define the geometrical design problem. Remarkably, the optimization results in a photonic topological insulator reminiscent of the shrink-and-grow approach to quantum-spin-Hall photonic topological insulators but with notable differences in the topology of the crystal as well as qualitatively different band structures and with significantly improved performance as gauged by the band-gap sizes, which are at least 50 % larger than previous designs. Furthermore, we find a directional beta factor exceeding 99 %, and very low losses for sharp bends. Our approach allows for the introduction of fabrication limitations by design and opens an avenue towards designing PTIs with hitherto unexplored symmetry constraints.