Large-Parameter-Space Optimization of Photonic Crystal Slab Cavities

Abstract

In the recent past, considerable progress has been made in the optimization of various figures of merit of photonic crystal cavities (PhC) and waveguides [1–5], leading to record performance in terms of quality factors in various materials [6,7], broad band slow light [8], second and third harmonic generation [9, 10]. Here, by combining large scale numerical simulations to a hybrid stochastic-global optimization strategy, we demonstrate that there is still considerable room for further improvement of these figures of merit when increasing the size of the optimization parameter space. In particular, we optimize a silicon-on-air L3 PhC cavity to a quality factor Q > 80 × 106. While this result may have limited impact on applications, as the fabrication disorder sets a statistical upper bound of roughly Q ∼ 4 × 106, we show that the same optimization strategy leads to unprecedented results for buried Si/SiO 2 PhC cavities, where we demonstrate a theoretical Q-factor in the 10 million range. This result sets a new record for the L3 cavity and opens the way to a new class of optimized PhC designs for monolithic structures or for structures made of low-index-contrast materials, such as AlN, GaN or silicon nitride, holding great promise for enhanced optical nonlinearity, sensing, and solid-state quantum optics.