Automatic optimization of miniaturized bound states in the continuum cavity

Area of Science:

• Photonics and optical engineering
• Computational physics
• Materials science

Background:

• Bound states in the continuum (BICs) are a promising mechanism for efficient light trapping.
• Confining light in three-dimensional compact volumes using BICs is challenging due to lateral boundary energy leakage.
• Conventional design methods struggle with the high degrees of freedom (DOFs) involved in optimizing BIC cavities.

Purpose of the Study:

• To develop a fully automatic optimization method for enhancing lateral light confinement in miniaturized BIC cavities.
• To overcome the limitations of conventional design approaches in addressing lateral boundary problems.
• To improve the performance of BIC cavities for applications requiring compact optical confinement.

Main Methods:

• Integration of a random parameter adjustment process with a convolutional neural network (CNN).
• Automated prediction of optimal boundary designs within a high-DOF parameter space.
• Utilizing CNNs to navigate complex design landscapes for photonic structures.

Main Results:

• A significant increase in the quality factor (Q-factor) from 4.32 × 10^4 to 6.32 × 10^5.
• Demonstrated substantial improvement in lateral light confinement.
• Validated the effectiveness of the automated CNN-based optimization approach.

Conclusions:

• The proposed CNN-based method effectively optimizes miniaturized BIC cavities for superior light confinement.
• This approach significantly enhances the quality factor by mitigating lateral leakage.
• The findings pave the way for developing advanced compact optical cavities for lasers, OLEDs, and sensors.