Binary particle swarm optimized 2  ×  2 power splitters in a standard foundry silicon photonic platform

Area of Science:

• Photonics and Optical Engineering
• Integrated Optics
• Computational Electromagnetics

Background:

• Silicon photonics is a key technology for integrated optical circuits.
• Power splitters are fundamental components in photonic integrated circuits.
• Ab initio design methods are crucial for optimizing complex photonic devices.

Purpose of the Study:

• To design compact power splitters using binary particle swarm optimization (BPSO).
• To investigate the performance of power splitters on a standard foundry silicon photonic platform.
• To explore the impact of cell dimensions on device performance and manufacturability.

Main Methods:

• Utilized binary particle swarm optimization (BPSO) for ab initio design of power splitters.
• Implemented designs on a standard foundry silicon photonic platform with varying cell sizes (200nm x 200nm and 100nm x 100nm).
• Characterized devices for insertion loss, bandwidth, and wafer-level consistency.

Main Results:

• Demonstrated compact power splitter designs with a 4.8μm x 4.8μm footprint.
• Designs utilizing smaller 100nm x 100nm cells exhibited lower insertion losses and broader operational bandwidth compared to 200nm x 200nm cells.
• The smaller cell design showed consistent performance across the wafer, despite not strictly adhering to design rules.

Conclusions:

• BPSO is an effective method for optimizing compact silicon photonic power splitters.
• Reducing cell dimensions can lead to improved optical performance (lower loss, wider bandwidth).
• Further research is needed to understand the impact of minimum feature dimensions on device performance and validate design rule deviations.