Inverse design and computational optimization of nanophotonic devices

This technology is a software for optimizing the design of waveguides and beam splitters in nanophotonic circuits.

Unmet Need: Optimization algorithms that mitigate fabrication errors

To optimize the design of nanoscale components, such as waveguides and beam splitters, increasingly complex software has been developed. Most current methods rely on topology optimization, which optimizes for design geometry, but does not predict performance or optimize against fabrication errors. Since fabrication errors are inherent to the manufacturing process, optimization processes that explicitly consider fabrication errors and manufacturing tolerances must be developed to improve device performance.

The Technology: Discrete, meta-design approach to optimize against fabrication errors

This technology uses an innovative, meta-design approach to delineate waveguides as concatenations of discrete, primitive sections to enable optimization against fabrication errors. The geometric design of each section is selected by a computational optimization algorithm to achieve a specific performance goal, while ensuring that the chosen parameters are robust against minor fabrication errors. This technology is compatible with silicon-on-insulator fabrication methods, outperforms other optimization algorithms without increasing computation time, and may significantly improve nanophotonic device performance by effectively mitigating fabrication errors.

This software was used to design a beam splitter with a significantly wider bandwidth (>200 nm) than existing designs, as well as a 1-to-4 mode converter capable of maintaining a conversion efficiency above - 1.55 dB (70%) despite a 5% deviation in geometric design.

Applications:

• Fiber optics
• Optoelectronics
• High speed telecommunications
• Sensors

Advantages:

• Automated inverse design
• Optimizes discrete sections with specific performance goals
• Considers manufacturing tolerances
• Outperforms state-of-the-art computational methods in terms of device bandwidth
• Improves nanophotonic device performance

Lead Inventor:

Michal Lipson, Ph.D.

Patent Information:

Patent Pending

Related Publications:

• Zhu D, Dave UD, Lipson M, Zheng C. “Inverse Geometric Design of Fabrication-Robust Nanophotonic Waveguides.” OSA Technical Digest. SF1J.3.

Tech Ventures Reference:

• IR CU21166

• Licensing Contact: Greg Maskel