Mode splitting in optical microcavities for speckle-free wavelength reconstruction

Area of Science:

• Optics and Photonics
• Machine Learning Applications
• Metrology

Background:

• Accurate wavelength measurement is crucial for spectroscopy, optical communications, semiconductor manufacturing, and quantum research.
• Traditional wavemeters face limitations in resolution, bandwidth, or cost.
• Reconstructive wavemeters offer compact, cost-effective solutions using computational techniques.

Purpose of the Study:

• To propose a novel reconstructive wavemeter synergizing optical microcavities and machine learning.
• To achieve high-resolution and broadband wavelength measurement with enhanced sensitivity.
• To enable chip-scale integrability and reduce calibration efforts.

Main Methods:

• Integration of thousands of high-quality-factor optical microcavities with deformed whispering gallery modes.
• Development of a hybrid machine learning approach using boosting methods and variational autoencoders for wavelength interpretation.
• Implementation of wavelength interpretation as a regression task.

Main Results:

• Demonstration of a novel reconstructive wavemeter concept.
• Ensured uniqueness of wavelength patterns over an ultra-wide (~100 nm) spectral window.
• Achieved high intrinsic sensitivity (~100 fm), comparable to integrating sphere-based wavemeters.
• Offered superior miniaturization options and chip-scale integrability with reduced calibration needs.

Conclusions:

• The proposed hybrid approach offers a significant advancement in wavemeter technology.
• This novel wavemeter provides a compact, cost-effective, and highly sensitive solution for accurate wavelength measurement.
• The technology holds promise for widespread adoption in spectroscopy, optical communications, and quantum research.