Kerr Optical Frequency Division Achieves 300GHz Millimeter-Wave Signal Purity

Millimeter-wave technology is crucial for advances in spectroscopy, radar and astronomy, but current oscillators struggle with spectral purity. Scott C. Egbert, Brendan M. Heffernan, and James Greenberg, all from Boulder Research Labs, IMRA America, Inc, alongside William F. McGrew and Antoine Rolland, demonstrate a groundbreaking method to surpass these limitations using Kerr optical frequency division. Their research combines a chip-scale microresonator with a dual-wavelength Brillouin laser to generate a 3.3THz signal with unprecedented coherence , achieving a phase-noise floor of -152 dBc/Hz and a remarkable RMS timing jitter of just 135 attoseconds. This innovative approach establishes optical frequency division as a versatile technique for creating highly coherent sub-terahertz carriers, effectively removing the constraints of traditional direct-generation methods.

This innovative approach generates a coherently divided 300GHz carrier with phase noise significantly below the quantum limit of conventional 300GHz dual-wavelength Brillouin lasers and far surpassing the thermo-refractive noise typically observed in microring resonators. The team achieved this by injection-locking a Kerr soliton microcomb with a 3.3THz optical reference, effectively dividing the frequency down to a remarkably stable 300GHz signal.

The study reveals a novel oscillator architecture based on optical frequency division, utilizing a dual-wavelength Brillouin laser to provide two mutually coherent optical tones separated by 3.3THz. These tones simultaneously pump and injection-lock a Kerr microresonator comb with a 300GHz repetition rate, inheriting the low phase noise of the optical reference while benefiting from the frequency division. This division reduces the impact of fundamental noise processes in direct proportion to the frequency ratio, enabling a substantial improvement in signal coherence. This level of timing precision is below that of even the quietest signals achieved with full optical frequency division, and was realized without the complexity of a self-referenced frequency comb. To directly verify these results, the researchers developed and implemented the first application of cross-correlation phase-noise measurements in the millimeter-wave regime using photonics-based local oscillators, providing conclusive evidence of the oscillator’s exceptional performance. This work establishes optical frequency division as a generic method for generating sub-terahertz carriers with coherence no longer constrained by direct-generation limits, opening new possibilities for high-capacity wireless, coherent imaging, molecular spectroscopy, and astronomical observations. The demonstrated attosecond-level integrated timing jitter represents a new benchmark in oscillator spectral purity and promises to unlock advancements in areas where precise timing and spectral control are paramount.

Kerr Combs and Brillouin Lasers for 300GHz Generation

Scientists engineered a novel millimeter-wave oscillator leveraging Kerr-induced optical frequency division to overcome limitations in spectral purity. The research team combined a chip-scale microresonator with a large-spacing dual-wavelength Brillouin laser operating at 3.3THz to achieve unprecedented coherence. This innovative approach directly addresses the fundamental noise constraints hindering millimeter-wave generation, particularly near the technologically important 300GHz band. Experiments employed a 3.3THz dual-wavelength Brillouin laser, which simultaneously pumped and injection-locked a Kerr microresonator comb with a 300GHz repetition rate.

The team carefully selected this repetition rate to coherently divide the 3.3THz carrier, effectively reducing the impact of fundamental noise processes proportional to the frequency ratio. A dispersion-compensated soliton pulse train was then generated and photodetected using a uni-traveling-carrier photodiode (UTC-PD), delivering a spectrally pure 300GHz carrier signal. This configuration enables a significant reduction in phase noise compared to traditional direct-generation methods. The study pioneered cross-correlation phase-noise measurements in the millimeter-wave regime, utilizing photonics-based local oscillators for direct verification of performance.

These measurements revealed a phase-noise floor of -152 dBc/Hz at 1MHz offset, consistent with photodetection shot noise, a benchmark indicative of exceptional signal quality. Integration of the measured spectrum from 1kHz to 1MHz yielded an RMS timing jitter of 135 attoseconds, corresponding to a timing-noise floor of approximately 18 zs/√Hz at 300GHz. This technique achieves a level of timing stability surpassing even the quietest signals obtained through full optical frequency division, and does so with a simplified system architecture. The team demonstrated that optical frequency division serves as a generic method for generating sub-terahertz carriers with coherence no longer constrained by direct-generation limits, opening new possibilities for applications in spectroscopy, radar, astronomy, and high-capacity wireless communication.

Microresonator Brillouin Laser Achieves Record Phase-Noise performance

Scientists have achieved a breakthrough in millimeter-wave oscillator technology, demonstrating unprecedented spectral purity for applications in spectroscopy, radar, and astronomy. The team measured a phase-noise floor of -152 dBc/Hz at 1MHz offset, a value consistent with photodetection shot noise and representing a significant advancement in signal coherence. Experiments revealed that this performance was attained by combining Kerr-induced optical frequency division within a chip-scale microresonator with a large-spacing dual-wavelength Brillouin laser, effectively overcoming limitations imposed by thermal phase-noise processes. The research focused on generating a 300GHz carrier through optical frequency division from a 3.3THz optical reference, injection-locking a Kerr soliton microcomb in the process.

Results demonstrate that this approach delivers a coherently divided 300GHz carrier with phase noise far below the thermo-refractive noise of a conventional microring resonator. Cross-correlation phase-noise measurements, developed specifically for this work, confirmed the exceptional performance of the resulting oscillator in the millimeter-wave regime. Integration of the measured spectrum yielded an RMS timing jitter of 135 attoseconds from 1kHz to 1MHz, establishing a new benchmark for timing stability. Tests prove that dividing the frequency from 3.3THz down to 300GHz significantly reduces the impact of fundamental noise processes, proportional to the frequency ratio.

Measurements confirm a timing-noise floor of approximately 18 zs/ √ Hz at 300GHz, surpassing the performance of even full optical frequency division techniques and achieved with a simplified system architecture. The oscillator architecture utilizes a dual-wavelength Brillouin laser providing two mutually coherent optical tones separated by Δ= 3.3THz, which simultaneously pump and injection-lock a Kerr microresonator comb. This breakthrough delivers a new method for generating sub-terahertz carriers with coherence no longer constrained by direct-generation limits. The team’s work establishes optical frequency division as a generic technique, paving the way for advancements in high-capacity wireless communication, coherent imaging, molecular spectroscopy, and astronomical observations. This innovative system achieves a coherently divided 300GHz carrier with phase noise below the limits of conventional direct-generation methods and far below the thermo-refractive noise typically seen in microring resonators. The resulting oscillator reached a phase-noise floor of -152 dBc/Hz at 1MHz offset, consistent with photodetection shot noise, and exhibited an RMS timing jitter of 135 attoseconds from 1kHz to 1MHz.

These findings establish optical frequency division as a versatile technique for generating sub-terahertz carriers with coherence no longer constrained by the limitations of direct generation, a substantial advancement in the field. The authors acknowledge that the performance is ultimately limited by photodetection noise and comb relative intensity noise, as detailed in their supplementary material. Future research could focus on mitigating these noise sources to further enhance the performance of this optical frequency division technique and explore its application in diverse areas such as spectroscopy, radar, and astronomy.