Guangdong Juyao Intelligent Technology Co., Ltd. · Laser Software & Process R&D
23°07′N 113°15′E · GUANGZHOU · PURE STATIC MONOGRAPH
← Back to news
Research Watch2026.10.31· 3 min read

Research Watch: Photon-pairing dual-comb spectroscopy crosses 400 km of fibre — spectra reconstructed at just 5.5×10⁻¹⁵ W per comb line

The farther a spectrum is measured, the weaker the light and the harder its phase is to hold steady. This paper changes the comparison itself: instead of forcing the wavefront phases of two beams into alignment, it pairs photons by their arrival times and recovers the spectrum from the statistical correlations of those pairs. Across 400 km of fibre, at 5.5×10⁻¹⁵ W per comb line, a spectrum can still be reconstructed — because pairing lets a weak signal be accumulated for longer, and more effectively.

JUYAO editorial visual: a glowing fibre line runs across a dusky mountain landscape, with paired bright light nodes along it and coloured spectral bands above
COVERJUYAO editorial visual (concept schematic, not a measured photograph or a paper figure).
EXECUTIVE SUMMARY

On 31 October, Jinping Lin, Lai Zhou and colleagues at the Beijing Academy of Quantum Information Sciences published a photon-pairing-enhanced distributed spectroscopy sensing study in Nature Communications. Tackling the phase-fluctuation problem in photon-counting dual-comb spectroscopy, the team replaced first-order interference with a second-order-correlation photon-pairing scheme, removing the need to phase-align interferograms and allowing long integration times. Over a 400 km fibre link in a distributed configuration, they accurately reconstructed intensity spectra at an average received power of only 5.5×10⁻¹⁵ W per comb line, with a signal-to-noise ratio above 23.5 dB and strong resilience to fibre-induced fluctuations. The reconstructed spectra’s signal-to-noise ratio grows linearly with photon count rate, unlike the square-root dependence of conventional photon-counting methods. The authors point to long-haul dual-comb spectroscopy, frequency metrology, remote sensing and quantum networking as prospective uses.

01 / The bottleneck: interferograms fear “shaking” more than “weakness”

The basic idea of dual-comb spectroscopy is to beat two optical frequency combs with slightly different spacings against each other, producing an interferogram from which the spectrum is recovered. Single-photon detection lets it work at extremely low photon fluxes — even when photons arrive one at a time, the signal can be built up by accumulation. But that accumulation has a precondition: the interferograms must first line up.

The difficulty named in the paper’s abstract is exactly this: photon-counting schemes reconstruct spectra by coherently accumulating interferograms, and phase fluctuations between the interfering beams disrupt that accumulation. The longer the fibre, the harder environmental, thermal and mechanical disturbances make it to suppress phase wander — and distributed measurement suffers most, because the phase scrambles before enough signal has been gathered.

02 / The change: from first-order interference to second-order correlation

The team’s change sidesteps phase alignment itself: a photon-pairing scheme exploiting second-order correlations, rather than relying on first-order interference. Put plainly, it no longer demands that the two beams’ wavefronts stay aligned at every instant; instead it reads the paired statistics of photon arrival times — the pairing events themselves carry the spectral information, so phase jitter no longer directly destroys a measurement.

Two benefits follow directly. First, long integration times become possible: with no need to maintain phase alignment continuously, a weak signal can be accumulated for longer. Second, the scaling law changes: the reconstructed spectra’s signal-to-noise ratio grows linearly with photon count rate, whereas conventional photon-counting methods grow only with its square root — so the same increase in count rate buys a faster rise in signal-to-noise ratio.

03 / The demonstration and JUYAO take: 400 km at femtowatt scale — how far from “usable”?

The demonstration figures are the paper’s hardest numbers: over a 400 km fibre link in a distributed configuration, the team accurately reconstructed intensity spectra at an average received power of only 5.5×10⁻¹⁵ W per comb line — femtowatt scale — with a signal-to-noise ratio above 23.5 dB and strong resilience to fibre-induced fluctuations. Note the boundaries of that claim: what was reconstructed is intensity spectra, the demonstration is a distributed configuration on a fibre link, and the abstract gives no commercial system, field deployment or specific sample under test.

For laser-industry readers, the value here is not “another quantum experiment” but a methodological generality: the dual comb is a core tool of laser spectroscopy, and if “correlation instead of alignment” holds, every scenario where light must travel far and phase is hard to stabilise — long-distance fibre sensing, remote gas and environmental monitoring, cross-site frequency comparison — gains an additional technical route. What to track next is how the scheme performs on real field fibre, with more complex samples and in multi-node networks — none of which is within this abstract’s disclosure.

PRIMARY REFERENCES & LINKS

  • Lin, Zhou et al.: Photon-pairing-enhanced distributed spectroscopy sensing over 400 km optical fiber
    Nature Communications · 31 October 2026 · Open access (CC BY-NC-ND 4.0)

    Paper abstract and publication record verified (received 7 July 2025, accepted 28 September 2026, published 31 October 2026; co-first authors Jinping Lin and Lai Zhou, corresponding author Lai Zhou; affiliations Beijing Academy of Quantum Information Sciences, Beijing University of Posts and Telecommunications, Peking University, Institute of Physics of the Chinese Academy of Sciences, and University of Chinese Academy of Sciences); used to verify the method, demonstration figures and prospective uses. Parameters beyond the abstract are not cited.

    Open source link ↗