Sunlight is quantum networking's loudest neighbor. Free-space quantum links — the kind you'd eventually use to connect satellites or bridge distant cities — collapse the moment the sun is up, because ambient photons flood the detector and bury the signal. Researchers have spent years engineering around this: spectral filters, timing gates, better detectors. A team at INRS in Montreal thinks they're solving the wrong side of the problem.
The INRS Énergie Matériaux Télécommunications Research Centre group, led by Benjamin Crockett during his PhD and supervised by professors José Azaña and Roberto Morandotti, published results in Science Advances last week showing that quantum coherence itself can do the filtering work that external optics have been attempting for years. The method, called a quantum Talbot Array Illuminator (qTAI), borrows from an effect well-known in spatial optics: when light passes through a periodic structure at specific distances, it recreates its own pattern — the Talbot effect. The quantum version maps that behavior onto time and frequency for single photons, using a phase modulator and a linearly chirped fiber Bragg grating to implement Talbot phase modulation in the temporal domain. The result is that a single photon or an entangled pair will self-focus into a narrow time window. Incoherent noise will not.
The numbers are real. Under moderate noise injection, the coincidence-to-accidental ratio — essentially the signal legibility for photon pairs — jumped from 2.2 to 21.3, according to ScienceBlog's reporting on the paper. Quantum state tomography fidelity recovered from 0.62 to 0.86 under heavy noise. Quantum interference visibility improved by nearly 50 percent in some conditions. The paper, posted to Science Advances (DOI: 10.1126/sciadv.ady8981), reports an order-of-magnitude CAR increase for time-bin entangled photon pairs using standard telecommunications infrastructure.
What the paper calls the "quantum coherent energy redistribution" effect works on individual photons and on time-entangled pairs, and it operates at 1550 nanometers in the standard telecommunications C-band — the same band used by essentially all fiber optic infrastructure. The phase modulation required is bounded to roughly pi radians per photon, which is within standard electro-optic modulator capabilities, according to ScienceBlog. Critically, unlike conventional filtering approaches, the method does not require prior knowledge of where in time the signal photons will arrive or what their central frequency is — the information the filters normally need before they can reject anything.
That last detail is the elegant part. Standard filtering is essentially a gatekeeper: it needs to know what to let through before it can work. The qTAI doesn't. It exploits the fact that entangled photons are, by definition, correlated — they respond to the device's phase manipulations in a coordinated way, focusing into narrow temporal peaks, while random noise photons, lacking that coordination, fail to focus and stay spread out. The coherence itself is the filter. A team including Benjamin Crockett, Nicola Montaut, James van Howe, Piotr Roztocki, Yang Liu, and Robin Helsten carried out the work. Crockett, now a Banting postdoctoral fellow at the University of British Columbia, previously became the first scientist from a Canadian university to win the Tingye Li Innovation Award at the OFC conference and received the SPIE D.J. Lovell Scholarship, along with additional recognition from Optica and the IEEE Photonics Society.
The implication the field has been quietly circling: if you can recover quantum signals from bright environments, you might finally run satellite quantum key distribution during daylight hours instead of scheduling experiments around the opposition of the sun and target receiver. That's the real hook here, and it's not small. Practical quantum satellite links are currently effectively overnight-only operations. A method that survives daylight doesn't require exotic new detector technology or heroic ambient noise subtraction — it works because the quantum state itself behaves differently under the same conditions.
The INRS team is upfront about what comes next. The next phase involves integrating the method onto a chip, testing it in optical fibers and free-space channels, and combining it with other denoising techniques to extend the range and reliability of future quantum links. That roadmap is where the caveats live.
The current setup has insertion losses around 5 dB, which could be halved with better components but still represent real optical budget. The timing jitter of superconducting nanowire single-photon detectors places a practical floor on the achievable resolution. Noise contributions that share the exact same time-frequency profile as the signal cannot be separated — a fundamental limitation, not an engineering gap. Multi-photon artifacts from the source also cannot be filtered by this method. And the demonstrations so far are in a controlled laboratory context, not across a deployed city-scale quantum network.
This is not a deployed system. It is not a product. It is a clever piece of lab work that points at a real problem and suggests a new way to think about it. The coherence-as-filter principle is genuinely elegant, which is rarer than it should be in quantum photonics papers. Whether it survives chip integration and a free-space channel with atmospheric turbulence and background sun is the question the next two years of lab work will answer. That answer matters for anyone building quantum satellite constellations — which is, increasingly, several governments and at least three well-funded startups. The sun will not cooperate. Whether this does remains to be proven.
