Researchers at the National Institute of Standards and Technology (NIST), together with the Joint Quantum Institute and the New York-based company Qunnect, distributed entangled photons over 62 kilometers of fiber strung from utility poles in the Maryland suburbs of Washington, DC, in early 2025. The link sustained entanglement distribution 92.8% of the time over a 24-hour run, transmitting 1,500 entangled photon pairs per second through fiber that expands in the daytime heat, contracts at night, and sways in the wind.

The team reported the results in the Journal of Optical Communications and Networking on July 15, 2026 (DOI: 10.1364/JOCN.592521), with Yicheng Shi of NIST as lead author.

The experiment matters because of what the fiber was not: conduit. Most of the 62 km hung above streets and sidewalks, fully exposed to the mechanical and thermal disturbances that real deployed fiber experiences. A European group distributed entangled photons over 248 kilometers of underground fiber in 2022 — a longer link, but a quieter one. The NIST work trades distance for environmental realism.

"It's about as bad a connection as you can possibly have," said NIST physicist Oliver Slattery, one of the study's authors.

The physics at stake is polarization. Classical networks encode information by modulating optical power, which tolerates mechanical disturbance and temperature drift. Entangled states, by contrast, are typically encoded in photon polarization — the orientation of the electric field vibration. Aerial fiber twists as it expands, contracts, and moves in the wind, and those twists rotate photon polarizations, destroying the shared quantum state that entanglement depends on.

The team's countermeasure was a pair of polarization-stabilization units developed by Qunnect. Each unit sent reference beams through the fiber, measured how the fiber's polarization transformation had changed the reference light, and applied the exact inverse transformation to the experimental photons in real time. A commercial source generated the entangled pairs, with each pair sharing a quantum state linking one photon's polarization to the other — parallel or at right angles. One photon from each pair went to an analyzer in the NIST lab; its partner traveled the 62 km to a second lab at the University of Maryland in College Park. A statistical test at the two endpoints confirmed that the detected photon pairs had remained entangled.

The measured numbers frame both the achievement and the gap. At 1,500 pairs per second, the distribution rate is, in the researchers' own assessment, respectable but short of what practical quantum networking will demand. The 92.8% uptime left 7.2% of the 24-hour period devoted to polarization correction — overhead that any deployment on similar infrastructure must budget for.

The application space the work serves is broad. Entangled telescopes thousands of kilometers apart could combine light from the same astronomical target for images sharper than any single instrument can produce. Arrays of entangled sensors could detect faint seismic disturbances and help pinpoint impending earthquakes or volcanic eruptions. Networks of linked quantum computers could run algorithms beyond any single machine's reach, with drug and materials simulation as candidate workloads. And any eavesdropping on an entanglement-based communication channel would be detectable.

All of these applications require keeping fragile entangled states alive outside the laboratory, and the economics of quantum networking require doing it on existing telecom fiber rather than purpose-built plant. That is the specific question the Maryland experiment answers: quantum networking protocols can function on the noisy aerial infrastructure that already carries internet traffic.

"I would call this a stress test of quantum networking systems," said Shi. "We put this to an extreme test in an environment that's really noisy. Amazingly, it turned out it still worked. It's a demonstration that quantum networking protocols can work in real-world environments."

The development raises a deployment question for network operators and standards bodies alike: if entanglement distribution now survives 62 km of wind-blown aerial fiber at 92.8% availability, what uptime, rate, and integration requirements must a metropolitan quantum network standard specify before carriers can be expected to provision it on fiber they already own?