Expanding Interferometry’s Potential with Quantum Memory
The more starlight a telescope collects, the crisper the image it can produce—hence, larger receivers result in finer-resolution pictures. But you can only build a telescope receiver so big before it warps under its own weight. Astronomers circumvent this limitation using a technique called interferometry, where they combine light gathered at multiple telescopes to create an interference pattern from which they can extract a single image.
Radio astronomers of the Event Horizon Telescope (EHT) Collaboration have used interferometry to great effect. In 2017, they combined signals collected from eight observatories over five continents to capture the first-ever image of a black hole, 55 million light-years away. Collectively, the observatories acted as a virtual radio receiver with a diameter, or “baseline length,” spanning the entire planet.
But so-called long-baseline interferometry is much more difficult with optical light. The reasons are complex, but optical photons oscillate much faster than radio photons, which limits the options for detecting and subsequently combining two optical signals [1]. The largest optical interferometer, CHARA, consists of six telescopes on a mountain in California, with an effective baseline of 330 m.
A team at Harvard University, led by Mikhail Lukin, is investigating how the baseline of optical interferometry can be expanded using devices called quantum memories, which are essentially two-qubit quantum computers. At the APS Global Physics Summit in Denver, team member Maxim Sirotin presented a recent demonstration in which the researchers created an interference pattern from optical photons detected at two sites separated by 1.55 km of optical fiber—nearly 5 times CHARA’s baseline [2]. Their demonstration is a proof of principle, as they produce the interference pattern from weak laser light simulating faraway starlight rather than actual astronomical signals.
“It’s the dawn of a new field where you apply quantum computing and quantum advantage to astronomical tasks,” Sirotin says. With more development, the researchers think astronomers can use the technique to map the surface of stars or image exoplanets.
The basic idea of astronomical interferometry involves collecting light from a source at two (or more) telescopes. The light travels a slightly farther distance to one telescope versus the other, and this lag introduces a phase difference between the two light signals. That phase difference, which provides information about the source’s location, can be measured by spatially overlapping the two signals and recording the interference pattern. CHARA, for example, produces interference patterns by physically shuttling photons from different telescopes to a single location via vacuum tubes. (Long-baseline radio interferometry, like that of the EHT, employs a different strategy, producing interference patterns at each telescope using a “local oscillator” as a kind of reference wave.)
In contrast, Lukin’s group does not spatially combine photons to make them interfere. Instead, when a photon arrives at one telescope location, the researchers record its quantum information in a quantum memory. Using quantum entanglement, this stored information can interfere with information in another memory at a second telescope location.
The two telescopes are located in two labs 6 m apart in the same building but are connected via one-and-a-half kilometers of spooled fiber. Each telescope’s quantum memory is a chip with a nanometer-scale cavity made of diamond. The diamond contains a defect known as a silicon vacancy, where two carbons in the diamond crystal are replaced by a silicon atom and a hole. An electron spin and a nuclear spin in the defect each serve as a qubit that stores and manipulates quantum information. Prior to each observation run, the researchers entangle the nuclear spins using light signals.
The team simulates an astronomical signal using a weak laser beam. An incoming photon interacts with an electron in one quantum memory. That electron spin then interacts with the silicon’s nuclear spin, thereby “imprinting” the incoming photon’s quantum information on the nuclear spin, Sirotin says. Because this nuclear spin is entangled with the nuclear spin at the second quantum memory, the researchers can reproduce the photon’s state at the second quantum memory by making particular measurements of the electron and nuclear spins. This so-called quantum teleportation allows them to combine the photon signals from the two telescopes and produce an interference pattern.
The quantum method uses a “nonlocal” transfer of information and thus avoids the photon loss that would occur in the conventional method of sending light signals from each telescope to some central interferometer. Photon loss degrades the interference pattern and becomes worse the farther the signals must travel. By contrast, the quantum-teleported interference pattern does not degrade with distance.
The efforts behind quantum-enhanced astronomy began at least 15 years ago, says Brian Smith of the University of Oregon, who was not involved in the work. “The field has made tremendous progress,” he says. Three years ago, Smith and his colleagues demonstrated quantum-assisted interferometry but without quantum memories [3]. The use of quantum memories is a significant step forward, Smith says, as these devices make it possible to observe interference between photons that arrive in a larger overall time window.
Lukin’s lab has been collaborating with NASA on a preliminary project exploring possible ways to incorporate quantum methods in the Habitable Worlds Observatory, the space telescope proposed to succeed the JWST. “We want to enhance the performance of a small telescope to make it act like a large one,” says Babak Saif of the Goddard Space Flight Center, one of Lukin’s collaborators. The researchers are investigating whether related quantum techniques could help them reduce the physical size of the telescope, potentially saving the mission billions of dollars [4].
Researchers in the field are also exploring the merits of a large-scale facility for quantum-assisted optical telescopes on the scale of gravitational-wave detectors, such as LIGO, over the next decade or so, Smith says. To that end, much work remains. For one, they would need to gather photons at a much higher rate, which would likely require them to scale up to at least two dozen quantum memories per location. They also need to improve the stability of the quantum networks that would connect the memories over long distances. “We’re in the early days of basic testing,” Smith says. “It’s going to take time and investment.”
–Sophia Chen
Sophia Chen is a freelance science writer based in Columbus, Ohio.
References
- M. Tsang, “Quantum nonlocality in weak-thermal-light interferometry,” Phys. Rev. Lett. 107, 270402 (2011).
- P.-J. Stas et al., “Entanglement-assisted non-local optical interferometry in a quantum network,” Nature 651, 326 (2026).
- M. R. Brown et al., “Interferometric imaging using shared quantum entanglement,” Phys. Rev. Lett. 131, 210801 (2023).
- A. Mokeev et al., “Enhancing optical imaging via quantum computation,” PRX Quantum 7, 010318 (2026).