����

Quantum technologies deliver practical results by harnessing some of the smallest things in the universe.��In contrast, quantum experiments generally involve large machines that manipulate and cool minuscule test subjects.��Scientists are working to shrink quantum devices and implement them in more practical applications.

Researchers have already discovered many ways to put quantum effects to practical uses, like using light and matter interactions to measure time and magnetic fields with extreme accuracy. But for practical applications, engineers would generally prefer small and portable devices, not a lab full of cumbersome equipment.

Working toward smaller quantum devices, the groups of Prof. James K. Thompson (at JILA, the University of Colorado, Boulder, and NIST) and of John Kitching (a Fellow of NIST),��teamed up to discover if a small and robust device that provides streams of atoms just above room temperature could generate quantum light��when combined with existing tools for manipulating light, specifically mirrored chambers called optical cavities.��Quantum light is light that cannot be adequately described as electromagnetic waves but instead must be described as individual particles of light—photons.

“Chip-scale instruments based on atoms are already important for precision timing and magnetic field sensing,” Kitching says. “Extending this technology into the quantum domain opens up new opportunities for future research and applications.”

In an�� published June 3, 2026, in the journal Science Advances,��the team demonstrated how the interaction between atoms and laser light inside a cavity could reliably produce quantum light. Their approach of combining compact streams of atoms and mirrors has potential applications in quantum devices that might aid in a variety of tasks including accurately measuring time, navigating without a GPS signal and securely sending messages.

“This demonstrates that there's a path in the future to scaling and creating mass-manufacturable sources of quantum light,” Thompson says. “That would open up new science and technology pathways for quantum sensing as well as for quantum communication in these systems. This really is a first demonstration that these technologies are going to be compatible with each other.”

Existing approaches to manipulating individual atoms often rely on a technique called laser cooling, which requires bulky vacuum chambers and several laser beams. In contrast, the demonstrated approach only requires a single laser beam and works at warm temperatures, which are producible with small devices, so it could potentially replace equipment that fills most of a laboratory with a single device that could fit in someone’s hand.

The new approach used a device that was developed by Kitching’s group members William McGehee and Alex Staron. It creates a stream of atoms by filling a small chamber with a gas of rubidium atoms that randomly bounce around and then giving them one way out: straight, narrow channels pointing in the direction they want the atoms to go. Each channel is three millimeters long, which is about twice the thickness of a penny, but is only a tenth of a millimeter across, which is around the width of a human hair. Only the atoms that happen to bounce straight toward the narrow hallways can escape—producing a narrow beam.

The Thompson group used interactions between light��and��the beam of atoms to whittle the crowd of photons in a laser down to a stream of individual photons.��Their experiment relied on the way atoms absorb certain photons with a specific wavelength, which are then released after a short time.��Absorbing a photon can prevent an atom from absorbing another until after it has released the first.

In the experiment, the team shone a laser at the atoms to start the process and then used an optical cavity to retrieve the individual released photons. Optical cavities are made from pairs of mirrors that are precisely separated to encourage certain wavelengths of light to get caught bouncing back and forth between them. The spacing of the cavity will encourage some wavelengths of light that neatly fit between the mirrors, while other wavelengths will rapidly disappear from the cavity.

The researchers set up the cavity so that the rubidium atoms would consistently release photons into it. In an optical cavity, there is a small chance each time the light reaches the mirror that it will pass through like a window instead of being reflected. The group aligned the beam of atoms to cross between the cavity’s mirrors so that when light was released into the cavity, it traveled perpendicular to the original beam. They put a single-photon detector outside the cavity to detect any photon passing through the mirror.

“In some sense, the role of the cavity is to suck that photon right out of the atom,” Thompson says. “The atom grabs one from an applied laser beam, and then the cavity resonates with the atom to essentially allow us to very quickly suck that photon out of the atom. Because there was only one photon inside the atom, we know we will only get one photon.”

For light that isn’t quantum, like that from their laser, physicists expect photons to arrive randomly, meaning sometimes one will arrive immediately on the tail of another. So, in a laser, the chance of seeing a photon immediately after seeing another is the same as at any other time. However, for a quantum light source, there will be a pause between seeing two photons, meaning that the photon stream will be more regularly spaced.

In their tests, the researchers measured that there was usually around a 30-nanosecond pause between photons being released by an atom. The fact that light was spaced out and that seeing a photon meant another was unlikely to follow for a period is a reliable signature of quantum light.

Even with the successful detection of quantum light, the group still had a potential issue they were concerned about: Would the device keep running with extended use?

The mirrors must be clean to reliably hold the desired photons and then transmit them to the detector. So, if rubidium atoms strayed from their path and built up on the mirror, the experiment would stop working.

“These mirrors are very, very high reflectivity,” says JILA graduate student Hagan Hensley, who is one of the lead authors of the article. “Of the light that hits them, 99.999% of it is going to reflect. If any kind of contaminant gets on the surface of the mirrors, that reflectivity is going to drop. And if the reflectivity drops from 99.999% to 99.998%, that is already enough for us to notice a severe degradation in the performance of the system.”

To avoid dirtying their mirrors, the team carefully aligned the beam to pass straight across the gap between the mirrors, and then constantly pumped away air so that the rubidium would get sucked away after traversing the cavity. They expected the narrow hallways would keep all the atoms on the straight and narrow path away from the mirrors, but they couldn’t see the individual atoms to confirm.

The group checked the longevity of the device by running the experiment for about six months. They estimate that approximately 10 trillion atoms passed through the cavity during that time, and they didn’t detect any contamination on the mirrors.

Their results demonstrate that the compact device offers a reliable way to produce quantum light without the hassle scientists have traditionally put into cooling down individual atoms to achieve similar results.

“With our experiment, it's dead simple,” says JILA graduate student Braden Larsen, who is one of the lead authors of the article. “You can remove all the complex infrastructure, like vacuum chambers, a lot of the optical complexity and things like that, and you can still do great science.”

Moving forward, the group is working on making their experiment small enough��that it could be held in the palm of your hand.��Their plans include completely enclosing the beam source and cavity and using graphite to capture the rubidium atoms instead of pumping them away. They hope to see these technologies applied in new experiments and eventually implemented outside of a laboratory.