Light is incredibly useful in daily life. We use light to see objects and determine details about them. Light is similarly valuable in probing the quantum world. It is often critical for both observing quantum objects and interacting with them.
When scientists need to precisely control atoms or molecules, light is often the only tool for the job. Selecting the correct frequency—color—of laser light and projecting it in the right configuration allows scientists to detect, trap, and even manipulate individual quantum particles.
However, keeping a laser stable at the right frequency is challenging. Even the most stable lasers randomly shift to slightly different frequencies and experience noise—random spurts of different frequencies similar to static on a radio signal. This frequency noise is currently one of the main limitations of lasers in many experiments. As researchers improve lasers, the improvements reliably produce better experiments and technologies, including more precise atomic clocks and quantum computers that experience fewer errors.
“Every quantum scientist dreams of having a laser that can keep driving quantum systems without introducing errors,” says Lingfeng Yan, a graduate student at JILA.
A team of researchers, led by JILA and National Institute of Standards and Technology Fellow and University of Colorado Boulder Physics professor Jun Ye, took on the challenge of tailoring a laser system to an unprecedented level of stability and showing the improvements it could deliver for practical applications. Achieving this new level of stability required them to make multiple lasers work together.
In an article published in the journal Physical Review X on August 26, 2025, they described their laser setup and showing the improvements it could deliver for practical applications. They showed that the laser delivered practical advantages by putting many neutral atoms through their paces working as a qubit—the basic building block of a quantum computer—and achieving an unprecedented low error rate for the particular design of qubit used.
The bespoke laser was needed because lasers aren’t all equal. Even with the best available designs, lasers of some colors are more stable than others in particular situations, and it’s impossible for any particular laser to do every job.
Fortunately, researchers can impart the stability of one laser onto another. It is like a dance teacher who has one student who is perfect at keeping their timing no matter how long the dance and another who is great at performing the necessary steps but frequently speeds up or slows down randomly. The teacher pairs them up, and whenever they notice the student messing up, they remind them to follow the lead of their partner. Properly directed, the group exceeds the performance of the individuals.
The group has access to a laser that can stay stable for extended periods—a prima ballerina. The researchers decided to test how well they could do at transferring its stability to a less stable dancer—specifically a laser compatible with altering the quantum states of strontium atoms. Such lasers are used to manipulate strontium in certain atomic clocks and quantum computers.
The lab’s stable dancer was a laser cavity made from a silicon crystal. The crystal’s rigidity makes it very stable over extended periods of time, but it must be kept at frigid temperatures to not be negatively impacted by temperature fluctuations.
“It is one of the best lasers in the world,” says Yan, who is the first author of the paper. “It provides an excellent long-term stability, but it's a specialized cavity.”
The specialized design means it is expensive and works for just a specific set frequency. So, to get similar performance at other frequencies, the team needed to become dance instructors and get other lasers to follow the silicon cavity’s lead over the long term.
Unfortunately, you can’t just yell dance instructions to a laser. The researchers had to use a specialized tool, called a frequency comb, to coordinate their lasers. A frequency comb is a device that, instead of producing a single laser beam, produces many precise, evenly spaced frequencies of light. The regular spacing makes frequency combs ruler-like tools for comparing different lasers and maintaining the frequency spacing between them.
However, even with the silicon cavity and frequency comb in the loop, the final beam would still experience high-frequency noise that would impair its use. This is largely because even the silicon cavity contributes a little noise, including some introduced by vibrations from the necessary cooling equipment.
To tamp down this residual noise, the researchers added another cavity to the dance: a simpler cavity that operated at the same laser frequency used to manipulate the strontium atoms. The second cavity is less stable over long times but doesn’t need to be cooled and therefore doesn’t experience the remaining troublesome noise over shorter periods. This second cavity handled suppressing their high-frequency noise issues while letting the silicon cavity steer the frequency over the long term.
The team carefully coordinated the appropriate set of correction procedures and technological connections between the two cavities, the optical frequency comb and the final laser, but that was just the beginning. The group still needed to confirm if their laser setup worked as intended. Was the meticulously tailored custom laser actually stable and could it deliver improved results?
The team created a test for themselves: Shining the laser at strontium atoms. The atoms’ sensitivity to specific light frequencies made them a precision tool for checking exactly how the laser was behaving. The researchers essentially turned the atoms into a tool for measuring the laser-frequency noise of the laser.
In the test, the strontium atoms reacted to subtle fluctuations in the light and could catch details that are otherwise easily missed. For example, during one test, they discovered an unexpected spike in the noise despite the laser seeming to run correctly. They discovered the noise was because a device designed to prevent the silicon cavity from vibrating had accidentally been turned off.
“What we trust most are measurements of the atomic response,” says Max Frankel, a graduate student at JILA and a co-author of the paper. “Atomic measurement should have the final word on our laser frequency noise model.”
Their test confirmed that their new setup delivered the improved performance they had predicted. Then, they moved on to demonstrating the practical advantage of all their effort by using the laser to make the atoms perform as qubits in a standardized test.
Using the stabilized laser, they performed strings of many gates—the basic operations of quantum computers—on each of 3000 qubits. They used gates that essentially signal an atom’s quantum state to spin around to various positions, which physicists call performing state rotations. Then, the researchers performed the gate that should reverse the whole string of operations. As long as noise didn’t interfere, the laser guided all the qubits through the set of steps to the same final position. By analyzing how well the qubits returned to their initial state over many runs, the researchers determined how reliably the laser executed the gates on average. They established a new record for the fidelity achieved using a laser to optically manipulate neutral atoms to perform state rotations.
The results of their test also match well with their model of the laser noise, which they say suggest that further laser improvements will likely deliver even better results. The team says that other researchers should be able to use the same techniques to tailor lasers with different frequencies to have similar refined stability.
“Lasers are central to manipulating quantum systems, which are very sensitive to imperfections, so improving lasers benefits scientists and engineers all over the world.” says Stefan Lannig, a JILA postdoctoral researcher and co-author of the paper. “To benefit from many ideas put forward by modern science, we need to enhance our control over intricate quantum systems, which requires first improving our tools.”
Written by Bailey Bedford, Freelance Science Communicator