A Sandia National Laboratories team trapped cesium atoms on a 420-nanometer optical fiber using just 5 milliwatts of optical power, roughly 2,000 times less than an LED bulb, according to a paper published 16 July 2026 in AVS Quantum Science. With 150 nanowatts of fiber-coupled light, the same setup also produced measurements that mimic atom interferometry.

The result reduces the optical power budget to one-sixth to one-fourth of what previous nanofiber approaches required, the authors report. That matters for any R&D group trying to fit a quantum inertial sensor onto a photonic integrated circuit — the platform Sandia sees as the route to a field-deployable device.

What does a chip-scale quantum inertial sensor buy you?

Today's aircraft and ground vehicles rely on GPS for course correction. Jam or block the satellite signal, and onboard accelerometers must carry the load until the receiver recovers. Over hours, drift accumulates and the vehicle strays from its intended path.

Quantum sensors use quantum-mechanical effects to measure acceleration and rotation with substantially lower drift than classical inertial units. The trade-off has historically been size, vacuum hardware, and laser power — the exact items Sandia is now trying to shrink.

How does guided atom interferometry differ from free-space designs?

Free-space atom interferometers drop ultracold atoms through a vacuum chamber and probe them with pulsed lasers. Vibration can briefly break the line of sight between laser and atom cloud, corrupting the data.

The Sandia approach replaces free fall with a physical channel. Evanescent light — the halo wrapping a 420-nanometer fiber — holds the atoms in place like marbles in a pipe. Even when the device is jolted, the atoms stay inside the laser's view.

Lead author Jongmin Lee, a quantum sensing scientist at Sandia, called the nanofiber result "a clear potential path toward chip-scale quantum inertial sensing."

How is Sandia handling the heat problem?

The same lasers that trap atoms also warm the guide. At 200 times thinner than a human hair, the structure can crack under thermal load. Designers have had to choose between suspended waveguides, which load atoms efficiently but shatter under heat, and substrate-mounted waveguides, which survive heat but load atoms poorly.

The new membrane-waveguide splits the difference. Two silicon pins anchor each end of the nano-thin membrane. The pins dwarf the membrane under a microscope, and they act as heat sinks, drawing thermal energy away from the laser-heated region.

Lee's group then plans to feed cold cesium into a hole in the membrane or a gap between two silicon needles, where the same evanescent-field trick that worked on the nanofiber should apply.

What work remains before the device is field-ready?

Lee tested his measurement protocols on the nanofiber rather than the membrane because the membrane still needs calibration. The next experiments will:

• Run atom-trapping and power-consumption tests directly on the membrane-waveguide.
• Apply momentum kicks to the trapped atoms during measurement.
• Integrate the guide with adjacent photonic components on one chip.
• Scale toward a multi-sensor array.

"Our concept is not fully demonstrated yet, but we're very close," Lee said.

Who is paying, and what is the commercial path?

Sandia's Laboratory Directed Research and Development (LDRD) program funded the work. Sandia operates as a multimission lab for the U.S. Department of Energy's National Nuclear Security Administration, with defense, global security, and energy technologies among its chartered areas.

The release names no commercial partner, foundry, or licensing deal. Applications highlighted are military: navigation through GPS jamming, rough terrain, or strong turbulence. Secondary markets — autonomous vehicles, underground surveying, seismology — would each need a separate qualification track and almost certainly a photonic-integration partner not identified here.

If the membrane-waveguide platform reproduces the nanofiber's 5-milliwatt trapping and 150-nanowatt measurement numbers, the next milestone to watch is a chip-integrated atom interferometer holding coherence long enough to deliver navigation-grade accuracy in a shaken, vibration-heavy enclosure.