Quantum Technology · Photonics
, quantum memories, and fiber-optic networks all speak different languages of light.
A hollow tube filled with xenon gas might finally translate between them — without losing what makes the message quantum.
Researchers from UCLA, SLAC National Accelerator Laboratory, the University of Rochester, and the University of Ottawa have modeled a way to convert light between wavelengths while preserving its quantum phase — the property that carries quantum information. Using a gas-filled hollow-core fiber and a process called four-wave mixing, the team showed phase correlations above 0.95, and in some cases above 0.99, across several wavelength conversions relevant to real quantum hardware. It's a step toward solving one of the quantum internet's most stubborn plumbing problems: getting incompatible quantum devices to talk to each other at all.
A gas-filled hollow-core fiber acts as the meeting point where infrared, ultraviolet, and visible-light quantum signals could one day be translated into each other.
The Quantum Internet's Missing Translator
Quantum technology promises secure communication, powerful new computing, and sensing tools we don't yet have equivalents for.
But there's a plumbing problem hiding underneath that promise, and it's more basic than most people realize.
Different quantum systems don't operate at the same wavelength of light. They don't even come close.
Quantum memories, trapped ions, and similar devices tend to work best in the ultraviolet or visible range.
Long-distance communication over optical fiber, meanwhile, works far more efficiently at telecommunications wavelengths — a completely different part of the spectrum.
Building a genuinely useful quantum network means finding a reliable way to translate between these optical bands — without losing the information the light is carrying.
A new study published in Advanced Photonics Nexus explores exactly that translation problem, through a process called four-wave mixing.
The team behind it spans four institutions: UCLA, SLAC National Accelerator Laboratory, the University of Rochester, and the University of Ottawa.
Why Quantum Devices Speak Different Wavelengths
To understand why this mismatch matters so much, it helps to picture what a real quantum network would need to connect.
Optical clocks, trapped-ion quantum devices, Rydberg-atom platforms, and rare-earth quantum memories are all genuinely promising quantum technologies.
Each one, however, has its own preferred wavelength — dictated by the physics of the atoms or materials it relies on.
Fiber-optic networks, the backbone of long-distance communication, are optimized for telecommunications wavelengths — chosen decades ago for entirely classical reasons.
None of these choices were made with each other in mind. They evolved independently, for different physics, in different labs.
That's precisely why connecting them requires an interface — something that can accept light at one wavelength and hand off equivalent information at another.
The stakes are high: without that interface, quantum networks stay stuck as isolated islands of hardware, each brilliant on its own and unable to cooperate.
Four-Wave Mixing: Turning One Color of Light Into Another
The technique at the center of this study has a technical name — four-wave mixing, or FWM — but the underlying idea is fairly intuitive.
The researchers modeled the process inside a gas-filled hollow-core capillary fiber — essentially a hollow optical tube filled with xenon gas.
Inside that fiber, an input light signal interacts with a separate, much stronger laser pulse.
That interaction generates a brand-new output wavelength — effectively translating the original signal into a different color of light.
The team tested three conversion routes chosen specifically for their relevance to real quantum hardware.
| Input Wavelength | Output Wavelength | Relevance |
|---|---|---|
| 1,030 nm (infrared) | 343 nm (ultraviolet) | Links infrared sources to UV-based quantum systems |
| 1,550 nm (telecom) | 308 nm (ultraviolet) | Links fiber networks to UV-based quantum memories |
| 1,550 nm (telecom) | 516 nm (visible) | Links fiber networks to visible-range quantum devices |
Each pairing represents a real bridge this technology could eventually build, connecting specific, named categories of quantum hardware.
The Real Test Wasn't Efficiency — It Was Phase
Most wavelength-conversion research focuses on one metric above all: how much of the light survives the conversion.
This team deliberately looked past that number, at something more fragile and arguably more important: phase.
Phase describes part of a light wave's structure, and it plays a central role in how quantum information gets encoded and transmitted.
If a conversion process scrambles phase unpredictably, the quantum information riding on that light can be degraded or lost entirely — even if most of the light itself gets through.
To test this, the researchers simulated different phase structures, including linear phase patterns and two forms of quadratic phase modulation.
By comparing the phase of the input signal against the phase of the newly generated output, they calculated correlation values — a direct measure of how faithfully that structure survived the trip.
That correlation number is effectively a report card for the conversion process: how much of the original quantum "handwriting" made it through intact.
Where Coherence Held Up Best — And Where It Didn't
The results were genuinely encouraging across a wide range of tested conditions.
In many cases, phase correlations between input and output exceeded 0.95 — and under favorable conditions, climbed above 0.99.
That's a very high degree of fidelity for a process that fundamentally reshapes the light involved.
The clearest pattern: phase fidelity was highest when the intermediary laser pulse had relatively low energy and a narrow bandwidth.
Under those gentler conditions, the conversion process barely disturbed the phase structure at all.
Performance varied notably by pathway. Converting 1,550 nm telecom light to 516 nm visible light proved especially robust across a wide operating range.
Telecom-to-ultraviolet conversion, by contrast, was more sensitive — phase fidelity declined as pulse energies and bandwidths increased.
The researchers traced this to a genuine trade-off: stronger laser pulses boost conversion efficiency, but also strengthen nonlinear effects that reshape the light's phase.
Efficiency numbers reflected that same tension — infrared-to-ultraviolet reached 28%, telecom-to-ultraviolet reached 8.4%, and telecom-to-visible reached about 10.8%.
Higher pulse energies generally pushed efficiency up, but often at the direct expense of phase quality — exactly the balance future quantum transducers will need to manage.
Manohar Insight
What stands out to me is the discipline of chasing phase instead of the flashier efficiency number.
It would have been easy to headline this work around conversion percentages. Those numbers are real, but they're not actually the hard part.
Phase preservation is the quieter, more demanding achievement — and it's the one that actually determines whether a quantum network could work at all.
It's also worth being clear-eyed about what this study is and isn't. It's a model, not a direct quantum-state measurement.
Corresponding author Hao Zhang frames the contribution as groundwork for integrating heterogeneous quantum platforms across widely separated optical bands.
That's an honest description. The next step — testing whether this phase-preserving behavior holds for actual quantum states and entanglement — hasn't happened yet.
A hollow tube of xenon gas won't build the quantum internet by itself. But it might be the connector that finally lets its pieces talk to each other.
Join the Signal
If quantum networks need translators between wavelengths just as much as they need faster , which do you think will be the harder problem to solve — the hardware, or the connections between it? Share your take in the comments.
Discussion