Deep in a lab in Erlangen, Germany, a research team did something that sounds almost backwards for photonics: they took a working fiber-optic cable and froze it solid. What they got back wasn’t a broken piece of glass. It was a fundamentally new kind of optical component — one that couples light and sound roughly a thousand times more efficiently than the fiber running through your internet connection right now.
The work, published in the journal Optica by researchers at the Max Planck Institute for the Science of Light (MPL) in Erlangen, Leibniz University Hannover (LUH), and the Leibniz Institute for Photonic Technologies (IPHT) in Jena, matters for reasons that go well beyond a neat physics trick. Photonic computing — the idea of processing information with light instead of electricity — has spent years running into the same wall: light and sound don’t naturally “talk” to each other efficiently enough to build useful memory or logic elements out of that interaction alone. This result attacks that wall directly, and it does it with a method elegant enough to describe in a sentence: freeze the fiber’s core and the physics changes underneath it.
Key Takeaways
- Researchers at MPL Erlangen, Leibniz University Hannover, and IPHT Jena froze a liquid-core optical fiber (LiCOF) filled with carbon disulfide (CS₂) in liquid nitrogen at -196°C, converting the liquid core to a solid, crystalline state while it kept guiding light.
- The frozen fiber core couples light and sound — an effect called Brillouin-Mandelstam scattering — more than 1,000 times more strongly than conventional silica optical fibers.
- The team used this effect to demonstrate optoacoustic memory: information carried by light is briefly converted into slower-moving sound waves, then converted back into light.
- The result is described as a new physical platform for photonic neuromorphic computing, quantum signal processing, microwave photonics, and high-precision sensing — largely because it could cut the energy cost of those systems by orders of magnitude.
- The research builds on years of prior liquid-core fiber work at IPHT Jena and was led by the “Quantum Optoacoustics” group at MPL, headed by Prof. Dr. Birgit Stiller.
Why a Frozen Fiber Behaves So Differently
To understand why freezing a fiber’s core is such a big deal, it helps to start with what a liquid-core optical fiber actually is. A standard telecom fiber is a solid glass strand with a light-guiding core running through its center. A liquid-core optical fiber, or LiCOF, swaps that solid core for a hollow glass capillary filled with a liquid — in this case, carbon disulfide, a common nonlinear optical liquid already used in fiber research for its strong light-matter interactions. IPHT Jena researchers, notably Prof. Markus Schmidt and Prof. Mario Chemnitz, have spent years developing and refining this liquid-core platform, and their prior work forms the foundation this new experiment builds on.
Liquids are useful in fiber cores because their optical and mechanical properties can be far more extreme, and far more tunable, than solid glass. But they also have limits: liquids don’t hold a rigid structure, their acoustic properties are relatively soft, and there’s only so much density and confinement you can engineer into a fluid at room temperature.
The MPL-led team’s insight was to push past that limit by triggering a phase transition. Submerging the CS₂-filled capillary in liquid nitrogen at -196°C forces the liquid core to freeze into a dense, ordered crystalline solid. It’s the same underlying physics that turns a lake into ice in winter, or molten lava into basalt as it cools — a liquid reorganizing into a solid lattice as thermal energy drops. What’s notable is what doesn’t break in the process: the fiber keeps its ability to guide light through the now-solid core, and — more importantly for this work — both the liquid and the frozen sections continue to guide hypersonic sound waves, acoustic vibrations far beyond the range of human hearing but very much in the operating range of nonlinear fiber optics.
Any phase transition changes a material’s density and refractive index, and those two properties are exactly what govern how efficiently light and sound move through, and interact within, a medium. By pushing CS₂ into its solid phase inside the confined geometry of a fiber core, the researchers created an environment with a far denser, far more tightly confined optical and acoustic medium than anything achievable in the liquid phase — or in ordinary solid-core silica fiber, for that matter.
The Physics at the Center of It: Brillouin-Mandelstam Scattering
The specific effect the team is exploiting is called Brillouin-Mandelstam scattering, a well-established nonlinear optical phenomenon in which photons — the particles of light — couple with phonons, the quantized vibrations that make up sound waves and heat-carrying lattice vibrations in a material. This isn’t a new discovery; Brillouin scattering is already a workhorse effect in fiber-optic sensing, distributed temperature and strain measurement, and certain types of fiber lasers.
What’s new is the scale of the coupling. In the frozen LiCOF, the optoacoustic coupling strength is reported at over 1,000 times stronger than in standard silica fiber. That’s not an incremental gain — it’s the kind of jump that can turn an effect from “measurable in a lab” into “usable as an engineering building block.” A recent supporting paper describing this work in detail — titled “Giant Brillouin gain in frozen CS₂ capillaries” — frames the achievement in exactly those terms: an extreme, dense, highly confined physical regime engineered specifically to maximize this light-sound coupling.
The mechanism matters because light and sound travel at wildly different speeds. Light in a fiber moves at a meaningful fraction of the speed of light in vacuum; hypersonic sound waves in a solid or liquid medium move many orders of magnitude slower. That speed mismatch is normally a liability for anyone trying to build fast, all-optical devices — but the MPL team turned it into an asset.
From Coupling Strength to Optoacoustic Memory
The most concrete demonstration to come out of this frozen-fiber platform is optoacoustic memory — and it’s the part of the story that turns a materials-science curiosity into something with a plausible product roadmap.
Here’s the underlying idea. Because light moves so much faster than the hypersonic sound waves the frozen fiber core also guides, information encoded in a light pulse can be transferred into the slower-moving acoustic wave, held there briefly, and then converted back into an optical signal on demand. In effect, the fiber becomes a very short-term storage medium — a way of “parking” data that arrived as light, without ever converting it into an electronic signal, and then releasing it back into the optical domain later.
That might sound like a narrow trick, but it solves a genuine bottleneck. Modern computing and telecommunications systems spend an enormous amount of energy converting signals back and forth between the optical domain (where data travels efficiently over long distances) and the electronic domain (where it’s processed, stored, and routed). Every one of those conversions costs energy and introduces latency. A fiber that can briefly store and then re-emit optical information — without an electronic intermediary — chips away directly at that overhead.
Simon Seiderer, one of the three lead authors of the paper and a researcher in Stiller’s Quantum Optoacoustics group at MPL, put the core finding plainly: the frozen section of the fiber keeps its ability to guide light, and both the liquid and frozen sections also guide hypersonic sound waves. That dual guiding capability — light and sound moving through the same confined channel, coupling far more strongly than before — is what makes the optoacoustic memory demonstration possible in the first place.
Why This Could Matter for Computing, Not Just Optics Labs
The applications the research team points to fall into a few overlapping categories, and it’s worth walking through why each one benefits specifically from stronger optoacoustic coupling rather than just any optoacoustic coupling.
Photonic neuromorphic computing. Neuromorphic architectures try to mimic the way biological neural networks process information — with many small, distributed, interacting elements rather than a centralized processor executing sequential instructions. Photonic versions of this idea promise speed and parallelism advantages over electronic chips, but they’ve been held back partly by the difficulty of building efficient, low-loss elements that can store and manipulate light-encoded information locally. A frozen fiber segment that can act as a light-sound-light memory element is a plausible building block for exactly that kind of architecture — and because the coupling is so much stronger, the energy cost of running such elements drops sharply.
Energy efficiency in photonic architectures more broadly. This is arguably the throughline connecting all the proposed applications. Whenever a system can complete a signal-processing operation through direct light-sound interaction in the fiber rather than routing the signal out to electronics and back, it avoids the conversion losses and the energy draw of electronic components. Birgit Stiller, who leads the project, described the frozen liquid-core platform as combining extreme nonlinear behavior with practical ease of use — an entirely new physical platform that provides extreme nonlinearities while remaining easy to handle.
Quantum signal processing. Strong, controllable optoacoustic coupling is also relevant to quantum photonics, where researchers are looking for ways to manipulate and transfer quantum information between different physical carriers — light and mechanical/acoustic modes among them — without introducing excess noise. A platform with dramatically stronger and more confined coupling gives quantum-photonics researchers a more workable interaction to engineer around.
Microwave photonics and high-precision sensing. Fiber-based Brillouin scattering already underpins a range of microwave-photonic filtering and signal-processing techniques, plus distributed sensing applications like fiber-based temperature and strain measurement. A thousand-fold increase in coupling strength, achieved in a comparatively simple, cryogenically frozen fiber segment, gives those existing techniques a substantially more sensitive underlying physical effect to build on — potentially without requiring exotic, chip-scale conversion hardware.
Stiller was explicit that the team sees the optoacoustic memory result as a first step rather than an endpoint, noting that this level of light-sound coupling opens up possibilities spanning neuromorphic computing, quantum information processing, microwave photonics, and high-precision sensing.
Building on a Decade of Liquid-Core Fiber Research
None of this emerged from nowhere. Liquid-core optical fibers have been an active research platform for over a decade, prized for the ability to tune their optical and nonlinear properties simply by changing the liquid filling the core — something impossible to do with a fixed solid-glass core. IPHT Jena’s Markus Schmidt and Mario Chemnitz have been central figures in that line of research, developing the liquid-core fiber techniques and characterizing the nonlinear photonic behavior of CS₂-filled capillaries specifically.
The MPL team’s contribution was recognizing that an additional, deliberately induced phase transition — freezing rather than simply cooling — could unlock nonlinearities beyond what the liquid phase allows on its own. That’s a meaningfully different research direction: rather than optimizing within the liquid regime, the team stepped outside it entirely, into a solid-but-still-fiber-compatible state that hadn’t previously been systematically exploited for this purpose. The Quantum Optoacoustics group’s broader research portfolio — which includes prior work on Brillouin-Mandelstam scattering thermodynamics in nanoliter fiber volumes — set up the theoretical and experimental groundwork for treating a frozen fiber core as a controllable optoacoustic medium rather than a failure mode to avoid.
What’s Still Unproven
It’s worth being precise about what has and hasn’t been demonstrated at this stage. The team has shown a working optoacoustic memory effect and characterized a dramatic increase in Brillouin-Mandelstam coupling strength in the frozen fiber segment. What hasn’t been demonstrated yet, at least based on the published results, is a scaled, chip-integrated, or room-temperature-stable version of any of this. The frozen-core effect, by definition, requires cryogenic cooling to liquid-nitrogen temperatures — a real practical constraint for any near-term commercial computing or telecom application, even if it’s a routine condition in many quantum-photonics and superconducting-computing labs already.
There are also open questions about durability and repeatability of the freeze-thaw cycle, the maximum practical length of frozen fiber segment that can be manufactured and maintained, and how the memory storage time and fidelity scale as researchers try to push toward more complex logic operations rather than simple store-and-release demonstrations. Those are the kinds of engineering questions that typically follow a foundational physics result like this one — not reasons to doubt the finding, but reasons the timeline from lab demonstration to deployable technology is likely to be measured in years, not months.
What to Watch Next
The most immediate next step for the field is likely to be scaling: can the frozen-core effect be reliably reproduced across longer fiber segments, integrated with existing photonic-chip fabrication approaches, or combined with other liquid-core fiber chemistries beyond CS₂ to tune the effect further? Groups working on photonic neuromorphic computing and quantum photonics elsewhere are likely to take note of this coupling-strength result specifically, since a 1,000x improvement in a foundational nonlinear effect is the kind of number that tends to attract fast follow-up work.
It’s also worth watching whether the cryogenic requirement becomes a genuine adoption barrier or a non-issue — much depends on whether the eventual applications (quantum computing, specialized sensing) are already operating in cryogenically cooled environments for other reasons, in which case a frozen fiber segment is simply one more cold component among many, rather than an added system-level cost.
Closing Analysis
What makes this result notable isn’t just the raw coupling-strength number, striking as it is — it’s that the method used to get there is conceptually simple and grounded in decades-old physics: cool a liquid past its freezing point and watch its optical and acoustic properties transform. The novelty is in recognizing that a fiber-optic context is exactly the right confined, controllable geometry to exploit that transition for information processing rather than treating it as an incidental material change. The unresolved questions are practical rather than fundamental — how far the effect scales, how it integrates with existing photonic hardware, and how quickly cryogenic operation stops being a barrier and starts being a design assumption. Given how directly this speaks to the energy-efficiency bottleneck facing photonic and quantum computing architectures, expect this specific coupling mechanism to show up as a reference point in follow-up research well before it shows up in any deployed system.
