A team at the University of Oxford has coaxed a single trapped ion into a quantum state so exotic it has no classical counterpart, marking the first experimental demonstration of “quadsqueezing,” a fourth-order quantum interaction that had existed only in theoretical proposals until now. The result, published in May 2026 in Nature Physics, pushes the boundaries of what physicists can do with the quantum motion of atoms and opens a new route toward more powerful quantum computers and ultra-precise sensors.
“We’ve shown that you can engineer these highly nonlinear quantum interactions in a trapped-ion system and do it fast enough that the delicate quantum features survive,” said Razvan Bazavan, the study’s lead author and a researcher in Oxford’s Department of Physics.
Squeezing, but turned up to four
To understand quadsqueezing, it helps to start with ordinary quantum squeezing, a technique physicists have used for decades. In standard (second-order) squeezing, the quantum uncertainty in one property of a particle, say its position, is compressed at the expense of increased uncertainty in another property, its momentum. The result is a state that can beat the usual quantum noise limits, which is why squeezed light already plays a role in gravitational-wave detectors like LIGO.
Trisqueezing (third-order) and quadsqueezing (fourth-order) go further. Instead of simply compressing uncertainty along one axis, these higher-order operations reshape the quantum state’s probability distribution into complex, multi-lobed patterns that cannot be described by the smooth, bell-curve-like distributions of classical physics. Physicists call these “non-Gaussian” states, and they are considered a missing ingredient for universal quantum computing, the kind that could tackle problems no classical machine can solve efficiently.
The catch has always been speed. Traditional methods for generating these higher-order states in trapped ions rely on driving so-called motional sidebands, transitions that couple an ion’s internal energy levels to its vibration in the trap. Each step up in order requires exploiting a weaker and weaker coupling, governed by a quantity called the Lamb-Dicke parameter. By the time you reach the fourth order, the interaction is so feeble that decoherence, the tendency of quantum states to blur into classical noise, destroys the state before it fully forms.
A shortcut through two forces at once
The Oxford group found a way around that bottleneck. Rather than driving a single, weak fourth-order sideband, the team applied two non-commuting spin-dependent forces to a single strontium-88 ion simultaneously. Each force on its own produces only a linear (first-order) coupling between the ion’s spin and its motion. But because the two forces do not commute, meaning the order in which they act matters, their combination generates an effective nonlinear interaction. By tuning the relative strengths and phases of the two forces, the researchers could dial up second-, third-, or fourth-order squeezing at will.
The theoretical blueprint for this approach appeared in a 2021 Physical Review A paper proposing a hybrid quantum computing scheme for trapped ions. The Oxford experiment is the first to put that idea into practice and extend it all the way to the fourth order.
The payoff in speed is dramatic. According to the Oxford team, their method operates more than 100 times faster than the conventional sideband approach for generating equivalent higher-order states. That margin is not incremental; it is the difference between a quantum state that survives long enough to be measured and one that never fully materializes.
Proving the states are genuinely quantum
Generating an unusual state is one thing. Proving it is genuinely quantum is another. The team used a technique called characteristic-function tomography to reconstruct the full quantum state of the ion’s motion in phase space, the abstract landscape where position and momentum are plotted together. The resulting maps, known as Wigner functions, displayed the telltale signatures of deep quantum behavior: multi-lobed structures and regions where the Wigner function dips below zero, something flatly impossible for any classical probability distribution.
The raw experimental data behind these reconstructions are publicly available as downloadable spreadsheet files attached to the Nature Physics paper, an unusual level of transparency that allows other research groups to independently verify the analysis. An earlier version of the work appeared as a preprint on arXiv, and comparison between that draft and the final published paper shows the central claims and methods remained consistent through peer review, with refinements concentrated in error analysis and presentation.
What the experiment does not yet show
For all its novelty, the demonstration leaves significant questions unanswered. The Nature Physics paper does not report explicit gate fidelity numbers or detailed error rates for the quadsqueezing operation, metrics that are standard currency in quantum information science. Oxford’s press materials describe the results as “high fidelity” without attaching a percentage, and the publicly available data files cover state reconstruction rather than a full error budget.
No independent laboratory has yet replicated the result, so the performance rests entirely on one team’s apparatus and calibration. The 100-times speed claim, while consistent with the known physics of Lamb-Dicke scaling, has not been tested under standardized comparison conditions by an outside group.
Perhaps the biggest open question is practical relevance. Non-Gaussian states like those produced by quadsqueezing are theoretically valuable for fault-tolerant quantum computing, quantum error correction, and precision measurement. But the Oxford experiment involved a single trapped ion. Scaling the technique to the multi-ion chains, photonic interconnects, and repeated gate cycles required by a working quantum processor is a separate engineering challenge that has not been addressed. How these fragile fourth-order states hold up against real-world noise sources, including motional heating, laser phase fluctuations, and crosstalk between ions, remains unmapped.
Competing platforms also loom. Groups working with superconducting circuits and photonic systems are pursuing their own routes to non-Gaussian state generation, and it is too early to say whether trapped ions will hold a lasting advantage in this arena.
Where quadsqueezing fits in the larger picture
The result is best understood as a proof of principle that clears a specific theoretical barrier. For years, physicists knew that non-Gaussian operations would be needed to unlock the full power of continuous-variable quantum computing, an approach that encodes information in the smooth, wave-like properties of quantum systems rather than in discrete qubits. Generating those operations efficiently in a real experiment was the sticking point. The Oxford work shows it can be done, and done fast, in at least one well-controlled setting.
The enabling technology, a bichromatic laser field driving an optical quadrupole transition in strontium-88, is documented in a separate accepted manuscript from the same Oxford group. That paper establishes the control primitives on which the quadsqueezing demonstration relies, including the ability to engineer tailored spin-motion couplings and suppress unwanted decoherence. Its existence signals that the result is not an isolated trick but part of a systematically developed experimental platform.
For the broader quantum technology community, the milestone reframes what is experimentally accessible. Fourth-order squeezing has moved from a line in a theorist’s proposal to a measured reality in a working ion trap. Whether it becomes a building block of future quantum machines will depend on the next round of experiments: scaling to multiple ions, benchmarking against noise, and integrating quadsqueezing into actual quantum algorithms. Those tests will determine whether this proof of principle becomes a practical tool or remains a striking but isolated laboratory achievement.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
