Oxford reshapes quantum cat states with trapped ions

Oxford physicists have demonstrated a new class of Schrödinger’s cat-like quantum states, pushing laboratory control of quantum superposition beyond the familiar model of two classical-looking wave packets held in a fragile balance.

The advance, published in Physical Review X on 3 June, shows that cat-like states can be built from components that are already strongly quantum, rather than from the coherent states traditionally used to mimic the famous alive-and-dead thought experiment. The result gives researchers a more flexible way to engineer quantum systems that may prove useful for error correction, sensing and continuous-variable quantum computing.

The work was carried out by Sebastian Saner, Oana Băzăvan, D. J. Webb, Gabriel Araneda, D. M. Lucas, C. J. Ballance and Raghavendra Srinivas at the University of Oxford’s Department of Physics. Their experiment used the motion of a single strontium-88 ion held in an ion trap, a platform prized because it combines two systems: an internal electronic state that can act as a qubit, and physical motion that behaves like a quantum harmonic oscillator with many possible energy states.

Schrödinger’s cat, proposed in 1935, was intended to expose the oddness of applying quantum rules to everyday objects. Modern laboratory “cats” are not animals but superpositions in which light, atoms or trapped-particle motion occupy two distinct states at once. Conventional oscillator cat states usually combine two coherent wave packets separated in phase space, making them quantum in their relationship but comparatively classical in their individual structure.

The Oxford team changed that architecture. It generated superpositions whose components include squeezed, trisqueezed and quadsqueezed states, where quantum uncertainty is redistributed or reshaped rather than merely displaced. These states are more intricate than standard coherent-state cats and can carry symmetries and interference patterns valuable for computation.

The method relied on engineered interactions that entangled the ion’s internal spin state with different possible motional states. A mid-circuit measurement then projected the ion’s motion into the target superposition while preserving coherence. By adjusting experimental settings, the researchers controlled the relative size, phase, orientation and separation of the components, including combinations of different nonclassical states in the same superposition.

“This approach gave us a tool to sculpt quantum superpositions into almost any shape,” Saner said. The team reconstructed the states through quantum tomography, a measurement process that maps the state in detail. The reconstructions showed interference fringes and Wigner negativity, indicators that the outcomes were genuine quantum states with no classical equivalent.

The technical importance lies partly in the move from binary qubits to richer oscillator-based systems. A qubit can store information in a superposition of 0 and 1, but a harmonic oscillator has many levels and offers a wider design space. Such systems underpin proposals for bosonic quantum error correction, where information is stored in oscillator states designed to resist or reveal noise.

Error correction remains one of the main obstacles to useful quantum computers. Qubits are highly sensitive to their environment, and today’s processors require elaborate safeguards before they can run long calculations reliably. Cat-state and Gottesman-Kitaev-Preskill codes are among the approaches being explored to reduce the overhead needed to protect quantum information. The Oxford states add new building blocks to that search, though the work remains a physics demonstration rather than a ready computing module.

Srinivas, who supervised the study, said the team believed it was “still scratching the surface” of what could be done with the technique. The authors argue that the approach should apply to any system in which a quantum harmonic oscillator is coupled to a spin-like degree of freedom, including superconducting circuits, atoms in cavities, optical tweezers and some nanoparticle systems.

The finding lands as quantum computing moves from laboratory milestones towards larger industrial bets. More than 300 companies are engaged in quantum-technology activity, while private investment in quantum start-ups jumped in 2025 amid stronger interest in computing, sensing and secure communications. IBM, Google, Quantinuum, IonQ, Rigetti and specialist groups working on bosonic and cat-qubit architectures are among the players trying to turn fragile quantum effects into reliable hardware.