A quantum processor probes a transition beyond the usual rules #

The standard theory of phase transitions organizes matter by the symmetries its phases break. A direct, continuous transition between phases that break different symmetries can fall outside that conventional picture. Researchers have now used superconducting quantum processors to investigate such a transition in a one-dimensional spin chain, between ferromagnetic order and a valence-bond solid, in which neighboring spins form a patterned arrangement of pairs.

Combining classically assisted state preparation with hardware measurements, the team tracked the weakening of long-range magnetic correlations and changes in the fluctuations of the competing orders. Correlations at different separations and comparisons across system sizes showed signatures consistent with critical behavior, rather than establishing it conclusively. The preprint also follows the evolution of initially paired states, demonstrating a way to study both competing phases and their dynamics on quantum hardware.

When a Rydberg blockade springs a leak #

In chains of highly excited Rydberg atoms, strong interactions can prevent neighboring atoms from being excited together. This “blockade” makes a complicated system easier to describe—but a real blockade is never infinitely strong. A theoretical study examines how that imperfection changes a periodically driven chain, disrupting a special flat structure in its spectrum of driven states.

The researchers identify two routes to this breakdown: resonances that produce neighboring excitation pairs, and virtual processes that pass through such pair configurations. Calculations checked against exact numerical results show that the outcome depends strongly on the starting state. A chain with no excitations is unusually resistant to resonant pair formation, while other patterns display slow oscillations or long-lived special states absent from the ideal blockade model. Measuring neighboring excitation pairs offers a practical way to test these predictions in Rydberg-atom experiments.

Shaping entanglement with a programmable light pump #

Topological photonic structures can provide optical modes protected by symmetry, but the entangled states they generate are often constrained by the structure’s geometry. An experimental preprint demonstrates a more flexible approach: shaping the light that pumps an integrated silicon photonic lattice to select the structure of the photon-pair state it produces.

The team tailored pump superpositions to address particular two-photon states, steer measured correlations and target different numbers of entangled modes. Measurements were reproducible across nominally identical devices despite fabrication-induced disorder. The significance is not that topology removes every source of error, but that it may coexist with useful programmability—a combination needed if such devices are to supply tailored states for quantum communication and photonic computing.

Time crystals as ingredients for classical logic #

A discrete time crystal responds to periodic driving with a rhythm that repeats more slowly than the drive itself. In a theoretical study of coupled, dissipative Dicke models—systems describing collective light–matter interactions—researchers find that coupling two such oscillators reduces their amplitude and, beyond a sufficiently strong coupling, destroys their time-crystalline motion. A growing relaxation time near that boundary suggests a continuous transition.

The same study shows how briefly coupling an oscillating system to a nonoscillating one can seed a new time crystal, either in step or out of step with the original. Using those possibilities, the authors propose sequences of couplings that implement classical NAND and NOR logic. They examine noise and quantum fluctuations to identify where the scheme remains viable. This is a proposal supported by simulations, not a demonstrated logic device.

A tunable quantum choice between swapping and staying put #

Researchers have experimentally implemented an exponential-SWAP gate between two vibrational modes of trapped ions. Rather than simply exchanging the modes’ states or leaving them untouched, the gate creates a coherent superposition of those two operations, with continuously adjustable weights.

The implementation combines spin-independent and spin-dependent beam-splitter interactions. Measurements of selected portions of the modes’ joint characteristic function—a mathematical description of their quantum state—tested the operation for different inputs and demonstrated its tunability. The result adds a controllable interaction for architectures that combine ordinary qubits with information stored in oscillator modes.

A better quantum memory—and a warning from long runs #

Quantum error correction spreads information across physical qubits so that a larger code can, under suitable noise conditions, make a more reliable logical memory. On a cloud-accessible IBM superconducting processor, researchers report improved surface-code memory performance by combining shorter readout, fused measurement and reset, and protective echo operations on data qubits during readout. Increasing the code’s protected distance from three to six, while keeping its other distance fixed, reduced per-round logical error rates from about 0.5 percent to roughly 0.2 percent in both tested memory bases.

The improvements came with important qualifications. Performance varied with the chosen region of the processor, and the advantage of larger codes weakened in experiments lasting up to 150 error-correction rounds. Long-lived correlations may indicate a persistent error component. The preprint therefore offers both evidence of logical error suppression and a useful account of its limits: better circuits help, but dependable memories also require attention to hardware variation and errors that persist over time.


These six preprints, drawn from the 68 papers announced in the field this week, share a practical question: how much control can researchers gain over complex quantum behavior without losing the features they want to exploit? The answers range from programmable entanglement and tunable gates to sharper accounts of imperfect blockades and persistent memory errors. All remain preliminary, not yet peer-reviewed reports, but together they show why useful quantum science depends as much on understanding limits as on demonstrating new capabilities.

Was this letter useful to you?

Written by AI from the 68 papers of the week in this field. Most are preprints, not yet peer reviewed: the text says what their authors report.

Get one letter by email every week with a free account, all of them with Premium. Create a free account