Analysis: Science & Technology — 16 October 2026
Atomic steps one atom high steer superconducting vortices more than 1,000 times more easily
Researchers at Japan’s National Institute for Materials Science report that steps only one atom high can act as rails for quantum vortices in an ultrathin superconductor. Four-terminal resistance measurements showed the vortices moved more than 1,000 times more easily along the steps than across them at intermediate magnetic fields. Scanning tunneling microscopy confirmed parallel steps and vortices sitting on them. The result, published in Physical Review B, is a rare demonstration of directional control in two-dimensional superconductors.
Superconductors carry current without resistance below a critical temperature, but magnetic flux enters as vortices—tiny cores ringed by circulating supercurrent. When those vortices move, they generate resistance and heat, which is why their motion often sets the practical current limit. Directing vortices in atomically thin films has been especially hard because conventional pinning sites are scarce and poorly ordered.
The guiding effect can be tuned with temperature and magnetic field. Between about 0.10 and 0.20 tesla the team saw one-dimensional, pinning-free flow along the steps. At the lowest temperatures, motion appeared to proceed by quantum tunneling rather than classical hopping. Fact: the anisotropy is large and field-dependent. Analysis: whether engineered vicinal surfaces can be used in devices, rather than remaining a laboratory curiosity, is still open, as is how heat would flow along the same atomic rails.
Sources: ScienceDaily; Physical Review B (Qian et al., 2026); NIMS/MANA.
Silicon spin qubits keep 99.99% gate fidelity with three qubits driven at once
A RIKEN-led team reports in Nature Physics that five silicon spin qubits, controlled through a single shared microwave line, reached single-qubit gate fidelities well above 99.99% when driven one at a time, with the best approaching 99.999%. Using a pairwise phase-calibration protocol, they kept fidelities above 99.99% while operating up to three qubits simultaneously. With all five driven together, fidelity stayed near 99.9%. Silicon spin qubits match existing chip fabrication, but parallel microwave control has historically degraded accuracy.
The dominant error source was the AC Stark shift: shared microwave drives shift qubit energy levels and add unwanted phase. Instead of calibrating every combination of qubits, the researchers used pairwise calibrations to cancel those phases. Remaining loss at five-qubit drive was attributed to decoherence from higher microwave power, not the phase error they had corrected.
Fact: three-qubit parallel gates now sit above the 99.99% mark often cited as a practical threshold for fault-tolerant architectures. Analysis: shared control lines help scaling, but drive-induced decoherence still grows with the number of qubits. Device redesign, not only software calibration, will decide whether the method survives larger arrays.
Sources: Phys.org; Nature Physics (Wu, Camenzind, Tarucha et al., 2026); RIKEN.
Highly charged ion spectra show lutetium nuclei are smaller than ytterbium
Using extreme-ultraviolet spectroscopy of highly charged ions at Tokyo-EBIT, an international group compared nuclear charge radii of lutetium and ytterbium directly. Emission lines from Lu⁵⁹⁺, Lu⁶⁰⁺, Yb⁵⁸⁺ and Yb⁵⁹⁺, converted with high-precision atomic-structure calculations, indicate that lutetium (Z = 71) is slightly smaller than ytterbium (Z = 70). The finding, in Nature Communications, reverses a long-standing anomaly in which odd-proton lutetium appeared larger than its even-proton neighbor.
Nuclear size is usually inferred from electron scattering, muonic atoms, or laser spectroscopy of neutrals and singly charged ions—methods that do not easily compare different elements. In highly charged ions, remaining electrons sit close to the nucleus, amplifying the volume shift. The new radii restore the expected odd-even staggering for N = 94 in the rare-earth region.
Fact: the Lu–Yb “reversal problem” is resolved under this method. Analysis: the same data still show even-odd effects that current nuclear-structure models do not fully capture. Whether the pattern is general will require neighboring rare earths and radioactive nuclides.
Sources: Phys.org; Nature Communications (Takacs et al., 2026); National Institutes of Natural Sciences.