Physics News — August 2026
2026年8月の最新研究・発見(全127件) 最終更新 / Last updated: — 査読論文・一次ソース付き / Peer-reviewed, with primary sources —
📰 2026年8月 のニュース / August 2026 (全127件)
2026年8月(August 2026)に発表・注目された基礎物理学の最新ニュースと研究解説。一次ソース(DOI・arXiv・機関発表)付きで月内の項目を掲載しています。Recent physics news and research explanations from August 2026, with primary sources.
A photonic time crystal (PTC) is the temporal analogue of an ordinary photonic crystal: instead of a refractive index that repeats in space, its optical properties are modulated periodically in time. Realizing one all-optically has long been out of reach, because the modulation must be both extremely strong and faster than a single optical cycle.
An international team from École Polytechnique, Collège de France and Helmholtz-Zentrum Dresden-Rossendorf (HZDR) now demonstrates the first all-optical photonic time crystal, built from a surface-plasmon cavity metamaterial — micrometre-scale gold resonators on an indium-antimonide semiconductor — driven at terahertz frequencies by HZDR’s superradiant TELBE source. The drive achieves near-unity, sub-optical-cycle modulation, dynamically changing the carriers’ kinetic energy and effective mass by up to 80%. Spectroscopy reveals a transition into the PTC regime mediated by an exceptional point at which two Floquet-driven optical eigenmodes coalesce, and in the PTC regime an emergent gain reduces plasmonic losses by more than 50%, putting plasmonic lasing within experimental reach. The result opens a route to ultrafast optical computing, terahertz lasers and time-domain photonics. Published in Nature (press releases 30–31 July 2026).
A glass nanoparticle levitated in vacuum by a focused laser is a promising quantum sensor: cooled near its motional quantum ground state, it becomes exquisitely sensitive to tiny forces. But there is a catch — the zero-point fluctuation of the ground state itself masks small displacements, and the high trap frequencies best suited to cooling are the worst for acceleration sensitivity.
Mitsuyoshi Kamba, Sotatsu Otabe, Ken Funo, Takahiro Sagawa and Kiyotaka Aikawa (University of Tokyo) resolve this dilemma with a dynamical protocol: after cooling a ~300-nm silica nanoparticle close to its ground state (an effective temperature of about 17 microkelvin), they rapidly quench the trapping-light intensity from the high value optimal for cooling to a low value optimal for sensing. The nonequilibrium dynamics amplify the acceleration-induced displacement beyond the zero-point limit of the original trap, improving acceleration sensitivity by two orders of magnitude and demonstrating detection of a weak gravitational acceleration of 1.2 mm/s² — about one ten-thousandth of Earth’s gravity — with the observed dynamics and Fisher information well reproduced by quantum Langevin simulations. Because the particle is electrically neutral, the scheme is immune to electromagnetic noise, and it points toward dark-matter detection, neutrino-mass studies, high-frequency gravitational-wave searches and quantum-gravity tests. Published in Physical Review Letters (28 July 2026); press release 31 July 2026.
“Is there a single tile that can cover the plane only aperiodically?” This long-standing mathematical riddle, known as the einstein problem (from the German ein Stein, “one stone”), was finally solved in 2023 with the discovery of the hat monotile. Yet no experiment had probed what physics this peculiar aperiodic order produces.
Yuto Moritake (Institute of Industrial Science, University of Tokyo) and Masaya Notomi (Institute of Science Tokyo / NTT Basic Research Laboratories) and colleagues fabricated an artificial nanostructure with the hat-monotile arrangement — hundreds of thousands of holes patterned in a silicon-nitride membrane — and measured its optical diffraction. They observed a pinwheel-shaped diffraction pattern reflecting the tiling’s chiral symmetry, and found that the diffraction intensity differs between left- and right-handed circularly polarized light — a response absent in conventional, mirror-symmetric quasiperiodic structures and not anticipated by prior theoretical studies. The work opens the science of chiral quasiperiodic order and suggests new routes for light control in nanophotonics and metasurfaces. Published in Nature Communications, 29 July 2026.
「一種類のタイルで、非周期的にしか平面を敷き詰められないものはあるか?」——この長年の数学の未解決問題はアインシュタイン問題(独語の ein Stein「1つの石」に由来)と呼ばれ、2023年にハット型モノタイルの発見によってついに解決された。しかし、この特異な非周期秩序がどんな物理現象を生むのかを実験で確かめた研究はなかった。
Heavy-fermion metals are compounds in which conduction electrons hybridize with localized f-electrons (the Kondo effect), acquiring effective masses hundreds of times that of a free electron. CeSiI stands out as a rare van der Waals heavy-fermion metal — it can be exfoliated into atomically thin layers like graphene — hosting antiferromagnetic order (TN = 7.5 K) inside a Kondo-coherent state (T* ≈ 50 K).
Tong Shi and colleagues (Institute of Physics, Chinese Academy of Sciences, and collaborators) now use high pressure as a tuning knob and map CeSiI’s full temperature–pressure phase diagram. Suppressing the antiferromagnetic order at a critical pressure of about 6 GPa makes a superconducting dome emerge, with a maximum Tc of 240 mK, accompanied by an unusual V-shaped, nonmonotonic evolution of the Kondo coherence temperature. Normal-state transport shows non-Fermi-liquid behaviour and a diverging effective electron mass — hallmarks of quantum criticality. The results establish CeSiI as a heavy-fermion superconductor and a unique two-dimensional platform for exploring the interplay of strong correlations, Kondo hybridization, magnetism and unconventional pairing. Published in Nature Physics, with a News & Views on 28 July 2026.
Gold has been prized for millennia precisely because it does not tarnish. The textbook explanation is chemical: gold holds its electrons tightly, so oxygen binds only weakly. But that alone does not account for how much more inert gold is than the numbers suggest.
Santu Biswas and Matthew M. Montemore (Tulane University) used first-principles simulations of how atoms and electrons behave when oxygen molecules meet two of the most common gold surface structures. They find that the decisive factor is what happens the instant a fresh gold surface is created: the exposed atoms of Au(110) and Au(100) shift from looser, square-like arrangements into denser quasihexagonal patterns — the well-known surface reconstructions. Their simulations show that on unreconstructed surfaces oxygen molecules would split apart and react with gold far more easily, whereas on the reconstructed, quasihexagonal surfaces O2 dissociation is slowed by a factor of a billion to a trillion, leaving the surface essentially unchanged. So gold’s inertness is not merely passive weak binding but an active, self-protecting structural response. The flip side is practical: gold catalysts are poor at splitting O2, so preventing or reversing these reconstructions could make gold a far better oxidation catalyst. Published in Physical Review Letters 136, 206203 (21 May 2026); widely covered by science media through July 2026.
Santu Biswas・Matthew M. Montemore(米テュレーン大学)は、酸素分子が最も一般的な2種類の金表面構造に出会ったときの原子と電子の振る舞いを第一原理シミュレーションで調べた。決定的だったのは、新しい金表面が生まれた瞬間に何が起こるかである。Au(110)とAu(100)では、露出した原子が緩い正方形状の配列からより密な準六方構造へと移行する——よく知られた表面再構成だ。シミュレーションによれば、再構成していない表面では酸素分子がはるかに容易に解離して金と反応するのに対し、再構成した準六方表面ではO₂の解離が10億〜1兆分の1に遅くなり、表面はほぼ変化しないまま保たれる。つまり金の不活性さは単なる受動的な弱い結合ではなく、能動的に自らを守る構造応答なのだ。裏を返せば実用的な含意もある。金触媒はO₂の解離が苦手だが、この再構成を防いだり逆転させたりできれば、はるかに優れた酸化触媒になりうる。Physical Review Letters 136, 206203(2026年5月21日付)掲載。2026年7月にかけて科学メディアで広く報道。
Liquid-xenon detectors like XENONnT (Gran Sasso, Italy) normally require both a scintillation flash (S1) and an ionization signal (S2) to reconstruct an event. Dropping the S1 requirement and using ionization-only (S2-only) data lowers the energy threshold dramatically, opening sensitivity to much lighter dark-matter particles — at the cost of larger, harder-to-model backgrounds.
The XENON Collaboration now reports a blinded S2-only search with a 7.8 tonne-year exposure accumulated over 579 days across three science runs, enabled by dedicated background-suppression techniques and the first complete S2-only background model for XENONnT. No significant excess is observed. The analysis sets 90%-confidence upper limits on spin-independent dark matter–nucleon and spin-dependent dark matter–neutron scattering for light dark-matter masses of a few GeV/c², on dark matter–electron scattering, and — most notably — world-leading bounds on sub-keV axionlike particles and dark photons absorbed by electrons. The sensitivity now pushes toward the region where coherent elastic neutrino–nucleus scattering (CEνNS) becomes an irreducible background — the so-called neutrino fog. Published in Physical Review Letters 137, 051003 (30 July 2026).
Topological physics has given wave systems — electrons, photons, sound, mechanics — robust one-way edge channels, epitomized by the Haldane model of a Chern insulator. But heat diffusion is not a wave: it does not respond to magnetic fields and carries no intrinsic spin, so realizing a Chern insulator for heat has long been considered challenging, if not impossible.
Jiaxin Li and colleagues now propose and realize a thermal analogue of the Haldane model. The trick is to endow the temperature field with a degree of freedom playing the role of a wave function’s phase, and to synthesize the complex next-nearest-neighbour couplings of the Haldane model by tailoring intralayer and interlayer “hopping” between thermal sites. Fabricating the thermal Haldane lattice with actively controlled, programmable thermal sources, the team experimentally observed the evidence of one-way chiral edge states of heat. The work extends Chern-insulator physics into purely diffusive systems and hints at flexible, robust manipulation of heat and mass transfer — a fresh direction for thermal management. Published in Physical Review Letters 137, 056301 (28 July 2026).
Zitterbewegung (“trembling motion”) is a jittering of the centre of mass predicted for relativistic Dirac electrons, arising when operators fail to commute. In non-Abelian gauge theories — the mathematical backbone of the Standard Model — the order of operations matters, and engineering such fields for photons has become a central goal of synthetic-dimension photonics.
Shu Yang, Bengy Tsz Tsun Wong, Jinbing Hu and Yi Yang (University of Hong Kong) experimentally created non-Abelian electric fields on a photonic frequency chain — a synthetic lattice whose “sites” are optical frequencies — using a polarization-multiplexed, time-modulated fibre ring resonator. Engineering spin-orbit coupling via modulation dephasing, polarization rotation and retardation gives programmable control over synthetic Floquet bands. With self-heterodyne coherent detection they observed photonic Zitterbewegung induced by the non-Abelian electric field — the oscillation of a wave packet’s centre along the frequency axis caused by the noncommutativity of scalar and vector potentials — and, switching on an Abelian field via modulation detuning, its interference with Bloch oscillations. The platform bridges synthetic dimensions with non-Abelian gauge theory for emulating relativistic quantum mechanics, with applications in frequency-domain optical computation and frequency-comb control. Published in Physical Review Letters 137, 056901 (2026); selected for a Viewpoint in Physics and as an Editors’ Suggestion.
High-frequency mechanical oscillators with long coherence times are a prized resource: they can serve as quantum memories, transducers between microwaves and optics, and precision sensors. But pushing their coherence into the range quantum applications demand first requires knowing what destroys it — and the origins of phonon decoherence in crystalline media have been poorly pinned down.
Combining non-invasive laser (Brillouin) spectroscopy with materials analysis, researchers used micro-fabricated high-overtone bulk acoustic-wave resonators (μHBARs) in crystalline quartz as a test bed and identified phonon–surface interactions as the dominant source of decoherence. Reducing those interactions through chemical mechanical polishing then yielded μHBARs with Q-factors above 240 million at 12 GHz, corresponding to phonon coherence times beyond 6 ms and record-level f–Q products, with a path identified to >100 ms. The work is from Yizhi Luo, …, Peter T. Rakich and colleagues (Yale University). Because the figure of merit for decoupling a resonator from its thermal environment is the frequency–Q product, this points a concrete materials-engineering route toward long-lived solid-state quantum memories. Published in Nature Physics; featured in a Research Briefing on 30 July 2026.
An ensemble of atoms coupled through their highly excited Rydberg levels can behave as a single giant two-level system — a “superatom” — because the Rydberg blockade forbids more than one excitation within a blockade radius. Superatoms can encode qubits and emit single photons on demand, making them attractive nodes for quantum networks. But real ensembles are large and positionally disordered, so the blockade is never perfect, and describing the residual imperfection accurately has been hard.
Valentin Magro, Sébastien Garcia and Alexei Ourjoumtsev (Collège de France) derive such a description from first principles: a treatment of interactions in a large, disordered ensemble that is simultaneously accurate, physically informative, and numerically scalable. They successfully test it against brute-force numerics and against experimental data. The model proves essential for making quantitative predictions of gate fidelities and photon-emission efficiencies, and thus for guiding experiments toward large-scale superatom-based quantum systems. Published in Physical Review Letters 137, 053605 (28 July 2026).
The quantum theory of light describes multiphoton states whose phase-space distributions are non-Gaussian — a property tied to quantum advantage in continuous-variable computing, since Gaussian states and operations alone are classically simulable. Several hierarchies rank non-Gaussian states by resource content; a prominent one is the stellar rank, conventionally read as counting how many photons must be added to a Gaussian state.
Nicolas Moulonguet, Eloi Descamps, Pérola Milman and colleagues (Université Paris Cité/CNRS, ENS-PSL, Paris-Saclay) give non-Gaussianity a physical interpretation by showing that the stellar rank emerges as a limiting case of the roots of Majorana polynomials — polynomials that uniquely represent bosonic states defined with a quantized phase reference. The consequence is a revised reading of both quantities: once superselection rules (here, photon-number conservation relative to a quantized phase reference) are properly accounted for, quadrature non-Gaussianity and nonzero stellar rank act as witnesses of particle entanglement, rather than signatures of photon addition. They further show the stellar rank is inherently basis-dependent, tied to the choice of coherent states and quadrature eigenstates as the computational basis — which qualifies claims about its relation to computational resources. Published in Physical Review Letters 137, 050203 (29 July 2026).
Topological insulators come in two flavours. First-order ones (Chern insulators, quantum spin Hall systems) host propagating edge states; higher-order ones host localized corner states. The two have been thought mutually exclusive: a nonzero first-order index (Chern or spin-Chern number) makes the conventional higher-order topological indices ill-defined.
Jiancheng Zheng, Zhenhang Pu, Jiuyang Lu, Weiyin Deng, Manzhu Ke and Zhengyou Liu (Wuhan University) show the exclusion is not fundamental. Even when the first-order indices are nonzero, such systems can still support higher-order corner states as Jackiw–Rebbi-type modes arising from the topology of the edge bands themselves. Realizing this in an acoustic crystal, they observe traveling edge states and trapped corner states coexisting within the exact same energy band — higher-order topology embedded inside first-order topological bands. The result enlarges the design space for topological wave devices, where robust one-way transport and robust localization can now be engineered in a single band. Published in Physical Review Letters 137, 056605 (28 July 2026).
Future fusion reactors face a dilemma: the high-confinement mode (H-mode) that keeps the plasma hot tends to produce violent edge bursts (ELMs) that can damage the wall, while taming them often costs performance. The quasicontinuous exhaust (QCE) regime is a promising “Goldilocks” operating point — high confinement with small, benign, quasi-continuous heat exhaust — but its underlying turbulence mechanism was unclear.
Kaiyu Zhang, Wladimir Zholobenko, Frank Jenko and colleagues (Max Planck Institute for Plasma Physics, with the ASDEX Upgrade Team) performed global two-fluid turbulence simulations of the QCE regime on the ASDEX Upgrade tokamak. The simulations reveal that a quasicoherent mode drives mesoscopic oscillations of the pedestal boundary across the magnetic separatrix and ejects ballistic filaments (“blobs”), reproducing both the mean profiles and the turbulent fluctuations observed experimentally. The behaviour arises from a synergistic interplay between kinetic ballooning modes and resistive X-point modes straddling the separatrix. Understanding this mechanism supports extrapolating the reactor-relevant QCE regime to future devices such as ITER and beyond. Published in Physical Review Letters 137, 055102 (30 July 2026); featured in a Physics Synopsis.
Galactic cosmic rays arrive at Earth almost — but not perfectly — isotropically. Tiny deviations at the 10−3–10−4 level, mapped over the sky, encode how charged particles are scrambled by magnetic fields on their way to us. Medium-scale anisotropies, spanning tens of degrees, are a particularly sensitive probe of the local magnetic environment within the cosmic-ray scattering length.
The LHAASO Collaboration, using the KM2A square-kilometre array on the Tibetan Plateau, reports precision observations of medium-scale anisotropies in cosmic rays above 10 TeV, identifying four excess and four deficit regions, each spanning roughly 10°. Crucially, the analysis detects, for the first time with high significance, energy-dependent shifts in the centroids of two excess regions — the known Region B and a newly identified region — and characterizes how the fractional intensity of each region evolves with energy. These findings imply the anisotropies are shaped by the specific realization of the local turbulent magnetic field, imposing strict constraints on models of local turbulence and cosmic-ray propagation. Published in Physical Review Letters 137, 051004 (30 July 2026).
When an excited nucleus relaxes, it emits gamma rays with a probability described by the gamma-ray strength function (γSF). For decades, some nuclei have shown a puzzling low-energy enhancement (LEE) — far more low-energy gamma rays than theory predicted — and whether its electromagnetic character is electric or magnetic remained an open question with direct consequences for how elements are forged in stars, supernovae and neutron-star mergers.
An international collaboration of 25 institutions led by researchers at the Facility for Rare Isotope Beams (FRIB, Michigan State University) settled the question for zinc-70. Using FRIB’s LEBIT Penning trap, the team prepared — for the first time — isotopically and isomerically pure beams of the parent copper-70 in its ground state and in an excited isomeric state, then recorded the gamma rays of the daughter zinc-70 with the SuN total-absorption detector. Analysing the two datasets with the β-Oslo and Shape methods yielded two independent strength functions whose comparison shows conclusively that the low-energy enhancement is of magnetic (dipole) character. The result provides a benchmark for nuclear theory and sharpens neutron-capture rate calculations used in models of heavy-element nucleosynthesis. Published in Nature (DOI 10.1038/s41586-026-10758-3); press releases in late July 2026.
“Quantum advantage” has two hard requirements that are rarely met at once: the computation must be beyond the practical reach of the best classical simulations, and there must be a way to trust that the quantum machine actually did it correctly. Earlier random-circuit-sampling demonstrations satisfied the first but struggled with the second.
IBM and researchers from the University of Chicago (including Bill Fefferman’s group, with Jay Gambetta of IBM Research) report a demonstration that meets both criteria. In one of the largest error-correction demonstrations to date, they executed 70 logical qubits, running 2,415 logical two-qubit operations and 468 logical “T gates” in roughly 15 minutes. The key ingredient is a spacetime code: ancilla qubits distributed across both space and time detect errors mid-computation, and post-selecting the runs that pass the consistency checks cut the effective gate error to about one-tenth of the raw physical rate. The task is a structured “T-doped” circuit that keeps random-circuit-sampling hardness while remaining verifiable, so the result comes with statistical fidelity bounds — a lower bound on how faithfully the computation was executed — where leading classical simulation runtimes were prohibitive. Announced 30 July 2026.
The Riemann Hypothesis (RH), posed in 1859, asserts that all nontrivial zeros of the Riemann zeta function lie on a single critical line. It remains unproven, yet more than a thousand theorems — and parts of modern cryptography — rest on it. The Hilbert–Pólya idea that these zeros might be the spectrum of some physical operator has motivated decades of attempts to give RH a physical body.
Shijie Wei, Tao Xin, Guilu Long, Franco Nori and colleagues (Beijing Academy of Quantum Information Sciences, SUSTech, Tsinghua University, RIKEN) now establish a direct correspondence between the nontrivial zeros and dynamical quantum phase transitions (DQPTs) in two complementary engineered many-body systems — characterized respectively by the average accumulated phase factor and the Loschmidt amplitude, both of which vanish exactly at evolution times matching the zeros. This recasts RH as the occurrence of phase transitions at a unique temperature, identifying a previously unknown transition mechanism. In a proof-of-principle experiment on a five-qubit spin-based quantum processor, the signal collapsed at the first five nontrivial zeta zeros and only on the critical line. The team also proposes a quantum computational framework implementing both systems with polynomial resources, suggesting a quantum advantage in probing the hypothesis. Note: this is a physical correspondence and a numerical demonstration, not a proof of RH. Published in Nature Communications (peer-reviewed version of arXiv:2511.11199).
In monolayer transition-metal dichalcogenides such as WSe2, electrons carry an extra label — the valley index — that behaves like a pseudospin and can be addressed with circularly polarized light. Placing such a monolayer in an optical microcavity and applying a magnetic field turns its Landau levels into a many-body playground for cavity quantum electrodynamics.
Xinyue Zhang and colleagues report the realization of Landau polaritons in a WSe2 monolayer integrated into an optical microcavity. By embedding a two-dimensional electron gas in the monolayer and applying a perpendicular magnetic field, they achieve strong coupling between cavity photons and interband Landau-level transitions, producing hybrid light–matter quasiparticles. These polaritons show valley-dependent oscillations in both coupling strength and resonance energy, driven by the electrical and magnetic filling of valley-contrasting Landau levels. Moreover, intervalley correlations between the Landau-level transitions and opposite-valley electrons produce a nonlinear renormalization of transition energies and coupling strengths. The work establishes a new paradigm for cavity QED in correlated two-dimensional systems and demonstrates light-mediated control of quantum phases via the valley degree of freedom. Published in Physical Review Letters 137, 046904 (24 July 2026).
Altermagnets are the newly recognized third class of collinear magnets: they break a combination of time-reversal and rotational symmetries yet carry no net magnetization. That absence is exactly what makes them hard to study — the order parameter of a d-wave altermagnet has the same symmetry as magnetic multipoles — coupling to the product of a magnetic field and uniaxial strain — so conventional magnetometry has little to grip.
Rahel Ohlendorf, Elena Gati and colleagues (Max Planck Institute for Chemical Physics of Solids) attack the problem thermodynamically. By combining symmetry-guided uniaxial strain and magnetic-field tuning with highly sensitive elastocaloric-effect measurements — which detect the temperature change a material undergoes when strained adiabatically, and are exquisitely sensitive to symmetry-breaking order — together with first-principles theory, they probe the multipolar order in the candidate altermagnet MnF2 directly — achieving the first unambiguous bulk confirmation of altermagnetism in MnF2 and establishing a thermodynamic probe of its predicted finite-temperature altermagnetic critical point. With colleagues at TU Dresden, the University of Minnesota, St Andrews, UIUC and KIT, the approach supplies a general route to identifying altermagnetic order, a prerequisite for stray-field-free spintronics. Published open access in Physical Review Letters 137, 056702 (28 July 2026).
Singly ionized ytterbium (Yb+) is a workhorse of precision measurement and trapped-ion quantum information, prized for its extremely narrow “clock” transitions and its abundance of stable isotopes. That isotope richness enables King-plot analyses of isotope shifts, in which a deviation from linearity can signal a new boson coupling electrons to neutrons — physics beyond the Standard Model. A statistically significant King-plot nonlinearity has already been reported, and disentangling it from higher-order Standard Model effects requires more narrow transitions to feed into the analysis.
Patrick McMillin, Hassan Farhat, William Liu and Wesley C. Campbell report the observation of three semiforbidden transitions in Yb+ from the metastable 2Fo7/2 state. Because both the upper and lower states are long-lived, the transitions are sub-hertz narrow, complementing those already in routine use. The team reports the absolute frequencies of these electric-quadrupole transitions along with their isotope shifts, hyperfine structure and quadrupole transition moments, and finds that the spontaneous lifetimes of the excited states are limited by slow magnetic-dipole emission to lower-lying odd-parity states. Published in Physical Review Letters 137, 053002 (31 July 2026).
When heavy ions collide at the LHC or RHIC, they briefly create a quark–gluon plasma (QGP) — the deconfined state of matter that filled the microsecond-old universe. High-energy jets punching through it are the sharpest available probes, but the imprint of the medium on a jet mixes two effects that are hard to separate: the medium’s perturbative modification of the parton shower, and its hydrodynamical backreaction (the wake the jet leaves behind).
Energy correlators — correlations of energy flow between detector directions — have become popular because they connect measurements robustly to quantum field theory. Theory work so far focused on the simplest two-point correlator; mapping the QGP’s dynamics requires going further. This Letter presents a systematic theoretical study of multipoint energy correlators for jets fragmenting in a dense QGP, accounting for both the perturbative modification and the hydrodynamic response. Examining both the scaling of projected correlators and the shape dependence of the three-point correlator, the authors show how each provides distinct insight into jet–medium interaction, and discuss how modifications depend parametrically on the medium’s scales — opening a concrete route to experimentally separating jet modification from medium response. Published in Physical Review Letters 137, 052302 (31 July 2026).
Silicon is transparent to short-wavelength infrared (SWIR) light because those photons carry less energy than silicon’s band gap. SWIR detection therefore relies on III–V compounds such as InGaAs — expensive, and not monolithically compatible with silicon CMOS. Hyperdoping silicon with deep-level impurities beyond the solubility limit creates an intermediate band that absorbs sub-band-gap photons, but the resulting quantum efficiency has been too low to matter.
E. García-Hemme, D. Pastor and colleagues (Universidad Complutense de Madrid, with Universidad Politécnica de Madrid) report a Si-CMOS-compatible photodiode for room-temperature SWIR detection that integrates tellurium-hyperdoped silicon with light-trapping structures. By combining surface texturing with a back reflector, they force sub-band-gap photons to make many passes through the thin hyperdoped layer, boosting absorptance and delivering a breakthrough in sub-band-gap quantum efficiency. Because the device is built in silicon, it points toward CMOS-compatible night-vision and SWIR imaging without exotic materials. Published in Physical Review Letters 137, 057002 (29 July 2026); featured in a Physics Synopsis.
Highly charged ions (HCI) — atoms stripped of most or all of their electrons — are the best laboratories we have for testing quantum electrodynamics in extreme fields. But there is a catch: the only efficient way to make heavy HCI is to strip them in flight at relativistic speeds, whereas precision spectroscopy and mass measurements demand ions that are slow, or ideally at rest.
The HITRAP project at GSI Helmholtzzentrum für Schwerionenforschung, together with TU Darmstadt, has now closed that gap. Working with GSI’s Decelerator Division, the team produced fully stripped argon nuclei at roughly 30% of the speed of light, then decelerated them through a multi-stage chain — a double drift buncher, an interdigital H-type structure, a rebuncher and a radio-frequency quadrupole — and captured them in a Penning trap whose 387 mm capture region is mechanically compensated for a harmonic potential. They then performed the first in-trap electron cooling of highly charged ions. Lead author Simon Rausch earned his PhD in Wilfried Nörtershäuser’s group at TU Darmstadt. The result unlocks precision tests of strong-field QED, high-accuracy mass spectrometry of few-electron systems, and slow-HCI materials science that were previously out of reach. Published in Physical Review X 16, 031022 (29 July 2026).
Topological photonics can force light to travel in a single direction, routing optical signals around corners and past defects with no backscattering. Until now, however, that one-way flow lived only at the boundary between two engineered “topological insulator” regions — meaning most of the material was dead weight, unavailable for transport, and each protected path could carry only a thin single channel.
A team in China led by Xiaohan Cui and Che Ting Chan at the Hong Kong University of Science and Technology has now obtained the same unidirectional behaviour without any insulating regions at all. Using a honeycomb lattice of magnetic rods, they created four adjacent one-way channels — two carrying signals in each direction — packed densely across the bulk rather than confined to an edge. In microwave experiments the signals propagated around sharp bends and through constrictions with negligible backscattering and negligible leakage between channels, and the one-way transport survived deliberate structural distortion. Extending the scheme to optical frequencies could sharply increase the information density of photonic circuits. Published in Nature (2026).
Every wire loses energy as heat because electrons scatter — and in transmission lines that loss reaches up to 8% of the power generated. Yet physicists still argue about how resistivity is assembled from individual microscopic collisions, because in a real metal the electron-electron contribution is hopelessly entangled with lattice vibrations, disorder and structural change.
A team from the University of Toronto, the École Normale Supérieure in Paris and Lehigh University, with Joseph Thywissen as senior author, sidestepped that problem by replacing electrons with fermionic potassium-40 atoms cooled to near absolute zero and loaded into a three-dimensional cubic optical lattice of period 0.53 µm. A magnetic Feshbach resonance let them dial the interaction strength continuously — something no real metal permits. Resistivity rose with collision rate at first, then saturated: there is a ceiling. The saturation traces to a quantum enhancement of the effective particle size, so that atoms only nanometres across collide as though they were far larger — the authors liken it to ducks colliding at the size of the bubbles around them. The resulting bound, which they name lattice unitarity, is a clean diagnostic for strange metals: if a strange metal’s electron-electron scattering exceeds this ceiling, the excess cannot be blamed on stronger interactions, and its charge carriers are probably not conventional electrons. Published in Physical Review Letters136, 213401 (26 May 2026); press release 16 June 2026, with further coverage in late July.
CPT symmetry — the combination of charge conjugation, parity inversion and time reversal — predicts that matter and antimatter are perfect mirror images, obeying identical physical laws. Antihydrogen, an antiproton orbited by a positron, is the sharpest tool for testing that. Its 1S–2S transition has already been pinned to parts per trillion, but the ground-state hyperfine splitting — the tiny gap created by the magnetic interaction between antiproton and positron spins — had stalled at 400 parts per million since ALPHA first observed it in 2017.
The ALPHA Collaboration at CERN has now improved that by two orders of magnitude. A new trapping technique producing roughly 15,000 antihydrogen atoms within hours enabled microwave spectroscopy on about 24,000 anti-atoms, yielding a1S/h = 1,420,404.8 ± 1.1(stat.) ± 5.6(sys.) kHz in a 1 T field — a 4 ppm measurement, consistent with hydrogen. That precision matters for a specific reason: the antiproton is not a point, but a bundle of antiquarks and gluons whose charge and magnetism distribution shifts the splitting by roughly 40 ppm. The measurement is therefore now sensitive to the antiproton’s internal structure, and is approaching the limits of existing theory. Spokesperson Jeffrey Hangst calls it the culmination of work begun when the team first trapped antimatter in 2010. Published in Nature 653, 1022 (27 May 2026); APS Physics commentary followed in early June, and Physics World coverage appeared in late July 2026.
Hydrogen moves through vanadium — a leading candidate material for hydrogen storage and separation — by hopping between interstitial sites of the crystal lattice. Sometimes it behaves as a classical particle that must climb over energy barriers; at other times it takes a quantum shortcut, tunneling through the barriers like a wave. What decides which regime applies had remained unclear.
Researchers at the Institute of Industrial Science, The University of Tokyo combined nuclear reaction analysis, hydrogen hopping-rate (electrical resistance) measurements on an epitaxial vanadium thin film, and quantum mechanical calculations, and found that crystal symmetry acts as the switch. At low hydrogen concentration the lattice stays highly symmetric and hydrogen forms delocalized tunneling states spread over neighboring sites; at higher concentration, symmetry-lowering lattice distortions suppress tunneling and enforce thermally activated classical hopping. The result gives a concrete design rule — control symmetry to control hydrogen’s quantum behavior — for solid-state hydrogen storage and separation materials. Published in Nature Communications (15 July 2026); featured by science media on 4 August 2026.
🚀 / 地上最強クラスの「加速度」がクォーク・グルーオン・プラズマに潜んでいた——重イオン衝突の火の玉をAMPT・UrQMD輸送模型+ガウス平滑化で連続流体場に変換し、固有加速度を初めて系統的にマッピング。ピークは数百MeV相当で火の玉外縁を外向きに指し、加速度が「QCD相図の新しい軸」になり得ると提案(Zhong・Deng・Huang・Ma、復旦大学、Nuclear Science and Techniques掲載)
When heavy nuclei collide at nearly the speed of light, their matter briefly melts into quark–gluon plasma (QGP) — the hottest fluid in the universe. Its vorticity and electromagnetic fields have been studied intensively, but the acceleration driving its violent expansion had received far less attention, even though in hydrodynamics acceleration is as fundamental as vorticity.
A team led by Xu-Guang Huang and Yu-Gang Ma at Fudan University combined the AMPT and UrQMD transport models with Gaussian smearing to convert discrete particle data into continuous energy, momentum and velocity fields, and tracked the fluid acceleration across collision energies from 3.5 GeV to 2.76 TeV. The simulations show peak proper accelerations reaching the equivalent of several hundred MeV — among the strongest accelerations ever produced on Earth — with the largest transverse acceleration pointing outward near the fireball’s edge. The authors argue that acceleration may act as a thermodynamic control parameter of QCD matter (via Unruh-like non-inertial quantum effects), potentially opening a new axis of the QCD phase diagram alongside temperature and density. Published in Nuclear Science and Techniques (25 July 2026).
Mars, lacking a global magnetic shield, is directly scoured by the solar wind. Localized plasma clouds in its upper atmosphere were known to carry away large amounts of ionized gas — episodes of “bulk escape” — but their origin could not be pinned down, because no single spacecraft could watch the incoming solar wind and the escaping ions at the same time.
A Boston University-led team solved this with simultaneous two-point observations: China’s Tianwen-1 monitored the undisturbed upstream solar wind while NASA’s MAVEN tracked ions near the planet. The data provide direct evidence that the plasma clouds are nonlinear wave packets generated by the Kelvin–Helmholtz instability — the same shear-flow instability that makes wind raise waves on water — ruling out solar-wind gusts as the trigger. Ion fluxes inside the clouds are one to two orders of magnitude (10–100×) higher than in the known steady-state escape channels, the process concentrates on one hemisphere depending on the solar-wind electric-field direction, and the spatial scale of the KH waves is constrained for the first time by the two-point measurement. The mechanism likely operates at other unmagnetized planets; NASA’s twin-orbiter ESCAPADE mission is set to follow up. Published in Science Advances (31 July 2026).
Trapped-ion quantum simulators are usually operated as close to absolute zero as possible, because thermal motion causes errors. But the chemistry and biology one wants to simulate — electron transfer, photosynthetic energy transport — happens at finite temperature, where thermal fluctuations are not noise but an essential ingredient.
Guido Pagano’s group at Rice University has now engineered thermal reservoirs with independently tunable temperature and dissipation rate for the motional modes of a trapped-ion system, using two independent knobs: randomized electric-field kicks that set the heating rate, and laser cooling that sets the cooling rate. Balancing the two dials places the ions in a controlled thermal state at will. In demonstrations with small ion chains, the team simulated charge transfer (an electron hopping from donor through a barrier to an acceptor) and vibrationally assisted exciton transfer, directly observing how temperature reshapes the transfer pathways and activates transport across barriers that stay frozen at low temperature. The scheme makes robust thermal-state preparation and open-system quantum simulation at chemically realistic temperatures broadly accessible. Published in Physical Review Letters (2026).
Color centers in diamond — atomic-scale defects that emit single photons — are workhorses of quantum networking, but nearly all of them suffer from coupling to lattice vibrations: phonons broaden their emission lines and siphon brightness into useless sidebands.
Researchers at the University of Illinois Urbana-Champaign (led by graduate student Swetapadma Sahoo in Simeon Bogdanov’s group, with Oak Ridge National Laboratory, UCLA and international collaborators) have discovered a new diamond color center, named IL1 after the university. IL1 emits exceptionally bright, ultranarrow-band single photons while remaining remarkably insensitive to the crystal vibrations typical of the diamond lattice — a “strong broadband phonon decoupling” that sidesteps the central weakness of established emitters. Such a vibration-immune, spectrally pure single-photon source is a promising building block for quantum communication and quantum networks. Published in Nature Communications (press release 31 July 2026).
Rutile RuO2 was once hailed as the flagship candidate altermagnet — a material with momentum-dependent spin splitting but no net magnetization — until mounting evidence (neutron diffraction, muon spin rotation, ARPES) indicated that bulk and thick-film RuO2 is simply nonmagnetic. The ultrathin limit, where substrate-imposed strain is strongest, had remained unexplored.
A team from Rice University, the University of Minnesota and the Paul Scherrer Institute grew atomically smooth, fully strained ~2-nm epitaxial RuO2 films on titanium-dioxide substrates by hybrid MBE and probed them with spin-resolved angle-resolved photoemission spectroscopy (spin-ARPES), supported by ab initio calculations. They observed an unconventional spin texture in which mirror-even and mirror-odd momentum-dependent components coexist; using two measurement geometries and a comprehensive symmetry analysis, they ruled out nonmagnetic origins and known artifacts. The findings point to an emergent nonrelativistic spin structure enabled by epitaxial strain in the ultrathin limit — a distinct departure from nonmagnetic bulk RuO2 — suggesting strain as a control knob for magnetic states in oxide films, with implications for next-generation spintronics and the ongoing altermagnetism debate. Published in Science Advances.
In quantum chromodynamics (QCD), the strong force that binds quarks inside protons and neutrons is carried by gluons. Unlike photons, gluons attract one another, so QCD predicts they can bind into an entirely new kind of particle made purely of force carriers — a glueball — with no matter particles at all. Despite nearly 50 years of searching, no glueball had been unambiguously established.
At a special plenary session of the International Conference on High Energy Physics (ICHEP 2026) in Natal, Brazil, the BESIII Collaboration reported that, after 15 years of sustained work at the Beijing Electron–Positron Collider (BEPCII), the dominant constituent of the particle X(2370) is a pseudoscalar glueball with spin-parity 0⁻⁺. X(2370) was first seen in 2011 in radiative J/ψ decays; in 2024, using a sample of 10 billion J/ψ events, its spin-parity was fixed at 0⁻⁺, matching lattice-QCD predictions for the lightest pseudoscalar glueball. Recently the team observed new decay modes and, crucially, established the state's flavor-singlet nature — the hallmark of a glueball, since gluons couple democratically to all quark flavors. Together these results form a complete experimental chain: a pseudoscalar-glueball component must dominate X(2370). It is the clearest result yet from nearly half a century of glueball hunting, and a decisive low-energy validation of QCD. Formally presented at an ICHEP plenary session on 5 August 2026 (Brazil local time).
Quantum electrodynamics (QED) predicts that a vacuum is not truly empty: it teems with short-lived virtual particle–antiparticle pairs. In an extreme magnetic field these pairs give the vacuum polarization-dependent refractive indices, so empty space acts like a birefringent optical filter. This effect, vacuum birefringence, was proposed by Heisenberg and Euler in 1936 but had never been confirmed observationally, because the critical field (~4.4 × 10¹³ G) is far beyond any laboratory magnet.
A team led by Rachael E. Stewart reports the strongest astrophysical evidence yet, from the radio-emitting magnetar 1E 1547.0−5408 (surface field exceeding 10¹⁴ G). By coordinating X-ray polarimetry from NASA's IXPE with the NICER X-ray timer and the Parkes/Murriyang radio telescope, they obtained phase- and energy-resolved measurements and found a large linear polarization degree — reaching a phase-averaged ~65% at 2 keV in the thermally dominated soft band — that then drops between 2 and 4 keV, near where QED predicts a vacuum resonance. Surface-emission models alone cannot reproduce such a high polarization; vacuum birefringence in the magnetosphere provides a natural explanation, offering an important observational test of QED in extreme fields. Published in Nature (5 August 2026).
Rachael E. Stewartらは、電波を放つマグネター1E 1547.0−5408(表面磁場は10¹⁴ガウス超)から、これまでで最強の天体観測的証拠を報告した。NASAのX線偏光観測衛星IXPE、X線タイミング観測装置NICER、Parkes/Murriyang電波望遠鏡を協調させ、位相・エネルギー分解の測定を実施。熱放射が支配する軟X線帯で位相平均約65%(2keV)という大きな直線偏光度を検出し、2〜4keVではQEDが真空共鳴を予言する領域付近で偏光度が低下することを捉えた。これほど高い偏光は星表面の放射モデルだけでは再現できず、磁気圏中の真空複屈折が自然な説明を与える。強磁場下でのQEDを検証する重要な観測的手がかりである。Natureに掲載(2026年8月5日)。
The Kelvin–Helmholtz instability (KHI) arises wherever two fluids slide past each other at different speeds, curling the shear boundary into breaking-wave–like vortices. It is seen across nature — from ocean waves and clouds to the atmospheres of Jupiter and Saturn — and had long been predicted on the Sun, but never directly resolved on its visible surface.
Using the NSF Daniel K. Inouye Solar Telescope (4-m aperture, Haleakalā, Maui), an international team from the National Solar Observatory (NSO), the NSF NCAR High Altitude Observatory (HAO) and the Max Planck Institute for Solar System Research (MPS) captured the highest-resolution images of the solar photosphere ever taken, resolving structures only ~20 km across (at 416 nm). At the edges of magnetic flux concentrations they found ubiquitous magnetized Kelvin–Helmholtz vortices — the first experimental confirmation of KHI in the photosphere — corroborated by high-resolution numerical simulations. The instability can drive small-scale plasma mixing and rapid magnetic diffusion, and may help explain how the corona is heated and how magnetic energy builds up to power flares and eruptions. Led by David Kuridze; published in Nature (5 August 2026).
Entangled photons underpin quantum communication, sensing and computing, but they are almost always produced with energy-hungry, highly coherent lasers. It was widely assumed that phase-stable, laser-like light is required to drive the nonlinear process that creates entanglement — leaving many to dismiss ordinary sunlight, which is spatially and temporally incoherent, as unusable.
A collaboration between the University of Ottawa and the Max Planck Institute for the Science of Light (MPL), building on earlier theory and LED experiments from R. W. Boyd's group, has now generated quantum entanglement directly from sunlight. Using a cone-shaped, all-glass solar concentrator to focus light gathered over ~1.4 m² onto a nonlinear crystal, they drove spontaneous parametric down-conversion (SPDC) with polarized but incoherent sunlight. The resulting photon pairs reached a fidelity of ~94% to an ideal entangled state and violated Bell's inequality, confirming genuine entanglement; when normalized to pump power and bandwidth, the efficiency was on par with laser-driven sources. Because it removes the electrical-to-optical conversion (and its waste heat), the approach points toward simpler, more energy-efficient quantum light sources — attractive for satellites and deep-space missions. Published in Optica (press releases 6 August 2026).
オタワ大学とマックス・プランク光科学研究所(MPL)の共同チームは、R. W. Boyd研の従来理論とLED実験を土台に、太陽光から直接量子もつれを生成することに成功した。円錐型の全ガラス製ソーラー集光器で約1.4m²分の光を非線形結晶に集め、偏光させたインコヒーレントな太陽光で自発パラメトリック下方変換(SPDC)を駆動。得られた光子対は理想的なもつれ状態に対して忠実度約94%に達し、ベル不等式も破って真のもつれであることを確認した。ポンプ出力と帯域で規格化すると、その効率はレーザー駆動と同等だった。電気→光の変換(とその廃熱)を不要にするため、より単純で省エネな量子光源への道を開く——衛星や深宇宙ミッションに魅力的だ。Opticaに掲載(プレス発表は2026年8月6日)。
Many of the most interesting quantum materials host multiple electronic phases at once — magnetism, superconductivity, charge order — and a leading idea is that how these phases interact governs their exotic properties. Disentangling coexisting phases, and understanding how each one forms, is therefore a central problem.
A team led by Nuh Gedik at MIT (co-led by Alfred Zong, now at Stanford; first authors Yifan Su and Bai-Qing Lv) studied the rare-earth tritelluride ErTe₃, which hosts two charge density waves (CDWs): a “dominant” wave along one direction and, at lower temperature, a “subdominant” wave perpendicular to it, forming a checkerboard. Cooling a sample far below both transitions, they used a two-pulse, “shake-then-listen” scheme — a first laser pulse melts the checkerboard, and a delayed high-energy probe pulse ejects electrons (time-resolved photoemission) to snapshot the recovery. The dominant CDW reformed gradually and uniformly — a textbook second-order transition. The subdominant CDW, unexpectedly, reformed like water crystallizing into ice: in isolated pockets that grow (nucleation-and-growth, a first-order transition). The result settles a long-debated mechanism and offers a template for untangling far more complex materials such as high-temperature superconductors. Published in Nature Physics (7 August 2026).
Elemental (single-element) superconductors are prized as clean model systems, but their transition temperatures (Tc) are generally low — the record is scandium at 36 K, and only under an extreme pressure of 260 GPa. Two-dimensional boron sheets (borophenes) show rich structural variety, yet their predicted Tc values (3.7–27.6 K) have been limited by relatively weak electron–phonon coupling.
In a first-principles (density-functional) study, Meng-hui Wang and colleagues computationally surveyed more than 9,000 stacked arrangements of bilayer borophene. The best structure — an AA-stacked ν1/7 configuration featuring direct boron–boron bonds between the two layers — reshapes the lattice vibrations and switches on a cooperative, dual-channel electron–phonon coupling, yielding a predicted superconducting Tc of about 68 K at ambient pressure. That nearly doubles the elemental record and effectively maps a new theoretical ceiling for what a pure two-layer element can achieve. This is a prediction from quantum-mechanical calculations, not a laboratory measurement; whether the required stacking and interlayer bonding can be synthesized reproducibly remains to be seen. Published in Physical Review Letters (4 August 2026).
Bose–Einstein condensates (BECs) in microgravity are a powerful tool for fundamental physics: in extended free fall, ultracold atoms float undisturbed, enabling long interrogation times for precision tests. But such experiments have been limited by how many atoms — and how fast — a compact, transportable apparatus can deliver, especially for two-species mixtures needed to compare how different atoms fall.
An international team reports the MAIUS-B apparatus, a fully integrated, sounding-rocket-class setup that generates dual-species BEC mixtures of ⁴¹K and ⁸⁷Rb (potassium and rubidium) at record atom flux. The team compared the release and free expansion of the mixtures on the ground and in free fall in the Einstein-Elevator at Leibniz University Hannover, and — using a dedicated trap switch-off protocol modeled to minimize residual magnetic fields — characterized the intra- and interspecies interactions governing the dynamics. A key enabler is a highly compact optical system built on Zerodur (an ultra-low-thermal-expansion glass-ceramic), developed with major contributions from Johannes Gutenberg University Mainz, that controls and cools the atoms without adding bulk. The result sets a new benchmark for ultracold mixtures on mobile platforms and lays groundwork for the German-American BECCAL laboratory on the ISS and for testing Einstein's equivalence principle — whether different atomic species fall with exactly the same acceleration. Coordinated by ZARM Bremen; published in Nature Communications (paper 28 July 2026; press 5 August 2026).
Quantum cryptography can stop votes being tampered with after they are cast — but it does not solve a deeper problem: if the central authority collecting the ballots is corrupt or compromised, the whole election is undermined, and earlier quantum protocols did not guarantee that voters stay anonymous to that authority.
Two independent teams have now implemented, with photons, a protocol proposed in 2022 by Federico Centrone and co-workers that removes the need to trust election administrators. The scheme distributes a Greenberger–Horne–Zeilinger (GHZ) entangled state with one qubit per voter: measuring it gives each voter a random 0 or 1, while entanglement fixes the parity of the total, so a single designated voter can flip the parity and have their vote counted with no link to their identity. Crucially, voters can verify that they really received a GHZ state, and each qubit can be used only once — for voting or for verification. F. Joseph Marcellino, Mingsong Wu and Rob Thew (University of Geneva) randomly assign each round to be either verification or voting, reaching an 87% verification success rate; Nicolas Laurent-Puig, Matilde Baroni, Federico Centrone and Eleni Diamanti (Sorbonne University and collaborators) use a more reliable GHZ source and reach about 96%, deferring the verification step to future work. Both produce GHZ states by spontaneous parametric down-conversion; the world record is 14 photons, so realistic use is likely to start with small, high-stakes settings such as boardroom votes rather than national elections. Published in Physical Review Letters 137, 060802 and 060803 (7 August 2026), with an APS Physics Focus story.
A proton is not simply three quarks. In quantum chromodynamics (QCD) its internal structure is described by distribution functions, and beyond the familiar quark densities there are quark–gluon–quark correlations — genuinely quantum interference terms in which a quark and a gluon combine. These are notoriously hard to extract from data and have usually been assumed to be small.
Alexey Vladimirov, Guillermo Portela and Simone Rodini present a proof-of-concept global QCD fit that determines the quark-gluon-quark interference contribution inside the proton for the first time. They find it is comparable in size to the polarized quark density contributions — not a small correction at all. The result implies the proton is a more strongly quantum-interfering object than standard treatments assume, and it bears on how the proton’s spin is apportioned among quarks, gluons and their orbital motion. Published as an Editors’ Suggestion in Physical Review Letters 137, 061902 (3 August 2026).
Nuclear fuel and its fission products emit antineutrinos, which stream out of a reactor building unimpeded. For two decades, nuclear safeguards agencies have explored whether this leakage could give an independent, tamper-proof read-out of what is actually inside a reactor. Until now, all such work studied reactors while they were running.
The Double Chooz Collaboration has now made the first quantitative measurement of the residual antineutrino flux from shutdown reactor cores and nearby spent-fuel pools at the Chooz nuclear power plant in France. Over 17.2 days in 2017 with both reactors offline, the detector recorded antineutrinos from beta decays of fission products accumulated in irradiated uranium-dioxide fuel, using inverse beta decay in liquid scintillator. Because the shutdown flux is roughly one hundredth that of an operating reactor, the measurement hinged on precise background modelling, tight control of systematics and state-of-the-art simulation. Since the spectrum’s intensity and shape depend on how much spent fuel is present, its isotopic makeup and how long it has been cooling, comparing a measured spectrum with simulation could reveal diverted material. Dedicated detectors and validation across different reactor and storage configurations are the next steps. Published in Physical Review Letters 137, 061803 (4 August 2026).
Molecules built around heavy radioactive nuclei are among the most sensitive probes of physics beyond the Standard Model. Radium is the star candidate: its pear-shaped nucleus and high atomic number amplify the effects of parity violation by more than three orders of magnitude compared with spherical nuclei, and embedding it in a molecule adds the enormous internal electric fields molecules provide. The catch has been supply — until now, radium-bearing molecules were made only in trace amounts at accelerator facilities such as CERN’s ISOLDE.
A team led by Nick Hutzler at Caltech has now done it on a laboratory bench. Tiny quantities of radium chloride or nitrate were mixed with water and the sweetener xylitol and heated to yield a gel-like radioactive target. Placed in a cryogenic cell filled with helium buffer gas alongside reagent pellets, the target was hit with laser pulses; the released precursors were collisionally cooled to about 4 K, and further laser light drove reactions producing RaF, RaOH and RaOD. At these temperatures the molecules occupy few quantum states, and successively narrower laser-induced-fluorescence scans located individual transitions at high resolution. Measured molecular properties largely matched predictions, though one discrepancy suggests theory does not yet fully capture the excited state. The group is now working with MIT and Harvard to extract the molecules into a beam for laser cooling and trapping — a route toward far more sensitive searches for the electron’s electric dipole moment. Published in Science 393, 319 (2026).
A stellarator confines fusion plasma entirely with external magnets, avoiding the disruptions that plague tokamaks — but that shifts the burden onto the coils, which must be twisted into shapes that are simultaneously good physics and buildable hardware. Coil design has therefore long been a bottleneck.
Pedro F. Gil and colleagues recast the problem as a highly nonconvex equality-constrained optimization and attack it with an augmented Lagrangian method, in which constraints are enforced through penalty terms plus explicit multipliers rather than being folded into a single weighted objective. Applied to modular coils surrounding the plasma in magnetic confinement fusion, the approach yields a class of stellarator coils with enhanced physics performance that also meets essential engineering design requirements — a combination that weighted-sum approaches typically have to trade off. Selected as an Editors’ Suggestion and featured on the journal cover, Physical Review Letters 137, 065101 (7 August 2026).
Pedro F. Gilらは今回、この問題を強く非凸な等式制約付き最適化問題として定式化し直し、拡張ラグランジュ法で解いた。制約をペナルティ項と明示的な乗数の両方で扱う手法で、すべてを重み付き和の単一目的関数に押し込める従来アプローチとは異なる。磁場閉じ込め核融合でプラズマを取り囲むモジュラーコイルに適用した結果、物理性能を高めながら、必須の工学設計要件も満たす新しいコイル群が得られた——重み付き和法では通常トレードオフになる組み合わせである。Editors' Suggestionに選出され、誌面の表紙も飾った。Physical Review Letters 137, 065101(2026年8月7日付)掲載。
In a rainbow, each wavelength of sunlight takes a different path. Rainbow trapping imports that idea into wave engineering: a designed material slows different frequency components of a broadband signal to a halt at different positions, sorting the signal in space. It has been hard to do for elastic waves — vibrations that deform the material as they travel — because most designs relied on modes confined to edges or interfaces, and offered no way to retrieve energy once localized.
Two teams now solve complementary halves of the problem, both using synthetic gauge fields: a pseudomagnetic field creates flat elastic Landau levels (near-zero group velocity, hence localization) while a pseudoelectric field shifts their frequency across the sample. Yafeng Chen (Tongji University) and colleagues patterned an aluminium plate so that two perpendicular geometric gradients control the two pseudofields independently; making the pseudomagnetic field nonuniform widened the flat momentum range. Sending in waves near 1.1 MHz and mapping the plate by laser scanning, they saw different frequencies stop at different positions — direct real-space evidence of a bulk elastic rainbow. Riyi Zheng (South China University of Technology) and colleagues built theirs on a silicon chip and added a pair of counter-propagating topological edge states: the upper edge state travels until it meets the bulk Landau mode of matching frequency, couples into it, crosses the sample and continues along the lower boundary — turning the rainbow into a frequency-dependent routing map, with no obvious backscattering from defects or disorder. Both in Physical Review Letters 137, 066601 and 066602 (3 August 2026), with a Physics Viewpoint.
Arrays of neutral atoms held in optical tweezers — tightly focused laser beams, each gripping a single atom — have become a leading platform for quantum computing and quantum simulation. Alkaline-earth-like atoms such as ytterbium are especially prized because their two valence electrons give narrow clock transitions and metastable states useful for qubit encoding and for optical clocks. The practical limit has been how many tweezers can be reliably filled at once.
Jiawen Zhu and colleagues report the stable loading of 2400 ytterbium-174 atoms into an optical tweezer array — the largest alkaline-earth-like atom array to date. The loading technique is not specific to ytterbium and is applicable to other atomic species, making it a general route to scaling up neutral-atom platforms. Selected as an Editors’ Suggestion, Physical Review Letters 137, 063201 (4 August 2026).
1T-TaS2 is a textbook playground for correlated electrons: it hosts a charge density wave (CDW) — a periodic modulation of electron density locked to a lattice distortion — that in its commensurate phase drives the material into a correlated insulating state often discussed as a Mott insulator and as a candidate quantum spin liquid. How that physics survives, or changes, when the crystal is thinned toward a single layer has been contested.
Gan Liu and colleagues combine Raman spectroscopy with transport measurements across a thickness series and find that approaching the two-dimensional limit markedly enhances both the CDW order and the associated correlated insulating state. Rather than being suppressed by reduced screening or interlayer decoupling, the ordered state is strengthened — a clean demonstration that dimensionality itself is a control knob for strongly correlated phases in van der Waals materials. Selected as an Editors’ Suggestion, Physical Review Letters 137, 066502 (5 August 2026).
Altermagnets — the third class of collinear magnetic order, alongside ferromagnets and antiferromagnets — have zero net magnetization yet show spin-split electronic bands without relativistic spin-orbit coupling. In the altermagnets identified so far, that splitting is an even function of momentum, set by crystal rotations relating the two spin sublattices. An odd-parity counterpart, in which the splitting reverses with momentum, would enable different responses again — but how to realize it has been unclear.
Yu-Ping Lin and Marc Vila show that sublattice currents — loop currents circulating within the sublattice structure — provide a feasible route. Starting from the Haldane-Hubbard model, in which complex next-nearest-neighbour hoppings encode exactly such currents, they generalize the mechanism to arbitrary bipartite lattices, obtaining nonrelativistic collinear spin splitting that is an odd function of momentum. The work turns odd-parity altermagnetism from a symmetry classification into a concrete design principle for materials and cold-atom simulators. Selected as an Editors’ Suggestion, Physical Review Letters 137, 066702 (4 August 2026).
Yu-Ping Lin と Marc Vilaは、サブラティス・カレント——副格子構造の内部を循環するループ電流——が実現可能な道筋になることを示した。出発点はハルデン・ハバード模型で、この模型の複素次近接ホッピングはまさにそうした電流を表現している。両氏はこの機構を一般の二部格子へ拡張し、運動量の奇関数となる非相対論的な共線スピン分裂が得られることを示した。奇パリティ・アルターマグネティズムを、対称性による分類から一歩進めて、物質設計や冷却原子シミュレータのための具体的な設計原理へと変える成果である。Editors' Suggestion選出、Physical Review Letters 137, 066702(2026年8月4日付)掲載。
Some frustrated magnets never order, even at absolute zero. Instead they settle into a quantum spin liquid whose low-energy physics looks like an emergent electromagnetism: a U(1) Coulomb liquid with its own photon and its own charges. A more exotic possibility is a rank-2 version, where the emergent gauge field is a tensor rather than a vector and its charges are fractons — excitations with restricted mobility.
Lasse Gresista and colleagues construct a minimal, materials-relevant spin-1/2 model on the breathing pyrochlore lattice — a pyrochlore network in which up- and down-pointing tetrahedra differ in size, as realized in several real compounds. Tuning the model solely by symmetry-allowed Dzyaloshinskii–Moriya interactions, they find robust quantum realizations of both rank-1 and rank-2 U(1) Coulomb liquids in the same setting. That a single, chemically plausible lattice can host both, controlled by an interaction that is present anyway, makes higher-rank spin liquids a much more realistic experimental target. Selected as an Editors’ Suggestion, Physical Review Letters 137, 066504 (4 August 2026).
Topology in physics describes properties that survive continuous deformation. In non-Hermitian systems — those with gain or loss, such as lasers — energies become complex numbers, and as parameters are driven in a loop around an exceptional point the eigenvalues trace out braids, forming links and knots of arbitrary complexity. Controlling and directly seeing that process has been the hard part.
Wenbo Mao, Bofeng Zhu, Y. D. Chong, Lan Yang and colleagues (Washington University in St. Louis / Nanyang Technological University, Singapore) demonstrate non-Hermitian braiding of laser modes on an integrated photonic chip. By actively steering the parametric trajectories of gain and detuning, they watch the braiding unfold directly in the evolution of the lasers’ frequencies and intensities — no reconstruction required. The result is a rich menagerie of topological structures including Hopf links, trefoil knots and Solomon links, generated programmably on a chip. Because the platform is a pair of coupled chip-scale lasers rather than a bespoke apparatus, it turns eigenvalue braiding from a theoretical classification into a tunable device function for light manipulation. Published online in Nature Physics (12 May 2026); brought back into the spotlight by a News & Views, “Braids of light,” by König and Bergholtz on 7 August 2026.
In isolated quantum many-body systems, temperature and entanglement usually travel together: hot states are extensively entangled, while low-temperature states have short-ranged correlations and modest entanglement. That “area law” simplicity is what makes ground states tractable for methods like DMRG and tensor networks. Exceptions are therefore precious — and rare.
Matteo Ippoliti (University of Texas at Austin) and David M. Long (Stanford) construct a family of static, geometrically local Hamiltonians whose eigenstates inherit the properties of periodically driven (Floquet) systems. Their tool is a variation on the Feynman–Kitaev clock — the standard mapping from quantum circuits to local Hamiltonians — with the clock register given periodic boundary conditions. If the input circuit obeys the eigenstate thermalization hypothesis (ETH), every eigenstate of the resulting Hamiltonian, including the ground state, acquires infinite-temperature characteristics such as volume-law entanglement entropy. They then build exactly solvable Floquet circuits — drawing on ideas from classical pseudorandom number generation — that provably satisfy ETH at infinite temperature. Combining the two yields local Hamiltonians with provably volume-law entangled ground states, and the first construction where the volume law holds for all contiguous subsystems. Since driven systems cannot conserve energy and can only equilibrate to infinite temperature, these models exhibit an exotic coexistence of infinite-temperature and zero-energy physics. Open access in Physical Review X 16, 031030 (7 August 2026).
Matteo Ippoliti(テキサス大学オースティン校)とDavid M. Long(スタンフォード大学)は、周期駆動(フロケ)系の固有状態の性質を受け継ぐ、静的で幾何学的に局所なハミルトニアンの族を構築した。道具はファインマン・カイタエフ時計模型——量子回路を局所ハミルトニアンへ写す標準的な対応——の変形版で、時計レジスタに周期境界条件を課す点が新しい。入力回路が固有状態熱化仮説(ETH)を満たすなら、得られるハミルトニアンの基底状態を含む全固有状態が、体積則のもつれエントロピーという無限温度的な性質を獲得する。さらに両氏は、古典的な擬似乱数生成の発想を借りて、無限温度でETHを満たすことが厳密に示せるフロケ量子回路の族も構成した。この2つを組み合わせることで、体積則もつれを持つ基底状態が証明できる局所ハミルトニアンが得られ、しかもすべての連続部分系について体積則が成立する初の構成となった。駆動系はエネルギーを保存できず「無限温度」にしか熱平衡化できないため、これらの模型では無限温度の物理とゼロエネルギーの物理が奇妙に共存する。Physical Review X 16, 031030(2026年8月7日付)にオープンアクセスで掲載。
Since the discovery of superconductivity in nickelates, a central question has been how closely they parallel the cuprate high-temperature superconductors. Low-valence nickelates such as Pr4Ni3O8 — a trilayer member of the square-planar nickelate family — are prized because their nickel valence and orbital occupancy sit close to the cuprate case, making them a sharp test of the analogy.
Y. Shen and colleagues use resonant inelastic x-ray scattering (RIXS), which probes charge and spin excitations with momentum resolution, to reveal propagating plasmons — collective oscillations of the conduction-electron density — in Pr4Ni3O8. Crucially, these plasmons show dynamics distinct from those of cuprate superconductors. Because plasmon dispersion encodes the effective long-range Coulomb interaction and the degree of interlayer coupling, the difference is a direct handle on how correlations in nickelates depart from the cuprate template. Open access in Physical Review X 16, 031031 (7 August 2026).
Y. Shenらは、電荷・スピン励起を運動量分解で捉えられる共鳴非弾性X線散乱(RIXS)を用い、Pr₄Ni₃O₈の中を伝播するプラズモン——伝導電子密度の集団振動——を観測した。重要なのは、このプラズモンが銅酸化物超伝導体とは明確に異なるダイナミクスを示した点である。プラズモンの分散関係には実効的な長距離クーロン相互作用と層間結合の強さが刻まれているため、この違いはニッケル酸化物の電子相関が銅酸化物の枠組みからどうずれているかを直接測る手がかりになる。Physical Review X 16, 031031(2026年8月7日付)にオープンアクセスで掲載。
Symmetry does not merely classify states — it restricts what you are allowed to do to them. The resource theory of asymmetry formalizes this: when your operations must respect a symmetry group, states that break that symmetry become a consumable resource, needed for tasks such as reference-frame alignment, quantum metrology and covariant error correction. The practical question is the conversion rate: given many copies of one asymmetric state, how many copies of another can you produce?
Koji Yamaguchi, Yosuke Mitsuhashi, Tomohiro Shitara and Hiroyasu Tajima prove that the quantum geometric tensor — the object combining the quantum Fisher information metric with the Berry curvature — completely dictates the pure-state asymptotic (i.i.d.) conversion rate under any compact Lie group symmetry. Previous results were largely confined to specific groups such as U(1); this establishes a single geometric quantity as the answer in full generality. Because the quantum geometric tensor is already a workhorse of quantum metrology and band theory, the result ties the limits of symmetric quantum operations directly to a familiar and computable object. Open access in Physical Review X 16, 031028 (5 August 2026).
Koji Yamaguchi・Yosuke Mitsuhashi・Tomohiro Shitara・Hiroyasu Tajimaの4氏は、量子幾何テンソル——量子フィッシャー情報計量とベリー曲率を束ねた量——が、任意のコンパクト・リー群対称性のもとで純粋状態の漸近的(i.i.d.)変換レートを完全に決定することを証明した。従来の結果はU(1)など特定の群に限られていたが、本研究は単一の幾何学量が完全な一般性のもとで答えを与えることを確立した。量子幾何テンソルは量子計測やバンド理論ですでに定番の道具であるため、対称性に縛られた量子操作の限界が、なじみ深く計算もできる対象に直結したことになる。Physical Review X 16, 031028(2026年8月5日付)にオープンアクセスで掲載。
Chemistry is usually driven by heat, light or electricity. Mechanochemistry instead uses mechanical force to push reactions along — and at the nanoscale, a sharp probe tip can concentrate a modest applied load into an enormous local stress.
Shuai Zhang and colleagues exploit exactly this. Working in the encapsulated space between two-dimensional materials, they use a small tip to generate a large stress that drives a reaction between two materials, producing a new, atomically thin metal with interesting electronic properties. Because the reaction happens only where the tip presses, the product can be written into patterns as small as 50 nm — direct-write synthesis rather than deposit-then-etch lithography. The approach points toward building conducting circuitry inside van der Waals heterostructures at the point of use. Open access in Physical Review X 16, 031029 (6 August 2026).
Entangling matter at long distance — not just photons — is the hard part of building a quantum internet. Quantum memories must be linked over fibre, and every kilometre of glass eats photons, so entanglement rates collapse exponentially with distance. There is also a hard benchmark to beat: the repeaterless bound, the maximum rate at which entanglement can be sent directly down a lossy channel without any repeater.
A team at the University of Science and Technology of China now reports entanglement between two atomic-ensemble quantum memories separated by 420 km of fibre. They use the DLCZ (Duan–Lukin–Cirac–Zoller) scheme and convert the memory-emitted photons into the telecom S band to exploit the lowest available fibre loss. Holding the optical phase steady over that distance required a two-part stabilization strategy: full-time far-off-resonant locking to suppress high-frequency noise, plus intermittent dual-band locking to cancel slow drift. Crucially, the memory–memory entangling probability beats the repeaterless channel capacity for direct entanglement distribution — the regime in which a quantum repeater actually pays for itself. Published in Physical Review Letters, 11 August 2026 (Editors’ Suggestion).
Cold dark matter (CDM) predicts a halo mass function that continues down to extremely small masses. Warm, fuzzy or interacting dark matter would instead cut the function off at some scale. Finding where — or whether — that cutoff sits is one of the sharpest available tests of dark matter’s particle nature, but the smallest halos contain no stars, so they are invisible to conventional surveys.
A team led by A. M. Nierenberg, D. Gilman and T. Treu uses 28 strong gravitational lenses to sidestep that problem: lensing responds to mass alone, so dark halos both in the lens galaxy and along the line of sight leave measurable imprints regardless of whether they host baryons. Modelling an extreme scenario in which the mass function and mass–concentration relation follow CDM but terminate sharply at some mlow — and accounting for tidal stripping, which populates masses below the nominal cutoff — they place an upper limit of mlow < 108.3 solar masses at 10:1 odds using a subhalo-mass-function prior from the semianalytic model galacticus, and 108.2 using an N-body prior. These bounds are comparable to or stronger than existing Milky Way satellite constraints, and the authors forecast more than an order-of-magnitude improvement from a sample of 200 quadruply imaged quasars — a small fraction of what Rubin, Euclid and Roman are expected to deliver. Published in Physical Review Letters, 11 August 2026.
Keywords: dark matter, 暗黒物質, cold dark matter, 冷たい暗黒物質, CDM, halo mass function, ハロー質量関数, strong gravitational lensing, 強重力レンズ, quadruply imaged quasar
🔦 / 水素のHα輝線で暗黒物質を探す新手法——静かなガス豊富な矮小銀河Leo TのMUSE観測から、対消滅・崩壊に対する世界初のHα由来制限を導出。eV〜GeV質量域の一部で最高感度(Rebecca K. Leane、Phys. Rev. Lett.掲載)
Indirect dark-matter searches usually look for gamma rays, X-rays or antiparticles. Rebecca K. Leane proposes an entirely different messenger: the Hα recombination line of hydrogen at 656 nm, the same red glow that lights up nebulae in amateur astrophotographs.
The logic is that annihilation or decay products ionize neutral gas; when the freed electrons recombine, the cascade produces Hα photons via the n = 3 → 2 transition. In quiet, gas-rich dwarf galaxies the n = 2 population is negligible, so Hα escapes essentially unabsorbed and traces exactly where energy was injected. Using the non-detection of extended Hα emission in the dwarf galaxy Leo T with Multi Unit Spectroscopic Explorer (MUSE) observations, Leane derives the first Hα-based limits on dark-matter annihilation and decay, reaching leading sensitivity over parts of the eV–GeV mass range — a window that gamma-ray telescopes handle poorly. Because optical spectrographs are abundant and improving, the work establishes Hα imaging as a genuinely new and extensible dark-matter search strategy. Published in Physical Review Letters, 11 August 2026.
暗黒物質の間接探索といえば普通はガンマ線・X線・反粒子を狙う。Rebecca K. Leaneはまったく別の使者を提案した。波長656nmの水素のHα再結合線——アマチュア天体写真で星雲を赤く輝かせる、あの光である。
論理はこうだ。暗黒物質の対消滅・崩壊で生じた粒子が中性ガスを電離し、解放された電子が再結合するとき、そのカスケードが n = 3 → 2 遷移によってHα光子を生む。静かでガスの豊富な矮小銀河では n = 2 の準位に居る原子がほとんどないため、Hαは事実上吸収されずに脱出し、エネルギーが注入された場所をそのまま描き出す。Leaneは矮小銀河Leo Tにおける広がったHα放射の「非検出」(MUSE分光装置による観測)を用い、暗黒物質の対消滅・崩壊に対する世界初のHα由来の制限を導出。eV〜GeV質量域の一部で最高感度に到達した——ガンマ線望遠鏡が苦手とする窓である。光学分光器は数が多く性能も向上し続けているため、この研究はHα撮像を新しく拡張性のある暗黒物質探索戦略として確立した。Physical Review Letters 2026年8月11日付掲載。
About 58% of Higgs bosons decay to a bottom–antibottom quark pair, yet that channel is notoriously hard to see at a hadron collider because ordinary QCD produces overwhelming numbers of b jets. The high transverse-momentum regime is especially interesting: new physics coupling to the Higgs would show up first as a deviation in the pT spectrum’s tail.
The ATLAS Collaboration now reports the first evidence for inclusive high-pT Higgs production in the bb̄ final state, with the boosted Higgs reconstructed as a single large-radius jet. The analysis uses 13 and 13.6 TeV proton–proton data totalling 301 fb−1, and leans on two machine-learning advances: a new transformer-based jet-flavour tagging algorithm for background suppression, and a dedicated regression model that sharpens the jet mass and pT resolution. For Higgs bosons produced above pT = 450 GeV, the measured yield relative to the Standard Model prediction is 1.53 (with statistical, experimental and theoretical uncertainties of roughly ±0.27, +0.33/−0.27 and ±0.17), an observed significance of 3.8σ against an expected 2.5σ. Results in three pT intervals are compatible with the Standard Model. Published in Physical Review Letters, 11 August 2026.
ヒッグス粒子の約58%はボトムクォークと反ボトムクォークの対に崩壊する。ところがハドロン衝突型加速器では、通常のQCD過程が圧倒的な数の b ジェットを作るため、この崩壊モードの観測は極めて難しい。とりわけ興味深いのが高い横運動量の領域だ。ヒッグスに結合する新物理があれば、まず pT スペクトルの裾に食い違いとして現れるからである。
How small can a nuclear collision be and still form quark–gluon plasma? The 2025 light-ion run at the LHC — oxygen–oxygen and neon–neon — was designed to answer exactly that, sitting between proton collisions (traditionally assumed too small) and lead collisions (definitively hot enough).
The LHCb Collaboration measures the ratio of D0 meson production between NeNe and OO collisions at a centre-of-mass energy per nucleon pair of 5.36 TeV, differentially in transverse momentum from 0.5 to 20 GeV and in the forward rapidity range 2.0 < y < 4.5, normalized to the number of recorded inelastic collisions in each sample. The production ratio shows evidence of variation with pT — a pattern inconsistent with predictions based on nuclear modification of nucleon structure alone. Instead it matches what one expects if quark–gluon plasma production switches on progressively as the colliding nuclei get larger. Because charm quarks are produced early and then traverse whatever medium forms, they are a particularly clean thermometer. Published in Physical Review Letters, 11 August 2026.
Two-neutrino double-beta decay (2νββ) is the rarest process ever directly observed, and it is the irreducible background for the hunt for its neutrinoless cousin — the decay that would prove neutrinos are their own antiparticles. Extracting physics from 2νββ spectra therefore demands theory of matching precision. In ordinary single-beta decay, that precision comes from the Sirlin function, a universal radiative-correction factor independent of nuclear structure. No such object existed for double-beta decay.
Jordy de Vries, Emanuele Mereghetti, Saad el Morabit and Stefan Sandner supply it. Using heavy-nucleus effective field theory, they derive the first “double-weak Sirlin function” — a universal correction that depends on the two individual electron energies and on their relative angle, and which differs significantly from the naive approximation of simply adding two single-beta Sirlin functions. They also compute the nuclear-structure-dependent piece and find it still negligible at present sensitivities. The practical consequence is sharp: the new correction distorts the electron energy and angular spectra by an amount comparable to the leading nuclear-structure correction parametrized by the matrix-element ratio ξ31, so recent extractions of ξ31 should be revisited. Published in Physical Review Letters, 11 August 2026.
In a thin superconducting film, magnetic flux enters as a Pearl vortex — a two-dimensional cousin of the Abrikosov vortex, with a much longer-ranged current pattern. Vortices dictate dissipation and critical current, but they are normally studied only in aggregate; watching a single one move has been out of reach.
A group at ETH Zurich (Marta Perego and colleagues, with Thomas Ihn and Klaus Ensslin) turns a gate-defined Josephson junction into a single-vortex sensor in magic-angle twisted bilayer graphene, giving direct access to individual vortex entry and exit events. Above about 100 mK the vortices cross their energy barriers by classical thermal activation. Below about 90 mK, the team observes macroscopic quantum tunnelling through those same barriers — a whole vortex, a collective object of many electrons, tunnelling as one. The data are consistent with a sharp, first-order-type quantum-to-classical transition. From the measurements they extract entry and exit barriers of a few kelvin and a barrier thickness of roughly 100 nm. Published in Physical Review Letters, 11 August 2026 (Editors’ Suggestion).
📡 / 触らずに超流動剛性を測る——オンチップ超伝導マイクロ波共振器で原子層superconductor 4Hb-TaS₂の位相剛性を非接触測定。鏡映対称性が破れているのに「ノードなし」で、表面ノーダル超伝導説を否定(Chistolini・Feng Wang・Siddiqiら、UCバークレー、Phys. Rev. Lett.掲載・Featured in Physics)
Superfluid phase stiffness — how much energy it costs to twist the superconducting phase — is one of the most fundamental properties of a superconductor, and it encodes the gap structure. In two-dimensional van der Waals materials it has been almost unmeasurable, because the flakes are mesoscopic and conventional probes need contacts.
A Berkeley-led team (Trevor Chistolini and colleagues, with Feng Wang, Irfan Siddiqi and James Analytis) introduces a contact-free technique: place the flake near an on-chip superconducting microwave resonator and read the electrodynamic response, and hence the stiffness, from the resonator. They apply it to 4Hb–TaS2, a van der Waals superconductor whose gap structure under broken mirror symmetry has been actively debated. Their cleanest few-layer device retains a critical temperature comparable to the bulk, and the temperature evolution of the phase stiffness is nodeless even though mirror symmetry is broken — inconsistent with the proposed nodal surface superconductivity. Because the method needs minimal fabrication, it should open microwave measurements across a wide range of atomically thin superconductors. Published in Physical Review Letters, 11 August 2026; Featured in Physics and an Editors’ Suggestion.
On 12 August 2026 the Moon’s shadow crossed the Arctic Ocean and Arctic Russia, Greenland, Iceland, a small corner of northeastern Portugal, northern Spain and the Balearic Islands, delivering mainland Europe’s first totality since 1999 — Iceland’s first since 1954, and Reykjavík’s first in nearly six centuries. The umbra was roughly 290 km across. Totality lasted at most 2 minutes 18 seconds, over the ocean west of Látrabjarg in Iceland; in Spain it reached at most about 1 minute 50 seconds and fell just before sunset with the Sun very low in the sky.
Beyond the spectacle, a total eclipse remains the only way to see the inner corona in white light from the ground, and this one was heavily instrumented. The NASA-supported Nationwide Eclipse Ballooning Project launched roughly 80 scientific balloons in Iceland and six in Spain to measure how the shadow’s sudden cold and darkness perturb the atmosphere. A NASA-funded team chased totality in a WB-57 high-altitude jet, whose nose-cone cameras imaged the corona at least 20 times per second and, at 460 mph, stretched the observing window to nearly three minutes. Meanwhile ESA and modellers at KU Leuven published advance predictions of the coronal structure — digital reconstructions of the Sun’s magnetic environment driven by near-real-time surface magnetograms — so that the observed corona could be used to grade, and improve, the models behind space-weather forecasting. With the sunspot number near 102, the corona was expected to look intermediate between the flattened solar-minimum shape and the fully structured solar-maximum display.
Twisted bilayer graphene is famous for what happens near the magic angle of about 1.1°. Large twist angles have been assumed to be the boring end of the story: interlayer coupling negligible, no band reconstruction, effectively two independent graphene sheets stacked by accident.
A team at National Cheng Kung University in Tainan (Kuo-En Chang, Ming-Hao Liu and Tse-Ming Chen), with National Taiwan University and Japan’s NIMS, shows that this paradigm breaks down near a twist angle of 29°, where the system sits at the crossover between quasicrystalline and commensurate order. Atomic-resolution transmission electron microscopy directly reveals the coexistence of near-dodecagonal (12-fold) quasicrystalline symmetry with an emerging moiré periodicity — an intermediate, genuinely nonperiodic structural regime. Magnetotransport then uncovers strong interlayer hybridization mediated by Umklapp scattering, visible as magneto-intersubband oscillations and a highly unconventional Landau-level spectrum. Most striking, the Landau-level degeneracy evolves from fourfold to twelvefold as temperature rises — behaviour flatly incompatible with two decoupled monolayers. Large-angle twisted graphene thus becomes a platform where quasiperiodic symmetry reshapes low-energy electronic states outside the usual moiré framework. Published in Physical Review Letters, 11 August 2026 (Editors’ Suggestion).
The Edelstein effect — an in-plane current generating a net spin polarization — is a workhorse of spintronics, but it requires broken inversion symmetry and produces a single, global magnetization. A team led by Binchang Zhou, Pan Zhou and Lizhong Sun introduces its layer-resolved sibling.
The layer Edelstein effect (LEE) is defined by layer-resolved spin magnetizations with opposite components on the top and bottom layers of a bilayer, driven by an in-plane charge current and tunable by an external electric field. It is the real-space counterpart of the layer Hall effect. Working from a minimal bilayer k·p theory and a general stacking framework, the authors derive a model-independent symmetry criterion showing the LEE is generically allowed in a broad class of nonmagnetic bilayer stackings — no magnetism required. It comes in two flavours: components mandated outright by symmetry, and components switched on when an electric field lowers the symmetry. First-principles calculations on stacked bilayer MoSSe, MoTe2 and WTe2 confirm the prediction and suggest it is experimentally within reach, offering a unified route to electrically generating and manipulating layer-resolved spin polarization. Published in Physical Review Letters, 11 August 2026.
Magnetic skyrmions are nanoscale swirls of spin with two dynamical degrees of freedom: the centre of mass, and the helicity — the internal twist angle. In conventional ferromagnets these are locked together, so the helicity is fixed once the skyrmion Hall angle is known.
Yang Liu, Zhejunyu Jin, Jie Liu and Peng Yan show that in frustrated d-wave altermagnets — the newly recognized third class of magnetic order, with nonrelativistic momentum-dependent spin splitting — that lock is broken. Under a driving current, the skyrmion helicity is not tied to the Hall angle; instead it rotates unidirectionally with a global angular velocity proportional to the square of the current density. The rotation rate is strongly anisotropic, depending on the direction the current flows. They further find helicity oscillation in the terahertz regime, where nonlinear mixing between fast and slow modes generates a comblike spectrum — a magnon frequency comb. Full atomistic spin-dynamics simulations confirm the analytics. Frustrated altermagnets thus become a promising platform for skyrmionics and terahertz technology. Published in Physical Review Letters, 11 August 2026.
The electronic heat capacity of a bulk metal or semiconductor is a textbook quantity. In nanoparticles it deviates strongly from bulk behaviour, which matters for laser processing, thermal management and nanoscale energy conversion — but pinning down why requires resolving contributions atom by atom, which is computationally brutal.
A. Aryanpour and Ali Sadeghi combine ab initio calculations with a machine-learned local density of states, exploiting the extensivity of heat capacity to build a framework that resolves individual atomic contributions in silicon nanoparticles from 1 to 100 nm. The result is a clean physical picture: the atomic heat capacity increases by up to an order of magnitude for undercoordinated surface atoms, and a surface layer only 3–4 Ångström thick — whose thickness is independent of structural order and cluster size — carries a distinct electronic density of states and comes to dominate the total electronic heat capacity as the particle shrinks. That same surface layer governs the nonlinear temperature dependence. The work bridges atomic-scale electronic structure and nanoscale thermal behaviour; extending it to vibrational contributions remains open. Published in Physical Review Letters, 11 August 2026.
Why do some two-note intervals sound consonant and others dissonant? The question has occupied physicists and music theorists since Pythagoras, and explanations have ranged from simple frequency ratios to beating between overtones to learned cultural convention. Modern accounts tend to place the effect in the cortex.
Florian Gomez and Ruedi Stoop argue that a large part of it happens much earlier — in the ear itself. Modelling the network of activated cochlear amplifiers as coupled nonlinear oscillators, they show that nonlinear sound processing produces a close correlate of the psychoacoustic consonance–dissonance judgement, matching behavioural data as well as the best phenomenological approach. The mechanism they identify is combination-tone avalanches triggered from the coupled oscillators, paired with biological parsimony in how much information the system bothers to encode. The broader claim is methodological: behavioural phenomena routinely attributed to “the cortex” may in fact originate in the nonlinear physics of the sensors themselves. Published in Physical Review Letters, 11 August 2026.
The dynamical structure factor is what neutron and light scattering experiments actually measure: the full space- and time-resolved density correlation of a many-body system. For strongly correlated quantum systems it is almost never known exactly.
Oleksandr Gamayun and Miłosz Panfil obtain an exact analytic expression for the dynamical structure factor of a one-dimensional quantum gas of hard rods, valid for an arbitrary many-body state — with finite-temperature states and the ground state as the important special cases. They verify that it satisfies the f-sum rule and detailed balance, and reveal a hidden fermionic structure behind the correlator. The most striking result is in the static limit, where the answer can be written in terms of universal functions that, at zero temperature, coincide with the level-spacing distribution of the Gaussian unitary ensemble — the same random-matrix statistics that appear in quantum chaos and in the Riemann zeta zeros. The Letter provides a full, exact characterization of a dynamic correlation function in a strongly correlated interacting many-body system, and captures edge singularities and diffusive long-time behaviour along the way. Published in Physical Review Letters, 11 August 2026 (Editors’ Suggestion).
An electron in a molecule has no fixed position; quantum mechanics describes it by a wavefunction, and inside molecules these are the molecular orbitals whose shapes determine how a molecule absorbs light, interacts with its surroundings and reacts. Capturing the complete three-dimensional wavefunction would be enormously useful — and has been an enormous experimental problem.
An interdisciplinary group at the University of Göttingen has now imaged the 3D wavefunction of a nanometre-sized organic molecule. As Stefan Mathias notes, the wavefunction cannot be observed directly, so the team used photoemission orbital tomography — measuring the momentum of emitted electrons, which yields one half of the wavefunction — and reconstructed the missing half computationally. Two advances made it practical, explains co-lead Matthijs Jansen: the algorithm was redesigned from the ground up so that reliable 3D images need far less experimental data, and the experiment runs on a lab-based soft-X-ray source delivering ultrashort pulses rather than a synchrotron. The reconstruction of PTCDA’s highest occupied molecular orbital resolved features smaller than the spacing between carbon atoms. First author Wiebke Bennecke notes the next step: stroboscopic videography of wavefunctions changing with femtosecond resolution. Published in Nature Communications on 19 June 2026.
Axion haloscopes use the strongest magnets humans can build to convert axions into detectable photons — but even the best of them operate over a small volume. Atsushi Taruya (Yukawa Institute for Theoretical Physics, Kyoto University) with Kōta Nomura (Kyoto University), Atsushi Nishizawa (Hiroshima University) and Yoshiaki Himemoto (Nihon University) asked whether the planet could do better: “We asked ourselves whether we could use the Earth itself as a giant detector,” he says. The Earth–ionosphere cavity is a natural resonator — best known for the lightning-driven Schumann resonances — and it amplifies electromagnetic waves right in the frequency band of interest.
For axions and dark photons roughly 19 to 21 orders of magnitude lighter than an electron, the surrounding halo behaves less like specks of matter than a faint, continuously oscillating field. Earlier models were reliable only below 1 Hz; the team built a framework including the atmosphere’s varying electrical conductivity, which predicts a strong boost near 8 Hz and reliable predictions up to about 30 Hz. The model also separates the two candidates: an axion signal needs Earth’s magnetic field, so its strength and direction vary with location (Southeast Asia is most sensitive), while a dark-photon signal should look the same worldwide. Applying this to 2012–2022 geomagnetic data from the British Geological Survey’s Eskdalemuir Observatory, they found no axion, but tightened the ground-based bound on the axion–photon coupling by about a factor of 100 over the previous best terrestrial result — competitive with Chandra and NuSTAR X-ray limits, without those limits’ astrophysical assumptions. The dark-photon search turned up several persistent candidate signals that instrumental or environmental effects may yet explain; multi-site observations will decide.
Light can push a material into a hidden state — a configuration with properties unlike anything reachable by heating or cooling. Such photoinduced states are attractive for optical control of materials, but the earliest moments of their formation unfold on the femtosecond timescale, which has kept the actual pathway hidden in its own right.
A team led by Assistant Professor Tadahiko Ishikawa (Institute of Science Tokyo), with then-doctoral student Samiran Banu (now at RIKEN) and collaborators at Tohoku University and Nagoya Institute of Technology, applied 6-femtosecond time-resolved reflection spectroscopy plus model calculations to a donor–acceptor type metal–organic framework (MOF). The Tohoku side was led by Hitoshi Miyasaka (Institute for Materials Research) and Shinichiro Iwai (Graduate School of Science), with Akira Takahashi at Nagoya Institute of Technology. Transient spectra show an additional absorption band developing on a 30 fs timescale, marking the ultrafast birth of the photoinduced hidden state, and time–frequency analysis captures spectral-weight transfer between phonons. “We found that the photoinduced hidden state forms within 30 fs through a previously unknown intermediate electronic state,” Ishikawa says; the paper identifies a bond-order wave — a spatially periodic modulation of the electronic bonding strength between neighbouring sites — as the driver, with the photoinduced state forming afterwards together with a lattice modulation. The press release adds that this state may be a polar state with broken inversion symmetry, which would open a route to ferroelectric-like functionality. As Ishikawa puts it, revealing intermediate states should help design materials that can be efficiently controlled by light. Published in Physical Review Letters, 22 July 2026.
Topological quantum computation stores information in the fusion space of non-Abelian anyons, where braiding — moving one quasiparticle around another — implements logical gates that are geometrically protected from local noise. The catch has been that for the simplest non-Abelian topological orders, braiding alone cannot reach universality: some gates are simply unreachable.
A collaboration spanning the University of Chicago Pritzker School of Molecular Engineering, Harvard, Stony Brook and Quantinuum closes that gap by treating anyon fusion — merging anyons and measuring their collective charge — as a computational primitive in its own right, implementing a proposal made by Carlos Mochon in 2003. On Quantinuum’s H2 trapped-ion processor they prepare a 54-qubit topologically ordered wavefunction for S3, the smallest non-Abelian group, then combine braiding with charge measurement and fusion to realize a complete universal gate set. The demonstration also shows that non-Abelian anyons can prepare a logical magic state directly through topological operations, suggesting a path to fault tolerance with far less reliance on resource-hungry magic-state distillation. Published in Nature 655, 591–597 (15 July 2026).
Quark–gluon plasma (QGP) forms at temperatures over 100,000 times hotter than the Sun’s centre, where protons and neutrons dissolve into their constituent quarks and gluons. It filled the universe for the first millionths of a second after the Big Bang, and physicists recreate it by smashing heavy nuclei such as lead. Small systems were long assumed to be too small.
One year after the LHC’s first-ever oxygen run, all four main collaborations — ALICE, ATLAS, CMS and LHCb — have each reported signs of QGP in oxygen–oxygen and neon–neon collisions. The key evidence is parton energy loss: fast quarks and gluons shedding energy in the hot medium. ATLAS sees it as an imbalance between pairs of jets, stronger in more central collisions where a larger plasma volume means greater loss, with a matching pattern for charged particles recoiling against photons. CMS observes suppressed charged-particle production relative to proton–proton collisions. LHCb finds that suppression of charm-plus-light-quark hadrons grows in the heavier neon system, as expected if QGP volume scales with collision size. ALICE reports anisotropic flow with three-quark baryons emitted more strongly in a preferred direction than two-quark mesons. Together the four results push the frontier of “how small can a droplet of early-universe matter be” considerably further down.
Symmetry constrains what quantum operations are allowed, and time-translation symmetry is the one tied to energy conservation and to how well a system can keep time. The resource theory of asymmetry formalizes this, and is well developed for discrete-variable (qubit-like) systems. Continuous-variable systems — the natural language of quantum optics — have lagged behind.
Xueyuan Hu, Lea Lautenbacher, Giovanni Spaventa, Martin B. Plenio, Nelly H. Y. Ng and Jeongrak Son close the gap with a rigorous classification of Gaussian covariant operations — Gaussian quantum operations covariant under time translations. The surprise is how much fails to carry over: in the Gaussian optical setting, discrepancies appear between physical and thermodynamic implementation, in the extensivity of asymmetry, and in catalytic advantages — all results taken for granted in the discrete-variable case. Their toolkit includes a peculiar pair of asymmetry measures that are completely nonextensive. The wider moral is that real-world settings with several simultaneous constraints (symmetry, Gaussianity, thermodynamics) have structure that examining each constraint separately simply cannot reveal. Published in Physical Review Letters, 11 August 2026.
Xueyuan Hu・Lea Lautenbacher・Giovanni Spaventa・Martin B. Plenio・Nelly H. Y. Ng・Jeongrak Sonは、時間並進の下で共変なガウス共変操作の厳密な分類を与えてこの空白を埋めた。驚くべきは、離散変数系の結果がどれほど引き継がれないかである。ガウス光学系では、物理的実装と熱力学的実装のあいだにずれが生じ、非対称性の示量性が成り立たず、触媒的優位性も破れる——いずれも離散変数系では当然視されてきた性質だ。彼らの道具立てには、完全に非示量的という奇妙な性質を持つ2つの非対称性尺度が含まれる。より一般的な教訓はこうだ。対称性・ガウス性・熱力学といった複数の制約が同時にかかる現実的な状況には、個々の制約を別々に調べていては見えてこない豊かな構造がある。Physical Review Letters 2026年8月11日付掲載。
In cavity quantum electrodynamics, a photon that leaves the cavity can be counted as lost or counted as work — reused as a power source. Both bookkeeping conventions define valid thermodynamic frameworks, and the choice has usually been treated as a matter of accounting taste rather than physics.
Marcelo Janovitch, Sander Stammbach, Matteo Brunelli and Patrick P. Potts (University of Basel) show it is not. They formulate a rigorous semiclassical limit of cavity QED and find that the resulting thermodynamic description can differ qualitatively from that of the fully quantized model. The sharpest diagnostic is the thermodynamic uncertainty relation, which bounds how precise a current can be for a given entropy production, and whose violations are a hallmark of quantum behaviour: in the semiclassical limit, those violations are recovered by only one of the two frameworks — the one that treats part of the photon flux as a power source. The authors illustrate the result in a three-level system coupled to a driven cavity. The message is that the bookkeeping convention encodes real physical assumptions about what the light field is doing. Published in Physical Review Letters, 11 August 2026.
Electric dipole moments (EDMs) are among the sharpest probes of physics beyond the Standard Model. A permanent EDM along a particle’s spin axis violates both parity and time-reversal symmetry, and hence — via the CPT theorem — CP symmetry, the very ingredient needed in excess to explain why the universe contains matter rather than equal parts matter and antimatter. The Standard Model predicts a muon EDM far too small to see, so any detection would be unambiguous new physics.
A year after delivering its final measurement of the muon’s magnetic anomaly, the Muon g-2 collaboration at Fermilab has turned the same apparatus to the EDM. The storage ring — the 50-foot superconducting magnet shipped from Brookhaven in 2013 — is tuned for the magnetic moment, but it retains incidental EDM sensitivity: an EDM would tilt the muon spin’s precession plane, producing an up-down asymmetry in where decay positrons emerge. Two straw-tube trackers comprising 32 layers of aluminium-coated Mylar straws measure the average vertical decay angle that encodes this tilt. Using 2019–2020 data — only 25% of the full dataset, yet already far more muons than Brookhaven ever collected — the collaboration reports dμ = (−0.35 ± 0.19stat ± 0.34sys) × 10−19 e·cm, consistent with zero, setting a direct limit of |dμ| < 1.10 × 10−19 e·cm at 95% confidence. This is the most sensitive direct muon-EDM search ever performed and only the third worldwide in 50 years. It also fixes the reference point that the next-generation dedicated experiments now under construction in Japan (J-PARC) and Switzerland (PSI) must beat. Announced 12 August 2026; preprint arXiv:2608.11124.
When an energetic quark or gluon plows through the quark-gluon plasma produced in a heavy-ion collision, it does not simply lose energy — it dumps energy and momentum into the medium, which responds. Hydrodynamics predicts that the medium’s reaction has two parts: a Mach cone of excess particles, and behind it a diffusion wake, a region depleted of particles on the side opposite the jet’s direction of travel. The depletion is the harder half to see, because ordinary background dwarfs it.
The CMS Collaboration now reports firm establishment of the diffusion wake. Rather than looking at a single jet, they used dijets — back-to-back jet pairs — and compared dijet-hadron correlations in lead-lead against proton-proton collisions at √sNN = 5.02 TeV. The key move was to sort events by the pseudorapidity separation of the two jets: as the dijet axis tilts away from the transverse plane, the wake shifts with it in a characteristic way that background does not mimic. The depletion signal reaches a significance greater than 5 standard deviations for charged particles with transverse momenta between 1 and 2 GeV. Comparison with models that do and do not include jet-wake effects provides direct new constraints on how the plasma transports the energy a jet deposits in it. Published in Physical Review Letters 137, 071902 (13 August 2026), featured in Physics.
The Moon has no atmosphere, no weather and no plate tectonics, so its surface layer — the regolith — is one of the best archives in the solar system of what has rained down on it. Live iron-60, a radioisotope forged in supernovae, has already been found in Apollo cores, recording nearby stellar explosions within the last few million years. But the archive is not a tidy stack of layers: crater-forming impacts constantly churn the soil, a process planetary scientists call gardening. Reading the timeline requires knowing how the churning works.
Emily S. Costello, John Ellis, Brian D. Fields, Rebecca Surman and Xilu Wang present a unified stochastic model that treats gardening as a competition between impact-driven advection (bulk downward or upward transport) and diffusion (random mixing). The model reproduces the maturity profiles of Apollo cores across more than two orders of magnitude in time, from 1.4×107 to 4.5×108 years, and describes the measured depth profiles of live 60Fe well — implying that supernova dust capture does not depend on the local abundance of native iron, and that the influx was uniform across the latitudes of the Apollo landing sites. Extending the model, the authors predict lunar depth profiles for r-process isotopes that may come from supernovae or kilonovae: 244Pu (linked to terrestrial detections), 129I, 182Hf and 247Cm. Crucially, the 244Pu/60Fe depth profile can discriminate between the possible origins of 244Pu — which makes it worth digging: they argue Artemis sampling should reach depths of order 100 cm. Published in Physical Review Letters 137, 071005 (14 August 2026), featured in Physics.
The electromagnetic coupling α is not a constant: it runs with energy, growing from about 1/137 at low energy to roughly 1/128 at the Z boson mass. Predicting the running requires knowing the hadronic contribution Δα(5)had(MZ2), which comes from quark loops in a regime where perturbation theory fails. Conventionally it is estimated from measured e+e− → hadrons cross sections. That number is one of the main limiting inputs to global electroweak fits — and hence to any indirect search for new physics.
Alessandro Conigli, Harvey B. Meyer, Hartmut Wittig and colleagues compute the hadronic running from first principles instead. Using lattice QCD in the low-energy regime, they reach permille precision for virtualities Q2 ≲ 12 GeV2, then bridge to high energy with perturbative QCD via the Euclidean split technique. The result, Δα(5)had(MZ2) = 0.027821(34)lat(35)pQCD, is more than twice as precise as recent phenomenological determinations — a total uncertainty of about 1.7 per mille (0.17%) that already meets the target set for next-generation electroweak measurements at FCC-ee. There is also a tension worth watching: at Q2 ≈ 1 GeV2 the lattice result departs from e+e−-based estimates by up to 7σ, echoing the data-versus-lattice discrepancy that has reshaped the muon g−2 story. Published in Physical Review Letters 137, 071901 (13 August 2026).
The negatively charged nitrogen-vacancy (NV−) center — a missing carbon atom next to a substitutional nitrogen in diamond — is the workhorse of solid-state quantum sensing, prized because its spin stays coherent for milliseconds at room temperature. For two decades nothing else has come close. A candidate has been waiting in the wings: a defect labelled WAR5 when discovered in 2009, later identified as the neutral oxygen-vacancy (OV0) center. Since oxygen has one more electron than nitrogen, OV0 is isoelectronic with NV− — same outer configuration, and so, in principle, the same virtues without needing to hold onto an extra charge.
Sounak Mukherjee, Nathalie P. de Leon and colleagues (Princeton University, with collaborators including Chris G. Van de Walle, Michael E. Flatté and Stephen A. Lyon) have now measured whether the promise holds. Using electron spin resonance — which works in the microwave band, where OV0’s spin transitions can be cleanly separated from those of other defects — they extracted the relaxation times T1 and T2. The numbers compare favourably with NV−: at room temperature T1 is 1 millisecond for OV0 versus 6.7 ms for NV−, while at 4 K OV0 reaches 14.4 minutes against an average of 3.3 minutes for NV−. One thing is still missing: the zero-phonon line, the optical transition that makes a color center readable by laser, sits somewhere in a thicket of other optical lines and has not yet been pinned down. Find it, and diamond quantum sensing gains a second workhorse. Published in Physical Review B 114, 074105 (13 August 2026).
Sounak Mukherjee・Nathalie P. de Leonら(プリンストン大学、共同研究者にChris G. Van de Walle、Michael E. Flatté、Stephen A. Lyonら)は、この期待が実際に成り立つかを測定した。用いたのは電子スピン共鳴である。マイクロ波帯で働くため、OV0のスピン遷移を他の不純物欠陥の遷移からきれいに分離できる。得られた緩和時間T1・T2はNV−に見劣りしない。室温でのT1はOV0が1ミリ秒、NV−が6.7ミリ秒。4KではOV0が14.4分に達し、NV−の平均3.3分を上回った。残る課題は一つ、ゼロフォノン線——色中心をレーザーで読み出せるようにする光学遷移——が多数の光学線に埋もれてまだ特定できていないことだ。これが見つかれば、ダイヤモンド量子センシングは第二の主役を得ることになる。Physical Review B 114, 074105(2026年8月13日)掲載。
The dark photon is one of the most-studied dark matter candidates: a hypothetical vector boson that mixes weakly with ordinary light. In the early universe, dark photons of the right mass would hit a resonance as the expanding plasma’s frequency swept past their mass, converting efficiently into ordinary photons (plasmons) and heating the plasma. Because such heating would leave traces in the cosmic microwave background and in primordial nucleosynthesis, this argument has been used to exclude broad swaths of dark photon parameter space.
Anson Hook, Junwu Huang and Mohamad Shalaby now revisit that argument and invalidate all of it. The exclusions assume the resonance transfers a substantial fraction of the dark photon energy density into the Standard Model plasma. But the conversion produces k ≈ 0 Langmuir waves in the electron-ion plasma, and once the Langmuir-wave energy approaches the plasma’s thermal energy, nonlinear effects driven by the ponderomotive force take over. Using dedicated particle-in-cell simulations, the authors show that large-amplitude k = 0 Langmuir waves excite higher-k Langmuir and ion-acoustic waves, producing strong spatial variations in density and hence in plasma frequency. These inhomogeneities destroy the resonance condition, saturating further conversion. The deposited energy is capped at roughly the thermal energy of the electrons at conversion time — orders of magnitude below any observable cosmological threshold. Consequently the constraints weaken by factors of 3000 to 107 across ten orders of magnitude in dark photon mass, reopening a vast region for experimental searches. Published in Physical Review Letters 137, 071004 (13 August 2026).
Every gravitational-wave measurement rests on calibration: knowing precisely how a given strain in spacetime translates into a number coming out of the interferometer. Calibration is normally established in situ, by pushing on the mirrors with radiation pressure from auxiliary lasers and measuring the response. The residual uncertainty then propagates into every inferred source parameter. But general relativity fixes the phase and amplitude evolution of a coalescing binary’s waveform very tightly — which raises a tempting inversion: for a loud enough signal, could the astrophysical source calibrate the detector instead?
The LIGO-Virgo-KAGRA collaborations report the first informative astrophysical calibration measurements, using two exceptionally loud binary-black-hole signals: GW240925 (network signal-to-noise ratio ≈32) and GW250207 (SNR ≈69), both seen by the LIGO Hanford–LIGO Livingston–Virgo network. For GW240925 the method is validated: the Hanford calibration inferred from the signal is cross-checked against known calibration errors from in situ measurements, and they agree. GW250207 is the case where it mattered — the Hanford detector was not fully stabilized at that time, leaving elevated calibration uncertainties, so astrophysical calibration was essential to obtain accurate data and to localize the source at all. Both events, being well-localized and high-SNR, are exactly the kind that can deliver precision source properties, stringent tests of general relativity, and informative dark siren cosmology — but only if calibration is handled correctly. As detector sensitivity improves, the authors argue, astrophysical calibration will become an increasingly valuable complement to in situ methods. Published in Physical Review Letters 137, 071401 (13 August 2026).
Diamond is not only a gemstone. It is the material of the capsule that holds the fuel in inertial confinement fusion, and it is believed to rain through the interiors of ice giants such as Neptune and Uranus, where carbon separates out of methane under crushing pressure. Both applications need to know where diamond melts — and the melting curve at multi-megabar pressures has remained poorly determined, with different theoretical approaches disagreeing about both the temperature and whether melting happens at all before other transformations set in.
Marius Millot, Federica Coppari, Jon H. Eggert and colleagues address this with shock compression experiments reaching 1 TPa — ten million atmospheres. Driving samples to these conditions and diagnosing the resulting states, they obtain strong evidence for shock-induced melting of diamond, locating the melting curve in a regime that had been the province of extrapolation. The result anchors equation-of-state models used both to design fusion targets, where the ablator’s behaviour under drive determines implosion symmetry, and to model the deep interiors of ice giant planets. Published in Nature Physics (13 August 2026).
Marius Millot・Federica Coppari・Jon H. Eggertらは、1テラパスカル(1000万気圧)に達する衝撃圧縮実験でこの問題に挑んだ。試料をこの条件まで駆動し、生成した状態を診断することで、ダイヤモンドの衝撃誘起融解の強い証拠を得た。これまで外挿に頼るしかなかった領域で融解曲線を位置づけたことになる。この結果は状態方程式モデルの錨となる。核融合ターゲットの設計では、駆動下でのアブレーターの振る舞いが爆縮の対称性を左右するし、氷惑星深部のモデル化にも同じ知識が要る。Nature Physics(2026年8月13日)掲載。
Proving that a quantum many-body system really breaks a symmetry — rather than merely appearing to in simulations — is hard, and the standard tools assume an energy gap. Gapless and frustrated systems, which is to say many of the interesting ones, fall outside those tools. The classical counterpart is well understood: Peierls’ argument shows that in two dimensions the Ising ferromagnet resists thermal fluctuations because flipping a domain costs energy proportional to its boundary.
Chao Yin and Andrew Lucas prove spontaneous symmetry breaking in suitably low-energy eigenstates of gapless, frustrated quantum systems: symmetric quantum perturbations of classical models that already break a finite group at some positive temperature. Notably the classical model need not even be local in space, provided it satisfies a quantum analogue of the Peierls condition. Their technique establishes quantum bottlenecks — something like a many-body WKB method for evaluating tunneling rates between symmetry-broken sectors. As a concrete case they establish robust ferromagnetism in the two-dimensional random-bond Ising model with sufficiently biased random couplings under a weak transverse field. The same machinery yields new proofs of metastability and of the slow decay of the false vacuum, now applicable to gapless metastable states. The authors present the work as a first step toward a rigorous classification of stable gapless quantum phases. Published in Physical Review Letters 137, 070405 (13 August 2026).
Quantum simulation — evolving a system under a Hamiltonian on a quantum computer — usually proceeds by Trotter-Suzuki decomposition, slicing the evolution into small steps whose ordering error must be controlled. Separately, two quantities are known to govern how hard a quantum state is to simulate classically: entanglement entropy and nonstabilizerness, often called magic. The relationship between these resources and the accuracy of quantum simulation itself has been an open question.
Xiangran Zhang, Jue Xu, Qi Zhao and You Zhou establish a rigorous connection, and the sign of the effect is the surprising part. Analysing ensembles of states with fixed entanglement entropy or fixed magic, they find first that the variance of the Trotter error decreases as entanglement entropy increases — errors concentrate more tightly for entangled states. Second, the kurtosis of the error depends linearly and negatively on magic: high-magic states have lighter-tailed error distributions, and therefore a reduced probability of large deviations. The moral is a genuinely counterintuitive one. The very quantum resources that obstruct classical emulation also make quantum simulation more robust, revealing a constructive interplay between computational complexity and numerical stability. Published in Physical Review Letters 137, 070202 (13 August 2026).
A quantum internet needs quantum memories inside repeater nodes, to buy time against transmission loss and lift the entanglement-distribution rate. Three properties matter simultaneously: multimode capacity (how many modes can be stored at once, which sets channel capacity), efficiency, and fidelity. Devices are usually optimized for one at the expense of the others, and — a subtler problem — there has been no agreed way to compare memories that trade these off differently.
Hao-Xuan Luo, Hui Yan, Shi-Liang Zhu and colleagues address both problems. They introduce the quantum interconnect rate, a single figure of merit that folds all the relevant metrics into one comparable number, and then build a memory that scores well on all three at once. Operating on 11-dimensional spatial modes, the device achieves a uniform efficiency above 80% and qubit storage fidelities above 99%, enabling efficient storage of high-dimensional qudits rather than just qubits. From the measured performance the authors estimate distribution of 3.56 ± 0.16 bits of quantum information over a 1000-km repeater link in one minute. Published in Physical Review Letters 137, 070802 (12 August 2026), featured in Physics.
Axion quasiparticles in topological materials produce a magnetoelectric coupling — the condensed-matter analogue of the axion term proposed in particle physics and long hunted as a dark matter candidate. The trouble is that their optical signatures are intrinsically feeble, which has held back both detection and any practical use.
Chang-Yin Ji, Chong Wang, Jiafang Li and Yugui Yao amplify those signatures by orders of magnitude using degenerate quasibound states in the continuum (qBICs). They integrate three-dimensional topological insulator films with achiral photonic crystal slabs — the achirality matters, since any intrinsic chirality of the photonic structure would swamp the effect being sought. The axion quasiparticles then induce an effective Zeeman effect that lifts the degeneracy of the slab’s achiral qBICs, splitting them into two nondegenerate high-Q chiral qBICs of maximally opposite handedness. The consequences are large: more than a 103-fold increase in polarization conversion efficiency and a 200-fold increase in Faraday rotation angle, relative to bare 3DTI films. The work also identifies circular dichroism and nonreciprocal optical response as new observables for probing axion quasiparticles — and, incidentally, offers a platform for controlling optical polarization, spin, phase and nonreciprocity. Published in Physical Review Letters 137, 076901 (10 August 2026).
Rare events — a hundred-year flood, a protein escaping a conformational trap — are usually modelled with Arrhenius laws: waiting times are exponentially distributed and successive occurrences are independent. That picture requires the underlying process to forget its history quickly. Many real systems do not. Geophysical time series and protein dynamics both display long-term memory, and empirically extreme events are observed to arrive in clusters rather than independently. Until now no analytical framework quantified how memory correlates successive rare events.
Apurba Biswas and Thomas Guérin supply one, for non-Markovian Gaussian processes. Using a perturbation approach, they determine analytically how long-term memory reshapes the distribution of the first and second passage times to a rarely reached threshold. The distribution comes out nonexponential — already a departure from the Arrhenius paradigm — and, more usefully, they obtain an explicit expression for the covariance between first and second passage times. From it follows a prediction of how the mean waiting time to the next extreme event depends on when the previous one occurred: the mathematical content of event clustering. Extensive stochastic simulations validate the analytics. Published in Physical Review Letters 137, 077101 (10 August 2026), with a Viewpoint in Physics.
The defining strangeness of the fractional quantum Hall effect is that its excitations carry a fraction of the electron charge — e/3, e/5, and stranger values still. Measuring that fractional charge is a delicate business. The classic route is shot noise: current through a constriction fluctuates in units of the tunneling charge, so the noise spectrum betrays it. But shot-noise experiments require care in interpretation, and the tunneling geometry is not fully under experimental control.
Mario Di Luca, Emily Hajigeorgiou and Mitali Banerjee demonstrate a different approach: a transport technique using a quantum Hall antidot — a deliberately introduced, gate-controlled impurity around which the fractional quantum Hall edge circulates. Quasiparticles tunnel through this controlled impurity, and the resulting transport signal reads out the effective charge directly. The authors describe the device as a fractional coulombmeter: an instrument whose output is the quasiparticle charge itself, with the scattering centre engineered rather than accidental. Published in Nature Physics (14 August 2026, open access).
Mario Di Luca・Emily Hajigeorgiou・Mitali Banerjeeは別の道を実証した。量子ホール・アンチドット——意図的に導入されゲートで制御される不純物で、その周りを分数量子ホール端状態が周回する——を用いた輸送測定である。準粒子はこの制御された不純物を通ってトンネルし、得られる輸送信号から実効電荷が直接読み出される。著者らはこの素子を分数クーロン計(fractional coulombmeter)と呼ぶ。散乱中心が偶発的ではなく設計されたものであり、出力そのものが準粒子電荷になる測定器というわけだ。Nature Physics(2026年8月14日、オープンアクセス)掲載。
Decays in which a b quark turns into a d quark plus a lepton pair are flavour-changing neutral currents: forbidden at tree level in the Standard Model and proceeding only through loops. That makes them exquisitely sensitive to heavy new particles circulating in the loop, and it also makes them extremely rare, at the 10−8 level. The corresponding muon mode B+ → π+μ+μ− has been observed; the electron mode is harder, because electrons radiate as they traverse the detector and their energy is more difficult to reconstruct.
The LHCb Collaboration now reports the first evidence for B+ → π+e+e−, using proton-proton collision data at centre-of-mass energies of 7, 8 and 13 TeV corresponding to an integrated luminosity of 9 fb−1. The excess has a significance of 3.2σ — evidence rather than observation, in the conventional terminology — and the branching fraction is measured as (2.4+0.9−0.8+0.4−0.2) × 10−8, where the first uncertainty is statistical and the second systematic. The result is consistent with the Standard Model expectation. Establishing the electron channel matters because comparing it against the muon channel is precisely how lepton flavour universality is tested in b → d transitions — the same class of comparison that produced, and then dissolved, the celebrated RK anomalies in b → s. Published in Physical Review Letters 137, 071804 (14 August 2026).
Several theoretical proposals describe horizonless compact objects — boson stars, gravastars and their relatives — that would mimic black holes closely enough to pass as one in a gravitational-wave signal. Distinguishing them requires an observable sensitive to internal structure. One such handle is the spin-induced quadrupole moment (SIQM): how much an object bulges when it spins. For a Kerr black hole the SIQM takes a unique value fixed by the no-hair theorem; exotic objects generally deviate.
N. V. Krishnendu, Tamara Evstafyeva and a large collaboration exploit the tight SIQM bounds from GW241011 to constrain the nature of its primary component. Across the exotic-compact-object models considered, the verdict is mixed but informative: rotating boson stars with quartic self-interactions cannot explain the primary and are excluded, whereas models of sufficiently large compactness, C ≳ 0.24, remain viable contenders. The broader lesson is methodological — individual loud events with well-measured spin effects are now sharp enough to prune the exotic-compact-object landscape one model family at a time. Published in Physical Review Letters 137, 071402 (14 August 2026).
N. V. Krishnendu・Tamara Evstafyevaら大規模な共同研究チームは、GW241011から得られた厳しいSIQM制限を利用して、その主星の正体を絞り込んだ。検討されたエキゾチック・コンパクト天体模型に対する判定は明暗が分かれるが、有益である。4次自己相互作用を持つ回転ボソン星では主星を説明できず排除される一方、十分に大きなコンパクトネスC ≳ 0.24を持つ模型は依然として候補として残る。より広い教訓は方法論的なものだ。スピン効果がよく測定された個々の大信号イベントは、いまやエキゾチック天体の理論空間を模型族ごとに刈り込めるだけの鋭さを持つに至った。Physical Review Letters 137, 071402(2026年8月14日)掲載。
The Hatano-Nelson model — a lattice with asymmetric left/right hopping — is the canonical example of a non-Hermitian system with nonreciprocal transport, and underlies much of the recent excitement about non-Hermitian topology and the skin effect. Non-Hermitian Hamiltonians arise physically from dissipation, but the standard derivations idealize the dissipation as frequency-independent. Real experimental environments are not like that: dissipation that varies with frequency is non-Markovian, and a Markovian treatment cannot capture it.
Sumit Kumar Jana, Ryo Hanai, Tan Van Vu, Hisao Hayakawa and Archak Purkayastha show how a quantum non-Markovian Hatano-Nelson model arises microscopically in a quasi-one-dimensional dissipative lattice. They work with nonequilibrium Green’s functions, requiring neither weak system-bath coupling nor timescale separation — the two approximations a Markovian treatment would have needed. The resulting effective system displays the defining nonreciprocal hopping of the Hatano-Nelson model along with uniform dissipation, both frequency-dependent, in bosonic and fermionic settings alike. Two genuinely non-Markovian phenomena emerge that have no Markovian description and no reciprocal analogue: unidirectional frequency blocking in the bosonic case, and a distinctive nonequilibrium dissipative quantum phase transition in the fermionic case. Published in Physical Review Letters 137, 070404 (14 August 2026).
A quantum droplet is a self-bound blob of ultradilute quantum fluid, held together not by classical surface tension but by a balance between mean-field attraction and quantum fluctuations. Droplets are established in Bose-Bose mixtures and dipolar gases. Whether they can exist in a Bose-Fermi mixture — where the stabilizing pressure would come from Fermi degeneracy rather than fluctuations — has been an open question, complicated by the tendency of strongly attractive boson-fermion pairs to simply bind into dimers instead.
Sam Foster, Olivier Bleu, Jesper Levinsen and Meera M. Parish study the canonical problem of a Fermi gas interacting with a weakly repulsive Bose-Einstein condensate at zero temperature. To cover the full range of boson-fermion interactions they construct a versatile variational Ansatz incorporating pair correlations and correctly reproducing the different polaron limits. The result: self-bound quantum droplets can exist in the strongly interacting regime, preempting boson-fermion dimer formation, when Fermi pressure balances the resonant boson-fermion attraction. This happens for a range of mass ratios near equal masses — conditions achievable in existing experiments. Raising the fermion density instead produces phase separation between mixture and excess fermions, along with behaviour reminiscent of a liquid-gas critical point. The broader implication is that first-order quantum phase transitions play a crucial and underappreciated role in the Bose-Fermi phase diagram. Published in Physical Review Letters 137, 073402 (14 August 2026).
Sam Foster・Olivier Bleu・Jesper Levinsen・Meera M. Parishは、絶対零度で弱く斥力的なボース・アインシュタイン凝縮体と相互作用するフェルミ気体という定番の問題を調べた。ボソン・フェルミオン相互作用の全域をカバーするため、対相関を取り込みつつ異なるポーラロン極限を正しく再現する汎用の変分Ansatzを構成している。結論はこうだ。フェルミ圧が共鳴的なボソン・フェルミオン引力と釣り合うとき、強相互作用領域で自己束縛量子液滴が存在しうる——しかもボソン・フェルミオン二量体の形成に先んじてである。これは質量比が等質量近傍のある範囲で起こり、現行の実験で到達可能な条件だ。一方フェルミオン密度を上げると、混合系と余剰フェルミオンとの相分離が生じ、気液臨界点を思わせる振る舞いも現れる。より広い含意は、1次量子相転移がボース・フェルミ相図で決定的かつ過小評価された役割を果たしているということである。Physical Review Letters 137, 073402(2026年8月14日)掲載。
A nucleus is not a static ball of nucleons. Its density fluctuates, and the character of those fluctuations — how fast, how far, how random — encodes the many-body correlations that ab initio nuclear theory exists to describe. Until recently, computing the spatiotemporal pattern of these fluctuations from realistic interactions was out of reach.
Francesca Bonaiti, Gaute Hagen and Thomas Papenbrock compute them for 16O, 24O and 48Ca using nuclear interactions from chiral effective field theory together with the time-dependent coupled-cluster method — an approach that scales gently enough with mass number to reach medium-mass nuclei while remaining systematically improvable. The finding is a clean characterization: two-particle–two-hole excitations generate small-amplitude fluctuations that are fast, short-ranged, and stochastic in character. Each adjective carries information — the timescale separates these fluctuations from collective modes, the short range distinguishes them from surface vibrations, and the stochasticity connects them to the statistical description of nuclear excitation. Published in Physical Review Letters 137, 072502 (13 August 2026).
Quantum spin liquids are magnets that refuse to order even at absolute zero, hosting fractionalized excitations — spinons carrying spin-1/2 with no charge — and emergent gauge fields. Because there is no static magnetic order to measure, the decisive experimental fingerprint is the spin excitation spectrum. Computing that spectrum theoretically has been a serious obstacle: dynamical quantities are far harder than ground-state energies, and the kagome lattice is among the most frustrated settings there is.
Jiahang Hu, Runze Chi, Yibin Guo, B. Normand, Hai-Jun Liao and T. Xiang use state-of-the-art tensor-network methods to obtain the full dynamical spectral function of the J1–J2 kagome Heisenberg model, benchmarking by tracking how it evolves across the magnetically ordered and spin-liquid phases. As |J2|/J1 is reduced, spin-wave modes suffer increasingly strong renormalization, flatten, and then merge into a continuum — the signature of deconfined spinons — at all finite energies in the spin-liquid phase. The low-energy continuum, together with gap closure at multiple high-symmetry points, identifies the phase as the gapless U(1) Dirac spin liquid — strong evidence in a ground-state debate that has run for more than a decade. The calculated spectra provide clear fingerprints for neutron scattering on candidate kagome materials. Published in Physical Review Letters 137, 076504 (14 August 2026).
Crystals are defined by periodic order. Amorphous solids are defined by its absence — and that absence has resisted a unifying description, because disordered, nonequilibrium structures come in bewildering variety. A central question is whether the rigidity of an amorphous solid comes from the same physics as the glass transition, or from jamming, the point at which particles first make enough contacts to resist deformation.
Jianhua Zhang, Jiaqi Si, Ning Xu and Hua Tong answer this by systematically tuning jammed elastic packings from perfectly crystalline all the way to fully disordered. They find that mechanical properties are universally governed by jamming criticality, showing the characteristic scaling behaviours near the jamming transition — everywhere except the singular close-packed point. The mechanism is random nonaffine elasticity arising from contact-level disorder, which survives even when the particle positions are nearly crystalline. A striking consequence: the jamming density can approach close packing, which implies a fundamental decoupling between jamming criticality and glass transition physics. The authors further uncover a universal coordination-number distribution and contact hyperuniformity in marginally jammed states, independent of particle-level structure. The upshot is a general organizing mechanism for emergent rigidity in disordered solids, and a much wider domain of relevance for jamming than previously assumed. Published in Physical Review Letters 137, 078201 (14 August 2026).
Controlling magnetism with an electric field rather than a magnetic one is the central goal of magnetoelectrics, because electric fields are cheap to generate locally and dissipate little energy. The coupling that makes it possible — spin-electric coupling — is usually weak, especially at the molecular scale where devices would be smallest.
Yan Lu, Xilong Xu and Li Wang predict, from first-principles simulations combined with spin superexchange theory, giant spin-electric coupling in heterometallic porphyrin dimers — two porphyrin rings hosting different transition-metal centres. The coupling coefficients reach 1 to 4 orders of magnitude beyond those of conventional two-dimensional magnetic systems. The mechanism is the asymmetry itself: the two distinct metal centres create asymmetric spin superexchange pathways, which respond asymmetrically to an applied electric field. The practical consequence is that local magnetic moments can be modulated with experimentally reachable field strengths, enough to drive transitions between ferromagnetic and antiferromagnetic coupling — magnetic switching at the single-molecule level. Published in Physical Review Letters 137, 076704 (14 August 2026).
Topology and nonlinearity are each well understood on their own; their interplay is one of the harder open problems in modern physics, because the topological invariants that classify linear band structures are not obviously defined once the Hamiltonian depends on the state it acts on.
Chong-Xiao Chen, Zheng-Wei Zhou, Han Pu and Xi-Wang Luo attack this with a synthetic Su-Schrieffer-Heeger lattice carrying all-to-all nonlocal interactions. The key structural fact is that this particular nonlinearity preserves an effective chiral symmetry, which is what allows a quantized nonlinear winding number and Berry phase to be defined at all; they corroborate this with a Bogoliubov nonlinear adiabatic theory developed for the purpose. Increasing the nonlinearity then drives a sequence of topological transitions, announced by characteristic swallowtail band structures at intermediate interaction strengths and by band swapping in the strongly nonlinear regime. Band swapping in turn produces quantized fractional windings and double-period Bloch oscillations — closely related to discrete time crystals. Most striking: even starting from a topologically trivial linear system, nonlocal nonlinearity can induce an emergent topological phase with fractional winding. The model is realizable with photons in a degenerate optical cavity using Rydberg-mediated interactions. Published in Physical Review Letters 137, 073803 (14 August 2026).
When electrons are made dilute enough that Coulomb repulsion overwhelms their kinetic energy, they crystallize into a Wigner crystal — a lattice made of electrons alone, predicted by Eugene Wigner in 1934. Electron Wigner crystals have now been observed in atomically thin transition metal dichalcogenides, where a charge-tunable monolayer lets the density be dialled through the crystallization point. But observing that a crystal exists is not the same as observing how it moves: its collective excitations — the phonons of an electron lattice — and its internal dynamics have remained largely inaccessible.
Two independent teams now report access to precisely that, in back-to-back papers. L. Wang, F. Menzel, T. Smoleński and colleagues reveal signatures of the collective excitations in the optical spectra of a charge-tunable monolayer semiconductor, identifying them as Wigner crystal polarons — composite objects in which an optical excitation drags the surrounding electron lattice. Separately, Lifu Zhang, Liuxin Gu, You Zhou and colleagues use Wigner polarons as a probe to track the dynamics of the crystal in a monolayer semiconductor. Together the two experiments turn the polaron into a spectroscopic handle on an electron solid, opening the internal life of Wigner crystals to measurement. Published in Nature Physics (11 August 2026).
Estimating several parameters at once from a quantum system is fundamentally harder than estimating one, because the optimal measurements for different parameters need not commute. The ultimate precision limit in this setting is the Holevo-Nagaoka bound — mathematically well characterized, and known to be attainable in the asymptotic limit, but attainable by what apparatus? The bound’s proofs are abstract, typically invoking collective measurements on many copies without saying what those measurements physically are.
Mankei Tsang proposes a concrete physical scheme. Given multiple independent and identically distributed quantum objects, the objects are made to interact physically with bosonic ancillae, and the ancillae are then subjected to a general-dyne measurement — the family of Gaussian measurements that includes homodyne and heterodyne detection as special cases and is routine in quantum optics laboratories. The result is a far more concrete description of the experimental setup needed to approach the ultimate multiparameter precision limit, replacing an abstract existence argument with something an experimentalist can read as a blueprint. Published in Physical Review Letters 137, 070804 (14 August 2026).
A central question in QCD is whether femtoscale droplets of quark-gluon plasma form in small collision systems — projectiles far lighter than lead. Collisions of light ions such as 16O and 20Ne offer a uniquely clean handle, because these nuclei have well-characterized and different intrinsic shapes: oxygen-16 is nearly spherical with alpha-cluster structure, while neon-20 is strongly deformed — often described as bowling-pin shaped.
The ALICE Collaboration presents the first measurements of elliptic (v2) and triangular (v3) flow of charged particles in 16O–16O and 20Ne–20Ne collisions at √sNN = 5.36 TeV. Hydrodynamic model predictions that explicitly incorporate the nuclear structures of the two species agree well with the data. The decisive observation is that v2 is larger in central Ne–Ne collisions than in O–O, and the enhancement is driven by nuclear geometry: a deformed projectile creates an anisotropic overlap region even in head-on collisions, which hydrodynamic expansion converts into an azimuthal momentum anisotropy. This supports the presence of nuclear geometry-driven hydrodynamic flow in light-ion collisions at the LHC, and demonstrates that light nuclei with well-defined shapes are a practical tool for constraining the initial conditions of heavy-ion collisions. Published in Physical Review Letters 137, 082301 (17 August 2026).
Published alongside the ALICE measurement, the CMS Collaboration reports its own observation of long-range collective flow in the same light-ion systems, with a different detector, a different analysis strategy and complementary sensitivity — which is exactly what one wants when the claim is that nuclear shape leaves a fingerprint on the quark-gluon plasma.
CMS collected O+O and Ne+Ne data at √sNN = 5.36 TeV with integrated luminosities of 7 nb−1 and 0.8 nb−1 respectively, and measured two- and four-particle azimuthal correlations over nearly five units of pseudorapidity. The long rapidity reach matters: it suppresses short-range “nonflow” correlations from jets and resonance decays that can mimic collectivity in small systems. Significant v2 and v3 harmonics are observed in both systems, and the ratios of vn between Ne+Ne and O+O show sensitivity to quadrupole correlations in the nuclear wave functions. Hydrodynamic models fed with ab initio nuclear structure inputs — nuclear lattice effective field theory, variational Monte Carlo and extended quantum molecular dynamics — qualitatively reproduce both the centrality dependence of the vn and the Ne+Ne to O+O ratios. The measurement provides new constraints on how nuclear structure and collective dynamics interlock. Published in Physical Review Letters 137, 082302 (17 August 2026).
Shell closures — the “magic numbers” that make nuclei unusually bound — are normally associated with spherical shapes. But in the heavy actinides, where nuclei are strongly deformed, gaps can open in the single-particle spectrum at deformation, producing deformed magic numbers. Z = 100 and N = 152 have long been suspected of being such a pair, which would make 252Fm a deformed doubly magic nucleus. Testing this requires measuring the rotational band that a deformed nucleus builds on its ground state — and the more rigid the nucleus, the more its rotational energies reveal.
R. Orlandi and colleagues investigated the ground-state rotational band of 252Fm using two different experimental setups at the JAEA Tandem Accelerator Laboratory in Japan. The 2+→0+ and 4+→2+ transitions were seen in the α decay of 256No via α–γ coincidence spectroscopy, while the (6+)→(4+), (8+)→(6+) and (10+)→(8+) transitions were identified in the prompt γ-ray spectrum of 252Fm produced by the 18O + 249Cf multinucleon transfer reaction. Compared with neighbouring fermium isotopes and other N = 152 isotones, 252Fm turns out to have the lowest first excited 2+ energy, 41.88(16) keV, and the largest kinematic moment of inertia — both signatures of enhanced stability against rotation. These constitute new evidence for deformed shell gaps at Z = 100 and N = 152. Published in Physical Review Letters 137, 082501 (17 August 2026).
Lorentz invariance — the statement that physics looks the same to all uniformly moving observers — is the bedrock of both relativity and quantum field theory, and it is one of the few principles fundamental enough that many approaches to quantum gravity predict tiny violations of it. Systematically searching for such violations requires a framework, and the standard one is the Standard-Model Extension (SME), which parameterizes every possible Lorentz-violating term. In the nonrelativistic limit of the SME’s minimal matter sector, there are 132 degrees of freedom. Dozens of experiments over several decades have chipped away at them — yet 43 remained entirely unconstrained, a persistent blind spot in an otherwise systematic program.
Marshall J. Basson, Jay D. Tasson and colleagues close that gap. They place limits on all 43 previously unconstrained degrees of freedom, and improve on 13 prior limits along the way — completing a program that has run for decades. The approach introduced also carries forward: the authors identify 49 degrees of freedom that could be improved further in suitable future experiments, and note additional discovery potential from combining data taken at different geographical locations — since Lorentz violation would manifest as sidereal variations whose phase depends on where on Earth the apparatus sits. Published in Physical Review Letters 137, 081601 (17 August 2026).
Marshall J. Basson・Jay D. Tassonらはこの空白を埋めた。未制限だった43自由度すべてに制限を与え、ついでに13個の既存制限を改善した——数十年続いた計画の完結である。導入された手法はさらに先へも続く。著者らは適切な将来実験でさらに改善しうる49自由度を特定し、異なる地理的位置で取得したデータを組み合わせることによる追加の発見可能性を指摘している。ローレンツ対称性の破れは恒星時変動として現れ、その位相は装置が地球上のどこにあるかに依存するからだ。Physical Review Letters 137, 081601(2026年8月17日)掲載。
The layered compound γ-PtBi2 is a topological semimetal whose surface hosts Fermi arcs — open segments of Fermi surface that connect bulk Weyl points and exist only because of the bulk topology. Recent work found signatures of surface superconductivity in this material, with gap openings implying a critical temperature orders of magnitude above the bulk value. The claim was exciting but incomplete: no superconducting vortices had been identified, and without vortices one cannot be sure the surface state possesses genuine macroscopic phase coherence rather than merely a spectroscopic gap.
Jose Antonio Moreno, Isabel Guillamón, Paul C. Canfield, Hermann Suderow and colleagues settle the question with very low temperature scanning tunneling microscopy. They find robust superconductivity with TC = 2.9 K and an upper critical field HC2 ≈ 1.8 T, linked to the Fermi arcs. Decisively, they observe quantized superconducting vortices and the Josephson effect — the two textbook manifestations of a well-defined superconducting phase. Together these demonstrate genuine two-dimensional macroscopic quantum phase coherence confined to the surface of a Weyl semimetal, a combination of topology and superconductivity that has been sought for its potential to host unconventional pairing. Published in Physical Review Letters 137, 086001 (17 August 2026).
The word “wave” carries an assumption: that a disturbance propagates. Waves on a shore, sound in air, light in a fibre — and gravitational waves, ripples in spacetime itself, which have been detected directly for a decade now. Yet there is a known exception in wave physics. Some perturbations remain spatially localized forever, coexisting with propagating modes at the same frequency and refusing to radiate away. These are bound states in the continuum (BICs), a concept born in quantum mechanics in 1929 and now a workhorse of photonics, where quasi-BICs deliver ultrahigh-Q resonances.
Rodrigo Berté asks what BICs would look like for gravitational waves, and formulates the corresponding theory. The motivation is concrete. Photons and gravitational waves — and the hypothetical graviton — can convert into one another in strong magnetic fields, but the two waves fall out of phase in that setting, spoiling the conversion. In photonics this kind of dissonance is routinely circumvented using quasi-BICs; the proposal is to do the same by localizing gravitational waves. If it works, gravitational-wave quasi-BICs would become a tool for probing both what exists within spacetime and the nature of spacetime itself — including the effects on quantum systems of shaking flat spacetime at very high frequencies. Published in Physica Scripta (2026); presented via Science X Dialog on 14 August 2026.
Rodrigo Bertéは、重力波にとってBICとは何かを問い、対応する理論を定式化した。動機は具体的である。光子と重力波——そして仮説上のグラビトン——は強磁場中で互いに変換されうるが、その状況では2つの波の歩調が合わなくなり、変換が損なわれる。フォトニクスではこの種の不協和を準BICで日常的に回避している。ならば重力波を局在させることで同じことができるのではないか、という提案だ。これがうまくいけば、重力波の準BICは時空の中に何があるかと、時空そのものの性質の双方を探る道具になる——平坦な時空を非常に高い周波数で揺さぶったとき量子系に何が起きるか、という問いも含めてである。Physica Scripta(2026年)掲載、2026年8月14日にScience X Dialogで紹介。
Simulating electrons and nuclei moving together out of equilibrium — nonadiabatic molecular dynamics (NAMD) — governs everything from photocatalysis to how a qubit loses coherence. The bottleneck has been the excited-state landscape: a general, accurate, efficient method for computing potential energy surfaces, forces and nonadiabatic couplings for multiple electronic states at once. Machine learning has transformed ground-state simulation, and has shown promise for excited states in isolated molecules, but a unified framework for general condensed matter systems has remained out of reach.
Changwei Zhang, Weibin Chu, Hongjun Xiang and colleagues introduce on-the-fly neural network NAMD for solids. An equivariant neural network predicts the system Hamiltonian itself rather than energies alone, from which excited-state energies, forces and nonadiabatic coupling vectors follow at a fraction of the cost of ab initio calculation. Crucially, the approach makes simulations at hybrid functional accuracy feasible, a level previously inaccessible for NAMD in solids. Three demonstrations make the case: correcting order-of-magnitude errors in the carrier dynamics that conventional procedures predict for a MoS2/WS2 heterostructure; simulating photoinduced ferroelectric switching, previously out of reach entirely; and capturing real-time polaron formation in TiO2 at the hybrid-functional level. Published in Physical Review Letters 137, 076905 (14 August 2026).
📡 / 100チャンネル同時の量子テレポーテーションを実現——プログラマブル・ホログラフィで10×10=100個の「空間的に分離可能な量子チャンネル」を生成し、測定不要の全光学フィードフォワードで並列転送。100画素の画像(文字「Q」)を古典限界を超える忠実度でテレポート(Louら、Phys. Rev. Lett. 137, 080801)
Quantum teleportation transfers a quantum state from sender to receiver using shared entanglement, without the state itself travelling across the channel. Almost every demonstration so far has moved one state at a time, which is a poor match for networks that will need to carry many modes in parallel.
Y. Lou and colleagues now report hundred-channel reconfigurable quantum teleportation. Programmable holographic engineering, based on a weighted Gerchberg–Saxton algorithm, shapes a beam into a 10×10 array of 100 independently addressable spatial modes that act as naturally spatially separable quantum channels. A measurement-free, all-optical feedforward then transfers quantum information across every channel at once, with no detection-and-correction step in the loop. The team teleported the complete 100-mode array and, as a demonstration, a 100-pixel image of the letter “Q”, with fidelities beating the corresponding classical limits. Because the channel count and layout are reconfigurable in software, the scheme offers a concrete route to scaling continuous-variable quantum networks. Published in Physical Review Letters on 20 August 2026 and highlighted by APS Physics.
An optical microcomb is a set of light frequencies spaced as evenly as the teeth of a comb, generated inside a microresonator on a chip. Beat two comb lines together on a fast photodiode and you get a microwave or millimetre-wave tone whose purity inherits the optical comb’s precision. Millimetre waves carry far more bandwidth than today’s wireless bands, but generating them stably has been the bottleneck.
A team led by Luke Peters, Antonio Cutrona, Alessia Pasquazi and Marco Peccianti at Loughborough University’s Emergent Photonics Research Centre, with collaborators at the University of Sussex, City University of Hong Kong and QXP Technologies, take a different route to the comb: instead of simply pumping a chip microresonator with a laser, they place the rice-grain-sized microresonator inside a much larger loop of optical fibre. Light circulates through both, so the desired states start on their own and stay locked — the team reports the comb surviving mechanical disturbance of the optical table. From that comb they generate several precisely spaced millimetre-wave frequencies simultaneously, rather than the single tone targeted by earlier work, and show that individual comb lines can be boosted or suppressed to tailor the output. Applications range from higher-capacity 6G links to radar, radio astronomy and the precision timing needed by quantum technologies. Published in Nature Communications, 20 August 2026.
Symmetry is the rulebook of solid-state physics. In a crystal, selection rules forbid harmonic coupling between zone-centre modes belonging to inequivalent irreducible representations — vibrations of the wrong symmetry simply cannot mix. Anharmonic processes could in principle bridge symmetry sectors, but direct evidence for such a mechanism in thermodynamic equilibrium had been missing.
An international team with first authors Francesco Barantani and Xinyue Peng, led by Edoardo Baldini (University of Texas at Austin) with theory from Angel Rubio’s group at the Max Planck Institute for the Structure and Dynamics of Matter (MPSD) and collaborators including Kiel University and Sapienza, now provides it. Their material is the van der Waals crystal 4Hb-TaS2, which hosts a ferroaxial charge density wave: electronic modulation intertwines with a rotational lattice distortion, forming star-of-David clusters that can point either way and thereby break multiple mirror symmetries while leaving the crystal with a built-in sense of rotation. Using helicity-resolved Raman micro-spectroscopy, which resolves individual ferroaxial domains, the team finds a temperature-dependent interplay in which amplitude fluctuations of the charge order act as a resonant intermediary between phonons of different symmetry, hybridizing them and strongly enhancing the chiral (Raman optical activity) response — a channel they name resonant chiral dressing, reproduced by a full theoretical treatment. Because the effect survives to room temperature, it offers a practical handle for detecting and eventually controlling ferroaxial states, which are hard to see precisely because they carry no net electric or magnetic moment. Published in Nature Physics 22, 720–727 (1 May 2026); re-reported in the general press on 20 August 2026.
Keywords: ferroaxial order, フェロアキシャル秩序, ferro-rotational, 強回転秩序, charge density wave, 電荷密度波, 4Hb-TaS2, van der Waals crystal, ファンデルワールス結晶, amplitude mode
🌌 / 「宇宙の大きさと膨張率は同時には決められない」——量子論的な不確定性を宇宙全体に適用した新提案。修正フリードマン方程式は暗黒エネルギーを入れずに後期宇宙の加速膨張を生み、ビッグバン特異点を置き換える可能性も(S. M. Koushiappas、Physical Review D)
Two of the deepest open problems in physics — how to quantize gravity, and what drives the accelerating expansion of the Universe — are usually attacked separately. A new single-author study argues they may be two views of the same thing.
Savvas M. Koushiappas proposes a cosmological uncertainty relation: the scale factor of the Universe and its rate of expansion cannot both be specified with arbitrary precision, in the same way that position and momentum cannot in ordinary quantum mechanics. Imposing that relation modifies the Friedmann equation governing cosmic expansion, and the modification behaves, at late times, exactly like the accelerated expansion that cosmologists currently attribute to dark energy — here it emerges instead as a macroscopic imprint of quantum gravity, with no new energy component added by hand. The paper presents the resulting Hubble diagram and the residuals relative to ΛCDM, and notes that depending on the detailed form of the relation the same construction can replace the Big Bang singularity. It remains a theoretical proposal that must be confronted with precision cosmological data. Published in Physical Review D (2026); press coverage 21 August 2026.
Savvas M. Koushiappasが提案するのは「宇宙論的不確定性関係」だ。通常の量子力学で位置と運動量を同時に確定できないのと同じように、宇宙のスケールファクター(大きさ)とその膨張率も、両方を任意の精度で指定することはできないとする。この関係を課すと宇宙膨張を支配するフリードマン方程式が修正され、その修正項は後期宇宙において、現在の宇宙論が暗黒エネルギーに帰している加速膨張とまさに同じ振る舞いを示す——つまり加速膨張が、新しいエネルギー成分を手で付け加えることなく、量子重力の巨視的な痕跡として自然に現れるのである。論文では得られたハッブル図とΛCDMからの残差が示され、関係式の具体的な形によってはビッグバン特異点そのものを置き換えうることも指摘されている。もちろん現段階では理論的提案であり、精密宇宙論データとの突き合わせが今後の課題だ。Physical Review D(2026年)掲載、報道は8月21日。
In an integrated photonic circuit, light is guided in a core — usually silicon nitride, prized for low loss and strong Kerr nonlinearity — while the surrounding cladding, typically silica, is treated as mere packaging whose job is to keep the light confined. Silicon nitride, however, provides too little Raman gain on its own, so Raman lasing had never been achieved in a silicon nitride integrated platform.
Arghadeep Pal, Alekhya Ghosh, Pascal Del’Haye and colleagues borrow the missing capability from the cladding. Because a fraction of the guided field always extends beyond the core, they engineered silica-clad silicon nitride ring resonators so that roughly 31% of the optical field sits in the cladding, then pumped them with a continuous-wave laser. A new line appeared 11 THz from the pump — the signature of stimulated Raman scattering in silica — with an on-chip lasing threshold of 143 mW. Letting the two materials work together, so that silica supplies Raman gain while the silicon nitride core supplies the Kerr nonlinearity that builds optical frequency combs, optimized devices generated combs spanning more than 400 nm with over 32% conversion efficiency. The same design philosophy — choosing core and cladding for complementary nonlinearities rather than treating the cladding as inert — should extend to other material pairs. Published in Advanced Photonics 8(4), 046008 (23 June 2026); publicized in August 2026.
⚛️ / アト秒科学30年の「カットオフの壁」を電子相関が突き破る——ヘリウムで相関した2電子が同時に再結合することで、従来カットオフ(120 eV超)を超える弱い第2プラトーがウォーターウィンドウ(280 eV)まで伸長。カットオフ則も3.2Up→5.5Upへ(Wang・de las Heras・Popmintchevら、UCサンディエゴ/ウィーン工科大、Nature Photonics)
High-harmonic generation (HHG) — an intense laser driving atoms to emit light at far higher frequencies, up into the X-ray range — is the foundation of attosecond science, work recognized by the 2023 Nobel Prize in Physics. For three decades it has been described by the single-active-electron picture: one electron tunnels out, is accelerated by the field, and recombines, which fixes a sharp cutoff photon energy at the ionization potential plus about 3.2 times the ponderomotive energy (Ip + 3.2Up), beyond which emission collapses.
Siyang Wang, Jieyu Yan, Alba de las Heras, Luis Plaja, Dimitar Popmintchev and Tenio Popmintchev (University of California San Diego, and the Institute of Photonics at TU Wien) now show experimentally that this limit is not fundamental. Driving helium, they observe a weak secondary plateau that carries coherent emission past the conventional cutoff above 120 eV and all the way to the water window at 280 eV. The mechanism is double-electron recombination of strongly correlated electron pairs, and it obeys a new cutoff scaling of up to 5.5 times the ponderomotive energy instead of the textbook 3.2. The effect appears in helium, where electron correlation is strongest, and not in the valence electrons of argon or neon — direct evidence that the two electrons are acting together. Beyond extending coherent X-ray sources, the secondary plateau becomes a probe of attosecond-timescale electron–electron correlation, with a route toward molecules and correlated solids. Published in Nature Photonics, 7 August 2026.
Siyang Wang・Jieyu Yan・Alba de las Heras・Luis Plaja・Dimitar Popmintchev・Tenio Popmintchevら(カリフォルニア大学サンディエゴ校/ウィーン工科大学)は今回、この限界が本質的なものではないことを実験で示した。ヘリウムを駆動すると、120 eV超の従来カットオフを越えてコヒーレント放射を運ぶ弱い第2プラトーが現れ、ウォーターウィンドウ(280 eV)まで到達したのである。機構は強く相関した電子対の二電子同時再結合で、カットオフ則は教科書的な3.2倍ではなく最大でポンデロモーティブエネルギーの5.5倍に従う。この効果は電子相関が最も強いヘリウムで現れ、アルゴンやネオンの価電子では現れない——2個の電子が協調して働いている直接的な証拠だ。コヒーレントX線源の高エネルギー化にとどまらず、この第2プラトーはアト秒スケールの電子相関そのものを探るプローブとなり、分子や強相関固体への展開も視野に入る。Nature Photonics 2026年8月7日付掲載。
Push a metal into a strong enough magnetic field and its carriers collapse into the lowest Landau level — the quantum limit. Beyond it, the textbook Lifshitz–Kosevich description says the familiar magnetoresistance oscillations, periodic in 1/B, should simply stop. The pentatellurides have stubbornly refused to comply, showing log(B)-periodic and other anomalous oscillations that different groups have attributed to different many-body effects, with no unified picture.
C. Kaufmann Ribeiro, J. Larrea Jiménez, S. M. Thomas and J. C. Palmstrom and colleagues (Institute of Physics, University of São Paulo; Los Alamos National Laboratory; University of Washington) measured ZrTe5 single crystals down to 700 mK and in pulsed magnetic fields up to 60 tesla (work supported by Los Alamos National Laboratory and the National High Magnetic Field Laboratory). They find robust non-1/B oscillations that persist well past the quantum limit and whose temperature and field dependence violates the Lifshitz–Kosevich framework. The explanation is structural rather than correlation-driven: competition between the Zeeman energy (field coupling to spin) and the cyclotron energy (orbital motion) drives a nonlinear evolution and back-bending of the Landau levels, so that low-index levels re-cross the Fermi energy at high field and produce fresh oscillations where none should exist. Carrier density and Fermi-surface size then decide which regime a given sample shows, reconciling the conflicting reports across the pentatelluride family. Published in Nature Communications 17, 6728 (22 May 2026); press coverage 17–22 August 2026.
C. Kaufmann Ribeiro・J. Larrea Jiménez・S. M. Thomas・J. C. Palmstromら(サンパウロ大学物理学研究所、ロスアラモス国立研究所、ワシントン大学)は、ZrTe₅単結晶を700 mKまで冷やし、60テスラまでのパルス磁場中で測定した(研究はロスアラモス国立研究所と米国国立強磁場研究所の支援を受けている)。その結果、量子極限をはるかに超えても持続する1/B周期でない振動を確認し、その温度・磁場依存性がリフシッツ–コセビッチの枠組みを破ることを示した。原因は多体相関ではなくバンド構造にある。ゼーマンエネルギー(磁場とスピンの結合)とサイクロトロンエネルギー(軌道運動)の競合がランダウ準位を非線形に発展させて後方屈曲(back-bending)させ、低指数の準位が高磁場で再びフェルミ準位を横切る——本来なにも起きないはずの領域に新たな振動が生まれるのだ。どのレジームが見えるかはキャリア密度とフェルミ面の大きさで決まり、ペンタテルライド系で食い違っていた報告群を統一的に説明できる。Nature Communications 17, 6728(2026年5月22日付)掲載、報道は8月17〜22日。
Electron microscopes image the very small by counting electrons that pass through a sample. From a physicist’s point of view that is wasteful: each electron is a quantum object, and if all you do is count it, the quantum information it carries is thrown away. The waste matters most for fragile specimens — biological molecules above all — which are destroyed long before enough electrons have been counted to form a sharp image.
Elias Pescoller and colleagues at TU Wien, with teams at the University of Vienna, JKU Linz and the University of Innsbruck, propose and analyse a concrete way to keep that information: coherently coupling the free electrons of a transmission electron microscope to a trapped-ion quantum processor. In the proposed setup, an ion trap sits in a plane conjugate to the specimen plane; a passing electron couples through the Coulomb interaction to the centre-of-mass motion of a trapped 40Ca+ ion prepared in a superposition of coherent states, imprinting a relative phase that is read out on the ion qubit. Their analysis shows that single electrons can induce resolvable qubit excitations, enabling non-destructive, quantum-coherent detection and the accumulation of information across many electrons — the physics behind dose-efficient, quantum-enhanced electron microscopy. The authors note the scheme could also serve coherent state preparation and readout in quantum free-electron lasers and nanoscale electron accelerators. A quantum-computer electron microscope based on the idea is now being built at TU Wien. Published in Physical Review Letters (2026).
Elias Pescollerらウィーン工科大学のチームは、ウィーン大学・ヨハネス・ケプラー大学リンツ・インスブルック大学と共同で、その情報を活かす具体的な方式を理論的に提案・解析した。透過型電子顕微鏡(TEM)の自由電子をイオントラップ量子プロセッサにコヒーレントに結合させるのである。提案する構成では、試料面と共役な面にイオントラップを置く。通過する電子はクーロン相互作用によって、コヒーレント状態の重ね合わせに用意された⁴⁰Ca⁺イオンの重心運動と結合し、その相対位相をイオン量子ビット上に刻み込む。解析によれば電子1個でも分解可能な量子ビット励起を誘起でき、非破壊かつ量子コヒーレントな検出と、多数の電子にわたる情報の蓄積が可能になる——低線量で量子的に増強された電子顕微鏡を支える物理である。著者らは、量子自由電子レーザーやナノスケール電子加速器におけるコヒーレントな状態準備・読み出しにも応用できると指摘する。この構想に基づく「量子コンピュータ電子顕微鏡」は現在ウィーン工科大で建設が進んでいる。Physical Review Letters(2026年)掲載。
An ordinary crystal repeats in space; a time crystal repeats in time. In a continuous time crystal the rhythm is not imposed by a pulsed drive at all: the system is illuminated steadily and settles into an oscillation whose frequency it chooses itself.
The TU Dortmund group of Alex Greilich and Manfred Bayer had previously realized an unusually robust continuous time crystal in a semiconductor — gallium arsenide lightly doped with indium and silicon, cooled to roughly −270 °C, in which each localized electron spin talks to about a million nuclear spins; continuous optical pumping polarizes the electrons, that polarization is transferred to the nuclei, and feedback between the two sustains oscillations coherent for hours. In the new work they excite several such time crystals at once. Microscopic disorder gives each region a slightly different natural frequency, yet nearby crystals adjust until they share one frequency — the quantum echo of Huygens’ 1665 observation that two pendulum clocks on a shared beam fall into step. Here nothing mechanical connects them: the messenger appears to be spin-polarized electrons diffusing through the crystal, and synchronization survives out to about 40 micrometres, closely matching the electron spin diffusion length and more than a thousand times the size of a single oscillator. Beyond that range the oscillators go their own way. A synchronized ensemble can thus be read as one larger, composite time crystal, opening solid-state studies of collective dynamics and information transfer in networks of spin oscillators. Published in Nature Communications, 22 July 2026; press coverage 24 August 2026.
Keywords: time crystal, 時間結晶, continuous time crystal, 連続時間結晶, synchronization, 同期, Huygens, ホイヘンス, electron spin, 電子スピン
🎯 / 「探していた粒子」は見つからず、別の2つが現れた——ジェファーソン研究所GlueXが8.0〜11.6 GeVの光子ビームで陽子を叩き、γp→φ(1020)π⁺π⁻p反応の断面積を世界で初めて測定。ストレンジオニウム候補Y(2175)(現在はφ(2170)に改称)を光生成過程で初めて探索したが有意な信号は現れず、断面積に上限を設定。代わりにm(φπ⁺π⁻)=2.24 GeV/c²に約5シグマの構造を発見(GlueX国際共同実験、Phys. Rev. Lett. 136, 251902)
Since the early 2000s, accelerator experiments have turned up a growing crowd of hadrons whose quantum numbers do not sit comfortably in the 1964 quark model. Physicists lump them together as XYZ states: they may be hybrids containing excited gluons, tetraquarks, or loosely bound molecules of ordinary hadrons. One of them, Y(2175) — recently renamed φ(2170) by the Particle Data Group — was reported by BaBar in 2006 near 2.16 GeV in the strangeonium region (states built from a strange quark and its antiquark) and later seen by BES and Belle, but always through electron–positron annihilation and never through any other production mechanism.
The GlueX Collaboration in Experimental Hall D at the Thomas Jefferson National Accelerator Facility set out to change that. CEBAF’s electrons are converted by an ultrathin diamond wafer into an intense beam of spin-aligned high-energy photons, which strike protons in a liquid-hydrogen target millions of times per second — a photoproduction channel available at this intensity nowhere else. From 334 pb−1 of data taken with photon energies between 8.0 and 11.6 GeV, the team measured for the first time the cross section of the exclusive reaction γ+p → φ(1020)π+π−p, reconstructed through the final state K+K−π+π−p.
Using the resonance parameters quoted by the Particle Data Group, no evidence for Y(2175)/φ(2170) appeared, and the paper instead sets upper limits on its photoproduction cross section. What did appear was a structure at m(φπ+π−) = 2.24 GeV/c2 with a statistical significance of about 5σ, whose parameters differ from those of the state being sought; Jefferson Lab’s account of the work refers to it as Y(2240) and to a second, weaker structure near 1.82 GeV at 3σ as X(1830). Because photoproduction probes the strong force differently from e+e− annihilation, the result gives theorists a fresh handle on which exotic quark–gluon configurations nature actually realizes; the collaboration notes that a full partial-wave analysis of this five-body final state would need a larger data set. Published in Physical Review Letters136, 251902 (June 2026); Jefferson Lab feature 26 August 2026.
ジェファーソン研究所(Thomas Jefferson National Accelerator Facility)実験ホールDのGlueX国際共同実験はここに挑んだ。CEBAFの電子ビームを極薄のダイヤモンド薄膜で偏極のそろった高エネルギー光子ビームに変換し、液体水素標的中の陽子へ毎秒数百万回打ち込む——この強度の光生成チャンネルを持つ施設は世界に他にない。光子エネルギー8.0〜11.6 GeVで取得した334 pb−1のデータから、終状態K+K−π+π−pを再構成し、排他反応γ+p → φ(1020)π+π−pの断面積を世界で初めて測定した。
A quantum droplet is not held together the way a raindrop is. In ultracold gases, a mean-field attraction that would ordinarily collapse the cloud is arrested by a quantum effect — usually beyond-mean-field fluctuations — leaving a self-bound blob that survives with no trap at all. Droplets of this kind have been seen in dipolar gases and in Bose–Bose mixtures. Mixtures of bosons and fermions, however, were widely expected not to droplet in the strongly interacting regime, and existing theories could only describe them when the interactions were weak.
Sam Foster, Olivier Bleu, Jesper Levinsen and Meera M. Parish (Monash University, with the Institut für Theoretische Physik at Heidelberg) revisit the canonical problem of a Fermi gas immersed in a weakly repulsive Bose–Einstein condensate at zero temperature. They build a variational ansatz that incorporates pair correlations and reproduces the known polaron limits, which lets them map the phase diagram across the full range of boson–fermion interactions rather than only the weak-coupling corner. The result: near a resonant boson–fermion interaction, self-bound droplets appear for boson–fermion mass ratios in the vicinity of unity — a range accessible in existing cold-atom laboratories — with the attraction balanced by the fermions’ degeneracy pressure. Raising the fermion density instead drives phase separation between the mixture and excess fermions, together with behaviour reminiscent of a liquid–gas critical point. The prediction hands experimentalists a concrete new state of matter to hunt for. Published in Physical Review Letters 137, 073402 (14 August 2026).
Sam Foster・Olivier Bleu・Jesper Levinsen・Meera M. Parish(モナッシュ大学、ハイデルベルク大学理論物理学研究所と共同)は、絶対零度で弱く斥力的なボース・アインシュタイン凝縮体の中にフェルミ気体を浸すという古典的な問題を改めて解いた。彼らは対相関を取り込んだ変分アンザッツを構築し、既知のポーラロン極限を正しく再現させることで、弱結合の隅だけでなくボース・フェルミ相互作用の全領域にわたって相図を描くことに成功した。その結果、共鳴的なボース・フェルミ相互作用の近傍では、ボース粒子とフェルミ粒子の質量比が1に近い領域——現在の冷却原子実験で到達可能な範囲——で自己束縛した液滴が現れる。引力はフェルミ粒子の縮退圧と釣り合っている。一方、フェルミ粒子の密度を上げると、混合気体と余剰フェルミ粒子との相分離が起こり、液相・気相の臨界点を思わせる振る舞いも現れた。実験家にとっては、探しに行くべき新しい物質状態が具体的に示されたことになる。Physical Review Letters 137, 073402(2026年8月14日付)掲載。
Magnetism has traditionally come in two flavours: ferromagnets, whose spins align and produce a net moment, and antiferromagnets, whose spins cancel. Altermagnetism has recently been proposed as a third class — the net moment cancels as in an antiferromagnet, yet the electronic bands are spin-split in momentum space without relying on spin–orbit coupling. Ruthenium dioxide (RuO2) became the most-studied candidate and then the most contested one: muon-spin-rotation, neutron and photoemission studies on bulk crystals and thick, strain-relaxed films have converged on the conclusion that RuO2 shows no magnetic order.
A study in Science Advances led by Yichen Zhang and Seung Gyo Jeong, with corresponding authors Ming Yi (Rice University), Bharat Jalan (University of Minnesota) and Milan Radovic (Paul Scherrer Institute), argues that the ultrathin limit is a different regime. Below roughly 4 nm, RuO2 grown on TiO2(110) is fully clamped by the substrate and carries substantial epitaxial strain. Using spin- and angle-resolved photoemission spectroscopy supported by ab initio calculations, and two measurement geometries to separate intrinsic signal from artefacts, the team mapped the electronic structure of 2-nm-thick epitaxial RuO2 heterostructures.
They observe an unconventional spin texture in which mirror-even and mirror-odd momentum-dependent components coexist. A comprehensive symmetry analysis rules out nonmagnetic origins for it, pointing to an emergent non-relativistic spin structure enabled by epitaxial strain — a clear departure from relaxed or bulk RuO2. The authors frame this as strain-engineered magnetism in oxide heterostructures rather than as a settled confirmation of altermagnetism in RuO2; given how sharply that question has been debated, independent verification on other samples and with other probes still matters. Published in Science Advances12(31), 29 July 2026; Rice University coverage 27 August 2026.
Water coming to a boil and a magnet losing its magnetism look like completely unrelated events, yet close to their transition points both are described by the same handful of numbers. Physicists call this universality: the microscopic mess washes out, and only a few essential features survive. The mathematical machinery behind much of it is conformal field theory (CFT), which predicts, among other things, a characteristic spectrum of energy levels at a critical point — a fingerprint of the universality class. Testing that fingerprint directly is hard, because real materials at criticality are difficult to isolate and measure level by level.
The experimental group of Manuel Endres and the theory group of Jason Alicea at Caltech, working with researchers at Université Paris-Saclay and the Technical University of Munich, instead used quantum simulators — engineered quantum systems tuned to sit at a critical point — to investigate two different conformal field theories experimentally. Measuring the universal energy spectra of synthetic matter at criticality confirms decades-old predictions and, more usefully, establishes a method: a way to identify which CFT governs a quantum system by reading out its level structure directly, rather than inferring it from bulk exponents. Published in Nature, 19 August 2026; Caltech announcement 28 August 2026.
Diamond is not only a gemstone. It forms the capsule that holds the fuel in inertial confinement fusion experiments, and planetary scientists believe it rains through the interiors of ice giants such as Neptune and Uranus. Both settings demand knowing how carbon behaves at extreme pressure — and there, for two decades, laboratory measurements and quantum-mechanical simulations disagreed about diamond’s melting temperature by up to 20%, a discrepancy worth more than 1,000 kelvin. No amount of theoretical refinement closed it.
Marius Millot, Federica Coppari, Jon H. Eggert and colleagues at Lawrence Livermore National Laboratory, using laser-driven dynamic compression with improved diagnostics developed with the University of Rochester’s Laboratory for Laser Energetics, shock-compressed microcrystalline diamond and simultaneously measured its atomic structure (X-ray diffraction), temperature (pyrometry), density (velocimetry) and optical reflectivity at pressures around 1 TPa — roughly ten million atmospheres, about three times the pressure at Earth’s centre, at temperatures above the surface of the Sun. The revised melting temperature, near 7,300 K, lands more than a thousand degrees below the influential earlier experimental value and close to what simulations had predicted; the older numbers, Eggert notes, were simply off. The data also show that under a single shock the sample does not have time to pass through the predicted intermediate BC8 crystalline phase on its way to liquid. Applied to fusion capsule design, the corrected melting curve suggests up to threefold greater energy gain. Published in Nature Physics, 13 August 2026; LLNL release the same day, with further coverage 29 August.
Black holes need not come from dying stars. General relativity also allows microscopic black holes to form when a concentration of energy is tuned to sit exactly at the threshold between dispersing and collapsing — the domain of critical collapse. In 1993 Matthew Choptuik discovered numerically that the critical state has a startling property: discrete self-similarity, in which the same pattern of spacetime curvature repeats at ever smaller scales, like a crystal whose lattice runs in space and time. Since then this “spacetime crystal” has been studied almost entirely through demanding numerical simulations.
Christian Ecker (Goethe University Frankfurt) with Florian Ecker and Daniel Grumiller (TU Wien) found an analytic route. Their trick is the large-D expansion: solving the Einstein–massless–Klein–Gordon system — gravity coupled to a massless scalar field — in the regime of a large number of spacetime dimensions D, where the equations simplify enough to be handled analytically. They construct an infinite family of discretely self-similar solutions in closed analytic form, characterize their structure, and compare them against numerical critical solutions at finite D, identifying both universal features and behaviour peculiar to large D. As Grumiller describes it, the crystal is an unstable intermediate: nudge the energy slightly one way and it dissolves back into ordinary spacetime filled with freely streaming particles; nudge it the other way and it collapses into a tiny black hole. Analytic control over this threshold gives a new handle on primordial and microscopic black hole formation without relying wholly on numerics. Published in Physical Review Letters (12 May 2026); featured coverage 30 August 2026.
Quantum simulators are built to reproduce the behaviour of quantum systems too complex for classical computers. That is also their weak point: once a simulator passes the size where a classical machine can check it, how does anyone know the answer is right? “No real experiment is perfect,” as Tristan Kraft puts it — interactions differ from design, the environment intrudes, measurements carry uncertainty.
A collaboration led by Tristan Kraft (TU Munich) and Peter Zoller (University of Innsbruck and IQOQI, Austrian Academy of Sciences), with Barbara Kraus (TU Munich) and an experimental team under Manoj Joshi and Christian Roos, stops assuming the device works as designed and instead measures how it actually behaves. From the experimental data they learn the relevant Hamiltonian (the coherent interactions) and Lindbladian (the noise and dissipation), then propagate the uncertainties of that learned model through to the simulation output. The simulator therefore returns not a bare number but a result with quantified error bounds. The method was first validated on ten ions, small enough for classical computation to serve as an independent check, then extended to a chain of 51 ions. Zoller notes a longer-term consequence: quantum advantage could be judged not only on speed or size but on which machine solves a problem with the smaller verifiable error. Two-dimensional simulators are the next target. Published in Physical Review X, 13 August 2026; Innsbruck release 31 August 2026.
Thought experiments have shaped quantum mechanics: Schrödinger’s cat, the double slit. In 1948 Richard Feynman added another, and it became a cornerstone of modern physics. In his path integral formulation, a quantum particle travelling from A to B does not follow one trajectory; every conceivable path contributes a probability amplitude, and the amplitudes are summed with phases that mostly cancel, leaving the familiar behaviour. The formulation underpins quantum field theory and statistical mechanics — yet the two postulates Feynman built it on had never been tested directly.
A team led by Shi-Liang Zhu at South China Normal University in Guangzhou, with first author Yong-Li Wen, closed that gap using single photons. They first developed a rigorous propagator-based approach that turns the postulates into something an experiment can address, then performed comprehensive measurements of a single photon’s probability amplitudes across more than 1.4 million paths (175 discretized trajectories), reaching high fidelity in the measured propagators and enabling complete reconstruction of the path probability amplitudes. The postulates held. Beyond confirming a foundation that decades of calculations have quietly relied on, the technique is portable: the group hopes it can be adapted to other physical systems — for instance, to test how paths combine when photons travel through a material rather than empty space. Published in Science Advances; coverage 31 August 2026.
In 1879 Edwin Hall showed that a magnetic field applied perpendicular to a current-carrying material deflects the moving charges and produces a measurable transverse voltage. The Hall effect and its magnetic cousin, the anomalous Hall effect (AHE), became workhorses for characterizing materials and are built into sensors in cars, keyboards and medical instruments. Symmetry, however, imposes a rule: the anomalous Hall response is proportional to the out-of-plane component of the magnetization. An in-plane version had been proposed theoretically but never demonstrated, because it requires a magnetic material with just the right (very low) symmetry.
Researchers in the Lab for Investigating Quantum Materials, Interfaces and Devices at Carnegie Mellon University engineered one. Simranjeet Singh and Jyoti Katoch built atomically precise heterostructures pairing the low-symmetry topological semimetal TaIrTe4 with the layered ferromagnetic insulator Cr2Ge2Te6 (CGT). Interfacing them leaves only a single mirror plane intact; once the magnetization acquires a component within that plane, even that symmetry is broken, and a Hall response appears that depends on both in-plane and out-of-plane magnetization. Measurements across multiple devices show the effect is gate-tunable, i.e. electrostatically controllable. Shubhayu Chatterjee’s minimal symmetry-constrained model attributes it to interfacial spin–orbit coupling and exchange interaction that switch on when CGT becomes ferromagnetic at low temperature. Practically, as Singh notes, one such sensor could read magnetic field along two directions where two were needed before. Published in Nature Materials; CMU announcement 19 August 2026, wider coverage 31 August.
Keywords: in-plane anomalous Hall effect, 面内異常ホール効果, anomalous Hall effect, 異常ホール効果, Hall effect, ホール効果, TaIrTe4, Cr2Ge2Te6, CGT, topological semimetal
❄️ / 「1つ」に見えた超伝導ギャップは、実は激しく混ざり合う「2つ」だった——数層のNbSe₂とTaS₂のトンネル分光データが、バンド間散乱がギャップより大きい領域のマクミラン2バンド模型でよく再現されると判明。2つのギャップは遷移金属由来のΓ点・K点フェルミ面に乗り、磁場による対破壊も同じ理論で説明できる。バルクの2H-NbSe₂は3バンド超伝導体である可能性(Simon・Klang・Millo・Steinberg、エルサレム・ヘブライ大学、Phys. Rev. Lett. 137(1)掲載)
A superconductor is usually characterized by its energy gap — the energy needed to break a Cooper pair. Some materials have more than one, because different electronic bands become superconducting with different gap sizes; bulk magnesium diboride is the textbook two-band case. The ultrathin transition-metal dichalcogenides NbSe2 and TaS2, heavily studied as platforms for two-dimensional superconductivity, appeared by contrast to be simple single-gap superconductors — a picture that never quite fitted all the data.
Shahar Simon, Maya Klang, Oded Millo and Hadar Steinberg at the Hebrew University of Jerusalem took tunneling data on thin exfoliated samples of both materials. The spectra turn out to match the McMillan two-band model very well — provided the interband scattering parameters are large compared with the gaps themselves. That is precisely the regime in which two gaps are smeared into what looks like one, which is why the single-gap reading survived so long even though its detailed shape resisted single-band theories. The authors place the two heavily cross-scattering gaps on the Γ and K Fermi surfaces derived from the transition-metal atoms, and show that pair breaking under magnetic field is described by the same two-band theory. A further implication: at bulk thickness, 2H-NbSe2 is likely a three-band superconductor. The number of bands and the strength of the scattering between them govern how a superconductor responds to field, disorder and reduced dimensionality, which matters for anyone engineering these layers into Josephson junctions or superconducting qubits. Published in Physical Review Letters137(1) (29 June 2026); Hebrew University coverage 31 August 2026.