Research articles

The study of frustrated magnet systems has unveiled a range of novel physical phenomena and continues to attract interest for in fields such as quantum spin liquid theory and high-temperature superconductors. Here, the authors use ab-initio calculations and a spin-wave analysis to demonstrate that an order-from-disorder phenomenon contributes to the columnar antiferromagnet ordering of BaCoS2.

Narrow-gap semiconductors with gate-controllable spin-splitting provide an ideal platform for novel spintronic and topological devices. The authors observe a large spontaneous spin-splitting energy, reaching 18 meV and widely tunable by a gate voltage, in an InAs quantum well that is magnetically proximitized by a ferromagnetic semiconductor (Ga,Fe)Sb.

Migdal-Eliashberg theory is a method for describing conventional superconductors. Here, the authors present an implementation that goes beyond the widely used constant density of states approximation by accommodating scattering processes beyond the Fermi surface, and they show its importance in two classes of near room temperature superhydrides.

Impedance theory grants insight to design metasurfaces for controlling acoustic fields, but such theory imposes great limitation on boundary conditions. The authors propose a generalized acoustic impedance theory connecting arbitrarily conservative acoustic fields, and design a beam splitter as an example of power flow processing.

A Chern insulator has non-trivial bulk topology with a quantized Chern number defined in the Brillouin zone (BZ), yielding robust gapless edge states. The authors introduce the concept of reduced Chern number, defined in subregions of the BZ, and construct a family of Chern dartboard insulators with quantized reduced Chern numbers but trivial bulk topology, exhibiting distinct pseudospin textures.

The turbulent skin friction drag at the solid/liquid interface results in high electric energy consumption when conveying liquids through hydraulic networks, and efficient drag reduction strategies are still unavailable. The authors coat a channel with a magnetic fluid film and achieve almost complete wall drag reduction of up to 90% across laminar and turbulent flow regimes.

Most physics-informed deep learning models alleviate the poor generalizability of pure data-driven models by minimizing residuals of the governing equations. In this work the authors propose to leverage physics priors, by embedding the discretized governing equations into the neural network architecture, which improves the generalizability and long-term prediction accuracy.

Reservoir computing (RC) is energy-efficient due to its simple architecture, allowing physical implementation. The proposed self-modulated RC extends the capabilities of RC, exhibiting superior learning abilities and complex dynamics, including attention and chaos, while retaining the advantage of physical implementation.

One main obstacle of flow cytometry techniques is the inability to image internal structure of live cells on the go, posing challenges in deciphering their biological mechanisms. To overcome this limit, the authors devise a light-sheet-based multichannel, multisheet and multicolor volume imaging cytometry for interrogating cells flowing simultaneously through microfluidic channels.

Understanding the thermal transport properties of tungsten nitrides formed on the divertor surface of the tokamak is crucial, as they will be subjected to continuous heat flux. In this article, the authors theoretically calculated the influence of vacancy defects on the electrical and thermal conductivities of tungsten nitrides, providing an understanding of the mechanism behind the effects in terms of electron behavior.

Generating ultrashort spectrally tunable pulses via high-harmonic generation (HHG) enables matching the excitation photon energy to the characteristic resonance of the sample to study its ultrafast dynamics. The authors demonstrate a compact continuously tunable high-intensity VUV HHG source that can rival state-of-the-art seeded FELs.

Harmonics produced when structured lasers hit isotropic targets inherit the topological properties of the driver, but such properties become more complex in the case of anisotropic nonlinear targets. The authors exploit topological analysis of the high-order harmonics to spatially resolve the information about the nonlinear response of the target.

Twisted bilayer systems have become a popular tool to investigate strongly correlated phenomenon but such a setup can also be relevant on the macroscopic scale. Here, the authors investigate the transport of macroscopic magnetic particles between magnetic square patterns and consider the impact the twist angle and the impact of disorder exerts on the observed behaviour.

Understanding the material strength of heterogeneously wetted granular systems is challenging due to the presence of a liquid modifying the interaction between the grains. Here, the authors investigate the mechanical properties of grains wetted with silicon oil finding an increase in Young’s modulus and an increased resistance to compressive strains.

Discovering nonlinear differential equations from empirical data is a significant challenge, often requiring manual parameter tuning. This paper introduces a machine learning method integrating denoising techniques, sparse regression, and bootstrap confidence intervals, which shows consistent accuracy in identifying 3D dynamical systems with moderate data size and high signal quality.

A key problem in non-Hermitian lattice physics is restoring bulk-boundary correspondence. Here, the authors study two non-Hermitian models with a generalized boundary condition, proving that these systems can continuously be tuned from a periodic to an open boundary condition, showing a phase transition characterized by an exceptional point.

Surrogate networks are synthetic alternatives to real world networks that avoid expensive data collection and privacy issues, but they often lack information on the temporal or topological properties of the input network. The authors propose a method to construct realistic surrogate network, outperforming the existing ones in accuracy and execution time.

Quantum many-body systems may not thermalize due to the phenomenon of many-body localisation. Its theoretical underpinning is given by observables, the l-bits, which could not as of now be probed by experiments. The authors define experimentally relevant quantities to retrieve spatially resolved entanglement information, allowing to probe the l-bits.

Key players are a small number of vertices that govern the structure and function of a complex network, and their identification can grant answers in many disciplines. By tapping into quantum information, the authors introduce an entanglement-based metric capable of quantifying the perturbations caused by individual vertices on spectral entropy.

High-intensity vortex and vector laser pulses can enable a wide range of applications; however, their generation remains challenging. The manuscript demonstrates that the strongly-coupled stimulated Brillouin scattering in plasma can be used as a promising amplification technique for both vortex and vector laser pulses, and thus paves the way for novel optical devices.