Papers

We study how unsupervised autoencoders trained on microscopic spin configurations from the Ising model learn macroscopic, theory-relevant variables underlying the data-generating process. We quantify learning across multiple spatial (coarse-graining) scales and reveal two distinct dynamical regimes that appear sequentially, controlled by the main hyperparameters (model depth, width, and learning rate): one in which magnetization and another in which energy is learned across scales. The first exhibits error fluctuations ordered to scale and learns global averages only; The second gradually resolves smaller scales relevant for the energy representation. Deep models trained at moderate and fast rates become arrested before reaching these regimes. We connect reconstruction errors with the latent representations using a novel analysis of self-recursive trajectories. These intrinsic dynamics are induced by prediction errors, exposing how training drives representation changes for macroscopic concepts. We utilize the intuition that learning operates as a process driven far from equilibrium by fluctuations from the training data to provide an interpretive basis grounded in both the physical world and the machine models that represent it.

Defect lines in 3D active nematic systems are intriguing topological singularities whose out-of-equilibrium dynamics remain elusive in confined settings. Here, we numerically study 3D active nematics confined within closed cylinders to elucidate the roles of geometry and activity. We reveal a competition between passive elasticity, which causes localisation of defects near edges, and activity, which endows defects with motility and gives rise to disorderly, delocalised dynamics. Varying boundary curvature, activity strength, and cylinder radius reveals a state space of static and dynamic localisation states, including handle-like configurations and chaotic motion bounded within the cylinder endcap. As activity is tuned to induce delocalisation, we identify phase transition signatures, including pronounced fluctuations and an emergent power law scaling of defect number and average defect length. We find that these scaling properties are strongly altered by confinement: unlike in bulk systems where activity governs length distributions, confinement tunes an activity-independent characteristic length, with an exponent reminiscent of self-avoiding confined polymers. These results establish confinement of inhomogeneous curvature as a versatile mechanism for controlling active topological dynamics.

Anharmonic lattice dynamics shapes phonon spectra and thermal transport, yet first-principles linewidth calculations typically assume phonon quasiparticles with sharply defined frequencies. Here, we introduce a self-consistent spectral function framework that determines interacting phonon spectral functions by dressing the internal lines of the three-phonon bubble self-energy with the spectra generated by the same interaction. The method self-consistently determines the real and imaginary parts of the self-energy on equal footing, eliminates the need for externally chosen smearing parameters to enforce energy conservation, and extends lattice dynamics and thermal transport calculations beyond the quasiparticle approximation. Applied to zincblende HgTe, self-consistency substantially broadens acoustic and optical phonon spectra, activates scattering channels inaccessible in the one-shot quasiparticle calculation, and reduces the lattice thermal conductivity by approximately a factor of five to the experimental scale while reproducing its temperature dependence, without explicitly invoking higher-order interactions. Mode-resolved spectral functions are further validated against the power spectra from direct molecular dynamics simulations. These results establish self-consistent spectral functions as a practical bridge between quasiparticle perturbation theory and fully dynamical simulations, and provide a framework readily generalizable to higher-order anharmonic processes.

We study vacancy-mediated hopping in a two-dimensional active Brownian-particle crystal with a single vacancy. Under active driving, the hopping rates are not organized by the free-particle effective noise amplitude alone. In contrast, in the passive limit, the hopping rate follows an Arrhenius-like activated trend with respect to the translational diffusion coefficient. Effective-noise comparisons show systematic dependence on the active fraction, indicating that the composition of thermal and persistent active fluctuations affects the hopping kinetics. These results demonstrate a breakdown of an effective-temperature description for microscopic defect-mediated transport in an active crystal.

We investigate the transverse response under the out-of-plane magnetic field for magnetic heterostructures with Cu/Bi2O3 or Ag/Bi2O3 Rashba interfaces. We detect opposite contributions on the transverse resistance by the Cu/Bi2O3 and the Ag/Bi2O3 interfaces, which interestingly coincide well with the opposite signs of the spin/charge interconversion from the two interfaces. We suppose the opposite influences on the transverse resistance feature spin swapping occurrence of the converted spin current. The transverse spin flow emerges due to the spin swapping in both Cu and Ag layer, but the direction of the spin flow is opposed dependent on the spin direction of the converted spin current.

The emergence of chirality in the charge density wave (CDW) phase of 1$T$-TiSe$_2$ has recently attracted significant attention, yet its microscopic origin remains debated. The prevailing interpretation attributes chirality to a relative phase shift between the charge density components. Here, using density functional theory and symmetry analysis, it is demonstrated that such phase-shifted states entail prohibitive energy costs, rendering them unlikely candidates for the ground state. Instead, a nearly degenerate metastable CDW phase stacking order with monoclinic symmetry is identified, that naturally breaks mirror and inversion symmetries. The CDW in 1$T$-TiSe$_2$ possesses a dual physical nature: the electronic charge redistribution is intrinsically bond-modulated, while the periodic lattice distortion is atom-centered transverse. This spatial separation allows a specific optical coupling mechanism where circularly polarized light breaks the energetic degeneracy between chiral domains. These results reconcile conflicting experimental observations, identifying the "spontaneous" chirality observed in local probes as a statistical distribution of CDW phase stacking domains, while establishing the mechanism for macroscopic, light-induced gyrotropic order.

Anyons in $2+1$ dimensions are not only characterized by exotic braiding statistics but also by intricate fusion properties. Two anyons may fuse into multiple topological charge sectors, and associativity of fusing three anyons to produce a fixed charge sector is governed by $F$-symbols. While braiding invariants, such as the modular data, are typically complex valued, a complete description of general anyon models requires understanding the arithmetic properties of its fusion associativity data as well. The $F$-symbols for many of the most common $2+1$d topological orders, including all Abelian anyon models as well as Fibonacci and Ising anyons, can be made real valued. We show this phenomenon is not universal by exhibiting braided fusion categories whose $F$-symbols cannot be made real. We call such $F$-symbols "inherently complex." The examples we study lack a charge-conjugation symmetry and our results are therefore consistent with the converse of a statement proved in a companion work linking real $F$-symbols in braided fusion categories with the existence of a suitable charge-conjugation symmetry. We analyse the smallest-rank braided fusion categories we know of with inherently complex $F$-symbols: ${\rm Rep}(\mathbb{Z}_7\rtimes\mathbb{Z}_3)$ and ${\rm Rep}(\mathbb{Z}_5\rtimes\mathbb{Z}_4)$. Consequently, the corresponding $\mathcal Z({\rm Rep}(\mathbb{Z}_7\rtimes\mathbb{Z}_3))$ and $\mathcal Z({\rm Rep}(\mathbb{Z}_5\rtimes\mathbb{Z}_4))$ anyon models also have inherently complex $F$-symbols. Our presentation connects these examples with recent results on classifying anyons beyond modular data.

We theoretically investigate the influence of hexagonal boron nitride (h-BN) on the electronic properties of an $\alpha$-$\text{T}_3$ lattice driven by an off-resonant elliptically polarized light field. The staggered potential $M$ breaks the sublattice inversion symmetry, transforming the initial semimetal into a trivial insulator with Chern number $C = 0$. We identify a fundamental geometric singularity at $\alpha = 1/\sqrt{2}$, independent of $M$, where the valley-resolved lower-gap threshold diverges, bounding a finite topological window where conduction–flat band inversion yields a Chern insulator with $C = 1$ carried by the flat band. Increasing the drive further closes the lower gap at the $K'$ valley, transferring the index to the valence band so that the flat band becomes trivial while the system remains $C = 1$. For $\alpha > 1/\sqrt{2}$ the lower gap closes at finite intensity, allowing a transition to $C = 2$ as the dice limit ($\alpha = 1$) is approached. The topological phases are characterized by quantized anomalous Hall plateaus at $\sigma_{xy} = e^2/h$ ($C = 1$) and $\sigma_{xy} = 2e^2/h$ ($C = 2$). The $C = 1$ plateau sits in a narrow gap and is the most fragile, while the $C = 2$ plateau is protected by a wider gap and remains robust to room temperature. A highly asymmetric thermoelectric Seebeck response further serves as an experimental fingerprint of each phase, providing a realistic framework for realizing stable high-Chern-number phases in substrate-supported $\alpha$-$\text{T}_3$ materials.

Quasicrystals emerge from competing interactions with incommensurate characteristic length scales that inhibit translational periodicity. Here, we show that such multiscale interactions can be realized in a dipolar Bose-Einstein condensate through the interplay between intrinsic dipole-dipole interactions and photon-mediated interactions generated by coupling to an excited-state manifold together with a suitably engineered optical feedback. This interplay gives rise to two pronounced roton instabilities in the Bogoliubov excitation spectrum, leading to a rich ground-state phase diagram including dodecagonal quasicrystals with twelvefold rotational symmetry for experimentally realistic parameters. We further propose a protocol to access these quasicrystals dynamically and develop a general numerical framework for calculating their collective excitation spectra.

Superlattices with twisted configurations, such as moire lattices, have recently been extensively exploited for their unique electronic, magnetic, and optical properties. One remarkable feature of nanoscale twisted superlattices is the distinct lattice symmetries and the continuous phase transitions between periodic or aperiodic phases, representing a unique opportunity to study many emerging physical phenomena. Here, we report a correlated light and matter interaction between the collective polarization effect of nematic plasmonic superstructures and the plasmonic excitation of individual constituent nanorods in reconfigurable twisted plasmonic superlattices. Using hybrid Fe3O4 and Au nanorods as building blocks, we assembled plasmonic nematic liquid crystals with unidirectionally aligned nanorods, which could be further assembled into moire plasmonic lattices through a vertical stacking assembly method. A twist angle dependent plasmonic excitation is recognized in the twisted bilayer of two plasmonic superlattices, featuring enhanced transverse and longitudinal plasmonic excitation at a twisting angle of 0 degree and 90 degree, respectively. Such correlated plasmonic excitation in twisted plasmonic superstructures is induced by the correlation between the collective polarization effect of the liquid crystal phases and the anisotropic plasmonic excitation of individual nanorods. The magnetic orientation control allows for precise alignment of hybrid Fe3O4 and Au nanorods in polymer substrates and enables the coding of nematic domains and plasmonic patterns in each sublattice. The correlated plasmonic excitation and light polarization create reconfigurable photonic moire superlattices with well-defined domain colors, feature sizes, periodicities, symmetries, and dimensions determined by twist angles and displacements in the twisted plasmonic lattices.

We study a class of diffusion processes on $\mathbb{R}^n$ interacting through a symmetric matrix $X\in\mathbb{R}^{n\times n}$. When eigenvectors of $X$ are Haar-uniform on the orthogonal group, we derive a dynamical mean-field limit for the empirical law of sample paths, extending the classical Sompolinsky–Zippelius characterization for $X\sim\mathrm{GOE}$. The limit takes the form of a generalized Langevin equation with correlated Gaussian noise and memory, whose correlation and response kernels relate to those of the original dynamics through convolution equations involving the free cumulants of the eigenvalue distribution of $X$. For the overdamped Langevin diffusion associated with $\mu(\boldsymbol{\theta})\propto \exp\!\big(\frac12\boldsymbol{\theta}^{\top}X\boldsymbol{\theta}\big)\prod_{i=1}^n\nu(\mathrm{d}\theta_i)$, we analyze the mean-field limit under a rapid-mixing assumption. The correlation and response kernels admit time-translation-invariant approximants satisfying a fluctuation-dissipation relation. The generalized Langevin equation admits a Markovian approximation coupled to an auxiliary multivariate OU process and converges to a replica-symmetric prediction for the empirical coordinate law under $\mu$. This auxiliary correlation structure is characterized through the infinitesimal generator of a Markov semigroup for the lifted path-history process. Consequently, the free energy converges to a replica-symmetric limit under an explicit high-temperature condition, which for an Ising model is $\|X\|_{\mathrm{op}}<1/2$. By recent dynamical universality results, the same free-energy characterization holds for deterministic models without random disorder when $X$ satisfies a set of deterministic delocalization conditions.

Nonrelativistic spin splitting (NRSS) in collinear antiferromagnets, including altermagnets and compensated ferrimagnets, offers a route to high-frequency spintronics immune to stray fields, but its dynamic control has remained elusive. We demonstrate, using density functional theory (DFT) and nonlinear phononics, that THz laser pulses can achieve ultrafast, reversible control of NRSS on picosecond timescales in antiferromagnets. We derive two mode-selection criteria, accounting for phonon and magnetic wavevector compatibility and inversion parity, to identify which phonon modes can activate NRSS. Our calculations illustrate the creation of NRSS in spin-degenerate NiO and the enhancement of existing splitting in LaFeO$_3$. While practical implementation depends on excitation conditions, lattice dynamics, and mode coupling, these results establish design principles for identifying materials with stronger and more accessible responses. Limitations imposed by driving fields, phonon damping, and anharmonic detuning motivate further materials screening based on accessible excitation strengths, lattice stability, resonance frequency matching, and favorable anharmonic coupling.

We introduce a unified entropic framework for real fluids that encompasses the van der Waals, Berthelot, Redlich Kwong, and Peng Robinson equations of state within a common thermodynamic description. The corresponding microscopic interactions are then explored using Geometrothermodynamics, GTD, through the scalar curvature $mathcal{R}$ of the equilibrium manifold. We show that curvature singularities accurately reproduce macroscopic critical phenomena, while vanishing curvature $\mathcal{R}=0$ identifies specific thermodynamic states where attractive and repulsive intermolecular forces effectively balance. Furthermore, we introduce a set of dimensionless critical-amplitude ratios $Q^i_{j}$, which reveal universal geometric features of the critical regime. Although individual critical amplitudes exhibit a logarithmic dependence on the system size, these invariant ratios organize different molecular species according to the strength of criticality and encode universal scaling features, suggesting their potential as robust classification parameters. Finally, employing Bayesian inference and Markov Chain Monte Carlo, MCMC methods, we statistically reconstruct the zero-curvature curves. The posterior distributions support the consistency of the geometric scaling behavior, demonstrating that the GTD manifold encodes non-trivial information about the underlying thermodynamical models.

Bismuth monolayer is the first two-dimensional elemental ferroelectric and an appealing platform for coupling polar order to spin-orbit-driven topology. However, its microscopic switching mechanism remains elusive. Here, using first-principles lattice dynamics and symmetry-adapted mode analysis, we identify a previously overlooked rotational pathway for in-plane polarization switching. Its energy barrier is more than four times lower than that of direct reversal, naturally explaining the vortexlike domain textures observed in molecular dynamics simulations. Remarkably, this polarization rotation also drives a spin-topological transition, changing the spin Chern number from $C_s=-2$ to $0$. Directional uniaxial strain further steers the polarization orientation and tunes the associated topological transition. These results establish polarization rotation as the switching mechanism of ferroelectric Bi monolayer and as an efficient route to electrically and mechanically programmable topology in two-dimensional ferroelectrics.

Two-dimensional topological insulators (2D TIs) have emerged as a cornerstone of next-generation spintronic technologies due to their robust, dissipationless edge states protected by time-reversal symmetry. Initial realizations of 2D TIs have primarily focused on materials with strong intrinsic spin-orbit coupling capable of driving band inversion, an approach that significantly constrains the accessible materials landscape. More recently, a paradigm shift has occurred toward engineering topological phases in van der Waals (vdW) heterostructures, where nontrivial band topology can arise from interfacial coupling rather than relying solely on intrinsic material properties. This framework provides an exceptionally versatile platform with multiple tunable degrees of freedom, including stacking configuration, twist angle, and chemical functionalization, allowing systematic manipulation of the band topology. Furthermore, external stimuli, such as electric fields, strain, and light-matter coupling, enable dynamic and reversible control of the topological character. The combined use of vdW interface engineering and external modulation allows the realization of 2D TI phases even in otherwise topologically trivial systems, substantially expanding the accessible materials landscape. This Research Update reviews key milestones in the development of vdW-engineered 2D topological quantum materials, critically assesses outstanding theoretical and experimental challenges, and outlines promising directions for future breakthroughs.

Epoxy-based materials are inherently hygroscopic, absorbing moisture from the environment, which can significantly alter their short and long-term performance. The presence of graphene is often considered as a potential candidate to act as a microscopic barrier, mitigating the adverse effects of hydration on the matrix. This study investigates the impact of hydration on the glass transition and elastic mechanical properties of epoxy resins and their graphene nanocomposites, focusing on water content up to 5 %wt. Using large-ensemble molecular dynamics simulations, we analyze the temperature-driven glass transition and mechanical response of both neat epoxy and epoxy-graphene systems under varying hydration levels. Our results reveal a distinct threshold at 3 %wt water content: below this, hydration primarily reduces the glass transition temperature, while mechanical properties remain unaffected. Beyond 3 %wt, however, the mechanical properties deteriorate, highlighting a non-linear sensitivity to water uptake. Furthermore, we emphasize the critical role of ensemble size in ensuring the reliability of molecular dynamics predictions for such heterogeneous systems. Our simulations demonstrate that ensembles substantially larger than current state-of-the-art standards are necessary to achieve converged distributions of the predicted mechanical properties, particularly in highly heterogeneous hydrated epoxy-graphene nanocomposites. These findings provide novel insights into the hydration behavior of epoxy-based materials and underscore the potential of graphene to enhance their environmental resistance. This work also advances the understanding of structure-property relationships in polymer nanocomposites, offering guidance for the design of more robust materials in humid environments.

Thermal rectification (TR) typically relies on structural asymmetry or material heterogeneity. Here, we show that gravity alone can induce and modulate TR in gaseous systems. Using a minimal model of a single gas particle confined in a two-dimensional channel between heat baths at different temperatures, we analytically demonstrate that gravitational fields generate TR, enabling perfect unidirectional heat conduction across broad gravitational parameter ranges. This effect exhibits an intrinsic trade-off between rectification efficiency and heat power. Extending beyond the single-particle limit, numerical simulations confirm that gravitationally induced TR persists in interacting many-particle systems. Notably, in interacting gas mixtures, the rectification direction can be reversed. Gravity-mediated control of heat currents thus provides new fundamental insights into thermal transport and suggests design principles for gaseous thermal diodes.

We present a quantitative framework for predicting how kinetic barriers governing temperature dependent relaxation arise in finite square magnetic networks. By formulating a series of non homogeneous transfer matrices from the adjacency spectrum of the underlying graph, we show how local coordination shapes the energy landscape and identifies geometric regions that act as bottlenecks for flux transport. Our approach predicts bimodal kinetic behavior, in which long intervals of trapping within charge compensated manifolds are interrupted by sudden, avalanche like relaxation episodes. Representing finite systems with a Husimi tree, we find that low temperature data from forty generations collapse onto a single curve consistent with a power law scaling form, implying that boundary truncation alone can give rise to scale invariant flux arrest, analogous to athermal granular jamming. By employing transition probabilities constructed from local Boltzmann factors, this framework connects equilibrium energy landscape concepts to non equilibrium phenomena, including kinetic arrest and telegraph noise, thereby enabling the prediction and design of intermittent transport in finite, frustrated networks.

Two-dimensional magnetic materials offer unique opportunities for exploring low-dimensional spin phenomena and next-generation spintronic devices. Chromium trihalides CrX3 (X = Cl, Br, I) belong to an important family of these materials, where alloying opens pathways for tailoring their electronic, magnetic, optical properties, and thermodynamic stability. In this work, we present a density functional theory study of CrX3 compounds and their ternary alloys. Our results show that for the pure compounds, the ground state is ferromagnetic (FM), with the antiferromagnetic-zigzag (AFM-Z) and paramagnetic (PM) phases being close in energy. For these pure systems, the band gap variation among different magnetic phases does not exceed 0.16 eV, and the average magnetic moments on Cr atoms increase from Cl to Br to I. For the alloys, the FM state remains the lowest-energy configuration, but the energy difference towards the AFM-Z phase decreases for compounds with lower iodine concentration. The calculated band gaps reveal a pronounced bowing along the compositional edge connecting CrCl3 and CrI3. The Curie temperatures show a smooth variation across compositions, consistent with the nearly linear behavior of the magnetic exchange parameters. Based on the calculated mixing enthalpy and configurational entropy, the approximate Gibbs free energy indicates that alloy formation becomes thermodynamically favorable at finite temperatures, which is important to overcome the intrinsic experimental instability of these compounds.

We investigate the magnetic-field response of interstitial anionic electrons (IAEs) in two-dimensional electrides, using monolayer Ca$_2$N as a prototypical system. By computing the Landau-level (LL) spectrum of the electride bands forming the Fermi surface, we find a linear LL evolution with magnetic field that closely resembles the behavior of a nearly-free 2D electron gas (2DEG). The extracted cyclotron effective mass and Landé g-factor deviate moderately from their free-electron values, indicating that the IAEs retain a remarkably free-electron-like character. Furthermore, the energy dispersion of the electride bands remains insensitive to the choice of exchange-correlation functional (LDA vs.~PBEsol), indicating that local exchange and correlation effects have minimal influence on the IAEs. Overall, our findings provide fundamental insight into the quantum nature of electrides and open new avenues for exploring magnetic confinement, correlation effects, and emergent quantum phenomena in low-dimensional interstitial electronic systems.

The interplay between nonlinear optical transitions and topological band structure offers a route to control photocurrents. We reveal a fourth-order optoelectronic response that emerges due to an interlink between the circular photogalvanic effect (CPGE) and the Berry curvature dipole (BCD) in noncentrosymmetric 2D materials. Using monolayer $\Td$-WTe$_2$ as a prototype, we predict that circularly polarized mid-infrared light produces a steady dc injection current that induces an internal electric field, which in turn drives a transverse nonlinear Hall response through BCD. The resulting cascaded photovoltage scales as the fourth power of the optical field $E_0^4$. By mapping the full injection current tensor, we show that this cascaded voltage is strongly tunable by the optical geometry: normal incidence drives an in-plane resonance $\mathrm{Im}(\eta_{yxy})$, whereas oblique illumination ($\theta = 45^{\circ}$) recruits a dominant out-of-plane component $\mathrm{Im}(\eta_{yyz})$ and amplifies the signal by more than two orders of magnitude (${\sim}10^2~\mu$V). While the massive linear Drude background typically screens nonlinear responses in semimetals, we argue that the amplitude modulation of the optical pump allows lock-in detection to cleanly isolate the frequency-doubled cascaded response. The proposed mechanism converts mid-infrared light into a gate-tunable transverse signal, providing a route for probing quantum geometry and realizing topological photodetectors and frequency doublers.

Coarse-graining (CG) lowers the computational cost of atomistic simulations by representing groups of atoms as effective interaction sites, reducing the degrees of freedom of the system but often compromising structural fidelity or requiring system-specific parameterization. Here, we introduce a novel anisotropic machine learning CG potential that extends the point particle representation of atomic nuclei to massive ellipsoidal beads with orientation-dependent features, enabling the learning of energies, forces, and torques directly from atomistic data. The anisotropic representation is physically motivated for polar and asymmetric molecules, where directional interactions and shape anisotropy play important roles in determining structure and dynamics. Using an equivariant message-passing neural network, the model accurately reproduces radial and angular distribution functions as well as relative orientation correlations in liquid water, demonstrating that both translational and rotational dynamics are well captured. Comparison with an isotropic baseline reveals that the lack of orientation information leads to systematic errors in short and long range order and degradation of angular correlations, proving orientation features are essential for accurate coarse-graining. The anisotropic model also exposes rotational structural observables fundamentally inaccessible to isotropic representations, with minimal computational overhead. Even for coarse-graining just three degrees of freedom, CG simulations achieve 7-27$\times$ speedups while preserving structural fidelity, highlighting the efficiency gains of this systemic reduction. This framework establishes the feasibility and necessity of learned equivariant representations for anisotropic CG modeling and provides a path towards accurate and efficient mesoscopic simulations of complex molecular liquids, polymers, and biomolecular systems.

Growing complex shapes requires control over both where growth begins and how it evolves in time. Here, we introduce a geometric framework for growing prescribed 2D shapes using a disk packing algorithm. In this approach, a target geometry is filled by disks whose centers define where growth is initiated and whose radii define how long each region is allowed to grow. The allowed disk sizes are constrained by the physics of the process, including the growth velocity, the time required to initiate each growth event, and the number of initiations that can occur in parallel. To generate physically realizable packings, we introduce the Largest Gap Algorithm (LGA), which sequentially fills the largest remaining gaps in a target shape with the largest disk that satisfies both geometric and kinetic constraints. We show that this method produces high coverage packings for a variety of geometries and that the resulting packings can be directly converted into spatiotemporal packing instructions. We then demonstrate that these instructions can be realized experimentally using multi-point initiation of frontal polymerization in viscosified dicyclopentadiene (DCPD) resin using CO$_2$ laser. Our results show that complex shapes can be grown by programming a small number of local initiation events, providing a simple connection between geometry and dynamics of growth.

Ni nanoparticles are frequently used as catalysts for hydrogenation reactions as well as in hydrogen storage applications. Recently, we have shown that small Ni nanoparticles can absorb hydrogen at < 10 bar pressure to form Ni hydride. During this process, the hydride growth is anisotropic, and a coherent Ni/NiH interface is formed. In order to explain the anisotropy and to comprehensively account for the coupling chemical, mechanical and thermal effects, we develop in this study a simplified chemo-thermo-mechanical enthalpy model for Ni/NiH interfaces in thin films. This model captures the combined influence of extent of hydride formation x, temperature T, pressure P, size effect ($l$), and the ${\alpha}$/${\beta}$ interface energy ${\lambda}$. Two different ${\alpha}$/${\beta}$ interface orientations, namely (100) and (111), are investigated. The model is shown to correctly predict the enthalpy and volume changes over a wide range of length scales, from atomically thin layers (~1 nm) to micron scale and larger. This work provides the basis for the development of similar enthalpy models for other solid-state hydrogen storage nanostructured materials where anisotropic growth of hydride phases is also observed.

Nonreciprocal forces are often a consequence of asymmetry in the properties of the interacting objects. However. even if all objects are identical and isotropic, and the pairwise interactions between two objects are completely reciprocal, nonreciprocal forces can still appear when an arrangement of many objects breaks configurational symmetry in the presence of non-pairwise, multi-body interactions. Here we demonstrate that such emergent nonreciprocity can activate a particle ensemble to behave like a liquid, albeit with unique traits. In our experiments, passive microspheres are acoustically levitated in air, where they form a freely floating monolayer containing up to a couple hundred particles and collectively behave like a two-dimensional liquid droplet. The particles interact via nonreciprocal multi-body forces that arise from the combination of acoustic scattering and sound-induced viscous microstreaming. We find that these forces drive superdiffusive particle motion with non-Gaussian tails in the particles' speed distribution. Using probes that reach laterally into the levitation plane, we perform liquid pendant drop and pinch-off experiments. Compared to ordinary liquids, the droplets are found to have a kinematic viscosity similar to that of water, but in combination with an extremely low interfacial tension. The pinching-off is driven by nonreciprocity-induced active fluctuations and exhibits the self-similar double-cone neck profile seen also in liquid nanojets close to rupture, however here characterized by power law behavior with a scaling exponent that is anomalously small.

Thanks to bosonization, bosonic lattice models can offer a lattice regularization of chiral fermions. We construct chiral lattice fermion operators in the 2D modified Villain scalar model and evaluate their correlation functions. This microscopic bosonic model has an ultra-local action and an ultra-local symmetry that realizes the fermionic chiral symmetry under bosonization. The reconstructed lattice Dirac operator has no doublers, but is consistent with the Nielsen-Ninomiya theorem because it turns out to be non-local. The non-locality of this derived quantity at finite lattice spacing does not pose any obstructions to gauging the non-anomalous symmetries of the model, which is itself ultra-local.

Polycatenanes are mechanically interlocked polymers composed of ring molecules linked through mechanical bonds. Here, we investigate the single-molecule dynamics of linear polycatenanes under strong steady shear flow using coarse-grained Brownian dynamics simulations with hydrodynamic interactions. We identify a stable tank-treading state in which individual rings rotate continuously while the overall polymer remains highly extended and aligned with the flow direction, adopting conformations typically associated with extensional flow. Remarkably, tank-treading is observed almost exclusively in odd polycatenanes, polycatenanes with an odd number of rings. We attribute this odd-even effect to the orientation of the terminal rings within the flow-gradient plane, where they act as anchors that stabilize polymer extension, a configuration accessible only to odd polycatenanes. Furthermore, our simulations and a Markov state model demonstrate that this dynamic behavior remains stable over long timescales. Finally, we show that tank treading polycatenanes can be used to fully extend other molecules in strong shear flow, hinting at applications of this behavior to manipulate molecules in rotational flows.

Heavy-fermion superconductors are mostly associated with f-electron materials with Kondo lattices, while known d-electron heavy-fermion-like systems are often linked to orbital-selective local moments, Hund-metal physics, or a charge-density-wave mechanism. Here we report Mo4PtGa17, a noncentrosymmetric itinerant d-electron superconductor with a geometrically frustrated breathing-pyrochlore Mo lattice. Thermodynamic, transport and NMR measurements reveal heavy-fermion-like behavior superconductivity and dominant ferromagnetic spin fluctuations near a ferromagnetic instability. Theoretical calculations identify nearly flat bands, van Hove singularities and Kramers nodal lines near the Fermi energy, derived intrinsically from Mo-4d states and are robust against on-site electronic correlations. These results suggest that the geometrically frustrated lattice in Mo4PtGa17 generates an intriguing electronic structure that enhances the density of states, spin susceptibility and quasiparticle mass. Mo4PtGa17 therefore identifies a unique route to heavy-fermion-like superconductivity in d-electron materials through geometrical frustration, different from the previously reported systems.

Metal halide perovskite semiconductors are promising materials for optoelectronic and photonic devices, including solar cells and next-generation coherent light sources. However, their low thermal conductivity limits the practical operation of devices under high excitation levels. Integrating thermally conductive, large band-gap two-dimensional (2D) materials into perovskite devices could suppress heat accumulation, while preserving their optical properties. Here, we show that planar hot-pressed (PHP) cesium lead bromide (CsPbBr${_3}$) thin films capped with few-layer 2D hexagonal boron nitride (hBN) are less affected by laser-induced heating under high-power continuous-wave excitation than uncapped perovskite samples. A large-scale semidry transfer method was developed to integrate 2D hBN onto PHP CsPbBr${_3}$ thin films. The process is chemically and thermally compatible with perovskites. Microscopic and spectroscopic analyses show that the hBN capping layer does not alter the morphology and optical properties of the perovskite thin film. The PHP CsPbBr${_3}$-hBN heterostructure exhibits a thermal conductivity of 3 W/(m * K), approximately seven times higher than that of the bare perovskite films of 0.45 W/(m * K). Heat diffusion simulations confirm enhanced heat dissipation in the heterostructure relative to bare perovskite films. Our experiments demonstrate an effective approach to enhancing heat dissipation in perovskite devices using transparent, thermally conductive 2D materials.

Machine-learning models for materials properties rest on two assumptions that standard validation never tests: that a model's features reflect the physics of the property rather than accidents of the training distribution, and that the training labels are themselves correct. We introduce a model-agnostic audit protocol for both, combining SHAP attribution, counterfactual partial dependence analysis, and Rashomon-style cross-model verification, with every finding adjudicated by targeted density functional theory (DFT). Demonstrated on intrinsic spin Hall conductivity using a composition-only Random Forest, the model needs no relaxed crystal structure, reaching accuracy competitive with structure-aware graph networks while remaining applicable to the far larger space of compositions for which no structure has been computed. The model audit reveals that the average p-valence descriptor becomes statistically entangled with Pt content - a property of the learned representation rather than the physics; DFT confirms the consequence, a Pt-free compound (HgOsPb$_2$) whose true SHC is nearly four times the prediction. The data audit exposes a thirtyfold error in the HfC training label, inherited undetectably by every black-box model trained on the same data. The protocol audits a model and its training data for the cost of a few DFT calculations, wherever one element dominates the high-property regime.