Papers

The phase behavior of a single type of colloid C suspended in near-critical solvents is known to be very rich. Motivated in part by recent experiments we consider a mixture of two colloidal types C1 and C2 in a binary solvent close to its demixing critical point. We extend a mean-field description of a lattice model, previously used to investigate systems with a single type of colloid in two dimensions, to the binary colloid case in three dimensions. The model treats the system as a full four-component mixture. For simplicity we choose C1 and C2 to be hard spheres with the same radius but with different affinities for one species, B, of the AB binary solvent. We show that intricate interplay between couplings of C1 and solvent, C2 and solvent as well as solvent-solvent interactions and hard sphere packing drive significant changes in the topology of the colloidal phase diagram when the relative volume fractions of the two different colloid types change. The behavior of the two lines of triple points is particularly interesting. Our results can provide some insight into the control of the self-assembly process for colloidal 'alloys' mediated by a near-critical solvent and therefore controlled by temperature in a reversible manner

We present an integrated iontronic memristor circuit that reproduces biologically inspired Spike Rate-Dependent Plasticity (SRDP) and functions as a physical nonlinear frequency kernel, which we demonstrate can be used to classify natural auditory data. The fluidic circuit integrates two parallel memristive membranes containing short and long conical memristive channels with opposite orientations, giving rise to heterogeneous internal timescales and different potentiation responses. As a result, the circuit exhibits a nonlinear frequency response in which low-frequency inputs decrease the overall conductance, whereas higher-frequency inputs increase it, thereby emulating biological SRDP. Our experimental measurements are inspired by and consistent with predictions of a theoretical model. We demonstrate the functionality of the device by separating encoded sound signals from different insects that cannot be linearly separated. By unifying theoretical predictions with experimental realisation of coupled iontronic memristors, this work moves beyond isolated components and demonstrates how heterogeneous iontronic dynamics can unlock nonlinear time-series processing capabilities, essential for future iontronic neuromorphic computing.

DNA self-assembly is a well-understood nanotechnology to obtain extremely ordered structures from the nanometer to up to the hundred of microns scale. By contrast, DNA hydrogels rely on the disordered assembly of DNA building blocks to reach macroscopic volumes. However, in order to hold the promise of DNA bulk materials, the sequence designer needs a systematic understanding of how macroscopic properties emerge from disorder. Here, we show a method to study systematically the mechanical response of a simple DNA nanostar hydrogel. This method mobilises bulk rheology, dynamic light scattering microrheology, mechanical modeling, as well as thermodynamic calculation and DNA sequence alteration. At low temperatures, we demonstrate a systematic deviation from Maxwell behaviour that is symptomatic of disordered materials. At temperatures much higher than the percolation of the DNA network, we characterise a surprising solid behaviour that we attribute to a glass transition. Our results show the importance of disorder in DNA materials. Furthermore, the method we showcase in this article can be widely applied to more complex DNA materials.

In solid-state batteries, ceramic solid electrolytes are penetrated by dendrites when plating above a critical current density $J_\mathrm{crit}$. A dendrite will propagate by metal deposition at a pre-existing dendrite tip if the mechanical energy required to crack the ceramic open is less than the electrical energy (Joule heating) wasted by forcing the current to detour around the dendrite to the flat electrode surface. Based on this principle of minimal power dissipation, a dependence of $J_\mathrm{crit}\propto c_\mathrm{max}^{3/2}$ is derived in an analytical fashion. $c_\mathrm{max}$ is the length of the longest pre-existing, sufficiently thin interfacial defect. Consequentially, scattering of dendrite growth between samples must follow a Weibull-distribution, similar to the tensile strength of ceramic components but at smaller Weibull-modulus.

We study the coupling mechanism of surface acoustic waves (SAW) with spin waves (SW) using micromagnetic analysis. The SAW magnetoacoustic excitation field is fully implemented, i.e., all strain and lattice-rotation terms are included. A realistic CoFeB film with a weak in-plane uniaxial anisotropy is considered. We investigate the conditions for efficient SAW–SW coupling, with particular emphasis on the case where the SAW propagates parallel to the external magnetic field, a configuration of special interest for magnonic applications. Remarkably, we find that the anisotropy orientation serves as a knob to tune the parallel resonant interaction. Overall, this work provides a unified and practical picture of SAW–SW coupling in thin magnetized films.

The ST1 diamond color center was experimentally demonstrated to involve a substitutional oxygen atom (O$_C$) and carbon vacancy (V$_C$), has a spin singlet ground-state, and a metastable electron spin ancilla: a triplet. ST1's structure was left unsolved for more than a decade. With embedded multiconfigurational quantum mechanical theory, we investigate O$_C$-V$_C$-derived diamond defects, specifically both 0 and +2-charged coupled O$_C$V$_C$, and O$_C$ surrounded by V$_C$s along the [110] axis (V$_C$O$_C$V$_C$). We found both O$_C$V$_{C}^{2+}$ (C$_{3v}$) and V$_C$O$_C$V$_{C}^{2+}$ (C$_{2v}$) to have a spin-singlet ground state (1$^1$A$_1$) and metastable spin triplets. We demonstrate ST1 to be V$_C$O$_C$V$_{C}^{2+}$. The calculated vertical excitation energies of V$_C$O$_C$V$_{C}^{2+}$'s first (1$^1$B$_2$) and second (2$^1$A$_1$) bright spin-singlet excited states closely match ST1's experimental zero phonon line (2.2-2.3 eV). O$_C$V$_{C}^{2+}$ ($^1$E) absorbs much higher (2.8 eV). The two O lone pairs favor V$_C$O$_C$V$_C$ over O$_C$V$_C$, in a similar manner as the single N lone pair favors formation of N$_C$V$_C$ centers.

Over the last decades, two distinct approaches have been instrumental to our understanding of the computational complexity of statistical estimation. The statistical physics literature predicts algorithmic hardness through local stability and monotonicity properties of the Franz–Parisi (FP) potential \cite{franz1995recipes,franz1997phase}, while the mathematically rigorous literature characterizes hardness via the limitations of restricted algorithmic classes, most notably low-degree polynomial estimators \cite{hopkins2017efficient}. For many inference models, these two perspectives yield strikingly consistent predictions, giving rise to a long-standing open problem of establishing a precise mathematical relationship between them. In this work, we show that for estimation problems the power of low-degree polynomials is equivalent to the monotonicity of the annealed FP potential for a broad family of Gaussian additive models (GAMs) with signal-to-noise ratio $\lambda$. In particular, subject to a low-degree conjecture for GAMs, our results imply that the polynomial-time limits of these models are directly implied by the monotonicity of the annealed FP potential, in conceptual agreement with predictions from the physics literature dating back to the 1990s.

Quantum noise has long served as a powerful probe of quantum transport in mesoscopic junctions. Recently, temperature-driven noise, or $\Delta_T$ noise, has attracted growing interest due to its presence even in the absence of average charge current. In this work, we investigate a normal metal-insulator-iron-pnictide junction and demonstrate how zero temperature quantum shot noise, finite temperature quantum noise and $\Delta_T$ noise can discriminate between $S_{++}$ and $S_{+-}$ pairing symmetries, which are relevant to iron-based superconductors. We introduce $\Delta_T$ noise as a novel probe for distinguishing between the two pairing symmetries. In contrast to conductance, which exhibits a single peak for both $S_{++}$ and $S_{+-}$ states with only a difference in magnitude, the $\Delta_T$ noise reveals qualitatively distinct features: a twin-peak structure for the $S_{++}$ pairing symmetry and a single-peak profile for the $S_{+-}$ state. A similar symmetry-dependent contrast is observed in both zero temperature quantum shot noise and finite temperature quantum noise, where the $S_{++}$ state consistently exhibits a twin-peak structure, while the $S_{+-}$ state shows a single-peak response. Our results demonstrate that noise-based measurements form a mutually reinforcing set of probes that enables reliable identification of superconducting gap symmetry in Iron Pnictide superconductors.

Chiral active matter, which breaks both parity symmetry and time-reversal symmetry, is ubiquitous in living systems. Here, we introduce a minimal two-dimensional chiral active lattice gas by incorporating stochastic, biased local rotations. At low temperatures, the system coarsens into condensates with chiral orientations and faceted, crystal-like shapes. The interfaces align at characteristic angles with respect to the lattice axes and exhibit edge currents that are persistent, unidirectional, and angle-dependent. To generalise these findings, we propose a continuum theory by adding an active chiral edge current term to Model B, which reveals the essential role of active chiral transport in the interfacial dynamics of phase separation. Edge currents with $n$-fold symmetry produce condensates whose shapes resemble regular $n$-sided polygons. In the thin-interface limit, we construct an effective interface potential governing edge currents, from which the steady-state condensate geometry can be obtained, both in the lattice model and the continuum description.

This review examines multiscale modelling approaches for cellulose nanocrystals (CNCs) and lignocellulosic plant cell walls, with a focus on hemicellulose and lignin interactions in aqueous environments. The three-dimensional reference interaction site model with the Kovalenko-Hirata closure (3D-RISM-KH) is highlighted as a powerful molecular solvation theory applied in nanochemistry and biomolecular simulations. The method has been successfully employed to investigate hemicellulose hydrogels, the influence of hemicellulose composition on nanoscale forces in primary cell walls, and lignin-lignin and lignin-hemicellulose interactions. Findings indicate that these interactions are predominantly hydrophobic and entropy-driven, arising from water exclusion effects. Insights gained through this modeling framework deepen the understanding of molecular-scale forces in plant cell walls and inform strategies for biomass valorization, including genetic engineering and pretreatment technologies aimed at enhancing cellulose extraction and utilization.

Protein conformational stability and function depend on non-covalent interactions that are strongly influenced by the surrounding environment. To explore protein properties, amino acids are often utilized as model systems. In this study, we determined the densities of seven $\alpha$-amino acids in aqueous solutions between 278.15 K and 308.15 K and calculated the apparent molar volumes. Linear extrapolation yielded standard molar volumes, which were analyzed to characterize amino-acid hydration. The contributions of side chains to the standard molar volume were determined relative to glycine. The standard molar volume increased with temperature, indicating reduced electrostriction of water around the amino acids, consistent with lower hydration numbers at higher temperatures. We employed the Ornstein-Zernike integral equation with hypernetted-chain closure and a coarse-grained Lennard-Jones bead model to calculate pair correlation functions and Kirkwood-Buff integrals, from which standard molar volumes were obtained. The model reproduced the experimental standard molar volumes very well.

Thermomigration is the driving force for hydrogen transport due to a temperature gradient. It can compete with hydrogen transport induced by stress gradients. While stress-driven hydrogen migration is well established, thermomigration remains comparatively underexplored, largely due to limited mechanistic understanding and a scarcity of experimental data. In this work, we develop a thermodynamically consistent framework for hydrogen transport, incorporating a mechanistic model for thermomigration. This is implemented within a finite element framework using an effective chemical potential. Using case studies of iron and nickel heat exchangers and zirconium alloy nuclear fuel cladding, we quantify the competing and synergistic effects of thermomigration and stress-driven transport. We show that thermomigration often dominates hydrogen redistribution in heat-carrying components, even in the presence of significant thermal incompatibility stresses. However, stress-driven transport is shown to become decisive near sharp stress concentrators. A graphical method is introduced to rapidly identify the dominant transport mechanism without requiring fully coupled simulations. The results provide practical guidance for assessing hydrogen redistribution and embrittlement risk in heat-carrying structural components.

Computer simulations are employed to investigate the adsorption mechanisms of ethane on both homogeneous and inhomogeneous substrates. For homogeneous surfaces, the full range of surface phase transitions - from incomplete to complete wetting - can be accessed by tuning the strength of the surface potential. The resulting layering transition temperatures show excellent agreement with experimental measurements of ethane on graphite. By contrast, although all inhomogeneous substrates exhibit a prewetting transition, the adsorption mechanisms are strongly influenced by the stripe width.

Thermodynamic uncertainty relations (TURs) impose a universal trade-off between current precision and entropy production in autonomous steady states, constraining in particular the power, efficiency, and constancy of heat engines. We demonstrate strong violations of the long-time TUR in a minimal autonomous heat engine composed of a discrete ratchet generating work against a constant bias and an underdamped harmonic oscillator acting as an internal stochastic control. In the regime of time-scale separation, the model becomes exactly solvable and yields a closed analytical expression for the TUR ratio, where the influence of the continuous degree of freedom is fully captured by the Fano factor of oscillator zero crossings. We show that increasingly deterministic internal control drives the TUR ratio arbitrarily close to zero while the engine operates near maximal current and efficiency. In an appropriate limit, the model reduces to the classical pendulum-clock system of Pietzonka, Phys. Rev. Lett. 128, 130606 (2022).

We explore the pressure dependence of a set of properties of simple monohydric alcohols, namely of methanol, ethanol and 1-propanol, by using isobaric-isothermal molecular dynamics computer simulations. A recently proposed united atom, non-polarizable force field for each of alcohols [V. García-Melgarejo et al., J. Mol. Liq., 323, 114576 (2021)] is applied. Accuracy of the force field is evaluated by comparing the simulation results and available experimental data from the literature. Specifically, the density of alcohols upon increasing pressure, the isothermal compressibility, the static dielectric constant and self-diffusion coefficient are investigated starting from 1 bar up to 3 kbar. Evolution of the microscopic structure under pressure is discussed in terms of the pair distribution functions and some coordination numbers. Conclusions of the present modelling and necessary developments to consider in future work are commented on.

We study the behavior of aqueous surfactant solutions in the bulk phase and in slit-like pores by molecular dynamics. Adsorption and self-assembly of nonionic surfactants C$_7$E$_3$ that mimic alkyl poly(ethylene oxide) molecules are investigated. We consider pores with the same walls and Janus-like slits. The individual walls are inert, hydrophilic, or hydrophobic. We focus on the morphology of the surfactant solution confined in different slits. The influence of a pore type and its width is discussed. The aggregative adsorption of surfactants was found. Our simulations show that in slits surfactants assemble into structures that do not occur in the bulk phases.

We explore the dependence of a wide set of properties of monohydric alcohols on temperature by using the isobaric-isothermal molecular dynamics computer simulations. Namely, methanol (MeOH), ethanol (EtOH) and 1-propanol (PrOH) alcohols are studied. The recently proposed united atom, non-polarizable force field for each of alcohols [V. García-Melgarejo et al., J. Mol. Liq., 2021, 323, 114576] is applied for this purpose. Accuracy of the force field is discussed comparing predictions from simulations and experimental data for density, dielectric constant, surface tension, and self-diffusion coefficient. Supplementary insights concerning applicability of the model are obtained by exploration of the composition dependence of various properties for MeOH-PrOH mixtures. Peculiarities of mixing of species in this system are elucidated in terms of density, excess mixing volume and excess mixing enthalpy. Static dielectric constant of the mixture and the corresponding excess are obtained. Perspectives of modelling are commented finally.

Symmetric mixtures characterized by high negative geometric and energetic non-additivity do not exhibit phase separation in the bulk. However, the phase separation occurs when such mixtures are confined in slit pores with selective walls. It is demonstrated that the wall selectivity affects the pore filling. When the difference of the interaction energies between the mixture components and pore walls is lower than a certain threshold value, condensation occurs between a dilute phase and the mixed liquid. When this difference exceeds the threshold value, the pore filling may occur in two steps. The first is the condensation of a dilute phase into the demixed liquid, and the second step leads to the formation of the mixed liquid. We have elucidated the changes in the phase behavior caused by non-additivity of symmetric mixtures, and by the difference in the interaction energies of the components with pore walls.

I analyze the nonlinear Hamiltonian equations of motion for a one-dimensional chain of transverse magnetic nano-islands, seeking solutions for different types of static domain-walls (DWs) connecting uniform static states. The system of elongated magnetic islands oriented transverse ($y$-direction) to the chain direction ($x$-direction) experiences an applied magnetic field transverse to the chain. The macro-spin model includes dipole interactions between islands, their uniaxial and easy-plane anisotropies, and Oersted energy of the applied field. DWs can form most easily between pairs of degenerate uniform states, described by their local magnetizations as oblique, $y$-parallel, and $y$-alternating. The DWs between oblique states are well-described with scalar $\varphi^4$ theory. General DW structures are found via a numerical energy relaxation scheme. At some anisotropy and field parameters, nearest-neighbor dipole interactions drive antiferromagnetic order inside the DW itself. The variety of DWs present in the model might be exploited for their sensitivity to parameter changes in detectors or switching technology.

We analyze the soft committee machine with Rectified Linear Unit (ReLU) activation by means of the replica method. In a realizable teacher–student setting, we compute the quenched free energy within a replica-symmetric ansatz and obtain the typical generalization behavior from the saddle-point equations for the macroscopic order parameters. The system exhibits a transition from an unspecialized symmetric phase to a specialized phase in which the permutation symmetry among hidden units is broken. We determine the critical training-set size as a function of the inverse training temperature and derive analytic expressions both near the transition and in the asymptotic large-sample regime. Unlike the corresponding model with sigmoidal activations, which undergoes a first-order transition, the ReLU soft committee machine shows a continuous specialization transition. These results show that the activation function plays a decisive role in the phase structure and generalization behavior of multilayer networks.

Machine-learning electronic Hamiltonians achieve orders-of-magnitude speedups over density-functional theory, yet current models omit long-range Coulomb interactions that govern physics in polar crystals and heterostructures. We derive closed-form long-range Hamiltonian matrix elements in a nonorthogonal atomic-orbital basis through variational decomposition of the electrostatic energy, deriving a variationally consistent mapping from the electron density matrix to effective atomic charges. We implement this framework in HamGNN-LR, a dual-channel architecture combining E(3)-equivariant message passing with reciprocal-space Ewald summation. Benchmarks demonstrate that physics-based long-range corrections are essential: purely data-driven attention mechanisms fail to capture macroscopic electrostatic potentials. Benchmarks on polar ZnO slabs, CdSe/ZnS heterostructures, and GaN/AlN superlattices show two- to threefold error reductions and robust transferability to systems far beyond training sizes, eliminating the characteristic staircase artifacts that plague short-range models in the presence of built-in electric fields.

Localized surface plasmon resonances are self-sustained, collective oscillations of free electrons in metallic nanostructures. They have a wide range of applications. The most common plasmonic metals are noble metals, such as gold and silver. However, there are applications, such as surface-enhanced Raman spectroscopy, in which using non-noble metals is advantageous. This review summarizes the investigation of localized surface plasmons in non-noble metal nanoparticles, providing an overview of the plasmonic properties of non-noble metals. We cover the following metals: aluminium (Al), antimony (Sb), bismuth (Bi), chromium (Cr), copper (Cu), gallium (Ga), indium (In), lead (Pb), magnesium (Mg), molybdenum (Mo), nickel (Ni), potassium (K), selenium (Se), sodium (Na), tellurium (Te), tin (Sn), titanium (Ti), tungsten (W), and zinc (Zn). Our summary therefore compares the plasmonic properties of non-noble metals and briefly introduces their potential to the readers.

In the Comment by Bordag et al. [Phys. Rev. B 113, 207401 (2026) and ArXiv:2506.10792], concerns are raised regarding the validity of the results presented in [Phys. Rev. B 111, 115428 (2025)], where the theoretical descriptions of the electric conductivity of graphene obtained from the Kubo formula and from quantum field theory via the polarization tensor are compared. In this Reply, we show that these concerns arise from misinterpretations of Phys. Rev. B 111, 115428 (2025), in which the results are either inaccurately represented or applied outside the domain of validity of the model. We address the comments concerning the derivation of the Luttinger formula for the electric conductivity from the Kubo formula and clarify why the results of Phys. Rev. B 111, 115428 (2025) cannot be arbitrarily extended to make claims on the gauge invariance. We further demonstrate that our findings are fully consistent with the established and widely accepted literature cited in the Comment. We confirm that the model for electric conductivity discussed in Phys. Rev. B 111, 115428 (2025) correctly predicts a vanishing electric current in the absence of an external electric field, as physically required, and in contrast with the model advocated by the Authors of the Comment. We also show that the electric permittivity does not exhibit a double pole in $\omega$, contrary to the claim made in the Comment. Finally, we emphasize that the inclusion of losses is a standard and well-established approach in the study of transport properties of materials, including graphene, and we take the opportunity to correct a few minor typographical errors in Phys. Rev. B 111, 115428 (2025). We show and maintain that all results derived in Phys. Rev. B 111, 115428 (2025) are fully valid and correct.

We investigated the lattice dynamics of the unconventional superconductor LiFeAs using inelastic neutron scattering experiments and density-functional theory (DFT) calculations. By comparing the neutron scattering intensities with lattice-dynamics simulations we can identify the polarization symmetry of all modes along the main-symmetry directions yielding a complete experimental picture of the phonon dispersion. Overall there is good agreement between the experimental and DFT results, which renders an overlooked strong electron phonon coupling unlikely. Our DFT calculations reveal only a small averaged electron-phonon coupling constant. The transversal acoustic in-plane branches exhibit a normal dispersion for small propagation vectors indicating the absence of a nematic instability. Several modes exhibit considerable hardening upon cooling that can be attributed to the anisotropic shrinking of the LiFeAs lattice.

Quantum Fisher information (QFI) is a measure of multipartite quantum entanglement that can be obtained from inelastic neutron scattering data on quantum magnets. In this work, we demonstrate that the QFI can distinguish an unconventional quantum critical point (QCP) with fractionalization and emergent gauge structure from conventional ones within the Landau paradigm. We compute the QFI, via large-scale quantum Monte Carlo (QMC) simulations and exact diagonalization, in a kagome lattice quantum spin liquid (QSL) model with an XY and a cluster-Ising interactions. When the XY interaction is ferromagetic, the QFI obtained by QMC reveals a large anomalous dimension, which is a fingerprint of the (2+1)d XY$^\ast$ universality class for the transition from the ferromagnetic phase to the $\mathbb{Z}_2$ QSL. The investigation of thermal and dynamical properties of QFI is further extended to the case of antiferromagnetic XY interaction via exact diagonalization. In this regime, a transition to a possibly distinct QSL phase is suggested via both entanglement-based probes, such as QFI and genuine multipartite negativity, and analyses of the energy spectrum and structure factors. These results not only demonstrate the versatility of QFI in identifying QSL states and unconventional QCPs but also provide useful guidance for future theoretical and experimental studies of frustrated magnets.

Atomic electron tomography (AET) determines the three-dimensional (3D) coordinates and chemical identities of individual atoms from a series of scanning transmission electron microscopy images taken at different tilt angles. However, under the low dose conditions required to mitigate beam damage, the reduced signal-to-noise ratio forces a trade off among accuracy, robustness, and throughput, which ultimately limits the broader application of AET. Here, we introduce a physics aware, two stage neural networks (PANN) that incorporates physical constraints throughout its workflow to achieve accurate AET under low-dose imaging. First, a global local 3D ResUNet refines the initial reconstruction and corrects geometric distortions in the volume. Second, the local density around each identified atom is encoded using 3D Zernike moments. These feature descriptors, along with the atomic coordinates are then processed by a graph attention Transformer to classify the elemental species. We benchmark the PANN workflow using a dataset of 42,588 reconstructed volumes, covering diverse noise models, materials morphologies, and dose settings. Under low dose conditions, PANN significantly improves performance, reducing the atomic coordinates error and leading to an increase in the atomic recovery rate. The framework's performance on experimental lose dose AET data across nanoparticles of varying morphology and composition demonstrate robust generalization. We anticipate this approach will extend the applicability of AET, particularly in investigating materials sensitive to electron dose or chemical state, including halide perovskites, zeolite, and quantum dot.

In this work, we demonstrate interfacial charge transfer-driven transport enhancement in few-layer graphene monolayer MoS2 vertical heterostructure field-effect transistor. Raman scattering and Raman intensity mapping results confirm the successful stacking of FL graphene on ML MoS2. Pronounced photoluminescence (PL) quenching of MoS2 and spectral redshift in the heterostructure suggest efficient interlayer charge transfer and strong electronic coupling at the vdW interface. Electrical measurements show enhanced drain current, field-effect mobility, and conductivity in Gr-MoS2 device compared to pristine MoS2 transistor with Ag contacts. The energy band considerations under equilibrium and gate bias conditions suggest improved Fermi-level alignment and reduced effective Schottky barrier effects at the graphene-MoS2 interface, enabling efficient carrier injection. Temperature-dependent transport (300-400 K) reveals phonon-dominated mobility and conductivity degradation in both devices; however, the heterostructure exhibits significantly suppressed performance degradation. The mobility enhancement factor increases from ~1.6 at 300 K to ~4.0 at 400 K, accompanied by a corresponding improvement in conductivity stability, demonstrating superior thermal robustness for the Gr-MoS2 heterostructure. The power-law analysis indicates that transport in pristine MoS2 is influenced by both intrinsic phonon scattering and additional thermally activated extrinsic processes such as contact and interfacial effects, whereas the weaker temperature dependence in the Gr-MoS2 device reflects moderated extrinsic contributions and transport behaviour approaching a predominantly phonon-limited regime. These findings demonstrate graphene contact engineering as a viable pathway toward improved performance and thermally stable two-dimensional semiconductor electronics.

In this contribution paper, we construct a two-dimensional non-Hermitian (NH) photonic crystal (PhC) to prototype its anisotropic non-Hermitian skin effect (NHSE) for experimental proposal. Based on the tight-binding model for Lieb lattice with NH coupling, a nontrivial spectral winding number is pinpointed for certain eigenstates, which translates to geometry-dependent skin modes with tilt boundaries. For ease of implementation, complex refractive indices are employed for the Lieb unit cell of PhC to emulate the NH coupling. Validated by full wave simulation, our work underscores the boundary dependence of skin effect, and provides a concrete prototype design of NHSE implementable by state-of-the-art of topological metamaterial platforms.

Differential scanning calorimetry (DSC) is a standard tool for studying precipitation and phase transformations in aluminum alloys, yet its relation to mechanical performance has so far remained mostly indirect. Here, we demonstrate that DSC curves themselves act as fingerprints that directly encode mechanical properties. Four representative 6xxx series alloys (Al-Mg-Si) were subjected to different natural and artificial aging regimens, followed by DSC heat-flow measurements and tensile testing. Machine learning models trained on the thermograms predicted yield strength, ultimate tensile strength, and uniform elongation in five-fold grouped cross-validation, with the best model (Lasso) achieving R^2 values of 0.93, 0.86, and 0.87 and mean absolute errors of 14.3 MPa, 11.1 MPa, and 1.5 percent, respectively. Leave-one-alloy-out evaluation with sparse calibration using anchor samples further demonstrated generalization across alloy chemistries. While direct prediction on unseen alloy data degraded performance substantially, inclusion of as few as one to two anchor conditions from the target alloy recovered predictive accuracy, approaching that of the standard cross-validation. Feature importance analysis revealed that the 230 to 270 C region, associated with precipitation of the primary hardening phase beta”, contributed most strongly to predictive accuracy, providing direct mechanistic validation of the model. These findings establish DSC as a diagnostic tool that can serve as a rapid proxy for mechanical property evaluation, enabling accelerated alloy screening, process optimization, and integration of thermal analysis into data-driven manufacturing.

Federated learning enables collaborative training without sharing raw data, but struggles under client heterogeneity and streaming distribution shifts, where drift and novel data can impair convergence and cause forgetting. We propose a federated associative-memory framework that learns shared archetypes in heterogeneous, continual settings, where client data are independent but not necessarily balanced. Each client encodes its experience as a low-rank Hebbian operator, sent to a central server for aggregation and factorization into global archetypes. This approach preserves privacy, avoids centralized replay buffers, and is robust to small, noisy, or evolving datasets. We cast aggregation as a low-rank-plus-noise spectral inference problem, deriving theoretical thresholds for detectability and retrieval robustness. An entropy-based controller balances stability and plasticity in streaming regimes. Experiments with heterogeneous clients, drift, and novelty show improved global archetype reconstruction and associative retrieval, supporting the spectral view of federated consolidation.