Browse Articles
Discover research articles across all indexed journals
Quantum machine learning for predicting properties of van der Waals bilayers
Layering two-dimensional (2D) materials into van der Waals bilayers provides an effective method to achieve innovative quantum states and tunable electronic properties. Exploring the extensive configurational space resulting from various layer combinations, twist angles, and stacking patterns presents a significant computing challenge. As the volume of accessible datasets expands, training machine learning models on conventional hardware may ultimately become excessively expensive, prompting the advancement of quantum machine learning (QML) methodologies for materials discovery. This study presents a quantum-enhanced machine learning framework for predicting the bandgaps of van der Waals bilayers utilizing quantum simulators. We start with conventional feature selection to ascertain the most important descriptors affecting electronic properties. The properties are subsequently encoded into quantum states utilizing parameterized quantum circuits. We evaluated three quantum feature maps, namely ZFeatureMap, ZZFeatureMap, and PauliFeatureMap, and observed task-dependent performance. For the classification task, ZFeatureMap with two repetitions gave the strongest quantum-classifier performance, whereas for Quantum Support Vector Regressor (QSVR) bandgap regression, PauliFeatureMap with one repetition achieved the lowest average root-mean-square error (RMSE). These results indicate that feature-map choice should be optimized separately for classification and regression rather than being interpreted as the universal superiority of a single feature map. Using these feature-map comparisons, we employ a QSVR and a variational quantum regressor to predict the bandgaps of 1850 semiconducting vdW bilayers, illustrating that QML may attain significant predictive accuracy despite limited training data. Overall, the results demonstrate the feasibility of a simulator-based QML workflow for materials-property prediction. The quantum classifiers are competitive with the classical support vector classifier (SVC) baseline for zero-gap/non-zero-gap classification, whereas the classical Support Vector Regression (SVR) achieves the lowest RMSE in the regression task. Therefore, the present study should be interpreted as an exploratory benchmark of hybrid classical–quantum models for van der Waals (vdW) bilayers rather than as evidence of a definitive quantum performance advantage.
Element-specific dynamical decoupling and local structural ordering in liquid NiCoCr medium-entropy alloy
The Stokes–Einstein relation (SER) links diffusion and viscosity, but its applicability is challenged in systems with complex structure and chemical properties. Medium-entropy alloys (MEAs), with their inherent chemical complexity and propensity for diverse local ordering, offer a compelling platform to study such coupling phenomena in the liquid state. However, how element-specific structural preferences influence SER breakdown remains unexplored. Here, we investigate the microscopic mechanisms underlying SER breakdown in liquid equiatomic NiCoCr MEA using high-energy x-ray scattering and molecular dynamics simulations. A dynamic transition is observed near 1700 K (melting point 1682.9 K). While the self-diffusion coefficients of all elements retain Arrhenius behavior down to the deeply supercooled regime, the viscosity and structural relaxation times exhibit a clear cross-over between two Arrhenius regimes below this temperature, leading to the breakdown of the inverse scaling between diffusion and viscosity. This decoupling is primarily governed by the anomalous temperature dependence of the viscosity. Fractional SER analysis reveals element-specific decoupling, with Cr showing a fundamentally distinct departure from classical scaling compared to Ni and Co. Concurrently, non-Gaussian parameters reveal growing dynamic heterogeneity upon cooling. Structurally, short-range order strengthens with significant increases in icosahedral-like and mixed clusters, especially those centered on Cr, accompanied by an enhancement of local five-fold symmetry. The formation of these rigid, Cr-centered ordered domains amplifies local geometric constraints, which severely hinder cooperative atomic rearrangements, while leaving single-atom diffusion less affected. These results connect element-specific ordering to the viscosity-driven breakdown of SER in NiCoCr, providing a structural perspective on diffusion-viscosity decoupling in MEA liquids.
Electrowetting on thick substrates: A modified Young–Lippmann equation
Electrowetting is a versatile technique for controlling the apparent contact angle of drops on thin insulating films through an externally applied potential. In the light of slide electrification—the spontaneous charging of sliding drops on dielectric substrates—electrowetting has gained new importance. Drops can spontaneously attain kilovolt potentials and undergo spontaneous electrowetting. Often, drop sizes and substrate thicknesses are comparable, violating the assumptions underlying the Young–Lippmann and the plate capacitor equations. In this work, I numerically determine the capacitance of sessile drops on thick substrates and introduce an analytical approximation. I derive a modified Young–Lippmann equation that quantitatively describes electrowetting ranging from thin films to substrate thicknesses many times the drop size.
Coupled-cluster theory for positron binding in anions and polyatomic molecules
We present the positron coupled cluster singles and doubles (POS-CCSD) method to calculate positron binding energies in molecules. This framework treats electrons and positrons on an equal footing and includes up to simultaneous double-electron–single-positron excitations. We benchmark the approach by computing binding energies for atomic anions and several polar and non-polar polyatomic systems, comparing the results with independent theoretical studies and, where available, experimental data. The fully converged results for H− are in excellent agreement with quantum Monte Carlo and multi-reference configuration interaction results. Quantitative agreement with experiments is not reached in the present study due to the slow convergence of the binding energy with respect to the size of the orbital bases for the electrons and the positron. However, the POS-CCSD results underscore the critical role of electron correlation in the description of electron–positron systems required for a balanced description of these complex systems. In addition, we examine nuclear relaxation effects following positron attachment in LiH.
Local thermal probe in a model molecular chain: A dissipaton-based approach
We study a system consisting of an infinite one-dimensional model molecular chain locally coupled to a probe. Starting from the Hamiltonian of the chain–probe composite and the corresponding spectral densities, we evaluate the heat current between the probe and the chain. Inspired by the perspective of quantum dissipative dynamics, we adopt a dissipaton-based quantum approach that is nonperturbative and non-Markovian. The dissipaton algebra yields a set of hierarchically coupled equations of motion for the c-number dissipaton moments, with iterative cross-tier connections when higher-order chain–probe interactions are included. Numerical results reveal the effects of temperature, probe frequency, on-site energy modification, and higher-order couplings on heat transport.
Surrogate functionals for machine-learned orbital-free density functional theory
We introduce surrogate functionals: machine-learned energy functionals for orbital-free density functional theory (OF-DFT) which are defined not by universal fidelity to a physical reference, but merely by the requirement that density optimization with a fixed procedure yields the true ground-state density. Helpfully, training surrogate functionals requires only ground-state densities as labels, no energies or gradients away from the ground state. We here propose a gradient-descent-improvement loss that guarantees exponential convergence of the density to the ground state, and combine it with an adaptive sampling scheme that concentrates learning around the optimization trajectories actually visited during inference. On the QM9 and QMugs benchmarks, surrogate functionals achieve density errors competitive with or improving upon the state of the art for fully supervised machine-learned OF-DFT, while eliminating the need for the O(N3) orthonormalization step required by prior work, yielding improved runtime scaling for larger systems.
Decoding vibrational energy transfer in CO–O2 collisions: Vibrational state-to-state rate coefficient datasets on a new potential energy surface
Energy exchanges in collisions between carbon monoxide (CO) and molecular oxygen (O2) are investigated in this paper, and comprehensive datasets of state-to-state rate coefficients for vibration-to-vibration (V–V) and vibration-to-translation/rotation (V–T/R) energy transfer processes are obtained. The data are generated using a mixed quantum–classical (MQC) dynamical method on a newly developed potential energy surface, which is optimized against available experimental data and high-level ab initio calculations. The MQC dataset covers a broad temperature range from 100 to 5000 K, with additional extension through a Gaussian process regression technique to incorporate key V–T/R and near-resonant V–V processes, which include vibrational states of vCO up to 30, as well as the full vibrational ladder of O2 up to its dissociation limit. The results reveal the dominance of V–V kinetics at lower temperatures and the increasing importance of V–T/R processes at higher temperatures. The present dataset addresses a critical gap in the kinetic modeling of CO + O2 collisions in high-temperature gaseous environments.
Single chain expulsion from diblock copolymer micelles with dense corona
We use self-consistent field theory to investigate the free energy landscape for single-chain expulsion from a diblock copolymer micelle with a dense corona. Using the distance from the micelle center-of-mass to the hydrophilic–hydrophobic junction of the chain as the reaction coordinate, we compute the free energy landscape for chain exchange. Our results show that the expulsion free energy barrier scales linearly with both the hydrophobic block length and the solvent selectivity, consistent with recent experiments. To accurately resolve the chain conformation, we introduce a second reaction coordinate: the distance between the junction and the free end of the hydrophobic block, and construct a two-dimensional free energy surface. Using the string method to identify the minimum energy path, we find that all pathways converge to a nearly degenerate reaction channel, irrespective of the initial path. Within this channel, the end-to-end distance of the hydrophobic block exhibits a broad distribution, yet the corresponding expulsion barriers remain nearly indistinguishable. Together, these findings establish a continuum-level theoretical foundation for understanding the hyperstretching mechanism and the transition state ensemble in micellar chain exchange.
Non-equilibrium conformations of dilute star polymers in shear flow
Polymer topology influences the structural and dynamic behavior of macromolecules, particularly under non-equilibrium conditions such as shear flow. As a result, conventional Gaussian chain models fail to capture the complex deformation and relaxation dynamics of polymers with branched architectures. In this study, we present a combination of Brownian dynamics simulations with Gram–Charlier (G–C) expansion analysis to quantify non-Gaussian features using two cumulant-based metrics: the standard deviation σ of the one-dimensional projected configurational distribution, where larger values indicate greater deformation, and the normalized fourth cumulant κ4/σ4, where larger values indicate stronger non-Gaussianity. Focusing on linear and star polymers in dilute solution, we systematically investigate how molecular architecture, finite extensibility, and hydrodynamic interactions (HI) influence their deformation and conformational response in shear flow. Our findings reveal that star polymers exhibit constrained global extension but enhanced local stretching near the core, resulting in reduced deviations from Gaussian behavior compared to linear chains. Finite extensibility imposes an upper limit on bond extension, which suppresses configurational deformation at high shear rates and leads to a non-monotonic trend in κ4/σ4. By contrast, hydrodynamic interactions introduce long-range segmental coupling that enhances coordinated motion and amplifies non-Gaussian character, resulting in higher values of κ4/σ4 under strong shear flow. Together, these results establish a robust framework for characterizing the influence of polymer topology on non-equilibrium conformations, offering new insights into the mechanics of branched polymer systems under flow.
A sensitive method for determining dehydrogenation probabilities at a metal surface
We report a sensitive new approach for tracking surface hydrogen coverage using the novel high-repetition rate implementation of velocity-resolved kinetics in an investigation of methanol-d4 dehydrogenation on Pt(111) to CO and D2. The reactant beam of methanol-d4 had an incidence kinetic energy of 0.46 eV, and the reaction was studied over a surface temperature range of 700–950 K. The reaction probability of methanol dehydrogenation was determined by probing the second-order reaction kinetics of D2 and HD formation, which is directly correlated to the surface hydrogen coverage. Under the conditions of our study, the methanol dehydrogenation probability was found to be low (∼10−3), and the lack of strong temperature dependence is consistent with dehydrogenation via a direct mechanism. Under conditions that favored methanol clustering in the incident methanol beam, the dehydrogenation probability increases significantly at low surface temperatures, which suggests a precursor-mediated mechanism. The possible role of cluster formation in enhancing the trapping probability of methanol is discussed.
Tree tensor networks methods for efficient calculation of molecular vibrational spectra
We develop and employ general tree tensor networks to compute the vibrational spectra for two model systems: a set of 64-dimensional coupled oscillators and acetonitrile. We explore various tree architectures, ranging from the simple linear structure of Matrix Product States (MPS), to trees where only the leaf nodes carry a physical leg—as commonly seen in the underlying ansatz of the multilayer multiconfiguration time-dependent hartree method—and further to more general trees in which all nodes are allowed to possess a physical leg. In addition, we implement locally optimal block preconditioned conjugate gradient methods and inverse iteration methods as eigensolvers. Benchmarking runtime and accuracy shows that all tested topologies can reach high accuracy. For acetonitrile, inverse-iteration refinement brings all 84 computed states below 1 cm−1 error, while the fork-4 tree, a comb-like tree with four backbone nodes, provides the best overall balance between accuracy and cost. MPS remains computationally attractive, whereas more connected trees generally improve accuracy at fixed bond dimension. All numerical simulations were performed using PyTreeNet, a Python package designed for flexible tensor network computations.
Interpretation of permittivity values in vapor deposited thin films of organic glasses
Permittivity values of thin glassy films obtained by physical vapor deposition onto interdigitated electrode cells have been reported to increase considerably after annealing the sample above its glass transition temperature, which cannot be explained by effects of density. While such a change might be interpreted as a transformation to a different structure or altered dynamics, we employ a nonpolar molecular liquid to demonstrate that the apparent permittivity derived from impedance experiments can increase without changes in properties inherent in the sample material. More specifically, it is shown that annealing induces surface flattening that leads to material relocating from on top of the metallic electrode digits to between these digits. For films of thicknesses below 10% of the digit spacing, we demonstrate that the impedance measurement senses predominantly the material located between electrode digits, while being blind to the part of the sample deposited onto the digits. This implies that the apparent permittivity can increase without a change in material properties. A relation that determines film height and deposition rate from the observed capacitance increment is provided.
From full dynamic to pure static: A family of <i>GW</i> -based approximations
We introduce a systematic hierarchy of one-body Green’s function methods derived from the GW approximation, constructed by progressively reducing the dynamical content of the self-energy. Starting from the fully dynamical Dyson formulation, we generate a family of approximations that interpolates between the standard GW approximation and purely static effective single-particle Hamiltonians. This framework enables a controlled investigation of the role of dynamical effects and particle–hole coupling in the description of ionization potentials. Within this unified formalism, the hole and particle branches can be selectively decoupled through downfolding strategies into reduced one-particle spaces. By benchmarking the different members of this hierarchy on molecular ionization energies, we assess their accuracy, numerical robustness, and algorithmic complexity. We demonstrate that consistently derived partially static schemes can yield reliable quasiparticle energies while significantly simplifying the underlying eigenvalue problem. We further introduce a novel static Hermitian self-energy obtained as the static limit of this hierarchy. Despite its conceptually distinct origin, it produces results remarkably close to those of qsGW, thereby providing an alternative static route toward partial self-consistency.
(SOS1-)PBE-DH-INVEST double hybrid functionals for INVEST systems: Validation for azaphenalene compounds
INVEST systems (INVErted Singlet–Triplet excited-state energies) have attracted increasing interest due to their potential relevance in organic light-emitting diodes and photocatalytic applications. However, their accurate theoretical description remains challenging because the singlet–triplet energy gap, ΔEST, is strongly governed by electron-correlation effects and double-excitation contributions, which are not properly captured by conventional time-dependent density functional theory approaches. In this work, the recently developed double-hybrid density functional PBE-DH-INVEST, together with its spin-opposite-scaled variant SOS1-PBE-DH-INVEST, is assessed for the AP117 benchmark set of azaphenalene derivatives. Excitation energies for the S1 ← S0 and T1 ← S0 transitions, as well as the corresponding ΔEST values, were computed using the aug-cc-pVDZ basis set within the Tamm–Dancoff approximation and compared against CC2 reference data. The results show that PBE-DH-INVEST provides a robust description of both excitation energies and singlet–triplet gaps, with low deviations relative to the reference values and a marked improvement over the corresponding self-consistent hybrid treatment lacking double-excitation corrections. In particular, the inclusion of the Δ(D) contribution for double excitations is shown to be essential for reproducing the inverted energetic ordering characteristic of INVEST systems. The SOS1-PBE-DH-INVEST variant also yields competitive results at a reduced formal computational cost, although with slightly larger errors.
Effect of pre-shear and dispersity on crystallization of a model polymer with soft pair interactions using molecular dynamics simulations
Polymer crystallization is a process of great interest in both fundamental theory and industrial settings, particularly in polymer processing and applications involving semi-crystalline materials. The effect of processing on the initial stages of crystallization is not fully understood. Our study investigates the influence of pre-shear on monodisperse melts and bidisperse blends of a generic, segmentally coarse-grained polymer model. Through molecular dynamics simulations, we explore how polydispersity affects crystallization, where we found that the addition of short chains to a melt of longer chains increased the end-of-simulation crystallinity by about 10% and increased the initial growth rate by roughly a factor of two. In contrast, however, pre-shearing the hot melt before quenching only showed a minor increase in both growth rates and final crystallinity, except in monodisperse melts of short chains. Crystal grain shapes were most influenced by pre-shearing monodisperse melts, where both asphericity and prolateness decreased. In addition, we determined topological connectivity of crystal grains through tie- and loop-chain analysis. Again, only monodisperse melts showed a significant increase of tie chain fractions with pre-shear, while all other systems showed only modest increases. Our findings provide insight into the changes of crystallinity and cluster morphologies that emerge when pre-sheared, offering a deeper understanding of the initial crystallization processes in polymer melts when subjected to pre-shear.
Tribological universality in compressed polymer brushes
Compressed polymer brushes exhibit mutual interpenetration that is linked to the tribological properties of brush-coated surfaces. Interpenetration is quantified by the overlap function, defined as the product of the density profiles of two opposing brushes. Using numerical self-consistent field calculations, we show that the shape of the overlap function is remarkably invariant across a broad class of symmetric and asymmetric brush pairs. The brush parameters that varied include chain length, grafting density, chain stiffness, and degree of branching. In the non-draining approximation and at low shearing rates, the overlap integral defines the sliding friction force. We find that normalizing the friction force vs external pressure curves by their values at a fixed reference pressure collapses the data onto a universal master curve for all brush pairs sharing the same equation of state, regardless of all the other parameters, which should simplify the design of lubricating coatings.
Long-term cultural continuity across the Neanderthal–modern human sequence at Üçağızlı II Cave, northern Levant
We present evidence from Üçağızlı II Cave in the northern Levant, Türkiye, documenting the sequential occupation by Neanderthals ( Homo neanderthalensis ; 77–59 ka) and modern humans ( Homo sapiens ; 59–47 ka). The combined evidence of human fossils, faunal and floral remains, lithics, and manuports demonstrates significant behavioral and technological continuity across this taxonomic sequence, characterized by consistent subsistence strategies and the persistent selection and transport of specific nondietary mollusk shells. The lithic assemblages largely align with late Middle Paleolithic traditions, exhibiting characteristic Mousterian variants. The association of diagnostic human fossils with these archaeological remains demonstrates that the shift from Neanderthals to modern humans occurred within a cultural continuum. Our findings suggest shared behaviors between Neanderthals and modern humans that extended beyond subsistence to include nonutilitarian behaviors within the specific geographic and temporal context studied here.
On the role of symmetry in quenching OH tunneling in 2,6-dimethylphenol
We model the spectroscopy of the two methyl torsional degrees of freedom coupled to the OH torsional motion in 2,6-dimethylphenol. Recent gas-phase rotational transitions [Welsh et al., J. Phys. Chem. Lett. 17, 3749–3758 (2026)] have been interpreted to be the result of certain symmetry states of the methyl groups leading to the quenching of the OH torsional motion. Both a three-dimensional model and an adiabatic model, in which the OH torsion is treated as the slow mode, are developed to test these conjectures. Good agreement is found between the exact and approximate adiabatic models. The adiabatic model is developed to interpret and elucidate the key couplings leading to the quenching of the tunneling splitting.
Extraction of slip velocity in NEMD Couette flow systems using frictional dissipation
Velocity slip at the solid–fluid (SF) interface plays a key role in fluid transport at the nanoscale, and the SF friction coefficient has been extensively studied because it indicates the degree of slippage. Owing to the scale of this phenomenon, molecular dynamics (MD) simulations are commonly employed using two major approaches: the Green–Kubo integral method in equilibrium MD (EMD) and the direct calculation of friction force and slip velocity in non-equilibrium MD (NEMD) systems under shear. Regarding the latter, a strict definition of the slip velocity is missing due to the nonzero thickness of the boundary at the microscale, and the average velocity of the first adsorption layer or the velocity at the boundary obtained by extrapolation or interpolation is often used. In this study, we propose an alternative description of the slip velocity based on a thermal perspective from the two different scales, i.e., at the macroscale, frictional heat is defined as the product of the friction force and slip velocity, whereas at the microscale, it can be expressed as the sum of the works exerted on the fluid and solid by each other. By combining the two different scales, we defined the slip velocity based on the dissipation induced at the SF interface under shear, which avoids the arbitrariness in the slip velocity at the microscale.
Efficient grand canonical global optimization with on-the-fly-trained machine-learning interatomic potentials
The characterization of nanostructured materials under reactive environments is challenging due to the complexity of the structural motifs involved and their chemical transformations. Global optimization approaches allow predicting stable structures for targeted materials, but addressing the configurational and compositional search spaces is both computationally demanding and inefficient, especially when first-principles calculations are required. In this work, we implement and evaluate a computationally efficient grand canonical global optimization algorithm able to identify stable structures and chemical states of targeted systems under given reaction conditions (e.g., reactant pressure and temperature). The algorithm leverages an on-the-fly trained machine-learning interatomic potential based on sparse Gaussian process regression and the smooth overlap of atomic positions descriptor to reduce the number of first-principles energy evaluations carried out during global optimization searches. The ab initio thermodynamics framework is incorporated to approximate the Gibbs energy of evaluated candidates, performing environment-aware optimizations over multiple stoichiometries. We demonstrate the computational performance of this approach and its ability to reproduce some literature examples.