Browse Articles

The density isobar of water: A comparative study of vdW-DF-cx and RPBE-D3

Accurately modeling volume-dependent properties of water remains a challenge for density functional theory (DFT), with widely used functionals failing to reproduce key features of the water density isobar, including its shape, density, and temperature of the density maximum. Here, we compare the performance of the RPBE-D3 and vdW-DF-cx functionals using replica exchange molecular dynamics (MD) driven by machine-learned force fields. Our simulations reveal that vdW-DF-cx predicts the water density more accurately than RPBE-D3 and reproduces the isobar closely between 307 and 340 K. In contrast, RPBE-D3 underestimates the density across the entire temperature range. However, vdW-DF-cx predicts the maximum density temperature to be ∼30 K higher than experiment. Using the local structure index, we attribute this shift to an onset of low-density, ice-like structures in the vdW-DF-cx-based MD at too high temperatures. Static DFT calculations on water dimers and representative high- and low-density water structures reveal that key features of the density isobars are reflected in the static energy-volume curves. In particular, the equilibrium intermolecular distance and curvature correlate with the maximum density and curvature around the maximum of the density isobar. Similarly, the early onset of the low-density structure is connected to the energetic preference for more structured, low-density water over a less ordered, high-density water structure. Decomposing the exchange–correlation energy reveals that the non-local dispersion energy decisively influences the predicted equilibrium intermolecular distances, whereas the semi-local part governs the balance between low- and high-density liquid structures.

Models for polymer dynamics from dimensionality reduction techniques

Polymer dynamics is analyzed through the lens of linear dimensionality reduction methods, in particular principal and time-lagged independent component analysis (tICA). For a polymer undergoing ideal Rouse dynamics, the slow modes identified by these transformations coincide with the conventional Rouse modes. When applied to the Fourier modes of the segment density, we show that tICA generates dynamics equivalent to dynamic self-consistent field theory (D-SCFT) with a wavevector-dependent Onsager coefficient and a free energy functional subject to the random phase approximation. We then introduce a hidden variable method and a time-local approach to include temporal memory in the tICA-generated dynamics and generalize it to construct continuum models for the nonequilibrium case of spinodal decomposition of a symmetric diblock copolymer melt.

Control simulations of many-body quantum systems by a synergism of discrete real-time learning and optimal control theory

We present a self-consistent algorithm for optimal control simulations of many-body quantum systems. The algorithm features a two-step synergism that combines discrete real-time machine learning (DRTL) with Quantum Optimal Control Theory (QOCT) using the time-dependent Schrödinger equation. Specifically, in step (1), DRTL is employed to identify a compact working space (i.e., the important portion of the Hilbert space) for the time evolution of the many-body quantum system in the presence of a control field (i.e., the initial or previously updated field), and in step (2), QOCT utilizes the DRTL-determined working space to find a newly updated control field for a chosen objective. Steps 1 and 2 are iterated until a self-consistent control objective value is reached such that the resulting optimal control field yields the same targeted objective value when the corresponding working space is systematically enlarged. To demonstrate this two-step self-consistent DRTL-QOCT synergistic algorithm, we perform optimal control simulations of strongly interacting 1D as well as 2D Heisenberg spin systems. In both scenarios, only a single spin (at the left end site for 1D and the upper left corner site for 2D) is driven by the time-dependent control fields to create an excitation at the opposite site as the target. It is found that, starting from all spin-down zero excitation states, the synergistic method is able to identify working spaces and convergence of the desired controlled dynamics with just a few iterations of the overall algorithm. In the cases studied, the dimensionality of the working space scales only quasi-linearly with the number of spins.

Comparing abstraction and exchange channels in the H + HBr reaction: A stereodynamical control perspective

This study investigates the stereodynamical control of the H + HBr (v = 0, j = 1) reaction within 0.01–1.50 eV collision energy using the time-dependent wave packet method. The key findings reveal a clear β-dependent (β is the angle of alignment) scattering behavior: the β = 90° configuration in the abstraction channel enhances reactivity and dominates the formation of the products at lower vibrational states with increasing collision energy. In contrast, the β = 0° configuration promotes all vibrational states in the exchange channel. Notably, the β = 45° configuration displays the smallest cross sections in both channels due to destructive quantum interference, contrasting with the constructive interference in the β = 90° configuration. Channel competition analysis demonstrates that β = 0°/45° configurations enhance exchange channel dominance, whereas β = 90° favors the abstraction channel. The differential cross section shows that the products in the abstraction channel shift from backward to forward in the β = 0°/45° configuration, while maintaining sideways distributions in the β = 90° configuration, and in the exchange channel, it is always backward scattering. The highest reaction rate in the abstraction channel occurs at the parallel alignment in the temperature region between 200 and 1000 K.

Femtosecond-laser-induced optical confinement with ping-pong motion

We introduce a novel method using a kilohertz (kHz) amplified 800 nm laser for the first experimental confinement of microparticles within a single beam. This study demonstrates that high-energy kHz pulses can confine 1-μm-radius polystyrene beads in water within ∼26 μm. This approach utilizes the unique properties of high-energy pulsed lasers, distinct from continuous-wave and megahertz pulsed lasers traditionally used in optical trapping. The pulsing nature of the kHz laser generates strong instantaneous forces that attract and confine particles within a specific region, inducing a “ping-pong” motion within the confined space. When the laser pulses strike the microparticles, the strong gradient forces pull the particles toward the laser focus, while the scattering force from the laser pushes them away. This interaction creates a dynamic equilibrium, causing the particles to oscillate continuously in a back-and-forth motion until the laser is blocked. This phenomenon differs from conventional optical trapping, which offers unique particle confinement possibilities. When combined with optical trapping, especially at femtomolar concentrations or single-particle conditions, this novel development shows that the kHz laser draws particles from significant distances toward the focal point, enhancing its trapping efficiency. All experiments were conducted on a single setup, varying only laser characteristics, ensuring high credibility in the results.

Tree tensor network hierarchical equations of motion based on time-dependent variational principle for efficient open quantum dynamics in structured thermal environments

We introduce an efficient method, TTN-HEOM, for exactly calculating the open quantum dynamics for driven quantum systems interacting with highly structured bosonic baths by combining the tree tensor network (TTN) decomposition scheme with the bexcitonic generalization of the numerically exact hierarchical equations of motion (HEOM). The method yields a series of quantum master equations for all core tensors in the TTN that efficiently and accurately capture the open quantum dynamics for non-Markovian environments to all orders in the system–bath interaction. These master equations are constructed based on the time-dependent Dirac–Frenkel variational principle, which isolates the optimal dynamics for the core tensors given the TTN ansatz. The dynamics converges to the HEOM when increasing the rank of the core tensors, a limit in which the TTN ansatz becomes exact. We introduce TENSO, tensor equations for non-Markovian structured open systems, as a general-purpose Python code to propagate the TTN-HEOM dynamics. We implement three general propagators for the coupled master equations: two fixed-rank methods that require a constant memory footprint during the dynamics and one adaptive-rank method with a variable memory footprint controlled by the target level of computational error. We exemplify the utility of these methods by simulating a two-level system coupled to a structured bath containing one Drude–Lorentz component and eight Brownian oscillators, which is beyond what can presently be computed using the standard HEOM. Our results show that the TTN-HEOM is capable of simulating both dephasing and relaxation dynamics of driven quantum systems interacting with structured baths, even those of chemical complexity, with an affordable computational cost.

Native crystal growth in 60 nm Sb2S3 amorphous film: A joint microscopy–calorimetry study

Joint direct microscopy–calorimetry measurements of crystal growth were performed for a 60 nm amorphous Sb2S3 film deposited either on a Kapton foil or on a soda-lime glass. Calorimetric crystallization proceeded in two steps, originating either from mechanical and stress-induced defects (230–275 °C) or from homogeneously formed nuclei (255–310 °C); both processes exhibited an identical activation energy of 200 kJ mol−1. At temperatures <230 °C, a Sb2O3 crystalline phase formed along the rhombohedral Sb2S3 structure. The normal growth model with the activation energy of ∼250 kJ mol−1 was used to describe the microscopic crystal growth rate data, and the viscosity–diffusivity decoupling was characterized by Ediger’s parameter ξ varying between 0.40 and 0.55. The crystal growth rate was slightly higher in the film deposited on the glass substrate, with the compressive stress introduced at higher T having only a small effect. Meanwhile, the deposition on the glass substrate led to a significantly higher (especially below the glass transition temperature) nucleation rate, which underlines the key aspect of the crystallization process in very thin chalcogenide films: the formation of nuclei due to the internal stresses arising from the difference of the film/substrate thermal expansion coefficients.

Dynamics of polymer rings: The Rouse ring chain with attractive harmonic potential of spherical symmetry

The static and dynamic properties of a cyclic Rouse chain modified by the introduction of an effective, spherically symmetric, attracting potential of entropic nature are studied. It is shown that a relatively weak potential can lead to a strong contraction of the polymer chain: the radius of gyration becomes much smaller compared to the size of the free cyclic chain. The pronounced decrease in the terminal relaxation time of cyclic macromolecules in the presence of a harmonic potential compared to the Rouse relaxation time leads to a lengthening of the time interval for the transition to the normal, i.e., the Fickian, diffusion regime, generating a quasi-plateau at increasing molecular mass for the time dependence of segmental mean squared displacement.

Erratum: “A basis-free phase space electronic Hamiltonian that recovers beyond Born–Oppenheimer electronic momentum and current density” [J. Chem. Phys. 162, 144111 (2025)]

Strain-induced instabilities of graphene under biaxial stress

The mechanical properties of graphene are investigated using classical molecular dynamics simulations as a function of temperature T and external stress τ. The elastic response is characterized by calculating elastic constants via three complementary methods: (i) numerical derivatives of stress–strain curves, (ii) analysis of cell fluctuation correlations, and (iii) phonon dispersion analysis. Simulations were performed with two interatomic models: an empirical potential and a tight-binding electronic Hamiltonian. Both models predict that the Poisson’s ratio ν of graphene increases monotonically with applied stress. The range of stresses studied spans the entire domain of mechanical stability of the planar structure. Two mechanical instabilities are identified. Under tensile stress, fracture occurs, signaled by the softening of a phonon mode at the Brillouin zone boundary (K point). At the studied temperatures (T < 1500 K), auxetic behavior (ν < 0) appears only at high tensile stresses, near the fracture threshold. Under compressive stress, a spinodal instability associated with long-wavelength wrinkling is observed. Finite-size analysis of this instability at 300 K reveals the existence of a finite surface tension σ in the unstressed membrane, which arises from the anharmonic coupling between out-of-plane and in-plane fluctuations. The surface tension stabilizes the membrane’s flat morphology. In the thermodynamic limit, the onset of the wrinkling instability occurs when the compressive stress matches the surface tension (τ = σ). Under this spinodal condition, the area compressibility modulus B is characterized by a scaling law N−1/2, where N is the number of atoms in the simulation cell.

Bridging electrostatic screening and ion transport in lithium salt-doped ionic liquids

Alkali salt-doped ionic liquids are emerging as promising electrolyte systems for energy applications, owing to their excellent interfacial stability. To address their limited ionic conductivity, various strategies have been proposed, including modifying the ion solvation environment and enhancing the transport of selected ions (e.g., Li+). Despite the pivotal role of electrostatic interactions in determining key physicochemical properties, their influence on ion transport in such systems has received relatively little attention. In this work, we investigate the connection between ion transport and electrostatic screening using atomistic molecular dynamics simulations of 1-butyl-1-methylpyrrolidinium bis(trifluoromethanesulfonyl)imide ([pyr14][TFSI]) doped with lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) at molar fractions xLiTFSI ≤ 0.3. We find that the charge–charge and density–density correlation functions exhibit oscillatory exponential decay, indicating that LiTFSI-doped [pyr14][TFSI] is a charge- and mass-dense system. The electrostatic screening length decreases with increasing LiTFSI concentration, whereas the decay length of the density–density correlation functions remains nearly unchanged. Notably, we find that the xLiTFSI-sensitive screening length serves as a central length scale for disentangling species-specific contributions of ion pairs to collective ion transport upon LiTFSI doping. This framework provides a unifying perspective on the interplay between structure and transport in ionic liquid systems.

Cationic, anionic, and global dynamics in 1-hexyl-3-methylimidazolium halide supercooled ionic liquids studied by 1H and 35Cl magnetic resonance and oscillatory shear rheology

The dynamics of the different constituents of the ionic liquid 1-hexyl-3-methylimidazolium chloride (HmimCl) is investigated using nuclear magnetic resonance including chlorine relaxometry, line shape analysis, and proton-detected diffusometry, as well as frequency-dependent shear mechanical measurements. This combination of techniques is useful to probe the individual motions of the anions and the cations, and the sample's overall flow response. The 35Cl− dynamics appears to be close to the structural (or α-) relaxation as seen by rheology. To examine possible sub-α responses, we scrutinize different representations of the viscoelastic response, including the shear modulus G*, compliance J* = 1/G*, fluidity F* = iωJ*, and viscosity η* = 1/F*, with some of these quantities being more susceptible to low-frequency features than others. This way, we are able to detect supramolecular rheological signatures not only for HmimCl but also for 1-hexyl-3-methylimidazolium bromide and 1-hexyl-3-methylimidazolium iodide. These results call for caution in the course of choosing particular response functions when estimating the degree of decoupling between the mesoscale dynamics and the structural rearrangements in ionic liquids.

Sub-1 cm−1 high-resolution broadband sum-frequency generation vibrational spectroscopy (HR-BB-SFG-VS) with significantly improved sensitivity and signal-to-noise ratio (SNR)

Sum-frequency generation vibrational spectroscopy (SFG-VS) has been well-established as a unique spectroscopic probe to interrogate the structure, interaction, and dynamics of molecular interfaces, with sub-monolayer sensitivity and broad applications. Sub-1 cm−1 High-Resolution Broadband SFG-VS (HR-BB-SFG-VS) has shown advantages with high spectral resolution and accurate spectral line shape. However, due to the lower peak intensity for the long picosecond pulse used in achieving sub-wavenumber resolution in the HR-BB-SFG-VS measurement, only molecular interfaces with relatively strong signal have been studied. To achieve detailed understanding and broader applications in molecular interfacial studies with HR-BB-SFG-VS, higher sensitivity and better signal-to-noise ratio (SNR) for HR-BB-SFG-VS is required. In this report, we present a systematic effort on the significant improvement of sensitivity and SNR for HR-BB-SFG-VS. Through optimization of laser pulse characteristics, automatic sample height control, and detection conditions, the sensitivity of HR-BB-SFG-VS was improved, reaching a level of 3 × 10−6 of the SFG signal from the α-quartz standard. The high SNR spectra of various molecular interfaces are thus obtained with exquisite line shapes and fine spectral features. To name a couple of examples, a new hydrogen-bonded water band around 3300 cm−1 can be explicitly identified in the air/neat-water interface spectra and pure chiral spectral peaks at the level of 1 × 10−5 of the quartz signal were measured at the air/Leucine aqueous solution interface etc. Such improvements in sensitivity and SNR in HR-BB-SFG-VS have brought and shall bring new opportunities and new discoveries with broad applications to molecular interface studies, in addition to the advantage of HR-BB-SFG-VS for its sub-wavenumber spectral resolution and the ability for intrinsic spectral line shape.

Rethinking the Kohn–Sham inverse problem

Density functional theory (DFT) is a cornerstone of modern electronic structure theory. In the Kohn–Sham scheme, the many-electron Schrödinger equation is replaced by a set of effective single-particle equations. Thus, the full complexity of the quantum mechanical many-particle effects is mapped to the exchange–correlation potential vxc(r). Almost all DFT calculations done in practice rely on approximations to vxc(r). However, numerical representations of the quasi-exact vxc(r) can be obtained from quasi-exact densities by inverting the Kohn–Sham procedure. This inverse Kohn–Sham scheme is an important source of insight into exact DFT. Here, we review the inverse Kohn–Sham problem and explain in detail several aspects of why Kohn–Sham inversion is intrinsically difficult. We then present several inversion schemes and discuss their pros and cons, specifically addressing the effects of statistical uncertainties that are inevitable in quantum Monte Carlo reference densities. We use these schemes to obtain representations of vxc(r) that correspond to the ground-state densities that have become available from accurate diffusion Monte Carlo calculations on real space grids for the Li2 and N2 molecules, and the C atom. In the latter, the highest occupied orbital has a nodal line and the exchange–correlation potential goes to a different asymptotic value in this direction. As an outlook, we discuss the possibility of interlacing quantum Monte Carlo and Kohn–Sham theory by using the quasi-exact Kohn–Sham determinant to fix the nodes in a diffusion Monte Carlo calculation.

Adversarial training for dynamics matching in coarse-grained models

Molecular dynamics simulations are essential for studying complex molecular systems, but their high computational cost limits scalability. Coarse-grained (CG) models reduce this cost by simplifying the system, yet traditional approaches often fail to maintain dynamic consistency, compromising their reliability in kinetics-driven processes. Here, we introduce an adversarial training framework that aligns CG trajectory ensembles with all-atom (AA) reference dynamics, ensuring both thermodynamic and kinetic fidelity. Our method adapts the adversarial learning paradigm, combining a physics-based generator with a neural network discriminator that differentiates between AA and CG trajectories. By adversarially optimizing CG parameters, our approach eliminates the need for predefined kinetic features. Applied to liquid water, it accurately reproduces radial and angular distribution functions as well as dynamical mean squared displacement, even extrapolating long-timescale dynamics from short training trajectories. This framework offers a new approach for bottom-up CG modeling, offering a systematic and principled way to preserve dynamic consistency in complex coarse-grained molecular systems.

Author Correction: Causal disentanglement for single-cell representations and controllable counterfactual generation

Temporal control of human DNA replication licensing by CDK4/6-RB signalling and chemical genetics

Abstract Cyclin-dependent kinases (CDKs) coordinate DNA replication and cell division, and play key roles in tissue homeostasis, genome stability and cancer development. The first step in replication is origin licensing, when minichromosome maintenance (MCM) helicases are loaded onto DNA by CDC6, CDT1 and the origin recognition complex (ORC). In yeast, origin licensing starts when CDK activity plummets in G1 phase, reinforcing the view that CDKs inhibit licensing. Here we show that, in human cells, CDK4/6 activity promotes origin licensing. By combining rapid protein degradation and time-resolved EdU-sequencing, we find that CDK4/6 activity acts epistatically to CDC6 and CDT1 in G1 phase and counteracts RB pocket proteins to promote origin licensing. Therapeutic CDK4/6 inhibitors block MCM and ORC6 loading, which we exploit to trigger mitosis with unreplicated DNA in p53-deficient cells. The CDK4/6-RB axis thus links replication licensing to proliferation, which has implications for human cell fate control and cancer therapy design.

Greenland ice sheet runoff reduced by meltwater refreezing in bare ice

Abstract The contribution of Greenland Ice Sheet meltwater runoff to global sea-level rise is accelerating due to increased melting of its bare-ice ablation zone. There is growing evidence, however, that climate models overestimate runoff from this critical area of the ice sheet. Climate models traditionally assume that all bare-ice runoff enters the ocean, unlike porous firn, in which some meltwater is retained and/or refrozen. We used field measurements and numerical modeling to reveal that extensive retention and refreezing also occurs in bare glacier ice. We found that, from 2009 to 2018, meltwater refreezing in bare, porous glacier ice reduced runoff by an estimated 11–17 Gt a−1 in southwest Greenland alone, equivalent to 9–15% of this sector’s annual meltwater runoff simulated by climate models. This mass retention explains evidence from prior studies of runoff overestimation on bare ice by current generation climate models and may represent an overlooked buffer on projected runoff increases. Inclusion of bare-ice retention and refreezing processes in climate models therefore has immediate potential to improve forecasts of ice sheet runoff and its contribution to sea-level rise.

Flexynesis: A deep learning toolkit for bulk multi-omics data integration for precision oncology and beyond

Abstract Accurate decision making in precision oncology depends on integration of multimodal molecular information, for which various deep learning methods have been developed. However, most deep learning-based bulk multi-omics integration methods lack transparency, modularity, deployability, and are limited to narrow tasks. To address these limitations, we introduce Flexynesis, which streamlines data processing, feature selection, hyperparameter tuning, and marker discovery. Users can choose from deep learning architectures or classical supervised machine learning methods with a standardized input interface for single/multi-task training and evaluation for regression, classification, and survival modeling. We showcase the tool’s capability across diverse use-cases in precision oncology. To maximize accessibility, Flexynesis is available on PyPi, Guix, Bioconda, and the Galaxy Server (https://usegalaxy.eu/). This toolset makes deep-learning based bulk multi-omics data integration in clinical/pre-clinical research more accessible to users with or without deep-learning experience. Flexynesis is available at https://github.com/BIMSBbioinfo/flexynesis.

Matrix directs trophoblast differentiation in a bioprinted organoid model of early placental development

Abstract Trophoblast organoids can provide crucial insights into mechanisms of placentation, however their potential is limited by highly variable extracellular matrices unable to reflect in vivo tissues. Here, we present a bioprinted placental organoid model, generated using the first trimester trophoblast cell line, ACH-3P, and a synthetic polyethylene glycol (PEG) matrix. Bioprinted or Matrigel-embedded organoids differentiate spontaneously from cytotrophoblasts into two major subtypes: extravillous trophoblasts (EVTs) and syncytiotrophoblasts (STBs). Bioprinted organoids are driven towards EVT differentiation and show close similarity with early human placenta or primary trophoblast organoids. Inflammation inhibits proliferation and STBs within bioprinted organoids, which aspirin or metformin (0.5 mM) cannot rescue. We reverse the inside-out architecture of ACH-3P organoids by suspension culture with STBs forming on the outer layer of organoids, reflecting placental tissue. Our bioprinted methodology is applicable to trophoblast stem cells. We present a high-throughput, automated, and tuneable trophoblast organoid model that reproducibly mimics the placental microenvironment in health and disease.