Browse Articles
Discover research articles across all indexed journals
Physics-informed transfer learning via frontier orbital pretraining for prediction of polymer electronic properties
Accurate prediction of electronic properties, including bandgap, ionization energy (IE), and electron affinity (EA), is central to the design of polymer electronic materials but is hindered by the vast chemical space and the high cost of reliable reference data. Here, a frontier orbital-guided learning framework is proposed that integrates low-cost quantum chemical pretraining with transfer learning to enable efficient and physically consistent prediction of polymer electronic properties. The model is pretrained on GFN2-xTB-derived frontier orbital properties of polymer trimers and subsequently fine-tuned using limited highfidelity data to predict chain bandgap (bandgap-chain), bulk bandgap (bandgap-bulk), IE, and EA. The resulting models exhibit consistently high predictive accuracy across all target properties, with test-set mean absolute errors of 0.246 eV for bandgap-chain, 0.269 eV for bandgap-bulk, 0.169 eV for IE, and 0.136 eV for EA, corresponding to RMSE values below 0.360 eV, while maintaining strong correlation with reference data (R2 > 0.90) and preserving key physical behaviors, including chain-length scaling and inter-property consistency. Leveraging this framework, electronic properties of ∼12 × 106 polymer repeat units are predicted, enabling statistically robust fragment-level analysis in which the observed trends remain consistent with established physical intuition and known structure–property relationships. This work provides a scalable and data-efficient framework for machine learning-assisted screening and design of polymer electronic materials.
Higher order Magnus expansion for driven two-level quantum dynamics
We investigate the Magnus expansion for a generic time-dependent two-level system under single-axis driving. By virtue of the su(2) Lie algebra, the expansion is decomposed into a commutator-free form. To illustrate the usefulness of the gained expression, we then revisit the Landau–Zener–Stückelberg–Majorana model, with a focus on non-adiabatic transitions as well as the Stokes phase. In addition, the semiclassical Rabi model is systematically treated by determining the Floquet quasienergy up to different orders. We demonstrate how to employ suitable picture transformations as well as how to enforce the symmetry of the underlying model to guarantee convergence of the expansion as well as to achieve satisfactory agreement with the exact results. For both models that we studied, it turns out that a third order approximation yields results that are in next to perfect agreement with exact analytical ones. Surprisingly, in the case of the semiclassical Rabi model, even the second order Magnus approximation in the adiabatic picture produces almost exact results for a large parameter range.
Holistic simulation of iron–sulfur cluster electronic and physical structures with hybrid density functional approximation reduced density matrix functional theory
[Fe–S] clusters are privileged and highly conserved metallocofactors that perform a wide range of biological functions, including redox catalysis and small molecule activation. Their reactivity is largely owed to their manifold of energetically low-lying, near-degenerate d-orbitals, resulting in a highly multi-reference, or strongly correlated, electronic structure. This results in not only a large number of electronic degrees of freedom but also a delicate interplay with the geometric configuration of the cluster core. Due to the size and computational complexity of these clusters, their larger-scale simulation has traditionally been limited to single-reference density functional theory (DFT), which struggles to capture strong-correlation effects. This approach leads to significant uncertainties not only in the predicted electronic properties of the [Fe–S] cluster but also in their optimized geometries, resulting in limitations to the ability of simulations to serve as a predictive tool in [Fe–S] chemistry. In a step to overcome these limitations, we employ a methodology based on combining existing, traditional density functionals with a 1-electron reduced density matrix functional (DFA 1-RDMFT), which captures strong correlation effects via fractional orbital occupation, while retaining the low computational scaling of DFT. We apply this approach to both simulate the electronic structure and optimize the geometries of a set of site-differentiated [Fe4S4]+ clusters coordinated by a series of electronically diverse ligands, demonstrating the ability of DFA 1-RDMFT to capture the delicate interplay between the electronic and physical structures in [Fe4S4] clusters.
An argument why the spinterface model cannot explain the chirality induced spin selectivity effect
In the context of chirality induced spin selectivity effect, it has been argued that a chiral molecule when adsorbed on a metal facilitates the formation of a local spin moment at the interface between the metal and molecule, given a strong spin–orbit coupling in the metal. The possibility for such spin moment formation is analyzed in terms of general arguments and effective modeling of a pertinent setup. The conclusion from this analysis is that a strong spin–orbit coupling in the metal does not provide a sufficient mechanism to sustain a stabilized spin moment at the interface. It is, moreover, shown that an electron flux into or out of the molecule does not provide conditions for a spin moment formation, regardless of whether the flux is spin-polarized or not.
Estimating memory time within the frameworks of generalized quantum master equation and transfer tensor methods
Simulating long-time nonadiabatic dynamics in condensed-phase systems is computationally demanding due to the inherent non-Markovianity of the electronic reduced density matrix evolution. While the generalized quantum master equation (GQME) and transfer tensor method (TTM) allow for the reconstruction of long-time dynamics from short-time projection-free inputs, their accuracy hinges on the rigorous estimation of the memory time, a parameter often determined by heuristic trial-and-error. In this work, we establish a comprehensive framework for estimating memory time and benchmarking propagation accuracy using semiclassical and numerical exact inputs on both standard spin-boson models and general multistate harmonic models. We develop an error estimation scheme that reveals a characteristic three-stage decay pattern in the non-Markovian propagation error: an initial transient drop, an exponential decay, and a saturation plateau. This estimator serves as a critical diagnostic tool for GQME and TTM, successfully distinguishing between converged predictions and reliability failures in complex systems, such as the carotenoid–porphyrin–fullerene triad. These findings provide a robust, quantitative protocol for validating memory-kernel-based simulations of nonadiabatic dynamics.
<i>mrfmsim</i> : A modular, extendable, and readable simulation package for magnetic resonance force microscopy experiments
We present mrfmsim, an open-source Python package that facilitates the design, simulation, and analysis of magnetic resonance force microscopy (MRFM) experiments. MRFM is a scanning-probe technique that detects magnetic resonance from nanoscale ensembles of nuclear or electron spins with a force sensor. Because MRFM experiments are complex and operate at sensitivity limits, numerical simulation is essential for designing experiments and estimating per-spin sensitivity and imaging resolution from measured signals. In this paper, we highlight the challenges of developing MRFM simulations and show that software designed to simulate specific experiments only in a rapidly evolving experimental field can yield erroneous results. The mrfmsim package addresses these challenges by supporting post-definition customization without rewriting the internal model and by employing a plugin system for extending functionality. We show that the package’s modular, extendable, and readable architecture improves reproducibility and accelerates development.
An efficient hybrid spectral-compact difference scheme for rod–coil diblock copolymers in slit confinement
Self-consistent field theory simulations of rod–coil diblock copolymers in slit confinement present significant numerical challenges due to sharp density gradients near hard walls. To rigorously resolve these systems utilizing the Gaussian and wormlike chain models, a hybrid spectral-compact finite difference scheme is developed on a non-uniform Chebyshev–Gauss–Lobatto grid. Shen’s Chebyshev spectral method is employed for the flexible blocks. For the semiflexible blocks, a second-order upwind compact scheme together with an L-stable TR-BDF2 contour-stepping algorithm is adopted. This hybrid framework effectively suppresses spurious numerical oscillations. This unconditionally stable formulation strictly preserves propagator non-negativity and achieves up to a two-orders-of-magnitude speedup over uniform-grid implementations while maintaining linear spatial scaling. Simulations utilizing this advanced framework under neutral wall conditions reveal that the confining walls naturally induce preferential wetting of the semiflexible blocks at the impenetrable boundaries. As the incompressibility penalty increases, the compressible system progressively approaches the incompressible limit. For the selected physical parameters, decreasing the slit width induces a sequence of structural transitions from a smectic-C morphology with three internal periods (SC3) to morphologies with two and one internal periods (SC2 and SC1), and ultimately to a highly compressed smectic-P morphology (SP1). The equilibrium thickness of these confined structures deviates from exact integer multiples of the bulk spatial period. This deviation arises from the volume compensation associated with boundary depletion layers, together with adjustments in the molecular tilt angle and the degree of molecular interdigitation.
Uniaxial order parameters associated with surface SFG spectra. II. Distributions with polar and azimuthal ordering
When a molecule or sub-molecular entity is ordered on a surface in a manner that can be described by a tilt- and azimuthal-angle distribution, the projection of the molecular hyperpolarizability into the laboratory frame can be described using spherical harmonics. We illustrate that this lends itself to the construction of ten achiral order parameters associated with vibrational sum-frequency generation. We describe how these order parameters can be extracted from spectral data, first ignoring and then including the dispersion of the local electric fields. We then use these order parameters to determine the most probable orientation distribution without any assumptions about the surface alignment and in the absence of any other type of experimental data. This constitutes a flexible framework for describing a wide array of surface orientation distributions.
What lies between crystal and randomly packed structures? A general characterization of non-periodic order
In this paper, we address the characterization of the structure of condensed materials, periodic and non-periodic. Carrying out an extensive study of over 7000 different ground-state structures of a 2D lattice model of binary packing, we find a predominance of non-periodic structures (over 96%) that extend across the entire range of possible diversities. These non-periodic structures are resolved by establishing whether a structure will accommodate or reject additional local structures. This property, structural selectivity, is treated as a signature of an underlying ordering principle. The major result of this paper is the determination that roughly 35% of the non-periodic structures are selective and, hence, ordered in some way. This selectivity extends up to a diversity of ∼9, well beyond the upper threshold for diversity in periodically ordered states.
WMS-Rot: From quantum-chemical predictions to rotational spectral assignment and refinement
We present WMS-Rot and its fitting companion WMS-FitRot as an integrated framework for the early stages of rotational spectral analysis, starting from spectroscopic parameters obtained from electronic-structure computations and progressing to assignment-aware local refinement driven by the same theoretical catalog used for prediction. The framework provides a practical and internally consistent route connecting modern composite quantum-chemical predictions to first-pass assignments and controlled refinement. More fundamentally, it reformulates the incorporation of theoretical information into the spectroscopic inverse problem: calculated parameters act not only as initial guesses but also as active constraints that stabilize assignments and guide early-stage refinement within a unified simulation–fit cycle. Applications to nicotinic acid and thiopronine show that accurate composite inputs markedly improve starting points compared to low-level models, enabling robust assignment, reliable conformer discrimination, and consistent refinement. The approach reproduces matched reduced-Hamiltonian fits while remaining fully compatible with standard SPCAT/SPFIT practice and provides diagnostic insight into parameter correlations, identifiability, and model conditioning.
Role of the Casimir force in the capacitive radio frequency microelectromechanical switches
We determine the role of the fluctuation-induced Casimir force acting between a membrane of cylindrical shape and a bottom electrode in microelectromechanical capacitive switches. For this purpose, the Casimir force is computed by taking into account the real properties of both the materials of a membrane and a bottom electrode with an account of surface roughness. The obtained results are compared with those found for the smooth surfaces using the idealization of ideal metal. It is shown that an account of both the real material properties and surface roughness is crucial for obtaining the correct values of the Casimir force. According to our results, at the shortest separations, when the switch membrane is in contact with the transmission line, the magnitudes of the Casimir force may exceed the magnitudes of the electric one, depending on the value of the operating voltage. The obtained values of the Casimir force can be used for determining the thickness of the switch membrane, which ensures the necessary magnitude of the restoring elastic force required for a stable cyclic functioning of the micromechanical switch with no pull-in.
Operational bounds and diagnostics for coherence in energy transfer
Excitation energy transfer in light-harvesting aggregates is highly efficient, yet whether quantum coherence plays an operational role in transport remains debated. A central challenge is that coherence is usually inferred from spectroscopic signatures, whereas transport performance is assessed through specific observables and depends on both the open system dynamics and the initial state preparation. Here, we develop a resource theoretic approach that quantifies the maximum change that initial site-basis coherence can induce in a chosen readout under fixed reduced dynamics. The central quantity is the resource impact functional, which yields state independent, readout specific bounds on coherence-induced changes in signals and transport figures of merit. We apply the framework to two models. For a donor–acceptor dimer, we analyze coherence sensitivity across coupling and bath-timescale regimes and bound trapping efficiency and average transfer time in terms of the impact functional. For a multi-site chain with terminal trapping, we derive rigorous criteria that distinguish population placement from sensitivity to initial state site-basis coherence. These include upper bounds on the largest advantage over incoherent preparations, necessary delocalization requirements for achieving a prescribed improvement, and a simple pairwise sufficient condition that can be checked from local information. For quasi-local reduced dynamics, we further obtain a Lieb–Robinson-type bound that constrains when coherence prepared in a distant region can influence a localized readout at finite times. Together, these results provide operational diagnostics and rigorous bounds for benchmarking coherence effects and for identifying regimes in which they are necessarily negligible or potentially relevant in excitonic transport models.
Simulation of homogeneous electrochemical proton-coupled electron transfer using the hybrid-bath hierarchical equations of motion
The dynamics of homogeneous electrochemical proton-coupled electron transfer (PCET) are governed by the complex interactions among the continuous electronic states of the electrode, molecular vibrational modes, and the solvent environment. Here, we study this process within a Newns–Anderson model using the hierarchical equations of motion (HEOM) method combined with matrix product states (MPS) for hybrid fermionic and bosonic baths. The simulations reveal how the reaction dynamics depend on a variety of parameters, including the proton-transfer distance, electrode chemical potential, molecule–electrode coupling strength, and solvent reorganization energy. Comparison with Fermi’s Golden Rule shows that the perturbative rate theory is reliable in the weak-coupling regime, but may become inaccurate at strong molecule–electrode coupling. Rates extracted from population dynamics yield Tafel plots whose shapes depend on both solvent and electrode couplings. The calculations also reproduce a primary kinetic isotope effect, with hydrogen transfer faster than deuterium transfer and with a larger effective transfer coefficient. These results highlight the capability of the hybrid-bath MPS-HEOM method to provide a unified description of electrochemical PCET in a wide range of parameter regimes.
How to improve the accuracy of semiclassical and quasiclassical dynamics with and without generalized quantum master equations
Semi- and quasi-classical (SC) theories can handle anharmonic interactions and are thus well-suited to predict atomistic quantum dynamics in condensed phases that encode energy and charge transport, spectroscopic responses, and chemical reactivity. However, SC theories can be computationally expensive and inaccurate. When combined with generalized quantum master equations (GQMEs), the resulting SC-GQMEs can enhance the efficiency and accuracy of SC dynamics. Yet, while the origin of improved efficiency is clear, the mechanism that improves accuracy remains elusive. Even worse, SC-GQMEs can yield unphysical dynamics in challenging parameter regimes—a shortcoming that might be avoided if the mechanism of accuracy improvement were understood. Here, we uncover this mechanism. We leverage short-time analyses to prove that exact, “left-handed” time-derivatives delay the onset of SC inaccuracy, even without the GQME. However, these derivatives are a double-edged sword: while offering greater short-time accuracy, they become unphysical in challenging parameter regimes. Because short-lived SC-GQME kernels combine short-time accuracy with long-time stability, we develop a protocol to unambiguously determine the memory kernel cutoff, even in challenging cases where previous treatments had failed. Our protocol employs only SC calculations and combines self-consistency with mixed-accuracy auxiliary kernels to triangulate a propitious kernel cutoff, yielding SC-GQMEs with greater accuracy than SC theory alone, while remaining physical and accurate over arbitrary times. Our insights into accuracy improvement, identification of when the SC-GQME is advantageous, and kernel cutoff protocol are general and can be expected to apply to complex systems that go beyond simple models.
Determination of zero-field splitting in magnetically diluted cobalt(II)-based single molecule magnets using circular dichroism Fourier-transform IR spectroscopy
Single-molecule magnets (SMMs) with total electron spin S = 3/2 are promising platforms for optical magnetization control. Their study in such a context is relevant in both magnetically concentrated and diluted samples and assumes the excitation of magnetic dipole transitions by resonant radiation. Such experiments require precise determination of the energy splitting between Kramers doublets (EZFS), since this value determines the resonant radiation frequency. The direct spectroscopic method available for such measurements is frequency-domain Fourier transform terahertz electron paramagnetic resonance (FD-FT THz-EPR) spectroscopy. It usually requires a setup that includes an FTIR-spectrometer and a superconducting magnet to create a magnetic field in the sample area. In this work, we propose a method for the rapid determination of EZFS of SMMs without using a superconducting magnet. The method is based on circular dichroism of an SMM placed in a static external magnetic field. It was implemented using a standard FTIR spectrometer with a sample holder modified by inserting a permanent magnet. Circular polarization was created using a grid polarizer and a quarter-wave plate. Model Co(II)-based SMMs with a varying degree of magnetic dilution were investigated by the proposed method. The results were confirmed by FD-FT THz-EPR. The obtained EZFS values of 43.4–43.6 cm−1 demonstrate a slight downward trend with a decrease in the paramagnetic center content. Such preliminary measurements of EZFS in magnetically diluted SMMs pave the way for the resonant excitation of magnetic dipole transitions using narrowband sources and even for coherent control of magnetic quantum states.
Molecular dynamics insights into water confined in zeolite-templated carbon nanomaterials
Water confined in zeolite-templated carbons (ZTCs) exhibits properties fundamentally different from those of bulk liquid, with profound implications for energy storage, separation technologies, and catalysis. Despite the technological importance of water behavior in ZTC nanopores, molecular-level understanding remains limited. This work presents comprehensive molecular dynamics (MD) simulations investigating the structure, dynamics, and hydrogen bonding characteristics of water confined within faujasite-derived ZTC. Classical MD simulations were developed with validated force fields to characterize radial and spatial distribution functions, hydrogen bond networks and lifetimes, cluster size distributions, domain formation, translational and rotational dynamics, and velocity autocorrelation functions. Systematic comparison with bulk liquid water reveals confinement-induced modifications to tetrahedral hydrogen bonding networks, spatial organization into discrete domains, hydrogen bond dynamics, and transport properties. The three-dimensional hierarchical pore topology of ZTC creates unique confinement environments distinct from one-dimensional nanotubes or two-dimensional slit pores. These findings provide molecular-level insights essential for the rational design of ZTC-based materials for electrochemical energy storage, water desalination membranes, proton exchange systems, and aqueous-phase catalysis, thereby advancing fundamental understanding of water confinement in complex carbon nanostructures.
Computational hole mobilities in (pseudo-)amorphous organic semiconductors
The development of novel organic semiconductors with enhanced conducting properties is often hindered by the challenge of accurately describing and modeling charge transport within the (pseudo-)amorphous films typically found in optoelectronic devices. In this study, we present a multiscale computational protocol to predict hole mobilities of non-crystalline hole-transporting materials with order-of-magnitude accuracy. Our approach, which integrates density functional theory, molecular dynamics, docking, and kinetic Monte Carlo simulations, reveals the impact of modeling different film morphologies—amorphous and pseudo-amorphous films as well as docking aggregates—on the charge transport properties of these materials. In particular, we demonstrate that experimentally observed mobility trends across a family of ten hole-transporting molecules, including the enhancement associated with increased aromatic core planarity and extended π-conjugation, can only be reproduced when both amorphous disorder and locally ordered molecular aggregates are explicitly considered. This work establishes a robust, morphology-aware framework for the rational, in silico design and optimization of next-generation organic semiconductors.
Brownian dynamics with soft constraints in soft matter systems
Stiff forces, which bind objects together or otherwise confine motion, are found widely in soft-matter systems—e.g., colloids with short-range attractions, ligand–receptor contacts, particles in optical traps, and fibers that resist stretching. To assess the long-term effect of these stiff forces on dynamics and structure, it is useful to consider the limit where they are treated as constraints, so the system evolves strictly within allowed configurations. Efforts to derive equations involving both constraints, and the stochastic motion appropriate at the scales of soft matter, began around 50 years ago, yet we are still lacking a straightforward way to extract the projected equations and apply them in modern formulations of mesoscale dynamics. Here, we address this gap with two key contributions: (1) a practical summary of the constrained Brownian dynamics equations with “soft” constraints, i.e., constraints imposed by stiff forces, which is illustrated through several representative examples, taking care to highlight the nontrivial effects of the constraints, and (2) a derivation using singular perturbation theory, establishing the validity of these equations over timescales exceeding the relaxation of stiffly constrained degrees of freedom. We further extend our approach to “soft–soft” constraints, where mobility varies on lengthscales comparable with the restraining forces—a scenario typical for particles in fluids experiencing hydrodynamic interactions. We hope our results will be useful for soft matter research, as a robust toolkit for studying tethered or confined systems.
One nitrogen atom as electron source is enough for efficient intermolecular charge transfer in organic cocrystals toward NIR-photothermal conversion
In organic donor–acceptor (D-A) materials, a nitrogen atom can serve as a skeleton electron source to tune the inter-/intra-molecular charge transfer (CT) nature and photophysical/photochemical properties. Here, we employ the model molecule stilbene modified with non-/one-/two-nitrogen atoms in the benzene ring skeleton (TSB/TSP/DPE) as a donor and F4TCNQ as an acceptor to synthesize three cocrystals by a one-step mechanochemical method toward NIR photothermal conversion (NIR-PTC). The results show that “One Nitrogen Atom Is Good, Two Are Too Many.” That is, one nitrogen atom in the donor (TSP) is enough to induce optimal performance of the CT effect with the degree of charge transfer of ∼1.16, redshift absorption of ∼2500 nm, and NIR-PTC efficiency of ∼89.4% @ 808 nm. Femtosecond transient absorption spectra exhibit that superior NIR-PTC of TSP-F4TCNQ cocrystal can be attributed to the first ultrafast internal conversion (IC) with a lifetime of 1.7 ± 0.4 ps and subsequent nonradiative decay (NR) via rotation/twisting of cyano-groups out of the plane of central ring of F4TCNQ with a lifetime of 17.1 ± 3.4 ps, followed by the third vibrational relaxation (VR) with a lifetime of 754 ± 220 ps. The individual partition contribution factor of IC/NR/VR channels is resolved to be about 0.33:1.51:1.12 based on a simple linear approximation treatment for the first time. This work provides a simple and potential approach for regulating photophysical properties of D-A cocrystals by introducing a skeleton electron source.
Two-dimensional fluorescence spectroscopy with quantum entangled photons: Idler-referenced timing without pump detection
Entangled photons have attracted increasing interest as resources for developing time-resolved spectroscopic techniques. Theoretical studies suggest that their non-classical correlations enable time-resolved spectroscopy with monochromatic pumping and can selectively isolate specific Liouville pathways in nonlinear optical signals. In an earlier study, we proposed a fluorescence detection scheme that could, in principle, be implemented using existing single-photon detectors [Y. Fujihashi et al., Sci. Adv. 12, eaed7026 (2026)]. In that design, the time origin was defined by detecting the arrival of the pulsed laser used to pump the nonlinear crystal for spontaneous parametric down-conversion, a requirement that made the overall experiment cumbersome. This study theoretically examines an alternative protocol that defines the reference time based on the arrival of idler photons. We demonstrate that this idler-referenced scheme functions effectively when the entangled photons exhibit either negative or negligible frequency correlations. Eliminating the pump-timing channel simplifies the optical layout and lowers the experimental barrier to realizing time-resolved two-dimensional fluorescence spectroscopy with entangled photons. Although the photons may exhibit frequency correlations in isolation, their frequency-time degrees of freedom can behave as effectively uncorrelated when considered over the full measurement timescale. Therefore, fully exploiting non-classical correlations requires an entangled photon source whose temporal characteristics are carefully matched to the overall timescale of the experiment.