In this work, the topology of the scalar field at four-component (4c) level, including bond critical points (BCPs) and basin populations, is studied for SnH4-nIn (n = 0, …, 4) and UF6 molecular systems. The basin population of the 4c scalar fields is calculated for the first time using the near-grid method combined with the Lebedev and Chebyshev quadratures, which improves the efficiency of the methodology applied in our previous work [A. D. Zapata-Escobar, A. F. Maldonado. Definitions of atoms in molecules with the modified Dirac equations. J. Chem. Phys. 164, 024114 (2026)]. The scalar fields used are the Dirac electron density ( ρ D ) and the modified Dirac scalar field ( ϕ MD ). In Sn-containing molecules, it was found that the Sn-BCP distance in the Sn-H bond increases when relativistic effects are included, and there is also an increase in the electron density at the BCPs, indicating that the covalent character of the Sn-H bond is enhanced. In the case of the Sn-I bond, the Sn-BCP distance also increases with relativistic effects, but a decrease in the covalent character is observed as the electron density decreases. For the UF6 molecule, relativistic effects move the BCPs closer to the uranium (central) atom and far away to the fluorine atoms, and this behavior is opposite to that obtained for the tin (central) atom in Sn-containing systems. Furthermore, relativistic effects produce changes in the electronic populations, with respect to the nonrelativistic values, exceeding 10% when heavy atoms are involved. In summary, it is observed that the basin associated with the heaviest atom in the studied systems is contracted by relativistic effects.

The exceptional stability and broad application potential of zirconium-based metal-organic frameworks call for reliable atomistic methods capable of describing their structural, thermal, and mechanical properties. In this work, we benchmark the recently introduced extension of the DFTB3/3OB parameter set for the transition metal zirconium by evaluating a representative set of Zr-containing metal-organic frameworks spanning different secondary building units and topologies, namely UiO-66, UiO-67, MOF-801, MIL-140A, and MIL-163. Equilibrium lattice parameters, X-ray diffraction patterns, thermal expansion coefficients, and volumetric stiffness measures have been computed from DFTB3/3OB molecular dynamics simulations and compared to literature reference data. The results have been further assessed using hybrid DFT (HSEsol) reference calculations and contextualized through an independent comparison with neural-network-based MACE-DAC-1 simulations. Across the benchmark set, DFTB3/3OB provides a consistent structural description and reproduces the dominant experimental diffraction features, while the deviations of the lattice parameters remain small. DFTB3/3OB correctly reproduces the negative thermal expansion behavior of the UiO frameworks and predicts positive thermal expansion for MOF-801; for the investigated MIL-type systems, the observed anisotropic expansion highlights a stronger method dependence. For the cubic frameworks UiO-66 and UiO-67, the bulk moduli obtained via DFTB3/3OB fall within reported reference ranges and generally indicate intermediate stiffness relative to MACE-DAC-1 and HSEsol. For the non-cubic MIL-140A and MIL-163, the corresponding MD-derived fixed-shape isotropic volumetric moduli provide a consistent basis for comparison between DFTB3/3OB and MACE-DAC-1. Overall, the presented results support the applicability and transferability of DFTB3/3OB to molecular dynamics simulations of zirconium-based MOFs beyond the prototypical UiO frameworks.

We present vibe-qc, a general-purpose open-source electronic-structure code for molecules and solids, developed almost entirely with agentic AI under expert human direction. The code is released under the Mozilla Public License 2.0 (MPLv2). The implementation comprises a C++17 numerical core and Python calculation drivers. Molecular methods include Hartree-Fock, Kohn-Sham DFT, Møller-Plesset theory and double hybrids, coupled cluster, active-space methods, and excited states. Dispersion corrections, continuum solvation, and effective core potentials are also available. Periodic calculations use Gaussian-basis HF and DFT with Brillouin-zone sampling. Molecular and periodic method coverage differs; the feature matrices specify analytic and numerical derivatives and distinguish validated and experimental implementations. The validation set is a key contribution of this work. It combines independent program comparisons, published benchmarks, and internal consistency tests. Molecular mean-field energies agree at the nanohartree scale, and conventional CCSD(T) energies agree within 3.3 × 10 - 7 E h for the tested systems. Local-correlation and periodic limitations are reported with the comparisons. Numerical results, calculation metadata, and analysis scripts are supplied as machine-readable Supporting Information. The independently maintained vibe-view viewer, vibe-queue scheduler, and qvf toolkit support visualization and job management; trexio provides wavefunction interchange.

A combined topological and kinetic approach to modeling abundant fullerene isomer formation is proposed. This approach considers a fullerene isomer formation event as a time trajectory of topological changes in the fullerene structure, represented as a sequence of bond rearrangement reactions (or transitions between fullerene isomers). Each reaction in the sequence is chosen on the basis of its calculated reaction rate from the complete list of all possible reactions for the current isomer and the reaction types corresponding to the atomistic formation mechanism under consideration. A detailed description of the proposed approach is provided, including the program structure and workflow. The result of the performed calculation is the yield of an abundant isomer for the considered atomistic mechanism as a function of synthesis conditions (for example, temperature and carbon vapor pressure). Algorithms for generating a complete list of reactions for a given isomer are developed based on topological description in the dual space for a set of possible reaction types, and methods for calculating the corresponding reaction rates are described. The sensitivity of the modeling results to the accuracy of the description of reaction rates is tested using the example of sp-atom migration in C58-C2v fullerene, which is a precursor for the most abundant C60-Ih fullerene isomer. Possible applications of the proposed approach and auxiliary tasks required for its development are discussed.

The first-order correction for the one-body electronic properties can be a powerful tool for increasing the accuracy of their calculations in the framework of multireference perturbation theory (MRPT) methods. The individual contributions for the evaluation of the one-body properties are derived for the fully uncontracted MRPT. A detailed algorithm for their simple and cost-effective calculation is also presented. A way to construct a sum of zeroth- and first-order contributions in operator form is demonstrated. Results obtained in the framework of uncontracted MRPT CDAS-PT2(IPEA) implemented in the NOPT quantum chemistry package for test cases of several small molecules show the improvement of the precision of the calculation for the permanent and transition dipole moment.

Rhombic Be2X2 clusters (X = O, S, Se, and Te) provide electron-deficient Be centers capable of coordinating noble gas (Ng) atoms. Here, the structures, energetic stability, and bonding of Be2X2 and their homoleptic Ng2Be2X2 complexes (Ng = He, Ne, Ar, Kr, Xe, and Rn) were investigated using correlated wavefunction calculations and complementary bonding analyses. Structural searches identified D2h-symmetric rhombic geometries as the global minima for all Be2X2 and Ng2Be2X2 systems. The bare clusters are strongly bound against dissociation, although their stability decreases from Be2O2 to Be2Te2. Ng bonding strength increases from Ar to Rn but generally weakens as X changes from O to Te, while the sequential dissociation energies reveal modest negative cooperativity between the two Ng⋯Be interactions. The He and Ne complexes exhibit very small dissociation energies. Natural population analysis indicates a gradual increase of Ng-to-Be electron donation toward the heavier noble gases. QTAIM descriptors indicate the presence of some degree of covalent character in the Ng⋯Be contacts. EDA-NOCV analysis confirms that stabilization results from a combination of orbital, electrostatic, and dispersion effects. In the Ng-Be interactions of Ng2Be2X2, orbital interactions dominate over electrostatics and arise mainly from two Ng(ρσ) → Be2X2 donation channels (in-phase and out-of-phase). Overall, these complexes are best described as polarized donor-acceptor assemblies, whose bonding character depends jointly on the Ng and the Be2X2 core.

Polyoxometalates (POMs) are versatile metal-oxo clusters with remarkable structural diversity and applications spanning catalysis, energy storage, and materials science. However, accurately predicting their thermodynamic properties remains computationally demanding due to the large system sizes and the cost of density functional theory (DFT) frequency calculations required to obtain Gibbs free energies. In this work, we present POMs-GibbsNet, a deep neural network designed to predict the thermal free energy correction (G - E) of polyoxometalates by combining graph-based molecular representations with DFT-derived descriptors. Molecular structures were encoded using variants of the AABBA autocorrelation graph kernel, which captures atom-atom, bond-bond, and atom-bond correlations together with quantum-derived features including QTAIM descriptors. POMs-GibbsNet was trained on a dataset of 1419 POM structures spanning 19 chemical systems to obtain good results (R 2 = 0 . 993 and RMSE = 1.9 kcal · mol[-1]. To evaluate the performance, a benchmark against 10 classical regressors was performed. Additionally, an uncertainty study was added to the methodology. External validation using POMSimulator reaction networks comprising over 10,819 reactions shows excellent agreement between predicted and DFT reaction energies (R 2 > 0 . 99), with errors typically within a few kcal mol[-1]. Error cancellation was also studied during this step. Notably, simplified atom-atom autocorrelation descriptors already provide robust performance while reducing model complexity. These results demonstrate that POMs-GibbsNet enables efficient estimation of thermodynamic corrections in POM systems, providing a powerful framework for accelerating the computational exploration of polyoxometalate speciation and self-assembly across large chemical spaces.

Ion-pair binding energy is a fundamental property that governs the physicochemical behavior of ionic liquids (ILs), influencing ion dissociation, transport, interfacial behavior, and field-driven extraction processes across a wide range of applications. Accurate prediction of this quantity is therefore essential for rational IL design. While density functional theory (DFT) delivers precise energetic characterization of ion pairs, its computational demands render it impractical for high-throughput evaluation of ionic liquid (IL) candidates. Here, we present a machine learning (ML) approach for direct prediction of DFT-derived ion-pair binding energies from SMILES strings of component ions. To resolve the mapping from molecular structure to binding energy, we considered seven machine learning surrogates: (1) kernel ridge regression (KRR), (2) Gaussian process regression (GPR), (3) random forest (RF), (4) XGBoost, (5) LightGBM, (6) CatBoost, and (7) graph neural network (GNN), trained on a set of 100 ion pairs and tested on a held-out set of 20 ion pairs. The gradient-boosted trees (XGBoost, LightGBM, and CatBoost) achieved the best predictive performance, yielding an average coefficient of determination ( R 2 ) of about 0.844 and an average Spearman's rank correlation coefficient ( ρ ) of about 0.879. The results from this study highlight the strong potential of descriptor-driven ML approaches to estimate ion-pair binding energies, establishing their utility as computationally efficient alternatives to DFT for preliminary screening of ionic liquids across varied application domains.

The introduction of non-fullerene acceptors (NFAs) in bulk heterojunction organic solar cells has advanced their power conversion efficiencies beyond 20%. NFAs consisting of a fused ring electron-donating core, capped with electron-withdrawing end groups, are being intensely investigated both experimentally and computationally, since they demonstrate superior performance. 1,1-dicyanomethylene-3-indanone (IC) is a widely used end group due to its strong electron-withdrawing ability and ease of chemical modification. The functionality of this end group can be further enhanced by employing two common strategies. Halogenation, that adjusts energy levels, and addition of π-extensions that red-shifts optical absorption. In the present work we improve the General Amber Force Field to model fused ring NFAs with a thiophene unit at the edge of the core. To this end, 10 different NFA structures that incorporate the IC end group along with its halogenated and π-extended variants are employed. They are classified into three popular families according to their structural characteristics. Using ab initio computations, the dihedral angles between the fused ring core and the end group are reparametrized, while partial atomic charges are derived using the standard RESP procedure, thus creating complete force field models. Using experimentally resolved crystal structures available in the literature we validate the new parameterization and demonstrate its suitability for modeling a variety of fused ring NFAs.

The frontier-orbital (HOMO-LUMO) gap is a compact descriptor of the electronic structure of transition-metal complexes (TMCs) and a recurrent screening target for catalysis, photophysics, and functional material design; however, its evaluation by density functional theory (DFT) is costly at the library scale. Working entirely from the public tmQM database, with no additional DFT calculations and no experiments, we examine three questions that are seldom treated together: how much structural detail a model requires, whether it generalizes across the periodic table, and how far its uncertainty estimates can be trusted. For 108,541 mononuclear TMCs, we construct a hierarchy of representations of increasing cost: composition-and-coordination descriptors; the same descriptors augmented with inexpensive cheminformatics (RDKit) features; and a SchNet graph neural network that operates directly on the three-dimensional geometry. The predictive accuracy increases monotonically across the hierarchy (test R[2] ≈ 0.57 → 0.63 → 0.77; mean absolute error 0.44 → 0.30 eV), and Shapley value analysis attributes the predictions to chemically reasonable factors. We then show that this in-domain accuracy is a poor proxy for reliability. Leave-one-series-out and leave-one-metal-out evaluations reveal a sharp loss of accuracy on unseen chemistry (median leave-one-metal-out R[2] ≈ 0.34; worst case d[10] zinc, R[2] < 0), and, contrary to a natural expectation, neither of two geometry networks (SchNet, DimeNet++) extrapolates better than the simpler tabular model across metal series-both are often worse-although the directional architecture alone extrapolates with positive skill to a single withheld metal. Finally, although standard split-conformal intervals are well calibrated in the domain (≈90% coverage), they undercover (81%) for an unseen series; we demonstrate that series-conditional (Mondrian) calibration restores per-series coverage and that a distance-based applicability-domain filter flags essentially all out-of-domain complexes, recovering calibrated coverage on the retained subset. The study offers a transparent, reproducible recipe for deciding when and how confidently such models may be applied.

Controlling and observing electron motion in chiral molecules is a fundamental challenge spanning chemical synthesis, materials science, and spintronics. The electronic origin of chiral asymmetry is revealed by combining real-space analysis of the total electronic charge density with time-dependent density functional theory (DFT) electron-nuclear dynamics simulations. Using simulated ultrafast circularly polarized laser pulses, the correct enantiomeric assignment for (S)-1-chloro-1-fluoroethane is demonstrated. Modulating the carrier envelope phase (CEP) angle altered the so-called extrinsic electronic chirality in both chiral (S)-1-chloro-1-fluoroethane and achiral chloroethane. Helical bound electron motion, quantified as intrinsic electronic chirality, however, was uniquely observed in the structurally chiral molecule and tuned to induce a significant degree of electronic achiral character via CEP variation. Furthermore, eigenvector-space trajectories exhibited distinct 3-D and 1-D morphologies for chiral (S)-1-chloro-1-fluoroethane and achiral chloroethane, respectively. This framework provides the molecular-level design principles needed to transition from static structural descriptors to dynamic electronic properties that can directly impact future practical applications, particularly in asymmetric synthesis and the design of next-generation chiral materials, such as those that exploit the chirality-induced spin selectivity (CISS) effect.

The results of the exhaustive conformational sampling are contrasted with metadynamics-based CREST simulations for a set of 13 gas-phase forms of clinical drug molecules retrieved from the CD25 dataset. [Chem. Sci. 2021, 12 (19), 6551] The GFN1- and GFN2-xTB methods complement each other in the identification of the most stable conformer. The GFN-FF performs worse due to fundamental differences between the force field and the density functional theory (DFT) potential energy surfaces. Conformational entropies (Sconf) from the systematic search as well as from the standard CREST entropy mode calculations with the single-point Hessian (SPH) deactivated are in reasonable agreement with their molecular dynamics counterparts that account for torsional correlation [J. Chem. Theory Comput. 2022, 18, 7166-7178]. The CREST entropy mode evaluations on the DFT-refined geometries using the activated SPH routine yield too large Sconf values for some species due to the overestimation of the Boltzmann entropy averaging ( S ¯ ° - S 1 ° ) contribution.

Single-electron transfer (SET) and energy-transfer (EnT) events play a central role in contemporary mechanistic proposals across photochemistry, electrochemistry, and catalysis. However, in routine computational practice, these steps are often treated less systematically than conventional bond-forming or bond-breaking events because their barrier evaluation usually requires a Marcus-theory workflow built from multiple quantum-chemical calculations, solvent-state manipulations, and careful post-processing of energies. Here, we present EASIQ-Marcus (https://github.com/maserasgroup-repo/easiq_marcus/), a Python package that automates this practical workflow for Gaussian users. The code implements several literature approaches for estimating reorganization energies and barriers, including hard-sphere, simplified hard-sphere, and four-point Marcus strategies in both symmetric and asymmetric forms. By turning the required file preparation and data extraction into explicit command-line steps, EASIQ-Marcus is designed to make the practical evaluation of SET and EnT events transparent, reproducible, and suitable for routine mechanistic studies.

Matrix isolation is an experimental technique which allows spectroscopists to observe highly reactive or weakly bound systems by trapping them in frozen noble gas matrices. In this work, we studied the viability of using ab initio molecular dynamics (AIMD) to simulate several H2S·amine complexes isolated in solid argon crystals, comparing the results with conventional density functional theory (DFT) harmonic and anharmonic calculations to assess the most practical and accurate method to calculate the vibrational spectrum of such systems. While AIMD accurately predicted the intensity ratios of certain peaks, the frequencies themselves were no more accurate than conventional harmonic DFT calculations. Modeling an explicit argon matrix did not improve the results. Based on the computational cost of each method, a gas phase anharmonic or scaled harmonic frequency calculation at an appropriate level of theory was determined to be the most accurate and efficient way to simulate matrix isolation spectra.

Self-consistent-field calculations require an inexpensive initial electronic structure that nevertheless reflects the molecular environment. Standard superposition of atomic potentials (SAP) is efficient, but its component potentials describe isolated atoms. We show that the missing environmental information is already encoded in the first molecular SAP density and can be returned through one damped atom-resolved update of the fitted screening potentials. With the feedback strength frozen at 0.25, the intrinsic-atomic-orbital (IAO) realization reduces median density distance by 25.10%, 25.89%, and 25.96% for RHF, PBE, and PBE0 on an independent 80-structure hold-out. IAO charge correlations remain at least 0.99989 from def2-SVP through augmented triple- and quadruple-zeta bases. The original symmetric-Löwdin realization gives the same response direction in compact bases but exposes the expected diffuse-basis pathology. No target-functional information, converged target density, or auxiliary fragment calculation enters either guess.