Journal of Computational Chemistry最新文献

This work introduces an interactive computational model built in Microsoft Excel using the ScienSolar platform to simulate classical electric fields for all periodic table elements. The approach uniquely integrates quantum-inspired concepts—including Slater's rules for electron shielding and Bohr-model orbital distributions—within an accessible spreadsheet environment that requires no programming expertise. Users can dynamically assemble multi-atomic systems, visualize 3D electric fields in real time, and directly modify physical equations to explore custom scenarios. Designed for both education and research, the tool enables rapid conceptual validation and hands-on exploration of electrostatic phenomena in atomic and molecular systems. By bridging theoretical models with practical computation, this framework offers an intuitive and flexible platform for teaching and prototyping in chemistry, materials science, and physics.

Amorphous materials exhibit certain advantageous properties compared to their crystalline counterparts, yet structural characterization is challenging due to their lack of long-range order. This issue is further compounded with amorphous nanomaterials, where strong probe-sample interactions are required. Although electron scattering measurements can address these challenges, extracting detailed structural insights remains difficult. To address this issue, we introduce pyHRMC, a Python package for Hybrid Reverse Monte Carlo (HRMC) analysis of electron total scattering. pyHRMC iteratively refines atomic positions to match experimental electron scattering, generating experimentally-derived atomistic structures. In contrast to previous HRMC software, which uses x-ray and neutron scattering, pyHRMC uses total electron scattering to facilitate investigations of amorphous nanomaterials. We demonstrate that pyHRMC outperforms conventional Reverse Monte Carlo when modeling Al₂O₃ by more closely replicating a target structure from the electron total scattering. The pyHRMC package provides researchers with a powerful tool for gaining structural insights into nanoscale amorphous materials.

Large language models (LLMs) demonstrate strong performance in natural language tasks, but their capacity for genuine in-context learning (ICL) in scientific regression remains unclear. We systematically assessed seven LLMs on molecular property prediction using a controlled framework of 56 transformed tasks that isolate shortcut learning and are designed to induce functional out-of-distribution (OOD) behavior. LLMs performed nearly perfectly on raw molecular weight prediction via shortcut cues but deteriorated under nonlinear transformations, whereas machine learning (ML) baselines showed greater robustness, yielding a performance crossover. Meta-analysis revealed that distributional descriptors and structure–activity landscape indices (SALI) predict task favorability, providing a framework for selecting between LLM- and ML-based approaches in chemistry.

A gradient-boosting based atomic-charge scheme, BoostCha, is introduced. The BoostCha model operates in three steps: it first predicts pseudo-charges for individual atoms based on their local environments, represented by three-dimensional descriptors of Kocer–Mason–Erturk type, then refines these values using global molecular information, and finally restores the charge conservation. The BoostCha charges are employed as input features in two independent machine-learning models for predicting solvation free energies in organic solvents: ESE-Boost, a gradient-boosting model, and ESE-ANN, a dense artificial neural network. Both approaches yield strong predictive performance, with average root-mean-square errors of 0.49 and 0.52 kcal/mol, respectively. The methods demonstrate consistent performance across diverse solvent classes and are particularly accurate for alkanes, alcohols, ethers, esters, ketones, and aromatic and haloaromatic solvents.

Outcomes from fragment-based assembly of small organic molecules are highly dependent on the number and type of fragments available for growth. Here, we present a new chemical searching strategy for the DOCK6 de novo design (DOCK_DN) engine which logically expands the palette of available fragments compared to the current method. Termed DOCK_SWAP, the new routine employs principles of isosteric swapping and a customized library infrastructure (iso-libraries) in which topologically related sidechains, linkers, and scaffolds have been pre-aligned and rank ordered relative to a parent fragment (iso-tables). The primary objectives of the current work are: (I) introduce the DOCK_SWAP infrastructure and algorithms, (II) characterize iso-library expansion outcomes and impact of different alignment and ranking protocols, (III) assess search performance, molecular scores, heterogeneity, and growth path coverage and (IV) evaluate lead refinement. Depending on the protocol, the new iso-libraries are roughly an order of magnitude larger (N = 2197 to 2708) than the existing library (N = 382). Large-scale simulations (N = 3420) across 57 protein–ligand systems, using three different iso-libraries each, confirm that the DOCK_SWAP code base and infrastructure is robust. Although the use of larger DOCK_SWAP iso-libraries come at an increased cost in terms of simulation time, the results highlight the many advantages over the existing DOCK_DN method for design of drug-like ligands with improved complementarity to the sites being targeted.

The global-minimum (GM) structure of alloys and compounds determines their physical, chemical, and mechanical properties, making its accurate identification critical for materials design and engineering. However, the exponential growth of possible atomic configurations with system size and the complexity of interatomic interactions pose major challenges to traditional structure-search methods. This work develops a new GMSAC package to address these challenges by integrating an improved genetic algorithm (GA) with the embedded atom method (EAM) for the structural prediction of alloys. GMSAC optimizes the GA workflow with symmetry-aware duplicate identification, adaptive operator probabilities and maximizing the fitness of population. The EAM potential is employed to rapidly calculate interatomic energies, balancing computational efficiency and prediction accuracy. Validation tests on eight binary alloy bulks (Al–Ag, Al–Cu, Au–Cu, Au–Pd, Pd–Al, Pd–Cu, Pt–Cu and Pt–Pd) demonstrate the good performance of GMSAC, which successfully maps convex hulls, identifies the stable structures per system, and locates GM structures with the lowest formation energies (e.g., Al₂₀Cu₁₂ for Al–Cu with −0.108 eV/atom). This work provides a new tool to accelerate the discovery of high-performance alloys and compounds.

Hydrogen bonding was initially identified by comparing the distance between the heavy atoms D and A in a DH•••A hydrogen bond, to the sum of their van der Waals (vdW) radii as locating the H atom was not possible. In recent years, hydrogen bond radii (HBR) have been defined for typical hydrogen bond donors DH and acceptors, A. However, both these approaches make an unreasonable assumption of treating all the atoms as spheres. This study advances the understanding of hydrogen bond distances by estimating the shape of hydrogen atoms in HD molecules and in A•••HD complexes through electron density profiles. Results reveal that hydrogen exhibits an elliptical shape in both isolated HD molecules and A•••HD complexes, with the minor axis aligned along the HD bond. In non-linear hydrogen bonds, HBR exhibits angle dependence due to anisotropy in H atomic structure, validated through quantum chemical calculations and Atoms in Molecules (AIM) theory. This study underscores the necessity of adopting directional and environment-dependent HBR, short and long radii, over one spherical vdW or HBR for hydrogen bond identification. The proposed findings are also validated using the data for OH•••N and NH•••N hydrogen bonds identified in the CCDC database. We have demonstrated the utility of long and short radii using several illustrative examples.