Journal of Computational Chemistry最新文献

The number of carbon allotropes is ever increasing with a variety of novel network topologies. Among several that were recently proposed, the K₄ structure draws particular attention due to its similarities with diamond, but it is a chiral periodic graph. However, the proposed K₄ was shown to be dynamically unstable. As a result, despite the fact that such a graph is uncommon in the allotropes of elements, the graph of K₄ is disregarded as a potential structure for elemental carbon. Herein, by means of first principles calculations, we have attempted stabilizing such a chiral graph for carbon by modulating phonon eigenvectors associated with imaginary phonon frequencies that were responsible for the structure instability. This has led us to generate six novel stable carbon allotropes, with 4- and a mix of 3,4-connected carbon atoms, that have the history of chiral K₄. The newly predicted carbon allotropes retain chirality, exhibit lower symmetry than K₄ carbon, and show improved thermodynamic stability. Among these, the thermodynamic stability increases as the ratio of 3:4 connectedness decreases. Analysis of the electronic structure suggests that the 3-connected K₄ net would prefer to be colored with elements of three valence electrons. The computed band gap suggests that all the six novel K₄ modifications predicted are semiconductors with band gaps ranging from 0.72 to 5.14 eV. In addition, the calculated elastic stiffness constants indicates that the K₄ carbon modifications are mechanically stable.

Identifying the most stable, or global-minimum (GM), structure of coordination complexes is critical across numerous scientific fields. In this work, we developed a new program called the SGMCC package, designed to search for the most stable structures of coordination complexes containing single-core or multi-cores metal centers. We introduced several improvements to the traditional genetic algorithm (GA), including the integration of isomer-specific chemical constraints, alongside the random perturbation (RP) algorithm followed by the geometry relaxation (GR) algorithm. The SGMCC package facilitates the study of micro-scale coordination structures formed by metal ions bound to ligands, with applications in areas such as seawater uranium extraction, rare earth element extraction and separation, and the separation of lanthanides/actinides and other nuclides. Additionally, the SGMCC program can accelerate the discovery of new ligands and materials for metal extraction and separation, investigate nucleation mechanisms, and analyze the dynamic structural evolution of metal complexes during nucleation. This paper presents an overview of the SGMCC package's applications.

The rotational dynamics of protonated cyanogen (NCCNH⁺) is studied for collision with both para (p-) and ortho (o-) hydrogen (H₂) in the temperature range $��1�-�100�$ K. Such cold collisions with H₂ are essential to get state-to-state rate coefficients for rotational transitions of newly detected NCCNH⁺. First, the ab initio 4D potential energy surface (PES) for NCCNH⁺-H₂ collision is generated at CCSD(T)-F12b level of theory using augmented triple zeta basis, considering rigid rotor approximation. The 4D PES is further fitted into a neural network (NN) model to augment the PES by $��\sim �130�$ folds. The PES is then expanded using bispherical harmonics functions to get radial terms, which are expressed as analytic functions. The cross-sections and rate coefficients for NCCNH⁺ are calculated using the exact close-coupling method for rotational states up to $��j�=�16�$, considering both p- and o-H₂ collisions. The rates show similar behavior to the earlier studied NCCN-H₂ collision, with the latter having larger rate coefficients.

We have combined the linear-scaling divide-and-conquer (DC) method with the semi-empirical extended tight-binding (xTB) method to develop a general-purpose large-scale quantum molecular dynamics simulation platform called DCxTBMD. Although the DC method is a fragmentation-based acceleration scheme, it is possible to perform highly accurate calculations even when adopting formal grid-based fragmentation, thanks to the mechanism called the buffer region. The program was developed by incorporating xTB Hamiltonian construction modules in the xTB program into the DCDFTBK package, a DC-based density functional tight-binding (DFTB) calculation suite. The accuracy and computational time, including the parallel scalability, of the present method were investigated on the Japanese supercomputer Fugaku, and it was found that the present method is suitable for the efficient calculation of large systems, including multiple heteroelements. We further implemented the energy gradient of the present method to enable the molecular dynamics simulation of such systems.

Tubular g-C₃N₄ nanotubes (CNNTs), experimentally realizable counterparts of 2D g-C₃N₄, present promising opportunities for tuning electronic, optical, and mechanical properties, which remain largely underexplored. In this work, density functional theory calculations are utilized to evaluate the band gaps and elastic moduli of CNNTs with varying diameters and chiralities under applied strain. The results reveal that the Young's moduli of these CNNTs span from 40 to 230 GPa, while the band gaps show an inverse correlation with the aspect ratio. This systematic study advances the fundamental understanding of CNNTs and provides valuable insights for the rational design of nanostructured materials aimed at applications including photocatalysis and electrocatalysis.