Significant progress has been achieved in recent years with the development of high-dimensional permutationally invariant analytic Born-Oppenheimer potential-energy surfaces, making use of polynomial invariant theory. In this work, we have developed a generalization of this approach which is suitable for the construction of multi-sheeted multi-dimensional potential-energy surfaces exhibiting seams of conical intersections. The method avoids the nonlinear optimization problem which is encountered in the construction of multi-sheeted diabatic potential-energy surfaces from ab initio electronic-structure data. The key of the method is the expansion of the coefficients of the characteristic polynomial in polynomials which are invariant with respect to the point group of the molecule or the permutation group of like atoms. The multi-sheeted adiabatic potential-energy surface is obtained from the Frobenius companion matrix which contains the fitted coefficients. A three-sheeted nine-dimensional adiabatic potential-energy surface of the (2)T2 electronic ground state of the methane cation has been constructed as an example of the application of this method.
Download full-text PDF |
Source |
---|---|
http://dx.doi.org/10.1063/1.4808358 | DOI Listing |
Chemphyschem
January 2025
Fachbereich Chemie, Philipps-Universität Marburg, 35032, Marburg, Germany.
Both, molecular chemical reactions and transport of atoms in solid media are determined by the energy landscape in which the seemingly different processes take place. Chemical reactions can be described as cooperative translocation of two chemical entities on a common potential energy surface. Transport of atoms in a solid can be envisaged as the translocation of a single particle in the potential energy landscape of all other particles constituting the solid.
View Article and Find Full Text PDFJ Phys Chem A
January 2025
Université Paris-Saclay, CNRS, Institut des Sciences Moléculaires d'Orsay UMR 8214, 91405 Orsay, France.
This study deals with the understanding of hydrogen atom scattering from graphene, a process critical for exploring C-H bond formation and energy transfer during atom surface collision. In our previous work [Shi, L.; 2023, 159, 194102], starting from a cell with 24 carbon atoms treated periodically, we have achieved quantum dynamics (QD) simulations with a reduced-dimensional model (15D) and a simulation in full dimensionality (75D).
View Article and Find Full Text PDFPhys Chem Chem Phys
January 2025
Wolfson Department of Chemical Engineering, Technion - Israel Institute of Technology, Haifa 3200003, Israel.
The pressure-dependent reactions on the NH potential energy surface (PES) have been investigated using CCSD(T)-F12/aug-cc-pVTZ-F12//B2PLYP-D3/aug-cc-pVTZ. This study expands the NH PES beyond the previous literature by incorporating a newly identified isomer, NHN, along with additional bimolecular reaction channels associated with this isomer, namely NNH + H and HNN(S) + H. Rate coefficients for all relevant pressure-dependent reactions, including well-skipping pathways, are predicted using a combination of transition state theory and master equation simulations.
View Article and Find Full Text PDFAdv Sci (Weinh)
January 2025
Department of Chemistry, Yonsei University, 50 Yonsei-ro, Seodaemun-gu, Seoul, 03722, Republic of Korea.
Machine learning interatomic potentials (MLIPs) promise quantum-level accuracy at classical force field speeds, but their performance hinges on the quality and diversity of training data. An efficient and fully automated approach to sample chemical reaction space without relying on human intuition, addressing a critical gap in MLIP development is presented. The method combines the speed of tight-binding calculations with selective high-level refinement, generating diverse datasets that capture both equilibrium and reactive regions of potential energy surfaces.
View Article and Find Full Text PDFJ Chem Inf Model
January 2025
College of Information Science and Engineering, Huaqiao University, Xiamen 361021, China.
Machine learning methods for fitting potential energy surfaces and molecular dynamics simulations are becoming increasingly popular due to their potentially high accuracy and savings in computational resources. However, existing application models often rely on basic architectures like artificial neural networks (ANNs) and multilayer perceptron (MLP), lagging behind cutting-edge technologies in the machine learning domain. Furthermore, the complexity of current machine learning frameworks leads to reduced interpretability and challenges for improvement.
View Article and Find Full Text PDFEnter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!