Nonadiabatic Derivative Couplings Calculated Using Information of Potential Energy Surfaces without Wavefunctions: Ab Initio and Machine Learning Implementations
In this work, we implemented an approximate algorithm for calculating nonadiabatic coupling matrix elements (NACMEs) of a polyatomic system with ab initio methods and machine learning (ML) models. Utilizing this algorithm, one can calculate NACMEs using only the information of potential energy surfa...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-05-01
|
| Series: | Molecules |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1420-3049/28/10/4222 |
| _version_ | 1827740454521667584 |
|---|---|
| author | Wen-Kai Chen Sheng-Rui Wang Xiang-Yang Liu Wei-Hai Fang Ganglong Cui |
| author_facet | Wen-Kai Chen Sheng-Rui Wang Xiang-Yang Liu Wei-Hai Fang Ganglong Cui |
| author_sort | Wen-Kai Chen |
| collection | DOAJ |
| description | In this work, we implemented an approximate algorithm for calculating nonadiabatic coupling matrix elements (NACMEs) of a polyatomic system with ab initio methods and machine learning (ML) models. Utilizing this algorithm, one can calculate NACMEs using only the information of potential energy surfaces (PESs), i.e., energies, and gradients as well as Hessian matrix elements. We used a realistic system, namely CH<sub>2</sub>NH, to compare NACMEs calculated by this approximate PES-based algorithm and the accurate wavefunction-based algorithm. Our results show that this approximate PES-based algorithm can give very accurate results comparable to the wavefunction-based algorithm except at energetically degenerate points, i.e., conical intersections. We also tested a machine learning (ML)-trained model with this approximate PES-based algorithm, which also supplied similarly accurate NACMEs but more efficiently. The advantage of this PES-based algorithm is its significant potential to combine with electronic structure methods that do not implement wavefunction-based algorithms, low-scaling energy-based fragment methods, etc., and in particular efficient ML models, to compute NACMEs. The present work could encourage further research on nonadiabatic processes of large systems simulated by ab initio nonadiabatic dynamics simulation methods in which NACMEs are always required. |
| first_indexed | 2024-03-11T03:26:15Z |
| format | Article |
| id | doaj.art-39bebb7da1df494b83ad43a65e5b67cf |
| institution | Directory Open Access Journal |
| issn | 1420-3049 |
| language | English |
| last_indexed | 2024-03-11T03:26:15Z |
| publishDate | 2023-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Molecules |
| spelling | doaj.art-39bebb7da1df494b83ad43a65e5b67cf2023-11-18T02:41:07ZengMDPI AGMolecules1420-30492023-05-012810422210.3390/molecules28104222Nonadiabatic Derivative Couplings Calculated Using Information of Potential Energy Surfaces without Wavefunctions: Ab Initio and Machine Learning ImplementationsWen-Kai Chen0Sheng-Rui Wang1Xiang-Yang Liu2Wei-Hai Fang3Ganglong Cui4Hebei Key Laboratory of Inorganic Nano-Materials, College of Chemistry and Materials Science, Hebei Normal University, Shijiazhuang 050024, ChinaKey Laboratory of Theoretical and Computational Photochemistry, Ministry of Education, College of Chemistry, Beijing Normal University, Beijing 100875, ChinaCollege of Chemistry and Material Science, Sichuan Normal University, Chengdu 610068, ChinaKey Laboratory of Theoretical and Computational Photochemistry, Ministry of Education, College of Chemistry, Beijing Normal University, Beijing 100875, ChinaKey Laboratory of Theoretical and Computational Photochemistry, Ministry of Education, College of Chemistry, Beijing Normal University, Beijing 100875, ChinaIn this work, we implemented an approximate algorithm for calculating nonadiabatic coupling matrix elements (NACMEs) of a polyatomic system with ab initio methods and machine learning (ML) models. Utilizing this algorithm, one can calculate NACMEs using only the information of potential energy surfaces (PESs), i.e., energies, and gradients as well as Hessian matrix elements. We used a realistic system, namely CH<sub>2</sub>NH, to compare NACMEs calculated by this approximate PES-based algorithm and the accurate wavefunction-based algorithm. Our results show that this approximate PES-based algorithm can give very accurate results comparable to the wavefunction-based algorithm except at energetically degenerate points, i.e., conical intersections. We also tested a machine learning (ML)-trained model with this approximate PES-based algorithm, which also supplied similarly accurate NACMEs but more efficiently. The advantage of this PES-based algorithm is its significant potential to combine with electronic structure methods that do not implement wavefunction-based algorithms, low-scaling energy-based fragment methods, etc., and in particular efficient ML models, to compute NACMEs. The present work could encourage further research on nonadiabatic processes of large systems simulated by ab initio nonadiabatic dynamics simulation methods in which NACMEs are always required.https://www.mdpi.com/1420-3049/28/10/4222nonadiabatic couplingsmachine learningexcited states |
| spellingShingle | Wen-Kai Chen Sheng-Rui Wang Xiang-Yang Liu Wei-Hai Fang Ganglong Cui Nonadiabatic Derivative Couplings Calculated Using Information of Potential Energy Surfaces without Wavefunctions: Ab Initio and Machine Learning Implementations Molecules nonadiabatic couplings machine learning excited states |
| title | Nonadiabatic Derivative Couplings Calculated Using Information of Potential Energy Surfaces without Wavefunctions: Ab Initio and Machine Learning Implementations |
| title_full | Nonadiabatic Derivative Couplings Calculated Using Information of Potential Energy Surfaces without Wavefunctions: Ab Initio and Machine Learning Implementations |
| title_fullStr | Nonadiabatic Derivative Couplings Calculated Using Information of Potential Energy Surfaces without Wavefunctions: Ab Initio and Machine Learning Implementations |
| title_full_unstemmed | Nonadiabatic Derivative Couplings Calculated Using Information of Potential Energy Surfaces without Wavefunctions: Ab Initio and Machine Learning Implementations |
| title_short | Nonadiabatic Derivative Couplings Calculated Using Information of Potential Energy Surfaces without Wavefunctions: Ab Initio and Machine Learning Implementations |
| title_sort | nonadiabatic derivative couplings calculated using information of potential energy surfaces without wavefunctions ab initio and machine learning implementations |
| topic | nonadiabatic couplings machine learning excited states |
| url | https://www.mdpi.com/1420-3049/28/10/4222 |
| work_keys_str_mv | AT wenkaichen nonadiabaticderivativecouplingscalculatedusinginformationofpotentialenergysurfaceswithoutwavefunctionsabinitioandmachinelearningimplementations AT shengruiwang nonadiabaticderivativecouplingscalculatedusinginformationofpotentialenergysurfaceswithoutwavefunctionsabinitioandmachinelearningimplementations AT xiangyangliu nonadiabaticderivativecouplingscalculatedusinginformationofpotentialenergysurfaceswithoutwavefunctionsabinitioandmachinelearningimplementations AT weihaifang nonadiabaticderivativecouplingscalculatedusinginformationofpotentialenergysurfaceswithoutwavefunctionsabinitioandmachinelearningimplementations AT ganglongcui nonadiabaticderivativecouplingscalculatedusinginformationofpotentialenergysurfaceswithoutwavefunctionsabinitioandmachinelearningimplementations |