1
|
Kalamatianos KG, Flenga ON. Prediction of Activation Energies of Organic Molecules With at Most Seven Non-Hydrogen Atoms Using Quantum-Chemically Assisted ML. J Comput Chem 2025; 46:e70083. [PMID: 40110645 DOI: 10.1002/jcc.70083] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/23/2024] [Revised: 02/23/2025] [Accepted: 03/01/2025] [Indexed: 03/22/2025]
Abstract
In this study, a hybrid machine learning (ML) approach is presented for accurately predicting activation energies (Ea) of gas-phase elementary reactions involving organic compounds with up to seven non-hydrogen atoms. Given the importance of activation energies in reaction studies and modeling, ML composite models were created that effectively integrate molecular descriptors with semi-empirical and single energy density functional theory (DFT) calculations. The dataset, containing 300 randomly selected elementary gas-phase reactions, was assembled using accurate DFT (ωB97X-D3/def2-TZVP) values for activation energies Ea from a database alongside semi-empirical computations. For accurate predictions, this approach required the inclusion of both physical organic and geometric/empirical descriptors in the training procedure. The best two ML models demonstrated efficient Ea prediction capability, achieving a mean absolute error (MAE) of 1.314 kcal mol-1 and R2 of 0.992 (Model 3) and (MAE) of 1.949 kcal mol-1 and R2 of 0.979 (Model 2) in validation tests. Notably, this performance approaches the threshold of "chemical accuracy" of 1 kcal mol-1. Model's 3 robustness was tested across the reaction types present in the dataset, demonstrating its ability in properly predicting activation energies, which is critical for the study and optimization of chemical processes.
Collapse
Affiliation(s)
- K G Kalamatianos
- Analytical Chemistry and Cheminformatics Laboratory, Athens, Greece
| | - Olga N Flenga
- Analytical Chemistry and Cheminformatics Laboratory, Athens, Greece
| |
Collapse
|
2
|
Vadaddi SM, Zhao Q, Savoie BM. Graph to Activation Energy Models Easily Reach Irreducible Errors but Show Limited Transferability. J Phys Chem A 2024; 128:2543-2555. [PMID: 38517281 DOI: 10.1021/acs.jpca.3c07240] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 03/23/2024]
Abstract
Activation energy characterization of competing reactions is a costly but crucial step for understanding the kinetic relevance of distinct reaction pathways, product yields, and myriad other properties of reacting systems. The standard methodology for activation energy characterization has historically been a transition state search using the highest level of theory that can be afforded. However, recently, several groups have popularized the idea of predicting activation energies directly based on nothing more than the reactant and product graphs, a sufficiently complex neural network, and a broad enough data set. Here, we have revisited this task using the recently developed Reaction Graph Depth 1 (RGD1) transition state data set and several newly developed graph attention architectures. All of these new architectures achieve similar state-of-the-art results of ∼4 kcal/mol mean absolute error on withheld testing sets of reactions but poor performance on external testing sets composed of reactions with differing mechanisms, reaction molecularity, or reactant size distribution. Limited transferability is also shown to be shared by other contemporary graph to activation energy architectures through a series of case studies. We conclude that an array of standard graph architectures can already achieve results comparable to the irreducible error of available reaction data sets but that out-of-distribution performance remains poor.
Collapse
Affiliation(s)
- Sai Mahit Vadaddi
- Davidson School of Chemical Engineering, Purdue University, West Lafayette, Indiana 47906, United States
| | - Qiyuan Zhao
- Department of Medicinal Chemistry, University of Michigan, Ann Arbor, Michigan 48109, United States
| | - Brett M Savoie
- Davidson School of Chemical Engineering, Purdue University, West Lafayette, Indiana 47906, United States
| |
Collapse
|
3
|
Kirkland JK, Kumawat J, Shaban Tameh M, Tolman T, Lambert AC, Lief GR, Yang Q, Ess DH. Machine Learning Models for Predicting Zirconocene Properties and Barriers. J Chem Inf Model 2024; 64:775-784. [PMID: 38259142 DOI: 10.1021/acs.jcim.3c01575] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 01/24/2024]
Abstract
Zr metallocenes have significant potential to be highly tunable polyethylene catalysts through modification of the aromatic ligand framework. Here we report the development of multiple machine learning models using a large library (>700 systems) of DFT-calculated zirconocene properties and barriers for ethylene polymerization. We show that very accurate machine learning models are possible for HOMO-LUMO gaps of precatalysts but the performance significantly depends on the machine learning algorithm and type of featurization, such as fingerprints, Coulomb matrices, smooth overlap of atomic positions, or persistence images. Surprisingly, the description of the bonding hapticity, the number of direct connections between Zr and the ligand aromatic carbons, only has a moderate influence on the performance of most models. Despite robust models for HOMO-LUMO gaps, these types of machine learning models based on structure connectivity type features perform poorly in predicting ethylene migratory insertion barrier heights. Therefore, we developed several relatively robust and accurate machine learning models for barrier heights that are based on quantum-chemical descriptors (QCDs). The quantitative accuracy of these models depends on which potential energy surface structure QCDs were harvested from. This revealed a Hammett-type principle to naturally emerge showing that QCDs from the π-coordination complexes provide much better descriptions of the transition states than other potential-energy structures. Feature importance analysis of the QCDs provides several fundamental principles that influence zirconocene catalyst reactivity.
Collapse
Affiliation(s)
- Justin K Kirkland
- Department of Chemistry and Biochemistry, Brigham Young University, Provo, Utah 84604, United States
| | - Jugal Kumawat
- Department of Chemistry and Biochemistry, Brigham Young University, Provo, Utah 84604, United States
| | - Maliheh Shaban Tameh
- Department of Chemistry and Biochemistry, Brigham Young University, Provo, Utah 84604, United States
| | - Tyson Tolman
- Department of Chemistry and Biochemistry, Brigham Young University, Provo, Utah 84604, United States
| | - Allison C Lambert
- Department of Chemistry and Biochemistry, Brigham Young University, Provo, Utah 84604, United States
| | - Graham R Lief
- Research and Technology, Chevron Phillips Chemical Company, Highways 60 & 123, Bartlesville, Oklahoma 74003, United States
| | - Qing Yang
- Research and Technology, Chevron Phillips Chemical Company, Highways 60 & 123, Bartlesville, Oklahoma 74003, United States
| | - Daniel H Ess
- Department of Chemistry and Biochemistry, Brigham Young University, Provo, Utah 84604, United States
| |
Collapse
|
4
|
Döntgen M, Wildenberg A, Heufer KA. Theoretical Investigation of Key Properties of the Pyrolysis of Methyl, Ethyl, and Dimethyl Dioxolane Isomers. J Phys Chem A 2022; 126:8326-8336. [DOI: 10.1021/acs.jpca.2c06564] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/07/2022]
Affiliation(s)
- Malte Döntgen
- Chair of High Pressure Gas Dynamics, Shock Wave Laboratory, RWTH Aachen University, 52056Aachen, Germany
| | - Alina Wildenberg
- Chair of High Pressure Gas Dynamics, Shock Wave Laboratory, RWTH Aachen University, 52056Aachen, Germany
| | - K. Alexander Heufer
- Chair of High Pressure Gas Dynamics, Shock Wave Laboratory, RWTH Aachen University, 52056Aachen, Germany
| |
Collapse
|
5
|
Spiekermann K, Pattanaik L, Green WH. High accuracy barrier heights, enthalpies, and rate coefficients for chemical reactions. Sci Data 2022; 9:417. [PMID: 35851390 PMCID: PMC9293986 DOI: 10.1038/s41597-022-01529-6] [Citation(s) in RCA: 14] [Impact Index Per Article: 4.7] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/18/2022] [Accepted: 06/30/2022] [Indexed: 12/13/2022] Open
Abstract
Quantitative chemical reaction data, including activation energies and reaction rates, are crucial for developing detailed kinetic mechanisms and accurately predicting reaction outcomes. However, such data are often difficult to find, and high-quality datasets are especially rare. Here, we use CCSD(T)-F12a/cc-pVDZ-F12//ωB97X-D3/def2-TZVP to obtain high-quality single point calculations for nearly 22,000 unique stable species and transition states. We report the results from these quantum chemistry calculations and extract the barrier heights and reaction enthalpies to create a kinetics dataset of nearly 12,000 gas-phase reactions. These reactions involve H, C, N, and O, contain up to seven heavy atoms, and have cleaned atom-mapped SMILES. Our higher-accuracy coupled-cluster barrier heights differ significantly (RMSE of ∼5 kcal mol-1) relative to those calculated at ωB97X-D3/def2-TZVP. We also report accurate transition state theory rate coefficients [Formula: see text] between 300 K and 2000 K and the corresponding Arrhenius parameters for a subset of rigid reactions. We believe this data will accelerate development of automated and reliable methods for quantitative reaction prediction.
Collapse
Affiliation(s)
- Kevin Spiekermann
- Department of Chemical Engineering, Massachusetts Institute of Technology, 77 Massachusetts Ave, Cambridge, MA, 02139, USA
| | - Lagnajit Pattanaik
- Department of Chemical Engineering, Massachusetts Institute of Technology, 77 Massachusetts Ave, Cambridge, MA, 02139, USA
| | - William H Green
- Department of Chemical Engineering, Massachusetts Institute of Technology, 77 Massachusetts Ave, Cambridge, MA, 02139, USA.
| |
Collapse
|
6
|
Lewis‐Atwell T, Townsend PA, Grayson MN. Machine learning activation energies of chemical reactions. WIRES COMPUTATIONAL MOLECULAR SCIENCE 2022. [DOI: 10.1002/wcms.1593] [Citation(s) in RCA: 8] [Impact Index Per Article: 2.7] [Reference Citation Analysis] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 12/11/2022]
Affiliation(s)
- Toby Lewis‐Atwell
- Department of Computer Science, Faculty of Science University of Bath Bath UK
| | - Piers A. Townsend
- Department of Chemistry, Faculty of Science University of Bath Bath UK
| | | |
Collapse
|
7
|
Panaget T, Mokrani N, Batut S, Lahccen A, Fenard Y, Pillier L, Vanhove G. Insight into the Ozone-Assisted Low-Temperature Combustion of Dimethyl Ether by Means of Stabilized Cool Flames. J Phys Chem A 2021; 125:9167-9179. [PMID: 34636244 DOI: 10.1021/acs.jpca.1c05583] [Citation(s) in RCA: 3] [Impact Index Per Article: 0.8] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/29/2022]
Abstract
The low-temperature combustion kinetics of dimethyl ether (DME) were studied by means of stabilized cool flames in a heated stagnation plate burner configuration using ozone-seeded premixed flows of DME/O2. Direct imaging of CH2O* chemiluminescence and laser-induced fluorescence of CH2O were used to determine the flame front positions in a wide range of lean and ultra-lean equivalence ratios and ozone concentrations for two strain rates. The temperature and species mole fraction profiles along the flame were measured by coupling thermocouples, gas chromatography, micro-chromatography, and quadrupole mass spectrometry analysis. A new kinetic model was built on the basis of the Aramco 1.3 model, coupled with a validated submechanism of O3 chemistry, and was updated to improve the agreement with the obtained experimental results and experimental data available in the literature. The main results show the efficiency of the tested model to predict the flame front position and temperature in every tested condition, as well as the importance of reactions typical of atmospheric chemistry in the prediction of cool flame occurrence. The agreement on the fuel and major products is overall good, except for methanol, highlighting some missing kinetic pathways for the DME/O2/O3 system, possibly linked to the direct addition of atomic oxygen on the fuel radical, modifying the product distribution after the cool flame.
Collapse
Affiliation(s)
- Thomas Panaget
- University of Lille, CNRS, UMR 8522-PC2A-Physicochimie des Processus de Combustion et de l'Atmosphère, F-59000 Lille, France
| | - Nabil Mokrani
- University of Lille, CNRS, UMR 8522-PC2A-Physicochimie des Processus de Combustion et de l'Atmosphère, F-59000 Lille, France
| | - Sébastien Batut
- University of Lille, CNRS, UMR 8522-PC2A-Physicochimie des Processus de Combustion et de l'Atmosphère, F-59000 Lille, France
| | - Amaury Lahccen
- University of Lille, CNRS, UMR 8522-PC2A-Physicochimie des Processus de Combustion et de l'Atmosphère, F-59000 Lille, France
| | - Yann Fenard
- University of Lille, CNRS, UMR 8522-PC2A-Physicochimie des Processus de Combustion et de l'Atmosphère, F-59000 Lille, France
| | - Laure Pillier
- University of Lille, CNRS, UMR 8522-PC2A-Physicochimie des Processus de Combustion et de l'Atmosphère, F-59000 Lille, France
| | - Guillaume Vanhove
- University of Lille, CNRS, UMR 8522-PC2A-Physicochimie des Processus de Combustion et de l'Atmosphère, F-59000 Lille, France
| |
Collapse
|