Prediction of organic homolytic bond dissociation enthalpies at near chemical accuracy with sub-second computational cost

Abstract

Bond dissociation enthalpies (BDEs) of organic molecules play a fundamental role in determining chemical reactivity and selectivity. However, BDE computations at sufficiently high levels of quantum mechanical theory require substantial computing resources. In this paper, we develop a machine learning model capable of accurately predicting BDEs for organic molecules in a fraction of a second. We perform automated density functional theory (DFT) calculations at the M06-2X/def2-TZVP level of theory for 42,577 small organic molecules, resulting in 290,664 BDEs. A graph neural network trained on a subset of these results achieves a mean absolute error of 0.58 kcal mol −1 (vs DFT) for BDEs of unseen molecules. We further demonstrate the model on two applications: first, we rapidly and accurately predict major sites of hydrogen abstraction in the metabolism of drug-like molecules, and second, we determine the dominant molecular fragmentation pathways during soot formation.

Quantum chemical calculations for over 200,000 organic radical species and 40,000 associated closed-shell molecules

“…These frameworks therefore continue to increase in accuracy as more data is collected far beyond traditional machine learning approaches. ML approaches to quickly and accurately predict enthalpy 14 , ground state energy 15 , bond dissociation energy 16 , and even transition-state activation energies 17 have been developed by leveraging increasingly large databases of DFT calculations. The public distribution of large quantum chemistry databases, such as ioChem-BD 18 , is an important part of advancing the field of machine learning research in computational chemistry, as prior publications of equilibrium 19 , off-equilibrium 20 , and transition-state structures 21 have found applicability beyond their original purpose.…”

“…Geometry optimizations and enthalpy calculations were performed at the M06-2X/def2-TZVP level of theory 26 , which was previously found to have a favorable trade-off between experimental accuracy and computational efficiency. For calculating the bond dissociation enthalpies specifically, results from this DFT methodology were benchmarked against experimental bond dissociation energies and calculations at higher levels of theory 16 . The calculation pipeline showed similar performance to CCSD(T), and is able to capture changes in enthalpy relative to experiment with an accuracy of approximately 2 kcal mol −1 .…”

tmQM Dataset—Quantum Geometries and Properties of 86k Transition Metal Complexes

“… 9 − 11 Neural networks 12 − 16 and other ML models have been used successfully in a wide range of applications, with numerous examples in materials science 17 − 21 and drug discovery. 22 − 26 ML and data-driven approaches are also making rapid progress in catalytic, 27 − 41 organic, 42 − 47 inorganic, 48 , 49 and theoretical 50 − 56 chemistry.…”

Influence of the double bond position in combustion chemistry of methyl butene isomers: A shock tube and laser absorption study

“…Methyl butene isomers have different types of bonds and their bond dissociation energies (BDEs) are presented in Table 2. The BDE values in Table 2 are computed using the machine learning tool developed by John et al 34 . It is to be noted that, allylic bonds are systematically weaker than their vinylic counterparts as shown in Table 2, which basically direct the initiation reactions for fuel decomposition.…”