Accurate Thermochemistry for Organic Cations via Error Cancellation using Connectivity-Based Hierarchy

2020

WIREs Comput Mol Sci

Self Cite

In computational thermochemistry, “isodesmic‐type” reactions play a significant role for obtaining accurate thermochemical quantities using low‐cost methods that can be applied to large systems. This review touches on some of the examples. For instance, a series of relative bond dissociation energies (BDEs) have been devised to calculate absolute BDEs with near‐chemical‐accuracy (~5 kJ mol−1) using density functional theory (DFT) methods. To facilitate the applicability of isodesmic‐type reactions, the connectivity‐based hierarchy (CBH) has been developed to automate the systematic generation of isodesmic‐type reaction schemes, and applied to large organic and biomolecular systems. The related netCBH scheme yields accurate reaction energies in complex organic reactions, achieving coupled‐cluster quality results at DFT cost. Isodesmic‐type reactions have been used to obtain heats of formation for medium‐sized fullerenes, with uncertainties of ~20 kJ mol−1 up to C180. In comparison, the literature C60 heat of formation has an uncertainty of 100 kJ mol−1. Importantly, it fills the gap in which heats of formation for those larger fullerenes are not available. These studies showcase how isodesmic‐type reactions propel the accuracy of quantum chemistry computations to a level that rivals or even betters modern experimental determinations, particularly for systems that are difficult to study experimentally. This article is categorized under: Structure and Mechanism > Reaction Mechanisms and Catalysis Electronic Structure Theory > Combined QM/MM Methods Theoretical and Physical Chemistry > Thermochemistry

Section: Performance Of Cbh For Thermochemistrymentioning

confidence: 99%

Applications of isodesmic‐type reactions for computational thermochemistry

Chan

2020

WIREs Comput Mol Sci

Self Cite

“…For example, we have developed the Connectivity-Based Hierarchy (CBH) of error cancellation schemes that provides an automated protocol to generate isodesmic-type reactions, which increasingly preserve the chemical environment on both sides of a reaction. , CBH reactions can be used to eliminate certain systematic errors present in approximate levels of theory via a corrective term derived from the reactants and products of the CBH reaction calculated at low and high levels of theory. The CBH approach has been utilized to study a broad range of thermochemical problems with an accuracy comparable to cWFT methods at the cost of low-fidelity DFT calculations, including heat of formation of charged and neutral organic and biomolecules, − redox potentials, p K a s, and bond dissociation energies . Thus, where the direct computation of accurate energies is not possible, the exploitation of systematic error cancellation provides a viable (and often the only) alternative to achieve high accuracies in thermochemistry .…”

Section: Introductionmentioning

confidence: 99%

A Fragmentation-Based Graph Embedding Framework for QM/ML

2021

J. Phys. Chem. A

Self Cite

We introduce a new fragmentation-based molecular representation framework "FragGraph" for QM/ML methods involving embedding fragment-wise fingerprints onto molecular graphs. Our model is specifically designed for delta machine learning (Δ-ML) with the central goal of correcting the deficiencies of approximate methods such as DFT to achieve high accuracy. Our framework is based on a judicious combination of ideas from fragmentation, error cancellation, and a state-of-the-art deep learning architecture. Broadly, we develop a general graph-network framework for molecular machine learning by incorporating the inherent advantages prebuilt into error cancellation methods such as the generalized Connectivity-Based Hierarchy. More specifically, we develop a QM/ML representation through a fragmentationbased attributed graph representation encoded with fragment-wise molecular fingerprints. The utility of our representation is demonstrated through a graph network fingerprint encoder in which a global fingerprint is generated through message passing of local neighborhoods of fragment-wise fingerprints, effectively augmenting standard fingerprints to also include the inbuilt molecular graph structure. On the 130k-GDB9 dataset, our method predicts an out-of-sample mean absolute error significantly lower than 1 kJ/mol compared to target G4(MP2) calculated energies, rivaling current deep learning methods with reduced computational scaling.

“…CBH-1 fragments are at maximum two heavy atoms in size and can cause a severe mismatch of bond types for aromatic systems or in molecules containing nonlocal effects such as hyperconjugation, protobranching, 65 or charge delocalization. 59 Additionally, increasing the fragment size should cause the extrapolated energy to monotonically converge toward the high level of theory. Thus, cases where CBH-1 outperforms CBH-2 are likely due to fortuitous cancellation of errors at the CBH-1 rung.…”

Section: Resultsmentioning

confidence: 99%

Effective Molecular Descriptors for Chemical Accuracy at DFT Cost: Fragmentation, Error-Cancellation, and Machine Learning

J. Chem. Theory Comput.

2020

Self Cite

Recent advances in theoretical thermochemistry have allowed the study of small organic and bio-organic molecules with high accuracy. However, applications to larger molecules are still impeded by the steep scaling problem of highly accurate quantum mechanical (QM) methods, forcing the use of approximate, more cost-effective methods at a greatly reduced accuracy. One of the most successful strategies to mitigate this error is the use of systematic error-cancellation schemes, in which highly accurate QM calculations can be performed on small portions of the molecule to construct corrections to an approximate method. Herein, we build on ideas from fragmentation and error-cancellation to introduce a new family of molecular descriptors for machine learning modeled after the Connectivity-Based Hierarchy (CBH) of generalized isodesmic reaction schemes. The best performing descriptor ML(CBH-2) is constructed from fragments preserving only the immediate connectivity of all heavy (non-H) atoms of a molecule along with overlapping regions of fragments in accordance with the inclusion−exclusion principle. Our proposed approach offers a simple, chemically intuitive grouping of atoms, tuned with an optimal amount of error-cancellation, and outperforms previous structure-based descriptors using a much smaller input vector length. For a wide variety of density functionals, DFT+ΔML(CBH-2) models, trained on a set of small-to medium-sized organic HCNOSClcontaining molecules, achieved an out-of-sample MAE within 0.5 kcal/mol and 2σ (95%) confidence interval of <1.5 kcal/mol compared to accurate G4 reference values at DFT cost.