Issue 46, 2022

Combination of explainable machine learning and conceptual density functional theory: applications for the study of key solvation mechanisms

Abstract

We present explainable machine learning approaches for the accurate prediction and understanding of solvation free energies, enthalpies, and entropies for different salts in various protic and aprotic solvents. As key input features, we use fundamental contributions from the conceptual density functional theory (DFT) of solutions. The most accurate models with the highest prediction accuracy for the experimental validation data set are decision tree-based approaches such as extreme gradient boosting and extra trees, which highlight the non-linear influence of feature values on target predictions. The detailed assessment of the importance of features in terms of Gini importance criteria as well as Shapley Additive Explanations (SHAP) and permutation and reduction approaches underlines the prominent role of anion and cation solvation effects in combination with fundamental electronic properties of the solvents. These results are reasonably consistent with previous assumptions and provide a solid rationale for more recent theoretical approaches.

Graphical abstract: Combination of explainable machine learning and conceptual density functional theory: applications for the study of key solvation mechanisms

Supplementary files

Article information

Article type
Paper
Submitted
22 Sept. 2022
Accepted
10 Nov. 2022
First published
10 Nov. 2022

Phys. Chem. Chem. Phys., 2022,24, 28314-28324

Combination of explainable machine learning and conceptual density functional theory: applications for the study of key solvation mechanisms

I. Ho, M. Matysik, L. M. Herrera, J. Yang, R. J. Guderlei, M. Laussegger, B. Schrantz, R. Hammer, R. A. Miranda-Quintana and J. Smiatek, Phys. Chem. Chem. Phys., 2022, 24, 28314 DOI: 10.1039/D2CP04428E

To request permission to reproduce material from this article, please go to the Copyright Clearance Center request page.

If you are an author contributing to an RSC publication, you do not need to request permission provided correct acknowledgement is given.

If you are the author of this article, you do not need to request permission to reproduce figures and diagrams provided correct acknowledgement is given. If you want to reproduce the whole article in a third-party publication (excluding your thesis/dissertation for which permission is not required) please go to the Copyright Clearance Center request page.

Read more about how to correctly acknowledge RSC content.

Social activity

Spotlight

Advertisements