Physics-informed graph neural networks for predicting cetane number with systematic data quality analysis

Kim, Yeonjoon; Cho, Jaeyoung; Naser, Nimal; Kumar, Sabari; Jeong, Keunhong; McCormick, Robert L.; John, Peter; Kim, Seonah

doi:10.1016/j.proci.2022.09.059

Cited by 19 publications

(21 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We selected four global features after testing various molecular descriptors; two surface area descriptors were utilized in Vermeire et al, 53 and two hydrogen bond descriptors were adopted in our predictive model for cetane number. 70 Of note, accuracies comparable to Vermeire et al were still achieved (Details in the next section) after the hyperparameter tuning, truncation, and modification of the model explained above. More details about the hyperparameter tuning procedure are available in the Methods section.…”

Section: Resultsmentioning

confidence: 57%

“…The overall architecture of two GNNs (GNN-Solvent and GNN-Solute) is similar to our previous GNNs for predicting bond dissociation enthalpy and cetane number. 59,70 It consists of three blocks representing the atom, bond, and global state of a molecule. Initial atom, bond, and global features are embedded as 128-dimensional vectors and pass through five message-passing layers.…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Designing solvent systems in chemical processes using self-evolving solubility databases and graph neural networks

Kim

Jung

Kumar

et al. 2023

Preprint

View full text Add to dashboard Cite

Designing solvent systems is the key to achieving the facile synthesis and separation of desired products from chemical processes. In this regard, many machine-learning models have been developed to predict the solubilities of given solute-solvent pairs. However, breakthroughs in developing predictive models for solubility are needed, which can be accomplished through a remarkable expansion and integration of experimental and computational solubility databases. To maximize predictive accuracy, these two databases should not be separately trained when developing ML models. In addition, they should not be simply combined without reconciling the discrepancies between different magnitudes of errors and uncertainties. Here, we introduce self-evolving solubility databases and graph neural networks developed through semi-supervised self-training approaches. Solubilities from quantum-mechanical calculations are referred to during semi-supervised learning, but they are not directly added to the database. Such methodologies enable the augmentation of databases while correcting the discrepancy between experiments and computation and improving the predictive accuracy against experimental solubilities. The resulting model was successfully applied to two practical examples relevant to solvent selection in organic reactions and separation processes: (i) linear relationship between reaction rates and solvation free energy for three organic reactions, and (ii) partition coefficients for lignin-derived monomers and drug-like molecules.

show abstract

Section: Resultsmentioning

confidence: 57%

Section: Resultsmentioning

confidence: 99%

Designing solvent systems in chemical processes using self-evolving solubility databases and graph neural networks

Kim

Jung

Kumar

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…The overall architecture of two GNNs (GNN-solvent and GNN-solute) is inspired and modified from our previously implemented GNNs for predicting bond dissociation enthalpy and cetane number (reactivity of fuel compounds). 61,83 It consists of three blocks representing a molecule's atom, bond, and global state. Initial atom, bond, and global features are embedded as 128-dimensional vectors and pass through five message-passing layers.…”

Section: Resultsmentioning

confidence: 99%

Designing solvent systems using self-evolving solubility databases and graph neural networks

Kim,

Jung,

Kumar

et al. 2024

Chem. Sci.

Self Cite

View full text Add to dashboard Cite

show abstract

“…In other studies, PINN was applied to predict the cetane number. 25 Chinifrooshan Esfahani 26 introduces a PINN model to predict the viscosity of the nanofluids. This research applies a positive output constraint on the layer corresponding to the group contribution values.…”

Section: Common Regression Loss Functions Include Mean Absolute Error...mentioning

confidence: 99%

“…Kim et al suggest the PINN approach to predict the minimum film boiling temperature. In other studies, PINN was applied to predict the cetane number . Chinifrooshan Esfahani introduces a PINN model to predict the viscosity of the nanofluids.…”

Section: Introductionmentioning

confidence: 99%

Physics-Informed Neural Networks with Group Contribution Methods

Babaei

Stone

Knotts

et al. 2023

J. Chem. Theory Comput.

View full text Add to dashboard Cite

Thermophysical properties of organic compounds are used in countless scientific, engineering, and industrial settings in developing theories, designing new systems and devices, analyzing costs and risks, and improving existing infrastructure. Often, due to costs, safety, prior interest, or procedural difficulties, experimental values for desired properties are not available and must be predicted. The literature is filled with prediction techniques, but even the best traditional methods have significant errors compared to what is possible considering experimental uncertainty. Recently, machine learning and artificial intelligence techniques have been applied to the property prediction problem, but the examples to date do not extrapolate well outside the data set used for training the model. This work demonstrates a solution to this problem by combining chemistry and physics when training the model and builds upon prior traditional and machine learning methods. Two case studies are presented. The first is for parachor which is used for surface tension prediction. Surface tensions are needed to design distillation columns, adsorption processes, gas−liquid reactors, liquid−liquid extractors, improve oil reservoir recovery, and undertake environmental impact studies or remediation actions. A set of 277 compounds is divided into training, validation, and test sets, and a multilayered physics-informed neural network (PINN) is developed. The results demonstrate that better extrapolation by deep learning models can be developed by adding in physics-based constraints. Second, a set of 1600 compounds is utilized for training, validating, and testing a PINN to improve normal boiling point predictions based on group contribution methods and physics-based constraints. The results show that the PINN performs better than any other method with a normal boiling point mean absolute error of 6.95 °C on training and 11.2 °C on test data. Key observations are that (1) a balanced split by compound type is important to have representative compound families in each of the train, validation, and test sets and (2) constraining group contributions being positive improves predictions on the test set. While this work demonstrates improvements for only surface tension and normal boiling point, the results offer significant hope that PINNs can improve prediction of other relevant thermophysical properties over existing approaches.

show abstract

Physics-informed graph neural networks for predicting cetane number with systematic data quality analysis

Cited by 19 publications

References 21 publications

Designing solvent systems in chemical processes using self-evolving solubility databases and graph neural networks

Designing solvent systems in chemical processes using self-evolving solubility databases and graph neural networks

Designing solvent systems using self-evolving solubility databases and graph neural networks

Physics-Informed Neural Networks with Group Contribution Methods

Contact Info

Product

Resources

About