Enhancing Hansen Solubility Predictions: A Combined Approach Using Integrated Datasets and Optimized Machine Learning Techniques

Abstract:

Accurate prediction of Hansen Solubility Parameters (HSPs) is important for understanding chemical compatibility in fields like pharmaceuticals, cosmetics and chemical engineering. This study aims to enhance HSP prediction by employing machine learning techniques and using a large, extended dataset from the Hansen Solubility Parameter in Practice (HSPiP) software. Models like XGBoost, CatBoost, LightGBM, as well as ensemble methods, were used for regression, optimized through hyperparameter tuning, feature selection and evaluated using Root Mean Squared Error (RMSE), Mean Absolute Error (MAE) and R-squared (R²) metrics. The results indicated that using a wide variety of molecular components improves prediction accuracy and enhances the model’s applicability across different compounds. The findings additionally show that advanced machine learning methods can significantly improve HSP prediction accuracy, facilitating more precise solubility estimates and advancing applications in chemical and materials science.

• Periodical: Key Engineering Materials (Volume 1027)
• Pages: 53-60
• DOI: https://doi.org/10.4028/p-v9fyrI
• Online since: October 2025
• Authors: Konstantina Pantelidou*, Stavros Devrelis, Antonios Lalas, Konstantinos Votis, Dimitrios Tzovaras
• Keywords: Hansen Solubility Parameters, Hyperparameter Optimization, Machine Learning, Regression Modeling, Solubility Prediction

References

[1] Hansen, Charles M. "The three dimensional solubility parameter." Danish Technical: Copenhagen 14 (1967).

[2] Lee, Sumin, et al. "Novel solubility prediction models: Molecular fingerprints and physicochemical features vs graph convolutional neural networks." ACS omega 7.14 (2022): 12268-12277.

DOI: 10.1021/acsomega.2c00697

[3] Belmares, M., et al. "Hildebrand and Hansen solubility parameters from molecular dynamics with applications to electronic nose polymer sensors." Journal of computational chemistry 25.15 (2004): 1814-1826.

DOI: 10.1002/jcc.20098

[4] Hildebrand, J. H., and R. L. Scott. "The Solubility of Non-Electrolytes, Reinhold, New York." (1950).

[5] Ahmed, Darya Rasul, and Fahmi F. Muhammadsharif. "A Review of Machine Learning in Organic Solar Cells." (2024).

[6] Wu, Xiaotong, et al. "Machine learning in the identification, prediction and exploration of environmental toxicology: Challenges and perspectives." Journal of Hazardous Materials 438 (2022): 129487.

DOI: 10.1016/j.jhazmat.2022.129487

[7] He, Lei, et al. "Applications of computational chemistry, artificial intelligence, and machine learning in aquatic chemistry research." Chemical Engineering Journal 426 (2021): 131810.

DOI: 10.1016/j.cej.2021.131810

[8] Hastings, Janna, et al. "Learning chemistry: exploring the suitability of machine learning for the task of structure-based chemical ontology classification." Journal of Cheminformatics 13 (2021): 1-20.

DOI: 10.1186/s13321-021-00500-8

[9] Terrell, Evan. "Estimation of Hansen solubility parameters with regularized regression for biomass conversion products: An application of adaptable group contribution." Chemical Engineering Science 248 (2022): 117184.

DOI: 10.1016/j.ces.2021.117184

[10] AlQasas, Neveen, and Daniel Johnson. "The use of neural network modeling for the estimation of the Hansen solubility parameters of polymer films from contact angle measurements." Surfaces and Interfaces 44 (2024): 103721.

DOI: 10.1016/j.surfin.2023.103721

[11] Chi, Mingzhe, et al. "Atomistic descriptors for machine learning models of solubility parameters for small molecules and polymers." Polymers 14.1 (2021): 26.

DOI: 10.3390/polym14010026

[12] Lee, Min-Hsuan. "Interpretable machine-learning for predicting power conversion efficiency of non-halogenated green solvent-processed organic solar cells based on Hansen solubility parameters and molecular weights of polymers." Solar Energy 261 (2023): 7-13.

DOI: 10.1016/j.solener.2023.05.050

[13] Li, Chunrong, et al. "Machine learning approach to predict Hansen solubility parameters of cocrystal coformers via integrating group contribution and COSMO-RS." Journal of Molecular Liquids (2024): 125319.

DOI: 10.1016/j.molliq.2024.125319

[14] Wigh, Daniel S., Jonathan M. Goodman, and Alexei A. Lapkin. "A review of molecular representation in the age of machine learning." Wiley Interdisciplinary Reviews: Computational Molecular Science 12.5 (2022): e1603.

DOI: 10.1002/wcms.1603

[15] Pang, Jiayun, Alexander WR Pine, and Abdulai Sulemana. "Using natural language processing (NLP)-inspired molecular embedding approach to predict Hansen solubility parameters." Digital Discovery 3.1 (2024): 145-154.

DOI: 10.1039/d3dd00119a

[16] Choi, Phillip, Tom A. Kavassalis, and Alfred Rudin. "Estimation of the three-dimensional solubility parameters of alkyl phenol ethoxylates using molecular dynamics." Journal of colloid and interface science 150.2 (1992): 386-393

DOI: 10.1016/0021-9797(92)90208-4

[17] Panayiotou, Costas. "Solubility parameter revisited: an equation-of-state approach for its estimation." Fluid Phase Equilibria 131.1-2 (1997): 21-35

DOI: 10.1016/s0378-3812(96)03221-9

[18] Perea, J. Darío, et al. "Combined computational approach based on density functional theory and artificial neural networks for predicting the solubility parameters of fullerenes." The Journal of Physical Chemistry B 120.19 (2016): 4431-4438.

DOI: 10.1021/acs.jpcb.6b00787

[19] Sanchez‐Lengeling, Benjamin, et al. "A Bayesian approach to predict solubility parameters." Advanced Theory and Simulations 2.1 (2019): 1800069.

[20] Cvetković, Darja, et al. "Enhancing Hansen Solubility Predictions with Molecular and Graph-Based Approaches." Chemometrics and Intelligent Laboratory Systems (2024): 105168.

DOI: 10.1016/j.chemolab.2024.105168

[21] Pang, Jiayun, Alexander WR Pine, and Abdulai Sulemana. "Using natural language processing (NLP)-inspired molecular embedding approach to predict Hansen solubility parameters." Digital Discovery 3.1 (2024): 145-154.

DOI: 10.1039/d3dd00119a

[22] Li, Chunrong, et al. "Machine learning approach to predict Hansen solubility parameters of cocrystal coformers via integrating group contribution and COSMO-RS." Journal of Molecular Liquids (2024): 125319.

DOI: 10.1016/j.molliq.2024.125319

[23] Saini, Vaneet. "Machine learning prediction of empirical polarity using SMILES encoding of organic solvents." Molecular diversity 27.5 (2023): 2331-2343.

DOI: 10.1007/s11030-022-10559-6

[24] Abbott, Steven. "Solubility science: principles and practice." University of Leeds: Leeds, UK (2017): 109-110.

[25] Chen, T. & Guestrin, C. (2016). XGBoost: A Scalable Tree Boosting System. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining.

DOI: 10.1145/2939672.2939785

[26] Dorogush, A.V., Ershov, V. & Gulin, A. (2018). CatBoost: gradient boosting with categorical features support. arXiv:1810.11363.

[27] Ke, G. et al. (2017). LightGBM: A Highly Efficient Gradient Boosting Decision Tree. Advances in Neural Information Processing Systems.

[28] Dietterich, T.G. (2000). Ensemble methods in machine learning. Multiple Classifier Systems.

[29] Akiba, T. et al. (2019). Optuna: A Next-generation Hyperparameter Optimization Framework. Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining.

DOI: 10.1145/3292500.3330701

[30] RDKit: Open-source cheminformatics. https://www.rdkit.org

[31] Moriwaki H, Tian Y-S, Kawashita N, Takagi T (2018) Mordred: a molecular descriptor calculator. Journal of Cheminformatics 10:4

DOI: 10.1186/s13321-018-0258-y

[32] Rofik, Rofik, and Nurul Hidayat. "Improving the Accuracy of the Logistic Regression Algorithm Model using SelectKBest in Customer Prediction Based on Purchasing Behavior Patterns." Future Computer Science Journal 1.1 (2023): 9-17.