Machine Learning-Guided Prediction of Glass Transition Temperatures in Sustainable Polymer Blends
Keywords:
glass transition temperature; sustainable polymer blends; machine learning; uncertainty quantification; applicability domain; polymer informaticsAbstract
Glass transition temperature (Tg) governs the service window, processability, and long-term stability of polymer blends, yet its prediction in sustainable blends remains difficult because bio-based components introduce variable molecular architecture, moisture sensitivity, and complex phase behavior. Existing additive and group-contribution models are transparent but rely on tuned interaction parameters and extrapolate poorly when composition, processing history, or blend miscibility depart from calibration conditions. This manuscript develops a machine-learning framework for Tg prediction in sustainable polymer blends that couples physically grounded feature construction with regularized regression, uncertainty quantification, and an explicit applicability domain. The framework represents each blend through component mass fractions, molecular descriptors, interaction proxies, and processing variables; trains predictive models with a regularized objective; estimates predictive variance using ensembles; and flags unreliable queries using Mahalanobis distance. Evaluation is organized around composition-extrapolation and leave-one-blend-out splits so that reported errors reflect the design scenarios encountered in bio-based blend discovery. The analysis links common error modes—phase separation, plasticization by moisture, molecular-weight effects, and non-equilibrium glassy states—to specific features and validation choices, and it positions classical inverse-additivity and Gordon-Taylor relations as interpretable baselines rather than competitors to be dismissed. The resulting methodological contribution is a reproducible decision framework that can guide blend selection, identify data gaps, and support uncertainty-aware design of sustainable polymer formulations without overstating the evidence obtainable from sparse blend datasets.
References
1. Debenedetti, P. G., & Stillinger, F. H. (2001). Supercooled liquids and the glass transition. Nature, 410(6825), 259–267.
2. Angell, C. A. (1995). Formation of glasses from liquids and biopolymers. Science, 267(5206), 1924–1935.
3. Zhu, Y., Romain, C., & Williams, C. K. (2016). Sustainable polymers from renewable resources. Nature, 540(7633), 354–362.
4. Schneiderman, D. K., & Hillmyer, M. A. (2017). 50th anniversary perspective: There is a great future in sustainable polymers. Macromolecules, 50(10), 3733–3749.
5. Miller, S. A. (2013). Sustainable polymers: Opportunities for the next decade. ACS Macro Letters, 2(6), 550–554.
6. Utracki, L. A. (2002). Polymer blends handbook. Kluwer Academic Publishers.
7. Fox, T. G. (1956). Influence of diluent and of copolymer composition on the glass temperature of a polymer system. Bulletin of the American Physical Society, 1, 123.
8. Gordon, M., & Taylor, J. S. (1952). Ideal copolymers and the second-order transitions of synthetic rubbers. I. Non-crystalline copolymers. Journal of Applied Chemistry, 2(9), 493–500.
9. Van Krevelen, D. W., & Te Nijenhuis, K. (2009). Properties of polymers: Their correlation with chemical structure; their numerical estimation and prediction from additive group contributions (4th ed.). Elsevier.
10. Bicerano, J. (2002). Prediction of polymer properties (3rd ed.). Marcel Dekker.
11. Ramprasad, R., Batra, R., Pilania, G., Mannodi-Kanakkithodi, A., & Kim, C. (2017). Machine learning in materials informatics: Recent applications and prospects. npj Computational Materials, 3(1), 54.
12. Kim, C., Chandrasekaran, A., Huan, T. D., Das, D., & Ramprasad, R. (2018). Polymer genome: A data-powered polymer informatics platform for property predictions. The Journal of Physical Chemistry C, 122(31), 17575–17585.
13. Tao, L., Varshney, V., & Li, Y. (2020). Benchmarking machine learning models for polymer informatics: An example of glass transition temperature. Journal of Chemical Information and Modeling, 60(11), 5454–5462.
14. Miccio, L. A., & Schwartz, G. A. (2020). From chemical structure to glass transition temperature: A machine learning approach. Polymer, 202, 122614.
15. Chen, L., Pilania, G., Batra, R., Huan, T. D., Kim, C., Kuenneth, C., & Ramprasad, R. (2021). Polymer informatics: Current status and critical next steps. Materials Science and Engineering: R: Reports, 144, 100595.
16. Kuenneth, C., & Ramprasad, R. (2023). polyBERT: A chemical language model to enable fully machine-driven ultrafast polymer informatics. Nature Communications, 14, 4099.
17. Rasmussen, C. E., & Williams, C. K. I. (2006). Gaussian processes for machine learning. MIT Press.
18. Lakshminarayanan, B., Pritzel, A., & Blundell, C. (2017). Simple and scalable predictive uncertainty estimation using deep ensembles. In Advances in Neural Information Processing Systems (Vol. 30, pp. 6402–6413).
19. Sahigara, F., Mansouri, K., Ballabio, D., Mauri, A., Consonni, V., & Todeschini, R. (2012). Comparison of different approaches to define the applicability domain of QSAR models. Molecules, 17(5), 4791–4810.
20. Vovk, V., Gammerman, A., & Shafer, G. (2005). Algorithmic learning in a random world. Springer.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 International Journal of Materials and Chemical Sciences

This work is licensed under a Creative Commons Attribution 4.0 International License.