2025 AIChE Annual Meeting

(398n) Thermodynamics-Informed Learning of Universal Fundamental Relations to Evaluate Polymer-Solvent Binary Mixture Phase Properties

Authors

Oliver Xie, Massachusetts Institute of Technology
Bradley Olsen, Massachusetts Institute of Technology
The ability to accurately predict mixture phase properties is imperative to addressing core questions of chemical engineering, such as the conditions necessary for polymerization reactions1,2 and the formation of biomolecular coacervates.3-5 A novel approach to thermodynamic mixture modeling has accurately predicted mixture phase properties for binary small molecule mixtures using only pure component properties.6 It does so by combing a theory-driven functional form with experimental data to learn the universal fundamental relations (UFR) of mixtures and the thermodynamically-consistent pure component parameters. Results indicate that by training the model on only infinite dilution activity coefficient (IDAC) data, the UFR model is able to predict VLE and LLE curves outside the training data set. This sets it apart from all other pure component thermodynamic mixing models.6 This work aims to extend the model from small molecule binary mixtures to polymer-solvent binary mixtures. Adaptation of the UFR model to polymer-solvent mixtures is crucial as there is currently no thermodynamics-informed models for predicting polymer phase separation. The model extension is achieved by first compiling property data for a set of polymers as well as polymer-solvent mixture data and adding it to the graph network. Then, additional layers, including a polymer molar mass adjustment, are added to the Gibbs free energy model to capture polymer-solvent mixture behavior that is not present in small molecule mixtures. Future work will include formulation of the loss function to address penalties necessary for polymer-solvent mixtures and comparing the predicted phase diagrams to literature data.

References:

  1. Warren, N. J. & Armes, S. P. Polymerization-Induced Self-Assembly of Block Copolymer Nano-objects via RAFT Aqueous Dispersion Polymerization. J Am Chem Soc 136, 10174–10185 (2014).
  2. Zhang, F., Fan, J. & Wang, S. Interfacial Polymerization: From Chemistry to Functional Materials. Angewandte Chemie International Edition 59, 21840–21856 (2020).
  3. Alberti, S. & Hyman, A. A. Biomolecular condensates at the nexus of cellular stress, protein aggregation disease and ageing. Nat Rev Mol Cell Biol 22, 196–213 (2021).
  4. Alberti, S., Gladfelter, A. & Mittag, T. Considerations and Challenges in Studying Liquid-Liquid Phase Separation and Biomolecular Condensates. Cell 176, 419–434 (2019).
  5. Ladouceur, A.-M. et al. Clusters of bacterial RNA polymerase are biomolecular condensates that assemble through liquid–liquid phase separation. Proceedings of the National Academy of Sciences 117, 18540–18549 (2020).
  6. Xie, O. & Olsen, B. D. Thermodynamics-informed machine learning of universal fundamental relations. In-prep.