This project focuses on advancing thermodynamic modeling techniques for accurately calculating solubility and solution free energies, with particular emphasis on electrolyte systems. The focus is on improving and expanding SCM’s current methodologies, specifically COSMO-RS, Pitzer-Debye-Hückel, and machine learning–based approaches, so they can better predict solubility across a broader range of chemical environments, specifically for charged species. 

A central challenge lies in properly describing short-range molecular interactions, captured by COSMO-RS, and the long-range electrostatic interactions described by the Pitzer-Debye-Hückel models. Achieving this requires developing consistent and transferable parameter sets that can represent both effects simultaneously while maintaining chemical accuracy and computational efficiency. In addition to improving theoretical models, the project aims to leverage data-driven methods to enhance prediction quality and reduce computational effort. 

Ultimately, the project contributes to the PREDICTOR’s broader objective of high-throughput screening of active materials for flow batteries, by improving the prediction of solubility, solution-free energy and other relevant thermodynamic properties, critical for the development of materials for electrochemical storage.