Prediction of Solvents and Non-Solvents for Polymers Using Machine Learning Techniques

As renewable energy sources are increasingly gaining traction as an alternative to scarce fossil fuels, exploratory research on energy storage devices which are reliable, cost-effective and have low energy losses has become prolific. This has spearheaded the development of solid polymer electrolytes for safer lithium-ion batteries with comparable properties as conventional liquid electrolyte-based lithium-ion batteries. Synthesizing solid polymer electrolytes involves preparing an electrolyte film from a solution of polymer and Li-ion salts in an appropriate solvent. Identifying appropriate solvents for the polymer solution is usually done by trial-and-error, and therefore, is time-consuming. To mitigate this problem, quantitative measures of solvent-polymer miscibility known as solubility parameters have been developed in the past. In the present study, we first assessed the performance of the Hildebrand and Hansen solubility parameters to predict solvents and non-solvents for a set of benchmark polymers. Based on the results of the assessment, a machine learning model was trained on a dataset of known polymer Hildebrand solubility parameters to predict the solubility parameter of a queried polymer. Matching the predicted value with known solvent solubility parameters was then utilized to identify suitable solvents and non-solvents for the queried polymer. This capability has been implemented at www.polymergenome.org.