Unveiling the Secrets of Dissolving Polymeric Materials with Machine Learning
Originally Published 2 years ago — by Phys.org
Scientists have developed a machine learning system that can predict the miscibility of polymers with organic solvents, a crucial step in recycling plastic waste. By integrating data from quantum chemistry calculations and real experiments, the model can accurately determine the compatibility of different polymer-solvent mixtures. This enables the selection and design of solvents for separating specific types of plastics in recycling processes. The model is about 40 times faster than conventional calculations and has the potential to streamline materials development and improve recycling rates. The researchers have made part of the source code and data available to the public to encourage open innovation and crowd-sourcing of data for further improvements.