Making polymers with tailor made properties is often iterative. Typical development cycles are slow, even when a candidate polymer class has been identified. In this case, we used state-of-the-art machine learning tools to design and validate 17 new polyoxazoline variants that meet targeted cloud points over a range of temperatures. All this 10x faster than the traditional R&D approach. The impact of such materials could be realized in personal care, packages, mining and other applications.
A scientific article published by some of our cofounders in the journal npj – Computational Materials describes how machine learning enables the design and creation of polymer cloud point engineering through inverse design. Learn more here:
