2.2 MEuro FWO grant for international consortium with Remco Havenith

The consortium consisting of researchers from the Universities of Antwerp (Billen, Marlen), Ghent (D’hooge, Van Steenberge), Brussels (Van Assche, De Proft), Oak Ridge National Laboratory (Cunha), and the RUG (Havenith) has received 2.2 Meuro from FWO for the development of the science for recycling polyurethanes.

Polyurethanes are versatile materials essential for modern comfort and insulation, but their complexity poses major challenges for circularity due to the diversity of (macro)monomers used. Recent efforts in chemical depolymerization revealed that the variety of polyols and isocyanates complicates separation, and reliable thermodynamic data for these (macro)monomers and derivatives are absent. As a result, predicting recoverability of the macromonomers remains impossible. Moreover, typical analytical descriptors for recycled resins are insufficient to ensure acceptance of high-recycled content in formulations. Discussions with polyol and polyurethane manufacturers show that even adapting virgin formulations is cumbersome, let alone for recycled resins, often leading to costly internal and external feedback loops, showing that structure-property relations in formulations are inadequately understood. To address this, MOSA-PUR develops a quantum-to-plant modeling workflow to elucidate structure–process–property relations, enabling optimized chemical recycling and more efficient formulation design. Kinetic insights are obtained via (conceptual) density functional theory, calibrated against advanced thermal analysis, feeding into kinetic Monte Carlo reaction engineering to predict polyol structures and depolymerization pathways. Thermodynamic data of (depolymerized) resins and molecular-scale physicochemical insights are generated via molecular dynamics, enabling process modeling and in-silico estimation of recycling process energy—an unprecedented step. To facilitate the translation to industrially relevant software, MOSA-PUR will build surrogate models trained on a variety of experimental and multi-scale modeling outputs.

More info? Contact Dr. Remco Havenith