PLASTILOOP

The PLASTILOOP 2. 0 industrial chair, coordinated by Professor Sébastien Paul (UCCS (Unité de Catalyse et Chimie du Solide) / Centrale Lille), coordinator of the REALCAT platform (integrated platform applied to the high-throughput screening of catalysts for biorefineries, Lille), brings together 4 academic laboratories (UCCS, ICV (Charles Violette Institute), the CRIStAL laboratory (Lille Research Centre for Computer Science, Signal and Automation and E2P2L) and an industrial company (Solvay) around the issue of recycling high-performance polymers.

Work on catalysis

The work of the Chair will be based on the REALCAT catalytic screening platform, which enables high-throughput research into chemical, biological or hybrid catalysis.

The scientific programme will be organised into three work packages:

• Work package 1 will study the biocatalyzed degradation of aromatic polymers with a view to reintroducing value into these wastes through the production of chemical synthons that can be used as a basis for the manufacture of new polymers.
• Work package 2 will focus on the modification of these monomer precursors in order to transform them into molecules that can be used directly in the polymerisation processes already implemented by Solvay.
• Work package 3 will complete the cycle by focusing on verifying the characteristics of these new monomers, to ensure that they correspond to the specifications expected by the industry.

This experimental set will be accompanied in its entirety by a final batch focusing on the analysis of the environmental impact of the processes involved using innovative human science tools.

Innovation within PLASTILOOP 2.0

The main innovation of PLASTILOOP 2.0 is the use of machine learning tools to understand and predict the catalytic systems involved. Indeed, in addition to developing new degradation and synthesis pathways using novel catalysts, PLASTILOOP 2.0 will have as its main task the programming of new prediction algorithms allowing the activity of these catalysts to be linked to the fundamental descriptors that characterise them. This approach will benefit from the unrivalled high-throughput screening capacity of the REALCAT platform, generating the amount of data necessary for their learning, which will ultimately considerably accelerate the development of these catalysts compared to the more traditional approaches of pre-existing projects.