- Kemira and CuspAI used generative AI to design novel metal-organic frameworks (MOFs) targeting PFAS molecules GenX, PFBS, and PFOS.
- CuspAI screened about 300 trillion MOF structures, delivered over 5,000 candidate designs and narrowed these to roughly 20 priority candidates.
- The discovery phase was completed in six months and the programme is moving into development and testing.
- Objective: selective, stable, sustainable and manufacturable PFAS removal at trace concentrations (sub-ppb/parts-per-trillion) amid tightening US EPA and EU drinking-water limits.
Overview
Kemira and CuspAI used generative AI to design novel metal–organic frameworks (MOFs) targeting removal of PFAS molecules GenX, PFBS and PFOS from drinking and process water at trace concentrations.
Discovery results
The platform explored an estimated 300 trillion MOF structures, generated over 5,000 candidate designs with property data for the three targets, and narrowed these to roughly 20 priority candidates; the discovery phase took six months and revealed new functional-group chemistries with adsorption potential.
Technical targets and regulatory context
The project brief prioritized materials capable of removing PFAS at sub-ppb/parts-per-trillion levels that are water-stable, environmentally compatible, synthesizable, manufacturable and cost-effective, responding to tightening limits such as US EPA parts-per-trillion MCLs and the EU Drinking Water Directive and seeking alternatives to granular activated carbon.
Next steps
Selected candidates will proceed to further development and testing, and additional programs across other material classes are being scoped under the partnership framework.
