A research collaboration led by Chalmers University of Technology has won funding to develop a method for destroying PFAS chemicals using artificial intelligence and photocatalysis, according to a Chalmers news article published on 12 November 2026.

The project, named Enabling PFAS Destruction Through Neuro-Symbolic AI Discovery of Organic Photocatalysts (END-PFAS), brings together researchers in chemistry, AI and materials science from Chalmers, Örebro University and Karlstad University. It is one of four projects selected for funding through a joint call by the strategic Wallenberg initiatives WASP (Wallenberg AI, Autonomous Systems and Software Program) and WISE (Wallenberg Initiative Materials Science for Sustainability). A total of 43 applications were submitted.

The selected projects address strategic, high-priority research challenges with international impact and visibility, the article states.

PFAS, or per- and polyfluoroalkyl substances, are synthetic chemicals widely used in everyday products such as water- and stain-resistant textiles and leather, cleaning products, smartphones and certain building materials. Known as "forever chemicals," they are extremely stable and do not break down easily in nature. PFAS have been shown to accumulate in animals and humans and can have toxic effects, including adverse effects on the immune system, the article notes.

The END-PFAS project focuses on destroying PFAS after they have been removed from contaminated water and soil, rather than simply storing them. Photocatalysis is a chemical process in which light activates a material, called a photocatalyst, accelerating a chemical reaction without consuming the catalyst itself.

"We want to use organic photocatalysts to reduce the environmental impact. But designing or identifying the best candidate among countless possible structures is a time-consuming challenge. It requires many hours of laboratory work. The idea behind this project is to use machine learning to predict which photocatalyst structures are most promising. We chemists can then synthesise them and test them experimentally," said Ergang Wang, Professor at the Department of Chemistry and Chemical Engineering at Chalmers and active within WISE, as quoted in the article.

Moa Johansson, Professor of Artificial Intelligence at Chalmers' Department of Computer Science and Engineering and affiliated with WASP, is co-leading the project. "This is an area where we get to test our work in neuro-symbolic AI on a problem of real importance for the natural environment. We have a very good team of researchers coming together and I very much look forward to working together," she said, according to the article.

The project partners include Ergang Wang, Moa Johansson, and Maria Abrahamsson at Chalmers; Moyses Araujo at Karlstad University; and Zuidberg Dos Martires and Luc De Raedt at Örebro University. The funding amounts to SEK 25 million over five years from WISE and WASP.

For university students, particularly those studying chemistry, materials science, AI, or environmental engineering, this project represents a concrete example of how interdisciplinary research can tackle a pressing environmental challenge. It also highlights growing career opportunities at the intersection of AI and sustainability, where machine learning is increasingly used to accelerate materials discovery. Students interested in these fields may find relevant thesis topics or future research positions within the WASP and WISE networks.