Department Environmental Chemistry

Chemical pollution is a major threat to ecosystems and human health and current regulations seem
insufficient to prevent widespread chemical contamination. Further, environmental chemistry, (eco-)toxicity and environmental engineering oftentimes focus on monitoring and assessing chemicals already in commerce and prioritize them for risk mitigation, which mostly entails end-of-pipe measures or bans. In contrast, our goal is to build a prototype framework (experimental and in silico) that enables designing non-hazardous chemicals in the first place. We develop novel assessment methods that are fast, yet robustly predictive of key environmental hazards and that can be fully integrated into early stages of the chemical design process.

Therefore, we develop:

• high-throughput screening experiments for two key hazard properties of chemicals, namely persistence and (eco-)toxicity.
• in silico models that will make it possible to predict relevant hazard properties, also for untested chemicals.
• generative models for computer-aided identification of novel molecular structures that combine an acceptable environmental hazard profile with targeted chemical functionality.

While we aim to provide generally valid experimental and machine-learning methods, we use the substance class of antioxidants, as they are high production volume substances with partially concerning environmental hazard profiles, as a proof of principle to validate our methods and demonstrate their application.