Henrik Hupatz. Photo: Private.

What is your project about? 
"The project aims to develop a sustainable workflow for the fast and sensitive detection of polar pesticides using a derivatizing reagent. We use an inverse design approach: rather than synthesizing and testing reagents in the lab, we include the desired properties in the reagent design. This involves a generative machine learning model that creates reagent structures, supported by predictive machine learning models for property estimation. This inverse design approach minimizes the need for experimental lab work, making the process more sustainable. By integrating environmental properties, like toxicity, early in the design phase, we enhance the sustainability of the reagents."

What is exciting about this project?
"It is exciting but also challenging to explore the potential of generative models in chemistry for sustainable design. I am very interested in investigating how the prediction of machine learning models can be put into practice in the lab. This project also gives me the opportunity to broaden my expertise from organic and analytical chemistry into machine learning and environmental chemistry—and I’ve already learned a tremendous amount in the process."

What were the expectations before starting and how did it evolve during the last year of ongoing work?
"Initially, I expected to be working in the lab very quickly and test out different reagents. However, after a thorough literature research and initial tests, we shifted to prioritize developing the machine learning workflow, which has since become a central focus of the project."

What are the challenges when you are responsible for an interdisciplinary project?
"One of the biggest challenges is building a strong theoretical foundation across all areas involved. For example, I needed to develop a solid grounding in programming and statistics early in the project. Effective communication across disciplines is also essential; it’s crucial to understand each discipline’s methodologies and limitations to facilitate collaboration."

What can SUCCeSS do to improve the connection between academia and society?
"SUCCeSS built some connections between academia, industry, and sustainability networks. We, researchers, often find ourselves in an academic “bubble” with limited interaction beyond it. Strengthening connections with society could involve attending non-academic events, engaging on social media, or writing for non-academic publications to share insights more broadly."