Cooperation of Kyrion student research group and Stanford University

Józef Koszewski, a student of the Faculty of Chemistry at the Warsaw University of Technology and President of the Kyrion Student Research Group of Materials Engineering, met Dr Amir Eskanlou during a scholarship stay at Stanford University in California. That was when he began collaborating with researchers from the foreign university, including by recording an interview with Dr Eskanlou, later published on Kyrion’s online channels. This marked the beginning of the group’s collaboration with Stanford – particularly with its research team Mineral-X.

The collaboration grew into a research project focused on critical materials – strategic resources that are essential for the economy but limited in availability. To support the Mineral-X team, three members of Kyrion – Józef Koszewski, Łukasz Cieszyński, and Zuzanna Kluczek – developed an algorithm that predicts the zeta potential of a sample with a given composition under specific conditions.

“It is a machine-learning model built on manually selected literature data and a series of potential measurements for known ores. In this way, we combined materials engineering, chemistry, and computer science. The three-member team from WUT included students with backgrounds not only in chemistry but also in geology and physics. Kyrion approaches materials engineering from multiple perspectives – this algorithm was created through collaboration as a student initiative,” explains Józef Koszewski, who led the WUT team in the project.

Extracting rock is only the first step in the processing chain – before a pure substance reaches factory production lines, it must be properly prepared. Flotation technology is one of the cornerstones of modern mining and enables the separation of valuable resources from impurities and unwanted components of ores.

The excavated material is crushed and then separated based on the differences in the wettability of its components. The efficiency of flotation depends on the composition of the ore as well as the environment in which it is placed – the entire process takes place in an electrolyte (liquid).

“Depending on the properties of the liquid used, chemical substances interact with it in different ways. This means there are optimal conditions under which flotation occurs smoothly and efficiently. Mining and chemical engineers work to find and improve these conditions. However, the number of factors affecting this equation is surprisingly large. One of them is the zeta potential. Although crucial to the process, its empirical measurement is time-consuming and resource-intensive. Predicting the zeta potential using an ML model makes it easier to estimate results and set expectations for researchers in geology, chemical, and mining engineering,” adds Józef Koszewski.

From the very beginning, the project has been a student-driven initiative. As members of the Kyrion group emphasize, their goal is to inspire other students to pursue their own ideas and bring them to life. Combining mining, chemistry, and machine learning may sound abstract, but as our students show, it all started with Józef conducting an interview with a foreign researcher.