Since their invention in 1984, permanent magnets based on Nd-Fe-B (neodymium, iron and boron) have been the material of choice for high-performance applications. They are now used, for example, in electric vehicles and wind turbines. However, heavy rare earths such as dysprosium and terbium – highly critical and scarce raw materials – are used to operate the magnets at temperatures of over 100°C. This makes the technology very costly, environmentally damaging and with supply chain risks.

Despite decades of research, no practical alternative to Nd-Fe-B magnets has been found. Tozman now wants to change that. She is researching magnets with components made mainly from iron along with samarium, some of which even have better magnetic properties than Nd-Fe-B and don’t require heavy rare earths and have less light rare earth. The researcher intends to optimise the magnetic properties of the components with innovative alloys.

As part of her MAG-TOOL project, Tozman, who works in the Functional Materials division (Department of Materials and Geosciences) under the leadership of Professor Oliver Gutfleisch, also wants to use artificial intelligence in the form of machine learning to predict the properties of different compositions. A cutting-edge toolbox will drastically reduce the number of experiments required. The combination of innovative experimental material processing and machine learning makes the research project unique. The aim is to develop a new medium- and high-performance magnet that protects resources and whose elements are safe, affordable and sustainable.