Dr Michael Böhme is a chemist who has developed a theoretical approach to predict the magnetic...

Image: Jan-Peter Kasper (University of Jena)

**Most of us probably know magnets as the colourful badges that we stick on our fridge door, as the buttons on our handbag, or as moveable com-pass needles. However, numerous electrical devices—from the telephone to the particle accelerator—only function thanks to their electromagnetic components. Chemists at the University of Jena are investigating a special type of magnets—they are meticulously working on magnetic polymers that could be used to store large volumes of data. **

Dr Michael Böhme is using computational chemistry to work on molecular chains that behave like tiny magnets. The chemist describes the object of his research: »These are polymers in which a large amount of magnetic metal ions, such as cobalt, are lined up like a pearl necklace«. The individual metal centres form their own magnetic domains, which can store magnetic information.

If, at some point in the future, we want to use these magnetic molecules as data storage systems, we have to exactly understand and predict their properties, which is technically infeasible at the moment. »These systems are highly complex,« explains Böhme. The chains are not infinitely long in reality, which means their ends also affect the properties. »And the metal centres are not identically structured. The order in which they are arranged also has an effect on the magnetism that we can ultimately observe in the experiment«. This puts to the test all previous theoretical models that researchers have used to interpret and predict the properties of the tiny »bar magnets«

Therefore, Böhme is simplifying his calculations by firstly looking at molecular rings of various sizes instead of an endless chain of molecules. He has been working with Prof. Dr Winfried Plass from the Institute of Inorganic and Analytical Chemistry at the University of Jena to develop a computer model that can be used to better interpret the experimental data of the real molecules and predict their magnetic properties with greater accuracy.

Until then, however, there is another problem to be solved: »The computers at our disposal are not powerful enough to calculate the properties of long chains. They need around one week to carry out ab initio calculations for one single metal centre. It simply isn’t feasible to calculate a complete domain from several centres with the current technology,« explains Prof. Plass.

The »Ising model« was developed way back in the 1920s to study magnetic molecular chains in a highly simplified manner. »The Ising model has essentially been used for a hundred years,« says Plass.

Michael Böhme has now developed a less idealized model on the basis of ab initio calculations, which is closer to reality than the Ising model. »In addition to the actual metal centres, the links that facilitate interaction between the magnetic centres are also important,« explains Böhme. »We can obtain this information by adapting the theoretical model to the actual measurement data, which ultimately enables us to calculate the domain properties. This also allows us to make predictions as to the behaviour of previously unknown sin glechain magnets«.

Instead of calculating an endless chain, Böhme has applied his model to rings with three, six, nine, and twelve members. »Twelve is the highest possible amount for us, because there are 4,096 possible states that have to be calculated,« explains Michael Böhme. »However, we can then extrapolate this data to accurately predict the properties of longer chains«. Winfried Plass highlights some of the potential future applications: »Magnetic materials are highly suitable for storing information. Individual magnetic molecules can store much more information than the current storage media, where individual areas are mag netized«.

By Marco Körner

Information

Original-publication: How to link theory and experiment for single-chain magnets beyond the Ising model: magnetic properties modeled from ab initio calculations of molecular fragments. Chemical Science (2019), DOI: 10.1039/C9SC02735A