Studying Magnets the Size of Atoms

Therese Eriksson
Materials Chemistry
Uppsala University
October 2003

I make new chemical compounds that are magnetic by combining different metals. My research is part of a collective effort between physicists and chemists at our materials research laboratory, where we study properties of selected metallic compounds to understand what makes them more effective magnets. Materials with improved magnetic properties can be used to create new computer memories that work faster and can store more information.

An ordinary bar magnet has two opposite poles. If we break it in half, each half also has two poles. When we have divided the magnet into its very smallest building blocks, what we are looking at is actually the individual atoms, each acting as a small bar magnet, all arranged with the poles in the same directions. The reason why the bar magnet sticks to your fridge is that all the atomic bar magnets point in the same direction. But why do they do that in the material of the bar magnet and not in any material? To understand that, we need to look at the interactions between the individual magnetic atoms. In our common effort to increase the understanding of why a certain material gets certain magnetic properties, my contribution is on the atomic level. I study what the distances are between an atom and its neighboring atoms, determine the geometry of how the atoms organize around each other, and identify of which chemical elements the nearest neighbor atoms are. These three factors together are what we call the atoms 'surrounding'. I then move on to determine in what direction each atomic bar magnet points. They can lay parallel, point in opposite directions, or lay with an intermediate angle between them.

It is not possible to directly view the atomic interactions. The main experimental method I use to reconstruct that information is called neutron diffraction. A neutron is one of the small particles that compose the atom. It has no charge, but it is magnetic and that is why it can be used to look at the directions in which the atomic bar magnets point. A beam of neutrons can be produced by a nuclear reactor. When it hits a sample magnet, the neutrons bounce on the atoms in the sample and come out on the other side with a characteristic pattern. By analyzing the pattern, I can determine the distances between the atoms, the geometry of how the atoms surround each other, and the directions of the atomic bar magnets.

So far, my results are that I have been able to prepare two new compounds by melting the elements iridium, manganese and silicon together in proportions that no one has studied before. For these compounds, I can describe the atoms' surroundings, and I have also been able to identify how the atomic bar magnets point relative to each other. In co-operation with physicists, some colleagues are now measuring the strength of the magnetic interactions and are also studying how the interactions are affected by temperature changes. Other colleagues have been able to create a theoretical model for one of the compounds, which can reproduce the magnetic properties found from these experiments. For the compound under study, we found that the particular distance between the nearest neighbors´ atomic bar magnets give them a strong will to lie in opposite directions to each other, but that the geometry of how the atoms are organized prevents this, and leaves the compound frustrated. Thus it is a weak magnet.

So far this discovery is very specialized knowledge valid only for this compound, but by studying more and more compounds, more generalized conclusions can be drawn. By combining knowledge of the atoms' surroundings with the direction that its bar magnet is pointing relative to the directions of the neighboring atomic bar magnets, we hope to gain a better understanding of how this local interaction governs the collective magnetic properties of materials in general. If we can understand this, we will be able to design materials that have exactly the magnetic properties we want, at the temperatures we want to use them.