S is for spin

The ancients observed that certain materials had a property they called 'magnetism'. Navigators figured out that they could use lumps of material that had this property to find their way, as if left to their own devices these lumps would orient themselves to point North-South. When travelling up the East Coast of Australia in the early 1770s Captain James Cook found a whole island that had enough magnetic material in it to throw off his compass. He called it 'Magnetic Island'.

But it wasn't until the verification of the physical existence of atoms about a century ago that we started to understand where magnetism came from. Indeed, it really wasn't until we understood that electrons have spin that we began to realize the deep connection between that property of subatomic particles and the macroscopic phenomenon of magnetism. At the level of Physics 253 we teach that materials become permanent magnets because of the fact that atomic spins line up together. Prof. Art Smith just finished teaching Physics 253, and that part of the course must've had particular interest for him, because his research involves investigating the details of how magnetism arises at the atomic level. Prof. Smith's laboratory uses polarized scanning-tunnelling microscopy to map the spin-dependent electron densities in solids. As he described in his talk on Friday, the ability to make detailed maps at the atomic scale allows us to get data on how magnetic properties are correlated with structural and electronic properties of the materials in question. In other words, to understand how microscopic dynamics produces macroscopic magnetism.

Prof. Smith showed a number of nice examples of such experiments (and associated theoretical interpretation). I am afraid though that I left the talk without an overall picture of what those structural and electronic properties that are crucial to magnetism are. But maybe that is part of the point. Whether something is a ferromagnet or an anti-ferromagnet may ultimately be a very delicate thing, dependent on many details of the doping, the way the material is produced, the electronic structure of the atoms in question, etc. etc. Indeed, as Prof. Kordesch pointed out in his question after the talk, it may even be affected by whether one is looking at a 2d piece of material (a surface) or a 3d chunk of it (bulk). Or are there some generic things one can say about magnetism that I missed?

But the quest for such understanding is of more than just academic interest. The ability to probe the behavior of electron spin in solids at the atomic level holds out the hope that one day we could control the spin at that level. As Prof. Smith explained in his Introduction, this could lead to a new generation of electronic devices, based not on the transport of overall electron number (=charge), but on the transport of electron spin. It does seem, though, that that it will only be (at least) the next generation of experiments in Dr Smith's lab that tell us how to build such 'spintronic' devices. At this point he is investigating the structural properties of electron spin, rather than the transport properties. So, for the time being we will have to content ourselves with a better understanding of just what determines why some materials can be used to make fridge magnets and some can't.