Shameless self-promotion part II. The actual scientific result that just came out in Nature is rather surprising. There are two "ordinary" settings for Kondo physics: a magnetic impurity in an otherwise nonmagnetic host (e.g., dilute Mn atoms in Cu), or a quantum dot containing an unpaired electron. In the former case, the conduction electrons of the host metal can lower their kinetic energy by trying to occupy a singly occupied d orbital of the magnetic impurity. However, because of the Coulomb repulsion of the other electrons on the impurity atom, really doing this is classically forbidden by energy conservation. Still, quantum mechanics lets that forbidden state exist as a virtual intermediate state in a scattering process that takes a conduction electron from the host, flips the spin of the impurity atom, and spits out an electron into a different conduction band state. In the quantum dot case, an analogous magnetic dance takes place, in which the spin of the unpaired electron on the dot is flipped, and an electron is transferred across the dot. This Kondo scattering process affects the electronic conduction through the dot in a particular, identifiable way.
The surprising result in our case is that we see indications of this Kondo process in atomic-scale junctions between chemically homogeneous (e.g., all the atoms are Ni) ferromagnetic metals. The data are pretty clear, and indicate that this spin-related process competes with ordinary ferromagnetic exchange in these nanostructures. It would appear, from accompanying theory calculations by our coauthors, that the very act of whittling the ferromagnetic metal down to the atomic-scale junction is enough to mess with the electronic properties of the metal that we'd ordinarily consider to be intrinsic. The bottom line is, when worrying about the magnetic properties of truly nanoscale structures (with many surface atoms), one may need to keep track of relatively exotic ("strong correlation") physics like the Kondo effect.