Yesterday I spent much of the meeting talking with collaborators and old friends, and seeing some invited talks at the sessions associated with some of the APS prizes. There were several really excellent talks.
The first that really stood out was Daniel Fisher's talk on the occasion of his winning the Onsager Prize. Fisher is a statistical mechanician, and he gave a very clear all about randomness, using domain walls in random magnets as an illustrative case for his ideas. In ferromagnets, there is an energetic cost associated with having a domain wall between regions of differently oriented magnetization. That acts like a surface tension, with the system tending to try to minimize the length of such a boundary, all other things being equal. Now if you allow the magnetic coupling between neighboring spins to have a random variation, the domain walls take on funny shapes, "finding" the lowest exchange locations because that also lowers the energy cost. Fisher talked about the statistical physics of this system, including the characteristic slow, history dependent kinetics of equilibration. The tails of the distribution of exchange values are really important here. Fisher then finished up talking about evolution as a statistical mechanics problem, where instead of minimizing an energy, the system tries to maximize "fitness", which is essentially the difference between birth and death rates.
The other talks that were exceptional were those in the Buckley Prize session. The prize this year was awarded to John Slonczewski, who predicted, quantitatively, the existence of the effects of spin transfer torque, which I've indirectly discussed before. Since spin really is angular momentum, flowing a spin polarized current into a magnetized material exerts a torque on the magnetization, if the flowing spins are not aligned with M. This is a way of using currents to cause magnetic domains to precess (ferromagnetic resonance) or flip altogether. Luc Berger gave a very good talk outlining the history of this field in a very pedagogical way, harkening all the way back to work done eighty years ago. Dan Ralph in the same session spoke about their incredibly beautiful results demonstrating all of these effects with quantitative agreement with theory. Further, Ralph showed how one can pump spin currents like this and drive such systems using the spin Hall effect rather than just direct current flow. That's worthy of a blog post all of its own, which I will do sometime soon.