More interesting talks on the third day of the meeting. Two invited talks stand out in particular for me. The first was by Prof. Peumans at Stanford, ostensibly on using surface plasmons to try to enhance photovoltaics, though it was actually more broad than that. In organic solar cells, there's a competition between trying to get good absorption of incident light (thus favoring thick polymer films) and trying to extract the charge efficiently (leaning toward thin polymer films). It would be great if there was a way to enhance absorption of light while preserving the thin film layers that improve charge transport. With conventional ray optics (the far field limit) in Si solar cells, this can be done by roughening the material surfaces. That way, you enhance the likelihood of total internal reflection, and you can capture a lot more light (proportional to the square of the index of refraction of the semiconductor). The question is, can one gain something by working in the near field and using interference, as opposed to working in the far field limit. The natural approaches that come to mind are: (1) use plasmonic structures as optical antennas, trying to focus and concentrate the light into the PV material; (2) make some kind of broad-band cavity to hold the light in; and (3) make some sort of waveguiding structure to increase the interaction of the light and the PV material. In the end, Peumans made a convincing argument that the near field doesn't have any fundamental advantage over ray optics approaches, but it can actually deliver better results in practice.
The other cool talk was by Dan Rugar at IBM, speaking about the latest results on using a form of cantilever-based force microscopy to do nuclear magnetic resonance imaging at the nanoscale. They now have the sensitivity to see ~ 1000 nuclear spins, with a spatial resolution of around 4 nm. It's a real tour de force experiment. The part that really impressed me was not actually the imaging that they did of the hydrogen (protons) in a tobacco mosaic virus. Rather, I was blown away that they could see, very clearly, the signal from the one or two monolayers of physisorbed hydrocarbon contamination (or, as we technical types call it, "goo") on the surface of their cantilever. Rugar has been working on this idea for years, and the progress has been very impressive.