Follow-up: Workshop on "Surface Plasmons, Metamaterials, and Catalysis"

The workshop here was, I think, very successful.  It was great to get so many knowledgeable people together in one place to talk about these issues.  Here are some key points that I learned from watching the talks and speaking with people:

• While it may take years and $$ to build, single-molecule tip-enhanced Raman spectroscopy in variable-temperature UHV looks like an amazing capability.  Turns out that vibrational lineshapes that look gaussian at room temperature look like Lorenztians at cryogenic temperatures, as one would expect for true (not inhomogeneously broadened) resonances with well-defined lifetimes. 
• It's possible to get large electromagnetic field enhancements out of high-index materials even without plasmon resonances.  Clearly it is worthwhile to try to create geometries where the local high intensities occur close enough to a surface interface that charge carriers produced there are able to diffuse to the surface to do chemistry.
• Careful engineering of structures can lead to very high (say 90%) absorption even in extremely thin coatings.  That is, you can imagine making anti-reflection coatings that are tuned to produce absorption in materials where the loss mechanism can be used to do chemistry.
• By driving plasmon resonances detuned from the resonance, it is possible to favor net motion of charge (generating photovoltages that can be measured in experiments).  Conversely, gating structures to alter their total charge density can tune their plasmon resonances (though it's still not clear to me how one would make this a big effect).  Shifting the energies a little with charge is something I understand; drastically changing the intensity of the resonances with charge is much more mysterious to me.
• Very specific defect sites can drive very particular catalytic reactions.  In some special cases theory can show how this works, but I can't help thinking that there are many issues left to resolve.  For example, in discussing catalysis most people draw energy level diagrams, correctly showing that energy conservation is really important when, e.g., a hot electron in some solid is able to occupy (transiently) an unfilled molecular orbital in a molecule of interest.  However, there are other issues that affect which electronic processes can happen (e.g., momentum conservation; how incoherent hot electrons or holes can excite plasmons, which are coherent e-h excitations) that seem to be given short shrift.
• When an electron transiently occupies an empty molecular orbital, you can think of that as delivering an impulse to the molecule's nuclei.  (The equilibrium bond lengths differ when the molecule has an extra electron; thus passing through that state is like kicking the nuclei.)
• Single-particle vs. ensemble studies (via plasmons and optics) can give insight into processes like the evolution of phase transitions with size - such as the formation of palladium hydride.
• Doped Si can support nice plasmon excitations out in the mid-far IR, and could be very interesting from several points of view.
• While it's very trendy to worry about water splitting, there are a huge number of other reactions that are important and complicated!  For example, converting CO₂ to methane is eventually an eight electron process (!).

Those are just a few points.  Teaching, etc. made me miss some of the talks as well.  All in all, good talks and good discussion.  There is clearly an enormous amount still to be done in this field, both on the experimental and theory sides.