Last Agora of the month, a message from Jeff
Unless someone can think of a way of adding more days to the month of March, this will unfortunately be our last session on optical antennas. That means that we have a lot of material to cover in one hour! Well actually, most of the detailed math in the paper is behind us, so I’m hoping we can have a more qualitative discussion on sections 3.8 – 3.10 and specially section 4 of the paper. We didn’t go through section 3.6 yet but we can just briefly look at that (and actually I’m more interested in exactly HOW one goes about holding an 80nm gold ball 5 or 10nm from a radiating atom in order to obtain the enhancement shown in figure 6!). The linear dipoles shown in 3.10 seem of somewhat more interest, and in this case he didn’t present any equations for us to ponder. Also I find very surprising him mentioning that metals can have highly non-linear properties discussed in 3.9 — do any of you have experience or knowledge of such nonlinearities and why they aren’t observed in normal situations?
I am still puzzled by one thing I brought up last time, concerning the downwards effective wavelength scaling of very thin linear elements according to eq. (39) which he doesn’t really explain at all. So I did find another paper by the same group (Novotny, Phys. Rev. Let. 98 266802, 29 June 2007) on that subject, and reading it didn’t help much at all. I find number of things troubling, including that the effective wavelength goes o zero for lambda ~= 400nm in silver and gold (unless that is just extending an approximation beyond its validity) as given by eq (39) where _1 < 0. That’s worrisome, especially if I were trying to build one of hese to work in the UV! But my bigger puzzle, if anyone can possibly shed ome light on this, is understanding QUALITATIVELY why there should be any uch reduction in effective wavelength. Why doesn’t this happen at lectronic frequencies? The reduction only becomes significant when the ntenna diameter R is of the order of the penetration depth. But that does NOT happen at RF when you have a very thin wire, so what gives?
Jeff
