photoelectrochemical cells, quick & dirty

Alright, already. I wasn't planning on it, but there were enough curious comments on the last post that I'm actually going to tell you guys what that little device was.[1] However, I'm not in the lab now (else, I would be working and not writing, duh) so I'm short on notes and lit references (which are currently lying on my desk, in the lab...where I am not). Soooo...this is the quickie version from off the top of my head. Sadly, I have no pictures for you here. Feel free to correct me. (Especially YOU, Dave. Or YOU, Lakshmi. Or anyone else, really.)
Remember all those posts from a while ago on solar cells?[2] Well, I've been meaning to do a post on Graetzel (dye-sensitized) cells as well. Not that I've ever worked on them, but they're pretty cool. There's a back electrode (read: not transparent), and usually there's a lot of nanocrystalline TiO₂ in them. This functions as an n-type semiconductor. Attached to the nanocrystalline titania is a dye. The "nanocrystalline" bit here is important--it provides a higher surface area for dye adsorption. Since titania's bandgap is too high to have much to do with absorbing sunlight, the dye is rather useful. After all, if it's a dye, it's not white like TiO₂, so it's absorbing something, somewhere in the visible spectrum. Often, the dyes used for this purpose are ruthenium complexes, although price concerns may drive the development of alternatives. Usually there is some carboxylic acid functionality present, which enables the dye to adhere to the titania. Other (frequently oxide-based) semiconductors can be used in place of titania, which has the associated problem (advantage, maybe, elsewhere?) of nibbling on organic things.
So the dye absorbs light, and transfers an electron to the titania. And then...yeah, pretty much nothing, until I mention the electrolyte that is essential to the functioning device. What's generally used is I^-/I₃^-. This, along with the front electrode, provides a cute little redox shuttle to restore the dye to its original state.
As I said, no pictures, since I have never made one of these and am not in the mood to Photoshop a nice little figure for other people to steal.[3] However, if you're interested, there are instructions for DIY dye-sensitized solar cells on the cheap here. Things like this are especially nice for outreach (HINT HINT).
Back to the original purpose of this post (which has thus far been neglected), the nanopopsicle I showed you a few days ago is not, in fact, a Graetzel cell. It is similar, though--the major difference is that the semiconductor used in my little device has a sufficiently low bandgap to absorb light in the visible spectrum.[4] That, and all the materials are completely different. The only appearance carbon makes in the nanopopsicle is in the insulating coat of epoxy.
Speaking of which, epoxy is not such a terrific thing for XPS, so I would be thrilled if someone had other quick & easy ideas for an alternative.

[1] In the early days of CBC, no one said much of anything to a device post. I know many of you are just bored synthetic chemists looking to see neat little structures, but...when there are no comments, we get lonely. And we stop doing device posts. Have the audience members experienced a change of heart?
[2] I'm not terribly satisfied with it, so a follow-up may be in the works. You have been warned.
[3] Not that you would want to, but anything of mine that appears on CBC is free for the taking as long as I am credited (and, yeah, it would be nice if you let me know you were going to use it).
[4] Never mind, y'know, all that infrared light from the sun. Bandgap's not THAT low.