A comment on the most recent post got me thinking about photochemistry. I've done that, but just for one project, and I'll bet that many other chemists have had the same pattern.
A lot of people who've never done light-powered chemistry find the idea neat - no nasty reagents, no masses of inorganic salts to remove, just shine a light on the flask and you're done. It seems like Chemistry Of The Future - you feel as if you should be wearing some sort of white jumpsuit with padded shoulders if you ever get around to setting some up.
The idea is appealing, but the reality is less so. Photochemistry often isn't as clean as you picture it being - a look at your clear starting solution gradually turning orange and brown under the punishing glare of the UV lamp tips you off about that. Things get hot in those setups, too, which also doesn't fit the sleek, cool, futuristic template. When you take a hot flask of darkened gunk off the lamp, it's hard not to wonder what would have happened if you'd just cooked it in the oil bath the old-fashioned way.
Probably not what happened under illumination, though. Light does do some odd stuff, and there are some neat-looking reactions that can be run that way. The problem is, many of those neat reactions are free-radical mechanisms, and that's what leads to a lot of that colorful crud. There are a lot of concerted mechanisms that can be driven photochemically, and those should (in theory) be cleaner, but in my experience, it can be hard to keep a lot of radical chemistry from going on, and it can swamp the cleaner stuff right out.
Radical reactions were all the trend back when I was in grad school (get off my lawn!), but while they've never disappeared, they've never caught on to become an essential part of every organic chemist's toolkit. There are several well-used reactions that run (or can run) by single-electron processes, but as a class, free radicals still have an exotic, slightly disreputable look to them. People will look at a potential transformation on the board, and say "Hmm, I bet I could do that by a dipolar cycloaddition", or "I'll bet that I'll able to do an olefin metathesis to get that". There are dozens of reaction classes that you reach for without thinking twice: metal-catalyzed coupling, epoxide opening, reductive amination, electrophilic ring substitution. But do you reach for a free-radical closure to a five-membered ring, a well-trodden radical process if ever there was one? Well, I don't, anyway - and I've done the things. Do you?