Looking over the chemical literature with an RSS reader can really give you a sense of what the hot topics are, and what's cooling off. Remember when it seemed as if every third paper was about ionic liquids? You still see work in the area, but it's nowhere near as crazy as it was. (I had a colleague come by my office the other day and ask "Did anyone ever find out what to do with those things?") Similarly, gold catalysts have been all over the place in recent years, but seem, to my eye, to be calming down.
Some of these things are research areas that look promising, but die off when their limits become apparent. Some of them are almost sheer fads, with papers coming out from all sorts of odd places because the authors want to get in on the hot, publishable topics while they can. Others keep going because the topics themselves are important but ver hard to exhaust (metal-catalyzed couplings come to mind).
And there are areas that keep going in the literature because they look like they should be important and useful, and eventually will, but no one can quite get them to either work generally enough or get people to recognize that they do. The metal-catalyzed coupling literature was in this shape back in the 1970s and into the 1980s - there were a lot of disparate reactions that you could do with palladium, but none of them had exactly taken over the world. My vote for a current field in this protostar state is engineered solid-phase catalysis.
That may sound odd, since work on solid-phase catalysts has been going on for decades, and is of huge industrial importance. But many of the important catalysts have been arrived at either by luck or by an awful lot of hard slogging. The field is complicated enough - fiendishly so - that it's hard to draw general conclusions. If you have a good solution-phase catalyst, how do you make a solid-supported variety that works just as efficiently? Well. . .if you really want one, you make about a zillion variants and hope for the best, as far as I can see.
Part of the problem (as with the metal-catalyzed coupling world) is that there are just so many variables. The solid supports alone are enough to keep a person occupied for life, what with all the various aluminas, silicas, zeolites, polymers, mesoporous engineered thingies, and so on. Then you have the uncountable schemes for linking these surfaces to active catalysts - what functional groups to use, what density things should be on the surface, what distance you need between the surface and the catalyst, etc. And just linking up to the known catalysts is no light work, either, since most of these things were not made with convenient handles hanging off them.
As we get better at making (and characterizing) new kinds of surfaces and new kinds of macromolecular assemblies, we might start to get our hands around this subject. For now, though, it seems to be mostly in the descriptive stage: papers are of the "Hey, we made this thing and here's what it does" variety, with further work in the series being "Hey, remember that stuff we made? Turns out you can do this with it, too - who knew?" What you don't see, or not too darn often, is a paper describing the general principles of these processes. For the most part, we don't know them yet.
But if I had to pick an area that will eventually blossom into a host of applications, this would be high on the list. It's a mixture of surface chemistry, materials science, nanotechnology, and organic synthesis, and it's got a lot of promise. But then again, it's had a lot of promise for a long time now. . .