One of the themes of chemistry over the last few decades has been the ability to pay attention to increasingly fine-grained details. We have new instruments that measure smaller and smaller samples and in shorter slices of time, and we’re seeing that the world that we’ve been used to is just an averaged-out look at reality, which is really a superposition of many smaller, stranger worlds.
What we’re used to is compounds in solution in a flask, or solid samples in a bottle, but as you zoom in on these things it all gets odd. Think of a powder sample. Maybe it’s crystalline, but even if it’s amorphous there could be amorphous material interspersed with several different crystalline forms, present as microcrystals. Either way, imagine a cubic crystal sitting there, looking like a tiny grain of salt. The molecules in the middle of that crystal are actually having a very different experience than the ones on the surface faces, aren’t they? And those in turn are in a different world than the ones out on the exact corners, which are much more weakly bound to the lattice than the others – they will likely be the first to dissolve off if the crystal is dropped into a solution, for one thing.
Or think of a pure compound dissolved in a solution. For similar reasons, the molecules up near a liquid interface are also having a very different experience than the ones out in the middle are. That goes for the solvent-glass interface as well as the solvent-air one. There’s been a lot of study directed to these effects in recent years, but here’s a particularly dramatic example in a recent Nature Chemistry paper. The authors are looking at the photochemical reaction of phenol, which is known to ionize under ultraviolet light in water into a phenoxy radical, a solvated electron, and a proton. That reaction has been studied in detail in the bulk phase, but what about the surface layer? The paper uses a technique whose full name I am only going to type out once: ultraviolet-excited time-resolved heterodyne-detected vibrational sum frequency generation (UV-TR-HD-VSFG) spectroscopy. This allows spatial resolution a few molecule layers thick at the surface and femtosecond time resolution of changes in the OH stretching spectrum – just what you need to answer this sort of question.
The authors pick up transient signals that can (quite reasonably) be assigned to the hydrated electrons, protonated water molecules, and the phenoxy radical, and these things come on very, very quickly in the surface layers as opposed to the bulk phase: ten thousand times faster. In physical-organic-chemistry-geek terms, the potential energy surface of the phenol molecule must change pretty drastically, along with the gap between various excited states. If you use shorter UV wavelengths, you can get fast phenol photoionization with no problem, but these experiments were done with far less energetic 267nM light. But even that is apparently more than enough to set things off for phenol molecules up near the interface. The unique arrangement of the surface water molecules around a given phenol seems to stabilize the excited state much more than the bulk phase does.
There’s no reason that this sort of thing can’t apply to many other reactions – some of them will be far slower at the interface, and some far faster. Different products might be produced or inhibited as well, and all of this could (and should) vary depending on the solvents used, ionic strength, and other factors. And don’t think that this is some weirdo physical chemistry effect with no relevance to the real world, because the active sites of proteins are, in fact, another interfacial zone with bulk solvent. Enzymes indeed depend on arranging individual water molecules and amino acid side chains in ways that affect potential energy surfaces of their substrates and lower the energy gaps between their starting states and the transition states of the desired reactions.
What’s happening up at the water surface with phenol is a “wild-type” version of what goes on in an enzyme. If we understood these things better, we could design all sorts of odd nanoscale cavity reactors that would do unusual and selective things to molecules, but we’re not there yet. But we’re walking around, breathing, thinking, and writing and reading blog posts thanks to an untold number of those little nanoscale cavity reactors floating around in our cells. We’ll figure it out, eventually. . .