This is one of those papers that’s going to seem to non-chemists (and perhaps to some chemists as well!) like a how-many-angels-can-dance-on-the-head-of-a-pin stuff. But it’s the details like this that we have to understand in order to figure out what’s really going on in our reactions.

Most everyone (chemist or not) has heard of hydrochloric acid. We tend to be a bit loose with the nomenclature, but the term is properly applied to solutions of hydrogen chloride (HCl) in water. The saturated solution is about 37% HCl, and many a blue-capped bottle of it have I lifted in the lab over the years. There’s one jug of it that I wish I’d bothered to lift off the floor in grad school, but no, I had to leave it where I could kick it over as I turned around from the sink, leaving me to quickly broad-jump over the expanding lake of acid in an effort to get over to a big container of sodium carbonate.

But I digress. Straight hydrogen chloride itself is of course a gas, and you can buy it in various-sized cylinders. These can at times be a bit too exciting to work with (see “Ping! Ping! Ping!”), not least because the valves and regulators on the top of them tend to get stuck. Those two links are both grad school stories as well, one of which does me no credit and the other of which does none for a guy in the lab next door. A third one, which at least was no one’s fault in particular, involved a full-sized cylinder of HCl whose regulator was in much worse shape than I realized. I opened up the main valve on that one, making sure that the regulator itself was closed down, and noted with respect the 1200 psi full-tank reading on the first gauge. Then without my touching anything, suddenly the dial on second gauge pinged hard over to its stop, and my hair practically stood on end as I realized that the flippin’ thing was broken somehow and that little adjustable valve at the end was quite possibly now getting the tank-neck pressure too and was all that was standing between my person and a massive room-filling cloud of corrosive poisonous gas. I shut off the main valve with alacrity and declared a lunch break.

But I digress again. You can dissolve HCl gas in a lot of different solvents, and some of these are shelf-stable enough to be commercial products (HCl in ether and HCl in dioxane are two of the most common). But what is HCl in ether, when you get down to details? I always had a hazy mental picture of a bunch of 1:1 pairs of HCl molecules with ether molecules, vaguely hydrogen-bonded, with these floating around in a larger number of bulk ether molecules, but that’s about as far as I went.

The paper linked to in the first paragraph is looking at the question more systematically. They’re not the first to do so, but results have been inconclusive. The 1:1 hydrogen-bonded complex has some solution evidence for it in infrared and Raman spectra, but solid ether-HCl complexes can also show complete ether protonation, giving you a diethyloxonium chloride species. Things get even more complicated when you think about ion pairing: if you have that oxonium chloride species, are the anion and cation together, surrounded by solvent (an “inner-sphere” ion pair), or are the two of them solvated to some degree off by themselves (an “outer-sphere” pair)? As I led off this post with, such questions can seem weirdly or needlessly picky, but it turns out that all of these can have somewhat different properties and reactivities. If you really want to understand acid solutions (and acid catalysis), then you need to make these distinctions, and you need to figure out what’s actually happening in the real flasks. As the authors note in their references, studies of other protic acid solutions have yielded some pretty varied and unpredictable structures in the relatively few cases where people have put in the effort.

The IR evidence in this work, together with molecular dynamics simulations, cover protonated and deuterated species in several solvents. And it all points to the protonated ether structure as an inner-sphere contact ion pair. That goes for HCl-ether in bulk ether solution, as well as in toluene or dichloromethane (those latter two making far fewer appearances in most labs!) And this leads to a couple of actionable conclusions. The first is that HCl/ether solutions are probably less reactive than HCl in non-ethereal solutions, because the actual acid in the latter isn’t hydrogen chloride itself, but a protonated oxonium species. And second, it implies that you could generate a number of different such oxoniums under controlled conditions, which might well have usefully different reactivities in catalytic systems (which is clearly where this work is headed).

So I’m going to have to revise my thinking about these things as well. My line for years has been to rip off Tolstoy’s opening to Anna Karenina and say that protic acids are all alike, while every Lewis acid is an acid in its own way. That’s never been completely true, of course (not that it slowed me down much) because protic (Brønsted) acids do have different characters according to their counterions. But those distinctions apparently run even finer than I’d realized. Lewis acids, for their part, can vary obviously and crazily in their reactivity and utility for different reactions, to the point that sometimes you just have to try a whole list of them and pretend later on that you knew which one would do the job. But apparently they’re all like that!