Azides have featured several times in the Things I Won't Work With series, starting with simple little things like, say, fluorine azide and going up to all kinds of ridiculous, gibbering, nitrogen-stuffed detonation bait. But for simplicity, it's hard to beat a good old metal azide compound, although if you're foolhardy enough to actually beat one of them it'll simply blow you up.
There's a new paper in Angewandte Chemie that illustrates this point in great detail. It provides the world with the preparation of all kinds of mercury azides, and any decent chemist will be wincing already. In general, the bigger and fluffier the metal counterions, the worse off you are with the explosive salts (perchlorates, fulminates, and the others in the sweaty-eyebrows category). Lithium perchlorate, for example, is no particular problem. Sodium azide can be scooped out with a spatula. Something like copper perchlorate, though, would be cause for grave concern, and a phrase like "mercury azide" is the last thing you want to hear, and it just might be the last thing you do.
As fate would have it, though, none of this chemistry is simple. You can get several crystalline forms of mercuric azide, for one thing. The paper tells you how to make small crystals of the alpha form, which is not too bad, as long as you keep it moist and in the dark, and never, ever, do anything with it. You can make larger crystals, too, by a different procedure, but heed the authors when they say: "This procedure is only recommended on a small scale, since crystalline α-Hg(N3)2 is very sensitive to impact and friction even if it is wet. Heavy detonations occur frequently if crystalline α-Hg(N3)2 is handled in dry state".
Ah, but now we come to the beta form. This, by contrast, is the unstable kind of mercury azide, as opposed to that spackle we were just discussing. These crystals are not as laid-back, and tend to blow up even if they're handled wet. Or even if they're not handled at all. Here, see if you've ever seen an experimental procedure quite like this one:
After a few minutes, the deposition of needle-like crystals starts at the interface between the nitrate and the azide layer (β-Hg(N3)2). After some time,
larger crystals tend to sink down, during this period explosions frequently occur which leads to a mixing of the layers, resulting in the acceleration of crystal formation and the growth of a mat of fine needle-like crystals. . .
Hard to keep a good smooth liquid interface going when things keep blowing up in there, that's for sure. Explosions are definitely underappreciated as a mixing technique, but in this case, they are keeping you from forming any larger crystals, a development which the paper says, with feeling, "should be avoided by all means". But it's time to reveal something about this paper: all this mercury azide stuff is just the preparation of the starting material for the real synthesis. What the paper is really focused on is the azide salt of Millon's base [Hg2N+].
Now that is a crazy compound. Millon's base is a rather obscure species, unless you're really into mercury chemistry or really into blowing things up (and there's a substantial overlap between those two groups). A lot of the literature on it is rather old (it was discovered in the early 1800s), and is complicated by the fact that it usually comes along as part of a mixture of umpteen mercury species. But it really is a dimercury-nitrogen beast, and what it's been lacking all these years - apparently - is an azide counterion.
There are two crystalline forms of that one, too, and both preparations have their little idiosyncracies. Both forms, needless to say, are hideously sensitive to friction, shock, and so on - there's no relief there. For the beta form, you take some of that mercuric diazide and concentrated aqueous ammonia, and heat them in an autoclave at 180C for three weeks. No, I didn't just have some sort of fit at the keyboard; that's what it says in the paper. I have to say, putting that stuff in an autoclave has roughly the same priority, for me, as putting it under my armpits, but that's why I don't do this kind of chemistry.
But the alpha form of the Millon's azide, now that one takes some patience. Read this procedure and see what it does for you:
Nitridodimercury bromide [Hg2N]Br (0.396g, 0.8mmol) is suspended in a saturated aqueous solution of sodium azide NaN3 (dest. ca. 3mL) at ambient temperature, resulting in an orange suspension which was stirred for ten minutes. The solution is stored at ambient temperature without stirring under exclusion of light. After one week, the colourless supernatant was removed by decantation or centrifugation and the orange residue was again suspended in a saturated aqueous solution of sodium azide NaN3. This procedure was repeated for 200 to 300 days, while the completion of the reaction was periodically monitored by PXRD, IR and Raman spectroscopy. . .
So you're looking at eight months of this, handling the damn stuff every Monday morning. The authors describe this procedure as "slightly less hazardous" than the other one, and I guess you have to take what you can get in this area. But the procedure goes on to say, rather unexpectedly, that "longer reaction times lead to partial decomposition", so don't go thinking that you're going to get a higher yield on the one-year anniversary or anything. What way to spend the seasons! What might occur to a person, after months of azidomercurial grunt work . . .surely some alternate career would have been better? Farm hand at the wild animal ranch, maybe? Get up when the chickens would be getting up, if they'd made it. . .head out to the barn and slop the wolverines. . .hmm, forsythia's starting to bloom, time to neuter the hyenas soon. . .
No, no such luck. The hyenas will have to remain unspayed, because it's time to add fresh azide to the horrible mercury prep. Only three more months to go! Sheesh.