This article was published in Scientific American’s former blog network and reflects the views of the author, not necessarily those of Scientific American
The Raspberry Pi computer sat innocently in the glove box. This particular glove box was military-grade, enclosed on all sides except one by an inch of reinforced steel, with a narrow porthole made from Pyrex for viewing and manipulation. A robotic arm allowed you to punch the keys. We gratifyingly thought of the $35 units that we had purchased for this project; there’s only so many octa-core Dell Precision Towers that you can blow up every day.
I winced as Alex gingerly started adding yet another nitrogen atom to the ring. It was conventional wisdom, known for ages and duplicated in laboratories around the world. Most explosives including TNT and RDX contained a generous dose of nitrogen atoms; add enough nitrogens to a molecule – preferably a ring – in certain strategic positions and you would almost certainly make a big bang. The bang came from a fundamental property of nitrogen, its tendency to cling to its own kind and eschew others with a fanatical tenacity. It was just basic chemistry, except that it had been put to good use in the service of war and killing for decades.
But nobody had pushed the principle to its limits. Not like this.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
The idea came from a recent publication from Prof. Klapötke’s group in Munich. They had synthesized azidoazide azide. That tongue twister gave away the identity of the beast; “azide” was chemical lingo for a group of three nitrogens strung together in a line. Azides are notoriously explosive; lead azide for instance consists of a single lead atom decorated with six nitrogens (two azides), waiting to blow anyone who approaches them to kingdom come. Azidoazide azide sported no less than eight nitrogens tightly knit around a ring. The resulting molecular entity was so unstable that it literally disintegrated the moment it was born. As veteran chemist and blogger Derek Lowe described it,
“The compound exploded in solution, it exploded on any attempts to touch or move the solid, and (most interestingly) it exploded when they were trying to get an infrared spectrum of it. The papers mention several detonations inside the Raman spectrometer as soon as the laser source was turned on”.
Basically the thing exploded no matter what you did or didn't do to it. The infrared spectrum was a harmless thing, a common experimental technique only supposed to aid in deciphering the structure of a molecule based on the vibrations of bonds between specific atoms. But the mercurial fiend was so remarkably unstable that it was a miracle it could yield itself to being described in a respectable journal, let alone be subjected to the indignities of infrared radiation.
It was after reading the paper that Alex had a brainwave. Azidoazide azide clearly blew up as soon as it was made. What if we designed an explosive that actually blew up before it was made? Preferably the moment it was drawn on a computer and optimized into a realistic structure with normal bond lengths and angles. We could call it a pre-explosive. You could always run the risk that such a compound blew up when it was doodled on a paper napkin by that workaholic chemist even as his kids were playing scrabble in the living room, but everyone knew that decades of graduate school training had done nothing to obliterate chemists' bad molecular drawings with nonsensical bond lengths and angles. No risk there.
It would be the perfect weapon. Ideally we would want tantalizing features of the molecule – perhaps an infrared spectrum, maybe a melting point, even a few steps of the enabling reaction – to somehow fall into the hands of the enemy. This could be accomplished using a double agent or a spy who willingly allowed herself to be captured with key documents. Once the enemy located the details of the characterization, they would no doubt think that they now possessed the recipe for a key weapon of strategic importance; perhaps a new chemical or debilitating agent, a revolutionary material for armor or a life-saving battlefield drug. All resources would be focused on figuring out the molecular structure of this potential bonanza by working backwards from its properties.
Now all that we would have to do is wait for them to try to reverse-engineer the molecule. Of course, everything’s that’s reverse-engineered these days has to first pass through a computer model. There is no better way to unearth a plausible structural gem from the dross of incomplete data than by using one of those new neural network-enhanced quantum genetic algorithms. The number of possible structures corresponding to such sparse data is astronomical and only a computer can cycle through these endless wannabes. But thanks to D-Wave’s pioneering work, computing power has progressed to such an extent now that commercial software can search roughly ten trillion molecular possibilities in a matter of hours. Once the enemy lets its computing power loose on our data, their computer transforms itself into a literal time bomb even as it cycles through the list of possible structures. Since the algorithm is random, the correct structure may come up within a few seconds or it may be the last one on the list. But what’s certain is that at some point it will appear, and the rest will literally be history. History scattered around in the form of discrete particulate matter.
We waited with bated breath the first time we did it, allowing the molecular structure to relax and optimize its energy. We don’t know exactly what happened as the convergence cycles came to a standstill and the initially deformed bonds started to look normally formed. They found both of us stretched out on the floor. Fortunately this first attempt had only resulted in a design with a modest PEDI (Pre-Explosive Detonation Impact) factor of 5.2, hardly sufficient to cause maiming or death; theory predicts that we would need a PEDI of at least 40 to cause damage equivalent to that caused by the most powerful non-nuclear explosive. But after the mishap the computers were duly installed in a robot-controlled glove box with reinforced walls. The plan was to ramp up the PEDI in a respectable, controlled manner.
It’s a particularly nice day outside as Alex is about to fire up a calculation on a potential molecular candidate. “Remind me what you are doing, again. I have to admit I was too groggy when you excitedly called me in the middle of the night yesterday”. “Well, it was one of those obscure new open-access journals they keep emailing us about. Usually I delete the emails the moment I see them, but this one had something in the title about azides so I took a look. There was a paper from some group in Latvia, from a university I have never heard of. Conforming to the shoddy standards in these journals, there was apparently a lot of characterization but few specific structures except for two general scaffolds, both similar to the work done at Toulouse on low-yield azides in the 80s; I guess Raman spectrometers are cheap and they were obsessed with patent filings. Thought I would reverse-engineer one of them just to see what it looks like.”
The computer displays elegant ball-and-stick structures in front of us as I absent-mindedly listen to Alex and tap my fingers on the side of the glove box. Alex’s quip about open-access journals makes me think of that article in Nature published a few months back which talked about the nuisance created by the proliferation of spurious open-access journals. Many of these use fake fronts, email addresses and entire made-up office locations to convince readers of their authenticity. Most ask you to foot the publishing fees and then disappear. I bet you would find no trace of their whereabouts if you actually decided to look for them. Of course you probably deserved it if you were credulous enough to fall for an unknown publication from a made-up source.
Suddenly something snaps inside me. I get up, startled, and look at Alex, even as he watches the computer flash a structure with a particularly beautiful geometric arrangement of atoms. Oxygens are flaming red, nitrogens are a tranquil blue.
Six thousand miles away in Kiev, a man sits sipping coffee in a café near the old town square. His cell phone rings and a gruff voice communicates the message. The man hangs up and mutters to himself, “45”. His face breaks into a faint, satisfied smile. He goes back to sipping his coffee.
This post was inspired by a spoof article by Isaac Asimov. In the 1940s Asimov was working on a rather thankless Ph.D. at Columbia University. Part of his work involved investigating the properties of compounds which were highly soluble in water. Some of these chemicals were so highly soluble that they seemed to dissolve almost instantly. This behavior encouraged Asimov to pen a spoof article titled "The Endochronic Properties of Resublimated Thiotimoline" about a compound that actually dissolves before it hits the water. Recent articles on new explosives that seem to literally detonate as soon as they are formed lead me to similar thinking...