A good short history of Tabun and other nerve agents, largely based on this book, can be found here. To summarize, in 1937 a report on Tabun made its way to the chemical warfare branch of the German military, and its value was recognized quickly. Gerhard Schrader's group was moved to new laboratory space and set to developing new agents in the same chemical class. Money and material was put on scaling up the synthesis of Tabun itself.
Now, that was no easy project to be assigned to. The biggest problem, of course, was the hideous toxicity of the product. The pilot plant had quite a containment system (double-walled enclosures with positive pressure and so on, very sophisticated for its time.) The workers wore rubber containment suits, which were rigorously cleaned and changed. People still got killed. The histories above give some examples - the one that sticks with me is the unfortunate who had two liters of Tabun suddenly pour down inside his rubber suit.
Even without the awful product, the chemistry by itself was pretty foul. To pick a major issue, it involved hot hydrofluoric acid. You really don't want HF around if you can help it. It attacks glass, for one thing, and goes after a number of metals. It leaves the really expensive ones untouched, though - if you read the old literature on the stuff, you find references to exotica like platinum dropping funnels and the like. The Germans had to use silver-lined reaction vessels in part of the plant; the sort of thing we'd use high-nickel alloys for today. On top of the corrosion problem, HF is also very toxic, and inflicts extremely dangerous time-delay burns. The idea of working in a process plant where hydrogen fluoride isn't the nastiest thing in the house gives me the shivers.
And just to top things off, the Tabun process also involves cyanide and produces HCN vapors, which have to be dealt with somehow. I spent a paragraph the other day talking about how HCN wasn't a very useful war gas (which it isn't,) but being cooped up in a factory with it is another matter entirely. All in all, this is a one-damn-thing-after-another process that no one would scale up under normal circumstances.
The route worked, though, although it took a couple of years to get it going reliably enough to where it wouldn't kill everyone in the vicinity. By the end of the war, Germany had produced 12,000 tons of Tabun, a figure to give anyone pause. (We'll talk about why they never used it in the next post.) Meanwhile, Schrader's group continued to work in the area, producing Sarin (or GB) in 1938 and Soman (GD) in 1944. While Tabun has largely disappeared (except for nations just getting into the nerve gas synthesis business,) Soman and Sarin are still very much with us. The chemical routes were just as nasty, though - they still hadn't found a way around the fluorine reagents, and Sarin has a fluorine group directly bonded to the phosphorus. At least the cyanide part was gone. Still, production of Sarin by the end of the war was merely a few tons, and Tabun still has the reputation, despite all the nasty steps, of being the easiest nerve agent to synthesize in bulk.
In the 1950s, post-war research led to the discovery of several other effective compounds, including VX, which still sets the standard. These were discovered in several countries, more or less simultaneously. VX has a structure reminiscent of the others, but where the fluorines (or cyano) groups are on the phosphorus, there's an aminoalkane linked through a sulfur atom. The Soviets stockpiled their own, very similar compound with almost identical properties. VX's structure wasn't publicly disclosed until the early 1970s, but as it turns out, a 1960s German patent had made its way into the various open databases which (to everyone's embarassment and surprise) had VX in it. The inventor was. . .Gerhard Schrader, still at the phosphorus chemistry after thirty years, and obviously a man with very good lab technique to have survived that long.
Improvements were made over the years to the syntheses of all these compounds, but I'm not going to go into the details. It's an interesting set of process chemistry problems, but you have to be conversant in that sort of thing to get the most out of the discussion, and I assume that a majority of my readers aren't. There are many discussions in the open literature. For example, see this large PDF file entitled "Technical Aspects of Chemical Weapon Proliferation," which has a vast amount of detail.
No matter what the route, anyone outside of a serious industrial setting who wants to try this chemistry runs a mortal risk. I've worked with some really toxic stuff in my years in the lab, but I wouldn't touch any of the nerve agents with a platinum pole. 10 milligrams of VX on the skin is the approximate lethal dose - is my lab technique really that good? Well, I think so. . .but is that the phrase that I want them to put on my tombstone?
No one knows if there's been much subsequent R&D in this area, but I doubt it. Chemical weapons have the reputation of being living fossils compared to how other weaponry has developed. A Russian report from the early 1990s spoke of a completely new chemical class of lethal agents, which is certainly possible - but why bother? The phosphorus-based ones are about as bad as it's possible to get. (An overview of their comparative properties with some similar historical background is here, and the link I provided yesterday goes into great detail, too.)
Of the agents that have been used in the real world, Sarin's the most volatile, and does a lot of its work by inhalation (it's likely that if Schrader's lab had made that one first that they all would have died.) Meanwhile, VX is more persistant, rather like mustard gas, and works mainly by contact. Both can be formulated so that the last chemical step in the syntheses takes place in the shell or bomb, after it's been fired or dropped. To the best of my knowledge, there has been no hostile use of these "binary" weapons, except for an alarmingly crude dump-and-seal technique tried by the Iraqis during the war with Iran.
So how does one deal with these things? In the way of organic chemistry, the same reactivity that allows these compounds to inactivate cholinesterase also gives you a handle to break them down. All these reactive groups attached to phosphorus can be hydrolyzed by things like sodium hydroxide or bleach, or reacted with a nucleophile like ammonia. The resulting phosphoric acids or phosphoramides are basically harmelss. (VX's hydrolysis product, though, is unusually toxic. Fortunately, it doesn't penetrate the skin.)
That's all very well for decontaminating a concrete wall, but what about decontaminating yourself? The fast action of the nerve agents makes speed the main consideration. Even potentially lethal exposures can be compensated for if treated quickly enough. There are two therapeutic approaches, which are generally used simultaneously: Oxime compounds can actually react with the cholinesterase-bound form of the nerve agent, knocking it out of the active site in the process and regenerating the enzyme. Meanwhile, out in the synapse, a cholinergic antagonist can block the receptors and keep the signaling at the neurons from getting out of hand.
One antagonist that's usually provided for this purpose is atropine, which under normal conditions is quite poisonous itself (since blocking acetylcholine signaling for no reason is arguably just as bad as overloading it.) Really heroic doses of the stuff can be used in nerve gas poisoning cases, though. An important part of any chemical-warfare supply kit is a supply of these antidotes, ready to inject. There's a picture of a standard apparatus in another of the links from yesterday's posting .
Finally, giving a reversible inhibitor of acetylcholinesterase can protect against nerve gas before any exposure. That seems rather odd at first, but the idea is that the inhibitor ties up a certain proportion of your enzyme (but not enough to cause trouble.) The reversible chemical equilibrium causes it to gradually be freed up, even after nerve gas exposure, and this gives you a reserve of active enzyme coming on line that hasn't been hit by the irreversible nerve agent. Used properly, this can be enough to prevent much of the damage. Here's a military manual on chemical agents that goes into more detail on treatments for exposure to all of them.
In the next post, which I hope will be tomorrow, I'll try to wrap things up with a strategic discussion - no, I'm not turning into Den Beste (he fills that role just fine!) - but now that I've talked about the history and properties of these things, it's time to see what's been done with them, and what might still be waiting.