A news article about the UN’s Clean Development Mechanism caught my eye this week because it reminded me of a similar problem that arose at the OPCW a few years back.
Perhaps the single most challenging aspect of keeping a disarmament and non-proliferation agreement like the Chemical Weapons Convention on the rails is ensuring that it achieves its objectives with the minimum possible impact on legitimate industry. Squaring that circle is extraordinarily complex and requires a comprehensive, in-depth knowledge and appreciation of how the global chemical industry works, and how its myriad components fit together. This sort of balancing act is not possible in every area of endeavour; the problems associated with deconflicting treaty requirements and the needs of the domestic and international biotech and pharmaceutical industries, for example, were the key factor in the failure of the 2001 BTWC Review Conference. In the chemical world, juggling competing interests often means getting down into the nitty-gritty of how chemicals are actually produced.
Dealing with Schedule 1 chemicals - the nerve gases, the blister agents, and a few others - is fairly straightforward, because all of these are intended to be used as warfare agents and have no commercial utility (Note 1). The problem that arose at the OPCW circulated around two concepts: “low concentration limits”, and “captive use”. Simply put, some processes for synthesizing commercial chemicals - some of which are produced in vast quantities - actually result in the production of extremely toxic compounds either “in the pot”, or as a byproduct of the reaction process. The classic “in the pot” example is the chemotherapy agent Mustargen (methchloramine hydrochloride). This drug is also called HN2-hydrochloride because it is the hydrochloride salt of a warfare agent known as HN-2 (Bis(2-chloroethyl)methylamine), one of the three “nitrogen mustards” listed in part 6 of Schedule 1. The Germans and British both produced this agent during the Second World War, but it was never used in combat. Producers stopped stockpiling the nitrogen mustards when it was discovered that they tend to be unstable in storage, off-gassing hydrogen, which has a lamentable tendancy to explode. Mustargen (the drug) is synthesized by chlorinating the precursor alcohol, methydiethanolamine, producing HN2 (a viscous liquid). Instead of decanting and distilling the CW agent, the solution is basified using sodium or potassium hydroxide, producing the hydrochloride salt of HN2, a solid that is precipitated out of the reaction mass, then stored and administered to chemotherapy patients by saline injection.
The production of Mustargen represents a case of “captive use” of a Schedule 1 chemical; the prohibited chemical is both produced and consumed in the course of a linear synthetic process. What’s interesting about this case is that the complaint was originally raised by a group of German academics working for the German National Authority for the CWC. First (unusually for Germans, who have traditionally been the class of the world when it comes to inventing new toxic chemicals), the academics got the reaction process wrong in their submission to the OPCW, inverting the chlorination and basification steps. Their technical errors were refuted when certain other States Parties helpfully produced the reports obtained when Allied intelligence agents interrogated German chemists captured during the Second World War - including their hand-drawn diagrams of how the German HN2 production equipment was set up, and how the reaction process worked. Another resounding victory for good, solid historical record-keeping.
Another interesting example of “captive use” is the case of BZ. BZ is a code name for 3-Quincuclidinyl benzilate, a chemical hallucinogen similar in structure and effect to LSD (lysergic acid diethylamide, or “acid”). The US produced and trialled BZ during the 1960s, roughly the same time Timothy Leary and his acolytes were doing the same with LSD, and the Beatles were rhapsodizing about “Lucy in the Sky with Diamonds”. Although not as powerful as LSD, BZ was a lot cheaper to make in quantity, and bombs and submunitions were filled with the stuff. The agent was taken out of service shortly thereafter, however, when it was discovered that while some personnel exposed to it became calm and detached, others descended into a “berserker-like rage”. Clearly this was not the sort of thing you’d want to be spraying over enemy troops just before engaging them in battle.
What’s interesting about BZ is the fact that, as seen with HN2 in the process for producing Mustargen, BZ shows up “in the pot” during the production process for the drug Clidinium Bromide. This is an anticholinergic / antispasmodic agent used to treat irritable bowel syndrome, and it results from the same sort of process that turns HN2 into Mustargen - the basification of a toxic chemical (in this case, BZ) to produce its hydrochloride salt for pharmaceutical use.
Given these sorts of synthesis issues, three proliferation-related questions arise. First, is the CW agent ever present in an isolatable form in the reaction stream? Second, is it present in a concentration that is sufficiently high to enable it to be produced in militarily relevant concentrations and quantities (or even to be used “as-is”)? And third, is the production equipment designed to allow the agent to be tapped off, i.e. diverted from the production stream, before the final basification step, either for further purification and storage, or for filling into munitions? These are all questions that can be asked and answered during Article VI inspections, and they highlight the importance of such inspections as the Convention’s key counter-proliferation mechanism.
Where the question of “concentration” becomes a little more problematic is when there IS no final step, i.e. when a militarily-relevant chemical appears at the end-stage of a synthesis process, instead of in the middle of it. For an example we must move from the pharmaceutical to the fluoropolymer industry. Teflon is possibly the most important fluorine-based chemical produced in the world today. It’s a trade name; the chemical itself is polytetrafluoroethylene, usually abbreviated as PTFE. As a heavy, solid, thermally stable and chemically inert compound, Teflon is ideal for non-stick coatings. Teflon production is complex, but all production methods have one thing in common: they produce, as a byproduct of the process, a compound known as perfluoroisobutylene, or PFIB. And PFIB, unlike Teflon, is a liquid with a low vapour point, and an inhalational toxicity about ten times that of Phosgene, the CW agent that caused most of the chemical-related deaths in the First World War. For this reason, PFIB is on Schedule 2, right next to BZ.
Fluoropolymer plants deal with PFIB in different ways. In large-scale, linear-production setups, the waste products are isolated and incinerated, with scrubbing to remove harmful compounds from exhaust gases. One very problematic plant in the UK, however, actually bottles the PFIB produced in Teflon manufacture, then trucks it many miles overland - in nice, small, easily-moved and therefore easily-stolen pressure cylinders - to an incineration facility. Is this plant producing a CW agent? Well, no, because the definition of “CW” in the CWC is intent-based, and nobody suspects England of making Teflon just to get PFIB for use as a CW agent when the UK is technologically capable of making much more toxic chemicals, like the G- or V-agents, or more advanced CW. But in an international treaty, this is a problem, because anything that’s legal for Britain is also legal for Iran and other ‘mercurial’ States Parties.
But is that Teflon plant actually a potential CW production facility? Well…it could be. The nature of chemical synthesis is that output responds to input and synthesis conditions. It’s possible to ‘tweak the pot’ - i.e., to modify the synthesis conditions to produce, for example, a much higher proportion of PFIB in the waste stream. 0.1% PFIB in solution is not a problem, nor is even 1%. But 10% PFIB…that’s concentrated enough to cause severe respiratory harm. In other words, simply by changing the temperature, feedstock, catalysts and other conditions, it would be relatively simple to produce much larger amounts of chemicals very different from what the plant is supposed to be producing. For example, the process could be tweaked to produce a lot more PFIB, and a lot less PTFE. It would make the plant much less efficient at making Teflon - but that wouldn’t matter if Teflon wasn’t the product you were really interested in producing.
This is why the OPCW has industry inspections. And - to finally get to the point of this message - it’s why this week’s news article about the UN’s Clean Development Mechanism (UN CDM) caught my eye.
One of the key precursors in the production of polytetrafluoroethylene is a chemical called trifluoromethane – a refrigerant commonly known as HFC-23. According to an AP report, the UN CDM is considering closing a loophole in what looks like a pretty poorly-thought-out part of the international “carbon offsets” trading mechanism. Under the CDM program, developed countries, rather than cutting their own carbon emissions, have been able to buy “carbon credits”, profits from which are used to pay developing countries to cut their emissions instead. There are a great many philosophical and sociological concerns with this sort of scheme, but it now looks as if there’s also a very practical one as well. It seems that five chemical plants in China have found an innovative means of making money off of the programme. The plants in question produce R-22 (chlorodifluoromethane), an “ozone friendly” refrigerant that is being phased out under the Montreal Protocol. However, as a byproduct of the production process, the plants also produce the abovementioned HFC-23 (Fluoroform or trifluoromethane), a chemical that is said to be 11,700 times as potent a “greenhouse gas” as carbon dioxide.
Because “carbon credit” prices for the elimination of harmful gases are scaled according to their degree of awfulness, the plants in question are being paid $100,000 for every tonne of HFC-23 they destroy. As a point of comparison, the commercial retail price for R-22 bought in bulk is about $6.00 per pound, or about $13,200 per metric tonne. Presumably the manufacturer gets less than this for the raw product.
The carbon credit program has, in effect, made the harmful byproduct seven times as valuable as the beneficial main product.
You can guess the rest. It seems that the plant managers have been ‘tweaking the pot’ to produce excess HFC-23, because - thanks to the carbon credit scheme - producing the undesirable byproduct of their production process and then destroying it immediately is earning them vastly greater amounts of money than producing the product the plant is actually designed to be making. According to the AP piece, “The evidence is overwhelming that manufacturers are creating excess HFC-23 simply to destroy it and earn carbon credits.”
When you’re trying to modify human behaviour, you have to look very closely at the incentive structures you’re creating. People are driven by self-interest. If you make it more lucrative for them to produce noxious waste than to produce useful product, chances are you’re going to get a lot less useful product, and a lot more noxious waste. We really need to keep a close eye on chemical production facilities, especially in countries where government oversight and regulatory controls are rather more lax than they are in North America.
There's a bit of a happy epilogue to the concentraiton problem. The OPCW member states solved the “low concentration limits” problem for Schedule 2A (Amiton and PFIB) and 2A* (BZ) chemicals last December at the 14th Session of the Conference of States Parties, declaring that declarations would not be required for plants producing Schedule 2 chemicals in solutions of 1% or less concentration, or for plants producing such chemicals in concentrations between 1% and 10%, provided that the total quantity produced is less than the relevant verification thresholds specified in Para 12 of Part VII of the Verification Annex (C-14/DEC.4 dated 2 December 2009).
And there was much rejoicing.
(1) Obligatory caveat: S1A(7) is Saxitoxin, a biological toxin produced by a marine dinoflagellate. Although the US attempted to weaponize it in the 1950s, it DOES have an important commercial use. Very small (i.e., milligram) quantities are used in diagnosing paralytic shellfish poisoning. The importance of being able to move small quantities of Saxitoxin around on short notice for medical purposes was the impetus behind Canada’s request, early in the Convention’s history, for a Technical Change to the Convention, inserting Paragraph 5bis into Part VI, Section B of the Verification Annex, in order to exempt quantities of Saxitoxin in amounts less than 5 milligrams from the requirement for 30-day advance notification of the impending transfer of a Schedule 1 chemical.