In the old days before the first Gulf War, most nuclear export controls were pretty cut-and-dried: Exporters checked for items on a short ”trigger list” that could be used for making nuclear bomb fuel. But after they learned what Iraq had secretly been up to for about a decade before 1991, the Nuclear Suppliers Group came up with a second list, Infcirc/254/Part 2, and it got into the tricky business of policing a panoply of dual-use goods sought by proliferators.

The stream of dual-use goods is virtually endless, and so has been the internal debate at the NSG about whether items should be listed or not. That’s even more so right now, because as I explained in this report published by Carnegie last month, the NSG has launched a comprehensive review of both its lists, and the decisions it makes about what to control and how to do it will profoundly impact how we attack procurement threats in years to come.
One of the items which is not on the NSG dual-use list is a chemical substance which Iran has been keenly and furtively trying to import since the mid-1990s. This is anhydrous hydrogen fluoride, or AHF. (The term “anhydrous” is just a fancy way of saying that hydrogen fluoride [HF], otherwise known as hydrofluoric acid, is more or less pure, having less than 400 ppm of water.)
Iran needs AHF to process UO2 into UF4 for two programs that the UN Security Council says should be suspended: uranium enrichment and the Arak heavy water reactor.
The open literature on Iran’s nuclear program contains a few misleading references asserting that back in 1985, when the U.S. and China were negotiating a bilateral civilian nuclear cooperation agreement, the U.S. believed that efforts by Iran to import AHF were a tell-tale signature of Iranian interest in reprocessing, and that China therefore agreed not to supply AHF to Iran because the NSG controlled exports of AHF.
In fact, some papers in my files, authored in 2003 by scientists from the Atomic Energy Organization of Iran, spelled out for anyone who cared to know that AEOI wanted AHF for two reasons: First, to produce F2 gas by an electrolysis process, and then use the F2 to fluorinate UF4 and generate UF6; and second, to use HF to hydrofluorinate UO2 to produce UF4 (also known as green salt), which would be subsequently reduced to uranium metal in a reduction vessel. These disclosures match findings in John Garver’s recent and comprehensive tome on Sino-Iran relations, namely, that China before that had “supplied Iran with blueprints, test reports, and design information” for a UF6 production plant and a second plant to produce uranium metal. In 2005, the AEOI reported in fact that the uranium conversion plant it built at Esfahan using Chinese technology features both chemical processes. UF4 can be processed into UF6 in Building 101C, or instead into uranium metal in Building 104.
Garver points out that when the U.S. in 1997 objected to China about its plans to sell AHF to Iran, the Chinese riposted that the transaction was perfectly legal because AHF wasn’t on the NSG dual-use list. That might have been just a tad disingenuous since China at that time was not a member of the NSG anyway. But the Chinese statement was correct.
It’s also correct today. AHF is still is not on the NSG list, regardless of its critical role in UF6 production in some enrichment programs, including Iran’s. If Iran wants to use its Chinese technology to provide the F4 in UF4 or the F6 in UF6 for enrichment feedstock or uranium metal fuel, it needs the AHF.


Iranian procurement agents were not deterred by assurances which China provided to the U.S. in 1998 that planned sales of AHF to Iran would be indefinitely “suspended.” Customs intelligence files of at least three Western governments testify to repeated efforts by Iran–some successful, some not–to import Chinese-origin AHF for a period of nearly 15 years. In the beginning, Iran tried to import the AHF directly from Chinese producers. As time went on, Iran added transit destinations to hoodwink export controllers.
In December 2004, while Iran was negotiating with the EU-3 over a deal to suspend its enrichment program, Iran announced that it had approved construction of its own AHF production plant, also at Esfahan–a point that stuck in the craw of EU negotiators who saw that step as a demonstration of bad faith by Iran. Until that plant was up and running, however, Iran would need to import AHF and throughout the last decade they kept trying to get their fingers on it.
Iran’s AHF production plant, with a capacity of 5,000 tons of AHF/yr, far more than what is needed to produce UF6 at current rates in Esfahan, was scheduled to be finished and working in just two years. Some intelligence information however suggested that Iran encountered delays. For whatever reason, Iran has continued to try to snag large amounts of AHF, and to disguise its efforts by organizing complex transactions involving multiple transit points. In 2008, for example, in Thailand, the Islamic Republic of Iran Shipping Lines loaded a large consignment of Chinese AHF onto a Liberian-flagged vessel bound for the Iranian port of Bandar Abbas on the Persian Gulf, with an intermediate stopover in the UAE where it would again change hands.  Before that, Iranian procurement agents had arranged for the AHF to be purchased from a Chinese producer by a Korean firm, and then forwarded to Thailand. Western customs intelligence agencies got wind of the transaction, and the AHF was intercepted in the nick of time at the Port of Dubai before it headed to Iran.

NSG List vs Catch-All Approach

So why, if AHF is a potential bottleneck in Iran’s designs to produce UF6 for its centrifuges, and Iran is going to great lengths to deceive us about its efforts to import the stuff, isn’t AHF on the NSG dual-use list?
AHF is just one of many items which could be considered nuclear dual-use and which are not routinely controlled. The way the NSG works, all 46 members must agree to put an item on a control list. In some cases, business pressure groups in key NSG states object to their governments and an item may be left off the list. (That’s why my report last month urged that the NSG closely monitor efforts by these lobbies to interfere with consensus-formation during the control list review).
There are a lot of companies making AHF, in a lot of countries worldwide. Producing it doesn’t require much if any proprietary know-how. And its nuclear use accounts for just a fraction of demand. These data from NNSA suggest that any efforts to put AHF on the NSG list would kick up a fuss, because uranium fuel processing accounts for just 3% of global use of HF. A major consumer is the petroleum industry–which would not be amused by a new requirement that AHF be subject to nuclear-use controls whenever it was sought for use in a gasoline refinery anywhere in the Middle East.
Furthermore, there’s the China factor to consider. By its own admission, the biggest producer of HF in the world is a U.S. firm, Honeywell. But according to industry market intelligence for this sector, there is now a shift underway in HF production away from advanced industrial countries. One reason for this globalization may be that making this stuff is a nasty health hazard. AHF is highly toxic and corrosive. If you breathe it in, deep in your lungs the gas precipitates to hydrofluoric acid. China presently makes HF cheaply and with comparatively little oversight and has half of the world’s fluorite reserves used to make HF. According to one market survey, “the growing tight supply of fluorite resources forced the tycoons in [the] fluorine chemical industry… to transfer the production of HF to China, and establish their plants there successfully.” (In one case, this process may have been just a little too successful: a German AHF producer last year sued its Chinese partners for secretly building 12 unauthorized AHF plants in China using the German firm’s technology.) Would China consent to tightening controls on AHF?
AHF is on the control list of the Australia Group because it is also identified as a precursor for nerve agents containing fluorine. My NSG report recommended that the NSG intensify contacts with the Australia Group and other multilateral export control arrangements in the interest of efficiency. Perhaps the CWC people have a best practice or two that would help the NSG keep AHF out of the wrong hands.
Just because AHF is not on the NSG control list, that doesn’t mean no one at the NSG is paying attention to it. Given repeated attempts by Iran to import it, AHF is on a so-called “watch list” of unlisted items which the U.S. government has provided NSG members to trigger catch-all controls for shipments to Iran.
The “watch list” is there to flag NSG members to halt exports of non-listed goods to suspicious end-users. But if Chinese-origin AHF is headed for Iran via non-NSG intermediate destinations like Thailand, it won’t be stopped if the intermediary entrepots have no catch-all controls on their books. What’s more, catch-all controls won’t be effective unless they are applied by all NSG states themselves when they are needed. South Korea is a member of the NSG; that 2008 transaction involved a South Korean AHF vendor.
In years ahead, the NSG will be challenged by a massive global trade increase in unlisted goods. That means that effective application of catch-all controls will be absolutely critical to halt proliferators–something for the NSG to think hard about as it reviews its control lists over the next three years.
This article originally appeared in Arms Control Wonk.