Foreign Policy Research Institute A Nation Must Think Before it Acts Iran, Radioisotopes, & the Joint Comprehensive Plan of Action

Iran, Radioisotopes, & the Joint Comprehensive Plan of Action

“Whether or not what you do has the effect you want, it will have three at least you never expected, and one of those usually unpleasant.”[1]

– Robert  Jordan


The 159-paqe Joint Comprehensive Plan of Action (JCPA) signed in Vienna on 14 July 2015 by the Islamic Republic of Iran and the E3/EU+3[2] is undergoing a thorough dissection in many quarters.[3] Some analysts defend the JCPA and claim it will retard Iran’s progress toward nuclear weapons, while others vigorously dispute that assertion. Likewise, defenders cite an intrusive inspection regime while opponents allege that regime is seriously deficient in aspects including on-demand access to suspect sites. Finally, defenders hail the “comprehensive” nature of the agreement while opponents note the JCPA is silent on many subjects, omitting, for example, any mention of several suspect nuclear facilities. Some of these are questions of fact.  Others are questions of perspective that will be endlessly gainsaid by each side, neither of which is better than the other at seeing into the future.  If in the realm of nuclear weapons, the alternative to what has been negotiated is a deteriorating status quo, then the JCPA’s merits in terms of containing Iran likely outweigh its deficiencies. It serves no good purpose to deny those deficiencies, however.

A thorough reading of the JCPA makes clear that its scope is not strictly limited to constraining Iran’s ability to develop nuclear weapons. One particularly important aspect has received little to no serious attention so far. It relates to Iran’s domestic production of radioisotopes.[4] The JCPA enables Iran markedly to expand radioisotope production, with the active assistance of at least some members of the E3/EU+3.  Radioisotopes have many legitimate commercial and medical uses.  There is nothing about their responsible production that is per se objectionable. An indelible direct line nonetheless connects radioisotope production to radiological weapons.

As noted in an earlier essay, the JCPA’s most abject deficiency is perhaps its least discussed—the consummate failure to address the question of radiological weapons. This essay considers two central considerations in the context of Iran’s production of radioisotopes. The first is the JCPA’s provision to reengineer Iran’s problematic Arak research reactor[5] for commercial and medical radioisotope production. The second is the E3/EU+3 commitment to assist Iran to build a modern radioisotope production infrastructure, including acquisition of state-of-the-art (and heretofore embargoed) particle accelerator technology.

The production and use of radioisotopes for commercial and medical applications is widely understood and practiced. The author has no issue with this. That being said, a caveat must be quickly added: many of these same radioisotopes are usable in improvised weapons of terror known as dirty bombs. They can also be used in military-grade radiological weapons. So the question arises: can an Iran that for good and just reasons is denied nuclear weapons be trusted with the means of producing another class of weapons—radiological weapons—the use of which has been called the “poor man’s nuclear warfare”[6]?

The author has written extensively on the subject of radiological warfare, dirty bombs, and the rising risk of radiological warfare in the Middle East, and refers readers seeking a detailed treatment of those subjects to those essays. This essay will quickly set the stage with an overview of the c.1950s deployment of radiological warheads by the Soviet Union. That is intended to clear away any question whether radiological weapons have a place in the arsenal of an ambitious, near-nuclear state. Moving on then, the balance of the essay will parse and discuss relevant provisions of the JCPA pertaining to radioisotope production. It concludes with some thoughts as to how Iran’s new opportunity to produce radiological weapons managed to escape the restrictions imposed on nuclear weapons production under the JCPA.

“A Film Come True”

“They were intending to show that it was not at all necessary to drop atomic bombs from airplanes.” – Boris Evseyevich Chertok

Lest anyone doubt radioisotopes can be weaponized, the Soviet Union managed to do it sixty years ago. The USSR developed and deployed radiological warheads in limited numbers at a time when it was still perfecting nuclear ones for its missile force. These warheads worked in much the same manner evaluated by Israel’s recent Greenhouse experiments, dispersing radioactive material in a liquid medium[7] into the environment by means of an explosive detonation.

The 1953 Russian film Serebristaya pyl (Silver Dust) depicts a demented American scientist who wants to test a new invention—”silver dust”—on human subjects. He plans to spread his “silver dust”—in reality, lethal radioactive powder—over a large area by detonating a special aerial bomb at high altitude, dispersing the powder to rain down on the unsuspecting people below.

The legendary Soviet rocket scientist Boris Chertok wrote in his memoir, Rockets and People:[8]

“It all started in the crowded conference room at our hotel at the test range, when they showed us the film… This was one of the first semi-fictional films dealing with the horrors of a future war…The radiation dose was lethal for everything living in the zone affected by the silver dust. No protective clothing or gas masks saved the population.”

“The film was created in consultation with specialists who had studied the effect of nuclear explosions. They were intending to show that it was not at all necessary to drop atomic bombs from airplanes…and after a certain amount of time, without having to fight, the victor could occupy the territory with all its resources preserved. There is an old adage about a ‘dream come true’.  This was a ‘film come true’.”

Chertok was part of the team that in the early 1950s developed two warheads for the Soviet R-2[9] short-range ballistic missile, which had a flight range of some 600 km (373 miles). The warheads—assigned the cryptonyms Geran[10] and Generator[11]—contained “military radioactive agents”—in reality, liquid radioactive waste—that, as in the film, was dispersed when the warhead detonated at altitude. The “military radioactive agents” payload was likely liquid radioactive waste from nearby plutonium production plants, which contained high levels of strontium-90 and cesium-137, both choice radionuclides for radiological weapons.[12]

The Geran contained a single vessel of the radioactive liquid that was released in an airburst to disperse it over a wide area.  Its sister warhead, the Generator, contained multiple small vessels that were released when the warhead detonated at altitude.  These vessels then fell to the ground where they burst open upon impact. Both warheads were retired from the Soviet arsenal in the late 1950s when a nuclear warhead was finally perfected for the R-2 and its successor, the R-5.[13]

Radioisotope Production Under the Joint Comprehensive Plan of Action

“Since the end of human action…can never be reliably predicted, the means used to achieve goals are more often than not of greater relevance to the future world than the intended goals.” – Hannah Arendt

Upon reading the JCPA, it is surprising how much of the document deals with the matter of radioisotope production in one fashion or another.[14] This seems little noticed so far. The author is aware of no analysis in which it merits more than peripheral mention, if indeed any at all. However, it comes up repeatedly in the JCPA’s discussion of first-line issues including the Arak research reactor, spent reactor fuel reprocessing, and civil nuclear cooperation.  The JCPA also establishes a framework for Iran’s development of a modern radioisotope production infrastructure over the next several years.  So it is puzzling that malevolent extensions of radioisotope production—weaponization and proliferation—are nowhere mentioned.[15]

The JCPA explicitly seeks to shift radioisotope production from one mode—nuclear research reactors—to another—particle accelerators.[16]  The reason is simple: to reduce or eliminate Iran’s use of highly enriched uranium[17] (HEU) targets and nuclear research reactors to produce radioisotopes.[18]  Research reactors have been integral to covert nuclear weapons programs in many countries: for example, India, Iraq, North Korea, Romania, and Yugoslavia all used them to produce plutonium for their domestic nuclear weapons programs. If un-, or under noticed, the JCPA’s effort to shift radioisotope production from research reactors to particle accelerators is a nonetheless foundational article of the agreement.

This is no small task: research reactors produce four-fifths of man-made radioisotopes worldwide. The remainder is produced by devices called particle accelerators—either circular ones called cyclotrons, or less commonly, straight or “linear” accelerators known as LINACS. The nuclear reactor method uses uranium-235 fission to produce neutrons in large numbers, called neutron flux.[19] When certain target materials are exposed to neutron flux inside a nuclear reactor, radioisotopes are formed. The process involves irradiating the target material, which is encapsulated in a specially designed container and loaded into a predetermined location in the reactor core or reflector. The target may be a stable material, which captures neutrons when the bombarding particles interact with the target material to form radioisotopes. Alternately, it can be uranium-235 that neutron flux causes to fission and form short- and long-lived fission products, some of which may undergo successive decays to form decay products.

Particle accelerators utilize magnets and alternating electrical fields to accelerate charged particles to very high speed and collide them into a stable target material.  This collision causes nuclei in the target material to take up high-energy protons, then to de-excite by emitting subatomic particles and radiation to form radioisotopes. Another approach is to bombard a primary target to produce neutrons or photons, which collide with a production target to form radioisotopes of interest.  Particle accelerators have several advantages: first, they produce far less (<10 percent) and far less hazardous radioactive waste than nuclear reactors; second, they tend to be less centralized, i.e., isotope production is spread over a greater number of facilities, meaning radioisotopes are produced locally or regionally rather than transported over long distances; and third, there is zero risk of a nuclear accident since they are powered by electricity rather than uranium fission.  In the context of the JCPA, their perceived overriding advantage is that particle accelerators do not pose a proliferation risk[20] since they utilize neither highly enriched uranium targets nor controlled chain reactions that produce weapons-grade nuclear material.

Iran’s IR-40 nuclear reactor at the Arak Nuclear Complex is a heavy water research reactor.[21] It is a “fundamental principle” of the JCPA that “Iran will redesign and rebuild” the Arak reactor to operate on low enriched (≤3.67 percent by mass) uranium in order to “support peaceful nuclear research and radioisotope production for medical and industrial purposes.”[22] Converting the Arak IR-40 to low enriched uranium fuel is intended to reduce its production of weapons-grade plutonium by between 73 percent and 85 percent (the range is a function of the LEU’s enrichment level).[23] A proliferation-resistant enrichment level still results in the production of some plutonium. Its diversion by Iran would be susceptible to IAEA detection and discovery.

It is important to note that while the JCPA states explicitly “The redesigned and rebuilt Arak reactor will not produce weapons grade plutonium,”[24] [emphasis added], modifying the Arak IR-40 reactor to use low enriched uranium fuel does not ipso facto eliminate the production of plutonium isotopes, most of which would be weapons-grade.[25] The November 2013 Joint Plan of Action spoke to design modification so that the Arak reactor would produce less, but not no, plutonium.[26] An authoritative 2014 paper assessing the feasibility of converting the Arak IR-40 to low enriched uranium fuel similarly concluded “it is possible to produce a 5-percent enriched LEU core with the same fresh-core reactivity and core life as the original, unmodified reactor, but which produces weapon-grade plutonium at only 19 percent the rate of the original reactor design.”[27]  This would represent a substantial reduction, but not a reduction to zero.

This perhaps underlies the JCPA’s provision that the redesigned Arak reactor “will be operated under IAEA monitoring.”[28] Iran makes a related commitment for a period of fifteen (15) years not to “engage in any spent fuel reprocessing…with the sole exception of separation activities aimed exclusively at the production of medical and industrial radioisotopes from irradiated enriched uranium targets.”[29] This refers to the practice of placing highly enriched uranium “targets” in high neutron fluxes near the cores of high-powered research reactors.[30]

So far, so good. Things may not be as simple as they seem, however, as Jeremy Bernstein writes:

“In Obama’s universe, the reactor at Arak would have its core redesigned so it would no longer pose a threat as a site for plutonium production. […] In Rouhani’s universe, heavy water will continue to be used at Arak, and no mention is made of plutonium…Inspections by the IAEA will be strictly limited, certainly when it comes to military facilities; in any case, the inspections are irrelevant because there is nothing to inspect.”[31]

The relevant part of the JCPA reads as follows:


8. Iran will redesign and rebuild a modernised [sic] heavy water research reactor in Arak, based on an agreed conceptual design, using fuel enriched up to 3.67%, in a form of an international partnership which will certify the final design…The redesigned and rebuilt Arak reactor will not produce weapons grade plutonium.”[32]

There are related limits on spent fuel reprocessing, further heavy water reactors, and heavy water accumulation. The JCPA’s use of the term weapons grade plutonium is unambiguous—it means plutonium in which the concentration of the fissile isotope plutonium-239 is at least 93 percent (and consequently, the plutonium-240 concentration is less than 7 percent)—so the only question is whether the Arak reactor is, or is not, producing it, which is a function of how long uranium-238 targets remain in the reactor core (where they are irradiated with neutrons). Producing weapons grade plutonium requires relatively brief irradiation, called “low burn-up.”[33] However, absent Iran’s abject failure to comply with the JCPA and/or intention to frustrate IAEA inspections, answering the Arak plutonium question seems clear cut.[34]

Moving on, Annex III addresses radioisotope “technologies, production and facilities.” It is perhaps the document’s least studied provision, probably because the language seems relatively innocuous in the context of an agreement where the prime objective is to limit Iran’s development of nuclear weapons. It is nonetheless consequential, providing “[T]he E3/EU+3 and Iran will seek cooperation and scientific exchange in the field of nuclear science and technology.” This includes “Research reactor applications at the TRR [the Tehran Research Reactor], modernized Arak reactor, or at other future research reactors in Iran,” including among such application “Radio-isotope production and utilization.”[35]  This raises two sites that have not been discussed so far—the existing Tehran Research Reactor, and the research reactors at Iran’s Esfahan Nuclear Technology Center— as well as the open question of Iran’s future plans to add research reactors [see footnote].[36]

Tehran Research Reactor. The Tehran Research Reactor (aka IR-0001) is a light water research reactor supplied to Iran in 1967 under the United States’ “Atoms for Peace” program. Iran built a separation facility for handling radioisotopes at the Tehran Nuclear Research Center in 2005 with the intention of synthesizing up to 20 different radioisotopes.  In the late 1980s Iran contracted with Argentina’s Investigaciones Aplicadas[37] to convert it from 93 percent enriched uranium fuel to slightly less than 20 percent enriched uranium fuel, just below the cutoff for it to qualify as highly enriched uranium.[38] As discussed above, the diversion of low enriched uranium fresh fuel is a safeguards concern (albeit a lesser one) because it can be further enriched to highly enriched uranium or irradiated in a clandestine reactor for plutonium production. Plutonium in spent low enriched uranium fuel also can be separated using well-known processes. There have been past allegations that the Tehran Research Reactor was used to conduct activities related to nuclear weapons development, including undeclared plutonium experiments[39] and the production of polonium-210.[40]

Esfahan Research Reactors. The Esfahan Nuclear Technology Center opened in 1984 and is Iran’s largest nuclear research complex. There are four Chinese origin research reactors at Esfahan:


  • The Esfahan Light Water Sub-critical Reactor (IR-0002) is a zero power reactor that runs on uranium metal fuel.  It was supplied by China in 1988 and first became operational in 1992, and currently is used for training purposes.[41]
  • The Esfahan Miniaturized Neutron Source Reactor (IR-0005) runs on 900 grams of 90% highly enriched uranium supplied by the China National Nuclear Corporation. The reactor was supplied by China in 1991 and first reached criticality in 1994. IR-0005 is suitable for the production of short-lived radioisotopes, and is used today for neutron activation analysis, a nondestructive nuclear process used to determine the concentration of elements in physical materials.[42]  Other Chinese origin MNSRs (a copy of the Canadian SLOWPOKE-2[43] research reactor) of the same design were built in Syria, Pakistan, Ghana, and Nigeria; all except the Syrian reactor later converted to low enriched uranium fuel.
  • The Esfahan Heavy Water Zero Power Reactor (IR-0004) runs on natural uranium fuel. It was constructed by China in 1991 and became operational in 1995. IR-0004 is used for heavy water research and while it is not suitable for plutonium production, has been used to gain experience in the control of larger heavy water reactors such as the Arak IR-40.[44]
  • The Esfahan Graphite Sub-Critical Reactor (IR-0003) is a decommissioned zero power light-water reactor (with graphite reflector) that was constructed by China in 1991 and first became operational in 1992.[45]

Esfahan is said to be the primary location of Iran’s nuclear weapons program. In January 1990 China and Iran signed a nuclear cooperation agreement that reportedly included construction of a plutonium production reactor. American satellite imagery first detected construction activities at Esfahan in September 1991.  This was reported in some open sources to be the IR-0005 Miniaturized Neutron Source Reactor.  Esfahan is also the reported site of Iran’s largest missile assembly and production plant, built with North Korean assistance. According to Russian reports attributed to the Russian FSB (Federal’naya sluzhba bezopasnosti) Federal Security Service, Esfahan is a production site for Scud-B and Scud-C surface-to-surface missiles, which Iran assembles from North Korean and Chinese-origin components.[46]

Radioisotope Technologies, Production & Facilities

Accelerators are a viable route to acquiring fissile material. – R. Scott Kemp

The discussion going forward focuses on Iran’s capacity over the next decade to produce radioisotopes with accepted industrial and/or medical applications.  It currently has three research reactors capable of radioisotope production—the Arak IR-40 Research Reactor, the Tehran IR-0001 Research Reactor, and the Esfahan IR-0005 Miniaturized Neutron Source Reactor, respectively.  The JCPA contemplates Iran will expand the production capacity of these three research reactors by acquiring particle accelerators.

There is no reasonable objection to Iran expanding its domestic radioisotope production capacity per se, with two exceptions. The first pertains to weaponization and proliferation. This means the production of radioisotopes that have legitimate industrial and/or medical uses, but where Iran’s intent is to weaponize the radioisotopes itself or to transfer them to a third party which intends to do so. The end result would be either a military-grade radiological weapon or an improvised radiological device of some sort. The second exception pertains to Iran’s acquisition of accelerator-based production technologies that pose a dual-use risk, and its continued use of reactor-based production technologies. The term “dual-use” means a device or technology that has legitimate application in the production of industrial and/or medical radioisotopes, but also can be used for some illicit purpose such as producing radioisotopes for weapons.

The discussion now moves to radioisotope weaponization and proliferation, and dual use technologies used in radioisotope production. It begins by identifying radioisotopes Iran is likely to seek to produce that raise weaponization and proliferation concerns. It then addresses the question of dual-use technologies, identifying those with a known or theorized use outside legitimate radioisotope production that is inconsistent with the JCPA’s letter or spirit.

Radioisotopes Raising Weaponization & Proliferation Concerns. Many radioisotopes are commonly used in medical imaging.[47] Others have therapeutic applications or are used in medical research.[48]  Still others are commonly used in various industrial and commercial applications.[49]  Of these only a few pose a genuine weaponization and proliferation risk.

The weaponization concern extends to radiological weapons, and to improvised devices intended that disperse radiological material (a Radiation Dispersal Device) or to cause passive exposure to dangerous levels of radiation (a Radiation Exposure Device). The proliferation concern pertains to the transfer of radioisotopes to a third party, e.g., a violent non-state group like Hezbollah, for use in an improvised radiological weapon.

The ideal malevolent-use radioisotope strikes a balance between persistence (measured by its half-life) and radiation emission. Three widely used radioisotopes—cobalt-60, strontium-90, and cesium-137—are strong radiation emitters and have suitable half-lives, making them good candidates for use in a radiological weapon. A fourth, iridium-192, is a strong radiation emitter though it is somewhat compromised by a relatively short half-life. All four radioisotopes satisfy a fourth criterion—they are widely used in medical and/or industrial applications—which means their production is not inherently suspicious.

In the context of Iran expanding its production of medical and industrial radioisotopes, these four radioisotopes raise concerns of possible weaponization and proliferation[50]:

  • Cesium-137. A beta and gamma emitter with a half-life of 30.2 years, cesium-137 is a byproduct of nuclear fission. It is among the most commonly used industrial radioisotopes, with applications including moisture-density, leveling and thickness gauges, and well-logging devices. It has been used in medicine as a source of high-energy x-rays for external therapy and brachytherapy, though it has largely been supplanted by iridium-192. The United States has embargoed cesium-137 exports to Iran (along with Cuba, Iraq, Libya, North Korea and Sudan).
  • Cobalt-60. A gamma emitter with a half-life of 5.27 years, cobalt-60 is produced in a nuclear reactor by bombarding a target of non-radioactive cobalt-59 pellets with neutrons. Cobalt-60 is used in medicine for radiotherapy and brachytherapy; and has many industrial applications, including industrial gamma radiography, industrial gauging and leveling devices, and equipment sterilization and food irradiation.
  • Iridium-192. A beta and gamma emitter with a half-life of 74 days, iridium-172 is produced from neutron bombardment of nonradioactive iridium-191 (usually in the form of natural iridium wire exposed to neutron flux). Iridium-192 is usually produced in a nuclear reactor but can be produced with an accelerator as well.  Its applications are high-dose brachytherapy,[51] and non-destructive and industrial gamma radiography.
  • Strontium-90. A beta emitter with a half-life of 28.8 years, strontium-90 is a fission byproduct and a major radioisotope in spent nuclear fuel. It is used industrially as an isotropic energy source and oxygen eliminator, and is used in certain optical materials.

Iran for at least a decade has made known its intention to produce cobalt-60 and iridium-192 (along with molybdenum, iodine and xenon radioisotopes) at the Arak IR-40 Research Reactor.[52]

Dual-Use Technology Concerns. Other radioisotopes do not themselves raise weaponization or proliferation concerns, but their production using highly enriched uranium targets does trigger such concerns. A good example is technitium-99m, which is used in myocardial perfusion and other noninvasive medical imaging studies.[53] Iran today imports its entire supply of molybdenum-99 (which decays to technitium-99m), about half of which is wasted in transit due to decay. Were Iran to commence molybdenum-99 production at its Tehran or Arak research reactors, it would raise a proliferation concern even if low enriched uranium fuel is used since these reactors produce plutonium-239.

In conjunction with its provision requiring Iran to redesign the Arak research reactor to run on low enriched uranium, the JCPA contemplates “upgrades to the infrastructure associated with existing cyclotron facilities, including for medical radioisotopes production” and “facilitating Iranian acquisition of a new cyclotron”.[54] It does not contemplate Iran shifting all radioisotope production to from nuclear reactors to particle accelerators: they cannot produce all medical and industrial radioisotopes for which Iran could claim a legitimate use. There is an explicit pledge by the E3/EU+3 regarding “cooperation and scientific exchange in the field of nuclear science and technology” including “research reactor applications at the TRR, modernized Arak reactor, or at other future research reactors in Iran” including “radioisotope production and utilization.”[55] This factor—Iran’s nuclear reactor-based radioisotope production—and the imperative to monitor Iran’s radioisotope production generally undergird the JCPA’s explicit reference to “IAEA involvement” in Appendix III activities.[56]

Iran’s expanded production of radioisotopes by means of particle accelerators—something explicitly encouraged under the JCPA—raises the matter of dual-use technologies. Particle accelerators constitute a viable route to acquiring fissile material. Any technology that can produce radioisotopes can transmute uranium-238 into weapon-usable plutonium-239, which occurs when a uranium-238 atom captures a neutron. So a dual-use technology with an acknowledged place in the legal production of radioisotopes for medical and industrial applications—a particle accelerator—can also be used for illicit activities that raise proliferation concerns.[57]

It has been shown that relatively small commercial cyclotrons are capable of producing amounts of fissile material greater than the IAEA “screening limit,” which is a capability to produce at least 100g of plutonium per year.[58] As a decade-old IAEA report concluded, “It has been shown that this technology is now mature enough for fissile material production.”[59] In the specific context of the JCPA, however, it is unclear whether, and if so, to what extent, the accelerator-based radioisotope production infrastructure contemplated for Iran would be subject to proactive IAEA inspections to detect clandestine attempts to produce fissile material. And though there is nothing in the JCPA requiring countries to transfer accelerator technology to Iran, in a significant apparent deficiency, it does not constrain Iran to the acquisition of accelerators with minimal to no proliferation risk.

It is broadly acknowledged that particle accelerators can be used to fabricate plutonium-239[60]—it was first proved in 1941, and many countries at one time or another pursued (and eventually abandoned) this production route—but generally not in weapon-scale quantities.[61] It is nonetheless possible to produce small quantities of plutonium-239 by modifying particle accelerators that were intended for medical use, either in radioisotope production or as a source of ionizing radiation. It is also possible to amplify a particle accelerator’s plutonium-239 production rate by as much as a factor of 20 through the use of a sub-critical assembly of natural uranium around the accelerator beam target. Linear proton accelerators or isochronous cyclotrons are able to produce fissile material at scale and at a reasonable production cost.

It is true that doing so would invite detection, given that the clandestine use of a sub-critical assembly requires the diversion of large quantities of uranium. It also would require Iran to dedicate a large number of particle accelerators to plutonium-239 production given the modest production rate achievable with this mode. Given Iran’s past activities and persistence in the face of international opprobrium (and punitive economic sanctions), its acquisition and use of particle accelerators must be closely and continuously scrutinized. It is worth noting that this concern is neither novel nor unique to Iran: in 2012 the Nuclear Threat Initiative warned regarding Syria’s Damascus Nuclear Medicine Center, “Of potential proliferation concern, particle accelerators of sufficient power are a viable route to acquiring fissile material. Theoretically, Syria could use the facility to conduct research on uranium enrichment or reprocess plutonium.”[62]

The JCPA also addresses the construction of “hot cells” used for separating medically and/or commercially useful radioisotopes from irradiated targets.[63] Hot cells are “dual-use” facilities, that is, they can be used for radioisotope processing as well as for plutonium separation. Iraq, Romania, Yugoslavia, and North Korea are all examples of countries where research reactors and hot cells have been used in connection with covert nuclear weapons programs.

Iran earlier declared to the IAEA plans to construct a building at Arak with hot cells for the production of what it called “long-lived” radioisotopes in addition to the hot cells for the production of “short-lived” isotopes. Many experts at the time interpreted “long-lived” radioisotopes to include significant amounts of plutonium. Iran revised its declaration in May 2004 to eliminate hot cells for long-lived isotopes.  Analysts remained concerned nevertheless that Iran still retained the option to build hot cell facilities at Arak or elsewhere for use in separating plutonium.[64] The JCPA appears to resolve this concern.

Does the JCPA Lower, or Raise, Weaponization & Proliferation Risks?

We are ready for any unforeseen event that may or may not occur. – Dan Quayle

The JCPA scrupulously addresses the proliferation risk associated with the use of highly enriched uranium in the production of medical and industrial radioisotopes by requiring the conversion of Iran’s Arak IR-40 research reactor from highly enriched uranium fuel to low enriched uranium fuel. It also subjects the Tehran Research Reactor to scrutiny. The JCPA also promotes the partial conversion of Iran’s mode of radioisotope production from nuclear reactors to particle accelerators, which are presumed to present a lower, but not zero, proliferation risk.

Paradoxically, the JCPA at the same time creates conditions under which the threat of weaponization could escalate, were Iran inclined to do so, and were the JCPA’s expectation of an omniscient inspection regime unfulfilled so far as radioisotope production. Those conditions arise from the expansion of Iran’s radioisotope production infrastructure—something intended under the JCPA—and its consequent decentralization as Iran acquires particle accelerators from foreign suppliers and places them in operation. These same conditions raise the possibility of an associated proliferation risk, were Iran to make weapon-suitable radioisotopes available to malevolent third parties. It is not entirely clear whether the JCPA contemplates an oversight regime with respect to Iran’s production of radioisotopes that is as global and intrusive as it does with respect to nuclear reactors. On this question, deference is owed the IAEA, although explicating how it will work with respect to particle accelerator radioisotope production would be beneficial.

None of this is to say that Iran intends to engage in activities in the realm of radioisotope production that raise weaponization and/or proliferation concerns. The author has no particular insight into the collective mind of the Iranian leadership, though its past belligerent statements regarding radiological weapons offers no comfort. Nor it must be conceded, has Iran’s past conduct earned it the benefit of the doubt with respect to its weaponization or proliferation intentions.

Surely, if the international community is determined to deny Iran the “N” category of the CBRN weapons of mass destruction quartet—chemical, biological, radiological, and nuclear—then it should be equally determined to do so with respect to radiological weapons. It is true that radioisotope production does not ipso facto mean a country will build radiological weapons: Australia, Belgium and the Netherlands among other radioisotope-producing nations do not. But radioisotope production remains the indispensable precursor to radiological weapons. Nor does the domestic production of weapons-usable radioisotopes mean a country will become a proliferator. But proliferation concerns are especially acute with respect to Iran given its long-time and continuing support of violent proxies such as Hezbollah.

The author’s judgment is that between weaponization and proliferation, the latter poses the more insidious risk since weapon-usable radiological material in the possession of a violent non-state actor is more likely actionable, meaning the likelihood of its malevolent or threatened use. If correct, it puts a premium on ensuring that Iran does not divert weapon-usable radioisotopes from bona fide medical or industrial uses. Such oversight is not explicitly contemplated under the JCPA. The second proliferation risk is Iran’s potential misapplication of particle accelerators to produce illicit radioisotopes, or to otherwise evade oversight of its radioisotope production, which likely will focus on research reactors.

This is not to dismiss the weaponization risk altogether. Iran has an oft-demonstrated penchant for testing the robustness of oversight regimes, and international willingness to enforce the letter of the law. The JCPA’s omission of any explicit reference to radiological weapons is a meaningful, but in the author’s opinion, not in and of itself, a disqualifying flaw.

On balance, then, the JCPA raises the proliferation risk with respect to weapon-usable radioisotopes by expanding Iran’s production capacity. This is by no means the same as concluding Iran will exploit this opportunity. It is also the case that the Arak IR-40 research reactor will operate under IAEA monitoring. The case for weaponization is less clear since Iran would have to divert large quantities of weapon-usable radioisotopes to such a program, and this fact alone would risk exposing any weaponization program to detection. Iran’s past behavior begs the question of how far it might be willing to go, however.

Trust, writes Andrew Kydd, is a belief that the other side prefers mutual cooperation to exploiting one’s own cooperation, while mistrust is a belief that the other side prefers exploiting one’s cooperation to returning it. So technical matters aside, the question is should we trust—or mistrust—Iran? The dilemma, as Ian Anthony of the Stockholm International Peace Research Institute observed, is the mismatch between the observed characteristics of Iran’s nuclear program, and Iran’s persistence in representing its peaceful intent. Trying to assess Iran’s intentions is at best an imperfect science. As Robert Jervis warned 40 years ago, a “view later proved incorrect may be supported by as much evidence as the correct one, or even more.”[65] At the same time, bellicose factions of the Iranian leadership seem determined to erase ambiguity about Iran’s intentions by reinforcing doubters’ least favorable preconceptions.[66]

The JCPA’s objective is to “ensure that Iran’s nuclear program will be exclusively peaceful,” the plain meaning of which is that “under no circumstances will Iran ever seek, develop or acquire any nuclear weapons.”[67] Proscribing Iran’s potential to weaponize and/or proliferate radiological materials, while congruent with that objective, is not perforce wholly part of it. Iran should be compelled to make a full disclosure of any radiological weapon program when answering the IAEA’s concerns about the possible military dimensions (PMD) of its nuclear programs. The status of radiological weapons under the PMD disclosure is unclear, however. Regarding proliferation, extraordinary diligence is required, given Iran’s not infrequent use of violent non-state proxies as the sharp end of its geopolitical ambitions. The disruptive effect of radiological material in the hands of Hezbollah and other violent Iranian proxies—let alone its effect on target populations should they use it—would be far more acute today than the over the horizon risk of an Iranian nuclear weapons program.

With respect to the JCPA, it seems de rigueur for technical analyses to end with a political judgment. There seems no purpose in doing so here, however, as the Security Council has already spoken and Congress has begun its formal assessment. Suffice it to say the JCPA, if imperfect in its treatment of radioisotope-related weaponization and proliferation, contains nothing in that realm at least that rises to the level of a disqualifying flaw. Iran has actively pursued expanded domestic radioisotope production and self-sufficiency, which the JCPA attempts to channel away from research reactors and toward particle accelerators so far as practicable. There is an inherent risk in doing so, of course. That risk, however, is more easily rated (and in the right circumstances, more easily managed) than an unsupervised program whereby Iran aggressively pursues radioisotope production using research reactors and highly enriched uranium targets, especially where alternate modes exist. The JCPA is in the end a political solution to a political problem, and that is the way history will likely judge it.



[1] The quote is from a novel by Robert Jordan, nom de plume of the late American author James Oliver Rigney, Jr.

[2] Also called the P5-+1, it is comprised of China, France, Germany, the Russian Federation, the United Kingdom, and the United States, along with the High Representative of the European Union for Foreign Affairs and Security Policy

[3] The full text can be accessed at  On the same day that the E3/EU+3 and Iran signed the JCPA, the Iranian government and the International Atomic Energy Agency (IAEA) signed a second document called the “Road-map for the clarification of past and present outstanding issues regarding Iran’s nuclear program” [].

[4] An isotope of any element has the same number of protons in its atom (atomic number) but a different mass due to the different number of neutrons (i.e., atomic mass is the sum of the protons and the neutrons). When a combination of neutrons and protons that does not exist in nature is produced artificially, the resulting atom will be unstable.  Such an atom is called a radioactive isotope or radioisotope, and emits radiation during its decay to a stable form. Unstable natural isotopes arise from the decay of primordial uranium and thorium. Overall, there are some 1800 known radioisotopes.

[5] One type of research or non-power nuclear reactor is used primarily as a neutron source and not for power generation.  Neutron source research reactors used for radioisotope production are broadly classified as one of two types: (1) enriched uranium, light water moderated, swimming pool type reactors; or (2) natural uranium, heavy water moderated and cooled tank type reactors.  Iran’s IR-40 nuclear reactor at the Arak Nuclear Complex is a heavy water research reactor.  Other types of research reactor are used as training or prototype reactors; or as zero-power critical assemblies used to evaluate proposed nuclear power reactor designs.

[6] From a 1964 New Scientist article by the same name on the same subject.

[7] Here, a solution containing technitium-99m, a radionuclide widely used in medical imaging.  For a detailed discussion of Israel’s Greenhouse experiments (and the companion Redhouse experiments on indoor detonations), see the author’s essay, “Iran, DAESH & the Rising Specter of Radiological Warfare in the Middle East” /articles/2015/07/iran-daesh-rising-specter-radiological-warfare-middle-east.

[8] Борис Евсеевич Черток (1996; 1999). Книга 2. Ракеты и люди. (Москва: Машиностроение, 1999).  Chertok’s entire four-volume memoir is available online in Russian.  For Volume II., see:  In 1999 NASA published an English translation of Chertok’s memoir.  See: Boris Evseevich Chertok (1999). Rockets and People, Volume II: Creating a Rocket Industry. (Washington, DC: National Aeronautical and Space Administration). Last accessed 15 July 2015. .Ракеты и люди

[9] R-2 [NATO designation: SS-2 Sibling] in the English language designation for a Soviet single-stage rocket, the P-2, the prefix of which stands for ракета (raketa or “rocket”).  The GRAU put the R-2 into service in November 1951 under the cryptonym “Article 8Zh38” [Russian: 8Ж38] using the MAD’s standard number-letter-number designation.  The acronym GRAU stands for Glavnoye raketno-artilleriyskoye upravleniye [Russian: Главное ракетно-артиллерийское управление (ГРАУ)] or the Main Missile and Artillery Directorate in the Soviet Defense Ministry.

[10]  Geran [Russian: герань] is the Russian transliteration of Geranium.

[11] Generator [Russian: генератор] is the transliteration of the same word in English.  The warhead was later deployed on a successor to the R-2, the R-5, which was first deployed in 1955.  Considerations of mass and radiation meant the Generator had to be handled by a specially designed, heavily shielded vehicle built by the OKB-156 factory.

[12] Strontium-90 and cesium-137 are fission byproducts.

[13] The R-5 missile [NATO designation SS-3 Shyster] was a medium range ballistic missile, and the first Soviet missile with nuclear delivery capacity.

[14] Iran has for some time declared its intention to become self-sufficient in radioisotope production.  In October 2012, the Iranian Students’ News Agency (ISNA) quoted Iran’s National Security Committee spokesman Hossein Naqavi Hosseini saying that Iran will become self-sufficient in the current calendar year which began on 21 March 2012 [].  In May 2011, Mohammad Qannadi, deputy director of the Atomic Energy Organization of Iran announced that Iran was now self-sufficient in production of medical radioisotopes.  “We have reached the final stage in the production of medical radioisotopes… and now there is no need for importation of medical radioisotopes,” Qannadi said. He also said that the project to produce industrial isotopes was already well underway. []

[15] There is no organized legal framework at the international level to control radiological weapons despite calls by the United Nations General Assembly’s First Special Session on Disarmament as far back as 1969, when the General Assembly requested the Conference of the Committee on Disarmament to consider effective methods of control against the use of radiological methods of warfare.  [UNGA (1969).  Resolution 2602 C. Last accessed 16 July 2015] In 1978 the UNGA called for a convention to prohibit ” the development, production, stockpiling and use of radiological weapons.” [UNGA (1978). Resolution Adopted on the Report of the Ad Hoc Committee of the Tenth Special Session, Article 76. See p. 9 at Last accessed 16 July 2015]  The UNGA has approved additional resolutions, for example, equating armed attacks on nuclear facilities to the use of radiological weapons [UNGA Resolution 43/75 adopted 7 December 1988.$FILE/A-44-621.pdf. Last accessed 16 July 2015]  Other multinational bodies have called for barring the use of radiological weapons, e.g., the North Atlantic Treaty Organization. [NATO (2002). Resolution 321 on Terrorism with Chemical, Biological, Radiological and Nuclear Weapons. Last accessed 16 July 2015]

[16] Radiochemical separation techniques comprise a third production method.

[17] There is good reason to be concerned: a number of countries that were supplied with HEU research reactor fuel were later discovered to have covert weapons programs, including Israel, Libya, Romania, Taiwan, South Africa, South Korea, and Yugoslavia.

[18] Radioisotopes are produced in a nuclear reactor either as fission products—fission of uranium results in some 80 different fission products—or through neutron capture, which results in one specific radioisotope

[19] This can be either highly enriched uranium (HEU)—a target with more than 90%mass of 235U—or low enriched uranium (LEU) [19]— a target with less than 20%mass of 235U.

[20] This statement is not categorically true, however, as discussed later in this essay.

[21] The use of so-called “heavy water”—the common name for water enriched with deuterium oxide, a form of water in which the most of hydrogen atoms contain a neutron in addition to a proton—as a moderator “slows” neutrons in a nuclear reactor and allows the use of unenriched uranium—uranium with a higher proportion by mass of uranium-238—as the fuel.  Most analysts consider heavy water reactors as a proliferation risk by: heavy water reactors do not use enriched uranium, which can allow a malevolent operator to evade IAEA supervision; and they produce more plutonium than light water reactors (the absorption of a neutron by uranium-238 leads to the production of plutonium). Experts believe the Arak IR-40 is based on a Russian design and that Iran received assistance in its design from unspecified Russian companies.

[22] JCPA paragraph 8. There is a similar provision in JCPA Annex I. Nuclear-related materials, paragraph 2.

[23] The Arak reactor is notable for being an exception: while some 46 countries have a total of 703 research reactors which are used for training, research and radioisotope production, only four—Canada, India, France and Algeria—use heavy water reactors for isotope production.

[24] Ibid.

[25] For a detailed discussion of plutonium production in a modified reactor core, see: Thomas Mo Willig (2011). “Feasibility and benefits of converting the Iranian heavy water research reactor IR-40 to a more proliferation-resistant reactor.” Norwegian Defense Research Establishment report dated 14 December 2011 (FFI-rapport 2011/02283), 62-65. Last accessed 21 July 2015.

[26] A similar situation arose in the early 1990s with the construction of a Chinese-origin research reactor in Algeria, the design of which was modified to minimize plutonium production along with an additional commitment to place the reactor under IAEA safeguards.

[27] R. Scott Kemp (2014). “Two Methods for Converting a Heavy-Water Research Reactor to Use Low-Enriched-Uranium Fuel to Improve Proliferation Resistance After Startup” (March 2014), 22. Last accessed 21 July 2015.

[28] JCPA Annex I. Nuclear-related materials, paragraph 12.

[29] JCPA paragraph 9. There is a similar provision in JCPA Annex I. Nuclear-related materials, paragraph 19.

[30] Molybdenum-99 is the most commonly used radioisotope produced by this method. Its decay product technetium-99m is the most widely used radioisotope in nuclear medicine, accounting for 80% of all diagnostic nuclear medicine procedures. For a complete list of radioisotopes produced in research reactors, see: IAEA (2003). Manual for reactor produced radioisotopes. (Vienna: IAEA). Last accessed 21 July 2015.

[31] Jeremy Bernstein (2015). “Which Iran Deal?” New York Review of Books [published online 15 July 2015]. Last accessed 18 July 2015.  Bernstein is referring to statements by Iranian President Hassan Rouhani during a 14 July 2015 televised address, the full video of which is available here:

[32] JCPA, paragraph B.8.

[33] This is not to say that lower-purity reactor-grade (80%-93% plutonium-239) plutonium cannot also be used in nuclear weapons, but that pathway is technically more demanding. The only isotopic mix of plutonium which cannot realistically be used for nuclear weapons is nearly pure plutonium-238, which generates so much heat that the weapon would not be stable.

[34] Plutonium production in a research reactor can be optimized if the core is loaded with a driver fuel to maintain reactor criticality and a target fuel to generate plutonium. The driver fuel can be HEU or LEU, whereas the target fuel may be either natural or depleted uranium. [National Nuclear Security Administration (2012). “Next Generation Safeguard Initiative. NGSI-SBD-003 (September 2012), 16. Last accessed 21 July 2015] However, while it is possible to irradiate low enriched uranium fresh in a clandestine reactor for plutonium production, it is difficult to envision how such a scheme would evade IAEA detection.

[35] JCPA Annex III., paragraph C.

[36] Ali Akbar Salehi, head of the Atomic Energy Organization of Iran, said in June 2010 that Iran planned to build four new research reactors at undisclosed locations in different parts of the country for the production of medical isotopes. See:

[37] Its name translates as the Applied Research Institute, which is known by the acronym INVAP.

[38] The United Nuclear Corporation supplied HEU to Iran starting in 1967 under a United States government export license until the United States stopped authorizing HEU shipments to Iran in 1979 after the Islamic Revolution.  Iran retained and stored several kilograms of irradiated United States-origin HEU fuel once the Tehran Research Reactor was converted to LEU.

[39] The Tehran Research Reactor’s produced inadequate plutonium for nuclear weapon production but sufficient amounts for testing experimental ways of extracting plutonium from irradiated research reactor fuel.

[40] Polonium-210 is a strong alpha emitter that is used in a beryllium-polonium neutron initiator that starts the chain reaction in a nuclear weapon.

[41] Sources: International Atomic Energy Agency [] and the James Martin Center for Nonproliferation Studies at the Monterey Institute of International Studies [].

[42] Sources: International Atomic Energy Agency [] and the James Martin Center for Nonproliferation Studies at the Monterey Institute of International Studies [].

[43] The SLOWPOKE-2 (an acronym for Safe Low-Power Kritical Experiment) is a low-energy research reactor designed in 1976 by a Canadian government-owned corporation, Atomic Energy of Canada Limited.  Low power reactors like SLOWPOKE-2 cannot routinely produce radioisotopes for medical use but can produce some tracers and very small amounts of radioisotopes for medical research, including sodium-24, argon-41, and technitium-99m.

[44] Sources: International Atomic Energy Agency [] and the James Martin Center for Nonproliferation Studies at the Monterey Institute of International Studies [].

[45] Sources: International Atomic Energy Agency [] and the James Martin Center for Nonproliferation Studies at the Monterey Institute of International Studies [].

[46] Source:

[47] Accelerator-produced examples include carbon-11, fluorine-18, gallium-67, indium-111, iodine-123, krypton-81m, nitrogen-13, oxygen-15, rubidium-82, technetium-99m, thallium-201, and xenon-127.

[48] Reactor-produced examples include chromium-51, copper-64, dysprosium-165, erbium-169, holmium-166, iodine-131, iron-59, lutetium-177, palladium-103, phosphorus-32, potassium-42, rhenium-186, samarium-153, selenium-75, sodium-24, and yttrium-90. Accelerator-produced examples include cobalt-57 and indium-111.

[49] Commonly used radioisotopes include americium-241, cadmium-109, californium-252, carbon-14, cesium-137, curium-244, iron-55, krypton-85, nickel-63, polonium-210, promethium-147, radium-226, sodium-24, sulphur-35, thallium-204, tritium, and thorium-229.

[50] A less persuasive case can be made for californium-252, which is an alpha and gamma emitter with a half-life of 2.6 years. It is available only in very small quantities (the element does not occur naturally) limiting its use in medicine (external neutron brachytherapy) and industry (as a portable neutron source and in moisture gauges). Similarly, polonium-210 is a strong alpha emitter but compromised by a short half-life (140 days) and, like californium-252, scarcity. It is produced artificially in a nuclear reactor by bombarding nonradioactive bismuth-209 with neutrons inside a nuclear reactor to form bismuth-210, which decays to polonium-210 through beta decay.

[51] Brachytherapy is a method in which sealed sources are used to deliver a radiation dose at a distance of up to a few centimeters by surface, intracavitary (implant), or interstitial application.

[52] Iran made its initial declaration in a 5 May 2003 letter to the IAEA, when it first informed the IAEA of its intention to construct a heavy water research reactor at Arak.

[53] Technetium-99m is produced by bombarding molybdenum-98 with neutrons in a nuclear reactor or a particle accelerator.  This produces the radioisotope molybdenum-99, which decays to technitium-99m (the “m” stands for metastable, meaning that it has a 6 hour half-life and decays by emitting a gamma ray).  It is not practical to produce technitium-99m directly because its 6.6-hour half-life does not allow for the radioisotope to be distributed efficiently from the production site to point-of-use. So instead, it is produced from the decay of molybdenum-99, which has a 66-hour half-life. In a research reactor, molybdenum-99 is typically created as a fission product of low- or highly enriched uranium targets exposed to a thermal neutron flux. Afterwards, the molybdenum-99 is removed from the targets and into “generators,” in which the molybdenum-99 decays with a 66-hour half-life into technitium-99m. Nine research reactors currently produce all the technetium-99m produced worldwide, which is subject to periodic shortages due to unexpected disruptions.

[54] JCPA Annex III, paragraph E.

[55] JCPA Annex III, paragraph C.

[56] JCPA Preamble and General Provisions, paragraph xiii.

[57] The author has purposely omitted a technical discussion on the subjects of radioisotope production and particle accelerators. Suffice it to sat that not all particle accelerators can produce radioisotopes of interest for weaponization purposes, nor do all pose a proliferation risk.  That being said, neither poses a particularly high bar to an Iran determined to go down the weaponization and/or proliferation pathways.

[58] IAEA (2004). “Implications of Partitioning and Transmutation in Radioactive Waste Management.” Technical Report Series No. 435. (Vienna: IAEA), 34. Last accessed 20 July 2015.

[59] Ibid., 35.

[60] This can be accomplished either by using accelerator-produced neutrons, or by directly bombarding a uranium-238 target with an energetic proton beam. It should be noted that the resulting production are low and so have been considered inefficient given the alternative of a nuclear reactor.

[61] On this point, however, Scott Kemp offered the following qualification: “Although established nuclear states have favored reactors for reasons of cost and reliability, these case examples cannot inform questions about entry-level proliferation. Established nuclear powers and technologically-sophisticated states need not contend with export controls, limited material resources or technical expertise, and clandestine objectives; the entry-level proliferator must.” See: R. Scott Kemp (2005). “Nuclear Proliferation with Particle Accelerators.” Science and Global Security. 13, 186. Last accessed 20 July 2015.

[62] Last accessed 20 July 2015.

[63] A “hot cell” is a shielded radiochemical laboratory outfitted with remote handling equipment for examining and processing radioactive materials.

[64] Institute for Science and International Security (2013). “Update on the Arak Reactor.” ISIS Report (15 July 2013). Last accessed 21 July 2015.

[65] Robert Jervis (1968), “Hypotheses on Misperception.” World Politics. XX (April 1968), 487. Quoted in Barry Steiner (2015). When Images and Alarm Collide: The Significance of Information Disparity. International Journal of Intelligence and Counterintelligence. 28:2.

[66] Suggested by Steiner (2015), 320.

[67] JCPA Preface.