This reliance on such a limited - and ageing - resource results in an extremely delicate supply chain, the vulnerability of which was highlighted late last year when an extended shutdown of the NRU reactor lead to a critical Mo-99 shortage in North America. Similarly, the shutdown of the HFR Petten in August 2008 caused Mo-99 shortages in North America and Europe. Clearly, the situation needs to be addressed.

With this in mind, TRIUMF - Canada's national laboratory for particle and nuclear physics - has taken things in hand. Together with spin-off Advanced Applied Physics Solutions and the University of British Columbia, TRIUMF set up a task force (supported by Canada's Ministry of Natural Resources) earlier this year to evaluate the options for ensuring a continued supply of Mo-99. In particular, a group of about two dozen experts recently convened to explore the feasibility of using high-power accelerators - rather than nuclear reactors - to generate large quantities of medical isotopes.

Currently, Mo-99 is created by bombarding U-235 with neutrons from a nuclear reactor, with the resulting fission reaction producing Mo-99 around 6% of the time. The four main Mo-99 manufacturers use targets made from highly enriched uranium (HEU). But nuclear non-proliferation and security concerns have led to increased worldwide pressure to migrate all non-military applications away from the use of weapons-grade HEU and towards low-enriched uranium (LEU). Only the OPAL reactor in Australia (currently in the final phases of commissioning) uses LEU target material at a level capable of significant production, and the processing and licensing of that approach is not yet completed.

But there are other ways to produce Mo-99, and without needing weapons-grade uranium. The neutron-capture process, for example, involves using an intense neutron beam to add one neutron to a Mo-98 target, while the photo-neutron process uses an intense photon beam to remove a neutron from Mo-100. The most promising approach, however, and the main focus of attention for TRIUMF's task force, is photo-fission. Here, a highly intense photon beam generated by an electron accelerator is used to split natural U-238. The U-238 photo-fission process offers the same fractional yield of Mo-99 (around 6%) as neutron-induced fission of U-235. Having said that, the probability of photo-fission is around 3000 times less, which means that an extremely high photon flux is needed to equal the production rate.

The key advantage of the photo-fission technique is that natural or depleted uranium can be used as the target material, eliminating the security issues of transporting, storing and disposing of HEU. In addition, it removes the reliance on a nuclear reactor, with accelerators subject to more straightforward licensing procedures, able to be turned on and off at will, and comparatively inexpensive to decommission at end-of-life. On the downside, an accelerator-based production facility will require substantially more electrical power than a reactor-based facility.

In the ensuing report - Making Medical Isotopes - the Task Force on Alternatives for Medical-Isotope Production concludes that, based on preliminary calculations and simulations, significant amounts of Mo-99 can be produced by photo-fission of natural uranium using accelerators. The task force affirms that the technology exists to build an electron accelerator of suitably high beam power (2-3 MW) and predicts that the radiochemistry needed to recover and refine the generated Mo-99 will likely resemble that used for HEU-generated Mo-99. The report suggests that a single multi-megawatt machine could supply the Canadian market, or 5-7% of the total North American market. To ensure high reliability of supply, half a dozen multi-megawatt machines could be built, which would meet about 30-50% of North American demand.

The expense of setting up such a facility also has to be considered. With substantial uncertainties in the capital and eventual operating costs for a reliable accelerator-based facility, the report notes that further assessment is needed, using experience from lower power experiments and feasibility tests. The task force estimates that construction of a photo-fission accelerator will take three to four years, at a cost of somewhere between C$50 and C$125 million ($40-100). In addition, the existing facilities employed for the total medical-isotope production cycle (including manufacture of targets, storage of radioactive waste and hot-cell facilities to recover and refine the Mo-99) need replacement, which would cost at least C$50 million. Operational costs, meanwhile, will likely be dominated by power requirements.

So could the future of Mo-99 production lie in this new technique? While the task force did not state a preference for either nuclear-reactor or photo-fission-accelerator technology, it did conclude that accelerator-driven photo-fission of U-238 "has a sufficient number of attractive features that it warrants further attention by public and private enterprises." The next step is to establish a steering group of public and private partners that will develop the technology, oversee a proof-of-principle demonstration and then assess the commercial viability. As for the key scientific, engineering and operational challenges, the task force recommends an R&D programme that focuses on the following tasks:

  • Production, over about six months, of a short conceptual report describing the optimal design of a high-power electron linear accelerator using photo-fission for production of Mo-99.
  • Calculation of the capital and operating costs, based on the conceptual design report and site considerations.
  • Verification of photo-fission accelerator production of Mo-99 equivalency to the present product using laboratory experiments.
  • Design of a target facility that's capable of handling 2-3 MW of electron-beam power.

Completion of these work packages should enable the steering group to present a recommendation on the photo-fission technology within three to four years.

TRIUMF itself is currently planning to build a new electron-beam accelerator for general research purposes. While the total power output of this facility will be lower than ideal for large-scale Mo-99 production (initially 100 kW in 2013 with an upgrade path to 0.5 MW), the device will utilize the same basic technology, enabling laboratory validation tests with power densities equivalent to a full-power Mo-99 machine. TRIUMF predicts that the accelerator should be ready for low-power testing of Mo-99 generation in a few years, with the generated samples used to validate the beam-power requirements, isotope yields, target performance, chemical recovery, and refinement and purity of the Mo-99.

It's unlikely, however, that TRIUMF will go on to build its own dedicated production accelerator. "TRIUMF is primarily a basic research laboratory," Timothy Meyer, head of strategic planning and communications, told medicalphysicsweb. "While the lab would undoubtedly be a partner in any enterprise constructing a commercial Mo-99 facility using accelerators, we don't see TRIUMF being the sole actor." Meyer suggests that the lab's current arrangement with Canadian radioisotope supplier MDS Nordion - in which Nordion owns and maintains three small cyclotrons for producing short-lived "light" isotopes, while TRIUMF oversees the safety, licensing and regulatory aspects, and supplies employees to help operate the cyclotrons - could prove an ideal model for a future photo-fission facility.