Medical Isotope production - NRX reactor

Andrew Mason

Canada is currently the largest supplier of the world's radioactive medical isotopes. That is until AECL's aging NRU reactor at Chalk River sprang a leak and had to be shut down. AECL had built two reactors in the 1990's, Maple I and II, to produce medical isotopes but due to design problems (the reactors have a positive reactivity co-efficient that will require a complete redesign and rebuilding of the reactor cores to correct) these reactors have been scrapped by the Canadian government. There is now talk that the US will start its own isotope program using existing nuclear facilities.

I am not clear why the NRU is such a good reactor for producing these isotopes. Is there an advantage to using a heavy water moderated reactor to produce them? If so, why?

How easy would it be for the US to get into the field of producing medical isotopes using an existing light water reactor in the US?

AM

Astronuc

There is a program by GE in the US to produce radioisotopes (for medical use) in a commercial LWR. The details are proprietary at this point, and I don't know what has been released publicly.

Tritium has also been produced by the DOE in a commercial PWR.

I think that NRU was simply convenient for isotope production.

Fuel cycle may be a factor. Commercial reactors run 18 to 24 month cycles which would not be good for isotopes that reach equilibrium at shorter times. Research reactors can be shutdown on irregular schedules, so are not constrained by commercial generation demand.

The university where I studied nuclear engineering had a TRIGA reactor. They used to produce radioisotopes there, and IIRC some were for medical purposes.

Andrew Mason

Thanks. Given that the desired isotope, Molybdenum 99, has a half-life of about 66 hours, I would think that the reactor would have to produce useable quantities within a few hours and allow the target to be extracted immediately. The target (HEU) would have to be processed to separate the Mo99 and ship it for further processing and distribution within a few days.

There is another way to make Mo99 using particle accelerators and U238. I am not sure how you would fission Uranium using high energy electrons from a particle accelerator but apparently it can be done. The electrons strike something to produce gamma rays which, in turn, can induce fission of the U238 nucleus. It strikes me as being a much slower process and I am not sure how much Mo99 you could produce in a few hours but I expect not much.

AM

hywel

Hi there. Yes, like most research reactors NRU is handy compared to a commercial power reactor because targets can be inserted, irradiated, and removed more readily in a research reactor; reactors with empty slots are good for this. Also, to get a high equilibrium activity in the irradiated target (usually some kind of highly-enriched U235 foil or similar in a tube), you want as high a neutron density as possible, something you typically get in a high-power (greater than, say, 20 MW) research reactor. The irradiated target has to be processed (dissolving and chemical separation) in about a day to preserve the activity at useful levels for compact hospital generators.

You can make Mo-99 using an accelerator. In fact, the first Mo-99 was made by Emilio Segre using the 37-inch Wilson cyclotron - and he observed the decay of the Tc-99. There are a variety of possible nuclear reaction methods that utilise different projectile particles (protons, neutrons, deuterons, electrons and gammas), but the best one is the one used widely to make radioactive beams. You take a beam of electrons, hit a target with it to make gamma rays via Bremsstrahlung, then those gammas cause giant resonance fission in either U-235 or U-238. You need a LOT of electrons though, so accelerator people are working on that aspect of it. You are right that yields are lower, but you don't have to build a reactor (c. 1 B dollars) to do it, just an accelerator (c. 10 M dollars).