Particle Accelerators for Tritium Production

These would work by firing protons at targets, probably lead but possibly bismuth or others [1]. Various things can happen when a high-energy proton hits a lead or bismuth nucleus, but by far the most common (apart from simply bouncing off) is that a single neutron will be ejected. (See [2]) This converts a lead nucleus into a bismuth nucleus (or a bismuth nucleus into a polonium one) of the same atomic mass. There are four naturally occurring isotopes of lead, and one of bismuth:

Isotope Percentage p,n → Isotope Half-life Decay
(years) (MeV)
208Pb 52.4 208Bi 368,000 2.880
207Pb 22.1 207Bi 31.55 2.398
206Pb 24.1 206Bi 0.017 3.758
204Pb 1.4 204Bi 0.001 4.438
209Bi 100 209Po 102 4.979
208Bi * 208Po 2.898 5.215

* This does not occur in nature, but will be present in lead in increasing quantities in this kind of particle accelerator, as a result of the reaction in the first row of the table. 208Po is the only secondary product that’s a potential problem.

All the decays of the bismuth isotopes are electron captures back to the isotope of lead they were produced from. Almost all (>99.7%) of the energy is emitted as gamma rays.

There’s another complication, in that although 205Pb doesn’t occur naturally, 204Pb is a fairly good neutron absorber (at all energies) so in a particle accelerator that’s producing neutrons, some 204Pb will be converted into 205Pb. With a half-life of 15.3 million years, this is a problem waste in itself (albeit much less so than the 208Bi or the 207Bi). It can also be transmuted to 205Bi, which has a half-life of 15.31 days and so is only a problem in the event of a particle accelerator accident.

The decay of 209Po is the emission of an energetic alpha particle (4.979MeV), so with a half-life of 102 years this is a nasty waste product – highly radioactive but around for centuries. The product of its decay is 205Pb, which is much less radioactive (a 51keV gamma emitter with a half-life of 15.3 million years) but nonetheless a long-term disposal problem.

209Bi has a significant cross-section for radiative capture of intermediate energy neutrons, producing 210Bi, which decays rapidly (just ~5 day half-life) to 210Po, an energetic (5.407MeV) alpha emitter with a half-life of 138 days – the nasty stuff they murdered Alexander Litvinenko with.

[1] There’ll be another page about other possibilities soon.

[2] The data for p,n reactions are given at: https://t2.lanl.gov/nis/data/endf/endfvii-p.html. The information for lead is right at the bottom. The pdf plots are much easier to understand than the raw data! In each case it’s the twelfth graph down that you want, to see the cross-sections for all the various particle-generating reactions. (The cross-sections for elastic collisions are very much larger. You can find them in the earlier graphs. Elastic collisions don’t reduce the proton’s energy much, because the nucleus is so much more massive than the proton, and they randomize the direction of travel of the proton. A proton could engage in several elastic collisions and still have enough energy for a p,n reaction.)

Sadly, there’s no data for 204Pb or bismuth, but the pattern for all the main isotopes of lead is similar, and something very similar probably applies to 204Pb and 209Bi as well: a modest cross-section for p,n reactions and negligible cross-sections for other particle-generating reactions. However, 204Pb only constitutes 1.4% of natural lead, and bismuth probably wouldn’t be the target material of choice because of the production of 209Po.

It’s worth noting that the particle production cross-sections are zero up to about 10MeV (you need that much to overcome the Coulomb repulsion between the incident proton and the nucleus) and don’t become significant until 20-40MeV. This consumes a significant part of your fusion reactor’s output (17.6MeV per D-T fusion reaction). That’s not quite as bad as it first appears, because each fast neutron produced in a p,n reaction is quite likely to be converted into two intermediate energy neutrons in an n,2n reaction with another lead or bismuth nucleus. That’s the end of it, though: the intermediate energy neutrons don’t have enough energy for another n,2n reaction (unless you’re giving your initial proton a much higher energy to start with).

On the other hand, you have to remember that the particle accelerator will consume a good deal more electrical energy than it imparts to the accelerated protons as kinetic energy (that is, its efficiency will be less than 100%), and the fusion reactor will generate a good deal less electrical energy than the thermal energy produced by the fusion reaction (again, efficiency < 100%).

Most of the energy consumed by the particle generator will end up as heat in the target (some will end up as internal energy in the waste products – for the top four rows in the table, the decay energy in the last column is precisely that internal energy increase, which we know because their decay takes us back to the original lead isotope), and a percentage of that energy could possibly be recovered, but the efficiency of recovery would be limited.