Sir

In your News Feature 'The hybrid returns' (Nature 460, 25–28; 2009), you discuss the feasibility of creating a fusion–fission hybrid reactor to generate greenhouse-gas-free and waste-free nuclear energy. However, there is another challenge to add to the factors to be considered: where would the hundreds of grams of tritium needed daily to fuel the deuterium–tritium fusion reaction be produced?

For every neutron produced in this reaction, and in theory destined to burn the radioactive waste from the fission reaction, a tritium atom must be bred in a lithium (or lithium/lead) reactor blanket. The conventional, pure-fusion design demands a breeding ratio that is greater than one, to allow for losses and the decay of tritium; this might be extremely difficult to achieve even with essential neutron multipliers such as beryllium. The significant fraction of neutrons required to treat fissile material means that there could not possibly be enough to breed tritium as well.

At present, gram-scale quantities are available from Canadian sources, but the net production rate is only about 500 grams per year — a trivial amount compared with fusion-reactor requirements. For example, starting up any type of fusion reactor will need several months' supply of tritium just to bring the breeding process on stream. This issue is better addressed now than in 30 to 50 years' time.