There's two other differences you'd need to account for: * One Pu-238 nucleus weighs 238 amu; a triton only weighs 3 amu * Tritium decays 7 times faster -- a

There's two other differences you'd need to account for: * One Pu-238 nucleus weighs 238 amu; a triton only weighs 3 amu * Tritium decays 7 times faster -- a 12.32 year half-life, vs. 87.7 years for Pu-238 Put together, the specific power (watts/kg) of pure tritium is about twice as high. That's chemically just hydrogen gas, which I'd guess isn't practical. The solids with the highest hydrogen density, like polyethylene ([C2H4]n), would be less power-dense, but not by much. But I think there's some slack here. The ESA is designing their future RTGs to use Americium-241, which is drastically worse (in terms of these figures), yet still usable.