Pu-239 is viewed as a desirable weapon constituent, while the military value of U-233 is problematic due to hard radiation from a very contaminant product associated with U-233 production in reactors. Only one U-233 weapons test was ever conducted by the United States, and that was not considered successful. In addition any deliverable U-233 weapon would require heavy shielding, making it extremely awkward for use by terrorists or by a military organization. Nations wishing to acquire nuclear weapons for military purposes have inevitably preferred Pu-239 or U-235 over U-233 although U-233 is quite easy to manufacture in low cost and technically unchallenging to build reactors. Pu-239 can be manufactured in such reactors and has always been preferred by weapons developers.
In the "Military Effects Test" of April 15, 1955 Weapons designers reportedly found that by substituting U-233 for U-235 in a standard test weapon, they lowered the weapon yield from 31 kt to 22 kt, a yield reduction of nearly a third.
A secret 1963 document, titled U-233 and prepared by Hanford weapons designer, A.E. Smith, set out Smith's concerns about U-233 weaponization. Smith indicated that in order to produce U-233 with low levels of U-232 contamination, a weapons material productions reactor at Hanford or Sevannah River would require three years. A test using high U-232 contaminated U-233 could be conducted more quickly, but it would involve problems that were not known, but which would probably involve "real sacrifices and risks to overcome." People assigned the task of purifying the U-233 would "take high radiation exposures." The U-233 would have to be purified to remove U-232 daughters, followed by rapid processing and assembly. A U-233 bomb test itself would cost between $7.5 and $15 million in 2011. Expected shelf life of a U-233 weapon would be 4 years, but disassemble and refurbishment would be more complicated and difficult than with a plutonium weapon.
Nations wishing to produce U-233 based nuclear weapons, would already have the resources to produce Pu-239 weapons, or would be capable of developing technology to produce U-235 weapons. U-235 or Pu-239 weapons would have superior military characteristics relative to U-233 based weapons. The potential militarization of U-233 seems unlikely to increase the risk of nuclear proliferation by nations, while the expenses, risks, costs and unknown challenges of militarizing U-233 would prove daunting to a terrorist organization.
U-235 was developed as a weapons material at the same time Pu-239 was. U-235 is extracted from Raw uranium which contains 99.3% U-238 and 0.7% U-235. It is possible to design quite simple weapon with U-235, and if a nation can design a low cost U-235 separation technology, U-235 is the royal road to nuclear power. South Africa proved, during the 1960's, 70's and 8Os's that developing a low cost U-235 separation technology was a relatively insignificant challenge for even a small country. Pakistan, a larger country, but one with even fewer industrial resources than South Africa, also developed U-235 separation technology at about the same time. Both nations used information that was readily available in other countries, as well as parts clandestinely obtained from other countries. Libya also undertook to obtain U-235 separation technology, and Iran has had developed a successful Uranium enrichment program prior to the appearance of the Stuxnet virus.
It should be assumed then that if a nation possesses a low cost, easily mastered route to nuclear weapons, the acquisition of a second, higher cost, less easily mastered route to nuclear weapons, will not increase the likelihood of nuclear proliferation. And given the choices of two routes, on difficult and one easy, nations can be expected to chose the easy cheap route over the more difficult and expensive path.
In addition to other fissionable materials, fairly small amounts of neptunium and americium can at least in theory can be used to power nuclear explosive devices according to the DoD. Reference to nuclear explosive devices rather than nuclear weapons suggest that the path to use of these materials in nuclear weapons.
Np-237 is produced by a double neutron capture conversion process which converts U-235 into Np-237. About 19% of U-235 fails to fission after a slow neutron capture, thus conversion of U-235 to NP-237 is fairly common. 19% of U-235 slow neutron capture leads to conversion rather than fission. In addition Np-237 is also produced via an (n,2n) reaction with U-238 after it encounters a high energy neutron. Np-237 is the product of alpha decay of Am-241.
In theory about 10 times as much Np-237 will be produced in a Uranium fuel cycle reactor than in a Thorium fuel cycle reactor. However, when NP-237 remains in a reactor for a significant period of time, it undergoes conversion to Pu-238 by the further capture of another neutron.
All uranium cycle reactors are capable of producing large amounts of plutonium will also produce Np-237. The ratio of plutonium to neptunium will be very high, over 20 to one, but enough Np-237 will be left in the nuclear waste to build a nuclear device every couple of years, if the nation who possess the nuclear waste. However, no nation is known to have built an Np-237 weapon, although the French reportedly may have done some research in that direction. However, no data has been published on Np-237 explosive devices or weapons designs, leaving the would be weapons designers to guess. Different statements on how much Np-237 is required to produce a nuclear explosion, with 60 kg asserted for a gun type device, and perhaps 30 kg for an implosion type device.
David Albright and Kimberly Kramer state,
Thus, large quantities of neptunium 237 are found in spent nuclear fuel. Each year, a typical 1,000-megawatt-electric light-water reactor produces about 25 tonnes of spent fuel containing about 10 kilograms of neptunium 237. The same spent fuel contains about 230 kilograms of plutonium. By weight, neptunium 237 discharges are about five percent of plutonium discharges and about 0.05 percent of spent-fuel discharges.It should be noted that both uranium fuel cycle and thorium fuel cycle MSRs produce some Np-237, with LFTRs can be expected to produce less than 10% of the Np-237 produced by conventional LWRs.
David Albright and Lauren Barbour note that as a nuclear explosive, americium would be considerably more problematic than Np-237.
The three most important isotopes are americium 241, americium 242m, and americium 243. All three have bare-sphere critical masses, but they vary widely and are uncertain. Americium 241 has a bare-sphere critical mass between about 60 to 100 kilograms. Americium 242m has the lowest bare-sphere critical mass, about 9 to 18 kilograms. The bare-sphere critical mass of americium 243 is somewhere between about 50 and 150 kilograms, indicating that public estimates of its critical mass vary enormously.Despite its explosive capacity. Americium has civilian uses, including in smoke detectors, as medical diagnostic tracers, and in neutron sources. In all of these uses very small amounts of americium are used. It would appear that would be proliferators who might chose americium weapons would have strong motives for preferring U-235 or Pu-239 weapons. Any state which possess the capacity to manufacture americium would also possess the capacity to produce weapons grade Pu-239. Weapons grade Pu-239 is far more predictable that americium, and thus far more desirable for weapons purposes.
Finally, in addition to Np-237 and americium, Protactinium-231 might also be mentioned as a further fissionable minor actinides. Pa-231 is very rare and can occur naturally as a U238 or U-235 decay product. Pa.231 is the longest lived Protactinium isotope, having a half life of 32,000 years. Small amounts of Pa-231 as well as Pa-233 would be produced in thorium cycle reactors. Even less is known about the explosive and military value of Pa-231, than is known about americium, and thus any weapons designer would be left to guess amount important details such as critical mass, and some authorities appear to deny that any amount of Pa-231 can explode. Once again, any state capable of building a Pa-231 weapon would also be capable of producing a Pu-239 weapon, and any sane weapons designer would prefer to use Pu-239.
It would be highly desirable to dispose of actinides in fast neutron reactors, and liquid sodium cooled reactors, either burners or breeders, would be very serviceable for this task, although fast MSRs would be equally serviceable, and might offer some advantages. Even thermal MSRs can be used for actinide disposal, although less energy would be extracted in the process than would be the case with fast reactors and other aspects of the process might turn out to be less efficient.