Concept for a simple "gun-type" nuclear bomb.

For a reliable military weapon that might be built by a state, a gun-type bomb should also have a neutron generator to set off a shower of neutrons to start the chain reaction going at the appropriate moment.[3] Terrorists who just wanted a good chance of some nuclear yield, and were not worried about reliability and predictability, might dispense with this somewhat challenging aspect and rely on the neutrons from spontaneous fission in the HEU to start the chain reaction. It is impossible to make a gun-type bomb with plutonium, because the rate of spontaneous fission, and the resulting neutron background, is so high that the chain reaction will start as the two pieces in the gun get close to each other, blowing the weapon apart before any significant yield results.

A gun-type weapon is highly inefficient (meaning that only a small fraction of the nuclear-explosive material used actually fissions) and so requires a substantial amount of nuclear material—in the range of a bare-sphere critical mass of HEU.[4] The Hiroshima bomb, which was a gun-type weapon, used approximately 60 kilograms of HEU metal.

An implosion type weapon uses a set of precisely shaped explosives arranged around a less-than-critical mass of nuclear explosive material to compress the mass to a smaller volume and higher density. When the atoms are closer together, every neutron has a higher chance of hitting another nucleus before leaking out of the nuclear material—and a mass that was subcritical at normal density becomes supercritical after sufficient compression.

In such an implosion, the transition from a less-than-critical mass to a more-than-critical mass happens very rapidly, making it possible to use plutonium in an implosion bomb despite its high rate of spontaneous neutron production. HEU can also be used in implosion weapons, and although it does not require this more sophisticated approach it does benefit from it in the form of a reduced critical mass.

The Trinity and Nagasaki implosion bombs involved explosive "lenses" arranged around a 6 kilogram sphere of plutonium metal, with detonators arranged all around the sphere so that the explosives were set off from every side at the same time, creating a spherical shock wave moving inward that crushed the sphere to a much smaller (and supercritical) volume. In this approach, if the explosives on one side go off much before the explosives on the other side, the nuclear material will be flattened rather than crushed to a smaller sphere, and no nuclear explosion will result. In addition, an implosion device using either weapon-grade plutonium or HEU requires a means for generating a burst of neutrons to start the chain reaction at the right moment, before the material "bounces" back to a subcritical state

Solving these technical challenges of implosion weapons was a major part of the Manhattan Project effort at Los Alamos during World War II. It had never been done before, and the whole approach had to be invented from scratch.[5] Getting implosion right remains a significant technical challenge today for anyone who is not a nuclear-weapons expert, even though explosive lenses and other shaped explosive charges are now in wide use for conventional military and even commercial applications.

Implosion devices are significantly more efficient than gun-type devices. The amount of HEU needed for a basic solid-sphere implosion device like the Nagasaki bomb is less than half the amount needed for a gun-type bomb—and still higher efficiency can be achieved with more sophisticated designs (such as hollow spheres, used in many modern nuclear weapons, and designs using faster explosives than were available in the 1940s).

Who could make such a bomb? Most states would be able to make at least a crude nuclear explosive, with a yield in the range of the Hiroshima bomb—either a gun type or an implosion type—if they had enough plutonium or HEU. States such as Iraq or North Korea—and still more, Iran, which has a substantially larger technical and industrial base—would certainly be able to. Indeed, a senior participant in Iraq's nuclear weapons program has estimated that Iraq could fabricate HEU components for its implosion design and build an HEU bomb within two months if it got stolen HEU from abroad; if the stolen material was plutonium, the existing design would have to be modified, but, he estimated, Iraq could still make a bomb in less than a year.[6]

A more difficult question is whether a terrorist group could make a nuclear bomb if it got the requisite nuclear material. Unfortunately, for a gun-type bomb, there can be little debate—it is quite plausible that a resourceful and well-organized terrorist group such as al Qaeda would be able to make at least a crude gun-type device, with a yield perhaps as large as the bomb that obliterated Hiroshima. An implosion-type device, as would be necessary for plutonium, or if the terrorists had not managed to acquire enough HEU for a gun-type explosive, would be a much more substantial challenge—but here, too, the possibility cannot be excluded.

Detailed examinations by U.S. nuclear weapons experts have concluded again and again that, with enough nuclear material in hand, it is plausible that a sophisticated terrorist group could build at least a crude nuclear explosive of either type. These conclusions were drawn before September 11 demonstrated the sophistication and careful planning and intelligence gathering of which Al Qaeda is capable.[7] Setting off a nuclear explosion with HEU is simple enough that DOE internal security regulations envision the possibility of an "improvised nuclear device"—a nuclear bomb the terrorists might be able to put together while they were still inside the facility where they stole the HEU.[8]

For a terrorist group to make a nuclear bomb, it would be helpful to have help from individuals with experience in the design and construction of nuclear weapons. But this would not be essential, particularly in the case of a gun-type bomb. A group of U.S. nuclear weapons designers has estimated that a team of 3-4 people, or possibly more, with some knowledge of physics, machining, explosives, and the chemical and physical properties of the nuclear material to be used—all of which is available in the unclassified literature, but not all of which would be easy for a terrorist group to put together—could potentially accomplish the job of building a nuclear bomb from stolen nuclear material, without anyone on the team with prior nuclear weapons experience.[9]

Historically, however, the effort that has been required for states to design and manufacture either gun-type or implosion-type weapons has been very much larger—typically dozens to hundreds of scientists and technicians, and years of effort. In significant part, this is because of the substantial difference between a nuclear weapon that a state would be willing to include in its military arsenal and one that might be satisfactory to a terrorist group. For a military weapon, it is important that the explosive yield be reasonably predictable, and that there be very, very high confidence the bomb will not go off accidentally. Simply propelling one slab of HEU into another and hoping for the best would not be satisfactory. Terrorists, on the other hand, might be quite satisfied with a bomb they believed would go off in some substantial way even if they had no idea how big a yield it would have, and might be willing to tolerate a substantial chance that it would go off accidentally. These much lower requirements for safety and predictability mean a significantly lower technical challenge.

Weapon-Related Properties of Specific Nuclear-Explosive Nuclides and Mixtures

There are only a few nuclear-explosive nuclides among the hundreds of nuclides found in nature and the thousands producible by technology. Two isotopes of uranium have this property (U-233 and U-235), as do all of the isotopes of plutonium (most importantly Pu-239, Pu-240, Pu-241, and Pu-242). More obscure nuclear-explosive nuclides include neptunium-237 and americium-241; these are less important than the indicated isotopes of uranium and plutonium because they exist today in much smaller quantities and because producing them in quantity is more difficult.[10]

Table 1 gives the approximate bare-sphere critical mass for the metallic form at normal density, the radioactive half-life, the rate of heat generation by radioactive decay, and the rate of generation of neutrons by spontaneous fission for the most important nuclear-explosive nuclides. The subsections that follow provide additional information relating to the usability of these nuclides, and commonly encountered materials containing them, in nuclear weapons of various types.

Table 1: Properties of Nuclear-Explosive Nuclides

Notes:

The critical masses given are for a bare sphere of metal at normal density. Plutonium metal can exist in six different forms corresponding to different crystalline configurations, with different densities. The two of these that are most germane for nuclear weapons are alpha phase (density 19.6 grams per cubic centimeter) and delta phase (density 15.7 grams per cubic centimeter). The indicated critical masses are for alpha-phase plutonium. For delta-phase plutonium, the critical masses would be about 60 percent larger.

Uranium Enriched in U-235

Uranium-235 is the only nuclear-explosive nuclide that occurs naturally in significant quantities. It constitutes 0.7 percent of natural uranium, but its nuclear-explosive properties emerge only if its proportion of the uranium atoms present is much higher than in the natural element. Uranium with more than the natural concentration of U-235 is termed "enriched", and the process of producing such uranium is called uranium enrichment (discussed below in the section on Production of Nuclear Materials).

While nuclear explosives can in principle be made with material containing somewhat less than 20 percent U-235, the amount of material required increases rapidly as the U-235 concentration falls below that level. (The critical mass for uranium with 20-percent U-235 surrounded by a 4-centimeter thick beryllium neutron reflector is over 400 kilograms; for uranium with 15-percent U-235, the corresponding figure is well over 1,000 kilograms.) In international practice, all uranium with a concentration of U-235 of 20 percent or more is referred to as "highly-enriched uranium" (HEU) and is subject to special safeguards measures. For fission explosives, nuclear-weapon designers prefer a U-235 fraction over 90 percent, and HEU in this concentration range is called "weapon-grade".

Weapon-grade HEU is a versatile nuclear-weapon material. As described above, it lends itself to use in simple "gun type" designs, as well as in the more efficient but also technically more demanding implosion designs that are needed to make a nuclear weapon with plutonium. The 60 kilograms of weapon-grade HEU metal used in the gun-type bomb dropped by the United States on Hiroshima would fit easily in a one-gallon container. An implosion weapon made with weapon-grade HEU might use only 12 to 15 kilograms.[11] This quantity of HEU metal would fit in a one-quart milk carton.

Weapon-grade HEU has the further advantage, for weapon-makers, of being only weakly radioactive, resulting in negligible heat generation (easing weapon design and extending shelf life), minimal radiation hazards to those working with the material and handling weapons made from it, and high difficulty of detection from its gamma-ray and neutron emissions. Table 2 illustrates these points by comparing heat generation rates, neutron emissions, gamma-ray doses, and an index of radiological hazard among natural uranium, low-enriched uranium typical of commercial reactor fuel, weapon-grade uranium, and reactor-grade and weapon-grade plutonium. The gamma dose rates and radiological hazards from uranium at all levels of enrichment in U-235 are so low as to make radiation exposure a negligible consideration for anyone stealing it or trying to make a weapon from it.

Table 2: Heat, Radioactivity, and Radiation from Various Nuclear Materials

Notes:

Compositions (percentage by weight) of the indicated materials are assumed to be as follows:

• Natural U = 99.275% U-238, 0.7193% U-235, 0.0057% U-234
• LEU (low-enriched uranium) = 96.475% U-238, 3.5% U-235, 0.025% U-234
• HEU (high-enriched uranium) = 5.88% U-238, 94.0% U-235, 0.12% U-234
• Weapon Pu = 0.01% Pu-238, 93.8% Pu-239, 5.8% Pu-240, 0.13% Pu-241, 0.02% Pu-242, 0.22% Am-241
• Reactor Pu = 1.3% Pu-238, 60.3% Pu-239, 24.3% Pu-240, 5.6% Pu-241, 5.0% Pu-242, 3.5% Am-241

Abbreviations: Ci = curie, g = gram, kg = kilogram, m³ = cubic meter, n = neutrons, W = watt. The gamma-ray dose is calculated at the surface of a sphere of the metal with a mass of a few kilograms. The radiological hazard index is the volume of water that a gram of the material could contaminate to the drinking-water standard.

Indeed, because of the suitability of weapon-grade HEU for use in simple and robust bomb designs that would not require nuclear-explosive testing, and because of the ease of handling and concealing it and weapons made from it, it is by far the most dangerous of the nuclear-explosive materials that society has produced, in relation to the threat of theft and use by terrorists or proliferant nations with limited access to nuclear-weapon expertise. On the other hand, from the technical standpoint HEU is quite easy to "denature"—by simple dilution with U-238 obtained from natural uranium or uranium-enrichment "tails" – so that it cannot then be used as a nuclear explosive except after technically demanding and costly re-enrichment.[12]

Plutonium

Plutonium is virtually nonexistent in nature, but it can be produced by exposing uranium-238 to neutrons in a nuclear reactor. Indeed, it is produced automatically in any nuclear reactor containing U-238 in its fuel, because the U-238 cannot be prevented from absorbing spare neutrons from the chain reaction and transforming itself into plutonium as a result.[13] (More is said about this below under the heading of Production of Nuclear Explosive Materials).

When plutonium is produced in a reactor, it is, by the nature of the process, intimately mixed with fission products as well as with uranium-238, and in that form it cannot be used to make a nuclear weapon. It must first be separated from the fission products and the uranium. The term "separated plutonium" is used when this has been accomplished to a degree such that the plutonium will support a nuclear explosion.

This separation is effected by chemical means (discussed further elsewhere here), which is possible because the plutonium has different chemical behavior than the other elements with which it is mixed. The fact that the separation process can be chemical rather than based on isotopic masses makes it technically easier, in principle, than uranium enrichment is. The very high radioactivity of the fission products is an added complication in plutonium separation, however, that has no counterpart in uranium enrichment.

Although all plutonium isotopes can support a fast-fission chain reaction and virtually any combination of them is therefore usable for making a nuclear weapon,[14] nuclear-weapon designers prefer to work with plutonium containing more than 90 percent Pu-239 (which is accordingly called "weapon-grade") This high Pu-239 concentration is generally achieved by removing the plutonium from the reactor where it has been produced before the higher isotopes (which result from successive neutron absorptions) have a chance to build up.[15] If a reactor is being used primarily for electricity generation, the longer refueling intervals optimum for this purpose generally result in plutonium with higher concentrations of Pu-240, Pu-241, and Pu-242 and only 60 to 70 percent Pu-239. This is called "reactor-grade" plutonium.

As noted earlier, plutonium can not be used as the nuclear explosive in a gun-type weapon, no matter what the plutonium's isotopic composition, because of its high rate of neutron emission. The implosion alternative to a gun-type design overcomes this "pre-initiation" liability of plutonium by achieving a much faster transition from sub-criticality to super-criticality in the nuclear fuel.

The key point that is not always appreciated is that this implosion approach can be effective enough at overcoming the problem of a high spontaneous rate of neutron generation, to cope with even the extremely high neutron-production rates associated with Pu-238, Pu-240, and Pu-242. The consequence is not simply that plutonium of virtually any isotopic composition can be used in implosion weapons. The fact is that, with sufficient sophistication in design and manufacturing, the less desirable mixtures of isotopes (such as the mixture in "reactor grade" plutonium) can be used to make nuclear weapons with performance very similar to what is achievable with "weapon-grade" plutonium.

The difficulties that must be overcome (or suffered) in using "reactor-grade" plutonium in weapons include not only the pre-initiation tendency arising from the high neutron-generation rate, but also the reactor-grade material's higher rate of heat generation (which may affect the stability and performance of the weapon's components) and its increased radiation levels (which result in greater dangers to those fabricating and handling weapons produced from reactor-grade plutonium, as well as greater difficulty of smuggling such weapons past detectors).

The clearest official unclassified statement to date about the extent to which the problems of reactor-grade plutonium affecting weapon performance can be overcome was issued in 1997 by the Department of Energy (which runs the US nuclear-weapon program). The DOE said:[16]

The degree to which these obstacles [to using reactor-grade plutonium in weapons] can be overcome depends on the sophistication of the state or group attempting to produce a nuclear weapon. At the lowest level of sophistication, a potential proliferating state or subnational group using designs and technologies no more sophisticated than those used in first-generation nuclear weapons could build a nuclear weapon from reactor-grade plutonium that would have an assured, reliable yield of one or a few kilotons (and a probable yield significantly higher than that). At the other end of the spectrum, advanced nuclear weapon states such as the United States and Russia, using modern designs, could produce weapons from reactor-grade plutonium having reliable explosive yields, weight, and other characteristics generally comparable to those of weapons made from weapons-grade plutonium. ...Proliferating states using designs of intermediate sophistication could produce weapons with assured yields substantially higher than the kiloton range possible with a simple, first-generation nuclear device.

Obfuscation on this point, which continues in some debates concerning recycle of plutonium for civilian nuclear energy generation, is irresponsible and dangerous. Separated plutonium of any isotopic composition other than nearly pure Pu-238 ought to be referred to as "weapon-usable" material, rather than making the potentially misleading distinction between "weapon-grade" and "reactor-grade". Depending on design and procedures, weapons made with reactor-grade plutonium might suffer some penalty in shelf-life and in elevated genetic and cancer risks for those manufacturing and handling them, but such shortcomings would be unlikely to discourage either countries bent on proliferation or terrorists.

The physical size of a "bomb quantity" of plutonium is astonishingly small. The 6 kilograms of weapon-grade plutonium used in the core of the implosion bomb dropped on Nagasaki would fit in a soda can.[17] Various authoritative sources state 4 kilograms as a more typical quantity in modern plutonium weapons. Even if the material is "reactor-grade" rather than "weapon-grade" plutonium, 5 or 6 kilograms is enough.

To these unsettling characteristics of plutonium, one must add the fact that there is no plutonium analog to the technically easy isotopic "denaturing" that is available for HEU. That is because, as noted earlier, all mixtures of plutonium isotopes remain nuclear explosives. And while it is true that a very high Pu-238 fraction in such a mixture might make its use as an explosive impractical in an engineering sense even if theoretically possible, studies of the matter indicate that it does not seem practical to produce enough Pu-238 to denature a significant fraction of the existing separated plutonium in this way.

The best available way to render separated plutonium unusable in nuclear weapons is to "un-separate" it by re-mixing it with fission products. Re-mixing it just with uranium-238 would not be as good, because re-separating it chemically again is too easy without the complication of the intense radioactivity of the fission products. The re-mixing with fission products to a suitable degree of difficulty for reversing the process could be achieved by embodying the plutonium in mixed-oxide (MOX) reactor fuel and then using that fuel in a power-generating reactor, or by mixing the plutonium with fission products that already exist in storage from prior military or civilian processing of spent nuclear fuel.[18] Both approaches are much costlier and more difficult than the isotopic denaturing of HEU with U-238, and the reprocessing technology needed to reverse them is probably not as demanding as the technology needed to re-enrich uranium. (More about this later.)

As described earlier, however, making an implosion bomb, as would be necessary to make a bomb from plutonium, requires considerably more sophistication from the designers and more trouble and skill from the builders than making a gun-type bomb from HEU. Plutonium also has a much higher rate of heat generation than does HEU, and plutonium metal itself is much more hazardous radiologically—and in other ways more difficult to work with—than HEU is.[19] The difficulties of heat generation and radiological hazard are much larger, though by no means insurmountable, for reactor-grade plutonium as compared to weapon-grade plutonium.

Uranium-233

The nuclear-explosive nuclide uranium-233 is, like plutonium, essentially non-existent in nature. Also like plutonium, however, it can be produced by exposing a common, naturally occurring nuclide to a flow of neutrons in a nuclear reactor or an accelerator. The "fertile" nuclide in this case—which, upon absorbing a neutron, undergoes two subsequent nuclear transformations to become U-233—is thorium-232. As noted above, U-238 is the analogous fertile nuclide for the production of Pu-239. The thorium-232 isotope constitutes 100 percent of naturally occurring thorium, which is about three times as abundant in the Earth's crust as uranium is.

The properties of U-233 as a nuclear explosive would seem to make it quite attractive to bomb-makers. Its bare-sphere critical mass is very similar to that of delta-phase plutonium-239, which is to say it's only 50 percent greater than that of alpha-phase plutonium-239 and about one third that of uranium-235. (Its greater reactivity compared to that of U-235 is reflected in the fact that a mixture of only 12 percent U-233 in U-238 is considered weapon usable, as opposed to the corresponding figure of 20 percent for U-235.) The rate of neutron generation from spontaneous fission in U-233 is low enough, moreover, to permit its use in the simpler gun-type weapon designs as well as in implosion designs; and its decay heat is one sixth that of Pu-239. Its gamma-ray dose rate is in the same range as that from weapon-grade plutonium.

Notwithstanding these apparent attractions, no country has yet produced U-233 for nuclear weapons as far as is publicly known. Neither has U-233 been produced in significant quantities in civilian nuclear-energy operations (although its use as the fissile component in a "thorium fuel cycle" has been much analyzed and discussed). There are two main reasons that the U-233 option has not been exercised so far:

• First, it's much easier to produce plutonium-239 than U-233 in reactors fueled by natural uranium, which are the kinds of reactors with which countries seeking weapons tend to start. If a country seeking weapons has the capacity to make enriched uranium (which is needed to fuel the type of reactor best suited for making U-233), that country's easiest route to bombs is to use that capacity instead to enrich natural uranium to weapon-grade in U-235.
• Second, U-233 is invariably accompanied by at least a small admixture of U-232, whose radioactive-decay chain contains a powerful emitter of a very penetrating gamma ray (thallium-208). Even at an initial U-232 concentration as low as 1 part per million, the gamma dose-rate after two years of build-up of its decay chain is twice as great as the gamma dose-rate from correspondingly aged reactor-grade plutonium.[20] This would pose significant problems in the form of radiation exposure to workers in a weapon program (or a nuclear-power program) using U-233, and it would also make weapons containing U-233 quite difficult to conceal.

These reasons continue to make U-233 production an unlikely path for new proliferators to try to take to a bomb, and the fact that no significant stocks of this material appear to exist anywhere in the world mean it is also not a theft risk.

How Big Is a Nuclear Bomb? How Could It Be Delivered?

The physical size and weight of nuclear weapons of different designs—from crude to sophisticated—are important determinants of how easy or difficult it is to handle these weapons, to hide them, and to transport (or smuggle) them. These characteristics also determine what means are practical for delivering nuclear weapons to their targets.

Size and Weight

The exact dimensions of currently deployed nuclear weapons are classified, but authoritative information is available about weapons from earlier eras and plausible estimates are available for more current designs.[21] In all instances, the dimensions of nuclear weapons greatly exceed those of the nuclear-explosive material they contain, because the complete weapon must also incorporate neutron-reflecting material around the nuclear explosive, chemical high-explosive, electronics for fusing and detonation, a casing, and so on.

The gun-type uranium bomb dropped by the United States on Hiroshima in August 1945 was about 10 feet long and 2.5 feet in diameter, weighed some 8,000 pounds, and contained about 60 kilograms of weapon-grade HEU; it yielded between 12 and 15 kilotons. The plutonium-implosion bomb dropped on Nagasaki a few days later was egg-shaped, about 5 feet in diameter, weighed about 10,000 pounds, and contained about 6 kilograms of weapon-grade plutonium; it yielded about 20 kilotons. By 1951, the United States had been able to reduce the weight of a plutonium-implosion bomb in this yield range to under 2,000 pounds.

It subsequently proved possible to make nuclear weapons considerably smaller. Indeed, the United States deployed nuclear artillery shells for cannons as small as the 155 millimeter howitzer; that warhead cannot have exceeded 155 millimeters in diameter—about 6 inches. It is reported to have weighed about 100 pounds and to have had a yield around a tenth of a kiloton. The lightest nuclear weapon the United States ever deployed was the W54 warhead, which in its "Davy Crockett" version weighed about 50 pounds and reportedly had a yield in the range of a quarter of a kiloton.

Unclassified "model" calculations carried out in 1990 to estimate the composition and weight of relatively simple implosion designs based on plutonium and HEU came up with diameters in the range of 18 inches and weights in the range of 300-400 pounds.[22] Estimates of the weights of modern thermonuclear warheads for US ballistic missiles and aircraft-delivered bombs fall in the range of 400 to 800 pounds, and these weights necessarily include not only the fission "primary" but also the separate thermonuclear "secondary" and the casing that connects the two.²³

Cou