On Enriched Uranium and (Iranian) Nuclear Bombs

News accounts dealing with how much uranium Iran might possess of a composition close to what is needed for making a nuclear bomb―and, relatedly, how quickly Iran might be able to produce what it would need for one such bomb or more―can be confusing to those not deeply familiar with the relevant physics and technology. This essay seeks to clarify the matter for those willing to put up with a bit of arithmetic.

On Uranium Fission

Natural uranium—the form of the element found in nature―is composed of two isotopes¹: U²³⁸ (making up 99.3 percent of natural uranium) and U²³⁵ (0.7 percent). Because the two isotopes have the same number of protons and, thus, the same number of electrons surrounding their nuclei, their chemical behavior is essentially identical.  But the difference in their numbers of neutrons―143 for U²³⁵ and 146 for U²³⁸―means not only that the two isotopes differ slightly in mass but also that they differ significantly in how they behave in nuclear reactions.

Specifically, the U²³⁵ nucleus is quite unstable in terms of how readily it fissions when struck by a free neutron, breaking into two large fragments of intermediate atomic weight (“fission products”) and releasing two or three more neutrons and large quantities of nuclear energy.  The result is that, in suitable circumstances, U²³⁵ can sustain a nuclear chain reaction. (A chain reaction requires that an average of at least one of the neutrons released from the fission of one U²³⁵ atom induces another fission in a second U²³⁵ atom, and so on.)

If the concentration of U²³⁵ atoms is high enough, and if the accompanying concentrations of U²³⁸ and other elements that absorb neutrons without fissioning are not too high, U²³⁵ can sustain a chain reaction. It has this capability, moreover, whether the neutrons impinging on it are “slow” (with velocity close to that corresponding to the temperature of the environment where the reactions are taking place) or “fast” (with velocity closer to the much higher speed the neutrons carried upon emission).

By contrast, U²³⁸ has negligible probability of fissioning when struck by a slow neutron; and, while it may fission when struck by a fast neutron, the likelihood that an incident fast neutron will be captured without inducing fission is too high for U²³⁸ to sustain a chain reaction on its own, even in a fast-neutron environment.

Despite its inability to sustain a nuclear chain reaction by itself, U²³⁸ is not useless in nuclear-reactor and nuclear-weapon applications. U²³⁸ fission by fast neutrons  adds something to the total energy produced in both nuclear reactors and nuclear bombs. And non-fission absorption of slow or fast neutrons by U²³⁸ produces unstable U²³⁹, which, in a reactor, transforms itself in two steps, over hours and days, into useful plutonium-239 (Pu²³⁹).  

Once created, some of the Pu²³⁹ (and other plutonium isotopes produced by additional neutron absorption) undergo fission in the reactor, thus contributing to its nuclear-energy production. Those plutonium isotopes that have not yet undergone fission remain in the “spent” nuclear fuel after it is removed from the reactor. This plutonium can be separated in a fuel-reprocessing plant from the fission products and unreacted uranium in the spent-fuel, for subsequent use either in nuclear reactors or nuclear weapons.²  

On Uranium Enrichment for Nuclear Reactors and Bombs

Nuclear reactors based primarily on the U²³⁵ chain reaction can use either slow or fast neutrons, depending on reactor design. Reactor types relying on slow neutrons (also termed “thermal” neutrons, with reactors relying on them likewise called “thermal” reactors), require incorporation of materials called “moderators” to slow down the fast neutrons emitted by fission. The best moderators are graphite and “heavy water” (deuterium oxide, D₂O), although ordinary water (“light water," H₂O) suffices in that role in the Light Water Reactors (LWRs) that dominate global nuclear-electricity generation today. Fast-neutron nuclear reactors and nuclear weapons (the latter depending entirely on fast neutrons) lack moderators.

Achieving a chain reaction in LWRs and most other reactor types³ requires the U²³⁵ fraction be increased beyond its 0.7% level in natural uranium, by means of “uranium enrichment.” The U²³⁵ concentration deemed optimal in today’s LWRs is around 5%, but some advanced reactor designs work best with U²³⁵ at a concentration approaching 20%.⁴ That happens to be the concentration at which making a nuclear explosive from uranium becomes theoretically possible (although hardly practical, because of the very large mass of 20%-enriched uranium required for an explosive chain reaction). To minimize the mass of enriched uranium needed for a fission bomb or warhead, a U²³⁵ concentration above 90% is preferred, with 93% often cited as ideal.

Practical approaches for uranium enrichment, whether for energy or weapon purposes, depend on selective removal of U²³⁸ by exploiting the very small mass difference between it and U²³⁵. Currently, the most cost-effective technology for achieving this separation consists of arrays of hundreds to thousands of very specialized, chained centrifuges, each one operating on gaseous uranium hexafluoride (UF₆). The rotors in the centrifuges spin at 50,000 rpm or more, driving the slightly heavier U²³⁸F₆ molecules more strongly toward the cylinder’s circumference, leaving the UF₆ near the axis slightly enriched in U²³⁵. 

Uranium gas centrifuges operate continuously rather than in the batch mode typical of centrifuges for other purposes, with the slightly enriched UF₆, extracted near the centrifuge axis and passed forward to the next centrifuge in the chain, while the slightly depleted UF₆ is extracted near the circumference and passed backward to the previous centrifuge. That the mass difference between the two uranium isotopes is so small means the degree of enrichment that can be accomplished in a single centrifuge is also small, which is why chaining multiple centrifuges together is necessary to achieve a useful degree of enrichment in the final product.

The magnitude of an enrichment task can be measured by:

1. the amount of unenriched or low-enriched uranium feed required to obtain the desired quantity of uranium product at a specified higher level of enrichment;
2. the amount of “separative work” that the laws of thermodynamics require for the degree of sorting of the heavy and light nuclei needed to provide the specified concentration of U²³⁵ in the desired quantity of product;
3. the number of centrifuges to produce the desired quantity in a specified time; and
4. the amount of electrical energy needed to power the production process.

We consider these factors in order, providing example cases where enrichment, starting from natural uranium, is to 5% U²³⁵ for use in a contemporary light-water power reactor (LWR) or to 93% U²³⁵ for use in a fission explosive.

The amount of uranium input required can be calculated from simple “balance” equations, requiring only the total quantities of U²³⁵ and U²³⁸ isotopes are conserved in the process. The answer depends only on the U²³⁵ concentration in the feed, the U²³⁵ concentration desired in the enriched product, and the residual concentration of U²³⁵ in the depleted-uranium waste stream (called the “tails”); it does not depend on the technological process used for enrichment, assuming only that material losses in the process are negligible. With that assumption, the balance equations reveal that the mass of the required uranium feed is larger than the mass of enriched product by the ratio, F/P = (xp – xt)/(xf – xt), where xp is the desired enrichment level of the product, xf is the enrichment level of the feed, and xt is that of the tails.⁵

The separative work that thermodynamics requires in order to produce a kilogram of product at enrichment level xp, from feed at enrichment level xf, assuming tails enrichment xt, also depends only on those concentrations, not on any details of the technology used. The measure of this quantity is the Separative Work Unit (SWU), for which the quite complicated and opaque formula need not trouble the reader here because an online SWU calculator is available at https://www.urenco.com/swu-calculator. Upon specification of xp, xf, and xt, it provides the needed quantity of SWU and the needed kilograms of feed associated with making a kilogram of product at enrichment xp. 

An important implication of the thermodynamics behind the SWU calculation is that separation gets easier the higher the U²³⁵ concentration in the feed. Thus, for example, if one has already produced some Low Enriched Uranium for use in a light-water power reactor and then decides to use that LEU as feed for a nuclear-weapon project, a surprising fraction of the total needed enrichment work has already been done.

Turning to the number of centrifuges involved, consider first the modern, commercial, centrifuge technologies in use in the United States and Europe. A typical centrifuge in this class―4 meters (13 feet) tall with a rotor diameter of 20 centimeters (8 inches)―can deliver 50 SWU per year. 

Modern commercial centrifuges are said to use 100 to 150 kilowatt-hours (kWh) of electricity per SWU. At the cost of electricity paid by industrial users in the United States, currently averaging about $0.10/kWh, the corresponding electricity cost per SWU would be $10-15.  

The Case of Iran

Prior to the June 2025 U.S. strikes on Iran’s uranium-enrichment facilities, that country was estimated to possess 20,000-22,000 centrifuges across multiple sites. Their centrifuges were of four different generations, all much less capable than the modern commercial centrifuges considered in the examples above. The capacities of the Iranian centrifuges were thought to range from 0.8-1 SWU/yr per centrifuge in the oldest generation to 6-10 SWU/yr per centrifuge in the newest. Iran’s total pre-strike enrichment capacity was estimated at 60,000-100,000 SWU/yr. 

That would not have been enough to support Iran’s single nuclear power reactor, which is a Russian-designed LWR at Bushehr with generating capacity around 1 million kilowatts, thus an enrichment need of around 150,000 SWU/yr. That reactor’s LEU is supplied (and removed after use) by Russia under the agreement through which the reactor was provided to Iran, however. Thus, it must be assumed that Iran’s development of its own enrichment capacity was intended to provide a nuclear-weapon option, a hedge against cutoff of the LEU supply for Bushehr from Russia, a starting point for a capability to fuel future power reactors domestically, or (most likely) a combination of all these rationales. 

At the lower limit of Iran’s estimated enrichment capacity before the 2025 U.S. strikes, Iran would have had the theoretical capacity to enrich, per year, the uranium for about six gun-type bombs using 50 kg each of HEU at 93% U²³⁵ (60,000 SWU/yr / 10,000 SWU/bomb). If Iran’s weapon designers had mastered implosion-type designs, the number of nuclear weapons theoretically producible using an enrichment capacity of 60,000 SWU/yr would have been in the range of twice as great. 

Information leaked from a confidential 2025 report by the IAEA indicates that the highest level of uranium enrichment actually produced by Iran prior to the strikes was 60% U²³⁵ and that Iran’s pre-strike inventory at this concentration was 440 kilograms. The online enrichment calculator indicates that enrichment to 60% U²³⁵ starting from natural uranium needs 125 SWU per kilogram of product, assuming 0.3% U²³⁵ in the tails. Thus, the 440 kg at 60% enrichment represents a past investment of 440 kg x 125 SWU/kg = 55,000 SWU.  

Enrichment at 60% U²³⁵ is well into the HEU range and is, in fact, sufficient to make quite powerful nuclear explosives without any investment in further enrichment. Doing so would entail either significantly lower yields than attainable with 93% U²³⁵ or, to get the same yields, a significant increase in the mass of the uranium charge. Unclassified calculations have suggested that simply substituting 60% U²³⁵ for 93% U²³⁵, kilogram for kilogram, in a relatively simple implosion-weapon design developed for the latter enrichment level would reduce the explosive yield by a factor of about four—from 10-15 kilotons in the 93% U²³⁵ case to 2.5-4 kilotons with the same mass of 60% U²³⁵. (That is still enough yield to cause massive devastation.)  

Alternatively, increasing the mass of 60% U²³⁵ 4-fold above the 20-25 kg 93% U²³⁵ in the relatively simple implosion design could, with modest other changes, give the same 10-15 kiloton yield as the 93% U²³⁵ weapon. The unfortunate conclusion is that Iran’s original 440 kilograms of 60% U²³⁵, if they still had access to it, could yield 4-5 weapons with yields of 10-15 kilotons each, albeit too heavy to be delivered by missile.

If, on the other hand, the Iranians wished to invest in the pursuit of more compact, lighter weapons—and if they still had or could rebuild a very modest fraction of their pre-strike number of centrifuges―they could subject their 60% U²³⁵ inventory to further enrichment. 

So where does Iran stand today in terms of the time it would need to acquire HEU of the quality needed for a weapon? The answer is that it’s impossible to say, not least because we don’t know (at least not publicly) whether any of Iran’s pre-2025-strike inventory of 60% U²³⁵ is still easily accessible…or how of what’s buried in rubble might be recoverable in a short time. Nor is it publicly known how many Iranian centrifuges survived the 2025 strikes and any subsequent strikes on nuclear facilities in the currently ongoing war, either intact or in repairable shape. The few public statements by U.S. officials familiar with classified intelligence estimates leave open the possibility that Iran could have its hands on a weapon-relevant quantity of 60% U²³⁵ in a month or less, as well as the possibility that Iran could have access now or soon to perhaps hundreds of operable centrifuges. 

Three bottom lines:

1. If Iran already has or could fairly quickly acquire access to as much as a quarter of their pre-2025-strike inventory of 440 kilograms of 60% U²³⁵, that would be enough, without further enrichment, to make a single 10-15 kiloton implosion- or gun-type nuclear weapon or 4-5 implosion-type weapons with about 25% of that yield.
2. If Iran already has or could fairly quickly acquire as much as 1% of their pre-2025-strike inventory of 20,000-22,000 centrifuges, that would give the country an enrichment capacity of 1,000-2,000 SWU/year. Even the lower number would suffice to yield about 70 kilograms of 93% U²³⁵ in 4.5 months, starting from a quarter of the pre-strike inventory of 60% U²³⁵. That would be sufficient for three 10-15 kiloton implosion-type weapons.
3. Respectable estimates of the amount of time Iran would need to fabricate such weapons, once the needed nuclear material was in hand, range from a month to a year. Iran’s incentives to make such efforts are surely greater now than they were before the U.S. and Israeli attacks on Iran that began on February 28, 2026. Whether events in the weeks and months ahead will alter either those incentives or the relevant capacities remains to be seen. 

Author's Note: I thank Matthew Bunn, Dan Poneman, and Steve Fetter for helpful suggestions on an earlier draft. The responsibility for any remaining errors or misjudgments is mine.