The yellowcake that is produced as a result of mining and milling cannot directly be made into fuel pellets to be burned in a reactor. The reason is that natural uranium comes in different isotopes—99.3% is 238U and 0.7% is 235U (and to be completely accurate, a very small fraction is 234U)—but only 235U is useful for fis­sioning in a nuclear reactor (see Chapter 6). But it takes 3-4% of 235U to allow fission to efficiently occur in the most common types of reactors. That is a prob­lem, since the natural uranium in yellowcake is only 0.7% 235U. In order to get a sufficiently high concentration of 2 35U, the uranium has to be enriched in 2 35U compared to the 238U. How can that be done, since the chemical properties of all uranium isotopes are identical?

That was the big problem facing the scientists and General Groves of the Manhattan Project who were trying to build an atomic bomb. The only physical difference between the two isotopes is a small difference in mass—238 to 235 mass units—due to the 3 additional neutrons in 238U, and this turns out to be the secret to enrich for 235U. Several different approaches were used during the Manhattan Project, but the one that became the standard for uranium enrichment in the

United States was gaseous diffusion. To accomplish the enrichment of enough 235U to build a bomb (which requires about 90% enriched 235U), the most auto­mated and complex plant in the world at the time was built along the Clinch River west of Knoxville, Tennessee: Clinton Labs, now known as Oak Ridge National Laboratory. The building housing the gaseous diffusion enrichment columns was nearly half a mile long, one-fifth of a mile wide, and four stories high, covering 42.6 acres (2). The idea is that lighter molecules will diffuse through membranes with extremely tiny pores more rapidly than heavier molecules. Each enrichment column has a porous barrier that slightly enriches the lighter 235U isotope in one side and has depleted uranium on the other side. The slightly enriched uranium is then fed into another column where it is further enriched, and then another, and so on, for thousands of columns to get the desired enrichment.

Uranium first has to be turned into a gas before it can be separated by gaseous diffusion or other processes. Yellowcake (U3O8) is modified through a series of chemical reactions to form uranium hexafluoride (UF6), a very reactive substance which has the desirable property that it is a solid white powder at room tempera­ture but volatilizes into a gas at 56.5°C (134°F) (34). It is the gaseous phase that is used in the gaseous diffusion process. The slight difference (less than 1%) in mass between the two hexafluoride isotopes of uranium means that the lighter 235UF6 will diffuse slightly more rapidly through a porous barrier than the 238UF6. That makes separating the isotopes a very inefficient and energy-demanding process, and is the main source of CO2 production from nuclear power because coal-fired power plants are often used to power the gaseous diffusion plant. In France, nuclear power is used to drive the gaseous diffusion plant, so CO2 production is very small. The energy to run a gaseous diffusion facility accounts for about half of the cost of nuclear fuel and about 5% of the electricity generated from nuclear power plants (35). The United States has a single gaseous diffusion plant currently in operation in Paducah, Kentucky, which is licensed by the NRC to provide enriched uranium for nuclear power plants (36). It uses 1,760 enrichment stages in four buildings that cover 74 acres and has a power capacity of 3 GW to power the plant (37).

Gaseous diffusion was not the only enrichment method developed by scientists in the Manhattan Project, however. Another method that is also based on the mass difference in uranium hexafluoride isotopes is gas centrifuge enrichment. A large number of rotating cylinders (centrifuges) are connected in series, and gaseous UF6 is fed into them. The centrifuges rotate at extremely high velocities of 50,000-70,000 rpm and must be made to stringent tolerances, which is why it was not technically feasible to use these during the Manhattan Project. The heavier 238UF6 moves toward the walls by centrifugal force, while the lighter 235UF6 remains near the center and is bled off from one centrifuge and fed into the next centrifuge. Thousands of stages of centrifuges are used to get the desired enrich­ment. Gas centrifuge plants can be built in phases with more centrifuges added to increase the capacity. The huge advantage of this technology is that it is far more energy efficient, using only about 2% as much energy as a gaseous diffusion plant and therefore only generating 2% as much CO2 and costing far less.

This has been the enrichment technology of choice for Russia, Japan, and Europe (excluding France). There is one gas centrifuge plant in the United States in Eunice, New Mexico; two others were under construction but are currently on hold. Worldwide, gaseous diffusion accounted for 50% of uranium enrichment capacity in 2000 while gas centrifuge was 40%, with 10% from retiring nuclear weapons. By 2010, however, only 25% was done by gaseous diffusion while gas centrifuge accounted for 65%. By 2017 it is projected that nearly all enrichment will be done by gas centrifuge (35, 36).

It is this technology that is causing such heartburn for the Western countries because of the gas centrifuge plants built by Iran. Iran is already enriching ura­nium to the 3-4% necessary to provide fuel for a nuclear power plant as well as some to 20% for a medical research reactor (38). But there is nothing besides world pressure to stop the enrichment from going on to 90%, which then can be used in an atomic bomb. Even if Iran actually does enrich uranium sufficiently to make a bomb, it is likely to cause them more headaches than advantages. In his book Atomic Obsession, John Mueller argues forcefully that joining the semi-exclusive nuclear club does not give a country power over its neighbors but instead tends to focus the world’s attention on the country and shackle it with economic embar­goes, as in the cases of North Korea and Iran. A country may bluster and posture, but it is actually constrained because it knows that if it actually used a nuclear bomb on a neighboring country, it would be obliterated by the United States. “It seems overwhelmingly likely that, if Iran does develop nuclear weapons and bran­dishes them to intimidate others or to get its way, it will find that those threatened, rather than capitulating to its blandishments, will ally with others (including con­ceivably Israel) to stand up to the intimidation." (39)

A new generation of enrichment technology developed in Australia is known as global laser enrichment (GLE). It uses lasers to specifically ionize 235UF6 instead of 238UF6, creating a charged molecule that can then be separated electrostatically. This is the only separation technology that does not depend on the mass differ­ence between the two isotopes but instead depends on molecular energy bands that are different between the two types of uranium hexafluoride molecules. The laser can be tuned to the proper frequency so that only the 235UF6 will be ionized and hence separated. General Electric-Hitachi submitted a license application to the NRC to build a plant in the United States, which, if approved, could be opera­tional in 2014. This technology is more efficient than gas centrifuge enrichment, leading to lower energy inputs, lower CO2 emissions, and lower cost (35, 36).