Until 1942, bombardment of targets composed of uranium compounds with deuterons produced by a cyclotron was the only source of plutonium. In December 1942, Fermi’s controlled chain reaction at the University of Chicago produced tiny quantities of plutonium in addition to energy and fission products. In November 1943, a larger pilot reactor at Oak Ridge, Tennessee, known as X-10, went critical and began to produce plutonium. Fuel rods consisting of natural uranium were inserted into the reactor for a given period of time. At the end of this period, the rods were removed and allowed to cool. Plutonium, at a concentration of approximately 250 parts per million. was found mixed with quantities of fission products and unreacted uranium. In 1945, the first of three large-scale reactors began to produce plutonium at Hanford, Washington.

X-10 graphite reactor at Oak Ridge
(Courtesy of the Department of Energy)

The U-238 in natural uranium captures neutrons to produce Pu-239, while the U-235 sustains the chain reaction required to produce more neutrons. Heavier plutonium isotopes require additional neutron capture and thus increase more slowly than Pu-239. Uranium that has been in the reactor for a short period of time has a significant amount of Pu-239 relative to the heavier Pu isotopes. However, Pu-239 is fissile and as its concentration increases in the reactor, it also begins to undergo fission. Thus, the uranium-containing fuel rods have to be pushed out of the reactor after several days to maximize the yield of Pu-239.

The next problem was the quantitative recovery of the small amounts of plutonium from large amounts of uranium and fission products. These radioactive products of U-235 fission had to be reduced to a concentration of less than one part in 10⁷ parts plutonium. This step was required to reduce the gamma radiation from the fission products to make the plutonium safer to handle and to remove neutron-absorbing impurities.

Because of the low concentration of plutonium, plutonium compounds could not be precipitated directly. Thus, any separation process involving precipitation had to be based on coprecipitation techniques that employed carriers for plutonium. The bismuth phosphate process was first used for large-scale separation of plutonium from uranium and the fission products. The keys to this process were the quantitative, selective coprecipitation of Pu⁴⁺ from an acid solution by a bismuth phosphate carrier and the ability of Pu⁶⁺ to remain in solution in the presence of the bismuth phosphate carrier.

(Courtesy of the Department of Energy)

The neutron-irradiated, aluminum-clad fuel rods were removed from the reactor to a pool of water, where they remained for a number of days to allow the short-lived, high-activity fission products to decay. The aluminum casing was removed from the rods with a concentrated sodium hydroxide solution. The contents of the fuel rods were dissolved in nitric acid and, after the addition of sulfuric acid to keep the uranium and other fission products in solution, Pu⁴⁺ was coprecipitated with bismuth phosphate. The precipitate containing the plutonium was then dissolved in nitric acid and the Pu⁴⁺ was oxidized to Pu⁶⁺ with sodium dichromate. This time, the plutonium remained in solution as Pu⁶⁺ while any residual uranium and fission products were precipitated with additional bismuth phosphate. The Pu⁶⁺ was then reduced to Pu⁴⁺ and the cycle was repeated. At this point, the carrier was changed to lanthanum fluoride (LaF₃) and a similar oxidation-reduction cycle was performed to achieve further purification and concentration of the plutonium.

This radiochemical processing facility, also known as "Queen Mary", processed irradiated fuel from the B- and D- reactors using the bismuth-phosphate process to produce plutonium nitrate, which was shipped to Los Alamos for the Trinity test bomb and the Nagasaki bomb.

The concentration of plutonium was high enough so that no carrier was required for the final purification step, the precipitation of plutonium peroxide from a basic solution. The peroxide was converted to a plutonium nitrate paste that was shipped to Los Alamos. The overall recovery of plutonium by this process was greater than 95% and the plutonium contained less than 1 part impurity in 10⁷ parts plutonium. The process generated large amounts of chemical and radioactive wastes (approximately 10,000 gallons per metric ton of uranium processed). A major fault, however, was that only plutonium could be extracted; the uranium remained in the waste.

By the 1960s, the plutonium/uranium/recovery/extraction (PUREX) process replaced the bismuth phosphate process. It was capable of recovering plutonium, uranium, and other materials with a continuous extraction using organic solvents. After dissolution in nitric acid, the plutonium and uranium nitrates were transferred into the organic phase, while the fission products were removed in the aqueous phase. This process also generated less waste than the bismuth phosphate process.

Complete Bibliography on Plutonium from the ALSOS Digital Library for Nuclear Issues