A history of the world's most dangerous element

by Jeremy Bernstein

Excerpt of Chapter IX

Plutonium is generated in a two-step process that begins with the capture of a neutron by a uranium-238 nucleus. But there will also be some uranium fission going on. This means that fission fragments will be created, but there are also impurities in the reactor fuel, elements such as boron. Plutonium decays by the emission of alpha particles. When these alpha particles collide with an element like boron, neutrons are produced. The question Seaborg posed was, What concentration of these impurities could one toler- ate so that the neutrons produced in such collisions did not generate a fizzle? His calculations showed that the concentration of impurity would have to be reduced to something like one part in 100 billion. Both General Groves and Oppenheimer were so informed. To make things worse, the British had realized the same thing, but their calculations showed that the impurity concentration would have to be reduced by another factor of 10. A discussion ensued of how such purity could be achieved, and the conclusion was reached only with great difficulty. Before anything was decided, samples of reactor-produced plutonium began to arrive at Los Alamos. As I explain shortly, these samples raised a new problem, which was of such a serious character that it definitely ruled out a gun-assembly plutonium weapon. The impurity question turned out to be, more or less, irrelevant.

Meanwhile, William Zachariasen had begun the study of the structure of plutonium and its compounds that he would pursue throughout the war and, indeed, afterwards. In a report written in 1946, he noted, "For the past three years within the plutonium project, I carried out partial or complete crystal structure determinations of 140 different compounds of plutonium, neptunium, uranium, thorium or rare earth elements. My collaborator Dr. Rose Mooney made similar determinations of an additional 20 compounds." Nearly all of the plutonium compounds were studied using X-ray diffraction, and even as late as 2005, Zachariasen's work on the structures of the oxides of these elements in these three years represented more than half of the total output of everyone else. As he later noted, "I remember working like heil on New Year's Day and all holidays; often I worked late for many, many hours to get the work done. I had a wonderful time ...." One of Zachariasen's early discoveries was that plutonium has "allotropes." Allotropes are different crystal structures of the same element. The canonical example is carbon. Depending on how it has been treated, carbon can manifest itself, for example, as graphite or diamond. Allotropes are different from what we usually call "phases," which refer to whether the element is found in a liquid or a solid state, for example. Nonetheless, you will frequently find the term "phase" used for different allotropes. I also use it, from time to time, since I can't think of a better term.

The first two allotropes of plutonium that Zachariasen found were labeled with the Greek letters α and δ. Crystallographers label allotropes by Greek letters in order of the increasing temperatures at which the allotrope is question is stable. For plutonium, stability is a relative concept since it does not take much of a jar to cause an allotropic transformation. When he first discovered these allotropes, Zachariasen did not know there were four more. The full labeling is α, β, γ, δ, δ' and ε. We can worry about the rest later and concentrate here on α and δ. The first thing to emphasize is that an allotrope is not a property of a single atom. A plutonium atom is a plutonium atom is a plutonium atom. If you have seen one, you have seen them all. It is rather a property of the crystal structures that can be built out of these atoms. It is these structures that are, or are not, stable in a given temperature range.

Let's begin with the α-allotrope. It is stable up to a temperature of 122 °C. This means that it is stable at room temperature. Zachariasen used X-ray diffraction to find the structure of its unit cell. While crystals can be exceedingly complex and unique — snowflakes, for — example they are nonetheless built out of a limited number of unit cell types. In the case of snowflakes they are hexagonal. , which you The α-allotrope turned out to be "monoclinic" — a crystal structure in which all of the axes in the unit cell are not perpendicular to each other and may have different lengths. The 16-atom unit cell for α-plutonium. looks perversely complicated. Moreover, it has less symmetry and hence little plasticity or pliability. Thus if you tried to bend α-plutonium metal, it would break like a piece of chalk. It behaves more like a mineral than a metal. On the other hand, the δ-phase is quite something else. It is stable between 317 and 453 °C and has a nice symmetrie unit cell, what the crystallographers call a face-centered cubic . There are eight atoms in the corners and six in the center of each face, making 4 in all. We can imagine displacing this structure along a plane and preserving it. Indeed, δ-plutonium is as malleable as an ordinary metal, perfect for making into a bomb, except for the fact that at lower temperatures it readily morphs into the α-phase, presenting a much greater engineering challenge. The densities are interesting. At 25 °C, the α-phase density is 19.86 grams per cubic centimeterâ — very dense indeed — while at 320 °C, the δ-phase density is 15.92 grams per cubic centimeter. The strange results that Zachariasen first found for the densities are explained by the mixture of different phases. Clearly, if you intend to use metallic plutonium to make a bomb, you will be confronted with a very significant metallurgical challenge. But worse is to come.

DuPont had been contracted to construct the production reactors. After Seaborg pointed out the impurity problem, there was some reluctance to proceed. However, once General Groves had decided to do something, it was next to impossible to stand in the way. Thus, beginning in February 1943, construction started on a pilot project located near Clinton, Tennessee what later became Oak Ridge. It was designed to use the bismuth phosphate method of separation that had been developed at the Met Lab. The Oak Ridge reactor went critical in November and by April 1944, it was shipping grams of plutonium to Los Alamos; but it soon became clear that a disaster had occurred. To understand the issue let us review how plutonium is produced in a reactor.

The basic fuel in these reactors was natural uranium, more than 99 percent uranium-238, the rest being mainly the fissile isotope uranium-235. To enhance fission reactions, the neutrons created in fission are slowed down by a moderator in this case, highly purified graphite, the same moderator that Fermi had used in his reactor. But some of the neutrons are absorbed by uranium-238 nuclei, producing neptunium-239, which beta-decays to plutonium-239. To get a substantial yield of plutonium-239, the reactor must be allowed to run for a reasonable amount of time. The longer the reactor is allowed to run before plutonium is separated from uranium, the more plutonium you get. However, while plutonium-239 remains in the reactor, it can absorb another neutron and become plutonium-240; but this isotope of plutonium spontaneously fissions, producing fast neutrons. There is now a balancing question, How much plutonium-240 can you tolerate without producing a weapon that will predetonate?

The fact that plutonium-240 would be produced was already known from the cyclotron production of plutonium. However, there was so little material to work with that measurements of the occurrence of this isotope were ambiguous. But now there were gram quantities, and Emilio Segrè was given the job of measuring the rate of spontaneous fission caused by the plutonium-240 in the sample they had. By late spring, Segrè reported that the spontaneous fission rate for this sample was at least five times as high as had been observed for the cyclotron-produced plutonium. By July 4 it had become clear that the gun-assembly method was not going to work for plutonium. It was just too slow. Neutrons would trigger a chain reaction before the material became supercritical. There was also a spontaneous fission issue for uranium-238, but in a bomb like Little Boy, some 90 percent of the material would be uranium-235,which had a spontaneous fission rate that was some 1200 times lower. This is why the gun-assembly method worked for uranium. There was no realistic way of separating plutonium-239 from plutonium-240. They differed by one mass unit, while uranium-235 and uranium-238 differed by three, which makes a huge difference when you are trying to separate isotopes. My guess is that if the people working on the bomb had not been persuaded that they were in a desperate race with the Germans, and if General Groves had not shared this obsession, the project might have stopped right there and then. As it was, Oppenheimer got discouraged and considered resigning as directer of Los Alamos. He didn't, but now the laboratory faced up to the two problems: metallurgy and assembly. I will begin with metallurgy. Enter into our story Cyril Stanley Smith .

Smith was born in Birmingham, England, in 1903. He got a degree in metallurgy from the University of Birmingham in 1924 and then a doctor of science from MIT in 1926. A year later, he began working at the American Brass Company in Connecticut's Naugatuck Valley. There he remained until the war, at which time he went to work for the War Metallurgy Committee in Washington, D.C. In February of 1943, while attending a meeting of the American Institute of Mining, Metallurgical, and Petroleum Engineers in New York, he was approached by the chemist Joseph Kennedy, who was one of Seaborg's collaborators in the discovery of plutonium and had been recruited to go to Los Alamos to head up its newly formed chemistry department. It is not clear why Kennedy contacted Smith in particular, although Smith had published a substantial amount of work and held several patents. It is also not clear what Kennedy could have told Smith about what would be going on at Los Alamos because Smith had no clearance. However, he told him enough, so that Smith saw going to Los Alamos as a way of escaping a desk job in Washington for which, as he later recalled, he had a "general distaste."

Not long after Smith's encounter with Kennedy, Oppenheimer had a recruiting talk with Smith on a park bench in Washington. Oppenheimer was very good at this sort of thing. By March of 1943, Smith was among the first group of scientists at Los Alamos. He was put in charge of creating a metallurgy group, without a clear idea of how big a job this was going to be. His first job was to find metallurgists who were not otherwise engaged in the war effort. This was not an easy task, but by 1945 when the war ended, he was running a department with 115 people in it. One of the difficulties in recruitment was that neither Smith nor anyone else, in the beginning, knew what such a department was supposed to do.

It was decided that the Los Alamos metallurgical group would not work on plutonium until gram samples arrived from the reactors, so they did various odd jobs, such as studying the properties of compounds of uranium with hydrogen. One of the oddest arose out of a request to take 620 pounds of gold and cast it into two hemispheres. Later Smith found one of the hemispheres being used as a doorstop. Once the plutonium began arriving at Los Alamos in half-gram lots in March 1944, the work to make a usable metal of it began in earnest. The first assumption was that its chemistry must be like that of uranium, because by this time it was understood how to make uranium into a metal: You began with uranium tetrafluoride (UF4) and took advantage of the fact that if you heated it in the presence of calcium (Ca), the calcium would be more attractive to the fluorine than uranium would and you would induce the reaction UF4 + 2Ca > U + 2CaF2, leaving uranium metal and calcium difluoride. Calcium here has acted as what is called a reducing agent. This sort of reaction was the way in which the micrograms of metallic plutonium that Zachariasen had been using had been made. This work was being done at the Met Lab by two young metallurgists Ted Magel and Nick Dallas. At the end of 1943, Magel and Dallas were producing one-gram buttons of pure uranium metal from uranium fluoride. By early 1944, Oppenheimer had persuaded the Met Lab to relinquish Magel and Dallas, who arrived in Smith's group bearing their Met Lab equipment, which included a centrifuge. They had performed the uranium reduction in a centrifuge, which would then separate out the metal. They planned to do the same thing for plutonium. Later it turned out that a better idea was to use what was known as a stationary "bomb," a crucible specially lined so that it could contain the plutonium compounds. But Magel liked the thrill of the centrifuge; Smith referred to this approach as "excited, energetic, but slightly slap-dash." Prior to the arrival of Magel and Dallas, the Los Alamos people, using their experience with uranium, tried to reduce plutonium trifluoride with calcium. They got what has been described as a "grayish cokey mass containing no agglomerated plutonium." Then Magel and Dallas got into the act. Magel's 1995 description of what occurred may be in the se non è vero è ben trovato category, but it is very amusing to read.

The reduction of a gram quantity of plutonium was considered a very big deal, because that amount of metal would allow much improved measurements of many crucial material properties. The reduction was supposed to take place on March 24, 1944, and General Groves and several top administrators had been specially invited to observe us as we did it. Well, when does everything go wrong when you have a whole lot of observers, right? So on the 23rd I said to Nick [Dallas], "Let's go up to the lab and make the reduction tonight before all these people get here." Nick agreed, and we carried out the reduction using the hot-centered centrifuge bomb method. When it was done, we cut open the bomb, dropped the little button of plutonium metal in a glass vial and put it on Cyril Smith's desk with a note that read: Here is your button of plutonium. We have gone to Santa Fe for the day. Everyone was pretty mad at us and claimed that we had contaminated the lathe and the back shop, when we opened the bomb to retrieve the plutonium button. I don't believe that we had, but I understood how they feit. In any case, once they had the button, they immediately started measurements of the density and so forth ....

Magel and Dallas had produced the first sample of metallic plutonium that could be seen without the aid of a microscope. It enabled measurements of the allotropic phases. The thing to note here is that there are two phases, δ and δ', in which the volume decreases when the temperature is raised. This is totally counterintuitive and is another example of just how bizarre an element plutonium is. There were diagrams during the war hut they were rather rough-hewn. Most of the details were filled in after the war. To give some idea, in 1958 a Russian chemist named Eugenii Makarov published what became a standard text on the crystal chemistry of uranium, thorium, plutonium, and neptunium. It was translated and published in the United States the following year.

Even in 1958, as their text makes clear, the crystal structure of the β-phase was still unknown. The next year, Zachariasen showed that it had a similar structure as the α-phase. During the war, anything that involved simple scientific curiosity was put aside if it did not contribute to making the bomb. One of the essential things that using these gram samples of metallic plutonium enabled the Los Alamos people to do was to measure the melting point of plutonium: the temperature at which the metal melts. The early Met Lab experiments gave results that seemed to be consistent (inconsistent? HJS) with the kind of temperatures one finds for other metals. To give a few examples; iron melts at 1510 °C, while steel melts at 1370 °C, copper at 1083 °C. On the other hand, it was discovered that plutonium metal melts at 640 °C, an extraordinarily low temperature, which you had better know if you are going to use plutonium metal for something.

After making the first gram, Magel and Dallas made eight more grams of the superpure plutonium that was thought to be required for a gun-assembly weapon. But once it became clear that such a weapon was impossible, superpure plutonium was no longer needed, so they were out of a job. They decided to leave Los Alamos and join a small Manhattan Project group at MIT.

© Jeremy Bernstein, 2007.