The Apocalypse Factory

by Steve Olson

  In science, new ideas often occur to more than one person at about the same time. They are “in the air” because previous developments point in their direction. That was not the case for the chain reaction. For the next five years, of all the scientists in the world, only Szilard thought deeply about the possibility. When he mentioned the idea to others, they mostly dismissed it as unlikely or impractical. At the same time, Szilard did not mention it to many people, because he immediately recognized the idea’s danger. If a chain reaction were possible, and if German scientists learned how to create one, they might be able to build a bomb so powerful that Germany could conquer the world.

  Early in 1938, tiring of Europe and its politics, Szilard moved to New York City, where he took a room at the King’s Crown. He spent much of that year traveling from place to place, meeting with people he knew in various academic and industrial laboratories and trying to get them interested in the chain reaction. But most were uninterested or unable to help. By the end of 1938, he was back at the King’s Crown and ready to give up.

  Szilard met Fermi just as the news of fission was making its way to America. Suddenly, the chain reaction that Szilard had pursued for years seemed within reach. If uranium gave off neutrons when it fissioned, those neutrons could cause more fissions, which in turn would give off more neutrons. But did fissions produce neutrons? Within a few weeks of their meeting, both Fermi and Szilard were working on the problem. Fermi had struck up a partnership with a young graduate student at Columbia named Herbert Anderson, just 24 years old, a scholarship student from the Bronx who had been studying electrical engineering before getting interested in physics. Szilard, meanwhile, had convinced a Canadian-born physicist at Columbia named Walter Zinn, 32 that year, the son of a factory worker, to collaborate with him on fission research. Before long, both groups had found the extra neutrons, as had several other groups in the United States and Europe. For Szilard, the discovery was vindication that the idea he’d had while crossing the street in London was correct. Yet he also knew that the discovery brought the world one step closer to an atomic bomb. The evening he and Zinn discovered the extra neutrons produced by fission, he later wrote, “there was very little doubt in my mind that the world was headed for grief.”

  The discovery that uranium releases extra neutrons raised another question: why don’t chain reactions occur in uranium ore? This question, too, was answered in the first few months of 1939. Uranium ore consists almost entirely of two isotopes—one with 92 protons and 146 neutrons, known as uranium-238, and one with 92 protons and 143 neutrons, or uranium-235. Theory suggested, and experiments confirmed, that only uranium-235 was fissioning when it was hit with neutrons. But uranium-235 makes up only about seven-tenths of one percent of uranium ore. That’s too low to support a chain reaction.

  Physicists also discovered that uranium-235 was splitting not just into barium and krypton but into many different pairs of atoms—xenon and strontium, and iodine and yttrium, and cesium and rubidium—into more than 20 different elements altogether. These fission products, as they’re called, tend to be highly radioactive. That’s because they have too many neutrons for the number of protons they contain. Uranium atoms, to keep their 92 protons together, have proportionately more neutrons in their nuclei than do lighter elements. As a result, when uranium atoms split, their fission products have a sort of neutron sickness—they have too many neutrons to be stable. To get back into balance, these fission products begin to convert neutrons in their nuclei into protons while releasing electrons and gamma rays. In this way, the fission products start crawling their way down the list of elements, away from hydrogen and toward uranium, until their neutron sickness is resolved.

  When physicists realized that only uranium-235 fissions in the presence of neutrons, they immediately saw one way to create a chain reaction. If enough uranium-235 could be extracted from natural uranium ore, the neutrons from fissioning atoms would cause other atoms to fission in an ever-growing cascade. If this reaction could be controlled, uranium-235 could provide virtually unlimited quantities of energy, whether for electricity generation, propulsion, or other purposes. If the reaction were uncontrolled, and if it occurred quickly enough, the result would be a bomb of unprecedented power.

  But isolating uranium-235 from uranium ore was going to be extremely difficult. Because the two uranium isotopes are different versions of the same element, they behave exactly the same chemically. The only way to separate them is on the basis of their weights. But they differ in weight by only about one percent, which provides very little traction for a separation process. At the time, several technologies existed that might do the job, but they would have to become much more efficient and then be applied on a huge scale. As the Danish physicist Niels Bohr famously remarked, “It would take the entire efforts of a country to make a bomb.”

  But Szilard was not going to give up on the chain reaction. Maybe a way could be found to arrange natural uranium ore so that a chain reaction could be produced some other way. And as more was learned about fission, he began to glimpse an intriguing possibility.

  The neutrons emitted by a fissioning uranium-235 nucleus travel very quickly—at about one-fifteenth the speed of light. When a neutron traveling at this speed hits a uranium-238 nucleus, one of two things usually happens. The neutron can bounce off the nucleus and head in another direction. Or it can be absorbed by the nucleus, creating uranium-239. In fact, this capture of neutrons by uranium-238 is the real reason why natural uranium ore does not explode. The absorption of neutrons by the heavier isotope of uranium squelches any possible chain reaction.

  But if the fast-moving neutrons from fission are slowed down, something else happens. Slow neutrons are less likely to be captured by uranium-238 atoms. Instead, they bounce harmlessly off uranium-238 until they find a nucleus of uranium-235 to fission. That’s the trick to achieving a chain reaction with natural uranium, Szilard realized. If some sort of moderator could be found that would slow down the fast neutrons from uranium-235, fewer neutrons would be captured by uranium-238, leaving more than enough to produce a chain reaction.

  But how could neutrons be slowed down? Remarkably, Fermi had answered this question just a few years before. High-speed neutrons slow down quickly when they bounce off the nuclei of light atoms. So to find a good moderator, Szilard and Fermi, working together now as each recognized the potential of chain reactions, began making their way down the list of elements. Hydrogen atoms, with their single proton, work best at slowing down neutrons. But they occasionally absorb neutrons, slowing the chain reaction. The next heavier element, helium, does not absorb neutrons, but suspending uranium ore in a container of helium gas did not seem immediately practical. Lithium, with three protons, is a strong neutron absorber, so that would not work. Beryllium, with four protons, does not absorb many neutrons, but it is highly toxic when inhaled. Boron, with five neutrons, absorbs neutrons like crazy. But what about carbon, with six protons and six neutrons? It captures neutrons at one-hundredth the rate of hydrogen. And a source of concentrated carbon was readily available: graphite, the “lead” in lead pencils.

  By the summer of 1939, Szilard had convinced himself that graphite would work. “It seems to me now that there is a good chance that carbon might be an excellent element to use” as a moderator, he wrote in a letter to Fermi. “I personally would be in favor of trying a large-scale experiment with a carbon-uranium-oxide mixture if we can get hold of the material.” But getting hold of the material wasn’t going to be easy. A few days later, Szilard visited the National Carbon Company on East Forty-Second Street to ask about the cost of acquiring graphite blocks. It could be done, but the graphite had to be extremely pure, and that would be expensive. Fermi and Szilard also needed a large amount of graphite. A slow neutron bouncing around in a large block of graphite containing embedded masses of uranium—which is the design that Szilard and Fermi gradually worked out over the next couple of years—can do one of two things. It can find an a
tom of uranium-235 to split. Or it can reach the edge of the graphite and escape, in which case it can no longer fission uranium. For a chain reaction to occur, a mass of graphite and uranium would have to be large enough for a neutron to find a uranium-235 atom before it was lost to the surrounding space.

  In the spring of 1940, Columbia University received $6,000 from the federal government to cover the purchase of a large quantity of uranium ore and extremely pure graphite. Fermi, Szilard, Anderson, Zinn, and a steadily growing research team immediately began to experiment with the materials Szilard bought with the new funds. Using bench saws and planers, they machined the graphite into four-inch by four-inch by twelve-inch blocks. They then stacked the blocks in layers until they had a column of dusty black graphite. The physicists at Columbia “started looking like coal miners,” Fermi recalled, “and the wives to whom these physicists came back tired at night were wondering what was happening.” The experimenters put neutron sources under the graphite blocks to measure how quickly the neutrons slowed down and whether they were being absorbed by remaining impurities in the graphite. They then put uranium in the midst of the graphite blocks. Sure enough, when neutrons from the bottom of the column encountered the uranium, fissioning uranium-235 atoms generated new neutrons. Fermi and Szilard were not yet calling this device by the name its much larger descendants would acquire—a reactor. They called it what it was—a pile.

  Chapter 3

  ELEMENT 94

  EVEN AS FERMI WAS BUILDING HIS PILES AT COLUMBIA UNIVERSITY, Seaborg was beginning the experiments that would link the work of the two men forever. Since 1939, a physicist at Berkeley named Ed McMillan had been studying the atoms produced when uranium fissions, and one of them had caught his attention. Because of the energy generated when uranium splits, most fission products fly away from the site of the fission at high speeds. This one didn’t. It was radioactive, just like a fission product. But it stayed close to the place where it was created. McMillan began to suspect that it wasn’t a fission product at all.

  For help he called on Phil Abelson, the Rad Lab employee who had been so distraught at barely missing the discovery of fission. Abelson had a background in chemistry, and he soon was able to show that the atom McMillan was studying had chemical properties different from those of any other known element. Only one conclusion could be drawn. It must be a new element. The most plausible explanation was that the uranium-238 was capturing a neutron from fissioning uranium-235 atoms. That would make it uranium-239, an isotope of uranium with 92 protons and 147 neutrons. But this isotope of uranium was evidently unstable. It appeared to be converting one of its neutrons into a proton, just as fission products do. The result would be an element with 146 neutrons and 93 protons, a transuranic element, an element never before seen on Earth.

  As the discoverers of the new element, McMillan and Abelson had the privilege of naming it. Uranium had been named after the planet Uranus. McMillan and Abelson therefore named their new element neptunium, after the next planet out from the sun.

  Seaborg had been following McMillan’s work closely—he would even pester McMillan with questions when the two met in the shower room of the faculty club. Then, in the fall of 1940, McMillan suddenly moved to MIT. Scientists in the United States were gearing up for an event that many people thought was inevitable—America’s entry into World War II. Instead of studying radioactive elements, McMillan spent the next few years developing radar.

  When Seaborg learned that McMillan was giving up his work at Berkeley, he wrote a letter asking if he and his colleagues could continue McMillan’s research. McMillan wrote back to say that he would be happy to turn the work over to Seaborg’s group.

  Seaborg and his colleagues at Berkeley had reason to believe that the neptunium found by McMillan and Abelson was also undergoing radioactive decay. If so, it would convert another one of its neutrons into a proton. The result would be an element with 94 protons—another transuranic element unknown in nature. Because of the way its nucleus was configured, this element would probably be much more stable than either uranium-239 or neptunium. And if this was true, element 94 could be a very interesting element indeed.

  WELL PAST MIDNIGHT on February 24, 1941, a storm was roiling the waters of San Francisco Bay as Seaborg and Art Wahl, a 23-year-old graduate student, prepared for the decisive experiment in Room 307 of Berkeley’s Gilman Hall. Many years later, when Room 307 of Gilman Hall was being designated a National Historic Landmark, Seaborg said that “a less significant or historical looking room hardly existed on the campus of the University of California.” It had a small sink set in a heavily stained stone countertop. Discolored iron pipes nearly filled the space between the countertop and a glass-fronted cabinet filled with apparatus and chemicals. The room was on the top floor of Gilman Hall, and a set of glass doors led to a small balcony notched into the red clay roof tiles of the building. The chemical hood in the room did not have a fan, and when Seaborg and Wahl wanted to do a particularly noxious experiment they moved their equipment to the balcony. But on this particular night rain was spattering the west-facing doors, and Seaborg and Wahl braved the fumes of their experiment inside.*

  About two months earlier, Wahl, who had come to Berkeley about a year earlier after majoring in chemistry at Iowa State University, had spread a uranium paste onto a copper plate. He and Seaborg then placed the plate in the target area of the 60-inch cyclotron at Berkeley. After the plate had undergone four hours of bombardment, Wahl, back in Gilman Hall, scraped the powder off the plate with an ice pick. He dissolved the scrapings in hot nitric acid and was alarmed to see the solution turn green. The uranium was doing something it should not be doing. But then he remembered—the copper plate. Some of the copper must have come off with the irradiated uranium.

  He took a bottle of cerium fluoride from the shelf and mixed some into the solution. The white powder caused the uranium, the neptunium, and any elements derived from neptunium to form solid particles in the liquid. Wahl separated this precipitate from the solution with a filter and attached it to a piece of cardboard.

  For the next several weeks, Seaborg and Wahl plied their sample of bombarded uranium with chemicals, trying to figure out how to isolate the new element they suspected they had made. When they wanted to know what elements were present in a solution or a precipitate, they took it two doors down the corridor to Room 303 in Gilman Hall, where their colleague Joseph Kennedy had built an ingenious set of radiation detectors. Every radioactive isotope gives off a distinct set of signals—for example, electrons and gamma rays with a particular range of energies. By measuring these signals in his detectors, Kennedy could tell what radioisotope each sample contained.

  But Seaborg and Wahl had a problem. To isolate their suspected new element from the irradiated uranium, they needed a chemical that would combine only with that element and not with any other, but they couldn’t find a compound strong enough to do the job. Finally they asked Berkeley chemist Wendell Latimer. He recommended the strongest combining agent he knew, peroxydisulfate, a compound in which two sulfur atoms brandish eight oxygen atoms like spikes on a cudgel. Late in the evening on Sunday night, February 23, Wahl dissolved some of the irradiated uranium in acid, added some peroxydisulfate, and let the solution stand for about half an hour. He then added hydrofluoric acid to the solution, which caused the remaining uranium and neptunium to form solids that he separated from the solution with a filter. But Kennedy’s detectors showed that the new element remained in the solution, meaning that it “can be separated from all the known elements,” as Seaborg wrote in his journal.

  Seaborg later recalled stepping onto the balcony of Gilman Hall after the momentous experiment. Ahead, the sky over San Francisco Bay reddened with the rising sun.

  “I was a 28-year-old kid and didn’t stop to ruminate about it,” he later told an Associated Press reporter. “I didn’t think, ‘My God, we’ve changed the history of the world.’”

  But Seaborg and Wahl had discovere
d an element that would change the history of the world.

  EVEN BY THE STANDARDS of the Berkeley chemistry department, Seaborg made for a strange sight as he struggled up the steps of the chemistry building. In addition to his lab coat he wore heavy lead-impregnated gloves and goggles. At the end of a long pole he carried a lead bucket. He was trying to protect himself from the radioactivity being emitted by the fission products in the bucket. Whether he succeeded is questionable. Years later, Room 307 in Gilman Hall had to be thoroughly decontaminated to temper the radioactivity emanating from its walls, floor, and countertops.

  Seaborg and his colleagues were conducting one of the most consequential experiments of the nuclear age. Could the new element that they had created be used to build atomic bombs? The idea seemed plausible. The reason uranium-235 fissions while uranium-238 does not is that the former has an odd number of protons and neutrons while the latter has an even number. Like dancers, protons and neutrons like to pair up in nuclei, which makes even-numbered nuclei less likely to fission from exposure to neutrons than odd-numbered nuclei. But the radioisotope Seaborg’s group had created, like uranium-235, had an odd number of nuclei—94 protons and 145 neutrons, for a total of 239. Furthermore, that isotope of element 94 is relatively stable and remains largely unchanged for thousands of years. If their new element fissioned like uranium-235 fissions, it, too, could serve as the fuel for atomic bombs.

  In the last week of February 1941, Seaborg and his colleagues had placed about two-and-a-half pounds of uranium in the target of the cyclotron and had bathed the uranium in neutrons. By the time they were done, the sample—200 times larger than the sample used to discover element 94—was filled with fission products and dangerously radioactive.