3—
Following up Fission

Back in Berkeley

When news of fission reached the Laboratory through the daily press about January 29, the staff could not be kept at the assembly of the 60-inch cyclotron, which had been its preoccupation. "All of us couldn't resist our curiosity to look into the matter a bit."

There were various approaches. Oppenheimer proved that fission is impossible. Alvarez wired East for further information, learned that Tuve's group had detected fission fragments, and immediately did the same, for the benefit of Oppenheimer and others who would not accept "so revolutionary [an effect] . . . until it had been confirmed in several laboratories."^[53] Alvarez and his collaborator, G.K. Green, in Berkeley as a National Research Fellow, saw the effects of thermal and fast neutrons and went beyond other hasty confirmers by estimating the time delay between irradiation and fission. They obtained an upper limit of 0.003 sec using a pulsed neutron beam modulated in Alvarez's manner; the limit was soon lowered to 0.001 sec by the intermittent neutron generator at Imperial College.^[54] They all missed the so-called "delayed neutrons," emitted a few seconds after cessation of the irradiation in a small percentage of fissions. These dilatory particles, ejected during the rearrangement of fission fragments, give a handle for the control of the fission process.^[55]

Green and Alvarez used Alvarez's preferred detector, a fast electronic affair, a thin-walled ionization chamber (on one of whose plates the uranium sat) connected to a linear amplifier and oscilloscope. Dale Corson, then in charge of the Laboratory's cloud chamber, teamed with Thornton to see fission fragments in a slower, but more evocative, way. They put uranium oxide on a collodion foil in a chamber filled with air, water vapor, and alcohol; irradiating the whole with neutrons procured from the 37-inch cyclotron, they obtained the nice result shown in plate 9.2. Assuming the short side tracks to belong to recoiling ions from the chamber gases, they inferred that the lower fission fragment must have had an atomic weight of at least 75 to have continued its course without apparent deviation. Two weeks after Corson and Thornton sent off the best of their picutures to the Physical Review

(they could choose among 25 obtained in 885 exposures to cyclotron neutrons), Joliot, a past master of cloud chamber technique, exhibited at the Paris Academy of Sciences the single photograph of a recoiling fission product that he had got in 902 exposures to Rn-Be neutrons. To physicists, the several quick observations of strongly charged, heavy particles constituted "conclusive evidence" of the phenomenon brought to light by Hahn and Strassmann after months of tedious, masterful analytical chemistry.^[56]

Seaborg heard about fission at the Journal Club meeting on January 30. "The news excited me very much. After the seminar, I spent hours walking the streets in Berkeley, chagrined that I had not recognized that the 'transuranic' elements . . . were not really 'transuranic' elements; I felt stupid." (He had judged the results of Curie and Savitch to be "rather strange.") His follow-up, or, rather, that of Joseph Kennedy, to whom he assigned the task, was not a chemical but a physical investigation, a search for the fast beta particles that they thought might be released in the successive transformations of the fission products. Kennedy spent some weeks at the quest, but to no avail. Nor could fast beta particles be detected concurrent with fission. They gave up and notified their colleagues of their failure.^[57]

In March, reinspired by the "elegant chemical separations" and "startling conclusions" of Hahn and Strassmann's definitive demonstration that uranium can give birth to barium, Kennedy and Seaborg tried another way to obtain information about the fission fragments. Once the chemical nature of one partner in a fission is known, the nature of the other follows from subtraction of atomic numbers; but a similar inference from an isotope of the one to an isotope of the other is not possible if free neutrons come out from a bursting uranium nucleus along with the fission partners. Kennedy and Seaborg looked for evidence of these neutrons by placing a Ra-Be source on the axis of a cylinder,

surrounding it with two concentric shells, the inner of uranium oxide and the outer of water, and counting the neutrons reaching the water's surface. They could not confirm their expectation that more neutrons should reach their detectors with the uranium in place than without it.^[58]

There was another and more urgent reason for desiring to know not only whether, but also how many, free neutrons may be emitted during fission. Should the number be large enough, the process of fission, once instituted by the energy of the physicist and his machines, might continue on its own, perhaps with explosive violence. That had been plain to the Columbia group the moment they had confirmed Frisch's experiment. "Here is real atomic energy!" Dunning wrote in his laboratory notebook on confirmation day. And on the morrow, in the same exclamatory style: "Secondary neutrons are highly important! If emitted would give possibility of self perpetuating neutron reaction." Many others saw the same: the omniscient Fermi; the busy Szilard and Szilard's sour partner Arno Brasch, who twitted him about the now worthless patents on chain reactions with which he had hoped to direct the course of events.^[59] At Berkeley they also understood. "We are trying to find out whether neutrons are generally given off in the splitting of uranium; and if so, prospects for useful nuclear energy become very real!" "It may be that the day of useful nuclear energy is not so far distant after all."^[60] Alvarez spearheaded the hunt, though not with his customary thoroughness. He directed his beam of slow neutrons into a bottle of uranium oxide and looked for fast neutrons arising from fission. He saw none.^[61] He also tried, with the help of a chemist, Kenneth Pitzer, to detect an increase in the temperature of uranium irradiated by neutrons. Again the experiments were inconclusive. They were resumed by Malcolm Henderson, back in Berkeley for the

summer, who detected heat enough to indicate that the fission fragments recoiled with an average energy of around 175 MeV, in agreement with other estimates made in other ways.^[62] No further efforts to detect fission neutrons at the Laboratory have come to light, although Lawrence said publicly that the possibility of obtaining energy from uranium depended upon the number of neutrons released in fission.^[63]

His lethargy in prosecuting this possibility, with which he had been flirting for several years, may be accounted for as follows. For a change, time was more precious than money. Lawrence preferred not to delay completion of the 60-inch cyclotron for the chance that the Laboratory would be the first to find and advertise fission neutrons. Nor was a cyclotron the best tool for the search. The groups who first established the multiplication of free neutrons during fission worked with Ra-Be sources. The earliest in the field was Joliot's. They compared the neutron intensity at various distances from their source of photoneutrons (Rag -Be) when it sat in tanks containing plain water and water strewn with uranium oxide (hence they measured primarily the effects of slow neutrons). Fermi and two of his associates at Columbia used the same technique, which was adapted from one the Rome group had invented to measure neutron absorption; while another pair at Columbia, Szilard and Walter Zinn, observed fast neutrons from uranium struck by slow ones. In Paris they overestimated the average number of neutrons emitted during uranium fission as 3.5; in New York they made it 2, and then, in a more careful measurement by a combination of the two Columbia groups, who pooled their radium and uranium, 1.5.^[64] One preferred the higher or

lower figure according to whether one hoped for nuclear power or feared a nuclear bomb.^[65]

The only important immediate contribution of the Laboratory to the unravelling of the fission process was made by a junior member, Philip Abelson, who, by combining an undergraduate chemistry major with graduate training in physics, constituted an untried one-man interdisciplinary team. In 1937, in the course of the customary search for new activities, he had confirmed the general scheme of transuranics then recently put forward by Meitner, Hahn, and Strassmann, and added a new activity, of 17 hours, whose place he could not find. Then, under the inspiration of Alvarez's work on K-electron capture, he thought to identify "transuranics" by their characteristic x rays.^[66] In March 1938, in "transuranics" prepared in quantity at the cyclotron, Abelson isolated an activity of 77 hours, which he identified with the 66-hour body that the Berlin group had fixed as element 95, and whose characteristic x rays he managed to detect with a cheap spectrometer he had constructed. By extrapolating beyond uranium the increase with atomic number of the penetrating power of L rays—characteristic radiation involving the second or L shell of the atom—Abelson convinced himself that the x rays he detected belonged to an atom of nuclear charge 95. The revelation of fission dumbfounded him.

After a day in the dumps, Abelson looked again and learned that the "L rays" of 77-hr element 95 were in fact K rays of 2.4-hr iodine (atomic number 53). The connection, as Abelson worked it out: the iodine, Meitner et al.'s 2.5-hr element 96, descends from the 77-hr body, which must accordingly be tellurium; tellurium's beta decay can knock out a K electron, setting up the production of iodine's K ray. The process, taken all together, constituted "an unambiguous and independent proof of Hahn's hypothesis of the cleavage of the uranium nucleus."^[67] Abelson's rays and Dale Corson's pictures caused Oppenheimer to adjust his evaluation of

fission from "impossible" to "unbelievable." And if the unbelievable should give rise to some extra neutrons? "A ten cm cube of uranium deuteride (one should have something to slow the neutrons without capturing them) might very well blow itself to Hell."^[68]

Abelson carried on. In three months of hard work, reprieved from crew duty, he found in the products of uranium irradiated by neutrons no fewer than five activities ascribable to antimony, seven to tellurium, and four to iodine. He thereby not only showed the complexity of the German transuranics, but also demonstrated that fission fragments were ordinary isotopes. Two of his telluriums and one of his iodines had the same periods and beta-ray spectra as activities then recently identified in the Laboratory by Seaborg, Livingood, and Kennedy. Abelson could therefore specify decay sequence, half-life, element, and, in some cases, the isotope of the products he studied. In accomplishing his work he mobilized the great experience of the Laboratory in nuclear chemistry, as represented by Segrè and Seaborg, and, of course, the Laboratory's prop and pride: Abelson attributed his success "in large measure to the huge neutron intensities of the cyclotron."^[69] The cyclotron deserved the credit, and more. Its neutrons made fission products strong enough to swamp the activities of the natural descendents of uranium, which regrew during chemical analysis of the irradiated samples. With their "enormously weak preparations," Hahn and Strassmann had great difficulty finding iodine among the uranium products and succeeded only after Abelson's demonstration. Their confirmation resulted in what they called the "definitive suppression" of their long transuranic chains.^[70]

And what about the Curie-Savitch body that precipitated the studies that led to fission? As Meitner and Frisch observed, a barium isotope arising from uranium should be accompanied by

an isotope of krypton that might decay through rubidium, strontium, and yttrium to a stable zirconium. Hahn and Strassmann immediately started to look for these elements, of which they soon reported sightings.^[71] Now yttrium has chemical properties similar to those of lanthanum, from which Curie and Savitch had not been able to separate R_3.5h . The exact investigation of the sequence from strontium on was left to one of Hahn's assistants, who found an yttrium of the necessary period, but declined to pronounce definitively that it decayed to zirconium and constituted the inspirational French "lanthanide." It did, and does.^[72]

Utopia or Armageddon

There remain the connections among the processes that the Berlin group had tied up in the problematic isomers of U²³⁹ . Recall that the long chains in figure 9.5 arise from fast and thermal neutrons respectively, and the short chain from resonance capture. Meitner and Hahn kept the third heavy uranium, of half-life 23 minutes, as a true U²³⁹ and tacitly referred both the long chains to the fission of U²³⁸ . They clung to this interpretation despite its odd consequence, that U²³⁸ would be fissionable by both fast and thermal neutrons but not by ones with moderate speeds, because the two chains seemed to be present in about equal strengths. They knew that a lighter isotope of uranium existed, but since it makes up less than 1 percent of natural uranium, it did not appear to them to be the ancestor of either long chain. Bohr's highly skeptical associate, George Placzek, adduced the ambiguity of the role ascribed to U²³⁸ as a strong argument against the possibility of fission. Thus inspired, Bohr found a way to father thermalneutron fission on U²³⁵ . It was only necessary to detach the fissions from the old decay chains: experiment could not distinguish which uranium isotope gave rise to which fragment; the

genetic connections in the first steps of the long chains were illusory.

Bohr argued from the excitation energies of the compound nuclei U²³⁹ and U²³⁶ . In the former case, the impacting neutrons must bring in the explosive energy, because, as an unpaired nucleon in the odd-numbered isotope 239, it will not excite the compound nucleus very much by its binding energy; in the latter case, the pairing of the neutron in the even isotope 236 contributes enough to the general excitation to create a chance of fission. The slower the neutron, the longer it takes to pass a nucleus and the greater the chance of its absorption.^[73] Hence the final interpretation of the "transuranic" chains: one comes from fast-neutron fission of U²³⁸ , the other from thermal-neutron fission of U²³⁵ . Placzek did not think that Bohr's reasons were very compelling.^[74] Others thought them strong enough to undertake the very difficult task of separating enough U²³⁵ from natural uranium to investigate its fissile properties directly.

Among the bold were the enterprising pair Seaborg and Kennedy. In the summer of 1939 they began construction of a tube, eventually over twenty feet in length, which they fixed to the outside of the chemistry building. Their scheme adapted a recent serendipitous invention made by Klaus Clusius and Gerhard Dickel, chemists at the University of Munich, who subjected a mixed gas to thermal gradient between a hot wire running down the tube's axis and the cooler tube walls. The resultant motion of the gas is hard to calculate but easy to describe: lighter molecules move radially inward under the thermal stress and axially upward by convection, heavier ones radially outward and axially downward. In the circulation, light molecules concentrate at the top and heavy ones at the bottom of the tube. Clusius and Dickel reported an almost complete separation of the major isotopes of chlorine.^[75] Kennedy and Seaborg planned to work with uranium hexafluoride gas, or "hex," as it came to be known from its

unsavory character, and they set up a fluorine generator to cast the spell. A little hex eventuated; a graduate student, Arthur Wahl, joined the project; Clusius's column was installed and Clusius written, in the best and most naive tradition of open science, for advice about uranium separation. Then, on January 12, 1940, the fluorine generator exploded while the three would-be hex splitters were working on it. The next day Kennedy was ill. Seaborg consulted a chemistry book: "Uranium is an extremely powerful, slow-acting poison." Kennedy stayed sick.^[76] Although his problem turned out to be mononucleosis, the incident destroyed the group's enthusiasm for hex. They put their columns—eventually they had three in operation—to separating the uncursed isotopes of hydrogen, carbon, and chlorine.^[77]

Another reason that Seaborg's group did not pursue uranium splitting was that others were doing it with greater success. By early 1940, the mass spectrograph run by Alfred Nier at the University of Michigan had collected enough U²³⁵ for the Columbia group to make possible a positive test of Bohr's conjecture about its fission by thermal neutrons. By the summer of 1940, when Kennedy and Seaborg made a tour of laboratories including Nier's, they learned about several attempts to acquire U²³⁵ in bulk: Urey and Tuve by diffusion methods, Beams by centrifugation, all starting with hex. The Columbia group had the help of Aristid von Grosse, a former collaborator of Hahn's and an expert in hex manfacture.^[78] That did not exhaust the competition. There were also Frisch, then relocated in England, where he and Otto Blüh, a refugee from Prague, set up a Clusius tube late in 1939; Clusius himself, who had declined to supply the additional information that Seaborg had requested; and Wilhelm Krasny-Ergen, who had established a Clusius distillery in Stockholm, from which, "had his activities not been suspended by the political situation," he expected a sevenfold enrichment in eighty days.^[79] Although

Tuve volunteered to send enough hex for preliminary experiments, Kennedy and Seaborg wisely withdrew from uranium separation. Perhaps Lawrence's belief that centrifugation held the greatest promise for large-scale separation of heavy isotopes influenced their decision.^[80]

The confirmation that U²³⁵ has a large cross-section for fission by slow neutrons gave physicists greater confidence and worry that an explosive chain reaction could be achieved. "Physicists are anxious that there be no public alarm over the possibility of the world being blown to bits by their experiments," Science Service reported in melodramatic ignorance just after the confirmation of fission at the end of January 1939. The world was not alarmed, despite a revelation in the New York Times that a little U²³⁵ could wipe out New York City and leave a hole halfway to Philadelphia.^[81] During 1939 physicists calculated what might be possible, but in ignorance of the relevant cross-sections and reactions they could only conjecture. An effort to keep pertinent data secret assisted the progress of ignorance and speculation. Although Szilard's novel notion that physicists should censor themselves failed to persuade Joliot and so failed internationally, enough was withheld in the United States that physicists as well placed as Tuve and Frisch were "hard pressed to get some data on uranium fission."^[82] The most sanguine discussion of the future of fission came from a colleague of Hahn's at the Kaiser-Wilhelm-Institut für Chemie, Siegfried Flügge, who offered a path to a "uranium machine."

Flügge started from the obvious: for a chain to succeed, enough slow neutrons must be obtained and caused to provoke fissions before they are lost or captured ineffectually. To slow the fast

fission neutrons, it is necessary only to pass them through water, which, to be sure, also captures them; but, according to Flügge's calculations, based on the probability that a slow neutron will cause a fission as measured at Columbia and on his own estimate of the liability of a fast neutron to loss in water while slowing down, a mixture of fifteen kilograms of uranium oxide per liter of water will sustain a chain reaction. That assumed the conservative estimate that an average fission sets free two fast neutrons. It also assumed that the reaction once started could be controlled. Flügge planned a power plant, not a bomb. A method of control had been published by two members of Joliot's team. It rests on the principle that the faster the neutron, the lower its probability of provoking a fission in uranium, and on the fact that the appetite of cadmium nuclei for slow neutrons is almost independent of the temperature. Therefore sprinkle a little cadmium dust in the water along with the uranium oxide. The chain begins, the mixture heats up, the neutrons move faster, the number of fissions goes down; equilibrium will be reached at a temperature fixed by the amount of neutron-removing cadmium present. According to Flügge, with 0.2 gram of cadmium per liter as seasoning, his uranium stew would boil along at a safe and steady 350°. The reactor vessel would require a diameter of a little over a meter. From it heat could be removed and used to make steam to drive a turbine; a cubic meter of uranium thus exploited could generate electricity for eleven years at a rate equivalent to Germany's consumption on the eve of the Depression.^[83]

By the time Flügge's design was published, Fermi and his colleagues had decided that water absorbs too many neutrons to serve as decelerator, or moderator, in a uranium machine. Hopes of capitalizing quickly on fission faded; by August 1939, when Tuve managed to learn what had been learned, he concluded that "all indications are that no chain can occur but it is pretty close." At the same time, Bohr was calculating the amount of hydrogen moderator needed to slow the neutrons. He arrived at a ratio of ten atoms of hydrogen to one of uranium, which he thought would prohibit a fast chain or strong explosion. Recalculation reduced

the ratio to one to one, which allowed a possibility. The Paris group was more negative. Their experiments with a homogeneous mixture of three atoms of hydrogen to one of uranium gave evidence of fissions caused by secondary and tertiary neutrons, but not of a divergent multiplication of fissions; and by the end of October 1939, in a paper they withheld from publication, they concluded that "it is almost certainly impossible" to promote a divergent chain reaction in a homogeneous blend of naturally occurring uranium, oxygen, and hydrogen.^[84] Of course, a moderator other than ordinary water, or a heterogeneous mixture of uranium and water, might perform better. The published consensus of physicists, as expressed in the several reviews of the year's fission research composed toward the end of 1939, was that exploitation of nuclear energy would not occur in the near future and might not be possible at all.^[85]

The announcement on May 1, 1940, of the details of Columbia's fissioning of Nier's latest sample of U²³⁵ transformed the discussion. William L. Laurence, a science writer for the New York Times , informed its readers on May 5 that five or ten pounds of U²³⁵ could drive an ocean liner or submarine indefinitely. So much light uranium might not be hard to procure. Nier had been able to enlarge his sample two hundred times in a matter of weeks. "It is not impossible that a few months or a year hence may see the realization of this quest." Nothing would be simpler than exploiting the new fuel. "All that is needed to put it to work running motors and steamships is to place it in a tank of water. . . . The water would be turned into steam. . . . New water supplied would keep the process going indefinitely." Furthermore, the business would be automatic and self-regulating, since the

heating of the water speeds up the neutrons and slows or stops the fission process until more water, "the colder the better," is added. Thus utopia. But evil men were trying to suborn the natural good behavior of U²³⁵ . "Every German scientist in this field, physicists, chemists, and engineers, it was learned, have been ordered to drop all other researches and devote themselves to this work alone." Fortunately they lacked the essential machine for further investigation, the cyclotron. A race had started, a race with stakes that were incalculable, or almost so. According to Laurence, a pound of U²³⁵ improperly treated would have the same explosive power as 15,000 tons of TNT. An effective separation plant, therefore, was "a secret to be given only to the United States government."^[86]

The physicists were not pleased with Laurence's mixture of fact and fiction. Nier declared that his handiwork had little present commercial or military value, the amount of U²³⁵ so far isolated being "hardly enough to spring a mousetrap." S.K. Allison rated producing utilizable atomic energy as "just as feasible as getting gold out of the ocean." George Pegram, who, as chairman of Columbia's physics department, was constantly pressed by the press, urged his colleagues to "stress what seems to be the fact, namely, that energy from uranium, even if it became available, would apparently not be cheap energy by any means and would not be very explosive energy." Pegram sold this point of view to the informed and responsible chief of the Times 's science section, Waldemar Kaempfert, who squared accounts as follows. To make a pound of U²³⁵ would cost more than the expenses of the federal government for an entire year; to make a gram by Nier's improved technique would take over a century. "The prospect of using U²³⁵ in the present war is zero."^[87]

While Kaempfert calculated, Krasny-Ergen's plan was published in Nature . It set Laurence off again. According to him, 10,000 of Krasny-Ergen's units could make a pound of U²³⁵ in forty days; at $100 per unit, a uranium factory would cost only $10 million. "It may be expected, therefore, that Germany will take measures at

once to install such a plant." Kaempfert again flew to the rescue: Krasny-Ergen's method of thermal diffusion would require 17 million kWh to separate a gram of U²³⁵ ; to make a kilogram, 34 million tons of coal would have to be burnt, at a cost of $68 million. "The more we think this over the more we are convinced that we would not invest ten cents in a uranium public utility company. . . . We doubt if the Germans have the time or the stupidity to bother much about isolating uranium-235 in large quantities."^[88]

Meanwhile Lawrence was examining a plan for a German power plant driven by light uranium. He had this information from Clifford Williams of Shell Development, who had it from Peter Debye, who had recently left the directorship of the Kaiser-Wilhelm-Institut für Physik in Berlin. According to Debye, all his former staff were engaged in developing U²³⁵ as a power source (fig. 9.8); Germany was "frantically mining uranium in Czechoslovakia;" and the native metal was to be separated by diffusion. It was perhaps indirectly from this disclosure that the author of the article on Lawrence in Scientific American obtained the information that "the Nazis are trying to lay hands on all the uranium they can find." The least secret bit of science in the United States in the summer of 1940 was that (as the San Francisco Chronicle put it) atomic power "will transform the face of the earth the moment the production of the magic element U-235 can be cheapened."^[89]

[Full Size]

Fig. 9.8
Design for a German U²³⁵  plant. Williams to Lawrence, 23 May 1940 (18/26).