Making of the Atomic Bomb
In the beginning Fermi worked alone. He intended eventually to irradiate most of the elements in the periodic table and he started methodically with the lightest. His source, he calculated, supplied him with more than 100,000 neutrons per second.792 “Small cylindrical containers filled with the substances tested,” he would explain in his first report, “were subjected to the action of the radiation from this source during intervals of time varying from several minutes to several hours.”793 Fermi first irradiated water—testing hydrogen and oxygen at the same time—then lithium, beryllium, boron and carbon without inducing them to radioactivity. Laura Fermi says he wavered then, discouraged by the lack of results, but Fermi seldom talked shop at home and doubt seems unlikely: he knew from the Joliot-Curie work that aluminum, a little farther along, reacted with alphas, and neutrons should prove even more effective.
In any case he succeeded on his next attempt, with fluorine: “Calcium fluoride, irradiated for a few minutes and rapidly brought into the vicinity of the counter, causes in the first few moments an increase of pulses; the effect decreases rapidly, reaching the half-value in about 10 seconds.”
Soon he found a radioactivity in aluminum with a half-life of twelve minutes, different from the Joliot-Curies’ discovery. Putting aluminum first to link his work with theirs, he reported his findings in a letter to the Ricerca Scientifica on March 25, 1934.
A Roman numeral I distinguishes that first report on “Radioactivity induced by neutron bombardment.” The search was on. To move it along Fermi recruited Amaldi and Segrè and cabled Rasetti in Morocco to rush home. Segrè writes:
We organized our activities this way: Fermi would do a good part of the experiments and calculations. Amaldi would take care of what we would now call the electronics, and I would secure the substances to be irradiated, the sources, etc. Now, of course, this division of labor was by no means rigid, and we all participated in all phases of the work, but we had a certain division of responsibility along these lines, and we proceeded at great speed. We needed all the help we could get, and we even enlisted the help of a younger brother of one of the students (probably 12 years old), persuading him that it was most interesting and important that he should prepare some neat paper cylinders in which we could irradiate our stuff.794
The next letter that went to the Ricerca Scientifica (and in summary form to Nature) reported artificially induced radioactivity in iron, silicon, phosphorus, chlorine, vanadium, copper, arsenic, silver, tellurium, iodine, chromium, barium, sodium, magnesium, titanium, zinc, selenium, antimony, bromine and lanthanum.795 By then they had established a routine: they irradiated substances at one end of the second floor and tested them under the Geiger counters at the other end, down a long hall. That shielded the counters from stray radiation from the neutron source. But it also meant, whenever the half-life of an induced radioactivity was short, that someone had to run down the hall. “Amaldi and Fermi prided themselves on being the fastest runners,” Laura Fermi notes, “and theirs was the task of speeding short-lived substances from one end of the corridor to the other. They always raced, and Enrico claims that he could run faster than Edoardo. But he is not a good loser.”796 A dignified Spaniard showed up one day to confer with “His Excellency Fermi.” Rome’s young professor of theoretical physics, a dirty lab coat flying out behind him, nearly knocked the visitor down.
They came, finally, to uranium. They had roughly classified the effects they were seeing. Light elements generally transmuted to lighter elements by ejecting either a proton or an alpha particle. But the electrical barrier around the nucleus works against exits as well as entrances, and that barrier increases in strength with increasing atomic number. So heavy elements got heavier, not lighter: they captured the bombarding neutron, threw off its binding energy by emitting gamma radiation, and thus, with the addition of the neutron’s mass, but with no added or subtracted charge, became a heavier isotope of themselves. Which then decayed by the delayed emission of a negative beta ray to an element with one more unit of atomic number. Uranium did the same; after a delay it emitted a beta electron. That should mean, Fermi realized, that bombarding uranium with neutrons was producing first a heavier isotope, uranium 239, and then a new, man-made transuranic element, atomic number 93, something never seen on earth before.
It was necessary to purify their uranium sample (uranium nitrate in solution, a light yellow liquid) of the obscuring beta activity its natural decay products gave off. (Uranium decays naturally through a series of fourteen complex steps down the periodic table to thorium, protactinium, radium, radon, polonium and bismuth to lead.) Trabacchi in his generosity had by then even lent the group a young chemist, Oscar D’Agostino, fresh from training in radiochemistry on the Rue Pierre Curie; D’Agostino accomplished the laborious purification in early May. They were using stronger sources now, up to 800 millicuries of radon, about a million neutrons per second.797 Irradiating the uranium nitrate gave “a very intense effect with several periods [of half-lives]: one period of about 1 minute, another of 13 minutes besides longer periods not yet exactly determined”—thus their May 10 report.798
These several induced radioactivities were all beta emitters. They made whatever atom was emitting them heavier by one atomic number. It seemed to follow, then, that they were transmutations up the periodic table into the uncharted new region of man-made elements. To confirm that stunning possibility Fermi needed to demonstrate with chemical separations that the neutron bombardment was not unaccountably creating elements lighter than uranium. The one-minute half-life was too short to work with, so he concentrated on the thirteen-minute substance. D’Agostino diluted the irradiated uranium nitrate with 50 percent nitric acid, dissolved into the acid a small amount of manganese salt and set the solution to boil. By adding sodium chlorate to the boiling solution he precipitated crystals of manganese dioxide. When he filtered the crystals from the solution the radioactivity went with the manganese, much as the radioactivity the Joliot-Curies had induced in aluminum went off with the hydrogen gas. If the radioactivity could be precipitated out of the uranium solution along with a manganese carrier, then it must not be uranium anymore.
By adding other carriers and precipitating other compounds D’Agostino proved that the thirteen-minute substance was neither protactinium (91), thorium (90), actinium (89), radium (88), bismuth (83) nor lead (82). Its behavior excluded elements 87 (then known as ekacesium), and 86 (radon). Element 85 was unknown. Perhaps because the half-lives were different, Fermi made no attempt to check polonium (84). But he felt he had been sufficiently thorough. “This negative evidence about the identity of the 13 min-activity from a large number of heavy elements,” he reported cautiously in Nature in June, “suggests the possibility that the atomic number of the element may be greater than 92.”799
Corbino injudiciously announced “a new element” at the annual convocation, the King of Italy in attendance, that closed the academic year, which set the press baying and gave Fermi a few sleepless nights.800 Having so splendidly accomplished Szilard’s “rather boring task,” the weary physicist was happy to depart after that with his wife and their small daughter Nella for a summer lecture tour sponsored by the Italian government through Argentina, Uruguay and Brazil.
* * *
Leo Szilard had emerged from his bath that spring of 1934 to pursue his favorite causes, not yet joined, of releasing the energy of the nucleus and of saving the world. In a late-April memorandum condemning the recent Japanese occupation of Manchuria he seemed to look ahead to a far future: “The discoveries of scientists,” he wrote, “have given weapons to mankind which may destroy our present civilization if we do not succeed in avoiding further wars.”801 He probably meant military aircraft; the horrors of strategic bombing and even its potential for deterrence through a balance of terror were much bruited at mid-decade. But almost certainly he was also thinking of atomic bombs.
Several weeks earlier, looking for a patron, he had sent Sir Hugo Hirst, the founder of th
e British General Electric Company, a copy of the first chapter of The World Set Free. “Of course,” he wrote Sir Hugo with a touch of bitterness, still brooding on Rutherford’s prediction, “all this is moonshine, but I have reason to believe that in so far as the industrial applications of the present discoveries in physics are concerned, the forecast of the writers may prove to be more accurate than the forecast of the scientists. The physicists have conclusive arguments as to why we cannot create at present new sources of energy for industrial purposes; I am not so sure whether they do not miss the point.”802
That Szilard saw beyond “energy for industrial purposes” to the possibility of weapons of war is evident in his next patent amendments, dated June 28 and July 4, 1934. Previously he had described “the transmutation of chemical elements”; now he added “the liberation of nuclear energy for power production and other purposes through nuclear transmutation.” He proposed for the first time “a chain reaction in which particles which carry no positive charge and the mass of which is approximately equal to the proton mass or a multiple thereof [i.e., neutrons] form the links of the chain.”803 He described the essential features of what came to be known as a “critical mass”—the volume of a chain-reacting substance necessary to make the chain reaction self-sustaining.804 He saw that the critical mass could be reduced by surrounding a sphere of chain-reacting substance with “some cheap heavy material, for instance lead,” that would reflect neutrons back into the sphere, the basic concept for what came to be known (by analogy with the mud tamped into drill holes to confine conventional explosives) as “tamper.” And he understood what would happen if he assembled a critical mass, spelling out the results simply on the fourth page of his application:805
If the thickness is larger than the critical value . . . I can produce an explosion.
As if to mark in some distant inhuman ledger the end of one age and the beginning of another, Marie Sklodowska Curie, born in Warsaw, Poland, on November 7, 1867, died that day of Szilard’s filing, July 4, 1934, in Savoy. Einstein’s was the best eulogy: “Marie Curie is,” he said, “of all celebrated beings, the only one whom fame has not corrupted.”806
There is nothing in the documentary record to indicate that Szilard was yet thinking of uranium. His June amendment describes a possible chain reaction using light, silvery beryllium, element number 4 on the periodic table.
To study that metal Szilard needed access to a laboratory and a source of radiation. The beryllium nucleus was so lightly bound he suspected he could knock neutrons out of it not only with alpha particles or neutrons but even with gamma rays or high-energy X rays. Radium emitted gamma rays and radium was available conveniently at the nearest large hospital. So Szilard, an unusually practical visionary, dropped in to see the director of the physics department at the medical college of St. Bartholomew’s Hospital. Couldn’t he use St. Bart’s radium, “which was not much in use in summertime,” for experiments? Something of value to medicine might emerge.807 The director thought he could if he teamed up with someone on the staff. “There was a very nice young Englishman, Mr. [T. A.] Chalmers, who was game, and so we teamed up and for the next two months we did experiments.”
Their first experiment demonstrated a brilliantly simple method for separating isotopes of iodine by bombarding an iodine compound with neutrons. They then used this Szilard-Chalmers effect (as it came to be called), which was extremely sensitive, as a tool for measuring the production of neutrons in their second experiment: knocking neutrons out of beryllium using the gamma radiation from radium. “These experiments,” Szilard reminisces wryly, “established me as a nuclear physicist, not in the eyes of Cambridge, but in the eyes of Oxford. [Szilard had in fact applied to Rutherford that spring to work at the Cavendish and Rutherford had turned him down.] I had never done work in nuclear physics before, but Oxford considered me an expert. . . . Cambridge . . . would never had made that mistake. For them I was just an upstart who might make all sorts of observations, but these observations could not be regarded as discoveries until they had been repeated at Cambridge and confirmed.”808, 809
If Szilard’s summer work helped establish his Oxford reputation, it was also a personal disappointment: beryllium proved an unsatisfactory candidate for chain reaction. The problem, not settled until 1935, lay with the established mass of helium.810 The one stable isotope of beryllium consists of two helium nuclei lightly bound by a neutron. Its apparently high mass, which was calculated from Francis Aston’s measurements of the mass of helium, seemed to indicate that it should be unstable. But the mass spectrograph was a skittish instrument even in the hands of its inventor, and as Bethe, Rutherford and others were about to demonstrate, Aston’s measurements were inaccurate: he had set the mass of helium too high. One casualty of that error was beryllium’s candidacy for chain reaction, for nuclear power and atomic bombs.
* * *
Emilio Segrè and Edoardo Amaldi pilgrimaged to Cambridge early in July, short on English but carrying with them a comprehensive report on the Rome neutron-bombardment investigations.811 They met Chadwick, Kapitza and the other regulars at the Cavendish; observed the retired J. J. Thomson making his rounds; noted Aston, says Amaldi innocently, “going on improving the accuracy of his measurements of atomic masses”; and had a memorable meeting with Rutherford, whose “strong personality dominated the whole laboratory.”
The two young physicists had come to compare experiments with two of Rutherford’s boys. An unanswered question hung over the neutron work, a question that called existing nuclear theory into doubt.812 The Nature paper they brought with them discussed the difficulty frankly. It concerned what is called “radiative capture,” the typical reaction of the heavy elements to neutron bombardment: a nucleus captures a neutron, emits a photon of gamma radiation to stabilize itself energetically and thus becomes an isotope one mass unit heavier.
Theory at the time treated the nucleus as if it was one large particle. As such, it had a definite diameter, which was modest enough that a speeding neutron could go in one side and exit out the other in about 10-21 seconds, a billion times less than a trillionth of a second. Any capture process would have to work within that brief interval of time. Otherwise the neutron would be gone. Capturing a neutron means stopping it within a nucleus. To do that the nucleus has to absorb the neutron’s energy of motion. The nucleus in turn has to get rid of the excess energy. Which it does: by emitting a gamma photon.
But the gamma-emission times Fermi’s group had measured were different from what theory said they ought to be. The nuclei the Rome group had studied took at least 10-16 seconds to get around to gamma emission—one hundred thousand times too long. And that was unaccountable.
Definite proof of radiative capture would sharpen the challenge to theory. That required proving beyond doubt, by experiment, that a heavier isotope really forms when a heavy nucleus captures a neutron. The Cavendish team Segrè and Amaldi came to visit in the summer of 1934 accomplished the first part of the proof, using sodium, while the Italians were on hand. They then returned to Rome and enlisted D’Agostino’s help to perform the confirming chemistry. In the heat of Roman August they looked for additional clear-cut examples and won a double prize: “We also found a second case of ‘proven’ radiative capture,” Amaldi writes, “which was based on the discovery of a new radioisotope of [aluminum] with a lifetime of almost 3 minutes.”813
Fermi planned to stop off in London for an international physics conference on his way home from South America. His young colleagues sent him word of their aluminum discovery. He reported to the conference on the neutron work. (Szilard also attended, happy to hear praise for his summer experiments and well launched toward a paying fellowship at Oxford.) Fermi said his group had studied sixty elements so far and had induced radioactivity in forty of them. Discussing the radiative-capture problem he cited the Cavendish results “and those of Amaldi and Segrè on aluminium,” which were both, he said, “to be considered particularly important.”814, 8
15 Segrè describes the tempestuous aftermath:
Shortly afterwards I caught a cold and could not go to the laboratory for several days. Amaldi tried to repeat our experiments and found a different [half-life] for irradiated aluminum which showed that our so-called (n,γ) reaction [i.e., neutron in, gamma photon out] did not occur. This was hurriedly relayed to Fermi who resented having communicated a result which now looked to be in error. He strongly criticized us and did not conceal his displeasure. The whole business was becoming very troublesome because we could not find any fault with the various experiments which gave inconsistent results.816
The chastened junior members had their work cut out for them. A new recruit joined them, a tall, broad-shouldered, handsome tennis champion from Pisa named Bruno Pontecorvo, as they set about polishing their first rough work. Neutron bombardment activated some elements more intensely than others. They had previously categorized that activation only generally as strong, medium or weak. Now they proposed to establish a quantitative scale of activibility. They needed some standard intensity against which to measure the intensity of other activations. They chose the convenient 2.3-minute half-life period that neutron bombardment induced in silver.
Amaldi and Pontecorvo got the assignment. They immediately found, to their surprise, that their silver cylinders activated differently in different parts of the laboratory. “In particular,” writes Amaldi, “there were certain wooden tables near a spectroscope in a dark room which had miraculous properties, since silver irradiated on those tables gained much more activity than when it was irradiated on a marble table in the same room.”817