A Productive Error

The simplest of the new isotopes recognized by physicists in the 1930s was the nucleus of heavy hydrogen, a compound of a proton and a neutron. This deuteron made a particularly interesting projectile for the cyclotron since it does not occur among the rays from radioactive substances. Therefore, Lawrence, G. N. Lewis, and M. S. Livingston had no definite expectation, but many high hopes, when they loosed the world's first high speed deuterons at a lithium target in March 1933. They found the deuteron to be ten times as effective a disintegrator as the proton. It offered a glimpse into the millenium; it appeared to release much more energy than it brought. When striking lithium it produced the fastest alpha particles ever observed. When striking all other substances, it coaxed out protons with a range of 18 cm in air. It began to seem too much of a good thing. Positively charged deuterons of a few million electron volts could not be expected to penetrate the Coulomb barrier of so heavy a nucleus as gold even by the tunneling discovered by George Gamow, E. U. Condon, and R. W. Gurney. Lawrence's group concluded that the fast protons they found came not from target nuclei, but from the disintegration of the deuteron. They accordingly announced that the nuclei of heavy hydrogen are unstable and decompose in the electric fields inside atoms.

This "disintegration hypothesis" was made public by Lawrence in May 1933 at a symposium at Caltech in honor of Niels Bohr. It had far-reaching consequences. First, the reaction Lawrence proposed implied that James Chadwick had grossly overestimated the mass of the neutron (mn). Chadwick assumed that his neutrons had come into being when a boron nucleus absorbed an alpha particle according to the reaction B11(alpha,n)N14. He computed the neutron mass by subtracting the kinetic energy of the neutron and the rest energy of nitrogen-14 from the total energies of the incident particle and the target atom. The result, according to the crude contemporary values of atomic masses: mn =1.0067 mp, where mp is the proton mass. Lawrence and his colleagues worked from the deuteron disintegration hypothesis, d -> n + p; and they fixed the neutron mass at 1.0006mp from the known rest masses and the hypothesis that the 18 cm proton carried away all of the deuteron's kinetic energy. (Neither value agrees with the modern one, m = 1.0014 mp.) Second, the implied instability menaced the theory Werner Heisenberg had devised on the assumption that the nucleus contains only neutrons and protons.

Since nuclear theory would disintegrate along with the deuteron should Lawrence's hypothesis hold, the two principals at the Caltech celebration welcomed it, although for different reasons. Niels Bohr, always hoping for conceptual revolution, extolled it as "a marvelous advancement" and the cyclotron as "the dream of yesterday...come true." Robert Millikan, nationalistic and opinionated, enjoyed the discomfiture of European theorists and praised the deuteron hypothesis as "altogether extraordinary [which it was] and most intelligently announced." Shortly after the meeting in Pasadena Lawrence appeared at the Century of Progress Exhibition in Chicago as the creator of a "new miracle of science, the most powerful cannon yet found for liberating relatively enormous stores of energy locked up in the inner core of the atom." The newspaper account of Lawrence's speech for Progress concluded: "The newest developments give only an inkling of what lies in store."
Participants at seventh Solvay congress in Brussels, Belgium, in October, 1933. Lawrence is standing second from right; Rutherford is sitting sixth from right, Chadwick on far right; Bohr third from left; Heisenberg stands fourth from left, and Cockcroft fifth from right.

What lay in store was a tough time. In October Lawrence brought his results before the seventh Solvay Congress in Brussels. Attendance was a great honor; Lawrence was only the eighth American ever invited, and the sole one for 1933. He did not, however, have the burden and distinction of presenting a full report. He appended a few pages on the operation of cyclotrons and the disintegration of deuterons to a lengthy account of Cambridge work on accelerators. The author, John Cockcroft, declined to entertain Lawrence's hypothesis on the ground that too little data yet existed; but at the end of his report he allowed himself to hint that the cyclotron might be a wasteful and unnecessary machine.

The Cavendish physicists had come to Belgium in strength. After Cockcroft, Lawrence faced Ernest Rutherford, who declared that- no neutrons come from lithium under deuteron bombardment, and Chadwick, who insisted that the mass of the neutron is exactly what he had said. Then came the theorists. Heisenberg observed that if disintegration occurred in the electric field of the nucleus, the yield should decline for heavy targets since the deuteron's penetration, and hence the rate of change of force on it, must decrease with increasing atomic number.

That being the case, added Bohr, we might suppose that the deuteron splits after entering a nucleus, but then the speed of the ejected proton should increase with atomic number like the nuclear Coulomb field, contrary to Lawrence's results.

The debate continued when Lawrence stuck his head in the Cavendish lion's den on the way back to Berkeley. Cockcroft, Marcus Oliphant, and Rutherford all dismissed the deuteron hypothesis and advised Lawrence to look for contamination of his targets or his tank. Back in the friendly West, Lawrence hastily reviewed the possibility of systematic contamination with the help of chemical colleague Lewis. The resulting paper, sent to the Physical Review in December 1933, should have convinced "the most skeptical [according to Lawrence] that the deuteron is energetically unstable and disintegrates into a proton and a neutron." Some friends, for example Jesse Beams, thought Lawrence's answer decisive. Others, including Charles Lauritsen and Merle Tuve, repeated the experiments with immaculate apparatus and did not find Lawrence's protons. Then Rutherford found the fast protons, but only after deliberately contaminating his targets with deuterium. Lawrence's conviction waned. By April 1934, with artificial radioactivity claiming his attention, he had discarded the deuteron disintegration hypothesis.

The Cavendish had known about the deuteron hypothesis almost from its inception, through Lewis and through Cockcroft, who reported it after a visit to the Laboratory in June 1933. Cockcroft doubted the hypothesis and urged his colleagues to check it using the heavy water that Lewis had presented Rutherford in May, with generosity typical of Berkeley scientists. Not having on hand a machine capable of accelerating Lewis's gift to millions of electron volts, Rutherford sent alpha particles from polonium through heavy water in order to simulate a cyclotron's bombardment of helium by deuterons. He found no trace of the reactions expected on the disintegration hypothesis. Then, together with Oliphant and Paul Harteck, who prepared more heavy water, Rutherford neatly identified the agent of Lawrence's contamination. When protons fell on targets containing hydrogen, such as NH4 Cl, they gave precisely the same effect whether the hydrogen was ordinary or heavy; but when deuterons fell on the same targets, those containing deuterium gave a prodigious number of protons and only slightly fewer neutrons. The Cavendish group inferred that deuterons interact readily, and in two different ways, to make two isotopes previously unknown: the fusion of deuterons produces either hydrogen-3 (tritium) and a proton, or helium-3 and a neutron. According to their estimate of the masses, hydrogen- 3 should be stable and helium-3 should not.
Top: Schematic diagram of Lawrence's hypothesis for disintegration of the deuteron to a proton and a neutron in the electric field of a nucleus. Bottom: Rutherford's proposal that two colliding deuterons decay either to hydrogen-3 or helium-3, yielding protons and neutrons.

In explanation of his error, Lawrence told Beams that the cyclotron had been so prolific of results that it discouraged "methodical, quantitative measurements," or thorough investigation of one thing before running to another. Another cause for adherence to the disintegration hypothesis was the good press that accompanied its disclosure. Not only did the press's flattery blunt Lawrence's self- criticism, it also assisted his appeal for money. After his refutation of the objections of the Europeans, Lawrence wrote the Research Corporation that the explosion of the deuteron pointed both to a source of fuel and to the reform of nuclear theory: "This first definite case of an atom that itself explodes when properly struck is of great interest, not only as a possible source of atomic energy, but especially because it is not understandable on contemporaneous theories. The fact that the deuteron is energetically unstable promises to be a keystone for a new theoretical structure." What was required to place the stone, as Lawrence had told the Century of Progress Exhibition, was an "atomic gun which in comparison with the present [27-inch] one will be like a 16-inch [diameter] rifle alongside a mere one-pounder."

The conclusion of Rutherford's group, that tritium should be stable, set the proprietors of the world's best mass spectrographs to look for the heaviest hydrogen in pure samples of heavy water. Walker Bleakney and his associates electrolyzed 75 tons of ordinary water down to its heaviest cubic centimeter, which they tested in their sensitive spectrometer at Princeton. They found traces of mass-5 particles, which they declared to be molecules made of an atom each of hydrogen-2 and hydrogen-3; and they guessed that hydrogen-3 constitutes about one part in a billion of ordinary water. Rutherford's group tried to confirm the finding with 11 cubic centimeters of the heaviest remainder of the electrolysis of 13,000 tons of water. The heavy part of the job was done by Norsk Hydro, and the lighter part by the dean of mass spectroscopists, Francis Aston. The 13,000 tons of Norwegian water contained not a drop of hydrogen-tritium oxide. Several groups in England and the United States then tried to make tritium by fusing deuterons. Again the Princeton physicists succeeded; again the British failed. Rutherford wondered whether Americans knew how to do experimental physics.

Since Rutherford thought that tritium is stable, he required a reason why he could not obtain it from the plentiful interactions of deuterons. His answer: tritium disappears quickly by combining with the bombarding deuterons. As for helium-3 the consensus, as represented by H. A. Bethe, held it to be unstable, decaying into the elusive tritium by electron capture. It was precisely with this preconception -- that tritium is elusive but stable and helium-3 is radioactive -- that Luis Alvarez went to look for them in the summer of 1939.

The occasion of Alvarez's investigation was the temporary idleness of the just-completed 60-inch cyclotron, which had not yet acquired the shielding necessary for its high-energy work. He thought to use deuterons from the 37-inch machine to make helium-3 to feed the 60-inch, which would serve as a mass spectrometer. As a preliminary, to test whether the 60-inch would adventitiously deliver stray particles at the setting of the magnetic field for resonant acceleration of particles of mass 3, he had the operating crew reduce the field, which had been set and shimmed for alpha particles (helium-4), by one quarter. Nothing happened: apparently there would be no noise to complicate the experiment. To confirm the finding, Alvarez asked the crew to drop the field from full strength to zero and bring it back to the value of interest. They obliged, but, contrary to usual procedure, left on the radiofrequency oscillator when they switched off the field. As it dropped, Alvarez, still looking at the oscilloscope connected with the ionization chamber, saw a burst of particles apparently of mass 3. Eddy currents set up by the rapidly changing field had providentially shimmed the machine to accelerate ions of helium-3. All this happened before the intended experiment began, using the ion source installed for testing the 60-inch cyclotron. The source employed ordinary helium from a deep well in Texas, where it had lain for geologic ages. The radioactivity of helium-3 had been greatly exaggerated. In fact, as Alvarez declared, it is stable.
"Cornog's robot," a product of downtime on the cyclotron, ca. 1939. Model unknown.

The stability of helium-3 implied the radioactivity of tritium. Alvarez tested this inference with the help of Robert Cornog, a graduate student who worked on the oil vapor vacuum pumps for the 60-inch machine. They routed the issue of heavy water irradiated with deuterons into an ionization chamber attached to an amplifier and found a long-term activity whose carrier behaved like hydrogen The number of active atoms agreed roughly with the number of neutrons produced in the bombardment, confirming the formation of tritium and a proton from two deuterons. The new isotope was long- lived. No appreciable decay could be detected in a sample imprisoned for five months.

It happened that an estimate of the half life of tritium already existed. It had been made at the Rad Lab in 1936 by McMillan, who had followed the decline of a beryllium-aluminum target that had been used to make neutrons by the reaction Be9(d,n)B10. He found an activity whose half life he put at over ten years. After the disclosure of the true nature of tritium, R. D. O'Neal and Maurice Goldhaber at the University of Illinois corrected McMillan's guess that the 10-year activity belonged to beryllium-10. They found no activity in beryllium from a target similar to McMillan's, but they acquired a nice radioactive gas on dissolving a bit of the target in acid. They supposed that they dealt with a product of the reaction Be9(d,H3)Be8. Alvarez and Cornog accepted the suggestion and increased their estimate of half life accordingly.
Milt White beside the 60-inch cyclotron with which Alvarez showed the stability of helium-3.

Rutherford reported his negative results in his last scientific paper, published in August 1937. Had his Norwegian collaborators concentrated whatever tritium they might have gathered by repeated electrolysis of heavy water, his sample would have been radioactive from trace amounts of tritium made in the atmosphere by cosmic rays and brought down in rain. The half life of tritium is 12.6 years. At the end of the second world war Willard Libby easily detected the residual activity of the sample with a Geiger counter. Apparently no one at the Cavendish laboratory in Rutherford's time had thought to make such a measurement.

CHAPTER 3 Deflecting Physics for War