«national academy of sciences edWin mattison mcmillan 1907—1991 A Biographical Memoir by J. david Jackson and W.k.H. Panofsky Any opinions expressed ...»
national academy of sciences
edWin mattison mcmillan
A Biographical Memoir by
J. david Jackson and W.k.H. Panofsky
Any opinions expressed in this memoir are those of the author(s)
and do not necessarily reflect the views of the
National Academy of Sciences.
NatioNal aCademies press
Courtesy of the Lawrence Berkeley Laboratory
E D W I N M AT T I S O N M C M I L L A N
September 18, 1907–September 8, 1991 BY J. DAVID JACKSON AND W. K. H. PANOFSKY W Edwin Mattison McMillan on SepITH THE DEATH OF tember 8, 1991, the world lost one of its great natural scientists. We advisedly use the term “natural scientist” since McMillan’s interests transcended greatly that of his profession of physicist. They encompassed everything natural from rocks through elementar y particles to pure mathematics and included an insatiable appetite for understanding everything from fundamental principles.
Edwin McMillan spent a large part of his professional life in close association with Ernest O. Lawrence1 and succeeded Lawrence as director of what is now the Lawrence Berkeley Laborator y in 1958. Yet the two men could hardly be more different. Lawrence was a man of great intuition, outgoing, and a highly capable organizer of the work of many people.
Edwin McMillan was thoroughly analytical in whatever he did and usually worked alone or with few associates. He disliked specialization and the division of physics divided into theory and experiment. He remarked at an international high-energy physics meeting, “Any experimentalist, unless proven a damn fool, should be given one half year to interpret his own experiment.” McMillan’s first and last publications illustrate the unusual breadth of his interests. While still an undergraduate 216 BIOGRAPHICAL MEMOIRS student in 1927, he published a paper2 on the x-ray study of alloys of lead and thallium, clearly a topic in chemistry.
At the time, he took many more courses in chemistry than was customar y for a physics major, and this publication was undertaken at the suggestion of Linus Pauling. His last paper, 3 written together with the mathematician Richard P.
Brent, was on an improved algorithm for computing Euler’s constant: the limit of the difference between the sum of the inverse integers from 1 to n and the natural logarithm of n, as n → ∞.
One of us (J.D.J.) recalls an incident that illustrates Ed McMillan’s range in science. When Jackson corresponded at the beginning of 1957 with Luis Alvarez and his colleagues about muon-catalyzed fusion, he was startled to receive facsimile copies of handwritten notesby McMillan on a calculation of the mu-mesic molecular formation process!
At that time, he knew McMillan’s name as the discoverer of neptunium, the codiscoverer of plutonium, and the inventor of phase stability in accelerators but never dreamt that he was a molecular theorist! At the time, Ed was busy as associate director under Lawrence. His molecular physics Ph.D. thesis research with Condon could be the origin of such expertise, but with McMillan it could just as easily be knowledge acquired for the fun of it.
The son of Edwin H. McMillan and Anna Maria Mattison, Edwin M. McMillan was born on September 18, 1907, in Redondo Beach, California; both parents were Scots. He was brought up in Pasadena, California, beyond age one and a half. His father was a physician, as were the parents of his wife Elsie McMillan (born Blumer), who incidentally is the sister of E. O. Lawrence’s wife, Molly. McMillan is survived by his wife and their three children (Ann Bradford Chaikin, David Mattison McMillan, and Stephen Walker McMillan). They were a wonderful and harmonious family.
EDWIN MATTISON MCMILLANAs a child, McMillan built gadgets and made use of the proximity of the California Institute of Technology in attending lectures and seminars and getting acquainted with physicists there. After high school McMillan entered CalTech, where he had a first-rate academic record leading to both the B.S. and M.S. degrees. He completed his work leading to the Ph.D. at Princeton University in 1932.
McMillan’s work can be separated into five phases that exhibit a great deal of overlap—not surprising considering the universality of McMillan’s interests: (1) the early prewar period; (2) studies of the transuranic elements; (3) military work during World War II; (4) accelerator physics; and (5) laboratory director. These phases were paralleled by work on advisory committees and other roles as a statesman of science.
THE EARLY PREWAR PERIODMcMillan’s Ph.D. thesis, under Professor E. U. Condon, examined the generation of a molecular beam of hydrogen-chloride nuclei in a nonhomogeneous electric field.4 In parallel, McMillan received a thorough education in experimental nuclear physics at Princeton. He published a paper5 on the isotopic composition of lithium in the sun from spectroscopic observations immediately after receiving his Ph.D. He then won a highly prized National Research Council (NRC) fellowship, supporting him at any university of his choice.
He accepted the invitation of E. O. Lawrence to come to Berkeley, where Lawrence was at the time engaged in exploring the experimental potential of the cyclotron. After McMillan accepted Lawrence’s invitation, he dedicated his first two years to activities somewhat separate from the mainstream activities of Lawrence’s new Radiation Laboratory.
He intended to measure the magnetic moment of the proBIOGRAPHICAL MEMOIRS ton, but that plan came to naught when Otto Stern and collaborators in Germany did the measurement. He continued to work on hyperfine structure as revealed in optical spectroscopy and published papers on the nuclear magnetic moment of tantalum6 as well as on the hyperfine structure of the solar spectrum. 7 But McMillan became progressively more involved with the work on Lawrence’s cyclotron, which by early 1934 could produce a deflected beam of 2.3-MeV deuterons. His experimental skill was recognized by Lawrence and his collaborators and was put to increasing use on both the cyclotron and its instrumentation and physical experiments with the beam.
McMillan used the extracted deuteron beam in collaboration with M. Stanley Livingston to irradiate nitrogen to produce the positron emitting 15O. Again, McMillan’s skill as a chemist was put to work. He used a tracer technique in which first nitrogen gas was bombarded and then mixed with oxygen and an excess of hydrogen. This mixture was catalyzed to water over heated platinized asbestos, and the water was collected on anhydrous calcium chloride. The radioactivity was shown to be localized in the calcium chloride and absent elsewhere, proving that oxygen carried the activity.8 This work was followed by fundamental studies on the absorption of gamma rays,9 which revealed the (at that time new) process of electromagnetic pair production in the Coulomb field of a nucleus. The 5.4-MeV gamma ray produced by bombardment of fluorine with protons and also the gamma rays of other isotopes were absorbed by foils of aluminum, copper, tin, and lead, enabling McMillan to isolate the components of the absorption process. At 5.4 MeV, electron-positron pair production is about one-half the total absorption cross-section in lead.
In 1935, with Lawrence and R. L. Thornton, McMillan
EDWIN MATTISON MCMILLANstudied the radioactivity produced when a variety of targets are exposed to a deuteron beam.10 At deuteron energies below 2 MeV, the activity increases rapidly with energy, as expected from the quantum mechanical penetration of the Coulomb barrier, first used to explain alpha radioactivity lifetimes by George Gamow. The experiments of McMillan and coworkers on (d,p) reactions with energies up to 3.4 MeV showed that the yield curves flattened above 2 MeV, even though the Coulomb barrier effects were expected to be considerably steeper from conventional estimates of the effective nuclear radii. A deuteron seemed to be able to have its neutron captured by the target nucleus while its proton remained relatively far away. These data intrigued J.
Robert Oppenheimer and his student, Melba Phillips, who then developed the theoretical explanation of the phenomenon: the small binding energy, and therefore large size, of the deuteron permits it to be polarized in the nuclear Coulomb field; this polarization places the neutron within the deuteron close to the nucleus, accessible for capture, while the proton is away from it. In essence, the proton becomes a “spectator” of the process. The Oppenheimer-Phillips process gives a quantitative explanation of the energy independence of the yield curves and the predominance of the (d,p) reactions in deuteron bombardments.
Following this work McMillan investigated the properties of 10 Be, with its extraordinarily long half-life for a light element (approximately 2.5 million years). He pursued further details of the properties of 10 Be in later publications.11 During that period McMillan did several additional experiments in what today has become nuclear chemistr y, some of them successful and some unsuccessful. At the same period, he wrote a seminal paper12 on the production of X rays by the acceleration of very fast electrons, a subject in which he maintained a lifelong interest.
220 BIOGRAPHICAL MEMOIRS McMillan made numerous experimental contributions to the cyclotron, in particular to its beam-focusing properties, to beam extraction, and to vacuum gauges. His deep understanding of the factors that limit the energy attainable by conventional cyclotrons is illustrated by his correspondence in late 1937 and early 1938 with Hans Bethe. Bethe had worked with M. E. Rose at Cornell on the energy limit problem, and McMillan was carrying out calculations at Berkeley with Robert R. Wilson developing orbit-tracing methods. In 1937 Bethe sent an advance copy of the Bethe-Rose paper to McMillan. McMillan found some errors in the paper and showed that the electrostatic defocusing effect of the cyclotron dee’s could be counteracted by the insertion of grids. McMillan also understood clearly the focusing effect of the radial fall-off of the magnetic field and the magnitude of the deviation from the synchronicity condition in the cyclotron produced by that radial fall-off, added to the relativistic mass increase. Bethe suggested that McMillan publish his findings, but characteristically McMillan felt that an additional paper would be redundant. The correspondence demonstrates McMillan’s deep quantitative mastery of the subject while at the same time exhibiting his basic humility. He preferred making an input to the Bethe-Rose paper over cluttering up the literature with controversy.
STUDIES ON TRANSURANIC ELEMENTSThe discovery of fission of uranium by Hahn and Strassmann in 1939 initiated intense activity worldwide. At Berkeley McMillan first performed a simple experiment to measure the ranges of the energetic fission fragments by exposing a thin layer of uranium oxide on paper sandwiched between several thin aluminum foils on either side to the neutrons from 8-MeV deuterons striking a beryllium target in the 37-inch cyclotron. The amounts of radioactivity in
EDWIN MATTISON MCMILLANsuccessive foils established the maximum range of the fragments as equivalent to approximately 2.2 centimeters in air.
He also used cigarette papers instead of the aluminum foils in another sandwich and followed the radioactivity in different papers after bombardment, finding the same time dependence in all. In contrast, the activity associated with the layer of paper on which the uranium oxide had been placed had different components. In addition to the fission fragment activity, there was one component with a twentyfive-minute half-life and another of roughly two days.
McMillan speculated that the twenty-five-minute activity was 239U, identified earlier by Hahn and co-workers as a product of resonant neutron capture in uranium.13 The two-day nonrecoiling activity intrigued McMillan.
Accordingly, he bombarded thin ammonium uranate layers deposited on a bakelite substrate and covered with cellophane (to catch the energetic fission fragments). After exposure to the neutrons, the ammonium uranate was scraped off the bakelite and its activity followed. At long times the 2.3-day activity was dominant; at short times, the twentythree-minute half-life of 239 U predominated. In contrast, the cellophane showed the characteristic power law decay associated with a mixture of fission fragments of different lifetimes. With the new activity physically separated, it was possible to begin study of its chemical properties. As a putative new element next to uranium, the activity seemed likely to have chemical properties akin to rhenium. McMillan therefore enlisted Emilio Segrè, who was familiar with the chemistry of rhenium from his discovery of a homolog, technetium, in 1937. Segrè found that the 2.3-day activity behaved like a rare earth, not like rhenium. Since rare earths are prominent among the fission fragments, it appeared that the 2.3-day activity was one of those. After a gap in his pursuit, McMillan had become persuaded by early 1940 that 222 BIOGRAPHICAL MEMOIRS the nonrecoiling 2.3-day activity just could not be the decay of a fission fragment. He began a set of experiments with the new 60-inch cyclotron and its 16-MeV deuterons. Two obser vations confirmed his belief as a certainty. One, using cadmium absorbers to reduce the thermal neutrons, showed greatly reduced fission activity but left the two nonrecoiling activities in the same relative proportion. The other, a fission product experiment with extremely thin collodion catcher foils, showed that the range of the 2.3-day “fragments” was less than 0.1 millimeter of air equivalent. The 2.3-day activity could not be from fission; the twenty-threeminute and 2.3-day activities almost certainly were genetically related. The beta decay of 239U was producing atoms of a new element with Z = 93! McMillan found chemically that the 2.3-day activity had some, but not all, the characteristics of a rare earth.