sign in sign up
<<
Home , Discovery of the neutron, Fritz Strassmann, Margrethe Bohr

The discovery of fission Otto R. FrischJohn A. Wheeler

Citation: Physics Today 20, 11, 43 (1967); doi: 10.1063/1.3034021 View online: http://dx.doi.org/10.1063/1.3034021 View Table of Contents: http://physicstoday.scitation.org/toc/pto/20/11 Published by the American Institute of Physics

Articles you may be interested in On the belated discovery of fission Physics Today 68, (2015); 10.1063/PT.3.2817

The Discovery of Physics Today 42, (2008); 10.1063/1.881174 The Discovery of Fission

Initial formulations of nuclear fission are colored with the successes, failures and just plain bad luck of several scientists from different nations. The winning combination of good fortune and careful thought made this exciting concept a reality.

by Otto R. Frisch and John A. Wheeler

How It All Began by Otto R. Frisch

THE was discovered in 1932. had little respect for theory. Once, Why, then, did it take seven years be- when one of her students suggested an fore nuclear fission was found? Fission experiment, adding that the theoreti- is obviously a striking phenomenon; it cal next door thought it results in a large amount of radioactiv- hopeful, she replied, "Well, we might ity of all kinds and produces fragments try it all the same." Their disregard that have more than ten times the total of theory may have cost them the dis- Otto R. Frisch, professor of natural ionization of anything previously covery of the neutron. philosophy (physics) at Cambridge known. So why did it take so long? Cambridge is the second place wor- University, England, did research in (1927-30), (1930- The question might be answered best thy of discussion. , 33), London (1933-34), Copen- by reviewing the situation in whose towering personality dominated hagen (1934-39) and Birmingham from an experimentalist's point of Cambridge research, had split atomic (1939-40). During the war he view. nuclei in 1919; since 1909 he had, in worked on the A-bomb at Los fact, been keenly concerned with the Alamos. He was first to observe Research in Europe energy liberated in the fission of a observation and counting of individual single nuoleus. In Europe there were few laboratories nuclear particles. He first introduced in which nuclear-physics research was the scintillation method and stuck conducted, and I think the word firmly to it. His great preference was "team" had not yet been introduced for simple, unsophisticated methods, into scientific jargon. Science was and he possessed a strong distrust of still pursued by individual scientists any complicated instrumentation. who worked with only one or two stu- Even in 1932, when John Cockcroft dents and assistants. and Ernest Walton first disintegrated Paris harbored some of the most ac- nuclei by artificially-accelerated pro- tive research laboratories in Europe. tons, they used scintillations to detect It is the city in which radioactivity the process. By that time Rutherford had been discovered and where Ma- had realized that electronic methods dame Curie was working until her of particle counting must be devel- John A. Wheeler, one of the first death in 1934. She still dominated oped. The reason was that the scin- American scientists to concentrate the situation: Techniques were quite tillation method clearly had its short- on nuclear fission, worked at the similar to those used at the turn of the comings. It did not work for very low U. of in 1934 as a century; that is, ionization chambers or high counting rates and was not National Research Fellow with really reliable. This deficiency was . Wheeler received his and electrometers. This state of af- PhD in physics at the Johns fairs is good enough for performing ac- highlighted by the results that came Hopkins University prior to his curate measurements on natural ra- from the third laboratory I want to research in Copenhagen. In 1938 dioactive elements, but it is not really mention—Vienna. he joined Princeton's physics de- partment, where he remains active. adequate for much of the work on nu- Vienna is where I began my career clear disintegration. Madame Curie and it was in those days a sort of en-

PHYSICS TODAY • NOVEMBER 1967 • 43 tion method. Even at a few hundred fant terrible of . Sev- dent research. Hahn was working on particles a minute there were quite eral physicists were claiming that not various applications of radioactivity large corrections to be applied. only nitrogen and one or two others of for the study of chemical reactions, the nuclei could be disintegrated structures of precipitates and similar was the first to use by alpha particles but that practically subjects, whereas was the coincidence method, both in an at- all of them could and did give many using radioactive materials chiefly to tempt to do something about cosmic more than anybody else could elucidate the processes of beta and rays and also for measuring the energy observe. I still do not know how they gamma emission and the interaction of of gamma rays by the range of the sec- found these wrong results. Apparent- gamma rays with matter. ondary they produced. This ly they employed students to do the In addition, was in was really the first reliable method for counting without telling them what to . He had been with Ruther- measuring the energy of weak gamma expect. On the face of it, that opera- ford from 1909 onwards, in the early radiations. tion appears to be a very objective days before the nucleus was discov- Until 1932, the only source of par- method because the student would ered. Rutherford felt uncertain about ticles for doing atomic nuclear disinte- have no bias; yet the students quickly the scintillation method and asked gration was natural alpha particles: developed a bias towards high num- Geiger to develop an electric counter either , which was difficult bers because they felt that they would to check on it. But as soon as Ruther- to come by (in fact one practically be given approval if they found lots of ford saw that the two gave the same had to go to Paris) or sources of one of particles. Quite likely this situation results, Rutherford returned to the the short-lived decay products of radi- caused the wrong results along with a scintillation method, which appeared um, which were very clean but were generally uncritical attitude and con- to be simpler and more reliable when short-lived and usually had lots of siderable enthusiasm over beating the used with proper precaution. Geiger gamma radiation. English at their own game. went back to Germany and perfected I still remember when I left Vienna his electric counters, and in 1928, to- The year of discovery at just about that time (after having gether with a student named But in 1932, that annus mirabilis, not escaped the duty of counting scintilla- W. Miiller, he developed an improved only the neutron was discovered but tions). My supervisor, Karl Przibram, counter that could count beta rays. two other developments took place. told me with sadness in his voice, "You Earlier counters were inadequate for In the US Ernest O. Lawrence made will tell the people in Berlin, won't this purpose, and scintillation methods the first that showed prom- you, that we are not quite as bad as were also incapable of detecting beta ise of being useful, and in England they think?" I failed to persuade rays. However the new counters were Cockcroft and Walton built the first them. still very slow because the discharge accelerator for protons capable of pro- Germany had nuclear-physics re- between the central wire and the cy- ducing nuclear disintegrations. I search in several places. The team of lindrical envelope was quenched by a need not state that this was the begin- and Lise Meitner, which large resistor of many megohms placed ning of an enormous development; had been one of the first groups to in the circuit; consequently the count- most of nuclear physics as we know it study radioactive elements, had at that ing rate was limited to numbers not would have never come about without time separated to carry out indepen- much greater than with the scintilla- at least one of those two instruments. But the interesting thing is that they played practically no role in that nar- row thread that led to the discovery of nuclear fission. I do not want to dwell on the dis- 92U+ (92U -f- n) 93EkaRe > 94EkaOs^^^^ (I) ioSek. 2,2 Min. covery of the neutron very much be- cause it was discussed in several inter- —?-> 95EkaIr esting lectures in 1962 at the History 59 Min. 66 Std! 9G JHli of Science Congress held in Ithaca, New York. The published proceed- 97EkaAu 2,5 Std. ings contained interesting contribu- tions by Norman Feather and Sir P , who showed that the (2) 92U + n-> (92U + n) 93EkaRe —•-• 94EkaOs ^ 40 Sek. 94 neutron was discovered in Cambridge, 16 Min. not simply by chance with everybody 95EkaIr ? else having done the groundwork, but 57Std. because a search for the neutron had been going on in Cambridge (admit- tedly with wrong ideas). The people (3) 92U+ (92U + n) 93EkaRe ? 23 Min. at Cambridge were keyed up for this discovery. They had made one obser- COMPUTATIONS, indicating chains of radioactive elements, were published in a vation that was important and that 1938 Die Naturwissenschaften article by Hahn, Meitner and Strassmann. —FIG. 1 tends to be overlooked: H. C. Web-

44 • NOVEMBER 1967 • PHYSICS TODAY ster showed that those queer pene- trating rays that emitted when alpha particles fell on it were more intense in the forward direction than in the backward direction. This result was quite incomprehensible if the radiation were gamma rays as ev- erybody believed. Even the French physicists Curie and Joliot shared that belief in the teeth of all theoretical predictions. Then Chadwick's experi- ment showed clearly that the mysteri- ous radiation consisted of particles hav- ing approximately the of the pro- ton. There was a bit of confusion at the time because the word "neutron" had been used by and Wolfgang Pauli to indicate the particle that later came to be called the "neu- trino." After the neutron was discovered, there was of course a certain rush of activity, but nobody knew quite what to do. were rather few in number. They were, after all, secon- dary products of nuclear disintegra- tion. With only natural alpha sources available at first, neutron production GREAT AND GOOD FRIENDS. Lord and Lady Rutherford (left) with was low. Niels and Margrethe Bohr in Rutherford's garden. The photograph was Moreover the main instrument for taken about 1930. detection was essentially the cloud chamber. With cloud chambers only a limited number of tracks due to neu- had used them with good effect to pin cial radioactivity before the French. trons could be found. And it was down the neutron, were still too noisy It is astonishing that nobody ap- slow work to make any sense out of to be of much use. pears to have thought beforehand that the few detected tracks of recoil nu- the result of a nuclear disintegration clei. once joked that if a Artificial radioactivity might be an unstable nucleus although man suddenly does something unex- Things really got moving when, in the existence of unstable nuclei had, of pected there is usually a woman be- 1934, artificial radioactivity was found course, been known for thirty years or hind it, but if an sud- by Curie and Joliot. I think they must more. I have been told that, after the denly does something unexpected, have been very happy to have made discovery, Rutherford wrote to Joliot there is probably a neutron behind it. up for their failure to spot the neutron and congratulated him on his discov- Electronic counting methods had two years previously. Almost to the ery saying that he himself had thought only just been developed; largely as a day two years previously both discov- that some of the resulting nuclei might reaction to the wrong results coming eries came out in the middle of Janu- be unstable, but had always looked for out of Vienna that nobody else could ary. They had known for many alpha particles only because he was confirm, it had been decided that it months before that aluminum bom- not really interested in beta particles. really was necessary to build - barded with alpha particles emits posi- As soon as this work became known ic amplifiers and counters. Actually trons, but it had never occurred to in January 1934 a lot of people rushed the Viennese themselves started that them that this might be a delayed to repeat and extend the experiment. process. They had only observed the kind of work but were not very suc- But most of them rushed in a straight during bombardment. Law- cessful. The work was also started in line indicated by Curie-Joliot, bom- rence and his cyclotron people in Cali- barding other elements with alpha Switzerland with some success by fornia had made the same mistake. In particles. (So did I in Blackett's labo- Hermann Greinacher. Yet I think the fact they had noticed that the counters ratory in London.) main thread that led to the develop- misbehaved after the cyclotron was But in Fermi at that time had ment of decent counters took place in switched off. I am told that they already decided that nuclear physics England, where Charles Wynn-Wil- built in special gadgetry so that the was an important and interesting line, liams used proper screening and tubes counters were automatically switched and he had started to set up some in- with low noise level etc. to produce off together with the cyclotron! Oth- strumentation. So when this discovery electronic counters. Nevertheless erwise they would have found artifi- came along, he began working quite those counters, although Chadwick

PHYSICS TODAY • NOVEMBER 1967 • 45 fast to see whether neutrons would they must all derive from uranium 238 form radioactive nuclei. or possibly one of them from 235 I remember that my reaction and (which is already much rarer). So it probably that of many others was that looked as if there were at least two Fermi's was a silly experiment because parallel chains of isomeric elements. neutrons were much fewer than alpha This isomeric property had to be prop- particles. What that simple argument agated all along the chain of beta overlooked of course was that they are disintegrations. very much more effective. Neutrons Nuclear isomerism was still fairly are not slowed down by electrons, and new in 1938, and its interpretation they are not repelled by the Coulomb was not altogether clear. It had been field of nuclei. Indeed, within about suggested (as we now accept) that it four weeks of the discovery by Curie was due to high angular momentum, and Joliot, Fermi published the first but there were also proposals that it results proving that various elements might be due to the existence of rigid did become radioactive when bom- structures inside nuclei. One could barded with neutrons. Only another imagine that such a rigid structure month later he announced that bom- might survive a and might barding uranium produced some new influence the half-life of the subse- radioactivity that he felt must be due quent product. to transuranic elements. Because But then there was still the mystery both on theoretical grounds (Coulomb of the great length of those chains. barrier and all that) and as far as the Uranium, after all, was not beta un- experiments confirmed it, all heavier stable itself. The other elements in elements were known to absorb neu- that region never had more than two trons without splitting anything off. beta decays in succession; yet here And so it was felt that must also be the four or five had been found. So Hahn case with uranium. the was delighted by so many This work was of course considera- new elements, but Hahn the radio- bly interesting to radiochemists. Sev- or radiochemist was rather eral took it up, but once again, oddly worried about the mechanism that enough, one false result started things could account for them. really moving—a note by Aristid von All this work was made difficult by Grosse, a German-born chemist work- the political situation in Germany. ing in the US, who thought one of Hitler was in power and the institute these elements behaved like protactin- had to play a delicate game of politics ium. He had done some of the early to prevent racial persecution from re- work with Hahn on protactinium soon moving some of its personnel. In after it was discovered in 1917; so his 1938, when Austria was occupied by suggestion put Hahn and Meitner on the Nazis, Lise Meitner felt very inse- their mettle. They felt protactinium LINKS IN THE CHAIN. Cockcroft cure; rumors began to float around was their own baby and they were (top) and Walton contributed to the that she might lose her post and be going to check it. Lise Meitner per- new ideas when they disintegrated nuclei by artificially-accelerated pro- prevented thereafter from leaving suaded Hahn to join forces again. tons. Germany because of her knowhow. A They soon showed that von Grosse certain amount of panic resulted. was wrong: It was not protactinium. Dutch colleagues offered to smuggle On the other hand there were so many her to Holland without a visa. Thus odd things there that they were cap- ly, Hahn was excited to have a whole she left Germany in the early summer tured by this phenomenon and had to new lot of chemical elements to play of 1938, went from Holland for a brief go on. The results were most pecu- with and to study their properties. stay in Denmark, and was offered hos- liar. Today, of course, these elements after pitality by Manne Siegbahn at the Figure 1 shows one of the tabula- uranium are known as , plu- Nobel Institute in Stockholm. tions indicating the chains of radioac- tonium, americium etc., and are tive elements that Hahn and Meitner known to be chemically quite different Near misses had thought identified them. They from those that Hahn was studying. After that, the team that had already did not give new names to the trans- brought Strassmann in with Hahn as a Parallel chains uranic elements that they thought second chemist had to carry on with- they had identified, but they used the The results were astonishing for two out her. In the meantime some work prefix "eka" to indicate that they were reasons. In the first place, it ap- had been started in Paris. It is inter- higher homologues of rhenium, os- peared that there were three parallel esting that they had a different angle. mium, etc. up to ekagold. Obvious- series. And from the yields obtained They were at first not so interested in

46 • NOVEMBER 1967 • PHYSICS TODAY the transuranic elements; but they realized that if is bombarded with neutrons, one ought to find the beginning of the new and missing ra- dioactive chain with the atomic weight 4n + 1. One realizes that the others, 4n, 4n + 2, 4n + 3, are all represented by the natural radioactive series. But the 4n + 1 was missing, and so Irene Curie, the daughter of Madame Curie, together with Hans von Halban, an Austrian, and Peter Preiswerk, a Swiss, set out to search for that series and published some work on it. RUTHERFORD was the first to use scintillation methods to detect particles. Later that team broke up because Halban came to Copenhagen and, for a time, worked with me on the study of slow neutrons. Irene Curie found a There was another near miss at diminish the surface tension and found new collaborator in Pavel Savitch, a about the same time: Gottfried von that it would be down to zero around Yugoslav. They tried to disentangle Droste, a physicist working with Lise Z = 100 and probably quite small for the transuranic elements. Having Meitner, looked for long-range alpha uranium. Lise Meitner worked out realized that there was a great variety rays from uranium during neutron the energies that would be available of different materials, Irene Curie had bombardment. If he had supressed from the mass defect in such a break- the good idea of selecting one of them the ordinary alpha rays by applying a up. She had the mass defect curve simply by the high penetration of its bias to the amplifier, he would not pretty well in her head. It turned out beta rays. They covered their have failed to find fission. Unfortu- that the electric repulsion of the frag- samples with a fairly thick sheet of nately instead of using a bias he used a ments would give them about 200 brass and only studied the substance foil, and that foil was thick enough to MeV of energy and that the mass de- whose radiation penetrated. They did stop not only uranium alpha rays but fect would indeed deliver that energy not realize that even that method also the fission fragments; nor did he so the process could take place on a might not select a single substance al- find any long-range alpha rays, which purely classical basis without having though the substance appeared to had to be there if radium or actinium to invoke the crossing of a potential have a reasonably unique lifetime of were formed. barrier, which of course could never 3.5 hours. From the chemical behav- Then Hahn and Strassmann checked have worked. ior they first thought it looked like tho- the chemical properties of this "ra- We only spent two or three days to- rium. dium" with care and found that they gether that Christmas. Then I went This work was checked by Hahn, were identical with those of . back to Copenhagen and just managed who concluded that it was not thorium to tell Bohr about the idea as he was and wrote so to Paris. Curie and Sav- A propitious visit catching his boat to the US. I remem- itch continued the work and in a later This is where I came in because Lise ber how he struck his head after I had paper in the summer of 1938 acknowl- Meitner was lonely in Sweden and, as barely started to speak and said: edged that the 3.5-hour substance was her faithful nephew, I went to visit her "Oh, what fools we have been! We not thorium but behaved a bit more at Christmas. There, in a small hotel ought to have seen that before." But like actinium and even more like lan- in Kungalv near Goteborg I found he had not—nobody had. thanum. She had come very close in- her at breakfast brooding over a letter Lise Meitner and I composed a deed to the concept of nuclear fission from Hahn. I was skeptical about the paper over the long-distance tele- but unfortunately did not state it contents—that barium was formed phone between Copenhagen and clearly. She said that it was definitely from uranium by neutrons—but she Stockholm. I told the whole story to not actinium and that it was quite sim- kept on with it. We walked up and George Placzek, who was in Copenha- ilar to , "from which it could down in the snow, I on skis and she gen, before it even occurred to me to be separated only by fractionation." on foot (she said and proved that she do an experiment. At first Placzek But she did think it could be sepa- could get along just as fast that way), did not believe the story that these rated. The reason was probably that and gradually the idea took shape that heavy nuclei, already known to suffer she still had a mixture of two substan- this was no chipping or cracking of from alpha instability, should also be ces; in that case of course one does ef- the nucleus but rather a process to be suffering from this extra affliction. fect a partial separation. Then this explained by Bohr's idea that the nu- "It sounds a bit," he said,. "like the work was in turn checked by Hahn cleus was like a liquid drop; such a man who is run over by a motor car and Strassmann who discovered ra- drop might elongate and divide it- and whose autopsy shows that he had dioactive products that behaved partly self. Then I worked out the way the a fatal tumor and would have died like actinium, partly a bit like radium. of the nucleus would within a few days anyway." Then he

PHYSICS TODAY • NOVEMBER 1967 • 47 THE JOLIOT-CURIES discovered CENTRAL FIGURES in the discovery were Otto Hahn and Lise Meitner, artificial radioactivity. here shown in front of the institute that bears their names

said, "Why don't you use a cloud Serendipitous searches Berkeley team turned a blind eye chamber to test it?" I did not have a when their counters "misbehaved." cloud chamber handy and thought it To come back to my initial question: After the discovery there was a would be difficult anyway. But I Why did it take so long before fission sheep-like rush to repeat the experi- used an ionization chamber and it was was recognized? Indeed, why wasn't ment with only the most obvious vari- a very easy experiment to observe the the neutron found earlier? Ruther- ation (I was one of the sheep). Only large pulses caused by ion fragments. ford thought about it and foretold Fermi had the intelligence to strike I do not think chronology means some of its properties as early as his out in a different and tremendously very much and certainly cannot claim Bakerian lecture in 1920; but Joliot fruitful direction. any particular intelligence or original- did not read it, expecting a public lec- But then Fermi got on the wrong ity. I was just lucky to be with Lise ture to contain nothing new! When track: He felt sure that uranium, like Meitner when she received advance Curie and Joliot found that the "beryl- other heavy nuclei, would obediently notice of Hahn's and Strassmann's dis- lium radiation" ejected protons from swallow any slow neutron that fell on covery. Then I had to be nudged be- paraffin, they put it down to a kind of it. He did make sure that the radioac- fore I did the crucial experiment on 13 Compton effect of a very hard gamma tive substances that were formed from January. By that time our joint paper radiation (some 50 MeV), ignoring it were different from any of the was nearly written. I held it back for the objections of theoretical physicists. known elements near uranium. Ida another three days to write up the The neutron was finally observed in Noddack, a German chemist, quite other paper, and then they were both Cambridge, where such a particle was rightly pointed out that they might be sent to on 16 January but pub- expected and had been sought. lighter elements; but her comments lished a week apart. In the first paper At the time the neutron was found (published in a journal not much read I used the word "fission" suggested to in 1932 pulse amplifiers and ionization by and hardly at all by physi- me by the American biologist, William chambers were available for a facile cists) were regarded as mere pedant- A. Arnold, whom I asked what one detection of fission pulses. But that ry. She did not indicate how such calls the phenomenon of cell division. would have been too big a jump to ex- light elements could be formed; her The second paper also contained a pect. The liquid-drop model of the paper had probably no effect whatev- suggestion from Lise Meitner that fis- nucleus was born late; the compound- er on later work. sion fragments emerging from a bom- nucleus idea was conceived by Bohr In the end it was good solid chemis- barded uranium layer could be collect- only late in 1936. It would have been try that got things on the right track. ed on a surface and their activity a stroke of genius to think of fission Irene Curie and Pavel Savitch came measured. The same thought inde- then, and nobody did. very close to it; only the presence of pendently occurred to Joliot, and he The discovery of artificial radioac- two substances with maliciously simi- successfully did this experiment on 26 tivity in 1934 was again a chance dis- lar properties prevented them from es- January. About that same time the covery; no one had looked for it ex- tablishing uranium fission before news reached the US; what happened cept Rutherford, who looked in vain Hahn and Strassman finally accom- then is discussed by Wheeler. for . And indeed the plished it. D

48 • NOVEMBER 1967 • PHYSICS TODAY up and down in the colloquium and took a central part in discussions. Mechanism of Fission Liquid drops The story of the development of the by John A. Wheeler liquid-drop model and the compound- nucleus picture is a familiar one. IN EARLY JANUARY 1939 the Swedish- working with Gregory Breit, to whose What is not so clear and was certainly American liner, MS Drottningholm insights I owe so much. He and the not evident at the time is the distinc- carried a short message across the group of which I soon found myself a tion between these ideas: (1) The stormy sea from Copenhagen to New member accepted almost unconsciously compound-nucleus model shows, in York. This message symbolized the the model of the nucleus of that day: essence, that the fate of a nucleus is steady transfer of nuclear discoveries neutrons and protons moving in a com- independent of the mechanism by from Europe to the US that had been mon self-consistent potential, closely which it has been formed, and (2) the going on during the Hitler years. analogous to the electric potential of liquid-drop model is, so to speak, a Although these transfers were fate- the atom. "Unconscious" our accept- special case of the compound-nucleus ful for the US and the rest of the ance of the model was, yes; but also model, a particular way of making world, the act of relaying this particu- shadowy. None of us took it too liter- such a model of rea- lar message was simple: a few words ally, especially not Breit, With his cau- sonable. Bohr proposed that the mean spoken by Otto Frisch to Niels Bohr tion and insight. Thus he was always free path of nucleon is short in rela- on the pier in Copenhagen and a few willing to consider alpha particles in tion to nuclear dimensions instead of words spoken to Enrico Fermi and the nucleus as well as neutrons and being long, as assumed in all previous me by Bohr on the pier in New York. protons when that point of view made estimates. This new idea made some- As a junior participator in the events sense in considering a particular reac- thing like a liquid-drop model exceed- that occurred then and in subsequent tion. Breit also directed especial at- ingly attractive. months, I shall relate the activities that tention to areas of investigation as No one looking back on the situation led to the publication of a Physical Re- nearly free as possible of model-de- from today's vantage point can fail to view paper by Bohr and me. In this pendent issues. Thus much work was be amazed at "the great accident of paper we summarized the thoughts done on the penetration of charged nuclear physics"—the circumstance that expressed in the message: the liquid- particles into nuceli and how the cross the mean free path of particles in the drop model that Frisch had applied to section for a depends nucleus is neither extremely short com- the mechanism of fission and the de- on energy. The analysis of scattering pared with nuclear dimensions (as as- terminations of packing fraction that processes in terms of phase shifts also sumed in the liquid-drop picture) nor Lise Meitner considered when arriving received much attention. extremely long (as assumed in the at the first estimate of energy release in With Breit's warm endorsement I earlier model) but of an intermediate fission. spent the following year at Niels Bohr's value. Moreover, all the marvelous No one looking at such a novel institute in Copenhagen. Here I was detail of nuclear physics turns out to process at that time could fail to call initiated into the study of many new depend in such a critical way on the on everything he knew about nuclear ideas, but nothing was more impressive value of this parameter. As physics to seek an interpretation. For- in nuclear physics than the message and Ben Mottelson have taught us in tunately the key ideas for unraveling that M0ller brought back during the recent years, no one could have pre- the puzzle had already been de- spring of 1935 from a short Easter visit dicted the precise one among many veloped. It may be appropriate to to Rome: It told of Fermi's slow-neu- alternative regimes in which the recall what had been learned about tron experiments and the astonishing phenomenology would actually lie nuclear physics in the preceding half resonances that he had discovered. from any advance estimate of the a dozen years. Every estimate ever made before then mean free path. Only observation indicated that a particle passing could suffice! Knowing as little as Clues to the answer through a nucleus would have an ex- one did in 1935 about the value of 1933 was a fruitful year for someone tremely small probability of losing its this decise parameter, still less about like me, who was just earning his doc- energy by radiation and undergoing its cirticality, one had no option but tor's degree. It was the year of the capture if the current nuclear model to explore with all vigor the idea that and Werner was credible. Yet, directly in opposi- the mean free path is very short. Heisenberg's great paper on the struc- tion to the predictions of this model, The development of the liquid-drop ture of nuclei built out of neutrons and Fermi's experiments displayed huge model, which was applied to a variety protons. These discoveries made one cross sections and resonances that were of processes, took place in the hands feel that he might soon know as much quite beyond explanation. of Fritz Kalckar and Niels Bohr in about the nucleus as he already knew Of course a number of weeks went 1935-37. They applied it to a variety about the atom. by before the most significant results of processes. At the center of every such application stood the idealization Encouraged by the vision that in- of this discovery could be sorted out. of the compound nucleus, that is, the spired so many young men, me in- Everyone was actively concerned, but concept that a nuclear reaction occurs cluded, at that time, I spent 1933-34 no one more so than Bohr, who paced

PHYSICS TODAY • NOVEMBER 1967 • 49 mechanism of fission and, not least, analyze the barrier against fission and the considerations that determine its height. First of all, of course, we had to formulate the very idea of a threshold or barrier. How can there even be any barrier according to the liquid- drop picture? Is not an ideal fluid infinitely subdivisible? And therefore cannot the activation energy required to go from the original configuration to a pair of fragments be made as small as one pleases? We obtained guidance on this question out of the theory of the calculus of variations in the large, maxima and minima, and critical points. This subject we absorbed by osmosis from our environment, so thoroughly charged over the years by the ideas and results of Marston Morse. It became clear that we could find a configuration space to describe the de- formation of the nucleus. In this de- formation space we could find a variety of paths leading from the nor- mal, nearly spherical configuration over a barrier to a separated configuration. On each path the energy of deforma- tion reaches a highest value. This peak value differs from one path to another. Among all these maxima the minimum measures the height of the saddle point or fission threshold or activation energy for fission. While we were estimating barrier heights and the energy release in vari- STROLLING THINKERS, Fermi (left) and Bohr, are well known for ous modes of fission, the time came for their important applications and expansions of early ideas of nuclear fission. the fifth annual theoretical physics con- ference held in Washington on 26 Jan. Bohr felt a responsibility toward Frisch and Meitner and thought that word of in two well separated stages: First, marizing this general picture of fission their work-in-progress and their con- the particle arrives in the nucleus and in terms of formation and breakup of cepts should not be released until they imparts an excitation; then in some the compound nucleus. had the proper opportunity to publish, way the nucleus uses that energy for Rosenfeld had originally accom- as is the custom throughout science. radiation, neutron or alpha-particle panied Bohr to Princeton for several Even though this was the situation, at emission or any other competing pro- months of work on the problem of the outset Rosenfeld did not appreciate cess. measurement in quantum electrody- all the complications and demands of namics. During Rosenfeld's Princeton Bohr brings the news Bohr's position. On the day of Bohr's sojourn Bohr gave less than half a arrival in the US Rosenfeld went down The message that Frisch gave Bohr dozen lectures on that issue. Never- to Princeton on the train. (Bohr had as Bohr left Copenhagen opened up a theless, that and many other questions an appointment later that day in New new domain of application for this conspired to take much of his time. York.) Rosenfeld reported the new concept of the compound nucleus. By No one could go into his office without discovery at the journal club—the regu- the time Bohr had arrived in New York seeing the long list of duties and people lar Monday night journal club—and of he had already recognized that fission he had to give time to. That list made course everybody was very excited. is one more process in competition with it easy to appreciate the pleasure with Isidor I. Rabi, who was at the journal neutron reemission and gamma-ray which he came into my office to discuss club, carried the news back to Colum- emission. Four days after his arrival the work that we had under way. We bia, where John Dunning started to he and Rosenfeld finished a paper sum- were trying to understand in detail the plan an experiment.

50 • NOVEMBER 1967 • PHYSICS TODAY Nevertheless, even on 26 Jan., Bohr tions. Also we derived some useful was reluctant to speak about Frisch's information from cosmic-ray physics. and Meitner's findings until he re- Who does not recall the many detailed ceived word that they had actually calculations St0rmer and his associates been published. Fortunately that made on the orbits of cosmic-ray par- afternoon an issue of Die Naturwissen- ticles in the earth's magnetic field? schaften, which contained work by Fortunately Manuel Sandoval Vallarta Harm and , was and later workers were able to spare handed to him; thus he could tell about themselves almost all of these details. it. Of course everybody started his They had only to employ Liouville's experiments. The first direct physical theorem. It said that the density of proof that fission takes place appeared systems in phase space remains con- in the newspapers of the twenty-ninth. stant in time. Shaping the theory The same considerations of phase ROSENFELD, with Bohr, summarized space were equally useful for evaluat- The analysis of fission led to the theory the idea of fission. ing the rate of fission. It turned out of a liquid drop and this in turn led that we could express the probability back to a favorite love of Bohr, who, of going over the barrier as the ratio of for his first student research work, ex- two numbers. One of these numbers is perimented on the instability of a jet Between this case (the power series related to the amount of phase space of water against breakup into smaller about x = 0) and the other case (the available in the transition-state con- drops. He was quite familiar with the power series about x = 1) there was figuration as the nucleus goes over the work of John W. Strutt, the third Lord an enormous gap. We saw that it top of the barrier. We were forced to Rayleigh. This work furnished a start- would take a great amount of work to think of all the degrees of freedom of ing point for our analysis. However, calculate the properties of the fission the nucleus other than the particular we had to go to terms of higher order barrier at points in between. Conse- one leading to fission. All these other than Rayleigh's favorite second-order quently we limited ourselves to inter- degrees of freedom are summarized calculations to pass beyond the purely polation between these points. In the in effect in the internal excitations of parabolic part of the nuclear potential, 28 years since that time many workers the nucleus as it passes over the fission that is, the part of the potential that have done an enormous amount of barrier. In classical terms this con- increases quadratically with deforma- computation on the topography of the cept is well defined. It is a volume tion. We determined the third-order deformation energy as depicted over in phase space completely determined terms to see the turning down of the configuration space as a "base" for the by the amount of energy. potential. They enabled us to evalu- topographic plot. We are still far from The other quantity, appearing in the ate the height of the barrier, or at least completing the analysis. Beautiful denominator of the rate-of-fission ex- the height of the barrier for a nucleus work by Wladyslaw J. Swiatecki and pression, is linked with the volume of whose charge was sufficiently close to his collaborators at Berkeley has taught phase space accessible to the com- the critical limit for immediate break- us much more than we ever knew be- pound system. In all the complex up. fore about the structure of this fission motion short of actual passage over the Here we found that we could reduce barrier and has revealed many unsus- barrier the ensemble of systems under the whole problem to finding a func- pected features for values of x that consideration remains confined to the tion / of a single dimensionless variable are remote from the two simple, origi- narrow band of energies, AE, defined x. This "fissility parameter" measures nal limits. by the energy of the incident neutron. the ratio of the square of the charge to From we turned to What counts is this energy interval the nuclear mass. This parameter has fission rate. All of us have always multiplied with the rate of change of the value 1 for a nucleus that is already recognized that nuclear physics con- volume in phase space with energy unstable against fission in its spherical sists of two parts: (a) the energy of a for the undissociated nucleus. The form. For values of x close to 1, by the process and (b) the rate at which the beauty of this derivation is the fact that power-series development mentioned process will go on. The compound- these classical ideas lend themselves above one could estimate the height of nucleus model told us that the rate to direct transcription into quantum- the barrier and actually give quite a should be measured by the partial mechanical terms. Thus the Went- detailed calculation of the first two width of the nuclear state in question zel-Kramers-Brillouin approximation terms in the power series for barrier for breakup by the specified process. taught us that volume in phase space height, or /, in powers of (1 — *)• determines the number of energy Toward a simpler theory The opposite limiting case also lent it- levels. So we concluded that the self to analysis. In this limit the nu- How could we estimate this width? width—the desired width measuring cleus has such a small charge that the Happily, in earlier days, several per- the probability for fission—is given by barrier is governed almost entirely by sons in the Princeton community— a ratio in which the numerator is the surface tension. The Coulomb forces among them Henry Eyring and Eugene number of states accessible to the give almost negligible assistance in Wigner—had been occupied by the transition-state nucleus as it is going pushing the material apart. theory of the rates of chemical reac- over the barrier, that is, the number of

PHYSICS TODAY • NOVEMBER 1967 • 51 distribution from which one could de- So the resonance had to be due to 238 termine the K values of that channel. U , and the very fact that the reso- The cautious phrase that was used in nance neutrons did not bring about 238 reference to that channel number ap- fission proved that U was not sus- pears in the following quotation: "It ceptible to fission by neutrons of such should be remarked that the specific low energy. Thus if it was not sus- quantum-mechanical effects which set septible at that energy, it would cer- in at and below the critical fission tainly not be susceptible at lower energy may even show their influence energies; consequently low-energy fis- 235 to a certain extent above this energy sion must be due to U . and produce slight oscillations in the A few days later, on 16 April, beginning of the yield curve, allowing, Placzek, Wigner, Rosenfeld, Bohr, my- possibly, a direct determination of the self and others discussed whether one number of channels." Of course we could ever hope to make a nuclear know how later on in the 1950's these explosive. It was so preposterous then variations were observed by Lamphere to think of separating U235 that I can- PLACZEK was helpful in formulating and Green and others and how they not forget the words that Bohr used theories of fission. led to direct measurement of the chan- in speaking about it: "It would take nel number. the entire efforts of a country to make a bomb." He did not foresee that, in Bohr's epiphany truth, the efforts of thousands of states of excitation other than motion The most important part of this Prince- workers drawn from three countries in the direction of fission. In the de- ton period happened when I was not would be needed to achieve that goal. nominator appears the spacing be- in direct touch with Bohr. One snowy The theory of fission made it pos- tween nuclear energy levels, divided morning he was walking from the sible to predict in general terms how by 2?r. Thus we had attained the most Nassau Club to his office in Fine Hall. the for fission would de- direct tie with experimentally interest- As a consequence of a breakfast dis- pend upon energy. In Palmer Physical ing quantities. The formula that was cussion with George Placzek, who was Laboratory Rudolf Ladenberg, James obtained in this way for the reaction deeply skeptical of these fission ideas, Kanner, Heinz H. Barschall and Van rate, or the level width, applied to a Bohr began struggling with the prob- Voorhies, just at the time we were wide class of reactions as well as to lem of explaining the remarkable de- working on the theory, actually mea- fission, and was more general than pendence of fission cross section on sured the cross section of uranium in any that had previously been available neutron energy. In the course of the the region from two million to three in reaction-rate theory. The new walk he concluded that slow-neutron million volts—and also the cross section formula gave considerable insight into fission is caused by U235 and fast- for thorium, all of which fitted in with the rate of passage over the fission neutron fission by U238. By the time predictions. The same considerations barrier. he had arrived at Fine Hall and he and of course made it possible to predict At this particular point it is inter- I had gathered together with Placzek that 239 would be fissile. esting to note the caution with which and Rosenfeld, he was ready to sketch For this application of the theory we Bohr adopted the formula. He would out the whole idea on the blackboard. are especially indebted to Louis A. come in every other day or so, and we There he displayed the concept that Turner. One started on the way that would go at it for perhaps a half a day, U238 is not susceptible to division by ultimately led to the giant plutonium trying out first this approach and then neutrons of thermal energy, nor is it project having only this theoretical that approach. But his supreme cau- susceptible to neutrons of intermediate estimate to light and encourage the first tion was most evident when we wanted energy but only to neutrons with ener- steps. to interpret the number of levels ac- gies of a million electron volts or more. offered a most cessible in the transition state. Today Further, the fission observed at lower attractive application of these ideas in 235 that number is called "the number of energies occurs because U is pres- conjunction with the concept of barrier channels," and we use it as a formula ent and has a 1/v cross section for penetration. Another application dealt to describe the channel-analysis theory capture. We already knew experi- with the difference between prompt of fission rate. Also we apply similar mentally that neutrons of intermediate neutrons and delayed neutrons. In channel-analysis considerations to energy undergo resonance capture. conclusion, nuclear fission brought us a other nuclear reactions. But at that And, with the help of simple con- process distinguished from all the time the idea that each one of these siderations, we could show that the other processes with which we ever individual channels has in principle a resonance reaction of neutrons with dealt before in nuclear physics, in that definite experimentally observable sig- uranium could not be due to U23r\ We nificance was, for us, of dubious cer- we have for the first time in fission a concluded this because we knew that nuclear transformation inescapably tainty. Still less did we appreciate, the resonance cross section would ex- until the later work of Aage Bohr, the collective in character. In this sense ceed the theoretical limit given by the fission opened the door to the develop- possibility that each individual chan- square of the if U235 were nel would have its individual angular ment of the collective model of the responsible for the resonance effect. nucleus in the postwar years. 0

52 • NOVEMBER 1967 • PHYSICS TODAY