Oral History Transcript — Dr. Morrell Cohen
This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.
This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.
Please bear in mind that: 1) This material is a
transcript of the spoken word rather than a literary product; 2) An interview
must be read with the awareness that different people's memories about an event
will often differ, and that memories can change with time for many reasons
including subsequent experiences, interactions with others, and one's feelings
about an event. Disclaimer: This transcript was scanned from
a typescript, introducing occasional spelling errors. The original typescript
Access form | Project support | How to cite | Print this page
See the catalog record for this interview and search for other interviews in our collection
Morrel Cohen; June 5, 1981
ABSTRACT: At Berkeley as a graduate student after Kittel’s arrival 1950, Kittel’s development of the department (after the loyalty oath); focus on solid state physics, mainly resonance physics (ferromagnetic resonance, cyclotron resonance); University of Chicago and Berkeley relationship. Cohen at Chicago, Institute for the Study of Metals, from 1952. The Institutes for Basic Research f Institute for Nuclear Studies (Fermi), Institute for the Study of Metals (Cyril Smith, Andy Lanson, Stuart Rice), Low Temperature Laboratory (Earl Long)]; discussions of contributions in resonance physics, semiconductor physics (Kittel, Cohen, Overhauser, Jeffries), superconducting alloys (Berndt Matthias and John Hulm); semi-metals, crystal structure, band structure; Fermi surface and Fermi theory of liquids (Zener-Cohen anecdote); University of Chicago model of an interdisciplinary research institute for materials science; Onsager’s theory (1951) and its stimulating effect; Cohen’s encounter with Pippard; Cohen GE consultant (W. Harrison, free electron model interpretation); J. Phillip’s critical point spectroscopy; Pippard and Blount “missed” the Fermi theory of liquids. Philosophical summarizing on declaring fields “closed” importance of young people in positions of responsibility.
Session 1 from March 31, 1981 is untranscribed.
This is the second session of our discussion about your life and various episodes relating to the history of solid-state physics. Shall we just proceed through the list of topics we just prepared?
Cohen:Okay. We were going to talk first about developments in solid-state physics initiated by Charles Kittel after he arrived in Berkeley in the fall of 1950. He made his presence felt very rapidly. Because of the loyalty oath issue there essentially was very little opportunity to do theoretical research at Berkeley. People had either declared their intention to leave or had in fact left. As I had mentioned earlier, in my case it was because my advisor suggested that I work with Kittel. In other cases, the information just propagated around the department, and a substantial group of students sort of condensed around him very quickly. He initiated a seminar on solid-state physics and so, in effect, built up a theoretical group, even though it was on a temporary basis. That theoretical group carried itself through .by continuing the seminar and by interacting as a group in his absence during the second semester of the academic year of ‘50-51. When he returned in the summer the group was essentially all constituted and began to function immediately as a group. And we all — a significant fraction of the total — finished up a year later.
Hoddeson:Were the discussions in your group entirely theoretical? Or did they also concern experiments being done at Berkeley or elsewhere?
Cohen:If this were contemporary high-energy theory, Kittel would be called a phenomenologist. He was intimately and deeply concerned with experiment and the interpretation of experiment rather than the construction of formal theories. Re was very closely oriented towards experimental work, so there was constant discussion of the results of experiment and what they meant. It influenced me very heavily. For example, I did a few trial problems and then when it came time to set a thesis problem, he suggested that I work on the theory of the effect of the nuclear quadrupole interaction in the magnetic resonance — the nuclear magnetic resonance spectrum in ferroelectrics as a means of probing the internal fields in ferroelectrics and, hopefully, of getting some microscopic information about the origin of ferroelectricity. This sounded to me like a very interesting sensible thing to do. I had already worked on ferroelectricity; I had written a couple of papers which were my first publications. One was a study of local fields, and the second pointed out the importance of local fields for ferroelectricity and anti-ferroelectricity. So, it made sense, and I learned about nuclear resonance. And he said, “Work with Walter Knight. I suggested that he do the experiment, and he was very interested, so he’ll do the experiment, and you will interpret the data.” Well, I didn’t realize, I was far too young and inexperienced to realize that that was a very bad situation to be in — as a theorist sitting around and waiting for data that might or might not be available on one’s own time scale. And you know, the work got finished about five years after I left. It might have been a little bit earlier than that, maybe not quite so extreme as I’m putting it, but it was literally years after I left. Walter Knight had been hired independently. I guess there was an awareness of the need to build up solid-state physics within the department. Kittel was one example of that awareness and Walter Knight was another independent example. Kittel and Knight immediately made contact on this particular problem and I started spending a lot of time in Walter’s laboratory. I would come in and watch. When the data finally came, I actually assisted a little bit in taking the data; and when the data finally came out, I interpreted it. It was very difficult to be absolutely sure and I was never satisfied and he was not satisfied and things went on and on and on. And finally it was beginning to get towards the end of the year. At this point, I had a job. My career, at least for the next few years, was lined up. I had everything except my Ph.D. thesis. So the last two months before I was due to leave, I decided I had better face reality and figure out something to work on. I decided to do something that I discovered subsequently was mathematically impossible. I struggled manfully and it got to be about a month before I was due to leave. And then what I did was simply collect together a series of thoughts that I had had over the preceding year and put them together in the form of a thesis and had that typed up. I went on campus the day I was supposed to leave, and handed all the copies of my thesis to Elihu Abrahams and said, “Elihu, I have to go. Please take these around to my committee; I won’t be able to meet with them.”
Hoddeson:And what happened?
Cohen:Well, the members of my committee knew me, and I don’t think they had any doubts about whether I qualified for the Ph.D. Kittel promised to read my Ph.D. thesis on the train to the meeting we were both going to attend, and he never did, he got a good idea and thought about that instead. In fact, he never read the thesis, I discovered subsequently. So I got a Ph.D. without my thesis advisor actually having read my Ph.D. thesis. That was the seeds of a sort of tense relationship between us that developed. So that’s the story of my Ph.D. thesis, and I was ashamed of it, you know, as a sort of a last-minute, patched-together kind of thing. It took me several years to realize that some of the things in it were new and good. Other people called this to my attention and some of it got published in bits and pieces over the years, and it became the seeds of other things that I did. In any event, getting back to the original question about the relationship between Kittel’s theoretical interests and the development of experimental solid-state physics at Berkeley, Kittel had known Kip when he was at MIT just after the war. That was the time of the Kip-Arnold experiment, which I guess was one of the earliest experiments in spin resonance. There was another experiment by Halliday and Wheatley at Pittsburgh, which was the earliest spin resonance experiment. That’s the Halliday of the textbook and the Wheatley of low-temperature physics. There was also some Soviet work, also all about the same time. And those three pieces of work started the explosion of work in electron spin resonance. And it drew Charlie’s attention to resonance physics, which was something he became very interested in. He had already, during the war actually, written a very nice little paper explaining the thermal conductivity of glasses as due to the short phonon mean-free path associated with the incoherence of the structure. And I guess, he’d done a couple of other things, but when he was at MIT he explained ferromagnetic resonance, and that got him the Buckley award, got him his position at Bell, and got him his associate professorship at Berkeley. There were many other things he had done besides and that built it up, but basically he came to people’s attention because of that work. While he was at MIT, he got to know Kip and sort of looked over his laboratory. The first thing he did when he got to Berkeley was to bring Kip. Kip did not make a good impression on the students when he came.
Cohen:We just didn’t think he was very bright. And it became apparent over the years that we were very, very wrong. The first indication that I had that I was wrong was when Walter Knight said to me that Art Kip had micrometers in his fingertips. (Laughter) And the fact that he had chosen to work for Leonard Loeb for his Ph.D. thesis, for example, was not very impressive to me. And but you know, it was funny, the older I got, the smarter Kip got. It’s just like the proverbial relationship between father and son. It’s amazing how fast the father learns once the son passes twenty. In any event, I went to Chicago, and the work that Kittel had started, with a strong emphasis on resonance physics — for example, he had Al Portis working on the experimental problem before Kip came; Portis basically worked for Kittel — the work that was started then and the work that was started afterwards just led to this unbelievable blossoming, particularly the work on semiconductors and magnetic resonance. When Charlie came to the Bell Labs, he was impressed by semiconductor physics that was going on there, very impressed by it. In fact, one of the first things that he did when he got out to Berkeley was he made sure that everybody got their copies of the collection of Bell reprints and preprints. It was basically the key papers on semiconductor physics. And what Charlie did was to take the MIT exposure to resonance, which came naturally directly out of the Radiation Laboratory at MIT — the microwave techniques were all developed there — and then he put that together with the experience that he had at Bell, the observation of semiconductor physics at Bell. He hadn’t done any semiconductor physics himself while he was at Bell, but he put those two things together, and it led to the first observation of cyclotron resonance.
Hoddeson:Where was that?
Cohen:That was at Berkeley after I left. In fact, I was at lunch here in Chicago at the Quadrangle Club. Fermi came and sat down, or I sat down after he did, I don’t remember. In any event, Fermi was across this big round table where the physics department sat, and he looked up at me and said, “You must be very happy.” I was startled, and he said, “Well, you know, the article by Kittel et al. on cyclotron resonance in semiconductors. They measured the effective mass of electrons and holes in germanium for the first time.” And I thought to myself, “That’s the difference between a great man and an ordinary man.” I wasn’t happy, I was jealous. And you know regretful — this happened and here I was alone at Chicago struggling. In any event, I had known about it earlier, before it was published in the PHYSICAL REVIEW. Fermi got his PHYSICAL REVIEW that morning and read it instantly and found that article in it and recognized its significance. It was a bit surprising to me, because I had already gotten used to the idea that this was going on. So they really did marvelous things and Kittel’s important contributions were in those first few years. After that, several things went on, I think. The momentum was lost. They exploited the initial momentum and then new ideas and new directions were required and somehow this didn’t happen. Kip indeed went from semiconductors to metals and did a whole series of experiments and trained a whole series of important students. But Charlie didn’t seem quite as interested in the work in metals, although he got Overhauser to work on spin resonance in metals.
Hoddeson:But that was also in the early years, wasn’t it?
Cohen:That was in the early years. He got Dyson to make a very important contribution, but somehow there wasn’t that same kind of symbiotic relationship with Kip later on. Maybe Kip didn’t want it, I don’t know. In any event, I wasn’t there; I didn’t observe it directly, only at a distance.
Hoddeson:You’re suggesting then that there were a few great years in solid state physics by the group at Berkeley surrounding Kittel from about 1950 to — what, 1955?
Cohen:Maybe it lasted a little bit longer than that, but I would say he had around five to seven fantastic years, and then the group that he built up which then had its own life, continued that momentum, but he played a less important role. He was the sort of father figure naturally, but he was a complex man, a very complex man. Difficult to be close to, very anxious, always concerned with virtual catastrophes sometime in the future. Made no real distinction between the possible and the probable, and that absorbed a lot of his energies. He turned to teaching also at which, despite his speech handicap, he was extremely good. He was involved in science education for a while, and it was just a sort of gradual decrease in creative activity, and later on when the semiconductor electron-hole droplets came along, Jeffries work on that, he got himself very involved. But he got into a controversy about the nature of the state of the electron and whole plasma, and he had trouble getting a paper that he sent to PHYSICAL REVIEW LETTERS published. I guess the Bell Labs, with their usual vigorous opposition, were involved with it. I’ve forgotten the details of the story. And then I think he withdrew from the American Physical Society. That was his last attempt to work on what I could regard as a hot subject. Still, Marvin Cohen interacted with him, and found it rather important. Cohen is now the central figure there; he was the student of my student here, Jim Phillips. And I also guided Marvin’s work for a year before he left. So, the relationship between Chicago and Berkeley has been a close one over the years. I sent two students there: Jim Garland wrote a spectacular Ph.D. thesis on the isotope effect in superconductivity; he explained it. But he failed when he got out there. He’s another one of those complicated individuals.
Cohen:He failed to function effectively. I made the serious error, because he was so good, of letting him go out before he had submitted his Ph.D. thesis for publication. So he couldn’t finally satisfy the formal requirements for a Ph.D. here without that. So he didn’t actually qualify for the faculty position that he held, and went on and on and on, finally he finished the Ph.D. thesis. What happened was that he lost one of his boxes of material on the way, and that precipitated an emotional state in which he decided he had to redo the whole damn thing. So, they let him go after about three years or so, and he’s at Chicago Circle now. But he was very close friends with Marvin, and he was the one who convinced me to take on Marvin, despite Marvin’s abysmal record here.
Yes, so I placed Marvin with Jim Phillips because I had enough students, and that was really the start of the whole thing. I .think Marvin worked with Jim on some semiconductor problems, and then Jim was away for the next year, and Marvin came to my office when I got back from being away in the summer. I helped him during the next year with this very important Ph.D. thesis. I got him a job at Bell, and then Garland at Berkeley said to Charlie Kittel, you have to bring Marvin here. So Charlie invited him to lecture and he gave apparently a beautiful lecture. And that evening Charlie called me, almost distraught, and said, “He gave a beautiful talk.” He said, “It was very impressive. It might have been the best talk we’ve had this year at Berkeley. So I went this evening back to the laboratory and to the department office and looked up his record.” He said, “It was abysmal. And the worst thing about it was that it was right next to yours in the file. You had a perfect record by contrast.” He said, “What’s going on? Is this guy glib or something?” The record was that he had just sort of barely survived as an undergraduate, and went on into graduate school, was accepted, but failed the qualifying exam at Berkeley. And everyone had believed that this was almost impossible to do. He got to Chicago, and he failed the candidacy examination here too. Well, what had happened was that he just simply hadn’t concentrated on his studies yet, he was doing other things, and he had other things on his agenda. I think, was he a math major? I don’t remember whether he was a math or physics undergraduate major. He was a sort of big man on campus, president of the fraternity council, and he had his own band in which he played trumpet, this kind of thing. Finally, he decided he wanted to go into graduate work, but he couldn’t catch up in the time required to get to the exam, and the same thing happened in Chicago. Well, he took it the second time and he passed and Garland came to me and said, “He’s very good.” Since something very similar had happened with Garland anyway and he turned out to be very good, it was agreeable, and it worked. So, Marvin is an example of a late bloomer in the academic sense. All right, so Marvin has then played a key role in continuing the buildup of solid-state physics at Berkeley, the center of the theoretical group. Another person that I placed at Berkeley was Fred Reif. It was a serious error, the failure to promote Fred here at Chicago. There was a lot of internal stuff going on here which I won’t talk about. In any event, faculty pressure did not prevail, so Fred failed to get promoted to tenure. So then I called up Walter Knight and wrote a very strong letter for Fred, in which I made some comparisons with existing faculty members at Berkeley. They called me back and. said, “Ah, we can’t use that letter.” So I had to write another letter. In any event, Reif got the job and then he had three or four fantastic years after getting to Berkeley in which he had several very good students and produced some very important things including the quantization of vortex rings, gapless superconductivity — I mean, it was incredible. And then he went off into other things. And, also Leo Falicov. There was no problem since Marvin was already there, but in 1968, Leo had had enough of Chicago for a variety of reasons, and so Berkeley seemed just great. So we arranged that for him also. So I have made my contribution to Berkeley, so to speak. Okay, now, getting back to Chicago.
Hoddeson:Your move occurred in 1952?
Cohen:That’s correct. There are a couple of interesting anecdotes about the move, or before. In 1948, we had the little bit of tension in Berlin, and because of that, I had to register for the draft. I took a pre-induction physical. Despite the fact that I had been previously classified 4-F, I was classified 1-A. So rather than argue on physical grounds, I simply argued that I was going to get my Ph.D. — I should be granted an educational deferment. And the standard form was to state when you would be completing your education, so I said, in September or October of ‘48; I said I was going to get my Ph.D. in June of 1952. And, in the spring of 1952, I was called up for a physical. I was told in a sense that I would report for induction after I got my degree. So, by this time I had the job in Chicago, I wrote to Cyril Smith, who was the director of the Institute for the Study of Metals and explained what was happening. So he wrote a letter for me to submit to the draft board for an occupational deferment. The letter stated that the entire war effort in Korea would fall apart if I were not allowed to go to the University of Chicago (laughter). That was the implication, and it did this without lying, by the proper concatenation of a series of arguments which were totally irrelevant but sounded very impressive.
Hoddeson:Did you get to see it?
Cohen:The letter was sent to me, I have it in my files someplace.
Hoddeson; It would be fun for me to see sometime.
Cohen:That letter was used as a prototype by all kinds of people who had similar problems around the country. Elihu Abrahams, I remember, used it and there were two or three other people I have vague memories of.
Hoddeson:He used it for himself?
Cohen:Yes. Lots of people had that problem at that time. It was the tail end of the Korean War, 1952. Eisenhower soon finished it, but in any event, that was an amusing anecdote. Another amusing anecdote was that I had decided that when I got my Ph.D., I wanted to go to Europe for a year. In fact, before we got married, I had told my wife that that was what we were going to do, and my wife was very much looking forward to it. So I applied for a Fulbright to go to work with Frolich, who was working on superconductivity at the time, and I didn’t hear, and I didn’t hear, and the people at Chicago were beginning to be rather importunate. I had to make a decision, so I decided that I couldn’t throw away a job at Chicago, which was at that point in time the best place in the world, in order to wait for a pleasant possibility with Frolich in Liverpool. So I wrote and withdrew my application and months later, I got the file sent by Frolich’s secretary, and it turned out that I had been awarded the Fulbright.
Cohen:And I got to see all the letters, copies of the letters of recommendation, which were very gratifying. So I’ve often wondered —
Hoddeson:Was this before or after BCS?
Cohen:This was four years before.
Hoddeson:I mean when you got the file.
Cohen:I applied in the fall of ‘51, and this material came back in the spring of ‘52. I would have started there in the fall of ‘52. I often wondered what would have happened if I had gone to work on superconductivity.
Hoddeson:So do I!
Cohen:Well, I didn’t. Instead, I came to Chicago. Now my view was that Chicago was the center, and Berkeley was the provinces. When I was in Berkeley, I remember that the graduate students would get their Ph.D. at Chicago and then either were on the faculty at Berkeley or up on the hill. These students included Jack Steinberger, Owen Chamberlain, Geoffrey Chew, Goldberger for a year as a post-doc, Lee as a post-doe. I mean it was very, very impressive. And these were just the students. When the faculty came through — I remember Teller giving his theory of the origin of the elements — it was incredibly impressive. It was all wrong, but it was incredibly impressive. And Chandrasekhar gave a talk — to the electrical engineers yet — on randomness in the universe. An associate of Libby came and talked about carbon-14 dating. These things were mind-blowing. The stuff that was going on at Berkeley was nice, but it just wasn’t in the same class. We had a journal club at Berkeley, which was held Monday evenings every other week, and Segre would get up at the beginning of this meeting regularly, not every time, but quite regularly, and pull out a little handwritten note from Fermi, and read what was going on in physics. And by the time I was finishing as a graduate student, I had no hesitation in comparing myself with the faculty at Berkeley. I knew I would have qualified for a faculty position at Berkeley, I didn’t have any hesitation. For example, one of Chamberlain’s students would come to me to get answers to the questions that Chamberlain couldn’t answer. And I wasn’t working in that field. And Segre discovered Kittel and would come to Kittel’s office to discuss things with him. I remember once he came when I was in the office, and asked Kittel a question which Kittel hesitated on and I answered. This disconcerted Segre greatly. These are people who went on to win the Nobel Prize subsequently as experimentalists. And I had other experiences like this with various members of the faculty. There were people like Panofsky, who were clearly very bright, extremely bright, and there were people like Alvarez who weren’t so bright, but who were incredibly imaginative. I always thought Alvarez was one of the individuals who showed me how complicated intellectual and creative intellectual work actually was. Because I didn’t think he understood very much. I was a student, and I couldn’t really judge. Listening to him lecture in a course, I learned two things: one, that he had enormous ego needs, and that he wasn’t really very good on basic physics, at least as it came across in the course. And then, of course, over the years there was just this incredible outpouring of creative work, I mean his imagination was really unbelievable. But smart? No. Panofsky was ‘sort of the reverse. He was exceedingly smart, but he didn’t demonstrate Alvarez’ kind of imaginative vitality. So there were a few people in the faculty that I won’t say awed me, but whom I really regarded as people I would have to stretch to reach. But when I got to Chicago, everybody was like that. It was an overwhelming experience. The next most senior faculty members were Murray Gell-Mann and Valentine Telegdi. They were instructors. Then there were two assistant professors: Goldberger, Murph Goldberger and a fellow named Ed Adams, who was another of these exceedingly bright and facile individuals without very much imagination, we learned subsequently, without any real depth of understanding.
Hoddeson:Was Yang there too?
Yang was long since gone. And, of course, the senior faculty was just unbelievable. Nobel prizes were dropping with regularity. Maria Mayer was working on the shell model, at that time, sort of winding up the work. They were sort of finishing up the work on the Chicago cyclotron; they were in the final stages of that. That was all the landmark work. And, it was just an incredibly impressive place. Murray Gell-Mann was concerning himself with strangeness. Goldberger and Gell-Mann were developing dispersion theory. It was just a whole level of discourse, the way one was expected to behave in intellectual interactions was one I had not only not experienced, but hadn’t even imagined. And Fermi was very impressive to watch. I went to the general theoretical seminar. The questions would come up, and a bright young man like Cell-Mann would be just buzzing around all over this problem and Fermi would be sitting there, everybody would be talking and Fermi would be sitting there, not saying anything.) And then finally, the momentum would slow down, Fermi would be sitting there steadily nodding his head, and finally it would come to a dead stop. And Fermi would speak and he would have the answer. Or an insight, at any rate. So the whole thing was both terrifying and salutary. When I came, I went from Worcester Tech to Dartmouth to Berkeley. And in the first few weeks, I had to effect an enormous change in my own intellectual level, in order to be able to perform at the level of the other graduate students. I had passed essentially all of them except the best few within a couple months after that, and they didn’t help me, rather they hindered me from that point on. When I came to Chicago, much the same thing happened. I had to effect a complete renormalization, in the level of thinking. When I came I had an offer only in the Institute for the Study of Metals. The first day I was there I came with my wife to the Institute to sort of check in. I was sent upstairs to say hello to Andy Lawson who was then head of the physics section of the Institute and the head of the physics department. And I went looking for him, and I saw this kid in the hall, a 17-year old in chinos and a t-shirt, and I said, can you take me to Professor Lawson’s office, please? And he said, “Yes, I can.” And he took me to the office, and he went inside and sat down at the desk. (Laughter) And he said, “Do you want to join the physics department?” And I said, “of course.” So that was arranged. So that’s how I got to Chicago, and those were some of the early impressions. But I want to talk somewhat more systematically about what the institute was, what its significance was, and so on. It was started immediately after the war. It was one of three institutes, the Institute for Nuclear Studies, the Institute for Radiobiology and Biophysics, and the Institute for the Study of Metals. These formed the Institutes for Basic Research. They were developed; I think, as a means to re-create in whatever measure was possible, the extraordinary circumstances that existed during the presence of the Metallurgy Project at the University of Chicago. People had come here, had lived here, and found it was good here, and good to work at the University, and so they tried to bring that back. They identified certain figures around which the institutes were built. Obviously, the Institute for Nuclear Studies was built around Fermi. And I guess they were very impressed with Cyril Smith, who was a metallurgist, a most unusual metallurgist, and there was no place for him at the University of Chicago. There was no metallurgy department, there was no engineering school, so they created an Institute for the Study of Metals; he was the founding director for the Institute for the Study of Metals. He was also one of the first chairmen of the division of solid-state physics of the American Physical Society.
Hoddeson:I get the impression that he was one of the people mainly responsible for bringing together the metallurgists and physicists, whose thinking had been going on quite different tracks until then.
Cohen:His view was that one would bring together in this institute physics and chemistry and metallurgy. Metallurgy was the queen of the sciences and physics and chemistry were her handmaidens. (Laughter)
Hoddeson:A healthy attitude.
Cohen:Yes, a healthy kind of arrogance that leads to progress in any field. Then, in addition, we had a fourth section within the institute which was the low-temperature laboratory. And I believe — this has to be confirmed by going back to the early correspondence that would be detective work, I mean, someone would have to come here and do it. And I don’t even know where the pre-institute correspondence actually resides, somewhere in the archives of the university. I think Earl Long was another person who was important at Los Alamos and they wanted to bring him into the chemistry department; that meant a low-temperature laboratory. He was Giauque’s student, so he built a low-temperature laboratory which included Willard Stout and Lester Gutman. Then later they got Lother Meyer from Holland. That had a life of its own, somewhat separate from the rest of the Institute, it was not directly connected with metallurgy, but still very heavily influenced by the fact that it was in an Institute for the Study of Metals, because it was here that the first systematic studies of superconducting alloys were made of, very well annealed samples, that led ultimately to the discovery of Type-II superconductivity. I mean, they annealed them for years and couldn’t get rid of the heterogeneity. Literally, they didn’t properly interpret it, but it was really the key experiment to which one could go back and say Type-II was there. Heterogeneity of composition was no problem, they knew the microstructure, and there was no possibility that the phenomena observed were associated with physical heterogeneities in the material. And also, that was where Berndt Matthias and John Hulm overlapped, which led ultimately to all of Matthias’ work on high-temperature superconductivity, the search for new superconductors. He was trained by Busch in Zurich and I guess, I hope I have that right, that’s what I remember. And he worked on ferroelectricity, discovered a number of ferroelectrics; came here, and got in contact with Hulm and taught him low-temperature physics, and he taught Hulm the excitement of the search for new materials, and they started looking for new superconductors, and that led to the search for high-temperature superconductors subsequently. So this was a central experience for Matthias in a number of ways. I can come back to that. Now, so the Institute became almost by chance, because of the personality of one man whom the University wanted to attract — one had to have something like this if he were to be attracted it became a model, an interdisciplinary research institute on material science. It was clear that what was being done here was far broader than metallurgy.
Hoddeson:Was anything else like that going on elsewhere in the country? For example, was it at Berkeley or MIT?
Cohen:I’ll come back to that. Would you like me to describe what went on here for a little bit longer?
Cohen:What we had was a metallurgy section, with Charles Barrett as the senior person, and it was I think very highly successful. Cyril Smith was very suspicious of theoretical physics, but Zener was okay because he was thinking about metallurgical problems. They had a lot of theoretical post-docs who came through and worked on various aspects of statistical mechanics and low-temperature physics. For Cyril this reinforced his prejudice. However, Zener did a lot of very good things while he was here. Exciting, very imaginative, but he was a man who never stayed for very long in any one place. He stayed here; I don’t know whether it was four or five years. In 1951, he left to take on the directorship of the Westinghouse Research Laboratory. This was a mistake for Zener and a mistake for Westinghouse. Zener was not what one would call an ideal manager. And there was also a loss here. In any event, there was an opening here in solid-state theory. And that was the job that I got. Kittel had me read the quarterly progress reports of the Institute so I knew what people were doing in advance. I read the papers. I took particular note on papers in the Physical Review that had been written by people at Chicago, and when I came, I came with a sort of broad-minded approach, and I sort of went around and talked to everybody.
Hoddeson:These quarterly progress reports, were they sent out to many —
Cohen:— a very broad distribution, yes.
Hoddeson:They were. So they must be available in the library somewhere.
Cohen:They certainly are, and they’re available here. Yes, they’re available here, certainly. You may very well have them available at Urbana.
Hoddeson:Thanks. That’s certainly a good source.
Cohen:But these were mostly papers that were going to be published, and after a while, we stopped that; it was too expensive, and we just sent out the abstracts. I don’t know how broad the distribution is now. It’s probably not very broad.
Hoddeson:But that in that period —
Cohen:In that period, it was very broad. That was what people did. There was a real community. The low-temperature people were a little bit off to the side, but the people doing chemistry, Nachtrieb —
Cohen:They were concerned with metals, the chemistry of metals, and the physics of metals and of metallurgy as such. There was a real community of interest. Now, there were senior metallurgists and junior metallurgists at the assistant professor level, and there were the physicists and the chemists. And what we didn’t realize at the time was that the situation was inherently unstable. That is, we could not hire metallurgists of real promise and stature. The young people didn’t particularly want to come because of the competition with the industrial research laboratories. That was one of the major factors. We didn’t have a metallurgy department, there weren’t any students. And that was a strong deterring effect for the senior people. We tried very hard to bring some of the best people in the world, and made essentially no progress whatsoever. And without the senior people, and with the competition with the industrial laboratories and with teaching metallurgy departments, we had a very difficult time in bringing the junior people. Moreover, the junior people that we did have had a hell of a time getting promoted. In fact, only one promotion from that section took place while I was here. Moreover, there were appointments in the research institutes alone, and there was a substantial amount of feeling within the university that this wasn’t right, that people should be not be solely in research institutes, that everyone should participate equally in the teaching. You shouldn’t have physicists here and physicists there and physicists there with different classes of physicists, but there should just be physicists, they should all have appointments in the physics department. That just wasn’t with regard to the institute; it was also in regard to the Hutchins College. So finally, the people who felt this way succeeded in eliminating any such new appointments. I mean, all physicists had to have appointments in the physics department, had to qualify for appointments in the physics department otherwise they couldn’t come. Well, it turned out that at least one exception was made to that by a dean of the college. However, it was a policy of the dean of the division of the physical sciences. But, the main point was that all of the things that I have mentioned so far made it essentially impossible for the metallurgy section of the institute to flourish and grow. And after a while, Cyril Smith left and then there was a hell of a time concerning the next director, was it going to be Andy Lawson. Was it going to be me? It turned out to be Stuart Rice, finally; it took about a year to get this sorted out. We had Mark Inghram as acting director in the meantime. Well Stuart, who had come from Harvard at 25 after he was a junior fellow for a year with Kirkwood at Yale, was extremely vigorous, very interested in building. I was very interested in building. So what happened was that chemistry built up, physics built up, and metallurgy couldn’t. So after Stuart was director for a while, he decided to make a de facto recognition of the situation by changing the name of the institute to the name of an individual under which we could do whatever we wanted as time went on. If we wanted to do biology, we could do biology in the institute if, you know, as time changed and the focus of the faculty shifted. We weren’t caught up in an institute for the study of metals. I resisted this strongly and didn’t even come to the faculty meeting in which it was voted on. I knew there was no rational basis for my resistance. I resisted it internally, I didn’t say very much. I just felt it, I guess, as much for sentimental reasons as anything else.
Hoddeson:Were there different individuals proposed for the name?
Cohen:Well, they had a hell of a time finding an individual, and it wasn’t a competition among six names. James Franck was the only one. He was important in the early history of physics; he was important in chemistry; he was a Nobel Prize winner; he was a member of the faculty of the University. So his breadth and versatility as well as his distinction made his name appropriate.
Hoddeson:And he was the only serious name considered?
Cohen:As I remember it, that’s correct.
Hoddeson:It’s interesting, because usually people name an institute after a person or to honor him for some particular very relevant achievement.
Cohen:Well, anyway, that was what happened. So we ceased being an institute for the study of metals, and we are now an institute in which there is condensed matter physics going on, statistical mechanics, condensed matter chemistry, gas phase chemistry, molecular beam work, it’s basically condensed matter physics and physical chemistry with a strong slant towards condensed matter work. The old spirit lives on a great deal, you know, there are central facilities, there’s a considerable amount of shared equipment and shared funding, which goes back to the early days of the original formation of the institute. The historical importance is both for the work done and because it was a model of an interdisciplinary institute for materials research. The only other one at the time was the metallurgy department at the General Electric research laboratory.
Cohen:That was started soon after the war; management tapped Herb Holloman, who was a metallurgist at MIT, very young at the time, obviously exceedingly bright. What his research accomplishments were at the time that impressed them, I don’t know. I have no idea what his research accomplishments were. I do know that they certainly chose’ the right person. I don’t know to what extent he was influenced by the Institute for the Study of Metals, but the structure was extraordinarily similar; it was much larger. The department was almost as big as our entire division of physical sciences, and so it covered the area of material sciences much more generally. It was called metallurgy, but there was physics there, and chemistry there, and ceramics, and metallurgy and low-temperature physics. It was much broader than’ the physics department, which was mostly semiconductor physics and many, many important things happened about that same period during which both of these institutions I would say had their peak period of flourishing, roughly the same time.
Hoddeson:This is about ‘55?
Cohen:Well, Chicago started in ‘46, I don’t know when the GE department started up, but it was certainly already highly developed by 1952, when I became aware of it. We sent one of our graduate students there, Israel Jacobs, several of Andy Lawson’s graduate students, in fact, got jobs there. And about a year and a half after I got there, after I got to Chicago, they made me an offer. Somebody called me up on the phone, a fellow named Joe Becker made me an offer. John Cahn had been here; I had known him at Berkeley, and Cahn was a disaster as an experimentalist, in the view of the people here. But I told him that he was, I tried to tell people that he was very smart and a very good theorist, and that he should be kept, but they failed to promote him from instructorship to assistant professor, and let him go at the end of two years, which was, I felt, a terrible ‘mistake. And so when the people called me from GE, I said, “You must have confused me with John Cahn. I’m not looking for a job, he is. Why don’t you consider him?” So they said, “Well, we didn’t confuse you, but we will be happy to consider him on your recommendation.” So, they hired him. And he continued his deeper and deeper involvement with metallurgy. He did routine chemical kinetics out at Berkeley when he came here, and it was coming here that turned him basically into a metallurgist. We had gone, taken a year’s leave of absence at the Cavendish in Cambridge, so the Cahns, who were good friends of ours, decided that was not such a bad idea, and they too went to the Cavendish and Cambridge. There John heard lectures by Batchelder, when Batchelder was talking about hydrodynamic instabilities that was the theory of spinoidal decomposition right there on the spot. Now, ironically enough, the metallurgy department — I don’t even know if there is a metallurgy department at GE — the metallurgy department as it was has ceased to exist. There’s been a de-emphasis on basic research at General Electric laboratory; things are totally transformed. There’s no resemblance. There’s much more of a resemblance between this at Chicago and what went before than there was at GE. Those two institutions, this institute and that department, were prototypic in the country. I don’t know how the ARPA-IDL program started, you’ll have to go into the government archives.
And the ARPA-IDL program, Advanced Research Projects Administration of the Department of Defense, was set up with the idea of funding substantial programs of research which would ultimately have an important payoff in military technology. And I guess they identified as one of their key areas material science. So they decided, they called together a committee to find out what to do with it. Cyril Smith was on the committee, Jack Goldman, and they found that, I mean, they came up with a report in which they proposed a series of large interdisciplinary laboratories with material science very broadly defined. This was very good. And the committee then set up a selection committee, and we didn’t make it on the first round, which caused a hue and cry, and also I think somewhere along the line Cyril Smith resigned before the final decisions were made on the first round because he didn’t like the way things were going in the committee. Jack Goldman told me something about the inner workings of it, and told me that it was regarded as a national scandal that we had not received one of these, and so there was a lot of pressure, and we did get one on the second round. I remember Willie Zachariasen taking over the writing of the proposal, and I suspect that if he hadn’t done that, there would have been enough contention which led to enough diffusiveness in the whole process, that we wouldn’t have come up with an acceptable proposal. So, Willie made the proposal and it worked, but it was sort of ironic that the prototypic institution among universities didn’t get funded on the first round. And then, after some years, around 1970, the ARPA had decided it had fulfilled its function — that lots of people had been trained in material science at the universities — that it no longer needed to support the program, so it got passed over to the National Science Foundation, and became the present Materials Research Laboratory program of the NSF. This is a good time to break; I think we’d better go and have some lunch.
Hoddeson:We’re continuing after a nice lunch.
I want to return to the early days of the Institute for the Study of Metals at Chicago, which was quite important with regard to metallurgy, and I think, in the field of metallurgy at that time, the importance and influence of the institute was very widely recognized. On the other hand, it did have certain importance in solid-state physics as well. The low-temperature laboratory brought Jules Marcus to work for a couple of years at the Institute, and Marcus had gotten his Ph.D. at Yale, and he was the person who discovered the second metal in which the de Haas-van Alphen effect was manifested, namely zinc. And this was after an interval of a considerable period of time since the original discovery in bismuth. Marcus stayed at the Institute for a couple of years, then moved up to Northwestern, and established the low-temperature laboratory there, and the facilities for studying the de Haas-van Alphen effect. And he was plagued by health. Still he turned out a series of students and his own personal activities broadened into a fairly substantial research effort among a significant number of people in condensed matter physics at Northwestern. So I think that there’s some evidence that the Institute for the Study of Metals was a seed institution for solid-state physics, even in the very early days, as Berkeley was, and perhaps earlier than Berkeley in fact, because of its earlier start.
Now, I guess I’ll go on to an anecdote that I have about my first interaction with Clarence Zener. I had been at the Institute for perhaps less than three months, it was December of 1952, I was still pretty green at that time, and Zener was a very distinguished theoretical physicist. In fact, I even have a Charles Kittel — Clarence Zener anecdote that I will tell.
Hoddeson:(laughter) I saw Zener recently in Pittsburgh.
Cohen: I see. Zener was visiting; he had recently left the institution, but still had very close associations with Chicago. He invited me for lunch, and we went to lunch. I presume he did this because people in the Institute had talked to him about me, so we went to lunch. And Zener followed his characteristic pattern, which was to ask what on first hearing appeared to be terribly naive questions. His method was that he’d think with obsession about one or two or three things, and ask whomever he met, no matter who they were what they thought about it. And then he would engage then in conversation, and gradually his own views might emerge, but mostly he was just trying to suck people dry. And so he asked me why the electron theory of metals worked when there was such strong interaction between electrons. So without thinking twice, I told him. I explained to him why it worked. And I didn’t realize until a number, quite a number of years later, that on the spot, I had invented a very substantial part of the Fermi theory of liquids, and Zener knew what I had done even though I didn’t, and he immediately made me a job offer to come to Westinghouse. And I told him I wasn’t interested in leaving Chicago, so he invited me to come for the summer as a consultant, which was the start of my consulting career, and it was much needed extra money. I mean, my salary at that time was $4,750 on a twelve-month contract, so $400 a month. And he offered me $700 a month, which seemed to me at the time to be an immense amount of money. And also, it was a chance to visit, and indeed stay with the Keffer’s, good friends of ours. He had moved from Berkeley to the University of Pittsburgh to go to a place for a short time where there were more people doing and interested in solid- state physics than they were around Chicago. So, it was an interesting experience that summer, but nevertheless I didn’t follow up on that idea at all. What I had realized was the reason that the independent-particle picture worked was because the states in the independent-particle picture, the low-lying states were in one-to-one correspondence with the low-lying states in the interacting system, and that one could generate the correspondence and make it explicit by a particular form of perturbation theory, which I wrote down for Zener. If I had had just enough sense to go to Fermi, say, and talk to him about this, perhaps something might have happened, but I didn’t really I didn’t realize what I had done and what I had understood until I read Landau’s papers.
Hoddeson:When did this happen?
Cohen:This was in 1952. When were Landau’s papers published?
Cohen:‘56 and ‘57, yes. Of course, Landau had probably had these similar ideas, some kind of similar understanding in the thirties or forties, which finally matured and led to publication. That’s the way one normally thinks about these concepts. So that was the Zener anecdote, and it’s a very interesting illustration of the impact that the kind of mind someone like Zener has on young people. Just by asking the questions, just by pointing out that these things are important to think about, or not obvious, not trivial, one can transform the way young people think.
Cohen:There are several themes that I could take up that concerns the specific contributions of our theoretical group here in the Institute. I will do these not necessarily in chronological order. One activity which I think was important was the work on group V semi-metals. I began to think about the tremendous developments that occurred in the physics of semiconductors, germanium, silicon, and the three fives as work on materials of average valence 4. And I began to wonder where does one go next? Suppose one took this as a way of doing research, as a theme, as a style, maybe we should concentrate then on materials of average valence 5. I didn’t think about it exactly that way, but rather I knew that these materials were very interesting; I came across a few articles in the early literature, and recognized that very little had been done. Roughly at the same as this, use of the valence emerged as a unifying theme, much as one would do in chemistry. The two things came together, and I suggested to Andy Lawson that we form a joint theoretical and experimental group to exploit the opportunities that I felt were there. Now the subject never really took off in the same way that the materials of average valence 4 took off, because the semi-metals simply didn’t have the practical applications of the semiconductors. And in some ways the research was harder; in some ways it wasn’t. But in any event, what we did was to resurrect the subject that Mott and Jones, Harry Jones, had been very interested in in the thirties. Jones in particular. And we started work on bismuth, we started work on bismuth-antimony alloys, on antimony —
Hoddeson:When did this work start?
Cohen:Well, I’m trying to think. Had to have been in 1956, ’55-56.
Hoddeson:Did that cut you off? You mentioned just two substances.
Cohen:Antimony, also arsenic, and their alloys, the possibility of doping. How you made good crystals, we got Chuck Barrett interested in studying the crystal structure of these things as a function of antimony content in the alloys — the temperature dependence in the internal displacement parameter. We got band structure calculations going with Falicov, who was very interested, and his students and post-docs. Falicov, Golen, and myself wrote a paper, sort of summarizing the band structure, putting things together for a conference that was held at IBM some time ago. There was this whole series of experimental Ph.D. theses that got done by Lawson’s students while advised by Lawson and me that had to do with the skin effect, the magnetic field dependence of ultrasonic attenuation. A whole broad range of things, it was a very nice, well-balanced kind of activity, and you know, one can’t really be absolutely sure that our work was the stimulus, but I believe that it provided the stimulus for a considerable expansion of the field, or provided an important part of the stimulus. As I said, it never took off the way the semiconductor research did, because there weren’t all the people interested in the applied aspects to keep it pumped up. And while I believe it still continues, it’s not a really major area. Still, it was for me an important and interesting experience because of the cooperation with an experimentalist. That was my first experience of playing an active role in the management of a large research group including experimentalists. And I did play an active role, that is, 1 frequently set the topic of the experimental theses, and suggested a particular experimental method, even though I was not competent to handle the equipment myself.
Another thing that’s interesting has to do with the development of the study of Fermi surfaces. It has an interesting history. I suppose in a certain sense it goes back to Kapitza and the foundation of the Mond laboratory, in his interest in high magnetic fields and low-temperatures.
Hoddeson:What dates are we talking about now?
Cohen:Now we’re talking about the thirties and David Shoenberg’s youth, his early work on the de Haas-van Alphen effect, and then the post-war work with improved experimental techniques.
Cohen:No, they didn’t use microwave, not in the de Haas-van Alphen effect, but when they started really doing accurate things they used modulation techniques and detection systems that had been developed at the Radiation Laboratory during the war.
Hoddeson:The MIT Rad Lab?
Cohen:Right, the MIT Rad Lab. In fact, Dicke’s invention I guess of the cathode follower was very important — he might not have invented the cathode follower but I believe he was responsible for inventing the basis for phase sensitive detection. In any event, the electronic techniques that developed during the war transformed the way one made measurements. Now there are several things that came together. Onsager visited Cambridge. I don’t know how long he spent there, a year perhaps. In any event, he wrote a very beautiful little paper that was a seminal paper on the de Haas-van Alphen effect, in which he wrote down the general expression for the de Haas-van Alphen periods for an arbitrary Fermi surface.
Hoddeson:When did he do it?
Cohen:The date that comes to my mind is 1951; it certainly was before 1954. That in turn stimulated Lifshitz and Pogorelov to solve the general problem of magneto resistance. That’s something I’ll come back to later. And they showed how it related to the geometric properties of the Fermi surface. Now the observed weird magneto resistance behavior could be understood in terms of simple geometrical considerations. That came a little bit later, but it stimulated people to start thinking about general Fermi surfaces — it brought this unbelievably complex problem suddenly within reach and made it simple.
Hoddeson:The Sommerfeld-Bethe Handbuch article of 1933 contains pictures of ideal Fermi surfaces drawn for very simple energy functions. To draw then, they hired the man who had done the drawings for the book of functions by Jahnke and Emde. I think they’re the earliest pictures of Fermi surfaces, but of course they’re ideal, not real Fermi surfaces and the word Fermi surface isn’t used there. Do you know when the concept entered the general vocabulary of those who were working on solid state theory or whether it came out of anybody’s specific work?
Cohen:I don’t know the answer to that question, it must have been early. I had a feeling that Sommerfeld understood these things very early on.
Hoddeson:It’s not reflected in his early work on the electron theory of metals in 1927.
Cohen:You see, I am not conscious of not having known that from the time I started my training. The earliest sources were, of course, Kittel and his class, and in the seminar we read the Sommerfeld-Bethe article. I read the early literature at that time, and I don’t know whether I brought it in, you know, whether it’s something that I produced in my own mind and just read it with that in mind, and so saw that concept in it. I just am not conscious of being without it, even in graduate school when I first started studying the subject.
Hoddeson:It may be an outgrowth of the Wigner-Seitz work and its applications to lots of different materials.
Cohen:That was strictly spherical. You know, the Fermi radius, the Fermi sphere.
Hoddeson:But then it was applied to copper and other more complicated solids.
Cohen:But whenever one did any calculations in those days, one didn’t take that into account. A little bit of work on the effect of anisotropy on the Fermi surface, but not much. There were ellipsoidal approximations that one made. Landau did the de Haas-van Alphen effect for ellipsoids when he was visiting the Mond laboratory.
Hoddeson:Anyway, we’ll have to trace it back.
Cohen:That would be very interesting.
Cohen:Okay. So the Fermi surface was something that people thought about and they recognized that it could have arbitrary shape, but everybody was really afraid to treat it in realistic terms, and I think that the importance of the Onsager paper was that it gave one the strength, so to speak, to face the complexity. And Pippard started wondering, he was there at Cambridge. He had already introduced what he called the ineffectiveness concept in dealing with the anomalous skin effect for superconductors, and explaining the electromagnetic properties of superconductors. I remember Kittel saying to me while I was a graduate student, he thought Pippard was awfully smart, and commenting on, just pointing out, some papers that Pippard had written on superconductivity that he thought were very impressive. And so Pippard had started thinking in terms of his ineffectiveness concepts, and had written a paper at that time, it must have been around ’53-54, on the anomalous skin effect as a means of determining the Fermi surface. He showed how you could in fact determine the Fermi surface geometry by inverting the anomalous skin effect data as a function of orientation. Now in 1954, Cyril Smith called me into his office and said he was invited to attend a conference in England, in Bristol, that Mott was organizing, but he couldn’t go; would I like to go? Would I like to go? (Laughter) So I said yes, and I figured that if I was going to get my way paid there, I might as well make the trip worthwhile. You just can’t imagine my attitude at that time towards a trip to Europe, you know, something that happened out of the blue, it was just beyond my conception. And so, I hustled my wife and 8-month old baby to her mother in Los Angeles and went to England to attend the conference and spend the month travelling around, which included a little sightseeing, and visits to Oxford and Cambridge, which John Hulm, who was working here, who was my colleague here at the time, had arranged. During the Cambridge visit, Hulm was ill and couldn’t meet me there. And so I was there on my own, and I met .Pippard and he immediately challenged me. I didn’t understand why at the time, although I know him very well now. But it was on some particular philosophical issue in science which I happened to have thought about far more deeply than he did and more effectively, so that the conversation was very quickly terminated when he realized he could add nothing to what I had said, and which had also contradicted what he had said. The subject was simply finished, and this was just a sort of almost inadvertent response to me. I mean I had no intention to be argumentative or anything, I was just there to learn as a guest. But anyway, that was what happened, but this apparently changed Pippard’s attitude towards me instantly. And so I never had any problems with him from then on. And after we talked a little bit about science, and I don’t remember what we talked about, he suddenly asked me if the Institute for the Study of Metals would be willing to have him come and visit for a year. He told me that what he wanted to do was to measure the anomalous skin effect in copper as a function of orientation. As far as he was concerned, there was only one place in the world where this could be done at the present time, where there was a combination of the metallurgical facilities and the low-temperature facilities and that was at the University of Chicago. So I said yes, of course. I had no authority, but by that time I was already a member of the appointments committee, so I figured I wouldn’t have any problems with the Pippard appointment. I remembered what Kittel had said and had learnt more about him since coming to Chicago. So I came back and we very quickly arranged a visiting professorship for him. And by correspondence he gave us the specifications of what he needed in the way of samples, copper samples, the orientations that he wanted, the quality of the surface that he wanted. And we built his apparatus in advance; we had the doors open and everything ready. He brought some of it with him. The rest of it was built here, and when he came, he was ready to start. Well, it turned out that the equipment didn’t work quite as well as he had wanted it to, he had somewhat too old-fashioned a design, so he changed the design when he came here. That was done very quickly; he had the full year to do the, nearly the full year to do the experiment. He had a very strong impact on me, because he was an excellent geometer, and he was thinking in quite general geometric terms about Fermi surfaces, so I started thinking more generally about Fermi surfaces also because of his presence. He got the experiments done here, then he took the data with him back to Cambridge, and he analyzed it in detail. The next year, he published his paper, it came out in the Proceedings of the Royal Society with this gorgeous view of the Fermi surface, in which there’s a distortion caused by the draftsman or by the reproduction process. I don’t know which. (?) So instead of the Fermi surface of copper showing the proper three-fold symmetry, in fact, six-fold symmetry around the 1-1-1axis, it’s distorted: the cross-section of the neck is an ellipse instead of a circle which he found a little bit embarrassing but he also found it amusing. Now that was the first determination of the, an accurate, complete determination of the Fermi surface.
Hoddeson:Is it published that way?
Cohen:Yes, it was published that way. It wasn’t caught until afterwards. That was the first determination of the Fermi surface of a, metal; it was very hard work, but it was a landmark in the history of the physics of metals, and the anomalous skin effect was never really used again, because at that point, Shoenberg realized that he had all the tools in his hands necessary to do it, namely, the de Haas-van Alphen effect, and the theory, or maybe he had realized this before, and just hadn’t gotten around to do it. But anyway, he put several students on, and Andrew Gold came out I think with the Fermi surface of lead, and aluminum, and very quickly, a whole series of things started emerging based on the Onsager theory. Also, another thing that Pippard was interested in at the time that he was at Chicago was the magnetic field dependence of ultrasonic attenuation, and what it told you about — Bommel I guess had discovered that, maybe when he was here in ‘54, maybe Bommel’s measurements were in ‘53. Pippard did a little theory, a sort of simple geometric theory of the effect for transverse waves, and when I was at Cambridge, I got very interested in this. And before I went to Cambridge, Lawson and I started experiments on it, and I gave the problem to a Ph.D. student, Mike Harrison. When I was at Cambridge, Lawson was going to look after him, and Mike had some very interesting ideas about how to do it that gave the oscillations. And I talked to Pippard at the tea about it, and Pippard didn’t believe a word of it, and he said, it doesn’t distinguish between transverse and longitudinal waves; ‘I understand the transverse, but the longitudinal waves work.” I said, “It’s very straightforward.” He said, “I absolutely refuse to believe it.” So, you know, I felt that he wasn’t paying proper attention to what Mike had said, and that it was really obvious after all. He annoyed me, so I sat down and knew Mike could not have done what I was about to do without my assistance and that there was no point waiting until the time we got back. I could see that the field was trembling on the edge, and so I decided to simply to work it out. It was Pippard’s doubt that stimulated me to do it, and then I justified my doing it instead of waiting for Mike to do it in some strange way which I don’t recall very much now, except that I was worrying about his being scooped. So I sat down and worked out the theory in detail in two weeks, and showed the results to Brian, which sort of overwhelmed him momentarily. And then, I shifted everything back to Mike and said, “Clean it up.” And nothing very much happened. And on the way back from Cambridge, I had been working with David Turnbull at Cambridge on the nature of glass — he had left six months earlier — he asked me to stop off at Schenectady and we continued our work on glass. That’s another story. We arranged for me to consult at GE, which was very important in my development.
Now Mike, as I said, hadn’t done very much. I came back from GE and discovered that Walter Harrison was working on the problem, and he had completely the wrong idea about how to do it, so I showed him how to do it, what we had and so on, left him my notes, and he worked out a lot of details that I had not. So it wound up as the Cohen, Harrison, and Harrison paper. I was the first complete theory of ultrasonic attenuation by electrons in metals. That was also fairly important in the development of the Fermi surface subject. So we were building up a series of experimental techniques that could measure in detail what the shape of the Fermi surface was. I mean it was the de Haas-van Alphen effect. And the anomalous skin effect soon sort of got left in history, and all the oscillatory equilibrium properties or transport properties that related to quantization, or derived from quantization of the electronic orbits in the presence of a magnetic field. And then, I guess I have to get to that later, the development of Landau-Fermi liquid theory and the development of pseudo potential theory gave us the entire conceptual structure we needed to show us what was going on, and the subject took off. And I guess I was really the first, one of the first in the United States, to deal with Fermi surfaces in their full complexity. I remember being asked to give an invited paper before the American Physical Society soon after, let’s see, I guess, 1 gave an invited paper on the ultrasonic attenuation in ‘58, so I must have given an invited paper on Fermi surfaces in general in 1959. I guess that was in Detroit. And I remember laying out the entire subject systematically, the topological issues, the geometric issues, what kind of experiments told what; I also remember that there were a number of experimentalists at both meetings who came up to me and talked to me in detail about what they were doing, and recently, in fact, very recently, I discovered that these conversations were very important to them, because this was very early in the development of the subject, and they felt that they were sort of alone in the wilderness. Was it worth doing? Were they on the right track? And so on and so forth. And I encouraged a number of them, and they said that they felt that that encouragement was important, and it’s possible that they might have otherwise abandoned it or worked with less concentration. One of the people was, oh God, I’ve forgotten, he became Assistant Secretary of the Navy (?), Bob Morse at Brown on ultrasonic attenuation; he created a number, he had a number of very interesting students, Hank Bohm, who is now president of AUA out at Argonne, and Dave Gavenda is professor at Texas.. There are several others too. Now, concomitantly with this development, there was the work which in a certain sense originated in Jim Phillip’s thesis, which is the notion of critical-point spectroscopy, and the determination of electronic structure of crystals from their complex optical spectra. Taft and Phillips were the first to produce in the year that I was working on, the question was how you interpreted these broad optical spectra, and Phillips introduced the notion of critical-point spectroscopy about the same time that he introduced the Phillips pseudo potential. What actually happened was this: Phillips was a member of the first class I ever taught, it was an undergraduate course, and we had a very peculiar system of education. When it became time to do his Ph.D., he went to Fermi, and about that time, Fermi was very discouraged with progress in particle physics. It was also about the time that Kittel wrote the article I mentioned earlier. And so, when Phillips went to Fermi for advice, he said, “Go into solid-state physics. Particle physics is quiet and very difficult now, but solid-state physics seems to be booming.” There were two people doing solid-state theory here: Ed Adams and myself. And so Phillips came to me to work for me because, he said, I won hands down on the grounds of personality. He was not my first student, however. I had a series of students. The first student who came to work for me decided after being with me for a little while that solid-state physics was much too hard for him. He wanted to do field theory. Now, this is I think an interesting point, Fermi’s perception of the way the currents were flowing in the different parts of physics, and what was timely, and what wasn’t timely. He knew that Phillips was one of the better students in the department, if not the best student in the department at the time. The one who poked his head in here a few minutes ago got one of the best grades on the candidacy examination that was Bob Barron. But Bob bombed after the candidacy examination; Phillips was second, and Phillips had what would have been an extremely good grade, if it hadn’t been for comparison with Barron, who was a problem solving machine, but didn’t have the creativity or the real depth. And Phillips did very well, came to work for me. When I went away for the summer, I gave him a problem to think about, and when I came back, I realized it was a lousy problem. So in the meantime, I had recognized that the work of van Hove on singularities in phonon spectra associated with critical points was significant; he had used the Morse theory — that is, the calculus of variations in the large — to explain the presence of these critical points and their relationship to one another. And I saw four things. First, that he had taken into account only the topological requirements; he had not taken into account the symmetry requirements, and therefore, there had to be many more critical points, and that one could develop a minimal set of critical points based on the symmetry plus topological considerations. The topological equations came in the form of inequalities, and what was required was to have the things one knew had to be there on the basis of symmetry, to get a minimal set that would satisfy both the symmetry requirements and the topological requirements. So he could quickly take a big step beyond van Hove. The second thing that I realized was that van Hove had used a theory that was developed for analytic manifolds, and we always deal with degeneracy and therefore nonanalytic manifolds. The third thing that I realized was that the analytic or nonanalytic singularities in the density of states would remain. And the fourth thing that I realized was that it was not only in the density of states but it would be in the optical spectrum as well, and in the combined density of states, and that it could be the basis for a spectroscopy. Because it would replace sharp-line spectroscopy. And so I set Jim Phillips on the problem for the phonon spectrum. And Jim was a young man in a hurry. By God, three months later, he came back with the job done on analytic manifolds. Gene Blount who was already my student — Blount was older than I and very mature in many ways, very, very bright — we worked very hard to convince Jim that he had to do it for nonanalytic manifolds, that he could do it for nonanalytic manifolds, that he could get help from the mathematicians here, that it was generalizable, and he couldn’t leave it in the present state, which didn’t represent that much of an advance over what van Hove had already done. So he finally reluctantly agreed because he saw himself getting his Ph.D. in 6 months altogether. And he in three months more had done the whole thing — in ways that — the mathematicians had thought was clumsy and sloppy. I mean, he went to Spanier, and Spanier said, “Oh, I could have done it much more easily.” Spanier was a topologist, and could have done it much more easily, but it was done. And then he brought me his paper, and that, of course, was impossible for anyone in physics to understand. It was full of his own private language, full of mathematics, so we went through it on a word-by-word basis. Finally, the paper was written and it was understandable, published, he finished,. I guess, he did the whole thing from start to finish from fall to spring, so it was 9 months. That got him a job at the Bell Labs. Jim first worked with Mel Lax on problems in one-dimensional disordered materials. And then Conyers Herring suggested that he look at some very simple way of getting at band structure. And that led Jim to do his first work with the model potential. It wasn’t what Conyers suggested, but Jim found a very simple model potential in which he was able to reproduce what was known about the band structure with just a few parameters. Then he went out to Berkeley, and he stayed at Berkeley for a year, during which he advised Leonard Kleinman. And then Jim realized that if you take the OPW equations and you rearrange them, then you have something that looked like a model potential, which you call the repulsive potential, and the pseudo potential for the total on the other side of the equation. And then, he went to Cambridge and I thought about this while he was at Cambridge. I produced the cancellation theorem. That is, I was able to show that there was a basic mathematical reason for it, namely, if the core functions formed a complete set for the description of arbitrary wave functions within the region of the core, then there would be complete cancellation of the actual potential by the repulsive potential within the core. And to the extent to which the core functions were complete, the cancellation was complete, one had a very weak or much weaker pseudo potential. I showed this to Heine he made a few suggestions, so that became the paper with Heine on the cancellation of kinetic and potential energy in atoms, molecules, and solids, which Slater was heard to remark was the only paper, that made any sense in the whole pseudo potential literature. That was reported back to me. And so we understood then how it was that in metals the electrons could behave as though the electron-atom potential; was weak. And with Landau, ‘we understood how it was that electrons could behave in metals as though they were free when the interaction was strong. So we had a complete basis for what the metallurgists understood, and we could really then go on. And’ I then recognized that since the potential was weak, the pseudo potential could be screened, you could use approximate linear screening, and Phillips and I sat down with that. And we were able to show that in fact what we got by using simple linear screening theory was very close to the self-consistent potentials that he and Kleinman had already calculated. And so, all the machinery was there. And that brings us to Walter Harrison. I was a consultant for GE.
Hoddeson:How often did you go?
I spent about 20 days a, year, in the years that I didn’t go for the summer. In years that I did go for the summer, of course, I would spend whatever part of the 20 days that I used up outside of the summer plus the summer. I have to come back to that; I want to develop the pseudo potential theme now. The ‘second summer I spent at GE, Walter had developed this idea of constructing free electron Fermi surfaces, which Andrew Gold had also developed independently, but sort of left in his drawer, because he didn’t take it all that seriously. Walter was innocent enough, and had had’ a poor enough education as a graduate student, so that he took this seriously. So he studied zinc, and then he came to me and said, “you know, this works pretty well. Are there any other materials for which there’s de Haas-van Alphen data?” And I said, “Walter! You mean you haven’t read the literature?” And he said, “No,” so I gave him a long list of materials and I assured him that there were lots of them. So he then tried his free electron Fermi surfaces on all of these, and they worked pretty well. And so, at that point, while 1 was at Cambridge, Heine and I had introduced this notion that the nearly free electron’ model was reasonably not a few plane-wave model, but a few OPW (Orthogonal Plane Wave) model, and that there was a cancellation coming in the OPW scheme which explained ‘why it worked. So we were already part way towards what Phillips had discovered. All I needed was to see what Phillips had done, and everything else was clear. And I explained to him that there was a basic physical reason for it, that he wasn’t really doing a few plane-wave approximation, he was doing something a like a few OPW approximation. And then, time went by and all the pseudo potential stuff developed and we got the cancellation theorem, we understood perturbation theory, we understood the screening. Jim Phillips was visiting Schenectady at that time; I guess it was just after the 1960 Fermi surface conference. And Walter was talking about wanting to do a real calculation. He actually wanted to do a really detailed band structure calculation. He thought he would do one in zinc. And so, Phillips and I said yes, that was a good choice, but we didn’t think that he should do a real band structure calculation; we thought he should do a pseudo potential calculation. And he said, what’s that? And we explained in detail, and we spent a lot of time, I mean, we spent several hours with him explaining this, sort of trying to generate his enthusiasm. We told him how to do it, and what to expect, and so on and so forth. And that conversation just sort of receded into the back of Walter’s mind, and what he did, basically, was to reinvent all of the things that we told him and a great deal more besides, and moved into a hell of a lot of interesting work with all of this stuff. But the origin of it was definitely in that conversation with Phillips and me. And that led to his book on pseudo potentials with his enchantment with finding very simple ways of doing these complicated electron structure calculations.
At the same time, Phillips was developing ideas on the nature of covalency and the distinction between ionic and covalent bonding. And Marvin Cohen was a graduate student here, as I told you earlier. And he picked all this up and he has since told me that his students asked him what it was like to be in Chicago at that period when Phillips and Falicov were here and the whole pseudo potential formula was developed. He said, “Well, my office was downstairs, and whenever I had a problem, I would go upstairs and get the answer. It was just a marvelous feeling to know that you could always walk upstairs and get the answer.” And that was really what it was like, because all the things that had been puzzling us for years just fell apart, you could see how to do anything in that area of metal and semiconductor physics.
Hoddeson:This is now about ‘59?
Cohen:Well, it started in ‘59, and it carried through until 1968, when Phillips and Falicov left. We hired Phillips in 1960; we hired Falicov as a post-doc in 1960. Within a very short period of time, he and I had done the effect of spin-orbit coupling on Fermi surfaces and magnetic breakdown. You know Mike Priestley had found the giant orbit in magnesium, and I remember sitting in Leo’s office. You see, the way this happened was the following: when I was visiting GE, Walter Harrison asked me whether the gap on the surface on the hexagonal zone face in hexagonal close-packed materials really had to be zero. And I said, yes, it was rigorously zero, because of the time-reversal invariance. Unless you include spin-orbit coupling, which means that you have to bring the spin into the way you consider what time-reversal invariance does, and that leaves you with gaps vanishing only along lines, but small everywhere else because of the high symmetry. The gaps could get, at most, the size of the spin-orbit splitting. So there are large areas in which it was small, very small, even smaller than the spin-orbit splitting. And so, Walter said, “what happens if you have such small gaps.” And that was an interesting question, and I mused on it, and I decided that what happened was that the electrons ignored it. I mean, if you had a magnetic field on, for example, the orbits would essentially be the same as if the gaps weren’t there. That was as far as I got. And then when Leo came, I said, “Why don’t you calculate the gaps, the spin-orbit gaps? You have the whole magnesium band structure, why don’t you just put spin-orbit coupling in?” So he did, and we looked at it, and we indeed found these very small gaps everywhere, so we sent that off as a Phys. Rev later, it contained a mistake of a factor of 2. And then we got the Priestley Ph.D. Thesis, with the Priestley giant orbit in it. Everybody was puzzling about that, and I remember sitting in Leon’s office, it was this space right here, as a matter of fact, and they were wondering about it. And I said, “What’s the cross section of the free-electron sphere?” And so we calculated it out right then, and it fitted to four significant figures.
Cohen:So, it was just magnetic breakdown. We published that, and we published an incorrect version of the magnetic breakdown theory, which Blount corrected, in which Pippard picked up the mistake. When Pippard saw a preprint, he said he felt like an idiot. He felt he should have realized that, but he also recognized that we calculated the breakdown field incorrectly, and that made him feel ever so much better (laughter). So Blount then published the correct magnetic breakdown theory. And then we hired Royal Stark here, because he had done an absolutely brilliant job at Case studying magnetic breakdown phenomena in magneto resistance. So we built up a very powerful group. When Stark got here, he picked up immediately the pseudo potential technology, and he set up experiments in which he could measure in detail with great accuracy the shapes of the Fermi surfaces, and then pseudo potential calculations in which he would determine the pseudo potential which would fit the Fermi surface and get the whole band structure. And now with the notion of critical point spectroscopy, you could get at the band structure from the optical properties also.
So, now Marvin Cohen picked this up, having been a graduate student and he has done an enormous amount of work with his students and post-docs along these lines working out details of charge density, interface structure, etc., etc. But if you look at the themes, van Hove and my perception of what was in the van Hove paper, Herring’s invention of the OPW scheme, and then his suggestion years later to Jim that he use a simple model potential, and Jim’s discovery that that wasn’t so good, but that there was another model potential, which was basically a model version of the pseudo potential, which led him to start thinking about the OPW scheme, and could a model potential come out of that, while at the same time, Heine and I realized that the cancellation in the OPW scheme could be systematized, and it was the basis for the nearly free electron model, then the cancellation theorem — all of this stuff came together. And it would have been amusing if this had happened and I had also followed up that lead on the Fermi liquid theory.
Hoddeson:I’ve never been able to put all these steps together before. I’m really very grateful to you.
Cohen:Phillips gave a marvelous lecture, his Buckley prize lecture, in which he really did a beautiful job on the history and how it developed.
Hoddeson:I’ll have to get a copy of it.
Cohen:And Marvin Cohen also at his Buckley. award lecture, went quite far back, too, and brought it all the way up, so if you take these two lectures and lay them side by side, essentially you have the whole current development.
Cohen:How are we doing? We’re doing well as far as time is concerned. Now, let me talk a little bit about GE, my impact on GE and the impact of GE on me. I had a good time.
Cohen:I got literally hyperactive, the adrenalin flow was unbelievable. Because this was already started in the fall of 1958. That was before I had really succeeded in building a solid-state group here that was stable over a period of 8 years. I didn’t have that many people to talk to here, but when I went to GE, there were lots of people to talk to. I never slept mare than 4 hours a night, if I got to bed at my normal bedtime, I would toss and turn and think for at least four hours.
Hoddeson:Who were the outstanding people at GE in this period?
Cohen:The people that I interacted with, I found Walter Harrison and Henry Ehrenreich there as fresh Ph.D.’s. I worked with them. The work with Henry, I just sort of discussed whatever he happened to be thinking about at the time. Several years later, that led to the paper on the dielectric constant, in which we wrote down the time-dependent Hartree formulation, and showed that it was the same as the Brout-Sawada theory. So that in fact all of that pair theory of mesons that Wentzel had developed and Sawada’s transferring it over to the electron gas and so on, they weren’t doing anything more than ordinary Hartree self-consistent field theory. And Brout said to me once, “I know you think we’re all stupid,” but I didn’t. I was just doing what Kittel taught me to do, realize that that was what they were doing, and I remember talking to Henry about it, and saying, well we should collaborate on it. Henry was working on it, and we had many long phone calls and I remember one night talking to him for about half an hour in which most of the half hour was spent convincing him it could be done, it was easy, and spelling the whole thing out in detail. And I didn’t talk about actually working out in detail what the dielectric function was, but I talked to him about the first part, the formal part of the theoretical structure. He resisted me and resisted me and resisted me, and finally he agreed at the end of the conversation. The next visit, he had it all done, it was indeed right, it was simple, and so we wrote this nice paper. So that was the pattern of the relationship with Henry at that time. The pattern of the relationship with Walter I’ve already told you about. People knew that I enjoyed very much going out to GE, and they would say, “Why do you like to go out to GE so much?” And the remark that Henry most recently could still quote back to me 20 years or more later, was, “I like to go to GE because the graduate students are so good.” (laughter) And that was really true, in a certain sense, Henry and Walter were my post-docs, although it wasn’t a formal relationship. I thought of them as graduate students, but I’m mellower now and I’m willing to concede that they were post-docs. (laughter) I worked with Werner Kanzig and we did several things together following that work that I had done when he had discovered the VK center, this peculiar spin-resonance in the alkali halides. I wrote the paper that explained what it was, and made a mistake of a factor of 2 in the calculation of the g factor, which led to my misidentification of the center. And I was able to infer what the optical absorption was and that led me, because of the factor of two, to suppose that it was just the V center rather than the VK center. But Ted Kostner found the mistake of the factor of two, and they recognized down at Urbana that it had to be a new center, as did some Japanese who had also found also my mistake of a factor of two. But that was a sort of prototypic analysis and we used it again.in analyzing a number of other centers, spin-resonance centers, in the alkali halides that Werner was exploring at GE. We worked with Truman Woodruff in this. There again, it was a sort of pedagogical relationship to try to somehow bring Truman to become more effective, more creative, more something; the administration felt that there was a problem. Another person that I interacted with very strongly was John Fisher, and one of the ideas that John had was that it would be extremely interesting to study thin films, thin metallic films. And I said, “what for? You’re just going to discover that you have scattering off the surfaces. This is reasonably well understood, I mean, there are more exciting things to do.” So this seemed to depress him, he was silent for 15 seconds, and he said, “well, if we shouldn’t study the current along the film, we’ll study it across the film.” I had absolutely no comeback to that, and that led him to decide that there should be tunneling experiments. So, about that time, Ivar Giaever was deciding to stay in the research laboratory, and he had training as an applied mathematician, had a master’s degree, and was working with, I think, a fellow named Horvay, or Horvay, I can’t remember which, who was Hungarian, and who was very bright. John was manager of the group. Giaever asked John if he could work for him, John said, “well I’ll check,” and he called Horvay, and said, “How good is this Giaever?” Horvay said, “Well, Horvay is good, but Giaever is better.” So John took him on, and put him on this tunneling experiment the idea for which John had had earlier. And so, he had tunneling from the metal film, crossing an oxide layer to a metal electrode. Giaever made the experiment work, and Walter Harrison started thinking, about it. And then there’s a low-temperature group right down the hall, Mylian Fisk, and Rolan Schmidt, and people knew what was going on in low-temperature physics, and everybody kept saying to him, “why don’t you look at superconductors? Why don’t you see what tunneling in superconductors is like?” And this was after the BCS theory there was a lot of excitement. Well, Giaever is a stubborn guy, and he didn’t see any reason why it would be any more interesting to look at superconductors than ordinary metals, because he didn’t see that there was anything that he specifically could measure that would be interesting. And it would be a lot of trouble, cryostats, and all that.
Cohen:So he resisted. He had never had any training in physics; he was doing graduate study at Rensselaer.
Hoddeson:He never had any training in physics?
Cohen:Well, apart from what you would get with training in mathematics. He hadn’t any advanced training. He was taking a course with Hill Huntington, and —
Cohen:At Rensselaer. It was a solid-state theory course. Huntington was a professor at Rensselaer, and Giaever was working at GE and taking classes at Rensselaer. Huntington taught the BCS theory in his class. When he introduced the energy gap, Giaever translated it into volts and millivolts, and said, “my God, I can measure that!” He went home, did the experiment immediately, and discovered superconducting tunneling.
Hoddeson:That’s a good story.
Cohen:Bardeen was a consultant for GE and came through every so often, and he came and talked to Walter Harrison. And Walter explained to him all of the problems, all of the difficulties, I mean, what the basic theoretical situation was in normal metals, why it got difficult when you went into superconductors. The big problem was that you shouldn’t see the density of states because you get a factor of the velocity coming in, you know, the density of states is one over the velocity, the current goes as the velocity, so they cancel out any singularity associated with the density of states. And yet they were seeing the singularity associated with the density of states in the normal superconducting tunneling junction. And so that was this very deep and serious problem, and nobody understood superconducting tunneling. So Walter explained everything to Bardeen, and then Bardeen wrote a Phys. Rev letter on tunneling from a many-particle point of view sure there was something in that paper, but I was so annoyed with Bardeen because — didn’t give any credit to Walter’s patient exposition. Resort of grabbed the problem from Waiter, who was a struggling young theorist. I guess Bardeen, because he had been in superconductivity since the mid-30’s, took a very strong proprietary attitude towards it, and so he wrote this paper. I read it, and felt it said nothing. It did not teach me anything I didn’t know, and it didn’t answer any of my questions. And I dropped it; I was just left with a mild touch of annoyance towards John. Time went on.
Over the years, since I’ve been here, Bob Gomer, who does field emission, had been coming to me and talking to me about field emission, I could follow him always up to a certain point, and then he started talking about tunneling problems, and I would lose him. He was using a language I just simply couldn’t understand, and when I asked him to explain, the explanation was more difficult for me to follow than the original thing that was to be explained. So I got into the pattern of nodding, and listening in the hope that maybe I would have something to say; it was interesting stuff, but somehow I just couldn’t communicate effectively. I couldn’t answer any of his questions, and I could offer essentially no encouragement. After a while, he stopped coming to me, and when Jim Phillips came, he started going to Jim Phillips. So after talking to Bob a number of times, suddenly Jim started saying, “I understand Bob’s tunneling experiments and because of that I understand superconductive tunneling.” Next thing I knew, Jim had written, a Phys. Rev Letter, taken it to Leo, Leo said, “this is all garbage. This is all wrong. (laughter) This is what is right.” So they rewrote it together, brought it to me, and put it on my desk. I said, “Well, I mean, why don’t you stop ‘talking about it and do the problem. I mean this is all talk.” And Jim got very annoyed and said, “Well, if the problem could be done so easily it would have been done already.” And Leo was a little more patient with me, and said, “Well, what did you have in mind?” Or something to that effect, some kind of leading thing that only might have been a gesture or something. And I said, “Well,” and then suddenly a light bulb went on, and I invented the tunneling Hamiltonian, which is not a proper Hamiltonian. You know, it’s a left system and a right system, and a Hamiltonian which transfers electrons from right to left or left to right using an over complete set of wave functions. But one could deal with that, there are ways of dealing with it mathematically, but I didn’t bother. I said, “Let’s see what this particular system will do.” So, Jim threw up his hands and walked out of the office, and I said, “Leo, why don’t you come over to my house tonight, we’ll work it out.” So he brought some articles on superconductivity and he came over to my house after dinner and we finished the whole theory of superconducting tunneling that evening for the normal to superconductor case
Hoddeson:In one night?
Cohen:We brought it to Jim in the morning, and Jim said, “Fine! Fantastic! I’ll write the Letter.” So Jim wrote it up and I said, “You know, there’s something about this that bothers me. I get a very queasy feeling in my stomach. That term that we. set equal to zero, this term which couples a pair in the superconductor to a pair in the normal metal which we then set equal to zero because there aren’t any pairs in the normal metal; that bothers me. We ought to do the superconducting-superconducting case.” So they said no, and I persisted, and Jim said, “For God’s sake, we know what the answer is going to be. It’s just the convoluted density of states.’.’ And Leo said, “Oh, my God, 16-by-16 matrices!” And so I persisted for a little while, but they didn’t want to have anything to do with it, so I said, “Okay, you guys. I’ll go back to what I was doing.” So I went back to what I was doing, which was basically trying to recoup something on the Fermi liquid theory, that is, trying to redo many-body theory in a way different from the Russian literature. In fact, it had all been done in the Russian literature, and I was wasting my time. If I had had the least bit of gumption, you know, I would have sat down, and done that thing myself. That particular calculation that I was proposing was exactly the way Ambegaokar derived the Josephson Effect some several years afterwards. As soon as Pippard got our preprint, he handed it to Josephson. Josephson had thought through all the physics, but he didn’t know how to do the calculation. As soon as he saw our paper, he knew exactly how to do the calculation, and he did it. Thus we provided the missing technical component. We would certainly have encountered the Josephson current had we done the calculation. It isn’t clear to me that we would have understood it and it isn’t clear to me that we would have built up the beautiful piece of physics that Josephson built subsequently while Anderson was in Cambridge. We might have done one piece, somebody else another piece, somebody else another piece. But one thing about it that is very interesting is that neither Leo nor Jim remembers this the way I remember it. I remember this sinking feeling in the pit of my stomach, and I remember trying to talk them into going on with the calculation. . And they remember not having had a clue about the possible existence of the Josephson Effect. Of course, I didn’t have a clue about the possible existence of the Josephson Effect, but I wanted to keep working. I had the sense and judgment to know that there was something missing, that there was something still there; there was something not kosher about what was going on. But I didn’t have the guts, or the courage, or whatever, the character if you will, to stick with the thing and make sure that we understood every single aspect of it — the simple virtue of the old-fashioned physicist, to understand something down to the bone. And I went off., I had the attitude that this was sort of trivial anyway, you know, and that I was going back to a much deeper challenge, which of course was nonsense. That’s the story of the Josephson Effect. I don’t have any regrets.
Hoddeson:Well, that happens.
Cohen:To me it’s happened three times, to a .greater or lesser degree. There’s just enough time left for the cat in the cream story. You know the cat in the cream lecture?
Cohen:Well, Pippard gave sort of a featured after-dinner type of speech at a low-temperature conference in Canada around 1960, ‘61, something like that, in which he talked about the cat and the cream. It was later published in Physics Today, and I think you can find it. The basic theme was that there isn’t anything left in solid-state physics anymore. The cat has lapped up all the cream, and we should be going into applied areas, and that’s where the emphasis should be. This is very amusing, because at the same time in the same institution, Mott was struggling with the basic concepts that one needed to develop in order to think about disordered materials, which is one area which, has just boomed enormously. The whole theory of phase transitions, you just can go on and on and on. Whenever anyone says something like that, you realize that it is a statement about himself or herself. I realized that at the time, and I felt very sad. I tried to understand it. When Brian was a student at the Cavendish laboratory, the mythology of the Cavendish laboratory was that there were three great unsolved problems:
the nature of liquid helium, the nature of superconductivity, and magneto resistance. That was in the culture in the Mond laboratory. Another part of the culture was that these problems would not be solved within the ordinary framework of physics. That is just as in particle physics, one needed to go beyond to get new principles and new concepts, so one would have to do that in order to solve these problems of macroscopic quantization. Something new was needed. Well, superfluid theory came along, of helium, and Bogoliubov theory came along, and showed that there was a microscopic basis for it. And then slowly, one small step at a time, these very crude and simplistic microscopic theories got refined to the point where it became apparent to everyone that they were basically right, and that we really understood it. We didn’t understand it as a consequence of any dramatic, single step but rather in a slow buildup.
In superconductivity it was much more dramatic. I mean, before 1956, there was nothing, after 1956, there was the BCS theory. And one struggled against the BCS theory, but it won every time. And again, nothing new was needed. All that one needed to do was to be clever enough to recognize that electrons could interact by exchanging phonons and all we needed to understand that was a parallelism with electromagnetism or with meson theory or what have you. We should have understood this a long, long time ago. If we just did second-order perturbation theory a little more carefully, we would have understood it, and that the consequence of having an attractive interaction was that there was condensation. All of the ingredients were very simple, very beautiful, didn’t need any new concepts. And we hadn’t been clever enough to see it. And finally, magneto resistance just turned out to be a simple consequence of the geometry of the Fermi surface. Again nothing. So, all the great problems of his youth which really stimulated him perhaps drew him in to physics, which perhaps excited him as a physicist, fell by the wayside. So he tried for a while to introduce some novelty by going to very high magnetic fields, and built a high magnetic field magnet. That absorbed him for a while, but in the middle of this, really before that magnet worked, and before they had gotten very much out of it, high-field superconductivity was discovered and made the MIT magnet lab look sort of foolish. But they (MIT) were resilient, they just subsumed that. It also really didn’t produce anything dramatic. Then Pippard became Cavendish professor, and then he started changing the directions of the Cavendish lab, with a stronger emphasis on applied work. And it ain’t what it used to be. Just not what it used to be, there were some very, very good people there. The legacy of Bragg lives on in the NRC and in radio astronomy. It’s interesting, you see. Bragg’s attitude was well, what is going to happen in the future? What strengths do we have that will lead to important events in the future and where will they be? And who are the people who are inspired and what should we be pushing? Whereas Pippard decided that in the basic area, there wasn’t very much left, and felt we should move into areas which we really are not very good at. I mean, other people should be doing those things. So, I think he had a strong negative effect on the Cavendish laboratory, and I think that the loss of Phil Anderson which was really a major loss, was in part because, well you can get that from Phil, but in part because of disagreements over policy issues of that sort. In any event, it shows the importance of maintaining people in roles of major responsibility who are young in the important senses. It also shows the importance of having limited terms on administrative positions. It shows the importance of never having people who are full-time administrators as well.
Cohen:Not that Pippard was, but I’m saying these are the dangers and this is an illustration, even in a person in whom one least expected it to show.
Hoddeson:Thanks so much. You did an absolutely fantastic job.