John Bardeen - Session III

Hoddeson:

Before continuing, I would like to ask you a question I omitted last time when we were discussing your Harvard period, that relates to a comment Foster Nix made when I spoke with him several years ago. He recalled a Cornell summer school in Solid State Physics that he attended during the 1930's but couldn't remember any details about. He, at that time, thought that you might also have attended. If you did, perhaps you could tell me some things you remember about it. It would be interesting to document that, if in fact, it took place because indicates some early interest in the field in the thirties.

Bardeen:

I don't remember any summer school in Solid State Physics. I did visit Cornell to give a colloquium talk, perhaps on more than one occasion, but I don't remember being there for any extended period. The main thing I was interested in talking about at Cornell at that time was trying to apply some ideas of solid state to try to get the level density of nuclei. Bethe, of course, was the real expert in nuclear structure. I don't remember anything specifically related to solid state.

Hoddeson:

Let's return then to where we left off last time. We were at the beginning of your period at Minnesota. You had told me briefly about how you got the job. I would like to know, were there other possibilities for you at that time?

Bardeen:

I think there may have been other possibilities, but that was the only really good one, and that turned up through the influence of Vleck, who had been a professor at Minnesota in the twenties, be fore going to Wisconsin and Harvard. There were two retirements at Minnesota -- Erikson, who had been Department Head, and Zalomi[?] who was a member of the department. Both retired, so they had two openings. One of the post docs at Illinois, I think he had an NRC Fellowship, was Al Nier who worked with me on mass spectra. And, of course, they knew him and wanted to get him back and asked Vleck for other suggestions, and he suggested my name. So I went out there for an interview. I talked with Jack Tate and others about the position and looked like a good opportunity. So I accepted as an Assistant Professor.

Hoddeson:

So you did some teaching then.

Bardeen:

Yes.

Hoddeson:

What courses did you teach?

Bardeen:

Oh, a variety of courses ranging from upper class undergraduate level to graduate level. I think the only elementary course I taught was during the summer; one summer I taught Elementary Physics. In those days, to have the opportunity to teach during the summer and get a little extra money on the side was a real advantage.

Hoddeson:

How would you compare the physics program at Minnesota then with that at Harvard and Princeton in the same period?

Bardeen:

There wasn't any solid state physics when I was out there except for some experimental work. There was no theoretical work. I didn't go out there thinking of myself as a solid state physicist. I just went out there thinking of myself as a theoretical physicist ready to work on any problems. Some of the work I did there was to try to work out a theory for an isotope separation method which Al Nier was working on.

Hoddeson:

Did you get interested in that through Nier?

Bardeen:

Yes, through talk with Nier, I got interested in the problem.

Hoddeson:

I suppose that on this list, papers number 11 through 15 date from that period.

Bardeen:

Eleven is more a holdover from Harvard. Twelve was the one I was just talking about on isotope separation, which was stimulated by experimental work of Nier. This is a review article on electrical conductivity of metals which I wrote for the Journal of Applied Physics.

Hoddeson:

I'd like to ask you a question or two about that.

Bardeen:

Fifteen is also a result of this work with Nier. Thirteen and fourteen are on metals. I taught graduate level courses in beginning Solid State Physics and also taught upper graduate courses in Electro-magnetism, things like that. I taught a course in Geophysics, taught courses in Atomic Physics -- a variety of courses.

Hoddeson:

Did you have any graduate students there?

Bardeen:

None that reached the state of getting a Ph.D.

Hoddeson:

This review article appears to have been written at the same time that Seitz was working on his big book, his bible which laid down the solid state field. Were you aware of Seitz's effort at that time to write a book?

Bardeen:

I knew he was working on a book because we'd get together with him and if he had any spare time, he would put into writing on the book. He was the sort of person that never wasted any time, so if he had any free time, he would be writing away on the book. I knew he was writing the book but this article is a different sort of thing. This was not a treatise of the sort that he was writing, but just a review of the subject for the Journal of Applied Physics.

Hoddeson:

It indicates quite a bit of interest in this field to have a review article written in a journal like that.

Bardeen:

There had been a number of advances in understanding resistivity in metals and some of the work I did at Harvard was related to that. So this is written more from that point of view, starting at the beginning. So it's more an introductory sort of review but to try to bring it up to date with the most recent theoretical developments and then do an approximate comparison of theory and experiment.

Hoddeson:

There is an excellent bibliography at the end which I was particularly interested in. Also, I noticed that at the end, you point out three unsolved problems, the best known being superconductivity. Now very soon thereafter, you published an abstract on superconductivity.

Bardeen:

It is now known as the Kondo Effect which created a great deal of interest in the last 10 to 15 years particularly. What was the third one? That's the minimum in resistivity.

Hoddeson:

Semiconductors within completely filled D-bands.

Bardeen:

Oh yes. Mott pointed out a difficulty. If you just count valence electrons, semiconductors should be metals because they have incompletely filled bands according to the band theory, but they are not metallic but semiconductors on insulators. And so, this is an indication that the Bloch theory doesn't apply to these materials. This is another subject which has been developed. All three of these have been developed considerably in subsequent years.

Hoddeson:

What wee you thinking about most deeply in that period?

Bardeen:

I was working on superconductivity primarily. The only thing published was just an abstract. I sent around a few preprints for comments. I wrote a paper and sent it around for comments. It looked like quantitatively, it was off at least by a factor of 10 or so. And so I never published the full paper. About that time I left to go to Washington to work for the Navy so that got stopped. But some of the ideas are carried over into the present theory, that there is a small energy gap covering the entire Fermi surface and that was the basis for this sort of a model. But the way the energy gap was obtained was different than it was at that time.

Hoddeson:

Was this a subject that lots of people were very interested in at the time.

Bardeen:

I sent around preprints to people who were interested like Seitz and others and got comments from them.

Hoddeson:

: I'm interested in what the ideas going around in your head were here that were so similar to the Peiels Distortion and Frolick Mechanism later on.

Bardeen:

Well it was similar to that but I was trying to do it in regular three-dimensional model. All the same ideas were taken up by Frolick for a simpler one-dimensional model, and there it apparently does work. But it doesn't seem to work in three-dimensions as quantitatively. I image there might be some particular cases where you'd get a Peierls sort of distortion in a three-dimensional model. It's known that you can in two-dimension, layer structures.

Hoddeson:

Well then the war came and you were brought to the U.S. Naval Ordnance Lab in Washington, DC, where you remained from 1941-45 as a civilian physicist. How did this assignment come through?

Bardeen:

Originally I was just going to be there a year -- that's before we were in the was. It was the summer of 1941.

Hoddeson:

What brought you down there?

Bardeen:

Lynn Rombauer who had been a professor at Minnesota went to Washington to work on this program. He had worked earlier at the Department of Terrestrial Magnetism in Washington. He was actually building a high energy accelerator that would bend terrestrial magnetism but they knew him. The problem was to protect ships from magnetic mines by so-called degaussing them, putting coils on the ships, running current through the coils to try to counteract the magnetic field of the ships so that wouldn't be seen by the mines. This is a subject which was worked on very intensively in Great Britain because they had difficulties with magnetic mines. You couldn't reduce the field of the ship to zero, you could just reduce the magnitude which required them to make the mines more sensitive if the ship was going to trigger them. But that made them easier to sweep with mine sweepers. So it was a combination of degaussing and mine sweeping rather from degaussing alone which protected the ships. But that subject was getting pretty well completed by the time I was there and as the methods for doing this and the techniques were pretty well under way by the time I got there. So I got interested more in the other side of the problem. Well, one I was interested in was influencing fields of ships in general. was later put in charge of a group which was charged with determining the magnitude of various influence fields of ships, not only magnetic but also acoustic and later for pressure mines. Again, one wouldn't use one of these effects by itself, say acoustic. One would use acoustic in combination again with magnetic to make it harder to sweep. You have to get the right combination.

Hoddeson:

Did you work in a team at this lab?

Bardeen:

I was in charge of a group which eventually got up to about 90 people, on influencing fields of ships. This extended not only to mines, but also to influencing firing of torpedoes.

Hoddeson:

I didn't realize you were a director at that time with so many people working for you.

Bardeen:

Lynn Rombauer was in charge of that section of a Laboratory and I was a division head under him.

Hoddeson:

Did you enjoy the role of the group leader?

Bardeen:

No, not particularly.

Hoddeson:

Were the people in the group Ph.D. physicists?

Bardeen:

A lot of them were. This background which was useful here went back to my Geophysics days. I recruited some of my geophysics friends to come. The sorts of work we were doing was very similar to the problems you run into in geophysics.

Hoddeson:

Did it relate at all to the work you were doing in solid state physics in any way?

Bardeen:

No, it was just classical physics and boundary value problems, things of that sort.

Hoddeson:

Just a general question about communications during the war: was there much communication between the Naval Ordnance Lab and other laboratories at that time doing similar work?

Bardeen:

We got into work on torpedoes, because of a difficulty they ran into, and it influenced firing the torpedoes. They had magnetically fired torpedoes which were never really adequately tested by the Newport Torpedo Station, and when they took them out to the Pacific, they ran into trouble with premature firing. If the torpedo didn't behave right, if it popped out of the water or something like that, that triggered. They not only lost the torpedo but often gave away the position of the submarines. They turned off the magnetic firing device which would fired on impact. Then they found the impact firing device wouldn't work because I guess in the testing, they both were on, the impact firing device would work in combination with the magnetic firing device, but if they turned off the magnetic firing device, the impact firing device wouldn't work and so they had to make emergency repairs in the flux. And so you can imagine this caused a great amount of activity in the Naval Ordnance organization so they brought us in to see if we could help on the problem. I went up to the Naval Torpedo Station and talked with several people there involved in the program, and I made a number of trips up there. I also made trips to Westinghouse where they were involved in making torpedoes and in Seattle where they had some test stations for torpedoes. The interesting thing in that period was the interview with Einstein who sent the idea to Naval Ordnance for a method for firing torpedoes, a non-impact method of firing torpedoes. They sent me up to interview him, which I did and had a very interesting talk with him in his study on the second floor of a house he lived in at Princeton. He described a method based on change of inductance of a coil essentially, when it passed under the seal hull of a ship. To do this, you require alternating currents and alternating currents are absorbed by the shell hull of the torpedo assembly. We thought about this possibility and got around this by using plastic heads but we didn't like to go into that. So it worked out very slowly. When I brought this objection to Einstein, he said we should build a plastic window on the lower head or use a plastic head. As a result of this, we took up the problem more intensively to see if we could design a warhead with a plastic window. And some months later, we found a German torpedo which had come up on a beach somewhere which had a firing device of exactly the same. And they had apparently been using them for sometime. So the Germans were using the design which Einstein proposed.

Hoddeson:

How did he happen to be working on this problem?

Bardeen:

I think he was a consultant for the Navy. Just how he got to be a consultant for the Navy I don't know. I don't know he came up on this idea, whether he got some hint of it through some German source or came on it independently. I don't know.

Hoddeson:

I wonder if you could make a general comment about the effects of the war on your own scientific work.

Bardeen:

I don't think it added anything to my own scientific work. It made use of the background I had picked up in my geophysics days but my main interest was in atomic physics and quantum physics and I didn't use that at all.

Hoddeson:

It's often said that the war had a large effect on solid state physics in general through the work on materials. I wonder if you would comment on that.

Bardeen:

Solid state certainly did in semiconductors using silicon detectors in radar. It also did in metallurgy.

Hoddeson:

Do you think those fields would have gone in different directions had the war not intervened?

Bardeen:

If the war hadn't intervened, there probably wouldn't have been the incentive to develop microwaves as rapidly, and develop the detectors for microwaves. Vacuum tubes couldn't be used so they had to go back to the old cats' whisker detectors for radio.

Hoddeson:

In articles I have read it's stated that the two reasons for choosing silicon and germanium were that first they were simple and second, because of the tremendous amount of work that was done on them during the war. If that war work had not been done, do you think that these semiconductors would perhaps not have been chosen. Might they have stuck with the copper oxide and selenium?

Bardeen:

Before the war, he was working on silicon I think independently of the war effort so I think it would have gone ahead but perhaps a bit more slowly.

Hoddeson:

Then of course there wouldn't have been any interruption due to the war.

Bardeen:

I think the pressure was to try to develop an amplifying device using a semiconductor. I don't think the war had any direct effect on that. It probably slowed that down if anything because the people were all working on other subjects during the war. I think Seitz was working on semiconductors for radar detectors. I don't think he was thinking about amplifiers.

Hoddeson:

It's an interesting question, what impact did the war have on the field if any at all.

Bardeen:

It's hard to say. The direction which was followed was certainly different as a result of the war. We probably had better materials to work with due to the war efforts.

Hoddeson:

In what sense was the direction different?

Bardeen:

Well, in emphasizing detectors and looking for possible ways of getting gain, getting an active device.

Hoddeson:

Well in 1945, you chose to move on to Bell rather than to return to Minnesota. I'm interested in how you made this decision, what Fisk and Shockley's role was in hiring you at Bell, who recruited you and so on.

Bardeen:

I think Kelly had been Director of Research and later President of the Bell Telephone Laboratories and had a great deal of influence in getting this program on solid state physics started at Bell Labs after the war. Even before the war ended, they were beginning to make plans to get people together. I think it was Kelly's idea to get together a group which had both chemists and physicists so that they could get the interaction between chemistry and physics and materials development. He also wanted some people who had some of the modern ideas of applying quantum concepts and trying to understand the properties of materials and attack materials on a rather broad front from the theoretical and atomic structure through preparation of materials like chemists and metallurgists. And this group was formed just to look into semiconductors but into all aspects of solid state. Magnetism was a big part of it. There had been a lot of activity in magnetism before the war and this was a major element. The group also looked into dielectric materials ferro-electric materials and all sorts of materials. And when I went there I didn't just go there to join the group to work on the theoretical aspects but didn't have any particular branch in mind which I would concentrate on.

Hoddeson:

The story that you tell in some of your writings is that it was very crowded at Bell so they put you into an office with Brattain and Pearson and then you got involved in what they were doing. Is that the whole story?

Bardeen:

Just about the time the war ended in October or November 1945, Bell Labs brought in a lot of people for wartime research so the laboratories were overcrowded. This was Murray Hill. These wartime projects were still going on as the war ha just ended; the turnover to peace-time activity was just beginning, so these people were still there. They were building a new building at Murray Hill which hadn't been completed yet. We were due to move into the new building when it was completed but at the time I moved there we were still in the old building in rather crowded conditions.

Hoddeson:

The old building in Murray Hill?

Bardeen:

Murray Hill.

Hoddeson:

By the way, was Murray Hill named after the tow or was the town named after the Labs. Do you have any idea?

Bardeen:

I think Murray Hill was there before the labs. It was the name of the area.

Hoddeson:

Do you know if Fisk or Shockley had any direct role in hiring you or was it mainly Kelly who...?

Bardeen:

Oh yes, very direct. I met Kelly but my main contacts were through Shockley and Fisk.

Hoddeson:

Who actually invited you?

Bardeen:

I guess my first contact was with Shockley, and then Fisk, and then eventually I talked with Kelly. I think Shockley and Fisk tried to get me interested in the job and it was Kelly who actually made me the offer.

Hoddeson:

How did they explain the position to you? Did they tell you who would be working on mission-oriented work?

Bardeen:

No, it was to be a basic research study in materials.

Hoddeson:

So you expected a free hand?

Bardeen:

Since they worked on materials from a rather broad front, it wasn't very restrictive. I could work on whatever theoretical problems I wanted to in connection with materials, so from that point of view, it looked very good.

Hoddeson:

Was it a difficult decision to choose Bell over returning to Minnesota?

Bardeen:

From a financial viewpoint, I got about twice as much money at Bell Labs and really couldn't afford to go back to Minnesota for what they were offering me which wasn't much more than I was getting before the war. That was perhaps primarily because I was one of the first, probably the first, to return. Others returned later. I guess the authorities at Minnesota hadn't realized how things had changed. Tate, who was also involved in wartime research, and was really the prime figure at Minnesota arranged for an increased salary, and tried to get me to go back but the differential was too great. I'm not sure whether I would have gone back had it been even, but this looked like a very good opportunity at Bell Labs. I also enjoyed my associations at Minnesota. The major decision was made on economic grounds.

Hoddeson:

Very interesting. Was Bell in those days, considered to be a model of mission-oriented research, as it is now for example?

Bardeen:

I think it was always regarded as the prime industrial research laboratory and they do much more in the way of basic scientific research now which is not so directly related to missions. Of course, they did at that time. This was really an effort to get a research group started working on a broader range of things not directed toward any particular product but just to try to develop a field.

Hoddeson:

And yet, of course, all these problems had to be in some way tied to communications and materials which were being used in communications.

Bardeen:

Of course, materials were very widely used in communications which used all sorts of materials one way or another. So any advance you could make on almost any aspect would be useful. There had been a good bit of work in materials before the war, but it hadn't been particularly tied into the modern atomic viewpoint of matter, that's the applications of quantum theory to understand the properties of matter. So, the research was more empirical rather than trying to understand the properties from a basic point of view.

Hoddeson:

Now this began to change at Bell. I've learned that there were some study efforts going on where people got together to educate themselves. Even people who hadn't been trained in quantum theory such as Pearson and Brattain came to these sessions.

Bardeen:

Apparently, this had been a tradition for a long time at Bell; to have small study groups working on this problem or this area or that area.

Hoddeson:

Do you recall participating in any of these when you first arrived?

Bardeen:

Yes. None of use had worked on semiconductors during the war and one of the first things we did was to find out what went on during the war, what progress had been made. And the papers of Shockley, which were published just toward the beginning of the war, 1940 or 1941, on the properties of rectifiers. We were just beginning to get hold of those and study those. They had a good bit of material on semi-conductors. Aside from that, we went through Pauling's book on the chemical bond which had come out around 1940, I imagine. L.P. Pauling, The Nature of the Chemical Board and the Structure of Molecules and Crystals (Oxford, 1939).

Hoddeson:

Who was running the seminar? Who chose the books to study?

Bardeen:

It was just a group that got together and we would take turns leading the discussion.

Hoddeson:

How often did the group meet?

Bardeen:

At least once a week, sometimes twice a week.

Hoddeson:

Who else was in the group?

Bardeen:

The group working on semi-conductors -- Shockley, Walter Brattain, Gerald Pearson and I'm sure there were others there too, probably on the order of eight or ten together.

Hoddeson:

Alan Holden, Kittel perhaps?

Bardeen:

Alan Holden was there. Charlie Kittel? He came later.

Hoddeson:

Did Kelly ever come to these meetings?

Bardeen:

No. He may have in his earlier days. Earlier these courses were out-of-hours; people did the studying on their own time. Later the company allowed the courses on company time, so to speak.

Hoddeson:

Was the library fairly good in those days?

Bardeen:

I think it had a very good library.

Hoddeson:

Were there many visitors and visits to other places in that period?

Bardeen:

Well, we had visitors. We also had journal clubs and visitors giving talks, probably once a week or so. There was a good bit of activity. They also had consultants. Debye was a consultant.

Hoddeson:

How about Slater?

Bardeen:

I think he consulted on magnetrons and microwave devices, but not with our group. I think Debye is the main consultant that I remember in those days.

Hoddeson:

How did Debye interact with people? Did he give talks? Did he sit in a room and have people ask him questions? Did he run seminars? In what ways did he serve the researchers?

Bardeen:

We would tell him what problems we were working on and then try to get any suggestions he might have, what to do next or some literature he was familiar with and we weren't; he would refer us to that. Just get his general reactions on how we were proceeding on the problems.

Hoddeson:

Did he have a considerable impact on the quality of the research?

Bardeen:

I think he did at the start. Yes. But as we got into it more deeply, it was kind of outside of the field of his expertise. I think he was very valuable at the start.

Hoddeson:

How much more time do you have to spend on today's session?

Bardeen:

I should be leaving right now. I have a luncheon meeting I'm supposed to be at.

Hoddeson:

O.K. Let me save my next question for next time. Thank you