A. Brian Pippard

Hoddeson:

This interview is intended to be on the subject of superconductivity in the early 50s — mainly on your own work. But before we start, I think you wanted to make a comment on polishing.

Pippard:

The question arose in a letter from Paul Hoch: How did I know in 1954 that they knew how to polish copper and metals in Chicago? And I have just written to him to say that I didn’t think that Morrel Cohen told me, and probably, although I couldn’t remember, I just knew on the grapevine. The grapevine was much more efficient in those days simply because there were so fewer physicists actually in any given field. But on thinking it over, I may have told him an untruth, because it is just possible that I told Morrel my ideas about how one might measure the Fermi surface and said that the only problem was that I didn’t know how to polish copper well enough. And it may be that he said that that would be no problem in Chicago. But he certainly said to me years afterwards that he initiated the invitation for me to go to Chicago, and that is quite possible. I don’t think by the way it matters very much, but just for the record, I hope I haven’t misled Paul too much.

Hoddeson:

I’ll let him know. I’ll be seeing him at a meeting in a few days. Let’s move on then to superconductivity. I believe the start for you was in or before 1946. Is that right?

Pippard:

For me it was in October 1945.

Hoddeson:

And what happened?

Pippard:

I came back as a research student, having worked for four years on radar. But I had already, before leaving the radar establishment, had some discussion with Allen, Shoenberg and John Ashmead who were the three senior staff members in low temperature at that time.

Baym:

Jack Allen?

Pippard:

Jack Allen, yes. And now here we get indiscreet. They told me that before the war started, Hugh Barkla, who was a research student, had begun to do some work on microwave measurements following the work that Heinz London published before the war and that Hugh was anxious to come back and continue it, but John Ashmead had insisted that if anybody continued it, it must be me because I had been working on microwaves during the war, and Hugh had not. So Hugh was fobbed off with ferroelectric measurements of barium titanate and I was given a free run of the microwave measurements in superconductors following the London work. And that was in Autumn 1945.

Hoddeson:

I see. Did anybody else attempt to study superconductors using microwaves, or were you the only one to your knowledge?

Pippard:

No, John Slater at MIT. I heard him say about that time that he was extremely keen that the expertise in microwave work that the Radiation Lab had built up should not be lost to physics. And it had seemed to him that since hυ = kT would hold at 1° absolute with microwave frequencies it would be obvious to have a low-temperature program involving microwaves at MIT. So with that sort of general principle he’d got some low-temperature work started, and indeed Emanuel Maxwell and Paul Marcus and Slater began work on measuring microwave losses in superconductors.

Hoddeson:

Were you in correspondence with him?

Pippard:

Only when I had published a note did that become clear, and about the same time it became clear that Bill Fairbank, who was then at Amherst College, had started to work on measuring the Q’s of 3-cm cavities. He set up to measure losses in a rectangular 3-cm waveguide made of tin. His results were dreadful, for a reason which only became clear a few years later when Bob Chambers and I did some measurements of the anomalous skin effect, when we took a sample which was cast and had a good surface (and I think Bob had one which was electropolished). And then we rubbed the surface and also polished a copper surface with Brasso, a copper polish, and that ruined the results. The resistance went up considerably, and what Bill Fairbank found was that his tin, which had been carefully milled out of a solid block, had a very big residual loss at absolute zero. So he got the general effect of the resistance going down like in that graph you’ve got there, but instead of going down to nought, it went down to 10 or 20% of the value above the transition temperature. But I think one thing which was clear about all that work was that both Fairbank and the Maxwell/Marcus/Slater group had had far too much experience of ordinary electronics! And they sought to set up measuring gear which would work more or less automatically, and feed out results without human intervention. The result of this was that they lost a good year or so in setting up the instrumentation, whereas I who had always had a great aversion to electronic circuitry, set up just with some waveguides and transmission lines, a crystal rectifier and a galvanometer. And I got my first results I think two months after I came back to Cambridge, so that by using extremely simple techniques I got reliable results very much faster than the other two teams did. And that was quite a lesson for me about not trying to be too fancy about any experiments. But those I think were the first groups. Then a year or two afterwards there were a couple of characters called Grebenkemper and Hagen.

Hoddeson:

Do you know where they were?

Pippard:

I cannot remember. Somewhere in America. And they did very nice experiments where they made quite a nice cavity out of a superconductor and then they excited it and let it decay naturally, and they measured the time constant for the natural decay of the cavity rather than measuring Q by sweeping frequency, which was the technique which all the rest of us had been using. Actually it was worked out quite nicely. To my mind it was an extremely elegant and at the same time extremely time consuming and probably inaccurate way of doing the job. In other words, I don’t approve of it. I rather admire anybody who could make it work, rather than actually thinking it was a good way of doing it. No, the simple sweeping frequency was I think, at the stage of technique which was then available, the natural way of doing the job.

Hoddeson:

How closely did David Shoenberg work with you at that time or watch what you were doing?

Pippard:

He was officially my supervisor, but he was quite content to let me get on with the job and we just talked from time to time about results. But he would be the first to admit that in fact he left me to my own devices. I had had four years of experience with microwave design which he had not, and I knew far better what I was doing than he did. He gave very good advice about general ideas and about designing experiments, and also about the theory of superconductivity. So he helped a great deal in interpreting results, and also, I may say, I found him extremely useful as a critic of my English. There’s nothing like having a Russian to teach you how to write English sentences which are grammatical. And in particular I remember he taught me how to make a sentence lead on naturally to the next, so that the reader can follow what the thought is instead of having a lot of disparate and disconnected sentences. So I do owe quite a lot to David for his general help in physics and writing.

Hoddeson:

Were you also in communication with Heinz London, whose work you were picking up and improving?

Pippard:

At the very early stage, yes. I can’t remember when I first met Heinz but he was around in Cambridge. That is, he dropped in from time to time — he was in Harwell at the time — and he dropped in and talked, and he always kept an interest in what I was doing because it was really very much along the lines of his own experiments before the war. He was extremely helpful.

Hoddeson:

Was he continuing any work along those lines?

Pippard:

No, no. He had become mixed up with the war effort and he stayed on in Harwell, and as far as I know the major work which he was doing at the time was fractionating oxygen isotopes with a fractionating column. I don’t know why. But some years later when helium-3 was discovered, it was London, I think, who first suggested that you could make a good refrigerator by separating helium-3 and helium-4 mixtures.

Baym:

That was a dilution refrigerator.

Pippard:

I think London is the originator of that idea.

Hoddeson:

In one of the early papers, I guess it’s this 1947 paper in the Proceedings of the International Conference on Low-Temperature Physics (“Resistance of Superconductors at Microwave Frequencies,” 113-118, see p. 117) you express your belief that a complete classical solution of the problem is possible. How did you come to feel that way?

Pippard:

That it was possible? Goodness me, I have forgotten this altogether. Let me see what I said and then I’ll tell you why I said it. That’s right. I say the general case, though easily formulated, is difficult to solve. I had formulated and solved it for the one-dimensional case, that is, electrons that were going straight up the surface and back again normal but kept being deflected sideways and carrying a current. Because of that, I had formulated the three-dimensional case, but I thought enough about approximations and that sort of thing to be confident that it wasn’t beyond the power of mathematics to do it. And it was when I got in touch with E. Sondheimer, that he formulated it more tidily. He also found the problem too much for him, but at least he knew where to go for an answer, which was to Reuter.

Hoddeson:

This was dealing mainly with the anomalous skin effect.

Pippard:

Yes, that is the classical problem. I’m sure that that’s what I was talking about when I say I thought it was solvable. The skin resistance of a normal metal at low temperature and high frequency is not in agreement with measurements of the direct-current conductivity. I had been investigating theoretically London’s suggested explanation and I believed a complete classical solution of the problem was possible. That’s the reason I had been writing down the — well you see I didn’t know things like Boltzmann’s equation. It is very important to realize that we were not so maleducated in those days that our minds were full of a lot of junk which we wouldn’t use. At that time my mind was almost completely empty of junk, even that which I could have used. I had been educated much more in the manner of long, detailed historical developments of various ideas like fundamental particles, passage of charged particles through matter, ionization of gases, all that sort of thing, with a blow-by-blow account of the experiments. But little things like the Boltzmann equation had been regarded as far too high for undergraduates in physics in those days. The Cavendish tradition was wholly experimental and the amount of theory that was taught was very, very little indeed. So having then worked during the war on radar, I came back quite innocent of things like Boltzmann’s equation. I had to invent the formulations myself and simply follow the paths of electrons through the metal. And what Sondheimer did was to remark that there was a thing called the Boltzmann equation, and if you wrote that down, then you got the same answer, but more tidily.

Hoddeson:

So you got to the anomalous skin effect problem by studying the high—frequency resistance of superconductors, and then you and Sondheimer and Reuter developed a more rigorous mathematical formulation. And then you developed the “ineffectiveness concept” formulation for that problem. It seems that then, through that problem, you came back to improving the London theory, but using the non-local formulation that came out of the Reuter and Sondheimer work. Is that the sequence?

Pippard:

No.

Hoddeson:

Perhaps you could tell me the right story.

Pippard:

I’ll try to get the historical development clear, as I saw it. First of all, I started from London’s observation of losses in superconductors — that was my primary purpose, to show how the resistance dropped as you got below the transition temperature. At the same time I was aware of London’s work on the normal metals. And David Shoenberg had put me in touch with all his papers, and I had read them. So when I noticed that the normal resistance of the metal in no way corresponded to the theory of the skin effect, that was no surprise at all. But London had assumed that you could take the skin layer of the metal and treat it like a thin film with the electrons suffering inelastic scattering at one of the surfaces. And that’s as far as he’d got; his measurements were not good enough to test whether that was right or not. In other words, he realized it was a mean free path effect but he had been making an analogy with Fuchs’ theory. Fuchs was at Bristol at the same time London was.

Hoddeson:

Is that the Fuchs that had worked at Los Alamos, the well-known spy?

Baym:

Oh, you mean Klaus.

Pippard:

Klaus Fuchs. He had worked out a full theory of the conductivity of thin films.

Hoddeson:

Didn’t he consider ellipsoidal surfaces using specular reflection?

Pippard:

No, I think spherical surfaces, and using random scattering. There’s no effect with specular reflection. And that was one of the things which is clearly wrong in London’s analogy: that with specular reflection you still get an enormous anomalous skin effect. Though that we didn’t know. My model calculations had convinced me that in the one-dimensional simplification, the electrons could take momentum away from the surface layer back into the metal and that will still be dissipative in the way that would not be so with a thin film. Formally it’s the same thing in three dimensions. I had worked entirely by myself without any help at all, and had tried turning an integral equation into a set of algebraic equations. I had a finite difference formulation and I got some results out of that, some general results about the integral of ETI through the surface layer and that sort of thing, but it obviously wasn’t leading very far. And then the ineffectiveness concept came to me, and it’s that time that I was talking about — I can’t remember now — I think this paper must have had the ineffectiveness concept in it, but I can’t remember. Now, it was about that stage anyhow that I became friendly with Sondheimer who was a research student at Trinity and asked him if he could throw any light on the problem. So then what happened was he formulated it tidily with Boltzmann’s equation, all that sort of thing and, oh, here we get into some complication: he had to call in Reuter and he was very much ashamed of having had to do that. When Reuter showed specular reflection could be solved by the standard Fourier integral method which is in Titchmarsh’s book [Introduction of the Theory of Fourier Integrals (Oxford, 1937)], Sondheimer felt that he really had let himself down by not knowing that. He is not in the least bit ashamed about the other one but Reuter worked out the Wiener-Hopf solution — have you got letters from Paul Hoch, his letters from Sondheimer about this?

Hoddeson:

I don’t, but Paul has them.

Pippard:

Paul has got the correspondence relating to this. I had said that Reuter solved the problem overnight. Sondheimer said no, he didn’t solve the problem overnight. He got as far as he could rather quickly, but he was nonplussed to find that he couldn’t find where all the zeros and poles were in the complex plane. He never did find them all. So it wasn’t quite as simple as I made it out to be. But about the same time, I must have gone over to Germany, as I did in ‘47, I think to a conference in Gottingen where I talked with Heisenberg and with his student Koppe about that. Shortly after that conference, or shortly before — again, my memory’s sort of uncertain — Koppe wrote me to say that those equations which I had published in the Royal Society could a) be compressed into a much more tidy form, and that was the form that Sondheimer found and b) could be solved by the method of Fourier integrals. So Koppe comes in at that point as far as I’m concerned. Now where are we? Things go underground then for a while, and nothing happens in my neck of the woods until round about ‘51 or so when I began to do measurements on dirty superconductors to see what happened as the scattering became greater. And it was then that I noticed the skin depth began to increase markedly a percent or two in indium and tin.

Hoddeson:

You had also gone up in frequency.

Pippard:

That was trivial in a sense. I had done one or two other things in between like finding that penetration depth didn’t change very much with magnetic field and things of that sort, but that’s a slightly different story. Well, no, it isn’t a slightly different story, because finding then that the penetration depth didn’t change with magnetic field led me to believe that the effect of the field must be spread into a much greater depth than the penetration depth. And the argument for this was something that Heinz London had suggested to me quite independently before. When we were talking one evening after dinner, he pointed out to me that the penetration depth ought to change in a magnetic field. We didn’t know whether it did or not but he thought it ought to because the magnetic analogy of one of Maxwell’s equations was that dS/dH = dM/dT. Now we know that because the penetration depth changes with temperature, that therefore in a constant magnetic field the magnetization changes with temperature. Consequently, the entropy of the superconductor must change when a magnetic field is applied. He then went on to say — I think but I am not sure, I may have calculated this myself afterwards but I suspect that London would not have let the detailed calculation slip — I think at that stage we were aware that if all of that entropy change were contained within the superconducting penetration layer, it would amount to something like 25% of the entropy difference between the superconductor and the normal material. Therefore you would suspect that quite a lot would happen to the electrodynamics if the magnetic field was throwing the thing towards the normal state to the extent of 25%. So as soon as I found that nothing like 25%, but not more than 2%, change was occurring [Paul Richards or Millie Spirak (Dresselhaus) got the wrong direction a few years later at a lower frequency. ABP], it was a very natural conclusion to jump to that the effect of the magnetic field was extending into a much greater depth and it was from that that the coherence idea came into my mind.

Baym:

I’m not sure why you got into looking at dirty materials.

Pippard:

Nor am I (laughter). No, really I would guess that one of the motives — it clearly was not caprice because I spent an awful lot of time on it — one of the things was the anomalous skin effect, to see how the skin resistance depended on mean free path. [Clearly my memory was very sketchy on this point, but I’ve looked at a few papers since and am pretty sure that the interest in dirty metals came from coherence more than anything else. Right from the start I was fairly sure [Proc. Camb. Phil. Soc. 47 (1951), 617] that coherence would explain the surface energy between normal and superconducting states, and that the decrease of coherence length when impurity was added would account for the decrease of surface energy, and ultimately for the flux trapping in really dirty metals. Hence the field variation of penetration depth should be greater, and it was this that became the goal of the microwave experiments, leading to Proc. Roy. Soc. A216 (1953), 547. ABP] And having done this it was obvious that the right thing to do was to extend the thing into the superconducting domain in the hope of finding some explanation of the more detailed theory of resistance, because no model that I or anybody else had put together explained in detail the way in which the resistance dropped out as you got below the transition temperature.

Not just was it wrong in detail but the whole algebraic form was wrong, and then everything was in a mess. By the time I wrote my thesis in ‘49, all I could put in there was what was an anomalous skin effect plus London equations. And absolutely everything came out wrong in detail. So I suspect that it was to try to get further light on that, as well as on the anomalous skin effect, that it was natural then to put lots of junk in and deal with that. And then finding the penetration depth changed. I already had this idea of a coherence length, and, oh, I also knew quite a lot about Heisenberg’s theory of superconductivity. Now, Heisenberg’s theory of superconductivity was ingenious and wholly wrong. One of the things which I had pleasure in talking to Heisenberg about was the first form of the theory, where he got his thermodynamics incorrect. So I went to see him in Gottingen and tell him he couldn’t get away with thermodynamics like that, because what he’d managed to prove was not that superconductors obeyed the London equation but that all metals were superconductors (laughter). Incidentally this is the thing which this chap whose name I forget now has written recently in Physical Review Letters, in which he showed that collisionless plasmas should obey the London equation, although it’s wrapped up in appalling Lagrangian jargon and I hadn’t the patience or the intelligence to see what he’s doing.

I am convinced that he has fallen into exactly the same trap of using the wrong thermodynamic potential which a man called Cook did in 1937 in Physical Review. It’s one of these standard mistakes. Anyhow, I was delighted to find, not only that Heisenberg was quite happy to be told that he got it wrong, but he then set about modifying the theory so that he got his London equations without thermodynamics. He had this fantastic idea of a sort of condensed crust on the Fermi surface. And as the temperature was lowered, the crust extended and covered more of the surface. When you put a current through, the normal electrons carry the current at first. This displaces the Fermi surface, and the crust migrates round into such a position that it takes over the current-carrying and leaves the normal electrons, which suffer collisions, free to be isotropic as they want to be. Now I already noted in my thesis, which was in ‘49, that if this idea of Heisenberg’s had any truth in it, then when you got an anomalous skin effect happening, the general properties of the anomalous skin effect would also be possessed by the superconducting current, so that you get a similar set of equations governing the London current. When I found that the penetration depth changed with impurity, I went straight back to that, and sort of pinned up the theory in such a way that it could behave reasonably in pure material.

Baym:

The theory of Heisenberg?

Pippard:

Yes. Or rather my interpretation of Heisenberg’s theory in skin effect conditions, which he hadn’t dealt with. And I found that it gave a very nice little integral equation for the supercurrent which had the penetration depth increasing as impurity was added. And in effect, that’s the equation which I published. I didn’t for a moment believe that Heisenberg’s explanation was right but it was a very good model to lead one to an equation which made sense.

Hoddeson:

I found you’ve mentioned many people in your papers who were working in the 50s, except for Ginzburg and Landau. You don’t ever mention them. Were you aware of their 1950 work?

Pippard:

About ‘52 was it?

Hoddeson:

‘50.

Pippard:

Yes. I think that we probably got it fairly shortly afterwards and recognized its power, its importance. And David Shoenberg ran off the English translation which was duplicated and handed around and used. Now that does raise the question: Why didn’t we take it seriously? Well, we took it seriously in a sense. It was just about the time that I got these first results on the change of penetration depth with impurity. And if you took Ginzburg-Landau seriously, the relation between penetration depth and the number of superconducting electrons was unique and inviolate and implied that the number of superconducting electrons was changing with impurity. Since the thermodynamical properties were not changing with impurity, I didn’t believe it. It was not until Gor’kov, in fact, that this difficulty was resolved. It’s very important not to see Ginzburg-Landau with hindsight. Of course Landau was a highly respected figure with us. David Shoenberg had been in Moscow before the war and knew Landau and in fact had published Landau’s theories of the de Haas-van Alphen effect as an appendix of one of his own papers when Landau was in prison. There’s no question about lack of respect. And we all knew from Ginzburg’s writings during the war on superconductivity that he was a powerful enough man. No, the problem was that the first paper of Ginzburg-Landau was obviously imaginative, but it was also highly speculative and introduced quite arbitrarily a wave function whose meaning they did not know, and they ascribed to some rather unusual boundary conditions. They pulled rabbits out of the hat right, left and center, none of which had been experimentally verified. So you see one was bound to look on this paper as a speculative venture into the theory of superconductivity. And since there were so many speculative ventures by respectable people at that time, like Slater or Bardeen…

Hoddeson:

I was going to come to the Bardeen-Frohlich work next, how you felt about that.

Pippard:

That comes a little bit later.

Hoddeson:

That’s also 1950.

Pippard:

My goodness, was it? Frohlich was as early as 1950, was he?

Hoddeson:

The spring of 1950.

Pippard:

Yes, well, we took Fr5hlich seriously. Not, I mean, the details of the theory, but the isotope effect. As soon as Frohlich suggested that there should be an isotope effect, we started to work measuring it. In fact, I was at a meeting when —

Hoddeson:

Wait a minute, the isotope effect — you started measuring it from Frohlich’s unpublished work?

Pippard:

Yes.

Hoddeson:

Tell me that story; I don’t know this.

Pippard:

You see he gave a talk at a conference that I was at, in the course of which he remarked that it would be highly desirable to measure the transition temperature of different isotopic compositions. As soon as he finished, I said that we were all set up in the Mond to measure accurately transition temperatures of very small samples — that is, Shoenberg and Desirant had set up a magnetometer of high sensitivity. If only someone would make the pure isotopic samples, we would be happy to measure it. At which, Allen from Harwell, a nuclear physicist whom I’d known during the war, happened to be in the audience and got up and said “well, I’ll make them for you.” And then it started. At this stage the discussion becomes scurrilous because after some months when Allen had been working with his isotope separator at Harwell producing the things, he then wrote a somewhat embarrassed letter to David Shoenberg to say that it seemed that he had promised Kurt Mendelssohn the first batch of isotopes and would we mind very much if he let Mendelssohn have the first batch and then within ten days or a fortnight, he could make the second batch, which he’d let us have.

Well we actually minded very much (laughter) but we’re all gentlemen in Cambridge. So as soon as we heard what was happening, we got the apparatus, Shoenberg, Michael Jock, and I, and we got it going extremely well, absolutely tuned up. As soon as the isotopes arrived, they were put in, and Lock and Shoenberg started work. I didn’t do any of the experimental work because it wasn’t my apparatus. Lock was the workhorse; he was a research student, and an extremely competent one, and he worked very hard. We not only measured transition temperatures, we measured the transition field as a function of temperature at the same time with high precision, so we could follow the variation and we hoped to see how the electronic specific heat was changing with isotope composition.

So within I think about two weeks we got some very nice curves and we did some nice analysis and showed that the square root law was pretty well obeyed, there was no significant change in the electronic specific heat γ with isotopic composition and so on. And we then wrote up a letter to Nature and sent a copy of this to Mendelssohn, saying that we understood that he was working on the subject also, that this was what we proposed to send to Nature, but since it would be a mistake not to collaborate, if he cared to add anything to what we had to say, we were willing to hold up submission for publication for a week. It put the skids under poor Mendelssohn, and he and his research student wrote up another letter and we sent them in together and they were published. And I’m afraid that the contrast between the two indicates that he would have been much wiser to have kept quiet about it, because his results were pretty scrappy. So we got a very satisfactory revenge, and we felt much happier and we’ve never had any twinges of conscience whatever about it. So surely that was later than 1950, wasn’t it?

Hoddeson:

I only know the effect was reported in the United States by two groups.

Pippard:

Oh yes, there was Maxwell.

Hoddeson:

Maxwell and Serin.

Pippard:

And Serin, yes.

Hoddeson:

And that was in March 1950.

Pippard:

They were certainly before us, yes. The publications were. I think that we were doing the work before their publications were out. The whole thing was happening very fast at that moment. I think it would be a good idea to stop at this stage for dinner, it’s a good point, and then take things in due course.

Baym:

Wasn’t there was a general confusion through the 30s and early 40s over type I versus type II superconductivity?

Pippard:

No, there wasn’t. The phrases were not known.

Baym:

Exactly. Could you just give a one-paragraph overview of the extent to which theories were for type I and experiments for type II, and vice versa?

Pippard:

As far as I know, nobody took any of the experiments on what we now call type II or dirty superconductors seriously. Schubnikov did a lot of experiments before he was finally imprisoned and liquidated. And David Shoenberg was the leader in this country, I think, of the school of thought that felt that it was just dirt. They were potentially interesting but they were uninterpretable. And as far as I know nobody else took any notice whatever. Except Mendelssohn who did research on his own and also believed in Schubnikov’s results. And he had his idea of the sponge, with superconducting filaments filling the space. So the flux was trapped and couldn’t get out. But it seems to me that all the views on everything concerned with superconductivity then were highly phenomenological and very confused except for the absolutely clear-cut things like the thermodynamics of pure super-conductors, relating its specific heat and the critical field and all that sort of thing. As soon as you’ve got either hysteretic and irreversible phenomena, well that was just because it was dirty and the stuff couldn’t get out, and that’s about as far as the theory went.

Baym:

Now at which step did this become clear?

Pippard:

Oh, that’s very hard to say. I suspect that there was always much greater respect for these things in Russia than elsewhere, because Landau had worked on the intermediate state, and published the theory of it, actually two theories. There was always some good experimenting going on by various odd people trying to find the size of the superconducting domains and so on. Sharvin certainly was. Before that there were several people — Meshkovsky and Shalnikov — but Sharvin was certainly one of the later exponents round about 1950 and did some beautiful work on that. And I think as a result of that continuing interest in the intermediate state it was possible for Abrikosov to think of applying the Ginzburg—Landau theory, even though as you know Landau was extremely offended by the whole thing and cut up…

Baym:

I think it was Ginzburg who said that Abrikosov’s story that Landau prevented his publication of this simply isn’t true.

Pippard:

Yes, yes.

Hoddeson:

Lifschitz is one of the people who says it isn’t true.

Pippard:

I thought that Lifschitz and Abrikosov would have been absolutely poles apart in their attitudes, because Abrikosov is not well thought of for his political views on one thing and another, whereas Lifschitz is as liberal as you can allow yourself to be without being dissident. Shall we put it this way, that nobody can be neutral in Russia on the question of Abrikosov and his relations with Landau, and if there was to be any polarization I would expect it to be Ginzburg on the side of Abrikosov — but not very much. And Lifschitz against Abrikosov rather strongly.

Baym:

They are both against Abrikosov.

Pippard:

Ginzburg gives the impression of being politically very neutral. I don’t mean he’s innocent, but I think he’s a studied neutral in these things, whereas Lifschitz, given half a chance, is not neutral. Yes, well, anyhow whether or not Abrikosov was hindered or helped by Landau I think the whole atmosphere in Russia was much more favorable towards theories of this sort. Of course in Cambridge at the same time there was quite a lot of work going on about the intermediate state. David Shoenberg was working with various people like Andrew and others testing Landau’s theory and of course that was because David Shoenberg was highly sympathetic to the Russian school of theory and experiments on superconductivity because of his knowledge and friendship with them. But we were not theoretically inclined in the same way that Abrikosov was so that, as you pointed out, we didn’t take the Ginzburg-Landau theory seriously. It was bad. We share the blame for that but I’ve already explained how we were worried by the penetration depth changing.

As far as this country is concerned, it was Goodman who pointed out at a Conference about 1954 or thereabouts, 1956 probably, that we ought to look at the Abrikosov theory seriously because it did contain, it seemed, within it the seeds of an explanation of what happened in dirty superconductors. I was playing the line “shortened coherence”; when the coherence length became smaller than the penetration depth, then the interphase surface energy went negative. But of course Abrikosov had quantified the thing — he had picked up not merely the idea that flux penetration was favorable, but he had quantized flux and was far, far more advanced than I could ever have been on that. Goodman was the main instrument in this country in deflecting interest from the merely qualitative ideas of a shortened coherence length towards the quantitative models of Abrikosov. But from then onwards, as far as we were concerned in this country, we were entirely on the side that dirty superconductors were a perfectly respectable thing and that the idea of flux penetration in type II super-conductors was a perfectly proper thing to work on. Now I can’t remember when serious work started on type II super-conductors — I suspect after the work Kunzler and Klauder had done that suggested the use of superconducting magnets for high-field work. That was 1957. It was from then onwards that there was a big, big experimental program mounted at Bell Labs, Westinghouse, and G.E. in particular to sort out the dirty superconductors, and flux pinning and things of that sort. But that then becomes modern history. I

t’s somewhere between Abrikosov and Gor’kov that the picture changes. The work of Gor’kov really substantiates the Ginzburg-Landau equations and makes them respectable. It must be remembered that in the West the pressure in this direction has always been American, basically American pressure. That is, the Ginzburg-Landau theory had always been very much more to the taste of the American physicists than to us much more empirical practical experimentalists, and I think we’ve always had more interest in genuinely dirty, inhomogeneous, mucky superconductors than in homogeneous impure things, which are a theorist’s delight. We turn things over to the metallurgists and say “get on with high field magnets, and good luck” and this is rather fun and worth doing. But all the fine details about flux line structure and so on, on the whole have not been all that amusing for us. There is an enormous difference between the climates in the two countries. In this sort of matter England is still far more empirical in its attitude.

Baym:

That’s funny because Wilson left Cambridge to go to Leipzig despairing not only was there reputedly only nuclear physicists there but it was also too mathematical.

Pippard:

Is that so? But he was a member of the mathematics faculty and under the dominance of Dirac and probably would not have, in a sense, noticed the existence of Rutherford, who was physics faculty. Incidentally, it was Wilson’s lectures on thermodynamics that I attended as an undergraduate. It was them that inspired me to love the subject. He was a beautiful lecturer. The first time I ever met Wilson was a few years after the war. I was introduced to him by Sondheimer because he was Wilson’s student, and I told him I had attended his lectures and had enormously enjoyed them, and he said, “I always thought thermodynamics was the dullest of subjects.” It didn’t stop him from lecturing with superb clarity.

Hoddeson:

Well, thank you very much for everything.