Sidney Millman

Bromberg:

We should introduce to the tape the fact that this is Dr. Sidney Millman who is at his Bell Labs office, and I'm Joan Bromberg and it is the 5th of June, 1984.

Millman:

For the record, I am not now officially a member of Bell Laboratories. I retired 11 years ago.

Bromberg:

So let's talk about Townes. You arranged for a consulting agreement for Charles Townes from Columbia to be coming and consulting at Bell Labs. I do want to ask about one thing that's even prior to the time when Townes starts to work with Schawlow on what eventually became their optical and infrared maser paper and that is, Schawlow recalls or thinks he recalls that at one point around 1956 or 1957 you came to him and suggested that he change from doing research in superconductivity to working on masers. Is that something you have any memory of? He decided at that time he didn't really want to make that switch.

Millman:

I have no recollection of that, whatsoever. I might say sort of peripherally that after all, in Physical Research, the kind of people we had, such as Schawlow and others, the Director of Physical Research was not one who was telling the people or suggesting what they ought to be working on. In the course of visiting his laboratory I might have talked about what he was doing. I may have tossed out a casual comment, but not in any sense of directing, as might be done in another organization. This was not the way of the physicists in Physical Research.

Bromberg:

So, now I'd like to know your memories of the work on the laser. First of all, did this come to your attention as early as the period of collaboration? The collaboration must have started around October, 1957.

Millman:

Collaboration with Schawlow?

Bromberg:

Schawlow and Townes.

Millman:

The consulting arrangement probably got started earlier.

Bromberg:

Yes, of course.

Millman:

I've forgotten when. Maybe 1955.

Bromberg:

Well, Townes told me that even when he was on sabbatical from 1955 to 1956 he was consulting at that time already.

Millman:

That's probably right.

Bromberg:

And he said that one of the things he did was to come to alert the Bell Laboratories people to the importance that a solid state maser would have in communications.

Millman:

That's right. That sounds entirely reasonable, yes.

Bromberg:

Okay. So then he was here for a while and then it's Schawlow's memory that it was about October 1957 that he and Townes actually came together and began to discuss the optical maser. What I want to do is to begin to find out when this percolated up to you. I think the paper was written up around July 1958, the patent was applied for [about then], so I'd like to ask you if you have any memories during that time or if you have any memories as the paper became...

Millman:

I don't have any specific memories as to my first hearing of their work on optical masers, for one thing, I was not Schawlow's department head. There was a layer of organization between Schawlow and me. In other words, I was Director of Physical Research and Theodore H. Geballe was his department head. So he might have discussed things a little more fully with his colleagues in the department and particularly with Ted Geballe, and not necessarily with me. But almost certainly I might have heard about it either from Schawlow or from Geballe; if he had given a seminar, as we used to do, I almost certainly would have attended it, but I don't have any vivid recollection of this.

Bromberg:

Javan was coming in just about that time. I understand that you went to Columbia and spoke to Javan there, is that right?

Millman:

Yes. Well, we went periodically recruiting to the Columbia Physics Department, at that time Schawlow was probably there as a recruiter, too, in addition to Stanley Geschwind, who a few years earlier did his Ph.D. thesis research under Professor Townes. Javan was coming to the end of his stay at Columbia. I think he not only got his Ph.D. under [Townes] but he spent an additional postdoctoral couple of years. He was in the process of looking for some permanent post, and I remember talking to Javan for quite some time at a meeting of the American Physical Society, which happened to be in that year, at the Hotel New Yorker. I remember that. And we talked for a long time. I knew from Townes and I knew from him, that he was interested in going to a university. Now I'm not sure as to which university offer he was fielding, probably the University of Pennsylvania. And I talked to Javan in quite frank terms. I said to Ali Javan, that I knew of his interests and respected his interest in going to a university, but I thought that he would in the long run have a better choice of acquiring a more senior university position if he came to Bell Labs first for a few years and did the kind of work I would expect him to do, knowing of the recommendations. I heard about him from Charlie Townes. And on that basis, I believe he decided to join Bell Laboratories.

Bromberg:

Was he already thinking about gas lasers at that time?

Millman:

No, not at all. I think at that time he was interested in working on helium. There was something of physics value in polarizing helium atoms; he built up quite a laboratory with magnets and so on, which he later on didn't use, in order to work on helium, not on lasers at all. Now his interests probably changed soon after the publication of the Schawlow-Townes paper.

Bromberg:

Which he must have heard about as soon as he got here I would guess.

Millman:

When did he come here according to your records?

Bromberg:

That was just something I picked up from the archives this afternoon, August 1958.

Millman:

August 1958. Well, okay. Then he might have heard of it but he certainly didn't work on this much before December 1958, when the Schawlow-Townes paper was published. So when he came here, that might have been soon after they submitted their paper. I'm sure he heard about this, and evidently that didn't take fire immediately because he went on and built up a laboratory which was quite massive. A few months after he joined, he decided he was going to work on the laser. He must have gotten the idea on his helium-neon laser quite early in the game. He published a paper in Physical Review Letters, July 15, 1959 which is not many months after the publication of the paper of Schawlow and Townes and way before any optical maser worked. Maiman's laser came a year later. However, he was perhaps over-cautious or felt as if one would have to find the right experimental conditions for this population inversion and so on, in order to get laser action in this helium-neon scheme that he proposed. He did two things. He not only did some high level quantum mechanical calculations about collision cross-sections and life time of relevant atomic states, but he felt that that wasn't enough, that he needed to determine experimentally what conditions will be optimum for laser action. So he built up a laboratory.

Now he also realized that he himself was not especially gifted as a laboratory person, if by that one means somebody who makes the apparatus sing. Some people are unusually good at that and some people are quite poor. I can mention my Herr Professor Rabi who is second to none as a top-notch experimental physicist in the true sense of a physicist who knows what experiments are worth doing and on what you need to do them and so on, and at the same time, very poor in the laboratory. I, and I think Javan probably had similar talents and shortcomings as a laboratory working person. But nevertheless very good, a top-notch physicist. So he prevailed on us to hire Bill Bennett who was about to complete his PhD requirements in the Columbia physics department. Bill Bennett was not a Townes' student, but nevertheless, Javan and Bennett got to know each other; Bill was known as an excellent laboratory experimenter and Javan very much wanted his experiment to succeed. So he prevailed on us to hire Bill Bennett. As a matter of fact, Bell Labs standards for hiring were quite high and we had a pretty good account from the chairman of the department as to who are the very best people and I initially, as a recruiter, did not pick up on Bennett. But when Ali Javan was making good progress in his calculations on the He-Ne scheme for a laser and wanted an experimental high-level collaborator, and Bill Bennett was of course top-notch, in that respect, he prevailed on us to offer him an appointment.

We arranged to give Bennett a postdoctoral position which was at that time called a limited-term appointment, so Bill Bennett was hired [to help insure that they find the right conditions for the laser to be successful, i.e., the laser that he proposed and published]. Later on Bennett achieved regular status, anyhow, because of his accomplishments; he was quite a good man, too. I'll go a little further on this. This is why my memories are strong on this. Javan built up, initially, quite a laboratory, not for laser work. But then when his interest shifted he wanted to buy more apparatus. He wanted to buy a multichannel analyzer, which was rather expensive, and in fact the money for plant items was practically exhausted in the Physical Research laboratory at that time. Well, we prevailed on our upper management to borrow from this department, or that department, in order to equip Javan with a multichannel analyzer, which I remember cost $26,000.

Bromberg:

That means going up to Kompfner wouldn't it? Or was it Cutler?

Millman:

No. Addison White was my boss. A.H. White was at that time the Executive Director of Research, Physical Sciences. Later on, the Physical Sciences Division split in two parts, Materials and Physics, and I became an Executive Director of the Physics Division. But that was not until 1965. In 1958, 1959 and 1960 I was Director of Physical Research and my immediate boss was Mr. A.H. White and then the Vice President of Research was Mr. W.O. Baker. Well, I prevailed on my upper management to give us a little beyond what we would normally get for our laboratory. Anyway, the story I would like to tell sort of bears on it. Several years later, both Javan and Maiman got a Hertz Foundation prize for their laser work. Hertz Foundation was quite a prestigious prize, it was $10,000 each, and that prize was presented at a dinner at some hotel in Washington about the time of the Physical Society meeting. Anyway, we were in Washington and, obviously, I was invited to dinner. There was a presentation by the Hertz Foundation. I've forgotten who made the presentation, it could have been Townes. Anyway, they awarded the prizes to Maiman and to Javan and both made brief acceptance speeches. Javan came up and gave the usual thank-you remark and added, "I want to take this occasion to thank the Bell Laboratories management," and looked in the direction of the table where my wife and I were sitting, "for the support of this research. In particular when I decided to work on the laser they didn't ask me what I was going to do with my room-full of apparatus that I had and didn't use, magnets and all that."

Bromberg:

Now, can you tell me when you as a Director probably would go to some Council of Administration, I don't know what the situation is, is there some conversation about, well, there's some work going on the laser, and what was the attitude at this very early time or what were the potentialities you saw or didn't see? Do you have any recollection of how it looked to the top management?

Millman:

Well, as a matter of fact, yes; first of all we had great confidence in Javan. Because of the very strong recommendation from Professor Townes. In addition, since that did require a little bit apparatus money which we didn't have in our Physical Research budget at that time, I did consult Mr. A. M. Clogston, who was the head of another department in Physical Research, and quite knowledgeable in matters of theory, a little stronger let's say than Geballe, although Geballe is also a first-class physicist. So I asked Clogston, in preparation for asking for more support for what looked like an obvious good experiment, to look into the theoretical studies that Javan carried out, the quantum mechanical calculation about population inversion, lifetime and so on. Clogston's report was that it's a good idea and it's worth supporting." On that basis, we said we're on to something big and Javan got full support whenever he wanted, including our getting Bennett for him.

Bromberg:

I see. So really it was quite clear that this was something really important.

Millman:

Quite promising, that's right. Both the optical maser idea, as described in the Schawlow-Townes paper and Javan's idea of a helium-neon laser. There was no question about it.

Bromberg:

That's interesting, because I have the impression, which of course needs more checking, that the situation now is quite different from the maser. When the maser was proposed in 1951, I don't have the impression that it was as evidently an interesting thing. But the laser as you describe it seems to have been from the very beginning something of great interest.

Millman:

Well, let me give you a reason why. I personally was involved during the war and afterwards in this matter of coherent radiation as we use those words, in the region of radio, and microwave radar. Even before the war I was working with molecular beams and had occasion to use oscillators for molecular beam magnetic resonance. When the war started I got involved in microwave magnetrons and we learned to appreciate how hard it is to go to higher frequencies. Higher frequencies were important from two different aspects. From the point of view of radar, because with higher frequencies, of course, as you know, come shorter wavelengths, and shorter wavelengths gives you better resolution. From the point of view of the electrical engineer working on communications, the interest is focused on higher frequencies and the greater bandwidth, although wavelength and frequency are inversely related, with the product equal to the velocity of light. When you transmit voice you need something like a bandwidth of the order of 5,000 hertz.

Of course, if you want to transmit high fidelity music you may want 20,000 hertz. And if you go to television you need not thousandths of hertz but millions of hertz, e.g., five megahertz, and if you are going to have the same carrier transmit a lot of channels, you need more frequency bands. So the radio engineers, in particular Bell Laboratories scientists, are particularly interested in the high frequency end of things. Now the maser that Charlie Townes and his students, Gordon and Zieger invented, didn't boost the frequency, it was an ammonia maser which is in the frequency region that we did work during the war, what we called K band. In fact some of our magnetrons we developed went down even to shorter wave lengths. So there was no breakthrough in a sense of "here is something that will go to shorter wavelengths." It was true, that it became possible to obtain narrower resonance lines and it was very interesting for spectroscopic studies. There was a full appreciation of the scientific value of the maser and what's more, as many of our people felt, it had great potentiality for low noise, which is important in communication. In fact, it was used at Bell Laboratories for a low noise amplifier in Telstar. The actual amplifier used was not the precise suggestion of Townes, but more along the lines recommended by Professor Nicolaas Bloembergen of Harvard. You'll get it from talking to Scovil. Nickel sulfate was the material which gave you this low noise maser. In the last analysis, the maser was an excellent idea, particularly the idea of utilizing stimulated emission.

Although stimulated emission in physics is not that unusual but nevertheless, Townes was able to do it in an orderly way. And it gave you very narrow lines. It was good for spectroscopy. But we did have coherent radiation, we did have vacuum tubes, we did have magnetrons, we did have klystrons. So, there wasn't anything that unusual as a source of coherent radiation. Now came the Schawlow-Townes' paper. It talked about extending the frequencies, not just by a factor of three, as we were used to [do] and generally found it quite hard work, to go from X band to K band. I was involved in these matters during the war. We were already in X band, which is about ten thousand megahertz or 10 gigahertz, and our charter was to go to K band, which is 24,000 megahertz, not quite a factor of 3, and later on we went even to higher frequencies than K band. So everybody was interested in one way or another in shorter wavelengths or higher frequencies. That was the general trend in those years, after the war in 1945, 1946, 1947, how to go to higher frequencies, shorter wavelengths. And now came Schawlow and Townes, and in one fell swoop jumped to optical frequencies. We are now talking about frequencies not 10 but 10, that is 4 orders of magnitude higher. Now of course our people weren't particularly interested or didn't envisage the use of the laser for detached retina applications or cutting diamonds; those are a by-product. They had their eyes on communications, that is the boosting of the communication frequencies by a factor of 10,000. You see what's happening now with light-wave communication.

The real interest started when Schawlow and Townes came through with their paper saying here is the possibility, although they didn't propose any specific system which worked. The only part of the system that they proposed which we find in all lasers, is the resonator circuit of place parallel mirrors which is very important, I don't want to minimize that. But the specific gain medium didn't pan out. In fact, relatively speaking, Schawlow himself at Bell Laboratories wasn't that successful in coming up with a specific design for a working laser. At Bell Labs it was not Schawlow's idea, but Javan's proposal, the He-Ne laser, which worked first. The Schawlow and Townes' main idea was there of course, the Bible as I call it. Nevertheless, there was a very strong appreciation by me and people who worked with high frequencies [and knew the] tough time in getting things to go to higher frequencies that the optical maser idea was a great discovery. Often people ask me about Nobel prizes and so on. My feeling was, just pure conjecture of course, that the maser by itself probably would not have yielded a Nobel prize because after all we had those frequencies, they didn't boost them. Now with the optical maser, they showed the way of going to much higher frequencies, by a factor of 10. That's the great invention.

Bromberg:

There is a period in here when Javan's working with Bennett and also Herriott.

Millman:

Herriott came from a different background. He came with experience in optics and lenses, and as you may remember, that the laser required mirrors and so they teamed up together. Their two mirrors, were inserted in the vacuum envelope not outside; they didn't have the Brewster angles and so on. Javan was the principal contributor to the physics of population inversion, energy transfer from helium to neon and so on. Bennett is the one who did all the good physics experiments, to measure whatever they wanted to measure, time constants cross-sections and so on. On the other hand, Herriott contributed mostly to the mechanical design of inserting the mirrors in the vacuum envelope with provisions for lining them up from the outside and adjust them precisely parallel.

Bromberg:

But then there was a time when other things were going on beside Javan. Robert Collins, Donald Nelson, Wolfang Kaiser, and Geofrey Garrett. So here you were and Javan was working on it but all these other people were too, and Javan had not yet succeeded so one didn't really know what was going to happen. I wonder what that period was like. And then suddenly Maiman comes out with his. Do you have any memory of following what these other people were doing or that whole period in time?

Millman:

First of all, some of these people were not in my physical research laboratory.

Bromberg:

I had raised the point about this intermediate period, when Maiman's laser was announced and Javan's was not yet going, and you told me that a lot of this work was being done outside the Physical Research laboratory.

Millman:

I think we were generally aware of what was going on, and everybody applauded. Of course, they were a little stung by the fact that here was Maiman coming out first, and making a ruby laser work, and so on.

Bromberg:

Did that sort of get to you, as the person who was so actively sponsoring the Javan work?

Millman:

Well, one has to recognize the fact that Bell Labs can't do everything. Naturally there was a little disappointment. I'm sure the very first person to be disappointed was Art Schawlow himself. In general, there was a lot of excitement, a lot of people working, seminars, etc. And of course, the immediate accomplishment, I remember, was the actual getting the helium-neon laser to work, at Christmas time of 1960. Maiman's laser came out in mid-1960 and Javan's was at the end of 1960. I remember, he was scheduled to present a paper on February 4, 1961, the last day of the annual meeting of the American Physical Society. There was a heavy snow and we couldn't go to hear it. My wife and I also had tickets to the Metropolitan Opera and didn't get to hear the opera either. I did go to the meetings of the previous days, but not on that Saturday when the paper was to be given. But that's all right. I remember that very vividly.

Bromberg:

Well, that at least fixes the day when.... So then you must have seen it in the lab, though, around Christmas.

Millman:

Oh yes, well, Christmas, we all saw it worked, with great pleasure. I also heard Javan rehearse his upcoming talk. It was a period of great excitement. Here were the predictions of Schawlow and Townes coming to fruition.

Bromberg:

Then did you move to strengthen that research activity at all, or that wasn't the kind of thing that was done?

Millman:

I don't think it was necessary, because there was some spontaneous activity going on all over Bell Laboratories, materials and other ways and those who were interested got involved in it, worked on something. I don't think you had to go and organize an effort on it. Does Giordmaine figure at all in you history?

Bromberg:

Giordmaine came here during the summer of 1961.

Millman:

Right. I remember I had a discussion with Giordmaine. I remember that vividly too. Giordmaine of course was also a Townes student and also very highly recommended, and he asked me which department he should join, department 111 where Schawlow was and Javan was, where this laser activity first got started, and then there was considerable activity in an adjacent department, department 115, where there was more research work on actual working lasers, sort of slightly more with an applied flavor: How do you get a laser with that kind of property or this kind of property, a different laser, a different wavelength, and so on and so forth? Giordmaine, as I mentioned earlier, was highly recommended by Townes, and so essentially I gave him a free hand to join either one. I essentially asked him, what his interest was. If he's coming here because he wants to do some good physics, then you know I think the obvious place for him is in 111. If one the other hand, you think, now that you know what's happening with Javan and Schawlow-Townes and so on, you would really want to work on lasers, (or we called them optical masers at that time). My suggestion was to join the other laboratory which was not under my direction, not physical research but solid state electronics research, 115 because this is where most of the laser activity was. And he decided that he wanted to be full time on lasers, rather than think more generally about physics research. So he joined the other department.

Bromberg:

OK, now, going to one of the questions I asked before we started taping, I'm very interested in what I'd call the prehistory of quantum electronics, if you have any ideas about it, and I'll formulate my question—you may even want to respond by telling me that I'm not formulating the question right, which is also a good response. Here you have a situation in which quantum resonances are being pressed into the use of technology, and that comes in a number of ways, masers and atomic clocks, you know. Even when Townes was here [in the mid-1940s] he had an idea for other ways in which you could use quantum resonances as filters or for frequency stabilization, you know, in devices of this sort. And then as I said, Ramsey mentions that early on, there was even a little talk at Columbia about using molecular beam resonances in some way or other. I would have to check back.

Millman:

He mentioned the cesium clock?

Bromberg:

No, I think it had to do with magnetic field standards.

Millman:

Zacharias had an idea, as I remember, about using a cesium beam—well, there was a cesium atomic spectral line used for frequency standards in atomic clocks at the Bureau of Standards. Rabi once gave a paper on this, his Rydberg lecture. At any rate, in order to make the cesium line narrow, the slower the molecules move, the longer they spend in the radio-frequency resonance field, and you get a narrower line, and therefore, the frequency is known with greater precision. For a frequency standard this is very important.

I remember, Zacharias had a scheme for shooting this cesium beam up, and if the cesium beam starts with a Maxwellian velocity distribution, and moves against gravity, some atoms will get to the top, and the very slowest one will come down, so that if you get your radiation field right at the very top where the molecules move very slowly, that would produce an unusually narrow line for a frequency standard. I think he came to Bell Laboratories to peddle this as a possible frequency standard but it was not bought. First of all, we in the Physical Research Laboratory were not directly interested in this because we weren't involved in devices, but there were other Bell Labs organizations that were interested in frequency standards and clocks. They have to synchronize time for transmission of information from one city to another. Timing, better than one part in 10(superscript 9, leave space), is relatively important; therefore there was a possible interest. And as I remember Zacharias came and talked to those people. I got them together because I knew Zacharias and so on, but the interest wasn't pushed that hard. Maybe they didn't think much of the scheme. I don't think the scheme was ever used anyhow.

Bromberg:

I have a feeling talking to you that my analysis doesn't represent the way people were thinking at that time. I mean, I'm thinking in terms of quantum resonances, you know, all the possible uses you can make of quantum resonances, but the way you recall it makes me think nobody then was thinking in those terms. They were thinking of either creating shorter wavelengths or making a more accurate clock, but not thinking in terms of well, wouldn't it be nice if we could press quantum resonances into the service of technology.

Millman:

I don't think there was that much effort, certainly not in Physical Research. I mean, people were interested in various forms of lasers, and of course ultimately, I'm sure you'll get to this sooner or later, there is this very simple laser one uses, the semiconductor laser, which was invented very early not only at Bell Laboratories but at several other laboratories: IBM, Lincoln Labs and so on, and of course it wasn't very interesting because it required large threshold currents. In 1968 or thereabouts Parish and Hayashi at Bell Labs got the heterostructure laser, so that was when the great interest developed in using lasers for optical communication.