Notice: We are in the process of migrating Oral History Interview metadata to this new version of our website.
During this migration, the following fields associated with interviews may be incomplete: Institutions, Additional Persons, and Subjects. Our Browse Subjects feature is also affected by this migration.
Please contact [email protected] with any feedback.
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 is available.
In footnotes or endnotes please cite AIP interviews like this:
Interview of Rolf Landauer by Joan Bromberg on 1984 October 17, Niels Bohr Library & Archives, American Institute of Physics, College Park, MD USA, www.aip.org/history-programs/niels-bohr-library/oral-histories/4726
For multiple citations, "AIP" is the preferred abbreviation for the location.
The microwave spectroscopy group under William V. Smith in the late 1950s. Research directions and management attitudes in the early years of the IBM T. J. Watson Research Center. The Sorokin-Stevenson 4-level masers. The invention of the semiconductor laser in 1962. Research on optical parametric oscillators. Also prominently mentioned are: Bill Dumke, William S. Franklin, Seymour Keller, James Arthur Krumhansl, Gordon Lasher, Marshall I. Nathan, Arthur Leonard Schawlow, W. V. Smith, Peter P. Sorokin, Mirek Stevenson, Charles Hard Townes, John Von Neumann, Weinrich; Bell Telephone Laboratories, and International Business Machines Corporation.
...I am with Dr. Rolf Landauer at the Yorktown Heights IBM Thomas Watson Research Center. Shall we talk first about an overview, of laser R and D as you saw it? IBM was not at all involved in the laser, is that right?
No. I think that’s a fair statement. I think that our interest in anything that could be explicitly related to the subject started with Bill Smith, W.V. Smith, coming into IBM, I’m not sure I remember the exact year but probably ‘56 or ‘57, and organizing a crew which may have had the little microwave spectroscopy. Anyway, it was a center of gravity, and that’s where we first, in a sense, developed a non-trivial interest in anything related to that question. Indeed, paramagnetic resonance and things like that, took up much of the energy of that group.
Now, where was this physically?
Physically it was in Poughkeepsie. It was in Poughkeepsie where there was IBM Research. When you say IBM Research, it’s very much a matter of definition, what you define, but certainly to a marked degree the antecedents of our current research organization existed in Poughkeepsie, about 1956. In expectation that this would be a large and growing organization, 1955-56, Manny Piore, E.R. Piore, was brought in to head the newly forming research organization, so obviously some who had been there before had at an earlier time joined activities which very much had the character of this developing research organization.
Were you in this group that Smith was in?
No. No. Again, you stress my exact memory of the relative sequence of events. I took on my first management position at IBM in ‘57, and I believe when I took that on, there were four groups that came under me, and one of them was Bill Smith’s group which I believe had just barely shortly before that been formed. I’m certainly not the one who brought Bill Smith into IBM and started that group. I did help, I was a recruiter at Harvard at that time, and was very much involved in bringing Peter Sorokin to Poughkeepsie to be interviewed as an applicant.
And he was destined for that Smith group?
He was interviewed by Bill Smith, and I think, as he said on many occasions, you may be able to get that directly from him, he came to IBM basically because he sensed Bill Smith’s tremendous interest in him. Of course, Peter came to us highly recommended, he was one of Bloembergen’s students. For many years we recruited in the group many Bloembergen students. John Armstrong is one, Jim Wynne is one. And certainly the question was in the air at that time: can you make an optical laser: I remember many discussions at that point. I don’t believe that when Peter first joined us, that was specifically his interest.
Surely if you’ve got this group going doing microwave spectroscopy, IBM is at that point thinking of other things besides lasers?
Well, to a certain extent it was a service group. Paramagnetic resonance was essential to understanding all kinds of solid state effects, and to a certain extent of course we were a broadly based industrial research laboratory in the process of formation. We had in front of us the models of Bell Laboratories, Philips Laboratory in the Netherlands, at the time, organizations that had a broad spread of interests, and we were essentially following that model.
Partly bringing in groups that they had?
Partly diversifying somewhat beyond, I would say, the most obvious — semiconductors were obviously of interest to IBM, and getting materials for storage was obviously of interest to IBM, but partly I think intentionally diversifying somewhat beyond that.
I guess what I’m really saying is, if you can at all go back to before the laser and try to remember, how that group fit into the whole scheme that you were managing? And what its function was, what you hoped from it, how large you wanted it to be, what kind of resources, that kind thing, just to try to get back before the laser and then through the transition?
I think it’s hard to give you fairly crisp answers, and certainly hard to give crisp answers which really have a guaranteed validity about exactly what was in our minds, exactly what were our aims at the time. This was a time, remember — let me bring out something else — while our interest in the laser may have started as a response to the world’s interest as a whole in that structure, and the obvious question: can you make it work in the visible light range? Can you make stimulated emission work in the optical range? Was an obvious question. Everyone was sort of interested and concerned and aware of it. But let me bring you to another reason for interest in microwaves we had, which relates also to this whole question of interest in nonlinear electromagnetic processes, non-linear circuits, processes of parametric amplification — (John) von Neumann had come (to IBM), probably in 1954. He has a patent on how to use parametric excitation for accomplishing what you want to do in a computer, via generation of subharmonics. He’d come and said, “Hey, I know how to use nonlinear circuit effects, generating from a microwave source of power, a signal of half the frequency which has one of two possible phases. “Bistable devices, things which can go in one of two possible ways, that’s what computers are made out of.” I know how to make these things act and accomplish the things you need to do in a computer, and here’s my patent,” and he sold it to IBM. Independently just about the same time, that same invention was made in Japan by Goto. That patent developed over a period of a few years into a major development project at IBM, and also fairly large projects at other laboratories, particularly at RCA in this country, It was also followed up in Japan — in fact, it was followed up in Japan further and more successfully, in the sense that in Japan people made actual working computers out of this effect. Paramentron computer 1 and 2, which actually were built in Tokyo in the late fifties and worked. So that was another source for us to be very much interested in microwave radiation and effects, interesting things you can do with it in the way of signal handling.
And probably it was a partial motivation for bringing in somebody like Bill Smith who had a strong interest in, what’s some of the detailed physics, how do I make microwaves interact with spins, etc.
Now, it’s Smith I understand who began the laser activity. I’d like to know what happened as the laser activity began. Who was talking to whom, what were various people thinking, who was for and who was against it, if you remember any of that?
The simplest part to answer is, who was against. Nobody. All right? I think there was no opposition. There was no objection. There was obvious uncertainty on the part of all — could such a thing be made? How it could be made, how well it would work, would it be useful all those questions? I think there was nobody against it. There was no objection. I think you have to remember that in those days, they were exploratory days. I think in a sense we were probably a little less critical of what we undertook than we might today have been, in a sense. One had to learn what kinds of things would be successful and what would not.
That has to do with the newness of the operations?
The newness at IBM, the newness in learning what kind of science makes technology move and how you couple the two. Of course, in a sense, today in this laboratory, we’re very free. You find people with respectable activities, and you find people at the extreme who are doing things that no one knows whether they will ever have any relationship to computers. We have people who worry about formation of galaxies and astrophysics. Not very many but we have a few. We have neurophysiologists here. Well, you can argue that maybe you’ll learn something eventually for computers out of how nerve pulses interact, but it’s not that likely. Nevertheless, before we take something very seriously these days, I think we’ll probably have to be more critical perhaps and more searching and careful than we were then.
By the way, I’m assuming that these were reactions to the Schawlow-Townes paper. Is that a correct assumption?
Certainly the general question, can you make a maser work at much higher frequencies? I think that question has an origin independent of the Schawlow-Townes paper.
You’re now asking me a question which, since I’m not of these laser inventors, you stress my memory. I suspect it only got taken very seriously after the Schawlow-Townes patent or paper appeared or the equivalent maybe in a conference paper. Can’t be sure of that. I don’t remember the exact sequence of events. I’m not one of the inventors and therefore that formative history is not that clearly in my mind.
OK. But just anything you can recall about —
— but you know, we all understood that, you know, you look at the formulas for the ration of spontaneous emission to stimulated emission, there were frequency cubed in there and you say, “My God, how could that possible work at optical frequencies? Can that work in the optical range?” It was unclear. Now, again, whether Bill Smith decided, “Hey, we want to go after this” and went to Peter Sorokin and said, “What do you think about that?” or it worked in the other direction. I’m not really qualified to say that. Certainly not as of 1984 any more, even if I did know, OK? But I think it was in the air, to a certain extent, and I think there were no objections. I think there was no viscosity in the process, no difficulty in getting going. If there were any difficulties, it was just to develop conviction that it was really worth doing, and a great many difficulties in detail again which Peter can tell you all about, like getting his crystals and getting’ his reflecting surgaces made and all these things, which we weren’t all that well equipped to do. Being a new and developing laboratory I think in a sense had its influence.
The other early laser inventions that were made elsewhere, were all more common subjects of discussion. There was the ruby laser. Everyone in the quantum electronics business, all these people were very much aware of ruby, and there may have been some debate, Schawlow you know for a long time indicated that the transition that finally made the ruby laser work wouldn’t work. There may have been some debate about it, but that system was on everybody’s mind. Also the gas laser, again, perhaps not in the final exact way it was observed to work, but the general possibility had been discussed already by Fabricant in Russia a long time before them, and was a subject of discussion at conferences. I think in a sense, because we were perhaps a new and less sophisticated environment, in a sense I think Peter Sorokin was thrown on his own knowledge and search of the literature, to think through the possible spectroscopy of materials, and really in a system which people hadn’t discussed and hadn’t appreciated.
By the way, I have to be careful here when I refer to Peter Sorokin because he is the one who is here today, in continuing in the field. I don’t think it’s fair to characterize him necessarily as the senior partner. Peter Sorokin and Mirek Stevenson were a team. My judgment is that Peter was the idea man, the tone who understood the theory and the possibilities better, Mireck was more the implementor who got things done on time and figured out how to do them. But they worked together, eventually came up with the first four level lasers, second and third lasers altogether, first laser that pumped at energies at that time low compared to what the ruby took. In the accomplishments leading to that, if we disregard subsequent history, I’m not sure one can say, “hey, one of those two was the senior individual.”
Now, one story which I’m rather anxious to get and I think this is a story that you’re particularly able to give me is, as these lasers come into existence and as real honest to God lasers are beginning to be built elsewhere also, whether there’s any expansion of the effort that takes place, whether the applications people get interested, the whole story of trying to probe the possibility of using some sort of optical logic. From where you stood, which was sort of in between the research team and management, I think it would be very profitable to get as much as you remember. How the laser seemed to fit into IBM or didn’t, IBM’s needs, you know — this kind of part.
We certainly asked ourselves questions about that all the time, and I believe there is an assessment. Maybe the first assessment was made by Bill Smith before the invention of the laser. I think he asked himself probably at that time about the maser, but I’m not sure what the system was, “Hey, can we do logic with that?” Just asking about energies and time constants. These assessments were made all the time. Certainly my impression is that at the time of the invention of the ruby laser, even at the time of the invention of the four level laser, in no sense simply because it was such a complex expensive system, huge energy requirements — lasers originally were not continuously operating — in no sense did we think of them at that point as, to take the extreme, as something that would say replace the tube or the transistor as the work horse in a computer. That was totally out of the question, I think, in our minds. And of course, the complexity of the system and the search for a much simpler laser was one of our main reasons for being interested in an injection laser, a semiconductor laser, something that wouldn’t be a much more complicated structure than a transistor. Now, if you back away from that and ask, what applications did we see for the laser other than that, say for the earlier lasers? I’ll have to give that some thought after the conversation and see if I can think of something sensible to answer that. At the moment, I doubt if we really had anything in mind above sorts of things that everybody talked about. From the very beginning, people talked about optical communications, using lasers as a communications link. It was many years before anybody had the least idea of how to go about doing that. In fact, I remember a visit that some of the management people here in this laboratory paid to Bell Laboratories in the early sixties, convinced that they knew more than they’d said about how you were going to use this. Everybody talked about this tremendous optical band width, at the same time we knew perfectly well, (we understood electrical optical modulations and things like that), that nobody knew how to do it. I remember going down to Bell Laboratories. I suspect it was a visit related to our patent cross-licensing negotiations, convinced that they must have a deep level of insight they hadn’t published. I came away from that visit convinced they had nothing of the sort. Either that or they were awfully clever about hiding it. I was pretty convinced that they didn’t have it. So that took a long time. And of course, again, it was much of the reason for our being interested in the injection laser. Lasers at IBM have been put to use basically as scanners in input-output devices, to impact the electrophotographic drum in printers, as scanners in the supermarket — but did we understand any of those possibilities at the time? I’m not sure, but I’m inclined to doubt it. I think that all came a little later. Remember, the early lasers required a lot of power, they operated intermittently, they were expensive, and I think it took a few years to start talking about those things.
I guess what I’m a little curious about, at some point there was research actually carried out on optical logic. Now, I got these bibliographies very late and I’ve read almost nothing that was on them, but I know that there is a paper that John Armstrong wrote with Shiren on how much energy —
— with Keyes, Bob Keyes.
That much later. No, the interest in doing optical logic — now, I must say that there’s one paper that somebody pointed out to me recently, I think it was just an abstract for an oral paper — one paper, I remember Eli Snitzer was one of the three co-authors, when he was at American Optical still. I think it was a neodymium... — I don’t know which neodymium lasers existed at that time — but an early observation of bistability in a laser system. But certainly, these things got taken very seriously here at IBM, only after the invention of the injection laser — immediately, almost, after the invention of the injection laser. Then we had something very small. We know we can turn it on and off very quickly. Looked like an ideal tool for modulating, if you want to transmit a communication — of course the low-loss optical fiber hadn’t been invented yet. People played around with injection lasers as a method for communicating for many years at IBM and elsewhere, and really didn’t get very far until the development of the low loss optical fiber. And certainly people at IBM in the sixties played with that, on a number of occasions in a number of different locations, using the laser as a communications method. Using laser as an actual tool for doing what you do in a computer, for doing logic, that is, having two signals interact, and producing an output which depends upon both signals, or having a bistable system, a system which can be in one of two state, having it sit there. It may have been in a dim way a partial motivation for our interest in injection lasers, but certainly any detailed thinking came after the injection laser. It came relatively quickly. And then a number of people here, — I think those names are listed in the things I’ve sent you already — including Fowler and Lasher, also including in one effort Marshall Nathan, people played around with these things, and really very quickly we learned that would be a very difficult way; that optical schemes had a number of shortcomings, that transistor circuits driving other transistor circuits don’t have. We clearly lost interest within a few years after the invention of the injection laser.
Organizationally speaking, you’ve got a group of scientists and you’re manager of these groups — now, do these groups have the mandate that you should just go and work mostly on lasers, or do they work on whatever they want to, and somehow here in the organization, there some place else —
Well, I think it varied. Let me recollect a little bit of history. I was a manager of the group which included Bill Smith’s microwave effort from 1957 to 1958, for a year. At that point I decided, hey, I really didn’t want to spend my time being a manager, I wanted to go back to doing theoretical physics. Then for a few years, I was leader of a small theoretical group, which included Gordon Lasher, whose name keeps coming up. So I wasn’t part of the management team all the time, and particularly I was not part of the management team at the Sorokin-Stevenson four level laser, calcium-fluoride-samarium, calcium-fluoride-uranium lasers. I had been in very close contact with that group, and had a number of discussions with Peter Sorokin about, oh, the nature of modes in a laser and how they determined — but it wasn’t a serious collaboration. I was very interested in it, listened to it, everybody was. I wasn’t a serious collaborator of Peter’s. And I got back into the management chain in, probably, about early ‘61. Early ‘61 I resumed a more serious management role than just managing a very small group of theoreticians, and was put in charge, spring of ‘61, of our semiconductor physics group and another group, I don’t remember what its title was but it certainly was a group that was a laser group, included people who had just discovered the four-level laser, included some other things like microwave acoustics and remnants of our interest in microwave spectroscopy. I probably interrupted your question with that bit of additional history.
No, my question was really about getting some sense of how things are organized, so, for example, if Lasher is working on theory of semiconductor lasers at one point, or Nathan is working on semiconductor laser experiments, or somebody else is working on bistability — just to get a feeling for how these things are organized here, whether it’s just spontaneously, the topic arises?
Well, I think it varies somewhat from group to group, and individual to individual. A large laboratory like this, of course, has some efforts where we decide, there is this technology emerging and IBM had better have it. And then it cannot depend upon people’s spontaneous instinct. When you try to, say, demonstrate to IBM and eventually to the world — let’s take a case we’ve just given up on, the Josephson junction. The Josephson junction is going to be the tool for computer — well, you have to tell somebody, hey, you’re going to work on power supplies, you’re going to work on cooling, you’re going to study deterioration of junctions, you develop the logic circuit, you develop the memory circuit — you have to make assignments like that, and plan how things come together. In our more basic and freewheeling activities, it’s not like that, and there tends to be a subtle graduation. Now, I would say, in our more fundamental efforts, like we’re talking about, semiconductor physics and what started as a microwave spectroscopy group and became more and more a laser group. I would say, there was some give and take. Some people were left relatively alone.
Occasionally we’d nudge or push people. As I mentioned of course Peter Sorokin had started working on lasers before the invention of the ruby laser, before anybody knew a laser could be made, or Peter Sorokin and Stevenson, I should correct myself. Of course the demonstration of the ruby laser by Naiman gave that work tremendous impetus, and obviously everybody asked, hey, is ruby a unique system? Are there a lot of lasers? And then, when the calcium fluoride lasers were discovered here people said, hey, ruby isn’t the only system, it’s not a remarkable freak, what other forms can lasers take? This is about the time when people in the world on the whole began to start thinking about semiconductor lasers, about chemical lasers, etc., other than optically pumped lasers. I don’t quite know when we first started asking that question. Certainly we took it seriously, a result of what turned out to be an erroneous rumor. (Pierre) Aigrain was coming to this country in early ‘61. He had had a longstanding interest in the injection laser and had advocated that, and he was coming to talk to some meeting, I think it was a spectroscopy instrumentation conference in Pittsburgh but I could be wrong about that, in early ‘61.
The incorrect rumor was that he had an injection laser in his pocket to show. Now, of course, that triggered us, we though, now, if he’s got it, how can we make that, how could it have been done? That got us thinking very seriously. We developed more and more conviction, first informally, that such a thing could be done. The conviction toward the end of ‘61 became strong enough so I felt, we’ve got to organize a little more and do things systematically. I convened a study group at that time, and the study group — I’m not sure I remember the exact composition — it almost certainly included Marshall Nathan, Gerry Burns and Gordon Lasher, possibly one or two others. And certainly I dragged Gordon into this study group, perhaps not kicking and screaming, I think that would be the wrong implication, but certainly he was nudged toward that area, and out of that came his concern, and his eventual understanding that 3-5 compounds with their indirect transitions were much more promising materials than germanium and silicon. That understanding had been reach somewhat independently in this laboratory by Bill Dumke, who did not report to me. He was part of a different group. Bill, however, was very much oriented toward optically pumped gallium arsenide, and not particularly, in the other group, oriented to the question of pumping the laser by electrical injection. So there is some nudging. In one of the notes I gave you; you’ll find that at one point, Jim Krumhansl of Cornell, a consultant to us, came and listened to us and said, “Well,” I forgot the exact nature of the question, “what exactly would you expect the threshold for lasing to be? How much current do you have to feed into a gallium arsenide diode:” And made us realize that we haven’t exactly answered a question just like that specifically, and I said to Gordon Lasher, “Hey, we ought to know enough to be able to answer that in almost no time at all.” And he did, and we realized the thresholds weren’t very impressive, you know, they wouldn’t be hard to reach. Perhaps we hade injection lasers without knowing it already in our possession.
So there was some nudging involved — the nudging that you do as manager, the nudging that you get from knowing that other people are interested in the same thing, and the natural, I would say, intellectual curiosity of people. They sort of merge. I don’t think it was any terribly constrained situation. I don’t think at that point we dragged anybody into this effort, you know, very much against their opposition, as occasionally you do. A number of years later, I started the work in this building, organized the work on large scale integration. In that case, very much more than in the injection laser case, yes, we had to take physicists who wanted to work on other things, say, or scientists and say, “Hey, this is more important to us, do that.” I think in the early laser days, both the calcium fluoride laser and the injection laser, there was little or none of that. There may have been some nudging, yes. I remember, too, just a few weeks after this conversation, involving Jim Krumhansl and Gordon Lasher, I remember some conversation — I remember it in the evening because my wife remembers it and she reminds me of it, because it took place in the evening — I simply said to Marshall (Nathan), “Why don’t you do it?” I don’t remember exactly what “it” was but remember saying, “Hey, get moving.” So there was some nudging, but really not very much.
Good, that’s interesting. Now, on the other side of this, that is, in terms of your —
The main nudging, by the way, was management. Management, representing me, to a lesser extent my superiors, to a certain extent other managers who reported to me — management was very interested in this goal. All right?
That’s what I was going to ask you.
Management saying, “Hey, this is terribly important to us” had a strong influence? When management says, it’s terribly important to us, and the outside world is also interested, it becomes an overpowering influence in a laboratory like this.
Now, let me go back — when you said management was terribly interested in it, does that mean, when you talked to Piore about it he said, “Gee, this is really interesting?”
No. I think management at Piore’s level probably didn’t share our conviction, probably said, “Well,” — now I’m interpreting, of course. “These people don’t look stupid to me. I think, I believe their reasons for wanting to do what they want to do, let them do it for a while and see what happens, all right?” I was never aware how often I was OK. Certainly I would say, if Piore had been against it he could have stopped it. He certainly wasn’t against it. I don’t think he gave us any particular intellectual support or encouragement. Management at a slightly lower level I think was a little more encouraging — as you come down, you become more positive.
Who was that management?
When I first started pushing the injection laser, when I was put in charge of the semiconductor and, I’ll call it, laser groups, my manager in turn was Andrew Eschenfelder, Andy Eschenfelder, now retired from IBM, and he had a job at that time which was comparable today to, oh, — perhaps a somewhat smaller version of what John Armstrong or, I don’t know whether you’ve met Praveen Chaudhari, who runs one of the other large groups in this building. He had one of the large groups in the building, all right. I would say again, the support at his level was more positive than Piore’s.
Eschenfelder must have been strongly interested.
Eschenfelder was interested. Again, he wasn’t experienced. He was about my age, perhaps a little older, but essentially my age — actually came to IBM a year or so after I did. I got my PhD probably three years before he did, but roughly my age, my contemporary, wasn’t experienced either as a manager, but it seemed a reasonable thing and the support was somewhat more positive in his case. Then at the beginning of 1962, I took Eschenfelder’s position. Eschenfelder became assistant director of research, second in command of the research division, a position that doesn’t exist today any more. But essentially the relationship continued on that basis. Certainly there was more than tolerance on his part, probably not a strong personal technical interest on his part.
Now, you said management was very interested but none of these people are all that interested. Were there other people who were interested?
Well, I was management at that time.
You were very interested!
I was management at that time. I was very interested. And people who reported to me who were also managers, Seymour Keller and Bob Keys, were very interested. OK? I was very interested. I was determined that this thing was going to come and by God, we weren’t going to miss it. In 1961, I was manager of these two groups. In 1962, I was manager of a large subgroup of this building at that point. So management is perhaps a euphemism for myself, to a certain extent, and the people in turn, some of the people who reported to me.
By the way, was anybody else at IBM interested at this point, like sales people, people out making devices? I’m assuming that they’re not, but is there anybody else who’s coming in to you from —
I very much doubt it. Once the laser was invented, of course, fairly quickly, we had visits from all kinds of people and from all parts of the IBMs technical organization, wanting to ask, wanting us to do something. Mostly, I would say very premature discussions, in terms of the time scale on which we knew how to do anything sensible with a laser, in terms of which laser would find applications, but yes, those conversations started right after the invention of the laser. I know there were quite a few of them, most of them leading essentially to nothing and not getting very far.
OK, if any of those pop up that really played a part in the story it would be nice to get them in but it sounds as if it’s just kind of a general buzz of —
— yes, I think certainly. A non-trivial percent of those conversations were with our Federal Systems Division. I’m not sure that’s its correct name today. We have a division which does most of our military contract work. They of course from the very beginning asked themselves questions about — probably as a reflection of discussions with the military — “Is this stuff any good for radars?” and so forth. And then a high percentage of our early conversation about applications of lasers were with people from the Federal Systems Division. They also at an early stage went into attempts to play around with laser communications — early stage must mean probably about middle sixties, not that early. Later, some interest in the Federal Systems Division in laser logic, laser memory, a rather fleeting interest, and very much representing the interest of one particular individual. You know, there were notions in our minds — for instance, I know, very shortly after the discovery of the injection laser, I asked myself, “hey, so we get enough power out of that gadget to make a visible dot on a piece of thermofax?” We did a quick and dirty experiment. We couldn’t get a dot on a piece of thermofax. So these questions were on our minds. The conversation in that direction didn’t get very disciplined and serious and fruitful for very many years.
We should also talk a little bit about the fact that you had this Fort Monmouth contract and you said a little bit about that at lunch.
Let me go over that again. We started talking to the people at Fort Monmouth. Frank Braud, who long since has left Fort Monmouth, was a key individual we talked to, and he was here in the middle of ‘61. My memory is, summer of ‘61. I forget whether we brought him here physically for that reason or whether there was some other reason for his visit. We exposed him to our developing ideas about injection lasers, that they were likely to come into existence, and that we wanted to take them seriously. We got him and his organization very much interested. They, in turn, developed enough interest to put out a Request for Proposal indicating contract money would be available. These things take a bit of time, of course, to get all the internal red tape straightened out. The proposal request appeared.
We naturally bid on it. I’ve given you a copy of our proposal. The proposal was submitted early January, 1962. And the proposal among other things includes, as an attachment, some of Gordon Lasher’s early theoretical calculations, pointing to the desirability of a 3-5 compound. At that time, and really almost until the injection laser was in existence, we had a twofold attack on the injection laser. We knew that, in a normal sort of semiconductor, germanium, silicon, gallium arsenide, and gallium arsenide was one of the more promising, you could make diodes out of those. You could get injection into it. But nobody at that time certainly in the early stages had every got very efficient light emission out of those structures. So we said, well, — we don’t know how to inject into things like the typical sort of phosphor material, zinc sulfide, cadmium sulfide. People get good light out of those. So it looked to us like a horse race, what could we solve more easily — to get the good light out of the typical semiconductor narrow band structures? Or to get injection, where it had never been achieved, in wide gap semiconductors, which were known to be easily emit light? We followed those two in parallel, and one turned out to be just about a total dead end. We didn’t really give up that prong until the gallium arsenide injection laser was just about achieved. Certainly towards the end, as we sensed, hey, if really looks like that ought to be possible, by everything we understand, perhaps, we dropped our alternative effort or diminished it a little bit before then.
Was it the same group that was doing both?
Well, different people. They were different people. We had had quite a few people interested in a 2-6 compound. There’s a history. Seymour Keller, who’s in this building still today and is certainly available, had been brought in to work on phosphors in the early fifties. He was a physical chemist who shortly turned to phosphors. At that time, phosphors as a possible storage medium, you know, you excite a phosphor, it takes some time for the light to decay. Everybody knew that you could move carriers around in a phosphor from one trap site to another. Maybe you could store information that way. There were all kinds of proposals running around like that. He was into that for a number of years. Then another suggestion came around. People were trying to make what today we call integrated circuitry.
How could we make a lot of things simultaneously and easily and cheaply? And one of the notions that came up was, you could make logic devices and do everything you needed to do in a computer out the combinations of light-emitting devices and photoconductors; the light-emitting device is very much like a relay, the current would control a photo-conductor, the photoconductor in turn would control the current supply to other light-emitting devices. The light-emitting device could be a neon bulb, it could be an electroluminescent device. That turned out to be a dead end. But as a result of that, we had a considerable background interest and ability to deal with 2-6 compounds. Of course, in all these other materials, there’d always been phosphors now we probably wanted to do something with a single crystal, but we had orientation toward that area. And so it was natural for us to follow that up. Didn’t get any place. Our vision was far from perfect, and that one didn’t go anywhere. Now, we had a lot of other ideas. I know in the early days, it’s probably in some of the literature I’ve given you if we said that we were going to get emission out of semiconductors, band to band emission, isn’t it important to have narrow lines? Very narrow lines had been observed in exciton recombination, particularly by some people at Bell Labs — maybe that’s the way to go. We took that possibility too seriously.
Instead of —
— oh, I don’t think we got any work done on it. It just started out, discussions of what to do, alternatives — we probably took it too seriously.
You know that this kind of information, on lasers that didn’t pan out, is of the greatest interest, because what you try to do when you reconstruct the history is, to put yourself back at that moment, what was the intellectual constellation like then? And if it’s a constellation where all these various leads are being pursued, that’s what you want your history to show.
Let me go back to before the injection laser and bring up one event in the local history which I want to say a few sentences about. There are always alternatives in people’s minds. We just discussed the injection laser and how the 2-6 compounds got us no place. There was an alternative to the laser which was very much on my mind in 1959. It was strictly a personal interest and it got no place at the time. I had been, as I have mentioned to you before, even since the von Neumann patent and the effort on parametric computing, been interested in parametric effects. In 1959, undoubtedly under the influence of people like Peter Sorokin and Stevenson starting to think about lasers and trying to figure out how they could make them work, I said, “Hey, can’t we generate light parametrically?” I sat down and asked myself, can I have traveling wave, optical parametric amplifier, which I pump say with a strong spectral line, say a sodium light source, can I get that to work? Then with a little thinking and calculation, I decided it was quite hopeless. Now, I came to that conclusion for reasons which were partly correct and partly incorrect. The correct part of my reasoning was, you know, to me it was an alternative for the laser. I didn’t allow myself to use the laser as a pump. Now, that of course turned out to be necessary. But I also made stupid mistakes. I thought I understand what the source of nonlinearities were. You need nonlinear effects to get parametric excitation, amplification. I drastically under underestimated the scale of nonlinear optical effects, and didn’t wake up until in — oh, I think it was ‘61 — that Frankin and Weinreich did Reir experiment.
— ‘61 –-
Showing nonlinear optical effects, and then I woke up with a start and said, “My God, I’ve drastically underestimated those effects, I’ve been stupid about it.” But it was on my mind very much, and again, one of those sort of things which turned out to be a dead end. OK, shall we go back to the injection laser?
After the Franken et al. paper, just to continue a little bit more on this business, did you not go back to that problem or people here under you didn’t go back to that problem at all, of parametric?
Yes, I went back to it immediately. Said, “My God, I made a mistake.” I wrote a little internal technical note. I can easily get you a copy of that.
That would be nice.
And said, “Here’s what I think should happen.” And I understand immediately, of course, in view of my background, that unlike Franken and Weinreich did it, the way to make these things work effectively is to make the pump, the source of energy, and the signal either being generated or amplified, make them travel together in the structure. That much I understood immediately. That was obvious to anyone who had thought about the microwave version of those effects. What I missed is the most obvious trick for keeping them together that came out of some other work of some other groups, Giordmaine at Bell Labs and I forgot which the other group is at the moment. It came out of two groups, I guess in late ‘61, who realized that the way you make these waves travel together is by using the dependence of the index of refraction and speed of propagation on the polarization of the light.
You directed my attention to a trio of papers by (Norman) Kroll, (Robert) Kingston and Aknmanov — it was in your bibliography.
By Norman Kroll?
Yes, Normal Kroll, Robert Kingston and, by the way the Kroll paper I couldn’t make head or tall of, the Kingston paper was very simple, and the Aknmanov –-
Yes. I remember the Kroll paper. I directed you to the others also?
Well, if you didn’t, then references that I got through –-
— OK, because they don’t ring a tremendously strong memory, you know. They don’t ring a bell at the moment.
OK, maybe I got it out of Kroll.
And they were all papers which said that there’s a microwave way of handling parametric amplification, and oscillation, that we ought to be able to now extend them, that Franken and all these people have done this so that I was wondering if there was a parallel thing that was going on here?
Well, I did the things. I triggered some interest in further things immediately on the part of Dick Garwin and Bill Hardy, Wilton A. Hardy, who in a totally different connection was mentioned at lunch by John Armstrong. (He talked about the “virtual laser,” a laser for the stimulated Roman effect.) Those two people were interested. The three of us filed a patent. I can easily get the patent and send it to you — unfortunately the patent was filed rather slowly. Our patent of course missed this key trick of using (phase-matching) so you can make widely different frequencies travel together and interact. You know, — materials have a dispersion, light at different frequencies doesn’t like to travel together. You have to have a special trick to make light at different frequencies stay together, and a trick that people found elsewhere, we didn’t realize it. So we understood some other possibilities, but the most practical and principal way to do it was to simply use the fact that if you take the polarization of light, the direct electric field vector, you know, crystals where the velocity depends upon that, then you can make up for the frequency-dependence of the velocity and that way have light at different frequencies travel together. That was not our invention. That was a key defect of our notion.
But we understood, for example, something people came back to many years, later, that you could obviously enhance these effects by confining the light very tightly by using optical fibers, that’s something in our patent. I tried, I must say, I tried to get people interested here in getting an experimental effort going. I didn’t try very hard. I was managing a fair sized activity by then. You try very hard, oh, certainly if you think there’s something of tremendous importance for the corporation. I didn’t have that conviction, that it was something very important for IBM, and just because of my personal invention I wasn’t going to knock people over to head to get them to work on it. So I didn’t try very hard. I tried but I didn’t try very hard. There was one early piece of work. There was a paper by A.W. Smith and N. Breslau in the early sixties on nonlinear parametric traveling wave effects, in the IBM JOURNAL, which was sort of the result of my lobbying. I don’t want to do it an injustice because I don’t remember what the paper was about, but I think if the paper had been very significant I’d remember what it was about.
So I would say, due to my lack of energetic prosecution, the interest in nonlinear optics in IBM really didn’t get very far, at first. It had to be reborn through different people, particularly through John Armstrong coming in with that interest from Harvard. And in fact, shortly after this period, I got very much preoccupied with getting this Corporation into Large Scale Integration, and for a while forgot all about these more fundamental things.
So I would say, the essential advance we might have had in going after nonlinear optics got frittered away.