Herwig Kogelnik

Bromberg:

I’ve heard the story that Kompfner recruited you, is that…

Kogelnik:

That’s true.

Bromberg:

What were you doing at Oxford? And what was it about Bell that attracted you to leave?

Kogelnik:

I did my studies in Vienna, as you know, then I got a British Council scholarship to Oxford, that’s the same as a Rhodes scholarship in the United States. There are various countries that got the Rhodes scholarships taken away after World War I, and the British replaced it by British Common [???]… Austria, Japan, I don’t know who else. Anyway, after my Ph.D. in Vienna I went to Oxford for two years and I worked with Motz. So that was a post-doctoral period I was supposed to be like a post-doc there, a post graduate I worked with Motz on what he called then undulators, but I didn’t really want to do a thesis then, or work on undulators, so we worked on another hot topic, which was electromagnetic radiation in magnetoplasms. Also, as you know, worked on undulators before.

Bromberg:

I don’t know what an undulator is.

Kogelnik:

An undulator is now called a free electron laser. And it is a machine where you shoot an electron beam in and you make it wiggle its path by having magnets North-South, South-North, alternating, so it wiggles the path. And if you shoot the beam in fast enough it wiggles very fast, and it would be a little oscillator in the axial direction. And it would give off radiation, even light. And Motz had seen light as early as 1956, I think, in Stanford. Just now got a prize for it.

Bromberg:

That’s interesting, because…

Kogelnik:

Because it is a predecessor of a free electron laser… it is a free electron laser.

Bromberg:

I knew that he had done this work in the ‘50s, but I didn’t realize he was continuing.

Kogelnik:

He was trying to continue, yes. The first work topic he suggested to me at Oxford was an undulator inside a wave-guide. And he has worked out gain and so on, it was a complicated thing with [???]s functions all over the place Then the radiation in magnetoplasms looked much harder to both of us, actually. And so I switched to this other topic.

Bromberg:

Did he consciously associate this with maser work, and with the Shallow and Townes paper on lasers? This didn’t seem to be…

Kogelnik:

Oh, no, because the Shallow and Townes paper on lasers basically came through later, because I started in 1958, hadn’t quite made it there.

Bromberg:

When it came, was it perceived as close enough to what you were doing to make an impact?

Kogelnik:

I left in 1960.

Bromberg:

And you hadn’t really been…

Kogelnik:

That’s when the first lasers were made.

Bromberg:

The paper itself must have come out in early ’59.

Kogelnik:

No, that was much too late.

Bromberg:

And that didn’t make any impression on what you people were doing at that point?

Kogelnik:

No, it was just a theoretical paper. The big impression is always the first experimental demonstration, isn’t it? That was made when I wasn’t there anymore.

Bromberg:

You came here in ‘61, is that so?

Kogelnik:

Yes I came to the United States in ‘60, the fall of ‘60, October ‘60, and I came sort of directly from Oxford. I spent some vacation time in Austria in between, back home. Probably (I would have to look up in my diaries) but I probably went to Austria in the summer and then in October I came here. And I came here on an interview trip arranged by the U S government… you can’t say that out1oud these days any more… just part of the procedures. And they paid my trip over and they said if after three months I don’t find anything I like, they would ship me back home again. They would pay me a salary for three months, GS12 I think it was. And for that salary I had to have interviews. They arranged the interviews. So I had 52 interviews, all over the United States and (12???, 5???) offers, and I picked Bell Labs out of that five.

Bromberg:

Well then, how does Kompfner have a role in this story?

Kogelnik:

Kompfner has a role, two-fold: he already visited in Oxford, and he talked with me there. He talked with my thesis professor, Motz. He talked with me at length and told me what Bell Labs was all about, and so on. And I told him that I was going to take this deal they offered — this interview trip — so I could get an overview. Then I guess, before I came here I heard so much about California, I was really biased toward California. I was going to come here and go to California. Kompfner must have arranged that on this interview trip Bell Labs was included, so I came to see Bell Labs two or three times in fact. He hosted… He made the interview extremely pleasant. I liked other places quite a bit too, for other reasons. I had a point system that I eventually thought I would use because it was getting so complicated. You can imagine with 52 places. The technical level of the place counted a lot, and proximity to CHICHUCK (?) counted a lot, proximity to skiing, proximity to an ocean all kinds of things, I had sort of a point system. But I came biased towards California, and I remember after all these interviews I had almost made up my mind to go to California. I was going to go into the space thing…

Bromberg:

The Jet Propulsion Laboratory?

Kogelnik:

North American Aviation then seemed very popular ...Eventually put the first man on the moon. And there was a lot of plasma physics there, and my thesis was basically plasma physics… Tell you about Oxford, I went there on a post-doc but eventually Motz said, “Look, you did this work, you might as well turn it into a Ph.D. thesis.” ... from Oxford... I have two PhDs.

Bromberg:

Actually, I would like to go back a little to Oxford, because even though the Shallow and Townes paper was only a theoretical paper, it made a big impression on a lot of people here. So it sounds as if the soil for this idea in England was somewhat different from the United States. I really am interested in Oxford’s reaction if there is anything you remember.

Kogelnik:

You have to imagine that I was not in optics, I was not even in EE then, I was very close to the fusion people. And there were people at Oxford who were working at Harwell (the British fusion place) all the furor about making energy out of hydrogen, basically.

Bromberg:

A very exciting time for fusion of course.

Kogelnik:

A very exciting time for fusion and they didn’t read the Shallow/Townes paper.

Bromberg:

OK.

Kogelnik:

I don’t think I ever heard about it.

Bromberg:

Did Komfner, when he talked to you?

Kogelnik:

He did, of course, yes.

Bromberg:

I see.

Kogelnik:

Komfner basically talked me into switching fields and starting to work with lasers. It is as simple as that. The offer to me then later to come to Bell Labs included that proposal: If I would join his group, I would switch fields and start working on lasers.

Bromberg:

Did you get any sense of; he was just excited about lasers, or excited about communications?

Kogelnik:

Komfner was always excited about communications, of course, because he always talked about the immense band widths that lasers offer. Just think of the band widths. Right from the start, pushed lasers… he pushed lasers for communications right from the start. That’s not true of everybody. I know Townes would tell you that he pushed lasers because of the immense potential for spectroscopy. Komfner was clearly the… I mean… or even I interviewed at Bell Labs, that’s what it was for.

Bromberg:

That’s what I was wondering.

Kogelnik:

He switched groups who worked on communications devices — namely microwave tubes — to work on lasers. That’s the group that I joined that had worked on travelling-wave tubes and things like that.

Bromberg:

Was that Gordon’s group, or do I have...

Kogelnik:

Gordon was part of that group. It was Komfner’s. Komfner was the Director; Gordon was the Department Head in that laboratory. I joined Tien’s department, which was a parallel department to Gordon. Komfner was a Director of what was then the electronics research laboratory.

Bromberg:

So Patel was in your group, under Tien?

Kogelnik:

Patel joined at just about the same time that I joined, that’s right, also Tien… joined months apart. We both joined ‘61.

Bromberg:

You must have gotten back to Europe about... came in October; you must have gotten back about Christmas time.

Kogelnik:

I got back to Europe …

Bromberg:

Oh, you just took this trip and then you...

Kogelnik:

I took this interview trip and I picked to go to Bell Labs, and I joined Bell Labs January 16, ’61. To go to Europe was a big procedure then, expensive. No, I picked… I went to Europe later, maybe once a year, or maybe once every two years.

Bromberg:

I then asked you a question about the relation between your background and laser work. What parts of the background as you started out were going to be especially helpful to you, and whether you had to go into study of any particular things as you came to laser work?

Kogelnik:

Well, most of the things we did in the early laser days were new, so you couldn’t go and study them any place, anyway. What you worked out was it.

Bromberg:

Oh, I see.

Kogelnik:

I didn’t know any optics, of course.

Bromberg:

What about quantum mechanics? Was that something you were well-trained in at the time?

Kogelnik:

No unfortunately, not too well. I picked a lot of that up along the way. And the same is true for optics, and part of the electromagnetics that I used.

Bromberg:

That too?

Kogelnik:

I had a deeper background in this plasma stuff, and microwave tubes from Vienna. In Vienna I worked on microwave tubes. My thesis topic in Vienna, my first paper is basically on microwave tubes. Do you have…? It is written still in German, my first paper. It is about energy principles in electron beams, basically. And that is travelling wave tube stuff. Archive der Electrir…, it is a German publication. But that’s clearly for fancy microwave tubes. So there are lots of electrons in there and electron streams and stuff like that.

Bromberg:

So you came… Did you… How did things look? What did problems seem to be?

Kogelnik:

Well, after I did all this evaluation of where I should go, I was very fascinated about going to Bell Labs, because there were all these many clever people there. And there was the beginning of lasers there. You have to imagine when I interviewed — that was in October, I guess — Javan, Bennett and Herriott had just barely demonstrated the helium-neon laser. And Mayven (?) of course, before at Hughes had demonstrated the ruby laser They were all very excited, and I think it must have been Komfner who induced into this group that are now the people who used to work on microwave tubes and did a lot of good stuff on microwave tubes. More than half of the people, I think, were still working on microwave tubes. And the whole group was making the transition. So the people had similar backgrounds as I used to have, in part. Of course there were some other people there, like Jim Gordon, who already were old hacks at quantum electronics. He had demonstrated the first maser to Townes, and he knew all about lasers. But with the exception of Jim, and maybe a few other people — Yariv was there too at that time…

Bromberg:

Yes, he had done his graduate thesis on...

Kogelnik:

…was an exception to that. People were making a transition from old disciplines to lasers.

Bromberg:

So you fell right in?

Kogelnik:

I fell right in. I was not alone learning; it was everybody.

Bromberg:

OK. Were there a lot of seminars going on, or people studying together, or…

Kogelnik:

A lot of discussions going on, I don’t remember any specific seminars, but a lot of interaction.

Bromberg:

That is…

Kogelnik:

It was like cooking. I remember I had lots of contacts with Jim Gordon, Amnon Yariv, Sergio Porto, Gary Boyd, Louisell was there, and we did a lot, Herriott. And everybody learned from each other. Then we would go back, learn something new. It was sort of a new field evolving. No particular background helped you very much. Obviously quantum mechanics, you had to know a fair deal of electromagnetics... It was sort of a merging of all of those things.

Bromberg:

Yes, it’s an interesting slant, and one that you don’t really think of, that you have to put yourself back in the idea that the whole field is beginning to grow.

Kogelnik:

It wasn’t there. When I came really it wasn’t there...

Bromberg:

It’s not like reading into a new field.

Kogelnik:

There was this clumsy Javan helium-neon laser that basically you couldn’t do anything with. It had to get all the attention, and be babied so it would work. That was basically the gas laser that I saw on my interview. That was still the only CW laser developed when I came. Immediately one question was how can one transform this monster into something that you could do something with that you could work with. Very early I got into this project of trying to make a Brewster angle laser...

Bromberg:

Was that Rigrod’s idea, or how did that begin to develop? I guess he is the first man on the …

Kogelnik:

He is the senior man on the paper. He was also the oldest, he was the most experienced. I think the idea came, it was in the group there somewhere, I can’t pin it down exactly, but I think Kibler and Fox had a lot to do with the idea.

Bromberg:

Who was the first name, Kibler?

Kogelnik:

He was in Holmdale. The idea may have started...

Kogelnik:

Many people at that time bought themselves this Born and Wolf book and tried to learn some optics and Brewster angles are there in the first chapter or second chapter or something like that. I think the idea was around. Neither Rigrod nor I had the original idea. But many people were not quite sure how to handle it. I know some people wondered, gee what influence is the Brewster angle going to have on the quantum states of this thing. And went off to study that, and then and then other people were worrying what would happen about the roughness of the windows, was there going to be so much scattering loss that the whole laser wouldn’t work — we still called it optical maser at that time. In fact, I went off with Herriot to test the flatness of mirrors and windows to a very high degree just to make sure. Remember Javan’s laser had the mirrors inside. So, to put Brewster windows between the mirrors and the laser then seemed like a very strange thing. You didn’t know was the scattering going to kill it; you didn’t know what happened to the quantum mechanics, or... Some people didn’t know. And we decided to just go ahead and try it.

Bromberg:

Were you working very closely with Bennett or Javan, or just with Herriott on these things?

Kogelnik:

Mostly with Herriott. Herriott was a tie into the first laser (helium-non laser). We in a way built the second version of it with the Brewster windows, and then we had to write about the advantage of curved mirrors.

Bromberg:

Did you have any person-to-person discussion with Boyd at this point on the…

Kogelnik:

Yes, sure. In fact, I worked with him almost simultaneously on this other paper there.

Bromberg:

I see, on the generalized…

Kogelnik:

…generalized, what happened if…that’s generalized. Boyd and Gordon had basically done the confocal one, and now there were lots of questions. What if it is not confocal? What if the curvatures aren’t the same? What if the Brewster windows effectively change the curvature (which in fact they do)?

Bromberg:

Oh, I see.

Kogelnik:

It becomes astigmatic, right; this curvature becomes different than this. What if a Brewster window is in between? All kinds of questions like those. Very practical questions came up. And yes I worked with Gary, we wrote that paper together.

Bromberg:

That there were seven, But what I didn’t realize is that that this and this one were really going on simultaneously.

Kogelnik:

Yes, you see that this has concave mirrors on it. So this was Brewster and angles and the first use of curved mirrors.

Bromberg:

And then Boyd told me it had actually begun with plane mirrors and then you switched to concave as you were working along this?

Kogelnik:

On the experimental plane mirror?

Bromberg:

Begun the apparatus with plane mirrors and then…

Kogelnik:

To do this with plane mirrors would be a terrible thing. I don’t think it would work just that easily.

Bromberg:

I see. You think it probably started out in the very beginning with concave?

Kogelnik:

I thought that experimentally we tried at the outset curved mirrors just because we thought it would be too tough to do anything else. Any slight non-planeness of the Brewster window would make a curved mirror out of a flat one anyway. In the beginning… [Kogelnik is looking through lab notebooks] In the first week or two I tried something completely different... These notebooks are terrible because there are so many proposals in here. OK, this is March '61, that surely... Here it says, “Kopfner’s proposal.”

Bromberg:

Oh, that’s interesting; did Komfner have a role in it?

Kogelnik:

I worked on something else in the first month, obviously, and it had something to do with some properties of Javan’s laser. Why was it polarized? And then there was something about phase modulation of light; all kinds of stuff. And here it says “Komfner’s proposal, March 7, 1961.” So right after discussing something with Rudy Komfner I drew that big picture here and it had two Brewster windows in it and two curved mirrors. So I never have a drawing with flat mirrors in here. And the questions were: what is the influence of the loss of the mirrors? What is the influence of the roughness or the tolerance of the windows? Is it required that these windows are perfectly parallel? How can one measure those things?

Bromberg:

Do you think that Komfner then played an important role of any sort in getting that whole experiment…

Kogelnik:

Well, he was obviously my input.

Bromberg:

Of course, one doesn’t know who was discussing it with whom.

Kogelnik:

But later I think he may have transported the idea from someplace else. But he was certainly a link in the chain. He never claimed credit for that. But he was a very modest man. He often stood back so that ideas would flourish. And then right afterwards I seemed to have tried to analyze what happens when light goes through Brewster windows as you can see. What happens with light deviation from the Brewster angle …

Bromberg:

Any chief analyst on experimental...

Kogelnik:

…regarded more the experimental side, but there was no such thing as a chief analyst, we all did both. I think that was strength of the group. I had my hands in lots of experiments.

Bromberg:

The other people were Herriott, and I am leaving one man out here… Brancaccio…

Kogelnik:

Brancaccio was the chief mechanic technician. He learned how to put Brewster windows on to these vacuum envelopes very precisely and accurately.

Bromberg:

I noticed at one point he made up something close to forty of these for various people in the laboratory and that I was… I was curious about that because I wondered if it was common to have someone sit down and make forty lasers. But if you say he was principally a mechanic on the…

Kogelnik:

He was running the shop of quite a few people back there. He always had good ideas of how to do things. He…there was a shop that I think was doing the technology for the microwave tubes.

Bromberg:

Ok.

Kogelnik:

And that know-how came over. There were also vacuum tubes now that were filled with gases. As you can see I went right into looking at what happens when the windows were irregular, and all these worries. It is not even easy… part of this stuff is not even published…Then I went with Herriott — this was all done with Don Herriott — to figure out how to determine the flatness of surfaces down to, I don’t know, it was terrible… lots of plates.

Bromberg:

Was it something novel that you had to invent in order to do that, or did you simply…

Kogelnik:

Well, there was an interferometric scheme using many colors that Herriott had, and I invented a mess on how to get the deviations out. Hin did obviously a lot of that work because he had jumping gates (?)… April, then all of a sudden I am in July... how this thing must have already worked…

Bromberg:

Let’s see what the date on that is. That would be on the paper, the submission date. It was published in February of 1962 in the Journal of Applied Physics, but the submission date would tell us…

Kogelnik:

… that’s plasma physics, plasma physics... August...

Bromberg:

So that’s just when you say, July it is just about finished.

Kogelnik:

Well, we started working other things. And I can see, when a thing first worked, we always had confocal mirrors. So, I obviously didn’t understand things yet because to use confocal mirrors is not the thing to do. The thing goes unstable.

Bromberg:

So this one has confocal mirrors?

Kogelnik:

The first proposal was confocal. The thing that worked much better was not confocal, deviating significantly from the confocal condition. It turned out — that’s the paper with Gary Boyd — that right near the confocal resonator any slight deviation would make the thing unstable. And you always have slight deviation.

Bromberg:

Now that paper with Boyd is also talking about optical transmission lines as I remember it. When do you begin to consider optical transmission lines?

Kogelnik:

That idea was already around.

Bromberg:

Even as you were coming in? When you say that idea was already around, you mean as you were coming in to Bell they were already…

Kogelnik:

The idea that a resonator was the equivalent of a sequence of apertures or lenses — you know light bouncing back and forth is the same as light bouncing along —

Bromberg:

That’s just in the literature; the engineering literature?

Kogelnik:

It was already in the Boyd-Gordon paper I think and the Fox-Li paper. It was more around maybe as an analytical tool, as a concept to think about a resonator. Right, you have two mirrors facing each other. How do you analyze it? Well they say it is the same as if light were bouncing off.

Bromberg:

…around in Bell Laboratories, kind of a Bell Labs idea? Or was that something that was well-known when you talk about resonators?

Kogelnik:

When I came that idea was certainly around in that group in Bell Laboratories. Like Boyd or Gordon or Komfner would have it and I think in retrospect also Fox and Li would have it. I think it is in their papers also. And then later Goubau and Schwering from Fort Monmouth also said they had thought about that independently. Coming from the microwave side, I think.

Bromberg:

I see. So it is not a completely obvious idea that everyone knows, it is coming into being.

Kogelnik:

It is just coming out, yes. Within… let me see what we wrote about that one…

Bromberg:

It seems so completely tied to Bell interests in the laser…

Kogelnik:

Frankly, it’s not completely because of Goubau and Schwering…“equivalent sequence of lenses”…OK, there is a tie into the electron beams again, the microwave tubes. Now, we don’t seem to make a big spiel about it, we are just using an equivalent sequence of lenses in a way to analyze the stability of these resonators. And one cool, quick reference is to 7, which is Pierce, which is “Sequences of lenses used for electron beam.” Theoretical design of electron beams,” that’s a book. Old stuff, ‘54.

Bromberg:

OK, so then it is pretty much in the literature? Goubau and Schwering, were they people who you were in touch with?

Kogelnik:

No.

Bromberg:

Or were they just people in the literature?

Kogelnik:

They are in the literature. The use sequence of lenses too. And I think Fox and Li did, and that Boyd and Gordon did. These are all ’61 papers.

Bromberg:

So it sounds like an idea which is not as commonplace as Brewster windows, but that is somehow around.

Kogelnik:

But it was around. The idea of sequence of lenses representing effectively resonators, that certainly was around. It was a powerful concept to analyze things. And then later somehow it got translated to… And not even later, I think Goubau and Schwering certainly started from that side.

Bromberg:

And were you people interested as early as 1961-62 in thinking about…Was anybody in your group, or were you doing any work on transmission lines themselves at that point?

Kogelnik:

That was probably being done here at Holmdel, Fox and Li. You have to ask that question of Fox and Li. The answer is yes.

Bromberg:

But from you, from your group, you weren’t yet thinking about designing transmission lines, or how they were…?

Kogelnik:

Well, the transmission research was down here at Holmdel. Up there, you understand, the group I joined was the electronics research lab, we were after devices.

Bromberg:

I didn’t completely understand that actually.

Kogelnik:

We were after devices at Murray Hill. And down here they always did the transmission research. This lab where I am in right now is actually where it could have been if this was [???]

Bromberg:

I do know for example that before you came, but maybe while you were interviewing, there was some kind of experiment going down here, where they were using the earliest lasers — maybe the ruby lasers — to transmit the signal between two hills?

Kogelnik:

Yes, from this hill up here, we are here in Crawford Hill…

Bromberg:

Right.

Kogelnik:

and there is a signed up there on the hill (where there are lots of telescopes too)… From up there I think they pointed a laser to Murray Hill (that’s the other laboratory)… first lasers yes. It wasn’t before me, because we built the first practical laser for such purposes — this one.

Bromberg:

I see…

Kogelnik:

Before that there was no laser, practically, for communications because the Jovan laser was just a terrible cl—-, you couldn’t move it out of the laboratory.

Bromberg:

But I am almost sure I saw that in something like ELECTRONICS MAGAZINE for fall, 1960. I’ll check it.

Kogelnik:

OK. Maybe they used the ruby laser.

Bromberg:

I think they used ruby lasers.

Kogelnik:

But the Ruby laser was only pulsed, so it couldn’t transmit any messages.

Bromberg:

That’s one of the reasons why I wondered whether that was just a publicity stunt, or what was going on at that point.

Kogelnik:

Maybe detecting photons or something.

Bromberg:

Ok.

Kogelnik:

You couldn’t make a communication… At that time, to make a communications link with lasers you would have had to use this one because the ruby laser was only pulsed. OK. So you could transmit messages. I don’t know how often it pulsed; maybe once a second, or something. And the other one I don’t think you could move out of the lab, because it was tied to a vacuum station and all this stuff. It was all one vacuum station, had the mirrors inside. This was the first practical laser for such a purpose. (?) CW, so you could modulate it, and it had to be sort of a rugged device. That wasn’t all that rugged, but it was pretty portable.

Bromberg:

Well, already I have some very different ideas on what happened and what I came in with. As you worked on the resonator, on these various kinds of resonators and so on, what were some of the most key breakthrough concepts would you say? For example, there were papers on the conversion of modes from laser system into an optical line, or they were studying a b——ns and bai— media and so on.

Kogelnik:

This one we can get rid of. There are two key concepts in this one is the use of curved mirrors, the other one of Brewster angles. Both ideas were in a way around, but putting it all together, making it all work represented a big step because from now on you could have CW lasers that looked pretty practical. We gave a paper at the DEVICE RESEARCH CONFERENCE; I think that was the first disclosure. That must have been in June.

Bromberg:

That must have been June, 1961.

Kogelnik:

June, ’61, yes. And that excited everybody. I still remember that because the others were considered laboratory curiosities, or something. And now all of a sudden you can go and use it. And indeed, a lot of what I did later on was go and use it. After all, we had more or less the first usable radar. And we tried obviously to find out how it behaved, and I think this next paper is that. Right here we saw for the first time what is now called the single mode laser that is…

Bromberg:

Number 5 here.

Kogelnik:

That’s the first time the modes were ever seen — the laser modes were ever seen. And they obviously came out beautifully, you know right out of this laser. The lasers with the plane mirrors, they just wouldn’t give you these beautiful pictures. What came out of those was determined by some kind of irregularities: mirrors, crystals and so on. For the first time you could see a very regular behavior of the laser beams that were coming out and exciting. All of a sudden these beautiful patterns came out, depending on how we controlled it.

Bromberg:

Is that a difficult experiment to get to work, or can one pretty quickly…?

Kogelnik:

Well, having that laser that one could work with, it wasn’t difficult. You stuck little obstacles into the light path. The light path between the mirrors was now accessible; before it wasn’t. And we could trigger to do this. In fact, the resonator theory work helped too, because we could adjust the mirror spacing’s and mirror curvatures to conditions that would favor the purity of the laser light that would come out. You understand the simpler the pattern is, the purer the light. Other people call this coherence, right: Incoherence is a complicated superposition of all these modes. But here for the first time, by adjusting the resonator geometries right, and by putting obstacles in the light path inside the cavity, we could make the laser go in pure modes. And we photographed them. And I guess these pictures are still all over the place.

Bromberg:

Yes, indeed. We see these pictures everywhere still.

Kogelnik:

Because it supports all these theories. The theories were of course around before, and that’s basically… Now, single mode fibers, and single mode lasers are very important, right. This is the first observation, and first demonstration of the single mode laser; this one spot here.

Bromberg:

So that again was a very exciting result?

Kogelnik:

You ask for breakthroughs; I consider this a major breakthrough because you get single mode 1-yard, which means spatially pure light; light that you could work with sent over long distances. Right now we use it to feed fibers and send over long distances, but… Also, if you want to go very far, you need to be pure.

Bromberg:

When you finished with this result, did you follow up on it or did you put that aside for the time being and then went on to something rather different?

Kogelnik:

Well, that’s tied intimately into the theoretical paper here with Gary, Gary Boyd. This one.

Bromberg:

And what…

Kogelnik:

They went on parallel.

Bromberg:

And that generalized confocal resonator theory and will predict those modes that you pictured there?

Kogelnik:

Yes, and in fact, the exact size of those, and the spacing. Even in here there is an analysis of the spacing. You know these zeroes there? You could measure how far apart, how they are related to each other, how they increased slightly, how fat these things are. All these sizes, all worked out.

Bromberg:

I don’t suppose you remember at this point whether you worked them out first or got the pictures first?

Kogelnik:

Oh, no, no. We referenced Boyd-Gordon here. We worked them out first, or got the pictures first? The notes that go with this… I am still in July… We are observing first here, clearly; not just that. We looked at patterns, but we also looked at beats. Each of those things would run at a different frequency and we mixed that and measured the best frequencies, megahertz and so on. Come in the various obstacles, right. Obstacles in the aperture, “preferred orientation improves the direction…”

Bromberg:

Of course, this also gives me a very different picture. I, for some reason, had formed the picture of your being mainly a theorist. I don’t know why. And the picture of you in the lab putting obstacles in the path of the laser beam is…

Kogelnik:

It was a lot of fun. I wrote quite a few experimental papers. I always get inspired by experiments, I think. Then the theoretical papers get more noticed. But I’m very proud of this Brewster angel laser paper, and obviously I was the new guy. But Bill Rigrod and I sort of shared the leading…

Bromberg:

It’s absolutely just a gratuitous supposition on my part to look at these papers and say, well…

Kogelnik:

This happens to me often. You are not the only one.

Bromberg:

That people think of you mainly as a theorist?

Kogelnik:

But I will point out to you all the experimental papers. That one obviously is an experimental paper. And I’ll point out several others to you. Here, all these — experiments, as you can see, for months I did experiments here. Fooling around with obstacles, apertures inside… O, my God, and then already in September '61 I was trying to make amplifiers. A lot of experiments with amplifiers, and the laser gain was so awful — small — that I never got very far, there are one or two papers later on.

Bromberg:

Do you remember what the reason was that you started to look into amplifiers?

Kogelnik:

Well, communications. First you need a source.

Bromberg:

Repeaters and things like that?

Kogelnik:

Yes. Obviously in September’61 I was already working on amplifiers, here “optical maser amplifiers” it still says here. And the gain was so small, so I wanted to bounce the light several times through the tube to pick up a little more gain.

Bromberg:

So this whole interest then is what eventuated in your paper with Yariv? At one point…

Kogelnik:

There is a direct tie, yes. But there was clearly experiments first trying to make an amplifier.

Bromberg:

Now is it right to guess that people in your group would be thinking in general about this kind of thing? What do we need in a laser system, and how would we…?

Kogelnik:

Sure, that triggered a lot of ideas. What do we need for communication experiments? We need a pure laser, we need amplifiers, we need modulators, and we need detectors... All those questions were floating around all the time.

Bromberg:

So would I be right to think of this in terms of, it’s not really applications-oriented in any immediate sense but the general idea of communications is propelling the kind of directions…

Kogelnik:

That’s right. There was a very long term goal. But the goal is very clearly there, discussed all the time with a guiding light of sorts.

Bromberg:

Also discussed with the people here or that not as much? I mean, was there a kind of

Bromberg:

...some interaction between the transmission people down here and ...

Kogelnik:

…getting stronger because Komfner was appointed Director of both places. He was allowed in Crawford Hill here, and he was also in charge of the Murray Hill operation. So he was a key link between the systems people down here, who were at that time still mostly working on microwave transmission, particularly the millimeter wave guide project. But they obviously said, and knew that to transmit you need amplifiers or generators.

Bromberg:

…come around and talk to you people so that…

Kogelnik:

Yes. And then people talked too. I know that pretty soon they wanted some of those lasers that we had made, the Brewster angle lasers, to do transmission experiments down here.

Bromberg:

Was Komfner closer to some people in the group than others? I mean, were there particular intimates of…

Kogelnik:

Oh, I think he was mixing very liberally everywhere. But he was a very good trigger. Always full of ideas, and he knew there were very many ideas out there that had to be sorted out, and he was very generous. Spilling all these ideas around, letting other people sweat them out.

Bromberg:

Now, another … maybe I am jumping to fast. But I have a whole thing I would like to know about. I would like to know a little bit about this business of mode conversion and mode coupling, and what the principle breakthroughs or ideas might have been.

Bromberg:

OK. Maybe I should shift through here and get there. Here, there are lots of beat measurements between modes. Trying to understand modes experiment.

Bromberg:

OK. That is still September ‘61, and this mode paper, this visual display of isolated modes, is received in December ’61. So that must be still…

Kogelnik:

So that was working up towards it I guess. I did a lot of experimental work. Look at all these numbers. Beats, beats, beats as a function of everything. Mirror curvature… anti-bounce system for amplifiers… Here comes filter… interferometric amplify… [???] was going to be. The entries are a little spotty. Oh, this is already March ’62. There is a big jump here, I am afraid. Because I wrote all these papers in between, and I thought they were documented in the papers. I can see that you jump here from November ’61 to March ’62, and from now on it is probably only new ideas that get in here. This is a book which we have to keep, to sign new ideas. You see these signatures all over the place. And then have things witnessed, and it’s for patent purposes, this log book. And I think from then on, I did all of the work outside of the book.

Bromberg:

I see.

Kogelnik:

And only put key ideas in there. It is still kind of a log, but not as complete. I started out very thoroughly, and then things fizzled away.

Bromberg:

But then, some of the things we are trying to get at here, like significant novel ideas, ought to still be in there.

Kogelnik:

They are still there. There are lots of books, yes, a whole stack. You see I have all these other books...

Bromberg:

Yes these down under here behind your chair.

Kogelnik:

And I split things into subjects because I got interested in so many subjects, so I would develop subjects in there. But the key ideas would go here for patent purposes.

Bromberg:

— and these books are going to be part of Bell records eventually?

Kogelnik:

No, these are just mine. These are the Bell records.

Bromberg:

The reason I ask is because one of the things we are always on the lookout for is, what is happening to the paper in which the work is recorded. What is happening to the notebooks and memoranda and correspondence…

Kogelnik:

These I keep. These are Bell Lab’s property, these numbered notebooks.

Bromberg:

Ok. So that the person who is going to document your work is going to want to go to you at some point and say, “Can I…

Kogelnik:

The more important things would be filtered out in here.

Bromberg:

Ok.

Kogelnik:

Because for patent protection we need to enter things in here and sign things here. Theoretical calculations, like this stuff, I would never again put into these notebooks, you know I just measured along.

Bromberg:

Yes, fine.

Kogelnik:

… for months. You know, what do you do with all these numbers? They clog up the notebook. Buy my record keeping is obviously not the best. Ok, you had a very specific question: Major breakthroughs. I thought this one and this one, which are the first two papers which were pretty fundamental steps; particularly here that we get a pure single mode. Beams are very important… pure beams that you can get. Paper still an afterthought from Oxford, I guess. Just came out later.

Bromberg:

This one by you and Boyd, we ought to consider as a kind of companion to the visual display paper, or…?

Kogelnik:

No, no. This thing here went on at the same time, but I think this one is a major theoretical paper on the resonators.

Bromberg:

Of course, you already said it contains ideas on both…on the idea of stable versus unstable resonators.

Kogelnik:

That’s right. That’s probably the key advance. We tried to spread out from what was known. And in trying to spread out it turned out that not all resonators with two curved mirrors were useful. In fact, some were stable, and some were not stable. That is one key here…this stuff.

Bromberg:

Now I always think of unstable resonators as an idea that comes along later in the decade. Am I just confused about that? I think of Siegman.

Kogelnik:

…of unstable resonators. Siegman put them to use to get large mode volumes in some high gain lasers. The discovery of them is this one, right? It is a published paper, just as I told you before; there are things in here that I actually published that people overlook. But I think I named them. Let me see what I named them here already; the word stable and unstable to designate resonator classes.

Bromberg:

Siegman then rediscovered...

Kogelnik:

No, he knew that, obviously.

Bromberg:

Ok.

Kogelnik:

See here: “We named the region unstable when it was not stable, and we named it stable when it was stable.” And there was a very specific meaning to it, in terms of the light wanting to spread out when it was unstable, and the light being re-focused all the time by the mirrors when it was stable. And we produced these regions of stability, right? This plot here, where you plot the distance between the mirrors divided by the radius of curvature of one mirror, and the distance between the mirrors divided by the radius of curvature of the other mirror; these two axes here. Then you get what we call our stability chart: “Diagram of stable and unstable regions.” That’s Figure 6. That’s a very key figure. And the discover that resonators divided into classes, the stable ones in here and the unstable ones (the shaded regions) that’s this paper listed right here. And it says here “Two stable and unstable regions.” Right?

Bromberg:

Does any of this work — by the way, and this is just a general question –- reflect your plasma physics background in anyway, or not?

Kogelnik:

No. But the microwave tube background comes in here, right? I mentioned that just before. We did indeed refer to Pierce here. Reference 7 is a microwave tube reference.

Bromberg:

Yes, when we were talking about the lens sequence: And there were these sequences of lenses for electron beams that went unstable and stable. And now we translate it to resonators. Then we made them unequal too. There is a link here to microwave tubes. I obviously knew where to look this kind of stuff when I went to Pierce’s book on electron optics.

Bromberg:

By the way, did Pierce play any role in here, or was he too high up in the organizational ladder.

Kogelnik:

He played a very interesting role. He was Rudy Komfner’s boss, and on the hand he seemed to support whatever Rudy was doing. Rudy pretty soon moved up, only a half a notch below Pierce. So Rudy Komfner was pretty high up to… Executive Director and Komfner was Assistant Executive director. There seemed to be a [???]. So Rudy was sort of a half of a notch below Pierce and seemed to be in charge with everything that had to do with lasers.

Bromberg:

Because in the conversation Komfner is coming up all over the place, while Pierce isn’t appearing…

Kogelnik:

Well, Pierce came through, if anything, as rather skeptical of the whole thing. But as he believed in Rudy, he let him do his thing.

Kogelnik:

He was then composing computer music and stuff of that nature. His interest at that time was someplace else.

Bromberg:

I see.

Kogelnik:

And a long time he would think that optic communications may not have such a chance because of satellites.

Bromberg:

Ok.

Kogelnik:

Pierce is in a way the inventor of communications satellites.

Bromberg:

Do you think it makes sense to try to interview him, or do you think he was just…

Kogelnik:

Oh, it would be very interesting to see what he has to say now. After all, he was in charge. He was in charge of these projects; there is no question about that. Rudy Komfner reported to him… obviously the freedom to do a lot of work on this which turns out to be a very broad view, and a willingness to invest in a lot of research and a long time basis. After all, optical fibers are only coming in now. There is a 20 year time scale here, that people… a massive scale too.

Bromberg:

I have no way to judge how big that investment was in comparison to other research investments at Bell Labs. Can I begin to think of it as… you just said massive?

Kogelnik:

It was major. It was obviously massive.

Bromberg:

Because, you know, the financial numbers are not released yet.

Kogelnik:

Yes they may never be accurate. I don’t know what they mean, but just look at the papers and count the number of people. And you will see it is massive. And you know what an average salary is, and what a loaded salary is. It added all up when it comes to enormous numbers, I’m sure.

Bromberg:

What was a loaded salary at that point? About…

Kogelnik:

I don’t really know; I wasn’t in management. But you know from universities, it is like here, it is always a factor of something or other, times the actual salaries.

Bromberg:

I actually know where I think I can get those numbers.

Kogelnik:

Somewhere... universities are somewhere between the factor of 3 and 4. We are not much different.

Bromberg:

It’s important.

Kogelnik:

And, oh yes it was clearly massive. So, well, maybe I am right? I buried this in a big paper and some people think that Siegman discovered unstable resonators. Well, here it says “unstable,” the paper is about resonators.

Bromberg:

I haven’t read through this paper yet, so…

Kogelnik:

I have no quarrel with Tony. Tony recognized that by working in this unstable region you can force a large mode volume when you want it. And he used it for some applications. So the application of unstable resonators, that’s Tony.

Bromberg:

Ok.

Kogelnik:

But the fact that they exist, is clearly this one. And the discovery came about by trying to work out what happens to resonators in the general case. You have different mirrors when you space them in all kinds of spacing, and trying to work out what the beam size was for all these situations, which we work out here. Trying to work out what the frequencies were going to be and so on. Here are the beam sizes, and the frequencies and the diffraction losses that seem to play a big role at that time. Looking at a resonator like this really did it, right? We said r1 r2 and the distance. And what happens for every case you can imagine.

Bromberg:

There is a whole lot of stuff in your papers that I just hadn’t seen in the papers that I looked at. This may just be a function of the fact that I have looked at a limited number. And that is these considerations on the actual shape of the beam and the way the beam is behaving inside the resonator. I had just not seen so much of that.

Kogelnik:

Yes, looking more at the beam than what makes it, right?

Bromberg:

Exactly. Is that something that was, should I identify a group of people who were interested in that, or was something peculiarly that you were focusing on, that… I mean, where does that stand on the whole unrolling of laser information? People like Lam and so on, who were doing laser theory, you think of them as interaction of atoms and radiation…

Kogelnik:

Most people at that time, and they were already going when I came, looked at what would generate coherent light. Right. What would be the materials mechanisms, the quantum mechanical mechanisms that would generate the light? I looked at more how it would go on from there. Once you managed to do this, how does it go on? What would be the nature of the light that was forming in there in a little more detail, so I could figure out how it would go on? How would you propagate an injector laser beam into something else?

Bromberg:

Could you identify the universe of people who were interested in this kind of thing or if it started with you, how that universe spread, both in the laboratories and maybe your contacts. Who were you talking to and who were you…

Kogelnik:

Well, some of stuff that I did then is actually now used by everybody who uses lasers. It is the users of lasers, rather than the users or lasers, rather than the generators of laser light. Whenever you use a laser, you have to figure out what the beam does that comes out of there. So you know what to do with it.

Bromberg:

I am trying to think who the users might be in the early... middle sixties; people who were modulating the light.

Kogelnik:

I don’t think there were any, really. Because you don’t use it until you have one, and I guess it traces back to the fact that we had the first somewhat useful Laser.

Bromberg:

Oh, I see.

Kogelnik:

That was already behaving like a generator of pure light. The ruby laser was nice, but it didn’t have a pure beam coming out of it. That was the other one.

Bromberg:

So could I be thinking that it is going to be in the late ‘60s and early ‘70s when a lot of this stuff will be picked up, or…

Kogelnik:

Well, it got picked up more and more and more. But it would have got picked up right away for some things. But now some of this stuff which we wrote is taught in every in even undergraduate classes.

Bromberg:

Good. Why don’t we go on then? There is a paper that I did not look at yet. Called “Mode Suppression in Single Frequency Operation in Gas Laser,” where you were working with Patel. What was…

Kogelnik:

Same thing, right. We just talked about the purity of the light. I just showed you this spot here which was the first pure mode, spatially. And it still, however, had several frequencies in it because of longitudinal resonances. So the next step was trying to get a single frequency. We work on that now too, single frequency lasers. And this, I guess is the first single frequency laser. That is the first laser that is truly coherent.

Bromberg:

Were there any special steps that you had to take that were revolutionary?

Kogelnik:

Yes, you see there are three mirrors now. There are two cavities that resonate against each other and that sort of in a comb-like way suppress all spikes but one; if I can just schematically draw it for you. If you draw… What’s your background?

Bromberg:

I have a Masters in physics.

Kogelnik:

Ok. If you draw the frequency, mu… if you would have said double e, I would have said f. But anyway, frequency mu, a cavity consisting of two mirrors has resonances like this. Spaced regularly apart, and that thing is velocity of light divided by twice the distance (c/2 L). If you make a second cavity, with a different distance, you might have resonances here, here, here, right? Different spacing because it is a different distance. Now, if you couple these two cavities, only that frequency will go where both resonances coincide, in very crude terms, now. So, here is a coincidence, right, that will work. Now in the way that I have drawn this cleverly, in no other case is there coincidence so nothing else will work, and you will only get one frequency out.

Bromberg:

And that is what you were able to do experimentally.

Kogelnik:

That is the basic principle, and here are the two cavities, two coupled cavities. And we got a single frequency out.

Bromberg:

This may seem very formal, but the tape is not going to understand this unless I keep it. It’s going to be otherwise a complete mystery.

Kogelnik:

And Cumel and I got together and we did that, and I guess that’s the first single frequency laser. Now we need things like that in communications.

Bromberg:

I see. At that point it was a just a kind of…

Kogelnik:

How do you get a single frequency? Well, the laser is supposed to be a source of coherent light. And coherent light in the ultimate definition has to be pure spatially and has to be pure in the spectrum. And I think this is truly the first really coherent light that came out. Everything else had several frequencies, several spatial modes.

Bromberg:

Where was this? Was this presented at a meeting also?

Kogelnik:

Yes, it was presented at a talk also.

Bromberg:

And was this... Yes, you say it was at that American Physical Society in Seattle. Where there many reactions to it or was it lost in the general excitement?

Kogelnik:

Not really. This sort of disappeared again. Now, everybody is very hot after single frequency lasers, because the more advanced fiber transmission systems, they want them now. The early, more primitive ones didn’t, but now you want pure light. Now the interest is all over. Earl (?) was 20 years too early.

Bromberg:

Now this paper with Porto on continuance helium neon lasers, a ramen source that struck me as if it might have just been a kind of aside.

Kogelnik:

Oh, but I did lots of asides, of course.

Bromberg:

Yes.

Kogelnik:

I like asides. But it does fall into the pattern of trying to use the first Lasers that were useful that we had.

Bromberg:

I see.

Kogelnik:

I sat in my lab and I had a laser that beautifully worked, that was very accessible, very flexible…

Bromberg:

And Porto came by and said…

Kogelnik:

and I knew Porto, I was talking things with Porto and he had already done spectroscopy with a ruby laser, or something. I said, look, my spectrum is so much purer. If you want to do spectroscopy you need a CW laser, and we have one, and he wanted one, so we did this together. Demonstrated obviously that very pure spectrum came out. Then a lot of work happened that was using CW lasers. Porto was more the spectroscopy interested man.

Bromberg:

Yes, but I think of him as the person who is interested in rahmen (?) spectroscopy.

Kogelnik:

...And I was more interested in using lasers any which way. In that sense it fits right in. You can see even the beam theory was already thinking of using lasers somehow.

Bromberg:

I see.

Kogelnik:

… because you had to understand them to use them.

Bromberg:

I can see why, yes.

Kogelnik:

It wasn’t all that far out for me.

Bromberg:

Good, I mean, this again, as I said, makes more sense of what seemed to me — just coming from the outside — as an aberration, or as I said, an aside. But, becomes more integral.

Kogelnik:

But on the other hand it is absolutely true, that I liked breadth. I have been moving all over the place. Interested in various things.

Bromberg:

Ok. Now we get into this business of moving your optical modes from one system to another and there is a paper “Matching of optical modes,” and then that is separated by some other papers. But then there is a paper on coupling and conversion coefficients for optical modes, we’re mostly in the ’63 period.

Kogelnik:

Here is this Porto paper, right?

Bromberg:

Well, there is one I am jumping just for the moment, because… what am I jumping?

Kogelnik:

This laser amplifier paper.

Bromberg:

Oh, that was just because it was in my particular list; it was after the coupling paper. But we can talk about that first, and then…

Kogelnik:

Oh, maybe it is just stacked in here differently. Ok. There were sort of 3 coupling papers, 4, 5… already saw that I was trying to use lasers, right in many different ways.

Bromberg:

Right.

Kogelnik:

... and for example, the paper with Porto, right, you had to shield unwanted light? To do that you had to focus the light to a small volume. Then you could shield everything else. Already you had to know how to handle the beam to optimize experimental situations. It wants to use laser beams. You already saw that I tried to make amplifiers which might [???] amplifiers, which meant injecting the beam in a precise way so you could bounce it around several times without it blowing up, and things like that. Again, I had to know how to handle the beam. And, well, I worked it all out. So I need to couple modes to other structures.

Bromberg:

Ok.

Kogelnik:

It is really… one thing gave the other. What the heck was this? This was amplifier stuff. That I never pub1ished, I guess, but I did give a paper at the device Research Conference. There are these mighty pass [???] cells, called white cells, and I was bouncing around beams to amplify the many passages, but there was gain inside, for the amplification. So I wondered when this thing would start to oscillate all by itself. And that’s actually a beautiful theory. Why the hell I never published that, I will never know. Ok. Then came the question, in that amplifier… here it is, both sizes in white cell… how do you have to inject the beam so the beam will always stay nicely bundled and contained? That question is addressed here. And the answer is you have to inject the beam in a certain way; right? With a certain size, certain direction and the minimum spot at a certain location. So the white cell here specified very clearly how I had to inject the beam to get the thing to bounce back and forth and stay bundled.

Bromberg:

So that’s obviously already a coupling.

Kogelnik:

It’s already what I call mode matching. I had to match the beam that came out of the laser to the beam that the structure wanted.

Bromberg:

Ok. So that’s a real missing link in terms of…

Kogelnik:

Well, it really is this paper here, “Matching of optical modes.” It just says, ok, you have a laser with a beam size w1, at that location, and you want a laser beam with a beam size w2 at some other location. What do you do? How do you do that? And that gives the prescription, you know use a — example and theory and you adjust…

Bromberg:

… particular thing that you had to work out, I mean, any particular part of the analysis that wasn’t just easy, or which had to be invented by you? What was the point of novelty is really what I am asking?

Kogelnik:

Well, the point of novelty in this is this law here which is sort of equivalent of the imaging law. But it includes the diffraction of the beam. You know imaging laws don’t include any diffraction, just ray tracing. And you know you can say, ok, image this thing from here to here.

Bromberg:

Yes.

Kogelnik:

But that doesn’t give you the right beams, because the beams diffract because of diffraction. So it is an extension of the classical imaging laws in a way, for laser beams to include diffraction, to give you a wanted beam from a beam that you have.

Bromberg:

And what happened to this particular thing? Did that just then become part of a laser way of dealing with things?

Kogelnik:

This is now how to deal with lasers. These formulas are now in every handbook and so on.

Bromberg:

Born and Wolf doesn’t do this kind of thing? Or they never had to worry about this?

Kogelnik:

No, they don’t know about laser beams at all.

Bromberg:

I don’t know enough to know whether there might be other optical instruments where you would have to worry about beams and diffraction.

Kogelnik:

Well, you only need it for laser beams.

Bromberg:

Oh, I see.

Kogelnik:

… for pure light If the light is incoherent and a mess then you use rays. If the light is pure and coherent, then you can focus it much tighter, and it will stay tight for longer distance, then you need to take care of diffraction phenomena, which is what spreads the beam, and then you need to use this

Bromberg:

Was that tough to put together?

Kogelnik:

Well, it took a while. It was… it took a lot of work to get it into a simple usable form. In general principle it was probably always in Maxwell’s equations, but this is already a usable form, so you can practically do something about matching one beam into another.

Bromberg:

Because… there is something I am trying to get at in a very clumsy way. You trained as an engineer, you’re working right on the border between engineering and physics …

Kogelnik:

I’m not only trained as an engineer, though.

Bromberg:

Ok, then that’s… the plasma thing is really a physics training?

Kogelnik:

I have two PhDs, one in EE, one is in physics.

Bromberg:

I guess I am trying to get at how you conceive of your work in terms of these two fields.

Kogelnik:

I keep mixing them, of course. And I like the mixing. I like fundamentals, and they excite me, but I also like to think problems through to an eventual use, so I tend to go further than most physicists, I think.

Bromberg:

So that the concentration...

Kogelnik:

And that inspired some of this work, exactly.

Bromberg:

I mean, the concentration on getting the equations in a usable form, now, I would just say superficially, that that sounds there seems to be an engineering eye.

Kogelnik:

That’s probably right.

Bromberg:

But I don’t know… just fishing here…

Kogelnik:

That’s probably right. But had I not tried to do that, we wouldn’t have discovered that resonators are stable and unstable.

Bromberg:

Which is more of a physics…

Kogelnik:

Because I tried to push it to get it into a usable form, but on the way there were all these obstacles. All of a sudden, these funny equations wouldn’t want to play. Right, I couldn’t express the output in terms of the input, or something like that. Because there was this instability in between, and then it clicked and said, “Ah hah, these things are sometimes unstable.” So the desire to push things further actually forced discoveries. It has really helped. And, you know, the desire to want to match modes produced these laws that look like Newton’s laws, except diffractions included. Newton’s imaging laws are similar looking, but they don’t have the wave lengths inside, and I have the wave lengths inside.

Bromberg:

That’s very illuminating then. And the amplifier stuff in your notebook, somehow I think is illuminating because well, other people might have read this in a more sophisticated way. But I just sort of was thinking in terms of going from a laser into a transmission line, and not thinking about something like an amplifier.

Kogelnik:

Or something else… Oh, we also worked, I think one of the papers in fact is with interferometers, you know injecting into interferometers for measurement purposes. Same thing, you need to know how to put it in.

Bromberg:

And the interferometers were part of measuring how did that problem come up?

Kogelnik:

With the interferometer you measure spectra, with a very high threshold. How do you find out that there is a single frequency, for example, when the frequencies are spaced that close, megahertz or hundreds of megahertz (?) … comes out of an interferometer you get all these laser frequencies out. It is used a lot now too, by the way confocal canning. You have to inject light in a very specific way. In fact, this really was done before this, and you can see an interesting difference. These were also… facing each other, light bouncing around. And you see all these spots? Can you look at the quality of those beams, versus the quality of this one?

Bromberg:

Well, these seem to be sharper is that?

Kogelnik:

Sharper, and much smaller and cleaner. This was pure light already. I already knew how to precisely matching.

Bromberg:

I had better tell the tape recorder, what Dr. Kogelnik is saying is that number 13 was done before number 11 on the publications list. That’s just an aside.

Kogelnik:

Well this is just a huge quality improvement. I mean, you are referencing this one here.

Bromberg:

Yes.

Kogelnik:

This one references this one.

Bromberg:

How did Komfner get into that one? Herriott makes sense on that off access pass, because he was very much interested in making this kind of an interferometer measurer But Komfner?

Kogelnik:

Herriott was obviously working on scanning interferometers, and I think was using flat mirrors first. We must have gotten together figuring out what would happen if you put curved mirrors in. How does Komfner get into that? That particular one I don’t know. What time is this, '64?

Bromberg:

… he had the idea? Perhaps I made a note on how he got in there.

Kogelnik:

Maybe in sitting all in a room.

Bromberg:

These papers I think must have been written about ‘63, maybe the end of ‘63 because it was submitted…

Kogelnik:

… here the experiments. And again, look lots of experiment. This on getting single frequencies. This is the thing that Patel… the three mirrors and all these measurements that ran with it. Rudy Komfner was very geometrically oriented. He may have had some seminal ideas there, but I didn’t write anything down. In ‘62 there is an entry here that says, “Komfner /// amplifier, diffraction loss…” Let me go back now to later on. It would be quite a ways. This is all the experiment to get single frequencies out. I can see that these are… to adjust mirrors, which had to be done. That’s a question. September ‘62. What does Komfner have to do with this one? Where did you say I should look?

Bromberg:

Well, the paper must have been submitted in the fall of ‘63, so it was finished by then… maybe I’ll rephrase the question; it might be easier to…

Kogelnik:

I think Rudy was obviously not doing the experiment.

Bromberg:

Yes, that’s what I thought.

Kogelnik:

Because he was not in the lab with us. So he must have had some key ideas… index maybe it is in there… interferometer, here it is. Yes, that just works out what it should be doing. It doesn’t’ say what Rudy’s role was, but… We have a patent with him too, right? Can we look through the patent list; I think there was a patent too. He must have had a contribution to the multi-pass idea.

Bromberg:

Optical maser amplifier in ‘66.

Kogelnik:

No, no, multi-pass something or other, it must be.

Bromberg:

Well, there are only two on this list, with Komfner.

Kogelnik:

He obviously wasn’t in the experiment so it must have been in the idea state.

Bromberg:

Is there anything else in this whole business, in these mode conversions, mode coupling papers that we ought to think about? I know there was a complex beam parameter, and I wondered whether that was an important part of it, or…

Kogelnik:

Oh, that’s very important, but the jump from here in trying to inject a beam into something else, which needed another beam, which we called mode matching… The jump from here to mode conversion isn’t very far, right? Here you are trying to do it and the other one is analyzing what happens if you don’t do a good job. It’s as simple as that right? And here, clearly, the microwave background comes out because mode conversion from one wave guide to another is a routine thing that microwave people look at.

Bromberg:

And is that in fact how the conversion work was stimulated, by saying “what happens when I can’t couple them, when I can’t match them?”

Kogelnik:

Exactly. This is trying to match, and the other one is what happens when you don’t, and how much happens. It tells you how accurately you have to match. It is a very close tie in of course. The abstract says virtually that too, right? It says there are all kinds of optical structures, like resonators, laser oscillators, Interferometers, transmission lines, etc., whatever you want to use laser beams in. And if a mode emerging from one of those is injected into another one, then you excite the modes of the other one, and then there is coupling to the modes, and so on… much power you transfer from one to the other… It is not phrased in this kind of a practical language that we have just been using, but to a technical person it means the same thing.

Bromberg:

When you go into the conversion problem, where should we look for the important steps, new steps?

Kogelnik:

This is just… First of all, wanting to do it, and then formulating the problem right. Here you get these horrendous integrals as the result of the formulation of the problem and then seeing if there is a way to solve that integral in a way that doesn’t produce and all that complicated result, which happens to be hyper-geometric functions, to most people terrible. But when you get to the low order modes, you get these relatively sample expressions for how much power goes from one mode to the other [?]

Bromberg:

I’ll tell the tape we are talking about paper number 14.

Kogelnik:

So it is a combination of things. Wanting to do it, formulating the problem right, and finding a way through.

Bromberg:

Ok. And making, as you imply I think, liberal use of microwave wave guide precedents.

Kogelnik:

That’s in formulating the problem that this microwave experience comes in. Coupling and conversion coefficients for microwave modes, that kind of stuff we were finding… these integrals and that stuff, that doesn’t show up there… that’s different. That was mostly a conceptual and a mathematical accomplishment, I would say. It is used a lot now, because people started using lasers a lot. Even fiber to fiber; laser to fiber, all these things.

Bromberg:

At that time... Again, I have this question about… Now if we go back then, was there a community of people within the laser world that was considering that kind of thing, or was this pretty much special to you and to your group maybe?

Kogelnik:

Well, this particular one I did for myself, I think, at that time.

Bromberg:

Were you, by the way, very much in touch with European laser physicists, or not in touch? Whom were you seeing, beside the people at Bell? Just to give me an idea of your intellectual world.

Kogelnik:

Who was I seeing besides people at Bell, at that time?

Bromberg:

Well, now we are still in ‘63, ‘62, ‘63,’64. We are in the… In fact, I am really sort of thinking up until the time that you became, I guess, the head of the optical department. I am thinking of that as kind of a unit…

Kogelnik:

That was ’67…

Bromberg:

Yes, I don’t know if I am wrong or not…

Kogelnik:

When I was up in Murray Hill, and interacted with a wide group of people, if you just count my co-authors as a large group… there was obviously more than that.

Bromberg:

Outside of the Bell people, was that chiefly your scientific community, or was there …?

Kogelnik:

We always have lots of visitors at Bell Labs.

Bromberg:

Ah, ha, that’s true, of course.

Kogelnik:

Including Europeans and anybody else, but I think in the beginning it was pretty much a one way street, because nobody else was really doing very much yet.

Bromberg:

That’s really the answer I am looking for.

Kogelnik:

And now, it is not that way. Now we learn a lot from other people. But in the beginning, you know, with Bell Labs making such a big push I wouldn’t be surprised if for the first two years fifty percent of the laser work was in Bell Labs, something like that.

Bromberg:

… this kind of work too?

Kogelnik:

And even more specifically, that kind of work. There were probably people outside trying to make lasers, and they didn’t have time to think further, yet.

Bromberg:

Then the other question is…

Kogelnik:

… there are exceptions, I’m sure.

Bromberg:

The amount of contact, the kinds of contacts people have are very often also a function of their linguistic skills, and I would expect you to have more contact with European scientists than…

Kogelnik:

Yes, there were quite a few coming in; mostly to learn at that time, at the early time.

Bromberg:

You see I want to say, that from my own reading so far, this is the first work of this sort that I’ve been looking at.

Kogelnik:

Probably is, yes.

Bromberg:

And so I wanted to find out…

Kogelnik:

There was a little, again in Fort Monmouth. Goubau and Schwering got into that somehow at Fort Monmouth.

Bromberg:

Now you are not ever working on contract? You are always working on company funds…

Kogelnik:

No, I never saw Goubau or Schwering. In fact, the two had a little, I don’t know, conflict maybe with [???] and Lee. I only know this from third-, or fifth-hand or something. That [???] thought that maybe these were his ideas, or something like that… think there was a working contact at all. Certainly I never saw Goubau. I saw lots of people in Bell Labs, and American university people had a good interaction with us. Stanford in particular… Siegmund I think was on the scene very early… very early there sometime.

Bromberg:

Of course, Whinnery came I guess at some point close in here.

Kogelnik:

Whinnery spent a whole year with us, that’s right, and I am sure there are others.

Bromberg:

Good. That gives a picture of Bell Labs as including the universe. I mean, the people who come…

Kogelnik:

Well, in the early days Bell Labs represented a large part of the effort in quantum electronics, I think. But there were all these conferences. We went a lot to these Device Research Conferences in particular and others too. But lots of people coming in.

Bromberg:

Ok. If you think of any people who were especially important in terms of your own work…

Kogelnik:

Directly. Siegman is probably the first one I would think of. Hard to figure out any specific ideas. I think we both had in parallel the same idea on one of those early things that I never accomplished. He did put an acknowledgement to the detection of light business. Something like that. And I know I have had contact with Tony for a long time; Bruce (?) I knew even from Vienna.

Bromberg:

There are also people who are right on the border between engineering and physics.

Kogelnik:

Who all moved into lasers?

Bromberg:

They are all very much with two feet very firmly planted in both.

Kogelnik:

Yes. I am sure there are many others. It depends on what time you are talking about. It would tend to be — if you wanted to go outside Bell Labs — it would tend to be US University people first, and there it would tend to be Cal Tech, Stanford, Berkeley, MIT.

Bromberg:

As for example to Hughes aircraft, or …

Kogelnik:

Well, you know, these are competing industries.

Bromberg:

Because, for example, I think of someone like Eugene Gordon as being very much in touch with places like Hughes, and Spectra Physics, and so on.

Kogelnik:

But that’s later, much later.

Bromberg:

Well, maybe ’63-’64.

Kogelnik:

Spectra physics didn’t exist…

Bromberg:

Well, I don’t believe Gordon was in touch with them at that early a point.

Kogelnik:

I did have discussions with Blume from Spectra Physics. Yes, he worked sort of in that direction, as I did yes, that’s true. A long time ago.

Bromberg:

Yes, I know. It’s kind of unfair, me to be asking all these things.

Kogelnik:

But I don’t think I ever wrote a paper.

Bromberg:

The easiest way to do that is to look at the publications list, I would guess.

Kogelnik:

I really don’t think so.

Bromberg:

It’s not very often that people do write a paper out of that.

Kogelnik:

It’s complicated.

Bromberg:

Some people do. I know that Gordon wrote that one with Bridges, Bill Bridges, but that was really a different kind of thing.

Kogelnik:

Yes. That was really trying to make a CW-argon laser.

Bromberg:

What about the paper with Yariv?

Kogelnik:

You already saw that I was trying to make amplifiers. The most important thing of an amplifier, as you know from electrical engineering, is its noise. You need to figure out how to suppress the noise as much as possible that comes out of the laser. And the ultimate is to make the hole smaller and smaller, then less noise will come out. But then, the laser beam can’t come out either. And to find out the right compromise, to minimize the noise, is what these… kind of noise suppression structures. With a screen with a hole in it. We call that an aperture, with a lens and another screen. Very fundamental laws on what to do to minimize… A by-product of this is this one curve here. It is something, unfortunately, again that is hidden in this paper. Ask yourself the question, if you have an opening on one end and an opening some distance away, what do you do to be able to transmit the maximum amount of light from one opening to another? The problem comes up over and over again. It comes up in Starwars, I’m sure, because you have to figure out how big a laser beam you have to make where you send it off in order to get most of the light that is sent off onto your target, which is a certain size, a certain distance away; a very fundamental principle. You can work out that optimum. The optimum of the fraction of the light you can hit is plotted here.

Bromberg:

That is figure 7 in the …

Kogelnik:

It is used in the context of noise here, but it really has much more general application.

Bromberg:

I think let’s just stop for a moment …

Kogelnik:

We touched on number 80 on. Schultz, De Gras, waveguide techniques to masers. I knew Schultz DuBois, I didn’t know he was applying wave guide techniques to masers. I did not have much interaction. I did have a lot of contact with Schultz DuBois, in general terms. We talked about the Rigrod laser. That we did... lasers are related to communications. “Tell us about the work that went on and the thinking that was being done on optical communications.” Well, I told you basically that is why I think Rudy Komfner pushed lasers.

Bromberg:

Yes, you answered a lot of this question. In fact … And number 2 is really just a very general thing, which I don’t ever think you have to…

Kogelnik:

... “Each paper is motivated quite differently.” Management interest was always great. And they were funded simply by having me there. We picked our own projects, really, in detail. “What were the major breakthroughs?” Well, we went through some of those, bit by bit. “work on mode conversion, mode coupling originate.” We went on that one. It came through the mode matching to use lasers in general. “… and Komfner’s roles in research in non-focal interferometers.” Well, Herriott had the scanning interferometer, that he wanted to apply. Komfner’s were ideas. “The significance of the complex beam parameter.” That’s… Well, I still think privately that my biggest accomplishment is the ABCD law, which was quite imaging of optical modes… allows you to translate what you know about tracing rays, ordinary rays, through any optical system. Translate that into laser beam language, and a matrix that you use for tracing rays has four coefficients in it. They are called A, B, C, and D.… get down a law for laser beams that’s not a matrix law, but a… (???) fraction. Well, anyway, Q-out equals this, the complex beam parameter that you asked about. That contains in it both the size of the beam, and the curvature of the wave front.

Bromberg:

And that parameter, then, is necessary to make this translation from the…

Kogelnik:

You have a complicated, messy system. If you manage to trace rays, you need a ray-tracing matrix, which is ABCD. Input ray position, x-in, and the angle at which the ray goes, at the input and the output position of the ray out (x-prime). The angle of the output beam would be given by this matrix formula in very simple terms. It is a very complicated box. You don’t know what is in it. Here is a ray at a position x from the axis, x-in, and it goes at some sort of an angle, which is x-prime, out at a certain place at a certain angle, and you can get that from that matrix. That is well-known. And for laser beams, if you know these coefficients, these four, you can find out what the beam that comes out looks like. That’s a very fundamental, very simple law, which ties geometrical optics to fractionate laser beam optics.

Bromberg:

How did you...

Kogelnik:

… like simple, simple, simple laws, complicated rules and laws. And all of a sudden it clicked. So I must have been working in the vicinity for a long time. Mode matching is, you know, is a special case. Almost everything I did was a special case. But that was a very general… that came out.

Bromberg:

So it somehow had rays and diffraction in it and it comes out — am I wrong — it comes out of the same melding of ray analysis with diffraction that you need for lasers. Or is this?

Kogelnik:

Rays are much simpler, right. But to me you can look at laser beams in three steps of knowledge. The most primitive one is rays. You just draw rays through some optical system, and then you know something already. The next step is to do it in terms of beams. Then you have to ask, what’s the beam size? How do they expand? What is the radius of curvature? And the item at the third level is to write diffraction integrals down for things. Now this is the second level of knowledge basically, just beam language. Just asking how fat is the beam? And what’s the wave front curvature at each place. It’s amazing that you can get to the higher level of knowledge with this simple transformation; very exciting to me, in fact, that this is possible.

Bromberg:

Now when you say it, the way you said it made me think this was something that wasn’t picked up as quickly as you might have expected. Is that true? You said this is the most…

Kogelnik:

Well, the publication of that is hidden in this fat paper. It surely is known now.

Bromberg:

That’s number 18, “The imaging of optical [???].”

Kogelnik:

And anybody seeing this would first of all put it aside for better times, before they can read it.

Bromberg:

Is that true? Is that the way physicists work?

Kogelnik:

Well, they always read the fat papers later. And there is all kinds of things in here, on various levels, even diffraction theory and so on. But the ABCD law is somewhere in there too.

Bromberg:

That’s quite the reverse of the way historians operate, they’ll always read a book before an article.

Kogelnik:

Here it is, page 80 I mean, equation 80. I didn’t write that right. It is a little sub-section. I did this as if I was writing a book, I guess. It tells you many beautiful things about laser beams. But that may have well been the key thing. I worked on this for a long time. Has graded index media in there too.

Bromberg:

Was that actively being talked about at the time?

Kogelnik:

No, later. But the analysis is in here.

Bromberg:

I mean, were the transmission people already talking about that kind of thing, or is this?

Kogelnik:

No, the way it first came about was that in some lasers the gain was a function of the radius, and therefore the index, and there was some small effect seen in lasers. There is a reference to that in there. Then I thought I analyzed it, and then Beriman came into my office one day and he wondered… Then we knew that sequences of lenses had reflection losses that were small but significant if you put lots of them next to each other. And [???] discussions [???].

Bromberg:

I didn’t realize you were involved in that, at that inaugural stage of that.

Kogelnik:

Well, Berman invented it in my lab. We were talking, and I knew that you could do gradual focusing from looking at these other things. But he knew how to do it. He said, can’t we do something else with lenses? When I must have said, well probably if you can do some gradual thing, and then he came up with this gas lens. I’m pretty sure that Dwight Beriman invented the gas lenses either in my lab or walking out of it, or something of that nature. That was well known by the people around there. He clearly invented it, but his office was right next to mine in Murray Hill.

Bromberg:

Ok, well I think we have reached a good point. We are sort of in around ‘64-’65, and we might want to do the next one mostly on holography and what happened.

Kogelnik:

Holography, actually, is easy. I don’t know how much time we have.

Bromberg:

You have 5 minutes.

Kogelnik:

Holography is easy. You can do that in 5 minutes, because it was really just another attempt at applying lasers, to me. I was the only one in our group and it intrigued me and I played around with it for a little while. That was holography.

Bromberg:

But there was a lot of Bell work on holography.

Kogelnik:

Yes, but that wasn’t in my group, in our group. There was a group in Development under Collier, Bob Collier, they were doing holography work and I interacted with them. I told them all the ideas I had, and I interacted with them …group in the electronics research lab, I was all by myself, and I did it for a while. It was intriguing, and did some fun experiments. Again, experiments. That experiment I am proud of. Shining light through this bathroom window and projecting a clear text through it.

Bromberg:

You used a bathroom window?

Kogelnik:

Because it was fun. Normally, you see clearly only through an ordinary flat window, right?

Bromberg:

I see. Ok.

Kogelnik:

But that, by many people is regarded as the beginning of conjugate, or phase conjugation. You reverse. If you go through a messy medium the image gets totally distorted. But if you can reverse the phase of the incoming ray, it will go back the way it came through the first time.

Bromberg:

That’s this paper here, number 21?

Kogelnik:

Yes. That was a fun experimental paper, based on a fun theoretical too, of basically of time reversal. That the wave will go back the way it came, one way, going back the other way. So I made it come from a clear text, and it went through this messy bathroom window, with lots of [???] Coca Cola bottle, or whatever it was, and… hologram here. With a hologram you can do this phase reversal. Then I took the text away, and I have this bathroom window here and a screen here, and I have the hologram on the other side of the bathroom window, and any visitor who came in that wanted to see it, I said, Look here is the bathroom window, here is the screen. All of a sudden there was a text appearing here. Nobody knew where it was coming from. And was using phase conjugation. And of course if you didn’t have the right bathroom window, you couldn’t read the text at all.

Bromberg:

You had to have the same one that you, the original…

Kogelnik:

It had to be the same bathroom window. People said, Ah Ha, now you can code, right. You send the hologram; the other guy has the bathroom window… the scrambled text.

Bromberg:

Did anybody ever code that way?

Kogelnik:

No, I don’t think so. But you can also figure that the bathroom window is the atmosphere, it scrambles light around. And if you ever want to get to a… even though there is a scrambling medium in between you can make all the light go to this place. That’s what I did. I unscrambled it in the hologram by phase reversal, by phase conjugation.

Bromberg:

Well, I don’t want to waste our time now with physics, with physical explanations. Let me just say, this is where we will take up next time, and I very much appreciate this.