Oral-History:Herbert Kroemer
About Herbert Kroemer
Herbert Kroemer was born 25 August 1928 in Weimar, Germany, and studied at the Universities of Jena and Göttingen. During the Soviet blockade of Berlin in 1948, he worked a summer job at Siemens, and escaped the Soviet Zone to go to the West. After obtaining his PhD, Kroemer worked at house physicist at Central Communications Lab, where he got the idea for heterostructure bipolar transistors. After spending three years at RCA labs, he returned to Germany to head Phillips Semiconductor Group, working on gallium arsenide technology in 1957. Returning to the United States, he worked at Varian Associates, and studied the Gunn effect. In 1968 went to University of Colorado, and in 1975 to University of California Santa Barbara. He was awarded the Nobel Prize in Physics in 2000. In this interview, Kroemer describes the effects of World War II and post-World War II German politics on education and scientific research. He analyzes the roles of theory, experimentation, and knowledge in scientific discovery and in education. Kroemer narrates the stages of his industrial and academic careers and traces his research interests.
About the Interview
HERBERT KROEMER: An Interview Conducted by John Vardalas, IEEE History Center, 12 February 2003
Interview #424 for the IEEE History Center, The Institute of Electrical and Electronics Engineers, Inc.
Copyright Statement
This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.
Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, IEEE History Center, 445 Hoes Lane, Piscataway, NJ 08854 USA or ieee-history@ieee.org. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.
It is recommended that this oral history be cited as follows:
Herbert Kroemer, an oral history conducted in 2003 by John Vardalas, IEEE History Center, Piscataway, NJ, USA.
Interview
INTERVIEWEE: Dr. Herbert Kroemer, 2000 Nobel Physics Laureate
INTERVIEWER: Dr. John Vardalas
DATE: 12 February 2003
PLACE: Dr. Kroemer's office at the University of California at Santa Barbara
Childhood, family, and educational background
Vardalas:
It's quite an honor to interview a man of your distinction. I suppose the natural place to start is at the beginning. When were you born and where?
Kroemer:
I was born on August 25, 1928 in Weimar, Germany.
Vardalas:
Did you have any brothers or sisters?
Kroemer:
I had two younger brothers.
Vardalas:
Did you spend your formative years in Weimar or in that general area?
Kroemer:
I lived in Weimer through high school, and then my first university after high school was the University of Jena just 15 miles from my hometown. Then after one year – I think it was 1948 – I went to Göttingen in West Germany. Weimar was in what was then East Germany.
Vardalas:
You grew up during a period of great economic and political turmoil in Germany and Europe. What are your strongest memories during this period?
Kroemer:
After 1945 Weimar was occupied by Soviet troops. It had originally been taken over by the Americans. Then after June or July when Germany was divided that area was turned over to Soviet troops. I remember a time when the Americans were leaving and the Soviets coming in as a pretty bad time, because the Soviets tried to enforce a type of government there that was very, very repressive. It was really a Stalinist operation, and Weimar was actually a very liberal town.
Vardalas:
Was it?
Kroemer:
Yes, and it therefore drew their specific attention. It was also pretty scary at times when I studied physics at the University of Jena for two semesters during 1947/48.
Vardalas:
Yes, I wanted to get to that.
Kroemer:
The student body at Jena was basically very liberal, as was the faculty, so they really cracked down. I remember one day in a philosophy class the professor made some comments that were clearly in disagreement with the communist party line that got a big applause. When class was over someone grabbed hold of me at the exit and said, "We'll keep track of people like you who applaud."
Vardalas:
Really?
Kroemer:
It was a really threatening kind of atmosphere.
Vardalas:
Did you find yourself being interested in social and political issues at the time?
Kroemer:
One couldn't help but be interested in social and political issues, but not on the officially approved side.
Vardalas:
Yes, of course.
Kroemer:
The summer of 1948 was the Berlin blockade.
Vardalas:
Yes.
Kroemer:
Four of us went as summer students to the Siemens Electrical Laboratories in West Berlin that year. At that point the blockade was pretty much in force and the airlift was in force, so I then decided I was not going to go back home and left for West Germany.
Vardalas:
I would like to go deeper into that. Before I do that I want to find out something about your education. Were you in high school in the midst of the war?
Kroemer:
Nowadays it is called a gymnasium. I went through four years of elementary school. Then in the German educational system as it existed at the time students got split up at the 5th grade.
Vardalas:
What happens?
Kroemer:
About 15 percent go on to "oberschule" or gymnasium. That was intended to ultimately lead to a degree called abitur, which is required for admission to German universities.
Vardalas:
I see.
Kroemer:
It was really a pre-academic education. The remaining majority of people went on in elementary school for another four years. Then after eight years they went to some kind of a trade school – business school or something like that. But I moved on to the German-style high school or gymnasium.
Vardalas:
Was this still during the war years that you were in the gymnasium?
Kroemer:
Yes, it was during the war.
Vardalas:
How did the war affect the level of education and what was being offered at school?
Kroemer:
Many teachers of course had been drafted, so older teachers who had already officially retired were called back into service. And some subjects were dropped. For example the third foreign language was dropped. We started foreign language in 5th grade, which was the first high school year. We started with English and two years later they added Latin – which I hated. Normally French would have been added two years after that, but that never happened. And a few other things were dropped. However science and the first and second language courses continued. English continued.
Vardalas:
English continued?
Kroemer:
Yes, at full blast.
Vardalas:
How interesting.
Kroemer:
We had a very interesting instructor in English. He had spent all of his life as a German language and German literature instructor at one of the two big English universities. I do not remember whether it was Oxford or Cambridge, but it was one of those two. He had retired just before the war and came back to Germany. He was immediately called back into service as a teacher of English. And he had a wonderful knowledge not only of English literature but also of the English parliamentary systems. There was never any commentary with the reading material that we got. He would simply say, "Well, they do things differently." He was really subversive, and he got away with it.
Vardalas:
He got away with it.
Kroemer:
And after the war, under Soviet Occupation, our Latin teacher was also very interesting, though in a different sense. We were reading Caesar's Gallic War, and we didn’t go through the whole darn thing beginning to end. He specifically picked sections from the book that indicated how the Romans as occupation troops were dealing with people who were occupied. And so he said, "You see ladies and gentleman, there is nothing new in this world."
Vardalas:
And the students understood that.
Kroemer:
We all understood that. And he got away with it actually.
Vardalas:
Really? Okay.
Kroemer:
This was the high school Latin teacher.
Vardalas:
Yes. I gather from reading your autobiographical sketch that you exhibited quite a love of learning from an early age.
Kroemer:
Yes. And it was a good environment for that.
Vardalas:
The school was a good environment?
Kroemer:
Yes. Teachers come in different flavors, but there were some that were very good. Learning was encouraged in the fields that interested me.
Vardalas:
I understand that your early inclinations ran towards math, physics and chemistry.
Kroemer:
That's right.
Vardalas:
At the same time, your parents had little education and knew very little about the world of science as you knew it.
Kroemer:
That's right, yes.
Vardalas:
How were you first drawn into science? Do you recall? Was it something you read? What first piqued your interest?
Kroemer:
I couldn't really answer that. I don't know. There was nothing specific I can point to. It happened fairly naturally.
Vardalas:
Could you say that you grew to like science and math because it came easily to you?
Kroemer:
It certainly did come easily to me. I was surprised when one day I discovered that math is actually easy. And I couldn't understand why other people found it hard. But I never really seriously thought about making a career out of math. It was a discipline that I enjoyed and was good at, and I was good enough that I actually was bored in the classes and often disrupted the proceedings.
Vardalas:
Yes, I want to get to that too.
Kroemer:
For my career choice it was a sort of toss-up between chemistry and physics, and eventually physics won out.
Vardalas:
At that age in high school or gymnasium – which was before university – how did you express these inclinations? Did you go beyond formal subjects you learned in school? What were your interests in science?
Kroemer:
I did a lot of reading on my own. We had a good library in the city and I borrowed books voraciously. Typically I would go twice a week and return with two new books each time.
Vardalas:
Were these science books?
Kroemer:
Typically they were science, though not science in a hard textbook style. It was science in the Scientific American style but in book form. I remember some of those books quite well. I was genuinely interested. As far as physics was concerned, we got a textbook and I went through that textbook on my own. There was too much memorization. I didn't like that.
Vardalas:
I guess that's why you didn't like Latin either then.
Kroemer:
Oh boy. Yes.
Vardalas:
Did anyone guide you in your readings or were you just on your own?
Kroemer:
I had no guidance of any kind – none from my parents and not from anyone else.
Vardalas:
Was this a matter of you finding your own way?
Kroemer:
Yes.
Vardalas:
I see.
Kroemer:
Nor did I have any classmates that I was interacting with particularly closely in this field.
Vardalas:
Really? There was no sense of sharing your interests?
Kroemer:
That's right.
Vardalas:
It must have been quite lonely.
Kroemer:
It wasn't lonely. It was alone but not lonely.
Vardalas:
Okay. I understand.
Kroemer:
Those are two different things.
Vardalas:
Yes. In reading the Scientific American style kind of material, what did you view or imagine physics to be all about at this point in time? And why was it attracting you? What was being done?
Kroemer:
Well, it's hard to pin down.
Vardalas:
Did you feel a sense of discovery and adventure in this?
Kroemer:
Not really. It was more the realization that from a small set of very fundamental laws one could draw very, very far-reaching conclusions.
Vardalas:
And this intrigued you.
Kroemer:
I was puzzled about the textbook, why the textbook didn't put the law of conservation of energy on page one and then use this as the hook on which everything gets hung. Today I know about this process, but that gives you an idea of my response. I did have one individual with whom I interacted. It was a cousin of mine. He was a few years older than me, and I spent a few summers with him at his parents' place. And he was a very, very good student at his school, and I asked him all sorts of questions – typically chemistry questions. With infinite patience he gave me all sorts of detailed explanations on how these things were done. I had an encyclopedia and I looked up all sorts of diagrams and equipment. That gave me sort of the illusion that I knew what was going on. My father had bought that encyclopedia, which was something like twenty volumes.
Vardalas:
Why had he bought the encyclopedia?
Kroemer:
They had bought that specifically for me to read. They didn't say, "Read the encyclopedia." It was simply there, and I started poking around. I enjoyed it and spent hours looking up things in the encyclopedia.
Vardalas:
That was very perceptive of them.
Kroemer:
Yes. Both of my parents felt very, very strongly that I should get the best education that I could handle. If I had done poorly in school, they would probably have taken me out rather than saying, "You have to do this."
Vardalas:
I see. Did you siblings have similar interests or did they go in different directions than you?
Kroemer:
The second one had a similar interest. I know very little about the third one. He grew up in the post-war years, born in '41. I do not really know how he responded in school. He never went to university, but that may have been for political reasons.
Vardalas:
Did your other brother become a professional?
Kroemer:
My other brother studied physics.
Vardalas:
Oh, he studied physics too?
Kroemer:
Yes.
Vardalas:
Did you inspire him?
Kroemer:
I don't know whether I inspired him or whether my mother inspired him that the firstborn had obviously done well so the second born should follow in his footsteps. I think he would have made a much better engineer than physicist. He would be a very, very good engineer. He's retired.
Vardalas:
Okay. I see. You graduated from gymnasium in 1947 at the age of 19.
Kroemer:
Yes.
Vardalas:
I want to get to this question of your boredom. You made a point of saying how you found school so boring that you became disruptive.
Kroemer:
Yes, particularly in math. I found it boring because I already knew that stuff. I was way ahead of it. I had done a lot of reading. How does a 17- or 18-year-old act when he is bored? He is trying to draw attention to himself so he becomes disruptive and makes smart-alecky remarks.
Vardalas:
You said you kept showing off. How did you show off?
Kroemer:
By speaking up when others were supposed to answer and that sort of thing. One way I showed off that was in an indirect way was that I taught some classmates mathematical techniques that were not part of the curriculum. They didn't really understand what was going on, but they were sort of cookbook recipes.
Vardalas:
Was that seen in a bad light?
Kroemer:
- Audio File
- MP3 Audio
(424 - kroemer - clip 1.mp3)
The teacher, Willibald Wimmer, was a great man. I later met his two daughters when I visited my hometown. Anyway, he had been an instructor at a junior engineering college so he was used to a somewhat more mature group of students. After the war that college did not reopen, so he became a teacher in math and physics at our school with his principal field being mathematics. He was a good teacher, no doubt about it. It is not uncommon for math and physics to be taught by the same teacher. I'll come back to his physics aspect, but he was very good in math. He treated us as adults, and certainly all of us respected that, so we liked the man very much. He was not flamboyant or anything, but he treated us as if he expected us to be reasonable.
Anyway, he found my behavior a nuisance. He knew that I knew this stuff. He couldn't really tell me not to come to class because I had to be in attendance. Therefore he suggested to me that, "Kroemer, I'll promise you an A in math and you don't have to do the homework — under one condition: you shut up. You have to be here, but you can do anything you want to except you must be quiet and not speak up in class unless asked." I thought it was a good deal. And that's where it stayed. I did my English homework during the math class typically.
Vardalas:
Do you ever apply that technique to your students?
Kroemer:
I've been tempted at times. But it was very different in his physics class.
Vardalas:
In what way?
Kroemer:
It was different in the sense that he basically knew what was in the textbook but nothing else. I had done voracious reading on physics. I knew what was publicly available in nuclear physics and all sorts of things.
Vardalas:
Really?
Kroemer:
A lesser individual would have resented that. And there was another student named Klaus Meyer who was similarly very much interested and knowledgeable in physics. He [Wimmer] simply enlisted the two of us to help him. One of the forms that this help took was in digging through the collection of physics equipment that had somehow survived the war. During the war our school building had been used as a hospital, so much of the equipment was in bad shape and/or not cataloged. We helped him with that and also with preparing lectures. In one case he said, "Why don't you give the next lecture?"
I remember what it was. There was a thermos bottle with a known amount of water and there was a resistor and we applied a voltage. We knew the voltage and the current, so we knew how many watts had to dissipate, and we were measuring the rise in temperature. Experiments along those lines. I don't remember all the details. He was sitting in the front row and having a ball. I enjoyed that.
Vardalas:
You enjoyed that?
Kroemer:
I enjoyed that.
Vardalas:
Was it the teaching or having attention and being in front?
Kroemer:
I enjoyed all of it, the way he was treating me and others, and my own ability to teach – which is a form of showing off. That was an interesting experience.
Vardalas:
This experience must have also fueled your motivation to keep reading more and staying on top.
Kroemer:
Yes, it certainly was positive feedback By that time I had pretty much decided to go into physics rather than chemistry.
Career plans and studies at the University of Jena
Vardalas:
Okay. You write that when you told your father that you wanted to study physics at the university he wasn't sure what that was about.
Kroemer:
Yes. He asked, "What is that?" You see, he didn't have a high school education. He didn't know what physics was.
Vardalas:
And he was concerned about whether you could earn a living at it.
Kroemer:
That's right.
Vardalas:
And you replied – and I quote – "I certainly could become a physics teacher at a gymnasium, a fairly respectable profession." Was this an answer you gave your father to ease his mind or was this something you actually contemplated as a possibility?
Kroemer:
That was more than just easing his mind. That was certainly the most natural end of studying physics. Remember this was 1947 or so. The idea that there were such things as industrial physicists was not all that obvious. And I certainly had not thought about an academic career at that point in time. But it wasn't really something I was explicitly planning. I simply felt that was an option I would definitely have. I knew that there might be other options but was not terribly concerned about what those were.
Vardalas:
As a high school student did you ever have dreams of great discoveries? Did you think, "I'm going to do that some day"?
Kroemer:
Not as a high school student, but as a student at the university I certainly had some dreams.
Vardalas:
Those dreams had not been formulated during high school?
Kroemer:
I do not recall dreams of great discoveries as a high school student. I certainly did not have any crackpot ideas.
Vardalas:
What would be an example of a crackpot idea?
Kroemer:
High school students often have crackpot ideas.
Vardalas:
Such as building a time machine?
Kroemer:
Yes, or things that violate known laws – either because they don't know the laws or they feel they can violate them. I wasn't that type.
Vardalas:
You stayed at Jena for one year?
Kroemer:
Yes.
Vardalas:
Did your understanding, perspective or expectations of physics change during that year?
Kroemer:
That year had a great deal of influence on me, yes, because I was suddenly confronted with teachers who were really at the top level of their profession – certainly in math and physics. I don't remember any chemistry at Jena though I do remember some chemistry at Göttingen. I found one of the mathematicians at Jena very inspiring. His name was Brödel.
Vardalas:
In what way was he inspiring?
Kroemer:
He was an absolutely fabulous lecturer. I must admit that I didn't understand everything.
Vardalas:
That's the first time you weren't bored.
Kroemer:
It was the first time I wasn't bored but challenged. One of the reasons I was challenged was that in the first and second semesters I took courses intended for third and fourth semesters. I simply didn't bother with the first -semester courses.
Vardalas:
You had that option? They didn't force you to take the first as prerequisites?
Kroemer:
No, we could take anything we wanted. In physics, one of the professors was a gentleman by the name of Friedrich Hund. He was a spectacular lecturer and he had a great influence on my becoming interested in the deep fundamental principals of physics.
Vardalas:
Why was this?
Kroemer:
He is pretty close to the top of my list of people who should have gotten the Nobel Prize but never did.
Vardalas:
What was his area of expertise?
Kroemer:
He was one of the leaders —though not one of the founders — of quantum mechanics. There is something called "Hund's rule" which plays an important role in atomic physics even today. He was a wonderful teacher in whatever he touched, including quantum mechanics and thermodynamics. I took a thermodynamics course with him. And he was a wonderful person.
Vardalas:
You took this course with him your first year?
Kroemer:
That was during that one year when I was at Jena.
Vardalas:
Do you recall your curriculum at that time? What did you take?
Kroemer:
I don't remember.
Vardalas:
You took thermodynamics and quantum mechanics?
Kroemer:
I didn't take quantum mechanics in the first semester. That came later. I really didn't have the background for that. In physics I probably took a classical mechanics course the first semester. That wasn't terribly inspiring. And it was probably in the second semester that I took the thermodynamics course from Hund.
Vardalas:
In the German system did they have what would be called auxiliary courses here in the United States, such as humanities? Did they make you take any courses besides those that related directly to your major?
Kroemer:
We could take anything we wanted.
Vardalas:
Anything you wanted?
Kroemer:
Yes. To get a degree one had to show certain pieces of paper. I was very much interested in taking courses outside of physics – particularly philosophy.
Vardalas:
You did? Interesting.
Kroemer:
I took a heavy load of philosophy. That's how I got into that trouble with that communist functionary in that class.
Vardalas:
Oh really?
Kroemer:
I took a heavy load of philosophy. I don't remember what the different topics were, but more than one course. I remember two of the instructors. One taught metaphysics. He was the one we applauded.
Vardalas:
Oh, that was the one you got in trouble for applauding.
Kroemer:
Yes, and lots of other things. Another instructor was a philosopher named Max Bense, who taught formal logic. He was fascinating. I have never known anybody who could construct sentences so long that were grammatically correct to the end.
Vardalas:
That is also a particular peculiarity of German, isn't it?
Kroemer:
Yes.
Vardalas:
How long did you have to wait for the verb?
Kroemer:
No, no. This is really not true that the verbs are always at the end. That really is bad German. It's allowed, but not a good practice. That is not what I meant.
Vardalas:
Did you mean clauses inside clauses and so on?
Kroemer:
Yes, yes. And the funny thing was that he said it in a way that it was understandable.
Vardalas:
I see.
Kroemer:
I found him interesting as a person more than I found the course interesting.
Vardalas:
Let me go back for a second. You mention rather brilliant people on the staff and at the University of Jena and you only spent a year there. What do you think the effect of Soviet presence was on that environment?
Kroemer:
It didn't do anything to the scientific environment, certainly not in those days, and I do not think it did much in subsequent years. But actually I wasn't there of course. It did nothing whatsoever to the scientific environment. The influence was in the political environment and the everyday environment. I had fallen in with a group of students who were all very liberal and I had the suspicion I was being noticed. One of the depressing things was that over that year every week one or two people would disappear.
Vardalas:
Really?
Kroemer:
One never knew whether they had left by their own free will or ended up in the gulag.
Vardalas:
I see.
Kroemer:
The gulag in those days meant the uranium mines at the Czech border.
Vardalas:
This must have been an environment of great uncertainty and fear.
Kroemer:
Yes, yes. We could not trust anyone we did not know personally.
Vardalas:
You said your time at Jena was important in your formation and that it had changed you somewhat.
Kroemer:
It really exposed me to physics.
Vardalas:
Did it also expose you to other good students?
Kroemer:
Yes, absolutely. They played an important role.
Vardalas:
Were you considered a bright student at this level yet?
Kroemer:
Probably, yes. Nothing indicated this officially, but I think I was one of the better students.
There is one interesting difference in that education system compared to what we are doing here at UCSB. For example in the calculus course. Homework is an integral part of the course in the United States, but the way this was done there was that there were problem and homework sessions. They were different courses, and sometimes they didn't have too much in common with one another. I knew calculus pretty well already. We had calculus in high school and I had studied it on my own – but not in the rigorous sense in which it gets done at the university where one worries about all sorts of things that could go wrong. The way I learned it one did not worry about things could go wrong. This makes one better qualified as a physicist.
Vardalas:
Yes, I was going to say that.
Kroemer:
That instructor was pretty boring, so I went only to the problem course. I needed that certificate that I had attended that course two semesters in a row. And that one was fascinating. There I found a real challenge and I got one of the top grades in the class. They congratulated me. I remember it was a lady who was teaching that one. When she was handing out the certificates at the end she said, "I hope to see you again" in the sense that I was one of the ones they wanted to see. I enjoyed it.
Vardalas:
I'm curious. With two courses, one solving problems and one for just theory, how were you tested in both? The test would seem to be the problem solving.
Kroemer:
The test is the problem solving, yes. I never went through the mainstream course. First of all I discovered very quickly that I already knew everything that was needed to solve the problems and it required more imagination as to how they could be tackled. That was the good part. Also in the physics courses the problems session typically was separate. In that other course with a mathematician whom I admired there was also much homework.
Vardalas:
At this stage, after one year's exposure, did you start to see a sense of what kind of physics you wanted to do?
Kroemer:
Not really. It was a wide-open field, but it was very clear that I enjoyed physics.
Summer employment at Siemens in West Berlin
Vardalas:
I want to pursue this thing about you winding up in West Berlin. You said four of you had a summer job at Siemens.
Kroemer:
Yes. One of us had arranged that summer job at Siemens.
Vardalas:
At this point there was not much impediment to travel from East Berlin to West Berlin and work?
Kroemer:
No, it was still completely open although this was during the Berlin blockade. The blockade referred to truck traffic, rail traffic and bringing supplies into Berlin, but there were no limitations on people crossing into Berlin or within Berlin and crossing between East Berlin and West Berlin.
Vardalas:
Then you go on to say that at that point you decided you were not going back because of this uncertainty and fear you were experiencing in this environment.
Kroemer:
Berlin during the Berlin blockade was an exhilarating experience.
Vardalas:
In what way?
Kroemer:
First of all the West Berlin government was about as firmly anti-communist as could be, even though the mayor was a former communist. He had turned anti-communist.
Vardalas:
What was his name?
Kroemer:
Ernst Reuter. He was a fascinating man, and I remember going to one of his big speeches. One of the interesting experiences I had was one of the airports which served the airlift – the airlift was functioning pretty well by late summer – was the Tempelhof Airport, which is basically the airport downtown. There is a railroad track along the edge of the airport and that is somewhat elevated, and in the evenings we often took the train to Tempelhof Station just to watch the aircraft take off. It was a fascinating experience.
Vardalas:
What was so great about it?
Kroemer:
It was staggering how well it was organized. The Soviets could have stopped this whole thing with one fighter plane, but they didn't dare. There were two runways – one only for takeoffs and one only for landing – and they were bringing in planes roughly at the rate of one every 75 seconds.
Vardalas:
Wow.
Kroemer:
You could see them along the horizon like pearls on a string. The pilots didn't really know to which gate they had to go, so next to the runway were a number of jeeps and as soon as the plane had slowed down so that a jeep could keep up with it the jeep pulled in front of the plane with a sign, "Follow me."
Vardalas:
Really?
Kroemer:
That jeep driver knew which gate had just been opened.
Vardalas:
Oh, I see.
Kroemer:
That made you believe in Germany and America having something in common.
Vardalas:
The efficiency?
Kroemer:
No. You see, in '47 and '48 it was not clear whether Germany had any future, but it was very clear that if there was a future for Germany it was in alliance with the United States. Amazingly, even though Berlin had been badly bombed out by the Americans and the British there was basically no resentment by the people.
Vardalas:
That surprises me. Was that because the communists were there?
Kroemer:
Yes, well, the contrast was stark.
Studies at the University of Göttingen
Vardalas:
Okay. The four of you were watching these planes coming in and leaving and then the thought came to you that you were not going back?
Kroemer:
I decided I was not going back.
Vardalas:
Were you ever personally frightened? I know the man that stood at the door said, "We're going to watch you." Was it a serious consideration for you that, "Maybe I'll get in trouble if I stay in East Germany"?
Kroemer:
Yes, but nothing specific. Yes. It was more than just vigorously disliking the environment. I was also worried that I might be forced to do something.
Vardalas:
The physics education was good in itself.
Kroemer:
Yes.
Vardalas:
It was other considerations.
Kroemer:
Göttingen was the top university.
Vardalas:
But at that time you didn't know about Göttingen?
Kroemer:
I knew about Göttingen.
Vardalas:
I mean as an option for you. It wasn't a real option yet when you made the decision to leave?
Kroemer:
I had written to several universities in West Germany and applied to each. Göttingen was one of them. Göttingen had actually turned me down, but I never got that mail.
Vardalas:
Fortunate for you.
Kroemer:
I had a recommendation from one of my Jena professors. I had gone back to Weimar to pick up my baggage before leaving by way of Jena. Professor Buchwald suggested, "Why don't you see Professor König and see what he can do for you?" König was an old friend of his.
Vardalas:
I see.
Kroemer:
I was staying Kassel, which is where relatives of mine were living. It is just about a half hour from Göttingen by train.
Vardalas:
Was it a difficult decision for you to head out and leave your family behind? It must have been. Or was it something that you didn't have to anguish over much because you were too worried about staying?
Kroemer:
It wasn't really difficult, but probably one of the reasons it wasn't difficult was because I didn't appreciate how difficult it might turn out to be.
Vardalas:
What did you later find?
Kroemer:
Well, my parents could no longer support me. The currency differential was such that it was totally impossible. I had to find a job. It was an interesting experience.
Vardalas:
After the walls went up and other things that happened you must have been cut off from your family.
Kroemer:
I went back for a visit a few times.
Vardalas:
Weren't you worried about being caught?
Kroemer:
Oh, I was always absolutely terrified.
Vardalas:
Was your father supportive of your leaving?
Kroemer:
He didn't try to talk me out of it. He probably realized that this was perhaps the best thing for me. He didn't actively encourage me. However neither he nor my mother made any attempt to talk me out of it. They said, "All right. If that is what you want to do, then go and do it."
Vardalas:
You also write that you found Göttingen to be a wonderfully stimulating place.
Kroemer:
It was absolutely wonderful.
Vardalas:
Was it the same kind of stimulation you found in Berlin? What was it intellectual stimulation that you found in Göttingen?
Kroemer:
Yes, it was intellectually stimulating. Göttingen was always one of the top universities in physics in Germany, and in fact in science in general. During the twenties and thirties Göttingen, Berlin and München were the leading universities in physics. Göttingen had not been bombed out during the war, and as a result many academic people who were refugees from the East congregated there. There were a few what is now called Max Planck Institutes though they didn't have that name at the time. It was an absolutely fascinating collection, absolutely fabulous.
Vardalas:
Tell me about your process of your getting into Göttingen. You said you were initially rejected.
Kroemer:
I showed up in the office of Professor König and he said, "Admissions are closed and there is nothing I can do," but somehow he decided to take me on a tour anyway. Maybe he just wanted to be nice or maybe he had some afterthoughts. I don't know. I showed up in the office of Professor Richard Becker and had a nice long conversation with him and he asked me many questions. His assistant, Dr. Günther Leibfried, was also there. They spent a long time with me and then passed me on to Professor Wolfgang Paul [Nobel 1989] and Professor Robert Pohl. Gradually it dawned on me that these were not just social conversation. It was in an examination.
Vardalas:
Did you start to worry at that point?
Kroemer:
No. I remember one of the exam questions given to me by Professor Paul. I'll never forget it. He said, "Mr. Kroemer, you know that a mirror interchanges left and right." I said, "Yes." He asked me, "Why then doesn't it interchange top and bottom?"
Vardalas:
What was your answer to that?
Kroemer:
I gave him the answer. The answer is it doesn't invert left and right. That's all he wanted to know. However it must have been very obvious that I had to think about this first. At the end of all of this I was led back to Becker's office and they said that admissions were closed but they had received notification from two people who had been admitted that they were not coming. Therefore they had two openings and that within the next day or two they would decide who would get those two openings. They explained that if I was chosen I would receive notice. That was on a Thursday I think. The next Monday or Tuesday I received a postcard notifying me that I was admitted.
Vardalas:
How did you support yourself at that time?
Kroemer:
That was difficult. I found a job in a local aluminum cooking ware factory. Lots of students where employed there during the night shift.
Vardalas:
What did the students do?
Kroemer:
We operated the equipment and did whatever jobs had to be done. It paid well enough that if one was frugal one could survive on it.
Vardalas:
That must have been tough working night shift and studying days.
Kroemer:
Yes, it was tough. One summer I worked in a coal mine.
Vardalas:
I wanted to ask you about that. What were you doing one kilometer down in a coal mine? It wasn't a neutrino experiment, was it?
Kroemer:
It was not a neutrino experiment. It was a very interesting experience. I don't remember the proper English terminology for what I was doing. Hard coal was being taken out, and of course on the surface of all of this is the dense population of the cities in the Ruhr Valley. In order to keep the surface from collapsing they tried to refill this after the coal had been taken out. It was being refilled with sand typically, and sorts of other debris. I worked with that crew. And it was strenuous work but it paid well. For he first time in my life, I was embedded in an environment as a lone student amongst coal miners. I suddenly realized that these people didn't have as much.
Vardalas:
How did they treat you?
Kroemer:
Very, very nicely. They always assured me, "At least you know that you will get out of this one day. We are going to be stuck here the rest of our lives." I suddenly understood why people like this were voting communist. Not that I agreed with them.
Vardalas:
Right.
Kroemer:
I understood their demoralization and their expectations and their feeling that the political system would not take care of them. That changed later on, but this must have been '48 or '49.
Vardalas:
That was a tough time economically. All of Europe was in ruins in a sense.
Kroemer:
Yes. Now that was during the Marshall Plan year during which time Germany was recovering rather quickly. The coal industry was one of the driving engines of the recovery. It was a very interesting experience.
Vardalas:
Referring back to your autobiographical sketch, you highlighted very briefly the deep influence that Dr. Fritz Sauter had on you. He was your Ph.D. supervisor, wasn't he?
Kroemer:
Both my diploma thesis – which is sort of a master's degree – and Ph.D. supervisor, yes.
Vardalas:
Would you explain more about his influence on you?
Kroemer:
Ours was not a close personal or warm relationship. It was a purely professional relationship.
Vardalas:
You wrote, "Under influences such as these I never developed into a hardcore Theorist with a capital T but became basically a conceptualist."
Kroemer:
Yes.
Vardalas:
What does that mean?
Kroemer:
Let me go back a few steps on that one. One course that was very important for me throughout my years in Göttingen was the seminar in theoretical physics. This was not a regular course but a sort of special topics thing where the instructor handed out material and assigned papers. The students were required to read the papers and then report on them. I found this a very stimulating environment. I had participated in these kinds of courses right from the beginning of my Göttingen days. Sauter led one of those courses. I got some assignments, and one of those assignments led to a master's dissertation.
Vardalas:
What was that assignment?
Kroemer:
- Audio File
- MP3 Audio
(424 - kroemer - clip 2.mp3)
I was already going toward solid-state physics. That assignment was studying certain things that happened in periodic potentials when some of the parameters are changed. An interesting aspect of Sauter's style was that he didn't call on students to report once or twice a week or on any regular schedule. He basically was available when he was needed, but he left the students alone. He was basically watching and forming his own opinions as he watched. I remember one time coming to him with an idea I had, a physical concept. He listened and then said with a tinge of sarcasm, "Well, Mr. Kroemer, that's all very nice, but you ought to be able to formulate this mathematically." At another time I would report to him how I would formulate something mathematically. Obviously mathematics was important. Then with a slightly more sarcastic tone he responded, "Mr. Kroemer, this is just a piece of math. What does it mean physically?" Once one goes through this loop a few times one gets the message. One gets the message that one has to be able to move at ease from one to the other in order to live up to Sauter's standards. That was a deep and formative experience for me. I have later on encountered the same thing with Bill [William] Shockley, whom I knew quite well. Shockley had that same style, moving back and forth.
There is something else that happened too. We had already agreed on the topic for my Ph.D. dissertation while I was still working on my diploma thesis. One day Sauter came into my office and told me to stop working on the diploma thesis and simply write up whatever I already had and submit it. I protested, but he said, "Never mind. Let's move on to the real thing." He was also wonderful in the sense that he didn't believe that degrees should be awarded on the basis having served time.
Vardalas:
Yes. I like that statement.
Kroemer:
First of all he didn't have any money to support his students. Sauter's idea was that he was watching people and to him a degree was to certify that this person could do creative and independent work on a certain level and that as soon as this level had been reached – get out. That was very good for me, because I had to earn a living. I got my Ph.D. a few weeks before my 24th birthday.
Vardalas:
That's remarkable. Five years of total experience between entering and leaving. Was that unusual?
Kroemer:
That was unusual even for Göttingen, yes. I was one of the youngest Ph.D.s in physics after the war. And that would not have happened under any other professor.
Vardalas:
Right.
Kroemer:
The way I ended up with Sauter was different. I had originally attached myself to a person who at that time didn't even have the professorial rank. He was a lecturer and instructor: "Privatdozent" Dr. Hellwege. I also had a very close personal relationship with Hellwege and had signed up for a diploma thesis with him. However he had an awfully long waiting list. Then Sauter offered me the opportunity to get my diploma under him. I talked to Hellwege and he said to me, "Kroemer, take it. You will be finished with him before you can start with me." This is how I ended up a theorist. Sauter was a fabulous mathematician, but to him it was a tool. To him ultimately physics mattered.
Experimentalism, theorism, and conceptualism
Vardalas:
You made the distinction that you were a theorist but not a hardcore theorist.
Kroemer:
I wasn't. I was a conceptualist.
Vardalas:
What does that mean?
Kroemer:
There are theorists that know only theory and are heavily engaged in mathematical formulas, whereas to me the mathematical formulas were never more than a tool. Mathematics always represented something to express physical ideas and the physical ideas were always related to experimental facts even though I was a theorist. And I never was a good experimentalist.
Vardalas:
You were never a good experimentalist?
Kroemer:
No. I think I was good at picking good projects.
Vardalas:
What does it take to be a good experimentalist?
Kroemer:
I don't know. Not having been one I could not say.
Vardalas:
What couldn't you do? You must have come to the conclusion you're not good at it or didn't like it.
Kroemer:
It is not a matter of not liking it. Some people have the touch. They invent their own equipment and build their own equipment. I was better at thinking about what equipment one should build than at actually building it.
Vardalas:
Okay. A conceptualist.
Kroemer:
I was a conceptualist.
Vardalas:
I see. You made the interesting remark that your role model was more Niels Bohr than any other great physicist.
Kroemer:
That's right. Yes.
Vardalas:
What was it about Niels Bohr that attracted you to him as a role model?
Kroemer:
Conceptual depth combined with very simple mathematics. He was an ideas man.
Vardalas:
I don't want to be too simplistic about it, but what about Einstein who did a lot of mathematics and had conceptual ideas too?
Kroemer:
Actually Einstein is really a conceptualist. The mathematics of course came with the general theory of relativity. He got his start from Minkowski on that one. In his later years his work was very, very heavily mathematical. However that didn't really start until 1920s or thereabouts. He was really a conceptualist.
Just for the fun of it, I have been rereading some of Einstein's early papers on statistical thermodynamics. The concept of wave particle duality shows up in Einstein's writing before DeBroglie.
Vardalas:
That's amazing.
Kroemer:
Not in the same formulas.
Einstein and the history of physics; Einstein and the Nobel Prize
Vardalas:
Do you see anything else in Einstein's early papers that strike you after all these years? Are there any surprises?
Kroemer:
Yes. It surprising in the sense of just how brilliant Einstein was, and what a fabulous instinct he had. He anticipated ideas that were not accepted until much later on and helped others get on their way. For example deBroglie’s thesis. I don't remember who his thesis advisor was in France.
Vardalas:
I can't remember either.
Kroemer:
Thesis advisors played a different role in those days. One turned in one's thesis and the thesis advisor was the person who was supposed to judge it. In a way Sauter was like that with me. He simply watched and when he saw a final draft he criticized it.
Anyway, deBroglie’s thesis advisor didn't really know what to do with deBroglie’s thesis so he turned it over to Einstein. The remark that Einstein supposedly made in regard to deBroglie’s thesis was, "He lifts a great veil from the secrets of physics." I am certain that Einstein had anticipated something like this. The history of physics is fascinating to me.
Vardalas:
Yes, it's interesting. And yet Einstein is always attributed to saying that God doesn't play dice. His resistance to quantum mechanics.
Kroemer:
His resistance to quantum mechanics was first of all that he definitely did not like the probabilistic interpretation.
Vardalas:
Yes.
Kroemer:
Probably what disturbed him more than the probabilistic interpretation was what we today call the non-locality of the theory. It contained an aspect of action at a distance. That was probably more disturbing to him. I've done a little bit of reading on what exactly Einstein opposed. There's the famous Einstein-Podolsky-Rosen argument. It was entirely consistent with Einstein’s thinking in the early days. Having been a participant in writing a book on thermodynamics, I have seen that what is today attributed to Boltzmann was really first clarified by Einstein.
Vardalas:
Really?
Kroemer:
Einstein took Boltzmann’s ideas and put some rigor into those concepts.
Vardalas:
Very interesting.
Kroemer:
That is something I discovered only recently. I became interested in Einstein basically while trying to understand more about the history the Nobel Prize. Having received that prize I decided to find out more about what makes that system tick.
Vardalas:
Have you found out what makes it tick?
Kroemer:
I found out a lot, because I've been reading on that. Of course Einstein's Nobel Prize was an extraordinarily controversial thing.
Vardalas:
Was it?
Kroemer:
Yes. He had been nominated a number of times for the theory of relativity, and the people who controlled the physics Nobel Prize at the Royal Academy were absolutely opposed to this one. One of the key members of the physics committee is on record as saying, "Einstein will never get the Nobel Prize." Planck had trouble too. Anyway, this is why Einstein ended up getting the Nobel Prize for the photoelectric effect rather than the theory of relativity.
Vardalas:
Okay.
Kroemer:
The political shenanigans of how this was pulled off have been described rather beautifully in some books. I became interested in exactly what Einstein's role was outside of quantum mechanics and relativity because of his role in thermodynamics.
Vardalas:
Einstein had a role in thermodynamics?
Kroemer:
Absolutely.
Vardalas:
I see. Though I studied Einstein, I was not aware of that. From this kind of exploration you've been doing, do you feel that it is important for physics majors to take a course in the history of physics to understand the development of their profession?
Kroemer:
I don't know how important it is. I certainly do not believe that physics should be taught in the historical order at all. The historic development of the field of physics contains a staggering number of blind alleys.
Vardalas:
Isn't that important for people to understand?
Kroemer:
It is not important for understanding physics. However it is important in a cultural sense. I personally am fascinated by the history of physics – by all the blind alleys and by all the mistakes that were made. And it's a story in human culture. However I am opposed to teaching physics, particularly quantum mechanics, in a historical context.
Vardalas:
But as a course in history?
Kroemer:
Yes
Vardalas:
Would you recommend that physicists also take a course in the history of physics?
Kroemer:
I wouldn't want it to be a required course, but I would highly encourage taking such a course taught by someone who knows what he or she is talking about. Friedrich Hund, who I mentioned earlier, wrote a beautiful book on the history of quantum mechanics. I do not know whether it is available in English translation. I read that one twice, which is something I don't do very often. He actually argued that the natural road to quantum mechanics was through thermodynamics rather than through spectroscopy.
Vardalas:
Through thermodynamics?
Kroemer:
Yes, and I never accepted this thesis of Hund's – until I read the Einstein papers.
Vardalas:
Really? That is the conventional wisdom that the road to quantum mechanics was through spectroscopy.
Kroemer:
Yes. Max Planck's blackbody radiation was an exercise in thermodynamics.
Vardalas:
I see.
Kroemer:
The idea was that electromagnetic energy might be quantized, which Planck himself did not accept initially. That of course blocked his Nobel Prize for many years.
Vardalas:
It did?
Kroemer:
It was obviously ridiculous. All one had to do was look at the diffraction experiment to realize it was a continuous wave. This is the resolution of this one. Einstein was simply the first one who decided to simply ignore the discrepancy.
Preparing students for independent research
Vardalas:
You just mentioned earlier about Sauter's views about degrees – not serving time but proving a certain capability to execute independent work. Do you take this approach with your students in the system that you have here in the United States?
Kroemer:
To the extent I can, yes. I am of course under some constraints. There are certain rigid requirements. I encourage my students to spend at least part of their time working on projects other than their own research projects in order to get broader experience. The overwhelming majority of the Ph.D. dissertations that I have supervised were experimental work.
Vardalas:
Really?
Kroemer:
And that doesn't go quite as fast as a theoretical dissertation.
Vardalas:
That's an interesting contradiction for someone who said he wasn't going to be a good experimentalist.
Kroemer:
I think I was pretty good at identifying projects that are worthwhile, and in a semi-facetious way I would like to add that maybe having been a theorist and thereby not knowing how impossible it was to do those things being proposed has helped in actually getting them done. You see, if I had grown up as an experimentalist I would have had a far greater appreciation of the difficulties of doing the things that I wanted done than I have had as a theorist. This has probably helped me. I do not recommend this as a universal procedure. I remember for example how I got into molecular beam epitaxy.
Vardalas:
How was that?
Kroemer:
There was a technology that I felt was making things possible that had previously not been possible. Therefore I proposed to put gallium phosphide on silicon. And of course if I had been an experimentalist I would never have proposed to put gallium phosphide on silicon because it was too obvious that this would not work and what the difficulties were. Well, we did it.
Vardalas:
I see your point. Sometimes a little ignorance is useful.
Kroemer:
Yes, but you shouldn't count on it.
Educational approach to physics through methodology rather than knowledge
Vardalas:
Luck is good too. Along these lines about an educational approach to what becoming a physicist means, the chief editor of the IEEE Spectrum wrote about you in the 2002 issue of that publication. In it he says, "To this day his [Kroemer's] view of education is that accumulating methodology matters more than accumulating subject matter knowledge."
Kroemer:
Absolutely.
Vardalas:
Would you please explain? In my own mind a methodology is a form of knowledge.
Kroemer:
By accumulating I mean accumulating data facts and details. I think the question is of how one goes about solving a problem. How do you estimate what problems you have to solve on the way? How do you estimate the chance of success? This to me is far more important.
Vardalas:
Can that be learned or is that something one gains from experience? Is that something that can be formally taught or is it something that can only be gained by experience through trial and error?
Kroemer:
I think it can be taught to some extent.
Vardalas:
What procedures do you use to teach this to your students?
Kroemer:
By always insisting that whatever they are doing is not simply done following the recipes, but that they understand the rationale behind it. In this context I have a story to tell. In January of 2001 I was invited as a keynote speaker to a workshop held by the German Ministry of Research and Education at Stanford University to an audience of something like 150 to 200 German Post Docs.
Vardalas:
Were these Post Docs from all fields or just physics?
Kroemer:
From all fields, and under the jurisdiction of Madame Edelgard Buhlmann, who was then and is still now the German Minister of Research and Education. She was very, very much interested in university reform. At this workshop somehow it got mentioned that I had received my Ph.D. before my 24th birthday. One of the Post Docs in the audience burst out, "But you didn't know anything yet at that time." Which was referring to the deplorable tendency of having people study a lot of material rather than concentrate.
Vardalas:
Just for the sake of it you have to know it all.
Kroemer:
Yes. I was tempted to answer, "Well, it didn't stop me from getting the Nobel Prize," but I didn't do that. It would not have been nice. I simply said, "Listen. I had learned how to tackle a problem even if I had no previous background in the details. And I feel that was important."
Vardalas:
That's very interesting. That explains the distinction to me.
Kroemer:
Knowledge of detailed subject matter becomes obsolete. Why would you want to cram your brain with detailed subject matter long before you actually need it? Learn how to find it.
Vardalas:
I see.
Kroemer:
Richard P. Feynman once made a similar comment.
Vardalas:
He did.
Kroemer:
Somewhere in his writing there is a comment that he didn't see much point in reading up on what all the others had done because they had obviously not succeeded. I sympathize. I agree with his point of view. Now if you are interested in the history of the field then of course that becomes a different thing. I do not view the history of science as a tragedy; I view it as a comedy.
Vardalas:
But you know tragedy and comedy are very close together.
Kroemer:
Yes, I know. Maybe personal tragedy with the blind alleys and the narrow-mindedness with which representatives of the established power structure often suppress things that they don't like.
Vardalas:
Do you remember the famous quote from Max Planck, "Theories succeed because their opponents eventually die off"?
Kroemer:
The trouble is, it applies to him too. Yes, I remember this from his autobiography. He was advised against studying physics because there were no problems left.
Vardalas:
Yes.
Kroemer:
Philipp von Jolly was the name of his physics professor at München who advised him that with the discovery of the principal of energy there were no more problems left. All the rest was just working out the details. Fortunately Max Planck went on to pursue physics. He was basically a very conservative man. He struggled for years trying to overcome and undo the revolution he had started.
Vardalas:
He did?
Kroemer:
Oh yes. For years he tried to invent this back into a classical framework – in total contrast to Einstein. The two respected each other tremendously. Einstein seemed to enjoy overthrowing such concepts.
Post-doctoral career prospects in 1952; work at Central Telecommunications Laboratory and theoretical approaches
Vardalas:
I see. In '52 you graduated, at 24 you got your Ph.D., and you wrote that there were not many opportunities for advancement in academia at this point even for a bright theoretician like yourself.
Kroemer:
Yes. Zilch.
Vardalas:
Zilch. Nothing.
Kroemer:
First of all there were no new universities, no new departments being formed, and there was a long waiting list of people with first-rate credentials and more seniority waiting for any openings. It was sort of a dream but I ruled it out.
Vardalas:
But it was something you would have loved to do if you had the opportunity?
Kroemer:
Yes. I would have loved to do it. I simply considered the opportunity to be zero – which was a correct assessment for the day.
Vardalas:
It turned out the other option had ramifications for you. You said that at the Central Telecommunications Laboratory you were a "house theorist."
Kroemer:
That's a verbatim translation from the German. That's a common concept.
Vardalas:
What did this position entail and how much latitude were you given to be house theorist?
Kroemer:
The position entailed to be available to try to answer whatever theoretical questions came up. That was the minimum, but more than that, I was really expected to take an active role.
Vardalas:
To seek the theoretical issues to what they're doing.
Kroemer:
To seek and to try to take an active role in making suggestions. I was encouraged to actively poke my nose into the experimentalists' business.
Vardalas:
How did they react to your presence?
Kroemer:
I was strictly forbidden to touch any equipment, so I was not a competitor. It was a good relationship.
Vardalas:
They valued your input rather than saying, "What do we need this theoretical stuff for?"
Kroemer:
Another one of the things that I was expected to do – every week or every other week – was to give a lecture to anybody who wanted to come and listen on any subject of my choosing. Naturally I picked subjects that I felt were relevant to the work that was going on.
Vardalas:
Was this well received?
Kroemer:
Oh yes. That very often involved me having to learn new things. I had never in my university career seen a metallurgic phase diagram. I discovered very quickly when I tried to understand on what principles these recipes were based that we were using make transistors that I had to learn a little bit of metallurgic phase diagrams. Teaching others is a much better way to learn something, so I thoroughly enjoyed this work and I think I was good at it.
Vardalas:
You wrote that this job was an important landmark in your development as a physicist. You commented, "I ceased to be a real theoretical physicist, if I ever was one." I took that in a positive sense in that you found yourself having to bridge between theory and practice a lot.
Kroemer:
Yes. First of all I was in an environment where I was basically not interacting with other theorists. I was in an environment where I was interacting with experimentalists. I was the theoretical advisor. And of course I was encouraged to follow-up theoretical developments in the literature. I was never qualified as a "real" theoretical physicist, if I ever was one. Whatever that means. I think when it comes to a Professional Theorist with a capital P and T, I am not considered a theorist. And I do not consider this a negative assessment. The important question is not "Are you a theorist or are you an experimentalist?" The important question is, "Are you doing something that is useful? Are you contributing?"
Vardalas:
Let me see if I can provoke you to say something controversial.
Kroemer:
In general it's very easy.
Vardalas:
Do you feel that in some areas of physics there tends to be an overemphasis on theoretical formalism in pursuing theoretical issues?
Kroemer:
I don't know whether it is characteristic of the area. It's probably characteristic of some individuals in all areas. Some areas of course lend themselves more readily to this than others, but I would not want to disparage any single area of physics in particular. And I shouldn't comment on areas that I don't understand anyway. But it certainly is clear that in all areas you find a broad range of people with different interests, from the pure empiricist experimentalist technologist who has not the foggiest idea of the underlying theoretical principals and not caring. Then there are people at the other extreme who see only mathematical formalism.
Vardalas:
They like elegance.
Kroemer:
They like elegance. I like elegance too.
Vardalas:
But elegance for the sake of elegance as opposed to elegance for the sake of a physical principal.
Kroemer:
Yes. It is a broad spectrum.
Vardalas:
I was thinking more of the areas of the Grand Unified Theory.
Kroemer:
I find this a fascinating human exercise. I can see why it appeals to ambitious young people who are doing theoretical analyses that this is the great problem to be solved. I understand this. The question is what do they do when they realize that there is a fierce competition? Thus far no one is making any progress. I wish they were a little broader.
Vardalas:
There are a lot of shipwrecks on that rock.
Kroemer:
My recommendation to somebody who really is interested in these deep profound principals is sure, do that, but do not restrict to yourself to just that. View this as a part embedded in something much broader.
Vardalas:
I see. Getting back again to the Central Telecommunications Laboratory, in Germany its acronym is FTZ. What does that stand for in German?
Kroemer:
Fernmeldetechnisches Zentralamt. Fernmelde is telecommunications, technisches is technical, zentralamt means central office.
Vardalas:
Okay. I know you answered this before but I have to ask you this anyway. Did you have close interaction with these experimentalists and technologists?
Kroemer:
Yes.
Vardalas:
Do you think this sharpened your skills as a physicist in a way you would not have gotten if you had stayed in academia?
Kroemer:
Probably yes. I don't know whether it sharpened my skills as a physicist, but it certainly broadened the range of things I was thinking about. It was again methodology, and I was taking an active interest in how one could do this, how I could implement this. I was taking an active interest in trying to understand first of all what we had done and then going on from there and saying, "All right, now that I understand, how could I modify it to do better?" It was like a pendulum swinging back and forth. That has a great influence on me that continued in my subsequent jobs.
Vardalas:
Would you say that in a sense this experience brought out your versatile talents as an astute problem-solver -- what you call an opportunist? Was this the first seed of you acting like an opportunist?
Kroemer:
Yes.
Vardalas:
In a good sense.
Kroemer:
Yes. It always shocks people that I call myself an opportunist, because it's always a dirty word. But I deliberately use it – perhaps for the shock value – because an opportunist is somebody who is looking for good opportunities to do something.
Theoretical physics, applied physics and electrical engineering as influences on semiconductors
Vardalas:
The right problem. Yes. I want to get to that, because that's an important theme - choosing the right problem. I want to ask you a broader issue here on the different levels of knowledge and methodology. In your career your research has bridged the realms of theoretical physics with a small “t”, applied physics and whatever that means and electrical engineering?
Kroemer:
Yes, although I am not really an engineer. My degree in engineering is honorary.
Vardalas:
Your understanding of some of these issues has been an important part of your work. Looking back over the growth of science and technology in semiconductors, how would you characterize the body of knowledge in each of these areas and how they interact at the interfaces between theoretical physics, applied physics and electrical engineering? And has that direction changed over the last fifty years?
Kroemer:
I don't quite know in which terms to answer that one.
Vardalas:
After all, you are in an engineering school now.
Kroemer:
- Audio File
- MP3 Audio
(424 - kroemer - clip 3.mp3)
I am in an engineering school and I feel very comfortable in an engineering school because I bring to bear my scientific background to solve problems and to contribute to engineering developments. However I am basically still a physicist. Throughout my entire career whenever I was working on something and then discovered, "Hey there's something else you have to learn in order to able to do this," I have always tried to acquire this knowledge. Therefore I have a background – though an extraordinarily sketchy background – in electrical engineering. I know those parts of electrical engineering that I need and I know those parts of theoretical physics that I need and those parts of mathematics that I need. I think the field as a whole has required and has benefited from the broad assortment of people in all of those particular disciplines from theoretical physics to metallurgy to you name it.
Those laboratories where the important contributions were made typically had organizations with a broad range of people from different backgrounds – to the point that an individual's background was often not known or even a concern. I certainly learned that in the three years I spent at RCA Laboratories.
Vardalas:
Where the problem is the important thing. How the problem is tackled and one's background doesn't really matter.
Kroemer:
That's right.
Vardalas:
That's interesting. Is that something that you feel has implications as to how universities should train people in these fields?
Kroemer:
To some extent this is a matter of education at the university, but to a large extent it is also a matter of personality. I believe that an in-depth education in some specific discipline is important.
Vardalas:
Yes.
Kroemer:
In my case it was in theoretical solid-state physics. All the other things got added on later. I do not believe in giving people a superficial education in a lot of different things. An in-depth education in the specific area is important, but it also requires an attitude to be interested in what goes on in adjacent fields, trying to interact with people in other fields, trying to contribute to their experiments or projects and having them contribute to your projects.
Vardalas:
I guess the issue then is not the knowledge but the attitude to look to the other people.
Kroemer:
Yes. For example I think it is important that physicists not take the attitude that pure physics is something more elevated than applied physics. One sometimes finds that attitude. Then in engineering the opposite attitude can be found which looks down on anyone interested in theory or in something for which an application is not already known.
Vardalas:
Yes.
Kroemer:
You know how hostile I am to that idea.
Vardalas:
Yes. I'll get to that. That's interesting.
Kroemer:
I think universities can contribute to this in two ways. They can obviously contribute by providing the specific knowledge that is required. There is always hope that along with that specific knowledge there will be instilled a desire to do more than just a specialty. It is desirable that we combine these two aspects and see this embedded in a broader framework. Of course this will depend on the individual, the faculty and the university environment.
Heterostructure bipolar transistors
Vardalas:
Am I correct in judging from what I read that the embryo of your ideas on heterostructure bipolar transistors first came to you while you were in the Central Telecommunication Laboratory?
Kroemer:
That's right.
Vardalas:
And then you write "and the rest is history."
Kroemer:
Yes.
Vardalas:
How did the idea come to you? Was it a natural progression of your dissertation?
Kroemer:
It had nothing to do with my dissertation.
Vardalas:
Was it the environment you were in that prompted you to think about this?
Kroemer:
Yes. We were working on the very early transistors. They were so slow that they were basically useless for the applications on hand at the time. I realized that an earlier incarnation of transistors – the first so-called point contact transistor that didn't have junctions inside – talking about bipolar and not FETs - were significantly faster than the first junction transistors. And there were a variety of reasons for that. One of the thoughts that came to me was, "Well, we do know that in the point contact transistor the collector is very leaky." In other words it draws a rather large current by itself without an injection of holes or electrons or whatever it is from the emitter. Of course this leaky collector film introduces an electric field in the body of semiconductor around the collector in such a way that carriers of the opposite polarity were drawn towards the collector. So that was my point of departure. I'd say, "How could you build an electric field into a junction transistor?"
Vardalas:
That was your key idea?
Kroemer:
That was the key idea and not by doing s leaky collector. The first idea then was, "All right, we are putting a non-uniform doping into the base," and specifically I looked at an exponential doping profile. You can then show that this leads a built-in field that speeds up the carrier. And that required an understanding of band structure.
Vardalas:
This is where theoretical issues came into play.
Kroemer:
This is where my understanding of basic semiconductor physics told me that it should work.
Vardalas:
I see.
Kroemer:
Then at one stage the thought occurred to me that another way of putting in a field is to use a non-uniform energy gap. And this is the theory.
Vardalas:
Okay. Did this idea come immediately after the first idea? How did this idea emerge?
Kroemer:
I don't know what immediate means here.
Vardalas:
Did it take a year?
Kroemer:
Less than a year. It was part of the same work.
Vardalas:
Okay.
Kroemer:
The idea came then that a field could be built in by grading the energy gap.
Vardalas:
Did something prompt you to think of it?
Kroemer:
It was obvious.
Vardalas:
It was obvious. Why hadn't people thought of it earlier if it was obvious?
Kroemer:
It was not obvious to others. Well, let me say it was obvious because I had a goal. I wanted to put in a field, I needed a sloping band and I realized that I could create a sloping band. You see I always try to view those things from as fundamental a point of view as possible. The need for a band slope was the key idea. The field needed a slope in the band and I realized fairly quickly that a second way to introduce a slope in the band was to grade the energy gap.
Vardalas:
Okay, so that's the succession of ideas. Once you committed to a sloping electric field then that led to the next idea.
Kroemer:
- Audio File
- MP3 Audio
(424 - kroemer - clip 4.mp3)
Yes. There is a comment in that first German paper on that one and I estimate what kind of a potential drop could be obtained in the band. It was a bit optimistic. These things are always optimistic. I looked specifically at germanium-silicon, not realizing that there were very, very severe problems there. Anyway, my technological colleague, Mr. Hähnlein, was basically the physicist working on the device technology. He looked at this and said, "There is no way I can do it. The most I can do is perhaps put an emitter with a wider energy gap on the base region, but I cannot put a control field in the base." Of course that would mean a uniform gap in the base region – and that would mean that the field that I was trying to achieve would not exist. So that idea was out.
But then on the way home, out of sheer curiosity, I wondered, "Well, what would be the consequences if that was done?" And I realized that this would have unique benefits of its own. The repulsive barrier at the emitter side would be increased for those carriers flowing from the base back into the emitter. That could now also be traded off with other things. This is how the wide gap emitter and the graded gap arose within days of each other. They are both in that first German paper published in 1954, so the idea probably arose in late '53 — though, in that paper it was unfortunately not accompanied by a band diagram. The idea that one has to be able to draw a band diagram came shortly after that.
Vardalas:
It seems like you had considerable freedom at this place.
Kroemer:
I could work on anything I pleased.
Vardalas:
And you found the work quite stimulating I imagine.
Kroemer:
Yes.
Move to RCA
Vardalas:
And yet you wanted to move on.
Kroemer:
Well, I wanted to get involved closer to the action.
Vardalas:
I gather you first thought of the opportunity with William Shockley.
Kroemer:
Shockley visited our place and I had a long and wonderful long discussion with him. I asked him about the chances of coming to Bell Labs at that time. He was reluctant in his response, because he was an official visitor. Official visitors are not stealing people.
Vardalas:
Oh yes, of course.
Kroemer:
I think his reluctance regarding such matters later decreased. He said, "You will have to take the first step." And then there was Ed Herold from RCA whom I had met at the Physical Society Meeting in Innsbruck. He didn't even know about the existence of our laboratory but I got to have a long discussion with him. He had presented an invited paper which pretty much confirmed everything I had claimed about how what was going on with the metallurgy that some people didn't believe. It turned out he had a day open on his agenda and so he came also to Darmstadt to the FTZ [Central Telecommunications Laboratory] and I had a long chat with him. When I asked him about the chances he was not under the constraints that Shockley had. A good question is why I ultimately went to RCA rather than Bell Labs.
Vardalas:
Yes.
Kroemer:
That's a very, very good question. In fact Bell Labs was offering more money when it finally came to the details.
Vardalas:
And what was it?
Kroemer:
I don't know. Jim Early said to me, "Herb, you may have felt that at Bell Labs you would always be in Shockley's shadow whereas at RCA you would be your own man." That may have played role in my decision. I do not know. That is speculation. This is sort of tabletop psychiatry.
RCA Labs research environment
Vardalas:
Well, let's go forward. What were your impressions of the people in research when you got to RCA Labs?
Kroemer:
Oh, it absolutely fabulous.
Vardalas:
Can you recreate what you saw and felt when you first got there?
Kroemer:
It was simply a different world. There were lots of people, many of them very good people, working on all sorts of aspects. It was an environment where free discussion across disciplines was very, very much encouraged. A theorist who wanted to do an experiment was not talked out of it because people saw no need for it. I said, "All right. Go ahead." It was a wonderful lab and I was surrounded with wonderful people. The ones who influenced me most were perhaps not the ones who were best known.
Vardalas:
Who were they?
Kroemer:
One particular guy who influenced me tremendously was an office mate named Lou Pensak. He had been heavily involved in the technology of television tubes. He was a constant discussion partner and also introduced me to the idea that if you want to build something the right tools are needed – and the tools should be built first. He had a great influence on me. It was a fascinating laboratory. Ed Herold was a wonderful boss. It was a great experience. Now this discussion with Ed Herold in Darmstadt was interesting one. You see I understood the metallurgy of those transistors in those early days. They were pnp transistors that had indium alloy to two sides of a germanium wafer. I was curious about npn transistors. I tried to think about how might one do an npn? I couldn't get anyone interested in actually doing it, but I sort of figured out what I would do if I were asked to build an npn. So I ask Ed Herold whether they had also made npn transistors. He said yes. I asked, "What did you as the alloy metal?" He was a little bit reluctant but answered, "Lead." To me it was clear that it was either lead or tin, and for reasons that I do not remember I thought lead was the likely candidate. I said, "But lead is not a donor, so you must have added something else." Long pause, more reluctance. "Antimony," which I thought all along was they had. Then I asked, "Well, how much antimony and at what temperature did you alloy it?" He clammed up. So I told him "You use 9 percent antimony and alloy it at 600°." And his jaw dropped. It was sufficiently close to what they were actually doing. It was a living exercise in the old rule "Never mind how something is done. Knowing that it has been done is the biggest secret of all."
Vardalas:
Is that what convinced him to hire you?
Kroemer:
I don't know whether that convinced him, but certainly that was an interesting exchange we had that I fondly remember.
Vardalas:
When you went to RCA did you go there with your own research agenda they had approved?
Kroemer:
No. Well, I came with certain ideas.
Vardalas:
What did they want you to do? How did this process work?
Kroemer:
They gave me an office and told me to work on whatever I wanted.
Vardalas:
Oh, that's it?
Kroemer:
Yes. Herold explained to me later that when he hired people his interest was in how good they were at whatever they had done previously. What field it was didn't matter to him. His concern was only for the quality of their work. His next question would be, "Is this person interested in working on the problems that I can offer?" If the answer was also positive, he would take the person. He would rather take an individual like that who had no idea about the field than somebody who had previous experience in the field but did second rate work.
Vardalas:
That's interesting.
Kroemer:
He explained to me, "Listen. In the environment of our lab, the kind of people that I hire like to communicate. If I put them into this lab they are sitting in an intellectual feedback loop, where they cannot help but be influenced in being directed towards the topics in which the lab is interested. They are not going to go off on a wild tangent." I would have been permitted to do so. In fact later on I did go on a fairly wild tangent, but even then it was encouraged on the grounds, "Well, maybe something will come of it."
Vardalas:
I see. In a place like this where people were allowed to do what they wanted to do, within I gather rather broad constraints, how does accountability work? How did they decide whether to keep people on and determine whether people are good or not good? By the results?
Kroemer:
Basically people were judged by the quality of their work. And it was encouraged to go out of the established conventional wisdom.
Vardalas:
Was it a healthy environment where failure was encouraged? If you tried and failed was that not seen as a negative thing? In some places that would be treated as a very negative thing like, "That's terrible. We don't want that."
Kroemer:
No. I think failure was encouraged. I remember when I was working on a particular project, I was working very hard and it eventually became clear to me that this was not going to work. It had been clear to others long before that it wasn't going to work, but there was no influence on me. I do remember one day in the morning I drew my conclusion that it was not going to work. I cleaned up my workbench, cleaned up my desk, went to my supervisor and said, "Harwick this is not going to work. I’m going to change subjects." He said, "Yes? Well, that's fine. We knew you would come to that conclusion." That was it. There was no pressure. It was an experience.
Of course I was the kind of person who was willing to drop the subject. There are sometimes people who do not want to let go of it. The dumbest reason for continuing something is because you have made the mistake of starting it. I have always been willing to drop projects. In fact I have never been interested in milking the last bit of juice out of something. If I had succeeded in achieving the key point, suddenly I would discover I had lots of friends who were willing to do the rest. Why not let them do it?
Vardalas:
Did you have an opportunity at RCA, the kind of problems, to go back to heterostructure bipolar transistors?
Kroemer:
Actually that is something I did try to do. In fact that was the project I decided to drop. The technology was staggering, the equipment primitive. And remember, by that time I knew more metallurgy that most theoretical physicists. So I came up with the idea of putting a silicon/germanium alloy on a germanium base. This was all on germanium bases. How do you do this? I knew my phase diagrams quite well, so I made a silicon-gold eutectic. Silicon-gold has a relatively low melting point 360°C or something like that. I made the silicon-gold eutectic which is an unpleasantly brittle substance, put it on an anvil and smashed it into a powder, and with a pair of tweezers picked up little grains and put them on the germanium wafer and alloyed it at 500 or 600°C, the eutectic would melt, it would eat up germanium and the germanium would then recrystallize. And we had added some dopants.
Vardalas:
You did this yourself?
Kroemer:
I did the experiments. I didn't have a technician. I did all of this myself. My one complaint at RCA is that they did not give me a technician. I think they had the right instinct. I didn't know how to use a technician.
Vardalas:
That must have taught you new things trying to do this stuff yourself.
Kroemer:
Yes, yes. This is where Lou Pensak's advice on how one does things came in. We actually made some transistors, but it was clearly a technology that would never amount to anything. So this is that and I dropped it. At that point I decided to get out of transistors and move to something altogether different. That must have been in late '56. I did write two papers, and those papers played an important role.
I then became very much interested in hot electron transport at low temperatures – which is something totally different.
Assessment of research accomplishments at RCA
Vardalas:
When you look back at all the work you did at RCA, of what are you most proud?
Kroemer:
In hindsight I see that those two papers that I wrote clearly had the greatest impact. But I did other things that were at the time certainly important to me.
Vardalas:
What gave you the most satisfaction at the time?
Kroemer:
I don't know. I don't want to single out any specific things. Of course by hindsight it's very clear the heterostructure work was the important one. And that paper in RCA Review was probably one of the most papers I ever wrote. I made a mistake by publishing it in an obscure journal, with the result that no one read it. However Zhores I. Alferov, the Russian with whom I won the Nobel Prize, read it. He knew the paper.
Vardalas:
I see. I would have imagined that the RCA Lab's journal would have been read by a lot of people just like the Bell Labs' journal.
Kroemer:
That was RCA's idea. The RCA Review was sort of an imitation of the Bell System Technical Journal, and it never really acquired the same following.
Vardalas:
I see.
Kroemer:
It turned out to be a place where a lot of papers were put if you couldn't get them published elsewhere.
Vardalas:
Did you think this was an important paper when you were writing it?
Kroemer:
This is hard to say. In that paper I clearly spelled out the heterostructure design principal. If I were to rewrite that paragraph that I quoted in both my Nobel lecture and elsewhere I would clean up the English a little bit, but the idea is clearly there. I realized the power of the idea. What was not clear to me at all was how important the idea would become. The real triumph of this idea is the heterostructure laser, which draws exactly on the concepts that were outlined in that RCA paper. Incidentally, this is something very typical that the fruitfulness of a new idea will not come until later.
Vardalas:
Right.
Kroemer:
There is a wonderful quote by David Mermin in Physics Today a few years ago. He says something like – you have to look up the exact wording – "I'm looking forward to the day that people realize that discovery does not work by deciding what you want and then discovering it."
Vardalas:
Isn't that in your Nobel lecture?
Kroemer:
I quoted that in my Nobel lecture.
Vardalas:
This has implications for those who say funding for science should be based on what the applicant—
Kroemer:
That is total nonsense.
Vardalas:
Is there anything to be gained by trying to get scientists thinking about problems that have some obvious need and utility in society? In other words even in theoretical issues like attacking an important problem and applying good science to it?
Kroemer:
This flows two ways. On the one hand people who have problems that they want solved can and should try to get more fundamental-oriented people interested. I think we all acknowledge that one. However on the other side – and I can only speak for myself – I have always been interested in fundamental principles, but whenever some conceptual advance had been made I have also always asked myself, "What could be done with that?" I did not restrict myself in my explorations to things where I saw applications beforehand, but I periodically asked myself, "What kind of applications might this have?" Sometimes I could outline some and sometimes and I could not. And if I could not, I still went on. It didn't stop me pursuing a certain thing.
Vardalas:
You gave applications some thought.
Kroemer:
I always gave it some thought. Critical thought. One has be honest about it and not come up with something where a few more minutes of additional thought would show it to be nonsense.
Vardalas:
I think you spent three years at RCA Labs.
Kroemer:
Yes.
Phillips Semiconductor Group in Hamburg
Vardalas:
Then you returned to Germany to head up the Phillips Semiconductor Group in Hamburg.
Kroemer:
That's right.
Vardalas:
After reading some of the things you wrote, I was wondering if being homesick was one of your reasons for returning to Germany.
Kroemer:
Yes.
Vardalas:
Was there something more to it than that?
Kroemer:
No. It was simply homesickness – on the part of both my wife and myself, though more my wife.
Vardalas:
Is your wife German?
Kroemer:
She is from Berlin.
Vardalas:
We never discussed when you met your wife.
Kroemer:
I met her when I was a student at Göttingen.
Vardalas:
Okay. So she was homesick to get back. She must have left Germany with you almost immediately.
Kroemer:
She followed me after a year. She stayed behind because we had a young child. She stayed behind with her parents. Then she followed me, but then she was terribly homesick. And I was offered what looked like a rather attractive job at Philips. So I went back. I must say that I have nothing but good things to say about the Philips Company. They treated me wonderfully.
Vardalas:
What was the undertaking?
Kroemer:
They had a new research laboratory in Hamburg – a lovely city incidentally – and I was the head of the semiconductor group. I decided to steer this group towards gallium arsenide. That was in 1957. I got the people in Eindhoven who were a bit astonished and saying, 'Yes, why not?" One person was assigned to do this technology. When I decided to go back to the United States this project was shut down.
Vardalas:
Why did you leave? Were you unhappy with the way things were going there?
Kroemer:
I had no problem with Philips. Germany simply was not the same. Our perception of Germany changed.
Vardalas:
What does that mean?
Kroemer:
My wife and I both suddenly realized that we wanted to go back to the United States. That was basically it. Philips tried all sorts of things to get me to stay. They were very nice about it. It was a reverse homesickness.
Vardalas:
You had to leave the U.S. to realize that you wanted to return to the U.S.
Kroemer:
Yes, so I book this under mental health expenses.
Vardalas:
Did you think you did anything in terms of your own professional development or ideas in physics there?
Kroemer:
While I was in Hamburg? No. Nothing.
Vardalas:
That must have been a disappointing time for you professionally.
Kroemer:
Well, I did some thinking and I think I wrote a couple of theoretical papers, but it does not show up in my productivity resume.
Varian Associates; semiconductor lasers research
Vardalas:
You came back to the U.S. and Varian Associates. Why Varian?
Kroemer:
I knew the head of central research at Varian, Lou Malter. I knew him from RCA. And the people at RCA actually wanted me back.
Vardalas:
Oh, they did?
Kroemer:
Yes. I don't really quite know why I didn't want to go back to RCA. I asked Lou Malter a little and he said, "Oh yes. Sure. Come." And we never agreed what I would work on.
Vardalas:
You never did?
Kroemer:
We never talked about it. He said, "Just come. There are plenty of interesting things around." He knew me well enough that he realized it would probably work. And I spent several years then at Varian Associates trying to build up a semiconductor activity there.
Vardalas:
You were doing the diode laser, right?
Kroemer:
Well, this is how the idea came about. I was actually not allowed to work on it.
Vardalas:
Yes, I wanted to get to that. How did you come upon this topic?
Kroemer:
I had worked on what we now call heterostructure bipolar transistors.
Vardalas:
And you put it aside for a while.
Kroemer:
I put it aside, yes. I think it was in 1962 at the Device Research Conference in Durham, New Hampshire, which was the big annual device meeting, that all hell broke loose about semiconductor lasers. The first gallium arsenide laser was reported and that dominated the conference. And I really wasn't interested in those things. I knew that the theoretical principle permitted that and I was still astonished that it actually worked, but I was totally occupied with something else. But a colleague of mine, Sol Miller, was also at that conference and he took a deep interest in it, and so he started working on that one after we returned to Varian. And it must have been in March of '63 that Ed Herold, who had come from RCA, the director of research at Varian.
Vardalas:
Kroemer:
The same Ed Herold. Lou Malter had hired him too.
Vardalas:
He was the director of research?
Kroemer:
- Audio File
- MP3 Audio
(424 - kroemer - clip 5.mp3)
He [Herold] was not director but vice president of research. Anyway, he insisted that things get done RCA style, and that included weekly colloquia. Therefore all of us had to give talks – which I think is a good idea. Sol Miller was asked to give a talk and he picked that laser topic. He gave a beautiful talk pointing out all that had been done and also that these things didn't work at room temperature and didn't even work continuously. It required very, very short pulses, and certainly a low temperature had to be used.
At the end of the talks Herold said, "That's all very nice, but what are the chances of getting this to operate cw and at room temperature? Because that's where the applications are." And Sol Miller replied suddenly something to the effect that, "No, this has been looked at." Then he quoted some paper. I do not want to repeat the quote, but the quote said that this was fundamentally impossible.
Vardalas:
Fundamentally impossible.
Kroemer:
Ed Herold was not about to put up with the statement that something was fundamentally impossible without explaining why it was fundamentally impossible. Then Sol basically gave an explanation that boiled down to that, “You first of all need a population inversion, so you need a degenerate doping on both sides, and if you bias it to the point that you actually get stimulated emission the electrons leak out to the p-type side very, very rapidly. Holes leak out to the n-type side."
Vardalas:
You can't maintain the population inversion?
Kroemer:
You just cannot get a decent population inversion except at low temperatures where the statistics is in your favor or pulse where you have transient effects. And I do not know whether Sol Miller was finished, but I certainly said, "That's a pile of crap."
Vardalas:
In those words?
Kroemer:
In those exact words. All you have to do is put a wider energy gap on the two sides. It was obvious. The moment I was told that there was a problem the answer was obvious because at that point I had been thinking enough about heterostructures. Anyway, everyone was astonished and we did a number of things. I wrote a paper on this one and submitted it to Applied Physics Letters.
Vardalas:
Yes.
Kroemer:
They rejected it.
Vardalas:
Before you go on, do you have any idea why it was rejected?
Kroemer:
All of this correspondence got lost.
Vardalas:
Do you recall?
Kroemer:
I do not remember exactly why. Ed Herold, who was a big shot in the IEEE, did not like the idea that I had submitted it to Applied Physics Letters in the first place. "Well, send it to Proceedings of the IEEE. They will publish anything." That is of course why I hadn't wanted to send it there. The letter section there was not very good in those days.
Vardalas:
Do you think that was why that paper was ignored?
Kroemer:
I submitted it, it was accepted, and it was published. No one read it. A reviewer had pointed out to Panish and Hayashi, who subsequently did this, that this paper of mine existed. When they published they were gentlemen and acknowledged the idea. And we wrote a patent.
Vardalas:
Yes. That patent was assigned to Varian, wasn't it?
Kroemer:
Yes. It has safely expired. We wrote a patent. I wasn't allowed to put a band diagram into the patent because the head of the patent department was an electron tube man who did not understand semiconductors. He wouldn't put anything into a patent that he himself did not understand.
Vardalas:
Oh my goodness.
Kroemer:
He argued however that science really does not matter; all that matters is that the correct prescription of what to do is given.
Vardalas:
For a patent, yes.
Kroemer:
That is technically correct, but if you look at Bill Shockley's patents there is always a very explicit elaboration on the science.
Vardalas:
Was your work in this paper purely theoretical?
Kroemer:
Purely theoretical.
Vardalas:
There were no prototypes, no fooling around with experiments, it was all theoretical arguments?
Kroemer:
It was all theory. And in fact it wasn't at all clear how one would go about building one. We didn't really have the technology. We got a number of criticisms. There were a few people who said, "Your physics is wacky. It doesn't work." But that wasn't really what bothered me. I had been through this sort of mode of operation before and had learned not to pay attention unless everybody tells me that it does not work. I prefer to rely on my own judgment. I was convinced the physics was right. The other argument of course was, "There is no technology." And that was true. But then came the killer. That was, "There is no point in developing the technology because this device will never have any practical applications." End of statement.
Vardalas:
Was that the reason why Varian did not want to pursue this further?
Kroemer:
That's why. They didn't say, "It's not in our business field." They said, "It has no applications."
The Gunn effect
Vardalas:
Did you fight this or did you give in to it?
Kroemer:
I gave in. I probably would not have given in if the Gunn effect had not come along at the same time.
Vardalas:
What is the Gunn effect?
Kroemer:
The Gunn effect is the phenomena where if one takes a piece of gallium arsenide, applies a high voltage, then under certain conditions high-frequency oscillations result. That was fascinating physics and I was the first to offer an explanation for the Gunn effect. I worked on that for a number of years. That was my consolation prize for not having been able to work on the laser. Well, I should not say consolation prize. It was sort of an alternative since I could not work on the laser. And we had the technology for the Gunn effect. In a way of course this is regrettable, because as a result of this work I never played any role in the subsequent realization of the DH Laser.
Vardalas:
Is that a source of disappointment for you?
Kroemer:
A little bit, yes.
Vardalas:
You also mentioned that as an alternative to this rejection of the laser rekindled your longstanding interest in high-electron negative resistance effects.
Kroemer:
This is the Gunn effect.
Vardalas:
I see. But you had worked enough for your doctorate dissertation in this broad area.
Kroemer:
In the broad area, but the Gunn effect is something quite different.
Vardalas:
Okay, but you say "with a longstanding interest in this."
Kroemer:
I had a longstanding interest in transport properties under obscure or unusual conditions. In my dissertation I worked on hot electron phenomena. During my last year at RCA I was working on hot electrons and hot holes in germanium. One of the reasons why I became interested in gallium arsenide was because theory predicted that the thing that we were looking for in germanium would be much easier to find in gallium arsenide.
Vardalas:
Okay.
Kroemer:
That influenced me in starting a gallium arsenide project. Then I went to Philips, and we wanted to look at high field transport properties as the first application. We did not have transistors or lasers in mind. Typical for my kind of thinking, I was simply convinced that the three-five compounds held a tremendous amount of future promise. So let’s be amongst the ones to do it.
Vardalas:
Did you go to Philips and pursue this idea of transport issues?
Kroemer:
That's not why I went to Philips. I wanted to do something that they were not doing in the main lab in Eindhoven. And I wanted to do something that was at the forefront of solid-state technology. I wanted to get into compound semiconductors and I wanted to work with gallium arsenide specifically.
Vardalas:
You spent ten years on research and engineering around the Gunn effect. Would that be where your biggest theoretical accomplishments lie?
Kroemer:
I certainly have more papers in that area than in anything else. It was a significant accomplishment. I would say I was one of the handful of leaders in this field from day one, and it was only when I came to Santa Barbara that I put an abrupt end to it. I decided to do that.
Nobel Prize for the development of heterostructures for high-speed and opto-electronics
Vardalas:
Was your Nobel Prize award related to the Gunn effect?
Kroemer:
No. It was for the development of heterostructures for high-speed- and opto-electronics.
Vardalas:
Thinking back, do you think the reason they gave this science such a high significance was because of the practical technological effect it has had?
Kroemer:
Absolutely.
Vardalas:
And that if that hadn't happened they probably would not have given you that award?
Kroemer:
Absolutely. Yes. The 2000 Nobel Prize in Physics really was a break with the tradition. The tradition for ninety years had been to award the prize only for discoveries, even though Nobel's will specified discoveries or inventions. There are certain reasons why the restriction to discoveries was made around the time of World War I.
Vardalas:
Why was that?
Kroemer:
It probably had to do with the Nobel Prize for Gustaf Dalén in 1912 which was purely a technological invention. That caused a great deal of protest. Many felt that the Nobel Prize in Physics should not be for something like this. If one simply looks at the Nobel awards, they are almost all for discoveries and those that were inventions were typically inventions essential for research discoveries.
Vardalas:
Right. Okay.
Kroemer:
The bubble chamber is a good example.
Vardalas:
Instrumentation.
Kroemer:
Instrumentation. So when friends and colleagues mumbled about me getting it someday, I looked at the statistics and took the attitude, "Well, they do not know the rules of the game." That was probably good for my mental health not to take those comments too seriously. Then, since people had been talking about it I sort of was thinking about the possibility and thought to myself, "Well, they are not going to give it to the HBT. They are not going to give it to me for the laser. I may have been the first one to spell out the idea, but I did not do the first laser. If however they decide to give it for the heterostructure concept, then I have a chance and I will probably share with Alferov."
Vardalas:
You said that to yourself?
Kroemer:
Yes, I said that to myself. I never talked about this to anyone. I think you are the first one to whom I have told this. And this is of course what happened. But it was clearly a break with tradition.
Vardalas:
But yours is less of a break than it was for them to give it to engineers like Jack Kilby.
Kroemer:
Yes, for Kilby it was even more so a break with tradition. However I think it was well deserved. They must have gone back to read Nobel's will.
Moore's Law and CMOS
Vardalas:
Okay. I want to cover some more ground before our time runs out. I am very interested in your interpretation of Moore’s Law. You gave a plenary talk called "Speculations about Future Directions."
Kroemer:
This must have been the MBE paper published last year. All right.
Vardalas:
In it you say, "Moore’s law reflects the triumph of parallel assembly."
Kroemer:
Yes.
Vardalas:
Would you please elaborate on that?
Kroemer:
"Moore’s law is an observation; it's not a law.
Vardalas:
Of course. Yes.
Kroemer:
It is an observation that spans several decades. The number of devices, per chip and per processing step processed in parallel has increased exponentially or approximately exponentially. In order to be able to do this a reduction in dimensions was necessary. But the reduction of dimensions was an enabler. If the dimensions were reduced and then these devices were done serially, one at a time, then you would not have Moore’s Law. This is not going to make a Pentium type chip doing one device at a time. This is my perspective on Moore’s Law. In that sense it is a triumph of parallel assembly. I raise this issue periodically whenever people say, "As the dimensions get smaller quantum mechanics will become more important. We will build a quantum device then. Once we have a quantum device Moore’s Law continue for a little bit longer." But actually quantum devices are all put together serially.
Vardalas:
Aha. That's your point.
Kroemer:
This is the wet blanket that I put over the subject. Moore was extraordinarily perceptive, because he didn't formulate it this way but he realized the trend and that we were very, very far away from physics limitations. In my perspective, Moore’s Law was based on the development in the infrastructure. If one asks, "Why didn't we do it right away?" that is a very easy question to answer: there was no infrastructure, instrumentation, manufacturing, and there were no crystals of the proper perfection.
Vardalas:
I'm interested in this in my own research. Someone told me that Moore’s Law is essentially a scaling up issue – not indefinitely, but one can scale up in this way.
Kroemer:
A straight line cannot be drawn on log-paper forever.
Vardalas:
Yes. But I had the impression that one of the reasons it really took off was the emergence of CMOS; that CMOS technology is geared to this kind of thing whereas bipolar is not.
Kroemer:
Yes, that is true. Let's put it this way. Take cause and effect the other way around, it is only with CMOS that we could do it. And of course Moore predicted CMOS.
Bipolar technology and CMOS
Vardalas:
I'm interested in the relation with bipolar technology.
Kroemer:
The principal reason this could not be done with bipolar is that thermal load is unbearable. Bipolar is a hot plate.
Vardalas:
Okay. Bipolar cannot scale up the way CMOS can.
Kroemer:
No.
Vardalas:
I see.
Kroemer:
Bipolar is obviously very, very important. Your cell phone is loaded with bipolar. Certainly on the transmitting side it's bipolar. Probably HBT is on the transmitter side, and it may be HBT is on the receiving side.
Vardalas:
I was thinking in terms of computer development, because at one point a bipolar was a dominant form of device in processors.
Kroemer:
Yes. Big mainframe computers used to run on ECL, Emitter Coupled Logic.
Vardalas:
Yes.
Kroemer:
And there is not a device that generates more heat per function than Emitter Coupled Logic.
Vardalas:
I'm writing a book now, Control Data Corporation, and they were so wed to the bipolar because of the speed advantage they had.
Kroemer:
Yes. They had a speed advantage, but that advantage is gone.
Vardalas:
Cray didn't stop using bipolar until the late 1990s.
Kroemer:
Yes. And of course as the devices got smaller they got faster too.
Heterostructures, germanium and CMOS
Vardalas:
For someone who is not in the field, does the value of the model of heterostructures have an equal technological importance for the CMOS device?
Kroemer:
No. At least I do not see it. I think there is a good reason why CMOS continues to be silicon, though we may see silicon-germanium in silicon technology. I know that many of the silicon houses are working on silicon-germanium technology though I do not know any details. In that sense that is heterostructure.
Vardalas:
One of the traditional explanations for the rise of silicon was its importance to the military in terms of its stability to damage.
Kroemer:
Yes.
Vardalas:
And that is the reason that silicon can dominate more than germanium.
Kroemer:
Certainly the military were the ones who initially supported it and helped it get off the ground, but I think what was probably more important was organizations like Bell Labs that realized that germanium was too limited.
Vardalas:
Too limited in what respect?
Kroemer:
Temperature capability, processing capability. Jack Morton probably did more than anyone else to promote silicon technology early in the game when it was still largely bipolar. Then of course there is the importance of that oxide. Silicon is a rather unusual material. All the attempts to imitate silicon in compound silicons have in my opinion been a waste of time. They fail to realize that silicon is the abnormality.
Vardalas:
It's not the rule but the exception?
Kroemer:
Not the rule. It's the exception. It has that wonderful stable oxide which serves for protection as a masking aid, it has a good thermal conductivity and it has the right kind of an energy gap.
Vardalas:
You can't find that in the other materials?
Kroemer:
You do not find this in anything else. It has a lousy mobility. That was the principal reason we thought silicon would never fly – the mobility. I remember back in the days at RCA we didn't believe in silicon. First of all one could not get rid of that stupid oxide. And certainly the mobility is still lousy. This is why we believed in germanium.
Vardalas:
Oh, I see.
Kroemer:
That was wrong of course.
Vardalas:
Do you think there are any limits to the switching speeds of bipolar transistors? Have we reached the limits?
Kroemer:
You will have to ask my colleague, Mark Rodwell. He is the leader on this one and he expresses things in terms of fmax, which is not the same as switching speed. Of course typically it is not useful to switch applications anyway. And his F-maxes are so high that he does not measure them. These are extrapolated figures.
Employment at Fairchild, 1966-1968
Vardalas:
You left Varian in '66 and joined the University of Colorado in '68.
Kroemer:
That's correct.
Vardalas:
That leaves a two-year gap. What did you do during those two years? Were you a consultant?
Kroemer:
I was at Fairchild.
Vardalas:
That's not in your biographical sketch.
Kroemer:
Yes. Those were two very unhappy years.
Vardalas:
Why was that?
Kroemer:
I do not want to talk about it.
Vardalas:
All right. You were in a silicon world?
Kroemer:
I was a misfit in a silicon world. They were unhappy years, but that did not in any way influence my good relations with Gordon Moore.
Vardalas:
Okay. That explains why you went to the University of Colorado. You wanted to get out of Fairchild.
Kroemer:
Yes.
Transition to academia at the University of Colorado, 1968
Vardalas:
This was the first time you worked in a university. After all those years in industry or relating to industrial people, what was it like to come back to academia as a physicist?
Kroemer:
I enjoyed it. I was a bit concerned about being on a fixed schedule, at least for teaching, but the intellectual freedom that one has at universities is just wonderful.
Vardalas:
Okay. Were there any attitudes in the industrial environment that you would like to have found in the academic environment?
Kroemer:
No.
Vardalas:
What did you do at the University of Colorado that is significant in your mind?
Kroemer:
I went there in '68 and was still very heavily involved in Gunn effect. I was also doing a number of theoretical pieces of work, though I didn't really do anything terribly important at that time. The Gunn effect work was good work, but the best part of the Gunn effect was done already while I was still at Fairchild. That is in fact something that I did at Fairchild. I started working on the Gunn effect at Varian and a couple of my most important papers on that subject were written while I was at Fairchild.
Vardalas:
Actual publications?
Kroemer:
Yes. They are under the Fairchild byline. And I continued to work on the Gunn effect at the University of Colorado but it was getting into increasingly subtle details. When I came to UCSB I decided, "That's it."
Compound semiconductors and research at University of California Santa Barbara
Vardalas:
When you left the University of Colorado to go to Santa Barbara you insisted that they choose a specialty in which they could realistically be successful rather than being like all the others. What was that about?
Kroemer:
Yes. I had an interview at CU with Ed Stear. We had a long discussion. He was at that time chairman of the department here at Santa Barbara. He had heard that I was interested in returning to the west coast so he visited me in Boulder. I knew a little bit about UCSB. They had a very, very good silicon technology teaching lab. But I wasn't at all impressed by the research they were doing. When Ed Stear visited me in Boulder he said, "Well, you know our solid-state laboratory," and I said, "Yes." He asked me, "What would you do with this laboratory?" I very quickly forgot this was a job interview and said, "Well sure as hell not what you are doing." He acted very upset and asked, "Why? What do you mean?" I criticized their research work. So this led to a horrible, unfriendly discussion for a while. I had decided what the hell. And then at one point he looked at me and said, "Oh, shut up." I thought, "Well, all right. He doesn't want to hear what I have to tell him." But very quietly he said, "Herb, I am looking for someone to rock the boat. You sound like you are my man. Back to my question. What would you do?" So I said, "All right. I know what everybody else is advising. Everybody is advising you to get into the mainstream silicon technology. Don't. It's too late. It's too expensive. And most importantly, the graduate program depends on being able to attract top graduate students. You will not be able to do this one and they will all go to other places."
Vardalas:
Right.
Kroemer:
I said, "My own interest is in compound semiconductors. I think this is going to be an important field. Different from silicon. It's not something in which everyone has to be involved. This discussion took place in the fall of '75. At this point there were three universities that had critical mass – critical mass meaning more than two professors. Stanford, Illinois and Cornell. There is a place for a fourth. If you are willing to put all your eggs in one basket and you are going to gamble, you have a 50/50 chance of making number four. If you go into silicon technology you have a 100 percent chance of being an also-ran. And that still was going to be a gamble. So that's what we did. We still don't have any silicon technology.
Vardalas:
That's what actually brought you here.
Kroemer:
I was basically the one who set the strategy for this concept.
Vardalas:
I see. Okay. In terms of your work here, what interests and fires you now?
Kroemer:
Well, it's a variety of things. My most recent papers have been on superconductor-semiconductor combinations basically taking an indium arsenide/aluminum antimonide quantum well that is heavily modulation-doped, so that there are large electrical concentrations in the quantum well but still very high electron mobility.
Vardalas:
Okay.
Kroemer:
Then on that indium arsenide is put niobium electrodes. Indium arsenide has the fascinating property of not making Schottky barriers, so these are true ohmic contacts. Down to 9 Kelvin these are simply ordinary resistors. Then as soon as the niobium becomes superconducting the conductivity properties of the indium arsenide itself change. Eventually you have a form of induced superconductivity through the indium arsenide. That's a fascinating phenomenon.
Vardalas:
What are the theoretical issues that interest you in that?
Kroemer:
The theoretical issues are that we don't understand the details. We do not understand the temperature dependence. I think we understand the basic physics that is behind the electron transport. It's a phenomenon called Andreev reflections that plays an important role. Others have studied Andreev reflections, but in other systems the transport is diffusive. In other words the diffusion process, how the carriers get from one electrode to the other.
Vardalas:
Right.
Kroemer:
And the indium arsenide is ballistic.
Vardalas:
Ballistic?
Kroemer:
Yes. Therefore we are building what in the jargon would be called ballistic weak links. That topic is still fascinating. We have done quite a bit of experimental work here, but that was basically finished in around '99. It was finished when the Office of Naval Research pulled the rug.
Research funding and applications
Vardalas:
How much of your funding comes from the defense- and government-related work?
Kroemer:
At that point my ONR funding was the basic money. I have done other work under industrial funding.
Vardalas:
How does one sell that kind of research to ONR?
Kroemer:
You would have to ask ONR why they cancelled it.
Vardalas:
Okay.
Kroemer:
They initially supported it.
Vardalas:
What did they see in it? Obviously they saw something as being useful for them initially.
Kroemer:
I do not know whether the original reason for funding it had anything to do with utility. ONR has a strong tradition in supporting things that are quite fundamental. I never made any promise for an application. I simply felt it was a sufficiently crazy phenomenon that it deserved study. That was my last major experimental project here. A couple of Ph.D. dissertations came out of that.
Vardalas:
I am fascinated by Kroemer's lemma on new technology. A lot of people quote it in one form or another.
Kroemer:
Yes, because I've been citing it often enough now. It was originally formulated at a NATO Advanced Research Workshop in France.
Vardalas:
For the purpose of this transcript let me quote it here. Let's see if I got it right. "The principal applications of any sufficiently new and innovative technology always have been and will continue to be applications created by that new technology."
Kroemer:
And the emphasis is on the word created.
Vardalas:
What prompted you to come up with this and how?
Kroemer:
What prompted me to come up with this is that, particularly in engineering research projects, too much emphasis is put on the applications that can be envisaged. But in the really big stuff the applications are always created. The computer was created by silicon. The portable computer was created by liquid-crystal development.
Vardalas:
Was this a defensive move on your part?
Kroemer:
The laser created the compact disc and fiber communications.
Vardalas:
Obviously you came up with this to counter some other perception.
Kroemer:
Absolutely. Yes.
Vardalas:
Which has funding implications.
Kroemer:
I grew up in a research environment where you were quite free and it was not difficult to get funding without having to promise immediate applications. To a large extent that has disappeared.
Vardalas:
Even in universities?
Kroemer:
In the funding for universities, yes, particularly on the engineering end. I think physics doesn't see that to the same extent. If you realize that any sufficiently new invention or discovery the principal applications always have been created, then you understand the importance of this type of research. In addition, the creation typically happens through somebody other than the original researcher. It takes a different kind of mentality. It takes somebody who says, "Hey, this is an interesting discovery. I know exactly what to do with it." If the researcher himself must tell the potential funding organization the applications, then progress is actually being restricted. It is not being advanced.
Vardalas:
That's a good point.
Kroemer:
That is what I keep hammering on. The Nobel Prize has helped me in getting a few more people listening to that.
Vardalas:
Yes. A certain credibility grows around you now.
Kroemer:
My success was basically achieved by following this principle. I never gave a damn what the applications were. I would ask myself what might the applications be, but that did not control what I was doing.
Vardalas:
Right. I see. Another lemma of yours is the proof of ignorance.
Kroemer:
Kroemer's lemma of proven ignorance. "If, in discussing a semiconductor problem, you cannot draw an Energy Band Diagram, you do not know what you are talking about."
Vardalas:
Did you formulate this lemma expressly for your students or for colleagues?
Kroemer:
It was not formulated for anyone in particular.
Vardalas:
Why was there a need for this? How did you come up with this?
Kroemer:
Because I have seen it too often that people cannot draw Energy Band Diagrams, and in reality those same people do not understand how things actually work. I really do feel that knowing how things work is absolutely essential. And without Energy Band Diagrams one does not understand heterostructures. I had this in my Nobel lecture. There is also the corollary that, "If you can draw it but don't, then your audience won't know what you are talking about." The students are always chuckling. If someone gives a talk and doesn't draw an Energy Band Diagram they know Herb Kroemer is going to raise the question, "Would you please draw an Energy Band Diagram?" And my colleagues agree with me. They absolutely agree with me.
Vardalas:
This would be something that perhaps would be something to apply to engineers working in the area who might be tempted to discuss things without ever going back to these first principals.
Kroemer:
That's right. Engineers are particularly prone to not doing Energy Band Diagrams, but they are not the only ones.
Vardalas:
Physicists will do it too?
Kroemer:
You will find this among physicists too.
Vardalas:
Interesting. I have one last question for you. In generations to come some young physicists will look to the greats for a role model and some will choose you. What traits do you hope they will try to emulate? You chose Niels Bohr for certain reasons. Why would you like them to see in you as a physicist?
Kroemer:
That's hard to answer. One of the things that made me tick, certainly skepticism towards what authorities say and a deep interest in really understanding on a fundamental level what I have done. These are probably the two important things.
Vardalas:
I want to thank you very much for a most fascinating interview.
Kroemer:
You're welcome.
- People and organizations
- Scientists
- Universities
- Materials
- Conductivity & superconductivity
- Semiconductor materials
- Computing and electronics
- Solid state circuits
- Transistors
- Communications
- High-speed electronics
- Integrated circuits
- CMOS integrated circuits & microprocessors
- Bipolar integrated circuits
- Compounds
- Gallium compounds
- Silicon compounds
- Lasers, lighting & electrooptics
- Lasers
- Engineering and society
- Military applications
- World War II
- News
