Notice: We are in the process of migrating Oral History Interview metadata to this new version of our website.
During this migration, the following fields associated with interviews may be incomplete: Institutions, Additional Persons, and Subjects. Our Browse Subjects feature is also affected by this migration.
We encourage researchers to utilize the full-text search on this page to navigate our oral histories or to use our catalog to locate oral history interviews by keyword.
Please contact [email protected] with any feedback.
Credit: Science Philanthropy Alliance
This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.
This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.
Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.
In footnotes or endnotes please cite AIP interviews like this:
Interview of Marc Kastner by David Zierler on May 5, 2021,
Niels Bohr Library & Archives, American Institute of Physics,
College Park, MD USA,
For multiple citations, "AIP" is the preferred abbreviation for the location.
In this interview, David Zierler, Oral Historian for AIP, interviews Marc Kastner, Donner Professor of Physics at MIT and senior science advisor to the Science Philanthropy Alliance. Kastner explains the nomenclature transition from solid state to condensed matter physics, and he surveys the interplay between theory and experiment in his field. He recounts his childhood in Ottawa and the influence of his father, who was an experimental physicist, and he explains the opportunities that led to his admission to the University of Chicago. Kastner explains his decision to remain at Chicago for graduate school to work under the direction of Hellmut Fritzsche on optical properties of semiconductors under pressure. He discusses his postdoctoral appointment at Harvard to work with Bill Paul on amorphous silicon, and his connection to David Adler who facilitated his faculty appointment at MIT. Kastner describes his work on amorphous semiconductors and transient excitation and his collaboration with Bob Birgeneau on high Tc. He discusses Joe Imry’s work on heterostructures and subsequent research on the Kondo effect, and how he came to understand the significance of his discovery of the single-electron transistor. Kastner discusses his tenure as department chair, director of MRSEC, and dean of science, and he explains his decision to retire and to join the Science Philanthropy Alliance. He describes his current work with his former student David Goldhaber-Gordon and his excitement over the current research on twistronics. At the end of the interview, Kastner reflects on the role of luck in his career, the centrality of technological advance in his research and what we can learn about physics more broadly as a result of the single-electron transistor.
OK, this is David Zierler, oral historian for the American Institute of Physics. It is May 5th, 2021. I am delighted to be here with Professor Marc A. Kastner. Marc, it’s great to see you. Thank you for joining me today.
Great to see you too.
Marc, to start, would you please tell me your titles and institutional affiliations? You’ll note I put an “S” on everything because I know you’re at more than one place.
[laugh] I’m Donner professor of physics emeritus at MIT, I’m an adjunct professor of physics at Stanford, and I’m a senior science advisor to the Science Philanthropy Alliance.
Now, in your advisory capacity, is this full-time work?
Not anymore. I was the first president of the Science Philanthropy Alliance from March of 2015 until December of 2019. I retired, and I was recently called back for a short stint to be the interim executive director. But, fortunately this week, France Córdova was named the president of the alliance, so I’m off the hook.
I’m still—I still work with the alliance as an advisor.
Now, in your role as an adjunct professor, what are you teaching these days, presumably over Zoom, I guess?
Everything’s over Zoom, but I’m not actually teaching. The way I say it is my former graduate student, David Goldhaber-Gordon, who’s a professor at Stanford, has taken me on as a postdoc.
I’m not a very good postdoc because I don’t work in the lab, even before the pandemic. But I do talk to David’s grad students, and advise them, and help them write their papers and so on. It’s been a lot of fun.
Marc, tell me about the Donner professorship. Who is or was Donner or the Donner family?
I don’t actually know. This was a professorship that has been at MIT for a very, very long time.
I think the Donner foundation created professorships at MIT and Harvard. When I was named the Donner professor, which I think was ’87 or ’89, I learned that the first holder of the Donner chair was Shannon, and the second holder of the Donner chair was Marvin Minsky. So, what I said at the time was that the first holder of the chair was the father of information theory, and the second holder was the father of artificial intelligence, and the third holder, which was me, was the father of two kids in Newton. [laugh]
[laugh] Well, based on the conversation that we’re about to have, I’m certainly going to reshape your self-perception, I can assure you of that.
Marc, to start, I’d like to ask, before we develop your personal narrative, and go back to Canada, I’d like to ask some sort of broad questions that I think will punctuate our discussion, and give a greater sense of your perspective on things, both sociologically and scientifically. So, the first is one simply of nomenclature. Where do you see the transition between solid-state and condensed matter, and where is that simply just words, and where is there a scientific heft and thoughtfulness behind this transition?
I think this happened in the 1970s, about the time that I was starting my career at MIT. There were two themes. One was solid-state physics, which was dominated by the electronics industry. But there was another community that was studying liquids, and foremost superfluid helium. Great strides were being made theoretically in superfluid helium and other liquids to understand their phase transitions. And it became clear that phase transitions were more universal than those things that you see in liquids or gasses. We knew that in solids, there were also phase transitions, magnetism the foremost among them. And it was the unification of those ideas of phase transitions that made the community realize that you should really think of all condensed systems in somewhat the same way.
Of course, in condensed matter, as in every area of physics, fundamental advance and discovery requires the interplay of theory and experiment. Where in your career have you seen bursts of advancement where sometimes the theory is providing guidance to the experiments, and where have you seen when the experiments or the experimenters have provided guidance to the theorists?
I would say in my experience, it has been mostly the latter. And I think if you look at the flow of physics as a whole, I think you could argue that most of the time it’s new technology that makes it possible to learn new things experimentally, and the theory comes in later. Everyone thinks of Einstein, which is the exception, and there’s always a very large community of theorists, especially young ones, who think that they can figure out everything in advance. But I think if you look at the history of quantum mechanics, or you look at the history of my field, condensed matter physics, it’s usually experimental discoveries that come first. There are a couple of examples recently where the opposite has been true. One is topological insulators, which were predicted theoretically before they were found experimentally. It turns out they were materials that we knew about for a very, very long time, but we didn’t know that they were topological insulators until the theory was developed. Another example is this new area, which my colleague Pablo Jarillo-Herrero has invented, which is now called twistronics, which is the properties of graphene when you take two layers and put them together at a small twist angle. Many of those properties were predicted by Allan MacDonald and his colleagues, and then Pablo went into the lab and made it happen. So, there’re some examples where theory comes first. But in the things that I’ve worked on—I can think of only one example where the theory came first. In my early work on amorphous semiconductors, theory almost never played a role [laugh], maybe a really minor role. In high-temperature superconductivity we still don’t have an accepted theory after 30 years. In nanoelectronics we did some things on the Kondo effect, and for that there were predictions before we did the experiments. But the actual making of, what I called at the time, artificial atoms came about accidentally. So, I would argue that most of the time, it’s experiments that come first.
Marc, as an academic scientist coming from an intellectual tradition that prizes basic research, in other words, just figuring out how nature works, I assume for you and your research, the motivation is essentially that. But where have you seen in your career, particularly your area of expertise, opportunities of application, even patents, even places where there might be commercial viability for the things you’ve done? What have been some of those instances, and how have you navigated that decision-making in the landscape?
I’ve always liked working on things where one could imagine that there might be at some distant time in the future applications. Whenever it got close, I moved on to something else.
Not your bag?
No, you know, I really wanted to understand what was going on, and I lost interest when it got to be so detailed that you had to make things really work reliably.
I did see some of my work lead to applications in amorphous semiconductors. We developed an understanding of materials which were very important in optical memories. More recently, they’re being used in phase-change memories in the computer industry. So, I’ve seen things evolve out of our work that became of practical importance, but I went on to other things long before that happened.
[laugh] Well, Marc, let’s take it all the way back to the beginning. Let’s go to Toronto, and start first with your parents. Tell me about them and where they’re from.
My father was born in Germany but came to Canada when he was about 6 or 7 years old. His father told him that he saw the Holocaust coming even in the early 1920s because he had actually fought in the First World War in the Austrian army, and he saw antisemitism growing dramatically in Germany after the war, and decided he had to get out. He came to Canada, and so my father came at the age of 7.
Why Canada? Was there family there?
[laugh] My grandfather’s an interesting character. He had studied opera in Vienna before he got drafted into the Austrian army. And he failed at almost everything. He didn’t make it as an opera singer. He went to Palestine to try to be a pioneer, and failed at that. And then he came back and tried to sell furniture in Germany, and failed at that. So, he just then decided that he wanted to get out of Germany. There was no immigration possible to the United States in the early 1920s. And the only way he could go to Canada is if he had someone sponsoring him with a job. So, he told them he was a cantor, and he got a job as a cantor in a synagogue in Winnipeg [laugh].
[laugh] That’s pretty good.
That’s how he got out of Germany. It’s a good story, right? [laugh]
It was wonderfully lucky because it meant he had a paying job throughout the Depression.
And he had 13 brothers and sisters, and he managed to bribe authorities to get a number of them out of Europe and to Canada before the Holocaust.
What about your mom?
My mom was born in Canada. Her father left Russia just before the First World War. He had been in the Russian army earlier, and was afraid he would get drafted again, so he fled to Canada. He had relatives there, so he was able to get there. But he had to leave his wife and his oldest daughter behind in Poland, and they lived in Poland through the First World War. And then shortly after the war, they came over to join him, and my mother was born soon after that.
Where did your parents meet?
They met in Toronto. My father was in the Canadian army, and I think there was something like an officers’ club for Jewish soldiers or something, and I think that’s where they met.
Did your father serve overseas during the war?
No, he was doing research in Canada. He was doing work to try to figure out what tank treads the tanks should use when they invaded Japan. And if the bomb had not been dropped, he would’ve been one of the first people on the beach [laugh] in Japan. So I am grateful [laugh] to the US atomic energy effort.
What was your father’s job post-war when you entered the scene?
Well, he was a graduate student then when I was born. The family joke was that my initials are M.A. because he got his master’s degree when I was born. And my sister was supposed to be Paul Henry David, but she didn’t work out.
[laugh] What neighborhood did you grow up in Toronto?
It was right downtown. I don’t know Toronto very well. But it was close to the university.
Was it a particularly Jewish neighborhood? Was it more diverse?
I think so. But we moved to Ottawa when I was about 4 years old or 5 years old, so I don’t remember Toronto much at all.
And then you spent the rest of your childhood in Ottawa?
No, we lived there only till I was 7. My father had a job at the National Research Council in Ottawa. He actually worked with Herzberg when he first came there. And then there were really no permanent jobs in Canada, so he decided he had to go to the United States. And he started applying for jobs. Even that was difficult. There was a case where he applied at an oil company, and they had offered him a job, and then they reneged and said they wouldn’t hire Jews. This was 1952, and there was still a lot of antisemitism. He wound up getting a job at General Electric in Cleveland, where there was a huge lamp division. That’s where they made a lot of the GE lightbulbs, and he got a job doing research there. And, so, I really think of growing up in Cleveland because I lived there from age 7 until 16. Then my father got a job at Argonne, and we moved to the Chicago area.
Would you have considered his career track that of an engineer, a material scientist?
No, he was an experimental physicist. He focused on radiological physics, studying the interaction of radiation with matter, and in the later years especially with biology, with humans. He became an expert at measuring radiation dosage. Even in Ottawa, he was working with radioactivity. The most exciting thing I remember from Ottawa was that some kid broke into his lab, and stole some radium needles.
Everybody was very worried about this kid getting sick. I remember the Mounted Police picking up my father in the morning to take him with his Geiger counter to look for these needles. I didn’t believe he was a Mountie because he wasn’t wearing a uniform. So, he opened the trunk of his car, and showed me the red coat—so I’d believe him [laugh].
Was your father’s style to include you? In other words, did you grow up knowing what an experimental physicist did?
I did. I did. And, you know, the most important thing to say about all of that was that we lived in a what you would now call a lower middle-class neighborhood. It was very small houses—individual houses, but very small. And our neighbors were working class people. Our next-door neighbor with whom my father was close was a plumber. The guy that lived across the street worked in the steel mills. It was a time when if you had a high school education, you could get a job that could allow you to live in a nice house. But my father was so far more educated than anyone I knew, any parents of my friends. I thought he was a kind of genius. He taught me how to work with tools and took me to his lab and showed me how he built instruments. He was very much an experimentalist. And I had just decided early on I was not going to be a physicist because—I couldn’t compete with this guy who was such a genius.
[laugh] Did your mom work outside the home at all when you were growing up?
Not at first, no. By the time I was in high school, she was going back to college. She had started college, but stopped when she had to take care of her mother. And then she completed her college degree only when I was a graduate student. So, she didn’t work until after that.
Was your family Jewishly connected when you were growing up? Were you members of a synagogue? Would have you—?
—Pesach dinner, those kinds of things?
Oh, yes, absolutely, absolutely. I was very much into it. I went to a Hebrew school through high school and the first year of Hebrew college, and went to a Hebrew-speaking summer camp, and, you know, conducted services. Even when I was a graduate student, I made some extra money conducting High Holiday services.
Really? Can you lain [read from the Torah]?
I could at the time [laugh]. I don’t remember anything. [laugh]
[laugh] Marc, when did you start to get interested in science yourself, and think about this might be a career track for you?
My father was very smart. He didn’t try to push me to be a scientist. I was planning to be a lawyer. I had developed this plan quite thoroughly. I had a friend whose father was a lawyer, and he took us to see some trials, and I had read something about law, and I just thought this would be a great career. But my father said, “Fine. It’s great to be a lawyer, but most people who are lawyers start out by studying history or English. And then if law school is not for them, they have nothing with which to get a job. So, you should get a degree in science because then you could be a patent lawyer. And if you decide not to be a lawyer—you can get a job.” So, I thought that made sense. I never liked writing long papers anyway. So, I decided to major in chemistry because that’s the subject I had done best in in high school. And, so, I majored in chemistry. And it was really my first physics course. Maybe it was my first chemistry course where we started learning some physical chemistry. That really made me start thinking about being a scientist.
Where did you apply for school for undergraduate?
I only applied to the University of Chicago. It was partly because I had some friends through a Jewish youth organization who had gone there, and I’d heard about it. A big reason was that my father was then working at Argonne, and so we got half tuition there. But it just seemed like an exciting place to go, and I had just moved to Chicago. I liked the city of Chicago very much at the time. I only applied there early admission, and got in, and never applied anywhere else.
Was the draft something you needed to contend with?
Not at that point. At that point, there were student deferments. And it was only when I became a graduate student that I had to start worrying about the draft.
At what point as an undergraduate did your interest in physical chemistry start to seep into pursuing physics itself?
It was really taking electricity and magnetism. As a chemistry major, I had to take physics courses. And I had taken a terrible physics course in high school, which was just mechanics, and I didn’t like mechanics at all. And, so, in my first term of physics, I wasn’t very excited. But when I started studying electricity and magnetism, I was just hooked. And [laugh] I went to my dad and I said, “OK, I want to switch and become a physics major.” And he said, “If you switch majors, you’re going to probably have to take an extra year of college, and I don’t want to pay for an extra year of college.” [laugh] “You can change in graduate school,” he said. [laugh]
Do you remember who your E&M professor was?
I do. It was Courtenay Wright. He actually taught a three-quarter sequence—Chicago’s on the quarter system, and it was an extremely tough first physics course. It was one quarter of mechanics, one quarter of E&M, and one quarter of waves and oscillations. And he was a brilliant lecturer. He was just brilliant. I really loved it.
Did you get any specific advice about perhaps not staying at Chicago for graduate school, that it might be better to expand your horizons a bit, and go somewhere else?
I did. I talked to Robert Platzman. He’s a famous atomic physicist who was at Chicago. I went to talk to him because I was actually thinking about going into biophysics, and he was doing things in biophysics. He helped me apply to UC San Diego, and I was going to go there and work with a famous biophysicist. I actually thought about that, and I just decided biophysics wasn’t mature enough. That it was, you know, it was too nascent, and it wasn’t clear what the discipline was. And that’s probably still [laugh] true. So, I applied there. I applied to Stanford because my father had been in the research group of Art Schawlow. They were graduate students together, and I thought that might help me get into Stanford. I wrote to Schawlow, but he didn’t help.
So, I did apply to a couple of different places. Since I was shifting departments, there was nobody who said, you know, you really shouldn’t stay here. When I told Stuart Rice, who is a famous physical chemist, that I was going to go to graduate school in physics, he was upset. He said, “You should stay a chemist.”
And then there were other people. I did summer work at Argonne first, with a chemist who was doing nuclear chemistry, and then with a guy named John Erskine, who was doing nuclear physics. And when I asked the chemist what I should do, he said, “You should go into physics because there are lots of chemists who know physics, but there are no physicists who know any chemistry.”
[laugh] Marc, on the social and political side of things, were you involved at all in all of the excitement that was happening in Chicago? Were you at the convention of ’68? Did you march on campus or any of those kinds of things?
No, I was really nervous about getting involved in any of this stuff. When we came to the United States, I remember we got our first television. There actually was no television in Canada when we left. And, so, we got our first television set, and my mother had it on watching the McCarthy hearings all the time. The theme in our family was that it was dangerous get involved [laugh], right? I was really very conservative in terms of being an activist at the time. I was liberal. I supported Gene McCarthy when he ran. We went to McCarthy rallies. But I wasn’t a radical. I wasn’t going to be involved in protests.
Now, intellectually, administratively, academically, how did you go about narrowing your interests in graduate school to experiment and solid-state? I imagine that it was not called—there was no condensed matter at the time.
That’s right. It was solid-state. Well, there were several aspects to it. Again, my father had a big influence. He advised me to do experimental work because, as I mentioned, he faced a lot of discrimination. And he said that the attitude was Jews can only do theory [laugh]. Of course, most Jewish kids who’d gone into physics wanted to do theory, right? [laugh]
He also said theory’s just too hard. So, he influenced me in that direction. And then he said absolutely do not do high-energy physics. The culture is terrible. So, those were the two pieces of guidance he gave me. Once I had decided not to do biophysics, I thought the way to get into biophysics was to first get a good grounding in solid-state physics, and then move on from there. And I think actually later on, that turned out to be a good strategy. I know a number of people who did that. But once I decided to do solid-state physics, and I got admitted to Chicago, I just went around and visited all the solid-state physicists, and there weren’t many. There was one guy who was about to leave, and then there was Royal Stark. I visited his lab. It turns out he was away, so he asked his graduate student to show me around. His graduate student was Dan Tsui, who later won the Nobel Prize for the quantum Hall effect. And then I visited Hellmut Fritzsche, and I just really liked him. Also I started reading some of the work he was doing, and I just found the idea of working on amorphous semiconductors very exciting.
What was the sense of opportunity you had in this field? What was there to do, and how might it be something that would be viable as a career track?
Well, you know, as I said before, I was attracted by the idea that it might have applications. The paper that really sparked excitement in this field was a paper by Stan Ovshinsky, who only had a high school education, but invented a new kind of electrical switch. And the claim was this could replace the transistor, so this sounded very exciting. And it raised immediately some very interesting intellectual questions because the paradigm for solid-state physics is that you start with a lattice, and then you solve Schrödinger’s equation on this lattice, and a huge amount of progress is made that way. What happens when you don’t have the lattice? What do you do? And it appealed to me because chemistry now becomes very important, right, because you really have to think about the bonding between atoms rather than thinking about plane waves, electrons moving through a lattice.
So, you really took advantage of your undergraduate background?
Absolutely. It played a really important role. There were ideas that I came up with, which any chemist would’ve come up with, but none of the physicists [laugh]—could think about things that way.
[laugh] What was the experiment or the project that ultimately informed your thesis?
There were a couple of things. Hellmut gave me a first project to work on, and I really hated it. We got a paper out of it, but I really hated it, and I said I really want to do something else. So, he suggested another direction, which was to study the optical properties of these amorphous semiconductors under high pressure. We discovered a compressor up in the attic that came from Willard Libby’s lab [laugh] when he was at the University of Chicago in the 1940s or something, and we bought a small amount of equipment, and I put that together, and I was doing those measurements. But, in the meantime, I was thinking about these chemical bonding ideas that I mentioned, and I actually came up with a new way of thinking about this particular class of materials called chalcogenide glasses. They’re glasses. They’re amorphous. They’re called chalcogenide because they contain sulfur, selenium, or tellurium, the chalcogens. And these were the first materials Stan Ovshinsky worked with. And I came up with an idea about the chemical bonding of these materials, and why they would be different from other kinds of semiconductors like silicon and germanium. I told Hellmut about this idea, and, at first, he didn’t understand it and, you know, he didn’t know what to make of it. He was very polite, but it was clear it didn’t have a resonance. He was working as a consultant for Ovshinsky’s company, and he saw Ovshinsky on a monthly basis. On his next visit to Ovshinsky, apparently he told Ovshinsky about this, and Ovshinsky went wild. This was the way he thought about things too. So, Hellmut came back, and told me I should write it up. So, I wrote up this completely theoretical model. It was very simple. And then I used those same ideas to try to understand the optical properties of these semiconductors under pressure. So, my thesis was a combination of those things.
Now, how closely was this related to what Hellmut was doing himself at this point?
Very much, very much. His entire group was working on the properties of these chalcogenide glasses at that time. There were seven or eight graduate students, and we were all doing things related to that.
Who else was on your thesis committee?
Oh, Morrell Cohen, who was a famous theorist. You must’ve—probably have him in your interviews.
I think there was a high-energy physicist. I think it was Peter Freund. He fell asleep during my thesis [laugh] review. That’s all I remember. I think there were just the three of them.
What postdoc opportunities were available and compelling to you at this point?
I had quite a few. I visited Jan Tauc who was a famous experimentalist in amorphous semiconductors at Brown. Marvin Cohen wanted me to come for a Miller Fellowship at Berkeley. And then there was Bill Paul at Harvard, which was the one I finally wound up taking.
What was compelling about Harvard?
Well, it wasn’t—you know, each one of these had some flaws [laugh] in my view. The Miller Fellowship sounded great except I would have to have found a lab to work in. There was nobody working on amorphous semiconductors there that I would obviously work with, and I was a little nervous about not having a lab to go to. Jan Tauc was a very nice guy, but I wasn’t that excited about what he was doing. I asked Hellmut’s advice, and he said, “Well, you can’t go wrong if you go to Harvard. It’s such a great place.”
And what was Bill Paul doing at that point?
Bill Paul was working on amorphous semiconductors, but not chalcogenide glasses. He was working on amorphous germanium. And that was not a great experience for me, that time with Bill Paul, but I got out of there very fast [laugh].
Did you see this time at Harvard as an opportunity to expand what you were working on, or to improve and refine your thesis research?
No, I was trying to do something different. The history was that in 1972, Spear and LeComber announced that they had made amorphous silicon, and had been able to make a field-effect transistor with it. My first experiment with Hellmut was an effort to make a field-effect transistor with amorphous germanium, and it was a complete disaster. I was convinced you just couldn’t make a transistor with amorphous germanium, so when this announcement came that they did it with amorphous silicon, I couldn’t believe that either because amorphous silicon was so similar to amorphous germanium.
Now, Marc, just to interject, is this—when you say you couldn’t believe it, is it because you just thought the science didn’t work, or were there limitations with the technology available at the time?
No, I just thought the science didn’t work.
Basically, if you think about trying to take silicon atoms, and put them together in a disordered way, you’re always going to have some dangling bonds. In crystalline silicon, every silicon atom is surrounded by four other silicon atoms at the corners of a tetrahedron. You can’t do that if you make it amorphous. There’s going to be some tetrahedra that don’t have all four bonds. And, so, I believed that you would always have so many dangling bonds that it would soak up all the extra electrons, and you couldn’t get any electrons that would move around. It’s as simple as that.
So, when Spear and LeComber announced this, that they had made the silicon out of silane, which is a molecule with one silicon atom bonded to four hydrogen atoms, I believed immediately that it must be that you have a lot of hydrogen in there satisfying the dangling bonds. And that turned out to be the right answer, but they didn’t want to believe it for, somehow, there was this prejudice in the semiconductor community that the best silicon had to be pure. And if it had hydrogen in it, it wasn’t pure. I discussed this with Hellmut, and he put a student to work, and they proved that there was hydrogen in this stuff. The way it happened was that a friend of mine, Steve Hudgens, who was my fellow graduate student, was trying to measure the magnetic susceptibility of silicon made from, silane. He put it in a sealed quartz ampoule to put it in the susceptometer. And when he warmed it up, the ampoule exploded [laugh] because so much hydrogen came out. So, Hellmut then got to work, and proved that there was hydrogen in all of this stuff. This was happening just at the time I moved to Bill Paul’s lab. Bill Paul believed that you had to make the purest, in this case, germanium instead of silicon. And I said, “We should try to put hydrogen in it,” and he said, “No, you can’t do that.” He would not allow me to do it because he thought the germanium should not be contaminated with hydrogen. My job was to build a sputtering machine to make the pure germanium, so I made an extra access port to the sputtering machine, which allowed me to put hydrogen in in the middle of the night, when Bill was not around. Then I left to my job at MIT just as I had achieved this, and my fellow postdoc then carried it on. Bill Paul’s whole research program for the next few years focused on hydrogenated amorphous germanium.
What was your first point of contact at MIT? Was it sort of in your world, just by virtue of being at Harvard, or was there a specific connection you had there?
There was a connection. It was David Adler.
You probably have heard of him.
There’s an Adler Award at the American Physical Society. He was a theorist who was actually in the electrical engineering department. At MIT in the ’60s, the high-energy physicists decided that solid-state physics was engineering, and so people like David Adler and Millie Dresselhaus had to get jobs in the electrical engineering department. But I got to know David quite well because he was also working with Ovshinsky, and he knew about my ideas about the chemistry of chalcogenides. But my real contact with MIT came with Peter Wolff, who was trying to build up solid-state physics in the physics department after they recovered from this craziness. He was looking for job candidates, and he contacted Hellmut, and Hellmut, you know, said nice things about me. So, Peter was my first actual formal contact with MIT.
In what ways immediately did it dawn on you that MIT was working for you in ways that Harvard was not?
Well, you know, it was just talking to Peter, even during the job interview. Peter came out of Bell Labs, and he was—he had been a manager at Bell Labs, and he was the best recruiter I’ve ever met [laugh]. He just made me feel that I was important and I was doing great things. My first couple of years at MIT were tough, but it was just so exciting to be on my own and not have Bill Paul telling me what I couldn’t do.
In what ways personally did you click with Peter Wolff so well?
He was just charming and brilliant. He was soft-spoken, but really very thoughtful. You just had the sense that he really wanted you to be successful. He helped me a lot when I got started. He got some internal funding for me. These were not days when you had big startup funds [laugh]. I had $20,000 or something like that, but I just felt I could trust him.
And this was a second postdoc, or was it a proper—
Oh, no, no.
This was a faculty appointment. The amusing part about that was I interviewed there in April. I went to Harvard in August of 1972. And by December, I told Hellmut that I really needed to get a faculty position [laugh] quick. So, he started telling people about me. I had interviews at Illinois, and City College of New York, and at MIT. And I really wanted to go to MIT, but I didn’t hear anything and I didn’t hear anything. And then Peter sent me a note saying—or called me and said that they decided to make an offer to somebody else. So, I was about to accept a job at Illinois when he called back and said their first choice turned them down, and would I still be interested? Interestingly, because of the Bell Labs influence, he said there was a condition, and the condition was that I had to not work on amorphous semiconductors.
Wow. What’s the story there?
The story is that the Bell Labs people, particularly Jim Phillips and, I think, even more Phil Anderson, really thought Stan Ovshinsky was a charlatan, and that there was nothing to this amorphous semiconductor stuff. They thought Stan’s switching was just really thermal runaway, and that he was just trying to make money. All the Bell Labs community felt that amorphous semiconductors wasn’t exciting. [laugh] Ironically, in 1975, it was Phil Anderson who revitalized it as a field of importance in physics, and I can tell you that story later. But Peter was worried that I would be going into a field which had no future, you know. Bell Labs didn’t think it was good, so it wasn’t good. And he was really just trying to protect me. I had done some work using these chemical bonding ideas on crystalline materials, and he thought that was a better way to go. So, he said if I wanted to do something along those lines, I could have a job. So, I figured, well, try something new.
And what was the state of play more broadly at that point with lone-pair semiconductors?
You know, I think it was generally accepted that this was an important idea for chalcogenide glasses. But most of the amorphous semiconductor community by this time was focused on hydrogenated amorphous silicon. So, people weren’t paying that much attention to it, except Stan Ovshinsky, who kept talking to me about it [laugh]. As you know, I was an assistant professor, and he invited me out to his company with David Adler, and we spent, hours just talking about the properties of these materials, and how you might do something with them.
Now, I know you were forbidden from direct research, but in what ways were you intellectually connected with what was going on in amorphous semiconductors during these early years?
It really is all because of Stan Ovshinsky. Because he really believed that my ideas were important, he invited me to visit him, as I said. And then in 1976, he invited Nevill Mott to visit Energy Conversion Devices, and created a workshop for people to talk to him. It was all of the leading people in the field. Morrel Cohen was there, Artie Bienenstock was there, of course Hellmut, David Adler, a whole bunch of all my heroes. These were all people much older than I, much more senior than I. But Stan insisted that I be included because he liked my ideas. Just before that, Anderson had published a paper on negative correlation energy. And this was motivated by the problem of amorphous semiconductors. He thought that maybe the reason that these chalcogenide glasses don’t have dangling bonds—because we knew that silicon and germanium did, but these chalcogenides didn’t—maybe there was an attraction between the electrons that paired them up, so you didn’t have unpaired electrons. And that inspired Mott and a young colleague named Bob Street to write a paper explaining in a more microscopic way where this negative U (negative correlation energy) was came from. As I remember, the whole discussion at this workshop was about Mott’s model, and it started with the idea that there were dangling bonds. Hellmut pointed out that there was a flaw in this argument, that it just couldn’t be right. And the two of them, Hellmut and Mott, were at the blackboard, and—
Was your sense that they were at the blackboard as equals, as peers?
Yes, they were. Well, it was the theorist and the experimentalist, right? Now, Hellmut was saying, “Here are the facts,” right?
“How can this be? Here are the numbers.” And Mott was arguing, well, sort of theoretically, that he thought might be the case. As you know, Mott was not a very mathematical theorist. His approach was very phenomenological, and that was his strength. And, suddenly, in the middle of this discussion, I had an idea of how to resolve this. And I almost blurted it out, but I didn’t say anything because Hellmut had told me many times that Mott took credit for anything that he was involved in. For example, there’s a large number of books whose authors are Mott and X, and nobody remembers X, right?
There’s Mott and Jones, and Mott and Davis, and Mott and Gurney. He constantly worked with people, and got credit for everything. So, I knew if I mentioned this idea to Mott, it would be Mott and Kastner, and I would never be known. So, I kept my mouth shut. And on the plane home with David, we worked out this idea in detail. The reason why Ovshinsky’s workshop was on that week was because there was a conference in Virginia the next week, and Hellmut was there too. So, then Hellmut and I and David spent that conference polishing it up, and had this paper within a week.
So, and that really changed things for me at MIT. I mean, at that point, I began working on amorphous semiconductors at MIT, and started doing experiments to follow up on these ideas. But that paper on the valence alternation model, it still is one of my most cited papers.
And where is Phil Anderson in all of this? How are his ideas influencing what’s going on at this point?
Well, you know, he threw out this idea of negative correlation energy, and our paper was a microscopic model of how you can actually get a negative correlation energy, and it comes from the chemical bonding. It comes from new bonds being formed. And, so, we made it more specific and precise. And, you know, I don’t think actually the idea that there’s generally a negative-U is accepted very much anymore. People every once in a while talk about negative-U. But I think [laugh] in the end what negative-U means is that electrons like to form bonds [laugh]. Atoms like to form bonds.
What is your sense of why the research on valence alternation models resonated so well?
[pause] That’s a really good question. You know, I think that it just helped explain a lot of mysteries. And people who were working on these chalcogenide glasses for switching, for optical memory, which, you know, Ovshinsky had already invented, there was a good sized community. And there were things that you just couldn’t understand without this model.
In what ways did this research really set you up on a long-term research agenda at MIT?
It did. You know, there were a lot of experimental questions that we could ask. How do these valence alternation pairs show up in a material when you excite it with light? So, I really started doing a lot of optical experiments with transient excitation. Would you have similar defects in silicon dioxide? Silicon dioxide is a pretty important material, right, it is the insulator in all the transistors in our electonics. And defects in silicon dioxide are critical in the semiconductor industry. So, we started a program trying to measure similar defects in silicon dioxide. We measured photoconductivity and photodarkening, all kinds of things. So I really built up a group based on those ideas.
Did you take on graduate students right away?
Ah [laugh]. MIT at that time had a horrible system. Graduate students were admitted mostly into research groups, so when they arrived, they had a research assistantship right away. There was very little money for teaching assistants. That was used mostly for students in high-energy theory, or students who lost their support and had to move from one group to another. There wasn’t a pool of graduate students for new faculty. And, so, I actually, in my first year, had to recruit students for my second year [laugh]. So, I didn’t have any graduate students in my first year. But once this work came out, I started collecting students who were interested in it. I had one student who had been working with another group, and he just came around to find out what I was doing, and he got excited about it. And I had another student who had worked with me as an undergraduate, and decided to stay in my group as a graduate student. So, the group grew slowly. When I became department head, one of my proudest accomplishments was getting enough fellowship money so every first-year graduate student could have a fellowship, and not be forced to join a group before they came.
Tell me about the work of Joe Orenstein, and why it was important for your research.
So, Joe’s the student who came from another group. He was working [laugh] on something he called breast-o-physics. He was using microwaves to do mammography, and he was really not happy. And he came around to visit me. Before the work on the valence-alternation model, Peter Wolff had helped me hire a postdoc. I convinced my fellow graduate student at Chicago, Steve Hudgens, to join me at MIT. Peter Wolff got money to support him to do some of the experiments on chemical bonding in crystals. But then as soon as this valence alternation model came out, Steve amd I both began working on chalcogenide glasses. I remember we were doing an experiment, and Joe came into the lab, and he looked at the strip chart recorder, and immediately had an insight. And I thought, “This is [laugh] a smart guy.” So, he joined my group, and we began doing optical experiments on chalcogenide glasses. I’ve forgotten where we got the money. Maybe Peter helped again. We got money to get a pulsed laser. And Joe did beautiful experiments on photoconductivity and photo-induced optical absorption in chalcogenide glasses. This was a way of really studying the density of states in the material. That was a beautiful experiment, and we went on doing things like that. We did things like that in other materials, and the next student worked on photoluminescence. So, his work was pivotal in my early days.
And just to zoom out, in the late 1970s going into the 1980s, what opportunities for collaboration were there in industrial research, places like Bell, IBM, things like this?
There was a lot. I didn’t do much with Bell in my earliest years at MIT. I went to visit there a couple of times. And, of course, MIT was like a branch of Bell because so many of our faculty had come from Bell: Bob Birgeneau, Patrick Lee, Peter Wolff. And people from Bell were coming through visiting all the time. I remember Bill Brinkman visiting in my very early years. And, similarly, IBM, I remember meeting Alex Müller before he was famous. My first collaboration with one of those labs came later, in the 1990s, when I started working on the single-electron effects. I didn’t do much until I started working on nanostructures. When we first started working on nanostructures, I was trying to make one-dimensional metals. Again, Peter Wolff played a critical role. By then, he was director the Research Laboratory for Electronics at MIT, which had a block grant from the Defense Department. He brought Hank Smith to MIT’s campus from Lincoln Labs. Hank was the father of x-ray lithography. From my earliest years at MIT, Bob Birgeneau taught me a huge amount. We used to go to lunch every day at Joyce Chen’s Small Eating Place, which was this restaurant up on Mass. Ave. When I first came to MIT, everybody was going to the student cafeterias. Bob came two years after me, he announced that this was just not adequate [laugh].
It may have been Jens Als-Nielsen who was visiting Bob who said, “We have to go to a better place.” We would go to this restaurant every day, and I remember Peter Wolff joined us one day. I was the only real solid-state physicist. Bob and all the others were experts in phase transitions. But Peter said, “This guy, Hank Smith, can really make wonderful nanostructures. You guys ought to think about doing something.” I had been reading Thouless’s paper about how all metal wires are insulators if you make them long enough. And I told Peter I thought we could make a one-dimensional wire to test Thouless’s predictions. And he put me together with Hank, and we made a one-dimensional wire. We didn’t have a dilution refrigerator at that point. This was now a couple of years later. We were making these one-dimensional transistors. And one of my students went down to visit Gerry Dolan and Dave Bishop, and they helped us do some really beautiful measurements. At that point, it was the era of universal conductance fluctuations, and we observed universal conductance fluctuations in a transistor for the first time. So, that was an important collaboration with Bell Labs.
I wonder if you can talk more broadly about the significance of what was happening in phase transitions at this point.
The heyday of phase transitions was earlier. It was really the late ’60s, early ’70s. The group at MIT that was there before I came, which was dominated by George Benedek and his former students, Dave Litster and Tom Greytak. They were all studying critical phenomena. What are the exponents as you go through a phase transition? And this was a really hot topic. Bob Birgeneau had been studying similar things in magnetism, and I knew nothing about it. So, these lunches were a wonderful education for me. Dave Litster was studying liquid crystals. Bob was getting into x-ray scattering, and studying phase transitions in two-dimensional systems, like xenon on graphite. Paul Horn came for a sabbatical, and worked with Bob on that. And Tom Greytak had been working on phase transitions in liquid helium, and was just beginning his efforts with Dan Kleppner to try to see a Bose condensation in hydrogen. So the whole community around me was working on phase transitions.
Now, on reading Thouless’s paper, and your idea to develop a one-dimensional transistor, what was provisional about this idea? In other words, what did you not know until you actually did it, and what were you sure of at the outset of this project?
Well, the idea was to try to have a wire in which you could control the electron density, and therefore the conductivity. And to test Thouless’s ideas that if you added electrons to a wire and make it more conducting, it would still become an insulator but at a lower temperature. We were never able to test any of his ideas because other things happened. What we immediately were confronted with was that disorder dominated everything, and the wires were either strongly insulating or weakly insulating, and we were studying the transition between those two cases. It was all surprising. But the biggest surprise, of course, was when we discovered a periodic oscillation of the conductance as a function of gate voltage, which we didn’t know at the time but was really single-electron effects.
In what ways were your graduate students really not just learning, but they were adding to this enterprise, they were helping you?
Oh, my graduate students always were on the front lines of discovering things. You know, I was never very good in the lab. My father never could believe that I was going to be an experimentalist because I—
[laugh] And he would know [laugh].
—because I was just all thumbs, you know. When I started [laugh] working with Hellmut, as I mentioned, he wanted me to try to make a field effect transistor, and his idea was to deposit the amorphous semiconductor on top of a ferroelectric to get a lot of charge into the semiconductor. Bit I kept breaking the ferroelectrics [laugh], and hehe complained to the other students, “This guy Kastner breaks everything,” . So, I was never very good in the lab. But I was really good at getting other people involved [laugh], even as a graduate student. My fellow student, Mike Paesler, made all the samples for me. And when I became a faculty member, Hellmut said, “You know, you should have one experiment to do yourself, and have another experiment for your students.” And I tried that, and I just never had enough time to get my own experiment going. So, I just decided I would work through my students. I haven’t actually done an experiment with my own hands since 1973 or so.
[laugh] Marc, how did you and Bob decide to come together to work on high Tc?
Well, we were standing in the hall [laugh]. His office was almost next door to mine. It was a wonderful configuration of a building with the labs in the middle, in the core, and then a hallway that went around in a circle, and the offices were all on the outside, both faculty and graduate students. So the order was Bob’s office, several student offices, and then my office. And when the discovery was announced, Bednorz and Müller had discovered this material, and it was superconducting, everybody was just so excited. And Bob instantly knew that magnetism was important because of the chemical composition of these materials. He knew that a copper compound’s magnetism was going to be important. And he knew that to do magnetism right, you’re going to have to do neutron scattering, which was his specialty. And to do that, you’d need a big crystal.
Fortunately, being at MIT, there was an outstanding crystal growth lab one floor above us. It had been developed to make laser rods for the Defense Department. In fact, they did some of the early development of the material for rods that are used at the National Ignition Facility, a giant facility at Livermore to try to create fusion. Bob knew these guys had the expertise to grow the crystals that we would need of Lanthanum copper oxide, the material in which Bednorz and Müller had discovered superconductivity. Bob went to the provost, and asked for some money to support the crystal growth, and the provost agreed. Instantly gave him 70K or something like that. There was a visitor from Australia who was working in the crystal growth lab, who agreed to shift his focus to grow these crystals. So Bob very quickly had the first and only cubic centimeter sized crystal of lanthanum copper oxide in North America. Bob was going to go off and do neutron scattering, and asked his student to measure the conductivity of the crystal. He wanted to make sure it was superconducting. So the student asked if he could borrow equipment in my lab, and I said sure. He made some measurements, and I came into the lab, and the student really looked depressed. And I said, “What’s wrong?” And he said, “Not only is it not superconducting, it’s not even a metal.” I looked at the data, and it looked just like amorphous semiconductors. It looked like a dirty semiconductor. And I told Bob, “This is my business. I can help,” and we started collaborating. You know, this would not have happened if we weren’t together in the same hallway with the students there.
What was some of the feedback that you were getting outside MIT among your peers? What were they saying about this work?
About the high-temperature superconducting work?
Oh, the neutron scattering got enormous attention. Anderson had already proposed the theory. Everybody proposed a theory, you know. Every theorist went to their desk, and opened a drawer, and pulled out a 10-year-old theory and [laugh] published it, right. Anderson had one from about 1973. It was his resonating valence bond theory, and it made predictions about the magnetism. The first paper that we wrote on the magnetism basically said it doesn’t look like Anderson’s model is right. And, unfortunately, Bob had a family crisis, and was not able to go to the international conference where this work was first discussed, and so he sent me. I had to give a talk on the neutron scattering for which I was really like a first-year graduate student. I knew very little. But I was able to give this talk, and did it in such a way that Phil Anderson was not angry—
—which was difficult.
When did you first meet Horst Störmer?
I think we interviewed him. I think Peter Wolff wanted to hire him. He came to give a talk at MIT long before he had worked on the quantum Hall effect. It was the early work on modulation doping, and so I knew him. I think he may have even been a postdoc at that point. So, it was his early years at Bell Labs.
What were some of his ideas about your work? In what way was he useful to the things that you were doing?
The interaction with him that I remember the most is when we discovered these periodic oscillations in these narrow silicon transistors, and I was trying to understand what it was. I went to my colleague, Patrick Lee, and he said, “That’s physically impossible.” Horst was visiting. I think he had the Nobel Prize by then, or at least we knew he was going to get it. And he came to give a talk, and I showed him the data, and he said, “You know, if I saw that, I would look for the amplifier that’s oscillating.” So he wasn’t very helpful [laugh].
But, you know, I always really liked him. He and Dan Tsui are just some of the most marvelous people in the world.
I remember Horst saying that what they did was not really very impressive. He said, “Anybody who had Art Gossard’s samples would’ve discovered the fractional quantum Hall effect.”
You know, it’s true, but it takes a big person to admit that [laugh].
Marc, I only interviewed Moty Heiblum on Monday, and so I’m curious—
About how we came to work together?
I did, yeah. So, for your perspective on the idea to focus on inverted heterostructures?
Well, you know, this story goes back to Joe Imry. I don’t know if you know of Joe. Joe was a wonderful guy. He was a theorist who started out as an experimentalist, and then switched to theory. He came to give a talk at MIT. We went out to dinner. And after dinner, we were walking to the car, and he pulled me aside, and he said, “There’s a graduate student at MIT who was an undergraduate at Tel Aviv. We would love to get him back to Israel, but he’s studying high-energy theory. And if he’s a theorist, he’s not going to get back to Israel. If he ever changes his mind, and chooses to do experiment, if he comes to see you, take him.” So, sure enough, one day, this kid arrives in my office, and I’m a little nervous because, you know, the Pauli effect, right? You don’t want to have a theorist near your experiments.
So, I started asking him, “What experimental experience do you have?” And he said nothing. And I said, “Did you ever work on cars?” And he said no. It turned out he spent one summer driving an 18-wheeler across the country to make money one summer while he was at the Center for Theoretical Physics [laugh]. I said, “Do you have any hobbies?” He said no. It turns out he played guitar well enough to make enough money to live in Paris for a year on his own. So, he had a lot of skills but he didn’t tell me about them. Then finally I said, “What did you do in the army?” And he said, “I can’t tell you what I did in the army.” So, I knew that was theory [laugh].
But I took a chance, and he was one of the most enjoyable students I ever had, brilliant and talented. So, it then turns out Moty Heiblum came to MIT for a visit, and asked to see this kid, Udi Meirav because Joe Imry was recruiting Moty for the Weizmann, and told Moty about Udi. They got together, and the next thing I knew we were collaborating with Moty’s group. Moty had the idea of making this inverted heterostructure. I’ve forgotten the motivation that he had at the time. It made a lot of sense to me. But it turned out it was a horrendously difficult way of making devices because you had anneal the contacts so they would go through the insulator to contact the semiconductor, but not go too far or you would get a short to the heavily doped material on the bottom. Amazingly, Udi managed to make this work, and when successive generations of students and postdocs, like Paul McEuen, looked at this, they said they would never try it.
Did you ever get to the Weizmann yourself?
I went to visit once. I visited Moty and Udi. Udi was then an assistant professor there. He had been recruited to the Weizmann following Joe Imry’s plan. He was an assistant professor for a while, and then decided to go into the business world.
What was going on more generally with the Kondo effect at this point?
Once everybody was making lateral quantum dots with semiconductors, everyone was looking for the Kondo effect. Patrick Lee and a colleague had written a paper years earlier based on some research that Paul McEuen had done in silicon, in which he had seen tunnelling through an impurity. That was even closer to the Kondo effect than our later work, because it was actually an impurity. Ned Wingreen and Yigal Meir, who were at MIT, were foremost among the theorists working out the details of how a quantum dot would have a Kondo effect. So, it was in the air. Everybody was looking for it. And I was convinced that you couldn’t see it unless you made smaller dots. Just about that time, David Goldhaber-Gordon had joined my group as a graduate student, and Udi came for a visit. I think he was already looking at management consulting firms at the time. But he invited David to come back to spend some time at the Weizmann, and to make very small single-electron devices. David was the first to see the Kondo effect in a dot, along with Sara Cronenwett in Charlie Marcus’s group. But then it just exploded. It’s hard to understand in retrospect why we didn’t see it earlier [laugh].
I’d love to hear about this Woodstock of physics at the APS March meeting. I’ve heard this before. This is a common term associated with this event at that point.
It must’ve been March of ’87, I guess, because in ’86, Bednorz and Müller had announced their discovery. And by the following March, everybody was studying this stuff, just thousands of people all over the world. Everybody submitted abstracts for the March APS meeting, and there were so many they had to have this special session which lasted all night. Hellmut and I were there sitting together for a while. Finally, I got tired and went to bed. And I think it just went on forever. I came away exhilarated but also depressed because this was before Bob had made his single crystal, and I knew this was probably the biggest discovery that would happen during my career. It was really going to change everything. But I didn’t know how to do anything with it because you can’t give graduate students at MIT a project which any high school student can do in their basement, right?
[laugh] You just can’t do that, right. You can’t compete unless you have something special to do. And, so, I actually came back depressed, and that’s why I got so excited when Bob had this crystal, and that was something unique.
Now, you were collaborating with Bob this whole time?
Not really. We never really collaborated until that crystal came to my lab. That was the beginning. We were good friends. We socialized. We went to lunch almost every day. We had department business we did together, and stuff like that. But we never actually collaborated until he got that single crystal growing.
Marc, to put 1990 into context, obviously it’s very hard to separate out what you saw in real time and looking back. But at what point did you recognize that the discovery of the single-electron transistor was sort of in a class by itself?
[pause] That’s a really good question. You know, when we first discovered these oscillations, we didn’t know what they were, and I came up with this crazy explanation that it was a Wigner crystal or a charge density wave. And it was only after Paul McEuen joined my group, and Udi had made his devices in gallium arsenide that we had much better control over it. When Wingreen and Meir were working closely with Paul McEuen and my students it really became clear pretty quickly that it was a single-electron phenomenon. Udi’s devices were so clean that we were able to really get beautiful physics out of it. And then it became an industry, you know. People were doing it all over the world, and it became clear that was something important.
Did you recognize that it would be your calling card, or did you actually resist that, that you wanted to move on to other things?
Oh, no, I just loved it. Most of the work that I’ve done in my career, certainly all the work on amorphous semiconductors, a lot of the work on high Tc, was a matter of taking somewhat ugly data, and getting something interesting out of it [laugh]. But the single-electron phenomena were so beautiful, just aesthetically beautiful, that I just enjoyed it tremendously. Bob and I were still working away on high Tc, and doing interesting things. But I’ve never enjoyed aesthetically anything so much as the work on single electrons.
What was your contribution more broadly to what was going on in neutron experimentation?
Well, I actually did one neutron experiment myself with Bob and Gen Shirane looking over my shoulder [laugh] and telling me what I was doing wrong. But most of the time, my students did optical and transport measurements to characterize the system more thoroughly, and then Bob and I would discuss the neutron measurements. There was one case where we actually discovered a magnetic transition in transport, and that’s the neutron experiment that I actually participated in, to characterize what the phase transition was that we were seeing.
There’s a range of responses when you’re asked to become department chair or department head, ranging from, “Oh, it’s my turn. I guess I’ll have to do it,” all the way to, “This is an opportunity for me to put my stamp on things, and to improve things as I see them.” When you were asked to do this in 1998, where did you fall on that spectrum?
I was ready to do it. At first, I should say that I had already taken on an administrative job, which was director of the Center for Material Science and Engineering. It was a National Science Foundation Materials Research Science and Engineering Center (MRSEC). And I found that I got some satisfaction out of doing administration because I could solve problems in a different way. So, administration wasn’t so daunting for me by that time. The other thing to say is that department heads at MIT are very powerful. My colleague Boris Altshuler said when he came to MIT that the difference between the department head at MIT and the department chair at Harvard is that the department heads at MIT determine the faculty salaries, and the department chairs at Harvard don’t even know the faculty salaries.
[laugh] And that really characterizes the difference. So, I think if I had been at an institution where the chair was just a rotating thing where you don’t have much authority, it wouldn’t have been as attractive to me. But I saw that as department head, I really would be able to have a big impact.
In what ways was your experience as a director of MRSEC useful administratively as you assumed this position at MIT?
I had a wonderful administrative officer at the Center for Material Science and Engineering, and she just basically taught me how to manage things [laugh]. She taught me about budgets. She taught about staffing, and how to manage staff, especially staff who are not functioning. It was really very educational, and I used all of that later on. And she actually came with me to be administrative officer of the physics department.
I wonder what cues you got from Bob Birgeneau. You learned so much from him scientifically, but what might you have learned from him in terms of not allowing administrative responsibilities to subsume your science?
I definitely learned that. You know, of course, Bob is totally unique in having maintained his research [laugh] through two university presidencies, right. It’s pretty amazing. But, yeah, he was a great example, you know. I think he became department head in ’88, if I’m not mistaken, and then dean in ’91. . But even as dean he would show up in my office in the middle of the day, and have some idea about physics, and we would talk about it for a while. And I realized that it was crucial to maintain sanity when you have these administrative jobs. Also, it was obvious that you have much more credibility in talking to other department heads, talking to the deans, if you’re doing research. You have much more credibility than if you’d given up research because you understand what the faculty are dealing with.
On that note, what were some of the challenges that you had to deal with as department head that you might not have even been aware of before this point?
So, there were budget problems. When I first came [laugh] to MIT, the department had over 100 faculty members. And all of the faculty members had to pay some of their salaries out of research grants. This was left over from an earlier era when the government funding was so flush that the department could rely on it. And this was changing. The department budget had to shrink and I felt that the budget cuts made the quality of life not what it should be for the faculty. I didn’t appreciate, as much as when I became department head, that, the department wasn’t doing enough for the faculty. As I mentioned already the place where I think that was most important was in graduate student support. So, I worked really hard to get lots of private donations to the department. When I started, there was really only a trickle of alumni contributions to the department, and I realized that there were things we needed money for. At that point the whole school of science with about 280 faculty members had just one development person. Bob had an assistant dean for development, and she was going to raise money for all of the departments, including biology and neuroscience, and I knew physics would never get her attention when the low hanging fruit was in the life sciences. So, I went to Bob, who was the dean, and said, “We want to have our own development person.” And he went to the provost. The provost first said no. And then I promised that we would never use any MIT money to pay for this person. We’d only use money from the gifts, and that this person would work very cooperatively with the central development people. That was what his biggest concern was—competition with central development. And we quickly started raising money, and we were enormously effective, and that was mostly luck. It was mostly because people who were of the age to give money had come to MIT in the 1960s, and everybody who came to MIT in the 1960s wanted to be a physicist. Most of them then went off to do other things, and some of those made a lot of money, but they never lost their love of physics. And, so, we were able to exploit that to raise a lot of money for the department.
Marc, beyond the budget, what work did you do regarding the structure, the organizational structure of the department, and how that may have impacted, for better or worse, hiring practices?
You know, I didn’t do a lot with the structure. Herman Feshbach had created a structure in the department, which worked extremely well. His idea was that the department was divided up into divisions, and each division had a representative on a council, and that council really governed the department. I brought in some additional members to the council who were the directors of the major laboratories where the physicists do their research. One covers particle and nuclear physics, and one covers astrophysics. I also created an associate department head for education, which is something that I think Herman had but then had disappeared over the years, and I reinstated that. And that was really important because I felt that our undergraduate education wasn’t getting enough attention, and I couldn’t give it enough attention. My colleague Tom Greytak did a fabulous job of that. So, I think the extra emphasis on undergraduate education really helped the department a lot. The junior faculty really came to take it seriously. It was taken very seriously in tenure and other promotion cases. The culture really evolved to the point where everybody, including Nobel Prizewinners like Wolfgang Ketterle, take undergraduate education very seriously.
Now, becoming dean, was this something that was on your radar at all? Were you looking forward to no longer being department head, and just being a civilian again? How did all of this come about for you, and what were your decisions?
Oh, I think I wanted a new challenge, and I had been department head for nine years. It’s a long time, and I had seen how the dean operated within the school, and some of the things I had learned in the department, particularly the fundraising, I wanted to carry that over to the school as a whole.
What did you see as your mandate as dean, and what was relevant from coming in as department head, and what was a totally different ballgame?
Well, I think the most important job of the dean is the decisions about tenure and promotion. That’s where universities have their quality control. The way it works at MIT is the dean has to make a recommendation to the provost on every tenure and promotion case. And the way the school of science worked is that the department heads would all read the cases for tenure and promotion, and vote on them, and then the dean would decide what to do. That is always the most important decision. Of course, the department heads make that decision at the department level. And if a department head brings weak cases to the dean, that department gets a bad reputation, and then every case is scrutinized more carefully. And if the dean brings a weak case to the provost, the same thing happens. So, that’s the most important thing. The way I say it is at both MIT and Harvard, there’s a quadratic dependence of power on rank. At Harvard, the maximum is at the dean. At MIT, the minimum is at the dean. [laugh] So, at MIT, the provost and the department heads are actually much more powerful than the dean in most cases. So, what the dean does is coach the department heads, and keeps them from making mistakes that will hurt their departments. I felt I played an important role in what is called academic council, where the deans meet with the provost and the president. Especially at MIT, engineering is so dominant that it’s really important for someone to speak up for basic research and science.
Did you harbor any hope that you could maintain your research like Bob had when he was dean?
It was declining already, and I was having trouble. One of the great things I did as department head is I hired some condensed matter physicists who were just wonderful. And the best students wanted to work with the new, young faculty, as they should. So, I wasn’t going to get the same caliber of students that I was getting when I was younger. And, as I told you, I really worked through my students. So, the writing was on the wall. We did one really beautiful experiment in my early years as dean. But after that, it declined a bit.
Tell me what it felt like when you won the Buckley Prize in 2000.
I was so excited. You know, it’s such an honor, right, and you look at the previous winners, and it’s just a wonderful group. And I was particularly excited to win it with Gerry Dolan and Ted Fulton from Bell Labs who made the first single-electron transistor.
When did you start thinking about going emeritus? Was that a process for you, or that was sort of a quick decision because you wanted to ease up on the administrative responsibilities?
When I became department head, 25% of the faculty were over 65. And I knew when I came in that I had to do something about this because elderly physics departments are not great physics departments.
I probably did it too fast and too aggressively, but I convinced a whole bunch of my colleagues to retire. And that allowed me to make a lot of appointments. I think we hired 30 people during that time I was department head. By the way, only one of those was at the senior level, and that was Frank Wilczek. MIT likes to hire people at the junior level. And I didn’t want to get close to 70, and have somebody tell me it was time to retire. So, I had been thinking about this. Probably by age 65, I was starting to think about how to make a graceful exit somehow.
You want to lead by example as well?
Exactly, well, exactly. But, also, I didn’t want to be one of these people who retires and hangs around because I wanted to be relevant wherever I was. I had actually been approached by Steve Chu in 2009 when he first was appointed as energy secretary about being the director of the Office of Science. And at that point I wasn’t ready to do it. I was thinking about other things, other opportunities, and I decided to remain as dean. But then four years later, Steve was outgoing secretary of energy, and we didn’t know that Ernie was going to be appointed. But Steve asked me if I’d be interested now, and I said, “Yes, I might.” And, so, when Ernie came in, he had a list from Steve of people that Ernie should think about. Ernie knew me of course very well. And, so, when Ernie asked me to be considered for director of the Office of Science, I thought this was a nice way to have a capstone of my career. I went through this process of being nominated, and having a Senate hearing, and getting a unanimous vote of the committee. And then I waited for over a year, and never got confirmed. At that point, the Republicans had taken over the Senate, and I would have had to start all over again. It was at that point that Jim Simons called me and asked me if I’d be interested in heading the Science Philanthropy Alliance. So, that was a different way of making an exit.
Oh, I didn’t realize that the Science Philanthropy Alliance was a Simons project.
Well, it was partly a Simons project. It was created by six foundations: the Simons Foundation, Kavli, Sloan, Moore, RCSA (Research Corporation for Science Advancement) and Howard Hughes Medical Institute. It was really a project of the presidents of all of those foundations, or, in Jim’s case, the founder. I think Bob Birgeneau mentioned to Jim that I might be vulnerable because of this experience with the DOE. And, so, Jim called me and asked me if I’d be interested, and it happened then very quickly.
Did you know Jim prior to this?
Oh, yes. Any dean of science at MIT is going to know Jim. [laugh]
Actually, he helped us with a fundraising event. When I was department head in physics, and Mike Sipser was department head in math, Jim hosted a fundraiser at his apartment for MIT math and physics alums on Wall Street. And it was the hottest ticket in town [laugh]. Everybody wanted to meet Jim, and go to his apartment. It was really great.
What was exciting to you about this opportunity, broadening out in terms of working to support basic science?
Obviously basic science was my thing, right—
—first as department head, and then as dean. When MIT began its recent campaign, I was the one who convinced the president that we had to have a component that was basic science in the campaign. I’m passionate about it, and I think it’s undervalued. I saw that in the government agencies that when funding got tight, even the NSF would argue for applications rather than arguing for the importance of basic science. And from my experience at MIT of working with philanthropists, I saw first of all that they could have a big impact and, secondly, that they enjoyed it. You know, we never got money from anybody who didn’t want to do it, and we made sure they got satisfaction out of it by helping them see how their money nurtured young people, young faculty, young postdocs or graduate students. And, so, I saw that this was a place I could help the community by encouraging philanthropists to give more for basic science. I didn’t have a clue how to do it at the beginning, I have to say [laugh]. I just thought the mission was great, and somehow we’d figure something out.
Marc, in what way have you remained connected with your field, with the science, with the particulars that you’ve been working with? What has been interesting you, even if you’re following on the sidelines, so to speak?
Well, I’m very lucky I’m no longer on the sidelines. When I moved out to California to start the alliance, I connected with my former graduate student David Goldhaber-Gordon, who’s on the faculty at Stanford, and he very kindly arranged a visiting faculty position for me. In my last year at the alliance, I reduced my time to three days a week, and started spending more time with his group. And then I was able with connections at the DOE to help increase David’s budget, and I’ve now become a member of his group, and I’m collaborating with his students, helping them to write papers. And it’s just really exciting to be back in the game.
And, so, what’s going on? What are some of the—what’s the cutting edge in the field right now?
You know, by far the most exciting thing to me is this twistronics, the twisted graphene layers that Pablo Jarillo-Herrero, one of my last hires in the MIT physics department, created. This is a unique phenomenon in condensed matter physics where you mechanically control two atomic layers, and you can change the properties of the material. The usual way of controlling properties of materials in solid-state physics is by growing the material with molecular beam epitaxy—that’s the way Horst Störmer’s Nobel Prize came about—or growing crystals to do neutron scattering, like Bob Birgeneau does, or doping them to make high-temperature superconductors. It’s really all the chemistry and material science of the materials. This is totally new where you take this atomic layer, and you bring it down, and if you change the angle, you get totally new properties. It’s just an amazing thing to me.
Marc, now that we’ve reached right up to the present in the narrative, for the last part of our talk, I’d like to ask a few broadly retrospective questions about your work. Then we’ll then look into the future. So, in any successful scientific career, there’s always an element of luck, right, being in the right place at the right time, reading the right paper just when you need to read it, meeting the right person. What are some of the areas in your memory that really stick out where not only were you rewarded with hard work or innate ability, but just being lucky?
Everything I did is luck [laugh]. It’s all luck [laugh]. It’s all luck. We’ve discussed some of them. So, early on, meeting Stan Ovshinsky, and having him appreciate my work was a crucial piece of luck. Having him invite me to the workshop with Nevill Mott was an amazing piece of luck. And, you know, I don’t think I would have got tenure at MIT if I hadn’t had the idea about valence alternation at that meeting. Being at lunch with Peter Wolff when he said, “Hank Smith’s here. He can do things. You should think about doing something with nanostructures.” And then, of course, having a postdoc and student who were good enough to see that this oscillation of the conductance of the narrow transistor was not a fluke, it was something real. I’ve been so lucky with my students, with Udi Meirav connecting with Moty Heiblum, I mean, at every stage. Bob Birgeneau growing the crystal, and then his student finding that the crystal was insulating. But suppose the student had found the crystal was superconducting, I never would’ve been involved with Bob at all. The only time we clearly set out to find something, and we found it, was the Kondo effect [laugh]. And there the luck was having the theory community at MIT that was talking about this, Yigal and Ned, really working out elaborately what we should find, so we would know what to look for. But, in most cases, it was just stumbling on things.
Marc, where have technological advances, either from computational power, from instrumentation, when have you appreciated and taken advantage of technological advance over the course of your career to advance the research, to advance the science?
So, at every stage, technological advancement was critical. In the early days, we were doing pulsed optical experiments. These pulse lasers had just become available. We wouldn’t have been able to do it without them. Cheap lab computers became available [laugh], so we could take data, you know. We had these little things with floppy disks. For high Tc superconductivity, the advancement relied in an essential way on the improvement in crystal growth. And there were new furnaces that the Japanese first used in which the material was heated with lamps rather than from a melt because you couldn’t make pure enough material if you pulled the crystals from the melt. All of the work we’ve done on nanostructures came about because of the advancements in the electronic industry. The whole nanoscience field really came about—or at least the fabrication part—from inventions for the semiconductor industry. You think about all the things that are being done, you just couldn’t do them without electron beam lithography, without optical—high-quality optical lithography and all these things. You know, I really believe that technology drives everything, and then, you know, it opens new windows.
I wonder if you can convey, obviously to our broad audience that’s not going to have your breadth in your area of expertise, what scientifically, intellectually, is such cause for wonderment with the development of the single-electron transistor?
Transistors have been around since 1948, and they get smaller and smaller and smaller. Our computers work based on having transistors which are switches that are either on or off. The ones and zeroes that we use for computation are really transistors being on or off. In a typical computer, it still takes a couple hundred electrons to turn a transistor on, and you have to add those electrons to turn it on, and take them off to turn it off. And it was really a phenomenal surprise when we found that if we made a transistor in a new way, that it would turn on and off every time we added one electron. So, we add the first electron and it turns on, and then we add a second electron, and in between the first and second, it turns off. And, so, it turns on and off again every time you add an electron. It behaves entirely differently from the transistors in our computers.
What does that tell us more broadly about physics?
The fundamental reason for this behavior is that the electrons of course repel each other because they have the same negative charge. What is causing the behavior is the quantization of charge. It’s the fact that the electrons actually are individual particles with charge. You can think about a conventional transistor as the electrons just being a fluid, and they flow on and off. But when you make the transistor small enough, and force the electrons to tunnel on and off, then the charge quantization becomes important. So, it really is an essential aspect of quantum mechanics that makes this happen.
Have there been any problems in the field that you just keep running up against a wall, things that seem solvable, and are just always out of grasp? What’s gnawed at you after all these decades in the field?
Are you talking about the single-electron phenomena or—?
All of the research that you’ve been involved in generally.
So, high-temperature superconductivity is the problem, I mean.
Yeah, what’s the holy grail? What’s it going to take?
Let me put it this way. In solid-state physics, condensed matter physics more generally, there’s a new discovery every decade. Now, the quantum Hall effect was one, and then the fractional quantum Hall effect. And the history has been that after three or four years, there’s a theory that explains it. Conventional superconductivity took a long time, 40 or 50 years, because quantum mechanics had to be invented first. We didn’t have quantum mechanics. Once we had quantum mechanics, it didn’t take that long that—to develop the theory of conventional superconductivity. Here we have a new phenomenon, which was discovered in 1986. Here we are, what, 30 years later, and we still don’t have a theory that quantitatively describes what’s going on. Every theorist who works on the problem thinks they’ve solved it. But nobody agrees that the other theorist has solved it [laugh].
And it’s not clear to me what it will take. I think it may be just a problem where we need new kinds of theory, and maybe that there’s some experiment that will be done that will break it open. But after this long a time, I’m not so optimistic about that.
Marc, last two questions, looking to the future, one on the administrative side, and one on the science side. Given how well positioned you are, I wonder if you can comment on just generally prospects for support of basic science in the United States, perhaps specifically looking at the transition from the Trump to the Biden administrations.
So, historically, Republicans have been much better for basic science than Democrats. And the reason is that Democrats want things to change quickly, so they want more money for more applied things. And I think there’s some concern that that will happen now with the Biden administration as well. But they’re proposing such large increases in budget that probably basic research will be OK. But, you know, I think that’s always a danger. I think the fact that there is a group of philanthropists who want to support basic research is wonderful and will continue to be very, very important. Did I answer that?
Yeah. So, you’re optimistic generally, you’re saying?
Generally, I’m optimistic right now, yeah. I think there’s some reasons to be concerned. I mean, you see this huge increase proposed for the NSF budget for applied research for, you know, technology. And they say that they’re going to keep 15% of it for basic research. But, you know, Vannevar Bush was right. Applied research always pushes out basic research.
[laugh] Marc, last question. It’s so wonderful to hear that you’re a postdoc all over again, that you get to see what’s happening at the cutting edge of the field. So, on that basis, what are you most optimistic about? Where is the field headed, and where do you see all of your contributions over the course of your career contributing to future fundamental discovery?
I think condensed matter physics is spectacular because we do see these new things happening every decade that causes just a new resurgence. And that really makes me optimistic that we’ll keep having new discoveries. There are other fields where surprises haven’t come, right? Nuclear physics, particle physics, we’ve invested huge amounts of money, and we’ve learned things, but not real surprises. In condensed matter physics, there’s just a surprise every five or ten years.
And I see that the students are doing amazing things with topological insulators, and topology becoming important is a big idea theoretically. It’s really fun.
Marc, thank you so much for spending this time with me. It’s been a great pleasure hearing your perspective on all things. And, once again, Bob Birgeneau’s a gift that keeps on giving, so I’m so happy he connected us.
Thank you. I enjoyed this very much.