Oral History Transcript — Dr. Stuart Wolf
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.
See the catalog record for this interview and search for other interviews in our collection
Interview with Dr. Stuart Wolf
Stuart Wolf; March 23, 2006
ABSTRACT: Interview done with solid-state physicist and former DARPA manager Stuart A. Wolf. Interview done as joint effort between AIP and the Center for Nanotechnology in Society at the University of California, Santa Barbara. Interview covers Wolf's early career including his research on superconductivity. It then addresses Wolf's role as a science manager at DARPA. The final and largest part of the interview concerns Wolf interest in and support of research in spintronics, a particular area of nanoelectronics.
McCray:Not to sound like a therapist, but let’s start with the obvious. Why don't we talk about where you grew up and what your childhood was like?
Wolf:Well, I grew up in Brooklyn, New York. I was the son of American parents. They were first-generation Americans. My grandparents were all from Eastern Europe, born in Russia, or born in Russia and Poland, so they came from that part of Europe.
McCray:What year were you born?
Wolf:I was born in 1943.
Wolf:And I always had an interest in science from the earliest days. When I would read something that I wanted to read rather than something that I had to read, it was science related. Even in the comic books.
McCray:Did you want to have a career in science?
Wolf:Well I wasn't really sure but that was the way I was gravitating. I was gravitating toward science even then, and I always participated in science fair projects starting in elementary school.
McCray:What kind of science fair projects did you do?
Wolf:It's hard to remember, but a few of them I remember — I built a motor very early on. Very early. I used to love to take things apart. I didn't always get to put them back together. I used to crave — almost anything that was mechanical. I would disassemble them and figure out how to put it back (together).
McCray:What did your parents think of this? Did they encourage discovery?
Wolf:Well, no, they wanted me to become a doctor or a dentist. Right. That was their dream. But I never was much interested in going in that direction. I wasn't sure I wanted to be a physicist, but chemistry, I had a chemistry set from early on, and I tried to develop my own things. Got me into some trouble with it.
McCray:What kind of trouble?
Wolf:Well, I fumigated a neighbor's house, [Laugh] and they had to move in with us for a night. [Laugh] I think we killed every bug in the house, however. [Laugh]
McCray:Okay. You went to Columbia for school, right?
McCray:Why did you choose Columbia?
Wolf:Well, at the time I really didn't want to move out of the house. So, Columbia was the best school in New York City. And I also went to a special high school, Stuyvesant High School, which is also in Manhattan.
McCray:Yes. I think you're the second or third person I've talked to this week, actually, that went to Stuyvesant.
Wolf:Went to Stuyvesant?
McCray:Stephan Von Molnar went to Stuyvesant.
Wolf:Oh, he did? I didn't know. Yes. Okay.
McCray:I'm pretty sure he did. Why did you decide to major in physics for undergraduate work?
Wolf:Well, it was really because of Stuyvesant. I had a rapport with the physics teachers there. So, it was in high school that I was really steered to physics rather than the other sciences. I was never that much interested in biology, so it would have probably been between physics and chemistry. And so, the fact that I had a very good physics teacher at Stuyvesant, and I also did very well in physics. I won the Physics Medal at Stuyvesant when I graduated.
McCray:Okay. When you were say twenty years old, did you have a particular vision of what a career in physics would look like?
Wolf:No. I really didn't know what it was like. And I really liked school. So, actually when I had to leave school that was a traumatic experience. [Laugh] But fortunately, I stayed in careers that were close to academic, even though I worked for a government lab for thirty years.
McCray:After your bachelors work then you went on to Rutgers for both masters and PhD?
McCray:I wanted to ask you about your dissertation topic. If I have it right, it was on superconductivity in silver.
Wolf:Silver. Yes, that’s correct.
McCray:Yes. Why that particular topic? What directed you toward that?
Wolf:Well, it was really motivated by my interaction with one of the postdocs who was at Rutgers at the time. I was interested in two areas, actually., I was asked by two professors to work for them. I was interested in solid state. So, one of the professors was the one I ultimately went to work for. His name is Peter Lindenfeld. The other was Elihu Abrahams. Elihu had taught a course in quantum mechanics and I had done really well in his course, so he had approached me. But he's a theorist. Actually, he's a very well known theorist, probably the most well known theorist at Rutgers. But at that time I wasn't really thinking of going into a career in theory, so I really wanted to work in the lab. So, I took the other job. Well, for now I was interested in superconductivity. So, Peter was part of the Rutgers Superconductivity Group, which was led by Bernie Serin. In fact, the physics building at Rutgers is now named after him. He was the head of that group. There were several faculty who were part of it. He was he was one of the ones of that group.
McCray:Was this more of an experimental group of researchers?
Wolf:Yes, this was an all-experimental group. Its claim to fame is the discovery of the “isotope effect in superconductivity,” which was a really major breakthrough. It was really one of the factors that led Bardeen, Schrieffer, and Cooper to come to their understanding of superconductivity, basically the BCS Theory was fundamentally connected to the Isotope Effect (the isotope effect is the shift of the superconducting transition temperature with isotopic mass), which meant that there was a clear connection between the lattice vibration frequencies and the superconducting transition temperature.
Wolf:So, my original project, which was — I started off on a slightly different bend. It was actually to measure low temperature transport properties of the Noble metals. And it turned out that very soon after I started the explanation of some of the unusual low temperature behavior was published. The explanation was the Kondo Effect which explained the unusual anomalies at low temperatures. So, I dropped that direction ,and moved on to a study of the proximity effect in silver mainly because at that time I had become friendly with one of the postdocs. His name was Guy [pron. Gee] Deutscher. By the way I've collaborated with him off and on over the years even until just a few years ago. So, he came out of Pierre DeGennes group in Orsay. His thesis also was on the proximity effect and he was sort of a bridge between theory and experiment. He never actually was in the lab with me but we discussed the project everyday. So, we had decided to look for proximity effect-induced superconductivity in the noble metals. And so, that was the direction we went, and we weren't sure exactly what answer we would get. But, we actually found that you could certainly induce superconductivity in silver, and by studying the properties of that induced state you could infer something about intrinsic silver. So the inference was that silver would be a superconductor but at extremely low temperatures.
McCray:So, we're talking below liquid helium?
Wolf:Oh, we're talking in the milliKelvin range, maybe even less than a milliKelvin. So, very, very low temperatures. In other words, it, the electron-phonon interaction responsible for superconductivity was extremely weak but positive. That was the result of my thesis.
McCray:Who was funding this kind of work and what kind of applications were people seeing from it?
Wolf:Well, this was in the days where you didn't really need to have an application identified — it was basically fundamental science. The work was actually funded by the Army. Bernie Serin, who was, as I said, the leader of the group had for many, many years a grant from the Army Research Office. I never saw any of that. I mean the group was well funded. Bernie was sort of the person who got the funds. Graduate students were not concerned with fundraising in those days.
McCray:But, were there any ideas that this would be used for particular applications or this really was just funding the science?
Wolf:This was basic research. We never – well of course I was a student so I didn't know what went on at the higher levels. But as far as I could tell the justification was just, "We want to understand superconductors better."
McCray:Because one of the things I was thinking of looking over your profile was superconductivity in the Naval Research Laboratory, and I was thinking about whether there were applications for submarine detection and that kind of thing?
Wolf:Well, I did get into that. And, once I got into the Navy, then we did have to think about where this would all be used. At NRL it was clear that there was a much more applied bent to what I was doing.
McCray:You graduated in 1969 and then you spent some time at Case Western. I guess, what for a couple of years, and then you started at NRL in '72, right? What led you to NRL and what kind of research were you starting, or joining there?
Wolf:Okay. Well, so what I did at Case, it was a very nice situation. What happened was I was hired by B.S. Chandrasekhar. Not, not the Nobelist — but he's also, I mean he always used to get irritated when he was mistaken for the “other” Chandrasekhar. [Laugh] But he also had a very good reputation in superconductivity. There's a very famous book called Park's Book, which is sort of the bible of superconductivity, ala early '60s, and B.S. Chandrasekhar wrote the first chapter, the introductory chapter. Guy co-authored a chapter with Pierre DeGennes on the proximity effect. And Bernie Serin authored a chapter. So, three of the people I was associated with were all major contributors to that bible of superconductivity. So, when I went to Case to work for Chandrasekhar, it turned out he was promoted to the dean of science. So, he was pretty much out of the lab and so I basically ran his laboratory and I managed his students.
McCray:Okay. And this was again continuing your research superconductivity?
Wolf:It was continuing superconductivity. In fact, I followed up on a few projects that I was starting at Rutgers but never really finished. I actually worked on several different projects in fact, and I also built a dilution refrigerator.
Wolf:A dilution refrigerator. It's a refrigerator — in those days you couldn't really go out and buy them. Now you just buy it. But in those days you had to build it yourself.
McCray:What is it?
Wolf:It's a way of getting down to the milliKelvin temperature range.
Wolf:So, you can use it to get helium to get to four degrees. You can pump on it to get to one degree. You can actually replace the helium-4 with helium-3 and get to three-tenths of a degree, and then you could actually make a mixture of helium-3 in helium-4 and when you make that mixture you dilute, actually the process of diluting the helium-3 and the helium-4, basically it's like a free expansion of gas and that gives you cooling. So, that takes you from the .3 all the way down to a few milliKelvin. So, it's a way of being able to do experiments at very low temperatures. And so, again, this was really just a scientific challenge for me, there were a lot of professors there who were interested in looking at things at very low temperature and I was the facilitator. In fact, a colleague of mine, Arnold Dahm, who did beautiful work on electrons and helium, he was basically the major recipient of that. So I really built it frankly just to build it.
McCray:Okay. Did it take you a long time to build it?
Wolf:It took me almost the full time that I was there. I didn't really get to use it much but they used it. I mean I used to get reports that said, "It's working great. It's working great." It worked great for about ten years. They used it for at least ten years after I left.
McCray:Okay. From Case then to NRL: how did that transition take place?
Wolf:Well, I was looking for jobs. Jobs were pretty scarce in the early '70s.
McCray:Yes, there was a boom of physics degrees produced in that time.
Wolf:The fellow who hired me at NRL knew Chandrasekhar, and so it was based on personal contacts –believe it or not, I actually had another job offer at that time. Actually, I had three job offers two for “real jobs” and one for another postdoc.
McCray:That's pretty good.
Wolf:Which, at that time was very good — so I had the one from NRL, I had one from Brookhaven, and I had one from Livermore.
McCray:Why did you choose NRL then?
Wolf:The Brookhaven job was basically more of a senior postdoc job. It wasn't a permanent position. The job at NRL sort of fit me. I liked the people there. They offered me a nice salary. Livermore was actually a better offer but my wife didn't want to move to California. It was just too far.
McCray:Were you married at this time?
Wolf:Yes, I was married at that time. And so, I took the NRL offer.
McCray:Okay. Well, tell me about your time at NRL. Let me preface that: at some point I want to transition from your research to becoming more of a research manager into the rest of your programs. I'd like to know how a person moves from one area to the other.
Wolf:Well, at NRL, I was a researcher. Again, I was hired in Applied Superconductivity to be part of our Applied Superconductivity Group. So, the idea was here – now at NRL we were not just looking at basic sciences, we were trying to figure out how to apply superconductivity. And the area that I started to work in was superconducting films. And, in particular to make something called a SQUID which stands for Superconducting Quantum Interference Device.
McCray:Okay. Yes, this is a device I've heard of.
Wolf:Yes. It was before you could actually buy them. So, I was actually working on making the SQUIDs. I came up with novel ways of making junctions without actually having to make the physical junction. I perfected ways of using granular superconductors to actually make an effective junction.
McCray:What were the materials you were using?
Wolf:I was using initially niobium, and then niobium nitride, and then niobium carbon nitride. That was the main direction, but of course, I was always involved in a few other things. I made a number of friends and one of my colleagues whose name was Vince Folen. He got me interested in working on a hydrogen maser. It was interesting because I would both be presenting work on superconductivity at one conference, then I'd be presenting work on magnetic shields. We had a project where we actually were funded to build a compact hydrogen maser for satellite applications.
McCray:What would you use it for in a satellite?
Wolf:For a clock. It's a very stable clock. The hydrogen masers are used by the National Bureau of Standards as a time reference. And so, the idea was — but they were big kluge things. So, our project was to build one that basically would fit on a table top eight inches in diameter and a foot high. So, that was what we did. We succeeded. I was responsible for the shielding, the magnetic shielding. I was responsible for all of the vacuum system. It had to be a self-contained vacuum system because the object ultimately would be on a satellite. We were working with the Smithsonian Astrophysical Observatory, so it was actually a jointly funded project.
McCray:Who did you work with at the Smithsonian?
Wolf:Roger Vessot. I don't know if you know his name. He was a partner in that. That was a very interesting venture. So I became associated with the satellite people at NRL at that time.
McCray:Oh, okay. They were doing a lot of work in x-ray astronomy if I remember?
Wolf:They certainly did at NRL, but the folks I was associated with they were part of the Navy Satellite Center. So I became associated with those guys and I have to say that was one of the highlights of my career. Unfortunately most of the work was highly classified at that time. However Vince Folen and I solved a major problem for GPS allowing the satellites to be put into orbit. mo .
McCray:Was Irwin Shapiro involved? He was a big name in some of the early GPS satellite work?
Wolf:Maurice Shapiro was a major name at NRL in the space arena but he was not part of our group he was in a separate Space Science Division not the Navy Satellite Center. The idea for GPS actually originated at NRL. The inventor’s name was Roger Easton. He also designed the space surveillance system, which actually is the system that keeps track of all the satellites. That was his baby, too. He was in on that. I don't know if he's still alive. He retired a long time ago.
McCray:Since you mentioned GPS, I want to ask a question.
McCray:The research that you were doing at NRL: was most of it classified at this point?
Wolf:No. The only classified piece was a part of the work that I did on the — so it was a very small fraction.
McCray:So you could present your research at American Physical Society meetings?
Wolf:Everything else could be presented at meetings. So, all the work I did on this hydrogen maser was all open. I presented it at conferences. And, anyway it's sort of funny. One of the fellows who was a graduate student with me at Rutgers also worked in timing, in other words in clocks, at Fort Monmouth, and we would see each other at all of the timing meetings, and he thought that this was my major research. [Laugh] It turned out it was only sort of a small fraction. So, my main research was really superconductivity and building SQUIDS and testing SQUIDS. I actually worked on a project to build an antenna for very low frequency communication, submarine communications. And that also wasn't a classified project, work on this superconducting antenna. I don't know if you have my whole resume but I have a whole bunch of articles on that.
McCray:Yes. At some point, according to this, you became a supervisory research physicist? I'm curious about that.
Wolf:Right, I worked very closely with a colleague by the name of Don Gubser who became my supervisor probably a few years after I started working there. I also started working on the SQUIDs. I got very interested in superconducting materials in general, and so I moved out of this applied section into a more basic section where Don was the boss. I probably did that about 1977 or so, maybe five years after I started working there. I moved, but still within superconductivity, still within the same — we call it a branch. So, it was basically in the same, more or less the same organization, and so that's when I also started developing new superconducting materials. So, not only was I working on this SQUID but I started working on superconducting materials.
McCray:What were some of the new materials?
Wolf:Well, we did a lot of work on niobium carbonitride. That's when I moved into the carbonitrides. There was a theoretical prediction that if you could make molybdenum nitride in a particular crystal structure, which was the B-1 (sodium chloride) crystal structure the theorists predicted you would have a record-high TC (critical temperature).
McCray:When you're talking high TCs . . .
Wolf:At that time I was talking about thirty Kelvin. Okay?
Wolf:The record at that time was twenty-three Kelvin. And so I spent a couple of years trying to make moly-nitride in that structure. I succeeded. It took about two years. The good news was, it was a superconductor, and it was a reasonably high TC (13K) superconductor but the bad news was it wasn't a record.
Wolf:It was thirteen Kelvin. Of course the theorists were really happy because they actually predicted that this material would be superconducting, but I was not so happy because it wasn't a record. [Laugh] So, it was just a nice publication.
McCray:What was difficult about making these materials?
Wolf:Well, I was making it using thin-film techniques, basically sputtering, and the trick was to find exactly the right conditions. It wants to form in another crystal structure. And so, the way I approached it was to do basically combinatorial chemistry. It might have been one of the earliest combinatorial approaches to materials science. So, I had a split target, of molybdenum on one side and niobium on the other, and I reactively sputtered it in nitrogen. And what I would do — so if you look at all the substrates as you go across, you know, here's the boundary, here on one side you've have very niobium-rich material. Here on the other side you'd have very moly-rich material. And, what would happen, if you studied the structure, you would actually see where the boundary went from the B1 crystal structure to the MO2N structure. So, you could actually track the boundary, and then by changing the deposition parameters I could find conditions where this boundary moved. Temperature, pressure, various things that you can change for example you can vary the sputtering rate. I had a number of knobs. And so what I would do is try to move this phase boundary closer and closer to the pure moly side and eventually we were able to find conditions where we could get it to go all the way for us.
McCray:Okay. So, you're at NRL at this time. What was your reaction to these new high TC superconductivity materials that were discovered in '86-'87?
Wolf:Oh, we jumped in with three feet. [Laugh] We, in fact for a while we were in contention for the patent for YBCO (Yttrium-Barium-Copper oxide). So, as soon as the discovery from Bednorz and Mueller about the forty Kelvin superconductor was circulated, we became very heavily involved in research in this area. When we knew and of course we had a very good grapevine, and there was a material around ninety Kelvin, we began working in that area. And our main contribution was to figure out exactly what the structure was. So, I wouldn't say we initiated it. We didn't discover the fact that there was superconductivity but we determined exactly what was responsible. In fact, Paul Chu’s patent application was very flawed, even though he clearly was the one who found the high TC.
Wolf:Because he didn't know his sample. His sample was mixed phase, it had the high TC phase in it but it wasn't a pure phase. And they didn't actually correctly identify the structure. Even though they had made the discovery that "Yes, there was superconductivity in this system," they hadn't clearly said what the material was. And the Patent Office decided what the patent would be for and it was not only a method for producing material that was some percentage, and I think it was ninety, ninety or eighty percent of the proper phase, but making it and knowing what it was. So, we had determined exactly what the material was and the fact that you had an oxygen deficiency was a major contribution of ours. So the Houston (Paul Chu’s) patent application was thrown out. There were three groups in contention. It was us at NRL. It was IBM, and it was AT&T Bell Labs. And ultimately we all were pretty close. AT&T had, came very close to understanding what it was but they had missed a key feature. They had missed the fact that there were these separate planes and chains. They knew, they pretty much had the structure but they didn't realize that it was an ordered defect structure, in a sense, because you had these planes and the chains and the chains actually — they thought that the chains were just an oxygen deficiency but not an ordered structure. So, we actually came up with exactly the right structure.
McCray:Okay. So, you had it on x-ray diffraction?
Wolf:It was x-ray, a lot of x-ray, and of course you had to have good samples. So, we had good sample prep and very good x-rays. So now the question was AT&T also had good samples but they didn't really get the fact that it was an ordered structure. So, it was down to the fact whether, if you did have a disordered structure, would it still be ninety percent, or eighty-five percent the superconducting phase? And, of course, that's a fine line. And I guess ultimately the patent office said that that was good enough. So, they granted AT&T the patent. But, this took like five or six years of depositions and testimonies and things.
McCray:Okay. At this time in the late 1980s, there was all this interest on the part of the federal government, but also the state in sponsoring high TC superconductivity. These are just in transcripts of hearings from '88, '89, about this, and seeing there was all this interest in U.S. competitiveness with Japan, and this idea that the Japanese were going to take over this area. What was your reaction to all this interest and excitement that was…
Wolf:Well, I was part of it. We were very much involved with trying to make the materials. In fact, I was part of something called HTSSE [pron. hit-see].
McCray:HTSSE? How do you spell it?
Wolf:High Temperature Superconducting Space Experiment. So we actually, again, because of the connection I had with the NRL Space Center we got this project funded. — it was HTSSE, that was the acronym. We called it HTTSE.
McCray:What was that going to do?
Wolf:Oh, we were going to devote a major part of an experimental satellite to superconductivity put it in orbit and measure and evaluate the performance of a variety of superconducting devices. See, if you could use this high TC in space, how it would stand up. So, I worked with the satellite people. We got funding from one of the agencies that supports satellite research and we started this project. I was one of the four program managers.
McCray:Okay. One of the agencies? Which agency?
Wolf:Well, I can't really say.
McCray:Okay. I assume something like the NSA, NRO, or CIA.
Wolf:So, the funding actually came from one of the three-letter agencies, but the program itself was unclassified, just the source of the funding.
McCray:Was the experiment done?
Wolf:Yes. Oh yes. It was done. It was successful.
McCray:When was it launched?
Wolf:It was launched in the early nineties. In '93, something like that. And, I mean the data was coming out for quite a while. Actually it was HTSSE II that got — we actually had a HTSSE I but that satellite blew up, so it was one that didn't make it into orbit. So, we actually improved that one.
McCray:Were these launched from Vandenberg?
Wolf:These were launched from Vandenberg, right.
McCray:Yes. At this time, around 1987, there was all this media attention given to superconductivity and of course a lot of academic professors, directed their research to it. It seems to me that it sort of went through this boom and bust cycle, but maybe that's not accurate. Maybe that was just sort of the presumption of people looking at it from the outside in respect to now.
Wolf:Well, the problem is — I wouldn't say it went to bust. I mean, there are applications of superconductivity. But, there are some really good ones, but the problem is the expectations were quite high. At DARPA I was involved with a program called "Super HYPE".
Wolf:Which, it was a superconductivity program. [Laugh] It was, it’s called superconducting hybrid power electronics and it was for moving superconductivity into the power industry. And of course, the first aspect would be actually to use it on ships for power distribution. You know, actually it can save size and weight, considerable size, weight, and energy if you move to superconducting components. The Navy, the reason that this was really timely was that the Navy is investing heavily in a superconducting motor for use on future ships. And again, the reason is that the motor could be considerably smaller and lower weight than a conventional motor. As they move to electric motors on ships, instead of turbines, first they'll probably go to conventional motors and maybe transition to superconducting motors. So, my vision was, "Well, if you're already going to be committed to superconductivity, you're already going to have a cooler — you're already going to be committed to the low temperatures then why not look at where else you can put superconductivity and would there be a benefit?" It turned out we funded some studies that showed that there is a significant benefit for moving the superconductivity into the power electronics.
McCray:What effect did these high expectations have, as you saw it, on the materials community?
Wolf:Well, of course, huge numbers of people moved in on superconductivity. So, for several years it really dominated the condensed matter community. I mean, if you follow the trend, I mean one good benchmark is looking at the number of talks at the March APS meeting. And so, I mean, you could follow that trend. It really went up, I mean, for a while there were eight parallel sessions on superconductivity. From maybe where in the past there was one or two. So, it really peaked and now it's down, but it hasn't gone away. It's now down to more than it was before high TC. So, the interest is still larger than it was pre-high TC but of course it's nowhere near the peak.
McCray:Yeah. This is sort of a side question - I want to come back and talk about spintronics. But, one of the things I've noticed in looking over the background is some of the people who are here for this meeting did research in superconductivity and they're now doing work in spintronics. Is that a coincidence or is there a connection?
Wolf:No, but there's, I mean you might think of superconductivity as being very closely related to magnetism. In principle it's a type of a magnetic state. Because the Cooper pair is really a special magnetic entity where the spin, you have a pair, the spin-up and a spin-down electron. And, you know people who studied – I mean that's, superconductivity and magnetism are always two things that are usually taught together. And so, people who are well versed in one are typically well versed in the other. So that transition is not too hard.
McCray:Okay. That's actually very helpful because this is something that I've noticed in looking over peoples' backgrounds. And, I wasn't sure if that was just coincidental.
Wolf:Yes. So, that boundary is diffused enough so that people can move readily. And, of course, one of the transitions, there were these colossal magneto-resistant materials that were discovered, the manganites. They're very similar. In fact, the way you make them is very similar to the way you make high TC. So, people could very easily move from the high TC materials into the manganites. They're both oxides. And, in fact, one of my main interests now, even in magnetism is oxides. Because I think they represent a really interesting class of materials that can be applied to some of these magnetic devices. So, I'm really, I'm high on oxides.
McCray:High on oxides? [Laugh]
Wolf:That's right. [Laugh] Right.
McCray:In '93 you started to work for DARPA on just, I wasn't clear on this. Were you still at NRL at the time?
Wolf:Yes. Okay. Here's what happened. So, I had moved up the ladder at NRL. So, I was working for Don Gubser, and he was what's called a Section Head, a first-line manager. And then he was promoted to branch head, and then he promoted me, so I took over his old slot. So, I became Section Head, and I was in charge of the superconductivity section, and then in 1986 just before high TC broke, he was promoted to Division Head and so I followed, I basically just followed. He promoted me to be what his old job was, the Branch Head.
McCray:So, you have sections, branches, and divisions?
Wolf:And so as he moved up I followed. We had worked very closely together and so then I became a branch head. And the branch head is the first position where you have to spend a considerable amount of time doing, I'll call it, bureaucracy rather than science. And so even though I kept my fingers in the lab, the amount of time that I had in the lab decreased considerably. Not right away because that was high TC and I was in the laboratory. In fact, Don was even in the laboratory even though he was the division head, making samples and making measurements.
McCray:How many people say in '86, '87, and '88 are working in superconductivity at NRL?
Wolf:I would say fifty people.
McCray:That's a lot.
Wolf:Not only in my branch but other people, other groups, and other — wherever superconductivity was being worked on. And if you count the people in the satellite group then it's even more. They weren't working on superconductivity per se but they were working on the infrastructure that would make it possible to actually put this experiment in space. I mean, there are a lot of other aspects to that and so quite a few people on this.
McCray:Did you like the bureaucracy part of it?
Wolf:Yes. Part of that I actually started a relationship with DARPA. One of the program managers at DARPA had asked m to help him manage the superconductivity program so I became a consultant to DARPA. I was helping them.
McCray:Was DARPA heavily funding high TC?
Wolf:Yes. They had a very big program. I think it was maybe $30 million a year, maybe slightly more.
McCray:And this is funding that they would give to universities?
Wolf:To universities and companies.
Wolf:Right. It was a rather big program. So, in the late '80s that was quite a bit. That was quite a bit of money in that.
McCray:Okay. And, so how did you find working with DARPA?
Wolf:Oh, it was great. So, I was looking for a change — NRL has a very nice sabbatical program, and Don Gubser, who was my boss, had taken a sabbatical. He went to NSF for a year. And so, I was thinking that, "Well, I might again follow in his footsteps and take a year." I had taken a year at UCLA. I spent one year between '81 and '82 working at UCLA. I was invited there by someone whose name you'll probably recognize, Ray Orbach.
Wolf:So, I knew him. Now he's the Deputy Secretary of Energy. But he was organizing a big meeting, the LT conference, and so he had contacted me and asked me if I wouldn't come and spend a year at UCLA and actually help him with that conference.
Wolf:It was LT, I think the LT 11. Low Temperature Physics Conference. It's a major conference, which was held in Los Angeles in 1981 and Ray was the chairman. So actually I spent a great year. That was also a great year. I sort of was like the Pied Piper, I spent most, of course, almost all my time at UCLA but I did have some connections at Stanford. And so, I'll tell you, I'll tell you how good that year was. So, not only did I bring back to NRL several of the students from UCLA to be postdocs for my group, I brought back three students from Stanford. During that year I gave a talk at Stanford. I gave a colloquium at Stanford and in the audience was someone who, and that's the fellow I'm going to visit on Sunday, his name is Vladimir Kresin. He is a really good theoretical physicist trained by Landau. He was one of the Landau Group, and he had he had immigrated to the U.S in 1978. He was one of these, I guess, refuseniks and he was hired by Lawrence Berkeley Laboratory. And, it turns out our theses were parallel. My experiment used the thermal conductivity of lead silver sandwiches to determine the energy gap of the proximity effect induced superconductivity in the silver. The analysis of the data was related to the theoretical paper that he had done on the thermal conductivity of superconductors when he was in the Soviet Union. In particular, what I did was thermal conductivity measurements, and he had done the theory.
McCray:Why do that?
Wolf:Well, we analyzed the data of thermal conductivity to get the temperature dependence of the induced energy gap in silver. I mean, it was a complicated analysis to extract all the information and he had done the theory. And so, we knew each other's names and of course obviously had no clue as to who we were. So, he came to that talk I gave at Stanford and then afterwards he introduced himself and that was the start of a collaboration that's lasted until today. In fact, we just wrote a review article together. We've probably published close to hundred papers together. [Laugh]
McCray:That's a lot.
Wolf:Right. So, that was the start of this great collaboration, my being at UCLA for that year.
McCray:Okay. So that was '81, '82. And then the transition to DARPA was about ten years later?
Wolf:The transition to DARPA came after high TC. So, after high TC I began – okay, so I started to tell you what happened. So, I had taken that sabbatical. At NRL you're allowed to have one sabbatical every ten years, and they paid for it. It's a very nice program. So, I was thinking of another one and as I say my boss Don, who I was also friendly with, had taken one at NSF. So, I had gone to NSF for an interview and they actually made me an offer to come to spend a year there, you know, as a rotator. But I was helping DARPA. So, one day I just said, you know, "I'm going to be switching." And they said, "Why?" And so the office director at DARPA said, "Why are you going to NSF?" I said, "Well, you know, I'm looking for change." He says, "So, why don't you come here?" And so they basically said, "Don't go to NSF, we'll take you here." [Laugh] So, that's what happened. So, I left, so I basically took a job being a rotator at DARPA instead of a rotator at NSF. It was a temporary appointment.
McCray:What would your duties be as a rotator?
Wolf:Well, it's the same as anyone else. At DARPA everybody's a rotator. I mean, it's running programs. And so, I was only looking for one year, but DARPA said "No. You have to do two years." And I said, "Okay. I'll do two years if I can be part-time." So, I actually maintained the branch head position at NRL and committed myself to DARPA for two years. In the meantime, interestingly enough, I didn't want to leave NSF holding the bag so I convinced one of the colleagues that I had worked very closely with at NRL and I asked him if he wanted to take that slot. His name was Ulrich Strom. And now he's — actually, so he went to NSF and he became permanent, he's now in charge of the MRSEC (Materials Research Science and Engineering Center) Program. So, that's sort of an interesting chain of circumstances.
McCray:Yes. That worked out pretty well.
Wolf:Yes. So, it worked out well for everybody. And so, what happened was I went to DARPA for a couple of years and after the two years I basically had to renew my commitment, but at that time I was really into DARPA, into my role as a program manager at DARPA.
McCray:What were some of the early programs that you ran at DARPA then?
Wolf:Okay. So, actually it was this – so one of the visions that I had when I finally decided to go there, I went there really to do superconductivity and work with one of the other program managers whose name was Frank Patten. But, I had it in my head that I wanted to start a program on magnetism because the branch that I was managing at that time had more than just superconductivity. Superconductivity was only about a third of what was going on in the branch. And, there was work in basically, we called it magnetoelectronics and the head of that section was Gary Prinz.
McCray:That's a familiar name. Is he at NRL now?
Wolf:No, he retired.
Wolf:So, he was, he was in my branch and had a very good program on what we called magnetoelectronics, which is basically very similar. So, I had it in my head that for my contribution at DARPA, I'd start a program in magnetic materials and devices. And so, I presented, so the way DARPA works is you basically come up with an idea or come up with a plan and present it to the director. The director makes those decisions. And so, I pitched the program, actually a very large one, I pitched a program that was $50 million.
McCray:Okay, and this was for how long and what topic?
Wolf:Over five years. This was for magnetic memory and sensors, based on GMR (Giant Magnetoresistance). And, when I pitched it, the idea was mainly GMR but by the time we started it became GMR and Spin Dependent Tunneling, which was really perfected in around 1995. So, I started pitching this program in 1994.This was when I started actually going ahead with a project on magnetism. So, I had gone there in October of 1993. That's when I started at DARPA.
McCray:Okay, when you're pitching one these programs to the DARPA director, how does it work? I'm imagining you're describing the program, what benefits it's going to produce, do you also have to provide a list of "Well, these are the groups that I want to engage?"
Wolf:Sure. Yes. I came up with this. So, what I had done was I convinced — there was a program at that time called the TRP, Technology Reinvestment Program. This was a big DARPA program. It was called TRP and it was probably $200 million to work on dual-use technologies.
McCray:Yeah, that was a big thing at the time.
Wolf:It turns out that the fellow who was my boss, his name was Buchanan. Lee Buchanan. He also was not only the office director of the Defense Sciences Office where I worked but he was also the head of TRP. So, I twisted his arm into giving me a little bit of money. I mean, of course a little bit of money in DARPA is, for other agencies is a lot of money. So, I twisted his arm to basically give me $5 million to start up a little effort to see where this was going. So, I had some money from TRP and I formed what was called a GMR consortium.
McCray:I want to talk about GMR just for a second. So, you have the discoveries that were made, what, in Paris, and I can't remember where else in like '87, '88?
Wolf:Yes, by Grunberg, Fert, and Baibich.
McCray:Right. And then you have Stuart Parkin's work, from '88, '89. Were these things that were on your radar?
Wolf:Yes, right. These were things that were on my radar screen, and because we were following up with that at NRL. I mean, Gary Prinz was following up on that work. And, in fact, you might say he stimulated some of that work. I mean, he missed making those discoveries although he was headed in that direction. He was really the first to start growing thin films of these magnetic materials on semiconductors. So, in the early '80s, he was one of the pioneers in doing that.
McCray:Okay. So, now you've got this $5 million program at DARPA. Then what?
Wolf:Yes. So, yes. So, of course TRP, the requirement was that you have fifty percent cost match. So, I basically had a project with several companies. Honeywell, NVE. At that time it was called NonVolatile Electronics. A sensor company, Federal Products. And…
McCray:That's actually the name of the company, Federal Products?
McCray:Boy, that's an innocuous name. [Laugh]
Wolf:Yes. They make a lot of position sensors for machine tools. And well it was Hughes Research Lab which I think at that time was part of General Motors.
McCray:Okay. So, these where you got the matching funds then for this?
Wolf:So, I got the matching funds, and the whole intent of that project was to explore where GMR was going to make an impact. And, that was a short-lived project and out of that project we identified memory — we identified a non-volatile radiation-hard memory. And some sensors that were going to be particularly of interest to the DoD.
McCray:So, you're talking about ways then to prevent satellite memory from being damaged by cosmic rays and other radiation.
Wolf:Right. And, in fact, a memory is on the space shuttle. It's a very old technology, but in fact on the Challenger, the memory that was recovered worked perfectly. Because I know, I worked closely with the Honeywell people and they made all of the memory. They made the systems, the navigation systems for just about every satellite.
McCray:Okay. I understand.
Wolf:And this particular old form of memory is a 128kb. It weighed forty pounds and cost a quarter of a million dollars. And, I actually, so I went to my friends in this, in the satellite part of NRL and I asked them, "Do they have one?" And, in fact, they did but it was actually on a satellite that they were putting together. They actually took it off the satellite, put it in one of these boxes and I actually brought it with me to DARPA. So, when I made this presentation I plopped it on the director's desk and I said, "Okay. We're going to take this and I'm going to replace it with a fifty-cent chip." [Laugh]
McCray:So this is something that obviously doesn't weigh forty pounds and is a lot smaller?
Wolf:No, this is just a semiconductor chip.
McCray:So, what was their reaction whenever you came in with this forty pound box?
Wolf:Well, so they gave me $50 million to develop this chip. [Laugh] So, and that's basically the chip then. So, this was the original. And, of course this was, and I also said we'd build sensors for various things. But this was the main — and of course this was science fiction then. I had projected what might be possible. And actually, believe it or not it came pretty close.
McCray:So, what was involved in going from this large piece of equipment to this small chip?
Wolf:Ten years. Ten years and a lot of work and actually hundreds of millions of dollars. Ultimately, investment not only by DARPA but by the companies. So, there are two companies now. Well, if you go to, there are now probably twenty companies that are interested, but Freescale actually has a product and IBM is pretty close to a product now.
McCray:Now, you mentioned here in your presentation that you gave this week about applications for avionics and missiles. Are there other applications?
Wolf:Yeah, there are — I mean ultimately it might appear in every computer.
Wolf:And, it's a matter of cost. Again it's cost volume. If the volume goes up. So, right now it's sort of niche applications because they're still pretty expensive. But, as the volume goes up the cost will come down and the application space will improve.
McCray:So, are these MRAM devices (magnetic random access memory)?
Wolf:This is MRAM. This was what I was proposing. Okay. And so, what happened was, so I got this, I got this money to start a program, and we actually funded quite a bit of — I mean, we funded three companies and a lot of universities.
McCray:What were some of the universities where the research was being done?
Wolf:We had Cornell which was one of the major ones where some really good things happened. We funded – oh my memory. Well, we funded, there was money going to NRL. There was money going to Minnesota.
McCray:And how long were these grants then before then? For a couple of years?
Wolf:Well, this was a five-year program.
McCray:It was five years?
Wolf:It was a five-year program. At that time, DARPA did commit to five years’ worth of funding. So, we had five-year funding. And actually that program went into several phases. So, it started in '96 and it ended in 2002. So, it actually went for a little bit longer than five years. But, in the middle the funding got bumped up. So, in the end DARPA was spending, I would say, closer to a $100 million total.
McCray:Wow. That's a lot of money. Okay. One of the things I'm curious about is how the various program managers at DARPA and other similar agencies interact with each other. I'm curious about the formal and informal networks. I mean, so you're in DARPA. You guys obviously have a lot of connections at NRL, and knowing the little that I know about Washington, I know there are just social networks? How did communications work between yourself and other groups?
McCray:Yes. Can you say more?
Wolf:I worked very well with the program managers at the Navy, at ONR, and at ARO and less so, but also had pretty good relationships with the Air Force. So, we knew what each other was doing. We were invited to each others’ reviews. The Navy also had a very nice program in this area. They were doing less-focused work, and this program was really focused on building this device. And, the university work was really focused on basically supporting the work that was going on in the companies.
McCray:Okay. So, would you all get together for weekly or monthly meetings?
Wolf:We would have one large program review. And so, this is how the name spintronics arose. So, we typically at DARPA, in addition to site visits where I would go and have an informal visit at the various contractors. We had one big meeting a year, a program review where all the folks would come. In this program we had a public presentation where everybody spoke to everybody and then I had private presentations from the three big performers, which were the three companies. They would have a private review.
McCray:And this would involve Honeywell?
Wolf:Honeywell, IBM, and Motorola.
McCray:Okay. Did these companies have to put up matching funds as well?
Wolf:Yes, in this program they did put up matching funds. We had a sliding scale. Okay. The way it worked was that this is an arrangement I said, "Okay. When the program is over after five years I'd like the contributions to be about 50/50. However since you're taking a lot of risk in the beginning, because we don't know whether this will pan out, DARPA will in the early years put up most of the money you'll have to put up very little. And as the program moves and it looks like a success DARPA contributions will shrink, and your contributions will increase." So, at DARPA, so we had these two slopes, right, and that's…
McCray:And this is over a five-year period?
Wolf:This was over a five-year period.
McCray:Is that a normal procedure for DARPA?
Wolf:No. This was, this was my — I mean, a DARPA program and every program manager is his own person. They can basically construct a program the way they want.
Wolf:So I felt that this was a reasonable, sensible way to structure the program.
McCray:Okay. I want to talk to some of the industry people, who at Honeywell, IBM, and Motorola would be good contacts?
Wolf:Honeywell would be Hassan Kaakani. At IBM it's probably Bill Gallagher. And, at Motorola, now it's Freescale. I mean Motorola spun out Freescale. Saied Tehrani.
McCray:Okay. So, DARPA's was putting in roughly a hundred million dollars?
Wolf:Well yeah, initially it was about fifty, and then as it was clear that we needed more money, we boosted that. And probably, I would say, when all was said and done it was probably close to, well, I don't remember the numbers but somewhere more than fifty and maybe less than a hundred but somewhere in the middle.
McCray:And then the companies contribute?
Wolf:And the companies were putting in roughly comparable amounts.
McCray:Okay. So, where does the term spintronics come from then?
Wolf:Okay. So then the term came. So, for the very first meeting, which was set for the end of '96, I was actually discussing the name with the people who organized the meetings. I didn't like calling it the Magnetic Materials and Devices Workshop.
Wolf:Well, I don't know. It just didn't roll off -– this didn't seem — I like acronyms so all my programs have, sort of had good names.
McCray:Yes. You've got some good ones. [Laugh]
Wolf:Right. Okay. So, I was saying, "Well this really is a program on spin transport because of GMRand spin-dependent tunneling. So, I said, "Okay, spin transport electronics." And then all of a sudden — just out of the blue — Spintronics came out. And so, I said, "Okay, we're going to call it the spintronics review." And so that's how the name came about — there were two people in the room, Daryl Treger and Alice Burgess. They liked the name so much they wanted me to copyright it. [Laugh] I said, "Nah." So that's where the term first arose. Then it appeared on the DARPA website and I actually was called by a writer for one of the British popular journals on physics.
McCray:Physics World? Or was it one of the others?
Wolf:Yes. And he liked that term. So he interviewed me. And then they put it on the cover, Spintronics. So, that's where it got popularized. Probably more so from the cover of this Physics World.
McCray:Do you wish you had patented it?
Wolf:No. Well no. I'll tell you. You know why? Because if I had copyrighted it it wouldn't have become so popular. I don't think so. Well, maybe. Maybe.
McCray:So, I'm curious about this program to develop dual-use technologies. Were there aspects of the spintronics program that were classified?
Wolf:Well, the spintronics program was never classified.
Wolf:So, it was always an open program. There was interest in the program from the agencies, but the program itself was never classified.
McCray:Okay. Are there particular difficulties or challenges in order to develop this dual-use technology?
Wolf:Well, I mean what ultimately happened is TRP only lasted for a few years and it was for issues like those, I guess Congress felt there probably shouldn't be a dual-use agency.
Wolf:They felt that DARPA's mission was for the war fighter and that the commercial industry can fend for themselves. So the way programs were structured was changed and they started the ATP program at the Commerce department, which was sort of the civilian version of what the TRP was. Their role was to do the commercial but DARPA should stick to the military. In fact, at that time we were called ARPA. And Congress got upset that we were doing things because for some of those programs, it was clear what the commercial aspect was. It wasn't so clear what the military aspects were. And so, they looked at some of these programs and said, "Wow, dual use," but, I mean the dual, the boundary had gone, went to the commercial (
McCray:Okay.) and the military part didn't look as obvious. And so, they said "Okay." So, they said, “DARPA, you're going to do defense. We're giving you a "D" [Laugh] and that's your mission.” They just kept going back and forth. I don't like that. So, it was actually the TRP Program that stimulated the switch to the DARPA, from the ARPA.
McCray:Okay. So, the spintronics program ran until about 1999?
Wolf:The spintronics program ran until 2002.
McCray:2002. Okay. Did you fund work here at Santa Barbara as well?
Wolf:Yes. Okay, not on the spintronics but I had an overlapping program called SPINS.
McCray:Okay, you're going to have explain what that is.
Wolf:Okay, DARPA's a funny place. They don't like to have programs that last forever. I mean some of the way ONR does things — once they start in an area, there's inertia. So, when you start a program in an area, it's very difficult to kill it. DARPA has the prospect that typically you have a five-year program. You have to show something significant. If it looks like it's worth extending then you just extend it for a limited time, but nothing runs forever. So, the only way of really continuing an area is to come up with something that may be related but really has new goals. Okay? So it was only a few years after the spintronics program started that I was very excited by some things that were going on here, by David Awschalom, and also in Japan by Hideo Ohno.
McCray:Okay, what were the things that they were doing?
Wolf:The work here was the work on the very long spin lifetimes.
McCray:The spin coherence research?
Wolf:Spin coherence times were very long on semiconductors, and the work at Tohuku University by Ohno was the fact that you could get a dilute magnetic semiconductor to be ferromagnetic at a really high temperature. It wasn't room temperature but this is gallium manganese arsenide.
McCray:Yes. I was talking to Ohno a little bit about that yesterday. I'm not sure I understand the significance. Why is it of interest that a dilute magnetic semiconductor is ferromagnetic, because it can be a semiconductor and storage at the same time?
Wolf:It can be a semiconductor, but the thing that intrigued me was the fact that you have a dilute magnetic semiconductor where the magnetism is mediated by carriers, and you can change the magnetism. If you have iron, I mean, iron is iron. It has its magnetic properties and whatever they are they remain the same. Materials where the magnetism is strongly connected to the carriers and the carriers are in semiconductors. Semiconductors have the property that you can manipulate the carriers. And their numbers are not fixed, you know, a metal has 1023 electrons and you can't do much about it. With a semiconductor you have various ways of controlling the carriers, and therefore you can control the magnetism. So, you have a carrier-controlled magnetism. And so, this was something that I felt should really be exploited.
McCray:How did you come across Ohno's work? Just by the literature? How did you become aware of it?
Wolf:It was the literature, and again discussing things with other program managers. So they would point out, "Hey, did you see this paper?" "Hey did you see that paper?" And so I was directed. I think it was Larry Cooper, in fact, who said, "Hey did you see that paper?" And I looked at it and I said, "Hey yes, this is really interesting."
McCray:Larry Cooper is who?
Wolf:He was at ONR.
McCray:ONR. Okay. All right. I'm curious about these yearly reviews that you would have. So, you would basically get all the different people from the companies that you're supporting and the universities?
Wolf:And the universities and the national labs together.
McCray:How long would the reviews last?
Wolf:The reviews would last for a few days, typically two or three days.
McCray:Okay. People would give their presentations?
Wolf:Right. We'd have posters. Yes, right. It would also be somewhat of a social event. I encouraged interaction at those. The idea was really to encourage interaction.
McCray:How many people would come to one of these?
Wolf:Well, that depended on the size of the program. For the spintronics program, maybe sixty people.
McCray:Okay. And then the SPINS project?
Wolf:Was even bigger, maybe a hundred people.
McCray:Okay. So I got you off track here but you were talking about the overlap between the Spintronics program and SPINS.
Wolf:So, I pitched a program to explore spins. So SPINS is an acronym for Spins in Semiconductors which became SPINS. And again, the whole motivation was the fact that magnetism in semiconductors, in principle, could be intimately connected with the carriers and therefore you have the new handle on magnetism. It's one idea. And the others, there were now several proposals for interesting spintronics devices using semiconductors. So, the idea was going to be basically physics. So, whereas the spintronics program was actually applied research, in Defense terminology 6.2, the spins program was actually 6.1.
McCray:I'm sorry, what’s the terminology?
Wolf:That was DoD speak.
Wolf:We have, so basic research is called 6.1. Applied research is 6.2.
McCray:Okay. Is there a 6.3?
Wolf:Development – yes there's 6.3, there's 6.3a, there's 6.4 [Laugh] and there's 6.5. So. [Laugh]
McCray:Okay, what's at the end of the rainbow? 6.5?
Wolf:The end of the rainbow is a gun pointing at you. It's the end product. So, the 6.5 is really the production version.
McCray:It sounds like a very linear model of basic research to finished product.
Wolf:It isn't, because if you look at the dollars, the dollars grow exponentially. The 6.1 dollars are usually pretty small, 6.2, there's probably something like a half an order of magnitude to an order of magnitude difference in the amount of money that goes into 6.2. So, as you go up to higher levels — so when we get to 6.5, you know, this is multi-billion dollars. 6.4 is on the order of hundreds of millions of dollars.
McCray:Just so I don't forget later on, where are spintronics devices in this continuum? Where would you put them on this scale?
Wolf:Well, I would put them, well actually where they are now is about 6.4.
McCray:Okay. Which was sort of – I mean, if 6.5 is actually making something?
Wolf:Well, they're not in production, so here's where the DoD is with this. Let me see if I have a picture of it [searches laptop]. So, okay, so the end result of spintronics was exactly the chip that I predicted in the beginning. It was a rad-hard MRAM chip. It was a joint project between Honeywell and Freescale and I brokered that marriage a few years ago. It's not funded by DARPA. It's funded by Navy Strategic Systems and it's going to go into real hardware in probably a few years in an upgrade program. So, it's a fully functional one-megabit chip. So, it actually has all of the characteristics of the 40 lb satellite memory , it’s a megabit instead of being 128kb. Let me go back. So, this is really the culmination of that spintronics project that I showed you before. Yes. So I go back. So, this is it. So, this is the chip which I’m showing you a picture of is actually now a reality. It works.
McCray:Brokering a marriage between Honeywell and Freescale: is that something that DARPA normally does? To get in there and encourage industry to work together?
Wolf:Yes. We do.
McCray:Okay. Is that typical?
Wolf:This was a very complicated thing because other agencies were involved.
McCray:Okay. Can you say who the other agencies are?
Wolf:No, well, I prefer not to. [Laugh] It was a matter — Honeywell's memory project, which I supported, had some significant issues. And so it was clear to me that they were the only ones that actually produced hardware for the DoD, dedicated rad-hard hardware. And so, I worked with some people at some of the agencies who really wanted this memory. And, I said, "Basically we have to make sure that these companies work so that you get what you want." And so the incentive was money. They were willing to pay for it if these companies, if Honeywell would work with Freescale.
McCray:Okay. Again, this is another naďve question, but would spintronics have developed without DARPA support?
Wolf:Well, again I think you should ask the companies that. I think the answer is, "It wouldn't have really developed without DARPA support." But, again, I'm not the right person to say that for sure.
McCray:Okay. You said that this intersection of those two curves that you described where the risk is high at the beginning, I understand, the companies are reluctant to invest in things that are expensive, and risky, and where they don't see an immediate payoff. So, I'm wondering if DARPA hadn't filled in that gap in the beginning, would companies have bothered?
Wolf:Yes. I mean, I think you'll find the answer is that they probably would not have invested in it without the DARPA partnership. But, I think that's a very good question to ask them.
McCray:Okay. Where does IBM fit into all this, simply because again I'm looking around at the people who are at this meeting this week and it seems that there's a lot of people who have some sort of connection with IBM. I mean Awschalom was at IBM, Ohno was at IBM, Von Molnar was at IBM, Stuart Parkin's still at IBM? Is that just a coincidence because of how this conference was put together?
Wolf:Well, I mean, you could say that IBM, at one time, could have dominated in this. Because they were working in magnetism, and their support of this area sort of fluctuates. And I mean, again, they're not sure because IBM is sort of rediscovering itself. And so, they sold their laptop business. They sold their hard drive business. It's not clear. I mean they still are in the chip business but it's not clear that they want to continue making memory. They don't really know where they're going to — see, I mean they're so big a company that they can't really afford a niche market. They don't really want to play in a niche market. Freescale is willing to play in a niche market, and particularly a niche market that they already have, that they already play in.
McCray:And, by this kind of market you're talking about making these MRAM type of chips for specific applications?
Wolf:Yes for specific applications, right.
Wolf:They do a lot of basic work, specialty chips, application-specific integrated circuit.
McCray:Okay. So, comparing IBM to Freescale obviously is apples to oranges?
Wolf:Right. That's right. Yes, it's not clear where IBM is going to go with this because their business model is changing all the time.
McCray:Okay. One of the things also that I'm interested in, and I'm curious to hear if you can say anything to this, are these roadmaps that the semiconductor industry prepares. Are these something you paid attention to when you were at DARPA? Also, were you involved in preparing them? And I'm also curious, just to throw out multiple questions. When you look at these roadmaps from 1990 to the present, there's always this sort of over-the-horizon area that they're talking about, and I'm curious about where spintronics sits now and is it still over the horizon or is it actually coming into view on these roadmaps?
Wolf:Well, it's coming into view.
Wolf:Right. So, I have not played a role but of course I do pay attention to roadmaps because they tell, and you really have to know where CMOS (complementary metal-oxide semiconductor) is going. But, CMOS is the big gorilla. I mean, you're going to be working with it. I mean, so you have to develop — I mean, that's one of the nice things about MRAM is it's very compatible with CMOS. So, that's always an issue. I always worry about all these exotic memories and things that people come up with. If they're not compatible with CMOS they're really not going to go anywhere at least for ten years.
McCray:What do you mean when you say "compatible"?
Wolf:It means that you can build all of the control circuitry that you want in CMOS. You want something like a back end of the line, which means that you build it on CMOS. If you're compatible you could build your devices or your structures on, using the best of CMOS, then you have a shot. But, if you're going to come up with, and have to develop a whole new infrastructure to control the chips to do all of the housekeeping functions for everything else you need on a chip— you just, no one can afford it. Think of how many billions have been invested in CMOS. For example, one foundry now is $2 billion. So this, to my mind, I mean the way to go is to come up with something that's going to work well with CMOS so it's going to compliment CMOS. All right? It's not going to replace CMOS at least for the near future. Now, of course, spintronics has come up on the roadmap as sort of post-CMOS as a potential. But, I mean, if it's ever going to get there. The point is it's not all of a sudden there we're at the end of CMOS and the beginning of spintronics. There has to be some transition. So, the only way even spintronics is going to have a chance to be a real successful technology say ten years from now is if we already start integrating things with CMOS.
McCray:Okay. Is that beginning to happen?
Wolf:And it's beginning to happen. Well, the MRAM is one, and advanced MRAM, but we have to make sure that we stay compatible with CMOS and that we continue to be able to scale with CMOS. And eventually, maybe we'll be able to do some things that are better than CMOS. So, my feeling is that you're not going to see an abrupt change. CMOS until 2016, then in 2017, we see billions of dollars of spintronics devices and that's not the way it's going to happen. At least in my opinion. [Laugh]
McCray:I was curious about this QUIST Program or Quantum Information Science and Technology. That's another program at DARPA?
Wolf:That's another program that I started, yes. And, it spun out of the SPINS Program.
McCray:Okay, we have the initial spintronics program, leading to SPINS, leading to QUIST?
McCray:What was QUIST?
Wolf:Okay. So here's the way that came then. When I pitched the SPINS Program it had three parts. We were going to look at what I called "spin quantum devices," which is different from “quantum spin devices". There's a difference. Okay.
McCray:What's the difference?
Wolf:Okay. The difference is that a regular transistor is a quantum device. In order to understand how it works you have know quantum mechanics, you have to know energy levels and all that. So, transistors, LEDs, lasers, semiconductor lasers are all what I'd call quantum devices, in a sense. Okay? And so, the first part of the program is really, "Can we add spin to all these devices that we know, transistors, and make them better?" So, that was the first part of the program. Look at adding the spin degree of freedom to conventional semiconductor electronics. This is, again, utilizing the dual magnetic semiconductors and the appropriate properties that can be controlled electrically. Can we actually make the devices that are going to be competitive? Just adding spin and making a device that looks and operates very similar to the regular devices. So that was one. That was the first part. And the second part, what I call "spin coherent devices," is where you now take advantage of the first. So, all these devices, I mean even though you need quantum mechanics to understand them, they don't really operate in a quantum regime. They operate in a very classical regime. You know, you're making zeros and ones. Right? A coherent device is using the fact that you now can put a spin in a superposition state. So, there are two aspects of spin that go beyond the normal devices. One is the fact that you can put the spins in a superposition state. The second is that you can entangle them. Okay. So, these are the two additional features of quantum mechanics that you have to explore. And so this was going to explore just the coherence part of that or the first part of using superposition. When I funded the group here, they were the major group exploring coherent effects.
McCray:Okay. So, that would be David Awschalom’s group here at UCSB?
Wolf:Right, David's group. The idea is, could we come up with some novel devices and even optimal switches that could take advantage of the spin coherence. But the program here is really not for the single spins. So, the third part of spins was to look at what I call the real quantum mechanical effects. And here you actually are beginning to look at single spins.
McCray:So who was part of this program?
Wolf:Well, I funded, under that aspect, UCLA, MIT, David Cory’s people doing the NMR. I funded Daniel Loss. Well, I funded a few people. It was initially the smallest part of the program. I mean, they sort of went this way-the first two parts were a lot of money. This last part was less money but it was clear that this last aspect of the SPINS program was just going to explode. But then I realized, in talking to other program managers at DARPA that using entanglement was becoming pervasive, I mean there were a whole bunch of efforts to build a quantum computer using entangles states. And so what I said was, "Well, I mean, I only have the charter in this SPINS program to look at these quantum mechanical spin effects, you know, and but there's a much bigger world out there.” Other people are looking at all sorts of other ways of building a quantum computer, building quantum information systems using, photons, ions, superconductivity.
Wolf:So with a couple of other people we said, "Okay, we really, really have to expand our horizons and do more than just spins." And so we had to worry about algorithms and everything else if you're going to build some sort of a quantum information device. So, that's how QUIST arose.
McCray:Okay. Just because this is nice and laid out here, roughly what would the funding then for something like QUIST be?
Wolf:So, whereas the SPINS program, in its heyday, was between twenty and thirty million dollars a year. For everything, the QUIST program is also between twenty and thirty million dollars a year.
McCray:Okay. So, would it fair, on this is a little graph I'm drawing here, to say that QUIST came out of Part III of the SPINS Program?
Wolf:Yes. At least in my mind that's how it evolved.
McCray:Okay. And, when did that start?
Wolf:So, that program started around 2000 and 2001.
McCray:Okay and it runs for?
Wolf:And it ran until 2005.
Wolf:This program, so this program started — well I don't know if you want to count the GMR Consortium, which started in 1995.
McCray:Okay. So, circa 1995?
Wolf:And the full-blown program started in 1996. The SPINS program started in 1999. So you can see that I used this to build quite a hierarchy of programs.
McCray:Okay. So what are some of the main groups involved with QUIST then?
Wolf:Oh, there was Santa Barbara. There was HRL and UCLA, and we had a program with TRW, and, IBM. There was NRL, Michigan and Georgia Tech. And, there were also a number of small players who did algorithms, the University of Michigan, Columbia, Princeton, UC Berkeley etc. We had a very big consortium organized by the University of Illinois, a involving six universities. A very big program at almost $2 million a year.
McCray:Okay. Now, just two weeks ago there was an announcement from the University here. Two things. One was this formation of this Western Institute for Nanoelectronics (WIN), and the other was a DoD grant to David's group to develop something called a multifunctional chip. Are you connected with it?
Wolf:I'm actually one of the subs on the multifunctional chip project which is a MURI. Okay. And the WIN folks would like me to be an affiliate but there are still some things to be worked out. And that – I actually helped, I know the people at Intel very well.
McCray:Okay. Tell me about the multifunctional chip project. How did that come about?
Wolf:Well, the program manager at ONR basically wanted to follow up on spins, and a MURI is, one of the ways that ONR had to get a large program. And most of the programs that they funded are rather small, a couple hundred thousand dollars a year. But a MURI is nice because it's a million dollars a year for five years.
McCray:And MURI stands for?
Wolf:Multi University Research Initiative.
McCray:Okay. Now, who would the person at ONR be?
Wolf:His name is Chaagan Baatar.
McCray:And this is roughly what, I think this is a $5 million?
Wolf:It's $5 million over five years. So, it's a million a year.
McCray:Okay. So, what will your role in this be?
Wolf:New materials. New magnetic materials and some novel structures, actually novel structures for doing spin momentum transfer.
McCray:I have a couple of general questions. Do you see the work that you're doing as nanotechnology?
Wolf:Well, some of it is. Certainly I would say QUIST would fall under the heading of nanotechnology. I would really call it advanced, or you know if you really want, it's really, just, it's just a roadmap of electronics. I mean, it's sort of in vogue to call it nanoelectronics. The industry is beginning to call it that mainly because they see a way of leveraging resources by going this route, you know, there's this National Nanotechnology Initiative to which we want to steer some of those resources to our projects. So, if you call it nanoelectronics. I mean, I do the same thing because the initiatives all call for work in nano. So, you call your work that —
McCray:Not to sound cynical but a lot of people have adopted the term.
Wolf:What I did wasn't because it was nanotechnology. I did it because I saw some path to something interesting. The question is what's your primary motive? Is your primary motive to do something nano or to do something that's going to be realistic in terms of a real technology, a real innovative technology. And so, I would say that if I'm trying to develop some ideas I'm really looking more toward the innovative technology and then the fact that it may be nano is secondary but, you know, you exploit it. There are all these things that mention nano. I mean "nano" is the way to get [Laugh] macro money. [Laugh]
Wolf:But there are a lot of people who just would say, "Okay. I have to do research in nano because then there's money there, and then try to figure – what they're doing, make it nano and then figure it out. So, my motivation is the reverse. I would actually not think about size until there's some idea or some direction that you head in and then if it happens to be nano you certainly should exploit it.
McCray:Okay. Now, you were at DARPA when the National Nanotech Initiative was being put together. Were you aware of the discussions that were being held about getting this big initiative together?
Wolf:Well, we came in at the tail end of that. And, of course, what we were asked to do was this: everybody had to do an accounting, so every program manager had to say how much of his money he was spending in nanotechnology, and this was all rolled in, and then the president could say, "We're investing so much in nanotechnology." But it wasn't as if we specifically invested in it, we just counted what we were investing, and so I would put in a large part of these programs as being work in nanotechnology. But, it wasn't the motivation of doing the work in the first place. We weren't doing this because it was nanotechnology.
McCray:Sure. Okay. And, again this comes back to my earlier question about how the various science managers in the DC system interact with one another. A follow-up would be, while you were enjoying the SPINS program, and spintronics, and QUIST programs, were you interacting with people at NSF?
Wolf:Not so much at NSF. I had some bad experiences with NSF. I'll be honest. Maybe you don't want to publish it. But, I interacted with other program managers in the DoD.
McCray:Was it particular things about how the NSF runs their program?
Wolf:Yes. I'll tell you an experience I had. I mean, I knew who the NSF program managers were and of course I knew Ulrich and I was friendly with Ulrich and a few of the others, people from NRL who had worked for me who were now over at NSF. I managed a program called VIP, which was called Virtual Integrated Prototyping. And again, part of that was leveraged for spintronics. A third of that program was related to spintronics. It was developing new deposition tools for use for making spintronics devices. That was part of it. Okay, so the original idea was, I mean NSF was very interested in this idea of doing this. And, it was a lot of applied math to basically do simulations of how you actually would grow, a lot of film-growth simulations and then trying to understand how to design a system for making the films in an optimal way. So, we had one for the magnetic films. We had one for superconductors. There was a superconducting project. And, the third one was the growth of basically III-V materials. Not silicon because silicon already had a lot of folks working in that arena. And so, NSF was interested in the original idea, so we spoke to the NSF program managers and we decided to have a joint program. We would put in half the money and they would put in half the money. And I think we, DARPA, were putting in $5 million. And NSF said when they put their money in they have to go through their process. So we sent the DARPA money to NSF and they managed it, they got the proposals and they had the panel. Well, I mean the program took three times as long to actually get into place than it would have if it was just the DARPA program, number one. They had their panels and it was – I mean I was on that. I sat through their panels and it was painful for me.
Wolf:Well, because they would have these fifteen people who could never agree and would spend hours arguing. Right? Okay, number one. We went through this tedious process of who could read it, who has to be recused. I was sitting in the room thinking, "What are we doing?" Okay, so that one never worked. Number two, at the end they only put in a million dollars. So, what was originally what was supposed to be a bigger program for some reason they had a budget cut, blah, blah, blah, anyway so they wound up only putting in one. We put in five million and yet they managed, they got the control of the whole process. [Laugh] And so, I was really mad. Right. I was not a happy camper and I decided that was the last time that I worked closely with them.
McCray:When was this program, the VIP?
Wolf:That was in 1997.
McCray:Okay. What you just described matches my understanding of how DARPA works versus NSF, because DARPA's a much more nimble agency in terms of direct funding towards things.
Wolf:Yes. I'll tell you it was just a very frustrating experience.
McCray:Okay. All right.
Wolf:So even though, you know, I liked the people at NSF but they have this process that really drove me crazy because in the end, I mean, I don't think the choices were -– I mean, I think we did okay but the process is very flawed, frankly. I'll be honest with you.
McCray:Yes. One of the things that struck me about NSF is that anytime political winds blow in DC it seems like they try to respond to that, but also their mission has shifted from basic research to now doing some kind of applied?
Wolf:And, of course, not only that but it's all these other things, all these other requirements about education and training and diversity. I mean, it seems like that's diverting from – I mean at DARPA we wanted to get something done. You know, we have a mission. We have something that we want to accomplish at the end of the day and that's the most important thing. And at NSF there are all of these other distractions that I think become issues. I mean, the problem is that, you know, you say the peer review is very, very fair. Well, the problem is that there are personalities. It would be fair if there were no personalities that come into the decision-making process. You have fifteen people and some of them just have axes to grind with some of the proposals. And, you have to be able to sort that out. And, sometimes you can, sometimes you can't.
McCray:So, how do you review at DARPA then whenever something is submitted?
Wolf:We basically review them amongst ourselves, between the managers. But primarily we'll review it, we'll get three of the program managers who are somewhat familiar, or we'll get some people from the DoD. We usually don't get academic reviewers. We strictly speak -– and so therefore nobody has a real ax to grind. Because these are the peer review is really academic so they do have a — you know, the point is it seems like it's backwards. They have more of an ax to grind than the DoD reviewers who are not competing at all. They're not in any of the programs. They're not competing for any funds. So, I think you get a better process.
McCray:Okay. This has been really helpful. This really packages things up nicely and I want to make sure you get over to ITP with plenty of time. One last question, which was, as I'm trying to understand the development of spintronics, from the 1980s through today, who are key people that you would suggest I talk to?
Wolf:Well, for sure the people at the companies that I mentioned. And you might also want to talk to Stuart Parkin, who was a major player.
McCray:Yes. He's on my list.
Wolf:And, of course, David. You might talk to Eli, Eli Yablonovitch. He's sort of marginal but he has some very novel ideas in terms of the quantum applications on spin.
McCray:Where is he?
Wolf:He's at UCLA.
Wolf:If you want to get an Intel perspective you might talk to George Bourianoff.
McCray:Okay. Anyone else?
Wolf:Well who else — I mean, Stephan von Molnar. You might talk to Larry Cooper.
McCray:Okay. Where is he?
Wolf:He's now at Arizona State, but he was the program manager at ONR for a long time in the area of magnetics. And you might talk to John Prater. He’s at ARO.
McCray:Okay. The person who you wrote this paper with, Treger, and I won't even begin to try to pronounce . . .
McCray:Would they be good to talk to as well?
Wolf:Yes. You could talk to them. They really were colleagues of mine. They worked for one of the small companies, support companies. So, she's actually a PhD and he's an engineer. But, they worked very closely with me when managing the programs.
Wolf:So, they were, they were contractors. But, I relied on them for a lot of the legwork.
McCray:One name that someone suggested I ask you about is Jane Alexander.
Wolf:It’s actually Xan Alexander.
McCray:Xan, okay. Is that a program or a person that you had any interaction with?
Wolf:Yes, she ultimately became the Deputy Director of DARPA. And, she was always a supporter but not directly in a hands-on way. But, she had a program called the ULTRA Program.
McCray:What was ULTRA?
Wolf:ULTRA was basically the precursor to nano. It was ultra fast, ultra small, ultra performance.
McCray:Oh. Okay. Yes, that might be interesting, too. Okay.
Wolf:Yes, she's now in Homeland Security. Jane Alexander is her real name. It's Jane Alexander, but she goes by Xan to her friends. Xan is sort of her nickname. See, the Xan actually comes from "Alexander," not from her first name. Yes, I know her very well.
Okay. Well, let's get you over to KITP so you've got time to get to your meeting. Thank you very much. As I noted, once we get the transcript done and edited, we’ll send it to you. Thanks again for your time.