Notice: We are in the process of migrating Oral History Interview metadata to this new version of our website.
During this migration, the following fields associated with interviews may be incomplete: Institutions, Additional Persons, and Subjects. Our Browse Subjects feature is also affected by this migration.
Please contact [email protected] with any feedback.
This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.
This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.
Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.
In footnotes or endnotes please cite AIP interviews like this:
Interview of David Arch by William Thomas on 2008 December 4,
Niels Bohr Library & Archives, American Institute of Physics,
College Park, MD USA,
For multiple citations, "AIP" is the preferred abbreviation for the location.
David Arch is a physicist who has worked primarily on solid-state devices for Honeywell in Minnesota. This interview was done as part of the American Institute of Physicists in Industry project, and is a follow-up interview to one conducted in November 2005 by Joe Anderson and Orville Butler. The interview discusses Arch's family and youth in rural Illinois, undergraduate education at St. John's College in Minnesota, graduate work on lattice dynamics at Iowa State University under Constantine Stassis, experience working with the Ames Laboratory, and on neutron beams at Oak Ridge National Laboratory. Arch was hired by Honeywell in 1980, and worked as a researcher in different laboratories in the Minneapolis area. The interview discusses his research and development project, his transition to management and business development positions, some corporate business and research history, and his recent transition to the assessment of prospective technologies.
Joe Anderson and Orville Butler have previously interviewed [David Arch] on November 14, 2005 for the History of Physics in Industry Project, and so we’re continuing on with a life and career interview today. So I guess if you want to tell us, we can just start at the beginning with family background, education, what got you into science, and that sort of thing. That’s as good a place to start as any, I suppose.
Well, that’s going back a long ways. I guess I got interested in science mainly through some good high school teachers. I had the stereotypical absent-minded professor for a physics teacher in high school, who just was a brilliant person, and I guess cultivated my interest in science at the time. Then my older brother was in college; he was six years older than I am, and he actually majored in physics.
Where was that?
At Loras College. It’s a small liberal arts college in Dubuque, Iowa. So there was a little bit of a history in the family already, but it was primarily I think the academic background in high school, both in mathematics and science. We had teachers that really encouraged us to pursue it. So when I went as an undergrad, I went in with the idea of majoring in physics. I changed — The first year and a half, you always wonder if you’re making the right decision and all that, but ultimately I settled into doing physics. This was back in the early ’70s, and whether I had a choice or not when I graduated from college — the economy was similar to now, the job situation wasn’t very good — but again, I had a couple of professors that thought I was quite capable of going to graduate school. They encouraged me to go to graduate school, so I decided to pursue a graduate degree in physics. I went to Iowa State University, which at the time there was one of the Department of Energy’s laboratories, Ames Laboratory, and had very good solid-state physics department, really in rare earths. They developed a reputation, actually from the war, about refining and purifying rare earth metals, and the solid-state department had kind of built up a reputation around that. So typically your first year or two years in grad school, you get exposed to a lot of different research areas. Really, the solid-state physics was the area that interested me most, so I pursued solid-state physics. They had a pretty active research group doing neutron scattering, doing lattice dynamics and magnetic properties of materials. Ames happened to have a nuclear reactor on campus at the time; they actually shut it down a year or two after I joined the group. I worked with Professor Constantine Stassis, who was maybe four or five years out of graduate school. His thesis advisor was Cliff Shull, who won the Nobel Prize back in the mid-’90s  for all of his work on neutron scattering, neutron diffraction. He was an eccentric Greek, Constantine Stassis, but he kind of exuded physics. He could get concepts across very well, so I learned a lot from him. I spent the next five years doing research on lattice dynamics and materials, and at the time, there weren’t too many places in the world that were able to do this. Ames was one. Oak Ridge was another, and we actually developed a relationship with Oak Ridge, and I spent a lot of time in Oak Ridge National Labs my last three years of grad school, where they had a high-flux isotope reactor that we could use for neutron scattering experiments. So the types of materials I was working on, it accelerated my research by years by being able to use the HFIR. So I did research on lattice dynamics of group — I can’t remember the classification now. Zirconium, titanium, and hafnium, the Group 4A metals, I guess. They’re all hexagonally close-packed materials at room temperature, and they undergo a phase transition to body-centered cubic at high temperatures. So we were starting to investigate what happens with the lattice dynamics as you go through that transition. So we built furnaces and looked at the temperature properties of these, and tried to understand what happens with the electron-phonon interactions as you cross this phase transition. My thesis work was on the lattice dynamics of those three materials. This was kind of a long-term project with the group, and I carried on the work until we felt like I had enough data for a thesis. We actually were able to look across the phase transition and do some measurements on the BCC phase, but it was still pretty early on in the research in looking at that phase when I left. At the time, my goal was actually to go in academics — I wanted to teach. And I got several teaching offers at more liberal arts schools that had research places. I also applied for a post-doc at the time the National Bureau of Standards, now NIST. But none of those came through. Well, actually I had a couple of academic offers, but just for references, the pay for teaching at a liberal arts school in 1980 was about $12,500 or $13,000 a year. My wife and I spent days trying to figure out how we could live on that little amount. I was probably making $8,000 as a grad student, and we had a son at the time, and we just couldn't figure out how to make it. So at the same time I was looking in academia I was looking at industry, too, and interestingly enough, at that time the oil companies were hiring like crazy. They were looking actually for physics with a lot of background in signal processing or computer capabilities. Of course back then, the PC wasn’t around; it was still using mainframes. And I did a lot of computer modeling in my graduate work. So they were hiring like crazy and I actually got a couple of job offers from people like Texaco and… I can’t remember.
Shell by any chance?
It wasn’t Shell. The names have all changed. But I do remember the computational abilities they had at the time were phenomenal. But it wasn’t going to be solid-state physics; it was going to be really geological surveying and looking at seismic data. While it was interesting, it wasn’t really what I wanted to do. It turns out I had a couple of contacts that worked at Honeywell at the time. Honeywell in 1980 — research has changed dramatically in industry in the last 25 years, but they literally had what we called a research laboratory at the time. It was less connected to immediate product needs.
They were doing semiconductor research, and gosh, I can’t remember the other things that were going on. A lot of materials research, basic materials research. It wasn’t necessarily “Hey, we’ve got a product here that’s going to benefit from it.” They did have at the time a strong business in infrared materials, and Honeywell was big in infrared detectors and infrared imaging systems. At the time, there was a material called mercury cadmium telluride. It’s an alloy; you can adjust the composition to change the band gap in the material, but the band gap in the material is anywhere from maybe 2 microns out to 30 microns. Of course, that’s the ideal window for infrared. I hope I’m remembering my dimensions right here. But the material is extremely difficult to grow and produce. So they had a very active research effort and they had a whole business that was focused on building these arrays. We were looking alternatives to that to get lower cost detectors and that. So when I came in, they were continuing research in mercury cadmium telluride. They were looking for a means of a better growing the material and also better fabricating devices on it, but they were also looking for alternative materials. The first project I got was to look at, could we dope silicon in the matter to create infrared band gap material. There were some theoretical papers out about what they called x centers in silicon. You may be really straining my memory here in all the things that are involved in it. But my first project here was to look at these x centers in silicon and see if we could get the doping levels up high enough in silicon to make them a worthwhile infrared detector. So I spent the first year or year and a half trying to grow this material, so I was learning, and again, some of this came from my graduate work. When I did zirconium, hafnium, and titanium work, we worked with the materials group at Ames Lab to grow this material. So we had hands-on involvement in growing single crystal material. So I was aware of the basic methods of growing this material, but when I got to Honeywell, we built the furnaces to grow to see if we could create these high doping levels in silicon. I spent the first year and a half trying to do that. And then also worked with the group on the mercury cadmium telluride and trying to develop better fabrication processes. This is still back in the early days — I shouldn’t say early days of ICs, integrated circuits, but even then, we want to fabricate these things similar to integrated circuits. The integrated circuit capability at the time, compared to what it is now, was very immature, but they were making fairly large for — Mer-Cad-Tel you built a single device, where ICs at the time, they were building 4K RAMs or whatever, and of course now they’re doing 16 megabit. They may be even higher; I’ve lost track of it. But part of my job was also to develop the fabrication technology to build this, so that’s where I started learning IC processing. So it’s interesting. When they hired me, they hired me because, the reason they gave me was because of my physics background. I had learned a broad array of capabilities, and they felt that the physics background provided a much diversified capability. You know, I could do materials work, I could the theoretical work, I could do the fabrication. I could figure all this stuff out, and that’s what I was doing the first couple of years. I was learning growth methods in materials and fabrication, the theory of infrared detectors and that sort of thing. So I slowly developed a capability actually in both silicon and the Mer-Cad-Tel about how to fabricate these devices. So that was my first exposure to integrated circuits. They weren’t integrated circuits, but the fabrication processes we used, the materials and whatever were different, but I learned a lot of fab capabilities that way. Basically after a year and a half I kind of showed that you just were never going to be able to get the doping levels in silicon high enough to be able to do this. Again, there was a big drive at the time. Mer-Cad-Tel was a hard material to deal with, and by the way, when you operate Mer-Cad-Tel as an infrared detector, you have to operate at about 77 Kelvin. You have to cool it down because of noise in the materials and the band gap is so small. So people were continuing to look for alternatives. We had a very good infrared group at the time, both theoretical and experimental. At the same time that we were working infrared, we had another group that was working silicon sensors and using micro machining in silicon. I don't know if you’ve heard of MEMS, microelectromechanical systems. It’s a very big area in electrical engineering, a very big area in places like UC Berkeley, Wisconsin, MIT. The whole idea behind this is there are certain properties of silicon, the crystalline properties, and they’ve also developed fabrication techniques that allow you to actually machine structures into silicon. So rather than just build an IC, you build a MOS fit or bipolar transistor by implanting things into the silicon and changing the conductivity. You literally can carve out the silicon into different forms. We had a group at the labs that was doing work in micro machining silicon, and they were trying to build a mass flow detector. Basically, it was an equivalent to a hot wire anemometer on silicon. The way they did this was basically they built a bridge of silicon that was all the silicon underneath was etched out, and that bridge was thermally isolated from the rest of the device. They were able to show that you can do mass airflow with that. Well, we were looking at it and said, “Gosh, if you can thermally isolate this, by putting the right material on top of it, you can make an infrared detector out of it,” what’s called a bolometer, an infrared bolometer. We started it in an effort in that area, and actually, I had set up a test lab for testing these infrared sensors and test capability. It wasn’t a full lab, but we could measure the properties of this. We took one of the mass airflow sensors, and the first bolometer we ever built is, I took one of those mass airflow sensors and lit a match and then blew it out and let the soot onto the infrared detector. Basically, I put the match underneath the mass airflow sensor, and the soot basically deposited on the mass flow detector. But what the soot did is basically it was an infrared absorber. So then we exposed this to infrared radiation and looked at the response properties of the detector, and found out it actually operated as an infrared detector. Now that by itself, it was an interesting experiment, but again because of silicon and again thinking about the parallels to integrated circuits — You know, the reason why integrated circuits had such an impact is you don't just build one transistor; you build millions of them very small. Well, we were able to build arrays of these infrared detectors now. And that research led to a whole area of uncooled infrared detection that now is a big business — but not for Honeywell. We actually sold the business in the late ’80s, but that research led to that. So it was kind of fun being part of the early stages of that and building the first detector and all that, but that’s kind of the first impact I had I guess. To me, the x center work that I did was interesting physics and interesting work, but it just was a null result. You know, no, this is not going to work. That kind of led to a whole area where for the next seven or eight years we had a huge group doing work in this, and doing really some just state-of-the-art research in developing new detection technology. I don't think there’s any infrared business out there now that doesn't have this capability that was built on that. So the first three or four years at Honeywell were spent doing materials research, fabrication of infrared technology. I don't know how much detail you want me to go into this.
Yeah, I guess I didn't want to interrupt your flow there. You were going pretty good. I suppose it would be nice if we could kind of go back and sort of fill in details if that’s all right.
Okay, so maybe just cover the top level all the things that I’ve done?
I don't quite understand the top level.
Well, I don't have to go into that much detail. You asked how I got interested in science and whatever and I just kept going. But anyway, that’s kind of how I evolved from — It really was mentors and people instilling interest in me in certain areas. Individuals do make a difference. I was with Stassis, who was my thesis advisor. I can remember we were up at 4:00 in the morning drinking coffee and looking at data night after night when we had time on the HFIR at Oak Ridge. It was an exciting time to me. It was just fun to do. Quite frankly, I could have been a perpetual graduate student. I think it was just fun to do. It’s just doing the research and collecting the data and analyzing the data. I’d love to be able to do that. The pressure was less, “You’ve got to have something by next week.” It was just focusing on the research that I really enjoyed.
So just kind of covering this initial part, you grew up in Illinois I think I was reading in the previous interview.
Yeah, Princeton, Illinois. It was a small town about 100 miles straight west of Chicago, an agricultural town, [population] about 7,000. A very good high school, very good academics back in the ’60s, and very dedicated faculty. They were very good at instilling the desire to go learn more, encourage people to go to college. You know, we were kind of the post-war baby boom.
Did your parents do anything technical at all?
No. My dad was the only one in his family who went to college, and he got a degree in marketing from the University of Illinois.
Yeah. He couldn't convince any of my brothers or sisters or me to go to University of Illinois, and in retrospect, for solid-state physics it would have been an ideal place.
That would have been all right, yeah.
John Bardeen was there. But [my parents] were very much supportive. One of the greatest gifts my parents ever gave me was the fact that I came out of college not owing a dime. All of my brothers and sisters, my parents sacrificed a lot. They just thought the education was extremely important.
Did your brother do anything with his physics degree?
Well, my brother was going to teach. He actually had a job lined up, and this was 1968. He also had a draft number that was pretty low, so he ended up enlisting in the Navy and spent four years in the Navy. When he came out, his enthusiasm and whatever for that had kind of drifted away. So he went into the insurance industry, and he’s been there ever since. He lives out in Connecticut. But at the time he was ready to teach, and all he was looking for was a — I forget what the Draft Board calls it — a dispensation or whatever. It was very clear in 1968 that the Draft Board wasn’t going to do that. So he had a choice probably of either being drafted or enlisting with his number, and he chose to enlist. But he left all of his books at home, so I was always looking through his physics books and that sort of thing. But I had a younger brother that went into industrial engineering; that was about the only science. I’ve got another brother who’s the chairman of the English department at Michigan State. He got his degree in English, but really it’s history. He did his graduate work on Cotton Mather. I don't know if they can separate history and English, going back that far. But anyway, I was the only one that went for “pure science,” and Tom went for industrial engineering.
How many brothers and sisters do you have?
I’ve got three brothers. So one was in physics, one in industrial engineering, and one was in English. Then I’ve got three sisters.
Oh, it’s a big family then.
Yeah, yeah. It was a big Catholic family, and it was back in the ’60s. My sisters all went into education. One’s a guidance counselor, and actually she’s a guidance counselor in Batavia, Illinois, so she gets all the Fermilab students, which is interesting, hearing all the stories from there. Then my other sister is a principal of an elementary school, and my other one just went back to work. She majored in nutrition. Her kids now are old enough that she’s decided to go back to work, and she’s working at one of the elementary schools. She actually lives here in town. She helps with planning the menus and all that for the kids and that sort of thing. Steve is the English major. We’re the only ones that went for a Ph.D., but Mary Paula, the principal, has her master’s degree, and Susan, the guidance counselor, has her master’s degree. So education has always been important in my family, and most of the kids, my nieces and nephews, that’s something we’ve encouraged. Now it was instilled in us by our parents, but education was extremely important.
Were there any sort of extra-curriculars available as far as science-sort of things went?
Not really. I guess the only thing I remember is our eighth grade class trip was always to the science museum in Chicago. The only thing I ever remember about that is they have a World War II submarine. You go down into a German submarine. There was one other thing I was thinking of; it escapes me now. But it was go to class and probe the teacher, and if you had a teacher that was interesting, they’d push you a little bit further. But there were never any extracurricular type of things that we did. Really, even as an undergrad, there wasn’t that much in terms of extracurricular activities to get involved in.
Yeah, because you came up to St. John here [in Minnesota].
Yeah, St. John’s.
What made you decide to go to St. John’s from Illinois?
Well, my interest, when I was in high school, I was fairly active in sports, and I love to run. I played football, which was in retrospect probably not the — It was more the big thing in town, so the peer pressure or whatever, I felt it. But I love to run, and so I ran track. And one of my running partners, who was a year older than me, won the Illinois State Mile Championship when I was a sophomore, he was a junior. There were no classes in Illinois at the time, so he was the top runner in the state of Illinois. So his senior year was going to be a really great year. That was the plan. So I started running with him, and we ran every day through the winter. He ended up finishing second in the state his senior year. But all of that training and whatever, I just enjoyed the sport. So I was looking for a place to go where I could also participate in sports. And it turned out — I sent letters out and whatever, but there are always admissions counselors from universities and colleges stopping by. One of the guys stopped by, his name was Paul Barnaby, and he actually lived in a town 15-20 miles from us, or that’s where he grew up, but he was with St. John’s. He talked about, hey, they had a great team, so I wrote the coach. I wasn’t a fantastic runner; I just loved to run. And he invited me to the NCAA Division III championships in Chicago in November of that year. This was my senior year. My dad drove me up there. If anything was going to turn me off to go to St. John’s, they ran it in six inches of snow, and guys with beards were coming in with icicles pulling down from their beards. But anyway, all the guys from St. John’s came up and just treated me like I was a member of the team. So that, and they had a physics major, they offered a physics major. They had a very good science department.
So you were already interested in physics specifically.
Yes, I was already interested in physics specifically, but not just physics. They had a very good science department and very good pre-med. I wasn’t quite sure what exactly I was going to do at the time. But that all kind of just convinced me that this would be a good place to go. I had never visited there. St. John’s was run by the Benedictine monks, and there was a Benedictine abbey 20 miles away from us, too. In fact, that’s where Paul Barnaby went to high school. So I actually knew some of the monks over there because they would come over to Princeton whenever they needed help at the church there. So my parents thought this was great. You know, “Dave’s going to go to a Catholic school.” So that’s how I ended up at St. John’s.
Were there any other places that you applied to or were looking at seriously?
University of Illinois, and I was accepted there. As a parent now who sent one of his kids through private school versus public school, my dad should have put his foot down more because it would not have been a mistake going to University of Illinois from a science point of view. I mean, it would have been better. I think I would have probably gotten a better science education, but I don't think I’d have gotten the other stuff that I got from St. John’s. I did a lot of things. I was all over the map, and they kind of let me do it. I took organ lessons. I took piano lessons. I explored a bunch of different areas that I would have never done if I had gone to the University of Illinois. But it was a combination of the science that they offered and the small school and the running. I ran my freshman to junior years. It just was too much by the time I got around to my senior year. In fact, I injured my foot between my junior and senior year, so I ended up not running at all my senior year. But that’s kind of what drove me in that particular direction. Interestingly enough, when I went to grad school, after my first two years of grad school, there were a couple of guys there that got me involved in running again. That was my release valve the last three years of grad school. That was the only thing I did other than focus on my research, to the deprivation of my family, too, that I did for any sort of exercise was I’d run. When we were at Oak Ridge, the solid-state laboratory, Oak Ridge National Labs was about eight miles out of town — I’d run from the labs to the apartment every night; that would be my exercise. So that’s kind of how I ended up at St. John’s is there was no way I could go to University of Illinois and run. I just wasn’t that good. St. John’s was a Division III school, and the emphasis wasn’t on performance, although we had some really good runners.
Were there a lot of people doing physics at St. John’s?
The typical graduating class was maybe six to ten people a year that graduated in physics, so it wasn’t enormous. The faculty was maybe five or six faculty members.
And it’s mostly teaching, I have to expect.
It’s mostly teaching. A couple of them would go away for summers and do their work. Actually, interestingly, one of the nuns was a student of [James] Van Allen at Iowa, and she would go back there during the summers and work. But most of it was teaching.
Do you remember her name?
Gibson. Sister something Gibson. I don't remember her first name. She actually left after my sophomore year, and I think she left primarily because she — Of course she was a nun, so she didn't have much choice, but she wanted to stay closer to the research. But it was primarily teaching. One of them actually tried to get some more advanced projects going. I do remember there was a group of students that was a year or two years ahead of me that built a radio telescope on top of the physics building, and they were trying to do radio astronomy with the physics professor. When I was a senior, again, we didn't have a lot of extra, outside of the classroom type of experiments.
Opportunities for —
Right. But interestingly enough, my senior year, one of the professors I had had worked for Honeywell research labs and gotten laid off back in the early ’70s, back when the economy wasn’t very good. That’s when he took the job at St. John's.
What was his name?
Dr. James Levine, I believe his name was. But he had made an arrangement with Honeywell so get some of their old cryogenic stuff, so that’s how I made my first contact at Honeywell. We went down to the labs and picked up this cryogenic stuff, and I set that up, and we started doing Hall measurements the last semester that I was there. So I was able to actually do some experimental work with the Hall setup before I left.
Did you do anything in the summers?
No. It was going home and working whatever job I could find during the summers, but there was little opportunity. They always have something posted on the walls in the physics department that you can apply for this internship during the summer, and I did that a couple of times. But I just never got accepted into it, or sometimes it wasn’t an exact match or my qualifications weren’t right for it. So I did not get the opportunity to do some hands-on stuff. Interestingly enough, since I’ve been here, although the last four or five years hasn't been the case, but I managed several different research groups here over the course of 20 years. We always had student interns coming in. We’d go out and seek them. You know, sophomores, juniors in college who had interest in a particular area, either in modeling or doing experimental work and get them in during the summer and pay them as interns. I would have loved to have an opportunity like that, but I just didn't get the exposure to it. I kind of went in — it was all kind of the theoretical, not a lot of hands-on stuff, until I really got in the laboratory my second year in graduate school.
So the grad school transition, then. You went to Iowa State. Were you looking elsewhere?
Well, I applied to several schools, but to be honest with you, I don't remember who else I applied to. Iowa State — again, a couple of the professors at St. John's, one went to Iowa State, and the other one had worked with a couple of professors at Iowa State. So I got letters of recommendation there. But I don't recall where else I applied. It’s just been too long ago to remember. I did apply to several schools, though.
It’s just kind of obvious that grad school, you were just good to go kind of straight on?
Yeah. Interestingly enough, I interviewed with 3M. That was the only company that I interviewed with as an undergraduate, and I remember being so nervous in that interview that I could hardly talk. But I think it was probably more the influence of the professors that I worked with that said, “You know, if you’re really going to do anything in physics, you’ve got to at least have a master’s degree.” They kind of convinced me not to — I did not pursue finding a job that hard after my undergraduate. I went through the 3M interview; my heart really wasn’t in it, especially after talking to the professors. When I got accepted and got a teaching assistantship at Iowa State, I was a single person, so there would be no problem living on that. So it was just kind of a natural thing to go on.
Okay, just for the sake of the transcript of the recording, I want to get the dates on there. B.S. from St. John's was in 1974. [Yes.] So you would have graduated from high school in 1970. [Yes.] So then you went on to Iowa State and ended up in the solid-state as you were telling me earlier. [Yeah.] So when you got to Iowa State, I guess there was sort of a general period of classes preparing for a comprehensive exam of some sort?
Yeah. Typically what you did was the first year was all academics. We had a seminar maybe every week or every month where each of the professors would come in and talk about their research, and it gave you the opportunity, if you will, to start thinking about what you wanted to do. They were really trying to line people up for the following summer, because the summer is when you start diving in to do the research. Or you’d pick a professor, pick a topical area, nuclear physics or whatever, and pursue that. So it was exposure to that the first year. I guess I had at the time already decided solid-state physics was going to be the focal point, although it didn't necessarily — you could have switched, but I ended up focusing on solid-state physics the first year. I think most of the graduate students picked either solid-state or nuclear or elementary particles or astrophysics — those are kind of the four big areas. You pick one of those, and you’d also be taking quantum mechanics and thermodynamics and electrodynamics also — the courses you have to take no matter what major or what focus you were in. So I took the basic introduction graduate course in solid-state physics that first year.
Were there a lot of people in the Ph.D. program?
Yeah. I couldn't tell you what the total was. I would say there were probably 70-80 students in the program at that time. I don't know what it is now, but maybe 70 or 80. We had maybe 20 come in in my class. I remember the first day sitting with the chairman of the department and the other 20 guys — you’re kind of looking all these over, and here someone’s from MIT, and I’m sitting here from little lowly St. John's. You know, everybody came with a different background, and some had a lot of experience and some were brilliant. It was a pretty diverse background class, and also diverse in terms of their physics capabilities. But I think it was somewhere around 70-80 graduate students for the whole department. The physics department at Iowa State was fairly large, primarily because of the association with the Department of Energy and Ames Lab.
So you kind of cast your lot then with the solid-state the first year?
Yes, I kind of cast my lot. I guess again maybe it was the influence of the some of the previous teachers that that was going to be a growing area and there would be a lot of opportunity in it, and it interested me. Again, through these weekly meetings with professors, the work on lattice dynamics and working — Again, they had a reactor on campus, so as part of the seminar, we went out to the reactor and saw the work that they were doing and what they were trying to do. That interested me. Maybe it was the big machine physics or whatever, you know, “Wow, a nuclear reactor!”
Had you had much exposure to solid-state theory at all at St. John's? Or that’s not kind of the staple of the undergraduate physics curriculum?
No, no. Only through the work we did the last semester on Hall measurements on materials. Levine was a solid-state physicist, so he had his prejudices I guess, but he definitely suggested that be an area that I should take a look at.
So then you get involved with the lattice dynamics, the reactor there. [Yes.] One thing [I’d like to clarify is] you were a research associate for most of your time at Iowa State at the Ames Laboratory. It says Minneapolis on there. Is that just a typo? [Yes.] Yes, I thought it might be. It was all there?
It was centered out of Ames, but in ’76, the Department of Energy decided to shut down the reactor at Ames. I’m not sure what the reasons were; probably financial. But interestingly enough, they thought that the group there was a very good experimental group, and what they offered was a couple of beams at Oak Ridge. There were two reactors at Oak Ridge: Oak Ridge reactor, which was a ’50s-vintage reactor; and then a high flux isotope reactor, which was fairly new. They offered us beams on the Oak Ridge reactor. So we had to build — What we did is we took the spectrometers that we had at Ames and we moved those down to Oak Ridge, and I was involved in some of that and setting that stuff up. It wasn’t the ideal setup at the Oak Ridge reactor, but the promise was that we’d get beam time on the HFIR. Like I said, some of the materials that you work with scatter neutrons more difficultly than other materials. The hafnium, which was the third element that I was working on, the titanium, zirconium, and hafnium has a pretty bad neutron cross-section. So if I were going to do the work at Ames, I’d still be there, because the flux of the Ames reactor was not very high. The Oak Ridge reactor was a little bit better, but it was still going to be difficult. The high flux isotope reactor, though, if I could get like a month or two months’ worth of time on that, it would allow me to collect a lot of data.
When you say high flux, you just mean the number of neutrons that go through?
Number of neutrons coming out, yeah. It was amazing the difference, and when you were working at 100 times faster than what Ames was. I may have my numbers off, but it was at least two orders of magnitude better. So the move was inconvenient, but the opportunity for the research actually expanded because, still we had two beam lines, although it took us a long time to get the things up and running. But then we could also get beam time on the high flux reactor. Oak Ridge had a solid-state physics department, and they had their own neutron diffraction group, a very well known, internationally known group. The agreement was that we got time when they weren’t using it. Of course, they were trying to use it all the time, but between experiments, we would get a weekend. They would give it to us, and my thesis professor —
Yeah. If he could get four days on it, we would work continuously four days, and that happened several times. My butt was dragging. But we ended up setting it up down in Oak Ridge. So what I would do is go down to Oak Ridge when we had experiments ready to be set up and work for a period of time and then come back. That extended — I had two six-month visits down there. We actually ended up renting an apartment, we the group rented an apartment down in Oak Ridge and had a car, so almost all the time, one of the grad students was down at Oak Ridge running the experiments and setting things up. Ultimately we hired a post-doc to come down and take care of it full time. So I’d go down and stay at the apartment. My wife and son came down with me a couple of times during the long visits.
Did you meet her in Iowa?
No, I met her when I was at St. John's. So they would come down on the long visits, but there were several times when I’d be down there for six to eight weeks by myself. Then all of a sudden I’d get a call from Stassis on Thursday [saying], “Hey, we’ve got beam time Friday to Monday. Go set it up. I’ll be down there tomorrow.” We would literally look at the data as it was coming out, so we hardly ever left the reactor. I’d just be glad when I drove him back to the airport because I could go home and sleep. So while I was still at Iowa State, we established this collaboration with Oak Ridge, which to my knowledge is still going. Stassis retired a couple of years ago, but Ames Lab still had those beams down at the Oak Ridge reactor, the high flux reactor, which has been upgraded since then. I kind of lost contact. I kept in contact with Stassis but kind of lost contact with regard to what exactly is going down in terms of the research.
It reminds me, actually, a little bit, when I was an undergraduate, I worked in a nuclear magnetic resonance lab at Northwestern, and we would always go down to the — or they would always go down. I only went down once to the High Magnetic Field Laboratory in Tallahassee. So it’s sort of a similar thing when you have time on one of the magnets that you just make the most of it.
Yeah, there’s a high magnetic field lab at MIT or Harvard. No, at Lincoln Labs I think. At Lincoln Labs I think there was a high magnetic laboratory. Again, that’s 25 years old. [Francis Bitter Magnet Laboratory]
Yeah, I’m not sure when they set the one up in Tallahassee, but I think it might be related to that one. Anyway, that’s my story and not yours. I just thought I’d mention that. So you mentioned you did a lot of computer modeling as well as a grad student?
Yeah. Maybe computer modeling is the wrong word. We had a theoretical group that worked with us, and they had developed models for the electronic band structure of these materials. We were trying to see — we thought some of the properties of these materials were related to electron-phonon coupling, so the theoretician that I worked with, Bruce Harmon, he had electronic band gap models for these materials, and we were trying to actually look at the impact of phonon behavior on the electronic structure of the material. So I was running his models. Then also, what we could do based on the phonon properties of the materials is we can make some — again through modeling and through the development of a… The theory and the data that we had, we could come up with what’s called a phonon density of states that based on that you could calculate heat capacities from basically the phonon measurements. So there was a lot of code that I had to write to do those phonon density of states. It was pretty straightforward; it wasn’t new theory or anything like that. I do remember, this was before the days you had your own computers, so I did it all on cards. When we were doing the electronic band structure calculations, what we tried to do is… These materials have a particular band structure associated with the crystalline structure, but when the atoms move because of phonon motion, we argued that it would lower the symmetry of the electronic structure. We wanted to take a look at what happens to the electronic structure as the atoms moved, and what is the energy lower — We were thinking about the HCP-BCC phase transformation and what happens there. Does it happen because the electronic structure changes? By the way, these materials are superconductors, too, although very low temperature. Does the screening change and all that? So we were trying to get a handle on that. By lowering the symmetry in these electronic structures, we wanted to see what happens with the total energy of the system, and see if as you went up in temperature the energy became more and more preferred for the structure to drive it towards a BCC phase. So it was really again kind of in the early phases of this, but I do remember that when we started doing heavy-duty band structure calculations, I’d fly down to Argonne and use their computers, because they had much more capability and I could run the models 20 times in a day. The way we did it is we’d give them the box of cards and they’d run it. You’d come back and then you make your changes on the cards and give them back. So it was not like today where you can submit it from a remote desktop into a mainframe. So in Ames, it was, I’d submit a thing and I’d get the answer back two days later. So when we were doing the heavy-duty stuff, I’d go down to Argonne to be able to get access to their high-speed computing capabilities relative to ours, but we always had to work from one of these (I don't know what they call them now), these computer stations that had access to the mainframes. But there was maybe one on each floor of the physics building, so all the physics students wanted to use this. I remember lots of times, I’d wait until midnight and go in and work until 5:00 or 6:00 in the morning so I could get the models run or do my phonon density state calculations. So I wrote all the code for the phonon density states to do heat capacity calculations. We were able to replicate what the experimental data said from specific heat measurements, or whatever, from the phonon measurements. So that kind of corroborated the data that we were taking. FORTRAN was the code that we used a lot. It was just part of the thesis work at the time. I was not a heavy-duty computer guy.
You just kind of got saddled with those duties?
It was just you had to do it. I had to do it to validate my data and show consistency with the phonon measurements with specific heat measurements. So you had to do it. So you just sat down and you figured out how to do it.
Now at Argonne, was that just some sort of thing where you had to apply for computer time and then when they had it…?
Well again, Argonne was a DOE lab.
Right. Or was it sort of more of a special relationship like you had with Oak Ridge?
Again, Bruce had the relationships with Argonne. Bruce actually was a Northwestern grad, and he did a lot of work at Argonne when he was there as a graduate student. So he made the connections and got me computer time. So I’d fly down in the morning. I had to fly out of Des Moines, the earliest flight out, get to Chicago, and I would sit and run until they would throw me out of the computer lab. But the way you did it is you did your punch cards, you dropped it in, and you gave it to a person at the desk, and they took it in and ran the program. Then you’d wait for the results and it would come out in a big printed sheet. You’d go through it all and “Oh, I made a mistake on that card.” You’d have to make the change, and you’d just keep going like that for as many iterations as you could push through in the time you were there. So this was all part of trying to formulate the theory behind what we were measuring, so that was my exposure to doing the modeling. It was really running all of our existing models for the electronic band structure, but we were modifying it and changing the symmetry properties of the model to see what would happen in terms of the energy — when you’d have these phonons go soft. The electron-phonon interaction plays a large role in superconductivity, at least type-1 superconductors. We were trying to understand the relationship between the electronic structure and the phonon structure to say what would cause that.
So this whole project that you were working on was it sort of up and running then when you joined the group?
Actually, one of the grad students that I — he was actually the post-doc that ended up moving to Oak Ridge — started the work on zirconium. He finished the work on zirconium below the BCC phase transition, so he did it at, I can’t remember how high in temperature it went. We built furnaces that went up to 1200-1300 C to be able to look at these. So he did the HCP zirconium. I came in and picked that work up, and then we started looking for trends between zirconium, titanium, and hafnium, and then started looking at it as phase transition with my work. That work was actually continuing on when I left, but it was very clear the hafnium work had to be done at the HFIR. It was just a progression of experimental work that they had planned over a long period of time to understand better how these electron-phonon interaction, better understand the lattice dynamics of these materials, and why do they go HCP to BCC, that sort of thing. Very basic type of research.
Was there any sort of motivation behind these particular materials? Was there sort of an eye towards possible utilities, or was it just sort of…?
No, no, no. It was really pure academics, and the main reason was to try to understand the electron-phonon interactions in superconductors, and also in classes of materials. You know can we understand the physical properties of the materials in terms of the phonon structure and the electronic structure of the materials? That basically was the thrust behind it.
I guess just one more question that I have in mind, before we move on, is how big was this group? Were you the only person who was working on these particular materials, or were there other people working on those? Were there other graduate students working on other materials who also worked under Stassis?
There were other graduate students working on other materials. I was doing this. I can’t remember Chung’s last name, but he was working on some more novel alloys. The research group consisted of Costas [Stassis] as the senior scientist/experimentalist, and Bruce [Harmon] as a senior theorist. We had maybe two or three post-docs that worked with them, then I would say maybe five to six grad students, of which one or two of them were always theory, and then there were maybe four of us at various stages in our careers that were experimental.
I just have a couple of names from the papers that were on the CV that you sent me. So we have you, Harmon, Stassis. There’s also Wakabayashi?
Nobu Wakabayashi. Nobu was actually at Oak Ridge. He was in the solid-state physics department at Oak Ridge, and we collaborated with him a lot. He ended up going back to Japan. I’m not sure exactly where. But yes, he was an experimental physicist at Oak Ridge.
Okay, and then there’s McMasters.
Dale McMasters. Dale was in the materials department at Ames Lab, a material scientist who was a guru at growing materials. So I worked with Dale in growing these titanium and hafnium. The hafnium especially was an extremely difficult material.
These are all crystals that you’re working with?
These are all crystals that we’re working with, but with neutrons, you need a fairly good size single-crystal material. It’s extremely challenging. The melting points for these materials are really high, and I think hafnium is maybe 1100 C or so. It may be even higher than that. But to grow good single-crystal material is you’ve got to get it up above the melting point, and then depending on the methods that you’re using, you have to be very careful on how you cool the thing down. Generally the way we’d search for single crystals in these materials — You can’t use X-rays on some of these because the X-rays couldn't penetrate. We’d have to actually take the crystals down and we’d put them on a goniometer, which allowed you to rotate in three dimensions, and put it on the neutron beam, and literally we’d spend days looking for a single crystal. You’d have to find a big enough crystal. You’d locate that crystal, and then you’d try to cut it out of — The material may be this big, but the single crystal may be only that size, and it’s in the middle of the thing. So you’d like to isolate it and get its orientation in space so that you could — Doing neutron diffraction where you’re measuring phonon structure, you had to know the orientation of the crystal very well. There was a lot of time spent trying to identify those single crystals. So Dale and I would — he ran the equipment, but I’d spent a lot of time with him actually monitoring the growth. He’d provide me four or five of these ingots, and I’d take them down to Oak Ridge and spend time trying to find single crystals. It was a challenge, especially on a low flux reactor. You don't get a lot, so you’re looking for a significant increase in signal when it goes through a plane, and you’re not sure where that plane is. You have to kind of twist the crystal all around. Of course, you’ve got a neutron beam coming out, which is supposed to be all thermal neutrons and is not supposed to — We probably weren’t the most health-conscious group in the world. Costa — it never bothered him. But we would use the low flux, the Oak Ridge reactor, to get the orientation. When you went to the high flux reactor, you didn't go in front of the beam at all, but again, thermal neutrons shouldn't have any impact. But I had several quarters there where I exceeded my radiation dose limits in trying to find the crystals. But Dale was the materials guy we’d go to and say, “How can we grow this material?” It wasn’t just my stuff. Chung (I wish I could remember Chung’s last name), but he was working on a material called nickel-3 aluminum, which is a much more complicated structure from a phonon point of view. You’ve got many, many more modes of vibration that you’re looking at, and he needed a fairly good size single crystal. You went to Dale to get it grown. He was the metallurgist that had all the tricks. “Here’s the best way to grow titanium.” In the case of titanium, he gave me a couple of crystals that gave me everything I needed for a couple of years. Hafnium, I think we had found one crystal that we could use, and it was a bear. The high flux reactor was about the only way that we could actually do any measurements on it. So at the time, it was a first-of-the-kind measurement because, first of all, it was a very difficult material to grow and second of all, you had to have something like high flux to be able to do it. [Break]
We were just wrapping up with the graduate school experience down at Ames and the transition to Honeywell, which you were telling us a little bit about before. I think you mentioned in your earlier interview that you had known that when you had started graduate school in the 1970s that it was kind of a poor job market but that by 1980, it was all right again?
1980 was great. There were academic jobs. I had two academic jobs offered. I had a couple of offers from oil companies. Again, that was all going to be doing seismology work, more again doing processing of geological information. I remember I had the Honeywell interview, and they had already made an offer, and 3M called me up and I had three job interviews at 3M on three separate days that they were going to offer. They were in different areas, but they almost begged me to come.
Yes, I spoke to the other two people on this trip at 3M, and that was all between 1980 and 1985 or ‘87 or something like that. So they seemed to have been doing a lot of hiring at that time.
Yeah, and everybody that I worked with or was getting out at that time was having opportunities. I only pushed one post-doc position, and that was at NBS [National Bureau of Standards]. If I was going to stay in the research area, that was probably the only way I was going to be able to stay in, and there weren’t a lot of opportunities there for the types of work that I was doing. Oak Ridge didn't have anything. NBS did have a couple of positions that they were looking at. But from an academic point of view and an industry point of view, the jobs, basically it was a really good market. Coming to Minneapolis was kind of an ideal location for our family. My wife is from Moorhead and my family is in Illinois, so we’re kind of in between, and having gone to St. John's, I knew the area well. That also was kind of a plus/minus type of a thing. You know, “Am I ever going to get out of the Midwest?” That did intrigue me, too, was going someplace totally different. But my son was four at the time, and it just was an ideal situation for us. Texaco was down in Houston and…
It has a reputation.
It has a reputation, yeah.
Okay, so you started here in Minneapolis. I guess organizationally, what part of the company were you in? You were in this sort of general research laboratory?
Actually, the lab was situated at 106th and Lyndale in Bloomington, overlooking the Minnesota River. And it was a corporate lab, so there was no business unit that was directing this. It was called the Corporate Technology Center. There are so many names it has gone through in the years that I don't remember what came first. But anyway, it was Corporate Technology Center. And quite frankly, there was very little tie-in to the divisions, very little direct instruction like, “You guys need to do this because we need this product in three years.” It was much more of, “You guys should be looking further out.” For example, we had this division doing mercury cadmium telluride infrared detectors and sensors, and they wanted us to develop new growth methods. Or they wanted us to develop new detector ideas, but there wasn’t anything like “I’ve got to have this in two years.”
There were things for which there wasn’t a clear path forward.
Exactly. And there was a recognition that, “It may not get there, but you guys need to look at that. Explore the landscape.”
So Honeywell, then, at this time is business divisions based upon certain core technologies?
Some probably core technologies, but certainly there was — I don't even remember the business structure at the time. There were underlying products. Some of it was really technology-driven that had come out of the labs, but I don't know if you’d call the divisions really technology-driven. Now at the time, Honeywell had an avionics division. They had a home and buildings, you know the old thermostat on the wall. That’s the home controls and industrial controls, and we also were in computers. Honeywell had a computer division at the time that we ended up divesting back in the early ’80s. We had a joint relationship with Bull of France; it was called Honeywell Bull. The timing on all this I’ve kind of lost. So it was avionics and home and building controls and industrial controls, so pretty diverse businesses. Again, the tie-in to the labs was less than it is now, much, much less than it is now. But we were trying to solve problems, but they were much longer-term. It was kind of like, “Well, you guys go off and work these areas, and if you come up with something interesting, fine.” But we were given license to do that, and that’s a huge difference from today. Today, they have to see a clear path before you even start, which really concerns me from a research point of view [because] it’s hard to predict where things are going to be 15 years from now. To say things are going to be the same and that we shouldn't be working on new stuff to me leaves us exposed and vulnerable to new things in the technology arena. But the structure was is that we had this corporate laboratory that basically got 90% of our money from the company.
Sorry, you have written down here Physical Sciences Center. Is that the —?
It could have been the name. It was Physical Sciences Center sometime in the early ’80s. But we got most of our money from the company and some money from contracts, and it was only maybe two years later that the contracts really started going up. We started relying more and more on contracts. But we were doing a lot of work in solar energy research at the time. We had a huge project in —
You in the laboratory or the company?
The laboratory. We had a group working on solar energy, and there was no obvious business there for us, but we thought this was going to be important for the future. So they were learning how to make amorphous silicon in big sheets, and when Reagan was elected, all the solar energy research in this country got kind of killed. So it was very much an investigation of possibilities, looking at what we thought was going to be important, but there wasn’t a clear tie to present divisions. Some of it was. The Mer-Cad-Tel was for our electro-optics division, but they were growing this using bulk growth methods, and they wanted an epitaxial method, so we were working on epitaxial methods. But it wasn’t like “I’ve got to do this tomorrow.”
What’s the difference between bulk growth and epitaxial?
Bulk growth is basically you throw in some mercury; you throw in some cadmium, and throw in some telluride in a certain ratio into a crucible of some sort. Of course, with mercury, you’ve got to put it under some sort of a pressure environment to protect the mercury from boiling away. But then you heat it up and then you slowly cool it down, and hopefully it will crystallize on some defect or some small crystal and you grow a large piece. They were able to grow ingots of mercury cadmium telluride, maybe about the thickness of my finger and six to eight inches long. Then you’d slice these things into little wafers and you’d build the sensors on top of that. Epitaxy — there are several forms of epitaxy. We were looking at liquid phase epitaxy, but there’s also metal organic where what you do is that you’d have a substrate of some single crystal material. In this particular case it was cadmium telluride, which you could grow bulk bigger than you could mercury cadmium telluride. But you put cadmium telluride and then you’d have a molten liquid of this mixture, but it was a very small amount. You can do it in a much more controlled manner, and you’d pass it over that substrate and it would align with the cadmium telluride, and you’d grow a thin layer of mercury cadmium telluride on top of it. The whole trick was pressure and temperature and chemical composition. Ultimately, they go to a CVD technique, chemical vapor deposition, where instead of passing a liquid over it you pass a gas over it. CVD is very common today in electronics industry, especially the III-Vs and LED technologies. Chemical vapor deposition is a molecular beam epitaxy where you’ve got a substrate, and you basically throw atoms at the surface of it by boiling the material and evaporates, and it comes on to the surface in a very controlled fashion in very high vacuums. We were exploring the liquid phase and the vapor phase epitaxy methods to grow mercury cadmium telluride, because it’s very difficult to grow. Mercury has a high vapor pressure and mercury’s a liquid at room temperature. So how do you get this into lattice in a single crystal form? Because you need the single crystal form. So we were exploring methods all the way through the ’80s on how to grow this material better, because if you’re going to grow wafers that big, you can’t make big arrays. When you think about the arrays in a digital camera now, they’re huge! You’ve got 10 megapixels on there. You’ve got a fairly large piece of silicon. Of course, they’ve shrunk the pixel size down. The pixel size is for these mercury cadmium telluride photoconductors and photodiodes are much bigger. So you’d want like an inch square material, and you can’t get that from here. Because of the bulk methods for growing were just — Again, mercury high pressure, you’ve got to do these in high pressure containers. It’s dangerous — safety issues, all that sort of stuff, so they were looking for new ways of being able to do that. So we explored those to see what we could do. And we eventually developed the LPE so that it was transferred into production out at our electro-optics division. We never got there with the CVD techniques. I think liquid phase epitaxy still is quite used a lot. You can grow cadmium telluride fairly large. You can grow these like in inch-size wafers. They may be bigger now; I’ve lost contact with that. But at the time, you wanted large substrates so you could make large arrays of detectors on it. The larger the arrays and detectors, the more sensitive your infrared system was. So we’d explore that. I was kind of the detector guy, so I’d take the LP material and build the detectors and we’d evaluate what their performance was. The other thing is that once you’re able to grow these things by layers rather than bulk, there are all sorts of different structures you can grow. You can grow different chemical compositions on top of each other and start tailoring the band gap of the materials. So we were looking at all sorts of new detectors that that growth methodology allowed you to do. We came up with a couple of detector designs that we thought would improve the detector performance by this method and were able to validate it in a lab, and wrote a couple of papers on it and have a patent on it. We did have a patent on it; I’m sure it’s expired by now. So the growth methods, the same thing has happened in the silicon industry and the III-V industry: as growth methods improve and you’re able to control growth molecularly or atomically, it gives you the ability to develop new structures that give you better performance or give you different performance.
Would you be looking to patent the process as well as the sensors or whatever that you had come up with?
We patented the LPE process itself. Again, my patent, that particular patent was on the actual structure and putting it into practice. So it was on how to fabricate the material, how to build the device, and using that particular structure to operate the device.
Tell me a little bit about the research team that you kind of had around you at that time. Obviously, you’re a physicist. [Were there] other physicists, chemists, or engineers involved?
You know, at the time, very few engineers. We started bringing in a lot more electrical engineers. I think electrical engineering in the ’70s and the ’80s started expanding into the integrated circuit area and we started getting that expertise in. The people I worked with when I first started there, let’s see. Dave was a physicist. Joe was a physicist.
Who are these people?
Joe Schmit, he just had his undergraduate degree, but he had 30 years of experience in growing this material. Andrew Wood, who actually ended up taking all this uncooled technology — I built the first detector, but he took it and built it into the capability that really showed the power of the technology. He was a physicist. He came from England, and I’m trying to think of whom he worked for before he came to Honeywell. He and I started on the same day, but he was eight or nine years older than I was. So we had a lot of physics background at the time. It was interesting. It was kind of like in a university. We had a couple of theoreticians that were working with us on this. Now they worked on some other stuff, but we had some theoretical physicists at the laboratory that we could go and talk to any time. We had a couple of material scientists that worked not only on this one but on other ones. So they operated more like consultants there.
Yes, I was just about to ask you. It seems like there’s a lot of flexibility between projects and the personnel.
There was a core group of people that were responsible for infrared, and then there was this kind of a fellow — We have a fellow organization now, but we had it then also, that these people would only spend 10% of the time on your project and then they’d go work on some other project. They were there. You know I could go into Tom’s office any time and talk to him about “What does this mean? How does this work?” He’d spend some time figuring it out, if he didn't know at the time. So they were almost like consultants that we could tap into any time to get better understanding. So we had that capability in the laboratory at the time. So there was a set of senior scientists that had their own core projects, but they had plenty of time and flexibility to help us on these projects. We had a growth team that consisted of two scientists and maybe three to four technicians and we had a growth laboratory. This is just for Mer-Cad-Tel. We had a growth laboratory where they would run out experiments on growing. We actually grew that because we were growing the mercury cadmium telluride substrates, too, so we had to grow the cadmium telluride. So there was another scientist that joined the group a little bit after I did that all he focused on was growing cadmium telluride.
Was this a fairly kind of natural transition from the Ames work then with the hafnium and…?
Well, the liquid phase epitaxy is totally different, but the principles are the same. So again, the concepts — you knew the concepts. And because it was so early in the development of this, everybody didn't know —
They were learning on the spot.
You were learning as you went along, so a lot of trial and error. Of course you had to understand the underlying behavior of the materials and that sort of thing. So we had a group of maybe five to seven people doing crystal growth, and then I had myself and I had one to two technicians all the time that would fab devices and test these devices.
What sort of tests would be done at this level of research?
Well, first of all you test for electrical properties to make sure you just have continuity in all that. But we’d actually test them as infrared detectors. So we had a detector setup where we could cool the stuff down and look at the behavior of the photoconductor or the photodiode over time and its infrared response. There were figure of merits for these detectors so we could measure, and we’d take a look at those figures of merit and report back. I can’t even remember what the —
The figures of merit are like standards?
Yeah. They call it, I may end up muddling this, but the D-star (D*?) of a detector was basically a combination of its efficiency as a photo detector. There were theoretical limits to what D-star could be, and you’d measure that D-star. The other thing we measured was its infrared response over a wavelength. So that way, we’d be able to correlate that with the growth. We’d measure uniformity across the layers by looking at the D-star uniformity wave length, or band gap uniformity across the layers, which was always — there’s be variation or whatever. So it was characterizing the growth as well as characterizing the detectors. So for the theoreticians, that would constitute 10-20% of their time, so that group then, between the growth and experiment, fab and test and the theoreticians, we probably had eight to ten people that were working on this over the course of the first three years that I was there.
Was there much contact with other projects that were going on in this laboratory?
No, no. In retrospect, I could kick myself for not getting more involved in what was going on elsewhere. There was a lot of interesting work going on in the lab in artificial intelligence, in media recording, in III-V materials. We’d start work on gallium arsenide, which is III-V, and alloys all on, both from an electronics point of view.
Sorry, when you say III-V? You’ve mentioned it several times, and I thought I’d better ask.
Oh, it’s just if you look at the periodic chart, group III and group V. Silicon is group IV. In III there’s gallium, and V is arsenic. Then there’s also indium. There’s a whole range of materials that they call III-Vs. It’s the basic constituents of the compound that they’re growing. All this stuff was going on, and you know how people give seminars and all that. Once in a while I’d go to it, but I was so focused on just trying to get started that I had enough trouble with my own stuff. But I wish I would have stuck my nose into more and just been aware of things to know what was going on. There was a lot of activity. I mentioned the solar energy research. We had five senior scientists working at it one time. It was a huge project from Solar Energy Research Institute, and it would have just been neat to know what was going on there. But I was kind of oblivious. You come in and I didn't know 90% of the people in the laboratory. I went up to my office and I had my little lab outside. When I was doing extrinsic silicon, I had a technician…
Overall, how many people are in the laboratory?
This is researchers and technicians, the whole lot of them?
Researchers and technicians. We had a lot of technicians. They do all the grunt work. I don't want to say grunt work, but you know they were in the laboratories taking the directions from the scientists. I don't know. Maybe it was 50/50 or 60/40 technicians to scientists. Some scientists had three or four techs working for them, more the senior ones when they’ve got their programs established. Sometimes I had a tough time keeping one technician busy. Later on, I could keep a whole group busy as you expand your understanding and knowledge and the directions you want to go in than what the program calls for.
So then presumably you have a project leader in this Mer-Cad-Tel and infrared?
Yeah. The way the group was organized — and again, you’re relying on my memory here. But we had technical sections, and there was an infrared group that was led by a section manager who would manage, basically interface with customers and manage the money and manage the workloads and manage the different projects. Then each project had a project leader. So I was a project leader for the extrinsic silicon program.
From the get-go, or did you become that?
I became that when I walked in the door. They gave that project to me. That was actually a contract from the Air Force, so I had a $150,000 contract from the Air Force to investigate the extrinsic silicon. I worked with my section manager, but then I’d also work with the Air Force contract manager on progress on the program and what I was doing, what the plans were, give reviews to them, that sort of thing. The section manager was always there kind of mentoring what we should be doing but gave me a lot of freedom in how to approach the problem.
What was the name of the section manager?
I can see his face — Walter Scott! Sir Walter Scott. Walter Scott was the manager, and there was a department manager above Walter, so Walter was maybe one of three or four section managers that were in like the electronics department. Olbert Tufty[?] was the department manager. He was a St. Olaf grad. Again, both these guys had their PhDs in physics. So we had an electronics group, and the other groups were doing things like III-Vs. The III-Vs was very, very early on research that they were doing. Then Andrew Wood led the liquid phase epitaxy project, and he had two or three technicians working for him on that. So I led the device fab part of it after the first year. I was just doing extrinsic silicon for the first year. But then as they started developing this LPE, they started needing devices made and I took over making the devices to evaluate the LPE material. That required some new fab processes, so I had a fab tech that worked with me on that. So it was kind of a hierarchical structure. You had the lab; then you had these department managers around certain areas, and I’ll be honest with you. I cannot remember all the departments. There was an electronics department, but I’m sure there was something around software and computers. There was an RF group, and I don't know if it was in our electronics department or if it was somewhere else. So that was the structure of it. So we had a section under Walter probably of 12-15 people, and we met on a monthly basis and talked about the normal things in a section meeting. That was how information flowed down from the rest of the laboratories, and we’d talk about problems and that sort of thing.
So you didn't feel too much influence then from even higher up?
I didn't. I was pretty isolated from them. I never really got involved with the higher-ups until I actually took one of the section manager’s jobs in I think ’85. That’s the first management job I took.
At what level would it interface with the rest of the company? Would that have to be all the way up at the top, or would the department manager…?
I’d say department manager, a little bit at the section manager. Walter would interface a lot with the opto-electronics division.
So if you were going to see if some of your work was of interest to some of the technology people in the divisions…
You’d start with Walter. You’d start with the section manager —
And you would go and speak with them.
Yes. They provided the contacts. They were the primary point of communication I think for a specific — Mer-Cad-Tel was all OED. There wasn’t anybody else in the company you had to worry about, so Walter took that on. Olbert spent a lot of time not only with the divisions but also with corporate justifying our expenditures. Corporate didn't spend — at least it didn't seem at that time — they didn't question a lot about what we were doing. They may come through once a year and take a tour of the lab to see that things were going okay, and the department managers and the section managers would do those tours. They were always able to give a good story of why we were doing what we were doing.
Should we discuss anything else then before we move on to gallium arsenide?
I guess the only other thing from my perspective that was kind of interesting from the Mer-Cad-Tel days was that molecular beam epitaxy at the time — People started looking at — DARPA had funded MBE, and they actually brought a guy over from France to the University of Illinois at Chicago and built him a laboratory to look at MBE of this material. I always questioned whether it made any sense. It wasn’t MBE of material. They were trying to grow multiple layers of different compositions of this material. I wasn’t involved in it at the start. Joe Schmit was, actually. He was doing some of the materials growth, and he got some of the material from Jean-Pierre Faurie, I think his name was. I may have the middle name wrong. But he got some of the material and Joe started looking at this using X-rays, and he saw some really weird structure. He started to come up with some crazy ideas why the structure was there. Well, again from my neutron diffraction days, I said, “No. No, no. That’s perfectly explainable by having multiple layers; you’ve got a super lattice here.” Joe said I was nuts. I went and did the calculations and showed, “You’ve got a super lattice here.” The X-ray scattering calculations are rather sophisticated. They involved Bessel functions and all that sort of stuff, which I’ve totally forgotten. But I wrote a paper up on it, and said, “I can explain all of this by this.” The response was, “Well, maybe you ought to work with Faurie.” So I actually started working with Faurie, getting material from him. From my neutron scattering days, one of the post-docs down there now was a junior scientist at Ames Laboratory in X-ray scattering, and I actually started a collaborative program with Iowa State for a couple of years where we’d take this material down. You needed a fairly good X-ray scattering capability to do this, so what we started doing was looking at this material. My concern was the stability of this material, because again, mercury’s got a high vapor pressure, and is it going to stay still, especially when you heat these detectors up? So we started looking at inner diffusion of these layers over time using X-ray diffraction, which was fairly unique. It wasn’t used that much. Some people had done some work in gallium arsenide at Bell Labs on this. Another interesting thing that came out of that was Faurie and I published a paper in ’84 or ’85 on inner diffusion, and we actually did several different papers on inner diffusion in these materials that raised a lot of questions about whether these super lattices in Mer-Cad-Tel would ever work. But it was an interesting physics exercise, I guess. So anyway, I was actually able to use some of the work that I had done in graduate school, knowing about scattering methodology and that sort of thing, which I was actually able to bring back and apply to clarify some results that we had gotten out of this new material.
So you started out with this Air Force contract, and you mentioned that it moves more towards contracts as you go along. Was this kind of during this initial period that you were there that you acquired more contracts?
It slowly evolved. When I first got there, really all there was were individual scientists or the section managers, and if they saw an opportunity, they’d go out and try to get a contract on it. It was decided really maybe in the first couple of years that I was there that we needed to establish a formal marketing arm within the organization to get more money in the research —
Within the laboratory?
Within the laboratory. It was clear that corporate wasn’t going to be giving us much more money, and if we ever wanted to increase our research efforts — There was also a feeling that by going out after contracts, our ideas would be more competitive. Having to compete on ideas would be a good thing to do. So there was a goal to increase our contract base. Not decrease the R&D base, but increase our contract base over time. It started with like one or two guys being in marketing and we’d just market to Air Force research laboratories in Dayton and DARPA. I don't know if there was anybody else. But they started chasing after these programs and making people aware of the research that we were doing at the laboratory.
Dayton — that’s Wright-Patterson?
Wright-Patterson, yeah. I spent a lot of time at Wright-Patterson on projects that we had with them. When I ran the group for about ten years, they funded a lot of research that my group did. We were very successful in winning contracts with them. So it allowed us to push research a lot faster and explore areas that we just couldn't do otherwise. So that was a slow transition during the ’80s, and I’d say by the end of the ’80s we were 50/50 in terms of contract and Honeywell money. I don't think really that the total dollars decreased from Honeywell during that time. Maybe in real dollar terms there was a little bit of a decrease, but I don't think it —
So it was mostly just a growing overall budget.
It was a growing budget. The infrared group didn't grow that much, but we grew a lot in the area of gallium arsenide during that period of time. One of the guys I worked with in the infrared group went over to the gallium arsenide group in ’83, and that’s just when they were starting to really do a lot of work in — Silicon was the bread and butter in the IC industry, but at the time, people were always arguing that there are limitations to what silicon can do. So we should be looking at gallium arsenide and related compounds. The other thing is gallium arsenide is an opto-electronic material. Its optical properties are much more beneficial than silicon. Silicon’s got an indirect band gap, so you can’t really do optical detection with it. There was this whole idea of merging optics and electronics together. So they had started doing some preliminary work in gallium arsenide in the early ’80s.
This is a nice coincidence, actually, because one of the former heads of AIP who I interviewed last year after he had retired, doing another one of these life histories, but it wasn’t explicitly part of this project. But he had run the gallium arsenide lab at IBM.
What was his name?
Really? The name definitely rings a bell from papers.
Yes, he had been doing amorphous silicon, so the fact that I have any knowledge of any of the things that you’re talking about is kind of a result of my talking with him.
Well, back in the early ’80s there was a belief that silicon was going to run out of gas. Ignore the pun. Gallium arsenide is called GAs. But the people thought it was going to run out, and it was based on what they thought limitations of optical lithography were going to be and understanding of silicon, and of course we ain’t there yet. But gallium arsenide does have a role, but we were in a computer business. So that was one area, and then also, gallium arsenide, some of the III-V compounds you can make from it, is also a very good analog material, so making not just digital electronics but analog electronics like for RF. We had an RF group that was very interested in this. Saw a lot of applications within Honeywell for this, although again in retrospect, we were a computer company, but we weren’t an IC company. That was a lot of the motivation behind the gallium arsenide is that, well; you can integrate optics and electronics. You can do digital electronics on it. You can do analog electronics on it, and the electron mobility in gallium arsenide is much higher than it is in silicon, so you could get to higher switching speeds. So naturally, you can get faster ICs out of this thing. That was one of the primary reasons why we started pushing it, and we were kind of at the forefront. IBM was another one. Bell Labs was another one, all doing work in this area, and DARPA jumped in. DARPA was very interested in pushing this, so they pumped a lot of money in the 1980s and we were one of the receivers of it. Air Force Dayton pumped a lot of money into this. So then Nick, the guy I mentioned that was in our group, went over and became a scientist in that group. Within a couple of years, he was building —
Nick Cirillo is his name. He’s in some of the papers on there. Nick joined the labs about six months after I did. He actually came from Naval Research Laboratories out of Washington. He, within a couple of years, had been successful in building what’s called modulation doped FETs out of gallium arsenide, aluminum gallium arsenide, and that required molecular beam epitaxy or CVD epitaxy to grow very thin layers and different compositions. You could take advantage of those different compositions to confine the electrons and get very, very high mobility. Honeywell, for the next six or seven years in this modulation doped FET area, was kind of at the forefront. We were building circuits that operated faster than any circuits ever built. So Nick went in and started this, and within two years he was running the section that was doing this work in the modulation doped FETs. His position that he left as a scientist opened up, and he suggested that I come on down, that my experience in fab, my experience in growth, and device testing and all that —
This is still the same facility in Bloomington?
Same facility. So I decided that this would be fun to do. It’s a growing group, a lot of interesting stuff, and the group was really growing. We had an analog group, we had a digital group, and both were having a lot of success. So I joined the group in ’85 and basically took over the modulation doped FET work. Again, the whole point of this was trying to build better FETs and trying to build larger arrays of these integrated circuits. So it was process technology, it was device physics, and we had a materials group within the digital electronics group that grew all this material. It was working with them to try to grow different structures. We were looking at super lattices in this material, so I had some work done in super lattices before. So all of this stuff kind of built on what I’d already been doing. So in ’85, I moved to that group, and in ’86, Nick moved up and took over the department chair, which included digital and RF, and I can’t remember what else was under there, all the opto-electronics stuff. So I ended up taking over the digital electronics group at that time. This was a time when the group was growing very rapidly. We were being very successful in winning programs, because every six months we would come out with a new world record for speed of a transistor or level of integration that we accomplished with this stuff. Again, for about a year and a half, I was assigned as a principal scientist working in this area trying to improve the processes, trying to understand the processes. There were some characteristics of gallium arsenide that you had to modify your processes so you didn't mess the devices up. I mentioned about the Mer-Cad-Tel inner diffusion and these super lattices. If you heated these materials up too high, the interfaces would diffuse out and you’d lose the electronic properties you wanted. So you had to develop methods to avoid this inner diffusion. Now typically, the circuits are going to run at less than 200C, but you’re processing these things at much higher temperatures. So we developed rapid thermal optical annealing techniques. People had used it before, but we applied it to gallium arsenide. So it was learning all of what happens in the materials and all that sort of stuff. So the next year and a half I spent developing rapid thermal optical annealing for this, what the impact was on the materials that we had on these devices. We continued to push the technology, and also wrote a couple of proposals to Air Force that we won on using super lattice structures to improve the electronic performance of these devices. We also during that time developed a — which was at the time kind of revolutionary in gallium arsenide — we developed a P-type and an N-type FET that you could put on the same wafer, so it was going to be similar to CMOS. You can’t grow a really good oxide on gallium arsenide, so you had to use some sort of an insulator, but we were able to come up with an analog to a CMOS process on gallium arsenide. As a result, again we got funding from — DARPA gave us funding and the Air Force gave us funding. So after about a year and a half, Nick moved up and he asked me to take that section over. So I ran that section for the next two and a half or three years (I can’t remember) where the focus was on proving out gallium arsenide as a digital electronic source. In retrospect, there’s very little I think done today in digital in gallium arsenide because silicon just kept going. It was just a moving target the whole time. Now in analog, gallium arsenide and the III-V compounds, you’ll find it in every cell phone and the communications area, and there are a lot of opto-electronics devices out there with III-Vs now. We had a group at the lab that worked on developing lasers out of III-V materials that we were doing integration of electronics with. So that’s where that technology’s gone. Digital electronics, though, the group that I had, we were very successful in levels of integration and very successful in meeting all these world records, but it was a technology that was trying to run against silicon, which was really difficult. So in retrospect, it was probably the wrong — The opto-electronics or the RF area would have been the areas to be in if you’re going to be in the III-V compounds. But for the three years that I managed that group, we continued to get a lot of support from the DARPAs of the world, because again, they were pushing for larger integration and higher speed. We had a lot of classified programs because again, they were trying to do very high-speed signal processing and they thought this was going to be the way to do it.
Do you have clearances or anything all throughout this period?
Through that period, yeah.
Through the gallium arsenide period or since you started?
I had clearances starting with the Mer-Cad-Tel. In fact, yeah, I picked up my first clearance with the Mer-Cad-Tel work because there was some of the work that we weren’t even allowed to publish that we developed. The Army stopped us from it, which I never quite understood. I picked up the first clearance there and ended up getting secret clearances. DOD has certain clearances; other agencies have different clearances, so I had them with both of them. But the primary interest in all of this was just high-speed signal processing, and their fear was that they weren’t going to be able to get it with silicon ultimately. There may be certain cases where there may be an advantage, but again, I’ve kind of lost contact. Seymour Cray, I don't know if you remember Cray Research here in town, but he tried to build a computer out of gallium arsenide because of the very fact of the high speed. It was a tough job, because you were trying to build a computer at the same time you were trying to mature the underlying technology, where with silicon, that’s not the case. Silicon, you take a mature generation of electronics and you build your computer and the next generation comes along and you build another computer. He was trying to do it in parallel. Of course, Cray got caught up in all the changes in the computer technology in the 1980s, too, but he tried for a long time. They did a lot of interesting work. But in the end, gallium arsenide in the digital world just couldn't compete with silicon because silicon just kept shrinking and shrinking and shrinking, which offset anything we were doing with gallium arsenide.
Let me just ask you. You mentioned publishing restrictions on that and so forth. Especially with the gallium arsenide, I’m interested in terms of conferences, publishing, and patents. You were only in it for a couple of years, I guess, but what’s that like? There are papers that are published on the subject, but obviously, that’s after you’ve secured your intellectual property rights on it. I’m wondering in terms of competition just kind of how that works.
Well, we published quite a bit in the gallium arsenide area. We had underlying patents on the basic devices, so lots of times we’d publish if we got increased in speed or if we set a world record. At the time, there was a big race between Bell Labs and ourselves, and there were a couple of other companies that were doing work in this area. So any time we came up with a new device structure, generally speaking, we wouldn't tell them how to build the device in the paper. It was always kind of a fine line of how much information we released in the paper. But we were encouraged to publish. We gave papers at a lot of conferences during those years. The intellectual property — that was always our first consideration. “Can we publish this based on where we stand? Do we reveal anything?” If we felt like there wasn’t anything being revealed or that we had the IP in place, we were encouraged to publish, in fact. We just never released the details of how we built the device because that was a key figure. Lots of times, we wouldn't give any process recipes away, and we’d only give a notional idea of what the device structure looked like. So it was a modulation doped FET, but we wouldn't go much beyond saying that it was aluminum gallium arsenide, gallium arsenide, but we wouldn't give the composition. I’d have to go back and look.
That’s kind of the impression that I’ve been getting, speaking with the people at 3M as well, is that when you're in industrial research, there is a tendency to publish, to go to conferences and what not, and then to say there what you’ve accomplished to give kind of the general idea like that. But then there is a certain amount that you withhold and occasionally there will be an academic there who will ask a question.
Oh, it wasn’t occasionally; it was all the time. And the Japanese were — Like NEC at the time was the big foreign competitor, if you will, in the research area, and NEC would get up there and they’d do the same thing. It was interesting at some of the gallium arsenide research conferences, they’d get up and give this paper and then someone would stand up and ask, “Well, what were the thicknesses of these layers?” or “I don't understand how this could work because the composition of that alloy doesn't seem to be right.” And all of a sudden, they didn't understand English. [Laughter] That happened a lot.
Yeah. Sometimes they would be shouting, I’ve heard.
Oh yeah, yeah. They did it to us, too. “Well, I can’t tell you. It’s a process detail, and that’s undergoing intellectual property stuff now and we can’t tell you,” and people would be upset, and that wasn’t just the academics. They felt that if you’re publishing a paper, you should have this information in it. So we were always going up against the limits of what you could do, and the same thing with publishing. You’d get comments back from the reviewers saying, “You need to say more about how you built this device,” and we’d send back and say, “We’re not telling any details about the process.” Because the paper isn’t about the process; the paper is about the device. All the magic is in building the device. Once you build the device, it’s obvious how it’s going to work.
I have it here in the CV that you sent me, but as far as where you had looked to publish, you’ve got Physical Review.
That’s more in grad school.
Oh, that’s more grad school. Okay, sorry. Journal of Applied Physics, Electron Device Letters.
You’ll see a significant shift going from physics journals to what I call IEEE journals, because that’s where most of the work was published. Unless it was seminal, then it would maybe end up in —
Journal of Crystal Growth.
— Journal of Applied Physics or Physical Review.
Yeah, I do see that transition.
Yes. Much more of the industrial stuff, especially electronics.
Yes, a lot of Electronic Device Letters, Electronic Letters. That makes sense. And then as far as conferences, that would be IEEE presumably.
Yeah. There was a lot of IEEE sponsored. I went to the AIP meetings, the March meeting for a couple of years after I was here. But then it was hard getting management to approve it because it was so general.
Oh yeah. American Physical Society.
Yeah, but again, that was probably on graduate work.
Well no, this is ’86. But yeah, there are others as well.
I’ve probably been to three or four of the American Physical Society meetings in the last 20 years. Gallium arsenide during the 1980s and 1990s, IEEE sponsored a gallium arsenide integrated circuit IC symposium every year. There was also a Materials Research Society and the IEEE sponsored an electron devices and materials meeting every year, It was always at Santa Barbara. Device Research Conference. I guess it wasn’t just always there [Santa Barbara]. So it depended on the topic area. When I switched over to doing MEMS in the 1990s, IEEE sponsored a biannual conference at Hilton Head that was probably the premier MEMS conference in the world. So you just go depending on what the topics were. Again, it may be IEEE-sponsored, but the people who are working in these areas in the early days, like the MEMS area or the gallium arsenide area, a lot of physicists were involved, and it seems like as the field grew and the applications grew, that’s when they started getting the electrical engineering groups more involved in it.
Okay. Should we talk more about gallium arsenide, or should we just move along?
It’s up to you.
I don't have anything else immediately, I suppose. So the next thing I have on your resume here is Acting Section Head of the Process Laboratory.
Yeah, I did that concurrently with the digital electronics. So I ran both sections for 12 or 18 months. I can’t remember.
Is this a bit more of an interaction with the production processes, or is it just kind of…?
No, no. By the second year I was running the group. Again, if you look at what’s happening in electronics, what’s happening in Mer-Cad-Tel, what’s happening in III-Vs, what’s happening in silicon, what’s happening in MEMS, everything is becoming integrated-circuit-like-based using processes that were percolating up from the integrated circuit industry. So we had to have a facility that allowed us to do that, and we had kind of a makeshift facility in the lab when I came. Basically, they built a clean room inside the building, and it wasn’t a very good clean room and it didn't have the space to be able to do everything that we wanted to do. So when I started my work in infrared detectors was growing. We were building more devices. The work in gallium arsenide was growing. The work in silicon MEMS was growing, these mass air flow sensors, and it evolved to doing infrared bolometers, where we were building very large arrays. We needed a much better capability, so in ’85, we actually built a wing on to the building with a new fab facility, class 10 IC processing facility. State of the art. When that facility came up and running, the guy who was responsible for running it left, and again, I had IC processing experience and I was running the digital electronics group, which was one of the big groups using the laboratory. Basically, I ran a process lab, so I oversaw the operations of the laboratory. So it involved capital equipment, involved all the technicians. There were probably 20 people in the laboratory that were running equipment or responsible for maintenance and repair. We had about 10,000 square feet of processing facility. So we were supposed to hire a person to be able to run the laboratory, but we didn't have anybody. So I ended up running the lab for that period of time there at the same time I was running the digital electronics group. So I basically had responsibility to ensure that the lab was up and running and available and that we had the right equipment for everybody else that was using the laboratory. So I was the punching bag when people couldn't get things through the lab and that sort of thing.
Okay, I see how it goes. Then up there [later on the resume], you have gallium arsenide again. I guess that would lead up to 1989 and Business Development Manager. That seems like it’s going to be a bit of a shift of gears here.
Oh, it was a big shift. It was probably just a respite from what I was doing, but those three years where I ran that group were pretty intense. What I had demonstrated I think while I was running that group was the ability to work with outside customers and bring money in. Again, I mentioned before the laboratory had shifted. We had grown this marketing group, business development. Marketing is probably not the right word — business development group to bring in outside money. It was very clear from the people that we had in those jobs that you couldn't just have somebody that was in marketing; you had to have someone with technical backgrounds. At the time, I was spending so much time with DARPA and the Air Force research labs and all that, that I was asked by the director of the business development to come over, and, “Why don't you run this for this side of the house, the whole gallium arsenide area?” So I started there, but at the same time, the labs reorganized. They reorganized along basically — It’s hard to describe, but there were some other research groups within the company. They decided to align along a military and commercial alignment, and of course most of the gallium arsenide stuff by that time was military or security, that sort of thing. The commercial side, there wasn’t any gallium arsenide, but there was a lot of sensor work, and also on the sensor side was all this MEMS work that was going on. And for some reason — I don't know why — but the decision at the time was just to put all sensors under that group, even if it was a military sensor. So basically the lab split in two, military side/commercial side, except the commercial side also had the sensors in it for the military side. I ended up going and taking over that sensor part, so they asked me to develop some business in that particular area. Well I mentioned this uncooled bolometer stuff to you. That was a military application, but it was a sensor, so that was on my side. So I had the bolometers plus the mass air flow sensors and what we were trying to do in that particular area.
I guess I’d better ask you at this point exactly what a bolometer is.
All it is is an infrared detector that relies on temperature change to measure infrared. So you’ve got this device, and you’ve got infrared radiation going on and it raises its temperature. That sounds simple, but it’s very difficult to build because how do you separate out infrared from all the other temperatures that are going around? So you have to build a device that’s thermally isolated from the environment and all it sees is infrared radiation. There are ways of doing that that we came up with — having the right material that absorbs infrared radiation over a certain wavelength. So I started doing business development in that area, and we had a silicon group that was doing this micro machining in silicon. They had determined that this was going to be a good growth area, and Honeywell is very big in sensors and has been. People don't know how big we are in sensors. It’s huge, a huge business. So I started working with the sensor group in this particular area, and we came up with some device concepts that would, I don't want to say revolutionize, but advance pressure sensor technology, that would advance accelerometer technology, and these are all types of sensors that Honeywell needs in their businesses. My relationships with the Air Force and with DARPA were such that I had doors into these places, so for the next couple of years, we were writing proposals. In fact, I ended up writing one of the proposals with one of the chief scientists in the group, probably the one I spent the most time on ever, and we won a program that led to a whole development effort for the next eight or nine years in the area of resonant sensors. We still do a little bit of work in that area, but the underlying fabrication technologies that we developed during that time are critical now to parts of our business. So as a business development manager, I kind of oversaw that. Well, business development — you can only do so much technology. There was a lot of paper work and a lot of glad-handing and that sort of stuff that went with it. I kind of had to wear a suit. [Laughs] So that went on for a couple of years, and I was just integrally involved with that particular group. They were doing some really good stuff. The uncooled bolometer stuff was just growing rapidly. We were getting contracts all over the place. We were actually doing the inverse of a bolometer. We were taking an array of these things, maybe 250,000 pixels, and using it as an infrared TV. So each pixel we’d change the temperature and then you create an infrared image. You say, “Well, where can those be used?” Well, they were used for calibrating infrared systems, so we developed a whole technology behind that that got funding for eight or nine years, and we were actually building it for the Air Force and the Army to calibrate all other infrared instruments. So we had a lot of success there. Well, a year and a half in or two years in, the section manager for the silicon MEMS group left; he went to Medtronic. I saw it as a way of getting back into technology, but into sensors. So I applied for the job, was actually asked to take it over, so I took that job in ’91. So I went back into technical management then, running the microstructure technologies, which is another way of saying MEMS. Microelectromechanical systems is really actually making moving parts in silicon, which we were able to do with this resonant device. So I took that over in ’91 and it was a growing group, a group of 15-20 people.
Sorry. Just on the business development, was that then the all-military, or was it commercial customers?
No, commercial, too. Commercial. We were going down…[laughs] I mentioned this mass air flow sensor. One of the scientists had developed an approach using several of these mass airflow sensors that he could measure the composition of natural gas with it. So we used it as a composition sensor. So we were talking to the Gas Research Institute. He’s making an electronic gas meter.
Sorry. Who are they, the Gas Research Institute?
They are a consortium. They are a research group funded by the natural gas producers in the United States. Probably if you were in west Bloomington [where Thomas grew up], you were heated by natural gas in your house and I’m sure you had a gas meter outside your house, which was a bellows type of gas meter and it would measure how much cubic feet of natural gas. Well, what we developed is a concept where you don't measure the cubic feet of natural gas; you measure the thermal content of it because it varies. If you buy natural gas from Texas, the thermal content is different than Canada, and of course you rely on the gas company to tell you what the BTU content is. Usually what they do is average it and you get billed for it. Well, if you’re an industrial plant consuming $100,000 worth of natural gas every month, you’d like to know am I really getting my money’s worth? So we were actually going to Gas Research Institute and we actually went overseas to some of the major gas meter manufacturers to try to convince them that they should take a look at that. None of them wanted to do it. The gas companies didn't want to do it because all of a sudden you’d have people questioning their bills. Now you’re going to measure by BTU content. So we went through that. I spent a couple of years during that time visiting every meter manufacturer in the world trying to talk to them about where they were going, and these are places where they bend pipe. We were talking about doing this all electronically, and we had demonstrated it in the laboratory; we had built this. So we went to the commercial side, too, trying to build cases for that. Again, we were not that successful with trying to make that transition, but we were technology pushing and we couldn't convince people the value proposition was there. I still think it’s a great idea, especially for when you’re trying to control your energy usage. But at the time, it just didn't fly. But we went on the commercial — Gas Research Institute — the electrical industry has one called EPRI, Electrical Power Research Institute, that’s in Palo Alto. So we were always involved a lot in home and building controls for energy conservation and that sort of thing. Those two the labs themselves got a lot of money from, generally not in sensors, though. It was more in systems applications, that sort of thing. So we did commercial as well as military.
Was it domestic or international as well?
In that case, we were doing international, and we do international now with the labs. We’ve got a group, a couple of people over in Europe, because we’re an international company. We have research centers in Czechoslovakia and in India. The Czechoslovakian one, because we’ve got a location over there, we can go after European Union funds, and it has to usually be spent in the European Union, which is in Czech Republic. I guess I shouldn't say Czechoslovakia anymore. So we do that now. At the time, we were just doing the marketing commercially. There were a lot of opportunities through international agencies, for example, especially in Eastern Europe when that opened up. So we take our technology and see where we can take it.
Is this kind of the real first interaction with the business side of the company, or had that been a feature of the gallium arsenide work as well?
There was a little bit with the gallium arsenide. We had users out there that wanted the stuff, but they would use what was available and they’d say, “If you get there, fine. Here’s what we need.” But that was it. They would never pull. There was not a lot of pull. On the business development side, we started looking very closely at how this would apply and where could we use this. It wasn’t just me. That kind of coincided with the nature of the research lab changing. We’ve got to become more tied into the business.
So this is also happening in this period, that you definitely have the shift away.
Yeah. And it certainly in the ’90s happened. Every year the money got tighter, and the rationale got more extreme. So I would say by the end of the ’90s, if you couldn't rationalize that this is going to impact the business in some manner, you were not going to get Honeywell funding.
In certain terms, yeah.
We could still go after outside contracts that weren’t clear. We’ve even got tougher on that in the last seven or eight years that we don't want to pursue contracts that we don't see tied in to the product line roadmaps, which I think is a mistake, just from the very nature of how long it takes technology to mature and the threats from new technologies coming in and how long it takes for them to mature. I think we’ve probably gone a little bit too far in that, but that may just be unique to Honeywell. I don't think it’s unique; I think a lot of places are like that. So we did both commercial and military business development. We always have. Quite frankly, especially during the ’90s, military dominated. You had a few areas you could go commercially. I actually had a couple of programs we negotiated directly with companies. We got money directly from companies to do research for them.
But that’s the only way that you can really do more long-term research is by procuring outside funding for it after they started clamping down in the ’90s.
As they started clamping down, it’s become more and more difficult to do the long-term stuff. The tendency I would say — I mean you can’t say absolutely there is some stuff that is, but if you look at the distribution, the distribution has gotten much narrower.
And just on a detailed point, was this a new facility then when you went into business development?
No. It was the same — I was at that facility till ’95, from ’80 to ’95. The name changed about six times, but the underlying group of people there and research being done there stayed the same. Our roles and responsibilities changed over that time, but…
Okay, we’ve got about an hour left, so we’d probably better head on to MEMS. So I think I interrupted you as you were starting to launch into what that was all about.
Well, again, I mentioned the uncooled bolometer and the mass air flow sensor were MEMS devices. Again, they were using the structural part of silicon to build the devices, and when I had the business development job, we got a project where we were actually building moving parts in the silicon. We were building basically — I don’t know if you’re familiar, like your watch has probably got a quartz oscillator in it for time keeping. It’s a little oscillator that vibrates back and forth at a certain frequency. We actually had a project where we were building those oscillators in silicon, and you had to build it inside a vacuum so that the air didn't damp it. Then we were using these oscillators to measure stress in the silicon, and if it was a pressure sensor, you had to build a diaphragm. You built this oscillator on the diaphragm and you can measure pressure with it or you can measure acceleration with it. So it required very sophisticated micromachining of the silicon. This area kind of evolved during the late ’70s and early ’80s, and we were kind of at the forefront of it. We commercialized a mass airflow sensor in ’87. One of our divisions builds those now. So there was a lot of interest in pursuing this for other sensors, so when I took that group over, we continued work on these resonant sensors and built up a fairly good size capability in that area because we thought it would maybe be the next generation of pressure sensors, next generation accelerometers, and in some respects it has, although it’s not in the form that we ended up developing. DARPA at that time realized the importance of MEMS and started a major effort in microelectromechanical systems. This was in ’92 maybe is when they started. They had a little bit going on before that, but we were successful in winning several programs over the course of — We still are, we’re still involved with them. But we developed this micromachining capability in silicon as a result over the next seven or eight years. I ran the group; I had a group of maybe 20 people, some very good top-notch scientists. These are becoming more the electrical engineers. They’re coming out of school with the expertise, but I had several physicists on the team, too. The whole idea was to understand the physics of these devices and know how to build them and get the performance out of them. So I had a fellow that supported my group, Dave Zook. He was actually I think interviewed by —
Yeah, the name seems familiar to me.
Yeah. Brilliant guy. A physicist. He basically supported me full-time for those seven or eight years in terms of doing these devices and exploring new device concepts. That’s when we really started working close with the divisions, trying to understand what their needs were, what kind of performance we needed to have, what the requirements were to transition technology into product, and that sort of thing. So that went on through ’98. Well, I’d run that group until 2002. Interestingly, in ’98 we had been pushing one of our divisions. We’ve got a division here in town that builds navigation systems for aircraft, so 99.9% probability that you flew out on an aircraft that’s guided by a Honeywell navigation system.
Is that over in St. Paul?
It’s in Minneapolis, over in Industrial Blvd.
I have a childhood friend who works for Honeywell, and I think he might work in that division.
It’s over what’s called Ridgeway. It’s on 35W just north of —
Is that where they had that chimney?
Oh. Yeah. I always remember that.
Yes. That’s the facility, or one of the facilities, and they build ring laser gyros there. The ring laser gyro is a gyroscope, and you need a gyroscope to be able to navigate with and to measure the orientation space, you need accelerometers with that, but a navigation system has gyros and accelerometers, and basically through integration it can tell you where you’ve gone over a period of time. So a really good gyro, a real good navigation system, if I take off from here, you can navigate totally with that system for a particular period of time, and four hours later, you’re flying above the clouds, you know exactly where you’re at. That’s a simplistic description. That’s the ideal system. Now there are errors that you’ve got to account for, and over a period of time the errors grow, but the ring laser gyro has been kind of the key sensor in a navigation system. Honeywell came in with this. They started developing it in the mid-’60s after the laser was developed, but it was 1980 before we were actually getting it into a product in the Boeing 757, 767, and the business just exploded. Well, they’re pretty expensive to build — they’re pretty sophisticated. We were looking at can you build a micromechanical gyro out of silicon. We won’t go into the details, but there are ways of being able to do that. Now everybody kind of poo-pooed the idea in the late ’80s when we first started talking about this, and we always wanted to do it. We thought you could do it, but performance was always an issue. Some people had built some. Draper Labs in Boston, Cambridge, had built one that was six orders of magnitude or eight orders of magnitude off of where we needed to be, but they showed you could build it.
Orders of magnitude of accuracy?
Accuracy, yeah, and stability. People always say, “Well, you’ll never get there.” Well, that’s the same thing they said back when they built the first transistor. “So what?” Then all of a sudden technology evolves. We did several studies back in the mid-’90s saying that we should be looking at this, and maybe not for really high performance, but for maybe areas where you don't need to do guidance for more than a couple of minutes, like a bomb, that you could apply this to. We had the capabilities in the lab to do it, certainly for building the device. The physics device is pretty sophisticated and the electronics are really challenging, but we were one of the people that actually had the capability of doing that. Well, about 1996, the division started getting nervous about this, and we started taking a look at what could be done. The problem was is that there are only a couple of ways of building these things, and there are some key patents out on it that we weren’t quite sure how we’d get around it.
Who had those?
Key patents were by Draper. There was one other one by — For our discussion, Draper was the key ones. It turned out in 1998 Draper had licensed the technology to Rockwell. Rockwell had an automotive business at the time, and they were looking at these things for automotive applications, using gyros in automotive. They’re used today. Gyros are used for anti-skid and whatever. So they were trying to build a business around automotive, Rockwell was, and Draper licensed them this technology. Well, we weren’t interested in automotive, although in retrospect we probably should have been because we had a commercial division that actually sells a lot of automotive sensors. But we couldn't get anybody interested in the gyro, until these guys became very nervous. Their customers are saying, “You’d better be looking at MEMS, because there’s a whole application area here, a tactical application.” Not aircraft navigating for eight hours, but let’s say navigating a bomb for 2-10 minutes. And you don't need that type of accuracy. You need something three or four orders of magnitude poorer, which still was a long way off from where the initial reports were on this. But three or four orders of magnitude after you go and build the first device is theoretically possible. So we started investigating this working with the division there, and we decided at the time it was too late for us to start it from scratch.
Sorry. What division are we talking about now?
This is the group that builds the navigation systems. So we started going out and talking to people about this. And actually it turned out at the same time; Rockwell sold off their automotive division and their military division to Boeing. Boeing is sitting here saying, “We don't make automobiles; we make airplanes,” and they decided to sell this gyro business. We had been out there to talk to them. In fact, I had gone out with the VP of the division to Rockwell before this thing even happened to talk to them about a joint program with us on this, and I had taken a look at the device data they had gotten, and I said, “We need to do something here.” Well, this became available, so in 1988 [sic] we bid on this and basically purchased all the IP and the stuff from Rockwell, which had the original IP from Draper. So it gave us an in, and all the micromachining that we had developed over the last eight years just fit like gloves to be able to build this device. So in 1988 [sic], all of a sudden, it wasn’t just the pressure sensors we were working on.
’98 is when we made the offer. It became official. Yeah, it was official in ’98, I think. All of a sudden, we had a direct tie-in to a divisional application. I mean immediate. And my group was given the responsibility of transferring all of this technology into Honeywell and building up the capability and being able to fab those devices. So my group grew from just doing resonant sensors to doing gyros and accelerometers, which just solidified our connections with the operating division. Well, it’s still going. We now have those in production. But all that micromachining capability in terms of resonant sensors and machining silicon and doing the electronics and all that, it was a capability that we had developed in the ’90s that just — We were building gyros within six months after the purchase. We were building them in our lab and having them functional, had performance that was close to where we needed to be for some of these tactical applications. So there was a lot of work there in working with the Boeing guys in terms of understanding the process and transferring the process, a lot of device physics. We built up the group here. We probably had a group of 25 people just doing the gyro work within a year. So my group expanded significantly during this time.
Let me just ask you. You referred to the military/commercial division, and you ended up in the commercial division. Was that structure persisting, or did that go away?
Oh, okay. All right. No. I forgot one thing. In ’93, ’94, they merged them all back together again, and it became one research center supporting the whole Honeywell structure.
This is the Honeywell Technology Center now?
Yeah. So that covered the entire organization. Because the sensors group covered the entire organization anyway, I didn't see much change; we had already been working with everybody. So that did not change. What’s interesting, though, is there was a navigation group that was on the military side that we started working with when we merged back together again because they were doing a lot of the testing of these devices, and they were more of a systems people. They were the ones that poo-pooed us getting into this thing. So we ended up having to fight them for a couple of years to convince them that this technology could get there. It’s always the case. A lot of this technology, you know you look at a snapshot and yeah, it may not be there, but you have to look at trajectories and what could be possible. If we wouldn't have jumped at that particular time, I think we’d have been pushed out of the business, quite frankly — certainly the tactical business. So we merged back together at that point in time. Allied Signal bought Honeywell back in 1999. We stayed as the same group until 2002. In 2002, basically they split off all aerospace business from the other businesses, and I stayed on the aerospace side. So at that point in time, the gyro work, the accelerometer work, anything to do with aerospace sensors went with me. All the other sensors went on the Automation and Controls. This building is owned by Automation and Controls. All that went to the other side. At the time, I was the director of the advanced sensors laboratory, so I had all sensors. I had a gyro group underneath me. I had advanced sensor group underneath me. I had the process lab underneath me. What we did is we split anything that was non-military, and they went off to the ACS side. So now there’s an ACS sensors group that is actually fairly big. ACS is all over the map in terms of sensors. Then on the aerospace side, there’s a sensors group. They’re focused much more on gyros, accelerometers. They’re also doing a lot of work on atomic clocks using MEMS technology, atomic-based sensors, high-performance sophisticated sensors for navigation and time keeping and that sort of thing. So they split back apart. In 2002 when we did that split, I took over the aerospace side, and basically from that time on, the bulk of our work was in gyros. It’s a huge business for Honeywell, and that stuff now isn’t going to product, the work that we did back then. If the 787 ever flies, if Boeing ever gets it off the ground, the flight controls have got these devices in it. They’re being used now in the war in Afghanistan for missions guidance. So that technology, we ended up perfecting it. Perfecting is probably the wrong word, but maturing it, although we had pushed for it a long time before that. But it was a result of the capabilities that we built up. You never know how your research is going to turn out, so a lot of the research we did in the 1990s just allowed us to be able to do this. We would never have been able to do it without doing that work. That other work we ended up licensing to other people because we never were able to get it to show a significant advantage over our present technologies.
Now when you refer to the other work, you’re referring to —?
Oh, like the resonant sensors that I mentioned. We already make some of the world’s best pressure sensors, so we’re competing against ourselves. It’s kind of like the gallium arsenide-silicon analogy. So we always struggled to argue, “Why would you go with this?” Our competitors, though, were going with it. The competition out there actually has sensors out there using technology that we developed. So the research gain is always a challenge because it’s hard to predict where things are going to be, and you have to know where other technologies are going to go. But in this particular case, it allowed us then to jump in and within a couple of years, we were able to produce these new sensors. It’s an area that continues to grow within Honeywell. In 2002 when we split the labs, I took over the directorship of the aerospace labs and ran that until 2006, and that’s when I just took over this — I’d been managing then for almost 20 years, and I needed a break from having responsibilities for people’s time cards and personnel reports and all that. What I saw, again as time progressed, it seemed like we were getting much, much more short-term in terms of applications and the research.
All along, is this kind of a linear development, more short-term?
It’s kind of a linear development. There isn’t anything abrupt. It’s much more applications oriented, much more short-term oriented, much more has got to be on the product road map. Still, the outside contracts — we won a program in 2003 maybe on the chip scale atomic clock that I was able to argue that this could be important, but there’s no product out there and that would still go. Today, that’s a much tougher argument, although everybody thinks the chip scale atomic clock is something we should continue on with. But the mentality among the organization is we’ve got too many things that’s got to be done short-term and too few people to do it, so that’s where we ought to be focusing. The concern always is are you going to miss the next revolution that’s going on there, and the way things develop, if you don't get in early, you’ve got to buy your way in. And buying your way in can be awfully expensive, after IP has been developed and that sort of thing. My technical management went through 2006 when I decided to take a position in the tech strategy group. And so now I’m back to — And I think all that experience had been very helpful, all the contacts I’ve made and whatever. I work across all of aerospace now. I’m doing some work now on what the cockpit of the future is going to look like.
Now you said we’re in Automated Controls right now.
The building that we’re in is owned by Automation and Controls, and we’ve got an aerospace group here of about 100 people that we basically rent the space from them. We were in a different building up until March that was all just labs, but this place had such a huge space availability that they decided to consolidate space here.
The original split that you’re talking about, that’s as a result of this Allied Signals?
Yeah, when Allied Signal bought Honeywell — although some people say it was a merger; I’d say it was a buy-out — aerospace didn’t reorganize. Some of the other groups, ACS reorganized, but we did not reorganize. I take that back. We didn't reorganize the labs. So we basically still had a laboratory that supported the ACS side and the aerospace side from ’98 to 2002. So it was still more like a corporate lab. The money came from corporate, and that was not normal for Allied Signal. Allied Signal liked to put the research into the divisions, which has got its own issues I think. But at the time, the CEO of the new Honeywell was the ex-CEO of Honeywell. That was part of the deal, and the old CEO of Allied Signal retired. Then in 2001, the merger, the financials didn't appear to be great and the new CEO, who was the old Honeywell guy, basically tried to sell the company, because we weren’t doing well and he felt like this was the only way that he could keep the company going. He tried to sell to United Technologies, and GE got wind of it at the last minute and GE made a bid. So GE went after Honeywell in 2001. I think the times may be a little bit off here.
We can look that up if necessary.
The European Union ended up squelching that deal, and the deal fell through. As a result, the CEO basically lost all credibility during the process. So they brought in the old CEO of Allied Signal in late 2001 or early 2002, and he made the decision like that [claps hands], “We’re going to break the labs up and we’re going to put the labs underneath the divisions.” So it was kind of a delayed reaction because the merger wasn’t going well. So in 2002, the ACS part of the lab went to ACS; the aerospace lab went to the aero side. Within aero, there were still some — Aerospace hadn’t really totally reorganized, and it was not until 2005 that they finished the reorganization of aerospace and all the labs were consolidated. There were parts of the old Allied Signal labs that weren’t part of the labs that I was part of in 2002. That all got merged in 2005, so it’s one advanced technology laboratory now under aerospace. That’s the organization that’s in place now that supports all of aerospace. ACS has got their own labs. They’ve got people here, too, and also people across town. We’re kind of spread out now. That’s another thing that’s changed. It used to be everybody was in a building, and now from aerospace, we have people here, we have people in Phoenix, we have people in Tempe, we have people in Torrance, we have people in South Bend, Indiana, we have people in Moorestown, New Jersey, we have people in Bernau, and we have people in Bangalore.
So this is all within the last ten years or so?
Yes. Actually for me, it’s all in the last — well, not all of it, but this final structure has been since 2005.
Oh, okay, even since we last spoke to you. [Yes.] So you mentioned that it was in ’95 that you actually moved out of the Bloomington facility?
’95, what happened was that a new VP came in, and he said, “We’ve got too much space and we need to consolidate space.” In Plymouth [Minnesota], we have an IC facility. They do radiation-hardened integrated circuits there. It’s kind of a special IC process. There are only a couple of them in the world. But they hadn’t claimed they had space up there and they wanted to consolidate. They also build all the sensors for aerospace up there, so they thought it would be an ideal merger. So they ended up building a new lab for us up in Plymouth and moved us from Bloomington in ’95 up to Plymouth, and then sold the space in Bloomington, which is now a residential area. They tore the building down. I know because I live six blocks away from there, which has made my commute a little bit more miserable. So they moved in ’95 to that facility over there. There was another building in town that was over in northeast Minneapolis, the Camden area along the Mississippi River that had all the systems people in it. So we had a Plymouth facility and a Camden facility. Now the Camden facility closed in March and everybody moved over here, so both from ACS and aerospace moved over to here. So that’s the present structure now. So it’s been more of a consolidation of space as things change.
Even as the overall structure of the R&D has moved outward to all these different inter-city, international…
Right. So that’s another big change. In ’98 when we started transferring the gyro in, we had a group in Minneapolis and we did everything. Now if we’re doing something, I’m doing a net meeting with five different organizations over ten different time zones trying to figure out what to do.
This is a little bit of a tangent, but the net meetings — do you use a video technology for that?
Very rarely do we ever use video. We have a videoconference room here, but it’s almost all net or live meetings, and it’s people presenting slides. I think I’ve been in a total of five videoconferences in the last ten years. We set one up over in Camden and used it for five or six months, and it was just so complicated. It ended up that what people were doing was showing their slides on the thing anyway, so the net meeting or live meeting works as well.
So I think we’ve got a pretty good overview then of the business and technology history at this point. How has your job changed through the 1990s and into the current decade?
Well, I had a section manager job starting in ’86 and other than that business development job, had that until 2002. So for 16 years I was a section manager. In ’86, I was still intimately involved in the technology, not just planning programs out, but I was in there brainstorming with them and doing the technology. So I was in the lab doing some work, not that I had to, but I wanted to and was able to. I was able to brainstorm new ideas with the scientists and when I’d go out to customers be able to talk intelligently about it, because I had the time. I would spend a lot of time with the scientists. What’s definitely changed between that and 2002 was the time you spent interfacing with upper management rationalizing investments, rationalizing the number of people you had, doing performance assessments, capital equipment. The rationalization behind every dollar being spent went way up, so my job went from being 25% administrative to 90% administrative.
I see. This is kind of over the ’90s? It’s sort of a gradual —?
Over the ’90s. I’d say from ’86 to 2002 that’s what happened, 80-90%. The only way I was able to compensate for that was just work more hours, but definitely from a physicist’s point of view, I was getting further removed from the technology, and not because I wanted to; it was because the job, the continuing justification. I also spent a lot more time going after money. You’re talking to customers and that sort of thing, both inside Honeywell and outside Honeywell. You don't get to spend the time looking at the literature or spending time with your people, really understanding what’s going on and what the issues are. That to me was the biggest thing. And of course, we went through the Allied merger at that time. We went through the attempted GE thing, which was a catastrophe from my point of view. I mean GE was calling shots in here six months before the thing was supposed to go through, and we were making decisions based on that it was going to happen. So when it didn't happen, and we had lost like a year of basically good work because we were spending all this time. We had people from GE in every other week talking about what we were doing and a lot of presentations and that sort of thing. So there was a lot more upwards paying attention to make sure that those people are taken care of and a lot less focusing on the technology. You just had to rely on your scientists to do more of that.
They kind of tell you what their expectations were for what would be accomplished and then you would sort of liaise with the business side of things then and the customers?
Yeah, yeah. But again, unless you really dig into the details, everybody views things differently. One of my jobs is to present a comprehensive review, so like when we were doing this gyro technology for the eight years that I was involved with it, I’m dealing with a number of people at Ridgeway and their customers all the time. For me to be effectively communicating to them where we were going, I needed to know what was going on in the lab and I needed to explain why this wasn’t working and this was working, which meant I had to spend a lot of time with those people. Boy, it didn't leave me much time to do anything else. And fortunately in the gyro case, it was easy to rationalize. But you’re still justifying capital equipment. You’re doing performance reviews for 30 people, and you’ve got to justify why you did this because your compensation is based on this — just much more time spent on the administrative part of it. You’d get a call down mid-year that budgets have got to be cut by 5%, and where are you going to pull this and why are you going to pull it. You’d have to go through and rearrange things, and on and on and on. So when I started as a section manager, it was really a technology manager. I had a budget and I had a business administrator to support me on it, but I’d take a look at it once a month and that was about it. We focused on getting technical results. Now, man, we’ve got milestones. You’ve got SPI and CPI that you’ve got meet and…
So what’s involved in formulating these project programs, which is what they would call it back in the 1960s?
Yeah, and justification. Well, obviously we’re interested in the transition.
Well back then they said, “Dave, I’m going to give you $2 million next year. Tell me what you’re going to do with it.” We’d put a program together and say, “Here’s what our objectives are. We need to increase integration levels this much. We need to increase the yield on these processes because this is going to be important if we’re ever going to transfer to this production.” Now it’s, “Dave, what can you do to enhance our product lines in these particular areas? Give me a proposal.” No money mentioned.
Is it your job to come up with a figure?
You have to come up with “Here’s how much it’s going to cost” and, “Oh by the way, we need this in a certain period of time.” So you put this together, and there are games played because people know that if you tell them it’s really going to cost this amount, you’ll never get it. So you try to increase risk and that sort of thing. You may have a bogey of some sort, but then there’s a review process that starts on the first of June, and it isn’t finalized until the first of February of the following year. So what you’ll do is, “Here’s a list of projects that I’d like to fund.” You prioritize those, do some internal prioritization, and then you have to go in front of — We no longer have money that comes directly to technology. It comes from marketing and product management. So there’s a whole group that controls all the money within aerospace for product development, including R&D, so now we’ve got that whole group that we have to convince that this is going to benefit them. And of course, they’re much more short-term oriented. So we go through a long process of going through these projects, rationalizing them. “What are we going to get out of it?” Doing return on investments and all of this sort of stuff, even on early-on research.
Is that largely experience-based, or is there a lot of research involved in putting these things together?
There’s kind of a mixture. Obviously, people that have been around a long time know what needs to be done, but there’s still a lot of — We rank technologies now. We measure technologies by a scale that NASA developed called technology readiness levels. We’ve got a formal process we go through to measure the maturity of the technology. It needs to be a TR level 6 before you can go to a product, so that’s one of the goals of the research is to get it to these TR levels during a particular year. But sometimes it’s a crapshoot on whether you can get there. So the process is just long and drawn out in terms of just determining what we should be proposing. But you give them a list, and then they decide on affordability where the line is drawn. That affordability usually isn’t determined till this time of the year. So we still haven’t got a final — There are estimates. They start with using last year’s budget plus 5%. So you know if you’ve got $6 million this year, don't go propose $10 million. So you’d better have a pretty good idea of what you want. But you end up talking a lot to the marketing and product management groups all the way through the year about what’s driving them. But they are very much driven by shorter-term goals — they’re worried about what they’re going to do in the next five years. So what’s missing there is what are the threats 15 years out? The problem with that is that they’re not focused on that, but it’s the 15-year-out stuff that can really come and bite you. If things are just going to be incremental all along and there are no big game changers, it doesn't bother you, but if you see something coming along from the outside, and let’s say it’s an order of magnitude — I don't know if you’ve read The Innovator’s Dilemma. It’s a book written by Christensen. I can’t remember his first name. [Clayton M.] He’s a Harvard Business guy. He uses several industries as examples in there. But you’re moving along incrementally, and you’re very focused on your business, and you have to keep doing that. But something comes along and it’s like it’s an order of magnitude off in performance and you say, “I’m not going to bother with it.” Well, the problem is they forget about the trajectory, and once it gets above a certain value, maybe it can come in and replace this. Christensen goes through disk drives as an example of this, that it’s just totally changed the disk drive industry and obsoleted companies because all of a sudden, this technology is enough to meet the customer’s requirements, and you’re off there trying to perfect what you did before but it’s costing more. There are areas like that that may not mature for ten more years, but they could definitely threaten one of our mainstream businesses, that you’re sitting there watching and there’s still an order of magnitude, two orders of magnitude off, but all the IPs being developed are on them now. If you’re not playing in those areas, that’s a challenge. And that’s the danger I see in this type of a funding mechanism. We should be doing some work in there, even if it’s just developing intellectual property to protect ourselves.
Do these changes come with changes at the highest level, like if new people come in from the outside or people with different training in business?
I think it’s a combination of everything. I think it’s the demand by Wall Street on returns. It’s the philosophy towards research that Allied brought was much — Honeywell gave us a lot more freedom to explore longer-term. When Allied came in, it wasn’t immediate, but over a period of time it’s become much shorter-term, which may help the bottom line in the next couple of years. The danger is when you have these disruptive innovations that come along that you have to be aware of. The MEMS gyro is a classic case. We argued for it and couldn't get anybody to go until their customers started banging on the door, and it was only because Boeing had that thing for sale that we got in at the right time. But if we had waited a couple more years, we would have been shut out. Then it still was another eight or nine years before it got into a product, but you wouldn't have been able to get in! You wouldn't have had the IP that allowed you to play in that. So it would have knocked you out of an entire market. It’s a tough problem because you’re kind of guessing, but I think there are certain technologies that we end up sacrificing because of this more shorter-term outlook, because the people who are controlling the money have a more short-term outlook. You’ve got product management and marketing group controlling that. Now in the past, it used to be a corporate level, and product and marketing, their job is the next five-year road map. So there is less of an inclination to worry about ten years out until it’s too late.
I noticed that you are a Black Belt in Six Sigma. Is that a big deal around here, or is that just something else?
It was. When Allied Signal bought us out, they were almost evangelical about it. Now it wasn’t that we weren’t doing it. We had Black Belts in the company; we just didn't make a big deal about it. Larry Bossidy, who ran Allied Signal at the time, was a protégé of Jack Welch at GE. Maybe it was the other way around. I think Jack may have gotten it from Bossidy or whatever, and when GE came in, it was the same thing. They were just nuts about this stuff. The first thing Allied Signal did when they came in is that we’re going to apply all this Six Sigma. As a physicist, as a scientist, I kind of looked at it and it’s a fad. I understood what the rationale was, but doing it in research I thought was ridiculous, and yet they were trying to apply it to research. It was also used as a guise, if you will; to I think cover up stupid decisions, too. I’ve had some arguments with one of the divisions about the decisions that were made. “Well, we used Six Sigma methodology.” I said, “Yeah, but the decision was stupid!” But you know they show all the charts and all that. But anyway, it was encouraged from the start. All engineering was required to be Green Belted after a couple of years. We were doing a lot of stuff with the gyro at the time that I thought would help us in the long term that I felt we could apply these Six Sigma tools. So I ended up doing my Black Belt on a bunch of different subject matters for the gyro, where all it was doing was just applying the tools to get the answers that we were looking for. It was a natural fit, so I went for it. And it was encouraged among all upper management that they had to have at least a Green Belt. It wasn’t much. It was four weeks of training. I spent a lot of time on it, but it allowed me to do technical work with the people in a group. We did a lot of statistical analysis of data, improving wafer uniformity that we needed to transfer this into production, so it just fit naturally, and it allowed me to keep my fingers in the technical till. So it just kind of came about naturally in that particular case. So I am certified. I haven’t done much work in it. It’s calmed down a bit within the company. They’re not evangelical about it anymore. There is recognition that especially in manufacturing it’s important. The language is used a lot. There are certain tools we use on a regular basis within advanced technology.
It’s just interesting. In my own research in the 1950s, the Rand Corporation was interested in program planning and budgeting systems and that sort of thing goes to the Department of Defense, notoriously, in the 1960s. So then when you see the Six Sigma coming onside, well it’s kind of just a similar sort of — I mean it’s not the same thing, but a repackaged version of a set of tools for planning.
Yeah, yeah, a lot of it.
This happens to be an import, but I think we originally exported it in the first place.
Yeah, yeah, yeah. We were doing FMEAs all the time before that. It just gave it more visibility. There are a lot of cynics about Six Sigma around here because of trying to force fit it.
How about this most recent job then? Or is there anything to talk about with the transition from section head to director?
Well, at that time it became much more Phoenix-centric because aerospace headquarters is in Phoenix. So I was on a plane to Phoenix at least once a month, interfacing a lot more with upper management, rationalizing why we’re spending the money, talking a lot with the business units, what their reasons were and whatever. So there was a lot more upwards attention paid, a lot less being able to do technology at all, just making sure that everything was running and that the funding was there and that the funding was justified and it was going to stay there, and keeping upper management aware of what was going on and what was required. So I spent a lot more time in Phoenix, a lot less time with the groups. Any customer interface, I still try to spend a lot of time with that. But the biggest difference was that all of a sudden I had 60 people that I was responsible for rather than 20, and all the stuff that goes along with that, from health benefits to disciplinary actions and all that sort of stuff. It was a pretty intense time, and it was an exciting time. This was when we finally combined all of the research in aerospace into one organization, advanced technology, so I was part of that. That’s all centered now in Phoenix. I’m still part of that organization now, but I actually report to one of the directors. I’m a staff scientist now, and they’ve given me latitude to just do a lot of things. I’ve been doing patent analysis trying to figure a way of how we can better evaluate the quality of our portfolio. We’ve got a lot of patents, but what does that mean? So we’ve looked at that. I’m putting together intellectual property strategy for each of the groups. You mentioned Rand Corporation. I’m doing scenario planning; I’ve done two projects this year with a couple of groups. And also evaluating some new tools for speeding up ideation, that sort of thing. So I’m kind of all over the map. I’m doing technical intelligence, which is just understanding what’s going on out there. It’s a real challenge. There are a lot of things that are peripheral to the business that we need to understand. We do what’s called these key intelligence topics. We did one earlier this year on alternative fuels for jet engines. That’s going to be a real issue in the next 10-20 years, and we need to understand what alternative fuels could possibly happen and what the implications are for engine design or the fuel systems, because you’ve got to build that into your research. You’ve got to know that there is not going to be any change, or that all of a sudden you may have to have different seals on the engines. You’ve got to develop those now. Same thing with electrical power on board an aircraft. The aircraft are going in a direction that it looks like more and more electrical power is required, and how do you deliver that on an aircraft, especially with fuel prices going out the window, although they seem to have come down lately. But they’ll eventually go back up. So I’m kind of working all over the board on there. They’ve given me a lot of latitude with the technology strategy group, but basically the strategy group is determining where we should be investing our money. So we need to have the intelligence. We need to know what the future is going to look like so that we’re placing our bets in the right spot. We have a group within the strategy group. All the technology councils are under the strategy group, so we work with the technology councils and work with the intellectual property councils, work with all the fellows in the organization. So I’m in areas I have known nothing about that I’m learning the avionics business or the aerospace business and learning about jet engines. The physics background — there are always parts of it that I utilize. When people start talking about thermal barrier coatings on jet engines, it’s similar to doing epitaxy and why you’re doing it. It’s a layer put down to protect. Well, we do that same type of stuff elsewhere, so you kind of know the rationale and what the issues are behind those things.
I wanted to ask you a little bit about you have customer contacts. But then do you have sort of joint research projects? I think you mentioned Sandia in your previous interview, that there was some interaction with them.
We had in the past. Now again because I’m — Oh, I’m also responsible for running the university programs with advanced technology, so we have research programs with universities that I coordinate. I’m basically the guy that makes sure that we have research agreements in place and we’re kosher on the Honeywell rights and that sort of thing. Then I control the pot of money, kind of control it. When I had the department manager’s position, we had joint projects. We do contracts with the universities, and we also fund the universities directly. We also belong to consortia, and consortia I control the funding for, too.
Okay. A consortium is?
Well for example — we don't belong to it anymore I don't think — but Berkeley had one called the Sensors and Actuators Center. It was set up by National Science Foundation. For $50,000 a year you could join that. Basically, you get access to all the research that’s being done in this. Usually, there are 20-30 companies that are funding this, and there’s about seven or eight professors at Berkeley and 100 graduate students. We meet twice a year at Berkeley. You get to see all their research up front. You get access to all their IP, and you get access to all of the early, up-front R&D. So when we were doing MEMS, I was out there at least once a year to see what they were doing, because they had great ideas, and we took some of their technology. We have consortia like that in materials, composites, polymers — a wide range of those that are being funded. And it’s usually $15,000 to $30,000, $40,000 a year membership. So we have this budget to do that. We’re selective on which ones we do, but then that gives us access to research that’s being done by that group.
Are those long lasting or are they kind of ephemeral?
Some are long. Berkeley went on for almost 20 years. I don't think we’re still members of it, but it’s partly because the nature of our research has changed and partly because it’s matured. Some of them are just maybe for a year or two years. I’d like to see them longer than that to establish long-term relationships, but it kind of depends. So we have some groups that fund universities directly out of their R&D budgets. They may think that this is worthwhile looking at, this new coating or something like that. We don't have the people to do it, so we pay the university and say, “Go take a look at this material and tell us what it does,” and that’s it. Other ones we’ll have a longer-term relationship. We want to explore… there’s possibly a new catalyst out there that we want to examine, a new way of forming the catalyst, and we want to see if it will work. So we’ll give the university money to go do that and if it works, then we may transfer the technology in.
When you work with the universities, is it sort of on an à la carte basis, or are there ones that you establish relationships with?
Well, we’re trying to establish relationships, but I think it’s really tough. You may have a firm relationship, and then a professor moves to another place, and you want that relationship, so you’ve got to set up — We actually try to do what’s called master research agreements with certain universities that once we get negotiated, we just go in and do project after project, but what I’m finding out is that the next person who comes along doesn't want a project with them. He wants to go with this other person and just do a single project. The biggest issue there is just negotiating the agreement to make sure the intellectual property is covered, because universities, they think you’re going to invent the next Google. The numbers of patents that turn into something like that are like 1 in a 1000, so you spend an enormous amount of time with legal and contracts trying to get an agreement. In fact, I have a meeting this afternoon on agreements with a couple of universities who insisted upon some protection for IP and the universities don't want to give it, and I can see both sides.
So they’re actually quite grabby about the IP then.
Oh absolutely. Absolutely. They see universities like Stanford getting $90 million a year for PCR. Wisconsin got it for years for pasteurization of milk.
I think we’re coming to the close of [the interview]. One thing that we try and establish — you might be the wrong person to ask, so I’ve been pushing it off into the shorter time span at the end of the interview, but I’m sure we quizzed you about record keeping and that sort of thing before. We’re always very interested in that. One of the things over at 3M, for example, that was interesting that developed since we talked to them was that a lot of their technical reports had just gotten purely to PowerPoint as opposed to being more of an explanatory document. So I was wondering if there have been any developments over the last several years.
It’s a mix. It’s all over the board. I don't think we have a very good — I still keep a data notebook. I write everything down in there. It’s a mixture of electronic files and data books, and when people leave, there’s no —
So even at your level, you’re still using the notebooks?
Yeah. It’s a challenge, collecting and archiving that information. I know in the past when people would leave — When Dave Zook retired, Dave gave me two boxes full of lab notebooks. Well, who’s going to go look through those lab notebooks? There are tools out there now that are coming online. I’m evaluating a tool now; it’s just taking me forever. Some companies do it now. You can optically scan this material and archive it and index it using this, and you can use that indexed information then for training or for active research. The problem is you’ve got to scan all the material in. If you’ve got an optical scanning system you can do it, although people are getting more and more electronically, keeping their stuff. It can scan that, too. So I’m actually working with the librarian next year on a project where we’re going to try to do that so that when people leave, we can capture electronically all their important stuff and index it and keep it on a server. So people can now come in and say, “Okay, what did Joe Blow do on this particular sensor while he was here?” It’s not perfect, but right now, when people archive and throw it in a box, you don't get anything.
Yeah, there’s nothing at all.
There’s nothing at all.
Do you have certain databases that you use, or is it just sort of ad hoc on drives that are in servers or something like that [for what?] for archiving reports distributing and so forth?
Well, that’s what we’ve got to establish. Right now it’s on PCs and maybe up on the server some place. But when a person leaves, I think other than the section manager; we don't know where the information is.
So when it comes to technical reports and that sort of thing, it’s just sort of ad-hoc. There’s no central system?
If it’s a real technical report, which is still fuzzy, it’s supposed to go to the library and the library keeps an electronic database. But all that stuff in between…
Is there a move toward electronic notebooks for R&D people?
No. No, there’s not a concerted move. I think that some people do better jobs than others, but that’s a real weak link we have now in my opinion. That’s something that nobody’s really paying a lot of attention to. It gets better when you get a little bit further along in the development process. When you have to start making drawings and all that, there is a formal system to do that. But the R&D stuff, I can’t tell you the number of times — You know, the old round thermostat has a mercury switch in it, and I bet you we went back and revisited the mercury switch a dozen times over the last 25 years on what can we do better, and I bet you we reinvented it 20 times. With some sort of an archiving system that you can actively retrieve, a lot of that extra work would go away. But it’s…
It’s just too much of a mess to reconcile.
It’s hard to coordinate that. It’s a real challenge. I was hoping this tool would allow me to do it, but it’s a system problem. It’s just not one guy can come in here and — We’re going to do a couple of experiments next year to show how we can utilize it and then try to expand it out. There’s just no other way of doing it, from what I see.
It’s funny. When I was working out in Minnetonka, there was a company of about 150 people; that includes everyone who was working for it. But I was kind of the guy, actually. That’s what I was hired for. I only worked there for a year or two. I would try to marshal things, get them into some level of order. That was lab notebooks, technical reports, and all that sort of thing. So it’s a little bit ironic that I’ve been asked to ask these questions, because it’s very familiar to me. Just before we break off, is there anybody else that you think that would be especially pertinent for us to do life interviews with or that we ought to talk to to kind of get a hold of things? I mean it was an excellent business history that you gave us.
Dave Zook would be a great guy, but he’s retired. But he still may be willing to talk.
Yeah, that’s entirely possible. He’s around here in the Twin Cities?
He actually lives about two miles away from here. I don’t know who else you guys talked to when you were here. Dave Zook would be a good guy from a technical point of view. He could give you a really good history going back to the late ’60s, early ’70s, more from a technical point of view. If you could get a hold of him — I could probably send you his phone number.
That would be great.
But other than that, I can’t think of anybody.
Okay, that sounds good. Thanks so much for your time.