# Oral-History:J. Ross Macdonald

Dr. J. Ross Macdonald was born in Georgia and attended both Williams College and the Massachusetts Institute of Technology as an undergraduate. He received a B.A. in physics from Williams and a S.B. in electrical engineering from M.I.T. in 1944. During World War Two he served as a Navy Radio-Radar officer and then returned to M.I.T. in 1946. He received a Rhodes Scholarship after earning a masters degree in EE from M.I.T. Oxford University granted him a D. Phil degree in solid state physics in 1950 and the D.Sc. degree, based on published papers, in 1967. Macdonald worked at Armour Research Foundation and the Argonne National Laboratory before joining Texas Instruments in 1953. He was Director of the TI Physics Research Laboratory for several years and then became Director of the TI Central Research Laboratories. He served as TI's Vice President of Corporate Research and Engineering from 1968 until 1972, and was Vice President of Corporate Research and Development from 1973 until 1974. He was named the William Kenan Professor of Physics at the University of North Carolina, Chapel Hill, in 1974. His research has been in areas such as ferromagnetic resonance, the electrical behavior of solids, electrical circuits, dielectric and mechanical relaxation, electrolyte double layers, and numerical analysis of experiments. He has published over 100 papers and is a member of the National Academy of Sciences and the National Academy of Engineering.

The interview spans most of Macdonald's career, beginning with his college education. Macdonald describes his undergraduate years and M.I.T. and Williams College and then discusses his service in the Navy during World War Two. He characterizes his career as marked by his interest in both engineering and physics, and demonstrates how this influenced his research and applied work. Macdonald discusses his graduate studies at M.I.T. and Oxford as well as his work in areas such as ferromagnetic resonance, photoelectric and photoconductive effects, field-effect transistors,, computers, Impedance Spectography, and biomedical engineering. He evaluates his time at Texas Instruments and his research there, the company's research priorities, and his colleagues Gordon Teal and Pat Haggerty. As the interview closes, Macdonald appraises his connection with IEEE and describes possible applications for theoretical studies of Impedance Spectography.

J. ROSS MACDONALD: An Interview Conducted by Rik Nebeker, IEEE History Center, May 7, 1993

Interview # 154 for the IEEE History Center, The Institute of Electrical and Electronics Engineers, Inc.

This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.

Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, IEEE History Center, 445 Hoes Lane, Piscataway, NJ 08854 USA or ieee-history@ieee.org. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.

It is recommended that this oral history be cited as follows:

J. Ross Macdonald, an oral history conducted in 1993 by Rik Nebeker, IEEE History Center, Piscataway, NJ, USA.

## Interview

Interview: J. Ross Macdonald

Interviewer: Rik Nebeker

Place: Dr. Macdonald's home, Chapel Hill, NC

Date: May 7, 1993

### Education

Nebeker:

This is the 7th of May 1993. I'm talking to J. Ross Macdonald at his home in Chapel Hill. This is Rik Nebeker.

Macdonald:

I got a bachelor's degree in physics from Williams and a bachelor's degree in electrical engineering from MIT, both in 1944, which has its difficulties at this point. Nineteen ninety-four is the 50th anniversary of my class in both places, and they're both having celebrations at the same time. So I probably will end up not going to either one, but who knows? Anyway, after I left Williams I went to MIT and worked on a bachelor's in electrical engineering, and finished in February 1944. I then became a graduate student in electrical engineering, and among other things taught a little bit at the radar school, the MIT radar school for Army and Navy officers.

Nebeker:

That was not the Rad Lab?

Macdonald:

That was not the Rad Lab. It was the school that prepared people to be radar officers in both these areas of the armed services. But I was also working on a master's degree. That only lasted for one term, and then I joined the Navy.

### World War II and Navy

Nebeker:

What field?

Macdonald:

Well, it was electrical engineering, and I was taking courses primarily at that time. But I was concentrating, if anything, certainly on electronics. I joined the Navy in June of 1944, and, after officers' training in New York State, near New York City, the Navy sent me to pre-radar school at Harvard for a few months. Although I'd been teaching radar school, I didn't object too much to this because at that time I had a girl who lived in the Cambridge area, and I was not mad to go off to war. So I went through that school. Then the normal progression from that was to go to the MIT radar school, which I went through as a student although I'd been a professor there earlier. That was of some interest and I certainly learned some things as a student that I hadn't known as a professor.

After I finished that school, I was posted into night-fighter radar in the aircraft part of the Navy. I was an ensign by then. I went to a school in Corpus Christi, Texas, where we learned more about the specific radars that they used for night-fighting planes such as the F6F — I believe it was the main one in those days. After that they sent me to night-fighter squadron training, which was at the Charleston, Rhode Island Auxiliary Navy Base. I learned there to sleep with a pillow over my head because the barracks were near the airfield, and of course night fighters meant that they were coming in and going out quite a bit at night.

Nebeker:

Is this still during the war?

Macdonald:

Yes. This was, I guess, the summer of 1945. While I was there, they dropped the atomic bomb. They'd had tests, but they dropped the first one on Japan in August, I think. I and a girl had gone to spend a few days or a day or two at Nantucket, and had taken the ferry over there, and rented bicycles and had ridden to the beach and gone swimming. I guess it was just a day excursion. When we came back toward the end of the day, we heard all this news. We had a fine day, but that was not a very nice sort of conclusion.

Nebeker:

You didn't know because it wasn't publicly known what that was.

Macdonald:

Exactly what it was.

Nebeker:

Did you know?

Macdonald:

Audio File
MP3 Audio
(154 - macdonald - clip 1.mp3)

No, not really. I mean I'd heard the words "atomic bomb" or "explosion" or whatever, but at that point I didn't know much. In fact, they kept everything quite secret. I remember back when I was first at MIT as an undergraduate in 1943, they had these domes up on top of some of the buildings, and I often wondered what was inside them. Of course they covered radar antennas, among other things. Anyway, the dropping of the bomb wiped out the plan of our squadron to go to the Pacific, within a month or two, which had been in the works, and they didn't do that. So I was transferred to an air base, a Navy airfield, in Boca Raton, Florida. This was just a holding pattern, and I didn't do anything there of any interest particularly. One of the higher officers got me in and forced me to spend time writing history of some activity that had gone on there. I am a reasonably literate and fluent writer, but I was in those days much more interested in science and in learning more and doing it, than just writing about the history of the base. I don't mean to demean your area, but I it didn't appeal too much to me.

After a while — I forget whether I was able to do this myself or what happened, but I think I had some hand in it I got transferred to the Naval Research Laboratory in Washington with something called the Airborne Coordinating Group, which was a group of Navy officers specializing in radio and radar who were traveling field representatives. If there was a problem somewhere, they'd send us off to try and solve it. This was in the spring of 1946, and the war was of course winding down. I certainly didn't contribute much to the Navy then. I was posted to be on an aircraft carrier, The Princeton. CV37, I think, was its designation. It was one of the big ones, but later on I think they had another one with the same name that was even bigger. I was on that for a couple of months maybe. We had a nice cruise in the Caribbean down to Trinidad and back. I was not on the main duty roster of the ship, so I didn't have much to do. I learned a little Spanish on my own and a few other things. From my point of view it was a nice trip but a waste of time. I did have some scientific books with me, and I studied relativity and a few things like that. It's a bit hard to do on your own.

### MIT and Digital Storage Tubes

Macdonald:

When that ended, and I was able to get out of the Navy in the summer of 1946, I immediately got married and went back to MIT. I can remember an interview in Washington, DC with Jay Forrester and his second in command, Bob Everett, who became head of the Sage Systems Laboratory later on and is now a member of NAE. Jay Forrester, who was the head of a project at MIT to work on building a computer to simulate aircraft flying interviewed me in Washington for a job on his program. When I got to MIT, I worked part time on that and part time on finishing up a master's degree in electrical engineering, and both were of considerable interest.

But the thing that got me into all of this was this paper here, which I'm now going to speak about. I did do a master's degree thesis, and although I'd wanted to do it on dielectric response, I was, in my mind, forced to do it within the group at MIT that was working on the Whirlwind computer. Forrester, who was the head of that project and only about six or seven years older than I perhaps (maybe a little more), insisted that he be the one who would oversee us all through the master's degree. So I did a thesis on storage tubes for digital computers, where one has a beam of electrons, as in a present-day TV tube, that goes to a surface, and the electrons hopefully get stored in discrete patches on the surface. You use another beam to weed them out so you can see whether they're there or not there. It was a problem of how to make it have high resolution. We made a few and did a few things. But that was something that only lasted for about a year, and I got my degree. The thesis was of some interest, but it certainly did not shake the world; and the tubes that we developed did not turn out to be competitive with some other possibilities.

Nebeker:

Did you design the tubes?

Macdonald:

Well, yes and no. I'm trying to look in this paper to show you. This is not the thesis I was looking for, but I couldn't find a copy of it here. I'll remind you that almost all of my papers are in storage at the University of Texas at Austin, Texas. They have a large repository of scientific papers. Here's a picture of the tube that I worked on. I was particularly concerned with the surface itself. We did build some tubes, and I did have a little to do with building it, but I was much more concerned with making a surface that could store the electrons, doing a few theoretical calculations, and doing practical work on that. Anodizing aluminum to make the surfaces, and so on.

Nebeker:

Was the main effort to increase the length of time one of these tubes could store the signal?

Macdonald:

It's been a long time ago now, and I haven't refreshed my memory on this. But my feeling is that they had another beam that could be used to refresh it from time to time. Therefore, I don't think that was a strong limitation. I have the feeling, but I may be wrong, that one of the limitations was that it was hard to get enough resolution and to store as many bits as one would like to do. My point in giving all this background is that in addition to doing a thesis at that time MIT required that one write a long seminar paper to get the master's in EE. This was one called "The Storage of Information — Its Evolution and Future."

It has a lot of background about where things stood at the time, where they might go, and what the future might hold. If you look at the beginning of it, you see a couple of the letters, one from. There was a lot of interest in the area. Unfortunately, Van Bush said that he didn't have time to look at it then. He said he would later, but I don't recall ever getting anything more from him. I got to know him quite well in later years. I never asked him about this again. I found I had three or four copies of this. One copy at the time was deposited in the Vail Library at MIT. It might still be there, but who knows how long they keep such things. So it occurred to me that with these extra copies I might see where it would be most reasonable to go, and my friend Harvey Cragon that I told you about, who is still very much in the field, suggested that the Babbage Institute would be a reasonable place and would be interested in this. I've sent him a copy that he wanted to keep, and he has said that he will write me a letter, suggesting that it should go to Babbage. Then I will send a copy to them. So I wanted you to be aware of this because I think that it's an important checkpoint.

Nebeker:

Oh, yes.

Macdonald:

In computers we now talk about memory rather than storage so much, and even at the time I was strongly against that because I thought that it was wrong and I preferred "storage". But I lost that battle long ago.

I think it is a checkpoint of what was going on at the time. There was an article that came out of this in Electronics sometime thereafter. I lent them a copy, and they wrote an article on it. They lost the copy, but they Xeroxed some kind of a copy that they'd made of it, but lost my original. It was from the summer of 1946 to the fall or summer, I guess, of 1947 that I worked on my master's degree at MIT and did these two things. Then I moved back into physics, having the master's degree in electrical engineering. I decided that what really interested me more was creative and, to some degree, theoretical work. I decided that I would like to get a Ph.D. in physics, so I shifted from EE to the physics department and started taking more courses there. I had been very lucky in my academic work at MIT up to that point. I had something like a 4.95 average out of 5.0 in EE. I was pleased with that. I didn't do quite as well in physics, I'm sorry to say.

### Rhodes Scholarship and Oxford

Macdonald:

But I was going along quite well, and I had an aunt who lived in England, married to an English clergyman, and we happened to see them around this time, perhaps the fall of 1947. She said, "Why don't you come to England? Why don't you apply for a Rhodes Scholarship?" Another piece of background: I should say that I'd been very lucky in my education up to that point in that Williams College had given me a scholarship when I'd first gone for tuition. Later on they gave me a Tyng Scholarship, which was one that paid more of one's expenses and had some money for graduate school involved in it. When I was out of the Navy, I also had the GI Bill. But at MIT for the master's degree I was married by then, and we were living in Westgate, one of those married student housing units and didn't have much money. I remember having to get a loan from MIT to keep going there because in those days I'd stopped working in the computer lab, so I didn't have that income any more. In those days you weren't necessarily a teaching assistant or a research assistant. You weren't necessarily paid anything by the department. Since my mother and father were divorced and neither of them had much money, it struck me that anything that could help support my education being attained might be a good thing. I decided I had very little to lose and possibly much to gain by applying for a Rhodes Scholarship, so I did.

But at that dinner each of us was asked to say something about our association with Williams and what had happened in connection with the Rhodes or Marshall Scholarship. So when my turn came, I had to say that I was there on false pretenses, that I was very pleased to be there, but that I was not a Williams Rhodes Scholar, and told them what had happened at my first Rhodes interview, I told the interviewers this: The head of the interviewing committee is not a Rhodes Scholar himself. He's a distinguished educator. He happened to be Phinney Baxter, who was the president of Williams at that time and also a distinguished historian of the war. So when that first session was over, and I'd been elected that far anyway, (although there was another session later to go through), Phinney, who I knew to be at Williams, came by and said, "Ross, I was disappointed that you didn't apply through Williams." I was pleased to tell him the story.

Later on I went through the Division Rhodes Scholar selection and was lucky enough to get selected. The other point was that Rhodes Scholars have a choice of being nominated from their own state or where they go to school. The state where I came from, lived more or less, is Georgia. While I knew that Georgia was likely to involve less competition than Massachusetts, which had MIT and Harvard and all those schools, I took an attitude of what will be, will be, and just decided that if I got it, fine. If I didn't, that's too bad. So I decided it would be cheaper and easier to do it from Massachusetts. That's what I did, and luckily enough I was able to get it. This was right after the war and for two or three years they allowed you to be married even your first year. So I was married, and that meant that my wife went with me when we went to Oxford. We were there for two years, and I worked on solid-state physics, ferromagnetic resonance, and got a D.Phil. there in 1950. That's the background for that and for some of my education. I have a few other things here that come later in my career that I'll tell you about later.

Nebeker:

What was England like then?

Macdonald:

It was right after the war, and the rationing was extreme. We practically lived on boxes from home. But I will leave that behind.

Nebeker:

What people did you work with there at Oxford?

Macdonald:

I was in the Clarendon Laboratory there, and I did a combined theoretical/experimental thesis on ferromagnetic resonance, which had just been discovered a couple of years previously by one of the people there named J.H.E. Griffiths. He was my main thesis advisor for the experimental part. But there was a theorist there named Maurice (M.H.L.) Pryce. I did consult him several times, at least in the theoretical part of my thesis, and I did sit in and take a number of his theoretical courses. He was a very brilliant guy. So those were the two principal ones. Since I already had a master's degree, I became an advanced student immediately and didn't have to take any courses. When I'd been at MIT, I had to pass two languages for my Ph.D. language requirement, and French was pretty easy. I'd studied some of that in high school, prep school, maybe even a little in college. But I'd studied no German to speak of. So I sat in on a German course at MIT and eventually took the exam and passed it. But I found it very difficult to do it that way. When I got to Oxford, I found that I didn't have to have any language tests. So in a way I'd wasted my time, although it wasn't quite wasted. I'm a great lover of German Lieder, and so knowing German to some degree has not been bad for me.

Nebeker:

That's interesting.

Macdonald:

I have not had to translate many German scientific papers, I'm glad to say. Even when I knew it best, it was difficult for me.

### Electrical Engineering and Physics

Nebeker:

That's interesting that you moved from EE into physics.

Macdonald:

But I'd been in physics to begin with, remember.

Nebeker:

I see. And you went into EE.

Macdonald:

I went into EE partly because of the war — practical things — and partly because I wanted to have a combination of basic research and practical applications. In fact, in my Rhodes Scholar application essay, I made that point, that I wanted both. I'd gone to Williams to begin with to get a liberal arts education as well as physics, and I'd gone to MIT primarily to learn more about the application of knowledge.

Nebeker:

Still thinking that physics was your main interest?

Macdonald:

Not necessarily. But I did learn eventually that I thought that physics was more appropriate for me. Certainly when I got into Oxford I did physics there. It's turned out through my life that I've never regretted having the EE and electronics, and I've used a great deal of it over the years. I've not regretted the Ph.D., the physics background, either. So let me stop for a moment and see if you have anything else before I go on.

Nebeker:

I've talked quite a bit with Charles Townes, and like you did, he wanted more of the EE than physics.

Macdonald:

Yes. I've met him a few times and certainly admire his work very greatly.

Nebeker:

He talks about how in those postwar years the electronics developed for radar during the war made a big difference to physics.

Macdonald:

They did. In fact, my work on ferromagnetic resonance could not have been done without the radar work that preceded it because I worked at K-band, where you use waveguides and silicon or germanium detectors. What one was doing was looking at the absorption of electromagnetic energy and ferromagnetic thin films as a function of the static magnetic field, which you would change over a wide range, but slowly, and over the frequency of the electromagnetic waves. Essentially you would get a resonance curve by having a fixed frequency and change the magnetic field strength over a wide range. This was microwave spectroscopy. It just happened to be a magnetic rather than a nuclear or ferromagnetic resonance. Here is my thesis. There's the waveguide. I'll show you a resonance curve in its magnetic field.

Nebeker:

I don't know anything about ferromagnetics.

Macdonald:

It has never turned out to be of magnificent importance, though work on it led to magnetic storage, which became important for a while and then became less so — although rotating-disk magnetic storage in hard disks in computers is still very important. Even that, I believe, is on the way out. I think it'll all be electronic sooner or later, without moving parts. But nevertheless, work on ferromagnetism has been of some importance. At one point I knew more about ferromagnetic resonance than anyone in the world, I think. That only lasted two or three years after my thesis, however.

### Armour Research Foundation

Nebeker:

What did you do after your thesis?

Macdonald:

After my thesis I was anxious to get a job and support my family. We had had one daughter while in England, using public health, which didn't cost anything, which was nice. That daughter is now a scientist at Bell Laboratories, named Antonina Hansell Macdonald. We came back to the United States, and unfortunately I didn't have much of a network of friends and professors because I'd been in England. So I had a little more trouble finding a job than one might have expected, although I think I'd done quite well at Oxford. I'd finished the Ph.D. in two years, which is much shorter than it usually takes. I interviewed at a number of places and got an offer from the Armour Research Foundation in Chicago, which is associated with Illinois Institute of Technology, to be on the technical staff. I joined them, was there for a couple of years, and worked on a variety of things, mostly government contracts to do research of one kind or another.

Then as part of one of the contracts, I discovered some interesting effects in alkali halides, some photoelectric, photoconductive effects. There was a distinguished ex-German professor at Argonne National Laboratory at that time who had a group working in somewhat the same area. I went over and talked to him about some of the things I was doing and was interested in, and I was invited to take a year's leave-of-absence from Armour to go to Argonne to work with his group. His name was Peter Pringzheim. He's now dead. He had a group of four to five, six people working with him. So I went there, and I did both some theoretical and experimental work in the area, and wrote a couple of papers, and decided that I didn't much want to go back to Armour Research. I wanted to find more of a permanent job, and even Argonne was not that much of a permanent job. So at that point I interviewed at IBM, Bell Labs, and a few other places. Luckily, it turned out that Texas Instruments was in the process of building up a research lab that they'd just gotten started that year.

### Texas Instruments and Transistors

Nebeker:

Was that with Gordon Teal?

Macdonald:

Gordon Teal had been hired by TI to do this. They had tried to get Shockley but had not succeeded. Shockley wanted to make a million dollars quickly, and they were not prepared to make that possible for him, quickly anyway. It turned out that had he come there, he would have made his million or more, but what he did instead, he didn't make it. So perhaps he made a mistake. Anyway, Teal came, and he happened to know someone at Brown who worked in solid-state physics and who was a friend of one of my professors at Williams. Somehow this professor knew I was looking around, and he told the man at Brown. Teal, who is a graduate of Brown, went there and asked them about people and so on. This man gave him my name. That's how I was approached by Teal, I guess it was while I was still at Argonne. I remember he came and interviewed me in Chicago. This was a very long shot to go to TI at that point. It had $25 million in sales; when I left it in 1974, it had$1.5 billion, and it's now on the order of six or seven billion. But at that time it had been in geophysics, and during the war it had done breakout production of electronic equipment. It had hired a brilliant guy named Pat Haggerty who'd been in the Navy.

Pat Haggerty became vice president and decided that TI ought to get into transistors. In about 1952, I think it was, Bell Laboratories had a kind of a course or informational gathering for a couple of weeks in which all companies that wanted to be involved had to pay $25,000, and they would go and learn about transistors. This was only a few years after the invention of transistors, but Bell had one or two books and other things that people there had written. TI was the smallest company that did this, in my remembrance, and$25,000 was a lot of money to them. It wasn't so much for GE or RCA. So TI went ahead and did that and learned enough to be able to start doing a little bit on transistors and hired Gordon Teal. He and they decided to build up some research in the engineering area. He made me an offer, and I guess it was another time when I decided to go all out and take a chance. It was a very good thing that I did. I made a good decision, although there were lots of people at the time, relatives and others, who thought it was a terrible decision. This little company off in Texas. I later talked to people from RCA who said, "Hey, why did you join TI?" or "At RCA we've got a much bigger program, we're doing great, and you've joined this little company? Why don't you leave there and join us?" It wasn't too many years later that RCA had no program, and a few more that they went out of business completely. I do remember being invited as a representative of TI to the 25th anniversary of RCA research laboratories. They had a day of visiting the lab and a big dinner with speeches.

Anyway, I did make the decision. I joined TI. I remember that it was very informal. There were very few top-level people at that time. I was lucky in that one of the people that I put down for my background references was named John Stuart Dudley, who was a lawyer in New York City. It turned out that the lawyer at TI had known him and perhaps had worked fairly closely with him. That gave me some cachet to begin with. I remember going down there and being interviewed. I was offered the job, and it paid, in fact, appreciably more money than I'd been making. It seems to me that when I came back from England and joined Armour I was making $6,000 a year. When I joined TI, they offered$9,000. By then I was making somewhat more than $6,000, but it was not$9,000. Now this seems like nothing at all, but then it was not too bad.

Nebeker:

What was your initial work when you got there?

Macdonald:

There were only five or six of us in the group initially. These people included: Malcolm Brachman, Mort Jones, Ed Jackson, Willis Adcock and one or two others. I remember when I first got there, I had a few ideas of some theoretical work on diodes and semiconductors, and I wrote a paper on that. Malcolm Brachman was a brilliant Ph.D. in physics who'd gone to Yale and then to Harvard and gotten a Ph.D. with Schwinger there. He had worked a bit in Los Alamos on the opacity problem. But when he had been found by Teal and offered a job, he came from Fort Worth, and it probably appealed to him to be back home. As soon as we were hired, we were sent off by TI to go to a summer school on transistors held at the University of Illinois, in 1953 I guess it was. It was taught in part by John Bardeen, and some other people in transistors came there and gave lectures. Bardeen was a terrible lecturer. He talked very quietly and perhaps didn't organize his things in a way that was very clear to people. But he was an exceptionally brilliant person, and a very nice guy, and I certainly learned a good deal from him and the others. At that point, of course, I had not had much background — a little solid-state theory and solid-state courses, but very little on transistors. I'd finished school before transistors got into the curriculum.

So this was a good thing to do, to go there. Malcolm and I were roommates, and I was amazed at how quick and smart he was. So he and I were both there at TI sitting at desks next to each other. I seemed to be somewhat more creative than he was, but he was perhaps more brilliant in being able to solve things. And I suggested a number of problems, and we collaborated on several papers in those days, mostly theoretical. One important one was on linear system response, integral transforms, and things of that nature — by me and Brachman. It was a long review article published in the Reviews of Modern Physics in 1955. That paper has been used and quoted quite a bit over the years, but of course now it's pretty much over and gone. Other people have done much more since then. After a while, Malcolm decided that he was going to go back into the oil business. His father was a rich oil man in Fort Worth, an oil equipment supplier. So Malcolm gave up science and worked for his father and set up a separate group. In later years he became one of the team of American bridge players, international bridge players. Still is, I guess. He has made appreciable money from his oil dealings, but I felt that it was a shame that a guy who got his Ph.D. from Schwinger and was as smart as he was got out of physics. I don't think it was the association with me that did it. He knew Murray Gell-Mann who'd been at Yale also, and I guess he figured that he was not as smart as Gell-Mann. It probably was true, but Gell-Mann was smarter than 99.9 percent of all physicists.

Anyway, Malcolm left TI, and while we were writing these papers and doing all this, I was asked to be the head of a very small group, practically just me but with a technician and some other help, to work on field-effect transistors. Most of the other group was working on silicon grown-junction transistors. There's an interesting story to tell about that. I built up some equipment and got some furnaces, and began working, making field-effect transistors which had been invented primarily by Shockley at Bell Labs. We were able to make some. But we didn't have very high resolution techniques of making very small regions and junctions Although I have several patents in the general area, I feel that the decision to work on that, which was made by Teal and perhaps higher-ups, was premature. It was only later that metal oxide, silicon junction, MOS and MOSFETs and all that came along, when they had much better fabrication techniques. We did make ones with pretty high transconductance, and they worked pretty well, but it clearly was not a competitive technology at that point, compared with the other possibilities.

### Grown-Junction and Silicon Transistors

Macdonald:

While I was working on field-effect transistors, which was almost entirely experimental, the other group was working on grown-junction transistors. There were two kinds of transistors, both germanium and silicon ones. TI was making transistors at that point, selling primarily germanium ones, which had the difficulty that they weren't operative at reasonably high temperatures. This made them no good for the military. It was clear that one needed to make silicon transistors. As I've already said, it was too early for silicon field-effect transistors, in my opinion. So the obvious thing was to make grown-junction p-n-p and n-p-n silicon transistors. There were two or three ways of doing this: by melt growth and by growing the crystals and doping them in various ways. At this point I couldn't tell you the distinctions in great detail, but any history of semiconductors will make clear the various things that were going on at that time.

The story that was of interest — and I can tell it pretty much dispassionately because I was on the side watching what the other people were doing, and was carrying on with trying to make field effects work and not working much (other than maybe giving a little advice now and then) in the grown-junction stuff. As you know, Teal had a lot of experience in growing crystals, and at one point the technique of pulling crystals out of a melt was known as the Teal-Little method. That was what Bell Labs tried to get people to call it. Little and Teal were then at Bell Labs. Dr. Little eventually went to IBM. It was really the Charlroffsky method, to which they added a few bells and whistles, but not very much in my opinion. So Teal certainly knew a good deal about that technique. Chemical engineers such as Willis Adcock, who was in the group, were able to build up equipment to do this.

It turned out that by both good luck and good planning and I don't know what else, TI followed a line that was not of much popularity by RCA, Philco, GE, and others — particularly their line of grown-junction transistors. Other companies had tried to do this and had not succeeded very well, so they were stuck with the kind of fused-junction transistors that other people were making, and they weren't working very well for silicon. So this story of some interest was that sometime after TI had been able to make a few working models of silicon transistors, Teal and many other people were invited to give talks in the area of transistors at a meeting sponsored by the military, possibly the Air Force. He went and the speaker before him talked about silicon transistors, described the method they were using, and indicated it wasn't good and that it was going to be a good many years before silicon grown-junction ones would come along. Then Teal spoke and he brought some apparatus with him. What it was was a simple record player with a speaker and a very small amplifier, and the main part of the amplifier involved a transistor. He put it on the end of a wand with a wire going back to the amplifier, and he said, "Well, I want to demonstrate some things to you here. Here is a germanium transistor." He had a bunsen burner and some oil that was heated quite hot, and he took the wand with the transistor and stuck it into the oil, and one could hear the music gradually dying out. He said, "we've been working on grown-junction silicon transistors for some time. We've made a good deal of progress, and here's what they do." He stuck it in, and it kept on playing with no change at all.

Nebeker:

That must have been quite a sensation.

Macdonald:

That was quite a sensation. [Chuckling]

Nebeker:

Were you at that meeting?

Macdonald:

No. I wish I had been. I wasn't working in the area. Teal was there and maybe another person from the group was there, but probably not.

Nebeker:

That was the big push at that time, to get these transistors that could stand higher temperatures?

Macdonald:

Yes. They were needed by the military. They had to go higher than germanium would go. Even germanium would get too hot for some consumer electronics, such as in the middle of a radio or something. So there was a big push for that. The military was supporting a lot of this, and so they redoubled their support of TI.

Nebeker:

TI was getting a lot of military support?

Macdonald:

They were getting support, and they got more as time progressed. TI therefore was able to come out with a silicon transistor and sell it to everybody. I won't say it was a monopoly, but theirs was the main product for several years. This was the base of their very rapid growth in transistors and was very important to their future. They were able to capitalize on that.

### Gordon Teal

Nebeker:

Macdonald:

[Chuckling] All right.

Macdonald:

Gordon Teal is a very nice person. Even in those days, when he was in his forties, he was in a sense very slow. Everyone wanted to complete his sentences. I'm particularly fast, and this was a great travail for me. I'm going to try to be honest here, and being honest comes in conflict with being nice, I'm afraid. And yet, I guess it may show me in a somewhat less perfect light, but I prefer honesty to being a nice guy. I'm going to tell you my own feelings about the situation. I got a small amount of help in the early days on the field-effect work from Gordon Teal. He would make a suggestion about where I might be able to buy a furnace or something like that, once in a while. But he was of very little general help. He sat in his office, and I guess he pushed papers around. Perhaps he was a little closer to the other group. Certainly in terms of the crystal-growing, I'm sure he was.

Over the years I reached the conclusion that Gordon Teal was an excellent technician. Although he had a Ph.D., he was not, to my mind, what I consider a research scientist. He has a number of patents, and he certainly did a lot of things at Bell Labs in various materials-related areas. I guess what I'm saying is that I think almost all those were manipulations of materials and things that I believe an excellent technician could do. When he came to TI, he wasn't as involved with that aspect. I don't think he ever got a patent at TI, or certainly very few. He was more of a pure administrator, and I want to say that I think that he did a pretty good job of building up the research lab in the early days when there were only five or six of us. There were several people that he got, and I'll probably include myself and Malcolm Brachman, who were good, hard-working, and knew what was going on. Mort Jones and Willis Adcock were excellent, too. Mort Jones was a chemistry Ph.D. from Linus Pauling, who, unfortunately didn't live up to the promise that he might have. But he still became head of the materials research lab at TI and was not, himself, too bad an administrator. He's now retired from TI.

Yet he did hire some good people. Once the rest of us were there, we could help with the hiring, and I remember many times going out to dinner with Gordon and Lyda, his wife, and me and my wife, and a candidate. We were able to hire some more very good people that way. But that's hardly enough to make you a good administrator. I guess I feel that Gordon was not himself enough of a scientist anymore to make any important contributions beyond the crystal-growing stuff. I may be doing him an injustice here. Perhaps I wasn't close enough to it to know, but certainly while we were a small group and I reported to him, these are my feelings. To be brutally honest, I know that there was a time when he was making a big push to get John Bardeen to nominate him for some important things, the Nobel Prize, I think, but some others as well. I'm not sure just what Bardeen did, but none of it came true. To my mind, Teal did not deserve for it to come true. He did become a member of the Patent Hall of Fame, because of his 50 or more patents in materials science areas. I don't want to throw any cold water on that at all. Obviously he did do some excellent things.

Nebeker:

But when he came to TI he didn't do much?

Macdonald:

He didn't do lab work at all himself at any time that I saw. He may have in those early days when we were in the first big bull pen, everybody all together in a bull pen except for Teal, who had a small office with a door on it. We were sitting at desks next to each other in a big bull pen with equipment there, too. I had a good chance to see what he did and didn't do, and he was in the office most of the time. I know he didn't do much to help me, and I don't think he did too much, except conceivably he did do something on growing crystals and helping Willis Adcock make progress on that. Willis Adcock who now also has one of these professorships that I spoke of at the University of Texas, would be a good one to get more background on Teal. In fact he told me a year or two ago that he was thinking of and probably going to start writing a history of research at TI, in which case he would be very much worth your while following up on.

Nebeker:

I want to.

Macdonald:

At one time they sent Gordon Teal to Europe to be a European representative of TI, to go to meetings and things like that. They helped get him a job at the Bureau of Standards. I know very little about what he did there. Perhaps he was able to get into the materials areas. But I personally doubt a) that he did a lot of lab work or experimental work; b) he might have been administrator of a group. I have no experience as to how well or badly he did on that.

Nebeker:

Are your views of Teal as an administrator shared by other people there at TI?

Macdonald:

Yes. I think there's no question of that. If you talked with Willis Adcock, he might not be as strong in his opinion as I am. He reported to Teal for perhaps not quite as long as I did, but still for quite some time. He also might not be as forthcoming. Also he might have entirely different views, but I doubt it. If you really wanted to talk to somebody who has a very, very poor opinion of Teal, you should talk to Mark Shepherd, who until recently was on the board of directors of Texas Instruments, and is now also retired as president and CEO. He gets back to Dallas a good deal. He has a home near Santa Fe. Teal was hired by Haggerty. Shepherd had very little patience with Teal. Another person who had a good deal of contact with Teal off and on, but not as close as I did certainly, is Frederick Seitz, who was the president of Rockefeller University. He had been a physicist and head of the Department of Physics at Illinois. He is a very fine, distinguished, solid-state physicist who had written the first main book in the area. Seitz became a member of the board of Texas Instruments in the sixties, and I saw a good deal of him in those days. I'm sure he had some contact with Gordon Teal. He is also a very nice person, nicer than I am, and I don't know to what degree he would be forthcoming on Teal. He's still at Rockefeller. He's retired, but he's recently bought a condominium here in Chapel Hill at a retirement community. So we expect to see him about half of every year.

Teal made some bad decisions as head of the group, and Shepherd was really disgusted with him in meetings. We had a lot of meetings at TI in which we would talk about what research and other things were going on. Teal would sit there and never say a word, and Shepherd was pretty disgusted with that, I'm sure. I have to say that I didn't say very much either, but I was at a lower level, and many scientists only talk when they have something to say. Anyway, I'm not sure that there's much more I can say about Gordon except that eventually he was head of the research group called the Central Research Labs. Before too long, it was split into a number of parts: the Materials Research Lab, the Physics Research Lab, the Geophysics Research Lab, and things like that. At that time I became head of the Physics Research Lab, and I still reported to Gordon Teal. He had these four or five labs under him. Then the head of the geophysics lab left. I think there was some pressure for him to go, but I'm not quite sure. He was moved to another part of TI, where he lasted only a few months and then left. I was given the job of running that lab as well as the Physics Research Lab for a year or more.

Nebeker:

What was that lab?

Macdonald:

The Geophysics Research Lab. I found it quite interesting. But it was fairly tough to do both jobs.

### Pat Haggerty

Nebeker:

Macdonald:

Well, in a way that's a harder one, to try to be very honest. Pat Haggerty was a genius; there's no doubt about it. But like everyone, he had a few feet of clay, which I'll be happy to try to say something about as well as the other end. He was able to motivate people exceptionally well. He was exceptionally well-read himself in management areas. He wrote some books and papers and so on in management. Because it was one of his specialties, he was very much concerned with how companies grow, the best way for them to grow, and strategy and tactics and things of that nature. (At one time I was head of the OST, the Office of Strategies and Tactics, near the end of my time at TI, after I was no longer head of the Central Research Labs. But I will still vice president of R&D.) But anyway, back to Pat Haggerty. One sees other people from one's own limited, special point of view. My special point of view was that I was a researcher and head of a research lab and later of the Central Research Lab, which is off in a building by itself, and Pat Haggerty was off in another building running the company. I was disappointed that there were very few times — I could certainly count them on the fingers of one hand — that I remember him ever coming over and talking much to the researchers. This happened less and less as time went on.

But he was a brilliant guy. Before Malcolm Brachman came to TI, he had spent a summer or so teaching a course on theoretical physics at SMU, and Haggerty had been one of his students there. I guess he figured that, being in transistors and all, he needed to know more about this. He was originally an electrical engineer, if I recall correctly. So Malcolm told me that Haggerty was brilliant in this class. Anyway, it was clear that he was on top of things technically and knew what was what, and had very good ideas. But I guess my feeling was that in a position of power he was more self-effacing in the research area than I would have liked. I would have liked to have seen him come over and interact not only with me, but also with scientists lower down. Perhaps he was too busy with other things to hardly ever get to do that.

In terms of organization and administration, I can say that there could hardly have been a better manager than he. He certainly had a fantastic reputation for that. One of his feet of clay, though, I feel, was that he saw his underlings to some degree through rose-tinted glasses. He was not so objective about people he hired or kept on. There were several people that he brought with him when he came from the Navy, and one or two of them became vice presidents, of whom I thought very little. Yet he kept them on and so on, being very nice. There were several others that he inherited that turned out not to be all that good. I can also remember him saying, "Mark Shepherd is the best middle-area manager in the country, and he's really great." Of course Haggerty died early, Mark Shepherd took over, and he ran TI into the ground. When I left, TI was going great. It was the top semiconductor manufacturer in the world when I left in 1974.

But the seeds of decay were already there, and Mark Shepherd, who had been the president at that time under Haggerty, became CEO and became the man who really ran the company with no inputs from Haggerty any more. Haggerty died two or three years after I left. Now TI is maybe number six or seven in the world in terms of the size of its semiconductor activities and has been relatively undistinguished. Its stock has been a great disappointment to me. I don't think there's much cause and effect between my leaving and its going downhill, but I do think that a lot of it has got to be blamed on Mark Shepherd and to some degree on his successor, President Fred Bucy, neither of whom evidently did the right things to keep TI in the forefront. But this was after my time mostly. I guess I feel rather bitter about them letting a great company become a not-so-great company, especially since I still have appreciable TI stock that's gone nowhere.

### Administering Research Projects at TI

Nebeker:

Could you summarize the work you oversaw and took part in yourself in those 20 some years that you were at TI.

Macdonald:

Yes. First of all, let me tell you about my own work because I'm obviously more familiar with it than with anybody else's work. When I went to TI, I did a little bit of experimental work, certainly at the beginning, on the silicon transistor and on a few other things. But particularly when I became head of the Physics Research Lab, it was clear to me that to have an ongoing significant experimental program of my own as well as management responsibilities was not going to work very well for me. But I didn't want to give up research. So because all of my past work had been mixed theoretical and experimental, I was able to decide to switch over and do pretty much theoretical work from then on. And I did do that. Over the years I had a few of the people at TI work with me. First of all, with Malcolm Brachman. We wrote four or five papers together. Then came Tony Barlow, C.A. Barlow, who worked with me in electrochemistry areas. TI was working on fuel cells and things like that, and that gave me enough of a justification to shift to some degree from semiconductors to electrochemistry. Some of the things that I'd done overlapped both areas. Barlow and I wrote some very good papers, one of them 200 pages long, in that area. I also worked with a few other people at TI.

But a lot of my work was all on my own, and if you look through that period of my publication list, you'll find that most of the papers are by me alone. It wasn't until after 1974 that you'll see what I did when I came here to UNC. There I rather quickly got post docs and even one or two undergraduates who wrote papers with me. I only had one graduate student, though, and he and I wrote one paper. Anyway, the rationale for my work at TI was that I did not want to give up research. I enjoyed doing research a lot, and I wanted to keep doing it. So I kept doing it. I consider myself to be a good administrator and well organized. I was able to do my administrative work pretty much in half a day every day, and I would keep the other half pretty much open to write papers, to do theoretical calculations, and to write up results.

Nebeker:

And take phone calls and so on in the other half of the day?

Macdonald:

Nebeker:

How often did you have to have somebody drop some line of research?

Macdonald:

Sometimes when times got very tight, we actually had to let people go from the lab. Once a year we would have evaluations. Those people who were in the lowest part would not be candidates for raises and might even have to go. I had some unhappy times when I had to fire a few people.

Nebeker:

Did you often have to tell someone, stop this line of research because of needing something more useful to TI?

Macdonald:

I was head of the Physics Research Lab where I had such people directly under me for five or six years. While I was doing that, I was of course close to the people, and I would have lunch with them, play chess with them, and have other close contact with them. I think we would never let it go to the point where I would have to do it. Often they would come to me and say this line is petering out, and ask what shall we do. I'd say that we'd better find another, better line. So I don't remember much of having to close down research projects there, unilaterally.

Nebeker:

What about the other side of it, people getting an idea, something they'd really like to pursue, rather than the project that they were given?

Macdonald:

This was called bootlegging, and we did a fair amount of this. We were lucky to be in a building off from the other people in the company. I personally believed in bootlegging. I did some of it myself. But it had to be something that had some hope of being valuable. Even a long shot that might have a big payoff was worth doing, in my opinion.

Nebeker:

When you say "bootlegging," does that mean that the person is doing this sort of on his own time?

Macdonald:

Audio File
MP3 Audio
(154 - macdonald - clip 2.mp3)

Well, he was doing it on his time, but at the lab, not necessarily off at home. He might be supported primarily by another project which he was supposed to be working on, but still putting part of his time into something different. We had some of that, and some of it led to some of our best results. When I became head of the Central Research Labs, and also as vice president for R&D, I had another layer of insulation because there were people who were heads of the labs under me. Even there, I feel that I didn't give as much direction as I might have to people under me, but I did what I could. It's a very hard decision to know how to balance such things. I was not the kind, I guess, who could go in and talk to somebody and immediately see a million possibilities that they ought to be working on. My point was if I'd done a good job in hiring a good person, I wanted it to be his work more than my own work.

Our lab was pretty highly thought of when I was head of either the Physics or the Central Research Labs. You can talk to people who were there at the time and perhaps get some feeling for it. From time to time there were personnel surveys at TI to obtain people's opinion about how things were going in the labs and everything else, not only the Central Research Labs, but other labs as well. Our lab generally got the highest rating of any with respect to personal satisfaction and achievement. This was bad in a way. It meant to the higher-ups that I wasn't being hard enough on the people. You can't win, you see. But from my point of view, I had, I think, very good relationships with everybody. We were a compatible group. We were hard-working. I pushed them, but I pushed them in what I considered to be a reasonable way. I tried to get the best people, and many of them have turned out to be really outstanding people, most of whom are now not at TI anymore.

### Leaving TI

Macdonald:

I became a member of the National Academy of Engineering in 1970, and of the NAS in 1973. I would never have become a member of NAS had I not continued my personal research all along. From that point of view it was a big plus. I think my research itself has certainly been of some value to the scientific community. But it helped me to become a member of the National Academy of Sciences, which I always felt I was damned lucky to get in to, and I might very well not have, but anyway I did. Once you're a member, you have a good deal of scope to go where you want to, particularly in the university sector.

I looked at one or two other companies. Some of them would initially say, "Well, we're really interested in you and you can be a member of the board," but when I would interview with them, that commitment was gone. I was not going to be a member of the board. It was just a come-on by the recruiter.

So I gave up that sort of thought and decided I would go to a university because I knew I could do research there, and I thought I might enjoy teaching. On that score I decided that I wanted to be on the East Coast because I grew up there, but not too far north. Although I'd been at Williams and MIT and Chicago, it was cold up there, and I didn't want to be in such cold weather any more. I didn't want to go too far south because the farther south you went, the worse the quality of the school would be generally. So that meant from Georgia to Virginia perhaps. I did apply to various places, and I had offers, including one from Georgia Tech. They were talking about a big endowed professorship, but it wasn't quite ready when I came, and they said, "we'll take you, and then you can get that later." Though I'd grown up in Atlanta part of my life, I decided there might be other opportunities. I decided that I wanted to be at a public university rather than a private one if I could, primarily because I thought their funding might be more solid. That didn't entirely work out. Duke has done better than UNC in that regard. But the funding didn't matter to me in terms of my own salary because I was able to be in a position to take a cut in salary wherever I went and not worry about it much. UNC offered me a position with their best endowed professorship, the William R. Kenan, Jr. Professorship, and my wife and I decided that Chapel Hill was an exceptionally nice place to live.

So that's how we came to come here. It was all possible because I continued to do research myself, which very few lab directors have been able to do or have wanted to do over the years. My successor at TI, Bob Stratton, whom I hired from England back in 1963, is an exceptionally good theoretical physicist. He wrote two or three papers after he came to TI. But since he became director of the Central Research Labs, he's written very few, if any. He only recently became a member of the National Academy of Engineering and is not a member of the National Academy of Sciences. Part of the reason, I believe, is because he didn't continue to do personal research. People at Bell Labs have trouble doing this, too. For example, the managers there generally don't write many more papers when they become managers. But I was still writing them when I was head of the whole Central Research Labs with a couple of hundred people under me. In a way, I'm proud of that. I feel, in looking back, I might have been able to give a little more direction to people under me, but it was not entirely my style. I might have been able to do it, but it might not have helped many things much.

Macdonald:

Back to the research work that I didn't personally do: Over the years the labs, the Physics Research Lab for example, came up and developed a very sensitive magnetometer using germanium: A resonance magnetometer that was used for some geophysical work. Another important thing was a germanium bolometer, developed by Frank J. Low, who put it to use in infrared astronomy areas. He soon left TI and is now a member of the National Academy of Sciences because of his work in that area, all due to seeing that one could use germanium at low temperatures as a bolometer to measure infrared radiation. Another thing that turned out to be very important was plasma etching techniques in semi-conductors. The person responsible, Al Reinberg, is someone I personally didn't have much to do with in terms of saying, "hey, this is a great project", though I certainly was for it and pushed it. He eventually went to Perkin Elmer and developed this idea even further there.

There were some basic advances, too, that people made in ferromagnetic resonance and other things. Of course, there were some developments in semiconductors. At one point in the very early days I was head of a transistor test lab and Mort Jones was working for me. We had people there from the production line essentially building transistors. We were doing statistics and trying to find ways to improve the techniques and so on. That was probably around 1955, 1956 or 1957. That went on for some time, and I think turned out to be a very valuable activity. In those days I also suggested silicon oxidation as an insulator for building silicon devices, and we put together a ozonator to try to do that. Unfortunately, we didn't push it, didn't have the right people working on it, and other people eventually came up with the same idea and pushed it into what became integrated circuit technology. Another thing that we tried to do: I had them build a very fast, rotating mirror to measure the response of semiconductors at very short times. It was powered by high air pressure, and rotated very quickly. You had to worry that it didn't explode and so on. It was of some value for a while, but it didn't turn out to be of any great, long-term importance.

It's a little hard to cast my mind back to the period and remember other things that the other people did. We had an excellent analytical laboratory which was able to make a wide variety of measurements on semiconductors of all different kinds, with good people in it. They did what they needed to do to help the rest of the company. We had an excellent Materials Research Lab, under Mort Jones mostly, and it was able to develop some materials, and worked on gallium arsenide a good deal and on other things. We had some other labs, and we did quite a lot on infrared detectors of one kind or another for the military. I pushed that quite a bit. That work shifted over to the equipment division in a natural way because they were using these things for detection in military equipment, infrared seekers and all that. Then we had a Systems Research Lab that developed certain things. We were always working on new business possibilities for the company. Our labs, let's say the Systems Lab for example, would look into something like computer-aided education. We'd do a big, long report on that, and I remember working quite a lot on that myself. Another topic, we worked on was cable TV, and we considered whether TI should get into the area. We worked on a number of other possible things: electronic cameras and various other electronic possibilities.

### Unsuccessful Projects

Nebeker:

I have a number of questions about your work at that time and also about the research more generally then.

Macdonald:

Go ahead. I've lost the train of thought on some of this research that I was going to say, but maybe it will come back later.

Nebeker:

You were going to tell me about an aborted project?

Macdonald:

### Small Scale Computers

Macdonald:

The next important area I want to talk about was the area not of big computers, not the advanced scientific computers, but of small computers. I pushed very hard to get TI into both software and small computers. It wasn't entirely me that did it, I'm sure, but we were able eventually to get TI started on a home computer. I also pushed and had a little contact with the development of the pocket calculator. I remember that at a NAE annual meeting I attended, a newly-elected member was Grace Hopper, Admiral Hopper eventually. This was probably in 1971 or 1972. She was by herself there with all these men. So I went over and had a chat with her and found her to be exceedingly interesting and nice. I invited her down to TI to give a talk, and she did. Part of the talk was the future of computing, and she made a very strong point at that time that a lot of the future was going to be small distributed computing. That surely has been borne out already very strongly. I believed her, and I strongly pushed this at TI. But the development of the home computer at TI was out of my hands and was done in another part of the company, and I had very little to do with it. It shared a common characteristic at TI, namely the hubris of the top managers who thought that anything that TI did, and TI's way, was always the best. So they developed their own CPU for the computer, which made for a very distinctive home computer. We sold a few, a fair number to begin with. But it was rapidly pushed out by the Apple and the IBM PC. It was a breakthrough in a way. It was perhaps the first 16 bit home computer, whereas the others were 8 bit at that time. But it was too different, and they didn't make any concessions to compatibility. Eventually they switched over and decided to make their computer like the others. So they did make one like the IBM PC, but unfortunately it still was not entirely compatible. That was the crowning blow, that it was not compatible. It couldn't run most — if any — of the programs, so....

Nebeker:

That they hadn't foreseen how important that was?

Macdonald:

I presume that was it. They had to do it TI's way. One of the people who'd been a lab director under me in the Central Research Lab, head of the Systems Lab, eventually became head of that computer activity. He was a very good guy, very smart, hard-working. Eventually, unfortunately, home computers didn't sell enough; they let the area decay, and he left the company. I'm not sure what's happened to him. So that's a few illustrations.

### Basic and Applied Research

Nebeker:

Can you summarize from your many years of experience how direction gets imparted to all the work at TI? It's not like an academic environment where so much of it is from the individual. I don't know how large a part of it is from contract work, how much is from the top management, and so on.

Macdonald:

When I was head of the Central Research Labs, we had about 30 percent perhaps supported by outside government contracts. We were pleased to have roughly that percent. We wouldn't have wanted 50 or 60 percent or higher because it would have hampered us in the sort of things we could work on. On things that we wanted to work on and could get government support for, that was fine. I'm not sure of the percentages anymore, but we had both applied and basic work. Some of the applied work would be very closely related to other work in other research laboratories, in particular the semiconductor or the equipment research laboratories. Some work would be not so close but still very much in areas of probable interest to TI. Then the basic research would be somewhat long-term, on various areas of physics and materials and on improving techniques.

Nebeker:

Did you make that distinction at that time?

Macdonald:

Yes. We felt that it was very important to be able to make clear the distinction between basic and applied research, and we had definitions that everybody knew. As time went on, we were pushed by higher management to have more applied work. By the time I left, there was much less basic and much more applied research, and probably more government support. And very close contact with the divisions. There's nothing wrong with that. One of the best ways is to get some of their people over to your place and have them work with your people, and vice versa. Certainly some of that was happening. But the longer-term basic work was primarily being pushed out, and after I left there the number of published papers in the literature from TI decreased greatly.

Nebeker:

Were they done by other people besides yourself?

Macdonald:

Oh, yes! In the early days we put together a thick book of papers by various people in the laboratory. Certainly there were several by me in there, including the one on linear systems already mentioned. But there were lots of other people, particularly from the physics and the materials research labs whose papers appeared in that book. They were continuing to publish.

Nebeker:

But it does seem counter to the interests of business to be having people do research and putting it in the public domain.

Macdonald:

You could look at the earlier days of Bell Labs where they published extensively and where their people won four or five Nobel Prizes over the years. Anything that would seem to be patentable, we would have to discuss with the patent people and we would delay publication. But a lot of the longer-term, more basic work was obviously not patentable. While it had to go up the chain of command to some degree to be allowed to be submitted for publication, that generally was not much of an impediment. The rationale was to learn more about the area, the materials, the techniques, and how to improve them. Maybe you'll find something entirely new. Maybe it will be of great importance. That certainly paid off over the years, in my opinion, at Bell Labs, and to slightly less of a degree at IBM. I was just reading an article in Physics Today about IBM's lab and the fact they were probably going to cut it by a factor of two and make what was left primarily applied, how many people they were losing, and what the morale is like. I was happy that I was never involved with that sort of a situation. We did have tight times. And toward the end of my time at TI there was certainly more and more pressure to have more applied and less basic research, which I still think is a bad technique over the long term. Maybe we had a little too much basic research; I don't know. But you need some if only to be able to read what the rest of the world is doing and understand it, and be able to apply it to your own stuff. People who are doing applied work may miss things like that because they're so much involved in short-term work.

Nebeker:

It would also be valuable to have in the same organization people at all points in that spectrum.

Macdonald:

Yes.

Nebeker:

Like Bell Labs benefits more from that.

Macdonald:

Yes. Some of my philosophy of research support is in this very short talk I gave when I got the Pake Prize, which is partly for management of research.

Nebeker:

Who gives the prize?

Macdonald:

That was the American Physical Society. But the Edison Award of the IEEE is in some sense partly the same thing. It may involve management, too.

### Writings, Research and Papers

Nebeker:

I was just glancing over the titles of some of your publications in those years, and a lot of it seems, quite basic.

Macdonald:

Yes, a lot of it was. No question of that. I did write a paper or two on semiconductor diodes and their response, and on transistor transient response, and things of that nature. Once I got over into electrochemistry, and a good deal of that was pretty basic, but it still had implications for batteries and fuel cells, which are practical things.

Nebeker:

In your own case, were these topics that you worked on usually suggested by the previous work? Was it a fairly continuous work coming out of different things that you were involved with?

Macdonald:

I find it very hard to answer that question. If we talk about my own personal research work, I've always had more ideas than I personally have been able to work on. I have to have ideas that are sufficiently simple that I personally am able to do something with them. I have a few that are too hard for me, and I obviously have to let them go. But I've always had a number of good ideas of things that I think are worth analyzing and writing up. As time has gone on, I find that although I'm still somewhat creative, I'm less creative and less capable than I used to be. I don't have as many good ideas at 70 as I had at 40. I'm not as good a manipulator of concepts and mathematics as I was in those days. But I have a little more experience, and I'm still writing four or five papers a year. I now tend to collaborate a little more with other people. I'm now doing a paper with a guy in Belgium, for example; we communicate by e-mail. I'm working with one or two of the people in my department. And I'm doing two or three things on my own.

Nebeker:

How many of these papers you wrote were related to work being done by other people at TI?

Macdonald:

Very, very few. I tend to be a self-starter, and I prefer to work on my ideas. If somebody else gave me a good idea, I'd be happy to work with them on it. But usually it was my idea. I won't say my ideas were the greatest, but I was lucky to have enough of the kind of ideas that I personally was smart enough to do something on.

Nebeker:

Ernst Weber, for example, told me that a lot of his research in the thirties came out of his graduate student projects. One could imagine a manager at TI getting a lot of his ideas that way.

Macdonald:

I did, however, have enough money in my contracts to support undergraduates, and several of my papers were co-authored with undergraduates, three or four at least. I was able to get some very smart undergraduates to work with me for a year or so, and they contributed considerably to the work over the years. Even among the post docs I had, who've gone on to other things, only one of them, Don Franceschetti, has stayed in the field, and he eventually became vice president or dean at Memphis State University, so he is doing very little research these days. Although he's still in the area, and we have written one paper together in the last four or five years, I doubt that we'll write any more. As for projects at TI coming from other parts of TI, none of them that came in were ones that immediately made it obvious that it was something I personally ought to work on.

Nebeker:

Can you characterize the topics in solid-state physics, the topics, that you were most involved in?

Macdonald:

Originally I was interested in transistors and in diodes and their response. That led me into the electrochemical double layer. That's some work primarily that Barlow and I did at TI. We wrote this big review paper on that, and eight or ten other papers in that area. I gave a keynote address in Australia on some of that work. It wasn't immediately obvious that this was ever going to be of any great interest to TI, although later TI did in fact get into electrochemistry. So it could have been of value, too.

Nebeker:

You told me on the phone that many or most of your papers are at the University of Texas Archives.

Macdonald:

Yes. About 25 boxes or so.

Nebeker:

Those cover the TI years as well?

Macdonald:

Yes. They go from some very early notebooks and other things from my grade school, through high school, and there are a good many Williams examination papers, essays and stories. I won second prize in the freshman story-writing contest at Williams, and I took a course on creative writing instead of regular English. So I had to write a lot of stories, and they're there. And poems and notes from courses at Williams. Notes from courses at MIT. A good deal of stuff from my various papers in the early days; I kept a lot then. Lists of people who requested reprints. Some of the original handwritten work of either analysis or the writing of the paper, or both.

Nebeker:

Macdonald:

Nebeker:

Are these records accessible?

Macdonald:

It is definitely accessible, and I told Willis Adcock, when he asked me questions about what he was planning to do for a history of research at TI, that it was right there where he was, and that he could look it up. Whether he has or not, I don't know. I haven't talked to him since. But it is definitely freely available. Although some of it may show me in a not perfect light, we all have warts, and I saw no reason to try to hide some of mine.

### Jay Forrester

Nebeker:

Can you give me your impressions of Jay Forrester? You're one of the few people around who've worked with him.

Macdonald:

One of them committed suicide later. He came to Texas, worked for LTV, and committed suicide — at the time I was in Texas, as a matter of fact. Anyway, I learned a great deal from Jay Forrester, in terms of management of research, and nearly all of it was negative. I learned what not to do. I'm trying to be honest here. I would not like to make him unhappy, though he perhaps knows his own strengths and weaknesses anyway, and mine is my own view of it, which may or may not be right. But I did have the chance of working with him reasonably closely for a year. Let me just give you one example. He would have high-ups come in from outside from RCA, or from the military — who were supporting the contract. He'd bring them into the lab and ask you to tell them what was going on. But he'd never introduce you to them. Never say anything about you. He was a cold individual, perhaps frightened, I don't know. But very cold in general. A smart guy, hard-working, probably a good administrator though I didn't have as close contact on that aspect of it as I had with Teal. I think a good electrical engineer. Later I had contact with him. He wanted to know what notes I had that would help him win the patent suit with the guy at RCA [Editor's note: Jan Rajchman] who had developed some magnetic storage and other storage activities there.

Nebeker:

Was he helpful to you in the year that you worked with him?

Macdonald:

As you know, Jay Forrester to some degree invented the magnetic core memory, and he had a patent interference with this guy at RCA. Jay wanted more information to help establish his rights. I was not able to help him on that. I couldn't find anything among my papers. I did have a few things from the lab there, and they might very well be in my Texas papers now, a few reports and so on. But nothing that was germane to that particular area. Personally I felt that An Wang of Wang Labs, who had been at Harvard, had invented the main ideas. In any case, I was unhappy to some degree that Jay Forrester got millions of dollars from IBM for this. I thought Wang deserved something at least. But I was not close to that, and he did it independent of me. Although when I did my research on ferromagnetic resonance when I came back to this country, I think he would have been glad to have me join the group because they were still working on magnetic storage at that time. But I'd had enough of that. When I became a director of research, I could look back on the way he had run things and try not to do it the same way myself.

I certainly don't want to take anything away from his a) being, I think, a very good administrator, and b) being a good electrical engineer and having good ideas, particularly his development of the core-type magnetic memory. I think it was excellent. He definitely had a very big hand in it himself and deserved to get something out of it. I was not involved with it at all. That came mostly after I'd left. He also got heavily involved with the growth of the world and the economy and so on [Editor's Note: Club of Rome Report, The Limits of Growth] A couple of people who were at Dartmouth wrote a book about the economy and world growth. I wrote an article about this. Forrester was involved in this and was behind it, and he pushed it very hard and made a whole area at MIT in which he developed a new kind of programming for it. In later days that was all that he did pretty much, I think. I can't quite remember the title. Anyway, I wrote a paper about that blasting it as being computer manipulation I felt it was virtually meaningless because of the lack of adequate data. It may be that he and other people over the years have been able to improve it some, and it may indeed have some relevance. But even so, it's so many variables, so many interactions, so many nonlinearities that I still highly suspect anything that's that total and complete. That's my only other contact with things that he has done.

### Involvement in IEEE

Nebeker:

I see here that you joined IRE.

Macdonald:

While I was at MIT.

Nebeker:

Macdonald:

I joined the IRE back when I first went into electronics and so on at MIT. I'd say in 1943, or 1944.

Nebeker:

What was your involvement with IRE in the early years?

Macdonald:

You'll find it all on the list in terms of the committees and other things that I've done

Nebeker:

I see on your resume, things like vice chairman of the program committee and so on.

Macdonald:

As you go through that list you will see that it lists everything I did that I can remember with IRE and IEEE. One of the things that I got quite involved in in the early days was audio, and I was quite active in the Professional Group on Audio, PGA. I was invited to be head of it at one point, but I turned it down because I felt that it was only a sideline of mine — a hobby. I did work on audio at home at night, built a lot of equipment, and invented a few new things, had a few patents in that area. But I felt that it was a sideline and that it was therefore not appropriate for me to be head of it. But I was on its executive committee and various committees for some years. I guess Ed David took a different point of view. He also had been in audio, and he was out of it, but he did accept to be head of the PGA for a while. That was part of my involvement. But in a way it had very little to do with TI, though I did develop a very fine adjustable band-pass filter, and there was an article in Electronics describing it. I tried to get TI to build it commercially, but they never did, though I did have some contact from other people outside that wanted to do so. It helped the development of the field, but that's about all it did. I still have it down in my basement, as a matter of fact.

Nebeker:

I was thinking not only about the committees and editorial boards but what you valued about IRE and IEEE. Some publications I imagine at times were valuable.

Macdonald:

Yes. I certainly kept very close watch on them, and I published in Proceedings and in the PGA and other journals over the years quite a bit. I won a prize for one of my papers in the IRE publication, one about a new kind of amplifier. I read these faithfully, kept up with what was going on, and did get stimulus from that. When you asked me earlier about stimulus from other people in other areas of TI, I got far more stimulus from my own work and interests, and from reading many different journals, of which the IEEE ones were very important. But physics journals and others were, important too. I valued the existence of the professional organization and what it was doing, and the chance to play a small role in some of its activities. I played some role in the local chapter for some years. I was on some of the selection committees for various awards of the IEEE and the IRE. I thought that it filled a large void in the scientific arena that was quite important, and it was very important that it filled it.

### Biophysics and Biomedical Engineering

Nebeker:

I also wanted to ask about your connection with biophysics and biomedical engineering. I see that you were a professor in that capacity.

Macdonald:

Yes. Well, it's not as good as it sounds. But it was of some value and interest. Soon after I got to TI, I was asked by some people at the medical school to help them with various problems. Because of that they made me an adjunct professor, an assistant professor, and eventually I became a full professor. But all I ever did was work with them on their problems, some of them analytical and some of them experimental, in the areas of cardiology and radioactive tracers. Well, I guess I did a little on lung function and cardiology. But the radioactive tracer, which you can find traces of in my papers here, had to do with the injection of radioactive tracer in one of the body liquid volumes, and it then diffuses into other volumes, particularly in people that have an acites, which is a sort of accumulation of liquid in various places. You want to know how long a tracer atom takes to get from one place to another, the reaction rates, the diffusion constant, and how much is in various places. So what you do is to inject it in there, and then as time goes on, you take samples from various parts of the body. You thus get growth and decay curves. This is a more or less linear system, which I was to some degree an expert on. While the solutions could have been computed directly, and would be these days, we didn't have very much computing power then. And to solve the matrix and other equations that would have been necessary, it would have been a big labor. I've forgotten exactly how it came about, but when they showed me this stuff and asked, what they could do with it, after doing some analytical work, I decided that perhaps we could produce an analog computer solution. So we built one.

Nebeker:

Macdonald:

In the late fifties. You can find it in one of my papers because we wrote up something on it. I got one of the people at TI, who was from geophysics but a really great experimentalist, to do one aspect of it. We built a big rack of equipment which would take a graph of this response, and you'd shine a light through it, and we'd move the spot of light around. From this you would get an input which would go into the electronics. We used vacuum tube operational amplifiers in those days. From the whole system you were able to solve the problem. In other words, from the graphs we would get new material that would solve this linear problem.

Nebeker:

That would explain the concentrations?

Macdonald:

Right. In various different places. It depended on how you figured it was connected together. But it essentially solved the problem. We transferred the equipment over to the medical school, and it may still be there, but I'm sure it isn't still being used.

Nebeker:

But it worked?

Macdonald:

It worked. The optics part was done by the guy from the geophysics group. He eventually left TI and joined another company, that did optical character recognition and got fairly rich from that. I had a summer student working with me, too, when we built that equipment. That was done, I guess, on TI time. I can't remember who paid for the equipment. Maybe they did, I expect. But it was then turned over to them. Except for building that one piece of equipment, this faculty position involved talking with them and perhaps now and again giving them some advice. A lot of it was largely honorary, I'd say. They kept me on for years when I didn't do very much of anything.

### Impedance Spectrosopy

Nebeker:

You wanted to tell me something more about the last 20 years.

Macdonald:

Yes. Back when we were talking about the work at TI, I didn't mention one area (the electrochemical double layer) that got me into the question of pressure effects in liquids. I wrote two or three papers in that area on isothermal equations of state, which certainly had very little to do with TI. Still it was of some interest. In fact I wrote a couple of papers in the Review of Modern Physics in that area. When I came here, there were aspects, primarily of electrochemistry and the double layer and the response of systems, that I wanted to continue to explore. It has to do with the small signal a.c. response of materials, which involves usually both the material and the electrode. The material could be a semiconductor or amorphous glass or a liquid electrolyte or a solid electrolyte, or a superionic sort of electrolyte. In any of those you have charge carriers which are either electrons, holes, ions or vacancies. By measuring the frequency response over a very wide range, from, say, 10^-3 or 10^-4 Hz up to 10^5 or 10^6 or10^7 hertz, one can learn a lot about the various processes that are going on in the material and in the electrodes. This has come to be called Impedance Spectroscopy. While a lot of it is oriented toward electrochemistry, it also applies to dielectric measurements, semiconductor measurements of this kind, and transient response.

I somehow got into that area, wrote some of the first theoretical papers, and solved some of the theoretical problems, but not all of them by any means. That got me into the area, and I kept on working and doing theoretical work. I got interested in data analysis, which I'd really been interested in back from the time of the high pressure work at TI, where I developed some new tools of doing least squares, particularly least squares fitting when the X and the Y variables both have errors. That was developed at that time, and it turned out to be useful in other things later. But it got me interested in data analysis, and I wrote several papers in that area. In recent years I've written some more. Impedance Spectroscopy is nearly pure electrical engineering, although the actual treatment of the material in the cell may be electrochemistry and the interpretation of results may be physics. I got interested in both advancing the area in theoretical ways by solving some of the theoretical problems associated with it, and also with analyzing the data.

Some years ago I was asked to be involved with co-editing a book on this subject. The guy who got us going was in Australia, he had some problems, and he never followed through. But a couple of us went to Australia, and I spent three months starting this thing. The way it finally turned out was that I became the editor of the book, which is in fact the first book in the area. It's called Impedance Spectroscopy. I am also one of the main contributors to the book. There are about ten contributors. Published by Wiley in 1987. There are a lot more than a thousand copies around the world at this point. It's still the only main book in this area, and anybody working in the field has a copy of it, or certainly should have.

Nebeker:

I'm pleased to see you have a section on history.

Macdonald:

[Chuckling] Yes. A little bit. I also was asked to write a section in the Academic Press Encyclopedia of Science and Technology on Impedance Spectroscopy. First it was for the yearbook of 1990 or 1991. Then I think they put it into their new edition in the encyclopedia itself. That had even more history in it. That particular paper that was in that book you are holding was taken up, and people wanted me to contribute a paper in the area of biochemistry, bioelectrolytes in cells and membranes, with emphasis on Impedance Spectroscopy. I decided I didn't have time to do that, but told them they could reprint this one if they wanted. So they did, and I made a few changes and improvements in it. That came out in the Journal of Biomedical Engineering a year or so ago. Same paper more or less. I've had a fair number of reprints for that. The point is the subject applies not only to physical science, but to medicine as well. I was pleased about that. Over the years I've had a number of post docs, and we worked on a variety of things, not all by any means in electrochemistry, but a lot of them in that area.

One of the things we did was to try to develop a technique for analyzing this data as well as possible. So over the years I and my colleagues of this kind, including undergraduates, developed a big computer program, which is now called LEVM, L-E-V-M, which stands for Levenberg-Marquardt, who are the two people who did some of the particularly useful kind of solution of least squares problems. The last M could also stand for Macdonald if one wanted it to. Anyway, this is a program to do fitting of Impedance Spectroscopy data. It's a large program which over the years has grown. I've made a number of revisions to it. It's now at number six. The total number of files and so on amounts to nearly 2 megabytes. The manual is about 110 pages long. I decided from the beginning that since the government helped support the work, post docs, undergraduates, and one graduate student, I didn't want to make any money personally out of it at all. So I made it freely available at cost, which turned out to be of the order of $50. In recent years I've had the department issue it to people who wanted it at$50. The $50 includes free updates for a while. I have not made any money out of it, and I guess there are at least a couple hundred copies of it around the world, in use. That's nice. When I saw myself approaching 70 and this thing still taking an appreciable amount of my time, (to make copies of the two disks to hold the program, and get them ready for the department to send out, and correspond with people, and so on), I decided I was obviously not going to be able to do that forever, and yet it was a very good program with very wide applicability, and I didn't want to see it go down the drain. So, in order for it to continue, I approached Schlumberger. They are one of the main companies making measuring equipment in this area. It's originally a French company. They bought an English company called Solartron, who made the original equipment in this area, and they're still making it in England, but it's part of Schlumberger now. I approached them at a scientific meeting. They always had people there demonstrating their equipment. I was an invited keynote speaker at both the first and the second conference on Electrochemical Impedance Spectroscopy. The last one of these was held last summer in Santa Barbara and the first in France. At Santa Barbara I saw these people and said, "Are you interested in this program?" They said, "Yes. We're using it already with our equipment, and it helps sell the equipment a great deal." It isn't the only one in the field, but there's really only one other one, and mine is far more general, and far more powerful, and better in many ways than the other one. So I said, "Okay. I will do one more revision of this and try to put it in as perfect order as I can, and you all take it over and distribute it at as low a cost as you can. And you give my department some money." Which turned out to be$10,000. In fact, together with the money from my Edison Award, which was $10,000 and which I gave to the Department of Physics and Astronomy at UNC, a trust called the J.R. and M.T. Macdonald Trust was set up, which accumulates money to send graduate students to meetings. That way I kept it from going through me and having to pay taxes on it. Nebeker: That was a very nice thing for the university. Macdonald: Yes. I thought it was nice, and I was pleased that Schlumberger would come up with this$10,000. They are now just in the process of copyrighting the manual, which I never bothered to do, and getting ready to distribute it. Unfortunately, they're probably going to have to charge $150, which still, for that much material, is not too bad. Unlike most programs, I've always distributed this one with full source code, so people can make their own changes. The$150 is if they include the source code. The cost is probably $100 or so if they don't include the source code. Many of the users don't need to change it, so they would not have to pay$150. Here is the manual of that program. You can page through it quickly. Over the years it turned out that two graduate students in other places have produced plotting programs that they sent me, which that they were willing for me to incorporate. The last one was particularly good: 2-D and 3-D color plotting programs from an Indian in Iowa, Dr. S. Amarasinghe. It was a very large addition to the total program. That helped make it as complete as it now is. It can fit tens, even hundreds of thousands of different circuits, and a number of theoretical solutions which can't even be represented by equivalent circuits. It has many different kinds of important weighting possibilities. It can solve these problems very, very quickly and well. It analyzes the data and tells you what's there and something about it. You can learn a great deal from such analysis. I recently gave a talk at the University of Florida in the Department of Chemical Engineering, in Gainesville, and a similar talk in Charlottesville at the University of Virginia in the Materials Science Department. It was on Impedance Spectroscopy, what it can do, and what the possibilities and problems of it are. My first slide is this one, and you can read it out if you wish.

Nebeker:

"Socrates: 'The unexamined life is not worth living.' Macdonald: 'An unanalyzed data set is not worth generating.'"

Macdonald:

So a lot of my work and all this program is involved in analyzing the data. Since I'm the editor of this book, I'm to some degree a father of this whole field. I also did some of the earliest theoretical work in it.

Nebeker:

What are the main applications, what areas?

Macdonald:

Let me show you one area. Materials, corrosion. It's used a great deal in that. It's used in real-time testing of paints. It's used in testing batteries. It can be used to look at very small surface defects much smaller than you can see with almost any kind of a microscope except an atomic resolution one.

Nebeker:

What you're doing is measuring the impedance at different frequencies?

Macdonald:

Yes. Impedance or admittance or dielectric constant. When you do that, you get curves. Here's the simplest example. It shows a circuit, and it shows what you see, a 3-D plot, which I also developed for the first time: 3-D plots, projections and various planes. But the actual data is more complicated in a way than that. Here's an illustration of what's called complex nonlinear least squares fitting, which is what the program does. It shows a circuit where you measure the individual pieces of it, and then you measure its total impedance over a wide range of frequencies and with the usual equipment. Then you analyze it with complex nonlinear least squares, preferably using LEVM, and down below you see the shapes that you get, which don't show very much structure. This is an impedance one and the other is admittance. The figures on these circuits are the values that you get from the complex nonlinear least squares fitting. You see they're very close, and they also include standard deviations which give you some idea of how important each circuit element is. Thus, this figure shows the result of a test of the accuracy and appropriateness of complex nonlinear least squares fitting and demonstrates that many parameters can be well estimated from rather featureless data.