Oral-History:Robert N. Hall

About Robert N. Hall

Robert N. Hall became interested in electricity at an early age when his uncle, a "career inventor," took him to a technical fair in New Haven, Connecticut, and he became fascinated with how things worked. He received his bachelors degree in physics from Caltech and was hired by General Electric as a test engineer at Schenectady. Hall returned to Caltech for his doctorate in nuclear physics (1948), then went back to GE. He was involved in many projects during his career at GE, including work with purifying germanium leading to Shockley-Read-Hall recombination, and involvement with PIN rectifiers, transistors, nuclear detectors and the semiconductor laser. Hall’s work on this laser earned him the Marconi Foundation Prize in 1989. He stayed active in the field after retirement when he consulted with other labs.

In his interview, Hall talked about how people in the industry had to be very careful when talking to colleagues from other companies because of antitrust concerns. Conversations had to center upon material already published or patented, not only because of antitrust, but also to keep others from taking ideas. Hall also talks about the publication of a paper about the semiconductor laser he developed coming out in the same magazine issue as a paper about a laser from IBM. Both companies were granted patents since they were deemed simultaneous inventions, but many years later the German patent was granted solely to GE. Hall also discusses the rapid development of ideas throughout the industry, as well as GE’s contributions to his field and status as a progressive company.

About the Interview

ROBERT N. HALL: An Interview Conducted by Hyungsub Choi, IEEE History Center, 5 March 2004

Interview # 443 for the IEEE History Center, The Institute of Electrical and Electronics Engineers, Inc.

Copyright Statement

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

Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, 39 Union Street, New Brunswick, NJ 08901-8538 USA. 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:

Robert N. Hall, an oral history conducted in 2004 by Hyungsub Choi, IEEE History Center, New Brunswick, NJ, USA.

Interview

Interviewee: Robert N. Hall

Interviewer: Hyungsub Choi

Date: 5 March 2004

Location: Schenectady, NY

Childhood and education

Choi:

This is an oral history interview with Dr. Robert Noel Hall by Hyungsub Choi of the Johns Hopkins University for the Center for the History of Electrical Engineering, IEEE. The interview began at 1:15pm, 5 March 2004 at Dr. Hall’s residence, Schenectady, NY. Let’s just begin with the earlier period of your life. How did you first get interested in physics and how did you choose your major of physics at Caltech, or how did you get to Caltech in the first place.

Hall:

OK. Well, go back quite a way. I had an uncle who was sort of an inventor, career inventor. And he took me to a technology fair when I was a small boy in New Haven, CT. And there were a lot of electrical exhibits, bouncing steel ball bearings and tin can motors were spinning on a flat table, stroboscopes. He got my attention, it seems like these were fascinating little things and I would like to know how they worked and he tried to explain them to me and showed me where to find books in the library. And later on when I went to high school, my mom let me have a little laboratory in the bedroom and I set up a lot of experiments and see if I can duplicate a lot of these things. It all kind of, to me, very interesting experiments. I had a lot of fun. I went into astronomy, and found a book that showed me how to make an astronomical telescope. So I ground my mirrors and made one, 8-inch telescope. Got some beautiful sights of Saturn and some of the other celestial wonders. Lot of fun, very educational too. Anyhow, I have always been interested in how things work and science and doing experiments. And the interviewer from Caltech came through, and I talked to him and ran through some tests and I got a scholarship to go to Caltech. So I went there for four years, actually three years, and I ran out of money, and stayed out for a year to catch, refresh my budget a little bit. Worked at Lockheed Aircraft, testing aircraft. This was just before the war. And then went to my final year at Caltech, got my physics degree, and was hired by GE to go out as a test engineer at Schenectady.

GE employment; Ph.D. studies

Hall:

And I worked for about four years in Schenectady doing a number of things. Worked on magnetrons, I also worked on crystal diodes with a fellow named Harper North. And Dr. Hull suggested that I go back to Caltech and try to get my PhD. I got a Research Council fellowship and went to Caltech, took some graduate studies and graduated in 48. I liked research labs so I decided to come back to Schenectady. I arrived here in the summer of 48, trying to figure out what to do, and in December Bell Labs announced the transistor. And Dr. Hull suggested I look into it, and it sounded like a fascinating new field. We already had a project making germanium high back-voltage diodes. This is my friend Harper North had his program, point contact diodes, and they felt that they can make diodes that would handle almost a hundred volts, which was rather remarkable. They didn’t know what made them so good, but the key seemed to be having high quality germanium, which was awfully hard to get. It was very variable. So I decided to keep things look into first would be a means of getting reliable reproducible high purity germanium. And I tried some initial attempts using chemical methods, and this proved unsuccessful, very difficult. And I looked into fractional crystallization, which Madam Curie used to purify radium, you know, many years ago. And it’s a method of purifying material which ends up with a crystal that’s already in crystalline form. It wouldn’t take any further processing.

Fractional crystallization; p-n junctions

Hall:

So I decided to look into fractional crystallization. The handbook showed that most elements are rejected very strongly by germanium as it solidifies. The distribution coefficient is very, very small. So most impurities would be rejected as the crystal froze. So I set up a program of drawing crystals in a horizontal boat, and directional freezing them from one end. So all the impurities would concentrate in the end. So all the impurities would concentrate in the end that froze last.

Choi:

That’s what they call zone refining?

Hall:

No, that’s not it. It’s fractional crystallization. And it was one-step process. We used this to purify germanium and cut the clean parts out, put them in another boat and do it again. And this was somewhat laborious. But it did give me very high purity material, and we actually had intrinsic samples of germanium. Now, I believe that Bill Pfann at Bell Labs got this idea of fractional crystallization, and had the very clever idea of zone refining, which is a much more efficient way of doing it. And he deserves a lot of credit for that. Unfortunately he died some time ago. That was a very ingenious idea. I wish I though of it. But I didn’t. So, I was studying the purification process, and beginning to add impurities to determine which elements were donors and acceptors. And I tried to run little crystal, where I added a little bit of arsenic, and lo and behold this crystal did indeed come out very strongly n-type at one end, and there was a mysterious impurity which always made the other end p-type, the end that solidified first came up p-type. We didn’t know at first what that impurity was. We later found that it was boron. But this crystal that I had doped with arsenic happened to come out as a single crystal. It was one from p-type to n-type, and you could even see their crystal facets on the surface. It was very smooth and you could see the crystal faces trying to form. To evaluate these crystals, what I did was to pass a current through it, and use a potential measurement, a voltmeter, to measure along the links to get the resistance profile, the voltage profile. From that you calculated the impurity distribution from the resistivity of the material. Well, this crystal when I measured it with one polarity, showed a very high impedance region in the middle. And when I reversed the current it was very highly conducting in the middle. And I thought I had a new principle of rectification. And I called it, I thought of it in my notebook, as a barrier-less rectifier. I realized sometime later that this was a very broad p-n junction, and obeyed the same laws that Bill Shockley had worked out for p-n junctions. But to me this was, I thought it was a brand new idea. I was modulating this middle region by injecting carriers that intermingled and made it highly conducting in one direction, and very high impedance in the other direction. So it was a p-n junction of course, but I realized later. But I was studying this, and I thought I thought I would try to do would make it much more efficient structure by taking away for the germanium and try to put donor impurities on one side and acceptor impurities on the other. That’s what I wanted to make. I thought this would make a much more efficient rectifier. You need to change anything?

Choi:

No.

Hall:

<flashmp3>444 - hall - clip 1.mp3</flashmp3>

So what I wanted to do was to diffuse impurities in the opposite sides, and I needed to pick out the right impurities. Well, of the acceptor elements, the one that seemed most suitable from a metallurgical properties was indium, because it’s easily melted, it’s soft. We assumed it would diffuse into the germanium and make a p-type layer on one side. And antimony is a donor element, which would melt at about the same several hundred degrees centigrade. So I set up an experiment. I took purified germanium, put a little drop of indium on the top, and some antimony on the bottom, and on a metal plate. And I put this in a [courts] tube with hydrogen going through it, and heat it with a [busten] burner, till I could see everything starting to melt and alloy together, fuse together. And I took it out and probed it, and by golly it would rectify very well. It was very leaky in the reverse direction, but it had remarkably good forward characteristics. It would carry several amperes at only a volt or so, which was very remarkable. Previous rectifiers had only been able to pass a few mili-amperes. So I could this had possibilities as a power device, and I did more experiments. Soon I could find that by etching it, I could get much better reverse characteristics. I began making large area cells, half a centimeter squared or so, by melting indium and antimony on opposite sides of a wafer. And I made these into rectifiers that were water cooled by soldering copper blocks on opposite sides. And I set up test experiments where you could pass hundreds of amperes in one direction, and block a hundred volts in the other. So I could handle many kilowatts of power, with these rectifiers, which were rather phenomenal. So I proved out in practice that these would handle large amounts of power. Then we wanted to find out how they worked. So I measured the electrical forward characteristics very carefully. Measured the current-voltage characteristic. And I did this as a function of temperature, because I found that it didn’t obey the usual formula that people believed that rectifiers would show. In other words exponential ev over kT. They were not quite that sharp. Actually they fit the characteristic of exponential of ev over 2kT. In other words, there was a factor of 2 that needed to be explained. And I puzzled over this for a long time. I went through Shockley’s book and studied all his analyses as well as I could, and tried to make an analysis that would account for the ev over 2kT behavior. Because the rectifiers obeyed this for over a wide range of currents and voltages and temperatures all the way from 77 degrees up to a hundred degrees centigrade. They showed this very consistent and surprising behavior. Well, it finally occurred to me that instead of electron meeting a hole and annihilating each other, which would be a two-body cohesive square of the concentration of carriers. It was linear in the concentration of carriers. And I proposed that the re-culmination happened by an electron dropping into a state midway into the gap, and then meeting a hole where it would fit complete the re-culmination process. In others words, a re-culmination via mid gap impurity defects. So, this, I analyzed this kind of kinetics and found that it did explain the ev over 2kT very accurately. It accounted for the much longer life time that was being observed experimentally for minority carriers in germanium, and also their variations with impurity concentration and so on.

Shockley-Read-Hall recombination; PIN rectifier

Hall:

So I was pretty well convinced that I had a new model of recombination, which I discussed at a meeting at the Physical Society. Bill Shockley was there. After the meeting he quizzed me at length about this, and apparently he was convinced, as a mechanism was quite important. So he and [Thorton] Read went through a very extensive analysis and produced a very formidable paper on electron-hole recombination. So this process is known either as a Shockley-Read-Hall recombination or Hall-Shockley-Read, depending on who you’re talking to. This was a real important breakthrough in semiconductor physics because it is a new mechanism for how minority carrier lifetimes behave in semiconductors. Well, I went on and proceeded to develop, publish this work on PIN rectifier, that is called, and also I’ve given that analysis of the minority carrier lifetime in oral paper, I guess published also in Phys. Rev. Letters, or the equivalent. And this work on the power rectifiers was of interest to several GE departments, of course to Electronics Lab in Syracuse. This was where John Saby and, what’s his name, trying to think of it, Paul Jordan were working, and they were commercializing these rectifiers as power converters for TV sets. And they made rectifiers that would handle several hundred volts, with little pellets of indium soldered to the germanium or alloyed to it. This was the commercial product that GE was producing. And John Saby became interested in the possibility of making transistors, and he built some structures with alloyed indium on both sides across the thin wafer, and produced some working transistors. I was working along similar directions at GE, but I believe his records predated mine as far as a working transistor, and he has GE patents as far as alloyed transistors go.

Choi:

Saby.

Hall:

John Saby.

GE Lab collaborations; joint publications

Choi:

Where does Dr. Crawford Dunlap fit into this story?

Hall:

The people in the GE research lab when I first came there in 48, Harper North was working with these high back-voltage diodes, and Crawford Dunlap was studying impurity diffusion, trying to measure diffusion. I guess it was pretty well established that diffusion was very, very slow in germanium. I don’t think that he even had numbers for it at this point. Well, I made these PIN rectifiers. I assumed I was doing diffusion to make the n-plus and p-plus layers. That’s what I was trying to do. So in the experimental write-up, I referred to these as diffused rectifiers. Dr. Dunlap became aware of the work I was doing, and he called me into his office and proposed that we produce a joint paper on diffused rectifiers. And I was rather taken aback by this, because as far as I can see, he hadn’t contributed anything to this. I did not make use of any of his work. I don’t know if he had any diffused results. Anyhow, we ended up producing a very short joint paper, describing these initial experiments on diffused junction rectifiers. I also mentioned that somewhat later, these power rectifiers were of interest to the work at Lynn, Massachusetts. They made rectifiers using selenium and copper oxide, and they became very interested in making welders based on high powered germanium diodes. One of the men working there was Bert English, A. C. English, and he was making large area germanium rectifiers. He opened some up and etched away the indium, and discovered a very interesting crystal structure underneath the indium. And he realized that what was going on was alloy solution of germanium into the indium, and as the crystal cooled the germanium would crystallize back out onto the substrate. So we were doing liquid phase epitaxy, what we would call now. But he realized that instead of producing a diffused layer, this was re-crystallized very heavily doped by indium, recrystallized back to the wafer. So when the patent disclosure for the PIN rectifier was put in final form, the wording was always fused and diffused junctions, sort of trying to stretch the language to include either, diffusion, which would be making a very thin diffused layer, but the recrystallization from the alloying process was one of the key ingredients to make these rectifiers work. But English did explain this structure, elucidated what was really going on. Let’s see, where are we going now. Trying to go over some of my notes. I thought you were primarily interested in this story from the early purification work up to the point where we were making PIN rectifiers and transistors. I was summarizing the various steps here. But you want to go beyond that?

Choi:

The technical story is of course my interest. But I’m also interested in how researchers worked together in the lab, and how the lab cooperated with Syracuse.

Alloyed junction transistor; rectifier production

Choi:

I know that you published this paper in 1950 titled “p-n junctions prepared by impurity diffusion.” There is also a jointly written paper by Dunlap and Saby, around the same time, on the same topic.

Hall:

I’ve heard about this, but I don’t think I remember seeing it.

Choi:

I think have it here somewhere. There we go. That was 53.

Hall:

Here they’re calling it fused-impurity already. [Hall skimming paper.] I guess they were trying to decide whether impurity diffusion played a role or not. I don’t want to take time to read it right now, but. Did diffused layer really is big enough to play a dominant role, OK. Maybe the case, I don’t know.

Choi:

So, is it fair to say that all these works formed the basis of what they called the alloyed junction transistor in the early 50s?

Hall:

Yes, I think so.

Choi:

And was GE manufacturing these kinds of transistors at the time?

Hall:

Well, GE…well let me. [brings in diode sample] These are the things that Electronics Park, Syracuse had made. These, inside there is a thing about the size of a quarter inside there which is a hermetically sealed capsule with a germanium diode inside it.

Choi:

So this is a single rectifier?

Hall:

Well, there’s two rectifiers, push-pull structure here. So it was made to be a full-way rectifier to convert power for a TV set. So GE manufactured these. What’s inside this is a wafer of germanium with little pellet of indium on top and leads, of course, connected to it. They were also working on larger power devices. These are some… These might have been made by Lynn [Hoover] works after they moved down to some place in Philadelphia. Great big mass of copper, studs on the bottom, but inside is a germanium wafer about a square centimeter or so. And at the time, whenever that was, these were considered pretty… the label says the world’s largest but it’s no longer true. [chuckles] Yeah, these were manufactured. GE got into the production of the power devices both in the… well, these were developed and phyrestors were made for high voltage converters for power transmission. So they had great big stacks of hundreds of these in series, phyrestors-controlled, silicon, not germanium but silicon. The history there is sort of interesting. I have developed large area PIN rectifiers, published the papers, and the people in Siemens [Stuttgart] laboratory were decided they were purifying silicon. They had their own float zone system to make silicon, and they cut up wafers and used a very similar process to make alloyed large area PIN devices of silicon. These were much more effective because they would stand a lot higher temperatures, and they worked beautifully. It was picked up here in Syracuse also, and Nick Holonyak, who was a young fellow who joined GE. He had then a student of John Bardeen’s, I believe. And then went to Bell labs for a while, then got drafted, went over to Japan for a while, and came back and was hired by GE. He knew about three terminal gadgets that Bell labs was developing as a switch. And he heard about this large area silicon work and found that he could make silicon rectifiers, large area PIN rectifiers, and that he could make them as a switch controlled rectifier. And he sort of single-handedly produced these thyrestors [sp?] in Syracuse. Very capable guy, Nick Holonyak. But we were closely with him, but he deserves the credit primarily for making thyrestors with silicon. So anyhow, this original idea I had with PIN germanium rectifier has grown tremendously. These were silicon things made in Europe and GE had quite a production line going for quite a while. They are made all over the world now, as controlled rectifiers of silicon. They’re used on power transmissions, they’re used on electrical automobiles, locomotives, motor controls, everywhere, thyrestor controls. This is not my work. This was a work that grew out of what Nick Holonyak had done.

Choi:

So in terms of the alloyed type of rectifiers, I assume that John Saby was the man in charge of that program at Syracuse.

Hall:

I guess he was in charge. I don’t know the management structure there very well. He worked at the Electronics Lab, Nick Holonyak was also there, some others I have trouble remembers these names. But they were more closely tied to the manufacturing group in Syracuse. They were supporting… industrial labs supporting them, but they did some excellent work, there is no doubt about that.

Choi:

So did they have to learn about the new way of making rectifiers from the Schenectady Lab?

Hall:

Well, when we had an idea for a power rectifier that looked practical, part of out job is to transfer the technology. So, I would go on a visit to Syracuse, and they in turn would come visit the lab when we’d show them what we’re doing. They had their own ways of doing things, which in many ways were more efficient for manufacture and develop a more manufacturable structure, commercialized it. That was their job was to turn it into a commercial product, and study how it worked and commercialized it.

GE executives and transistors

Choi:

Now, GE was starting to produce transistors at some level, at least. How much do you think the corporate executives put an emphasis on transistors as opposed to vacuum tubes? I assume it was not a smooth process of transition.

Hall:

Well, at the upper levels of GE there, came to be a decision whether GE would get into the communications field, or stay in industrial power devices, where they had a much more established position. And I guess they could see a lot of competition coming from people like Texas Instruments and Raytheon, RCA, Fairchild out in California. integrated circuits were coming on in a big way, and GE did not have a very strong position. Rather than pour a lot of money into electronics, I think they decided it was better to purchase these electronic transistors and circuits and so on, and not try to compete in semiconductor work. This is a disappointment to many us because GE had already done pretty well, you know, getting started with a lot of integrated circuits, but the decision was to back off and let the competition take over that field and GE would not be a major player in the integrated circuit field.

Choi:

So, that kind conservative business move was quite apparent in those days in the 50s?

Hall:

Well, it was a decision, you know. You can’t do everything. And there was so many players in the integrated circuit field that I think… probably made sense from a business point of view to invest in other kinds of fields like aircraft engines and power devices, and not try to compete in silicon integrated circuit technology. It’s pretty expensive business. Tremendous competition.

Nuclear physics

Choi:

Let’s backtrack a little bit, and go back to when you started work at GE again. I read that you did your PhD degree in nuclear physics? What you did during your degree, did it have any relationship between what you did at GE after your degree and what you did during your studies?

Hall:

Well, when I returned to Caltech, there was a lot of interest in nuclear physics, because the atomic energy, the reactors, atomic bomb had been developed, and a lot of nuclear physics working, a lot of training needed to be done to train people in this field. There were some excellent courses being taught. I found a thesis project which was being directed by Dr. Fowler, who was very interested in cosmology. And he wanted me to study the reaction of protons bombarding carbon to produce one of the fusion reactors that goes on in the sun. He was trying to understand the solar energy cycle. And I had some experience with ion sources and working in gas discharges and so on. So I took on as a thesis project, making a proton source, ion source, to bombard high voltage accelerator. Not a very high voltage, this was about a hundred kilovolts… to bombard carbon with protons at high currents but very rather modest energies to try to study the energy cycles that were going on in the interior of the sun. One of the steps being the capture of protons by carbon. So I set up this experiment, and bombarded carbon cut some detectors and measured this very very low background of a few counts above the noise level. And eventually did succeed in measuring the cross section for the reaction, which made… that was my thesis project. It got me interested in training in nuclear physics, but also in working with ion sources and plasma work. That’s what I though I would pursue when I came back to GE. I thought I’d get into gas discharge work. I was trying to find a suitable subject to study, and then the transistor came along.

Introduction of the transistor; germanium purification

Choi:

Do you remember that day when it was announced?

Hall:

Oh yes.

Choi:

What was your reaction to that?

Hall:

Well, this hit the Physical Review, letters to the editor, I guess it was called. And I remember Dr. Hull, who was the director of our section of the laboratory, came in with this reprint from the Phys. Rev, Physical Review Letters, and he says, Robert here’s an interesting development from Bell Laboratories. It looks like something pretty new and exciting. Would you like to look into it and see if there’s anything interesting there. So I looked at it, I was skeptical at first. And I thought, well, is this a real effect, is it really going to produce power, and I studied it and analyzed the data that it contained, and concluded, yes indeed it is something new and a basis of an amplifier. And I got interested and we decided it was a good subject to pursue. We already had a group working in germanium diodes, these high back-voltage diodes. And Crawford Dunlap was already studying Hall measurements in germanium and conducting his diffusion experiments. So I decided, this looks like an interesting field, and I looked at what was going on and what the real problems were that they had to understand in order to pursue this field. And they needed a reliable source of germanium. So I took on the project of trying to purify germanium. So that’s… What surprises me right now is that I don’t remember anybody advising me what to do. I was just told look and see if you’re going to find something to do that’s worth doing and do it. You know, find out what needs to be done and go to work on it. So that’s what I did. I don’t remember offering advice or telling me how to go about doing it. It was just up to me. So I was a young kid just out of school left to my own devices. And I made a false start trying chemical methods first, and decided that was hopeless, and hit upon this crystallization idea that proved to be very fruitful and efficient. And everything worked from that. It’s one thing after another. You get one problem solved that covers something new. It’s purification when I tried to measure distribution coefficients, I doped this crystal with arsenic to study what it would do and happened to have this single crystal, turned out that had this remarkable characteristics. The reason that it worked so well was that it happened to be a single crystal with excellent structure, so it didn’t have much recombination, very few defects to cause recombination. So it had a very long minority carrier lifetime, so these conductivity modulation phenomena were very pronounced, so you could easily measure them. It was very lucky. But, you know, if you have a piece of good luck you capitalize on it.

Choi:

I assume that most researchers in the early years of transistors had their background in working with vacuum tubes. And you didn’t…

Hall:

No, I didn’t have that kind of background at all.

Choi:

Is it a correct statement that most researchers had a background in vacuum tubes?

Hall:

Well, the people who were applying these things in circuits, that’s certainly true, pretty much of them. But there were also students coming out. Purdue had a pretty strong graduate study in germanium, they were doing crystal study and purity, trying to get… setting the properties of germanium, understand the band structure, conductivity and mobility and so forth. Lark-Horovitz was one of the leaders of the group at Purdue. Lot of their work was a little strange but they were guys who were learning how semiconductors behave. This was a very new field and the old oxide semiconductors were just hopeless. The textbooks were pretty hard to understand. Illinois, Johns Bardeen’s group, no wait a minute, it wasn’t Bardeen at that point. No, maybe it was. Cause right after the transistor at Bell Labs, Bardeen had left and gone to Illinois. And he was training, Nick Holonyak was one of his students. In fact, I was invited out there to give some lectures on PIN rectifiers and germanium. Nick Holonyak took one of my courses there.

He, by the way, has been getting all kinds of awards lately, light emitting diodes and he’s… you see him up in Spectrums and a lot of journals right now because he’s done a lot of excellent work and turned out lot of good students in light emitting diodes and lasers. He was very interested in making visible LEDs and lasers. And he made the first semiconductor, visible laser.

Institutions in the electronics industry

Choi:

Taking a more broader perspective. Within the electronics industry, what was the relationship between researchers from different institutions, say RCA, GE, Bell Labs, and say Purdue people and Illinois people?

Hall:

OK, well, Bell Labs, since they invented the transistor, they put a tremendous effort into semiconductor work. They had a lot of very good people and a very big budget and did… They were called Ma Bell, you know, they were sort of this source of everything in semiconductors. And they put on a conference on germanium technology sponsored by the Air Force, I think. And a lot of us from the industry attended these conferences and learned about crystal drawing and doping and semiconductor phenomena. It was quite an educational process.

Choi:

Were you at that conference?

Hall:

I attended that series, yes.

Choi:

52?

Hall:

Is that 52? It might have been. Probably was. Anyhow, there was a good education. And a lot of people in industry from Raytheon, RCA, Philco, GI, all sent representatives and got training in these fields. So, Bell Labs was sort of a focus of semiconductor work but were being picked up in a lot of smaller labs around the country, and we were one group that, a rather small group, but you can always find new things, there’s such a… a new field, there’s always new areas to explore.

Choi:

We’re back again. And we were talking about the Bell Labs symposium of 1952 and you were present there. My general question was about the relationships between researchers from different institutions. I have been looking at the papers at the RCA labs recently. And I found a number of instances where researchers go on trips to other labs, even to Europe, exchanging information, exchanging sample…

Antitrust concerns and professional communication

Hall:

Well, let’s see. This… We were cautioned in head to observe a lot of rules. In other words, you can talk about things that had been published. In other words, it was a published material, you can discuss it at length with competitors. But there was a lot of antitrust concerns. We had to be very careful, in discussions that we didn’t disclose to one industry one commercial party without excluding somebody else. In other words, you could generally talk in groups with different people, but talking individually with someone from another commercial organization you had to stick to things that had been published. That’s pretty carefully… quite a concern about that. You could often talk quite freely with university people, and discuss research results there, quite freely. And I was often very helpful because they had explanations for a lot things that might have been puzzling to us. And we were often invited to universities to give talks. As far as being invited or visit commercial labs, I don’t remember I ever went to RCA. I don’t think this is ever done, very rarely. We would meet at conferences and we would discuss subjects, things that we were working to some extent, but we had to be careful not to cover stuff that had not been published or patented already.

Choi:

And that was the case even to Bell Labs, you didn’t visit them?

Hall:

Well, for this conference of course, they were disclosing material which they were asked to by the Air Force, I think they were sponsored by the Air Force, some military group, anyhow. They helped organize the conference. I think what most of what they showed was covered by their own patent work. They had it all documented. And it was being disclosed to anybody who sent a representative. I think the companies paid a fee for this, but it was monitored by the Air Force and nothing was done exclusively to any one individual commercial company. Anybody who sent a representative would get whatever was being talked about in the course.

GE achievements in solid state electronics

Choi:

What do you think was the major contribution of GE in solid state electronics?

Hall:

In solid state electronics? That’s a hard one to answer because you tend to focus on your own work. I know GE had a pretty good program in silicon integrated circuits in the early stages. They had a [inaudible] micron technology that was coming along quite along in making… Well there was a group doing very good work in surface physics. Got [Peter Gray] and [Dale Brown], some others had done some very excellent work in understanding of surface states in silicon, oxidized silicon surfaces, which were critical of… crucial importance for silicon integrated circuits. This group had a very good reputation in understanding of surface states and reliability, and analyzing surface conditions of silicon. There was quite a substantial effort in thermoelectric generators. Lead-telleride, bismuth-telleride, I forgot what the compounds were, but trying to produce high efficiency converters to generate electricity by thermal differences. The Russians had made electrical generators using thermoelectrics. And some of this technology was picked up this… lot of work on thermoconductivity of solids gone slack, and some very good studies of thermoconductivity of solids trying to develop materials which were good conductors of electricity but poor conductors of heat, as a way of making efficient thermoelectric converters. This was quite a significant project for quite a while, but I think was finally abandoned because it never really proved out. I think they’re used for, some of these thermoelectric coolers are used for refrigerators. I’ve seen these recently for lasers for these kinds. I have a little trouble thinking, but right now… Oh, for nuclear detectors. By the way, there is a very interesting story I can tell you about nuclear detectors, if we get to change the subject. But, thermoelectric coolers are used for lot of instrumental work and laboratories and nuclear detector, particle detector physics. This nuclear detector story, do you want to start on that?

Choi:

Yeah, sure.

Nuclear detectors

Hall:

<flashmp3>444 - hall - clip 2.mp3</flashmp3>

OK. I had a friend of mine in nuclear physics at Caltech named Dave Lind. He stayed in nuclear physics, we were friends, and I’d often visit him as a friendly visit. One time he invited me to go skiing with him out in Colorado. And we were sitting up on top of a mountain, cross country skiing and so on. I was just thinking, down below there you can see in the distance the airport, and I was due to be back there in about two hours. And here I am sitting up in my ski clothes on top of a mountain. So we decided we better pack up and get going. So we skied back down to his car, and while he was driving me down to the airport, he got to talking about the germanium problem. He had been using lithium drifted detectors, which are germanium diodes used for detecting gamma rays. They’re wonderful things for doing nuclear physics and reactor physics. But something had happened to the nuclear detector manufacturer all over the United States. None of the germanium, what they were using was a lithium drift process, where you would take a block of n-type… p-type germanium and put lithium on one side and diffuse it in to make a drifted, lithium drifted detector, which made a diode with an effectively intrinsic layer in between, which could be a centimeter thick. I have never heard of these things before, totally astonishing to me. But we knew about the drift process because one of the guys in our laboratory invented the idea, [Eric Pell]. So I knew the process, but I didn’t know it was being used in making detectors. Well, Dave Lind told me that something had happened to the germanium manufacturing process in the US, and in fact all around the world. Nobody could make drifted detectors any more, the lithium would no longer drift in an electric field. This was a calamity. And he said this was terrible, we’re doing nuclear physics, we cannot get these detectors anymore, they can’t make them. You guys, I suppose, know something about germanium, why don’t you see if you can fix the problem. Well, it was totally new subject, I have never heard of these devices before. So I had to study up, learn the literature, and study up what has been going on, and it was fascinating. These guys had learned how to make germanium detectors, these were a centimeter thick, and fully depleted diodes. So that when you apply a thousand volts to them, the electric field will go all the way across a centimeter thick. If gamma ray came through, they ionized a lot of atoms, and the ions would drift across and give a pulse. You can measure the pulse height and measure the energy as a gamma ray. So this was wonderful for nuclear physics, but something happened to the germanium and they couldn’t make these anywhere in the world, and it was a calamity. So what I decided was, if you can make germanium pure enough, you can make diodes with thick depletion layers, and you would not need the lithium process to do it. You could just use high purified germanium. And it’s easy to calculate what you need. You had to have a much higher purity than anybody ever had achieved before. And most people think that this was impossible. In fact I had some publications saying that this was an interesting idea but of course totally impossible to accomplish. Well, I wasn’t sure it was impossible, and I got approval and a contract to see go ahead and see if it was possible. And we found that we could indeed get germanium sufficiently pure, that we made diodes with depletion layers of half centimeter or so of thickness, which was good enough to make detectors. The trouble with lithium detectors, if you could make them, you always had to keep them near liquid air temperature. If you let them warm to room temperature, the lithium would precipitate and your detector would no longer work. So you drifted germanium, cool it down to liquid nitrogen, and always have to keep it cold, which is a big nuisance. But the high purity detectors would be stable. And what you had the key was to get germanium sufficiently pure, and we established it was possible… a group at the [Lawrence Brook’s] laboratory picked up on this and some of the detector makers began to develop their own ways of doing it, and pretty soon in a few years laboratories all over the world were developing, making detectors out of this high purity germanium. We learned why the lithium wouldn’t drift, because of some contamination of oxygen in the germanium. That process is no longer used. I had learned from Dave Lind that these detectors existed and needed help, and we got approval and were able to make high purity germanium that would produce these gamma ray detectors. And it has since been commercialized by laboratories all over the world. People in [Ortech] and Princeton Gamma Tech, well it’s a different name now… the detector makers are all making their own germanium. The people in Belgium, Hoboken Chemical, has drawing lots of high purity germanium for sale to detector makers. So it has revolutionized the technology of making detectors, and a lot of the detectors and satellites going around… astronomical observatories and so on, are using high purity germanium for detecting gamma rays. So, it was really satisfying feel… GE’s dropped out of it because we’re not in the detector business, and I was only allowed to get into it because I could get a contract to do the work… and we thought for a while it might be a way of making a gamma imaging camera for nuclear medicine. But never proved out. They were too hard to make bigger rays. So that never was commercialized.

Choi:

Let me clarify one thing. Your conversation with David Lind was around what time?

Hall:

I could check on that, but… It must have been around 1960. I can look it up easy enough. I have a list of publications, I can figure it out.

GE research contributions

Choi:

So all these contributions made by GE… Do you think… This question is supposed to be a little bit provocative. Do you think that GE’s contribution was underestimated by other people in any way?

Hall:

Oh, I don’t think I would say that. Well, GE’s research lab had a very excellent reputation over the years…

Choi:

But in terms of solid state electronics?

Hall:

Maybe as applied to electronics, it’s not so outstanding. But they had a lot of work in ceramics and metallurgy. Some of these advanced refractory metals for turbine rotors. Pretty important stuff. [Lukolock’s] lamps. There has been a lot of work in metallurgy and ceramics, structural materials. I think GE’s got a pretty solid reputation is the field. I find it a very diversified laboratory, and funded most fields. I think they do very excellent work. It’s changed a lot. It’s very different now than it was a few years ago. As the field matures, and it’s pretty well worked out, it’s time to shift into something different. So, it’s… they keep moving. And it’s one of the laboratories that’s well renowned worldwide now. They have branches all over the world. It’s called GE Global Research, and they have laboratories in India and China. There’s one opening up in Germany now. So it’s a progressive organization.

Evolution of solid state research; semiconductor laser

Choi:

Going back to your work, as opposed to the lab as a whole. I can see that you have made some notes…

Hall:

I have tried to refresh my memory about this period… You are interested in this early development of leading up to the transistor. You can have this if you want. These were the publication dates I could find. And this is sort of the chronology of how these things worked in my own career.

Choi:

These were what you talked about in the first part of the interview?

Hall:

Yes, I was…sort of went through this list of… sort of one thing led to another all the way down here. It was… as soon as you found one interesting thing and figure it out how that worked, you’d learn something surprising to be uncovered again. One thing leads to another. Such a rapidly growing field that there’s such an abundance of new things to look into. It was like a kid in a candy shop, you know. You could always find something to work on.

Choi:

That’s the joy of doing science, I guess. So your work extends far beyond this list, beyond 52. Out of those, what do you feel that your major contribution to solid state was?

Hall:

Well, the semiconductor laser was one that was very exciting thing that happened very quickly. I had been interested in trying to make semiconductor devices that produced light. We played around a bit with silicon carbide, trying to get some visible light, but that’s tough material to work with. We had a little fun but nothing good came of it. And I was aware of gallium arsenide devices that produced infra-red light. I knew about radiated recombination because I’d worked quite a lot with electron-hole recombination. I knew that the III-V compounds should be much more efficient by producing direct recombination of electrons and holes. They didn’t need a phonon to be involved, so their radiative recombination coefficient was much higher than the III-V compounds like gallium arsenide. And then at one of those device conferences, I heard about reports from Lincoln Lab, which is somewhat related to MIT. Two fellows, [Keez and Quist], found that they were getting very intense infra-red light from gallium arsenide diodes. And they had really remarkable statements they made. They believed they were getting as much as a kilowatt per square centimeter of energy radiated from these junctions. That was just an unheard of level of light intensity. And they said these diodes, the light, even though it’s in the infra-red, it was so intense that if you look at them you’d get the visual sensation of red light. It was infra-red but it stimulated the eye to give you the impression it was red light. When I heard about this, it seemed to me that… Well, I have had a visitor from France, this was in 1962. Because, we visited foreign laboratories and sometime foreign representatives which come to the lab, and we talk to them about what we were doing in public, published areas, work related to that. And he said, Dr. Hall, you make all kinds of semiconductor devices. Why don’t you make a semiconductor laser? I said that’s a nice idea, but there’s lot of problems, efficiency in optical and light emitting diodes has always been terrible. It’s like a fraction of a percent. We wouldn’t have anything like shock line spectrum that you have in most lasers. I just don’t see any way of making a laser out of a semiconductor. Nevertheless, he sort of prompted me to look into the subject a bit. So I browsed around in the library to see what people were doing in the lasers, because it was a brand new field. So I became a little bit familiar with more conventional lasers. When I heard about this gallium arsenide, very intense infra-red, I put some numbers in. And to be able to calculate the carrier concentrations in these diode must have been up in 10 to 18 per cc or so. That would be enough to make it degenerate populations of electrons and holes that were intermingled, which means that you might be able to get stimulated emission efficiently, and a laser might be possible. So I saw an exciting prospect, and I had the idea that if you are going to produce this intense radiation in a junction, you want to have the radiation going back and forth in the plane of the junction, and you want to put mirrors on the opposite edges of it. So I decided this might be a possible device, and I talked to some friends and they sounded enthusiastic about it. I told my boss that there was this idea which sounds… It might work. We don’t know what a laser would do. It would be fun to try, and even fit in work with learning a lot about these high efficient light emitters… could be very useful. So he said go ahead and try it. We all set to work and we had a scoop of five of us, I guess, and we divided up the work among the different fellows. I was going to, [Ted Salties] and I was going to make the laser structures. I designed them and he tried to make the design I proposed. Gunther [Fenner] was good at electronics, so he was going to set up test equipment, and he had an image tube so he will observe the infra-red radiation and look for any possible changes. Dick Carlson was going to diffuse crystals, try to find more gallium arsenide crystals and diffuse them in different impurities, so we could have starting material to work with. So we all set to work in different phases of the project, and in a very short space of time, we had got one that actually worked. Gunther had gotten some results on a Friday night. It looked promising, so he came in that Sunday and tried some more, and something very strange going on. He called his boss to come in and look at it. And next Monday, he sure enough had a laser thing that clearly was a working laser, pulsed operation and liquid nitrogen temperature, so very impractical, commercial kind of device. But still a new phenomenon and very exciting. So we all set to work on it, and worked real hard, got a lot of data, made a lot more better diodes, and got to study the results, got a paper ready for publication and the patent drawn up so that we can take it up to the patent office and get coverage, then we’d be able to publish the work and talk about it. And got it ready for a publication for the November 1st issue of Phys Rev Letters. Well, the same month, same November 1st issue of Applied Physics Letters, came out a paper from IBM, claiming that they had almost a laser. This was very pronounced stimulated emission from gallium arsenide, a very pronounced spectrum narrowing. And it was so close to getting a laser that the newspapers took it as both companies announcing the laser simultaneously. And as far as patents go, the patent department figured out a way to allow a patent from both companies simultaneously.

Choi:

Another case of simultaneous invention…

Hall:

Yes, often it happens. We actually had a [fabrie poro] geometry, so it was a true laser. They had not figured out how to do that, but they had such very pronounced spectrum narrowing that they were so close that there was no point in fighting about it. There was a curious maneuver that the lawyers achieved to get both patents to issue. We thought it was band-to-band radiation, where the energy of the light was slightly greater than the band gap’s operation. And the IBM people thought it was recombination [inaudible] electrons and split-off band of acceptors, which would be less than the band gap. So all of our claims specified radiation greater than the band gap, and all the IBM claims say radiations slightly less than the band gap. So they could distinguish the two. All the claims could be clearly distinguished. Well, as far as I’m concerned, my knowledge at the time, these heavily doped materials have bandages are smeared out so much you can’t specify what the actual band gap is anyhow. So it’s a moot point. They allowed both patents, and they deserve a lot of credit just as we do for inventing the first laser. Now, we also had a contest in Germany. In Germany technical judges are quite astute in technical matters, they know their physics pretty well. They needed to pin down who were the… they were being required to adjudicate who actually had the invention of the semiconductor laser. So I presented our evident before the German patent court, and we were able to persuade them that we were indeed the true inventor of the lasers. So in Europe, the German patents gave us as the sole inventor, GE as the sole inventor of the semiconductor laser. So that was kind of a nice victory over there. That happened several years later though. And by the way this fellow Nick Holonyak I mentioned, he knew of our work in lasers, trying to develop a laser, we were close friends. He had been trying to make visible diodes, light emitting diodes with gallium arsenide phosphide, with a mixture such that the band gap was still direct, but it was bigger than gallium arsenide so that it would produce visible light. He was making these mixed crystals, and he followed out procedures in these mixed crystals of gallium arsenide phosphide. He actually succeeded in making a visible laser, later that same year. And I guess November-December of 1962. So he went right to work on it, and he gets credit for the first visible diode laser. So… he’s a very efficient experimenter, very good man.

Influence of semiconductor laser and fiber optic communication

Choi:

So, this work earns you the Marconi Foundation Prize in 1989. How do think that this invention has contributed the progress of communication?

Hall:

Oh, fiber optic communication has taken over all over the world, you know. And you needed a good source of light to fit into these fiber optic cables, the optical fibers. And a laser is one that can focus its light down a little fiber of glass and be transmitted around the world, can be modulated very rapidly. So semiconductor played a tremendous role in communications. Of course, it is used in lots of other things, CD scanners, copiers… gallium arsenide lasers are used, I believe they’re lasers, used for Xerox copiers for scanning the print with infra-red light. But only for communications, diodes are used everywhere… and for CD players.

Choi:

How do you think that semiconductor lasers fit into the path of communications progress?

Hall:

Oh, I should have, I did tell you that ours were very primitive. They worked under pulse conditions and liquid air temperature only. Well, GE did not see much of a future. We weren’t in the communications business, and nobody knew quite what to do with these crazy little lasers that worked under these very strange conditions, low temperature, pulse conditions. We didn’t see much of a future as a product within GE. I think the group at Syracuse did manufacture lasers for a few months as sort of a novelty, thousand dollars apiece, but soon dropped out of the competition. Other labs began to work the problem of getting a CW operation and improving this structure so you could get efficient injection, good carrier confinement, and getting it to work continuously at room temperature. This took a lot of very ingenious contributions. People had to work, understand the problem and work very hard at getting more efficient structures. They had to use hetero-junctions, they had to use three carefully doped MOCVD methods where you could deposit layers of vaporing, and get structures tailored to the exact impurity distribution. They would give you this very efficient results. A lot of different labs worked very hard, and a number of important breakthroughs were done, to get to where you had a laser that would work at room temperature, work continuously, and work reliably for a long lifetime. Because a lot of these lasers would just disintegrate in a few hours. They were something moving around in the laser to cause them to go bad. And these other labs had to figure out what the defects were that were causing this trouble and get rid of them so that they can have a good operating life. So a lot of men worked very hard, and made a lot of contributions to bring the laser to commercial level. And GE would not involve in much of that work.

Choi:

So, were there cases were used in ways that you didn’t expect to happen?

Hall:

Well, at the time we invented it, we had no idea where it might be applied. We didn’t know what to do with it. So, these recent develops since have been, to me, very impressive and quite surprising. I never expected that they would be able to get one to operate continuously, and I never thought that they would get it to work at room temperature. But these guys, clever guys worked hard and figured out how to do it.

Gallium nitride lasers

Hall:

And even more recently, there was a young fellow in Japan. I can give you his name if… But he found out a way of making gallium nitride lasers, which worked not only throughout the visible, in the blue and even the ultra-violet. These are amazing things. And this was almost a solo contribution by this young fellow, who talked to his boss, his is a very small company making phosphorous in Japan. You know this story, I think, don’t you?

Choi:

Is this a recent invention?

Hall:

This is quite recent. I have reprints if you want to see them, but…

Choi:

I know that the LEDs people usually use are red and orange.

Hall:

Yes, but gallium nitride lasers are able to work into blue and green and ultra-violet even. This is an astonishing breakthrough. People have tried to make lasers into the yellow and green, so forth. One group is even using the II-VI compounds, zinc selenide, some of these crystals, which are awful things to work with. They did get a sort of a blue green laser, but it was tough making it last any length of time. And when this fellow, I can get you his name easily, he found out that he could make gallium nitride lasers using MOCVD that would work in the blue and even ultra-violet. And this is a total surprise, that’s picked up in a lot of different places now, and he has since gone to the University of Santa Barbara. I guess is pursuing that work at UC Santa Barbara. That’s an astonishing achievement by a young fellow who somehow talked his boss into giving him a million dollars or so, so he can go to Florida and learn how to run MOCVD apparatus, take it back to Japan, and use this on gallium nitride and try to make lasers, and he succeeded. It’s an amazing story.

Research and development as a career

Choi:

OK, I exhausted all my questions that I had prepared, but there are some other things that I want to ask. Did you ever think about being an academic physicist?

Hall:

I thought about it. I was tempted. I did have a very good offer from Caltech, at one point. But I had a lot of things under way already at the GE lab, lot of close friends there, and… I was more of an experimentalist than an academician, I think, and wasn’t this sure how I would fit in at Caltech. It might have been a good move, but I decided against it. I’ve had a very good career at the research lab in Schenectady. I was treated well, and had good problems to work on and, a lot of freedom. So, I didn’t make the jump. Some guys did. Lot of them worked out very well. Anyhow, that’s the way it worked.

Antitrust concerns in industrial research

Choi:

You mentioned the company policy of not allowing you to talk about things that haven’t been published.

Hall:

That’s widespread in the industry. You’re not allowed to do that for antitrust reasons. You had to be very careful.

Choi:

Can you tell us a little bit more about the antitrust reasons?

Hall:

Well, I think it’s just a pretty well established principle, at least in industrial research, that you can’t play favorites, you cannot conspire with some other industry, some other industrial organization, in an area that would exclude other people. You get into antitrust problems. It’s illegal. So you had to be pretty careful about that. And the lawyers cautioned us many times about this. We must never talk individually with fellow scientists from one other company in a sort of a closed discussion situation. You could talk sort of casually in the hallway, but you would stick to published information pretty much. I think you might stray a little bit, and sometimes hinted some things… but you know, you don’t want to undercut your own work either. You don’t want to spill the beans to some competitor, who’s going to scoop you with whatever he’s working on, because we have a lot of rivalry. This fellow at RCA, Jacques Pankove. He is a fellow that… his work and my work paralleled quite a bit. And I’ve never talked with him individually about things, but we were aware of each other’s works. So, we studied his publications, and he studies mine, I’m sure, and figure out what he’s up to now and try to… try to stay ahead of him. Isn’t that easy.

Choi:

Do you have any other researchers in your field, that you remember personally? C. W. Mueller is the guy that I’m working with, and E. W. Herold.

Hall:

Yeah, Mueller. Ed Herold, I knew of him, I didn’t… We didn’t mix much with RCA. I think mostly we talked with university people. I know that I never visited RCA lab, I can’t remember anyhow. There were some of the Air Force labs, they were fairly open, because their function is to encourage new investigation. So they’re pretty free, able to talk with people, and we talked with some of the researchers in the Air Force laboratories.

Choi:

What about the Signal Corps?

Hall:

Signal Corps, yeah. And when I was in the semiconductor… detector work, we would visit out at Berkeley, and talk with the folks at Lawrence Berkeley laboratory, who were very strong in instrumentation, this particle detector physics, electronics, they were very good at it. I learned a lot from them. But in that work there were no commercial rivalry. We were trying to work with… get the technology picked up in Lawrence Berkeley, so we talked quite freely with what’s going on, and there we had a lot of good arguments about how to do things and they’re doing it wrong this way and they think we’re doing something wrong and somebody’s explanation is either right or wrong, and needs to be extended. So we had a lot of good discussions of that kind. After I retired, then I did consult with other laboratories. I consulted with Princeton GammaTech, and later on I talked with the… what’s the group down in Oak Ridge? I would not talk with OrcTech, because they were rivals of this Oxford Instruments, as it is called now. So I would consult with one of the guys in Oxford Instruments, but not with any other organizations, because of confidentiality agreements.

Spread of new ideas, 1950 article

Choi:

It’s really quite interesting, because it’s a very different impression I get from the records. You said you didn’t have much interaction with the RCA people. That would mean… As soon as your Phys Rev article in 1950 was published, RCA people started on their own program of alloyed process, and they quickly scaled that up to small scale manufacturing process of making alloy junction transistors using indium dots on each end of the germanium…

Hall:

Well, you know, Saby, I guess, published in about that time. And he probably gave some oral talks. He may have given at a device conference, talked about alloyed transistors. So, once you do that, you got RCA guys in the audience. And they see what’s going on, and take the idea home and work on it.

Choi:

But still, the speed of progress and diffusion of ideas in the field is quite amazing.

Hall:

It is amazing, yeah. It’s amazing how fast things do get developed. This list, this list I gave you, happened in a very short period of time. Once you give a published talk on something, and people are in the field have workers already active in related areas, they can pick up quite rapidly and apply their own ideas and extend it. So things do happen pretty fast. Or for that matter, you know, during the war, just think how rapidly nuclear technology evolved from a, just a, beginnings of an idea in 1940, cutting it all the way through to a successful nuclear device in just four five years.

Choi:

But that took hundreds and thousands of top class scientists at one place. But here we have a small number of researchers in different locations.

Hall:

It’s not a tougher problem either…

GE and Japan

Choi:

What about Japanese connections, as far as you know, with the lab?

Hall:

The part played by Japan?

Choi:

Obviously a lot of knowledge went from the United States to Japan and I know that a couple of Japanese companies licensed GE technologies off…

Hall:

There was one Japanese, Nishizawa, in… what’s the island just north of Japan… Hokkaido? I think. He’s a very aggressive, he’s a self promoter, sort of a big operator. But he has a lot of very creditable work. He is in a university out of the mainstream of Japan, and has his own another empire which he’s pushing. But he’s getting very remarkable results from his laboratory and he had an early idea for a semiconductor laser, claimed to be the first inventor, maybe a year before the actual, we actually made the laser. Well, his claim for invention was an idea, which was not practical. It would never work. I guess he got a lot of internal credit as being the inventor of the semiconductor laser in Japan but it never really paid off because it was not a workable idea. But the Japanese picked up very rapidly in semiconductors. Leo Esaki was working in one of the laboratories. And Nick Holonyak, when he was in the services, he was stationed in Japan. He would make contact with Leo Esaki. Did he work for Sony, I forget. Anyhow, Leo had just made the first tunnel diodes, Esaki diodes he called them. And Nick Holonyak heard about this and told us about it. I don’t know if it was published by then or not, but anyhow he got us off to an early start studying these, we called them tunnel diodes or also called Esaki diodes. Pretty amazing device, we realized how it worked, and we found out how to make them out of other materials like silicon and gallium arsenide. Worked very well in gallium arsenide. And silicon, you could measure actual phonon energies. One of the lab newcomers to the GE laboratory had Jerry [Teeman] picked up this investigation of phonons from tunnel diodes of silicon, showing a remarkable phonon structure and being able to measure them electrically. And he got off on making electric devices based on tunnel devices. But the Japanese picked up very quickly on semiconductors because it fit their needs. Very limited amount of raw materials needed. They don’t have much power in Japan. But they had a lot of young fellows willing to work hard and pursued semiconductors very strongly, got a very strong position in semiconductors. And I guess Asia in general, you folks have been doing pretty well too. [LAUGH]

Choi:

Esaki eventually came over to the United States to IBM and spent his entire career there. Now he is the president of the University of Tsukuba.

Hall:

Is that right? University of Where?

Choi:

Tsukuba.

Hall:

Oh, back in Japan? OK.

Choi:

And that is one of the big technical universities. There is a science city in Tsukuba and at the center of the city is this university, which is very technical. So he’s doing well…

Rensselaer Polytechnic Institute

Hall:

RPI is trying to do this here, the tech valley… bring a lot of high tech stuff into this area now, pushing upstate New York.

Choi:

Why wasn’t RPI doing semiconductor transistor research at that time?

Hall:

Oh, I don’t know. It would take some guy that sees a new field, and it’s just a confluence of a number of things. You got to have funding and… what not. I guess they were in other technical areas, and never picked up in semiconductors until recent years. I don’t know. You get some guy who’s a spark plug and gets you a lot of big names drawn in or young fellows with good abilities, and get some courses started and get a nucleus going, then you can build up a department. I don’t know how you do it. That’s what managers are for, I guess. That’s not my department.

GE labs and factory

Choi:

So your experience was mostly around the lab… Were you aware of what was going on in the factory?

Hall:

Oh yeah. I would go there and visit some of the engineering groups in Syracuse. Mainly I was working on… Let’s see… I didn’t spend a great deal of time visiting other laboratories in GE. I would occasionally and often in selected areas, for example when they were needing this intrinsic germanium to start making their rectifiers. We were trying to get our developments of purifying germanium transferred to their production line so they could make their own germanium. We’d come visiting, they’d tell us their problems and we would try to figure out answers or maybe go home and work on that problem and try to solve it for them if we could. Yeah. We would visit there maybe 3 or 4 times a year, trying to see what’s going on. This was many years ago. It’s hard to remember all these things.

Choi:

And the factory was down at Syracuse, so it was about an hour…

Hall:

An hour and a half, yes.

Choi:

So it was not a distance you can go very frequently…

Hall:

It was a day’s trip, yeah. Not too bad. But I also talked to the people down… it used to be west Lynn and they moved down to Pennsylvania—farther down than that I can’t remember what the name of the city is—where they were making these high powered [thyrestors] and things like that. Because they had a lot of problems with getting a silicon prepared right and diffused to make the rectifier structures, how to get them sealed. There were a lot of manufacturing problems. We listened to their problems and tried to help out any way we could, but usually these were places where maybe we can make a suggestion or two, but we couldn’t take that kind of a problem back and work on it at home. They were production problems that… maybe we can offer suggestions and might follow up on these… sometimes we have ideas they will listen to and take advantage of it. But that’s sort of a… part of our job was to spread technology around and to go to the production, manufacturing units and listen to their problems and try to offer help if we can see where it might help to offer suggestions. Sometimes we would get a very interesting problem that we can take back and go to work on, but not too often.

Choi:

Can you give us a case of such instance?

Hall:

[Pause] I’m having trouble coming up with anything.

John Saby

Choi:

What was John Saby like? I’ve seen his photograph and he looks like a very lively person.

Hall:

Well, he was. He’s a very active guy. He bounced around different places in GE after he left Syracuse. I guess he went to the lighting department, didn’t he? I’m not sure. He came down with polio, and went around with a brace quite a bit. When I was a small boy, I also had polio, so I have a left leg that doesn’t work quite right. So I have a little sympathy for Saby in that regard. We were good friends, and we talked to each other quite often on our various projects in semiconductors.

Choi:

Thank you for the oral history.