Oral History Transcript — Dr. Jack Scaff
This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.
This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.
Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.
See the catalog record for this interview and search for other interviews in our collection
Interview with Dr. Jack Scaff
Dr. Jack Scaff; August 6, 1975
ABSTRACT: Work at Bell Laboratories, pertaining to development of the transistor. Main topics include his early education, work at Bell Labs under Earle A. Schumacher and work on dielectrics, copper oxide rectifiers, silicon, and microwave detectors.
Hoddeson:This is an interview with Jack Hall Scaff at his home in Bernardsville, New Jersey. Today is August 6th, 1975. You were born in Jackson, Tennessee --
Hoddeson:-- in 1908. What did your father do?
Scaff:He was a railroad engineer, with the Illinois Central Railroad. He was killed when I was 18 months old.
Hoddeson:I see, and your mother?
Scaff:My mother had moved back to Fulton, Kentucky, where she had grown up. She remarried, when I was about five years old and I grew up with a perfectly marvelous stepfather, one Samuel B. Smith, who was quite a successful tobacco exporter. He bought tobacco from the producers, the farmers. Packed it in Hogsheads, and shipped it abroad. It was an export trade. We sold to Italy, France, Spain, and the West Coast of Africa. Quite large quantities. .
Hoddeson:And then, were you educated in the local schools?
Scaff:I was educated in the local high school, which was in my opinion an extremely good school. They stuck to basics: science, mathematics, English, languages, and none of the things you might call frills. We had no physical education program at all, for example. There were outdoor athletics I mean, baseball, football, all of that -- but there was none of the fancy gym programs.
Hoddeson:Did you have any brothers or sisters?
Scaff:I had one sister.
Hoddeson:From your same father?
Scaff:Same father, yes. Older than I am.
Hoddeson:Where did your scientific interests stem from?
Scaff:I suppose it may have come from my father, who was actually trained as a lawyer, but he liked mechanical things. My father was trained as a lawyer, and had his legal degree, but he liked manual-things so much that he became a railroad engineer. I think the whole idea of mechanical gadgets and engineering just appealed to me right from the day I was born.
Hoddeson:Do any specific incidents stick out in your memory, from your childhood?
Scaff:Oh, just building everything you could imagine with my hands. You know, bird houses, you name it.
Hoddeson:What about your sister?
Hoddeson:Strictly a housewife.
Scaff:She was a musician, some training on the violin. But like so many people, she dropped it after a while.
Hoddeson:Let's see -- you went to the local high school and graduated around 1920 or so?
Hoddeson:'25, excuse me.
Scaff:That's an interesting thing there. I graduated from high school when I was 16. I'd skipped two grades in elementary school, something that I think most children could do if the school system were adapted to it. So I went I went to the University of Michigan when I was 16 years old.
Hoddeson:How did you make the choice to go all the way to Michigan?
Scaff:My stepfather had given me the choice of going to any engineering school in the country outside of the South. And the reason for that was that he felt that I'd grown up in the South, and part of education was to broaden your background by getting out of your local community. And when I entered college, the logistics of going to school were something of a problem. It took me a day and a half by train -- a night and a half a day -- to get to Ann Arbor from my home town. And the only other schools that I considered were the University of Minnesota, Cornell and MIT. But Cornell and MIT were just nearly impossible, in those days, from the western part of Kentucky. It would have taken three days, to get to Cornell. I still don't know how I'd have gotten there. But bear in mind, you took all of your worldly goods in a trunk for the winter, and had to move it to college. It's not like people getting into automobiles today, just driving. They didn't have cars for such distances.
Hoddeson:Did your father support you?
Scaff:Yes. He did, in part. I always had employment in the summer time, but I soon acquired some university assistance, which was modest in those days, but it did pay all my laboratory fees and $250 per year, which, for an undergraduate, was quite a good stipend because laboratory fees amounted to almost $500.
Hoddeson:Let's see, I've come across the fact that you were an assistant in physical chemistry at Michigan --
Hoddeson:That's what you're referring to now.
Scaff:Yes, it is.
Hoddeson:Was that for some particular professor-instructor?
Scaff:It was for Professor Lee Case, who was an experimental physical chemist. He put me in charge of the physical chemistry laboratory there, which meant preparing absolute alcohol of proper boiling point and other kinds of experiments. of that kind. A lot of the work involved precise laboratory preparations, something I enjoyed very much.
Hoddeson:Had you applied for that job?
Scaff:Yes. But I'd been a good student and I belonged to a chemical fraternity, Alpha Chi Sigma. And so many of the graduate students who were then junior instructors were also active in that fraternity. They had recommended that I apply for the job, because they thought I could do it. And while it was unusual for an undergraduate to do it, they thought it would be possible. And it was possible, and something that meant a good deal to me -- prestige, recognition, having a private laboratory when I was only a junior in college, you know.
Hoddeson:Oh, sure. About how much of your time did you spend on the job?
Scaff:I'd say it took two afternoons a week.
Hoddeson:That certainly must have been a significant experience, I'm sure, in channeling you along your career direction.
Scaff:Yes. Influences in college...
Hoddeson:Were there any other that were also important?
Scaff:Well, I think only the fact that we had some extraordinary professors. I had a professor of organic chemistry by the name of Moses Gomberg, who was enormously impressive. I had a professor of mathematics, by the name of Clyde Love, who had written all of the Midwestern textbooks on mathematics, differential and integral calculus. And he made mathematics so simple that I couldn't understand why anyone would fail to pass the course. An extraordinary teacher.
Hoddeson:Were there other teachers who impressed you?
Scaff:Well, those two stand out in my mind.
Hoddeson:Any physics teachers?
Scaff:Yes. We had Randall as professor of physics, and F. Lawrence Bigelow as professor of chemistry, and these were all extraordinary people.
Hoddeson:This period when you were in college was an extraordinary time in physics.
Scaff:But you see, I was studying chemical engineering, so physics was just a requirement, something I had to endure. But we did endure it, and I think we had good basic training in physics.
Hoddeson:I see, then you probably weren't aware of the new developments.
Scaff:No. I had heard the words quantum mechanics, but this course was just being offered in graduate school. At that time, these subjects just weren't in the regular curriculum.
Hoddeson:There was an exciting summer school going on just at the time you were at Michigan, a summer symposium in theoretical physics.
Scaff:They always had summer schools there. During my tenure there as a student assistant in physical chemistry, they brought over a famous Dutch professor in colloid chemistry Professor De Kruyt as I remember the name.
Hoddeson:So, they did that in chemistry, as well as in physics.
Scaff:Yes. But I didn't go to school in the summer. I did attend one school summer, because I needed to take physical chemistry laboratory, before I could become the assistant. So I had to stay in the summer and take the course.
Hoddeson:What did you do during the other summers?
Scaff:A professor of chemical engineering there, by the name of Warren Badger, was a very well-known man, and had an outstanding chemical engineering laboratory. This had been given by one of the industrial companies, Swenson Evaporator Co., with the understanding that Badger could do technical work for them in the summer time, with the equipment installed in the university building. So we operated evaporators and did all kinds of crazy things, made sugar out of old brown molasses.
Hoddeson:Sounds like fun.
Scaff:It was fun. Especially it was fun when we made rum.
Hoddeson:Right. Well, then you got your BS in chemical engineering, in 1929, and went to Bell --
Scaff:-- went to Bell --
Hoddeson:-- immediately after graduation.
Scaff:Right, in July.
Hoddeson:How did you hear about the possibility of a job at Bell?
Scaff:Well, I'll tell you the whole gory truth. We had a graduate from chemical engineering who'd gone to work for Western Electric Co. at the Hawthorne Works in Cicero, Illinois. And before the telephone recruiters came to the campus, they asked recent graduates who the promising possible employees might be. And Calvin Corey had given my name as a possible employee. They were hiring so many people in those days that they were there for three days. And I deliberately waited until about 3 o'clock in the afternoon of the third day, because I just wasn't interested in the Western Electric Co. or the telephone company, anything about it. I deliberately waited. But I wanted to go in and at least see these people, as a courtesy to my old friend Corey. So I did at 3 o'clock, and the Western people said, "Well, too bad, we've filled all our openings, but please talk to Mr. Long of Bell Laboratories." I said, "I'm not interested in Bell Labs. They said, Oh, you don't know what it's all about." So as a courtesy I did talk with them. They offered me a job. I didn't answer the letter, because the salary was not in line with other engineering graduates and chemical engineering graduates at the time, so I didn't even bother to answer it. Finally I got a follow-up letter from, I forget his name now. It doesn't matter. But he invited me to dinner, and I went back on a kind of a follow-up, and he talked me into coming to Bell Labs for the summer. I came with the clear understanding that I could go back to school in the Fall if I wanted to, without any prejudice. In those days, I think you just didn't feel as mobile as people do now.
Hoddeson:So you were thinking about going to graduate school --
Scaff:Oh yes. I was intrigued with some of the oil companies and their offers for foreign employment. The pay was extremely good.
Hoddeson:Do you remember the ball park of salaries in those days?
Scaff:Oh yes. I know exactly what my salary was. When I came to Bell it was $30 per week.
Hoddeson:This was in '29.
Scaff:Yes. And believe you me, apartments in New York were not that different in cost from what they are now, either. And a suit of clothes, a decent suit, would be $60, $75. That's not a very munificent sum. But the oil companies were paying much better. They were paying $50 a week and up. Which doesn't sound like much, but in those days it was a big difference.
Hoddeson:But you stayed with Bell.
Scaff:Yes. I got interested in what I had an opportunity to do.
Hoddeson:You began working on magnetic materials.
Hoddeson:In the early days?
Hoddeson:And whom did you work with initially?
Scaff:Well, I worked for Earle Schumacher.
Hoddeson:You worked with him for over 20 years.
Hoddeson:Were you immediately placed in his group?
Scaff:No, I had a supervisor, I guess I was a little difficult to supervise, because I seemed to get a new one every year, until finally I didn't have any.
Hoddeson:I'm not sure I understand that exactly. Wasn’t Schumacher your supervisor?
Scaff:He was a department head. See, my first supervisor was Lawrence Ferguson, who interfered with my work so much, I went and told the boss. I said, "Look, if you'll keep him out of my laboratory, I'll get the job done." He was interfering with a glass blower. I was building some glass equipment, and he irritated the glass blower no end. So after that, I immediately was moved to report to George Bouton.
Scaff:George Bouton, yes. I guess he was my last supervisor.
Hoddeson:Let's talk a little bit about the research atmosphere in those early days. Did you work six day a week or five days a week?
Scaff:When I started out, we worked 5 1/2 days. And shortly thereafter, we --
Hoddeson:-- the Depression came.
Scaff:Came the Depression, we went to five days a week and later four.
Hoddeson:And your salary was cut proportionately?
Scaff:Salary was cut proportionately. But at least you had a job. I was one of the few people hired in 1929 to survive the Depression.
Hoddeson:I don't understand what you mean. lots of people survived.
Scaff:Not at Bell Labs, as employees. People hired in 1929-30 were terminated. Yes.
Hoddeson:I wasn't aware of that.
Scaff:Well, they had the misfortune in 1930 of having some 450 acceptances. And these offers had all been based on expecting only 50 percent of them to accept. Well, by the time the Depression had gotten that far down the line, jobs were so scarce that practically every offer that they had outstanding was accepted. So they ended up over-hiring greatly, particularly in 1930.
Hoddeson:There was just before the Depression a surge of hiring at Bell.
Scaff:Yes. There was then.
Hoddeson:They must have thought that they were going to expand.
Scaff:Well, they were expanding, business was growing.
Hoddeson:So Bell was hit doubly, because first of all, they hired so many people and then because in a time of Depression everyone accepted.
Scaff:Right. But you know, to put that all in perspective, I know any number of people who were required to change jobs during the Depression. In fact, every one of them improved his long term prospects. It's perfectly true. Most of them ended up as presidents or vice presidents.
Hoddeson:The Laboratory was in New York?
Scaff:It was at 463 West St.
Hoddeson:Yes. And I gather you lived in the Village?
Scaff:I lived at 15 West 12th St., in the Village. After I was married, I moved out to Queens. I lived in Forest Hills and Kew Gardens long before the subway opened up. I commuted by Long Island Railroad.
Hoddeson:How large was your Lab? How many people were it? I'd like to get some sense of the physical atmosphere.
Scaff:Well, the facilities at West St., and particularly for the people in chemistry, were quite poor indeed, because the New York City fire laws required that all chemical laboratories be on the top floor.
Scaff:So because of the fire problem we were scattered. from very old sections of the building, to newer sections, but all different levels. My own laboratory had wooden floors, the crummiest setup you'd ever want to see. But it was adequate in terms of the space and availability of facilities, something we never lacked.
Hoddeson:You never had any trouble getting supplies?
Scaff:No, or equipment.
Hoddeson:Did they give you assistants?
Scaff:Yes. Oh yes.
Hoddeson:So these were easy to come by too?
Scaff:Reasonably easy to come by. Yes. There were three great people in Bell Labs, associated with me at one time. Henry Theuerer is one. He was my second technical assistant, replacing Frank Schnettler, who'd left Bell, just as the semiconductor work started, to go to. Columbia to get his Ph.D. in metallurgy under Professor Jette. Later he went to the Manhattan Project at Los Alamos. He's back at Bell now, hired by Jim Fisk, by the way, from Los Alamos.
Hoddeson:About how many people were working together in the same room? There were ten people in Schumacher's group in 1931.
Scaff:Oh no, no, no. For example, I was in a seperate room. One, two, three, four -- there was more than one room per person. Because the nature of the metallurgic facilities was such that some of them required large areas. You couldn't have a rolling mill sitting on top of a postage size desk.
Hoddeson:Were you encouraged to write scientific papers at that time?
Scaff:I've must say yes and no, because the publication policy of Bell was by no means as open as it is now. Particularly, when it related to technological processes, and things that might give you a proprietary advantage, in making better magnetic materials for example. So, science was always published but the kind of applied science and technology that I was concerned with was not so frequently published. And in fact, there's some work on purification of magnetic materials that we never did publish, and in retrospect, it should have been, because it was an interesting bit of work on chemical refining reactions for eliminating harmful impurities.
Hoddeson:Was this written up as a technical memorandum?
Scaff:Yes, they were all written up in technical memos.
Hoddeson:I see, so they're filed away some place.
Scaff:They're around there, somewhere.
Hoddeson:I hope they're catalogued.
Scaff:Oh, I’m sure they are.
Hoddeson:In those days, were there many seminars, talks, and visitors generally?
Scaff:Yes. Yes. Again, I don't think these things were as frequent as they are now, but there were fewer people, so they didn't need as many seminars. But we had our technical symposia that met in the West St. auditorium once a week with invited internal speakers, to talk about their programs. I gave a talk there on gases in metals at one time. Not only that, but we were encouraged to give informal work seminars, and I think one that will interest you is an individual study program, on Mott and Jones' book on solid state physics. I don't know who organized it. It may have been Walter Brattain, but one of the --
Hoddeson:-- oh, I know which group you're talking about.
Scaff:Bill Pfann was a member.
Hoddeson:I didn't realize he was a member.
Scaff:He was. And they went through that book, week by week, with assignments and discussion, classwork and problems. Completely informal, but nevertheless, a very elegant study program.
Hoddeson:Did you attend any of the meetings?
Scaff:Did I? No, I did not, because was just not my forte. I was a chemist, engineering type. solid state physics and particularly an engineering type.
Hoddeson:Who would you discuss your work with most often? How closely did you work with Schumacher, for example?
Scaff:Well Schumacher. He's dead now, but he would know my work.
Hoddeson:On a day-to-day basis, when you did something you were happy about, who would you tell?
Scaff:Well, I didn't worry about that, I would say. I tell you, we had this man here (pointing to name on chart).
Scaff:Yes. Periodically, we would have work reviews with him. And we had a consulting metallurgist from Yale University by the name of C. H. Mathewson metal for whom the Metallurgical Society is named, a very famous man. And he would come down once a week. And we would be paraded to discuss our work with him. I’d say I got more satisfaction out of interaction with him, because he was a very perceptive man, and was interested in what you’d been up to, why you were doing it, and what your whole thinking process was, you know.
Hoddeson:Any others visitors who carne by regularly to talk to you besides Mathewson?
Scaff:Well, of course, we always had technical visitors from various other companies, like Westinghouse, GE, --
Hoddeson:Can you tell me more about these?
Scaff:Well, BTL had a patent license arrangement with both of those companies, and particularly, with GE a very close working relationship, because our interest in magnetic materials was different from theirs. But nevertheless, they did impinge on each other. GE had a program that was quite complementary to ours, in the sense that they were developing so-called grain-oriented silicon steel for transformers, and distribution transformers particularly. We found uses for these alloys in the Bell System, too. The same sort of technology applied -- that is, control of the low quantities of impurities, and new impurities that we were concerned with were the same ones that they were also, but different alloy systems. And they were carbon, oxygen, nitrogen and sulfur, so-called "interstitial” impurities, that in some funny way adversely affects the magnetic softness, if you want to call it that -- the permeability and those things, and low field strength properties. So we had very frequent technical interchanges and discussions, because our respective technologies tended to complement each other. And were not in any sense competitive.
Hoddeson:Did you interact at all with Bozorth?
Scaff:Oh yes, sure --
Hoddeson:-- in his magnetic work?
Scaff:Oh yes, indeed. He was in physics. But from the materials preparation and purification points of view, we had a very strong input into his general activities. Magnetics research was, generally speaking, dominated by the physics people. And we were more the technological people, you know.
Hoddeson:Did you talk with him frequently?
Scaff:Oh yes. There was very, very frequent interaction, because he was interested in alloy development, and we were the people that prepared the alloy. And heat treated them. Bozorth was actually a physical chemist himself.
Hoddeson:Did you ever participate in any professional summer schools?
Scaff:R. N. Burns, who was one time chemical director encouraged me to go down to one of the programs they have in the summer, the Gordon Conferences. In those days, the Gordon Conferences were held down at Fisher's Island, outside of Baltimore, and it was quite a honor to be invited to go down there. But I never went because I just didn't want to take the time. In retrospect, that was a mistake.
Scaff:Well, I think you broaden your contacts. The technical interchange was stimulating.
Hoddeson:Did Schumacher go?
Scaff:No, he didn't. In fact, those conferences still are dominated by individual contributors, and not by supervision and management. They're run by AAAS now. In fact, they were then run by AAAS also.
Hoddeson:Who chose your research problems in the early years?
Scaff:Well, I did, although Earle Schumacher had a thing about gases and metals, as being something that ought to be investigated. I think it went back to his early days of working with vacuum tube materials, where gas evolution of an alloyed filament could be deleterious to the life of the vacuum tube.
Hoddeson:Had he worked in the vacuum tube department?
Scaff:Yes, he had. In fact, that's why he was hired. He had been hired shortly after the discovery of the three element vacuum tube, and the recognition of its importance in transcontinental communication. They began a program to develop improved vacuum filament materials. He was particularly concerned with rare earth thermionic emitters, their preparation and the technology for putting them on filament. When this whole program started, the ordinary vacuum tube filament would be driven by currents of one ampere or more. This was reduced to fractions of an amp, over a period of years, before the transistor came along and revolutionized electronics. But he'd studied gas evolution from glasses, and from various materials used in the vacuum tube itself, and I think he was just interested in gases and metals. So I undertook such a project. But the specific work and the specific directions were things that I chose. I never had detailed technical instructions.
Hoddeson:Did he continue in his interest in vacuum tubes?
Scaff:No, like everything else, the things grew up, and then They were moved into the vacuum tube development organization, then headed by Mervin Kelly. So gradually, while for many years we -- not we, but this group here, H. T. Reeve -- produced all the filaments that went into the Bell System tubes that were manufactured.
Hoddeson:Did you interact much with Mervin Kelly?
Hoddeson:Or did your group?
Scaff:Certainly these people did, all of them.
Hoddeson:Schumacher, Reeve and White, J. H. White. In a sense they were servicing the vacuum tube, weren't they?
Scaff:That's right. These two here, Harris and Schumacher, developed the filaments that Kelly's tubes used -- platinum nickel alloys, as I remember. This fellow (pointing) fabricated them --
Scaff:And this one (pointing) produced all the early permalloy.
Hoddeson:White. Did Mervin Kelly stop in at the Labs and see what was going on once in a while? Did you see him much?
Scaff:I didn't really have any contact with Mervin Kelly in this program, because I wasn't concerned with it. What I know about vacuum tube filaments is just what I learned indirectly, at this time. I did interact with Kelly very strongly during the war, though.
Hoddeson:We're going to come to that, I'd like to know about that. I was wondering about the degree of interaction between the various departments. I'm getting a picture of a rather strong interaction, between the physicists and the chemists and the vacuum tube people.
Scaff:There was indeed, because as I say, the whole permalloy development was based on work that these two did. And the development of the present -- even the present receiver diaphragm material, which is an alloy of equal parts of iron cobalt with 2 percent vanadium, 48-48.2 percent vanadium, so called vanadium permandur -- the thing that made that material, workable, that is something that could be rolled into strip, from which you could make a telephone diaphragm -- was controlled by a patent issued to Charles Wahl and J. H. White.
Hoddeson:Were you also aware of the work being done by Morgan and others in dielectrics?
Scaff:Well, yes, but to a much lesser extent at that time. I interacted more with Kohman, a physical chemist. I had various interesting little projects with him.
Scaff:He was concerned with dielectrics too, but particularly with paper capacitors. Some of the capacitors were made from mica with electrodes from electro deposition of a metallic film. Frequently some electrolyte was incorporated in the deposit, and this led to deterioration of the condensor. Well, Kohman thought there might be some way of depositing metals from a non-electrolytic solution, non-aqueous solution, and it occurred to me that you might do this by decomposing carbonyl compounds. Nickel carbonyl is a compound that can easily be formed at a low temperature and decomposed at a somewhat higher temperature, to form thin nickel film, involving no aqueous electrolyte. And we did experiments with Kohman like that. Didn't work, because we couldn't get these deposits free of pinholes. I suspect that if we'd known how to clean surfaces then the way we do now, that we could have made a success of it.
Hoddeson:But I guess those earlier experiences helped later on anyway.
Hoddeson:At this time, were there colloquia that people were attending in the Bell Labs research groups? Or did those become more important in the later years?
Scaff:They were held regularly. But as I recall, always after hours on our free time.
Hoddeson:I'm curious to learn of your interactions with some of the other people who were around in those days. Joe Becker, for example.
Scaff:Yes. Well, there was something that I think really should interest you, because Becker and Brattain were very much involved in the copper oxide rectifier development They developed the copper oxide varistor. The basic technology that had come from a man by the name of Grondahl, who worked with Union Switch and Signal Co., a subsidiary of Westinghouse. Becker and Brattain did adapt this to Bell System needs, worked up the processes that we utilized. Now, these things required a very special grade of copper from Chile, and therefore we were involved in investigating the nature of this Chile copper, why some lots worked and why some didn’t. We never solved this one. We were on the verge of solving it after the War, after mass spectrometers for solids became available, but then the transistor came along and we didn't need the solution. I'm sure we could have solved it. But in any event, they did have to buy special grades of Chile copper that required special processing, and this chap here, Smith.
Hoddeson:G. O. Smith under Reeve?
Scaff:Yes. Developed a thallium copper alloy that was used in telephone click reducers, which is a symmetrical varistor, that has a very high resistance at low voltages, and very low resistance at slightly higher voltages. And it bridges the telephone receiver, so as to discharge heavy voltage surges as, for example, from lightning. The principle is that of a thyristor for much lower voltages. A pair of these was in every telephone set. And I believe the copper oxide varistors, the modulators, and these click reducers, were the first uses of solid state devices in the Bell System.
Hoddeson:This was in the late thirties, then?
Hoddeson:Did you interact at all with Teal, or Gibney?
Scaff:After the transistor discovery.
Hoddeson:We'll come back to that then later on. How about Darrow? Did you see Darrow much?
Scaff:We read his articles in the Bell System Technical Journal, which ended up I guess as a book volume as he was a very gifted writer. So I would say we all interacted with him by reading his papers, more than by personal contact.
Hoddeson:I see. You didn't see him very often?
Scaff:No. In fact, I don't know how often he was at Bell Labs, really. He was writing and doing things. Very much of an independent soul.
Hoddeson:How about Alan Holden? He was off in publications then I think.
Scaff:I'm familiar with Alan's early work on crystals, but I did not interact with him very much. In fact, not at all.
Hoddeson:Now, let's see. Pfann joined in '35 or ‘36.
Hoddeson:Did you start working with him right away?
Scaff:No, he was assigned to Evan Thomas in metallography. And while I was an observer, and recognized him as a very bright young lad, and one that had a great deal to offer, it wasn't until we became involved in this silicon project.
Scaff:And that was a job that had a high priority, even before the War. I managed to get Pfann assigned to do the metallographic work, and that preceded his direct association with me.
Hoddeson:I see. Now, Theuer was working in '36, doing work on finishes and then he was working on surface chemistry in ‘37.
Hoddeson:And then in ‘38 he joined your group.
Scaff:Surface finishing -- he developed a means for testing the mechanical properties of organic films, coatings, paints. It involved rotating the film around a tapered Mandrel, so he could measure the exact extension at the point of fracture. This test was used world-wide. And I doubt very much whether many people would ever think of that, but he was a very clever person, and one who showed even at that early day how creative he was. And his particular creativity was unusual, because it usually resulted from a very specific need. No sooner did you point out to him a very strong need for something, than he would devise a very clever way to provide the answer. The result is that while his inventions, in total number, were not extraordinary by Bell System standards, -- I believe he had 28 or 30 patents -- the fraction of them that actually received practical use was very high indeed.
Hoddeson:How did you hear about his work?
Hoddeson:How did you happen to choose him as an assistant?
Scaff:He had moved from the finish department over to work in vacuum tubes. When Frank Schnettler left me to go to Columbia to get his Ph.D. in metallurgy, the chemical director decided that Henry Theuerer would be a good one to take his place. And this began a long and very successful relationship, I think.
Hoddeson:I think so too, judging from the papers that came out. Well, now we're actually moving into the later thirties, and I'm just wondering, before we start talking about the silicon work, whether there’s anything important that I left out in the earlier period.
Scaff:I don't think so. But bear in mind that Henry and I, at the time the silicon work started, were studying the removal of trace quantities of sulfur and other deleterious trace elements in magnetic alloys. And we tried any number of approaches to it, and studied each one of them, with the whole idea of coming up with a technological process for making purer and better magnetic materials. And so we had developed facilities for producing metals in a very pure state, furnaces, reduction furnaces and things of this kind. And it was the availability of these facilities that got us involved in the silicon work.
Hoddeson:I was going to ask you about that. Friis was exploring the high frequencies in radio communications.
Scaff:Before we get started on that, I think I'll go get myself a glass of water.
Hoddeson:OK, fine. (Break)
Scaff:In the development of copper oxide varistors, Walter Brattain and Joe Becker were the theoreticians who interpreted the performance of these gadgets in terms of' physical theories. And they were quite well aware of the role of stoichiometry in controlling --
Scaff:Yes. Exact chemical proportions of cuprous oxide, Cu20 rigorously -- but a departure from that, having either a deficiency of oxygen or an excess of oxygen, could affect conductivity. And they referred to the type of semiconductor as excess and defect semiconductors. The defect semiconductor being I believe one that would have a deficiency in oxygen in the cuprous oxide structure.
Hoddeson:This was in about the mid-thirties?
Scaff:About '36, I'd say. That could easily be documented. 36 or '37. The point is that you couldn't think by analogy to explain the unusual effects we had observed in the early silicon ingots we had prepared. When I first suggested that these effects were due to impurities, Walter Brattain told me he's a very direct fine person and he didn't mince words -- he told me I was crazy. It just didn't fit the accepted pattern of thinking at the time. I said, "Well, I may be crazy, but that's what I think it is." And of course, in retrospect, we were absolutely right on it.
Hoddeson:Now, that's a little bit later, isn't it?
Scaff:Yes, it is later.
Scaff:Yes, Now, that is fairly well documented, in that paper.
Hoddeson:Yes, that's where I got it from. I'd like to go back to how you got involved in silicon in the first place. You were working on --
Scaff:Well, several people at Holmdel, and I'm sure this early history is not entirely clear -- but Russell Ohl was one, A. P. King was another, George Southworth another, somehow, they all got interested, in crystal detectors. They were then working at what we thought to be very high frequencies, 40 centimeter wavelengths. The inter-electrode capacitance and transit-time effects were beginning to bother them, in their circuit technology and in their electrical measurements. And someone got the idea of coming back to the old crystal detector. Now, I'm not sure which one of these three first got the idea of going back to crystals, but Ohl certainly is the one that began the systematic exploration of crystal detectors, and especially from the point of view of materials to be used, because he was an electro-chemist by training. And he was the one who found that iron pyrites or iron sulfide was a very fine point contact rectifier material, but the surfaces weren't uniform -- they were spotty. He decided silicon was ideal material for this use.
Hoddeson:You say he tested about 100 materials.
Scaff:Yes. It was actually --
Hoddeson:This is in the Bell Technical Journal Article.
Scaff:Yes. That article was a great disappointment to me, because I wrote it for the Bell Systems book on the War contribution. And I was unhappy to be a joint author with Ohl because so much of the work in my department had been done by Bill Pfann, Bob Treuting, Karl Olsen, Henry Theuerer and others, and I didn't want to report their work or preclude their publication later. But they wouldn't permit me to have those people as co-authors on this article, so what I did was to leave this open enough so that we could write seperate papers later, to fill in the gaps, you know.
Hoddeson:Oh, I see.
Scaff:So this is a very broad brush treatment, and there is ample room for a half a dozen more papers. Then the transistor was discovered and we just got busy -- so the whole War development was never fully published but it's all available in technical memos.
Hoddeson:Now, I interrupted you --
Scaff:Well anyway, Ohl had gotten some silicon, and he'd decided, chemically, and for various other witchcraft reasons, that silicon was the element to work with in these point contact detectors. He went to R. O. Grisdale, but Grisdale didn't have the facilities. He did however fuse a small piece of silicon in an oxygen-hydrogen torch, but this is not the way to do it. It was however a means of making a start. But Ohl knew I had the facilities for melting pure metals, and he came to me and asked me to melt some of this pure silicon for him. I had the facilities for preparing fairly pure substances, from the magnetics work. I had the furnaces, various kinds of crucibles and the like. We were probably the only people who were set up to produce small quantities of fairly pure substances in Bell Labs at the time. And it is a very interesting tale, how narrowly we missed our discovery, because by dumb luck the first sample that Ohl examined was n type-silicon with superior rectifying properties. We had difficulty in repeating this first result, and particularly we had problems with the silicon cracking on freezing. We didn't realize then that silicon expands on freezing. We thought the cracking was due to thermal shock from cooling too rapidly. So we did the logical thing, and cooled it slowly. We sent some beautiful ingots down to Ohl.
Hoddeson:You're talking about you and Grisdale.
Scaff:No me and Theuerer. By this time, Grisdale was out of it. He just started the program and was glad to turn the job over to us.
Scaff:I think if he'd realized what was going to happen he wouldn't have done that. We were really just doing this fellow down at Holmdel a favor, melting stuff for him. It was not our regular job.
Hoddeson:Was Grisdale the natural person to ask to do this?
Scaff:Yes, for he was involved with Stan Morgan in some of the dielectric work and some of the semiconductor work involving silicon carbide. So he was the logical one, yes. He also by the way, was a chemical engineer. In any case, we sent these beautiful ingots down to Ohl and The interesting thing is that Ohl would send these ingots out to be cut into pieces for physical measurements and other experiments. We didn't have a diamond saw in Bell Laboratories at that time. He would send them out to a jewelry outfit down in Perth Amboy, to grind them and cut them and make the samples, and had produced some beautiful little samples, oh, about an eighth of an inch in diameter, and an inch long. Nice cylinders. And he tried to get Joe Becker to study the electrical resistance of these samples, and Joe would have no part of that either, because he couldn't see what was to be gained from measuring these samples. So finally Ohl in very great disgust decided to measure the samples himself. They'd laid on the shelf for months.
Hoddeson:Why didn't Ohl want to do his measurements himself?
Scaff:Well, I think he was trying to bring expert people into the pictures. Brattain and Joe Becker were the obvious ones to bring into the picture. But he got nowhere with them. And finally, in great disgust, he decided to measure the things himself. And every voltage Ohl ever measured was measured with an oscilloscope. He never used a voltmeter. And he had this thing set up on the oscilloscope, current and voltage leads on it, a light, a bench light nearby with an electric fan cooling the room off, and the fan interrupted the light, to the place where, first thing you know, he realized he was developing a voltage in this sample, and the voltage was being interrupted by the pulsating lights from the fan interrupting. And that's how Mr. Ohl discovered the pn junction and photovoltaic effect in silicon.
Hoddeson:I guess he was a little bit older?
Scaff:Yes. Yes. I’d say ten years.
Hoddeson:That's fascinating. Did you all realize right away that these pn junctions might be very important in rectification?
Scaff:Certainly, after reading Kingsbury's memo. Incidentally I turned that memo over to the Bell Labs historical archives --
Hoddeson:Who did you turn it over to?
Scaff:Mort Fagan. Yes. It was in my files, and I didn't see any reason why I should keep it, and I turned it over to them, because Kingsbury did a very thorough study on the photovoltaic properties of this cell. I've got, in my paper there, I've got three -- Here’s one of the curves, on the paper, but that's only one of them. I thought I had it --The minute we realized that this silicon cell had its --
Scaff:The minute we realized that this cell had a peak of sensitivity in the infra-red, we sensed its importance to solar energy. So, the truth is that this was the first siliconphoto cell -- I've still got the sample, by the way in my desk upstairs.
Hoddeson:I'd love to see it later. Brattain then got involved, didn't he?
Scaff:Yes. This work was interrupted by the war. And that, the war program, which is also documented in here, was done wholly in my department, with electrical collaboration from the people in Holmdel, Harold Friis' organization.
Hoddeson:Let's turn to that now -- unless there's something here that isn’t down in the publications.
Scaff:-- I think it's pretty well described there
Hoddeson:Perhaps some of the social or human aspects aren't described --
Scaff:I well remember the excitement I felt, when Russell Ohl called me one Friday afternoon, to tell me about the discovery of this photocell. And I was more interested in getting out on the golf course than I was in listening to this photocell stuff, you know, that particular afternoon. But we had almost daily telephone calls, interchanges, because we were doing the preparative side and he was doing the electrical side. And this was, it was in such a telephone conversation that we decided to name these two types of silicon p type and n type. And we did it for reasons that are documented in there. By that time, we knew that p type material was developing a positive thermal emf against metals, current was flowing in this little internal battery, it was going from positive to negative, so -- I later got into a dispute with Professor Lark-Horovitz at Purdue, because in a paper, a talk I gave after the War, I had referred to these impurity effects, and he wrote me a letter and said, "I'm surprised you didn't cite my prior work…” He gave me a classified conference he had attended up at MIT. I wasn't even there. And he had apparently quite independently chanced on the same ideas and the same terminology. But in a letter to him I explained, "Well, look, I'm sorry to tell you, but this term first appeared in a Bell Labs memorandum,1J and I gave him the date of it. And it settled the dispute. I guess that memo's referred to in here.
Hoddeson:Then, the War came, and as you say in one of your articles, Hour scientific studies of semiconductors were interrupted at this point because of the emerging war effort, implying that instead of scientific studies, you were --
Scaff:-- we were building the first detectors for microwave radar, and seeing to it that they got manufactured for the war effort.
Hoddeson:How big an operation was that?
Scaff:It was pretty large. We must have had 15 people involved at Bell Labs. And of course, Western set up, as I think is documented in there:" one of these places -- they set up, first at Kearny, later at Clifton -- to produce these detectors on a commercial scale -- and were highly successful at it. But all of the technology, and there's a lot of it, that isn't documented here at all, the techniques for putting these things together, for protecting them from moisture, for carrying out the electrical adjustment, the electrical evaluation, quality control to assure high production yield and all that was something that required our day-to-day attention. This was a six day program. It got to be pretty rugged, too, because this detector here, the IN 25, was intended for classified Navy project, and I don’t really know to this day --
Hoddeson:(Figure 13, in the historical article).
Scaff:I’m not sure it’s declassified to this day. But it was a very fascinating program. It was intended for joint use by the combined Allied and American fleets in the pacific, in case an invasion of Japan were required. And after VE Day, the Navy I think was fearful of a morale letdown or an emotional letdown, so I think they just manufactured reasons for insisting that this project get all sorts of priority. And the result is, I frequently spent three nights a week out at the Clifton factory because with the factory running full blast, it was difficult to do any experimental work, and we had to get some answers, to what some of their problems were up there. So several of us would go up there late in the afternoon, work up there until 11, 12 at night. We'd do that three days a week and then work six days besides. And after that you've just about had it, I’ll tell you.
Hoddeson:Was more money poured into the research during this period?
Scaff:You didn't need money. I mean, the heck with money. It was just a question of getting the technical problems solved and the manufacturing problems solved, so that this gadget would be available. It was a real interesting project. One thing I had not alluded to in this article, but I think it's of historical interest, especially to physicists: we had quite regular interchanges of information with the Radiation Lab and with the British equivalents. And they had liaison people that --
Hoddeson:Who were some of them?
Scaff:Well Charles Purcell was one. He later became a Nobel Prize winner, at Harvard, I think, and Bleany at Oxford was another, Nobel Prize winner. These fellows were visitors every two or three weeks, and they were carrying back and forth the latest technology. So our work benefited by what the British were doing, and I'm sure they benefited by what we were doing, too.
Hoddeson:So there was this triangle, Cambridge, Mass., the British and Bell Laboratories.
Hoddeson:What about Columbia? Did they get involved too?
Scaff:Columbia was involved in magnetron work, but not in the silicon detector work. They were involved in radar.
Scaff:They were involved later. Incidentally, during the very early stages of the War, long before we declared war, after Britain was involved, and before they had gotten aboard their own technology for making silicon, we were shipping silicon of this type over to them. Not the same size, but the same general technology. We were sending samples to them, as well as samples of our detectors, over in the diplomatic pouches as well as complete technical information.
Hoddeson:What are diplomatic pouches?
Scaff:Well, it's a way to get classified material hand carried, over to your buddies over on the other side, without going through the mails. It was essentially a locked up case, that some trusted military courier had chained to his arm to carry it.
Hoddeson:That's fascinating. You were going to tell me how Mervin Kelly got involved with your work in this period.
Scaff:Mervin Kelly was in overall charge of Bell Labs, period.
Hoddeson:He was director of research at the time.
Scaff:Yes. We'll have to go back to the chart -- but I think that's right. He may even have been president. I think he was. In any event, he was very very much involved in the microwave radar program. And one of the key things in the microwave radar program was this silicon detector… The vacuum tube detectors were from my point of view passé: We improved the detectors so much during the War that we picked up six DB in receiver noise figure, which is equivalent to a vast increase in radar transmitter power. Now, the magnetron people were doing the same thing in terms of increasing the power output of the magnetron. All this meant increased range for radar, lower weight, especially important for airborne radar. What is not widely remembered is that the Western Electric Co. produced something like 60 percent of the radar used during World War II. And the radars they produced, Western, always were the more complicated types, more precise and more difficult to manufacture. The easier ones, they farmed out to subcontractors.
Hoddeson:So Mervin Kelly directed this.
Scaff:He did the whole bit, yes, and this was important enough to him that he very frequently came to see us once a week, to see what the latest progress was.
Hoddeson:Did he keep up with the technical progress?
Scaff:Oh, I'll say he did. Certainly he did. In planning a new radar system, he would have Harold Friis who generally was an overall systems man, he'd have Jim Fisk, who was in charge of the magnetron work, frequently Gordon Thayer, who was another systems man in charge of one of the airborne radar, certainly I was included, because we were trying to estimate how much each part could be improved for the next generation of radar systems. There was a planning meeting, prior to getting the military to approve a development project of a particular system. But let me tell you, he followed it closely... There's a very interesting tale in this regard. If you think about it, the fact that you have this very small point contact sitting on a wafer of silicon, and you realize the voltage you can develop by wearing a tweed suit and wiggling around in a chair, you realize that a spark discharge could easily destroy the detector. We had to find that out by just agonizing experimentation. And the problem was that we had a set of 100 crystal detectors that had been measured as Holmdel's measurement standard, because they wanted to replace these in a frequency converter without changing the tuning, so that people in the field could readily pull one out and put another one in. It meant having the test equipment organized around the characteristics of these particular detectors. We had no better way to specify them so we had 100 samples that we kept as a kind of a standard sample. We were setting up the measurements program at Murray Hill, to support our own development effort, rather than having to ship all the crystals to Holmdel This was again in the charge of a Western Electric Co. engineer. But no sooner than this program would start, we'd bring up 100 samples from Holmdel , and this fellow Cliff Corbett would run through the whole 100, and 25 of them would be completely knocked out of adjustment. They'd go back to Holmdel, and repeat this thing time and again. They could be measured for a week, every hour on the hour, and no changes were observed. We finally discovered, just inadvertently, that Merlin Sharpless, who was doing these measurements at Holmdel, systematically put his fingers on a grounded plate, and thereby destroyed any static that he had accumulated. Whereas if he had not done that, holding these detectors by the base and plugging them into a grounded piece of equipment, it would discharge thousands of volts through this small point contact, and literally blow it up. We had to establish a routine, where the test operator was required to keep one hand on a grounded electrical plate. Well, Mervin Kelly heard this tale and he came storming into Earle Schumacher's office one Saturday afternoon.
Hoddeson:1940, was it, or ‘41?
Scaff:About then. He said, fir have just heard that these silicon detectors can be destroyed by static discharges such as just wiggling around in your chair and touching them. Earle said, “That’s right.” He called me on the carpet. “What has been done,” he said, “to avoid a real catastrophe in the field?” Fortunately, we had started a little warning program, each box shipped to the field had a big sign warning that they could be damaged by misuse. We’d put these detectors for shipment in a lead container, lead shielding, and then had an instruction booklet which described specifically how to avoid this difficulty. I think if that step had not been taken, on that particular Saturday Mervin Kelly would have had my head. But we had fortunately anticipated it.
Hoddeson:Did you work closely with Ohl during the War too?
Scaff:Ohl was the inventive type, and the person that Harold Friis assigned to this program during the war was a fellow by the name of Mervyn Sharpless, who was in charge of electrical measurements.
Hoddeson:Was the work that you were doing before the War completely dropped, or were you still doing some of it?
Scaff:Oh, we were doing a little bit of it. You couldn't help but think about it once in a while.
Scaff:In fact, it had not been completely dropped. We had correct ideas as to which impurities were responsible for p type and n type conductivity. And the decisive experiment that we had developed, the technology for adjusting these specific resistances of the silicon, by the addition of boron and in fact, this table, Table 1, shows the range of electrical resistivities that we utilized. Here -- both .03 to .04 centimeters -- .05 to .09 for that one -- .01 for the 126 detector -- hm? These were specification numbers. The point is, to make this detector work right, that S band, 3000 megacycles, X band, 10,000 megacycles -- They had to have these, didn’t they?
Scaff:They had to have these, and to get the uniformity of properties we desired, we needed to hold the specific resistance of the silicon within these limits. So what we did was to buy a new batch of silicon, standardize it by making a melt, measure its resistance, then adding an alloy, 1 percent boron, so many milligrams, we adjusted the resistivity to come within these limits -- these were the impurities, we chose boron. Now the point is, boron's a group three.
Hoddeson:This was mostly for frequency conversion?
Scaff:That's right, for essentially the first detectors in radar, operating at 3000 to, this on the K band was 21,000 megacycles. Never got into commercial production. S band was 3000, X band 10,000. The general technique was to take the radar transmitter frequency, 3000 megacycles, which would be reflected from an object, an airplane or battleship or submarine or something, it came back to the receiving antenna, and there it was mixed with an intermediate B frequency, to down-convert to a lower frequency, which could then be amplified and finally processed in other ways. So when we speak of a frequency converter, it's really a down converter, from the transmitter frequency down to something that can be amplified and handled by regular electronic techniques.
Hoddeson:I guess behind my question was really the thought whether or not the idea of a solid state amplifier was already in the back of some people's minds.
Scaff:It was. But in quite a different connection. It was in Bill Shockley's mind the day he came to Bell Laboratories, because he realized the analogy between the copper oxide varistor, and the vacuum tube diode, so he reasoned that if he could take this solid state device and put a third electrode in it, same as putting a grid.in a vacuum tube, then we could build a solid state amplifier. And he tried in fact experiments in which a nichrome grid was fired into a copper oxide varistor -- the cuprous oxide layer on Chile copper was produced by thermal oxidation -- and he tried this, that type of experiment, to get a grid right at the seat of the so-called junction between the cuprous oxide and the copper, to control the flow of electrons by grid action. But what he didn't reckon with there was an effect called the Kirkendall effect, and the grid didn't end up in the right place. The oxide migrated right by it.
Hoddeson:But the connection with silicon really wasn't seen until after the war?
Scaff:Oh, Bill went to some duty in Washington, some kind of operations research work -- But then after the war he returned and Kelly, seeing how much we'd been able to do empirically during the war with these semi-conductors, and particularly with the newer ones, germanium, he set up the group headed by Shockley and Stan Morgan, to do research in solid state physics. And this is when Gerald Pearson joined the team, and Sandy Goucher, and many others.
Hoddeson:The new solid state group that was formed in '45 was really Mervin Kelly's baby, was it?
Scaff:Yes. Yes, it was. But based on part on a postwar review of what had gone on during the war. Well, in fact the whole top Bell Labs management was involved in that one.
Hoddeson:Where could I find out about some of that?
Scaff:Well, Stan Morgan is still alive.
Hoddeson:Are there any documents I might find, relating to the decision to establish the solid state group?
Scaff:No. I wouldn't know that. I do know that our part was purely to give the material expertise, the samples they needed to work with. But we were encouraged to get out of the field entirely, because after all, we were empirical, experimentalist, you know. And so we did get out of it, Henry Theuerer started working on thin films, something he took up later, again, very successfully, and Bill Pfann started to work on electrical contact materials. And we were just essentially tapering off on our Whole semiconductor program when the transistor was discovered.
Hoddeson:They were using your materials.
Hoddeson:And were you in contact with them frequently?
Scaff:Oh yes, absolutely. Sure. Almost daily.
Hoddeson:Who, for example, did you talk to?
Hoddeson:What kind of question would Shockley ask you?
Scaff:Oh, he'd want a sample of germanium.
Hoddeson:This is of course before the single crystals were developed.
Hoddeson:That didn't come till three or four years later.
Scaff:Yes. There's some funny terminology here, though. These early germanium ingots -- I don't know if I've got a picture of one of them here or not. I think I do. Yeah, here --
Hoddeson:This is Fig. 13
Scaff:They are heavily twinned, so that the area that these point contacts received, really was one single crystal. But these were twin boundaries providing recombination centers, reducing the minority carrier lifetimes as we now know. I'm not depreciating the single crystal development at all, but I am really pointing out that this was not a polycrystalline piece, in the sense of a polycrystalline sample of steel.
Hoddeson:I hadn't understood that. I want to ask a question about the interactions in the wartime period between Bell and MIT and the British. Did these interactions continue after the war, or did they taper off also?
Scaff:Let me add one other laboratory there. Frederick Seitz headed up a research program at the University of Pennsylvania, much smaller than the one at MIT, but he was specifically concerned with silicon as a semiconductor, and research in solid state physics. Lark-Horovitz was doing the same thing at Purdue on germanium. At MIT they were more device-oriented, looking towards practical applications. But the research program on silicon as a semiconductor was done largely at the University of Pennsylvania under Seitz.
Hoddeson:This was during the war?
Scaff:Yes, before he went to Carnegie. Another interesting bit of Seitz's visits to Bell Laboratories during the war: the silicon detectors were adjusted by beating them on the side with a little selected mallet, a process known as “tapping.” This is described briefly in that other article. But to this day, I don't think anyone has the faintest idea of why this process worked. But it did work. From informal discussions, Seitz had decided that we were bouncing the point around mechanically, until it found an active spot. But I told him, “This just isn’t so. This is a perfectly systematic and not a random process. You hit it once, twice three times, and the forward current increases progressively, and the reverse current correspondingly decreases. Every one of them behaves similarly, and if this were a random popping around process looking for an active spot, it wouldn't work this way." And we had to get him up to Murray Hill one day to demonstrate this to him. And he became a convert. Although no one ever explained the mechanism. We still don't understand it nor even have a good hypothesis for the mechanism.
Hoddeson:Did Seitz have a large group at Penn?
Scaff:No, it wasn't a large group. I'd say, oh, just recollection now -- about eight people. But it was a solid state research group documented in NDRC reports. And I think you ought to dig out some of the reports, and see in retrospect how little solid state physicists knew about semi-conductors, because at that time it was in a very early and preliminary state.
Hoddeson:What reports are you referring to now?
Scaff:These are reports issued by the National Defense Research Committee. I had some of those, too, and I don't know whether I kept them -- I’m sure I didn't keep them. But I kept them for many years, because I was so anxious to show them to our erudite solid state physics people, who sometimes gave the impression that we arrived at the present sophisticated state from pure theory. Instead there was a great interaction between experiment and theory.
Hoddeson:Where do you think I might be able to get information? Do you think Seitz might be able to --?
Scaff:You can ask. The whole thing's complicated, because all of this was classified, during the War.
Scaff:And when the War stopped, they began to terminate these war projects, abruptly. They set up a project at MIT, to write the history of MIT's and related programs and this resulted in the book on crystal rectifiers by Torrey and Whitmer. It's got a lot of inaccuracies in it but it still is very valuable contribution. Again, it was written rather broadly, trying to make everyone happy, you know. But the result is that many of those reports were never properly declassified. And I kept many of them for years, in my safe over at Bell, because I knew the information itself was declassified, but the document per se wasn't, and it did drive our security people crazy, when they came across one of these secret documents in my locked files.
Hoddeson:You have no idea what happened to your collection of papers.?
Scaff:I really don't know where they might be. You might ask Mort Fagen. Do you know him?
Scaff:I've turned over a lot of that stuff to him. Certainly, I wouldn't have destroyed these reports. But they were kind of funny items, because they still had the classification on them, and storage under present circumstances was just a real trouble. The same is true of our memoranda, too, all written during the war. They were all classified.
Hoddeson:Coming now to the postwar period. Did some of these interactions get a little bit less intense when things got back to normal? Interactions say between MIT and Purdue and --
Scaff:Well, yes, because the War-time functions of these organizations ceased. They were all closed up, and the university professors went back to their activities. We hired a number of people from MIT. Mort Fagen's one of them, A. M. Clogston another and there are many others. Fagen was a liaison man on this germanium rectifier program, monitoring Purdue’s, MIT's and our own program.
Hoddeson:I didn't know that.
Scaff:He came down from MIT, the NDRC and sat in on all of our conferences with Purdue University. That's how I first knew him. Outside of writing the history, like Torry and Witmer's book, these people all just disbanded. Purcell went to Harvard. Al Clogston, who worked on magnetrons I think at MIT, came to Bell Labs. Shockley came -- well, Shockley was not at the Radiation Lab, he was in Washington, but they all dispersed.
Hoddeson:The question I'm asking is leading into a more general issue, the effect of what you learned during the War, technologically, on the research that followed.
Scaff:Well, the very fact that we had the technology for preparing fairly pure silicon, and certainly for preparing very pure germanium, of controlled properties, was rather vital to -- Have you seen Walter Brattain's story in Physics Today on the history of the transistor?
Hoddeson:Let's see, I have two by Brattain here. This one is called "The Genesis of the Transistor," and then there's one that's a bit longer and a little more detailed.
Scaff:Yes. This is the one that I'm thinking of.
Hoddeson:This one in The Physics Teacher.
Scaff:This is a very good picture of the situation, here. As I know it, anyway.
Hoddeson:If the War hadn't occurred, how would the developments have changed? You were still working on trying to get the materials pure and with controlled impurities.
Scaff:Yes. I don't think there's any question, though, but that the wartime emphasis on developing first detectors forced us to acquire a great deal of technology and know-how for producing these materials. I don't think there's any doubt but that it had a major impact on postwar work. I doubt that we would have brought solid state physics and that group into such sharp focus if Kelly hadn't followed this development so closely, all during the War, and realized its potential. And don't forget, he's the man that hired Bill Shockley in the first place. So he knew his talents and capabilities. I remember meeting Bill when he first came to the Lab. I do not know what Kelly's thinking was, that got him to hire him. But certainly he knew what he was doing because he was a vacuum tube man, and I think he himself saw the analogy between solid state copper oxide varistors, and vacuum tube diodes. Probably felt it was a field for exploitation, and he was certainly right.
Hoddeson:Did you interact much with Shockley, after the War?
Hoddeson:How about Dean Wooldrige? We haven't talked about him at all.
Scaff:I had no contact with Dean during the War. I knew by reputation what his work was.
Scaff:I've known Gordon since he first came to Bell Labs, and of course he was always interested in semi-conductors and things of that kind, and did begin to do some work after the war, on different ways of making semi-conductors. He was another man who was very patent-conscious.
Hoddeson:Did you ever work with him?
Scaff:In what way?
Hoddeson:I mean, on a common project?
Hoddeson:But you would talk --
Scaff:Oh yes, we would exchange ideas. We were familiar with each other's programs.
Hoddeson:Was the work on the germanium diode a direct outgrowth of the radar war work?
Scaff:The group at Purdue had started looking at germanium as a possible competitor to silicon for microwave detectors. And in fact, to some degree this was realized, because when we began to work in these very high frequencies, a chap by the name of Harper North at GE took up technology that Purdue had started, I’m making a germanium microwave diode. In that program, which had started with the microwave use in mind they had discovered the so-called high-back voltage rectifier, a very unusual point contact device which would withstand something in excess of 50 volts in a reverse direction, whereas the silicon detector would withstand maybe three or four volts without being destroyed.
Hoddeson:About when was this?
Scaff:It was during the War. And MIT, in looking at possible useful applications of this knowledge, hit on the idea of making the germanium second detector. And since we had been quite successful at Bell in developing the silicon microwave detector for manufacture, they asked us, independently of Purdue to have a separate NDRC contract to develop germanium second detectors. And this was indeed done, with NDRC money. And it set up an interaction between ourselves, the Purdue group, and MIT again.
Hoddeson:After the War?
Scaff:During the war. I don't -- I may have that other article, "The Germanium Detector" Let me go upstairs and get a document for you.
Hoddeson:OK., We can wind up for now. In so doing, I'd like to ask you a few general questions. Do you recall when you heard about the invention of the transistor?
Scaff:Well, certainly. That's recorded. There's that note. That note was my --
Hoddeson:-- (that note is Figure 17-in your 1970 article.)
Scaff:I was hosting a conference on ferrites. And Bill opened my office door, and pushed that note in.
Scaff:Yes. And of course, you recognize these names Bown, Ralph Bown was director of research. Kelly himself. And Quarles, he was then vice president of development, the job that Morris Tannenbaum has now. He later became Secretary of Defense, Donald Quarles. Needless to say, I did break away, and I did go to the meeting. The funny part is, for some reason or other, the true date of this discovery was never known exactly on the outside. This note had a date on it. I dated it. And the original one that's turned over to Mort Fagen has the date on it, but since I was giving this invited paper, and I wasn't too sure of how sensitive that date might still be, I took it out of the publication. It was December 19 --
Hoddeson:'47. By the way, did you know Bardeen?
Scaff:Yes. I knew him after he came to Bell Labs. I did not know him before then. I knew him quite well, at Bell.
Hoddeson:During this transistor period?
Scaff:Oh yes. Sure. I remember once the transistor program was organized, the collaboration between us had intensified. And Bill Shockley had a planning conference about every week, to review what had gone on, and we were always part of that meeting.
Hoddeson:You, and --?
Scaff:Bill Pfann, Henry Theuerer.
Hoddeson:Were these small meetings?
Hoddeson:Who? Bardeen was there and Shockley was there.
Scaff:Frequently we had --
Scaff:Teal, Joe Burton -- all those people that were involved in it and were around were there at those meetings. Frequently we had Peter Debye Sr., from Cornell as consultant. I'll tell you a true anecdote. Bill Shockley was well known for deriving equations on the board, relating to hole transport or some esoteric subject, equations running all over the board, you know, and Walter Brattain was equally interested in walking back and forth in the back of the room and watching what was going on. I guess he felt better when he was walking around, moving around, than he did just sitting on a chair. And on this particular day, Bill Shockley turned around and snapped at him and said, "Walter, I wish you'd quit jiggling those coins in your pocket. I can't think when you make money jingle." And Brattain answered him and said, "Look, I can't think when I don’t have money jingling." Bill said, "OK -- will you please only jingle bills, after this?" (Laughter)
Hoddeson:Well, things changed then, abruptly, is that right, or not? After the invention?
Scaff:Oh, they changed abruptly. The whole program grew very very rapidly. Ralph Bown had set up a time scale for us to get as much patent coverage as we could within six months, because he was confident that the information would need to be widely disseminated in order to have the transistor fully exploited, and he wanted to establish our own proprietary position before this information began to get out. And those small meetings of Shockley's grew as that program grew to an auditorium sized meeting. And this all in a matter of months. Incidentally, Jack Morton and Joe Becker were given the official responsibility for making a "housed" transistor, with the idea of furnishing samples to the various circuit groups around Bell. We realized that, from the War effort, we still had piece parts and tools for the so called IN26 detector program, we being Bill Pfann and myself. The program was almost wholly Pfann's because this was a major project for him during the war. This was a principal assignment of his. And we realized that we could take this device with a single point contact and replace it with two. And by assembling this like so, with two point contacts here, you had the physical housing for a transistor. And this was actually the prototype of the "Type A" transistor. Now, Bill Pfann and Dan Dorsi, his technical aide, pending the Becker program, -- after all it took some time to learn the technology, to get piece parts, to get going, in other words. In the meantime, all of this big demand came up for internal transistors. They were all produced by Bill Pfann, using this technology. And most of the transistors that went into the first press show, where they demonstrated the vacuum tubeless radio, amplifiers and all kinds of gadgets, practically all of those were made by Bill.
Hoddeson:This was quite some time before the work on zone refining, wasn't it?
Scaff:Yes. Not quite some time, but before the zone refining. A year or two maybe.
Hoddeson:I think this is probably a good time to stop.
Scaff:Good. Would you like some coca cola, a drink, or some coffee?
Yes, thank you. And particularly, I want to thank you so much for being so very helpful in this interview. I’ve learned a great deal today.
The Role of Metallurgy in the Technology of Electronic Materials" by J. H. Scaff, Metallurgical Trans., Vol. 1 March 1970, 561-73
"Development of Silicon Crystal Rectifiers for Microwave Radar Receivers, Tf J. H. Scaff and R. S. Ohl, Bell Syst. Tech XXVI, Jan '47, No.1 p. 1
P. 565 Metallurgy Transactions Vol; March 1970
Ibid. p. 565
M. J. Kelly was director of research 1936-44, executive vice president 1944-51, and president 1951-58.
P. 568 Metallurgy Transactions Vol; March 1970
Development of Silicon Crystal Recti:fiers for Microwave Radar Receivers. J. H. Scaff and R. S. Ohl. Radar Systems and Components - Bell Labs. Slatt and van Nostrand M4-1949
The Physics Teacher, March 1968 Pp 109-114
G. L. Pearson and W. H. Brattain, Proc. Inst. Radio Engrs., 1974 (1955)
Ibid. p. 571