My First Impressions of William Shockley

Taken from the R. VICTOR JONES, interview by DAVID C. BROCK via telephone, on 18 April 2006 Philadelphia: Chemical Heritage Foundation, Oral History Transcript # 0336 (Pages 7-12).

BROCK: Okay.  When and how did William Shockley [William Bradford Shockley] emerge in the mix?

JONES: That was an odd experience. I had already accepted a job at Bell Labs, but was finishing up my thesis research.  My research room at Berkeley was in LeConte Hall (the physics building), just off the front hall. One day this little man walked into my lab and, without introducing himself, started asking me searching questions about my experiment.  I was used to getting lots of madmen visiting my lab because it was right across from the front entrance to LeConte Hall (just opposite the Campanile) and often crazy people would just walk in and start asking questions about all kinds of crazy science fiction stuff.

BROCK: Okay. [laughter]

JONES: So I was guarded when this little guy walked into my lab and started asking questions.  But, the questions were reasonably perceptive and I started to pay more attention to him.  Then, he confessed that he had sought me out following a conversation with Charlie Kittel.  Do you know that name, Charlie Kittel?

BROCK: I do. Yes.

JONES: Charlie Kittel was an outstanding solid state theorist at Berkeley and former associate of Shockley at Bell Labs. Shockley told Kittel about his attempts to recruit people and Kittel said, ―Why don't you go and have a try at this guy Jones?  Thus, Shockley’s visit to my lab. I was in my working togs in the lab, so Shockley said, ―Why don't you go get cleaned up and we'll go out for lunch, and, flash forward in Shockley’s Jaguar, we were soon at the Fairmont Hotel in San Francisco for lunch. [laughter] For a starving graduate student the luxury was impressive and a bit of a shock, but there followed what it was probably the most intense afternoon of physics I have ever enjoyed.  Shockley and I spent, I would say, the next six or seven hours in very intense discussions and, as it happened, he asked me about subjects about which I knew quite a lot.  I guess I showed up pretty well.  He thought he was asking very hard questions, but they were in areas of comfort for me.  One of the greatest and most interesting subjects, to me, has always been electromagnetic theory.  I had a couple of very good courses in the subject at Berkeley that had strongly triggered my interest.  Shockley had just come from a visit with Feynman [Richard Feynman] at Caltech [California Institute of Technology], and he (Shockley) had learned of a number of electromagnetic paradoxes that he was interested in quizzing me about.  It turned out that I was—not because of my brilliance but just because I happened to know the subject pretty well—able to cope with the paradoxes that he posed. [laughter] I don't mean this to sound immodest since my supposed brilliance was essentially accidental.  However, Shockley was impressed by me and he made a real pitch to get me to work for him.  I said, ―No, I've already taken a job at Bell Labs and it's a good one.  Of course, that was even a greater incentive for him because he wanted to win a victory over Bell Labs. [laughter]

This debate went on for several days.  I don't know if you've picked up on this or not, but one of the things that Shockley was enamored with was having psychological profiles of his people.  He said, ―Well, you don't have to come with me, but I'd like you to submit to a psychological profile because I'd like to calibrate the guys that are doing it.  As a personal favor would you do this?  I did it and that was an interesting part of the whole recruitment.  At that point, I really wasn't inclined to go with him.  I had already established my interest with Bell Labs.  But, I liked the West and would have preferred to stay in California.  I had family in California, and my mother was getting along into her eighties, (although she did eventually live to a hundred). I also had a semiserious girlfriend who lived in California.  The idea of going East was not attractive for me.  However, there weren't many opportunities in California at that point.  Certainly nothing like Bell Labs existed in California.  So, by hook or crook, I ended up going with Shockley.  As result, I became a persona non grata at Bell Labs for the rest of my career. [laughter] Whenever I met somebody from Bell Labs in subsequent years they would say, ―Oh, you're that guy.  In those days, you just didn't turn down a Bell Labs job.

BROCK: It must have captured their attention, certainly.

JONES: That's how I got to the Shockley Semiconductor Laboratory.

BROCK: Were you aware of Shockley's reputation?

JONES: Not to the degree that I became aware of it later. [laughter] In retrospect it was a very na´ve decision process, but I was young.

BROCK: What about his technical or scientific reputation?

JONES: I knew it was high.  I was fairly aware of what was going on in contemporary solid-state physics since I had audited Kittel’s course and one of my closest friends at Berkeley, Gene Dresselhaus, was one of Kittel's  most outstanding students.  Kittel brought solid-state physics to Berkeley.  I was aware of a lot of the work in his group, and I knew Arthur F. Kip who was an experimentalist working with Kittel.  I was aware of the personas in solid-state physics, and one of the reasons I did cyclotron resonance in gases was the fact that cyclotron resonance in semiconductor materials was a very hot subject.  Gene Dresselhaus did an extensive theoretical analysis of cyclotron resonance in semiconductors.  So I was inclined to have a crack at it in gases. After a lot of effort, it actually worked. My thesis had a couple of pieces but that was the dominant piece.  I was well aware of who Shockley was, but I didn't know about all the crazy stories.  I'm not sure that if I knew them it would have changed my mind. [laughter]

BROCK: At the time he recruited you, how did Shockley describe the nature of his new operation and his intent?

JONES: I think his view was that interesting devices were invented at Bell Labs using all the wondrous opportunities that it offered for doing pure research, or at least directed research of a certain kind.  Yet, Bell Labs wasn't able to capitalize on these by moving forward solid-state technology.  His feeling was that the opportunity was there to do things that weren't being done, and that the technology should move forward.  In particular he and Morrie Tanenbaum [Morris Tanenbaum] invented and developed the so-called mesa transistor.  The mesa transistor, to me, may be characterized as a major metaphorical development in modern technology—I can expand on this contention if it would be useful. Shockley put forward, in our discussions, the fact that there were new opportunities in semiconductor device development which were not being exploited.  Based on my conversations with Shockley, I came to believe his contention.   I believe to some extent he was triggered by some of the Hewlett Packard concepts of getting together a group of people who were research-level scientists but who would be motivated to get something out the door.  As you probably know Hewlett Packard worked for many years on the philosophy of people taking a device from the basics to the market.  I knew Shockley was very familiar with Hewlett Packard's operation, whether he picked that up from them or whether it was something that came from him, I don't really know.  Shockley’s idea was that he would recruit people who could make their living as research scientists but had the willingness to bring stuff to utility.  To me, that was an attractive thought.  I've always viewed myself as an applied physicist in some sense.  I'm not sure I could have articulated it at that point, but as life has gone along that is what I enjoy.

BROCK: Let’s talk about what you mean when referring to the mesa transistor, the diffused-junction transistor, as a metaphorical development?

JONES: I have not done the historic research to know whether Shockley is the unique figure in this regard, but I suspect he is.  What Shockley articulated at that time in the development of transistors was the need to make things extremely small.  The specific big issue was that in the physics of junction transistors, the thickness of the base region is the crucial parameter.  The base region has to be kept thin, for a whole variety of reasons.  If it's too thick then the electrons injected from the emitter into the base won't get to the collector junction and transistor gain will be impaired.  The base region has to be physically thin.  The technology available at that point was developed around various mechanical or semi-mechanical means for slimming down the devices into very small thin structures.  Let me note parenthetically that Bob Noyce [Robert N. Noyce], before he came to join the Shockley Lab had been working at Philco [Philco Corporation] on a methodology for making very thin base regions on germanium transistors, which was a mechanical marvel in some sense.  It was obvious that they we weren't going to be able to keep making things smaller and smaller by basically mechanical means. Shockley said, ―We must use controlled natural processes to arrive at small transistor structures.   What became the technology for the diffused base transistor was to start with a plain semiconductor surface and then to use processes of sequential diffusion controlled by time, pressure and temperature to differentially build structures in the vertical direction from the plane down.  I think that was the advent of one of the key metaphors of our time because if you look at any technology now—acoustics, biotechnology, optoelectronics—the idea of building structures with planar-oriented processes – diffusion, ion implantation, lithography, selective etching—is a dominant metaphor.  I have not come across any examples of the use of this metaphor before the advent of the mesa transistor.  I could be wrong, but when I posed my view to Jay Last [Jay T. Last] at our recent luncheon, he thought it was an interesting thought, but he wasn't quite sure he believed it. [laughter] But, it’s been a view that I've have bandied about for some time.  Photolithography and a whole arsenal of planar-oriented techniques now permeate every facet of technology.  In fact one of my sons has worked on interesting biological applications of planar structures of this sort.  In my view, this powerful metaphor first emerged in the context of the diffused-based, mesa transistor.

BROCK: Right.  And certainly it sounds as if Shockley was putting that foremost, to getting—

JONES: Oh yes.  The diffused-base technology was one of the fundamental design ideas that he put forth.  He felt that—I don't know what the history at BTL was—but he felt it was insufficiently appreciated.  I had conversations with Morrie Tanenbaum in later years about this, and he recognized that it was a new metaphor that had been developed with Shockley.  He didn't deny that it probably wasn't made useful by Bell Labs in the way that Shockley had hoped.  But I really don't have any other insight into the history.

BROCK: Did Shockley discuss a strategy for his new operation with respect to crystal material at all?

JONES: Not at that point, but it was quite clear when we got in the business that it was a serious issue.  I think the diffusion technology, as a basis of device development, was the key matter as well as the idea of taking research-level people to do the development.  I think the success of Intel [Intel Corporation] demonstrates the validity of the assumption.  The success of Fairchild [Fairchild Semiconductor Corporation] and Intel certainly bore that point out.  Most of the people that were active in those early developments would certainly be of the caliber who could operate successfully in a research laboratory.

BROCK: Right.  When was your first day at Shockley Semiconductor?

JONES: [laughter] I'm not sure of the first day. It was early 1956 or late 1955.  I hadn't submitted my thesis yet but I started working there.  My thesis was done but I was putting together a few final things.  I think I appeared sometime in early 1956
or late 1955.  I don't really have a precise recollection of the date, but I remember having a very depressing impression.  With all of Shockley’s high-minded rhetoric, putting together the laboratory was something less than an exciting process—Jay Last once characterized the early days in an old garage as life in the putty knife factory. [laughter] In fact, when I started with Shockley there was nothing.  We were sitting around in a store front somewhere in Palo Alto.  We were a strange group of people.  Most of the later outstanding people had not yet been recruited.  There was a fellow by the name of Bill Happ [William Happ], there was another fellow by the name of Leo Valdes, and there was someone by the name of Smoot Horsley. Those were the people. I was taken aback.  This was not Berkeley or Bell Labs, but of course Shockley was there.  Shockley put me to work on a device which had great complications, and that was the four-layer diode.  He immediately said, ―You're a good theorist, so why don't you work on understanding the four-layer diode because we're going to make them.  I have got to say that the four-layer diode is still, to me, one of the most complicated devices in semiconductor technology. [laughter] It is a remarkable device and it obviously resembles some gaseous discharge devices that I knew about, but the theory was wild and very nonlinear.  For other reasons I spent time looking at that theory in subsequent years and I still believe it was a overwhelming task to put on a guy who knew nothing about semiconductor physics, but that's what Shockley did  .I spent most of my time reading and trying to learn about four-layer diodes.  The reasons for Shockley's interest in the four-layer diode, which triggered some of the later problems with the staff, was that he saw that it could be used for telephone crossbar switching.  Indeed, it could have been a very interesting possibility for that application.  I think Shockley saw the four-layer diode as a unique but fairly simple use of his diffusion-based planar technology.  Namely, what had to be done was to diffuse in four layers, and then contact could be easily made at the top and the bottom, of the four-layer stack.   I think what he saw—I'm somewhat projecting here—was the fact that one could put a whole bunch of these on one chip and then have a crossbar.  Whether he really saw it that way or not, I don't know, but it would not have been a stretch to imagine that.  I think even though it seemed crazy at that time the idea of building a chip of crossbar switches was not too big a stretch.

BROCK: I've heard varying things about the four-layer diode.  One of which was that the real issue with it, from a manufacturing point of view, was that it was hard to get uniformity from diode to diode because the diffusions were so delicate?

JONES: That's correct.  But to build a CMOS [complementary metal oxide semiconductor] device, I think, had similar challenges and the challenges have been met.  I think the three-layer junction transistor device was simpler, but on the other hand contact had to be made with the intermediate layer.  That made it harder. [Laugher] At this distance, I don't think one could make a good parse of which device is the most desirable. I think that probably what was the more pressing point was that the four-layer diode was a fantastic device for telephone switching, but if one was trying to sell a general purpose device which would go into many different products, I don't think it was a very good choice.  After all, the bipolar transistor is a little bit like a general computer. It's a device that can be used for many different purposes, whereas this four-layer diode has only a very unique function.

BROCK: I see.

JONES: I once played with the idea of semiconductor optoelectronic devices which mimic the functions of the retina and the visual system.(
Photoneural Systems: An Introduction, Applied Optics 26(10) (15 May 1987): pp 1948-58)  In such an application the four-layer diode could well play the crucial role of a synaptic-like element—it has the response of an axion.  It is a very interesting device and I don't think it should be pooh-poohed out of existence, but it certainly was not a good way to build a market, which I think was the real criticism of the people that left Shockley.

BROCK: Okay. Just to make sure that I understand you exactly, this was still in the days when Shockley Semiconductor Laboratory was a handful of people in a storefront when Shockley gave you the tall order of getting your head around the four-layer diode theory?

JONES: That's correct.  I think it was basically a smart thing.  He knew that I was bridled by the less than inspirational people I was dealing with and I think he was probably afraid that I would take off because I did not have any particularly strong bonds there.  He was trying to give me something challenging.  I also spent a lot of time on things like helping to find lab space we eventually ended up with an old garage that became the putty knife factory, in Jay Last’s term—and recruiting people.  I traveled a lot with Shockley.  He was trying to recruit people in various places, and again, he had recruited me away from Bell Labs so he used that as an argument.