AT THE AT&T Bell Labs in Murray Hill, New Jersey -- the place where in, December of 1947, the transistor was invented -- there is an entity which is known locally as "The Blue Zoo." It is a sizeable laboratory with specially filtered air in which the workers move around like surgeons in azure gowns and surgical gloves. The laboratory is devoted to research on microchips -- those silicon wafers on which hundreds of thousands of electronic elements are etched and which are the heart and soul of the modern electronics industry. Just outside the laboratory there is a visitor's gallery and on a table a high powered microscope has been set up. When one stares into it, at first, one does not know what to make of what one is seeing. Focused in one way, a series of lines appear that look like a Martian city. They are the "wires" that interconnect the components. In the background there is a fuzzy object that, compared to the lines, looks like a gigantic reptile. Only when one refocuses the miscroscope onto the "reptile" does one realize that it is a single strand of human hair.
The whole thing makes such an eerie impression that one's curiosity is aroused as to where this technology came from. Who got the idea of putting all of these components together on a single chip and how does it work? T.R. Reid's book, The Chip, is an attempt to answer both of these questions. I very much enjoyed Mr. Reid's answer to the first question -- who did it? -- but I must say that I found his answers to the second -- how does it work? -- pretty unsatisfactory.
First the good news. The idea of making an "integrated circuit" -- all the components of a circuit on one chip -- first came to the Kansas-born inventor Jack Kilby in July of 1958 and then, independently, to Robert Noyce, six months later. I was fascinated by Reid's description of these two men -- totally different personalities, and both admirable in their own ways. Kilby worked for Texas Instruments and Noyce was one of the founders of the Fairchild Semiconductor Corporation, which he later left to create Intel, the company that is arguably the most innovative semiconductor company in the world. Eventually Noyce and Kilby became locked in a patent suit which had to do with the wording of Kilby's patent. The issue hinged on how the various components of the circuit were to be connected. (Incidentally, Reid does not mention that the practical construction of these microchips depends on a discovery made in 1954 at Bell Labs that a coating of silicon dioxide -- a kind of glass -- could be made to adhere to the surface of the silicon. Holes are made in this surface into which the impurities that make up the circuit elements can diffuse). The suit was finally decided in favor of Noyce in 1969, but by that time it was largely irrelevant since Texas Instruments and Fairchild had reached an agreement among themselves to cut the pie -- a $100 million pie in royalties over the years. All of this was new to me, as was the lively impressions one gets of the two principals Reid interviewed. So far, so good.
THE PROBLEM with the book is the science. These semiconductor devices are not easy to understand. Very distinguished solid state physicists have confessed to me that from, one year to next, if they are not actively working on them, they forget how they work. A very reasonable question on a physics PhD qualifying oral is to ask the candidate how a transistor works. Its workings require a clear mastery of the quantum theory of solids -- nothing less. With this is mind it is understandable, albeit, unfortunate, that Reid's science is rather a shambles. Here are a couple of examples: "The mass of each particle electron would be less than one-thousandth the size of a hydrogen atom." Comparing a "mass" with a "size" is not like comparing "thee to a summer's day." It is more like comparing a whale to a radio. The mass of an electron is about one two-thousandth of the mass of a hydrogen atom.
On a less pedantic level; "Under these rules (the Bohr quantum theory of the atom) the orbit that is farthest from the nucleus can have from one to eight electrons." This statement is reasonably clear, albeit wrong. Palladium, for example, has 10 electrons in its outer shell. It is not really correct to say that these electrons are in the same "orbit," but this is not the place to try to explain the quantum mechanics of the atom.
I wish Reid had run the technical sections of his book by a few critical physicists. I think if he had done so he would have had a better book. Although I am a professional physicist, when I had occasion to write about these solid state devices which are not in my field of competence -- I had some nine of my colleagues -- including two Nobel Prize winners -- vet the material, and I am still not sure if we got all the bugs out.
None of these remarks are meant to take away from what I think is the real value of the book. The very good idea that Reid had, was to focus on Kilby and Noyce and what their remarkable idea has meant for us. Even those of us who are old enough to remember the slide rule and the wind-up watch have now come to take for granted that an instrument costing not very many dollars can simultanously tell time, do arithmetic, sound an alarm, give the date, take our temperature and God knows what. We have come to feel that having these devices is part of our birthright. When an idea or technological innovation takes on dimensions like this it is very easy to lose sight of where it came from and who was responsible. No reader of this book will forge Kilby and Noyce. They are the Edison and Ford of the present.