Encyclopedia of 20th Century Technology

(Note: Sample material is taken from uncorrected proofs. Changes may be made prior to publication.)

Transistors

In 1906, the American inventor Lee De Forest developed a triode, a three element vacuum tube (or "thermionic" valve). Dubbed the De Forest Audion, it was a device that could detect and electronically amplify radio and telephone signals. In 1909 the American company Bell AT&T bought De Forest's patent and improved the tube so that it could be used to amplify signals in long-distance telephony. A practical problem was that the vacuum tubes were often unreliable, slow, used too much power, and produced too much heat. For years, researchers in Western countries tried to make a solid-state amplifier-what became the transistor-in an attempt to enable the creation of smaller, faster, less power-hungry electronics.

Early experiments in transistor technology were based on the analogy between the semiconductor and the vacuum tube: the ability to both amplify and effectively switch an electrical signal on or off (rectification). By 1940 Russell Ohl at Bell Telephone Laboratories, among others, had found that impure silicon had both positive (p-type material with holes) and negative (n-type) regions. When a junction is created between n-type material and p-type material, electrons on the n-type side are attracted across the junction to fill holes in the other layer. In this way, the n-type semiconductor becomes positively charged and the p-type becomes negatively charged. Holes move in the opposite direction thus reinforcing the voltage that is built up at the junction (Figure 1). The key point is that current flows from one side to the other when a positive voltage is applied to the layers ('forward biased').

The transistor is also a solid-state device that could amplify electrical current. Transistor stands for transit resistor. Its development went along two lines: basic research in solid-state physics to replace the old vacuum tubes (such as De Forest's audion) that failed to solve technological problems, and multi-disciplinary research activities at several industrial and university research labs. It was in December 1947 that John Bardeen and Walter Brattain working at Bell Telephone Laboratories in New Jersey, USA in a research team headed by William Shockley, demonstrated the first transistor, a semiconductor device based on germanium. However, the German scientist Julius E. Lilienfeld from New York had patented the first field-effect transistor in 1926. It was a patent on a 'Method and apparatus for controlling electric currents'. It was unlikely, however, that he ever got it to work. Nevertheless, in the early 1930s solid-state physics, and later semiconductor technology, was a promising research field for a broad range of researchers. Semiconductors were interesting to the radio and telephony industry because of their ability to rectify electrical current (allowing current to flow in one direction and not the other) and they were useful as electronic switches.

In the late 1930s, Bell's director of research Mervin Kelly recognized that a better amplifier was needed for the telephone business. He gave a group of researchers headed by Shockley the freedom to carry out scientific work in the field of solid-state physics. In 1939 Shockley further developed the principle of the field-effect transistor (FET), in which instead of the wire "grid" of De Forest's triode, an electric field controlled the stream of charge carriers between electrodes. This principle started a line of inquiry that led to new experiments. However, Shockley himself went off in other directions and was hardly involved in the further experimental research of this Bell group. Meanwhile Robert W. Pohl and Rudolf Hilsch from Gottingen University made a solid-state amplifier in 1938 using salt as the semiconductor. This was a working device, but it reacted to signals too slowly. Karl Lark-Horovitz and his research team at the physics department of Purdue University, Indiana also became involved in solid-state physics, working on improving the crystal rectifiers that were used as radar detectors in World War II. The team at Purdue worked with both silicon and germanium crystals. Such crystal detectors had no signal gain, but the work on germanium and techniques of growing and doping semiconductor crystals were important to later semiconductor researchers.

The Bell researchers did the most extensive work on crystal rectifiers in the radar program both in America and England. In 1945 John Bardeen, a theoretical solid-state physicist, joined Shockley's group and a semiconductor subgroup was formed within the Bell Lab. Shockley filled out his team with a mix of physicists, chemists, and engineers. In this subgroup Walter Brattain was an experimental physicist. Bardeen and Brattain continued the research on Shockley's earlier design sketches for the field effect transistor. The substitution of the Fleming triode tube (developed by Ambrose Fleming in 1904) by a solid-state device, the transistor, formed the most important outcome of this semiconductor subgroup's research efforts.

In their experiments they placed the electric circuit contacts on two strips of golden foil, since Bardeen suggested that greater amplification could be obtained by placing the two point contacts closer to each other. Bardeen also suggested replacing silicon with high purity germanium that made better rectifying contacts. Germanium is an n-type semiconductor (excess of electrons), and when current flowed in from the gold foil contact, holes were "injected" into the germanium surface. This created a p-n junction as described above. In the junction, current started to flow from one side to the other. In the case of their little construction, current flowed towards the second gold contact. The outcome was that a small current changed the nature of the semiconductor so that a larger, separate current started flowing across the germanium and out the second contact. In other words, a small current was able to alter the flow of a much bigger one, thus effectively amplifying it. The first device was called a point-contact transistor because the wires stood directly in contact with the semiconducting material (Figure 2). Later Shockley developed the junction transistor (also called the sandwich transistor), of which there are two types, called pnp and npn (depending on which material forms the inside layer). The field effect transistor was not built until the 1960s, but today, most transistors are field-effect transistors.

In 1956 Shockley and his colleagues shared the Nobel Prize for their invention of the point contact transistor. Inner competition broke the Bell Lab team apart, but their invention was of great importance for Bell, of which numerous patents and licenses with amongst others General Electric, IBM, Texas Instruments, Philips, and later Sony Electronics bear witness.

Following Bells announcement in 1948 of the first working transistor, the transistor quickly became popular in industry as Bell licensed their transistor, and the first commercial product with transistors, a hearing aid, was sold by Raytheon in 1952. Military applications, as a replacement for the vacuum tube in communications and computing, swiftly followed. Transistors began to replace fragile vacuum tubes in consumer electronic devices, and the first US transistor radio was sold by Texas Instruments in 1954. Sony Electronics especially was able to mass-produce miniaturized transistor radios from 1957. The transistor became the key to further developments in electronic technology and the consumer electronics industry. Considerable developments were made in the 1950s as a result of open sharing of technology between various industrial and university labs. These developments, like the means of introducing dopants (impurities) to very shallow depths using vapor phase diffusion, the use of silicon dioxide as a diffusion mask, and an all diffused silicon transistor enclosed in oxide, led to a wide range of transistors.

A practical problem remained, however. Like the elements in the vacuum tubes, the electric components that formed the transistor needed to be soldered together. The more complex the electric circuits became, the more complicated the construction of the transistor. Computer technology in particular needed complex circuits. Because of this problem, practical application of transistors was slowed down. However, in 1958 Jack Clair Kilby of Texas Instruments developed the first integrated circuit or chip using some key achievements from the 1950s transistor research activities. His invention combined a collection of transistors arranged on a single chip of silicon, in order to save space. This was the first step to integrated circuits that replaced individual transistors in computers—a refinement that led to the development of the modern microprocessor.

See also: Integrated Circuits; Radio Receivers, early; Radio Receivers, Valve and Transistor Circuits; Semiconductors; Vacuum tubes (valves)

Kees Boersma

Further reading

Bardeen, J., "The early days of the transistor", Raju, G.V.S., ed. Proceedings Stocker Symposium, (1979): 3-10

Brattain, W.H., "Genesis of the Transistor", The Physics Teacher, 6 (1968): 108-14

Braun, Ernest and Stuart Macdonald. Revolution in Miniature: The History and Impact of Semiconductor Electronics. Cambridge and New York: Cambridge Press, 1978

Brinkman, William, Douglas Haggan and William Troutman, A History of the Invention of the Transistor and Where It Will Lead Us, IEEE Journal of Solid-State Circuits, 32(12) December 1997 (and on the WWW at www.sscs.org/AdCom/transistorhistory.pdf)

Hoddeson, L., "Innovation and basic research in the industrial laboratory: the repeater, transistor and Bell Telephone System" in Between Science and Technology, edited by Sarlemijn, A., P. Kroes, Amsterdam: North Holland, 1990

Riordan, M, L. Hoddeson, Crystal Fire: the birth of the information age, London and New York: Norton, 1997.

Shockley, W., "The theory of p-n junctions in semiconductors and p-n junction transistors", Bell System Technical Journal, 27 (1949): 435-89

http://www.pbs.org/transistor

Sample Entries