Tecnico

The invention of the transistor
Michael Riordan
Department of Physics, University of California, Santa Cruz, California 95064

Lillian Hoddeson
Department of History, University of Illinois, Urbana, Illinois 61801

Conyers Herring
Department of Applied Physics, Stanford University, Stanford, California 94305
[S0034-6861(99)00302-5]

Arguably the most important invention of the pastcentury, the transistor is often cited as the exemplar of how scientific research can lead to useful commercial products. Emerging in 1947 from a Bell Telephone Laboratories program of basic research on the physics of solids, it began to replace vacuum tubes in the 1950s and eventually spawned the integrated circuit and microprocessor—the heart of a semiconductor industry now generating annual salesof more than $150 billion. These solid-state electronic devices are what have put computers in our laps and on desktops and permitted them to communicate with each other over telephone networks around the globe. The transistor has aptly been called the ‘‘nerve cell’’ of the Information Age. Actually the history of this invention is far more involved and interesting than given by this ‘‘linear’’account, which overlooks the intricate interplay of scientific, technological, social, and personal interests and developments. These and many other factors contributed to the invention of not one but two distinctly different transistors—the point-contact transistor by John Bardeen and Walter Brattain in December 1947, and the junction transistor by William Shockley a month later.1 Thepoint-contact transistor saw only limited production and never achieved commercial success. Instead, it was the junction transistor that made the modern semiconductor industry possible, contributing crucially to the rise of companies such as Texas Instruments, SONY, and Fairchild Semiconductor. Given the tremendous impact of the transistor, it is surprising how little scholarship has been devoted to itshistory.2 We have tried to fill this gap in recent publications (Herring, 1992; Riordan and Hoddeson, 1997a, 1997b). Here we present a review of its invention, emphasizing the crucial role played by the postwar under1 This paper is based in large part on Riordan and Hoddeson (1997a). The best scholarly historical account of the pointcontact transistor is that of Hoddeson (1981); on the invention ofthe junction transistor, see Shockley (1976). 2 In addition to the above references, see Bardeen (1957), Brattain (1968), Shockley (1973,1976), Weiner (1973), Holonyak (1992), Riordan and Hoddeson (1997b), Ross (1998), and Seitz and Einspruch (1998b). Scholarly books that cover the topic well include those of Braun and MacDonald (1978) and Seitz and Einspruch (1998a).

standing of solid-statephysics. We conclude with an analysis of the impact of this breakthrough upon the discipline itself.
I. PRELIMINARY INVESTIGATIONS

The quantum theory of solids was fairly well established by the mid-1930s, when semiconductors began to be of interest to industrial scientists seeking solid-state alternatives to vacuum-tube amplifiers and electromechanical relays. Based on the work of Felix Bloch,Rudolf Peierls, and Alan Wilson, there was an established understanding of the band structure of electron energies in ideal crystals (Hoddeson, Baym, and Eckert, 1987; Hoddeson et al., 1992). This theory was then applied to calculations of the energy bands in real substances by groups of graduate students working with Eugene Wigner at Princeton and John Slater at MIT. Bardeen and Frederick Seitz,for example, wrote dissertations under Wigner, calculating the work function and band structure of sodium; studying with Slater, Shockley determined the band structure of sodium chloride (Bardeen, 1936; Shockley, 1936; Herring, 1992). By the mid-1930s the behavior of semiconductors was widely recognized to be due to impurities in crystals, although this was more a qualitative than quantitative...