Transistor
When radio was in its infancy, crystals were used as rectifiers, allowing alternating current to pass in only one direction. These were unreliable, however, and such crystal sets were quickly replaced by sets containing radio tubes (see 1904), which had now been in use in various electronic instruments, including the new computers, for over forty years. Radio tubes have to be large enough to enclose a vacuum, however. They are fragile, spring leaks, and frequently have to be replaced. They are also energy-hungry, and one must wait for the filament to heat up before the machine starts working. In 1948 three physicists, William Bradford Shockley (b. 1910), Walter Houser Brattain (1902-1987), and John Bardeen (b. 1908)—all American, though Shockley was British-born—discovered a new kind of crystal. It consisted chiefly of germanium, which conducted an electric current less well than metals but better than such insulators as glass and rubber. Germanium—and silicon, which in a few years replaced germanium, since silicon was both cheaper and better—were considered examples of semiconductors. If traces of certain impurities were added to the semiconductor, the crystal could serve as a rectifier or as an amplifier. It could, in short, perform any function that tubes could. These semiconductors were solid (hence solid-state devices) and required no vacuum, so that they could be made quite small. They were rugged, so that they almost never needed replacing. They used very little energy, and they required no heating, so that they began working at once. A fellow worker, the American electrical engineer John Robinson Pierce (b. 1910), suggested that the device be called a transistor, because it transmitted current across a resistor. With time, transistors replaced tubes completely. Transistors and the improvements that followed may well prove to be the most significant technological advance of the twentieth century.