Bipolar junction transistor

The bipolar junction transistor emerged one month after the point-contact transistor because William Shockley understood that fragile mechanical contacts on a germanium slab could never be manufactured at scale. On January 23, 1948—exactly one month after Bardeen and Brattain demonstrated their point-contact device—Shockley conceived a fundamentally different approach based on the p-n junction discovered by Russell Ohl in 1940.

The adjacent possible had aligned during World War II. MIT's Radiation Laboratory managed extensive radar research that required crystal diode rectifiers as frequency mixer elements. Purdue University produced high-quality germanium semiconducting crystals. Karl Lark-Horovitz's research team advanced understanding of semiconductor properties. When Mervin Kelly reactivated Bell Labs' solid-state amplifier project in 1945, this wartime knowledge infrastructure was waiting.

Shockley's motivation was partly professional jealousy—he resented not being directly involved in the point-contact breakthrough—but his engineering insight was sound. The point-contact transistor relied on delicate surface effects: two gold contacts held by a plastic wedge, touching a germanium slab. Manufacturing these with consistent quality was nearly impossible. The BJT used minority carrier injection, where positively charged holes penetrate through bulk semiconductor material—a solid three-layer sandwich with no fragile contacts.

Shockley invented the grown-junction transistor on June 23, 1948, and filed his patent three days later. But Bell Labs waited until July 4, 1951 to announce it publicly, spending three years perfecting the manufacturing process. Gordon Teal developed crystal-pulling techniques for high-purity germanium in 1950-1951. William Pfann invented zone-refining for even higher purity in 1951.

Convergent emergence appeared immediately. In late 1948, Herbert Mataré and Heinrich Welker, German radar program physicists working at a Westinghouse subsidiary in Paris, claimed independent invention of their 'transistron.' By mid-1949, thousands were manufactured for the French telephone system. The conditions had aligned simultaneously on both sides of the Atlantic.

Bell Labs shared its knowledge deliberately. At the April 1952 Transistor Technology Symposium, 40 companies paid $25,000 licensing fees to attend a nine-day intensive course including a visit to Western Electric's production line. General Electric, RCA, IBM, Texas Instruments, and Sony all attended. The published proceedings became the 'bible of the semiconductor industry.' Jack Morton believed Bell could benefit from advances made elsewhere—and AT&T feared antitrust action.

The cascade was transformative. Texas Instruments produced the first commercial silicon BJT in April 1954, quickly dominating military markets. The Regency TR-1 transistor radio launched that holiday season. By 1957, IBM CEO Thomas Watson Jr. decreed that no new machines would use vacuum tubes. The November 1953 University of Manchester Transistor Computer was the first in the world. Robert Moog's synthesizer used silicon transistors' exponential voltage-current relationship to create voltage-controlled oscillators.

Today, MOSFETs dominate digital circuits, but BJTs persist in specialized applications: audio amplifiers, RF circuits, low-noise analog systems, and precision matching. Even CMOS chips use parasitic BJTs for bandgap voltage references, temperature sensors, and electrostatic discharge protection. The original insight—that bulk semiconductor physics could replace surface effects—made everything digital possible.