In the early 20th century, scientists knew how to make diodes at both ends by contacting sharp metal probes with semiconductor crystals. These point-contact diodes can turn an oscillating signal into a stable signal and are widely used as detectors in crystal radio receivers. By the 1920s, inventors began working on the use of semiconductors to amplify and switch signals.
Some of the earliest semiconductor amplifier work came from Eastern Europe. In 1922-23, Russian engineer Oleg Losev of the Nizhegorod Radio Laboratory in Leningrad discovered a special mode of operation of contact zinc ore (ZnO) crystal diodes at the point of contact with zinc ore (ZnO) that supported signal amplification up to 5 MHz. Although Losev has been experimenting with this material in radio circuits for many years, he died during the siege of Leningrad in 1942 and cannot justify his place in history, so his work is largely unknown.
Oleg Losev
The Austro-Hungarian physicist Julius E. Lilienfeld moved to the United States and in 1926 applied for a patent for "Method and apparatus for controlling electric current," in which he described a three-electrode amplification device using copper sulfide semiconductor materials. Lilienfeld is credited with inventing the electrolytic capacitor, but there is no evidence that he made a working amplifier. However, his patent bears enough resemblance to later field-effect transistors to reject future patent applications for the structure.
Julius E. Lilienfeld
German scientists also contributed to this early study. While working at the University of Cambridge in England in 1934, German electrical engineer and inventor Oskar Heil applied for a patent that controlled the current in semiconductors — essentially field-effect transistors — through capacitive coupling on electrodes.
In 1938, Robert Pohl and Rudolf Hilsch experimented with potassium bromide crystals with three electrodes at the University of Göttingen. They reported amplification of low-frequency (about 1 Hz) signal. None of these studies resulted in any applications, but Heil is remembered today in audiophile circles for its use of air motion transformers in high-fidelity loudspeakers.
Genius Shockley
Because of its poor reliability and high power consumption, AT&T engineers knew by the late 1930s that vacuum tube circuits could not meet the company's rapidly growing demand for telephone capacity. Mervin J. Kelly, director of research at Bell Labs, assigned William Shockley to investigate the possibility of replacing tubes with semiconductor technology.
In early 1945, Shockley experimented with a field-effect amplifier similar in concept to Heil and Lilienfeld's patent, using improved semiconductor materials developed for radar detectors during the war, but failed to work as he had hoped. Physicist John Bardeen proposed that electrons on the surface of semiconductors might prevent electric fields from penetrating into materials.
Under Shockley's guidance, Bardeen, together with physicist Walter Bratattain, began studying the behavior of these "surface states."
John Bardeen、William Shockley 和 Walter Brattain
Walter Houser Brattain (1902–1987) followed his parents to settle on a ranch near Tonaskeete, Washington. He received his master's degree from the University of Oregon and his Ph.D. from the University of Minnesota. Brattain joined Bell Labs in 1929 as a research physicist, where he was considered a skilled experimentalist.
Theoretical physicist John Bardeen (1908-1991), born in Madison, Wisconsin, skipped third grade in school as a child prodigy. He received his master's degree from the University of Wisconsin and his Ph.D. from Princeton University, where he developed an interest in solid-state physics.
Secretly codenamed "The Surface State Job," their project was an important priority for AT&T research agency Bell Labs, which sought to find smaller, lower-power alternatives to bulky, power-hungry vacuum tubes. Mervin J. Kelly, the lab's director of research, thinks crystalline semiconductor materials, such as germanium or silicon, may offer a solution. To this end, in 1936, he recruited William Shockley from the Massachusetts Institute of Technology (MIT) to study solids.
William Bradford Shockley, born in London, England, to American parents, spent his youth in Palo Alto, California, just yards from the famous HP garage. As a precocious child, he was "short-tempered, spoiled and almost uncontrollable, which made the life of his doting parents miserable". He received his B.A. from the California Institute of Technology and his Ph.D. in theoretical physics from the Massachusetts Institute of Technology. The talented Intel company Gordon Moore commented that Shockley "can see electronics", but Shockley is egotistical and capricious, so while he has the support of management, he is less popular among his peers. Nobel laureate Charles Townes said even more bluntly: "He knows everything, but he doesn't understand people." ”
In 1939, convinced that he could find a solution based on solid materials, he wrote: "Today it occurred to me that, in principle, amplifiers could be made using semiconductors instead of vacuum rolls. "Brattain assisted Shockley in experimenting with what we today call field-effect transistors (FETs), but no useful results were achieved.
But the ensuing Second World War disrupted the work, but it was restored in 1945 when Shockley hired John Bardeen and asked him to see if he could spot anything wrong with his design. Bardeen initially concluded that it should work.
A FET is a device that uses an electric field to control the flow of current in a semiconductor material. Shockley published a paper during the MIT era that hypothesized that electrons near surfaces could move freely like electrons in the body of the material. On March 19, 1946, Bardeen theoretically determined that this claim was not valid. He concluded that electrons in this region must be trapped, creating a surface state that creates a movement barrier.
Bardeen and Brattain, with the help of physicist Gerald Pearson and chemist Robert Gibnety, worked to figure out if he was right. By the beginning of 1947, in laboratory experiments, they demonstrated the existence of a barrier. As their manager, Shockley advises on how to break down barriers, but is not involved in their day-to-day work.
Magical November
On Monday, Nov. 17, Gibney suggested that Brattain apply voltage between the metal plate on the upper surface and the contacts on the back of the germanium crystal plate to create a strong electric field perpendicular to the surface. A drop of liquid electrolyte at the electrical contact contact material neutralizes the surface state and creates measurable field effects in the structure.
Following Bardeen's suggestion to probe the surface with a sharp metal dot surrounded by electrolyte, on November 21, Brattain made a functional amplifier, albeit at a very low frequency. During what Shockley calls the "Magic Month," weeks of long and frenetic activity took place on blackboards and lab benches, combining the occasional "accident" of handling material with clever intuition, using what they had learned in the absence of electrolytes.
Bardeen calculated that reducing the distance between the two contacts would enhance the effect. Brattain came up with an ingenious way to bond gold leaf to a plastic wedge and then surgically precisely cut the tip with a razor blade to create two contact points separated by the width of a sheet of paper.
On the afternoon of Tuesday, December 16, 1947, they connected a spring that pressed the rough device firmly against the germanium surface. Brattain found that if he swung it just right, "I had an amplifier with a magnification of 100, clear to the audio range." ”
Thus, solid-state semiconductor amplifiers were born.
Components of Bardeen and Brattain transistors
He and Bratton agreed: "We should tell Shockley what we did today. ”
Bardeen rarely discusses his work at home; That night, however, he casually said to his wife, who was peeling carrots in the kitchen: "I found it today." "That's great," she replied subconsciously. After some time, Jane discovered that the thing was a transistor.
It is worth mentioning that at the beginning of the same year, the German physicist Herbert Mataré and his colleague Heinrich Welker, while studying a phenomenon he called "interference", independently created a germanium-based amplifier with two point contact points on the surface of the Westinghouse laboratory in Paris, France. When they learned of Bell Labs' announcement, Mataré and Welker patented their own device, which they called "transistors."
The transistor was finally born
Shockley admits that Bardeen and Brattain's news stirred conflicting emotions in his heart. "My joy at the team's success was offset by the frustration of not being one of the inventors." Shockley said. However, he was equally aware of the importance of their breakthrough and planned to arrange an amplifier demonstration for Bell executives on the afternoon of Tuesday, December 23, 1947.
Brattain's record of the December 23, 1947 presentation
Brattain recorded in his notebook with a microphone and headphones, "This circuit has actually been discussed and... It can be heard and seen in the oscilloscope demo. Sadly, no one remembers what was said, only that it worked. Shockley called it a "wonderful Christmas present."
Within days of Christmas, Bell Labs' patent attorneys began documenting their work and preparing for public announcements. Because Shockley's self-drive and self-promotion made him Bell Labs' most visible spokesperson, orders were issued that Bardeen and Brattain not be photographed in his absence. Promotional photos from the time show him on the front and center of the scene.
At the first press conference in New York on June 30, 1948, a spokesman claimed that transistors "may have far-reaching significance in electronic and electrical communications." Unmoved, the New York Times downgraded the story to a "broadcast news" page — below the announcement from soap opera sponsors.
Brattain's colleague John Pierce is believed to have come up with the name. He also realized that it worked according to the cross-resistor principle, and Pierce derived transistors from related electronic components called resistors.
AT&T's equipment division, Western Electric, began manufacturing point-contact transistors in 1951 and by mid-1952 was producing more than 6,000 devices per month, primarily for telephone switching systems and hearing aids.
Sonotone 1010 (1952) The first commercial hearing aid to use transistors
According to Bratattain, Shockley is pushing to incorporate some of his ideas into their patent applications, "Shortly after the presentation, Bardeen and I called us separately and told us that sometimes the person doing the work didn't get me telling him, 'Oh hell, Shockley, that's enough glory for everyone.'" But he left alone, worked from home, and in a way ceased to be part of the research team anymore.
However, based on his theoretical contributions to understanding semiconductor physics and his invention of the junction transistor, Shockley received the 1956 Nobel Prize in Physics along with Bardeen and Bratton for "the discovery of semiconductor research and the transistor effect."
Jealous of not participating more pronounceably in the invention of the transistor, as well as the need to maintain his position relative to his subordinates, Shockley began a month-long intensive theoretical activity. He determined that point-contact transistor operation was not due to a hypothetical near-surface field effect, but due to a completely different structure in the crystal body called the PN junction.
As a result of this work, Shockley conceived on January 23, 1948, a radically different type of component called a junction transistor, which proved to be more reliable and easier to mass-produce than point-contact devices. Until Bell Labs announced progress on July 4, 1951, manufacturing working transistors remained a formidable challenge. His version became the main active electronic building block for decades to come by enabling a new generation of powerful computers.
It is worth mentioning that based on his theoretical contributions to understanding semiconductor physics and his invention of the junction transistor, Shockley received the 1956 Nobel Prize in Physics with Bardeen and Brattain for "the discovery of semiconductor research and the transistor effect."
Bardeen and Brattain accepted the Nobel Prize
And in the ensuing development, many different manufacturing methods were developed to make transistors faster, cheaper and more reliable. An important advance in 1954 was the silicon transistor, first developed by Morris Tanenbaum of Bell Labs and soon after by a team led by chemist Willis Adcock of New Guide Texas Instruments. By the end of the 2050s, silicon had become the industry's material of choice, and TI became a major semiconductor supplier.
Subsequently, the legendary announcement was made that Fairchild's dual-diffused silicon countertop transistor introduced in 1958 was a huge commercial success. Swiss physicist Jean Hoerni's revolutionary planar process solves reliability problems that threaten the company's future. Hoerni's planar technology not only transformed transistor manufacturing from semi-manual to high-volume automated production. It has also facilitated the development of modern integrated circuits (ICs).
The arrival of MOSFETs
After several years of development, the idea of Lilienfeld and Heil and Shockley's failed early experiments finally bore fruit in 1959. At that time, Korean electrical engineer Dawon Kahng worked for Egyptian engineer Martin M. (John) Atalla at Bell Labs to study semiconductor surfaces and built the first successful effector transistor (FET: field-effect transistor), consisting of metal (M–gate), oxide (O – insulation) and silicon (S – semiconductor) layers. This is also known as MOSFETs (often referred to simply as MOS), which made smaller, cheaper, and lower power transistors possible.
Dawon Kahng
Fairchild and RCA introduced commercial MOS transistors in 1964. But in the decade of solving early manufacturing problems with MOS processes, individual transistors in computer systems have largely been replaced by ICs. In the long run, MOS transistors are proving to be the most practical way to build high-density ICs such as microprocessors and memory. Nearly 100% of the billions of transistors manufactured every day are MOS devices.
As with most technological developments, the invention of the modern transistor followed a baconian model, that is, gradually emerged from a "growing knowledge base" built by a truly international group of engineers and scientists, rather than from the individual efforts of a heroic "inventor". ”