The Invention of the Transistor

Photo: Mehmet Turgut Kirkgoz / Pexels
The transistor is the fundamental switching element of all modern electronics, invented on December 16, 1947 at Bell Telephone Laboratories in Murray Hill, New Jersey by John Bardeen and Walter Brattain — with William Shockley providing theoretical guidance. A transistor controls the flow of electrical current using a small control signal, acting as either an amplifier or a switch. The device that Bardeen and Brattain first demonstrated — made of gold foil, a plastic wedge, and a slab of germanium — launched the semiconductor age. Today's computer chips contain up to 100 billion transistors, each smaller than a virus.
What Bell Labs Was Trying to Solve
Bell Telephone Laboratories in the 1940s was attempting to solve a practical engineering problem: the telephone network was built on vacuum tubes, and vacuum tubes were failing constantly. The US telephone system depended on amplifiers to boost long-distance signals, and each amplifier station used vacuum tubes — glass cylinders the size of a lightbulb, filled with metal electrodes and evacuated of air. They were hot, large, energy-hungry, and short-lived. A transcontinental telephone call passed through thousands of tubes. When any tube failed, the call dropped.
Vacuum tube — an electronic component consisting of metal electrodes in an evacuated glass envelope, used to amplify or switch electrical signals; the predecessor to the transistor.
Bell Labs leadership understood that solid-state physics — physics of crystalline materials like silicon and germanium — might offer an alternative. The quantum mechanical description of electrons in solids had been advancing rapidly since the late 1920s, and theorists were beginning to understand how certain materials conducted electricity through two types of carriers: electrons (negative) and electron holes (positive). William Shockley, head of the solid-state physics group at Bell Labs, believed a field-effect device — where an electric field perpendicular to a semiconductor surface modulates current flow — was achievable.
Shockley's first field-effect attempts in 1945 and 1946 failed. He could not get the external field to penetrate the semiconductor surface effectively. The surface was trapping charges and screening the field. This problem, the "surface state" problem, was what Bardeen was hired to explain theoretically, and what Brattain was tasked with probing experimentally.
Source: Computer History Museum (2023). "1947: Invention of the Point-Contact Transistor" — Silicon Engine exhibit. Primary source documentation of the research timeline and device specifications.
How a Transistor Actually Works
The transistor exploits the electrical properties of semiconductors — materials whose electrical conductivity falls between that of conductors (metals) and insulators (glass or rubber). Pure silicon or germanium conducts electricity poorly, but adding small amounts of impurity atoms — a process called doping — creates two distinct types of semiconductor.
N-type semiconductor — a semiconductor doped with atoms that donate extra electrons (negative charge carriers), such as phosphorus in silicon.
P-type semiconductor — a semiconductor doped with atoms that create "holes" — absences of electrons that behave like positive charge carriers — such as boron in silicon.
The original point-contact transistor worked differently from the transistors in modern chips, but the underlying physics is related. In the junction transistor that Shockley developed shortly after — and which became the standard design — a thin layer of one type of semiconductor is sandwiched between two layers of the opposite type. This creates a p-n-p or n-p-n structure with three terminals: the emitter, base, and collector.
The transistor's genius is this: a very small current injected into the base terminal controls a much larger current flowing between the emitter and collector. This is amplification — the transistor takes a weak signal and produces a proportionally stronger one using power from a separate supply. In switching mode, the base current either permits or blocks the emitter-collector current entirely, producing binary 1 (current flows) or 0 (current blocked).
This is analogous to a valve controlling a water pipe. A tiny turn of the valve handle — representing the base current — can open or close a pipe carrying a large volume of water. The handle does not supply the water; it controls whether the pipe's reservoir can flow.
Source: Bardeen, J. and Brattain, W.H. (1948). "The Transistor, A Semi-Conductor Triode" — Physical Review, Vol. 74, pp. 230–231. The paper announcing the discovery to the physics community.
In a modern MOSFET (Metal-Oxide-Semiconductor Field-Effect Transistor), the mechanism is slightly different. A voltage applied to the gate terminal creates an electric field through a thin insulating layer of silicon dioxide, which induces a conducting channel between the source and drain terminals. The gate draws essentially no current while controlling substantial source-drain current — making the MOSFET extraordinarily energy-efficient and ideal for digital logic. This is the transistor architecture that dominates computer chips today.
December 16 and December 23, 1947
The invention happened in two stages. On December 16, 1947, Bardeen and Brattain made the first successful demonstration of amplification using a solid-state device. They had built a point-contact device: two strips of gold foil separated by a razor blade's width, pressed onto the surface of a germanium crystal. A signal applied to one gold contact was amplified by the other. The device worked, but barely, and only intermittently.
Over the following week, Brattain refined the device and increased its reliability. On December 23, 1947 — one week later, and two days before Christmas — Bardeen and Brattain demonstrated the improved device to Bell Labs management in a private session. They connected a microphone to one terminal and a speaker to another, and the device amplified speech. The executives in the room understood they were watching something significant. The demonstration was the formal institutional moment of invention.
The device was called a "transistor" — a name coined by Bell engineer John Pierce, who combined "transconductance" and "varistor" to produce a word that captured what the device did: it transferred resistance across the semiconductor material.
William Shockley was not in the room for the December 23 demonstration. He had been kept at arm's length from the experimental work because of professional tensions within the group. He learned of the success that evening. His reaction was not celebration but something closer to jealousy. Shockley immediately set to work on an improved design he could call his own.
Source: Computer History Museum (2023). "December 23: Bardeen and Brattain Demonstrate the Transistor to Bell Labs Supervisors" — This Day in History entry, with primary source citations from Bell Labs records.
Within six months, Shockley had developed the junction transistor — a more robust design using a thin layer of p-type germanium sandwiched between two n-type layers. Unlike the point-contact device, which was mechanically delicate and electrically noisy, the junction transistor was reliable, reproducible, and manufacturable. The junction transistor became the commercial product. Shockley's design is the conceptual ancestor of every transistor manufactured since.
The Three Men and Their Diverging Fates
The transistor story involves three men of radically different character, and their subsequent trajectories illustrate how personality shapes scientific legacy.
John Bardeen was reserved, methodical, and widely regarded as one of the most talented theoretical physicists of the twentieth century. After the transistor work, his relationship with Shockley deteriorated, and he left Bell Labs to join the University of Illinois. There, working with Leon Cooper and J. Robert Schrieffer, he co-developed the BCS theory of superconductivity — explaining how electrons pair up in certain materials at low temperatures to flow without resistance. For BCS theory, Bardeen shared the 1972 Nobel Prize in Physics. He remains the only person ever to have won two Nobel Prizes in Physics. He was 64 when he received his second.
Walter Brattain was an experimentalist's experimentalist — skilled with his hands, patient with finicky equipment, and productive in the lab. He spent his career at Bell Labs after the transistor work, was known for his collegial nature, and never sought the spotlight. He received the Nobel jointly with Bardeen and Shockley in 1956 and continued working in solid-state physics until he retired.
William Shockley was a physicist of genuine brilliance and a manager of extraordinary toxicity. He had an uncommon ability to see where physics was pointing, as demonstrated by his theoretical frameworks for the transistor and his eventual junction transistor design. But he was also competitive, suspicious of colleagues, and unable to tolerate the success of others. When he founded Shockley Semiconductor Laboratory in Mountain View, California in 1956 — the move that seeded Silicon Valley — he managed his team so poorly that eight of its most talented members, including Robert Noyce and Gordon Moore, resigned en masse. They became the founders of Fairchild Semiconductor, and later Intel.
In his later years at Stanford, Shockley devoted increasing energy to arguing that Black Americans were genetically inferior in intelligence and advocated for a "Voluntary Sterilization Bonus Plan" that would pay low-IQ women to undergo sterilization. These views had no scientific basis, were condemned by the scientific community, and cost him his professional reputation. He died in 1989 estranged from most of his family and colleagues.
Source: Wikipedia contributors (2025). "William Shockley" — synthesizing material from: Riordan, M. and Hoddeson, L. (1997). Crystal Fire: The Invention of the Transistor and the Birth of the Information Age — Norton. The authoritative history of the invention and its aftermath.
The Three Nobels Connected to the Transistor's Story
The transistor's story touches three separate Nobel Prizes:
The 1956 Nobel Prize in Physics went to Bardeen, Brattain, and Shockley "for their researches on semiconductors and their discovery of the transistor effect." It was awarded nine years after the discovery — unusually delayed, perhaps because the full significance of the transistor was still being absorbed.
The 1972 Nobel Prize in Physics went to Bardeen, Cooper, and Schrieffer for BCS superconductivity theory — making Bardeen the unprecedented two-time Physics Nobel winner.
The 2000 Nobel Prize in Physics went partly to Jack Kilby for the invention of the integrated circuit — the technology that allowed transistors to be fabricated by the billions on a single chip of silicon. Kilby's co-inventor Robert Noyce (who independently developed the IC at Fairchild, the company Shockley's toxic management had inadvertently created) had died in 1990 and could not share the prize.
From Point-Contact to MOSFET to 2 nm: The Scaling Story
The transistor evolved through several distinct generations of design, each enabling a step change in capability.
The point-contact transistor (1947) was a laboratory curiosity — hand-assembled from foil and germanium, large, noisy, and inconsistent.
The junction transistor (1951, Shockley) was the first commercially manufacturable design. Using layered semiconductor materials, it was reliable and reproducible enough to replace vacuum tubes in hearing aids, radios, and eventually computers.
The planar process (1959, Jean Hoerni at Fairchild) allowed transistors to be fabricated in a flat layer of silicon using photolithography — the same process used to print patterns on film, but used to etch semiconductor structures. This made it possible to build multiple transistors on a single chip and connect them with deposited metal lines.
The MOSFET (Metal-Oxide-Semiconductor Field-Effect Transistor, 1959, Kahng and Atalla at Bell Labs) was the design that would eventually dominate digital logic. It uses a gate separated from the semiconductor by a thin oxide layer, draws essentially no current in steady state, and scales with extraordinary efficiency.
Moore's Law — the observation by Gordon Moore (1965) that the number of transistors on a dense integrated circuit doubles approximately every two years — is not a law of physics but a self-fulfilling prediction that organized the semiconductor industry's roadmap for 60 years. Each generation of miniaturization required new materials, new lithography techniques, and new physics.
| Year | Process Node | Transistors per Chip (approx.) | Key Innovation |
|---|---|---|---|
| 1947 | — | 1 | Point-contact transistor |
| 1971 | 10,000 nm | 2,300 (Intel 4004) | First commercial microprocessor |
| 1989 | 1,000 nm (1 µm) | ~1 million | Sub-micron lithography |
| 2003 | 90 nm | ~77 million | Strained silicon |
| 2012 | 22 nm | ~1.4 billion | FinFET architecture |
| 2022 | 3 nm | ~15 billion (Apple M2) | Gate-all-around transition |
| 2025 | 2 nm | ~490 million/mm² density | GAA nanosheet, TSMC N2 |
The transistor shrinkage timeline — from millimeter-scale to sub-nanometer structures in 78 years.
Source: TSMC (2025). "TSMC 2nm Process Technology" — product announcement; EE Times (2024), "TSMC's Chip Scaling Efforts Reach Crossroads at 2nm."
At the 2 nm node, transistor gate lengths are approximately 2 nanometers — 10 silicon atoms wide. At this scale, quantum tunneling becomes a design constraint: electrons can literally tunnel through the gate insulator rather than flowing through the intended channel. Managing quantum effects is now as important as managing classical electrical characteristics.
TSMC began mass production of its N2 (2 nm class) process in late 2025, using Gate-All-Around (GAA) nanosheet transistors, where the gate material wraps completely around the conducting channel on all four sides, giving engineers precise electrostatic control at atomic scales. The first products using N2 include Apple's upcoming silicon and NVIDIA accelerator chips.
The Patent Dispute and Bell Labs' Unusual Decision
Bell Labs filed patents on the transistor in 1948. Then, in an unusual decision driven partly by antitrust concerns — AT&T was already under scrutiny as a regulated monopoly — Bell Labs licensed the transistor patents broadly and cheaply to any company willing to pay a $25,000 fee. This decision, arguably one of the most consequential technology policy choices of the twentieth century, meant that dozens of companies could build transistors without negotiating exclusive deals with AT&T.
The beneficiaries included Sony (which used the transistor license to build the first mass-market transistor radio in 1955), Texas Instruments (where Jack Kilby invented the integrated circuit), Fairchild Semiconductor, and eventually Intel. The semiconductor industry that emerged from this open licensing regime is the economic foundation of the modern digital economy.
Had AT&T maintained exclusivity over transistor technology, the development of the integrated circuit, the microprocessor, and the personal computer would likely have been delayed by years or decades, and concentrated in a single vertically integrated corporation rather than the competitive ecosystem that actually emerged.
Source: Engineering and Technology History Wiki (ETHW). "Bell Demonstrates Transistor" — entry documenting the licensing decision and its competitive consequences.
Unintended Consequences: What Nobody Predicted
The transistor's inventors did not anticipate the Internet. They did not anticipate the smartphone. They did not anticipate that their device would eventually fit 100 billion copies onto a chip the size of a fingernail. The Bell Labs team's horizon was the telephone network — they wanted a better amplifier for long-distance calls.
The unintended consequences of the transistor are so vast as to encompass essentially all of modern civilization's information infrastructure. Some specific examples:
The transistor made digital computing practical. Mechanical computers and vacuum-tube computers were too large, hot, and unreliable for widespread use. The transistor's reliability and miniaturizability made the mainframe possible, then the minicomputer, then the personal computer, then the smartphone.
The transistor created Silicon Valley. Shockley's 1956 move to Mountain View, and his team's subsequent exodus to found Fairchild, established the geographic and cultural template for the tech industry. The venture-capital model, the spinoff culture, and the cluster dynamics of the Bay Area tech economy all trace to the eight engineers who walked out on Shockley.
The transistor enabled the GPS system, the MRI machine, satellite communications, and the global financial system — all of which depend on digital signal processing at speeds that vacuum tubes could not achieve.
The energy cost of transistors at scale has become a major environmental concern. Data centers consumed approximately 240–340 TWh of electricity globally in 2022 (International Energy Agency, 2023) — roughly one percent of global electricity consumption. As AI model training and inference workloads grow, this figure is rising. The transistor's efficiency gains, which have historically offset the energy cost of adding more compute, are slowing as physical scaling limits are approached.
Where It Stands Today
The transistor is not a solved problem. At sub-5 nm dimensions, semiconductor physics requires constant material and structural innovation. The move from planar MOSFET to FinFET to GAA nanosheets reflects the necessity of new architectures when simple geometric scaling becomes insufficient.
Beyond silicon, research is active in carbon nanotube transistors (demonstrated to work at sub-1 nm gate lengths), two-dimensional material transistors based on molybdenum disulfide, and quantum dot transistors that exploit quantum confinement. These are laboratory demonstrations, not production technologies — but the transistor's history shows how laboratory demonstrations become the products of the next decade.
At the system level, the most significant development is the heterogeneous integration of transistors — stacking multiple chips in three-dimensional packages (3D IC), connecting different types of chips (CPU, memory, analog) using high-density interconnects. This is a response to the slowdown of single-chip scaling: rather than making transistors smaller, engineers are connecting more of them more densely.
The transistor that Bardeen and Brattain assembled from gold foil and a germanium crystal in December 1947 generated a power gain of roughly 18 dB. A modern MOSFET operates at frequencies exceeding 300 GHz and switches in under 1 picosecond. The underlying physics — a small signal controlling a large current through a semiconductor junction — is recognizably the same. The scale of everything else has changed by a factor of roughly a billion.
Source: International Energy Agency (2023). "Data Centres and Data Transmission Networks" — factsheet on global electricity consumption.
FAQs
Q: Who actually invented the transistor — Bardeen, Brattain, or Shockley?
A: All three contributed. Bardeen provided the theoretical understanding of surface states that made the device possible. Brattain built and tested the point-contact device with Bardeen. Shockley's junction transistor design became the commercially dominant form. The 1956 Nobel correctly recognized all three, though the division of credit was a source of professional conflict throughout their careers.
Q: Why did Shockley not participate in the first demonstration?
A: Shockley had been pursuing a different design approach (field-effect rather than point-contact) and was frustrated when Bardeen and Brattain achieved success with their approach. He was kept at some distance from the experimental work due to professional tensions, and the December 23 demonstration happened without him.
Q: What is Moore's Law and is it still true?
A: Moore's Law is Gordon Moore's 1965 observation that the number of transistors on a dense chip doubles approximately every two years. It held reasonably well for about 50 years. Today it is slowing: the 3nm-to-2nm transition at TSMC took approximately three years rather than two, and physical scaling is encountering quantum limits. The "law" continues in modified form through architectural innovations like 3D stacking rather than flat geometric shrinkage.
Q: What is the difference between a transistor and a microchip?
A: A transistor is a single electronic switching or amplifying element. A microchip (integrated circuit) is a piece of semiconductor material containing millions to billions of transistors, along with resistors, capacitors, and metal interconnects, all fabricated in a single manufacturing process. Jack Kilby at Texas Instruments and Robert Noyce at Fairchild independently invented the integrated circuit in 1958–1959.
Q: How small is the smallest transistor in production today?
A: TSMC's N2 process, entering mass production in 2025, achieves a transistor density of approximately 490 million transistors per square millimeter. The "2 nm" label refers to a marketing classification, not the literal gate length — actual gate lengths in N2 transistors are closer to 6–8 nm. The 2 nm label indicates the process generation, not a single physical dimension.
Q: Did Bell Labs give away the transistor patent?
A: Not exactly. Bell Labs licensed the transistor broadly under an unusual arrangement driven partly by antitrust concerns. Any company could license the technology for a $25,000 fee, which was modest for the era. This open licensing policy is credited with enabling the rapid proliferation of transistor-based products and the emergence of the competitive semiconductor industry.
Read next:
Follow new articles
Add SciHub101 to your feed reader to see new science and technology stories.
About the Author
SciHub101 is an independent science and technology publication founded and edited by Buğra Sözeri, a software engineer and app developer based in Türkiye. AI tools assist with drafting and research. Source documentation in the existing archive is uneven, and articles are being reviewed and corrected. The Editorial Team byline identifies editorial responsibility; it does not certify specialist or independent human review. Where we compare products, anything built by our own team is labelled as such — see the editorial guidelines and About page for how corrections and disclosures work.
Related Articles

Velcro: How a Dog Walk Became a $2.8 Billion Fastener Industry
How Swiss engineer George de Mestral turned burdock burrs stuck to his dog into the hook-and-loop fastener that NASA used, kids wear, and trademark lawyers fight over.

The Invention of the Steam Engine
How Thomas Newcomen's 1712 atmospheric engine and James Watt's 1769 separate condenser transformed fire into motion and launched the Industrial Revolution.

How the Transistor Changed Everything
Demonstrated at Bell Labs in December 1947, the transistor became the building block of every digital device on earth.

Microchip Invention: Kilby, Noyce, and the Integrated Circuit
The microchip emerged from parallel breakthroughs by Jack Kilby and Robert Noyce. This guide explains the invention that made modern computing possible.

The Invention of the Solar Cell
How Bell Labs' 1954 silicon breakthrough, a 1958 satellite, and a 99.9% price collapse turned sunlight into the world's cheapest source of electricity.

The Microchip: How Two Men Independently Shrank a Room
Jack Kilby and Robert Noyce independently invented the integrated circuit in 1958–59. Here's how a germanium sliver and a silicon wafer changed everything.