A glow with no heat
Long before anyone used the word LED, scientists noticed that some solid materials could glow when electricity passed through them, without getting hot. This effect, called electroluminescence, is the basis of the light-emitting diode. It is different from the glow of an incandescent filament, which depends on heat.
First observations
In 1907 the British engineer Henry Joseph Round, working for the Marconi company, a firm built on wireless telegraphy, reported a yellowish glow when current passed through a crystal of silicon carbide, which was being tested as a radio detector. His note was brief and the observation was not pursued at the time.
In the 1920s, the Russian researcher Oleg Losev studied the same kind of glow in crystal detectors more carefully. He published papers describing light emission from silicon carbide junctions and is often credited with the clearest early work, though his research did not lead to a product, and his life was cut short during the Second World War. In the 1930s, the French physicist Georges Destriau studied electroluminescence in zinc sulfide powder, work sometimes called the Destriau effect. These are separate lines of research, and historians treat them as early steps rather than as the invention of the LED itself.
The semiconductor era
The key ingredient for a practical LED was the semiconductor junction. After the invention of the transistor in the late 1940s, researchers understood much better how electrons and holes move across the boundary between two types of semiconductor and release energy as they combine. In certain materials, such as gallium arsenide, much of that energy is released as light.
In 1961 Robert Biard and Gary Pittman, working at Texas Instruments, reported a gallium arsenide diode that gave off infrared light, which is invisible to the eye. Infrared diodes had uses in sensors and remote signaling, and Texas Instruments pursued patents on the device. Around the same time, other laboratories also worked on semiconductor light emission, so priority is shared in the record.
The first visible LED
In 1962 Nick Holonyak Jr., working at General Electric, demonstrated a diode made from gallium arsenide phosphide that gave off visible red light. It is often described as the first practical visible-spectrum LED, and Holonyak is sometimes called the father of the light-emitting diode in popular accounts. That description is a simplification, since the infrared work of Biard and Pittman and the research of others came earlier or at about the same time, but Holonyak’s red LED opened the way to visible displays.
Early red LEDs were dim and expensive. They produced a small, deep red glow suitable only for indicators and not for lighting a room.
Making them practical
Through the 1960s, improved materials and manufacturing slowly reduced the cost. The company Monsanto began mass-producing red LEDs for use as indicator lamps in the late 1960s, and Hewlett-Packard produced LED displays in the same period. In 1972 M. George Craford, a former student of Holonyak, developed a yellow LED and a brighter red one, expanding the range of colors. Green and orange followed, though efficiency was still modest.
LEDs found their first markets in small, specialized devices:
- indicator lights on laboratory instruments and electronic equipment
- seven-segment numeric displays on early pocket calculators and digital watches in the 1970s
- remote controls, which used infrared LEDs
- optical couplers that passed signals between circuits
For these uses their small size, low voltage, quick response and long life were more important than brightness. LEDs also began to replace small filament bulbs in panel lamps because they rarely burned out.
Why they stayed in the shadows
Although LEDs were useful, nothing in the 1960s and 1970s suggested they would one day compete with bulbs for general lighting. The available colors were red, yellow and green, and efficiency was low compared with fluorescent lamps. White light needed a different approach. In the 1970s and 1980s, researchers improved brightness, and by the late 1980s and 1990s red, orange and yellow LEDs could be very bright, which allowed them to be used in vehicle brake lights and traffic signals.
The missing color was blue. Without it, an LED could not be combined with others to make white light, and it could not excite a phosphor to produce white. That obstacle took until the 1990s to overcome and has its own story.
Disputes and credit
As with many inventions, priority claims are tangled. Round observed the effect, Losev analyzed it, Biard and Pittman made a practical infrared diode, and Holonyak made a visible one. Patent filings, company announcements and later honors have each been used to assign the title of inventor to different people. A careful summary is that the LED came from several decades of research by many scientists, with 1961 and 1962 usually named as the years when practical devices first appeared. Its connection to integrated circuits and other semiconductor devices is part of why progress accelerated once manufacturing methods matured.