How LEDs Are Made: The Engineering Behind a Lighting Revolution

If you've walked past the Sydney Opera House at night recently, you've stood in front of one of the most visible advertisements for LED technology in the country. In 2014, the heritage-listed sails were re-lit with energy-efficient LED floodlights — a switch that, according to the ABC, slashed the building's lighting energy use dramatically while giving stage designers a palette of colours the old metal-halide lamps could never produce.

That same year, the Royal Swedish Academy of Sciences awarded the Nobel Prize in Physics to three researchers — Isamu Akasaki, Hiroji Amano and Shuji Nakamura — for inventing the blue light-emitting diode. Without their work, that Opera House light show, the screen you're reading this on, and the bulb above your kitchen bench would not exist in their current form.

So how is an LED actually made? And why was one particular colour — blue — so fiendishly difficult that it took three decades and a Nobel Prize to crack?

The basic recipe: a sandwich of semiconductors

An LED, at its heart, is a tiny semiconductor chip. When electric current passes through it, electrons drop from a high-energy state to a lower one, releasing the leftover energy as a photon — a particle of light. The colour of that light depends entirely on the energy gap between the two states, which in turn depends on the material the chip is made from.

Red and green LEDs were figured out in the 1950s and 60s using semiconductors like gallium arsenide and gallium phosphide. The manufacturing process is, in principle, straightforward:

That's it. The catch is in step one and step two. The crystals have to be almost atomically perfect, and the doping has to be precise to within a few atoms per million. For most of the 20th century, no one could do that with the one material capable of producing blue light: gallium nitride (GaN).

Why blue was a 30-year problem

As AZoM explains in its retrospective on the 2014 Nobel, the prize specifically recognised the breakthrough in blue LEDs — not red, not green, but blue. The reason that one colour mattered so much comes down to physics: blue light has the highest energy of the visible spectrum, which means you need a semiconductor with a wide bandgap to produce it. Gallium nitride has exactly that bandgap, but growing it as a usable crystal proved extraordinarily hard.

The problems were piled on top of each other:

As The Guardian noted when the Nobel was announced, Akasaki and Amano at Nagoya University, working separately from Nakamura at the Japanese company Nichia, plugged away at these problems through the 1980s and into the 90s when most of the industry had given up. Their solutions were partly chemistry, partly metallurgy, and partly sheer stubbornness — Nakamura famously built much of his own equipment because his employer wouldn't fund him.

The breakthroughs that made manufacturing possible

Three specific innovations turned blue LEDs from a curiosity into a manufacturable product:

1. A buffer layer. Akasaki and Amano discovered that depositing a thin layer of aluminium nitride on the sapphire substrate first — at a low temperature — gave the GaN crystal a surface it could grow on with far fewer defects. Suddenly the crystal quality was good enough to work with.

2. Activating the dopant. The team found that bombarding GaN doped with magnesium with a low-energy electron beam — almost by accident — "woke up" the magnesium atoms and finally produced working p-type material. Nakamura later showed that simply heating the material in the right atmosphere did the same job, which made mass production viable.

3. Heterostructures and quantum wells. Nakamura refined the layered sandwich further, sandwiching ultra-thin layers of indium gallium nitride between GaN layers. These "quantum wells" trap electrons and holes in the same tiny region, dramatically boosting the brightness and efficiency of the resulting light.

By the mid-1990s, Nichia was selling commercial blue LEDs. The factory process today is a refined version of the same approach: wafers of sapphire (or, increasingly, silicon carbide or GaN itself) are loaded into a metal-organic chemical vapour deposition (MOCVD) reactor, where gases containing gallium, indium, nitrogen and dopants are flowed over the heated wafer. Layer by layer, each only nanometres thick, the chip is built up. The wafer is then patterned with electrical contacts, diced into individual chips, and packaged.

Why blue was the key to white

You can't light a stage, a screen or a kitchen with blue alone. The reason the blue LED was Nobel-worthy is that it unlocked white light. As Science News Explores set out, white LEDs are made in one of two ways: combine red, green and blue LEDs in the same package, or — far more commonly — take a blue LED and coat it with a yellow phosphor. The phosphor absorbs some of the blue photons and re-emits them at lower energies; mix the leftover blue with the broad yellow glow and your eye perceives white.

That phosphor-coating step is itself a piece of clever manufacturing: the yellow material (usually cerium-doped yttrium aluminium garnet) is suspended in silicone and dispensed in precisely metered drops over each chip, then cured. Adjust the phosphor mix and you adjust the colour temperature, from a warm 2700K reading-lamp glow to a cold 6500K daylight white.

Why this changed lighting forever

The numbers explain why councils, businesses and icons like the Opera House have switched. A modern white LED converts roughly half of the electrical energy fed into it into visible light. An incandescent bulb manages around 5 per cent — the rest is heat. LEDs also last for tens of thousands of hours rather than a thousand or two, contain no mercury (unlike compact fluorescents), can be dimmed and colour-tuned electronically, and are small enough to be embedded into almost any object.

For Australia, where lighting accounts for a meaningful slice of household and commercial electricity demand, the mass adoption of LEDs has quietly become one of the simpler decarbonisation wins of the past decade. Every street light retrofitted, every shop ceiling re-fitted, every Opera House sail re-lit translates directly into less coal burned at the grid.

A Nobel-worthy piece of factory floor

It is easy, in 2024, to take LEDs for granted — they are the default in everything from torches to traffic lights. But the technology sits on a remarkably specific foundation: a wide-bandgap semiconductor that nobody could grow properly until two Japanese university researchers and one company engineer cracked the recipe in the early 90s.

The Nobel committee summed up the importance bluntly when announcing the 2014 prize: incandescent bulbs lit the 20th century, and LEDs will light the 21st. Every time you flick a switch and a chip the size of a grain of rice produces a useful amount of light, you are watching the output of one of the most demanding, atomically precise manufacturing processes humans have ever industrialised.

Related on Bleen

Sources