The Royal Swedish Academy of Sciences will award this year's Nobel Prize in Physics to Isamu Akasaki (Meijo University, Nagoya, Japan), Hiroshi Amano (Nagoya University), and Shuji Nakamura (University of California, Santa Barbara).
From left to right: Isamu Akasaki, Hiroshi Amano, and Shuji Nakamura.
Starting in the 1970s, the three researchers tackled a range of challenges in device physics and materials science to create light-emitting diodes that could shine blue light. Red and green LEDs were already available by the late 1960s. The advent of the first luminous blue LED, which took place in 1993, completed the visual spectrum. A wide range of potential applications, from domestic lighting to optical storage, opened up.
Compared with incandescent light bulbs, LEDs are 10 times more energy efficient, last 100 times longer, and are much more resistant to vibration and shock. Given that 20–30% of the world's electricity is consumed by lighting, the widespread adoption of LEDs will significantly reduce the world's energy consumption and, with it, its emission of carbon dioxide into the atmosphere.
Echoing the words of Alfred Nobel's will, the Nobel selection committee remarked that the invention of the blue LED is "of great benefit to mankind."
Toward blue LEDs
Blue LEDs work in the same way as their red antecedents. Two layers of semiconductor, one p-doped, the other n-doped, abut each other. Applying voltage across the layers, from p to n, drives the extra electrons from the conduction band of the n-doped layer to fill holes in the valence band of the p-doped layer.
If the electrons can cross the bandgap without having to gain or shed momentum—that is, if the conduction band’s minimum and the valence band’s maximum face each other across momentum space—each electron–hole recombination yields a photon whose energy matches the bandgap.
Materials that have such "direct" bandgaps make efficient LEDs, but they are the exception rather than the rule among semiconductors. The world's preeminent semiconductor, silicon, has an indirect bandgap. Most LEDs—from the original red LEDs to the prize-winning blue LEDs—are made of direct-bandgap compounds drawn from elements from groups III and V of the periodic table.
Red and green LEDs are made from gallium arsenide and gallium phosphide. In principle, extending the family to achieve shorter wavelengths entails pairing Ga with a lighter element from group V, nitrogen, whose smaller size yields tighter binding and, with it, a wider bandgap.
The quest to harness GaN's bandgap for light emission began in the 1950s even before the red LED made its debut in 1962. By the early 1970s, progress had foundered. Making pure GaN device-sized crystals, let alone doped crystals, proved too difficult.
Prospects brightened in the mid 1970s when a new technique came online for building crystals layer by layer: metalorganic vapor phase epitaxy (MOVPE). Amano and Akasaki set themselves the goal of using MOVPE to make crystals of p- and n-doped GaN. In 1986, after a decade of effort, they had found a successful recipe: Deposit GaN with its dopants on top of a layer of aluminum nitride that is itself deposited on a sapphire substrate. The sapphire–AlN foundation guides the formation of a crystalline GaN layer. Working independently, Nakamura hit on a similar recipe in 1991.
Doping GaN with magnesium or zinc yielded p-doped crystals, but not ones that could accept electrons efficiently. Fortuitously, Amano and Akasaki found in the late 1980s that samples they had examined with an electron microscope became better acceptors. The cause, Nakamura discovered, arose during crystal growth: Dopants formed efficiency-sapping complexes with hydrogen atoms, whose presence as a contaminant arises from the use of organic precursors in MOVPE. Irradiating the crystals with electrons breaks up the complexes. Annealing has the same beneficial effect.
The final step toward making efficient blue LEDs was to exploit the concept of heterostructures. In GaN LEDs, as in GaAs LEDs before them, different semiconductors from the same groups of the periodic table are combined in layers. Family membership ensures that the layers, which have different bandgaps and refractive indices, are structurally compatible with each other. With a judicious choice of layers, the electrons and holes that combine to emit photons can be squeezed into a narrower volume, thereby boosting efficiency. Further gains in efficiency come from exploiting the layers' optical properties.
For their first blue LEDs, Amano and Akasaki layered GaN with aluminum gallium nitride; Nakamura paired GaN with indium gallium nitride and InGaN with AlGaN. By 1993, Nakamura had made a tiny blue LED that shone as brightly as a candle. Light emission in the device took place in a layer of zinc-doped InGaN sandwiched between n- and p-doped AlGaN, which, in turn, was sandwiched between n- and p-doped GaN. To date, the paper describing the landmark device has been cited more than 3000 times.
Besides the potential for slashing the world's electricity bill, GaN-based LEDs have other important and widespread applications. The devices deliver light to the screens of cell phones, computers, and TVs. In poor countries, solar-powered LED lights are supplanting lamps fueled by kerosene.
