Blue LEDs given Nobel Prize in physics

[rhoenix]

Each year, roughly a quarter of the electricity we generate goes to lighting. For decades, that lighting came in the form of an incandescent light bulb, which produced 16 lumens for every Watt it was fed. Fluorescent bulbs are roughly five times as efficient, but recent LEDs do nearly 19 times better than incandescents, producing 300 lumens for each Watt.

The first LEDs date back to 1907, but it's only recently that their incredible efficiency has been brought to bear on the lighting market. One of the key holdups was our inability to generate a broad spectrum of colors. Specifically, we couldn't make white light because we lacked the ability to produce blue LEDs. Now, the Nobel Prize in Physics is being given to three materials scientists who overcame this roadblock.

The people receiving the honor are Isamu Akasaki and Hiroshi Amano, both faculty at Nagoya University in Japan, and Shuji Nakamura, now of UC Santa Barbara, who did much of his key work while at Nichia Chemicals, a small company in Japan.

LEDs rely on the combination of electrons and holes (areas that lack an electron and thus have a positive charge). When the two are combined, the energy difference between them is converted to a photon. This energy difference is set by the material the electrons and holes reside in—the bandgap between the energy of the conducting and stationary electrons in a semiconductor varies based on the semiconductor being used. The key to the efficiency of LEDs is the fact that, with the right materials, all of the energy difference gets converted into light (though there are still inefficiencies involved in generating the electrons and holes).

The first report of light being emitted by a semiconductor dates back to 1907, but it took a while for the theoretical understanding described in the last paragraph to be built. Once it was, it was relatively easy to generate LEDs that efficiently produced light at longer wavelengths. Red LEDs were developed in the late 1950s and swept into the consumer market with the first electronic calculators in the 1970s. But producing shorter wavelengths remained a significant hold up.

Gallium nitride (GaN) was recognized early on as a potentially useful LED material, since it has a large bandgap that can be tuned by mixing in different materials. But it also presented a number of distinct challenges. To begin with, it was extremely difficult to grow crystals of the material that are large enough to be of any use. Initial work on overcoming this problem began in the 1960s. While some progress was made, the material was still considered difficult to work with in the 1970s when Akasaki began studying it.

By the 1980s, Akasaki had moved from industry to academia and started working with his colleague at Nagoya University, Hiroshi Amano. Decades of trying different conditions allowed them to develop a technique that produced useful crystals. This involved starting a population of small crystals at 500°C, then gradually raising the temperature to 1,000°C, at which point the crystals would merge and grow. The intermediate temperature allowed the formation of a thin layer of semi-disordered material that acted as a foundation for growing ordered crystals above it.

On its own, GaN isn't a good host for either electrons or holes; to prepare it for use as an LED, specific impurities (called dopants) need to be added. This involved adding small amounts of either aluminum or zinc to different layers of GaN crystals. By chance, Akasaki and Amano found that putting their material through an electron microscopy—a standard way to look at a material's surface properties—boosted the amount of light it emitted.

While this was going on, Nakamura developed his own method of making high-quality crystals. When Akasaki and Amano published their electron microscopy results, Nakamura was able to explain them, showing that it involved removing hydrogen that had been shielding some of the aluminum. He quickly developed an easier method of achieving the same end, pioneering the use of indium as a dopant in GaN crystals. By 1994, GaN was being used to make the first blue LEDs, although these weren't impressively efficient. With the basic process in place, however, efficiencies soon followed, opening the door to blue LEDs.

This year's prize stands in sharp contrast to last year's award in physics, which went to the development of the theory that suggested the existence of the Higgs boson. For this year's recipients, the theory was understood decades before they began their work. Getting the theory to be useful, however, required decades of materials science and engineering, a lot of it trial and error in areas that lacked a rigorous theory to provide researchers with a clear direction for their trials. And it wasn't enough for the resulting materials to simply work; the processes that produced them had to be able to run at massive scales.

But as the Nobel Committee points out in its announcement, the results of all that hard work will be on display everywhere as the recipients make their way through Stockholm: "the light from their invention [will be] glowing in virtually all the windows of the city." And they can be assured that, as the century progresses, the ascendancy of the LED will only become more dramatic.