Colin Humphreys of Cambridge University has the 21st-century equivalent of the philosopher’s stone. He is working with something that sounds like an alchemist’s dream: a substance that could turn base metal into gold, or rather cash. It could transmit light without wasting energy as heat, amplify cellphone signals tenfold and make computers 10 000 times faster.
The substance in question is called gallium nitride. It exists nowhere in nature and has to be forged in state-of-the-art laboratories. But it could bring sunlight indoors, alleviate seasonal depression, assist cancer surgeons, scale CDs down to an inch in diameter, and help solve the energy crisis.
It may sound like the stuff of sorcery, but gallium nitride has gone from curious fabric to big business in just four years: 184 companies and 293 universities around the world have now invested in it, and experts predict a £2,7-billion market for it by 2007. Britain, through Humphreys, is among the pioneers.
Light-emitting diodes (LEDs) — little devices that convert electricity directly into light — already count out the hours, minutes and seconds in digital clocks and light up the jumbo-sized advertising screens in city centres. Other light sources such as tungsten-based light bulbs waste energy as heat. But LEDs based on gallium nitride work with maximum efficiency: that is, almost all of the energy emerges as light, not heat.
For once, cool is the right word. Work it out: ordinary light bulbs tend to burn out after six months. Gallium nitride-based lights will gleam at least 100 times longer than traditional light bulbs and consume only 10% of their energy. The material is big in Japan and California. In the United Kingdom, however, so far it is providing a set of experimental traffic lights in London.
Gallium is a silvery white metal that, like mercury and caesium, tends to melt into a liquid at around room temperature. It is a by-product of the process of refining aluminium from bauxite. Metallurgists long ago predicted that if forced into a compound with other elements it might have unusual properties. Gallium arsenide became one of those bits of metal magic that could turn electricity directly into light and is now the basis of new industries.
Gallium nitride, first exploited by a Japanese professor called Shuji Nakamura, and taken up by Humphreys shortly afterwards, may become even more important.
Unlike all previous LEDs, it shines in the blue, or high-frequency, spectrum. It is this match of properties — high efficiency, high frequency — that brings a gleam to the eye of researchers. Traffic lights are all very well, but the big prize is in the light from the bulb in the ceiling.
”The next big thing in terms of energy saving is white light. To get white light you take one of these blue LEDs and coat it with a phosphor that emits in yellow, and this combination of blue and yellow gives white light,” says Humphreys.
”These are very efficient, already more efficient than a light bulb, but the total light they emit isn’t as great as a light bulb. So to replace home and office lighting, at the moment, you’d have to cover a ceiling. But if we can do more research, we could increase the total light output.”
In about 10 years, probably five, the amount of light emitted from white LEDs will be greater than light bulbs, Humphreys predicts. Swap these super-bright LEDs for light bulbs and the resulting saving in energy means we could close down five power stations in England.
This would also herald a revolution in interior design. Researchers could make the LEDs look like light bulbs, or have them flush with the ceiling or distributed on walls. In the daylight they would be more or less invisible. ”But the main thing is the energy saving, your bills should go down and CO2 emissions should fall by 10%.”
The next revolution would be in data storage. Normal CDs are written with red lasers. Blue light has half the wavelength of red light, at the very least you could get four times as much information on a CD. In practice you could get 10 times as much, Humphreys says.
”So your CD doesn’t have to be the size it is to hold a lot of music. Philips is working on a CD that is an inch in diameter and will probably just go in somebody’s earpiece, so you won’t need a Walkman, just a one-inch CD in your headphones. And that will happen, with blue lasers based on gallium nitride you can make right now.”
The information revolution could be extended. What Humphreys and others call ”quantum dots”, made with gallium nitride, have remarkable properties that could open the way for quantum computers that could operate 10 000 times faster than any supercomputers now in the world. But quantum computing so far remains only a dream, enticing in theory, still seemingly improbable in practice.
Meanwhile, back in the real world, computer scientists have begun to reach the limits of silicon: how much they can pack on a chip, how robustly it performs. So, years ago, they began looking at gallium arsenide as the next big thing.
The reality is that just as gallium arsenide is a much better amplifier than silicon, gallium nitride is 10 times more powerful an amplifier than gallium arsenide. Its use in cell- phones could mean that cellphone base stations could be situated 10 times further apart. It could also revive the dream of a truly portable satellite phone, with a signal powerful enough to go directly from mouthpiece to satellite and then down again to anywhere in the world, handling huge volumes of data at astonishing speeds.
There are other glints in Humphreys eye. One is of using gallium nitride-powered microwaves to replace spark plugs in car engines, increasing combustion efficiency by at least 20%. Another is of sunlight at night. Make an ultraviolet LED, coat it with a phosphor that emits in the red, blue and green, tweak the system a bit and you would have indoor light that mimicked sunlight, but without the outdoor risk of sunburn. It could help people with seasonally affective disorder.
Humphreys even sees a role for gallium nitride in cancer surgery: surgeons aim to take out tumour tissue without also taking healthy tissue, which means sending samples to a laboratory during the operation, and adding at least half an hour to the ordeal under anaesthetic. But suppose the patient was doped with a drug taken up only by cancerous cells, a drug that would fluoresce when a gallium nitride light was shone upon it? Then the surgeon would know exactly where his knife should stop.
Right now, gallium nitride is a $1,35-billion market. Humphreys is working with one small firm in Cumbria and another in Cambridge. ”We are exploiting this to some extent. What would be really nice,” he says, ”would be more major electronics companies in this country exploiting it even more.” — Â