Engineers take aim at efficiency barrier in LED technology

20 February 2012 FREMONT, Calif. — In a brand-new factory here, Eric Kim, chief executive of Soraa Inc., cradles a palm-size light that he refers to as “LED 2.0.” The light has a circular snowflakelike cooling frame surrounding a lens…

20 February 2012

FREMONT, Calif. — In a brand-new factory here, Eric Kim, chief executive of Soraa Inc., cradles a palm-size light that he refers to as “LED 2.0.” The light has a circular snowflakelike cooling frame surrounding a lens that emits a bright white light.

But it also radiates a mystery — and a continuing controversy.

Over the past few years, energy-saving LED lights have popped up nearly every place where low power is required. They provide the backlighting for cellphones, smartphones and laptops as well as for headlamps for hikers, for instance.

But in the United States in particular, LED lights have not yet caught on for home lighting, still a bastion of the incandescent light bulb — which to this day is not much more efficient than when it was invented by Thomas Edison in 1879.

The problem is what’s called efficiency droop. LEDs function most efficiently at low currents. Turn the current up to levels needed for room lighting, and the efficiency falls off markedly. The lights don’t dim, but as you turn up the amount of electricity, you don’t get more light, so the efficiency goes down, a problem that has made it impossible for LED bulbs to be as cost-effective as incandescent or fluorescent home lighting.

“The efficiency droop is one of the most severe and most interesting problems and controversies in science and engineering,” said E. Fred Schubert, a professor of electrical engineering at Rensselaer Polytechnic Institute in Troy, N.Y.

“Considering that LEDs are the winning future in lighting,” he added, it’s important “for industry and society that the efficiency droop be understood and solved.”

Modern LEDs — or light-emitting diodes — are semiconductor chips that produce light when electrified. LEDs typically use less than 20 percent of the energy required by comparable incandescent lights and only a little more than half the energy required by compact fluorescents.

The diodes allow current to flow easily in one direction. Charge-carriers — electrons — pass through a junction within the chip, and when an electron meets a hole, it drops to a lower energy level, in the process emitting energy in the form of a photon — light.

Most of the time, that is. But not always. And explaining why a photon is emitted in some cases, but not in others, is a persistent theoretical debate among physicists and electrical engineers. The question is: What causes the photon not to be released and causes the droop?

That is exactly what physicists don’t agree on.

Announcements this month by Cree, one of the major makers of LEDs in the United States and by Soraa, a start-up here making next-generation LEDs, indicate that significant progress continues to be made in efficiency — and in avoiding droop.

The high brightness blue LED, the kind used today but too expensive for extensive use, was invented in 1993 by Shuji Nakamura, a Japanese physicist who is one of the founders of Soraa.

This month, the firm claimed a breakthrough in circumventing LED droop. However, it is keeping the details of its new technology proprietary.

Despite Soraa’s secrecy, last April a group of researchers at the University of California, Santa Barbara, where Dr. Nakamura took a faculty position after leaving Japan in 1999 and where two of his co-founders, Steve DenBaars and Jim Speck, also teach, published a research report indicating that they had found evidence that LED droop could be explained by a process known as Auger recombination. The Auger effect was first discovered in 1922 by two European physicists, and it describes a subatomic process in which an electron displaces another electron but does not emit light.

Still, the debate has not been put to rest. Late last year, two other groups put forward competing theories that suggest that Auger recombination plays little or no part in efficiency droop.

Other theories suggest a process called carrier leakage, in which at high currents the carriers start to spill away from areas of high performance, or another process, called carrier delocalization, where electrons simply fail to find a paired hole at high electric currents because they are pulled away from the active region where electrons and holes are supposed to recombine and emit light.

Soraa’s chief technology officer is Mike Krames, an electrical engineer, who while at a previous company, Lumileds, was one of the pioneers of the Auger recombination theory.

“For me the only model that is consistent with all of the experimental data is Auger recombination,” he said.

One scientist who is familiar with the company’s technology said that Soraa’s advantage comes from changing the orientation of the quantum well where the interactions that generate luminescence take place. Strong electrical fields that form along the plane of the crystal structure interfere with the process that typically generates photons. By tilting the structure out of that plane, it is possible to minimize the droop effect, he said.

From Dr. Krames’s perspective, the battle between the theorists may not be over, but the war has been won.

“The debate goes on in other people’s heads, but it doesn’t go on in mine,” he said. “It doesn’t change the fact that all the evidence points in this direction, which leads to a strategy which we’ve adopted, which leads to better devices.”