Leading LED chipmakers faced off and offered their best shots (that is, claims of class-leading devices) at the 7th International Conference on Nitride Semiconductors recently held in Las Vegas.

There are two sides to this story: the first is the "where are LED products today" side. The second is the "where do LED products want to go" side. Let’s start with the where we are today by running down some of the key products presented at the conference.

Art Berman
Insight Media Consultant

Nichia’s Yukio Narakawa kicked-off the day by announcing the development of a 1 mm x 1 mm white LED delivering 169 lm/W efficacy at 20 mA. This beats the 161 lm/W figure that Nichia reported last year and resulted from research to improve current spreading and reduce the forward voltage.

According to Narakawa, the latest LED also offers improved performance at high drive currents. It produces 134 lm/W at 350 mA and a forward voltage of 3.1 V and can deliver 361 lm at 97 lm/W when driven at 1 A.

Philips Lumileds’ best result at 350 mA still stands at 115 lm/W. This was the value announced earlier this year for its "droop-free" 1 mm x 1 mm chip, which features a double heterostructure design incorporating a thick InGaN region.

Frank Steranka, who was speaking on behalf of Lumileds, pointed out that by increasing the size of the chip, this LED can deliver higher efficacies. A 4 mm x 4 mm LED emitted 142 lm/W.

Cree’s John Edmond said that the chip maker’s best result was 133 lm/W at 350 mA. This chip is also bigger than 1 mm x 1 mm, although Edmond did not disclose its exact dimensions.

Meanwhile, SemiLEDs, the Taiwanese chip maker that incorporates a metallic mirror into its emitters, revealed that it had also reached the 100 lm/W efficacy milestone at 350 mA. The higher efficacy comes from adding a several-micron-wide step to the textured surface of the LED structure.

Seoul Semiconductor CEO Chung Lee championed the Korean company’s "Acriche" product line. Although these devices "only" operate at 60 lm/W, they run directly from an AC source, which Lee sees as a key advantage.

Now, in terms of where LED products want to go, a good road sign are the future targets proposed by the US Department of Energy. These targets are pretty stiff and include reaching 150 lm/W at 2 A by 2012.

To hit this particular target will require a hike in internal efficiency to 90%. In addition, phosphor technology needs major improvement. Currently up to one fifth of the emission can be lost in conversion.

Although the 2012 target is demanding, Steranka believes achieving it holds the key to LED adoption in solid-state lighting. That accomplishment, by all measures, is the ultimate goal and would be the big win for the LED industry.

At present, the cost-per-lumen for incandescent and fluorescent lamps are $0.03-0.05 and $0.06, respectively. To compete, the corresponding figure for LEDs needs to drop by a factor of about 20.

How can this be accomplished? Increasing the drive current to 2 Amps and the efficacy to 150 lm/W will deliver a nine-fold improvement. Since that is not enough, it will also require reducing the cost of producing these improved LEDs by half. Only then will LED chip manufacturers be in the right corral.

Although a classic "easier said than done", the financial and societal incentives to attain these targets are so compelling and LED technology is so vibrant, that our bet is that LED’s will make it. Check back in 2012 to see if IM was right.

Reply to the author