Zeroing in on the elusive green LED

Apr 25, 2011
Researchers at Rensselaer Polytechnic Institute have developed a new method for manufacturing green LEDs with greatly enhanced light output. Led by Professor Christian Wetzel, the research team etched a nanoscale pattern at the interface between the LED’s sapphire base and the layer of gallium nitride (GaN) that gives the LED its green color. Overall, the new technique results in green LEDs with significant enhancements in light extraction, internal efficiency, and light output. Credit: Rensselaer/Robbins

Researchers at Rensselaer Polytechnic Institute have developed a new method for manufacturing green-colored LEDs with greatly enhanced light output.

The research team, led by Christian Wetzel, professor of physics and the Wellfleet Constellation Professor of Future Chips at Rensselaer, etched a nanoscale pattern at the interface between the LED's base and the layer of (GaN) that gives the LED its green color. Overall, the new technique results in green LEDs with significant enhancements in light extraction, internal efficiency, and light output.

The discovery brings Wetzel one step closer to his goal of developing a high-performance, low-cost green LED.

"Green LEDs are proving much more challenging to create than academia and industry ever imagined," Wetzel said. "Every computer monitor and television produces its picture by using red, blue, and green. We already have powerful, inexpensive red and blue LEDs. Once we develop a similar green LED, it should lead to a new generation of high-performance, energy-efficient display and illumination devices. This new research finding is an important step in the right direction."

Sapphire is among the least expensive and widely used substrate materials for manufacturing LEDs, so Wetzel's discovery could hold important implications for the rapidly growing, fast-changing LED industry. He said this new method should also be able to increase the light output of red and blue LEDs.

Results of the study, titled "Defect-reduced green GaInN/GaN on nanopatterned sapphire," were published last week in the journal , and are featured in today's issue of the Virtual Journal of Nanoscale Science & Technology, published by the American Institute of Physics and the American Physical Society.

The research program is supported by the U.S. Department of Energy National Energy Technology Laboratory (NETL) Solid-State Lighting Contract of Directed Research, and the National Science Foundation (NSF) Smart Lighting Engineering Research Center (ERC), which is led by Rensselaer.

Researchers at Rensselaer Polytechnic Institute have developed a new method for manufacturing green LEDs with greatly enhanced light output. Led by Professor Christian Wetzel, the research team etched a nanoscale pattern at the interface between the LED’s sapphire base and the layer of gallium nitride (GaN) that gives the LED its green color. Overall, the new technique results in green LEDs with significant enhancements in light extraction, internal efficiency, and light output. Credit: Rensselaer/Robbins

LED lighting only requires a fraction of the energy required by conventional light bulbs, and LEDs contain none of the toxic heavy metals used in the newer compact fluorescent light bulbs. In general, LEDs are very durable and long-lived.

First discovered in the 1920s, LEDs – light-emitting diodes – are semiconductors that convert electricity into light. When switched on, swarms of electrons pass through the semiconductor material and fall from an area with surplus electrons into an area with a shortage of electrons. As they fall, the electrons jump to a lower orbital and release small amounts of energy. This energy is realized as photons – the most basic unit of light. Unlike conventional light bulbs, LEDs produce almost no heat.

The color of light produced by LEDs depends on the type of semiconductor material it contains. The first LEDs were red, and not long thereafter researchers tweaked their formula and developed some that produced orange light. Years later came blue LEDs, which are frequently used today as blue light sources in mobile phones, CD players, laptop computers, and other electronic devices.

The holy grail of solid-state lighting, however, is a true white LED, Wetzel said. The white LEDs commonly used in novelty lighting applications, such as key chains, auto headlights, and grocery freezers, are actually blue LEDs coated with yellow phosphorus – which adds a step to the manufacturing process and also results in a faux-white illumination with a noticeable bluish tint.

The key to true white LEDs, Wetzel said, is all about green. High-performance red LEDs and blue LEDs exist. Pairing them with a comparable green LED should allow devices to produce every color visible to the human eye – including true white, Wetzel said. Today's computer monitor and television produces its picture by using red, blue, and green. This means developing a high-performance green could therefore likely lead to a new generation of high-performance, energy-efficient display devices.

The problem, however, is that green LEDs are much more difficult to create than anyone anticipated. Wetzel and his research team and investigating how to "close the green gap," and develop green LEDs that are as powerful as their red or blue counterparts.

Explore further: What is Nothing?

More information: In Their Own Words: Christian Wetzel on Green LEDs blogger.rpi.edu/approach/2009/… etzel-on-green-leds/

Defect-reduced green GaInN/GaN light-emitting diode on nanopatterned sapphire, Appl. Phys. Lett. 98, 151102 (2011); doi:10.1063/1.3579255

Abstract
Green GaInN/GaN quantum well light-emitting diode (LED) wafers were grown on nanopatterned c-plane sapphire substrate by metal-organic vapor phase epitaxy. Without roughening the chip surface, such LEDs show triple the light output over structures on planar sapphire. By quantitative analysis the enhancement was attributed to both, enhanced generation efficiency and extraction. The spectral interference and emission patterns reveal a 58% enhanced light extraction while photoluminescence reveals a doubling of the internal quantum efficiency. The latter was attributed to a 44% lower threading dislocation density as observed in transmission electron microscopy. The partial light output power measured from the sapphire side of the unencapsulated nanopatterned substrate LED die reaches 5.2 mW at 525 nm at 100 mA compared to 1.8 mW in the reference LED.

Related Stories

Researchers aim to close 'green gap' in LED technology

Aug 23, 2006

A team of researchers from Rensselaer Polytechnic Institute has received $1.8 million in federal funding to improve the energy efficiency of green light-emitting diodes (LEDs). As part of the U.S. Department ...

Osram LEDs for Mini Projectors

Mar 04, 2010

New, particularly high-performance light-emitting diodes (LEDs) from Osram make it possible to build mini projectors. These LEDs produce enough light to project images measuring over one meter in diagonal ...

Green LED is bright enough for large projector

Mar 17, 2011

Osram Opto Semiconductors has developed an extremely bright, green light-emitting diode (LED) that makes LED projectors in office environments possible. Projectors in conference rooms have to be very bright ...

Cheap, efficient white light LEDs new design

Apr 07, 2009

Roughly 20 percent of the electricity consumed worldwide is used to light homes, businesses, and other private and public spaces. Though this consumption represents a large drain on resources, it also presents ...

Recommended for you

What is Nothing?

Aug 22, 2014

Is there any place in the Universe where there's truly nothing? Consider the gaps between stars and galaxies? Or the gaps between atoms? What are the properties of nothing?

On the hunt for dark matter

Aug 22, 2014

New University of Adelaide Future Fellow Dr Martin White is starting a research project that has the potential to redirect the experiments of thousands of physicists around the world who are trying to identify the nature ...

Water window imaging opportunity

Aug 21, 2014

Ever heard of the water window? It consists of radiations in the 3.3 to 4.4 nanometre range, which are not absorbed by the water in biological tissues. New theoretical findings show that it is possible to ...

User comments : 18

Adjust slider to filter visible comments by rank

Display comments: newest first

Eikka
2.5 / 5 (8) Apr 25, 2011
LEDs produce almost no heat.


That is a blatant lie.

The maximum theoretical limit for efficiency for a LED is 38.1...43.9% in absolute terms, and the best LEDs by far are just approaching 15% of photons out versus energy in.

Heat is a massive problem for LEDs, causing them to dim rapidly in use, because they can't operate at high internal temperature without failing. Typically just 150 C on the inside of the chip means that it stops working immediately, which is why some high power LED bulbs even employ tiny fans inside the enclosure. (And which is why they won't work for 25,000 hours)

The only reason they appear to produce no heat is because the bulbs are designed to shed heat rapidly to stay cool on the inside. If the heatsink is hot to the touch, that means the diode junction inside is even hotter, which means the bulb will fail soon.
Eikka
1.8 / 5 (6) Apr 25, 2011
The other reason is that LED bulbs on the market come in just 4-7 Watts, so they don't have the power to become hot in the first place.

Trying to find and use a 20+ Watt LED bulb is alltogether more difficult, because they do have the power to become hot, yet they don't tolerate it at all.
Eikka
2.7 / 5 (10) Apr 25, 2011
Pairing them with a comparable green LED should allow devices to produce every color visible to the human eye including true white


Another lie.

Having mostly monochromatic red, green and blue light doesn't produce "true white".

What happens when you view a bright yellow object under that light? Same thing as what happens when you look at something red or green under the blue-yellow faux white. They look desaturated or greyish because those specific parts of the spectrum are missing.
deatopmg
1 / 5 (1) Apr 25, 2011
THis "report" has a lot of verbiage and little information. I would not have expected that from RPI.
Sonhouse
not rated yet Apr 25, 2011
The thing I see coming, if the green LED's get more power and they imply the same patterning will up the output of red and blue LED's, it looks like the same situation will happen because the red and blue LED's will still have a performance edge on the green. And like the guy above said, you would need probably 5 LED's, a color between red and green and another between green and blue to get more of a pure white. Those wavelengths may also prove to be as hard as green is to get high power and efficiency. BTW, there is work proceeding on high power LED's, maxed to the max and giving nanosecond pulse widths to take the place of spark plugs in engines, which will be another use for the LED, squeezing a few more percent out of petrol fuels.
PinkElephant
4.5 / 5 (10) Apr 25, 2011
@Eikka,

Your criticism is misplaced and disingenuous. They didn't say just "almost no heat", they said almost no heat WHEN COMPARED TO regular light-bulbs of same brightness.
Having mostly monochromatic red, green and blue light doesn't produce "true white".
Yes it does: ON A DISPLAY. You are confusing display applications with illumination applications.
Eikka
1.8 / 5 (6) Apr 26, 2011
@Eikka,

Your criticism is misplaced and disingenuous. They didn't say just "almost no heat", they said almost no heat WHEN COMPARED TO regular light-bulbs of same brightness.


Nope. I see no comparison of equal brighness. Claims that LEDs produce "almost no heat", while they do produce quite a lot of heat, are disingenuous. Try putting a high power LED bulb in a canister fixture and see it overheat itself. A 20 Watt LED bulb will produce about as much heat as a 20 Watt CFL if not more.


Yes it does: ON A DISPLAY. You are confusing display applications with illumination applications.


The article says:
The white LEDs commonly used in novelty lighting applications
(...)
The key to true white LEDs...


The article mixes and matches things. In the same paragraph they make claims about RGB displays that don't relate to what was just said about true white LEDs.

Full RGB on an LED display is already possible anyways.
Eikka
2.7 / 5 (7) Apr 26, 2011
And RGB displays do not produce "every possible color that the eye can see". They can simulate much of it, but not all.

The RGB primary wavelenghts are chosen because the eye is very sensitive to those wavelengths, but we can still see light that is bluer than the primary blue and redder than the primary red.

It's a compromize of efficiency.
Eikka
2.6 / 5 (5) Apr 26, 2011
Besides, there is no such thing as "true white". What is percieved as white depends on how much light there is, and how long you have been staying under it.

The only thing that approximates "true white" is something that produces a continuous spectrum of light, which LEDs don't.
bart_laws
2 / 5 (1) Apr 26, 2011
Err, excuse me, but if green LEDs are so hard to make, why does every GFI and smoke detector in my house have a green LED?
Bonkers
5 / 5 (2) Apr 26, 2011
I'm not sure about the white light thing. true white light is a pure flat rainbow of all wavelengths, but or eyes respond to blue green and red, therefore we can produce a white that's good enough, or even indistinguishable (to humans) using RGB monochromatic sources. Well. that's the accepted theory - and good enough for the point the article is trying to make. There are some second-order terms though - humans vary a little in their chromatic vision, some can see a difference between hues other's can't - particularly in the indigo/violet and in the deep red. Actually the wierd "glow" of venus (when not illuminated) - looks like the "earthshine" on the crescent moon, is only visible to some and is because it radiates in the deep red, just below 780nm - where we conventionally say that vision stops.
d_robison
5 / 5 (2) Apr 26, 2011
@Eikka

You have to look at the pros and cons of each type of light bulb to really get a good picture.

100W Incandescent Light Bulb.
Output: ~1700 lm or ~17 lm/W
Heat: ~335 degrees F
Lifespan: ~750 hours

13W CFL Light Bulb.
Output: ~800 lm or ~62.5 lm/W
Heat: ~180 degrees F
Lifespan: ~8,000 hours

For LED I chose the EarthLED Evolux S LED Light (13W).
Output: ~1000 lm or ~76.9 lm/W
Heat: ~87 degree F
Lifespan: ~50,000 hours

Just a few numbers for everyone, you can find out these numbers online or on the packaging, and what tests they did to obtain these numbers. In general LED's and CFL's are both very good choices over incandescent and both have advantages and disadvantages in certain applications.

You can also look up more information on websites like Scientific American, Popular Mechanics, etc. or you can choose a more technical website.
GSwift7
2 / 5 (3) Apr 26, 2011
I have a joke forming in my mind. Something about certain people who read articles through green colored lenses and dismissing the need for green LED's because ALL LED's are 'green'.

Err, excuse me, but if green LEDs are so hard to make, why does every GFI and smoke detector in my house have a green LED


Because those are relatively large and dim compared to the relatively tiny and bright LED's they are talking about here. Heck, they might even be blue LED's coated in yellow plastic? A true 1080P screen has just over 2 million pixels, so that's 6.2 million LED's if you use the above stated RGB method to make the picture in native 1080P mode. If each of the 2 million green LED's costs one thenth of a cent to make, that adds $2,000 to the cost to make the screen. Reducing the cost per green LED by one hundredth of a penny would save the manufacturer $200 per screen. That's significant.
that_guy
5 / 5 (1) Apr 26, 2011
So close Gswift. I wanted to give you a better score, but led tvs refers to backlighting, not pixels - all led tvs are LCD with LED backlighting, and oled tvs are a slightly different technology than traditional leds.

I know you all covered all the bad points that eikka made, but I'd like to reiterate them together.

1. Green LEDs are either energy inefficient (using the phosphor method like the "white" LED) compared to red and blue leds, and/or are expensive.
2. White light is a rainbow of spectrum, very well represented by the red, blue, and green pixels in a tv or computer screen. So while a true white led would be the holy grail, an efficient RGB setup would be nearly as good for most applications.
3. LEDs are tremendously more light efficient, and considerably more heat efficent than other forms of lighting. Even though you could get a significant fraction of heat as a lightbulb if you pumped all that wattage through them, it would still be less, and is unnecesary
that_guy
5 / 5 (1) Apr 26, 2011
4. The primary issues with LEDS are their SENSITIVITY to heat, as well as reduced effeciency at higher voltages (IE to get brighter light)

I'm sure that eikka ruined a bunch of other facts here, but I'm tired of scanning trying to pick them out.
GSwift7
2.5 / 5 (2) Apr 27, 2011
So close Gswift. I wanted to give you a better score, but led tvs refers to backlighting, not pixels - all led tvs are LCD with LED backlighting,


yeah, that's why I said the following:

so that's 6.2 million LED's IF YOU USE THE ABOVE STATED RGB METHOD (I don't know how to bold, so I used caps. I wasn't trying to shout :) )

Once LED's become advanced enough it will become possible to make a true LED display using RGB LED's. It would be a much better display with much better off-angle and bright light viewabilatiy, especially for displays on portable devices like cell phones. I'm aware of the LCD/LED backlighting, and I have a top LG model like that; she's my baby and she's beautiful. :)
d_robison
not rated yet Apr 27, 2011
@that_guy

Good sum of the advantages/disadvantages with LED's. I think everyone here definitely agrees that there needs to be more R&D for LED's before they can reach their full potential. It's a great technology, it just needs some work.
AmritSorli
1 / 5 (1) May 01, 2011
At the moment public opinion is: 73% is for time is numerical order of change in space and 27% time is a 4th dimension of space, vote on www.timelessuniverse.net