Cheaper and better white LEDs on the horizon

[werepossum]

Here's some interesting news. A working group from Rutgers University delivered a presentation at a meeting of the American Chemical Society detailing a method of producing white LEDs which uses a combination of abundant Earth metals and organic luminescent molecules to produce white LEDS which not only cost up to 90% less than existing rare Earth phosphors, but are also finely customizable to eliminate band gaps and therefore can deliver better color. (For those not familiar with doping phosphors, they are usually doped to radiate in three fairly narrow spectra, with those three combining to make more or less white light.)
http://www.techbriefs.com/component/content/article/22908

To achieve the common, soft white light that consumers expect, current LED technologies typically use a single semiconductor chip to produce light, usually blue, and then rely on a yellow-emitting “phosphor” coating to shift the color to white. That’s because LEDs do not emit a white light. The phosphor is made from materials, such as cerium-doped yttrium aluminum garnet, that are composed of rare-earth elements. These elements are expensive and in limited supply, since they are primarily available only from mining operations outside the U.S. Additionally, the light output of these phosphors tends to be harsh, “cold” colors.

Li’s team is developing hybrid phosphor-based technologies that are much more sustainable, efficient and low-cost. They combine common, earth-abundant metals with organic luminescent molecules to produce phosphors that emit a controllable white light from LEDs. By varying the metal and organic components, the researchers can systematically tune the color of the phosphors to regions of the visible light spectrum that are most acceptable to the human eye. The team is continuing to experiment and develop other rare-earth-free LED phosphors based on different metals and organic compounds.

Many material combinations are possible, so they use a computational approach to initially sort through the possibilities and to predict what color of light the various metals and organics combinations will emit. They then test the best combinations experimentally. Their approach allows a systematic fine tuning of band gaps and optical emissions that cover the entire visible range, including yellow and white colors. As a result, their LEDs can be fine-tuned to create a warmer white light, similar to cheaper but inefficient incandescent lights. Their approach shows significant promise for use in general lighting applications.

This could also be used for fluorescent lamps, although given current life and efficiencies I doubt anyone will spend the time to come up with the correct compounds for UV radiation. But for LEDs, this should be huge. Our rare Earth elements are mostly from China and Russia, and although they aren't actually all that rare one has to mine a LOT of earth to get usable quantities of rare Earth elements. Even if it was any cheaper, the ability to not be dependent on China and Russia coupled with the ability to mine much more compactly would make this an important development. Here's hoping that it pans out and the methods they have developed scale well.

 

[glenn1]

Good news. Plus the low power requirements probably means it's a good application for consumer grade solar.

 

[shira]

Here's some interesting news. A working group from Rutgers University delivered a presentation at a meeting of the American Chemical Society detailing a method of producing white LEDs which uses a combination of abundant Earth metals and organic luminescent molecules to produce white LEDS which not only cost up to 90% less than existing rare Earth phosphors, but are also finely customizable to eliminate band gaps and therefore can deliver better color. (For those not familiar with doping phosphors, they are usually doped to radiate in three fairly narrow spectra, with those three combining to make more or less white light.)
http://www.techbriefs.com/component/content/article/22908


This could also be used for fluorescent lamps, although given current life and efficiencies I doubt anyone will spend the time to come up with the correct compounds for UV radiation. But for LEDs, this should be huge. Our rare Earth elements are mostly from China and Russia, and although they aren't actually all that rare one has to mine a LOT of earth to get usable quantities of rare Earth elements. Even if it was any cheaper, the ability to not be dependent on China and Russia coupled with the ability to mine much more compactly would make this an important development. Here's hoping that it pans out and the methods they have developed scale well.

I was under the impression that the main cost of LED lights was the expense of putting them together into a usable device. For example, most LED light bulbs consist of many LEDs, each separately soldered to the device framework. So I'm reading this article to be saying that the cost of the LEDs will eventually come down significantly, but it's not clear how much that will affect the overall price of LED devices.

 

[werepossum]

I was under the impression that the main cost of LED lights was the expense of putting them together into a usable device. For example, most LED light bulbs consist of many LEDs, each separately soldered to the device framework. So I'm reading this article to be saying that the cost of the LEDs will eventually come down significantly, but it's not clear how much that will affect the overall price of LED devices.

Depends on the type of fixture. I've questioned some manufacturer reps and there seems to be no consensus. On a good quality can light, the LEDs might be as much as a third of the cost. On a high end decorative cast light, that might drop to a tenth. On a replacement light bulb, I'm guessing maybe half. Basically for most fixtures you have the LEDs, the driver, the housing, and the heat sink (which more and more often is the housing.) Typically though the soldering is automatic or semi-automatic, so it's pretty cheap.

 

[zephyrprime]

The phosphor is not the primary cost so this will only have a small effect. Compact flourescent lights have way more phosphors in them and they're cheap compared to leds. I was under the impression that the primary cost of leds was the chip.

 

[werepossum]

The phosphor is not the primary cost so this will only have a small effect. Compact flourescent lights have way more phosphors in them and they're cheap compared to leds. I was under the impression that the primary cost of leds was the chip.

LEDs are pretty cheap doped or undoped. I think we'll see the most benefit in remote phosphor designs, which use a LOT more phosphor but have better efficiency. I know a couple of years ago, a typical 900 lumen downlight might cost $15 in doped LEDs, out of a manufacturing cost of perhaps $30. A typically remote phosphor disc for a downlight was around $10 bucks, whereas the Cree royal blue LEDs would run $12 depending on run. (It's basically the same LED, just without the white triphosphor doping.) At that disc price, there is a premium (call it $7 or around 20% for our very rough estimate) for remote phosphor designs, but they typically deliver better color, better color stability, and better efficiency - around 20% better is what I've seen quoted. Of course, the reflector for a remote phosphor design is also a bit more expensive, which this won't help, but if this development can make remote phosphor designs compete with doped-chip designs, that would be great.

The other big winner might be light bulbs. Instead of a very complex design around a very bright point source, imagine a bulb frosted with an inexpensive but efficient phosphor coating driven by a bright blue LED. Efficiency, low cost, and much less glare.

 

[sm625]

I bought like 40 of those LED bulbs that were on sale at home depot for $2.50 each a few months back. They are great frickin lights. I cant say they've paid for themselves yet, because they mostly replaced CFLs. But a few of them have replaced incandescants. They arent really going to get much cheaper than this for at least another 3 years.

 

[shira]

I bought like 40 of those LED bulbs that were on sale at home depot for $2.50 each a few months back. They are great frickin lights. I cant say they've paid for themselves yet, because they mostly replaced CFLs. But a few of them have replaced incandescants. They arent really going to get much cheaper than this for at least another 3 years.

I have 50 recessed lights in my house, all with BR30 (medium flood) LED bulbs. At the time, those bulbs were $28 apiece. They provide 50% more light from the same fixture as the CFL bulbs that came with the house. I definitely prefer the LED bulbs, though their CRI (color rendering) rating isn't that great. I wouldn't be surprised if in the next five years or so it will be possible to purchase CRI 90+ BR30 LED bulbs that output 150 lumens/watt for less than $10 apiece.

 

[RampantAndroid]

From what I remember (and I was dorking with LED lights back in college 10 years ago when my college was building a cabin that was capable of being 100% off the grid, and powering lighting off 12V no inverter needed was perfect) - the higher output LEDs, while small, are also MUCH harder to keep cool - and you need special PCBs that will help wick heat off the LED and let you put a heatsink on the PCB itself.

I think the LEDs I was playing with were Phillips made - run with no cooling, they'd burn themselves out in 15-20 seconds. The LEDs in the bulb that use multiple LEDs each put out a lot less light than the LEDs I played with.

 

[Murloc]

LEDs are still worth the price because they therotically last forever.

I have yet to see one break on me.

Anyway this is nice mainly for the chance of getting better color, although I found out that it's mostly a matter of habit, you get used to it.

 

[flexy]

You can't compare LEDs from 10 years ago to recent ones.

Even today when you buy LED bulbs for your living room, I only found ONE brand which can replace my halogens in terms of "light quality", eg. color temps. There are still many "cheaper" LEDs out there which are also labelled as emitting 2500K ("warm light") but which give off horrible light which IMHO is entirely unsuitable for inside a home. (Even with that brand which I am using now, "Osram", I found massive differences in "light quality" with the SAME type of light. You might buy 5, and only 3 or 4 of them have the right light while another one is just "too cold" for, say, a living room).

10 years ago, LEDs were total shiate..I remember buying one for testing back then, it was just a bulb with 30-40 "normal" white LEDs, not only was the color temp way off..it was also WAY too dim to even remotely replace halogen/incandescent.

The good thing is...the entire LED tech is progressing very fast, newer 5W or recently 10W single LEDs give off an insane amount of light even if cooling for those higher Watt LEDs still is sort-of an issue.

The OSRAMs I use for replacing the Halogen beamers, I would say they're pretty much 100% equal (unless you get one of the duds as mentioned)...and for the other lamps in the house I use Phillips, Toshiba and Samsung "bulb-style" LEDs, but their color rendering is 'slightly' different as compared to incandescent but still acceptable. I had no luck with "store brand" LEDs, all of those were really awful.