The combination of blue light and yellow light looks white to humans.
But where do the phosphors come from?
One answer is, from a Californian company called Intematix, which is behind some developments that are improving efficacy, colour-rendering, and lifetime amongst LEDs.
“In the past, it was a blue chip plus a YAG [yttrium aluminium garnet] or silicate phosphor. These are fairly narrow spectrum phosphors, which make fairly cool light with a lot of blue going through,” Intematix marketing director Julian Carey told Electronics Weekly. “They are still popular for outdoor and automotive lights. We still sell tonnes of them.”
The light from these LEDs has a tall narrow blue peak in its spectrum, and a broader yellow peak that has little blue-green or red. As such, it is difficult to produce warmer whites, which have a high red content, and colour-rendering can be poor.
For interior lighting, there is increasing interest in LEDs with a more balanced spectrum, and with more red to reduce the colour temperature.
A useful, if flawed, measure of colour rendering capability is colour-rendering index (CRI) where sunlight scores 100 and cool white LEDs score, very roughly, 70-80.
One way to improve CRI, and get a warmer white, is to mix in a second phosphor, and even a third, to take more of the blue and turn it into red.
“With a three phosphor strategy, including one red and one 650nm deep red, we can get 98CRI and ‘R9’ 99CRI,” said Carey.
R9 is a saturated red colour – the colour of meat and some fruit – and is not used in calculating normal CRI.
If lighting groceries, or Ruby Woo lipstick, is important, you need to think about the rendering of this red. R9 99CRI means this particular LED scores 99 out of 100 doing that.
Standard red phosphors are ‘red nitrides’, producing oranges and reds from 620 to 670nm.
Intematix has a range of red nitrides and, to fill out the middle of the spectrum, it has ‘green aluminates’ (515-560) branded ‘GALs’ for which it has patents.
“Where greens are needed, a broad spectrum green is green aluminate. If you add red to YAG phosphors, the spectrum is still too narrow. Green aluminate is much broader, you get more points in the CRI; so GAL plus red nitride for high CRI,” said Carey, whose firm also has patents for GALs and the latter combination.
Lutetium aluminum garnet, related to YAG and called LuAG (they are both ‘garnets’), is another broad spectrum yellow-green emitter.
“We make LuAG phosphors and others make them too,” said Carey. “LuAG is more green than YAG. GAL is broader that LuAG, generally gives higher CRI, and is more expensive.”
LED phosphors have a hard life – they are subject to a high flux of blue (high-energy) light, and they are frequently deposited on a die which may be running at over 100°C.
Do modern phosphors fade?
“They are quite indestructible. The lighting phosphors today: GAL, red nitride and YAG, have terrific lifetimes and high thermal stability independent of lifetime. You can go to 150°C with GAL and only lose 5% output,” said Carey. “‘Silicate’ phosphors, in mobile phone displays, will have lumen maintenance lower than GAL or nitride. They are for shorter life application, but they are blindingly bright.”
A key metric in phosphors is quantum efficiency – the ability to absorb one photon of the pump wavelength and emit another photon of the desired wavelength.
“We have phosphors as high as 93% QE. But this is only part of the story. Particles are as small as 5µm, and their shape and polish and colouration make a difference. That is where we have our experience,” said Carey. And “everything is great when a photon hits a particle head-on. Scatter losses and engineering the spectrum are two more angles that would be our contribution to the LED industry.”
Phosphors can deposited onto an LED die surface (‘local phosphor’), or made into a free-standing structure – like a dome – and positioned over a blue LED (‘remote phosphor’).
“Remote is our main phosphor technology. It is more expensive, but solution cost can be lower because efficiency is higher,” said Carey.
The drawback local phosphor is scattering loss.
“Phosphor is creating light in all directions, half the light going backwards. There is substantial loss on die surface – a lot of the die surfaces are absorbing,” said Carey.
Remote phosphors can keep scattering losses low because very little of the phosphor emission hits the die. Most reverse emission hits other parts of the remote phosphor. “Phosphors only respond to certain excitation wavelengths. A red photon hitting red nitride phosphor – basically, nothing happens,” said Carey – it is not absorbed.
And, as much as efficiency droop with temperature is low in modern phosphors, a few more percent are lost because local phosphors run hotter.
Remote phosphors can also be shaped to replace the cosmetic diffuser fitted in some LED light sources – incandescent bulb replacements, for example. Carey puts diffuser loss at greater than 10%.
“Efficiency benefit can be as high as 30% compared with coated LEDs behind a diffuser,” he said.
ChromaLit is Intematix range of remote phosphors, shaped to form the outside of luminaires. They come in various flat, domed and other shapes. There is a specific toroidal version for a light bulb-replacing reference design aimed at meeting Energy Star light distribution.
Why make red LEDs using blue die plus phosphor?
More than one LED company is avoiding traditional AlInGaP die when making red and amber LEDs, and is instead using blue InGaN die plus a phosphor layer to turn blue into red or amber.
“A red LED under ideal conditions is more efficient than a phosphor converted blue LED – it is immutable physics,” said Carey. “One issue is system stability. Red and blue LED die have different thermal curves, so you need control electronics to maintain colour balance. And you still need phosphor for green.”
What about getting the rare earths vital in phosphors emission, and come mainly from China?
“The base materials are plentiful materials. The dopants are where you find rare earths,” said Carey. The amount varies. It is usually very small, or can be more significant. I have not seen any effect of rare earth pricing in LEDs. The amount of rare earth as a fraction of the LED is very small.”
He points out that fluorescent lamps have multiple grams of phosphor in each tube, and that rising Chinese rare earth prices are affecting fluorescent tube prices.
Additional sources of rare earths are appearing. Mountain Pass mine in California was largest producer of them in the world before it was closed, producing material to go into electron beam-energised phosphors for CRTs. It is now open again.
“Mountain Pass is going to play a role,” said Carey.