Different types of blue emitters used in White LEDs

[Anders Hoveland]

An LED emitter generally emits just one distinct color wavelength. The type of white color LED that has become ubiquitous involves using a blue wavelength LED emitter with phosphor, to convert some of the blue light into other wavelengths.

What I would like to discuss here is something we often do not consider. Phosphor-converted white LEDs may use different wavelengths of blue emitters, depending on the design.


Many white LEDs use a 465nm emitter.
I have some cheap Chinese "sky blue" color LEDs that seem to be using a 470-475nm emitter with some green phosphor.
The Cree LS (using TrueWhite technology) uses a 440nm blue emitter.
I have a Philips Luxeon lime color LED (highly efficient, 190 lumens when run at 1W) which uses a 430nm emitter with greenish-yellow phosphor.
(only a very small amount of 430nm gets through)

Of course, underdriving a blue LED emitter can shift the wavelength a little. When run at 30% of its rated power, the wavelength will typically increase by about +3nm.

These different emitter wavelengths have an effect on the quality of light emitted. The color shift resulting from even a 3 nanometer shift in blue wavelength can be noticeable to the human eye. That is why manufacturers are often concerned about color shift in dimmable white LED products.
(the human eye is much more sensitive to wavelength changes in this blue-green region of the spectrum than it is in other parts of the color spectrum)

Then there is color rendering. Shorter wavelengths of blue typically render blue colors more of a pure vibrant indigo blue, whereas much longer blue wavelengths can render blue colors more faded, or sometimes even a little greenish-blue. There is a clear difference in how jeans look under a white LED using a 465nm emitter compared to 470-475nm.

The third effect has to do with perceived harshness. I have never seen anything written about this effect, but my personal observations suggest that shorter blue wavelengths seem to be "harsher" on my eyes, more difficult to concentrate on. Using a multi-emitter LED array to generate white light for reading, it seems much easier on the eyes when using longer wavelengths for the blue, and I experimented with several different wavelengths.

This range of "blue" wavelength emitters can really span from azure blue to indigo, even violet. There are advantages and disadvantages to using different blue emitters.

 

[Harold_B]

I'm not sure I'm all too clear on the point(s) you are trying to make so in the spirit of furthering your goal of discussing the blue emitters I'll put some of what I have learned through experience in the thread. The impression I am left with from your post is that you assume the LED designer selects the die and specifies the output wavelength first. That is, the die and the wavelength are established and then the phosphor and encapsulant are selected to yield a specific performance. If that is the case then that is somewhat the opposite of the preferred method.

Typically the output spectrum of the LED including the CRI or most likely target x,y, and Y target values selected. Then a phosphor blend that is available is selected or a new blend is designed. This has everything to do with the die selection. Once the phosphor blend has been selected then the designer will know the excitation curve and the efficiency curve for the phosphor. The die is selected for maximum excitation of the phosphor without being too close to a drop-off "knee" where the output can change significantly, usually depending on temperature. If you are curious about the excitation curves, just look through the Intematix catalog.

The blue does contribute to CCT and CRI so it's not all about down conversion. But there you go, my two cents....

 

[SemiMan]

Generally 450-455nm is most common for white LEDs as it maximizes LED radiometric efficiency and phosphor conversion efficiency combined of common phosphors. Keep in mind that phosphor deposition is slightly adjusted on a wafer basis to maximize color point accuracy so there will be shifts in both blue wavelength and spectrum to maximize perception of the target color accuracy.

Harshness could be related to total blue cone response especially off axis as glare is related to blue and not only in the central 2 degrees. Could also be blue scatter and subtle focus effects which would be amplified by low light levels and a wide open pupil.

Philips lime is a special case to achieve high output of green for color mixing. It works well. Lots of work on quantum dot and other phosphor techniques in this area .... For red too.

Using 470nm may limit the color gamut ...ie range of color. Not a good thing normally. R8 cri may be okay but its like not having deep red on the other end of the spectrum.

 

[Xe54]

The Cree LS (using TrueWhite technology) uses a 440nm blue emitter.

Do you happen to know if the Cree LS fixtures use a stock Cree emitter?

 

[Anders Hoveland]

When it comes to some of the newer white LEDs, high CRI isn't all just about the phosphor.
Many designs are using blue emitters with slightly wider bandwidth emission. It's basically still a spike peaking at a discrete wavelength, but the spectral distribution trails off more and further towards the longer wavelengths, and this is evident from a careful look at the spectral graphs (even though much of the effect is hidden beneath the phosphor emission).

The most extreme example of this was the Spectrafill Blue LED design used by Electrospell (which apparently never actually got used in a commercial product).

Do you happen to know if the Cree LS fixtures use a stock Cree emitter?

No, but I would imagine it's very similar to the Oslon EQ-white.